Abstract
Cisplatin treatment (9 mg/kg) causes bloating of the stomach, an increase in gastric acid, and ulceration in rats. Gastrin, a gut peptide, plays an important role in regulating gastric acid production. To study the role of gastrin in this increased gastric acid production after cisplatin treatment, male Wistar rats (100–150 g) were treated with cisplatin (9 mg/kg) in five divided doses over 5 consecutive days. The rats were sacrificed 1, 6, 10, or 15 days after the last treatment. As measured by immunocytochemistry, in situ hybridization, Northern blot, and dot-blot techniques, gastrin was found to be below detectable limits just 1 day after cisplatin treatment. However, 10–15 days after the last injection, the levels for both gastrin and its mRNA gradually recovered to normal. Northern blot studies showed that decreased somatostatin mRNA parallels the changes of gastrin and its mRNA. These results suggest that after cisplatin treatment the increased gastric acid production in rat stomach is independent of gastrin. This decrease of gastrin production is not under the influence of somatostatin, which also decreased after cisplatin treatment. (J Histochem Cytochem 47:1057–1062, 1999)
Keywords
C
Under low chloride ion concentrations inside the cell, cisplatin hydrolyzes into monoaqua (monovalent) and diaqua (divalent) species. The diaqua form of cisplatin with its divalent charge has been demonstrated to bind to calmodulin and inhibit its binding to calcium. Without the calcium-calmodulin complex, acetylcholine release is inhibited, resulting in bloating of the stomach and an increase in its acid content (Jarve and Aggarwal 1997).
Gastrin, one of the gut peptides, which is primarily produced and secreted in the stomach and proximal duodenum, is a potent stimulant of gastric acid secretion and proliferation of the acid-secreting oxyntic cells of the gastric mucosa. It acts both directly on gastrin receptors of the parietal cells and indirectly on gastrin receptors of the enterochromaffin-like (ECL) cells to produce histamine which, in turn, stimulates the gastric acid production by parietal cells. Somatostatin is known to influence gastrin production in rat stomach through a paracrine action (Walsh 1994). This study was undertaken to investigate the changes in gastrin and somatostatin after cisplatin treatment and to correlate the changes to the lowered pH in rat stomach.
Materials and Methods
Animals and Tissue Preparation
Male Wistar rats (Charles River Laboratory; Wilmington, MA) weighing 100–150 g were housed on a 12-hr light/12-hr dark cycle with free access to food and water. Cisplatin at a clinical dosage of 9 mg/kg in physiological saline was injected (IP) in five divided doses over 5 days. The controls received only the injection vehicle. The rats were anesthetized with equithesin and either perfused with buffered 4% paraformaldehyde (0.1 M phosphate buffer, pH 7.4) or stomach tissues were excised and frozen (−70C) 1, 6, 10, or 15 days after last cisplatin treatment. Each interval had a minimum of three animals. Perfused stomach tissues were postfixed with buffered 4% paraformaldehyde (0.1 M phosphate buffer, pH 7.4) for 12 hr, dehydrated through ethanol gradients, and embedded in paraffin at 56C. Sections (10 (μm) were cut and placed on gelatin-coated slides for in situ hybridization (ISH) and immunocytochemical studies.
Immunocytochemistry
Immunocytochemical study for gastrin was performed using the avidin-biotin-peroxidase complex (ABC) (Hsu et al. 1981) with a Vectastain ABC-AP Kit (Vector Laboratories; Burlingame, CA). Paraffin sections were deparaffinized in xylene, hydrated, and sequentially incubated in gastrin-specific primary antibody (1:1000) (Dako; Carpinteria, CA) at 4C overnight and the biotinylated antibody for 45 min at room temperature (RT). Sections were then incubated in ABC-alkaline phosphatase (AP) complex for 45 min at (RT). AP activity was revealed by using the Vector Red Substrate Kit. After rinsing in tapwater and proper dehydration, sections were mounted in Permount and viewed for intensity differences on a visual basis. The tissues were directly stained with Vector Red Substrate to check the effect of the endogenous AP activity. The liver tissue served as a negative control.
