parabiosis experiments

Type search request and press enter

Young Blood and the Search for Biological Immortality

How scientists are devising remedies for the ills of human aging..

Reading time min

By Deni Ellis Béchard

E ternal youth is among humanity’s oldest quests. For centuries, from the British Isles to Japan, people searched for fountains and elixirs. In ancient China, wealthy individuals consumed potions made from gemstones to metabolize the endurance of jade or cinnabar, or from gold to keep from “tarnishing.” But in Europe, a touted panacea was human blood, bought fresh from executioners during Medieval and Renaissance times, or—much earlier, in ancient Greece—harvested from virgins and gladiators. As has so often been the case, the Hellenic people were on to something. 

The first murmurings that young blood might mitigate the effects of aging echoed through the scientific community in 2005. A team of Stanford researchers had paired old mice with young mice, linking their circulatory systems, and within five weeks, the muscle and liver tissues of the old mice began to resemble those of the young mice. The common belief that aging results from simple wear and tear suddenly seemed questionable. Rather, the body’s cells appeared to receive signals from the young blood telling them to regenerate. 

The gold standard of science is reproducibility, and over the next 15 years, trials repeatedly showed health improvements in old mice dosed with young blood. Few scientists questioned its restorative effects, but many asked which of its components were responsible and who would be the first to distill those and bring them to market. 

“Just a couple decades ago, if you postulated that you could slow down or reverse aging, you were really out there,” says Stanford professor of neurology and neurological sciences Tony Wyss-Coray, “but now, just in the past 10 years, there are so many interventions—genetic interventions, diets, environmental interventions—that show that you can slow down aging or even reverse aging, measured with many different tools.”

In recent years, life-extension therapies have hit the mainstream. Unlike their enthusiasts who promise the imminence of biological immortality, Wyss-Coray stands out for the rigor with which he has researched how young blood affects the brain. Chosen by Time in 2018 as one of 50 people transforming health care, the 55-year-old Swiss-born American is also a skeptic. “I am not one of these researchers who would take a drug they work on because they believe in it so much,” he says. He points out that trial after clinical trial with humans—many of which were based on promising mouse experiments—has failed. Yet through his work at Stanford and Alkahest, the biotech company he co-founded, he has brought us closer to understanding how young blood can rejuvenate the brain and which of its components do so. (Hint: They’re not specific to virgins or gladiators, or even mice.)

Making Cells Young

Parabiosis (“living beside”), a state in which two organisms share physiological systems, as in conjoined twins, was surgically induced for the first time by French zoologist Paul Bert in 1864. As a proof of concept, he linked the circulatory systems of two rats. The technique won him some attention but largely fell out of use until the 1950s, when researchers joined rats to evaluate whether the metabolic needs of one affected the other. (They did.) During that same period, Clive McCay, a gerontologist at Cornell, devised heterochronic (“different in time”) parabiosis, in which old rats were conjoined with young rats, which resulted in minor increases in life span for the old rats. Half a century later, the technique was revived, this time at Stanford under the leadership of Thomas Rando, professor of neurology, and Irving Weissman, MD ’65, professor of developmental biology. “We had been studying the decline in tissue regeneration with age,” Rando recalls, “and we had been trying to understand the mechanisms by which specifically the stem cells in old animals failed to engage in tissue repair like stem cells from young animals.” 

The results were startling: Old mice soon exhibited significantly enhanced repair of muscle and liver tissue. Something in young blood was activating stem cells, the cells responsible for tissue repair. The aged cells themselves began rejuvenating—a word that Rando uses cautiously, since measuring a cell’s biological age is difficult. “But certainly what we seem to be doing,” he says, “is conferring youthful properties to old cells.”

The regeneration appeared to involve epigenetics. “What makes a liver cell different from a skin cell,” Rando says, “is not the DNA—the DNA is the same—it’s the epigenome reading that DNA in a different way. We have evidence that with age you get the same kind of thing happening, where the DNA is essentially the same but the readout of that DNA is different, and it’s possible to reprogram that so that the old cell becomes younger by the way it reads its DNA.” 

The Plasma Swap

Wyss-Coray joined Stanford’s faculty in 2002, recruited by Rando, who was spearheading a research program on aging and age-related diseases. Wyss-Coray had completed his PhD in immunology at the University of Bern and a postdoc at Scripps Research Institute and the Gladstone Institute and was developing ways to diagnose Alzheimer’s—a disease to which he had a personal connection through his father-in-law. “I saw how he disappeared,” Wyss-Coray recalls, “and how in the end there was just a shell.” 

Wyss-Coray had conducted his research with mice engineered to express Alzheimer’s symptoms, but the lack of progress in diagnosing and treating the disease in humans frustrated him. At Stanford, he focused on how to detect the disease in people. “You can’t study the brain at the molecular level unless a person has died,” he says. “But you can probe the blood. The idea is that if something happens in the brain, it will leave molecular signatures in the blood.” 

‘Just a couple decades ago, if you postulated that you could slow down or reverse aging, you were really out there.’

When he analyzed human blood for clear markers specific to the disease, he found them as well as biomarkers of old age—which appeared to grow stronger as a patient’s Alzheimer’s advanced. “The aging connection,” he says, “came from following the trail to understand Alzheimer’s and realizing that the strongest signature we kept seeing was an aging signature.” After publishing these results in 2007, he increasingly found himself being invited to conferences on aging, a field in which he’d previously held no ambitions.

Rando and Wyss-Coray then combined their efforts to investigate how heterochronic parabiosis affects the brain. In a 2011 paper in Nature , they showed that young mice infused with the blood of old mice had impaired learning and memory. In 2014, in Nature Medicine , Wyss-Coray’s lab went further, publishing a paper that stated: “Exposure of an aged animal to young blood can counteract and reverse preexisting effects of brain aging.” 

During this time, Wyss-Coray’s lab showed that parabiosis wasn’t necessary to conduct these experiments. Plasma—the liquid part of the blood—could simply be injected. It didn’t even have to come from the same species. “If we take plasma from old people and put it into young mice,” he says, “we make the brains of mice more inflamed, we reduce stem cell activity, and we impair cognitive function. If we take plasma from young people and put it in an old mouse, that old mouse has more stem cell activity, has less inflammation and their memory function is better.”

While Rando and Wyss-Coray were conducting their research, Amy Wagers, a postdoc with the 2005 Rando-Weissman research team who is now professor of stem cell and regenerative biology at Harvard, was doing her own investigation into heterochronic parabiosis. In 2013 and 2014, her lab published research showing that parabiosis promoted muscle regeneration in older mice and made their enlarged and inefficient hearts resemble those of young mice. 

Speaking of the chronic diseases prevalent in older individuals, Wagers says, “A hypothesis that’s being tested not just by my lab but by many around the world is that the common denominator is aging, that there are fundamental mechanisms of aging that are seeding these diseases. Often there’s a confounding of life-span extension and strategies targeted at improving the health of older individuals. It’s entirely possible those two things will be connected, but it’s also possible that they are not. You could have an impact on health without changing life expectancy. And that, I think, would also be a win.”

Now, Wait a Minute

Around this time, the idea that young blood had rejuvenating qualities started generating excitement in Silicon Valley, where start-ups began charging tens or hundreds of thousands of dollars for plasma transfusions. Cashing in on the buzz, HBO’s Silicon Valley featured a scene in which Gavin Belson—chief innovation officer and tech supervillain—watches a PowerPoint presentation while receiving blood from his “transfusion associate.” Even Joe & the Juice, a global chain of juice bars with several locations in Palo Alto, got in on the fun, including “Young Blood” on its menu (it contains celery, lemon and apple). 

The FDA, responding to companies bringing young blood to market before either the mechanism underlying its short-term effects or its overall long-term impact was understood, issued a warning in February 2019, stating that “some patients are being preyed upon by unscrupulous actors touting treatments of plasma from young donors as cures and remedies. Such treatments have no proven clinical benefits for the uses for which these clinics are advertising them and are potentially harmful.” 

Wyss-Coray sees premature commercialization of young blood overshadowing the research aims of Alkahest, which he co-founded in 2014 and on whose board he and Rando serve. “We are very different,” he says. “We use clinical trials to demonstrate whether this really works. In a clinical trial, you cannot charge the subject.”

The challenge now was legitimizing what had previously looked to be one of the most promising breakthroughs in the field of aging.

The Philosopher’s Sponge?

The story of Alkahest begins with Chen Din Hwa, a philanthropist in Hong Kong who owned the Nan Fung Group—one of the city’s largest privately held property developers. In 2009, at the age of 86, Din Hwa learned he had Alzheimer’s and also began receiving blood transfusions for cancer. His grandson, Vincent Cheung, who holds a bachelor’s in molecular and cell biology from UC Berkeley, noticed that after each transfusion, his grandfather’s lucidity temporarily increased. When he shared this observation with Karoly Nikolich, a family friend who was an adjunct professor of psychiatry and behavioral sciences at Stanford and had a long history of involvement in biotech, Nikolich told him about Wyss-Coray’s work. Not long afterward, the Nan Fung Group expressed interest in seeding a company. 

Din Hwa died in 2012, but two years later, Alkahest was founded, with Nikolich as CEO and Wyss-Coray as chair of the scientific advisory board. The company’s name came from Wyss-Coray’s readings on Paracelsus, the 16th-century Swiss alchemist who claimed to have invented alkahest—a universal solvent that could dissolve any substance to its individual parts. It was supposed to be the philosopher’s stone—a centuries-old notion of a material that could transmute base metals to gold and restore youth. 

“Here we have plasma,” Wyss-Coray says, “which is this complex soup, and if we can figure out the rules and the individual components, we understand life, if you will.” But invoking Paracelsus conveyed a subtler message. “Paracelsus is credited as the founder of pharmacology because he discovered, or claimed to discover, that the dose makes the toxin. It depends on how much you take of anything whether it kills you or it has a beneficial effect.” 

‘[T]he dose makes the toxin. It depends on how much you take of anything whether it kills you or it has a beneficial effect.’

To avoid incorrect dosing, Alkahest used an albumin-rich plasma fraction that contained many proteins commonly found in young blood and that had already received FDA approval for transfusions, which facilitated the rapid start of clinical trials. (Due to the demand for millions of transfusions each year, donated blood is separated into the fractions that recipients need—only the red or the white blood cells, the plasma, or specific plasma proteins.) 

While Alkahest does not disclose the reasoning behind the proteins it chooses for treatment, albumin has been of special interest due to its ability to stabilize other factors, preventing them from degrading. “Albumin is a sponge protein that’s the most abundant protein in the blood,” Wyss-Coray explains. “It binds a lot of different factors and acts as an antioxidant to some extent, and with age, it changes its function and it’s not as effective.”

In August 2019, Alkahest completed a six-month phase 2 trial. (Only after phase 3—a much longer trial with a larger cohort and a control group—does the FDA give approval.) Forty patients with mild to moderate Alzheimer’s who had been treated with the plasma fraction over a period of six months showed no significant decline in cognitive function. “If you look at historic controls of people who have Alzheimer’s disease,” Wyss-Coray says, “they go down very gradually on average, so there’s a noticeable effect from the normal decline to no decline at all.” 

Wyss-Coray also embarked on another study that would help legitimize the use of blood plasma in treating cognitive decline. The vessels nourishing our gray matter are highly impermeable compared with those elsewhere, and historically, this blood-brain barrier was thought to keep out many elements in the blood, allowing only water and essential nutrients to pass through—not the types of blood factors that might cause regeneration. “It’s still an enigma what this blood-brain barrier is,” Wyss-Coray says. “We’re basically postulating all you need to do to make the brain function better and have less inflammation and more stem cell activity is change the composition of the blood. And the first reaction from neuroscientists is, ‘Are you crazy? These factors can’t get into the brain.’ ”

Using molecular-labeling techniques, Wyss-Coray’s lab tagged thousands of protein species in plasma and then injected it into mice. When the researchers looked at the mice’s brain tissue under the microscope, they saw tagged proteins inside the blood-brain barrier and in the neurons themselves. There was no longer any question that the blood was sending molecular signals to the brain. 

Just a Matter of Time

Understanding aging scientifically demands a system of biological measurement. Chronological age is an imprecise gauge of health, as a quick comparison of vigor among octogenarians, much less among people in their 30s or 40s, would reveal. Precisely assessing the protein composition of plasma at different ages, however, might provide a yardstick for life span. 

“When I started 15 years ago, I was looking at about 100 proteins in the blood, and now I can look at almost 3,000 different proteins,” Wyss-Coray says, emphasizing how advances in technology have benefited his research. Recently, in a study of 4,263 people aged 18 to 95, his team measured how the blood’s protein content changes with age, then created an aging clock. The team discovered that people’s biological age is generally within three years of their chronological age, with the exception of those who are unusually healthy or unhealthy. 

The largest change in blood proteins takes place around 78 years old, when the concentrations of approximately 1,000 proteins decrease or increase. Smaller waves occur around 60 and even around 34. “That’s super interesting to us,” Wyss-Coray says, “because it suggests that the aging process is uneven, that the proteins that change then are not the same ones that change later. So you can start to ask what the biological processes are that change at an early age in people. How do they affect aging 20, 30, 50 years down the line? And if you want to have an impact on aging, would you actually have to intervene much earlier?”

Remember Me

Wyss-Coray and his wife, Christina, the clinical coordinator at Stanford’s Alzheimer’s Disease Research Center, have two grown daughters—an urban planner and a PhD student in biology at Stanford—and a 16-year-old son. Wyss-Coray observes the rapidity with which his youngest child learns to use a new computer or phone. “I get frustrated when I see how quickly he picks up this stuff,” he says and laughs. “I start to feel aging, and it gets annoying, especially cognitive aging—that your brain is not as fast.” 

He acknowledges the ticking clock but remains skeptical of solving aging for his generation. “I’m too realistic,” he says. “We still have a limited understanding of biology in general. For a lot of these proteins, we know very little about them.” 

By mapping proteins associated with aging, he has come closer to identifying which of young blood’s ingredients have regenerative potential. Still, the reasons that proteins are created at certain ages and how they affect the rest of the body remain poorly understood. He points out that while the life span of worms has been dramatically extended in labs, the same techniques haven’t worked in humans or even mice, and he reiterates that mouse studies of the brain rarely produce results in humans. “Many clinical trials start with good, solid preclinical data—otherwise, you wouldn’t put $100 million into a phase 3 trial. And yet they fail, one after another.”

Aside from the feasibility of life extension, Wyss-Coray also weighs the ethical considerations. Given the global demand for plasma and the already limited supply for people critically in need, synthetic plasma proteins would have to be made and might be available only for the wealthy. “There are huge socioeconomic implications,” he says. “If we all of a sudden find something that prolongs life span to 120 in the average population, I don’t think we could deal with that. There aren’t enough resources, and the population would increase so rapidly that we could probably not cope with it without starting to kill each other or having massive famines.”

In light of these concerns, Wyss-Coray’s focus is on treating diseases and allowing people to have healthy lives (what both Rando and Wagers refer to as increasing “health span” rather than life span). As Alkahest raises funds for a phase 3 study on mild to moderate Alzheimer’s, it is running other trials investigating the impact of young plasma on Parkinson’s, severe Alzheimer’s and recovery after surgery. It is still years away from definitively knowing whether young blood can treat age-related cognitive diseases.

As for Wyss-Coray’s own cognitive decline, would he eventually consider using plasma to prevent it if the trials are successful? He hesitates and then says, “If the phase 3 data shows a positive effect? Yeah.”

Deni Ellis Béchard is a senior writer at S tanford . Email him at  [email protected] .

You May Also Like

The uber ‘rich’.

Some rideshare drivers have an advantage.

Stanford, Still

Photos of a campus made quiet by the novel coronavirus.

Diary of a Hospitalist

What it’s like to provide medical care during a pandemic.

Stanford Alumni Association

• Current Issue
• Past Issues
• Class Notes

Collections

• Recent Grads
• Mental Health
• After the Farm
• The Stanfords

Get in touch

• Letters to the Editor
• Submit an Obituary
• Accessibility
• Privacy Policy
• Terms of Use
• Code of Conduct

• Stanford Home
• Maps & Directions
• Search Stanford
• Emergency Info
• Non-Discrimination

© Stanford University. Stanford, California 94305.

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings
• My Bibliography
• Collections
• Citation manager

Save citation to file

Email citation, add to collections.

• Create a new collection
• Add to an existing collection

Add to My Bibliography

Your saved search, create a file for external citation management software, your rss feed.

• Search in PubMed
• Search in NLM Catalog
• Add to Search

Parabiosis modeling: protocol, application and perspectives

Affiliations.

• 1 Institute of Neurology, Sichuan Provincial People's Hospital, University of Electronic Science and Technology of China, Chengdu, Sichuan 610072, China.
• 2 Chinese Academy of Sciences Sichuan Translational Medicine Research Hospital, Chengdu, Sichuan 610072, China.
• 3 Department of Neurology, General Hospital of Western Theater Command, Chengdu, Sichuan 610083, China.
• 4 Department of Neurology and Center for Clinical Neuroscience, Daping Hospital, Third Military Medical University, Chongqing 400012, China.
• 5 Department of Neurology and Center for Clinical Neuroscience, Daping Hospital, Third Military Medical University, Chongqing 400012, China. E-mail: [email protected].
• 6 Department of Neurology, Sichuan Provincial People's Hospital, University of Electronic Science and Technology of China, Chengdu, Sichuan 610072, China. E-mail:[email protected].
• PMID: 33723928
• PMCID: PMC8175960
• DOI: 10.24272/j.issn.2095-8137.2020.368

Abstract in English, Chinese

Parabiosis is a surgical method of animal modeling with a long history. It has been widely used in medical research, particularly in the fields of aging, stem cells, neuroscience, and immunity in the past two decades. The protocols for parabiosis have been improved many times and are now widely accepted. However, researchers need to consider many details, from surgical operation to perioperative management, to reduce mortality and maintain the parabiosis union. Although parabiosis has certain inevitable limitations, it still has broad application prospects as an irreplaceable animal model in the medical research field.

并联共生（parabiosis）是通过手术方法将两只实验动物进行外科连接的一种历史悠久的动物模型。在过去的二十年里，该模型被广泛应用于生物医学研究，特别是在衰老、干细胞、神经科学和免疫等领域。虽然并联共生模型已经得到了不断改进和完善，但从手术操作到围术期管理，仍有诸多细节值得研究者们关注，以降低死亡率和提高并联共生的稳定性。尽管存在着自身的局限性，但作为一种不可替代的研究手段，并联共生动物模型在生物医学研究领域仍有广阔的应用前景。.

Keywords: Application; Parabiosis; Perioperative management; Protocols; Surgical technique.

PubMed Disclaimer

Surgical preparation of parabiosis mouse…

Surgical preparation of parabiosis mouse model

Formation of surgical incisions and…

Formation of surgical incisions and management of scapulae and femurs (left mouse taken…

Suturing scapulae and securing femurs…

Suturing scapulae and securing femurs together

Suturing dorsal skin incisions

Surgical treatment of ventral structures

Suturing ventral skin incisions

Re-checking all sutured incisions

Similar articles

• A revival of parabiosis in biomedical research. Eggel A, Wyss-Coray T. Eggel A, et al. Swiss Med Wkly. 2014 Feb 4;144:w13914. doi: 10.4414/smw.2014.13914. Swiss Med Wkly. 2014. PMID: 24496774 Free PMC article. Review.
• Heterochronic parabiosis: historical perspective and methodological considerations for studies of aging and longevity. Conboy MJ, Conboy IM, Rando TA. Conboy MJ, et al. Aging Cell. 2013 Jun;12(3):525-30. doi: 10.1111/acel.12065. Epub 2013 Apr 10. Aging Cell. 2013. PMID: 23489470 Free PMC article. Review.
• Research Techniques Made Simple: Parabiosis to Elucidate Humoral Factors in Skin Biology. Spencer CA, Leung TH. Spencer CA, et al. J Invest Dermatol. 2019 Jun;139(6):1208-1213.e1. doi: 10.1016/j.jid.2019.03.1134. J Invest Dermatol. 2019. PMID: 31126426 Free PMC article.
• Comments on the history of medical-biological studies of aging, the birth of scientific gerontology. Robert L, Labat-Robert J. Robert L, et al. Curr Res Transl Med. 2017 Jan-Mar;65(1):44-47. doi: 10.1016/j.retram.2016.08.004. Epub 2016 Nov 3. Curr Res Transl Med. 2017. PMID: 28340696 Review.
• A new medical research model: ethically and responsibly advancing health for humans and animals. Olson PN, Ganzert RR. Olson PN, et al. Annu Rev Anim Biosci. 2015;3:265-82. doi: 10.1146/annurev-animal-022114-110759. Epub 2014 Nov 3. Annu Rev Anim Biosci. 2015. PMID: 25387116 Review.
• High-mobility group box 1 fragment ameliorates chronic pancreatitis induced by caerulein in mice. Hokkoku D, Sasaki K, Kobayashi S, Shimbo T, Kitayama T, Yamazaki S, Yamamoto Y, Ouchi Y, Imamura H, Kado T, Toya K, Fujii W, Iwagami Y, Yamada D, Tomimaru Y, Noda T, Takahashi H, Tamai K, Doki Y, Eguchi H. Hokkoku D, et al. J Gastroenterol. 2024 Aug;59(8):744-757. doi: 10.1007/s00535-024-02112-z. Epub 2024 May 10. J Gastroenterol. 2024. PMID: 38727823
• Fate-Mapping Macrophages: From Ontogeny to Functions. Ahlback A, Gentek R. Ahlback A, et al. Methods Mol Biol. 2024;2713:11-43. doi: 10.1007/978-1-0716-3437-0_2. Methods Mol Biol. 2024. PMID: 37639113 Review.
• Remyelination and Ageing: Ethical Considerations of Using Surgically Joined Animals in Research. Horn CM, Pop N, Anderson M. Horn CM, et al. J Undergrad Neurosci Educ. 2023 May 19;21(2):R1-R4. doi: 10.59390/SBPM5289. eCollection 2023 Spring. J Undergrad Neurosci Educ. 2023. PMID: 37588652 Free PMC article.
• Heterochronic Plasma Transfer: Experimental Design, Considerations, and Technical Challenges. Ansere VA, Bubak MP, Miller BF, Freeman WM. Ansere VA, et al. Rejuvenation Res. 2023 Oct;26(5):171-179. doi: 10.1089/rej.2023.0035. Epub 2023 Sep 25. Rejuvenation Res. 2023. PMID: 37551981 Review.
• The effects of high plasma levels of Aβ 1-42 on mononuclear macrophage in mouse models of Alzheimer's disease. Li C, Liu K, Zhu J, Zhu F. Li C, et al. Immun Ageing. 2023 Jul 31;20(1):39. doi: 10.1186/s12979-023-00366-4. Immun Ageing. 2023. PMID: 37525137 Free PMC article.
• Ajami B, Bennett JL, Krieger C, Tetzlaff W, Rossi FMV Local self-renewal can sustain CNS microglia maintenance and function throughout adult life. Nature Neuroscience. 2007;10(12):1538–1543. doi: 10.1038/nn2014. - DOI - PubMed
• Ajami B, Bennett JL, Krieger C, McNagny KM, Rossi FMV Infiltrating monocytes trigger EAE progression, but do not contribute to the resident microglia pool. Nature Neuroscience. 2011;14(9):1142–1149. doi: 10.1038/nn.2887. - DOI - PubMed
• Bert P Experiences et considerations sur la greffe animale. Journal de l'Anatomie et de la Physiologie. 1864;1:69–87.
• Brack AS, Conboy MJ, Roy S, Lee M, Kuo CJ, Keller C, et al Increased Wnt signaling during aging alters muscle stem cell fate and increases fibrosis. Science. 2007;317(5839):807–810. doi: 10.1126/science.1144090. - DOI - PubMed
• Bu XL, Xiang Y, Jin WS, Wang J, Shen LL, Huang ZL, et al Blood-derived amyloid-β protein induces Alzheimer's disease pathologies. Molecular Psychiatry. 2018;23(9):1948–1956. doi: 10.1038/mp.2017.204. - DOI - PubMed

Publication types

• Search in MeSH

Related information

Grants and funding, linkout - more resources, full text sources.

• Editorial Office of Zoological Research
• Europe PubMed Central
• PubMed Central

Other Literature Sources

• scite Smart Citations

Miscellaneous

• NCI CPTAC Assay Portal
• Citation Manager

NCBI Literature Resources

MeSH PMC Bookshelf Disclaimer

The PubMed wordmark and PubMed logo are registered trademarks of the U.S. Department of Health and Human Services (HHS). Unauthorized use of these marks is strictly prohibited.

Has the fountain of youth been in our blood all along?

BY  KAT MCGOWAN

He’s a bit muddled, but that’s not his fault. Few new neurons are being born inside his itty-bitty brain. The cells that once exuberantly branched, sending lush offshoots to interweave and connect with others, are now sparse and barren.

This Lego test indirectly measures those physical changes by monitoring his behavior. When mice of a certain age become forgetful, they spend more time checking out little trinkets they’ve seen before—objects that should warrant only a quick “Oh yeah, that thing again” glance. Cameras and laser-based detectors mounted on the ceiling capture and quantify those pauses and vacillations.

Alana Horowitz, the University of California, San Francisco graduate student conducting this FaceTime lab tour, puts her phone camera right up to the mouse’s muzzle. His eyes are bleary, like an old barfly’s. He probably hasn’t groomed himself recently, she says. His coat looks shabby and worn. You’ve likely never looked an elderly mouse in the face, but if you did, all of this—the thinning fur, the dim eyes, the hesitation—would be depressingly familiar. He inspires pity. Like sands through the hourglass, little fella.

But in this lab, headed up by neurobiologist Saul Villeda, nobody is sighing and moping over graybeard mice. Here, aging is not a sad fate to bemoan; it’s a problem to be solved. And for mice, at least, this team has already figured out how to reverse the damage time brings.

The secret is somewhere within those tiny veins. In a series of studies over the last 15 years, Villeda and others in a few like-minded labs at places like Stanford and Harvard have shown that, when infused with blood from young mice, old ones heal faster, move quicker, think better, remember more. The experiments reverse almost every indicator of aging the teams have probed so far: It fixes signs of heart failure, improves bone healing, regrows pancreatic cells, and speeds spinal cord repair. “It sounds sensational, almost like pseudoscience,” says Villeda. It’s some of the most provocative aging research in decades.

These studies, which use a peculiar surgical method called parabiosis that turns mice into literal blood brothers, show that aging is not inevitable. It is not time’s arrow. It’s biology, and therefore something we could theoretically change. The attempt to turn back the clock in living bodies “is probably the most revolutionary experiment that biologists have done,” says Stanford professor of neurology Tony Wyss-Coray, who was Villeda’s graduate supervisor and still leads blood-based research on Alzheimer’s disease and cognitive decline. “It supports this notion that it is possible to reassemble and fix things that we thought are doomed to die.”

We know blood does indeed perform some kind of alchemy that can restore and remodel the flesh or hasten its decay.

Benjamin Button-ing, of course, isn’t natural. But Villeda counters that getting old isn’t either: “It is the most artificial construct.” Previously, only a very few rare individuals reached 90 or 100. Now, in wealthy nations, it’s becoming downright common. With antibiotics, vaccines, public health measures, and a steady food supply, the industrialized world made the long, slow goodbye of aging commonplace—and, along with it, the consequences, such as brittle bones, Alzheimer’s disease, diabetes, and heart failure. Young-blood research, like some gory fairy tale, whispers to us that there could one day be a magic pill that can fix it all. The plot twist: That bloody fountain of youth was inside our bodies all along.

Biologists like Villeda just haven’t yet figured out why all this trading works.

Blood itself will not become a treatment for old age. It’s too messy, too complicated, too dangerous. But because of these labs’ findings, we know that somewhere swirling around in young veins are signals that awaken the natural mechanisms to repair and restore the body. These mystery factors, once researchers can identify and fine-tune them, could become precious medicine.

Villeda’s group in particular is applying parabiosis to address the toughest task of all: fixing an elderly brain. His team is also testing whether other physiological benefits, like those brought on by fasting or exercise, could be spotted in blood and distilled into a remedy for the aged. “We know there’s a needle in this haystack,” he says. “We just have to figure it out.”

THE IDEA THAT BLOOD can impart vigor and vitality has a long and stomach-turning history. Pliny the Elder, writing in first-century Rome, describes people with epilepsy guzzling the gore of wounded gladiators. Similar motifs reappear frequently in European lore: The sickly 15th-century pope Innocent VIII allegedly traded blood with three shepherd boys; all four died shortly thereafter.

Once British physician William Harvey mapped the circulatory system in 1628, swapping fluids became a fad. Across France and England, enterprising proto-scientists linked animals to animals and animals to people, and on and bloody on. Their hypothesis was that blood could remodel the flesh. In 1666, for instance, the legendary natural philosopher Robert Boyle proposed that introducing blood from a cowardly dog into a fierce one might temper the savage beast’s nature.

In 1667, London’s Royal Society hosted a public experiment in which a surgeon paid a man suffering from mental illness to be linked to a living sheep for a few moments via feather quills and silver pipes. Perhaps the gentle lamb’s essence might ease his agitation, was the thinking. Afterward the fellow indeed “found himself very well,” at least according to the surgeon, and he allegedly went on to spend his fee in the tavern. (The sheep’s feelings were not recorded.)

Months later, a Frenchman died following a transfusion, taking the wind out of these blood-spattered sails. The pope himself (Innocent XI, this time) put an end to the practice in 1679.

A new round of transfusion science emerged in the early 19th century, this one with much more scientific rigor. These experiments helped establish the first real knowledge about how to keep injured soldiers from bleeding out or mothers from dying in labor. But it wasn’t until 1864 that a Parisian physician working on skin grafts developed true parabiosis: a sustained commingling of the blood supplies of two living creatures.

Knowing that the red stuff flows through every organ and tissue, scientists have used the technique ever since to study bodywide states like obesity and systemic diseases like radiation sickness. If you divert blood from a sickly animal into a healthy one, and that one also becomes ill, it suggests some soluble factor in the blood plays a role. That knowledge, in turn, helps you narrow down what causes the illness or condition. For example, in 1958, scientists linked up rats from a strain prone to tooth decay to rodents from another strain that’s naturally resistant to cavities, to test whether something in the blood might account for the differences. In this case, at least, blood swapping made no difference.

Heterochronic parabiosis, in which researchers pair two animals at different points in the lifespan, was first used to study aging in the 1950s. But by the 1990s, it was largely forgotten—until Stanford put it back on the map.

