The genome of the toxic invasive species Heracleum sosnowskyi carries an increased number of genes despite absence of recent whole-genome duplications.

Heracleum sosnowskyi, belonging to a group of giant hogweeds, is a plant with large effects on ecosystems and human health. It is an invasive species that contributes to the deterioration of grassland ecosystems. The ability of H. sosnowskyi to produce linear furanocoumarins (FCs), photosensitizing compounds, makes it very dangerous. At the same time, linear FCs are compounds with high pharmaceutical value used in skin disease therapies. Despite this high importance, it has not been the focus of genetic and genomic studies. Here, we report a chromosome-scale assembly of Sosnowsky's hogweed genome. Genomic analysis revealed an unusually high number of genes (55106) in the hogweed genome, in contrast to the 25-35 thousand found in most plants. However, we did not find any traces of recent whole-genome duplications not shared with its confamiliar, Daucus carota (carrot), which has approximately thirty thousand genes. The analysis of the genomic proximity of duplicated genes indicates on tandem duplications as a main reason for this increase. We performed a genome-wide search of the genes of the FC biosynthesis pathway and surveyed their expression in aboveground plant parts. Using a combination of expression data and phylogenetic analysis, we found candidate genes for psoralen synthase and experimentally showed the activity of one of them using a heterologous yeast expression system. These findings expand our knowledge on the evolution of gene space in plants and lay a foundation for further analysis of hogweed as an invasive plant and as a source of FCs.

Origin and diversity of Capsella bursa-pastoris from the genomic point of view.

BACKGROUND: Capsella bursa-pastoris, a cosmopolitan weed of hybrid origin, is an emerging model object for the study of early consequences of polyploidy, being a fast growing annual and a close relative of Arabidopsis thaliana. The development of this model is hampered by the absence of a reference genome sequence. RESULTS: We present here a subgenome-resolved chromosome-scale assembly and a genetic map of the genome of Capsella bursa-pastoris. It shows that the subgenomes are mostly colinear, with no massive deletions, insertions, or rearrangements in any of them. A subgenome-aware annotation reveals the lack of genome dominance-both subgenomes carry similar number of genes. While most chromosomes can be unambiguously recognized as derived from either paternal or maternal parent, we also found homeologous exchange between two chromosomes. It led to an emergence of two hybrid chromosomes; this event is shared between distant populations of C. bursa-pastoris. The whole-genome analysis of 119 samples belonging to C. bursa-pastoris and its parental species C. grandiflora/rubella and C. orientalis reveals introgression from C. orientalis but not from C. grandiflora/rubella. CONCLUSIONS: C. bursa-pastoris does not show genome dominance. In the earliest stages of evolution of this species, a homeologous exchange occurred; its presence in all present-day populations of C. bursa-pastoris indicates on a single origin of this species. The evidence coming from whole-genome analysis challenges the current view that C. grandiflora/rubella was a direct progenitor of C. bursa-pastoris; we hypothesize that it was an extinct (or undiscovered) species sister to C. grandiflora/rubella.

Interspecific comparison of gene expression profiles using machine learning.

Interspecific gene comparisons are the keystones for many areas of biological research and are especially important for the translation of knowledge from model organisms to economically important species. Currently they are hampered by the low resolution of methods based on sequence analysis and by the complex evolutionary history of eukaryotic genes. This is especially critical for plants, whose genomes are shaped by multiple whole genome duplications and subsequent gene loss. This requires the development of new methods for comparing the functions of genes in different species. Here, we report ISEEML (Interspecific Similarity of Expression Evaluated using Machine Learning)-a novel machine learning-based algorithm for interspecific gene classification. In contrast to previous studies focused on sequence similarity, our algorithm focuses on functional similarity inferred from the comparison of gene expression profiles. We propose novel metrics for expression pattern similarity-expression score (ES)-that is suitable for species with differing morphologies. As a proof of concept, we compare detailed transcriptome maps of Arabidopsis thaliana, the model species, Zea mays (maize) and Fagopyrum esculentum (common buckwheat), which are species that represent distant clades within flowering plants. The classifier resulted in an AUC of 0.91; under the ES threshold of 0.5, the specificity was 94%, and sensitivity was 72%.

Local fitness landscape of the green fluorescent protein.

