Gastric Biopsies in Gastroparesis: Insights into Gastric Neuromuscular Disorders to aid treatment

Introduction

Gastroparesis is a chronic disorder of the upper gastrointestinal (GI) tract characterized by delayed gastric emptying of solids and/or liquids in the absence of mechanical obstruction of the gastric outlet (1, 2). The signs and symptoms include nausea, vomiting, abdominal pain, early satiety and post-prandial fullness (3), which result in significant impairment of quality of life and high healthcare expenditures (4).

In one study, diabetic gastroparesis (DG) accounted for 29% patients, while gastroparesis acquired after a surgery (post-surgical gastroparesis) accounted for 13% of the cases (3). An unknown primary cause or idiopathic gastroparesis (IG) accounts for >50% of patients (3, 5). Other etiologies include connective tissue disorders, end stage renal disease, Parkinson’s disease, or medication-induced (3, 6). An epidemiological survey by Jung et al. shows the incidence of definite gastroparesis ranging from 6.3 to 17.2 cases per 100,000 persons in an age and gender-adjusted population and a 4-fold higher prevalence of gastroparesis was observed in women when compared with men (6), although other estimates suggest a considerably higher prevalence.

Gastric emptying is considered to be abnormal/delayed when >60% of administered contents are retained at 2 hours and/or >10% at 4 hours on a solid meal gastric scintigraphy (7). Ongoing research in the field of gastroparesis using human gastric biopsy specimens has provided deeper insights into pathological alterations associated with this disease. In this review, we aim to highlight the recent advancements in our understanding of gastroparesis using the data from human gastric biopsy specimens and also from complimentary animal models studies. Additionally, we propose future directions for clinical utility of human biopsies and strategies for tissue procurement that are minimally invasive and can be applied more broadly.

Cellular control of normal gastric function:

Following the completion of gastric phase, partially digested food (chyme) present in the stomach is emptied into the duodenum by coordinated gastric motility. Gastric emptying is a complex physiological process, which involves the coordinated interactions of extrinsic nerves, the enteric nervous system (ENS), smooth muscle cells (SMCs), and the ICC within the muscularis and myenteric layers of the stomach as well as coordination between different parts of the stomach and feedback loops between the small intestine and stomach. A characteristic feature of the GI tract is the presence of its own intrinsic neuroglial circuits (the ENS). Nerve fibers innervate the muscular layers of the stomach (8). The myenteric plexus situated between the circular and longitudinal muscle layers is known to regulate contraction and relaxation of smooth muscles (9). Inhibitory and the excitatory neurons are categorized based on the expression pattern of neurotransmitters. For example, excitatory neurons predominantly express choline acetyltransferase (ChAT) (10), substance P (SP) and neurokinins (NKA, NKB & neuropeptide Y) (11) while, inhibitory neurons express vasoactive intestinal polypeptide (VIP) (12) and, neuronal nitric oxide synthase (nNOS) which generates nitric oxide (NO) (13).

Smooth muscle of GI tract is required to mix and churn intraluminal contents, enabling breakdown of ingested food. This is followed by propulsion of chyme into the duodenum primarily by the contraction of SMCs present in the muscularis externa. This process of gastric emptying is accomplished by phasic contractions of SMCs (14). The SMCs are linked to the neighboring pacemaker cells via gap junctions creating a syncytium (15), which drives coordinated contractions and relaxation of SMCs (16).

ICCs are the pacemaker cells which create the bioelectrical slow wave potential in the GI tract (17). Syncytial network of ICCs generate electrical slow waves in a spatiotemporal manner driving rhythmic contraction of SMCs (18), (19). ICCs are also involved in integrating and mediating sequential excitatory and inhibitory neuro transmissions and in mechanotransduction – the critical components of normal gastrointestinal motility (20). Variants in Ano-1 transcripts, a protein specifically expressed in ICC has been associated with symptoms in DG (21).

