Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2025 Feb;39(2):346-359.
doi: 10.1038/s41375-024-02454-w. Epub 2024 Nov 17.

The glycosyltransferase ST3GAL4 drives immune evasion in acute myeloid leukemia by synthesizing ligands for the glyco-immune checkpoint receptor Siglec-9

Affiliations

The glycosyltransferase ST3GAL4 drives immune evasion in acute myeloid leukemia by synthesizing ligands for the glyco-immune checkpoint receptor Siglec-9

Vignesh Krishnamoorthy et al. Leukemia. 2025 Feb.

Abstract

Immunotherapy has demonstrated promise as a treatment for acute myeloid leukemia (AML). However, there is still an urgent need to identify new molecules that inhibit the immune response to AML. Most prior research in this area has focused on protein-protein interaction interfaces. While carbohydrates also regulate immune recognition, the role of cell-surface glycans in driving AML immune evasion is comparatively understudied. The Siglecs, for example, are an important family of inhibitory, glycan-binding signaling receptors that have emerged as prime targets for cancer immunotherapy in recent years. In this study, we find that AML cells express ligands for the receptor Siglec-9 at high levels. Integrated CRISPR genomic screening and clinical bioinformatic analysis identified ST3GAL4 as a potential driver of Siglec-9 ligand expression in AML. Depletion of ST3GAL4 by CRISPR-Cas9 knockout (KO) dramatically reduced the expression of Siglec-9 ligands in AML cells. Mass spectrometry analysis of cell-surface glycosylation in ST3GAL4 KO cells revealed that Siglec-9 primarily binds N-linked sialoglycans on these cell types. Finally, we found that ST3GAL4 KO enhanced the sensitivity of AML cells to phagocytosis by Siglec-9-expressing macrophages. This work reveals a novel axis of immune evasion and implicates ST3GAL4 as a possible target for immunotherapy in AML.

PubMed Disclaimer

Conflict of interest statement

Competing interests: The authors declare no competing interests. Ethics approval and consent to participate: Primary AML bone marrow and peripheral blood samples were obtained from the hematology cell bank at BC Cancer Research Centre in a manner complying with all relevant guidelines. The study complies with ethical regulations as approved by the University of British Columbia Human Ethics Board (H24-01726). Informed consent was obtained from all subjects.

