Differential regulation of monocyte exhaustion by distinct TLR agonists and underlying mechanisms

Abstract

Monocyte exhaustion is a dysfunctional immune state marked by persistent inflammation and immune suppression, reflected in STAT1-mediated expression of pathogenic inflammatory mediator CD38 and immune suppressive PD-L1, as well as a suppression of Akt signaling and a reduction of immune-enhancing mediator CD86. While prolonged Toll-like receptor 4 (TLR4) stimulation induces monocyte exhaustion, the roles of other TLRs remain unclear. Here, we systematically evaluated the ability of TLR-2, TLR-3, TLR-7, and TLR-9 agonists to induce murine bone marrow-derived monocytes exhaustion in vitro. Although all tested agonists promoted exhaustion phenotypes to varying degrees, characterized by upregulation of STAT1 mediated expression of CD38 and PD-L1, only TLR-2 and TLR7 agonists drastically suppressed Akt and CD86. In contrast, TLR-3 or TLR-9 agonists preferentially sustained Akt activation and CD86 expression. Mechanistically, STAT1/STAT3 activation mediated by mTROC1 was common across TLR agonists responsible for elevated expression of CD38 and PD-L1. On the other hand, Akt signaling mediated by mTORC2 responsible for the expression of CD86 was preferentially suppressed by TLR-2 and TLR-7 agonists, but retained by TLR-3 and TLR-9 agonists. Deletion of Rictor, a key component of mTORC2, blocked the activation of Akt/CD86 triggered by TLR-3/9 agonists, and further elevated mTORC1 mediated activation of STAT1/3 as well as CD38 expression. Conversely, Fumagillin or Rapamycin treatment, which has been associated with reduced mTORC1 signaling activity, mitigated TLR-2/7-induced STAT1/STAT3 activation and CD38 expression. These findings reveal that monocyte exhaustion is a shared but differentially regulated outcome of distinct TLR pathways, with the mTOR axis potentially serving as a key therapeutic target for immune dysfunction.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12964-026-02955-4.

Introduction

Toll-like receptors (TLRs) are essential sensors of the innate immune challenges, enabling myeloid cells such as monocytes and macrophages to detect and respond to diverse pathogen-associated molecular patterns (PAMPs) [1–4]. These receptors are located either on the cell surface or within intracellular vesicles [5]. To date, at least 10 functional TLRs have been identified in humans and 12 in mice. With varying degrees of expressions across cell types, each TLR recognizes distinct pathogen-associated molecular patterns (PAMPs) derived from bacteria, mycobacteria, viruses, fungi, or parasites [2, 6]. For example, lipoproteins are preferentially detected by TLR-1, TLR-2, and TLR-6; double-stranded RNA (dsRNA) by TLR-3; lipopolysaccharide (LPS) by TLR-4; single-stranded RNA (ssRNA) by TLR-7; and unmethylated DNA by TLR-9. Notably, TLR-9 is poorly or not expressed in human monocytes compared with murine myeloid cells, which contributes to reduced responsiveness of human monocytes to TLR-9 stimulation.

Upon recognizing their respective ligands, TLRs recruit specific adaptor molecules to initiate signaling cascades that trigger transcriptional programs and robust inflammatory responses critical for host defense [3, 4, 7–13]. However, repetitive or excessive TLR stimulation can lead to immune dysfunction, including monocyte exhaustion-a state marked by the paradigm of pathogenic inflammation (manifested by sustained expression of pro-inflammatory mediators such as CD38) and immuno-suppression ( reflected in elevated expression of immune suppressive molecules such as PD-L1, along with downregulation of immune-activating molecules such as CD86) [14–17]. This dysfunctional exhausted state contributes to chronic inflammatory tissue damages and compromised immune defense in conditions such as sepsis, chronic infections, and cancer. Our previous work demonstrated that repetitive high-dose stimulation of TLR-4 with LPS induces monocyte exhaustion via the TRAM adaptor and downstream mTOR signaling [16–18]. However, it remains unclear whether such exhaustion paradigm is a unifying principle for innate immunity when challenged by diverse TLR agonists, and whether differential exhaustion phenotypes may share or utilize distinct molecular mechanisms.

In this study, we systematically evaluated the ability of distinct TLR agonists to induce monocyte exhaustion through comprehensive phenotypic and functional analyses. Although repetitive stimulation with all tested TLR agonists induced monocyte exhaustion to varying extents, each agonist exhibited distinct regulatory signatures, including differences in surface marker expression, cytokine production, and effects on T cell responses. Notably, the signaling pathways driving exhaustion varied substantially among TLRs: TLR-2 and TLR-7 preferentially activated mTORC1, whereas TLR-3 and TLR-9 primarily engaged mTORC2 signaling. Furthermore, we uncovered opposing roles for mTORC1 and mTORC2 in regulating monocyte exhaustion and immune function. Genetic deletion of Rictor, a key component of mTORC2, exacerbated exhaustion under TLR-3 and TLR-9 stimulation. In contrast, pharmacological inhibition of mTORC1 with Fumagillin or Rapamycin attenuated TLR-2 and TLR-7 induced exhaustion by rebalancing mTOR signaling. Our findings expand the framework of innate immune exhaustion, provide novel mechanistic insights for future therapeutic targeting in sepsis and chronic inflammatory diseases.

