Glucocorticoid regulation of the mTORC1 pathway modulates CD4⁺ T cell responses during infection

Abstract

Objectives

Conventional glucocorticoid (GC) treatment poses significant risks for opportunistic infections due to its suppressive impact on CD4⁺ T cells. This study aimed to explore the mechanisms by which GCs modulate the functionality of CD4⁺ T cells during infection.

Methods

We consistently measured FOXP3, inflammatory cytokines and phospho‐S6 ribosomal protein levels in CD4⁺ T cells from patients undergoing conventional GC treatment. Using Foxp3^EGFP animals, we investigated the dynamic activation of the mechanistic target of rapamycin complex 1 (mTORC1) pathway and its correlation with the immunoregulatory function of CD4⁺ T cells under the influence of GCs.

Results

GCs dynamically altered the expression pattern of FOXP3 in CD4⁺ T cells, promoting their acquisition of an active T regulatory (Treg) cell phenotype upon stimulation. Mechanistically, GCs undermined the kinetics of the mTORC1 pathway, which was closely correlated with phenotype conversion and functional properties of CD4⁺ T cells. Dynamic activation of the mTORC1 signaling modified the GC‐dampened immunoregulatory capacity of CD4⁺ T cells by phenotypically and functionally bolstering the FOXP3⁺ Treg cells. Interventions targeting the mTORC1 pathway effectively modulated the GC‐dampened immunoregulatory capacity of CD4⁺ T cells.

Conclusion

These findings highlight a novel mTORC1‐mediated mechanism underlying CD4⁺ T cell immunity in the context of conventional GC treatment.

This study underscores the long‐term immunomodulatory impact of conventional glucocorticoid treatment on CD4⁺ T cells during infection, revealing a novel role of the mTORC1 pathway in this immune regulatory process. It highlights the necessity for careful utilisation of glucocorticoids in the clinical management of autoimmune and inflammatory diseases.

Introduction

The conventional glucocorticoid (GC) regimen is a widely recognised strategy for rapidly and effectively achieving complete remission in a variety of steroid‐responsive autoimmune and inflammatory diseases, surpassing existing alternative immunosuppressants in effectiveness. However, an increased risk of opportunistic infections, a significant drawback, arises from prolonged GC usage due to their potent immunosuppressive effects. CD4⁺ T cells, serving as essential elements of the adaptive immune system, orchestrate both cellular and humoral immune responses. ¹ Clinical evidence substantiates that persistent reduction in CD4⁺ T cell counts occurs following conventional GC therapy, enduring even after treatment cessation, particularly in patients on extended steroid regimens. ² , ³ Given the pivotal role steroids play in T cell development and functionality, ⁴ dysregulations of CD4⁺ T cells might considerably impact immunity during long‐term GC exposure. However, the precise mechanisms underlying these effects largely remain elusive, thus underscoring the importance of comprehensive investigations into these dynamics.

CD4⁺ T cells differentiate into specific subpopulations upon activation. ⁵ These include a heterogeneous group of T regulatory cells (Tregs) characterised by specific surface markers and varying levels of FOXP3 expression. ⁶ Thymus‐derived Tregs.

specialise in potent immunosuppression with stable FOXP3 expression. In contrast, peripheral Tregs generally exhibit transient or low FOXP3 expression unless antigenically stimulated. ⁷ , ⁸ Recent studies have demonstrated that GCs regulate the survival, maturation and differentiation of Treg subsets in a complex and context‐dependent manner. ⁹ Notably, the dynamics of FOXP3 expression in CD4⁺ T cells strongly align with the long‐term results of glucocorticoid therapy in patients suffering from glomerular disease. ¹⁰ Furthermore, our latest findings underscore the critical epigenetic regulatory role of long‐term glucocorticoid treatment in the biology of CD4⁺ T cells. ¹¹ Collectively, current evidence suggests that GCs promote anti‐inflammatory Th2 and Treg responses while suppressing pro‐inflammatory Th1 and Th17 responses. However, these effects are influenced by multiple factors, including the dose and duration of GC treatment, the timing of glucocorticoid administration (before or after stimulation), and the activation state of the T cells. ¹² Despite these insights, it remains unclear how GCs uniquely influence the various subsets of CD4⁺ T cells and their response to infections in vivo. This is especially complex during conventional GC treatment for human diseases.

In this scenario, the mammalian target of rapamycin complex 1 (mTORC1) could potentially play a pivotal role. mTORC1 controls CD4⁺ T cell survival and differentiation, and its disruption could lead to autoimmune or inflammatory diseases. ¹³ mTORC1 inhibition can promote FOXP3 expression and suppress effector T cell development, ameliorating autoimmune phenotypes in various models. ¹⁴ From a mechanistic standpoint, the inhibition of mTORC1, either through the use of rapamycin or by deletion of its essential component, Raptor, has been shown to stabilise FOXP3 mRNA, enhance TGF‐β signalling, and boost FOXP3 expression, consequently driving Treg cell differentiation. ¹⁵ Nonetheless, the interaction between mTORC1 activity and Treg function is intricate and context dependent. Specifically, the combination of mTORC1 inhibition with immunosuppression could potentially impair the regulatory responses under certain conditions. ¹⁶ , ¹⁷

