Abstract

Introduction

   Hypopharyngeal squamous cell carcinoma (HSCC) is one of the malignant
   tumors with the worst prognosis in head and neck cancers. The
   transformation from normal tissue through low-grade and high-grade
   intraepithelial neoplasia to cancerous tissue in HSCC is typically
   viewed as a progressive pathological sequence typical of tumorigenesis.
   Nonetheless, the alterations in diverse cell clusters within the tissue
   microenvironment (TME) throughout tumorigenesis and their impact on the
   development of HSCC are yet to be fully understood.

Methods

   We employed single-cell RNA sequencing and TCR/BCR sequencing to
   sequence 60,854 cells from nine tissue samples representing different
   stages during the progression of HSCC. This allowed us to construct
   dynamic transcriptomic maps of cells in diverse TME across various
   disease stages, and experimentally validated the key molecules within
   it.

Results

   We delineated the heterogeneity among tumor cells, immune cells
   (including T cells, B cells, and myeloid cells), and stromal cells
   (such as fibroblasts and endothelial cells) during the tumorigenesis of
   HSCC. We uncovered the alterations in function and state of distinct
   cell clusters at different stages of tumor development and identified
   specific clusters closely associated with the tumorigenesis of HSCC.
   Consequently, we discovered molecules like MAGEA3 and MMP3, pivotal for
   the diagnosis and treatment of HSCC.

Discussion

   Our research sheds light on the dynamic alterations within the TME
   during the tumorigenesis of HSCC, which will help to understand its
   mechanism of canceration, identify early diagnostic markers, and
   discover new therapeutic targets.

   Keywords: hypopharyngeal squamous cell carcinoma, carcinogenesis,
   dynamic transcriptomic mapping, tissue microenvironment, single-cell
   sequencing

1. Introduction

   The hypopharynx is crucial for physiological functions like swallowing
   and speech. Hypopharyngeal squamous cell carcinoma (HSCC) often
   develops unnoticed, leading to late-stage diagnoses. Despite advanced
   treatments, the 5-year survival rate is below 40%, marking it as one of
   the most severe malignancies in the head and neck region ([35]1).
   Squamous cell carcinoma is the predominant form of hypopharynx,
   strongly linked to smoking and alcohol ([36]2). The transformation from
   normal mucosa to cancer, involving stages of low-grade intraepithelial
   neoplasia (LGIN) and high-grade intraepithelial neoplasia (HGIN), is
   poorly understood due to the complexity of signaling networks and
   molecules involved ([37]3). Thus, studying the cellular and molecular
   alterations in HSCC development is vital for understanding its
   molecular mechanisms, identifying diagnostic markers, and discovering
   therapeutic targets.

   Understanding the biological mechanisms of HSCC carcinogenesis requires
   detailed characterization of the molecular, cellular, and acellular
   components involved in cancerous transformation, which exhibit
   pronounced spatial and temporal heterogeneity throughout tumor
   initiation and progression ([38]4). This heterogeneity is marked by the
   emergence of distinct cellular entities, each with unique molecular
   signatures and functional differences ([39]5).

   Traditional transcriptome sequencing methods, which average RNA
   expression, can mask the disparities in gene expression among diverse
   cells within a group. The advent of single-cell RNA sequencing
   (scRNA-seq) has overcome this, allowing for the extraction and
   preparation of RNA libraries at the single-cell level and providing
   comprehensive transcriptomic insights. This technology aids in
   identifying cell types, states, and functions, and illuminates cellular
   heterogeneity and evolutionary pathways. Integrated with immunome
   library analysis, it offers a complete examination of the T-cell
   receptors (TCR) and B-cell receptors (BCR) of all immune cells,
   providing a detailed immune profile of gene expression in specific
   tissues or diseases. This analysis is crucial for uncovering tumor
   heterogeneity, tracing cell lineage, and understanding the complex
   mechanisms of tumor clonal evolution, paving the way for early
   detection, therapeutic stratification, prognostic evaluation, and
   monitoring of recurrence ([40]6).

   In this study, we explored the multi-stage carcinogenesis of
   hypopharyngeal mucosa by conducting scRNA-seq on tissue specimens from
   various pathological stages, aiming to analyze the composition and
   expression variations of cells within different tissue
   microenvironments (TME). This approach, simulating the progression of
   hypopharyngeal carcinogenesis, aspires to construct a comprehensive
   dynamic transcriptome map to depict the evolution of cellular and
   molecular expressions throughout tumorigenesis. The insights gained are
   expected to elucidate the trajectory and potential molecular mechanisms
   of hypopharyngeal carcinogenesis and reveal key regulatory molecules,
   aiding in the identification of predictive molecular markers for
   hypopharyngeal carcinogenesis, thereby contributing significantly to
   the advancement of HSCC analysis.

