Introduction
The endothelial blood–brain barrier (BBB) in the central nervous system (CNS) microvessels protects the CNS from changes in the blood, thus ensuring CNS homeostasis, which is a prerequisite for proper function of CNS neurons. The anatomical basis of the BBB is represented by unique intercellular tight junctional (TJs) complexes formed between BBB endothelial cells, inhibiting paracellular diffusion of water-soluble molecules [
1,
2]. In addition to sealing the paracellular pathway, BBB endothelial cells keep out unwanted compounds from the brain by their lack of fenestrae and low pinocytotic activity [
3]. To meet the high demand of the neuronal cells for energy and to drive efflux of toxic metabolites from the CNS, BBB endothelial cells express unique combinations of enzymes and transporters [
4].
The TJs of the BBB are distinct from those of other endothelial cells with respect to their high complexity and continuity of junctional strands that rather resemble those of the epithelial cells [
5]. Freeze fracture studies demonstrated that in contrast to the peripheral vasculature, which shows predominantly E-face associated TJs, TJs of the BBB are mainly P-face associated [
6]. P-face association of TJs-strands has been shown to correlate with barrier properties of the BBB [
7]. Additionally, the protein expression signature at the TJs is very unique, being composed of occludin, junctional adhesion molecules and members of the claudins family [
2]. The claudins are integral membrane proteins exclusively found in the TJs of all epithelia and endothelial cells, and they are essential and sufficient in establishing TJs and thus paracellular diffusion barriers [
8]. In mammals, there are presently 27 known members of the claudin family, which exert different functions and present tissue and developmental stage specific expression patterns [
9]. Thus, a different combination of claudins establishes specific TJs, which ultimately regulates their tightness [
10]. It is known that some claudins can tighten the cleft between two adjacent cells, such as claudin-1 and claudin-3 [
11,
12], while others form paracellular pores that contribute to a controlled passage of ions and water through the TJs, e.g. claudin-2 and claudin-16 [
13,
14]. However, for some claudins, a precise function is not yet known. In the BBB endothelium, it was suggested that claudin-3, claudin-5 and claudin-12 contribute to the tightness of this barrier [
15]. Claudin-5 is the most predominant claudin expressed in the BBB TJs and is essential for the establishment of BBB TJs during development, since its absence leads to perinatal death of mice due to increased permeability of the BBB to small molecular tracers [
15,
16]. Moreover, with an inducible knock-down mouse model, it was seen that suppression of claudin-5 in the TJs leads to the disruption of BBB integrity and ultimately to seizures and behavioral changes [
17], which demonstrates the importance of claudin-5 in the maintenance of this paracellular barrier in the adult. However, claudin-5-deficient mice display morphologically intact TJs suggesting the presence of other proteins localized to BBB TJs. Claudin-3 was described to be involved in the induction and maintenance of the BBB [
18,
19]. However, expression of claudin-3 at the BBB TJs has been repeatedly questioned [
20,
21] and more recently, its absence from the BBB endothelium was confirmed by employing a combination of methods including immunostaining, Western Blotting and single cell RNAseq analysis (scRNAseq) of the brain endothelium [
1,
22]. scRNAseq of brain endothelial cells has also confirmed lack of expression of the TJ sealing claudin-1 at the BBB, as previously described [
4,
22‐
24]. As claudin-5 by itself induces E-face associated TJs this has raised the question if another member of the claudin family could contribute to the formation of P-face associated TJs in the BBB [
16]. Expression of claudin-12, an unusual member of the claudin family, has additionally been described at the BBB [
15] and its expression in brain endothelial cells was recently confirmed by us [
1]. Claudin-12 is an atypical member of the claudin family because it does not have a PDZ binding motif, which mediates the interaction of claudins with the cytoskeleton, by allowing the binding to the intracellular scaffolding proteins ZO-1, ZO-2 and ZO-3 [
25]. Thus, its potential contribution to BBB TJs still remains unknown.
