Downregulation of PGGHG and ODF3B gene expression in RCC
The expression levels of mRNAs were assessed in 40 tumor samples and their tumor's adjacent normal tissues using qRT-PCR. As shown in the Fig. 2A, the expression of PGGHG had significantly lower in tumor tissues (median = 0.378) compared to their adjacent normal tissues (median = 1.120) (Pvalue = 0.034). Also, the expression level of ODF3B was significantly lower in tumors tissues (median = 0.625) in comparison to the tumor's adjacent normal tissues (median = 0.977) (Pvalue = 0.042, Fig. 2B).
Association between PGGHG and ODF3B gene expression and clinicopathological and demographic features of RCC
Next, we investigated the relationship between PGGHG and ODF3B expression levels and the clinicopathologic status of patients with renal cell carcinoma. Our results demonstrated that ODF3B had lower expression in RCC patients who had tumor size > 4cm (Pvalue = 0.060, Fig. 3A). When we divided fold changes into two high and low expressions groups, chi-square test proved the significant relation between tumor size and ODF3B expression (Pvalue = 0.011, Table 3). Comparison of PGGHG expression level and tumor size did not show a significant difference (Pvalue = 0.744).
Table 3
The association of ODF3B and PGGHG expression levels with demographic and clinicopathological factors in RCC patients, according to categorizing patients in to two groups of high and low expressions.
|
ODF3B level
|
PGGHG level
|
Low
|
High
|
P-value
|
Low
|
High
|
P-value
|
N
|
%
|
N
|
%
|
N
|
%
|
N
|
%
|
Tumor size
|
≤ 4
|
6
|
30.0%
|
14
|
70.0%
|
0.011
|
9
|
45.0%
|
11
|
55.0%
|
0.527
|
> 4
|
14
|
70.0%
|
6
|
30.0%
|
11
|
55.0%
|
9
|
45.0%
|
Tumor focality
|
focal
|
11
|
44.0%
|
14
|
56.0%
|
0.327
|
12
|
48.0%
|
13
|
52.0%
|
0.744
|
unifocal
|
9
|
60.0%
|
6
|
40.0%
|
8
|
53.3%
|
7
|
46.7%
|
Fuhrman nuclear grade
|
1
|
4
|
80.0%
|
1
|
20.0%
|
0.157
|
4
|
80.0%
|
1
|
20.0%
|
0.093
|
2
|
12
|
57.1%
|
9
|
42.9%
|
10
|
47.6%
|
11
|
52.4%
|
3
|
2
|
22.2%
|
7
|
77.8%
|
|
2
|
22.2%
|
7
|
77.8%
|
|
4
|
2
|
40.0%
|
3
|
60.0%
|
|
4
|
80.0%
|
1
|
20.0%
|
|
Lymph vascular perineural invasion
|
No
|
4
|
50.0%
|
4
|
50.0%
|
1.000
|
5
|
62.5%
|
3
|
37.5%
|
0.429
|
Yes
|
16
|
50.0%
|
16
|
50.0%
|
15
|
46.9%
|
17
|
53.1%
|
BMI
|
≤ 25
|
5
|
55.6%
|
4
|
44.4%
|
0.793
|
6
|
66.7%
|
3
|
33.3%
|
0.508
|
25–29
|
12
|
46.2%
|
14
|
53.8%
|
12
|
46.2%
|
14
|
53.8%
|
≥ 30
|
3
|
60.0%
|
2
|
40.0%
|
2
|
40.0%
|
3
|
60.0%
|
Background disease
|
None
|
11
|
52.4%
|
10
|
47.6%
|
0.727
|
12
|
57.1%
|
9
|
42.9%
|
0.313
|
High blood pressure
|
7
|
58.3%
|
5
|
41.7%
|
6
|
50.0%
|
6
|
50.0%
|
diabetes
|
1
|
33.3%
|
2
|
66.7%
|
1
|
33.3%
|
2
|
66.7%
|
High blood pressure and diabetes
|
1
|
33.3%
|
2
|
66.7%
|
0
|
0.0%
|
3
|
100.0%
|
|
prostate problems
|
0
|
0.0%
|
1
|
100.0%
|
1
|
100.0%
|
0
|
0.0%
|
|
Kidney disease
|
No
|
14
|
48.3%
|
15
|
51.7%
|
0.723
|
15
|
51.7%
|
14
|
48.3%
|
0.723
|
Yes
|
6
|
54.5%
|
5
|
45.5%
|
5
|
45.5%
|
6
|
54.5%
|
Kidney stone
|
No
|
10
|
47.6%
|
11
|
52.4%
|
0.752
|
10
|
47.6%
|
11
|
52.4%
|
0.752
|
Yes
|
10
|
52.6%
|
9
|
47.4%
|
10
|
52.6%
|
9
|
47.4%
|
Since ccRCC is the most common of RCC and oncocytoma is the least common type in our study community (Table 1), our analysis showed that PGGHG expression level was higher in oncocytoma type in comparison to other type of RCC (Pvalue = 0.017, Fig. 3B). Similarly, the relationship between the ODF3B expression level and the tumor type showed significant. The expression level of ODF3B was higher in oncocytoma type (Pvalue = 0.004, Fig. 3C).
Bioinformatic Analysis
Functional enrichment analysis
Using Enrichr, we found down-regulation of ODF3B upon RUNX1 Knock-out mouse, PPARG deficiency mouse, SETDB1 Knock-down in THP1 human cells, IRF9 Knock-down in human cells, NANOG over-expressed mouse, and CREM Knock-out mouse (Table 4). These data can suggest the possible upstream transcription factors of ODF3B which regulate its expression in biological pathways. Among GO ontologies using Enrichr, ODF3B was found to be involved in cytoskeleton (GO: 0005856, p-value: 0.03000) in terms of cellular component (CC).