Emerging evidence highlights a critical interaction between GC and mTORC1, with far‐reaching implications across a range of biological activities and health conditions. For instance, vitamin D can enhance GC efficacy through mTORC1 inhibition in experimental models of multiple sclerosis, specifically affecting T cells. ¹⁸ In skeletal muscle homeostasis, GCs can trigger the suppression of mTORC1 activity, which contributes to muscle atrophy, while the activation of mTORC1 can inhibit the catabolic processes provoked by GCs. ¹⁹ GCs are understood to curtail pro‐inflammatory mediators through mTORC1 inhibition. However, it is critical to acknowledge that this same mTORC1 inhibition could potentially weaken the anti‐inflammatory potency of GCs. ²⁰ Therefore, the interplay between GC and mTORC1 strikes a delicate balance that is vital for managing the efficiency of immune responses in pathological conditions, such as infections. Understanding this complex balance could provide the foundation for innovative therapeutic strategies that optimise the curative potential of GCs while mitigating their adverse effects on the immune system.

Results

Modulation of CD4⁺ T cell inflammatory response by GCs through mTORC1 signalling

Multi‐parameter flow cytometry was used to assess the proportional changes in circulating CD4⁺, CD8⁺, CD4⁺CD25⁺ and CD4⁺FOXP3⁺ T cells in a cohort of MCD patients who had been receiving conventional GC treatment for more than 8 months (Figure 1a). GC treatment resulted in a decrease in the proportion of CD3⁺CD4⁺ T cells (Figure 1b) and an increase in the proportion of CD3⁺CD8⁺ T cells (Figure 1c), leading to a significant reduction in the CD4/CD8 ratio (Figure 1d). Furthermore, GC treatment significantly promoted regulatory CD4⁺ T cell subsets, as evidenced by an increase in CD4⁺CD25⁺ (Figure 1e) and CD4⁺FOXP3⁺ (Figure 1f) T cells. In vitro experiments showed that antigenic stimulation markedly increased IFNG mRNA expression, an effect that was inhibited in the GC‐exposed CD4⁺ T cells (Figure 1g). Conversely, GC treatment consistently enhanced IL10 mRNA expression (Figure 1h).

Figure 1.

Conventional GC treatment sustainedly promotes regulatory CD4⁺ T cells. A time‐course study was conducted on a cohort of five patients, who were diagnosed with biopsy‐proven MCD and presented with an initial episode of nephrotic syndrome. All the subjects received the conventional GC treatment for at least 8 months as described in the Methods section. At each predetermined time point, circulating T cells were isolated and subjected to staining with fluorochrome‐conjugated antibodies to identify the relevant cell population. (a) Gating strategies to discriminate different T cell subpopulations by multiparametric flow cytometry with markers indicated in the x and y axes. (b–f) Analyses of CD3⁺CD4⁺, CD3⁺CD8⁺, CD4/CD8 ratio, CD4⁺CD25⁺ and CD4⁺FOXP3⁺ in circulating T cells. For the in vitro experiments, CD3⁺CD4⁺ T cells were isolated at 0 month and 2 months after GC exposure and subjected to phytohemagglutinin stimulation (5 μg mL⁻¹). (g, h) qPCR analysis of mRNA expression of IFNG and IL10 at designated time points in the cultured CD4⁺ T cells. The results of all individuals (n = 5 in each group) are shown. Data were obtained from at least three independent experiments and are presented as means ± SEM. The differences between groups were analysed using a paired t‐test. P‐values: *P < 0.05, **P < 0.01.

Herpes zoster infection (HZI) is a common complication in patients with long‐term GC treatment. In a GC‐free condition, HZI could induce prominent expression of IFNG in the CD4⁺ T cells (GF/HZI, Figure 2a). This inflammatory response to HZI was observed to be similar in the GC‐exposed group (GCs/HZI). However, HZI enhanced IL10 production of CD4⁺ T cells that were particularly significant in the GC‐exposed group (Figure 2b). Consistently, a time‐course study confirmed that IL10 production upon antigenic stimulation markedly increased in the GC‐exposed CD4⁺ T cells during the entire treatment period (Figure 2c). We speculated that GC promoted FOXP3 expression, which determines the differentiation and suppressor functions of regulatory T cell subsets. In‐depth stratification analysis revealed that global FOXP3 protein level was markedly enhanced in the GC‐exposed CD4⁺ T  cells, which was mostly ascribed to proportional increases of FOXP3^low and FOXP3^medium subsets (Figure 2d). In a GC‐free condition, FOXP3 expression was slightly altered by HZI (GF/HZI), which, however, significantly promoted the subsets of FOXP3^medium and FOXP3^high particularly in the GC‐exposed CD4⁺ T cells. These findings imply that GC promotes a proportional increase of peripheral FOXP3^low Treg cells. The composition of Treg types can undergo a substantial shift in response to an infection. This shift is characterised by a surge in FOXP3^high subsets, which are possibly associated with increased IL10 production in the CD4⁺ T cells.

Figure 2.

