Ghrelin prevents development of diabetes at adult age in streptozotocin-treated newborn rats
- First Online:
- 524 Downloads
Ghrelin, a stomach-derived hormone, functions in multiple biological processes, including glucose metabolism and cellular differentiation and proliferation. In this study, we examined whether early treatment with ghrelin can regenerate beta cells of the pancreas in an animal model of diabetes mellitus, the n0-STZ model, in which neonatal rats are injected with streptozotocin (STZ) at birth.
Following administration of ghrelin to n0-STZ rats from postnatal days 2 to 8, we examined beta cell mass, mRNA expression levels of insulin and of pancreatic and duodenal homeobox 1 (Pdx1) gene, and pancreatic morphology on days 21 and 70. In addition, we investigated the effects of ghrelin on beta cell replication.
By day 21, ghrelin treatment increased pancreatic expression of insulin and Pdx1 mRNA in n0-STZ rats. The number of replicating cells was also significantly increased in the ghrelin-treated n0-STZ model. At day 70, n0-STZ rats exhibited hyperglycaemia, despite slight increases in plasma insulin levels. Ghrelin treatment resulted in the improvement of plasma glucose levels, which were associated with normal plasma insulin levels. Pancreatic insulin mRNA and protein levels were significantly increased in ghrelin-treated n0-STZ model animals.
These findings suggest that ghrelin promotes regeneration of beta cells in STZ-treated newborn rats. Thus, early administration of ghrelin may help prevent the development of diabetes in disease-prone subjects after beta cell destruction.
KeywordsDiabetes mellitus Ghrelin Insulin Pancreatic beta cell Pdx1 Streptozotocin
fasting blood glucose
growth hormone secretagogue receptor
pancreatic and duodenal homeobox 1
Ghrelin was originally isolated from rat stomach as a growth-hormone-releasing acylated peptide . This peptide hormone stimulates food intake and growth hormone secretion [2, 3], but also plays a regulatory role in glucose metabolism and insulin secretion. Ghrelin may have direct biological action within the pancreas, potentially functioning during pancreatic development. Although these results remain controversial, in vivo peripheral administration of ghrelin modulated plasma insulin and glucose levels [4, 5, 6, 7]. The growth hormone secretagogue receptor (GHS-R), for which ghrelin can serve as a ligand, is expressed in the pancreas [8, 9]. Ghrelin gene expression is markedly higher in the fetal pancreas than in the stomach , suggesting ghrelin might be involved in the development, proliferation or differentiation of the pancreas. Ghrelin has been reported to possess mitogenic and anti-apoptotic actions in adipocyte cells; treatment of pre-adipocytes with ghrelin induced cellular proliferation and differentiation into mature adipocytes [11, 12].
Diabetes mellitus is a common and serious disease. Of the abundant experimental trials designed to explore novel therapies for the disease, one attempted to increase insulin secretion from the pancreas by promoting the proliferation or regeneration of beta cells . In this study, we examined the effects of ghrelin on the development of hyperglycaemia and the regeneration of beta cells in streptozotocin (STZ)-treated newborn rats, which show rapid remission but gradual development of diabetes in adult animals. This model has been traditionally used to study pancreatic regeneration [14, 15, 16]. We investigated the changes in the islets of Langerhans morphology and pancreatic insulin mRNA and protein levels following ghrelin administration in this animal model. Further, we studied the resulting alterations in the expression levels of pancreatic and duodenal homeobox 1 (Pdx1), one of the major pancreatic transcriptional factors thought to play a key role in the development and regeneration of the pancreas [17, 18, 19].
Materials and methods
All animal experimental procedures were approved by the Kyoto University Graduate School of Medicine committee on animal research.
Female and male Sprague–Dawley rats, purchased from Charles River (Yokohama, Japan), were allowed free access to tap water and a standard pellet rat diet (352 kcal/100 g, CE-2; Clea, Tokyo, Japan). Females were caged with a male for one night; pregnancy was detected by abdominal palpation 14 days later. Natural birth occurred 22 days after mating.
We examined four experimental groups: (1) a Control group, in which newborn rats received a single i.p. injection of citrate buffer; (2) a Ghrelin group, which received twice daily s.c. injections of ghrelin (100 μg/kg body weight) for 7 days (from day 2 to 8) after birth; (3) an n0-STZ group, which received a single i.p. injection of streptozotocin; and (4) an n0-STZ/Ghrelin group, which received a single i.p. injection of streptozotocin followed by injections of ghrelin as in the Ghrelin group. Dams were randomly assigned to four groups and pups from the same litter were assigned to the same group. The numbers of dams in each of the four groups were five (Control), five (Ghrelin), seven (n0-STZ) and seven (n0-STZ/Ghrelin).
