Cancer cell-derived sialylated IgG interacting with Siglec-7/9/10 is a potential immunotherapeutic target in pancreatic cancer
Shenghua Zhang, Ming Cui, Xinmei Huang, Xiaoyao Feng, Ruiling Xiao, Qijia Liu, Jialu Bai, Xianlin Han, Xiaoding Liu, Weiyan Xu, Jing Huang, Quan Liao, Yupei Zhao, Xiaoyan Qiu
TL;DR
A cancer cell protein called SIA-IgG helps pancreatic tumors avoid the immune system, but blocking it could improve immunotherapy.
Contribution
SIA-IgG is identified as a novel immunosuppressive factor in pancreatic cancer that interacts with Siglec receptors.
Findings
SIA-IgG is overexpressed in pancreatic cancer and inhibits macrophage activity.
SIA-IgG and TGF-β1 form a feedback loop to suppress immunity in pancreatic tumors.
Blocking SIA-IgG reverses immunosuppression and shows therapeutic potential in PDAC.
Abstract
The limited effectiveness of T cell-based immune checkpoint blockade (ICB) therapy in most patients with pancreatic ductal adenocarcinoma (PDAC) is largely due to poor CD8+ T cell infiltration and a highly immunosuppressive microenvironment driven by excessive myeloid cell accumulation. This highlights the urgent need for new immunotherapy targets and strategies. In this study, an identified pro-cancer factor, cancer cell-derived sialylated IgG (SIA-IgG), is found to be significantly overexpressed in pancreatic cancer cells. SIA-IgG inhibits macrophage phagocytosis and induces an M2-like immunosuppressive phenotype through interactions with Siglec-7/9/10. SIA-IgG and TGF-β1, a key immunosuppressive factor, reinforce each other in a positive feedback loop, promoting immune evasion in PDAC. Blocking SIA-IgG with specific monoclonal antibodies shows significant therapeutic potential…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsGlycosylation and Glycoproteins Research · Monoclonal and Polyclonal Antibodies Research · Immunotherapy and Immune Responses
Introduction
Pancreatic ductal adenocarcinoma (PDAC) remains one of the deadliest cancers, with an average 5-year overall survival rate of around 10%.1^,^2 Radical surgical resection and systemic chemotherapy are the mainstays of PDAC treatment, yet their effectiveness is still inadequate.3 In the past decade, immunotherapy, particularly immune checkpoint blockade (ICB), has emerged as a promising treatment for various cancers. However, recent clinical trials have shown that PDAC is largely resistant to T cell-based ICB therapies, such as anti-programmed cell death protein 1 (PD-1) and anti-cytotoxic T cell lymphocyte-4 (CTLA-4) antibodies.4^,^5^,^6 The reason is that PDAC is characterized by an immunosuppressive tumor microenvironment (TME), which, in turn, is characterized by prominent infiltration of myeloid cells and scarcity of cytotoxic T cells.7^,^8 Thus, it is difficult for patients with PDAC to benefit from conventional T cell-mediated immunotherapy. Therefore, there is an urgent need to develop novel therapeutic targets and strategies for PDAC immunotherapy.
Immunoglobulins (Igs) in tumors have traditionally been considered anti-tumor antibodies produced by B cells.9 However, in the past decades, numerous lines of evidence have confirmed that non-B cells also widely express Igs, demonstrating not only antibody activity but also significant involvement in cellular biological functions.10^,^11^,^12^,^13^,^14^,^15 Notably, various cancer cells of non-B cell origin, including pancreatic cancer, widely overexpress IgG, which contributes to tumor progression.10^,^16^,^17 Further investigations have demonstrated that epithelial cancer cell-derived IgG shares the same genome and tetrapeptide chain structures as B cell-derived IgG but exhibits distinct glycosylation modifications. These unique glycoforms can be selectively and specifically recognized by the monoclonal antibody RP215, which was originally generated from a screen of over 3,000 hybridoma clones for antibodies that bind to tumor tissues but not to normal counterparts.18^,^19^,^20^,^21 Subsequent collaborative studies identified the RP215-recognized antigen as a non-classical glycoform of the IgG heavy chain. Based on its characteristic terminal sialylation and cancer cell origin, we defined this molecule as SIA-IgG, and, accordingly, renamed RP215 as the anti-SIA-IgG antibody (α-SIA-IgG). Importantly, the sialylation of terminal IgG-Fab at Asn162 confers tumor-promoting properties. For example, SIA-IgG is highly expressed in cancer stem cells and maintains cancer cell stemness.22 Additionally, SIA-IgG functions as a specific ligand for Siglec-7 and Siglec-10 on T cells, facilitating tumor immune evasion as an immune checkpoint.23^,^24
Siglecs are the most important receptor family for sialic acids. Sialic acids bind to most CD33-related Siglecs and induce immune cell tolerogenic programs by phosphorylating SHP-1/SHP-2 through intracellular ITIM or ITIM-like motifs.25^,^26^,^27 Cancer-associated sialylation promotes metastatic spread and immune evasion, including PDAC.24^,^28 We have previously reported that SIA-IgG is specifically overexpressed in PDAC cells and is positively correlated with a poor prognosis.17^,^29 We hypothesized that SIA-IgG may serve as a key ligand for Siglecs on tumor-associated macrophages (TAMs), conferring immune escape properties in PDAC.
In this study, we found that SIA-IgG binds to Siglecs on macrophages and induces M2-like macrophage polarization. Interestingly, SIA-IgG and TGF-β1, the key immunosuppressive factor in PDAC, induce each other to form a positive feedback loop that promotes the immune escape of PDAC. Importantly, utilizing specific antibodies targeting the SIA-IgG/Siglec axis reversed tumor immune evasion and presented a promising therapeutic effect in PDAC. The SIA-IgG/Siglec interaction could serve as a macrophage immune checkpoint and a potential target for immunotherapy in PDAC.
Results
SIA-IgG overexpression correlates with M2-TAM infiltration in PDAC
To investigate the expression of SIA-IgG in PDAC, we first used TCGA database to analyze IgG gene transcripts (IGHG1, IGHG2, IGHG3, and IGHG4) in PDAC and found that the levels of all IgG gene transcripts were higher than those in peritumoral tissues (Figure S1A). We collected tumor samples from 160 PDAC patients at Peking Union Medical College Hospital for tissue microarray (TMA); of the total, 130 cases contained matched adjacent non-cancerous tissues. We further analyzed SIA-IgG expression in PDAC tissues by immunohistochemistry (IHC) using α-SIA-IgG, which recognizes SIA-IgG Fab non-classical sialylation modified at Asn162. We found a high frequency of SIA-IgG positivity in PDAC cases (81.8%), especially in invasive cancer cells, but no or only weak positivity in peritumoral tissues (Figure 1A). SIA-IgG was highly expressed in 48 patients (30%) and was positively correlated with poor prognosis (Figure S1E and Table S1). We next detected SIA-IgG expression in pancreatic cancer cell lines, including cell lysis, extracellular matrix (ECM), and supernatant. We found that SIA-IgG was highly expressed in various pancreatic cancer cell lines, but not in normal pancreatic ductal epithelial HPNE (Figure S1B). SIA-IgG is reported to be highly expressed in cancer stem cells (CSCs) and has been proposed as a potential CSC biomarker.22 In PDAC tissues, we observed significant co-localization of SIA-IgG with CD44v6, a commonly used marker of pancreatic CSCs (Figure 1B).30^,^31 To further investigate this association, we isolated SIA-IgG ^high^ and SIA-IgG ^low^ T3M-4 cells by FACS sorting and found that CD44 expression was markedly elevated in the SIA-IgG ^high^ T3M-4 population compared to the SIA-IgG ^low^ cells (Figure 1C). In parallel, we generated tumor spheres form T3M-4 cells to model stem-like properties, which exhibited markedly elevated levels of both SIA-IgG and CD44 compared to their adherent counterparts (Figure 1D). These results suggest that SIA-IgG ^high^ pancreatic cancer cells may exhibit CSC-like properties.Figure 1SIA-IgG overexpression correlates with M2-TAM infiltration in PDAC(A) Representative images and quantification of IHC for SIA-IgG expression (using α-SIA-IgG) in intratumoral and peritumoral regions. SIA-IgG expression was quantified using an H-score (intensity [0–3] × percentage of positive cells [0%–100%]). Scale bars, 200 μm (n = 130). Statistical analyses were performed using paired t test (^∗∗∗∗^p < 0.0001).(B) Multiple immunohistochemical staining of PDAC sections for nuclei (DAPI), CSC marker CD44v6, and SIA-IgG. Scale bars, 200 μm.(C) Representative flow cytometry sorting method of SIA-IgG ^high^ and SIA-IgG ^low^ T3M-4 cells, and Western blot analysis of indicated proteins in T3M-4 cells after SIA-IgG sorting. Data are representative of two independent experiments.(D) Representative images of tumor spheres and western blot analysis of indicated proteins in T3M-4 cells between sphere cells and parental cells. Data are representative of three independent experiments.(E) Characterization of the immune cell population in PDAC, using ImmuCellAI analysis of PDAC tumor RNA-seq data from TCGA.(F) Representative images and quantification of IHC for M2-TAM infiltration (using anti-CD163 antibody) in peritumoral and intratumoral regions. M2-TAM infiltration was assessed by counting CD163^+^ cells, with the total number per TMA core used as the infiltration score. Scale bars, 200 μm (n = 130). Statistical analyses were performed using paired t test (^∗∗∗∗^p < 0.0001).(G) Correlation of SIA-IgG expression and M2-TAM infiltration in PDAC tissues. Statistical analyses were performed by using Pearson’s correlation analysis.(H) Multiple immunohistochemical staining of PDAC sections for nuclei (DAPI), M1-macrophage cells (CD68^+^ CD163^−^), M2-macrophage cells (CD68^+^ CD163^+^), and cancer cell-derived SIA-IgG. Scale bars, 200 μm.(I) Kaplan-Meier survival analysis of PDAC patients based on combined SIA-IgG expression and M2-TAM infiltration (n = 160). Statistical analyses were performed using Kaplan-Meier survival analysis (^∗∗∗∗^p < 0.0001).PDAC, pancreatic ductal adenocarcinoma; TAM, tumor-associated macrophage; TMA, tissue microarray.See also Tables S1 and S2.
