Journal of Molecular Evolution (2021) 89:639–655 https://doi.org/10.1007/s00239-021-10036-y ORIGINAL ARTICLE The Evolution of Hemocyanin Genes in Caenogastropoda: Gene Duplications and Intron Accumulation in Highly Diverse Gastropods Gabriela Giannina Schäfer1  · Lukas Jörg Grebe1 · Robin Schinkel1 · Bernhard Lieb1 Received: 22 April 2021 / Accepted: 15 October 2021 / Published online: 10 November 2021 © The Author(s) 2021 Abstract Hemocyanin is the oxygen transport protein of most molluscs and represents an important physiological factor that has to be well-adapted to their environments because of the strong influences of abiotic factors on its oxygen affinity. Multiple independent gene duplications and intron gains have been reported for hemocyanin genes of Tectipleura (Heterobranchia) and the caenogastropod species Pomacea canaliculata, which contrast with the uniform gene architectures of hemocyanins in Vetigastropoda. The goal of this study was to analyze hemocyanin gene evolution within the diverse group of Caeno- gastropoda in more detail. Our findings reveal multiple gene duplications and intron gains and imply that these represent general features of Apogastropoda hemocyanins. Whereas hemocyanin exon–intron structures are identical within different Tectipleura lineages, they differ strongly within Caenogastropoda among phylogenetic groups as well as between paralogous hemocyanin genes of the same species. Thus, intron accumulation took place more gradually within Caenogastropoda but finally led to a similar consequence, namely, a multitude of introns. Since both phenomena occurred independently within Heterobranchia and Caenogastropoda, the results support the hypothesis that introns may contribute to adaptive radiation by offering new opportunities for genetic variability (multiple paralogs that may evolve differently) and regulation (multi- ple introns). Our study indicates that adaptation of hemocyanin genes may be one of several factors that contributed to the evolution of the large diversity of Apogastropoda. While questions remain, this hypothesis is presented as a starting point for the further study of hemocyanin genes and possible correlations between hemocyanin diversity and adaptive radiation. Keywords Hemocyanin · Adaptation · Gene structure · Intron accumulation · Gene duplication · Caenogastropoda Background and terrestrial ecosystems, as well as all kinds of intermedi- ate environments. The numerous habitat shifts, which were Mollusca is the second largest animal phylum and includes undergone multiple times independently by different groups over 82,000 extant species (for numbers cf. WoRMS Edi- of Apogastropoda, were enabled by a multitude of adapta- torial Board 2020). The great diversity of this phylum is tions that resulted in enormous diversification. In addition represented best by the two large gastropod groups Het- to the evolution of a range of different lifestyles and mor- erobranchia and Caenogastropoda which together form the phological adaptations, modifications of respiratory systems clade Apogastropoda. They comprise over 64,000 species have been essential during habitat shifts. In addition to the living in various habitats including the sea, fresh waters evolution of new respiratory organs such as pneumostomes and lungs (Dayrat and Tillier 2002; Jörger et al. 2010; Kocot et al. 2013; Schrödl 2014), molecular adaptations that influ- Handling editor: John Bracht. ence respiration have been detected, e.g., adaptations of mitochondrial complexes of Panpulmonata to increase met- * Bernhard Lieb lieb@uni-mainz.de abolic efficiency (Romero et al. 2016b) or the evolution of multiple metabolic states using different levels of available Gabriela Giannina Schäfer schaefga@uni-mainz.de oxygen (Schweizer et al. 2019). Another very important factor of gastropod respiration 1 Institute of Molecular Physiology, Johannes Gutenberg- that has to be adapted to environmental conditions is the University of Mainz, Johann-Joachim-Becher-Weg 7, oxygen transporter hemocyanin. Previous studies have 55128 Mainz, Germany Vol.:(012 3456789) 6 40 Journal of Molecular Evolution (2021) 89:639–655 shown that oxygen affinity, which strongly influences the before the first/after the first/after the second nucleotide of function of hemocyanin, is temperature dependent (Brix a codon). Within all FUs of molluscan hemocyanins that et al. 1989, 1995; Mangum 1990; Miller 1985; Miller and have been analyzed so far, they lie at almost equivalent van Holde 1974). Thus, shifts to habitats with different positions just upstream of linker peptide coding regions temperatures must be accompanied by adaptations of these (Altenhein et al. 2002; Bergmann et al. 2006; Lieb et al. proteins to sustain a sufficient oxygen supply. In particu- 2001; Schäfer et al. 2021b). Accordingly, these introns lar, environments with varying temperatures, e.g., land and are termed linker introns, while those lying within FU- intertidal zones in contrast to solely marine habitats, require coding regions are termed internal introns. The numbers well-adapted oxygen transport proteins. Different hemocya- and positions of internal introns are less conserved and nin paralogs can have different oxygen affinities (Swerdlow differ between hemocyanins of different molluscan line- et al. 1996), and differential expression helps to adapt to var- ages (Fig. 1B). Previous studies showed that the number of ying oxygen conditions (e.g., low oxygen pressure in eggs of these internal introns varies greatly between Octopodoidea Sepia officinalis; Gutowska and Melzner 2009; Strobel et al. (5 internal introns) or Lepetellida (Vetigastropoda; 8 inter- 2012). We previously reported a multitude of hemocyanin nal introns) and Tectipleura (Heterobranchia; 46 internal gene duplications in different species of Tectipleura (Schäfer introns) but are conserved within these different groups et al. 2018). This large group of Heterobranchia comprises of molluscs (Altenhein et al. 2000, 2002; Lieb et al. 2001; very diverse snails that conquered land and freshwater sev- Yao et al. 2019). eral times independently in different lineages (Dinapoli Despite the strong conservation of the cylindrical hemo- and Klussmann-Kolb 2010; Jörger et al. 2010; Kano et al. cyanin structure, their subunits and their genes, deviations 2016; Kocot et al. 2013; Romero et al. 2016a). Therefore, we from these basic structures have been described for several hypothesized that hemocyanin duplications may have helped molluscan groups. These deviations mostly concern the to increase genetic variability by leading to a multitude of number of functional units which changed due to domain hemocyanin paralogs with potentially different properties duplications or losses, e.g., hemocyanins of Cephalopoda and/or varying expression patterns. Accordingly, they may lack FU-h (van Holde and Miller 1995). While the basic represent one of many factors that have enabled the exploita- structure of gastropod hemocyanins corresponds to the tion of new habitats and extremely large adaptive radiation typical eight functional unit domains, multiple variations (Schäfer et al. 2018). have been found for hemocyanins of Caenogastropoda and The overall shape of functional molluscan hemocyanin are discussed below. proteins is a partly hollow cylinder of 4 MDa formed by Within the extremely large and diverse Cerithioidea decamers of 35 nm in diameter which can assemble into (Cerithiida, Caenogastropoda), the so-called mega-hemo- di- or multidecamers (Fig. 1A). These large oxygen trans- cyanin has been identified (Lieb et al. 2010). It represents port molecules float freely within the hemolymph of most a hemocyanin tridecamer that includes two typical decam- molluscs (van Holde and Miller 1995). The basic struc- ers built from 400 kDa subunits and additionally one larger ture of a single 400 kDa polypeptide subunit encompasses decamer that is located between the two typical decam- eight paralogous domains called functional units a, b, …, h ers. This larger decamer is composed of subunits with a (FU-a to FU-h), which are connected by short linker regions molecular mass of 550 kDa. These 550 kDa mega-hemo- (Fig. 1B). The FUs have similar tertiary structures forming cyanin subunits lack FU-g and FU-h but encompass six 45 to 50 kDa large globular substructures of the polypeptides additional functional units which are paralogous to FU-f and comprising one oxygen binding site each. Thus, one (FU-f1, FU-f2 … FU-f6) (Gatsogiannis et al. 2015). These didecamer, which is the most common hemocyanin molecule additional FUs reach within the center of the molecule in gastropods, encompasses 160 oxygen binding sites (basic and fill the mega-hemocyanin cylinders. Therefore, they structure reviewed in Markl 2013 and Kato et al. 2018). The increase the oxygen transport capacity. The viscosity and basic composition of hemocyanin subunits, including mul- the colloid-osmotic pressure of the hemolymph, however, tiple FU domains as well as the primary structures of these remain the same as in a typical hemocyanin tridecamer, FUs are highly conserved across all different molluscan thus, the oxygen transport efficiency is increased (Gatso- classes that have been analyzed thus far (overview in Markl giannis et al. 2015). The ability to differentially express 2013). the 400  kDa and 550  kDa hemocyanin subunits most The segmentation of molluscan hemocyanin subunits likely facilitates variable ratios of typical hemocyanins in multiple FU domains is also represented by the highly and mega-hemocyanins. This may further help to adapt conserved basic exon–intron structure of their genes to different living conditions and may have accelerated (Lieb et al. 2001). Gene segments that encode for dif- the adaptive radiation of the extremely diverse group of ferent functional units are separated by phase 1 introns Cerithioidea (Lieb et al. 2010). (Fig. 1B; intron phases 0/1/2 are defined as being located 1 3 Journal of Molecular Evolution (2021) 