γGCs enhance the regulatory T cell response upon infection. The time‐course study and the in vitro experiment are as described in Figure 1. Meanwhile, human CD4⁺ T cells were isolated from a subgroup of MCD patients (n = 6 in each group) with the incidence of herpes zoster infection (HZI) or not, including GC‐free and uninfected (GF/uni), GC‐treated and uninfected (GCs/uni), GC‐free and infected (GF/HZI) and GC‐treated and infected (GCs/HZI). (a, b) Flow cytometry analysis of the intracellular expression of IFN‐γ and IL10 of circulating CD4⁺ T cells upon fresh isolation. (c) Time‐course of IL10 ELISPOT assay of the cultured CD4⁺ T cells with phytohemagglutinin stimulation (5 μg mL⁻¹). (d) Stratification analysis of FOXP3 expression in the circulating CD4⁺ T cells. One representative experiment and the results of all individuals are shown. Data were obtained from at least three independent experiments and are presented as means ± SEM. Differences between groups were compared using a Wilcoxon rank‐sum test or a Student t‐test. P‐values: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Regulation of immune function in GC‐exposed CD4 ⁺ T cells by the mTORC1 pathway

The mTOR complex 1 (mTORC1) signalling pathway has been shown to play a pleiotropic role in the differentiation of CD4⁺ T cells. ²¹ Using a real‐time qPCR array, we found that GC significantly altered the mRNA expression levels of multiple mTORC1 regulator genes in CD4⁺ T cells. Our expression pattern analysis revealed that GC suppressed the transcription within the mTORC1 pathway by predominantly inhibiting positive regulator genes while promoting negative ones (Figure 3a). Consistent with this, a time‐course flow cytometry assay on CD4⁺ T cells showed that GC sustainably dampened their p‐S6 activity (p‐S6^medium and p‐S6^high), a downstream effector of the mTORC1 pathway (Figure 3b). Notably, compared to the untreated cells (GCs‐0 m), GC increased the proportion of CD4⁺ T cells with baseline mTORC1 activity (p‐S6^low). Next, the CD4⁺ T cells were isolated from the time‐course MCD patients based on their levels of p‐S6 expression. Subsequent qPCR assays revealed that GC significantly reduced the mRNA expression of IFNG in both p‐S6^low and p‐S6^high subsets (Figure 3c). In contrast, GC increased the mRNA expression of IL10 (Figure 3d) and FOXP3 (Figure 3e) in p‐S6^low cells but inhibited these genes in the p‐S6^high subsets, suggesting distinct effects of GC on the regulatory function of CD4⁺ T cells in an mTORC1‐dependent manner.

Figure 3.

The mTORC1 pathway regulates the inflammatory function of GCs‐exposed CD4⁺ T cells. The analysis of circulating T cell population was described in Figures 1 and 2. The data are presented as relative fold changes to facilitate comparisons between groups. (a–e) The time‐course experiment was conducted on a cohort of 5 MCD patients undergoing conventional GC treatment. (a) Real‐time PCR array to show the fold changes of mTORC1 pathway regulator genes in the freshly isolated CD4⁺ T cells after 2 months of GC treatment. (b) Stratification analysis of the intracellular expression of phospho‐S6 ribosomal protein (p‐S6) in the freshly isolated CD4⁺ T cells. (c–e) Time‐course of mRNA expression of IFNG, IL10 and FOXP3 in the sorted CD4⁺ T cells based on their p‐S6 expression. The gene expression was normalised to age‐matched healthy controls (Ctrl, n = 6). (f, g) Human CD4⁺ T cells were isolated from a subgroup of MCD patients with the incidence of herpes zoster infection (HZI) or not. After fresh isolation, cells were sorted based on their p‐S6 expression and subjected to real‐time‐PCR assays on IFNG and IL10. (h, i) A portion of CD4⁺ T cells was isolated from the healthy or MCD subjects after 2 months of GC treatment and applied to the antigen stimulation experiment with the mTORC1‐targeted siRNA interventions (PTEN siRNA and PS6K siRNA), followed by western blot assay on protein expression of PTEN, PS6K, PS6, mTOR and p‐mTOR and flow cytometry analysis of intracellular expression of IFNG and IL10. The results of all individuals (n = 5 in each group) are shown. Data were obtained from at least three independent experiments and are presented as boxes and whiskers. Differences between groups were compared using a paired t‐test. P‐values: *P < 0.05, **P < 0.01, ***P < 0.001.

Furthermore, under conditions of HZI, mRNA expression levels of both IFNG and IL10 in CD4⁺ T cells were significantly higher in the p‐S6^high subset than in the p‐S6^low subset, regardless of GC exposure (Figure 3f and g). This indicates an mTORC1‐mediated regulation in response to infection. Subsequently, intervention experiments were conducted to elucidate the regulatory role of the mTORC1 pathway in CD4⁺ T cells. The effectiveness of siRNA‐mediated knockdown of PS6K and PTEN was validated by immunoblotting assays, which demonstrated that pathogen stimulation resulted in significant phosphorylation and activation of the mTORC1 pathway in cultured CD4⁺ T cells, which could be strongly suppressed by PS6KsiRNA, but promoted by PTENsiRNA (Figure 3h). Flow cytometry analysis revealed that inhibition of mTORC1 by PS6KsiRNA exerted potent inhibitory effects on the GC‐induced production of IL10, while activation of mTORC1 by PTENsiRNA further promoted the cellular production of both IL10 and IFNG in GC‐exposed CD4⁺ T cells (Figure 3i). These results collectively indicate an indispensable role of mTORC1 signalling in the regulation of CD4⁺ T cell‐mediated inflammatory response to pathogens during GC treatment.