Macrophages are increasingly recognized as key regulators of tissue homeostasis. Resident macrophages are highly heterogeneous and can acquire distinct phenotypes in response to changes in the tissue microenvironment. They display variable gene expression profiles (22). Indeed, macrophage gene expression varies from mucosa, submucosa, muscularis and serosa in mouse (23). Mouse muscularis macrophages predominantly express CX3CR1 (hi), MHCII (hi) and CD11c (lo), differentiating themselves from lamina propria macrophages which express CD11c (hi) (24). Similarly, human muscularis macrophages express high levels of CD11b and CD14 while, low levels of CD11b and CD14 are expressed by mucosal macrophages (25). Variability also persists in the expression pattern of well-established macrophage markers between mouse and humans (26, 27); for example, iNOS, Arginase-1 and Ym1 are predominantly expressed by mouse macrophages but not by human macrophages (28, 29). Phenotypically, circular and longitudinal muscle muscularis macrophages are bipolar in shape, whereas, myenteric macrophages are stellate shaped (23).

Muscularis macrophages live in symbiosis with the ENS wherein, macrophages produce bone morphogenetic protein-2 (BMP-2) – a secreted protein belonging to TGF-β superfamily (24). BMP-2 directly acts on enteric neurons expressing BMP receptor (BMPR), leading to oligomerization of type I and type II serine kinases followed by phosphorylation and nuclear translocation of SMAD proteins (24). Activation of BMPR promotes nitrergic enteric neuronal differentiation (30) and helps in regulating gastrointestinal motility (24). In turn, enteric neurons produce colony stimulating fator-1 (CSF-1) – a growth factor crucial for the differentiation and maintenance of muscularis macrophages (24), evident from the osteopetrotic (op/op) mice, which lack muscularis macrophages due to a mutation in their CSF-1 gene (31) and show a disorganized ENS architecture (24). Recently, Avetisyan et al., have reported ICCs as a non-neuronal source of CSF-1 which helps in macrophage homeostasis in RetKO mice which lack enteric neurons (32). In this way, cross-talk between immune cells and the cells of ENS plays essential role in the maintenance of gastric physiological processes.

Cellular abnormalities associated with the pathogenesis of gastroparesis:

Decades of research aiming to identify the physiological and pathological changes associated with the pathogenesis of gastroparesis has resulted in identifying several key abnormalities both at cellular (Figure.1) and molecular levels. In the subsequent sections we will describe some of the important findings which have impacted our understanding of the pathological basis of gastroparesis.

Figure 1: Cellular and molecular changes associated with gastroparesis.

Under normal circumstances (Left panel), the network of ICCs (c-Kit) produce electrical slow waves leading to the membrane depolarization of smooth muscle cells (SMC) followed by emptying of gastric contents. There is a balance of macrophages in the muscle layer and myenteric plexus with majority expressing anti-inflammatory proteins like CD206, HO-1, Arg-1, TGF-β. Neurotransmitters like nitric oxide (NO) and acetylcholine (Ach) released respectively by inhibitory (nNOS, VIP) and excitatory (ChAT, SP, NPY) neurons control the relaxation and contraction of smooth muscles. In gastroparesis (Right panel), loss of ICCs results in impaired electrical slow wave production leading to reduced chronotropicity and ionotropicity of smooth muscle contraction. The changes occurring in the tissue microenvironment polarize the macrophages to acquire pro-inflammatory phenotype which predominantly expresses IL-6, TNF-α, IL-1β and iNOS and these macrophages cause injury and loss of ICCs. Additionally, defect/loss of enteric nerves and neurons leads to impaired contraction and relaxation of SMC as well as impaired pyloric relaxation. These cellular abnormalities collectively lead to delayed gastric emptying.

ENS changes associated with gastroparesis:

In an initial case report by Paul Kassander (33), asymptomatic gastric retention of meal was observed on radiographic studies in diabetic patients. This putative disturbance of gastric motor function was attributed to peripheral neuropathy involving the vagus nerve, as the radiography results resembled to those of patients with gastric hypotonia after vagotomy (33). However, in an observation made almost two decades later, delayed gastric emptying was noted even in the absence of extrinsic diabetic neuropathy (34). Around the same time, a report described “intermittent gastric atony” in 5 non-diabetic patients experiencing the symptoms of gastroparesis (35), which gave rise to the entity of IG (36, 37). A study by Yoshida et al., (38) involving 16 long-term diabetic patients, of which 5 patients exhibiting gastroparesis, failed to show any changes in neuronal numbers or vagal morphological abnormalities (38). However, this study was limited by the use of only conventional histological markers available for testing.

