Somatic mutation in STAD
Figure 1 is the flow chart of this research. Firstly, we showed the top thirty frequently mutated genes in STAD samples of the TCGA cohort (Figure 2A). The ten most commonly mutated genes in the TCGA cohort are TTN, TP53, MUC16, ARID1A, LRP1B, SYNE1, FLG, FAT4, CSMD3, and PCLO. Then, we also suggested the top thirty frequently mutated genes in ICGC and found the ten most commonly mutated genes are TP53, TTN, MUC17, LRP1B, ZFHX4, ZNF814, OBSCN, MUC16, FLG, and CSMD3 (Figure 2B). All gene mutations are mainly missense mutations. We found that there were 16 genes that both mutated in the TCGA and ICGC cohorts (Figure 3A). These commonly mutant genes are TTN, TP53, MUC16, LRP1B, SYNE1, FLG, CSMD3, PCLO, FAT3, OBSCN, HMCN1, ZFHX4, SPTA1, CSMD1, USH2A, and KMT2C. We focused on these 16 commonly mutated genes in subsequent studies.
OBSCN mutation is related to TMB and clinical features in STAD
The TMB score of STAD patients ranges from 0.02-198.42/Mb (3.55/Mb as the median). We divided the STAD patients from the TCGA cohort into high-TMB and low-TMB groups based on the median. Survival analysis suggested that compared to the low-TMB group, patients in the high-TMB group had a better prognosis (Figure 3B). Our results are consistent with previous reports [10]. Compared to the WT group, all of the 16 commonly mutated genes enjoy higher TMB (Figure 4). Survival analysis showed that patients with OBSCN mutation had a favorable prognosis (Figure 5A), but no other gene mutations had a significant association with prognosis (Supplementary Figure 1). Univariate and multivariate Cox regression analyses suggested that age, TNM stage, and OBSCN mutation are independent risk factors for STAD patients (Figure 5B and 5C). Furthermore, we found that patients with OBSCN mutations tended to have higher age and lower pathological stages (Table 1). Our results indicated that the mutation of OBSCN may be a biomarker for higher TMB and better prognosis.
Table 1
Association between OBSCN mutations and clinical features.
Characteristic
|
OBSCN MT
|
OBSCN WT
|
p
|
T stage, n (%)
|
|
|
0.079
|
T1
|
8 (8.99%)
|
11 (3.99%)
|
|
T2
|
17 (20.48%)
|
61 (22.10%)
|
|
T3
|
34 (40.96%)
|
134 (76.14%)
|
|
T4
|
30 (36.14%)
|
70 (39.77%)
|
|
N stage, n (%)
|
|
|
0.056
|
N0
|
34(39.54%)
|
75 (27.99%)
|
|
N1
|
23 (26.74%)
|
73 (27.24%)
|
|
N2
|
19 (22.09%)
|
56 (20.89%)
|
|
N3
|
10 (11.63%)
|
64 (23.88%)
|
|
M stage, n (%)
|
|
|
0.593
|
M0
|
81(94.19%)
|
246 (92.48%)
|
|
M1
|
5 (5.81%)
|
20(7.52%)
|
|
Pathologic stage, n (%)
|
|
|
<0.001
|
Stage I
|
15 (18.07%)
|
36 (21.69%)
|
|
Stage II
|
31 (37.35%)
|
20 (12.05%)
|
|
Stage III
|
29 (34.94%)
|
80 (48.19%)
|
|
Stage IV
|
8 (9.64%)
|
30 (18.07%)
|
|
Gender, n (%)
|
|
|
0.104
|
Female
|
39 (42.86%)
|
94 (33.45%)
|
|
Male
|
52 (57.14%)
|
187 (66.55%)
|
|
Age, n (%)
|
|
|
0.003
|
<=65
|
25 (27.78%)
|
127 (45.85%)
|
|
>65
|
65 (72.22%)
|
150 (54.15%)
|
|
Histologic grade, n (%)
|
|
|
< 0.539
|
G1
|
1 (1.14%)
|
9 (3.27%)
|
|
G2
|
35 (39.77%)
|
102 (37.09%)
|
|
G3
|
52 (59.09%)
|
164 (59.64%)
|
|
Functional enrichment of OBSCN mutation
To explore the biological function of OBSCN mutation, we divided STAD patients into OBSCN MT group and WT group, and performed Gene Set Enrichment Analysis (GSEA) analysis. The results showed that OBSCN mutations are mainly enriched in E2F targets, MYC targets, peroxisome, MTORC1 signaling and Oxidative phosphorylation (Figure 6A), while OBSCN WT group is mainly enriched in hedgehog signaling, KRAS signaling, NOTCH signaling (Figure 6B).
