Ubiquitination is a post-translational mechanism of protein modification involved in multiple signaling pathways (1, 2). The 76-amino acid ubiquitin is covalently conjugated to lysine (K) residues of target proteins catalyzed by ubiquitin-activating (E1), ubiquitin-conjugating (E2), and ubiquitin-ligating (E3) enzymes (3, 4) ( Figure 1 ). Ubiquitin molecules are conjugated to target proteins either singly (monoubiquitination) or in the form of polyubiquitin chains. Polyubiquitin chains are formed by isopeptide bond connection between the carboxyl-terminal glycine residue of ubiquitin and seven internal K residues (K6, K11, K27, K29, K33, K48, K63) or the amino-terminal methionine (M1) of another ubiquitin (5). The various ubiquitin chains play different roles in cellular process while K48- and K63-linked ubiquitin chains are most extensively studied. K48- and K11-linked ubiquitin chains mainly target substrates for degradation by 26s proteasome (6). K63- and M1-linked ubiquitin chains mainly induce signal transduction and DNA repair. Deubiquitinases (DUBs) mediate the reversible modification by cleaving the conjugated ubiquitin from target peotein (7).The substrate specificity of ubiquitination is mainly determined by E3s, which mediate the binding of ubiquitin molecules to target proteins directly or catalyze the transfer of ubiquitin molecules from E2 conjugating enzymes to substrate proteins (5, 8). Over 700 mammalian E3 ubiquitin ligases were recognized (3). The most studied E3 ligase subfamilies are really interesting new gene- (RING-) and homologous to E6AP C terminus- (HECT-) domain containing ligases (9). E3s with RING domain catalyze the transfer of ubiquitin from E2 to substrate (10), while E3s with HECT domain transfer ubiquitin to substrate through covalent thioester intermediate between ubiquitin and the cysteine residue of HECT domain (4, 11).

T cells are implicated in multiple immune processes, including immune response to pathogens, tumors and also in inflammation, immune homeostasis, immune memory and self-tolerance maintenance (12, 13). Fetal liver or bone marrow derived thymus-seeding progenitors (TSPs) settle in thymus, developing into early thymic progenitors (ETPs), which give rise to γδT and αβT cells (13–15). Under the driving of the specialized thymic microenvironments, ETPs proliferate extensively and undergo several selection processes to mature into distinct functional T cell lineages. For the best-studied conventional CD4⁺ and CD8⁺ αβT cells, ETPs develop into non-self-reactive T cells with the ability to recognize MHC-antigen complexes along the CD4^-CD8^- double negative (DN), CD4⁺CD8⁺ double positive (DP) and CD4⁺ or CD8⁺ single positive (SP) stages (16–18) ( Figure 2 ). During thymocyte development, the strength of TCR-peptides-MHC interaction decides the cell fate (19, 20). Weak or absent TCR-peptides-MHC interaction leads to cell death, and low or moderate affinity of TCR-peptides-MHC interaction promotes T cell development through positive selection. High-affinity TCR-peptides-MHC interaction triggers the death of self-reactive thymocytes during negative selection. The combined positive selection and negative selection ensure the production of the conventional naïve CD4⁺ and CD8⁺ αβT cells. However, some self-reactive thymocytes survive and develop into regulatory T cells (Tregs) or unconventional T cells including invariant natural killer T (iNKT) cells, natural TCRαβ⁺ CD8αα⁺ intraepithelial T cells and natural T helper 17 (nTh17) cells, through a process called agonist selection (21, 22).

The ordered and sequential differentiation and selection during thymocyte development are tightly regulated and correlated with the generation of non-self-reactive T cells with functional TCRs. In the past few decades, increasing studies demonstrated the fundamental role of ubiquitination regulation in T cell development. Figuring out the ubiquitination regulatory mechanisms of T cell development not only enhances our understanding of how T cell fate is determined, but also provides potential new therapeutic approaches for T cell dysfunction-associated and autoimmune diseases. It is worth noting that drugs targeting the ubiquitin-proteasome system has shown promise in the treatment of autoimmune disorders and cancers, such as bendamustine (targeting LUBAC), thalidomide (targeting CRBN), and mitoxantrone (targeting USP11, USP15) et al. (23–25). In this review, we discuss the importance of ubiquitination in T cell development ( Figure 3 ) and self-tolerance establishment ( Figure 4 ), with an emphasis on the mechanism studies of T cell progenitors, thymic selections, proliferation, apoptosis, Treg development and TEC functions.

CD34 serves as a marker for human stem cells and progenitor cells, labeling T cell precursors in the thymus (26). Human DN thymocytes are divided into three subsets based on the expression of CD34, CD7, and CD1a, which correspond to the developmental order: CD34⁺CD7^-CD1a^- (Thy1), CD34⁺CD7⁺CD1a^- (Thy2), and CD34⁺CD7⁺CD1a⁺ (Thy3) (27–31)( Figure 2 ). The expression of CD1a marks T cell lineage commitment, and Thy1 cells are the fewest and the earliest T cell precursors with multi-lineage differentiation potential including T cells, B cells, NK cells and myeloid cells in vitro ( 27, 30, 32). Recent transcriptomics studies revealed the heterogeneity of Thy2 cells. Based on the expression of CD44 and CD2, Thy2 cells are divided into three subsets, CD44⁺CD2^- (Thy2a), CD44⁺CD2⁺ (Thy2b) and CD44^-CD2⁺ (Thy2c) (33). CD2 expression marks the stage of T lineage commitment and the multi-lineage differentiation potentials. CD2^- Thy2a cells are most similar to Thy1 cells, with multi-lineage differentiation potential including T, B, NK and myeloid cells in vitro. CD2⁺ Thy2b cells exhibit T, NK and myeloid cell differentiation potential, but no longer B cell differentiation capability (33). Downregulation of CD44 represents lineage commitment of T cells, thus Thy2c and Thy3 are T-lineage committed and have no other lineages differentiation potentials (33, 34). CD34⁺CD38^- cells have also been shown to contain the earliest progenitors in thymus. CD38 is expressed before CD7, and in CD7^- cells, more than half of the cells express CD38 (30, 35).

In mice, DN cells are divided into four populations: DN1, DN2, DN3, and DN4, with surface marker expression profiles of CD44⁺CD25^-, CD44⁺CD25⁺, CD44^-CD25⁺, and CD44^-CD25^- respectively (36–39)( Figure 2 ). DN1 contains mouse ETP (CD44⁺KIT^hiCD25^-), possessing multi-lineage differentiation potential (39, 40). DN2 cells contain DN2a and DN2b populations. DN2a also has multi-lineage differentiation potential, capable of differentiating into T, NK and ILC lineages. DN2b cells have been committed to T cell lineage and lose differentiation potential into other lineages (41–43). The onset of the rearrangement of gene encoding T cell receptor β (Tcrb), γ (Tcrg), and δ (Tcrd) occurs during the DN2a to DN2b transition (44, 45). From DN2 to DN3, γδT and αβT cell fate are determined (46, 47).

