T cell regeneration after immunological injury

Over the past decade, intensive work has been done to further optimize the efficacy of already identified approaches and to identify alternative regenerative mechanisms (Fig.  2 ). In addition to IL-7, other cytokines have demonstrated efficacy in preclinical mouse models to restore thymic function and/or expand immune cells in the periphery following immune insults. Administration of IL-12 not only induces thymocyte proliferation through increased IL-7 and IL-2 signalling but also enhances engraftment and haematopoietic reconstitution after transplantation 185 , 186 . IL-15 can also boost immunity primarily by promoting NK cell, NKT cell and CD8 + T cell proliferation and function. Overall, these effects enhance reconstitution of these cell subsets and graft-versus-tumour responses in mouse models of allogeneic HCT 187 – 189 . IL-21 can enhance thymic function in young and aged mice. These effects are primarily mediated by the impact of IL-21 on DP thymocytes, which express high levels of IL-21 receptor after glucocorticoid-induced thymic atrophy, and by activation of the IL-21 downstream target BCL-6 (ref. 190 ). Administration of recombinant IL-21 improved thymic regeneration and reconstitution of the peripheral naive T cell compartment in different models of immune damage, including glucocorticoid-induced thymic atrophy, ageing and allogeneic HCT 190 – 192 . IL-22 can also mediate thymic regeneration 64 , 193 . Following thymic damage, the loss of DP thymocytes can trigger the production of IL-22 by thymic ILCs in an IL-23-dependent manner (Fig.  2 ). IL-22 then acts on TECs to promote their survival and proliferation through activation of STAT3 and STAT5 and expression of the downstream antiapoptotic molecule MCL1 (refs 64 , 193 ). Administration of recombinant IL-22 to sublethally irradiated mice or allogeneic HCT recipients can promote the recovery of thymic function and the development of new thymic-derived peripheral T cells 64 , 80 , 193 . ILCs play a fundamental role in thymic reconstitution not only via IL-22 production but also through production of receptor activator of nuclear factor-κΒ ligand (RANKL), a potent factor known for its fundamental role in TEC maintenance and maturation 194 – 196 . Recent studies have characterized the role of RANKL in the regeneration of the thymic microenvironment and T cell recovery in mouse models of allogeneic HCT. RANKL is expressed early after thymic damage by CD4 + thymocytes and, to a great extent, by lymphoid tissue inducer (LTi) cells 64 , 197 . RANKL acts on its cognate receptor, RANK, expressed on LTi cells, and induces upregulation of lymphotoxin-α (LTα). Following thymic damage, LTα can bind to LTβ receptor on thymic epithelial progenitor cells and TECs and promote their regeneration. Exogenous administration of recombinant RANKL boosts regeneration of thymic epithelial progenitor cells and TECs and improves T cell progenitor homing and de novo thymopoiesis. Overall, these effects lead to enhanced peripheral T cell reconstitution 197 . LTα can activate LTβ receptor on intestinal dendritic cells to induce IL-23 production, which in turn acts on intestinal ILCs to promote IL-22 production 198 . As cells producing RANKL, IL-22, LTα or LTβ are responsive to IL-7R signalling, which can promote their expansion and function 199 , 200 , it is possible that these molecules contribute to the regenerative effects mediated by IL-7 in the thymus. Although IL-7 has not been shown to directly regulate IL-22, it can regulate LTα 1 β 2 expression by LTi cells 201 , thus providing a mechanism by which IL-7R signalling integrates regenerative pathways. In addition, IL-7 can directly induce RANKL expression by T cells and aid in thymic regeneration 202 . In addition to its well-described role during thymus organogenesis 203 , 204 , BMP4 can promote thymic regeneration after thymic damage. Indeed, BMP4 produced by thymic endothelial cells drives thymic regeneration by binding to its receptor expressed on TECs and stimulating the upregulation of FOXN1 and its target genes 65 , 205 . Importantly, adoptive transfer of ex vivo expanded thymic endothelial cells improves thymic reconstitution after a sublethal dose of total body irradiation through the delivery of BMP4.

At present, there are no approved therapies to enhance T cell function in patients with lymphopenia. The development of such approaches would not only benefit patients whose immune system has been decimated by multiple cycles of chemotherapy and radiotherapy or by viral infections but could also improve T cell responses in other clinical settings. Cancer immunotherapy with immune checkpoint inhibitors is emerging as one of the most promising new treatments for a variety of solid and liquid malignancies. To be effective, these treatments rely on the presence of an adequate pool of T cells capable of recognizing specific tumour antigens 206 . Previous studies demonstrated that the limited response rate to checkpoint inhibitor therapy may be linked to a restricted TCR repertoire observed in the periphery of patients with cancer before therapy. Non-synonymous mutations and neoantigens are associated with clinical efficacy of immune checkpoint blockade 207 – 209 . Thus, treatments capable of improving immune functions and enhancing TCR repertoire diversity may have the potential to significantly extend the clinical benefit of immune checkpoint blockade. Therapeutic approaches that can rejuvenate the peripheral T cell pool will also be relevant for the treatment of elderly patients not only to enhance responses to pathogens but also to increase the efficacy of vaccines. The progressive expansion of peripheral TCR clonality and loss of specific T cell clones observed during ageing contribute to the moderate success rate of vaccination in elderly patients. However, important aspects should be taken in consideration when one is designing approaches to restore immunocompetence in aged individuals. Recent work performed in old mice and non-human primates demonstrated that additional barriers limit the impact of immune regeneration in the periphery even when thymic regeneration is achieved 210 . Although the thymuses of old mice treated with sex steroid ablation or KGF can be rejuvenated, this did not translate into increased frequencies of naive CD8 + T cells and naive CD4 + T cells in peripheral blood. Age-related intrinsic defects of RTEs 210 , 211 and a defective thymic stromal microenvironment 113 , 114 , 177 , together with reduced responses to homeostatic cytokines 212 , could explain the defective maintenance and function of naive T cells in old recipients. One promising, recently identified strategy to expedite immune reconstitution following HCT is the use of non-genotoxic conditioning approaches. The use of depleting antibodies, such as anti-CD117 and anti-CD45, which can recognize and eliminate HSCs and other haematopoietic cells in a targeted manner, allows remarkably efficient HSC engraftment while sparing non-haematopoietic cells, with minimal off-target toxicity 213 – 215 . This novel approach has the potential to reduce HCT-related toxicity, promote faster immune recovery and significantly improve patient clinical outcome.