Next, we investigated the potential role of SIA-IgG in mediating immune evasion in PDAC. We used the TCGA database to analyze the infiltration of different immune cell populations in the PDAC TME. The PDAC TME was characterized by a marked lack of effector immune cells, including CD4^+^ T cells, CD8^+^ T cells, and NK cells, yet enriched with TAMs (Figure 1E and Table S4). We analyzed the correlation between SIA-IgG expression and TAM infiltration. M2-TAMs, characterized by CD163, constitute the predominant TAM subset in PDAC.28 IHC revealed abundant infiltration of CD163^+^ M2-TAMs within intratumoral regions, with minimal presence in peritumoral tissues (Figure 1F and Table S2). High intratumoral M2-TAM density was also associated with a poor prognosis (Figure S1F). Notably, SIA-IgG expression showed a significant correlation with intratumoral M2-TAM infiltration (R = 0.57, p < 0.05; Figures 1G and S1G). Moreover, M2-TAMs (CD68^+^ CD163^+^ macrophages) were predominantly distributed around SIA-IgG^+^ cancer cells, whereas M1-TAMs (CD68^+^ CD163^−^ macrophages) were found at greater distances (Figure 1H). Prognostic analysis showed that patients with both high SIA-IgG and high M2-TAM levels had the poorest prognosis, followed by those with high SIA-IgG but low M2-TAMs. Patients with low levels of both markers exhibited the most favorable outcomes (Figure 1I). These findings suggested that SIA-IgG may confer immune escape in PDAC via promoting M2 macrophages.
SIA-IgG polarizes macrophages into M2-like immunosuppressive phenotype
We then investigated whether SIA-IgG directly binds to macrophages. CD14^+^ monocytes were obtained from human peripheral blood mononuclear cells (PBMCs) and differentiated into human primary macrophages by using macrophage colony-stimulating factor (M-CSF) (Figure 2A). IgG was purified from PDAC tissues via protein G and further incubated with a SIA-IgG-specific affinity chromatography column. This process resulted in the purification of PDAC-derived SIA-IgG (Figures S1H and S1I). Compared with the human intravenous immunoglobulin, IgG_3_ and IgG_4_ were significantly enriched in SIA-IgG (Figure S1J). SIA-IgG was found to bind human primary macrophages. Although IgG is known to interact with Fc receptors, Fc receptor blockade only slightly reduced SIA-IgG binding, suggesting an Fc-independent mechanism (Figure 2B).Figure 2SIA-IgG is essential for macrophage polarization toward an immunosuppressive phenotype(A) Schematic of the generation of human primary macrophages from CD14^+^ monocytes.(B) Binding of SIA-IgG to human primary macrophages with or without FcR blocking. Data are representative of three independent experiments.(C) Schematic of conditioned culture of macrophages using extracellular matrix (ECM) derived from differently treated T3M-4 cells.(D) Western blot analysis of SIA-IgG in T3M-4 cells after siRNA-mediated knockdown.(E and F) Western blot analysis of indicated proteins in THP-1-derived macrophages after conditioned cultures with T3M-4 cell-derived ECM following IgG knockdown (E) or SIA-IgG blockade (F).(G) Relative mRNA expression of TGF-β1, IL-6, and IL-1β in human primary macrophages after conditioned cultures with T3M-4 cell-derived ECM with or without SIA-IgG blockade. Expression levels were normalized to M0 macrophage (set as 1) (n = 3). Data are presented as the mean ± SEM. Statistical analyses were performed by using paired t test (^∗∗^p < 0.01, ^∗∗∗∗^p < 0.0001).(H) Schematic of the macrophages cultured with ECM derived from sorted SIA-IgG ^high^ or SIA-IgG ^low^ T3M-4 cells.(I) Relative mRNA expression of TGF-β1, ARG1, IL-6, and IL-1β in human primary macrophages after conditioned culture with SIA-IgG ^high^ or SIA-IgG ^low^ T3M-4 cell-derived ECM. Expression levels were normalized to macrophages cultured with SIA-IgG ^low^ T3M-4 cells (set as 1) (n = 3–4). Data are presented as the mean ± SEM. Statistical analyses were performed by using paired t test (^∗^p < 0.05, ^∗∗^p < 0.01). Data are representative of three independent experiments (D–F).
To determine whether SIA-IgG can induce macrophage polarization, we cultured THP-1-derived macrophages and human primary macrophages with ECM derived from differently treated T3M-4 cells (Figure 2C). Knockdown of IgG in T3M-4 cells (Figure 2D) markedly impaired M2 polarization: THP-1-derived macrophages exhibited decreased expression of M2-related molecules, such as TGF-β1 and ARG1, and increased expression of M1-related molecules, such as IL-18 and IL-1β (Figure 2E). Blocking SIA-IgG significantly reduced ARG1 and TGF-β1 expression in THP-1-derived macrophages, while exogenous addition of SIA-IgG, but not non-SIA-IgG, reversed this phenotype (Figure 2F). We also observed a decrease in TGF-β1 and IL-6 mRNA expression and an increase in IL-1β mRNA expression in human primary macrophages after blocking SIA-IgG (Figure 2G). Macrophages cultured with ECM from SIA-IgG ^high^ T3M-4 cells showed significant higher levels of TGF-β1, IL-6, and ARG1 mRNA expression than those cultured with ECM from SIA-IgG ^low^ cells (Figures 2H and 2I).
To explore whether SIA-IgG can directly induce macrophage M2 polarization, we directly stimulated human primary macrophages with purified SIA-IgG. Mass spectrometry (MS) analysis indicated that SIA-IgG induced the upregulation of M2-related genes, such as CD163, CD206 (MRC1), ARG1, and TGF-β1 (Figure 3A and Table S5). Further experiments showed that SIA-IgG upregulated CD163 expression and increased TGF-β1, IL-10, IL-6, and ARG1 mRNA expressions in human primary macrophages (Figures 3B and 3C). Similar induction of TGF-β1 expression was observed in THP-1-derived macrophages (Figures S2A and S2B). Importantly, SIA-IgG also induced TGF-β1 expression in M1-polarized macrophage, indicating a capacity to reprogram M1-like macrophages into M2-like macrophages (Figure S2C). In addition, we obtained mouse bone marrow-derived macrophages (BMDMs) and treated them with SIA-IgG (Figure S2D), MS analysis indicated that SIA-IgG induces M2-like macrophages (Figure S2E and Table S6). Further experiments showed that SIA-IgG directly induced ARG1 and IL-6 mRNA expressions and increased the secretion of IL-10, TGF-β1, and IL-6 in BMDMs (Figures S2F and S2G). Moreover, SIA-IgG induced PD-L1 protein and mRNA expression in human and mouse primary macrophages (Figures S2H and S2I). Importantly, we found that mouse CD8^+^ T cell proliferation was significantly inhibited after co-culturing with SIA-IgG-treated mouse BMDMs (Figures 3D and 3E). These results suggest that SIA-IgG inhibits CD8^+^ T cell proliferation by inducing the macrophage immunosuppressive phenotype.Figure 3SIA-IgG directly induces the immunosuppressive phenotype of macrophages(A) Heatmap of significant differentially expressed genes in human primary macrophages stimulated with SIA-IgG versus non-SIA-IgG analyzed by mass spectrometry.(B) Relative mRNA expression of TGF-β1, IL-10, IL-6, and ARG1 in human primary macrophages after stimulation with SIA-IgG or non-SIA-IgG. Expression levels were normalized to M0 macrophages (set as 1) (n = 3). Data are presented as the mean ± SEM. Statistical analyses were performed by using paired t test (ns, not significant; ^∗∗^, p < 0.01; ^∗∗∗^, p < 0.001; ^∗∗∗∗^, p < 0.0001).(C) Flow cytometry analysis of CD163 expression on human primary macrophages stimulated with SIA-IgG or non-SIA-IgG (n = 5). Data are presented as the mean ± SEM. Statistical analyses were performed by using paired t test (^∗^, p < 0.05; ^∗∗^, p < 0.01).(D) Schematic of the mouse T cell proliferation assay, indicating co-culturing of T cells with differentially treated mouse primary macrophages.(E) Representative flow cytometry plots and quantification for mouse CD8^+^ T cell proliferation after 3 days of co-culture with macrophages pre-conditioned with SIA-IgG or non-SIA-IgG (n = 3). Data are presented as the mean ± SEM. Statistical analyses were performed by using unpaired t test (^∗∗^p < 0.01).(F) Schematic of the macrophage phagocytosis assay.(G and H) I**n vitro macrophage phagocytosis assay using CFSE- or pHrodo Red-labeled T3M-4 cells. (G) Representative flow cytometry plots and quantification of in vitro phagocytosis of CFSE-labeled T3M-4 cells in the presence of 20 μg/mL α-SIA-IgG or mIgG control (n = 3). (H) Representative fluorescence microscopy images of in vitro phagocytosis in the presence of 20 μg/mL α-SIA-IgG or mIgG control. Images are representative of two of four donors (n = 4). Scale bars, 200 μm. Data are presented as the mean ± SEM. Statistical analyses were performed by using paired t test (^∗∗^p < 0.01).CFSE, carboxyfluorescein diacetate succinimidyl ester.See also Figure S2.
Macrophages are professional phagocytic cells, and the CD47-SIRPα pathway is an important mechanism for macrophages to distinguish “self” from “non-self,”32 We demonstrated that SIA-IgG induced CD47 and SIRPα expression in the macrophages (Figure 3A). To determine whether SIA-IgG inhibits macrophage phagocytosis, tumor cells labeled with pHrodo Red or carboxyfluorescein diacetate succinimidyl ester (CFSE) were incubated with human primary macrophages under serum-free conditions (Figure 3F).33 Macrophage phagocytosis was assessed by fluorescence microscopy and flow cytometry, revealing that blockade of SIA-IgG with α-SIA-IgG significantly enhanced the phagocytic activity. These findings indicate that SIA-IgG suppresses macrophage phagocytosis (Figures 3G and 3H).
