Genomic, biochemical and expressional properties reveal strong conservation of the CLCA2 gene in birds and mammals

In silico sequence analysis of gCLCA2 and generation of antibodies

Detailed gene positions, sizes, gene and amino acid (aa) sequences from chicken, quail, ostrich, pig, cat, and mouse CLCA loci were extracted from the NCBI (https://www.ncbi.nlm.nih.gov/) and Ensembl (http://www.ensembl.org/index.html) databases as described by Plog et al. (2009). NCBI or Ensembl identifiers for CLCA2 sequences used are listed in Fig. S1. Exon-intron boundaries were established using WebScipio (Hatje, Hammesfahr & Kollmar, 2013) and aligned by GenePainter (Hammesfahr et al., 2013). Predicted protein domains were identified by the NCBI Conserved Domain Database (Lu et al., 2020), EMBL-EBI HMMER web server (Potter et al., 2018), Phobius webserver (Käll, Krogh & Sonnhammer, 2007), SOSUI (Hirokawa, Boon-Chieng & Mitaku, 1998), SignalP 3.0 (Bendtsen et al., 2004) algorithms and manual alignments. Asparagine (N)-linked glycosylation sites were predicted using the NetNGlyc webserver 1.0 (http://www.cbs.dtu.dk/services/NetNGlyc/). Turkey CLCA2 was not incorporated in the in silico analysis due to the low quality of the full-length gene and aa sequences stored in the NCBI and Ensemble databases (XP_031410715.1, XM_031554855.1, ENSMGAT00000009704.2). Phylogenetic relationship based on protein sequences of galline, quail, ostrich, feline, porcine and murine CLCA2 sequences was inferred by using the Maximum Likelihood method and JTT matrix-based model conducted in the MEGA X software package with 100 bootstrap replicates (Tamura, Stecher & Kumar, 2021) (S6).

Anti-gCLCA2 antibodies were generated similar to anti-porcine CLCA1 antibodies (Plog et al., 2009). In brief, an oligopeptide corresponding to aa 875 to 888 (WTAPGDDFDKGQAA) in the C-terminal region of gCLCA2 was synthesized and conjugated with Limulus polyphemus hemocyanin (LPH). The LPH- conjugated peptide was used for immunization of two rabbits. Specific IgG-antibodies were isolated from the antisera using a cyanogen bromide immunization-peptide coupled sepharose column and named gC2.

Animals and tissues

In accordance with the 3R principle for the reduction of animal experiments, all tissues used in this study were obtained from the veterinary diagnostic pathology tissue archive of the Department of Veterinary Pathology, Freie Universität Berlin, Germany (VetPathFU) and originated either from the veterinary clinical diagnostic service unit or previous experimental studies. No animal was bred, raised, kept or euthanized specifically for this study. For chickens, 45 freshly frozen or formalin fixed and paraffin embedded (FFPE) tissues (Table S2) from ten-week old female individuals (Gallus gallus domesticus, Hampshire x White Leghorn, n = 3) and the gonads of age-matched male chickens (Hampshire x White Leghorn, n = 3) were used from (Bartenschlager et al., 2022). In brief, the tissues were by-products from slaughtered animals intended for human consumption. The animals had been bred, housed, and slaughtered in the Albrecht Daniel Thaer-Institute of Agricultural and Horticultural Sciences of the Humboldt-Universität zu Berlin, Germany under the permission of the State Office of Health and Social Affairs (approval number IC 114-ZH70). Weight at harvest was 1–1.2 kg (females) and 1.3–1.5 kg (males). The animals were raised in groups of 25, with infrared heat lamps offered until week five. They were fed with fledgling rearing feed until week eight and young hen feed afterwards. Miscanthus litter was used for housing enrichment. Harvesting was conducted according to national guidelines, which includes anesthesia by head blow and rapid exsanguination via jugular veins and carotid arteries. Females were harvested in the morning and males in the morning of the following day. FFPE tissues for immunohistochemical analyses, including esophagus and skin from shanks, abdomen and foot from one female and two male ostriches (Struthio camelus), three male quails (Coturnix sp.), and skin from three turkeys (Meleagris gallopavo) and skin of cats (Felis catus) were provided by the veterinary clinical postmortem diagnostic service unit of VetPathFU with no association to animal experiments. The tissues were obtained from the routine diagnostic spectrum to determine the cause of death of animals kept by private owners and free of histopathological changes. The permission to further use these tissues for research purposes was given by signature on the necropsy submission form by the owners. Additionally, FFPE skin samples from each of three mice and pigs from previous experimental studies (Braun et al., 2010; Plog et al., 2012a approval numbers T 0104/06 and G 0323/06, respectively) were obtained from the archive of VetPathFU and used for immunohistochemical analyses. In brief, 10-weeks-old female C57BL/6J mice were kept in cages enriched with nesting material. All animals had unlimited access to standard pelleted food and tap water. The room temperature was at 22 ± 2 °C and the relative humidity at 45–65%. A 12-h light/dark cycle was maintained. For experimental procedures, all mice were sacrificed by cervical dislocation in accordance with the national guidelines. Furthermore, the piglets (Euroc  × Piétrain) were 18 days old, male, castrated and kept for four weeks in flatdeck compartments in groups of six piglets enriched with playthings. All animals had unlimited access to mash food and tap water. The room temperature was 28 °C at stabling and was gradually decreased to 22 °C within 10 days with air humidity at approx. 65%. The light programme consisted of a 16 h light and 8 h dark phase. With 45 days of age, all piglets were anesthesized with ketamine hydrochloride (Ursotamin^®, 10%; Serumwerk Bernburg AG, Germany) and azaperone (Stresnil^®, Jansen-Cilag, Neuss, Germany) and euthanized using tetracaine hydrochloride, mebezonium iodide and embutramide (T61^®, Intervet, Germany). All efforts were made to minimize animal discomfort and suffering.