Fitness landscapes depict how genotypes manifest at the phenotypic level and form the basis of our understanding of many areas of biology, yet their properties remain elusive. Previous studies have analysed specific genes, often using their function as a proxy for fitness, experimentally assessing the effect on function of single mutations and their combinations in a specific sequence or in different sequences. However, systematic high-throughput studies of the local fitness landscape of an entire protein have not yet been reported. Here we visualize an extensive region of the local fitness landscape of the green fluorescent protein from Aequorea victoria (avGFP) by measuring the native function (fluorescence) of tens of thousands of derivative genotypes of avGFP. We show that the fitness landscape of avGFP is narrow, with 3/4 of the derivatives with a single mutation showing reduced fluorescence and half of the derivatives with four mutations being completely non-fluorescent. The narrowness is enhanced by epistasis, which was detected in up to 30% of genotypes with multiple mutations and mostly occurred through the cumulative effect of slightly deleterious mutations causing a threshold-like decrease in protein stability and a concomitant loss of fluorescence. A model of orthologous sequence divergence spanning hundreds of millions of years predicted the extent of epistasis in our data, indicating congruence between the fitness landscape properties at the local and global scales. The characterization of the local fitness landscape of avGFP has important implications for several fields including molecular evolution, population genetics and protein design.

Studying RNA-DNA interactome by Red-C identifies noncoding RNAs associated with various chromatin types and reveals transcription dynamics.

Non-coding RNAs (ncRNAs) participate in various biological processes, including regulating transcription and sustaining genome 3D organization. Here, we present a method termed Red-C that exploits proximity ligation to identify contacts with the genome for all RNA molecules present in the nucleus. Using Red-C, we uncovered the RNA-DNA interactome of human K562 cells and identified hundreds of ncRNAs enriched in active or repressed chromatin, including previously undescribed RNAs. Analysis of the RNA-DNA interactome also allowed us to trace the kinetics of messenger RNA production. Our data support the model of co-transcriptional intron splicing, but not the hypothesis of the circularization of actively transcribed genes.

Coding palindromes in mitochondrial genes of Nematomorpha.

Inverted repeats are common DNA elements, but they rarely overlap with protein-coding sequences due to the ensuing conflict with the structure and function of the encoded protein. We discovered numerous perfect inverted repeats of considerable length (up to 284 bp) embedded within the protein-coding genes in mitochondrial genomes of four Nematomorpha species. Strikingly, both arms of the inverted repeats encode conserved regions of the amino acid sequence. We confirmed enzymatic activity of the respiratory complex I encoded by inverted repeat-containing genes. The nucleotide composition of inverted repeats suggests strong selection at the amino acid level in these regions. We conclude that the inverted repeat-containing genes are transcribed and translated into functional proteins. The survey of available mitochondrial genomes reveals that several other organisms possess similar albeit shorter embedded repeats. Mitochondrial genomes of Nematomorpha demonstrate an extraordinary evolutionary compromise where protein function and stringent secondary structure elements within the coding regions are preserved simultaneously.

Hidden cases of tRNA gene duplication and remolding in mitochondrial genomes of amphipods.

The evolution of tRNA genes in mitochondrial (mt) genomes is a complex process that includes duplications, degenerations, and transpositions, as well as a specific process of identity change through mutations in the anticodon (tRNA gene remolding or tRNA gene recruitment). Using amphipod-specific tRNA models for annotation, we show that tRNA duplications are more common in the mt genomes of amphipods than what was revealed by previous annotations. Seventeen cases of tRNA gene duplications were detected in the mt genomes of amphipods, and ten of them were tRNA genes that underwent remolding. The additional tRNA gene findings were verified using phylogenetic analysis and genetic distance analysis. The majority of remolded tRNA genes (seven out of ten cases) were found in the mt genomes of endemic amphipod species from Lake Baikal. All additional mt tRNA genes arose independently in the Baikalian amphipods, indicating the unusual plasticity of tRNA gene evolution in these species assemblages. The possible reasons for the unusual abundance of additional tRNA genes in the mt genomes of Baikalian amphipods are discussed. The amphipod-specific tRNA models developed for MiTFi refine existing predictions of tRNA genes in amphipods and reveal additional cases of duplicated tRNA genes overlooked by using less specific Metazoa-wide models. The application of these models for mt tRNA gene prediction will be useful for the correct annotation of mt genomes of amphipods and probably other crustaceans.

Trypanosomatid mitochondrial RNA editing: dramatically complex transcript repertoires revealed with a dedicated mapping tool.

