This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Fernandes, A. Articles by Teitelman, G. Search for Related Content PubMed PubMed Citation Articles by Fernandes, A. Articles by Teitelman, G.

Differentiation of New Insulin-Producing Cells Is Induced by Injury in Adult Pancreatic Islets¹

Department of Anatomy and Cell Biology (A.F., L.C.K., Y.G., G.T.), SUNY Health Science Center at Brooklyn, Brooklyn, New York 11203; Department of Cell Biology (C.V.E.W.), Vanderbilt University Medical Center, Nashville, Tennessee 37232; and the Department of Molecular Physiology and Biophysics (R.S.), Vanderbilt University Medical Center, Nashville, Tennessee 37232

Address all correspondence and requests for reprints to: Dr. Gladys Teitelman, Department of Anatomy and Cell Biology, SUNY Health Science Center at Brooklyn, Brooklyn, New York 11203.

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
The ability of the adult pancreas to generate new insulin (ß) cells has been controversial because of difficulties in unequivocally identifying the precursor population. We recently determined that ß cells were generated during development from precursors that expressed the homeodomain-containing transcription factor pancreas duodenum homeobox gene-1 (PDX-1). To investigate whether PDX-1⁺ stem cells are present in adult pancreas, we examined two animal models of diabetes. One model was produced by injecting adult mice with streptozotocin (SZ), a toxin that produces hyperglycemia due to rapid and massive ß cell death. After SZ-mediated elimination of existing IN⁺/PDX-1⁺ cells, a population of somatostatin (SOM)⁺/PDX-1⁺ cells, a cell type thought to represent an embryonic islet precursor cell, appeared in islets. The appearance of SOM⁺/PDX-1⁺ cells was followed in time by the differentiation to SOM⁺/IN⁺/PDX-1⁺cells. SOM⁺/PDX-1⁺ cells also appeared in islets of nonobese diabetic mice, a strain of mice in which ß cell destruction is immune-mediated. Our findings establish the existence of PDX-1⁺ ß cell precursors in the adult pancreas and indicate that their differentiation is induced by islet injury .

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
IT IS GENERALLY believed that pancreatic exocrine and endocrine cells develop from precursor cells present in the pancreatic duct (1). Embryonic endocrine cells aggregate and form the islets of Langerhans, which, in mice, achieve a typical adult configuration after birth. Insulin (IN)-containing ß cells form the core of the mature islets, whereas the periphery contains lower numbers of the other endocrine cell types: the , , and PP cells, which synthesize glucagon (GLU), somatostatin (SOM) and pancreatic polypeptide (PP) respectively.

To elucidate the cell lineage relationships of pancreatic islet cells, the biochemical properties of the progenitor cells have been examined by us (2, 3) and others (4, 5, 6). The first molecular marker expressed by islet stem cells during development is a homeodomain protein, pancreas duodenum homeobox gene-1 (PDX-1). PDX-1 plays a key transcriptional role during the development of both the endocrine and exocrine pancreas as demonstrated by targeted mutagenesis experiments in mice (7, 8). It has been found that PDX-1, previously termed STF-1 (9), IDX-1(10), or IPF-1(11), is transiently expressed by precursor cells present in the pancreatic duct and by each of the four endocrine cell types when they first differentiate (3). Pancreatic PDX-1 expression becomes progressively restricted during development such that it is almost exclusively localized to ß cells in adults (12, 13), where it is believed to play a role in the regulation of IN gene transcription (11, 12, 13, 14). In addition to ß cells, a small subset of the , , and PP cells express PDX-1 (12, 13)

There is considerable evidence suggesting that the differentiation of PDX-1⁺ progenitor cells into pancreatic islet cells occurs by a multistep process, involving successive changes in the antigenic profile of the stem cells. PDX-1 cells coexpressing IN and GLU appear at e-9.5 in the pancreatic buds of mice before full morphogenesis of the pancreas (3). Coexpression of IN, GLU, and PDX-1 by islet cells decreases during gestation, such that a significant number of cells either were GLU⁺ or IN ⁺ PDX-1⁺ by e-14.5 (3). SOM⁺PDX-1⁺ and PP⁺PDX-1⁺ cells are first seen at e-14.5 and postnatal day 1 (P-1), respectively (3). At these stages a subset of SOM ⁺ and PP⁺ cells coexpress IN (2). These observations, together with similar findings in other systems (4, 15, 16), suggest that the four principal islet cell types arise from common multipotential precursors that coexpress PDX-1, GLU, and IN (3).

The increase in ß cell number during embryonic development and early postnatal life is presumed to occur by continued differentiation of new islet precursors cells from the pancreatic duct and by proliferation of differentiated IN⁺ cells (1, 17, 18, 19, 20, 21). In adults, new IN⁺ cells arise mostly from division of existing ß cells (reviewed in Ref.19).

Neogenesis, the differentiation of IN cells from stem-like cells present in the pancreas, has been reported to occur in several animal models, including after ß cell destruction with toxins in neonatal rats (22), after cellophane wrapping of the pancreas in adults (23), after partial pancreatectomy, (24), and in some transgenic mouse models of autoimmune diabetes (25, 26). However, in all these studies, attempts to demonstrate how the new ß cells were generated were hampered by the inability to unambiguously identify the precursor cells. The recent identification of GLU⁺/IN⁺/PDX-1⁺ cells as a likely embryonic islet stem cells (3) prompted us to ascertain whether similar cells were found in adult pancreas.

High circulating blood sugar has been postulated to induce the differentiation of ß stem cells into IN⁺ cells (19). As a consequence, we examined pancreatic tissues of diabetic mice to determine whether islets contained newly differentiated ß cells. We reasoned that PDX-1⁺ ß precursor cells in adults might be located in the pancreatic duct, as in embryos. Alternatively, these ß precursor cells may be scattered in the exocrine tissue or clustered in islets. Furthermore, we hypothesized that these cells could be induced to differentiate by mechanisms unleashed by islet injury. To address this hypothesis, we analyzed ß cell neogenesis in pancreas in two different animal model systems. The first consisted of mice injected with SZ, a specific ß cell toxin (17, 27; reviewed in Ref. 28). SZ consists of 1-methyl-1 nitrosourea attached to the C-2 of D-glucose. It is believed that the basis of the selective effect of SZ on ß cells is its glucose moiety, which reacts with specific glucose sensing mechanisms (28, 29). A widely accepted view is that it acts within ß cells to deplete nicotinamide adenine dinucleotide (NAD), producing long-lasting impairment and death of ß cells (28, 30). To determine whether the appearance of new ß cells was correlated with the degree of islet injury, pancreatic tissues of mice injected either with 200 mg/kg SZ (200 SZ mice) or 100 mg/kg SZ (100 SZ mice) were compared. While the higher dose of SZ eliminated almost all ß cells within a few hours after exposure of the drug, the lower dose caused a partial reduction of the ß cell mass (28, 31). The second model examined was the NOD mouse strain, in which mature ß cells are lost through an autoimmune destruction (reviewed in Ref.28).