Molecular cloning and sequencing of gCLCA2

The gCLCA2 open reading frame (ORF) was cloned as described with minor modifications (Bartenschlager et al., 2022). In brief, the gCLCA2 ORF was amplified from a batch of tissues including pharynx, crop, proctodeum and footpad from animal #2 (Table S2). The gCLCA2 ORF was tagged with the enhanced yellow fluorescent protein (EYFP) at the C-terminus by cloning it into the pEYFP-N1 vector (Clontech, Mountain View, California, USA). The resulting plasmid (gCLCA2#2) was sequenced using the primer walking method (Data S5). Three plasmids from independent experiments yielded identical results.

RT-qPCR tissue localization of gCLCA2 mRNA

mRNA expression was analyzed using RT-qPCR as described (Bartenschlager et al., 2022). In brief, total RNA was isolated from galline tissues (Table S2), reverse transcribed, and the cDNA diluted to a final concentration of 1 ng/µl. Specific exon 13/14-boundaries spanning primers (upstream: 5′-CCAGGCTAACAGGACTACC-3′; downstream: 5′-GAAACCTCCTCTTCTGACCTGAAC-3′) were used to detect gCLCA2 or the reference gene phosphoglycerate kinase (PGK1, upstream: 5′-AAAGTTCAGGATAAGATCCAGCTG-3′; downstream 5′-GCCATCAGGTCCTTGACAAT-3′; Olias et al., 2014) using a SYBR green qPCR assay (Thermo Fisher Scientific, Waltham, MA, USA). The gCLCA2-PCR product corresponds to the gCLCA2 protein region from aa760 to aa809 (QANRTTVPQTAMPWSHAMYIPGYVENGKLKMNPSRPPAIENNVQVRRGGF). gCLCA2 mRNA was considered to be expressed when C_t-values of 35 or less were detected in at least two out of three tested animals.

Transient transfection of HEK293 cells

HEK293 cells (ATCC, Manassas, Virginia, USA) were transiently transfected as described with minor modifications (Bartenschlager et al., 2022). In brief, cells were grown in six-well plates in Dulbecco’s Modified Eagle’s Medium (DMEM) supplemented with 10% heat-inactivated fetal calf serum (FCS), 1% HEPES, and 1% penicillin/streptomycin. When reaching 80–90% confluence, the cells were transfected with 2 µg of a plasmid containing gCLCA2#2 or EYFP alone (mock) using 8 µl polyethylenimine (PEI) per well. 12 h post transfection, the cells were washed with phosphate buffered saline (PBS) and serum-free DMEM was added. 48 h after transfection, the cells of each well were lyzed using 500 µl radioimmunoprecipitation assay (RIPA) buffer supplemented with a protease inhibitor cocktail (complete Mini, EDTA-free, Roche Diagnostics, Rotkreuz, Switzerland). The protein concentration of supernatants and cell lysates were quantified using the bicinchoninic acid (Thermo Fisher Scientific, Waltham, Massachusetts, USA) method prior to freezing at −20 °C.

Endoglycosidase treatment

For glycosylation analysis, lysates from gCLCA2-transfected cells were deglycosylated by incubation with 25 U/ml endo H, 50 U/ml PNGase F or left untreated at 37 °C over night according to the manufacturer’s protocols (New England Biolabs, Ipswich, Massachusetts, USA).