mTORC1 pathway determines phenotype conversion of human CD4⁺ regulatory T cells

Using multi‐parameter flow cytometry, we aimed to evaluate the pattern of phenotype conversion of CD4⁺ T cells by analysing the co‐expressions of p‐S6 and FOXP3 under different conditions (Figure 4a). The quadrants of flow cytometry analysis, which were gated on the expressions of p‐S6 and FOXP3, indicate the activity of the mTORC1 pathway and the regulatory capacity of CD4⁺ T cells. Quadrant one (Q1) represents the proportion of active CD4⁺ effector T cells (FOXP3^low and p‐S6^high), which was not influenced by GCs but markedly promoted by an HZI under both GC‐free and GC‐exposed conditions (Figure 4b). The proportion of inactive CD4⁺ effectors (Q4) decreased significantly upon HZI (Figure 4c), indicating enhanced activation of CD4⁺ effectors against infection. Quadrant two (Q2) indicates the proportion of active CD4⁺ regulatory T cells (FOXP3^high and p‐S6^high), which was slightly suppressed by the GCs (Figure 4d). However, this proportion could be significantly promoted upon an HZI in the GC‐exposed T cells (GCs/HZI), while inactive Treg cells (FOXP3^high and p‐S6^low, Q3) differed slightly between the groups (Figure 4e), indicating the enhanced regulatory function of the GC‐exposed CD4⁺ T cells during infection. Further analysis showed that an infection markedly activated the CD4⁺ T cells, particularly in the presence of GCs (Figure 4f and g). Notably, GCs significantly suppressed the effectors (Figure 4h) but promoted the regulators (Figure 4i) of the CD4⁺ T cells during infection. Collectively (as presented in Figures 3 and 4), a combined analysis of p‐S6 and FOXP3 co‐expression patterns may highlight the key role of the mTORC1 pathway in the regulation of CD4⁺ effector and regulatory cells (Figure 4j).

Figure 4.

The co‐expression pattern of FOXP3 and p‐S6 indicates phenotype conversion of CD4⁺ T cells upon infection. Human CD4⁺ T cells were isolated from a subgroup of MCD patients as described in Figure 2 and subjected to in‐depth flow cytometry analysis (n = 5 in each group). (a) One representative experiment to show the gating strategies of FOXP3 and p‐S6 expression in CD4⁺ T cells by multiparametric flow cytometry. (b–i) Based on the co‐expression pattern of FOXP3 and p‐S6, the quadrants (Q1–Q4) and their combinations accordingly signified the proportions of effector and regulator of CD4⁺ T cells, including Q1: Active CD4⁺ effectors (b), Q4: Inactive CD4⁺ effectors (c), Q2: Active CD4⁺ regulators (d), Q3: Inactive CD4⁺ regulators (e), Q1 + Q2: Active CD4⁺ T cells (f), Q3 + Q4: Inactive CD4⁺ T cells (g), Q1 + Q4: CD4⁺ effectors (h), and Q2 + Q3: CD4⁺ regulators (i). (j) A schematic model to summarise dynamic regulation of mTORC1 pathway on the phenotype conversion of CD4⁺ T cells in response to GC exposure and an infection. Data were obtained from at least three independent experiments and are presented as violin plots. The differences between groups were analysed using one‐way ANOVA. P‐values: *P < 0.05, **P < 0.01, ***P < 0.001.

Chronic GC treatment impairs CD4⁺ T cell immunity during sepsis

Our recent study uncovers a causal correlation between sustained exposure to GCs and an exacerbated inflammatory reaction driven by compromised CD3⁺ T cells during infections. ¹² To further probe the potential involvement of mTORC1 in this T cell impairment, we implemented a sepsis mouse model using cecal ligation and puncture (CLP). Our results showed that GC‐treated animals had a significantly higher mortality rate following CLP induction than that of the vehicle controls (Figure 5a). Serum levels of IFN‐γ and IL10 were dramatically elevated at 24 h post‐CLP, particularly in the GC‐exposed animals (Figure 5b and c). Flow cytometry assays revealed a substantial increase in the number of peritoneal CD45⁺ leukocytes in the GC‐exposed group, but a significant reduction in the number of peritoneal infiltrating CD3⁺ T cells (Figure 5d and e). While both CD8⁺ and CD4⁺ subsets were markedly reduced, the proportional reduction of CD4⁺ subset was more pronounced, leading to a significant decline in the CD4/CD8 ratio in the GCs group (Figure 5f–h). Enhanced apoptosis and necrosis of CD4⁺ T cells were observed in the GC‐exposed group, which resulted in a reduction of peritoneal infiltrating CD4⁺ T cells (Figure 5i). Moreover, GC exposure suppressed the production of IFN‐γ but promoted that of IL10 in peritoneal CD4⁺ T cells under CLP challenge (Figure 5j and k). These results collectively demonstrate that chronic GC exposure exacerbates CD4⁺ T cell death and impairs their inflammatory response, ultimately leading to increased mortality in sepsis.

Figure 5.

Chronic glucocorticoid exposure compromises CD4⁺ T cells in a sepsis model induced by CLP. Male C57BL/6 mice were intraperitoneally administered with Dexamethasone (100 μg) twice weekly for a duration of 1 month. Then, mice were subjected to cecal ligation and puncture (CLP) or sham operation as described in the Methods, and samples were collected at 24 h after CLP. (a) Kaplan–Meier curve to show animal survival after CLP operation (n = 10 in each group). (b, c) ELISA assay of serum IFN‐γ and IL‐10 (n = 10 in each group). (d–k) Flow cytometry analysis (n = 6 in each group) of peritoneal CD45⁺ leukocytes, CD3⁺, CD8⁺, CD4⁺, CD4/CD8 ratio, cell death (apoptosis and necrosis) of splenic CD3⁺ T cells, and intracellular expression of IFN‐γ and IL‐10. The results of all individuals are shown. Data were obtained from at least three independent experiments and are presented as boxes and whiskers. The differences between groups were compared using an unpaired t‐test. P‐values: *P < 0.05, **P < 0.01.