AGING AFFECTS EVERYTHING EVERYWHERE, all at once. The hair grows gray, the bones weaken, the heart falters. Inside cells, DNA replication glitches and stutters, and proteins clump up into sticky globs. Meanwhile, natural repair mechanisms like adult stem cells no longer scurry to replace dead or injured tissues. All this happens more or less in sync, as if some systemwide signal has told the whole body to go down the tubes.

This organized process of decrepitude was still largely an enigma in 1993, when biologist Cynthia Kenyon, then at UCSF, discovered that mutating just one gene in a roundworm doubled its lifespan. Her finding helped launch the modern study of aging, but it soon became clear that a one-gene or one-protein approach wasn’t going to work, at least not for mammals. “We started to realize that the human body is not a simple assembly of individual molecules, but an incredibly complex physiological machine,” says Stanford’s Wyss-Coray.

But what is it that coordinates this systemic ruin? Fellow Stanford neurologist Thomas Rando reasoned that it made sense to look in the blood, that witch’s brew of biochemical whatnot that bathes the body, pinkie toe to pointer finger. Mostly water, nutrients, and red blood cells, what runs through our veins also transports a huge variety of signaling molecules that coordinate metabolism, immune responses, fight-or-flight reactions, and myriad other activities.

On the theory that blood-borne factors might orchestrate the transitions of aging, Rando and two postdocs in his lab, the husband-and-wife team of Michael and Irina Conboy, turned to heterochronic parabiosis. In the creepy but simple procedure, the surgeon slits two anesthetized mice down their flanks, then sutures and staples them together, side by side. Because these lab animals are so inbred, their immune systems don’t attack one another. As the incisions heal, their blood vessels connect, and the two share a supply.

Conjoined, the Frankenmice learn to eat together, make their little nests together, and ramble around as if they’re in a three-legged race. Their bodies begin to change. The old mouse’s fur gets thicker and silkier. It scrapes together its bedding more quickly. The junior partner loses speed, becomes tentative.

The team’s 2005 findings, published in  Nature , caused a stir. It’s like this: If an older mouse’s leg gets frozen with a piece of dry ice, the cells in charge of muscle repair don’t respond much; the number of active cells increases by just 10 percent or so. But after heterochronic parabiosis, twice as many cells activate in response to injury—a reaction like that of a young animal. Older mouse livers demonstrate a similarly sprightly cellular turnover.

The authors had brain data too, but it was too preliminary to be included in the paper. By 2005, the long-held dogma that adult brains cannot make new cells had softened: Research had shown that certain regions, including the hippocampus, could generate new neurons, but claims of actually restoring function still raised most eyebrows sky-high.

Soon after the Rando paper’s publication, Villeda, then just 25, was returning to his graduate studies in Wyss-Coray’s lab, one floor away in the same building at Stanford. The son of Guatemalan immigrants, Villeda had been educated in public schools in Los Angeles with little exposure to science until college, when he walked into a lab and saw a mouse embryo growing in a dish. It blew his mind. He loved science, the challenge, the craziness of it, the fun of it. He was curious and intellectually fearless. That is to say, exactly the type of person to grasp this particular third rail.

“It was very high risk,” says Wyss-Coray, who frequently collaborates with Rando. “Most people would say, ‘What does blood have to do with the brain? This will absolutely not work.’”

For three years, Villeda did the tiny surgeries and collected evidence. Soon, he could see that new brain cells were in fact surging in old mice. And they looked great.

“When a neuron is born in an old brain, it’s [usually] scrunched up,” he says, balling up his fist. “In these old brains they looked just like the young ones, beautiful,” he continues, stretching out his fingers. Those cells eagerly extended their long tendrils to make connections—the synapses that enable learning, memory, thinking, and everything else an elderly mind often struggles with.

In 2011, Villeda published a paper, also in  Nature , showing that mature mice in parabiotic pairings sprout two to three times as many new neurons as usual. But the bigger splash came in  Nature Medicine  in 2014, where he demonstrated that the access to young blood not only remodeled old nerve cells so that they looked and responded like younger neurons but also improved aged mouse learning and memory. A group led by Harvard’s Amy Wagers published similar results in  Science  at the same time, bolstering both claims.

Wagers and others at places like Columbia Medical Center soon showed that parabiosis could improve the function of heart, bone, and other tissues. These teams worked together to establish a working definition of what really qualifies as rejuvenation, including changes in DNA modification, gene activation, or protein levels characteristic of younger bodies.

As Villeda drew blood, he also collected plasma—blood with the cells removed—from young mice, drop by teeny-tiny drop, and transfused it into older ones. The effect was the same, strongly suggesting that whatever the magic was, it was something dissolved in the fluid itself, some code or key that signaled a fresh start.

“The way I think about it is that there’s a lot of information in the blood,” he says. Now, at last, they could work on cracking that code—and hopefully doing for humans what they’d already done for mice.

JUST TO GET THIS OUT OF THE WAY: Nobody’s sewing humans together. Our immune systems would wallop one another, with potentially deadly consequences (the lovely technical term is  parabiotic disharmony ). Transfusing seniors with young blood isn’t practical either; people would probably need repeat treatments, with each bringing a risk of infection, allergic reaction, and even injury to the lungs (transfusions sometimes cause a poorly understood immune reaction that ravages their lining). Because the dosing would restart cell division, it might also spark cancerous growths. And we don’t even know whether it would produce the desired results in a human being—or what mechanism would be behind the transformation.

Nonetheless, those two 2014 papers inspired a lot of wild ambitions. Rando got calls from cosmetics companies developing elixirs for youthful skin and from men’s magazines seeking secrets for reinvigorated muscles. A billionaire invited Wyss-Coray to an Oscar party. (He didn’t go.) “You get offers of a lot of money and no oversight,” says Villeda; people who owned property in nations with lax regulatory supervision on human research made what he refers to as “indecent proposals.”

Longevity enthusiasts eagerly discussed the findings, even though there is little evidence that heterochronic parabiosis extends life; even in rodents, all we know for sure is that it undoes some late-in-life decay. Captains of the tech world also took note. Reputed interest from billionaire founder Peter Thiel inspired a spot-on subplot in the HBO comedy series  Silicon Valley  in which an aging mogul takes a meeting while getting pumped full of blood from a fresh-faced athlete.

Meanwhile, a cottage industry began selling young plasma. Around 2016, Ambrosia, a California company, offered to infuse customers as part of a clinical trial that charged participants $8,000 to join. (So far, the team has not published any findings in the scientific literature.) Other entities and individuals launched similar efforts, such as a proposed study that would charge large sums to frail elderly people for doses of young plasma.

This “therapeutic plasma exchange” is a legitimate treatment for certain rare autoimmune diseases and problems with coagulation, so these providers are not necessarily required to obtain explicit approval from the Food and Drug Administration so long as they make no unsubstantiated health claims about their regimen. But, of course, they did: Companies marketed benefits for people with memory loss, heart disease, and even Parkinson’s. The FDA, now stepping into the regulatory role of the 17th-century pope, released a stern memo in 2019 that curbed the trend.

Ultimately, these projects made no progress toward the real prize, which is to convert the knowledge gained into a convenient, powerful, and predictable form, such as a pill. “Everyone recognizes this is an incredibly important experiment,” says Eric Verdin, CEO of the Buck Institute for Research on Aging, who closely follows parabiosis. “What has been lagging is: How do you translate these discoveries?”

The most straightforward path would be to pinpoint a pro-aging factor in old blood, mouse and human, that a drug could block. Many groups have identified such elements. Villeda and his collaborators, for instance, found that a protein called CCL11 increases in aged humans and mice and is correlated with reduced brain cell birth.

The other obvious tactic is to identify youthful plasma’s secret formula and optimize it. The Conboys’ research suggests the hormone oxytocin might be a candidate; Wagers has identified the protein GDF11. Combination therapies are also under consideration; the biotech company Wyss-Coray founded is exploring mixtures of hundreds of blood-borne proteins as therapies for a variety of age-related diseases. Villeda is on its board.

It’s also possible that the rejuvenating effects seen in experiments don’t arise from one magic ingredient, or even from some combination of a dozen or a hundred compounds, but happen simply because the procedure dilutes some unknown harmful substances that accumulate in old blood. From this perspective, there’s no particular need for young stuff: Any form of plasma replacement will do. It’s sort of like changing the oil in your car.

The Conboys, now both at the University of California, Berkeley, suspect this is the case and are moving forward with tests of the idea. Their recent experiments, published in the journal  Aging , replaced half the blood of some old mice with a mix of salt water and purified albumin (the main protein in plasma), which successfully rejuvenated the rodents’ hearts, livers, and brains. They too are starting a company and are aiming for human clinical trials to determine if simply flushing out the bloodstream can help with problems like frailty and declining cognition.

At this point, the quest for a treatment has no satisfying ending. We know blood does indeed perform some kind of alchemy that can restore and remodel the flesh or hasten its decay. But even as that core mystery lingers, Villeda and others are rushing forward with a bigger project: cracking all the other codes that might be written in blood.

SAUL VILLEDA IS NOW 40. His thick black hair still has no tinge of gray. He speaks quickly and laughs often and generally hums with energy. He still seems young, but he is no longer a newbie. At UCSF, he now oversees a group advancing a new era in rejuvenation research. It’s looking at other systemic bodily shifts, such as those caused by exercise or diet, to find the mechanisms that can turn back the clock—demonstrating that youthfulness alone is not the only fountain of youth.

Soon after Villeda started his lab in 2013, his postdoc Shelly Fan was eager to begin a risky project. It’s well known that exercise can reduce some of the effects of aging on the brain, increasing blood flow to the organ and boosting cell birth in one of the few regions that produce new neurons. The junior researcher wanted to see whether plasma from an active animal could transmit those benefits to a sedentary one—but it would take many years of work to find out.

Villeda was now the senior scientist, fretting that his fearless young-blood collaborator was taking too big a risk. But he gave her the go-ahead. Shortly after the project got underway, Horowitz took over working with those Lego-snuffing critters. She’s spent three years watching mice age, watching them run, watching them remember and forget. “It’s long and grueling and tiring,” she says.

Mature mice were allowed to sprint as much as they wanted on little exercise wheels for six weeks (these critters typically like a nice, brisk jog). She then collected their plasma and delivered it to aged couch-potato equivalents. These older animals’ brains produced extra new neurons, and they aced memory tests. The paper was published in  Science  in summer 2020.

The surprise was that the effects seemed to flow through the liver, which ramped up several factors including an enzyme called GPLD1 that is also plentiful in active elderly humans. Rando and Wyss-Coray, with others, published similar results in  Nature Metabolism ; they found that serum (plasma with clotting factors and platelets removed) taken from exercising older mice restarted the systems responsible for muscle repair.

In addition to exercise, Villeda has played with a regimen known as caloric restriction that cuts food intake by 20 to 30 percent. Historically, the practice has improved age-related declines in brain, metabolism, and cardiac function in lab animals. At Stanford, Rando’s group is testing a high-fat, low-carb ketogenic diet. Others are interested in the effects of short-term physical stress (such as from bursts of intense exercise, or maybe even a small dose of radiation).

This story originally appeared in the Youth issue of Popular Science. Current subscribers can access the whole digital edition  here , or  click here  to subscribe.

The Hot, Young Blood Transfusion Idea Taking Over Silicon Valley

Parabiosis has a sketchy history, but that's not stopping Peter Thiel from believing in its potential.

Silicon Valley has cemented its place as a disruptor in various fields, but its most revolutionary experiment is happening right under our noses with a modern-day quest for the fountain of youth.

It’s weird, but the fountain of youth might be filled with the blood of the young.

As a New Yorker article on tech’s quest for immortality highlights, venture capitalists and scientists have centered their focus on parabiosis, a process that boils down to a blood transfusion from a young person to an older person. The theory is radically simple: Young blood has stronger, more vibrant, more active cells, while older blood has tired ones that aren’t as capable or fresh. Take the young blood, pour it into an older human, and the latter should see a spring in their step, an extra beat in their hearts, and — hopefully — a delay in the eventual aging of their body.

Parabiosis, however, isn’t technically real — at least, in the literal sense. The word is derived from the Latin for “living beside,” a sort of symbiosis that morphs two different bodies together.

Parabiosis Has Been Around Since the Mid-1800s

And while parabiosis might just be picking up in Google searches thanks to the New Yorker ’s deep dive into the immortality business, it’s existed for at least a century, starting with French zoologist Paul Bert in the mid-1800s, who first suggested that parabiosis could save lives by switching circulatory systems . The next instance of parabiosis, however, didn’t go so well, as the New Yorker ’s Tad Friend recounts:

In 1924, the physician and Bolshevik Alexander Bogdanov began young-blood transfusions, and a fellow-revolutionary wrote that he “seems to have become seven, no, ten years younger.” Then Bogdanov injected himself with blood from a student who had both malaria and tuberculosis, and died.

Bogdanov’s death effectively killed off any further research into parabiosis and was forgotten until the 1950s, when Cornell University researchers conducted crude experiments with rats and their circulatory systems . Surprisingly, the study offered a glimmer of hope: older rats revived their decaying cartilage, rejuvenating them to youthful levels.

That experiment, too, fizzled and remained simply a weird biological quirk until 2004, when Thomas Rando at Stanford University stumbled on a fact that was the opposite of aging theory. Before, it was assumed that older tissue had fewer stem cells compared to their younger counterparts, and that that drove aging. But Rando’s research showed that older tissue seemed to contain the same amount of stem cells as younger tissue. Rando turned to the 1950s parabiosis experiments, and swapped the blood of young and old mice .

After five weeks, Rando found an astonishing reversal: The younger mice had started aging, their stem cells lagging and their muscles dragging. The older mice, however, were hyped up on new cells that made their livers youthful, their hearts stronger, and practically reversed aging.

What Peter Thiel Thinks of Parabiosis

Parabiosis, all of a sudden, was something that potentially held the key to eternal youth, and noted Silicon Valley venture capitalist/immortality obsessive Peter Thiel noticed.

But that’s where the problems actually begin: Not only are blood transfusions notoriously risky medical business, but the science of swapping blood is fuzzy at best. Tricking the body into rapid cell multiplication like that of these experiments can lead to cancer , and the transaction of blood from one individual to another is notoriously dangerous, with some bodies outright rejecting this new blood. At best, a person could get sick. At worst, they’re dead.

The risks and muddled science of parabiosis, however, haven’t stopped Silicon Valley from waging war against death. Its rise in popularity correlates with the meteoric rise of Thiel, notorious disbeliever of death and parabiosis champion , who once told Inc that he found parabiosis “really interesting.” “Why is everyone else so indifferent about their immortality?” he wondered aloud to the New York Times ’ Maureen Dowd.

He might be wrong on that — and much of the credit goes to Thiel’s relentless pursuit to chase what he views as a personal mission to not just live longer, but to be immortal . He’s singlehandedly brought parabiosis into the popular lingo thanks to his seemingly bottomless well of cash, funneling funds into obscure Silicon Valley bio-startups that focus entirely on not repeating Bogdanov’s deadly mistake and ensuring Thiel his dream of being an eternal vampire. An October 2016 Inc article linked Thiel to the biotech firm Ambrosia, which was seeking federal approval to conduct clinical trials of parabiosis using donor plasma from people 25 and under.

“The effects seem to be almost permanent … like there’s a resetting of gene expression ,” Thiel’s personal medical consultant — a physician named Jesse Karmazin — told Inc . (Kamazin changed his story in June of this year , telling Techcrunch that he was never contacted by anyone at Thiel Capital. Inc stood by its original reporting.)

Alkahest — a competing company — is neck and neck with Ambrosia in churning out results that might make aging seem obsolete, though its results have only been shown in mice. Last November, the company announced it had used human teen blood, injected it into old mouse blood, and remarkably reversed aging . From tired, slow, and decrepit, the elderly mice suddenly had sharp memories, renewed cognition, and were exercising with the vigor of youth. “It’s more or less what we would expect,” Victoria Bolotina of Boston University told New Scientist . “The blood of young people must have something in it that’s important for keeping them young.”

For now, it’s safe to say that Thiel’s status as trusted advisor to the Trump administration guarantee that parabiosis research will move forward. It’s hard to say whether his obsession with youth will make parabiosis a scientific reality. But we’re closer than ever before to a future where we might stop by a clinic, juice up on young blood, and walk out with no expiration date.

• Architecture and Design
• Asian and Pacific Studies
• Business and Economics
• Classical and Ancient Near Eastern Studies
• Computer Sciences
• Cultural Studies
• Engineering
• General Interest
• Geosciences
• Industrial Chemistry
• Islamic and Middle Eastern Studies
• Jewish Studies
• Library and Information Science, Book Studies
• Life Sciences
• Linguistics and Semiotics
• Literary Studies
• Materials Sciences
• Mathematics
• Social Sciences
• Sports and Recreation
• Theology and Religion
• Publish your article
• The role of authors
• Promoting your article
• Abstracting & indexing
• Publishing Ethics
• Why publish with De Gruyter
• How to publish with De Gruyter
• Our book series
• Our subject areas
• Your digital product at De Gruyter
• Contribute to our reference works
• Product information
• Tools & resources
• Product Information
• Promotional Materials
• Orders and Inquiries
• FAQ for Library Suppliers and Book Sellers
• Repository Policy
• Free access policy
• Open Access agreements
• Database portals
• For Authors
• Customer service
• People + Culture
• Journal Management
• How to join us
• Working at De Gruyter
• Mission & Vision
• De Gruyter Foundation
• De Gruyter Ebound
• Our Responsibility
• Partner publishers

Your purchase has been completed. Your documents are now available to view.

The Fountain of Youth: A tale of parabiosis, stem cells, and rejuvenation

Transfusion (or drinking) of blood or of its components has been thought as a rejuvenation method since ancient times. Parabiosis, the procedure of joining two animals so that they share each others blood circulation, has revitalized the concept of blood as a putative drug. Since 2005, a number of papers have reported the anti-ageing effect of heterochronic parabiosis, which is joining an aged mouse to a young partner. The hallmark of aging is the decline of regenerative properties in most tissues, partially attributed to impaired function of stem and progenitor cells. In the parabiosis experiments, it was elegantly shown that factors derived from the young systemic environment are able to activate molecular signaling pathways in hepatic, muscle or neural stem cells of the old parabiont leading to increased tissue regeneration. Eventually, further studies have brought to identify some soluble factors in part responsible for these rejuvenating effects, including the chemokine CCL11, the growth differentiation factor 11, a member of the TGF-β superfamily, and oxytocin. The question about giving whole blood or specific factors in helping rejuvenation is open, as well as the mechanisms of action of these factors, deserving further studies to be translated into the life of (old) human beings.

1 Introduction

In the 16th century, the famous Spanish explorer and conquistador Juan Ponce de León led the expedition around the Caribbean islands and eventually into Florida to find the Fountain of Youth, a magical water source supposedly capable of reversing the aging process and curing sickness [ 1 ]. Although the explorer made no mention of the Fountain of Youth in his letters, led by the rumors, the expedition continued the search and many perished. The Fountain was nowhere to be found, as locals were unaware of its exact location. Fontaneda wrote in his memories: “So earnestly did they engage in the pursuit, that there remained not a river nor a brook in all Florida, not even lakes and ponds, in which they did not bathe; and to this day they persist in seeking that water, and never are satisfied. […] …and it ended in all that numerous people who went over to Carlos forming a settlement: but to this day youth and age find alike that they are mocked, and many have destroyed themselves”.

This is just one example of the continuous search human beings have dedicated to finding a way to live forever. Interestingly, another commonly cited approach for rejuvenation was attempting to transfer the warmth and fluids of youth from young people to old. Some examples of this approach were sleeping with virgins, a practice prescribed also by scientific physicians during the 17 th and 18 th centuries [ 2 ], or bathing in or drinking blood [ 3 ].

Blood, plasma, and their derivatives are what modern medicine has produced to aid stem cell function and tissue regeneration and repair. Platelet-rich blood derivatives, such as platelet-rich plasma (PRP) and platelet-rich fibrin, produce and deliver growth factors with antiapoptotic and angiogenic properties, augmenting the regenerative capacity of stem and progenitor cells, either resident locally or administered exogenously [ 4 ]. PRP has become popular for use in various orthopedic surgical procedures to treat different conditions including osteoarthritis [ 5 , 6 ], in plastic surgery to improve graft survival [ 7 , 8 ] and to treat impending skin necrosis [ 9 ]. Thus, no doubt that the blood and derivatives can be employed with success in strategies of regenerative medicine, but even for the ‘holy grail’ of rejuvenation - the reversal of the aging process.

2 A tale of parabiosis

The claim that blood can rejuvenate our organs has been revitalized by one research group at the Stanford University School of Medicine in 2005 [ 10 ] and 2010 [ 11 ]. These studies stemmed out from observations which show that tissue regenerative capacity declines with age. In tissues such as muscle, blood, liver, and brain this decline has been attributed to a diminished responsiveness of tissue-specific stem and progenitor cells [ 12 , 13 , 14 , 15 ]. However, aged muscle successfully regenerates when grafted into muscle in a young host, but young muscle displays impaired regeneration when grafted into an aged host [ 16 , 17 ]. Either local or systemic factors could be responsible for these reciprocal effects. In order to test whether systemic factors can support the regeneration of tissues in young animals and/or inhibit regeneration in old animals, the the paper by Conboy and colleagues of 2005 reported an experimental setup in which – in contrast to transplantation – regenerating tissues in aged animals are exposed only to circulating factors of young animals, and vice versa [ 10 ]. Thus, they established parabiotic pairings between young and old mice (heterochronic parabioses), with parabiotic pairings between two young mice or two old mice (isochronic parabioses) serving as controls ( Fig. 1 ). In parabiosis, two mice are surgically joined, such that they develop a shared blood circulation with rapid and continuous exchange of cells and soluble factors at physiological levels through their common circulatory system [ 18 ]. Parabiosis was invented in 1864 by the physiologist Paul Bert in order to see whether a shared circulatory system was created. Clive McCay, a biochemist and gerontologist at Cornell University in Ithaca, New York, was the first to apply parabiosis to the study of ageing, but this technique fell out of favour after the 1970s, likely because many rats died from a mysterious condition termed parabiotic disease, which occurs approximately one to two weeks after partners are joined, and may be a form of tissue rejection. Only at the beginning of the XXI century, Irving Weissman and Thomas A. Rando at the Stanford University brought parabiosis back to life, to study the movement and fate of blood stem cells.

Parabiosis. Two mice are stitched together sharing a common bloodstream. Heterochronic parabiosis is when a young mouse is surgically joined to aged partners, while isochronic parabiosis is referred to pairs of young-young or old-old animals. Modified from ref. [ 36 ] with permission of Nature Publishing Group.

The Stanford group investigated muscle regeneration and liver cell proliferation in the parabiosis setting. After muscle injury, muscle regeneration was studied by the formation of myotubes expressing embryonic myosin heavy chain, a specific marker of regenerating myotubes in adult animals. Five days after injury, muscles in young animals in both isochronic and heterochronic parabioses had regenerated robustly. In contrast, injured muscle from old isochronic parabionts regenerated poorly. Notably, parabiosis with young mice significantly enhanced the regeneration of muscle in old partners. The regeneration of aged muscle was almost exclusively due to the activation of resident, aged progenitor cells, and not to the engraftment of circulating progenitor cells from young partners (as judged by the presence of less than 0.1% of green fluorescent protein [GFP] expressing cells derived from young partners transgenic for GFP). Since the loss of muscle regeneration with age is in part due to an age-related impairment in the up regulation of Notch ligand Delta after muscle injury [ 12 ], Delta expression was also studied. Notably, satellite cells from the aged partners of heterochronic parabionts showed a marked up regulation of Delta, comparable to that found in their young partners and in young mice not subjected to parabiotic pairings ( Fig. 2 ).

Ageing of muscles, liver and brain in old mice and rejuvenation by heterochronic parabiosis. The regeneration of skeletal muscle upon injury is linked with the up regulation of the Notch ligand Delta, that is lost with age (upper panels). Hepatocyte proliferation in young animals correlates with the decrease of cEBP- α -brahma (cEBP- α -Brm) complex as compared with aged mice (middle panels). While young animals can increase their neurogenesis and angiogenesis in the subventricular zone of the brain, where neural stem cells are present, aged animals cannot (lower panels). In principle, the heterochonic parabiosis reverts all phenotypic and molecular hallmarks of ageing by transferring soluble factors and cells.

In the case of liver studies, and as in muscle, while proliferation of albumin-positive cells in old isochonic parabionts was less than that observed in young isochronic parabionts, parabiosis to a young partner significantly increased hepatocyte proliferation in aged mice. As also in muscle, the enhancement of hepatocyte proliferation in aged mice was due to resident cells and not the engraftment of circulating cells from young partners. The decline of hepatocyte progenitor cell proliferation is due to the formation of a complex involving cEBP- α and the chromatin remodeling factor brahma (Brm) that inhibits E2F-driven gene expression [ 19 ]. In parallel with the effect on hepatocyte regeneration, the formation of cEBP- α -Brm complex was detected in livers from old heterochronic parabionts but not from young isochronic ones, and the complex was diminished in old heterochronicparabionts ( Fig. 2 ). Finally, in muscle and liver, they noticed a reduction of progenitor cell proliferation in young mice after parabiotic pairing with old mice, suggesting that old mice are enriched with inhibitory factors which are diluted upon parabiosis. Overall, this data indicated that there are systemic factors that can modulate the molecular signaling pathways critical to the activation or inhibition of tissue-specific progenitor cells, and that the systemic environment of a young animal is one that promotes successful regeneration, whereas that of an older animal either fails to promote or actively inhibits successful tissue regeneration. Finally, this work also demonstrated that tissue-specific stem/progenitor cells retain much of their intrinsic proliferative potential even when old, but that age-related changes in the systemic environment and/or niche in which progenitor cells reside, preclude full activation of these cells for productive tissue regeneration.

In the paper of 2010, Wagers and colleagues tried to figure out what is the role of local micro environmental niche-related and systemic factors in ageing of hematopoietic stem and progenitor cells (HSPCs), using the in vivo parabiotic mouse system and studying HSC frequency and number of long-term HSCs (LT-HSCs). Congenic markers were used to distinguish HSCs from aged versus young partners. Ageing is accompanied at the level of bone marrow by a considerable expansion of HSPCs coupled paradoxically with a reduced capacity for blood reconstitution and skewed differentiation potential after transplant [ 20 , 21 , 22 , 23 ]. Aged-heterochronic parabionts showed significant reduction of LT-HSCs, which approached normal ‘youthful’ levels. Notably, this effect arose from changes in the aged HSC population itself and not to trafficking of ‘young’ cells to the aged partners’ marrow. Moreover, heterochronic parabiosis also induced recovery of LT-HSC function in aged mice, as evidenced by engraftment potential and restoration of youthful ratios of B lymphoid to myeloid cells. As with HSCs, both the frequency and total number of osteoblastic niche cells isolatable from aged mice were increased compared to young controls. In vitro experiments of interaction between young bone marrow cells with aged osteoblastic niche cells also showing expansion of HSCs, suggested that the HSC rejuvenating effects of heterochronic parabiosis occur indirectly – by reverting age-related changes in osteoblastic niche cells. Indeed, osteoblast frequency and number were restored to youthful levels when aged animals experienced in heterochonic parabiosis a young systemic environment. Moreover, niche cells isolated from aged-heterochronic parabiosis showed a significantly reduced capacity to cause HSPC accumulation, in contrast to niche cells from aged-isochronic parabiosis. Interestingly, osteoblast niche cells isolated from young heterochronic parabionts induced a slight expansion of young HSPCs as compared to the ones from young-isochronic parabionts. These in vitro studies suggested a reciprocal effect of the aged circulatory environment on niche activity in young heterochronic partners and indicate that systemic signals restore aged niches. Further experiments have been done, aimed at evaluating the ability of young HSPCs to reconstitute hematopoiesis. The results, demonstrated that, similar to the impaired engraftment function of naturally aged HSCs [ 20 ], young HSPCs exposed in vitro to aged-isochronic niched cells exhibited a reduced capacity for hematopoietic reconstitution. This indicates that interaction with aged osteoblastic niche cells is sufficient to induce defects in HSC function. However, aged heterochronic niche cells did not alter the reconstituting activity of young HSCs. Together, this data demonstrated that age-induced, functional alterations in HSC-regulatory niche cells can be reversed by young circulating factors. Overall, these findings further suggest that the rejuvenating effects of a young circulation on HSC are communicated indirectly – by signaling from rejuvenated osteoblastic niche cells.

In order to understand which factor is involved in the regulation of niche cell function, the authors sought to investigate whether insulin-like growth factor-1 (IGF-1) could play a role. IGF-1 has been shown to be an evolutionarily conserved ageing and longevity regulator [ 24 ]. In vitro and in vivo experiments demonstrated that local, not systemic, IGF-1 seems to induce ageing of HSC-regulatory niche cells, and that neutralization of IGF-1 signaling in the bone marrow microenvironment reverts age-related changes in osteoblastic niche cells that impair their appropriate regulation of HSCs.

Overall, these findings suggested that while under youthful conditions osteoblastic niche cells promote homeostatic stem-cell maintenance, they are altered by ageing such that instead allow the enhanced accumulation of dysfunctional HSCs. These age-specific alterations in niche cells seem to be signaled by an uncharacterized circulating factors that act in part by altering IGF-1 signaling in the niche cell themselves ( Fig. 3 ). It is likely that IGF-1 has not a major role for all the aged tissue, since while its role in the osteoblastic niche is age promoting, in contrast in skeletal muscle local expression of IGF-1 maintains regenerative capacity in aged animals.

Proposed model describing age-related changes in osteoblastic cell niche and HSCs, and how these changes may be reverted by heterochronic parabiosis. Age-specific changes in autocrine or paracrine effects of IGF-1 on osteoblastic niche cells are signaled by circulating soluble factors which themselves change with age. IGF-1 signaling in aged osteoblastic niche cells ( a ) directly contributes to age-related dysfunction in HSCs, including HSC over-accumulation and skewed B lymphoid (B cell)/myeloid (My) fate choice. Following heterochronic parabiosis, or after neutralization of IGF-1 signaling in vivo ( b ), the “youthful” activity of aged niche cells is restored, such that they no longer induce over-accumulation or lineage skewing of HSCs. From ref. [ 11 ] with permission of Nature Publishing Group.

