TGF-β has a multi-faceted, negative impact on both CD4⁺ and CD8⁺ T cell proliferation and effector function, ultimately limiting clearance of infected cells (38). One primary mechanism by which it does this is through downregulation of IL-2 production, a cytokine essential for both CD4⁺ and CD8⁺ T cell proliferation and activation (39, 40). Combined with virus-driven CD4⁺ T cell loss (41), this TGF-β-driven reduced proliferative capacity and reduced responsiveness to co-stimulation (42) weakens the immune system’s ability to clear infected cells and control viral spread. In support of this, PBMCs from PWH exhibit defective proliferation in response to recall antigens, which can be partially restored by neutralizing TGF-β with monoclonal antibodies (31). Elevated TGF-β also promotes CD4⁺ T cell differentiation into regulatory T cells (Tregs) (43), which release additional TGF-β and suppressive cytokines like IL-10 (44), thereby amplifying immunosuppression. The expansion of Tregs in HIV has been directly linked to increased TGF-β production, establishing a positive feedback loop that suppresses effector T cell function (45, 46). Moreover, studies in SIV-infected ART-treated macaques have shown that Treg/Th17 balance is altered in gut-associated lymphoid tissues (GALT): higher Treg frequencies persist together with elevated TGF-β and IDO expression in mesenteric lymph nodes (MLN), even under therapy (47), suggesting that TGF-β–driven immunoregulation remains active in mucosal and gut-associated compartments despite viral suppression.

Beyond its effects on proliferation and Treg induction, TGF-β signaling also intersects with inhibitory checkpoint pathways and T follicular helper (Tfh) differentiation programs. TGF-β has been shown to increase PD-1 expression on T cells in chronic infection, reinforcing an exhausted, hyporesponsive phenotype (48, 49). PD-1^hi Tfh cells, in turn, represent a major reservoir for HIV in lymphoid tissues and B cell follicles (50, 51). Consistent with this, recent work demonstrates that TGF-β promotes Tfh differentiation and humoral responses at least in part by inhibiting expression of the chromatin organizer SATB1 (52). Together, these findings suggest that higher TGF-β not only dampens antiviral effector responses but also favors the generation and maintenance of Tfh cells that function as key tissue reservoir.

Importantly, TGF-β has been shown to directly inhibit terminal differentiation and cytotoxic activity of antiviral CD4⁺ T cells during chronic viral infections, supporting an undifferentiated memory phenotype through decreased inhibition of eomesodermin (53). Beyond the direct suppression of effector functions, emerging evidence indicates that TGF-β impacts CD4⁺ T cell metabolism, reducing mitochondrial respiration and glycolytic capacity (54). These metabolic changes are known to restrict T cell activation and cytokine production, although this mechanism remains relatively underexplored in the context of HIV infection.

In CD8⁺ T cells, TGF-β has been shown to suppress both proliferative capacity and cytotoxic activity (55). TGF-β directly reduces CD8⁺ T cell proliferation by inducing cyclin-dependent kinase inhibitors such as p21 (56) and p15 (57) to inhibit cell-cycle progression. In addition, TGF-β interferes with IL-15 signaling, which is essential for long-term CD8⁺ T cell survival and memory formation (58). Beyond reducing proliferative capacity, TGF-β also suppresses cytotoxic function by downregulating the expression of key effector proteins –including perforin, granzyme A and B, Fas ligand, and interferon-γ (31, 59). Neutralization of TGF-β in vitro has been shown to restore perforin expression in rectal CD8⁺ T cells, providing evidence that TGF-β directly limits effector function in HIV-targeted tissues (60). Importantly, in SIV-infected macaques, elevated TGF-β (and IDO) expression within intestinal lymphoid tissues was associated with increased death of effector/memory CD8⁺ T cells via a Bax/Bak/Puma–dependent apoptotic pathway; conversely, in vitro blockade of TGF-β enhanced T cell proliferation and reduced CD8⁺ T-cell death (61). This evidence suggests that TGF-β–mediated depletion of cytotoxic CD8⁺ T cells in mucosal/lymphoid reservoirs can contribute to inefficient viral control during chronic infection.

