PRIMARY AND SECONDARY DEFECTS OF THE THYMUS

Summary

The thymus is the primary site of T-cell development, enabling generation and selection of a diverse repertoire of T cells that recognize non-self, whilst remaining tolerant to self- antigens. Severe congenital disorders of thymic development (athymia) can be fatal if left untreated due to infections, and thymic tissue implantation is the only cure. While newborn screening for severe combined immune deficiency has allowed improved detection at birth of congenital athymia, thymic disorders acquired later in life are still underrecognized and assessing the quality of thymic function in such conditions remains a challenge. The thymus is sensitive to injury elicited from a variety of endogenous and exogenous factors, and its self-renewal capacity decreases with age. Secondary and age-related forms of thymic dysfunction may lead to an increased risk of infections, malignancy, and autoimmunity. Promising results have been obtained in preclinical models and clinical trials upon administration of soluble factors promoting thymic regeneration, but to date no therapy is approved for clinical use.

In this review we provide a background on thymus development, function, and age-related involution. We discuss disease mechanisms, diagnostic and therapeutic approaches for primary and secondary thymic defects.

Introduction

The thymus is the primary lymphoid organ responsible for the generation of a T-cell receptor (TCR) repertoire capable of responding to foreign antigens and providing surveillance against tumor cells whilst remaining tolerant to self.¹ The vital significance of the thymus in immune system development early in life has been known since the 1960s, when J. Miller reported infections, lack of rejection of foreign skin grafts and tumors in mice that were thymectomized during the neonatal period.¹ In the same decade, the first cases of thymus aplasia with fatal consequences were reported in nude SCID mice and humans with DiGeorge syndrome (DGS).^2-5 Inborn errors of the thymic stroma may be genetic, either intrinsic to the stroma and its precursor cells or due to interrupted lympho-stromal crosstalk, or can be a result of environmental causes, as seen in maternal diabetes (fig. 1).^6,7 The capacity to detect and treat thymic aplasia has improved remarkably with the availability of newborn screening (NBS) based on enumeration of T-cell receptor excision circles (TRECs) in dried blood spots collected at birth. However, the differentiation from primary hematopoietic cell defects versus thymic intrinsic defects remains challenging. Once identified, thymic aplasia can be treated with thymic tissue transplantation since its introduction in 1993.⁸

Fig. 1: Causes for primary and secondary defects of the thymus.

Primary thymic defects are cause genetically (red). Secondary defects can be due to intrauterine exposure to adverse conditions (green), or can arise postnatally following clinical interventions, diseases, or malnutrition (yellow). Prenatal, postnatal and age-related thymic hypofunction can have the same consequences with great variety in severity and persistance. A self-reinforcing cycle is possible in which primary or secondary defects of the thymus, e.g. secondary to cytoreductive therapy for HSCT, lead to clinical conditions, for example infections or acute GVHD, that, in turn, may aggravate thymic dysfunction, thereby increasing the risk for further clinical consequences, for example chronic autoimmune phenomena.

PP, pharyngeal pouch; HSCT, hematopoietic stem cell transplantation; endog., endogeneous; exog., exogeneous; TEP, thymus epithelial progenitor cell; TEC, thymus epithelial cell; GVHD, graft versus host disease; AIRE, autoimmune regulator.

Historically, the role of the thymus in later life has been considered insignificant due to its age-related involution.^1,9 Only recently has data on the importance of maintaining thymic function at later ages emerged.¹⁰ Age-related phenomena, such as increased risks for autoimmunity, cancer, infections and higher rates of mortality and morbidity of interventions like hematopoietic stem cell transplantation (HSCT) have indeed been linked to age-related decline of thymic function.^11,12 However, thymus atrophy and subsequent loss of function can occur also earlier in life as the thymus is highly sensitive to insults such as chemotherapy, radiation and immunosupressants¹³, as well as steroids¹⁴, malnutrition¹⁵, graft versus host disease (GVHD)¹⁶, and infection¹⁷ (fig. 1). Such insults can lead to thymus atrophy by harming hematopoietic cells prior to thymic immigration, by damaging the thymocytes in the thymus, or by harming the stromal compartment of the thymus. Because the thymocytes and stromal compartment of the thymus are highly interdependent both during prenatal development and postnatal maintenance and function, injury in one compartment affects the other as well.¹⁸ Insults can either harm thymic cells directly, for example in the case of HIV infection of thymocytes, or indirectly, as seen in systemically rising glucocorticoid levels during an organism’s stress response.¹⁷ The thymus has a remarkable capacity for self-regeneration.¹⁸ Yet, when thymus injury is prolonged or persistent, secondary thymic dysfunction may occur. Consequences of secondary thymic dysfunction resemble those of age-related thymic atrophy, namely lower naïve T-cell output and limited TCR diversity, ultimately causing increased risk for infections, malignancies and autoimmunity (fig. 1). The severity of thymus damage depends on the intensity of the insult and on the residual regenerative capacity, which decreases with age.^19,20 Despite the consequences, secondary forms of thymic hypofunction often remain undetected because measuring thymic function remains challenging. Even if detected, there are currently no approved therapies to enhance thymic function, but clinical trials are underway.¹³

In this review we provide a background on thymus development, function, and age-related involution, and discuss disease mechanisms and new diagnostic and therapeutic approaches for primary and secondary thymic defects.

Thymus development

In humans, the embryologic development of the thymus begins in the sixth week of gestation with the formation of the third pharyngeal pouches from the anterior ventral foregut endoderm (fig. 2A).²¹ Retinoic acid (RA) patterns the pharyngeal endoderm, as demonstrated in animal models wherein interrupted RA signaling results in absence of the third pharyngeal pouch.²² Accordingly, in vitro differentiation of human embryonic stem cells to thymic epithelial progenitor cells (TEPs) was impaired in the absence of RA.²³ The patterning and development of the third pharyngeal pouch, along with the subsequent formation of a primordium, are regulated by interacting transcription factors, including TBX1, HOXA3, PAX1, PAX9, SIX1, EYA1 and FOXI3 (fig. 2B).²² In mouse models, loss of these transcription factors leads to thymus aplasia.^22,24-30. The expression of Pax1, Pax9, Tbx1 and Hoxa3 has been shown to be dependent on RA levels.²² Upon formation of the third pharyngeal pouch, a common primordium of the thymus and parathyroid gland is generated, surrounded by mesenchymal cells of neural crest cell (NCC) origin. Ablation of NCC results in severely impaired thymus organogenesis.^31,32 Later on, mesenchymal cells develop into fibroblasts, building the thymic capsule and septae, endothelial cells, as well as pericytes, the cells located along the walls of capillaries which are involved in T-cells emmigration.^33-35 The interaction between mesenchymal cells and developing thymic epithelial progenitor cells (TEPs) is indispensable for thymic epithelial cell (TEC) development. This so-called epithelial-mesenchymal crosstalk involves morphogens such as bone morphogenetic proteins (BMPs), fibroblast growth factors (FGF), and hedgehog proteins.^34,36 Interruption of BMP4, for example, disrupts not only the formation of the thymic capsule, separation of thymus and parathyroid, and thymus migration, but also leads to a reduced size of the primordium.^36,37

Fig. 2: Early thymus organogenesis.

(A) Development of thymic epithelial cells (TECs) in humans. From the anterior foregut endoderm, which is patterend by morphogens including retinoic acid (RA), the third pharyngeal pouch arrises. A common thymus and parathyroid primordium arises and is surrounded by mesenchymal cells derived from the neural crest ectoderm. At that time, the thymic primordium contains a bipotent thymic epithelial progenitor (TEP) cell. Lymphocytes immigrate. The primordia seperate and migrate in a BMP4 dependend manner. The two thymic primordia meet in the anterior mediastinum and form the bi-lobed organ. (B) Formation and patterning of the third pharyngeal pouch and early TEP development is regulated by a network of transcription factors. An earlier, FOXN1 indipendend phase (in the endoderm, grey box) can be distinguished from a later, stage dependend on FOXN1, the master regulator of TEC development (in TEPs, red box). The arrows interconnecting the transcription factors and indicating a hierarchy are based on murine data. All displayed transcription factors lead to severe hypo- or aplasia if knocked out in mouse models. Transcription factors annoted with a star* were reported to cause human thymus aplasia if mutated in a loss of function manner, including HOXA3 (own unpublished data) (C) The development and function of the thymic stromal and lymphocyte compartment are highly interdependent. Development and maturation of TECs and subsequent differentiation into cTECs and mTECs requires crosstalk with lymphocytes. For mTEC development, single positive αβ thymocytes that have recently undergone positive selection by cTECs provide their ligands, including RANK, CD40 and lLT-β. Thereby, mTEC maturate and express the autoimmune regulator (AIRE) and costimulatory molecules. Mature mTECs enable establishmend of central tolerance by inducing apoptosis or Treg development of self-reactive T-cell clones.

Starting in the seventh gestational week, the left and right thymic domain separate from the parathyroid domains and migrate medially and caudally. During the ninth week, they merge and form the bi-lobed thymus organ located in the anterior superior mediastinum.³⁴

From midweek six FOXN1 is expressed in the thymic primordium (fig. 2A). In week eight differentiation of cortical and medullary TECs (cTEC, mTEC) begins, mediated by FOXN1 and its target genes, including Cbx4, which is associated with Tp63, a transcriptional regulator that maintains proliferative potential in epithelial cells.^38-43 Lymphoid progenitor cells colonize the thymus primordium, initially in a vasculature-independent manner guided by chemokines such as C-C motif chemokine ligand (CCL) 21 and CCL25, and later, from week 10, in rising numbers through the blood vessels.⁴⁴ The bidirectional interaction between thymocytes and the developing TECs, lympho-epithelial crosstalk, is indispensable for generation and maintenance of functional thymic epithelial compartments.^45,46 In particular, cTECs need interaction with immature double negative (DN) thymocytes to develop. In mice where thymocyte development is blocked prior to CD4− CD8− CD44+ CD25+ (DN2) stage, TECs remain immature and co-express both cytokeratin 5 (CK5) and CK8 (as opposed to mature cTECs that express CK8 only), and the thymus is severely hypoplastic with a disorganized epithelial compartment.^47,48 Blocks of thymocyte developmental at later stages dos not impair development of cTECs significantly.^49,50 By contrast, mTEC development requires interaction of stromal cells with thymocytes of a later differentiation stage, in particular single positive (SP) αβ thymocytes that have recently undergone positive selection by cTEC. The architecture of the medulla is severely impaired in mice where thymocyte development is interrupted at or prior to the CD4+ CD8+ double positive (DP) maturation stage, for example in TCRα deficient mice.^51,52 SP thymocytes provide ligands such as receptor activator of NFKB (RANK), CD40 and lymphotoxin-β (LT-β) receptor for the developing mTECs. Mature mTECs exprss the transcriptional activator autoimmune regulator (AIRE), which is necessary for the expression of tissue-restricted antigens (TRAs). It should be noted that mTECs represent the cell type with the largest diversity of gene transcripts in the entire body. Expression of TRAs in the context of major histocompatibility complex (MHC) molecules on the surface of mTECs and of dendritic cells in the thymic medulla enables negative selection of self-reactive SP T-cells, or their diversion to suppressive regulatory T-cells (fig. 2C). Accordingly, mutations in AIRE or in the genes involved in signaling pathways that regulate mTEC maturation and AIRE expression are associated with impairment of central tolerance and autoimmunity.^46,53-55

Postnatal T-cell development

Lymphoid progenitors originate in the bone marrow and migrate through the bloodstream to the thymus. Upon arrival, they immigrate at the cortico-medullary junction, a process regulated by CCL25 and further chemokines (fig. 3).⁴⁴ Subsequently, different steps of T-cell development occurs in different compartments of the thymus.⁴⁴ Trafficking of the thymocytes through the thymus is guided by chemokines produced by stromal cells and sequential expression of different chemokine receptors on the surface of the developing thymocytes. Early committed T-cells lack expression of T-cell receptor (TCR), CD4 and CD8. As cells migrate through the cortex and progress through the Pro- and Pre-T stages TCR δ, γ and β rearrangements occur. The stages are identified by their surface expression of CD34 and CD7, while lacking CD5 and CD1a (Pro-T), and by expression of CD5+ (Pre-T1) and CD5+ and CD1a+ (Pre-TII). cTECs promote development of the thymocytes by providing notch-mediated signals and interleukin-7 (IL-7).⁵⁶ Pre-TII cells differentiate further into αβ or γδ T-cells. In the development towards αβ T-cells, they proceed to the immature single positive stage (ISP) where they express a non-rearranging pre-TCR-α. Pre-TCR-α pairs with the TCR-β chain, which is the product of somatic DNA rearrangements that require expression of recombination-activation gene 1 (RAG1) and RAG2 proteins. At the cell surface, the pre-TCR-αβ pair is associated with a collection of proteins (the CD3/ζ complex).⁵⁶ Successful pre-TCR-αβ expression leads to recombination at the TCR-α locus to generate the α chain of the mature TCR. pre-TCR-α is no longer expressed. Low level of mature αβ-TCR assembled with CD3/ζ proteins are displayed on the cell surface. Co-receptor proteins, CD8 and CD4, are expressed and, eventually, a large population of double-positive αβ-TCR+ CD4+ CD8+ thymocytes (DP) is generated.⁵⁶ Up to young adulthood, these DP cells make up for around 99% of the cells in thymus.⁵⁵ During the subsequent selection, only the subset of the DP that is best suited to function in the host environment is permitted to mature. Firstly, the DP interact through TCRs with cTECs that express MHC class I and class II (MHC-I, MHC-II) molecules associated with self-peptides. Insufficient signalling, as is the case in around 90% of the DP thymocytes, results in apoptosis. The appropriate, intermediate level of TCR signalling initiates effective maturation and enables lineage-specific differentiation into either CD4+ or CD8+ mature T-cells (positive selection).⁵⁶ High signalling on engagement of TCRs with self-peptide–MHC ligands presented on mTECs, dendritic cells and B-cells, means potential autoimmune pathology and promotes rapid apoptotic death (negative selection) or generation of regulatory T-cells.⁵⁶ Macrophages play a role in the clearance of apoptotic thymocytes.⁵⁵ Less then 5% of immigrating thymocytes survive the repertoire selection and are eventually exported from the thymus to the periphery.⁵⁷

Fig. 3: Postnatal T-cell development in humans.

