Targeting Pancreatic Islet NLRP3 Improves Islet Graft Revascularization

Islet transplantation represents a promising therapeutic approach to improve glycometabolic control in patients with type 1 diabetes and in patients with diabetes suffering from chronic pancreatitis or after pancreatectomy (1). However, insufficient revascularization and, thus, hypoxia-induced inflammatory failure of grafted islets are a major obstacles preventing long-term graft survival (2–7).

The role of inflammasomes in inflammation is incontrovertible, and their activation has been recently described during ischemia-induced injury (8,9). Nucleotide-binding oligomerization domain–like receptor protein 3 (NLRP3) is the best-characterized inflammasome with respect to its pathophysiological roles and clinical implications (10). It consists of the sensor NLRP3, caspase-1, and the apoptosis-associated speck-like protein containing a CARD adapter protein (10). Proximity-induced conformational changes of caspase-1, triggered by its recruitment to the inflammasome complex, results in processing of pro-interleukin-1β (pro-IL-1β) and pro-IL-18 to the mature and biologically active forms IL-1β and IL-18 (11). The NLRP3 inflammasome is mainly expressed by innate immune effector cells and can be activated by various exogenous pathogen-associated molecular patterns or endogenous damage-associated molecular patterns (10). We have recently identified apolipoprotein C3 as a novel mediator leading to alternative inflammasome activation in human monocytes (12). On the other hand, NLPR3 is capable of interacting with the thioredoxin-interacting protein (TXNIP) in a reactive oxygen species–dependent manner, which mediates inflammasome activation (13).

Interestingly, NLRP3 is also expressed in pancreatic islets and even upregulated under hypoxia (13–15). In addition, Sokolova et al. (16) reported that a global knockout of NLRP3 promotes insulin secretion in isolated islets by a still unknown mechanism. Accordingly, the inhibition of NLRP3 in isolated islets before transplantation may be sufficient to improve islet transplantation.

Therefore, the aim of the current study was to analyze whether the genetic Nlrp3 deficiency or the exposure of isolated islets to the selective NLRP3 inhibitor CY-09 improves graft revascularization in an insulin-dependent manner. Moreover, we elucidated the underlying regulatory mechanism of the increased insulin secretion in Nlrp3-deficient islets.

Primary human umbilical vein endothelial cells (HUVECs) were cultivated in endothelial cell basal medium (100 units/mL penicillin and 0.1 mg/mL streptomycin) at 37°C under a humidified 95:5% (v/v) mixture of air and CO₂. The cells were passaged at a split ratio of 1:3 after reaching confluence.

The mouse pancreatic β-cell line MIN6 was cultivated in DMEM (10% [v/v] FCS, 100 units/mL penicillin, and 0.1 mg/mL streptomycin) at 37°C under a humidified 95:5% (v/v) mixture of air and CO₂. Cells were passaged at a split ratio of 1:3 after reaching confluence.

MIN6 cells were exposed to 10 μmol/L CY-09 or DMSO (vehicle) and cultivated in DMEM (1 g/L glucose) under hypoxic conditions (95% N₂, 5% CO₂, and 1% O₂) for 24 h. Isolated wild-type (WT), Nlrp3^−/−, and WT islets exposed to 10 μmol/L CY-09 or vehicle were cultivated in DMEM (1 g/L glucose) under hypoxic conditions (95% N₂, 5% CO₂, and 1% O₂) for 16 h.

Isolated WT and Nlrp3^−/− islets and MIN6 cells exposed to CY-09 (10 μmol/L) or vehicle for 24 h were harvested, and whole-cell, cytoplasmic, and nuclear extracts were generated as previously described in detail (17). Further details are provided in the Supplementary Material.

Hypoxic MIN6 cells exposed to CY-09 (10 μmol/L) or vehicle for 24 h were harvested, and whole-cell extracts were prepared. Coimmunoprecipitation was performed according to the manufacturer’s protocol (Pierce Crosslink IP Kit, Thermo Fisher Scientific). Bound proteins were eluted and separated through 12.5% SDS-PAGE, transferred to a polyvinylidene difluoride membrane, and analyzed by Western blot.