2. Materials and methods

2.1. Patient recruitment and sample collection

   From January to February 2023, five male patients, median age 57 (55-63
   years), with HSCC were recruited from the Cancer Hospital of the
   Chinese Academy of Medical Sciences. The study received approval from
   the hospital ethics committee (number: 22/454-3656), with informed
   consent obtained from each patient before examination. None had
   received any treatment (radiation, chemotherapy, or surgery) or had a
   history of other tumor diseases. Before treatment, all underwent
   laryngoscopy and gastroscopy. Based on laryngoscopic findings, multiple
   biopsies were taken from different hypopharyngeal regions of the
   patients, yielding nine samples: four from the left pyriform sinus,
   three from the right, and two from the posterior hypopharyngeal wall.
   Each sample was bifurcated; a fragment was preserved in 10% formalin
   for routine pathology, and the remainder was used for scRNA-seq and
   TCR/BCR-seq. Two experienced, blinded pathologists conducted the
   pathological assessments, agreeing on the final diagnoses.
   Histopathological diagnosis is the diagnostic gold standard. The final
   pathological diagnoses were two cases of normal squamous epithelial
   tissue, one of LGIN, three of HGIN, and three of HSCC. Subsequently, we
   classified the study categories into four groups based on pathological
   grading: Normal, LGIN, HGIN, and Tumor ([41]7, [42]8).

2.2. Preparation of single cell suspensions

   After sampling, tissues were rinsed of any blood stains with saline and
   stored in brown tubes with MACS^® Tissue Storage Solution (Miltenyi
   Biotec), transported to the laboratory at 4°C. Dissociation experiments
   began within 1 hour of arrival. Samples were segmented into 2-3 mm
   pieces and processed to single-cell suspensions using the Human Tumor
   Dissociation Kit protocol. Tissue pieces were transferred to a 5 mL
   tube with dissociation solution and dissociated at 37°C for 2 hours
   using a rotary mixer. Post-dissociation, 20 mL of DMEM was added, and
   the suspension was filtered using a 70 μm strainer. Cells were
   collected by centrifugation at 4°C. Cells were then resuspended in
   1×PBS and treated with Red Blood Cell Lysis Solution. Finally, cells
   were resuspended in 1×PBS + 0.04% BSA + 1U/μL RNase inhibitor. Cell
   viability, concentration, and aggregation rate were measured using
   AO/PI Fluorescent Dye on a LUNA-FL™ Cell Counter, with required
   standards of viability >75%, concentration between 700-1200 cells/μL,
   and aggregation rate <5%.

2.3. Library construction and Next generation sequencing

   We utilized Chromium Next GEM Single Cell 5’ Reagent Kits v2 (Dual
   Index) from 10× Genomics to construct single-cell libraries, aiming for
   10,000 captured cells per sample. After generating GEMs with the
   Chromium Controller, we adhered to the kit instructions for 5’
   single-cell RT-PCR amplification, 5’ cDNA amplification and
   purification, TCR and BCR sequence amplification and purification, and
   library construction.

   The quality and concentration of the library were assessed using a
   Qubit 4.0 and Qubit™ 1× dsDNA Assay Kits (high sensitivity) from Thermo
   Fisher Scientific. The molar concentration and fragment insertion of
   the library were evaluated using the StepOnePlus™ Real-Time PCR System
   from Applied Biosystems and detected by LabChip Touch. Sequencing was
   executed on Illumina’s Novaseq 6000 with a PE150 read length.

   For processing single-cell 5’ gene expression and TCR enrichment data,
   we employed Cell Ranger count and Cell Ranger vdj functions of
   Cellranger software (version 6.0.1) from 10× Genomics. Gene expression
   data were aligned to the human genome reference (GRCh38), and TCR and
   BCR enrichment data to the VDJ reference sequences available at 10×
   Genomics Reference Data.

2.4. Single-cell gene expression quantification and determination of cell
types

   We processed the sequencing data utilizing the Seurat R package
   (version 4.3.0) ([43]9). Data was converted into a Seurat object and
   quality filtered to exclude cells meeting specific conditions. Perform
   batch effect correction using the harmony package (version 1.0.3).
   Logarithmic normalization and linear regression were performed using
   Seurat package functions to construct the gene expression matrix. The
   COSG package (version 0.9.3) identified cell clusters, categorized into
   six primary cell types based on distinct marker genes: T cells, myeloid
   cells, B cells, epithelial cells, mast cells, and fibroblasts ([44]10).

   Subsequently, normalization, scaling, and clustering were repeated to
   further subdivide and label specific cell subtypes based on average
   expression of gene sets in each primary cell type. Cells with multiple
   labeled genes and elevated UMI counts were considered cellular
   contamination and excluded. Each cluster of a primary cell type was
   assigned a cluster identifier containing a marker gene, selected based
   on criteria including top ranking in differential gene expression
   analysis, high specificity of gene expression, and literature support
   validating their role as marker genes or functional genes associated
   with the cell type.

2.5. Pathway enrichment analysis

   We used the clusterProfiler software package ([45]11) (version 4.0.5)
   for R for Gene Ontology (GO), KEGG Pathway, and Reactome enrichment
   analyses to explore the functions and mechanisms of the identified
   cellular clusters. P-values were adjusted with the Benjamini and
   Hochberg method, with p.adjust values below 0.05 deemed statistically
   significant. Additionally, we performed Gene Set Variation Analysis
   (GSVA) using the GSEABase package ([46]12) (version 1.62.0) for R,
   primarily focusing on the 50 hallmark gene sets from the MSigDB
   database ([47]https://www.gsea-msigdb.org/).

2.6. CNV estimation

   To identify malignant cells in HSCC patients, we used inferCNV software
   (version 1.14.2) to infer CNVs from chromosomal gene expression
   patterns, setting a cutoff value at 0.1 and enabling the denoise
   option. We chose the expression profiles of T and B cells as
   references, including all epithelial cell clusters in the observation