Article Text
PDF
Abstract
Background The Tumor Immuno-Micro Environment (TIME) exhibits a complex interplay of diverse cellular and molecular components. When examining the TIME in cancers linked to viral infections, understanding the localized immune response and developing targeted therapies necessitates the comparison of immune cells proximal to and distant from the viral infection sites. Previously, this investigation was unattainable until the emergence of spatial transcriptomic techniques. In this study, we introduce a method for precisely localizing viral infection sites using SpaTial Enhanced REsolution Omics-sequencing (Stereo-seq)1 data, as depicted in figure 1. We demonstrated the efficacy of this approach through two representative samples: Epstein-Barr virus (EBV)-associated Nasophayngeal Carcinoma (NPC)2 3 and human Hepatitis B Virus (HBV)-associated hepatocellular carcinoma (HCC).4 5
Methods The Stereo-seq experimental protocol was executed separately for fresh-frozen NPC and LELC samples, with the inclusion of one virus-free cancer fresh-frozen sample as a control in each run. Given the limited read depth per Nanoball, we adopted a 100 x 100 Nanoball grid (BIN100) as the analytical unit to ensure adequate read depth (figure 1A). For each BIN100, we used Stereo-Seq Analysis Workflow (SAW) pipeline v5.5.2 for aligning its reads to human genome GRCh38.p12, enabling Louvain cell clustering using Seurat package and cell type annotation using Bioconductor packages EasyCellType and SingleR. Reads unaligned to the human genome were mapped to EBV-1 or HBV genome from NCBI RefSeq using STAR v2.7.10b. Subsequently, BIN100s with substantial viral reads were superimposed onto the SAW-registered ssDNA fluorescent-staining image. Additionally, Hematoxylin and Eosin (H&E) staining, as well as QuPath tissue annotation, were performed.
Results Overlaying images of virus-positive BIN100s and ssDNA tissue-staining vividly demonstrated a clear distinction between virus-positive and virus-negative cancer samples (figures 2A,3A). Most EBV1-positive BIN100s were located in the invasive margin, as indicated by QuPath annotation of H&E staining (figure 2A-B), aligning with expectations. Surrounding viral infection sites, B cell/plasma cell clusters 5,9,14 were prevalent for NPC, while monocyte clusters 13,15 were prominent for HCC (figures 2,3). The clear separation of these clusters from other immune cell clusters underscores significant influence of the distance from viral infection sites on the gene expression profile of immune cells.
Conclusions Our study introduces a robust method for localizing viruses within the tumor microenvironment, unveiling the intricate details of TIME shaped by viral infections. In the future, we plan to further validate this method by applying it to additional NPC and HCC samples, along with more virus-free and virus-positive cancer samples. Furthermore, we will continue conducting differential gene expression analysis for immune cell clusters at varying distances from viral infection sites.
References
Chen A, Liao S, Cheng M, Ma K, Wu L, Lai Y, et al. Spatiotemporal transcriptomic atlas of mouse organogenesis using DNA nanoball-patterned arrays. Cell. 2022;185:1777–1792.
Young L, Yap L, Murray P. Epstein-Barr virus: more than 50 years old and still providing surprises. Nat Rev Cancer. 2016;16:789–802.
Png YT, Yang AZY, Lee MY, Chua MJM, Lim CM. The Role of NK Cells in EBV Infection and EBV-Associated NPC. Viruses. 2021 Feb 15;13(2):300.
Chen CJ, Yang HI, Su J, Jen CL, You SL, Lu SN, Huang GT, Iloeje UH; REVEAL-HBV Study Group. Risk of hepatocellular carcinoma across a biological gradient of serum hepatitis B virus DNA level. JAMA. 2006 Jan 4;295(1):65–73.
Liu Z, Jiang Y, Yuan H, Fang Q, Cai N, Suo C, Jin L, Zhang T, Chen X. The trends in incidence of primary liver cancer caused by specific etiologies: Results from the Global Burden of Disease Study 2016 and implications for liver cancer prevention. J Hepatol. 2019 Apr;70(4):674–683.
Ethics Approval This study was approved by the Agency of Science, Technology and Research (A*STAR) Human Biomedical Research Office (A*STAR IRB: 2021-161, 2021-188, 2021-112).
Consent De-identified patient data was used in our work. Samples were collected with consent from patients.
