Analysis of aquaporins from the euryhaline barnacle Balanus improvisus reveals differential expression in response to changes in salinity

Abstract

Barnacles are sessile macro-invertebrates, found along rocky shores in coastal areas worldwide. The euryhaline bay barnacle Balanus improvisus (Darwin, 1854) (= Amphibalanus improvisus) can tolerate a wide range of salinities, but the molecular mechanisms underlying the osmoregulatory capacity of this truly brackish species are not well understood. Aquaporins are pore-forming integral membrane proteins that facilitate transport of water, small solutes and ions through cellular membranes, and that have been shown to be important for osmoregulation in many organisms. The knowledge of the function of aquaporins in crustaceans is, however, limited and nothing is known about them in barnacles. We here present the repertoire of aquaporins from a thecostracan crustacean, the barnacle B. improvisus, based on genome and transcriptome sequencing. Our analyses reveal that B. improvisus contains eight genes for aquaporins. Phylogenetic analysis showed that they represented members of the classical water aquaporins (Aqp1, Aqp2), the aquaglyceroporins (Glp1, Glp2), the unorthodox aquaporin (Aqp12) and the arthropod-specific big brain aquaporin (Bib). Interestingly, we also found two big brain-like proteins (BibL1 and BibL2) constituting a new group of aquaporins not yet described in arthropods. In addition, we found that the two water-specific aquaporins were expressed as C-terminal splice variants. Heterologous expression of some of the aquaporins followed by functional characterization showed that Aqp1 transported water and Glp2 water and glycerol, agreeing with the predictions of substrate specificity based on 3D modeling and phylogeny. To investigate a possible role for the B. improvisus aquaporins in osmoregulation, mRNA expression changes in adult barnacles were analysed after long-term acclimation to different salinities. The most pronounced expression difference was seen for AQP1 with a substantial (>100-fold) decrease in the mantle tissue in low salinity (3 PSU) compared to high salinity (33 PSU). Our study provides a base for future mechanistic studies on the role of aquaporins in osmoregulation.

Fig 1. Initial phylogenetic classification of the aquaporins in B. improvisus.

A phylogenetic tree was constructed using aquaporin sequences from B. improvisus and other arthropods. In addition, the human aquaporins were included for reference. The analysis was done using the program PhyML 3.0 at the Phylogeny.fr website, creating an unrooted tree. The four main subfamilies according to Stavang et al [22] are indicated to the right and different aquaporin subgroups to the left. Aqp8-type aquaammoniaporins are abbreviated to aquaammoniaporins. The B. improvisus aquaporins are marked with a dot. The numbers on the branches are aLRT SH-like support-values. The scalebar shows substitutions per site.

Fig 2. Gene structure of the eight aquaporin genes in B. improvisus.

The B. improvisus aquaporin genes contain 4–6 exons. AQP1 and AQP2 are expressed as two alternative splice variants where exon 5 is excluded in one of the isoforms (indicated by bent arrows). Coding parts of exons are indicated in black, 5’ and 3’ UTR regions in grey and introns by thin lines. Exons, but not introns, are displayed to scale. For indication of intron lengths see S2 Fig. The minimum lengths of the UTRs being part of the first or last coding exon are shown. An asterisk indicates that the UTR continues in an upstream or downstream non-coding exon not shown in the figure. Genomic data, cDNA clones and RNA-seq data were used to define the gene structure. In the case of the BIBL1 and BIBL2 mRNA, the sequence of the complete 3´end was determined by RACE.

Fig 3. Schematic overview of the B. improvisus aquaporins.

The length [number of amino acids (aa)] and main features of the aquaporins in B. improvisus are displayed. Included are also the splice forms for Aqp1 and Aqp2. The region spanning from transmembrane 1 to transmembrane 6 (see Fig 4), is indicated in dark grey and the N- and C-termini in light grey. The sequences are roughly aligned according to the position of the NPA sites. The total number of amino acid residues for each aquaporin is indicated to the right.

Fig 4. Protein alignment of the aquaporins in B. improvisus.