RNA editing by targeted insertion and deletion of uridine is crucial to generate translatable mRNAs from the cryptogenes of the mitochondrial genome of kinetoplastids. This type of editing consists of a stepwise cascade of reactions generally proceeding from 3' to 5' on a transcript, resulting in a population of partially edited as well as pre-edited and completely edited molecules for each mitochondrial cryptogene of these protozoans. Often, the number of uridines inserted and deleted exceed the number of nucleotides that are genome-encoded. Thus, analysis of kinetoplastid mitochondrial transcriptomes has proven frustratingly complex. Here we present our analysis of Leptomonas pyrrhocoris mitochondrial cDNA deep sequencing reads using T-Aligner, our new tool which allows comprehensive characterization of RNA editing, not relying on targeted transcript amplification and on prior knowledge of final edited products. T-Aligner implements a pipeline of read mapping, visualization of all editing states and their coverage, and assembly of canonical and alternative translatable mRNAs. We also assess T-Aligner functionality on a more challenging deep sequencing read input from Trypanosoma cruzi. The analysis reveals that transcripts of cryptogenes of both species undergo very complex editing that includes the formation of alternative open reading frames and whole categories of truncated editing products.

Active chromatin and transcription play a key role in chromosome partitioning into topologically associating domains.

Recent advances enabled by the Hi-C technique have unraveled many principles of chromosomal folding that were subsequently linked to disease and gene regulation. In particular, Hi-C revealed that chromosomes of animals are organized into topologically associating domains (TADs), evolutionary conserved compact chromatin domains that influence gene expression. Mechanisms that underlie partitioning of the genome into TADs remain poorly understood. To explore principles of TAD folding in Drosophila melanogaster, we performed Hi-C and poly(A)(+) RNA-seq in four cell lines of various origins (S2, Kc167, DmBG3-c2, and OSC). Contrary to previous studies, we find that regions between TADs (i.e., the inter-TADs and TAD boundaries) in Drosophila are only weakly enriched with the insulator protein dCTCF, while another insulator protein Su(Hw) is preferentially present within TADs. However, Drosophila inter-TADs harbor active chromatin and constitutively transcribed (housekeeping) genes. Accordingly, we find that binding of insulator proteins dCTCF and Su(Hw) predicts TAD boundaries much worse than active chromatin marks do. Interestingly, inter-TADs correspond to decompacted inter-bands of polytene chromosomes, whereas TADs mostly correspond to densely packed bands. Collectively, our results suggest that TADs are condensed chromatin domains depleted in active chromatin marks, separated by regions of active chromatin. We propose the mechanism of TAD self-assembly based on the ability of nucleosomes from inactive chromatin to aggregate, and lack of this ability in acetylated nucleosomal arrays. Finally, we test this hypothesis by polymer simulations and find that TAD partitioning may be explained by different modes of inter-nucleosomal interactions for active and inactive chromatin.

An update to database TraVA: organ-specific cold stress response in Arabidopsis thaliana.

BACKGROUND: Transcriptome map is a powerful tool for a variety of biological studies; transcriptome maps that include different organs, tissues, cells and stages of development are currently available for at least 30 plants. Some of them include samples treated by environmental or biotic stresses. However, most studies explore only limited set of organs and developmental stages (leaves or seedlings). In order to provide broader view of organ-specific strategies of cold stress response we studied expression changes that follow exposure to cold (+ 4 °C) in different aerial parts of plant: cotyledons, hypocotyl, leaves, young flowers, mature flowers and seeds using RNA-seq. RESULTS: The results on differential expression in leaves are congruent with current knowledge on stress response pathways, in particular, the role of CBF genes. In other organs, both essence and dynamics of gene expression changes are different. We show the involvement of genes that are confined to narrow expression patterns in non-stress conditions into stress response. In particular, the genes that control cell wall modification in pollen, are activated in leaves. In seeds, predominant pattern is the change of lipid metabolism. CONCLUSIONS: Stress response is highly organ-specific; different pathways are involved in this process in each type of organs. The results were integrated with previously published transcriptome map of Arabidopsis thaliana and used for an update of a public database TraVa: http://travadb.org/browse/Species=AthStress .

Characterization of the mitochondrial genome of the MAX1 type of cytoplasmic male-sterile sunflower.