Although most of the initial studies focused on extrinsic nervous system, subsequent work using animal models of diabetic gastroparesis displayed defects in intrinsic nervous system. A report by Belai et al., (39) involving streptozotocin (STZ) induced rat model of diabetes (without delayed gastric emptying) showed an increase in VIP-like immunoreactivity in nerve fibers, and intensely stained cell bodies in the myenteric plexus and circular muscle layer of both ileum and proximal colon (p<0.001) but no changes in the SP innervation (39). STZ induced diabetic rats developing delayed gastric emptying also displayed increased VIP-like immunoreactivity (40). These changes resolved with insulin administration in an in vitro setting, suggesting that, insulin replacement may restore the defects occurring during early stages of diabetes (41). Similar results were observed in non-obese diabetic (NOD) mice (42). Treating these mice either with insulin or with phosphodiesterase inhibitor (Sildenafil) increased the levels of NO and also restored gastric emptying (42). Interestingly, deletion of gene encoding nNOS in mice resulted in delayed gastric emptying of solids and liquids (43) suggesting the importance of nNOS and NO signaling in pyloric dysfunction associated with a subset of gastroparesis patients. In a study by Chandrasekharan et al., (44) enhanced apoptosis and loss of peripherin, nNOS, neuropeptide Y (NPY) and ChAT neurons were observed in the colons of patients with diabetes and this neuronal loss was associated with significant decrease in ganglion size (44). In a study on chronic estrogen deficiency using follicle stimulating hormone receptor knock-out female mice, Ravella et al., (45) have reported the chronic deficiency of estrogen negatively affecting the function of tetrahydrobiopterin (BH4, a co-factor for nNOS dimerization and enzyme activity) and nNOS thereby, contributing to the development of gastroparesis (45). This may relate to the finding that the gastroparesis is more common in females than in males with diabetes mellitus (46). Meanwhile, in a NOD mouse model of DG, Choi et al., have reported the restoration of delayed gastric emptying upon exogenous administration of IL-10 independent of nNOS activity (47). Further research aiming to understand the pathways that are involved in regulating nNOS expression would help to determine its usefulness as a molecular target to treat gastroparesis.

Studies using human tissues from gastroparetic patients have also revealed alterations in nNOS expression. A full thickness jejunal biopsy from a 38-year-old type 1 DM patient showed a substantial decrease in nerve fiber content (PGP9.5) in the circular muscle when compared with the tissues from six control subjects (p<0.05) (48) and, the inhibitory innervation (nNOS, VIP, and PACAP) was also found to be decreased (48). Similar findings were observed in a case report of 32 year old patient with severe IG where, the stomach corpus sections displayed loss of nerve cell bodies (PGP9.5) by 69% (49). A study on type 2 DM male patients with gastric cancer showed significant reduction in the expression of nNOS (p<0.01) and SP (p<0.01) in their antrum, when compared with gastric cancer patients without type 2 DM, suggesting the association of expression levels of nNOS & SP in the pathogenesis of DG (50). In a case report of two gastroparetic patients with type 1 DM (51), histological examination of full thickness gastric biopsy from one patient with long standing, poorly controlled diabetes displayed reduced nNOS, nerve fibers and myenteric neurons (PGP9.5). Despite, both patients had severe refractory symptoms with malnutrition, the other patient with short duration of diabetes that was well controlled, had no significant abnormalities, when compared with the controls. These observations suggest an association of poor diabetic control with histopathological alterations in the ENS (51). In another study with 28 full-thickness antral biopsies obtained from patients with refractory gastroparesis (14 each of type 1 DM & idiopathic gastroparesis), a significant reduction in the numbers of ganglia, NOS⁺ and NOS⁻ nerve cells was seen compared to controls (52). On the contrary, in the largest study to date carried out by the NIDDK GpCRC defined cellular changes using gastric biopsies obtained from control, DG and IG patients (20 patients in each group). There was only a 14–18% decrease in the expression of PGP9.5 in gastroparesis as compared to the controls (53). These biopsies were obtained from the gastric body in contrast to the antrum in the previous study. nNOS was found to be decreased in 20% of patients with DG and 40% of patients with IG. Nonetheless, overall quantification of nNOS did not yield a significant difference between any of the groups (53). Likewise, 20% of patients with DG and 15% of patients with IG displayed a reduction in the expression of VIP immunolabelling but, no statistically significant difference was found between the three groups (53). Immunolabeling for SP was found to be increased in one of 20 patients with DG while, decreased in four of 20 patients with DG and in two of 20 patients with IG. No statistically significant difference in SP expression between the groups was observed (53). These results suggest that alterations in the number of enteric nerves and neurons may be only present in a subset of gastroparesis patients and may depend on the gastroparesis etiology, disease duration and site of assessment in the stomach.

