Targeting SHP2 to reverse immune evasion and resistance to anti-PD-1 therapy in non-small cell lung cancer.

Introduction
Non-small cell lung cancer (NSCLC) remains a paramount global health concern, characterized by high incidence and its status as the leading cause of cancer-related mortality worldwide [1, 2]. The therapeutic landscape for NSCLC has been significantly advanced by the advent of immune checkpoint inhibitors (ICIs), particularly antibodies targeting the PD-1/PD-L1 axis, which restore T-cell mediated anti-tumor immunity and yield durable responses in a subset of patients [2, 3]. Despite this progress, a substantial challenge persists: many patients exhibit primary or acquired resistance, failing to respond initially, while others develop acquired resistance over time, curtailing the overall clinical benefit of ICI therapy [4, 5]. This underscores an urgent need to decipher the underlying mechanisms of resistance to enhance therapeutic strategies.
Emerging evidence consistently points to the critical role of the tumor microenvironment (TME) composition, particularly the presence and functional state of CD8 + T lymphocytes, as a key determinant of ICI efficacy [6, 7]. These effector T cells mediate tumor control largely through the secretion of cytokines, with interferon-gamma (IFN-γ) serving as a central orchestrator of anti-tumor immunity [8]. IFN-γ exerts pleiotropic effects, including enhancing antigen presentation, inducing tumor cell apoptosis, and inhibiting angiogenesis [9–11]. However, the biological impact of IFN-γ within the TME is complex; prolonged or high levels can paradoxically promote immune evasion by upregulating inhibitory ligands like PD-L1 and fostering an immunosuppressive milieu, potentially contributing to ICI resistance [12–14]. This paradoxical function of IFN-γ highlights a critical knowledge gap regarding the tumor-intrinsic mechanisms that govern its signaling pathway, and how their disruption contributes to immune evasion and therapeutic resistance to checkpoint blockade.
Intracellular signaling molecules within tumor cells are poised to modulate such responses. The ubiquitously expressed protein tyrosine phosphatase SHP2 (PTPN11) is a well-established oncogenic signaling node, primarily known for positively regulating receptor tyrosine kinase (RTK) signaling via the RAS-MAPK pathway, thus promoting tumor growth and survival [15–19]. Additionally, SHP2 functions within immune cells, acting as a negative regulator of T-cell activation, partly through interactions with inhibitory receptors like PD-1 [19–25], Our previous research underscored SHP2's significance in NSCLC, demonstrating that its upregulation via an exosomal circUSP7/miR-934 mechanism contributes to CD8 + T cell dysfunction and promotes resistance to anti-PD1 immunotherapy [26], establishing SHP2 as a key mediator of immune escape. Despite this understanding of an upstream regulatory axis, the precise downstream mechanisms by which tumor cell-intrinsic SHP2 directly subverts anti-tumor immunity remained largely unexplored. Specifically, how it might interfere with crucial pro-inflammatory signaling like the IFN-γ pathway remained largely unexplored. Elucidating this direct interplay is crucial for revealing novel mechanisms of immune subversion and forms a central aim of the present study, which further investigates the consequences for CCL5-mediated T cell recruitment and sensitivity to anti-PD-1 therapy.
Furthermore, effective IFN-γ signaling is crucial not just for direct tumor cell effects but also for shaping the immune landscape through chemokine induction [27, 28]. IFN-γ is a known inducer of CCL5 (RANTES), a chemokine vital for recruiting effector CD8 + T cells into the tumor [29]. Indeed, elevated CCL5 levels correlate with increased T cell infiltration and improved ICI outcomes [30, 31]. Consequently, a specific, mechanistically linked question arises: If tumor-intrinsic SHP2 does indeed negatively regulate IFN-γ signaling, does this suppression extend downstream to impair the production of essential T-cell chemoattractant like CCL5? Determining this link would provide a tangible mechanism for SHP2's influence on the TME and ICI response.
Therefore, this study was specifically designed to investigate whether and how tumor cell-intrinsic SHP2 modulates the IFN-γ signaling pathway in NSCLC, assess the downstream consequences on the key chemoattractant CCL5 and CD8 + T cell function, and evaluate its impact on sensitivity to anti-PD-1 therapy. We hypothesized that SHP2 negatively regulates IFN-γ signaling within tumor cells, thereby impairing anti-tumor immunity via downstream effectors like CCL5. Our findings, demonstrating that SHP2 suppresses the IFN-γ/STAT1/IRF1/CCL5 axis and that its inhibition enhances anti-PD-1 efficacy, directly address these unresolved questions and elucidate a novel mechanism of immune evasion in NSCLC.

