Animals

Adult cloudy catsharks (Scyliorhinus torazame) were transported from Ibaraki Prefectural Oarai Aquarium to the Atmosphere and Ocean Research Institute at The University of Tokyo. They were kept in 3000 L tanks with recirculating natural SW under a constant photoperiod (light: dark; 12 h: 12 h) at 16°C. Fertilized eggs were collected from captive individuals, and eggs and hatched juveniles were maintained in 1000 L tanks under the same environmental conditions. Both adults and juveniles were fed with chopped squid or sardine to satiation twice a week. All experimental procedures were approved by the Animal Ethics Committee of the Atmosphere and Ocean Research Institute of The University of Tokyo (P19-2). The present study was carried out in compliance with the ARRIVE guidelines.

Tissue preparation for laser microdissection (LMD)

Juvenile catsharks were euthanized with 0.02% (w/v) ethyl 3-aminobenzoate methanesulfonate (Sigma-Aldrich, MO, USA), and whole kidneys were dissected out. The dissected kidneys were sliced transversely into approximately 2 mm-thick kidney blocks. Some blocks were frozen immediately in liquid N2, and the other blocks were fixed using various protocols summarized in Table 1. The fixatives used in the present study are as follows: Methacarn (methanol: chloroform: acetic acid = 6: 3: 1) (Puchtler et al., 1970); UMFIX (methanol: polyethylene glycol 300 = 17: 3) (Vincek et al., 2003); Z7 (0.5% zinc chloride, 17.16 mM zinc trifluoroacetate, 0.05% calcium acetate in 0.1 M Tris-HCl, pH 6.4 to 6.7) (Lykidis et al., 2007); 4% PFA (4% paraformaldehyde in 0.1 M phosphate buffer [pH 7.4] containing 150 mM NaCl and 350 mM urea); 10% NBF (commercial 10% formalin neutral buffer solution [Wako Pure Chemical Industries, Osaka, Japan]), and 1% NBF (1% NBF in 20% sucrose).

Table 1.

Fixatives used in the present study and corresponding RNA integrity numbers.

Fixed blocks of kidney tissue were washed once in Optimal Cutting Temperature (O.C.T.) compound (Sakura Finetek Japan, Tokyo, Japan), and embedded in new O.C.T. compound. All of the embedded blocks were stored at –80°C until sectioning. Frozen blocks were cut at 10 µm using a cryostat maintained at –20°C and consecutive sections were mounted alternatively onto MembraneSlide NF 1.0 PEN (Carl Zeiss Micro Imaging, Jena, Germany) for LMD use or onto MAS-GP-coated glass slides (Matsunami Glass, Osaka, Japan) for reference. This allows a reference histological image of the adjacent section for each LMD-slide. The slides were dried in a cryostat for at least 2 min, followed by fixation in 70% ethanol for 5 min, and then washed with distilled water (DW) for 1 min twice.

Sections on the LMD-slides were stained with Mayer's hematoxylin solution (Wako Pure Chemical Industries) for 1 min, and hematoxylin coloration was enhanced by incubation in DW at 40°C for 1 min. Sections were washed with 70% ethanol for 5 min, dehydrated through graded ethanol of 70%, 95%, and 100% (1 min each), and rapidly dried with a hair dryer. After spraying Liquid Cover Glass (Carl Zeiss Micro Imaging) on the sections, the slides were placed in 50 ml tubes and stored at –80°C until use. All of the staining procedures were performed on ice, except for the coloration enhancement step after hematoxylin staining.

Sections on the reference-slides were stained with the Periodic acid Schiff (PAS) method. Sections were stained with 0.5% Periodic acid solution (Wako Pure Chemical Industries) for 5 min, washed with tap water for 3 min once and then with DW for 5 min three times. Sections were then incubated with Schiff's reagent (Wako Pure Chemical Industries) for 15 min. After rinsing with sulfurous acid solution (10% sodium bisulfite: 1 N HCl: DW = 6: 5: 100) for 3 min three times, the sections were washed with DW for 5 min three times, and then counterstained with Mayer's hematoxylin solution for 3 min. To promote hematoxylin coloration, the sections were washed with running tap water for 15 min. The sections were washed with DW for 3 min, and then dehydrated through the graded ethanol series, cleared in xylene, and mounted with Permount (Thermo Fisher Scientific, Waltham, MA, USA).

