Tissue- and Salinity-specic Na–Cl Cotransporter (NCC) Orthologues Involved in the Sea lamprey (Petromyzon Marinus) Adaptive Osmoregulation

Two ncc orthologues (termed ncca and nccb) were found in the sea lamprey genome, whereas nkcc2 was not. In a phylogenetic comparison among other vertebrate amino acids, NCC and NKCC deduced sequences, the sea lamprey NCC’s occupied basal positions within the NCC clade. In freshwater, ncca mRNA was found only in the gill and nccb only in the intestine, whereas both were found in the kidney. Acclimation to seawater increased nccb mRNA in the intestine and kidney. Intestinal nccb mRNA also increased during late metamorphosis. The electrophysiological approach in the Ussing chamber of intestinal tissue ex vivo showed signicant differences between freshwater and seawater-acclimated juveniles. Luminal application of indapamide (NCC inhibitor) resulted in 73 and 30% inhibition of short-circuit current (I sc ) in the proximal and distal intestine, respectively. The luminal application of bumetanide (NKCC inhibitor) did not affect intestinal I sc . Indapamide also inhibited ex vivo intestinal water absorption. Our results indicate that NCCb is likely the key passive ion cotransporter protein for ion uptake by the lamprey intestine to facilitate water absorption in seawater. As such, the preparatory increases in intestinal nccb mRNA expression during metamorphosis are likely critical to the development of whole animal salinity tolerance. (ii) 2R, an nkcc2 gene demonstrate that an indapamide-sensitive mechanism (i.e., NCCb) on the apical membrane to be a major contributor to anion absorption in the lamprey intestine. Overall lumen-to-serosa ion inux along the intestinal tract was observed, with marked differences depending on intestinal regions. In SW-acclimated sea lamprey juveniles, approximately 70 and 30% of the net absorptive currents in the AI and PI, respectively, are mediated by presumably NCCb. The remaining uncharacterized absorptive current might involve different apical transporters and/or bidirectional ion movement (e.g., cation secretion and anion absorption, or vice versa). Also, we cannot rule out that indapamide did not inhibit all NCCb I sc , even at the highest tested doses.