Early thymic progenitors migrate from the bone marrow to the thymus and enter at the cortico-medullary junction. They lack expression of T-cell receptor (TCR), CD4 and CD8. As the thymocytes progress through subsequent Pro- and Pre-T stages and reach the immature single positive stage where they express the pre-TCR. The pre-TRC is composed of a non-rearranging pre-Tα chain and a rearranged TCR β-chain. Successful pre-TCR expression leads to excessive cell proliferation and replacement of the pre-TCR α-chain with a newly rearranged TCR α-chain, resulting in a complete αβ TCR. The DP, αβ-TCR+CD4+CD8+ expressing thymocytes then interact with MHC class I and class II (MHC-I, MHC-II) molecules associated with self-peptides on cortical thymic epithelial cells (cTEC). The survival of the DPs depends on signaling that is mediated by interaction of the TCR with these self-peptide–MHC ligands. Low signaling leads to apoptosis. An intermediate level of TCR signaling enables further maturation (positive selection). Thymocytes that express TCRs that bind self-peptide–MHC -I complexes become CD8+ T-cells, whereas those that express TCRs that bind self-peptide–MHC-II ligands become CD4+ T-cells. During negative selection, occurring mostoly in the medulla where medullary thymic epithelial cells (mTECs), dendritic cells and B-cells present tissue restricted antigens (TRAs), too much signaling promotes apoptosis or generation of regulatory T-cells (Treg). Mature T-cells emigrate to the circulation. HSC, hematopoietic stem cell.

Physiological aging of the thymus

The human thymus begins to shrink after the first year of life, and size reduction occurs at a rate of 3% per year until middle age, when it decreases to a rate of less than 1% per year.⁵⁸ Decreased thymic output is already apparent by the age of 40 years, and by 55 years, only 5% of naïve T-cells are derived from thymic export.^59,60 Understanding the basic mechanisms of thymic age-related involution is important, because they overlap with mechanisms of secondary thymic atrophy, and may inspire therapeutic interventions.

With age, fewer progenitor cells migrate from the bone marrow to the thymus and total numbers of thymocytes are reduced.^18,61 Still, age-related changes of thymopoiesis seem to be mostly due stroma-intrinsic changes, not hematopoietic-intrinsic alterations.^62,63 Mouse models show that following transplantation of fetal thymus, early thymic progenitors have similar differentiation capacity in both young and aged mice.⁶⁴ Also, allogenic transplantation of thymocytes from young mice does not restore normal thymic output in aged mice. Conversely, HSCT from aged mice into mice with a young thymus is sufficient to partially reverse age-related immunological changes.^13,64,65 The stromal compartment of the human thymus in the elderly is characterized by loss of tissue organization, especially loss of the cortico-medullary junction, and augmentation of the perivascular space.^63,66 With age, total number of TECs and the ratio of total mTEC to cTEC declines.⁶³ Both mTECs and cTECS downregulate cell cycle-related genes, for example E2F3, a transcription factor that regulates cell proliferation, myc targets, and ribosomal genes.^67-70 At the same time, expression of inflammatory chemokines and cytokines such as IL-6 and IL-1β is upregulated in TECs and dendritic cells.^63,67,68,71 With age, key pathways in TECs are dysregulated, possibly mediated by micro RNAs (mRNA). ^71,72 Age-related alterations occur in the TGF-β pathway, where increased signaling was shown to inhibit thymopoiesis, while inhibition decelerates the process of thymic aging⁷³, and downregulates the Wnt pathway, essential for thympopoiesis^72,74. With age, TECs downregulate FOXN1, also possibly medated by mRNAs.^63,72 FOXN1 is not only pivotal for TEC development, but is expressed throughout life with utmost importance for TEC maintenance and function. FOXN1 regulates for example genes essential for early thymocyte development such as delta like Notch ligand 4 (Dll4), which promotes T lineage specification, C-X-C motif chemokine ligand Cxcl12, which is involved in intrathymic thymocyte trafficking, and Ccl25, wich directs thymic immigration of thymocyte precursors.^41-43,75,76 Accordingly, in mouse models induced expression of Foxn1 decelerates age-related thymic involution and restores age-related loss of TECs.^77,78 Furthermore, decreased expression of FGF21 may contribute to age-related changes in the TEC compartment. Loss of FGF21 delays the reconstitution of the thymus after injury.^63,79 In mice, overexpression prevents thymic age-related involution and atrophy.^63,79 With rising age, mTECs especially have impaired capability to support T-cell development. IL-7, a critical factor for survival, proliferation, and differentiation of thymocytes, is downregulated.^76,80 Expression of AIRE and MHC-II molecules is reduced, reflecting the diminishing expression of TRAs, impaired negative selection, and is a potential mechanism for age-related autoimmune phenomena.^63,81 Age-related changes are also notable in other types of stroma cells, including expansion of fibroblasts, potentially mediated by inflammation-induced TGF-ß or metalloproteinases, and adipocytes, the later possibly contributing to impaired thymopoiesis by secretion of cytokines.^63,82,83

Causative factors for age-related thymic involution are still poorly understood. Different factors may dominate in different age groups, supported by the observation that the speed of thymic involution varies throughout life.⁸³ Increased levels of steroid hormones may represent a key factor, as thymus degeneration peaks concomitant with a peak of sex steroid production after puberty.⁸⁴ This may also explain earlier onset of thymic involution in males, possibly mediated by androgens.⁸⁵ Furthermore, a decline of growth hormone may contribute to the involution.¹⁸ However, regrowth of the thymus after castration is transient even though androgen reduction is permanent, suggesting there are further mechanisms.⁸⁶ Another explanation may be damage from oxygen free radicals that accumulates over time, which was shown to contributes to the aging process in many organs.⁸⁷ Thymic involution may preceed aging in other organ systems as TECs lack catalase, a key antioxidant enzyme.⁸⁸ This goes along with the finding that reduced metabolic activity which results in less oxygen free radicals, for example in caloric restriction or administration of insulin-like growth factor (IGF), in turn reduces thymic age-related involution.^89,90

The result of age-related thymic involution is a progressive decline of total numbers of naïve T-cells, as reflected in decreasing TREC levels and TCR repertoire diversity. At the same time, homeostatic T-cell expansion and relative number of memory T-cells increase.⁶³ Furthermore, naïve T-cells are functionally impaired. TCR-dependent stimulation and responses to mitogens are weakened. Homing receptors such as CD62L and CCR7 are downregulated, impeding migration to sites of action.^91-94 The clinial phenotype of age-related thymic involution resembles inborn or aquired thymic hypofunction (fig. 1). Recently, a mathematical model showed that rising rates of cancer and infections with increasing age, are caused primarily by thymic involution and not immunosenescence of the immune system in general.¹¹ Viewing the thymus as key player in age-related immunological disorders is further strengthened by the observation that thymus transplantation could reverse autoimmune disease in aging mice.⁹⁵ In fact, thymic function is an independent predictor of mortality in healthy humans over 65 years.⁹⁶ Besides impaired immune function, age-related thymic involution leads to decreased regenerative capacity of the thymus after insults such as chemotherapy and radiation. This reduced regenerative capacity aggravates with age, ranging from a delayed to an incomplete recovery of thymic function.^13,97

Consequences of thymic hypofunction

The most severe form of inborn errors of the thymic stroma, thymic aplasia, causes severe combined immunodeficiency (T−/lowB+NK+ SCID). Patients present with severe, possibly fatal infections of a broad spectrum, as not only the cell-mediated immunity is impaired, but also the required support for antibody production.⁶ Patients may develop Omenn syndrome, an inflammatory phenotype characterized by erythroderma, desquamation, alopecia, diarrhea, failure to thrive, lymphadenopathy, and hepatosplenomegaly.^98,99

Less severe thymic hypofunction, either inborn, secondary, or age-associated, may causeincreased rates of morbidity and mortality of a variety of conditions, as discussed below and summarized in fig. 1. The incidence of infections and thymic output are inversely correlated.¹¹ Age-related thymic involution leads to increased susceptibility to infections.¹⁰⁰ For example, being older than 65 years is a risk factor for morbidity and mortality from SARS-CoV-2.¹⁰¹ Thymic atrophy secondary to malnourishment is associated with susceptibility to fatal outcome after infectious diseases.¹⁰² There is the risk for a self-enforcing circle, with thymic atrophy leading to higher rates of infections which, in turn, contribute to thymic hypofunction directly or further malnourishment, which aggravates the secondary thymic atrophy.¹⁵ Thymic atrophy secondary to cytoablative therapy in the setting of HSCT leads to incomplete T-cell immune reconstitution, especially of CD4+ T-cells, and a reduced TCR diversity, which in turn causes an increased risk of opportunistic infections and impaired immune responses to several antigens.^103-105 Age-related thymic involution is also associated with an impaired response to vaccines.^100,106 Only around 30-40 % of elderly people are capable of generating a protective immune response to the influenza vaccine.¹⁰⁷ Thymic atrophy increases the risk for autoimmune diseases.¹⁰⁸ Multiorgan autoimmune disease with reduced lifespan occurs also if NFKB signaling is impaired, which is essential for the development and differentiation of mTECs, causing thymic medullary atrophy. This is, for example, the case in mice deficient in RelB, an NFKB transcription factor subunit, RANK, which selectively stimulates the noncanonical NFKB, and NFKB 1 and 2 deficient mice and humans.^109,110 In another mouse model, thymic central tolerance was disrupted by murine roseolavirus infection, which induces autoimmune gastritis.¹¹¹ Thymic atrophy can also be secondary to acute GVHD, with the process of negative selection notably impaired, which may cause chronic autoimmune manifestations affecting the skin or the salivary gland, for example.^112-114 Also, thymic output and the incidence of cancer are inversely correlated.¹¹ Long-term immune reconstitution in patients that have undergone HSCT has been reported to be mainly thymus-dependent.¹¹⁵ Impaired thymic function prior to HSCT correlates with an increased risk of relapse of malignancy, severe and opportunistic infections, and mortality.^116,117 Mouse models of cancer where HSCT was combined with thymic transplantation, or with other interventions aimed at improving thymic regeneration, showed increased rates of tumor regression and overall survival.^118,119 Age-related thymic involution may be the cause for slower T-cell reconstitution after HSCT and higher morbidity and mortality in the elderly, and, conversely, higher total naïve T-cell numbers and TREC levels as well as greater TCR diversity in a pediatric population compared to adult recipients.^117,120-122

Primary thymic defects

Inborn errors of immunity associated with thymic stromal defects

DGS is characterized by a constellation of clinical features due to failure of the appropriate development of tissues and organs that have a common embryogenetic origin. The clinical features include congenital heart disease (CHD), thymic aplasia or hypoplasia, facial dysmorphism, palatal malformation and hypoparathyroidism.^123-125 The most frequent molecular defect associated with DGS is chromosome 22q11.2 deletion syndrome (22q11.2del)¹²⁶, but many other less frequent genetic abnormalities and environmental exposures during pregnancy can lead to overlapping clinical features. The spectrum of immunodeficiency in DGS is variable.^125,127 Most patients with DGS have an incomplete phenotypic spectrum (partial DGS); they have significant lymphopenia but residual presence of naïve CD4+ T-cells ≥50 cells/mL and an increased proportion of oligoclonal memory T-cells due to homeostatic proliferation.^125,128-130 By contrast, only 1% of DGS subjects have complete absence of thymic tissues and naïve CD4+ T-cells numbers below 50 cells/ mL, in addition to the other typical features of the disease (complete DGS). They are prone to severe viral and bacterial infections early in life and require cultured thymic tissue implantation.¹³¹ In a significant proportion of cases, patients with complete DGS develop in the first weeks of life an expansion of a markedly oligoclonal repertoire of T-cells that infiltrate the skin and other organs, mimicking Omenn syndrome. This condition is also referred to as “complete atypical DGS” and requires immune suppression to reduce the risk of organ damage and prevent graft rejection prior to thymus implantation.