In the healthy state, homeostasis of the immune system relies on a fine balance between cell production and cell death. During an infection, this dynamic equilibrium is altered to ensure pathogen clearance without unrestrained immune responses. Haematopoietic stem and progenitor cells (HSPCs) replenish immune cells by responding to infections either indirectly through sensing a depletion of downstream cells (a process termed ‘emergency haematopoiesis’) or directly through sensing pathogen-specific systemic inflammatory signals (such as cytokines or Toll-like receptor ligands) 21 . During acute inflammation, lineage commitment of HSPCs favours granulopoiesis over lymphopoiesis 22 . During sepsis , migration of thymic precursors from the BM to the thymus is decreased, leading to a depletion of early thymic progenitors and contributing to lymphopenia 23 . In addition, the thymus itself is a target organ of various pathogens, leading to thymic atrophy and lymphocyte depletion, which are both common features of infectious diseases 24 . Thymic haematopoietic and stromal compartments can both be directly targeted in viral and parasitic infections. CD4 + CD8 + double-positive (DP) thymocytes and their immediate precursors CD24 hi CD3 low CD8 + single-positive thymocytes are particularly vulnerable, whereas mature CD24 mid/low CD8 + SP cells are the most resistant thymic subsets during infection 25 , 26 . Although the precise mechanisms for infection-induced acute thymic involution remain to be further elucidated, stress-responsive hormones (such as glucocorticoids ), pro-inflammatory mediators (such as interferon-γ (IFNγ) and tumour necrosis factor (TNF)) and apoptosis pathways (such as those mediated by BAX, BCL-2, JUN amino-terminal kinase, p53, caspase 8 and caspase 9) have all been implicated 27 . DP thymocytes are particularly sensitive to the increased level of glucocorticoids occurring in infection, providing a potential mechanism by which infections induce depletion of this particular cell subset 28 . Functional changes in the thymic microenvironment are also observed following infection 24 . The thymic epithelium undergoes substantial phenotypic and functional changes after infection with the parasite Trypanosoma cruzi 29 and viruses such as HIV 30 , measles virus 31 and Zika virus 32 . For instance, localization of the TR5 and CK18 antigens restricted to medullary TECs (mTECs) and cortical TECs (cTECs), respectively, was altered in T. cruzi -infected mice along with increased extracellular matrix production by in vitro cultured TECs 33 , 34 . Other functional changes in the thymic epithelium, including cell cycle arrest, terminal cell differentiation and abnormal expression of cell adhesion and migration genes, were observed in experimental models of measles virus 31 and Zika virus 32 . The increases in the levels of pro-inflammatory cytokines, including IL-6, TNF and IFNγ 35 (which was previously shown to cause BM and thymic aplasia 36 – 38 ), together with other putative mechanisms, including direct damage of lymphoid tissues and apoptosis of lymphocytes, are presumably the primary causes of the profound lymphopenia observed in patients infected with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) 39 , 40 . Thymic atrophy during acute infection is usually a rapid yet transient response, with thymic rebound occurring within 2 weeks after infection 41 , 42 . Nonetheless, delayed or incomplete thymic recovery has been reported and is particularly associated with pathogens that cause chronic infections such as HIV 43 , 44 . Sustained immune activation and alteration in T cell homeostasis during chronic infections have been associated with progressive and possibly irreversible loss of thymic function as seen in advanced age. Thymopoiesis, as reflected by the levels of TCR excision circles (TRECs) and naive CD4 + T cells in the periphery (Box  2 ), is significantly impaired in HIV-infected patients 43 , 44 . In addition, the reductions in the ratios of CD4 + T cells to CD8 + T cells and naive T cells to memory T cells, the expansion in CD57 + CD8 + and CD28 – CD8 + senescent T cell populations, and the overall reduction in vaccine responsiveness have all been correlated with HIV-related disease progression and ageing 45 . This suggests that HIV-associated immunological abnormalities can induce early onset of immunosenescence and correlate with the higher risk of age-associated diseases typically observed in these patients 46 . The close association between T cell function and clinical response to cancer immunotherapy highlights the need for comprehensive methods to monitor a patient’s immune function and T cell receptor (TCR) repertoire changes and develop potential biomarkers predictive of clinical responses. Characterization of the main lymphocyte populations in peripheral blood is performed classically by flow cytometry and, more recently, by mass cytometry 270 using a combination of cell surface markers such as CD4, CD8, CD45RA, CD45RO, CCR7, CD62L and CD27 (ref. 271 ). Nonetheless, there are significant disparities among centres regarding the combination of markers to use and the tests to characterize reactivation of thymic function, TCR diversity and T cell function. Analysis of thymic output Analysis of recent thymic emigrants (RTEs) is performed in most immune monitoring studies. Several markers can be used to phenotypically characterize RTEs, when combined with CD45RA and CD45RO, including CD31, CCR7, PTK7 for CD4 + RTEs 272 and CD103 for CD8 + RTEs 273 , which can be used to enrich for RTEs carrying TCR excision circles (TRECs). However, not all CD31 + naive CD4 + T cells are RTEs, as T cells can undergo interleukin-7-driven homeostatic proliferation without downregulating their CD31 expression 146 . In addition, CD31 can be re-expressed by activated CD4 + T cells 274 . As an alternative, recent studies suggest that complement receptor 1 (CR1) and CR2 are novel markers enriched in RTEs and allow identification of this population with high purity 275 . Other approaches to monitor T cell output include molecular quantification of TRECs. Detectable by standard or real-time PCR, these non-replicating circles of DNA are enriched in RTEs 276 . However, division of T cells in response to cytokine stimulation in the periphery can complicate the assessment of RTEs and thymic output 141 , 217 . Imaging thymic mass by computed tomography scanning or magnetic resonance imaging has also been used to evaluate thymic regeneration in patients following antineoplastic therapy 66 , 70 and after haematopoietic cell transplantation 69 . A major limitation of this approach is its semiquantitative estimation of thymic output based on the size of the thymus, as these two parameters often do not follow a linear correlation. Analysis of TCR diversity A diverse TCR repertoire is generally associated with a favourable clinical outcome of immunotherapy for cancer. TCR diversity can be assessed by flow cytometry, which allows measurement of use of different TCR variable (V) genes at the protein level 277 , and spectratyping, which reveals clonal length polymorphism in the complementarity-determining region 3 of each TCR V gene family at a molecular level 278 . However, these methods estimate total TCR diversity of the whole αβ T cell population and do not measure the frequency of individual TCRs. The rapid advances in next-generation sequencing-based high-throughput TCR analysis have offered the opportunity to measure TCR diversity with increased resolution. As TCR sequencing is an indirect measurement of thymic function, other approaches should be used to gain more direct insights into the endogenous