SIA-IgG binding with Siglec-7/9/10 polarizes macrophages
Given that SIA-IgG relies on sialic acids to induce a macrophage immunosuppressive phenotype, we investigated whether SIA-IgG regulates macrophage function through sialic acid receptors. Sialic acid receptors on macrophages include Siglecs and DC-SIGN (CD209).34 We analyzed the single-cell RNA sequencing (scRNA-seq) data of PDAC35 and found that all sialic acid receptors, including Siglecs and DC-SIGN (CD209), were predominantly expressed in monocytes and macrophages within the pancreatic TME. Specifically, macrophages mainly express Siglec-1, -3, -7, -9 and -10, of which Siglec-3, -7, -9, and -10 mediate immunosuppressive signals through ITIM motifs (Figures 4A and S3A). Siglec-1, -3, -7, -9, and -10 were found to be mainly expressed in CD68^+^ CD163^+^ M2-like macrophages (clusters 4, 12, 17, and 27), with little or no expression in CD68^+^ CD163^−^ M1-like macrophages (clusters 5, 13, 14, 16, and 20) (Figure S3B), suggesting that Siglec signaling may confer M2 macrophage polarization.Figure 4SIA-IgG binding with Siglec-7/9/10 polarizes macrophages(A) Expression of representative sialic acid receptors in different cell populations from PDAC scRNA-seq analysis. See also Figure S3.(B) Western blot analysis of the interaction between His-Siglec-9 and GFP-SIA-IgG detected by co-IP in 293FT. Data are representative of three independent experiments.(C) SPR analysis showing specific binding of human Siglec-9 to SIA-IgG, but not to neuraminidase (α-2-3,6,8)-treated SIA-IgG or non-SIA-IgG. See also Figure S4.(D) Western blot analysis of the SHP1 phosphorylation in THP-1-derived macrophages treated with SIA-IgG or non-SIA-IgG.(E–H) Percentage of M2-like macrophages (n = 4) (E), and the relative mRNA expression of IL-6 (n = 4) (F), TGF-β1 (n = 4) (G), and ARG1 (n = 3) (H) in human primary macrophage treated with SIA-IgG, with or without 20 μg/mL goat anti-Siglec-1, -7, -9, or -10 antibody or goat IgG. Data are presented as the mean ± SEM. Statistical analyses were performed by using paired t test (ns, not significant; ^∗^, p < 0.05; ^∗∗^, p < 0.01; ^∗∗∗^, p < 0.001; ^∗∗∗∗^, p < 0.0001).(I) Flow cytometry quantification of in vitro phagocytosis of CFSE-labeled T3M-4 cells by macrophages with or without 20 μg/mL goat anti-Siglec-1, -7, -9, -10 antibody or goat IgG (n = 4). Data are presented as the mean ± SEM. Statistical analyses were performed by using paired t test (^∗∗^, p < 0.01).(J and K) Representative flow cytometry plots and quantification of Siglec-9 and Siglec-10 (n = 6–7) (J), and relative mRNA expression of Siglec-9 and Siglec-10 (K) in human primary macrophages after stimulation with SIA-IgG or non-SIA-IgG. Expression levels were normalized to M0 macrophages (set as 1) (n = 4). Data are presented as the mean ± SEM. Statistical analyses were performed by using paired t test (ns, not significant; ^∗^, p < 0.05; ^∗∗^, p < 0.01).co-IP, co-immunoprecipitation; SPR, surface plasmon resonance; CFSE, carboxyfluorescein diacetate succinimidyl ester.See also Figure S5.
We then sought to determine if SIA-IgG induces M2-like macrophage polarization via Siglecs. We previously reported that SIA-IgG interacts with Siglecs-7 and -10.23^,^24 In this study, we further confirmed that SIA-IgG interacts with Siglec-9 but not with Siglec-3 (Figures 4B and S4A). Importantly, SIA-IgG-CH1mut (N162Q) showed a nearly complete loss of interaction with Siglec-9 compared to wild-type SIA-IgG (SIA-IgG-WT) (Figure 4B). Surface plasmon resonance (SPR) and ELISA showed that human Siglec-9 binds to SIA-IgG with high affinity and that the interaction is sialic acid dependent, as enzymatic desialylation completely abolished binding (Figures 4C and S4B–S4F). Consistently, no binding was detected between Siglec-9 and non-SIA-IgG under identical conditions (Figure 4C). We also confirmed that the SIA-IgG interaction with murine Siglec-E and Siglec-G primarily relies on the sialic acid at Asn162, a mechanism conserved in human Siglecs (Figures S4G and S4H). Furthermore, SIA-IgG, but not non-SIA-IgG, phosphorylated SHP-1 to transmit inhibitory signals by interacting with Siglecs-7, -9, and -10 on THP-1-derived macrophages (Figure 4D). To investigate whether SIA-IgG induces polarization of M2-like macrophages via Siglecs-7, -9, and -10, we blocked Siglec signaling by using anti-Siglec-1, -7, -9, and -10 neutralizing antibodies, followed by treatment of human primary macrophages with SIA-IgG. We found that SIA-IgG-induced M2-like macrophage polarization was partially blocked by anti-Siglec-7, -Siglec-9, and -Siglec-10 antibodies and almost completely blocked by α-SIA-IgG (Figure 4E). Furthermore, we found that SIA-IgG induced IL-6 mRNA expression primarily through Siglec-10 and partially through Siglec-7 and Siglec-9 (Figure 4F). SIA-IgG also induced TGF-β1 mRNA expression via Siglec-7, Siglec-9, and Siglec-10 (Figure 4G) and ARG1 mRNA expression primarily through Siglec-10 (Figure 4H). Together, these results indicated that SIA-IgG induces M2-like immunosuppressive macrophages via interactions with Siglec-7, Siglec-9, and Siglec-10. Also, blocking the SIA-IgG-Siglec10 interaction significantly enhanced macrophage phagocytosis (Figure 4I).
We next explored whether SIA-IgG can regulate Siglec expression in macrophages. Human primary macrophages were cultured with T3M-4-derived ECM, and blocking SIA-IgG significantly reduced the expression of Siglecs-1, -3, -7, -9, and -10 on these macrophages (Figure S5A). Additionally, ECM from SIA-IgG ^high^ T3M-4 cells induced higher levels of Siglec-1, -9, and -10 protein and mRNA expression compared with SIA-IgG ^low^ T3M-4 cells (Figures S5B and S5C). To determine if SIA-IgG directly induces Siglec expression, we performed MS analysis and found that SIA-IgG induced Siglec expression in human primary macrophages (Figure S5D and Table S5). Specifically, SIA-IgG upregulated Siglec-9 and -10 protein and mRNA expressions (Figures 4J and 4K) but did not affect Siglec-1, -3, or -7 and DC-SIGN (CD209) expressions in human primary macrophages (Figures S5E–S5G).
SIA-IgG and TGF-β1 form a positive feedback loop in PDAC
We further explored the regulation of SIA-IgG in the TME. IL-10 and TGF-β are abundant in the TME and known to induce IgG expression in B cells.36^,^37 We hypothesized that IL-10 and TGF-β1 also induce SIA-IgG expression in PDAC cells. We added IL-10 or TGF-β1 to T3M-4 and PK-9 PDAC cell lines. Interestingly, we observed that both IL-10 and TGF-β1 promoted SIA-IgG expression in the T3M-4 and PK-9 cells (Figures 5A, 5B, and S1D).Figure 5SIA-IgG and TGF-β1 form a positive feedback loop in PDAC(A) Western blot of SIA-IgG treated with TGF-β1 in T3M-4 and PK-9 cells.(B) IGHG mRNA expression treated with TGF-β1 in T3M-4 cells. Expression levels were normalized to the control (set as 1). Data are presented as the mean ± SEM. Statistical analyses were performed by using unpaired t test (^∗∗^, p < 0.01; ^∗∗∗^, p < 0.001).(C) Western blot analysis of SIA-IgG after neuraminidase or pNGase F treatment.(D) Heatmap of differentially expressed genes involved in sialylation treated with TGF-β1 in the murine PDAC cell line KPC (GSE149130).(E) Western blot analysis of ST6GAL1 expression treated with TGF-β1 in T3M-4 cells.(F) Representative images of IHC for the expression of SIA-IgG and ST6GAL1 in human PDAC tissues (n = 3). Scale bars, 100 μm.(G) Western blot analysis of SIA-IgG and ST6GAL1 in T3M-4 cells after transfection with ST6GAL1 siRNAs.(H) Western blot analysis of intracellular and secreted TGF-β1 in T3M-4 cells treated with SIA-IgG.(I) Immunofluorescence images of SIA-IgG and ST6GAL1 after ST6GAL1 knockdown in T3M-4 cells. Scale bars, 100 μm. All data are representative of three independent experiments.PDAC, pancreatic ductal adenocarcinoma.
We then investigated the sialylation modification of SIA-IgG by treating PDAC-derived SIA-IgG with various neuraminidases. Glycopeptide analysis confirmed that SIA-IgG Fab contains a non-classical glycosylation modification at the Asn162 terminal with a single sialic acid residue.24 This suggests that the terminal sialic acid on SIA-IgG may be α-2,3 or α-2,6 linked.38 Treatment with α-2,3 neuraminidase did not affect SIA-IgG signaling, while α-2-3,6,8 neuraminidase treatments significantly reduced SIA-IgG signaling (Figure 5C), indicating that the sialic acid recognized by the α-SIA-IgG is predominantly α-2,6 linked. Importantly, TGF-β1 induced ST6GAL1 but not ST6GAL2 expression in both mouse39 and human pancreatic cancer cells (Figures 5D and 5E). SIA-IgG and ST6GAL1 were found to be co-distributed in PDAC tissues (Figure 5F). To further investigate whether ST6GAL1 is involved in SIA-IgG sialylation modification, we knocked down ST6GAL1 in T3M-4 cells and observed a significant reduction in SIA-IgG expression (Figures 5G and 5I). These results suggest that TGF-β1 promote SIA-IgG sialylation through upregulation of ST6GAL1. Notably, SIA-IgG itself promoted TGF-β1 expression in both cancer cells and macrophages (Figures 5H and 3B), establishing a positive feedback loop that accelerates PDAC progression.
