Alterations in ENS associated with gastroparesis

Paul Kassander stands first to describe diabetic gastroparesis . In his initial report involving case studies of six patients with diabetes, asymptomatic gastric retention of meal was observed on X-ray findings with no abnormalities in small intestine. This disturbance in the expulsion of gastric contents was attributed to the vagus nerve, and the radiographic results resembled to those of gastric hypotonia after vagotomy . However, two decades later, a case study by Norman Soler involving two diabetic patients showed gastroparesis even in the absence of extrinsic diabetic neuropathy . Around the same time, a report by Shellito & Warshaw described a syndrome of recurrent intermittent gastric atony in five non-diabetic patients experiencing the symptoms of gastroparesis , which gave rise to the entity of IG . A study by Yoshida et al., on 16 patients with long-standing diabetes involving five patients with gastroparesis, did not show any changes in the number and appearance of neurons or axons, and also no abnormalities associated with vagus nerve were observed in the myenteric plexus . However, the major drawback of this study was the use of only conventional histological staining techniques for visualizing myenteric plexus and the vagus nerve.

The initial studies were focused only on the extrinsic nervous system however; subsequent research exhibited defects in intrinsic nervous system in animal models of diabetic gastroparesis. Using streptozotocin (STZ) induced rat model of diabetes (without delayed gastric emptying), Belai et al., have observed a consistent increase in VIP-like immunoreactivity in nerve fibers, and more frequent intensely stained cell bodies in the myenteric plexus and circular muscle layer of both ileum and proximal colon ( P <0.001). Whereas, compared to controls, no significant changes were observed in the SP innervation in myenteric plexus and circular muscle of both ileum and colon of diabetic rats . Similarly, increased VIP-like immunoreactivity was also observed in STZ induced diabetic rats with delayed gastric emptying . Interestingly, continuous insulin administration in an in vitro setting reversed these effects, suggesting that, the availability of a steady level of insulin influences some of the changes occurring during early stages of diabetes . In a similar study by Watkins et al. , non-obese diabetic (NOD) mice and STZ induced diabetic mice showed defects mimicking the mice harboring a deletion of the nNOS by showing defects in gastric emptying and pyloric muscle relaxation . Insulin administration or intraperitoneal injection of phosphodiesterase inhibitor (Sildenafil) to these mice resulted in an increase in the levels of NO and also restored gastric emptying . Interestingly, chronic depletion of NO derived from nNOS resulted in delayed emptying of solids and liquids from the stomach but, depletion of NO derived from eNOS – a constitutive enzyme produced by endothelial cells, had no effect , suggesting the importance of nNOS and NO signaling in pyloric relaxation. Chronic estrogen deficiency in Follicle-stimulating hormone receptor knockout (FORKO) mouse has been reported to affect the function of tetrahydrobiopterin – a co-factor essentially required for nNOS enzyme activity, and thus contributing for the development of gastroparesis . In a report by Choi et al. delayed gastric emptying in NOD mouse model of DG was restored by exogenous administration of IL-10 and this effect was independent of nNOS activity . Further research aiming to delineate the pathways associated with regulation of nNOS expression would help in considering it as a molecular target to treat gastroparesis.

Alterations in expression of enteric nerve markers have also been observed in biopsy specimens from gastroparetic patients. In a case report of a 38-year-old type 1 DM patient, full thickness jejunal biopsy specimen showed a significant decrease in nerve fiber content (PGP9.5) in the circular muscle when compared with the control tissues from six subjects ( P <0.05) . In addition, the nNOS, VIP, and PACAP were also found to be decreased significantly . Yet another case report of 32 year-old IG patient showed similar findings where, there was 69% reduction in the number of nerve cell bodies (PGP9.5) in the stomach corpus sections . A report by Iwasaki et al. involving male gastric cancer patients with and without type 2 DM showed significant reduction in the expression of nNOS (p<0.01) and SP ( P <0.01) in the stomach antrum of patients with type 2 DM but not in patients without type 2 DM, indicating that the expression levels of nNOS & SP are associated with the pathogenesis of DG . In a case report of two gastroparetic patients with type 1 DM, Pasricha et al. have shown reduced expression of nNOS, nerve fibers and myenteric neurons (PGP9.5) in the full thickness gastric biopsy of a patient with long standing, poorly controlled diabetes. Although, both the patients were presented with similar symptoms, the other patient with well controlled diabetes over a short period showed no significant abnormalities, when compared with the controls . These findings suggest the association of histopathological alterations in the ENS with poorly controlled diabetes. In a study by Chandrasekharan et al. enhanced apoptosis and loss of peripherin, nNOS, neuropeptide Y (NPY) and ChAT neurons were observed along with a significant reduction in the ganglion size in the colons of patients with diabetes . In another study by Harberson et al. , aiming to histologically evaluate the pathologies in full-thickness antral biopsies from patients with refractory gastroparesis (14 each of type 1 DM & idiopathic gastroparesis), showed a substantial reduction in the numbers of ganglia, NOS ⁺ and NOS ⁻ nerve cells when compared to controls . However, in the largest study by the NIDDK GpCRC defining the cellular changes associated with gastroparesis using gastric biopsies obtained from control, DG and IG patients (20 patients in each group), only a 14–18% decrease in the expression of PGP9.5 in gastroparesis as compared to the controls . Although, nNOS was found to be decreased in 20% of patients with DG and 40% of patients with IG, no statistically significant difference in nNOS was observed when compared between the groups . Likewise, 20% of patients with DG and 15% of patients with IG displayed a reduction in the expression of VIP immunolabelling but, no statistical difference was found between the three groups . Immunolabeling for SP did not show any statistically significant difference either, when compared between the groups . The results of this largest study to date, involving more number of subjects in each group clearly indicates the changes in the number of enteric nerves and neurons observed may only be present in a subset of patients with gastroparesis. The changes may also be associated with the etiology, disease duration and site of assessment in the stomach. Remarkably, Faussone-Pellegrini et al. using transmission electron microscopy (TEM) of full-thickness gastric biopsies showed evidence for cellular damage to the nerves, even in patients with no apparent histological changes on light microscopy . The observed alterations include, loss of synaptic vesicles, thickened basal lamina and fibrosis around nerves. Interestingly, IG patients showed a more severe damage as compared to DG .

