The orderly differentiation of cell lineages within gastric glands is regulated by a complicated interplay of local mucosal growth factors and hormones. Histamine secreted from enterochromaffin-like cells plays an important role in not only stimulated gastric acid secretion but also coordination of intramucosal growth and lineage differentiation. We have examined histidine-decarboxylase (HDC)-deficient mice, which lack endogenous histamine synthesis, to evaluate the influence of histamine on differentiation of fundic mucosal lineages and the development of metaplasia following induction of acute oxyntic atrophy. Stomachs from HDC-deficient mice and wild-type mice were evaluated at 8 wk and 12 mo of age. DMP-777 was administrated orally to 6-wk-old mice for 1 to 14 days. Sections of gastric mucosa were stained with antibodies against Mist1, intrinsic factor, H/K-ATPase, trefoil factor 2 (TFF2), chromogranin A, and Ext1 and for the cell cycle marker phospho-histone H3. HDC-deficient mice at 8 wk of age demonstrated a prominent increase in chief cells expressing Mist1 and intrinsic factor. Importantly Mist1-positive mature chief cells were present in the midgland region as well as at the bases of fundic glands, indicating a premature differentiation of chief cells. Mice dually deficient for both HDC and gastrin showed a normal distribution of chief cells in fundic glands. Treatment of HDC-deficient mice with DMP-777 led to loss of parietal cells and an accelerated and exaggerated emergence of mucous cell metaplasia with the presence of dual intrinsic factor and TFF2-expressing cells throughout the gland length, indicative of the emergence of spasmolytic polypeptide-expressing metaplasia (SPEM) from chief cells. These findings indicate that histamine, in concert with gastrin, regulates the appropriate differentiation of chief cells from mucous neck cells as they migrate toward the bases of fundic glands. Nevertheless, histamine is not required for emergence of SPEM following acute oxyntic atrophy.
- enterochromaffin-like cell
- spasmolytic polypeptide-expressing metaplasia
in the normal gastric fundic mucosa, progenitor cells located in the upper gland neck give rise to four types of epithelial cells including pepsinogen-secreting zymogenic chief cells, acid-producing parietal cells, and two types of mucous cells, namely surface mucous cells and mucous neck cells. These lineages differentiate from three second-order progenitor cells, prepit, preparietal, and preneck cells (7–9). Prepit cells give rise to surface mucous cells, which migrate toward the lumen and secrete trefoil factor family 1 and mucin 5 types A and C. Parietal cells differentiate from preparietal cells and in mice most of these acid-secreting cells migrate toward the base. The mucous neck cells differentiate from preneck cells and secrete spasmolytic polypeptide/trefoil factor 2 (TFF2) and MUC6. As mucous neck cells migrate toward the bases of fundic glands, they redifferentiate into zymogenic chief cells, which secrete both pepsinogen and intrinsic factor in rodents (9). The prezymogenic cells display granules showing features intermediate between those of neck cells and zymogenic chief cells (9, 25). Importantly, the transition between mucous neck cells and chief cells occurs without an intermediate transiently amplifying cell and appears to involve the induction of the expression of the transcription factor Mist1 (25).
Recent investigations have noted the critical influence of intrinsic mucosal growth factors on the differentiation of fundic cell lineages. A number of studies have suggested that gastrin, released from antral G cells, leads to the expansion of parietal cell lineages (11, 18). Gastrin is also the major driver for the expansion of surface cell lineages in hypergastrinemic states (20). Similarly, elevation of transforming growth factor (TGF)-α, normally secreted by parietal cells, surface cells, and enterochromaffin-like (ECL) cells, leads to a marked expansion of surface cells and a reciprocal decrement in the differentiation of lineages in the deep glands (i.e., parietal cells, chief cells, and mucous neck cells) (2, 3, 5, 26). Parietal cells secrete a number of other growth factors including the EGF receptor ligands, amphiregulin (1), heparin-binding EGF (15), as well as sonic hedgehog (30). Loss of parietal cells leads to a number of alterations in gastric lineages including an inhibition of mucous neck cell to chief cell differentiation as well as transdifferentiation of chief cell in spasmolytic polypeptide-expressing metaplasia (SPEM) (17, 20, 21). The loss of either gastrin (20) or amphiregulin (17) or attenuation of EGF-receptor signaling (22) leads to accelerated emergence of SPEM following acute oxyntic atrophy. Thus loss of intrinsic growth factors may lead to global alterations in intramucosal signaling that are necessary for the normal response of the mucosa to injury or lineage loss.