An ultrastructural examination of full-thickness gastric biopsies from gastroparetic patients conducted by Faussone-Pellegrini et al. showed evidence for cellular damage to the nerves, even in patients with no apparent histological changes on light microscopy (54). These include a loss of synaptic vesicles, thickened basal lamina and fibrosis around nerves. IG patients showed a more severe damage as compared to DG (54).

Smooth muscle changes associated with gastroparesis:

In a report by Xue et al., (55) smooth muscles of STZ induced diabetic rats showed antral muscle inactivity, reduced norepinephrine sensitivity, decreased Na-K pump activity and increased sensitivity to acetylcholine without any change in their resting membrane potential. In a recent report by Herring et al., ablating the forkhead transcription factors - FOXF1 and FOXF2, in adult SMCs of mice resulted in impaired gastric emptying (56). In parallel, the authors have also observed a reduced expression of FOXF1 and FOXF2 in full-thickness gastric biopsies obtained from gastroparesis patients, suggesting the possible importance of FOXF1 and FOXF2 in normal gastric function (56). Interestingly, forkhead transcription factors are well known for their involvement in smooth muscle proliferation, migration and apoptosis (57). These results suggested that in addition to neuropathy, diabetes can also induce multiple alterations in gastric smooth muscles which can contribute to delayed gastric emptying (55).

In a study by Ejskjaer et al. (58), gastric histopathology of four type 1 DM patients with gastroparesis showed smooth muscle degeneration and fibrosis, with eosinophilic inclusion bodies. In a study of two patients with DG (51), histological examination of full thickness gastric biopsy from a patient with poorly controlled, long standing diabetes showed substantial increase in fibrosis in both circular and longitudinal muscle layers and, also around the myenteric plexus while, the same was not observed in another patient with well controlled diabetes over a short duration. This indicates that clinical course of diabetes may influence the smooth muscle changes (51). However, no significant fibrosis in the gastric wall was observed in the full-thickness gastric body biopsy specimens obtained from both DG and IG patients enrolled in the NIDDK GpCRC (53). Immunolabeling for smoothelin-A – a specific marker for SMCs, showed a reduced expression in 23% of patients. However, the electron microscopy study failed to show any marked abnormalities in SMCs in both DG and IG (53).

ICC changes associated with gastroparesis:

An initial study in an animal model of diabetes (NOD/LtJ mice) conducted by Ordog et al., (59) indicated a significant delay in gastric emptying of the meal (Control: 96 ± 1%, Diabetic: 45 ± 16%; p<0.02). The delay in gastric emptying was associated with reduced ICC content (~50% reduction; p<0.05) in the antral muscles as indicated by decreased Kit-immunoreactivity, and a lack of association between ICC and enteric nerve terminals was also observed (59). The study suggested loss of pacemaker ICCs being the basis of DG. A study using STZ induced rat model of diabetes showed reduced density of ICC in antrum and this effect was also associated with loss of synaptic connections and decreased gastric emptying (60).