Results

The expression of SHP2 is Upregulated in NSCLC patients and promotes progression
To identify novel regulators of immune evasion in NSCLC, we first performed a genome-wide CRISPR knockout screen in A549 cells. Following co-culture with activated CD8 + T cells under immune pressure (Fig. 1A), deep sequencing of sgRNAs in surviving cells identified SHP2 (PTPN11) as a significantly enriched gene, indicating its role in promoting A549 cell survival against T cell-mediated killing (Fig. 1B-C).
Next, we investigated SHP2 protein expression and its clinical relevance using tissue microarrays from 110 NSCLC patients, including paired tumor and adjacent normal lung tissues (patient characteristics in Table 1). Immunohistochemistry (IHC) staining, quantified using an Automatic H-score Quantification (AHSQ) online tool, revealed significantly higher SHP2 protein levels in NSCLC tissues compared to adjacent normal tissues (Fig. 1D-E). Elevated SHP2 expression in tumors correlated with larger tumor sizes (Fig. 1F) and a higher incidence of lymph node metastasis (Fig. 1H). Furthermore, SHP2 expression was significantly increased in stage III NSCLC compared to stage I, though differences between other adjacent stages (I vs. II, II vs. III) were not statistically significant (Fig. 1G).
We then assessed the prognostic value of SHP2 expression. Kaplan–Meier analysis showed that elevated SHP2 levels were significantly associated with shorter overall survival (OS; p = 0.0055, F ig. 1I) and disease-free survival (DFS; p = 0.019, Fig. 1J). Univariate analysis (Table 2) confirmed that higher SHP2 expression was linked to poorer OS (p = 0.007) and DFS (p = 0.032), alongside factors such as advanced tumor stage and lymph node metastasis. Importantly, multivariate analysis (Table 3) identified SHP2 expression level as an independent prognostic factor for both OS (HR: 1.391, 95% CI: 1.110–1.744, p = 0.004) and DFS (HR: 1.383, 95% CI: 1.080–1.771, p = 0.01), along with TNM stage. Collectively, these findings establish that SHP2 is upregulated in NSCLC, where its elevated expression significantly associates with aggressive clinicopathological features and serves as an independent predictor of poorer patient survival.

SHP2 is involved in the activation of the IFN-γ/STAT1/IRF1 signaling pathway
To establish cellular models for investigating SHP2's role in NSCLC, we first assessed endogenous SHP2 expression levels in human bronchial epithelial (HBE) cells and a panel of eight NSCLC cell lines by qRT-PCR (Figure S1A). A549 and NCI-H1299 cells exhibited relatively high SHP2 expression, whereas SK-MES-1 and NCI-H1573 cells showed low SHP2 levels. Based on these findings, we generated stable SHP2 knockdown (KD) cell lines using A549 and NCI-H1299 cells, and SHP2 overexpression (OE) cell lines using SK-MES-1 and NCI-H1573 cells. Successful modulation of SHP2 mRNA levels was confirmed by qRT-PCR (Figure S1B-C).
To uncover transcriptional programs regulated by SHP2, we performed bulk RNA sequencing (RNA-seq) on SHP2-KD A549 cells versus control cells. Principal component analysis (PCA) showed distinct clustering based on SHP2 status (Fig. 2A). Notably, Gene Set Enrichment Analysis (GSEA) revealed significant upregulation of both interferon-alpha (IFN-α) and interferon-gamma (IFN-γ) response pathways upon SHP2 knockdown (Fig. 2B-C), suggesting SHP2 negatively regulates these pathways.
Consistent with SHP2 influencing IFN-γ responses, functional assays demonstrated its impact on cell proliferation and survival under IFN-γ treatment. SHP2-OE cells (SK-MES-1, NCI-H1573) exhibited enhanced proliferation (Figure S2B) and increased resistance to IFN-γ-induced growth inhibition compared to control cells (Fig. 2D). Conversely, SHP2-KD cells (A549, NCI-H1299) showed suppressed growth (Figure S2B). Moreover, under IFN-γ treatment, SHP2 knockdown significantly impaired clonogenic growth, while SHP2 overexpression promoted it (Fig. 2E, S2A).
Given the GSEA results, we then directly examined the IFN-γ/STAT1/IRF1 signaling pathway. SHP2 knockdown in A549 and NCI-H1299 cells led to elevated IRF1 mRNA levels, whereas SHP2 overexpression in SK-MES-1 and NCI-H1573 cells suppressed IRF1 mRNA expression, particularly following IFN-γ treatment (Fig. 2F). Western blot analysis further showed that while basal levels of phosphorylated STAT1 (p-STAT1) and IRF1 were similar, IFN-γ stimulation revealed a clear negative correlation between SHP2 expression and the levels of both p-STAT1 and IRF1 (Fig. 2G, S2C). Supporting a direct regulatory role, proximity ligation assays (PLA) confirmed a physical interaction between SHP2 and p-STAT1 in NSCLC cells (Fig. 2H).
Together, these results demonstrate that SHP2 negatively modulates the IFN-γ/STAT1/IRF1 signaling axis in NSCLC cells, at least in part through direct interaction with p-STAT1. This identifies a critical mechanism by which high SHP2 expression in tumor cells may contribute to immune evasion and potentially to anti-PD-1 resistance.

