Age-Related Effects on Thymic Output and Homeostatic T Cell Expansion Following Depletional Induction in Renal Transplant Recipients

Abstract

Thymic output and homeostatic mature cell proliferation both influence T-cell repopulation following depletional induction, though the relative contribution of each and their association with recipient age have not been well studied. We investigated the repopulating T-cell kinetics in kidney transplant recipients who underwent alemtuzumab induction followed by belatacept/rapamycin-based immunosuppression over 36 months posttransplantation. We focused specifically on the correlation between repopulating T cell subsets and the age of patients. Substantial homeostatic Ki67-expressing T-cell proliferation was seen posttransplantation. A repertoire enriched for naïve T (T_Naïve) cells emerged posttransplantation. Analysis by generalized estimating equation linear models revealed a strong negative linear association between reconstituting T_Naïve cells and advancing age. A relationship between age and persistence of effector memory cells was shown. We assessed thymic output and found an increase in the frequency of recent thymic emigrants (RTEs, CD4⁺CD31⁺) at 12-month posttransplantation. Patients under 30 years of age showed significantly higher levels of CD4⁺CD31⁺ cells than patients over 55 years of age pre and posttransplantation. IL-7 and autologous mature dendritic cells (mDCs) induced CD57^- cell proliferation. In contrast, mDCs, but not IL-7, induced CD57⁺ cell proliferation. This study establishes the relationship between age and thymic output during T-cell homeostatic repopulation after alemtuzumab induction.

Introduction

Depletional induction with T cell-specific antibodies effectively reduces the incidence of allograft rejection ^(1–4), and in particular has been effective in reducing rejection occurring when belatacept has been used as an alternative to calcineurin inhibitors ^(2,6). Alemtuzumab produces profound lymphocyte depletion that is effective for all T- and B-cell phenotypes^{(3–4, 6–7)}. Lymphocyte reconstitution post–alemtuzumab induction is characterized by a rapid repopulation of CD27^-IgD⁺ naïve B cells⁽⁹⁾, followed by slow T cell recovery^(6–8). Specifically, CD4⁺ cells repopulate with a CD28⁺ enriched repertoire, a transient increase in CD4⁺CD25⁺Foxp3⁺ cells^{(6, 8, 10)}, and a reduction in CD57⁺ cells⁽¹¹⁾—cells implicated in belatacept resistant rejection⁽¹²⁾. Lymphocyte depletion-induced lymphopenia triggers T-cell reconstitution through thymic output to generate T_Naïve cells and homeostatic proliferation of residual peripheral cells⁽¹³⁾ in patients posttransplantation^(6–8). T-cell reconstitution during lymphopenia may in part be due to T-cell receptor (TCR) stimulation by self-HLA molecules⁽¹⁴⁾. Alternatively, homeostatic repopulation may occur without TCR-signaling⁽¹⁵⁾, due to either cytokine (particularly, IL-7)-induced proliferation ⁽¹⁶⁾, or via B cell–dependent mechanisms⁽¹⁷⁾.

Thymic output gradually decreases with age as evidenced by a reduction in RTEs over time⁽¹⁸⁾. This immunological aging is characterized by shifting T-cell phenotypes, including a reduction in T_Naïve cells and an increase in terminally differentiated memory cells (T_EMRA)^(19–20). Both naïve and memory cells are involved in post-depletional repopulation ^{(6–8, 21)}. However, the extent to which T-cell repopulation is influenced by thymic output and peripheral homeostatic proliferation in depleted patients remains poorly understood. Specifically, the association between age and the balance between reconstituting T_Naïve and T effector memory (T_EM) cells has not been investigated in patients with immunosuppression, and this balance has been implicated in belatacept resistant rejection. Therefore, we longitudinally interrogated the kinetics of repopulating T_Naïve and T_EM cells in 40 kidney transplant patients who underwent alemtuzumab-induction and belatacept/rapamycin-based immunosuppression (ABR) to elucidate their association with age. We found substantial homeostatic T-cell proliferation in all patients with an increased generation of T_Naïve through thymic output in younger patients, and established a strong negative linear association between T_Naïve/T_EM cells and advancing age in these patients. Additionally, we demonstrated that autologous mDCs, but not IL-7 alone, induce CD57⁺ cell proliferation. These data provide evidence of age-specific patterns of thymic output and peripheral T-cell proliferation post-alemtuzumab induction.

