MYPT1 Down-regulation by Lipopolysaccharide-SIAH1/2 E3 Ligase-Ubiquitin-Proteasomal Degradation Contributes to Colonic Obstruction of Hirschsprung Disease
W. Zhao,P. Wang,W. He,T. Tao,H. Li,Y. Li,W. Jiang,J. Sun,X. Ge,X. Chen,Y. Zheng,L. Wei,C. Chen,Y. Wang,C. Li,H. Chen,B. Yao,W. Tang,M. Zhu
DOI: https://doi.org/10.1016/j.jcmgh.2019.11.003
2020-01-01
Abstract:Smooth muscle contraction is evoked through multiple signaling pathways converging on myosin light chain phosphorylation.1Murthy K.S. Annu Rev Physiol. 2006; 68: 345-374Crossref PubMed Scopus (300) Google Scholar,2Somlyo A.P. et al.Physiol Rev. 2003; 83: 1325-1358Crossref PubMed Scopus (1678) Google Scholar Contraction-associated proteins such as MLCK, MYPT1, RLC, RhoA, ROCK, and CPI-17 are differentially expressed in different types of smooth muscle, leading to contractile diversity.3Patel C.A. et al.Am J Physiol Gastrointest Liver Physiol. 2006; 291: G830-G837Crossref PubMed Scopus (57) Google Scholar, 4Perrino B.A. J Neurogastroenterol Motil. 2016; 22: 213-225Crossref PubMed Scopus (33) Google Scholar, 5Woodsome T.P. et al.J Physiol. 2001; 535: 553-564Crossref PubMed Scopus (213) Google Scholar Herein we demonstrate that lipopolysaccharide (LPS) causes MYPT1 protein degradation via SIAH1/2 E3 ligases-ubiquitin-proteasomal pathway, which alters colonic contractile properties and results in obstructive phenotype. Isolated colonic smooth muscle was treated with increasing doses of LPS ex vivo, causing significant reduction of MYPT1 protein (Supplementary Figure 1A). Compared with the proximal colon, MYPT1 protein in the distal colon was also reduced by LPS local treatment (Supplementary Figure 1B). Meanwhile, a single intraperitoneal injection of LPS (5 mg/kg) induced a significant reduction of MYPT1 within 24 hours, and the effect lasted for 5 days (Figure 1A). After injection of LPS (0.5 mg/kg) sequentially with a 24-hour interval, MYPT1 protein was maintained at a low level and recovered within 3 days after stopping the injection (Supplementary Figure 1C). Summarily, MYPT1 protein can be reduced by LPS and reversed by removing LPS. We found MYPT1 mRNA level in colonic smooth muscle comparable with or without LPS (Supplementary Figure 1D). On treatment with MG132, MYPT1 protein was not affected by LPS (Supplementary Figure 1E). In pull-down by anti-MYPT1 antibody, the ubiquitinated MYPT1 level was increased by LPS (Figure 1B). Pulled-down MYPT1 protein complex contained E3 ligases (SIAH1 and SIAH2) that have a binding motif with MYPT1.6Twomey E. et al.Exp Cell Res. 2010; 316: 68-77Crossref PubMed Scopus (22) Google Scholar SIAH1 and SIAH2 levels were elevated by LPS or LPS plus MG132 (Figure 1B). Furthermore, C3H/HeJ mice with a mutant TLR4 showed no apparent reduction of MYPT1 by LPS (Supplementary Figure 1F). Therefore, LPS/TLR4 enhances MYPT1 degradation through the SIAH1/2 E3 ligases-ubiquitin-proteasomal pathway. LPS local treatment in the distal colon caused 45% of mice to have packed feces in the colon lumen (Figure 1D), and circular smooth muscles showed an enhanced sustained phase of KCl-evoked contraction (Supplementary Figure 1G). These phenotypes did not appear in MYPT1SMKO mice with the same treatment (Supplementary Figure 1H). Therefore, MYPT1 reduction by LPS causes an obstruction-like phenotype by altering colonic contraction. To investigate the role of MYTP1 in Hirschsprung disease (HD), we collected fresh colons from HD patients: circular (Cir-) and longitudinal (Long-) muscle strips from dilated (D) and narrow (N) segments. LPS concentration was significantly higher in Cir-N than in Cir-D and Long-D and