To explore the change of regulatory T cells, we repeated CLP experiments on Foxp3^EGFP mice that co‐express EGFP and the X‐linked regulatory T cell‐specific transcription factor Foxp3. Flow cytometry analysis of splenocytes confirmed that more than 95% of Foxp3⁺ T cells were EGFP‐positive in the transgenic animals (Figure 6a). The proportion of EGFP‐expressing CD4⁺ T cells decreased following CLP, indicating an enhanced inflammatory response to infection (Figure 6b and c). However, chronic GC treatment favoured the phenotype conversion of EGFP⁺ T cells, which was particularly prominent during sepsis. Furthermore, qPCR analysis of splenic EGFP‐positive T cells demonstrated that GCs significantly increased mRNA expression of Treg‐associated genes in the CLP animals, including Il10, Ctla4 and Il2ra (Figure 6d–f). For the splenic EGFP‐negative CD4⁺ T cells, GCs significantly reduced the expression of their inflammatory genes upon CLP, including Tnf and Ifng (Figure 6g and h). These findings indicate that GCs undermine CD4⁺ T cell immunity against infection by promoting the FOXP3⁺ Treg cells phenotypically and functionally.

Figure 6.

Characterisation of CD4⁺ regulatory T cells in a CLP model using Foxp3^EGFP mice. Dexamethasone treatment and the cecal ligation and puncture (CLP) procedure were described in Figure 5, using male Foxp3^EGFP mice that co‐express EGFP and Foxp3 in the CD4⁺ T cell population. Splenic T cells were isolated 24 h post‐CLP induction and subsequently stained with fluorochrome‐conjugated antibodies to identify the specific populations under investigation. (a) Gating strategies to discriminate different T cell subpopulations by multiparametric flow cytometry with markers indicated in the x and y axes. (b, c) Cell sorting and analysis of EGFP⁺CD4⁺ T cells. (d–h) Real‐time‐PCR analysis to reveal mRNA expression of Il10, Ctla4 and Il2ra in the EGFP‐positive CD4⁺ T cells and of Tnf and Ifng in the EGFP‐negative CD4⁺ T cells. The results of all individuals (n = 6 in each group) are shown. Data were obtained from at least three independent experiments and are presented as means ± SEM. Group differences were analysed using one‐way ANOVA. P‐values: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Impact of mTORC1‐targeted strategies on glucocorticoid‐induced immunosuppression of CD4⁺ T cells

In accordance with the clinical observations, time‐course analysis revealed that GC treatment significantly increased the frequency of EGFP‐positive Treg cells, and this effect might continue after the cessation of GC treatment (Figure 7a). Concomitantly, the activity of the mTORC1 pathway, as determined by p‐S6 expression, was suppressed in these Treg cells (Figure 7b). Western blot analysis confirmed that GCs dampened the increased activity of mTORC1 pathway in CD4⁺ cells upon CLP challenge (Figure 7c). In the Foxp3^EGFP mice, co‐expression of FOXP3 and EGFP was restricted to the T cell lineage, primarily to the CD4⁺ T cell population (Figure 6a). Based on EGFP expression, splenic T cells were isolated from the Foxp3^EGFP mice with or without GC treatment and subjected to targeting‐mTORC1 interventions. Inhibition of mTORC1 by PS6K siRNA reversed the GC‐induced production of anti‐inflammatory cytokines IL10 and TGFβ in EGFP‐positive CD4⁺ Treg cell upon antigenic stimulation, while mTORC1 activation induced by PTEN siRNA significantly augmented the GC‐enhanced production of these cytokines (Figure 7d and e). Conversely, inhibition of mTORC1 further suppressed the production of pro‐inflammatory cytokines IFN‐γ and TNF‐α in GC‐exposed EGFP‐negative CD4⁺ effector T cells, whereas activation of mTORC1 reversed the inhibitory effect of GCs on the inflammatory response of these cells (Figure 7f and g). These results, consistent with the clinical findings, demonstrate that conventional GC treatment modulates the phenotype conversion between CD4⁺ Treg and effector T cells during pathogen response by dynamically regulating the mTORC1 pathway (Figure 4j).

Figure 7.

siRNA‐based interventions distinctly alter the regulatory effects of GC on CD4⁺ T cell subpopulations. The animal experiment was conducted using male Foxp3^EGFP mice as described in Figure 6. (a, b) Time‐course of EGFP expression in splenic CD4⁺ T cells and p‐S6 activity in the EGFP⁺ cells after GC treatment. (c) Western blot assay on protein expression of p‐mTOR and p‐S6 in splenic CD4⁺ T cells after isolation at 24 h of CLP. (d–g) For the in vitro experiments, a portion of splenic CD4⁺ T cells was isolated from the Foxp3^EGFP mice, which had been treated with DEX as described in Figure 5 or not. Based on whether they expressed EGFP or not, CD4⁺ T cells were isolated and subjected to PHA stimulation (5 μg mL⁻¹) for 72 h. To target mTORC1 pathway, PS6KsiRNA and PTENsiRNA were applied to a part of the cells following mitogenic stimulation. After interventions, flow cytometry analysis was performed to reveal cytokine expression of IL10 (d) and TGFβ (e) in the EGFP‐positive, and of IFNγ (f) and TNFα (g) in the EGFP‐negative T cells. The results of all animals (n = 6 in each group) are shown. Data were obtained from at least 3 independent experiments and are presented as boxes and whiskers. Differences between groups were compared using one‐way ANOVA. P‐values: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Discussion

