Diabetic Gastroparesis The aim of this project is to understand the cellular mechanisms that lead to the development of gastroparesis and to develop new strategies to stop and reverse these defects. Project Progress A. Non-invasive method to measure gastric emptying in mice
Critical towards being able to carry out both specific aims was the need to develop a non-invasive method to measure gastric emptying in mice. This was needed because we had to be able to directly correlate gastric emptying with cellular changes. The lack of ability to follow changes in gastric emptying over time and obtain tissue directly correlated with gastric emptying has limited advances in our understanding of the cellular and molecular changes that lead to gastroparesis in diabetes. This is because of the inability to distinguish changes associated with diabetes from changes specifically associated with development of abnormal gastric emptying. We tested and validated a 13C octanoic acid-based solid gastric emptying test where mice are fed egg yolk containing 2.5 µmol 13C octanoic acid. The mice are placed in chambers (6 chambers in parallel) that allow them to turn around, which is important to prevent stress. We worked with the company that makes the Infra Red Isotope Analyzer (IRIS, Wagner, Germany) to modify the software that was developed for humans to enable use in mice. This included use of equations for mouse body surface area and CO2 production. We determined the reliability and responsiveness of the 13C-octanoic acid breath test in NOD mice and used the test to measure solid gastric emptying in this and other strains of mice. The 13C-breath test produced results similar to post-mortem recovery of a meal, showing the test is accurate. Bethanechol accelerated gastric emptying and atropine slowed gastric emptying, showing the test is responsive. Normal gastric emptying (T½) in non-diabetic NOD mice (8-12 wks) was 91±2 min. Onset of diabetes was accompanied by accelerated gastric emptying during weeks 1-2 of diabetes. Gastric emptying returned to normal by end of week 2 with no delay. After that, about 20% of mice (currently 30%) developed delayed gastric emptying. Initial delay was noted at 4 weeks and the median was 5.5 weeks. Mice with more than 8 weeks of diabetes rarely developed delayed gastric emptying (1 in 100), indicating that the mice that had normal gastric emptying by 8 weeks of diabetes were resistant to the development of delay. We concluded that the 13C-octanoic acid breath test accurately measures gastric emptying in NOD mice, is useful to study the time course of changes in gastric emptying in diabetic NOD mice, and is able to detect acceleration and delay in gastric emptying in diabetes. This methodology is now used by several laboratories. B. Develop methodology to obtian full thickness biopsies endoscopically
We have set up a series of experiments using pigs in collaboration with Dr. Liz Rajan and Dr. Chris Gostout to develop methodology to obtian full thickness biopsies endoscopically. In the first paper we tested different methods to endoscopically obtain full thickness biopsies. We tested endoscopic ultrasound-guided Tru-Cut biopsy of the gastric wall, jumbo biopsy of post-endoscopic mucosal resection site, jumbo biopsy of gastrotomy margin, serosal-side biopsy through a gastrotomy and double endoscopic mucosal resection. We found that double endoscopic mucosal resection technique was the only technique that resulted in a deep gastric wall sample and provided sufficient tissue to evaluate both muscle layers and myenteric ganglia. However, due to the need to create a 'hole' that is then clipped shut, we noted a high rate of peritonitis. Therefore, we modified the double endoscopic mucosal resection by using detachable endoloops and prototype T-tag tissue anchors prior to resection. We found we were able to obtain full thickness gastric biopsies from all (n=6) animals without perforation. There were no procedural complications. Four animals had an uneventful clinical course. Two animals were euthanized at days 3 and 6, respectively, due to suspected peritonitis. At necropsy, delayed perforations at the resection sites were noted with displaced endoloops and tissue anchors. We have now redesigned the endoloops and will test them in the coming year. If successful, we would have established a method for non-surgical, endoscopic full thickness biopsies, which could be used in a variety of diseases to aid with diagnosis and offer insight into etiology. - Rajan E, Gostout CJ, Lurken MS, Talley NJ, Locke GR, Szarka LA, Levy MJ, Sumiyama K, de la Mora-Levy JG, Bakken TA, Stoltz GJ, Knipschield MA, Farrugia G. Evaluation of endoscopic approaches for deep gastric muscle wall biopsies. Gastrointest Endosc 2008 Feb;67(2):297-303.
