A protocol to create gene modified human pseudoislets from dispersed human islet cells that are transduced by lentivirus carrying short hairpin RNA (shRNA) is presented. This protocol utilizes readily available enzyme and culture vessels, can be performed easily, and produces genetically modified human pseudoislets suitable for functional and morphological studies.
Various genetic tools are available to modulate genes in pancreatic islets of rodents to dissect function of islet genes for diabetes research. However, the data obtained from rodent islets are often not fully reproduced in or applicable to human islets due to well-known differences in islet structure and function between the species. Currently, techniques that are available to manipulate gene expression of human islets are very limited. Introduction of transgene into intact islets by adenovirus, plasmid, and oligonucleotides often suffers from low efficiency and high toxicity. Low efficiency is especially problematic in gene downregulation studies in intact islets, which require high efficiency. It has been known that enzymatically-dispersed islet cells reaggregate in culture forming spheroids termed pseudoislets. Size-controlled reaggregation of human islet cells creates pseudoislets that maintain dynamic first phase insulin secretion after prolonged culture and provide a window to efficiently introduce lentiviral short hairpin RNA (shRNA) with low toxicity. Here, a detailed protocol for the creation of human pseudoislets after lentiviral transduction using two commercially available multiwell plates is described. The protocol can be easily performed and allows for efficient downregulation of genes and assessment of dynamism of insulin secretion using human islet cells. Thus, human pseudoislets with lentiviral mediated gene modulation provide a powerful and versatile model to assess gene function within human islet cells.
The loss of functional beta cell mass is the central pathology for both type 1 and type 2 diabetes1. While beta cells are the producers of insulin in pancreatic islets, communication between beta cells and non-beta cells plays a critical role in the regulation of insulin secretion2. In addition, dysregulation of glucagon secretion contributes to hyperglycemia in diabetes3. Thus, there is strong interest to modulate gene expression of cells within pancreatic islets to address the mechanism behind the development of islet dysfunction in diabetes. A variety of approaches including transgenic mice are available to modulate gene expression of mouse islets. However, human and mouse islets show distinct innervation, cell distribution, ratio of beta to alpha cells, and response to secretagogues4. Therefore, direct assessment of gene function in human islets is extremely important for understanding the pathophysiology of human pancreatic islets.
Adenoviral vector is the most widely used viral vector to transduce pancreatic islets in vitro due to the high efficiency of transduction in non-dividing cells. However, adenovirus does not penetrate to the core of islets efficiently, especially in human islets5, and is cytotoxic at high doses6. Comparatively, lentiviral vector is less cytotoxic and delivers exogenous genes permanently into the chromosome of post-mitotic cells, making it a widely tested vehicle for gene therapy7. However, the ability of the lentivirus to penetrate the core of intact human islets is also limited, thus requiring partial dispersion by enzymatic digestion to increase the transduction efficiency8. The caveat with the dispersion of intact human islets is the interruption of cell-cell and cell-matrix communication, which compromises the dynamic regulation of insulin secretion critical for the maintenance of glucose homeostasis in humans9. Thus, it has been challenging to assess the impact of gene modulation on the dynamic regulation of islet function in a model of human islets.
