Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Animal Models

3-D imaging of islets in obesity: formation of the islet–duct complex and neurovascular remodeling in young hyperphagic mice



Obesity and insulin resistance lead to islet hyperplasia. However, how the islet remodeling influences the pancreatic environment and the associated neurovascular networks is largely unknown. The lack of information is primarily due to the difficulty of global visualization of the hyperplasic islet (>200 μm) and the neurovascular environment with high definition.


We modulated the pancreatic optical property to achieve 3-dimensional (3-D) whole-islet histology and to integrate transmitted light microscopy (which provides the ground-truth tissue information) with confocal fluorescence imaging. The new optical and imaging conditions were used to globally examine the hyperplastic islets of the young (2 months) obese db/db and ob/ob mice, which otherwise cannot be easily portrayed by the standard microtome-based histology. The voxel-based islet micrographs were digitally processed for stereo projection and qualitative and quantitative analyses of the islet tissue networks.


Paired staining and imaging of the pancreatic islets, ducts and neurovascular networks reveal the unexpected formation of the ‘neuro-insular-ductal complex’ in the young obese mice. The complex consists of the peri- and/or intra-islet ducts and prominent peri-ductal sympathetic nerves; the latter contributes to a marked increase in islet sympathetic innervation. In vascular characterization, we identify a decreased perivascular density of the ob/ob islet pericytes, which adapt to ensheathing the dilated microvessels with hypertrophic processes.


Modulation of pancreatic optical property enables 3-D panoramic examination of islets in the young hyperphagic mice to reveal the formation of the islet–duct complex and neurovascular remodeling. On the basis of the morphological proximity of the remodeled tissue networks, we propose a reactive islet microenvironment consisting of the endocrine cells, ductal epithelium and neurovascular tissues in response to the metabolic challenge that is experienced early in life.


Obesity is the primary risk factor for type 2 diabetes (T2D), an increasing health-care problem among adolescents and adults worldwide.1, 2 In the development of obesity and insulin resistance, the increase in metabolic loading and insulin demand drives β-cell expansion, leading to islet hyperplasia. Although the hyperplastic phenomenon has been demonstrated in both humans and animals,3, 4 how the islet remodeling influences the pancreatic environment and the associated neurovascular networks is incompletely understood.

In the pancreas, the endocrine islets are closely associated with the neurovascular networks, which consist of the autonomic nerves and microvessels: the former includes the sympathetic, parasympathetic and sensory nerves;5, 6, 7 the latter are formed by the endothelium and the perivascular pericytes.8, 9, 10 Physiologically, the neurovascular association allows the endocrine islets to receive circulatory and nervous signals to modulate hormone secretion, including the fasting glucagon secretion and the cephalic and postprandial insulin secretion.6, 11

In diabetes, neuropathies and microvascular complications have been well recognized.12, 13, 14 However, from the onset of obesity to insulin resistance to T2D, the early events of the islet neurovascular remodeling remain unclear. Although in diabetes the nerves and blood vessels may simply be damaged by the hyperglycemia-induced oxidative stress,15, 16 lines of evidence have emerged to indicate that, before their damages, the neurovascular tissues carry intrinsic plasticity in response to the change(s) in the islet microenvironment. First, in the rodent islet injury induced by streptozotocin (STZ) injection or lymphocytic invasion, both the islet neural tissues (sympathetic nerves and Schwann cells) and perivascular pericytes become reactive following the islet microstructural and vascular damages.17, 18, 19 Second, in the mouse islet transplantation, the donor islet Schwann cells and pericytes participate in graft neurovascular regeneration.20 Third, in the obese ob/ob mice, the islet microvessels dilate in response to insulin resistance.21 Overall, the reactivity (or adaptability) of the neurovascular tissues in response to the changes in the islet microenvironment suggests that they are not passive targets damaged in the progression of obesity and diabetes; rather, they participate in islet remodeling.

In addition to the intimate islet–neurovascular association, the islet endocrine cells in the pancreas also neighbor the exocrine acinar and ductal cells, with which the islet cells share the same progenitors in the ductal epithelium in development.22 Although it is generally agreed that the β-cell expansion in obesity and insulin resistance is a proliferative response of the existing β-cells,23 whether the ductal cells participate in and/or influence the expansion process requires further characterization. Particularly, because the pancreatic ductal replication also increases in obesity,24 it is important to investigate the islet, duct and the pancreatic neurovascular networks in an integrated manner to understand the tissue remodeling in islet hyperplasia.

Unfortunately, because of the depth limitation in microscopy, the large, hyperplastic islets (with a size of >200 μm) cannot be easily portrayed by the standard microtome-based 2-dimensional (2-D) histology (3–5 μm in tissue thickness). In this research, we modulate the pancreatic optical condition by tissue clearing18, 25, 26, 27, 28 to perform 3-dimensional (3-D) confocal microscopy for global visualization of the islet, duct and neurovascular networks with high definition. To demonstrate the associated pancreatic tissue remodeling in obesity, we targeted the young (2 months) hyperphagic db/db and ob/ob mice for characterization. Note that in human obesity, the risk factors have a genetic contribution.29, 30 Although the contribution is polygenic, monogenic animal models, such as the db/db and ob/ob mice with mutation in the leptin signaling pathway, help us understand the link between satiety and obesity.31, 32 The link is particularly important in childhood and juvenile obesity, in which the genetic predisposition contributes to low satiety responsiveness,33 leading to excessive calorie consumption in an environment rich with food. To date, because there is no suitable polygenic animal model for studying childhood and juvenile obesity, our 3-D examination of the islets in the young hyperphagic mice provides the first insight into the hyperplastic islet microenvironment for qualitative and quantitative analyses of the islet structural and neurovascular remodeling in the obesity that develops early in life.

