Abstract
The aim of this study was to evaluate the location of PP and δ cells in relation to the vascularization within human pancreatic islets. To this end, pancreas sections were analysed by immunofluorescence using antibodies against endocrine islet and endothelial cells. Staining in different islet areas corresponding to islet cells adjacent or not to peripheral or central vascular channels was quantified by computerized morphometry. As results, α, PP and δ cells were preferentially found adjacent to vessels. In contrast to α cells, which were evenly distributed between islet periphery and intraislet vascular channels, PP and δ cells had asymmetric and opposite distributions: PP staining was higher and somatostatin staining was lower in the islet periphery than in the area around intraislet vascular channels. Additionally, frequencies of PP and δ cells were negatively correlated in the islets. No difference was observed between islets from the head and the tail of the pancreas, and from type 2 diabetic and non-diabetic donors. In conclusion, the distribution of δ cells differs from that of PP cells in human islets, suggesting that vessels at the periphery and at the centre of islets drain different hormonal cocktails.
Introduction
Islets of Langerhans comprise different endocrine cell types. β cells secreting insulin and α cells secreting glucagon are the main islet cell types in most animal species, including human, and exhibit a clear and established function in glucose homeostasis. These two hormones have diametrically opposite actions on glucose regulation, and consequently, the secretion of one hormone is controlled by the other one. δcells secreting somatostatin and PP cells secreting pancreatic polypeptide are less abundant (Bosco et al. 2005), and less attention has been paid to their function in metabolism. Nevertheless, somatostatin has been shown to inhibit both insulin and glucagon secretions (Schuit et al. 1989) and thus participates, albeit indirectly, in glucose homeostasis. Pancreatic polypeptide is mostly involved in satiety (Batterham et al. 2003) and in the regulation of pancreatic exocrine secretion (Putnam et al. 1989, Morisset 2008), but is known to exhibit some glucose regulatory function as well, for instance by modulating insulin receptor expression in hepatocytes (Seymour et al. 1996) and inhibiting glucagon release in mouse pancreatic islets (Aragon et al. 2015). It is generally accepted that local interactions between the different islet hormones play fundamental roles in the islet physiology, and some definite interactions are facilitated by the specific positioning of cell types within the islets (Hauge-Evans et al. 2009, Unger & Orci 2010, Rodriguez-Diaz et al. 2011, Koh et al. 2012).
A prototypical islet description was conducted from studies in rodents showing a relatively simple islet organization with a core of β cells surrounded by a mantle of other cell types. It has been proposed that this specific architecture is functionally relevant. Indeed, taking into account the microvascularization and the direction of the blood flow within the islet, one can conclude that a given secreted hormone (e.g., insulin) may affect the secretion of another one (e.g., glucagon) within the same islet (Bonner-Weir & Orci 1982, Marks et al. 1990). In humans, islets have a more complex cell organization and exhibit heterogeneous cell composition that furthermore varies according to their location into the pancreas (Stefan et al. 1982, Kim et al. 2009, Bosco et al. 2010, Wang et al. 2013, Bonner-Weir et al. 2015). Whereas smaller human islets have a core–mantle structure quite similar to that of rodent islets, the larger human islets have a more complex organization with endocrine cells organized into a continuous folded epithelial band (Bosco et al. 2010), evoking fused or even discrete core–mantle subunits in pancreas sections. For reasons of simplification, we hereafter use the term “subunit” to refer to a cross section through folded epithelial band.
In human pancreas, blood vessels are adjacent to the periphery of islets and others are found to invaginate into the central part and transit through vascular channels between the subunits in larger islets. Therefore, the mantle cells in subunit clusters are adjacent either to peripheral or to invaginating vessels (Bosco et al. 2010). The blood flow in human islets is so complex that a complete understanding of the microcirculation is lacking. In addition, it is not known whether the vessels at the periphery and those inside the islets originate from a common or different vessel ramification (Brunicardi et al. 1996, Ballian & Brunicardi 2007). Consequently, whether and how microcirculation affects the function of islet cell types is not understood either. Variations in islet cell composition, mostly related to α and PP cells, have been shown between ventral and dorsal parts of the human pancreas (Orci et al. 1976, Baetens et al. 1979, Stefan et al. 1982). Interestingly, these two pancreas parts are known to be irrigated by two arterial systems (Orci et al. 1976, Bertelli et al. 1997), suggesting that the nonhomogeneous distribution of cell types in islets may have functional importance. Whether cell distribution is similar between subunit mantle regions adjacent to the peripheral and invaginating vessels has not yet been investigated. In this work, we investigated the distribution of α, β, δ and PP cells by immunofluorescence on pancreas sections according to three different areas of the islets: the periphery (adjacent to peripheral vessels), the area adjacent to intraislet vascular channels and the area distant from periphery and intraislet vascular channels (the core of subunits). We also analysed the relative frequencies of PP and δcells in the very same islets and studied whether the distributions of PP and δcells differed between islets from the head and the tail of the pancreas, and from type 2 diabetic and non-diabetic donors.