Targeting SIA-IgG reverses the suppressive TME in PDAC
Building on these findings, we next sought to evaluate whether SIA-IgG can serve as a potential therapeutic target for PDAC. We first confirmed that the KPC murine pancreatic cancer cell line also expressed SIA-IgG (Figure S1C). Next, we established a subcutaneous tumor model in C57BL/6 mice by using KPC cells and treated the tumors with α-SIA-IgG (Figure 6A). Treatment with α-SIA-IgG significantly inhibited tumor growth in vivo without noticeable toxicity or side effects (Figures 6B–6F). We analyzed immune cell infiltration in the TME of subcutaneous tumors. After α-SIA-IgG treatment, we observed a decrease in macrophage infiltration, especially those of M2 macrophages, in the TME (Figures 6G–6I and S6A). Notably, both CD8^+^ T and CD4^+^ T cell infiltration were found to increase following α-SIA-IgG treatment (Figures 6J and 6K).Figure 6. Targeting SIA-IgG reverses the suppressive tumor microenvironment in PDAC(A) Schematic of the experimental design for subcutaneous tumor model in C57BL/6 mice by using KPC cells.(B) Representative image of harvested subcutaneous tumors.(C and D) Plots for tumor weight (C) and tumor volume (D) of the mice intravenously injected with 5 mg/kg α-SIA-IgG or mIgG control (n = 9) at the end of study. Data are presented as the mean ± SEM. Statistical analyses were performed by using unpaired t test.(E and F) Curves for tumor growth (E) and body weight (F) of the mice intravenously injected with 5 mg/kg α-SIA-IgG or mIgG control (n = 9). Data are presented as the mean ± SEM. Statistical analyses were performed by using unpaired t test.(G–K) Flow cytometry and immunohistochemistry analysis of immune cells population (n = 4–7) in subcutaneous tumors.(G and H) Representative flow cytometry plots (G) and quantification (H) showing the percentage of macrophages within CD45^+^ cells population.(I) Immunohistochemical quantification of M2 macrophage infiltration using anti-CD206 antibody.(J and K) Representative flow cytometry plots (J) and quantification (K) of CD4^+^T cells and CD8^+^T cells within CD45^+^ cell population. Data are presented as the mean ± SEM. Statistical analyses were performed by using unpaired t test.(L) Schematic of the experimental design for orthotopic tumor model in C57BL/6 mice using KPC cells.(M–P) Representative image of harvested orthotopic tumors (M), tumor weight plots (N), spleen weight plots (O) and body weight (P) of the mice intravenously injected with 5 mg/kg α-SIA-IgG or mIgG control (n = 9). Data are presented as the mean ± SEM. Statistical analyses were performed by using unpaired t test.(Q–T) Flow cytometry analysis of immune cells population (n = 5) in orthotopic tumors.(Q and R) Representative flow cytometry plots (Q) and quantification (R) of M2-macrophages (F4/80^+^CD206^+^ cells) and M1-macrophages (F4/80^+^CD206^−^ cells) within CD45^+^ cell population.(S and T) Representative flow cytometry plots (S) and quantification (T) of CD8^+^T cells and CD4^+^T cells within CD45^+^ cells population. Data are presented as the mean ± SEM. Statistical analyses were performed by using unpaired t test.See also Figure S6.
To more accurately recapitulate the native TME and provide greater physiological relevance, we constructed an orthotopic tumor model by using the KPC cell line in C57BL/6 mice and treated the tumors with α-SIA-IgG (Figure 6L). α-SIA-IgG significantly inhibited the tumorigenic ability of KPC in vivo without obvious toxicity or side effects, and a reduction in pathological enlargement of the spleen was observed (Figures 6M–6P). We also observed a reduction in M2 macrophage infiltration following treatment (Figures 6Q and 6R), while no significant change was observed in monocytic myeloid-derived suppressor cell (M-MDSC) (Figure S6C). Importantly, α-SIA-IgG treatment significantly promoted CD8^+^ T cell infiltration (Figures 6S and 6T), although the exhaustion of infiltrated CD8^+^ T cells showed only slight improvement (Figures S6E and S6F). Meanwhile, no significant changes were observed in CD8^+^ Tregs or CD4^+^ Tregs (Figures S6E–S6G). ScRNA-seq data further confirmed that α-SIA-IgG treatment markedly reduced M2 macrophage infiltration and suppressed fibroblast function (Figure S6H). These results indicate that targeting SIA-IgG can reverse the suppressive TME in PDAC.
SIA-IgG/Siglec axis serves as a macrophage immune checkpoint for PDAC immunotherapy
Given that PDAC is characterized by prominent myeloid cell infiltration and scarce of cytotoxic T cells, we used T cell-deficient nude mice and NOD-SCID mice to construct tumor models and evaluate whether the SIA-IgG/Siglec axis serves as a macrophage immune checkpoint for PDAC immunotherapy. We humanized α-SIA-IgG and screened to generate humanized anti-SIA-IgG monoclonal antibody (HASA). The affinity of α-SIA-IgG for the antigen SIA-IgG was 2.051 × 10^−9^ M, and the affinity of HASA was 3.702 × 10^−9^ M, indicating that the humanization process did not affect the antibody’s affinity for SIA-IgG (Figures S7A and S7B).
We then used T3M-4 cells to establish a subcutaneous tumor model in nude mice and treated them with HASA (Figure S7C). HASA significantly inhibited tumor growth without causing obvious toxicity and side effects (Figures S7D–S7H). To further highlight the therapeutic effect of HASA in an orthotopic model, T3M-4 cells were orthotopically implanted into nude mice, followed by HASA treatment (Figure 7A). HASA significantly inhibited tumor growth without noticeable toxicity or side effects, and a reduction in pathological enlargement of the spleen was similarly observed (Figures 7B–7E). We also observed that HASA effectively suppressed M2 macrophage infiltration (Figures 7F, 7G, and S6B), and no significant change was observed in M-MDSC (Figure S6D).Figure 7SIA-IgG/Siglec axis serves as a macrophage immune checkpoint for PDAC immunotherapy(A) Schematic of the experimental design for orthotopic tumor model in nude mice using T3M-4 cells.(B) Representative image of harvested orthotopic tumors.(C–E) Plots for tumor weight (C), spleen weight (D), and body weight (E) of the mice intravenously injected with 5 mg/kg HASA or IVIG control (n = 5). Data are presented as the mean ± SEM. Statistical analyses were performed by using unpaired t test.(F and G) Representative flow cytometry plots (F) and quantification (G) showing the percentage of macrophages within CD45^+^ cells population. Data are presented as the mean ± SEM. Statistical analyses were performed by using unpaired t test.(H) Schematic of the experimental design for PDAC patient-derived PDX model.(I) Representative image of harvested PDX tumors.(J and K) Curves for tumor growth (J) and body weight (K) after intravenous treatment with 5 mg/kg α-SIA-IgG, HASA, or mIgG control (n = 10). Data are presented as the mean ± SEM. Statistical analyses were performed by using unpaired t test (^∗∗∗∗^, p < 0.0001).(L and M) Plots for tumor volume (L) and tumor weight (M) after intravenous treatment with 5 mg/kg α-SIA-IgG, HASA, or mIgG control (n = 10). Data are presented as the mean ± SEM. Statistical analyses were performed by using unpaired t test (^∗∗∗∗^, p < 0.0001).(N and O) Representative flow cytometry plots (N) and quantification (O) of M1 macrophages (F4/80^+^CD206^-^ cells) and M2 macrophages (F4/80^+^CD206^+^ cells) within CD45^+^ cell population (n = 5). Data are presented as the mean ± SEM. Statistical analyses were performed by using unpaired t test.(P) Schematic of the experimental design for macrophage depletion combined with anti-SIA-IgG treatment in tumor-bearing mice.(Q–T) Tumor growth curve (Q), representative tumor image (R), tumor weight plots (S), and tumor volume plots (T) of the harvested T3M-4 CDX tumors treated with 5 mg/kg HASA or IVIG control, with or without macrophage depletion (n = 8). Data are presented as the mean ± SEM. Statistical analyses were performed by using unpaired t test (^∗^, p < 0.05; ^∗∗∗∗^, p < 0.0001).HASA, humanized anti-SIA-IgG monoclonal antibody; IVIG, intravenous immunoglobulin; PDX, patient-derived xenograft.
To evaluate the therapeutic effects of HASA in a more clinically relevant contact, we established patient-derived xenograft (PDX) models, using tumor tissues from PDAC patients (Figure 7H). Both α-SIA-IgG and HASA significantly inhibited the growth of human PDAC PDXs in vivo without apparent toxicity or side effects (Figures 7I–7M). Notably, both α-SIA-IgG and HASA inhibited M2-like macrophage infiltration in PDAC PDXs, while having no significant effect on M1-like macrophage infiltration (Figures 7N and 7O).
Given the limited T cell infiltration in the TME of PDAC, we further investigated whether combining anti-SIA-IgG antibody with macrophage depletion could provide a therapeutic approach. Subcutaneous tumor models were established by inoculating T3M-4 cells into nude mice, followed by macrophage depletion using liposomal clodronate (Figure 7P). The combination of HASA and macrophage depletion resulted in significant inhibition of tumor growth compared with macrophage depletion or HASA treatment alone (Figures 7Q–7T). These results suggest that the SIA-IgG/Siglec axis functions as a macrophage immune checkpoint in PDAC, and HASA holds promise for future clinical application in PDAC immunotherapy.
Discussion
In this study, we demonstrated that PDAC cell-derived SIA-IgG promotes the polarization of M2-like macrophages through binding to Siglecs-7, -9 and -10 and confers cancer cell immune escape properties. We further showed that TGF-β1 markedly upregulates SIA-IgG expression in cancer cells, while SIA-IgG, in turn, induces TGF-β1 secretion from both cancer cells and macrophages, establishing a positive feedback loop that drives PDAC progression. Finally, using multiple in vivo models, we confirmed that the SIA-IgG/Siglec axis serves as a macrophage immune checkpoint and a promising target for PDAC immunotherapy.
Studies have shown that pancreatic cancer cells exhibit increased sialylation, which facilitates binding with Siglecs and promotes M2-TAM polarization.28^,^40 Additionally, sialylated ligands such as CD24/Siglec-10 and MUC-1/Siglec-9 pathways have been reported to inhibit macrophage phagocytosis and drive M2-like macrophage polarization in breast and ovarian cancers.33^,^41 However, the specific sialylated ligands on pancreatic cancer cells and their role in fostering an immunosuppressive TME in PDAC remain poorly understood. In this study, we demonstrated that PDAC cell-derived SIA-IgG serves as a crucial ligand for Siglecs-7, -9 and -10, inducing M2-like macrophage polarization through interaction with these Siglecs. Mutating SIA-IgG CH1 (N162Q) nearly abolished its interaction with Siglecs, indicating that the binding is primarily dependent on the terminal sialic acid at the Asn162 N-glycosylation site. Furthermore, while Siglec-9 is known to preferentially bind α-2,3-linked sialic acids, it can also recognize α-2,6- and α-2,8-linked sialic acids with lower affinity.42^,^43^,^44 Consistent with this, our results show that Siglec-9 binds SIA-IgG at Asn162 in a sialic acid-dependent manner, demonstrating a broader ligand recognition range than that reported previously. Similar to the CD24/Siglec-10 pathway,33 we observed that SIA-IgG/Siglec-10 interaction inhibited macrophage phagocytosis, which presents another potential mechanism for targeting macrophages in PDAC therapy. Notably, while little is known about Siglec inducers in the TME, we identified SIA-IgG as a key inducer of Siglec-9 and Siglec-10 expression.