In vivo, TGF-β blockade in SIV infected, ART-treated macaques enhanced SIV-specific CD8⁺ and CD4⁺ T cells responses providing direct evidence that inhibiting TGF-β can boost antiviral immunity (62, 63). Furthermore, FoxP3⁺ CD8⁺ Treg cells in PWH and SIV-infected NHPs have been found to express TGF-β along with inhibitory receptors, such as PD-1 and CTLA-4. These populations have reduced cytokine secretion and proliferative capacity, reflective of an exhausted phenotype (64). This occurs in part through lower expression of the transcription factor T-bet, a critical regulator of cytotoxic gene programs (65). On the other hand, TGF-β appears to be critical to the maintenance of stemness in pre-exhausted Tpex cells and to form long-term CD8⁺ T cell memory in chronic LCMV infection (49). A comprehensive of the impact of TGF-β on T cells can has been published by Chen W et al. in 2023 (5).

In myeloid lineage cells, particularly macrophages and dendritic cells, TGF-β significantly impairs antigen-presentation and inflammatory functions, creating additional barriers to effective antiviral immunity. TGF-β has been shown to inhibit Class II Transactivator (CIITA), the master regulator of MHC class II gene expression, thereby decreasing cell surface levels of MHC II and reducing antigen presentation capacity (66, 67). As a result, TGF-β inhibits the ability of macrophages and dendritic cells to effectively engage T cell receptors (TCRs) and stimulate adaptive responses. TGF-β also downregulates co-stimulatory molecules, such as CD80 and CD86, limiting the secondary signals required for full T cell activation (68). In the context of HIV infection, exposure to elevated TGF-β was shown to alter dendritic cell maturation, diminishing IL-12 secretion and skewing differentiation toward a tolerogenic phenotype that compromises the initiation of Th1-driven antiviral immunity (68, 69). Similarly, studies of monocyte/macrophage polarization show that TGF-β, alone or in combination with IL-10 and IL-4, drives cells toward an immunosuppressive “M2-like” phenotype characterized by reduced antigen-presentation capacity and enhanced tissue-repair function (70). A more recent report suggests that TGF-β leads to metabolic reprogramming that ultimately induces a distinct macrophage phenotype characterized by elevated expression of genes associated with a pro-repair phenotype (such as tissue remodeling, vasculature development, negative regulation of TNF production) and decreased expression of genes associated with inflammation (71). Of note, in SIV infected macaques on ART, blocking TGF-β signaling led to a more tolerogenic phenotype in lymph node macrophages (62). Collectively, these findings demonstrate that TGF-β not only directly suppresses lymphocyte effector functions but also fundamentally alters the activation state and functional capacity of antigen-presenting cells, thereby compromising immune coordination and antiviral responses in PWH.

Natural killer (NK) cells play a crucial role in controlling HIV-1 replication during acute and chronic infection through both direct cytotoxic activity and immunoregulatory functions (72–74). NK cell function has been shown to modulate virologic control in HIV-1 elite controllers, who maintain viral suppression in the absence of ART, and in long-term non-progressors, who control HIV-1 disease progression in the absence of ART (75). In NHP models, a subset of NKG2a/c^low CD16⁺ cytotoxic NK cells in the lymph nodes has been associated with virologic control in non-pathogenic SIV infection and in pathogenic SIV infection upon ART interruption (76). TGF-β restricts NK cell number and effector function in chronic viral infection and inhibits NK cell cytotoxicity and IFN-γ production (77). Notably, TGF-β specifically inhibits CD16 (FcγRIIIa)-driven IFN-γ production and NK cell antibody-dependent cellular cytotoxicity (ADCC) capability via SMAD3 signaling (78), with blockade of TGF-β being sufficient to upregulate CD16 expression on NK cells in vivo (62). This is particularly relevant in the context of current therapeutic and curative strategies involving anti-HIV broadly neutralizing antibodies (bNAbs), which leverage CD16 signaling to engage in NK cell-mediated ADCC (79, 80). Beyond these direct functional effects, TGF-β also profoundly impacts NK cell metabolism by suppressing glycolysis and oxidative phosphorylation, thereby limiting the bioenergetic capacity required for optimal NK cell effector responses (81, 82).

In addition to classical NK cells, TGF-β also modulates other innate lymphocyte subsets. In pathogenic SIVmac infection, the acute phase is marked by early elevations of both TGF-β and IL-18 in intestinal tissues, which drive the emergence of highly inflammatory IL-17–expressing NKT⁺ cells (83). These IL-17⁺ innate-like lymphocytes are absent in non-pathogenic SIVagm infection (83), which also does not lead to an increase in TGF-β/IL-18–associated pathways (84), supporting a link between elevated TGF-β and pathogenic outcomes.