Materials and methods

In vitro cell culture

Primary cells from female C57BL/6 (WT), LysM-Cre (doble Cre (D-Cre)) and Rictor KO (rictor^fl/fl LysM-Cre) mice with C57BL/6 background were used for in vitro cell culture in this study. C57BL/6 mice were obtained from Jackson Laboratory. D-Cre and myeloid specific Rictor KO mice were provided by Dr. Babak Razani at University of Pittsburgh School of Medicine. All animal procedures complied with the U.S. National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the Virginia Tech Institutional Animal Care and Use Committee (IACUC). Primary bone marrow (BM) cells were isolated from WT, D-Cre, and Rictor KO mice using previously established protocols [19, 20]. The cells were cultured in complete RPMI 1640 medium supplemented with 10% FBS, 1% L-Glutamine, 1% penicillin-streptomycin, and 10 ng/mL M-CSF (PeproTech, Rocky Hill, #351-02). Cultures were maintained at 37 °C in a humidified incubator with 5% CO₂. During the culture period, cells were treated for five days with either the TLR2 ligand Pam3CSK4 (InvivoGen, #tlrl-pms), TLR3 agonist Poly(I: C) (InvivoGen, #tlrl-picw), TLR7 agonist CL264 (InvivoGen, #tlrl-c264e-5), or TLR9 agonist CpG (ODN 1585, InvivoGen, #tlrl-1585). For the Fumagillin/Rapamycin function assay, cells were cultured with 10ng/mL Fumagillin (MCE, #HY-B0751) or 100nM Rapamycin (MCE, # HY-10219), or with a DMSO vehicle control, in the presence of the indicated TLR agonists. Fresh RPMI 1640 medium containing M-CSF and the designated treatments were replenished every two days, as previously described [19, 20]. Additional inhibitors used during the 5-day culture period included the Akt inhibitor MK-2206 (MCE, #HY-108232, 1 µM) and the JAK inhibitor AG490 (MCE, #HY-12000, 5 µM). Under these culture conditions, the bone marrow-derived cells exhibited a loosely adherent, monocyte-like phenotype and lacked expression of the mature macrophage marker CD71 [21].

T cell isolation and coculture

Mouse spleen CD3⁺ T cells were isolated using the MojoSort Mouse CD3 T Cell Isolation Kit (BioLegend, #480031), achieving a purity of over 95%. The purified T cells were labeled with 5 µM carboxyfluorescein succinimidyl ester (CFSE, Invitrogen, #C34554) for 10 min at room temperature (RT), followed by blocking with 3 mL of fetal bovine serum (FBS) for 10 min at 37 °C. After labeling, CFSE-stained T cells were resuspended in culture medium (RPMI 1640 supplemented with 10% FBS, 1% L-Glutamine, 1% penicillin-streptomycin, and 0.055 mM 2-mercaptoethanol) and stimulated with 2.5 µg/mL coated anti-CD3 antibody (Bio X Cell, #BE0001-1) and 5 µg/mL soluble anti-CD28 antibody (Bio X Cell, #BE0015-1) in U-bottom 96-well plates for two days. For coculture experiments, CFSE-labeled T cells (1 × 10⁶ cells/mL) were mixed at a 1:1 ratio with 5-day PBS- or TLR agonist-treated WT monocytes (monocytes were washed at least 3 times with PBS buffer prior to coculture with T cells) in the presence of coated anti-CD3 and soluble anti-CD28 antibodies. The cocultures were maintained in U-bottom 96-well plates (200 µL/well) for two days in a standard cell culture incubator.

NAD + and lactate assay

NAD⁺ levels were determined by Sigma-Aldrich™ NAD+/NADH Assay Kit (Colorimetric, MAK468) and the Lactate Assay Kit was used to determine the lactate levels following manufacturer’s instructions. Cytation³ Image Reader (Bio Tek) was used to quantify NAD⁺ and lactate levels.

Western blot analysis

Western blot analysis was conducted as previously described [20]. In brief, cultured cells were lysed using a 2% lysis buffer supplemented with a protease and phosphatase inhibitor cocktail (Sigma-Aldrich™, #P8340, #P5276, #P0044). Protein concentration in the lysates was determined using the DC Protein Assay Kit (Bio-Rad, #5000112). Proteins were resolved on a 10% acrylamide gel and transferred onto PVDF membranes. The membranes were blocked for 1 h in PBST containing 5% non-fat milk (Bio-Rad, #170–6404) and then incubated overnight at 4 °C with primary antibodies targeting: Phospho-STAT1 (Cell Signaling, #9177), total STAT1 (T-STAT1, Cell Signaling, #9172), Phospho-STAT3 (Cell Signaling, #9145), total STAT3 (T-STAT3, Cell Signaling, #12640), Phospho-Akt (Cell Signaling, #9271), total Akt (T-Akt, Cell Signaling, #9272), Phospho-Src (Cell Signaling, #2105), total-Src (Cell Signaling, #2110), total-Syk (Cell Signaling, #2712), S6K(P70) (Proteintech, #4485-1-AP), Phospho-4E-BP1 ( Cell Signaling, #9644), 4E-BP1 (Cell Signaling, #9452), and β-actin (Cell Signaling, #4970). Following primary antibody incubation, membranes were exposed to HRP-conjugated secondary antibodies, either anti-rabbit IgG (Cell Signaling, #7074) or anti-mouse IgG (Cell Signaling, #7076) for 1 h at room temperature. Protein bands were visualized using an ECL detection kit (VWR, #490005-020), and band intensity was quantified using ImageJ (NIH).