Materials and methods


Pancreata harvested from db/db (BKS.Cg-Dock7m +/+ Leprdb/J; JAX Mice stock number: 000642), db/+ (lean littermates), ob/ob (B6.V-Lepob/J; JAX Mice stock number: 000632) and wild-type C57BL/6 (B6) mice (2 months; National Laboratory Animal Center, Taipei, Taiwan) were used to acquire the images of the islets in the obese (average 70% more on the body weight) and the lean control animals. The obese db/db (serum insulin level: 1691±952 pmol l1; n=5) and ob/ob (serum insulin level: 2435±785 pmol l1; n=4) mice were also hyperglycemic, confirmed with 6-h fasting glucose concentration >350 mg dl1 (measured from the tail tip with an Accu-Check Performa glucometer; Roche, Indianapolis, IN, USA). In the low-dose STZ (Sigma, St Louis, MO, USA) -induced islet injury, STZ at 90 μg g1 body weight was injected into the B6 mice to induce β-cell damage and the subsequent replication while maintaining normoglycemia (non-fasting glucose levels <200 mg dl1). Overall, 192, 60, 91, 75 and 81 image stacks acquired from seven db/db mice, four db/+ lean littermates, five ob/ob mice, five B6 mice and five low-dose STZ-injected B6 mice, respectively, were used for qualitative and quantitative analyses. The National Tsing Hua University Institutional Animal Care and Use Committee approved all procedures with mice.

Islet transplantation

The db/db mice and their db/+ littermates (age 2 months) were used as the donors and the recipients, respectively, in islet transplantation to avoid rejection. The recipient db/+ mice were diabetic (non-fasting glucose levels >350 mg dl1), induced by a single high-dose STZ injection (200 μg g1 body weight) 2 weeks before islet transplantation. Islet isolation from the db/db donor mice was performed under sodium amobarbital-induced anesthesia with the pancreases distended with 2.5 ml of digestion solution (ductal injection of RPMI-1640 medium supplemented with 1.5 mg ml1 of collagenase; RPMI: Invitrogen, Carlsbad, CA, USA; collagenase: Sigma, from Clostridium histolyticum, type XI), excised and incubated in a water bath at 37 °C. Afterward, the islets were purified by a density gradient (Histopaque-1077; Sigma) and then handpicked under a stereo microscope. Three hundred islets were transplanted under the left kidney capsule on the same day of isolation. Reversal of diabetes of the db/+ recipients was confirmed 2 weeks post transplantation. The recipient mice were killed 3 weeks post transplantation to examine the engrafted islets.

Preparation of pancreatic specimens

Blood vessels of the pancreas were labeled by vessel painting34, 35 via cardiac perfusion of the lectin-Alexa Fluor 488 conjugate (30 μg g1 of body weight; Invitrogen) followed by 4% paraformaldehyde perfusion fixation. Afterward, pancreata were harvested and post-fixed in 4% paraformaldehyde solution for 40 min at 15 °C. The vibratome sections of the fixed tissue (~400 μm) were then immersed in 2% Triton X-100 solution for 2 h at 15 °C for permeabilization.

Five different primary antibodies were used to immunolabel the tissues following the protocol outlined below. The antibodies used were guinea pig anti-insulin antibody (GTX27842; GeneTex, Irvine, CA, USA), rabbit anti-tyrosine hydroxylase (TH, sympathetic marker) antibody (AB152; Millipore, Billerica, MA, USA), rabbit anti-cytokeratin 19 (CK19, duct epithelial marker) antibody (ab133496; Abcam, Cambridge, MA, USA), rabbit anti-neuron-glial antigen 2 (NG2, pericyte marker) antibody (AB5320; Millipore) and rabbit anti-ki67 (nuclear protein associated with cellular proliferation) antibody (ab15580; Abcam). Before applying the antibody, the tissue was rinsed in phosphate-buffered saline. This was followed by a blocking step, incubating the tissue with the blocking buffer (2% Triton X-100, 10% normal goat serum and 0.02% sodium azide in phosphate-buffered saline). The primary antibody was then diluted in the dilution buffer (1:100, 0.25% Triton X-100, 1% normal goat serum and 0.02% sodium azide in phosphate-buffered saline) to replace the blocking buffer and incubated for 1 day at 15 °C.

Alexa Fluor 647-conjugated goat anti-rabbit secondary antibody and Alexa Fluor 546-conjugated goat anti-guinea pig secondary antibody (1:200; Invitrogen) were used to reveal the immunostained structures. Nuclear staining by propidium iodide (50 μg ml1; Invitrogen) was performed at room temperature for 1 h to reveal the nuclei. The labeled specimens were then immersed in the tissue-clearing solution (FocusClear solution, RI: 1.46, CelExplorer, Hsinchu, Taiwan or RapiClear 1.52 solution, RI: 1.52, SunJin Lab, Hsinchu, Taiwan) before being imaged via confocal microscopy.20, 36 To quantify the tissue transparency, the percentage of light transmittance was measured by the microplate reader (SpectraMax M2e; Molecular Devices, Sunnyvale, CA, USA) with the unlabeled specimens immersed in saline or the clearing solutions.

Deep-tissue confocal microscopy

Imaging of the tissue structure was performed with a Zeiss LSM 510 Meta confocal microscope (Carl Zeiss, Jena, Germany) equipped with an objective of × 25 LD ‘Plan-Apochromat’ glycerine immersion lenses (working distance: 570 μm) (optical section: 5 μm; Z-axis increment: 2.5 μm) and an objective of × 40 LD ‘C-Apochromat’ water immersion lenses (working distance: 620 μm) (optical section: 3 μm; Z-axis increment: 1.5 μm) under a regular or tile-scan mode with automatic image stitching. The laser-scanning process was operated under the multi-track scanning mode to sequentially acquire signals, including the transmitted light signals. The Alexa Fluor 647-labeled structures were excited at 633 nm, and the fluorescence was collected by the 650- to 710-nm band-pass filter. The propidium iodide-labeled nuclei and the Alexa Fluor 546-labeled structures were excited at 543 nm, and the signals were collected by the 560- to 615-nm band-pass filter. The lectin-Alexa Fluor 488-labeled blood vessels were excited at 488 nm, and the fluorescence was collected by the 500- to 550-nm band-pass filter.

Image projection and analysis

The LSM 510 software (Carl Zeiss) and the Avizo 6.2 image reconstruction software (VSG, Burlington, MA, USA) were used for projection, signal segmentation and analysis of the confocal images. In the Supplementary Movies, image stacks were recorded using the ‘Movie Maker’ function of Avizo with the increase in display time in association with the depth of the optical section. In Supplementary Movie S1, the ‘Camera Rotation’ function of Avizo was used to adjust the projection angles of the 3-D images.