We found few PDX-1⁺ ß cells in the pancreatic duct in all of these animal models of diabetes. Importantly, injection of 200 mg/kg SZ with the resulting elimination of most of the existing ß cells induced the differentiation of significant numbers of SOM⁺/PDX-1⁺ cells, a transitional cell type in the endocrine cell lineage (3), in all pancreatic islets. The appearance of these transitional cells was followed by the differentiation of a large number of SOM⁺ cells also expressing IN. SOM⁺/PDX-1⁺ cells also differentiated in islets of 100 SZ and NOD mice. We conclude that mature islets contain a PDX-1⁺ ß stem cell population that can be induced to differentiate into IN⁺ cells following islet injury.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Animals and tissue processing
Six-week-old male CD-1 mice were purchased from Charles River (Wilmington, MA). Mice were injected ip with SZ (Upjohn Co., Kalamazoo, MI; USB/Amersham, Arlington Heights, IL) in 0.1 M citrate buffer, pH 4.5, after a 12-h overnight fast. The drug was administered immediately after preparation of the solution. Mice injected with SZ at 200 mg/kg (200 SZ mice) after an overnight fast became severely hyperglycemic (blood glucose levels over 400 mg/dl) 24 h later. Animals that received 100 mg/kg SZ (100SZ mice) also became hyperglycemic, but the increase in blood glucose levels occurred at a slower rate, reaching very high levels (over 400 mg/dl) 2–3 days after the treatment. Controls were injected with the equivalent volume of citrate buffer. Blood for glucose determination was collected by snipping the tail in the fed state. Blood glucose was measured with Tracer II Blood Glucose monitor (Boehringer Mannheim Corp., Indianapolis, IN). Animals with a plasma glucose of 400 mg/dl or higher were used.

One-month-old female NOD mutant and control (ICR) mice were purchased from Taconic Farms (Germantown, NY) and were kept for the required amount of time in the animal colony of SUNY.

Mice were perfused through the heart with Zamboni’s fixative (85 ml 2% paraformaldehyde in 0.1 M phosphate buffer, pH 7.4, plus 15 ml saturated picric acid) and postfixed for several hours in the same fixative. For semithin epon sections, tissues were stored overnight in 7% sucrose in phosphate buffer and dehydrated through a graded series of ethanols to propylene oxide. After dehydration, the tissues were transferred to a mixture of propylene oxide: epon (1:1) for 2 h, then placed in Epon 812-filled capsules at 37 C overnight, followed by a 24-h incubation at 45 C, and an overnight incubation at 60 C as previously described (32). For cryostat sections, the fixed tissues were infiltrated overnight in 30% sucrose, mounted in embedding matrix (Lipshaw Co., Pittsburgh, PA), and 15–20 µm cryostat sections were collected onto gelatin-coated slides.

Source of antibodies and purified peptides
Primary antiserum. Guinea pig antibodies to bovine IN were purchased from Linco Research Inc. (Eureka, MO). Rabbit antiserum to human glucagon was purchased from Calbiochem (San Diego, CA). Rabbit antisera to human PP and to somatostatin were supplied by Peninsula Labs (Belmont, CA). Affinity purified antiserum to the N-terminal domain of PDX-1 was raised against the first 75 amino acids of PDX-1 as a GST/XIH box 8 fusion protein (13, 33).

Secondary antibodies. Biotinylated goat antirabbit IgG and avidin-labeled peroxidase were purchased from Vector Laboratories (Burlingame, CA). All antisera was tested by immunoadsorption and the dot blot technique according to criteria described elsewhere (32, 34).

Immunolabeling of cryostat sections using peroxidase techniques
These techniques have been described in detail elsewhere (1, 2, 32). In brief, the sections were incubated sequentially in an empirically derived optimal dilution of control serum or primary antibody raised in species "X" containing 1% goat serum in Tris-saline solution (0.9% NaCl in 0.1 M Tris, pH 7.4) for 18 h; a 1:50 dilution of anti-(species x) biotinylated IgG solution in 1% goat serum in TS for 30 min; and a 1:100 dilution of peroxidase-avidin complex for 30 min (avidin-biotin complex: ABC technique). Following these incubations, the bound peroxidase was visualized by reaction for 6 min in a solution containing 22 mg of 3,3'-diaminobenzidine (DAB) and 10 µl of 30% H₂O₂ in 100 ml of 0.1 M TS. After the DAB step, sections were dehydrated and mounted with Permount. Antibodies were used at the following dilutions: guinea pig antibovine IN-1:400; rabbit antihuman glucagon-1:12,000; rabbit antihuman somatostatin-1:8,000; rabbit antihuman pancreatic polypeptide-1:20,000; rabbit anti-PDX-1 at 1:500.

Double label immunohistochemistry using two peroxidase substrates
Sections were incubated first with antisera to PDX-1, and the bound antibody was visualized by DAB (brown precipitate), followed by incubation with antisera to a hormone, which was visualized with the blue reaction product of the Vector SG substrate (Vector Labs).

Analysis of double label slides
Slides processed for double immunohistochemical staining or for combined immunohistochemistry and autoradiography were examined with a Nikon (Garden City, NY) Microphot SA microscope equipped with Nomarski optics and using a 10x ocular and an oil immersion 100x objective. Depth of focus was calculated as previously described (3). Briefly, the depth of focus was determined according to manufacture specifications and Klein and Furtak (35) as follows:

D = n x l/2 x (N.A.)² + n/7 x N.A. x M

where n = refractive index on object side

l = wavelength of light

M = total magnification

D = 1.515 x 550 nm/2 x (1.25)² + 1.515/7 x 1.25 x 1250¹

= 0.35807

¹ Due to 1.25 differential interference contrast (DIC) magnification factor in microscope.

With a depth of focus of approximately 0.4 µm, only objects located at ±0.4 µm are within the same plane of focus. Therefore, structures from different cells can be easily distinguished because they are at different focal planes and cannot be focused simultaneously for photography. The small depth of focus of the objective used allowed us to clearly distinguish the presence (or absence) of staining in the nucleus and cytoplasm of individual cells. Only cells that contained immunostained nucleus and cytoplasm in the same plane of focus were scored as double labeled.

Determination of relative islet volume and ß cell mass
The relative islet volume was determined in sections immunostained for IN by the point sampling method (36) using a 300 point ocular grid at a total magnification of 400x. The average number of ß cells per islet volume was calculated according to the formula: F = h/n, in which h is the number of hits over ß cells and n is the number of points scored over islets (37).