Flow cytometry

Bone marrow (BM) cells along with T cells from in vitro cultures, were analyzed using flow cytometry. Cells were washed with PBS and preincubated with FcBlock (anti-mouse CD16/32, BD Biosciences, #553141) to prevent non-specific binding. Cells were then stained for 30 min with fluorochrome-conjugated antibodies targeting mouse CD11b (APC-Cy7, BioLegend, #101226), Ly6C (PE-Cy7, BioLegend, #128018), PD-L1 (PE, BioLegend, #124308), CD86 (FITC, BioLegend, #105006), CD38 (APC, BioLegend, #102712), CD3 (APC, BioLegend, #100236), CD4 (PE, BioLegend, #116006), CD8 (Alexa Fluor 488, BioLegend, #100723), and CD69 (APC-Cy7, BioLegend, #104526). After staining, cells were washed and resuspended in FACS buffer containing Propidium Iodide (Invitrogen, #P3566) to assess viability. Samples were acquired on a FACS Canto II (BD Biosciences), and data analysis was performed using FlowJo (Ashland, OR).

Quantitative real-time PCR

Total RNA was extracted from monocytes cultured in vitro for five days and treated with various TLR agonists. Briefly, total RNA was isolated using the RNeasy Plus Mini Kit (Qiagen, #74134), and cDNA was synthesized with the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, #4368813). mRNA levels were quantified by real-time polymerase chain reaction (RT-PCR). The primers used for RT-PCR are listed below: IL-6 forward, GGAGCCCACCAAGAACGATA. IL-6 reverse, CCAGCATCAGTCCCAAGAA. IL-10 forward, AGCTGAAGACCCTCAGGATG. IL-10 reverse, CACTCTTCACCTGCTCCACT. TGF-β forward, CAACCCAGGTCCTTCCTAAA. TGF-β reverse, GGAGAGCCCTGGATACCAAC. TNF-α forward, CAGCCTCTTCTCATTCCTGCTTGTG. TNF-α reverse, CTGGAAGACTCCTCCCAGGTATAT. IL-1β forward, ATGAAGGGCTGCTTCCAAAC. IL-1β reverse, TCTCCACAGCCACAATGAGT.

Statistical analysis

Graphs and statistical analyses were performed using GraphPad Prism v8.0. Comparisons between two groups were assessed with Student’s t-test, while one-way ANOVA adjusted for Tukey’s multiple comparisons (for multiple groups with a single control) or two-way ANOVA (for multiple groups with multiple variables) was applied where appropriate. Normality was assessed using the Shapiro-Wilk test, and homogeneity of variance was evaluated using Levene’s test. Statistical significance was denoted as follows: *p < 0.05; **p < 0.01; ***p < 0.001.

Results

Differential exhaustion of monocytes by distinct TLR agonists

To investigate how different TLR pathways influence monocyte exhaustion, bone marrow-derived monocytes (BMDMs) from wild-type (WT) mice were repetitively stimulated with TLR-2, TLR-3, TLR-7, or TLR-9 agonists for 5 days (Figure S1A-D). After stimulation, the proportion of Ly6C⁺CD11b⁺ monocytes-an exhausted monocyte population-along with the key exhaustion markers CD38, CD86, and PD-L1 were analyzed by flow cytometry (Fig. 1A-D). These salient markers representing the key exhaustion paradigm of pathogenic inflammation (elevated CD38) and immune-suppression (elevated PD-L1, reduction of CD86), were more closely associated with innate functional exhaustion as comparted to secreted cytokines, and were independently documented both in human and murine systems [16, 18, 22, 23]. We observed that the Ly6C⁺CD11b⁺ population increased after stimulation with each TLR agonist, suggesting an overall shared exhaustion profile among all tested TLR agonists. Furthermore, the pathogenic inflammation marker CD38 and the immune suppression marker PD-L1 were universally enhanced following stimulation with all TLR agonists, further suggesting shared features of monocyte exhaustion. Intriguingly, the immune-enhancing marker CD86 showed differential regulation: it decreased only upon TLR-2 and TLR-7 stimulation but increased following TLR-3 and TLR-9 stimulation (Fig. 1A-D). Our data reveal that only TLR-2 and TLR-7 agonists can drive full-blown exhaustion of murine monocytes. In contrast, monocytes repetitively challenged with TLR-3 or TLR-9 agonists still maintain robust expression of immune-enhancing CD86, reflecting limited exhaustion.

Fig. 1.

Differential exhaustion of monocytes by distinct TLR agonists. Bone marrow-derived monocytes (BMDMs) from wild-type (WT) mice were stimulated for 5 days with individual TLR agonists: Pam3CSK4 (TLR-2, 100 nM, A), Poly(I: C) (TLR-3, 500 ng/mL, B), CL264 (TLR-7, 500 nM, C), or CpG (TLR-9, 100 nM, D). The Ly6C⁻CD11b⁺ and Ly6C⁺CD11b⁺ monocyte subsets were analyzed and the mean fluorescence intensity (mFI) of CD38, CD86, and PD-L1, which were quantified and plotted separately. n = 3. Error bars represent mean ± SD. Statistical significance: *p < 0.05, **p < 0.01, **p < 0.001