To answer this question, we generated a claudin-12lacZ/lacZ C57BL/6J mouse, with a lacZ cassette inserted in the open reading frame (ORF) of claudin-12, which allows us not only to use it as a reporter gene for claudin-12 expression, but also to take advantage of the null allele to investigate the function of this protein. We observed broad expression of claudin-12 in numerous tissues and most prominently in smooth and striated muscle cells. Within the brain, expression of claudin-12 was detected in many different cell types with most prominent expression in neurons and astrocytes. Determining the subcellular localization of claudin-12 protein was prohibited due to the lack of antibodies specifically and selectively detecting claudin-12 protein. Nonetheless, our study rules out an essential role for claudin-12 in regulating BBB integrity under non-inflammatory or neuroinflammatory conditions, as we did not observe any aggravation of experimental autoimmune encephalomyelitis (EAE), an animal model of multiple sclerosis (MS). Yet, an in-depth phenotypic study of the claudin-12lacZ/lacZ C57BL/6J mice points to possible roles of claudin-12 in select neurological functions and, more prominently, in cardiovascular functions.
Materials and methods
Generation of claudin-12lacZ/lacZ C57BL/6J mice
Three gene-targeted embryonic stem cell (ES) clones (Cldn12_13208A-A6, Cldn12_13208AE2 and Cldn12_13208AF12) from KOMP project number KO1756 were requested from the KOMP repository at UC Davis. The mutation was generated on the Velocigen platform by Regeneron Pharmaceuticals (Tarrytown, NJ). The insertion of Velocigene cassette ZEN-Ub1 created a deletion of 760 bps between positions 5507663–5508422 of Chromosome 5 (Genome Build37) (
J:136110). This deletion replaces the entire coding sequence residing only within in exon 5 with a lacZ cassette and a LoxP flanked neomycin-selection cassette. The deletion includes 28 bps upstream of the translational start site and 21 bps downstream of the stop codon. Parental ES cells are VGB6, which are derived from the C57BL/6NTac inbred mouse strain. The sequence of this allele is available here:
https://www.i-dcc.org/imits/targ_rep/alleles/38187/escell-clone-genbank-file. ES clone Cldn12_13208A-A6 was injected into BALB/c blastocysts at the Institute of Laboratory Animal Science, University of Zurich. Five chimeric males were born and mated to C57BL/6J albino females. One of the chimeric males produced a single black male out of 40 offspring. The single black male was mated to C57BL/6J females and the offspring genotyped by PCR using the following primers: Cldn12_GT_SD (5′-CTCCTAGCCTCATCCGACTGAAACG-3′), Cldn12_ ΔGT3_TDF (5′-CTGCTGTTCGTTTGGTATTGTGCATG-3′) and PGK 3′UTR FW1 (5′-GGGTGGGATTAGATAAATGCCTGCTCT-3′). PCR-cycling conditions were: 94 °C for 4 min., and 35 repeats of 94 °C for 30 s., 62 °C for 30 s., 72 °C for 60 s., followed by 94 °C for 4 min. The wild-type (WT) allele was detected by a 586 bp band, while the neo-allele gave rise to a 775 bp band. To avoid any influence of the ubiquitin-C promoter driving the neomycin-resistance gene on the claudin-12 phenotype, the LoxP flanked neo-cassette was deleted by crossing claudin-12 neo heterozygous mice to ZP3-Cre (Tg(ZP3-cre)93Knw) mice expressing cre in the female germ line (oocytes) [
26] (a gift from Pawel Pelczar). After deletion of the neo-cassette, claudin-12 mutated mice were genotyped by PCR using the following primers: Cldn12_FW2 (5′-TTTCTGATAGGATGGGTAGGTGGT GG-3′), Cldn12_REV2 (5′-CAGGCCCGTGTAAATCGTCAGGT-3′), LacZ-5′REV1 (5′-GAGCGAGTAACAACCCGTCGGATTCT-3′). PCR-cycling conditions were: 94 °C for 4 min., and 35 repeats of 94 °C for 30 s., 62 °C for 30 s., 72 °C for 60 s., followed by 94 °C for 4 min. The WT allele was detected by a 425 bp band, while the claudin-12-lacZ-allele gave rise to a 607 bp band.