Semiquantitative assessment of insulinitis in islets of NOD mice
Pancreatic islets of NOD mice were scored as described in Ref. 38. Sections stained for IN were examined using Nomarski interference optics, which allows the visualization of stained and unstained cells At least three mice per stage and six or more islets per animal were scored for islet infiltration and alteration of islet morphology. Islets with normal morphology, and no visible sign of infiltration were scored as 0. The score was 1 when the islets were surrounded by lymphocytes with little or no perturbation of islet structure, 2 when islets contained considerable lymphocyte infiltration and had significant alteration in islet structure, and 3 when islets were completely engulfed by lymphocytes and had only small clusters of endocrine cells.

Determination of labeling index (LI) using ³H-thymidine
Mice received a single dose of the isotope (5 uCi/g, ip; NEN, specific activity approximately 80 uC/mmol). The animals were perfused with fixative solution 1 h or 24 h later, the pancreas was removed, and the sections processed for immunohistochemistry. After the DAB step, sections were dehydrated with ascending alcohols, cleared in xylene, rehydrated, air dried, and dipped in Ilford L-4 emulsion (Polysciences, Warrington, PA) at 50 C diluted 1:1 with distilled water. The autoradiographic preparations were air dried and exposed in light-proof boxes with dessicant for 4 C for 2–3 weeks. At the end of the exposure period, the slides were developed for 5 min in Kodak D-19 developer, rinsed in water, fixed with Ektaflo fixer for 10 min, washed in running water for 1 h, dehydrated, cleared, and mounted with a coverslip using Permount.

Analysis of light microscopic autoradiography
Cells that were at the S phase of the cell cycle incorporated the isotope into the nucleus. The range of ß particles from tritium does not exceed 2 µm (39), such that only ³H-thymidine containing nuclei located within this distance from the emulsion will be decorated with silver grains, allowing the examination of cells that were located immediately below the emulsion. Examination of immunostained slides processed for autoradiography was performed using the same optical system and magnification described in the preceding section.

Processing of semithin sections
Semithin sections (1-µm thick) were treated to remove the Epon (40) and were incubated at 4 C for 2–3 days with SOM, IN, or GLU antisera, respectively. Then, the sections were processed for visualization of the bound antibody by the ABC technique as described above. Antibodies were used at the following dilutions: guinea pig antibovine IN-1:50; rabbit antihuman glucagon-1:1000; rabbit antihuman somatostatin-1:1000. Using this technique, we visualized IN⁺ and SOM⁺ cells, but not PP⁺ cells or PDX-1⁺ cells because the procedure used to remove the epon abolished PP and PDX-1 immunoreactivity. In addition, we were unable to visualize IN in epon sections of NOD islets. Although this technique gave clear cellular localization of immunoreactive products, it precluded the survey of a large number of pancreatic islets. This was due to the small size of the islets present in SZ-treated mice, which prevented their visualization before embedding and sectioning.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
Reappearance of ß cells expressing PDX-1 and IN after SZ
One day after injection of SZ (1 day after SZ injection = 1 day post SZ), most islet ß cells were destroyed in mice injected with 200 mg/kg SZ (200 SZ mice) (Fig. 1 and Table 1). In addition to the ß cells, some GLU⁺ cells of 200 SZ mice were also necrotic (Fig. 1). In contrast, there was no evidence of necrosis of or PP cells. The core of 1 day post SZ islets contained IN⁺ granule-like substance that probably included IN secretory vesicles. It has been shown that the membranes of IN secretory granules are more resistant to degradation during necrosis than cell membranes (41). The number of IN⁺ cells in some, but not in all islets, increased with time (Figs. 1 and 2). Some ß cells stained darkly for IN, whereas others were degranulated and weakly reactive, a typical feature of cells that have responded to extremely high levels of glucose (27). At 5 days post SZ, a substantial number of ß cells had reappeared (Fig. 1 and Table 1). However, the average number of ß cells per islet volume was lower than in controls (Table 1). In addition, some islets were fibrotic, and the newly differentiated IN⁺ cells failed to fill the core of the islet. These studies strongly suggest that SZ eliminated almost all preexisting ß cells and that new IN⁺ cells gradually reappeared in the islets after injection of the drug.

View larger version (109K): [in this window] [in a new window]   Figure 1. SZ deletes most ß cells. Cryostat sections of pancreas immunostained with IN antisera. A, Control, Bar, 40 µm; B, 1 day post SZ. At this stage most ß cells have been eliminated. Bar, 40 µm. C, Staining of a cryostat section of a 1 day post SZ islet with GLU antisera illustrates the presence of necrotic cells (indicated with arrows). C, 5 days post SZ. Note the reappearance of cells containing IN in the SZ-treated islets. Bar, 40 µm.

View larger version (126K): [in this window] [in a new window]   Figure 2. ß cells expressing IN and PDX-1 reappear in SZ-treated islets. A, Localization of PDX-1 in cryostat sections of control islets. Most cells present in the core of the islet express the transcription factor. Bar = 45 µm. B, At 1 day post SZ, cells expressing PDX-1 persist in the periphery of SZ-treated islets. Note the absence of PDX-1+ cells in the core of the islet. Bar = 45 µm. C, Localization of PDX-1 and IN in cryostat sections of 1 day post SZ islet. Note that PDX-1+ cells in the periphery (brown label) lack IN staining and that the core of the islet do not contain IN+ cells. IN immunoreactive product (blue label) is dispersed throughout the islet. D, Photomicrograph illustrates a cryostat section of an islet of 1 day post SZ mice injected with 200 mg/kg. Section was sequentially incubated with antisera to PDX-1, which was visualized with DAB (brown reaction product in nuclei) and with a cocktail of antisera to SOM, GLU, and PP, which was visualized with a blue chromogen. Note the presence of cell debris, reacting with peroxidase, in the core of the islet (indicated with an asterisk). Cells present in the periphery of the islet, which were not affected by the toxin, contain immunostained cytoplasm and nuclei. Bar = 4 µm. E, Photomicrograph of an islet of a 5 days post SZ mice illustrates the presence of cells containing IN (bluelabel) and PDX-1 (brown reaction product). Note the location of IN+PDX-1+ (arrow) and of IN+PDX-1- cells (arrowhead). The location of IN+ cells in the periphery of the islet is a typical configuration of islets at this stage. Bar = 4 µm.

In 200 SZ mice, cells containing PDX-1 and IN reappeared within the islets by 5 days post SZ (Fig. 2E). In contrast to noninjected littermates, only a subset (44.5% ± 5.47, N = 20) of the IN⁺ cells expressed the transcription factor at detectable levels (Fig. 2E). A decrease in the percentage of ß cells expressing PDX-1 was also observed in 100 SZ mice at 5 days post SZ (data not shown). Similarly, almost 50% of IN⁺ cells present in severely affected islets lacked PDX-1 expression in NOD mice (data not shown). These observations indicated that the decrease in the percentage of ß cells expressing PDX-1 is not specific to SZ treatment. Because hyperglycemia dramatically decreases PDX-1 levels (42), we presume that we were unable to detect the homeoprotein in these IN⁺ cells. We conclude that an injection of 200 mg/kg SZ eliminated almost all IN⁺PDX-1⁺ cells present in islets and that new IN⁺ cells differentiated 5 days later.

