2.1 MTFR2 was upregulated and correlated with sex, age, smoking history, cancer stage, histological subtype and TP53 mutation status in LUAD patients
MTFR2 was abnormally expressed in pan-cancerous tissues, and was mainly over-expressed in a variety of tumor tissues (Fig. 1). In detail, in the Oncomine database, MTFR2 was highly expressed in breast cancer, colorectal cancer, gastric cancer, lung cancer and other tumor tissues, while lowly expressed in brain nerve tumor, breast cancer, leukemia and other tumor tissues (Fig. 1A). In Timer database, MTFR2 was highly expressed in BLCA, BRCA, CHOL, LUSC, LUAD and other tumor tissues (Fig. 1B).
In addition, we found similar results in the Ualcan database, where MTFR2 was highly expressed in LUAD tissues (Table 1). In addition, in the TCGA database, the expression of MTFR2 was increased in unpaired LUAD patients (Fig. 2A) and in 57 pairs of LUAD patients (Fig. 2B).
Table 1
The expression of MTFR2 in Ualcan database was correlated to the clinicopathological characteristics in LUAD patients.
Clinicopathological characteristics
|
p value
|
Expression
|
|
Normal-vs-Primary
|
1.62447832963153E-12
|
Gender
|
|
Male-vs-Female
|
4.185800E-02
|
Age
|
|
Age(41-60Yrs)-vs-Age(61-80Yrs)
|
1.089890E-02
|
Smoking
|
|
Non smoker-vs-Smoker
|
7.3849000004067E-07
|
Non smoker-vs-Reformed smoker2
|
6.560800E-03
|
Smoker-vs-Reformed smoker1
|
1.53600000496468E-08
|
Smoker-vs-Reformed smoker2
|
1.978210E-03
|
Reformed smoker1-vs-Reformed smoker2
|
1.955780E-04
|
Cancer stages
|
|
Stage1-vs-Stage2
|
4.427500E-02
|
Tissue types
|
|
NOS-vs-Mixed
|
2.900200E-02
|
NOS-vs-LBC-Non Mucinous
|
2.038200E-03
|
NOS-vs-Solid Pattern Predominant
|
2.599600E-02
|
NOS-vs-Acinar
|
5.378900E-03
|
NOS-vs-LBC-Mucinous
|
1.70860000003881E-06
|
NOS-vs-Papillary
|
1.027880E-04
|
Mixed-vs-Solid Pattern Predominant
|
1.196000E-02
|
Mixed-vs-LBC-Mucinous
|
4.00150000001709E-06
|
Mixed-vs-Papillary
|
4.149500E-02
|
Clear Cell-vs-LBC-Non Mucinous
|
4.028200E-02
|
Clear Cell-vs-Papillary
|
2.063500E-02
|
LBC-Non Mucinous-vs-Solid Pattern Predominant
|
4.452100E-02
|
LBC-Non Mucinous-vs-LBC-Mucinous
|
1.174210E-02
|
Solid Pattern Predominant-vs-LBC-Mucinous
|
1.926450E-02
|
Solid Pattern Predominant-vs-Mucinous carcinoma
|
4.480600E-03
|
Solid Pattern Predominant-vs-Papillary
|
3.801800E-02
|
Acinar-vs-LBC-Mucinous
|
1.693590E-02
|
LBC-Mucinous-vs-Papillary
|
1.804030E-02
|
TP53 Mutant
|
|
TP53-Mutant-vs-TP53-Non Mutant
|
1.81166193158333E-12
|
In the Ualcan database, we found that the expression of MTFR2 was related to the clinicopathological features of LUAD patients (Table 1). In detail, the level of MTFR2 was related to the sex (Male vs Female), Age (Age(41-60Yrs) vs Age (61-80Yrs)), Smoking (Non smoker vs Smoker; Non smoker vs Reformed smoker2; Smoker vs Reformed smoker1; Smoker vs Reformed smoker2; Reformed smoker1 vs Reformed smoker2), histological subtype (NOS vs Mixed; NOS vs LBC-Non Mucinous; NOS vs Solid Pattern Predominant; NOS vs Acinar; NOS vs LBC-Mucinous; NOS vs Papillary; Mixed vs Solid Pattern Predominant; Mixed vs LBC-Mucinous; Mixed vs Papillary; Clear Cell vs LBC-Non Mucinous; Clear Cell vs Papillary; LBC-Non Mucinous vs Solid Pattern Predominant; LBC-Non Mucinous vs LBC-Mucinous; Solid Pattern Predominant vs LBC-Mucinous; Solid Pattern Predominant vs Mucinous carcinoma; Solid Pattern Predominant vs Papillary; Acinar vs LBC-Mucinous; LBC-Mucinous vs Papillary) and TP53 mutation (TP53-Mutant vs TP53-Non Mutant) of LUAD patients.