For the tube formation assay, HUVECs were seeded in a 96-well plate (1.5 × 10⁴ cells/well), which contained 50 μL Matrigel per well. The cells were exposed to insulin (1.5 μmol/L), CY-09 (10 μmol/L), linsitinib (800 nmol/L), or vehicle. Phase-contrast light micrographs were taken after 24 h. Tube formation was quantified by measuring the number of tube meshes (i.e., areas completely surrounded by endothelial tubes) per high-power field using ImageJ software (National Institutes of Health, Bethesda, MD).

Total RNA from WT, Nlrp3^−/−, and WT islets exposed to 10 μmol/L CY-09 or vehicle and MIN6 cells exposed to 10 μmol/L CY-09 or vehicle for 24 h was isolated using QIAzol Lysis Reagent (QIAGEN). Further details are provided in the Supplementary Material.

MIN6 cells were seeded on coverslips and incubated for 24 h with vehicle or 10 μmol/L CY-09. Moreover, WT and Nlrp3^−/− islets were seeded on coverslips. The cells or islets were loaded with 1 μmol/L Fura-2 AM for 40 min in Krebs-Ringer buffer without glucose. After incubation, the cells or islets were washed with PBS twice and imaged using excitation 340/380 nm and emission 505 nm. Further details are provided in the Supplementary Material.

Isolated islets cultivated for 24 h in 0.2% H₂O₂ (positive control) were incubated for 20 min at 37°C with calcein (1 μg/mL) and propidium iodide (5 μg/mL). Cell nuclei were stained for 10 min at 37°C with Hoechst 33342 (2 μg/mL). The islets were washed with PBS, resuspended in glycerin gelatin, and covered on slides. The cellular stainings were visualized by using fluorescence microscopy 15 min after mounting.

Animals were maintained on a standard 12-h day/night cycle. Water and standard pellet chow (Altromin, Lage, Germany) were provided ad libitum. Transgenic Nlrp3^−/− (B6.129S6-Nlrp3^tm1Bhk/J) and Casp1^−/− (B6N.129S2-Casp1^tm1Flv/J) mice, which we have already used in a previous study (12), were purchased from The Jackson Laboratory. C57BL/6J WT, Nlrp3^−/−, and Casp1^−/− mice aged 8–10 weeks and with a body weight of 25–30 g served as donors for islet isolation and aortic ring assays. Male C57BL/6J WT mice with a body weight of 30–35 g served as donors for microvascular fraction (MVF) isolation. C57BL/6J WT mice with a body weight of 23–25 g were used for the dorsal skinfold chamber model. Diabetes was induced in 6–8-week-old male C57BL/6J WT mice.

All experiments were performed according to German legislation on the protection of animals and the National Institutes of Health Guide for the Care and Use of Laboratory Animals. The experiments were approved by the local governmental animal protection committee (permission nos. 45/2018 and 06/2020).

Mice were anesthetized by intraperitoneal injection of ketamine (80 mg/kg body weight) and xylazine (12 mg/kg body weight). Following cervical dislocation and midline laparotomy, the pancreatic duct was injected with 1 mg/mL collagenase NB 8 containing 25 μL/mL neutral red solution, and pancreatic islets were isolated as previously described in detail (18). Isolated islets were cultivated in DMEM (supplemented with 10% [v/v] FCS, 100 units/mL penicillin, and 0.1 mg/mL streptomycin) for 24 h at 37°C and 5% CO₂ for further experiments.

Mice were anesthetized and euthanized by cervical dislocation. Subsequently, MVFs were isolated by mechanic and enzymatic digestion (collagenase NB 4G) of epididymal fat pads of mice, as previously described in detail (19). Further details are provided in the Supplementary Material.

Mice were anesthetized and euthanized by cervical dislocation. Subsequently, aortic rings of WT and Nlrp3^−/− mice were excised and embedded in Matrigel. After 1 h, the Matrigel was polymerized, and DMEM (10% FCS, 100 units/mL penicillin, and 0.1 mg/mL streptomycin) with or without insulin (1.5 μmol/L), linsitinib (800 nmol/L), and CY-09 (10 μmol/L) was added. The aortic rings were then incubated at 37°C for 6 days, including a medium change on day 3. The sprouting area of the rings was analyzed by phase-contrast microscopy.

Histological sections were performed as previously described in detail (20). Further details are provided in the Supplementary Material.

Isolated islets were incubated for 2 min at room temperature with neutral red or trypan blue and washed with PBS. The cellular stainings were visualized by bright field images.