All aquaporins from B. improvisus, including splice variants, are aligned. Dark grey indicates identical amino acids and light grey functionally similar amino acids. The positions of the transmembrane spanning regions, based on TMHMM predictions of Aqp1_v1, are indicated with a black line under the sequences. The two NPA motifs (or the variants thereof) are indicated with boxes, and the four amino acids in the constriction site (see Fig 5) are indicated with an asterisk. Some sequence differences are apparent between Aqp1_v1 and Aqp1_v2 even outside the C-terminal portion corresponding to exon 5 and 6. This is because the two sequences are obtained from different individuals.

Fig 5. Amino acids in the Ar/R constriction site of aquaporins in B. improvisus.

The four amino acids in the Ar/R constriction motif in the Balanus aquaporins were identified based on protein alignments with the human water aquaporin AQP1 and the E. coli aquaglyceroporin Glpf, which have well characterized constriction regions. All human aquaporins are included for comparison. Positions and residues for human AQP1 and E. coli Glpf are indicated at the top. The names of the B. improvisus aquaporins are shown in bold.

Fig 6. The Ar/R constriction sites in the 3D homology models of the B. improvisus aquaporins.

A-C: The aquaporins are viewed from the extracellular side and the four residues of the constriction site are shown as a ball-and-stick model. A) Aqp1; B) Aqp2; C) Glp2. D: Side view of the Aqp12 constriction site, showing the protrusion of R135 into the pore at the extracellular entrance.

Fig 7. Functional analysis of AQP proteins.

The B. improvisus aquaporins Aqp1, Glp1 and Glp2 were heterologously expressed in the yeast P. pastoris, purified and reconstituted into liposomes. The spinach aquaporin SoPIP2;1 was included as control. The transport of water and glycerol out of the proteoliposome vesicles were measured as light scattering (indicates liposome swelling/shrinkage) using stopped flow spectroscopy measured at 436 nm. For each experiment, at least three traces were averaged and fitted to either single or double exponential functions using the method of least squares. One typical stopped flow spectroscopy trace is displayed. Reported rate constants are an average of three independent measurements. Error bars show standard deviation.*** = significantly different from control (ANOVA, p<0.001). A-B) water transport. C-D) glycerol transport. The y-axis in B and D shows the rate constant for the curve fitted to the data.

Fig 8. mRNAexpression of aquaporins in B. improvisus.

The mRNA expression of the B. improvisus aquaporins was determined by RNA-seq in an adult (A) and in cyprid larvae (B) cultivated at seawater salinity (≈ 30 PSU). Total RNA was prepared and sequenced by paired-end Illumina sequencing. Normalized expression levels for the aquaporins were estimated by mapping reads to the aquaporins ORFs using the program RSEM. The expression levels for the different AQP genes are shown as FPKM (Fragments Per Kilobase of transcript per Million mapped reads) for the adult and TMM-normalized FPKM for cyprids. The sample sizes were n = 1 for the adult and n = 4 for the cyprids, with 300 cyprids pooled in four independent replicates. Error bars for the cyprid batches in B shows standard deviation. AQP2 was significantly higher expressed in cyprids compared to all the other aquaporins (ANOVA, P<0.001; see supplementary S1 Table for significant changes for all pair-wise comparisons). Absolute expression values should not be compared between life stages due to lack of normalization between the life stages.

Fig 9. Expression of aquaporins of B. improvisus in adults during exposure to various salinities.

Adult individuals were incubated at three different salinities for 14 days (3, 20 and 33 PSU). For RNA preparation, soma, cirri and mantle of the adults were separated. For each salinity, the tissues (soma, cirri or mantle) from eighteen adults were pooled three-by-three to give six independent samples (n = 6). Quantitative PCR was used to determined aquaporin expression levels relative to actin. In case of cirri at 20 and 33 PSU only 5 independent samples were used in the qPCR due to very low RNA amounts obtained from one of the samples in each case. For some of the samples the expression was below the level of detection (N.D., not detected). Asterisks indicate significant levels (ANOVA): *** p<0.001, ** p<0.01, * p<0.05. Error bars show standard deviation.