In the neurovascular density analysis, the islet images were acquired from four obese and lean control mice. Four or five image stacks were taken from each animal to assess their sympathetic nerve, blood vessel and pericyte signals. Quantitation of the neurovascular tissue density was illustrated in Juang et al.37 The same tissue labeling, imaging and quantitation processes were conducted on the obese and lean control specimens to compare the tissue densities on the same basis. In estimation of the density, feature extraction and image segmentation were first performed by the ‘Label Field’ function of Avizo to collect the voxels of the islet (including the peri-islet ducts attached to the islet mantle) and those of the sympathetic nerves (TH signals, excluding the globular TH+ endocrine cells) or vasculature (NG2 signals of pericytes and microvessel signals from vessel painting). The number of pericytes in the segmented volume of interest is counted using the NG2 and nuclear signals; a pericyte is defined by its NG2 immunoreactive cell body with at least two processes contacting the blood vessels.19, 20 Afterward, voxels of the nerves (or NG2) in the acquired image stack were divided by those of the islet (or the number of pericytes) × 100% to estimate the sympathetic nerve density (or the level of pericyte surface marker expression). The number of pericytes was divided by the voxels of blood vessels × 100% to estimate the perivascular pericyte density. Signal or pericyte densities derived from different image stacks of the same animal were first normalized and then averaged over the other animals in the same group.

Statistical analysis

The quantitative values are presented as means±s.d. Statistical differences were determined by the unpaired Student’s t-test. Differences between groups were considered as statistically significant when P<0.05.


Modulation of pancreatic optical property for 3-D islet examination

Unlike the transparent crystalline lens of the eye, the mouse pancreatic specimens strongly scatter light. Thus, the specimens require tissue clearing (or optical clearing—the use of an immersion solution with a high refractive index (RI) to improve light transmission)38, 39, 40 to allow for deep-tissue microscopy. Figures 1a–f show an embedded db/db islet with a diameter of ~300 μm becoming transparent in the clearing solution (RI: 1.52, close to the refractive indices of phospholipids and cholesterol) for 3-D confocal imaging. Importantly, whole-islet imaging was achieved using the transparent islet as shown in Supplementary Movie S1. The anatomic information of the hyperplastic islet microstructure, vasculature and sympathetic innervation was continuously presented with the resolving power to distinguish the islet blood vessels, nerves and the adjacent nuclei with high definition.

Figure 1

3-D examination of hyperplastic islets via modulation of pancreatic optical condition. (a–f) Whole-islet histology. An embedded db/db islet (a) becomes transparent in the clearing solution with RI at 1.52 (b) for deep-tissue microscopy (cf and Supplementary Movie S1). The islet is labeled with blood vessels (red), sympathetic nerves (white in c, d and f) and nuclei (green in c). Note: tissue clearing in the RI: 1.52 solution allows the autofluorescence from the exocrine acini to be detected (green in c, along with the nuclear signals), serving as a counterstain to reveal the pancreatic parenchyma and the acinar-islet boundary. Projection of the autofluorescence and the nuclear signals is shown in (a) and (d) (blue). (g-i) Matching the transmitted light and fluorescence signals in islet characterization. A hyperplasic ob/ob islet is visualized with improved contrast in the RI: 1.46 solution via transmitted light microscopy (g). The transmitted light information matches the fluorescence signals derived from confocal microscopy (h: 2-D image; i: 3-D projection). (j) Transmitted light image of the opaque pancreatic specimen in saline (RI: 1.33). (k) Comparison of light transmission among the pancreatic specimens immersed in saline and the clearing solutions. Three wavelengths used in confocal imaging were tested. Tissue thickness: 400 μm. n=6. Data are presented as means±.s.d. (l) Signal decay in deep-tissue microscopy of the control and the optically cleared pancreatic specimens. Tissues were labeled with lipophilic dye DiD and imaged with 633-nm laser excitation via confocal microscopy. n=3. Data are presented as means±s.d.

Application of both transmitted light and fluorescence microscopy in islet characterization

Although the transparent specimens allow global islet visualization, the increased light transmission decreases the contrast of the transmitted light image (Figure 1b). In histology, transmitted light microscopy is essential for providing the ground-truth information to verify the size and locations of the fluorescently labeled tissue structures. Thus, to improve the contrast, we next used the clearing reagent with a lower RI at 1.46 to modulate the tissue transparency, thereby allowing both the transmitted light and confocal fluorescence images to be simultaneously taken to demonstrate the matched information (Figures 1g–i). However, due to the decreased light transmission, the image depth was limited at ~100 μm (Supplementary Movie S2). In this research, we use both clearing conditions (Figures 1b and g) for identification and characterization of the islet structure and tissue networks.

Comparing with the optically cleared tissues, Figure 1j shows the opaque pancreatic specimen in saline (RI at 1.33). The opaque specimen allows limited light transmission (Figure 1k), causing a drastic decline in fluorescence signals against the focal depth in deep-tissue imaging (Figure 1l). In comparison, the increased light transmission of the optically cleared pancreas enables a slower fluorescence decline under the same imaging conditions. Note that in practice, the gradual decline in the fluorescence signals can be automatically compensated by adjusting the detector gain in confocal microscopy.

Pancreatic islet–duct association and the intra-islet ducts in db/db islets

To visualize the islet with the surrounding tissue networks, we next extended the imaging field to provide a complete view of the islet microenvironment in the wild-type and obese mice. Figures 2a–c illustrate the approach of using 3-D microscopy with tile scanning to panoramically reveal the pancreatic ductal network and its close association with the islets. Particularly, in the ob/ob and db/db mice, we routinely observed a prominent presence of the large-diameter (interlobular and/or hypertrophic) ducts around the hyperplastic islets, making the islet–duct complexes the landmarks in the pancreas (Figures 2b and c). When zooming into the islet–duct complex, surprisingly we identified the intra-islet ducts in the db/db islets, with the ductal epithelium extending from the peri-islet domain into the core (Figures 2d–i and Supplementary Movie S3).

Figure 2

Pancreatic islet–duct association and intra-islet ducts. (a–c) Tile scanning of islet–duct association in B6, ob/ob and db/db pancreas. White: insulin staining. Red: blood vessels. Green: ductal epithelial marker CK19. Blue: pancreatic parenchyma. Dotted line: stitching of tile-scanned micrographs. Projection depth: 280 μm in (a); 320 μm in (b); 250 μm in (c). (d–i) Two examples of the peri- and intra-islet ducts in db/db pancreas. Both the large, hyperplastic (d, e) and the medium (f–i) islets are identified with the ingrowth of ductal epithelium (confirmed by the CK19 and nuclear signals). Green: CK19. Blue/cyan: nuclei.