High expression of SHP2 inhibits CD8 + T cell function and promotes resistance to anti-PD1 therapy
To investigate how SHP2 expression within NSCLC cells modulates interactions with CD8 + T cells, we established an in vitro co-culture system (Fig. 3A). Therefore, analysis of co-culture supernatants by ELISA revealed that CD8 + T cells co-cultured with SHP2-knockdown NSCLC cells secreted significantly higher levels of TNF-α, IFN-γ, Granzyme B, and perforin (Fig. 3B–C). In contrast, SHP2 overexpression in tumor cells suppressed the secretion of these key anti-tumor effector molecules by CD8 + T cells (Figure S3A). These data indicate that tumor cell-intrinsic SHP2 inhibits the effector functions of CD8 + T cells, likely contributing to immune evasion.
To establish the in vivo relevance of SHP2's role, we utilized a syngeneic Lewis lung carcinoma (LLC) model in C57BL/6 mice. First, we confirmed the efficacy of SHP2 modulation in LLC cells in vitro: SHP2 knockdown (shSHP2#3) significantly decreased LLC cell proliferation and colony formation, and concurrently increased IRF1 mRNA expression (Figure S3B). Building on this, mice bearing these SHP2-knockdown LLC tumors exhibited markedly suppressed tumor growth and a significantly enhanced therapeutic response to anti-PD-1 antibody treatment when compared to mice bearing control shRNA-transduced tumors (Fig. 3D-F). Consistent with this improved anti-tumor efficacy, analysis of harvested tumor tissues from this model using multiplex immunofluorescence revealed that SHP2 knockdown tumors were characterized by significantly increased infiltration of IFN-γ-producing CD8 + T cells and TNF-α-producing CD8 + T cells (Fig. 3G). Furthermore, immunohistochemistry (IHC) confirmed a corresponding trend of enhanced overall CD8 + T cell infiltration in these SHP2-knockdown tumors (Figure S4).

SHP2 regulates CCL5 secretion in NSCLC, driving CD8 + T cell dysfunction
Given SHP2's modulation of the IFN-γ/STAT1/IRF1 pathway, we next investigated downstream chemokines. RNA sequencing of SHP2-knockdown A549 cells (Fig. 4A) and subsequent Luminex assays on conditioned media from NSCLC cells with differential SHP2 expression (Fig. 4B) revealed that SHP2 knockdown significantly altered the expression and secretion of CCL5, CXCL5, and CXCL9.
Validation by qRT-PCR confirmed these expression changes in SHP2-knockdown A549 cells (Fig. 4C) and demonstrated inverse effects in SHP2-overexpressing SK-MES-1 and NCI-H1573 cells (Fig. 4C, S5A). Notably, only CCL5 showed consistent, significant modulation by SHP2 across all cell lines. Subsequent ELISAs corroborated this at the protein level: SHP2 knockdown significantly increased CCL5 secretion, while SHP2 overexpression markedly reduced it (Fig. 4D-E), identifying CCL5 as a key SHP2-regulated chemokine. To assess CCL5's functional importance, we employed siRNA to knock down CCL5 in SK-MES-1 and NCI-H1573 cells (Fig. 4F, S5B). Critically, CCL5 knockdown in these tumor cells significantly impaired the subsequent secretion of key effector molecules (TNF-α, IFN-γ, Granzyme-B, perforin) by these T cells (Fig. 4G). These findings underscore that tumor-derived CCL5 is essential for robust CD8 + T cell cytotoxic function.
In vivo, C57BL/6 mice bearing SHP2-knockdown LLC tumors (Fig. 3D) also exhibited significantly elevated circulating CCL5 levels during anti-PD-1 therapy compared to controls undergoing the same treatment (Fig. 4H), mirroring our in vitro observations.
Collectively, these data establish SHP2 as a critical negative regulator of CCL5 secretion in NSCLC cells. By suppressing CCL5, SHP2 likely impairs both the recruitment and effector function of CD8 + T cells, thereby promoting immune evasion and contributing to anti-PD-1 resistance, highlighting the SHP2-CCL5 axis as a promising therapeutic target.