process of T cell development. In addition, measurement of total TCR diversity also comes with significant challenges due to the high heterogeneity and dynamism of naive T cells. Cytoablative treatments, such as radiotherapy, cell-depleting antibodies and chemotherapy, are designed to target malignant cells as well as the host haematopoietic system to allow successful HCT and markedly deplete all haematopoietic lineages, especially T and B cells 47 . Whereas many cytoablative therapies target highly proliferative cells while sparing the quiescent haematopoietic stem cells (HSCs) numerically, the resultant immunodepletion impairs haematopoietic cell function by selectively inducing premature senescence of HSCs through upregulation of the cyclin-dependent kinase inhibitors p19 ARF and p16 INK4a (refs 48 , 49 ). Recent studies demonstrate that chemotherapy induces fundamental changes in the expression of prohaematopoietic factors by the BM stromal niche, including downregulation of the Notch ligands Delta-like 4 (DLL4) and DLL1 by the vascular endothelium 50 . Consistent with the key role of DLL4 in sustaining common lymphoid progenitors and providing T-lineage specification before thymic entry 51 , reduced Notch signalling induces a premature myeloid skewing of HSPCs and can contribute to delayed recovery of lymphoid cells following chemotherapy. In addition to depleting haematopoietic and lymphoid precursors, cytoablative therapies induce prolonged deficiency in adaptive immunity by damaging the normal process of T cell development in the thymus and impairing B cell lymphopoiesis in the BM (Box  1 ). The reduction in BM progenitors and T cell development can lead to long-term suppression of thymic function even after a single sublethal irradiation dose and can accelerate age-associated thymic involution 52 , 53 . In one study, chemotherapy induced transient thymic atrophy in as many as 90% of patients 54 . Both alkylating agents, such as cyclophosphamide, and irradiation reduce the numbers of all thymocyte subsets in mice, with DP thymocytes and their immediate progenitors CD4 – CD8 – double-negative thymocytes being particularly sensitive to such insults 55 , 56 . Further contributing to impaired thymopoiesis, cytoablative agents damage TECs 57 . Irradiation and cyclophosphamide both deplete cTECs and mTECs which are important for, respectively, positive and negative selection of developing T cells, and this interferes with the generation of a broadly reactive TCR repertoire 58 , 59 . Activation of signal transducer and activator of transcription 3 (STAT3) signalling has been linked to the induction of apoptosis in TECs after irradiation 60 . However, more recent studies show that STAT3-mediated signalling is essential for growth and architectural organization of mTECs but is dispensable for the biology of cTECs 61 , 62 . Coinciding with their higher rate of proliferation 57 , the greatest numerical reduction occurs in the subset of TECs expressing the highest levels of MHC class II, particularly the tolerance-inducing AIRE + MHC-II hi mTECs 63 . This imbalance in TEC subsets persists even after thymic cellularity is restored and may promote the development of autoimmunity 58 . Importantly, certain stromal subsets, such as innate lymphoid cells (ILCs) and endothelial cells, are relatively resistant to cytoreductive therapies and play an important role in thymic regeneration by secreting factors, such as IL-22 and bone morphogenetic protein 4 (BMP4), which promote the survival, proliferation and maintenance of TECs 64 , 65 . Full, albeit delayed, recovery of thymic function is possible up until middle age, but the peripheral naive repertoire is never fully restored in older patients 66 . Peripheral T cell depletion, especially of CD4 + T cells, is the main cause of clinical immunodeficiency observed in patients with cancer 67 . T cell recovery following cytoablative therapy is predominantly achieved through two mechanisms: de novo production in the thymus, particularly for CD4 + T cells, or homeostatic expansion of peripheral T cells, preferentially for CD8 + T cells 68 , 69 (Fig.  1 ). The relatively rapid CD8 + T cell recovery through extrathymic clonal expansion 70 , 71 is typically associated with limited breadth of the TCR repertoire and diminished immune responses 72 . Although allogeneic HCT offers a potential cure for various malignant and non-malignant disorders, its wider application is limited by the significant morbidity and mortality associated with GVHD and prolonged post-transplantation immunodeficiency 73 . In GVHD, donor alloreactive T cells in addition to targeting the host gastrointestinal tract, liver and skin also directly destroy primary lymphoid organs (BM and thymus) and delay immune reconstitution after allogeneic HCT. Both mouse models and clinical studies have demonstrated that GVHD targets the BM 74 , and the resultant medullary aplasia delays donor-derived haematopoiesis, especially of the lymphoid lineage 75 . IFNγ and TNF disrupt donor haematopoiesis both directly and indirectly through their effects on host BM stromal cells, including osteoblasts and mesenchymal stem cells. As a result, acute GVHD is associated with a significant reduction in the numbers of residual and de novo B cell precursors, in the numbers of mature B cells and in antibody production 76 , 77 (Box  1 ). During acute GVHD, the thymus is exquisitely sensitive to damage by alloreactive T cells, and its acute atrophy reflects the precipitous contractions of the lymphocytic and stromal compartments 78 , 79 . Mouse models of acute GVHD have demonstrated that, in addition to the reduced number of early thymic progenitors migrating from the BM to the thymus, reduction in thymic cellularity is primarily due to loss of the large DP thymocyte population via glucocorticoid-independent apoptotic cell death 78 . Within the stromal cell subset, TECs are targets of GVHD-mediated apoptosis. TECs also act as antigen-presenting cells and prime alloreactive T cells directly through their intrinsic expression of MHC class I and class II molecules, such that depletion of host-derived professional antigen-presenting cells does not prevent activation of alloreactive donor T cells and thymic injury 79 . This raises the possibility that GVHD is restricted to the thymus and remains subclinical despite detrimental consequences for T cell reconstitution. In addition, alloreactive T cells eliminate IL-22-producing thymic group 3 ILCs (ILC3s), resulting in impaired thymic recovery 80 . Moreover, donor alloreactive CD8 + T cells preferentially target tolerance-inducing mature mTECs, thereby impairing negative selection and allowing de novo generation of autoreactive CD4 + T cells, linking alloimmunity during acute GVHD to the development of autoimmunity during chronic GVHD 81 , 82 . These studies underline the importance of functional preservation of TECs as well as numerical restoration of lymphocytes as the goals for immune-regenerative strategies following allogeneic HCT. Clinically, GVHD-induced immunosuppression has been linked to decreased thymic output of naive T cells, as measured by TRECs, and a distorted TCR repertoire 83 – 86 . Paradoxically, pharmacological immunosuppressants, including corticosteroids and cyclosporine A, used in allogeneic HCT to prevent allograft rejection, as well as GVHD prophylaxis and/or treatment, can induce thymic involution, deplete thymocytes and ablate AIRE + MHC-II hi mTEC subsets, further contributing to thymic damage 58 , 87 . These findings highlight the multifactorial nature of post-transplantation immune disorders and represent the clinical challenges in preventing and