Blocking siglec-10hi tumor-associated macrophages improves anti-tumor immunity and enhances immunotherapy for hepatocellular carcinoma

Background Tumor-associated macrophages (TAMs) promote key processes in the modulation of tumor microenvironment (TME). However, the clinical significance of heterogeneous subpopulations of TAMs in hepatocellular carcinoma (HCC) remains unknown. Methods HCC tissues from Zhongshan Hospital and data from The Cancer Genome Atlas were obtained and analyzed. Immunohistochemistry and flow cytometry were performed to detect the characteristics of sialic acid-binding immunoglobulin-like lectin 10high (Siglec-10hi) TAMs and explore their impact on the TME of HCC. The effect of Siglec-10 blockade was evaluated in vitro based on fresh human tumor tissues. Results Our data revealed that Siglec-10 was abundant in a large proportion of HCC specimens and prominently distributed on macrophages. Kaplan–Meier curves and Cox regression analysis showed that intratumoral Siglec-10+ cell enrichment was associated with unfavorable prognosis in patients with HCC. Notably, multiple anti-inflammatory cytokines and inhibitory receptors were enriched in Siglec-10hi TAMs. RNA sequencing data also revealed that numerous M2-like signaling pathways were significantly upregulated in Siglec-10hi TAMs. High infiltration of Siglec-10hi TAMs was associated with impaired CD8+ T cell function in HCC. Of note, blocking Siglec-10 with the competitive binding antibody Siglec-10 Fc led to decreased expression of immunosuppressive molecules and increased the cytotoxic effects of CD8+ T cells against HCC cells. Moreover, blocking Siglec-10 promoted the anti-tumor efficacy of the programmed cell death protein 1 (PD-1) inhibitor pembrolizumab. Conclusions Siglec-10hi TAMs are associated with immune suppression in the TME, and indicate poor prognosis in patients with HCC. Targeting Siglec-10hi TAMs may serve as a promising immunotherapy approach for HCC. Supplementary Information The online version contains supplementary material available at 10.1186/s40164-021-00230-5.


Introduction
patients with HCC are diagnosed in the advanced stages, and therefore have poor prognosis [4].
Immunotherapy targeting the tumor microenvironment (TME) has revolutionized tumor treatment. The results of the CheckMate 040 and KEYNOTE-224 studies were extraordinary milestones in the battle against HCC. However, the response rate to monoclonal anti-programmed cell death protein 1 (PD-1) antibody has been disappointing due to the high heterogeneity of immune evasion mechanisms [5]. Thus, it is urgent to identify novel immune modulators to restore anti-tumor immunity in HCC.
Tumor-associated macrophages (TAMs) are the most abundant immune cells within the TME, and promote vital processes in tumor progression [6]. Increased TAM infiltration is associated with poor patient outcomes in HCC [7,8]. TAMs can generally be classified into proinflammatory (M1) or anti-inflammatory (M2) phenotype depending on diverse environmental stimulus. M1 TAMs act as prominent antigen-presenting cells and play critical roles in anti-tumor activity, whereas M2 TAMs are featured by higher production of anti-inflammatory cytokines and active metabolic pathways, which suppress the adaptive immune response [9]. Actually, M1 and M2 polarization are two extremes of a spectrum of activation states. TAMs often exhibit a mixed M1/M2 phenotype, with their exact point on the scale based on the specific blend of cytokines present in the TME [9]. Targeting TAMs to remodel the TME is a hotspot in tumor immunotherapy. Blocking anti-inflammatory receptors on specific macrophage subsets can reactivate anti-tumor immunity [10][11][12].
Sialic acid-binding immunoglobulin-like lectin 10 (Siglec-10) is widely expressed on subsets of human leukocytes, and plays an important part in distinguishing between self and non-self in the immune system [13]. As an immune checkpoint, Siglec-10 is closely involved in the regulation of macrophage phagocytosis [14], B-cell tolerance [15], impaired T-cell activation [16], and pathogen immune evasion [17]. Of note, multiple studies have shown the clear distribution of Siglec-10 on macrophages [13,14]. Thus, we hypothesized that Siglec-10 hi macrophages may promote immune evasion in HCC.
In this study, we found that Siglec-10, which was associated with poor overall survival (OS) and recurrence-free survival (RFS) in patients with HCC, was predominately expressed on macrophages. Furthermore, we demonstrated that intratumoral Siglec-10 hi TAM infiltration was positively associated with CD8 + T cell inactivation. Notably, inhibition of Siglec-10 restrained the secretion of anti-inflammatory cytokines from macrophages and suppressed the expression of inhibitory receptors on CD8 + T cells, resulting in restored cytotoxic activities against tumor cells. Our study identifies targeting Siglec-10 hi TAMs as a promising therapeutic option for HCC.