Methods

Patients, Protocol Therapy, and Follow-Up

Forty patients enrolled in this study under an Institutional Review Board-approved, Food and Drug Administration–sponsored clinical trial following informed consent^(6–7). All patients were donor-specific alloantibody free at enrollment and assessed for alloantibody posttransplantation as described previously⁽⁶⁾. Patients received a kidney transplant from either living related (n=30) or deceased donors (n=10) and were treated with ABR regimen as previously reported^(6–7). The clinical outcomes of these patients have been recently reported⁽⁷⁾.

Cells, Flow Cytometry and Proliferation Assays

Peripheral blood mononuclear cells (PBMCs) were isolated from blood by centrifugation, and stored at −145°C until immune-assay. Frozen PBMCs were defrosted, and surface stained with monoclonal antibodies (see supplementary methods), as described in detail previously^(6–7). To detect Ki67 expression, surface-stained cells were permeabilized/fixed followed by intracellular staining with anti-Ki67 mAb.

As CD57 expression has been associated with belatacept resistant rejection⁽¹²⁾, the expansion of purified CD57^- and CD57⁺ cells was assessed using (see supplementary methods) a VPD450-based lymphocyte proliferation assay (n=6). Briefly, CD57^- and CD57⁺ cells were labeled with VPD450, and 2 × 10⁵ cells were added into a 12-well plate in the presence or absence of 1 × 10⁵ autologous mDCs (see supplementary methods and Figure S1). Cells incubated with IL-7 (50 ng/mL) were used as controls. In selected experiments, belatacept was added to the culture medium. Cells were incubated for 10 days, and stained with mAb followed by flow cytometry analysis.

Statistical Analysis

T-cell frequencies and absolute counts over time were summarized with descriptive statistics. To quantify the changes in frequencies or absolute counts over time and account for repeated measures for the same patient, we used longitudinal generalized estimating equation (GEE)-specifying linear models with an exchangeable correlation structure and times as covariates. The main effects of time were tested, and the phenotype measurements at the posttransplantation time points were compared with those pretransplantation. To protect against multiple comparisons, pairwise comparisons between the time points were made only if the overall omnibus test was statistically significant.

Next, we examined the functional form of age in association with T_Naïve cell frequencies over time based on a longitudinal linear model with GEE and restricted cubic splines of three knots for age placed at the 5th, 50th, and 95th percentiles of age. Because the non-linear association tests were not significant, age was treated as a continuous variable in all linear models with GEE to assess the relationship between age and T-cell reconstitution. These linear models included age, time, and interaction between age and time. The interaction between age and time was examined and removed from the final model if not significant. All statistical tests used a two-sided significance level of 0.05. Analyses were performed using Prism, R 3.5.1 (R Core Team, Vienna, Austria) and SAS 9.4 (SAS Institute, Cary, NC).

Results

Alemtuzumab induction produces profound lymphocyte depletion and prevents acute rejection

The clinical outcomes of these patients in this study have recently been reported in detail⁽⁷⁾. Briefly, alemtuzumab resulted in profound pan-lymphocyte depletion and clinical rejection was prevented. Twelve/nineteen patients who attempted to wean their oral immunosuppression were successfully maintained with belatacept monotherapy.

The T-cell reconstitution, particularly of CD4⁺ cells, was markedly delayed with significant CD4⁺ lymphopenia lasting for more than 36 months posttransplantation. Thus, repopulation was studied arising from a substantial lymphopenic nadir and progressing during a period of general clinical stability.