BACKGROUND: More than 70 cytoplasmic male sterility (CMS) types have been identified in Helianthus, but only for less than half of them, research of mitochondrial organization has been conducted. Moreover, complete mitochondrion sequences have only been published for two CMS sources - PET1 and PET2. It has been demonstrated that other sunflower CMS sources like MAX1, significantly differ from the PET1 and PET2 types. However, possible molecular causes for the CMS induction by MAX1 have not yet been proposed. In the present study, we have investigated structural changes in the mitochondrial genome of HA89 (MAX1) CMS sunflower line in comparison to the fertile mitochondrial genome. RESULTS: Eight significant major reorganization events have been determined in HA89 (MAX1) mtDNA: one 110 kb inverted region, four deletions of 439 bp, 978 bp, 3183 bp and 14,296 bp, respectively, and three insertions of 1999 bp, 5272 bp and 6583 bp. The rearrangements have led to functional changes in the mitochondrial genome of HA89 (MAX1) resulting in the complete elimination of orf777 and the appearance of new ORFs - orf306, orf480, orf645 and orf1287. Aligning the mtDNA of the CMS sources PET1 and PET2 with MAX1 we found some common reorganization features in their mitochondrial genome sequences. CONCLUSION: The new open reading frame orf1287, representing a chimeric atp6 gene, may play a key role in MAX1 CMS phenotype formation in sunflower, while the contribution of other mitochondrial reorganizations seems to appear negligible for the CMS development.

Highly efficient primed spacer acquisition from targets destroyed by the Escherichia coli type I-E CRISPR-Cas interfering complex.

Prokaryotic clustered regularly interspaced short palindromic repeat (CRISPR)-CRISPR associated (Cas) immunity relies on adaptive acquisition of spacers-short fragments of foreign DNA. For the type I-E CRISPR-Cas system from Escherichia coli, efficient "primed" adaptation requires Cas effector proteins and a CRISPR RNA (crRNA) whose spacer partially matches a segment (protospacer) in target DNA. Primed adaptation leads to selective acquisition of additional spacers from DNA molecules recognized by the effector-crRNA complex. When the crRNA spacer fully matches a protospacer, CRISPR interference-that is, target destruction without acquisition of additional spacers-is observed. We show here that when the rate of degradation of DNA with fully and partially matching crRNA targets is made equal, fully matching protospacers stimulate primed adaptation much more efficiently than partially matching ones. The result indicates that different functional outcomes of CRISPR-Cas response to two kinds of protospacers are not caused by different structures formed by the effector-crRNA complex but are due to the more rapid destruction of targets with fully matching protospacers.

High-quality genome assembly of Capsella bursa-pastoris reveals asymmetry of regulatory elements at early stages of polyploid genome evolution.

Polyploidization and subsequent sub- and neofunctionalization of duplicated genes represent a major mechanism of plant genome evolution. Capsella bursa-pastoris, a widespread ruderal plant, is a recent allotetraploid and, thus, is an ideal model organism for studying early changes following polyploidization. We constructed a high-quality assembly of C. bursa-pastoris genome and a transcriptome atlas covering a broad sample of organs and developmental stages (available online at http://travadb.org/browse/Species=Cbp). We demonstrate that expression of homeologs is mostly symmetric between subgenomes, and identify a set of homeolog pairs with discordant expression. Comparison of promoters within such pairs revealed emerging asymmetry of regulatory elements. Among them there are multiple binding sites for transcription factors controlling the regulation of photosynthesis and plant development by light (PIF3, HY5) and cold stress response (CBF). These results suggest that polyploidization in C. bursa-pastoris enhanced its plasticity of response to light and temperature, and allowed substantial expansion of its distribution range.

RNA-seq highlights parallel and contrasting patterns in the evolution of the nuclear genome of fully mycoheterotrophic plants.

BACKGROUND: While photosynthesis is the most notable trait of plants, several lineages of plants (so-called full heterotrophs) have adapted to obtain organic compounds from other sources. The switch to heterotrophy leads to profound changes at the morphological, physiological and genomic levels. RESULTS: Here, we characterize the transcriptomes of three species representing two lineages of mycoheterotrophic plants: orchids (Epipogium aphyllum and Epipogium roseum) and Ericaceae (Hypopitys monotropa). Comparative analysis is used to highlight the parallelism between distantly related fully heterotrophic plants. In both lineages, we observed genome-wide elimination of nuclear genes that encode proteins related to photosynthesis, while systems associated with protein import to plastids as well as plastid transcription and translation remain active. Genes encoding components of plastid ribosomes that have been lost from the plastid genomes have not been transferred to the nuclear genomes; instead, some of the encoded proteins have been substituted by homologs. The nuclear genes of both Epipogium species accumulated nucleotide substitutions twice as rapidly as their photosynthetic relatives; in contrast, no increase in the substitution rate was observed in H. monotropa. CONCLUSIONS: Full heterotrophy leads to profound changes in nuclear gene content. The observed increase in the rate of nucleotide substitutions is lineage specific, rather than a universal phenomenon among non-photosynthetic plants.