slightly higher than in Long-N (Figure 2A). The LPS average concentration in narrowed segments was higher than in the dilated (Figure 2B). Meanwhile, compared with Cir-D, MYPT1 protein in Cir-N was decreased (Figure 2C), and PP1cδ was accordingly down-regulated (Figure 2C). In Long-N and Long-D, the contractile proteins were comparable (Supplementary Figure 2A). Moreover, Cir-N muscles produced a strong contractile response with a large, robust, and sustained tension by KCl, whereas Cir-D, Long-N, and Long-D displayed a phasic contraction (Figure 2D). H1152 and GF109203X inhibitors could not relax the KCl-contracted muscle, whereas sodium nitroprusside (SNP) and nifedipine could in Cir-N (Supplementary Figure 2B–E). On KCl stimulation, Cir-N smooth muscle from HD patients exhibited significantly higher regulatory light chain (RLC) phosphorylation than Cir-D at the sustained phase (Figure 2E), in accordance with its maintained force tension. We generated MYPT1 smooth muscle–specific knockout mice, Mypt1f/f; SMA Cre-, Mypt1f/+; SMA-Cre+ and Mypt1f/f; and SMA-Cre+7. On KCl stimulation, the colonic circular muscle from homozygotes showed enhanced initial and sustained tension, whereas the vehicle from heterozygotes displayed a modest enhancement (Supplementary Figure 3A). Meanwhile, H1152 and GF109203 could not relax the mutant muscle, but SNP and nifedipine could (Supplementary Figure 2B–E). The number of ganglia and ganglionic cells in the mutant colon was significantly reduced (Supplementary Figure 3F). Collectively, MYPT1-deficient colon showed similar phenotypes including altered contractile properties and ganglia to Cir-N from HD patients, although the MYPT1SMKO mice showed no colonic obstruction, comparable bowel motility, transit velocity, and eating/defecation activity.7He W.Q. et al.Gastroenterology. 2013; 144 (1465 e1–e5): 1456-1465Abstract Full Text Full Text PDF PubMed Scopus (37) Google Scholar Because circular smooth muscle layer constricts the bowel lumen and longitudinal muscle dilates the lumen and propels feces,8Smith T.K. et al.J Physiol. 1998; 506: 563-577PubMed Google Scholar,9Trendelenburg P. Naunyn Schmiedebergs Arch Pharmacol. 2006; 373: 101-133Crossref PubMed Scopus (20) Google Scholar the disruption of this coordination might cause colonic obstruction. We also established an Ednrb-/- line (Supplementary Figure 4A–D), an HD model with colonic obstruction and few ganglions. The high concentration of LPS, MYPT1 deficiency, and altered contractile property were also confirmed (Supplementary Figure 4E–G). In summary, normal intestinal function relies on multiple factors such as smooth muscle and enteric nervous system. MYPT1 is a primary regulator of smooth muscle contraction.7He W.Q. et al.Gastroenterology. 2013; 144 (1465 e1–e5): 1456-1465Abstract Full Text Full Text PDF PubMed Scopus (37) Google Scholar LPS can degrade MYPT1 in colonic circular smooth muscle, thereby altering the contractile property and leading to colonic obstruction. Removing local LPS or targeting the SIAH/E3-mediated protein degradation pathway might be a useful strategy to treat colonic obstruction in HD. Immunoprecipitation was performed according to the method as previously described.1Yang W.L. et al.Science. 