Homeostatic proliferation and thymic emigration contribute to T-cell reconstitution post-depletional induction

Profound lymphopenia is critical in driving T-cell reconstitution via thymopoiesis for newly generated naïve cells and homeostatic proliferation of peripheral, depletion-resistant T cells⁽¹³⁾. T cells upregulate intracellular Ki67 expression during homeostatic T-cell expansion-associated activation⁽²²⁾. As shown in Figure 1, an increase in homeostatic proliferation of CD4⁺Ki67⁺ and CD8⁺Ki67⁺ cells was observed between months 6 and 24 post-depletion when compared with baseline, consistent with initial reports from the first 20 live donor patients in this trial^{(6, 8)}.

Figure 1. Increasing homeostatic proliferation post-alemtuzumab induction over the course of renal allograft transplantation.

CD3⁺ T cells are segregated by CD4⁺ (left) and CD8⁺ (right) cells. T cell homeostatic proliferation is measured by intracellular staining for Ki67 expression. Patients demonstrate significant increasing homeostatic proliferation within 18–24 months posttransplantation. The box borders indicate the 75^th and 25^th percentiles, and the line within the box indicates the median. The upper and lower whiskers represent the 90^th and 10^th percentiles. The dots represent outliers. (* p≤0.01, ** p≤0.001, *** p≤0.0001)

Previous studies have defined CD31 as a marker for RTEs in evaluating thymic output⁽²³⁾. We longitudinally evaluated the dynamics of repopulating CD4⁺CD31⁺ cells in 28 patients during the first year. The frequency of CD4⁺CD31⁺ cells significantly increased post-induction (Figure 2a) consistent with prior reports^{(6, 8)}. These CD4⁺CD31⁺ cells were predominantly T_Naïve cells. In contrast, a significant reduction in the CD4⁺CD31^- subset, largely T_EM, was observed (Figure 2b). Furthermore, although CD4⁺CD31⁺ T_Naïve cells repopulated posttransplantation, their absolute numbers remained significantly below baseline (p=0.0493), and the repopulating T_EM cells (either CD31⁺ or CD31^-) remained markedly suppressed (Figure 2c).

Figure 2. Increased recent thymic emigrants during CD4⁺ cell reconstitution at 12 months post-alemtuzumab induction (n=28).

Shown are the percentage (a,b) and absolute numbers (c), of recent thymic emigrants, defined as CD31⁺ cells gating on CD3⁺CD4⁺ cells. Patients show a significant increase in the frequency of CD4⁺CD31⁺ cells following alemtuzumab induction, and the recent thymic emigrants are phenotypically CD45RA⁺CCR7⁺ naïve cells (a). In contrast, CD4⁺CD31^- cells, phenotypically defined as CD45RA^- CCR7^- memory cells, decrease significantly post-alemtuzumab induction (b). Repopulating CD4⁺CD31⁺ naïve cells were still significantly below baseline levels (p=0.0493). In contrast, the repopulation of effector memory cells in CD31⁺ or CD31^- populations was dramatically suppressed posttransplantation (c). The box borders indicate the 75^th and 25^th percentiles, and the line within the box indicates the median. The upper and lower whiskers represent the 90^th and 10^th percentiles. The dots represent outliers.

The reconstitution profile of T-cell subsets is influenced by age post-alemtuzumab induction

The depletional induction–induced lymphopenia triggers homeostatic T-cell reconstitution in patients undergoing allotransplantation^(1–4). It is believed that thymic output from a functional thymus and homeostatic proliferation of existing peripheral T cells are two major mechanisms for replenishing T cells^{(13, 18–20)}. The immunological aging is characterized by changes in T-cell subsets, such as a reduction in the naïve population and an increase in T_EMRA cells. As recently reported⁽⁷⁾, the frequency of CD4⁺ T_Naïve cells increased significantly over baseline with slow recovery of absolute T_Naïve but not T_EM counts to baseline by 24 months post-depletion (Figure S2a). A significant increase in the frequency of CD8⁺ T_Naïve cells with a concomitant decrease in the frequency of CD8⁺ T_EM cells occurred posttransplantation (Figure S2b). Though the absolute number of CD8⁺ T_Naïve cells returned to baseline by 18 months, CD8⁺ T_EM cells remained suppressed for up to 36 months.