Materials and methods
Pancreas procurement
Human pancreata were harvested from adult heart-beating brain-dead donors and designed to be processed for islet isolation and transplantation. Within the last 15 years, for diverse adverse motives (prolonged time for ischaemia, suspicion of tumours, type 2 diabetes, unforeseen technical difficulties, etc.), some pancreata were not processed for islet isolation and small specimens were taken for histology before organ destruction. Specimens from 25 donors were used for these analyses. Of these specimens, 18 were from non-diabetic donors with a mean age of 45±16 years (range 15–69 years) and a mean body mass index (BMI) of 25±4 (range 20–34). Specimens from non-diabetic donors were obtained from the head (n=3), body (n=2), tail (n=4) and unspecified regions (n=8) of the pancreas. For two pancreata, specimens from both the head and the tail were available. Specimens from type 2 diabetic donors were obtained from the head (n=2), body (n=2), tail (n=4) and unspecified regions (n=3) of the pancreas. Type 2 diabetic donors had a mean age of 58±11 years (range 26–65 years) and a BMI of 31±8 (range 26–51). Glycaemia at the time of hospital admission was 7.5±3.7 mmol/L (mean±s.d., n=13) for non-diabetic donors and 14.4±6.2 mmol/L (mean±s.d., n=6) for type 2 diabetic donors. Some control and diabetic specimens were the same used in a precedent work (Bosco et al. 2010). Histopathological analysis of the diabetic specimens revealed that several β cells were positive for amylin (Bosco et al. 2010).
Histological sections of pancreata
Pancreas samples were incubated for 24 h in 10% neutral buffered formalin, then dehydrated and embedded in paraffin before being sectioned at 5μm. For each pancreas sample, one to three sections, separated by 500μm, were randomly performed in order to analyse at least four islets per pancreas sample. All islets on the pancreas sections that fit with the requirements (presence of at least one vascular channel, one δ cell, or one PP cell) were analysed.
Immunofluorescence
Sections were deparaffinized and rehydrated with a series of alcohol solutions of decreasing concentrations. Sections were then washed with phosphate buffer saline (PBS) and treated for 10 min at 37°C with a 0.1% trypsin solution. They were washed again with PBS and successively treated for 20 min with 0.5% Triton X100 in PBS and 20 min with 0.1% BSA in PBS at room temperature. Then, the sections were incubated at room temperature with antibodies diluted in PBS as follows. Sections were first incubated for 1h with a mouse anti-CD34 antibody (AbD; Serotec, Bio-Rad), diluted 1:20, washed and incubated with an Alexa 488-conjugated anti-mouse antibody (1:500; Invitrogen). After washing, the sections were incubated for 1 h with a guinea pig anti-insulin antibody (1:300; Dako), together with one of the following rabbit antibodies: anti-glucagon, diluted 1:100, anti-somatostatin, diluted 1:200 and anti-PP, diluted 1:200. After washing, the sections were incubated for 1h simultaneously with aminimethylcoumarin-conjugated anti-guinea pig antibody, diluted 1:400 and rhodamine-conjugated anti-rabbit antibody, diluted 1:200 (Jackson ImmunoResearch Laboratories).
Morphological analysis and quantifications
Microscopic sections were analysed using a Leica DM2000 microscope (Leica Microsystems) equipped with ultraviolet illumination and filters for blue, red and green fluorescences. Images were captured using a Leica digital camera (Leica Microsystems) and recorded on computer through the software (Leica Microsystems). Quantifications of stained areas were performed on digital images using the Metamorph imaging software for microscopy (Universal Imaging, West Chester, PA, USA). This software was programmed to automatically quantify stained areas within defined regions of interest. Insulin, glucagon, somatostatin and PP-stained areas were measured in islets comprising at least one intraislet vascular channel identified by CD34 staining. The islets displaying no vascular channel, which were excluded from the analysis, were mostly with a diameter of about 50μm or smaller (Bosco et al. 2010). Their frequencies were very variable between the pancreata, but did not exceed 20% of all islets. In each analysed islets, three regions were defined: peripheral region, the region around intraislet vascular channels and the core of islet subunits. The islet peripheral region was defined as the most outer layer of the cells, adjacent to islet surrounding blood vessels. Thus, a demarcation line was drawn around and 15μm distant from the outer islet border. The area around intraislet vascular channels was defined as one layer of cells adjacent to islet invaginating vessels. Thus, a demarcation line was drawn around and 15μm distant from the border of intraislet vascular channels. The remaining islet area (beyond the 15μm demarcation lines) was considered as the core of islet subunits. In each region, pixel area occupied by glucagon, somatostatin, or PP staining was measured (pixel area A) and expressed relative to that of insulin staining (pixel area B) by the formula 100× pixel area A/(pixel area A pixel area B). To study the relative frequencies of δ and PP cells in the same islets, pixel areas labelled for somatostatin or PP staining (using different fluorochromes) were measured in islets displaying at least one δ or PP cell. Pixel areas are then expressed relative to the pixels of the whole islet area measured manually by drawing the contour of the islets. Data were expressed as the mean±s.e.m. of islets (from all pancreata) or the means of islets by pancreata. Differences between means were assessed either by the Student’s t-test or, when required, by one-way ANOVA. When ANOVA was applied, Fisher’s least significant difference post hoc analysis was used to identify significant differences.