Isolated islets were dispersed into single cells by accutase. Subsequently, the cells were washed in PBS, resuspended in incubation buffer, and stained for 15 min with propidium iodide and annexin V (100 μg/mL) according to the manufacturer’s protocol (Roche). The stained cells were analyzed by flow cytometry using FACScan (BD Biosciences), and the number of vital, apoptotic, necrotic, and necroptotic cells was given as the percentage of all measured cells.

The amount of secreted insulin was measured by intraperitoneal glucose tolerance test (IPGTT) as previously described in detail (20,21). Further details are provided in the Supplementary Material.

Mice were anesthetized, and the dorsal skinfold chamber was implanted, as previously described in detail (22). After 48 h of recovery, the mice were anesthetized. Thereafter, the cover glass was removed, and the tissue was washed with saline. Subsequently, eight isolated islets were transplanted onto the exposed striated muscle tissue. Finally, the chamber was sealed with a new cover slip for further intravital fluorescence microscopic analyses. Further details are provided in the Supplementary Material.

Diabetic phenotypes were induced by a single intraperitoneal injection of 180 mg/kg streptozotocin (STZ) 8 days before islet transplantation. Body weights and nonfasting blood glucose levels of STZ-induced diabetic mice were measured twice a week during the entire observation period of 28 days. Blood samples were taken from the tail vein and analyzed by a portable blood glucose monitoring system (GL50; Breuer). Mice with a nonfasting blood glucose level ≥350 mg/dL served as recipients for islet transplantation (23). Two hundred, 300, and 400 isolated islets were injected under the left kidney capsule of diabetic mice using a 10-μL Hamilton syringe, as previously described in detail (24). Normoglycemia was defined by blood glucose levels <200 mg/dL.

All in vitro experiments were reproduced at least three times. For in vivo studies, we used at least five animals per group, and no mice were excluded from the statistical analysis. After testing the data for normal distribution and equal variance, differences between two groups were assessed by unpaired Student t test. To test differences among multiple groups, one-way ANOVA was applied, followed by the Tukey post hoc test. GraphPad Prism 8 software was used for the analysis. All values are expressed as mean ± SEM. Statistical significance was accepted at P < 0.05.

The data sets generated and/or analyzed during the current study are available from the corresponding author upon reasonable request.

In a first set of experiments, we assessed the viability of isolated islets from WT and Nlrp3^−/− mice and of isolated WT islets exposed to vehicle (WT + vehicle) or the NLRP3 inhibitor CY-09 (WT + CY-09) for 24 h by means of calcein/propidium iodide staining (Fig. 1A) and neutral red/trypan blue staining (Supplementary Fig. 1). We did not detect any differences between the groups. Using flow cytometric analyses of propidium iodide/annexin V–stained islet cells, we confirmed that neither the Nlrp3 deficiency nor the pharmacological inhibition of NLRP3 triggered apoptosis or necrosis (Fig. 1B and C). Moreover, we studied the cellular composition of islets within the pancreas of WT and Nlrp3^−/− mice and isolated WT, Nlrp3^−/−, WT + vehicle, and WT + CY-09 islets by immunohistochemical assessment of the number of insulin-, glucagon-, somatostatin-, and CD31-positive cells (Fig. 1D and Supplementary Fig. 2A). The quantitative analyses revealed no differences between the groups (Fig. 1E and F and Supplementary Fig. 2B).

Next, we analyzed the revascularization of transplanted WT, Nlrp3^−/−, WT + vehicle, and WT + CY-09 islets by means of the dorsal skinfold chamber model in combination with intravital fluorescence microscopy (Fig. 2A). We found that the genetic abrogation of Nlrp3 or inhibition of NLRP3 before transplantation improved the take rate of the grafts (i.e., the number of engrafted islets in relation to the overall number of transplants per group) on day 14 (Fig. 2B and C). In line with these results, the application of FITC-labeled dextran revealed a higher revascularized area of Nlrp3^−/− and WT + CY-09 islets over time compared with their corresponding controls (Fig. 2D–F). The additional measurement of microhemodynamic parameters demonstrated that the loss or inhibition of NLRP3 does not markedly affect the diameter of intraislet blood vessels compared with controls (Supplementary Fig. 3A and B). However, the microvessels of transplanted Nlrp3^−/− and WT + CY-09 islets exhibited a higher centerline red blood cell velocity and volumetric blood flow (Supplementary Fig. 3C–F). This can be explained by the accelerated revascularization and, thus, more developed microvascular networks within transplanted Nlrp3^−/− and WT + CY-09 islets compared with corresponding controls at individual time points. We further detected a larger rhodamine 6G–positive area within Nlrp3^−/− and WT + CY-09 islets, indicating an increased endocrine tissue perfusion (25) (Fig. 2G–I).