RNA extraction and evaluation of the quality of RNA

Total RNA was extracted from two sections (10 µm each) directly from the cryostat, or six scraped-off sections after hematoxylin staining with a FastGene RNA Basic Kit according to the manufacturer's protocol (Nippon Genetics, Tokyo, Japan). RNA quality was assessed by microcapillary electrophoresis with the RNA 6000 Pico LabChip kit (Agilent Biotechnologies, Tokyo, Japan) and the Agilent 2100 bioanalyzer (Agilent Biotechnologies), which assigns an RNA integrity number (RIN) to each RNA electropherogram. This number ranges from 1 (completely degraded RNA sample) to 10 (intact RNA sample) (Schroeder et al., 2006).

LMD and Quartz-Seq

The PALM MicroBeam System (Carl Zeiss Micro Imaging) at Toyo University Itakura Campus was used to excise the selected tissue areas on the LMD slides. The selected tissue areas were determined according to the morphological identification in the reference slides that contain the adjacent sections after PAS-staining. The frozen LMD slides were transferred to the laser microscope (PALM) immediately after being brought to room temperature. The narrow beam ultraviolet laser cut out the areas of interest from surrounding tissue, and the laser was then used to catapult the micro-dissected areas into collecting caps of AdhesiveCap 500 clear tubes (Carl Zeiss Micro Imaging). In the present study, the bundle zone (“Bundle”; a mixture of the first loop, third loop and final segment), the PII segment of the second loop (“2nd”) and the late distal tubule (LDT) of the fourth loop (“4th”) were excised from 12 cryosections, and stored at –80°C until being extracted.

Total RNA from LMD-excised renal segments was extracted using an RNeasy^® Micro Kit (Qiagen, Tokyo, Japan) according to the manufacturer's protocol. Total RNA concentration was measured using the QuantiFluor RNA System (Promega, Madison, WI, USA). Quartz-Seq (Sasagawa et al., 2013) was performed to amplify cDNA from the low-yield total RNA. Firstly, total RNA was added to Quartz-RT primer (5′-TATAGAATTCGCGGCCGCTCGCG ATAATACGACTCACTATAGGGCG(T)24-3′), dNTPs (Thermo Fisher Scientific), and RNasin Plus (Promega) at 12 µl solution and the mixtures were treated at 70°C for 90 s and at 35°C for 15 s to prime the RNA with RT primers. Then, SuperScript III (Thermo Fisher Scientific) master mix was added to a final volume of 20 µl and the reverse transcription reaction was performed at 35°C for 5 min and 45°C for 20 min, followed by a 10 min treatment at 70°C to inactivate the reverse transcriptase. The resulting first-strand cDNA was purified with Agencourt AMPure XP beads (Beckman Coulter, Brea, CA, USA) and treated with Exonuclease I (Takara Bio, Shiga, Japan) at 37°C for 30 min to digest the excess RT primers. Exonuclease I was inactivated by incubation at 80°C for 20 min. After that, the Poly-A tailing reaction was performed by adding reaction mixture containing dATP (Thermo Fisher Scientific), RNase H (Thermo Fisher Scientific), and Terminal Transferase (Roche Applied Science, Mannheim, Germany). A reaction mix containing Quartz-Tagging primer (5′-TATAGAATTCGCGGCCGCTCGCGA(T)24-3′) and MightyAmp DNA polymerase (Takara Bio) was prepared on ice and added to the cDNA templates. The double-stranded DNA synthesis reaction was carried out at 98°C for 2 min 10 sec, 40°C for 1 min, and 68°C for 5 min. After the reaction, the solution was placed on ice to stop the reaction. cDNA was then amplified in a 107 µl reaction system with Quartz-PCR primer (5′-NH2-GTATAGAATTCGCGGCCGCTCGCGAT-3′). cDNA amplification was carried out in 18 amplification cycles at 98°C for 10 s, 65°C for 15 s, and 68°C for 5 min. To verify the quality of the amplified cDNA, 4 µl of the reaction product was electrophoresed on a 1.5% agarose gel. A successful cDNA amplification resulted in a smear of DNA at various lengths. Subsequently, the product was purified using Min-Elute PCR Purification Kit (Qiagen) and the cDNA concentration was measured using Qubit 4 (Thermo Fisher Scientific) and the Qubit dsDNA HS Assay Kit (Thermo Fisher Scientific) and Qubit Assay Tubes (Thermo Fisher Scientific).

RNA-seq analysis

RNA-seq library preparation, sequencing, mapping, and gene expression analysis were performed by DNAFORM (Yokohama, Kanagawa, Japan).