German Mouse Clinic
The claudin-12
lacZ/lacZ mouse line was backcrossed to C57BL/6J four times prior to the German Mouse Clinic (GMC) Primary Screen. Health status was confirmed to be specific pathogen-free according to FELASA recommendations. Using the platform established at the GMC, we performed a primary phenotypic analysis of a total of 60 (15 mice each of each genotype and sex) [
27‐
29] exactly as described in the Additional file
6.
Mouse housing
Mice were housed in individually ventilated cages (IVC) under specific pathogen-free conditions at 22 °C and 55% relative humidity with free access to chow and water. Animal procedures executed were approved by the Veterinary Office of the Canton Bern (permit no. BE55/09, BE42/14, BE72/15, BE31/17). At the GMC mice were maintained in IVC cages with water and standard mouse chow (Altromin no. 1314) according to the GMC housing conditions and German laws. All tests performed at the GMC were approved by the responsible authority of the district government of Upper Bavaria.
Experimental autoimmune encephalomyelitis (EAE)
Active EAE was induced in 8–12-week-old female claudin-12
lacZ/lacZ C57BL/6J mice, claudin-12
lacZ/+ C57BL/6J mice and their WT C57BL/6J littermates exactly as previously described [
30,
31]. Weights and clinical disease activity were assessed twice daily and scored as follows: 0, healthy; 0.5, limp tail; 1, hind leg paraparesis; 2, hind leg paraplegia; 3, hind leg paraplegia with incontinence implementing the 3R rules as described in [
32]. Two independent experiments were performed.
Isolation of brain microvessels
Primary mouse brain microvessels were isolated from WT and claudin-12
lacZ/lacZ C57BL/6J mice as previously described in detail [
33]. The only modification to this previous protocol was that instead of the final plating step, microvessels were incubated in a red blood cell lysis buffer (0.83% ammonium chloride and Tris–HCl, pH = 7.5), for 5 min, RT. After two washing steps, microvessels were lysed in HES lysis buffer (10 mM HEPES, 1 mM EDTA solution, 250 mM sucrose solution), in the presence of protease inhibitor cOmplete ULTRA Tablets, Mini, EDTA-free, EASYpack (1 tablet/10 mL) (Roche Diagnostics, Mannheim, Germany), and kept at − 20 °C.
SDS-PAGE
In accordance to the isolation of brain microvessels, muscle tissue from WT and claudin-12lacZ/lacZ C57BL/6J mice was also lysed in HES lysis buffer (10 mM HEPES, 1 mM EDTA solution, 250 mM sucrose solution), in the presence of protease inhibitor cOmplete ULTRA Tablets, Mini, EDTA-free, EASYpack (1 tablet/10 mL) (Roche Diagnostics, Mannheim, Germany). Protein concentration was measured using the Pierce™ BCA Protein Assay Kit (Thermo Scientific™ Pierce™ Protein Biology, Waltham, USA), according to the manufacturer’s instructions. 20 μg of each sample were boiled at 95°, for 5 min, and loaded onto a 10% SDS-polyacrylamide gel and transferred to a nitrocellulose membrane (Amersham Protan, GE Healthcare, United Kingdom), using a Trans-Blot Turbo transfer system (BioRad Laboratories, Hercules, CA, USA), according to the manufacturer’s instructions. Membranes were blocked with Rockland Buffer (Rockland, Limerick, PA, USA) for 1 h at RT and incubated overnight at 4 °C with three different rabbit anti-mouse claudin-12 antibodies (IBL, cat. no 18801; Abcam cat. no ab107061; Invitrogen, cat. no 388200), mouse anti-mouse β-actin (Merck, cat. no A5316), rabbit anti-α-tubulin (Abcam, cat. no ab4074), rabbit anti-claudin-5 (Thermo Fisher Scientific, cat. no 34-1600), rabbit anti-occludin (Thermo Fisher Scientific, cat. no 71-1500), or rabbit anti-ZO-1 (Thermo Fisher Scientific, cat. no 61-7300). On the following day, membranes were washed and incubated with secondary antibodies goat-anti-rabbit Alexa Fluor® 680 (Thermo Fisher Scientific, cat. No A21109) and goat anti-mouse IRDye® TM 800 (Rockland Immunochemicals, cat. no 605-732-125), for 1 h at RT. Proteins were detected by the Odyssey near infrared imaging system and software (LI-COR Biotechnology, Lincoln, NE, USA). Band intensity for the three independent experiments was quantified using the ImageJ software (NIH, Bethesda, MD, USA) and normalized against β-actin.