Another important finding of this study is that SIA-IgG and TGF-β1 form a positive feedback loop that sustains the immunosuppressive TME and promotes immune escape in PDAC. TGF-β signaling is known to accelerate pancreatic tumorigenesis by enhancing epithelial-to-mesenchymal transition (EMT) and fibrosis and evading cytotoxic immune surveillance.45 In this study, we demonstrated that TGF-β1 strongly induces SIA-IgG expression in PDAC cells. More importantly, TGF-β1 promotes IgG heavy chain sialylation at Asn162 by upregulating ST6GAL1 expression, thereby ensuring that IgG is highly sialylated. As observed, the high level of SIA-IgG expression in T3M-4 cells may be due to SMAD4 mutations in T3M-4, which cause abnormal TGF-β signaling, in combination with the KRAS Q61H mutation in T3M-4 cells that leads to increased ST6GAL1 expression.46 Additionally, we demonstrated that SIA-IgG significantly promotes the expression and secretion of TGF-β1 in PDAC cells. These findings suggest that SIA-IgG and TGF-β1 form a positive feedback loop that sustains the immunosuppressive TME in PDAC.
Given its critical role in shaping the immunosuppressive TME, SIA-IgG represents a promising target for PDAC immunotherapy. The primary objective of most immunotherapies is to activate the adaptive immune response, particularly CD8^+^ T cells. However, overcoming the immunosuppressive PDAC TME presents a significant challenge for T cell-based ICB therapies.4^,^47^,^48 Notably, suppressive myeloid cells are key initiators of the PDAC immunosuppressive TME. Thus, targeting TAMs has emerged as a potential strategy for PDAC immunotherapy.49 Given that SIA-IgG features terminal sialylated IgG-Fab N-glycosylation at Asn162, which is distinct from B cell IgG and normal non-B cell-derived IgG, targeting this sialylation modification at the Asn162 epitopes holds significant promise as a therapeutic approach for PDAC. Particularly, anti-SIA-IgG may exhibit enhanced efficacy when combined with CCR2 inhibitors49 or anti-TGF-β1 agents.50 The humanized modification of therapeutic antibodies targeting SIA-IgG has been successfully achieved and demonstrated comparable efficacy in PDAC, highlighting its potential for clinical application.
In conclusion, the SIA-IgG/Siglec axis represents a potential immunotherapeutic target in PDAC, with the potential to reshape current therapeutic strategies. Targeting this pathway could enhance immune response by modulating macrophage polarization, offering an approach for overcoming the immunosuppressive TME in PDAC.
Limitations of the study
While our study demonstrates that the SIA-IgG/Siglec axis has potential as an immunotherapeutic target in PDAC, several aspects require further investigation. First, the role of SIA-IgG in pancreatic cancer stem cells remains unclear. Future studies with more in-depth in vitro and in vivo experiments are needed to elucidate its specific effects within this important subpopulation. Second, the impact of SIA-IgG on other immune cell populations in the TME, including its effects on key immune receptors, remains to be fully explored. In addition, the therapeutic efficacy of combining SIA-IgG targeting with conventional chemotherapy, immune checkpoint inhibitors, and targeted therapies warrants further evaluation. Finally, while the safety of α-SIA-IgG is suggested, phase 1 clinical trials are necessary to establish both the safety and therapeutic efficacy of this antibody. In particular, optimal dosing and the evaluation of clinical benefits in PDAC patients will be crucial for its potential therapeutic application.
Resource availability
Lead contact
For further information and requests, please direct all inquiries to the lead contact, Dr. Xiaoyan Qiu ([email protected]).
Materials availability
Anti-SIA-IgG antibodies or other unique reagents generated in this study will be available from the lead contact upon request with a materials transfer agreement.
Data and code availability
- •Original proteomic data generated in this study are available within the article and its Tables S5 and S6.
- •This paper does not report original code.
- •Any additional information required to reanalyze the data reported in this work paper is available from the lead contact upon request.
Acknowledgments
The authors thank Xiajuan Zou, Peking University Health Science Center (Beijing, China), for assistance in LC-MS/MS analysis. This work was supported by the following research grants: 10.13039/501100001809National Natural Science Foundation of China, 82030044 and 82341227 to X.Q. and 82302076 to M.C.; Key R&D Program Disruptive Technology Innovation Projects of the 10.13039/501100002855Ministry of Science and Technology (China) of China, 2022YFF1501800 to X.Q.; Fundamental Research Funds for the Central Universities, 3332024199 to M.C.; National High Level Hospital Clinical Research Funding, 2022-PUMCH-D-001 to Y.Z.; CAMS Innovation Fund for Medical Sciences (CIFMS), 2024-I2M-ZD-001 to Y.Z.; Beijing Natural Science Foundation, 7224340 to M.C.; and Peking Union Medical College Hospital Talent Cultivation Program Category D, UHB12625 to M.C. It was also supported by the Milstein Medical Asian American Partnership (MMAAP) foundation.
Author contributions
Conceptualization, S.Z., M.C., X. Huang, J.H., and X.Q.; resources, M.C., X. Han, X.L., Q. Liao, Y.Z., and X.Q.; data curation, S.Z.; formal analysis, S.Z.; validation, S.Z.; investigation, S.Z., M.C., X.F., Q. Liu, and J.B.; visualization, S.Z., R.X., and W.X.; methodology, S.Z., M.C., X. Huang, and W.X.; writing – original draft, S.Z.; writing – review & editing, S.Z., M.C., Y.Z., and X.Q.; supervision, M.C., J.H., Q. Liao, Y.Z., and X.Q.; project administration, S.Z., M.C., and X.Q.; funding acquisition, M.C., Y.Z., and X.Q.
Declaration of interests
The authors declare no competing interests.
STAR★Methods
Key resources table
REAGENT or RESOURCESOURCEIDENTIFIERAntibodiesAnti-human CD68 antibodyZSGB-BioCat# ZM-0060; RRID: AB_2904190Anti-human CD163 antibodyZSGB-BioCat# ZM-0428; RRID: AB_3714707Anti-human CD44v6 antibodyZSGB-BioCat# ZM-0052; RRID: N/AAnti-human CD163-PE/Cy7(Clone: RM3/1)BiolegendCat# 326513; RRID: AB_2565244Anti-human CD14 FITC (Clone: 61D3)Thermo Fisher ScientificCat# 11-0149-42; RRID: AB_10597597Anti-human PD-L1 APC (Clone: 29E.2A3)BiolegendCat# 329707; RRID: AB_940358Anti-human Siglec-1 PerCP/Cy5.5 (Clone: 7-239)BiolegendCat# 346019; RRID: AB_2860903Anti-human Siglec-3 (CD33) BV510 (Clone: P67.6)BiolegendCat# 366609; RRID: AB_2566402Anti-human Siglec-7 PE (Clone: 194211)R&D systemCat# FAB11381P; RRID: AB_3645881Anti-human Siglec-9 AF405 (Clone: 191240)R&D systemCat# FAB1139V; RRID: AB_3645893Anti-human Siglec-10 APC (Clone: 5G6)BiolegendCat# 347606; RRID: AB_11203899Anti-human Siglec-10 PE (Clone: 5G6)BiolegendCat# 347603; RRID: AB_2270417Anti-mouse PD-L1 PE/Cy7 (Clone: 10F.9G2)BiolegendCat# 124313; RRID: AB_10639934Anti-mouse CD-45 PerCP/Cy5.5 (Clone:104)Thermo Fisher ScientificCat# 45-0454-80; RRID: AB_953592Anti-mouse CD-45 APC (Clone: 30-F11)Thermo Fisher ScientificCat# 17-0451-82; RRID: AB_469392Anti-mouse CD-45 FITC (Clone: 30-F11)BiolegendCat# 103108, RRID: AB_312973Anti- mouse/human CD11b FITC (Clone: M1/70)NovusCat# NB100-77375, RRID: AB_1083233Anti-mouse/human CD11b APC/Cyanine7 (Clone: M1/70)BioLegendCat# 101226, RRID: AB_830642Anti-mouse Ly6G PE (Clone: HK1.4)Thermo Fisher ScientificCat# 17-5932-82; RRID: AB_1724153Anti-mouse Ly6C APC (Clone: RB6-8C5)Thermo Fisher ScientificCat# 12-5931-82; RRID: AB_466045Anti-mouse F4/80 PE/Cy7 (Clone: BM8)Thermo Fisher ScientificCat# 25-4801-82; RRID: AB_469653Anti-mouse F4/80 PE (Clone: W20065B)BioLegendCat# 111603, RRID: AB_3082990Anti-mouse Siglec-1 PE (Clone: 3D6.112)BiolegendCat# 142403; RRID: AB_10915470Anti-mouse Siglec-E FITC (Clone: M1304A01)BiolegendCat# 677111; RRID: AB_2721583Anti-mouse Siglec-G APC (Clone:SH2.1)Thermo Fisher ScientificCat# 17-5833-80; RRID: AB_2573229Anti-mouse CD4 PerCP/Cy5.5 (Clone:RM4-5)Thermo Fisher ScientificCat# 45-0042-82; RRID: AB_1107001Anti-mouse CD4 APC/Cyanine7 (Clone:RM4-5)BiolegendCat# 100526, RRID: AB_312727Anti-mouse CD8 PE/Cy7 (Clone:53-6.7)Thermo Fisher ScientificCat# 25-0081-81; RRID: AB_469583Anti-mouse CD8 Brilliant Violet 510 (Clone:53-6.7)BioLegendCat# 100752, RRID: AB_2563057Anti-mouse CD206 APC (Clone: C068C2)BiolegendCat# 141707; RRID: AB_10896057APC anti-mouse CD279 (PD-1) (Clone: RMP1-30)BioLegendCat# 109112, RRID: AB_10612938PE anti-mouse FOXP3 Antibody (Clone: MF-14)BiolegendCat# 126403, RRID: AB_1089118Anti-mouse CD366 (Tim-3) PE/Cyanine7 (Clone: RMT3-23)BiolegendCat# 119716, RRID: AB_2571933Anti-human IgG Fab antibodySolarbioCat# SE206Anti-human IgG1 antibodyAbcamCat# ab108969; RRID: AB_3665955Anti-human IgG2 antibodyAbcamCat# ab134050; RRID: AB_3665956Anti-human IgG4 antibodyAbcamCat# ab238320; RRID: N/AAnti-human IgG3 antibodyAbcamCat# ab169323; RRID: N/AAnti-M-CSF antibodyAbcamCat# ab99178; RRID: AB_11143555Anti-TGF-β1 antibodyAbcamCat# ab215715; RRID: AB_2893156Ant-ARG1 antibodyProteintechCat# 16001-1-AP; RRID: AB_2289842Anti-IL-18 antibodyCell SignalingCat# 67775S; RRID: AB_3105845Anti-IL-1β antibodyAbcamCat# ab9722; RRID: AB_308765Anti-CD133 antibodyProteintechCat# 66666-1-Ig; RRID: AB_2801586Anti-CD44 antibodyCell SignalingCat# 37259; RRID: AB_2750879Anti-OCT4 antibodyCell SignalingCat# 2890S; RRID: AB_2167725Anti-His antibodyProteintechCat# 66005-1-Ig; RRID: AB_11232599Anti-GFP antibodyAbmartCat# M20004F; RRID: AB_2619674Anti-ST6GAL1 antibodyR&D systemCat# AF5924; RRID: AB_2044637Anti-SHP-1 antibodyAbcamCat# ab32559; RRID: AB_777912Anti-p-SHP-1 antibodyCell SignalingCat# 8849; RRID: AB_11141050Anti-Mannose Receptor (CD206) antibodyAbcamCat# ab64693, RRID: AB_1523910Anti-human Siglec-1 antibodyR&D systemCat# AF5197; RRID: AB_2239277Anti-human Siglec-7 antibodyR&D systemCat# AF1138; RRID: AB_2189416Anti-human Siglec-9 antibodyR&D systemCat# MAB1139; RRID: AB_2270258Anti-human Siglec-10 antibodyR&D systemCat# AF2130; RRID: AB_355164anti-mouse CD3 antibodyBioLegendCat# 317303; RRID: AB_571924anti-mouse CD28 antibodyBioLegendCat# 302913; RRID: AB_314315FITC-conjugated goat anti-mouse IgG (H+L)ZSGB-BioCat# ZF-0312; RRID: AB_2716306TRITC-conjugated goat anti-mouse IgG (H+L)ZSGB-BioCat# ZF-0313; RRID: AB_2571577FITC-conjugated goat anti-rabbit IgG (H+L)ZSGB-BioCat# ZF-0311; RRID: AB_2571576TRITC-conjugated goat anti-rabbit IgG (H+L)ZSGB-BioCat# ZF-0316; RRID: AB_2728778Goat anti-Rabbit IgG (H+L) Secondary Antibody, HRPThermo Fisher ScientificCat# 31460; RRID: AB_228341Goat anti-Mouse IgG (H+L) Secondary Antibody, HRPThermo Fisher ScientificCat# 31430; RRID: AB_228307Anti-Siglec-9 monoclonal antibodyThis paperClone: 1G12anti-SIA-IgG monoclonal antibody (α-SIA-IgG)Generated from RP215N/A (see Gregory18 and Tang et al.20)humanized anti-SIA-IgG mAb (HASA)Zencore BiologicsN/ABacterial and virus strainsE.coli DH5α Competent CellsSolarbioCat# C1100-10Biological samplesFormalin-fixed paraffin-embedded (FFPE) PDAC tissue samplesPeking Union Medical College HospitalN/APDAC tissuesPeking Union Medical College HospitalN/AHuman peripheral blood from PDAC patients and healthy donorsPeking Union Medical College HospitalN/AChemicals, peptides, and recombinant proteinsTrisGenviewCat# BT350EDTA-2Na·2H_2_OHeownsCat# E−84000Tween-20SolarbioCat# T8220ArginineHARVEYBIOCat# AA0993KanamycinSolarbioCat# K8020AmpicillinSolarbioCat# A8180Lymphocyte Separation MediaBiotoppedCat# ZK100222PBSSolarbioCat# P1010LPSSolarbioCat# L8880PMASigmaCat# P8139BSASolarbioA8020Mouse IgGSolarbioCat# SP031Human IFN-γPeprotechCat# 300-02-20Human M-CSFPeprotechCat# 300-25-2Mouse M-CSFPeprotechCat# 315-03-5Human EGFPeprotechCat# AF-100-15Human FGF-basic (FGF-2/bFGF)PeprotechCat# 100-18BB27GibcoCat# 17504044Collagenase IVSolarbioCat# C8160DNase ISolarbioCat# D8070PNGase FNEBCat# P0704α 2-3,6,8 NeuraminidaseNEBCat# P0720α 2-3 NeuraminidaseNEBCat# P0743Human Siglec-9, His TagACRO BiosystemsCat# SI9-H52H4CFSEThermo Fisher ScientificCat# C34554pHrodo Red succinimidyl esterThermo Fisher ScientificCat# P36600HoechstSolarbioCat# C0031TRIzol ReagentLife TechnologyCat# 155960181×Hieff qPCR SYBR Green Master MixYeasen BiotechnologyCat# 11203ES08PEI ReagentPolysciencesCat# 23966Fetal Bovine SerumHyCloneCat# SH30396.02RPMI 1640GibcoCat# 31800-022DMEMGibcoCat# 12100061Cell lysatesLeageneCat# PS001310mM sodium acetate pH4.5CytivaCat# BR-1003-5010mM sodium acetate pH5.5CytivaCat# BR-1003-5210mM Glycine-HClCytivaCat# BR-1003-5650mM NaOHCytivaCat# BR-1003-58PBS-P+ Buffer 10×CytivaCat# 28995084Critical commercial assaysOpal 6-Plex Manual Detection KitAkoya BiosciencesCat# NEL811001KTDAB Detection KitGene TechCat# GK600705Protein G Sepharose 4 Fast Flow columnGE HealthcareCat# 17-0618-01CNBr activated Sepharose 4B71-7086-00 AFGE HealthcareCat# 71-7086-00 AFEasySep Human CD14 Positive Selection Kit IIStemCell TechnologiesCat# 17858RevertAid First Stand cDNA Synthesis KitThermo Fisher ScientificCat# K1622Dynabeads His-Tag Isolation and PulldownThermo Fisher ScientificCat# 10103DClophosome-A and Control LiposomesFormuMaxCat# F70101C-NC-10Amino Conjugation KitCytivaCat# BR-1000-50EZ-Link Sulfo-NHS-LC-Biotinylation KitThermo Fisher ScientificCat# 21435Lipofectamine 3000Thermo Fisher ScientificCat# L-3000015LEGEND MAX Mouse IL-10 ELISA KitBiolegendCat# 431417Human/Mouse TGF beta-1 Uncoated ELISA KitThermo Fisher ScientificCat# 88-8350-88Mouse IL-6 Uncoated ELISA KitThermo Fisher ScientificCat# 88-7064-88Deposited dataPDAC scRNA-SeqPeng et al.35GSA: CRA001160RNA seq of KPC treated with TGF-β1Hou et al.39GSE149130Original proteomic data Tables S5 and S6This paperTables S5 and S6Experimental models: Cell lines293FTCell Resource Center, Peking Union Medical CollegeRRID: CVCL_6911hTERT-HPNEATCCRRID: CVCL_C466THP-1ATCCRRID: CVCL_0006T3M-4Prof. Jörg Kleeff (Martin-LutherUniversity Halle-Wittenberg)RRID: CVCL_4056PANC-1ATCCRRID: CVCL_0480SW1990ATCCRRID: CVCL_1723MIA PaCa-2ATCCRRID: CVCL_0428Pa18CATCCRRID: CVCL_1637PK-9Prof. Jun Yu (Tianjin Medical University)RRID: CVCL_D721KPCSpontaneous pancreatic cancer KPC (LSL-Kras^G12D/+^; LSL-Trp53^R172H/+^; Pdx-1-Cre) miceN/AExperimental models: Organisms/strainsKPC (LSL-Kras^G12D/+^; LSL-Trp53^R172H/+^; Pdx-1-Cre) miceThe Jackson LaboratoryRRID: IMSR_JAX:032429NOD-SCIDVital River Laboratory Animal Technology CompanyRRID: IMSR_CRL:394C57BL/6Vital River Laboratory Animal Technology CompanyRRID: IMSR_JAX:000664NudeVital River Laboratory Animal Technology CompanyRRID: IMSR_CRL:088OligonucleotidesPrimers for quantitative real-time (qRT-)PCR, see Table S3 in supplemental informationThis paperTable S3siIGHG-1 (GGUGGACAAGACAGUUGAG)This paperN/AsiIGHG-2 (AGUGCAAGGUCUCCAACAA)This paperN/AsiST6GAL1-1 (CAGCCAACUUCCAACAAGAdTdT)This paperN/AsiST6GAL1-2 (ACUCAGAUAUCCCAAAGUG)This paperN/ANC (GUAUGACA ACAGCCUCAAGTT)This paperN/ARecombinant DNA6His-Siglec-3 (CD33)-ECD in pCMV3 plasmidThis paperN/AGFP-V_H_5-51/D_H_3-16/J_H_4-IGHG1 (N162Q) in pEGFP-C3 plasmidThis paperN/AGFP-V_H_5-51/D_H_3-16/J_H_4-IGHG1 in pEGFP-C3 plasmidThis paperN/AFlag- V_H_5-51/D_H_3-16/J_H_4-IGHG1 in pcDNA3.1 plasmidThis paperN/AFlag- V_H_5-51/D_H_3-16/J_H_4-IGHG1 (N162Q) in pcDNA3.1 plasmidThis paperN/AMyc-Siglec-E in pcDNA3.1 plasmidThis paperN/AMyc-Siglec-G in pcDNA3.1 plasmidThis paperN/ASoftware and algorithmsFlowJo (v10)N/ARRID:SCR_008520R (v4.2.1)https://www.r-project.org/RRID:SCR_001905GraphPad Prism 8.1N/ARRID:SCR_002798X-tile 3.6.1Camp et al.51RRID:SCR_005602GSEA_4.3.2https://www.gsea-msigdb.org/RRID:SCR_003199ImageJN/ARRID:SCR_003070OtherStreptavidin -PEBiossCat# bs-0437P-PEDynaMag-2Thermo Fisher ScientificCat# 12321D
Experimental model and study participant details
Mice
NOD-SCID (RRID: IMSR_CRL:394) and C57BL/6 (RRID: IMSR_JAX:000664) female mice were purchased from Beijing Vital River Laboratory Animal Technology Company. Nude (RRID: IMSR_CRL:088) female mice were obtained from the Beijing Vital River Laboratory Animal Technology Company or Department of Laboratory Animal Science at the Peking University Health Science Center. The animal experiments were approved by the Animal Care and Use Committee of Peking University (No: LA2020262). Purchased mice were housed in a specific pathogen-free facility with four or five mice per cage at the Peking University Health Science Center. And mice were maintained on a fixed day-night lighting cycle (12:12 light: dark) with AD libitum access to food and water. Chopped corncobs were used as bedding.
Experimental design of mouse PDAC model
For each animal, two different investigators were involved as follows: the first investigator assigned animals to treatment according to randomization, and this investigator was the only person who knew the treatment group assignment. The second investigator was responsible for recording observations. The maximum size of the tumors allowed to grow in the mice before euthanasia was 2000 mm^3^.