Collectively, TGF-β-mediated suppression of T cell responses and cytotoxic activity, impaired antigen presentation by myeloid cells, and compromised NK cell effector functions establish TGF-β as a critical mediator of HIV-driven immune dysfunction and viral persistence (Figure 2). Evidence from clinical studies with PWH and NHP studies indicate that robust CD8⁺ T cell and NK cell responses can synergistically eliminate infected cells, contribute to virologic control on ART, and delay viral rebound following ART interruption (85–87). Hence, strategies to counteract TGF-β-mediated immunosuppression may be key to restoring immunological function and achieving host-mediated virologic control in the absence of ART.

One of the clinical hallmarks of long-term HIV infection is the development of fibrosis throughout multiple organ systems, with lymphoid tissues being particularly affected (31). In both untreated and treated HIV infection, extensive collagen deposition in lymphoid tissues drives profound pathological changes, leading to complete loss of tissue architecture, paracortical T cell zone damage and depletion of CD4⁺ T cells (88, 89). Even in PWH who have achieved virological suppression on ART, tissue fibrosis progresses through persistent pro-fibrotic signaling and chronic low-grade inflammation (90). The expansion of fibrotic regions impairs immune reconstitution and tissue repair mechanisms, thereby compounding HIV-associated tissue damage and pathogenesis while contributing to viral persistence (91).

Fibrosis in the context of HIV infection is strongly driven by TGF-β, which promotes the excess deposition and limited turnover of extracellular matrix (ECM) proteins that form the basis of fibrotic tissue (92, 93). In healthy individuals, the transient deposition of ECM proteins facilitates wound healing with matrix metalloproteinases (MMPs) promoting ECM turnover to prevent pathological scarring. However, when TGF-β remains chronically elevated during long-term HIV infection, ECM proteins can accumulate pathologically and lead to the development of fibrotic tissue (Figure 3). TGF-β promotes the differentiation of epithelial cells and fibroblasts into myofibroblasts (92, 93), which are markedly more active in producing ECM proteins, such as collagen, in comparison to resident cells in healthy tissues (94). Furthermore, TGF-β also inhibits ECM degradation by downregulation of MMP expression, which impairs ECM clearance and amplifies fibrotic remodeling (88). Pharmacologic targeting of the TGF-β pathway, such as with anti-fibrotic drug pirfenidone, significantly reduced myofibroblast-driven ECM production, reinforcing TGF-β’s central role in HIV-associated lymphoid fibrosis (24). Importantly, treatment of rhesus macaques with pirfenidone at the time of SIV infection prevented lymphoid tissue fibrosis in vivo and was associated with preservation of CD4⁺ T cells in both lymph nodes and blood, while administration post-infection was less effective (95). In these studies, interference with TGF-β signaling by pirfenidone was responsible for the effect, confirming TGF-β as a target for prevention or reversal of lymphoid fibrosis and improved immune reconstitution in HIV (95).

While lymphoid tissue fibrosis is a hallmark of long-term HIV infection (31), TGF-β is also strongly implicated in fibrogenesis in several non-lymphoid tissues (Figure 3). In the liver, elevated TGF-β acts synergistically with IFN-γ to enhance inflammatory and fibrotic pathways, contributing to non-alcoholic steatohepatitis (NASH) and metabolic dysfunction-associated steatotic liver disease (MASLD) in both humanized mouse models and in PWH (96, 97). Notably, MASLD in PWH has been reported to exhibit more advanced stages of fibrosis despite lower inflammatory activity compared to uninfected controls with MASLD, suggesting an outsized role for TGF-β in driving disease in PWH (98). Recent studies in SIV-infected macaques further support a central role for TGF-β in hepatic inflammation and fibrogenesis. These analyses revealed that, in addition to innate immune populations (99), hepatic CD4⁺ T cells are directly infected during SIV infection and exhibit elevated TGF-β mRNA expression (100). The presence of TGFB1-expressing infected CD4⁺ T cells within the liver microenvironment provides an additional localized source of TGF-β, potentially amplifying pro-fibrotic signaling and contributing to NASH-like pathology in chronic infection. Clinical trials of hydronidone, a modified pirfenidone derivative designed to more specifically target hepatic fibrosis via inhibition of TGF-β signaling, have shown significant benefit in Hepatitis B-driven liver fibrosis (101), suggesting that similar pharmacological approaches may improve liver fibrosis in the context of HIV.