BRCA1-associated protein-1 (BAP1) is a member of the ubiquitin C-terminal hydrolase (UCH) subfamily of DUBs, the function of which in cancers is well studied (48, 49). BAP1 deletion in mice leads to severe thymic atrophy in both the cortex and medulla area. ETPs, DN, ISP and DP thymocytes are decreased in tamoxifen-induced Bap1 deletion mice (Bap1 ^fl/fl Rosa26 ^CreERT2) (50). Further study reveals that BAP1 deficiency results in the expansion of hematopoietic progenitor cells in bone marrow, but inhibits the development of B cells. In vitro culture experiments show that BAP1-deficient thymic precursors progress normally at DN1, DN2 and DN3 stages, but blocked at DN3 stage. In the absence of BAP1, DN cells lose the capability of transition to DP stage (50). Reconstruction of Bap1 ^fl/fl Rosa26 ^CreERT2 lineage-negative bone marrow progenitors with WT or mutant BAP1 (C91A, with no catalytic activity) shows that the mutant BAP1 (C91A) is not able to restore T cell development, suggesting the deubiquitinating activity of BAP1 is necessary for T cell differentiatio (50). Mechanism study shows decreased phosphorylation of H3S10 and enhanced ubiquitination of H2AK119 in BAP1-deficient cells, which is associated with G₂-M transition in cell cycle, suggesting BAP1 regulates cell proliferation in thymocytes by facilitating the deubiquitination of H2AK119 and allowing G₂-M transition (50).

The RING-containing E3 ubiquitin ligase F-box and WD repeat domain-containing 7 (FBW7, also called FBXW7) functions as the substrate recognition component of the Skp-cullin-F-box (SCF) E3 ubiquitin ligase. Knockout of FBW7 is embryonically lethal (51, 52). FBW7 mediates the ubiquitination and degradation of NOTCH1 and c-MYC, thereby involving in hematopoietic cell development (53–55). Hematopoietic-specific knockout of FBW7 impairs the self-renewal of hematopoietic stem cells (HSCs). Fbw7^-/- HSCs have defective cell cycle quiescence, failing to enter the resting state and thereby impacting self-renewal (56). As total c-MYC deficiency completely abrogates HSC differentiation, the single Myc allele was deleted in Fbw7^-/- Lin^-Sca-1⁺c-Kit⁺ (LSK) cells, and resulted in restored LSK numbers, cell-cycle status and in vitro self-renewal capacity, indicating the HSC reduction in Fbw7^-/- mice is due to accumulation of c-MYC (57). Notably, hematopoietic-specific knockout of FBW7 also impairs T cell development, with significantly reduced numbers of thymocyte at 4 weeks after polyI-polyC-induced deletion of Fbw7. In hematopoietic-specific FBW7 knockout mice, ETPs and DN cells are reduced significantly, while DN2 and DN3 thymocytes are nearly completely blocked (56).

Myb-like SWIRM and MPN domain containing 1 (MYSM1) belongs to the metalloprotease protein family and have deubiquitinase activity (58, 59). By coordinating chromatin structure and transcriptional programs, MYSM1 plays integral roles in hematopoietic stem cell activity, blood cell development, and immune system (58, 60, 61). In addition, cytosolic MYSM1 is shown to regulate innate immune signaling pathways by promoting the deubiquitination of tumor necrosis factor receptor-associated factor 3 (TRAF3), 6 (TRAF6), and receptor interacting protein 2 (RIP2) (62, 63). MYSM1 also functions in the early development stage of thymocyte (64). Mysm1 ^−/− mice show severely hypoplastic thymus due to decreased ETPs in context with lymphoid-primed multipotent progenitor (LMPP) defects in bone marrow, partial block of DN1-DN2 transition, and increased apoptosis of thymocytes (64). Further mechanism study reveals that increased apoptosis is correlated with the upregulation of p19^ARF in Mysm1-deficient thymocytes, which is an important regulator of the p53 tumor suppressor protein. Mysm1-deficient thymocytes exhibit increased expression of p53 and p53 target genes such as Bax and p21^Waf/Cip , as well as decreased expression of pro-survival proteins MCL-1 and BCL-X_L. In p53 ^-/- mysm1 ^-/- double-deficient mice, defects caused by Mysm1 deficiency are almost completely restored, including reduction of lymphoid-primed multipotent progenitors and decreased number of thymocytes, confirming the involvement of p53-mediated apoptotic pathway in early thymocytes development (64).

During development, T cells undergo TCR rearrangement and selection, eventually becoming mature, self-tolerant CD4⁺ or CD8⁺ T cells (13). TCR is divided into αβTCR and γδTCR. The variable (V), diversity (D), and joining (J) gene segments of the α, β, γ, and δ chains undergo rearrangement, forming a diverse TCR repertoire in a process called V(D)J recombination (13, 14, 65). V(D)J recombination is mediated by the recombination activating gene 1 (RAG1) and 2 (RAG2) (66). In human thymus, TCR rearrangement follows the sequential order of TCRD > TCRG > TCRB > TCRA ( 35, 67, 68). TCRD gene rearrangement is initiated at the CD34⁺CD38^-CD1a^- stage (35). TCRB gene rearrangement begins at the CD34⁺CD38⁺CD1a^- stage, followed by the expression of CD4 and the immature single positive CD4⁺ T cells (CD4 ISP) (35). CD4 ISP have the potential to develop into αβT and γδT cells (13, 37). β-selection occurs at CD34⁺CD38⁺CD1a⁺ stage in human and DN3 stage in mice (40, 47), which is an important checkpoint in αβT cell development. At this point, the rearranged β chain pairs with pre-TCRα and CD3 to form the pre-TCR complex (47, 69), which provides survival, proliferation and allelic exclusion signals without ligand binding (70–74). Thymocytes with a successfully rearranged TCRβ chain are selected for further differentiation, while thymocytes with failed TCRβ chain rearrangement undergo apoptosis (75, 76). Pre-TCR signaling induces a burst of proliferation, the rearrangement of TCRα chain and silence of TCRδ chain, as well as differentiation of CD4 ISP to DP cells (77–79). After entering the DP stage, the expression of pre-TCR is rapidly downregulated. DP thymocytes express CD4 and CD8. CD8 is a heterodimer consisting of α and β chains. CD8α chain expression precedes β chain, and thymocytes expressing CD8α chain are termed as early DP cells, which further develop into CD4⁺CD8α⁺β⁺ DP cells (47, 80–82).

Pre-TCR induces rearrangement of TCR α chain. The successfully rearranged α chain combines with the β chain to form the αβTCR (35, 83). Thymocytes with successfully rearranged TCR α chain recognize self-antigens presented by MHCI or MHCII with low or moderate affinity, thus activating the TCR signaling to provide proliferation and survival signals, which is known as positive selection (84, 85). Positively selected thymocytes upregulate expression of the early activation molecules CD69, CC-chemokine receptor 7 (CCR7), and subsequently CD27, while the expression of RAG1 and RAG2 is downregulated (86–90). Through binding of CCR7 expressed by thymocytes and CC-chemokine ligand 21 (CCL21) expressed by medullary TECs (mTECs), positive-selected thymocytes migrate to the thymic medulla for negative selection (87–89). High-affinity TCR-self-peptides-MHC interaction triggers cell death by apoptosis to establish self-tolerance, which is known as negative selection. After positive and negative selection, T cells acquire the ability to recognize alloantigens-MHC without being self-reactive, differentiating into mature CD4⁺ or CD8⁺ SP cells with naïve marker CD45RA⁺CCR7 +  (84, 91), and then enter the periphery to perform immune functions ( Figure 2 ).