To characterize the cellular composition of the islets on day 14 after transplantation, the expression of insulin, glucagon, somatostatin, and CD31 was assessed by immunohistochemistry (Fig. 3A–C). We did not observe any differences in the fractions of endocrine β-, α-, and δ-cells between Nlrp3^−/− and WT islets and between WT + vehicle and WT + CY-09 islets (Fig. 3B and C). However, the genetic deficiency of Nlrp3 or inhibition of NLRP3 resulted in a significantly higher fraction of intraislet CD31⁺ endothelial cells compared with controls (Fig. 3B and C). Furthermore, the fraction of myeloperoxidase-positive neutrophils, CD68⁺ macrophages, and CD3⁺ lymphocytes was analyzed within the grafts on day 14 (Fig. 3D). We detected neutrophils and macrophages but not lymphocytes within the transplanted islets. Of interest, loss or inhibition of NLRP3 significantly lowered the fraction of these cell types (Fig. 3E and F).

Vascular endothelial growth factor A (VEGF-A) is expressed in β-cells and plays a crucial role in the development of islet vascularization (26). Moreover, NLRP3 and VEGF-A are thought to be regulated in a reciprocal fashion (27). Therefore, we determined VEGFA gene expression in hypoxic WT or Nlrp3^−/− islets and hypoxic WT islets exposed to vehicle or CY-09. We found that neither the genetic abrogation of Nlrp3 nor the pharmacological inhibition of NLRP3 affects VEGF-A expression (Fig. 4A and B).

We next analyzed insulin secretion from these islets because NLRP3 inhibition has been shown to improve insulin secretion (16). In fact, the loss or inhibition of NLRP3 in isolated islets promoted insulin secretion under normoxia and hypoxia (Fig. 4C and D). It is well known that insulin triggers angiogenic pathways in endothelial cells (28). Therefore, we hypothesized that the increased insulin secretion from Nlrp3^−/− or WT + CY-09 islets rather than the reduced NLRP3 activity in islet endothelial cells stimulates graft revascularization. To clarify this, we performed a panel of in vitro angiogenesis assays. As expected, insulin stimulated angiogenic tube and sprout formation of human endothelial cells, mouse aortic rings, and MVF spheroids (Fig. 4E–H and Supplementary Fig. 4A–C). The simultaneous inhibition of insulin receptor kinase and IGF-I receptor by linsitinib abolished the proangiogenic effect of insulin (Fig. 4E–H). Of note, neither the inhibition of NLRP3 in HUVECs nor genetic deficiency of Nlrp3 in cells of aortic rings triggered angiogenic stimulation or prevented an insulin-induced angiogenic response (Fig. 4E and H). These results support our hypothesis that the elevated insulin secretion due to a decreased NLRP3 activity improves the revascularization of transplanted islets.

The increased insulin release after loss or inhibition of NLRP3 may be caused by an enhanced Ca²⁺ influx and/or an upregulated insulin gene expression. To test this, isolated WT and Nlrp3^−/− islets were cultivated under hypoxia for 16 h. Subsequently, K⁺ and glucose-dependent Ca²⁺ influx were determined. The intracellular Ca²⁺ levels did not differ between the two groups (Fig. 5A and B). However, an increased expression of the insulin genes Ins1 and Ins2 was detected in Nlrp3^−/− islets compared with WT islets (Fig. 5C and D). We also verified both experiments in hypoxic MIN6 cells exposed to vehicle or CY-09 (Supplementary Fig. 5A–D).