- Rajan E, Gostout CJ, Lurken MS, Talley NJ, Locke GR, Szarka LA, Levy MJ, Sumiyama K, Bakken TA, Stoltz GJ, Knipschield MA, Farrugia G. Endoscopic “no hole” full thickness biopsy of the stomach to detect myenteric ganglia. Gastrointestinal Endoscopy, In press.
C. Examine the role of nNOS and NO as a factor underlying the ICC
We examined the role of nNOS and NO as a factor underlying the survival and maintenance of ICC. To do this, ICC volumes were determined in nNOS knockout mice and compared to control mice in eight carefully delineated regions of the gastric body using immunohistochemistry and laser scanning confocal microscopy followed by three-dimensional reconstructions. Also, ICC numbers were determined in ICC-enriched cell cultures after treatment with S-nitroso N-acetylpenicillamine (SNAP) as an NO donor or NG nitro-L-arginine (L-NNA) as a nNOS inhibitor. We looked at Kit expression using Western blots and showed that Kit expression was decreased in nNOS knockout mice. The volume of ICC in the myenteric region of the gastric body was also significantly less. As it was possible that the decrease in ICC in the myenteric plexus region was due to stretching of the stomach in the nNOS knockout mice, we also determined the number of ganglia and the number of neurons within the ganglia. No significant differences between control and nNOS knockout mice were observed. As nNOS is absent throughout development in nNOS knockout mice, it is possible that changes in ICC may be secondary to changes in other cell types in the muscle wall that may also be altered by a decrease in nNOS availability. To determine if NO had a direct effect on ICC, ICC cultures were exposed for two days to either 100 µM SNAP or 200 µM L-NNA. Treatment with SNAP increased the number of Kit positive ICC by 37±8% while treatment with L-NNA decreased the number of Kit positive ICC by 56±27%. These data suggest that nNOS derived NO is cytoprotective to ICC. These data link loss of ICC with the known decreased expression of nNOS seen in diabetes as loss of the cytoprotective effects of NO would make ICC more vulnerable to injury. D. Establish the use of the human appendix as a diabetes mode
In another set of experiments, we established the use of the human appendix as a diabetes model for the rest of gut. Appendixes were collected from patients with type 1 diabetes and gastroparesis and from non-diabetic controls. Volumes of nerves and ICC were determined using 3-D reconstruction and nNOS expressing neurons were counted. ICC were found throughout the muscle layers but no myenteric ICC were found. In diabetes, Kit positive ICC volumes were significantly reduced, as were nNOS expressing neurons. However, volumes of PGP9.5 immunoreactive fibers were not down within the muscle layers overall, although they were slightly decreased in the longitudinal muscle layer. These data therefore suggest that the human appendix, a readily available source of human tissue, may be useful model for the study of motility disorders, especially diabetic gastroenteropathy. They also suggest that the loss of nNOS in diabetes occurs throughout the gut and is not associated to a significant degree with neuronal cell loss. This raises the possibility of being able to intervene to restore nNOS expression, given that the neuron itself is not lost. - Miller S, Narasimhan R, Schmalz P, Soffer E, Walsh R, Krishnamurthi V, Pasricha P, Szurszewski J, Farrugia G: Distribution of interstitial cells of cajal and nitrergic neurons in normal and diabetic human appendix. Neurogastroenterol and Motil. 20:349-57, 2008.
- Sagstetter AM, Camp JJ, Lurken MS, Szurszewski JH, Farrugia G, Gibbons SJ, Robb RA. Computer aided classification of cell nuclei in the gastrointestinal tract by volume and principal axis. Proc of SPIE 6514:65140E-1 – 65140E-9, 2007.