It has been known that dispersed islet cells from human and rodent islets autonomously reaggregate into islet-like structures called “pseudoislets”. Pseudoislets show beta and non-beta cell distribution similar to native islets10,11. Additionally, after long-term culture, native islets progressively lose robust first phase insulin secretion5,10,11,12. Yet, pseudoislets demonstrated better preservation of first phase insulin secretion in response to glucose compared with native islets after the same culture period5. In addition to having better preservation of insulin secretion, size-controlled reaggregation of human islet cells in low attachment plates11 provides a window of opportunity to introduce lentivirus vectors prior to their reaggregation into pseudoislets. Several studies have demonstrated the utility of pseudoislets combined with lentiviral mediated transduction. Caton et al.13 reported that the introduction of the green fluorescent protein (GFP) expressing lentivirus had little effect on insulin secretion while achieving homogenous expression of GFP in rat pseudoislets compared with non-infected control. They also demonstrated the specific effect of different connexins on insulin secretion by overexpressing connexins 32, 36, and 43 via lentivirus13. Human pseudoislets prepared with a commercially available 96-well ultra-low attachment plate demonstrated that lentiviral-mediated overexpression of transcription factor SIX3 improves insulin secretion assessed by static incubation14. Recently, human pseudoislets prepared with a 96-well ultra-low attachment plate were used to downregulate glucokinase via lentiviral short hairpin RNA (shRNA) as a proof of principle to show that glucose-stimulated insulin secretion is reduced, while KCl-stimulated insulin secretion was preserved5. The study also demonstrated that human pseudoislets are similar to native islets in gene expression and secretory profiles, further supporting the utility of human pseudoislets to dissect the regulation of islet function5. Although perifusion was not performed, a bioengineered microwell culture plate that recently became commercially available, was also reported to be compatible for lentiviral transduction and produced human pseudoislets that exhibited excellent insulin secretion in vitro and in vivo after transplantation11. Collectively, human pseudoislet formation combined with lentiviral transduction is a simple and efficient approach to investigate human islet pathophysiology, providing a valuable tool to perform mechanistic studies in human islets.
In the current report, a protocol to form human pseudoislets transduced with lentivirus using two commercially available platforms, a 96-well ultra-low attachment plate and a microwell culture plate is presented. Both achieve efficient modulation of gene expression and create human pseudoislets that are compatible for downstream assessments including static incubation and perifusion.
Prior to commencement of studies, a human subjects research determination was made by the University of Iowa Institutional Review Board, who determined that the study did not meet the criteria for human subjects research. Consult the local review board before the initiation of the study to determine if the source of islets and planned study requires prior approval.
NOTE: Typically, 1,200−1,400 islet equivalent (IEQ) of human islets are required for the formation of 192 pseudoislets at the size of 3,000 cells/pseudoislets in a 96-well ultra-low attachment plate or 1,200 pseudoislets at the size of 500 cells/pseudoislets in a microwell culture plate. IEQ of islets required varies between different preparations of human islets as donor factors (age, health, weight), isolation efficiency, and culture conditions affect the yield of the single cell suspension. In this protocol, lentivirus containing shRNA targeting a gene of interest is used. The cytomegalovirus (CMV) and human phosphoglycerate kinase (hPGK) promoter based lentiviral vectors are reported to down-regulate gene efficiently in human pseudoislets5,15. The use of lentivirus requires precaution as biohazard16. Contact the local biosafety committee prior to the initiation of the use of lentivirus.
1. Overnight Culture of Human Islets for Recovery After Shipment
2. Preparation of Single Cell Suspension from Human Islets
3. Pseudoislet Formation and Transduction by Lentivirus
4. RNA Extraction for Evaluation of Gene Silencing Efficiency
Figure 1 illustrates key steps in the production of pseudoislets using a 96-well ultra-low attachment plate and a microwell culture plate. Figure 2a shows sequential changes in morphology during the formation of pseudoislets from 3 x 103 human islet cells in a 96-well ultra-low attachment plate. Monolayer or loose clumps of cells observed in day 1 changed into solid aggregates with a smooth, round border by day 5 to 7 (Figure 2a). In a microwell culture plate, the formation of solid pseudoislets is usually visible within 4 days (Figure 2b). When 600 cells/microwell were plated in the microwell culture plate, human islet cells were condensed into spheroids of uniform size. It is observed that a microwell culture plate allows the successful formation of pseudoislets from a small number of cells compared with a 96-well ultra-low attachment plate. Typically, over 1,500 cells/pseudoislet are needed for a 96-well ultra-low attachment plate, while 500 cells/pseudoislet are sufficient for a 24-well microwell culture plate. Successfully formed pseudoislets remain as spheroids after recovery from a 96 ultra-low attachment plate or a microwell culture plate and are compatible for downstream applications including static incubation (Figure 3a) and perifusion (Figure 3b). The uniform size of pseudoislets reduces the variation within a test group and allows static incubation using as little as 5 pseudoislets per measurement (Figure 3a). Also, human pseudoislets maintained robust first phase insulin secretion in response to glucose after 7 days of culture when the original human islets cultured for a similar period of time showed blunted first phase glucose-stimulated insulin secretion (Figure 3b)5. The introduction of lentivirus into a single cell suspension ensures the efficient and homogenous transduction of islet cells and achieves highly efficient down-regulation of genes as shown in Figure 3c. All results shown were obtained using human islets from non-diabetic donors.