Elevated TGF-β levels have also been associated with HIV-linked cardiac fibrosis and increased cardiovascular risk in PWH (102, 103). Cardiac magnetic resonance imaging (MRI) studies have demonstrated that myocardial fibrosis is three to four times more prevalent in PWH compared to PWoH (104). Pirfenidone has demonstrated efficacy in reducing myocardial fibrosis in heart failure patients, as evidenced by decreased myocardial extracellular volume after 52 weeks of treatment (105). However, because TGF-β also contributes to tissue repair following myocardial infarction, larger clinical trials are required to determine whether TGF-β inhibition represents a safe and effective strategy for managing HIV-associated cardiac fibrosis.

Additionally, TGF-β has been linked to worsened renal disease and chronic nephropathy in PWH through progressive fibrotic accumulation (25, 106). Unlike hepatic and myocardial fibrosis, however, clinical strategies targeting TGF-β in renal disease have been largely unsuccessful. Clinical trials with LY2382770, a TGF-β-specific monoclonal antibody, were terminated early due to lack of efficacy in improving diabetes-associated nephropathy (107). Pirfenidone demonstrated modest improvement in glomerular filtration rate, but these studies were limited by small sample size and high dropout rates (108). Clinical studies on the therapeutic targeting of TGF-β in the context of renal disease have not focused on PWH as a treatment group, and further large-scale investigations are required to determine whether TGF-β inhibition has therapeutic potential in HIV-associated nephropathy.

Beyond its role as a driver of tissue fibrosis, a process tightly linked to the development of aging-related disorders (109), TGF-β contributes in several other ways to the broader processes of immune and systemic aging in PWH. Long-term HIV infection, even in the context of ART-mediated viral suppression, is associated with the premature onset of age-related comorbidities relative to age-matched HIV-negative controls, including cardiovascular disease, renal disease, neurocognitive decline, and metabolic multimorbidity (110–112). The mechanisms driving this accelerated aging phenotype among PWH appear to be multifactorial, but TGF-β-mediated fibrosis, chronic inflammation, dysregulated cytokine signaling (as evidenced by elevated IL-6, TNF-α, and sCD14 levels (113)), and cellular senescence (88, 114) all appear to be central contributors. Persistent TGF-β signaling promotes several of these factors that accelerate immunologic and systemic aging (89, 115, 116), providing a mechanistic link between chronic inflammation and accelerated aging observed in PWH.

TGF-β-driven fibrosis mirrors several core mechanisms of organ aging. The resulting ECM stiffening and architectural distortion impair tissue elasticity, disrupt stem-cell niches, and limit regenerative capacity, mechanisms that align closely with hallmarks of aging such as stem-cell exhaustion, cellular senescence, and impaired immune regeneration. In PWH, these fibrosis-aging dynamics emerge early and persist despite ART. For example, fibro-collagenous remodeling in lymph nodes and gut-associated lymphoid tissue has been shown to restrict access to key T cell survival signals (e.g., IL-7), compromise CD4⁺ T-cell homeostasis, and interfere with immune reconstitution, functionally “aging” the immune microenvironment even in virologically suppressed individuals (115). Furthermore, stiffening of the ECM has been shown to induce senescence in alveolar epithelial cells in pulmonary fibrosis models (117) and mesenchymal stromal cells (stem/progenitor cell populations) in PWH show functional deficits in immunological non-responders (PWH who achieve viral suppression with ART, but do not fully restore their CD4⁺ T cell counts) (118). Thus, fibrosis reflects a convergent pathway through which TGF-β-based signaling reshapes tissue mechanics and repair programs, linking physiologic aging and HIV to a shared framework of premature multimorbidity in PWH.

Beyond fibrosis, TGF-β may also directly contribute to the epigenetic and metabolic reprogramming of immune and stem cell compartments in PWH. CD4⁺ T cells from PWH on suppressive ART exhibit transcriptional and metabolic signatures resembling those of older uninfected individuals, including elevated exhaustion marker expression, reduced mitochondrial function, and impaired proliferative capacity (119). These dysfunctions may be amplified in TGF-β-rich environments, potentially promoting effector-to-regulatory shifts and activating SMAD-mediated transcriptional programs that reinforce cellular senescence.