Oligonucleotide Probes
5′-GACCTTGGGGCCCCAGCTGTCTCCGAT-3′, a 27-mer hybridization sequence complementary to the coding sequence of rat gastrin mRNA (position 212–238) (Larsson and Hougaard 1993) and 5′CCAGAAGAAGTTCTTGCAGCCAGCTTTGCGTTCCCGGGGTGCCAT-3′, a 45-mer hybridization sequence complementary to the coding region of rat somatostatin mRNA (bases 286–330) (Rage et al. 1994), were synthesized using the ABI 3948 Synthesis and Purification System. The sense probe was synthesized at the same time to serve as control. Both sense and antisense oligonucleotides were labeled with the digoxigenin (DIG) Oligonucleotide Tailing Kit (Boehringer Mannheim; Indianapolis, IN) according to the manufacturer's instructions. A 100 pmol probe was incubated with tailing buffer, CoCl2 solution, DIG-dUTP, dATP, and terminal transferase solutions at 37C for 18 min, then placed on ice and the reaction stopped by adding 4 μ1 of stop solution. The tailed probe was precipitated in 5 μ1 4M LiC1 and 150 μ1 prechilled (−20C) ethanol at −20C for 3 hr and then centrifuged at 12,000 × g for 15 min. The supernatant was discarded and the pellet was allowed to air-dry and kept at −70C till use.
In Situ Hybridization
The protocols for in situ hybridization were followed according to Larsson and Hougaard (1993). Deparaffinized sections (10 μm) were soaked in chloroform (Mallinckrodt; Paris, KY) and hydrated through descending ethanol gradients. Proteolytic treatment was carried out with 0.015% pepsin (Boehringer Mannheim; Indianapolis, IN) in 0.2 M HCl for 20 min at RT. Sections were fixed with buffered 4% paraformaldehyde (0.1 M phosphate buffer, pH 7.4) at 4C for 5 min. After washing twice with PBS, the sections were soaked in freshly made 0.25% acetic anhydride in triethanolamine (TEA) buffer (pH 8.0). The sections were pretreated for 30 min at 42C in the prehybridization buffer, which is identical to the hybridization buffer but without the probe. Hybridization was performed at 42C overnight with 50 × l hybridization buffer covering each slide and containing 5 ng of the labeled probe, 50% formamide, 1 × Denhardt's solution, 10% dextran sulfate, 10 mM dithiothreitol (DTT) (Sigma; St Louis, MO), 150 μg/ml tRNA (Boehringer Mannheim), 100 μg/ml denatured sheared salmon sperm DNA (GIBCO BRL; Gaithersburg, MD) and 3 × standard saline citrate (SSC). Stringency wash was carried out in 0.1 SSC (four changes × 20 min) at 42C. Immunocytochemical detection of DIG was performed with DIG Nucleic Acid Detection Kit (Boehringer Mannheim). Sections were washed in maleic acid buffer (pH 7.4) for 10 min, then sequentially incubated in blocking buffer in the kit for 30 min and 1:1000 anti-DIG antibody for 45 min. After washes in PBS and 100 mM Tris-HCl buffer (pH 9.5), sections were incubated in colorsubstrate solution, nitroblue tetrazolium/bromochloroindolyl phosphate (NBT/BCIP) overnight. Finally, the sections were dehydrated and mounted in Permount. All the H2O used in above study was diethyl pyrocarbonate (DEPC)-treated double-distilled water.
Statistical Analysis
Numbers of gastrin and its mRNA-staining positive cells were counted in 15 random visual fields of five different tissue sections from each group. The data were averaged and plotted. All data were statistically analyzed by the Student's t-test when comparison between control and cisplatin treatment was made (Steele and Torrie 1980).
(A) Immunocytochemical localization of gastrin in normal rat stomach (arrows), basal portion of gastric mucosa. ∗, basal lamina. Bar = 40 μm. (B) Immunocytochemical study showing complete absence of gastrin-positive cells from the pylorus of the rat gastric mucosa 1–6 days after cisplatin (9 mg/kg) treatment. Bar = 20 μm. (C) Light micrograph showing the immunocytochemical distribution of gastrin-positive cells 10 days after cisplatin treatment. Note the appearance of the gastrin-positive cells. ∗, basal lamina. Bar = 10 μm. (D) ISH study demonstrating the presence of gastrin mRNA-positive cells (arrows) in the basal portions of the pylorus of normal rat stomach. The cytoplasm of these cells is positively stained, whereas the nuclei are negative. The distribution pattern of gastrin mRNA-positive cells is the same as in the immunocytochemical study shown in A. ∗, basal lamina. Bar = 20 μm. (E) In situ hybridization study showing complete absence of gastrin mRNA-positive cells from the gastric mucosa 1–6 days after cisplatin treatment. Bar = 20 μm. (F) Light micrograph showing the gastrin mRNA-positive cells 15 days after the last cisplatin treatment. Staining appears similar in intensity and distribution to that of normal tissues. ∗, basal lamina. Bar = 20 μm.
Northern Blot Analysis
RNA was extracted from the stomach tissues by a guanidinium isothiocyanate solution and purified by the CsCl cushion method (Chirgwin et al. 1979). Final RNA recovered from each sample was quantified by its UV absorbance at 260 nm. Intensity of ribosomal RNA bands was studied after ethidium bromide staining.