CD4⁺ T cells are major targets of synthetic GC therapy, with the aim of controlling autoimmunity and inflammation. Early studies have demonstrated the potent lymphotoxic effects of high‐dose GCs through their genomic suppressive mechanisms. ²² , ²³ Recent findings have shown that GCs exert transcriptional inhibitions of Th1 and Th2 cells ⁴ but phenotypically favour Treg cells by promoting the mRNA expression of FOXP3 and IL‐10. ²⁴ The latest evidence further elucidates that synthetic GCs exert regulatory actions by directly targeting Treg cells through a microRNA‐mediated mechanism. ²⁵ However, there remains a lack of evidence, particularly in human studies, to clarify the dynamic interference of chronic GC exposure with the CD4⁺ Treg cell subsets. In this study, we demonstrate that chronic GC exposure disrupts the effector/regulator balance of CD4+ T cells, favouring immune suppression over immune defence against pathogens. This shift occurs, at least in part, due to increased IL‐10 production by CD4+ Treg cells. Intriguingly, the conventional GC regimen fosters the proportion of resting pTreg (FOXP3^low) while mildly suppressing the FOXP3^high subset in the periphery (Figure 2d). However, in the event of HZI, there is a marked increase in this fraction of active regulators in the CD4⁺ T cells subjected to GC exposure (Figure 4i). We propose that chronic exposure to GCs promotes FOXP3 gene expression in the resting CD4⁺ pTreg cells, which, upon encountering pathogens, can be functionally activated to exhibit heightened regulatory characteristics. This observation aligns with previous research suggesting that FOXP3^low quiescent Treg cells can rapidly transition into FOXP3^high proliferative effectors in response to inflammation. ²⁶ Consequently, an increase in resting pTreg cells potentially undermines T cell immunity against invasive microorganisms. The altered phenotype of Tregs among total CD4⁺ T cells thereby underlies the impairment of adaptive immunity associated with GC upon infection.

An increasing body of evidence underscores the crucial role of mTORC1 activation in guiding the T cell differentiation into Th1, Th2 and Th17 effector cell lineages. Conversely, the temporal deletion or inhibition of mTORC1 seems to foster the conversion of peripheral conventional T cells into Treg cells. ¹⁵ , ²⁷ , ²⁸ , ²⁹ , ³⁰ This viewpoint, however, seems to contrast with the emerging findings that mTORC1 activity is necessary for the maintenance of Treg functions. ²¹ , ³¹ An explanation to reconcile these observations has recently been proposed in animal models by classification of Treg cells into two different subsets based on their mTORC1 activities: the mTORC1^high effector Treg and the mTORC1^low resting Treg. mTORC1 inhibition favours the generation of the resting Treg but dampens the activity of the effector Treg. ³² This study complements this model in humans in the context of GC treatment. In brief, the mTORC1 pathway's baseline activity is critical in maintaining the balance of effector CD4⁺ T cells and Treg cells. During an infection, the mTORC1 pathway rapidly escalates in activated effector CD4⁺ T cells, thereby boosting their inflammatory cytokine production. Concurrently, mTORC1 activation enhances Treg functionality to circumvent an excessive inflammatory response, reinstating immune homeostasis. However, long‐term GC exposure could disrupt this mTORC1 regulation, leading to a decrease in effector CD4⁺ T cells and an increase in Treg conversion upon pathogen stimulation. In response to infection, mTORC1 activation in GC‐exposed effector and Treg cells could lead to delayed bacterial clearance and excessive inflammatory injury due to dysregulation of CD4⁺ T cell homeostasis. Mechanistically, our latest study uncovers that conventional GC treatment significantly undermines this regulatory mechanism, decreases effector CD4⁺ T cells' survival and disrupts Treg homeostasis. It consequently impairs the CD4⁺ T cell immunity against pathogens and their capacity to restore tissue homeostasis as well. ¹¹

Based on collective analysis of p‐S6 and FOXP3 expression in CD4⁺ T cells (as shown in Figure 4), we propose a novel model to interpret the dynamic regulation of mTORC1 signalling in CD4⁺ T cell biology. Normally, the baseline activity of the mTORC1 pathway is essential to maintain homeostasis of inactive effector CD4⁺ T cells (FOXP3^negativeP‐S6^low) and Treg cells (FOXP3^lowP‐S6^low). Once an infection occurs, the mTORC1 pathway can be rapidly promoted in the activated CD4⁺ effector (FOXP3^negativeP‐S6^high) and Treg cells (FOXP3^highP‐S6^high). In this process, the mTORC1 pathway is fundamental for appropriate regulation of both effector and regulatory T cells, which orchestrate a pro‐inflammatory milieu to eliminate the invasive pathogens and induce well‐timed inflammation resolution to facilitate tissue repair. As a result, mTORC1 crucially balances effector and regulatory T cells, facilitating inflammation resolution and tissue repair post‐pathogen elimination.