To investigate the relationship between age and T-cell reconstitution post-depletion, we first fit longitudinal linear models created by GEEs to the frequencies of naïve CD4⁺ and CD8⁺ cells over time. The analysis of repopulating naïve B cells was used as a control. The tests of non-linear association were not significant, suggesting a linear relationship between age and T-cell frequencies. Our models revealed a strong negative linear correlation between the frequencies of CD4⁺ (p=0.0034) and CD8⁺ (p=0.0001) naïve cells and advancing age posttransplantation (Figure 3a). In contrast, no significant association was detected between advancing age and the frequency of repopulating naïve B cells in these patients (Figure 3b). There was no significant association between patient age and acute rejection episodes.

Figure 3. Functional form of age in relation to the frequency of repopulating naïve T cells.

(a) The association between a patient’s age and reconstituting CD4⁺ (left) and CD8⁺ naïve cells (right). Age is coded using restricted cubic splines with three knots at the 5^th, 50^th, and 95^th percentiles of age. The Y-axis represents the difference in the naive cell frequency between individuals of any age and individuals 47 years of age. The dashed lines indicate 95% confidence intervals. The knots are presented by dots. The frequencies of naïve cells are significantly different between the ages, and the association between naïve cell frequency and age is linear (p=0.0034 for CD4⁺ and p=0.0001 for CD8⁺). (b) The dynamics of repopulating naïve B cells post-alemtuzumab induction (left). The frequency of naïve cells changes significantly over time posttransplantation. The box borders indicate the 75^th and 25^th percentiles, and the line within the box indicates the median. The association between age and the frequency of naïve B cells (right). No association between age and the frequency of naïve cells was found during repopulation posttransplantation (p=0.506). The dashed lines indicate 95% confidence intervals. The dots represent the knots of restricted cubic splines at the 5^th, 50^th, and 95^th percentiles of age. (**** p≤0.00001)

To further characterize the association between advancing age and repopulating T-cell subsets, patients were segregated into three age groups: ≤ 30 (n=6), 31–54, and ≥ 55 (n=4) years of age. First, CD4⁺CD31⁺ RTEs were evaluated. The results showed that patients ≤ 30 years of age and 31–54 years of age had significantly higher frequencies (Figure 4a) and absolute counts (Figure 4b) of CD4⁺CD31⁺ cells over time when compared with patients ≥ 55 years of age. The repopulating CD4⁺CD31⁺ cells in patients ≤ 54 years of age were predominantly naïve cells (CD45RA⁺CCR7⁺), which was significantly higher than in patients ≥ 55 years of age (Figure 4). Indeed, CD4⁺CD31⁺ cells in patients ≥ 55 years of age were largely T_EM cells (Figure S3).

Figure 4. Dynamics and phenotype of repopulating CD31⁺CD4⁺ cells post-alemtuzumab induction in three age groups.

(a) Patients ≤ 30 years of age show higher frequencies of CD31⁺ cells than patients ≥ 55 years of age. In the ≤ 30 year and 31–54 year age groups, a significant increase in the frequency of CD4⁺CD31⁺ cells over time is observed when compared with patients ≥ 55 years of age posttransplantation (left). CD4⁺CD31⁺ cells prior to depletional induction contain large fractions of T_Naïve cells in the ≤ 30 and 31–54 year age groups but not in the ≥ 55 year age group (right). The frequency of T_Naïve cells remained unchanged following the repopulation of the CD4⁺CD31⁺ subset following depletional induction. (b) The absolute CD4⁺CD31⁺ cells and naïve cells (CD45RA⁺CCR7⁺) in CD4⁺CD31⁺ population repopulated in patients ≤ 30 but not ≥ 55 years of age. The circle represents each patient. (* p≤0.01, ** p≤0.001, *** p≤0.0001)