Fundic endocrine cell lineages can also influence through release of hormones such as somatostatin and histamine. Somatostatin-deficient mice develop hyperplastic polyps (29). H2-histamine receptor knockout mice develop a hypertrophic gastropathy with marked foveolar hyperplasia (23). Furthermore, previous investigations of mice with targeted deletion of histidine decarboxylase (HDC), the key enzyme for the production of histamine, have shown an increase in ECL cell numbers and expansion of parietal cells that appear small and less mature (16). Although previous investigations have suggested that parietal cell-derived influences are required for maturation of chief cells (12), no investigations have identified particular paracrine influences that promote chief cell differentiation. We have now sought to characterize the lineage changes in HDC-deficient mice in more depth. Our studies have revealed a previously unrecognized marked increase in chief cell numbers in HDC-deficient mice. HDC-deficient mice showed a pattern of premature differentiation of mucous neck cells into mature chief cells in the neck region of the fundic glands. However, dual knockout mice deficient in both histamine and gastrin showed an amelioration of the premature chief cell differentiation phenotype. The chief cells that differentiated in neck region still appear to be susceptible to transdifferentiation into SPEM following acute oxyntic atrophy. All of these results indicate that the process of chief cell differentiation from mucous neck cells is regulated by the combined influence of both histamine and gastrin.
MATERIALS AND METHODS
HDC-deficient mice were genotyped as previously described (24) (primers used were sense, 5′-GAG CAC TGT CAG CGA ATC CAC-3′; wild-type allele antisense, 5′-GGC CGT GAG ATA AGC GTG ACC-3′; and HDC-deficient allele antisense, 5′-TGG GAT TAG ATA AAT GCC TGC TCT-3′). Gastrin-deficient mice were genotyped as previously described (20) (primers used were wild-type allele sense, 5′-TCC ATG CCT CTT TGT TGT TG-3′; gastrin-deficient allele sense, 5′-TCG TCA AGA AGG CGA TAG-3′; and antisense, 5′-CCA GAG GTA AAG GGC TGA CCA G-3′). HDC-deficient mice and gastrin-deficient mice were crossed, and dual HDC/gastrin-deficient mice (HDC−/−/Gastrin−/−) were established by genotyping.
HDC-deficient mice and Balb/c mice (Jackson Laboratories, Bar Harbor, ME) were euthanized at 6 wk, 2 mo, 6 mo, or 12 mo of age. HDC−/−/Gastrin−/− dual knockout mice were euthanized at 8 wk of age. DMP-777 (a gift of DuPont Pharmaceuticals), formulated in 0.5% methylcellulose, was administrated orally to 6-wk-old mice (6 per group) as a gavage once daily at 350 mg/kg per day. Stomachs were excised and opened along the greater curvature and fixed with 4% paraformaldehyde and embedded in paraffin for immunohistochemical analysis. All breeding and procedures were performed under Vanderbilt University IACUC-approved animal protocols.