22q11.2del is the most common microdeletion syndrome in humans affecting almost 0.1% of fetuses.¹³² The incidence is 1 every 3000–6000 births with male to female ratio of 1:1.^133,134 More than 90% of DGS patients carry this deletion on one allele; it can be inherited, but in most cases it occurs de novo.¹²⁵ This specific chromosomal region is enriched in low copy repeats and thus it is prone to genomic instability during meyosis with consequent high ratio of deletions and duplication events in this area.¹³⁵

90% of patients with 22q11.2del carry a 3 Mb deletion encompassing more than 80 genes while around 6% of patients carry a 1.5 Mb deletion that includes 30 genes.¹³³ There is no direct correlation between the size of the chromosomal deletion and the severity of clinical manifestations.¹³⁶ Furthermore, the same genetic abnormality can be associated with variable clinical phenotypes, although genetic drivers of specific clinical features of DGS have been reported. Lack of one copy of the TBX1 gene leads to more severe lymphopenia and CHD, and haploinsufficiency of the CRKL gene maybe implicated in both T-cell lymphopenia, neuropsychiatric impairment and kidney disease.^137-139 The clinical manifestation with the most significant comorbidity in 22q11.2del is CHD (ventriculo-septal defect, Tetralogy of Fallot, truncus arteriosus). It is present in half of the DGS patients, and it is the leading cause of death.^140,141 Hypoparathyroidism is also present in 50% of patients; it results in clinically significant hypocalcemia manifesting as neonatal seizures, feeding difficulties, and weight loss.¹⁴² Monitoring of calcium levels and prompt calcium supplementation are especially important in situation of stress, such as infections, pregnancy and puberty.¹⁴³ Other clinical manifestations that tend to increase with age in 22q11.2del patients include atopic disorders, immune dysregulation, malignancies and neuropsychiatric disorders. According to the severity of lymphopenia, patients are at increased risk of viral infections and of bacterial superinfections; however, this increased risk of infections is not only secondary to immunodeficiency, but also to the anatomical and behavioral abnormalities characteristic of the disease.^124,125 Changes in thymus size and architecture are important determinants of the severity of the immune defect in 22q11.2del.¹⁴⁴ The degree of hypoplasia of the thymic stroma is directly correlated to thymic output, as reflected by decreased generation of naïve T-cells, and low number of recent thymic emigrant (RTE) and T-regulatory cells. Defects of the thymic medulla and of AIRE expression have been observed in DGS, and may lead to impairment of negative selection and persistence of self-reactive T-cell specificities, accounting for the high rates of autoimmunity.¹⁴⁴ Moreover, B-cell defects have been also reported in 22q11del. In particular, reduced antibody production, impaired response to immunization, likely secondary to defective T-cell helper activity^145-147, and increased proportion of autoreactive B-cells, may contribute to the increased rate of infections and of autoimmune manifestations (rheumatoid arthritis, idiopathic thrombocytopenia, autoimmune hemolytic anemia, and thyroid disease)¹²⁴ that have been described, especially in older patients. Some of these autoimmune manifestations correlate with low CD4 counts (autoimmune thyroiditis) and low naïve T-cell numbers (autoimmune thrombocytopenia).¹⁴⁸

Heterozygous pathogenic mutations in TBX1 and TBX2 genes are also a rare cause of DGS.¹⁴⁹ The proteins encoded by these genes are transcriptional regulators that belong to the T-box family and are necessary drivers of embryonic and thymus development. TBX1 expression is controlled during the organogenesis of the third and fourth pharyngeal pouch.¹⁵⁰ This gene regulates the expression of more than 2000 genes through epigenetic changes.^133,151 In mice, Tbx1 is expressed early in embryogenesis in the pharynx and plays a crucial role in pharyngeal arch artery formation and pharyngeal segmentation.^28,152,153 The effect of Tbx1 abnormal functioning on mouse thymus is time dependent. Deletion of Tbx1 before 11 days during mouse embryogenesis leads to a severe defect in thymus development, whereas its deletion after 11.5 days has no consequences on thymus formation.²⁸ Later in embryogenesis, expression of Tbx1 in the third pharyngeal pouchbecomes more restricted to the parathyroid domain¹⁵² and Tbx1 expression in the thymus primordium is turned off to allow for TEC differentiation.¹⁵⁴ While heterozygous Tbx1 mutations in mice are not associated with thymic defect¹⁵⁵, several monoallelic mutations of this gene in humans have been associated with a clinical phenotype of hypoparathyroidism, hypocalcemia, CHD, velopharyngeal insufficiency but also thymic hypoplasia and lymphopenia in around 30% of patients.^156-160 The variable phenotype of the disease is most likely due to the impact of epigenetic modifications of TBX1 and of other genes that affect TBX1 expression. TBX2 together with TBX1 and TBX3 coordinates the development of the pharyngeal and the arterial pole morphogenesis.¹⁶¹ Liu et al. reported a family with a dominant TBX2 pathogenic variant. The 3 affected family members manifested with variable expressivity of skeletal, cranial, cardiac abnormalities and thymic hypoplasia. One sibling required cultured thymic tissue implantation due to minimal thymic output.¹⁶² Interestingly a fourth, unrelated patient, with a different monoallelic variant, did not suffer from any evident thymic dysfunction.¹⁶²

Another genetic cause of DGS is partial monosomy 10p. Around 50 patients with this genetic anomaly have been identified so far. They present with a DGS-like phenotype, but also other distinctive features like deafness, genito-urinary abnormalities and developmental and growth retardation. Analysis of the breakage points in affected patients has demonstrated that two regions on chromosome 10p— DiGeorge Critical Region 2 (DGCR2) and HDR1— are associated with different clinical characteristics.^163-165 Deletion of the DGCR2 region leads to cardiac defects and thymic hypoplasia¹⁶⁴, while loss of the HDR1 region is associated with sensorineural hearing loss, kidney dysplasia and hypoparathyroidism.^165,166

Pathogenic variants in the CHD7 gene lead to a complex syndrome defined as CHARGE association (Coloboma, CHD, Atresia choanae, Retarded growth and development, Genital and Ear anomalies).¹⁶⁷ CHD7 codifies for an enzyme that is crucial for chromatin remodeling¹⁶⁸ and regulates the expression of multiple transcription factors that control embryonic development. CHD7 is expressed in mesenchymal cells and in TECs and finely tunes the development of the third pharyngeal pouch as well as of the thymus, by regulating the expression of the master regulator transcription factor FOXN1.¹⁶⁹ The incidence of CHARGE is 1 in 16000 live birth.¹⁷⁰ Almost 1000 causative mutations have been reported so far; the vast majority have occurred de novo on the paternal allele, and only few are representative of mosaic or germinal mosaicism.¹⁷⁰ Thymic hypoplasia and consequent immunodeficiency are present in around 50% of the patients with CHARGE syndrome. The immunological phenotype is variable, ranging from marked T-cell lymphopenia and Omenn syndrome to mildly decreased naïve T-cell count, reduced T-cell function, with normal B and NK cell numbers.^99,171-173 Of note, even CHARGE patients without documented thymic aplasia or hypoplasia can display low or absent naïve T-cells.^174,175 While athymia and SCID phenotype are rare, and usually the T-cell lymphopenia improves overtime, some patients with CHD7 haploinsufficiency have been identified by NBS and have required cultured thymus tissue implantation early in life to survive.⁸

Features of DGS without CHD have been linked to chromosome 2p11.2 microdeletion in 5 unrelated families. Four of the probands were identified due to a positive NBS and displayed both hypocalcemia and lymphopenia early in life.¹⁷⁶ The analysis of the deletion breakage points in 13 subjects highlighted that several overlapping microdeletions in the 2p11.2 area are all encompassing the forkhead box I 3 transcription factor (FOXI3) gene.¹⁷⁶ FOXI3 is a composed of 2 exons and encodes for a 420 amino acid transcription factor with a DNA binding domain that is functionally active during early stages of embryogenesis. FOXI3 is expressed in the pharyngeal pouch endoderm and ectoderm concurrently with TBX1 with which genetically interacts.^29,177 Haploinsufficiency for both Tbx1 and Foxi3 in mice causes parathyroid and thymus gland development defects. Foxi3-null mice have impaired segmentation of the pharyngeal pouch with thymus, craniofacial and ears abnormalities¹⁷⁸ while the phenotype of Foxi3^+/− mice is restricted to a smaller thymus size.¹⁷⁶ In 2022, two unrelated individuals with FOXI3 haploinsufficiency were described; this condition is now included among the genetic causes of T-cell lymphopenia at birth.¹⁷⁹ Both probands presented with abnormal TRECs at birth and moderate CD4+ and CD8+ T-cell lymphopenia. Genetic testing revealed in one subject a novel heterozygous c.280C>T (p.Gln94Ter) nonsense variant and in a second subject a heterozygous c.74_75dup (p.Ala26Profs∗116) frameshift variant. Both mutations were in exon 1 of the FOXI3 gene. Interestingly the first patient shared the mutation with his father who however has remained clinically asymptomatic, and with two older sisters whose NBS and lymphocyte counts were normal, supporting variable expressivity of the phenotype of this disease. The second proband was born with no visible thymic shadow; however, his lymphopenia improved overtime. He had normal antibody responses and T-cell proliferation and lived the first 8 years of life with normal growth and development with no significant history of infections.¹⁷⁹

FOXN1 is a member of the forkhead/winged helix family of transcription factors that controls the development, differentiation and survival of epithelial cells in the skin as well as of cTECs and mTECs both during embryogenesis and in post-natal life.^41,180,181 Biallelic loss-of-function (LOF) mutations in FOXN1 lead to the so called Nude SCID phenotype, originally described in mice², and later identified in infants presenting with athymia, SCID, congenital alopecia universalis, and nail dystrophy.^182,183 The immunological phenotype is consistent with decreased thymic output: low TRECs, decreased count of RTE, naïve CD4+ and naïve CD8+ T-cells, impaired T-cell proliferation, and absence of thymic shadow. The B-cell numbers are usually preserved, but B-cell function is often impaired due to lack of T-cell helper activity. These patients are at high risk of life-threatening infections since early in life and require cultured thymus tissue implantation to correct the disease.^182,184-187 Compound heterozygous cases of FOXN1 have also been described in 2 patients with selective thymic hypoplasia but no alopecia or nail dystrophy. These patients have a milder phenotype compared to the homozygous affected subjects with moderate lymphopenia, low or absent naïve T-cells and normal B and NK cells. One patient died as a newborn of Parainfluenza infections.¹⁸⁸