SHP2 Inhibition Enhances Anti-Tumor Immunity and Synergizes with PD-1 Blockade in NSCLC
We next investigated the therapeutic potential of pharmacologically inhibiting SHP2 using JAB-3312 (SHP2i). In vitro, we first determined the half-maximal inhibitory concentration (IC50) of JAB-3312 in A549 and NCI-H1299 cells to establish effective concentrations for subsequent assays (Figure S6A-B). Treating SHP2-high human NSCLC cell lines (A549 and NCI-H1299) with JAB-3312 effectively reversed the SHP2-mediated suppression of STAT1 phosphorylation and IRF1 expression upon IFN-γ stimulation, as confirmed by Western blot (Fig. 5A) and qRT-PCR (Fig. 5B). Functionally, JAB-3312 significantly inhibited the proliferation of these cells (Fig. 5C, S6C-D). Beyond these direct effects on tumor cells, JAB-3312 treatment also modulated their immunogenicity, significantly enhancing CCL5 secretion (Fig. 5D). Consequently, co-culture of these JAB-3312-treated tumor cells with CD8 + T cells resulted in boosted secretion of TNF-α, IFN-γ, Granzyme-B, and perforin by the CD8 + T cells (Fig. 5E), indicating an enhanced T cell effector response.
To further establish the link between SHP2 activity and CCL5 regulation within the context of our murine in vivo model system, we also performed mechanistic studies in LLC cells. In vitro, SHP2 knockdown in LLC cells significantly upregulated CCL5 expression and secretion, effects that were reversed by simultaneous siCCL5 treatment (Figure S7A–B). Co-culturing murine CD8 + T cells with SHP2-knockdown LLC cells, which exhibit upregulated CCL5, led to a significant increase in the T cells' secretion of TNF-α, IFN-γ, Granzyme-B, and Perforin. This effect was abrogated when CCL5 was silenced in the SHP2-knockdown LLC cells, confirming the critical role of tumor-derived CCL5 in this process (Figure S7C). Moreover, consistent with its effects in human cells, SHP2 inhibition in LLC cells led to increased IRF1 expression and activation of the pSTAT1/STAT1 axis (Figure S6E), further supporting the immune-stimulatory role of SHP2 inhibition. To translate these mechanistic findings into a more clinically relevant context, we successfully established patient-derived organoid (PDO) models from two non-small cell lung cancer (NSCLC) patients, we utilized these models to evaluate the efficacy of the SHP2 inhibitor, JAB-3312. The resulting IC50 curves clearly demonstrate that this inhibitor exerts significant anti-tumor activity in PDOs from both major NSCLC subtypes (Fig. 5F-G). These data provide preliminary validation of the SHP2 inhibitor's efficacy and offer a valuable in vitro reference for its potential clinical application.
Given these promising in vitro results across both human and murine cell lines suggesting that SHP2 inhibition could improve responsiveness to immune checkpoint inhibitors, we then evaluated JAB-3312 in vivo using the LLC syngeneic tumor model in C57BL/6 mice. Treatment with JAB-3312 significantly enhanced the therapeutic efficacy of anti-PD-1 monoclonal antibody (mAb) therapy, leading to a marked reduction in tumor growth without adversely affecting mouse body weight (Fig. 5H–J, S8A). Notably, the combination of JAB-3312 and anti-PD-1 mAb demonstrated a synergistic anti-tumor effect, significantly suppressing tumor volume compared to monotherapy with either agent alone (Figure S8B). To dissect the mechanism driving this synergy, we performed an adoptive T-cell transfer experiment. This revealed a potent T-cell-intrinsic effect, as the transfer of CD8 + T cells that were pre-treated ex vivo with the SHP2 inhibitor was sufficient to significantly delay tumor growth (Figure S9A-B). This provides strong evidence that the overall therapeutic efficacy of JAB-3312 is attributable to a dual mechanism, which involves not only a direct inhibitory effect on tumor cells but also a potentiation of T-cell anti-tumor responses. Consistent with this findings, systemic SHP2 inhibition with JAB-3312 also led to increased CCL5 levels in the peripheral blood of these tumor-bearing mice, which was associated with enhanced tumor sensitivity to PD-1 blockade (Fig. 5N).
To further elucidate the immune mechanisms underlying the efficacy of JAB-3312 in this model, harvested tumor tissues were analyzed. Immunohistochemistry (IHC) confirmed a corresponding trend of enhanced overall CD8 + T cell infiltration in these combination-treated tumors (Fig. 5K-L, S10).
Furthermore, Multiplex immunofluorescence revealed that tumors from mice treated with the JAB-3312 and anti-PD-1 mAb combination were characterized by significantly increased infiltration of IFN-γ-producing CD8 + T cells and TNF-α-producing CD8 + T cells (Fig. 5M, S11).
In conclusion, the SHP2 inhibitor JAB-3312 not only directly inhibits NSCLC cell proliferation but also enhances anti-tumor immunity by restoring IFN-γ pathway signaling and CCL5 secretion in tumor cells, thereby boosting CD8 + T cell cytotoxicity and effector molecule production. Crucially, these effects, including enhanced functional T cell infiltration, translate to synergistic anti-tumor activity when JAB-3312 is combined with PD-1 blockade in vivo, supporting this combination as a promising therapeutic strategy for NSCLC.