Further linear models using GEEs were constructed to assess the relationship between T-cell frequency or absolute count with age, time, and the interaction between age and time. Though the relationship between age and T-cell frequencies or absolute counts was assumed to be linear in the models (age was treated as a continuous variable), we used the age groups to create figures for ease of illustration. As shown in Figure 5a, naïve CD4⁺ and CD8⁺ cells demonstrated rapid repopulation in patients ≤ 30 and 31–54 years of age when compared with patients ≥ 55 years of age. The aging effect on the absolute counts and frequencies of naïve cells was significant, and the effect of age differed by time (table in Figure 5a). In contrast, the absolute counts of CD4⁺ and CD8⁺ T_EM subsets demonstrated delayed repopulation (Figure 5b). The effect of age on the absolute CD4⁺ T_EM cell counts was not significant. However, the effect of age on the frequency of CD4⁺ T_EM cells was significant, and this effect differed by time posttransplant (table in Figure 5b). The effect of age on the frequency of CD8⁺ T_EM cells was significant, and this effect differed by time (Figure 5b).

Figure 5. Dynamics of repopulating naïve and effector memory cells post-alemtuzumab induction in three age groups.

(a) The effect of age on the absolute counts and frequencies of CD4⁺ and CD8⁺ naïve cells is significant, and the effect of age differs at different time points posttransplantation. (b) The effect of age on the absolute CD4⁺ T_EM cell counts is not significant. The effect of age on the frequency of CD4⁺ T_EM cells is significant, and the effect of age differs at different time points. The effect of age on the frequency of CD8⁺ T_EM cells is significant, and the effect of age differs at different time points posttransplantation. The data are expressed as mean ± SD (standard deviation).

The CD57 expression is much less common on CD4⁺ cells, but its expression on CD4⁺, but not CD8⁺ cells, has been shown to be associated with belatacept resistant rejection in non-depleted patients⁽¹²⁾. In this study, we focused on the repopulating CD57⁺ and CD57^- subsets and the association with advanced age. We noted that patients ≤ 30 years of age showed higher absolute counts and frequencies of CD4⁺CD57^-PD-1^- cells at baseline when compared with other groups, and the reconstituting CD4⁺CD57^-PD-1^- cells were significantly higher in patients ≤ 30 years of age than in other age groups (Figure 6a). In contrast, the absolute counts and frequencies of CD4⁺CD57⁺PD-1^- cells were lower among all age groups at baseline and posttransplant. Though the absolute counts of CD4⁺CD57⁺PD-1^- cells remained low posttransplantation, no significant reduction in the frequency of these cells was observed. As shown in Figure 6b, patients ≤ 30 years of age demonstrated higher absolute counts and frequencies of the CD8⁺CD57^-PD-1^- subset when compared with other age groups pre- and posttransplantation. Additionally, the reduction in the frequency of CD8⁺CD57⁺PD-1^- cells over time by age group was significant. The reconstitution of absolute CD8⁺CD57⁺PD-1^- cells in patients ≤ 30 years of age appeared to be dramatically suppressed posttransplantation.

Figure 6. Dynamics of repopulating CD57^- and CD57⁺cells following depletional induction in three age groups.