Patients and follow-up
We retrospectively evaluated 246 patients with HCC who underwent hepatectomy from 2008 to 2010. The exclusion criteria were: pathological diagnosis other than HCC or combined with other pathological types, distant metastatic disease, or unavailability of tumor tissues or follow-up information. None of the patients enrolled in this study had received neoadjuvant therapy before surgery. Two patients with mixed HCC/intrahepatic cholangiocarcinoma (ICC) and five patients with unavailable follow-up data were excluded. Four specimens were lost on tissue microarray (TMA) when immunohistochemistry (IHC) was performed. Finally, 235 eligible HCC patients were randomly divided into discovery set (n = 115) and validation set (n = 120). Patients' clinical characteristics are shown in Additional file 1: Table S1. A detailed flow chart of patient selection is shown in Additional file 1: Fig. S1.
The clinical tumor stage was determined according to the American Joint Committee on Cancer/International Union Against Cancer (AJCC/UICC) Tumor-Node-Metastasis (TNM) staging system 8th edition. All follow-up data were collected from the date of surgery to December 2018. OS was defined as the time period between surgery and death or last follow-up. RFS was defined as the time period between surgery and the first documented recurrence or death, whichever occurred first. This study was approved by the Clinical Research Ethics Committee of Zhongshan Hospital, Fudan University (Shanghai, China). An informed consent form was signed by every patient.

IHC and immunofluorescence
IHC and immunofluorescence (IF) staining were performed on TMAs constructed with formalin-fixed and paraffin-embedded HCC specimens. IHC staining was performed according to a previously described protocol [11]. In brief, TMAs were incubated with primary antibodies overnight at 4 °C. The staining was carried out with DAB reagent. For IF staining, the sections were incubated with two primary antibodies at 4 °C overnight, followed by incubation with FITC-and TRITCconjugated secondary antibodies at 37 °C for 2 h. Finally, the slides were mounted with Antifade Mounting Solution containing DAPI. The details of the antibodies are listed in Additional file 1: Table S2. For negative controls, primary antibody was omitted. All stained tissues were evaluated under the Leica DM6000 B microscope (Leica Microsystems, Wetzlar, Germany) independently by two pathologists who were blinded to the patients' characteristics.

Fresh samples and flow cytometry
Seventy-seven fresh human specimens were collected from HCC patients, who underwent liver resection at Zhongshan Hospital. Single-cell suspension was prepared as previously described [18]. After removing blood and necrotic tissues, samples were digested in RPMI Medium 1640 with 1 mg/mL Collagenase (Sigma Aldrich) and 0.1 mg/mL DNase I (Roche) for 30 min at 37 °C. Then, the tumor suspension was filtered through a 70 mm cell strainer (BD Falcon) and washed with Stain Buffer (BD Biosciences). For intracellular cytokine detection, cells were stimulated for 4 h with phorbol myristate acetate (50 ng/mL) and ionomycin (1 µg/mL) in the presence of BD GolgiStop Protein Transport Inhibitor (1:1000). After red blood cells were removed with Lysing Buffer (BD Biosciences), Fixable Viability Dye(eBioscience)was applied to label dead cells in the samples. Then, samples were incubated with Human Fc receptor blocking reagent (BD Biosciences) and stained with the indicated fluorochrome-conjugated antibodies against surface markers for 30 min at 4 °C in dark. For staining with antibodies against intracellular proteins or transcription factors, surface-stained cells were treated with the Fixation/Permeabilization Solution Kit or Transcription Factor Fixation/ Permeabilization Buffer (BD Biosciences) respectively according to the manufacturer's instructions, and then incubated with the indicated fluorochrome-conjugated antibodies for 40 min at 4 °C in dark. Flow cytometry analysis was performed with the BD FACS Celesta flow cytometer and the data were analyzed using FlowJo 10.0 software (Tree Star). All flow cytometry antibodies and reagents are summarized in Additional file 1: Table S2.

RNA sequencing
Siglec-10 hi TAMs (CD45 + CD14 + Siglec-10 hi ) and Siglec-10 lo TAMs (CD45 + CD14 + Siglec-10 lo ) were freshly isolated from three HCC specimens using the MoFlo XDP Cell Sorter (Beckman Coulter, Sykesville, MD, USA) and directly lysed in lysis buffer. The cDNA libraries were constructed using the TruSeq Stranded mRNA LT Sample Prep Kit (Illumina, San Diego, CA, USA) according to the manufacturer's instructions. Then the libraries were sequenced on the Illumina HiSeq X Ten platform. About 6G raw reads for each sample of fastq format were processed using Trimmomatic [19] and low-quality reads were removed to obtain clean reads.