Quantitative real-time PCR analysis (qRT-PCR)
RNA was extracted from the heart tissue of WT and claudin-12
lacZ/lacZ C57BL/6J mice by using the High Pure RNA Isolation kit (Hoffman-La Roche, Basel, Switzerland). cDNA was obtained from each sample’s total isolated RNA with the SuperScript III First-Strand Synthesis System (Invitrogen, Carlsbad, CA, USA) and the qRT-PCR was done as previously described [
4]. The pairs of primers that were used in this study are the following: 5′-CTGAGTTCACTAAGCTGACTTTGG-3′ (sense primer within exon 3) and 5′-CCTGTCTGCGCCTCTGAT-3′ (anti-sense primer within exon 4) for the 5´prime untranslated region (UTR) of claudin-12 mRNA; 5′-TGCTTGGAGAAACGCTGATT-3′ (sense primer within exon 4) and 5′-GTGGCTGCGTGGACATCT-3′ (anti-sense primer within the ORF integral to exon 5) for the open reading frame (ORF); 5′-TGCTTGGAGAAACGCTGATT-3′ (sense primer within exon 4) and 5′-GTCTGTCCTAGCTTCCTCACTG-3′ for lacZ detection. ΔC
T value was obtained (average C
T value of target gene − average C
T value of S16) and relative expression values of three independent experiments were calculated according to the comparative 2
−ΔΔCt method (ΔΔC
T = Δ C
T sample − Δ C
T WT).
LacZ staining of tissue sections
WT, Tie2-lacZ and claudin-12
lacZ/+ C57BL/6J mice were anesthetized with Isoflurane Baxter (Arovet, Dietikon, Switzerland) and perfused with 1% paraformaldehyde (PFA). As positive controls, Tie2-lacZ (B6.FVB-Tg (TIE2-lac)182Sato) transgenic mice were used [
34]. Brain, retina, liver, heart, tongue, skeletal muscle, intestine and kidney were removed, washed twice in PBS (pH = 7.4) and fixed with 1% PFA + 5 mM EDTA + 2 mM MgCl
2, for 4 h at 4 °C. The tissue was next incubated in 18% sucrose, overnight at 4 °C, prior to embedding in Tissue-Tek O.C.T. compound (Sakura Finetek, The Netherlands) and snap-freezing [
35]. 20 µm cryosections were cut from these tissues, air dried at RT for 4 h, and fixed in 1% PFA + 5 mM EDTA + 2 mM MgCl
2 for 3 min. Afterwards, a rehydration step was performed with PBS (pH = 7.4). For LacZ staining, sections were incubated overnight with 0.1% X-Gal at 37 °C in the dark. Next, sections were washed once in distilled water, incubated for 1 min in 1% Neutral Red (Sigma-Aldrich, St. Louis, Missouri, USA) and differentiated in distilled water. Then, cryosections were dehydrated in consecutive steps in EtOH 75%, EtOH 85%, EtOH 95% and EtOH 100%, and finally in Xylol before mounting in Entellan
® (Merck Millipore, Darmstadt, Germany). Three independent experiments were performed and were analysed using a Nikon Eclipse E600 microscope connected to a Nikon Digital Camera DXM1200F with the Nikon NIS-Elements BR3.10 software (Nikon, Egg, Switzerland). Images were processed and mounted using Adobe Illustrator software (Adobe Systems, CA, USA).