Differential analysis between the OBSCN MT group and WT group
To further explore the molecular events in the OBSCN MT group, we conducted gene differential analysis between OBSCN MT and WT groups at transcription levels. Our results demonstrated that compared with the OBSCN WT group, four genes were highly expressed in the OBSCN MT group, and 26 genes were lowly expressed (Figure 7A and 7B) (Adjusted P <0.05 and log2FC> 1). Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis suggested that these 30 genes are mainly associated with the immune-related biological process such as cytokine-cytokine receptor interactions, Chemokine signaling pathway, immune cell migration, and immune response (Figure 7C). Then we studied the relationship between the expression of these 30 genes and the prognosis of STAD patients. The results showed that there is a good prognosis in STAD patients with high expression of CES3 and MMP12 (Figure 8), which were highly expressed in the OBSCN MT group. The increased expression of APOD, BEX2, CARD11, CLIC6, DKK1, and LRRN1 indicates a poor prognosis (Figure 8), which were lowly expressed in the OBSCN MT group.
The relationship between OBSCN mutations and immune-related features
Given that the differentially expressed genes between the OBSCN MT and WT group are significantly enriched in inflammation-related biological functions, we explored whether the OBSCN mutation is associated with the immune cell infiltration in the TME of STAD by the ssGSEA algorithm. Our results suggested that compared with the OBSCN WT group, the OBSCN MT group enjoyed higher T cell follicular helper (Tfh), NK cells, Th1 cells, and Th2 cells but lower mast cells (Figure 9A). In addition, the immune status, such as cytolytic activity and inflammation-promoting were higher in the OBSCN MT group (Figure 9B). Then, we explored the correlation between ONSCN mutation and infiltration of immune cells in STAD by the TISIDB database. The results suggested that the STAD samples with OBSCN mutation infiltrated more activated CD8+ T cells, central memory CD8+ T cells, and activated CD4+ T cells (Figure 9C). This indicated that STAD patients with OBSCN mutation might have more excellent antitumor immune activity. We also used the TISIDB database to explore the correlation between OBSCN mutations and immunostimulators. Results demonstrated that the expression of most immunostimulators is up-regulated in STAD with OBSCN mutations (Supplementary Figure 2A). CD70, ICOS, LAT, MICIs, TNFRSF4, TNFRSF9, TNFRSF18, TNFSF9, and TNFSF13B were obviously up-regulated in the OBSCN MT group (Supplementary Figure 2B). Our results further implied that the OBSCN mutation is associated with antitumor immune responses in STAD.
The relationship between OBSCN mutations and drug sensitivity
We explored the relationship between the mutation status of OBSCN and the expression of common immune checkpoint genes. Our results showed that compared with the OBSCN WT group, CD274 (PD-L1), CTLA-4, HAVCR2, LAG3, and PDCD1 (PD-1) expressions and MSI scores were highly expressed in the OBSCN MT group (Figure 10A and 10B). We then used the Tumor Immune Dysfunction and Exclusion (TIDE) algorithm to explore the relationship between OBSCN mutation and ICIs treatment sensitivity. Generally, patients with high TIDE scores are less sensitive to ICIs treatment. Our results show that the OBSCN mutation group had a lower TIDE score and more responders to ICIs treatment (Figure 10C and 10D). This indicated that the patients with OBSCN mutation might be more sensitive to ICIs treatment in STAD. Subsequently, we studied the relationship between OBSCN mutations and the susceptibility of STAD to standard chemotherapeutics. Our results indicated that patients with OBSCN mutations were more sensitive to cisplatin, 5-Fluorouracil, and Docetaxel (Figure 10E-G).