A Practical Guide to Rodent Islet Isolation and Assessment
© to the author(s) 2009
Received: 18 June 2009
Accepted: 22 October 2009
Published: 3 December 2009
Skip to main content
© to the author(s) 2009
Received: 18 June 2009
Accepted: 22 October 2009
Published: 3 December 2009
Pancreatic islets of Langerhans secrete hormones that are vital to the regulation of blood glucose and are, therefore, a key focus of diabetes research. Purifying viable and functional islets from the pancreas for study is an intricate process. This review highlights the key elements involved with mouse and rat islet isolation, including choices of collagenase, the collagenase digestion process, purification of islets using a density gradient, and islet culture conditions. In addition, this paper reviews commonly used techniques for assessing islet viability and function, including visual assessment, fluorescent markers of cell death, glucose-stimulated insulin secretion, and intracellular calcium measurements. A detailed protocol is also included that describes a common method for rodent islet isolation that our laboratory uses to obtain viable and functional mouse islets for in vitro study of islet function, beta-cell physiology, and in vivo rodent islet transplantation. The purpose of this review is to serve as a resource and foundation for successfully procuring and purifying high-quality islets for research purposes.
Pancreatic islets are thought to play a key role in the pathophysiology of Type 1 and Type 2 diabetes through the failure of islet beta cells to secrete sufficient quantities of insulin to regulate blood glucose (1). In recent years, increasing interest surrounding islet replacement therapies in humans has provided the drive for advances in the methods used to isolate islets from humans as well as a host of animal research models (2, 3). Although there are many published islet isolation protocols specific to mouse and rat, few provide the necessary details for researchers to successfully perform the complex procedures.
The primary goal of isolating pancreatic islets, whether for in vivo transplantation or in vitro studies, is to obtain viable purified islets that respond in a manner consistent with their function in vivo. The key elements of a successful islet isolation procedure are: (1) enzymatically digesting the tissues connecting the islets to the exocrine tissue, (2) separating islets from non-islet tissue, and (3) culturing isolated islets in an environment that maintains cell viability. We review these key elements and provide methods for evaluating islet quality. In addition, we present a common method for the isolation of rodent islets used regularly in our islet isolation facility to consistently procure viable and functional islets for in vitro study of islet function, beta-cell physiology, and in vivo islet transplantation. Our protocol is, by no means, the only successful method for isolating pancreatic islets; however, the additional details provided in this protocol are intended to provide the rationale for each step in the process in order to assist researchers in their efforts to obtain healthy islets for study.
The two most prevalent approaches for isolating islets from rodent pancreatic tissue differ, primarily, in the way digestive enzymes are introduced to the pancreatic tissue surrounding the islets. In the first approach, the pancreas is excised from a euthanized animal and cut into 1–2 mm pieces, thus increasing the surface area and providing conditions for the digestive enzyme collagenase to break down the tissue surrounding the islets (4, 5). The pancreatic pieces are enzymatically digested in a collagenase solution and concurrently mechanically digested by being stirred or shaken. In the second approach, described by Gotoh et al., collagenase is injected into the common bile duct (CBD) of a euthanized animal. The pancreas is then excised and digested at 37°C without being cut into pieces or mechanically digested (6).
Although these two approaches form the foundation for many islet isolation techniques, there are considerable variations in the details among published methods, as well as alternative methods for isolating islets (7–11). The advantages of the CBD method described by Gotoh et al. are twofold: (1) perfusing the pancreas through the CBD allows collagenase to access the islets using anatomical structures, and (2) stationary digestion reduces mechanical damage to the islet. A comparison of these two methods showed that either method can be used to procure viable and functional rat islets; however, the Gotoh et al. method produced an islet yield approximately 50% higher and was more cost effective (12).
