LAMC1 expression in human pan-cancers
GTEx datasets showed various levels of LAMC1 gene expression in humans and a low expression level in brain tissues (Fig. 1A). However, a relatively high expression level of LAMC1 was found in CNS tumors (Fig. 1B). In TCGA data, LAMC1 was highly expressed in CHOL, ESCA, GBM, HNSC, KIRC, KIRP, LIHC, LUAD, LUSC, STAD, and THCA. Conversely, LAMC1 expression was downregulated in tumors in BLCA, BRCA, and KICH relative to normal tissue (Fig. 1C). After adding normal tissues from the GTEx dataset as controls, we found enhanced expression of LAMC1 in DLBC, LGG, and THYM (Fig. 1D).
Multifaceted prognostic features of LAMC1 across pan-cancers
Cox proportional risk model analysis showed that LAMC1 expression levels were significantly correlated to OS of patients with KIRP, LGG, MESO, and UVM (P < 0.001) and many other cancer types (P < 0.05) (Fig. 2A). LAMC1 expression was associated with DSS of various cancers, including BLCA, LGG, MESO, and UVM (P < 0.001) (Fig. 2B). High LAMC1 expression was correlated to DFI in patients with BLCA, CESC, OV, and PAAD (P < 0.05) (Fig. 2C). In terms of the association between high LAMC1 expression levels and PFI, a forest map showed a poor prognosis of patients with ACC, BLCA, LGG, and UVM (P < 0.001) (Fig. 2D). Kaplan–Meier survival analysis also showed that LGG patients with high LAMC1 levels had poor OS, DSS, and PFI (P < 0.001, Fig. 3).
Clinicopathological analysis of LAMC1 expression in Chinese glioma patients
Bioinformatics analysis of LAMC1 expression in gliomas was performed using the CGGA database[17]. ROC curve analysis showed that LAMC1 expression was a predictor of 1-year (AUC = 0.727), 3-year (AUC = 0.781), and 5-year (AUC = 0.797) survival with a good predictive value (Fig. 4A). Kaplan–Meier survival analysis showed that high LAMC1 expression was associated with a poor prognosis of patients with glioma (Fig. 4B). Univariate Cox analysis showed that LAMC1 expression, primary-recurrent-secondary (PRS) type, histology, grade, age, and chemotherapy were high-risk factors (HR > 1), and IDH mutation and 1p19q co-deletion were low-risk factors (HR < 1) (Fig. 4C). Multivariate Cox analysis showed that LAMC1 expression (P < 0.001, HR = 1.163), PRS type (P < 0.001, HR = 1.918), grade (P < 0.001, HR = 2.619), and 1p19q co-deletion (P < 0.001, HR = 0.402) may also be independent prognostic factors (Fig. 4D).
Correlation between LAMC1 expression and clinical phenotypes of various cancers
Next, we examined differential expression of LAMC1 in patients with each tumor type in accordance with age and the tumor stage in TCGA database. We found that patients older than 65 years of age with BRCA (P = 0.00032), COAD (P = 0.007), KIRC (P = 0.011), LIHC (P = 0.0078), or PAAD (P = 0.0018) had low LAMC1 expression levels. LAMC1 expression was higher in THYM (P = 0.012) and UCEC (P = 0.041) patients over 65 years of age (Additional file 1: Fig. S2A). We also found statistical significance between LAMC1 expression and the partial tumor stage in eight cancer types, including BLCA, COAD, HNSC, KICH, KIRC, LUSC, PAAD, and UVM. Notably, LAMC1 expression increased with the stage in most tumors, suggesting that high LAMC1 expression is associated with tumor progression (Additional file 1: Fig. S2B). Clinicopathological characteristic correlation analysis of the CGGA data demonstrated that the expression level of LAMC1 was remarkably related to age (Fig. 5A), grade (Fig. 5B), PRS type (Fig. 5C), chemotherapy (Fig. 5D), and histology (Fig. 5G) in glioma samples. Recently, it has been widely recognized that IDH mutations and 1p/19q co-deletion suggest a favorable prognosis of gliomas[18]. In this study, high expression levels of LAMC1 were found in IDH-wildtype and 1p19q non-coding glioma compared with IDH-mutants or 1p19q co-deletions (Fig. 5E-F). These findings suggest that LAMC1 participates in the clinical development of glioma and may be a novel biomarker.
