Effect of the intratumoral microbiota on spatial and cellular heterogeneity in cancer

معا للقضاء على التشيع

The presence of bacteria within individual host cells of the TME has been reported across a range of human cancer types2,13. However, we have little information on the identity of invasive bacteria, the host cells that they interact with and how these host–bacterial associations affect cellular function within the TME. To investigate bacterial–host cell-to-cell interaction within the TME and the effect on host cell transcriptomics, we developed INVADEseq (invasion–adhesion-directed expression sequencing) by introducing a primer that targets a conserved region of bacterial 16S rRNA, facilitating the generation of cDNA libraries with bacterial transcripts from the bacteria-associated human cells (Extended Data Fig. 4a). Addition of this bacteria-targeting primer did not affect the gene-expression profile of human CRC cells (Extended Data Fig. 4b), and validation co-culture experiments with non-adherent and non-invasive Escherichia coli DH5α (Extended Data Fig. 4c) showed specificity for cell-associated bacteria.

To further validate this approach, the human CRC cell line HCT116 was infected with three invasive bacterial species—F.nucleatum, Porphyromonas gingivalis and Prevotella intermedia—at a multiplicity of infection (MOI) of 100:1 and 500:1, and processed for INVADEseq (Extended Data Fig. 4d). Confocal imaging indicated the presence of intracellular bacteria in cancer cells after bacterial co-culture (Extended Data Fig. 4e). Using INVADEseq, we mapped bacterial reads to single human cells (Extended Data Fig. 4f,g). At the cell-cluster level for these epithelial single cells (clusters 1–10), most F.nucleatum– and P.gingivalis-positive single cells were distributed in cancer cell clusters 5 and 6, respectively (Extended Data Fig. 4g). Both cell clusters (clusters 5 and 6) were very minor cell populations in the uninfected control group (Extended Data Fig. 4f). When compared to uninfected controls (MOI = 0), the appearance of cell clusters 5 (Fusobacterium cluster) and 6 (Porphyromonas cluster) coincided with a relative reduction in the percentage of cluster 1 (uninfected control cluster) (Extended Data Fig. 4g). This finding suggests that F.nucleatum and P.gingivalis affect cancer cell heterogeneity by altering distinct transcriptional programs that contribute to specific cell clusters (Extended Data Fig. 4g).

After integrating data from the three HCT116 samples (Extended Data Fig. 4h,i), we compared the gene expression of F.nucleatum– or P.gingivalis-associated single epithelial cells to that of the bacteria-negative epithelial single cells (Total Bac). We noted that the number of differentially expressed genes increased when a bacterial UMI cut-off (≥3), a proxy for bacterial transcriptional load, was applied (Extended Data Fig. 5a–d and Supplementary Table 5). Furthermore, a comparison of cells from cluster 5 (Fusobacterium cluster) and cluster 6 (Porphyromonas cluster) to bacteria-negative cells from cluster 1 (uninfected control cluster) showed that bacteria-infected cells exhibited a significant upregulation of signalling pathways that are involved in the response to bacterial infection, such as the TNF pathway and pathways related to inflammation and hypoxia, as well as cancer cell progression via the epithelial–mesenchymal transition (EMT) and the p53 signalling pathway22,23. Bacteria-infected cells also showed a downregulation of cell-cycle signalling pathways that relate to the formation of the mitotic spindle and the G2–M DNA damage checkpoint, as compared with cells from the uninfected control cluster (Extended Data Fig. 5a–d). At the gene-expression level, bacteria-associated single epithelial cells showed significant increases in the expression of molecules that are positively associated with metastasis, such as PLAU, PLAUR, RELB and AREG, along with an upregulation of the chemokines CXCL1, CXCL2, CXCL3 and the neutrophil chemoattractant CXCL8, along with members of the TNF family (Extended Data Fig. 5a–d). We also noted a significant upregulation of transcription factors including NFKBIA, NFKB2, NEAT1, SAT1 and members of the JUN and FOS family, with a downregulation of the cyclins CCNB1 and CCNA2 (Extended Data Fig. 5a–d). Similar findings were observed when CRC-derived HT-29 cells were treated with F.nucleatum at a MOI of 100:1; that is, an increase in the expression of genes that encode molecules related to inflammation through TNF, hypoxia, the EMT and p53 signalling pathways, and a reduction in the expression of genes that are involved in DNA repair (Extended Data Fig. 5e–g and Supplementary Table 5).