Peri-ductal sympathetic innervation and formation of the ‘neuro-insular-ductal complex’

In the normal mouse pancreas, the sympathetic nerves follow the blood vessels into the islet pole and subsequently enter the islet mantle and core, forming the peri-islet and perivascular sympathetic innervation.28 In the db/db islets, the two innervation patterns remain prominently seen. However, when we zoomed into the db/db islet-duct boundary, a third pattern of the peri-ductal sympathetic innervation was identified (Figures 3a–c and Supplementary Movie S4). In-depth imaging and projection of the db/db islets show the unique ‘neuro-insular-ductal complex,’ featuring the abundant sympathetic axons and varicosities concentrating at the islet-duct boundary (Figures 3d–f and Supplementary Movie S5). Morphologically, the complex leads to heterogeneity in islet sympathetic innervation, recruiting more nerves to the islet domain around the duct(s) than the distal area (Figures 3g and h). In comparison, Figures 3i and j and the last part of Supplementary Movie S4 show the sympathetic innervation of the control islet in the db/+ lean littermate. Overall, the neuroanatomy indicates the plasticity of the pancreatic innervation in response to the changes in the islet microstructure (the ingrowth of ductal epithelium) and the immediate environment (the peri-islet ductal network) in obesity. A gallery display of the db/db islet sympathetic innervation is presented in Supplementary Figure S1.

Figure 3

Peri-ductal and the elevated islet sympathetic innervation in obesity. (a–c) Peri-ductal sympathetic innervation of the islet–duct complex in db/db pancreas. The sympathetic nerves are labeled by TH staining. TH is a critical enzyme in the production of the sympathetic neurotransmitter norepinephrine. The staining also reveals the TH+ endocrine cells, which carry the enzyme but do not belong to the sympathetic nervous system (their role in islets is unclear). White: TH staining. Cyan: nuclei. Red: blood vessels. Projection depth in (c): 45 μm. (d–f) The neuro-insular-ductal complex. Abundant sympathetic axons and varicosities concentrate at the islet–duct boundary, forming the unique complex in the pancreas. Blue: insulin. White: TH. Red: blood vessels. (g, h) A second example of the neuro-insular-ductal complex formed with a large, hyperplastic db/db islet. The yellow line indicates that the islet domain facing the ducts (right side of the line) receives more sympathetic nerves than the distal area (left side). (i, j) Sympathetic innervation of the control islet in the db/+ lean littermate. The islet and a nearby duct are presented in the last part of Supplementary Movie S4. (k–m) NGF signals in the obese db/db and lean db/+ pancreas. (k) A panoramic view of the db/db pancreas with two hyperplastic islets. Red: blood vessels. White: nuclear signals and autofluorescence. (l) Prominent NGF signals with the hyperplastic islets and ducts in comparison with the signals in (m) are shown. (n) Increase in the sympathetic nerve density of islets in the obese db/db and ob/ob mice. Lean littermate: islets in the db/+ lean littermates. B6 control: islets in the wild-type B6 mice. The difference in the islet sympathetic nerve density (TH+ signals of axons and varicosities divided by tissue volume) is statistically significant between the obese mice and the lean controls (*P<0.05). n=4. Data are presented as means±s.d.

The apparent increase in the islet sympathetic innervation can be explained by the elevated immunostaining signals of the nerve growth factor (NGF) with the islet–duct complex in the obese mice (Figures 3k–m). The elevation of NGF expression in the ductal epithelium has been previously seen in the experimental pancreatitis.41 Here, we identify that both the hyperplastic islets and ductal epithelium are stained with an elevated NGF expression to recruit the sympathetic nerves. In quantitation, Figure 3n shows a 49% increase in the sympathetic nerve density of the db/db islets against that of the db/+ islets in the lean littermates; the increase is 61% in the ob/ob islets against that of the wild-type B6 control islets. No statistical difference was found between the db/db and ob/ob islets and between db/+ and B6 islets.

Formation of islet–duct complex is associated with islet cell proliferation

Because the donor ductal cells are positively linked with long-term metabolic success in human islet transplantation,42 we suspect that the formation of the islet–duct complex provides a favorable environment for islet cell proliferation. This condition could occur both in situ (that is, in the pancreas of young obese mice; Figures 2 and 3) and after transplantation. To test this concept, we performed immunostaining of the ki67 nuclear protein to detect the proliferating cells in the islet–duct complex in the db/db mouse pancreas (Figure 4a) and after transplantation under the kidney capsule of the db/+ mice (Figures 4b and c). In both conditions, the ki67+ cells were found in the islet endocrine domain close to the ductal epithelium as well as in the ductal epithelium, suggesting the mutual influence of the islet and ductal cells on their proliferation while forming the complex.

Figure 4

Detection of proliferating cells in the db/db islet–duct complex in situ and after transplantation. (a) Immunostaining of ki67 nuclear protein in the db/db mouse pancreas. Green: nuclear signals and exocrine. Yellow: overlap of ki67 (red) and nuclear signals (green). Blue: blood vessels. Arrows indicate the peri- and intra-islet ducts. (b, c) Ki67 staining of db/db islet–duct complex after transplantation under the kidney capsule of db/+ littermate. Inset: isolated islet attached with ductal epithelium (cyan, CK19 staining). Three weeks after transplantation, the ki67+ cells were detected in the graft domain between the kidney capsule and parenchyma. Both the transplanted islet and ductal epithelium were found in proliferation. (b) Overlay of transmitted light and fluorescence image. (c) Projection of fluorescent signals. Images were derived from stitching of the tile-scanned image stacks (dotted lines represent the stack boundaries).

Furthermore, we injected the low-dose STZ (90 μg g1 body weight) into the wild-type B6 mice to induce islet injury and test the influence of the β-cell replication on the ingrowth of ductal epithelium (note: at this dose, STZ is able to induce β-cell damage and the subsequent regeneration but insufficient to cause hyperglycemia). Two weeks after the STZ injection, Figures 5a–c show the formation of the intra-islet duct, which is similar to that presented in the young obese db/db mice (Figures 2f–i). Importantly, this result indicates: (1) the formation of the islet–duct complex is associated with β-cell replication and (2) hyperglycemia is not essential to the formation of the complex. In addition, in this islet injury/regeneration model, Figures 5d and e show the prominent sympathetic nerve fibers and varicosities associated with the islet–duct complex, indicating that the sympathetic innervation is coupled with the β-cell replication.