Jasmonate-responsive MYB factors spatially repress rutin biosynthesis in Fagopyrum tataricum.

Jasmonates are plant hormones that induce the accumulation of many secondary metabolites, such as rutin in buckwheat, via regulation of jasmonate-responsive transcription factors. Here, we report on the identification of a clade of jasmonate-responsive subgroup 4 MYB transcription factors, FtMYB13, FtMYB14, FtMYB15, and FtMYB16, which directly repress rutin biosynthesis in Fagopyrum tataricum. Immunoblot analysis showed that FtMYB13, FtMYB14, and FtMYB15 could be degraded via the 26S proteasome in the COI1-dependent jasmonate signaling pathway, and that this degradation is due to the SID motif in their C-terminus. Yeast two-hybrid and bimolecular fluorescence complementation assays revealed that FtMYB13, FtMYB14, and FtMYB15 interact with the importin protein Sensitive to ABA and Drought 2 (FtSAD2) in stem and inflorescence. Furthermore, the key repressor of jasmonate signaling FtJAZ1 specifically interacts with FtMYB13. Point mutation analysis showed that the conserved Asp residue of the SID domain contributes to mediating protein-protein interaction. Protoplast transient activation assays demonstrated that FtMYB13, FtMYB14, and FtMYB15 directly repress phenylalanine ammonia lyase (FtPAL) gene expression, and FtSAD2 and FtJAZ1 significantly promote the repressing activity of FtMYBs. These findings may ultimately be promising for further engineering of plant secondary metabolism.

Colonizing the Wild West: Low Diversity of Complete Mitochondrial Genomes in Western North Pacific Killer Whales Suggests a Founder Effect.

In the North Pacific, fish-eating R-type "resident" and mammal-eating T-type "transient" killer whales do not interbreed and differ in ecology and behavior. Full-length mitochondrial genomes (about 16.4 kbp) were sequenced and assembled for 12 R-type and 14 T-type killer whale samples from different areas of the western North Pacific. All R-type individuals had the same haplotype, previously described for R-type killer whales from both eastern and western North Pacific. However, haplotype diversity of R-type killer whales was much lower in the western North Pacific than in the Aleutian Islands and the eastern North Pacific. T-type whales had 3 different haplotypes, including one previously undescribed. Haplotype diversity of T-type killer whales in the Okhotsk Sea was also much lower than in the Aleutian Islands and the eastern North Pacific. The highest haplotype diversity for both R- and T-type killer whales was observed in the Aleutian Islands. We discuss how the environmental conditions during the last glacial period might have shaped the history of killer whale populations in the North Pacific. Our results suggest the recent colonization or re-colonization of the western North Pacific by small groups of killer whales originating from the central or eastern North Pacific, possibly due to favorable environmental changes after the Last Glacial Maximum.

A high resolution map of the Arabidopsis thaliana developmental transcriptome based on RNA-seq profiling.

Arabidopsis thaliana is a long established model species for plant molecular biology, genetics and genomics, and studies of A. thaliana gene function provide the basis for formulating hypotheses and designing experiments involving other plants, including economically important species. A comprehensive understanding of the A. thaliana genome and a detailed and accurate understanding of the expression of its associated genes is therefore of great importance for both fundamental research and practical applications. Such goal is reliant on the development of new genetic and genomic resources, involving new methods of data acquisition and analysis. We present here the genome-wide analysis of A. thaliana gene expression profiles across different organs and developmental stages using high-throughput transcriptome sequencing. The expression of 25 706 protein-coding genes, as well as their stability and their spatiotemporal specificity, was assessed in 79 organs and developmental stages. A search for alternative splicing events identified 37 873 previously unreported splice junctions, approximately 30% of them occurred in intergenic regions. These potentially represent novel spliced genes that are not included in the TAIR10 database. These data are housed in an open-access web-based database, TraVA (Transcriptome Variation Analysis, http://travadb.org/), which allows visualization and analysis of gene expression profiles and differential gene expression between organs and developmental stages.

Comparative analysis of inverted repeats of polypod fern (Polypodiales) plastomes reveals two hypervariable regions.