treating GVHD after allogeneic HCT. The functionality of the immune system progressively declines with age 88 , contributing to increased incidence of infections, inadequate vaccine responses and decreased immunosurveillance of malignant cells observed in the elderly population 89 , 90 . It is estimated that only ~30–40% of elderly people are capable of mounting sufficient immune responses to the influenza vaccine 91 , 92 . Ageing impairs the normal process of thymic and BM lymphopoiesis at multiple levels. With age, there is an accumulation of HSCs with reduced homing and engraftment capacity 93 – 95 . In patients receiving allogeneic HCT, younger donor age has been associated with faster immune reconstitution 96 . Although several intrinsic changes in HSCs and lymphoid progenitors have been identified (including myeloid skewing 97 , 98 , defects in DNA repair 99 , 100 , epigenetic alterations 101 and loss of cell polarity 102 ), substantial changes in the BM stromal microenvironment also contribute to defective lymphopoiesis 103 . Transplantation of old HSCs into a young microenvironment is sufficient to partially reverse myeloid skewing 104 , 105 . As T cell development depends on the constant output of T cell progenitors from the BM, defective lymphopoiesis, along with impaired developmental potential of intrathymic progenitors 106 , results in a significant reduction in the number of intrathymic lymphoid progenitors contributing to the reduced T cell output observed in older individuals 97 , 106 . The progressive decline of thymus function during ageing represents one of the most important causes of immune degeneration in the elderly population 107 . However, the precise kinetics of this chronic physiological process is still under debate. Compared with the mouse thymus, which undergoes a progressive reduction in its volume with age, the human thymus remains almost unchanged in size under normal circumstances and, instead, it is characterized by profound perturbation of the thymic stromal cell microenvironment, including loss of epithelial cells, increased volume of the perivascular space and progressive replacement of healthy tissue with adipose tissue 108 . The thymic stroma is one of the main targets of the effects of ageing, as demonstrated by HCT and parabiosis experiments in mice with different age-mismatched donors 109 , 110 . For example, intrathymic injection of early thymic progenitors derived from young mice into old mice is not sufficient to restore normal thymic lymphopoiesis 111 . At a molecular level, several studies have revealed extensive transcriptional changes in mouse thymic stromal cells as early as 3 months of age, particularly in genes associated with cell cycling and inflammatory responses 112 – 114 . Although ageing significantly erodes its functionality, the thymus maintains a proportion of functional cortical and medullary regions and active thymopoiesis. Indeed, T cell output, as measured by TREC level in the peripheral blood, is maintained in older individuals, albeit substantially reduced 115 , 116 . In addition, recent studies demonstrate that the involution of thymic tissue is not as dramatic as previously reported and that residual thymic tissue can be detected by computed tomography scanning in individuals up to 70 years of age 117 . As the thymus involutes, the production of T cells is progressively impaired, and with age the naive T cell pool is increasingly maintained by homeostatic expansion 3 , 118 , 119 . Mathematical modelling suggested that thymic T cell export declines exponentially over time with a half-life of 15.7 years; therefore, by the age of 55 years only 5% of naive T cell production is derived from thymic export 120 . More recently, the average TREC content of the naive T cell population was used to estimate thymic T cell output. This study demonstrated that up to 90% of naive T cells in healthy human adults are generated through proliferation 118 . By measuring the average TREC content in naive T cells as well as the turnover of T cells, this work also showed that, unlike in humans, the naive T cell pool in mice is almost exclusively sustained throughout life by thymus output 118 . The decline in naive T cell production results in a shift towards an oligoclonal pool of memory T cells and an almost linear decrease of T cell diversity with age 119 , 121 . These effects are particularly pronounced after the age of 60 years 3 , 122 . Some long-lived individuals (average age of 82 years) display a significantly higher percentage of naive CD4 + T cells, decreased abundance of expanded clones and increased TCR diversity compared with a younger cohort, suggesting that these key immune parameters may represent hallmarks of longevity 122 . Mathematical modelling studies demonstrate a significant inverse correlation between thymic export and the incidence of both infectious diseases and cancer 123 . During the coronavirus disease 2019 (COVID-19) pandemic, being older than 65 years was identified as a risk factor for morbidity and death from COVID-19 (ref. 124 ). The age-associated decline of thymic function significantly impairs and delays the endogenous process of thymic repair after immunological insults, resulting in a prolonged recovery time following common cancer cytoreductive therapies 12 , 15 . This is particularly problematic in the HCT setting, which is increasingly used in older patients. An inverse relationship exists between the transplant recipient’s age and T cell recovery after transplantation 9 , 125 . Insufficient recovery of thymopoiesis has been directly correlated with an increased risk of opportunistic infections, leukaemia relapse and adverse clinical outcome 126 , 127 .

Over the past few years, several approaches have been proposed to enhance immune function through recovery of the T cell pool (Fig.  2 ; Table  1 ). These strategies include the stimulation of T cell development and expansion using cytokines, such as IL-7, IL-12 and IL-21; the administration of cytokines and growth factors, such as stem cell factor (SCF; also known as KITLG), keratinocyte growth factor (KGF; also known as FGF7), IL-22 and FMS-like tyrosine kinase 3 ligand (FLT3LG); the modulation of hormone levels by suppression of sex steroids or by administration of thymosin-α1, growth hormone (GH; also known as somatotropin), insulin-like growth factor 1 (IGF1) and ghrelin; the adoptive transfer of lymphoid progenitors such as precursor T cells and ex vivo expanded thymus-derived endothelial cells; and the use of artificial BM or thymus-like grafts. Fig. 2 Simplified overview of T cell generation with regenerative strategies after immune injury. T cell development begins when T cell progenitors, originating from common lymphoid progenitors (CLPs) in the bone marrow, migrate into the thymus and progress through a series of well-characterized developmental steps. Thymocytes go through the double-negative (DN; CD4 – CD8 – ) and double-positive (DP; CD4 + CD8 + ) stages to form single-positive (CD4 + CD8 – or CD4 – CD8 + ) T cells. During this process, approximately 95% of developing thymocytes produced daily are deleted through β-selection, positive selection and negative selection, resulting in the formation of self-restricted and self-tolerant naive CD4 + and CD8 + single-positive cells that can exit the thymus and migrate to peripheral lymphoid organs. Approaches to enhance T cell recovery act at multiple levels. Factors and approaches such as interleukin-7 (IL-7), IL-12, IL-15, IL-21, FMS-like tyrosine kinase 3 ligand (FLT3LG), growth hormone (GH), insulin-like growth factor 1 (IGF1), sex steroid ablation (SSA), thymosin-α1, stem cell factor (SCF), administration of precursor T cells (pre-T cells) and delivery of ex vivo-generated thymic epithelial cells (exTECs) primarily promote recovery of the haematopoietic compartment. By contrast, keratinocyte growth factor (KGF), IL-22, receptor activator of nuclear factor-κB ligand (RANKL), IGF1, lymphotoxin-α (LTα) and bone morphogenetic protein 4 (BMP4) produced by ex vivo-generated endothelial cells (exECs) enhance reconstitution of the thymic stromal compartment. Question marks denote approaches where the effects on specific targets are not fully understood. ETP, early thymic progenitor; CMP, common myeloid progenitor; cTEC, cortical thymic epithelial cell; DC, dendritic cell: ILC, innate lymphoid cell; HSC, haematopoietic stem cell; MPP, multipotent progenitor; MSC, mesenchymal stromal cell; mTEC, medullary thymic epithelial cell; RTE, recent thymic emigrant. Table 1 Approaches to enhance T cell recovery, their targets and progress towards the clinic Approach Target cells Clinical trials Refs Cytokines IL-7 HSPCs, thymocytes, peripheral T cells NCT00477321 , NCT01190111 , NCT01241643 , NCT00839436 , NCT00684008 136 – 144 , 216 , 217 IL-12 Thymocytes, HSPCs? Not currently in clinical trials 185 , 186 IL-15 NK cells, NKT cells, CD8 + T cells Not currently in clinical trials 187 – 189 IL-21 Thymocytes, HSPCs Not currently in clinical trials 190 – 192 IL-22 TECs Not currently in clinical trials 64 , 80 , 193 RANKL TECs Not currently in clinical trials 197 , 218 , 219 Growth factors KGF TECs NCT01233921 , NCT03042585 , NCT02356159 , NCT00593554 , NCT01712945 150 – 155 , 157 , 159 , 220 – 223 FLT3LG BM HSPCs, thymocytes Not currently in clinical trials 224 – 226 IGF1 TECs Not currently in clinical trials 227 , 228 SCF Thymocytes Not currently in clinical trials 229 – 231 Hormones and hormone-like mediators Thymosin-α1 Thymocytes NCT00580450 164 , 232 GH and ghrelin TECs, thymocytes NCT00071240 , NCT00287677 , NCT00119769 , NCT00050921 180 – 182 , 184 , 233 – 235 Sex steroid ablation TECs, BM HSPCs, thymocytes NCT01746849 , NCT01338987 55 , 58 , 167 , 169 – 173 , 175 , 176 , 236 – 240 Cell-based approaches Precursor T cells TECs, thymocytes Not currently in clinical trials 241 – 246 Ex vivo expanded endothelial cells TECs Not currently in clinical trials 65 Ex vivo expanded MSCs HSPCs, TECs, T cells Not currently in clinical trials 247 , 248 Ex vivo generated TEC graft TECs, thymocytes Not currently in clinical trials 249 – 254 Injectable thymus-like scaffolds CLPs, peripheral T cells Not currently in clinical trials 253 , 255 – 257 BM, bone marrow; CLP, common lymphoid progenitor; FLT3LG, FMS-like tyrosine kinase 3 ligand; GH, growth hormone; HSPCs, haematopoietic stem and progenitor cells; IGF1, insulin-like growth factor 1; KGF, keratinocyte growth factor; IL, interleukin; MSC, mesenchymal stromal cell; NK, natural killer; NKT cells, natural killer T cells; RANKL, receptor activator of nuclear factor-κB ligand; SCF, stem cell factor; TEC, thymic epithelial cell. T cell development begins when T cell progenitors, originating from common lymphoid progenitors (CLPs) in the bone marrow, migrate into the thymus and progress through a series of well-characterized developmental steps. Thymocytes go through the double-negative (DN; CD4 – CD8 – ) and double-positive (DP; CD4 + CD8 + ) stages to form single-positive (CD4 + CD8 – or CD4 – CD8 + ) T cells. During this process, approximately 95% of developing thymocytes produced daily are deleted through β-selection, positive selection and negative selection, resulting in the formation of self-restricted and self-tolerant naive CD4 + and CD8 + single-positive cells that can exit the thymus and migrate to peripheral lymphoid organs. Approaches to enhance T cell recovery act at multiple levels. Factors and approaches such as interleukin-7 (IL-7), IL-12, IL-15, IL-21, FMS-like tyrosine kinase 3 ligand (FLT3LG), growth hormone (GH), insulin-like growth factor 1 (IGF1), sex steroid ablation (SSA), thymosin-α1, stem cell factor (SCF), administration of precursor T cells (pre-T cells) and delivery of ex vivo-generated thymic epithelial cells (exTECs) primarily promote recovery of the haematopoietic compartment. By contrast, keratinocyte growth factor (KGF), IL-22, receptor activator of nuclear factor-κB ligand (RANKL), IGF1, lymphotoxin-α (LTα) and bone morphogenetic protein 4 (BMP4) produced by ex vivo-generated endothelial cells (exECs) enhance reconstitution of the thymic stromal compartment. Question marks denote approaches where the effects on specific targets are not fully understood. ETP, early thymic progenitor; CMP, common myeloid progenitor; cTEC, cortical thymic epithelial cell; DC, dendritic cell: ILC, innate lymphoid cell; HSC, haematopoietic stem cell; MPP, multipotent progenitor; MSC, mesenchymal stromal cell; mTEC, medullary thymic epithelial cell; RTE, recent thymic emigrant. Approaches to enhance T cell recovery, their targets and progress towards the clinic BM, bone marrow; CLP, common lymphoid progenitor; FLT3LG, FMS-like tyrosine kinase 3 ligand; GH, growth hormone; HSPCs, haematopoietic stem and progenitor cells; IGF1, insulin-like growth factor 1; KGF, keratinocyte growth factor; IL, interleukin; MSC, mesenchymal stromal cell; NK, natural killer; NKT cells, natural killer T cells; RANKL, receptor activator of nuclear factor-κB ligand; SCF, stem cell factor; TEC, thymic epithelial cell. Here, we focus on immune-boosting strategies that have been translated into clinical studies and provide a brief update on novel approaches that have shown regenerative potential in preclinical models. We put particular emphasis on approaches that can promote de novo T cell formation through the regeneration of thymic function, which is the primary mechanism to generate a pool of naive T cells with a diverse TCR repertoire. We also provide a brief summary of the different methods used to estimate thymic function and the dynamics of T cell reconstitution in Box  2 . IL-7 is classified as a type 1 short-chain cytokine crucial for the development of innate and adaptive immune cells 128 . It is secreted mainly by non-haematopoietic cells, including epithelial cells and fibroblasts in the thymus, BM stromal cells, lymphatic endothelial cells, fibroblastic reticular cells and enterocytes 129 – 131 . IL-7 is particularly important for the differentiation of T and B cells from common lymphoid progenitors and for the maintenance and survival of mature T cells. IL-7 receptor (IL-7R) is a heterodimer of two chains: IL-7Rα (also known as CD127) and cytokine receptor common subunit-γ (also known as CD132 or IL-2RG). The γ-chain is expressed by all haematopoietic cell types, whereas IL-7Rα is expressed mainly by developing B and T cells, naive and memory T cells, NKT cells, ILC2s and ILC3s. The crucial role of IL-7 in lymphopoiesis is demonstrated by the development of severe combined immunodeficiency disease in patients carrying mutations affecting the α-chain 132 , 133 or the γ-chain of IL-7R 134 . While studies in Il7 –/– mice showed that IL-7 is a non-redundant cytokine for both T and B cell lymphopoiesis, B cell development in humans does not appear to require IL-7, as B cells are maintained in patients with severe combined immunodeficiency disease with mutations in the gene encoding IL-7Rα 132 . The important role of IL-7 in T cell biology is also supported by the inverse correlation between circulating IL-7 levels and peripheral T cells observed in patients with lymphopenia 135 . The degree of available IL-7 controls the size of the peripheral T cell pool and plays an important role in regulating overall T cell homeostasis. The effects of exogenous administration