Immunofluorescence staining of tissue sections
WT and claudin-12
lacZ/lacZ C57BL/6J mice were anesthetized with Isoflurane Baxter (Arovet, Dietikon, Switzerland) and perfused with 1% PFA. Brains and liver were removed, embedded in Tissue-Tek
® OCT compound (Sakura Finetek, The Netherlands) and snap-frozen. Cryosections were cut at 6 μm or 10 μm thickness and fixed in either ice cold acetone or 2% PFA for 10 min and air-dried. Cryosections were stained as described before [
36,
37]. Sections were incubated for 1 h, RT, with the following primary antibodies: two different rabbit anti-mouse claudin-12 antibodies (IBL, cat. no 18801; Invitrogen, cat. no 388200), rat anti-PECAM-1 (in house, clone Mec13.3), rabbit anti-β-galactosidase (Thermo Fisher Scientific, cat. no A-11132), rat anti-CD140b (eBioscience, cat. no 14-1402-82), mouse anti-GFAP (Sigma-Aldrich, cat- no G3893), mouse anti-NeuN (clone A60, Millipore, cat. no MAB377), rabbit anti-fibronectin (DAKO, cat. no A0245) and biotinylated goat anti-mouse IgG (Vector, cat. no BA-9200). After the washing steps, the sections were incubated with the following secondary antibodies: goat polyclonal IgG Cy3 anti-rabbit (Jackson ImmunoResearch, cat. no 11-165-144), donkey anti-rat IgG (H+L) Alexa Fluor 488 (Thermo Fisher Scientific, cat. no A-21208), donkey anti-rabbit IgG (H+L) Alexa Fluor 488 (Thermo Fisher Scientific, cat. no A-21206), donkey anti-mouse IgG (H+L) Alexa Fluor 488 (Thermo Fisher Scientific, cat. no A-32766), Streptavidin-Cy3 (Vector, cat. no SA-1300) and with DAPI (1:1000, Thermo Fisher Scientific, Carlsbad, CA, USA), for 1 h, RT. Fluorescence stainings were performed in at least three independent experiments and were analysed using a Nikon Eclipse E600 microscope connected to a Nikon Digital Camera DXM1200F with the Nikon NIS-Elements BR3.10 software (Nikon, Egg, Switzerland). Images were processed and mounted using Adobe Illustrator software (Adobe Systems, CA, USA).
Statistics
Statistical analysis was performed using GraphPad Prism 6.0 software. To compare two groups, an unpaired t-test with Welch’s correction was performed. For the analysis of the EAE experiments, a Mann–Whitney U-test was performed. Results are shown as mean ± SD and a p < 0.05 was considered significant. If not stated otherwise, data generated by the German Mouse Clinic was analyzed using R (Version 3.2.3). Tests for genotype effects were made by using t-test, Wilcoxon rank sum test, linear models, or ANOVA and posthoc tests, or Fisher’s exact test depending on the assumed distribution of the parameter and the questions addressed to the data. A p-value < 0.05 has been used as level of significance; a correction for multiple testing has not been performed.
Discussion
The BBB plays a crucial role in maintaining CNS homeostasis, by preventing free diffusion of solutes into the brain, with the BBB TJs playing an essential role in sealing the paracellular cleft between the BBB endothelial cells. The claudin family of TJs proteins are crucial in establishing and maintaining TJ function at the BBB [
2]. During recent years, there is an increasing understanding of the role of each individual claudin in contributing to tissue specific TJ functions [
40]. Thus, to understand BBB TJ function it is essential to understand which claudins are expressed in brain endothelial cells and localized to BBB TJs. Bearing in mind the important role each claudin plays in maintaining barrier function under physiological conditions and how disturbance of their localization at the TJs can contribute to barrier breakdown, it is urgent to fully clarify the molecular claudin makeup of BBB TJs. Claudin-5 is the most enriched claudin in the vascular endothelium of the BBB. Its absence is associated with BBB integrity breakdown, an early feature in the development of MS [
41], and it is also implied in the pathology of stroke, traumatic brain injury and schizophrenia [
17,
42]. However, it was seen that claudin-5 knock-out mice still present morphologically intact TJs [
16]. As claudins are essential and sufficient for TJ induction this means that another claudin that is present at the BBB TJs contributes to TJs formation in the absence of claudin-5. In the past years, claudin-1 and claudin-3 were repeatedly suggested to be part of the BBB TJs [
3,
18]. However, our recent studies have shown that these claudins are not expressed in brain endothelial cells and can thus not contribute to BBB TJs [
1,
23]. Claudin-12 is an additional claudin, expression of which has been described in BBB TJs [
15]. Claudin-12 is an unusual claudin, lacking a PDZ binding motif in its C-terminal domain, which is required to connect claudins via scaffolding proteins such as ZO-1 to the cytoskeleton [
25]. The possible function of claudin-12 at the BBB has remained unknown to date.