Figure 1: Process of human pseudoislet preparation. (a) The suspension containing 4,000 IEQ of human islets becomes cloudy after digestion by a proteolytic and collagenolytic enzyme mixture and mild pipetting. (b) Human islets after dispersion are passed through a strainer. Undigested islets remaining on top of the strainer are dispersed using a 1 mL syringe plunger. (c,d) Microscope images of the single cell suspension containing 3,000 cells/well in a 96-well ultra-low attachment plate before (c) and after (d) centrifugation. (e,f) Microscope images of the single cell suspension containing 500 cells/microwell in a 24-well microwell culture plate before (e) and after (f) centrifugation. Scale bar = 250 µm. Please click here to view a larger version of this figure.
Figure 2: Morphology of human pseudoislets. Sequential changes in morphology of human pseudoislets created (a) in a 96-well ultra-low attachment plate from 3,000 cells and (b) in a 24-well microwell culture plate from 500 cells. Scale bar = 100 µm. Please click here to view a larger version of this figure.
Figure 3: Examples of functional assays using human pseudoislets. (a) Representative static incubation performed using human pseudoislets created in a microwell culture plate from a single donor at the size of 500 human islet cells per pseudoislet. Four sets of 5 pseudoislets were incubated for 1 h in Krebs-ringer bicarbonate buffer supplemented with either 2 mM or 16.8 mM glucose. Each symbol represents insulin secretion from one set of 5 pseudoislets. Mean ± standard error of the mean (SEM) is shown. *, p < 0.05 by student’s t test. Representative results from three donors. (b) Representative perifusion testing insulin secretion from human pseudoislets created in a microwell culture plate in response to 16.7 mM glucose and 30 mM KCl. Method for perifusion was previously published5. Mean ± SEM of insulin secretion from two sets of 40 pseudoislets created in a microwell culture plate from a single donor at the size of 500 human islet cells per pseudoislet plate is shown. Representative data from six donors. (c) Pseudoislets were created with lentivirus carrying shRNA targeting human ATGL (targeting CCTGCCACTCTATGAGCTTAA, left) or PLIN5 (targeting GACAAGCTGGAAGAGAAGCTT, right). Control pseudoislets were transduced with lentivirus expressing scrambled sequence (Scr) previously published5. The mRNA expression of each gene was determined by real time polymerase chain reaction (PCR) as previously published5. Data were expressed using 2–DDCT taking peptidylprolyl Isomerase B (PPIB) as an internal control17. Each dot represents data from each donor for an indicated primer and a data set from the same donor is connected by a line. N = 3 donors. *; p < 0.05 by student’s t test. Please click here to view a larger version of this figure.
Here, a detailed protocol to generate human pseudoislets that are transduced by lentivirus using a 96-well ultra-low attachment plate or a microwell culture plate is presented. Pseudoislets have been reported to demonstrate morphology and secretory functions similar to native human islets and can be cultured for prolonged time in vitro5,11,18. Unlike native human islets that show a wide variation in size, pseudoislets are relatively uniform in size, reducing variation between donors and experimental replicates5,11. Downregulation of genes requiring high efficiency of transduction can be performed easily prior to the formation of pseudoislets in single cell suspension. This method avoids the difficulty of viral penetration through layers of cells in intact islets. Thus, this simple, highly efficient, and reproducible protocol for the creation of human pseudoislets has wide applications.