At birth, newborns received a single i.p. injection of 100 mg/kg body weight of STZ (Sigma, Steinheim, Germany), freshly dissolved in citrate buffer (0.05 mmol/l,pH 4.5). Pups were left with their mothers. All neonates were tested on day 2 for glycosuria with Multistix SG (Bayer Medical, Leverkusen, Germany). Only those animals that were glycosuric (3+ value with the Multistix SG test) at day 2 after birth were included in the n0-STZ model group. Ghrelin treatment was started after glycosuria was confirmed.
Animals were killed at 21 or 70 days after birth by bleeding following anaesthesia with an i.p. injection of sodium pentobarbital (50 mg/kg). Blood samples were collected from the inferior vena cava, centrifuged immediately at 20,000 ×g for 2 min at 4°C, and stored at −80°C until assayed.
Pancreas removal and treatment
After excision, pancreases were removed and weighed. To measure insulin content, pancreases (35–50 mg) were homogenised and centrifuged in 5 ml acid-ethanol (0.15 mol/l HCl in 75% [vol/vol] ethanol) at 1,000 g for 20 min; the supernatants were stored at −80°C. For immunohistochemistry, additional pancreases were fixed in 4% paraformaldehyde fixative for 24 h and embedded in paraffin.
Insulin and Pdx1 were detected immunohistochemically on 3-μm-thick tissue sections using an indirect peroxidase-labelling technique. Sections were incubated for 1 h at room temperature or overnight at 4°C with primary antibodies (guinea-pig anti-porcine insulin [Dako, Glostrup, Denmark], or rabbit anti-mouse/rat IDX-1 [Chemicon, Temecula, USA]). Insulin- and Pdx1-positive staining was visualised using a 3,3′-diaminobenzidine tetrahydrochloride kit (DakoCytomation) and fluorescence, respectively. Quantitative evaluations of insulin- and Pdx1-positive areas were performed using a computer-assisted image analysis procedure using an Olympus BX 51 microscope connected via a DP 12 digital camera and Mac SCOPE Ver 2.6 software (Mitani, Fukui, Japan). The insulin- and Pdx1-positive area, as well as the area of the total pancreatic section, was evaluated on ten non-consecutive sections of the pancreas of five animals per group at each point. For semi-quantitative analysis, more than six fields were observed per section. The relative volume of insulin- and Pdx1-positive cells was determined by a stereological morphometric method, calculating the ratio between the areas occupied by immunoreactive cells and that occupied by total pancreatic cells.
Beta cell replication
Pancreatic sections that had not already been used for morphometric studies were used to examine beta cell replication. Sections were double-stained for phospho-histone H3 (Ser10) and insulin. Sections were incubated with a 1/50 dilution of anti-phospho-histone H3 (Ser10) antibody (Cell Signaling Technology, Beverly, MA, USA) overnight at 4°C. Sections were then incubated with guinea-pig anti-insulin antibody for 1 h at room temperature. After six washes with PBS, the sections were incubated with the secondary antibody conjugated with fluorescence (Alexa fluor 488 and 546; Molecular Probes, Eugene, OR, USA) in blocking solution for 30 min at room temperature. After rinsing with PBS, sections were mounted with Mounting Medium with DAPI (Vector Laboratories, Burlingame, CA, USA) and examined with confocal laser scanning microscopy (Leica Microsystems, Mannheim, Germany). A minimum of 1,000 beta cell nuclei were counted per section.
RNA extraction and RT-PCR methods
Total RNA was extracted from rat pancreases as previously described . To synthesise the first-strand cDNA, the extracted products were used as templates in reactions containing RT (Invitrogen, Carlsbad, CA, USA). RT-PCR analysis was performed as previously described . The first-strand cDNA was used in subsequent PCR analyses, in combination with the following oligonucleotide primers: insulin-1: 5′ TAGACCATCAGCAAGCAGGTC, 3′ CACACCAGGTACAGAGCCT; Pdx1: 5′ AGGAGGTGCATACGCAGCAG, 3′ GAGGCCGGGAGATGTATTTGTT.
The identity of the PCR products was confirmed by agarose gel electrophoresis.
We mixed 1 μl of the amplified cDNA with 25 μl 2×PCR Master Mix (Applied Biosystems, Foster City, CA, USA), 23 μl sterile distilled water and 0.5 μl 10 pmol/μl sense and antisense primers specific for either insulin or Pdx1. Forty cycles of PCR amplification were carried out using the Thermal Cycler System (ABI PRISM 7700; Applied Biosystems) for 15 s at 95°C followed by 60 s at 60°C. The concentration of each mRNA product was quantified using calibration curves expressing the fluorescence intensity against the amount of standardised PCR product. All expression data were normalised to the amount of 18S ribosomal RNA (Applied Biosystems) amplified from the same individual sample.