of IL-7 on immune reconstitution have been widely investigated 128 . Several preclinical mouse models have demonstrated the beneficial effects of exogenous IL-7 in promoting immune reconstitution through thymus-dependent and thymus-independent mechanisms. In the setting of HCT, exogenous IL-7 accelerates the reconstitution of donor-derived thymocytes and the peripheral T cell pool, leading to enhanced T cell recovery after both syngeneic and allogeneic HCT 135 – 139 . A phase I/IIa dose-escalation study ( NCT00477321 ) of repeated administration of a glycosylated recombinant human IL-7 (rhIL-7; CYT 107) in HIV-1-infected patients demonstrated that rhIL-7 treatment was safe, well tolerated and transiently promoted the expansion of naive and memory CD4 + and CD8 + T cells, and decreased the proportion of exhausted PD1 + T cells 140 . Importantly, rhIL-7 therapy also increased the numbers of CD4 + recent thymic emigrants (RTEs), the signal joint to β TREC ratio and TCR repertoire diversity in some participants, effects that imply enhanced thymic activity (Box  2 ). Subsequently, two other clinical trials ( NCT01190111 and NCT01241643 ) demonstrated that repeated doses of rhIL-7 were well tolerated and resulted in sustained CD4 + T cell numbers in the majority of HIV-infected participants 140 – 143 . Similarly, in a phase I/IIa dose-escalation trial ( NCT00839436 ) in patients with idiopathic CD4 + lymphopenia at risk of disease progression, rhIL-7 led to an increase in the number of circulating CD4 + and CD8 + T cells and tissue-resident CD3 + T cells in the gut mucosa and BM. Importantly, enhanced thymopoiesis, measured by TRECs, was observed only in the youngest patients of the cohort (aged 23 and 34 years) 144 . In a phase I clinical trial ( NCT00684008 ) that evaluated the immune-regenerative properties of rhIL-7 in patients receiving T cell-depleted allogeneic HCT 138 , rhIL-7 induced a rapid increase in peripheral CD4 + and CD8 + T cell numbers. While the estimated half-life of rhIL-7 in this study was 9–35 hours, the biological effects on T cell numbers persisted for several weeks after the circulating levels of IL-7 returned to the baseline. In addition, although rhIL-7 administration resulted in increased numbers of RTEs only in some young patients, most participants had enhanced TCR repertoire diversity that persisted several weeks after the end of rhIL-7 therapy 138 . The limited effects on thymic output in this study, as represented by minimal changes not only in the numbers of RTEs but also in the levels of TRECs, could be explained by the age and lymphopenic condition of the patients at the time points analysed. It is also possible that extended duration of rhIL-7 administration is necessary to have a greater effect on thymic function. However, compared with mice, in which IL-7 has both thymic-dependent and thymic-independent regenerative effects 135 , it is still unclear what the direct impact of exogenously administered IL-7 is on the thymus in humans and non-human primates. On the basis of clinical and preclinical observations, it appears that most of the effects on TCR diversity following rhIL-7 treatment are primarily driven by extrathymic sources, including the expansion of less frequent but highly diverse RTEs and naive T cells, which preferentially respond to IL-7 stimulation 145 , 146 , as well as their recirculation from lymphoid organs 141 , 147 , 148 . Lymphopenia, particularly with regard to CD4 + and CD8 + T cell subsets, in patients with COVID-19 on admission to hospital is emerging as one of the key clinical signs of COVID-19 and is closely associated with disease progression 39 . Enhancing T cell immunity could be a worthwhile strategy for treating these patients. Thus, a clinical trial has recently begun to investigate the possibility that IL-7 can restore T cell immunity in patients with COVID-19. KGF is a potent growth factor for TECs and is expressed under physiological conditions in the thymus primarily by mesenchymal cells 149 . KGF binds to its receptor, fibroblast growth factor 2 variant IIIb (FGFR2IIIb), on TECs and induces TEC proliferation through activation of the PI3K–AKT–nuclear factor-κB and p53 pathways 150 – 152 . Studies using knockout animals found that although KGF is redundant for thymopoiesis in steady-state conditions, it is crucial for thymic regeneration and peripheral T cell reconstitution after injury such as that caused by total body irradiation and syngeneic or allogeneic HCT 153 . The impact of exogenous administration of KGF on TEC function and thymic regrowth has been extensively evaluated in several mouse studies. Administration of recombinant KGF transiently accelerated thymic recovery after immune insults such as irradiation, cyclophosphamide therapy and dexamethasone therapy, and enhanced recovery of thymic and peripheral T cell numbers after HCT 150 , 152 , 154 , 155 . KGF reversed thymic involution and restored thymopoiesis in aged mice for up to 2 months after treatment 156 . A study performed in rhesus macaques showed that KGF-treated animals displayed accelerated haematological recovery, improved thymopoiesis and enhanced naive T cell recovery following HCT 157 . Although the effects on thymic function were modest, as measured by minimal changes in thymus mass, KGF-treated animals showed increased numbers of TREC-positive T cells up to 3 months following KGF treatment. Human recombinant KGF (palifermin; trade name Kepivance, marketed by Biovitrum) is approved by the US Food and Drug Administration for the prevention of mucositis in patients receiving high-dose chemotherapy. Several trials ( NCT01233921 , NCT03042585 , NCT02356159 and NCT00593554 ) are exploring its effects on T cell reconstitution, but no results have been reported yet. As demonstrated in previous preclinical work 158 , the benefits of palifermin on immune reconstitution in transplant recipients may derive from its synergistic effects with other immune-boosting therapies rather than as a sole therapeutic agent. However, a recent study of the use of KGF to promote immune reconstitution in patients with relapsing–remitting multiple sclerosis treated with the anti-CD52 lymphocyte-depleting agent alemtuzumab showed reduced thymic output in KGF-treated patients as measured by evaluation of naive CD4 + T cells, RTEs and TRECs ( NCT01712945 ) 159 . Given that human cTECs express CD52, one possible explanation for these clinical data is that palifermin exacerbates the negative effects of alemtuzumab on thymic function, perhaps through upregulation of CD52 expression on cTECs, rendering these cells more susceptible to antibody-mediated elimination. Thus, the combination of KGF with other drugs should be assessed cautiously as synergistic or deleterious effects on immune regeneration may occur depending on dosing, timing and mechanisms. Thymosins are a group of low molecular weight peptides originally isolated from bovine thymus 160 . Thymosin-α1, derived from prothymosin-α, is produced by TECs 161 and can increase lymphocyte maturation, boost T cell function and promote recovery following immune insults, although its mechanism of action is not completely understood. The receptor for thymosin-α1 is expressed by developing thymocytes, in which it regulates their survival and proliferation. Thymosin-α1 can antagonize dexamethasone-induced apoptosis of DP thymocytes in vitro, as well as the hydrocortisone-induced decrease in thymus and spleen mass 162 . Thymosin-α1 can also enhance the production of IL-7 by TECs 163 . Given promising preclinical results, multiple clinical trials have been initiated to evaluate the immunomodulatory effects of thymosin-α1 in