, , ,

Yes, T cells mature/develop in the thymus and yes, the thymus shrinks with age through a process called Involution. Occurring in many vertebrates, thymic involution is an ancient and evolutionarily conserved process (1). Let’s examine the thymus in step-wise fashion. How does it shrink? Why is it important? Could mouse models be inconsequential for understanding human thymus function? Finally, why does it shrink with age?

Before proceeding, let’s define a few immunological terms pertinent for this discussion.
Peripheral: Immunologists define the thymus as the site of T cell development/maturation. Everywhere else we find T cells, we call the periphery, and naturally, such T cells we call peripheral, as opposed to thymic, T cells.
Germline: The key identifier of T cells is their T cell receptor (TCR). Immature T cells start out with a TCR we call germline TCR. During T cell development/maturation, this TCR undergoes re-arrangement through genetic recombination to become entirely different from the one the immature T cell started out with.
Senescence: Permanent cell cycle arrest. Hayflick originally identified it in human fibroblasts (2). Senescent cells either no longer divide/can divide. Same concept applied to immune cells is called Immunosenescence (3).
Repertoire: Typically refers to the spectrum of unique TCRs (and BCRs; B cell receptor) in an individual. Naive repertoire refers to the spectrum of new (naive) T and B cells and their receptors.

How does the thymus shrink with age?
Current consensus says thymus shrinks in an age-dependent program. Though the process starts during puberty or even earlier, in middle age adipocytes (fat cells) infiltrate and take over. In short, with age, the thymus changes from a primary lymphoid organ into a fatty tissue (4).

C, cortex; M, medulla
From 5 (figure is based on original data from 6)

Why is the thymus important?
As the thymus shrinks with age, fewer naive (new) T cells come out of it. The unique feature of T and B cells that  renders them responsive to billions of antigens is key for understanding the ramifications of this decline. Somatic, i.e. non-germ (eggs and sperm) cells, usually do not undergo genetic recombination. Uniquely among somatic cells, T and B cells undergo genetic recombination during their development. This happens in the thymus for T cells and in the bone marrow for B cells. As a result of this recombination, mature T cells coming out of the thymus bear re-arranged TCRs, each specific for one antigen (peptide)* out of a vast universe of antigens. A picture is worth a thousand words? With T cells, numbers reveal their awe-inspiring scope. We humans have approximately 3X1011 T cells in circulation, capable of recognizing approximately a billion unique antigens, and our thymus needs to generate approximately 3X109 T cells daily to replenish this circulating pool (7).

Coming out of the thymus, T cells circulate in our body, those that get activated by their specific antigen (peptide) expand and become memory T cells while those that don’t, die. Let’s remember there’s a universe of antigens out there, and we don’t know ahead of time which ones our T cells would need to respond to. With age, with fewer new T cells coming out of our thymus, the range of antigens recognized by our circulating T cells shrinks and we become less capable of clearing new infections. When thymus is removed (thymectomy) within a few weeks of birth, circulating T cells of even 22 year olds resembled those of 75 year olds (8, 9), suggesting early-life thymus is key for long-term healthy T cell function.

The human is not just an overlarge mouse.
Mouse models suggest thymus involution weakens T cell function. Does the same hold for humans? Recent data suggests not. Phew! Rather, this mouse-human difference is a difference in kind, not degree.

  1. Human naive T cells self-renew from the existing peripheral T cell pool (10).
  2. A mathematical (in silico) model (11) also suggests our peripheral T cells self-renew.
  3. Our thymus starts to shrink, and reduces T cell output already in our 20s, yet our circulating naive CD4 T cell repertoire stays more or less steady for decades, abruptly collapsing only in our 80s (12, 13). Reasons for this late-age abrupt collapse are still unknown.

An important caveat about human studies on T cell function? Most of our data is from circulating blood cells. Problem? Yes, for two reasons.