The INVADEseq method was subsequently implemented to examine bacteria–host interactions in fresh tumour tissues from seven patients with OSCC. After the tissues were dissociated to single cells, confocal imaging showed that single cells from a tumour from a patient with OSCC contained cell-adherent and intracellular bacteria (Fig. 3a). Integrated scRNA-seq from the seven tumours revealed that the intratumoral microbiota is dominated by bacterial species that belong to the Fusobacterium (34%) and Treponema (29.8%) genera (Fig. 3b). Mapping bacterial reads from INVADEseq analysis to annotated single cells showed that Fusobacterium and Treponema were predominantly associated with the epithelial and monocyte-derived macrophage-v1 (referred to as the macrophage cluster) cell clusters in these patient tumours, with a total bacterial infection rate of 25% and 52%, respectively (Fig. 3c and Extended Data Fig. 6a). INVADEseq cannot distinguish whether bacteria are actively invading the macrophage cells or whether the macrophages are phagocytizing the bacteria; however, we refer to these cells as ‘macrophages with bacteria engulfed’. Within the epithelial cell clusters, cells in cluster 3 were identified as aneuploid, confirming that these are tumour cells with severe chromosomal instability (Extended Data Fig. 6b–d). Notably, this aneuploid epithelial cell cluster contained most of the bacterial UMI transcripts, as compared to other euploid epithelial cell clusters (Extended Data Fig. 6d). Gene set enrichment analysis (GSEA) confirmed that the cells from the bacteria-dominant epithelial cell cluster 3 were indeed cancer cells, with gene-expression signatures characterized by an upregulation of signalling pathways involved in cancer progression, including EMT, PI3K–AKT–mTOR, hypoxia and the interferon (IFN) response, among others (Extended Data Fig. 6e–g).

Fig. 3: Effect of cell-associated intratumoral bacteria on transcriptomics in host single cells.
figure 3

a, RNAscope-FISH (left) shows the distribution of intratumoral bacteria in a tumour from a patient with OSCC. Confocal images (right) show bacteria-associated single cells after tissue dissociation. Scale bars, 1 mm (left); 5 μm (right). b, Microbiome composition at the genus level after integration of tumour scRNA-seq data from seven patients with OSCC using the INVADEseq method. c, UMAP plots indicate host cell annotation and bacteria transcripts (UMI) from total bacteria and Fusobacterium– and Treponema-associated cells in integrated tumour single-cell data from seven patients with OSCC as indicated. Colour bars indicate the bacterial UMI transcripts for total bacteria and for each bacterial species as indicated. DCs, dendritic cells; MSCs, mesenchymal stem cells; Treg cells, regulatory T cells. d, GSEA analysis showing the signalling pathways that are differentially regulated in cells that contain ≥3 Fusobacterium UMI (High Fuso) or ≥3 Treponema UMI (High Trep) transcripts versus (vs) total bacteria-negative cells (Total Bac) from the epithelial cell cluster. e, Volcano plots showing the differentially expressed genes between cell populations described in d. Dashed lines indicate the threshold of significant gene expression, defined as log2-transformed fold change ≤ −0.58 and ≥ 0.58 with −log10(P) ≥ 1.301. f, GSEA analysis showing the signalling pathways that are differentially regulated between total Fusobacterium (Total Fuso+) or total Treponema (Total Trep+) associated cells versus bacteria-negative cells (Total Bac) in the monocyte-derived macrophage-v1 cell cluster. g, Volcano plots showing the differentially expressed genes between cell populations described in f. Dashed lines indicate the threshold of significant gene expression, defined as log2-transformed fold change ≤ −0.58 and ≥ 0.58 with −log10(P) ≥ 1.301. The normalized enrichment scores (NESs) in d,f were calculated using the Wilcoxon rank sum test. LMM analysis followed by Benjamini–Hochberg multiple-correction testing was used to calculate the fold change and P values for each gene in e,g.

