Non-human Primate Models of T-cell Reconstitution

Introduction

Non-human primates (NHP) have become an indispensable model in studying the common and dangerous human chronic infections, including HIV/SIV, Hepatitis C virus, and tuberculosis. More recently, we and others have used aged NHP to model human immune aging. Chronic infections and aging are both characterized by a significant depletion of defined lymphocyte subsets and the compensatory attempts to regenerate the immune system. As efficacious antiviral drugs and novel methods to improve and boost the immune system emerge, therapeutic immune regeneration has become a realistic goal in both the physiologic and pathologic settings. This article will summarize our current knowledge on this topic and will discuss future research directions as well as the potential and power of translational studies in non-human primate models of infection, aging and bone marrow transplantation.

Model discussion

Most of the data describing lymphocyte development, lymphocyte homeostasis and the age-related changes in homeostasis and function of the immune system, come from studies of the immune system in specific pathogen-free (SPF) housed rodents. This model is popular among immunologists for obvious reasons – defined immunogenetics, versatile genetic manipulation, a plethora of reagents and tools and the ability to expand and house large animal numbers. These advantages, along with a number of established infectious disease models make incisive mechanistic studies realistically achievable in inbred laboratory mice. Consequently, laboratory rodents handily outcompete any other model by a margin of nearly 10:1 with regard to the number of scientists using them, the number of the NIH awards employing the model, and, on the average, the number of speakers presenting the data from a given model at any major meeting.

Therefore, it is not surprising that rodent models have been invaluable in elucidating various fundamental facets of the structure and function of the immune system. When it comes to translating the findings into human health and clinical medicine, however, rodent models are less informative than large animal models. There are reminders that the results from rodent models do not automatically translate to humans ¹, including failure in humans of vaccine approaches that were validated in rodents ². Therefore, the vaccination efficacy in primates is frequently used as a stringent, “real-life”, validation of results obtained in rodents, and pharmaceuticals must be tested in both rodents and non-rodents for safety and toxicity prior to human trials [XXX}. Old world monkey species (mangabeys, macaques, baboons, mandrills, patas monkeys etc.) have the closest evolutionary relationship to humans among the approachable animal models. This makes them the most appropriate, and often the only, model for a variety of infectious diseases, including AIDS, Hepatitis C virus (HCV), emerging diseases and diseases caused by select (bioterrorism) agents^3-7. The organization and function of the immune system in these animals closely resembles that of humans ^8-12. Thus, NHP provide an invaluable in vivo model of pathogenesis and immunity to, infection, but also a key tool to develop vaccines and therapies to protect against such agents. Furthermore, this model is essential to investigate fundamental immunological questions relevant to the primate immune system, which cannot be ethically studied in humans.

In case of immunity to infection and of the aging of the immune system, there are additional inherent differences between rodent models and humans. First, the span of chronological aging is widely different between rodents and humans (differences in median and maximal lifespan between mice/rats and humans are ~30-fold). Second, inbred laboratory rodents, different from genetically diverse humans, are maintained under SPF) conditions which constitutes an enormous difference in the microbial environment that shapes the state of activation of the immune system. Finally, rodents and humans diverged from common ancestor ~210 million years ago. The non-human primate (NHP) model scores much better in all of these points (difference in lifespans only 3-fold; outbred genetics and exposure to environmental pathogens, unless deliberately manipulated, are comparable to humans; extremely close evolutionary relationship with ~30 million year divergence between the old world monkeys and humans ¹³; shared markers of T-cell phenotype, differentiation and function ^8-10), and thereby represents a highly desirable and relevant model of infection, immunity and immunogerontological conditions observed in humans.