The RNA (20 μg) was size-separated by electrophoresis in a 1.3% agarose gel containing formaldehyde and electroblotted onto nylon membrane (GeneScreen; New England Nuclear, Boston, MA). The blots were prehybridized with the hybridization buffer without probes for 3 hr at 42C. Blots were incubated with fresh hybridization buffer in the presence of labeled probes (40 ng/ml) at 42C for 2 days. The hybridization buffer for the gastrin mRNA contained 50% formamide, 10% dextran sulfate, 50 mM Tris (pH 6.8), 3 × SSC, 100 μg/ml sonicated salmon sperm DNA, and 5 × Denhardt's solution. The buffer for the somatostatin was the same as that for gastrin except without dextran sulfate. The blots were washed with two changes of 2 × SSC/0.1% SDS at 42C for 30 min each and 0.1 × SSC/0.1% SDS at 42C for 15 min. The probes were detected by the same kit for in situ hybridization and following the same procedure.
Dot-blot Analysis
Stomach tissues were homogenized (Sambrook et al. 1989) and the protein was quantified by its UV absorbance at 280 nm (Deutscher 1990). Total denatured protein sample (100 μg) was dot-blotted on the nitrocellulose membrane (ImmunoSELECT; GIBCO BRL) by Hybri-slot Manifold (Bethesda Research Laboratories; Bethesda, MD). The blot was incubated in the primary antibody to gastrin at 4C overnight and was detected by the same method used in the immunocy-tochemical study described above. Liver tissue served as negative control.
Results
Immunocytochemistry and In Situ Hybridization
Using a gastrin-specific primary antibody, immunocy-tochemical study showed that gastrin-positive cells were mostly localized in the basal portion of the pylorus of rat stomach (Figure 1A). Control sections stained only with Vector Red substrate showed no significant endogenous AP activity. The liver tissue was also negative. At 1 and 6 days after the last cisplatin treatment, no gastrin-positive cells were observed in the gastric mucosa (Figure 1B). However, 10 days after cisplatin treatment, gastrin-positive cells were evident (Figure 1C). The intensity and distribution of staining were back to normal on Day 15 after the last cisplatin treatment.
Gastrin-specific primary antibody crossreacts with cholecystokinin (CCK) octopeptide, so ISH and Northern blot tests were applied to the adjacent tissues for immunocytochemical and dot-blot studies to confirm the results from these studies. The intensity and distribution of gastrin mRNA-positive cells were similar to those observed after immunocytochemical study, mostly in the basal portion of the pylorus of the rat stomach, with intense positive cytoplasmic staining (Figure 1D). The number of positively stained cells was less than that after the immunocytochemical study. This is probably because the CCK-positive cells did not stain after ISH. No gastrin mRNA-positive cells were noted 1 and 6 days after cisplatin treatment in the gastric mucosa (Figure 1E). However, 15 days after cisplatin treatment, the intensity and distribution of gastrin-positive cells were similar to those of the normal tissues (Figure 1F). The numbers of gastrin and gastrin mRNA-positive cells in each group were significantly different (Figure 2), p <0.01.
Northern Blot and Dot-blot Analysis
Total RNA from rat stomach tissues was isolated and size-separated by electrophoresis on formaldehyde-1.3% agarose gels. The RNA blot was hybridized with DIG-labeled oligonucleotide probes for either gastrin mRNA or somatostatin mRNA. At the corresponding RNA molecular marker level, gastrin mRNA was detected (Figure 3). At 1 and 6 days after cisplatin treatment the gastrin mRNA bands were almost negative. However, 10 days after the last cisplatin treatment the gastrin mRNA levels became significant, reaching normal levels 15 days after treatment. Somatostatin (Figure 4) mRNA followed similar patterns as described for gastrin mRNA.
Numbers of gastrin-positive cells after immunocy-tochemical study (broken line) vs gastrin mRNA positive cells after ISH (solid line). Gastrin and its mRNA drop to negligible levels 1 day after the last cisplatin injection. These levels remain depressed for about 6 days and then gradually rise to normal after 15 days (p < 0.01).
For dot-blot studies, 100-μg protein samples were dot-blotted onto the nitrocellulose membrane. The membrane was incubated with rabbit anti-gastrin antibody and detected by the ABC method (Figure 5). At 1 and 6 days after cisplatin treatment, gastrin levels were undetectable, as in case of gastrin mRNA levels described above. However, 10–15 days after cisplatin treatment the gastrin levels showed a gradual increase. The rats in the same groups demonstrated a similar change in Northern and dot-blot tests.