To elucidate the underlying regulatory mechanism of the upregulated insulin gene expression, we analyzed the subcellular localization of pancreatic and duodenal homeobox-1 (PDX-1) and MafA, the two major transcription factors of insulin. Immunohistochemical stainings revealed a higher number of PDX-1–positive nuclei in hypoxic Nlrp3^−/− and WT + CY-09 islets compared with controls (Fig. 5E–G). We further generated cytoplasmic and nuclear fractions from hypoxic WT and Nlrp3^−/− islets and hypoxic MIN6 cells exposed to vehicle or CY-09. As expected, the loss or inhibition of NLRP3 triggered the shuttling of PDX-1 and MafA into the nucleus (Fig. 5H–K and Supplementary Fig. 5E–H). The NLRP3 inflammasome interacts with the redox regulator TXNIP (13). In turn, TXNIP suppresses MafA-dependent insulin expression (29). Therefore, we next assessed the protein level of TXNIP. Our analysis showed a reduced protein level of this redox protein in hypoxic Nlrp3^−/− islets and hypoxic MIN6 cells exposed to CY-09 (Fig. 5L and M and Supplementary Fig. 5I and J). Additional coimmunoprecipitation experiments demonstrated that the inhibition of NLRP3 lowers the interaction of TXNIP with NLRP3 (Fig. 5N). This can be explained by the fact that pharmacological NLRP3 inhibition results in a decreased NLRP3 protein expression (30).

Finally, we analyzed whether loss or inhibition of NLRP3 in transplanted islets also results in an improved restoration of normoglycemia in diabetic animals. For this purpose, islets were transplanted under the kidney capsule of STZ-induced diabetic mice, and blood glucose levels and body weights were determined over 28 days (Fig. 6A). Nondiabetic animals served as negative controls. First, we determined the marginal mass of WT islets, which does not restore normoglycemia in diabetic animals. We found that the transplantation of 200 and 300 islets does not reverse hyperglycemia, whereas the transplantation of 400 islets restores normoglycemia (Supplementary Fig. 6A–D). Therefore, 300 islets were used as the marginal islet mass in the current study.

We detected significantly lower blood glucose levels of mice transplanted with Nlrp3^−/− islets or WT + CY-09 islets a few days after transplantation compared with controls (Fig. 6B and F). More importantly, we measured physiological blood glucose levels in both groups on day 28, whereas the transplantation of WT islets and WT + vehicle islets did not restore normoglycemia during the entire observation period (Fig. 6B and F). Accordingly, the area under the curve of both groups was significantly reduced compared with controls (Fig. 6C and G). The body weights of the animals did not differ between the groups over the 28-day observation period (Supplementary Fig. 7A–D). An IPGTT demonstrated that the blood glucose levels of mice transplanted with Nlrp3^−/− and WT + CY-09 islets were lower compared with control animals (Fig. 6D, E, H, and I). As expected, we also found higher plasma insulin levels in mice receiving Nlrp3^−/− and WT + CY-09 islets 15 min after glucose injection (Supplementary Fig. 8).

To further analyze whether the observed effects are influenced by the NLRP3 downstream effector caspase-1, we transplanted WT or Casp1^−/− islets under the kidney capsule of diabetic recipient mice. No differences were observed in plasma insulin levels and body weights between the groups (Supplementary Figs. 8 and 9A and B). The loss of Casp1 in transplanted islets slightly ameliorated hyperglycemia after transplantation compared control (Fig. 6J–M). However, we did not detect physiological blood glucose levels in the group of Casp1^−/− transplanted islets at the end of the observation period (Fig. 6J).

NLRP3 is the most studied inflammasome sensor because of its involvement in pathogen-driven and sterile inflammation (10). For instance, we found that apolipoprotein C3 is an endogenous mediator that leads to alternative NLRP3 inflammasome activation in human monocytes by forming a heterotrimer among toll-like receptors 2 and 4 and SLP65/SLP76, Csk-interacting membrane protein (12). Furthermore, we identified an intronic variant of Nlrp3 (rs10754555), which is associated with increased systemic inflammation, inflammasome activation, prevalent coronary artery disease, and mortality (31). The current study shows that Nlrp3 deficiency or inhibition of NLRP3 in isolated islets is capable of restoring normoglycemia in diabetic animals. Analyses of the underlying mechanism revealed that the suppression of NLRP3 promotes the revascularization process by upregulation of insulin gene expression via PDX-1 and MafA.