More recently, the NBS program based on TRECs enumeration has allowed the identification heterozygous loss of function FOXN1 mutations as a cause of T-cell lymphopenia at birth.^187-189 Bosticardo et al. described a cohort of 25 pediatric patients carrying heterozygous FOXN1 variants; 21 of them were flagged by the NBS program and received genetic testing. Most mutations were localized in the Fork-head domain -FKD- and in the C-terminal domain. They were mostly frameshift (65%) with consequent early stop codon, followed by missense (20%), nonsense and splice site mutation. 10 out of 13 patients evaluated by imaging had small or absent thymic shadow. 72% of the patients suffered from viral infections that were mostly non severe with only 5 individuals having recurrent or severe infections including EBV diffused large B-cell lymphoma in 1 case. Nail dystrophy was present in 50% of patients while atopy occurred in 10% of the cohort. Immunological phenotype was characterized by T-cell lymphopenia at birth that tend to normalize during adulthood in the CD4 T-cell compartment but not in the CD8. Three patients received HSCT prior to the genetic diagnosis. One patient died 2 years after transplant while the 2 other patients survived but despite achieving good T-cell engraftment remained with low T-cells after the procedure, consistent with a thymic stromal defect.^189,190 In another cohort of 5 patients with heterozygous FOXN1 variants, 4 of the patients belonged to the same family. Three of them presented with a SCID-like phenotype at birth while the fourth suffered only from mild to moderate T lymphopenia supporting the mechanism of variable expressivity in FOXN1 heterozygous mutations.¹⁸⁷ Furthermore FOXN1 compound heterozygous mutations are associated with T^−/lowB⁺NK⁺ SCID without evidence of alopecia and nail dystrophy.¹⁸⁸ Finally a recent study identified 3 individuals with the same FOXN1 c.1370delA frameshift mutation. Devoted functional studies elucidated the mechanism by which this and other C-terminal mutants exert a dominant negative effect on wild-type FOXN1.¹⁹¹ The broad spectrum of mechanism by which monoallelic FOXN1 variants lead to different degrees of thymic hypoplasia has been recently studied by Moses et al. in 35 selected FOXN1 variants that were tested with transcriptional reporter assays, imaging studies and reaggregate thymus organ cultures. This approach allowed to categorize the variants into benign, loss- or gain-of-function, and/or dominant-negatives. Of note the dominant negative activities mapped to frameshift variants impacting the transactivation domain while a nuclear localization signal was mapped within the DNA binding domain.¹⁹²

Homozygous variants in PAX1, a gene that belongs to the paired box family of transcription factors, cause otofaciocervical syndrome type 2 (OTFCS2), a clinical syndrome characterized by external ear anomalies/ear impairment, facial dysmorphic features, skeletal abnormalities in the shoulder girdle and vertebral bodies and mild intellectual impairment.^193-197 PAX1 is expressed in the sclerotome, from which the vertebral column develops, and in all pharyngeal pouches. Due to its expression also in the third pharyngeal pouch, it regulates thymus development.¹⁹⁸ PAX1 is expressed at an earlier timepoint during thymus development than FOXN1, but its expression continues at later time points in the thymic enlage and in cTECs as a FOXN1-dependent gene.¹⁸¹ Yamazaki et al. reported 6 patients whose immunophenotype was consistent with T⁻B⁺NK⁺ SCID and clear evidence of impaired thymic output with low RTE, reduced count of naïve T-cells, and absent thymic shadow.¹⁹³ Two patients presented with multiorgan T-cell infiltrates due to oligoclonal T-cell expansion consistent with Omenn syndrome. Four patients were treated with HSCT but failed to reverse the immune defect and lack of naïve T-cells, consistent with a thymic stromal cell intrinsic defect that would be corrected only by thymus transplantation. Molecular studies have shown that when PAX1 is mutated, the DNA binding is affected, and the gene transcriptional regulatory function is impaired. Differentiation of patient-derived induced pluripotent stem cells (iPSC) into TEPs, allowed to show that PAX1 mutations lead to impaired expression of multiple genes, including some that are crucial for TEC development such as FOXN1, TP63 and BMP4, but also genes involved in skeletal, cartilage, pharyngeal, and ear development, that would account for the broad range of malformations presented by patients carrying PAX1 mutations.¹⁹³ Recently, another cohort of 6 PAX1 deficient patients has been described. All patient presented with classical dysmorphic features of PAX1 deficiency. The immunophenotype was either SCID (n=3) or combined immune deficiency (CID) (n=3) and manifested as recurrent infections in half of the patients, Omenn syndrome in 1, failure to thrive in 4, autoimmune cytopenia in 3 subjects and EBV-related lymphoproliferative disease in another patient. Interestingly, some of these patients presented features of DGS that had not been previously reported in PAX1 deficiency. Five out of 6 patients suffered from primary hypoparathyroidism, and 3 patients presented with neonatal hypocalcemia leading to seizure and tetany. Two patients had atrial septal defect that was hemodynamically neutral and did not require surgery.¹⁹⁶

Variants in the Exostosin-like 3 (EXTL3) gene, which encodes a glycosyltransferase involved in heparan sulphate proteoglycan biosynthesis (HSPG), are associated with a rare syndrome characterized by T-cell lymphopenia of variable severity, skeletal abnormalities and neurodevelopmental delay, and often liver and kidney cysts.^199-202 HSPG are crucial modulators of growth factors for both hematopoietic progenitors and thymic stromal cells. While no human data describing the thymus tissue defect are available Volpi et al. showed that the extl3-mutant zebrafish has a thymus reduced in size, consistent with a role for the EXTL3 protein in thymus organogenesis.¹⁹⁹ Experiments with EXTL3-mutant iPSCs showed that patient’s cells were impaired in their differentiation towards both early lymphoid progenitors and TEC lineages.¹⁹⁹ In particular, when differentiated to the TEC lineage, patient’s iPSCs displayed decreased expression of TEC-specific genes, such as FOXN1, KRT5 and EYA1, while retaining higher expression of SOX17, a marker of definitive endoderm that is physiologically downregulated during differentiation from DE cells to TEPs.¹⁹⁹ HSCT has been curative for the immune defect in one case of EXTL3 deficiency, but only partially corrective in another patient^199,200 consistent with the likely dual origin of the immune defect in this disease.

A hybrid defect in both lymphoid and thymic stromal cell compartments has also been postulated in TTC7A deficiency. Biallelic LOF variants in TTC7A, which encodes a tetratricopeptide repeat domain containing protein, have been described in individuals with multiple intestinal atresia and combined immune deficiency affecting T, B, NK cells.^203-209 On the other hand, hypomorphic variants of the TTC7A gene have been associated with features of ectodermal dysplasia including alopecia and nail dystrophy but also autoimmunity, IBD and increased risk of infections.²⁰⁶ TTC7A is expressed not only in hematopoietic cells and gut epithelial cells, where it regulates actin cytoskeleton function crucial for both T-cell activation, proliferation, migration²¹⁰ and gut epithelium polarity, but also in TECs. This has been demonstrated by the analysis of a thymic autoptic specimen²⁰⁵ which showed dysplastic changes and a vague cortico-medullary demarcation, along with severe lymphoid depletion.²⁰⁵ Altogether, this evidence suggests that a thymic stromal defect may contribute to the immune defect present in these patients.

The AIRE gene is expressed by MHC-II^hi mTECs, and is responsible for the expression of TRAs, allowing negative selection of self-reactive thymocytes and induction of central T-cell tolerance. Biallelic LOF variants in this gene cause the Autoimmune Polyglandular Syndrome type 1 (APS1), also known as Autoimmune Polyendocrinopathy, Candidiasis, Ectodermal Dystrophy (APECED).²¹¹ Patients with APS1 present with chronic mucocutaneous candidiasis and early onset autoimmune polyendocrinopathy including hypoparathyroidsm and adrenal insufficiency, ectodermal dystrophy including enamel anomalies and nail dystrophy, elevated titers of autoantibodies against IL-17, Il-22²¹² and type I interferons.²¹¹ Additional autoimmune manifestations are primary gonadal insufficiency, autoimmune thyroiditis and hepatitis, diabetes mellitus type 1, pernicious anemia. Recent studies have shown that there is also an autosomal dominant form of the disease, due to heterozygous dominant-negative variants, that manifests with milder clinical features and variable expressivity.^213-217

Studies in mouse models have provided insights on the pathophysiology of AIRE deficiency in humans and on the consequent alterations of the thymic structure even though the mouse phenotype does not fully recapitulate the human disease. Reduced Aire expression in mTECs leads to impairment in the distribution and differentiation of mTECs and Hassall's corpuscle-like structures in the medulla, supporting the notion that Aire controls the differentiation program of mTECs.²¹⁸

Secondary defects of AIRE expression and lymphostromal cross-talk

Secondary defects of AIRE expression have been reported in several conditions in which abnormalities of mTEC compartment are due to defective crosstalk with developing thymocytes¹⁴⁹, or because of defective AIRE signaling, as in the case of mutations in NFKB2 and RELB, both of which affect NFKB signaling in mTECs.²¹⁹

A recent study on NFKB2 and RELB patients has documented thymic stroma abnormalities in these IEIs and has linked defective AIRE expression in these patients to the generation of anti-Interferon Type I antibodies.²²⁰ In human fetal thymuses, NFKB2 and RELB transcripts are highly abundant in AIRE-expressing mTECs. However, the impact of deleterious variants affecting the alternative NKKB pathway on AIRE expression was undefined. Studies in patients aged 4 to 16 years with deleterious NFKB2 variants (n = 11) revealed that the total thymic volume was smaller compared to age-matched controls. Immunofluorescence analysis of thymic biopsy samples from a patient with complete autosomal-recessive RELB deficiency and a 27-year-old deceased patient with NFKB2 deficiency revealed a dysplastic organ with vague corticomedullary demarcation and atrophic medulla. A residual epithelial cell population (pan-keratin-expressing cells) with disorganized keratin 5 (K5)- and keratin 8 (K8)-positive cells was detected in the thymuses of both patients. mTECs were rare, but not entirely absent, in the thymus of the patient with autosomal-recessive RELB deficiency. However, no AIRE or keratin 10 (K10)-positive Hassall’s corpuscles were detected. These findings suggest that RELB deficiency does not completely block mTEC specification but, rather, prevents differentiation into AIRE-expressing mTECs. The analysis of the thymus from the adult patient with an autosomal-dominant NFKB2 variant showed lack of mTECs altogether, including AIRE-expressing cells and Hassall’s corpuscles. These results suggest that the NFKB2 deficiency impairs the maturation of human AIRE-expressing cells. However, thymic involution in this older patient makes it difficult to draw definitive conclusions. In any case, these data support the concept that human NFKB2 and RELB control the development of mature mTECs and the thymic expression of AIRE²²⁰

Hypomorphic mutations in RAG1/2 lead to a broad variety of clinical phenotypes including Omenn syndrome (with oligoclonal and activated autologous T-cells infiltrating the skin, and other organs) and CID with autoimmunity and granulomas.^221,222The leading autoimmune manifestation in these patients is cytopenia including autoimmune hemolytic anemia, immune thrombocytopenia, or autoimmune neutropenia. In addition, other manifestations of immune dysregulation may affect the skin, vasculature, gut, endocrine glands, CNS and kidneys. Granulomas are not restricted to the skin, but also involve internal organs and bones. Furthermore, similar to what was observed in APECED patients, hypomorphic RAG mutations in humans are frequently associated with neutralizing autoantibodies to interferon-α (IFN-α), and IFN-ω, and in a smaller proportion of cases to IL-12, IL-17 and IL-22.²²³ While it is not clear if these anti-cytokine antibodies are the cause or the consequence of an increased rate of infections in these patients, it is interesting to observe that the thymus of patients with hypomorphic RAG mutations has a smaller medullary compartment, loss of corticomedullary demarcation, lacks Hassall’s corpuscles, and is also characterized by reduced AIRE expression, which could be an important driving factor for the immune dysregulation phenotype of this condition.^45,224 The hypothesis that these thymic abnormalities lead to an impairment of negative selection is supported by the observation that peripheral blood T-cells of hypomorphic RAG-mutated patients carry a high frequency of TCR clonotypes containing central cysteine residues²²⁵ (indicative of impairment of wave 1 deletion of self-reactive thymocytes in the thymus cortex) and hydrophobic amino acids at positions 6 and 7 of the CDR3 (a molecular signature of impairment of the second wave of negative selection).²²⁵ Moreover, regulatory T-cells are present in low numbers, display a restricted repertoire and have reduced suppressive function.^226,227

A significant reduction in thymus size, with few or absent Hassal’s corpuscles and absent demarcation of corticomedullary junction was also observed in multiple other forms of SCID or CID^45,228, including IL2RG/CD132 deficiency²²⁹, JAK3 deficiency²³⁰, IL7R deficiency²³¹, AK2 deficiency²³², CD3D, CD3E and CD3Z genes^45,233,234, and ATM deficiency^235-237. By contrast an hypoplastic thymus with presence of Hassall’s corpuscles and differentiated germinal epithelium has been described in adenosine deaminase (ADA) deficiency.^238,239

MHC-II deficiency includes several genetic defects of transcription factors that lead to expression of MHC-II molecules on TECs and thymic dencritic cells.^240-242 Antigen presentation within MHC-II pouches is crucial for positive and negative selection of the CD4+ developing lymphocytes. B-cell antibody production and immunization response is also affected due to lack of CD4+ T helper function.^243,244 Thymic samples from MHC-II-deficient patients showed a small-sized thymic medulla, and reduced AIRE expression in TECs, possibly leading leading to impaired negative selection.²⁴⁵ As this CID involves defective expresssion of MHC-II molecules on both thymic stromal tissue and on dendritic cells, HSCT is only partially curative, and persistent CD4+ T-cell lymphopenia is often observed after transplantation.²⁴⁶