Sequence analysis
The clean reads were aligned to the human genome (GRCh38) using HISAT2 [20]. The FPKM [21] of each gene was calculated using Cufflinks [22], and the read counts of each gene were obtained by HTSeq-count [23]. Differential expression analysis was performed using R software (version 3.6.0) and edgeR package (version 3.26.8) [24]. P < 0.05 and fold change > 2 or fold change < 0.5 was set as the threshold for significantly differential expression. Hierarchical cluster analysis of differentially expressed genes (DEGs) was performed to demonstrate the expression pattern of genes. Gene Ontology (GO) enrichment and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis of DEGs between Siglec-10 hi TAMs and Siglec-10 lo TAMs were performed respectively using R software based on the hypergeometric distribution. Gene set enrichment analysis (GSEA) was performed by the Molecular Signature Database for gene functional annotation.

Bioinformatics analysis
The Cancer Genome Atlas Liver Hepatocellular Carcinoma (TCGA-LIHC) mRNA and clinical data, including RNA sequencing (RNA-seq) and clinicopathological data for 372 tumors, were downloaded from https:// www. cbiop ortal. org on July 15, 2020. Six cases were excluded due to lack of RNA-seq data. Finally, data from 366 patients were enrolled in this study. The median value of Siglec-10 expression was set as the cut-off value.

In vitro neutralizing assay
In vitro intervention studies were performed according to previously described methods [25]. Dissociated from fresh human HCC tissues, single-cell suspension containing tumor cells, immune cells, and other cells were co-cultured and randomly divided into four groups (isotype control, Siglec-10 Fc, pembrolizumab, Siglec-10 Fc plus pembrolizumab). The usage of recombinant human Siglec-10 Fc chimera was according to previous research [26,27].
The antibodies used in this experiment were IgG1 isotype control (10 μg/mL, ab206198; Abcam, Cambridge, UK), recombinant human Siglec-10 Fc chimera (5 μg/ mL; R&D Systems), and pembrolizumab (5 µg/mL; Selleck Chemicals, Houston, TX, USA). After overnight culture in RPMI 1640 medium containing 10% fetal bovine serum and corresponding antibodies, cells were subjected to flow cytometry analysis to examine the apoptosis of tumor cells by using the FITC Annexin-V Apoptosis Detection Kit I (BD Biosciences) or corresponding fluorescence-activated cell sorting antibodies.

Statistical analysis
Results are expressed as the mean ± standard deviation. Categorical variables were reported as numbers and percentages, and were analyzed by the Pearson's chi-squared test or Fisher's exact test. Student's t-test or one-way analysis of variance was used for continuous variables. Spearman's correlation was employed to evaluate the correlation between different variables. Kaplan-Meier method was used to determine OS and RFS. Cox proportional hazards regression model was used for multivariate analysis with hazard ratios (HRs) and 95% confidence intervals (CIs). To obtain the best prognostic efficacy, X-Tile Software (version 3.6.1; Yale University, New Haven, CT, USA) was used as previously described [28]. The cut-off values of Siglec-10 as prognostic biomarkers were defined according to the OS and RFS data of the discovery set, and then applied to the validation set. In flow cytometry analysis, the median value of intratumoral Siglec-10 hi TAM infiltration(Siglec-10 hi CD68 + cells / CD68 + cells)were defined as cut-off value. Calculations were done with SPSS 22.0 (IBM, Armonk, NY, USA). Graphical illustrations were carried out with GraphPad Prism v7 (GraphPad Software, Inc., La Jolla, CA, USA). A two-tailed P value less than 0.05 indicated statistical significance.

Intratumoral Siglec-10 + cell enrichment associates with poor prognosis in patients with HCC
To investigate the clinical significance of intratumoral Siglec-10 + cells in HCC, Kaplan-Meier curves were used to compare the prognosis of patients stratified by intratumoral Siglec-10 + cell infiltration. Patients with low Siglec-10 + cell infiltration had better OS (P = 0.005 and 0.003, respectively; Fig. 1a, c) and RFS (P = 0.004 and 0.016, respectively; Fig. 1b, d) in both the discovery set and validation set. Multivariate analysis revealed that intratumoral Siglec-10 + cell infiltration was an independent prognostic factor of unfavorable OS and RFS (P = 0.004 and 0.008, HR = 1.691 and 1.577, respectively; Fig. 1e, f ). These findings suggest that intratumoral Siglec-10 + cells may lead to poor prognosis in patients with HCC.