TGF-β can also influence the stem and progenitor cell compartments altering differentiation potential and limiting self-renewal. Hematopoietic stem cells (HSCs) in aging individuals and PWH often exhibit skewed differentiation potential (with a notable bias towards myeloid cells), reduced self-renewal, and functional exhaustion—phenotypes linked to increased TGF-β signaling in the bone marrow microenvironment (120, 121). Likewise, in pigtailed macaques with asymptomatic SIV infection, elevated plasma TGF-β levels were inversely correlated with hepatic thrombopoietin (THPO) transcription and bone marrow megakaryocyte density, suggesting inhibitory effects on megakaryopoiesis and broader hematopoietic dysfunction (122).

Together, these findings position TGF-β as a potential central regulator of premature aging in PWH, affecting immune function, metabolic balance, epigenetic programming, and stem-cell maintenance. However, additional research is needed to determine whether therapeutic modulation of TGF-β can safely and effectively alleviate HIV-driven accelerated aging, particularly within dysfunctional and fibrotic tissues.

TGF-β signaling causes multiple downstream effects that can either enhance or inhibit HIV replication depending on the cell type and context. During the acute phase of HIV infection, activated CD4⁺ T cells that express high levels of the CCR5 co-receptor are preferentially infected (1, 123, 124). The level of CCR5 on the cell surface is strongly influenced by cytokines and T cell activation signals. For example, IL-6 and several Th1-type, pro-inflammatory cytokines (such as TNF-α, IL-2, and IL-12) have been reported to upregulate CCR5, whereas other T cell stimulation and other cytokines have been reported to downregulate it in a cell type dependent manner (125–127). TGF-β has been shown to increase CCR5 expression and to promote infection of both resting and activated memory CD4⁺ T cells in vitro (Figure 4), suggesting that TGF-β can enhance CCR5-tropic virus spread under certain conditions (128, 129). This has been similarly observed in myeloid cells, with one report showing that TGF-β increases viral replication in monocyte-derived macrophages (130).

By contrast, several other reports have shown that TGF-β suppresses HIV expression and spread in cells that are already chronically infected—for example, potent inhibition of virus production was observed in the promonocytic U1 cell line after TGF-β treatment (131–133). This inhibitory effect has been linked to direct modulation of HIV LTR activity by the TGF-β signaling cascade [for example, by regulation of non-canonical NF-κB (134) or BLIMP-1 (135)]. A more detailed discussion of the impact of TGF-β on HIV transcription is included in the next section. Beyond these direct transcriptional effects, broader phenotypic changes in TGF-β-treated HIV-1-target cells may likewise alter viral restriction factors and/or susceptibility to infection [i.e., by impacting cell differentiation (43, 136, 137), triggering changes in cell adhesion and trafficking (137, 138), or via metabolic reprogramming (54, 139)], but much more work in this area is still required.

These apparently discordant observations can be reconciled with some attention to experimental context: studies that report TGF-β–mediated increases in viral replication measured infection after exposing uninfected primary cells to virus in the presence of TGF-β, whereas the reports of suppression measured virus production from cell lines or primary cells already harboring integrated provirus. Taken together, these data are consistent with TGF-β playing a multifaceted, context-dependent role in HIV infection. During active replication (i.e., during acute infection or in the absence of antiretroviral therapy), TGF-β may increase susceptibility to infection and facilitate viral spread by inducing the expression of CCR5 (Figure 4), whereas in already infected cells, TGF-β signaling may inhibit viral gene expression and contribute to latency.

HIV persists in long-lived latent reservoirs composed primarily of quiescent CD4⁺ T cells that harbor integrated, intact DNA proviruses, but that do not actively produce replication-competent virions due to transcriptional or post-transcriptional blocks to viral gene expression (140, 141). Given that immune recognition largely depends on viral antigen expression, these latently infected cells can evade immune clearance and persist even under long-term suppressive ART (142, 143). This persistent viral reservoir represents the main barrier to an HIV cure (144, 145).