Double-stranded cDNA libraries (RNA-seq libraries) were prepared from the cDNA amplified by Quartz-Seq using the SMART-Seq Stranded Kit (Takara Bio) according to the manufacturer's instruction. The “Bundle”, the “2nd”, and the “4th” samples of catshark kidney were analyzed (n = 3). RNA-seq libraries were sequenced using 150 bp paired end reads on a NextSeq 500 instrument (Illumina, San Diego, CA, USA). Qualified reads were mapped to the catshark Genome assembly: Storazame_v1.0 using STAR (version 2.7.2) (Dobin et al., 2013). Reads on annotated genes were counted using featureCounts (version 1.6.1) (Liao et al., 2014). Transcripts per million (TPM) values were calculated from the mapped reads and normalized to total counts of transcript, and the TPM values for each segment were arranged in a descending order. The expression levels among the three segments (Bundle, 2nd, and 4th) were then statistically analyzed. All of the coding nucleotide sequences predicted on the catshark genome were subjected to NCBI blastx (Expect threshold: 0.05) against the non-redundant protein sequences to identify the putative proteins. The gene names, protein names, and Gene Ontology (GO) terms corresponding to all of the identified proteins were acquired through Uniprot database. The genes containing GO terms of molecular function “transporter activity” and cellular component “plasma membrane” were then short-listed. The raw sequenced reads were deposited in the DNA Data Bank of Japan (DDBJ) under the accession numbers DRR413293–DRR413301.

cDNA cloning

Complementary DNA cloning was performed as previously described in detail (Takagi et al., 2017) using frozen kidney of juvenile catshark. Total RNA was extracted using ISOGEN (Nippon Gene, Toyama, Japan). DNase-treated total RNA (1 µg) was reverse-transcribed into cDNA using a High Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific), following the manufacturer's instructions. Putative nucleotide sequences of mRNAs encoding catshark SGLT2, NKCC2, NCC, UT, Slc1a3, Slc5a8, Slc5a11, Slc13a3, and β-actin were obtained by BLAST search on catshark transcriptome database Squalomix ( https://transcriptome.riken.jp/squalomix/) (Hara et al., 2018). Partial cDNA fragments, which were amplified with Kapa Taq Extra DNA polymerase (Kapa Biosystems, Boston, MA, USA) and gene-specific primer sets (see  Supplementary Table S1 (zs220092_TableS1.pdf)), were ligated into pGEMT Easy plasmid vector (Promega). The nucleotide sequences were determined using a DNA sequencer (ABI PRISM 3130, Life Technologies, Carlsbad, CA, USA). The sequence data were deposited in the DDBJ under the accession numbers LC739024 (ncc), LC739025 (sglt2), LC739026 (nkcc2), LC739027 (ut), LC739028 (slc1a3), LC739029 (slc5a8), LC739030 (slc5a11), and LC739031 (slc13a3).

In situ hybridization

Kidneys of adult female catsharks fixed in Bouin's solution without acetic acid were embedded in Paraplast (McCormick Scientific, Richmond, IL, USA). Serial sections were prepared at 7 µm and mounted onto MAS-coated slides (Matsunami Glass). The cloned cDNAs were amplified from the plasmids with Primestar GXL DNA polymerase (Takara Bio) using M13 forward and reverse primers. The PCR products were subsequently purified using Wizard SV Gel and PCR Clean-up System (Promega) and used to synthesize digoxigenin (DIG)-labeled antisense and sense cRNA probes with a DIG RNA labeling kit (Roche Applied Science), following the manufacturer's protocols. Hybridization and washing were conducted using a previously described protocol (Aburatani et al., 2022). Stained sections were counterstained with Kernechtrot Stain Solution (MUTO PURE CHEMICALS, Tokyo, Japan) and mounted using Permount (Thermo Fisher Scientific) (sglt2, nkcc2, ncc, and ut), or mounted using CC/Mount (Diagnostic BioSystems, Pleasanton, CA, USA) without counterstaning for weakly stained sections (slc1a3, slc5a8, slc5a11, and slc13a3).

Quantitative real-time PCR (qPCR)

The gene expression levels were measured by qPCR using 7900HT Fast Real Time PCR System (Applied Biosystems, Forster city, CA, USA) with a KAPA SYBR FAST qPCR Kit (Kapa Biosystems) and gene-specific primer sets (see  Supplementary Table S1 (zs220092_TableS1.pdf)). The cloned plasmids containing the target sequences were serially diluted as standard templates for quantification. The expression levels of β-actin mRNA were used as internal control.

Statistical analysis

Values are presented as means ± S.E.M. The qPCR results were analyzed by one-way ANOVA followed by Tukey's multiple comparison test or Kruskal-Wallis test followed by Dunn's multiple comparison test. The TPM values among the three segments (Bundle, 2nd, and 4th) were analyzed by one-way ANOVA followed by Tukey's multiple comparison test. P values < 0.05 were considered as statistically significant. All analyses were performed using Prism Ver. 9 for Windows (GraphPad, San Diego, CA, USA).