In October 2010, three of the four authors, including Amy J. Wagers, retracted this paper, in particular for the role of the osteoblastic niche cells in the rejuvenation of HSCs in aged mice [ 25 ]. It was found that the first author manipulated the images for bone nodules formed in osteoblastic niche cells from young and aged mice (Retraction Watch, http://retractionwatch.com/2012/08/29/ori-finds-harvard-stem-cell-lab-post-doc-mayack-manipulated-images/ ). Thus, further confirmation about his issue should be obtained, considering also that the parabiosis model was exploited for studying rejuvenation of other old organs. Indeed, two papers subsequently appeared showing that exposing ayoung mouse to an old systemic environment can inhibit myogenesis [ 26 ] and neurogenesis [ 27 ].

In 2013, the team led by Amy J. Wagers, published another work by which they demonstrated using the parabiosis model, that age-related cardiac hypertrophy can be reversed by exposure to a young circulatory environment with only 4 weeks of parabiosis [ 28 ]. The measurement of blood pressures and of circulating levels of angiotensin II and aldosterone in the various groups, clearly demonstrated that the reversal of cardiac hypertrophy in old mice exposed to a young circulation could not be explained by a simple reduction in blood pressure or in the modulation of known effectors of blood pressure in the older mice. Interestingly, heterochronic parabiosis induced no changes in heart weight-to-tibia ratio, cardiomyocyte size, or blood pressure in young mice joined to aged partners. This data implicated an antihypertrophic factor produced by young mice (rather than dilution of a prohypertrophy factor produced by old mice) in the cardiac remodeling induced by heterochronic parabiosis. In “sham parabiosis”, whereby mice are surgically joined while leaving the skin intact, such that they do not develop a shared circulation, no significant difference in heart weight-to-tibia length ratio in aged mice was found, further indicating that cross-circulation and exchangeof blood-borne factors are required for reversal of age-related cardiac hypertrophy. In order to identify these factors, a broad-scale proteomics analysis using aptamer-based technology revealed 13 analytes that reliably distinguished young mice from old mice. One of these candidates, the growth differentiation factor 11 (GDF11), a member of the activin/TGF- β superfamily, was confirmed in further analyses. GDF11 was reduced in the plasma of old isochronic compared to young isochronic mice and was restored to youthful levels in old mice after exposureto a young circulation. A daily 30-day treatment of old mice with GDF11 led to a significant reduction in the heart weight-to-tibia length ratio compared to the saline-injected control group. This data suggested that at least one pathologic component of age-related diastolic heart failure is hormonal in nature. However, the observed regression of cardiac hypertrophy in old mice exposed to a young circulation is unlikely to be attributable entirely to the replenishment of a single factor, and other factors should be recognized.

Two subsequent studies by Wagers and colleagues found that GDF11 boosted the growth of new blood vessels and neurons in the brain [ 29 ] and spurred stem cells to regenerate skeletal muscle at the sites of injuries [ 30 ]. In one of these papers [ 29 ], the mouse heterochronic parabiosis model revealed an increase in cerebral blood vessel volume and blood flow in response to young systemic factors, together with a higher self-renewal and differentiation in subventricular zone neural stem cell population, bringing an improvement in olfactory discrimination ( Fig. 2 ). Furthermore, they found that GDF11 could increase blood vessel volume as well as neurogenesis in old mice. Interestingly, blood from 15-month-old mice did not decrease neural stem-cell populations in the young brain, whereas older blood (21 months) provoked a detrimental effect, suggesting that older the animals higher the accumulation of deleterious systemic factors and/or lower protective young factors. On the other hand, in the paper authored by Rando and Tony Wyss-Corey as senior scientists, the chemokine CCL11/eotaxin was identified as an age-related blood factor associated with decreased neurogenesis and impaired learning and memory in mice [ 27 ]. What is to be determined is if CCL11 interacts directly with neural progenitor cells during aging influencing their differentiation capacity, or it has indirect actions by interactions with other neurogenic niche cell types.

In the other paper by Wagers and colleagues [ 30 ], it was demonstrated that satellite cells sorted from aged-heterochronic mice had improved myogenic differentiation capacity as well as lower DNA damage when compared with satellite cells from aged-isochronic controls. As for the reversion of age-related cardiac hypertrophy [ 28 ], the treatment of aged mice with daily intraperitoneal injections of recombinant GDF11 for 4 weeks increased numbers of satellite cells with intact DNA, as compared with cells from aged mice receiving vehicle alone. Moreover, in a model of muscle injury, GDF11 treatment of aged mice 28 days before injury and continued for 7 days thereafter restored more youthful profiles of myofiber caliber in regenerating muscle. Aged mice treated with GDF11 also showed increased average exercise endurance and grip strength.

In this last paper, they also found that in vitro exposure of aged satellite cells to GDF11, but not myostatin (another member of the TGF- β superfamily) or TGF- β 1, produced dose-responsive increases in satellite cell proliferation and differentiation, suggesting that GDF11, in contrast to myostatin, can act directly on satellite cells to alter their function.

3 A tale of surprise and new drugs

The results about the role of GDF11 in the rejuvenation of muscles were not without effect. The surprise came from the prior knowledge that myostatin, a close related protein to GDF11, was known to reduce myogenesis [ 31 ]. Indeed, a more recent paper has denied the results obtained by those investigations. Egerman et al. [ 32 ] carefully re-assessed this hypothesis and discovered that the previously used reagents to detect GDF11 were nonspecific, i.e. the antibodies could not distinguish between myostatin and GDF11. It also appeared that the total levels of myostatin/GDF11 actually increase with age, contradicting the prior reports [ 28 , 30 ]. A more specific assay for GDF11 showed a trend towards increased GDF11 levels in serum from older rats and humans compared to younger individuals. Moreover, GDF11 mRNA levels rose in rat muscle with increased age. In vitro experiments also showed that both GDF11 and myostatin induce the same signaling pathways (SMAD 2/3 and MAPK activation) at similar degrees in primary and immortalized human skeletal muscle cells and that differentiation of human primary myoblasts into myotubes was inhibited by GDF11 and myostatin. By using the same treatment protocol established by Sinha et al . [ 30 ], these authors found no differences in the regenerative capacity of skeletal muscle of aged mice treated with GDF11 or vehicle at the 7-day time point post-cardiotoxin injury. Rather, higher systemic levels of GDF11 were associated with impaired regeneration in young mice, as indicated by a greater number of very small myofibers in the GDF11-treated muscles. Finally, they also showed that treatment with GDF11 decreased the growth of adult and aged satellite cells cultures in a dose-dependent manner. Consistent with the demonstration in the present study that GDF11 increases with age, the same group had previously demonstrated that the downstream myostatin/GDF11 signaling pathway, characterizedby SMAD3 phosphorylation, was also elevated in the aged rat [ 31 ]. The implication of this study is that if old individuals are found with very high levels of GDF11, and this is coincident with muscle loss (sarcopenia), they are candidates for either GDF11-specific blockade or for a more general blockade of the GFD11, myostatin, and their receptors.

Although at first glance the data generated by Egerman and colleagues seemed to conflict with Amy Wagers team’s results, there could be multiple forms of GDF11 and only one could decrease with age, as reported by The Scientist in an e-mail correspondence with Amy Wagers [ 33 ]. Moreover, the Novartis group injured the muscle more extensively and then treated it with more GDF11 than Wagers’ group had done, so the results may not be directly comparable (the Novartis team used young animals and a dose of GDF11 three times higher). The fact is that the results published by Egerman and colleagues could help to explain the mechanism behind bimagrumab, an experimental Novartis treatment for muscle weakness and wasting [ 34 ]. The drug, which is currently in clinical trials, blocks myostatin — and perhaps GDF11 as well [ 35 ].

In brief, it is not doubted that young blood renews old mice, but Novartis team says that the Harvard group’s explanation is wrong. Probably the truth stays in the middle, and maintaining GDF11 levels in an appropriate physiological range would be essential for muscle health. It is also important to recall what Amy Wagers said: “We’re not de-ageing animals. We’re restoring function to tissues” [ 36 ]. Alternatively, other factors may act in this context. In 2014, Irina and Michael Conboy identified [ 37 ] one of the anti-aging factors circulating in the blood: oxytocin, a nonapeptide produced by hypothalamus, which is involved in parturition and bonding. They observed that oxytocin levels declined in old mice (18-24 months), and when injected subcutaneously into aged mice, oxytocin recovered the regenerating capacity of muscle cells upon cardiotoxin injury.

In the field of identification of specific rejuvenation factors, there are still many unresolved issues. For example, how CCL11 or GDF11 improve aged tissue specific stem cell microenvironment are largely undetermined. Moreover, some observations are difficult to be reconciled one with the other. For example, the blood from aged individuals, which reduces GDF11 in the serum [ 28 ], negatively affects neurogenesis and cognitive function in young individuals [ 27 ]. This is in contradiction with the fact that the old serum is diluted in young serum, calling in question whether microRNAs (miRNAs), or ther protein factors, could target the GDF11 pathway in order to increase or decrease its expression. What is relevant to our review is that on the basis of Wagers’ results, at least one company is attempting to replicate the effect in humans using blood plasma from healthy young people to treat patients with Alzheimer’s disease. It is interesting to recall that in 1972, two researchers at the University of California studied the lifespans of old–young rat pairs. Older partners lived for four to five months longer than controls, suggesting for the first time that circulation of young blood might affect longevity [ 38 ]. In September 2014, a start-up company, Alkahestin Menlo Park, California, began an open label, single group assignment clinical trial assessing the safety and efficacy of 1 unit of plasma from young donors (males, aged 30 or younger) to treat mild-to-moderate Alzheimer’s disease (ClinicalTrials.gov Identifier: NCT02256306). Results concerning the primary (symptoms and adverse events) as well as secondary outcomes (MRI and blood biomarkers) are awaited in this year. However, although this trial is designed over 4 weeks, concerns of this practice may arise: for example, the possibility that activation of stem cells over a long period of time would result in an increase in cancer incidence.

4 Pros and cons of whole blood versus specific factors

The advantage of giving young blood to an old person is that it may contain different rejuvenating factors that could have pleiotropic effects on many diseased organs at the same time. Indeed, whole blood is a mixture of cells, colloids and crystalloids. Each of these components can be separated from the others in order to obtain packed red blood cell (PRBC) concentrate, platelet concentrate, fresh frozen plasma and cryoprecipitate, used in different indications [ 39 ]. The transfusion of whole blood or its components is a safe clinical practice today and has many applications in regenerative medicine [ 40 ]. However, some of its components, such as plasma proteins, leucocytes, red cell antigens, plasma and pathogens, may give rise to adverse effects that may range from mild allergic manifestations to fatal reactions [ 39 ]. Moreover, we do not know for humans if, for example, plasma from a young donor contains factors beneficial to patients with muscular dystrophy or Alzheimer’s. For these reasons, the identification of specific factors helping specifically old organs and tissue to rejuvenate or heal may be a safer approach. Rando, Wagers and other would prefer to see testing for a specific blood factor or combination of known factors synthesized on the bench in the lab [ 36 ]. The disadvantage in this case is that the mechanism of action of identified factors (CCL11, GDF11, or oxytocin) is far to be fully understood. It can also be that these factors do not act directly, but they do depend on epigenetic mechanisms (e.g, miRNAs), or that other elusive blood components acts with them to compound their rejuvenation effects.

5 Conclusions

This is a tale of long search by human beings of the ‘fountain of youth’. Parabiosis has suggested over a long period of time that factors from young blood may help diseased or aged tissues to regenerate. Some would prefer to administer whole blood or its derivatives, such as plasma, while others are more akin to deliver specific factors or cocktails of factors. The best scenario would be to use patient’s own plasma or platelet-derived cytokines and growth factors to stimulate wound healing and tissue regeneration. Some hints are coming out from animal studies, but the link to humans is still to be found.

Conflict of interest

Conflict of interest statement: Authors state no conflict of interest

[1] de Escalante Fontaneda H. Fontaneda’s Memoir, Translation by Buckingham Smith, 1854, Available from: http://keyshistory.org . Accessed July 18, 2016 Search in Google Scholar

[2] Shapin S, Martyn C. How to live forever: lessons of history. BMJ 2000; 321: 1580-1582 10.1136/bmj.321.7276.1580 Search in Google Scholar PubMed PubMed Central

[3] Learoyd P. The history of blood transfusion prior to the 20th century--part 1. Transfus Med 2012; 22: 308-314 10.1111/j.1365-3148.2012.01180.x Search in Google Scholar PubMed

[4] Masoudi E, Ribas J, Kaushik G, Leijten J, Khademhosseini A. Platelet-Rich Blood Derivatives for Stem Cell-Based Tissue Engineering and Regeneration. Curr Stem Cell Rep 2016; 2: 33-42 10.1007/s40778-016-0034-8 Search in Google Scholar PubMed PubMed Central

[5] Lee JW, Kwon OH, Kim TK, Cho YK, Choi KY, Chung HY, et al. Platelet-rich plasma: quantitative assessment of growth factor levels and comparative analysis of activated and inactivated groups. Arch Plast Surg 2013; 40: 530-535 10.5999/aps.2013.40.5.530 Search in Google Scholar PubMed PubMed Central

[6] Patel S, Dhillon MS, Aggarwal S, Marwaha N, Jain A. Treatment with platelet-rich plasma is more effective than placebo for knee osteoarthritis: a prospective, double-blind, randomized trial. Am J Sports Med 2013; 41: 356-364 10.1177/0363546512471299 Search in Google Scholar PubMed

[7] Cervelli V, Palla L, Pascali M, De Angelis B, Curcio BC, Gentile P. Autologous platelet-rich plasma mixed with purified fat graft in aesthetic plastic surgery. Aesthetic Plast Surg 2009; 33: 716-721 10.1007/s00266-009-9386-0 Search in Google Scholar PubMed

[8] Choi J, Minn KW, Chang H. The efficacy and safety of platelet-rich plasma and adipose-derived stem cells: an update. Arch Plast Surg 2012; 39: 585-592 10.5999/aps.2012.39.6.585 Search in Google Scholar PubMed PubMed Central

[9] Kim YJ, Jeong JH. Clinical application of adipose stem cells in plastic surgery. J Korean Med Sci 2014; 29: 462-467 10.3346/jkms.2014.29.4.462 Search in Google Scholar PubMed PubMed Central

[10] Conboy IM, Conboy MJ, Wagers AJ, Girma ER, Weissman IL, Rando TA. Rejuvenation of aged progenitor cells by exposure to a young systemic environment. Nature 2005; 433: 760-764 10.1038/nature03260 Search in Google Scholar PubMed

[11] Mayack SR, Shadrach JL, Kim FS, Wagers AJ. Systemic signals regulate ageing and rejuvenation of blood stem cell niches. Nature 2010; 463: 495-500 10.1038/nature08749 Search in Google Scholar PubMed

[12] Conboy IM, Conboy MJ, Smythe GM, Rando TA. Notch-mediated restoration of regenerative potential to aged muscle. Science 2003; 302: 1575-1577 10.1126/science.1087573 Search in Google Scholar

[13] Fuller J. Hematopoietic stem cells and aging. Sci Aging Knowledge Environ 2002; 2002: pe11 10.1126/sageke.2002.25.pe11 Search in Google Scholar

[14] Sigal SH, Brill S, Fiorino AS, Reid LM. The liver as a stem cell and lineage system. Am J Physiol 1992; 263: G139-148 10.1152/ajpgi.1992.263.2.G139 Search in Google Scholar

[15] Kuhn HG, Dickinson-Anson H, Gage FH. Neurogenesis in the dentate gyrus of the adult rat: age-related decrease of neuronal progenitor proliferation. J Neurosci 1996; 16: 2027-2033 10.1523/JNEUROSCI.16-06-02027.1996 Search in Google Scholar

[16] Carlson BM, Faulkner JA. Muscle transplantation between young and old rats: age of host determines recovery. Am J Physiol 1989; 256: C1262-1266 10.1152/ajpcell.1989.256.6.C1262 Search in Google Scholar

[17] Carlson BM, Dedkov EI, Borisov AB, Faulkner JA. Skeletal muscle regeneration in very old rats. J Gerontol A Biol Sci Med Sci 2001; 56: B224-233 10.1093/gerona/56.5.B224 Search in Google Scholar

[18] Wright DE, Wagers AJ, Gulati AP, Johnson FL, Weissman IL. Physiological migration of hematopoietic stem and progenitor cells. Science 2001; 294: 1933-1936 10.1126/science.1064081 Search in Google Scholar

[19] Iakova P, Awad SS, Timchenko NA. Aging reduces proliferative capacities of liver by switching pathways of C/EBPalpha growth arrest. Cell 2003; 113: 495-506 10.1016/S0092-8674(03)00318-0 Search in Google Scholar

[20] Rossi DJ, Bryder D, Zahn JM, Ahlenius H, Sonu R, Wagers AJ, et al. Cell intrinsic alterations underlie hematopoietic stem cell aging. Proc Natl Acad Sci U S A 2005; 102: 9194-9199 10.1073/pnas.0503280102 Search in Google Scholar PubMed PubMed Central

[21] Rossi DJ, Jamieson CH, Weissman IL. Stems cells and the pathways to aging and cancer. Cell 2008; 132: 681-696 10.1016/j.cell.2008.01.036 Search in Google Scholar PubMed

[22] Morrison SJ, Wandycz AM, Akashi K, Globerson A, Weissman IL. The aging of hematopoietic stem cells. Nat Med 1996; 2: 1011-1016 10.1038/nm0996-1011 Search in Google Scholar PubMed

[23] Liang Y, Van Zant G, Szilvassy SJ, Effects of aging on the homing and engraftment of murine hematopoietic stem and progenitor cells. Blood 2005; 106: 1479-1487 10.1182/blood-2004-11-4282 Search in Google Scholar PubMed PubMed Central

[24] Musaro A, McCullagh K, Paul A, Houghton L, Dobrowolny G, Molinaro M, et al. Localized Igf-1 transgene expression sustains hypertrophy and regeneration in senescent skeletal muscle. Nat Genet 2001; 27: 195-200 10.1038/84839 Search in Google Scholar PubMed

[25] Retraction. Systemic signals regulate ageing and rejuvenation of blood stem cell niches. Nature 2010; 467: 872 10.1038/nature09474 Search in Google Scholar PubMed

[26] Brack AS, Conboy MJ, Roy S, Lee M, Kuo CJ, Keller C, et al. Increased Wnt signaling during aging alters muscle stem cell fate and increases fibrosis. Science 2007; 317: 807-810 10.1126/science.1144090 Search in Google Scholar PubMed

[27] Villeda SA, Luo J, Mosher KI, Zou B, Britschgi M, Bieri G, et al. The ageing systemic milieu negatively regulates neurogenesis and cognitive function. Nature 2011; 477: 90-94 10.1038/nature10357 Search in Google Scholar PubMed PubMed Central

[28] Loffredo FS, Steinhauser ML, Jay SM, Gannon J, Pancoast JR, Yalamanchi P, et al. Growth differentiation factor 11 is a circulating factor that reverses age-related cardiac hypertrophy. Cell 2013; 153: 828-839 10.1016/j.cell.2013.04.015 Search in Google Scholar PubMed PubMed Central

[29] Katsimpardi L, Litterman NK, Schein PA, Miller CM, Loffredo FS, Wojtkiewicz GR, et al. Vascular and neurogenic rejuvenation of the aging mouse brain by young systemic factors. Science 2014; 344: 630-634 10.1126/science.1251141 Search in Google Scholar PubMed PubMed Central

[30] Sinha M, Jang YC, Oh J, Khong D, Wu EY, Manohar R, et al. Restoring systemic GDF11 levels reverses age-related dysfunction in mouse skeletal muscle. Science 2014; 344: 649-652 10.1126/science.1251152 Search in Google Scholar PubMed PubMed Central

[31] Trendelenburg AU, Meyer A, Jacobi C, Feige JN, Glass DJ. TAK-1/p38/nNFkappaB signaling inhibits myoblast differentiation by increasing levels of Activin A. Skelet Muscle 2012; 2:3 10.1186/2044-5040-2-3 Search in Google Scholar PubMed PubMed Central

[32] Egerman MA, Cadena SM, Gilbert JA, Meyer A, Nelson HN, Swalley SE, et al. GDF11 Increases with Age and Inhibits Skeletal Muscle Regeneration. Cell Metab 2015; 22: 164-174 10.1016/j.cmet.2015.05.010 Search in Google Scholar PubMed PubMed Central

[33] Grens K. Studies Conflict on Regenerative Molecule. The Scientist. May 19, 2015. Accessed April 22nd, 2017 Search in Google Scholar

[34] Lach-Trifilieff E, Minetti GC, Sheppard K, Ibebunjo C, Feige JN, Hartmann S, et al. An antibody blocking activin type II receptors induces strong skeletal muscle hypertrophy and protects from atrophy. Mol Cell Biol 2014; 34: 606-618 10.1128/MCB.01307-13 Search in Google Scholar PubMed PubMed Central

[35] Reardon S. ‘Young blood’ anti-ageing mechanism called into question, 19 May 2015, Available from: http://www.nature.com/news/young-blood-anti-ageing-mechanism-called-into-question-1.17583 Accessed 10.1038/nature.2015.17583 Search in Google Scholar

[36] Scudellari M. Ageing research: Blood to blood. Nature 2015; 517: 426-429 10.1038/517426a Search in Google Scholar PubMed

[37] Elabd C, Cousin W, Upadhyayula P, Chen RY, Chooljian MS, Li J, et al. Oxytocin is an age-specific circulating hormone that is necessary for muscle maintenance and regeneration. Nat Commun 2014; 5: 4082 10.1038/ncomms5082 Search in Google Scholar PubMed PubMed Central

[38] Ludwig FC, Elashoff RM. Mortality in syngeneic rat parabionts of different chronological age. Trans N Y Acad Sci 1972; 3: 582-587 10.1111/j.2164-0947.1972.tb02712.x Search in Google Scholar PubMed

[39] Basu D, Kulkarni R. Overview of blood components and their preparation. Indian J Anaesth 2014; 58: 529-537 10.4103/0019-5049.144647 Search in Google Scholar PubMed PubMed Central

[40] Corigiano M, Ciobanu G, Baldoni E, Pompa G. Biochemical assessment of growth factors and circulation of blood components contained in the different fractions obtained by centrifugation of venous blood. J Biol Regul Homeost Agents 2014; 28: 301-315 Search in Google Scholar

© 2017 Massimo Conesea et al.

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.

• X / Twitter

Supplementary Materials

Please login or register with De Gruyter to order this product.

Journal and Issue

Articles in the same issue.

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 09 March 2023

Heterochronic parabiosis reprograms the mouse brain transcriptome by shifting aging signatures in multiple cell types

• Methodios Ximerakis   ORCID: orcid.org/0000-0002-2815-7558 1 , 2 , 3   na1 ,
• Kristina M. Holton 1 , 2 , 3   na1 ,
• Richard M. Giadone   ORCID: orcid.org/0000-0003-4523-3062 1 , 2 ,
• Ceren Ozek 1 , 2 ,
• Monika Saxena 1 , 2 ,
• Samara Santiago   ORCID: orcid.org/0000-0002-9502-7104 1 , 2 ,
• Xian Adiconis 3 , 4 ,
• Danielle Dionne 4 ,
• Lan Nguyen 4 ,
• Kavya M. Shah   ORCID: orcid.org/0000-0002-8722-4095 1 , 2 ,
• Jill M. Goldstein 1 , 2 ,
• Caterina Gasperini 1 , 2 ,
• Ioannis A. Gampierakis 1 , 2 ,
• Scott L. Lipnick 1 , 2 , 3 ,
• Sean K. Simmons 3 , 4 ,
• Sean M. Buchanan 1 , 2 ,
• Amy J. Wagers 1 , 2 , 5 , 6 ,
• Aviv Regev   ORCID: orcid.org/0000-0003-3293-3158 4 , 7 ,
• Joshua Z. Levin   ORCID: orcid.org/0000-0002-0170-3598 3 , 4 &
• Lee L. Rubin   ORCID: orcid.org/0000-0002-8658-841X 1 , 2 , 3  

Nature Aging volume  3 ,  pages 327–345 ( 2023 ) Cite this article

22k Accesses

16 Citations

72 Altmetric

Metrics details

• Alzheimer's disease
• Blood–brain barrier
• Gene regulatory networks
• Neural ageing
• Neuro–vascular interactions

Aging is a complex process involving transcriptomic changes associated with deterioration across multiple tissues and organs, including the brain. Recent studies using heterochronic parabiosis have shown that various aspects of aging-associated decline are modifiable or even reversible. To better understand how this occurs, we performed single-cell transcriptomic profiling of young and old mouse brains after parabiosis. For each cell type, we cataloged alterations in gene expression, molecular pathways, transcriptional networks, ligand–receptor interactions and senescence status. Our analyses identified gene signatures, demonstrating that heterochronic parabiosis regulates several hallmarks of aging in a cell-type-specific manner. Brain endothelial cells were found to be especially malleable to this intervention, exhibiting dynamic transcriptional changes that affect vascular structure and function. These findings suggest new strategies for slowing deterioration and driving regeneration in the aging brain through approaches that do not rely on disease-specific mechanisms or actions of individual circulating factors.

Similar content being viewed by others

Pharmacologically reversible zonation-dependent endothelial cell transcriptomic changes with neurodegenerative disease associations in the aged brain

Molecular hallmarks of heterochronic parabiosis at single-cell resolution

Rejuvenation of the aged brain immune cell landscape in mice through p16-positive senescent cell clearance

Aging is a complicated process that is far from being completely understood, although a number of hallmarks have been recognized 1 . The brain itself is affected substantially by aging, with processes such as cellular respiration, protein synthesis, oxidative stress, neurotransmission, myelination, neurogenesis, inflammation and blood flow being compromised 2 , 3 . None the less, a series of recent observations demonstrate that several aspects of aging can be delayed or even reversed by a variety of interventions, including exercise 4 , 5 , caloric restriction 6 , elimination of senescent cells 7 , 8 , administration of rapamycin 9 or metformin 10 , transient cell reprogramming 11 and young bone marrow transplantation 12 .

One of the most robust methods of improving the function of aging tissues is that of heterochronic parabiosis, a surgical procedure whereby young and old mice are joined together so that they share a common circulatory system 13 , 14 . Multiple publications have led to the surprising conclusion that exposure of old mice to the young circulatory environment improves the function of various tissues and organs 15 , 16 , 17 , 18 , 19 , 20 , 21 , including the central nervous system (CNS) 22 , 23 , 24 , 25 . In the CNS specifically, studies from our lab 25 , 26 and others 24 , 27 , 28 , 29 , 30 have shown that circulating factors in young blood stimulate functional improvement in aged and diseased brains. Conversely, it has been shown that systemic factors in old blood drive aging phenotypes in young tissues 22 , 31 , 32 , including the brain. The molecular underpinnings of these changes remain to be fully elucidated, but a great deal of recent work has focused on measuring age-related changes in serum blood proteins 33 , primarily based on the hypothesis that function-improving factors decline with age. Although this is a plausible approach 25 , 28 , 29 , 34 , 35 , 36 , it ignores nonprotein factors (blood cells 23 , exosomes, lipids) and factors that either do or do not change in unexpected directions with aging.

Following our previously published work employing single-cell RNA-sequencing (scRNA-seq), which describes changes that occur in the brain during aging 37 , we have now quantified changes in the transcriptomes of young and old mouse brains after parabiosis. The comprehensive single-cell datasets we generated allowed us to detect brain cell types, the transcriptional states of which are affected by parabiosis. We also identified changes in inter-/intracellular molecular pathways, gene regulatory networks, cell–cell interactions and senescence. Endothelial cells (ECs) were found to be highly affected by parabiosis, exhibiting strong dynamic changes in their transcriptome, with several of these validated by orthogonal assays. Overall, this work shows that heterochronic parabiosis regulates several canonical hallmarks of aging 1 , 3 by shifting aging-induced changes of the transcriptome in a cell-type-specific manner.

Single-cell profiling to study rejuvenation and aging acceleration

We employed high-throughput scRNA-seq to examine the transcriptional profiles of young and old mouse brains after parabiosis (Fig. 1a,b ). We generated heterochronic pairs in which 3- to 4-month-old mice were joined with 20- to 22-month-old mice. We also generated age-matched isochronic pairs of young and old mice as controls. All pairs were maintained for 4–5 weeks before tissue collection and analysis. We confirmed successful parabiosis and establishment of blood crosscirculation as previously described 38 (Extended Data Fig. 1 ). We dissociated the brain tissues using our recently developed protocol 37 and analyzed the transcriptomes of 158,767 single cells (Extended Data Figs. 2 and 3 ). On stringent filtering and batch effect examination ( Methods ), we retained 105,329 cells, of which 67,992 cells derived from 34 parabionts (7 isochronic young (YY), 9 heterochronic young (YO), 7 isochronic old (OO), 11 heterochronic old (OY)) and 37,337 cells derived from 16 unpaired animals (8 young (YX) and 8 old (OX)) (Fig. 1 and Extended Data Figs. 4 and 5 ).

a , Schematic representation of the animal types used in the present study. Sequencing data from isochronic (YY and OO) and heterochronic (YO and OY) parabiosis pairs were generated and integrated with sequencing data from young (YX) and old (OX) unpaired mice from our previous work 37 which were generated simultaneously with those of the parabionts. b , Schematic representation of the experimental workflow (see Methods for details). c , UMAP projection of 105,329 cells across 31 clusters derived from 34 parabionts (7 YY, 9 YO, 7 OO and 11 OY) and 16 unpaired animals (8 YX and 8 OX). For the cell-type abbreviations please see the text and Methods (Supplementary Table 1 ). d , UMAP projection of five major cell populations showing the expression of representative, well-known, cell-type-specific marker genes (OLGs: Cldn11 ; ASCs: Gja1 ; NEUR: Syt1 ; ECs: Cldn5 ; MGs: Tmem119 ). The numbers reflect the number of nCount RNA (UMI) detected for the specified gene for each cell. e , Violin plot showing the distribution of expression levels of well-known, representative, cell-type-enriched, marker genes across all 31 distinct cell types. f , g , Boxplot showing the distribution over n  = 50 biologically independent animals of the number of detected cells per cell type ( f ) or number of detected genes per cell type ( g ). Boxplot minimum is the smallest value within 1.5× the interquartile range (IQR) below the 25th percentile and maximum is the largest value within 1.5× the IQR above the 75th percentile. Boxplot center is the 50th percentile (median) and box bounds are the 25th and 75th percentiles. Outliers are >1.5× and <3× the IQR. Panel b was created with BioRender.com .

Single-cell atlas reveals that cell types are preserved by parabiosis

By combining cells in unpaired and parabiotic brains and using cell markers from our previous work 37 , we identified 31 major cell types with distinct expression profiles (Fig. 1c ). For a complete list of cell types and abbreviations, see Methods and Supplementary Table 1 .