In TCGA database, ROC analysis showed that the level of MTFR2 had a diagnostic value in LUAD patients (Figs. 2C and D). Further Kaplan-Meier analysis showed that the LUAD patients with increased MTFR2 had a poor prognosis (Fig. 2E). In addition, in the Kaplan-Meier Plotter database, the expression levels of MTFR2 (228069_at) and MTFR2 (234944_at) were associated with the overall survival and early disease progression of LUAD patients (Figs. 2F-I). Univariate Cox analysis of complete clinical data of LUAD patients in TCGA database showed that cancer stage, T stage, distant metastasis as well as MTFR2 expression were risk factors for the prognosis of LUAD patients (Table 2). Further multivariate Cox analysis showed that cancer stage was an independent risk factor for the prognosis of LUAD patients (Table 2).
Table 2
COX regression analysis of the correlation between clinicopathological features with prognosis.
Clinical characteristics
|
HR
|
HR 95% CI low
|
HR 95% CI up
|
p value
|
a
|
|
|
|
|
Age
|
1.057406203
|
0.713393673
|
1.567308372
|
0.781014797
|
Gender
|
1.00097432
|
0.698837323
|
1.433737948
|
0.995761563
|
Stage
|
1.64465101
|
1.396688
|
1.93663649
|
2.42E-09
|
T
|
1.623091548
|
1.309819761
|
2.011289072
|
9.57E-06
|
M
|
1.681168333
|
0.923680619
|
3.059853055
|
0.08910352
|
N
|
1.792676516
|
1.464854278
|
2.193862653
|
1.47E-08
|
MTFR2
|
1.096892408
|
1.002541345
|
1.200123029
|
0.043876089
|
b
|
|
|
|
|
Stage
|
1.37421707
|
1.094758223
|
1.725013356
|
0.006135636
|
T
|
1.223211397
|
0.971744105
|
1.539753227
|
0.086187124
|
N
|
1.283860776
|
0.991173421
|
1.662976889
|
0.05838382
|
MTFR2
|
1.062548364
|
0.964968929
|
1.169995211
|
0.217045634
|
2.2 Screening of MTFR2 co-expressed genes
We screened 1127 MTFR2 co-expressed genes in TCGA database by screening conditions, most of which were positive related genes. Among them, there were 840 positive and 208 negative related genes (S1 Table). The top 10 genes that were positively and negatively related to MTFR2 (Table 3). The expression of the top 10 positive and negative related genes of MTFR2 in LUAD tissues from TCGA were abnormally expressed (Fig. 3).
Table 3
The top 10 genes with positive and negative correlation to MTFR2.
Gene
|
Corgene
|
Cor
|
p value
|
MTFR2
|
CENPW
|
0.876
|
1.08E-170
|
MTFR2
|
TTK
|
0.849
|
7.63E-150
|
MTFR2
|
NCAPH
|
0.848
|
4.00E-149
|
MTFR2
|
RAD51
|
0.835
|
1.43E-140
|
MTFR2
|
CENPA
|
0.827
|
2.59E-135
|
MTFR2
|
KIF2C
|
0.817
|
2.83E-129
|
MTFR2
|
BUB1
|
0.817
|
1.98E-129
|
MTFR2
|
CDCA8
|
0.816
|
5.28E-129
|
MTFR2
|
HJURP
|
0.814
|
7.81E-128
|
MTFR2
|
CDCA5
|
0.814
|
1.26E-127
|
MTFR2
|
C16orf89
|
-0.655
|
7.35E-67
|
MTFR2
|
CRY2
|
-0.628
|
3.92E-60
|
MTFR2
|
C1orf116
|
-0.585
|
1.60E-50
|
MTFR2
|
CYP4B1
|
-0.583
|
5.72E-50
|
MTFR2
|
SELENBP1
|
-0.58
|
2.16E-49
|
MTFR2
|
CACNA2D2
|
-0.576
|
1.56E-48
|
MTFR2
|
ADGRF5
|
-0.568
|
4.68E-47
|
MTFR2
|
NAPSA
|