Evaluation of tissue preparation methods

For successful LMD dissection and subsequent RNA-seq, an optimal fixation method that provides substantial histological resolution and RNA integrity is required. To avoid possible cross-linking damage to RNA integrity by formaldehyde, several formaldehyde-free fixatives, Methacarn, UMFIX, and Z7, were examined (Table 1). The RNA quality was assessed by the RNA integrity number (RIN) in an RNA electropherogram. For each fixation method, RNA samples were prepared from 1) the tissues following cryostat sectioning without mounting on glass slides (“Cryostat”) and 2) the tissues scraped off from glass slides following hematoxylin staining (“Scraped off”). The RNA samples prepared from the unfixed tissues had RIN values between 5.73 ± 0.66 (“Cryostat”) and 4.60 ± 1.10 (“Scraped off”). Most RNA samples from fixed tissues with formaldehyde-free fixatives had RIN values higher than 5, but some samples from the Methacarn-fixed for 1 h, UMFIX-fixed for 3 h, and Z7 for 12h had RIN values lower than 5 (Table 1). On the other hand, RNA samples prepared from the formalin-fixed paraffin-embedded (4% PFA overnight) tissues had RIN values less than 3.5, which was considerably lower than those of formaldehyde-free fixatives.

Figure 1A shows the transverse section of the formaldehyde-fixed kidney of adult catshark. The kidney consists of the sinus zone and the bundle zone, with single nephrons traversing repeatedly between the two zones (see Fig. 1D). The sinus zone is filled with the second and the fourth loops (Fig. 1B), while five tubular segments (the descending and ascending segments of the first and third loops and the collecting tubule) are aggregated in the bundle zone. The five bundle segments were closely wrapped by a sac-like peritubular sheath (circled in Fig. 1C). Figure 1E, G represent the sinus zone of UMFIX- and Z7-fixed kidney samples, respectively, and histological preservation by the formaldehyde-free fixatives was too low for identifying specific segments for the subsequent LMD. In addition, during the preparation of the adjacent sections for PAS staining, we noticed that histological structures were easily distorted in the formaldehyde-free fixed sections, and thus confident identification of specific tubular segments on the reference slides was not possible (Fig. 1F, H).

Fig. 1.

Anatomical features of the cartilaginous fish kidney and histological preservation by the formaldehyde-free and formaldehyde-based fixatives for LMD. (A) Transverse section of the formaldehyde-fixed adult catshark kidney stained with PAS. SZ, sinus zone; BZ, bundle zone. (B) Magnified image of the sinus zone. The number labels 2 and 4 mark the PII segment and the LDT, respectively. Note that, in the case of PII segment, the brush border of the apical surface is stained with PAS. (C) Magnified view of the bundle zone. The circle represents five tubular segments of a single nephron. (D) Schematic illustration of cartilaginous fish nephron. CT, collecting tubule; EDT, early distal tubule; IS, intermediate segment; LDT, late distal tubule; NS, neck segment; PI and PII, proximal segment I and II, respectively; RC, renal corpuscle. (E–L) Histological images of catshark kidney for LMD (hematoxylin-stained [E, G, I, K]) and reference (PAS-stained [F, H, J, L]) slides. Tissues were fixed with UNFIX for 12 h (E, F), Z7 for 12 h (G, H), 10% NBF for 1 min (I, J), or 1% NBF for 15 min (K, L). The number label 2 represents the PII segment of second loop. Scale bars: 1 mm (A), 100 µm (B, C), 300 µm (E–L).

The above results suggest that the formaldehyde-free fixatives examined so far are not suitable for excising specific nephron segments using LMD. Therefore, the use of formaldehyde-based fixatives was examined (Table 1). The low RIN score in the 4% PFA overnight samples might be due to excessive cross-linking and modification of nucleic acids by formaldehyde, and possible damage to nucleic acids by formic acid produced by the oxidation of formaldehyde. Therefore, we next examined various fixation conditions for frozen tissues using a commercial 10% neutral buffered formalin (NBF; approximately 3.7% paraformaldehyde) solution. Four conditions were examined: 1) 10% NBF for 5 min under negative pressure, 2) 10% NBF for 1 min, 3) 1% NBF in 20% sucrose for 1 min, and 4) 1% NBF in 20% sucrose for 15 min (Table 1). Fixation in 10% NBF under negative pressure is considered to facilitate formaldehyde penetration into the tissues, thereby minimizing the exposure time to formaldehyde. However, the RIN scores were 5.80 and 1.00, indicating that the negative pressure protocol was not compatible with RNA extraction. On the other hand, the reduced fixation time for 1 min in 10% NBF achieved high RIN scores of 7.90 and 8.00. Furthermore, a low concentration of NBF at 1% for 1 min or 15 min also improved RIN score to over 7.0 except for one sample fixed in 1% NBF for 15 min (Table 1).