Clinical and transcriptomic analyses reveal SHP2-mediated CCL5 suppression is linked to immune evasion and anti-PD-1 resistance in NSCLC
To assess the clinical relevance of our findings, we first evaluated the association between SHP2 expression, CCL5 levels, and response to anti-PD-1 immunotherapy in a cohort of 29 NSCLC patients. Patients were categorized by treatment response (Progressive Disease (PD), Stable Disease (SD), or Partial Response (PR)) using RECIST 1.1 criteria following six cycles of therapy (Fig. 6A). qRT-PCR analysis of tumor tissues from these patients revealed significantly higher SHP2 mRNA expression in the PD group compared to the PR and SD groups (Fig. 6B). Conversely, ELISA quantification of CCL5 in matched peripheral blood samples showed markedly lower CCL5 levels in patients with PD than in those with PR or SD (Fig. 6C). These results from our cohort suggest that elevated tumor SHP2 expression and reduced systemic CCL5 levels are associated with resistance to anti-PD-1 therapy in NSCLC.
To further validate these observations in a broader clinical context, we analyzed public transcriptomic data. From the GEO database, we utilized a dataset (GSE207422 [32]) comprising bulk RNA sequencing data from patients with resectable NSCLC treated with neoadjuvant PD-1 blockade. Analysis of this dataset showed that CCL5 mRNA expression was significantly elevated in patients achieving a complete response (CR) compared to those with SD or PR (Fig. 6D). Furthermore, a strong negative correlation was observed between SHP2 and CCL5 mRNA expression within this cohort (Fig. 6E), reinforcing their inverse regulatory relationship.
Finally, leveraging larger publicly available datasets, we investigated the prognostic significance of SHP2 and CCL5 and their correlation with immune cell infiltration. Kaplan–Meier survival analysis of 2,166 NSCLC patients from The Cancer Genome Atlas (TCGA) revealed that high SHP2 expression was associated with worse overall survival (Fig. 6F), whereas high CCL5 expression correlated with improved prognosis (Fig. 6G). Importantly, analysis using GSCA datasets [33] demonstrated that higher SHP2 (PTPN11) expression in NSCLC negatively correlated with CD8 + T cell infiltration, while CCL5 expression positively correlated with CD8 + T cell infiltration (Fig. 6H, S12).
Together, these integrated analyses from our patient cohort and multiple public datasets consistently highlight an inverse relationship between SHP2 and CCL5 expression. Elevated SHP2 and consequently reduced CCL5 are associated with impaired CD8 + T cell infiltration, poorer prognosis, and resistance to anti-PD-1 therapy in NSCLC, supporting the crucial role of the SHP2-CCL5 axis in tumor immune evasion.

