Computationally predicted gene regulatory networks in molluscan biomineralization identify extracellular matrix production and ion transportation pathways

Abstract

Motivation The molecular processes regulating molluscan shell production remain relatively uncharacterized, despite the clear evolutionary and societal importance of biomineralization. Results Here we built the first computationally predicted gene regulatory network (GRN) for molluscan biomineralization using Antarctic clam (Laternula elliptica) mantle gene expression data produced over an age-categorized shell damage-repair time-course. We used previously published in vivo in situ hybridization expression data to ground truth gene interactions predicted by the GRN and show that candidate biomineralization genes from different shell layers, and hence microstructures, were connected in unique modules. We characterized two biomineralization modules of the GRN and hypothesize that one module is responsible for translating the extracellular proteins required for growing, repairing or remodelling the nacreous shell layer, whereas the second module orchestrates the transport of both ions and proteins to the shell secretion site, which are required during normal shell growth, and repair. Our findings demonstrate that unbiased computational methods are particularly valuable for studying fundamental biological processes and gene interactions in non-model species where rich sources of gene expression data exist, but annotation rates are poor and the ability to carry out true functional tests are still lacking. Availability and implementation The raw RNA-Seq data is freely available for download from NCBI SRA (Accession: PRJNA398984), the assembled and annotated transcriptome can be viewed and downloaded from molluscDB (ensembl.molluscdb.org) and in addition, the assembled transcripts, reconstructed GRN, modules and detailed annotations are all available as Supplementary Files. Supplementary information Supplementary data are available at Bioinformatics online.

Gene network analyses support subfunctionalization hypothesis for duplicated hsp70 genes in the Antarctic clam

“…Fragments of the duplicated inducible forms of hsp70 in the Antarctic clam L. elliptica were previously identified via degenerate PCR (Clark et al 2008) and designated hsp70A (accession number AM293598.1) and hsp70B (accession number AM293600.1). In the present paper, these short fragments were BLAST searched against a recently published in-house mantle transcriptome database for L. elliptica (now available at MolluscDB: https://molluscdb.org/ (Caurcel C (2017)) to identify full-length transcripts for further investigation via the computationally predicted GRN (Sleight et al 2019).…”

“…Briefly, 199,321 transcripts generated by a Trinity mantle transcriptome assembly were collapsed into 18,862 expression clusters with shared expression profiles using a selforganizing tree algorithm (SOTA); i.e. transcripts in the same cluster have tightly correlated expression over all the experimental samples in the dataset and were shown to be coexpressed in vivo using mRNA in situ hybridization (Sleight et al 2019). A brief schematic overview of methods is available in the supplementary material (Supplementary information S1).…”

“…Whilst these types of network calculations have been used to infer in vivo transcriptional regulation in the biomedical literature for some years (Theodoris et al 2015;Walsh et al 2017), the approach is only beginning to be applied to non-model organisms wellpoised to answer questions of an environmental or evolutionary nature (Antczak et al 2015). A computationally predicted gene regulatory network (GRN) has recently been produced using ARACNe for the Antarctic clam Laternula elliptica (Sleight et al 2019). This network was produced from mantle-specific shell damage-repair experimental data and, in addition to its contribution to the study of biomineralization, it also provides a resource to investigate questions related to cellular stress responses in L. elliptica, for example, the identification of putative differentiated functions of paralogous genes through examination of neighbouring genes.…”

Deciphering mollusc shell production: the roles of genetic mechanisms through to ecology, aquaculture and biomimetics

“…These approaches are particularly powerful when combined with analyses of expression modules and gene networks, such as correlative weighted gene co‐ expression network analysis ( WGCNA ) and programs such as the algorithm for the reconstruction of accurate cellular networks (ARACNe) (e.g. De Wit et al ., 2018; Sleight et al ., 2020). These in silico techniques visualize gene interactions and provide major progress towards defining biochemical pathways in non‐model species.…”

“…These in silico techniques visualize gene interactions and provide major progress towards defining biochemical pathways in non‐model species. They are particularly useful for identifying upstream control genes and incorporation of species‐specific transcripts that have little or no annotation, into biochemical pathways (De Wit et al ., 2018; Yarra, 2018; Sleight et al ., 2020; Ramsøe, Clark, & Sleight, 2020).…”