Immunoblotting

Cell lysates and supernatants of gCLCA2 transfected cells were analyzed using immunoblotting as described with minor modifications (Bartenschlager et al., 2022). In brief, samples of cell lysates or concentrated cell culture supernatant were reduced in 1,4-dithiothreitol (DTT) and separated using a 10% SDS-polyacrylamide gel electrophoresis. Proteins were transferred to a (PVDF)-membrane (https://www.linguee.de/englisch-deutsch/uebersetzung/polyvinylidene+fluoride.html) and blocked with 5% non-fat milk. Membranes were probed with antibody gC2 in a three-fold dilution series from 5 µg/ml to 0.05 µg/ml, or mouse monoclonal anti-YFP (cat. G163; ABM, Vancouver, Canada) diluted at 1:500, or mouse monoclonal anti-beta-actin (A5441, Sigma-Aldrich, St. Louis, Missouri, USA) diluted at 1:1,000. Membranes were incubated with horseradish peroxidase-conjugated goat anti-rabbit (115-035-068, Jackson Immuno Research Laboratories, Inc., West Grove, Pennsylvania, USA) or goat anti-mouse (111-035-144, Jackson Immuno Research Laboratories, Inc.) secondary antibodies and developed using enhanced chemiluminescence (Supersignal West Pico Plus, Thermo Fisher Scientific, Waltham, MA, USA). The gCLCA2 protein was only detected by this technique when using the anti-YFP antibody; however, it was undetectable when the gC2 antibodies were used.

Immunocytochemistry of transfected HEK293 cells

Immunocytochemistry was performed as described with minor modifications (Bartenschlager et al., 2022). In brief, HEK293 (ATCC) cells were grown on 8-well tissue chamber slides and transfected with gCLCA2#2 or EYFP- mock plasmids. 48 h after transfection, the cells were briefly fixed in ice-cold methanol followed by a 4% paraformaldehyde fixation for 10 min. After permeabilization with 0.1% Triton X-100 in PBS and blocking with 10% goat normal serum (GS) and 0.05% Tween 20 in PBS, cells were probed with untreated or pre-absorbed antibody gC2 each used at 2 µg/ml or irrelevant affinity-purified rabbit polyclonal anti-porcine CFTR antibody (Plog et al., 2010) (S3). Alexa fluor 568 conjugated goat anti-rabbit (AB_143157, Invitrogen, Carlsbad, California, USA) were used as secondary antibodies followed by 4′, 6-diamidino-2-phenylindole (DAPI) nuclear counterstain. All in vitro experiments were repeated three times.

Tissue and cellular localization of gCLCA2 protein using immunohistochemistry and immunofluorescence

All galline tissues in which gCLCA2 mRNA was detected at C_t-values below 35 were analyzed via immunofluorescence to identify gCLCA2 expressing cell types. Furthermore, skin and esophagus from chicken, ostriches, quails, as well as skin of turkeys, mice, pigs, and cats were analyzed via immunohistochemistry. Immunofluorescence and immunohistochemistry were performed as described with minor modifications (Bartenschlager et al., 2022). In brief, FFPE-tissues were cut, mounted on adhesive glass slides, and dewaxed. For immunohistochemistry, endogenous peroxidase was blocked by adding 0.5% H₂O₂ in methanol. For immunofluorescence analysis, tissue sections of chickens were permeabilized with 0.1% Triton X-100 in PBS. Antigen was retrieved using 1 mg/ml recombinant protease from Streptomyces griseus. Slides were blocked with 10% Roti-ImmunoBlock and 20% GS in PBS for immunohistochemistry and 10% GS and 0.05% Tween 20 in PBS for immunofluorescence, both for 30 min. The slides were probed with the immunopurified gC2 or irrelevant affinity-purified rabbit polyclonal (anti-porcine CFTR, Plog et al., 2010) antibodies at 2 µg/ml. Additionally, mouse monoclonal anti-cytokeratin (AE1/AE3, M3515, Agilent Dako, Santa Clara, California, USA) antibodies were used at 1:400. For immunofluorescence, Alexa fluor 568-conjugated goat anti-rabbit (AB_143157; Invitrogen, Waltham, MA, USA) secondary antibodies were used diluted at 1:200, followed by DAPI nuclear counterstain. For immunohistochemistry with AE1/AE3 primary antibodies, 3,3′-diaminobenzidine (DAB) was added after incubation with goat anti-mouse biotinylated secondary antibodies (BA-9200; Vector Laboratories, Burlingame, California, USA) diluted at 1: 200 and an avidin-biotin complex. For immunohistochemistry with the gC2 primary antibody, DAB was added after incubation with the Histofine Simple Stain Mouse MAX PO anti-rabbit polymer kit (414341F; Nichirei Biosciences Inc., Tokyo, Japan). Potential cross reactivity of the gC2 antibody with pig, cat, mouse, turkey, quail and ostrich CLCA2 orthologues was tested by epitope sequence alignment using the NCBI Protein Blast (https://blast.ncbi.nlm.nih.gov/Blast.cgi?PAGE=Proteins, Table S4).