Northern blot analysis of gastrin mRNA. Total rat stomach RNA (20 μg) from different groups was fractionated by electrophoresis on a 1.3% agarose gel, electroblotted onto a nylon membrane, and hybridized with rat gastrin probe, which was detected by a nonradioactive method. Gastrin mRNA levels were hardly detectable at Day 1 and Day 6 after cisplatin treatment. However, the levels gradually returned to normal 10–15 days after the last cisplatin treatment. 28S ribosomal RNA served as control for any variations in sample size. C, control; 1, 6, 10, and 15 represent days after the last treatment.
Northern blot analysis of somatostatin mRNA. Total rat stomach RNA (20 μg) from different groups was fractionated by electrophoresis on a 1.3% agarose gel, electroblotted onto a nylon membrane, and hybridized with somatostatin probe, which was detected by a nonradioactive method. The somatostatin mRNA level dramatically decreased 1 and 6 days after cisplatin treatment, and gradually returned to normal 10–15 days after cisplatin treatment. 28S ribosomal RNA served as control for any variations in sample size. C, control; 1, 6, 10, and 15 represent days after the last treatment.
Dot-blot analysis of gastrin after cisplatin treatment. One hundred μg protein sample was dot-blotted onto the nitrocellulose membrane and incubated with rabbit anti-gastrin antibody (1:1000). The antibody was detected by the ABC method. Gastrin was negative 1 and 6 days after the last cisplatin treatment. Note a gradual increase of gastrin 10 days after the cisplatin treatment. C, control; 1, 6, 10, and 15 represent days after the last treatment.
Discussion
Gastrin regulates the secretion of gastric acid through its action on its receptors on parietal cells and ECL cells. ECL cells produce histamine, which is a potent stimulant of gastric acid production. The half-life for rat gastrin in circulation is around 10 min, so the circulating gastrin level is basically maintained by the continuous production and secretion of gastrin from G-cells (Walsh 1994). Gastrin mRNA transcription and gastrin production were inhibited after the first day of cisplatin treatment, and the acid content of the stomach builds up soon after Day 2 (Aggarwal et al. 1994). Gastrin release in rats is strongly dependent on paracrine regulation by somatostatin-secreting cells (D-cells) in the stomach (Walsh 1994). Northern blot studies show that somatostatin mRNA was hardly detectable at 1 and 6 days after cisplatin treatment, but gradually recovered 10 days after the last treatment. This change in the somatostatin mRNA level is parallel to that of gastrin mRNA. Therefore, gastrin inhibition after cisplatin treatment is independent of somatostatin. Otherwise, we would have observed an increase in somatostatin mRNA.
The cellular toxicity of cisplatin is mainly caused by its ability to bind covalently to DNA to form intra-strand and/or interstrand crosslinks, which in turn prevents DNA replication and transcription (Andrews and Howell 1990). Suppressed DNA replication and transcription of DNA after cisplatin treatment might represent a nonspecific inhibition of gastrin and somatostatin mRNA production.
Ulceration of the stomach is probably caused by an imbalance between gastric protection against injury and the erosive acid/peptic factors that exist in the normal stomach (Mertz and Walsh 1991). Recently, both exogenous and endogenous gastrin have been shown to exert a gastroprotective action (Konturek et al. 1995; Stroff et al. 1994, 1995). Gastrin plays an important physiological role in the maintenance of gastric mucosal integrity. Mediated by the nitric oxide pathway, gastrin also maintains the gastric mucosal blood flow, which plays a crucial role in gastroprotection (Leung et al. 1985). It appears that after cisplatin treatment there is a significant decrease in gastrin.
Normal stomach motility involves contraction of the stomach smooth muscle and relaxation of the pyloric sphincter, both of which are controlled by the release of acetylcholine from the nerve terminals. Acetylcholine release is dependent on calcium-calmodulin. Under the low chloride ion concentrations inside the cell, cisplatin hydrolyzes into monoaqua (monovalent) and diaqua (divalent) species. The diaquatic form interferes with binding of calmodulin to calcium, which is further affected by hypocalcemia induced by cisplatin (Blachley and Hill 1981). Lack of the calcium-calmodulin complex inhibits acetylcholine release from the synaptic vesicles which, in turn, causes bloating of the stomach and increased production of gastric acid. It has been demonstrated that stomach bloating and a significant increase in gastric acid in rats can be prevented by administration of calcium (Aggarwal and Fadool 1993; Aggarwal et al. 1994; Jarve and Aggarwal 1997), whereas gastrin and somatostatin have no direct involvement in the gastric acid increase responsible for ulceration. However, it is probable that gastroprotective effects are compromised in the absence of gastrin.
Footnotes
Acknowledgments
Our sincere thanks to Drs Neal Band and Will Kopachik for their kind guidance during the conduct of our experiments. Cisplatin was a gift from Andrulis Pharmaceutical Corporation, Beltsville, Maryland.