Given the observed upregulation of CD38-a NAD⁺-hydrolyzing ectoenzyme-in exhausted monocytes stimulated by all TLR agonists tested (Fig. 1A-D) [24–26], we next measured intracellular NAD⁺ and lactate levels. NAD⁺ levels were reduced in monocytes treated with TLR-2, TLR-3, TLR-7, or TLR-9 agonists (Figure S2A). Notably, however, lactate production was significantly increased only in monocytes stimulated with TLR-2 and TLR-7 agonists, but not with TLR-3 or TLR-9 agonists (Figure S2B). Proliferation assays further demonstrated that stimulation through TLR-2, TLR-3, TLR-7, or TLR-9 enhanced monocyte proliferation compared to unstimulated controls (Figure S2C). Next, we examined the expression profiles of selected pro- and anti-inflammatory cytokine genes following stimulation. Pro-inflammatory cytokines IL-6 and TNF-α were upregulated by all TLR agonists, with TLR-9 inducing particularly strong IL-6 expression. All TLR agonists also elevated the anti-inflammatory cytokine IL-10, whereas only TLR-3 stimulation increased TGF-β expression. Interestingly, IL-1β expression was increased by TLR-2 and TLR-3 stimulation but decreased by TLR-7 and TLR-9 stimulation (Figure S2D). Together, these results demonstrate that distinct TLR agonists differentially regulate monocyte exhaustion, proliferation, metabolic reprogramming, and cytokine production, with TLR-3 and TLR-9 agonists-trained monocytes retaining the expression of immune-enhancing CD86.

Differential modulation of T cell responses by distinct TLR agonists

One of the key functional implications of monocyte exhaustion is their broad spectrum suppression of T cells, contributing to compromised overall immune defense against polymicrobial infections during sepsis [27, 28]. To evaluate the impact of exhausted monocytes on broadly suppressing T cell activation, splenic T cells were co-cultured with bone marrow-derived monocytes that had been treated for 5 days with different TLR agonists (Fig. 2A). Consistent with clinical observations of immunosuppressive septic monocytes, prolonged exposure to TLR-2 and TLR-7 agonists markedly suppressed the expression of the activation marker CD69 on both CD4⁺ and CD8⁺ T cells, while T cell proliferation was not affected (Fig. 2A-C). Interestingly, monocytes repetitively challenged by TLR-3 or TLR-9 stimulation reduced CD4⁺ and CD8⁺ T cell proliferation but, paradoxically, increased CD69 expression on T cells (Fig. 2A-C). These findings differ from our previously published results with high-dose TLR-4 agonist (LPS), where prolonged LPS exposure strongly suppressed both T cell proliferation and activation [17]. Our data reveal differential exhaustion of monocytes by distinct TLR agonists, with TLR-2, 4, and 7 agonist inducing full-spectrum of exhaustion. In contrast, monocytes repetitively challenged with TLR-3 and 9 agonists still maintain a robust immune enhancing phenotype.

Fig. 2.

Differential modulation of T cell responses by distinct TLR agonists. Splenic T cells were co-cultured for two days with monocytes pre-stimulated with PBS or various TLR agonists. Recipient T cells, pre-labeled with CFSE, were stained for CD4, CD8, and CD69, and analyzed by flow cytometry on day 2. CFSE dilution histograms of CD4⁺ and CD8⁺ T cells, along with the quantification of their proliferation, are presented in (A) and (B), respectively. C CD69 expression levels in CD4⁺ and CD8⁺ T cells after two days of co-culture were quantified and plotted. Data are representative of at least three independent experiments. Error bars represent mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001

The differential exhaustion phenotypes closely correlate with our defined cellular surface markers of monocytes. For example, CD86 expression was differentially modulated and was particularly elevated following TLR-3/9 stimulation (Fig. 1B and D), which may explain the pronounced increase in CD69 expression on CD4⁺ and CD8⁺ T cells. CD38 and PD-L1 were consistently elevated across all conditions, with PD-L1 being most strongly upregulated after TLR-3 and TLR-9 stimulation (Fig. 1B and D), potentially accounting for the marked suppression of T cell proliferation observed in these two co-culture groups.

STAT1/STAT3 and mTOR signaling pathways regulate monocyte exhaustion induced by distinct TLR agonists

To investigate the signaling mechanisms underlying monocyte exhaustion triggered by various TLR agonists, we analyzed key exhaustion-associated signaling molecules previously implicated in high-dose LPS (TLR-4 agonist)-induced monocyte exhaustion [17, 20]. BMDMs were stimulated for five days with TLR-2, TLR-3, TLR-7, or TLR-9 agonists, and the activation status of STAT1, STAT3, Src, Syk, and components of the mTOR pathway were assessed by Western blotting. Phosphorylated and total STAT1 levels were elevated across all TLR stimulation conditions, indicating a consistent activation of the STAT1 signaling axis during monocyte exhaustion (Fig. 3A). Similarly, STAT3 activation, as evidenced by increased phosphorylated and total STAT3, was observed in all groups, suggesting a broad and conserved role of STAT3 signaling in TLR-induced exhaustion (Fig. 3B).

Fig. 3.