Results
The aim of this study is to compare islet cell distribution between peripheral and invaginating vessels; therefore, only islets presenting at least one recognizable intraislet vascular channel throughout pancreas sections were analysed. We first analysed islets from random sections of the pancreas of non-diabetic donors. In human islets, endocrine cells are organized into a continuous trilaminar folded epithelial band, which looks like discrete subunits in cross section (Bosco et al. 2010). As previously shown, α cells were more frequently found at the periphery of these subunits, and β cells were mainly present in the core of these subunits, i.e., at least one cell diameter distant from blood vessels (Bosco et al. 2010). Similarities and differences in topographical distributions were observed for δ and PP cells. Similar to α cells, δ and PP cells were rare in the core of islet subunits and were more often observed at their periphery (Fig. 1A, B, C, D, E, and F). Interestingly, unlike α cells, δ and PP cells were not symmetrically distributed between the intraislet vascular channels and the islet periphery: δ cells were more frequently found adjacent to the intraislet vascular channels (Fig. 1B, D, and F) whereas PP cells were more frequently found in the islet periphery (Fig. 1A, C, and E). No difference in cell organization was observed according to the size of islets (Fig. 1A, B, C, D, E, and F).
Different locations of PP and δ cells in human pancreatic islets. Pancreas sections were double labelled either for PP and insulin (A, C and E) or for somatostatin and insulin (B, D and F). (A, C and E) PP staining was mostly observed at the islet periphery compared with the area around intraislet vascular channels (v.c.). (B, D and F) Somatostatin was mostly observed at the area around (v.c.) compared with the islet periphery. Similar distribution patterns were observed in small (50–100μm diameter, A and B), medium-sized (100–200μm diameter, C and D) and bigger islets (>200μm diameter, E and F). Scale bar: 25μm in A, C and E; 40μm in B, D and F.
Citation: Journal of Endocrinology 229, 2; 10.1530/JOE-15-0542
To confirm differential islet cell distribution between α, δ and PP cells, immunofluorescence staining on digital images of pancreatic islets was quantitatively analysed by computer-assisted morphometry. To minimize the effect of islet size on results, areas stained for glucagon, somatostatin and PP were expressed relative to the area stained for insulin (hereafter referred as stained area ratio). We found that the stained area ratio for glucagon was similar between the islet periphery and the region around intraislet vascular channels and higher in the core subunit region (Fig. 2A). These results confirm our previous work reporting that human islet cells are organized into trilaminar epithelial bands with a symmetrical α cell distribution (Bosco et al. 2010). The somatostatin-stained area ratio was higher in the region around the intraislet vascular channel than in the peripheral or core regions (Fig. 2C). The somatostatin-stained area ratio in the peripheral region was similar to that in the core subunit region (Fig. 2C), which mainly contained βcells, confirming that a limited amount of δ cells was present at the peripheral region. On the contrary to what observed for somatostatin, the stained area ratio for PP was higher in the peripheral region than around the intraislet vascular channel region (Fig. 2B). The PP-stained area ratio in the area around the intraislet vascular channel was slightly higher than that in the core subunit region (Fig. 2B), further confirming that limited amount of PP was present around the intraislet vascular channel region. These results were not biased by the fact that stained areas in each islet regions were expressed relative to the area stained for insulin. Indeed, similar results were obtained when stained areas were expressed relative to the whole areas instead of the insulin-stained areas (Supplementary Fig. 1, see section on supplementary data given at the end of this article). This similarity is further rationalized by the relative weak variations in insulin staining observed between the different islet regions (periphery, core and vascular channel) (Supplementary Fig. 2).
