Cancer Res: Tang Fu-Hui/Cui Heng/Ji Jia-kui collaborate to reveal molecular regulatory mechanisms for high-grade serous ovarian cancer

Ovarian cancer is one of the most fatal gynecological tumors, with high-grade serous ovarian cancer (HGSC) being the most common subtype

Intratumoral heterogeneity is one

In order to fully reveal the molecular characteristics and interplay between different omics levels of high-grade serous ovarian cancer in the process of occurrence and development, Tang Fuhui Research Group of Peking University Biomedical Frontier Innovation Center cooperated with Cui Heng Research Group of Peking University People's Hospital and Ji Jiakui Research Group of Tsinghua University to conduct high-precision single-cell multi-omics sequencing analysis (genome (copy number variation), chromatin status group, DNA methylation group, transcriptome

The research results were recently published online in the journal Cancer Research under the title of Single-cell dissection of the multiomic landscape of high-grade serous ovarian cancer

Figure 1.

The main findings of the study are:

1.

Tumor tissue is composed of many types of cells, and accurate identification of cells is an important part

By comparing differences in gene expression between tubal epithelial cells (FTEs) and primary tumor cells, the study found that pathways associated with eukaryotic transcription and protein ubiquitination were significantly downregulated in high-grade serous ovarian cancer tumor cells, suggesting drastic changes

To further clarify the typical characteristics of high-grade serous ovarian cancer, the study also systematically compared

Figure 2.

2.

High-grade serous ovarian cancer cells often experience an increase in copy number on chromosome 8, so the study analyzed whether the genes (361) specifically upregulated in tumor cells were associated

3.

DNA methylation is one of the most important epigenetic modifications that undergo extensive reprogramming during tumorigenesis and can affect gene expression

The study found that the vast majority of ovarian cancer patients (9/11) had significantly reduced genomic methylation levels in tumor cells, suggesting that large-scale DNA demethylation is also a distinctive feature

Further analysis of genes regulated by promoter methylation shows that metallothionein-bound metal-related genes and interferon pathway-related genes are likely to be similarly affected

The study used MIMICCs to analyze the effects
of methylation of remote regulatory elements (e.
g.
, enhancers) on gene expression.
A total of 16,941 regulatory region-target gene pairs were detected in the
study.
Among them, 53.
9% were inversely correlated
with the expression of target genes.
Most target genes may have fewer than 5 regulatory regions regulated by DNA methylation
.
The above results suggest that some genes may be regulated not only by methylation in their promoter regions, but also by methylation in their distal regulatory regions
.
For example, the expression of the gene IFI27 (Figure 3) specifically upregulated by tumor cells is negatively correlated with its promoter DNA methylation and also negatively correlated with the methylation of its distal regulatory element (approximately 235 kb from the IFI27 transcription initiation site
).
In addition, this distal regulatory element region overlaps with a known enhancer region whose flanking region exhibits a higher chromatin openness in primary tumor cells than fallopian tube epithelial cells, highlighting its potential regulatory effect
on IFI27 expression.

Figure 3.
Schematic diagram showing methylation in the distal region (WCG7649168: chr14: 94,577,079-94,583,033) may regulate the expression of IFI27

4.
It was found that the regulatory activity of important transcription factors such as FOXK1, TFAP2C, NR2F6, DDIT3 and other important transcription factors was significantly enhanced
during the tumorigenesis of high-grade serous ovarian cancer.

The specific binding of transcription factor proteins to gene cis-regulatory elements is the basic mechanism of transcriptional regulation, while the binding of transcription factors is usually regulated
by the state of chromatin.
The study found that the global chromatin status of tumor cells is more open than that of tubal epithelial cells, while the gene promoter region (the region adjacent to the transcriptional initiation site) is less open than that of tubal epithelial cells (the open area is narrower and less open
).

To explore potential key regulators in the development of high-grade serous ovarian cancer tumors, the study used chromVAR for transcription factor binding motif enrichment analysis
of chromVAR.
A total of 107 transcription factor motifs are more open in tumor cells, while the motifs of another 126 transcription factors are more closed
.
To further explore which transcription factors are more likely to be involved in tumor-specific gene expression regulation, the study used Lisa to infer transcription factors
that regulate tumor-specific expression genes.
Among these 107 transcription factors, the expression levels of foxk1, TFAP2C, NR2F6, and DDIT3 themselves were significantly higher in tumor cells than in fallopian tube epithelial cells (Figure 4
).
FOXK1 is an important regulator of glycolysis induction, considering that the glycolytic pathway is upregulated in high-grade serous ovarian cancer, and the binding motif of its transcriptional regulator FOXK1 shows high accessibility, the study believes that FOXK1 is likely to be a candidate therapeutic target for
high-grade serous ovarian cancer.

Figure 4.
Chromatin accessibility and transcriptome information were used to comprehensively explore transcription factors that may play an important regulatory role in high-grade serous ovarian cancer

5.
It was found that the gene expression pattern of tumor cells in high-grade serous ovarian cancer during intraperitoneal metastasis was basically unchanged
.