89:639–655 641 (A) side view top view side view collar (B) Tecpleura (Heterobranchia) Lepetellida (Vegastropoda) Octopodoidea Naulus FU-a FU-b FU-c FU-d FU-e FU-f FU-g FU-h = linker pepde = linker intron; all in Phase 1 = internal intron; Phase 0, 1 and 2 Fig. 1 Molluscan hemocyanin: molecules and genes. A Typical gas- functional units. Hemocyanin genes in Cephalopoda do not contain tropod hemocyanin didecamer and based on the 9 Å model of KLH1 FU-h. Large boxes represent functional units (FU-a, FU-b, …, FU-h), (Gatsogiannis and Markl 2009, PDB: 4BED). The wall (FU-a– and small boxes between them represent linker peptides. Arrows FU-f) is colored in dark blue. The collar is restricted to both sides symbolize intron positions in hemocyanin genes with respect to the of the didecamer and built by 10 FU-g (cyan) and 10 FU-h (light coding sequences. Linker introns are conserved in phase 1 within all blue). Side and top views are depicted with one hemocyanin subu- known hemocyanin genes (yellow arrows). Internal introns occur in nit dimer highlighted in gold (wall) and gray/yellow (collar: FU- all intron phases and are color-coded accordingly: located before the g/h). B Exon–intron structure of molluscan hemocyanins. Shown first (phase 0, white), after the first (phase 1, gray) or after the second are coding sequences of molluscan hemocyanins. Their genes typi- (phase 2, black) nucleotide of a codon. Hemocyanin genes in Tecti- cally contain ~ 10,200–10,300 nucleotides coding for the eight func- pleura comprise a significantly larger number of internal introns than tional units (FU-a, FU-b, …, FU-h) of one hemocyanin polypeptide those in Lepetellida, Octopodoidea or Nautilus adapted from Schäfer subunit. FU-h is approximately 300 nucleotides longer than the other et al. 2021a (Color figure online) Recently, we reported an additional variation in one of mass within the hemocyanin didecamer of up to 800 kDa the two hemocyanin subunits of Rapana venosa (RtH2 and may influence the function of this hemocyanin mol- derived from the synonym R. thomasiana) and Nucella ecule within these species (Schäfer et al. 2021a). lapillus (NlH2) (Schäfer et al. 2021a). Both species belong Chiumiento et  al. (2020) analyzed hemocyanins of to Muricidae, which represents another main group of Pomacea canaliculata, a species that belongs to Ampul- Caenogastropoda (Fig. 2). We identified 118 (in RtH2) and lariida and represents a third main group of Caenogas- 340 (in NlH2) highly hydrophilic amino acids within the tropoda (Fig. 2). They identified four hemocyanin subunits N-terminal region of FU-g in addition to the highly con- that correspond to the basic polypeptide structure of this served amino acids within a typical molluscan hemocya- oxygen transporter but encompass a remarkably larger nin. These additional amino acids seem to build an extra number of introns than hemocyanins of Vetigastropoda 1 3 642 Journal of Molecular Evolution (2021) 89:639–655 Hypsogastropoda Sorbeoconcha Caenogastropoda Cephalo- poda Apogastropoda Gastropoda = newly derived gene structures = reannotaon of published gene structures = already published gene structures Fig. 2 Broad systematics of Gastropoda focusing on Caenogas- species included in this study and their numbers of hemocyanin (Hc) tropoda. Despite a large number of phylogenetic studies on Caeno- paralogs are shown in brackets. Groups in which hemocyanin gene gastropoda, many phylogenetic relationships within that large group structures are newly derived in this study, reannotated or already pub- of gastropods remain unresolved. The depicted systematics are com- lished are indicated bined from different studies by Ponder et al. (2019). Caenogastropoda or Cephalopoda (27–32 in P. canaliculata, only 9–15 in and N. lapillus but does not belong to the siphonate clade Vetigastropoda or Cephalopoda; Altenhein et al. 2002; (Fig. 2; Ponder et al. 2019). Bergmann et al. 2006; Lieb et al. 2001; Yao et al. 2019). Since hemocyanin genes of Tectipleura include 53 introns each (Schäfer et al. 2021b), large numbers of introns in Methods hemocyanin genes may be a feature of Apogastropoda in general. According to Chiumiento et al. (2020), the Animal Sampling and DNA Isolation exon–intron architectures of the four hemocyanin genes of P. canaliculata differ from each other. This contrasts with One individual of M. tuberculata was taken from a fresh- the exon–intron structures of hemocyanins that have been water aquarium at the Institute of Molecular Physiology in analyzed for Vetigastropoda (Lepetellida), Heterobranchia Mainz. Three specimens of N. lapillus were collected at the (Tectipleura) and Octopodoidea, which are highly con- western Atlantic coast of Brittany, France (Schäfer et al. served across different hemocyanin paralogs (Altenhein 2021a). DNA of one individual of both species was iso- et al. 2002; Schäfer et al. 2021b). lated from foot tissue using the E.Z.N.A.® Mollusc DNA Motivated by the above exceptions in the organization Kit (Omega Bio-Tek, Norcross, GA, USA). Via a Nanodrop of caenogastropod hemocyanins, we analyzed the evolu- (Thermo Fisher Scientific, Waltham, MA, USA), the DNA tion of the genes coding for this oxygen transporter within was checked for purity and quantified. Subsequently, the Caenogastropoda in more detail. Therefore, we inferred DNA was sent to StarSeq in Mainz, Germany, for next- the evolutionary background of caenogastropod hemocya- generation sequencing (NGS, Illumina Next Seq500) and nin genes by reconstructing phylogenetic trees and com- library preparation to subsequently enable the reconstruction pared their exon–intron structures. These results revealed of hemocyanin gene structures (see below). a detailed, novel scenario of intron evolution in gastropod hemocyanins. Therefore, our analysis included the previ- In Silico Assembly of Hemocyanin cDNAs of L. ously described hemocyanins of the Cerithioidea Mela- saxatilis and P. canaliculata noides tuberculata [MtH400 and MtH550 (Lieb et al. 2010), the Muricidae species R. venosa and N. lapillus (RtH and Hemocyanin cDNA sequence assemblies were per- NlH; Gebauer et al. 1999; Schäfer et al. 2021b) and three formed with Geneious 9.1.8 (Kearse et  al. 2012) using hemocyanins of P. canaliculata (Ampullariida; Chiumiento publicly available transcriptomic raw data of L. saxatilis et al. 2020)]. Additionally, we characterized hemocyanins (SRR9651721, SRR9651722, SRR9651724) and P. canali- from Littorina saxatilis, a species that belongs to the same culata (SRR6429145, SRR6429146, SRR6429153) to obtain large group of Hypsogastropoda as the Muricidae R. venosa hemocyanin coding sequences. Paired-end reads were set, 1 3 Journal of Molecular Evolution (2021) 89:639–655 643 sequencing adapters were removed, and transcriptomic raw raw reads. Trimmed and paired-end reads were then mapped reads were quality trimmed with Geneious 9.1.8 (Kearse to coding sequences of R. venosa (BK014286, BK014287), et al. 2012). Processed reads of L. saxatilis were mapped N. lapillus (MT939254, MT939255), M. tuberculata to the previously published cDNA sequence of the 400 kDa (KC405575, KC405576) and L. saxatilis (obtained in this hemocyanin of M. tuberculata (KC405575, overlap identity: study, BK014376, BK014375). The mapping results showed 70%). Those of P. canaliculata were mapped to the cDNA that some parts of the cDNA sequences were not covered by sequences that we deduced from the published hemocyanin genomic NGS data or showed inconsistencies. These sec- gene structures (Chiumiento et al. 2020). Reads that mapped tions were used to subdivide the cDNA sequences into dif- against the known references were used as references for ferent sections representing hypothetical exons. To obtain iterative mappings of the remaining reads to elongate cDNA all splice sites, these sections were extended by repetitive fragments and to obtain the full-length coding sequences mappings of genomic NGS data until their 3’ and 5’ ends [minimum overlap: 60 nucleotides; minimum overlap iden- deviated by at least ten base pairs from cDNA sequences tity: 99%; maximum mismatches: 1%; using Geneious 9.1.8 (procedure explained in more detail in Schäfer et al. 2021b). (Kearse et al. 2012)]. This mapping process was reiterated In this way, we derived intron positions within hemocyanin until the isolated fragments resulted in complete hemocyanin genes. Therefore, our analyses did not include characteriza- coding sequences. tion of intron lengths or sequences. The corrected cDNA The existence of multiple hemocyanin genes per species sequences of P. canaliculata (BK014379, BK014378, and the repetitive structure of their cDNAs, which con- BK014377) were compared with genomic sequences by tain sequences coding for functional units that share some Chiumiento et al. (2020; cf. https:// doi. org/ 10. 