2009; 325: 1134-1138Crossref PubMed Scopus (436) Google Scholar Colonic smooth muscles were lysed by lysis buffer (50 mmol/L Tris.HCl (pH 7.4), 150 mmol/L NaCl, 1 mmol/L EDTA, 1% Triton X-100, phenylmethylsulfonyl fluoride, and protease inhibitor cocktail [Roche, Basel, Switzerland]), and the debris was removed by centrifugation. The resultant lysates were pre-cleared with protein G Sepharose beads (lysates: protein G, 10:1) for 1 hour, followed by incubation with anti–MYPT-1 antibody overnight at 4°C. After adding protein G Sepharose beads and incubating for 3 hours at 4°C, Sepharose beads were collected and washed twice by centrifugation. Then we added 1 × loading buffer (10 mmol/L Tris-HCl, pH 8.0, 1% sodium dodecyl sulfate, 50 mmol/L dithiothreitol, 10% glycerol) to the beads, followed by incubation at 100°C for 10 minutes, and the protein sample was prepared by centrifugation at 12,000g for 1 minute. To determine the expression of MYPT-1, ROCK-2, PKC, PP1cδ, and their related proteins, we isolated and homogenized circular and longitudinal smooth muscle layers. This procedure was performed as previously described.2Matthew J.D. et al.J Physiol. 2000; 529: 811-824Crossref PubMed Scopus (80) Google Scholar The following primary antibodies were used for immunoprecipitation and immunoblot: anti–β-actin antibody (Sigma-Aldrich, St Louis, MO), anti–MYPT-1 antibody (Millipore, Burlington, MA), anti–ROCK-2 antibody (Santa Cruz Biotechnology, Dallas, TX), anti-PP1cδ antibody (Millipore), anti-PKC antibody (Millipore), anti-RhoA antibody (Santa Cruz Biotechnology), anti-ubiquitin antibody (Abcam, Cambridge, UK), anti–SIAH-1 antibody (Signalway Antibody Co, College Park, MD), and anti–SIAH-2 antibody (Santa Cruz Biotechnology). To produce smooth muscle–specific knockout mice, Mypt1flox/flox mice and SMA-Cre transgenic mice were crossed.3He W.Q. et al.Gastroenterology. 2013; 144 (1465 e1–e5): 1456-1465Abstract Full Text Full Text PDF PubMed Scopus (55) Google Scholar The resultant mice were Mypt1+/+ (Mypt1flox/flox)，Mypt1+/- (Mypt1+/flox；SMA-Cre) and Mypt1-/- (Mypt1flox/flox；SMA-Cre). The strategy of genotyping was described in our previous report.3He W.Q. et al.Gastroenterology. 2013; 144 (1465 e1–e5): 1456-1465Abstract Full Text Full Text PDF PubMed Scopus (55) Google Scholar Endothelin B receptor (Ednrb) knockout mice were generated by CRISPR/Cas9 technology. In brief, the Cas9 protein containing a C-terminal SV40 nuclear localization signal was synthesized and inserted into pUC57-sgRNA expression vector. Two gRNAs (gRNA1 and gRNA2) targeting sites of exon1 of Ednrb were designed, and sgRNA scaffold containing T7 promoter was amplified from pUC57-sgRNA. The gRNAs together with Cas9 mRNA were injected into C57BL/6 mouse zygotes. Primer pairs for genotyping were forward, 5′-CCA GTT GGT CTC CAG ACT GAA-3′; reverse, 5′-AAG GAT CTT GGC GGG ACT CCA GC-3′. All animal procedures were conducted in accordance with the guidelines of the Institutional Animal Care and Use Committee of the Model Animal Research Center of Nanjing University (Nanjing, China). For measuring the force produced by mouse colonic muscles, the circular (3–5 mm in width) muscle strips from the distal colon (10 mm proximal to the anus)4Perrino B.A. J Neurogastroenterol Motil. 2016; 22: 213-225Crossref PubMed Scopus (33) Google Scholar,5Woodsome T.P. et al.J Physiol. 2001; 535: 553-564Crossref PubMed Scopus (213) Google Scholar were prepared and then mounted on a force transducer. Force measurements were performed according to our previously described methods.3He W.Q. et al.Gastroenterology. 2013; 144 (1465 e1–e5): 1456-1465Abstract Full Text Full Text PDF PubMed Scopus (55) Google Scholar The resting tension was set to approximately 0.5 g before force measurement. The muscles were stimulated by using a KCl-depolarization buffer containing 87 mmol/L KCl. The chemicals used for the contraction measurement, such as H1152 (Calbiochem, San Diego), GF109203X (Tocris, Bristol, UK), SNP (Sigma-Aldrich), and nifedipine (Sigma-Aldrich), were diluted with HEPES-Tyrode (H-T) buffer to achieve the indicated concentrations. The dilated and narrow colon biopsies were collected from 87 HD patients who received transanal endorectal pull-through or laparoscope-assisted transanal endorectal pull-through operations6Twomey E. et al.Exp Cell Res. 2010; 316: 68-77Crossref PubMed Scopus (22) Google Scholar,7He W.Q. et al.Gastroenterology. 