Correlation of LAMC1 protein expression to clinicopathological parameters and prognosis in the glioma cohort
We explored the correlation between LAMC1 protein expression and clinicopathological parameters based on the information of tissue microarray cases. The calculated total immunoreactive scores of LAMC1 expression ranged from 2 to 24 in all samples. We divided the cases into subgroups of high (IHC score > 10) and low (IHC score ≤ 10) LAMC1 expression based on the cutoff point determined by X-tile software related to survival time and status. LAMC1 expression was correlated to the degree of pathological grade and postoperative recurrence of patients (P < 0.05). Additionally, no association was found between LAMC1 expression and gender (P = 0.798) or age (P = 0.314) (Additional file 1: Table S1). Representative images of the IHC staining intensity of LAMC1 in gliomas with different pathological grades and normal brain tissue are shown in Fig. 6A. LAMC1 was nearly negative in normal brain tissue (Fig. 6A-B). As shown in Fig. 6C, glioma patients with higher pathological grades (G2–4) had higher total immunoreactive scores for LAMC1 than those with a lower grade (G1) (P < 0.05). Survival analysis by the Kaplan–Meier method with the log-rank test indicated that patients with a high level of LAMC1 had worse outcomes and shorter OS (P = 0.0482) and DFS (P = 0.0033) than those with low LAMC1 expression (Fig. 6D). Age and pathological grade were independent prognostic indicators in both univariate and multivariate Cox analysis models (Additional file 1: Table S2).
LAMC1 knockdown inhibits glioma cell proliferation, migration, and invasion
Hs683 cells with high LAMC1 expression were chosen for subsequent experiments (Additional file 1: Fig. S1). Fluorescence microscopy showed that lentiviral particles with the third targeted site had a higher infection efficiency in glioma cells (Fig. 7A). Western blotting verified its efficacy (Fig. 7B). CCK-8 assays showed that cell proliferation in the LAMC1 knockdown group was significantly slower than that in the NC group (P < 0.01, Fig. 7C), together with a reduced ability for colony formation (Fig. 7D). Furthermore, LAMC1 knockdown decreased the number of cells that migrated through transwell chambers (P < 0.001, Fig. 7E-F). Invasion assay results were consistent with the migration assay results (Fig. 7G). These data collectively indicate that LAMC1 promotes glioma cell proliferation, migration, and invasion.
HIF-1α expression correlates positively to LAMC1 expression in glioma
HIF-1α expression was significantly upregulated in various tumors, including GBM, GBMLGG, LGG, CESC, ESCA, STES, COAD, STAD, HNSC, LUSC, THCA, PAAD, TGCT, ALL, LAML, and CHOL. Significant downregulation of HIF-1α expression was observed in KIRP, KIPAN, KIRC, SKCM, ACC, and KICH (Fig. 8A). Correlation analysis showed that HIF-1α expression in normal brain tissue and glioma was positively correlated to LAMC1 (expression R = 0.58, P < 0.001; Fig. 8B).
Effects of hypoxia on LAMC1 expression in glioma
LAMC1 and HIF-1α expression varied with the change in hypoxic exposure time (Fig. 9A-B, Additional file 1: Fig. S3A). After hypoxic treatment for 12 h, LAMC1 and HIF-1α protein expression was the highest. Notably, administration of HIF-1α inhibitor YC-1 suppressed hypoxia-stimulated LAMC1 and HIF-1α expression (Fig. 9C-D, Additional file 1: Fig. S3B). Bioinformatics predicted four binding sites for HIF-1α protein on the human LAMC1 gene promoter (Fig. 9E). The results of promoter luciferase assay showed that HIF-1α directly regulated activity of the LAMC1 promoter (Fig. 9F). It was noteworthy that we observed strong luciferase activity in cells co-transfected with LAMC1 Pro-luc and HIF-1α-NC plasmids, suggesting other regulators of LAMC1 activation in addition to HIF-1α (Fig. 9F).