The IgCAM CLMP regulates expression of Connexin43 and Connexin45 in intestinal and ureteral smooth muscle contraction in mice

ABSTRACT CAR-like membrane protein (CLMP), an immunoglobulin cell adhesion molecule (IgCAM), has been implicated in congenital short-bowel syndrome in humans, a condition with high mortality for which there is currently no cure. We therefore studied the function of CLMP in a Clmp-deficient mouse model. Although we found that the levels of mRNAs encoding Connexin43 or Connexin45 were not or were only marginally affected, respectively, by Clmp deficiency, the absence of CLMP caused a severe reduction of both proteins in smooth muscle cells of the intestine and of Connexin43 in the ureter. Analysis of calcium signaling revealed a disordered cell-cell communication between smooth muscle cells, which in turn induced an impaired and uncoordinated motility of the intestine and the ureter. Consequently, insufficient transport of chyme and urine caused a fatal delay to thrive, a high rate of mortality, and provoked a severe hydronephrosis in CLMP knockouts. Neurotransmission and the capability of smooth muscle cells to contract in ring preparations of the intestine were not altered. Physical obstructions were not detectable and an overall normal histology in the intestine as well as in the ureter was observed, except for a slight hypertrophy of smooth muscle layers. Deletion of Clmp did not lead to a reduced length of the intestine as shown for the human CLMP gene but resulted in gut malrotations. In sum, the absence of CLMP caused functional obstructions in the intestinal tract and ureter by impaired peristaltic contractions most likely due to a lack of gap-junctional communication between smooth muscle cells.


Figure S2
Enrichment of smooth muscle cell layer and villi from intestine.
Western blotting demonstrating the enrichment of the smooth muscle cell layer or villi preparations from intestine using antibodies to pan-Cadherin (a marker for the villi) and an antibody to smooth muscle actin. In the upper panel 10 µg of protein per lane was loaded.
GAPDH and clathrin served as additional loading control. In the lower panel 1 µg of protein was loaded from the smooth muscle layer and 10 µg from villi. Molecular mass markers are indicated at the left of the panels.

Figure S3
Localization of CLMP in the embryonic intestine and ureter.
Localization of CLMP in cross sections of intestine and ureter at different embryonic stages using affinity purified antibody 102 to the extracellular domain of mCLMP. CLMP is primarily localized in the developing smooth muscle layer and more weakly in the mucosa of the intestine and in the developing mesentery (see left panel of the intestine row). In ureter sections CLMP is also primarily expressed in the developing smooth muscle layer and in the surrounding fat tissue. Sections were counter-stained with DAPI. Bar, 50 µm.  with BglII or with NcoI using 5' or 3' probes, respectively. C) PCR analysis of genomic DNA with primers P1 and P2 or P3 and P4 for amplification of the wild type or the mutant allele, respectively. A 459 bp product is generated from wild type allele and a 428 bp product from the targeted allele. D) RT-PCR of RNA extracted from brain tissue reveal the absence of CLMP encoding mRNA in CLMP-deficient mice. Actb amplification was used as control to verify integrity of total RNA isolation and cDNA reverse transcription. E) Western blot of membrane enriched fractions from brain tissue using an antibody to the cytoplasmic domain of CLMP demonstrate the absence of CLMP in mutant mice. A 46 kDa band is revealed in wild type or heterozygous mice. F) Quantitative RT-PCR of intestinal smooth muscle tissue (4 weeks old) from wild type and CLMP knockout mice. A 1050 fold higher value was detected in wild type in comparison to knockout indicating the absence of mRNA encoding Clmp in the CLMP knockout (p=0.0007). Intestinal length from adult CLMP-deficient mice does not differ from wild types.
Intestines of adult mice (mixed genetic background) were dissected; gut lengths were measured and normalized to body weights. A) Examples, B) total intestine length and C) total intestine length related to body length. Numbers in brackets indicate numbers of analyzed animals. SI, small intestine; LI, large intestine; st, stomach; ce, cecum. Scale bar, 1 cm.

Figure S7
Urine composition and increased NGAL in CLMP-deficient mice.

Figure S9
Localization of Connexin45 in the duodenum in in wild type and CLMP-deficient mice.
A and B) Connexin45 plaques are reduced in the circular smooth muscle cell layer of CLMPdeficient duodenum (A). Quantitative data are presented in Figure 7D. Bar, 50 µm. Clmp Cx45 DAPI Cx45 Disease Models & Mechanisms 11: doi:10.1242/dmm. Table S2 Wet weight analysis of organs of the Clmp-deficient mice (mixed background) at P97 -
Significant changes in kidney and the gastrointestinal tract are printed in bold.

Table S1
Configural frequency analysis of recovery of Mendelian ratio in newborn pups (P0.5) from heterozygous-to-heterozygous mating. Animal numbers with a Chi² value > 3.841 are underrepresented. Monitoring was done at noon.