Initial studies of HIV latency were performed in immortalized cell line models. T cell line models including CEM-derived (i.e., 8E5, ACH-2) and Jurkat-derived (i.e., J-Lat, J1.1, Jurkat E4) lines demonstrated that cells with an integrated provirus can remain transcriptionally silent without producing infectious virions (146). These cell lines differ in their baseline levels of viral gene expression, in the site(s) of proviral integration, and in viral genotype, all of which can influence maintenance of the latent state and reactivation in response to stimulation (146). A large number of blocks to proviral gene expression have been described in these models, including epigenetic, transcriptional, and post-transcriptional blocks. Some of these blocks can be overcome by exposure to latency-reversing agents (LRAs) that stimulate viral gene expression through a variety of mechanisms. Well-characterized LRAs include: PKC agonists such as PMA (phorbol 12-myristate 13-acetate) (147), canonical NF-kB agonists such as TNF-α (Tumor necrosis factor alpha) (148), non-canonical NF-kB agonists such as AZD5582 (149), histone deacetylases (HDAC) inhibitors such as vorinostat (150), BET bromodomain inhibitors such as JQ1 (151), etc. Analogous models have also been established in myeloid lineages. The U1 cell line (derived from U937 promonocytes) and the OM10.1 cell line (derived from HL-60 promyelocytes) similarly maintain integrated, but transcriptionally inhibited, proviruses, which can be reactivated by cellular stimulation or cytokine exposure (152).

Early work with the U1 cell line showed that PMA-induced viral reactivation could be dampened by co-treatment with TGF-β (132). Interestingly, the TGF-β driven decrease in reactivation was not observed when TNF-α was used as the LRA in the same study, even though both PMA and TNF-α are thought to act through activation of the NF-kB pathway (153, 154). Similar results have been reported in the ACH-2 T-cell model, with TGF-β dampening viral reactivation upon PMA treatment (133). The determinants underlying the repressive effect TGF-β has on some LRAs in some cell line models of latency remain to be fully described. While the mechanistic basis for TGF-β’s inhibitory effects on reactivation requires further exploration, these findings suggest that TGF-β’s latency-enforcing properties may extend across diverse cell types that compose the reservoir. Comprehensive investigations examining TGF-β effects in diverse latency models, encompassing both myeloid-derived and T-cell-derived systems, will be essential to fully understand how this cytokine influences HIV latency dynamics in different cellular contexts.

TGF-β’s involvement in HIV latency has also been examined in primary cells and cells from PWH, which offer more physiologically relevant insights than immortalized cell lines. Primary cell latency model development has concentrated predominantly on memory CD4⁺ T cell populations, as memory cells exist in a quiescent state and constitute the majority of the HIV reservoir in vivo (155–157). The first documented example of TGF-β contributing to HIV latency in primary T cells occurred during early efforts to establish a T_cm differentiation protocol for latency induction (158, 159). In these early studies, researchers aimed to differentiate T_cm cells from naïve CD4⁺ T cells by activating them with αCD3/αCD28 in presence of αIL-4, αIL-12, and TGF-β. Following HIV infection, these differentiated cells maintained viral latency and could be reactivated upon stimulation with αCD3/αCD28 and IL-2. While TGF-β’s latency-promoting properties were not the intended focus of these experiments, this work provided the first evidence that TGF-β could facilitate HIV latency establishment in primary T cells (158, 159).

This model of latency generation was subsequently refined by a different group, who developed the QUECEL (quiescent effector cell latency) model to generate well-defined, quiescent memory T cells through controlled differentiation and polarization (160). In this approach, naïve CD4⁺ T cells are first activated with αCD3/αCD28 in the presence of specific polarization cytokine cocktails to produce Th1, Th2, Th17, and Treg effector populations. These expanded effector cells are then infected with HIV-1 and driven into a memory-like latent state using a combination of TGF-β, IL-8, and IL-10. Although IL-8 and IL-10 alone were able to induce partial latency, the inclusion of TGF-β was required for full latency establishment in this model (160), underscoring its critical role in promoting the transition from active effector phenotypes to resting, memory-like states across multiple T cell subsets. Notably, the transcriptional and reactivation profiles of cells derived from the QUECEL model closely mirror those of ex vivo PBMCs isolated from PWH on long-term ART, highlighting its physiological relevance as a primary cell model for studying HIV latency.

Other primary cell latency systems have been developed based on similar principles, including the LARA (latency and reversion assay) model (155, 161). This model was based on the observation that the majority of the inducible viral reservoir in vivo is present in effector memory (T_em) subsets rather than central (T_cm) or transitional (T_tm) memory CD4⁺ T cells (155) (Figure 5). This is in alignment with other studies that have reported an enrichment of the HIV reservoir in T_em cells compared to other memory subsets (162, 163). However, several recent studies have suggested that the enrichment of intact proviruses in specific cell subsets may depend on the timing of ART (164, 165), immune responses (166) and sex differences (167). That being said, the association between an effector memory signature and higher reservoir inducibility is also supported by modeling studies (168), cross-sectional studies (169), and evidence suggesting that cellular differentiation promotes viral expression and immune recognition (170, 171).