Siglec-10 is enriched in HCC and mainly distributed on TAMs
IHC of HCC TMAs showed that Siglec-10 + cells were abundant in tumor tissues compared with peritumor tissues (Fig. 2a). IF analysis of human HCC tissues revealed that most Siglec-10 was co-localized with macrophage marker CD68 (Fig. 2b). To detect the specific distribution of Siglec-10 in the TME, the co-localization of Siglec-10 with different immune cell markers was examined. Flow cytometry showed that Siglec-10 was predominantly expressed by macrophages (Fig. 2c, d), and that HCC specimens exhibited higher levels of Siglec-10 hi TAMs compared with peritumor tissues (Fig. 2e, f ).  (Fig. 3a). In flow cytometry analysis, patients with high (≥ 25%) or low (< 25%) intratumoral Siglec-10 hi TAM infiltration (Siglec-10 hi CD68 + cells / CD68 + cells) were defined according to median cut-off value. HCC with high Siglec-10 hi TAM infiltration exhibited higher levels of CD4 + FoxP3 + regulatory T cells compared to tumors with low Siglec-10 hi TAM infiltration (Fig. 3b). Flow cytometry analysis confirmed that the levels of CD45 + CD8 + T cells and CD45 + CD56 + NK cells were significantly higher in specimens with low Siglec-10 hi TAM infiltration compared with high Siglec-10 hi TAM infiltrating samples (Fig. 3b).

Blocking Siglec-10 improves the anti-tumor activity of CD8 + CTLs and the efficacy of the PD-1 inhibitor
Then we analyzed the effects of blocking Siglec-10 in HCC with recombinant human Siglec-10 Fc chimera. After incubating with Siglec-10 Fc for 12 h (Fig. 6a), single-cell suspension was subjected to flow cytometry analysis to detect the phenotypes of Siglec-10 hi TAMs. The expression of pro-inflammatory cytokines (IL-12, TNF-α) significantly increased after blocking Siglec-10, whereas the levels of anti-inflammatory molecules were markedly decreased compared with the isotype group (Fig. 6b).
Additionally, the mRNA level of Siglec-10 was positively correlated with that of PD-1 and PD-L1 in TCGA cohort (Fig. 7a, b), indicating that Siglec-10 is involved in the PD-1 inhibitory pathway and may induce CD8 + CTL inactivation. Thus, we speculate that inhibiting Siglec-10 may produce synergistic benefits with the PD-1 inhibitor pembrolizumab, facilitating CD8 + CTL activation and tumor cell death.
We found that treatment of Siglec-10 Fc or pembrolizumab in monotherapy groups can promote CD45 − Epcam + tumor cell death and inhibit the proliferation of tumor cells compared to the isotype control (Fig. 7c, d). Furthermore, the combination of Siglec-10 Fc and pembrolizumab had synergistic effects on promoting tumor cell apoptosis (Fig. 7c) and impairing tumor cell proliferation (Fig. 7d) compared to monotherapy groups. Overall, these findings indicate that blocking Siglec-10 may exert synergistic effects with pembrolizumab to restore the anti-tumor activity of CD8 + CTLs and promote tumor cell death.