In a series of our previous investigations, we found that targeting mTORC1, either biochemically (e.g. using rapamycin and PTEN inhibitors) or genetically (e.g. through transgenic animals and siRNA), has significant regulatory effects on immune cells, especially CD4⁺ T cells. These interventions have consistently demonstrated significant impacts on outcomes of various human diseases, including AKI, DKD and MCD. ³³ , ³⁴ , ³⁵ This study consistently supports the idea that strategies targeting mTORC1 can sufficiently alter the GC‐dampened immunoregulatory capacity of human T cells. However, formulating personalised treatment strategies based on mTORC1 expression characteristics and activity levels remains a significant clinical challenge. To achieve this, further comprehensive mechanistic studies and robust preclinical data are needed for a careful evaluation of various factors, such as the timing of treatment and implementation of combined drug regimens, when devising treatment strategies that target mTORC1 signalling.

The present clinical study does have certain limitations. First, due to the relatively low frequency of HZI among patients receiving GC treatment, we utilised two distinct cohorts for analysing Treg subpopulations. While data from both cohorts show consistency, there may still be potential variations due to population differences. In addition, technical and ethical constraints limit the extensive evaluation of IL10 expression in human Treg subsets, given their typically low production under normal conditions. Future application of more advanced techniques like single‐cell transcriptomics could offer greater detail, permitting functional validation of shifts in Treg subpopulations during GC treatment.

In summary, this research uncovers a new mechanism, mediated by mTORC1, which underlies the compromised function of CD4⁺ T cells responding to an infection under the influence of long‐term GC therapy. A traditional GC regimen alters the mTORC1 activity of CD4⁺ T cells in a cell‐type‐specific manner. As a result, it substantially impairs the survival capacity of the CD4⁺ effector T cells while phenotypically promoting the resting Treg cells (mTORC1^low), which can be functionally transformed into the effector Treg (mTORC1^high) upon an infection. Of scientific importance, it remains necessary to further explore the molecular mechanisms underlying the distinct effects of mTORC1 signalling on diverse T cell subpopulations, especially in response to different antigenic stimulation. Our latest research has revealed the prolonged inhibitory impacts of a conventional GC treatment on CD4⁺ T cells in MCD patients, attributed to an epigenetic modulation of the mTORC1 pathway genes that is dose‐independent. ¹¹ This investigation bolsters the significance of comprehending the intricate interactions between glucocorticoids, the immune system and infectious diseases. Moreover, these findings could potentially pave the way for the development of more tailored and effective GC treatment strategies for patients with autoimmune and inflammatory conditions.

Methods

Study approval

The Human Research Ethics Committee of the Second Xiangya Hospital of Central South University reviewed and approved the study, which was conducted in accordance with the World Medical Association Declaration of Helsinki. All participants provided written informed consent upon admission.

Time‐course study

A longitudinal study was performed involving a group of five patients diagnosed with biopsy‐confirmed minimal change disease (MCD), all of whom were experiencing an initial episode of nephrotic syndrome. The study spanned 8 months and the patients were all well‐informed and underwent a standard conventional GC regimen. ³⁶ This protocol was initiated with high‐dose oral prednisone (1 mg/kg/day, up to a maximum of 80 mg/day) or an equivalent dosage of methylprednisolone for a duration of 6–8 weeks. This was followed by a tapering process, wherein prednisone dosages were progressively diminished by 5 mg every 2 weeks following the attainment of complete remission. The patients reached the medium (0.5 mg/kg/day at prednisone doses) and low dose (daily 5–10 mg at prednisone doses) at about 4 and 6 months into the tapering protocol, respectively. Patients who did not achieve complete remission of nephrotic proteinuria by 8 weeks or who suffered from early relapse were excluded from the study and subjected to other immunosuppressive therapy. The enrolled patients were not treated with any immunosuppressants other than GCs. Blood samples were collected and laboratory assessments were conducted at each designated time point. Upon admission, demographic information and clinical data of the patients were collected.

Infection cohort study

A comparable cohort of GC‐treated MCD patients was selected, with some patients experiencing acute herpes zoster infection (HZI) and others not (n = 6–9 in each group). The diagnosis of HZI was clinically determined by a dermatologist based on the characteristic symptoms of a unilateral, painful vesicular rash in a confined dermatomal pattern. The study excluded individuals exhibiting incidents of sepsis or chronic infectious diseases such as hepatitis or tuberculosis. Furthermore, based on their medical history, patients who had been prescribed any immunosuppressive drugs apart from glucocorticoids within the preceding 12 months were also disqualified from participation in this study. For all enrolled subjects, the GC regimen and supportive management of glomerular disease adhered to the general principles of the KDIGO 2012 clinical practice guideline, ³⁶ as determined by the treating physicians.

Human CD4⁺ T cell isolation and stimulation experiments

CD4⁺ T cells were isolated from PBMCs using a magnetic bead separation kit (Miltenyi Biotec, Bergisch Gladbach, Germany), as previously described in our studies. ³³ , ³⁵ Briefly, lineage‐specific biotin‐conjugated antibodies and anti‐biotin MicroBeads were used to remove non‐CD4⁺ T cells, and CD25⁺ PE‐labelled cells were magnetically removed from purified CD4⁺ fractions, leaving unlabeled CD4⁺CD25⁻ cells for stimulation experiments. The CD4⁺ T cells were activated for 72 h in RPMI‐1640 medium supplemented with 10% FBS and a T Cell Expansion Kit (Miltenyi Biotec), containing biotinylated antibodies against human CD2, CD3 and CD28. For pathogen stimulation experiments, cells were stimulated with phytohemagglutinin (PHA, 5 μg mL⁻¹, Sigma‐Aldrich, Missouri, USA) for 72 h, followed by a further sorting analysis. To target mTORC1, commercial non‐targeted siRNA, PTEN siRNA and PS6K siRNA were used following mitogenic stimulation according to the manufacturer's instructions (#6568, #6251 and #6566, Cell Signalling, Danvers, USA).