Excised stomachs embedded in paraffin were used for immunohistochemistry analysis. Deparaffinized sections were pretreated with antigen retrieval using Target Retrieval solution (Dako Cytomation, Glostrup, Denmark) or Trilogy antigen retrieval solution (Cell Marque, Austin, TX) at 120°C for 15 min, followed by immediate cooling using iced water. Sections were then treated with 2% blocking serum and incubated with the primary antibody overnight at 4°C. Immunostaining was performed with the following primary antibodies: murine monoclonal immunoglobulin M (IgM) anti-TFF2 (a gift from Sir Nicholas Wright; Cancer Research, London, UK; 1:1,000), rabbit anti-Mist1 (a gift from Dr. Jason C. Mills; Washington University, St Louis, MO; 1:2,000), murine monoclonal anti-H/K-ATPase (a gift from Dr. Adam J. Smolka; Medical University of South Carolina, Charleston, SC; 1:100,000), rabbit anti-chromogranin A (Zymed Laboratories, San Francisco, CA; 1:500), rabbit anti-exostoses 1 (Ext1; ProteinTech, Chicago, IL; 1:1,000), murine monoclonal anti-phospho-histone H3 (Cell Signaling Technology, Danvers, MA; 1:1,000), and rabbit anti-intrinsic factor (a gift from Dr. David Alpers; Washington University; 1:2,000). For immunohistochemistry with detection with diaminobenzidine, the sections were incubated with biotinylated secondary antibody followed by horseradish peroxidase-conjugated streptavidin. Chromogen was developed with diaminobenzidine (BioGenex Laboratories, San Ramon, CA). For immunohistochemistry with alkaline-phosphatase detection, the sections were incubated with biotinylated secondary antibody followed by alkaline phosphatase-conjugated avidin-biotin complex. Chromogen was developed with Vector Red Substrate (Vector Laboratories, Burlingame, CA). All sections were counterstained with Meyer's hematoxylin. For immunofluorescence analysis, Cy3-goat anti-mouse IgM antibody, Cy5-goat anti-mouse IgG antibody (Jackson ImmunoResearch, West Grove, PA), and Alexa488 goat anti-rabbit IgG (Invitrogen, Carlsbad, CA) were used, and ProLong Gold Antifade Reagent with DAPI (Invitrogen) was used for nuclear counterstain and mounting medium. Sections were viewed and photographed on a Zeiss Axiophot microscope equipped with an AxioVision digital imaging system (Zeiss, Jena, Germany) or a FluoView FV1000 confocal microscope system (Olympus, Tokyo, Japan).
For quantitation of early maturation of chief cells, numbers of cells expressing only intrinsic factor cells between the first and last TFF2-staining mucous neck cells were determined for 10 glands in dual-labeled sections from three animals in each group (wild-type, HDC-deficient, and dual HDC/gastrin-deficient mice). Statistically significant differences were determined by a Student's t-test.
For quantitation of cell numbers, well-oriented sections from three animals from each group (wild-type, days 0, 1, 3, 7, and 14 of treatment; HDC knockout, days 0, 1, 3, 7, and 14 of treatment) were analyzed. Three gland units from the lesser curvature of the fundic mucosa, in each slide, were counted under fluorescent microscope (Zeiss). The average and standard deviation for cell numbers for each cell type were determined, and statistically significant differences were analyzed by Mann-Whitney U-test.
HDC-deficient mice have prominent increases in ECL cells.
Previous investigations of HDC-deficient mice demonstrated mucosal hypertrophy with increases in parietal cells and ECL cells (16). We observed a similar pattern of hypertrophy in the mucosa of all HDC knockout mice with a prominent increase in mucosal height in 8-wk-old animals (Fig. 1, A and B). We noted a prominent increase in ECL cells marked by both chromogranin A (Fig. 1, C and D) and Ext1 staining (Fig. 1, E and F). Whereas ECL cells in wild-type mice were distributed more toward the bases of fundic glands, in HDC-deficient mice the ECL cells were distributed more broadly along the gland length. Nevertheless, we did not observe any instances of either nodular or linear hyperplasia patterns.
HDC-deficient mice show a pattern of premature chief cell differentiation.
In examining the fluorescence stains for lineages in 8-wk-old mice, it became clear that, in addition to the broader distribution of ECL cells, there were significant alterations in other lineages. Most notably, there appeared to be two zones of TFF2-expressing mucous neck cells (Fig. 2). The mucous neck cells in both zones were morphologically normal, but there also was a zone of intermediate cells that did not stain for TFF2; also, not all of these cells were accounted for by H/K-ATPase-expressing parietal cells (Fig. 2E). We therefore investigated whether the intervening cells might be differentiated chief cells by assessing staining with antibodies against intrinsic factor (Fig. 3), which is only expressed the zymogen granules of mature chief cells (25). In wild-type mice, intrinsic factor-expressing chief cells were observed only at the bases of glands, completely separate from TFF2-expressing mucous neck cells in the midgland (Fig. 3, A–C). As noted previously, this phenotype is consistent with the differentiation of mucous neck into chief cells during migration toward the bases of glands (9, 25). In contrast, in 8-wk-old HDC-deficient mice, intrinsic factor-expressing chief cells were noted in both the midgland and at the bases of fundic glands (Fig. 3, D–F). This pattern of early chief cell differentiation was confirmed using staining for Mist1, a transcription factor, which is only expressed in mature chief cells (Supplemental Fig. S1). Supplemental information for this article is available at the American Journal of Physiology Gastrointestinal and Liver Physiology website. To quantitate this alteration in cell differentiation along the gland axis, we determined the number of intrinsic factor-positive cells lying between the first and last TFF2-expressing mucous neck cell in fundic glands from wild-type and HDC-deficient mice. Figure 4 demonstrates that, although few mature chief cells expressing intrinsic factor were identified within the mucous neck cell zone in wild-type mice, many mature chief cells were observed within the mucous neck cell region in HDC-deficient mice.