Similar results depicting the reduction in ICC content in human biopsy specimens of jejunum (48, 61), muscle layer of colon (62, 63) and antrum (64) were reported in gastroparesis patients with or without diabetes. The diabetes induced loss of ICCs was not due to hyperglycemia; instead the effect was due to reduced insulin and insulin like growth factor 1 (IGF-1) signaling (65). This indicates that, ICCs require insulin or IGF-1 for their maintenance. There are several reports depicting reduced Kit expression, indicating partial or total loss of ICCs in gastric biopsy specimens as one of the most common abnormality in gastroparesis patients. In a study by the NIDDK GpCRC (53), 50% of patients either with DG or IG had more than 25% reduction in Kit expression. Both DG and IG patients exhibited similar loss with a mean ICC count of ~50% compared to controls. A well-defined myenteric plexus of ICC was not seen in the human stomach in contrast to the mouse stomach. On ultrastructure, all DG and IG patients showed signs of ICC damage. These include apoptotic features, intracytoplasmatic vacuoles, swollen mitochondria and extended rough endoplasmic reticulum. As compared to controls, a physical separation of ICC from other ICC and nerves was seen. Both intramuscular and myenteric ICC had similar changes (54). In a report by Lin et al., (66) analysis of antral biopsies from 41 patients with refractory gastroparesis (34 diabetic, five idiopathic and two postsurgical) showed severe loss of ICCs in 36% of patients while, rest of the patients had no visible ICCs identified (66). Similarly, patients enrolled in the GpCRC had >60% loss of ICC, more prominent that the loss seen in gastric body (67). These studies from the GpCRC and others suggest that ICC loss is a hallmark feature of gastroparesis in humans.

Changes in macrophage population associated with gastroparesis:

Recently, macrophage dysregulation has been found to be a key factor in the pathogenesis of gastroparesis. Various stimuli and cytokines present in the tissue microenvironment are known to polarize macrophages into a spectrum which has been dichotomized as M1 (pro-inflammatory) or M2 (anti-inflammatory) (68). A number of disease states have been associated with an alteration in the macrophage milieu (69, 70). Expression of various surface markers has been used to differentiate the M1 and M2 spectrum of macrophages. For instance, M1 macrophages express TNF-α, IL-1β and iNOS whereas, M2 macrophages predominantly express the mannose receptor or CD206, Arginase-1 (mouse only) and TGF-β (71). In the context of gastroparesis, there is a reduction in the number of anti-inflammatory CD206+ macrophages in the muscularis layers (72). CD206+ macrophages are well known to express heme oxygnase 1 (HO-1) and induction of the enzyme HO-1 is a cellular defense mechanism against oxidative stress (73). In the NOD mouse model, gastric macrophages (CD206+) displayed upregulated expression of HO-1 during early diabetes and the levels were consistently maintained in the animals that were resistant to develop delayed gastric emptying (73). However, in mice developing delayed gastric emptying, there was loss of CD206+ macrophages expressing HO-1 in addition to the loss of Kit (ICC) (73). Also, loss of HO-1 expression increased oxidative stress in mice (73) and treating these mice with hemin (73) or IL-10 (47) resulted in HO-1 induction and restoration of Kit expression, indicating that CD206+ macrophages expressing HO-1 are crucial for the prevention of DG.

Mice lacking macrophages (CSF1^op/op mice) do not develop delayed gastric emptying despite severe diabetes (74). Supplementing CSF1^op/op mice with an intraperitoneal injection of CSF1 restores the muscularis macrophages and makes the susceptible to developing gastroparesis upon induction of diabetes (75). In an in vitro study, treating cultured ICCs with conditioned media from M1 macrophages led to a reduction in ICC numbers by 41% whereas, no change in ICC was observed when treated with conditioned medium from M2 macrophages (76). Of the 40 markers tested by immunoblot, M1 macrophage conditioned medium was found to be having increased amounts of 12 pro-inflammatory cytokines and chemokines when compared with the M2 conditioned medium (76). These results indicate that, the presence of macrophages and their phenotype plays a central role in the onset and progression of gastroparesis. M2 macrophages protect ICC while M1 macrophages are required to damage ICC and result in delayed gastric emptying. The molecular mechanisms that are mediated by macrophages in orchestrating ENS and other cellular abnormalities still need to be understood.