Quantification of islet cell locations in human pancreatic islets. Stained area ratios for glucagon (A), PP (B) and somatostatin (C) are shown according to the islet peripheral (white columns), core (black columns) and vascular channel (vascular ch.; grey columns) regions. (A) Glucagon-stained area ratios are similar between the islet periphery and the area around intraislet vascular channels and are higher (*P=0.004) than in the core of islet subunits; n=5 pancreata for a total of 18 islets. (B) PP-stained area ratio is higher in the islet periphery compared with the core of islet subunits (*P<0.0001) and the area around intraislet vascular channels (**P=0.02); n=12 pancreata for a total of 55 islets. (C) Somatostatin-stained area ratio is higher (*P<0.0001) in the area around the intraislet vascular channels than in the islet periphery and the core of islet subunits; n=14 pancreata for a total of 83 islets.
Citation: Journal of Endocrinology 229, 2; 10.1530/JOE-15-0542
As some variations in islet cell composition have been reported between different regions of the pancreas, islets from the tail and head regions of the pancreas were compared in terms of δ and PP distributions. The pancreatic uncinate processes with PP cell-rich islets were excluded from the analysis. As shown in Fig. 3A, B, and C, distributions of α, PP and δ cells around the intraislet vascular channel and the peripheral regions of islets were similar between the head and the tail of the pancreas.
Similar islet cell locations in islets from the tail and the head of pancreas. Stained area ratios for glucagon (A), PP (B) and somatostatin (C) are shown according to the islet peripheral (white columns) and vascular channel (vascular ch.; grey columns) regions in the head and tail pancreas regions. (A) Glucagon-stained area ratios are similar between the islet periphery and the area around intraislet vascular channels, and this pattern is similar between the islets from the head (n=38 islets from four pancreata) and the tail of the pancreas (n=34 islets from four pancreata). (B) PP-stained area ratio was lower (*P<0.001) in the area around the intraislet vascular channels than the islet periphery, and this pattern is similar between the islets from the head (n=28 islets from four pancreata) and the tail of the pancreas (n=26 islets from four pancreata). (C) Somatostatin-stained area ratio was higher (*p<0.05) in the area around the intraislet vascular channels than the islet periphery, and this pattern is similar between the islets from the head (n=41 islets from four pancreata) and the tail of the pancreas (n=41 islets from four pancreata). When compared with the results in Fig. 2, values here are higher; this is due to a lower staining for insulin (taken into account in the calculation of the stained area ratio) in this series of experiments.
Citation: Journal of Endocrinology 229, 2; 10.1530/JOE-15-0542
δ and PP cell distributions were also analysed in pancreatic islets from type 2 diabetic donors. As control, pancreas sections from non-diabetic donors were additionally and concomitantly processed by immunofluorescence to minimalize the impact of technical variations of staining on the results. As shown in Fig. 4A, B, and C, distributions of α, PP and δ cells around the intraislet vascular channel and the peripheral regions of islets were similar in islets from non-diabetic and type 2 diabetic donors. Taken together, these results show that the islet cell organization is not affected by type 2 diabetes even if a consistent decrease in β cell mass has been demonstrated in these islets (Butler et al. 2003, Yoon et al. 2003).
Similar islet cell locations in islets from non-diabetic and T2D donors. Stained area ratios for glucagon (A), PP (B) and somatostatin (C) are shown according to the islet peripheral (white columns) and vascular channel (vascular ch.; grey columns) regions, in non-diabetes (Control) and type 2 diabetes (T2D) donors. (A) Glucagon-stained area ratios are similar between the islet periphery and the area around intraislet vascular channels, and this pattern is similar between the islets from non-diabetic (control) donors (n = 147 islets from 11 pancreata) and T2D donors (n = 189 islets from five pancreata); these data are obtained from analyses previously made in (Bosco et al. 2010). (B) PP-stained area ratio is lower (*P < 0.001) in the area around the intraislet vascular channels than the islet periphery, and this pattern is similar between the islets from non-diabetic (control) donors (n = 31 islets from six pancreata) and T2D donors (n = 28 islets from five pancreata). (C) Somatostatin-stained area ratio is higher (*P < 0.01) in the area around the intraislet vascular channels than the islet periphery, and this pattern is similar between the islets from non-diabetic (control) donors (n = 83 islets from four pancreata) and T2D donors (n = 59 islets from nine pancreata).