We also examined whether the mucosal phenotype could be explained by reactivation of mucosal progenitor cells. Supplemental Fig. S2 demonstrates that phospho-histone H3-expressing progenitor cells in both wild-type and HDC-deficient mice were only observed in the neck region of fundic glands. Similar results were also observed with Ki-67 staining (data not shown). These results suggested that a population of chief cells was maturing prematurely in the midgland region in the fundic mucosa of HDC-deficient mice.
The secretion of histamine and gastrin are coregulated in the gastric mucosa to provide control of acid secretion. Previous investigations have determined that HDC-deficient mice have profound hypergastrinemia (16, 27). Therefore, to analyze the influence of gastrin on the phenotype in histamine-deficient mice, we also examined the fundic mucosa of mice dually deficient in both histamine and gastrin (HDC−/− × gastrin−/− mice). In these dual knockout mice, we observed a normal distribution of lineages. Mucous neck cells were located in the midgland region distinctly separated from intrinsic factor-expressing chief cells that were located predominantly in the deep gland region (Fig. 3, G–I). Although occasional intrinsic factor-expressing mature chief cells were still observed in the midgland, the cross of the HDC-deficient mice onto the gastrin knockout background significantly ameliorated the phenotypic aberrations in the HDC-deficient mice as reflected by a decrease in intrinsic factor-expressing cells in the mucous neck cell zone of fundic glands (Fig. 4). Similar results were also seen with Mist1 staining (Supplemental Fig. S1). These results indicate that the premature maturation of chief cells accrued from the coordinate influences of both a loss of histamine and hypergastrinemia.
Effects of acute parietal cell loss in HDC-deficient mice.
We have previously noted that acute loss of parietal cells using treatment with oral gavage of DMP-777 leads to mucous cell metaplasia through transdifferentiation of chief cells into SPEM. Following acute parietal cell losses, we observed marked increases in gastrin (4, 20), and, because gastrin strongly stimulates histamine release, one would expect to observe elevated levels of tissue histamine release. Whereas we have recently demonstrated that loss of either gastrin or amphiregulin can promote the formation of SPEM (17, 20), the influence of histamine in promotion of metaplasia has not been examined. We therefore administered DMP-777 to HDC-deficient mice for 1–14 days (Fig. 5). As with wild-type mice, DMP-777 elicited a rapid loss of parietal cells that was maximal by 3 days of treatment (Fig. 5A). Although HDC-deficient mice showed almost twice the number of intrinsic factor-labeling chief cells compared with wild-type mice before treatment, in both groups we observed only small changes in intrinsic factor-expressing cells with DMP-777 treatment. We have previously noted that the appearance of cells expressing both intrinsic factor and TFF2 is the best reflection of the induction of SPEM following acute oxyntic atrophy (17, 20). In HDC-deficient mice treated with DMP-777, the appearance of dual labeling cells was accelerated, and a significantly greater number of dual-expressing cells was observed in the HDC-deficient mice at all days of treatment. Figure 6 demonstrates that, although intrinsic factor-expressing chief cells were present in the untreated HDC-deficient mice, few cells were observed with dual staining for intrinsic factor and TFF2 (Fig. 6C). In contrast, in DMP-777-treated HDC-deficient mice, dual intrinsic factor/TFF2-labeling cells were observed along the entire length of the fundic glands (Fig. 6, F and I). These results suggest that histamine is not required for development of metaplasia induced by acute parietal cell loss. In addition, the findings indicate that chief cells that have matured prematurely in the neck region also can develop into metaplastic SPEM cells.