Introduction
Lampreys are one of two surviving lineages of jawless shes. Agnathans appear to have diverged during the late Ordovician period ~ 450 Mya (million years ago) 1,2 and persisted through several con rmed mass extinctions on Earth 3,4 . Adult lampreys have remained morphologically conserved based on a limited fossil record and have inhabited marine and estuarine environments from at least the late Devonian period ~ 360 Mya 5 . Two episodes of whole-genome duplication likely occurred prior to the divergence of Agnatha 6 . Additional whole-genome duplications have been reported in the ancestors of modern salmonids 7 , castomids (suckers) 8 , and carp 9 , and additional partial genome duplications have been reported elsewhere in shes 10 . Given Agnathans' basal phylogenetic position and the many genomic and phenotypic evolutionary transitions that have occurred since the divergence of this group, molecular studies in this group are important for understanding and contextualizing physiological shifts throughout vertebrate evolution.
Modern lampreys are iono-and osmoregulators, in that they maintain their plasma osmolality at approximately one-third that of seawater (SW) regardless of external salinity. This osmoregulatory strategy is found in nearly all other extant vertebrate lineages, with the notable exception of elasmobranchs, coelacanths, and lung shes 11,12 . The persistence of an osmoregulatory system over evolutionary time underscores its adaptive importance, particularly considering the extensive increases in phenotypic complexity that has occurred in derived shes 13 .
Regardless of the pathway for water absorption, the driving force is the same: water absorption from the lumen is driven by a luminal-to-serosal osmotic gradient, particularly those related to Cl − movements 21 .
In this process, ingested SW lls lateral intercellular spaces between ion-transporting enterocytes. The intestinal water absorption from the lumen into the serosa likely occurs by osmosis following Cl − molecular transport 21 . The luminal Cl-uptake mechanism is species-and tissue-dependent but always includes a member of the SLC family of solute carrying proteins such as NKCC2, NCC, and Cl − /HCO 3 − exchanger [23][24][25] . The active transport of Cl − across the intestinal epithelium is associated with high levels of Na + /K + -ATPase (NKA) activity 26 . The basolateral NKA energizes a favorable electrogenic gradient for the absorption of Cl − to occur via the SLC cotransporters localized in apical and basolateral membranes of enterocytes 14,20 . Excess of monovalent ions that are incorporated into the blood plasma is secreted across the gills 27 .
The NCC is perhaps most well-known for its crucial role in renal ion reabsorption in the mammalian distal convoluted tubule, where 5-10 % of the ltered Na + and Cl − are reabsorbed 28-30 . Expression of an orthologous NCC (termed NCC1) in teleost shes acclimated to freshwater (FW) environments has been shown to be highly expressed in the kidney, presumably ful lling a similar role in NaCl absorption 31,32 .
Fishes have another NCC gene (termed NCC2), which is classically referred to as a ' sh-speci c NCC', although it has not been found in sharks 32 . In teleost shes, NCC2 is localized to the apical membrane of branchial ionocytes and is responsible for branchial NaCl absorption in FW [33][34][35][36] . Within the NKCC1/NKCC2/NCC gene superfamily, NCC1 and NCC2 form distinct lineages, indicating that this gene family has functionally diverged on an evolutionary timescale 37 .
In teleost shes, NKCC2 is the predominant cation-chloride cotransporter that facilitates apical ion absorption in the SW intestine. The nkcc2 increases following seawater exposure in a number of teleost species 25,26,38 . Ion and water uptake by the intestine has been shown to be inhibited by bumetanide, a selective inhibitor of NKCC 20 . The NCC2 appears to have a role in Na + and Cl − removal from ingested SW in the esophagus and intestine of the Japanese 39,40 and European eel 41 . However, information regarding the importance of NCC in intestinal ion and water uptake in other sh species is still scarce 20 .
Although these underlying mechanisms for intestinal SW osmoregulation have been elucidated in teleost shes, whether such molecular mechanisms are present in the lampreys' intestine was unknown prior to the present study. Thus, we aimed to functionally characterize the putative molecular pathway for intestinal ion absorption during SW osmoregulation in the sea lamprey (Petromyzon marinus). Sea lamprey has a complex anadromous life cycle marked by a dramatic metamorphosis from a benthic lter-feeding larva (ammocoete) into a parasitic juvenile that migrates from FW to SW. After ~ 2 years at sea, the adults migrate upstream in search of a place for spawning. Although larval-juvenile metamorphosis has been morphologically conserved from at least the early Cretaceous, ~ 125 mya 42 , recent studies on fossil records suggest that ammocoetes are specializations of modern lampreys rather than a relic of ancient lampreys 43 .
We hypothesized that intestinal ion absorption in SW-acclimated sea lamprey occurred by an NKCC and/or NCC. We present the discovery of two functionally divergent NCC isoforms in lamprey, which appear to be basal in structure and function to the NCC isoforms of more derived shes. Moreover, we show that NCC and not an NKCC2 is the likely pathway for ion absorption in marine lampreys' intestine.

Tissue pro le and molecular characterization
We comprehensively surveyed cDNA sequences encoding relevant ion-transporting proteins in the sea lamprey genome in preliminary genomic analyses. Partial sequences of orthologous cDNAs encoding two distinct NCC-like proteins were discovered. The accession numbers of these newly identi ed sea lamprey NCC sequences are BK014291 and BK014292, which we have subsequently named "NCCa" and "NCCb", respectively. We could not identify a gene encoding an NKCC2-like protein in any available genome for different lamprey species, whereas we did for NKCC1-, NCCa-, and NCCb-like proteins. Sea lamprey NCCa and NCCb amino acid deduced sequences shared the highest sequence identity to known NCC families; molecular phylogenetic analysis grouped NCCa and NCCb with other NCC sequences (Fig. 1). NCCa occupied a basal position among the NCC1 and NCC2 clades. NCCb was basally positioned within the clade of the conventional, kidney-speci c NCC1 sequences.
In tissue pro les, mRNA abundance of ncca and nccb showed drastically different expression patterns. The ncca was expressed predominately in the gills and kidney at ~ 400-fold higher than other tissues in FW-acclimated juvenile sea lamprey ( Fig. 2A). Remarkably, ncca mRNA was even greater expressed in larvae than juveniles, both acclimated to FW (Barany, unpublished data). Acclimation to SW resulted in ~ 90 % reduction of ncca in the gills and kidney. The nccb was predominately expressed in the anterior intestine (AI), posterior intestine (PI), and kidney at ~ 1,000-fold higher than other tissues in SWacclimated sea lamprey (Fig. 2B). The nccb mRNA in the AI and kidney of SW-acclimated sea lamprey was 5-and 2-fold higher, respectively, than in FW-acclimated sea lamprey.
In additional metamorphic and salinity experiments, nccb in the intestine increased at the late stages of metamorphosis and then increased further after SW acclimation. The nccb mRNA levels increased 6-and 5-fold in the AI and PI, respectively, of eld-collected post-metamorphic migrating juveniles, compared to juveniles (Fig. 3A, B). In lab-acclimated sea lamprey, intestinal nccb mRNA was signi cantly higher in both the AI (9-fold) and PI (5.5-fold) in FW-acclimated juveniles compared to larvae (Fig. 3C, D). Acclimation to SW further increased sea lamprey nccb mRNA abundance in the AI (2-fold) and PI (1.5fold).