Our laboratory uses a modification of the CBD protocol when isolating rodent islets (see Appendix A for a annotated murine protocol and Appendix B for an abbreviated version). Although cannulation of the rodent bile duct requires technical skill, we are partial to this approach. Collagenase may interact more closely with the connective tissue surrounding the islets when delivered through intact anatomical structures, which results in a higher islet yield as suggested by others (12). Szot et al. provide a detailed video account of a rodent islet isolation using a method of bile duct cannulation (13). We describe an alternative method of injecting lobes of the pancreas in situ for cases when the bile duct has been compromised or cannot be used (see Appendix A , 12B). When evaluating any islet isolation protocol, one must consider that the outcome is influenced by differences in the type and concentration of collagenase used, the method of collagenase administration, the temperature and duration of pancreas digestion, the method for purifying islets from pancreatic acinar tissue, and the culture conditions following isolation.
Collagenase enzymes are routinely used in digesting connective tissue that binds islets to other pancreatic tissue. Variability exists between manufacturers and between each lot of collagenase product from the same manufacturer. Enzyme activity, purity, and formulation, therefore, strongly influence the outcome of the islet isolation. The composition of collagenase and other enzymes in each lot must be ideal to the task of isolating islets specifically. Wolters et al. have described, in detail, the differences in isolation with purified collagenase types 1 and 2 separately, and the combination that yields the most effective rat islet isolation (14, 15). De Haan et al. formulated criteria for evaluating each lot of commercially available collagenase to ensure proper digestion of rat islets (16). Formulations with increased collagenase activity, a specific range of both neutral proteases and clostripain, and with low levels of trypsin activity may produce the most viable islets (14, 16). We have found that optimal collagenase formulations for rat islet isolations also provide acceptable criteria for rating collagenase used in our mouse islet isolation procedures.
Digestive enzyme formulations for islet isolation range from crude collagenases to highly purified combinations used extensively in human islet transplants. Brandhorst et al. suggest that the differences from lot to lot may be due to the lack of proper accounting of tryptic-like activity, even in the most pure mixtures of collagenases, such as Liberase HI (Roche, Indianapolis, IN, USA) and collagenase NB1 (Serva, Heidelberg, Germany) used in isolation of human islets for transplantation (17). The tryptic-like activity in enzyme blends may work in concert with the other enzymes to increase the activity of the digestion, although there is some debate about the damage tryptic-like activity has on the islets (18). Enzyme blends with high purity and precise notation of the components are used in human islet isolation to ensure consistent activity and reproducibility (17, 19). Enzymes with higher purity, and consequently a higher price, are also used in human pancreatic islet isolation to reduce incidence of contamination by endotoxins.
Endotoxins correlate with increased proinflammatory cytokines in models of transplantation (19). Endotoxins have been identified in both the digestion and gradient separation steps. Collagenase formulations as well as different types of gradient compounds have been identified as containing endotoxins in varying amounts (19). Endotoxins are of particular concern in human-to-human islet transplantation procedures. Infiltration of transplanted islets by inflammatory cytokines has been attributed to endotoxin contamination (19).
Almost as important as the specific formulation of collagenase is ensuring that a protocol is optimized to the collagenase type and activity and to the animal species and strain in a given protocol. Factors that influence the process include digestion time, digestion temperature, collagenase concentration, and the route of administration of collagenase, which vary widely among protocols. Perfusing the pancreas through the common bile duct allows collagenase to access the islets using biological structures, which may change the duration of digestion when compared to other methods. Differences between collagenase batches and other factors that influence digestion facilitate the need for testing each protocol for optimal islet viability and islet function, which remains paramount to success and reproducibility of islet isolation.
Our laboratory follows published guidelines for collagenase enzyme formulations (16), which provide consistency for expected outcomes during digestion. We use Collagenase P (Roche, Indianapolis, IN, USA) enzyme at 1.4 mg/mL in a modified Hank’s Balanced Salt Solution (HBSS; Invitrogen, Carlsbad, CA, USA) injected into the pancreas via the CBD. The pancreas is then excised whole and placed in modified HBSS for stationary digestion at 37°C for 8–11 min as described in detail in Appendices A and B .