Older HDC-deficient mice show hyperplastic and metaplastic changes in the mucosa.
Previous investigations have reported that 9-mo-old HDC-deficient mice can develop hyperplastic changes in the fundus (16). We also examined HDC-deficient mice at over 1 yr of age. We did observe that all older mice showed regions of hyperplastic polyps along the greater curvature containing numerous parietal cells, chief cells, and mucous neck cells (Supplemental Fig. S3). Nevertheless, in 20% of the mice, we also observed other regions, especially along the lesser curvature, which showed marked mucous cell metaplasia with glands dominated by mucous cells that were reactive for both diastase-resistant periodic acid-Schiff (PAS) and Alcian blue (Fig. 7, A–D). We have previously noted this phenotype in some forms of SPEM, and indeed the mucous cell metaplasia was strongly positive for TFF2 (Fig. 7E). In contrast with the hyperplastic polyps on the greater curvature, these glands contained few parietal cells (Fig. 7F). Chromogranin A-labeled ECL cells were observed along the gland length intermixed with the metaplasia (Fig. 7, G and H). These results indicate that chronic absence of histamine in HDC-deficient may lead to global changes of both hyperplasia and metaplasia.
Investigations over the past several years have led to the realization that cell lineage differentiation in fundic gastric glands is critically influenced by both hormonal factors as well as intrinsic mucosal growth factors. Intramucosal growth factors may have variable influences that are spatially heterogeneous as cells migrate from the progenitor zone in the neck to regions in the deeper glands. In particular, mucous neck cells differentiate initially in the neck from preneck cells and undergo a further differentiation into chief cells as they migrate toward the gland base (9). Importantly, mucous neck cell redifferentiation into chief cells occurs without any proliferating cell intermediate although a morphological prezymogenic cell can be identified (9, 25). Maturation of chief cells requires the expression of the transcription factor Mist1, and loss of Mist1 leads to a failure of complete differentiation of zymogenic cells in the deep glands (25). As chief cells migrate toward the base, they come under the influence of secreted paracrine factors from other lineages including parietal cells and ECL cells. These factors include a number of EGF receptor ligands and sonic hedgehog from parietal cells, as well as histamine and other growth factors secreted from ECL cells. A number of studies have indicated that the loss of parietal cells can alter the full differentiation of chief cells (2, 12). However, the present investigation suggests that endocrine and intrinsic mucosal histamine production also have important influences on the differentiation of chief cells. Indeed, loss of histamine led to premature differentiation of chief cells in the neck region, before full migration to the base. The effects of histamine loss were also dependent on elevation of gastrin because breeding of HDC-deficient mice onto the gastrin-null background led to an amelioration of the premature chief cell differentiation phenotype. All of these results support the concept that the orchestration of fundic gland cell lineage differentiation requires a complex and coordinated influence of intrinsic and extrinsic growth factors.
Few studies have addressed the role of histamine as a regulator of gastric cell proliferation and differentiation. Previous investigations have suggested that histamine can regulate the proliferation of ECL cells (14). Pharmacological inhibition of the H2-receptor leads to inhibition of acid secretion and elevated levels of gastrin but is not associated with alteration in chief cell lineages. Similarly, knockout of the H2-histamine receptor leads to marked elevations in gastrin as well as increases in TGF-α and massive foveolar hyperplasia (16). This phenotype has been compared with that for transgenic overexpression of TGF-α, which also leads to massive foveolar hyperplasia (2, 5, 19, 26).