Citation: Journal of Endocrinology 229, 2; 10.1530/JOE-15-0542
When pancreas sections were labelled concomitantly for somatostatin and PP, we observed only few islets displaying both δ and PP cells (Fig. 5B). Of the islets displaying at least one δ or one PP cell, 23% comprised both δ and PP cells, whereas the majority (77%) were positive for δ or PP cells only (Fig. 5A and C). Then, somatostatin and PP staining in islets was quantified by computer-assisted morphometry and expressed as percentage of total islet area. Somatostatin staining was 3.8±0.33% (mean±s.e.m. of 132 islets) and PP staining was 2.7±0.23% (mean±s.e.m. of 132 islets). In each pancreas, islets were arbitrarily classified into three groups according to the somatostatin staining: 0%, 0 to 5% and more than 5%, and the corresponding PP staining calculated (Fig. 5D). In islets with no somatostatin staining, the PP staining was the highest (11±2.8%), and in islets with the highest somatostatin staining, the PP staining was the lowest (0.52±0.32%) (Fig. 5D). Similar results were obtained with T2D islets (not shown). These results indicate that in islets, the amount of δ cells negatively correlates with the amount of PP cells.
Non-concomitant occurrences of PP and δcells in islets. (A, B and C) Islet sections were triple labelled for PP (red), somatostatin (green) and insulin (blue). (A) One islet displaying insulin (blue) and PP staining (red). (B) One islet displaying insulin (blue), PP (red) and somatostatin staining (green). (C) One islet displaying insulin (blue) and somatostatin staining (green). Scale bar, 20μm. (D) PP and somatostatin staining in islets was quantified by morphometry and islets were categorized into three groups according to the somatostatin staining (0, >0–5 and >5% of staining); the corresponding mean values for PP and somatostatin staining was calculated from islets (10–20) in each pancreas. Columns are the mean±s.e.m. of 8–12 different pancreata. In islets with no somatostatin staining (0%), the PP staining was the highest, and in islets with the highest somatostatin staining (>5%), the PP staining was the lowest. *P=0.001 compared with 0% somatostatin staining; **P 0.0001 compared with the two other groups.
Citation: Journal of Endocrinology 229, 2; 10.1530/JOE-15-0542
Discussion
The main finding of this study is the asymmetric distributions of δ and PP cells between two different islet regions: the periphery and the intraislet vascular channels. Considering the trilaminar epithelial band model previously described (Bosco et al. 2010), one can conclude that one side of the epithelial band, bordered by invaginating vessels, is enriched in δ cells and the other side, bordered by peripheral vessels, is enriched in PP cells. The question arising from this observation pertains to the functional significance of this unique spatial organization of δ and PP cells. From a paracrine point of view, the consequence is that intercellular interactions are more limited between δ and PP cells than between the other cell types. The role of δ and PP cell interactions in islet physiology is not well understood, even if it has been shown that both δ and PP cells can inhibit one another (Bloom et al. 1980, Kim et al. 2014). With regard to their effects on insulin and glucagon release, both somatostatin and PP have been shown to have inhibitory effects (Lundquist et al. 1979, Bloom et al. 1980, Aragon et al. 2015). Somatostatin is a short living peptide with a half-life of <1 min in the circulation, suggesting that its action is mediated principally through a local intraislet paracrine route. As both insulin and somatostatin secretions from βand δ cells, respectively, are increased by glucose, the resulting effect of somatostatin on β cells is to moderate the glucose-induced insulin secretion. The effect of somatostatin on α cells could be even more critical considering that the secretion of somatostatin by δ cells at high glucose could be required to inhibit glucagon secretion. In somatostatin knockout mice, the inhibition of glucagon secretion in response to glucose was absent (Hauge-Evans et al. 2009), indicating that δ cell somatostatin plays an important role in the inhibition of glucagon secretion in response to increases in plasma glucose and suggesting that glucose-induced somatostatin release is a major contributor to the inhibitory effects of glucose on islet α cells. Considering the preferential location of δ cells within the central part of the islets, it is tempting to suggest that all α or βcells within an islet are not similarly regulated by somatostatin. Both α and βcells have been shown to display different types of somatostatin receptors (SSTR) with variable affinities for the somatostatin secreted by δ cells (Mandarino et al. 1981, Francis et al. 1990, Kumar et al. 1999). It is therefore conceivable that islet cells could be heterogeneous for SSTR expressions and it would be interesting to investigate whether βor α cells lining intraislet vessels and contacting δ cells express distinct types of SSTR as compared to α or β cells at the islet periphery. PP has a most probable paracrine effect on αcells than on β cells. Indeed, one main PP receptor (PPYR1) was shown to be expressed on α cells and not on β cells. Furthermore, PP was shown to decrease glucagon secretion and to have no effect on insulin secretion (Aragon et al. 2015), suggesting that PP cells can mediate the inhibitory effect of glucose on glucagon secretion. Similarly to what proposed for δcells, we can speculate that α cells close to PP cells, i.e. αcells at the periphery of islets, are more sensitive to PP than the other α cells. A functional significance of the specific δ and PP cell location in human islets can also be hypothesized taking into account the normal endocrine fate of islet hormones through vascularization. Blood vessels lining the islet periphery and those penetrating inside the islet might be from distinct branches and convey blood with different hormone compositions, i.e. enriched with either PP or somatostatin (a model is depicted in Fig. 6). As a consequence, parenchymal cells might be exposed to different hormone cocktails enriched with either somatostatin or PP, depending on the vascular branches. The secretory activity and the amylase content were shown to be different between peri-insular and other acini (Putzke & Said 1975, Aughsteen & Kataoka 1994), suggesting that hormones secreted by islets may affect the function of adjacent acinar cells. Considering the possible role of PP on pancreatic exocrine secretion (Putnam et al. 1989, Morisset 2008) and that hormone delivery to target cells may be mediated via the interstitial compartment, it makes sense that PP cells are mostly located at the periphery of islets.