To tackle this question, in the present study we created claudin-12
lacZ/+ and claudin-12
lacZ/lacZ C57BL/6J mice allowing to explore claudin-12 expression as well as the impact of claudin-12 deletion on BBB TJ integrity. Making use of the lacZ reporter allele, we found a characteristic punctate β-galactosidase activity in numerous cells throughout the entire CNS. Our findings are in accordance to the staining pattern previously observed by the International Mouse Phenotyping Consortium (IMPC) when analysing reporter gene expression of a lacZ allele of different claudin-12 mutant mice (Cldn12
tm1b(EUCOMM)Wtsi) by whole mount lacZ staining of adult Cldn12
tm1b(EUCOMM)Wtsi mice (
http://www.mousephenotype.org/data/search/impc_images?kw=%22cldn12%22). This previous dataset thus underscores ubiquitous expression of claudin-12 throughout the brain and furthermore shows that claudin-12 expression is enriched in the grey matter of the spinal cord.
In our present study, vascular expression of claudin-12 based on β-galactosidase activity was clearly visible in the smooth muscle cell layer of larger vessels in the brain parenchyma, as well as in the meninges rather than in the endothelial cells of these vessels. In turn, it could be specifically detected by endothelial-cell-specific lacZ activity in Tie2-lacZ reporter mice. Furthermore, we failed to detect lacZ activity in endothelial cells of microvessels forming the BBB in our brain tissue sections. This suggests that if claudin-12 is expressed in brain endothelial cells, this is at a very low level. This observation was confirmed by our immunofluorescence stainings, which allowed to localize expression of the β-galactosidase reporter in sparse PECAM-1-positive brain endothelial cells and few PDGFRβ-positive pericytes. At the same time, positive immunostaining for the β-galactosidase reporter was found in GFAP-positive astrocytes and most abundantly in NeuN-positive neurons further underscoring low but ubiquitous expression of claudin-12 in the brain of the mouse.
In fact, previous scRNAseq transcriptomic analysis of brain vessels isolated from developing (P7) and adult mice confirms low levels of expression of claudin-12 mRNA in a wide range of vascular and CNS vessel associated cell types (
https://markfsabbagh.shinyapps.io/vectrdb/ and
http://betsholtzlab.org/VascularSingleCells/database.html) [
22,
38,
43]. Average reads for claudin-12 mRNA were found to be low and similar in endothelial cells along the vascular tree and were found to be higher in smooth muscle cells compared to brain endothelial cells. In addition, low expression of claudin-12 was also detected in microglial cells, oligodendrocytes, astrocytes and neurons [
22,
24] (
http://www.brainrnaseq.org/). Last but not least, brain endothelial expression levels of claudin-12 were found comparable to those detected using the same methodology in lung endothelial cells. Our present observations underscore low to undetectable expression of claudin-12 in brain endothelium and visible expression in vascular smooth muscle cells, pericytes, astrocytes and neurons. As these cells do not typically form TJs, this suggests that claudin-12 plays a role other than forming classical TJs. Expression data on claudin-12 mRNA in isolated brain microvessels and the detected regulation of claudin-12 mRNA expression must therefore be carefully interpreted as it may reflect regulation of claudin-12 expression in pericytes, smooth muscle cells or even in astrocytes, rather than the brain endothelial cells proper [
44].
To determine the subcellular localization of claudin-12 in the CNS vasculature we made use of commercially available antibodies to claudin-12 and performed side-by-side immunostainings on brain sections of WT and claudin-12lacZ/lacZ C57BL/6J mice. We could not observe any difference in the staining patterns produced by the anti claudin-12 antibodies in the brain sections of WT and claudin-12lacZ/lacZ C57BL/6J mice. This observation therefore prohibits reliable subcellular localization of claudin-12 in the CNS vasculature and suggests that antibodies recognizing claudin-12 cross-react with other molecules present in CNS microvessels. Cross-reactivity of claudin-12 detecting antibodies was confirmed by Western Blots, which showed detection of a similar sized band in tissue samples from claudin-12lacZ/lacZ C57BL/6J mice.