5061/ dryad. highly conserved amino acid motifs, may challenge correct 15nd8 v3) to deduce splice site positions. assemblies. To preclude such misassemblies, we verified the sequences by (i) low sensitive mappings that enable misas- Sequence Alignment and Phylogenetic Tree sembly detection, (ii) analyzing the sequences for highly Generation identical sections between hemocyanin sequences of a spe- cies to enable manual double checking for correct assem- We used MEGA 7 (Kumar et al. 2016) to align amino acid blies and (iii) using paired-end reads. For a more detailed sequences by applying the implemented Muscle algorithm description of sequence assembly and verification, see and to determine LG + G + I + F as the best evolutionary Schäfer et al. 2021a. model. This model was used to compute the maximum like- lihood tree with branch supports based on 100 bootstrap rep- Reconstructing Exon–Intron Structures licates using MEGA version 7 (Kumar et al. 2016). Hemo- of Hemocyanin Genes cyanin sequences of the Cephalopoda Nautilus pompilius and Enteroctopus dofleini were used as outgroups. We used For the reconstruction of gene architectures, we used MrBayes 3.2.6 (Huelsenbeck and Ronquist 2001), which Geneious 9.1.8 as a bioinformatic tool (Kearse et al. 2012) is implemented in Geneious 9.1.8 (Kearse et al. 2012), to to map genomic NGS data to hemocyanin coding sequences. conduct Bayesian inference based on two parallel runs of For M. tuberculata and N. lapillus, we used NGS data four Monte Carlo Markov chains (MCMC) with one million sequenced by StarSeq in Mainz, Germany (see above), generations, a subsampling frequency of 500 and a burn-in which included 104,512,762 and 195,550,720 sequences, of 250,000. respectively. Public genomic NGS data were used for R. venosa (SRR5371534; 661,123,146 sequences) and L. saxa- tilis (SRR7976330; 502,027,256 sequences). All genomic NGS data were processed as described for transcriptomic Table 1 Hemocyanins of LisaH1 LisaH2 PcH I PcH IIb PcH III L. saxatilis (LisaH) and P. canaliculata (PcH) Accession number BK014376 BK014375 BK014379 BK014378 BK014377 CDS (nt) 10,308 10,278 10,278 10,287 10,272 Primary structure (aa) 3436 3426 3426 3429 3424 Deduced molecular mass (kDa) 392 391 391 391 390 Accession numbers; the lengths of coding sequences (CDS) in nucleotides (nt); the number of amino acids (aa) for the deduced primary structure of the polypeptides; and the calculated molecular weight in kDa are shown 1 3 644 Journal of Molecular Evolution (2021) 89:639–655 Caeno- */* NlH1RtH1 Muricidae gastropoda */* NlH2 (siphonate Hypsogastropoda)*/* */* RtH2 98/* LisaH1 Littorinida93/- LisaH2 (asiphonate Hypsogastropoda) */* MtH400 */* MtH Cerithiida550 a-f PcH I 87/* PcH IIb Ampullariida */* PcH III */* LsH1 */* */* LsH2 CaH alpD */* HpH alpD Tectipleura */* */* CaH alpN (Heterobranchia) HpH alpN Apogastropoda*/* CaH beta*/* HpH beta */* HrH1 */* HtH1 KLH1 Lepetellida */* KLH2 (Vetigastropoda) */* HrH2 */* */* HtH2Cephalopoda 0,10 / = hemocyanin gene duplications in Caenogastropoda / Tectipleura and Lepetellida Fig. 3 Maximum likelihood (ML) tree of gastropod hemocyanins. Megathura crenulata KLH1 + 2). Gene duplications within hemo- The phylogenetic tree is based on an amino acid sequence alignment cyanins from Tectipleura and Lepetellida are represented by gray and conducted with MUSCLE (Edgar 2004) implemented in MEGA7 arrows. The tree was rooted with hemocyanins of the Cephalopoda (Kumar et  al. 2016) using the LG + G + I + F model. Independ- Nautilus pompilius (NpH) and Enteroctopus dofleini ( OdHA and ent hemocyanin gene duplications (symbolized by orange arrows) OdHG). Nodes are congruent with those obtained by Bayesian infer- occurred within all main groups of Caenogastropoda included in this ence except for the position of LisaH2, which is grouped together study, namely, within Ampullariida, Cerithiida and Hypsogastropoda. with hemocyanins of Muricidae in the Bayesian analysis tree (Supple- Therefore, the tree includes hemocyanins from the following Cae- ment 1). Hemocyanin gene duplications in Caenogastropoda are not nogastropoda: Pomacea canaliculata (PcH I + IIb + III), M. tubercu- affected by the previously described deviations from maximum like- lata (MtH400+550), Littorina saxatilis (LisaH1 + 2), Rapana venosa lihood or Bayesian inference. Nodes are labeled with bootstrap (BS) (RtH1 + 2) and Nucella lapillus (NlH1 + 2). It further encompasses percentages based on 100 replicates from ML analyses and posterior hemocyanins from Tectipleura (Helix pomatia HpHαD + αN + β; probabilities (PP) computed by MrBayes (BS/PP). Asterisks indicate Cornu aspersum CaHαD + αN + β; Lymnaea stagnalis LsH1 + 2) and nodes supported by BS ≥ 99%/PP ≥ 0.99 Lepetellida (Haliotis tuberculata HtH1 + 2; Haliotis rubra HrH1 + 2; Results (two hemocyanin paralogs: KC405575, KC405576). Addi- tionally, we revised the hemocyanin sequences of the Amp- To reveal relationships between different hemocyanin para- ullariida species P. canaliculata (three hemocyanin paralogs: logs and to identify hemocyanin gene duplication events BK014379, BK014378, BK014377) and newly assembled within Caenogastropoda, we inferred the phylogeny of the cDNA of the hemocyanins of L. saxatilis (two hemocya- these proteins, including eleven hemocyanin sequences of nin paralogs: BK014376, BK014375), which belongs to the over 10,000 nucleotides each from species that belong to asiphonate Hypsogastropoda (Fig. 2). four different main groups of Caenogastropoda. To further investigate the evolution of hemocyanin genes in Caenogas- Canonical Hemocyanin Coding Sequences Identified tropoda, we compiled exon–intron structures for all eleven for L. saxatilis and P. canaliculata caenogastropod hemocyanin genes. These analyses include the previously published hemocyanin cDNA sequences We obtained two complete hemocyanin coding sequences of the Muricidae species (siphonate Hypsogastropoda) N. for L. saxatilis by assembling public transcriptomic NGS lapillus (two hemocyanin paralogs: MT939254, MT939255) data (LisaH1, LisaH2). For P. canaliculata, four hemocy- and R. venosa (two hemocyanin paralogs: BK014286, anin cDNAs were published by Chiumiento et al. (2020). BK014287) and of the Cerithioidea species M. tuberculata As previously reported, these sequences contained some 1 3 Journal of Molecular Evolution (2021) 89:639–655 645 Table 2 FU-internal introns of caenogastropod hemocyanins FU Hc RtH1 NlH1 RtH2/NlH2 LisaH1 LisaH2 MtH400 MtH550 PcH I PcH IIb/ PcH III HpH HtH Sign 2 2 2 2 2 2 2 2 2 1 2 -a 5 5 5 5 5 6 7 4 3 5 3 -b 4 4 4 4 5 6 4 4 3 6 0 -c 2 3 2 2 3 5 5 2 1 6 0 -d 6 6 4 5 5 7 5 1 1 5 0 -e 5 5 5 5 6 6 6 5 3 5 0 -f 3 3 3 4 4 6 5 4 4 5 1 -g 4 4 4 4 4 7 – 2 1 5 2 -h 5 5 4 5 4 9 – 3 3 7 0 -f1 – – – – – – 2 – – – – -f2 – – – – – – 6 – – – – -f3 – – – – – – 4 – – – – -f4 – – – – – – 5 – – – – -f5 – – – – – – 3 – – – – -f6 – – – – – – 3 – – – – Σ(total, incl. sign.) 36 37 33 36 38 54 57 27 21 45 8 Σ(a–f) 25 26 23 25 28 36 32 20 15 32 4 Σ(FUs) 34 35 31 34 36 52 55 25 19 44 6 Ø Intron ~ 4.3 ~ 4.4 ~ 3.9 ~ 4.3 ~ 4.5 ~ 6.5 ~ 4.6 ~ 3.1 ~ 2.4 ~ 5.5 ~ 0.8 /FU Internal introns vary between hemocyanin genes of different Caenogastropoda groups and between different genes within the same caenogastro- pod species. These results contrast with the highly conserved exon–intron structures of Tectipleura (Heterobranchia) and Lepetellida (Vetigas- tropoda) represented in this table by the hemocyanins of Helix pomatia (HpH) and Haliotis tuberculata (HtH). The table shows the numbers of introns lying within the functional units of hemocyanins of R. venosa (RtH, two paralogs), N. lapillus (NlH, two paralogs), L. saxatilis (LisaH, two paralogs), M. tuberculata (MtH, two paralogs) and P. canaliculata (PcH, three paralogs). The numbers of internal introns are shown for the specific FUs and the signal peptide (sign.). Additionally, the table includes the sum of the internal introns of FU-a to FU-f (these FUs are included in all represented hemocyanins), as well as the total number of internal introns and the respective average number (Ø) of internal introns per functional unit in each hemocyanin inconsistencies (cf. Schäfer et al. 2021a). By assembling well-supported (Fig. 3, Supplement 1). These trees are transcriptomic NGS data, we identified and corrected three largely congruent with each other and share the same of those hemocyanin cDNA sequences (PcH I, IIb, III). This nodes, except for minor differences in one hemocyanin facilitated comparative phylogenetic analyses and helped to from L. saxatilis (LisaH2), which is grouped with hemo- elucidate the correct gene architectures (see below). cyanins from Muricidae in the Bayesian analysis. The The obtained hemocyanin sequences of L. saxatilis and obtained phylogenies of caenogastropod hemocyanins P. canaliculata include eight canonical FUs (a, b, …, h) are in accordance with the currently accepted systematic and comprise approximately 10,250–10,300 nucleotides, as relationships of the four groups Muricidae, Littorinida, is typical for gastropod hemocyanins (cf. Lieb et al. 2000, Cerithiida and Ampullariida (Fig. 2; Bouchet et al. 2017; 2004; Schäfer et al. 2018). Accession numbers, lengths and Ponder et al. 2019). The phylogenies obtained by maxi- molecular weights for all hemocyanin cDNA sequences and mum likelihood and MrBayes both reveal that the mul- primary structures of L. saxatilis and P. canaliculata are tiple hemocyanin genes from different analyzed species shown in Table 1. resulted from independent gene duplications that occurred after the diversification of Caenogastropoda into Ampul- Phylogenetic Analyses Reveal Multiple Independent lariida, Cerithiida and Hypsogastropoda (orange arrows Gene Duplications in Different Caenogastropods in Fig. 3). Although the position in the various phyloge- netic trees of LisaH2 is uncertain within the hemocyanins Molecular phylogenetic trees based on amino acid align- of Hypsogastropoda, the results of maximum likelihood ments of full-length protein sequences and inferred and Bayesian inferences both support multiple independ- by maximum likelihood and Bayesian analyses are ent duplications. The gene duplication that led to the two 1 3 646 Journal of Molecular Evolution (2021) 89:639–655 Caeno- Muricidae Hc 1 NlH1, RtH1 gastropoda Muricidae Hc 2 NlH2, RtH2 Liorinida Hc 1 LisaH1 Liorinida Hc 2 LisaH2 Cerithiida Hc400 kDa MtH400 Cerithiida Hc550 kDa MtH500 Ampullariida Hc 1 PcH I Ampullariida Hc 2+3 PcH IIb, PcH III Hetero- Tecpleura Hcs (LsH, AcH, HpH, CaH) branchia Ve - Lepetellida Hcs KLH,HdH, HrH, HtH gastropoda Cephalo- Octopodoidea Hcs OdH, OvH, ObH poda Naulus Hc NpH = linker intron, all in Phase 1 = internal intron = addional intron in only one ofin Phase 0, 1 and 2 the represented gene structures Fig. 4 Hemocyanin gene structures of Caenogastropoda, Het- 2: black). Smaller arrows with a star on top represent internal introns erobranchia, Vetigastropoda and Cephalopoda. The comparison that were obtained in only one hemocyanin from one structure type of exon–intron structures indicates that nine of the eleven analyzed (one intron within NlH1, Muricidae; and one within LsH1, Tecti- hemocyanin genes in Caenogastropoda vary in their gene