Bioelectrical properties of sea lamprey intestine
The transepithelial potential (TEP) and I sc values observed for this tissue are consistent with a preferential anion-absorbing tissue. Chloride is the primary physiologic anion and the ion suspected of carrying the majority of the current. Thus, both FW-and SW-acclimated sea lamprey were preferentially anion absorptive. In the AI, I sc was signi cantly less negative after SW acclimation (Fig. 4A). I sc of the AI of SW-acclimated sea lamprey was signi cantly less negative than that of the PI. The transepithelial electric resistance (TER) did not change across intestinal regions and/or salinities (Fig. 4B). The magnitude of the indapamide effects varied between intestinal regions and with environmental salinity (Fig. 5). In FW, approximately 40 % and 25 % of the baseline anion absorption was indapamidesensitive in the AI and PI, respectively. In SW, indapamide-sensitive anion absorption was signi cantly higher in the AI than in the PI. Indapamide-sensitive anion absorption in the AI was 73 % of baseline anion absorption in SW-acclimated juvenile sea lamprey, compared to only 30 % of that PI.

Discussion
It has been hypothesized that a rst-round (1R) of whole-genome duplication in vertebrates led to an initial divergence of NKCC from NCC and that a second round (2R) of whole-genome duplication in the vertebrate lineage resulted in multiple NKCC and NCC isoforms, NKCC1/NKCC2 and NCC1/NCC2 44 .
Our molecular phylogenetic analysis showed that the NCC1 and NCC2 of bony shes were grouped into distinct NCC1 (conventional, kidney NCC) and NCC2 (ray-nned sh-speci c gill NCC) clades. Sea lamprey NCCb was clearly positioned at a basal position within the NCC1 clade. Sea lamprey NCCa was not explicitly located within either the NCC1 or NCC2 clades and was instead positioned basal to both NCC clades. All retrieved NKCC1-, NCCa-, and NCCb-like proteins from other lamprey species were grouped with the respective identi ed sequences of the sea lamprey (Petromyzon marinus).
The large changes in gene transcription during development and response to environmental changes are good indicators of functional relevance. We found that nccb mRNA levels were signi cantly upregulated in the later stages of metamorphosis, particularly in post-metamorphic juveniles which were captured while actively migrating toward the sea. Barany et al. 14,46 observed increases in intestinal NKA transcription, mRNA, and activity in the late stages of metamorphosis. Both NKA and NCCb might be important to the development of SW tolerance that occurs late in metamorphosis. Based on our transcriptional and pharmacological experiments, NCCb in the intestine was further upregulated in juveniles after exposure to SW. Furthermore, the expression and activity of NCCb in SW were higher in AI than in PI. This salinity-responsive pattern for NCCb is similar to previously reported apical NKCC2 in the intestine of euryhaline teleosts transferred to higher salinities 25,47,48 , but also in the basolateral NKCC1 in gills of the sea lamprey 49 . These changes in intestinal NCCb expression and activity during metamorphosis and SW-acclimation re ect similar intestinal NKA activity and expression 14 , indicating that both of these transporters play a critical role in the development of salinity tolerance.
Although the present study was explicitly focused on functional characterization of the intestinal NCCb isoform, we posit a potential mechanism for branchial ion uptake and renal NaCl reabsorption in FWacclimated sea lamprey, the NCCa isoform. In our mRNA tissue pro le, ncca was highly expressed in FWacclimated juveniles' gills and signi cantly down-regulated after SW acclimation in agreement with what was found in the adult up-streaming sea lamprey 50 . This salinity-speci c regulation of gill NCC in order to adapt to different salinities is similar to what has been described of gill NCC2 in teleosts 36,51,52 .
The transcription of both NCCa and NCCb in the sea lamprey kidney is unique and warrants further discussion and investigation. The sea lamprey ncca was abundant in the kidney of FW-acclimated juveniles and strongly down-regulated after SW acclimation. By contrast, nccb mRNA levels in the kidney were upregulated after SW acclimation. The typical salinity-responsive pattern of the conventional kidney NCC1 in shes is that it is abundant in FW and down-regulated in SW 31 . This indicates that NCCa in sea lamprey kidney functions similar to the conventional, kidney-speci c NCC involved salt retention of teleosts in FW. The transcription of the sea lamprey ncca in both the gills and kidney is particularly interesting because, in teleost shes, the NCC in the gills (NCC2) and kidney (NCC1) are two distinct NCC isoforms 31,34,51 . It is important to note that the elasmobranch NCC also exhibits FW-speci c function in the gills and kidney 32,53 , similar to the expression pattern of the sea lamprey NCCa.