  1. T cells circulating in our blood account for only about 2% of our total T cell pool (14)
  2. As Jurgen Westermann, a German immunologist, has argued for years, immune cells don’t leave tissues such as bone marrow, spleen, lymph nodes at random to circulate in our blood. They do so in response to specific cues and to perform specific functions.

Yet, we freely, egregiously and foolishly extrapolate from circulating T cell function to entire body T cell function.

When it comes to aging and immune function, we are left with the inherent contradiction of the less active (immunosenescence) co-existing with the overactive (autoimmunity) (15). How so?

  1. With age, our immune function becomes senescent (immunosenescence) and less effective (15). Old age is associated with poor response to vaccinations (16, 17) and to chronic viral infections (18), and much lower acute transplant rejections (19, 20).
  2. At the same time, the older we are, more we are prone to autoimmune diseases such as rheumatoid arthritis (21). Autoimmune diseases represent overactive, not less active, immune function.

How to deal with this contradiction? Let’s examine possible reasons.

  1. Maybe elderly with autoimmunity have dysfunctional physiology, and T cells specific for the autoimmune protein targets expand at the expense of vaccine-specific T cells? For e.g., elderly autoimmune T cells have signaling pathway changes that render them more, not less, responsive (15).
  2. Maybe elderly do have similar numbers of vaccine-specific T cells, only they are dysfunctional or non-responsive, having undergone specific biochemical changes. Thus, even though immune senescence and autoimmunity appear contradictory on the surface, they have key similarities in specific processes such as chronic DNA damage like telomere erosion (15), though we don’t yet have definitive data for this being responsible for T cell aging effects (22).
  3. Elderly with poor response to vaccines could have persistent viral infections (23) such as Varicella-zoster, measles, HIV-1, CMV (cytomegalovirus), viruses that, for some reason, are not cleared but persist/reside inside cells such as epithelium, neurons, immune cells, sitting there quietly (latent), popping out every now and then to re-activate virus-specific T cells. Over time, such virus-specific T cells would expand at the expense of other, particularly naive T cells. After all, our body offers a finite space. T cells that get stimulated stick around, those that don’t, die.
  4. In some elderly, the gut becomes leaky, letting in more microbes and microbial products, which chronically stimulate their T cells, a process evocatively called Inflammaging (24, 25).

We finally come to the most important question.
Why does the thymus shrink with age?
Our thymic T cell output is highest immediately post-birth, and then steadily declines, with little output after 20 years of age. 20! That’s not old or elderly by any stretch of the imagination. While an intriguing 1999 study showed adult human thymus had similar thymopoiesis (T cell development) compared to those in fetuses (26), it had two main problems. One, few donors (looks like no more than 4 adult donors). Two, age not specified except to say above 23 years of age. Thus, puberty triggers thymic involution and a study showed it starts even earlier (27). What’s the link with puberty? Do sex steroid hormones initiate thymic involution? That’s debatable (28). Nevertheless they kill thymocytes (29). Thus, with puberty, there’s a sharp decline in thymus tissue. After this sudden sharp reduction, involution continues at a steady rate with about 3% loss per year until middle age, and about 1% loss per year thereafter (30, 31).

With age, why does the thymus change from lymphoid tissue into fat? To avoid the danger of mistaking effect for cause, let’s recognize that the thymus starts to shrink (involute) long before it gets fat deposits (32). However, aged thymi do indeed have fat deposits, and these adipocytes may secrete cytokines that alter/impair thymic function. Thus, while fat deposition may not start thymic involution, it may help set it in stone, or rather fat.

Fat deposition aside, two types of age-associated changes may trigger and/or maintain thymic involution.