Discussion
The protein tyrosine phosphatase SHP2 plays multifaceted roles in cellular signaling, positioning it as a critical mediator of the complex interplay between cancer cells and the tumor microenvironment (TME) [34]. While SHP2's function as a key positive regulator of the oncogenic RAS-MAPK pathway is well-documented [19], its intrinsic contributions within cancer cells to modulating immune responses, particularly in the context of immunotherapy resistance, have remained less explored [35]. IFN-γ, a crucial cytokine secreted by effector CD8 + T cells, exhibits a dual role: suppressing tumor growth while potentially driving an inflammatory milieu that facilitates immune escape and resistance to immune checkpoint blockade (ICB) [8]. Our study sought to elucidate the specific role of tumor cell-intrinsic SHP2 in NSCLC, focusing on its impact on tumor progression, immune evasion, and resistance to anti-PD-1 therapy. We propose a model (Fig. 7) wherein tumor cell-intrinsic SHP2 acts as a pivotal node driving immune resistance in NSCLC, primarily by suppressing the canonical IFN-γ/STAT1/IRF1 signaling axis and, consequently, CCL5-mediated immune cell recruitment and function.
Our investigation began by identifying SHP2 as a key regulator of tumor cell resistance to CD8 + T cell cytotoxicity through an unbiased genome-wide CRISPR screen. This finding was substantiated by clinical observations: SHP2 is significantly upregulated in NSCLC tissues compared to adjacent normal tissues, and its elevated expression correlates strongly with advanced tumor stage (larger tumor size, lymph node metastasis) and poorer patient outcomes (reduced OS and DFS). These correlations underscore SHP2's potential not only as a therapeutic target but also as a prognostic biomarker in NSCLC.
Mechanistically, our data compellingly demonstrate that SHP2 negatively regulates the IFN-γ signaling pathway within NSCLC cells. Although IFN-γ typically activates anti-proliferative pathways such as JAK2-STAT1 [36], we found that SHP2 overexpression dampens this response by inhibiting the phosphorylation and activation of STAT1 and its downstream target IRF1. This aligns with and extends existing literature, such as observations linking EMT-associated SHP2 activation to suppressed IFN-γ signaling or studies showing PI3K inhibition can enhance IFN-γ's anti-tumor effects [37–40]. Our findings suggest SHP2 acts as a crucial brake on IFN-γ's beneficial anti-cancer activities. Functionally, this SHP2-mediated suppression of IFN-γ signaling translated directly into impaired CD8 + T cell cytotoxicity against SHP2-overexpressing NSCLC cells, evidenced by decreased secretion of key effector molecules (TNF-α, IFN-γ, Granzyme B, perforin). This highlights a direct mechanism by which tumor-intrinsic SHP2 fosters immune escape.
Effective anti-tumor immunity critically depends on the efficient recruitment of effector T cells into the tumor. Building on previous work highlighting the importance of tumor-derived CCL5 for T cell infiltration [30, 31], we established SHP2 as a novel upstream negative regulator of CCL5 production in NSCLC cells, an effect likely mediated via its suppression of the IFN-γ/STAT1/IRF1 axis. The consequent reduction in CCL5 contributes to an immune-excluded or "cold" tumor microenvironment.
Furthermore, beyond its well-established role as a potent chemoattractant for CD8 + T cells, NK cells, and other leukocytes via receptors such as CCR1, CCR3, and particularly CCR5 [41–43], CCL5 is increasingly recognized for its direct contributions to T cell activation and effector function. Engagement of CCR5 on T cells by CCL5, for instance, can provide co-stimulatory signals that synergize with T cell receptor activation, leading to enhanced T cell proliferation, survival, and the amplification of cytotoxic responses, including the heightened production of IFN-γ, TNF-α, perforin, and granzymes [31, 44]. Our own findings, where CCL5 knockdown in NSCLC cells diminished effector molecule secretion by co-cultured CD8 + T cells, strongly support this broader role. Thus, SHP2-mediated suppression of CCL5 likely impairs anti-tumor immunity through a dual mechanism: reducing T cell infiltration and diminishing the effector capacity of those T cells present.
These mechanistic insights prompted our exploration of SHP2 inhibition as a therapeutic strategy. While SHP2 inhibitors have been combined with MAPK pathway inhibitors [45, 46], their role in augmenting immunotherapy in NSCLC is less established. Our study provides significant preclinical evidence using JAB-3312, a novel SHP2 inhibitor [3], demonstrating that its pharmacological inhibition not only suppresses NSCLC tumor growth but also markedly enhances the efficacy of anti-PD-1 therapy in vivo. This therapeutic benefit was associated with increased peripheral blood CCL5 levels in tumor-bearing mice, consistent with our proposed mechanism. Importantly, it is crucial to also consider the direct impact of systemic SHP2 inhibitors, such as JAB-3312, on immune cells themselves. SHP2 is a well-established negative regulator of T cell activation and effector functions, acting as a critical phosphatase downstream of multiple inhibitory receptors, including PD-1 [47]. By antagonizing activating signals or mediating inhibitory ones, SHP2 in T cells typically serves to dampen immune responses. Therefore, the pharmacological inhibition of SHP2 in T cells can be expected to have direct pro-inflammatory and T cell-enhancing effects, including lowering activation thresholds, promoting proliferation, and augmenting cytotoxic molecule production. The profound synergy we observed between JAB-3312 and anti-PD-1 therapy is thus likely a composite outcome: resulting from both the sensitization of tumor cells (via restored IFN-γ pathway activity and increased CCL5) and the direct bolstering of T cell effector activity due to SHP2 inhibition within T cells. This dual action underscores the promise of SHP2 inhibitors in combination immunotherapy. Dissecting the relative contributions of SHP2 inhibition in tumor versus various immune cell compartments remains a valuable direction for future research.
Our findings establish SHP2 inhibition as a potent strategy to overcome resistance to anti-PD-1 therapy. Contextualizing this approach alongside other emerging strategies in NSCLC highlights its distinct mechanistic advantages. One prominent strategy, for instance, involves targeting the immunosuppressive cytokine TGF-β [48]. Therapeutic agents such as bifunctional fusion proteins, which simultaneously block PD-L1 and sequester TGF-β, are designed to disrupt the fibroblastic, immune-excluded tumor microenvironment (TME) promoted by TGF-β signaling. Although mechanistically distinct, both approaches aim to convert immunologically "cold" tumors into "hot," T-cell-infiltrated phenotypes. However, SHP2 inhibition possesses a unique dual mechanism of action. It not only modulates chemokine profiles to enhance T-cell recruitment but also directly augments T-cell effector function by targeting intracellular pathways downstream of inhibitory receptors such as PD-1 [49].
A second major strategy involves reversing the metabolic dysregulation within the TME that drives T-cell exhaustion. This includes inhibitors of IDO1 or the adenosine pathway (targeting CD73/A2AR), which aim to restore nutrient availability and clear immunosuppressive metabolites, thereby restoring the metabolic fitness of exhausted T cells [50, 51]. In contrast, SHP2 inhibition acts upon a more proximal intracellular signaling node, directly counteracting the inhibitory signals that mediate T-cell exhaustion. Consequently, SHP2 inhibition may synergize with metabolic modulators by concurrently abrogating intrinsic T-cell inhibitory signaling while restoring the metabolic capacity required for a sustained anti-tumor response. This dual functionality positions SHP2 inhibition as a central node for reversing immunotherapy resistance, with clear potential for rational combination with other emerging therapeutic avenues.
Despite these promising preclinical results, the clinical translation of SHP2 inhibitors, including JAB-3312, requires rigorous validation through ongoing and future clinical trials to confirm their safety and efficacy in NSCLC patients [52]. Nevertheless, our findings provide a strong biological rationale for targeting SHP2, particularly in combination with ICIs. Patients with high tumor SHP2 expression, potentially identifiable via IHC, might represent a subgroup that could particularly benefit from such combination strategies.
We acknowledge several limitations in our study. Firstly, our initial in vitro mechanistic explorations utilized established cancer cell lines and allogeneic human T cells. While this system allowed for controlled investigation of tumor-intrinsic SHP2 functions, we recognize its limitations in fully recapitulating patient-specific immunity. Future studies employing autologous co-culture systems, patient-derived organoids with matched immune cells, or models incorporating defined antigen-specific T cell responses are necessary to further validate and extend our findings in a more physiologically relevant context. Secondly, due to experimental constraints, we were unable to establish robust anti-PD-1 resistant patient-derived xenograft (PDX) models, which would have allowed for a more stringent evaluation of JAB-3312's efficacy in overcoming acquired resistance. The lack of access to tumor tissues from clinically resistant patients also precluded deeper mechanistic exploration in that specific setting. These represent important avenues for future work.
In conclusion, this study establishes tumor cell-intrinsic SHP2 as a key node mediating NSCLC pathogenesis, immune evasion, and resistance to anti-PD-1 therapy. Mechanistically, SHP2 suppresses the IFN-γ/STAT1/IRF1/CCL5 signaling axis, reprogramming the tumor's response to IFN-γ to impair CD8 + T cell recruitment and function. Critically, the SHP2 inhibitor JAB-3312 demonstrated potent synergy with anti-PD-1 therapy. This enhanced efficacy stems from a dual mechanism: reversing tumor-intrinsic immune suppression while promoting T-cell activity. These findings provide a strong rationale for developing SHP2-targeted combination immunotherapies as a novel strategy to overcome resistance in NSCLC.