In our effort to functionally characterize the intestinal NCCb isoform, we used an electrophysiological approach to elucidate the underlying mechanisms contributing to intestinal ion transport in response to environmental salinity. The sea lamprey intestine's anion absorptive properties reported here agree with what has previously been shown in the sea lamprey intestine 14 . Surprisingly, absorptive I sc in the AI was higher in FW-acclimated juvenile sea lamprey. It is known that FW-acclimated sea lamprey exhibits some minimal drinking 14,54 . Thus, it could be that the AI might offer extra osmoregulatory support by facilitating ion uptake in ion-poor environments, suggesting that intestinal ion uptake from imbibed FW might be more important than previously thought. In SW, higher anion-absorptive I sc was detected in the PI than in the AI. This demonstrates that these intestinal regions have different functional responses to environmental salinity, similar to what has also been previously reported in teleosts 25,38,55 . Intestinal precipitates were observed only in SW-acclimated juveniles (Barany et al., unpublished). Thus, the reduction in anion absorption in the AI of SW-acclimated juveniles may also indicate a functional switch in intestinal physiology to secrete HCO 3 − and/or H + 56 into the AI lumen to lower luminal osmolality and facilitate net water absorption 57,58 .
In the juvenile sea lamprey, the apparent lack of sensitivity to bumetanide indicates that anion absorption by NKCC2 is unlikely to be present in the sea lamprey intestine. This is further supported by our inability to identify a gene encoding NKCC2 in the sea lamprey, western brook lamprey, Paci c lamprey, or Arctic lamprey genomes. We cannot rule out the possibility that a sea lamprey NKCC2 could be insensitive to bumetanide 59,60 and that nkcc2 is present but has yet to be detected in the available lamprey genomes. Note that no signi cant effects in I sc were obtained with either chlorothiazide or hydrochlorothiazide, which were initially employed to target a putative NCC. It also should be noted that NKCC1 has been identi ed and characterized in sea lamprey gills 49 but is present only at very low levels in the intestine. Moreover, we were able to identify NKCC1-, NCCa-, and NKCCb-like proteins in the available different lamprey species genomes. Therefore we can surmise two possible explanations for the lack of a sea lamprey nkcc2: (i) Agnathans diverged after 2R, but without a functional role, an nkcc2 gene was eventually lost; and (ii) Agnathans diverged before 2R, and an nkcc2 gene never existed in sea lamprey.
Our results demonstrate that an indapamide-sensitive mechanism (i.e., NCCb) on the apical membrane appears to be a major contributor to anion absorption in the lamprey intestine. Overall lumen-to-serosa ion in ux along the intestinal tract was observed, with marked differences depending on intestinal regions. In SW-acclimated sea lamprey juveniles, approximately 70 and 30% of the net absorptive currents in the AI and PI, respectively, are mediated by presumably NCCb. The remaining uncharacterized absorptive current might involve different apical transporters and/or bidirectional ion movement (e.g., cation secretion and anion absorption, or vice versa). Also, we cannot rule out that indapamide did not inhibit all NCCb I sc , even at the highest tested doses.
Additionally, we suggest that DIDS-sensitive mechanisms signi cantly contribute to I sc substantially The contribution of NCCb to anion-absorption in the AI, in particular, is about twice as high in SW compared to FW. Thus, increased environmental salinity may cause a shift of Cl − absorption mechanisms 38 , in keeping with previously reported increases in drinking rate, intestinal NKA activity, and intestinal water absorption 14 , as well as increased gill Cl − secretory mechanisms 49 . The higher sensitivity of I sc to indapamide in the AI than the PI is consistent with higher nccb transcription in this region.
To better understand how intestinal water absorption was related to NCCb activity, we characterized indapamide-sensitive intestinal water transport in SW-acclimated sea lamprey juveniles. As expected, the serosal addition of ouabain inhibited active water absorption, demonstrating NKA is critical involvement in this process, as has previously been shown in teleosts and sea lamprey 14,62 . The luminal addition of indapamide also inhibited water absorption to a similar degree as the addition of ouabain. These results indicate that the activity of NCCb in the sea lamprey intestine is a critical anion absorptive mechanism that is functionally coupled to water absorption.
In summary, our nding that an NCCb, and not NKCC2, functions in the intestine offers new insights into the diversi cation of the vertebrate NKCC/NCC protein superfamily as it might be related to partial and/or whole-genome duplication events. Although the results reported here are an important discovery and describe speci c differences in the osmoregulatory roles of NKCCs and NCCs between lamprey and laterderived shes, further studies are still needed to complete our understanding of ion-absorptive pathways in the gill, kidney, and intestine of the sea lamprey.