the treatment of patients experiencing viral infections, immunodeficiency or haematological malignancies. Safety and efficacy of thymosin-α1 administration were evaluated in recipients of allogeneic HCT in a phase I/II clinical study ( NCT00580450 ). Treatment with thymosin-α1 increased T cell numbers and resulted in earlier appearance of pathogen-specific T cell responses against pathogens such as cytomegalovirus and Aspergillus species. Importantly, thymosin-α1 did not exacerbate acute or chronic GVHD and was associated with significant improvement in phagocytosis and dendritic cell function 164 . Recently, thymosin-α1 was given to patients with COVID-19 showing severe lymphopenia to enhance immunity. Thymosin-α1 treatment increased T cell numbers and recovery of thymic function, measured by TREC analysis 165 . Importantly, thymosin-α1 administration was also associated with increased survival of patients with severe COVID-19 (ref. 165 ). In addition to their fundamental role in regulating sex dimorphism, sex hormones can impact haematopoiesis at multiple levels. One of the first observations regarding a relationship between T cell development and sex hormones dates back to 1898, when it was reported that the thymus enlarged after castration of male rabbits 166 . Several studies confirmed the enlargement of thymic tissue after gonadectomy in both sexes in different experimental animal models. Conversely, androgens and oestrogens induce atrophy of the thymus 167 , 168 . The increase in the levels of sex steroids, and in particular of androgens, during puberty has been directly linked to the age-associated deterioration of immune function and to the process of thymic involution. Although the connection between the increase in the levels of sex steroids after puberty and the initiation of thymic involution is still debated, the regenerative impact of the removal of sex steroids on both thymic and BM lymphopoiesis has been extensively characterized. Indeed, through the use of clinically relevant mouse models of immune reconstitution after haematopoietic injuries, such as chemotherapy and radiotherapy, it has been demonstrated that sex steroid ablation enhances HSC self-renewal and lymphoid differentiation capacity and increases the number of common lymphoid progenitors in the BM 169 – 171 . Sex steroid ablation also has a direct effect on the BM microenvironment, restoring expression of key haematopoietic factors that are downregulated with age, such as FOXO1 (ref. 169 ). Considerable rejuvenation effects in the thymus have been extensively characterized, demonstrating that sex steroid ablation reverses thymic atrophy, accelerates the recovery of all thymocyte subsets and elicits potent regenerative signals to the thymic stromal microenvironment 55 , 172 – 174 . At a molecular level, sex steroid ablation promotes the upregulation of the key thymopoietic factors CC-chemokine ligand 25 (CCL25) 175 and DLL4 (ref. 167 ) in mTECs and cTECs, respectively. Several drugs have been developed to transiently and reversibly block sex steroids for the treatment of precocious puberty, endometriosis, hormone-sensitive prostate cancer and breast cancer. Some of these sex steroid blockers have been tested clinically to boost immune reconstitution after HCT. A non-randomized pilot study demonstrated that administration of the luteinizing hormone-releasing hormone (LHRH) agonist goserelin (Zoladex) before HCT significantly increased neutrophil engraftment, as well as total lymphocyte numbers, particularly those of naive CD4 + T cells, and levels of TRECs and improved recovery of TCR repertoire diversity 176 . Importantly, an increase in disease-free survival was observed in autologous HCT recipients treated with goserelin. Two trials ( NCT01746849 and NCT01338987 ) are ongoing to evaluate the effects of the LHRH agonist leuprolide (Leuprorelin) and the LHRH antagonist degarelix (Firmagon) to promote immune reconstitution following allogeneic HCT. Notably, the latest androgen receptor inhibitors and LHRH antagonists have the advantage of immediately blocking sex steroids without an initial surge of sex steroids as seen with LHRH agonists 167 . These novel approaches may provide better therapeutic tools to suppress sex steroids and mediate immune reconstitution. The regenerative effects of sex steroid ablation on T cell development might continue only as long as the levels of sex steroids are suppressed. However, the duration of such effect, particularly in the setting of surgical castration, remains a subject of debate in the field. After the initial regrowth following castration, the thymus of aged animals has been reported to decline and return approximately to its pretherapy condition 1 month after sex steroid ablation therapy 177 . While these results support a model in which the regenerative effects induced after surgical sex steroid ablation are transitory and dynamic, additional studies should be done to better characterize the nature of these ‘transient’ effects and the precise kinetics of thymic regeneration, in particular, at later time points. For example, it would be interesting to evaluate whether removal of the gonads, in the long term, can induce additional hormonal changes that negatively impact the process of lymphopoiesis. GH is a small peptide hormone secreted primarily in the bloodstream by somatotropic cells in the anterior pituitary gland 178 . Apart from its anabolic effects and impact on height, GH is also implicated in the regulation of haematopoietic function. Expression of GH receptor (GHR), by which GH mediates most of its effects, has been found on T cells, B cells, NK cells, monocytes, thymocytes and HSCs in humans and in other species 179 . While the impact of its signalling seems to be dispensable for HSC function, as suggested by the lack of phenotypic defects in HSCs of GHR-deficient mice, GH has important effects on immune function in mice and humans, either directly or through its principal mediator, IGF1. In vivo administration of GH can reverse thymic involution 180 – 182 . Transgenic mice that overexpress GH have an enlarged thymus. Similarly, the administration of a recombinant form of GH or IGF1 promotes thymic regeneration, increases TCR diversity and enhances recovery of haematopoietic compartments in immunocompromised and aged animals 183 . GH administration to immunocompromised patients has been studied in several clinical trials. Daily administration of human recombinant GH (somatropin) enhances thymic function and peripheral immune function in HIV-infected patients ( NCT00071240 ) 180 , 181 . The effects on thymic output appear to be transient, as discontinuation of GH treatment is associated with the recurrence of thymic atrophy 181 . Results from the recently completed studies ( NCT00287677 , NCT00119769 and NCT00050921 ) are still pending. The use of GH to reverse chronological ageing of the immune system was assessed in a recent pre‐phase-I study performed in volunteers aged between 51 and 65 years. Magnetic resonance imaging of thymic density showed that the accumulated fat tissue in the thymus was replaced with regenerated tissue in seven of the nine participants receiving the treatment 184 . In the periphery, GH treatment was associated with a significant increase in both naive CD4 + T cells and naive CD8 + T cells, with a concomitant decrease in PD1 + CD8 + T cells 184 . However, there are several concerns about the side effects associated with GH treatment, such as increased risk of heart disease, diabetes, elevated cholesterol levels and, more importantly, tumour progression. Thus, additional studies are needed to evaluate its potential use in humans.