ZAP-70 is a kinase involved in TCR signaling; its deficiency leads to selective lack of CD8+ T-cells, while CD4+ cells fail to proliferate in response to TCR/CD3 engagement.^247-249 Interestingly, thymic biopsies of these patients revealed that cortico-medullary demarcation and presence of Hassall’s corpuscles are preserved. However, absence of ZAP-70 might impact terminal differentiation of mTECs, given that the medullary area is reduced and expresses minimal amounts of thymic stromal lymphopoietin and involucrin.^250,251

Secondary non-genetic prenatal thymic defects

Maternal diabetes mellitus

Among nongenetic determinants of inborn errors of thymic stroma, uncontrolled maternal diabetes is the most common, making up 29% of a cohort of 105 recipients of thymus tissue transplantation.²⁵² Still, with gestational diabetes affecting up to 10% of the pregnancies in the United States, there are most likely additional vulnerability factors yet to be identified.²⁵³ Uncontrolled maternal diabetes can severely impair thymic development leading to long-term thymic hypoplasia or aplasia in mice, rats and humans.^254-260 The clinical phenotype often includes renal agenesis and features resembling DGS, such as cardiac outflow tract and vertebral and limb malformations, hypoparathyroidism, and hearing loss.²⁵⁵ Dysregulated RA may be involved in the disease mechanism. Cyp26a1, encoding an enzyme required for local inactivation of RA, is downregulated in embryos of diabetic mice.²⁶¹ Susceptibility to diabetes-associated thymic hypoplasia is potentiated in murine embryos with genetically impaired RA metabolism, an association yet to be examined in humans.²⁶¹. Altered RA signaling alone leads to thymic hypo- or aplasia, as discussed below.²⁶² Impaired thymic organogenesis in diabetic embryopathy was furthermore suggested to be mediated by impaired migration and organization of NCCs and derivatives, possibly via RA dysregulation, as discussed below. Aptly, NCCs play a role in organogenesis of most of the anatomical structures apart from the thymus that are also malformed in this phenotype.^255,258

Retinoic Acid

RA, a metabolite of vitamin A, is a crucial morphogen in vertebrates.²⁶³ Exposure of the fetus to RA, for example due to isotretinoin treatment for acne, can induce thymic hypo- or aplasia and further branchial arch defects resembling a DGS phenotype.^264-266 At the same time, low levels of RA during gestation may lead to the same phenotype. Mice with a hypomorphic allele of the RA-synthesizing enzyme Raldh2 or RA receptor knockout develop a DGS phenotype.^267,268 Accordingly, TEC development was shown to be impaired in vivo in mice and in vitro in human cells upon alteration of RA signaling.^23,269 The impaired TEC development may be mediated by dysregulated transcription factors. In mice and chicks, changes of RA concentration alter the expression of transcription factors regulating thymus organogenesis, namely Tbx1, Pax1 and Hoxa3.^262,270 Tbx1 and Chd7 were in turn reported to regulate the RA metabolism, illustrating possible feedback mechanisms.^271-273 Another mechanism that may contribute to the RA-associated impaired TEC development may involve NCCs. NCCs migrate from the posterior hindbrain and give also rise to a cardiac NCC population, which is required for septation of the outflow tract. These cells were disorganized in the aforementioned mouse model with hypomorphic Raldh2.²⁶⁸ In the same way, RA may interrupt migration and organization of the thymic NCC population, thereby affecting anatomic structure, epithelial mesenchymal crosstalk and eventually TEC development.

Alcohol

Since the first description of the fetal alcohol syndrome in 1973, there has been increasing evidence liniking intrauterine ethanol exposure to immune deficits.^274,275 Patients have decreased lymphocyte numbers and function and increased incidences of infections and malignancies.²⁷⁶ In humans, thymic development was reported to be abnormal upon fetal ethanol exposure, leading to thymic hypo- or aplasia.^276-278 In mice and rats, the teratogenic effects of ethanol on the thymus were confirmed, as well as the clinical consequences described in humans and the persistence of the immunological abnormalities into adulthood.^279-290 In nonhuman primates, intrauterin ethanol exposure lead to decreased numbers and impaired function in the T-cell compartment, a 22% mortality due to infectious diseases compared to 0% in the control group and significantly decreased vaccination efficiency.²⁹¹ Both in humans and in animal models, the phenotype can be similar to DGS, with abnormalities of heart, great vessels, and fore- and midbrain.²⁷⁷ Prenatal alcohol exposure has been shown to cause cell death, abnormal migration and disorganization within NCC populations.^279,292,293 Furthermore, RA may contribute to the detrimental effects of ethanol on the developing thymus. Chronic alcohol consumption in non-pregnant and pregnant women and animals results in the depletion of vitamin A stores in the liver and deficiency of vitamin A in the peripheral blood of the women.^294-297 In a study in rats, intrauterine ethanol exposure lead to lower levels of RA in fetal hearts.²⁹⁸ RA supplementation in zebrafish embryo gastrulation and somitogenesis stages also partially reversed some of the dysmorphic features associated with fetal alcohol syndrome, although the thymus was not examined in this study.²⁹⁹ Intrauterine ethanol exposure may hamper vitamin A metabolism, thereby reducing the availability of RA and mediate the phenotypic features.^{299-301 302,303} Moreover, disease mechanism may involve hormonal systems. Fetal alcohol exposure is known to permanently increase activity of the hypothalamic-pituitary-adrenal axis in humans. ³⁰⁴ Different levels of glucocorticoid receptor sites per thymocyte in ethanol-exposed compared with control rats in the first 2 months of life, suggesting that the glucocorticoid hormones may mediate changes.³⁰⁵

Novel tools for diagnosis and treatment of primary and secondary prenatal thymic defects

NBS evaluates T-cell output at birth via quantification of TRECs (circularized DNA excision products that are formed during rearrangement at the TRA locus in the thymus) from dried blood spots.³⁰⁶ NBS allows the identification of infants with severe T-cell-lymphopenia (TCL) prior to onset of clinical manifestations.³⁰⁷

Usually, genetic testing is performed after a positive NBS test to identify or rule out SCID.¹⁴⁹ This approach has allowed the discovery of many new genetic defects associated with TCL at birth. However, there is still a significant proportion of patients in which genetic analysis is not able to identify mutations in a known IEI gene, or when variant(s) are identified in a novel immune-related candidate gene but the function of the gene is not known and its pattern of expression (thymic stroma or hematopoietic cells) is unclear. Distinguishing between hematopoietic cell intrinsic and thymic intrinsic defects is of cardinal importance to guide treatment to either HSCT or to thymic tissue implantation. To help with this, in vitro systems allowing development of CD34+ cells into T-cells have been recently developed and perfected. Among these assays, artificial thymic organoids (ATO), in which CD34+ cells from either bone marrow or peripheral blood are purified and co-cultured with a murine stromal cell line expressing the Notch ligand DLL4, allow the differentiation of the CD34+ cells into mature TCRαβ+ CD3+ T-cells in 5–6 weeks in culture.^308,309 This system also allows definition of the precise stages at which blocks in T-cell development occur, and is a powerful system to distinguish between lymphopenia due to hematopoietic intrinsic defects versus thymic instrinsic defects of lymphopoiesis. This tool has also revealed unexpected differences between T-cell development in humans and mice. In particular, in Rag ^−/− micethymocyte development is blocked at DN3 stage³¹⁰, but the ATO system has revealed that patients with RAG deficiency can generate DP cells, although with reduced efficiency.³⁰⁸ However, the ATO system has some limitations. In this platform, CD34+ cells from ADA- and PNP-deficient patients develop normally into CD3+ TCRab+ cells because the stromal cells detoxify in trans the patient cells. Furthermore, while the ATO system can identify hematopoietic intrinsic defects of T-cell lymphopoiesis, it cannot prove the thymic origin of the TCL. This evidence can be reached by differentiating patient-derived iPSCs into TEPs³¹¹ and demonstrating abnormalities in their gene expression profile, as shown for patients with EXTL3¹⁹⁹ and PAX1¹⁹³ deficiency. However, this approach is laborious and time-consuming.

For patients with severe T-cell lymphopenia due to thymic stromal defects, cultured thymus implantation is the only treatment that can effectively reverse the immunological phenotype. The thymic tissue used in the transplants is obtained from thymuses that are removed during cardiothoracic surgery of babies with CHD but who are otherwise healthy. After removal, the tissue is sliced and cultured for 2-3 weeks to remove all T lymphocytes from the stroma before transplantation.³¹² The thymus slices are then implanted into the quadriceps muscle^131,252 where the recipient hematopoietic precursors will migrate, differentiate into mature T-cells and finally move to periphery. More than 100 subjects have received thymus implantation so far, with an overall survival around 75%.^252,313 Infections (including viral reactivation), autoimmunity and autoinflammation have been reported as major complications in several of these patients after transplantation.^252,313 While the number of peripheral T-cells often remains lower than normal after thymus implantation, the quality of the TCR repertoire shows good diversity and evenness. This treatment has obtained FDA approval in October 2021. Unfortunately, only two centers in the world offer the procedure, making the only curative treatment of congenital athymia hardly accessible. A possible alternative to cultured thymic tissue implantation would be to generate ex vivo transplantable thymic organoids. A recent paper showed that decellularized human thymic tissue can be used as a tri-dimensional scaffold in which to seed TEC progenitor cells that have been previously cultured in vitro.³¹⁴ If successful, transplantation of thymic organoids could represent a valid treatment option not only for patients with congenital athymia, but also for secondary defects of thymic function due to aging, chemotherapy, irradiation, GVHD, infections, and thymectomy.

Secondary postnatal thymic defects

Cytoreductive Therapy

Radiation and chemotherapy are intended to target malignant cells or hematopoietic cells prior to HSCT. However, they can simultaneously damage the normal process of T-cell development in the thymus. Even low-dose-radiation (5–200 mGy) can cause thymic atrophy, with decreased thymopoiesis relative to nonexposed controls.³¹⁵ Also, even a single sublethal irradiation dose can lead to long-term suppression of thymic function, primarily due to permanent impairment of BM-derived progenitors, of T-cell lineage commitment and of intrathymic T-cell differentiation, rather than to changes in the thymic stroma. In the thymus, all thymocyte populations are affected by irradiation, with the DP thymocytes being particularly sensitive.^{316 317} Following chemotherapy, up to 90% of patients show thymus atrophy in radiologic studies.^318,319 Thymic atrophy peaks at the time of maximal myelosuppression, suggesting a major contribution of the thymocyte compartment, rather than the stromal compartment.³¹⁸ Histologically, in mice there is almost complete loss of cortical lymphocytes four days after administration of cyclophosphamide.³²⁰ Fewer thymocytes immigrate to the thymus, where they have reduced developmental potential and higher rates of apoptosis. In the bone marrow, chemotherapy leads to downregulation of DLL4, resulting in myeloid skewing and depletion of lymphoid precursors.¹³ In the thymus of mice treated with alkylating agents, such as cyclophosphamide, total numbers of DP thymocytes and their immediate progenitors CD4−CD8− DN thymocytes are severely reduced, while numbers of thymocyte subpopulations at later stages of development are less severely affected.^316,321 Irradiation and chemotherapeutic agents also damage thymic stromal in both mice and humans.^13,322-324 Following cyclophosphamide administration, histological studies reported reduced total numbers of TECs and degenerative changes with cytoplasmic vacuolization.³²⁰ The induction of apoptosis in TECs after irradiation has been linked to the activation of signal transducer and activator of transcription 3 (STAT3) signaling. In the stroma, AIRE+ MHC-II^hi expressing mTECs are most affected by cell depletion, likely due to their higher rate of proliferation.^322,324-326 The resulting imbalance in TEC subsets persists even after thymic cellularity is restored and may link thymic atrophy upon cytoreductive therapies with the development of autoimmunity.^13,327,328

The damaging effect of cytoreductive therapy on the thymus is often prolonged, but mostly transient, although can sometimes be persistent. Still, the effects of radiation on thymic function can be long-term. At 2.5 years after mediastinal irradiation, patients with Hodgkin disease who did not receive chemotherapy showed decreased counts of peripheral blood naïve CD4+ and CD8+ T-cells.³²⁹ In patients with a history of acute lymphoblastic leukemia 8.7 years after radiation and chemotherapy RTE were decreased but T-cell diversity was preserved.³³⁰ Generally, the thymus has a great regenerative capacity following cytoreductive therapies.^318,319 Regrowth of the thymus after chemotherapy has been observed in 90% of younger patients.³¹⁸ It has been shown that certain stromal cell populations, such as innate lymphoid cells and endothelial cells, are less susceptible to chemotherapy and radiation, and as such promote endogenous thymic regeneration.^331-333 The efficiency of thymus regeneration depends on the dose of the cytoreductive agent administered and on the age of the patient. For elderly individuals, it may take years, if ever, to fully recover thymic function and restore the peripheral naïve T-cell repertoire.^334-337 A study reported irradiation-induced acute TEC loss to be more dose-sensitive in females mice, suggesting an influence of sex on susceptibility and regenerative capacity of the thymus.³¹⁷ The lack of data regarding sex differences of thymic damage should encourage future studies to explore this variable.