We then examined the major markers for each cell type to ensure robustness of their transcriptional signatures. Each cell type expressed markers that match its cellular identity, as previously characterized 37 (Fig. 1d,e ). Unsupervised clustering also showed that the identified cell populations were represented in all batches and animal types (Extended Data Figs. 4a,b and 6 ), indicating that cell identity is preserved in parabiotic mice. We found that the proportions of each cell type represented in each group were only slightly smaller for the isochronic brains, which may be due to the smaller number of contributing animals (Extended Data Fig. 7 ). We quantified the difference in the total number of cells for each cell population by comparing the animal types using analysis of variance (ANOVA) with a threshold of P  ≤ 0.05, confirming that parabiosis does not significantly alter the number of cells per cell population between animal types, except for dopaminergic neurons (DOPA) in young heterochronic versus young unpaired mice ( P  = 0.002) (Fig. 1f , Extended Data Figs. 4c and 7 and Supplementary Table 2 ). However, data analysis of cell proportion changes should be cautiously considered in single-cell sequencing studies, especially when tissue dissociation is used. Across the different types of cells, we observed the largest number of total detected genes in ependymocytes (EPCs), choroid plexus epithelial cells (CPCs), GABAergic neurons (GABA), glutamatergic neurons (GLUT), neuroendocrine cells (NendCs) 37 and arachnoid barrier cells (ABCs) (Fig. 1g ).

For further investigation, we performed high-resolution subclustering analysis to uncover the heterogeneity of these cell types. To reduce the effects of drastically different cell identities, we first grouped the identified cells into five distinct classes based on their expression profile, lineage, function and anatomical organization (Fig. 2 ). We delineated 75 distinct cell populations in accordance with the literature and our previous results 37 (Fig. 2a–e ). As with the major cell populations, all the identified subpopulations were represented in each animal type (Extended Data Fig. 6 ). For instance, we identified five subpopulations of ECs (Fig. 2f,g ): ECs only expressing classic markers such as Cldn5 (EC_1), which represent the largest fraction of ECs; ECs positive for astrocytic markers, such as Slc1a3 (EC_2) 39 , potentially due to the presence of adherent RNA-containing astrocytic endfeet; ECs expressing the mitogenic/neovascularization marker Lrg1 (ref. 40 ) (EC_3); ECs positive for the olfactory marker Omp (EC_4), possibly reflecting RNA in olfactory axons still attached to ECs; and ECs denoted by the expression of Plvap , known to be expressed in fenestrated ECs in the choroid plexus 41 and circumventricular regions (EC_5). Although EC_2 has been characterized by others, EC_2 and EC_4 may be consequences of the mild dissociation protocol used (Fig. 2h ). As found in our previous study 37 and by others 41 , 42 , we did not observe distinct separation of arterial, capillary and venous ECs. However, select markers exhibited a zonation effect that further highlights the heterogeneity of ECs derived from different vascular beds (Extended Data Fig. 8 ). More specifically, as shown in Fig. 2i , probabilistic programming of cell class assignment using arterial/capillary/venous markers characterized in recent studies 41 , 42 , 43 similarly displayed a clear zonation of ECs along the arteriovenous axis (Extended Data Fig. 8 ).

a – e , Subpopulation analysis of cell types grouped in five distinct cell classes: OLG lineage and OEGs ( n  = 41,873 cells) ( a ), astroependymal cells and NSCs ( n  = 19,520 cells) ( b ), neuronal lineage ( n  = 20,869 cells) ( c ), vasculature cells ( n  = 10,438 cells) ( d ) and immune cells ( n  = 12,629 cells) ( e ). q , quiescent; p , proliferating; c , committed, nf , newly formed; mf , myelin-formin; mt , mature. f , UMAP subpopulation analysis of EC clusters ( n  = 6,218 cells). g , UMAP subpopulation of EC clusters, stratified by animal type. h , Violin plot of delineating markers of ECs, as Cldn5 , Slc1a3 , Lrg1 , Omp and Plvap . i , UMAP overlay of EC zonation markers along the arteriovenous axis curated from the literature 41 , 42 , 43 . Markers in left-to-right order: large arteries: Fbln5 ; arterial: Gkn3; capillary–arterial: Tgfb2 ; capillary: Mfsd2a ; capillary: Cxcl12 ; capillary–venous: Car4 ; venous: Slc38a6 ; large veins: Lcn2 ; and large vessels: Vcam1 .

Parabiosis reprograms the transcriptional landscape of brain cell types

Adult mouse tissues experience structural and functional improvement when exposed to young blood and deterioration when exposed to old blood. Our previous work detailed transcriptional changes that take place in brain cells during aging 37 , but not all these changes are necessarily of functional consequence. Heterochronic parabiosis might provide a first-pass filter to help identify those aging-related genes that drive the reversal of aging in the brain. We performed differential gene expression (DGE) analysis to identify the gene changes that are associated with the rejuvenation process in old heterochronic parabionts or with the aging acceleration process in young heterochronic parabionts. In the analysis, we attempted to compensate for the effects of the parabiosis surgery itself. We used a pseudobulk approach ( Methods ) to allow for chained pairwise comparisons of parabiosis-induced rejuvenation (RJV) or aging acceleration (AGA) with respect to the parabiosis surgery and the physiological aging process 44 , 45 , 46 . We also completed all relevant pairwise comparisons to directly compare the transcriptomes of any two animal types. Overall, our data allowed us to identify signatures that either shifted more toward reversal (RJV) or acceleration (AGA) of aging. More specifically, the RJV DGE dataset lists the old heterochronic genes (OY) taking into account isochronic mice (OO) and are associated with a transition to a ‘more youthful’ state with respect to normal aging (OX–YX). Conversely, the aging AGA DGE dataset lists the young heterochronic genes (YO) taking into account isochronic mice (YY) and are associated with a transition to a ‘more aged’ state with respect to normal aging (YX–OX) ( Methods ).

Of the 20,905 total detected genes, 700 were significantly changed with aging in at least one cell type (false discovery rate (FDR) ≤ 0.05), whereas 442 were significantly changed in RJV and 155 in AGA (Supplementary Tables 3 – 11 ). As expected, we did not capture significant transcriptional changes in all the identified cell types, potentially due to the low number of cells sequenced in some populations and the low levels of detected transcription and dropouts inherent to scRNA-seq analysis 47 . None the less, our analysis also showed that the proportions of total DGEs per cluster that were common to aging and RJV varied from cell type to cell type, as did those that were common to aging and AGA (Extended Data Fig. 9 ). The DGEs in common between RJV and aging and AGA and aging, along with genes unique to each paradigm, demonstrate that not all effects of aging are reversed or accelerated by parabiosis and that heterochronic parabiosis may also work through genes and pathways independent of the aging process (Supplementary Tables 12 and 13 ).

On examination of the DGE datasets, a clear pattern of gene expression changes in RJV and AGA (FDR ≤ 0.05) was evident among our large cell populations (>5,000 cells) (Fig. 3 and Supplementary Tables 3 – 11 ), potentially due to higher statistical power. Specifically, oligodendrocytes (OLGs), astrocytes (ASCs), GLUT, GABA, ECs and microglia (MGs) yielded the largest number of putative DGEs (Figs. 3 and 4a and Supplementary Tables 3 – 11 ). Although OLGs had the largest number of cells, the number of DGEs normalized to the total number of OLGs in RJV, AGA and aging were 0.461, 0.00587 and 0.896, respectively. Conversely, ECs, an average cluster in size, had the third highest percentage of normalized RJV DGEs (1.0936) behind only NendCs (1.229) and GLUT (1.190), and had the highest percentage of AGA (0.836) and aging (2.638) DGEs. These data indicated that the EC population, most directly exposed to circulating factors, is highly affected in both the RJV and the AGA paradigms. This is in accordance with our previous observations showing improvements in the brain vasculature after heterochronic parabiosis 25 and after growth differentiation factor (GDF)11 treatment 26 , and further emphasizes the key role that ECs play in regulating communication between blood and the brain 30 .

a – f , An FDR ≤ 0.05 was used to identify significant DGE genes, with n denoting the total number of genes meeting this threshold. The RJV framework depicts normalized gene expression changes across YX, OY, OO and OX. The AGA framework depicts normalized gene expression changes across OX, YO, YY and YX. DGE genes are log 2 ( z -scored) scaled across rows (all animals) and are ordered by descending log(FC), with OLGs ( n  = 156 RJV and n  = 2 AGA) ( a ), ASCs ( n  = 49 RJV and n  = 48 AGA) ( b ), GABA ( n  = 39 RJV and n  = 20 AGA) ( c ), GLUT ( n  = 61 RJV and n  = 6 AGA) ( d ), ECs ( n  = 68 RJV and n  = 52 AGA) ( e ) and MGs ( n  = 63 RJV and n  = 37 AGA) ( f ) in respective order. The color bar of the heatmap reflects the z -score, from negative (blue) to positive (magenta). The batch is denoted in the top annotation bar and animal type in the second annotation bar.

a , Rose diagrams (circular histograms of number of DGEs) of aging, RJV and AGA across all cell types at FDR ≤ 0.05, colored by direction of log(FC) (up magenta, down blue). b , Venn diagram of RJV and AGA DGEs across all cell types, demonstrating bidirectional log(FC) changes between the comparisons (depicted with arrows). c , d , Upset plot of FDR = 0.05 DGE with positive log(FC) (upregulation) and negative log(FC) (downregulation) in both RJV ( c ) and AGA ( d ). The top bar height reflects the number of DGEs in the intersection (in common between the barbells below), and the side bar width reflects the magnitude of the set size. e , EC RJV and AGA DGE Venn diagram split by log(FC) sign, revealing genes that reverse direction between comparisons. The arrows point to listed bidirectional genes. f , GSEA dot plots (Benjamini–Hochberg-adjusted P value for multiple comparisons ( P adj ) ≤ 0.25) of representative terms across cell types in RJV and aging, with the size of dot proportional to inverse P adj and color by NES from negative (blue) to positive (magenta).

Parabiosis shifts gene signatures associated with RJV and AGA

To find genes that are dysregulated across multiple cell types, we looked for key gene signatures that change in opposite directions in RJV and AGA. Across all cell types, a total of 41 unique genes was found to change their expression levels bidirectionally in RJV and AGA. Specifically, 34 genes were downregulated in RJV and upregulated in AGA (Fig. 4b and Supplementary Table 14 ). To identify cell types that change the direction of expression of the same genes in the parabiosis-induced RJV process and/or the parabiosis-induced AGA process, we created higher-dimensional Venn diagrams through upset plots (Fig. 4c,d ). This analysis identified certain pairs of cell types that shared multiple DGEs in the same direction in RJV (such as ECs–MGs and OLGs–ASCs) and AGA (ECs–pericytes (PCs) and ECs–MGs) (Fig. 4c,d and Supplementary Tables 15 and 16 ). In multiple cell types in RJV, two major apolipoproteins ( Apoe and Clu ) were dysregulated, similar to findings in neurodegenerative diseases such as Alzheimer’s disease 48 , 49 , 50 (Supplementary Tables 15 and 16 ).

ECs again demonstrated the greatest number of genes that change direction of expression (log(fold-change) (log(FC))) between RJV and AGA, further highlighting their susceptibility to the aging process and their potential for manipulation. Overall, a substantial percentage of genes with expression that was increased with aging decreased their ECs after heterochronic parabiosis. For example, the transcription factor (TF) Maff , which is highly upregulated with aging in ECs, was found to be downregulated in RJV and upregulated in AGA (Fig. 4e and Supplementary Table 14 ). Similarly, the aging-upregulated genes Hsp90aa1 and Hspa1a , which encode heat shock response proteins, Adamsts1 , which is induced on shear stress 51 , Apold1 , which is responsive to hypoxia/ischemia 52 and lipopolysaccharide treatment 30 , Cyr61 , which encodes an extracellular matrix protein involved in angiogenesis 53 , and Dusp1 , which participates in EC migration 54 along with Stmn2 , a gene involved in the microtubule organization 37 , 39 , were all downregulated in RJV and upregulated in AGA (Fig. 4e and Supplementary Table 14 ). There were only two genes with expression that changed in the opposite direction—down with aging and up after heterochronic parabiosis. One is Avp , a hormone that has been known for some time to signal to brain ECs, which, however, were not known to secrete it. The other is Ivns1ab , known to stabilize the cytoskeleton in some cell types and protect from cell death induction due to actin destabilization 55 (Fig. 4e and Supplementary Table 14 ). The DGE analysis also showed that the TF Klf6 , which was one of the most upregulated genes with aging in ECs, was differentially downregulated in RJV, as was its downstream target Smad7 (Supplementary Tables 3 and 4 ). Considering that Klf6 expression is known to be regulated by vascular injury 56 , 57 , we further characterized its transcriptional changes by RNA in situ hybridization. As shown in Fig. 5a,b , Klf6 expression was indeed found to be highly upregulated with aging in ECs ( Pecam1 + ). Heterochronic parabiosis reversed this change in the old parabionts, bringing its transcriptional levels of Klf6 close to those seen in young ECs (Fig. 5b ).

a , Representative RNA images of mouse cortices showing Klf6 puncta in Pecam1 + ECs in YX, OY, OO and OX mice. Scale bars, 20 µm. b , Violin and boxplot representation of RNA quantification ( n  = 3 biologically independent animals) by two-tailed Welch’s t -test with no multiple comparison adjustment for significance. P values for OY–YX: 0.473 (95% confidence interval (CI) −1.588, 0.737); OY–OO: 5.328 × 10 −19 (95% CI −10.260, −6.620), OY–OX: 4.900 × 10 −28 (95% CI −8.881, −6.244); OO–OX: 0.349 (95% CI −0.960, 2.714). Nonsignificant (NS) P  > 0.05, * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001. Boxplot minimum is the smallest value within 1.5× the IQR below the 25th percentile and maximum is the largest value within 1.5× the IQR above the 75th percentile. Boxplot center is the 50th percentile (median) and box bounds are the 25th and 75th percentiles. Outliers are >1.5× and <3× the IQR. c , Representative RNA images of mouse cortices showing Hspa1a puncta in Pecam1 + ECs in YX, OY, OO and OX mice. Scale bars, 20 µm. d , Violin and boxplot representation of RNA quantification ( n  = 4 biologically independent animals) by two-tailed Welch’s t -test with no multiple comparison adjustment for significance. P values for OY–YX: 3.535 × 10 −16 (95% CI 2.727, 4.373); OY–OO 0.008 (95% CI −2.603, −0.385); OY–OX: 0.006 (95% CI −2.177, −0.372); OO–OX: 0.686 (95% CI −0.847, 1.286). NS P  > 0.05, * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001, **** P  ≤ 0.0001. Boxplot minimum is the smallest value within 1.5× the IQR below the 25th percentile and maximum is the largest value within 1.5× the IQR above the 75th percentile. Boxplot center is the 50th percentile (median) and box bounds are the 25th and 75th percentiles. Outliers are >1.5× and <3× the IQR.

Source data

Taken together, our computational analysis identified numerous aging-related genes across multiple cell types, including ECs, with expression restored in old mice and/or disrupted in young mice after heterochronic parabiosis, suggesting their potential involvement in the rejuvenation and/or aging acceleration process, respectively. These data are again consistent with the notion that parabiosis is likely to act in part by regulating processes important to vascular structure and health.

Heterochronic parabiosis reverses aging-induced pathways

We next performed an implementation of gene set enrichment analysis (GSEA) 58 to reveal biological processes and molecular pathways associated with aging, RJV and AGA. Composite ranks were calculated for each gene based on FDR and log(FC). As with our previous study on old and young animals 37 , the ranking metric used yielded many significant terms for most cell types (Supplementary Table 17 ). Specifically, in the RJV model, the most gene sets were observed in oligodendrocyte precursor cells (OPCs), OLGs, olfactory ensheathing glia (OEGs), ASCs, GABA, GLUT, NendCs, ECs and MGs, whereas, in the AGA model, only OPCs, OLGs and PCs yielded significant terms. This further reinforced our observation that heterochronic parabiosis induces stronger aging-related gene signature changes in the old parabionts.

Among the key ontologies that were downregulated in aging and upregulated in RJV across multiple cell types were pathways related to mitochondrial activity, such as gene ontology (GO) oxidative phosphorylation, and GO electron transport chain, as well as oxidative stress homeostasis and metabolism pathways, such as GO detoxification, GO generation of precursor metabolites and energy and reactome metabolism of amino acids, and derivatives followed this trend (Fig. 4f and Supplementary Table 17 ). RJV-downregulated pathways such as GO DNA conformation change and GO peptidyl lysine methylation further demonstrated that the epigenetic machinery is functionally perturbed in various cell populations with parabiosis (Fig. 4f and Supplementary Table 17 ).

In addition to the above ontologies, ECs displayed a clear pattern of normalized enrichment score (NES) sign reversal between aging and RJV, corresponding to our DGE profiling. Changes in mitochondrial and metabolic pathways were found in RJV. Processes that were downregulated include inflammatory pathways such as hallmark tumor necrosis factor α (TNFA) signaling via nuclear factor-κB (NF-κB), and apoptosis and the senescence-associated hallmark P53 pathway. Likewise, proteostasis-associated pathways, such as Reactome HSF1 activation and GO de novo protein folding, were found to be upregulated with aging and downregulated in RJV (Fig. 4f and Supplementary Table 17 ). Collectively, these data suggested that heterochronic parabiosis changes the metabolic profile, improves proteostatic machinery and reduces aging-associated apoptosis or senescence to improve EC function, consistent with recent findings in aortic ECs 59 .

We then focused on further exploring the decline of proteostasis, an acknowledged hallmark of aging, in ECs. Our pathway analysis revealed upregulation of several stress-inducible pathways in ECs that are presumably activated in response to misfolded protein accumulation in aging and suppressed in RJV (Fig. 4f and Supplementary Table 17 ). For example, we examined the gene expression levels of Hspa1a , which encodes a stress-inducible heat shock protein. At the DGE level, as mentioned above, Hspa1a was found to be upregulated in aging and downregulated in RJV in ECs (Supplementary Tables 3 and 4 ). To verify this change, we performed RNA in situ hybridization. As shown in Fig. 5c,d , we detected a significant decline in the number Hspa1a puncta expressed by ECs in heterochronic parabionts.

Parabiosis activates global remodeling of GRNs

In an attempt to identify key regulatory TFs involved in RJV or AGA independent of changes in their own expression levels, we utilized the SCENIC approach to detect gene regulatory networks (GRNs) comprising TFs and their downstream effector genes 60 . Each animal type was profiled individually to identify putatively active GRNs, stratified by cell class. The young unpaired and young isochronic animals had fewer cell types with higher frequencies (>100) of higher (>1) GRN activity scores than young heterochronic animals (Fig. 6a ). Vascular smooth muscle cells (VSMCs), a recently identified perivascular-like cell type of the brain vasculature 39 , frequently had high activity across all animal types (Fig. 6a ). In old heterochronic animals, vascular leptomeningeal cells (VLMCs), NEUT, ASCs and DOPA had higher frequencies of high (>1) GRN activity scores than old unpaired animals (Fig. 6a and Supplementary Table 18 ). These putative changes in GRN activity may suggest an increase in global remodeling in the old and young heterochronic mice to reflect RJV and AGA, respectively.

a , Rose diagrams of GRN scores per cell type across YX, YY, YO, OX, OO and OY. The color scale from blue to magenta reflects the degree of GRN activity ( Methods ). b , Difference heatmap of active GRN TFs corresponding to RJV/AGA log(FC) change sign. Magnitude is the absolute magnitude of the difference, and direction is positive for upregulation in RJV (magenta) and negative for upregulation in AGA (blue). c , Heatmap of EC-active GRN TFs plotted by RJV and AGA log(FC), with upregulation magenta, downregulation blue, clustered with Euclidean distance, average linkage. d , e , Venn diagrams of EC RJV ( d ) and AGA ( e ) animal frameworks’ active GRN TFs. The arrows point to those TFs in common between OY and YX and YO and OX, respectively.

RJV via parabiosis has a distinct transcriptional landscape that follows the opposite direction of expression in AGA. The signed absolute difference in log(FC) identifies TFs that are bidirectional between RJV and AGA, meaning that they have increased expression in RJV and downregulation in AGA, or vice versa. DOPA demonstrated the most extreme instances of TFs that change direction, followed by VLMCs, then ECs (Fig. 6b and Supplementary Table 14 ). Considering that ECs displayed many bidirectional TFs, we further explored their entire transcriptional landscape. Of the EC TFs in old and young heterochronic brains, most demonstrated opposite regulation in RJV and AGA (Fig. 6c ). TFs in common between old heterochronic and young unpaired ECs that reflect RJV included essential regulators of EC function such as Tbx3 (refs. 61 , 62 ) and Foxo4 (ref. 63 ), a member of the FOXO family of TFs that are components of a fundamental aging regulatory pathway , Patz1 , which has been implicated in regulating p53 levels and senescence in ECs 64 , and Arnt , which participates in aryl hydrocarbon receptor signaling and is involved in several aspects of vascular biology 65 (Fig. 6d ). Likewise, young heterochronic and old unpaired ECs that reflect the AGA construct shared the hypoxia response genes Atf1 , Atf2 and Hif1a , indicating a stressed EC profile, as well as Atf4 which has been implicated in angiogenesis 66 (Fig. 6e ).

Parabiosis alters intercellular communication networks

Altered intercellular communication is one of the hallmarks of aging 1 . Although many studies have examined the actions of blood-borne factors on CNS cells, few have looked at factors secreted by the CNS cells themselves and how they are modified by aging. The importance of such secreted factors has been shown to be dysregulated in inflammation and degeneration 67 , 68 , 69 , 70 . To analyze changes in intercellular communication within the brain we used CellChat 71 . We first measured the total number of interactions for each animal type to elucidate the number of cell–cell communication connections (Fig. 7a and Supplementary Table 19 ). We found that the old heterochronic parabionts exhibited fewer putative connections than the old unpaired and old isochronic animals, whereas the young heterochronic parabionts exhibited more putative connections than in the young unpaired and young isochronic animals (Fig. 7a ). In RJV, we discovered connections triggered by VLMCs and ABCs, both of which act physiologically as barrier cells. For example, we found that VLMCs and ABCs potentially signaled to ASCs, ECs and MGs, whereas they received signals from neural stem cells (NSCs), neuronal-restricted precursors (NRPs) and VSMCs (Fig. 7b and Supplementary Table 19 ). Conversely, in AGA, more signaling was triggered by CPCs. Specifically, CPCs signaled to OPCs and various neuronal cell types (DOPA, GLUT, cholinergic neurons (CHOL)), whereas they received signals from VSMCs and GLUT. In this paradigm, we also observed more signaling triggered by EPCs, which are cells that also act physiologically as barriers because they form the epithelial lining of the ventricles (Fig. 7b and Supplementary Table 19 ). Taken together, this computational analysis highlighted the prominent roles for various barrier cells in the brain parenchyma, as well as CPCs, in processes accompanying or mediating the effects of parabiosis.

a , Summarization network graphs of the number of ligand–receptor interactions between cell types in YX, YY, YO, OX, OO and OY mice. Node size is proportional to cell population size. Edge width and transparency of color are proportional to the number of all edges between a set of nodes. b , Chord diagrams representing the informatically predicted unique source:target:receptor:ligand pairings identified only in the rejuvenation model of OY and YX (Venn diagram inset, left panel) or the aging acceleration model of YO and OX (Venn diagram inset, right panel). c , For all identified EC receptors, edgeR DGE QLF test metrics are shown for the aging, RJV and AGA paradigms. Node size is inversely proportional to the Benjamini–Hochberg-adjusted P value for multiple comparisons and node color is scaled by intensity of log(FC) from blue (negative, downregulation) to magenta (positive, upregulation).

Following our previously published approach 37 , we pursued ligand–receptor interactions determined by encapsulating FDR and log(FC) as the metric for putative association. This computational analysis showed that there are certain aging-induced changes in cell–cell communication that were reversed in the old heterochronic parabionts and/or potentiated in the young heterochronic parabionts. As an example, we detailed interactions involving EC ligands, which are known to secrete signals that regulate neurogenesis 72 in NRPs and, recently, were reported to be implicated in signaling networks with OLG lineage cells 73 . We delineated a clear pattern of connection direction reversal between aging and RJV for most ligands, as well as between RJV and AGA, with AGA exhibiting a direction similar to aging (Extended Data Fig. 10 ). Specifically, we identified several secreted factors that could mediate these intercellular relationships, including cytokines/inflammatory mediators such as CXCL12 (ref. 74 ) and growth factors such as brain-derived neurotrophic factor 75 . Another way of investigating the data is to look for aging/RJV-dependent changes in the levels of cellular receptors. We applied this analysis to ECs, known to be highly influenced by factors found in blood and recognized to be subject to interactions with PCs and ASCs. The genes encoding EC receptors found in the cell–cell communication analysis are shown in Fig. 7c , demonstrating the log(FC) reversal between aging and RJV, and same log(FC) direction between aging and AGA, in many receptors.

Our computational analysis identified numerous cell–cell communication networks that are perturbed during the aging process and modified on heterochronic parabiosis, and highlighted the significance of ECs as a potential target for therapeutics, because its intercellular interactions are affected by both aging and heterochronic parabiosis.

Heterochronic parabiosis regulates the senescence state

The effects of parabiosis on cellular senescence are beginning to be recognized 76 , 77 . To explore this in more detail, we performed an implementation of GSEA with a reference gene set of literature-defined, senescence-associated genes (Supplementary Table 20 ) 76 , 78 against the preranked aging, RJV and AGA gene sets to determine functional enrichment of senescence in these paradigms, and their directions. Across 20 of the 29 examined cell populations, we observed directionality reversal between aging and RJV based on NES (Fig. 8a and Supplementary Table 21 ). In the vast majority of cell types, aging was associated with increased senescence consistent with previous studies 79 , whereas RJV resulted in reduced senescence. The NES direction was the same in aging and AGA for 20 cell types, indicative of recapitulation of the aging process in AGA (Fig. 8a and Supplementary Table 21 ). To confirm the effects of parabiosis on senescence, we performed RNA in situ hybridization to evaluate the expression levels of Cdkn1a , a well-known senescence-associated gene. Specifically, we observed a significant decrease in Cdkn1a expression in ECs ( Pecam1 + ) in the old heterochronic parabiotic brains compared with the old isochronic and old brains (Fig. 8b-c ).

a , Dot-plot representation of senescence-associated marker genes curated from the literature 76 , 78 , permuted against each cell type in aging, RJV and AGA with fast GSEA. The inverse log 10 ( P adj ) values for multiple comparisons (Benjamini–Hochberg) reflect the size of the dot and NESs reflect color from blue (negative enrichment) to magenta (positive enrichment). b , Representative RNA in situ images of mouse cortices showing Cdkn1a puncta in Pecam1 + ECs in YX, OY, OO and OX mice. Scale bars, 20 µm. c , Violin and boxplot representation of RNA quantification ( n  = 6 biologically independent animals) by two-tailed Welch’s t -test with no multiple comparison adjustment for significance. P values: OY–YX: 0.810 (95% CI −0.930, 1.190); OY–OO: 0.0484 (95% CI −1.670, −0.006); OY–OX: 0.0000874 (95% CI −3.063, −1.025); and OO–OX: 0.0251 (95% CI −2.260, −0.151). NS P  > 0.05, * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001 and **** P  < 0.0001. Boxplot minimum is the smallest value within 1.5× the IQR below the 25th percentile and maximum is the largest value within 1.5× the IQR above the 75th percentile. Boxplot center is the 50th percentile (median) and box bounds are the 25th and 75th percentiles. Outliers are >1.5× and <3× the IQR.

Our computational analysis indicates that heterochronic parabiosis modulates the senescence signature in multiple brain cell types and suggests the possibility that the cellular senescence regulation itself may contribute, at least partly, to the positive and negative effects of parabiosis in the brains of old and young parabionts, respectively.

Over the past few years, many studies have demonstrated that aging is a highly dynamic and malleable process, with several types of treatment reported to rejuvenate tissues and organs and, in turn, extend organismal healthspan and lifespan 80 , 81 . Among these interventions, heterochronic parabiosis appears to be one of the most effective. Despite this, the mechanism(s) by which parabiosis acts at the cellular and molecular levels to improve tissue function remains elusive. In the present study, we present a comprehensive single-cell survey of the gene expression changes that occur in the aged and young mouse brain after heterochronic parabiosis as a step toward understanding how parabiosis improves brain function.

We first assessed the cellular complexity of the parabiotic brains and showed that cell identity and composition were largely retained. Thus, functional changes associated with parabiosis are probably due, at least substantially, to changes in the gene expression profiles of CNS cells. A limitation of the present study is that, as tissue dissociation is an inherent part of the single-cell sequencing workflow, this might have resulted in nonuniform sampling problems 82 . Therefore, we cannot completely rule out shifts in the numbers of certain cell populations 22 , 25 . Future studies employing single-nuclei sequencing approaches 83 and spatially resolved transcriptomics 84 , 85 may shed more light on this matter.

We further explored the gene expression changes associated with aging and heterochronic parabiosis in all identified brain cell types to discover genes and pathways that are implicated in the rejuvenation and aging acceleration. To this end, we identified the primary cell types exhibiting these changes and revealed the common and cell-type-specific, aging-induced signatures and transcriptional programs that were rescued after parabiosis in the old brains and/or disrupted in the young brains. As corroborated by recent reports, our data provide evidence that heterochronic parabiosis effectively modulates multiple manifestations of canonical aging hallmarks, including altered intercellular communication 69 , loss of proteostasis 86 , defects in mitochondrial dysfunction 59 , 87 , 88 , 89 and cellular senescence 76 , 77 , 90 , 91 , 92 .

Our analyses suggest that the modulation of the aging process is mediated by reprogramming of the associated transcriptional signatures across multiple brain cell types. In support of this notion, aging-induced changes in the epigenetic status of the aged mouse brain 93 , as well as of the aged liver and blood tissues 94 , were recently reported to be reversed after heterochronic parabiosis. The reprogramming of the transcriptome presented in the present study is somewhat similar to the reprogramming recently reported to take place in rats subjected to caloric restriction 95 . These findings raise the exciting possibility that both interventions could promote tissue rejuvenation by mitigating the appearance of similar aging-associated epigenetic alterations, and consequently their induced transcriptional changes. Thus, future studies investigating the exact epigenetic regulators and mechanisms that are responsible for these types of changes will be of high importance 81 , 96 .