Cryostat sections with formaldehyde-based fixatives improved histological preservation compared to those with formaldehyde-free fixatives (Fig. 1I, K). When adjacent sections were stained with PAS, the proximal segment of the second loop and the distal segment of the fourth loop were clearly distinguishable by the presence and absence, respectively, of brush border on the apical surface of lumen (Fig. 1J, L). Although both fixation times of 1 min and 15 min resulted in satisfactory RIN values, the longer fixation time reduced the hastiness of the procedures and thus produced a more consistent fixation condition. Collectively, fixation in a low-concentration formaldehyde (1% NBF) for 15 min maintains a reasonable tissue morphology and RNA quality, and thus allows LMD clipping of specific renal tubules for subsequent RNA-seq.

Fig. 2.

Expression of the marker genes used in the present study. (A–D) Na⁺-coupled glucose transporter (sglt2mRNA), (E–H) Na⁺,K⁺,2Cl^–cotransporter (nkcc2 mRNA), (I–L) Na⁺,Cl^– cotransporter (ncc mRNA), (M–P) urea transporter (ut mRNA). (A, E, I, M) Low magnification of catshark kidney showing both the bundle and sinus zones. Magnified views of the bundle zone (B, F, J, N) and the sinus zone (C, G, K, O). The marks represent the following: open arrowheads, the ascending part of the PIa segment (“ii” in the tubules); arrows, the EDT of the third loop (“iv” in the tubules); arrowheads, the collecting tubule (“v” in the tubules); aLDT and pLDT, anterior and posterior LDT, respectively. (D, H, L, P) Schematic images representing localization of sglt2 (D), nkcc2 (H), ncc (L), and ut (P) mRNAs that are shown in red color. aPIa, ascending part of PIa segment. Scale bars: 200 µm (A, E, I, M) and 50 µm (B, C, F, G, J, K, N, O).

Evaluation of segment specificity of LMD samples

Four membrane transporter genes were selected as markers as they are specifically and abundantly expressed in particular segments of the nephron according to previous studies. Segment-specific expressions of these marker genes were confirmed in the catshark kidney (Fig. 2). Signals for sglt2 mRNA were detected in the bundle zone, but not in the sinus zone (Fig. 2A–C). In the bundle zone, sglt2 mRNA signals were observed only in a single nephron segment (open arrowheads in Fig. 2B). Observation using serial sections revealed that the sglt2 mRNA-expressing tubule is the ascending (posterior) limb of the PIa segment (Fig. 2D). On the other hand, nkcc2 mRNA signals were found both in the bundle and sinus zones (Figs. 2E–G). In the bundle zone, intense nkcc2 mRNA signals were detected in tubules comprised of the largest columnar cells, indicating that the early distal tubule (EDT; filled arrows) of the third loop expresses nkcc2 mRNA (Fig. 2F, H). In the sinus zone, nkcc2 mRNA was detected in tubules comprised of small and thin squamous cells, indicating that nkcc2 mRNA is expressed in the fourth loop (the late distal tubule; LDT) (Fig. 2G, H). The ncc mRNA signals were observed in the LDT and the transitional region between LDT and CT (Fig. 2I, K). In the sinus zone, however, the posterior half of LDT was intensely stained by ncc mRNA signals, which was different from the anterior half of the LDT that was intensely stained by nkcc2 mRNA (Fig. 2G, H, K, L). The ut mRNA signals were found in the tubules in bundle zone (Fig. 2M), and the ut mRNA positive segment (filled arrowheads in Fig. 2N) did not overlap with segments expressing sglt2 (open arrowheads) or nkcc2 (filled arrows) mRNAs, indicating that the ut mRNA-positive segment is the collecting tubule and preceding transitional region between LDT and CT (Fig. 2P). These results confirmed that the four marker genes are effective for evaluating segment specificity of LMD samples.

In the present LMD study, three regions of catshark nephron were excised: 1) five tubular segments wrapped by the peritubular sheath in the bundle zone (“Bundle”, the descending and ascending segments of the first and third loops, and the collecting tubule), 2) the PII segment of the second loop (“2nd”), and 3) the LDT of the fourth loop (“4th”). Although no significant difference was observed in marker gene expressions among the LMD samples due to the large variations in observed data, qPCR analysis revealed that the expression of four marker genes in the LMD samples corresponded to the local abundance of mRNA shown by the results of in situ hybridization (Fig. 3). Expressions of sglt2, nkcc2, and ut were observed in the “Bundle” samples, in which the first loop, third loop, and collecting tubule were included. Expression of nkcc2 mRNA was also found in the “4th” samples. Expression of ncc mRNA was detected in the “4th” samples, while none of the four genes had high expression in the “2nd” samples. These results further demonstrated that contamination of other segments is negligible for the obtained LMD samples.