Recovery of immunocompetence after periods of haematopoietic stress or injury is crucial not only for efficient responses against pathogens and tumour antigens but also for optimal responses to immunotherapy for cancer. In contrast to the early recovery of innate cells, including neutrophils, natural killer (NK) cells and monocytes, adaptive immune cells, in particular T cells, recover at a much slower pace and are particularly sensitive to negative insults caused by infections or cytoreductive chemotherapy and radiotherapy. Constrictions in the diversity of the T cell pool have been associated with impaired immune responses to several antigens 1 – 3 and adverse clinical outcomes in patients receiving haematopoietic cell transplantation (HCT) 4 , 5 . The capacity of T cells to mount and maintain effective responses to a wide variety of antigens depends on a large repertoire of unique T cell receptors (TCRs) generated in the thymus during the process of T cell development. This process is dependent on crosstalk between bone marrow (BM)-derived T cell progenitors and the supportive thymic stromal microenvironment, which primarily consists of thymic epithelial cells (TECs), endothelial cells, mesenchymal stromal cells, dendritic cells and macrophages 6 . Although, for example, T cell proliferation, driven by interleukin-7 (IL-7) and IL-15, in response to lymphopenic conditions can contribute to numerical reconstitution of T cells, complete long-term recovery of a diverse and functional T cell pool requires reactivation of thymic function and de novo T cell generation (Fig.  1 ). However, the thymus is sensitive to various injuries, such as those caused by cytoreductive treatments, infection, septic shock and graft-versus-host disease (GVHD). Furthermore, progressive involution of thymic tissue during ageing leads to a decline in T cell output and T cell senescence with restricted TCR repertoire diversity and impaired immune responses. Fig. 1 Overview of the dynamics and determinants of T cell reconstitution after haematopoietic cell transplantation. In the first period following haematopoietic cell transplantation (HCT), immune cells follow a predictable course of reconstitution. In contrast to the relatively early recovery of innate immune cells, recipients of HCT experience prolonged deficiencies in T cells and B cells, which can take more than 2 years to fully recover. This is particularly evident in adult patients, whose thymic function is lessened owing to age-related thymic involution. The ‘first wave’ of T cells after HCT comprises donor T cells that undergo lymphopenia-induced homeostatic proliferation and alloactivation. This results in polyclonal T cells with a restricted T cell receptor (TCR) repertoire and limited antigen specificity, or with alloreactivity causing graft-versus-host disease (GVHD). Overall, the incomplete recovery of the T cell pool has been directly linked to increased risks of infection, malignancy relapse and adverse clinical outcomes. Optimal and complete T cell reconstitution requires the regeneration of thymic function and the reactivation of endogenous T cell development. This allows the generation of a new T cell pool with broad TCR diversity. Multiple pretransplant and post-transplant factors influence the overall process of T cell reconstitution. HSC, haematopoietic stem cell; IL, interleukin; NK, natural killer. In the first period following haematopoietic cell transplantation (HCT), immune cells follow a predictable course of reconstitution. In contrast to the relatively early recovery of innate immune cells, recipients of HCT experience prolonged deficiencies in T cells and B cells, which can take more than 2 years to fully recover. This is particularly evident in adult patients, whose thymic function is lessened owing to age-related thymic involution. The ‘first wave’ of T cells after HCT comprises donor T cells that undergo lymphopenia-induced homeostatic proliferation and alloactivation. This results in polyclonal T cells with a restricted T cell receptor (TCR) repertoire and limited antigen specificity, or with alloreactivity causing graft-versus-host disease (GVHD). Overall, the incomplete recovery of the T cell pool has been directly linked to increased risks of infection, malignancy relapse and adverse clinical outcomes. Optimal and complete T cell reconstitution requires the regeneration of thymic function and the reactivation of endogenous T cell development. This allows the generation of a new T cell pool with broad TCR diversity. Multiple pretransplant and post-transplant factors influence the overall process of T cell reconstitution. HSC, haematopoietic stem cell; IL, interleukin; NK, natural killer. Thymic damage and impaired T cell reconstitution are particularly detrimental in HCT recipients 7 . Defective quantitative and functional recovery of T cells, in particular of CD4 + T cells 8 – 11 , has been directly linked to increased risks of opportunistic infections 9 , 12 , malignant relapse 13 and overall adverse clinical outcomes 14 , 15 . Defective T cell responses are a clinical hurdle not only for patients receiving HCT but also for patients receiving other modalities of cancer immunotherapy, including immune checkpoint inhibitors, that exert their antitumour effects primarily through the activation of T cell effector function. Although the prognostic significance of this association has still to be further characterized in larger studies, a highly diverse pool of T cells before therapy correlates with improved outcome after immune checkpoint blockade therapy 16 – 19 . Thus, there is considerable interest in developing approaches to evaluate the quantity and quality of T cells before and during different forms of immunotherapy to guide treatment directions, monitor immune responses and ultimately identify functional biomarkers to predict clinical outcomes 20 . In this Review, we highlight the primary causes of impaired immune function, with special emphasis on HCT recipients, and discuss regenerative approaches that have been clinically translated to facilitate the recovery of adaptive immune function. We also provide an update on emerging new immune-boosting approaches that have demonstrated promising regenerative properties in preclinical models. We focus on approaches that can broaden the diversity of the T cell pool through the restoration of de novo T cell formation in the thymus and discuss the implications for other cancer immunotherapies. While this Review primarily concentrates on T cell immunity, a brief summary of the B cell defects associated with immunological insults is provided in Box  1 . Compared with our understanding of T cell reconstitution, less information is available regarding the kinetics of B cell recovery after immunological insults and the factors regulating this process 258 . Optimal B cell function is important not only for the generation of protective antibodies and efficient antigen presentation but also for proper immune tolerance. This is particularly crucial for recipients of haematopoietic cell transplantation (HCT) as delays in B cell reconstitution are associated with increased risk of infection 259 and development of chronic graft-versus-host disease (GVHD) 260 . B cell numbers generally return to normal levels within 12 months after HCT, but it can take up to 2 years for complete recovery of the B cell compartment 259 , 261 . Several factors such as conditioning regimens , total body irradiation, corticosteroid treatment and GVHD negatively impact B cell reconstitution 77 , 258 , 262 . B cell-targeting therapies such as the monoclonal anti-CD20 agent rituximab also significantly delay B cell recovery 263 . In allogeneic HCT recipients, immunoglobulins can be derived from recipient plasma cells that survive conditioning regimens and donor-derived mature B cells 261 . However, restoration of the B cell compartment is primarily mediated by de novo regeneration from bone marrow progenitors. B cell recovery following HCT is reminiscent of B cell ontogeny 264 , 265 . Transitional CD19 + CD21 low CD38 hi B cells are the first B cells to emerge following HCT. Their percentage decreases in the first 12 months as the proportion of mature CD19 + CD21 hi CD27 − naive B cells increases 262 . However, even when the total number of B cells recovers, their functionality can remain compromised for 1–2 years. Indeed, in addition to environmental or vaccine-based antigen exposure, insufficient recovery and signalling from donor CD4 + T cells can result in B cell maturation arrest and decreased responses to vaccines 259 , 266 . Thus, the degree of CD4 + T cell recovery impairs B cell differentiation and functional reconstitution. In the first period following HCT, due to lack of T cell help, memory B cells display a limited diversity of the IgH complementarity-determining region 3 repertoire compared with naive B cells 267 . However, B cell autonomous defects are also evident in HCT recipients as the capacity to accumulate somatic mutations is decreased in mature B cells even in the presence of an adequate pool of CD4 + T cells 268 . Limited information is available regarding the dynamics of B cell receptor repertoire evolution following HCT. Similarly to the T cell receptor repertoire, the sequencing of the VH1 repertoire of class-switched B cells revealed lower repertoire diversity after HCT than the pretransplantation status 269 . However, a more comprehensive study of B cell receptor reconstitution and repertoires after allogeneic HCT is needed. Analogously to the T cell compartment, the naive B cell pool also declines with ageing. In old individuals, B cells are replaced by antigen-experienced memory cells carrying substantial functional changes, including impaired affinity maturation and isotope switching 88 .