There is some debate regarding the use of a density gradient to purify islets from acinar tissue. Purifying islets from acinar tissue, regardless of the method, is important due to the nature of the pancreatic tissue. The cells of the exocrine pancreas secrete various digestive enzymes necessitating the separation of islets from pancreatic acinar tissue (20). Sodium diatrizoate, Histopaque (Sigma-Aldrich, St. Louis, MO, USA), is a hypertonic solution also used in isolating other cell types. Our laboratory has used Ficoll 400 (Sigma-Aldrich, St. Louis, MO, USA) and Histopaque at different densities. In isolations with Ficoll 400 layered in a discontinuous gradient of 1.109, 1.096, 1.070, and 0.570 g/mL islets were isolated from the interfaces of both the 1.070/1.096 and 1.096/1.109 g/mL layers. However, we found that the preparations were often contaminated with acinar tissue. Our purity results generally improved using aseptically filled and premixed Histopaque, which is also available in sterile preparations. Combining Histopaque 1.119 g/mL with the 1.077 g/mL preparation to produce a 1.100 g/mL gradient appears to enhance islet purity. It should be noted that these studies comparing Histopaque and Ficoll were not rigorous, and both gradients are widely used and accepted. We provide our anecdotal evidence for consideration in choosing a purification method.
The final purity of the product depends on the animal strain and the characteristics of density gradients. In our experience, lean rodents tend to yield a higher purity of the final preparation than those with more fat. There is also a strain-dependent difference in the outcome of the gradient purification, which is consistent with findings describing strain-dependent differences in islet isolation (21).
A second purification of islets from acinar tissue is often needed to further increase islet purity prior to culture. Our protocol includes using a microscope to identify islets, then handpicking those islets from one suspension culture dish into a second dish containing a buffered solution or culture medium, such as HBSS or RPMI, respectively. Islets can then be transferred to a dish containing culture media for overnight incubation. Once islets have been transferred to media, minimizing time outside the sterile incubator will limit exposure to contamination and pH changes while handpicking islets for experiments.
The total number of islets found in a rodent pancreas varies considerably among strains. Bock et al. compared seven different mouse strains and found the number of islets per pancreas ranged from 971 ± 88 (129S6 mice) to 2,509 ± 133 (B6 mice) (22). Using a mouse model of diabetes, Bock et al. identified a similar number of islets per pancreas (~3,200) for both ob/ob and ob/+ control mice; the islets from the diabetes-prone ob/ob mice were 3.6 times larger, however, than ob/+ controls (23). In young male Wistar rats, Inuwa et al. demonstrated that the total islet number increased with age, ranging from ~6,000 to ~20,000 islets per pancreas (24). Other studies have estimated the number of rat islets as low as approximately 3,000–5,000 per pancreas (25, 26). Therefore, the expected islet yield from an isolation procedure depends a great deal on the age and strain of the rodent.
The expertise of the technician, as well as the method of isolation chosen, will also influence the total islet yield. Consequently, a definitive expected yield is difficult to quantify. With an experienced technician yields from various mouse strains should range from 200–400 with average yields of 300 islets per mouse (6, 9). Rat yields range from approximately 600–800 islets per animal (21). Based on estimates of total islet numbers, this suggests that the islet yield ranges from 10% to 30% for the typical rodent pancreas. For comparison, the human pancreas is thought to contain over one million islets, and the typical isolation yields approximately 250,000–450,000 islets as estimated by islet equivalents (27).
After performing islet isolation, proper culture conditions are imperative to ensuring that the islets are able to recover from the insult of collagenase digestion. Examination of media with different glucose concentrations indicated that islets cultured with 11 mM glucose have lower apoptosis rates and increased viability than those in media with both higher and lower glucose concentrations for rodents (28). Media with glucose concentrations substantially below 11 mM can reduce islet insulin content and downregulate key genes related to glucose metabolism, whereas extended exposure to high glucose causes toxicity (28, 29). Studies of optimal culture conditions demonstrated that RPMI 1640 with serum maintains or augments glucose-stimulated insulin secretion in murine islets (30). Insulin secretion remained lower in islets cultured in five other types of culture media brought to comparable glucose concentrations (30). Thus, properties apart from its higher glucose concentration (11 mmol/L) make RPMI 1640 suited for studies of insulin secretion in murine islets (30). In another study CMRL1066, rather than RPMI 1640, was used to culture rat islets in order to preserve the insulin secretory capacity (16).