Materials and methods

Clinical specimens
A total of 110 paired NSCLC tissue samples and adjacent non-tumorous tissues, confirmed by two pathologists, were collected from patients who underwent lobectomy or segmentectomy at the Second Affiliated Hospital of Nanchang University between 2017 and 2019.

Human NSCLC tissue microarray (TMA) analysis
Immunohistochemistry (IHC) analysis was conducted on tissue microarrays (TMAs) comprising 110 paired non-small cell lung cancer (NSCLC) tissue samples and corresponding adjacent non-tumorous tissues, provided by Shanghai Biochip Co., Ltd. (Shanghai, China). The experimental procedure adhered to established protocols. Briefly, paraffin-embedded TMAs were deparaffinized in xylene and rehydrated through a graded ethanol series. Endogenous peroxidase activity was blocked by incubating the sections with 3% hydrogen peroxide (H₂O₂) for 30 min at room temperature. To minimize nonspecific binding, the slides were treated with 5% bovine serum albumin (BSA) (Yesen, Shanghai, China) for 1 h at room temperature. The TMAs were then incubated overnight at 4 °C with specific primary polyclonal antibodies (Supplementary Table 1). On the following day, the slides were treated with biotinylated secondary antibodies and incubated for 1 h at room temperature. After each step, thorough washes with phosphate-buffered saline (PBS) were performed to remove unbound reagents. Antigen–antibody complexes were visualized using diaminobenzidine (DAB)-H₂O₂ (GeneTech, Shanghai, China) as the chromogenic substrate, followed by light hematoxylin counterstaining, performed according to the manufacturer’s instructions. Finally, the slides were mounted with cover slips using neutral balsam (Yeasen, Shanghai, China), sealed, and examined under a light microscope. IHC results for SHP2 expression were quantified using the Automatic H-score Quantification (AHSQ) online tool.

CD8 + T cell isolation and Co-culture with tumor cell
CD8 + T cells were isolated from peripheral blood mononuclear cells (PBMCs) of healthy donors using the EasySep™ Direct Human CD8 + T Cell Isolation Kit (STEMCELL Technologies). The purified CD8 + T cells were seeded into 24-well plates and activated with anti-CD3 and anti-CD28 antibodies (BD Biosciences) for 48 h. The cells were then maintained in serum-free medium (Lonza, Switzerland) supplemented with 20 ng/mL IL-2 (SinoBiological, China). Activated CD8 + T cells were co-cultured with tumor cells. After incubation, supernatants were collected for ELISA to assess CD8 + T cell cytokine production.

