GFPT2 is highly expressed in colon cancer and correlates with poor patient pathological features.
To explore the roles of GFPT2 in patients with colon cancer, we first investigated the expression levels of GFPT2 in colon cancer. We collected tissues from 83 colon cancer patients and detected the protein levels of GFPT2 by immunohistochemistry (IHC). Figure 1a shows typical images of GFPT2 expression levels in colon cancer tissues and corresponding paraneoplastic tissues. Our results showed that GFPT2 was an oncogene highly expressed in colon cancer tissues compared to paraneoplastic tissues (Fig. 1a, 1b).
We further investigated the correlation between the expression levels of GFPT2 and the pathological characteristics of colon cancer patients. As shown in Table 1, high levels of protein expression of GFPT2 were positively correlated with unfavorable clinicopathological features in 83 samples of colon cancer patients. Specifically, increased GFPT2 levels were positively associated with advanced TNM (χ2 = 7.803, P=0.005 <0.01), advanced primary tumor size (χ2 = 5.896, P=0.015 <0.05) and excessive lymph node metastasis (χ2 = 10.008, P=0.007 <0.01). However, protein expression of GFPT2 showed no correlation in patient's gender, age, and cell differentiation (P>0.05). We show representative images of the protein levels of different GFPT2 (negative-, positive+, positive++, positive ++++) in Figure 1c. Our results show that more advanced TNM stages show higher GFPT2 expression levels than early TNM stages (Fig. 1d). Moreover, the different GFPT2 positive levels (-, +, ++, ++++) in early TNM stages were 9.80%, 49.02%, 33.33% and 7.85%, respectively; and 3.12%, 34.38%, 37.50% and 25.00% in late TNM, respectively (Fig. 1e). In addition, more cases of lymph node metastasis were positively correlated with higher GFPT2 levels (Fig. 1f). The positive levels of different GFPT2 (-, +, ++, ++++) were 9.80%, 49.02%, 33.33% and 7.85% in N0 stage, respectively; and 5.26%, 52.63%, 21.05% and 21.06% in N1 stage, respectively; and 0.00%, 7.69%, 61.54% and 30.77% in N2 stage, respectively (Fig. 1g).
Table 1
Association between GFPT2 protein expression and clinicopathological characteristics in colon cancer tissues.
Clinicopathological characteristic
|
n
|
GFPT2 expression
|
χ2
|
P value
|
Low expression
|
High expression
|
Total number
|
83
|
42
|
41
|
|
|
Gender
|
|
|
|
0.003
|
0.958
|
Male
|
49
|
23 (46.94)
|
26 (53.06)
|
|
|
Female
|
34
|
19 (55.88)
|
15 (44.12)
|
|
|
Age (years)
|
|
|
|
0.071
|
0.790
|
<65
|
38
|
21 (55.26)
|
17 (44.74)
|
|
|
≥65
|
45
|
21 (46.67)
|
24 (53.33)
|
|
|
Cell differentiation
|
|
|
|
2.049
|
0.359
|
Well
|
21
|
9 (42.86)
|
12 (57.14)
|
|
|
Moderate
|
36
|
19 (52.78)
|
17 (47.22)
|
|
|
Poor
|
26
|
9 (34.62)
|
17 (65.38)
|
|
|
Primary tumor size
|
|
|
|
5.896
|
0.015*
|
T1-2
|
48
|
30 (62.50)
|
18 (37.50)
|
|
|
T3-4
|
35
|
12 (34.29)
|
23 (65.71)
|
|
|
Lymph node metastasis
|
|
|
|
10.008
|
0.007**
|
N0
|
51
|
32 (62.75)
|
19 (37.25)
|
|
|
N1
|
19
|
8 (42.11)
|
11 (57.89)
|
|
|
N2
|
13
|
2 (15.38)
|
11 (84.62)
|
|
|
TNM
|
|
|
|
7.803
|
0.005**
|
I/II
|
51
|
32 (62.75)
|
19 (37.25)
|
|
|
III/IV
|
32
|
10 (31.25)
|
22 (68.75)
|
|
|
**P<0.01. |
To further explore the relationship between GFPT2 expression and prognosis of colon cancer patients, we investigated the overall survival (OS) of GFPT2 in the TCGA database using Kaplan-Meier. The results showed that high GFPT2 expression was significantly associated with poorer OS in 329 patients with colon adenocarcinoma (COAD) (P=0.028 < 0.05) (Fig. 1h). We conclude that GFPT2 expression levels are associated with poor pathological characteristics and poor prognostic features in patients with colon cancer.