Steroids

Exogeneous glucocorticoids, due to their wide range of immunosuppressive and anti-inflammatory effects, are used in a great variety of conditions including autoimmune, allergic and inflammatory disorders, prevention of allograft rejection, and treatment of GVHD and lymphoid neoplasms.³³⁸ Both exogeneous administration and endogenous rise of glucocorticoids³³⁹ and other steroid hormons, like androgens^340-342, estrogens^343-345 and progestin³⁴⁶ can trigger a rapid and dramatic reduction in thymus size. Endogenous rise of androgens and estrogens may occur with age progression and of progestin during pregnancy. An endogeneous rise glucocorticoids is measurable upon emotional, physical or biological stress events, including infection and starvation, thereby offering a common mechanism underlying thymic atrophy upon different insults.^347-351 When adult mice are implanted with corticosterone-releasing pellets, inducing serum glucocorticoid levels similar to those observed during infections, the thymus size is reduced by more than 80%.³⁵² In an avian model, intense glucocorticoid treatment leads to similar effects as observed after surgical thymectomy, with a profound reduction of naïve T-cell count in the periphery. Three weeks after steroid treatment, T-cells expressing high-affinity choline transporter chT1, a marker identifying avian RTEs, were rarely found in the periphery.³⁵³ Both in humans and in an avian model³⁵⁴, administration of dexamethasone lead to a rapid decline of γδ T-cells. γδ T-cells have less capacity for peripheral expansion than other T-cell subsets, thereby suggesting that peripheral lymphopenia upon glucocorticoids is due to decreased thymic output, rather than peripheral loss.³⁵³ Glucocorticoids lead to depletion of the thymic thymocyte compartment and increased deposition of extracellular matrix in vitro and in vivo in mice.^18,355 Dexamethasone, a synthetic glucocorticoid analogue, activates phosphatidylinositol-specific phospholipase C via PKC, thereby generating diacylglycerol, which activates sphingomyelinase, resulting in caspase 3 and caspase 8 activation and apoptosis of mouse thymocytes.³⁵⁶ Lymphocyte glucocorticoid-dependent apoptosis varies with stages of differentiation and is most pronounced among DP thymocytes. In humans, this susceptibility is due to high expression of glucocorticoid receptor in DP thymocytes, promoting apoptosis induction via Apaf-1 and caspase-9.¹⁸ Another mechanism may be that DP thymocytes, in contrast to DN and SP thymocytes and peripheral T-cells, lack expression of Bcl-2, an antiapoptotic factor, and Notch, wich upregulates Bcl-2.³⁵³ Physiologically, the different susceptibility to apoptotic signals of thymocytes at different stages of differentiation may be necessary for proper positive or negative selection.¹⁸ TECs are able to produce glucocorticoids locally.³⁵⁷ Depending on the level of TEC-derived glucocorticoids, TCR-mediated thymocyte activation or deletion can be pursued, thereby setting the thresholds at which TCR avidity results in positive or negative selection.^358,359 Thymus atrophy induced by glucocorticoid treatment in the avian model is fully reversible, with immediate increase in thymic output after cessation of treatment and full recovery of RTE count after one month, suggesting that intensive steroid therapy for limited periods of time has no lasting effect either on lymphoid progenitor cells or the thymic stroma when thymopoietic activity is at its peak.³⁵³

Types and concentrations of steroids vary largely between males and females and there are differences in overall thymic output and TCR repertoire composition and diversity. Correspondingly, there is variation in the burden of disease with >80 of cases of autoimmunity occurring in females, while males are more susceptible to cancer and infections. This also translates into differences in the regenerative capacity after secondary thymic atrophy.¹⁴ For instance, the extent of secondary thymic atrophy upon radiation varies with sex in mice.³¹⁷ Possible mechanisms for such sex-related effect include that androgens are immunosuppressive, while estrogens are immunostimulatory; alternatively, sex steroids may have different effects on thymus cells in females and males.¹⁴ Sex steroids reduce the number of thymocytes migrating to the thymus.^360,361 In the thymus, estradiol administration impairs the differentiation of thymocytes by disrupting early-stage development of DN thymocytes; it also causes thymocyte apoptosis via the membrane receptor GPER1 and via the nuclear receptor ERα that inhibits NFKB signaling, the latter beeing required for thymocyte survival and proliferation.¹⁴ While these data indicate important effects of steroids on hematopoietic and lymphoid progenitor cells, and similar observations have been made upon use of androgens³⁶², studies using bone marrow chimera experiments and sex steroid-resistant mice show that sex steroid- induced secondary thymic atrophy is mostly due to effects on the stromal compartment.¹⁴ Indeed, androgen-induced thymus atrophy is mostly due to direct effects on stromal cells expressing the androgen receptor.^363,364 TECs from male mice, (which are exposed to higher androgen levels than female mice), are characterized by lower proliferation rates. In addition, androgen-dependent signaling represses the expression of Foxn1, Psmb11, and Ctsl, genes involved in cTEC development and function.³⁶⁵ High levels of androgens suppress thymic stroma expression of Il-7 and Ccl21, which promote thymocyte survival and proliferation, as well as Ccl25, encoding for a homing chemokine, and Dll4.^342,366 Furthermore, reduced apoptosis of DP thymocytes has been observed in the thymus of androgen-receptor knock-out mice, indicating that androgens may inhibit positive selection of thymocytes.³⁶⁷ The disparity of autoimmunity between females and males led to the idea that negative selection in the medulla might be affected by sex steroids. The growth of mTEC^hi cells is promoted by androgens, and the expression of Aire and TRAs is higher in males than in females. Upon engagement, the AR binds the Aire promoter and directly upregulates its expression. At the same time, estrogen treatment downregulates expression of Aire and of genes encoding TRAs by inducing methylation of the Aire promoter. Therefore, the androgen/estrogen ratio determines the level of Aire expression in mice.¹⁴ These findings suggest that effects of exogenously administered androgens and estrogens and of physiological sex differences on autoimmunity are mediated by signaling onto thymus stroma cells, and not so much by peripheral modulation of immune responses. Studies on the effect of sex steroid administration in gender transition are yet to be conducted.

Increased levels of progesterone cause physiological thymic involution during pregnancy. TECs are a main target of progesterone. Signaling through the progesterone receptor leads to downregulation of Ccl25 and Dll4 expression and upregulation of Aire, possibly enhancing negative selection and regulatory T-cell generation to ensure immunological tolerance of the fetus.

Immunosuppressive Drugs

Immunosuppressive agents such as cyclosporine A or alemtuzumab are administered to target peripheral immune cells in patients with allograft rejection, autoimmunity, or hematological malignancies. Still, rising evidence shows that the drugs also target the thymus, with potential deleterious hypofunction.

Cyclosporine A (CsA) is commonly used to prevent allograft rejection. It inhibits peripheral T-cell activation by blocking calcineurin-mediated dephosphorylation of the nuclear factors of activated T-cells and subsequent IL-2 transcription of activated T-cells.³⁶⁸ At the same time, it causes atrophy and damage in the thymus.^322,369-371 Total number of CD4+ and CD8+ SP and of DP thymocytes are reduced.³²² The thymic medulla disappears with loss of the AIRE+ mTEC^hi subset, while cTECs are relatively unaffected.^368,372 When thymi from CsA-treated mice were transplanted into athymic nude mice, recipients developed autoimmunity and, in a dose-dependent manner, low-dose CsA can induce autoimmunity in animals and humans.^371,373-375 The autoimmune manifestations are thought to be due predominantly to the loss of mTEC cells, resulting in impaired negative selection of thymocytes.^322,370,376 Of note, studies in mice suggest that the deleterious effects of CsA on thymic stroma are reversible, as Aire expression is restored within 7 days, and the mTEC^hi subset is fully recovered within 10 days after interruption of CsA administration.³²²

Alemtuzumab, an humanized anti-CD52 monoclonal antibody, is a lymphocyte-depleting agent approved for relapsing remitting multiple sclerosis (RRMS) and leukemia. Given that TECs also express CD52 and that incomplete T-cell repertoire renewal and reduced thymopoiesis were reported upon alemtuzumab administration, it is tempting to speculate that thymic injury may be the reason for secondary autoimmune manifestations that have been reported in at least 50% of the RRMS patients after treatment with alemtuzumab.^13,377,378 Although the mechanism remains speculative, it is noteworthy that a study in which RRMS patients received both alemtuzumab and keratinocyte growth factor, a factor that, if administered alone, promotes TEC proliferation, reported deleteriously reduced thymic output.^13,377 (NCT01712945)

Malnutrition

Malnourished children have increased risk of dying from infectious diseases, contributing to malnutrition beeing the underlying cause of 45% of global deaths in children below 5 years of age.³⁷⁹ Malnutrition leads to atrophy of lymphatic tissue, particularly the thymus.¹⁵ The size of the thymus is an independent and strong predictor for survival in children in populations in West Africa and Southeast Asia.³⁸⁰ Thymus atrophy upon malnutrition was reported in autopsies to an extent termed “nutritional thymectomy” in severely malnourished children. Thymus atrophy has been reported even in mild malnutrition prospective studies using imaging to measure thymus size.¹⁵ Poor nutrition causes depletion of the intrathymic lymphocyte compartment by apoptosis, particularly of the DP cells.^15,347 Accordingly, even in mildly undernourished mice, there is a significant change of the intrathymic cytokines, with reduced levels of IL-6, IL-2 and IL-4 and increased levels of IFN- γ, IL-10 and TNF-α.³⁸¹ The degree of thymocyte depletion in severely malnourished mice and children correlates with an increase of thymic intralobular connective tissue composed of fibronectin, laminin, and type IV collagen.³⁴⁷ In both undernourished animals and children, the volume of the cortical and medullary epithelial tissue is decreased, and cortico-medullar differentiation is diminished.^15,347 Breastfeeding was associated with larger thymuses compared to formula feeding, possibly due to the presence of human IL-7 in breast milk.³⁸² Malnutrition-induced thymic atrophy may be mediated by hormonal factors. Leptin, prolactin, and growth hormone promote thymic growth and function and their levels are decreased in malnourished children.^383-385 Low leptin levels correlate with a higher risk of death in malnourished children.³⁸⁶ Importantly, nutritional rehabilitation may allow normalization of the thymus size even in severely malnourished children.^387,388 Thymic atrophy does not only occur in protein-energy deficiency but also in isolated deficiencies in vitamins and micronutrients, especially zinc. Zinc-deficient mice manifest with thymic atrophy and hypocellularity of the thymocyte compartment, possibly mediated by a systemic rise of glucocorticoids.³⁸⁹

Interestingly, although not further discussed here, diet-induced obesity can also cause secondary thymic atrophy due to thymocyte apoptosis, a possibly contributing factor to the increased rates of cancer and infections observed in obesity.³⁹⁰

Infections

Thymic export of naïve T-cells with a broad TCR repertoire is essential for effective immune response to infections.¹ However, infectious diseases can lead to severe thymus atrophy with impaired thymocyte development. The degree of thymus atrophy correlates with TREC levels in peripheral blood, as shown in SIV-infected macaques.^{391 392}