Loss of ICC associated with gastroparesis

In the context of diabetic gastroparesis, the very first study using an animal model of diabetes (NOD/LtJ mice), depletion of gastric ICC networks resulted in a significant delay in emptying of the gastric contents . ICC loss in antral muscle tissues by 50% was associated with reduced Kit-like immunoreactivity. In addition, lack of association between ICC and enteric nerve terminals was also noted . This report suggested the alteration in pacemaker ICC content being the basis of DG. A subsequent study by Wang et al. employing STZ induced rat model of diabetes showed reduction in the density of ICC in antrum and the effect was also accompanied with the loss of synaptic connections and delayed gastric emptying .

Reports concerning the loss of ICC in human biopsy specimens of jejunum , colon and antrum were reported in DG and IG patients. A report by Horváth et al. using organotypic cultures of murine gastric muscle showed that, the loss of ICC was not associated with diabetes induced hyperglycemia; instead, the effect was due to reduced insulin and insulin like growth factor 1 (IGF-1) signaling . In a study by the NIDDK GpCRC , more than 25% of ICC loss was documented in 50% of patients irrespective of DG or IG as shown by reduction in Kit expression. Loss was ICC was similar in both DG and IG accounting for ~50% mean ICC count when compared with controls. Ultrastructural studies using TEM displayed damage associated alterations in ICC structure in the tissues of both DG and IG patients. The observed alterations include apoptotic features, intracytoplasmatic vacuoles, swollen mitochondria and extended rough endoplasmic reticulum. A physical separation of ICC from other ICC and nerves was also observed when compared with the controls. In a similar study focusing on the ultrastructural differences between DG and IG, both intramuscular and myenteric ICC had similar changes . A study on investigating the associations between the status of ICC and electrogastrogram (EGG) parameters on gastric emptying and its symptoms in a large cohort of 41 patients with gastroparesis by Lin et al., showed, severe loss of ICCs in 36% of patients while, rest of the patients had no visible ICCs identified . EGG recordings showed significantly reduced normal slow waves in patients with severe loss of ICC both under fasting and fed states and these results correlated significantly with the loss of ICC . In a similar study by GpCRC, >60% loss of ICC was documented in the enrolled patients . All these studies clearly suggest that, loss of ICC as a hallmark feature of gastroparesis in humans.

Changes in fibroblast-like cells associated with gastroparesis

In a report aiming to determine the morphology and distribution of fibroblast-like cells (PDGFRα ⁺ ) in human gastric body biopsies obtained from control subjects (n=5) showed PDGFRα ⁺ cells having slender cell bodies with long and thin processes . And these PDGFRα ⁺ cells were more predominantly located in longitudinal muscle layers than circular muscle layer. In addition, in the myenteric plexus, these PDGFRα ⁺ cells displayed morphology with smaller and round cell bodies possessing 3–4 processes, forming a network around the ganglia . Although, fibroblast-like cells are a identified to be a sub-population of interstitial cells , no overlapping cell bodies of fibroblast-like cells with the ICC were observed despite, being located in close proximity to each other . In an attempt to identify their alterations in patients with DG (n=10) and IG (n=10), no changes in morphology, distribution or the number of cells were observed in patients with gastroparesis when compared with the controls . Interestingly, in a study involving a large cohort of patients with DM (n=32; without gastroparesis) and control subjects (n=26), significant number of fibroblast-like cells were reduced in the fundus and the body of stomach but, not in antrum of patients with DM when compared with the control group, as identified by PDGFRα ⁺ immunoreactivity . Hence, this creates an avenue for an in-depth analysis of changes in the population of PDGFRα ⁺ cells during DG and IG, and also to delineate their actual role in gastroparesis.