Increase in the percentage of SOM⁺/PDX-1⁺ cells during neogenesis
The observation that IN⁺ cells reappear after exposure to SZ raised the possibility that these cells were generated through the transient appearance of a population of IN⁺/GLU⁺/PDX-1⁺ precursor cells, reminiscent of those seen in embryonic islet ontogeny. If this were the case, we expected to find an increase in the percentage of GLU⁺/PDX-1⁺ cells after SZ injection, suggesting that these stem cells were initiating IN synthesis. However, we found that the percentage of GLU⁺ cells that coexpressed PDX-1 (2%) did not change after exposure to SZ (Fig. 3). In contrast, the percentage of SOM⁺/PDX-1⁺ cells increased dramatically 1 day after SZ injection (Fig. 4) reaching, at 2 days post SZ, a level similar to that previously found in embryos (46.5% ± 2.9; Ref.3). The level of SOM⁺/PDX-1⁺ cells decreased to control values by 30 days post SZ (Fig. 4). The percentage of PP⁺/PDX-1 ⁺ cells was also transiently higher at 2 days post SZ than in controls (Fig. 3) although it did not reached the level previously determined in embryos (40.2% ± 3.5; Ref 3). These observations suggest that in contrast to embryos, the IN⁺ cells are produced in adults from PDX-1⁺ precursors expressing SOM (and possibly also PP).

View larger version (21K): [in this window] [in a new window]   Figure 3. Percentage of GLU+PDX-1+ and of PP+PDX-1+ in islets after SZ. The percentage of GLU+PDX-1+ and of PP+PDX-1+ cells at different times after the injection of 200 mg/kg SZ is compared with littermate controls. At least three pancreas per stage per antibody combinations were examined. The total number of cells scored for the GLU+/PDX-1+ combination was 1176 from 6 islets for 1-day-old post SZ, 1032 from 8 islets for 2 days post SZ, and 1108 from 10 islets for 5 days post SZ. Total number of cells scored for the PP+/PDX-1+ combination was 1565 from 14 islets for 1 day post SZ; 1536 cells from 19 islets for 2 days post SZ and 2196 cells from 23 islets for 5 days post SZ. The number of cells expressing PDX-1 and hormone is expressed as the mean percentage ± SEM of the cells immunoreactive for the hormone. In this figure, the number of days after SZ injection = x days SZ. *, P < 0.001 vs. controls

View larger version (33K): [in this window] [in a new window]   Figure 4. The percentage of SOM+/PDX-1+ cells increases in islets after SZ injection. The percentage of SOM+PDX-1+ cells at different times after the injection of 200 mg/kg SZ is compared with the values determined in littermate controls. The percentage of double labeled cells in islets of mice injected with 100 mg/kg SZ, is indicated for comparison. Note the large increase in the percentage of SOM+PDX-1+ cells in islets of SZ-treated mice. At least three pancreas per stage per animal models were examined. Total number of cells scored for the SOM+/PDX-1+ combination in control islets was 1389 from 18 islets. IN 200 SZ mice, the number of cells scored was 1029 cells from 14 islets for 1 day post SZ, 1004 cells from 11 islets for 2 days post SZ, 2564 cells from 31 islets for 5 days post SZ, and 1496 cells from 18 islets for 30 days post SZ. Total number of cells counted for the SOM+/PDX-1+ (100 SZ) combination was 1308 cells from 13 islets for 2 days post SZ, 1058 from 16 islets for 5 days post SZ, and 1496 cells from 18 islets for 30 days post SZ. The number of cells expressing SOM and PDX-1 is expressed as the mean percentage ± SEM of the cells immunoreactive for SOM. In this figure, the number of days after SZ injection = x days SZ. *, P < 0.001 vs. controls.

SOM⁺PDX-1⁺ cells also appeared in islets of NOD mice
To ascertain whether the increase in the percentage of SOM⁺/PDX-1⁺ cells occurred in other cases of islet destruction, we analyzed islets from 1, 2, and 4-month-old NOD mice. All 1-month-old NOD mice examined were normoglycemic, and their islets had no visible evidence of lymphocytic infiltration (score = 0). Two-month-old NOD mice were also normoglycemic and had a combined islet cell destruction and insulinitis score of 2 (criteria for score determination is described in Materials and Methods). At 4 months, one of the NOD mice analyzed had mild hyperglycemia (156 mg/dl), whereas the other three mice were severely hyperglycemic (blood glucose over 400 mg/dl). Pancreas of both normoglycemic and hyperglycemic 4-month-old mutant mice contained islets with different degrees of lymphocyte infiltration and endocrine cell number with a combined islet destruction score 3. As expected, the number of affected islets was greater in hyperglycemic animals. Examination of sections of NOD pancreas processed for double immunostaining indicated that the percentage of SOM⁺/PDX-1⁺ cells was always higher in NOD mutant islets than in littermate controls (Figs. 5 and 6). However, there was a large variation in the percentage of SOM⁺/PDX-1⁺ cells, which ranged from 30% in islets with a normal morphology, to almost 80% in islets that were severely infiltrated. These findings suggest that the PDX-1⁺/SOM⁺ cells differentiate in islets of NOD mice and that the appearance of these cells occurs before the disruption of the structural organization of the islets.

View larger version (43K): [in this window] [in a new window]   Figure 5. Coexpression of PDX-1 and SOM by islet cells of NOD mice. The percentage of SOM+PDX-1+ cells in 1-, 2-, and 4-month-old NOD mice is compared with the values determined in littermate controls. Note the large increase in the percentage of SOM+PDX-1+ cells in islets of NOD mice at all stages examined. At least three pancreas per stage per animal models were examined. Total number of cells scored for the SOM+/PDX-1+ combination for the NOD mouse models were 1230 cells from 20 islets for 1-month-old mice, 1032 cells from 13 islets for 2-month-old mice, and 1437 cells from 21 islets for 4-month-old mice, respectively. For controls, at least 1000 cells from 12 different islets were scored per stage. The number of cells expressing PDX-1 and SOM is expressed as the mean percentage ± SEM of the cells immunoreactive for SOM. In this figure, the number of days after SZ injection = x days SZ. *, P < 0.001 vs. littermate controls

View larger version (111K): [in this window] [in a new window]   Figure 6. Most SOM+ cells of islets of NOD mice express PDX-1. Localization of SOM (blue reaction product) and PDX-1(brown precipitate in nuclei) in islets of 4-month-old NOD mice. These photomicrographs illustrates that the percentage of PDX-1+/SOM+ cells in islets of NOD mice is correlated with the degree of lymphocyte infiltration and islet cell disruption. A, Photomicrograph illustrates an islet in an initial stage of lymphocyte infiltration. Location of lymphocytes is indicated with an asterisk. A core of PDX-1+ cells, presumably ß cells, is surrounded by non-ß cells. Bar = 30 µm. B, High magnification photomicrograph of islet shown in A. In A and B, a PDX-1+SOM+ cell is indicated with an arrowhead. Bar = 4 µm. C, Photomicrograph illustrates an islet that is severely infiltrated with lymphocytes (asterisk) and contains few endocrine cells. Bar = 30 µm D, High magnification photomicrograph of islet shown in C illustrates that almost all cells of this islet express the homeoprotein in their nuclei. Bar = 4 µm.

