Type XX Collagen Is Elevated in Circulation of Patients with Solid Tumors

Abstract

In the tumor microenvironment, the extracellular matrix (ECM) has been recognized as an important part of cancer development. The dominant ECM proteins are the 28 types of collagens, each with a unique function in tissue architecture. Type XX collagen, however, is poorly characterized, and little is known about its involvement in cancer. We developed an ELISA quantifying type XX collagen, named PRO-C20, using a monoclonal antibody raised against the C-terminus. PRO-C20 and PRO-C1, an ELISA targeting the N-terminal pro-peptide of type I collagen, was measured in sera of 219 patients with various solid cancer types and compared to sera levels of 33 healthy controls. PRO-C20 was subsequently measured in a separate cohort comprising 36 patients with pancreatic ductal adenocarcinoma (PDAC) and compared to 20 healthy controls and 11 patients with chronic pancreatitis. PRO-C20 was significantly elevated in all cancers tested: bladder, breast, colorectal, head and neck, kidney, lung, melanoma, ovarian, pancreatic, prostate, and stomach cancer (p < 0.01−p < 0.0001). PRO-C1 was only elevated in patients with ovarian cancer. PRO-C20 could discriminate between patients and healthy controls with AUROC values ranging from 0.76 to 0.92. Elevated levels were confirmed in a separate cohort of patients with PDAC (p < 0.0001). High PRO-C20 levels (above 2.57 nM) were predictive of poor survival after adjusting for the presence of metastasis, age, and sex (HR: 4.25, 95% CI: 1.52−11.9, p-value: 0.006). Circulating type XX collagen is elevated in sera of patients with various types of cancer and has prognostic value in PDAC. If validated, PRO-C20 may be a novel biomarker for patients with solid tumors and can help understand the ECM biology of cancer.

Keywords: ECM; PDAC; biomarker; cancer; serum; type XX collagen.

Figure 1

Specificity of the PRO-C20 assay. Inhibition curve of the standard peptide (RHLEGRGEPGAVGQMGSPGQQGASTQGLWE), elongated peptide (QGASTQGLWES), truncated peptide (QGASTQGLW), and a non-sense standard peptide (SHAHQRTGGN) as well as a non-sense coater peptide (Biotin-SHAHQRTGGN). Peptides were diluted twofold in series from 16 nM to 0.25 nM. Signal is shown as the ratio between measured optical density of the sample (OD) and the measured optical density of a blank buffer sample (background OD) as a function of peptide concentration in nM on a logarithmic scale. Points correspond to means of duplicate measurements. Lines correspond to four-parametric logistic regression of each dilution series. Inhibition of the signal indicates competition for antibody binding.

Figure 2

PRO-C20 (left) and PRO-C1 (right) in cohort 1. Quantification of PRO-C20 and PRO-C1 in serum from healthy controls (n = 33), bladder cancer (n = 20), breast cancer (n = 20), colorectal cancer (n = 20), head and neck cancer (n = 20), kidney cancer (n = 20), lung cancer (n = 20), melanoma (n = 20), ovarian cancer (n = 19), pancreatic cancer (n = 20), prostate cancer (n = 20), and stomach cancer (n = 20). Biomarker levels are presented as Tukey-style boxplots with data-point jitter. Samples measuring below the lower limit of measurement range were given the value of that limit, as determined in the validation of the assays. Differences in biomarker levels between cancer groups and the healthy controls were evaluated by ordinary ANOVA followed by multiple comparisons to the controls with Dunnett test. **** indicates a p-value below 0.0001. *** below 0.001. ** below 0.01.

Figure 3

PRO-C20 in cohort 1 according to cancer stage. Quantification of PRO-C20 in cohort 1 was stratified into cancer types and plotted as a function of cancer stage. Biomarker levels are presented as Tukey-style boxplots with datapoint jitter. Differences in PRO-C20 according to cancer stage in each cancer type was evaluated by ordinary ANOVA.

Figure 4

PRO-C20 in cohort 2. Left: Quantification of PRO-C20 in sera from healthy controls (n = 20), chronic pancreatitis (n = 11), and pancreatic ductal adenocarcinoma (PDAC, n = 36). PRO-C20 levels are presented as Tukey-style boxplots with data-point jitter. Differences in PRO-C20 levels between the groups were evaluated by ordinary ANOVA followed by multiple comparisons with Tukey’s HSD test. **** indicates a p-value below 0.0001. *** below 0.001. ns no significance. Right: Quantification of PRO-C20 in cohort 2 plotted as a function of cancer stage. Biomarker levels are presented as Tukey-style boxplots with datapoint jitter. Differences in PRO-C20 according to cancer stage were evaluated by ordinary ANOVA.

Figure 5

Kaplan–Meier curves of survival in patients with pancreatic ductal adenocarcinoma grouped into high (above 2.57 nM) or low PRO-C20 (below 2.57 nM) shaded with 95% confidence intervals.

Figure 6

The tumor microenvironment accumulates collagens and forms fibrosis. The primary components of tumor fibrosis are fibrillar collagens, such as type I or type III collagen, and FACIT collagens, such as type XX collagen, are thought to organize the fibrillar collagens. Fragments of type XX collagen find their way to the circulation, where they can be detected in blood samples by a monoclonal antibody (mAb) raised against the C-terminus of type XX collagen and quantified in an ELISA. Figure made with BioRender.