3.1 alpha1H treatment affects protein-coding gene expression in patients with bladder cancer
We collected gene expression sequencing data from clinical experimental patients with alpha1H, a new bladder cancer drug, GSE172112, with simple preliminary preprocessing and statistics, in which 39 patients using the new drug and 42 patients using placebo were tested, and gene expression data from a total of 81 samples were tested. Subsequently, using the R package "DESeq2", we transformed the gene read count data detected by the samples into the corresponding expression values, combined with the RNA classification information in the R package "biomaRt" package, yielding a total of 20,356 PCG and 12,382 lncRNA (including 7,109 lincRNA and 5,273 antisense).
First, for 20,356 protein-coding genes, using the patient drug profile, we performed differential expression analysis between patients using the new drug alpha1H versus patients using placebo. We counted the mean expression of each protein-coding gene in both patient groups, calculated the corresponding FC values, and recorded the Wilcoxon rank-sum test p-values, finally obtaining 781 differentially expressed genes with a p-value < 0.01, containing 396 significantly up-regulated genes and 385 significantly downregulated genes (Table 1, Fig. 1).
Table 1
396 KEGG pathway associated with significantly upregulated genes
Term
|
Count
|
PValue
|
hsa 04550:Signaling pathways regulating pluripotency of stem cells
|
10
|
3.52E-04
|
hsa 04350:TGF -beta signaling pathway
|
6
|
0.010437931
|
hsa 04110:Cell cycle
|
7
|
0.013608859
|
hsa 03018:RNA degradation
|
5
|
0.033388814
|
hsa 03420:Nucleotide excision repair
|
4
|
0.037922436
|
hsa 00280:Valine, leucine and isoleucine degradation
|
4
|
0.037922436
|
hsa 04330:Notch signaling pathway
|
4
|
0.039997665
|
Furthermore, we distributed differentially expressed protein-coding genes to functionally annotated significantly up-and significantly down-regulated genes.among, For the 394 genes that were significantly upregulated, We found that their function is mainly associated with the following KEGG pathway: "hsa04550: Signaling pathways regulating pluripotency of stem cells" regulates signaling in stem cells, "hsa04350: TGF-beta signaling pathway" TGF-beta signaling pathway, "hsa04110: Cell cycle" cell cycle, et al. (Table 1, Fig. 2).
But for the 385 significantly downregulated genes, we found that their function was mainly with (Table 2,Fig. 3). like the following KEGG pathway-associated: "hsa 04010:MAPK signaling pathway" MAP K signaling pathway, "hsa04145:Phagosome" phagosomal, "hsa04668:TNF signaling pathway" TNF signaling pathway
Table 2
KEGG pathway associated with 385 significantly downregulated genes
Term
|
Count
|
PValue
|
hsa 05132:Salmonella infection
|
10
|
1.45E-04
|
hsa 04010:MAPK signaling pathway
|
15
|
0.002761576
|
hsa 05152:Tuberculosis
|
12
|
0.00324906
|
hsa 04145:Phagosome
|
10
|
0.009423262
|
hsa 04060:Cytokine -cytokine receptor interaction
|
13
|
0.013038165
|
hsa 04668:TNF signaling pathway
|
8
|
0.013680707
|
hsa 04910:Insulin signaling pathway
|
9
|
0.017016212
|
hsa04064:NF-kappa B signaling pathway
|
7
|
0.017281951
|
hsa 04062:Chemokine signaling pathway
|
10
|
0.033343318
|
hsa 05146:Amoebiasis
|
7
|
0.040529418
|
hsa 04621:NOD -like receptor signaling pathway
|
5
|
0.043877413
|
hsa 05020:Prion diseases
|
4
|
0.046418919
|
3.2 alpha1H treatment affects lncRNA expression in patients with bladder cancer
In addition, for the non-coding RNA molecule lncRNA, the protein detected in the GSE172112 dataset, we also performed a differential expression analysis. As with the analysis of protein-coding genes, for 12,382 lncRNA, we finally obtained 119 significantly differentially expressed lncRNA, including 53 significantly upregulated lncRNA and 66 significantly downregulated lncRNA (Table 2, Fig. 4).
Further, we obtained from the lnc2cancer database (http: / / bio-bigdata.hrbmu.edu.cn/lnc2cancer/index.The 369 bladder cancer-related lncRNA entries were downloaded from html) and two of these lncRNA were closely associated with bladder cancer and LINC00152 with TUG1 (Table 3).We found that both lncRNA were upregulated in bladder cancer patients in most literature reports, and in the GSE172112 dataset, we also compared the expression of patients using new drugs and placebo.