(a) The frequency of repopulating CD57^- cells is significantly associated with age (p≤0.0004), and younger age shows more CD57^- cells at all time points. Similarly, the effect of age on the absolute count of the CD4⁺CD57^- subset is also significant (p≤0.0104). In contrast, the association between age and the absolute counts of the CD4⁺CD57⁺ subset is not significant due to prolonged suppression of cell repopulation within first 36 months posttransplantation. However, the effects of age on the frequency of CD4⁺CD57⁺ cells is significant (p≤0.0002), and older age shows a higher frequency at all time points when compared with the other two age groups. (b) The absolute count of the CD8⁺CD57^- subset is significantly associated with advancing age (p≤0.0033), and younger age shows higher absolute counts of the CD8⁺CD57^- subset. The effects of age on the frequency of CD8⁺CD57^- cells is significant (p≤0.0004), and younger age shows a higher CD8⁺CD57^- cell frequency. However, the interaction between age and time is not significant. The absolute counts of repopulating CD8⁺CD57⁺ cells is not significantly correlated with age and time. The effect of age on the frequency of repopulating CD8⁺CD57⁺ cells is significant (p≤0.0001), and older age shows a higher CD57⁺ cell frequency pre- and posttransplantation. However, the interaction between age and time is not significant. The data are expressed as mean ± SD (standard deviation).

Autologous mature dendritic cells induce CD57⁺ T cell proliferation

T-cell homeostasis is critical in restoring T-cell numbers following depletion-induced lymphopenia. IL-7 plays an important role in inducing CD57^- cell proliferation post-depletion⁽¹¹⁾. Increased Ki67-expressing cells post–depletion suggest homeostatic replication of T cells. Growing evidence suggests that DCs may play an important role in naïve and memory cell proliferation^(24–25). We therefore investigated the role of autologous mDCs in inducing CD57⁺ cell proliferation. As shown in Figure 7a, CD57^- but not CD57⁺ cells proliferated in the presence of IL-7 confirming recent report⁽¹¹⁾. Both CD4⁺CD57⁺ and CD8⁺CD57⁺ cells demonstrated significant proliferation in the presence of autologous mDCs when compared with unstimulated cells. The potential influence of B7 blockade in autologous mDC-induced T cell proliferation was investigated in the presence of belatacept. As shown in Figure 7b, belatacept inhibited mDC-induced expansion of CD57^- cells. The inhibitory effects of belatacept on CD57⁺ cell expansion were limited.

Figure 7. Proliferation of purified CD57⁺ and CD57^- cells in response to autologous mature dendritic cells.

(a) CD57⁺ and CD57^- cells are sorted from normal human PBMCs after labeling with anti-CD57 FITC antibody. Cells are labeled with proliferation dye VPD-450 and incubated with IL-7 or autologous dendritic cells (mDCs). Unstimulated cells are used as negative controls. The proliferation of cells is analyzed by flow cytometry. Unlike unstimulated cells, CD57^- cells demonstrate proliferation in the presence of Il-7 or mDCs, respectively. In contrast, CD57⁺ cells demonstrate barely detectable proliferation in the presence of IL-7. However, autologous mDCs effectively induce CD57⁺ cell proliferation. (* p≤0.01, ** p≤0.001). (b) Proliferation of purified CD57^- cells following coincubation with autologous mDCs is inhibited by belatacept (100 μg/mL). In contrast, limited inhibitory effects of belatacept on CD57⁺ cell proliferation are observed. The data are expressed as mean ± SD (standard deviation).

Discussion

Lymphocyte depletion is used in the majority of kidney transplant patients ⁽²⁶⁾. While known to have an inhomogeneous depletional effect on T cells of varying phenotypes, much less attention has been paid to factors altering repopulation. In this study, we evaluated T_Naïve and T_EM cell repopulation, given their starkly different sensitivities to belatacept, and the respective contributions of thymic output and peripheral homeostatic proliferation. We focused specifically on the effect of age on the relative importance of these two mechanisms for T cell replenishment, positing that age-related thymic involution could alter the repopulation dynamics and thus influence the post-repopulation repertoire. Indeed, we found that substantial phenotypical changes related to thymic output and homeostatic expansion in association with age occurred in these patients. Finally, we identified a potentially contributory mechanism for the preferential repopulation of T_Naïve cells in an in vitro T-cell proliferation model.