While several different platforms have been reported for the formation of pseudoislets12,14,19,20, both a 96-well ultra-low attachment plate and a microwell culture plate are commercially available, allowing this technique to be adopted by any laboratory. Although the hanging drop method12 also allows the formation of human pseudoislets using common labware, potential limitations include the difficulty in controlling size and reaggregation duration of pseudoislets. These limitations were due to the limited volume per drop of pseudoislet and ongoing evaporation during the 5−7 day culture required for pseudoislet formation. Additionally, it is easier to contain lentivirus with the use of a 96 well ultra-low attachment plate or a microwell culture plate compared with the hanging-drop method.
Several steps within the protocol require close attention. Optimizing the digestion of intact islets with the proteolytic and collagenolytic enzyme mixture is critical since both under- and over-digestion will reduce yield of single cell suspension and subsequently affect aggregation of pseudoislets. During digestion, it is important to closely monitor islets for the disappearance of clumps and the increase in cloudiness as islets dissociate into single cells. It is important to note that the optimal time for digestion varies between islets from different donors. The optimal time is dependent on several factors including medical history and age of each donor, the length of ischemia time, the islet isolation procedure used, islet size, islet purity, islet viability, and shipping conditions. Typically, human islets with viability and purity higher than 80% and within 5 days of isolation are used. Careful and gentle pipetting during dispersion is also important to maintain cell viability and recovery that will ultimately affect cell aggregation and the final size of pseudoislets being formed. When dispensing islet cell suspension to wells (steps 3.1.4 and 3.2.7), gentle and thorough mixing of cells is important to achieve an even distribution of single cells into microwells. If cells form clumps after 1 h incubation with lentivirus, it requires gentle pipetting to break the clumps into single cell suspension prior to the final centrifugation.
We have had similar success in creating human pseudoislets after lentiviral transduction using both a 96-well plate and microwell culture plate. The choice between the two platforms depends on the size and number of pseudoislets desired. A microwell culture plate has small, pyramid shaped bottoms allowing condensation of a smaller number of cells compared with a 96 well round bottom plate. Thus, the number of cells per pseudoislet can be reduced for a microwell culture plate. Also, a single centrifugation step creates all pseudoislets simultaneously in a microwell culture plate while multiple pipetting is required for creating pseudoislets in a 96-well plate. Thus, scaling up the creation of pseudoislets is easier in a microwell culture plate. However, the currently available microwell culture plate does not offer flexibility in the number of pseudoislets being created. Currently, the minimum number of pseudoislets created using a microwell culture plate is 1,200 and can be increased only by the factor of 1,200. Thus, we typically use a 96-well plate for small scale pilot experiments and for an experiment in which small quantity of samples is sufficient such as an insulin secretion assay and RNA extraction for gene expression. We have used pseudoislets from a microwell culture plate for assays that require large numbers of cells such as Western blot, oxygen consumption rate determination by a metabolic analyzer, and triglyceride extraction.
The major limiting factor for the generation of human pseudoislets is the loss of cells during the preparation of single cell suspension. While 1 IEQ of human islet is considered to contain around 2,000 cells, the recovery of single cell suspension is typically 30% or lower due to multiple washing and passing through a strainer. Heterogeneity of islet size also makes it difficult to dissociate all islets simultaneously. While gentle pipetting and the use of the proteolytic and collagenolytic enzyme mixture in the protocol are efforts to combine mechanical and enzymatic forces for maximum recovery of single cells, there still is an inevitable loss of cells. Thus, the application of pseudoislets requires clear justification over studying intact human islets. It also needs to be reminded that insulin secretion from pseudoislets is more robust than islets cultured for the same period of the time but tends to be lower compared with freshly isolated islets5,11.