2013; 144 (1465 e1–e5): 1456-1465Abstract Full Text Full Text PDF PubMed Scopus (37) Google Scholar at Nanjing Children Hospital. The ages of patients were 4 ± 0.3 months, and the gender ratio (male: female) was 3:1. The percentage of patients with preoperative Hirschsprung-associated enterocolitis (HAEC) was 21.79%, and the percentage of patients with postoperative HAEC was 11.53%. Patients with preoperative HAEC were treated with antibiotic treatment before surgery, whereas the others were treated through enema with 0.9% NaCl. The percentage of patients with more than 15 cm length of narrow segments was 24.36%, and percentage of those with segment less than 15 cm length was 75.64%. All biopsies were immediately stored in pre-cold and pre-oxygenated H-T buffer3He W.Q. et al.Gastroenterology. 2013; 144 (1465 e1–e5): 1456-1465Abstract Full Text Full Text PDF PubMed Scopus (55) Google Scholar and then subjected to force measurement within 2 hours. After removing the mucosa from colon, strips of circular and longitudinal muscle (1.5 × 10.0 mm)8Gill R.C. et al.Gut. 1986; 27: 293-299Crossref PubMed Scopus (10) Google Scholar from the dilated colon and the narrow colon were cut along the direction of the muscle fibers and were mounted on a force transducer (MLT0202; AD Instruments, Spain) that was connected to a PowerLab recording device (AD Instruments, Australia). The strips were mounted in circular or longitudinal orientation and were equilibrated in H-T buffer for 30 minutes at 37°C before force measurement. The resting tension was set to approximately 1.0 g.9Choe E.K. et al.J Korean Med Sci. 2012; 27: 777-783Crossref PubMed Scopus (9) Google Scholar Urea/glycerol–polyacrylamide gel electrophoresis was performed to separate the non-phosphorylated RLC from the phosphorylated RLC.10Isotani E. et al.Proc Natl Acad Sci U S A. 2004; 101: 6279-6284Crossref PubMed Scopus (101) Google Scholar This procedure was performed as previously described.3He W.Q. et al.Gastroenterology. 2013; 144 (1465 e1–e5): 1456-1465Abstract Full Text Full Text PDF PubMed Scopus (55) Google Scholar Western blotting using an RLC-specific antibody was performed to visualize the RLC-containing bands.10Isotani E. et al.Proc Natl Acad Sci U S A. 2004; 101: 6279-6284Crossref PubMed Scopus (101) Google Scholar The amount of monophosphorylated RLC relative to the total amount of RLC protein was determined by using a Jieda 801 Image Analysis System 3.3.2 (JEDA Science-Technology Development Co, Ltd, Nanjing, China) and was expressed as a percentage. Eight- to twelve-week-old MYPT1SMKO mice were killed by cervical dislocation. The entire colon (proximal colon, 10 mm to cecum; distal colon, 10 mm to anus) was fixed in 4% formaldehyde at 4°C for 2 hours, dehydrated in butyl alcohol at 4°C overnight, embedded in paraffin, and cut into 7-μm sections.11Maeda S. et al.Science. 