Gene expression analysis of COL1A1, PRPF40A, and UCP2 in NSCLC
Our comparative qPCR analysis showed abundant COL1A1 overexpression in tumours (median RQ = 975.4, IQR 389.9–3172.7, N = 132) when compared to adjacent normal tissues (69.4, 29.8–235.6, N = 119, Mann–Whitney test, p < 1 × 10−4) (Fig. 1a). Similarly, PRPF40A was overexpressed in NSCLC samples (RQ = 3.49, 1.13–11.84, N = 135 vs 1.53, 0.98–3.16, N = 119, Mann–Whitney test, p < 1 × 10−4) (Fig. 1b). In the case of UCP2, there was only a trend of higher expression observed in lung tumours (RQ = 0.043, 0.025–0.08, N = 136 vs 0.034, 0.026–0.055, N = 122, Mann–Whitney test, p = 0.066) (Fig. 1c). UCP2 mRNA expression was higher in lung adenocarcinomas (RQ = 0.053, 0.03–0.087, N = 61) than in SqCLCs (RQ = 0.035, 0.019–0.072, N = 75, Mann–Whitney test, p = 0.015). Apart from that no other associations were observed between COL1A1, PRPF40A or UCP2 mRNA expression and patients’ gender, age, survival, smoking history, TNM classification, and tumour histology and differentiation.
Promoter methylation analysis of COL1A1, PRPF40A, and UCP2 in NSCLC
For the promoter methylation analysis, we set the hypermethylation threshold according to the previously described method (Normal reference range = mean MtI + 2 × standard deviation of the normal samples) (Oleksiewicz et al. 2011) at MtI = 19.9% for UCP2 and MtI = 13.0% for COL1A1. PRPF40A promoter was unmethylated in all samples tested. UCP2 promoter hypermethylation was observed only in 5/91 NSCLC samples and none (0/26) of the normal tissues. We noted elevated COL1A1 promoter methylation in 33/91 NSCLCs (Fig. 1D) and in 2/26 normal samples. Although the difference in methylation between grouped normal and tumour samples was insignificant, pairwise comparison showed significant hypermethylation of COL1A1 promoter in NSCLC (p = 0.035, Wilcoxon test, N = 26 pairs). The samples selected for the latter analysis featured clinicopathological profile representative of all tumour samples utilised in this study (Table 1), so that the pairwise comparison was not confounded by the clinical factors. Furthermore, higher methylation was observed in SqCLCs (median MtI = 8.72%, IQR 1.95–26.18, N = 56) than in adenocarcinomas (2.28, 1.36–5.5, N = 35, Mann–Whitney test, p = 0.024, Fig. 1e), as well as in the moderately and well-differentiated tumours (MtI = 5.19%, IQR 1.99–28.21, N = 65) in comparison with poorly differentiated NSCLCs (2.18, 1.39–5.45, N = 26, Mann–Whitney test, p = 0.01, Fig. 1f). Apart from that observation COL1A1 methylation did not correlate with patients’ gender, age, survival, smoking history, and TNM classification. Moreover, methylation was not associated with the mRNA expression values, even when potentially confounding factors (histology and differentiation) were taken into account.
COL1A1, PRPF40A, and UCP2 expression correlates with hypoxia markers
Next, we evaluated interrelationships between COL1A1, PRPF40A, UCP2, and hypoxia markers (CYGB, Hypoxia-Inducible Factor 1α—HIF1α and Vascular Endothelial Growth Factor—VEGFa), whose expression profiles were previously reported (Oleksiewicz et al. 2011). PRPF40A exhibited the strongest hypoxia association pattern among all genes under investigation (Table 2; Fig. 2a–c). Its mRNA expression level correlated with CYGB (ρ = 0.795, p < 1 × 10−4, N = 130), HIF1α (ρ = 0.841, p < 1 × 10−4, N = 129), and VEGFa (ρ = 0.677, p < 1 × 10−4, N = 95). The hypoxia dependence was observed as well in the case of COL1A1 (Fig. 2d–f), whose expression was associated with CYGB (ρ = 0.709, p < 1 × 10−4, N = 124), HIF1α (ρ = 0.646, p < 1 × 10−4, N = 127) and, to a lesser extent, with VEGFa (ρ = 0.356, p = 5.8 × 10−4, N = 90). Similar relationships were seen in the case of UCP2 (UCP2 vs CYGB: ρ = 0.495, p < 1 × 10−4, N = 129, UCP2 vs HIF1α: ρ = 0.559, p < 1 × 10−4, N = 130 and vs VEGFa: ρ = 0.343, p = 7.1 × 10−4, N = 94, Fig. 2g–i). The expression profiles of COL1A1, PRPF40A, and UCP2 genes correlated with each other. The strongest positive association was observed between PRPF40A and COL1A1 (ρ = 0.612, p < 1 × 10−4, N = 129), PRPF40A and UCP2 (ρ = 0.596, p < 1 × 10−4, N = 132), while the weakest between COL1A1 and UCP2 (ρ = 0.353, p < 1 × 10−4, N = 128).
COL1A1, PRPF40A, and UCP2 expression under stress conditions in vitro
Hypoxia response is activated not only with oxygen depletion, but also with nutrient deficiency, oxidative stress, and other signalling pathways. Therefore, we wanted to assess whether COL1A1, PRPF40A, and UCP2 might be regulated by hypoxia and/or oxidative stress in vitro. Cellular response to oxidative and hypoxic stress was confirmed by testing glutathione content (Fig. 3a) and the mRNA expression of VEGFa (Fig. 3b), respectively. COL1A1 became upregulated in hypoxic conditions in both cell lines; however, this was significant only in CALU1 (RQ = 3.15 ± 0.8 vs 1.0 ± 0.07 in normoxic cells, p = 0.02, Mann–Whitney), but not in H358 (RQ = 2.2 ± 0.7 vs 1.1 ± 0.06, p = 0.074). PRPF40A and UCP2 expression little changed at 1% O2, as only CALU1 exhibited modest upregulation of UCP2 (RQ = 1.5 ± 0.24 vs 0.9 ± 0.15, p = 0.04, Mann–Whitney). Similarly, oxidative stress did not evoke significant changes in the expression of COL1A1, UCP2, and PRPF40A genes. This lack of responsiveness to stress conditions was not caused by CpG methylation, as the promoters of COL1A1, UCP2, and PRPF40A showed low methylation level (MtI < 10%) in both cell lines (Fig. 3f).