Figures

Fig. 1
Fig. 1. AML cells express Siglec-binding glycans.
A The indicated cell lines were incubated with 1 μg/mL Siglec-Fc reagents precomplexed with 1 μg/mL AlexFluor488-antihuFc for 30 minutes on ice. For T-cell staining, primary peripheral blood mononuclear cells (PBMCs) were co-stained with Siglec-Fc reagents and an anti-CD3 antibody. The cells were then analyzed by flow cytometry to detect the expression of different Siglec-binding ligands. Representative plots for each cell type are provided. For T-cells, the plot shown indicates Siglec-Fc fluorescence in CD3+ cells. B Average median fluorescence intensity (MFI) of Siglec-Fc staining in the indicated AML cell lines. The average of n = 3 independent biological replicates is plotted. C Patient-derived primary AML cells were incubated and analyzed as in (A). Representative plots are shown. Median fluorescence intensity for Siglec-Fc staining is indicated directly on the flow cytometry plot. D PBMCs from multiple healthy blood donors and the indicated cell lines were stained with Siglec-9-Fc as in (A). For PBMC samples, CD14+ monocytes were gated by co-staining with a fluorescent anti-CD14 antibody. The relative fluorescence intensity (RFI) for each sample is indicated so as to normalize for differences in autofluorescence between cell types. RFI is equal to the MFI(Sig9-Fc)/MFI(hFc). Mean values plotted, error bars indicate SEM.
Fig. 2
Fig. 2. Functional genomics analysis reveals ST3GAL4 as a potential driver of Siglec-9 ligand expression in AML.
A K-562-dCas9KRAB cells were transduced with a genome-wide library of sgRNAs and stained with Siglec-9-Fc as in Fig. 1A. Cells exhibiting low binding were sorted by FACS and sequenced to reveal enriched sgRNAs. Screen hits were analyzed using MAGeCK. Selection score indicates the strength of sgRNA enrichment in the low-staining population vs. an unsorted control. A positive score indicates that sgRNAs for the indicated gene were more enriched in the low-staining population. A negative score indicates they were less enriched. B Biosynthetic pathway for the synthesis of Sialyl-LacNAc-containing glycans. Biosynthetic enzymes that were recovered as screen hits are indicated. C Relative mRNA expression levels of ST3GAL4 in patient AML samples with the indicated genetic alterations (BloodSpot Database) are shown. ST3GAL4 expression in hematopoietic stem cells (HSC), common myeloid progenitors (CMP), and monocytes from healthy donors are shown as a comparison. An unpaired Student’s t-test was performed to detect statistically significant differences in mRNA expression between each AML subtype and HSCs. *** indicates P < 0.001.
Fig. 3
Fig. 3. ST3GAL4 knockout reduces expression of Siglec-9 ligands in AML cell lines.
A The indicated AML cell lines were transduced with a plasmid encoding Cas9 and an sgRNA targeting an exon of ST3GAL4. Genomic DNA was extracted from both the wild-type (WT) and ST3GAL4 knockout (KO) cell lines. PCR was performed to amplify a sequence around the sgRNA annealing site. The amplified region was sequenced and analyzed by TIDE software to quantify editing at the ST3GAL4 locus. The plot depicts a representative Sanger sequencing trace at the ST3GAL4 locus of the WT and KO cells, along with the expected break site. B The graph indicates the average percentage of insertions and deletions (indels) in the ST3GAL4 gene for the indicated AML cell lines transduced with an sgRNA against ST3GAL4. Mean values are plotted for n = 3 biological replicates, error bars indicate SEM. C WT and ST3GAL4 KO AML cell lines were analyzed for Siglec ligand expression. The average mean fluorescence intensity (MFI) of Siglec-9-Fc staining is shown for WT and ST3GAL4 KO cells. Mean values plotted for n = 3 biological replicates. Statistical significance was determined by a Student’s two-tailed t-test. *P < 0.05, **P < 0.01. D A monoclonal MOLM-13 ST3GAL4 KO cell line was generated through clonal dilution. Sanger sequencing of the ST3GAL4 locus is depicted. E MOLM-13 ST3GAL4 KO cells were transduced with a ST3GAL4 WT cDNA containing silent mutations to eliminate the sgRNA binding site. WT, ST3GAL4 KO, and ST3GAL4-transduced cells were then stained with fluorescent Siglec-9-Fc. The indicated flow cytometry plots are representative of n = 3 biological replicates.
Fig. 4
Fig. 4. ST3GAL4 KO primarily affects the expression of N-linked sialoglycans on AML cells.