Casitas B-lineage lymphoma (Cbl) belongs to RING-containing E3s. Cbl-b and c-Cbl are highly homologous expressed in hematopoietic cells. Both c-Cbl and Cbl-b are expressed in thymocytes and mature peripheral T cells, with c-Cbl expressing higher in DP cells than mature T cells, while Cbl-b is expressed higher in mature CD4 and CD8 T cells, suggesting c-Cbl and Cbl-b play different roles in T cell development and functions (92). Cbl-b deficient mice have normal thymic T cell development, while c-Cbl deletion leads to significant increase of CD4⁺ SP thymocytes (93). Using H-Y TCR transgenic mice, it was found that loss of c-Cbl promotes positive selection of CD4⁺ T cells without affecting CD8⁺ T cells, meanwhile, peripheral c-Cbl^-/- CD4⁺ T cells are less responsive to antigenic stimulation (93). These results indicate that c-Cbl selectively inhibits positive selection of CD4⁺ but not CD8⁺ T cells. Mechanistic studies reveal that c-Cbl deletion enhances TCR activation in thymocytes, with increased phosphorylation of ZAP70 and downstream ERK1/2 activation (93). c-Cbl^-/- DP cells expressed higher level of CD3, CD5, and CD69 molecules in proportions (93), indicating increased DP cells undergoing positive selection after c-Cbl deletion. Therefore, it is hypothesized that c-Cbl regulates TCR signaling by controlling ZAP70 and ERK activation, thereby promoting positive selection of thymocytes. However, whether c-Cbl directly regulates the ubiquitination of ZAP70 has not been demonstrated. Considering the enhanced CD3 and CD4 expression after c-Cbl deletion, it is possible that c-Cbl deletion promotes TCR-MHC complex interactions and thus positive selection of CD4⁺ T cells (93). In addition, another homologue Cbl promotes ubiquitination of TCRζ chain with its functional variant Src homology 2 domain and intact RING finger, while ZAP70 acts as an adaptor protein. Disrupt the interaction of TCRζ/ZAP70 or ZAP70/Cbl reduces the ubiquitination of TCRζ, by which Cbl negatively regulates the activation of TCR and positive selection during T cell development (94).

CD4 and CD8 lineage commitment is highly dependent on MHCII and MHCI, respectively. Thymocytes with TCR recognizing MHCII develop into CD4⁺ T cells and with TCR recognizing MHCI choose CD8⁺ T cell fate (95). Simultaneous deletion of c-Cbl and Cbl-b by Lck-Cre enhances thymic negative selection (96). c-Cbl and Cbl-b double deficient thymocytes still generate CD4⁺ and CD8⁺ cells in the absence of MHCI and MHCII molecules (96). Further studies find that c-Cbl and Cbl-b double deficient DP thymocytes have high expression of pre-TCR, as well as spontaneous and constitutive activation pre-TCR signaling (96). Pre-TCR is briefly expressed after TCRβ chain rearrangement in DN cells and downregulated after thymocytes entering DP stage during T lymphocyte development (77). pre-TCRα, TCRβ and CD3 form pre-TCR complex, which is constitutive activated without ligand stimulation to provide proliferation and allelic exclusion signals to mediate the transition of DN to DP stage (97, 98). It has been reported that c-Cbl mediates ubiquitination and degradation of pre-TCR (99), therefore downregulating pre-TCR signaling and enabling normal T cell development in MHC-dependent manner. In the absence of c-Cbl and Cbl-b, pre-TCR signaling fails to downregulate properly, providing constitutive survival and proliferation signals for thymocytes, thus these thymocytes no longer require αβTCR activation signals induced by MHC interaction to generate CD4⁺ or CD8⁺ SP T cells (96, 100). Therefore, by utilizing the constitutive activation signals provided by pre-TCR, thymocytes still develop into CD4⁺ and CD8⁺ T cells in c-Cbl and Cbl-b double-deficient mice.

CYLD is a deubiquitinase that removes K63-linked ubiquitin of TRAF2, TRAF6 and nuclear factor (NF)-kappaB essential modulator (NEMO), thereby blocking NF-κB signaling pathway. CYLD deficient mice (exon 2 deletion, Cyld^Δ2 ) have impaired thymocyte development, indicated by increased DP and decreased SP cells, leading to significantly reduced peripheral T cells (61). Mechanistic studies reveal that CYLD positively regulates TCR signaling in thymic DP cells through binding and removing K63- and K48-linked ubiquitin chains from active LCK, facilitating LCK recruitment of ZAP70 and promoting thymocyte development (61). However, another study using mice with Cyld gene exon 2 and 3 knocked out (Cyld^Δ2/3 ) showed no effect on thymocyte development, with comparable percentage of DP and SP thymocytes between wild-type mice and Cyld^Δ2/3 mice, and these Cyld^Δ2/3 mice are more susceptible to colonic inflammation and tumor incidence (101). Molecular mechanism studies reveal that CYLD deletion (exon 2 and 3) enhances NF-κB signaling. Cyld^Δ2/3 T cells have increased NEMO ubiquitination and NF-κB activation upon anti-CD3 stimulation (101). Meanwhile, TNF-α activated macrophages also have increased TRAF2 ubiquitination and NF-κB activation in Cyld^Δ2/3 mice, indicating CYLD controls NF-κB pathway activation by deubiquitinating NEMO and TRAF2, thereby regulating inflammatory responses (101).

To verify the role of CYLD in T cell development, another study using Lck-cre mice to delete exon 9 of Cyld gene (Cyld^Δ9 ) in T cells, which encodes a region essential for deubiquitinase activity of CYLD. Deleting exon 9 of Cyld gene inhibits the catalytic activity of CYLD. Cyld^Δ9 mice show significantly reduced SP thymocytes and peripheral T cells (102). By verifying the expression of CD69, TCR and CD5, researchers found that positive selection is markedly blocked in Cyld^Δ9 mice, and occurrence of apoptosis in DP thymocytes is increased (102). Meanwhile, basal NF-κB and JNK activity is elevated in Cyld^Δ9 DP thymocyte. Inactivation of NEMO could restore the NF-κB activity and thymocytes development in Cyld^Δ9 mice, indicating CYLD controls NF-κB activation by modulating NEMO, thereby establishing proper activation threshold for thymocytes positive selection (102). The distinct targeting strategies employed in generating the three CYLD mutants result in phenotypic differences in thymocyte development. These differences may arise from the expression of different truncated CYLD proteins following exon deletion and the use of T-cell specific versus germ-line deletion strategies. The germ-line deletion approach may introduce interference from non-T cells, further contributing to phenotypic variations.

CHMP5 is an endosomal sorting complex required for transport (ESCRT) proteins, which are mainly involved in multivesicular body formation (103, 104).Indeed, CHMP5 promotes thymocyte survival during positive selection through post-translational modifications. DP cells express the highest mRNA levels of Chmp5 among the different thymocyte subsets. Cd4-cre mediated Chmp5 knockout in T cells leads to significant thymic disruption, with markedly reduced CD4⁺ and CD8⁺ SP thymocytes and mature CD24^loTCRβ^hi T cells, along with greatly diminished αβT cells in peripheral lymphoid organs. The remaining T cells display a CD44^hi phenotype, indicating severely impaired T cell development (105). Positively selected thymocytes initiate anti-apoptotic programs including expression of the pro-survival protein Bcl-2 (106–108). Bcl-2 is inactivated and degraded upon oxidation (109). It was revealed that positive selection induces CHMP5-USP8 interaction, thereby maintaining CHMP5 protein stability. CHMP5 in turn inhibits oxidation and degradation of Bcl-2, ensuring survival of positively selected thymocytes (105).Deletion of the pro-apoptotic gene Bim or transgenic overexpression of Bcl-2 can reverse the T cell development defects caused by CHMP5 deficiency (105). Further studies show CHMP5 binds USP8 in CD69⁺TCRβ^hi DP cells but not in CD69⁻TCRβ^neg-lo DP cells, which have not undergone positive selection, suggesting that positive selection induces USP8-CHMP5 interaction, thus CHMP5 is deubiquitinated and stabilized, enabling CHMP5 to inhibit oxidation and degradation of Bcl-2, thereby promoting survival of positively selected thymocytes (105).