In light of our previous observations that claudin-3 targeting antibodies produce an identical staining pattern in CNS microvessels in claudin-3 deficient and WT C57BL/6J mice [
1], it seems that claudin-targeting antibodies often produce false-positive detection of the respective claudin due to cross-reactivity with other claudins, which may be due to the highly conserved nature of the immunogenic domains of this family of proteins [
45]. Employing presently available claudin-12 antibodies we could therefore not reproduce the prominent immunostaining for claudin-12 originally observed at the BBB in mice when using a home-made anti-claudin-12 antibody [
15]. Lack of reliable antibodies detecting claudin-12 thus also questions previous observations of claudin-12 protein expression in the human brain endothelial cell line hCMEC/D3 [
46]. However, significant expression of claudin-12 mRNA has also been detected in this cell line [
47]. At the same time, transcriptome profiling of freshly isolated cells from the human brain (available at
http://www.brainrnaseq.org) shows even lower expression levels for claudin-12 mRNA in human brain endothelial cells as observed in the mouse, but at the same time confirms expression of claudin-12 in cells other than brain endothelium in the CNS [
48]. Thus, the final confirmation if claudin-12 is expressed at the protein level in CNS blood vessels and if so its precise cellular and subcellular localization within the CNS vascular cells in mouse and man remains to be determined.
In accordance to the available transcriptome profiles on brain cells [
22,
24], we observed that claudin-12 is expressed in various cells of the brain vasculature as well as in astrocytes and neurons. Thus, a functional impact of claudin-12 in the settings of neuroinflammation remains possible. We therefore hypothesized that its absence could still affect BBB integrity. BBB breakdown is a major hallmark of MS and is associated with loss of TJs (summarized in [
2]). It has previously been shown that impaired BBB integrity, e.g. in PECAM-1-deficient mice, aggravates EAE [
49], while endothelial cell-specific ectopic expression of claudin-1 inhibits BBB breakdown during EAE and ameliorates chronic disease [
23]. Therefore, we here compared development of clinical EAE in claudin-12
lacZ/lacZ C57BL/6J mice to that in their WT littermates. We did not observe any differences in clinical EAE between WT and claudin-12
lacZ/lacZ C57BL/6J mice. We also did not detect any differences in BBB integrity or immune phenotype, e.g. upregulation of adhesion molecules in claudin-12
lacZ/lacZ C57BL/6J mice versus WT C57BL/6J mice during EAE. Thus, although claudin-12 is expressed in various cells of the CNS including vascular cells, its absence does not affect BBB integrity during autoimmune neuroinflammation.
Conclusions about claudin-12 expression in the present study rely on available mRNA data sets [
22,
24,
38] and on the novel information provided in the present study based on claudin-12 lacZ reporter gene expression. Our present observations on claudin-12 expression using our claudin-12 reporter mouse allowed for robust detection of β-galactosidase staining in various tissues, such as skeletal muscle, liver, intestine and kidney, suggesting that while claudin-12 is expressed in different cells of the CNS, its expression is stronger in the periphery. These findings are consistent with the previous observations on claudin-12 expression in the Cldn12
tm1b(EUCOMM)Wtsi mice (
http://www.mousephenotype.org/data/search/impc_images?kw=%22cldn12%22).
Due to the presence of claudin-12 in a wide range of organs, we decided to perform a systematic multiparameter phenotypic analysis of our claudin-12lacZ/lacZ C57BL/6J mouse and compare it to WT littermates.