Expression profiles of membrane transporter mRNAs in the LMD samples

Segment-dependent mRNA expression profiles were examined using three individuals. In the present study, we focused on membrane transport proteins, including the solute carrier (Slc) family, for deducing the reabsorbing and/or secreting functions of each nephron segment. By searching GO terms as described in the Methods section, we shortlisted the top 30 highly expressed membrane transporter genes in each LMD segment (Tables 2 to 4), and we statistically analyzed their expression levels among the three LMD segments. As a result, five, 13, and four genes were found to be significantly expressed in the “Bundle”, “2nd”, and “4th” segments, respectively, which characterize the functions of those segments (Tables 2 to 4).

Fig. 3.

Expression of the marker genes in the LMD samples for “Bundle” (five tubular segments in the bundle zone), “2nd” (the PII segment of second loop), and “4th” (the LDT of fourth loop). Data are presented as means ± S.E.M. with each sample value as an open circle (n = 3). No significant differences were observed in expressions of the marker genes among the LMD samples due to the large variations.

Many segment-dependent genes were detected in the second loop samples, but generally these have not been investigated yet in the elasmobranch kidneys. Therefore, we selected four genes (slc13a3, slc5a8, slc5a11, and slc1a3) and further investigated their localization by in situ hybridization (Fig. 4). slc13a3 has the highest TPM values among Slc family genes in the second loop, and its expression was intensely detected in the sinus zone (Fig. 4A). The slc13a3-positive segments were relatively large and did not express nkcc2 and ncc mRNAs (Fig. 4P), indicating that slc13a3 gene expression is specific to the PII segment of second loop. The intense signals observed in the second loop by in situ hybridization were consistent with the results of RNA-seq, where the TPM value of slc13a3 mRNA was significantly higher in the “2nd” sample than that in the “4th” and “Bundle” samples.

Two genes encoding SLC5 members (slc5a8 and slc5a11) were listed as genes abundantly expressed in the second loop based on RNA-seq. However, the localization patterns of slc5a8 and slc5a11 mRNAs in the nephron were different. The slc5a11 mRNA signals were found only in the sinus zone (Fig. 4F). The slc5a11-positive tubules were the same tubules expressing slc13a3 mRNA, although the signal intensity of slc5a11 mRNA was much lower than that of slc13a3 mRNA. Meanwhile, slc5a8 mRNA signals were faintly detected in one segment within the bundle zone (arrows in Fig. 4M). The slc5a8 mRNA-positive segment co-expressed sglt2 mRNA (Fig. 4R) but was negative for nkcc2 (Fig. 4Q) and ut (Fig. 4S) mRNAs, suggesting that slc5a8 is expressed in the ascending part of the PIa segment of the first loop.

Fig. 4.

Expression of slc13a3 (A–E), slc5a11 (F–J) and slc5a8 (K–O) mRNAs in the sinus zone (A, B, F, G, K, L) and the bundle zone (C, D, H, I, M, N) of catshark kidney. Sections were hybridized with antisense (A, C, F, H, K, M) or sense (B, D, G, I, L, N) probes. (P–S) Adjacent sections were stained with ncc ([P] sinus zone), nkcc2 ([Q] bundle zone), sglt2 ([R] bundle zone), or ut ([S] bundle zone) mRNA. The labels ii, iv, and v mark the ascending part of PIa segment, the EDT, and the collecting tubule, respectively. Both slc13a3 and slc5a11 mRNA signals were localized in the PII segment of the second loop (A, F), while the weak mRNA signals of slc5a8 were found in the ascending part of PIa segment (arrows in [M]), where mRNA signals of sglt2 were co-localized (arrows in [R]). Schematic images represent localization of slc13a3 (E), slc5a11 (J), and slc5a8 (O) mRNAs that are shown in red color. Scale bars, 50 µm.

Although slc1a3 mRNA was listed as a gene characteristic of the second loop, expression levels were too low to be detected by in situ hybridization. In all in situ hybridization experiments, negative control slides using sense cRNA probes resulted in no signals (Fig. 4B, D, G, I, L, N).