Despite this potential, the scientific yield of immunological and, even more so, immunogerontological studies in NHP has historically been severely hampered by the lack of the necessary tools and expertise to perform and correctly interpret incisive studies of NHP immunobiology. Although human-specific reagents frequently cross-react with NHP cells, such cross-reactivity is often incomplete and/or inadequate, and many of these reagents have been applied without adequate optimization and validation in various NHP species. Moreover, procedures and protocols for longitudinal, systemic analysis of immunity were poorly developed in NHP, negating a key advantage of these animals as models of immune function and vaccine efficacy. This picture is rapidly changing with the latest phase of AIDS/SIV research, which brought state-of-the-art tools and their verification to the NHP model ^8,14,15}. Similarly, there has been a recent effort by our and other laboratories to characterize aging of the immune system in NHP and to popularize NHP as a model to study immune aging ^{10, 16-18}. More importantly from the standpoint of this review, induced and acquired immunodeficiencies and aging are all accompanied by depletion of certain lymphocyte subsets and are therefore ideal to study immune reconstitution in cases where the primary cause for such depletion can be retarded or eliminated.

A brief overview of T-cell homeostasis

For the purposes of this review, we shall define T-cell homeostasis as the maintenance of naïve and memory T-cell pool numbers and diversity and the ability to rebound to prior state after acute Ag challenge. Most of our knowledge on how homeostasis is maintained comes from rodent studies, with occasional contributions from larger animal or human models (rev. in ^19-22. Maintenance of the peripheral T-cell pools is an active and dynamic process ^23,24. Naïve and memory T-cells are believed to inhabit distinct, but somewhat overlapping, niches, which are ill-defined anatomically, but are likely to be discrete parts of T-cell microenvironment that can provide them with survival and trophic signals. Naive cells are predominant in secondary lymphoid organs (lymph nodes and spleen), while memory cells patrol the tissues. Moreover, experiments in young mice indicated that naïve and memory T-cells may be largely independently regulated ²⁵, although these results may be subject to alternative interpretations as well.

Life cycle of a peripheral T-cell begins at the time of its emigration from the thymus. Recent thymic emigrants (RTE) seed the periphery at a relatively constant rate (~1-2× 10⁶/day in mice, or ~1% of the total thymic cell content/day) throughout much of adulthood ²⁶. However, thymic involution begins almost immediately after birth in humans ²⁷ and a second decline in thymic production in humans may occur between the fourth and the fifth decade of life ²⁸. Parallel situation has been observed in mice, where the number of RTE drops by at least tenfold between young adult to 22-mo-old mice, paralleling the reduction of thymic cellularity and mass ^29,30. Despite that decline, in both mice and humans the thymus produces naïve T-cells longer than previously believed ^29,31. Which factors regulate this production? It appears that thymic output is regulated independently from the size of the peripheral pool^32,33, and it is likely that the output is regulated by the level of seeding by bone marrow precursors and the availability of thymic stroma. Upon emigration, RTE contribute to the pool of naïve peripheral T-cells, which have no pre-set life spans^34,35. Rather, naive CD8 and many of the naïve CD4 T-cells are maintained by trophic signals from interaction of their T-cell receptor TCR with self peptide-MHC (pMHC) complexes. Naïve T-cells also need the cytokine IL-7, and can survive indefinitely in serial transplantation experiments if these two contacts are provided^23,36. However, these cells are also regularly replaced by RTE in a random fashion, ensuring diversity of the naïve pool ³². Murine RTE proliferate faster than naïve peripheral T-cells in the first three weeks after export ³³, perhaps in order to maximize naïve T-cell diversity. After that, they equilibrate with other naïve T-cells. In vivo, naïve T-cells display very low levels of spontaneous (dubbed “homeostatic”) cycling. Homeostatic cycling is greatly increased in lymphopenia, where T-cells sense the “space” (most likely by sensing an excess of unused cytokines IL-7 and IL-15) and try to fill it by responding to signals that drive them into antigen (Ag)-independent homeostatic proliferative expansion (HPE) ^21,23. TCR:pMHC contact and the common γ chain cytokines IL-2, IL-7 and IL-15 have the potential to critically influence T-cell survival, maintenance and expansion. In mice, TCR:pMHC contact and IL-7 and IL-15 are believed to molecularly define “niches” or “space”, respectively, for conventional TCR αβ T-cells, and to allow HPE. It is unclear whether these signals are provided by a particular cell type (e.g. dendritic cells). It is also not clear whether the competition between T-cells for survival may be confined in the anatomic or histologic sense so that only certain cell types (e.g. dendritic cells) would be able to provide survival and trophic signals. IL-2 opposes some of the actions of IL-7 and IL-15³⁷ and its primary role appears to be in immune regulation and production and expansion of T regulatory cells ³⁸. While it is accepted that IL-2 production plummets with aging ³⁹, the evidence for age-related alterations in IL-7 or IL-15 levels is somewhat contradictory ^40-43, in part because it is extremely difficult to measure relevant IL-15 concentrations due to its highly localized mechanism of action. It is also controversial whether exogenously administered IL-7 exerts its main effects at the level of thymopoiesis ^40,44 or the level of peripheral T-cell maintenance ^42,45,46.