Although limitations exist, stable and highly efficient gene silencing combined with better preservation of glucose-stimulated insulin secretion for prolonged time in culture enable the assessment of gene function in human islet cells. Additionally, the complex intercellular communication between beta-beta and beta-non beta cells is proposed to have a regulatory role in islet function. However, there is currently limited information regarding intercellular communication within human islets. With increased availability of cell specific markers21, it is feasible to create human pseudoislets with defined cell composition, as recently reported in mouse islet cells22, which will facilitate improved understanding of the cell-cell communication between human islet cells. The recent advancement of imaging of three-dimensional tissues also potentially increases the utility of human pseudoislets as a model to unmask how cellular polarity and intercellular communication are regulated in human islets. Thus, human pseudoislets provide a useful model to dissect the functions of genes of interest and other questions in the field of islet biology.
The authors have nothing to disclose.
This work was financially supported by National Institutes of Health to Y.I. (R01-DK090490) and American Diabetes Association to Y.I. (1-17-IBS-132). J.A. and Y.I. are supported by the Fraternal Order of Eagles Diabetes Research Center. A.B. is supported by a National Institutes of Health training grant (T32NS45549). Authors utilized human pancreatic islets provided by the NIDDK-funded Integrated Islet Distribution Program (IIDP) at City of Hope (2UC4DK098085).
Anti-adherence rinsing solution | Stemcell technologies | 7919 | |
Biological safety cabinet | Thermo Scientific | 1300 Series Type A2 | |
cell strainer, 40 micrometer | Corning | 431750 | |
CMRL-1066 | ThermoFisher | 11530037 | |
CO2 incubator | Thermo Scientific | Heracell VIOS 160i | |
conical centrifuge tube, 15 mL | VWR | 89039-666 | |
conical centrifuge tube, 50 mL | VWR | 89039-658 | |
fetal bovine serum | ThermoFisher | 26140079 | |
guanidinium thiocyanate RNA extraction reagent | ThermoFisher | 15596026 | Trizol |
glutamine | ThermoFisher | 25030164 | |
Hemocytometer | Marien Feld | Neubauer-Improved Bright line | |
Human serum albumin | Sigma | A1653 | |
inverted microscope | Fisher brand | 11-350-119 | |
microcentrifuge | Beckman Coulter | Microfuge 20 | |
microcentrifuge tube, 1.5 mL | USA Scientific | 1615-5500 | |
microwell culture plate | Stemcell technologies | 34411 | Aggrewell 400, 24 well |
motor-driven pestle | GAMUT | #399X644 | |
non-tissue culture treated dish, 10 cm | Fisher Scientific | FB0875713 | |
PBS | ThermoFisher | 14190250 | |
Penicillin-streptomycin | ThermoFisher | 10378016 | |
Petri dish, 35 mm | Celltreat | 229638 | |
pipette, 5 mL | DOT Scientific, | 667205B | |
pipette, 8-channel | VWR | #613-5253 | |
pipette, 10 mL | VWR | 667210B | |
pipette, P10 | Denville | UEZ-P-10 | |
pipette, P200 | Denville | UEZ-P-200 | |
pipette, P1000 | Denville | UEZ-P-1000 | |
proteolytic and collagenolytic enzyme mixture | Sigma | A6965 | Accutase |
reagent reservoir, 50 mL | VWR | 89094-680 | |
reversible strainer, 37 micrometer | Stemcell technologies | 27251 | |
swing bucket plate centrifuge | Beckman Coulter | Allegra X-14R | |
swing bucket rotor | Beckman Coulter | SX4750A | |
tuberculin syringe, 1 mL | BD | 309659 | |
ultra low attachment microplate, 96 well | Corning | 4515 |