After accurately identifying tumor cells, the study used breakpoint information in chromosomes to construct different genetic lineages
of tumor cells in the same patient.
Although high-grade serous ovarian cancers carry a large number of genomic copy number variants, their genetic lineage composition is relatively simple, with most patients (8/9) having only 1-2 genetic lineages
in tumor cells.

After analyzing the differences between tubal epithelial cells and primary tumor cells of high-grade serous ovarian cancer, the study next explored the differences
in gene expression between primary tumor cells and metastatic tumor cells of the same genetic lineage.
The results showed that in most patients (5/6), only a very small number of gene expressions of tumor cells of the same genetic lineage changed after metastasis to the abdominal cavity (but in one patient who metastasized to lymph nodes, tumor cells of the same genetic lineage had significant changes in the expression of more than two thousand genes before and after metastasis; Figure 5).

This suggests that although tumor cells have metastasized to the abdominal cavity, the microenvironment of the abdominal cavity does not strongly affect the overall gene expression patterns of tumor cells, and suggests that the ability of tumor cells to metastasis to the abdominal cavity is likely to be obtained in the early stages of
tumorigenesis.

Although there are only a few genes with differential expression between celiac metastatic tumors and primary tumors, these differentially expressed genes provide clues to the discovery of key genes that promote intraperitoneal metastasis
.
The study validated the effect
of specific downregulation of the HSPA6 gene during metastasis of high-grade serous ovarian cancer.
The results showed that excessive expression of HSPA6 inhibited the migration and invasion of cancer cells, suggesting that the specific downregulation of HSPA6 promoted the tumor metastasis process of high-grade serous ovarian cancer (Figure 5
).

Figure 5.
Functional verification of differential genes and HSPA6 between primary and metastatic tumors

6.
It was found that the DNA methylation pattern of tumor cells in the process of peritoneal metastasis of high-grade serous ovarian cancer was basically unchanged
.

The study found that the global DNA methylation level of tumor cells remained essentially unchanged during tumor metastasis in most patients with high-grade serous ovarian cancer (Figure 6
).
Therefore, DNA methylation patterns can be used to infer the genetic lineage
of tumor cells in the same patient.
Unsupervised stratified clustering of DNA methylation confirmed this, and the results showed that genetic lineages of tumor cells consistent with the genome copy number variant were inferred based on DNA methylation patterns (Figure 6
).

Figure 6.
Cancer cells maintain DNA methylation levels and patterns during metastasis

7.
It was found that the differences in gene expression between tumor cells of different lineages in primary tumors can be used to find candidate genes
involved in tumor metastasis.

Reconstruction of tumor cytogenetic lineages can reveal intratumoral heterogeneity and provide more accurate tumor metastasis analysis
.
Tumor cells with different genetic lineages in primary tumors not only have different genomic copy number variations, different DNA methylation levels, and different gene expression patterns, but are also likely to have different metastatic potential
.
For example, patients with OC09 and OC14 both had two distinct tumor cytogenetic lineages (lineages A1 and A2
).
Tumor cells of lineage A1 metastasize, while tumor cells of lineage A2 do not metastasize
.
Differential gene expression analysis showed that A1 lineage tumor cells in OC09 patients and OC14 patients specifically upregulated the expression of CCN1 and HSP90AA1 compared to A2 lineages (Figure 7
).
After knocking out the CCN1 gene using the CRISPR-Cas9 system, the migration and invasion ability of ovarian cancer cells was significantly reduced
.
After specific inhibition of HSP90AA1 activity using the HSP90α inhibitor TAS-116, the migration and invasion ability of ovarian cancer cells was significantly reduced (Figure 7
).
The above results suggest that CCN1 and HSP90AA1 are likely to be involved in the tumor metastasis process of high-grade serous ovarian cancer and may serve as potential therapeutic targets for
ovarian cancer.

Figure 7.
Functional verification of CCN1 and HSP90AA1

In summary, this study is the first to systematically explore the molecular characteristics
of genomic copy number variation, DNA methylation group, chromatin status group, and transcriptome in high-grade serous ovarian cancer with single-cell and single-base resolution.
The molecular changes of high-grade serous ovarian cancer were comprehensively analyzed, and potential therapeutic targets such as CCN1 and HSP90AA1 were found, which helped to further understand the molecular characteristics of the disease and provide a theoretical basis
for the diagnosis and treatment of the disease.

Wang Yiorange, phD student of The School of Life Sciences and Biomedical Frontier Innovation Center of Peking University, Xie Haoling, PhD student, Professor Chang Xiaohong and Professor Li Yi of the Department of Gynecology of Peking University People's Hospital, as well as Dr.
Hu Wenqi and PhD student Li Mengyao of Tsinghua University are the tie-in first authors
of the paper.
Professor Tang Fuhui of the Biomedical Frontier Innovation Center of Peking University, Professor Cui Heng of the Department of Gynecology of Peking University People's Hospital, and Professor Ji Jiakui of Tsinghua University are the co-corresponding authors
of the paper.
The research project is supported
by the National Natural Science Foundation of China and the Beijing Advanced Innovation Center for Future Genetic Diagnostics (ICG).

Original source:

Yicheng Wang， et al.
Single-cell dissection of the multiomic landscape of high-grade serous ovarian cancer.
Cancer Res CAN-21-3819.