Segment-dependent gene expression profiles

In the present study, we successfully obtained lists of genes abundantly expressed in the second loop, fourth loop, and the tubular bundle from the RNA-seq analyses on the LMD samples. We focused on membrane transporter genes, including the slc family, as they play crucial roles in reabsorption and secretion of solutes across the epithelial cells of renal nephrons (Hediger et al., 2004). We found that the marker genes used in the present study were on the lists: nkcc2, sglt2, and ut genes on the “Bundle” list, and nkcc2 and ncc on the “4th” list.

Among the genes on the lists, we further focused on the Na⁺-coupled cotransporters abundantly expressed in the second loop because of the following reasons: 1) Na⁺coupled cotransporters are the secondarily active transporters directly contributing to reabsorption or secretion of solutes in nephron segments, and 2) little information has been available on the function of proximal segment(s) in the second loop except possible divalent ion secretion (Hasegawa et al., 2016). The most abundantly expressed slc gene in the second loop was slc13a3. In situ hybridization clearly showed that slc13a3 mRNA is abundantly expressed in the PII segment of the second loop. The Slc13 family is comprised of five members that encode electrogenic Na⁺coupled cotransporters for anions or di- and tri-carboxylates (Markovich and Murer, 2004). In mammalian kidneys, Slc13a3 is localized on the basolateral membrane of the proximal tubule, and transport intermediates of the Krebs cycle such as succinate, citrate, and α-ketoglutarate (Bergeron et al., 2013). Slc13a3 also plays an important role in the renal elimination of xenobiotics by secreting organic anions into urine.

Table 4.

List of top 30 highly expressed membrane transporter genes in the “4th” samples.

The renal secretion of organic anions is considered to be achieved with several steps as follows (Breljac et al., 2016): 1) primary active NKA creates an inwardly directed Na⁺ gradient, 2) secondary active Na⁺-coupled dicarboxylate cotransporter (Slc13a3) creates an outwardly directed gradient of α-ketoglutarate, and 3) tertiary active organic anion transporters (such as Slc22a6 and Slc22a8) transport organic anions into epithelial cells by exchanging organic anions with α-ketoglutarate. All three components (NKA, Slc13a3, and Slc22a6 or Slc22a8) were localized on the basolateral membrane (see Breljac et al., 2016). On the apical membrane, 4) multiple ATP-binding cassette (ABC) and Slc transporters secrete organic anions into the lumen of the proximal tubule (Schlatter et al., 2006). In this regard, our RNA-seq data supported such a scheme for the shark nephron since the transcripts of slc22a6 and abcc1 genes were on the list of highly expressed membrane transporter genes in the second loop. These results suggest that the PII segment of the second loop contributes to secretion of organic anions in the elasmobranch nephron, in addition to divalent ion secretion.

In the list of highly expressed transporters in the second loop, two Slc5 members were found, slc5a8 and slc5a11. The slc5a11 mRNA was localized in the PII segment of the second loop, while slc5a8 mRNA signals were co-localized with sglt2 mRNA in the ascending part of the PIa segment in the bundle zone (Figure 4). The Slc5 family is known as the Na ⁺ -coupled glucose cotransporter family, but these proteins also transport monocarboxylates such as lactate and pyruvate, short-chain fatty acids, and ketone bodies (Plata et al., 2007). Slc5a8 transports monocarboxylates and ketone bodies, and thus was named Na ⁺ -coupled monocarboxylate transporter (SMCT1) (Gopal et al., 2004). In mammalian kidneys, the reabsorption fraction of lactate from the glomerular filtrate is over 95%. In the renal tubular system, lactate and ketone bodies are also used as the major fuels for active reabsorption of electrolytes and nutrients (Yanase et al., 2008). Since an increased reliance on ketone bodies as an energy source was observed following prolonged starvation in elasmobranchs (Zammit and Newsholme, 1979), Slc5a8 could have a regulatory role in energy metabolism in the elasmobranch nephron. Since both sglt2 and slc5a8 mRNAs were expressed in the ascending part of PIa segment, it is likely that this segment is important for reabsorption of energy substrates such as glucose, lactate, and ketone bodies. On the other hand, Slc5a11 (SGLT6) is a Na ⁺ -coupled myo-inositol cotransporter in the apical membrane of renal tubule cells, where it mediates myo-inositol reabsorption from the primary urine (Kottgen et al., 2018). Myo-inositol and its derivatives have functions including regulation of ion channels, metabolic flux, mRNA transcription and translation, stress response, and embryonic development (Kiani et al., 2021). In the kidney, myo-inositol is also important as a protective osmolyte against osmotic stress.