Parabiosis is a complex process and stimulates a number of changes—positive and negative—in both parabionts 97 . It is interesting that it can be argued that the positive effects of young blood factor exposure can overcome not only aging-driven changes in gene expression, but also processes such as systemic inflammation and stress stimulated by the parabiosis surgery itself 14 , 98 . Our data support the view that parabiosis also exerts effects independent of the aging process, an aspect incorporated into our analysis. Moreover, parabiosis-driven effects may be mediated through means other than reversing the effects of normal aging. In recognition of this possibility, we designated the DGEs from the chained pairwise comparisons that have no overlap with aging DGEs (Supplementary Tables 12 and 13 ).

One of the primary discoveries of our study is that the EC transcriptome is dynamically regulated by both aging and heterochronic parabiosis. We found that ECs, when compared with other brain cell types, exhibited one of the highest fractions of aging-related genes that were rescued after heterochronic parabiosis in the old brain and, similarly, the highest fraction of aging-related genes that were disrupted after heterochronic parabiosis in the young brain. This finding supports our previous research that the vasculature is strongly affected by aging and disease and is capable of regrowth after heterochronic parabiosis 25 or systemic GDF11 treatment 26 . We observed that a subset of ECs was classified as mitogenic, expressing high levels of Lrg1 and Lcn2 (Fig. 2f,g ), and it is reasonable to speculate that the growth of these cells, the growth of which is probably prevented or suspended by the inflammatory environment of the aged brain, may be among the cell populations that respond to these interventions. Although proteostasis in brain ECs has not been thoroughly investigated, they are apparently long-lived cells 30 , 99 and, like neurons, might therefore accumulate protein aggregates with age 100 , potentially compromising their function. As previously shown, ECs become senescent with age 101 , 102 , but parabiosis may reverse that phenotype as well. Taken together, these findings underline the strong susceptibility and malleability of ECs, which are directly exposed to secreted factors in both brain parenchyma and blood, to adapt to changes in their microenvironment, which is consistent with pervious observations from our lab 25 , 26 and others 30 , 43 , 59 , 69 . Therefore, ECs, despite comprising <5% of the total number of brain cells 103 , are a promising and accessible target for the treatment of aging and its associated diseases 104 .

Collectively, our computational matrices and web portals provide unprecedented data that can be further explored to form working hypotheses for future studies. Similar to other recent reports 89 , 105 , 106 , our study advances fundamental understanding of the mechanisms underlying the aging process and potential interventions that go beyond descriptive studies of cell states. Our work extends current knowledge about the effects of heterochronic parabiosis on the aging process and supports the key role of the brain vasculature in mediating these effects. Effectively, in one, albeit complex procedure, parabiosis improves many of the individual processes, such as mitochondrial dysfunction, proteostasis collapse and cellular senescence, which are usually targeted separately by therapeutic interventions. Based on data in this article, future work aimed at developing therapeutics might focus on individual processes in specific cell types, on regulating levels of regulatory factors secreted from within the CNS or even on regulating the transcriptional states of cells in the brain.

C57BL/6J inbred male mice (JAX no. 000664; CD45.1 − CD45.2 + ) and B6.SJL-Ptprc a Pepc b /BoyJ male mice (JAX no. 002014; CD45.1 + CD45.2 − ) were housed in the Harvard Biolabs Animal Facility under standard conditions. All experimental procedures were approved in advance by the Institutional Animal Care and Use Committee of Harvard University (AEP no. 10-23) and are in compliance with federal and state laws. On the day of sacrifice, young mice were aged 3–4 months (13–15 weeks) and old mice were 20–22 months (80–87 weeks), analogous to human ages 18–20 and 65–70 years, respectively 107 .

Parabiosis surgeries were performed as previously described 25 , 108 with a few modifications. In brief, mice were sedated with controlled isoflurane anesthesia and placed on heating pads to prevent hypothermia. Ophthalmic ointment was applied to prevent dryness of the eyes. The lateral sides of the mice were then carefully shaved and aseptically prepared. Matched skin incisions were made to the shaved sides and the knee and elbow joints were tied together with nonabsorbable sutures (Reli no. SK7772) to facilitate coordinated movement. Surgical wound clips (BD Autoclips 9 mm, no. 427631) and absorbable sutures (Ethicon no. J385H) were then used to join the skins together. On surgery, parabionts were injected subcutaneously with the anti-inflammatory enrofloxacin (5 mg kg −1 ) to prevent infection and the analgesics carprofen (10 mg kg −1 ) and buprenorphine (0.1 mg kg −1 ) to manage pain (single injection every 12 h for up to 3 d post-surgery). In each injection, 0.5 ml of prewarmed 0.9% (w:v) sodium chloride was also provided to prevent dehydration. Pairs were then kept in clean cages and placed on to heating pads for up to 24 post-surgeries to maintain body temperature. Throughout the surgery and postoperative recovery, each pair was monitored continuously and potential signs of pain and distress were recorded, although several physical characteristics were also analyzed, including pair weights. After 15 d post-surgery, pairs were sedated again briefly to allow removal of the surgical wound clips and remnants of the absorbable sutures. Parabiotic pairs were maintained for 4–5 weeks (mean 31.5 d; Extended Data Fig. 2 ) before processing for tissue collection and subsequent analysis.

Blood chimerism analysis

To evaluate blood crosscirculation we followed the same approach as previously described 25 , in which the presence in both partners of heterochronic pairs of blood cells bearing congenic markers from both the aged (CD45.2 + ) and the young (CD45.1 + ) parabionts is demonstrated. Specifically, in the heterochronic pairs we used young mice carrying the congenic marker CD45.1 (JAX no. 002014) and old mice carrying the congenic marker CD45.2 (JAX no. 000664), whereas in the young isochronic pairs we used mice carrying either CD45.1 or CD45.2. For the blood chimerism analysis, spleens were extracted from the mice and single cells were mechanically isolated by passing the spleen through a 40-μm filter. Erythrocytes were lysed with ACK lysis buffer (Thermo Fisher Scientific, catalog no. A1049201) for 3 min on ice and single cells were resuspended in Hanks Balanced Salt Solution (Thermo Fisher Scientific, catalog no. 14025-134) containing 2% fetal bovine serum. Splenocytes were then filtered through a 40-µm filter and stained with an antibody cocktail (Pacific Blue anti-CD45.1 (BioLegend, catalog no. 110722; 1:100 dilution), APC anti-CD45.2 (BioLegend, catalog no. 109814; 1:100 dilution), PE anti-TER-119 (Thermo Fisher Scientific, catalog no. 12-5921-82; 1:200)) and the fixable viability dye Zombie Aqua (BioLegend, catalog no. 423101; 1:300 dilution) for 30 min on ice in the dark. Cells were then washed and fixed in 1% paraformaldehyde before analysis. Cells were gated on physical parameters to identify singlets followed by gating on the Zombie Aqua low TER-119 − population to identify live nonerythroid cells. These cells were subsequently gated as CD45.1 + or CD45.2 + to measure the frequency of donor-derived blood cells from one partner in the spleen of the other partner. Flow cytometry analysis was performed using a BD LSR II flow cytometer (BD Biosciences) and data were analyzed with FlowJo software (v.10). We found that the partner-derived cells represented 30–50% of splenocytes (mean 41.3%; Extended Data Fig. 1 ), consistent with successful establishment of parabiotic crosscirculation 38 . This analysis could not be applied to old isochronic mice, because old mice carrying the congenic marker CD45.1 were not available for purchase; however, the crosscirculation in old isochronic mice has been well characterized previously 23 .

Brain tissue dissociation

Brain tissue collection was performed at the same time of day (9–10am), processing one pair of mice per day, thus limiting circadian variation 109 . Brain tissue dissociation was performed as previously described 37 . Briefly, mice were CO 2 anesthetized and then rapidly decapitated. Brains were extracted and hindbrain regions removed. The remaining tissue was mechanically and enzymatically dissociated into single cells and kept on ice for no longer than 1 h until further processing.

For the scRNA-seq experiments, 8 YX, 9 YO, 9 YY, 8 OX, 11 OY and 12 OO brains were analyzed, with two animals (one pair) sacrificed per day as mentioned above. Briefly, after dissociation, cells were diluted in ice-cold phosphate-buffered saline (PBS) containing 0.4% bovine serum albumin at a density of 1,000 cells µl −1 . For every sample, 17,400 cells were loaded into a Chromium Single Cell 3′ Chip (10x Genomics) and processed following the manufacturer’s instructions. ScRNA-seq libraries were prepared using the Chromium v.2 Single Cell 3′ Library and Gel Bead kit v.2 and i7 Mutiplex kit (10x Genomics). Libraries were pooled based on their molar concentrations. Pooled libraries were then loaded at 2.07 pM and sequenced on a NextSeq 500 instrument (Illumina) with 26 bases for read1, 57 bases for read2 and 8 bases for Index1. Cell Ranger (v.2.0) (10x Genomics) was used to perform sample de-multiplexing, barcode processing and single-cell gene unique molecular identifier (UMI) counting, whereas a digital expression matrix was obtained for each experiment with default parameters, mapped to the 10x reference for mm10, v.1.2.0. After the initial sequencing, the samples in each pool were re-pooled based on the actual number of cells detected by Cell Ranger (Extended Data Fig. 2a ), aiming to sequence each sample to a similar depth (number of reads per cell) (mean 43,107 reads per cell; Extended Data Fig. 3c ). Multiple NextSeq runs were conducted to achieve >70% sequencing saturation as determined again by Cell Ranger (median: 75%).

Raw data processing and quality control for cell inclusion

Basic processing and visualization of the scRNA-seq data were performed using the Seurat package (v.3.2.1.9002) 110 in R (v.3.6.1). The initial dataset contained 158,767 cells with data for 21,876 genes (Extended Data Fig. 3 ) The average number of UMI (nCount_RNA) and nonzero genes (nFeature_RNA) are 2828.298 and 1206.153, respectively. The data were log(normalized) and scaled to 10,000 transcripts per cell. Variable genes were identified with FindVariableFeatures() function with the following parameters to set minimum and maximum average expressions and minimum dispersion: mean.cutoff(0.00125, 3), dispersion.cutoff(1,Inf). Next, principal component analysis (PCA) was carried out and the top 50 PCs were stored. Clusters were identified with FindNeighbors() by constructing a K-nearest neighbor (KNN) graph and clustered with the Louvain algorithm with FindClusters() at resolution 2, represented by Uniform Manifold Approximation and Projection (UMAP). All clusters with only one cell were removed and clusters with >8% mitochondrial genes, clusters with min nCount_RNA <1,000 and clusters with min nFeature_RNA 500 were flagged for exclusion, resulting in 80 initial clusters. Animals with low average number of genes >0 (<700), percentage mitochondria >1.5 and not having cell contribution to each cluster were assessed for exclusion. In total, five isochronic old and one isochronic young were removed from the dataset (to retain eight YX, seven YY, nine YO, eight OX, seven OO and eleven OY animals), and the above clustering steps were performed at resolution 2. For quality control (QC) filtering, we selectively removed clusters with minimum percentage mitochondria 0, maximum percentage mitochondria 5%, min_nFeature_RNA 250, max_nFeature_RNA 6000, min_nCount_RNA 200, max_nCount_RNA 30000, min_cells=5. After the second round of QC, we retained 130,889 cells and 20,905 genes. The average nCount_RNA, nonzero genes, percentage mitochondrial RNA and percentage ribosomal RNA were 2736.187, 1368.007, 1.149 and 5.135, respectively. We re-clustered at resolution 2 to identify 69 clusters. The final preprocessing step was to remove probable doublet artifacts arising from the cocapture of multiple cells in one droplet. After an initial round of cluster identity determination as assessed in the next section, we employed a doublet-finding technique by searching for the top differential markers of each identified cluster/subcluster with the FindMarkers() function, and marked doublets/multiplets as any cluster in which >40% of its cells express seven of the top ten genes specific to an initially identified cell type and any other outside of the class of the cell type with which it is associated. These clusters were removed from the downstream analysis and clustering was again performed at resolution 2, representing 105,329 cells with similar retention to other studies 39 , 111 and 69 clusters across 20,905 genes. We examined the UMAP space and all clusters are represented by all batches, so no further correction was warranted (Extended Data Fig. 4a,b ).

Determination of cell-type identity

We used multiple cell-specific/enriched gene markers that have been previously described in the literature to assist in determining cell-type identity 37 . We identified 31 major cell types with distinct expression profiles: OPCs, OLGs, OEGs, NSCs, astrocyte-restricted precursors (ARPs), ASCs, EPCs, hypendymal cells (HypEPCs), tanycytes (TNCs), CPCs, NRPs, immature neurons (ImmNs), GABA, DOPA, GLUT, CHOL, NendCs 37 , ECs, PCs, VSMCs, hemoglobin-expressing vascular cells (Hb-VCs) 37 , VLMCs, ABCs, MGs, monocytes (MNCs), macrophages (MACs), dendritic cells (DCs), NEUT, T cells (T_cell), natural killer (NK) cells and B cells (B_cell) (Fig. 1c and Supplementary Table 1 ).

We then arranged all the identified cell types based on their expression profile, lineage, function and topology into five classes of cells (OLG lineage and OEGs, astroependymal cells and NSCs, neuronal lineage, vasculature cells and immune cells). For each group, we re-clustered the subcategorized cell types using top 50 PCs at resolution 5. The annotation of the subclusters was performed similar to the identification of the main cell clusters.

DGE analysis

After initial QC preprocessing and determination of cellular identities, we utilized the muscat package (v.1.0.0) in R (v.3.6.1) to perform pseudobulk DGE analysis with edgeR’s quasi-likelihood F (QLF) test 44 , 45 , 46 . Seurat objects were exported to SingleCellExperiments and reads were collapsed per animal to ‘sum’ based on ‘counts’. The ‘rejuvenation framework’ RJV follows the design contrast (OY–OX)–(OO–YX), assigning in the design matrix: OY: 1, OO: −1; OX: −1; YX: 1. The ‘aging acceleration framework’ AGA follows the design contrast (YO–YX)–(YY–OX), assigning in the design matrix: YO: 1; YY: −1; YX: −1; OX: 1. Pairwise comparisons for OXvYX, OYvOX, OYvOO, OOvOX, YOvYX, YOvYY and YYvYX were also computed. Via muscat, edgeR generates a log(FC), log(counts per min), F , p_val ( P value), p_adj.loc and p_adj.glb. We used the Benjamini–Hochberg-adjusted q value p_val.loc in all downstream thresholding. Our ability to establish a baseline level of transcription is reliant on the number of cells measured and thus larger clusters’ variation can be more adequately modeled. HypEPCs and TNCs did not contain enough cells over multiple animals to successfully derive statistics. For all mouse types, raw normalized transcript per million (TPM) values were calculated and percentage of expression per animal type. For heatmap representations, the log 2 ( z -score) of each animal’s TPM in the rejuvenation or aging acceleration process was calculated gene-wise (by row).

Pathway analysis

GSEA was performed with the fgsea R package (v.1.12.0) 112 . Using the protocol previously implemented 113 , for each cell population and DGE comparison, genes were ranked by multiplying −log 10 (p_val) with the sign of the log(FC) and converted to Homo sapiens orthologs using biomaRt (v.2.46.2). From MSigDB, we used five gene sets: Hallmark pathways, GO biological process, Kyoto Encyclopedia of Genes and Genomes (KEGG), BioCarta and Reactome. In fgsea, 1,000 permutations were performed with minimum gene set size of 15 and maximum 500. Gene sets with FDR ≤ 0.25 were considered significantly enriched. Term annotations and grouping of those overrepresenting the same pathway were derived from Cytoscape software (v.3.5.1) and the AutoAnnotate app (v.1.2) as previously described 37 . The NES directionality was used to collate cell-type pathways per DGE comparison. Dot-plot representations are a composite of FDR and NES.

We employed SCENIC to assess GRNs and score their activity, using the R implementation (v.1.1.2-2) 60 . Each animal type was analyzed with respect to each lineage. Briefly, we used GENIE3 to identify genes that are coexpressed with TFs. Then, RCisTarget prunes these coexpression modules to create GRNs (regulons). The direct targets of each TF are found using cis -regulatory motif analysis. AUCell scores each regulon’s activity, binarized to on/off at threshold 0.7. The regulon activity per animal type per lineage is graphed with rose diagram histograms. Regulons with TFs that change log(FC) direction between RJV and AGA are identified and those regulons changing direction in at least eight cell clusters are further investigated. The magnitude of the log(FC) difference between RJV and AGA can take the form of the absolute difference, with the sign of the difference positive if rejuvenation is >0 and aging acceleration is <0, negative if aging acceleration is >0, and rejuvenation is <0 or zeroed out if they are in the same direction. Regulon matrices of each animal type across all clusters, along with row-wise regulon count, are reported.

Cell–cell communication

Cell–cell communication between cell types per animal type was assessed using the CellChat tool 71 . The number of interaction graphs per animal were thresholded at interactions reaching P  ≤ 0.05 and graphed with netVisual_circle. Rejuvenation-associated construct graphs were the subset of unique receptor:ligand:source:target combinations of interactions occurring only in OY and YX, and aging-associated construct graphs were the same combination occurring only in YO and OX, graphed via the circlize package (0.4.13.1001) 114 . EC receptors in all six animal types were collapsed into a master list, with DGE P adj /log(FC) graphed via ggplot2. EC receptor graphs per animal type were constructed via CellChat netVisual_aggregate.

Cell–cell communication per comparison was also conducted as previously described 37 using the CCInx package (0.5.1). Per-comparison plots were generated between ligand–receptor pairs using the CCInx tool.

Cellular senescence analysis

Cellular senescence was investigated using functional enrichment on preranked genes against known senescence marker genes as described in the literature 76 , 78 . Briefly, the preranked (−log 10 (p _ val)) multiplied by the sign of the log(FC) aging, RJV and AGA gene lists were permuted 1,000× against the gene set using the fgsea algorithm implemented in ClusterProfiler (v.3.14.3). The NES and P adj (Benjamini–Hochberg) are reported.

EC class assignment

EC ‘zonation’ was assessed through deep learning using the CellAssign framework 115 (v.0.99.21, tensorflow_2.2.0.9000). Gene markers from Zhao et al. 43 and other sources 41 , 42 , 43 were used to define arterial–venous–capillary markers. The learning rate used was 1 × 10 −2 , with a min_delta of 0.25 and 10 runs on a V100 graphics processing unit hosted on the FAS Cannon cluster.

RNA in situ hybridization

RNA in situ hybridization was performed on fresh-frozen brain tissue from at least three mice for each relevant condition (YX, OY, OO, OX). For sample preparation, mice were sacrificed via cervical dislocation and the brains were rapidly extracted and embedded in optimal cutting temperature (OCT; Tissue Tek) on dry ice, and subsequently stored at −80 °C until further processing. Brains were divided into 14-μm cryostat sections and RNA in situ hybridizations were carried out using the RNA in situ Multiplex Fluorescent Manual Assay kit (Advanced Cell Diagnostics (ACD)) per the manufacturer’s instructions. Briefly, thawed sections were fixed in 4% paraformaldehyde in PBS and dehydrated in sequential incubations with ethanol, followed by a 30-min Protease IV treatment and washing in 1× PBS. Appropriate combinations of hybridization probes were incubated on tissue for 2 h at 40 °C, followed by four amplification steps. Sections were subsequently stained with DAPI and mounted with Prolong Gold mounting medium (Thermo Fisher Scientific, catalog no. P36930). Brain regions were selected based on areas of high expression levels of assessed examined genes, according to the Allen Brain Atlas 116 . Commercially available and validated probes for Cdkn1a (ACD, catalog no. 408551), Hspa1a (ACD, catalog no. 488351), Klf6 (ACD, catalog no. 426901) and Pecam1 (ACD, catalog no. 316721) were utilized per the manufacturer’s instructions. For each mouse and tissue, three Bregma-matched sections were imaged. Images (four per tissue section) were acquired with a Zeiss LSM 880 Confocal Microscope with identical settings across sections and represented as maximum intensity projections of acquired confocal z -stacks. Analysis was done using a script within CellProfiler software (v.4.2.1), in which Cdkn1a , Hspa1a or Klf6 puncta with a diameter between 1 and 12 pixels located within the perinuclear space (100 pixels of DAPI-positive nuclei) were identified and quantified. Cells with two or more Pecam1 + puncta were designated Pecam1 + ECs. For Hspa1a and Cdkn1a experiments, the EC marker Pecam1 was labeled by fluorophore Atto 647, whereas target probes were labeled by Atto 488 ( Hspa1a ) and Atto 550 ( Cdkn1a ). For Klf6 experiments, Pecam1 was labeled by the fluorophore Alexa Fluor-488 whereas Klf6 was labeled by the fluorophore Atto 550. Lipofuscin granules largely associated with aged brain tissue were avoided utilizing the 1- to 12-pixel cutoff for identifying puncta. For each animal, an unstained tissue was imaged as a negative control and to assess levels of background fluorescence.

Statistics and reproducibility

No statistical methods were used to predetermine sample sizes; our samples sizes were determined iteratively. No randomization was performed. Data collection and analysis were not performed blind to the conditions of the experiments. Animals with low average number of genes >0 (<700), percentage mitochondria >1.5 and not having cell contribution to each cluster were assessed for exclusion. In total, five OO and one YY were removed from the dataset. Further, clusters of poor quality, over percentage mitochondria = 5%, under nFeature_RNA = 250, over nFeature_RNA = 6,000, under nCount_RNA = 200, over nCount_RNA = 30,000, fewer than 5 cells were removed (see Methods for full details). All statistical analyses were performed with R (v.3.6.1). To generate P values for cell counts, ANOVA was conducted between animal types per cell type (rstatix 0.6.0). For validation of gene expression changes by RNA in situ hybridization, two-tailed Welch’s t -test was conducted as indicated (rstatix 0.6.0). Data distribution was assumed to be normal with equal variance, but this was not formally tested.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

Raw data are available on the Gene Expression Omnibus under accession no. GSE222510 . Data exploration of this scRNA-seq study is currently available at https://rubinlab.connect.hms.harvard.edu/parabiosis and on the Broad Single Cell Portal at https://singlecell.broadinstitute.org/single_cell/study/SCP2011/aging-mouse-brain-parabiosis . Source data are provided with this paper.

Code availability

Code is available upon request: https://github.com/kmh005/rubin_parabiosis .

Lopez-Otin, C., Blasco, M. A., Partridge, L., Serrano, M. & Kroemer, G. Hallmarks of aging: an expanding universe. Cell https://doi.org/10.1016/j.cell.2022.11.001 (2022).

Burke, S. N. & Barnes, C. A. Neural plasticity in the ageing brain. Nat. Rev. Neurosci. 7 , 30–40 (2006).

Article   CAS   PubMed   Google Scholar  

Mattson, M. P. & Arumugam, T. V. Hallmarks of brain aging: adaptive and pathological modification by metabolic states. Cell Metab. 27 , 1176–1199 (2018).

Article   CAS   PubMed   PubMed Central   Google Scholar  

Navarro, A., Gomez, C., Lopez-Cepero, J. M. & Boveris, A. Beneficial effects of moderate exercise on mice aging: survival, behavior, oxidative stress, and mitochondrial electron transfer. Am. J. Physiol. Regul. Integr. Comp. Physiol. 286 , R505–R511 (2004).

Brett, J. O. et al. Exercise rejuvenates quiescent skeletal muscle stem cells in old mice through restoration of cyclin D1. Nat. Metab. 2 , 307–317 (2020).

Brandhorst, S. et al. A periodic diet that mimics fasting promotes multi-system regeneration, enhanced cognitive performance, and healthspan. Cell Metab. 22 , 86–99 (2015).

Chang, J. et al. Clearance of senescent cells by ABT263 rejuvenates aged hematopoietic stem cells in mice. Nat. Med. 22 , 78–83 (2016).

Xu, M. et al. Targeting senescent cells enhances adipogenesis and metabolic function in old age. eLife 4 , e12997 (2015).

Article   PubMed   PubMed Central   Google Scholar  

Arriola Apelo, S. I., Pumper, C. P., Baar, E. L., Cummings, N. E. & Lamming, D. W. Intermittent administration of rapamycin extends the life span of female C57BL/6J mice. J. Gerontol. A Biol. Sci. Med. Sci. 71 , 876–881 (2016).

Martin-Montalvo, A. et al. Metformin improves healthspan and lifespan in mice. Nat. Commun. 4 , 2192 (2013).

Article   PubMed   Google Scholar  

Ocampo, A. et al. In vivo amelioration of age-associated hallmarks by partial reprogramming. Cell 167 , 1719–1733.e1712 (2016).

Das, M. M. et al. Young bone marrow transplantation preserves learning and memory in old mice. Commun. Biol . 2 , 73 (2019).

Conboy, I. M. & Rando, T. A. Heterochronic parabiosis for the study of the effects of aging on stem cells and their niches. Cell Cycle 11 , 2260–2267 (2012).

Conboy, M. J., Conboy, I. M. & Rando, T. A. Heterochronic parabiosis: historical perspective and methodological considerations for studies of aging and longevity. Aging Cell 12 , 525–530 (2013).

Conboy, I. M. et al. Rejuvenation of aged progenitor cells by exposure to a young systemic environment. Nature 433 , 760–764 (2005).

Brack, A. S. et al. Increased Wnt signaling during aging alters muscle stem cell fate and increases fibrosis. Science 317 , 807–810 (2007).

Loffredo, F. S. et al. Growth differentiation factor 11 is a circulating factor that reverses age-related cardiac hypertrophy. Cell 153 , 828–839 (2013).

Salpeter, S. J. et al. Systemic regulation of the age-related decline of pancreatic beta-cell replication. Diabetes 62 , 2843–2848 (2013).

Sinha, M. et al. Restoring systemic GDF11 levels reverses age-related dysfunction in mouse skeletal muscle. Science 344 , 649–652 (2014).

Baht, G. S. et al. Exposure to a youthful circulaton rejuvenates bone repair through modulation of beta-catenin. Nat. Commun. 6 , 7131 (2015).

Lu, A. et al. Heterogenetic parabiosis between healthy and dystrophic mice improve the histopathology in muscular dystrophy. Sci. Rep. 10 , 7075 (2020).

Villeda, S. A. et al. The ageing systemic milieu negatively regulates neurogenesis and cognitive function. Nature 477 , 90–94 (2011).

Ruckh, J. M. et al. Rejuvenation of regeneration in the aging central nervous system. Cell Stem Cell 10 , 96–103 (2012).

Villeda, S. A. et al. Young blood reverses age-related impairments in cognitive function and synaptic plasticity in mice. Nat. Med. 20 , 659–663 (2014).

Katsimpardi, L. et al. Vascular and neurogenic rejuvenation of the aging mouse brain by young systemic factors. Science 344 , 630–634 (2014).

Ozek, C., Krolewski, R. C., Buchanan, S. M. & Rubin, L. L. Growth differentiation factor 11 treatment leads to neuronal and vascular improvements in the hippocampus of aged mice. Sci. Rep. 8 , 17293 (2018).

Middeldorp, J. et al. Preclinical assessment of young blood plasma for Alzheimer disease. JAMA Neurol. 73 , 1325–1333 (2016).

Castellano, J. M. et al. Human umbilical cord plasma proteins revitalize hippocampal function in aged mice. Nature 544 , 488–492 (2017).

Gan, K. J. & Sudhof, T. C. Specific factors in blood from young but not old mice directly promote synapse formation and NMDA-receptor recruitment. Proc. Natl Acad. Sci. USA 116 , 12524–12533 (2019).

Chen, M. B. et al. Brain endothelial cells are exquisite sensors of age-related circulatory cues. Cell Rep. 30 , 4418–4432.e4414 (2020).

Pishel, I. et al. Accelerated aging versus rejuvenation of the immune system in heterochronic parabiosis. Rejuvenation Res. 15 , 239–248 (2012).

Rebo, J. et al. A single heterochronic blood exchange reveals rapid inhibition of multiple tissues by old blood. Nat. Commun. 7 , 13363 (2016).

Lehallier, B. et al. Undulating changes in human plasma proteome profiles across the lifespan. Nat. Med. 25 , 1843–1850 (2019).

Khrimian, L. et al. Gpr158 mediates osteocalcin’s regulation of cognition. J. Exp. Med. 214 , 2859–2873 (2017).

Lourenco, M. V. et al. Exercise-linked FNDC5/irisin rescues synaptic plasticity and memory defects in Alzheimer’s models. Nat. Med. 25 , 165–175 (2019).

Sousa-Victor, P. et al. MANF regulates metabolic and immune homeostasis in ageing and protects against liver damage. Nat. Metab. 1 , 276–290 (2019).

Ximerakis, M. et al. Single-cell transcriptomic profiling of the aging mouse brain. Nat. Neurosci. 22 , 1696–1708 (2019).

Wright, D. E., Wagers, A. J., Gulati, A. P., Johnson, F. L. & Weissman, I. L. Physiological migration of hematopoietic stem and progenitor cells. Science 294 , 1933–1936 (2001).

Zeisel, A. et al. Molecular architecture of the mouse nervous system. Cell 174 , 999–1014.e1022 (2018).

Wang, X. et al. LRG1 promotes angiogenesis by modulating endothelial TGF-beta signalling. Nature 499 , 306–311 (2013).

Kalucka, J. et al. Single-cell transcriptome atlas of murine endothelial cells. Cell 180 , 764–779.e720 (2020).

Vanlandewijck, M. et al. A molecular atlas of cell types and zonation in the brain vasculature. Nature 554 , 475–480 (2018).

Zhao, L. et al. Pharmacologically reversible zonation-dependent endothelial cell transcriptomic changes with neurodegenerative disease associations in the aged brain. Nat. Commun. 11 , 4413 (2020).

Robinson, M. D., McCarthy, D. J. & Smyth, G. K. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics 26 , 139–140 (2010).

McCarthy, D. J., Chen, Y. & Smyth, G. K. Differential expression analysis of multifactor RNA-Seq experiments with respect to biological variation. Nucleic Acids Res. 40 , 4288–4297 (2012).

Crowell, H, Germain, P., Soneson, C., Sonrel, A. & Robinson, M. muscat: multi-sample multi-group scRNA-seq data analysis tools. R package v.1.4.0 (2020).

Kharchenko, P. V., Silberstein, L. & Scadden, D. T. Bayesian approach to single-cell differential expression analysis. Nat. Methods 11 , 740–742 (2014).

Yamazaki, Y., Zhao, N., Caulfield, T. R., Liu, C. C. & Bu, G. Apolipoprotein E and Alzheimer disease: pathobiology and targeting strategies. Nat. Rev. Neurol. 15 , 501–518 (2019).