GFPT2 is associated with tumor-associated enrichment pathways.
To explore the expression and pathway enrichment of GFPT2 in the TCGA database, we analyzed GFPT2 and related genes in COAD cases. We performed Gene oncology (GO) analysis using the ClusterProfiler R package and obtained GFPT2 significantly enriched functions and pathways (p<0.05). We displayed 30 pathways significantly enriched with GFPT2 that were associated with many tumor-related pathways, including pathways in cancer, Focal adhesion, Adhersion molecules cams, ECM receptor interaction, JAK-STAT signaling pathway and MAPK signaling pathway (Fig. 2a-e), suggesting that GFPT2 may play an important function in tumorigenesis and progression. In cell biology, focal adhesion and ECM receptor interaction mediate the regulation of cell adhesion to the extracellular matrix (ECM) mainly through its transmission between the ECM and interacting cells (17, 18). They, therefore, play a central role in cell migration (18). Increased GFPT2 levels were positively linked to the enrichment of focal adhesion and ECM receptor interaction (Fig. 2b, 2c), indicating that GFPT2 may have the essential position in stromal cell and invasive of tumors.
The JAK-STAT signaling pathway is an important pathway in cell biology that is involved in various vital physiological functions, including immunity, cell death and tumor formation (19). Abnormalities in JAK-STAT signaling will lead to various diseases, such as skin diseases, cancer and diseases affecting the immune system (20). It has four JAK proteins, namely JAK1, JAK2, JAK3 and TYK2 (20). Seven STAT proteins are involved, which can be identified as STAT1, STAT2, STAT3, STAT4, STAT5A, STAT5B and STAT6 (20). We demonstrated a remarkable positively related correlation between GFPT2 overexpression and proteins associated with the JAK-STAT signaling pathway (Fig. 2f). JAK-STAT signaling is able to interconnect with other cellular signaling pathways, such as the MAPK/ERK pathway. JAKs phosphorylated receptors can bind to the SH2 binding domain of Grb2, an important protein in the MAPK/ERK pathway (19). In addition, MAPK (mitogen-activated protein kinase), can phosphorylate STATs, thus allowing STATs to increase gene transcription, which in turn promotes the JAK-STAT signaling pathway (19). Interestingly, our results revealed that an increase in GFPT2 promotes the MAPK signaling pathway (Fig. 2e), a result consistent with previous published reports.
Interestingly, in addition to GFPT2 high expression enriching lots of tumor-related pathways, increased GFPT2 expression also enriched a large number of immune-related pathways, including Chemokine signaling pathway, Cytokine cytokine receptor interaction, Toll-like receptor signaling pathway, NOD-like receptor signaling pathway, Natural killer cell mediated cytotoxicity, B cell receptor signaling pathway and T cell receptor signaling pathway (Fig. 2a).
Association of GFPT2 levels with tumor microenvironment
We have known that GFPT2 expression related to ECM and immune pathways, since stromal cells and immune cells are major components of the tumor microenvironment (TEM), so we speculated whether GFBP2 would be involved in TEM. To confirm our hypothesis, we observed the correlation of GFPT2 expression with stromal cell and immune cell infiltrations. Our results revealed that increased GFPT2 expression was strongly and positively associated with stromal score (R=0.89, P<0.0001) (Fig. 3a), immune score (R=0.64, P<0.0001), and ESTIMATE score (a combined score of stromal and immune cells) (R=0.82, P<0.0001) (Fig. 3a-c), indicating that GFPT2 expression affects stromal and immune cell infiltrations. Interestingly, the high levels of GFPT2 were significantly and negatively linked to tumor purity (R=0.82, P<0.0001) (Fig. 3d), suggesting that GFPT2 mainly affects the predominant activity of TEM, which has essential position in stromal cell and immune cell infiltrations.
Correlation between GFPT2 levels and stromal cell infiltration.