As is known, alemtuzumab profoundly depletes lymphocyte followed by delayed T-cell repopulation^{(1–4, 6–9)}. Phenotypically, repopulating T cells shift toward CD28⁺ T_Naïve cells, which are highly susceptible to targeting by belatacept⁽¹²⁾. In contrast, the recovery of T_EM cells is greatly suppressed by this regimen. Our findings confirm that the depletion and suppression of allo-specific T_EM cells with concomitant CD28-B7 costimulation blockade, most appropriate for T_Naïve cells, are major mechanisms for donor hyporesponsiveness ^(6–8).

T-cell reconstitution post-depletion is driven by lymphopenia and occurs via thymic output generating naïve cells and homeostatic proliferation, yielding mature T cells^(13,18–20). Using intracellular Ki-67 expression, a proliferation signature marker ⁽²²⁾, we observed a significant increase in CD4⁺ and CD8⁺ homeostatic proliferation. To define the role of thymic output in response to alemtuzumab-induced lymphopenia, we focused on the evaluation of CD4⁺CD31⁺ cells, a subset representing RTEs⁽²³⁾. We found that CD4⁺CD31⁺ cells underwent significant expansion posttransplant. These were phenotypically T_Naïve cells, suggesting that thymic output plays a critical role in generating naïve cells post-depletion. In contrast, we demonstrated a decrease in the proportion of CD4⁺CD31^- cells, which are primarily T_EM cells.

With regard to immunologic aging, thymic output decreases gradually leading to a reduction in the generation of naïve cells and increased dependence on mature T-cell division to maintain peripheral T-cells^(19–20). The influence of age-related thymic involution on the balance between reconstituting naïve and memory cells under belatacept-based immunosuppression has not been directly studied. This study demonstrated a strong negative linear relationship between CD4⁺/CD8⁺ T_Naïve cells and advancing age in patients on belatacept post-alemtuzumab induction. We found an increase in CD4⁺CD31⁺ cells and T_Naïve cells during repopulation in patients less than 30 years of age, suggesting that thymic output dominates in younger patients. In contrast, patients greater than 55 years of age had repopulating T-cell subsets that were largely T_EM and CD57⁺ memory cells suggesting the predominant role of mature cell division in maintaining the T-cell pool in older patients. Although older patients receiving alemtuzumab reconstituted with memory cells, we found no association between rejection and age, but given the low event rate, were not powered to directly assess this. Previously, we reported the development of hyporesponsiveness in donor-specific T cells with intact T-cell immunity specific for CMV/EBV and third-party allo-antigens in these patients^(6–8). Overall, we propose that the elimination of costimulation-resistant alloreactive memory cells by alemtuzumab-induction and subsequent activation-induced cell death during de novo allostimulation under belatacept and mTOR inhibition⁽²⁸⁾ may be the two mechanisms to suppress allospecific T-cell responses.

Lymphopenia can induce lymphocyte reconstitution via multiple processes, including thymopoiesis and homeostatic expansion^{(13, 18–20)}. Previous studies suggest that cytokines, including IL-7⁽¹⁶⁾ and IL-15⁽²⁴⁾, play a critical role in governing these two pathways. Furthermore, increased IL-7 production has been shown in patients post-alemtuzumab induction⁽¹¹⁾. This study confirmed previous findings that CD57^- but not CD57⁺ cells respond to IL-7⁽¹¹⁾ indicating that the IL-7/IL-7R pathway may selectively induce expansion of T_Naïve cells. Furthermore, we found significant T-cell expansion regardless of CD57 status in the presence of autologous mDCs, which provide not only TCR stimulation and costimulation⁽²⁹⁾, but also cytokines to elicit cell expansion^{(24, 30)}. Our studies demonstrate that belatacept reduces autologous mDC-induced CD57^- cell proliferation in normal individuals.

