High-Mobility Group AT-Hook 1 Served as a Prognosis Biomarker and Associated with Immune Infiltrate in Hepatocellular Carcinoma

Figures & data

Figure 1 HMGA1 mRNA expression levels in normal and HCC tissues, as obtained from the (A) TCGA, (B) ICGC, (C) GEO, and (D) Guangxi cohort. Representative immunohistochemical staining images of HMGA1 protein in HCC specimens (E and F) and in normal liver tissues (G and H), which taken from the open HPA dataset.

Abbreviations: HMGA1, high-mobility group AT-hook 1; HCC, hepatocellular carcinoma; TCGA, The Cancer Genome Atlas; ICGC, International Cancer Genomics Consortium; GEO, Gene Expression Omnibus; HPA, The Human Protein Atlas.

Table 1 Association Between Clinicopathological Features of Patients and HMGA1 Expression in the Cancer Genome Atlas

Clinicopathological Parameters	Cases(n)	HMGA1 Expression Levels	T	P-value
Age
<60 years	125	43.01	1.342	0.181
≥60 years	110	32.00
Gender
Male	161	39.48	−0.234	0.815
Female	74	38.22
Stage
I+II	163	35.65	−2.081	0.039*
III+IV	72	46.85
Grade
G1+G2	132	32.57	−2.858	0.005*
G3+G4	103	47.43
T stage
T1+T2	167	35.68	−1.800	0.075
T3+T4	68	47.44
N stage
N0	231	38.58	0.252	0.801
N1	4	19.81
M stage
M0	231	38.61	0.195	0.845
M1	4	13.68

Note: *Statistically significant.

Table 2 Association Between HMGA1 Expression and Clinicopathological Features (Logistic Regression)

Clinical Characteristic	Total(N)	(OR) in HMGA1 Expression	P-value
Age	370	0.822(0.545–1.237)	0.348
Stage (II vs I)	258	2.103(1.247–3.581)	0.006*
Stage (III vs I)	256	2.387(1.407–4.095)	0.001*
Stage (IV vs I)	176	2.272(0.367–17.587)	0.376
Grade (II vs I)	234	0.414(0.213–0.776)	0.007*
Grade (III vs I)	206	0.359(0.179–0.698)	0.003*
Grade (IV vs I)	67	0.320(0.084–1.140)	0.081
Tumor status (II vs I)	275	2.496(1.503–4.192)	<0.001*
Tumor status (III vs I)	261	2.323(1.363–4.006)	0.002*
Tumor status (IV vs I)	194	1.808(0.578–5.824)	0.304
Lymph node (II vs I)	256	1.000(0.118–8.439)	1.000
Distant metastasis (II vs I)	300	3.045(0.384–61.996)	0.338

Note: *Statistically significant.

Abbreviation: OR, odds ratio.

Table 3 Univariate Cox Proportional Hazards Regression and Multivariate Cox Regression Analyses of HMGA1 and Clinical Features in HCC

Variable	Univariate Analysis			Multivariate Analysis
	HR	95% CI	P-value	HR	95% CI	P-value
Age	1.005	0.987–1.023	0.591	1.012	0.993–1.032	0.218
Gender	0.780	0.487–1.249	0.301	0.927	0.556–1.545	0.770
Grade	1.017	0.746–1.387	0.914	0.913	0.652–1.279	0.596
Stage	1.865	1.456–2.388	<0.001*	0.941	0.353–2.508	0.903
T stage	1.804	1.434–2.270	<0.001*	1.816	0.756–4.360	0.182
N stage	2.022	0.494–8.276	0.328	1.872	0.275–12.743	0.521
M stage	3.850	1.207–12.281	0.023 *	1.256	0.331–4.758	0.738
HMGA1	1.571	1.285–1.922	<0.001*	1.558	1.257–1.922	<0.001*

Note: *P<0.05 was considered statistically significant.

Abbreviations: HR, hazard ratio; CI, confidence interval.

Figure 2 Kaplan-Meier analysis of HMGA1 and OS time of HCC in the (A) TCGA, (B) ICGC, (C) GEO, and (D) Guangxi cohorts.

Abbreviations: HMGA1, high-mobility group AT-hook 1; HCC, hepatocellular carcinoma; OS, overall survival; TCGA, The Cancer Genome Atlas; ICGC, International Cancer Genomics Consortium; GEO, Gene Expression Omnibus.

Figure 3 Boxplot showing expression of HMGA1 correlate with different histological grades (A), T stages (B), pathologic stages (C), and p53 gene mutation status (D) in HCC. The t-test was used to estimate the significance of the difference in gene expression levels between groups. The application value of HMGA1 in HCC clinical parameters. (E) Forest plot illustrated the results of multivariate Cox regression analyses. ***Stands for P < 0.001. (F) ROC curves of pathologic stage, T stage, M stage, and HMGA1 expression.

Abbreviations: HMGA1, high-mobility group AT-hook 1; HCC, hepatocellular carcinoma; ROC, Receiver operating characteristic; T stage, Tumor stage; M stage, Metastasis stage; AUC, Area under Curve.

Figure 4 Significantly enriched GO annotations (A) and KEGG pathways (B) of HMGA1 in HCC based on TCGA database. (C) PPI analysis of HMGA1 by STRING. (D) HMGA1 co-expression network was inferred from Coexpedia and deep hue indicated a higher degree of association with HMGA1.

Abbreviations: GO, Gene Ontology; KEGG, Kyoto Encyclopedia of Genes and Genomes; PPI, Protein-protein interaction; HMGA1, high-mobility group AT-hook 1; TCGA, The Cancer Genome Atlas; STRING, search tool for recurring instances of neighboring genes.

Figure 5 HMGA1 mRNA expression was respectively significant positive correlation to the mRNA expressions of TP53 (R=0.4, P<0.001), RB1 (R=0.17, P<0.001), CCNB1 (R=0.54, P<0.001), PAK1 (R=0.42, P<0.001), CDK1 (R=0.37, P<0.001), CDK2 (R=0.34, P<0.001), CDK4 (R=0.55, P<0.001), E2F1 (R=0.22, P<0.001), E2F2 (R=0.39, P<0.001), HNRNPUL1 (R=0.45, P<0.001), while was significant negative correlation to GADD45A (R=−0.23, P<0.001) (A–L).

Abbreviations: HMGA1, high-mobility group AT-hook 1; TP53, tumor protein p53; RB1, RB Transcriptional Corepressor 1; CCNB1, Cyclin B1; PAK1, P21 (RAC1) Activated Kinase 1; CDK1, Cyclin Dependent Kinase 1; CDK2, Cyclin Dependent Kinase 2; CDK4, Cyclin Dependent Kinase 4; E2F1, E2F Transcription Factor 1; HNRNPUL1, Heterogeneous Nuclear Ribonucleoprotein U Like 1; GADD45A, Growth Arrest and DNA Damage Inducible Alpha.

Figure 6 The relationship between HMGA1 and immune invasion in hepatocellular carcinoma. (A) Relationship between the abundance of immune cells and HMGA1 expression in HCC (TIMER). (B) Different proportions of sorts of immune cells in high- and low-HMGA1 groups in HCC specimens (CIBERSORT).

Abbreviations: HMGA1, high-mobility group AT-hook 1; HCC, hepatocellular carcinoma; TIMER, Tumor Immune Estimation Resource; CIBERSORT, Cell type Identification through the Estimating Relative Subsets of RNA Transcripts.

Figure 7 Resulting heat map visualizing the correlation matrix of diverse tumor-infiltrating immune cells subpopulations. Each tile indicates coefficients calculated by Pearson’s correlation test.