Avian and mammalian CLCA2 genes and their overall protein structures are conserved

Similar to mammals (Bartenschlager et al., 2022; Mundhenk et al., 2018), avian CLCA2 (aCLCA2) are single-copy genes located directly adjacent to the ODF2L gene in all species analyzed here (Fig. 1A). While in mammals the region between CLCA2 and SH3GLB1 comprise the complex and divergent CLCA1/3/4 locus, birds contain only one single CLCA1 gene (Fig. 1A). The presence of only two CLCA homologues but also the shorter intronic and intergenic regions that correspond to the more compact organization of avian genomes (Ellegren, 2005) make the consensus avian CLCA gene locus much smaller than that in mammals (Fig. 1A, Bartenschlager et al., 2022). The aCLCA2 genes comprise 14 exons that encode for a single, putatively functional protein in all bird species analyzed here (Figs. 1B, 2), identical to all mammals investigated to date (Patel, Brett & Holtzman, 2009). The nucleotide numbers vary in only four exons by three to nine nucleotides between avian and mammalian CLCA2 genes (Fig. 1B), causing only slightly distinct protein lengths (Fig. 2). Overall, our findings establish a high degree of evolutionary conservation of the genomic CLCA2 architecture across birds and mammals (Fig. 1B). For all CLCA2 orthologous genes investigated here, the predicted proteins share the canonical CLCA protein domain architecture as described (Patel, Brett & Holtzman, 2009). In silico prediction suggests a signal peptide within the first 22 to 43 aa, indicating entry into the secretory pathway which is a highly conserved trait in all avian and mammalian CLCA2 sequences (Fig. 2). An N-CLCA domain (PFAM identifier: pfam08434) containing a proteolytic HExxH-motif and a cysteine-rich domain is prepended to a vWA domain (PFAM identifier: pfam13519) and a bsr domain (Fig. 2, S1, Patel, Brett & Holtzman, 2009). In contrast to other CLCA proteins but consistent with maCLCA2, the vWA domain of CLCA2 of birds does not contain an intact metal ion dependent adhesion site (MIDAS, DxSxS-T4-D5). In concordance to maCLCA2, an fn3 (PFAM identifier: PF00041) and a TM domain are predicted in the C-terminal cleavage product for aCLCA2 (Fig. 2, Fig. S1). The presence of designated beta4-integrin binding motif (IBM, consensus sequence F(S/N)R(I/L/V)(S/T)S, Abdel-Ghany et al., 2003) appears less consistent. Human and porcine CLCA2 show such an IBM within the vWA domain (human aa480-485: FSRISS (Patel, Brett & Holtzman, 2009), pig aa480-485: FSRISS, Fig. S1) while it is absent from chicken, quail, ostrich, feline and murine CLCA2 (Fig. S1). Another IBM motif is lacking in the C-terminal cleavage product of birds and pigs whereas such a motif was found in human, feline and murine CLCA2 (human aa480-485: FSRISS (Patel, Brett & Holtzman, 2009), cat aa741-746: FSRVSS, Fig. S1, mouse aa740-745: FSRVSS (Patel, Brett & Holtzman, 2009), Fig. S1).

Noteworthy, aCLCA2 amino acid (aa) sequences are 921 to 930 aa long and therefore shorter than all of their mammalian homologues with 942 to 944 aa (Fig. 2). In addition, aCLCA2 do not contain a predicted glycosylation site after the predicted transmembrane domain (Fig. 2).

A phylogeny based on galline, quail, ostrich, feline, porcine and murine CLCA2 protein sequences revealed a monophyletic aCLCA2 group separate from maCLCA2 (Fig. S6).