structures pleura). The comparison includes all of the hemocyanins analyzed concerning the number and positions of internal introns. The cod- in this study from the following Caenogastropoda species: Nucella ing sequences of hemocyanins (large boxes: functional units; small lapillus (NlH1 + 2), Rapana venosa (RtH1 + 2), Littorina saxatilis and gray boxes: linker peptides) are shown, and intron positions (LisaH1 + 2), Melanoides tuberculata ( MtH400; MtH550 FU-a–FU-f; within their genes are represented by arrows (cf. alignment in Sup- exon–intron structures of the additional functional units f 1, f2, …, plement 2). To enable the comparability of gene structures, we did f6 of the mega-hemocyanin of M. tuberculata are shown in Supple- not enlarge the box representing the FU-g coding sequence of Muri- ment 3) and P. canaliculata (PcH I + IIb + III). Additionally, con- cidae Hc2 despite 118 and 340 additional amino acids being iden- served gene structures are included for hemocyanins of Tectipleura tified for RtH2 and NlH2 (cf. Supplement 2, Schäfer et  al. 2021a). (Lymnaea stagnalis LsH, Aplysia californica AcH, Helix pomatia Instead, we included a violet triangular box on top of the box rep- HpH and Cornu aspersum), Lepetellida (Megathura crenulata KLH, resenting the additional amino acids. Introns are divided into linker Haliotis diversicolor HdH, Haliotis rubra HrH, Haliotis tuberculata introns, which are highly conserved within all molluscan hemocya- HtH), Octopodoidea (Enteroctopus dofleini OdH, Octopus vulgaris nins (bold yellow arrows), and internal introns, which differ between OvH and O. bimaculoides ObH) and Nautilus pompilius NpH (Color various genes (thin arrows in phase 0: white; phase 1: gray or phase figure online) 1 3 Journal of Molecular Evolution (2021) 89:639–655 647 hemocyanin paralogs of R. venosa and N. lapillus most Supplement 2 + 3), which are characteristic of molluscan likely took place in a common ancestor of both Muricidae hemocyanins (Lieb et al. 2001). species but after separation of siphonate and asiphonate Hypsogastropoda (Fig. 3). Discussion High Variability of Hemocyanin Gene Architectures Similar to most molluscs, Caenogastropoda use hemocya- in Caenogastropoda nin as an oxygen transporter. This respiratory protein has previously been identified within the Muricidae species Characterization of the gene structures of hemocyanins R. venosa (Gebauer et al. 1999) and N. lapillus (Schäfer in five Caenogastropoda (each species has 2–3 hemocya- et al. 2021a), within the cerithioid M. tuberculata (Lieb nin paralogs; Table 2) revealed that nine of the eleven et al. 2010) and within the apple snail P. canaliculata hemocyanin genes differ in both the number and relative (Chiumiento et al. 2020). All of these Caenogastropoda intragene positions of internal introns (Fig. 4). Therefore, species possess at least two hemocyanin genes like other we obtained exon–intron architectures for hemocyanin Gastropoda (Gebauer et  al. 1999; Markl et  al. 1991; genes from different caenogastropod lineages (variations Schäfer et al. 2018). This gene system may enable differ- between Muricidae, Littorinida, Cerithioidea and Ampul- ential expression of several hemocyanin genes as shown lariida and between different hemocyanin paralogs within for Cephalopoda (Oellermann et al. 2015a, 2015b; Thonig the same species, e.g., between RtH1 and RtH2 of R. et al. 2014). Similar regulatory mechanisms may help Cae- venosa or LisaH1 and LisaH2 of L. saxatilis). Identical nogastropoda to adapt to different living conditions by sus- exon–intron structures are only present for hemocyanin taining oxygen supply despite changes in partial oxygen gene 2 of the two Muricidae species (RtH2 and NlH2; pressure and temperature, as hypothesized for Tectipleura RtH1 and NlH1 vary in one intron but differ strongly from (further discussed below under “Evolutionary constraints RtH2 and NlH2) and for the hemocyanin genes PcH IIb on hemocyanin genes in Apogastropoda?” and in Schäfer and PcH III of P. canaliculata. et al. 2018). by analyzing public transcriptomic NGS data, The total numbers of internal introns of caenogastro- we corrected inconsistencies within published hemocya- pod hemocyanins vary between 21 and 57 (Table 2). The nin cDNA sequences from P. canaliculata and obtained average numbers of internal introns per functional unit two hemocyanins from L. saxatilis, which belongs to the domain vary between 2.4 introns/FU (PcH IIb/III) and 6.5 asiphonate Hypsogastropoda and thus represent another introns/FU (MtH400). The comparison with known gas- lineage of the large group of Caenogastropoda (Fig. 2). tropod hemocyanin genes shows that all analyzed hemo- Finally, we conducted molecular phylogenies based on cyanins of Caenogastropoda contain a greater number of maximum likelihood (Fig. 3; Felsenstein 1981; Kumar internal introns than those of Vetigastropoda (Altenhein et al. 2016) and MrBayes (Supplement 1; Huelsenbeck et al. 2002; Lieb et al. 2001). Hemocyanins of Caenogas- and Ronquist 2001; Mau and Newton 1997) and analyzed tropoda contain fewer internal introns per functional unit the hemocyanin gene structures of the following species: than those of Tectipleura (cf. HpH in Table 2 and Fig. 4; R. venosa, N. lapillus, L. saxatilis, M. tuberculata and Schäfer et al. 2021b), with the exception of M tH400 of M. P. canaliculata (Fig. 4). Our analyses revealed ongoing tuberculata. Hemocyanin genes of this cerithioid encom- hemocyanin gene evolution within all major groups of pass the greatest number of internal introns that have been Caenogastropoda that were analyzed within this study: identified thus far. siphonate and asiphonate Hypsogastropoda (Muricidae Linker introns are highly conserved across all molluscan and Littorinida), Cerithiida and Ampullariida (Fig. 2). hemocyanins that have been analyzed thus far (Bergmann Since we analyzed only one or two species per group and et al. 2006; Lieb et al. 2001; Schäfer et al. 2021b) and are basal groups as Cyclophorida and Viviparida were not also present in hemocyanins of Caenogastropoda (Fig. 4). included in the study (Fig. 2), our analysis will not provide However, we did not detect an intron within the linker pep- a comprehensive overview of hemocyanin gene evolution tide-coding region between FU-f1 and FU-f2 of the mega- within Caenogastropoda. Nevertheless, by including spe- hemocyanin gene in M. tuberculata (star in Supplement 3). cies of various large groups of Caenogastropoda, our study Thus, the M tH550 gene is not only the first mega-hemocyanin gives a first insight into hemocyanin gene evolution within with a gene structure that has been analyzed but also the the diverse group of Caenogastropoda. Furthermore, the first hemocyanin gene that has been detected to lack a linker observed phenomenon is similar to that in their sister intron between two FU-coding regions. All other analyzed group Heterobranchia and thus will be discussed for the hemocyanin genes in Caenogastropoda include typical phase whole group of Apogastropoda. 1 linker introns between all FU-coding regions (Fig. 3, 1 3 648 Journal of Molecular Evolution (2021) 89:639–655 1 1 1 siphonate 5–8 (Muricidae: 1,863 species) 2 1 2 2 12 2 2 1 212 1 1 2 1 asiphonate 1–10 (Liorina saxalis) Hypsogastropoda 7–12 (in total: ~27,000 species) 12 2 1 21 1 2 12 121 121 2 12 21 21 1 12 Cerithioidea 19–33 (M. tuberculata) Sorbeoconcha (in total: ~29,000 species) 3–5 1 1 1 1 2 1 1 1 21 Ampullariida 3–11 (P. canaliculata) Caenogastropoda 12–13 (in total: ~33,000 species) Heterobranchia 42–43 (Tecpleura: 26,000 species) Apogastropoda 4 8 Vegastropoda(Lepetellida) Gastropoda = linker = internal intron = presence in ancestor cannot be 1 2 = intron only presenn hc gene = intron in only one hc gene of one intron in phase 0/1/2 decided on parsimony principle structure 1 or 2 o he species species o he represented group Fig. 5 Maximum parsimony scenario of exon–intron architecture we indicated the range between the possible minimum and maximum. evolution within gastropod hemocyanins. The phylogenetic tree on We were unable to exactly assess the origin of every intron because the left shows the relation of gastropod species with known hemo- two independent intron gains within two descendant species represent cyanin gene architectures (Ponder et al. 2019). For groups of Apogas- the same number of evolutionary events as one gain within a com- tropoda, the number of included species according to the WoRMS mon ancestor together with an intron loss within one gene of one of Editorial Board (2020) is also shown. If gene structures are known the descendants. We highlighted these cases with smaller arrows in for only one representative of a lineage, the specific species is given. the ancestor genes. In contrast to the hemocyanin gene structures in On the right, hypothetical gene structures are shown for the common Tectipleura, those in Caenogastropoda vary not only between differ- ancestors. These structures are based on the maximum parsimony ent caenogastropod lineages but also within paralogous genes within principle, assuming that introns that are at the same position concern- the same species. We symbolized such introns that are present in the ing the coding sequences of hemocyanin genes of sister groups were gene structure of only one hemocyanin gene of the same species with already present in hemocyanin genes of a common ancestor. The small arrows with the number of the gene in which it is located on same applies for intron losses if both descendants do not include a top. Arrows with a star on top represent deviations from a conserved formerly present intron. The deduced model of gene structure evolu- group-specific hemocyanin gene structure [e.g., additional Hygroph- tion within gastropods shows a gradual gain of introns during the evo- ila-specific intron, see Fig. 4 and cf. Schäfer et al. (2018)]. As in all lution of Caenogastropoda. The level of intron gains varies between other molluscan hemocyanins, linker introns are