Figure 5

Low-dose STZ injection induces the formation of islet–duct complex and prominent sympathetic innervation. (a–c) Ingrowth of ductal epithelium to the injured islet. (a, b) Projections and optical section of islet–duct complex. (c) Zoom-in identification of the intra-islet duct via the nuclear signals. Green: CK19 staining. Red: blood vessels. Blue/cyan: nuclear signals and autofluorescence. Arrows indicate the empty areas in the injured islet core without the fluorescence signals. Similar empty space can be found in (d). (d, e) Prominent sympathetic innervation of the islet–duct complex in islet regeneration. Both panels (ac) and (d, e) were taken 2 weeks after the STZ-induced islet injury. (f) Frequency of islets to develop the islet–duct complex in the mouse lines and conditions analyzed in this research. Overall, 81, 155, 91, 40 and 62 islets were examined in five STZ-injected B6, seven db/db, five ob/ob, four db/+ and five B6 mice, respectively. In particular, the frequencies of the islets to develop the intra-islet duct are 30±9% and 83±13% in the STZ-injected B6 and the db/db mice, respectively.

Finally, Figure 5f summarizes the frequency of the islets to develop the islet–duct complex in the mouse lines and conditions analyzed in this research. The high frequency (>70%) of the complex developed in the models of the STZ-induced islet injury/regeneration and of islet hyperplasia (db/db and ob/ob mice) highlights the use of the epithelial ingrowth as the morphological feature to reveal the change in the pancreatic microenvironment.

Microvessel dilation and pericyte remodeling in ob/ob islets

To characterize the islet vascular remodeling in obesity, we combined vessel painting (fluorescent lectin perfusion)35 and NG2 (pericyte surface marker) immunostaining to simultaneously reveal the islet endothelium and the perivascular pericytes for 3-D examination. In the wild-type B6 islets, the pericytes reside on the walls of microvessels with a prominent cell body and processes extending toward the longitudinal direction of the capillary (Figures 6a–f). However, in the ob/ob islets, the microvessels and pericytes undergo associated remodeling: the microvessels dilate and the perivascular pericytes become hypertrophic with circular expansion of the processes to enclose the dilated vessel walls (Figures 6g–l). This associated vascular remodeling is most prominent in the islet core and is found in both the medium (100–200 μm) and large (>200 μm; Figures 6m–o) ob/ob islets.

Figure 6

Associated remodeling of ob/ob islet microvessels and perivascular pericytes. (a–f) Pancreatic islet pericytes in wild-type B6 mice. The projections show that the NG2+ pericytes (green) reside on the abluminal domain of blood vessels with prominent cell bodies and processes embracing the vessel wall. Note that the microvessels of the islet and acini are comparable in size. (g–l) Microvessel dilation and pericyte hypertrophy in obese ob/ob islet. The projections show a medium ob/ob islet with microvessel dilation in the core. The dilation is associated with perivascular pericyte remodeling, featuring circularly extended and expanded cellular processes (hypertrophy) with an apparent increase in NG2 staining signals in space. (m–o) Microvessels sheathed by pericytes in the obese ob/ob islet core. (m, n) Gross view of a large, hyperplastic ob/ob islet. (o) Zoom-in examination of the perivascular pericyte sheath. Asterisk denotes the luminal space of the enclosed microvessel (an example). Green: pericyte surface marker NG2. Red: blood vessels. Blue: pancreatic parenchyma. (p) Decrease in perivascular pericyte density in the ob/ob islet. The difference in the perivascular pericyte density (number of NG2+ pericytes/microvessel volume) was statistically significant between the B6 and ob/ob islets (*P<0.05), but not between the db/+ (lean littermate) and db/db islets. (q) Increase in pericyte staining signals in the ob/ob islet. The difference in the NG2 staining signals per pericyte (total NG2+ voxels/total number of pericytes) was statistically significant between the B6 and ob/ob islets (*P<0.05), but not between the db/+ (lean littermate) and db/db islets. n= 4. Data are presented as means±s.d. Note: the increased NG2 signals of the hypertrophic processes in the ob/ob islet could be confused with an increase in the pericyte density around the microvessels (the opposite shown in p). However, when we quantified the pericyte number against the microvessel volume, a decrease in perivascular pericyte density was shown in the ob/ob islet in comparison with that of the B6 islet.

Quantitative analysis of the ob/ob islet pericytes indicates a 39% decrease in the pericyte density on the microvessels, accompanied with a marked 96% increase in the NG2 staining signals per pericyte relative to those of the B6 control islets (Figures 6p and q). Interestingly, although similar islet hyperplasia occurs in the young ob/ob and db/db islets, the microvessel dilation and pericyte remodeling of the db/db islets are less significant. Because leptin has been known to influence the physiology of blood vessels,43, 44 this result underlines the fundamental difference of the two animal models—the leptin gene mutation versus the leptin receptor mutation—and the resolving power of our imaging approach to differentiate their difference in islet vascular remodeling.


The adult ob/ob and db/db mice have been widely used to study neuropathies and vascular diseases in T2D.45, 46, 47, 48 However, before the tissue damages, the early events of the islet neurovascular remodeling have not been systematically demonstrated, which is largely due to the lack of imaging tools to globally visualize the hyperplastic islets in obesity. In this research, we adjusted the pancreatic optical property to enable whole-islet histology via deep-tissue confocal microscopy (Figures 1a–f and Supplementary Movie S1). This approach was combined with tile scanning to visualize the pancreatic islets and the associated ductal and neurovascular networks in an integrated manner. Importantly, in the young (2 months) obese db/db and ob/ob mice, we identified the unique neuro-insular-ductal complex, featuring the prominent peri- and/or intra-islet ducts with rich sympathetic innervation at the islet–duct boundaries (Figures 2 and 3). In vascular characterization, we revealed a decreased population of the ob/ob islet pericytes, which adapt to ensheathing the dilated microvessels with hypertrophic processes (Figure 6). These novel anatomies highlight the advantage of using 3-D microscopy to explore the unknown details of the islet microenvironment in obesity.

