Expression profile of SR splicing factor family in multiple human adult tissues
SR splicing factors are RNA processing factors that play a critical role in various aspects of gene regulation including mRNA splicing, mRNA stability, transport, and translation [17]. SR splicing factors are known to be important for normal development but their precise function in adult healthy tissues remains unclear. Therefore, we examined the transcript levels of SR splicing factor family (SRSF1 to SRSF12) in 16 adult healthy tissues of different origins. We observed a tissue-specific expression of SRSF12 with increased expression in the testis (lower average delta Ct value) and lower expression in peripheral blood leukocytes amongst other healthy tissues. Similar results for SRSF 12 have been reported in the human protein atlas (https://www.proteinatlas.org/ENSG00000154548-SRSF12). SRSF6 and SRSF8 showed a lower transcript abundance in all the 16 healthy adult tissues. We also observed that amongst all the SR splicing factors, SRSF3, SRSF5, and SRSF9 were mostly expressed in all the healthy tissues (Fig. 1). Our findings suggest differential expression of SR splicing factor in different tissue types indicating their tissue-specific role (Supplementary Fig. S1). A recent study also demonstrated variable expression of SR splicing factors in undifferentiated and differentiated cells [18]. Owing to the multiple roles played by SR splicing factor family members in various cancers (Table 1), we further examined the expression of SR splicing factor family in several cancerous cell lines of different origin.
Expression profile of SR splicing factor family in cancerous cell lines of various origins
Expression levels of SRSF1-SRSF12 were examined in 27 cell lines of different origin, amongst which 23 were cancerous, one was transformed and three were non-transformed cell lines. We observed that the average delta Ct values of the SR splicing factor were quite similar in the non-transformed cell lines (VH10, HaCaT, IHH) despite their different origin – foreskin fibroblast, skin keratinocyte and hepatocyte (Supplementary Fig. S2A). Hence these three non-transformed cell lines were used as controls to study the expression profile of the SR splicing factor family in all the cancerous cell lines.
The various cancerous cell lines used in the study included oral cancer (n = 4), breast cancer (n = 4), glioblastoma (n = 2), cervical cancer (n = 3), gastric cancer (n = 2), leukemia (n = 2), pancreatic cancer (n = 2), lung cancer (n = 2), neuroblastoma (n = 1), colorectal carcinoma (n = 1) and human embryonic kidney transformed (n = 1). Details of all the cell lines under study are described in Supplementary Table S1. A scatter plot of delta Ct values of all the cancerous cell lines of different origin is provided in Supplementary Fig. S2B-L. A near significant difference (p = 0.057) in average delta Ct values of SRSF3, SRSF9 and SRSF10 was observed in oral cancer cell lines when compared to the control cell lines (Fig. 2a-c). These results could arise due to differences among the four oral cell lines. We could not find any significant difference in other splicing factor family members. These results suggested the involvement of SRSF3, SRSF9, and SRSF10 in oral cancer. Increased expression of SRSF3 has been observed in OSCC patients as well as oral cancer cell lines suggesting its role in tumour initiation, progress and maintenance [19]. Further, increased SRSF3 expression has also been reported in other cancers including cervical, lung, colon, stomach, and breast cancer [20, 21].
SRSF9 has also been implicated in cancers like hepatocellular carcinoma [22], colorectal cancer [23], and cervical cancer [24]. Frequent upregulation of SRSF10 in oral cancer has also been described earlier and it may have a role to play in its oncogenesis [25].
Interestingly, SRSF12 showed highly variable expression in various cancerous cell lines namely- SCC-4, SAS-H1, C-33 A, SH-SY5Y, and HEK293T (Supplementary Fig. S3). The role of SRSF12 in cancer is still less defined, except in an integrated analysis-based study in soft tissue sarcoma that reported its upregulation [26] along with another study on intravascular NK/T-cell lymphoma where copy number loss of the gene was reported [27]. Further studies are warranted to identify the role of SRSF12 if any in cancer biology.