located at the same the different hemocyanin lineages and is numbered on the left side of position within the sequences coding for linker peptides between all the taxonomic tree. If the exact number of intron gains is uncertain, canonical functional units Multiple Independent Hemocyanin Gene gene duplications in Tectipleura (Heterobranchia), which Duplications: A Phenomenon of Apogastropoda took place independently in different groups (e.g., Stylom- matophora, Hygrophila, Anaspidea; Schäfer et al. 2018) Both, maximum likelihood analysis and Bayesian phy- and thus may be a general phenomenon for Apogastropoda. logenetic inferences (Fig. 3, Supplement 1) revealed that These multiple independent gene duplications contrast with multiple hemocyanin paralogs identified for P. canalicu- the much more conserved hemocyanin genes 1 and 2 of lata, M. tuberculata, L. saxatilis and the two Muricidae Lepetellida analyzed within H. tuberculata, Haliotis diver- species R. venosa and N. lapillus resulted from independ- sicolor and Megathura crenulata (Lieb and Markl 2004; Yao ent gene duplications. This result is similar to the multiple et al. 2019). These paralogous hemocyanin genes resulted 1 3 Journal of Molecular Evolution (2021) 89:639–655 649 from one gene duplication that took place before the split of To analyze the evolution of hemocyanin gene structures Lepetellida into Haliotoidea and Fissurelloidea ~ 343 million of Gastropoda in more detail, we derived the most parsi- years ago (Lieb and Markl 2004) and has been maintained in monious scenario of intron evolution within these genes both lineages. Thus, the comparison of hemocyanin genes of (Fig. 5). This scenario is based on the parsimony principle Apogastropoda and Vetigastropoda indicates strongly differ- (Rogozin et al. 2006), which assumes that an intron shared ent evolutionary dynamics of hemocyanin gene duplications within a sister group was already present within their com- during the evolution of these gastropod lineages. mon ancestor. The same applies to missing introns detected within sister groups that are assumed to be lost in an ances- Extensive Variability of Hemocyanin Gene tor. Intron sliding was not considered because this phenom- Structures of Caenogastropoda Suggests a Higher enon is difficult to identify based on location alone if introns Continuous Intron Gain Rate than that Identified vary by more than one nucleotide (Rogozin et al. 2000). in any other Molluscs Positions of nearby FU-internal introns within the analyzed hemocyanin genes of Caenogastropoda vary by at least six In addition to multiple independent gene duplications, our nucleotides (Supplement 2). Furthermore, intron sliding results on exon–intron architectures also reveal a high rate most likely contributes little to gene structure evolution of evolutionary changes in hemocyanin genes in Caenogas- (Poverennaya et al. 2020; Stoltzfus et al. 1997). It should tropoda. We have previously reported that hemocyanin genes be noted that the presence of some introns within an ances- of Tectipleura encompass a significantly larger number of tor cannot be assessed because the two possible scenarios FU-internal introns (on average 5.6 introns per FU; Schäfer would have taken the same number of evolutionary steps et al. 2021b) than hemocyanin genes of Vetigastropoda or (smaller arrows in a hypothetical gene precursor, Fig. 5). Cephalopoda (≤ 0.8 introns per FU). This study also iden- Nevertheless, this model of intron evolution shows the most tified a larger number of internal introns in hemocyanin parsimonious explanation for the revealed exon–intron genes of Caenogastropoda (2.4–6.5 introns per FU), the structures and clearly indicates that ongoing changes within other branch of Apogastropoda (Fig. 4; Table 2). As typical hemocyanin gene structures during the evolution of differ- for internal introns of molluscan hemocyanins, they vary in ent Caenogastropoda are most likely. This scenario strongly phases. Across all hemocyanin gene structures of Caeno- suggests a gradual accumulation of introns, which led to gastropoda that we have analyzed, phase 0 was the most fre- gene structures with several internal introns, regardless of quent intron phase (≥ 50%). This result matches the results the exact origins of particular introns. Thus, our findings of Long & Deutsch (1999) and Fedorov et al. (1992) as well support the hypothesis that the accumulation of introns is a as the results for Tectipleura hemocyanins (Schäfer et al. general phenomenon within hemocyanin genes of Apogas- 2021b). Intron phases affect which exons may be spliced tropoda and contrasts with the hemocyanin gene evolution alternatively. Since we did not detect any splice variants, we of Vetigastropoda and Cephalopoda (Chiumiento et al. 2020; did not focus on analyzing these intron phases in more detail. Schäfer et al. 2021b). However, the color-coded phases in Fig. 4 accentuate the dif- Considered more closely, the presented model of ferences, which will be described below. The distinct hemo- intron evolution reveals four internal introns that are pre- cyanin genes of Caenogastropoda differ strongly from each sent within all analyzed species of Tectipleura (Hetero- other in terms of the number and position of internal introns. branchia) and Caenogastropoda and thus within both major This phenomenon applies to hemocyanins of different cae- groups of Apogastropoda. Therefore, intron accumulation nogastropod lineages (Muricidae, Littorinida, Cerithioidea most likely started within a hemocyanin gene of a com- and Ampullariida) as well as to paralogous genes within mon ancestor, and ongoing accumulation subsequently the same species (Fig. 4, Table 2). Large variations in gene led to various gene structures within different Apogas- architectures between paralogous hemocyanin genes within tropoda lineages. The model of intron evolution shows, for one molluscan species have not been reported before. To example, that 12 or 13 FU-internal introns and two introns date, only for Hygrophila, a group of Tectipleura, have two within the signal peptide-coding sequences are likely to paralogous hemocyanin genes been identified that vary in have evolved within a common ancestor of the analyzed one of 45 or 46 introns. Specifically, hemocyanin gene 1 Caenogastropoda groups and that accumulation most from Lymnaea stagnalis, Radix balthica and Biomphalaria likely continued gradually throughout the evolution of glabrata (LsH1, RbH1, and BgHcl-2, which are all ortholo- Sorbeoconcha, Hypsogastropoda and Muricidae (Fig. 5). gous to each other) has one additional intron along with We have previously shown that the high intron gain rate those conserved within all other analyzed Tectipleura hemo- is specific to the hemocyanin gene within Apogastropoda cyanins, including hemocyanin gene 2 of those Hygrophila (Schäfer et al. 2021b). Such lineage- and gene-specific evo- species (Fig. 4, arrow with star and Schäfer et al. 2021b). lutionary rates have also been described for other genes (Carmel et al. 2007; Rogozin et al. 2003). A large number 1 3 6 50 Journal of Molecular Evolution (2021) 89:639–655 of introns may be accompanied by regulatory advantages, constraint acts on hemocyanin genes in Caenogastropoda, which may cause this lineage- and gene-specific trend of this might explain the continuous accumulation of introns intron accumulation in hemocyanins in Apogastropoda in these genes during the evolution of Caenogastropoda, as (Schäfer et al. 2021b). proposed in the maximum parsimony scenario (Fig. 5). The The extensive variation of exon–intron structures of hemocyanin gene architectures suggested for the common caenogastropod hemocyanin genes reported in this study ancestors of Caenogastropoda, Sorbeoconcha or Hypsogas- contrasts with the highly conserved exon–intron structures tropoda include fewer introns than the gene structures of that have been detected in hemocyanins in Tectipleura Tectipleura hemocyanins (Fig. 5). Therefore, such hemo- (Heterobranchia). Hemocyanin gene architectures in Tec- cyanin genes comprise relatively large exons which may tipleura evolved most likely within a common ancestor and represent possible targets for intron insertion (Hawkins thus before the multiple gene duplications that occurred in 1988; Hwang and Cohen 1997; Roy and Irimia 2009). Due several Tectipleura groups. Subsequently, the exon–intron to lineage- and gene-specific intron gain/loss rates that may structure has remained mostly conserved within all ana- result from evolutionary constraints, the gene architectures lyzed species of this large group of gastropods (Schäfer et al. of many hemocyanin genes may have accumulated introns 2021b). Hemocyanin gene structures of Caenogastropoda, in during the evolution of Caenogastropoda until saturation is contrast, changed independently within the different lineages reached. This may be the reason for the continuous intron and within different paralogous genes. gains in hemocyanins in Caenogastropoda which contrasts Since the number of Tectipleura species corresponds with the conserved gene structure of Tectipleura hemocya- to ~ 80% of the number of Caenogastropoda species and nins. Although the rate of intron gains varies, intron accu- even ~ 96% of the number of Hypsogastropoda species mulation seems to be common within hemocyanins of both (Fig. 5; WoRMS Editorial Board 2020), the higher num- groups of Apogastropoda and may be caused by evolution- ber of differences within hemocyanin genes in Caenogas- ary constraints. tropoda cannot be explained by the relatively higher num- ber of species. Rather, we suggest that the intron gain rate Evolutionary Constraints on Hemocyanin Genes decreased within Tectipleura quite early in evolution because in Apogastropoda? their hemocyanin genes became saturated by introns more quickly than those of Caenogastropoda. Therefore, the Our results on the large number of gene duplications and large number of introns led to relatively small exon lengths, the various