The remodeling of islet microstructure to include the intra-islet duct has only been sporadically reported.49, 50, 51 In Figures 2d–i and Supplementary Movie S3, we reveal the intra-islet ducts in the db/db islets with two unique features. First, the intra-islet ducts spontaneously occur in these young animals in response to the obesity with intrinsic leptin signal deficiency, rather than being induced by pancreatectomy49 or transgenic growth factor expression in the pancreas.50, 51 Second, the intra-islet ducts connect to the pancreatic ductal network (Figures 2e and f), suggesting a direct participation of the ductal epithelium in budding and entering the peri- and intra-islet domains. However, we do not rule out the endocrine-to-ductal transdifferentiation in this unique islet microenvironment, as the acinar-to-ductal transdifferentiation has been demonstrated in the formation of pancreatic ductal lesions.52 Definitive proof for a ductal cell or endocrine cell of origin that participates in the formation of the intra-islet duct will require lineage-tracing experiments with appropriate labeling of the ductal and endocrine cells of the db/db mice.

Based on the known abilities of the pancreatic duct to release the angiogenic and neurotrophic factors,41, 53, 54 the intimate islet–duct association also suggests that the ductal epithelium could create an appropriate islet microenvironment for cell proliferation. In this research, we show that the release of NGF recruits the sympathetic nerves, leading to an increase in islet sympathetic innervation in the obese young mice (Figure 3). Note that in mouse development the pancreatic/islet sympathetic innervation is important for the islet formation and functional maturation,54 whereas in the adult animals the sympathetic nerves are crucial in modulating the islet hormone secretion in response to the stress conditions such as the marked hypoglycemia.6 Here, in the young obese animals, the islet–duct complex and the increased sympathetic innervation seemingly provide an appropriate environment, similar to that of the islet development, for β-cell expansion to produce more insulin with a safeguard against its risk. However, the increase in the sympathetic input could also cause the islet under chronic sympathetic stress, leading to accelerated β-cell apoptosis as observed in the adult humans and animals with T2D.55 This is an important insight that must be validated experimentally, but illustrates the value of using our imaging approach to study the hyperplastic islets in the young obese animals.

In addition to the new islet structure and innervation pattern, the 3-D images of the ob/ob islets also reveal the associated remodeling of the islet microvessels and the perivascular pericytes. The pericytes, also known as the mural cells, reside on the abluminal side of the endothelial cells to establish physical contacts and the paracrine signaling to maintain the vascular system. Using the developmental mouse brain, Hellstrom et al.56 have identified that the lack of pericytes leads to blood vessel dilation, hyperplasia of endothelial cells and formation of microaneurysm.56 Here, in the 2-month ob/ob mice, we confirm that a similar dilation of the islet microvessels occurs in association with a decreased density of pericytes (Figures 6g–p), which was not identified in the older 4-month ob/ob mice.21 Importantly, in the young mice, we demonstrate that the dilated microvessels are surrounded by the hypertrophic pericytes with an apparent increase in the contact areas between the endothelium and the pericyte processes (Figures 6c versus i). The morphology of these ‘sheath-like’ pericytes suggests a stop-gap mechanism to structurally support the deformed islet endothelium to maintain the vascular function.

In summary, we developed a penetrative islet imaging method for global visualization of the hyperplasic islets in the young obese mice. The high-definition islet images and movies reveal three levels of tissue remodeling in response to the metabolic challenge. First, at the islet structural level, the islet–duct complex is identified with the db/db and ob/ob islets. Second, at the neuronal level, we identify the rich sympathetic innervation of the islet–duct complex. Third, at the vascular level, we identify the association between the microvessel dilation and pericyte remodeling in the ob/ob islets. Overall, the 3-D imaging and illustration of the islet remodeling in the young obese mice will lay the technical and morphological foundations to explore the human islet remodeling in obesity.


  1. 1

    D'Adamo E, Caprio S . Type 2 diabetes in youth: epidemiology and pathophysiology. Diabetes Care 2011; 34: S161–S165.

    Article  Google Scholar 

  2. 2

    Chen L, Magliano DJ, Zimmet PZ . The worldwide epidemiology of type 2 diabetes mellitus—present and future perspectives. Nat Rev Endocrinol 2012; 8: 228–236.

    CAS  Article  Google Scholar 

  3. 3

    Linnemann AK, Baan M, Davis DB . Pancreatic beta-cell proliferation in obesity. Adv Nutr 2014; 5: 278–288.

    CAS  Article  Google Scholar 

  4. 4

    Okamoto H, Hribal ML, Lin HV, Bennett WR, Ward A, Accili D . Role of the forkhead protein FoxO1 in beta cell compensation to insulin resistance. J Clin Invest 2006; 116: 775–782.

    CAS  Article  Google Scholar 

  5. 5

    Woods SC, Porte D Jr . Neural control of the endocrine pancreas. Physiol Rev 1974; 54: 596–619.

    CAS  Article  Google Scholar 

  6. 6

    Ahren B . Autonomic regulation of islet hormone secretion—implications for health and disease. Diabetologia 2000; 43: 393–410.

    CAS  Article  Google Scholar 

  7. 7

    Ahren B . Islet nerves in focus—defining their neurobiological and clinical role. Diabetologia 2012; 55: 3152–3154.

    CAS  Article  Google Scholar 

  8. 8

    Richards OC, Raines SM, Attie AD . The role of blood vessels, endothelial cells, and vascular pericytes in insulin secretion and peripheral insulin action. Endocr Rev 2010; 31: 343–363.

    CAS  Article  Google Scholar 

  9. 9

    Hayden MR, Karuparthi PR, Habibi J, Lastra G, Patel K, Wasekar C et al. Ultrastructure of islet microcirculation, pericytes and the islet exocrine interface in the HIP rat model of diabetes. Exp Biol Med (Maywood) 2008; 233: 1109–1123.

    CAS  Article  Google Scholar 

  10. 10

    Bergers G, Song S . The role of pericytes in blood-vessel formation and maintenance. Neuro Oncol 2005; 7: 452–464.

    CAS  Article  Google Scholar 

  11. 11

    Ahren B, Holst JJ . The cephalic insulin response to meal ingestion in humans is dependent on both cholinergic and noncholinergic mechanisms and is important for postprandial glycemia. Diabetes 2001; 50: 1030–1038.

    CAS  Article  Google Scholar 

  12. 12

    Boulton AJ, Vinik AI, Arezzo JC, Bril V, Feldman EL, Freeman R et al. Diabetic neuropathies: a statement by the American Diabetes Association. Diabetes Care 2005; 28: 956–962.