Further, we observed ~ 2 fold upregulation of SRSF2 in almost all cancerous cell lines. Previous studies have also reported upregulation of SRSF2 in breast cancer [28], cervical cancer [29] and colorectal cancer [30] cell lines. We also observed upregulation (~ 1.8 fold and above) of SRSF7 in almost all the cell lines except, UM-SCC-84 and Kasumi-1 where we observed a downregulation of 1.5 fold. Elevated levels of SRSF7 in colon cancer and lung cancer patient samples as well as cell line have been previously reported [31]. Similarly, SRSF3 was also found to be upregulated by 1.6 fold and above in almost all the cell lines except, A549 and Kasumi-1 where a 1.5 fold downregulation was observed. It is important to note that few cell lines namely MDA-MB-453, MKN-45, MIA PaCa-2 and HEK293T had a significant upregulation of all the SR splicing factors at transcript level. In contrast, Kasumi-1 and A549 showed a significant downregulation of all the SR splicing factors at transcript level. MCF-7 also showed downregulation of all the splicing factors except SRSF5, SRSF7, SRSF9 and SRSF10. Figure 2d describes the relative expression profile of SR splicing factor family (SRSF1 to SRSF12) in cancerous cell lines when compared to the control cell lines. These results indicate that the SR splicing factor family members are dysregulated in most of the cancerous cell lines we have studied excluding A549 and Kasumi-1. Both cell lines had a downregulation for all SR family of splicing factors. Although we observed a lower expression of SRSF6 in lung cancer cell lines, a study group reported its overexpression in the said cancer cell line and resulted in enhanced proliferation [32].
HEK293T, a transformed cell line derived from an embryonic kidney, showed a significant upregulation in all the splicing factors SRSF1 to SRSF12 ranging from 3–10 fold; it could be because SR splicing factor expression is known to be higher during the developmental phase. Various studies on the SR splicing factors discuss their importance in the development process and their ability in regulating the embryonic pluripotent stem cells [33]. Studies on the conditional deletion of SR splicing factor in animal models have resulted in severe developmental defects stating that these proteins are important in normal development [17].
Expression profile of SR splicing factor family in patients with OSCC
Altered expression of SR splicing factor family members in oral cancer cell lines tempted us to examine their role in OSCC patients. A total of forty patients were enrolled in the study and were divided into 3 categories based on their clinical diagnosis: Pre-cancer (n = 15), Early Cancer (n = 11), and Late cancer (n = 14). Twenty-six age-matched controls were also enrolled in this study. The demographic characteristics of the study subjects are given in Table 2.
Table 2
Demographic characteristics of the oral squamous cell carcinoma patient tissue biopsy collected. Data is presented in the form of Mean ± SD and mean percent values. PDSCC- Poorly differentiated squamous cell carcinoma, MDSCC- Moderately differentiated squamous cell carcinoma, WDSCC- Well differentiated squamous cell carcinoma, SCC- Undifferentiated squamous cell carcinoma.
Parameters | Contols (n = 26 ) | OSCC Patients |
---|
Pre-cancer (n = 15) | Early Cancer (n = 11) | Late Cancer (n = 14) |
---|
Number | 26 | 15 | 11 | 14 |
Age | 48 ± 11.5 | 47.5 ± 9.8 | 48.2 ± 11.8 | 46.1 ± 11.3 |
% Male | 21(80.8) | 14(93.99) | 11(100) | 14(100) |
% Female | 5(19.2) | 1(6.67) | 0 | 0 |
Cancer stage/differentiation | | | PDSCC (n = 1) MDSCC (n = 5) WDSCC (n = 1) SCC (n = 4) | PDSCC (n = 0) MDSCC (n = 7 ) WDSCC (n = 2 ) SCC (n = 5) |
%Tobacco Consumers | 16(61.5) | 8(53.3) | 8(72.7) | 10(71.4) |
% Non- Consumers | 10(38.4) | 7(46.7) | 3(27.3) | 4(28.6) |
We observed a significant difference in the expression levels (average delta Ct) of SRSF1, SRSF3, SRSF7, SRSF9, SRSF10 and SRSF11 in late cancer patient samples compared to that of the controls (Fig. 3). Further, there was a near significant difference (p = 0.058) in SRSF10 levels in the late cancer patient group compared to pre-cancer patient group. Similar results were reported by a group studying SRSF3 expression in OSCC and observed its up-regulated expression in patients with higher grade cancer indicating their association in development of the disease [34]. A comprehensive study in OSCC on alternative splicing signatures revealed that the higher expression of SRSF9 was related to poor outcome of the disease [35]. Another study showed that SRSF10 overexpression was associated with poor outcome of the disease and that SRSF10 acts as an oncogenic driver [25]. These studies align with our data suggesting that there are more than one SR splicing factor associated with the progression of OSCC. There was no significant difference in transcript levels of SRSF2, SRSF4, SRSF5, SRSF6, SRSF8, SRSF12 among the three patient groups. Although a group studying SRSF5 found it to be overexpressed in OSCC and reported that it is essential for cell proliferation and development of cancer [36]. A study on HNSCC reported a positive correlation of expression in a VEGF isoforms (VEGFAxxx and VEGFA165b ), a proangiogenic factor linked to tumour growth and metastasis with SRSF1, SRSF5 and SRSF6, which also may have contributed alternative splicing of VEGFA gene. They also observed upregulation of SRSF1 and SRSF5, but in pharyngeal tumour. Lower expression of SRSF6 was also reported in advanced oral tumours [37]. Scatter plot of delta Ct values of all the controls, pre-cancer, early cancer and late cancer samples observed in our study is provided in Supplementary Fig. S4.