View larger version (141K): [in this window] [in a new window]   Figure 7. In controls, most cells do not express IN. Immunostaining of consecutive 1 µm epon sections of a control islet. High magnification photomicrographs of a pole of the islet containing many cells. A (SOM) and B (IN) illustrate that most, if not all IN+ cells of controls lack SOM. Some of the cells that contain SOM in A but lack IN in B are indicated with arrowheads. Cell indicated with an arrow contains only IN. Bar = 4 µm.

View larger version (137K): [in this window] [in a new window]   Figure 8. Cells coexpressing SOM and IN appear in 200 SZ mice. Immunostaining of consecutive 1 µm epon sections of SZ treated mice. A and B, 2 days post SZ. At this stage SOM+ cells (A) do not express IN (B). Note lack of double labeled cells, i.e. cluster of cells indicated with an arrow contain only SOM. Similarly, cells containing only IN is illustrated with an arrowhead. Bar = 15 µm. C and D, 5 days post SZ. Many SOM+ cells (C) coexpress IN (D). Note that the distribution of cells in these two sections is overlapping. Many cells coexpress both hormones and some of these cells are indicated with arrowheads. Right side of both photomicrographs is out of focus and the profile of individual cells is difficult to distinguish. Indicated with a cross in C is a cell that contains only SOM, whereas a cell that contains only IN (D) is indicated with an asterisk. Bar = 4 µm. In some islets of 5 days post SZ pancreas, shown in E and F, SOM+ cells do not express IN. E, SOM; F, IN. Note the lack of overlap in the distribution of SOM+ and IN+ cells, the presence of few ß and cells and the absence of cells expressing both hormones. Arrowheads indicate SOM+ IN- cells; asterisks indicate identical regions in both photomicrographs and are included to facilitate the comparison of the distribution of stained cells. Bar = 20 µm.

View larger version (141K): [in this window] [in a new window]   Figure 9. Most SOM cells in islets of 100 SZ mice do not express IN. Immunostaining of consecutive 1 µm epon sections at 5 days post SZ: A, SOM; B, IN. Note lack of double labeled cells, i.e. cells indicated with arrowheads in A contain only SOM. Cell indicated with an arrow contains only IN. Bar = 4 µm.

SOM⁺/PDX-1⁺ cells are generated from mitotically quiescent precursors
Next, we sought to determine whether the SOM⁺/PDX-1⁺ cells present at 2 days post SZ were generated from a proliferating stem cell-like population. To address this question, 200 SZ mice that were injected with ³H-thymidine at 1 day post SZ were killed 1 day later and the pancreas processed for SOM immunohistochemistry and autoradiography. At 2 days post SZ, we did not find SOM⁺ cells with ³H-thymidine in their nuclei (data not shown), indicating that the SOM⁺/PDX-1⁺ cells present at 2 days post SZ were produced from a quiescent precursor cell population.

While islet cells were mitotically quiescent at 1 day post SZ, a significant proportion of SOM⁺ and of PDX-1⁺ cells entered the cell cycle at 2 days post SZ. Thus, 2 days post SZ, mice injected with ³H-thymidine and killed 1 h later contained a large number of proliferating SOM⁺ and PDX-1⁺ cells (Table 2). The increase in the rate of proliferation was transient, and the labeling index of SOM⁺ and PDX-1⁺ cells decreased back to control levels by 5 days post SZ (Table 2). The mitogenic effect of SZ-induced ß cell depletion was not specific for SOM⁺ and PDX-1⁺ cells because the labeling index of GLU⁺ and PP⁺ cells of islets from 200 SZ mice also transiently increased at 2 days post SZ (Table 2). At this stage, most islets did not contain IN⁺ cells, and the occasional ß cell present at that stage did not proliferate. In addition, the islet core contained dividing cells that did not express either PDX-1 or islet hormones (data not shown). At 5 days post SZ, the rate of proliferation of the newly differentiated ß cells was similar to controls (Table 2).

Precursor cells in the pancreatic duct did not have a major role in ß cell neogenesis
It is generally believed that the pancreatic duct in adults contains islet precursor cells (1). In agreement with this hypothesis, we found some PDX-1⁺ cells in the pancreatic duct of SZ-treated and NOD mice. Some of these PDX-1⁺ cells coexpressed IN, GLU, SOM, or PP (Fig. 10), whereas other cells expressed GLU, SOM or PP, but not PDX-1 (not shown). We examined the mitotic index of these ductal cells in pancreata from 1, 2, and 5 days SZ mice. This analysis indicated that most of the PDX-1⁺ and the hormone-containing cells present in the duct had not incorporated ³H-thymidine (data not shown). In contrast, the acinar and connective tissues contained many cells that had incorporated the isotope (data not shown). The fact that the number of cells expressing PDX-1⁺ and/or pancreatic hormones in the pancreatic duct of 1, 2, and 5 days post SZ was very low, and that most of these cells did not proliferate, indicates that the pancreatic duct was not a major source of islet stem cells in SZ-treated mice.

View larger version (115K): [in this window] [in a new window]   Figure 10. Different endocrine cell types appear in the pancreatic duct of SZ-treated mice. Immunocytochemical localization of PDX-1 (brown precipitate) and a pancreatic hormone (blue reaction product) in cryostat sections of 5 days post SZ pancreas. A, PDX-1 and SOM immunostaining. Longitudinal section through a branch of the pancreatic duct. Cells indicated with arrows are shown with higher magnification in B. Bar = 20 µm. B, High magnification photomicrograph illustrates the presence of two PDX-1+/SOM+ cells. Nuclear staining in cells surrounding the cavity of the duct is an artifact due to folding of the tissue. C, PDX-1+ GLU+ cells. Oblique section of pancreatic duct. D, Photomicrograph of a cross section of a pancreatic duct documents the presence of a PDX-1+ PP+ cell. In C and D, unstained ductal cells are indicated with an asterisk and the border of the ductal epithelium is indicated with arrowheads. Bars of B–D = 4 µm

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

View larger version (18K): [in this window] [in a new window]   Figure 11. Model of endocrine cell lineage (modified from Guz et al., 1995). During embryonic development, stem cells coexpressing the transcription factor PDX-1 and multiple hormones generate the mature monospecific islet cells. The derivation of all a cells from PDX-1+ cells is uncertain because PDX-1 knockout mice contained a population of GLU+ cells (Offield et al., 1996). In adults, PDX-1 is expressed by 90% of ß cells and by small subsets of , , and PP cells. Our present findings suggest that, in the SZ model, neoformed ß cells are generated by SOM+ PDX-1+ precursor cells that initiate and IN expression eventually mature into monospecifc ß cells. This pathway of neogenesis is indicated with dashed arrows.