DP thymocytes interact with MHCII or MHCI molecules expressed on TECs and DCs during positive selection, determining CD4 or CD8 lineage development. TEC and DC present self-antigens to thymocytes through MHCII, facilitating positive selection and negative selection of CD4⁺ T cells. It was found that the E3 ubiquitin ligase MARCH8 targets and regulates the stability of MHCII in TECs (110, 111). MARCH8 deletion leads to increased MHCII protein levels in cTEC and AIRE^- mTEC, while the expression of MHCII in AIRE⁺ mTEC is not affected, suggesting MARCH8 modulates protein stability of MHCII (110, 111). However, the number and TCR repertoire of March8^-/- CD4⁺ T cells in thymus and spleen are comparable to wild-type mice (110, 111). CD83 is expressed by TEC and activated T cell, B cell, and DCs. MARCH8 regulates MHCII stability through antagonizing CD83 for MHCII binding in TECs (112, 113). CD83 deletion in TEC leads to impaired MHCII expression and CD4⁺ T cell development, with significantly reduced CD4⁺ SP cells in thymus, while DP and DN cells are normal (111). Notably, either MHCII^K225R mutation (preventing MHCII ubiquitination) or MARCH8 deletion can rescue defects of CD4⁺ T cell development in CD83 deficient mice, suggesting MARCH8 and CD83 modulate MHCII ubiquitination to regulate CD4⁺ T cell development. Additionally, CD83 binds and maintains MHCII protein stability, competing with another E3 ubiquitin ligase MARCH1 for MHCII binding in DCs, thereby antagonizing MARCH1-mediated MHCII ubiquitination (114, 115). In contrast, MARCH1 does not affect MHCII expression in TEC. In summary, MARCH8 antagonizes CD83 to control MHCII ubiquitination and stability in TEC, thereby modulating CD4⁺ T lineage commitment (110, 111).

The linear ubiquitin chain assembly complex (LUBAC) is composed of at least 3 proteins: RING-containing E3 ubiquitin ligase (RNF31, also called HOIP), RANBP2-type, C3HC4-type zinc finger-containing 1 (RBCK1, also called HOIL-1L) and SHANK-associated RH domain interactor (SHARPIN, also called SIPL1). LUBAC catalyzes linear ubiquitin chains of substrate proteins, thereby regulating various signaling pathways (116). The three components ensure optimal LUBAC catalytic activity with different extents. HOIP deletion is embryonically lethal and completely abrogates E3 activity of LUBAC (117). Cells lacking SHARPIN still retain substantial linear ubiquitination, as the HOIP-HOIL-1L complex is sufficient to maintain LUBAC E3 catalytic capability (118). Patients with loss-of-function mutations of HOIL-1L and HOIP have reduced T cells, suggesting HOIL and HOIP are involved in T cell homeostasis or development (119). T cell specific knockout of Rnf31 or Rbck1 genes (Rnf31^ΔCD4 or Rbck1^ΔCD4 mice) by Cd4-cre leads to nearly complete loss of CD4⁺ and CD8⁺ αβT cells in peripheral lymphoid organs (120). The proportion of Tregs among CD4⁺ αβT cells appears normal but the absolute number is significantly reduced (120). The remaining CD4⁺ and CD8⁺ αβT cells are mostly CD44^hiCD62L^lo phenotype, indicative of an effector/memory phenotype (120). In contrast, Sharpin deleted (Sharpin^cpdm ) mice have normal CD4⁺ and CD8⁺ T cells, but reduced Treg among CD4⁺ T cells (120). Analysis of thymocyte subsets shows that HOIP and HOIL-1L do not affect the positive and negative selection of early thymocytes, but regulate the maturation of late thymocytes. Rnf31^ΔCD4 and Rbck1^ΔCD4 mice have reduced CD4⁺ and CD8⁺ SP thymocytes with normal DPs (120). Sharpin^cpdm mice show normal percentages and numbers of DN, DP and SP thymocytes (120). Recent studies suggested two potential pathways of thymic Tregs development involving distinct Treg precursors (TregP): CD25⁺FOXP3^- (CD25⁺ TregP) and CD25^-FOXP3^lo (FOXP3lo TregP) (121). CD25⁺ TregP is believed exist at the TCR-instructive phase to generate CD25⁺FOXP3⁺ Tregs following cytokine stimulation in a two-step model of Treg differentiation (122), and is suggested more prone to FOXP3-induced apoptosis (123). Concurrently, both CD25+FOXP3+ Tregs and CD25-FOXP3+ Treg precursors are significantly reduced in the thymuses of Rnf31^ΔCD4 , Rbck1^ΔCD4 and Sharpin^cpdm mice, indicating HOIP, HOIL-1L and SHARPIN all promote Treg development in the thymus, while HOIP and HOIL-1L additionally regulate CD4⁺ and CD8⁺ T cell development (120).

Further studies revealed that HOIP and HOIL-1L regulate NF-κB activation in thymocytes. Thymocytes lacking HOIP show delayed IκBα degradation upon anti-CD3 and anti-CD28 stimulation. Additionally, Thymocytes lacking HOIP or HOIL-1L exhibit delayed and reduced p65 phosphorylation and IκBα degradation, as well as dampened p38 phosphorylation in MAPK signaling pathway, upon TNF stimulation (120). Rnf31^ΔCD4 IKKca mice with sustained IKKβ activation show rescued proportions of mature CD4⁺ SP T cells (CD4⁺CD24^lowCD62L^high) and FOXP3⁺ Tregs in thymus, while peripheral T cell numbers remains unreplenished (120), suggesting LUBAC impacts thymocyte development via NF-κB pathway as well as other mechanisms. Furthermore, constitutive activation of IKKβ promotes the development of Tregs by facilitating the binding of c-Rel to FOXP3 enhancer region, leading to the transcriptional induction of FOXP3 (124).

Thymocyte development relies on signals induced by both TCR and cytokine receptors. During thymocyte development, the sequential selections induce several rounds of proliferation and apoptosis, ensuring the diversification of TCR repertoire and elimination of self-reactive T cells. Thymocytes undergo two main rounds of proliferation during DN stage, one before TCRβ rearrangement mediated by stem cell factors (SCF), IL-7 and NOTCH. Mice with IL-7 or IL-7R deficiency are defective in thymocyte proliferation and DN2/3 cell number (125–129). Another is during β-selection when DN3 cells with successful β chain rearrangement express pre-TCR and massively expand (126, 130–132). Cell cycle progression from G0 to G1 and S/G1/M active phases during DN proliferation is controlled by Cyclin-dependent kinases (cyclin-CDK) complexes (133). Cyclin-CDK activity and cell cycle progression are inhibited by the Cip/Kip family including Cdkn1a, Cdkn1b and Cdkn1c, among which Cdkn1b plays a major role in β-selection (134–136). Cdkn1b mainly inhibits the cyclin A-CDK and cyclin E-CDK complexes. Cdkn1b-knockout mice have enlarged thymuses (134, 135), while Cdkn1b overexpression causes small thymuses with blocking T cell development at DN3 stage (137).