Alterations in smooth muscle cells associated with gastroparesis

A study on the effects of transmural nerve stimulation, acetylcholine, norepinephrine, or K-free solution on the electrical responses of gastric smooth muscles by Xue et al., , the STZ induced diabetic rats showed functional impairments associated with neuromuscular transmission. This abnormality was associated with reduced adrenoceptor (norepinephrine) sensitivity and Na-K pump activity whereas, increased sensitivity to muscarinic receptor (acetylcholine) was observed . These results suggests that, the alterations in SMC leading to the development of delayed gastric emptying is a result of diabetes .

Forkhead transcription factors are shown to be involved in smooth muscle proliferation, migration and apoptosis . Knocking out the forkhead transcription factors – FOXF1 and FOXF2, in SMC of adult mice has been shown to cause delayed gastric emptying . In addition, the authors have also observed a significant reduction in the expression levels of these transcription factors in full-thickness gastric biopsies obtained from DG and IG patients, suggesting the role of FOXF1 and FOXF2 in normal gastric function .

Smooth muscle degeneration and fibrosis with eosinophilic inclusion bodies has been reported by Ejskjaer et al. in four DG patients with a symptom of intractable vomiting and this indicates the presence of gastromyopathy. In a case study of two patients, one exhibiting poorly controlled long standing diabetes and the other with well controlled diabetes over a short duration reported by Pasricha et al. showed, extensive fibrosis in the myenteric plexus, circular and longitudinal muscle layers of full thickness gastric biopsy obtained from the patient with poorly controlled diabetes over long duration but, not in patient with well controlled diabetes over a short duration . This clearly indicates the SMC changes associated with the status and clinical course of DM. Unfortunately, in a large cohort study by NIDDK GpCRC involving both DG and IG patients, no such fibrosis was observed in the full-thickness stomach body specimens when compared with matched controls . In the same study, 23% of patients exhibited reduced expression of smoothelin-A – a specific marker of SMC. However, study using TEM, failed to show any marked abnormalities in SMC in the tissue specimens of both DG and IG patients .

Alterations in macrophage milieu associated with gastroparesis

Determining the role of macrophage dysregulation in the pathogenesis of gastroparesis is an active area of research. This is because; macrophages are prone to acquiring different phenotypes in response to subtle changes in the tissue microenvironment. Various stimuli and cytokines are known to polarize macrophages into M1 (pro-inflammatory) or M2 (anti-inflammatory) phenotype . A number of disease states have been associated with an alteration in the macrophage milieu . M1 and M2 spectrum of macrophages are differentiated based on the expression profiles of different pro- and anti-inflammatory markers expressed on their surface. For example, M1 macrophages typically express the classical pro-inflammatory markers TNF-α, IL-1β and iNOS. On the other hand, M2 macrophages predominantly express the CD206 or mannose receptor, Arginase-1 and TGF-β .

In the context of gastroparesis, the CD206+ macrophages are gaining much importance as the studies show alterations in the overall population of these anti-inflammatory macrophages within the muscular layers of the GI tract. CD206+ macrophages express heme oxygnase 1 (HO-1) – an enzyme well known for its role in cellular defense mechanism against oxidative stress . In a study on NOD mouse model by Choi et al., during early diabetes, CD206+ muscularis macrophages showed increased expression of HO-1 and these levels were consistently maintained in the animals that were resistant to develop delayed gastric emptying . However, loss of CD206+ macrophages was observed in mice developing delayed gastric emptying, and this loss was also associated with loss of HO-1 expression in addition to the loss of Kit (ICC) . Due to the reduced expression of HO-1, an increase in the oxidative stress was also observed and this effect was reversed upon treating these mice either with hemin or IL-10 . This provides evidence that HO-1 expressing CD206+ macrophages may be able to prevent or treat DG.

Despite being severely diabetic, CSF1 ^op/op mice that lack muscularis macrophages do not develop delayed gastric emptying . Supplementing these mice with an intraperitoneal injection of CSF1 restores the muscularis macrophages, rendering them susceptible to acquire gastroparesis upon induction of diabetes . Interestingly, subjecting the cultured ICC with M1 macrophage conditioned medium reduced the ICC cell counts by 41% while, no such effect was observed when the ICC were treated with conditioned medium from M2 macrophages . In an attempt to identify possible paracrine mediators, immunoblot assay showed, 12 pro-inflammatory cytokine and chemokine markers in M1 macrophage conditioned medium . This suggests a potential role for paracrine signaling in macrophage driven Cajalopathy. Further studies delineating the molecular mechanisms associated with macrophage-ICC interactions are needed.

Concerning human studies, an initial report by Harberson et al. on full thickness antrum biopsies from patients with DG (n=14) and IG (n=14), showed mild lymphocytic infiltrate positive for CD45 and CD3 in the myenteric plexus . Further studies by NIDDK GpCRC on full thickness gastric biopsies of the antrum and body obtained from DG and IG patients showed a significant reduction in the total number of CD206+ macrophages in the sections of antrum but not the body of the stomach. However, total population of immune cells (CD45+) was unchanged . Remarkably, the loss of CD206+ macrophages correlated significantly with the loss of ICC in circular muscle of gastric antrum , suggesting the importance of CD206+ macrophages in maintaining ICC population.