A Transcriptomic Approach to the Metabolism of Tetrapyrrolic Photosensitizers in a Marine Annelid

Abstract

The past decade has seen growing interest in marine natural pigments for biotechnological applications. One of the most abundant classes of biological pigments is the tetrapyrroles, which are prized targets due their photodynamic properties; porphyrins are the best known examples of this group. Many animal porphyrinoids and other tetrapyrroles are produced through heme metabolic pathways, the best known of which are the bile pigments biliverdin and bilirubin. Eulalia is a marine Polychaeta characterized by its bright green coloration resulting from a remarkably wide range of greenish and yellowish tetrapyrroles, some of which have promising photodynamic properties. The present study combined metabolomics based on HPLC-DAD with RNA-seq transcriptomics to investigate the molecular pathways of porphyrinoid metabolism by comparing the worm's proboscis and epidermis, which display distinct pigmentation patterns. The results showed that pigments are endogenous and seemingly heme-derived. The worm possesses homologs in both organs for genes encoding enzymes involved in heme metabolism such as ALAD, FECH, UROS, and PPOX. However, the findings also indicate that variants of the canonical enzymes of the heme biosynthesis pathway can be species- and organ-specific. These differences between molecular networks contribute to explain not only the differential pigmentation patterns between organs, but also the worm's variety of novel endogenous tetrapyrrolic compounds.

Keywords: Annelida; bile pigments; bioinformatics; heme; photodynamic; porphyrin metabolism.

Figure 1

Main pigments in the studied organs of Eulalia. The data were retrieved by high-performance liquid chromatography with a diode array detector (HPLC-DAD). Absorbance (AU) and retention time (min) are illustrated for each chromatogram for the extracts from (a) proboscis, (b) epidermis, (c) intestine, and (d) oocytes. Each pigment was labeled according to Martins et al. [22]. The selected wavelengths were 280, 400, and 700 nm to fit absorbance within the range of UV, violet (to detect yellow pigments), and red (to detect green pigments), respectively.

Figure 2

Common pigments between the different organs of Eulalia. The spectra were produced by high-performance liquid chromatography with a diode array detector (HPLC-DAD). (a) Yellow pigments Pr2 (proboscis) and Ep2 (epidermis); (b) Green pigments Int1 (intestine) and Oo3 (oocytes); (c) Green pigments Int3 (intestine), Oo5 (oocytes), and Ep4 (epidermis).

Figure 3

Representative pigment absorption spectra of Eulalia pigments. The data were retrieved using high-performance liquid chromatography with a diode array detector (HPLC-DAD). Each graph displays the Soret band (380–500 nm) and Q bands (500–750 nm) with different degrees of resolution. (a) Pr2, a yellow pigment from the proboscis; (b) Ep4, a green pigment from the epidermis; (c) Int1, a green pigment from the intestine; and (d) Oo2, a green pigment from the oocytes. The retention times were 4.05 min for Pr2, 6.63 min for Ep4, 5.47 min for Int1, and 4.89 min for Oo2.

Figure 4

Relative expression of total tetrapyrrole/porphyrinoid-related ORFs in Eulalia obtained via RNA-seq and de novo transcriptome assembly. Each Venn diagram represents a customized subset of the Uniprot database built from specific search terms. (a) “Porphyrin Eumetazoa”, (b) “chlorins”, (c) “heme biosynthesis”, and (d) “heme degradation”. Relative expression was arbitrarily identified as “high” and “low” when logTPM > 2 and logTPM < −2, respectively.

Figure 5

Heat maps illustrating the overall patterns of differentially transcribed mRNAs potentially pertaining to genes associated with the metabolism of tetrapyrroles and related compounds between the proboscis and epidermis of Eulalia. Separated by annotation subset retrieved from Uniprot. Three independent replicates (corresponding to three different individuals) per organ are represented (PR1-3 and EP1-3 for proboscis and epidermis, respectively). (a) “Porphyrin Eumetazoa”, (b) “heme biosynthesis”, and (c) “heme degradation”. Horizontal and vertical dendrograms plus sidebars represent clustering of biological replicates and genes, respectively. The criteria for the selection of DEGs were significant homology-matching (e-value < 10⁻⁵), FDR-adjusted p ≤ 0.05, and |logFC| > 2. Expression levels were row-normalized and are given as transcripts per million (TPM) and plotted as a color gradient from yellow (lowest) to red (highest). Cluster analysis was based on Euclidean distances as metric and dendrograms were built using complete linkage.

Figure 6

Interaction network of the enzymes involved in heme biosynthesis in the marine annelid Eulalia and validation of RNA-seq by RT-qPCR of a gene subset. (a) The eight enzymes involved in the heme biosynthesis pathway were identified in the translated transcriptome. (1) ALAS1, δ-aminolevulinic acid synthase 1; (2) ALAD, δ-aminolevulinic acid dehydratase; (3) HMBS, hydroxymethylbilane synthase; (4) UROS, uroporphyrinogen synthase; (5) UROD, uroporphyrinogen decarboxylase; (6) CPOX, coproporphyrinogen oxidase; (7) PPOX, protoporphyrinogen oxidase; and (8) FECH, ferrochelatase. Protein–protein interactions were retrieved from the STRING database with information from the “heme biosynthesis” subset. (b) Validation of RNA-seq data was done for mRNAs coding for ALAD, UROS, and PPOX, using RT-qPCR. Relative expression was determined via the 2^−∆∆Ct method following the use of primers designed to amplify the best homology-matched sequences. The results are presented as mean + standard deviation. Despite the trend for overexpression of selected genes in the proboscis, no statistical differences were found between the organs, in accordance with the RNA-seq results.

Figure 7

Protein–protein interactions and respective pathways or functions retrieved for the most abundant transcripts in the proboscis and epidermis. Schematic representation of the potentially highly expressed genes coding for proteins in the “porphyrin Eumetazoa”, “heme biosynthesis”, and “heme degradation” subsets. The criterion to select abundant transcripts was logTPM > 2. The various enriched pathways or functions of the proteins (based on GO terms) are represented by a different color, all with FDR-adjusted p < 0.05. Functional annotation and protein–protein interactions were retrieved from the STRING database.

Figure 8

Phylogenetic trees comparing the sequences of three main heme biosynthesis proteins. Each dendrogram was created using the maximum likelihood method and an LG gamma distribution model (consensus tree obtained from 1,000,000 iterations). The sequences used were from Eulalia predicted transcriptome and Metazoa sequences from Uniprot for (a) ALAD, δ-aminolevulinic acid dehydratase; (b) UROD, uroporphyrinogen decarboxylase; and (c) FECH, ferrochelatase.

Figure 9

A simplified representation of the essential heme metabolic pathways shared between Eulalia and higher-order Eumetazoa, especially vertebrates. ALAS, aminolevulinic acid synthase; ALAD, δ-aminolevulinic acid dehydratase; HMBS, hydroxymethylbilane synthase; UROS, uroporphyrinogen synthase; UROD, uroporphyrinogen decarboxylase; FECH, ferrochelatase; PPOX, protoporphyrinogen oxidase; CPOX, coproporphyrinogen oxidase; HMOX, heme oxygenase; BLVRA, biliverdin reductase.