FBXL1 and FBXL12, as the substrate recognition subunits of Skp-cullin-F-box (SCF) E3 ubiquitin ligases, regulate DN thymocyte proliferation (138–140). In thymocytes, NOTCH signaling induces FBXL1 expression while pre-TCR induces FBXL12 expression (140, 141). SCF complexes containing FBXL1 and FBXL12 collaboratively mediate K48-linked ubiquitination and proteasomal degradation of Cdkn1b, thereby promoting proliferation of DN cells after β-selection (140, 142, 143). Fbxl1-knockout mice have significantly smaller thymus, with normal DN1 and DN2 cells but severely impaired DN3 to DN4 development and reduced proliferative capacity of DN4, ISP and DP cells (140). Concurrent Cdkn1b deletion restores normal thymus size in Fbxl1-knockout mice (144). Germline deletion of Fbxl12 is embryonic lethal (145). Lck-cre mediated conditional Fbxl12-deletion results in a phenotype similar to that in Fbxl1 ^-/- mice (140). Lck-cre ⁺ Fbxl12^fl/fl mice have increased Cdkn1b protein level in total DN thymocytes, and Cdkn1b deletion also rescues the defects in DN3-DN4 development caused by FBXL12 deficiency (140). Additionally, combined FBXL12 and FBXL1 knockout further impairs DN4, ISP and DP proliferative capacity, indicating the collaborative roles of FBXL12 and FBXL1 (140). In summary, FBXL1 and FBXL12 relieve Cdkn1b inhibition on proliferation of DN cell, thereby promoting β-selection-associated proliferation through SCF-mediated Cdkn1b ubiquitination and proteasomal degradation (140).

The ubiquitin-specific protease 8 (USP8) belongs to USP deubiquitinase family. Cd4-cre mediated T cell specific knockout of Usp8 (Usp8-TKO) results in significantly reduced percentages and numbers of CD4⁺ and CD8⁺ SP thymocytes, with unchanged number of DPs (146). The proportion of CD25⁺FOXP3⁺ Tregs among CD4⁺ thymocytes is also markedly reduced (146). Researchers also found TCRβ the expression of CD4⁺ and CD8⁺ SP thymocytes is nearly absent in Usp8-TKO mice (146). Further studies show USP8 regulates thymocyte proliferation without affecting negative selection. T cells lacking Usp8 fail to proliferate upon stimulation in vitro. In Usp8-knockout DP cells, mRNA levels of Il7ra and Ccr7 are reduced (146), and FOXO1 binding to Il7ra promoter is significantly decreased in Usp8-knockout CD4⁺ thymocytes, which regulates expression of Il7ra, Ccr7 and Sell ( 147, 148). Similar to mice with T cell specific Il7ra knockout, Usp8-TKO mice have a greater reduction of CD8⁺ SP thymocytes in the TCR^hi subset, suggesting USP8 regulates T cell development through FOXO1-IL7Rα axis (146). However, the binding of FOXO1 to the Il7ra promoter by USP8 is not mediated thorough ubiquitination directly. In detail, USP8 is shown to interact with the adaptor Gads and the regulatory molecule 14-3-3β. 14-3-3β belongs to the 14-3-3 family, which regulates the localization, DNA binding, and transcriptional activity of FOXO1 (149). By co-expression of USP8, 14-3-3β, and ubiquitin, ubiquitination of 14-3-3β at Lys9 and Lys 189 by mass spectrometry is identified (146). In this regard, USP8 might regulate the binding of FOXO1 to DNA indirectly by the regulating the ubiquitination of 14-3-3β. However, there is no direct ubiquitination experiment to reveal that USP8 mediates deubiquitination of 14-3-3β. Otherwise, USP8 binds Gads via its SH3BM domain, and ectopic re-expression of SH3BM-laking mutant of USP8 do not restore T cell development. Moreover, the catalytic activity of USP8 is required for T cell development, as demonstrated by in vivo rescue experiment (146). Additionally, deletion of Usp8 also causes reduced CD4⁺ SP thymocytes. These findings suggest that USP8 regulates T cell development through both FOXO1-IL-7Rα-dependent and independent mechanisms.

The von Hippel-Lindau protein (pVHL) is the substrate recognition component of E3 ubiquitin ligase complex that targets protein for degradation. pVHL is well known as a tumor suppressor through regulating the proteasomal degradation of hypoxia-inducible factor alpha (HIF-α) under normal oxygen conditions (150, 151). pVHL dysfunction leads to VHL disease, a genetic disorder characterized by increased cancer risk (152). Inactivation of pVHL in mouse germ line is embryonic lethality. Using mice with thymocyte-specific Vhlh (encoding VHL) deletion by Lck-cre, it is revealed that pVHL regulates thymocyte apoptosis. Thymi of Vhlh-deficient mice is small and highly vascularized, and the number of DP thymocytes is severely reduce (153). Further study reveals decreased expression of BCL-X_L and increased caspase 8 activity in Vhlh-deficient thymocytes, which results in impaired survival and increased apoptosis (153). pVHL regulates the ubiquitination and degradation of HIF-α subunits, which has been shown to promote apoptosis (154). Conditional deletion of HIF-1α in pVHL deficient thymocytes restores the expression of BCL-X_L and caspase 8 activity in DPs (153). Furthermore, it is demonstrated that HIF-1-dependent caspase 8 activity results in increased apoptosis of thymocytes but not BCL-X_L through transgenic expression of BCL-X_L and caspase 8-specific inhibitor assays, suggesting the HIF-1-dependent caspase8 activity is required for thymocyte development and viability regulated by pVHL (153).

FOXP3-expressing regulatory T (Treg) cells have two major roles in immune system: the prevention of autoimmunity and maintenance of immune homeostasis (155, 156). Based on origin and mechanism of functions, Tregs are divided into tTregs (thymus-derived Tregs) and pTregs (peripherally derived Tregs). tTregs recognize self-antigens presented by MHCII and develop in thymic medulla (157). FOXP3 and IL-2 are critical for the development and maintenance of tTregs. Mature tTregs migrate to the periphery, and function in immune tolerance (158–160). pTregs differentiate from peripheral CD4⁺ T cells under the induction of TGF-β, and mainly secrete IL-10 and TGF-β, playing immune suppression functions at inflammatory sites (161, 162).

Although most self-reactive CD4⁺ T cells are eliminated in the thymic medulla, some self-reactive CD4 T cells can differentiate into Tregs (13, 122, 163, 164). The strength of TCR recognizing self-peptide–MHC complexes is the key factor determining the cell fate of thymocyte. Low-affinity TCR interactions are required for positive selection, while high-affinity TCR interactions lead to negative selection and T cell apoptosis (165). However, some T cell with high-affinity TCRs for self-antigens are not eliminated, but instead undergo clonal diversion to Treg cells (157, 166–168). It is widely accepted that within a certain range of TCR affinity for self-antigen, the likelihood of Treg generation increases. This range typically falls between the thresholds for positive and negative selection (85, 169–172). Treating humanized mice with anti-CD28 antibody increases the number of CD25⁺FOXP3⁺ Tregs in thymus (173, 174). Mice lacking CD28 have greatly reduced numbers and proportions of tTregs in thymus and spleen (160). In addition, patients lacking ZAP70 have significantly decreased numbers and proportions of tTregs. Given the critical roles of CD28 and ZAP70 in TCR activation, these results indicate a promoting effect of TCR signaling on Treg differentiation (175).

Cbl-b plays a key role in the regulation of TCR signaling thresholds. Cblb ^-/- T cells have enhanced IL-2 production and cell proliferation upon TCR activation (176, 177). It has been shown that Cbl-b is involved in regulating tTreg and pTreg development via CD28 co-stimulation, which induces expression of FOXP3 in DP thymocytes (178). Deficiency of Cbl-b results in increased tTreg. CD28 deficiency leads to dramatically reduced tTreg numbers in both thymus and periphery (160), while Cbl-b deletion can partially restore tTreg development in Cd28^-/- mice. Mechanistic studies found that Cbl-b binds FOXP3 upon TCR activation and, together with STUB1, mediates ubiquitination and subsequent proteasomal degradation of FOXP3. Proteasome inhibition can restore tTreg development in Cd28^-/- mice, similar to the effect of Cbl-b deletion in Cd28^-/- mice, indicating that Cbl-b regulates tTreg development by ubiquitinating and degrading FOXP3 (178).