STAT1/STAT3 regulate monocyte exhaustion induced by distinct TLR agonists. BMDMs from WT mice were stimulated for 5 days with individual TLR agonists: TLR-2, TLR-3, TLR-7, or TLR-9. A Western blot analysis was performed to assess phosphorylated STAT1 (p-STAT1) and total STAT1 (T-STAT1) protein levels. Densitometric quantification was normalized to β-actin. B Phosphorylated STAT3 (p-STAT3) and total STAT3 (T-STAT3) levels were evaluated by western blot and normalized to β-actin. Data are representative of at least three independent experiments, and error bars represent mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001

We next assessed upstream kinases involved in the regulation of STAT and mTOR pathways. Phosphorylated Src (p-Src) was markedly upregulated in TLR-2- and TLR-7-stimulated cells, but only modestly increased following TLR-3 and TLR-9 stimulation. Interestingly, total Src expression increased only in TLR-3-stimulated cells and decreased under the other TLR conditions (Fig. 4A). Additionally, total Syk protein levels were moderately increased in TLR-2 and TLR-7 treated monocytes, showed minimal change in the TLR-9 group, and were decreased in the TLR-3 group (Fig. 4B). Examination of the mTORC1 downstream target S6K revealed strong upregulation in the TLR-2 and TLR-7 groups, while S6K levels were substantially diminished following TLR-3 and TLR-9 stimulation (Fig. 4B), suggesting differential regulation of mTORC1 signaling.

Fig. 4.

Src-Syk and mTOR-Akt signaling pathways contribute to monocyte exhaustion induced by distinct TLR agonists. BMDMs from WT mice were stimulated for 5 days with individual TLR agonists: TLR-2, TLR-3, TLR-7, or TLR-9. A Protein levels of phosphorylated Src (p-Src) and total Src (T-Src) were assessed via western blot, with normalization to β-actin. B Total Syk (T-Syk) and S6K protein levels were measured by western blot and normalized to β-actin. C Western blot analysis was performed to evaluate phosphorylated Akt (p-Akt) and total Akt (T-Akt) protein levels. Band intensities were quantified and normalized to β-actin. D Schematic illustration summarizing the potential mechanisms by which distinct TLR agonists regulate monocyte exhaustion. Data are representative of at least three independent experiments, and error bars represent mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001

To provide causal evidence for the involvement of STAT1/3 in driving the pathogenic inflammatory aspect of monocyte exhaustion initiated by TLR-2 and TLR-7 agonists, we applied selective pharmacological inhibitor for STAT1/3 signaling. As shown in Figure S3, pharmacologic inhibition of STAT signaling significantly reversed the exhausted phenotype mediated by TLR-2 and TLR-7 agonists, as reflected by decreased CD38 and PD-L1 expression and restored CD86 expression (Figure S3). These results provide causal evidence that STAT activation is a critical driver of TLR-induced monocyte exhaustion.

To further dissect the involvement of mTOR signaling, we analyzed phosphorylation states of downstream effectors. TLR-2 and TLR-7 stimulation resulted in reduced levels of phosphorylated Akt, whereas TLR-3 and TLR-9 significantly increased Akt phosphorylation (Fig. 4C). Since mTORC2 regulates Akt phosphorylation [29, 30], these findings suggest divergent activation of the mTORC2-Akt axis depending on the specific TLR agonist. To further evaluate whether sustained CD86 expression induced by TLR-3 or TLR-9 stimulation is mediated through the mTORC2-Akt axis, we pharmacologically inhibited Akt activation during TLR-3 stimulation. Akt inhibition significantly enhanced the exhausted marker CD38 expression and markedly decreased CD86 expression following TLR-3 treatment (Figure S4). These findings support a functional role for Akt activation downstream of mTORC2 in regulating sustained CD86 expression, although additional studies would be required to fully establish direct mechanistic causality.

Overall, these results reveal that distinct TLR agonists induce monocyte exhaustion via both shared and divergent signaling mechanisms. While STAT1/STAT3 activation appears to be a common hallmark, the balance of mTORC1 versus mTORC2 signaling differs markedly between TLR pathways. Specifically, TLR-2 and TLR-7 predominantly activate mTORC1 and suppress mTORC2, whereas TLR-3 and TLR-9 preferentially enhance mTORC2 signaling (Fig. 4D). These distinct pathway preferences may underline functional differences in monocyte exhaustion phenotypes elicited by different TLR stimuli.

Rictor deficiency exacerbates TLR-3 and TLR-9 induced monocyte exhaustion

Based on our earlier finding that TLR-3 and TLR-9 preferentially activate mTORC2 signaling during monocyte exhaustion, we next examined whether the mTORC2 pathway is required for this process. To this end, we used monocytes lacking Rictor, a key component of mTORC2 [29, 31, 32], and cultured them for five days in the presence of TLR-3 or TLR-9 agonists. In D-Cre control monocytes, prolonged stimulation with TLR-3 or TLR-9 agonists induced an exhausted phenotype, characterized by increased expression of CD38, along with an immune-enhancing phenotype reflected by elevated CD86 expression (as shown in Fig. 1B and D). In contrast, Rictor-deficient monocytes displayed a more severe exhausted phenotype: CD38 expression was further increased, whereas CD86 expression was markedly reduced (Fig. 5A-B). These results suggest that Rictor, and by extension mTORC2, is required to maintain the balance between exhaustion and residual immune activation during TLR-3 and TLR-9 driven monocyte stimulation. Mechanistically, the loss of Rictor disrupts the balance between mTORC1 and mTORC2, leading to hyperactivation of mTORC1 signaling. Consistent with this, Rictor-deficient monocytes exhibited a substantial reduction in phosphorylated Akt levels, even after TLR-3 or TLR-9 stimulation (Fig. 5C-D), accompanied by increased mTORC1 downstream effector S6K and the phosphorylation of 4E-BP1 as compared to D-Cre controls (Fig. 5E-H).

Fig. 5.