The mutant mice showed some behavioral impairments. While it is currently not possible to draw conclusions concerning the exact nature of these alterations, our findings are bolstered by the reported behavioral changes in the aforementioned IMPC line (Cldn12tm1b(EUCOMM)Wtsi). Although there are differences between the two lines and the observed effects are subtle, behavioral alterations now described in both suggest that loss of claudin-12 likely affects brain function to some degree, most probably due to its lack of expression in neurons. It is therefore of interest to better understand the potential role of claudin-12 in the different cells of the brain. Claudin-12 has in addition been shown to be expressed in the inner ear as well as well as in retina, as confirmed by us in the present study [
50,
51]. Thus, the subtle alterations in auditory and retinal function might also be interesting for further analysis. Moreover, it is important to highlight that some morphological and functional alterations of the heart were detected in claudin-12
lacZ/lacZ C57BL/6J mice in the cardiovascular screen. Here, additional studies are required to elucidate the role and functional implications of claudin-12 in the murine cardiac muscle. Also, the fact that claudin-12 is strongly expressed in skeletal muscle and is reported to be involved in vitamin D-dependent calcium absorption [
52] asks for further clarifications in the role of this protein in the muscle, although under standard conditions no changes have been detected for muscle function analyzed by grip strength.
Acknowledgements
Special thanks go to Dr. Charaf Benarafa (IVI, Mittelhäusern, Switzerland) for help in establishing the claudin-12laz/lacZ mice and for valuable comments to the manuscript. We also owe special thanks to David Miguel Ferreira Francisco for his valuable advice on the correct interpretation of published RNAseq datasets. We thank Therese Périnat, Simone Ebener, Sara Barcos, Claudia Blatti for expert technical assistance, and Albert Witt and Mark Liebi for mouse genotyping. The contributions of the animal caretaker team in our animal facility are gratefully acknowledged.
The members of the German Mouse Clinic Consortium are listed in Declarations.
The German Mouse Clinic Consortium.
Antonio Aguilar-Pimentel1, Thure Adler1, Dirk H. Busch2, Nadine Spielmann1, Kristin Moreth1, Wolfgang Hans1, Oana Amarie1,3, Jochen Graw3, Jan Rozman1,12, Ildiko Radc4,13, Frauke Neff1, Julia Calzada-Wack1, Birgit Rathkolb1,5,12, Eckhard Wolf5, Thomas Klopstock6,7,8,9, Wolfgang Wurst3,7,8,10, Johannes Beckers1,11,12, Manuela Östereicher1, Gregor Miller1, Holger Maier1, Claudia Stoeger1, Stefanie Leuchtenberger1, Valérie Gailus-Durner1, Helmut Fuchs1
1German Mouse Clinic, Institute of Experimental Genetics, Helmholtz Zentrum München, German; Research Center for Environmental Health GmbH, Ingolstädter Landstrasse 1, 85764 Neuherberg, Germany. 2Institute for Medical Microbiology, Immunology and Hygiene, Technical University of Munich, Trogerstrasse 30, 81675 Munich, Germany. 3Institute of Developmental Genetics, Helmholtz Zentrum München, German Research Center for Environmental Health GmbH, Ingolstädter Landstrasse 1, 85764 Neuherberg, Germany. 4Institute of Molecular Psychiatry, Medical Faculty, University of Bonn, Sigmund-Freud-Strasse 25, 53127 Bonn, Germany. 5Ludwig-Maximilians-Universität München, Gene Center, Institute of Molecular Animal Breeding and Biotechnology, Feodor-Lynen Strasse 25, 81377 Munich, Germany. 6Department of Neurology, Friedrich-Baur-Institut, Ludwig-Maximilians-Universität München, Ziemssenstrasse 1a, 80336 Munich, Germany. 7Deutsches Institut für Neurodegenerative Erkrankungen (DZNE) Site Munich, Schillerstrasse 44, 80336 Munich, Germany. 8Munich Cluster for Systems Neurology (SyNergy), Adolf-Butenandt-Institut, Ludwig-Maximilians-Universität München, Schillerstrasse 44, 80336 Munich, Germany. 9German Network for Mitochondrial Disorders (mitoNET). 10Chair of Developmental Genetics, Technical University München-Weihenstephan, c/o Helmholtz Zentrum München, Ingolstädter Landstr. 1, 85764 Neuherberg, Germany. 11Chair of Experimental Genetics, School of Life Science Weihenstephan, Technische Universität München, Alte Akademie 8, 85354 Freising, Germany. 12Member of German Center for Diabetes Research (DZD), Ingolstädter Landstraße 1, 85764 Neuherberg, Germany. 13Present address: Clinic of Neurodegenerative Diseases and Gerontopsychiatry, University of Bonn Medical Center, Bonn, Germany.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.