The memory compartment arises from the naive T-cells following Ag encounters, and is broadly divided into central memory cells, which self renew but lack acute activation markers and need a short lag phase to perform the optimal spectrum of effector functions, and effector/effector memory cells that are acutely activated and able to perform immediate effector function⁴⁷. Neither central memory nor effector/effector memory T-cells require specific pMHC contact for survival; they cycle and self-renew in vivo three to four-fold faster than naïve T-cells and are capable of vigorous proliferation during lymphopenia ²³. In mice, this proliferation, as well as memory T-cell maintenance, is dependent on IL-15, or, in its absence, on IL-7⁴⁸. Therefore, in the face of repeated pathogen challenges, homeostatic forces must balance Ag-driven expansion of naïve Ag-specific T-cells, generation of effector population, their subsequent contraction, and the recruitment of the surviving cells into the memory compartment. This simultaneously preserves T cell repertoire diversity to combat new pathogens and enables vigorous recall response to reinfection. In young and adult mice, numbers and diversity of peripheral T-cells are remarkably similar in different individuals ^23,24,49, suggesting that clonal expansion and contraction must be tightly regulated in adulthood (rev. in²⁴). However, this balance likely depends upon the stable naïve T-cell pool, which, in youth, critically depends upon production of new T-cells by the thymus ^29,33, and also on the absence of manifest pathogen challenge in SPF animals.

T-cell reconstitution in adult NHP in the presence or the absence of SIV infection and antiviral therapy

HIV, and its simian homologue, SIV, are known to cause one of the most devastating immunodeficiencies, characterized by massive and acute depletion of memory CD4 T-cells from the gut ^50-52, from which the immune system usually cannot recover ^53-55. While the naïve CD4 T-cells and the entire CD8 compartment remain largely unaffected by the virus in the course of the acute HIV/SIV infection, they both undergo excessive activation (perhaps due to microbial translocation⁵⁶) and turnover, followed by eventual exhaustion. Recent work has linked the onset of chronic AIDS with the inability of central memory CD4+ T cells to maintain themselves and to produce sufficient effector memory CD4+ T cells to replace those lost by ongoing infection and immune activation Most recent results suggest that this exhaustion occurs due to progressive conversion of central memory CD4 T-cells into effector memory, rather than due to activation of naïve CD4 T-cells that then become infected by the virus and killed ⁵⁷. The advent of highly active antiretroviral therapy (HAART) has rendered the prognosis of HIV/SIV less bleak, and has provided impetus to find ways to reconstitute the wounded T-cell compartment. NHP represent the best experimental model for HIV infection, and much of the knowledge on T-cell reconstitution was gained using this model. Given the importance of the thymus and the key role of T-cell maintaining cytokines, it is not surprising that the first therapeutic agents used were the common γ chain cytokines aimed to improve and/or protect thymic function and to improve maintenance of different T-cell subsets. Biology of these molecules was reviewed elsewhere in this volume, and we will therefore only briefly mention their key characteristics.

IL-2 and T-cell reconstitution

IL-2⁵⁸ was mostly used in NHP as a vaccine adjuvant. In one study IL-2 was administered in the form of an IL-2/Ig fusion protein to test its effect on peripheral T-cells in uninfected and SIV-infected monkeys ⁵⁹. In healthy young adult animals, a significant increase in CD4 T-cells was observed, and these cells were found to express CD25. Similar induction of CD25 expression and a transient increase in CD4 T-cells was also found in SIV-infected animals ⁵⁹. Importantly, these results closely paralleled the findings from HIV patients, where intermittent treatment with IL-2 led to an increase in CD4 T-cells ^60,61. However, most of the NHP cells expressing CD25, also expressed FoxP3 ⁶², suggesting strongly that IL-2 induces expansion of T regulatory cells (Treg) in primates. According to this evidence and the propensity of IL-2 to expand Treg cells in several models ³⁸, it is unlikely that this cytokine will be useful for immune reconstitution in situations other than treatment of autoimmune diseases.