A OCI-AML-2 WT and ST3GAL4 KO cells were lysed and glycans were extracted using a mix of chemical and enzymatic methods. Free glycans were permethylated and subjected to MS-based glycomics analysis using MALDI-TOF. The graph indicates the percentage  of all N-linked structures that contain sialic acid. B The graph displays the percent abundance of the indicated  glycan structure as a proportion of all N-linked glycans in WT and ST3GAL4 KO cells. C The graph displays the percent abundance of the indicated terminally fucosylated glycan structure as a proportion of all N-linked glycans in WT and ST3GAL4 KO cells. D The graph displays the percent abundance  of the indicated biantennary glycan structure as a proportion of all N-linked glycans in WT and ST3GAL4 KO cells. E The graph displays the percent abundance of the indicated triantennary glycan structure as a proportion of all N-linked glycans in WT and ST3GAL4 KO cells. F OCI-AML-2 WT and ST3GAL4 KO cells were processed as in (A). The graph indicates the percentage of all O-linked structures containing sialic acid in WT and ST3GAL4 KO cells. G The graph displays the percent abundance  of the indicated disialyl core 2 structure as a proportion of all O-linked glycans in WT and ST3GAL4 KO cells. H The graph displays the percent abundance  of the indicated disialyl core 1 structure as a proportion of all O-linked glycans in WT and ST3GAL4 KO cells. In all cases, mean values are plotted for n = 3 biological replicates. Error bars indicate SEM.
Fig. 5
Fig. 5. Siglec-9 binds primarily to N-linked sialoglycans.
A Biosynthetic pathway for the synthesis of complex N-linked glycans. MGAT1 catalyzes an essential first step in this pathway. B OCI-AML-2-Cas9 cells were transduced with an sgRNA against MGAT1 and co-stained with fluorescent L-PHA (5 µg/mL) along with fluorescent Siglec-9-Fc (1.5 µg/mL) as described above. A representative flow cytometry plot and gating strategy is shown. C Mean fluorescence intensity of Siglec-9-Fc staining in L-PHA+ and L-PHA- cell populations is shown. 2′ indicates control staining with a fluorescent secondary antibody. D Biosynthetic pathway for the synthesis of terminally fucosylated N-linked glycans. FUT7 catalyzes an essential step in this pathway. E K-562-dCas9-VPR cells were transduced with sgRNAs targeting FUT7 and stained with Siglec-9-Fc as described above. The median fluorescence intensity of Sig9-Fc staining is shown. F Biosynthetic pathway for the synthesis of branched N-linked glycans. MGAT4A catalyzes an essential first step in this pathway. G K-562-dCas9VPR cells were transduced with sgRNAs targeting MGAT4A and stained with Siglec-9-Fc as described above. The median fluorescence intensity of Sig9-Fc staining is shown. Mean values are plotted for n = 3 biological replicates in all cases, error bars indicate SEM.
Fig. 6
Fig. 6. ST3GAL4 KO sensitizes AML cells to phagocytosis.
A Monocytes were isolated from PBMCs and differentiated into macrophages over a period of 7 days. Macrophages were incubated with fluorescent antibodies against CD14, CD11b, Siglec-7 and Siglec-9. The cells were then analyzed by flow cytometry. The indicated flow cytometry plots are representative of n = 3 biological replicates. B WT and ST3GAL4 KO MOLM-13 cells were labeled with a fluorescent dye (CTFR) and co-cultured with macrophages for 4 h. Phagocytosis was quantitated by analyzing the uptake of the fluorescent dye by macrophages. C A representative gating strategy for analysis of phagocytosis by primary macrophages. Macrophage co-cultures were stained with an anti-CD11b antibody. The phagocytic index is the median CTFR fluorescence intensity in CD11b+ cells. D The indicated flow cytometry plots depict the phagocytosis of MOLM-13 WT and ST3GAL4 KO cells. Plots are representative of n = 3 biological replicates. E The phagocytic index was determined as in (C) following the co-culture of macrophages with WT or ST3GAL4 KO cells. Cells were either untreated, pre-treated with a Siglec-9-blocking antibody (20 µg/mL), or pre-treated with a sialidase enzyme prior to co-culture. All values were normalized to the WT, untreated sample. F Patient-derived xenograft (PDX) AML cells were isolated, treated with sialidase, and co-cultured with primary macrophages as in (B). Two different PDX donors were analyzed. The mean value of all technical replicates for a given donor is indicated. G Co-cultures were established as in (B). After 24 h, the co-culture was stained with a fluorescent anti-CD86 antibody. The percentage of the total cells that were CD86 positive is graphed for the indicated conditions. H Co-cultures were established as in (B). After 24 h, the supernatants were collected and subjected to Luminex analysis to quantitate the secretion of TNF-α. The mean concentration of TNF-α in culture supernatants is shown for each of the indicated conditions. Mean values are plotted for n = 3 biological replicates in all cases, error bars indicate SEM. Statistical significance was determined by Student’s two-tailed t-test, *P < 0.05, **P < 0.01 ***P < 0.001.

References

    1. Munoz-Wolf N, Lavelle EC. Innate immune receptors. Methods Mol Biol. 2016;1417:1–43. - PubMed
    1. Kawai T, Akira S. The role of pattern-recognition receptors in innate immunity: update on Toll-like receptors. Nat Immunol. 2010;11:373. - PubMed
    1. Pardoll DM. The blockade of immune checkpoints in cancer immunotherapy. Nat Rev Cancer. 2012;12:252. - PMC - PubMed
    1. Carlino MS, Larkin J, Long GV. Immune checkpoint inhibitors in melanoma. Lancet. 2021;398:1002–14. - PubMed
    1. Yang H, Bueso-Ramos C, DiNardo C, Estecio MR, Davanlou M, Geng QR, et al. Expression of PD-L1, PD-L2, PD-1 and CTLA4 in myelodysplastic syndromes is enhanced by treatment with hypomethylating agents. Leukemia. 2014;28:1280–8. - PMC - PubMed

MeSH terms

Substances