Plasma and blood glucose levels were determined using a glucose analyser (Antisense 2; Sankyo, Tokyo, Japan). Insulin was extracted from pancreases as described . Insulin concentrations were measured using a Lebis insulin ELISA kit (Shibayagi, Gunma, Japan).
Values are expressed as means±SEM. Differences between each group of rats were evaluated using a Mann–Whitney test or a Bonferroni-corrected ANOVA as appropriate. A value of p<0.05 was considered statistically significant.
Effects of ghrelin treatment from day 2 to day 8 after birth: studies on day 21
Characteristics (means±SEM) of fasted 21-day-old rats
Body weight (g)
Plasma insulin (pmol/l)
Thus, the n0-STZ group exhibited diminished insulin production at 21 days after birth, although FBG levels achieved a spontaneous remission. The n0-STZ/Ghrelin group showed increased insulin expression in both mRNA and protein levels, Therefore, neonatal ghrelin treatment in this model may improve the reduction of insulin production at 21 days and prevent development of hyperglycaemia at 70 days.
Long-lasting effects of early treatment with ghrelin: studies at 10 weeks
Characteristics (means±SEM) of fasted rats at 70 days
Body weight (g)
Plasma insulin (pmol/l)
Pancreas weight (mg)
Pancreatic insulin (μg/pancreas)
The mRNA expression of insulin at the adult stage in the Ghrelin group was lower than that in the Control group (Fig. 4b). In the n0-STZ group, the mRNA expression of insulin at day 70 remained reduced. The n0-STZ/Ghrelin group demonstrated a similar insulin expression level to that of the Control group, as was also observed 21 days after birth (Fig. 1a and Fig. 4b).
We performed the experiments of immunostaining for phospho-histone H3 and insulin also in the adult stage. There were no significant differences between the four groups. But in the model of STZ, the ghrelin treatment tended to increase the double-positive cells for phospho-histone H3 and insulin, while in the Control group, the ghrelin treatment tended to decrease those cells (Fig. 3c).
Thus, in this n0-STZ model of diabetes, animals developed reduced insulin production in the pancreas and hyperglycaemia at 10 weeks after birth. Ghrelin treatment may prevent this deterioration through maintenance of insulin-positive cells and insulin production, which mechanistically may involve the increased expression of Pdx1.
Neonatal rats treated with STZ exhibit normal glycaemia at 3 weeks after birth by rapid remission, but then deteriorate gradually after 8 weeks [14, 15, 16]. This delayed deterioration probably follows from the inability of the insulin production of the pancreas to recover to the levels of control rats [14, 15]. In this study, we observed a similar clinical course, characterised by the reduction of insulin production and content in the pancreases of n0-STZ rats. This neonatal model showed more reduction of pancreatic beta cells compared with other reports, in which the reduction was approximately 50% that of controls [15, 16, 24]. This may be due to the different timing of STZ injection, which is known to result in type 2 diabetes models with different severity . In fact, we injected STZ at 24 h after birth, while others injected within 12 h of birth [15, 16, 24].
Moreover, we first demonstrated that ghrelin treatment prevented the development of hyperglycaemia in n0-STZ rats at adult ages. In n0-STZ/Ghrelin rats, both the transcription and translation of insulin in the pancreas were significantly increased at 21 and at 70 days. In addition, Pdx1 gene and protein expression levels at 21 days were restored to levels similar to those of the Control group. These findings strongly suggest that beta cells damaged by STZ administration during the neonatal stage can be regenerated or replicated following ghrelin treatment. Indeed, the immunohistochemical analysis for phospho-histone H3, which is a proliferation marker, demonstrated that ghrelin treatment significantly increased the staining in both Control and STZ-treated rats at the age of 21 days.
Similar findings have been reported for glucagon-like peptide-1 and exendin-4 by Tourrel et al. . They demonstrated that glucagon-like peptide-1 and exendin-4 stimulated beta cell neogenesis or regeneration in STZ-treated newborn rats. While these n0-STZ rats ameliorated plasma glucose levels at adult stages, such treatment could not improve insulin secretion during a glucose-tolerance test. In future, it will be necessary to examine the effects of ghrelin treatment on insulin secretion in this n0-STZ/ghrelin model in detail, including an examination of glucose tolerance. In addition, a possibility that ghrelin treatment might modulate STZ toxic effects should be considered, because ghrelin administration was performed shortly after STZ treatment. Finally, another possibility that STZ may have affected hepatic metabolism and/or insulin sensitivity in peripheral tissues should be taken in account. These effects could explain some of the discrepancies between pancreas and plasma insulin levels.
