Revised Exon Structure of l-DOPA Decarboxylase (DDC) Reveals Novel Splice Variants Associated with Colorectal Cancer Progression

Abstract

Colorectal cancer (CRC) is a highly heterogenous malignancy with an increased mortality rate. Aberrant splicing is a typical characteristic of CRC, and several studies support the prognostic value of particular transcripts in this malignancy. l-DOPA decarboxylase (DDC) and its derivative neurotransmitters play a multifaceted role in physiological and pathological states. Our recent data support the existence of 6 DDC novel exons. In this study, we investigated the existence of additional DDC novel exons and transcripts, and their potential value as biomarkers in CRC. Next-generation sequencing (NGS) in 55 human cell lines coupled with Sanger sequencing uncovered 3 additional DDC novel exons and 20 splice variants, 7 of which likely encode new protein isoforms. Eight of these transcripts were detected in CRC. An in-house qPCR assay was developed and performed in TNM II and III CRC samples for the quantification of transcripts bearing novel exons. Extensive biostatistical analysis uncovered the prognostic value of specific DDC novel exons for patients' disease-free and overall survival. The revised DDC exon structure, the putative protein isoforms with distinct functions, and the prognostic value of novel exons highlight the pivotal role of DDC in CRC progression, indicating its potential utility as a molecular biomarker in CRC.

Keywords: alternative splicing; biogenic amines; colon carcinoma; dopamine; l-aromatic amino acid decarboxylase (AADC); molecular biomarkers; next-generation sequencing (NGS); protein isoforms; serotonin; transcriptomics.

Figure 1

Expression analysis in the 3 cDNA pools derived from colorectal, hepatocellular, and lung cancerous tissues, in which the novel transcripts were more abundant. More specifically, the transcripts with L1 as the first exon containing one of the novel exons (X1 to X9) are shown in the first 8 photos. The transcripts starting with exon N1 are shown in the last photo (bottom right). The pair of primers that was used for the generation of each transcript and the expected product size (s) are written on the panel. The arrows point at each predicted product.

Figure 2

Depiction of the exon structure of the 20 DDC novel transcripts. The names of the novel exons are shown in pink. Exons are presented as boxes and introns as lines, while numbers inside boxes and above lines indicate the length of each exon or intron (in nucleotides), respectively. In the first-round PCR, either of the forward (green color) primers and the common reverse primer (red color) were utilized together. The forward primer (light blue color) in the nested PCR was used along with an exon-specific reverse primer (orange color) for the amplification of the left part of each transcript; the exon-specific forward primer (dark blue color) in the nested PCR was used along with a common reverse primer (brown color) for the amplification of the right part of each transcript. (A) Previously annotated DDC transcripts and the 7 DDC novel transcripts with an open reading frame (ORF) are depicted, accompanied by their accession numbers. Blue and grey boxes indicate coding and non-coding exons, respectively, while novel exons are colored with a more intense blue shade. Arrows (↓) indicate the position of the ATG codon, pentagrams (★) show the position of the stop codon, and question marks (?) represent an undetermined untranslated region (UTR). (B) DDC novel transcripts without an ORF, accompanied by their accession numbers. Novel exons are colored dark gray.

Figure 3

Predicted ligands bound to 4 of the deduced DDC isoforms (is.14, is.15, is.16, and is.18), illustrated with the use of the I-TASSER, COFACTOR (https://zhanglab.ccmb.med.umich.edu/COFACTOR/), and COACH (https://zhanglab.ccmb.med.umich.edu/COACH/) servers. Each ligand is depicted with a green-yellow sphere. Binding aa residues are shown by blue balls and sticks. Besides pyridoxal-5-phosphate (PLP), is.14 and is.16 bind carbidopa, a well-known inhibitor of DDC function. Is.15 binds the amino acid leucine and is.18 binds the amino acid glycine. All these predictions have relatively high confidence scores.

Figure 4

Comparison of DDC exon X1 expression levels among 40 CRC tissue samples and their adjacent non-cancerous counterparts. DDC exon X1 expression is downregulated in colorectal tumors as compared to non-cancerous paired tissues (p = 0.018). The p value was calculated using the Wilcoxon signed-rank test.

Figure 5

Kaplan–Meier survival curves for the disease-free survival (DFS) and overall survival (OS) times of CRC patients. (A) Positive DDC exon X3 expression status is a potential favorable prognostic biomarker of DFS in CRC (p = 0.043). (B) DDC X8-positive patients have a shorter DFS time period than the negative ones (p = 0.015). (C) Positive DDC exon X9 expression status leads to longer CRC patient DFS time (p = 0.014). (D) Patients with DDC exon X3-positive tumors have longer OS period than those with DDC exon X3-negative tumors (p = 0.033). (E) Positive DDC exon X8 expression status is a potential unfavorable prognostic biomarker of OS in CRC (p = 0.010). (F) DDC exon X9-positive patients are characterized by a longer OS time period than the negative ones (p = 0.007).

Figure 6

Stratified Kaplan–Meier survival curves of groups of patients with substantially different prognoses according to TNM stage. Patients with DDC exon X8-positive tumors in TNM stage II have (A) a shorter DFS and (B) OS time period compared to those with DDC exon X8-negative tumors. Patients with DDC exon X9-positive tumors in TNM stage II have (C) a higher DFS and (D) OS time period compared to those with DDC exon X9-negative tumors. Patients with DDC exon X3-positive tumors in TNM stage III have (E) a higher DFS and (F) OS time period compared to those with DDC exon X3-negative tumors.

Figure 7

Overview of the experimental workflow.