Table 4
Down-regulation of ODF3B mRNA upon transcription perturbations.
Index
|
Name
|
P-value
|
Adjusted p-value
|
Odds Ratio
|
Combined score
|
1
|
RUNX1 KO MOUSE GSE34292 CREEDSID GENE 2396 DOWN
|
0.00895
|
0.01826
|
19821
|
93478.06
|
2
|
PPARG DEFICIENCY MOUSE GSE23421 CREEDSID GENE 1231 DOWN
|
0.0103
|
0.01826
|
19794
|
90569.85
|
3
|
SETDB1 KD THP1 HUMAN GSE103409 SGRNA6DAY4 RNASEQ DOWN
|
0.01105
|
0.01826
|
19779
|
89111.01
|
4
|
IRF9 KD HUMAN GSE50588 CREEDSID GENE 2809 DOWN
|
0.0132
|
0.01826
|
19736
|
85408.48
|
5
|
NANOG OE MOUSE GSE48370 CREEDSID GENE 2980 DOWN
|
0.01565
|
0.01826
|
19687
|
81844.62
|
7
|
CREM KO MOUSE GSE29593 CREEDSID GENE 2559 DOWN
|
0.01905
|
0.01905
|
19619
|
77704.89
|
In relation to the PGGHG, Enrichr revealed its down-regulation upon ZNF503 shRNA in H1 human cells, NFYC Knock-out in mouse, BCL6 Knock-down in human cells, POU2AF1 over-expression in mouse, EPAS1 Knock-down in HUVEC human cells, NFYA Knock-out in mouse, FOXP3 activation in human cells, PPARG over-expression in mouse, and GATA4 over-expression in A549 human cells (Table 5).
Table 5
Down-regulation of PGGHG mRNA upon transcription perturbations.
Index
|
Name
|
P-value
|
Adjusted p-value
|
Odds Ratio
|
Combined score
|
1
|
ZNF503 SHRNA H1 HUMAN GSE69618 DAY2 RNASEQ DOWN
|
0.00795
|
0.01847
|
19841
|
95923.19
|
3
|
NFYC KD MOUSE GSE56838 CREEDSID GENE 1343 DOWN
|
0.0093
|
0.01847
|
19814
|
92684.96
|
5
|
BCL6 KD HUMAN GSE45838 CREEDSID GENE 1815 DOWN
|
0.0118
|
0.01847
|
19764
|
87745.54
|
6
|
POU2AF1 OE MOUSE GSE12421 CREEDSID GENE 2313 DOWN
|
0.0122
|
0.01847
|
19756
|
87051.43
|
7
|
EPAS1 KD HUVEC HUMAN GSE62974 RNASEQ DOWN
|
0.0136
|
0.01847
|
19728
|
84784.91
|
8
|
NFYA KD MOUSE GSE56838 CREEDSID GENE 1341 DOWN
|
0.0138
|
0.01847
|
19724
|
84479.77
|
9
|
FOXP3 ACTIVATION HUMAN GSE41087 CREEDSID GENE 1054 DOWN
|
0.01385
|
0.01847
|
19723
|
84404.16
|
10
|
PPARG OE MOUSE GSE10192 CREEDSID GENE 2733 DOWN
|
0.0161
|
0.01932
|
19678
|
81249.36
|
12
|
GATA4 OE A549 HUMAN GSE85001 RNASEQ DOWN
|
0.02545
|
0.02545
|
19491
|
71552.36
|
The interesting result was that while PPARG over-expression in mouse resulted in PGGHG down-regulation, PPARG deficiency mouse showed ODF3B down-regulation. These data may indicate PPARG as the same upstream transcription factor affecting ODF3B and PGGHG.
Among GO ontologies using Enrichr, PGGHG showed involvement in hydrolase activity, hydrolyzing O-glycosyl compounds (GO: 0004553, p-value: 0.001850, in terms of molecular function (MF).
As shown in Fig. 1, protein networks of PGGHG and ODF3B are connected and correlated. Therefore, we investigated the GO terms for all proteins involved in these networks to shed light into the identification of possible pathways of PGGHG and ODF3B.
Enrichr revealed significant KEGG, biological process (BP), MF, and CC for proteins involved in these networks given in Fig. 4. However, the highest significant terms were as follow: KEGG: Starch and sucrose metabolism (p-value: 3.058e-43); BP: glucan biosynthetic process (GO:0009250, p-value: 5.259e-24), glycogen biosynthetic process (GO:0005978, p-value: 5.259e-24), and glycogen metabolic process (GO:0005977, p-value: 8.723e-27); MF: UDP-glucosyltransferase activity (GO:0035251, p-value: 1.319e-8) and 1,4-alpha-oligoglucan phosphorylase activity (GO:0004645, p-value: 2.305e-7); CC: chromosome (GO:0005694, p-value: 1.921e-8). All information and proteins involved in these terms are provided in Supplementary file 1 and 2. These data may propose the possible roles of PGGHG and ODF3B, directly or indirectly, in these GO terms and KEGG pathways.
Overall Survival
To evaluate survival analysis, we used the Kaplan–Meier method based on observed survival times. However, we were not able to predict the prognosis of renal cell carcinoma patients based on PGGHG (ATHL1) and ODF3B gene expression using this data (Fig. 5A, 5B). The small proportion of participants evaluated, low duration of follow up and undetected exact cause of death could be some of the reasons.