Model of a double vascularization in islets. Three islets with increasing size (50, 120 and 200μm diameters) are depicted. PP cells are represented in red and δ cells in green. Blood vessels at the islet periphery and within intraislet vascular channels are shown in blue. Small islets (50μm) are without intraislet vascular channel. In this model, blood circulating at the periphery of islet is enriched in PP, whereas blood circulating within intraislet vascular channels is enriched in somatostatin.
Citation: Journal of Endocrinology 229, 2; 10.1530/JOE-15-0542
Distributions of islet cells around the vascular channels and at the islet periphery were similar between the head and the tail of the pancreas. This result does not exclude more subtle and/or quantitative differences between islets from tail and the head of the pancreas. Indeed, analysis were designed here to study differential distributions of cells between the different islet areas, and it was beyond the scope of this study to analyse other qualitative or quantitative differences in islet cell distribution between different pancreas areas.
From a developmental point of view, it would be interesting to study how and when this differential distribution of δand PP cells arises and whether it is related to differential expressions of transcription factors, and more specifically those that drive islet progenitors towards δ or PP cell differentiation. Notably, the expression of the Arx gene has been shown to be crucial in promoting islet progenitors towards the α and PP cell lineages (Collombat et al. 2007), as its inhibition induced by the Pax gene is required to form βand δ cells (Napolitano et al. 2015). It is totally unknown whether the location of δand PP cells in islets is also dictated by differential expressions of these or other genes. Our analyses further show that no major islet architectural default is associated with type 2 diabetes. Thus, our results suggest that the dysfunction of insulin secretion in type 2 diabetes is not originated by a structural defect involving paracrine interactions between different islet cell types. Previous works have shown that anomalies in islet structure, consisting of intermingled α and δ cells with βcells, were detected in mouse models of obesity and type 2 diabetes (Starich et al. 1991, Kim et al. 2009). Similarly, in transgenic mice overexpressing hepatocyte nuclear factor 6, α, δ and PP cells were abnormally mixed with β cells (Gannon et al. 2000). It was suggested that islet cells established inappropriate intercellular interactions, resulting in perturbed islet cell function and contributing to diabetes in these animal models (Starich et al. 1991, Kim et al. 2009) (Gannon et al. 2000). Our results do not confirm these conclusions; however, additional studies should be conducted to understand whether some subtle defects in islet cell organization could be associated with type 2 diabetes in humans. With regard to type 1 diabetes, a clear description of human islet cell organization has not been reported yet; however, it is obvious that the severe destruction of β cells within type 1 diabetic islets results in a perturbed arrangement of the other islet cell types. Precise arrangement of cell-to-cell as well as cell-to-vasculature interactions in type 1 diabetic human islets deserves further detailed investigations.
In summary, these results indicate that δ and PP cells are found mostly in distinct islets and, when present within the same islets, are most often in separate compartments. Thus, these cell types have a propensity to be located away from one another. Further studies are needed to determine whether cell-to-cell or paracrine interactions between δ and PP cells are essential or not for an adequate control of islet hormone release. Finally, these results raise the possibility that the blood hormonal composition in pancreatic microvascularization might differ according to the origin of the islet blood vessels.
Supplementary data
This is linked to the online version of the paper at http://dx.doi.org/10.1530/JOE-15-0542.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.
Funding
This work was supported by a grant (310030_143462/1) from the Swiss National Science Foundation and a grant from the Insuleman Foundation.
Author contribution statement
D B and C B researched and analysed the data and wrote the manuscript. T B contributed to interpretation of the data and revised the manuscript. G P, V L, E B, J M, M A A and E B researched and analysed the data. All authors have approved the final version of the paper.
Acknowledgements
The authors thank D Matthey-Doret, C Sinigaglia, L Perez, C Rouget and N Pernin for their excellent technical assistance.
References
Aragon F, Karaca M, Novials A, Maldonado R, Maechler P & Rubi B 2015 Pancreatic polypeptide regulates glucagon release through PPYR1 receptors expressed in mouse and human alpha-cells. Biochimica et Biophysica Acta 1850 343–351.
