When comparing the known picture of polyploidy in animals and in plants, it is possible to recognize some similarities, namely: (i) multiple and recurrent origins in several well-established taxonomic groups; (ii) a strong and regular association with hybridization events; (iii) the production of genotypic diversity; (iv) a rapid genomic reshuffling; (v) a very active role of transposable elements in allopolyploids; (vi) a comparatively privileged occurrence in harsher environments when compared with their diploid relatives, and (vii) gene silencing and divergence of duplicated genes without disruption of duplicated loci. Research on polyploidy was highly biased towards plants during the last century because polyploidy in animals was for long time considered rare, occasional and irrelevant from an evolutionary perspective. However, as empirically observed in plants, genome rediploidization starts in polyploid organisms immediately after the polyploid shock. Given the speed and dynamicity of this process, evidence of genome multiplication is completely erased over time, and hence, only the most recent events are likely to be acknowledged. Although varying in expression between and within taxonomic groups, polyploidy and hybridization are ubiquitous in animals and may be recurrent, fostering evolution. Since evolutionary allopolyploid genomes behave as biologically diploid, zoologists have to challenge the old paradigm of an irrelevant evolutionary role in animals using current genomic and cytogenomic tools. These methods are most likely to reveal the role of polyploid mechanisms in producing evolutionary novelties. Nonsexual complexes are the perfect models to bridge the gap between empirical and theoretical research, while the evolutionary process is in action. Such animal complexes represent a transient stage that, in general, moves towards a polyploid stage, where bisexuality might be recovered, ultimately giving rise to a new gonochoric species. These pathways are herein illustrated by the Iberian allopolyploid Squalius alburnoides. Some general aspects on this fish's complex are updated and reviewed, namely the reproductive modes of the distinct genomotypes, since variable ploidies and genomic combinations occur in natural populations. Most recent data on the mechanisms of gene expression regulation and the importance of the genomic context in driving allelic expression are also included. It was first demonstrated that a regulatory mechanism involving dosage compensation by gene-copy silencing exists in allotriploid females and that allelic expression patterns differed either between genomically equivalent individuals or within the same individual (between tissues and genes). Thus, instead of a whole haplome inactivation, a biased silencing towards repression of a specific allele was observed as well as a reduction of the transcript levels to the diploid state. See also sister article focusing on plants by Tayalé and Parisod in this themed issue

How allopolyploids are able not only to cope but profit from their condition is a question that remains elusive, but is of great importance within the context of successful allopolyploid evolution. One outstanding example of successful allopolyploidy is the endemic Iberian cyprinid Squalius alburnoides. Previously, based on the evaluation of a few genes, it was reported that the transcription levels between diploid and triploid S. alburnoides were similar. If this phenomenon occurs on a full genomic scale, a wide functional ‘‘diploidization’’ could be related to the success of these polyploids. We generated RNA-seq data from whole juvenile fish and from adult livers, to perform the first comparative quantitative transcriptomic analysis between diploid and triploid individuals of a vertebrate allopolyploid. Together with an assay to estimate relative expression per cell, it was possible to infer the relative sizes of transcriptomes. This showed that diploid and triploid S. alburnoides hybrids have similar liver transcriptome sizes. This in turn made it valid to directly compare the S. alburnoides RNA-seq transcript data sets and obtain a profile of dosage responses across the S. alburnoides transcriptome. We found that 64% of transcripts in juveniles’ samples and 44% in liver samples differed less than twofold between diploid and triploid hybrids (similar expression). Yet, respectively 29% and 15% of transcripts presented accurate dosage compensation (PAA/PA expression ratio of 1 instead of 1.5). Therefore, an exact functional diploidization of the triploid genome does not occur, but a significant down regulation of gene expression in triploids was observed. However, for those genes with similar expression levels between diploids and triploids, expression is not globally strictly proportional to gene dosage nor is it set to a perfect diploid level. This quantitative expression flexibility may be a strong contributor to overcome the genomic shock, and be an immediate evolutionary advantage of allopolyploids.

Assessing allele-specific gene expression (ASE) on a large scale continues to be a technically challenging problem. Certain biological phenomena, such as X chromosome inactivation and parental imprinting, affect ASE most drastically by completely shutting down the expression of a whole set of alleles. Other more subtle effects on ASE are likely to be much more complex and dependent on the genetic environment and are perhaps more important to understand since they may be responsible for a significant amount of biological diversity. Tools to assess ASE in a diploid biological system are becoming more reliable. Non-diploid systems are, however, not uncommon. In humans full or partial polyploid states are regularly found in both healthy (meiotic cells, polynucleated cell types) and diseased tissues (trisomies, non-disjunction events, cancerous tissues). In this work we have studied ASE in the medaka fish model system. We have developed a method for determining ASE in polyploid organisms from RNAseq data and we have implemented this method in a software tool set. As a biological model system we have used nuclear transplantation to experimentally produce artificial triploid medaka composed of three different haplomes. We measured ASE in RNA isolated from the livers of two adult, triploid medaka fish that showed a high degree of similarity. The majority of genes examined (82%) shared expression more or less evenly among the three alleles in both triploids. The rest of the genes (18%) displayed a wide range of ASE levels. Interestingly the majority of genes (78%) displayed generally consistent ASE levels in both triploid individuals. A large contingent of these genes had the same allele entirely suppressed in both triploids. When viewed in a chromosomal context, it is revealed that these genes are from large sections of 4 chromosomes and may be indicative of some broad scale suppression of gene expression.