-0.561
|
1.09E-45
|
MTFR2
|
SNED1
|
-0.558
|
4.89E-45
|
MTFR2
|
NR3C2
|
-0.557
|
6.44E-45
|
Note: Corgene, co-expressed gene; Cor, correlation coefficient. |
2.3 Analysis of MTFR2 and its coexpressed genes by GO, KEGG and GSEA
In order to further understand the potential function of MTFR2 in the development of LUAD, we carried out GO and KEGG analysis of MTFR2 co-expressed genes. MTFR2 coexpressed genes were involved in biological processes (BP), cellular component (CC) and molecular function (MF) (Figs. 4A-C), and in details, were mainly involved in DNA replication, chromosome segregation, mitotic nuclear division, regulation of mitotic cell cycle phase transition and other processes (S2 Table ). KEGG pathway analysis indicated MTFR2 co-expressed genes were mainly involved in cell cycle, DNA replication, proteasome, homologous recombination, spliceosome, nucleotide excision repair, human T-cell leukemia virus 1 infection, p53 signaling pathway and other signal pathways (Fig. 4D and Table 4). Secondly, GSEA found that cell cycle, DNA_replication, homologous recombination, P53 signal pathway, oocyte meiosis and base_excision repair signaling pathways were highly enriched in MTFR2 highly expressed group (Fig. 5 and Table 5). To sum up, these results suggested that MTFR2 might regulate the progress of LUAD through cell cycle, DNA replication, Homologous recombination and P53 signaling pathway.
Table 4
MTFR2 co-expressed genes by KEGG.
ID
|
Description
|
Count
|
p value
|
p adjust
|
hsa04110
|
Cell cycle
|
49
|
2.54E-26
|
7.53E-24
|
hsa03030
|
DNA replication
|
26
|
1.92E-23
|
2.84E-21
|
hsa03050
|
Proteasome
|
19
|
2.45E-11
|
2.42E-09
|
hsa03440
|
Homologous recombination
|
17
|
2.60E-10
|
1.59E-08
|
hsa03430
|
Mismatch repair
|
13
|
2.69E-10
|
1.59E-08
|
hsa03013
|
RNA transport
|
35
|
5.88E-09
|
2.90E-07
|
hsa03460
|
Fanconi anemia pathway
|
17
|
3.58E-08
|
1.52E-06
|
hsa03040
|
Spliceosome
|
28
|
4.24E-07
|
1.57E-05
|
hsa03410
|
Base excision repair
|
12
|
6.52E-07
|
2.14E-05
|
hsa03420
|
Nucleotide excision repair
|
14
|
1.25E-06
|
3.69E-05
|
hsa04114
|
Oocyte meiosis
|
24
|
2.97E-06
|
8.00E-05
|
hsa00670
|
One carbon pool by folate
|
8
|
2.25E-05
|
0.000555487
|
hsa03008
|
Ribosome biogenesis in eukaryotes
|
20
|
3.63E-05
|
0.000826201
|
hsa04218
|
Cellular senescence
|
24
|
0.000136465
|
0.002885268
|
hsa04914
|
Progesterone-mediated oocyte maturation
|
17
|
0.000252864
|
0.004989844
|
hsa00240
|
Pyrimidine metabolism
|
12
|
0.000297856
|
0.00551034
|
hsa05012
|
Parkinson disease
|
31
|
0.000505075
|
0.008794241
|
hsa05166
|
Human T-cell leukemia virus 1 infection
|
28
|
0.000610833
|
0.010044803
|
hsa00270
|
Cysteine and methionine metabolism
|
10
|
0.001213762
|
0.018909138
|
hsa04115
|
p53 signaling pathway
|
12
|
0.002598093
|
0.038451783
|
Table 5
GSEA displayed the main signal pathways derived from high expression of MTFR2 enrichment.