Cbl-b also regulates pTreg differentiation induced by TGF-β. In contrast to tTreg development, Cbl-b regulates pTreg development by modulating FOXP3 expression instead of degradation, suggesting different mechanisms in the thymus and periphery. Transcription factors Forkhead-box O (FOXO) family proteins FOXO3a and FOXO1 directly bind the promoter of FOXP3 and induce its transcription, and phosphorylation of FOXO3a and FOXO1 prevents its binding to the FOXP3 promoter (179). Cbl-b deficiency leads to reduced TGF-β-induced pTreg cell numbers (179, 180). Mechanism study revealed enhanced phosphorylation of FOXO3a and FOXO1 in Cbl-b-deficient CD4 T cells upon TGF-β stimulation, leading to decreased FOXP3 expression (179). Expressing FOXO3a in Cbl-b-deficient mice can restore expression of FOXP3, indicating that Cbl-b inhibits the phosphorylation of FOXO3a and FOXO1, thereby promoting the expression of FOXP3 and development of pTreg (179). However, how Cbl-b, as an E3 ubiquitin ligase, inhibits phosphorylation of FOXO3a and FOXO1 remains to be elucidated. Another study showed that AKT2 mediates phosphorylation of FOXO3a and FOXO1, and the phosphorylation of AKT is enhanced in Cbl-b-deficient T cells (181). AKT2 deletion eliminates excessive phosphorylation of FOXO3a and FOXO1 caused by Cbl-b deficiency, suggesting that Cbl-b regulates FOXO phosphorylation and subsequent pTreg development through modulating AKT activation, although the underlying mechanisms of how Cbl-b regulates AKT phosphorylation as an E3 ligase requires further investigation (181). Furthermore, recent study suggests that TGFβ signaling promotes Treg differentiation by disrupting weaker agonist TCR signaling to promote nuclear FOXO1 expression (182). Given that Cbl-b-deficient T cells show hyperactivation of AKT upon TCR stimulation (181, 183), it is possible that Cbl-b deficiency enhances T cell activation and therefore overcomes TGFβ-mediated suppression of TCR signaling, thus leading to enhanced AKT activation to enhance the phosphorylation of FOXO3a and FOXO1, and suppressed FOXP3 expression. Additionally, studies show that TGF-β stimulation leads to reduced phosphorylation of downstream SMAD2 and decreased FOXP3⁺ Treg generation in Cbl-b deficient mice, suggesting Cbl-b regulates FOXP3 expression and Treg induction via modulating the TGF-β/SMAD2 signal pathway (180).

MARCH1 is a RING domain-containing E3 ubiquitin ligase. Treg development in thymus requires antigen presentation by DCs and TECs (184, 185). MHCII-expressing DCs promote negative selection and Treg generation in the thymus (186, 187). MARCH1 regulates ubiquitination of MHCII and CD86 in DCs, leading to internalization, lysosomal targeting, and degradation of the indicated molecules, which dampens DC antigen presentation (186, 188). Upon exposure to maturation signals like microbes and inflammatory stimuli, MARCH1 expression is downregulated and MHCII/CD86 levels increase, enhancing antigen presentation of DCs (189, 190). Under immunosuppressive conditions like IL-10 stimulation, MARCH1 expression is upregulated on DCs, thus decreasing MHCII/CD86 protein levels and antigen presentation capacity (115, 186). By regulating ubiquitination of MHCII in thymic DCs, MARCH1 controls tTreg development (191). MARCH1 deficiency leads to dramatically reduced tTreg numbers without affecting the total CD4⁺ thymocytes, and MARCH1-deficient DCs fail to support antigen-specific Treg development in vivo and in vitro ( 191). MHCII mutant mice that cannot be ubiquitinated also exhibit reduced tTreg numbers (191, 192), highlighting the importance of MARCH1-mediated MHCII ubiquitination in tTreg development. Given the higher-strength signals compared to positive selection, and lower-strength signals compared to negative selection are required for Treg development, it is possible that the impaired Treg development in MARCH1 deficient mice is due to the strong TCR signaling caused by increased MHCII/CD86 and antigen presentation capacity.

Itch is a HECT-containing E3 ubiquitin ligase that mainly regulates T cell activation (193). Itch^-/- T cells are unresponsive to TGF-β stimulation, leading to impaired FOXP3 expression and reduced Treg differentiation (194). Mechanistic studies found that Itch binds the transcription factor TGF-β inducible early gene 1 (TIEG1) and mediates its mono-ubiquitination and poly-ubiquitination without affecting its protein stability (194). Ubiquitinated TIEG1 directly binds to the FOXP3 promoter and induces FOXP3 transcription. Deficiency of either TIEG1 or Itch dramatically reduces FOXP3 expression and Treg differentiation, suggesting Itch, together with TIEG1, promote the expression of FOXP3 and Treg differentiation (194).

Additionally, other studies show Itch regulates TGF-β-induced Treg differentiation via the adaptor protein Ndfip1 (195). Itch ^-/- T cells exhibit increased proliferation and preferential Th2 differentiation with elevated production of Th2-associated cytokines like IL-4 and IL-5 (196, 197). It has been studied that Itch binds and mediates ubiquitination and degradation of the transcription factor JunB, thus inhibiting IL-4 production (196–198). The adaptor protein Ndfip1 is critical for Itch-mediated JunB ubiquitination. Upon T cell activation, Ndfip1 associates with Itch and promotes Itch-mediated JunB degradation (199, 200). Deficiency of either Ndfip1 or Itch leads to prolonged JunB half-life and enhanced Th2 responses with elevated IL-4 production (199), and TGF-β-induced FOXP3⁺ Treg generation is impaired (195). Ndfip1 deficiency does not affect the binding of TIEG1 to the FOXP3 promoter, indicating Ndfip1/Itch regulate FOXP3⁺ Treg differentiation independently of TIEG1. Furthermore, subsequent studies show that increased IL-4 production in Ndfip1 ^-/- cells inhibits FOXP3 expression (195, 201). Therefore, it is proposed that Ndfip1 facilitates Itch-mediated ubiquitination and degradation of JunB, inhibiting IL-4 production and promoting FOXP3⁺ Treg differentiation (195).

TRAF3, a member of the tumor necrosis factor receptor-associated factor family, mainly participates in TNF receptor (TNFR) and Toll-like receptor (TLR) signaling pathway. TRAF3 contains TRAF domains and a RING domain, enabling its function as an adaptor protein to facilitate signaling transduction or E3 ubiquitin ligase to mediate ubiquitination (202). It is shown that TRAF3 negatively regulates tTreg development in thymus by modulating Treg precursor differentiation, a process requiring IL-2 signaling (122, 203, 204). IL-2 activates downstream pathways upon binding to IL-2 receptor (IL-2R), which is consisted of tree subunits, IL-2Rα (CD25), IL-2Rβ (CD122), and common γ chain (CD132) (205). JAK1 and JAK3 associate with CD122 and CD132, respectively, and mediate phosphorylation of themselves as well as IL-2Rβ and the transcription factor STAT5 (206, 207). Phosphorylated STAT5 binds to the FOXP3 promoter and induces FOXP3 expression (208, 209). TRAF3 recruits the tyrosine phosphatase TCPTP to the IL-2R complex, leading to dephosphorylation and inactivation of JAK1/3 and subsequent STAT5 signaling, thus inhibiting tTreg development (203). However, whether TRAF3 functions as an E3 ligase or adaptor in this process remains unclear. Additionally, Treg-specific TRAF3 deletion impairs Treg function, disrupts CD4 T cell homeostasis, increases Th1-like effector/memory T cells, and enhances Tfh activation and germinal center formation as well as IgG production (210). Another TRAF molecule, TRAF6, has also been reported to regulate Treg function by mediating K63-linked ubiquitination of FOXP3, ensuring nuclear localization of FOXP3 for its transcriptional activity. Traf6-deficient Tregs showed perinuclear accumulation of FOXP3, resulting in impaired immunosuppressive capacity and enhanced anti-tumor immunity (211).