It could be proposed that the SOM⁺/IN⁺ cells that appeared at 5 days post SZ were generated by the few ß cells that survived the toxic injury and that these cells proliferated and initiated SOM expression. Two observations argue against this possibility. First, in agreement with previous reports (17), our ³H-thymidine-labeling experiments indicate that the occasional ß cell that survived the SZ treatment did not proliferate during the first few days following the injection of the drug. In addition, we showed that a large number of SOM⁺ cells coexpressed IN at 5 days SZ. These observations are consistent with the proposition that the SOM⁺ PDX-1⁺ at 2 days post SZ serve as the precursor population for the SOM⁺IN⁺ cells that populated islets of 5 days post SZ mice.

Islets of the 100 SZ mice also had many SOM⁺PDX-1⁺ cells. However, in contrast to the 200 SZ model, islets of 100 SZ did not contain SOM⁺ IN⁺ cells. Conceivably, islets of the 100 SZ mice lack signals to promote the initiation of IN expression in SOM⁺PDX-1⁺ transitional cells. Islets of NOD mice also contain SOM⁺/PDX-1⁺ cells and lack SOM⁺/IN⁺/PDX-1⁺ cells. It is possible that the absence of demonstrable neogenesis of IN⁺ cells in NOD mice could be due to the lack of specific differentiation factors. However, we favor the possibility that islets of NOD mice contained SOM⁺/PDX-1⁺/IN⁺ cells but that these cells remained undetected in our studies because they had low levels of IN and/or they were rapidly eliminated by the immune system.

Source of PDX-1⁺/SOM⁺ cells
What is the source of the PDX-1⁺/SOM⁺ precursors that differentiate to IN⁺ cells? One source could be intra-islet hormone-null cells that do not express endocrine specific antigens. However, extensive histological examination of the islets has not yet revealed such a population of undifferentiated cells (44). A second source of ß precursor cells could be the SOM⁺/PDX-1⁺ cells present in control (untreated) islets. Because 17% of the cells of control islets expressed PDX-1, it seems reasonable to propose that these cells initiate IN expression and then generate monospecific IN⁺ cells, following the injection of SZ (Fig. 11). Finally, a third source of ß precursor cells could be the cells expressing SOM only. According to this possibility, monospecific SOM⁺ cells are induced by islet injury to reexpress PDX-1 and IN. In the 200 SZ model, some islets retained a near normal complement of SOM⁺ cells while other islets contained only a few SOM⁺ cells (Fig. 8). IN⁺ cells appeared only in islets that contained many SOM⁺ cells. This observation suggests that, in some islets, many SOM⁺ cells are ablated by high dosage of SZ and that the surviving population of SOM⁺ cells is unable to initiate PDX-1 and IN expression.

Embryonic and adult ß progenitor cells expressed different phenotypes
We have now identified two subsets of PDX-1⁺ ß precursor cells. The IN⁺/GLU⁺/PDX-1⁺cell type, which can also express the neuronal marker TH (45), lead to monospecific ß cells in embryos, whereas IN⁺/SOM⁺/PDX-1⁺cells, a cell type at a more advanced stage in the cellular hierarchy (Fig. 11), are proposed to generate IN⁺ cells in regenerating islets of adults.

Interestingly, embryonic SOM⁺/IN⁺/PDX-1⁺ cells differentiate into monospecific SOM⁺ cells (3), whereas we show here that these cells produce IN⁺ cells during islet regeneration in adults. Thus, it is likely that the PDX-1⁺ precursor cell type is committed to distinct developmental pathways in embryos and adults (Fig. 11). An alternate possibility is that IN/SOM⁺ stem cells are bipotential and produce monospecific IN and SOM cells during development and regeneration. It also remains to be determined whether the IN⁺ cells generated in SZ-treated islets from SOM⁺/PDX-1⁺ precursors contain all the properties characteristic of mature ß cells.

Possible signals for ß cell neogenesis in adult islets
Two precursor cell types appeared sequentially in adult islets during neogenesis. The first islet cell type to appear was the SOM⁺/PDX-1⁺ cells, which differentiated in islets of 100 SZ, 200 SZ, and NOD mice. Because the first two groups of mice were exposed to both high glucose leveIs and islet injury, it could be suggested that these two events cooperated in signaling the activation of PDX-1 expression by SOM⁺ cells. However, the finding that islets of normoglycemic young NOD mice contained large number of SOM⁺/PDX-1⁺ cells argues against a role of hyperglycemia in the differentiation of this precursor cell type. Taken together, these observations suggest that islet injury may be the principal event that initiated ß cell precursor differentiation. Our finding that structural normal islets of 1-month-old NOD mice contained a large percentage of SOM⁺/PDX-1⁺ cells appears to contradict this hypothesis. However, using an immunological rather than morphological approach, others identified infiltrating lymphocytes in islets of even younger NOD mice (46), indicating that the cellular environment of apparently normal NOD islets is perturbed very soon after birth. The identity of the injury-induced factors that stimulate PDX-1 expression in SOM⁺ cells is not known.

The second cell type that appeared during neogenesis in adult islets was the SOM⁺/IN⁺/PDX-1⁺ cells. The fact that the expression of PDX-1 by SOM⁺ cells precedes the appearance of IN suggests that this transcription factor is required for initiating IN gene expression. However, whereas islets of 200 SZ mice contained many SOM⁺/IN⁺/PDX-1⁺ cells, islets of 100 SZ mice had large numbers of SOM⁺/PDX-1⁺ cells but only occasional SOM⁺/IN⁺ cells. This finding indicates that other factors are required to induce the conversion of the transitional SOM⁺/PDX-1⁺ cells to IN-containing cells and that these signals are present in islets of 200 SZ but not of 100 SZ mice. One possible source of differentiation signals are the macrophages that invade the islet and phagocytose necrotic ß cells. Presumably, the number of macrophages in islets correlates with the degree of islet injury and is higher in islets of 200 SZ than in islets of 100 SZ mice. Macrophages produce cytokines involved in many crucial cellular processes (47, 48), and one or more of these substances could participate in the activation of the IN gene in SOM⁺/PDX-1⁺cells.