Overview of Stereo-seq technology and its experimental workflow. (A) Stereo-seq chip illustration: a DNA nanoball (DNB) array with sub-cellular resolution. Each template consists of a primer oligonucleotide, a unique coordinate identifier (CID), a capture oligo, and adapters on both ends. Each template is cyclized into a circular single-stranded DNA, and then undergone rolling cycle amplification separately to form a DNA nanoball (DNB). DNBs are then loaded into separate spots according to the MGI DNBSEQ-Tx sequencer manual. Each spot is 220nm in diameter, with a center-center distance of 500nm. In data analysis, square bins (BIN1) contain sequencing reads in the spot and surrounding area. To reduce sequencing errors due to low read counts, adjacent square bins are often combined into larger bins (BIN3 = 3×3 BIN1 spots). Common choices are BIN14 (close to the size of most cells), BIN20, BIN50, BIN100, or BIN200. Capture oligonucleotide probes with CID, UMI, and Poly-T are hybridized with DNBs and fixed onto the chip. Fresh-frozen tissue sections are placed on the chip, and tissue poly(A) mRNAs hybridize with the fixed probes. Subsequent steps include reverse transcription, amplification, cDNA library preparation, and sequencing. Stereo-seq Analysis Workflow (SAW) converts CIDs to x and y coordinates on the chip for spatial localization of sequencing reads using a chip specific mask file. (B) Stereo-seq results are pair-ended, with read 1 containing CID and UMI information and read 2 containing the cDNA sequence of interest. (C) Note that the current Stereo-seq protocol does not allow simultaneous H&E staining and in-situ sequencing on the same tissue section. H&E staining is performed on a nearby tissue section, which may result in tissue outline discrepancies (see figure 3). Future protocol versions aim to address this issue, enabling H&E staining and in-situ sequencing on the same tissue section.
Application of Stereo-seq and SAW to an Epstein-Barr Virus (EBV)-positive Nasopharyngeal Carcinoma (NPC) sample. (A) EBV-1 read localization on the Stereo-seq chip highlights the contrast between EBV-positive and EBV-negative cancers. Stereo-seq analysis was performed on the EBV-positive NPC sample and an EBV-negative Oral Squamous Cell Carcinoma (OSCC). The red circle identifies the top 10 BIN100s with the highest number of reads (≥100) mapped to the EBV-1 genome. The background image is the ssDNA fluorescence image from the tissue section used for in-situ sequencing. (B) QuPath annotation (right) of the NPC sample’s H&E image (left). Tumor regions are marked in red in the QuPath image and dark purple in the H&E staining, while stroma regions are marked in green in the QuPath image and light pink in the H&E staining. (C) 18 clusters of BIN100s were identified using the ‘Seurat::FindClusters(resolution=2.2)’ function. All clusters are spatially contiguous. (D) Seurat clusters of the NPC sample exhibit clear localization relative to the QuPath annotation. Clusters 2, 3, 4, and 15 primarily represent stroma (colored in green), clusters 5, 9, and 14, primarily consisting of plasma cells, are predominantly found in the invasive margin (boundary between stroma and tumor, colored in orange), and the remaining four clusters are colored in red. (E) The t-SNE plot demonstrates that the 18 clusters generated by Seurat are contiguous in both coordinate space (panel C) and t-SNE space (this panel), indicating the high quality of the clustering result. (F) t-SNE plot further illustrates the localization of clusters within the tumor, stroma, and invasive margin, with tumor clusters in red, invasive margin clusters in orange, and stroma clusters in green.
Applying Stereo-seq and SAW to HBV-positive HepatoCellular Carcinoma (HCC). (A) HBV read localization on the Stereo-seq chip distinguishes HBV-positive from HBV-negative cancers. We analysed the HBV-positive HCC sample alongside an HBV-negative ColoRectal Carcinoma (CRC). The red circle highlights the top 9 BIN100s with the most reads (≥360) mapped to the HBV genome. The background image is the ssDNA fluorescence image for in-situ sequencing. Some BIN100s extend beyond the ssDNA tissue contour due to tissue morphology variations. Panels A and C collectively reveal that clusters 13, 15 (monocyte) and cluster 14 (hepatocyte) are closest to HBV infection sites, validating Seurat clustering’s ability to differentiate monocytes in proximity to and distant from the HBV site. (B) QuPath annotation (right) of the H&E image (left) for the HCC sample. The tumor region is marked in red in the QuPath image and dark purple in H&E staining, while the stroma region is marked in green in QuPath and light pink in H&E staining. Discrepancies in HCC tissue outlines between H&E and the Stereo-seq chip result from current experimental limitations. (C) We identified 19 spatially contiguous clusters of BIN100s using ‘Seurat::FindClusters(resolution=2.2)’, With cell type annotations generated by Bioconductor packages EasyCellType and SIngleR. (D) Localization of BIN100 clusters relative to the QuPath annotation is less distinct compared to the NPC sample. Manual assessment reveals that clusters 1, 5, 9-12, 16-18 mainly represent stroma (colored in green), while the other 10 clusters are colored in red. (E) The t-SNE plot demonstrates that the 19 Seurat-generated clusters exhibit continuity in both coordinate space (panel C) and t-SNE space, affirming the high quality clustering results with labelled cell type annotations. (F) The t-SNE plot specifically highlights clusters within the tumor and stroma, with tumor clusters in red and stroma clusters in green.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See http://creativecommons.org/licenses/by-nc/4.0/.