Aughsteen AA & Kataoka K 1994 Immunohistochemical localization of amylase in peri- and tele-insular acinar cells of the human exocrine pancreas. Hiroshima Journal of Medical Sciences 43 115–118.
Baetens D, Malaisse-Lagae F, Perrelet A & Orci L 1979 Endocrine pancreas: three-dimensional reconstruction shows two types of islets of langerhans. Science 206 1323–1325.
Ballian N & Brunicardi FC 2007 Islet vasculature as a regulator of endocrine pancreas function. World Journal of Surgery 31 705–714. (doi:10.1007/s00268-006-0719-8)
Batterham RL, Le Roux CW, Cohen MA, Park AJ, Ellis SM, Patterson M, Frost GS, Ghatei MA & Bloom SR 2003 Pancreatic polypeptide reduces appetite and food intake in humans. Journal of Clinical Endocrinology & Metabolism 88 3989–3992.
Bertelli E, Di Gregorio F, Bertelli L, Orazioli D & Bastianini A 1997 The arterial blood supply of the pancreas: a review. IV. The anterior inferior and posterior pancreaticoduodenal aa., and minor sources of blood supply for the head of the pancreas. An anatomical review and radiologic study. Surgical and Radiologic Anatomy 19 203–212.
Bloom SR, Edwards AV & Jarhult J 1980 The effect of somatostatin on pancreatic endocrine responses mediated via the parasympathetic innervation in the conscious calf. Journal of Physiology 308 29–38. (doi:10.1113/jphysiol.1980.sp013459)
Bonner-Weir S & Orci L 1982 New perspectives on the microvasculature of the islets of Langerhans in the rat. Diabetes 31 883–889. (doi:10.2337/diab.31.10.883)
Bonner-Weir S, Sullivan BA & Weir GC 2015 Human islet morphology revisited: human and rodent islets are not so different after all. Journal of Histochemistry & Cytochemistry 63 604–612.
Bosco D, Meda P, Morel P, Matthey-Doret D, Caille D, Toso C, Buhler LH & Berney T 2005 Expression and secretion of alpha1-proteinase inhibitor are regulated by proinflammatory cytokines in human pancreatic islet cells. Diabetologia 48 1523–1533. (doi:10.1007/s00125-005-1816-1)
Bosco D, Armanet M, Morel P, Niclauss N, Sgroi A, Muller YD, Giovannoni L, Parnaud G & Berney T 2010 Unique arrangement of alpha- and beta-cells in human islets of Langerhans. Diabetes 59 1202–1210. (doi:10.2337/db09-1177)
Brunicardi FC, Stagner J, Bonner-Weir S, Wayland H, Kleinman R, Livingston E, Guth P, Menger M, McCuskey R, Intaglietta M et al., . 1996 Microcirculation of the islets of Langerhans. Long Beach Veterans Administration Regional Medical Education Center Symposium. Diabetes 45 385–392.
Butler AE, Janson J, Bonner-Weir S, Ritzel R, Rizza RA & Butler PC 2003 Beta-cell deficit and increased beta-cell apoptosis in humans with type 2 diabetes. Diabetes 52 102–110. (doi:10.2337/diabetes.52.1.102)
Collombat P, Hecksher-Sorensen J, Krull J, Berger J, Riedel D, Herrera PL, Serup P & Mansouri A 2007 Embryonic endocrine pancreas and mature beta cells acquire alpha and PP cell phenotypes upon Arx misexpression. Journal of Clinical Investigation 117 961–970. (doi:10.1172/JCI29115)
Francis BH, Baskin DG, Saunders DR & Ensinck JW 1990 Distribution of somatostatin-14 and somatostatin-28 gastrointestinal-pancreatic cells of rats and humans. Gastroenterology 99 1283–1291.
Gannon M, Ray MK, Van Zee K, Rausa F, Costa RH & Wright CV 2000 Persistent expression of HNF6 in islet endocrine cells causes disrupted islet architecture and loss of beta cell function. Development 127 2883–2895.