The differentiation and proliferation of the population of IN⁺ stem cells may also be induced by the dramatic alteration in the three dimensional architecture produced during the elimination of the ß cell core in the islet. This hypothesis is particularly compelling because IN⁺ cell neogenesis occurred in islets of mice injected with 200 SZ but not in the less damaged islets of 100 SZ mice. There appears to be a tight relationship between islet structure and function (49). Thus, alterations in intraislet cellular interactions may lead to profound changes in islet cell properties. These observations suggest that depletion of the normal complement of ß cells in islets of SZ treated and NOD mice may enable peripheral islet cell types to express novel phenotypic traits. Finally, the possibility that SZ itself promoted ß cell neogenesis has to be considered. As an alkylating agent, SZ induces DNA strand breaks and inhibits DNA repair (30) and may cause DNA recombinations that leads to the differentiation of islet precursor cells.

Although islets of 200 SZ mice contained a significant number of newly differentiated ß cells, the animals remained severely hyperglycemic at all stages, indicating that these cells fail to detect and/or respond efficiently to the increased blood glucose levels. It has been shown that excessive glucose stimulation permanently affects ß cell function (50, 51, 52). Conceivably, the new IN⁺ cells produced in SZ-treated mice are functionally defective because their differentiation occurs in a hyperglycemic environment. Because SZ impairs the response of mature ß cells to glucose (reviewed in Ref.28), it is also possible that the drug has a similar effect on ß stem cells. It would be of great interest to determine why these cells are dysfunctional and then to characterize conditions for appropriate maturation of precursor cells in adults to functionally normal glucose-sensitive, IN-producing ß cells. The answers to such questions could have important implications for the treatment of type I diabetes in humans.

	Acknowledgments

 
The authors thank Lyudmila Aleyner (SUNY, Brooklyn) for her help with the quantitative analysis.

	Footnotes

 
¹ Part of this work has been presented at the 55th Annual Meeting of the American Diabetes Association, Atlanta, Georgia, 1995, and as a "State of the Art" lecture at the meeting on "Immunology of Diabetes" in Orvieto, Italy, 1995. This work was supported by grants from the NIH, the American Diabetes Association, and in part by the Vanderbilt University Diabetes Research and Training Center Molecular Biology Core Laboratory and the Foundation for Medical Research.