In an initial study focusing on changes in immune cell population in full thickness gastric body biopsies obtained from patients with gastroparesis showed a mild lymphocytic infiltrate in the myenteric plexus as determined by H&E staining. Most of these infiltrated cells were positive for CD45 and CD3 (52). Further reports by the NIDDK GpCRC on full thickness gastric biopsies of the antrum (67) but not body (77) from DG and IG patients showed a significant reduction in the total number of CD206+ macrophages. The total population of immune cells (CD45+) was unchanged (67). Remarkably, the reduction in the number of CD206+ macrophages correlated significantly with the loss of ICCs in circular muscle of gastric body and antrum tissues (67, 77), suggesting an interaction between CD206+ macrophages and maintenance of ICC.

Deep molecular profiling in gastroparesis:

The reports describing molecular changes associated with the pathogenesis of human gastroparesis are scarce. A recent study by the NIDDK GpCRC describes the differentially expressing genes (log₂fold difference | ≥ 2|, FDR < 5%) in DG and IG by deep transcriptomic profiling of full-thickness muscle from the gastric body (78). Wherein, 111 genes in DG and 181 genes in IG were observed to be differentially expressed. Sixty-five of these genes were common in DG and IG. The Ingenuity Pathway Analysis showed most of the differentially expressing genes to be listed in the top five canonical pathways associated with immune signaling (78). Immune profile analysis using CIBERSORT revealed that genes associated with M1 (pro inflammatory) macrophages were enriched in tissues from IG tissues compared to controls. A recent report by Herring et al., describes the increased expression of MYH11, MYLK1 and ACTA2 - mRNAs encoding contractile proteins, in smooth muscle tissues obtained from IG patients as compared to lean controls (79). And the authors have also observed a decrease in platelet-derived growth factor receptor alpha (PDGFRα) and its ligand - PDGFB mRNA expression. An interaction with BMI was observed in this study and the lean BMI group had more males whereas the IG group was predominantly females. In our work on the gastric body, the numbers or distribution of PDGFRα expressing fibroblast-like cells was not altered in gastroparesis (80).

Recently, we have performed quantitative proteomic analysis of full-thickness gastric antrum biopsies from DG and IG patients using aptamer-based SomaLogic tissue scan that quantitatively identifies 1300 human proteins (81). We have found 73 proteins differentially expressing in DG, 133 proteins differentially expressing in IG and, 40 differentially expressing proteins were common between both DG and IG. The study also complements the results of deep transcriptomic profiling study (78), wherein “Role of Macrophages, Fibroblasts and Endothelial Cells” is the most statistically significant altered pathway. In summary, RNA- and protein-based molecular signatures complement histological studies in suggesting an immune based dysregulation and injury to the ENS in gastroparesis.

Correlation with clinical symptoms and gastric emptying:

A large study showed ICC loss to positively correlate with gastric retention of solids in patients with DG (82). Another study showed delayed gastric emptying to be associated with ICC counts in myenteric plexus (83). In a study by Forster et al., (64) gastroparesis patients exhibiting no or depleted ICCs (ICC- group) had significantly higher tachygastria when compared with patients with some or adequate number of ICCs (ICC+ group). Tachygastria in ICC- group was also associated with reduced postprandial rhythmic function. Along with these, the total symptom score in ICC- patients was significantly higher when compared with ICC+ group at baseline and also after 3 months of using gastric electrical stimulator (64). This was subsequently shown again in a larger study by the same group (66). The study from the GpCRC also found correlation between nausea and overall symptom severity and abundance of myenteric immune cells (82). Overall the correlation between symptoms and cellular defect is low suggesting that once established, symptoms may persist independently of the original cause. Use of next generation sequencing complemented by targeted validation can help identify molecules that can serve as biomarkers and allow target discovery for treatment of gastroparesis.