Clinical samples
Formalin-fixed, paraffin-embedded (FFPE) PDAC tissue samples, tissues from PDAC patients and peripheral blood from healthy donors, were collected from Peking Union Medical College Hospital. These collections were approved by the Medical Ethics Committee at Peking Union Medical College Hospital (No: S-K623, K5119). Neither the patients nor the public were directly involved in this research.
Cell lines and primary macrophage culture
The cell lines 293FT, hTERT-HPNE, THP-1, T3M-4, PANC-1, SW1990 MIA PaCa-2 and Pa18C were obtained from the ATCC and maintained by the Peking University Center of Human Disease Genomics (Beijing, China). The human PDAC cell line PK-9 was kindly provided by Dr. Jun Yu (Tianjin Medical University), The KPC cell line was derived from spontaneous pancreatic cancer KPC (LSL-Kras^G12D/+^; LSL-Trp53^R172H/+^; Pdx-1-Cre) mice. T3M-4 and THP-1 cells were grown in 90% RPMI1640 (Gibco), while PANC-1, SW1990, HPNE, PK-9, Pa18C, and KPC cells were grown in 90% DMEM (Gibco).
THP-1 cells were stimulated with 100 ng/mL PMA (Sigma) to generate THP-1-derived macrophages. Human CD14^+^ monocytes or murine bone marrow cells were cultured with M-CSF to generate human or mouse primary macrophages.
Method details
Primary macrophage culture
Human primary macrophages are generated from the differentiation of CD14^+^ monocytes isolated from peripheral blood mononuclear cells (PBMCs). CD14^+^ monocytes were then isolated from the PBMCs with a Human CD14 Positive Selection Kit II (STEMCELL), achieving a purity of over 90%. To generate M0 macrophages, the isolated CD14^+^ monocytes were cultured for 6 days in RPMI1640 (Gibco) supplemented with 10% fetal bovine serum (FBS) and 20 ng/mL human M-CSF (Peprotech). For M1 polarization, the M0 macrophages were subsequently stimulated for 2 days with 100 ng/mL LPS (Solarbio) and 50 ng/mL human IFN-γ (Peprotech).
For macrophage conditioned cultures, when T3M-4 cells reached 80–90% confluency, these cancer cells were gently removed by trypsin-free digestion, and macrophages were subsequently seeded into the same culture plates, which contained the cancer cell-derived extracellular matrix (ECM).
Tumor sphere formation assay
T3M-4 cells were seeded in ultra-low attachment plates and cultured in serum-free RPMI1640 (Gibco) supplemented with 2% B27 (Gibco) supplement, 20 ng/mL EGF (Peprotech), and 20 ng/mL bFGF (Peprotech) in a humidified incubator with 5% CO_2_ at 37°C. Tumor microspheres were formed within 7–10 days.
Immunostaining
For immunofluorescence staining, cells were fixed with fresh acetone at room temperature for 10 min, followed by blocking with 5% BSA in PBS for 30 min. The cells were then incubated with primary antibodies at 4°C overnight. Subsequently, the cells were incubated with FITC or TRITC conjugated fluorescent secondary antibodies (ZSGB-Bio) at room temperature for 1 h. Cell nuclei were counterstained with Hoechst (Solarbio). Images were acquired using a fluorescence microscope.
For IHC on formalin-fixed paraffin-embedded (FFPE) human PDAC tissue sections, antigen retrieval was performed using Tris-EDTA buffer (pH 9.0). Endogenous peroxidases were quenched by incubation with 3% H_2_O_2_ for 10 min, and nonspecific binding sites were blocked with 10% goat serum (ZSGB-Bio) for 30 min at room temperature. The sections were then incubated with the primary antibody at 4°C overnight, followed by incubation with horseradish peroxidase (HRP)-labeled secondary antibodies for 20 min at room temperature. Signal detection was performed with a DAB chromogenic substrate according to the manufacturer’s instructions.
For multicolor immunofluorescence staining, sections were blocked and incubated with the primary antibody as described for IHC. Subsequent staining was performed using the Opal 6-Plex Manual Detection Kit (Akoya Biosciences), according to the manufacturer’s instructions.
IHC scoring and classification methods
Each tissue microarray (TMA) core was divided into four equal regions, and SIA-IgG expression was quantified using an H-score (intensity [0–3] × percentage of positive cells [0–100%]), with the average score calculated across the four regions. M2-TAM infiltration was assessed by counting CD163^+^ cells, with the total number per TMA core used as the infiltration score. Optimal cutoffs (221 for SIA-IgG, 392 cells/core for M2-TAMs) were determined using X-tile software (v3.6.1, Yale University) based on prognostic stratification.
Flow cytometric analysis and sorting
Single-cell suspensions were prepared from cell lines, primary human/mouse macrophages and pancreatic cancer tissues of mouse models. For adherent cell lines and primary macrophages, cells were obtained using a sterile cell scraper or trypsin-free digestive solution. For mouse pancreatic cancer tissues, the samples were first minced and then digested into a single-cell suspension using collagenase IV and DNase I at 37°C for 1 h. Cells were subsequently incubated with 5% FBS in PBS on ice for 30 min to reduce nonspecific binding, followed by a 40 min incubation at 4°C in the dark with fluorochrome-conjugated antibodies. Flow cytometric analysis was performed on a FACSVerse or FACSCanto Plus flow cytometer (BD Biosciences), and the data were analyzed with FlowJo v10.
For flow cytometric sorting, single-cell suspensions were prepared, blocked, and stained as described above, following aseptic procedures. Cells were sorted on a FACS Aria II (BD Biosciences) and cultured in a humidified atmosphere with 5% CO_2_ at 37°C.
ELISA
According to the manufacturer’s protocols, the levels of IL-1β, TGF-β1, and IL-10 in the cell culture supernatant were measured using the respective ELISA kits: Mouse IL-6 Uncoated ELISA Kit, Human/Mouse TGF beta-1 Uncoated ELISA Kit, and LEGEND MAX Mouse IL-10 ELISA Kit. The absorbance at 450 nm was recorded with a spectrophotometer.
The binding affinity between SIA-IgG and Siglec-9 was evaluated using a direct ELISA. Briefly, microplates were coated with Siglec-9 at 4°C overnight. After blocking nonspecific binding sites with 5% BSA, the wells were incubated with serial dilutions of either native or neuraminidase-treated SIA-IgG for 1 h at room temperature. Following a washing step, bound SIA-IgG was detected using an HRP-conjugated anti-human IgG antibody. The reaction was developed with TMB substrate and terminated with 2 M H_2_SO_4_, after which the absorbance was measured at 450 nm with a spectrophotometer.
siRNA transfection
Transient transfection of siRNAs was performed using Lipofectamine 3000 (catalog no. L-3000015, Thermo Fisher Scientific) according to the manufacturer’s instructions. Specific siIGHG-1, siIGHG-2, siST6GAL1-1, siST6GAL1-2 or NC targeting the IGHG or ST6GAL1 were used to knock down IgG or ST6GAL1 expression.
Coimmunoprecipitation analysis
For coimmunoprecipitation analysis, the human 6His-Siglec-9 in the pcDNA 3.1, 6His-Siglec-3 (CD33) in the pCMV3, human SIA-IgG pEGFP-C3 or a control vector were transfected into 293FT cells using PEI Reagent (Polysciences, Inc). The 293FT cells were harvested and lysed with Co-IP cell lysate buffer (50 mM Tris-HCl, pH 7.5; 150 mM NaCl; and 0.05% NP-40). Cell lysates were incubated with Dynabeads His-Tag Isolation and Pulldown (Thermo Fisher Scientific) according to the manufacturer’s instructions. The beads were then captured on a magnetic rack and washed with washing buffer (50 mM Tris-HCl, pH 7.5; 150 mM NaCl, 0.05% NP-40). Finally, the bound proteins were eluted by boiling the beads in loading buffer at 99°C for 10 min and analyzed by SDS-PAGE or Western blot.
SDS-PAGE and western blot analysis
For SDS-PAGE analysis, Samples were denatured at 99°C and then separated by 12.5% SDS-PAGE, the separated proteins were visualized by staining the gel with Coomassie brilliant blue.
For Western blot analysis, cellular proteins and extracellular matrix (ECM) proteins were obtained from cell lysates (Leagene) and TSD lysates, respectively. The proteins were separated by SDS-PAGE and then transferred onto an NC membrane. Membranes were blocked with TBST containing 5% milk for 1 h at room temperature and incubated with primary antibodies at 4°C overnight, followed by incubation with the HRP-coupled secondary antibodies for 1 h at room temperature.
Purification of SIA-IgG
PDAC tissue-derived proteins were centrifuged at 13,000 rpm for 30 min at 4°C to remove debris and then diluted with PBS. The protein sample was applied to a Protein G Sepharose 4 Fast Flow column (GE Healthcare), and total IgG was obtained by washing and eluting following the manufacturer’s recommendations. The α-SIA-IgG was conjugated to CNBr activated Sepharose 4B (GE Healthcare) to prepare the α-SIA-IgG-coupled column. Total IgG was incubated with the α-SIA-IgG-coupled column, and non-SIA-IgG was obtained from flow-through fraction, and the SIA-IgG was subsequently obtained by washing with PBS (pH 7.4) and eluted with 0.5M arginine (pH 3.8).
Cell membrane binding assay
Single-cell suspensions of macrophages were prepared for membrane binding analysis. SIA-IgG and non-SIA-IgG were biotinylated using an EZ-Link Sulfo-NHS-LC-LC-Biotin Kit according to the manufacturer’s instructions. To block potential Fc receptor-mediated binding, macrophages were preincubated with excess Human TruStain FcX Fc Receptor Blocking Solution prior to incubation with SIA-IgG. Then biotinylated SIA-IgG or non-SIA-IgG was diluted to 20 μg/mL in PBS containing 2% FBS and incubated with macrophage suspensions at 4°C for 30 min to allow specific binding while minimizing internalization. Cells were then fixed with 1% paraformaldehyde for 30 min at room temperature. Following fixation, cells were stained with PE-conjugated streptavidin-phycoerythrin (1:200 dilution) in the dark at room temperature for 30 min. SIA-IgG binding was assessed using flow cytometry.