Received September 6, 1996.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Pictet R, Rutter WJ 1972 Development of the embryonic pancreas. In: Steiner DF, Frenkel N (eds) Handbook of Physiology, Section 7. American Physiological Society, Washington, DC, pp 25–66
2. Alpert S, Hanahan D, Teitelman G 1988 Hybrid insulin genes reveal a developmental lineage for pancreatic endocrine cells and imply a relationship with neurons. Cell 53:295–308
3. Guz Y, Montminy MR, Stein R, Leonard J, Gamer LW, Wright CVE, Teitelman G 1995 Expression of murine Pdx 1, a putative insulin gene transcription factor, in ß cells of pancreas, duodenal epithelium and pancreatic exocrine and endocrine progenitors during ontogeny. Development 121:11–18
4. Hashimoto T, Kawano H, Daikoku S, Shima K, Taniguchi H, Baba S 1988 Transient coappearance of glucagon and insulin in the progenitor cells of the rat pancreatic islets. Anat Embryol 178:489–497
5. Herrera PL, Huarte J, Sanvito F, Meda P, Orci L, Vassalli JD 1991 Embryogenesis of the murine endocrine pancreas; early expression of the pancreatic polypeptide gene. Development 113:1257–1265
6. Upchurch BH, Aponte GW, Leiter AB 1994 Expression of peptide YY in all four islet cell types in the developing mouse pancreas suggests a common peptide YY producing progenitor. Development 120:245–252
7. Jonsson J, Carlsson L, Edlund T, Edlund H 1994 Insulin-promoter-factor 1 is required for pancreas development in mice. Nature 371:606–609
8. Offield MF, Jetton TL, Labosky PA, Ray M, Stein RW, Magnuson MA, Hogan BLM, Wright CVE 1996 PDX-1 is required for pancreatic outgrowth and differentiation of rostral duodenum. Development 122:983–995
9. Leonard J, Peers B, Johnson T, Ferreri K, Lee S, Montminy MR 1993 Characterization of Somatostatin Transactivating Factor 1, a novel homeobox factor that stimulates somatostatin expression in pancreatic islet cells. Mol Endocrinol 7:1275–1283
10. Miller CP, McGehee Jr RE, Habener JF 1994 IDX-1: a new homeodomain transcripion factor expressed in rat pancreatic islets and duodenum that transactivates the somatostatin gene. EMBO J 13:1145–56
11. Ohlsson H, Karlsson K, Edlund T 1993 IPF1, a homeodomain-containing transactivator of the insulin gene. EMBO J 12:4251–4259
12. Peers B, Leonard J, Sharma S, Teitelman G, Montminy MR 1994 Insulin expression in pancreatic islet cells relies on cooperative interactions between the helix loop helix factor E47 and the homeobox factor PDX 1. Mol Endocrinol 8:1798–806
13. Peshavaria M, Gamer L, Henderson E, Teitelman G, Wright CVE, Stein R 1994 XIHbox 8, an endoderm-specific Xenopus homeodomain protein, is closely related to a mammalian insulin gene transcription factor. Mol Endocrinol 8:806–816
14. Petersen HU, Serup P, Leonard J, Michelsen BK Madsen OD 1994 Transcriptional regulation of the human insulin gene is dependent on the homeodomain proteins STF 1/IPF 1 acting through the CT boxes. Proc Natl Acad Sci USA 91:10465–10469
15. De Krijger RR, Aanstoot HJ, Kranenburg G, Reinhard M, Visser WJ, Bruining G J 1992 The midgestational human fetal pancreas contains cells co-expressing islet hormones. Dev Biol 153:368–375
16. Lukinius A, Ericsson JLE, Grimelius L, Korsgren O 1992 Electron microscopic immunocytochemical study of the ontogeny of fetal human and porcine endocrine pancreas, with special reference to co-localization of the four major islet hormones. Dev Biol 153:376–390
17. Logothetopoulos J 1972 Islet regeneration and neogenesis. In: Freinkel N (ed) Handbook of Physiology: Endocrinology, Vol 1. Waverly Press, Inc., Baltimore, pp 67–76
18. Hellestrom C, Andersson A, Gunnarsson R 1976 Regeneration of islet cells. Acta Endocrinol [Suppl 203] 83:145–158
19. Hellestrom C, Swenne I 1985 Growth pattern of pancreatic islet cells in animals. In: Volk BV, Arquilla ER (eds) The Diabetic Pancreas, ed. 2. Plenum Medical Book Co., New York, pp 53–80
20. Teitelman G, Alpert S, Hanahan D 1988 Proliferation, senescence and neoplastic progression of beta cells in hyperplastic pancreatic islets. Cell 52:97–105
21. Brelje TC, Parsons JA, Sorenson RL 1994 Regulation of islet ß-cell proliferation by prolactin in rat islets. Diabetes 43:263–273
22. Dutrillaux M C, Portha B, Roze C, Holland E 1982 Ultrastructural study of pancreatic ß cell regeneration in newborn rats after destruction by streptozotocin. Virchows Arch (Cell Pathol) 39:173–185
23. Rosenberg L, Vinik AI 1989 Induction of endocrine cell differentiation:a new approach to management of diabetes. J Lab Clin Med 114:75–83
24. Bonner-Weir S, Baxter LA, Schuppin GT, Smith FE 1993 A second pathway for regeneration of adult exocrine and endocrine pancreas. Diabetes 42:1715–720
25. Gu D, Sarvetnick N 1993 Epithelial cell proliferation and islet neogenesis in IFN-g transgenic mice. Development 118:33–46
26. Gu D, Lee MS, Krahl T, Sarvetnick N 1994 Transitional cells in the regenerating pancreas. Development 120:1873–1881
27. Like AA, Rossini A A 1976 Streptozotocin-induced pancreatic insulinitis: new model of diabetes mellitus. Science 193:415–418
28. Shafrir E 1990 Diabetes in animals. In: Rifkin H, Porte D (eds) Diabetes Mellitus. Theory and Practice. Elsevier, New York, pp 299–340
29. Junod A, Lambert AE, Stauffacher W, Renold AE 1969 Diabetogenic action of streptozotocin: relationship of dose to metabolic response. J Clin Invest 48:2129–2139
30. Eizirik DL, Sandler S, Ahnstrom G, Welsh M 1991 Exposure of pancreatic islets to different alkylating agents decreases mitochondrial DNA content but only streptozotocin induces long-lasting functional impairment of ß-cells. Biochem Pharmacol 42:2275–2282
31. Brodsky G, Logothetopoulos J 1969 Streptozotocin diabetes in the mouse and guinea pig. Diabetes 18:606–611
32. Teitelman G, Alpert S, Polak J M, Martinez A, Hanahan D 1993 Precursor cells of mouse endocrine pancreas coexpress insulin, glucagon, and the neuronal proteins tyrosine hydroxylase and neuropeptide Y, but not pancreatic polypeptide. Development 118:1031–1039
33. Wright CVE, Schnegelsberg P, De Robertis E 1988 XIHbox 8: a novel Xenopus homeoprotein restricted to a narrow band of endoderm. Development 104:787–794
34. Larsson LI 1981 A novel immunocytochemical model system for specificity and sensitivity screening of antisera against multiple antigens. J Histochem Cytochem 29:408–410
35. Klein MV, Furtak TE 1986 Optics. John Wiley and Sons, New York
36. Weibel ER 1979 Stereological Methods. Practical Methods for Biological Morphometry, chapter 4, pp 101–169, Academic Press, New York
37. Hellestrom C 1977 Growth pattern of pancreatic islets in animals. In: Volk BV, Wellman KE (eds) The Diabetic Pancreas, ed 1. Plenum Medical Book Co., New York, pp 61–97
38. Atkinson MA, Maclaren NK, Luchetta R 1990 Insulinitis and diabetes in NOD mice reduced by prophylactic insulin therapy. Diabetes 39:933–937
39. Rogers A 1973 Techniques of Autoradiography. Elsevier, New York
40. Maxwell M H 1978. Two rapid and simple methods used for the removal of resins from 1.0 µm thick epoxy sections. J Microscopy 112:253–255
41. Junod A, Lambert AE, Orci L, Pictet R, Gonet AE, Renold AE 1967 Studies of the diabetogenic action of streptozotocin. Proc Soc Exp Biol Med 126:201–205
42. Olson LK, Sharma A, Peshavaria M, Wright CVE, Towle HC, Robertson RP, Stein R 1995 Reduction of insulin gene transcription in HIT-T15 ß cells chronically exposed to a supraphysiologic glucose concentration is associated with loss of SFT-1 transcription factor expression. Proc Natl Acad Sci USA 92:9127–9131
43. Steiner H, Oelz O, Zahn G, Froesch E R 1970 Studies on islet regeneration, hyperplasia and intrainsular interrelations in long lasting streptozotocin diabetes in rats. Diabetologia 6:558–564
44. Fawcett DW 1960 A Texbook of Histology, ed 11. W. B. Saunders, Philadelphia
45. Teitelman G, Lee J K 1987 Cell lineage analysis of pancreatic islet cell development. Glucagon and insulin cells arise from catecholaminergic precursors in the pancreatic duct. Dev Biol 121:454–466
46. Yang Y, Charlton B, Shimada A, Dal Canto R, Fathman CG 1996 Monoclonal T cells identified in early NOD islet infiltrates. Immunity 4:189–194
47. Nathan C F 1987 Secretory products of macrophages. J Clin Invest 79:319–326
48. Merrill JE, Jonakait GM 1995 Interactions of the nervous and immune systems in development, normal brain homeostasis, and disease. FASEB J 9:611–618
49. Gepts W, LeCompte P M 1985 The Pathology of Type I (juvenile) Diabetes. In: Volk BW, Arquilla ER (eds): The Diabetic Pancreas. Plenum Medical Book Co., New York, pp 337–60
50. Bonner-Weir S, Trent DF, Honey RN, Weir GC 1981 Responses of neonatal rat islets to streptozotocin. Limited ß-cell regeneration and hyperglycemia. Diabetes 30:64–69
51. Bonner-Weir S, Trent DF, Weir GC 1983 Partial pancreatectomy in the rat and subsequent defect in glucose-induced insulin release. J Clin Invest 71:1544–1554
52. Wier GC, Bonner-Weir S, Leahy JL 1991 Adverse effects of hyperglycemia upon the islets of Langerhans. In: Samols E (ed) The Endocrine Pancreas, Raven Press, New York, pp 337–354

This article has been cited by other articles:

				Y. Sun, L. Zhang, H. F. Gu, W. Han, M. Ren, F. Wang, B. Gong, L. Wang, H. Guo, W. Xin, et al. Peroxisome Proliferator-Activated Receptor-{alpha} Regulates the Expression of Pancreatic/Duodenal Homeobox-1 in Rat Insulinoma (INS-1) Cells and Ameliorates Glucose-Induced Insulin Secretion Impaired by Palmitate Endocrinology, February 1, 2008; 149(2): 662 - 671. [Abstract] [Full Text] [PDF]

				M. Kodama, F. Takeshita, S. Kanegasaki, T. Ochiya, and G. Quinn Pancreatic Endocrine and Exocrine Cell Ontogeny From Renal Capsule transplanted Embryonic Stem Cells in Streptozocin-injured Mice J. Histochem. Cytochem., January 1, 2008; 56(1): 33 - 44. [Abstract] [Full Text] [PDF]

				S. Hanley and L. Rosenberg Transforming Growth Factor {beta} Is a Critical Regulator of Adult Human Islet Plasticity Mol. Endocrinol., June 1, 2007; 21(6): 1467 - 1477. [Abstract] [Full Text] [PDF]