Several reports have shown that both ghrelin and GHS-R are expressed in the pancreas [8, 9, 25, 26, 27]. Ghrelin is expressed in both the alpha  and beta [9, 29] cells of the pancreas. Recently, ghrelin was also identified in epsilon cells . The pancreas is one of the organs in which GHS-R expression has been detected by RNA protection assay . In addition, ghrelin gene expression is higher in the fetal pancreas than in fetal stomach , while expression is highest in the stomach in adults. Thus, ghrelin may act on beta cells in a paracrine or autocrine manner, potentially functioning in the differentiation, proliferation or development of beta cells.
Ghrelin has been shown to induce the proliferation of neoplastic cells in certain systems , while other studies have demonstrated  anti-proliferative or anti-apoptotic effects against neoplastic cells . Ghrelin treatment of pre-adipocytes or motor neurons induced cellular proliferation and differentiation into mature adipocytes or neurons [11, 12, 33]. While these findings suggest ghrelin may influence cell differentiation, proliferation, development or apoptosis, the mechanism by which ghrelin treatment exerts these effects is not clear. In the present study, expression of Pdx1, insulin and phospho-histone H3 in n0-STZ/Ghrelin rats was markedly increased in comparison with the level observed in 21-day-old n0-STZ rats. These findings strongly favour the conclusion that ghrelin is stimulating beta cell regeneration or replication in STZ-treated newborn rats, although further studies are necessary to clarify the mechanism including the contribution of apoptosis. The values obtained with phospho-histone H3 were higher than those obtained with 5-bromodeoxyuridine (BrdU) staining of beta cells at 21 days in Wistar rats (approximately 0.5–1.5%) [16, 24]. At this moment, we do not clearly know why we obtained higher phospho-histone H3 values in Sprague–Dawley rats of the same age and at 70 days, compared with the BrdU values. Although the correlation between anti-BrdU and -histone labelling indices was statistically significant, variation was noted in the percentage of the BrdU-positive cells double-labelled with phospho-histone H3, depending on cell type and length of BrdU incubation [34, 35]. Because, as far as we know, no phospho-histone H3 staining of pancreatic beta cells has been performed, it is difficult to directly compare our results with others. Nonetheless, ghrelin treatment increased the expression level of phospho-histone H3 at the age of 21 days but not 70 days. Finally, neonatal STZ treatment in earlier studies resulted in a slight increase of beta cell proliferation measured by the BrdU labelling index, in contrast to marked decrease in the present study. This discrepancy will be due to the fact that this neonatal model showed more reduction of pancreatic beta cells compared with other reports, as already discussed.
In addition, it was observed that ghrelin treatment in control rats appeared to result in different alterations from that in STZ-treated rats, for example, regarding body weight and FBG at day 21 and insulin gene expression level at day 70. This is probably because physiological and pathological responses to ghrelin treatment are different between healthy control and diabetic STZ-treated rats. Although it is hard to explain these apparent discrepancies at this time, it is necessary to study ghrelin action in healthy controls separately from morbid subjects. Although ghrelin treatment increased body weight in the Ghrelin group compared with the Control group in this study, such a significant difference was not observed in a previous report by Hayashida et al. . However, they also observed that body weight tended to increase after ghrelin treatment, albeit not statistically significantly (personal communication with N. Murakami, Miyazaki University, Miyazaki, Japan). They speculate that the increase in body weight might have reached statistically significance if the number of rats was increased.
From a clinical standpoint, results of the present study suggest that the n0-STZ/Ghrelin rat may be a valuable model to investigate novel treatment strategies for diabetes mellitus. In addition, ghrelin may be a useful treatment for diabetes. Recently, circulating ghrelin levels were discovered to be decreased in patients with type 2 diabetes . Ghrelin activity is involved in insulin resistance as well as the maintenance of energy balance . As multiple studies have demonstrated the importance of ghrelin in glucose and insulin metabolism [2, 3], ghrelin will be an important molecule to investigate the pathophysiology of diabetes.
In conclusion, we have provided evidence that ghrelin may assist the regeneration of beta cells, using STZ-treated newborn rats as a model of diabetes mellitus. In these rats, ghrelin treatment increased pancreatic insulin expression and resulted in the improvement of plasma glucose levels with normal plasma insulin levels. These findings suggest that early administration of ghrelin may prevent the development of diabetes in disease-prone subjects after beta cell destruction.
We would like to thank H. Hiratani, C. Ishimoto, N. Takehisa and K. Imai for their excellent technical assistance; M. Kouchi and K. Amemiya for their excellent secretarial assistance; and S. Teramukai (Kyoto University, Kyoto, Japan) for his statistical advice. This study was supported by funds from the Ministry of Education, Science, Culture, Sports and Technology of Japan, and the Ministry of Health, Labor, and Welfare of Japan.