Discussion
Siglecs are proteins that bind to sugar molecules, which are attached to many other proteins. Most Siglecs act as inhibitory receptors on innate and adaptive immune cells [29]. Cancer-associated aberrant sialoglycan expression causes immune evasion upon binding to inhibitory Siglecs [30][31][32]. In fact, there is an association between levels of tumor glycosylation and metastatic potential [33,34]. Notably, the high expression of Siglec-10 is associated with poor survival and immune dysfunction in patients with HCC [35]. Deficiency of Siglec-10 or blocking Siglec-10 on human macrophages increases phagocytosis and tumor cell clearance in breast cancer [14]. According to our data, intratumoral Siglec-10 was primarily co-localized with the macrophage marker CD68 in HCC tissues. This suggests that intratumoral Siglec-10 hi macrophages may act as an immune modulator in the TME of HCC. Indeed, our results revealed that intratumoral Siglec-10 hi TAMs expressed higher levels of suppressive cytokines and inhibitory receptors than intratumoral Siglec-10 lo (See figure on next page.) Fig. 5 RNA-seq analysis reveals significant transcriptomic differences between Siglec-10 hi and Siglec-10 lo TAMs. a Heat map analysis showed gene expression differences between intratumoral Siglec-10 hi and Siglec-10 lo TAMs. b KEGG pathway analysis of the upregulated DEGs between intratumoral Siglec-10 hi and Siglec-10 lo TAMs. c GSEA plots of gene expression changes in Siglec-10 hi TAMs compared with Siglec-10 lo TAMs. The signature was defined by genes with significant expression changes. d GO enrichment of the downregulated DEGs between intratumoral Siglec-10 hi and Siglec-10 lo TAMs TAMs. In line with these results, RNA-seq data confirmed that multiple signaling pathways involved in macrophage M2 polarization were specifically activated in Siglec-10 hi TAMs. Additionally, high levels of Siglec-10 hi TAMs infiltration were related to an immunosuppressive TME with restrained anti-tumor activity. Sialoglycan-Siglec immune checkpoints are attractive targets for anti-tumor immunotherapy, as aberrant glycosylation is a key feature of malignant transformation [32]. The use of blocking antibodies against Siglec-7 and Siglec-9 has been successfully tested in vitro to enhance the immune response against cancer [36][37][38]. Siglec-2 (CD22) and Siglec-3 (CD33)-targeting antibody-drug conjugates have been approved by US Food and Drug Administration for the treatment of lymphoma and leukemia [39][40][41]. Previous in vivo research has reported that sialic acid blockade enhances CTL-mediated killing of tumor cells, thus suppressing tumor growth [42]. A recent study revealed that Siglec-15 suppresses antigen-specific T-cell responses in vitro and in vivo. Blocking Siglec-15 enhances anti-tumor immunity and inhibits tumor growth in mice [43]. A humanized anti-Siglec-15 monoclonal antibody is being assessed in a phase I clinical trial in advanced solid tumors (NCT03665285). Blockade of PD-1/PD-L1 can reactivate the anti-tumor function of intratumoral T cells in HCC and has shown clinical efficacy in 20% HCC patients [5]. Importantly, combining blockade of PD-L1 with TIM3, LAG3, or CTLA4 further improved the responsiveness and was effective in more patients [44]. Recent study on microenvironment characterization indicated that the immune status of HCC could help to recognize patients with different prognosis and responses to immunotherapy in HCC [45]. Suppressive immune cells inhibit the immune response of tumor-specific T cells and could serve as biomarkers of anti-PD-1 therapy [46]. In accordance with these results, we found that blocking Siglec-10 led to reduced expression of immune inhibitory molecules in the TME and enhanced the anti-tumor activity of CTLs. In addition, compared with monotherapy groups, the combination of Siglec-10 blockade and pembrolizumab significantly enhanced tumor cell apoptosis and impaired tumor cell proliferation. Diversity and plasticity are hallmarks of the monocytemacrophage lineage. Macrophages associated with different TME exhibit a variety of phenotypes. While the description of M1 and M2 phenotypes suggests that TAMs can be either tumor-killing or tumor-promoting, the heterogeneity of TAM functions in the TME goes far beyond that. Transcriptome analysis has revealed that macrophages have an entirely different transcriptional profile distinct from M1 or M2 activation [47][48][49][50]. TAM subtypes generally perform mixed M1/M2 markers at different levels [48,50,51]. In a reassessment, Fernando [52] postulated that macrophages do not form settled subsets but constantly respond to a combination of cues present in the TME.
In our study, we also found that intratumoral Siglec-10 hi TAMs exhibited mixed M1 (CD80, CD86, HLA-DR) and M2 (CD206) markers. Compared to Siglec-10 lo TAMs, Siglec-10 hi TAMs showed higher levels of immune stimulatory cytokines (TNF-α, IL-12), and higher levels of immune inhibitory molecules (PD-L1, TIM-3, Arg-1, IL-10, TGF-β). Interestingly, the expression of immune inhibitory molecules was significantly suppressed after blocking Siglec-10, whereas the levels of TNF-α and IL-12 were further increased. It is known that despite functional impairment, exhausted T cells with overexpression of inhibitory receptors retain a crucial level of control over tumor growth [53][54][55]. Targeting these inhibitory receptors could enhance T cell-mediated immune responses against tumors [55]. Likewise, we speculate that TAM subtype with M2 features, such as Siglec-10 hi TAMs, retain the anti-tumor effect in some degree. Indeed, our findings indicate that Siglec-10 expression may facilitate the secretion of macrophagederived anti-inflammatory cytokines in HCC. However, once released from inhibition, Siglec-10 hi TAMs may be transformed into an active role in anti-tumor immunity. In summary, Siglec-10 represents a novel immune