IL-7 and T-cell reconstitution

IL-7 is the key differentiation factor for the very immature thymocytes (DN2-DN3 stage of the CD4^-CD8^- double-negative thymocyte subsets) ^63-65 and is secreted by resting epithelial cells, including the thymic stroma (rev in. ²¹. This tempted researches to postulate that IL-7 defects are at the heart of thymic involution ⁴⁰ and to attempt IL-7 therapy of thymic involution. Evidence that IL-7 improved thymic function was scant ^44,46. However, IL-7 is also critically involved in the maintenance of naïve (rev in. ^23,66 and, in some cases, memory, T-cell in mice ^48,67, and it can exert profound effects upon T-cell expansion and homeostasis in humans. This led to experimental administration of IL-7 to NHP ^42,45,68. Nugeyre et al. administered the cytokine twice a day over three weeks to two young uninfected and two young SIV-infected rhesus macaques, and noted an increase in both CD4 and CD8 T-cells that persisted over several weeks in both infected and uninfected animals, but returned to baseline by 11 weeks ⁶⁸. Naïve as well as memory T-cells increased within days of administration, and the number of recent thymic emigrants, as judged by the TCR recombination excision circle (TREC) values, decreased (particularly in infected animals), arguing against the authors’ interpretation that central (thymic) as well as peripheral effects are at play. In this and other studies ^42,68,69, IL-7 did not increase viral loads in infected animals and did not seem to promote B-cell expansion or tumorigenesis. This not only suggests that IL-7 is unlikely to induce enhanced viral replication, but also indicates that in primates (unlike in mice) IL-7 may not act on B-cells.

In a complementary experiment with cynomolgus macaques, IL-7 led to expansion of all T-cell subsets, and this too led to dilution of newly generated T-cells, suggesting that homeostatic proliferation, rather than thymic effects, are responsible for the IL-7 effects ⁴². Still, it is unclear whether the observation period was long enough to observe possible thymic effects. Similar effects were seen in both control, T-cell replete, and SIV-infected, moderately T-cell depleted animals – cell cycling was increased and T-cell numbers, both naïve and memory, went up ⁴². In another study using uninfected young adult macaques, IL-7 induced T-cell proliferation in a dose-dependent manner, and led to a marked phenotypic conversion of naïve T-cells to memory phenotype and function (the latter as assessed by TNFα and IFNγ secretion ⁴⁵. Remarkably, these T-cells reverted back to naïve phenotype after IL-7 therapy withdrawal ⁴⁵, suggesting that transient treatment did not convert these cells to the memory fate.

IL-7 was also used in a model of monkey reconstitution after autologous CD34⁺ cell transplantation ⁶⁹. Baboons were subjected to total body irradiation and autologous CD34⁺ cell transplantation and IL-7 was administered from weeks 6-10 post transplantation (twice/day). A remarkable increase in T-cell numbers was observed, which appeared to originate from peripheral expansion, rather than from an increase in thymic production ⁶⁹, as judged by TREC levels, and by the imaging and direct analysis of the thymic tissue.

IL-15 and T-cell reconstitution

The other critical cytokine in T-cell maintenance is IL-15. This cytokine is critically involved in the maintenance of memory T-cells and NK cells ^70,71 in mice. Likewise, it was shown that IL-15 is the most potent and perhaps critical cytokine for the maintenance of NK and memory T-cells in NHP ^72,73 and humans ⁷⁴. A recent study has examined the effects of IL-15 administration to SIV-infected or control rhesus macaques⁷⁵. The authors found a surprising and dramatic expansion of the effector memory cells, including the effector memory CD4 T-cells, which are precisely the population hit the hardest by the SIV and HIV in their respective hosts ⁷⁵. Effects of IL-15 were severely blunted in SIV-infected RM with uncontrolled SIV replication, but were robust in SIV-infected animals following HAART ⁷⁵. Furthermore, even in the uninfected rhesus macaques, the initial robust expansion was followed by waning of the response upon further continuous treatment, suggesting a feedback inhibition or exhaustion of the responsive cells ⁷⁵.