E. Determine gastric emptying in non-diabetic NOD mice and as diabetes progressed
We took advantage of the 13C octanoic acid-based non-invasive gastric emptying test to weekly determine gastric emptying in non-diabetic NOD mice and follow gastric emptying as diabetes progressed. This enabled us to directly correlate changes in Kit and nNOS expression with changes in gastric emptying. Tissue was harvested at 2 weeks of diabetes, 4-5 weeks of diabetes, and 10 weeks of diabetes. Mice started developing delayed gastric emptying after 4 weeks of diabetes (range 4-7 weeks) with a mean of 5.5 weeks. There was no difference in glucose levels or b-hydroxybutyrate levels between diabetic groups. Mice that developed delayed gastric emptying were assigned to a separate group. No mice developed delayed gastric emptying after 7.5 weeks; therefore, the mice with 10 weeks of diabetes were considered resistant to the development of diabetic gastroparesis. Analysis of Western blots from the gastric body and antrum showed that nNOS expression decreased within 2 weeks of development of diabetes and decreased further at 4-5 weeks and 10 weeks, as well as in the mice with delayed gastric emptying. In contrast, Kit expression was maintained in all mice that did not develop delayed gastric emptying, while it was markedly decreased in every mouse that developed delayed gastric emptying. These data suggested that loss of Kit expression is required for the development of delayed gastric emptying, while loss of nNOS expression is associated with diabetes and predisposes to development of delayed gastric emptying but is not directly causative of delayed gastric emptying. Oxidative stress levels were higher in the group that developed delayed gastric emptying compared to the ones that did not. Oxidative stress has been implicated in the development of end organ damage from diabetes. To test the hypothesis that induction of HO1 protects cells against oxidative damage, we next examined HO1 expression in the different mouse groups. In both the gastric body and antrum basal levels of expression of HO1 were low. HO1 expression increased markedly after 4-5 weeks of diabetes in all mice. In all mice that were resistant to development of diabetic gastroparesis, HO1 levels remained high. In contrast, in all mice that developed diabetic gastroparesis, HO1 protein expression levels were no longer elevated and had returned to baseline levels. These data may be interpreted that a required step to loss of Kit expression and subsequent development of diabetic gastroparesis is loss of up-regulation of HO1. However, it is also possible that the loss of up-regulation of HO1 was an epiphenomenon, not directly linked to loss of Kit or development of diabetic gastroparesis. Therefore, we next investigated whether induction of HO1 via hemin reversed delayed gastric emptying. In a separate series of experiments, diabetic NOD mice were longitudinally followed with weekly gastric emptying tests until development of delayed gastric emptying. After a 2-week period to ensure that the delay in gastric emptying was sustained, these mice were randomly assigned to receive hemin (40 µmol/kg) ip daily or vehicle (n=5 each group) for up to 6 weeks or sooner (mean 4.4 weeks) if gastric emptying normalized on two consecutive tests. All 5 mice receiving hemin normalized their gastric emptying; none of the vehicle treated controls did. Normalization of gastric emptying was accompanied by a decrease in oxidative stress, and a reversal of the decrease in expression of Kit. These data confirm that HO1 cannot only protect against development of delayed gastric emptying but can also reverse established delayed gastric emptying and reverse the loss of Kit expression, in a relatively rapid time course of less than 5 weeks. Given the significance of the findings, we confirmed these results by treating diabetic mice with an inhibitor of HO activity, chromium mesoporphyrin (CrMP, 3 µmol/kg) ip. Two weeks after onset of diabetes, mice were assigned to receive CrMP or vehicle for 4 weeks or until delayed gastric emptying was documented on two consecutive tests. CrMP inhibited gastric HO activity, increased oxidative stress, and resulted in development of delayed gastric emptying in all mice assigned to this group. None of the mice in the vehicle group developed delayed gastric emptying. Development of delayed gastric emptying was associated with loss of expression of Kit. - Kyoung Moo Choi, Simon J. Gibbons, Tien V. Nguyen, Gary J. Stoltz, Matthew S. Lurken, Tamas Ordog, Joseph H. Szurszewski, Gianrico Farrugia. Induction of Heme Oxygenase-1 Reverses Diabetic Gastroparesis. Gastroenterology. In Press.
F. Other publications related to diabetic gastroparesis |