Rictor deficiency exacerbates TLR-3 and TLR-9 induced monocyte exhaustion. BMDMs from WT or Rictor knockout (KO) mice were stimulated for 5 days with either TLR-3 or TLR-9 agonists. A The mFI of CD38 and CD86 were measured in TLR-3 agonist stimulated WT and Rictor KO monocytes and presented separately. B mFI of CD38 and CD86 in TLR-9 agonist stimulated WT and Rictor KO monocytes were similarly quantified and plotted. C-D Levels of phosphorylated Akt (p-Akt) and total Akt (T-Akt) were evaluated by western blot analysis, with β-actin serving as a loading control. E-F S6K protein levels were assessed by western blot and normalized to β-actin. G-H phosphorylated 4E-BP1 (p-4E-BP1) and total 4E-BP1 (T-4E-BP1) protein levels were assessed by western blot and normalized to β-actin. Data are representative of at least three independent experiments, and error bars represent mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001

In contrast, Rictor deficiency had no effect on phosphorylated or total STAT1 and STAT3 expression following TLR-3 or TLR-9 stimulation (Figure S5), indicating that STAT1/STAT3 signaling is not involved in Rictor-mediated regulation of monocyte exhaustion. Collectively, these data demonstrate that mTORC2/Rictor plays a key role in restraining excessive mTORC1 activity and preserving a balance between exhaustion and residual immune activation in TLR-3 and TLR-9 stimulated monocytes.

Fumagillin alleviates TLR-2 and TLR-7 induced monocyte exhaustion

Fumagillin is a secondary metabolite produced by Aspergillus species and is a natural antibiotic that specifically inhibits methionine aminopeptidase 2 (MetAP2), a key regulator of protein synthesis and cellular metabolism [33–35]. Recent studies uncovered that inhibition of MetAP2 can indirectly influence mTOR signaling by dampening excessive mTORC1 activity [36–38]. Since our previous results indicated that hyperactivation of mTORC1 and disruption of mTORC1/mTORC2 balance drive monocyte exhaustion-and that TLR-2 and TLR-7 stimulated monocytes exhibit an exhaustion phenotype similar to that induced by high-dose LPS via TLR-4, we next examined whether Fumagillin treatment could mitigate this phenotype. BMDMs from WT mice were treated with 10 ng/mL Fumagillin or DMSO control and then stimulated with TLR-2 or TLR-7 agonists for five days. Flow cytometry analysis revealed that Fumagillin treatment markedly reduced the expression of the exhaustion marker CD38 and restored CD86 expression in both TLR-2 and TLR-7-stimulated monocytes (Fig. 6A-B). Furthermore, the proportion of exhausted Ly6C⁺CD11b⁺ monocytes was also decreased after Fumagillin treatment (Figure S6A-B).

Fig. 6.

Fumagillin alleviates TLR-2 and TLR-7 induced monocyte exhaustion. BMDMs from WT mice were treated with 10 ng/mL Fumagillin or DMSO (vehicle control), followed by stimulation with TLR-2 or TLR-7 agonists for 5 days. A The mFI of CD38, CD86 were measured in TLR-2 agonist stimulated monocytes treated with either Fumagillin or DMSO and presented separately. B mFI of CD38 and CD86 were similarly quantified in TLR-7 agonist stimulated monocytes with or without Fumagillin treatment. C-D Levels of phosphorylated STAT1 (p-STAT1) and total STAT1 (T-STAT1) in TLR-2 agonist stimulated group were evaluated by western blot analysis, with β-actin serving as a loading control. E-F Levels of phosphorylated STAT1 (p-STAT1) and total STAT1 (T-STAT1) in TLR-7 agonist stimulated group were evaluated by western blot analysis, with β-actin serving as a loading control. Levels of phosphorylated STAT3 (p-STAT3) and total STAT3 (T-STAT3) in TLR-2 agonist stimulated group (G-H) or TLR-7 agonist group (I-J) were evaluated by western blot analysis, with β-actin serving as a loading control. Data are representative of at least three independent experiments, and error bars represent mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001

To further validate that these effects were causally mediated through mTORC1 inhibition, we performed parallel experiments using Rapamycin, a well-established and specific mTORC1 inhibitor. Consistent with the effects of Fumagillin, Rapamycin treatment similarly attenuated exhaustion-associated marker expression (Figure S7A-B), supporting a central role for mTORC1 in TLR-2/7 induced monocyte exhaustion. To explore the underlying mechanisms, we analyzed intracellular signaling pathways by western blot. In both TLR-2 and TLR-7 stimulated groups, Fumagillin treatment significantly reduced the phosphorylation and total levels of STAT1 (Fig. 6C-F) as well as the total level of STAT3 (Fig. 6G-J), correlating with the observed decrease in CD38 expression. Notably, Rapamycin treatment produced comparable effects on STAT1/STAT3 signaling, further supporting that modulation of these pathways occurs downstream of mTORC1 inhibition (Figure S7C). Furthermore, Fumagillin substantially decreased phosphorylation of Src, while increasing total Src levels (Fig. 7A-D). The elevation of Syk induced by TLR-2 and TLR-7 agonists was also reversed by Fumagillin co-treatment (Fig. 7E-F). Similar regulatory effects were observed with Rapamycin treatment (Figure S7D). Consistent with its role in modulating mTORC1 activity, both Fumagillin and Rapamycin markedly reduced levels of S6K, a downstream effector of mTORC1, in both TLR-2 and TLR-7 stimulated monocytes (Fig. 7G-H, Figure S7D).

Fig. 7.