2005; 307: 734-738Crossref PubMed Scopus (665) Google Scholar The sections were stained by hematoxylin/eosin, and the ganglionic cells in the entire colon were examined under microscopy images (Dotslide; Olympus, Tokyo, Japan). Data were presented as the mean ± standard error of the mean. The differences between 2 groups were evaluated by paired or unpaired t tests. Multiple group comparison was performed by using one-way ANOVA, followed by Tukey's test. P ≤ .05 was considered statistically significant. All statistical analyses were performed by using GraphPad software (San Diego, CA).Supplementary Figure 2Relaxant effects of ROCK, PKC, L-type calcium channel inhibitors, and SNP on KCl-evoked contraction of Cir-N from HD and colon from MYPT1SMKO mice. (A) Western blots showing relative amounts of proteins in Long- smooth muscles (left panel). Quantification of protein levels in the narrow segment as percentage of those in the dilated segment (right panel) (paired t test). MYPT1, n = 51; PP1c, n = 34; ROCK2, n = 29; PKC, n = 31; RLC, n = 19; RhoA, n = 15. (B–E) Representative tracings of Cir-N pre-contracted using 87 mmol/L KCl, followed by exposure to 0.3 μmol/L H1152 (B), 5 μmol/L GF109203X (C), 100 μmol/L SNP (D), or 3 μmol/L nifedipine (E). Relative ratios of relaxed force during the sustained phase, which were expressed as percent of force of contraction after addition of KCl at same time point as addition of the vehicle (force % = (Fvehicle – Finhibitor)/Fvehicle, F = force) in the left panel. Bars represent mean values ± standard error of the mean; n = 3. *P < .05; ***P < .001 (t test).View Large Image Figure ViewerDownload Hi-res image Download (PPT)Supplementary Figure 3Colon from MYPT1SMKO mice presented a similar phenotype compared with Cir-N from HD. (A) Representative tracings of circular smooth muscle of distal colon treated with 87 mmol/L KCl from Myptf/f; SMA Cre-, Mypt1f/+; SMA Cre+, and Mypt1f/f; SMA Cre+ mice. Quantification of force responses of circular smooth muscle to treatment with KCl (one-way ANOVA) (n = 6). (B–E) Representative tracings of circular distal colon from MYPT1SMKO mice pre-contracted using 87 mmol/L KCl, followed by exposure to 0.3 μmol/L H1152 (B), 5 μmol/L GF109203X (C), 100 μmol/L SNP (D), or 3 μmol/L nifedipine (E). Relative ratios of the relaxed force during the sustained phase, which were expressed as percent of force of contraction after addition of KCl at the same time point as addition of the vehicle (force % = (Fvehicle – Finhibitor)/Fvehicle, F = force) in left panel. (F) H&E staining of colons showed significant reduction in number of ganglionic aggregates and total ganglionic cells in Myptf/f; SMA Cre- and Mypt1f/f; SMA Cre+ mice. Quantitation of the ganglion and the total neuron cells count normalized by colon area (t test) (n = 6). Asterisk indicates a ganglion. Bars represent mean values ± standard error of the mean. **P < .01; ***P < .001; #P < .05; ##P < .01.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Supplementary Figure 4Colonic phenotype of Ednrb-/- mice. (A) Schematic representation of Ednrb knockout strategy by CRISP-Cas9 technology. (B and C) DNA sequencing isolated from chimeric (+/-) and wild-type (+/+) mice tail presented the deleted segment. (D) Western blot analysis of EDNRB protein expression in the colon from Ednrb+/+ and Ednrb-/- mice. (E) Limulus test showing concentration of LPS in Ednrb+/+ and Ednrb-/- mice (n = 6). (F) Western blots showing decreased expression of MYPT1 in Ednrb-/- mice (n = 4). (G) Representative tracings of responses of circular smooth muscle from colon of Ednrb+/+ (left panel) and Ednrb-/- (middle panel) mice evoked by 87 mmol/L KCl. Quantification of force tension of circular smooth muscle from HD colons that were evoked by KCl (right panel) (n = 6). Bars represent mean values ± standard error of the mean. *P < .05; **P < .01; ***P < .01 (t test). bp, base pairs.View Large Image Figure ViewerDownload Hi-res image Download (PPT)