Table 3. Related information of LINC00152 and TUG1 recorded in lnc2Cancer database
lncRNA
|
Cancer.name
|
Methods
|
Expression.pattern
|
Pubmed.ID
|
LINC00152
|
gallbladder cancer
|
qPCR, RNAi, Western blot, RIP, Luciferase reporter assay etc.
|
up-regulated
|
28077595
|
LINC00152
|
gallbladder cancer
|
qPCR, RNAi, Western blot, ChIP , Luciferase reporter assay, Cell proliferation assay etc.
|
up-regulated
|
27829993
|
LINC00152
|
bladder cancer
|
qRT -PCR, ect .
|
down-regulated
|
29988223
|
TUG1
|
bladder cancer
|
qPCR, Luciferase reporter assay, Western blot
|
up-regulated
|
30925453
|
TUG1
|
bladder cancer
|
qPCR, Western blot, in vitro knockdown
|
up-regulated
|
31308746
|
TUG1
|
gallbladder cancer
|
qPCR, RNAi, Western blot, Luciferase reporter assay, Cell proliferation assay etc.
|
up-regulated
|
28178615
|
TUG1
|
bladder cancer
|
qPCR, RNAi, Western blot, RIP, Luciferase reporter assay etc.
|
up-regulated
|
26318860
|
TUG1
|
bladder cancer
|
qPCR etc.
|
up-regulated
|
29029461
|
TUG1
|
bladder cancer
|
qRT -PCR, Luciferase reporter assay, in vitro knockdown
|
up-regulated
|
29321088
|
TUG1
|
bladder cancer
|
qPCR, Western blot, MIT
|
up-regulated
|
28503069
|
TUG1
|
bladder cancer
|
qPCR, Western blot etc.
|
up-regulated
|
28376901
|
TUG1
|
bladder cancer
|
qPCR, RNAi, MTT assay etc.
|
up-regulated
|
27460088
|
For LINC00152, we found that the mean expression was 5.8159 in patients using new drugs, compared to 6.5003 in placebo patients and a Wilcoxon test p-value of 0.0038, showing that LINC00152 expression was significantly reduced by using new drugs.
But for TUG1, we found that the mean expression was 12.3502 in patients with new drugs, but 12.0904 in placebo, and a Wilcoxon test p-value of 0.001, showing that using new drugs could not reduce TUG1 expression.
Combined with the two lncRNA analysis described above, we suggest that the new drug alpha1H most likely caused a series of gene expression changes by reducing LINC00152 expression and ultimately played a good role in killing tumor cells.
3.3 alpha1H treatment altered part of the hallmark pathway activation in patients with bladder cancer
Using differential expression analysis between the two classes of patients in GSE172112, we found that many differentially expressed genes were generated in alpha1H-treated patients, and that treatment of alpha1H at a single gene level had a huge effect on patient gene expression. Further, we evaluated whether both types of patients have a huge impact on the activation of certain pathways from the perspective of the overall pathways.
Therefore, we collected and collated gene members of the 50 hallmark pathways from the MsigDB database, rated each sample in GSE172112 for pathway activation using the "ssgsea" algorithm in the R package "GSVA", and assessed score differences between samples using the Wilcoxon rank-sum test.final, We found that in alpha1H treatment compared with placebo, Activation of the following pathways in patients: where "REACTIVE_OXYGEN_SPECIES_PATHWAY", "P53_PATHWAY", "UV_RESPONSE_UV", "HYPOXIA", "APOPTOSIS", "TNFA_SIGNALING_VIA_NFKB", "COMPLEMENT", "IL6_JAK_STAT3_SIGNALING", "INFLAMMATORY_RESPONSE" is all downregulated (Table 4).