  1. Bone marrow output decreases. It’s the source of hematopoietic stem cells (HSC). This includes T cell progenitor cells which, in response to as yet unknown cues, enter the blood circulation and from there the thymus, and interact with thymic epithelial cells to mature in a strictly defined developmental program. With age, bone marrow output of HSCs including T cell progenitor cells declines (33).
  2. The thymus tissue (stroma) becomes fibrotic. Specialized thymic epithelial cells provide signals necessary for thymopoiesis (cell maturation process) and with age, are replaced by fibroblasts (fibrosis; 34). Increasing fibrosis in many organs such as heart, kidney and liver is indeed associated with aging, and may be a common aging signature.

Donald B. Palmer and colleagues from the Royal Veterinary College at the University of London in the UK (35, 36) propose an intriguing idea about thymic involution, separating it into a two-stage process. The first one, in puberty, they call “growth dependent thymic involution”, occurring as it does during a period of physiological growth and development. The second one, they call “age-dependent thymic involution”, which occurs in tandem with other age-related changes in the body.

Thus, during puberty, we undergo “growth dependent thymic involution”, while the rate and degree of “age-dependent thymic involution” is likely influenced by our genetics and lifestyle choices. For example, many studies suggest moderate physical activity supports and maintains immune function in the elderly (37, 38, 39). How could exercise do this? After all, it doesn’t reverse thymic involution per se so it wouldn’t increase new T cell output. Maybe exercise correctly balances the milieu in which T cells operate, maybe it leads to healthier lymphatics which help T cells circulate properly and efficiently, maybe it better balances the neuro-endocrine system, helping allay or delay thymic involuting factors such as adipocyte infiltration? Open questions yet.

Regardless thymus involution and attendant reduction in new T cell output, we have among us the successfully aged, namely centenarians. What’s different about their thymus and their T cells? Data is limited, not extensive.

  1. Long-term Swedish studies (40, 41) found centenarians did not have an Immune Risk Profile (IRP). What’s the IRP? Inverted circulating CD4 to CD8 T cell ratio. High CD4 to CD8 T cell ratio is considered normal in the young. However, a newer data set from 151 healthy >65 year olds found two other parameters correlated with mortality (42). One was higher levels of a liver-derived protein, C-reactive protein (CRP), indicative of systemic inflammation. The other was lower thymic output, measured using a newer technique called signal-joint T cell receptor (TCR) excision circles (sjTREC). This  suggests active thymic output may remain relevant even into late old age.
  2. Another centenarian study found genotypic differences in a particular cytokine, IL-10 (43). We immunologists think IL-10 dampens exuberant immunity.

Proposed originally in the 1970s by the acclaimed French immunologist, Jean-Francois Bach, the most compelling idea yet about the thymus is a long-forgotten one, that it’s an endocrine organ. Unfortunately, we simultaneously discovered that it was also the site of T cell development, and since then its thymopoiesis function became the research focus at the expense of its endocrine function (44). For example, as an endocrine gland, the thymus secretes thymulin. Discovered and characterized by Bach, it’s a metallopeptide that, among other things, mediates thymus-pitutary communication (44). Thymus as an endocrine gland throws up many more open questions about its function. Circadian rhythms, seasonal changes, sex differences in autoimmunity. Careful, extensive and imaginative inter-disciplinary study of the centenarians among us will help resolve the mystery of thymus function in healthy aging.