Foster, E. M., Dangla-Valls, A., Lovestone, S., Ribe, E. M. & Buckley, N. J. Clusterin in Alzheimer’s disease: mechanisms, genetics, and lessons from other pathologies. Front. Neurosci. 13 , 164 (2019).

DeMattos, R. B. et al. ApoE and clusterin cooperatively suppress Abeta levels and deposition: evidence that ApoE regulates extracellular Abeta metabolism in vivo. Neuron 41 , 193–202 (2004).

Hohberg, M. et al. Expression of ADAMTS1 in endothelial cells is induced by shear stress and suppressed in sprouting capillaries. J. Cell. Physiol. 226 , 350–361 (2011).

Regard, J. B. et al. Verge: a novel vascular early response gene. J. Neurosci. 24 , 4092–4103 (2004).

Babic, A. M., Kireeva, M. L., Kolesnikova, T. V. & Lau, L. F. CYR61, a product of a growth factor-inducible immediate early gene, promotes angiogenesis and tumor growth. Proc. Natl Acad. Sci. USA 95 , 6355–6360 (1998).

Bellou, S. et al. VEGF autoregulates its proliferative and migratory ERK1/2 and p38 cascades by enhancing the expression of DUSP1 and DUSP5 phosphatases in endothelial cells. Am. J. Physiol. Cell Physiol. 297 , C1477–C1489 (2009).

Othumpangat, S., Noti, J. D., Blachere, F. M. & Beezhold, D. H. Expression of non-structural-1A binding protein in lung epithelial cells is modulated by miRNA-548an on exposure to influenza A virus. Virology 447 , 84–94 (2013).

Kojima, S. et al. Transcriptional activation of urokinase by the Kruppel-like factor Zf9/COPEB activates latent TGF-beta1 in vascular endothelial cells. Blood 95 , 1309–1316 (2000).

CAS   PubMed   Google Scholar  

Botella, L. M. et al. Transcriptional activation of endoglin and transforming growth factor-beta signaling components by cooperative interaction between Sp1 and KLF6: their potential role in the response to vascular injury. Blood 100 , 4001–4010 (2002).

Subramanian, A. et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc. Natl Acad. Sci. USA 102 , 15545–15550 (2005).

Kiss, T. et al. Circulating anti-geronic factors from heterochonic parabionts promote vascular rejuvenation in aged mice: transcriptional footprint of mitochondrial protection, attenuation of oxidative stress, and rescue of endothelial function by young blood. Geroscience 42 , 727–748 (2020).

Aibar, S. et al. SCENIC: single-cell regulatory network inference and clustering. Nat. Methods 14 , 1083–1086 (2017).

Czymai, T. et al. FOXO3 modulates endothelial gene expression and function by classical and alternative mechanisms. J. Biol. Chem. 285 , 10163–10178 (2010).

Krstic, M. et al. Isoform-specific promotion of breast cancer tumorigenicity by TBX3 involves induction of angiogenesis. Lab. Invest. 100 , 400–413 (2020).

Potente, M. et al. Involvement of Foxo transcription factors in angiogenesis and postnatal neovascularization. J. Clin. Invest. 115 , 2382–2392 (2005).

Cho, J. H., Kim, M. J., Kim, K. J. & Kim, J. R. POZ/BTB and AT-hook-containing zinc finger protein 1 (PATZ1) inhibits endothelial cell senescence through a p53 dependent pathway. Cell Death Differ. 19 , 703–712 (2012).

Yim, S. H. et al. Disruption of the Arnt gene in endothelial cells causes hepatic vascular defects and partial embryonic lethality in mice. Hepatology 44 , 550–560 (2006).

Fan, Z. et al. Exercise-induced angiogenesis is dependent on metabolically primed ATF3/4 + endothelial cells. Cell Metab. 33 , 1793–1807.e1799 (2021).

Zhang, G. et al. Hypothalamic programming of systemic ageing involving IKK-beta, NF-kappaB and GnRH. Nature 497 , 211–216 (2013).

Kosmidis, S. et al. RbAp48 protein Is a critical component of GPR158/OCN signaling and ameliorates age-related memory loss. Cell Rep. 25 , 959–973.e956 (2018).

Yousef, H. et al. Aged blood impairs hippocampal neural precursor activity and activates microglia via brain endothelial cell VCAM1. Nat. Med. 25 , 988–1000 (2019).

Dulken, B. W. et al. Single-cell analysis reveals T cell infiltration in old neurogenic niches. Nature 571 , 205–210 (2019).

Jin, S. et al. Inference and analysis of cell-cell communication using CellChat. Nat. Commun. 12 , 1088 (2021).

Shen, Q. et al. Endothelial cells stimulate self-renewal and expand neurogenesis of neural stem cells. Science 304 , 1338–1340 (2004).

Paredes, I. et al. Oligodendrocyte precursor cell specification is regulated by bidirectional neural progenitor-endothelial cell crosstalk. Nat. Neurosci. 24 , 478–488 (2021).

Patel, J. R., McCandless, E. E., Dorsey, D. & Klein, R. S. CXCR4 promotes differentiation of oligodendrocyte progenitors and remyelination. Proc. Natl Acad. Sci. USA 107 , 11062–11067 (2010).

Djalali, S. et al. Effects of brain-derived neurotrophic factor (BDNF) on glial cells and serotonergic neurones during development. J. Neurochem. 92 , 616–627 (2005).

Yousefzadeh, M. J. et al. Heterochronic parabiosis regulates the extent of cellular senescence in multiple tissues. Geroscience 42 , 951–961 (2020).

Mehdipour, M. et al. Plasma dilution improves cognition and attenuates neuroinflammation in old mice. Geroscience https://doi.org/10.1007/s11357-020-00297-8 (2020).

Coppe, J. P., Desprez, P. Y., Krtolica, A. & Campisi, J. The senescence-associated secretory phenotype: the dark side of tumor suppression. Annu. Rev. Pathol. 5 , 99–118 (2010).

Sikora, E. et al. Cellular senescence in brain aging. Front. Aging Neurosci. https://doi.org/10.3389/fnagi.2021.646924 (2021).

Campisi, J. et al. From discoveries in ageing research to therapeutics for healthy ageing. Nature 571 , 183–192 (2019).

Zhang, W., Qu, J., Liu, G. H. & Belmonte, J. C. I. The ageing epigenome and its rejuvenation. Nat. Rev. Mol. Cell Biol. 21 , 137–150 (2020).

Denisenko, E. et al. Systematic assessment of tissue dissociation and storage biases in single-cell and single-nucleus RNA-seq workflows. Genome Biol. 21 , 130 (2020).

Habib, N. et al. Massively parallel single-nucleus RNA-seq with DroNc-seq. Nat. Methods 14 , 955–958 (2017).

Allen, W. E., Blosser, T. R., Sullivan, Z. A., Dulac, C. & Zhuang, X. Molecular and spatial signatures of mouse brain aging at single-cell resolution. Cell https://doi.org/10.1016/j.cell.2022.12.010 (2022).

Stahl, P. L. et al. Visualization and analysis of gene expression in tissue sections by spatial transcriptomics. Science 353 , 78–82 (2016).

Liu, A. et al. Young plasma reverses age-dependent alterations in hepatic function through the restoration of autophagy. Aging Cell https://doi.org/10.1111/acel.12708 (2018).

Morrison, E. J. et al. Parabiosis incompletely reverses aging-induced metabolic changes and oxidant stress in mouse red blood cells. Nutrients https://doi.org/10.3390/nu11061337 (2019).

Gonzalez-Armenta, J. L., Li, N., Lee, R. L., Lu, B. & Molina, A. J. A. Heterochronic parabiosis: old blood induces changes in mitochondrial structure and function of young mice. J. Gerontol. A Biol. Sci. Med. Sci. 76 , 434–439 (2021).

Palovics, R. et al. Molecular hallmarks of heterochronic parabiosis at single-cell resolution. Nature 603 , 309–314 (2022).

Ghosh, A. K., O’Brien, M., Mau, T., Qi, N. & Yung, R. Adipose tissue senescence and inflammation in aging is reversed by the young milieu. J. Gerontol. A Biol. Sci. Med. Sci. 74 , 1709–1715 (2019).

Barinda, A. J. et al. Endothelial progeria induces adipose tissue senescence and impairs insulin sensitivity through senescence associated secretory phenotype. Nat. Commun. 11 , 481 (2020).

Ma, S. et al. Heterochronic parabiosis induces stem cell revitalization and systemic rejuvenation across aged tissues. Cell Stem Cell 29 , 990–1005.e1010 (2022).

Koellhoffer, E. C., Morales-Scheihing, D., d’Aigle, J. & McCullough, L. D. Heterochronic parabiosis reverses the epigenetic imbalance of aged central nervous system. Stroke 48 , AWP122 (2017).

Zhang, B. et al. Multi-omic rejuvenation and lifespan extension upon exposure to youthful circulation. Preprint at bioRxiv https://doi.org/10.1101/2021.11.11.468258 (2021).

Ma, S. et al. Caloric restriction reprograms the single-cell transcriptional landscape of Rattus norvegicus aging. Cell 180 , 984–1001.e1022 (2020).

Gontier, G. et al. Tet2 rescues age-related regenerative decline and enhances cognitive function in the adult mouse brain. Cell Rep. 22 , 1974–1981 (2018).

Eichwald, E. J., Lustgraaf, E. C. & Strainer, M. Genetic factors in parabiosis. J. Natl Cancer Inst. 23 , 1193–1213 (1959).

Castellano, J. M. et al. In vivo assessment of behavioral recovery and circulatory exchange in the peritoneal parabiosis model. Sci. Rep. 6 , 29015 (2016).

Brandes, R. P., Fleming, I. & Busse, R. Endothelial aging. Cardiovasc. Res. 66 , 286–294 (2005).

Fonseca, A. C., Oliveira, C. R., Pereira, C. F. & Cardoso, S. M. Loss of proteostasis induced by amyloid beta peptide in brain endothelial cells. Biochim. Biophys. Acta 1843 , 1150–1161 (2014).

Yamazaki, Y. et al. Vascular cell senescence contributes to blood-brain barrier breakdown. Stroke 47 , 1068–1077 (2016).

Kiss, T. et al. Single-cell RNA sequencing identifies senescent cerebromicrovascular endothelial cells in the aged mouse brain. Geroscience 42 , 429–444 (2020).

von Bartheld, C. S., Bahney, J. & Herculano-Houzel, S. The search for true numbers of neurons and glial cells in the human brain: a review of 150 years of cell counting. J. Comp. Neurol. 524 , 3865–3895 (2016).

Article   Google Scholar  

Lau, S. F., Cao, H., Fu, A. K. Y. & Ip, N. Y. Single-nucleus transcriptome analysis reveals dysregulation of angiogenic endothelial cells and neuroprotective glia in Alzheimer’s disease. Proc. Natl Acad. Sci. USA 117 , 25800–25809 (2020).

Ma, S. et al. Heterochronic parabiosis induces stem cell revitalization and systemic rejuvenation across aged tissues. Cell Stem Cell https://doi.org/10.1016/j.stem.2022.04.017 (2022).

Buckley, M. T. et al. Cell-type-specific aging clocks to quantify aging and rejuvenation in neurogenic regions of the brain. Nat. Aging 3 , 121–137 (2023).

Dutta, S. & Sengupta, P. Men and mice: relating their ages. Life Sci. 152 , 244–248 (2016).

Kamran, P. et al. Parabiosis in mice: a detailed protocol. J. Vis. Exp . https://doi.org/10.3791/50556 (2013).

Zhang, R., Lahens, N. F., Ballance, H. I., Hughes, M. E. & Hogenesch, J. B. A circadian gene expression atlas in mammals: implications for biology and medicine. Proc. Natl Acad. Sci. USA 111 , 16219–16224 (2014).

Stuart, T. et al. Comprehensive integration of single-cell data. Cell 177 , 1888–1902.e1821 (2019).

Hrvatin, S. et al. Single-cell analysis of experience-dependent transcriptomic states in the mouse visual cortex. Nat. Neurosci. 21 , 120–129 (2018).

Korotkevich G, S. V. & Sergushichev A. Fast gene set enrichment analysis. Preprint at bioRxiv https://doi.org/10.1101/060012 (2019).

Reimand, J. et al. Pathway enrichment analysis and visualization of omics data using g:Profiler, GSEA, Cytoscape and EnrichmentMap. Nat. Protoc. 14 , 482–517 (2019).

Gu, Z., Gu, L., Eils, R., Schlesner, M. & Brors, B. circlize implements and enhances circular visualization in R. Bioinformatics 30 , 2811–2812 (2014).

Zhang, A. W. et al. Probabilistic cell-type assignment of single-cell RNA-seq for tumor microenvironment profiling. Nat. Methods 16 , 1007–1015 (2019).

Lein, E. S. et al. Genome-wide atlas of gene expression in the adult mouse brain. Nature 445 , 168–176 (2007).

Download references

Acknowledgements

We thank T. Okino and his team at Ono Pharmaceuticals for fruitful discussions during the progress of this work. We thank K. Pritchett-Corning, F. Rapino, K. Pfaff, B. Mayweather, M. H. C. Florido, S. Ghosh and J. LaLonde for their advice and help in different aspects of the study, and F. Price for illustrations. We also thank the staff members of the Harvard Biolabs Animal Facility, the Harvard Center for Biological Imaging and the Harvard Stem Cell and Regenerative Biology Histology Core for their continuous support and assistance. This work was supported by Ono Pharmaceutical Co., Ltd (L.L.R.), the Stanley Center for Psychiatric Research (L.L.R, J.Z.L.), the Klarman Cell Observatory of the Broad Institute of MIT and Harvard (J.Z.L), Howard Hughes Medical Institute (A.V.), an award from the Glenn Foundation (A.J.W.), grants from the National Institutes of Health (NIH)/National Institute of Neurological Disorders and Stroke (grant no. 1R01NS117407) (L.L.R.), NIH/National Institute on Aging (grant no. 1R01AG072086) (L.L.R.) and the NIH (grant no. T32 DK007529) (J.M.G.), and the Simons Foundation (Collaboration on Plasticity and the Aging Brain) (L.L.R.). The funders had no role in the study design, experiments performed, data collection, data analysis and interpretation, or preparation of the manuscript.

Author information

These authors contributed equally: Methodios Ximerakis, Kristina M. Holton.

Authors and Affiliations

Department of Stem Cell and Regenerative Biology, Harvard University, Cambridge, MA, USA

Methodios Ximerakis, Kristina M. Holton, Richard M. Giadone, Ceren Ozek, Monika Saxena, Samara Santiago, Kavya M. Shah, Jill M. Goldstein, Caterina Gasperini, Ioannis A. Gampierakis, Scott L. Lipnick, Sean M. Buchanan, Amy J. Wagers & Lee L. Rubin

Harvard Stem Cell Institute, Cambridge, MA, USA

Stanley Center for Psychiatric Research, Broad Institute of MIT and Harvard, Cambridge, MA, USA

Methodios Ximerakis, Kristina M. Holton, Xian Adiconis, Scott L. Lipnick, Sean K. Simmons, Joshua Z. Levin & Lee L. Rubin

Klarman Cell Observatory, Broad Institute of MIT and Harvard, Cambridge, MA, USA

Xian Adiconis, Danielle Dionne, Lan Nguyen, Sean K. Simmons, Aviv Regev & Joshua Z. Levin

Joslin Diabetes Center, Boston, MA, USA

Amy J. Wagers

Paul F. Glenn Center for the Biology of Aging, Harvard Medical School, Boston, MA, USA

Howard Hughes Medical Institute, Koch Institute of Integrative Cancer Research, Department of Biology, Massachusetts Institute of Technology, Cambridge, MA, USA

You can also search for this author in PubMed   Google Scholar

Contributions

M.X. and L.L.R. conceived the study. M.X., K.M.H. and L.L.R. designed the study. M.X. performed the parabiosis experiments. M.X., X.A., D.D. and L.N. performed the scRNA-seq experiments. K.M.H., S.L.L. and J.Z.L. processed the scRNA-seq data. K.M.H. developed the computational framework and performed all associated analyses. M.X. and K.M.H. interpreted the data. R.M.G., C.O., M.S., S.S., I.A.G. and C.G. designed and/or performed validation experiments. J.M.G. performed the blood chimerism experiments and analysis. K.M.S. assisted in the development of transcriptional networks. K.M.H. built the online portal. M.X., K.M.H., A.J.W., A.R. and J.Z.L. supervised aspects of the study. L.L.R. directed the study. J.Z.L. and L.L.R. secured funding. M.X. and K.M.H. wrote the original draft of the manuscript. C.O., S.K.S., S.M.B., J.Z.L., A.J.W., A.R. and L.L.R. provided critical feedback and/or edited the manuscript. All authors reviewed the manuscript and approved its submission.

Corresponding authors

Correspondence to Methodios Ximerakis or Lee L. Rubin .

Ethics declarations

Competing interests.

L.L.R. is a founder of Elevian, Rejuveron and Vesalius Therapeutics, a member of their scientific advisory boards and a private equity shareholder. All are interested in formulating approaches intended to treat diseases of the nervous system and other tissues. He is also on the advisory board of Alkahest, a Grifols company, focused on the plasma proteome. None of these companies provided any financial support for the work in this paper. A.J.W. is a scientific advisor for Kate Therapeutics and Frequency Therapeutics, and is a founder of Elevian, Inc. and a member of their scientific advisory board and shareholder. Elevian, Inc. also provides sponsored research to the Wagers lab. A.R. is a founder and equity holder of Celsius Therapeutics, an equity holder in Immunitas Therapeutics and until 31 August 2020 was a SAB member of Syros Pharmaceuticals, Neogene Therapeutics, Asimov and Thermo Fisher Scientific. From 1 August 2020, A.R. has been an employee of Genentech, a member of the Roche Group. M.X. has been an employee of Merck & Co. since August 2020. The remaining authors declare no competing interests.

Peer review

Peer review information.

Nature Aging thanks Bo Peng and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data

Extended data fig. 1 confirmation of blood chimerism..

Representative flow cytometry analysis of CD45.1 and CD45.2 expression markers on splenocytes isolated from ( a ) young (CD45.1 + ) and ( b ) old mice (CD45.2 + ) following heterochronic parabiosis. c) The percentage of donor-derived blood cells from one partner in the spleen of the other partner is depicted by arrows.

Extended Data Fig. 2 Sample metrics.

Profiling of animals and their derived brain cells used for sequencing, before (N = 56) (a,c,e) and after (n = 50) (b,d,f) quality control filtering in which certain animals were omitted (see Methods). Boxplot minimum is the smallest value within 1.5 times the interquartile range below the 25 th percentile, maximum is the largest value within 1.5 times the interquartile range above the 75 th percentile. Boxplot center is the 50 th percentile (median), box bounds are the 25 th and 75 th percentile. Outliers are >1.5 times and < 3 times the interquartile range. a-b . Age of mice in weeks prior to parabiosis surgeries. c-d . Number of days joined across parabiotic pairs. e-f . Number of dissociated cells analyzed per brain across all animal types.

Extended Data Fig. 3 Sequencing metrics.

Violin plots with boxplots showing sequencing metrics of the distribution of animals from all sequenced animal types. Each dot represents one animal. Boxplot minimum is the smallest value within 1.5 times the interquartile range below the 25 th percentile, maximum is the largest value within 1.5 times the interquartile range above the 75 th percentile. Boxplot center is the 50 th percentile (median), box bounds are the 25 th and 75 th percentile. Outliers are >1.5 times and < 3 times the interquartile range. a . Number of cells sequenced by animal. b . Table of total number of animals and cells analyzed. c . Mean number of mapped reads per cell by animal. d . Median number of nCount RNA (UMI) detected per cell by animal. e . Median number of genes detected per cell by animal. f . Percent of sequencing saturation by animal.

Extended Data Fig. 4 Distribution of 50 animals across 5 sequencing batches, with respect to cell clusters, and cell count.

a . UMAP projection of color-coded batches over clusters that passed filtering criteria. b . Frequency of each color-coded batch representation in each cell type. All cell types are represented by cells from all batches, except for HypEPC in batch 5, probably due to its small size. c . Number of detected cells in each cell type.

Extended Data Fig. 5 Primary data analysis.

a. Violin plot and boxplot showing the number of cells analyzed by animal after cell filtering, in which all cells were successfully assigned to a specific cell type. Each dot represents one animal. Boxplot minimum is the smallest value within 1.5 times the interquartile range below the 25 th percentile, maximum is the largest value within 1.5 times the interquartile range above the 75 th percentile. Boxplot center is the 50 th percentile (median), box bounds are the 25 th and 75 th percentile. b-e . Violin plots showing QC metrics, plots in (b, c) showing aggregated data of cells of all brain types, while plots in (d, e) showing individual cell data separated by animal type: (b, d) showing nCount RNA (UMI) per cell type. (c, e) showing nFeature RNA (number of unique genes) detected per cell.

Extended Data Fig. 6 Representation of each animal type’s distribution within each cell type.

a . Dot plot representation of each cell type’s representation by each animal type. Size of the dot is proportional to the number of cells contributed by each animal type within a cell type. b . Dot plot representation of each subpopulation’s representation by each animal type. Size of the dot is proportional to the number of cells contributed by each animal type within a subpopulation.

Extended Data Fig. 7 Cell type composition and cell count from each animal type.

a . Frequency bar plot demonstrating composition of each cell type with respect to animal type. b . Boxplot of raw cell counts with respect to each animal. All animals contribute to all cell types. ANOVA p-values (one-tailed) for pairwise iterations can be found in Supplementary Table 2 . The only comparisons with unadjusted p-values < 0.05 are: OOvOX: DC (p = 0.041), YOvYX DOPA (p = 0.022), YYvYX OEG (p = 0.046), ImmN (p = 0.045), and PC (p = 0.016), Boxplot minimum is the smallest value within 1.5 times the interquartile range below the 25th percentile, maximum is the largest value within 1.5 times the interquartile range above the 75th percentile. Boxplot center is the 50th percentile (median), box bounds are the 25th and 75th percentile. Outliers are >1.5 times and < 3 times the interquartile range.

Extended Data Fig. 8 Animal type distribution and machine learning approaches to explore EC arteriovenous zonation.

a . Animal type cell distribution across EC subclusters. b-c . Probabilistic programming cell class assignment using EC marker genes described by Zhao et al 2020 43 (b) and others 41 , 42 , 43 (c).

Extended Data Fig. 9 Composition of DGEs per cell type between Aging-RJV, and Aging-AGA.

a . Bar graph of each cell type’s total FDR ≤ 0.05 DGEs split by logFC direction. The proportion of DGEs reflecting Aging and RJV is depicted, as well as the fraction of overlapping signatures (intersection in grey). b . Bar graph of each cell type’s total FDR ≤ 0.05 DGEs split by logFC direction. The proportion of DGEs reflecting Aging and AGA is depicted, as well as the fraction of overlapping signatures (intersection in grey).

Extended Data Fig. 10 Intercellular communication networks between EC-OLG and EC-NRP revealed aging-related interactions that were modified by heterochronic parabiosis.

Canonical EC ligands and their cognate receptors in OLG (a) or in NRP (b) are shown in each paradigm (Aging, RJV, AGA). In all panels of ligand-receptor interactions, node color represents the magnitude of the DGE (logFC as estimated by DGE) such that the most significantly up-regulated genes are in magenta, and the downregulated genes are in blue. Node borders indicate multiple testing corrected Benjamini-Hochberg FDR for statistical significance of DGE as calculated by edgeR. Edge color represents the sum of scaled differential expression magnitudes from each contributing node, while width and transparency are determined by the magnitude of the scaled differential expression (see details in the Methods section).

Supplementary information

Reporting summary, supplementary table 1.

List of abbreviations for all cell types and subpopulations and major markers delineating them. The cell type abbreviations and their major markers are listed in the first tab (Cell_Type_Abbreviations_Markers). In the second tab (Cell Type Seurat Markers), the top ten (by avg_log2FC) informatically derived markers from Seurat’s FindAllMarkers() for each cell type are listed, along with Wilcoxon’s rank-sum P value, Bonferonni-adjusted P value (p_val_adj), average log 2 (FC) (avg_log2FC), pct.1 (the percentage of cells in that cell type expressing the gene) and pct.2 (the percentage of cells in all other cell types expressing the gene).

Supplementary Table 2

Metrics of pairwise comparison of cell types by cell number. For each pairwise comparison, per cell type, ANOVA was applied (one tailed) with no multiple comparison correction, with significance designation (NS P  > 0.05, * P  ≤ 0.05, ** P  ≤ 0.01, *** P  ≤ 0.001).

Supplementary Table 3

DGE metrics of RJV, per cell type. edgeR/muscat metrics from the QLF test were computed for each cluster in a comparison, with log(FC), P value, Benjamini–Hochberg P -value adjustment for multiple comparisons (p_adj.loc is per cluster, p_adj.global is with respect to all clusters in class) reported. TPM values (‘tpm’) for each animal type, per cell type, are listed. Percentage expression (‘pct’) of each gene for each animal type, per cell type, is listed.

Supplementary Table 4

DGE metrics of AGA, per cell type. edgeR/muscat metrics from the QLF test were computed for each cluster in a comparison, with log(FC), P value, Benjamini–Hochberg P -value adjustment for multiple comparisons (p_adj.loc is per cluster, p_adj.global is with respect to all clusters in class) reported. TPM values (‘tpm’) for each animal type, per cell type, are listed. Percentage expression (“pct”) of each gene for each animal type, per cell type are listed.

Supplementary Table 5

DGE metrics of aging, per cell type. edgeR/muscat metrics from the QLF test were computed for each cluster in a comparison, with log(FC), P value, Benjamini–Hochberg P -value adjustment for multiple comparisons (p_adj.loc is per cluster, p_adj.global is with respect to all clusters in class) reported. TTPM values (‘tpm’) for each animal type, per cell type, are listed. Percentage expression (‘pct’) of each gene for each animal type, per cell type are listed.

Supplementary Table 6

DGE metrics of OYvOO, per cell type. edgeR/muscat metrics from the QLF test were computed for each cluster in a comparison, with log(FC), P value, Benjamini–Hochberg P -value adjustment for multiple comparisons (p_adj.loc is per cluster, p_adj.global is with respect to all clusters in class) reported. TPM values (‘tpm’) for each animal type, per cell type, are listed. Percentage expression (‘pct’) of each gene for each animal type, per cell type are listed.

Supplementary Table 7

DGE metrics of OYvOX, per cell type. edgeR/muscat metrics from the QLF test were computed for each cluster in a comparison, with logFC, P value, Benjamini–Hochberg P value adjustment for multiple comparisons (p_adj.loc is per cluster, p_adj.global is with respect to all clusters in class) reported. TPM values (“tpm”) for each animal type, per cell type are listed. Percentage expression (‘pct’) of each gene for each animal type, per cell type are listed.

Supplementary Table 8

DGE metrics of OOvOX, per cell type. edgeR/muscat metrics from the QLF test were computed for each cluster in a comparison, with log(FC), P value, Benjamini–Hochberg P -value adjustment for multiple comparisons (p_adj.loc is per cluster, p_adj.global is with respect to all clusters in class) reported. TTPM values (‘tpm’) for each animal type, per cell type, are listed. Percentage expression (‘pct’) of each gene for each animal type, per cell type are listed.

Supplementary Table 9

DGE metrics of YOvYY per cell type. edgeR/muscat metrics from the QLF test were computed for each cluster in a comparison, with log(FC), P value, Benjamini–Hochberg P -value adjustment for multiple comparisons (p_adj.loc is per cluster, p_adj.global is with respect to all clusters in class) reported. TPM values (‘tpm’) for each animal type, per cell type, are listed. Percentage expression (‘pct’) of each gene for each animal type, per cell type are listed.

Supplementary Table 10

DGE metrics of YOvYY, per cell type. edgeR/muscat metrics from the QLF test were computed for each cluster in a comparison, with log(FC), P value, Benjamini–Hochberg P -value adjustment for multiple comparisons (p_adj.loc is per cluster, p_adj.global is with respect to all clusters in class) reported. TPM values (‘tpm’) for each animal type, per cell type, are listed. Percentage expression (‘pct’) of each gene for each animal type, per cell type are listed.

Supplementary Table 11

DGE metrics of YYvYX per cell type. edgeR/muscat metrics from the QLF test were computed for each cluster in a comparison, with log(FC), P value, Benjamini–Hochberg P -value adjustment for multiple comparisons (p_adj.loc is per cluster, p_adj.global is with respect to all clusters in class) reported. TPM values (‘tpm’) for each animal type, per cell type are listed. Percentage expression (‘pct’) of each gene for each animal type, per cell type are listed.

Supplementary Table 12

RJV and aging common DGEs and unique DGEs, per cell type. DGE FDR ≤ 0.05 genes common between RJV and aging, but also those only found in RJV and those only found in aging are listed per cell type.

Supplementary Table 13

AGA and aging common DGEs and unique DGEs, per cell type. DGE FDR ≤ 0.05 genes common between AGA and aging, but also those only found in AGA, and those only found in aging are listed per cell type.

Supplementary Table 14

DGE bidirectional DGEs between RJV and AGA. For RJV and AGA DGE genes FDR ≤ 0.05 per cell type, report the genes that are RJV up (log(FC) > 0) and AGA down (log(FC) <0), or RJV down (log(FC) < 0) and AGA up (log(FC) > 0). Count is the number of times a gene is bidirectionally expressed across cell types.

Supplementary Table 15

Matrices of all DGE FDR ≤ 0.05 log(FC) values across cell types. Per comparison, per gene, clusters where the gene’s significance is FDR ≤ 0.05 have their log(FC) value reported. ‘Up’ column is the sum of clusters with log(FC) > 0, ‘Down’ column is the sum of clusters with log(FC) < 0.

Supplementary Table 16

DGE logFC values across all cell types. Per comparison, per gene, collated log(FC) values across all clusters reporting DGE, with no thresholding. ‘Up’ column is the sum of clusters with log(FC) > 0. ‘Down’ column is the sum of clusters with log(FC) < 0.

Supplementary Table 17

Matrices of all identified significant GSEA terms per comparison across cell types. Fast preranked GSEA (fGSEA) Benjamini–Hochberg-adjusted P value for multiple comparisons ≤0.25 significant terms are collated per comparison across all cell types by NES. Pathway and process metaclasses are described in Methods . ‘Up’ column is the sum of cell types with NES > 0. ‘Down’ column is the sum of cell types with NES < 0.

Supplementary Table 18

SCENIC regulon matrices per animal type. Per animal type, per cell type, SCENIC regulon activity scores are reported. Column ‘counts’ is the sum of cell type that have a regulon score.