Table 4
Differences in ssGSEA scores of 50 hallmark pathway activation in patients in the GSE172112 dataset
Hallmark pathway
|
mean ssGSEA in alpha1H treated
|
mean ssGSEA in placebo treated
|
wilcox.test pValue
|
REACTIVE_OXYGEN_SPECIES_PATHWAY
|
1.569619674
|
1.582734513
|
0.004120032
|
P53_PATHWAY
|
1.507696636
|
1.521012228
|
0.007985337
|
UV_RESPONSE_UP
|
1.424304396
|
1.434798911
|
0.012893519
|
HYPOXIA
|
1.42652941
|
1.439254409
|
0.021327705
|
APOPTOSIS
|
1.531946053
|
1.544429318
|
0.02244453
|
TNFA_SIGNALING_VIA_NFKB
|
1.428124292
|
1.465793485
|
0.023021538
|
COMPLEMENT
|
1.374210535
|
1.397833555
|
0.033326204
|
IL6_JAK_STAT3_SIGNALING
|
1.33525268
|
1.368464526
|
0.036656563
|
INFLAMMATORY_RESPONSE
|
1.281653791
|
1.317457585
|
0.04728863
|
MTORC1_SIGNALING
|
1.607157413
|
1.616700174
|
0.050597275
|
HEME_METABOLISM
|
1.405351602
|
1.409403022
|
0.063037602
|
KRAS_SIGNALING_UP
|
1.341771721
|
1.361403026
|
0.067223154
|
IL2_STAT5_SIGNALING
|
1.40733034
|
1.423663491
|
0.076272281
|
PI3K_AKT_MTOR_SIGNALING
|
1.536935306
|
1.542465415
|
0.103223941
|
ALLOGRAFT_REJECTION
|
1.30101474
|
1.328268003
|
0.115903328
|
COAGULATION
|
1.247726476
|
1.260215792
|
0.122681092
|
CHOLESTEROL_HOMEOSTASIS
|
1.492168982
|
1.49907447
|
0.158421731
|
ANDROGEN_RESPONSE
|
1.610609182
|
1.615768853
|
0.172916472
|
UNFOLDED_PROTEIN_RESPONSE
|
1.587649281
|
1.593578296
|
0.178980994
|
NOTCH_SIGNALING
|
1.483836676
|
1.479281759
|
0.185200472
|
INTERFERON_GAMMA_RESPONSE
|
1.49277404
|
1.513236111
|
0.198110796
|
ESTROGEN_RESPONSE_LATE
|
1.413869042
|
1.418128337
|
0.201437702
|
PROTEIN_SECRETION
|
1.708610457
|
1.70515225
|
0.211659698
|
EPITHELIAL_MESENCHYMAL_TRANSITION
|
1.409575339
|
1.432730605
|
0.211659698
|
GLYCOLYSIS
|
1.437199485
|
1.443256898
|
0.222247114
|
APICAL_SURFACE
|
1.298814008
|
1.308220393
|
0.22951047
|
ESTROGEN_RESPONSE_EARLY
|
1.455088631
|
1.457645328
|
0.244534355
|
APICAL_JUNCTION
|
1.368342672
|
1.377967747
|
0.260227498
|
ANGIOGENESIS
|
1.364058543
|
1.37392876
|
0.293644881
|
ADIPOGENESIS
|
1.542517571
|
1.547454615
|
0.344052034
|
XENOBIOTIC_METABOLISM
|
1.318189564
|
1.322982168
|
0.404902132
|
MYOGENESIS
|
1.206300049
|
1.219144025
|
0.41562863
|
INTERFERON_ALPHA_RESPONSE
|
1.54320002
|
1.557834653
|
0.421053516
|
SPERMATOGENESIS
|
1.002492991
|
0.99673011
|
0.437572636
|
KRAS_SIGNALING_DN
|
0.920369019
|
0.915648732
|
0.531644659
|
PANCREAS_BETA_CELLS
|
0.889961293
|
0.886293012
|
0.550346464
|
E2F_TARGETS
|
1.457823735
|
1.450085836
|
0.569362912
|
HEDGEHOG_SIGNALING
|
1.276771876
|
1.278887994
|
0.60829904
|
G2M_CHECKPOINT
|
1.461654487
|
1.454182453
|
0.614900788
|
PEROXISOME
|
1.456106141
|
1.454807063
|
0.731490581
|
TGF_BETA_SIGNALING
|
1.616556225
|
1.614138978
|
0.752815594
|
UV_RESPONSE_DN
|
1.540834642
|
1.541525932
|
0.803279015
|
MITOTIC_SPINDLE
|
1.536838934
|
1.533410379
|
0.81785756
|
OXIDATIVE_PHOSPHORYLATION
|
1.64021922
|
1.642747538
|
0.832498667
|
MYC_TARGETS_V1
|
1.729433959
|
1.728475121
|
0.84719746
|
BILE_ACID_METABOLISM
|
1.267742379
|
1.267085055
|
0.884164318
|
WNT_BETA_CATENIN_SIGNALING
|
1.378224136
|
1.378829549
|
0.966262551
|
DNA_REPAIR
|
1.496350045
|
1.498487779
|
0.996250303
|
FATTY_ACID_METABOLISM
|
1.464107524
|
1.464935918
|
0.996250303
|
MYC_TARGETS_V2
|
1.488252701
|
1.486866857
|
1
|
Based on the above results, we found that the downregulation of "P53_PATHWAY" indicates the decreased proliferation capacity of tumor cells; while "TNFA_SIGNALING_VIA_NFKB", "IL6_JAK_STAT3_SIGNALING" and "INFLAMMATORY_RESPONSE" pathways indicate the alleviated immunoinflammatory response in patients after using the new drug alpha1H, indicating the removal of tumor cells (Fig. 5).
