Gene expression profiles in stage I uterine serous carcinoma in comparison to grade 3 and grade 1 stage I endometrioid adenocarcinoma

Abstract

Background: Endometrial cancer is the most common gynecologic malignancy in the developed countries. Clinical studies have shown that early stage uterine serous carcinoma (USC) has outcomes similar to early stage high grade endometrioid adenocarcinoma (EAC-G3) than to early stage low grade endometrioid adenocarcinoma (EAC-G1). However, little is known about the origin of these different clinical outcomes. This study applied the whole genome expression profiling to explore the expression difference of stage I USC (n = 11) relative to stage I EAC-G3 (n = 11) and stage I EAC-G1 (n = 11), respectively.

Methodology/principal finding: We found that the expression difference between USC and EAC-G3, as measured by the number of differentially expressed genes (DEGs), is consistently less than that found between USC and EAC-G1. Pathway enrichment analyses suggested that DEGs specific to USC vs. EAC-G3 are enriched for genes involved in signaling transduction, while DEGs specific to USC vs. EAC-G1 are enriched for genes involved in cell cycle. Gene expression differences for selected DEGs are confirmed by quantitative RT-PCR with a high validation rate.

Conclusion: This data, although preliminary, indicates that stage I USC is genetically similar to stage I EAC-G3 compared to stage I EAC-G1. DEGs identified from this study might provide an insight in to the potential mechanisms that influence the clinical outcome differences between endometrial cancer subtypes. They might also have potential prognostic and therapeutic impacts on patients diagnosed with uterine cancer.

Figure 1. Hierarchical clustering of patient samples based on differentially expressed genes with at least 1.5-fold change obtained from USC versus EAC-G1 and USC versus EAC-G3, respectively.

In clustering heat map, red means up-regulated while green means down regulated. A) USC versus EAC-G1: In clustering dendrogram, blue stands for USC samples while yellow stands for EAC-G1 samples. B) USC versus EAC-G3: In clustering dendrogram, blue stands for USC samples while yellow stands for EAC-G3 samples.

Figure 2. Venn diagrams showing the number of differentially expressed genes (DEGs) derived from USC vs. EAC-G3 comparison is consistently less than that derived from USC vs. EAC-G1 comparison.

A) DEGs as defined by P-value <0.01. B) DEGs with at least 1.5-fold change. C) DEGs with at least 2-fold change.

Figure 3. Enriched function terms for differentially expressed genes (DEGs) with at least 1.5-fold change identified by microarray.

Function enrichment analyses are conducted using NCBI DAVID API server. The number following each enriched functional term is the number of annotated DEGs. A) Enriched functional annotation for DEGs specific to USC vs. EAC-G1 comparison. B) Enriched functional annotation for DEGs specific to USC vs. EAC-G3 comparison. C) Enriched functional annotation for DEGs shared by USC vs. EAC-G1 and USC vs. EAC-G3 comparison.

Figure 4. Log2 fold change of microarray estimated expression for genes from an investigated pathway are plotted as a cumulative distribution function (in red color) and compared with the corresponding cumulative distribution function for the rest of genes (i.e., not belonging to the investigated pathway, in blue color).

A) P53 signaling pathway in USC vs. EAC-G3 comparison. The two cumulative distribution functions are significantly different (P = 4.5e-3) by the Kolmogorov-Smirnov test. B) P53 signaling pathway in USC vs. EAC-G1 comparison. C) PTEN signaling pathway in USC vs. EAC-G3 comparison. D) PTEN signaling pathway in USC vs. EAC-G1 comparison.

Figure 5. qRT-PCR validations of selected differentially expressed genes with at least 1.5-fold expression change identified by microarray.

Blue bar is the fold change estimated by qRT-PCR, while red bar is the fold change estimated by microarray. The fold change is shown in log2 scale. A–B) For USC vs. EAC-G3, 15 of 15 selected genes showed at least 1.5-fold change in qPCR estimated expression levels and was concordant with the microarray data. C–D) For USC vs. EAC-G1, 7 of 9 selected genes showed at least 1.5-fold change in qPCR estimated expression levels and was concordant with the microarray data. The two genes with less than 1.5-fold change in expression level based on qPCR are MXN and TBX2.