BACKGROUND: Ferns are large and underexplored group of vascular plants (~ 11 thousands species). The genomic data available by now include low coverage nuclear genomes sequences and partial sequences of mitochondrial genomes for six species and several plastid genomes. RESULTS: We characterized plastid genomes of three species of Dryopteris, which is one of the largest fern genera, using sequencing of chloroplast DNA enriched samples and performed comparative analysis with available plastomes of Polypodiales, the most species-rich group of ferns. We also sequenced the plastome of Adianthum hispidulum (Pteridaceae). Unexpectedly, we found high variability in the IR region, including duplication of rrn16 in D. blanfordii, complete loss of trnI-GAU in D. filix-mas, its pseudogenization due to the loss of an exon in D. blanfordii. Analysis of previously reported plastomes of Polypodiales demonstrated that Woodwardia unigemmata and Lepisorus clathratus have unusual insertions in the IR region. The sequence of these inserted regions has high similarity to several LSC fragments of ferns outside of Polypodiales and to spacer between tRNA-CGA and tRNA-TTT genes of mitochondrial genome of Asplenium nidus. We suggest that this reflects the ancient DNA transfer from mitochondrial to plastid genome occurred in a common ancestor of ferns. We determined the marked conservation of gene content and relative evolution rate of genes and intergenic spacers in the IRs of Polypodiales. Faster evolution of the four intergenic regions had been demonstrated (trnA- orf42, rrn16-rps12, rps7-psbA and ycf2-trnN). CONCLUSIONS: IRs of Polypodiales plastomes are dynamic, driven by such events as gene loss, duplication and putative lateral transfer from mitochondria.

FtSAD2 and FtJAZ1 regulate activity of the FtMYB11 transcription repressor of the phenylpropanoid pathway in Fagopyrum tataricum.

Little is known about the molecular mechanism of the R2R3-MYB transcriptional repressors involved in plant phenylpropanoid metabolism. Here, we describe one R2R3-type MYB repressor, FtMYB11 from Fagopyrum tataricum. It contains the SID-like motif GGDFNFDL and it is regulated by both the importin protein 'Sensitive to ABA and Drought 2' (SAD2) and the jasmonates signalling cascade repressor JAZ protein. Yeast two hybrid and bimolecular fluorescence complementation assays demonstrated that FtMYB11 interacts with SAD2 and FtJAZ1. Protoplast transactivation assays demonstrated that FtMYB11 acts synergistically with FtSAD2 or FtJAZ1 and directly represses its target genes via the MYB-core element AATAGTT. Changing the Asp122 residue to Asn in the SID-like motif results in cytoplasmic localization of FtMYB11 because of loss of interaction with SAD2, while changing the Asp126 residue to Asn results in the loss of interaction with FtJAZ1. Overexpression of FtMYB11or FtMYB11D126N in F. tataricum hairy roots resulted in reduced accumulation of rutin, while overexpression of FtMYB11D122N in hairy roots did not lead to such a change. The results indicate that FtMYB11 acts as a regulator via interacting with FtSAD2 or FtJAZ1 to repress phenylpropanoid biosynthesis, and this repression depends on two conserved Asp residues of its SID-like motif.

Transcriptome-based phylogeny of endemic Lake Baikal amphipod species flock: fast speciation accompanied by frequent episodes of positive selection.

Endemic species flocks inhabiting ancient lakes, oceanic islands and other long-lived isolated habitats are often interpreted as adaptive radiations. Yet molecular evidence for directional selection during species flocks radiation is scarce. Using partial transcriptomes of 64 species of Lake Baikal (Siberia, Russia) endemic amphipods and two nonendemic outgroups, we report a revised phylogeny of this species flock and analyse evidence for positive selection within the endemic lineages. We confirm two independent invasions of amphipods into Baikal and demonstrate that several morphological features of Baikal amphipods, such as body armour and reduction in appendages and sensory organs, evolved in several lineages in parallel. Radiation of Baikal amphipods has been characterized by short phylogenetic branches and frequent episodes of positive selection which tended to be more frequent in the early phase of the second invasion of amphipods into Baikal when the most intensive diversification occurred. Notably, signatures of positive selection are frequent in genes encoding mitochondrial membrane proteins with electron transfer chain and ATP synthesis functionality. In particular, subunits of both the membrane and substrate-level ATP synthases show evidence of positive selection in the plankton species Macrohectopus branickii, possibly indicating adaptation to active plankton lifestyle and to survival under conditions of low temperature and high hydrostatic pressures known to affect membranes functioning. Other functional categories represented among genes likely to be under positive selection include Ca-binding muscle-related proteins, possibly indicating adaptation to Ca-deficient low mineralization Baikal waters.