Quantitative RT-PCR
Total RNA was extracted from human and mouse primary macrophage using TRIzol Reagent (Life Technology). cDNA was synthesized obtained by reverse transcription (RT) using the RevertAid First Stand cDNA Synthesis Kit (Thermo Fisher Scientific) according to the manufacturer’s instructions. Subsequently, quantitative real-time PCR (qRT-PCR) was performed using the synthesized cDNA as a template with 1×Hieff qPCR SYBR Green Master Mix (Yeasen Biotechnology) and gene-specific primers (working concentration: 0.2 μM; sequences are provided in Table S3). The PCR amplification protocol consisted of an denaturation at 95°C for 5min, followed by 40 cycles of 95°C for 10 s and 60°C for 30 s, using a LightCycler 480 system (Roche Applied Science). The relative expression of the target genes was normalized to the housekeeping gene GAPDH, and calculated using the2^−ΔCT^ method.
T cell proliferation assay
To assess the inhibitory effect of treated macrophages on T cell proliferation, primary mouse macrophages were cultured and pretreated with 20 μg/mL SIA-IgG or non-SIA-IgG for 48 h. Following removing SIA-IgG or non-SIA-IgG, CFSE-labeled mouse T cells were activated by adding 3 μg/mL anti-mouse CD3 antibody (BioLegend) and 1 μg/mL anti-mouse CD28 antibody (BioLegend) T cell proliferation was assessed using flow cytometry.
Macrophage phagocytosis assay
To investigate macrophage phagocytosis, target cells and human primary macrophages were seeded together at a ratio of 20,000–30,000 target cells to 10,000 macrophages in 96-well plates for co-culture. The co-cultured cells were maintained in serum-free RPMI1640 (Gibco) in a humidified atmosphere with 5% CO_2_ at 37°C for 48 h. Prior to co-culture, T3M-4 cells were harvested and fluorescently labeled with CFSE or pHrodo Red succinimidyl ester (Thermo Fisher Scientific) according to the manufacturer’s instructions. After co-culture, macrophage phagocytosis was assessed by both flow cytometry and fluorescence microscopy. For assays involving antibody treatments, specific antibodies or their corresponding isotype controls were added at concentration of 20 μg/mL. Phagocytosis of pHrodo Red-labeled target cells was evaluated using fluorescence microscope, whereas phagocytosis of CSFE-labeled target cells was quantified by flow cytometry, measuring the percentage of CD14^+^ CFSE^+^ cell within the CD14^+^ cell population.
Surface plasmon resonance
Binding affinities were determined using a Biacore 8K system (Cytiva). α-SIA-IgG, HASA, or recombinant Siglec-9 was immobilized on CM5 sensor chips via amine coupling in sodium acetate buffer (pH 4.5 for α-SIA-IgG/HASA, pH 5.5 for Siglec-9, Cytiva). The surfaces were blocked with ethanolamine-HCl (Eth-HCl) to deactivate remaining reactive groups.
For kinetic measurements, SIA-IgG at varying concentrations was injected using 1× PBS-P buffer (Cytiva) as running buffer. Following each binding cycle, the sensor surfaces were regenerated with glycine-HCl (pH 2.5) for α-SIA-IgG/HASA or 50 mM NaOH for Siglec-9. Binding kinetics and affinity constants were calculated using Biacore 8K evaluation software.
PDAC mouse models
For PDAC cell line-derived xenograft (CDX) tumor models, 1 × 10^6^ KPC cells were inoculated into the armpits of 6–8 weeks old C57BL/6 female mice, or 3 × 10^6^ T3M-4 cells were inoculated into the armpits of 6–8 weeks old nude female mice. When the tumor reached approximately 20 mm^3^ in size, α-SIA-IgG or the humanized anti-SIA-IgG mAb (HASA) (5 mg/kg, every two days) was administered via tail vein injection, mIgG (mouse IgG) served as isotype control for α-SIA-IgG and IVIG (Intravenous immunoglobulin) served as isotype control for HASA.
For orthotopic xenograft tumor models derived from PDAC cell lines, mice were anesthetized with 1% sodium pentobarbital, the abdominal cavity was opened. A total of 1 × 10^6^ KPC cells or 2 × 10^6^ T3M-4 cells were inoculated into the tail of the pancreas in 6–8 weeks old female C57BL/6 mice or nude mice. On the third day after tumor inoculation, α-SIA-IgG or HASA or (5 mg/kg, every two days) was administered via tail vein injection, mIgG and IVIG were used as the isotype controls for α-SIA-IgG and HASA, respectively.
For PDX tumor models, PDX originating from patients with PDAC were purchased from Beijing Vitalstar Biotechnology Company. The tumors were sectioned into 2 × 2 × 2mm^3^ pieces and implanted subcutaneously into the armpits of 6–8 weeks old NOD-SCID female mice. When the tumor reached approximately 30 mm^3^ in size, α-SIA-IgG, HASA or mIgG (5 mg/kg, twice weekly) was administered via tail vein injection, with mIgG serving as the isotype control for α-SIA-IgG.
For tumor-burdened macrophage deletion in mice, 3 × 10^6^ T3M-4 cells were inoculated into the armpits of 6–8 weeks old nude female mice. According to the manufacturer’s instructions for Clodronate Liposomes (Clophosome), 200 μL of clodronate liposomes or control liposomes (FormuMax) were injected intraperitoneally, with reinjection every three days. When the tumor reached approximately 20 mm^3^ in size, HASA or IVIG (5 mg/kg, every two days) was administered via tail vein injection, using IVIG as the isotype control for HASA.
Single-cell RNA-seq
Single-cell RNA sequencing (scRNA-seq) was performed at the laboratory of NovelBio Co., Ltd. Orthotopic tumors derived from KPC cells were excised, mechanically minced into approximately 1 mm^3^ fragments on ice, and enzymatically digested using a Tumor Dissociation Kit for 45 min at 37°C to obtain a single-cell suspension.
Transcriptomic data were captured using the NovelCyto Single-Cell Analysis System, and libraries were constructed following the NovelCyto single-cell whole-transcriptome amplification (WTA) protocol, which includes random priming and extension, amplification PCR, and WTA library index PCR. Library concentration and quality were assessed using an Agilent 4200 Bioanalyzer with a High Sensitivity DNA chip and a Qubit High Sensitivity DNA assay Kit (Thermo Fisher Scientific). All libraries were sequenced on a DNBSEQ-T7 platform (MGI, Shenzhen, China) in 150 bp paired-end mode.
scRNA-seq data were analyzed by the bioinformatics team at NovelBio using the NovelBrain Cloud Analysis Platform. Raw sequencing reads were processed with fastp (with default parameters) to remove adapter sequences and low-quality reads.52 Clean reads were then aligned to the mouse reference genome (mm10, integrated annotation version 100) using STARsolo (v2.7.10a) for gene expression quantification.
Quantification and statistical analysis
Single-cell transcriptomic data were analyzed using the R (v4.2.1) (RRID:SCR_001905). Quantification of M2 macrophages was performed using ImageJ (RRID:SCR_003070). All data were analyzed using GraphPad Prism 8.1 software (RRID:SCR_002798). Student’s two tailed t test was used to compare differences between groups, with data in figures expressed as mean ± standard deviation (SD). Immune infiltration was analyzed using ImmuCellAI (ImmuCellAI (wchscu.cn)). The correlation between SIA-IgG expression, TAM infiltration and clinicopathological features was analyzed by chi-square test. Gene set enrichment analysis (GSEA) was performed with GSEA software (RRID:SCR_003199). The cutoff value in the Kaplan-Meier (KM) survival analysis was determined using X-tile 3.6.1 (Yale University School of Medicine) (RRID:SCR_005602).51 A two-sided p < 0.05 was considered statistically significant, while p ≥ 0.05 was considered not significant (ns).
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Halbrook C.J.Lyssiotis C.A.Pasca Di Magliano M.Maitra A.Pancreatic cancer: Advances and challenges Cell 18620231729175410.1016/j.cell.2023.02.01437059070 PMC 10182830 · doi ↗ · pubmed ↗
- 2Han B.Zheng R.Zeng H.Wang S.Sun K.Chen R.Li L.Wei W.He J.Cancer incidence and mortality in China, 2022 J. Natl. Cancer Cent.42024475310.1016/j.jncc.2024.01.00639036382 PMC 11256708 · doi ↗ · pubmed ↗
- 3Mizrahi J.D.Surana R.Valle J.W.Shroff R.T.Pancreatic cancer Lancet 39520202008202010.1016/S 0140-6736(20)30974-032593337 · doi ↗ · pubmed ↗
- 4Royal R.E.Levy C.Turner K.Mathur A.Hughes M.Kammula U.S.Sherry R.M.Topalian S.L.Yang J.C.Lowy I.Rosenberg S.A.Phase 2 trial of single agent ipilimumab (Anti-CTLA-4) for locally advanced or metastatic pancreatic adenocarcinoma J. Immunother.33201082883310.1097/CJI.0b 013e 3181 eec 14c 20842054 PMC 7322622 · doi ↗ · pubmed ↗
- 5Brahmer J.R.Tykodi S.S.Chow L.Q.M.Hwu W.-J.Topalian S.L.Hwu P.Drake C.G.Camacho L.H.Kauh J.Odunsi K.Safety and Activity of Anti – PD-L 1 Antibody in Patients with Advanced Cancer N. Engl. J. Med.36620122455246510.1056/NEJ Moa 120069422658128 PMC 3563263 · doi ↗ · pubmed ↗
- 6O Reilly E.M.Oh D.Y.Dhani N.Renouf D.J.Lee M.A.Sun W.Fisher G.Hezel A.Chang S.C.Vlahovic G.Durvalumab With or Without Tremelimumab for Patients with Metastatic Pancreatic Ductal Adenocarcinoma: A Phase 2 Randomized Clinical Trial JAMA Oncol.520191431143810.1001/jamaoncol.2019.158831318392 PMC 6647002 · doi ↗ · pubmed ↗
- 7Liudahl S.M.Betts C.B.Sivagnanam S.Morales-Oyarvide V.da Silva A.Yuan C.Hwang S.Grossblatt-Wait A.Leis K.R.Larson W.Leukocyte heterogeneity in pancreatic ductal adenocarcinoma: Phenotypic and spatial features associated with clinical outcome Cancer Discov.1120212014203110.1158/2159-8290.CD-20-084133727309 PMC 8338775 · doi ↗ · pubmed ↗
- 8Vayrynen S.A.Zhang J.Yuan C.Vayrynen J.P.Costa A.D.Williams H.Morales-Oyarvide V.Lau M.C.Rubinson D.A.Dunne R.F.Composition, Spatial Characteristics, and Prognostic Significance of Myeloid Cell Infiltration in Pancreatic Cancer Clin. Cancer Res.2720211069108110.1158/1078-0432.CCR-20-314133262135 PMC 8345232 · doi ↗ · pubmed ↗