3.4 alpha1H and the TGF-beta signaling pathway
Combined with single gene-level differential analysis of overall pathway activation, we found that alpha1H affected some of gene members in TGF-beta signaling (including SMAD4, ACVR1C, ID4, SMAD9, SMAD6 and ACVR2B), but the activation of overall TGF-beta signaling was not significantly changed with Wilcoxon rank sum test p-value of 0.7528. However, SMAD4, SMAD9, SMAD6 Three SMAD family members also have important effects on TGF-beta signaling. Therefore, we focused on the expression of these three genes and found that all three SMAD family members were relatively highly expressed (Table 5) in patients treated with alpha1H.
Table 5
Expression of SMAD4, SMAD9, and SMAD6 in the GSE172112 dataset
gene
|
mean expression in alpha1H treated
|
mean expression in placebo treated
|
FC
|
wilcox.test pValue
|
SMAD9
|
8.654020345
|
7.850127802
|
1.102405026
|
0.003750499
|
SMAD6
|
7.889000669
|
7.319214986
|
1.077847923
|
0.006899118
|
SMAD4
|
11.30312093
|
11.10285469
|
1.018037366
|
0.006899118
|
To further validate the role of these three genes in patients with bladder cancer, we performed a prognostic analysis of these three genes separately, combined with the expression of BLCA patients in the TCGA program and the recorded clinical prognostic information. Among them, a total of 426 patients with detected expression and clinical prognostic information were analyzed.
For SMAD4, we divided patients into high expression groups (expression above 85.0865) and low expression group (expression below 85.0865), and the HR [95%CI] was 1.818[1.221–2.706] and the log-rank test p-value of 0.0029, indicating that the worse the prognosis of SMAD4 expression (Fig. 6).
For SMAD6, we divided patients into high expression groups (expression above 34.2596) and low expression groups (expression below 34.2596), and the HR of Cox proportional hazards [95%CI] of the final constructed proportional hazards model was 0.5052[0.3783–0.6747], and the log-rank test p-value < 0.0001, indicating lower SMAD6 expression (Fig. 7).
For SMAD9, we divided the patients into high expression groups (expression above 11.3209) and low expression group (expression below 11.3209) according to their expression value, and the HR [95%CI] of the final constructed Cox proportional-risk model was 1.521[1.064–2.174], and the log-rank test p-value was 0.021, indicating that the worse the prognosis of patients with higher SMAD9 expression (Fig. 8).
The above results show that the upregulation of SMAD6 gene expression after alpha1H treatment contributes to patient prognosis, but the upregulation of SMAD4 and SMAD9 genes is instead unfavorable to patient prognosis.
3.5 LINC00152 and patients with bladder cancer
In the analysis of differentially expressed lncRNA in the GSE172112 dataset, we found significantly downregulated LINC00152 expression in patients treated with alpha1H, and the results collected in the lnc2cancer database also showed generally high expression of this lncRNA in bladder cancer patients, we used this lncRNA to assess the prognostic value using the TCGA BLCA dataset, A total of 252 patients who examined lncRNA expression and recorded clinical prognostic information were analyzed.
We found that, given as the expression values of the LINC00152, Patients were divided into high expression groups (expression above 1.130234) and low expression groups (expression below 1.130234), The HR [95%CI] of the final constructed model of the Cox scale analysis was 3.681[1.354–10.01], The log-rank test p-value was 0.0062, It shows that the high expression of LINC00152 is unfavorable to the patient prognosis, The new drug alpha1H reduces LINC00152 and has a positive effect on curing cancer (Fig. 9).