    Article  Google Scholar 

  13. 13

    Creager MA, Luscher TF, Cosentino F, Beckman JA . Diabetes and vascular disease: pathophysiology, clinical consequences, and medical therapy: Part I. Circulation 2003; 108: 1527–1532.

    Article  Google Scholar 

  14. 14

    Stratton IM, Adler AI, Neil HA, Matthews DR, Manley SE, Cull CA et al. Association of glycaemia with macrovascular and microvascular complications of type 2 diabetes (UKPDS 35): prospective observational study. BMJ 2000; 321: 405–412.

    CAS  Article  Google Scholar 

  15. 15

    Giugliano D, Ceriello A, Paolisso G . Oxidative stress and diabetic vascular complications. Diabetes Care 1996; 19: 257–267.

    CAS  Article  Google Scholar 

  16. 16

    Vincent AM, Russell JW, Low P, Feldman EL . Oxidative stress in the pathogenesis of diabetic neuropathy. Endocr Rev 2004; 25: 612–628.

    CAS  Article  Google Scholar 

  17. 17

    Teitelman G, Guz Y, Ivkovic S, Ehrlich M . Islet injury induces neurotrophin expression in pancreatic cells and reactive gliosis of peri-islet Schwann cells. J Neurobiol 1998; 34: 304–318.

    CAS  Article  Google Scholar 

  18. 18

    Chiu YC, Hua TE, Fu YY, Pasricha PJ, Tang SC . 3-D imaging and illustration of the perfusive mouse islet sympathetic innervation and its remodelling in injury. Diabetologia 2012; 55: 3252–3261.

    Article  Google Scholar 

  19. 19

    Tang SC, Chiu YC, Hsu CT, Peng SJ, Fu YY . Plasticity of Schwann cells and pericytes in response to islet injury in mice. Diabetologia 2013; 56: 2424–2434.

    Article  Google Scholar 

  20. 20

    Juang JH, Kuo CH, Peng SJ, Tang SC . 3-D imaging reveals participation of donor islet Schwann cells and pericytes in islet transplantation and graft neurovascular regeneration. Ebiomedicine 2015; 2: 109–119.

    Article  Google Scholar 

  21. 21

    Dai C, Brissova M, Reinert RB, Nyman L, Liu EH, Thompson C et al. Pancreatic islet vasculature adapts to insulin resistance through dilation and not angiogenesis. Diabetes 2013; 62: 4144–4153.

    CAS  Article  Google Scholar 

  22. 22

    Slack JM . Developmental biology of the pancreas. Development 1995; 121: 1569–1580.

    CAS  PubMed  Google Scholar 

  23. 23

    Dor Y, Brown J, Martinez OI, Melton DA . Adult pancreatic beta-cells are formed by self-duplication rather than stem-cell differentiation. Nature 2004; 429: 41–46.

    CAS  Article  Google Scholar 

  24. 24

    Butler AE, Galasso R, Matveyenko A, Rizza RA, Dry S, Butler PC . Pancreatic duct replication is increased with obesity and type 2 diabetes in humans. Diabetologia 2010; 53: 21–26.

    CAS  Article  Google Scholar 

  25. 25

    Fu YY, Lin CW, Enikolopov G, Sibley E, Chiang AS, Tang SC . Microtome-free 3-dimensional confocal imaging method for visualization of mouse intestine with subcellular-level resolution. Gastroenterology 2009; 137: 453–465.

    Article  Google Scholar 

  26. 26

    Chung K, Wallace J, Kim SY, Kalyanasundaram S, Andalman AS, Davidson TJ et al. Structural and molecular interrogation of intact biological systems. Nature 2013; 497: 332–337.

    CAS  Article  Google Scholar 

  27. 27

    Marx V . Microscopy: seeing through tissue. Nat Methods 2014; 11: 1209–1214.

    CAS  Article  Google Scholar 

  28. 28

    Tang SC, Peng SJ, Chien HJ . Imaging of the islet neural network. Diabetes Obes Metab 2014; 16 (Supp 1): 77–86.

    Article  Google Scholar 

  29. 29

    Willyard C . Heritability: the family roots of obesity. Nature 2014; 508: S58–S60.

    CAS  Article  Google Scholar 

  30. 30

    Fernandez JR, Klimentidis YC, Dulin-Keita A, Casazza K . Genetic influences in childhood obesity: recent progress and recommendations for experimental designs. Int J Obes (Lond) 2012; 36: 479–484.

    CAS  Article  Google Scholar 

  31. 31

    Hinney A, Vogel CI, Hebebrand J . From monogenic to polygenic obesity: recent advances. Eur Child Adolesc Psychiatry 2010; 19: 297–310.

    Article  Google Scholar 

  32. 32

    Kanasaki K, Koya D . Biology of obesity: lessons from animal models of obesity. J Biomed Biotechnol 2011; 2011: 197636.

    Article  Google Scholar 

  33. 33

    Llewellyn CH, Trzaskowski M, van Jaarsveld CH, Plomin R, Wardle J . Satiety mechanisms in genetic risk of obesity. JAMA Pediatr 2014; 168: 338–344.

    Article  Google Scholar 

  34. 34

    Fu YY, Lu CH, Lin CW, Juang JH, Enikolopov G, Sibley E et al. Three-dimensional optical method for integrated visualization of mouse islet microstructure and vascular network with subcellular-level resolution. J Biomed Opt 2010; 15: 046018.

    Article  Google Scholar 

  35. 35

    Fu YY, Peng SJ, Lin HY, Pasricha PJ, Tang SC . 3-D imaging and illustration of mouse intestinal neurovascular complex. Am J Physiol Gastrointest Liver Physiol 2013; 304: G1–11.

    CAS  Article  Google Scholar 

  36. 36

    Liu YA, Chung YC, Shen MY, Pan ST, Kuo CW, Peng SJ et al. Perivascular interstitial cells of cajal in human colon. Cell Mol Gastroenterol Hepatol 2015; 1: 102–119.

    Article  Google Scholar 

  37. 37

    Juang JH, Peng SJ, Kuo CH, Tang SC . Three-dimensional islet graft histology: panoramic imaging of neural plasticity in sympathetic reinnervation of transplanted islets under the kidney capsule. Am J Physiol Endocrinol Metab 2014; 306: E559–E570.

    CAS  Article  Google Scholar 

  38. 38

    Genina EA, Bashkatov AN, Tuchin VV . Tissue optical immersion clearing. Expert Rev Med Devices 2010; 7: 825–842.