Depletion of the thymic stromal compartment upon infection was reported after bacterial (Francisella tularensis, Salmonella typhimurium), viral (HIV, simian immunodeficiency syndrome, rabies, SARS-CoV2), parasite (T. cruzi, plasmodium chaubi, schistosoma mansoni, trichinella spiralis), and fungal (paracoccidioides brasiliensis, histoplasma capsulatum) infections.^392-404 In the thymic stromal compartment, upon infection, total numbers of both cTECs and mTECs decrease in a mouse model of Trypanosoma cruzi infection.³⁹² Also, infections lead to increased deposition of extracellular matrix.^405-408 Infections may alter the expression of TEC-specific markers, such as CK18 and TR5, as well as expression of chemokines such as CXCL12 and CCL4. The expression of genes regulating proliferation, cell cycle, differentiation and migration is altered, as seen for example in in vitro models of measles virus infection, both in vivo and in vitro upon ZIKA virus infection and during SARS-CoV-2 infection.^409-412 Of note, all these three viruses can infect TECs directly. For example, SARS-CoV-2 enters TECs via their expression of angiotensin-converting enzyme 2. Chronic viral infections that directly infect TECs such as HIV and CMV have especially detrimental and lasting effects on TECs. HIV also infects dendritic cells in the thymic stroma.^413-416 TEC infection by these pathogens cause disorganization of TECs, resulting in disrupted thymic architecture and loss of a functional microenvironment.⁴¹⁷

In the thymic thymocyte compartment, massive apoptosis affects mainly DP thymocytes and their immediate precursors while other thymic subsets such as mature SP cells are less vulnerable.^{13,399,417-419} A key mechanism suggested is the rise in endogenous glucocorticoids as part of the organisms’ stress response.^18,392 DP thymocytes are particularly sensitive to increased levels of glucocorticoids, providing a potential mechanism of selective depletion of that subset of cells.⁴²⁰ Adrenalectomy and/or inhibition of the glucocorticoid receptors prevented intrathymic loss of thymocytes in mouse models of rabies or Trypanosoma cruzi infection.^421-423 Still, the depletion of intrathymic lymphocytes was not or was only partially reduced in adrenalectomized mice infected with mouse hepatitis virus, Trypanosoma cruzi, Francisella tularensis, or Listeria monocytogenes, suggesting that infection-induced apoptosis of thymocytes is not simply mediated by high endogenous glucocorticoid levels.^397-400,418

Further mechanisms involved in infection-induced thymus atrophy include the production of other systemic stress mediators, such as prolactin as demonstrated in Trypanosoma cruzi infected mice.⁴²⁴ Furthermore, increased levels of pro-inflammatory mediators such as IL-6, IFN-γ and TNF-α may also contribute to bone marrow and thymus atrophy, and loss of DP thymocytes.^425-427 Levels of TNF-α, for example, increase during infection with Francisella tularenis, Trypanosoma cruzi, and Salmonella enterica infection.^399,422,428 In Francisella tularensis infection-induced thymus atrophy and thymocyte depletion were prevented in mice deficient in TNF receptors 1 and 2.³⁹⁹ In humans, it has been suggested that increased levels of IL-6, TNF-α and IFN-γ may be the cause of profound peripheral lymphopenia observed in patients infected with SARS-CoV-2.^429,430 Furthermore, soluble virulence factors and endotoxins may play a role in infection-associated thymic atrophy. Trans-sialidase, an enzyme that trypanosome cruzi sheds into the bloodstream, activates the purinergic receptor P2X, thereby inducing thymocyte apoptosis.^431,432 In another study, lipopolysaccharide injection, which can be released from gram-negative bacteria such as Escherichia coli, caused thymic atrophy.^433-435 Direct infection of thymocytes by pathogens may also be detrimental for thymic function. HIV, for example, directly infects CD4+ SP thymocytes and progenitor cells, causing the release of cytotoxic viral products and apoptosis of both infected and uninfected thymocytes.^436,437

After resolution of the infection, thymic atrophy is reversed typically within 2 weeks.⁴³⁸ However, the regenerative capacity of an aged or pre-damaged thymus is reduced. Furthermore, chronic or severe infections such as HIV, have a lasting impact on thymic function and T-cell output.^439,440 There is a reduction of TRECs, peripheral naïve CD4+ T-cells, abnormal ratios of CD4+ T-cells to CD8+ T-cells and of naïve to memory T-cells, expansion of senescent T-cell populations and persisting abnormalities of the TCR repertoire.^441,442

Autoimmunity

Autoimmunity can be both a cause and a consequence of thymic atrophy. Human rheumatoid arthritis, juvenile idiopathic arthritis, and multiple sclerosis are correlated with thymic atrophy, as documented by decreased TREC levels and export of immature CD4 RTEs compared to age-matched controls.^108,443-447 In mouse models, autoimmune disorders, including experimental autoimmune encephalomyelitis, a model for MS, preceded thymic atrophy and decline in thymopoiesis.⁴⁴⁸ Activated T-cells recirculate into the thymus from the periphery. In in vitro organ cultures, activated T-cells, especially CD4+ T-cells, inhibit thymocyte development and proliferation. Thymic transplantation experiments suggest that declined thymopoiesis is due to severe perturbation of TECs after exposure to the activated T-cells, possibly mediated by excessive RANK signaling.⁴⁴⁹

Malignancy

The broad T-cell repertoire generated by thymic selection enables T-cells not only to detect foreign antigens, but also neoantigens arising from cancer cells. Consequently, prolonged thymic atrophy with reduced TCR diversity correlates with rising risk for malignancies.^11,450 At the same time, cancer can lead to thymic atrophy. Tumor-bearing mice develop thymic atrophy, independent of chemotherapy or other therapeutic interventions.⁴⁵¹ Depletion of DP cells has been observed in a mouse model of mammary tumor, and a possible mechanism is represented by downregulation of Bcl-XL and A1 in thymocytes (two anti-apoptotic members of the Bcl-2 family).⁴⁵² In addition, the thymus of mammary tumor bearing mice displays increasing disorganization of the stromal microenvironment ⁴⁵³, with functional impairment and reduced proliferative capacity of TECs.⁴⁵⁴ Another mechanism by which malignancy can lead to thymus atrophy may be insufficient provision of IL-7 by the stroma. Indeed, the phenotype of cancer-associated thymus atrophy resembles what seen in IL-7- or IL-7R-deficient mice, namely decreased thymic cellularity, and early developmental arrest at the pro-T 1 CD44+25- precursor stage.⁴⁵⁵ Furthermore, it was suggested that VEGF, elevated in advanced stage cancer patients, blocks the proliferative and migratory potential of thymocyte progenitors, leading to reduced numbers of the earliest observable thymocytes in the thymus.⁴⁵⁰

Graft versus host disease

Allogeneic HSCT offers a potential therapy for various malignant and non-malignant disorders. Still, the outcome is, besides prolonged post-transplantation immunodeficiency, limited by the morbidity and mortality associated with GVHD.⁴⁵⁶ In acute GVHD, activated donor T-cells trigger an inflammatory cascade.⁴⁵⁷ Beyond damage to intestines, liver, and skin, the alloreactive T-cells also directly target the bone marrow and the thymus.^457,458 Accordingly, acute GVHD leads to delayed immune reconstitution.^116,459-463 Besides the damaging effect on the bone marrow stroma, leading to decreased thymic input^464,465, acute GVHD causes severe loss in both the thymic thymocyte and the stromal compartments.^466,467 Structural changes include the disappearance of Hassall’s bodies and a loss in the distinction of cortex and medulla. TRECs and TCR diversity are decreased.⁴⁶⁸ Abnormalities of the thymic thymocyte compartment upon GVHD are primarily consisting of loss of DP thymocytes via glucocorticoid-independent T-cell death in mice.⁴⁶⁶ In addition, pro- and pre-T-cells are impaired in cell cycle progression.^468,469 TECs are also targets of GVHD-mediated apoptosis, irrespective of whether a conditioning regimen is used and of its nature.^467,468 TEC apoptosis may be mediated by STAT-1 driven IFN-γ secretion by alloreactive T-cells.⁴⁶⁷ Of note, TECs, in their function as antigen-presenting cells, can prime alloreactive T-cells directly.⁴⁶⁷ Donor alloreactive CD8+ T-cells preferentially target tolerance-inducing mature mTECs, causing aberrant TCR selection and persistence of autoreactive T-cells.^468,469 These autoreactive may cells link alloimmunity in the thymus during acute GVHD to the development of autoimmunity seen in chronic GVHD.^112-114,470

Chronic GVHD impairs the capacity of the thymus to recover. A mouse model showed that thymic regeneration may be impaired because alloreactive T-cells eliminate IL-22-producing thymic innate lymphoid cells. These innate lymphoid cells play an important role in promoting thymus recovery.⁴⁷¹

Novel tools for diagnosis and treatment of secondary postnatal thymic defects

Evaluating thymic function remains challenging, as all available markers are proxies. Procedures vary between different centers and studies.

RTEs, the youngest subset of naïve T-cells and a marker for thymic output, are most commonly characterized by the co-expression of CD45RA and CD31 by CD4+ cells. Age-specific standard values are available.¹²⁰ CD45RA can also be combined with phenotypical markers such as complement receptors 1 and 2, CCR7, PTK7, or CD103.¹³ However, homeostatic proliferation upon IL-7 stimulation without downregulation of CD31 on one hand, and CD31 re-expression by activated T-cells on the other hand, may affect the accuracy of estimating thymic output.^472,473

TRECs represent a useful tool to assess thymic function and immune reconstitution. However, the longevity of naïve T-cells and peripheral T-cell divisions warrants a cautious interpretation of this data.^474,475 In the presence of GVHD with altered homeostasis of naïve peripheral T-cells, for example, a mouse model showed the importance of interpreting TRECs in combination with assessment of cell division, for example by measuring Ki-67 expression, a proliferation associated antigen, CFSE labeling, or examination of telomeres.⁴⁷⁶

TCR diversity is an indirect measurement of thymic function. At the protein level, TCR diversity can be measured by flow cytometry quantifying the use of different TCR variable (V) gene families.⁴⁷⁷ At the molecular level, spectratyping allows determination of the clonal length in the third complementarity region (CDR3) of each TCR V gene family.⁴⁷⁸ The frequency of individual CDR3s can be obtained by high-throughput sequencing of TCR repertoire.⁴⁷⁹ Noteworthy, restrictions of the TCR repertoire with aging are aggravated in response to influenza and SARS-CoV-2 infections.^480,481

The thymic volume can be measured by ultrasound, especially used in low resource settings, and by computer tomography or magnetic resonance imaging, which were used in studies to evaluate regeneration after radiation, chemotherapy and HSCT (fig. 2C).^{20,318,334,482} However, imaging can neither be used to diagnose primary athymia, as lack of thymus visualization may also be secondary to hematopoietic-intrinsic forms of SCID, nor to assess secondary forms of thymic hypoplasia, as the size of the thymus and thymic output do not follow a linear correlation.^13,483

Despite thymus sensitivity to a wide range of potentially harmful stimuli, the organ has a remarkable capacity for repair.^331,484,485 The first therapeutic approaches to enhance thymic function were developed from exploiting pathways of endogenous regeneration (fig. 4).⁴⁸⁶ However, at present, there are still no approved therapies to enhance thymic function.¹³ Here below, we discuss approaches that target the thymocyte and thymus stromal compartment to promote thymic regeneration. Tools to enhance pre-thymus lymphopoiesis are reviewed elsewhere.¹³

Fig. 4: Selected endogeneous and exogeneous mechanisms of thymic regeneration in humans.

Involved factors are assigned to the producing cell type, either from the thymic stromal compartment (yellow) or from the thymocyte compartment (blue). Arrows indicate the target cells. Factors in pink were administered in clinical trials.

One of the earliest studies on mechanisms of endogenous thymic regeneration focused on keratinocyte growth factor (KGF), also known as FGF7⁴⁸⁷. KGF targets TECs and is primarily produced by fibroblasts.⁴⁸⁸ KGF binds to the FGF2 variant IIIb on TECs, activates Phosphoinositide 3-kinase (PI3K)–protein kinase B (PKB)–NFKB and p53 pathways, promoting TEP differentiation and TEC proliferation.^468,489,490 Studies in KGF knockout animals reported impaired endogenous thymic regeneration after damage caused by irradiation, HSCT and GVHD.^468,487 Administration of rKGF limits thymic damage and promotes thymic regeneration in mouse models exposed to irradiation^487,491,492, cyclophosphamide^487,491, dexamethasone⁴⁸⁷ or conditioning regimes or in case of GVHD^468,470 or age-related thymic involution⁴⁹³. rKGF even promoted TEC proliferation in vivo in young mice with unimpaired thymic function⁴⁸⁹. In rhesus macaques, KGF reduces thymic atrophy by limiting DP thymocyte apoptosis and improves immune reconstitution after HSCT with increased TREC levels.^494,495 Human rKGF (palifermin) is an FDA-approved drug to prevent mucositis due to high-dose chemotherapy. In a phase I study, a single dose of palifermin pre-HSCT decreased the risk for GVHD in a dose depended manner, possibly mediated by the effect of palifermin on the thymus. ⁴⁹⁶ (NCT02356159) Multiple clinical trials are underway and may inform about the actual effect of rKGF on thymic T-cell reconstitution, but the results have not yet been published. (NCT01233921, NCT03042585, and NCT00593554)

The loss of DP thymocytes during thymic damage can lead to the upregulation of IL-23 by thymic dendritic cells, which in turn promotes IL-22 upregulation by innate lymphoid cells.^331,497 IL-22 then activates STAT3 and STAT5 and the expression of the downstream antiapoptotic molecule MCL1 in TECs, thereby promoting TEC survival and proliferation.^{331,471,498,499} Administration of rIL-22 to mice whose thymus has been damaged by irradiation, GVHD or chemotherapy promoted the recovery of thymic function.^{331,471,498,500,501} A phase IIa clinical study showed safety and tolerability of human rIL-22 in conjunction with systemic corticosteroids in patients with acute GVHD (NCT02406651).