The TEM is composed of ECM, cancer-associated fibroblasts (CAFs), myofibroblasts, immune cells and other factors (21). In order to examine the relationship between GFPT2 levels and stromal cell infiltrations, we first surveyed the Tumor Immune Single-cell Hub (TISCH) (http://tisch.comp-genomics.org/) database (a single cell center) to investigate which cell subpopulations of GFPT2 are primarily expressed in. We explored a colon cancer single cell GSE dataset (GSC_GSE146771_Smartseq2) and found that GFPT2 was expressed in both immune and stromal cell single cell subpopulations (Fig. 4a-c). Since the tumor-associated stromal cells mainly include endothelial cells, fibroblasts and myofibroblasts. We next investigated which stromal cell components were the cell subpopulations with high GFPT2 expression in colon cancer. The findings indicated that fibroblasts were the major GFPT2 expressing cells, and the expression of GFPT2 was very high in fibroblasts compared to other cell subpopulations (Fig. 4c), indicating that GFPT2 has an important function in CAFs.
Next, we inspected the association among GFPT2 mRNA expression and biomarkers associated with CAFs. CAFs primarily express α-smooth muscle actin (α-SMA), fibronectin (FAP), cytoskeletal protein (Palladin), mucin-type protein (podoplanin), and prolyl 4-hydroxylase, while TGF-β, PDGF, HGF, FGF2 and THBS1 are the main factors that promote the activation of CAFs (22–24). S100A8/A9 could facilitate the proliferation of fibroblasts and worked in the differentiation of fibroblasts to myofibroblasts (25). In the current study, we noticed that the expression of GFPT2 had significantly positive correlation with the above mentioned markers related to CAFs (Fig. 4d).
CAFs exert huge contributions in tumorigenesis and development, which can mainly boost angiogenesis, promote the initiation of epithelial-mesenchymal transition (EMT) and affect the survival of tumor cells (26). To confirm the effects of GFPT2 expression on EMT, we examined the levels of GFPT2 in relation to EMT-related markers. The results showed that GFPT2 expression was highly correlated with EMT-related factors (Fig. 4e).
Association of GFPT2 levels with immune cell infiltration.
To better evaluate the influences of GFPT2 levels on immune cell infiltration, we calculated the correlation coefficients between GFPT2 expression and immune cells using the CIBERSORT method. The correlation between GFPT2 expression and immune cells was verified by spearman's correlation test. There was a statistically significant positive correlation between increasing GFPT2 and immune cell fractions, including Macrophages M0, Neutrophils and activated Mast cells (Fig. 5a-c). However, overexpression of GFPT2 adversely related to certain immune cell components, such as Monocytes, resting Mast cells, T cells follicular helper, plasma cells, activated CD4 T cells memory, activated NK cells, resting Dendritic cells, CD8 T cells (Fig. 5a-c).
In general, T-cell depletion is vital factor for the effectiveness of immune checkpoint blockade (27). To further discuss the relevance of GFPT2 levels to marker genes of T-cell exhaustion, our results displayed a remarkable association of GFPT2 levels with LAG3, BTLA, PDCD1, TIGIT and CTLA4 (Fig. 5d). Our analysis suggested that GFPT2 was mainly positively correlated with immunosuppressive cells, such as fibroblasts and macrophages, and therefore, we hypothesized that GFPT2 might be involved in the regulation of COAD as an immunosuppressive component. To verify this point, we conducted a correlation analysis between GFPT2 expression and key factors of immunosuppression. The results showed that GFPT2 levels were significantly and positively correlated with many immunosuppressive factors (Fig. 5e).
The relationship between GFPT2 expression levels and drug sensitivity.
We downloaded gene expression and drug sensitivity data from CellMiner and removed drugs without clinical trials or FDA approval and calculated the correlation coefficient between GFPT2 expression and drug sensitivity using the cor.test function and correlation in R language. We selected the top 16 drugs associated with GFPT2 by R value, and the results showed that GFPT2 expression was associated with a number of drugs, including Deforolimius, SGX-523, JNJ-38877605, Motesanib, Staurosporine, Itraconazole, CCT-128930, AZD-5363, AS-703569, AT-9283, Silmitasertib, Rigosertib, LY-294002, Rebimastat, PF-04217903 (Fig. 6), and the higher GFPT2 expression was linked to better sensitivity of tumor cells to these drugs. And the expression level of GFPT2 was associated with increased resistance to By-Product of CUDC-305 (Fig. 6). These results suggest that the expression of GFPT2 may be a judgment of the sensitivity of a certain class of drugs.