Supplementary Table 19

Cell–cell communication networks in RJV, AGA and per animal type . For RJV, the set of unique source:target:receptor:ligand pairs that are found only in OY and YX combined. For AGA, the set of unique source:target:receptor:ligand pairs that are found only in YO and OX combined. YX, YY, YO, OX, OO and OY display the CellChat 71 networks derived for each animal type (see details in Methods ). ‘Source’ is the cell type the ligand comes from, whereas ‘Target’ is the cell type found matching the ligand’s receptor. Probability and P value are the statistical measures derived by CellChat. Ligand–receptor pairs are given interaction names and assigned to a pathway. Annotation provides the type of interaction are secreted signaling, ECM–receptor, cell–cell contact. Evidence codes and relevant PubMed IDs are provided by CellChat.

Supplementary Table 20

Literature-curated senescence-associated genes. Senescence-associated genes were curated from the literature 76 , 78 , for use as a reference gene set to perform fGSEA. HGNC.symbol denotes Homo sapiens gene symbol, MGI.ID denotes MGI ID number, MGI.symbol is Mus musculus gene symbol, Name is the long name of the gene and Feature type denotes the type of gene or pseudogene.

Supplementary Table 21

Senescence status GSEA per comparison. For RJV, AGA and aging, fGSEA via ClusterProfiler was run against a literature-curated senescence gene set (Supplementary Table 20 ) to derive enrichment score, NES, P value, Benjamini–Hochberg-adjusted P value for multiple testing correction, rank and genes in the leading edge.

Source Data Fig. 5

Raw images of RNA in situ hybridization for Klf6 (a) YX.

Raw images of RNA in situ hybridization for Klf6 (a) OY.

Raw images of RNA in situ hybridization for Klf6 (a) OO.

Raw images of RNA in situ hybridization for Klf6 (a) OX.

Raw images of RNA in situ hybridization for Hspa1a (c) YX.

Raw images of RNA in situ hybridization for Hspa1a (c) OY.

Raw images of RNA in situ hybridization for Hspa1a (c) OO.

Raw images of RNA in situ hybridization for Hspa1a (c) OX.

Source Data Fig. 8

Raw images of RNA in situ hybridization for Cdkn1a YX.

Raw images of RNA in situ hybridization for Cdkn1a OY.

Raw images of RNA in situ hybridization for Cdkn1a OO.

Raw images of RNA in situ hybridization for Cdkn1a OX.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Ximerakis, M., Holton, K.M., Giadone, R.M. et al. Heterochronic parabiosis reprograms the mouse brain transcriptome by shifting aging signatures in multiple cell types. Nat Aging 3 , 327–345 (2023). https://doi.org/10.1038/s43587-023-00373-6

Download citation

Received : 19 May 2022

Accepted : 30 January 2023

Published : 09 March 2023

Issue Date : March 2023

DOI : https://doi.org/10.1038/s43587-023-00373-6

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

This article is cited by

Blood-brain barrier disruption: a culprit of cognitive decline.

• Yinying Sun

Fluids and Barriers of the CNS (2024)

Restoration of neuronal progenitors by partial reprogramming in the aged neurogenic niche

• Julliana Ramirez-Matias
• Anne Brunet

Nature Aging (2024)

Aging insights from heterochronic parabiosis models

• Francisco Alejandro Lagunas-Rangel

npj Aging (2024)

In platelet single donor apheresis, platelet factor 4 levels correlated with donor’s age and decreased during storage

• Anne Claire Duchez
• Marco Heestermans
• Fabrice Cognasse

Scientific Reports (2024)

Young blood-mediated cerebromicrovascular rejuvenation through heterochronic parabiosis: enhancing blood-brain barrier integrity and capillarization in the aged mouse brain

• Rafal Gulej
• Ádám Nyúl-Tóth
• Zoltan Ungvari

GeroScience (2024)

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Science-Based Medicine

Exploring issues and controversies in the relationship between science and medicine

Parabiosis – The Next Snakeoil

However, the media likes a good story, and one of their favorite narratives is the “new miracle cure.” They will often take preliminary basic science research and present it with headlines promising a cure for some horrible disease (sometimes they will add a question mark).

When we see these headlines, we know what will happen next – hucksters will ride the hype with a wave of snake oil products promising the same cure, and claiming to be based in science. Dr. Oz will probably promote it on his show, and Mike Adams will rant about the government conspiracy to keep this cure from the public (but he will sell it to you).

We have seen this pattern with antioxidants , stem cells , resveratrol , and countless others. Sometimes the hucksters manufacture their own hype, as with green coffee beans . They don’t wait for actual scientists, they corner the market on some worthless bean or berry, then invent health claims for it and try to hype demand through the usual channels. This sadly works.

The huckster cycle is beginning again with a treatment called parabiosis. Actually, it is a treatment based on parabiosis, but is not parabiosis itself. The term refers to experiments that were first conducted in 1864 by Physiologist Paul Bert. He cut the skin of two mice, then sewed them together. When they healed together their blood vessels combined, enough so that they essentially shared their circulatory systems.

This technique became popular for studying physiology. Researchers could discover if blood factors affected some physiological property. The technique could be used to study the effects of hormones, for example.

In the 1950s researchers connected old mice to young mice to determine its effects. They found that the old mice experienced numerous rejuvenating effects. Many biomarkers of youth returned, and the mice lived longer. The younger mice connected to older mice also had shorter lifespans.

Parabiosis experiments died out in the 1970s, partly because researchers had learned what they could from the technique, and regulations of animal research made it more challenging to conduct the experiments.

Elixir of Youth

Interest in parabiosis, however, is now coming back. A recent article on Inc.com is sure to spawn interest and another cycle of snake oil promises. The article focuses on how billionaire Peter Thiel is interested in plasma transfusions from young donors as a life extension and rejuvenation treatment, based on the science of parabiosis.

I can’t help but also see the supervillain angle to this story – an aging billionaire, desperate to live forever, is feeding off the blood of young healthy victims. Of course, to get the full effect, a simple transfusion will not do. He will have to connect their circulatory system to his own.

What is the current state of the science in terms of parabiosis and anti-aging effects? Any specific health claims for humans is definitely unproven at this time, but the research is intriguing (i.e., perfect for snake oil). The mouse experiments convincingly show a benefit from parabiosis for older mice. One question is, does this benefit come from specific factors in the blood or from the fact that they are sharing their entire circulatory system?

In other words, are the younger mouse’s kidneys, livers, and lungs just supporting the older mouse’s organs by doing the heavy lifting of cleaning and oxygenating the blood? This is probably a factor, but the other option is that there are proteins in the young mouse blood that exist in smaller amounts in the older blood.

The latter seems to be true also. Researchers have found that higher levels of oxytocin in young blood, for example, stimulate muscle growth. Factors in the blood also seem to stimulate stem cells in many organs to start dividing again. This may be the main rejuvenating effect, bringing stem cells throughout the body back to life, helping to heal damage, replace cells, and increase organ function.

We are in the preliminary research stage. In order to truly answer these questions we need to do carefully-controlled clinical research in humans. Open questions include whether one time transfusions of plasma from young donors (<30) will have any enduring benefit on older recipients. It seems from mouse studies that blood cells are not necessary, just the plasma, which contains all the proteins and hormones that are likely having any effect.

If plasma is enough, then what effect does it have? Can it treat diseases like Alzheimer’s? Can it increase lifespan? How long do the effects last?

Critically – are there any risks or side effects? The mouse research revealed what was called parabiosis disease. In an experiment with 69 mouse pairs, 11 of them died quickly. It is believed this was from some incompatibility, a form of rejection. This has been reduced in later experiments by using genetically matched inbred mouse strains. So – how carefully matched do the plasma donors have to be to the recipients?

Some researchers have also raised concerns about stimulating stem cells in older patients. Parabiosis researcher Thomas Rando said:

My suspicion is that chronic treatments with anything — plasma, drugs — that rejuvenate cells in old animals is going to lead to an increase in cancer. Even if we learn how to make cells young, it’s something we’ll want to do judiciously.

Some researchers object to the notion that young plasma transfusions should be considered “rejuvenation.” No cells are being rejuvenated, they are just being stimulated into working harder. This may provide a temporary benefit, but is unlikely to significantly affect aging.

There may be a real benefit in specific disease states that are missing some critical protein in the plasma. Again – this will take years of research to sort out.

Other researchers feel we should identify the specific factors in the plasma that are having specific physiological effects and then give them in purified and specified amounts for specific conditions. This, of course, would be ideal, but the more clumsy method of just giving plasma may have to do until we identify those specific factors.

Where do things stand?

As of right now, young blood transfusions as the next elixir of youth is enjoying its 15 minutes of fame. The science is genuinely interesting, and seems deserving of further research. What is clearly needed is high quality clinical research, before any clinical claims are made.

Given history, however, it is likely that young transfusions, or even some form of parabiosis, will now also take on a life of its own as the latest snake oil product. Already there is a company called Ambrosia who is running a “study,” and as Inc.com reports:

The study is patient-funded; participants, who range in age from late 30s through 80s, must pay $8,000 to take part, and live in or travel to Monterey for treatments and follow-up assessments.

This is ethically dubious, in my opinion. Subjects should not have to pay for a treatment that is part of research. This opens the door to selling unproven treatments under the guise of doing a study (which we have seen Burzynski do for years ).

Hopefully I will be able to tell you in 10-20 years if transfusions of plasma from young donors is of any clinical benefit. Until then the treatment will likely have a second life on the fringe as snake oil. Given that this is likely to be a very expensive treatment, it will probably be elite snake oil for the wealthy.

It will probably not be limited to the wealthy, however. Hucksters prey not only on the rich, but the desperate. Poor individuals with Alzheimer’s or other serious diseases will mortgage their house, or raise funds from friends and family, to pay for very expensive yet dubious treatments, even those costing hundreds of thousands of dollars. The international stem cell clinic fraud is one example of this.

The only thing that will end this cycle of exploitation and pseudoscience is for the scientific and academic communities to take proper interest and speak out, and for government regulators to do their job to protect the public from health fraud. Sadly, this is unlikely to happen. That is part of the repeating pattern also.

• Posted in: Medical devices , Medical Ethics
• Tagged in: ageing , mouse models , parabiosis , quackery

Posted by Steven Novella

• Jump to the Navigation Menu
• Login to My AlzForum
• Jump to the Search form

In Revival of Parabiosis, Young Blood Rejuvenates Aging Microglia, Cognition

Series - 2014 Zilkha Symposium on Alzheimer’s Disease and Related Disorders : Part 1 of 4: It’s Not All About You, Neurons. Glia, Blood, Arteries Shine at Symposium Part 2 of 4: Fluid Markers and Imaging Back Idea of Breached Blood-Brain Barrier Part 3 of 4: In Revival of Parabiosis, Young Blood Rejuvenates Aging Microglia, Cognition Part 4 of 4: Glymphatic Flow, Sleep, microRNA Are Frontiers in Alzheimer’s Research

Quick Links

Further reading.

• Add to my Library
• Follow Comments

How would you like to share?

05 May 2014

As the brain ages, its microglial cells turn sluggish in their task of ingesting and degrading toxic products, and the flow of blood through its micro vessels slows. Are there components in the blood that age the brain—and can renew it?  At the Zilkha Symposium on Alzheimer’s Disease and Related Disorders, held April 4 at the University of Southern California, Los Angeles, Tony Wyss-Coray of Stanford University, Palo Alto, shared unpublished results from an ongoing study that first characterized a microglial aging phenotype and then partially reversed it with still-unknown factors from the body’s systemic milieu. 

The data offered a provocative new twist on the old specter of rejuvenation with young blood. It also reflected the power of heterochronic parabiosis, a surgical protocol of conjoining the blood supply of a young and an old mouse to study complex pathophysiological processes. The new data presented at the Zilkha symposium receives support from a flurry of separate parabiosis papers published on May 4. Led by scientists at the University of California, San Francisco, Harvard Medical School, and other institutions, these three papers demonstrate striking benefits of young blood on cognitive function, synaptic plasticity, neurogenesis, and the cerebral vasculature of old mice. The beneficial effects are not limited to the brain, and the operative factors can be identified, as one study reports that growth differentiation factor GDF11 empowers young blood to bestow regenerative oomph to old muscle.

These two mice, one old one young, live with a shared blood supply. [Image courtesy of Tony Wyss-Coray.]

Aging brings with it not only a decline in cognition but also a smoldering inflammation within the innate immune system. “This low-grade chronic inflammation is bad news,” Terrence Town of the University of Southern California in Los Angeles said at the Zilkha conference. In the brain, this manifests as an abnormal state of that organ’s main resident immune cell, the microglia. For example, expression of the microglial activation marker CD68, a lysosomal protein, rises with age. Electron micrographs of aging brain show microglia with an enlarged, dense nucleus, shriveled Golgi cisternae, few lysosomes, and vesicles jammed with lipufuscin granules. “Aging microglia clearly look abnormal,” Wyss-Coray told the conference audience. They behave abnormally, too, hardly phagocytosing in culture when presented with their usual substrates.

To see whether this is an internal affair of the aging brain or influenced by the periphery, Wyss-Coray returned to a blood-sharing experiment called parabiosis. His lab had previously used it to show that a young systemic environment can essentially rejuvenate neurogenesis and other aspects of the aging brain (see Nov 2009 news story ,  Mar 2013 news story ). 

Parabiosis involves suturing the body walls of two mice together such that their capillaries fuse. The mice then live like Siamese twins joined through their blood supply. At the Zilkha conference, Wyss-Coray said that pairing an 18-month-old with a 3-month-old mouse, and letting them live together for five weeks, reversed microglial aging. Microglial activation as measured by CD68 expression was down in the brains of old mice exposed to young blood. In the electron microscope, the old mice’s microglia looked like those of young mice, with a normal-sized, light nucleus, larger Golgi cisternae, more lysosomes, and less lipofuscin.

To the eyes of some scientists at the conference, the nucleus in microglia from old mice looked as if it contained dense chromatin that would allow less protein translation. Wyss-Coray replied that he does not know for sure how the ultrastructural changes in aging microglia relate to gene expression and function.

That said, his lab did compare the microglial transcriptome from old mice paired with other old mice to that from old mice paired with young mice. They saw that blood supplied by a young mouse did indeed largely reverse the gene expression phenotype of microglial aging. This includes age-related increases not only in CD68 and in the complement component C1qB; but also age-related decreases in progranulin, the transcription factor EGR1, and many other expression changes. 

In a separate study, Ingenuity Pathway Analysis of gene expression profiles of the aging hippocampus pointed to a synaptic plasticity network anchored by the transcription factors Creb (cAMP response element-binding protein) and EGR1 as being most preserved in rejuvenated old mice. Golgi silver staining spotted more spines on the dendrites of old mice when each had each partnered with a young mouse. A paper published by Wyss-Coray and his former postdoc Saul Villeda and colleagues on May 4 in Nature Medicine reports additional findings to back up the claim of revitalized synapses in heterochronic old mice. These include strengthened long-term potentiation as seen with electrophysiology of cultured hippocampal slices, as well as mechanistic experiments using local expression of dominant-negative Creb and RNA interference of Creb to pinpoint this transcription factor as a hub in the requisite signaling network.  “There is some sort of reactivation of a synaptic plasticity network in an old mouse exposed to young blood, ” Wyss-Coray said.

Whether these changes at the molecular and cellular level amount to better function is difficult to assess in parabiotic mice. The pairs run the rotarod together, but rigorous behavior assays are not possible. Instead, the Stanford scientists decided to model parabiosis by transferring young plasma into an old mouse once every three days for three weeks. In this study, old mice injected with plasma from young mice outperformed untreated old mice in the radial arm water maze and a fear-conditioning test. The treated mice also recapitulate other previously shown parabiosis phenotypes, including more neurogenesis, synaptic plasticity, spine density, and less neuroinflammation. The effects are not due to steroid hormones, and happen equally in male and female mice, Wyss-Coray said.

“These results are stunning. Extremely interesting,” commented Berislav Zlokovic of USC.

Are they too good to be true? “That is what some reviewers said,” Wyss-Coray replied dryly. The latest findings of this line of research are not yet published, but previous work appeared in 2009 and has been awaiting independent replication. Villeda has started his own lab at UCSF, where new students have reproduced the findings. Beyond that, few other labs have independently replicated them yet. 

That may be beginning to change with a new paper, released also on May 4, in Science magazine. In this study, researchers led by Lee Rubin at the Harvard Stem Cell Institute in Cambridge, Massachusetts, report that young blood reinvigorated blood vessels of the neurogenic niche in the brain of old mice. In heterochronic old mice, the volume of cerebral blood vessels almost doubled. They formed new branches and allowed more blood to flow through (see movie below). This, in turn, supported a robust increase in the number of neural stem cells in the old mice’s sub-ventricular zone, a brain area that is a source of new neurons throughout life but runs dry in aging (see news story on Science Now ). 

Blood flow in an old mouse brain. [Courtesy of John Chen and Greg Wojtkiewicz] 

Blood flow in a rejuvenated old mouse brain. [Courtesy of John Chen and Greg Wojtkiewicz]  

This paper confirms Wyss-Coray’s 2009 findings that factors in young blood stimulate neurogenesis in the old brain and that factors in old blood slow neurogenesis in the young brain. Rubin’s paper refines Wyss-Coray’s by reporting that it is not until mice are truly old—in this case 21 months—that their blood impairs neurogenesis of young mice. Blood from 15-month-old mice did not, suggesting that it is only during aging that these negative factors accumulate in the blood. 

These new papers aside, why were scientists at large slow to catch on to the potential of parabiosis for the study of brain aging? It may be partly because parabiosis has fallen out of favor over the past two decades and appears only now to experience a small revival. Greek for “living alongside,” this experimental system has a storied history. Used widely in physiology and endocrinology research during the first 70 years of the 20th century, parabiosis advanced the fields of growth and sex hormones and set the stage for the discovery of parathyroid hypertensive factor. Parabiosis showed the presence of the satiety factor that was later called leptin, a discovery that garnered the 2010 Albert Lasker Basic Medical Research Award  ( Coleman 2010 ). In 1972, parabiosis showed that old rats lived longer and were more vigorous when conjoined to young rats ( Ludwig et al., 1972 . Incidentally, the “trans” in this citation stands for ‘Transactions,” not "Transylvania"). Alas, the “creep factor” of creating a surgical bondage in the service of science may have set animal care committees against the technique, Wyss-Coray said, and it faded from use. 

In a recent review article arguing for a return of parabiosis to study the pathophysiology of age-related disease, Wyss-Coray writes that paired mice fare better than many other animal models exposed to pathogens, traumatic injuries, cancer, or debilitating mutations ( Eggel  and Wyss-Coray, 2014 ). In Los Angeles, he noted unpublished recovery data showing that the pairs resumed grooming and nesting, and lived a full lifespan.

Wyss-Coray’s lab initially learned parabiosis from fellow Stanford scientist Tom Rando, who used it to stimulate regeneration in liver and muscle ( Conboy et al., 2005 ). Since then, heterochronic parabiosis has boosted recovery in models of multiple sclerosis and heart failure due to weakening cardiac muscle ( Ruckh et al., 2012 ;  Loffredo et al., 2013 ). Indeed, this last study, by the groups of Lee Rubin and Amy Wagers, also at the Harvard Stem Cell Institute, gave rise to the third paper published on May 4. In it, researchers led by Wagers report that the circulatory protein GDF11 rejuvenates not only heart but also skeletal muscle. The scientists first characterized how heterochronic parabiosis restored in old mice the muscle satellite cells that promote muscle healing, and then went on to show that daily injection of recombinant GDF11 generated much the same phenotype. The new paper by Rubin et al. on the effects of young blood on the neurovascular niche of aged mice also showed that daily injection of GDF11 alone pulled off about half the beneficial effect on the brain’s capillaries and neurogenesis as that noted for whole young blood in parabiosis. 

In the view of other scientists at the Zilkha conference, the apparent success of the plasma transfer protocol and even targeted injection of candidate humoral factors is likely to prompt more laboratories to take up parabiosis.

In a press release issued by Harvard University, Rubin is quoted extensively. “We think an effect of GDF11 is the improved vascularity and blood flow, which is associated with increased neurogenesis. [This] should have more widespread effects on brain function,” he said. "We do think that, at least in principle, there will be a way to reverse some of the cognitive decline that takes place during aging. It isn't out of question that GDF11, or a drug developed from it, might be capable of slowing some of the cognitive defects associated with Alzheimer's disease." 

Rubin is further quoted as saying that a future treatment for Alzheimer’s might be a combination of a therapeutic that reduces plaques and tangles with a potential cognition enhancer like GDF11.—Gabrielle Strobel 

• Posted: 09 May 2014
• Paper: Young blood reverses age-related impairments in cognitive function and synaptic plasticity in mice.

The new findings of Villeda et al. are remarkable and thought-provoking, as they suggest the possibility that transfusion of blood/plasma from young humans can restore cognitive function in patients with mild cognitive impairment or Alzheimer’s disease. Indeed, in a previous study it was reported that cognitive function was improved in Alzheimer’s disease patients who underwent plasma exchange ( Boada et al., 2009 ). It will be critical to identify the presumptive factor(s) in the plasma of young animals that stimulate(s) neuroplasticity in the brains of old animals. Possibilities range from a neurotrophic factor to a protein that promotes removal of toxic molecules from the brain.

References:

Boada M, Ortiz P, Anaya F, Hernández I, Muñoz J, Núñez L, Olazarán J, Roca I, Cuberas G, Tárraga L, Buendia M, Pla RP, Ferrer I, Páez A . Amyloid-targeted therapeutics in Alzheimer's disease: use of human albumin in plasma exchange as a novel approach for Abeta mobilization . Drug News Perspect . 2009 Jul-Aug;22(6):325-39. PubMed .

Make a Comment

To make a comment you must login or register .

News Citations

• Chicago: The Vampire Principle—Young Blood Rejuvenates Aging Brain? 20 Nov 2009
• Blessing or Curse? Peripheral Cytokines in the Brain 22 Mar 2013

Paper Citations

• Coleman DL . A historical perspective on leptin . Nat Med . 2010 Oct;16(10):1097-9. PubMed .
• Ludwig FC, Elashoff RM . Mortality in syngeneic rat parabionts of different chronological age . Trans N Y Acad Sci . 1972 Nov;34(7):582-7. PubMed .
• Eggel A, Wyss-Coray T . A revival of parabiosis in biomedical research . Swiss Med Wkly . 2014 Feb 4;144:w13914. PubMed .
• Conboy IM, Conboy MJ, Wagers AJ, Girma ER, Weissman IL, Rando TA . Rejuvenation of aged progenitor cells by exposure to a young systemic environment . Nature . 2005 Feb 17;433(7027):760-4. PubMed .
• Ruckh JM, Zhao JW, Shadrach JL, van Wijngaarden P, Rao TN, Wagers AJ, Franklin RJ . Rejuvenation of regeneration in the aging central nervous system . Cell Stem Cell . 2012 Jan 6;10(1):96-103. PubMed .
• Loffredo FS, Steinhauser ML, Jay SM, Gannon J, Pancoast JR, Yalamanchi P, Sinha M, Dall'Osso C, Khong D, Shadrach JL, Miller CM, Singer BS, Stewart A, Psychogios N, Gerszten RE, Hartigan AJ, Kim MJ, Serwold T, Wagers AJ, Lee RT . Growth differentiation factor 11 is a circulating factor that reverses age-related cardiac hypertrophy . Cell . 2013 May 9;153(4):828-39. PubMed .

External Citations

• Science Now
• 2010 Albert Lasker Basic Medical Research Award

No Available Further Reading

Parabiosis in Mice: A Detailed Protocol

Parabiotic joining of two organisms leads to the development of a shared circulatory system. In this protocol, we describe the surgical steps to form a parabiotic connection between a wild-type mouse and a constitutive GFP-expressing mouse.

Parabiosis is a surgical union of two organisms allowing sharing of the blood circulation. Attaching the skin of two animals promotes formation of microvasculature at the site of inflammation. Parabiotic partners share their circulating antigens and thus are free of adverse immune reaction. First described by Paul Bert in 1864 1 , the parabiosis surgery was refined by Bunster and Meyer in 1933 to improve animal survival 2 . In the current protocol, two mice are surgically joined following a modification of the Bunster and Meyer technique. Animals are connected through the elbow and knee joints followed by attachment of the skin allowing firm support that prevents strain on the sutured skin. Herein, we describe in detail the parabiotic joining of a ubiquitous GFP expressing mouse to a wild type (WT) mouse. Two weeks after the procedure, the pair is separated and GFP positive cells can be detected by flow cytometric analysis in the blood circulation of the WT mouse. The blood chimerism allows one to examine the contribution of the circulating cells from one animal in the other.

Introduction

Parabiosis, the surgical joining of two organisms, was first described in 1864 by Paul Bert as a way to develop a model to study shared circulatory systems and consisted of the joining of the skin and muscular walls of two rats 1 . Parabiosis promotes formation of microvasculature at the site of inflammation 3 and has had several applications in physiological studies, such as the hormonal communication between the pituitary gland and gonads as well as the role of the kidney in hypertension 4 . It has been further employed to investigate the recruitment and integration of progenitor cells in neovascularization 5 , migration of hematopoietic stem cells 6 , and lymphocyte trafficking 7 , as well as the role and kinetics of circulating inflammatory or stem cells in tumor metastasis 8,9 , and neurodegenerative disease 10 .

One significant advantage of parabiosis lies in that the partnered animals share common circulating antigens, allowing cell migration and neovascularization without triggering an immunological reaction. Importantly, Weissman et al. have shown that parabiosis between male and female mice does not lead to formation of anti H-Y antibodies 11 .

In the original protocol described by Paul Bert the two animals were joined together through connection of the skin and muscle walls 1 . This method however, caused significant strain to the animals and resulted in high mortality due to infection of the wound. Since then the parabiosis technique has been revised by several groups with the most predominant being the protocol proposed by Bunster and Meyer in 1933 2 . Their method included joining of the scapula joints, body cavities, and skin, permitting better support and less pain for the animals. At the same time, the new method resulted in minimal post-operative care and significantly decreased mortality rates. The protocol described herein is a modification of the Bunster and Meyer technique that is less invasive and allows firmer joining. Namely, mice are connected through the elbow and knee joints as well as the skin. This joining prevents extension of the skin and therefore causes less pain and complications. Here we describe the joining of a wild type (WT) adult mouse to a constitutive GFP expressing mouse. We show that two weeks following surgery we can achieve 50% of blood chimerism demonstrating the efficacy of this surgical procedure to create a shared circulatory system.

All animal studies were performed according to the guidelines of UCLA's animal care and use committee and the National Institutes of Health Guide for the Care and Use of Laboratory Animals. The duration of the procedure described below is approximately 45-60 min from beginning to end.

1. Preparation of Surgical Field

• Perform procedure in a clean animal surgery room.
• Equipment: isoflurane Vaporizer, Gaymer T Pump with heating pad.
• Sterile tools: two curved forceps, fine scissors, needle holder.
• Sterile gloves must be used during the entire procedure.

2. Preparation of Animals

• Place two male or female mice, from same genetic background, of similar weight and size in the same cage and monitor for at least two weeks to ensure harmonious cohabitation. Female mice are preferred due to their less aggressive behavior.
• Anesthetize animals by using an isoflurane vaporizer. Place mice in a Posi-Seal Induction Chamber connected to the isoflurane vaporizer (4-5% v/v). Once anesthesia is induced, transfer the animal to the fur shaving area and maintain the anesthesia throughout the procedure through a nose cone connected to isoflurane (1.5-2% v/v). Apply ophthalmic ointment with a Q-tip to prevent dry eyes.
• Place the animal on the supine position. Thoroughly shave the left side of the mouse placed on the left and the right side of the mouse placed on the right starting at approximately 1 cm above the elbow to 1 cm below the knee.
• Aseptically prepare the shaved areas by thoroughly wiping (2-3x) with Betadine-soaked wipes followed by alcohol wipes. Place the mice on a heated pad covered by a sterile pad.
• For analgesia, administer Carprofen and Buprenorphine intraperitoneally or subcutaneously at a dose of 10 mg/kg and 0.1 mg/kg respectively.
• Place animals on their side, back to back, with adjacent shaved areas facing up. To avoid any contamination of the surgical area, cover the mice with a sterile drape exposing only the operation area. Create a small drape opening to stay sterile when performing the surgery. We made the drape window large to have a better viewing during videotaping.

3. Parabiosis

• Using a sharp scissor, perform longitudinal skin incisions to the shaved sides of each animal starting at 0.5 cm above the elbow all the way to 0.5 cm below the knee joint ( Figure 1 ). Following the incision, gently detach the skin from the subcutaneous fascia by holding the skin up with a pair of curved forceps and separate the fascia with a second pair to create 0.5 cm of free skin. Perform this separation along the entire incision.
• Begin the joining by attaching the left olecranon of one animal to the right olecranon of the other. Both olecranons and knee joints are clearly distinguishable following the skin incision. To facilitate the joining, bend the elbow of the first mouse and pass the needle of the non-absorbable 3-0 suture under the olecranon. Similarly, bend the elbow of the second mouse and pass the same suture under it. Attach joints tightly by a double surgical knot.
• Connect the knee joints following the same procedure.
• Following the attachment of the joints, connect the skin of the two animals with a continuous absorbable 5-0 Vicryl suture starting ventrally from the elbow towards the knee. To prevent skin rupture and separation perform a tight suture closure of the skin in the area around the elbows and knees. Once the ventral skin attachment has been completed, perform a double surgical knot. Place the mice in the prone position and continue the suture dorsally ending with a double surgical knot. Verify the continuity of the suture and confirm the lack of openings.
• Administer 0.5 ml of 0.9% NaCl subcutaneously to each mouse to prevent dehydration.

4. Postoperative Recovery

• Keep animals on heated pad until recovery.
• Following recovery, provide analgesics carprofen and buprenorphine. Repeat intraperitoneally or subcutaneously every 24 and 12 hr, respectively, for 48 hr at the same doses described above (step 2.5). Monitor animals for signs of pain and distress such as shaking, lethargy, chewing of tail, arched back, lack of grooming, etc. daily for two weeks.
• Prophylactically, treat mice with Sulfamethoxazole /Trimethoprim oral suspension in their water bottle 2 mg sulfa/ml +0.4 mg trim/ml for 10 days to prevent bacterial infections.
• House each parabiotic pair in a clean cage with monolithic bedding material ( e.g. paper towel or absorbent sterile pad) to prevent aspiration of bedding material. Return the animals to a bedding filled cage when they are able to maintain sternal recumbency with their head up. To minimize the strain of reaching for food while adjusting to parabiotic existence, place the moistened food pellets on the cage floor. Provide nesting material. In 1-2 weeks parabiotic mice have the ability to ambulate normally on surgically paired fore-and hind-limbs.
• Blood chimerism occurs 10-14 days following the surgery.