Receptor activator of NFKB ligand (RANKL) is a member of the tumor necrosis factor (TNF) superfamily and is a key contributor to endogenous thymic regeneration.⁵⁰² RANKL gets upregulated on CD4+ thymocytes and innate lymphoid cells shortly after thymic injury.^331,500 In mice, the administration of RANKL stimulates thymic regeneration after irradiation and, especially in the setting of age-related thymic involution, enhances immune reconstitution after HSCT.⁵⁰⁰ RANKL binds to RANK on innate lymphoid cells, promoting generation of LT-α. LT-α binds the LT-β receptor on TECs and TEPs, thus enhancing their regeneration. LT-α is critical for thymic regenerative effects and thymocyte homing induced by administration of RANKLs in mice.⁵⁰⁰ In addition, RANKL binds to mTECs, activates the NFKB2 signaling pathway and promotes AIRE expression.⁵⁰³ RANKL also stimulates TEC production of IL-7.⁵⁰⁴ IL-7 has a regenerative effect in the thymus and may integrate various of the above-described regenerative pathways.¹³ IL-7 targets HSPCs, thymocytes, and peripheral T-cells and is a pro-survival factor for innate lymphoid cells.⁵⁰⁵ It stimulates proliferation and function of cells that express RANKL, IL-22, LT-α and LT-β ^506,507 and can upregulate RANKL expression by T-cells ⁵⁰⁸. Activation of the IL-7 receptor promotes cell differentiation through JAK-STAT signaling, and proliferation and survival through the PI3K-PKB pathway.⁵⁰⁹ IL-7 has been shown to be effective at boosting thymic function in both aged mice as well in a preclinical model with acute thymic damage induced by conditioning.^510-516 In a phase I clinical trial (NCT00684008) in recipients of allogeneic HSCT, administration of rIL-7 led to persistent increase of peripheral CD4+ and CD8+ T-cell numbers, RTE (although only in young patients), and increased TCR diversity.⁵¹² A clinical trial to evaluate the effect of rIL-7 on T-cell reconstitution in recipients of cord blood HSCT has not yet published results (NCT03941769). Increased thymic function upon IL-7 administration has also been reported in HIV-1-infected and in cancer patients.^517-521 Exogenous IL-7 furthermore increases antigen-specific T-cell responses to vaccines and viral infections.^522,523 In mouse models, exogenous IL-12 can reverse age-related thymus involution and promote engraftment and hematopoietic reconstitution after transplantation, especially after irradiation.^524,525 Some of these effects of IL-12 may be mediated through upregulation of IL-7 and IL-2, the latter being indispensable for developing regulatory T-cells, and thereby maintenance of central tolerance.^524-526 Another chemokine that improves thymic regeneration and reconstitution in mouse models is exogeneous IL-21, potentially mediated through activation of BCL-6, an anti-apoptotic gene required for early intrathymic thymocyte development. IL-21 primarily acts on DP thymocytes, which upon secondary thymic injury upregulate expression of the IL-21 receptor, a potential pathway of endogenous regeneration.^527,528 In case of thymic damage, endothelial cells produce BMP4, which binds to its receptor expressed on TECs and induces the upregulation of FOXN1 and its target genes, such as DLL4.^332,342,529 BMP4, administered by transfer of ex-vivo expanded thymic endothelial cells, improved thymic reconstitution in mice after sublethal total body irradiation, suggesting a potential clinical approach to support thymic regeneration and an innovative approach to deliver therapeutic factors specifically to the thymus.³³² FOXN1 is crucial for postnatal thymic function and maintenance.⁴¹ Forced expression of Foxn1 leads to regeneration of the aged thymus.^77,78 After thymic damage, Foxn1 expression correlates positively with IL-22 expression. IL-22 is, as discussed, an endogeneous repair cytokine, suggesting a role of Foxn1 in thymic regeneration. Accordingly, rFoxn1 administration leads to increase in TEC numbers and thymic output in mice after HSCT and in mice with age-related thymic involution.^77,530

Thymosins are immunomodulatory, low molecular weight peptides.^531,532 TECs produce thymosin-α1, which promotes thymocyte survival, proliferation, maturation and regeneration after insults.¹⁸ Thymosin-α1 can prevent dexamethasone-induced apoptosis of DP thymocytes in vitro, as well as glucocorticoid-induced thymus atrophy in vivo in preclinical models, possibly by upregulating IL-7 from TECs.^533,534 A phase I/II clinical study demonstrated that administration of Thymosin-α1 was safe and efficacious in recipients of allogeneic HSCT, as shown by an increase in T-cell numbers and earlier effective pathogen-specific T-cell responses (NCT00580450).⁵³⁵ In patients with SARS-CoV-2 and severe lymphopenia, thymosin-α1 promoted thymic regeneration, and improved TREC levels and patient survival.⁵³⁶

Administration of GH or its principal mediator IGF-1, can reverse thymic atrophy upon injury or age in mouse models.^537-541 The IGF-1 receptor is expressed by both the thymic stromal and thymocyte compartments and activates the JAK2/STAT5 pathway.¹⁸ Administration of rGH to mice post-HSCT leads to early but transient increase in CD8+ and CD4+ T-cells in the periphery and increased thymic cellularity, but no increase in TREC levels and no effect on GVHD development.⁵³⁹ A phase 1 clinical trial of GH administration post unrelated cord blood transplant showed safety and efficacy (NCT00737113). In HIV-1-infected patients administration of GH lead to, although transiently, increased thymic volume and function, quantified by TREC levels and numbers of total and naïve CD4+ T-cells (NCT00071240).^537,538 A pre-phase-I study assessed the effect of exogeneous GH on age-related thymus involution in patients between 51 and 65 years. In more than half of the cohort, the intrathymic fat tissue was replaced with regenerated tissue, and thymus size, along with the total numbers of peripheral naïve T-cells, increased.⁵⁴² A follow-up expanded trial is ongoing (NCT04375657). While these are promising results, severe side effects of GH administration, including cardiovascular events, diabetes and cancer, should advise caution.¹³

Sex steroid ablation in mouse models reverses secondary thymic atrophy, in both thymocyte and stromal thymic compartment.^{321,485,543-545} However, the regeneration did not last after the end of suppression. The effects of sex steroid ablation are mediated via the upregulation of thymopoietic factors in TECs, namely CCL25 and DLL4.^342,366 In humans, drugs that reversibly block sex steroids are well established in the treatment of precocious puberty, endometriosis and hormone-sensitive cancer. Sex steroid blocking for cancer treatment was shown to increase naïve CD4+ cells, TREC levels, and increased TCR diversity.^544,546 Studies in which sex steroids were blocked prior to HSCT and/or cytoablative therapy confirmed improved thymic output.^{360,361,485,547,548} Two ongoing trials examine immune reconstitution after HSCT upon sex steroid blocker mediated by luteinizing hormone-releasing hormone, leuprolide, and degarelix. (NCT01746849, NCT01338987) Another hormone, leptin, reversed endotoxin-induced thymic atrophy in mice. Leptin promotes TEC survival mediated by upregulation of IL-7 and reduced apoptosis, while increasing profliferation of DP cells. Potentially, the effects are mediated by decreasing cytosolic phospholipase A2 and the p38 MAPK signaling pathway.⁵⁴⁹

Following damage of the thymus due to irradiation or HSCT conditioning, zinc ions are released from thymocytes into the extracellular space where zinc upregulates endothelial cell production of BMP4 via the cell surface receptor GPR39. This promotes TEC regeneration via Foxn1. Consistent with these observations, thymi from zinc-deficient mice have less BMP4 protein.⁵⁵⁰ In aged mice, zinc supplementation leads to partial recovery of thymic involution.⁵⁵¹

Conclusion

The detection of thymic aplasia improved with the implementation of NBS for SCID. Artificial thymic organoids (ATOs) can help to distinguish between hematopoietic-intrinsic versus thymic intrinsic cause of T cell lymphopenia but they are not useful to differentiate genetic versus aquired causes of congenital athymia. For this purpose studies are needed that examine genes that are so far of unknown significance in human thymus organogenesis but may, if mutated, cause primary thymic defects. Despite the limitations of the ATOs approach, it is of cardinal importance to guide treatment towards thymic tissue implantation in all cases of thymic aplasia irrespective of the etiology.

The thymus is susceptible to various insults, including cytoreductive and immunosuppressive therapies, steroids, malnutrition, infection and autoimmunity, resulting in secondary atrophy. Both secondary thymic atrophy and age-related involution lead to decreased naïve T-cell output and T-cell receptor diversity. The detection of secondary thymic atrophy is limited by the unavailability of a comprehensive biomarker of thymic function. This makes secondary thymic atrophy an under-detected condition with potentially severe consequences. Even though the self-renewal capacity of the thymus is remarkable, it may be insufficient in elderly individuals and in case of chronic or severe insults. Secondary thymic atrophy due to chronic infections may infact generate a self-reinforcing cycle by leading to impaired thymic function that further contributes to infectious susceptibility. Similarly, acute GVHD may lead to an impaired thymic selection that predispose to autoimmunity.

At present, there are no approved therapies to enhance thymic regeneration. Clinical trials are ongoing, with IL-7 and KGF being among the most promising candidates. Such therapies could not only benefit patients undergoing medical interventions, such as HSCT, but also in patients with chronic conditions, such as infections or GVHD; moreover they could improve vaccination efficiency and age-related errors of immunity in the elderly.

Abbreviations

ADA

adenosine deaminase

AIRE

autoimmune regulator

APECED

autoimmune Polyendocrinopathy, Candidiasis, Ectodermal dystrophy

APS

autoimmune Polyglandular Syndrome

ATO

artificial thymic organoid

BMP

bone morphogenetic protein

CCL

C-C motif chemokine ligand

CDR3

third complementary region

CHD

congenital heart disease

CID

combined immune deficiency

CK

cytokeratin

CsA

cyclosporine A

cTEC

cortical thymic epithelial cell

CXCL

C-X-C motif chemokine ligand

DGCR

DiGeorge Critical Region

DGS

DiGeorge syndrome

DLL4

delta like Notch ligand 4

DN

double negative

EXTL3

exostosin-like 3

FGF

fibroblast growth factor

fig

figure

FOX

forkhead box

GVHD

graft versus host disease

HSCT

hematopoietic stem cell transplantation

HSPG

heparan sulphate proteoglycan biosynthesis

IFN

interferon

IGF

insulin-like growth factor

IL

interleukin

iPSC

induced pluripotent stem cells

K

keratin

KGF

keratinocyte growth factor

LOF

loss-of-function

LT

lymphotoxin

MHC-I, MHC-II

major histocompatibility complex class I, class Iinf

mTEC

medullary thymic epithelial cell

NBS

newborn screening

NCC

neural crest cell

NFKB

nuclear factor kappa-light-chain-enhancer of activated B cells

OTFCS

otofaciocervical syndrome

PI3K

phosphoinositide 3-kinase

PK

proteinkinase

r

recombinant

RA

retinoic acid

RAG

recombination-activation gene

RANK

receptor activator of NFKB

RANKL

receptor activator of NFKB ligand

RRMS

relapsing remitting multiple sclerosis

RTE

recent thymic emigrant

SCID

severe combined immunodeficiency

SP

single positive

STAT

signal transducer and activator of transcription

TCL

T-cell-lymphopenia

TCR

T-cell receptor

TEC

thymic epithelial cell

TEP

thymus epithelial progenitor cell

TNF

tumor necrosis factor

TRA

tissue-restricted antigen

TREC

T-cell receptor excision circle

Data Availability Statement

Data sharing is not applicable to this article as no new data were created or analyzed in this study.