A20 is a deubiquitinase that mainly mediates negative feedback regulation of NF-κB signaling and regulates TNF-induced apoptosis (212). A20 plays important roles in T cell activation, survival, and differentiation. Upon TCR activation, A20 removes ubiquitination of MALT1, thus inhibiting NF-κB activation (213, 214). A20 also inhibits receptor interacting protein kinase 3 (RIPK3)-induced necroptosis of naïve CD4 T cells (215). It was shown that the number of Treg cells is significantly increased in both the thymus and periphery in mice with T cell-specific A20 deletion (216). A20-deficient Tregs are less dependent on IL-2 with unaffected proliferation and apoptosis. Thymic Treg cell progenitors (CD4⁺CD25⁺GITR⁺FOXP3^-) are also markedly increased in mice with T cell-specific A20 deletion, along with increased GITR expression, which is a member of the tumor necrosis factor receptor superfamily (TNFRSF) protein and promotes Treg development (216). TNFRSF agonists enhance activation of IL-2R and STAT5, thus promoting the differentiation of Treg cell progenitors (217). GITR deficiency impairs Treg development, and combined inhibition of GITR, OX40, and TNFR2 abrogates the development of tTreg (217), highlighting the role of GITR in promoting tTreg development. Therefore, increased GITR expression in T cell-specific A20-deficient mice promotes tTreg development.

The deubiquitinase CYLD regulates both CD4⁺/CD8⁺ T cell and Treg development, although different CYLD deletions have differential effects. Mice lacking exons 2 and 3 of Cyld gene exhibit significantly increased peripheral and normal thymic Treg numbers (218). In vitro, TGF-β-induced differentiation of Tregs from naïve CD4⁺ T cells is enhanced upon Cyld deletion (exons 2 and 3) (218). Mechanistic studies found that CYLD binds SMAD7 and removes ubiquitination chains at Lys360/Lys374 sites of SMAD7, thus dampening the activation of TAK1, p38, as well as the transcription factor AP-1 (218). SMAD7 or p38 deficiency inhibits Treg differentiation, indicating that CYLD regulates TGF-β signaling and inhibits Treg differentiation by deubiquitinating SMAD7 and dampening AP-1 activation (218).

Lacking of exons 7 and 8 of Cyld (Cyld^ex7/8 ) abrogates the interaction of CYLD-NEMO and CYLD-TRAF2 (219). Cyld^ex7/8 mice, exclusively express a naturally occurring splice variant of CYLD lacking exons 7 and 8, exhibit constitutive NF-κB activation in T cells and produce higher levels of proinflammatory cytokines like IL-17A, IL-6, and IFN-γ (220). Concurrently, tTreg and pTreg numbers are significantly increased in Cyld^ex7/8 mice. However, these Tregs exhibit reduced CD25 and CTLA-4 expression associated with impaired suppressive capacity, indicating that while CYLD inhibits tTreg development, it promotes Treg regulatory function to control abnormal T cell responses (220).

Positive and negative selection are highly regulated processes dependent on the distinct thymic microenvironment composed of stromal cells including thymic epithelial cells (TECs), endothelial cells, mesenchymal cells, and non-lymphoid cells like dendritic cells (DCs) and macrophages (221, 222). Among them, TECs play critical roles. TECs exert important functions in selection of αβT cells, Treg generation, and development of other unconventional T cell subsets (223, 224). TECs provide migration, proliferative, differentiative and survival signals for developing thymocytes to shape a diverse and functional T cell repertoire without self-reactivity, thus establishing central tolerance, preventing peripheral autoimmune responses (225, 226). Based on anatomical localization, marker expression, and functional differences, TECs are classified into cortical TECs (cTECs) and mTECs (225, 227)。 cTEC resides in the thymic cortex region and expresses various molecules critical for early T cell development and positive selection, such as IL-7, MHC, the NOTCH ligands DLL1 and DLL4, proteases β5t, TSSP and CTSL, to constitute the unique cortical microenvironment supporting DN and DP cell development (85, 225, 228). mTEC localizes in the medulla region and plays vital roles in negative selection and self-tolerance establishment (229, 230).

mTECs promiscuously express tissue-restricted antigens (TRAs) to mediate removal of thymocytes bearing self-reactive TCRs, establishing central tolerance to avoid autoimmune attack of self-antigens by T cells exiting to the periphery (231, 232). The transcription factor autoimmune regulator (AIRE) is crucial for TRA expression of mTEC. Deficiency or loss-function mutations of AIRE reduce TRA expression in mice, leading to defective central tolerance and escape of autoreactive T cells (233–236). Loss-function mutations of AIRE in human also result in multiple organ-specific autoimmune diseases, such as autoimmune polyendocrinopathy-candidiasis-ectodermal dystrophy (APECED), otherwise known as autoimmune polyendocrine syndrome type 1 (APS-1) (237, 238). AIRE contains two plant homeodomains (PHDs) with similar structure to RING domains (239, 240). Some studies suggest the PHD1 domain possesses E3 ubiquitin ligase activity that impacts AIRE function in central tolerance (241), but whether AIRE functions as an E3 ligase remains controversial as other studies found no ubiquitin ligase activity (242).

As an E3 ubiquitin ligase, TRAF6 activates classical NF-κB downstream of TLR and TCR signaling. TRAF6 deficiency causes thymic atrophy, indicating its critical role in promoting organogenesis (243). Further studies found that TRAF6 deletion disrupts mTEC distribution in the thymus, reduces mature mTECs, and impairs thymic architecture (244). Additionally, transplantation of thymic stroma of Traf6-deficient embryos into athymic nude mice causes inflammatory response in lung, liver, pancreas, and kidney, as well as splenomegaly along with high levels of autoantibodies against pancreas and lung (244), indicating TRAF6 deficiency leads to autoimmunity.

NF-κB is activated through canonical and non-canonical pathways. Substantial evidence demonstrates the role of non-canonical NF-κB signaling pathway in promoting mTEC development. Deficiency of the key proteins in non-canonical NF-κB activation like RANK, CD40, LIGHT, and LTβR in mice impairs both mTEC development and function, leading to autoimmune symptoms (224, 226, 243). As the transcription factor of non-canonical NF-κB signaling, RelB deficiency reduces expression of AIRE and functional mTECs, which results in multi-organ inflammatory response (245–247), highlighting the importance of non-canonical NF-κB in mTEC development. FTOC experiments using TRAF6-deficient embryos found that RANKL and CD40L stimulation fails to induce UEA⁺Aire⁺ mTEC and inhibits AIRE and TRA expression, indicating the vital role of TRAF6 in mTEC development and function (248). Further studies revealed that RelB expression is significantly reduced in medullary cells of TRAF6-deficient thymus, and RANKL/CD40L stimulation of TRAF6-deficient thymic organ cultures cannot induce RelB expression either, suggesting TRAF6 regulates mTEC development by modulating RelB expression (244, 248).