Several studies already reported a crucial role of NLRP3 in the endocrine function and stress-induced inflammatory response of β-cells (14,16,32). Recently, it has been shown that the expression of NLPR3 in grafted islets is particularly upregulated in the first days after transplantation (15). Accordingly, we speculated that a reduced activity of NLRP3 before transplantation may improve islet engraftment. In line with this view, we found a markedly accelerated revascularization of transplanted Nlrp3^−/− and WT + CY-09 islets, resulting in a higher take rate of the grafts compared with controls.

Chai et al. (33) reported that NLRP3 overexpression promotes the secretion of proangiogenic cytokines in human retinal endothelial cells. This is caused by the upregulation of hypoxia-inducible factor-1α/VEGF. On the basis of these results, it is conceivable that the inhibition of NLRP3 suppresses angiogenesis. However, we found no differences in the expression of VEGF-A. Sokolova et al. (16) reported that a reduced NLRP3 activity triggers insulin secretion and that insulin, in turn, promotes blood vessel formation via binding to endothelial insulin receptor kinase/IGF-I receptor (34–36). In line with these findings, we detected an enhanced insulin secretion in Nlrp3^−/− and WT + CY-09 islets. The stimulation of endothelial cells with insulin promoted angiogenic tube and sprout formation, whereas this effect was abolished by blocking insulin receptor kinase/IGF-I receptor signaling transduction. Of note, the decreased NLRP3 activity in endothelial cells did not affect their angiogenic activity. Therefore, we assume that an increased insulin level after NLRP3 inhibition is a potential mechanism that stimulates blood vessel formation.

It is well known that the revascularization of transplanted islets requires several days and involves angiogenesis and, to a lesser extent, vasculogenesis (37–40). Vasculogenesis is mediated by differentiation of endothelial precursor cells (EPCs) into endothelial cells (41). Deng et al. (42) demonstrated that inhibition of NLRP3 improves EPC function, as shown by a higher migratory capacity and ability to form tube-like structures in vitro. Moreover, NLRP3 is activated by oxidized LDLs in EPCs, which reduces their proliferation, migration, and tube formation (43). Although the presence of resident EPCs in isolated islets has still not been proven, it is conceivable that resident EPCs in isolated Nlrp3^−/− islets may contribute to the improved islet engraftment we observed.

Ca²⁺ influx through voltage-gated Ca²⁺ channels is a prerequisite for proper glucose-stimulated insulin secretion (44). We investigated whether the loss of NLRP3 affects Ca²⁺ signaling. Although Ca²⁺ release from the endoplasmic reticulum and influx of extracellular Ca²⁺ are both required to activate the NLRP3 inflammasome (45), we did not find differences between WT and Nlrp3^−/− islets in glucose-induced Ca²⁺ influx. However, the inhibition of the inflammasome significantly upregulated insulin gene expression. To gain further insights into the regulatory mechanism, we determined the expression of TXNIP because this cellular redox regulator is upregulated under hypoxia (15) and interacts with NLRP3 (13). Xu et al. (29) demonstrated that TXNIP induces the expression of miRNA-204, which, in turn, blocks insulin production by directly targeting and downregulating MafA (29). Our results show a decreased protein level of TXNIP and an increased nuclear localization of MafA in hypoxic Nlrp3^−/− islets. Of note, MafA is capable of activating PDX-1 promoter-driven reporter gene expression (46), and knockdown of TXNIP in β-cells increases PDX-1 gene expression (47). As expected, a strong nuclear signal of this transcription factor could be detected in Nlrp3^−/− islets under hypoxic conditions. Hence, it can be assumed that an increased nuclear localization of MafA caused by a reduced TXNIP expression promotes the shuttling of PDX-1 into the nucleus.

To investigate the mechanism of the reduced protein level of TXNIP, we analyzed the interaction of NLRP3 with TXNIP. Our results demonstrated that less NLRP3 is immunoprecipitated after CY-09 treatment, resulting in a reduced amount of coimmunoprecipitated TXNIP. This can be explained by the fact that the inhibition of NLRP3 leads to a significantly reduced NLRP3 expression in hypoxic β-cells, which is in line with the findings of Matsuoka et al. (30) showing that the suppression of NLRP3 activity in cytokine-exposed β-cells downregulates the gene expression of NLRP3. Therefore, it is tempting to speculate that inhibition of NLRP3 decreases its own gene expression by a negative feedback loop.