  1. Shanley, Daryl P., et al. “An evolutionary perspective on the mechanisms of immunosenescence.” Trends in immunology 30.7 (2009): 374-381.
  2. Page on google.com=
  3. Page on nih.gov
  4. Page on nih.gov
  5. Page on nih.gov
  6. Leukemia Inhibitory Factor, Oncostatin M, IL-6, and Stem Cell Factor mRNA Expression in Human Thymus Increases with Age and Is Associated with Thymic Atrophy
  7. Goronzy, Jörg J., and Cornelia M. Weyand. “T cell development and receptor diversity during aging.” Current opinion in immunology 17.5 (2005): 468-475.
  8. Page on nih.gov
  9. Page on nih.gov
  10. Page on sciencedirect.com)
  11. Johnson, P.L.,Yates, A.J.,Goronzy, J.J., and Antia, R.(2012). Peripheral selection rather than thymic involution explains sudden contraction in naive CD4 T-cell diversity with age. Proc.Natl.Acad. Sci.U.S.A. 109, 21432–21437. doi:10.1073/pnas.1209283110.
  12. Page on jimmunol.org
  13. Czesnikiewicz-Guzik, M.,Lee, W.W., Cui, D.,Hiruma, Y.,Lamar, D. L., Yang, Z.Z.,etal.(2008).T cell subset-specific susceptibility to aging. Clin. Immunol. 127, 107–118. doi:10.1016/j.clim.2007.12.002
  14. Di Rosa, Francesca, and Reinhard Pabst. “The bone marrow: a nest for migratory memory T cells.” Trends in immunology 26.7 (2005): 360-366.
  15. The Janus Head of T Cell Aging – Autoimmunity and Immunodeficiency
  16. Page on jimmunol.org
  17. Page on oxfordjournals.org
  18. Severe cytomegalovirus infection in apparently immunocompetent patients: a systematic review
  19. Bradley, Benjamin A. “Rejection and recipient age.” Transplant immunology 10.2 (2002): 125-132.
  20. Trzonkowski, Piotr, et al. “Immunosenescence increases the rate of acceptance of kidney allotransplants in elderly recipients through exhaustion of CD4< sup>+</sup> T-cells.” Mechanisms of ageing and development 131.2 (2010): 96-104.
  21. Page on nih.gov
  22. Page on nih.gov
  23. Human T Cell Aging and the Impact of Persistent Viral Infections
  24. Franceschi, Claudio, et al. “Inflamm‐aging: an evolutionary perspective on immunosenescence.” Annals of the New York Academy of Sciences 908.1 (2000): 244-254.
  25. Page on www.ideal-ageing.eu)
  26. Page on cell.com
  27. Steinmann, G. G., B. Klaus, and H‐K. MÜLLER‐HERMELINK. “The involution of the ageing human thymic epithelium is independent of puberty.” Scandinavian journal of immunology 22.5 (1985): 563-575.
  28. Montecino-Rodriquez, Encarnacion, Hyeyoung Min, and Kenneth Dorshkind. “Reevaluating current models of thymic involution.” Seminars in immunology. Vol. 17. No. 5. Academic Press, 2005.
  29. The Effect of Age on Thymic Function
  30. Steinmann, G. G. “Changes in the human thymus during aging.” The Human Thymus. Springer Berlin Heidelberg, 1986. 43-88.
  31. George, Andrew JT, and Mary A. Ritter. “Thymic involution with ageing: obsolescence or good housekeeping?.” Immunology today 17.6 (1996): 267-272.
  32. Page on nih.gov
  33. Page on nih.gov
  34. Bertho, Jean-Marc, et al. “Phenotypic and immunohistological analyses of the human adult thymus: evidence for an active thymus during adult life.” Cellular immunology 179.1 (1997): 30-40.
  35. Aw, Danielle, and Donald B. Palmer. “It’s not all equal: a multiphasic theory of thymic involution.” Biogerontology 13.1 (2012): 77-81.
  36. The Effect of Age on Thymic Function
  37. Physical Activity and Hemostatic and Inflammatory Variables in Elderly Men; Page on google.com=
  38. The influence of exercise on brain aging and dementia
  39. Page on dcconferences.com.au
  40. Olsson, Jadwiga, et al. “Age-related change in peripheral blood T-lymphocyte subpopulations and cytomegalovirus infection in the very old: the Swedish longitudinal OCTO immune study.” Mechanisms of ageing and development 121.1 (2001): 187-201.
  41. Page on tara.tcd.ie
  42. Page on nih.gov
  43. Page on nih.gov
  44. Page on nih.gov

* The antigens T (and B) cells respond to (their specificity) is a much more complicated story, and incomplete without considering cross-reactivity, another topic altogether.

Why does the thymus shrink?