We use RPMI 1640 culture medium both for culturing islets and while purifying islets from acinar tissue after the density gradient separation. RPMI1640 is supplemented with 10% (v/v) fetal bovine serum to promote viability and with 100 U/mL penicillin and 100 μg/mL streptomycin to reduce the possibility of contamination. Islets are plated in 100 × 20 mm suspension culture dishes (Corning Inc., # 430591), rather than culture-treated dishes, to decrease islet attachment. Islets are cultured in 10 mL of RPMI 1640 media in these dishes.
To maintain islets for long-term culture, the optimal islet density is four islets per square centimeter in order to prevent competition for nutrients (16). Overnight incubation of 16–20 h provides islets time to recover from the harsh process of collagenase digestion. Recovery in a sterile incubator at 37°C with 5% CO2 infusion and humidified air is necessary for islet function prior to performing viability or functional assessment assays (30). Rodent islets can maintain glucose sensitivity for at least 1 week in culture with frequent medium changes (30) and perhaps even longer based on data from human islets (30, 31); however, changes in rodent islet function can occur between as little as between 1 and 4 days in culture (32).
Supplementing visual inspection with additional techniques can provide quantification of islet viability and functionality. We define viability as living versus dead or dying cells as assessed using cell exclusion or DNA-binding dyes. A common approach in the human islet transplant field is to measure the ratio of healthy living cells to dead cells within each islet with fluorescence microscopy. Fluorescein diacetate (FDA) incorporates into healthy cells by facilitated diffusion and fluoresces blue; propidium iodide (PI; Sigma-Aldrich, St. Louis, MO, USA) is a membrane impermeant red fluorescent dye that is excluded from viable cells and enters only dead or dying cells (34). Using these fluorescent dyes in combination, the health of islets can be assessed by the FDA/PI ratio. Unmanipulated isolated islets from healthy control animals generally have 90–95% viability, meaning blue FDA staining in 90–95% of the component cells of an islet and red PI staining is detectable in only 5–10% of the cells in any given islet. Additional techniques commonly used to measure islet viability include AnnexinV, SYTO-13/ethidium bromide, calcein AM/ethidium homodimer, fluorescein diacetate, and ethidium bromide which are more expensive but are also more sensitive to islet cell damage than FDA/PI (34–36).
We use a variation of the above approach by quantifying the mean intensity of PI fluorescence within an islet. Imaging software records the fluorescence intensity within a region of interest (ROI) designated by an encircled islet. Healthy islets typically display smooth round borders (Fig. 2a, top panel) and show minimal PI staining (Fig. 2a, bottom panel). Unhealthy cells, such as islets cultured overnight in a mixture of proinflammatory cytokines (Fig. 2b), exhibit cells protruding from the rough islet surface (top panel) and significantly greater PI staining (bottom panel). The mean pixel fluorescence intensity of the ROIs is used to quantify cell death as listed atop the lower panels of Fig. 2a, b.
A fundamental property of pancreatic islets is their capacity to regulate the release of insulin and other hormones in direct response to changes in extracellular glucose concentration. In large part, this ability defines islet function since insulin is produced and released in the body only from islet beta cells. Insulin is crucial to regulating blood glucose, and reduced insulin secretion is a key feature of both Type 1 and Type 2 diabetes (1). Glucagon, somatostatin, and other peptides are also produced by islet cells, but these are secreted in smaller amounts and more difficult to detect. Glucose-stimulated insulin secretion (GSIS) is thus a well-accepted measure of islet function.