TEC-specific TRAF6 deletion reveals normal cTEC development but dramatically reduced mature and immature mTECs (249). Additionally, TEC-specific TRAF6 knockout mice have reduced body weight, inflammatory response in the liver, lungs, and kidneys, and exhibit autoantibodies against the liver, lungs, kidneys, small intestine, and colon, while features of human autoimmune hepatitis (AIH) including increased plasmacytes and liver-reactive T cells are also observed in the livers (249). Transferring liver T cells into immunodeficient mice recapitulates AIH symptoms, confirming that TRAF6 directly regulates mTEC development to impact self-reactive T cells and autoimmunity (249). However, the detailed molecular mechanisms of how TRAF6, as an E3 ubiquitin ligase, regulates RelB expression and mTEC development remain unclear. Given the role of TRAF6 in classical NF-κB signaling, it is proposed that TRAF6 may modulate RelB expression and mTEC development via activating the classical NF-κB pathway (248).

mTEC promotes the induction of self-tolerance in developing T cells. In turn, development of mTEC requires signals from the developing thymocytes, a process termed crosstalk (250–252). As a negative regulator of non-canonical NF-κB pathway, TRAF3, together with TRAF2, facilitates the ubiquitination and degradation of NIK to maintain extremely low NIK levels. In Tcra^-/- mice, T cell development is arrested at the DP stage, with a complete absence of SP thymocytes as well as significantly decreased mTEC regions (253). TRAF3 was reported to enforce the requirement for T cell crosstalk in mTEC development (254). TEC-specific deletion of TRAF3 using Foxn1-cre mice in Tcra^-/- mice restores the development of mTEC, with confluent medullary areas and normal AIRE expression (254). Furthermore, TEC-specific deletion of TRAF3 in Ltbr^-/-/Ltbr^-/- double-deficient mice restores the defects in mTEC development (254). These results suggest TRAF3 deletion can overcome requirement for thymocyte crosstalk as well as LTβR and CD40 signals. However, TEC-specific TRAF3 deletion fails to rescue mTEC development in Relb^-/- mice, suggesting its role is RelB dependent (254). Given the critical role of non-canonical NF-κB pathway in mTEC development, it is likely TRAF3 regulates mTE development by facilitating the ubiquitination and degradation of NIK.

Besides its roles in thymic positive selection and Treg development mentioned above, CYLD also regulates mTEC development and function, thereby impacting thymic negative selection (255). In Cyld^ex7/8 mice, development from immature mTEC to mature mTECs is blocked, and mature mTECs are dramatically reduced. Regarding the role of CYLD in T cell development, CYLD^FL-CD4-cre mice are utilized, in which sCYLD (lacking exons 7 and 8) is overexpressed solely in αβT cells. The results show normal mTEC numbers, indicating that CYLD directly regulates mTEC development independent of effects on T cells. However, whether the development of thymocyte was affected in CYLD^FL-CD4-cre mice has not been shown. Furthermore, to assess the impact of CYLD deletion on negative selection, HYtg mice were crossed with Cyld^ex7/8 mice and the results show that HYtg Cyld^ex7/8 male mice contain detectable DP, CD4⁺ SP, and CD8⁺ SP thymocytes, indicating impaired negative selection and escape of autoreactive T cells upon CYLD deletion (exons 7/8). This demonstrates that CYLD regulates mTEC development and maturation to impact negative selection and establishment of central tolerance (255).

The role of LUBAC in thymocyte development has been discussed above. TEC-specific knockout of HOIL-1L or HOIP also affects the development and survival of TEC, thereby affecting T cell development in the thymus (256). Using Foxn1-cre mice to delete Rbck1 (Hoil1^ΔFoxn1 ) or Rnf31 (Hoip^ΔFoxn1 ) specifically in TEC, it is found that the thymus of adult mice (8 weeks old) is severely atrophied, and DN, DP and SP cells are significantly reduced (256). The absolute numbers of DN1, DN2, DN3 and DN4 thymocytes are significantly reduced, among which the proportion of DN3 is decreased, suggesting the development of DN3 thymocytes is blocked (256). Additionally, TEC is significantly reduced in adult Hoil1^ΔFoxn1 and Hoip^ΔFoxn1 mice (256). Immunofluorescence staining shows structural defects in K8⁺ cortical area and K5⁺ or UEA-1⁺ medullary area of Hoil1^ΔFoxn1 (or Hoip^ΔFoxn1 ) thymus (256). AIRE⁺ mTECs are almost completely disappeared, while ERTR7⁺ fibroblasts are prominently remodeled (256). mTECs including AIRE⁺ mTEC are also significantly decreased in embryonic and newborn Hoil1^ΔFoxn1 mice, revealing the dual roles of LUBAC complex in TEC differentiation and survival (256).

Further RNA sequencing reveals that the expression of programmed cell death-related gene is changed in HOIL-1L knockout TECs, among which Mlkl and Casp8 are upregulated in cTEC and mTEC^hi (MHCII^hiCD80^hi) populations (256). Deletion of Mlkl and Casp8 genes in Hoil-1^ΔFoxn1 mice revealed that the total number of thymocytes, DN, DP and SP cells recovers to normal levels, but the number of cTEC and mTEC only partially recovers (256). Immunofluorescence staining results show that the cortical structure is basically restored, but the medullary structure is only partially repaired, suggesting that HOIL-1L affects cTEC survival by regulating Caspase-8/MLKL-mediated cell apoptosis, while there are still other mechanisms regulating mTEC development (256). Otherwise, since Mlkl and Casp8 are germline deleted, the effects reflected in thymocyte development in Hoil-1^ΔFoxn1 mice may be caused through thymocyte-intrinsic way (256). Therefore, how LUBAC affects AIRE⁺ mTEC development and whether it affects the establishment of self-tolerance needs more research.

T cell plays critical role in adaptive immunity. Progenitors from the bone marrow enter the thymus and undergo TCR rearrangement, β-selection, positive selection, negative selection, and finally differentiated to mature T cells entering periphery, forming an abundant TCR repertoire that enables the recognition of diverse antigens (155, 257–260). Ubiquitination modulates thymocyte development through regulating several key events including T precursor homeostasis, cell proliferation and apoptosis, pre-TCR and TCR activation to ensure the optimal signals for normal T cell development. Meanwhile, several E3 ubiquitin ligases and DUBs regulate Treg development by modulating critical signals like IL-2 or TGF-β. Negative selection of thymocytes ensures mature T cells enter the periphery without attacking self-tissues, thereby establishing central tolerance, while mTEC plays an indispensable role in this process. E3 ubiquitin ligases or DUBs affect mTEC development and function mainly by regulating the NF-κB pathway, thereby modulating negative selection of T cells, ensuring the establishment of central tolerance.

Impaired negative selection leads to self-reactive T cells entering the periphery, leading to autoimmunity in multiple organs. Therefore, figuring out the regulatory mechanisms is meaningful for normal T cell development and guiding the treatment of autoimmune diseases. Besides role in T cell activation, regulatory role of ubiquitination in T cell development and self-tolerance establishment is relatively less studied. Previous studies mainly focus on the role of single E3 or DUB molecules, while the systematic mechanism underlying how ubiquitination regulates T cell development is absent. Ubiquitinomics advancements make it possible and resolve these questions, which benefits the therapeutic target development, providing novel therapeutic opportunity for autoimmune diseases and cancers.

HY: Writing – original draft, Writing – review & editing, Conceptualization, Investigation, Software. WY: Conceptualization, Funding acquisition, Writing – original draft. MC: Conceptualization, Funding acquisition, Writing – original draft. QL: Writing – review & editing, Validation. RY: Conceptualization, Formal analysis, Investigation, Writing – review & editing. HX: Conceptualization, Data curation, Investigation, Writing – review & editing. YC: Writing – review & editing, Validation. XC: Writing – review & editing, Validation. XS: Writing – review & editing, Validation. HZ: Conceptualization, Data curation, Funding acquisition, Investigation, Writing – review & editing. LL: Conceptualization, Data curation, Investigation, Writing – original draft, Writing – review & editing.