intron gains identified for hemocyanin genes in which may prevent further intron gains (Hawkins 1988; Caenogastropoda support the hypothesis of high dynam- Hwang and Cohen 1997; Roy and Irimia 2009). Gene- and ics within hemocyanin gene evolution in Apogastropoda lineage-specific saturation of intron densities have already (Schäfer et al. 2018, 2021b). Both Caenogastropoda and Het- been described for plants (Basu et al. 2008) and mammals erobranchia represent the most diverse groups of the phylum (Kordiš 2011). According to the most parsimonious scenario Mollusca and together encompass over 64,000 extant species of intron evolution in gastropod hemocyanins (Fig. 5), 42 (WoRMS Editorial Board 2020) that live in nearly all kinds introns were gained during the evolution from an Apogas- of habitats from the deep sea to deserts. The radiation that tropoda ancestor to a precursor of Tectipleura and may have led to the high diversity of Apogastropoda involved a great saturated the genes. This number is strikingly higher than number of adaptations, including the evolution of altered that for intron gains during the evolution from an Apogas- abilities for osmoregulation, novel respiratory organs (pneu- tropoda ancestor to the ancestors of Caenogastropoda mostomes and lungs) and reproductive behavioral strategies (12–13 gains), Sorbeoconcha (15–18 gains) or Hypsogas- (Vermeij and Dudley 2000; Vermeij and Wesselingh 2002). tropoda (22–30 gains, Fig. 5). Schäfer et al. (2021b) identi- As we have proposed previously, high evolutionary rates fied conservation in the exon lengths within hemocyanin within hemocyanin genes may represent molecular adapta- genes in Tectipleura. This phenomenon has been suggested tions that have enabled multiple habitat shifts and species to indicate evolutionary advantages (Davila-Velderrain diversification within Apogastropoda (Schäfer et al. 2018, et al. 2014; Fu and Lin 2012). Accordingly, our results 2021b). This hypothesis is developed from consideration of showing conserved exon–intron structures may imply that the strong temperature dependence of the oxygen affinity of the large number of introns and the regular distribution of hemocyanins (Brix et al. 1989; Burnett et al. 1988; Mangum exons within hemocyanin genes of Tectipleura may have 1990). Various adaptations of this oxygen transport protein evolutionary benefits, e.g., expanded possibilities of gene to different environmental conditions have been reported expression regulation (Schäfer et al. 2021b). More elaborate (González et al. 2017; Melzner et al. 2007; Oellermann et al. explanations of possible evolutionary advantages are further 2015a, b; Strobel et al. 2012; Yesilyurt et al. 2008; Zielinski discussed below under “Evolutionary constraints on hemo- et al. 2001). These adaptations appear necessary to ensure cyanin genes in Apogastropoda?”. If a similar evolutionary a sufficient supply of oxygen and therefore are fundamental 1 3 Journal of Molecular Evolution (2021) 89:639–655 651 for molluscs to survive. Consequently, they represent one Carmel 2012; Rose 2008, 2018) and include, among others, essential precondition for habitat shifts between strongly dif- temperature-dependent gene expression (Airoldi et al. 2015; ferent environments (e.g., from sea to land or freshwater). Evantal et al. 2018; Gotic et al. 2016; James et al. 2018). Strong variability, as previously identified for hemocyanin Differential expression of hemocyanin genes could help to genes in Tectipleura and now verified for five groups of Cae- control the availability of different paralogs that harbor dif- nogastropoda, may accommodate these required adaptations ferent properties (e.g., varying oxygen affinities at differ- (Schäfer et al. 2018, 2021b). ent temperatures). Thus, the regulatory functions of introns Gene duplications, as we have identified for hemocyanins could represent evolutionary advantages that promote intron in Tectipleura and Caenogastropoda (Fig. 3), play a major accumulation in hemocyanin genes of Apogastropoda (for role in genomic complexity and evolution (Magadum et al. more details see Schäfer et al. 2021b). To analyze potential 2013; Ohno 1970). They are a driving force in organismal advantages of introns in hemocyanin genes, further studies diversity (Lynch and Conery 2000) and can promote adapta- should characterize their nucleotide sequences, investigate tion (Qian and Zhang 2014). Hemocyanin gene duplications them for specific regulatory mechanisms and determine if could be followed by differential evolution of various genes there are differences in introns between paralogs within one and could eventually lead to hemocyanins with, for exam- species. ple, different oxygen affinities, varying pH or temperature Our results show that the accumulation rate of introns sensitivities or differential expression patterns. These dif- maintained in hemocyanins in Caenogastropoda is differ- ferences may represent the origin of genetic variability and ent within various lineages and highest within Cerithioidea adaptation, as has already been discovered for Cephalopoda (Fig. 5). In addition, the two paralogous hemocyanin genes (Oellermann et al. 2015a, b; Strobel et al. 2012; Thonig from the cerithioid gastropod M. tuberculata exhibit the et al. 2014). The squid S. officinalis, for example, possesses largest variations in exon–intron structures that have been multiple hemocyanin genes that underlie differential expres- found within one gastropod species (small arrows with num- sion (Strobel et al. 2012; Thonig et al. 2014). Thonig et al. bers on top in Fig. 5, cf. also Supplement 3). Furthermore, (2014) identified ontogeny-specific expression patterns the 550 kDa mega-hemocyanin subunit represents the only of hemocyanin genes in this squid species. For example, hemocyanin gene that lacks a linker intron between two dif- one hemocyanin gene is highly expressed within the egg ferent functional units. The loss of regions coding for FU-g and may help to sustain an adequate oxygen supply despite and FU-h, as well as the gain of six FU-f-coding sequence hypoxic and hypercapnic conditions arising within the eggs sections, additionally led to strong variations in the protein during embryogenesis. After hatching, however, the expres- structures which influence its physiological features (typi- sion of this gene is downregulated, whereas the expression cal 400 kDa hemocyanin and 550 kDa mega-hemocyanin of two other hemocyanin genes is upregulated (Thonig et al. subunits that differ in their oxygen binding capacity, cf. 2014). To examine whether the multitude of hemocyanin introduction above and Lieb et al. 2010). These extensive genes in Apogastropoda contributes to adaptive radiation differences at the gene and protein levels may represent and evolutionary benefits or if it has even led to neofunc- combined adaptations of hemocyanins that enable Cerith- tionalization of these proteins, further studies are needed ioidea to live in a variety of habitats. The members of this to examine the biochemical properties and physiological superfamily of Caenogastropoda are distributed in freshwa- implications of paralogous hemocyanins in Heterobranchia ter, brackish water and marine ecosystems in mostly warm and Caenogastropoda. temperate regions (for an overview, see Strong et al. 2011). In addition to gene duplications, our findings confirm Their habitats include some extreme biotopes, such as rocky that intron accumulation is a general phenomenon of hemo- intertidal shores, mudflats and mangrove forests, which cyanin genes in Apogastropoda (Chiumiento et al. 2020; include environmental conditions such as strongly changing Schäfer et  al. 2021b). As already proposed, this result temperatures, humidity differences and little oxygen avail- may indicate evolutionary constraints on a large number ability. The fact that the strongest differences in paralogous of introns in gene structures of Apogastropoda. Introns, in hemocyanins (including the gene and protein levels) have general, can increase species diversity (Calarco and Ellis been identified in a group of gastropods with such extremely 2020) and thus may also contribute to adaptation. The exten- diverse habitats may represent a further hint that the evolu- sive number of internal introns found within hemocyanin tion of hemocyanin plays a major role in the evolution of genes of Apogastropoda may facilitate the regulation of molluscs. Since both genes can be expressed in different differential expression (discussed in more detail in Schäfer ratios (Lieb et al. 2010), they can help to adapt to different et al. 2021b). Several regulatory mechanisms associated living conditions. with introns have already been identified (e.g., Chorev and 1 3 6 52 Journal of Molecular Evolution (2021) 89:639–655 Conclusions Data Availability The cDNA sequences and gene structures obtained during the current study are available at the NCBI under the follow- ing accession numbers: LisaH1: BK014376; LisaH2: BK014375; The oxygen affinity of molluscan hemocyanins is strongly PcH I: BK014379; PcH IIb: BK014378; PcH III: BK014377; NlH1: influenced by abiotic factors such as temperature (Mangum MT939254; NlH2: MT939255; RtH1: BK014286; RtH2: BK014287; 1990). A multitude of adaptations in these oxygen trans- MtH400: KC405575; MtH550: KC405576. port proteins have already been described (e.g., Strobel Code Availability Not applicable. et al. 2012) and seem to be indispensable for many mol- luscan species to ensure a sufficient oxygen supply. Our Ethical Approval Not applicable. findings reveal a strong diversity of hemocyanin genes of Caenogastropoda, including multiple independent gene Consent to Participate Not applicable. duplications within different caenogastropod groups Consent for Publication Not applicable. as well as a strong accumulation of FU-internal introns (21–57) within their genes. Since gene duplications and Open Access This article is licensed under a Creative Commons Attri- intron gains have also been discovered