To measure GSIS, islets are cultured in a ‘low’ glucose concentration, typically near 3 mM, to measure the amount of insulin secreted into the media under ‘basal’ or ‘unstimulated’ conditions. Stimulated insulin release is measured by exposing islets to a higher glucose concentration such as 11.1 mM (half-maximal) or >28 mM (maximal). In response to the glucose stimulation, the time course of the islet response is biphasic, consisting of a rapid spike in insulin release (first phase) followed by a decline to a prolonged second phase plateau of insulin that remains throughout the duration of the stimulus. GSIS can be measured either by static conditions or by perfusing islets to measure the kinetics of insulin release in response to glucose. Each technique has its advantages and disadvantages, which are reviewed elsewhere (37). The glucose stimulation index (SI) is the ratio of stimulated-to-basal insulin secretion. Healthy islets have an SI of 2–20 depending on several factors including strain, age, and body weight.
In nearly any cell type, the tight regulation of intracellular calcium ([Ca2+]i) is crucial to normal functioning of many cellular processes including metabolism, signal transduction, and exocytosis (38). As in the GSIS assay, analyzing islet [Ca2+]i in response to changes in extracellular glucose concentration provides supportive information about islet viability and function because calcium is an integral component of the insulin secretion pathway (39–41). Therefore, measuring the glucose-stimulated [Ca2+]i response (GSCa) in islets offers a reasonable reflection of insulin secretion (42) and more importantly, a potentially sensitive indicator of overall islet health and function in vitro (43). A number of fluorescent probes are commonly used to measure intracellular calcium including fura-2 acetoxy-methyl-ester (fura-2AM), fluo-3, fluo-4, and fura red.
There are several advantages of measuring GSCa over GSIS. Substantial time and costs are involved in GSIS since insulin must be measured by immunoassay following the GSIS experiment, whereas GSCa provides results in real-time without additional expense. Furthermore, GSCa utilizes frequent sampling, which allows for precise temporal analysis of the amplitude, latency, and trajectory of changes in response to glucose stimulation. Also, GSCa can be used to assess individual islets, so fewer than ten islets would be sufficient to characterize the function for the entire batch. In contrast, islets must be grouped together for GSIS to produce detectable quantities of insulin, especially for islet perfusion studies (we use 50 islets for each replicate in perfusion studies). One limitation, however, is that GSCa cannot be easily normalized to a standard value in the same way that insulin is normalized to a standard set of insulin concentrations. Coupled with potential differences in dye loading and changes in light source intensity/efficiency over long periods of time, these issues make batch-to-batch comparisons of different islet preparations using GSCa difficult. These issues can be somewhat mitigated by determining the stimulation index, which is commonly used to assess GSIS from donor to donor in human islets for transplantation purposes (45).
Islets are composed of several distinct cell types consisting of the glucogon-secreting alpha cells, insulin-secreting beta cells, somatostatin-producing delta cells, and others (46–48). The percentage of these cells, as well as their anatomical locations in islets, varies between species. In rodents, the majority of cells are beta cells (65–85%) and alpha cells (10–25%), with the remaining 5–10% of cells consisting of delta cells and other cell types (46–48). Isolating these cells requires either mechanically disrupting the bonds between cells or using a digestive enzyme to separate the cells (49). Once islets have been separated into their component cells, use of counterflow elution as described by Pipeleers (50), or light scatter flow cytometry as described by Rabinovitch et al. (51) may be used to identify and purify the cells. We provide a detailed protocol describing the dissociation and culturing of murine islet cells in Appendix C .
The ability to consistently procure viable and functional islets is crucial to effectively studying the physiology and pathophysiology of islets and their constituent cells. As stated previously, islet isolation is an intricate process. In this review, we have addressed the key factors to consider in the isolation and assessment processes to obtain both viable and functional islets. In the accompanying protocol, we provide a method developed by integrating the reports of many others in the research field with careful experimentation to optimize the islet isolation process for our laboratory. While following this protocol provides a start for islet isolation, any procedure must be optimized to the capabilities of the laboratory and the specific goals of the study.
This work was supported by National Institutes of Health Grant DK063609 to the University of Virginia Diabetes Endocrine Research Center (DERC) and K01-DK081621 to C.S.N. Mouse islets were acquired through the Cell and Islet Isolation Core Facility at the University of Virginia DERC (DK063609).