Differential gene expression during early development in recently evolved and sympatric Arctic charr morphs

• Published in: PeerJ
• Main Authors: Jóhannes Guðbrandsson, Sigríður Rut Franzdóttir, Bjarni Kristófer Kristjánsson, Ehsan Pashay Ahi, Valerie Helene Maier, Kalina Hristova Kapralova, Sigurður Sveinn Snorrason, Zophonías Oddur Jónsson, Arnar Pálsson
• Format: Article in Journal/Newspaper
• Language: English
• Published: PeerJ Inc. 2018
• Subjects: Divergence; Salmonid; Transcriptome; Evolution; Lake Thingvallavatn; Salvelinus alpinus; Medicine; R; Biology (General); QH301-705.5; Arctic; Arctic charr
• Online Access: https://doi.org/10.7717/peerj.4345; https://doaj.org/article/bd5abc5acdd043caa151afc1b6541ee1

Phenotypic differences between closely related taxa or populations can arise through genetic variation or be environmentally induced, leading to altered transcription of genes during development. Comparative developmental studies of closely related species or variable populations within species can help to elucidate the molecular mechanisms related to evolutionary divergence and speciation. Studies of Arctic charr (Salvelinus alpinus) and related salmonids have revealed considerable phenotypic variation among populations and in Arctic charr many cases of extensive variation within lakes (resource polymorphism) have been recorded. One example is the four Arctic charr morphs in the ∼10,000 year old Lake Thingvallavatn, which differ in numerous morphological and life history traits. We set out to investigate the molecular and developmental roots of this polymorphism by studying gene expression in embryos of three of the morphs reared in a common garden set-up. We performed RNA-sequencing, de-novo transcriptome assembly and compared gene expression among morphs during an important timeframe in early development, i.e., preceding the formation of key trophic structures. Expectedly, developmental time was the predominant explanatory variable. As the data were affected by some form of RNA-degradation even though all samples passed quality control testing, an estimate of 3′-bias was the second most common explanatory variable. Importantly, morph, both as an independent variable and as interaction with developmental time, affected the expression of numerous transcripts. Transcripts with morph effect, separated the three morphs at the expression level, with the two benthic morphs being more similar. However, Gene Ontology analyses did not reveal clear functional enrichment of transcripts between groups. Verification via qPCR confirmed differential expression of several genes between the morphs, including regulatory genes such as AT-Rich Interaction Domain 4A (arid4a) and translin (tsn). The data are consistent with a ...