Summary of the effects of common γ chain cytokines on NHP T-cell reconstitution

Overall, among the three main common γ chain cytokines involved in T-cell maintenance, IL-2 is unlikely to help in general T-cell reconstitution protocols due to its confined action on Treg cells. With regard to population maintenance, IL-7 and IL-15 appear to dominantly target non-overlapping T-cell differentiation stages (although some biological overlap may exist, particularly at higher cytokine doses). Thus, IL-7 dominantly acts on T-cells recirculating between secondary lymphoid organs (naïve and central memory), whereas IL-15 operates on cells differentiating towards the effector memory stage, which are likely not to express the IL-7R nor to react to IL-7. While this provides hope in rejuvenation of each of the three subsets, a cautionary note is in order. Based upon the NHP studies, IL-7, at doses most effective for reconstitution, seems to convert many naïve T-cells into memory-phenotype T-cells⁴², which would be an undesirable effect in any model of reconstitution. By contrast, IL-15 seems to quickly lose its effect after the initial dramatic increase in effector memory cells. Therefore, both refinement and optimization of treatment protocols will be necessary to achieve optimal effects of these cytokines. In that regard, it would be interesting to test the above cytokines administered in complex with antibodies, which may significantly prolong their half-life ⁷⁶.

Keratinocyte growth factor (KGF) and T-cell reconstitution

KGF was discovered as a factor that induces proliferation of epithelial, but not endothelial or fibroblastoid cells ⁷⁷, with a particularly potent effect on keratinocytes. This molecule, formerly also known as the fibroblast growth factor 7 (FGF-7), acts on the early stages of T-cell development, allowing differentiation up to the point of action of IL-7 (rev in ⁷⁸. That prompted its use in various models of thymic regeneration. In the mouse model of bone marrow transplantation following total body irradiation, KGF was shown to protect thymic epithelial cells, reduce inflammation and enhance T-cell reconstitution ^79-81. Moreover, it was shown to increase thymic reconstitution after cyclophosphamide and dexamethasone treatments and to improve thymic output in old mice ⁸². Based upon these studies, KGF was tested for the ability to enhance T-cell reconstitution in a rhesus macaque model of autologous bone marrow transplantation ⁸³. Young adult monkeys were treated before, or before and after, total body irradiation and autologous CD34⁺ cell transplant. This treatment led to preservation of thymic architecture and an increase in naïve cells, which was significant in the lymph nodes ⁶⁹. A decline in the expression of the proliferation-associated marker Ki-67 in peripheral T-cells (characteristic of successful reconstitution of the T-cell compartment), correlated with the preservation of thymic stroma in monkeys receiving multiple KGF injections ⁶⁹. Thus, KGF provides hope as a therapeutic agent capable of affecting thymopoiesis, at least in young adult individuals.

Age-related disturbances in T-cell homeostasis in NHP

Aging of the immune system leads to the common but poorly understood state of immunodeficiency. Unlike the severe congenital or acquired immunodeficiencies, aging is not associated with a decrease in lymphocyte numbers, but rather with gradual shifts in lymphocyte responsiveness, population ratios and repertoire (rev. in ^84-87). Specific changes include: (i) decreased or impaired lymphocyte responsiveness, including impaired or blunted signaling along several branches of the BCR/TCR and co-stimulatory signaling pathways; (ii) an increase in memory and a decrease in naive lymphocyte representation, likely due to combined effects of decreased bone marrow and thymic output, inadequate maintenance of naïve lymphocytes in the periphery and life-long exposure to pathogens; and (iii) a decrease in lymphocyte diversity as exemplified by the appearance and persistence of B-and T-cell clonal expansions (BCE and TCE). Importantly, it has long been known that old animals and humans experience greater morbidity and mortality from infectious diseases and respond poorly to immunization with new Ag, including vaccination against the most frequent and severe pathogens of the elderly: influenza virus, pneumococcus and varicella-zoster virus ^88-91. Although this is certainly compounded by alterations in other organ systems, the above immunological decline is likely a primary underlying cause.