This study has limitations. There is significant heterogeneity in the patient treatments late in the study, making it difficult to determine whether clinical events are due to phenotypic differences in T-cells or variations in care. Also, the definition of RTEs in humans by flow cytometry is incomplete. Though our in vitro experiments suggest that more mature/senescent T cells do not proliferate in the presence of IL-7 alone, we do not investigate the mechanism of this insensitivity or the minimum set of factors necessary for the proliferation of these cells. The investigation of memory T-cell proliferation using cells directly from depleted patients is limited by the low peripheral cell numbers characteristic of the repopulation period. We have used T cells response in vitro to mDC circumvent the challenge of cell availability, but acknowledge that this does not completely recapitulate the conditions encountered in an immunosuppressed patient. Further study is warranted to directly elucidate specific mechanisms of the preferential repopulating naïve and memory cells in patients post-alemtuzumab induction. Finally, it may not be possible to impute these findings to thymoglobulin induction, as the impact of polyclonal preparations may influence thymic function more than the highly specific targeting of CD52 by alemtuzumab.

In conclusion, this study demonstrates a substantial role of both thymic output and peripheral homeostatic proliferation with a reciprocal, age-dependent differential contribution of each. We find a strong negative linear association between reconstituting T_Naïve/T_EM cells and age in alemtuzumab-depleted patients. In addition, we demonstrate the possible role of mDCs in inducing CD57⁺ memory cell proliferation. This study establishes the importance of age-related thymic output and homeostatic expansion of T cells after alemtuzumab induction.

Supplementary Material

Figure S1

Figure S1. Verification of mDCs. mDCs, derived from CD14⁺ cells following stimulation by IL-4 and GM-CSF for 7 days and TNF-α in final 24 hours of culture, demonstrate upregulation of CD83, CD40, HLA-DR, and CD54 (dark gray histogram) when compared with immature DCs (light gray histogram)

Figure S2

Figure S2. Dynamics of repopulating naïve and memory cells post-alemtuzumab induction. CD3⁺ cells are segregated by CD4⁺ and CD8⁺ cells, and the naïve (T_Naive) and effector memory (T_EM) cells are determined by the surface expression of CD45RA and CCR7. (a) Patients demonstrate significant increase in the frequency of CD4⁺ T_Naive cells but not T_EM cells. The T_Naive cell counts return to baseline levels between 18 and 24 months posttransplantation. In contrast, the recovery of T_EM cell counts are significant suppressed during repopulation. (b) CD8⁺ cells demonstrate a significantly increased frequency of T_Naive cells and a decreased frequency of T_EM cells posttransplantation. The T_Naive cell counts return to baseline levels between 12 and 18 months posttransplantation, and the absolute counts for repopulating T_EM cells are dramatically delayed until 36 months posttransplantation. The box borders indicate the 75^th and 25^th percentiles, and the line within the box indicates the median. The upper and lower whiskers represent the 90^th and 10^th percentiles. The dots represent outliers. (* p≤0.01, ** p≤0.001, **** p≤0.00001). Absolute counts have been previously reported.

Figure S3

Figure S3. Memory subsets of CD4⁺CD31⁺cells in patients > 55 years of age. The analysis of CD4⁺CD31⁺ cells in patients > 55 years of age is based on CD45RA and CCR7 expression. The largely fraction of CD4⁺CD31⁺ cells in this group of patients is characterized as effector memory cells.

Abbreviations:

T_Naïve

naïve T cells

T_EMRA

terminally differential effector memory T cells

T_EM

effector memory T cells

T_CM

central memory T cells

mAb

monoclonal antibody

mDCs

mature dendritic cells

mTOR

mammalian target of rapamycin

TCR

T cell receptor

RTEs

recent thymic emigrants

EBV

Epstein–Barr virus

IL-7

interleukin-7

IL-4

interleukin-4

GM-CSF

granulocyte-macrophage colony stimulating factor

PMBC

peripheral blood mononuclear cell

ABR

alemtuzumab induction belatacept/rapamycin maintenance regimen

TNF-α

tumor necrosis factor- α

GEE

generalized estimating equation