The unique hypusine modification of eIF5A promotes islet beta cell inflammation and dysfunction in mice

Abstract

In both type 1 and type 2 diabetes, pancreatic islet dysfunction results in part from cytokine-mediated inflammation. The ubiquitous eukaryotic translation initiation factor 5A (eIF5A), which is the only protein to contain the amino acid hypusine, contributes to the production of proinflammatory cytokines. We therefore investigated whether eIF5A participates in the inflammatory cascade leading to islet dysfunction during the development of diabetes. As described herein, we found that eIF5A regulates iNOS levels and that eIF5A depletion as well as the inhibition of hypusination protects against glucose intolerance in inflammatory mouse models of diabetes. We observed that following knockdown of eIF5A expression, mice were resistant to beta cell loss and the development of hyperglycemia in the low-dose streptozotocin model of diabetes. The depletion of eIF5A led to impaired translation of iNOS-encoding mRNA within the islet. A role for the hypusine residue of eIF5A in islet inflammatory responses was suggested by the observation that inhibition of hypusine synthesis reduced translation of iNOS-encoding mRNA in rodent beta cells and human islets and protected mice against the development of glucose intolerance the low-dose streptozotocin model of diabetes. Further analysis revealed that hypusine is required in part for nuclear export of iNOS-encoding mRNA, a process that involved the export protein exportin1. These observations identify the hypusine modification of eIF5A as a potential therapeutic target for preserving islet function under inflammatory conditions.

Figures

Figure 1. STZ-induced hyperglycemia is partially blocked by IL-1Ra and by knockdown of eIF5A in immunocompetent mice.

(A) Schematic representation of the STZ, IL-1Ra, and siRNA injection protocol in male mice. (B) IPGTTs at day 7 in C57BL/6J male mice (n = 4 per group). *P < 0.05. (C) IPGTTs at day 7 in NOD/SCID/Il2rg-null male mice (n = 4 per group). *P < 0.05. (D) Scatter plot of individual fasting blood glucose levels at day 7 in C57BL/6J male mice. The mean fasting blood glucose for si-Control–injected mice is statistically different (P < 0.05) compared with untreated mice. Numbers over each set of symbols represent the mean fasting glucose level in mg/dl in each group. Individual symbols represent individual mice, and horizontal bars indicate the mean glucose level for each group. (E) IPGTTs at day 7 in C57BL/6J male mice (n = 13–17 per group). *P < 0.05.

Figure 2. Knockdown of eIF5A in mice preserves islet mass and attenuates iNOS induction following STZ treatment.

C57BL/6J male mice were injected with vehicle (untreated) or STZ plus siRNAs, per the protocol in Figure 1A. (A) Pancreata from treated mice were sectioned and stained for insulin (red) and counterstained with hematoxylin (blue). Sections from representative pancreata are shown at low (original magnification, ×100; upper panels) and high (original magnification, ×400; bottom panels) magnification. (B) β cell mass in treated mice (n = 3 mice per group). *P < 0.05. (C) At the end of the study, pancreata from treated mice were paraffin-embedded and stained for iNOS (red) and counterstained with hematoxylin (blue). Original magnification, ×630. (D) Immunoblots of islet extracts from treated mice following a single dose of STZ. Data are from pooled islets (n = 3 mice per group).

Figure 3. Knockdown of eIF5A causes relative preservation of islet function following cytokine exposure in vitro.

Male C57BL/6J mice were injected with vehicle (untreated) or siRNAs as indicated for 3 consecutive days and then euthanized for islet procurement. (A) Representative immunoblot of islet extract for actin (top panel) and eIF5A (middle panel). Islets from si-Control– and si-eIF5A–treated animals were pulsed for 4 hours with 3H-spermidine, and then extracts were subjected to electrophoresis and fluorography (bottom panel). (B) Quantitation of eIF5A protein levels (normalized to actin protein levels) from islets from injected mice (n = 3 per group). *P < 0.05. (C and D) Islets from injected mice were subjected to studies of (C) GSIS and (D) GSCa at the indicated glucose concentrations. *P < 0.05, compared with untreated islets. (E and F) Islets from injected mice were treated with a cocktail of cytokines (IL-1β, TNF-α, IFN-γ) for 4 hours, and then subjected to (E) GSIS and (F) GSCa at the indicated glucose concentrations. *P < 0.05, compared with untreated islets. The immunoblot in E shows islet extract for actin and eIF5A following cytokine exposure. 3H-eIF5AHyp, tritium-labeled eIF5A.