Hauge-Evans AC, King AJ, Carmignac D, Richardson CC, Robinson IC, Low MJ, Christie MR, Persaud SJ & Jones PM 2009 Somatostatin secreted by islet delta-cells fulfills multiple roles as a paracrine regulator of islet function. Diabetes 58 403–411. (doi:10.2337/db08-0792)
Kim A, Miller K, Jo J, Kilimnik G, Wojcik P & Hara M 2009 Islet architecture: a comparative study. Islets 1 129–136. (doi:10.4161/isl.1.2.9480)
Kim W, Fiori JL, Shin YK, Okun E, Kim JS, Rapp PR & Egan JM 2014 Pancreatic polypeptide inhibits somatostatin secretion. FEBS Letters 588 3233–3239. (doi:10.1016/j.febslet.2014.07.005)
Koh DS, Cho JH & Chen L 2012 Paracrine interactions within islets of Langerhans. Journal of Molecular Neuroscience 48 429–440. (doi:10.1007/s12031-012-9752-2)
Kumar U, Sasi R, Suresh S, Patel A, Thangaraju M, Metrakos P, Patel SC & Patel YC 1999 Subtype-selective expression of the five somatostatin receptors (hSSTR1-5) in human pancreatic islet cells: a quantitative double-label immunohistochemical analysis. Diabetes 48 77–85. (doi:10.2337/diabetes.48.1.77)
Lundquist I, Sundler F, Ahren B, Alumets J & Hakanson R 1979 Somatostatin, pancreatic polypeptide, substance P, and neurotensin: cellular distribution and effects on stimulated insulin secretion in the mouse. Endocrinology 104 832–838. (doi:10.1210/endo-104-3-832)
Mandarino L, Stenner D, Blanchard W, Nissen S, Gerich J, Ling N, Brazeau P, Bohlen P, Esch F & Guillemin R 1981 Selective effects of somatostatin-14, -25 and -28 on in vitro insulin and glucagon secretion. Nature 291 76–77. (doi:10.1038/291076a0)
Marks V, Tan KS, Stagner JI & Samols E 1990 Intra-islet cellular interrelationships. Biochemical Society Transactions 18 103–104. (doi:10.1042/bsbib180103)
Morisset J 2008 Negative control of human pancreatic secretion: physiological mechanisms and factors. Pancreas 37 1–12. (doi:10.1097/MPA.0b013e318161b99a)
Napolitano T, Avolio F, Courtney M, Vieira A, Druelle N, Ben-Othman N, Hadzic B, Navarro S & Collombat P 2015 Pax4 acts as a key player in pancreas development and plasticity. Seminars in Cell & Developmental Biology 44 107–114.
Orci L, Baetens D, Ravazzola M, Stefan Y & Malaisse-Lagae F 1976 Pancreatic polypeptide and glucagon: non-random distribution in pancreatic islets. Life Sciences 19 1811–1815. (doi:10.1016/0024-3205(76)90112-0)
Putnam WS, Liddle RA & Williams JA 1989 Inhibitory regulation of rat exocrine pancreas by peptide YY and pancreatic polypeptide. American Journal of Physiology 256 G698–G703.
Putzke HP & Said F 1975 Different secretory responses of periinsular and other acini in the rat pancreas after pilocarpine injection. Cell and Tissue Research 161 133–143.
Rodriguez-Diaz R, Dando R, Jacques-Silva MC, Fachado A, Molina J, Abdulreda MH, Ricordi C, Roper SD, Berggren PO & Caicedo A 2011 Alpha cells secrete acetylcholine as a non-neuronal paracrine signal priming beta cell function in humans. Nature Medicine 17 888–892. (doi:10.1038/nm.2371)
Schuit FC, Derde MP & Pipeleers DG 1989 Sensitivity of rat pancreatic A and B cells to somatostatin. Diabetologia 32 207–212. (doi:10.1007/BF00265096)
Seymour NE, Volpert AR & Andersen DK 1996 Regulation of hepatic insulin receptors by pancreatic polypeptide in fasting and feeding. Journal of Surgical Research 65 1–4. (doi:10.1006/jsre.1996.9999)
Starich GH, Zafirova M, Jablenska R, Petkov P & Lardinois CK 1991 A morphological and immunohistochemical investigation of endocrine pancreata from obese ob /ob mice. Acta Histochemica 90 93–101. (doi:10.1016/S0065-1281(11)80167-4)
Stefan Y, Orci L, Malaisse-Lagae F, Perrelet A, Patel Y & Unger RH 1982 Quantitation of endocrine cell content in the pancreas of nondiabetic and diabetic humans. Diabetes 31 694–700. (doi:10.2337/diab.31.8.694)
Unger RH & Orci L 2010 Paracrinology of islets and the paracrinopathy of diabetes. PNAS 107 16009–16012. (doi:10.1073/pnas.1006639107)
Wang X, Zielinski MC, Misawa R, Wen P, Wang TY, Wang CZ, Witkowski P & Hara M 2013 Quantitative analysis of pancreatic polypeptide cell distribution in the human pancreas. PLoS ONE 8 e55501.
Yoon KH, Ko SH, Cho JH, Lee JM, Ahn YB, Song KH, Yoo SJ, Kang MI, Cha BY, Lee KW et al., . 2003 Selective beta-cell loss and alpha-cell expansion in patients with type 2 diabetes mellitus in Korea. Journal of Clinical Endocrinology & Metabolism 88 2300–2308.