slc1a3, which is another slc gene examined in the present study, is a member of glutamate/neutral amino acid transporters. The TPM value of slc1a3 was less than 5% of that of slc13a3, and we could not confirm the gene expression site by in situ hybridization due to the low expression level. Slc1a3 is a glial-type glutamate transporter (also known as a glutamate-aspartate transporter, GLAST, or an excitatory amino acid transporter, EAAT1) abundantly expressed in the cerebellum in mammals (Kanai and Hediger, 2004). Slc1a3 is responsible for glutamate uptake in the brain, which is crucial for glutamatergic neurotransmission by removing glutamate from the synaptic cleft (Berger and Hediger, 2006). However, expression of the Slc1a3 gene has not been proven in mammalian renal tubules. Instead, another glutamate transporter, Slc1a1, is localized in the apical membrane of proximal tubules in the mammalian kidneys (Berger and Hediger, 2006). Slc1a1-knockout mice exhibited dicarboxylic aminoaciduria, confirming its crucial role for reabsorbing glutamate in the proximal tubule (Kanai and Hediger, 2004). Further investigations on slc1a1 and slc1a3 mRNAs are required to understand their roles in the kidney of elasmobranchs.

Seawater is an environment of excess divalent ions, and thus the kidneys of marine teleosts and cartilaginous fishes must excrete divalent ions. Excretion of NaCl is achieved by the branchial ionocytes (teleost fishes) and the rectal gland (cartilaginous fishes), whereas it has been well documented that kidneys are responsible for the excretion of divalent ions (see Takvam et al., 2021). We previously revealed co-localization of sulfate transporters Slc26a1 and Slc26a6 in the basolateral and apical membranes, respectively, in the PII segment of the elephant fish nephron, implying that the PII segment of the second loop is a secretory segment for excess solutes (Hasegawa et al., 2016). In the present study, we found intense expression of slc13a3 mRNA in the PII segment. As mentioned above, Slc13a3 most likely contributes to the renal elimination of xenobiotics by secreting organic anions into urine, further supporting the secretory role of the PII segment. On the other hand, expression of Na⁺-coupled myo-inositol transporter (slc5a11) mRNA was also found in the PII segment. Although the cellular localization of Slc5a11 is uncertain in the cartilaginous fish nephron, distribution of these transporters in the apical membrane has been well documented in the mammalian nephron. These results imply that the PII segment may not be a simple secretory segment, but it may also be involved in reabsorption of valuable solutes. In the catshark nephron, at least four segments can be identified in the proximal tubule: the descending PIa (the first loop with no expression of sglt2 mRNA), the ascending PIa (the first loop with sglt2 and slc5a8 mRNAs), PIb, and PII (with slc13a3 and slc5a11 mRNAs). To understand further the function of each segment, finer-scale clipping of each segment by the LMD is required, with the guidance of reference section stained with in situ hybridization and/or immunohistochemistry.

Among the four slc transporters examined in the second loop, expressions of slc5a11 and slc13a3 were only detected in the PII segment by in situ hybridization, confirming the quantitative results of RNA-seq from the LMD samples. However, expression of slc5a8 mRNA was detected in the ascending part of PIa segment in the bundle zone, which is inconsistent with the low TPM values of slc5a8 in the “Bundle” samples (4.4 ± 1.0). Since the TPM values of sglt2 (slc5a2) that marked the PIa segment were significantly high in the “Bundle” samples, it is uncertain why the TPM values of slc5a8 mRNA were low in the “Bundle” samples. Perhaps unannotated contig(s) encoding Slc5a8 exists in the RNA-seq data of the “Bundle” samples, leading to incomplete counting of transcripts.

Conclusion and perspectives

In the present study, we developed a protocol for clipping specific nephron segments of catshark kidney by LMD, and successfully conducted subsequent RNA-seq analysis for profiling the segment-dependent gene expressions. Most cartilaginous fishes are obligatory marine species, while a limited number of sharks and rays have the ability to inhabit low salinity environments (Ballantyne and Robinson, 2010). During the acclimation to low salinity environments, euryhaline sharks and rays produce large volumes of dilute urine to excrete excess water; they do this by increasing glomerular filtration and solute reabsorption in the renal tubules (Janech et al., 2006; Imaseki et al., 2019; Aburatani et al., 2022). Indeed, we found marked increase in the expression of the ncc gene in the LDT of FW-acclimated bull shark, Carcharhinus leucas (Imaseki et al., 2019), and of the nkcc2 gene in the EDT of FW-acclimated red stingray, Hemitrygon akajei (Aburatani et al., 2020). It is likely that water reabsorption in renal tubules would also be decreased in low salinity environments. Currently, regulatory mechanisms of renal function remain poorly resolved in euryhaline elasmobranchs, and therefore, the segment-dependent gene expression profiling that we report here will be a powerful technique for unraveling the renal mechanisms and regulation in euryhaline elasmobranchs.