Fumagillin alleviates TLR-2 and TLR-7 induced monocyte exhaustion through mTORC1 signaling pathway. BMDMs from WT mice were treated with 10 ng/mL Fumagillin or DMSO (vehicle control), followed by stimulation with TLR-2 or TLR-7 agonists for 5 days. A-B Levels of phosphorylated Src (p-Src) and total Src (T-Src) in TLR-2 agonist stimulated group were evaluated by western blot analysis, with β-actin serving as a loading control. C-D Levels of phosphorylated Src (p-Src) and total Src (T-Src) in TLR-7 agonist stimulated group were evaluated by western blot analysis, with β-actin serving as a loading control. Levels of Syk in TLR-2 agonist stimulated group (E) or TLR-7 agonist group (F) were evaluated by western blot analysis, with β-actin serving as a loading control. Levels of S6K in TLR-2 agonist stimulated group (G) or TLR-7 agonist group (H) were evaluated by western blot analysis, with β-actin serving as a loading control. Data are representative of at least three independent experiments, and error bars represent mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001

Collectively, these findings indicate that pharmacological inhibition of mTORC1, using either Fumagillin or Rapamycin, alleviates TLR-2- and TLR-7-induced monocyte exhaustion by rebalancing mTORC1 signaling, reducing S6K activation, and limiting CD38 upregulation, while also modulating STAT1/STAT3 and Src signaling.

Discussion

Monocyte exhaustion memory state is characterized by the dichotomy of persistent pathogenic inflammation and immune suppression, following prolonged or repetitive challenges with strong systemic inflammatory or infectious signals [16–18, 20, 39, 40]. It is important to distinguish monocyte exhaustion from other forms of innate immune adaptation such as endotoxin tolerance or macrophage polarization. Endotoxin tolerance typically describes a transient hyporesponsive state following acute TLR stimulation, manifested in reduced expression of pro-inflammatory mediators [41, 42]. On the hand, macrophage polarization (e.g., M1/M2) reflects differentiation programs driven by distinct cytokine environments, in which specialized cytokine such as IFNγ can drive a pro-inflammatory state while IL-4 drives an anti-inflammatory state [43]. While tolerance or classical polarization reflects either a transient state of anti-inflammatory adaptation or polarization by one-time challenge of distinct cytokines, monocyte exhaustion can only be achieved by repetitive and prolonged challenges with inflammatory signals [44]. Phenotypically, monocyte exhaustion phenotype is also distinct from the simple anti-inflammatory tolerance, or macrophage polarization (either M1 inflammatory or M2 anti-inflammatory). Instead, monocyte exhaustion is cardinally manifested in the dichotomy of reduced differentiation, immune suppression and pathogenic inflammation [16, 22]. The simultaneous presence of immune suppression and pathogenic inflammation present a double hit for the host, with elevated tissue damage and reduced ability to combat infections, collectively contributing to disease progression in settings like chronic infections, sepsis, and cancer [16–18, 23, 45]. Key cell surface features of monocyte exhaustion include an increased expression of key inhibitory and inflammatory markers such as PD-L1 and CD38, alongside reduced immune-enhancing markers such as MHCII molecules and CD86 [16, 22], observed in both murine and human systems.

Although immune exhaustion following repetitive and prolonged challenges has been well defined in the context of adaptive immune T cells [46, 47], particularly in the context of tumor immunology, innate exhaustion and underlying mechanisms are less studied nor understood. Past studies regarding innate immune memory dynamics largely focused on the transient “two-hit” model of training and tolerance [48, 49], which largely ignored the effects of repetitive and prolonged effects of exhaustion memory known in the adaptive immune field. Our current study fills this critical gap in knowledge and addresses the commonality of monocyte exhaustion following wide spectrum of repetitive challenges. To this regard, we tested the exhaustion profiles of monocytes following repetitive stimulations with various TLR agonists.

Previous studies have shown that individual TLRs activate distinct signaling cascades, resulting in varied immune outcomes in monocytes [50]. TLR-2 promotes inflammatory cytokine production and is implicated in chronic inflammatory disorders [51–53]. TLR-3 drives type I interferon responses, particularly during viral infections, but can also lead to excessive inflammation and cell death [54, 55]. TLR-7 reprograms monocytes toward a pro-inflammatory phenotype by modulating calcium signaling and enhancing responsiveness to secondary stimulation [56–58]. TLR-9, through recognition of unmethylated CpG DNA, triggers antiviral cytokine production [59–61]. Despite these well-characterized functions, it remains unclear how prolonged activation of different TLRs affects monocyte function over time-particularly in relation to the induction of monocyte exhaustion.

Our data reveal that monocyte exhaustion can be similarly generated following the repetitive challenges of diverse TLR stimulants. However, there are some unique differences in terms of the extent of monocyte exhaustion as well as underlying mechanisms. All tested TLR agonists can similarly induce the pathogenic inflammatory marker CD38 as well as the immune suppression marker of PD-L1, suggesting a conserved pattern of monocyte exhaustion. Consequently, we also observed a drastic reduction of intracellular NAD⁺, an indicator of metabolic rewiring and exhaustion. However, only the TLR-2 or TLR-7 stimulation led to a significant decrease of the immune enhancing marker CD86 and a rise in lactate levels.