To determine whether the dominant cell-associated bacterial genera, Fusobacterium and Treponema, affected epithelial signalling pathways, Fusobacterium– or Treponema-associated single cells (UMI ≥ 3) were compared to bacteria-negative cells (Total Bac) from the epithelial cell cluster. After GSEA analysis, we observed a significant upregulation of IFN and JAK–STAT signalling, with increased expression of molecules from the SERPIN family; chemokines such as CXCL10, CXCL11, CCL4 and CCL3; and metalloproteinases, including MMP9 and MMP3 (Fig. 3d,e and Supplementary Table 6). A comparison of general bacteria-positive epithelial cells (Total Bac+), independent of a specific genus, and bacteria-negative cells (Total Bac) showed that gene expression and cell signalling pathways related to cancer progression were  modestly affected in bacteria-positive epithelial cells, as compared to the effects that were observed in cells infected with specific taxa (Extended Data Fig. 6h–j and Supplementary Table 6). This is likely to be reflective of taxa-specific epithelial cell interactions or capabilities rather than a general bacteria-induced response.

At the specimen level, the total bacterial load from each sample was negatively correlated with the expression of TP53 and positively correlated with its negatively regulated target molecule, SAT1 (Extended Data Fig. 7a)—consistent with our findings from DSP (Fig. 2), in which bacteria colonized microniches with reduced levels of wild-type p53. In addition, the total bacterial load negatively correlated with the expression of the proliferation marker MKI67, which encodes Ki-67 (Extended Data Fig. 7a), providing support for our spatial microniche data (Fig. 2 and Extended Data Fig. 3a,b).

In the macrophage cell cluster, by comparing Total Bac+ to Total Bac cells, we found that macrophages with bacteria engulfed had significantly increased expression levels of genes that are involved in the inflammatory response through activation of TNF, INFγ and IFNα, and genes that are involved in the production of interleukins through the JAK–STAT signalling pathway, such as IL1B, IL6 and IL10. Macrophages with bacteria engulfed also overexpressed the chemokines CCL2, CCL4, CCL8, CCL7, CXCL1 and CXCL10 (Extended Data Fig. 6k,l and Supplementary Table 7). This gene-expression signature was observed when analysing cells associated with bacteria in general (Extended Data Fig. 6k,l), but also when assessing specific bacterial genera, including Fusobacterium and Treponema (Fig. 3f,g and Supplementary Table 7). Furthermore, at the specimen level, the bacterial load from each OSCC specimen was positively correlated with the potent neutrophil chemoattractant CXCL8 and negatively correlated with the expression of CD3E (Extended Data Fig. 7a), supporting the DSP findings that intratumoral bacteria-colonized microniches are immunosuppressive by recruiting neutrophils and excluding CD3+ T cells (Fig. 2 and Extended Data Fig 3c,d).

Unlike our findings in ‘macrophages with bacteria engulfed’ single cells, in which the response appears generalized to the presence of bacterial lipopolysaccharide or other widespread damage-associated molecular patterns, in epithelial single cells, specific dominant taxa such as Fusobacterium and Treponema enhanced signatures of cancer progression. Overall, this shows that the cell-associated members of the intratumoral microbiota can drive heterogeneity in patient tumours at the single-cell level within immune and epithelial populations.

An independent analysis of tumour single-cell data from the individual patients with OSCC revealed inter-patient heterogeneity in bacterial load, dominant cell-associated bacterial taxa and magnitude of the inflammatory gene-expression response (Extended Data Fig. 8a–d and Supplementary Tables 8–10). Similar to the integrated analysis, the percentage of bacteria-associated single cells is significantly higher in the aneuploid cancer epithelial cell cluster (cluster 3) compared to the euploid epithelial cell clusters (Extended Data Fig. 8e). This single-cell analysis of individual patients shows that specific cell-associated bacteria can significantly affect intratumoral heterogeneity at the single-cell level (Extended Data Fig. 8a–d and Supplementary Tables 8–10).

معا للقضاء على التشيع

اترك تعليقاً

لن يتم نشر عنوان بريدك الإلكتروني.

زر الذهاب إلى الأعلى