The LARA model begins with the isolation of T_cm, T_tm, and T_em CD4⁺ T cell subsets from HIV-negative donors, which are subsequently infected with HIV (161). Culture conditions are designed to mimic the lymph node microenvironment that supports long-term T cell survival and quiescence. Specifically, cells are maintained in media containing TGF-β and IL-7 (cytokines involved in memory T cell establishment and survival (172)) that is further supplemented with conditioned medium from the H-80 stromal cell line, which is rich in TGF-β1, TGF-β2, TGF-β3, and IL-9. Over two weeks, this culture system induces a marked decrease in T_tm and T_em populations, accompanied by a relative increase in quiescent T_cm cells (155, 161). Given that T_em cells more readily revert to a virus-producing phenotype (171), this shift toward a more undifferentiated memory state suggests a mechanism by which TGF-β signaling may facilitate HIV latency maintenance. TGF-β signaling is known to promote an undifferentiated, stem-like transcriptional profile in T cells through both epigenetic and metabolic mechanisms (49, 173, 174), though the exact contributions of these pathways to HIV latency remain to be determined (Figure 5).

Another line of evidence supporting TGF-β’s role in maintaining HIV latency comes from studies showing that pharmacological inhibition of TGF-β signaling enhances latency reversal across multiple model systems (62, 133). Galunisertib, a small-molecule inhibitor of TGFβRI (ALK5) that reached Phase I/II clinical development, prevents SMAD2/3 phosphorylation and downstream signaling through the canonical pathway (175). In vitro, TGF-β prevented PMA-mediated reactivation of viral transcription in both cell line models and in primary CD4⁺ T cells isolated from PBMCs of PWH, but TGF-β-blockade using galunisertib is sufficient to restore reactivation (133). Furthermore, in vivo, galunisertib promoted viral reactivation in a small cohort of seven SIVmac251-infected rhesus macaques on long-term ART (63). This observation was subsequently confirmed in a follow-up study of eight SIVmac239M-infected rhesus macaques on ART, where latency reversal was documented in both blood and tissue compartments using classical molecular assays as well as SIV-envelope immuno-PET/CT imaging (62). Mechanistically, TGF-β blockade in vivo was associated with transcriptional reprogramming toward a transitional effector phenotype characterized by increased expression of activation- and metabolism-related transcriptional programs (Figure 5), yet notably without upregulation of canonical activation or proliferation markers (62). These findings suggest that HIV latency reversal could be achieved through TGF-β blockade without risk of inducing systemic immune activation or inflammation.

Finally, recent clinical studies characterizing the transcriptionally silent, replication-competent reservoir in long-term virologically suppressed individuals by comprehensive single-cell sequencing and phenotypical analysis provide additional evidence for a role of TGF-β in shaping the HIV reservoir (166). These studies revealed enrichment of TGFβRI expression in CD4⁺ T cells harboring intact proviral transcripts, along with markers of T cell exhaustion, such as PD-1 and TIGIT (166). These findings suggest that TGF-β signaling contributes to maintenance of the latent reservoir and promotes the persistence of these cells in PWH on ART, even after decades of suppressive therapy.

Taken together, evidence from in vitro latency models, ex vivo studies, in vivo blockade experiments, and clinical specimen characterization converge to identify TGF-β as a key regulator that links T cell quiescence and viral persistence. TGF-β appears to contribute to HIV latency through at least two mechanisms: first, by facilitating the effector-to-memory cell transition (176) leading to latency establishment through transcriptional and epigenetic remodeling; and second, by maintaining cellular quiescence and stem-like properties through metabolic suppression, particularly via mTOR inhibition (173) and its effects on mitochondria respiration (54). These mechanisms likely act synergistically to both establish and stabilize HIV latency across diverse biological compartments, raising the threshold for latency reactivation and promoting the selection of long-lived reservoir cells that persist despite decades of suppressive ART (Figure 5). That being said, significant gaps remain in our understanding of TGF-β’s role across these different reservoir compartments, cell types, and experimental systems, necessitating further investigation to elucidate how TGF-β signaling shapes HIV persistence in vivo and how it can be targeted to eliminate the viral reservoir and induce virologic control.