In contrast, TLR-3 or TLR-9 stimulation largely sustained CD86 expression, did not enhance lactate production, and even promoted T cell activation, indicating a partial exhaustion phenotype with retained immune enhancing functionality. Our data extend our previous report showing differential tolerance effects of TLR agonists [62], and consistent with a recent independent study suggesting the differential effects of TLR agonists in modulating monocyte programming dynamics [63]. Our data complement previous studies regarding the translational applications of various TLR agonists in clinical development. With particular relevance, TLR agonists have been explored for serving as potential immune activators with varied efficacies for the development of innate adjuvants in broadly enhancing host adaptive immunity [64]. However, limitations remain significant: these agonists can trigger excessive inflammation or systemic toxicity, exhibit dose- and context-dependent variability, and may induce immune tolerance or exhaustion with repeated use. Our mechanistic studies revealing the exhaustion-inducing effects of TLR-4, -2, and -7 agonists further provide mechanistic explanations for these noted limitations. Intriguingly, TLR-9 agonists are considered more effective adjuvants given their potent immune-enhancing effects without immune suppressive effects [65]. Our data regarding the sustained immune-enhancing effects selectively induced by TLR-9 provide mechanistic insights and complements the translational adjuvant studies. Further dosing and time-course studies with TLR-9 agonist may yield more effective conditions for developing effective and safe adjuvant with optimum immune-enhancing efficacy while avoiding excessive inflammatory toxicity and adverse effects.

In terms of underlying molecular mechanisms, we noticed a universal activation of Src and STAT1 signaling in exhausted monocytes sustained by all TLR agonists, independently shown to be important for the elevated expression of CD38 and PD-L1 [40, 66–68]. Emerging studies reveal that mTORC1 signaling may play an important role in immune cell activation [69–73]. However, whether prolonged TLR stimulation in monocytes similarly involves mTOR signaling and whether this contributes to exhaustion has not been previously explored.

Our data suggests that mTOR signaling may play complex roles in monocyte exhaustion sustained by all tested TLR agonists. Intriguing, recent studies suggest that mTORC1 and mTORC2 may be mutually competitive, and differentially involved in the activation of immune cells [74–76]. Consistent with these observations, we found that all tested TLR agonists induced a drastic activation of mTORC1, known to be responsible for the inflammatory activation of Src/Syk pathway and correlated with the induction of CD38 as well as PD-L1 [77–79]. In terms of mTORC2-Akt path, only TLR-2 and TLR-7 agonists led to a significant reduction of mTORC2-Akt. In contrast, TLR-3 and TLR-9 agonists preferentially sustained the activation of mTORC-Akt, correlated with the sustained activation of CD86, as well as sustained T cell-stimulatory capacity. These observations were reinforced by both genetic and pharmacologic interventions. We observed that the disruption of mTORC2 via Rictor deficiency exacerbated exhaustion in TLR-3/9 stimulated monocytes by allowing exacerbated mTORC1 activity. Conversely, treatment with Fumagillin, a MetAP2 inhibitor that may rebalance mTORC1/mTORC2 signaling, reversed the exhaustion phenotype in TLR-2/7 stimulated monocytes. This reversal was associated with decreased STAT1/3 and Src/Syk activation and suppression of the mTORC1 downstream effector S6K.

Together, our initial attempt reveals intriguing conservations as well as distinct differences with regard to monocyte exhaustion induced by diverse TLR agonists, with shared as well as unique intra-cellular signaling processes. Our study laid a foundation for future in vivo and ex vivo studies of monocyte dynamics in disease settings to further define pathophysiological relevance as well as underlying mechanisms of various exhaustion phenotypes. Although the present study was conducted primarily using murine monocytes, many of the signaling pathways identified here, including TLR-mediated activation and mTORC1/mTORC2 signaling, are highly conserved in human monocytes. These findings therefore suggest that similar mechanisms may contribute to the development of monocyte exhaustion in human inflammatory and infectious diseases. Nevertheless, species specific differences in monocyte subsets and immune regulation may influence the extent and functional consequences of exhaustion. Future studies using human primary monocytes and clinical samples will be needed to determine the translational relevance of these mechanisms in human disease settings.

In physiological settings such as microbial infection, immune cells are often exposed to multiple pathogen-associated molecular patterns simultaneously, leading to the concurrent activation of several TLR pathways. Although the present study focuses on the effects of individual TLR agonists in driving distinct monocyte exhaustion phenotypes, it is likely that combinatorial TLR signaling further exacerbates these responses. The interaction between multiple TLR pathways may alter the balance of downstream signaling networks, including mTORC1 and mTORC2 activation, thereby influencing the magnitude or characteristics of monocyte exhaustion. A systematic analysis of combined TLR stimulation will therefore be an important direction for future studies to better understand how complex pathogen signals shape monocyte functional states in vivo. In addition, classical models of immune exhaustion often involve a progressive decline in cellular responsiveness following repeated stimulation. Although the present study focused on defining the phenotypic and functional states induced by prolonged stimulation with distinct TLR agonists, the temporal dynamics of cytokine production were not systematically examined. Future studies analyzing cytokine production kinetics and responses to secondary stimulation will be important to further clarify the functional progression and stability of TLR-induced monocyte exhaustion.

Authors’ contributions

L.L conceived the study, wrote the paper, and supervised the study. J.W performed the study, analyzed the data, and wrote the paper. Y.W, G.L and B.R performed the study and analyzed the data. All authors agreed to the final version of the manuscript.

Funding

This work is supported by NIH grant AI172133 to LL, and the American Heart Association Postdoctoral Fellowship Award to JW (23POST1027469, 10.58275/AHA.23POST1027469.pc.gr.161338).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Competing interests

The authors declare no competing interests.