The gCLCA2 protein shares many biochemical properties with mammalian CLCA2 proteins

Posttranslational cleavage

The cleavage of a precursor protein into a larger N- and a shorter C-terminal subunit belongs to the conserved properties of all maCLCA proteins (Patel, Brett & Holtzman, 2009). It is thought to be mediated by the zinc-binding HExxH motif in the N-CLCA domain, cleaving the protein at a canonical cleavage site (Bothe et al., 2012; Lenart et al., 2013; Pawłowski et al., 2006; Yurtsever et al., 2012). Consistently, canonical HExxH motifs and putative proteolytic cleavage sites were identified in aCLCA2 proteins (Fig. 2). The predicted cleavage was verified in vitro for gCLCA2 as a putative avian prototype by immunoblot analysis of lysates from heterologously transfected HEK293 cells. A band consistent with a precursor protein at approx. 145 kilodalton (kDa) and a band consistent with the C-terminal cleavage product of approx. 57 kDa were detected (Fig. 3), suggesting that posttranslational cleavage of CLCA2 also occurs in chicken.

N-glycosylation and cleavage in the medial Golgi

maCLCA2 proteins are multiple N-linked glycosylated (Braun et al., 2010; Elble et al., 2006; Gruber et al., 1999; Plog et al., 2012b). Consistently, our in silico analyses predicted five N-linked glycosylation sites for aCLCA2 (Fig. 2). To verify this prediction experimentally, lysates from gCLCA2- transfected HEK293 cells were treated with endoglycosidases endo H and PNGase F and immunoblotted to identify the kind and extent of glycosylation. The approx. 145 kDa precursor protein was sensitive to endo H and PNGase F treatments (Fig. 4), resulting in a size shift from approx. 145 kDa to approx. 130 kDa, suggestive of an immature high mannose-type glycosylation pattern. In contrast, the C-terminal cleavage product was resistant to endo H treatment but sensitive to PNGase F treatment, as suggested by a reduction in size from approx. 57 kDa to approx. 54 kDa (Fig. 4). Our results indicate that virtually all predicted sites might be glycosylated, presuming a molecular weight of approx. 3 kDa per glycosylation (Pult et al., 2011). Furthermore, the complex high mannose-rich glycosylation pattern of the C-terminal cleavage products, in contrast to the immature glycosylated precursor protein, suggests cleavage of gCLCA2 early in the medial Golgi, similar to what has been observed for the murine CLCA2 (Braun et al., 2010). Thus, glycosylation and cleavage in the medial Golgi also appear as conserved traits.

Anchoring in the plasma membrane

Identically to its maCLCA2 orthologues, a TM domain in the C-terminal subunit was predicted for aCLCA2, which anchors the protein to the plasma membrane (Fig. 2; Braun et al., 2010; Elble et al., 2006; Gruber et al., 1999; Plog et al., 2012b). Consistently, a prominent green autofluorescent EYFP signal was detected along the plasma membrane of HEK293 cells transfected with the gCLCA2 construct that contained an EYFP tag downstream to the TM domain (Figs. 5A and Figs. 5B1). A coinciding signal (Fig. 5B3) was found by immunocytochemistry using the anti-gC2 antibody directed against the fn3 domain of the protein, which is located upstream to the C-terminal transmembrane domain (Fig. 5B2). This is in contrast to the diffuse green signal of cytosolic EYFP protein (Fig. 5C). Therefore, anchorage of the CLCA2 protein in the cell membrane via a TM domain seems also conserved with mammals.

aCLCA2 is expressed in stratified squamous epithelia of skin and mucous membranes

The tissue and cellular expression patterns of gCLCA2 were examined at the mRNA- and protein-levels. gCLCA2 mRNA was detected by RT-qPCR in all tested locations of the skin (back, foot, wattle, ball of the foot, proctodeum) and skin appendages (feather follicle, beak, uropygial gland) as well as in organs with keratinizing mucosal membranes, such as the nose, pharynx, esophagus, crop, and proctodeum (Fig. 6, Table S2). Additionally, gCLCA2 mRNA was found in the trachea, cecum, kidney, bursa of Fabricius, thyroid gland, sciatic nerve, eye and in the liver (Fig. 6, Table S2). gCLCA2 protein was exclusively detected immunohistochemically in keratinocytes of the skin, skin appendages and keratinizing mucosal membranes of the larynx, esophagus and crop (Fig. 7A, Fig. S3). gCLCA2 was localized in all layers of the stratified epithelium similar to the epithelial cell marker cytokeratins (Fig. 7B). At the subcellular level, signals consistently appeared as multiple, evenly distributed dots within the cytosol, with no specific signal enrichment detected at the plasma membrane. However, we failed to detect the CLCA2 protein in other tissues with notable gCLCA2 mRNA presence (Fig. 6). Abundant expression in epidermal keratinocytes appears as a consistent hallmark of CLCA2 proteins, as verified in chicken, turkeys, quails, ostriches, cats, pigs and mice (Fig. 8).