Animals and experimental designs
All experiments were carried out in accordance with USGS guidelines and approved by the USGS were not offered food because they naturally stop feeding during metamorphosis and does not resume until they begin parasitic feeding once they reach the ocean.

Sampling protocol
For all tissue collection, lamprey were euthanized using MS-222 (400 mg L − 1 buffered with NaHCO 3 , pH 7.0) (Argent Chemical Laboratories, Redmond, WA, USA). Tissue was collected from the gills and two sections of the intestine: (i) the AI, corresponding to 1.5-2 cm after the end of the oesophagus; and (ii) the PI, which corresponds to a section of distal intestine, 2-3 cm in length, delimited by the rectal sphincter. All samples were immediately frozen and stored at -80°C until analysis, except in cases of measurements of intestinal water permeability and electrophysiology, where the tissues were used immediately after dissection.

Gene identi cation and molecular phylogenetic analysis
Gene-speci c primers were designed based on sea lamprey genomic information available from the SIMRbase (https://genomes.stowers.org/). The identity of annotated genes was con rmed using sequence identity comparison and NCBI blast analyses.
The deduced amino acid sequences of the obtained annotated genes were predicted using the Translate tool provided by the ExPASy bioinformatics resource portal (https://www. expasy.org). Phylogenetic analysis was carried out on a balanced selection of vertebrate NKCC1, NKCC2 and NCC peptides available from NCBI GenBank, Ensemble Genomes and SIMRbase, using a ClustalW alignment (https://www.ebi.ac.uk/clustalw) and the neighbour-joining method with bootstrap analysis for 1000 cycles implemented by Mega-X software. Accession numbers for the peptides used in the phylogenetic