within hemocyanin bution 4.0 International License, which permits use, sharing, adapta- genes of Tectipleura (Schäfer et al. 2018, 2021b), they tion, distribution and reproduction in any medium or format, as long most likely represent general phenomena of hemocyanin as you give appropriate credit to the original author(s) and the source, gene evolution within Apogastropoda. This contrasts with provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are the hemocyanin evolution of Vetigastropoda and may sup- included in the article's Creative Commons licence, unless indicated port the hypothesis that diversity in this oxygen transporter otherwise in a credit line to the material. If material is not included in may increase adaptation. Therefore, gene duplications may the article's Creative Commons licence and your intended use is not provide new genes to be adjusted by mutation and selec- permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a tion, and the accumulation of introns may contribute to copy of this licence, visit http://c reati vecom mons.o rg/l icens es/b y/4.0 /. regulatory opportunities. Although follow-up studies are required to analyze the biochemical and physiological properties of apogastropod hemocyanins (e.g., varying oxygen affinities and different temperature dependencies References of hemocyanin paralogs and differential expression pat- terns of their paralogous genes), our results indicate that Airoldi CA, McKay M, Davies B (2015) MAF2 Is regulated by tem- the evolution of hemocyanins in Heterobranchia and Cae- perature-dependent splicing and represses flowering at low tem- nogastropoda may be one of many influencing factors that peratures in parallel with FLM. PLoS ONE 10:e0126516. https:// doi. org/ 10. 1371/ journa l. pone. 01265 16 enabled the large diversity of Apogastropoda by facilitat- Altenhein B, Lieb B, Awenius C, Markl J (2000) Gene structure of ing adaptive radiation. gastropod hemocyanin. Zoology Suppl. III. In: Proceedings of the 93th annual meeting Bonn, Germany Supplementary Information The online version contains supplemen- Altenhein B, Markl J, Lieb B (2002) Gene structure and hemocya- tary material available at https://d oi.o rg/1 0.1 007/s 00239-0 21-1 0036-y. nin isoform HtH2 from the mollusc Haliotis tuberculata indicate early and late intron hot spots. Gene 301:53–60. https:// doi. org/ 10. 1016/ S0378- 1119(02) 01081-8 Author Contributions GGS was a major contributor to the conception Basu MK, Rogozin IB, Deusch O, Dagan T, Martin W, Koonin EV of the analyses. She prepared genomic DNA and RNA for NGS. Addi- (2008) Evolutionary dynamics of introns in plastid-derived genes tionally, she analyzed and interpreted nucleotide sequences and gene in plants: saturation nearly reached but slow intron gain contin- structures in cooperation with LJG and RS who conducted large parts ues. Mol Biol Evol 25:111–119. https:// doi. org/ 10.1 093/m olbev/ of the NGS analyses of MtH, RtH and LisaH. BL conceived the project msm234 and supervised this study as the principal investigator. GGS wrote the Bergmann S, Lieb B, Ruth P, Markl J (2006) The hemocyanin from a first version of the manuscript which was commented on, revised and living fossil, the cephalopod Nautilus pompilius: protein structure, approved by all authors. gene organization, and evolution. J Mol Evol 62:362–374. https:// doi.o rg/ 10. 1007/s 00239- 005- 0160-x Funding Open Access funding enabled and organized by Projekt Bouchet P, Rocroi J-P, Hausdorf B, Kaim A, Kano Y, Nützel A, DEAL. This study was supported by a grant from the German Research Parkhaev P, Schrödl M, Strong EE (2017) Revised classification, Foundation (DFG) to B. Lieb, project number LI998/15-1. Additional nomenclator and typification of gastropod and monoplacophoran support came from the Wagner and the Feldbausch Foundation (both families. Malacologia 61:1–526. https://d oi.o rg/1 0.4 002/0 40.0 61. to B. Lieb, University of Mainz). 0201 Brix O, Bårdgard A, Cau A, Colosimo A, Condò SG, Giardina B Declarations (1989) Oxygen-binding properties of cephalopod blood with special reference to environmental temperatures and ecological distribution. J Exp Zool 252:34–42. https:// doi. org/ 10. 1002/ jez. Conflict of interest The authors declare that they have no competing 14025 20106 interests. Brix O, Colosimo A, Giardina B (1995) Temperature dependence of oxygen binding to cephalopod haemocyanins: ecological 1 3 Journal of Molecular Evolution (2021) 89:639–655 653 implications. Mar Freshw Behav Physiol 25:149–162. https:// Gotic I, Omidi S, Fleury-Olela F, Molina N, Naef F, Schibler U (2016) doi. org/ 10. 1080/ 10236 249409 37891 4 Temperature regulates splicing efficiency of the cold-inducible Burnett LE, Scholnick DA, Mangum CP (1988) Temperature sensitiv- RNA-binding protein gene Cirbp. Genes Dev 30:2005–2017. ity of molluscan and arthropod hemocyanins. Biol Bull 174:153– https:// doi. org/1 0.1 101/g ad.2 87094.1 16 162. https:// doi. org/ 10. 2307/ 15417 82 Gutowska MA, Melzner F (2009) Abiotic conditions in cephalopod Calarco L, Ellis J (2020) Contribution of introns to the species (Sepia officinalis) eggs: embryonic development at low pH and diversity associated with the apicomplexan parasite, Neospora high pCO2. Mar Biol 156:515–519. https:// doi. org/ 10. 1007/ caninum. Parasitol Res 119:431–445. https:// doi. org/ 10. 1007/ s00227- 008- 1096-7 s00436- 019- 06561-x Hawkins JD (1988) A survey on intron and exon lengths. Nucleic Acids Carmel L, Rogozin IB, Wolf YI, Koonin EV (2007) Evolutionarily Res 16:9893–9908 conserved genes preferentially accumulate introns. Genome Res Huelsenbeck JP, Ronquist F (2001) MRBAYES: Bayesian inference of 17:1045–1050. https://d oi.o rg/1 0.1 101/g r. 59782 07 phylogenetic trees. Bioinformatics 17:754–755 Chiumiento IR, Ituarte S, Sun J, Qiu JW, Heras H, Dreon MS (2020) Hwang DY, Cohen JB (1997) U1 small nuclear RNA-promoted exon Hemocyanin of the caenogastropod Pomacea canaliculata exhib- selection requires a minimal distance between the position of its evolutionary differences among gastropod clades. PLoS ONE U1 binding and the 3’ splice site across the exon. Mol Cell Biol 15:e0228325. https:// doi. org/ 10. 1371/ journ al.p one. 022832 5 17:7099–7107. https:// doi.o rg/ 10.1 128/ MCB. 17.1 2. 7099 Chorev M, Carmel L (2012) The function of introns. Front Genet 3:55. James AB, Calixto CPG, Tzioutziou NA, Guo W, Zhang R, Simpson https://d oi.o rg/ 10. 3389/ fgene. 2012.0 0055 CG, Jiang W, Nimmo GA, Brown JWS, Nimmo HG (2018) Davila-Velderrain J, Servin-Marquez A, Alvarez-Buylla ER (2014) How does temperature affect splicing events? Isoform switching Molecular evolution constraints in the floral organ specification of splicing factors regulates splicing of LATE ELONGATED gene regulatory network module across 18 angiosperm genomes. HYPOCOTYL (LHY). Plant Cell Environ 41:1539–1550. Mol Biol Evol 31:560–573. https:// doi. org/ 10. 1093/ molbev/ https://d oi. org/ 10.1 111/ pce. 13193 mst223 Jörger KM, Stöger I, Kano Y, Fukuda H, Knebelsberger T, Schrödl Dayrat B, Tillier S (2002) Evolutionary relationships of euthyneuran M (2010) On the origin of Acochlidia and other enigmatic gastropods (Mollusca): a cladistic re-evaluation of morphologi- euthyneuran gastropods, with implications for the systematics cal characters. Zool J Linn Soc 135:403–470. https://d oi. org/1 0. of Heterobranchia. BMC Evol Biol 10:323. https:// doi. org/ 10. 1046/j. 1096- 3642. 2002.0 0018.x 1186/ 1471- 2148- 10- 323 Dinapoli A, Klussmann-Kolb A (2010) The long way to diversity— Kano Y, Brenzinger B, Nützel A, Wilson NG, Schrödl M (2016) phylogeny and evolution of the Heterobranchia (Mollusca: Gas- Ringiculid bubble snails recovered as the sister group to sea tropoda). Mol Phylogenet Evol 55:60–76. https:// doi. org/ 10. slugs (Nudipleura). Sci Rep 6:30908. https:// doi. org/ 10. 1038/ 1016/j.y mpev.2 009. 09. 019 srep30 908 Edgar RC (2004) MUSCLE: multiple sequence alignment with high Kato S, Matsui T, Gatsogiannis C, Tanaka Y (2018) Molluscan hemo- accuracy and high throughput. Nucleic Acids Res 32:1792–1797. cyanin: structure, evolution, and physiology. Biophys Rev 10:191– https:// doi. org/1 0. 1093/n ar/ gkh340 202. https:// doi.o rg/1 0. 1007/s 12551- 017- 0349-4 Evantal N, Anduaga AM, Bartok O, Patop IL, Weiss R, Kadener S Kearse M, Moir R, Wilson A, Stones-Havas S, Cheung M, Sturrock S, (2018) Thermosensitive alternative splicing senses and mediates Buxton S, Cooper A, Markowitz S, Duran C, Thierer T, Ashton temperature adaptation in Drosophila. bioRxiv. https://d oi.o rg/1 0. B, Meintjes P, Drummond A (2012) Geneious basic: an integrated 7554/ eLife. 44642 and extendable desktop software platform for the organization and Fedorov A, Suboch G, Bujakov M, Fedorova L (1992) Analysis of analysis of sequence data. Bioinformatics 28:1647–1649. https:// nonuniformity in intron phase distribution. Nucleic Acids Res doi.o rg/ 10.1 093/ bioin forma tics/b ts199 20:2553–2557. https:// doi. org/ 10. 1093/ nar/2 0.1 0.2 553 Kocot KM, Halanych KM, Krug PJ (2013) Phylogenomics supports Felsenstein J (1981) Evolutionary trees from DNA sequences: a maxi- panpulmonata: opisthobranch paraphyly and key evolutionary mum likelihood approach. J Mol Evol 17:368–376. https:// doi. steps in a major radiation of gastropod molluscs. Mol Phylogenet org/ 10. 1007/ BF0173 4359 Evol 69:764–771. https:// doi.o rg/ 10. 1016/j. ympev. 2013. 07. 001 Fu GC-L, Lin W-C (2012) Identification of gene-oriented exon orthol- Kordiš D (2011) Extensive intron gain in the ancestor of pla- ogy between human and mouse. BMC Genom 13(Suppl 1):S10. cental mammals. Biol Direct 6:59. https:// doi. org/ 10. 1186/ https:// doi.o rg/1 0. 1186/1 471- 2164- 13- S1- S10 1745- 6150-6-5 9 Gatsogiannis C, Markl J (2009) Keyhole limpet hemocyanin: 9-A Kumar S, Stecher G, Tamura K (2016) MEGA7: molecular evolution- CryoEM structure and molecular model of the KLH1 didecamer ary genetics analysis version 7.0 for bigger datasets. Mol Biol reveal the interfaces and intricate topology of the 160 functional Evol 33:1870–1874. https:// doi. org/1 0. 