Nevertheless, these mice do not show alterations in chief cell differentiation as seen in the HDC-deficient mice. All of these mice, as well as other models associated with hypergastrinemia, do not demonstrate early differentiation of chief cells (13). Importantly, because all of these models elevate gastrin, they would also be expected to raise histamine secretion. Thus most models of hypergastrinemia are likely dual models for hypergastrinemia with elevated histamine release (Fig. 8). It is interesting to note that, although alterations in chief cell differentiation are not observed in hypergastrinemic insulin-gastrin transgenic mice at an early age, these mice do go on at a later age to develop SPEM and gastritis cystica profunda (28). In the HDC-deficient mouse, the gastric mucosa experiences hypergastrinemia in the absence of histamine (16) (Fig. 8B). Interestingly, our studies have shown that the phenotype of premature differentiation and maturation in chief cells in the fundic gland neck was substantially ameliorated in dual HDC- and gastrin-deficient mice. These results suggest that the phenotype observed in HDC-deficient mice results from an imbalance in the influences of histamine and gastrin. Since a similar phenotype is not apparently observed in H2-receptor knockout mice (13, 23), it seems likely that the loss of histamine may lead to alterations in signaling through another histamine receptor. Perhaps the most likely candidate for this action would be through the H3-histamine receptor. The H3-histamine receptor is an autoreceptor on ECL cells and may regulate coordinated release of other ECL cell-paracrine factors (6, 10). Alternatively, the loss of histamine secretion may in turn change the secretory pattern of other growth factors from parietal cells. In any case, these results again suggest that loss of one paracrine regulator such as histamine in the stomach can lead to global effects on amine, peptide, or growth factor secretion through various feed back loops. Disruption of these signaling loops alters the balance of lineage differentiation within gland units.
Previous studies have shown that acute oxyntic atrophy from treatment with the parietal cell-toxic compound DMP-777 leads to a rapid rise in serum gastrin and marked alterations in mucosal dynamics, leading to the emergence of SPEM (4, 20, 21). We have previously noted that SPEM developed more rapidly following DMP-777 treatment in gastrin- or amphiregulin-deficient mice (17, 20). Similarly, in HDC-deficient mice, acute oxyntic atrophy led to more rapid induction of SPEM and a greater amount of metaplasia within glands. The latter finding likely accrued from the increased baseline numbers of mature chief cells in the fundic glands of HDC-deficient mice. Accumulating evidence suggests that SPEM arises from both transdifferentiation of mature chief cells into mucous cell metaplasia as well as an arrest of mucous neck cell differentiation into chief cells (17, 21). The results presented here indicate that mature Mist1-expressing chief cells, regardless of their position in the fundic glands of HDC-deficient mice, can undergo transdifferentiation manifested by dual expression of both intrinsic factor and TFF2 in separate granules within SPEM cells. Thus the ability to undergo transdifferentiation into mucous cell metaplasia appears to be an intrinsic characteristic of mature chief cells. Interestingly, whereas mucosal hyperplasia was the most characteristic pathological finding in older HDC-deficient mice, 20% of mice over 1 yr of age also showed metaplastic polyps along the lesser curvature. Thus it is possible that the expansion of the chief cell population may predispose over time to the development of oxyntic atrophy and metaplasia.
In summary, whereas previous investigations in HDC-deficient mice had documented age-dependent fundic mucosal hyperplasia, the present studies document that the absence of intramucosal histamine production also leads to a remarkable premature differentiation in chief cells from mucous neck cells. HDC-deficient mice demonstrate a phenotype of increased mature chief cells located throughout the gastric gland. Although gastrin appears necessary for the development of this phenotype, it is does not appear to represent the sole mediator because hypergastrinemia alone does not elicit such an effect on chief cell differentiation. The small size of parietal cells suggests that they also do not undergo complete maturation (16). It appears likely that the combination of hypergastrinemia without histamine leads to an alteration in intramucosal signaling that regulates the timely differentiation of lineages as they migrate down the fundic glands.
These studies were supported by grants to J. Goldenring from a Department of Veterans Affairs Merit Review Award, NIH RO1 DK071590, a pilot project grant from the Vanderbilt SPORE in Gastrointestinal Cancer (P50 CA95103), the AGA Funderburg Award in Gastric Biology Related to Cancer, and a Discovery Grant from the Vanderbilt-Ingram Cancer Center.
We thank Dr. Adam Smolka, Dr. Jason Mills, Dr. Steve Konieczny, Dr. Nicholas Wright, and Dr. David Alpers for the gifts of antibodies.
↵* K. Nozaki and V. Weis contributed equally to this work.
- Copyright © 2009 the American Physiological Society