5. Confirmation and Reverse Procedure

• After two weeks draw blood from the tail veins of each mouse for flow-cytometric analysis to confirm blood chimerism ( Figure 3 ).
• Blood processing for flow cytometric analysis: dilute 2-3 drops of blood in 500 μl of 10 mM EDTA. Add 500 μl of 2% dextran (diluted in PBS), mix thoroughly and incubate at 37 °C for 30 min. Transfer the supernatant in a new tube, spin at 1,200 rpm for 5 min and resuspend the pelleted cells in PBS (or other buffer) for flow cytometric analysis.
• Depending on experimental design, the parabiotic pair can be separated at later time points. In our experience, we have maintained parabiosed pairs for up to 9 months without any complications.
• Perform reverse procedure on a sterile surgical surface. Anesthetize mice by placing them in an induction chamber connected to an isoflurane vaporizer (4-5%) and maintain anesthesia with 1.5-2% isoflurane, as described above (step 2.2).
• Remove hair from the area surrounding the initial suture and aseptically prepare the shaved areas as described above (steps 2.3, 2.4).
• For analgesia, administer Carprofen and Buprenorphine intraperitoneally or subcutaneously at a dose of 10 mg/kg and 0.1 mg/kg, respectively.
• Cover surgical area with a sterile drape as described above (step 2.6).
• Using sharp scissors, separate the mice through a longitudinal incision along the lateral suture and gently detach the newly formed fascia between the two mice with a pair of curved forceps. To separate the joints, cut the knots of the suture connecting them. Trim off the skin along the incision to achieve smooth edges and reattach skin with an absorbable continuous 5-0 coated Vicryl suture.
• To prevent dehydration, administer 0.5 ml of 0.9% NaCl subcutaneously to each mouse. Keep animals on heated pad until recovery.
• After recovery, repeat analgesic administration, carprofen (10 mg/kg) every 24 hr and buprenorphine (0.1 mg/kg) every 12 hr for 48 hr.
• To prevent bacterial infections treat mice with Sulfamethoxazole /Trimethoprim oral suspension (2 mg sulfa/ml +0.4 mg trim/ml) in their water bottle for 10 days.

Representative Results

The anticipated outcome of parabiosis of two organisms is the equal contribution of each animal’s circulatory system to a common blood circulation ( Figure 2 ). One can easily verify the successful equilibration of the blood of the parabiosed WT and GFP positive mice by flow cytometric analysis. Here, venous blood was obtained from the tails of both parabionts at 2 weeks after the surgery and was fractionated for peripheral blood cells (to exclude erythrocytes). The fractionated hematopoietic-derived cell fraction was subsequently analyzed by flow cytometry for the presence of GFP positive and WT cells. Consistent with previous studies, blood from the WT mouse revealed presence of chimerism, as indicated by approximately half GFP positive and half WT blood cells (53% WT cells and 47% GFP cells) ( Figure 3 ).

The parabiosis method discussed here presents minimal technical difficulties and results in low mortality rates. The attachment of knee and elbow joints is a significant improvement of the Bunster and Meyer technique. However, the procedure remains invasive thus maintenance of sterile conditions throughout the surgery is imperative. To further prevent infection of the surgical site, it is important that the parabiosed animals receive a combination of antibiotics and be monitored regularly. To ensure firm support and prevent pain, the suture connecting the elbows and knees should surround the joints rather than passing through the tissue.

As an alternative to isoflurane inhalation, anesthesia may be induced with the use of intraperitoneal injection of ketamine/xylazine or other anesthetic drugs approved by institutional committees. The advantage of using isoflurane is that it can rapidly induce anesthesia and significantly reduce the recovery time. Furthermore, the level of anesthesia can be precisely controlled.

Parabiosis, which allows circulatory systems from two animals to commingle and equilibrate, is a powerful experimental procedure for physiological studies. It presents several advantages to other commonly used techniques such as bone marrow transplantation or cell injection. Cell delivery procedures, at times, provide a short window to examine the effect of transplanted cells. In addition, immunosuppression increases the risk of infection and subsequent morbidity and mortality. However, with parabiosis one can maintain a chimeric circulation for long periods of time and study circulating factors (cells, cytokines, etc. ) independent of their origin. This system can be used to determine the role of circulating cells in wound healing, tumor formation, aging, regeneration, and inflammatory response, among many others.

Disclosures

The authors have nothing to disclose.

Acknowledgements

The authors would like to thank Adriane Mosley and Libuse Jerabek (Stanford) for assistance with the surgical technique.

• Bert, P. Expériences et considérations sur la greffe animale. Journal de l’Anatomie et de la Physiologie . 1 , 69-87 .
• Bunster, E., Meyer, R. K. Improved methods of parabiosis. Anat. Rec . 57 , 339-380 (1933).
• Waskow, C. Generation of parabiotic mice for the study of DC and DC precursor circulation. Methods Mol. Biol . 595 , 413-428 (2010).
• Finerty, J. C. Parabiosis in physiological studies. Physiol. Rev . 32 , 277-302 (1952).
• Aicher, A., Heeschen, C. Nonbone marrow-derived endothelial progenitor cells: what is their exact location. Circ. Res . 101 , e102 (2007).
• Abe, S., Boyer, C., et al. Cells derived from the circulation contribute to the repair of lung injury. Am. J. Respir. Crit. Care Med . 170 , 1158-1163 (2004).
• Donskoy, E., Goldschneider, I. Thymocytopoiesis is maintained by blood-borne precursors throughout postnatal life. A study in parabiotic mice. J. Immunol . 148 , 1604-1612 (1992).
• Powell, A. E., Anderson, E. C., et al. Fusion between Intestinal epithelial cells and macrophages in a cancer context results in nuclear reprogramming. Cancer Res . 71 , 1497-1505 (2011).
• Duyverman, A. M., Kohno, M., et al. A transient parabiosis skin transplantation model in mice. Nat. Protoc . 7 , 763-770 (2012).
• Ajami, B., Bennett, J. L., et al. Local self-renewal can sustain CNS microglia maintenance and function throughout adult life. Nat. Neurosci . 10 , 1538-1543 (2007).
• Weissman, I. L., Jerabek, L., et al. Tolerance and the H-Y antigen: Requirement for male T cells, but not B cells, to induce tolerance in neonatal female mice. Transplantation . 37 , 3-6 (1984).

Cite This Article

To prove you're not a robot, please enter the text in the image below.

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• HHS Author Manuscripts

Parabiosis for the study of age-related chronic disease

Alexander eggel.

1 Institute of Immunology, University of Bern, 3010 Bern, Switzerland

Tony Wyss-Coray

2 Department of Neurology and Neurological Sciences, Stanford University School of Medicine, Stanford, California 94305, USA

3 Center for Tissue Regeneration, Repair and Restoration, VA Palo Alto Health Care System, Palo Alto, California 94304, USA

Modern medicine wields the power to treat large numbers of diseases and injuries most of us would have died from just a hundred years ago. In view of this tremendous achievement, it can seem as if progress has slowed, and we have been unable to impact the most devastating diseases of our time. Chronic diseases of age such as cardiovascular disease, diabetes, osteoarthritis, or Alzheimer’s disease turn out to be of a complexity that may require transformative ideas and paradigms to understand and treat them. Parabiosis, which mimics aspects of the naturally occurring shared blood supply in conjoined twins in humans and certain animals, may just have the power to be such a transformative experimental paradigm. Forgotten and now shunned in many countries, it has contributed to major breakthroughs in tumor biology, endocrinology, and transplantation research in the past century, and a set of new studies in the US and Britain report stunning advances in stem cell biology and tissue regeneration using parabiosis between young and old mice. We review here briefly the history of parabiosis and discuss its utility to study physiological and pathophysiological processes. We argue that parabiosis is a technique that should enjoy wider acceptance and application, and that policies should be revisited especially if one is to study complex age-related, chronic disorders.

Parabiosis – an experimental model inspired by nature

Conjoined twins have fascinated people ever since this naturally occurring physiologic condition gained worldwide publicity through the Siamese brothers Chang and Eng Bunker in the early 19 th century. Even though the term “Siamese twins” has been derived from their case, older reports describing conjoined twins date back to the year 1100. The occurrence of this condition is around 1:100000; however, only 26% survive birth [ 1 , 2 ]. The degree of conjunction and the points of attachment vary substantially between different cases and an anatomic terminology has been introduced to classify the types of unions [ 1 , 2 ]. Despite considerable progress in the medical field, the chance of a successful surgical separation of the two individuals still depends on how many vital organs are shared. There are several well-known cases in which a separation of the twins has not been possible or has been declined. For Chang and Eng Bunker, such a surgery was not an option and, therefore, they have adapted to living a conjoined life, staying together to an age of 63 years. Conjoined twins develop astonishing coordinative abilities, and for a long time one could only speculate on the physiologic mechanisms underlying this higher form of inter-communication between the two individuals.

In order to investigate the influences of an organism on its conjoined partner, scientists came up with an animal model that essentially copies the natural phenomenon of conjoined twins. The surgical technique to physically connect two living organisms that was later termed “parabiosis” (from the Greek words, para “besides” and bios “life”) was first introduced by the French physiologist Paul Bert in the 1860’s using white albino rats ( Fig. 1a ). In the beginning, parabiosis surgeries consisted of short skin incisions and a suturing together at the flank of each animal, but the technique has evolved over the years. Nowadays, the skin incisions typically extend along the whole body flank; additionally, in some models the limbs are sutured at the joints and the abdominal walls are joined in order to increase stability and the surface for vascularization. A detailed procedure of the surgery including reversal of the parabiotic pairing has recently been described by Conboy et al. [ 3 ]. Following the first experiments using rats, other animal species including axolotls [ 4 ]have been included in parabiosis experiments but it turned out that rodents recovered best from the surgery, displaying remarkable resistance against wound infections as opposed to higher mammals. Therefore, the majority of subsequent investigations have been conducted with rats or mice. In addition to connecting adult organisms for parabiosis, embryonic tissue has also been fused in amphibians and fish to study developmental processes (e.g. [ 5 ]). Early parabiosis studies using adult animals reported cases of parabiosis intoxication in which one of the two parabionts suddenly died [ 6 , 7 ]. While this intoxication has mainly been due to the lack of genetic uniformity resulting in tissue rejection, a survival rate similar to other invasive surgical procedures (>80%) can now be attained in mouse parabionts with appropriate precautions taken by a skilled operator, and even long-term survival seems unaffected (own observation). So far, most of the parabiosis studies have been conducted in the US and in Japan, whereas only few publications originate from Europe ( Fig. 1b ).

(a) The annual number of publications using parabiosis is shown from 1860–2013. Several studies are highlighted as they provided groundbreaking findings. (b) Publications including parabiosis experiments are listed for different countries. All values have been extracted from http://www.gopubmed.org

The early days of parabiosis

In his doctoral thesis “la greffe animale” Bert sutured the skin of two albino rats at their flanks and found that intravenously administered fluids passed from the circulation of one animal into the bloodstream of its adjacent partner. He therefore postulated that surgically connected animals spontaneously develop a single, shared circulatory system through anastomosis ( Fig. 2 ) [ 8 ]. For his pioneering work Bert was awarded the Prize of Experimental Physiology of the French Academy of Science in the year 1866. Thereafter, very few studies followed up on his approach until the early 20 th century.

Organism A and B share a common blood supply, which spontaneously develops through anastomosis post-surgery. Organisms with different physiologic conditions may be used for parabiosis in order to assess the systemic effect of one organism on a particular tissue of interest in its attached partner.

In 1908 the German surgeons Sauerbruch and Heyde revived the technique and introduced the term parabiosis for the artificially established symbiosis between two animals [ 9 ]. Researchers from a variety of different fields (e.g., endocrinology, metabolism, transplantation, nephrology, radiology, allergy and immunology) started to take advantage of the parabiosis model for their own scientific investigations. A main question at that time was whether transmissible, humoral factors present in one animal have a physiological effect on its adjacent partner. Rous, who won a Noble Prize in 1966 for his discovery of tumor-inducing viruses, used parabiosis to examine whether the presence of circulating anti-cancer antibodies in tumor-resistant rats would affect tumor susceptibility in attached non-resistant rats. He did not succeed in identifying such protective humoral anti-cancer factors in these experiments [ 10 ], but parabiosis was instrumental in his early studies. The most striking results that were obtained using the parabiosis model in this early era have been summarized in an extensive review [ 11 ].

More than 1700 articles related to parabiosis have been published (source: http//www.gopubmed.org ) since Bert’s original dissertation. A publication peak was reached in the years between 1960–1980 ( Fig. 1a ). In 1969, Coleman grafted mice with the mutation diabetes ( db/db ), which are prone to become obese and develop type II diabetes, to inbred wildtype mice [ 12 ]. He initially hypothesized that the db/db mouse would lose weight upon exposure to a systemic environment of a non-obese mouse. Surprisingly, he observed that the wildtype mouse significantly decreased food intake while the obese mouse continued to gain body weight. Coleman concluded that there must be a satiety factor involved to which only the wildtype but not the db/db mouse had been able to respond [ 13 ]. Almost three decades later, Friedman finally identified this satiety factor and called it leptin [ 14 ]. Today, leptin is known as one of the key hormones regulating body weight. Shortly after this remarkable discovery, Friedman and Leibel found that the db gene encodes for the leptin receptor and that mutations in this gene result in a non-functional molecule [ 15 , 16 ]. This finding, which earned Coleman and Friedman the 2010 Lasker award, clearly confirmed Coleman’s interpretation of his earlier experiments and underlines the importance of parabiosis models for the identification of new transmissible, humoral factors.

In 1969, another remarkable study using parabiotic pairings was performed by Lewis K. Dahl’s group [ 17 ]. They grafted wildtype rats to partners with constitutional predisposition for hypertension. As a result they found that renoprival hypertension occurred in both rats at the same frequency. Again this finding pointed towards a humoral factor inducing hypertension in the wildtype animal. Additionally, they described that nephrectomized rats with a predisposition to develop hypertension did not induce higher blood pressure in the wildtype parabiont, suggesting that the factor is produced in the kidney of hypertensinogenic rats. The presence of this factor has subsequently been confirmed in other studies [ 18 ] and, in 1993, Lewanczuk et al. identified it as parathyroid hypertensive factor (PHF) [ 19 ].

Parabiosis was not only helpful to discover and study individual humoral factors but also to assess the physiological consequences in an organism upon exposure to the systemic environment of its attached partner. Initially, parabiotic surgeries showed highest success rates when using young, sex- and age-matched littermates. Over time the procedure has improved and, in the early 70’s, scientists started to graft animals of different ages to each other. This heterochronic parabiosis set the basis for the investigation of effects induced through exposure of an aged organism to a youthful systemic environment. In their studies, Ludwig and Elashoff particularly focused on the extension of lifespan in the old heterochronic parabiont when attached to a young counterpart. Indeed, in 1972 their results provided the first evidence that the old organism in the heterochronic pairing lived longer in response to the young environment compared to the age-matched isochronic control animals [ 20 ]. Later, this model proved critical to study the physiology of aging and stem cells in different tissues and organ systems (see below).

Parabiosis for the study of aging and tissue regeneration

In spite of these remarkable findings, based in part on parabiosis, by the end of the last century the procedure had fallen out of favor with the research community with only a handful of papers using the technique ( Fig. 1a ). It was at that time when Drs. Weissman, Wagers, and Rando “rediscovered” parabiosis at Stanford University for the study of stem cell engraftment and trans-differentiation [ 21 , 22 ] as well as tissue regeneration in the aged organism [ 23 ]. Different studies have shown that the regenerative capacity of tissues and organs are dependent on the proliferative activity of progenitor cells derived from tissue-resident stem cells [ 24 – 28 ]. A major hallmark of aging is that the regenerative properties significantly decline in most tissues. This has partially been attributed to impaired stem cell function [ 29 – 31 ]. However, whether these age-related effects were due to cell intrinsic changes or alterations in the microenvironment of stem cells required further investigation. In 2005 Conboy et al. used heterochronic parabiosis experiments to address this question. They showed that factors derived from the young systemic environment are able to activate molecular signaling pathways in hepatic or muscle stem cells of the old parabiont leading to increased proliferation and tissue regeneration. These in vivo results were furthermore confirmed ex vivo by culturing muscle stem cells in medium containing serum from young animals [ 23 ]. Their findings clearly suggest that the age-associated impairment of stem cell function is induced to a significant extent by the molecular composition of the surrounding niche rather than by cell intrinsic changes alone.

In 2011 our group published a similar finding suggesting an old systemic environment can be detrimental for stem cell function and negatively regulate adult neurogenesis in brains of young heterochronic parabionts. This led to the discovery that factors in old blood are sufficient to decrease synaptic plasticity and impair contextual fear conditioning and spatial memory. Using a systematic proteomic approach ( Fig. 2 ) we were able to identify soluble factors that were significantly increased in blood plasma of old mice and humans. One of these factors was the chemokine CCL11 (eotaxin), known to chemotactically attract eosinophils to tissues. Indeed, application of CCL11 was sufficient to induce impaired adult neurogenesis [ 32 ]. Again, these findings provide evidence that the age-related decline in stem cell function can be attributed to changes in the systemic environment. Three more recent publications using heterochronic parabiosis further support this conclusion. Ruckh et al. reported that recovery from experimentally induced demyelination in the CNS is enhanced in old mice that were exposed to a young systemic environment [ 33 ]. Salpeter and colleagues showed that the decline in pancreatic β-cell proliferation in old mice can be reversed in old parabionts paired with young mice [ 34 ]. And most recently, Loffredo et al. demonstrated that age-related loss of normal cardiac function leading to diastolic heart failure is partially due to the lack of certain circulating factors in old mice. They reported that this hypertrophy is reversible upon exposure of an aged animal to a youthful systemic environment through heterochronic parabiosis. They identified growth differentiation factor 11 (GDF11), which is significantly reduced in the blood plasma of old mice, as a crucial factor to prevent cardiac hypertrophy.

The promise of parabiosis for regenerative medicine and the study of age-related diseases

The value of parabiosis as an experimental model is most evident for physiological or pathophysiological studies that affect the organism as a whole or that induce changes in the circulatory system. Naturally, such (patho)physiological studies are most relevant to understanding the complexity of higher organisms and disease processes, but they are also the most challenging to conduct and they cannot be replaced by in vitro experiments. Indeed, it becomes increasingly evident that many diseases and biological processes, including aging, result in organism-wide, systemic changes contributing to local tissue alterations. Thus, studying an individual organ or cell type in isolation may not lead to a holistic understanding of events. This shift in thinking has been particularly striking with respect to the brain, where decades of neuron-centric research has started to give way to include studies on other brain cell types as critical regulators of cognition and disease, and where a growing number of studies document effects of factors outside the brain including gut microbiota, diet, and other systemic changes on CNS function [ 35 – 39 ].

We think parabiosis is an ideal tool to ask whether alterations occurring in an organism as a consequence of disease, aging, genetic background, infection, diet, exercise etc. might result in circulatory changes altering the status of a healthy, young, uninfected or sedentary organism ( Fig. 3 ). Thus, parabiosis may help assessing the effects of any number of functional states of one organism on a partner organism through a shared circulatory system. This is, of course, only a first step in linking particular factors or cells to a newly discovered transmissible effect. But as the above cited reports show, it has indeed been possible to identify, for example, cells that regenerate an injured brain [ 33 ] or proteins that induce satiety [ 13 ], regenerate an aging heart [ 40 ], or accelerate aspects of brain aging [ 32 ]. A generalized approach to reveal such factors or cells using heterochronic parabiosis is to analyze systemic changes and correlate them with local alterations in a particular tissue of interest ( Fig. 2 ). Whether the identified candidates are necessary or sufficient to induce pathophysiology may subsequently be assessed by exogenous application or neutralization as well as endogenous overexpression or ablation experiments in suitable animal models.

Using heterochronic pairings of young (A) and old (B) mice allows assessment of the effect of a young systemic environment on a particular local tissue of interest in the aged partner and vice versa. Isochronic pairings (AA or BB) are important controls to exclude surgery-related observations and to determine age-related changes in the systemic environment. Briefly, systemic body fluids such as blood, lymph or CSF are collected, assessed with OMICS tools such as protein, lipid or hormone arrays and analyzed for differential levels of soluble factors pre- and post-parabiosis. A particular tissue of interest is isolated, phenotypically characterized by flow cytometry, immunohistochemistry or epi-/genetic measures and analyzed for parabiosis-induced phenotypic changes. The integration of these data leads to the identification of individual candidate factors or cells that can subsequently be tested in a suitable mouse model.

As many of the major untreatable diseases of our time are chiefly dependent on aging, understanding them will require more insight into the systemic changes and the resulting molecular alterations occurring with age. Animal models can replicate many aspects of chronic diseases including heart disease, stroke, or neurodegeneration, yet we know very little about the contribution of the systemic environment and aging to these conditions. Parabiosis and heterochronic parabiosis in particular could help answering some of the fundamental questions in this regard: are circulatory factors or cells in a young organism protecting against age-related disease, and vice versa, are factors or cells in the old organism predisposing or promoting disease in a younger organism? Parabiosis between mutant mice genetically manipulated to develop disease and age-matched or heterochronic wildtype littermates or between other genetically engineered mice can help address the importance of systemic factors in the disease process. Variations of this paradigm can help elucidate pathways and mediators in many other conditions ( Fig. 3 ).

Parabiosis has led to remarkable biological and medical discoveries over the last decades and over the past few years in particular. Given its success, it is surprising that this model is not used more extensively. Mice adapt remarkably to the paired living as they gain mobility quickly after surgery and start building nests. When performed with the appropriate refinements and considerations, they do not show overtly abnormal behavior and the survival rate is not affected by the new physiologic state. The highlighted studies underline the promise of the parabiosis model to study aging, stem cells, and tissue regeneration but the model can be employed to address many other aspects of physiology or disease. Considering the remarkable rejuvenating impact young mice have on aged tissues in heterochronic pairings, we predict that parabiosis will experience another revival over the next years and will hopefully accelerate our progress towards curing the most devastating diseases of our time.

Acknowledgments

We thank Drs. Kira Mosher, Jinte Middeldorp, and Joseph Castellano for insightful comments on the manuscript. This work was supported by a Fondation Acteria Award (A.E.) and a Swiss National Science Foundation Ambizione grant (PZ00P3_148185, A.E.), Anonymous (T.W.-C), Department of Veterans Affairs (T.W.-C), a California Institute for Regenerative Medicine Award (T.W.-C), and a National Institutes of Health Institute on Aging (R01 AG027505, T.W.-C).

Parabiosis Experiments Prove Bloodborne Aging Factors

Macabre experiments surgically attaching old animals to young ones show there is something in the blood that causes aging.

Below is an approximation of this video’s audio content. To see any graphs, charts, graphics, images, and quotes to which Dr. Greger may be referring, watch the above video.

One of the major hallmarks of aging is the decline of regenerative capacity of our tissues. There are stem cells residing in our muscles, for example, that can leap into service at the first sign of injury to repair any damage. Is the waning of tissue-renewing abilities due to some intrinsic property of aging stem cells, or a consequence of being trapped in an aging body? To find out, researchers grafted the muscles of old rats into young rats, and vice versa. Inside the young rat, even the weak atrophied muscles of extremely old rats at the end of their lives regained their strength, volume, and ability to regenerate. They became young again, so the capacity was still there all along. And young muscles in old rats lost renewal capacity. So, it appeared to be something about the surrounding milieu, rather than inherent defects with age.

To see if the critical elements were circulating in the bloodstream, old muscle stem cells were cultured in the blood of young animals. This alone had a rejuvenating effect, suggesting that there may be some sort of vitalizing factors in youth that we lose as we age (or inversely some repair-repressing component that builds up). Either way, this may be good news, because if we can find out what those factors are, we may be able to slow aging or even reverse it.

Could more than just muscle be restored? What about the brain and all the other organs? To see the extent in which circulating factors might play in affecting aging, researchers turned to a macabre procedure called parabiosis, from the Greek para , meaning “next to,” and bios , for “life.” It was an attempt to recreate the phenomenon of conjoined twins in a lab by sewing animals together to study the effects of transmissible factors.

Conjoined twins are often referred to as “Siamese twins,” due to the notoriety of a 19 th century pair of Siamese-American brothers joined at the chest. The Blažek sisters are the only conjoined twins on record ever having given birth. Josepha and Rosa Blažek were quite literally joined at the hip. When Rosa got pregnant, both of their breasts developed and started lactating, supporting the theory we now know to be true today that lactation is regulated by hormones that circulate in the blood. So, what about aging factors circulating in the blood?

Surgically, researchers can connect the skin, circulation, muscle walls, body cavities, and shoulder blade joints of two animals. Early attempts to graft different species failed—for example, mammals to birds, or a cat to a rat. But in 1862, a pioneering French scientist was able to successfully pair together two young rats.

The first heterochronic union ( hetero meaning “different,” khronos meaning “time”) was created in 1955 to answer the question: what would happen if you bathed the tissues of an old animal in the blood of a young one? The title of the paper was “Experimental Prolongation of the Life Span.” Old rats hooked up to young rats lived about 20 percent longer than old rats hooked up to one another. Subsequent experiments showed old mice coupled with young became healthier, stronger, and smarter. Aged tissues in a number of organs were rejuvenated, including the brain, heart, pancreas, skeleton, and muscles.

That doesn’t necessarily mean that aging was slowed, though. Rather than some sort of restorative bloodborne factor, maybe the older animals were just taking advantage of the reserve organ capacity in the younger animals––like having an extra set of youthful kidneys. To see whether there’s some aging or anti-aging transmissible element, rather than sharing organs and an entire circulatory system, what about just getting a transfusion of young blood? After all, as an American Aging Association journal review concluded: “The use of parabiosis in humans is currently not performed due to the surgical complications and resulting undesirable lifestyle.” Ya think? But getting a transfusion of young blood would be easy. Does it work? We’ll find out next.

Please consider volunteering  to help out on the site.

• Conboy IM, Conboy MJ, Wagers AJ, Girma ER, Weissman IL, Rando TA. Rejuvenation of aged progenitor cells by exposure to a young systemic environment. Nature. 2005;433(7027):760-764.
• Finerty JC. Parabiosis in physiological studies. Physiol Rev. 1952;32(3):277-302.
• McCay CM, Pope F, Lunsford W, Sperling G, Sambhavaphol P. Parabiosis between old and young rats. Gerontologia. 1957;1(1):7-17.
• Eggel A, Wyss-Coray T. A revival of parabiosis in biomedical research. Swiss Med Wkly. 2014;144:w13914.
• Ashapkin VV, Kutueva LI, Vanyushin BF. The effects of parabiosis on aging and age-related diseases. Adv Exp Med Biol. 2020;1260:107-122.
• Kang JS, Yang YR. Circulating plasma factors involved in rejuvenation. Aging (Albany NY). 2020;12(22):23394-23408.
• Zhang H, Cherian R, Jin K. Systemic milieu and age-related deterioration. Geroscience. 2019;41(3):275-284.
• Hofmann B. Young blood rejuvenates old bodies: a call for reflection when moving from mice to men. Transfus Med Hemother. 2018;45(1):67-71.
• Rebo J, Mehdipour M, Gathwala R, et al. A single heterochronic blood exchange reveals rapid inhibition of multiple tissues by old blood. Nat Commun. 2016;7:13363.
• Carlson BM, Faulkner JA. Muscle transplantation between young and old rats: age of host determines recovery. Am J Physiol. 1989;256(6 Pt 1):C1262-1266.
• Carlson BM, Dedkov EI, Borisov AB, Faulkner JA. Skeletal muscle regeneration in very old rats. J Gerontol A Biol Sci Med Sci. 2001;56(5):B224-233.
• Schuknecht HF. The Siamese twins, Eng and Chang: their lives and their hearing losses. Arch Otolaryngol. 1979;105(12):737-740.
• Sills ES, Vrbikova J, Kastratovic-Kotlica B. Conjoined twins, conception, pregnancy, and delivery: a reproductive history of the pygopagus Blazek sisters (1878-1922). Am J Obstet Gynecol. 2001;185(6):1396-1402.
• Pope F, Lunsford W, McCay CM. Experimental prolongation of the life span. J Chronic Dis. 1956;4(2):153-158.
• Ludwig FC, Elashoff RM. Mortality in syngeneic rat parabionts of different chronological age. Trans N Y Acad Sci. 1972;34(7):582-587.

Acknowledgements

Motion graphics by  Avo Media

• animal studies

View Transcript

Sources cited, republishing "parabiosis experiments prove bloodborne aging factors".

You may republish this material online or in print under our Creative Commons licence . You must attribute the article to NutritionFacts.org with a link back to our website in your republication.

If any changes are made to the original text or video, you must indicate, reasonably, what has changed about the article or video.

You may not use our material for commercial purposes.

You may not apply legal terms or technological measures that restrict others from doing anything permitted here.

If you have any questions, please Contact Us

Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0)

Doctor's note.

So, Can Getting Transfusions of Young Blood Slow Aging? . That’s the next video.

My new book, How Not to Age , is all about aging, and you can get your copy now at your local public library or wherever books are sold. If you haven’t seen them yet, check out the book trailer and my new presentation . (As always, all proceeds I receive from all of my books are donated to charity.)

If you haven’t yet, you can subscribe to my videos for free by clicking here . Read our important information about translations here .

Subscribe to our free newsletter and receive our Care for Your Skin as You Age infographic.

How Well Do Medicines Like Fosamax Work to Treat Osteoporosis?

Friday favorites: the optimal vitamin b12 dosage and type, subscribe to nutritionfacts.org videos.

By subscribing, you will automatically receive the latest videos emailed to you or downloaded to your computer or portable device. Select the subscription method below that best fits your lifestyle.

Or subscribe with your favorite app by using the address below:

iOS (iPhone, iPad, and iPod)

To subscribe, select the "Subscribe on iTunes" button above.

To subscribe, select the "Subscribe via E-Mail" button above.

Mac and Windows

Android and amazon fire.

To subscribe, select the "Subscribe on Android" button above.

Using Your Favorite Application

Copy the address found in the box above and paste into your favorite podcast application or news reader.

Bookmarking NutritionFacts.org

To bookmark this site, press the Ctrl + D keys on your Windows keyboard, or Command + D for Mac.

Pin It on Pinterest