Arhgap17, a RhoGTPase activating protein, regulates mucosal and epithelial barrier function in the mouse colon

Coordinated regulation of the actin cytoskeleton by the Rho GTPase family is required for the maintenance of polarity in epithelial cells as well as for their proliferation and migration. A RhoGTPase-activating protein 17 (Arhgap17) is known to be involved in multiple cellular processes in vitro, including the maintenance of tight junctions and vesicle trafficking. However, the function of Arhgap17 has not been studied in the physiological context. Here, we generated Arhgap17-deficient mice and examined the effect in the epithelial and mucosal barriers of the intestine. Reporter staining revealed that Arhgap17 expression is limited to the luminal epithelium of intestine. Arhgap17-deficient mice show an increased paracellular permeability and aberrant localization of the apical junction complex in the luminal epithelium, but do not develop spontaneous colitis. The inner mucus layer is impervious to the enteric bacteria irrespective of Tff3 downregulation in the Arhgap17-deficient mice. Interestingly however, treatment with dextran sulfate sodium (DSS) causes an increased accumulation of DSS and TNF production in intraluminal cells and rapid destruction of the inner mucus layer, resulting in increased severity of colitis in mutant mice. Overall, these data reveal that Arhgap17 has a novel function in regulating transcellular transport and maintaining integrity of intestinal barriers.


Identification of Arhgap17 Gt(RRN062)Byg allele and generation of Arhgap17 mutant mice
Gene trap ES cell line, Gt(RRN062)Byg (129P2/OlaHsd origin), was obtained from International Gene Trap Consortium. Using genomic DNA of ES cells as a template, the exact structure of Arhgap17 Gt(RRN062)Byg allele was identified by PCR sequencing. For 5' junction identification, 5'-GACCGAAGTCCTCAGTGAAGATC-3' (forward) specific for exon 2 of Arhgap17 and 5'-GGGAAAGAGGGCTCTGTCCT-3' (reverse) for gene trap vector (pGT0lxf) were used for PCR primers. For 3' junction identification, 5'-CTCACCGGCTCCAGATTTATCA-3' (forward) for pGT0lxf and 5'-CGTCAGAGCTCTAGGCTGGA-3' (reverse) for Arhgap17 were used for PCR primers. PCR sequencing of Arhgap17 Gt(RRN062)Byg allele of ES cells revealed that splicing acceptor from gene trap vector was inserted into intron 2 of Arhgap17 (Supplementary Figure 1A). Sequencing of RT-PCR product of gene trap allele showed that exon 2 of Arhgap17 was followed by En2 coding leader in gene trap vector as expected, which would code for fusion protein tagging with 31 amino acids of Arhgap17 N-terminal region.
ES cells were injected into C57BL/6J-derived blastocysts to generate chimeric mice and chimeric mice were further bred to C57BL/6J female to obtain heterozygous mice. Heterozygous mice were maintained congenic on C57BL/6J background more than 10 generations. Genotypes were determined by PCR with three primers (Supplementary Figure 1A): Arhgap17-S1, 5'-GAGATGCATGATGGGATTGCG-3' (forward); Arhgap17-AS1, 5'-CGTCAGAGCTCTAGGCTGGA-3' (reverse); En2-AS1, 5'-CCGGAGCGGATCTCAAACTC-3' (reverse). Arhgap17-S1 and Arhgap17-AS1 amplify the wild-type allele and Arhgap17-S1 and En2-AS1 3 amplify the mutant allele, respectively. Genotypes of offspring were determined by PCR followed by agarose gel electrophoresis (Supplementary Figure 1B). Homozygous mutant mice with Arhgap17 Gt(RRN062)Byg allele (here it named Arhgap17 m ) were produced at Mendelian ratios from heterozygous intercrosses and developed normally. RT-PCR analysis using three primers (Supplementary Figure 1A) to detect endogenous Arhgap17 and mutant mRNA expression showed that endogenous Arhgap17 mRNA was not expressed in tissues of homozygous mice (Supplementary Figure   1C). Whether any Arhgap17 expression occurred by alternative splicing from Arhgap17 m allele, qRT-PCR was performed using far downstream primers for exon 11 (forward) and exon 13 (reverse). The mRNA levels were approximately one-half in the heterozygous mice and less than 1% in homozygous mutant mice compared to wild-type mice (Supplementary Figure 1D), respectively. Moreover, immunoprecipitation-western blotting exhibited no protein band around 95 kD from the homozygous mutant mice (Supplementary Figure 1E), suggesting that Arhgap17 m allele is likely a null allele. Then, homozygous mutant (Arhgap17 m/m ) mice are stated as arhgap17-deficient mice.

Cell proliferation and apoptosis in the intestine
Tissues were labeled with 10 mM BrdU (Sigma-Aldrich) for 2 hours and flushed with PBS before collection of tissues. The paraffin sections were immunostained with a BrdU staining kit or a TUNEL staining kit as described previously. 2 BrdU or TUNEL stained slides were examined under ECLIPSE 80i light microscope (Nikon) and analyzed with NIS-Elements BR 3.2 imaging software. Twenty crypts per mouse were assessed for analysis.

Histological scoring after DSS treatment
The severity of mucosal injury was graded on a scale of 0 to 3: grade 0, normal; grade 1, partial destruction of crypts; grade 2, complete loss of crypts; and grade 3, complete loss of crypts and epithelial cells. 1,2 The histological score was determined by multiplying the portion of injured surface by the grade of severity. Two different regions of proximal, middle and distal colon were used for histological scoring, and the histological scores from individual mice were determined by adding all six values. Histological scoring was performed in a blinded fashion.

LacZ staining
The Arhgap17 expression in tissues was determined by β-galactosidase (LacZ) staining. For whole mount LacZ staining, embryos were fixed in 0.25% glutaraldehyde, 5 mM EGTA and 2 mM MgCl 2 in PBS for 10 minutes at 4°C. After rinsing with detergent solution (2 mM MgCl 2 , 0.02% Igepal CA-630, 0.01% sodium deoxycholate in PBS) three times, embryos were stained with X-gal staining solution 1 for 2 to 3 hours at 37°C and examined under dissecting microscope. To analyze the endogenous Arhgap17 expression in adult, tissues collected from arghap17-deficient mice were fixed, frozen-sectioned and stained with X-gal as described previously. 1 The X-gal-stained tissues were counterstained with 0.1× (v/v) Harris hematoxylin (Sigma-Aldrich) followed by eosinY staining not to disturb the X-gal staining.
The image was visualized with light microscope.

Migration of epithelial cells in the intestine
Tissues were labeled with 10 mM BrdU for 30 hours and flushed with PBS before collection of tissues. The paraffin sections were immunostained with a BrdU staining kit and cell migration was analyzed by measuring the length of far upstream BrdU-stained epithelial cells from the crypt using NIS-Elements BR 3.2 imaging software (Nikon). Twenty villus or ten crypts per mouse were assessed for analysis. Scale bar, 50 μm, n = 8. (B) Total RNA was extracted from the distal colon of indicated mice before