Several studies reported that 100–200 islets are capable of restoring normoglycemia in diabetic animals (48–50). However, this number crucially depends on the body weights of the recipients. Zmuda et al. (51) showed that the transplantation of 150–250 islets under the kidney capsule of diabetic mice with body weights of 18–20 g only marginally restores normoglycemia, whereas the transplantation of 300 islets reverses hyperglycemia. We herein used diabetic animals with a body weight of ∼25 g. Accordingly, a higher number of islets was required to restore physiological blood glucose levels. In fact, we demonstrated that the transplantation of 400 WT islets leads to normoglycemia, whereas 200 and 300 WT islets do not reverse hyperglycemia. On the basis of these findings, 300 islets were used as a marginal mass for islet transplantation. We found significantly lower blood glucose levels in STZ-induced diabetic mice receiving Nlrp3^−/− and WT + CY-09 islets, whereas the corresponding controls did not restore normoglycemia during the entire observation period of 28 days.

Hajmrle et al. (52) demonstrated that IL-1β potentiates insulin secretion from β-cells by increasing exocytosis subsequent to enhanced insulin granule docking at the plasma membrane. In contrast, IL-1β has also been reported to suppress insulin release (53,54). To analyze whether the improved restoration of normoglycemia following NLRP3 inhibition is mediated by caspase-1–dependent cleavage of pro-IL-1β to IL-1β, we additionally transplanted Casp1^−/− islets under the kidney capsule of diabetic recipients. Our results showed decreased blood glucose levels in mice receiving Casp1^−/− islets; however, we did not detect physiological blood glucose levels at the end of the observation period. This indicates that the enhanced islet transplantation after NLRP3 inhibition is caused not only by NLRP3 itself but also by its downstream effectors, such as caspase-1 and IL-1β.

Hypoxia induces inflammasome-driven inflammation and pyroptosis by NLRP3 activation in monocytes and macrophages (55). Islet-resident macrophages constitute the majority of leukocytes found in islets at steady state, and they can act as a potential modulator of T-cell activation (56). We found a markedly reduced number of macrophages and neutrophils in Nlrp3-deficient and WT + CY-09 islets compared with WT controls. Accordingly, it is conceivable that the diminished activity of NLRP3 in islet-resident macrophages, and thus their decreased release of IL-1β and IL-18, reduces the recruitment of neutrophils as well as infiltration of host macrophages. In line with this view, Matsuoka et al. (30) recently reported that the pretreatment of diabetic animals with the NLRP3 inhibitor MCC950 improves hepatic islet transplantation by reducing the number of infiltrated macrophages.

Isolated islets from human pancreata are commonly cultured before transplantation to 1) reduce isolation-induced cellular stress (57), 2) deplete the number of passenger leukocytes (58), and 3) reduce the transplanted tissue volume due to the elimination of acinar cells and/or dead cells (59). Accordingly, the Consortium of Islet Transplantation has proposed pretransplant cultivation as a standard procedure for clinical transplants (60). In the current study, we found that suppression of NLRP3 in isolated islets before transplantation markedly improves graft revascularization. Additional analyses of the underlying mechanism revealed that this could be due to the elevated insulin levels. Hence, the exposure of isolated islets to NLRP3 inhibitors during pretransplant cultivation may represent a promising strategy to improve clinical islet transplantation.

Acknowledgments. The authors thank Servier Medical Art for providing access to designed medical elements (https://smart.servier.com), supporting the generation of graphical items in this publication. The authors also thank Caroline Bickelmann and Ruth M. Nickels (Institute for Clinical and Experimental Surgery) for excellent assistance.

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. S.W., T.S., L.N., A.S.B., M.P., D.A., L.P.R., M.D.A.H., B.M.S., M.D.M., M.W.L., and E.A. analyzed data and interpreted the results. S.W., T.S., L.N., A.S.B., M.P., D.A., L.P.R., M.D.A.H., B.M.S., A.W., and E.A. performed the experiments. S.W., T.S., M.D.M., M.W.L., and E.A. designed the research and wrote the manuscript. M.W.L. and E.A. supervised the study. All authors read and approved the final version of the manuscript. E.A. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Prior Presentation. Parts of this study were presented in abstract form at the 24th Surgical Research Days, Leipzig, Germany, 2–3 September 2021.