Name
|
Size
|
Nes
|
Nom p value
|
KEGG_CELL_CYCLE
|
124
|
2.5325148
|
0
|
KEGG_SPLICEOSOME
|
126
|
2.342574
|
0
|
KEGG_OOCYTE_MEIOSIS
|
112
|
2.3013418
|
0
|
KEGG_MISMATCH_REPAIR
|
23
|
2.2582684
|
0
|
KEGG_DNA_REPLICATION
|
36
|
2.210251
|
0
|
KEGG_HOMOLOGOUS_RECOMBINATION
|
28
|
2.207216
|
0
|
KEGG_RNA_DEGRADATION
|
57
|
2.2033587
|
0
|
KEGG_PROGESTERONE_MEDIATED_
OOCYTE_MATURATION
|
85
|
2.1570013
|
0
|
KEGG_PROTEASOME
|
44
|
2.108183
|
0
|
KEGG_BASE_EXCISION_REPAIR
|
33
|
2.097559
|
0
|
KEGG_PATHOGENIC_ESCHERICHIA_COLI_INFECTION
|
55
|
2.096588
|
0
|
KEGG_GLYOXYLATE_AND_
DICARBOXYLATE_METABOLISM
|
16
|
2.0935147
|
0
|
KEGG_BASAL_TRANSCRIPTION_FACTORS
|
35
|
2.051244
|
0
|
KEGG_PYRIMIDINE_METABOLISM
|
98
|
2.0281448
|
0
|
KEGG_ONE_CARBON_POOL_BY_FOLATE
|
17
|
2.0059762
|
0
|
KEGG_P53_SIGNALING_PATHWAY
|
68
|
1.96642
|
0
|
KEGG_PURINE_METABOLISM
|
157
|
1.9532791
|
0
|
KEGG_NUCLEOTIDE_EXCISION_REPAIR
|
44
|
2.0500638
|
0.001831502
|
KEGG_UBIQUITIN_MEDIATED_PROTEOLYSIS
|
133
|
1.7833992
|
0.003868472
|
KEGG_CYSTEINE_AND_METHIONINE_METABOLISM
|
34
|
1.8469852
|
0.003883495
|
KEGG_FRUCTOSE_AND_MANNOSE_METABOLISM
|
33
|
1.821762
|
0.011516315
|
KEGG_PENTOSE_PHOSPHATE_PATHWAY
|
27
|
1.7594053
|
0.011650485
|
KEGG_N_GLYCAN_BIOSYNTHESIS
|
46
|
1.6938915
|
0.023346303
|
KEGG_BLADDER_CANCER
|
42
|
1.5152026
|
0.0251938
|
KEGG_RIBOFLAVIN_METABOLISM
|
16
|
1.5648233
|
0.027675277
|
KEGG_PANCREATIC_CANCER
|
70
|
1.4912151
|
0.03131524
|
KEGG_GLYCOLYSIS_GLUCONEOGENESIS
|
62
|
1.5601822
|
0.036398467
|
KEGG_HUNTINGTONS_DISEASE
|
177
|
1.6180334
|
0.046511628
|
KEGG_RNA_POLYMERASE
|
29
|
1.5426548
|
0.048732944
|
2.4 Construction of PPI network and the expressions of Hub genes and its prognostic value
By identifying the function of MTFR2 co-expressed genes, the potential biological function of MTFR2 was inferred. The top 10 Hub genes in the PPI network were selected by the CytoHubba plug-in as CDK1, CDC20, CCNB1, PLK1, CCNA2, AURKB, CCNB2, BUB1B, MAD2L1 and BUB1 (S Fig and Table 6). The expression of MTFR2 was correlated with that of CDK1, CDC20, CCNB1, PLK1, CCNA2, AURKB, CCNB2, BUB1B, MAD2L1 and BUB1 (Fig. 6). In addition, GEPIA online website was used to analyze the expressions of Hub genes and its prognostic value. We found that the expressions of CDK1, CDC20, CCNB1, PLK1, CCNA2, AURKB, CCNB2, BUB1B, MAD2L1 and BUB1 genes were significantly increased in LUAD tissues (Fig. 7). The top10 Hub genes were associated with overall survival in LUAD patients, and CDK1, CCNB1, PLK1, AURKB, CCNB2, BUB1B and BUB1 were associated with disease free survival (DFS) in LUAD patients with disease-free progression (Fig. 8). These results further indicated that MTFR2 played an vital role in the diagnosis, treatment and disease progression of LUAD.
Table 6
The top 10 Hub genes in PPI Network.
Name
|
Description
|
Score
|
CDK1
|
Cyclin dependent kinase 1
|
192
|
CDC20
|
Cell division cycle 20
|
160
|
CCNB1
|
Cyclin B1
|
138
|
PLK1
|
Polo like kinase 1
|
134
|
CCNA2
|
Cyclin A2
|
129
|
AURKB
|
Aurora kinase B
|
127
|
CCNB2
|
Cyclin B2
|
124
|
BUB1B
|
BUB1 mitotic checkpoint serine/threonine kinase B
|
118
|
MAD2L1
|
Mitotic arrest deficient 2 like 1
|
117
|
BUB1
|
BUB1 mitotic checkpoint serine/threonine kinase
|
105
|