Future directions:

Animal models have played a major role in providing us valuable information on the pathogenesis of delayed gastric emptying. Understanding of human gastroparesis is driven by work on full thickness gastric biopsies which has found several similarities between diabetic and idiopathic gastroparesis. An important shift in paradigm has been the understanding of innate immune system in driving injury to the ENS and ICC. A challenge to approaching immune-ENS interactions is a lack of atlas for either of these cell types in the gastric muscle layers. There are differences in the expression pattern of genotypic markers in mouse and human immune cells (84). Future studies will need to generate cellular and molecular atlases from healthy subjects that can be then used to compare the gastroparesis patients. From a diagnostic standpoint, we need to further develop minimally invasive methods for procuring gastric full-thickness tissue. In a recent study by Rajan et al.,(85) a no-hole gastric body biopsy using an over-the-scope clip was technically successful and provided adequate tissue (~1 cm) for ENS assessment. Endoscopic ultrasound based approaches have also been tried but the yield for full-thickness assessment of the ENS is low (86). From a therapeutic standpoint, strategies to influence macrophage polarization or regenerate gastroparesis may act a disease-modifying treatment for gastroparesis.

Paradigms that include identification of gastroparesis patients with changes in ENS and immune cell phenotypes followed by targeted therapies can be particularly helpful in altering the natural history as compared to current therapy which is focused solely on symptom alleviation.

Key points:

• Insulinopenic hyperglycemic animal models have provided significant insight into the pathophysiology of diabetic gastroparesis. Interstitial cells of Cajal (ICC) loss or damage is the key cellular abnormality leading to delay in gastric emptying in gastroparesis.

• A macrophage-based immune dysregulation leading to injury to ICC and other components of ENS is emerging as an important precursor step in the pathophysiology of gastroparesis.

• Full-thickness gastric biopsies obtained from patients in the NIDDK Gastroparesis Clinical Research Consortium (GpCRC) have been instrumental in delineating disease mechanisms and corroborating hypotheses generated in preclinical models.

• Future strategies employing minimally invasive methodologies for tissue procurement and use of systems biology approaches will help in identifying targets for diagnosis, prognosis and treatment of diseases associated with abnormal gastric function.

Synopsis.

The cellular and molecular understanding of human gastroparesis has markedly improved due to studies on full-thickness gastric biopsies. A decrease in number of interstitial cells of Cajal (ICC) and functional changes in ICC constitutes the hallmark cellular feature of gastroparesis. More recently, in animal models, macrophages have also been identified to play a central role in development of delayed gastric emptying. Activation of macrophages leads to loss of ICC. In human gastroparesis, loss of anti-inflammatory macrophages in gastric muscle has been shown. Deeper molecular characterization using transcriptomics and proteomics has identified macrophage-based immune dysregulation in human gastroparesis.

Abbriviation:

Ano-1

anoctamin-1

BH4

tetrahydrobiopterin

BMP-2

bone morphogenetic protein-2

BMPR

BMP receptor

ChAT

choline acetyltransferase

CSF-1

colony stimulating fator-1

CX3CR1

CX3C chemokine receptor 1

DG

diabetic gastroparesis

DM

diabetes mellitus

ENS

enteric nervous system

GI

gastrointestinal

HO-1

heme oxygnase 1

ICC

interstitial cells of Cajal

IG

idiopathic gastroparesis

IGF-1

insulin like growth factor 1

IL-1β

interleukin 1β

iNOS

inhibitory nitric oxide synthase

IP3

inositol triphosphate

MHCII

major histocompatibility complex II

nNOS

neuronal nitric oxide synthase

NO

nitric oxide

NOD

non-obese diabetic

NPY

neuropeptide Y

op/op

osteopetrotic mouse

PKC

protein kinase C

SMCs

smooth muscle cells

SP

substance P

STZ

streptozotocin

TGF-β

transforming growth factor β

TNF-α

tumor necrosis factor α

VIP

vasoactive intestinal polypeptide