    Article  Google Scholar 

  39. 39

    Liu YA, Chung YC, Pan ST, Hou YC, Peng SJ, Pasricha PJ et al. 3-D illustration of network orientations of interstitial cells of Cajal subgroups in human colon as revealed by deep-tissue imaging with optical clearing. Am J Physiol Gastrointest Liver Physiol 2012; 302: G1099–G1110.

    CAS  Article  Google Scholar 

  40. 40

    Hua TE, Yang TL, Yang WC, Liu KJ, Tang SC . 3-D neurohistology of transparent tongue in health and injury with optical clearing. Front Neuroanat 2013; 7: 36.

    Article  Google Scholar 

  41. 41

    Toma H, Winston J, Micci MA, Shenoy M, Pasricha PJ . Nerve growth factor expression is up-regulated in the rat model of L-arginine-induced acute pancreatitis. Gastroenterology 2000; 119: 1373–1381.

    CAS  Article  Google Scholar 

  42. 42

    Street CN, Lakey JR, Shapiro AM, Imes S, Rajotte RV, Ryan EA et al. Islet graft assessment in the Edmonton Protocol: implications for predicting long-term clinical outcome. Diabetes 2004; 53: 3107–3114.

    CAS  Article  Google Scholar 

  43. 43

    Cooke JP, Oka RK . Does leptin cause vascular disease? Circulation 2002; 106: 1904–1905.

    Article  Google Scholar 

  44. 44

    Sikka G, Yang R, Reid S, Benjo A, Koitabashi N, Camara A et al. Leptin is essential in maintaining normal vascular compliance independent of body weight. Int J Obes (Lond) 2010; 34: 203–206.

    CAS  Article  Google Scholar 

  45. 45

    Drel VR, Mashtalir N, Ilnytska O, Shin J, Li F, Lyzogubov VV et al. The leptin-deficient (ob/ob) mouse: a new animal model of peripheral neuropathy of type 2 diabetes and obesity. Diabetes 2006; 55: 3335–3343.

    CAS  Article  Google Scholar 

  46. 46

    Sullivan KA, Hayes JM, Wiggin TD, Backus C, Su Oh S, Lentz SI et al. Mouse models of diabetic neuropathy. Neurobiol Dis 2007; 28: 276–285.

    CAS  Article  Google Scholar 

  47. 47

    Su W, Guo Z, Randall DC, Cassis L, Brown DR, Gong MC . Hypertension and disrupted blood pressure circadian rhythm in type 2 diabetic db/db mice. Am J Physiol Heart Circ Physiol 2008; 295: H1634–H1641.

    CAS  Article  Google Scholar 

  48. 48

    Sharma K, McCue P, Dunn SR . Diabetic kidney disease in the db/db mouse. Am J Physiol Renal Physiol 2003; 284: F1138–F1144.

    CAS  Article  Google Scholar 

  49. 49

    Tulachan SS, El-Gohary Y, Gittes GK . Pancreatic duct-to-β-cell transdifferentiation represents the most likely source of new beta cells during postnatal growth and regeneration. Gastroenterology 2014; 146 (Suppl 1): S81–S82.

    Article  Google Scholar 

  50. 50

    Means AL, Ray KC, Singh AB, Washington MK, Whitehead RH, Harris RC Jr . et al. Overexpression of heparin-binding EGF-like growth factor in mouse pancreas results in fibrosis and epithelial metaplasia. Gastroenterology 2003; 124: 1020–1036.

    CAS  Article  Google Scholar 

  51. 51

    Krakowski ML, Kritzik MR, Jones EM, Krahl T, Lee J, Arnush M et al. Pancreatic expression of keratinocyte growth factor leads to differentiation of islet hepatocytes and proliferation of duct cells. Am J Pathol 1999; 154: 683–691.

    CAS  Article  Google Scholar 

  52. 52

    Strobel O, Dor Y, Alsina J, Stirman A, Lauwers G, Trainor A et al. In vivo lineage tracing defines the role of acinar-to-ductal transdifferentiation in inflammatory ductal metaplasia. Gastroenterology 2007; 133: 1999–2009.

    Article  Google Scholar 

  53. 53

    Xiao X, Prasadan K, Guo P, El-Gohary Y, Fischbach S, Wiersch J et al. Pancreatic duct cells as a source of VEGF in mice. Diabetologia 2014; 57: 991–1000.

    CAS  Article  Google Scholar 

  54. 54

    Borden P, Houtz J, Leach SD, Kuruvilla R . Sympathetic innervation during development is necessary for pancreatic islet architecture and functional maturation. Cell Rep 2013; 4: 287–301.

    CAS  Article  Google Scholar 

  55. 55

    Talchai C, Lin HV, Kitamura T, Accili D . Genetic and biochemical pathways of beta-cell failure in type 2 diabetes. Diabetes Obes Metab 2009; 11: 38–45.

    CAS  Article  Google Scholar 

  56. 56

    Hellstrom M, Gerhardt H, Kalen M, Li X, Eriksson U, Wolburg H et al. Lack of pericytes leads to endothelial hyperplasia and abnormal vascular morphogenesis. J Cell Biol 2001; 153: 543–553.

    CAS  Article  Google Scholar 

Download references


We thank the Brain Research Center in National Tsing Hua University for technical support of confocal imaging. This work was supported in part by grants from the Chang Gung Memorial Hospital (CMRP G3C0191 and CMRPG3D0601) to JHJ, and Taiwan National Health Research Institutes (NHRI-EX102-10044EI and NHRI-EX103-10332EI) and National Science Council (NSC 102-2628-B-007-002-MY2) to SCT.

Author contributions

All authors contributed to experimental design, data analysis and interpretation of data. HJC, SJP and TEH contributed to specimen preparation and data collection. CHK and JHJ contributed to the islet transplantation experiment. SCT directed the imaging project and contributed to the writing of the paper. All the authors approved the final version of the paper.

Author information



Corresponding author

Correspondence to S-C Tang.

Ethics declarations

Competing interests

The authors declare no conflict of interest.

Additional information

Supplementary Information accompanies this paper on International Journal of Obesity website

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Chien, HJ., Peng, SJ., Hua, TE. et al. 3-D imaging of islets in obesity: formation of the islet–duct complex and neurovascular remodeling in young hyperphagic mice. Int J Obes 40, 685–697 (2016).

Download citation

Further reading


Quick links