RNA isolation and quantitative real-time PCR
Total RNA was extracted from frozen tissues using the TRIzol reagent (Molecular Research Center Inc., Cincinnati, OH, USA) following the manufacturer's protocol. The total RNA concentration and purity of each sample were determined spectrophotometrically using a Take3 microvolume plate reader (BioTek Instruments, Inc., Winooski, VT, USA), and RNA integrity was assessed using gel electrophoresis. Only high-purity RNA (1.9 < A260/A280 > 2.2) were used for cDNA synthesis. First-strand cDNA synthesis was accomplished using a High-Capacity Reverse Transcription Kit following the manufacturer's protocol (Applied Biosystems, Carlsbad, CA, USA).
Quantitative real-time PCR (RT-qPCR) reactions were carried out in 10 µl reactions containing 5 ng cDNA template (estimated from the input of total RNA), 150 nM forward and reverse primers, and SYBR Select master mix (ThermoFisher, Waltham, MA, USA). All RT-qPCR reactions were performed in optical 96-well reaction plates in a StepOnePlus™ Real-Time PCR System (Applied Biosystems, Inc., Foster City, CA, USA) covered with adhesive seals. The thermocycling procedure was as follows: an initial step of 50°C (2 min) followed by 95°C (2 min), and then 40 consecutive cycles of 95°C (15 s), 60°C (1 min), and 72°C (30 s). Melting curves were used to ensure that only a single PCR product was ampli ed and to verify the absence of primer-dimer artefacts. A pool of concentrated cDNA from all samples for each tissue was used for a standard curve, which was a ve-point, serial 10-fold dilution from 50 ng to 5 pg. The linearity and e ciency of ampli cation for each primer combination was assessed. Control reactions with RNase free water (no template control) and RNA (no reverse transcriptase during cDNA synthesis) were included in the analysis to ensure the absence of primer-dimerization and genomic DNA contamination. For all pairs of primers, linearities (R 2 ) and ampli cation e ciencies were in the ranges 0.998-1 and 0.923-0.939, respectively. The relative gene expressions were quanti ed using the ΔΔC T method 64 , corrected for e ciencies 65 , and referenced to the relative mRNA expression of the difference between the target gene and an endogenous reference gene (eef1α; gene encoding eukaryotic elongation factor 1-alpha) for each sample. Primer sequence details are shown in Table 1. Table 1 Primers used for real-time qPCR expression analysis. GenBank accession number referred to the unique sequence identi er in the gene assembly and sizes of the ampli ed products (bp).

Chemicals
Bumetanide (a selective NKCC2 inhibitor), ouabain (a selective NKA inhibitor), and indapamide (a selective NCC inhibitor) were purchased from ThermoFisher (Waltham, MA, USA) and prepared as concentrated stocks in dimethyl sulfoxide (DMSO) except ouabain, which was dissolved in lamprey serosal saline. The nal volume of DMSO never exceeded 0.2 %, and no effect of DMSO at this concentration was observed on multiple tissues tested (data not shown).

Electrophysiology in Ussing chambers
Anterior and posterior intestinal tissues were dissected, washed, and mounted in Ussing chambers. The tissues were opened longitudinally, attened, and placed between the chambers containing 1.

Determination of iso-osmotic in vitro intestinal water absorption
Water absorption was determined following previous methods 14 . Brie y, the intestinal sections were isolated, ushed and then incubated for 30 min in lamprey serosal saline bubbled with a physiological gas mixture (humidi ed 0.5 % CO 2 in oxygen). Saline was maintained at 15°C and pre-gassed for at least 30 min before experimentation. The ends of the intestinal samples were sealed at each end using dental line and placed in a beaker containing serosal saline supplied with the gas mixture. The intestinal preparations were weighed every 15 min for 1 h. At the end of the experiment, each intestinal section was opened, placed on graph paper, and then photographed. The surface area of the intestinal segment within the tied region was determined using ImageJ software (NIH, Bethesda, MD).

Statistical analysis
Statistical comparisons across experimental procedures were performed using unpaired Student's t-test, one-sample t-test, one-way ANOVA, and two-way ANOVA, as indicated in the gure legends. All data are represented as the mean ± SEM. Prior to these statistical analyses, both normality and equal variance were con rmed. When ANOVA yielded signi cant differences, Tukey's post hoc test was used to identify these. Statistical signi cance was accepted at P < 0.05.  Figure 1 Molecular phylogenetic tree of vertebrate NKCC/NCC family. Rat KCC1 was used as an outgroup to root the tree.

Figure 2
Gene transcription of ncca (A) and nccb (B) isoforms in a tissue pro le of juvenile sea lamprey acclimated to either FW or SW. Abbreviations: B, brain; P, pituitary; G, gill; H, heart; L, liver; AI, anterior intestine; PI, posterior intestine; K, kidney; M; muscle. Values are relative to the FW brain tissue. Each column represents mean ± SEM (n = 3). *Signi cant differences within a given tissue between environmental salinities (two-way ANOVA, Tukey´s post hoc).