In addition, when we compared the fold change in the patient groups compared to the controls, we observed that the transcript levels of SR splicing factors had a lower expression in pre-cancer patients and this expression was upregulated in the late-cancer patient group. Early cancer patient groups showed a mixed expression profile. The relative expression of splicing factors especially, SRSF10 and SRSF11was found to be upregulated two-fold and above in late cancer samples compared to that of the control group. We also observed an upregulation of about 1.6 fold in SRSF1, SRSF2, SRSF3, and SRSF6 in the late cancer patient group compared to that of the controls. Similarly, upregulation by 1.6 fold and above was also observed in early cancer patient groups compared to controls in SRSF8, SRSF10, SRSF11 and SRSF12. We observed an increase in the expression of SRSF3, SRSF10 and SRSF11 from pre-cancer to early cancer to late cancer indicating that these splicing factors may play a role in cancer progression. Figure 4a describes the relative expression profile of SR splicing factor family (SRSF1 to SRSF12) in patients with pre-cancer, early cancer and late cancer stages of oral cancer when compared to controls. We also found a near significant difference in the fold change of the late cancer patient group compared to the pre-cancer patient group in SRSF3 and SRSF11 with an upregulation by 1.7 fold in both the splicing factors (Fig. 4b, c). It can be concluded from these results that more than one SR splicing factor may be involved in the oral squamous cell carcinoma progression.
Role of SRSF3, SRSF9 and SRSF10 in oral squamous cell carcinoma
Interestingly, we found SRSF1, SRSF3, SRSF7, SRSF9, SRSF10, and SRSF11 had a significant difference in their relative expression in patients with late-stage oral squamous cell carcinoma compared to controls (Fig. 3) and were upregulated by 1.5 fold and above. The similar expression profile was observed in oral cancer cell lines for SRSF3, SRSF9 and SRSF10; it was found to have a significant difference in their expression levels compared to control cell lines (Fig. 2a-c) and were upregulated by 1.5 fold and above in cell lines SCC-4, UM-SCC-84 and CAL-33. It can be concluded that upregulation of splicing factors SRSF3, SRSF9 and SRSF10 may be involved in oral cancer pathogenesis.
Over a period of time splicing modulators have emerged as a new class of tumor suppressor genes and oncoproteins. SRSF3 known to be a proto-oncogene is upregulated in various tumours. Mis-regulated splicing resulting from SRSF3 has been linked to various cancers including oral cancer [38] and has been associated with cell cycle, apoptosis, cell proliferation, drug resistance, cell migration and invasion among other functions [16]. Thereby, maintaining normal levels of SRSF3 or other splicing factors in a specific cell type is critical for preventing cell transformation [19]. In oral cancer with moderate or high grade dysplasia, overexpression of SRSF3 has been positively associated with initiation and development of the disease [34]. SRSF3 may also be involved in inhibiting autophagy by repressing FoxO1 and p65 and BECN1; an autophagy regulator, which may result in early development of OSCC [39, 40]. Another study in head and neck cancer revealed that the under expression of SRSF3 along with RBMX results in better outcome of the disease [41]. These evidences suggest that SRSF3 is definitely a promising new therapeutic target for a better management of OSCC. SRSF9 is another proto-oncogene frequently upregulated in many tumours. It has been found to promote Wnt signalling by increasing β-catenin accumulation and might result in cell transformations [42]. Increased expression of SRSF10 has been frequently observed in cancers like colorectal cancer [43], liver cancer, renal cancer and hepatocellular carcinoma and is mostly associated with poor outcome of the disease. There are less number of reports associating SRSF9 and SRSF10 with OSCC.
In conclusion, our findings strongly suggests tissue specific expression of different SR splicing factor family members in healthy adult tissues. Further, a dysregulated expression profile of these splicing factors was observed in various cancerous cell lines of different origin. Expression changes in the splicing factors were observed at different stages of oral cancer- Pre-cancer, Early and Late cancer. Our data strongly reflects the involvement of SRSF3, SRSF9 and SRSF10 in oral cancer pathology.