While defects were reported in several branches of adaptive and innate immunity, the most remarkable changes so far were detected in T and B lymphocyte adaptive responses. Often, reconstitution of the T-cell numbers and function can restore the overall immune competence in old animals ^18,92,93. Therefore, much of the ensuing discussion will focus upon T-cell numbers, phenotype and function. Age-related changes in the T-cell compartment of rhesus macaques follow the pattern observed in rodents and humans, with a possible exception that the changes may be occurring sooner and to a more dramatic extent in the outdoor-housed animals^10,16. By contrast, changes are more gradual, but also strongly significant, in the indoor-housed animals (ref. ¹⁸ and L. Cicin-Sain et al. in preparation), probably due to differences in pathogen exposure between the two housing situations. There is an early conversion, starting at birth, of naïve T-cells into memory phenotype T-cells ^10,16 and that is accompanied by an accumulation of memory cells that readily secrete proinflammatory cytokines ¹⁶. CD4/CD8 ratios also decrease, and there is evidence that the dominance of CD8 T-cells in the old age occurs due to their prolonged cycling and, perhaps, preferential survival, compared to the CD4 T-cells ¹⁶. Moreover, just like their human counterparts and unlike the laboratory rodents, macaque T-cells lose CD28 expression in the course of activation ¹⁶, and show a pronounced accumulation of the CD28^- subset with aging ^10,16. As aging progresses and the thymic output further declines, one would expect that paucity of naïve T-cells may prompt the remaining naïve T-cells to homeostatically proliferate in an attempt to maintain their numbers. We were recently able to document this phenomenon in aged rhesus macaques. We observed a strongly statistically significant inverse correlation between the naïve T-cell proliferation and the number of remaining naïve T-cells, particularly for the CD8⁺ subset (L. Cicin-Sain et al., in preparation). Therefore, aging in rhesus macaques exhibits all the expected characteristics seen in human T-cell aging to date.

Retardation of T-cell aging in NHP by caloric restriction

Caloric restriction (CR) has long been known to increase median and maximal life spans and to decrease mortality and morbidity in short-lived animal models, likely by altering fundamental biological processes that regulate aging and longevity (rev. in ⁹⁴. In rodents, CR was reported to delay the aging of the immune system ⁹⁵, and possibly even the retardation of thymic involution ⁹⁶, although results reported in that study could have alternative explanations. We investigated the effect of long-term caloric restriction upon T-cell subset distribution and function of old NHP ¹⁸ and J.N-Z et al., unpublished results). We noted a marked preservation of naïve T-cells, accompanied by an increase in TREC content and TCR repertoire diversity ¹⁸. It remained unclear whether CR delays thymic involution in primates, and experiments that are currently in progress, will hopefully resolve this issue.

Moreover, T-cells from CR old animals, but not form age-matched controls, proliferated better in response to polyclonal stimulation, and fewer of their memory cells were found to secrete pro-inflammatory cytokines (TNFα and IFNγ) in response to brief stimulation ¹⁸. Therefore, long-term CR was sufficient to retard the symptoms of T-cell aging in macaques, and has also improved the response of adult monkeys to vaccination (J. N-Z, et al., unpublished results).

Concluding comments

The above studies illustrate the power of the NHP model in validating findings from rodents that can be translated to primates, and in outlining findings where the divergent evolution limits the relevance of rodent models. More importantly, the studies above highlight a clear path towards the future studies in T-cell (and immune) reconstitution. To the best of our knowledge, KGF was the only treatment attempted so far in NHP to substantially affect thymic function; IL-7 and CR affected naïve T-cell turnover (albeit in a different manner), whereas IL-15 affected the effector memory T-cell subset. It remains to be determined whether the above treatments will bear their described effects in different models of T-cell depletion and whether their effects can be optimized by combining them. For example, it is particularly provoking to try to use KGF in aged animals and test whether the thymus can truly be rejuvenated. Likewise, it would be of utmost interest to test whether androgen blockade ⁹⁷ can affect thymus production in old NHP. Overall, development of treatments that can significantly increase T-cell reconstitution is coming of age, and one can look at the years in front of us with realistic optimism and excitement.