Flow cytometry analysis

Flow cytometry was performed as previously described in our studies. ³³ , ³⁵ A single‐cell suspension was prepared and incubated with directly conjugated antibodies at 4°C for 30 min. The following primary antibodies were used: CD3‐PerCP, CD4‐PE, CD8‐PE/Cy7, CD25‐APC/Cy7, FOXP3‐FITC, IFNG‐Pacificblue, IL10‐Pacificblue, rabbit isotype control IgG (BioLegend, San Diego, USA), Bcl‐xL‐FITC (Invitrogen, Waltham, USA) and Phospho‐S6^Ser235/236‐APC (Cell Signalling). Isotype and fluorescence minus one control were included to set the gates for the antibodies in the multi‐colour immunofluorescent experiments. For intracellular cytokine staining, the single‐cell suspension was subjected to fixation and permeabilization using the Foxp3/Transcription Factor Staining Buffer Set (eBioscience, San Diego, USA) according to the manufacturer's instructions. The FITC Annexin V/PI Apoptosis Detection Kit (Biolegend) was used for cell apoptosis assays. Cell sorting was performed using the BD FACSCanto II or FACSJazz and the acquired data were analysed with FlowJo software (Becton, Dickinson & Company, Franklin Lakes, USA).

Cecal ligation and puncture (CLP) experiments

Male C57BL/6 and Foxp3^EGFP mice (8–10 weeks, strain No. 006772) were acquired from the Jackson Laboratory and underwent CLP surgery as previously described. ³⁷ In summary, mice were anaesthetised with 60 mg kg⁻¹ pentobarbital and underwent a moderate cecal ligation and puncture (CLP) procedure, which was performed via a small longitudinal midline incision made to expose the cecum. The cecum was ligated at a distance of 1 cm from the blind‐ending and then punctured once using a 22G needle in a mesenteric‐to‐antimesenteric direction. A small amount of faeces was extruded from both the mesenteric and antimesenteric penetration sites to ensure patency after the needle was removed. For the control group, an identical procedure was performed but without the steps of ligation and perforation. Animal survival was monitored four times a day for the initial 3 days post‐surgery, and subsequently twice daily for a period of up to 10 days. For the induction of glucocorticoid treatment, mice were intraperitoneally administered with dexamethasone (DXM, 100 μg, twice weekly, Sigma‐Aldrich) for 1 month preceding the CLP or control operation. Tissue and cell samples were harvested 24 h following the surgical procedures. Mouse splenocytes were obtained by passing the homogenised spleens through a cell strainer. Flow cytometry was used to identify and sort T cell subsets, while doublets and dead cells were excluded.

Real‐time PCR assay

Real‐time PCR assays were performed as previously described in our studies. ³³ , ³⁵ After isolation, total RNA was immediately extracted from sorted cells using the RNeasy mini or FFPE kit (Qiagen, Germantown, USA) to minimise the impact of fixation solution on RNA quality. To guarantee samples devoid of genomic DNA contamination, total RNA underwent treatment with DNase (Qiagen, RNase‐Free DNase Set). RNA quantity and quality were ascertained using NanoDrop, ensuring that the ratios of 260/280 and 260/230 were above 2.0 for all samples. Subsequently, cDNA was synthesised employing a Synthesis Kit (Bio‐rad, USA). For each well, 1 μL of cDNA was incorporated and combined with a PCR master mix and pre‐designated primers (see Supplementary table 1). Gene expression, evaluated as fold changes for each target gene, was normalised to glyceraldehyde‐3‐phosphate dehydrogenase (Gapdh) following the 2^−▵▵Ct method. All assays were performed in triplicate. Additionally, a non‐template control was incorporated into the experiment to estimate any potential DNA contamination in the isolated RNA and reagents.

Statistical analysis

Statistical analyses were performed using the SPSS 22 software package (IBM, New York, USA). Data are presented as mean ± standard error of the mean (SEM). To compare differences between groups, appropriate statistical tests, including the two‐tailed Student's t‐test, the Log‐rank (Mantel‐Cox) test, one‐way ANOVA, the chi‐squared test (χ ²  test), or the Wilcoxon rank‐sum test, were performed. A two‐sided P‐value < 0.05 was considered statistically significant.

Author contributions

Huihui Chen: Formal analysis; investigation; methodology; visualization; writing – original draft. Zhiwen Liu: Methodology. Jie Zha: Methodology. Li Zeng: Data curation; methodology. Runyan Tang: Formal analysis. Chengyuan Tang: Methodology. Juan Cai: Methodology. Chongqing Tan: Data curation. Hong Liu: Methodology. Zheng Dong: Methodology. Guochun Chen: Conceptualization; formal analysis; funding acquisition; investigation; supervision; writing – review and editing.

Conflict of interest

The authors declare that no competing interests exist.

Supporting information

Supplementary table 1

Data availability statement

The data and materials analysed in the current study are available from the corresponding author upon reasonable request.