1093/ molbev/ msw054 units. J Mol Biol 385:963–983. https:// doi. org/ 10. 1016/j. jmb. Lieb B, Markl J (2004) Evolution of molluscan hemocyanins as 2008.1 0. 080 deduced from DNA sequencing. Micron 35:117–119. https://d oi. Gatsogiannis C, Hofnagel O, Markl J, Raunser S (2015) Structure of org/ 10.1 016/j. micron. 2003. 10.0 35 mega-hemocyanin reveals protein origami in snails. Structure Lieb B, Altenhein B, Markl J (2000) The sequence of a gastropod 23:93–103. https:// doi.o rg/ 10. 1016/j. str. 2014. 10.0 13 hemocyanin (HtH1 from Haliotis tuberculata). J Biol Chem Gebauer W, Stoeva S, Voelter W, Dainese E, Salvato B, Beltramini M, 275:5675–5681. https:// doi.o rg/ 10. 1074/j bc. 275.8.5 675 Markl J (1999) Hemocyanin subunit organization of the gastro- Lieb B, Altenhein B, Markl J, Vincent A, van Olden E, van Holde pod Rapana thomasiana. Arch Biochem Biophys 372:128–134. KE, Miller KI (2001) Structures of two molluscan hemocyanin https:// doi. org/ 10.1 006/ abbi. 1999. 1478 genes: significance for gene evolution. Proc Natl Acad Sci USA González A, Nova E, Del Campo M, Manubens A, de Ioannes A, 98:4546–4551. https:// doi.o rg/ 10.1 073/p nas.0 7104 9998 Ferreira J, Becker MI (2017) The oxygen-binding properties of Lieb B, Boisguerin V, Gebauer W, Markl J (2004) cDNA sequence, hemocyanin from the mollusk Concholepas concholepas. Biochim protein structure, and evolution of the single hemocyanin from Biophys Acta Proteins Proteom 1865:1746–1757. https://d oi.o rg/ Aplysia californica, an opisthobranch gastropod. J Mol Evol 10.1 016/j. bbapap. 2017.0 8. 017 59:536–545. https://d oi.o rg/ 10. 1007/ s00239- 004-2 646-3 Lieb B, Gebauer W, Gatsogiannis C, Depoix F, Hellmann N, Harasew- ych MG, Strong EE, Markl J (2010) Molluscan mega-hemocyanin: 1 3 654 Journal of Molecular Evolution (2021) 89:639–655 an ancient oxygen carrier tuned by a ~550 kDa polypeptide. Front supports independent terrestrial invasions. Mol Phylogenet Evol Zool 7:14. https:// doi. org/ 10. 1186/ 1742- 9994-7- 14 97:43–54. https://d oi. org/ 10. 1016/j. ympev. 2015. 12.0 14 Long M, Deutsch M (1999) Association of intron phases with con- Romero PE, Weigand AM, Pfenninger M (2016b) Positive selection servation at splice site sequences and evolution of spliceosomal on panpulmonate mitogenomes provide new clues on adaptations introns. Mol Biol Evol 16:1528–1534. https:// doi. org/ 10. 1093/ to terrestrial life. BMC Evol Biol 16:164. https://d oi.o rg/1 0.1 186/ oxford jour nals. molbev.a 0260 65 s12862- 016- 0735-8 Lynch M, Conery JS (2000) The evolutionary fate and consequences Rose AB (2008) Intron-mediated regulation of gene expression. Curr of duplicate genes. Science 290:1151–1155. https:// doi. org/ 10. Top Microbiol Immunol 326:277–290. https:// doi. org/ 10. 1007/ 1126/s cienc e.2 90. 5494. 1151 978-3- 540-7 6776-3_1 5 Magadum S, Banerjee U, Murugan P, Gangapur D, Ravikesavan R Rose AB (2018) Introns as gene regulators: a brick on the accelera- (2013) Gene duplication as a major force in evolution. J Genet tor. Front Genet 9:672. https://d oi.o rg/1 0.3 389/f gene.2 018.0 0672 92:155–161. https:// doi.o rg/ 10. 1007/ s12041-0 13- 0212-8 Roy SW, Irimia M (2009) Mystery of intron gain: new data and new Mangum CP (1990) Gas transport in the blood. In: Gilbert DL, Adel- models. Trends Genet 25:67–73. https:// doi. org/ 10. 1016/j. tig. man WJ, Arnold JM (eds) Squid as experimental animals, vol 211. 2008. 11.0 04 Springer, Boston, pp 443–468 Schäfer GG, Pedrini-Martha V, Schnegg R, Dallinger R, Jackson DJ, Markl J (2013) Evolution of molluscan hemocyanin structures. Bio- Lieb B (2018) Hemocyanin genes as indicators of habitat shifts chim Biophys Acta 1834:1840–1852. https:// doi. org/ 10. 1016/j. in Panpulmonata? Mol Phylogenet Evol 130:99–103. https:// doi. bbapap. 2013.0 2.0 20 org/ 10.1 016/j.y mpev. 2018.1 0.0 14 Markl J, Savel-Niemann A, Wegener-Strake A, Sding M, Schneider Schäfer GG, Grebe LJ, Depoix F, Lieb B (2021a) Hemocyanins of A, Gebauer W, Harris JR (1991) The role of two distinct subunit muricidae: new “insights” unravel an additional highly hydro- types in the architecture of keyhole limpet hemocyanin (KLH). philic 800 kDa mass within the molecule. J Mol Evol. https://d oi. Naturwissenschaften 78:512–514. https://d oi.o rg/1 0.1 007/B F011 org/1 0.1 007/s 00239- 020- 09986-6 31401 Schäfer GG, Pedrini-Martha V, Jackson DJ, Dallinger R, Lieb B Mau B, Newton MA (1997) Phylogenetic inference for binary data on (2021b) The evolution of hemocyanin genes in Tectipleura: a dendograms using Markov chain Monte Carlo. J Comput Graph multitude of conserved introns in highly diverse gastropods. BMC Stat 6:122–131. https://d oi.o rg/1 0.1 080/1 06186 00.1 997.1 04747 31 Ecol Evol 21:36. https:// doi.o rg/1 0. 1186/ s12862- 021-0 1763-3 Melzner F, Mark FC, Pörtner H-O (2007) Role of blood-oxygen trans- Schrödl M (2014) Opinion: time to say “bye-bye pulmonata”? Spixiana port in thermal tolerance of the cuttlefish, Sepia officinalis. Integr 37:161–164 Comp Biol 47:645–655. https:// doi.o rg/ 10. 1093/i cb/ icm074 Schweizer M, Triebskorn R, Köhler H-R (2019) Snails in the sun: strat- Miller K, van Holde KE (1974) Oxygen binding by callianassa cali- egies of terrestrial gastropods to cope with hot and dry conditions. forniensis hemocyanin. Biochemistry 13:1668–1674. https:// doi. Ecol Evol 9:12940–12960. https://d oi.o rg/ 10. 1002/ ece3.5 607 org/ 10. 1021/b i0070 5a017 Stoltzfus A, Logsdon JM, Palmer JD, Doolittle WF (1997) Intron “slid- Miller KI (1985) Oxygen equilibria of Octopus dofleini hemocya- ing” and the diversity of intron positions. Proc Natl Acad Sci USA nin. Biochemistry 24:4582–4586. https:// doi. org/ 10. 1021/ bi003 94:10739–10744. https:// doi. org/1 0. 1073/ pnas.9 4.2 0. 10739 38a015 Strobel A, Hu MYA, Gutowska MA, Lieb B, Lucassen M, Melzner F, Oellermann M, Lieb B, Pörtner H-O, Semmens JM, Mark FC (2015a) Portner HO, Mark FC (2012) Influence of temperature, hyper- Blue blood on ice: modulated blood oxygen transport facilitates capnia, and development on the relative expression of different cold compensation and eurythermy in an Antarctic octopod. Front hemocyanin isoforms in the common cuttlefish Sepia officinalis. Zool. https:// doi.o rg/ 10. 1186/ s12983- 015-0 097-x J Exp Zool A 317:511–523. https:// doi. org/ 10.1 002/j ez. 1743 Oellermann M, Strugnell JM, Lieb B, Mark FC (2015b) Positive selec- Strong EE, Colgan DJ, Healy JM, Lydeard C, Ponder WF, Glaubrecht tion in octopus haemocyanin indicates functional links to tempera- M (2011) Phylogeny of the gastropod superfamily Cerithioidea ture adaptation. BMC Evol Biol 15:133. https:// doi. org/1 0. 1186/ using morphology and molecules. Zool J Linn Soc 162:43–89. s12862-0 15- 0411-4 https://d oi.o rg/ 10. 1111/j. 1096- 3642. 2010.0 0670.x Ohno S (1970) Evolution by gene duplication. Springer, Berlin Swerdlow RD, Ebert RF, Lee P, Bonaventura C, Miller KI (1996) Key- Ponder WF, Lindberg DR, Ponder JM (2019) Biology and evolution of hole limpet hemocyanin: structural and functional characterization the mollusca. CRC Press, Boca Raton of two different subunits and multimers. Comp Biochem Physiol Poverennaya IV, Potapova NA, Spirin SA (2020) Is there any intron B 113:537–548. https:// doi.o rg/ 10. 1016/ 0305- 0491(95) 02091-8 sliding in mammals? BMC Evol Biol 20:164. https://d oi.o rg/1 0. Thonig A, Oellermann M, Lieb B, Mark FC (2014) A new haemocya- 1186/s 12862- 020-0 1726-0 nin in cuttlefish (Sepia officinalis) eggs: sequence analysis and Qian W, Zhang J (2014) Genomic evidence for adaptation by gene relevance during ontogeny. EvoDevo 5:6. https://d oi.o rg/1 0.1 186/ duplication. Genome Res 24:1356–1362. https://d oi.o rg/1 0.1 101/ 2041- 9139-5-6 gr.1 72098.1 14 van Holde KE, Miller KI (1995) Hemocyanins. Adv Protein Chem Rogozin IB, Lyons-Weiler J, Koonin EV (2000) Intron sliding in con- 47:1–81 served gene families. Trends Genet 16:430–432. https:// doi. org/ Vermeij GJ, Dudley R (2000) Why are there so few evolutionary transi- 10.1 016/S 0168-9 525(00) 02096-5 tions between aquatic and terrestrial ecosystems. Biol J Linn Soc Rogozin IB, Wolf YI, Sorokin AV, Mirkin BG, Koonin EV (2003) 70:541–554. https://d oi.o rg/1 0.1 111/j.1 095-8 312.2 000.t b0021 6.x Remarkable interkingdom conservation of intron positions and Vermeij GJ, Wesselingh FP (2002) Neogastropod molluscs from the massive, lineage-specific intron loss and gain in eukaryotic evo- Miocene of western Amazonia, with comments on marine to lution. Curr Biol 13:1512–1517. https:// doi. org/ 10. 1016/ S0960- freshwater transitions in molluscs. J Paleontol 76:265–270. https:// 9822(03)0 0558-X doi. org/ 10. 1666/ 0022- 3360(2002) 076% 3c0265: NMFTMO% Rogozin IB, Wolf YI, Babenko VN, Koonin EV (2006) Dollo par- 3e2.0.C O;2 simony and the reconstruction of genome evolution. In: Albert WoRMS Editorial Board (2020) World register of marine species: for VA (ed) Parsimony, phylogeny, and genomics. Oxford University numbers: uncheck marine only, check extant only. http:// www. Press, Oxford, pp 190–200 marin especi es. org. Accessed 4 Apr 2020 Romero PE, Pfenninger M, Kano Y, Klussmann-Kolb A (2016a) Molec- Yao T, Zhao M-M, He J, Han T, Peng W, Zhang H, Wang J-Y, Jiang ular phylogeny of the ellobiidae (Gastropoda: Panpulmonata) J-Z (2019) Gene expression and phenoloxidase activities of 1 3 Journal of Molecular Evolution (2021) 89:639–655 655 hemocyanin isoforms in response to pathogen infections in aba- not correlate with carbohydrate content. FEBS J 275:3625–3632. lone Haliotis diversicolor. Int J Biol Macromol 129:538–551. https:// doi. org/1 0. 1111/j.1 742- 4658.2 008. 06507.x https:// doi. org/1 0. 1016/j.i jbiom ac. 2019. 02. 013 Zielinski S, Sartoris FJ, Pörtner HO (2001) Temperature effects on Yesilyurt BT, Gielens C, Meersman F (2008) Thermal stability of hemocyanin oxygen binding in an antarctic cephalopod. Biol Bull homologous functional units of Helix pomatia hemocyanin does 200:67–76. https:// doi.o rg/ 10.2 307/ 15430 86 1 3