Mouse CD8 + T cell isolation and preparation
Under ethically approved protocols and sterile conditions, CD8 + T cells were prepared from the spleens of C57BL/6 mice. This involved creating single-cell suspensions, including a red blood cell lysis step for spleen-derived cells, followed by the purification of CD8 + T cells using a EasySep™ Mouse CD8 + T Cell Isolation Kit (Catalog # 19,853). These isolated CD8 + T cells were used directly in experiments activated in vitro for 2–3 days by culturing them on plates pre-coated with an anti-CD3e antibody in media supplemented with soluble anti-CD28 antibody and recombinant mouse IL-2. Concurrently, Lewis Lung Carcinoma (LLC) tumor cells were cultured and seeded into appropriate plates. These co-cultures were subsequently incubated at 37 °C in a 5% CO2 atmosphere for a designated period.

Multiplex immunofluorescence (mIF) staining and analysis
Formalin-fixed, paraffin-embedded (FFPE) tumor tissue Sects. (4 µm thick) from the LLC syngeneic mouse model were processed for multiplex immunofluorescence staining using a Tyramide Signal Amplification (TSA) based method. Sections were deparaffinized in xylene and rehydrated through a graded ethanol series. Heat-induced antigen retrieval was performed by immersing slides in sodium citrate buffer (pH 6.0) and heating in water bath. Endogenous peroxidase activity was then blocked by incubating sections with 3% hydrogen peroxide for 15–20 min at room temperature.
The staining protocol involved sequential rounds of antibody incubation and signal development. In each round, sections were incubated with a specific primary antibody overnight at 4 °C. The primary antibodies used in this study were:
Anti-CD8 (Rabbit anti-Mouse CD8α, [Cell Signaling Technology, Cat# 98941S, Dilution 1:400]).
Anti-IFN-γ (Rat anti-Mouse IFN-γ, [BioLegend, Cat# 517,901, Dilution 1:800]).
Anti-TNF-α (Goat anti-Mouse TNF-α, [R&D Systems, Cat# AF-410-NA, Dilution 1:400]).
Following primary antibody incubation, sections were washed and incubated with an appropriate HRP-conjugated secondary antibody for 1 h at room temperature. After further washes, the signal was developed using a specific TSA-Plus fluorophore for 5–10 min at room temperature. Between each staining round for a different primary antibody, antibodies from the previous round were stripped by heat treatment to allow for the next primary antibody incubation. This cycle of primary antibody, HRP-secondary antibody, TSA reaction, and stripping was repeated for each marker, using a distinct fluorophore for each target. After all rounds of staining were completed, cell nuclei were counterstained with DAPI. Slides were then cover slipped using an anti-fade mounting medium.

Image acquisition and analysis
Stained slides were scanned using a PanoVIEW VS200 slide scanner (Panovue, Beijing, China) at 20X magnification to acquire multispectral images. The acquired multispectral images were then processed using image analysis software OlyVIA software for spectral unmixing to isolate the signal from each fluorophore and for subsequent quantitative analysis. Cell segmentation was performed based on DAPI nuclear staining to identify individual cells. The expression of CD8, IFN-γ, and TNF-α was then quantified on a per-cell basis within the tumor microenvironment. For quantifying phenotypic TILs, CD8 + cells were first identified, and then the percentage of these CD8 + cells co-expressing IFN-γ (IFN-γ + CD8 +/CD8 +) or TNF-α (TNF-α + CD8 +/CD8 +) was determined across different treatment groups. Representative images were selected for presentation. Statistical analysis of quantitative data was performed as described in the Statistical Analysis section.

Supporting Information Doc. S1
The detailed information of Immunohistochemistry (IHC), Cell Culture and Transfection, Western Blot Analysis and RNA Extraction, Colony Formation and CCK-8 Assay for Cell Proliferation Evaluation, CRISPR-Cas9 Screening, In Vivo Tumor Growth and Anti-PD-1 Treatment, Establishment of NSCLC Patient-Derived Organoids, Patient-Derived Organoid (PDO) Drug Response Assay, Adoptive Transfer Models, Luminex Analysis, Proximity Ligation Assay (PLA) and Enzyme-Linked Immunosorbent Assay (ELISA), Bulk RNA Sequencing (RNA-Seq) and Datasets can be found in Doc. S1. Details of the antibodies, ELISA kits, and primers used are provided in Supplementary Table S1-3. The sequences of the shRNA and siRNA are listed in Supplementary Table S4.

Statistical analysis
Categorical variables were compared using the Chi-squared or Fisher's exact test. Differences between groups were analyzed with two-tailed Student's t-tests for independent samples or paired t-tests for matched data. Results are presented as mean ± SEM. The optimal SHP2 H-score cutoff was determined using R software. Kaplan–Meier survival analysis and log-rank test were used to assess survival differences. Univariate and multivariate analyses were conducted with a two-sided Cox proportional hazards regression model to identify independent prognostic factors. A p-value < 0.05 was considered statistically significant.

Supplementary Information