Figure 4. Cytokine signaling to Nos2 translation is impaired in islets depleted of eIF5A.

Male C57BL/6J mice were injected with vehicle (untreated) or siRNAs for 3 consecutive days and then euthanized for islet procurement. (A) Islets from injected mice were subjected to quantitative RT-PCR for the genes indicated. Data for each gene were normalized to Actb mRNA levels and then reported as expression relative to vehicle (untreated). (B) Islets from injected mice were exposed to cytokines for 4 hours and then subjected to quantitative RT-PCR for the genes indicated. Data for each gene were normalized to Actb mRNA levels and then reported as expression relative to non–cytokine-treated, vehicle injection. *P < 0.05. (C) Islets from injected mice were untreated or treated with cytokines for 4 hours and then subjected to quantitative RT-PCR for Nos2 mRNA. Data were normalized to Actb mRNA levels and then reported as expression relative to non–cytokine-treated, vehicle injection. (D) Representative iNOS and actin immunoblots of islet extracts from injected mice treated with cytokines for 4 hours. (A–C) Data are from 3 independent islet isolations.

Figure 5. Inhibition of hypusination impairs cytokine-induced Nos2 translation in INS-1 (832/13) β cells and human islets.

(A) Mouse islets and INS-1 cells that were treated with GC7 overnight were pulsed with 3H-spermidine for 4 hours and subjected to electrophoresis and fluorography. (B) Representative immunoblots of actin and eIF5A from INS-1 cell extracts following overnight treatment with GC7. The arrowhead identifies an upper band of decreasing intensity. (C) Representative immunoblot of iNOS and actin from INS-1 cell extracts. The bar graph shows quantitation of iNOS protein levels, normalized to actin levels (n = 3). (D) Nitrite levels in INS-1 cell medium (n = 3). (C and D) *P < 0.05, compared with the non-cytokine–treated sample. (E) Nos2 transcript levels in INS-1 cells. Data were normalized to Actb mRNA levels and are reported as fold induction relative to non-cytokine, non-GC7–treatment. Data are the mean ± SEM from 3 independent experiments. (F) Representative immunoblots of iNOS and actin from human islets. (G) Nos2 transcript levels in human islets. Data were normalized to Actb mRNA levels and are reported as fold induction relative to non-cytokine, non-GC7–treatment (n = 3 experiments from a single human islet donor). (H) Representative immunoblots for iNOS, actin, and eIF5A from INS-1 cells transfected with the siRNAs indicated.

Figure 6. Inhibition of hypusination preserves INS-1 β cell function during cytokine exposure.

INS-1 β cells were exposed to (A) no treatment, (B) 4-hour cytokine incubation, and (C) a combination of overnight 125 μM GC7 and 4-hour cytokine incubation and then loaded with Fura-2 dye and subjected to GSCa (left panels) and GSIS studies (right panels) at the indicated glucose concentrations. For the GSCa studies, traces of at least 8 individual cells from a total of 3 independent experiments are shown. GSIS data are from 3 independent experiments. *P < 0.05.

Figure 7. Nos2 mRNA nucleocytoplasmic shuttling in INS-1 β cells is dependent upon exportin1/CRM1 and hypusinated eIF5A.

(A) INS-1 β cells were transfected with GFP-eIF5A or GFP-eIF5A (K50A) mutant, and cellular extracts were immunoprecipitated with the indicated antibodies and then immunoblotted for GFP, exportin1/CRM1, and eIF5A. (B) INS-1 cells were treated for 3 hours with leptomycin B (Lep B), overnight with GC7, or untreated (NT) and then exposed to 4-hour cytokine treatment, followed by fractionation of cytoplasmic and nuclear fractions. RNA from these fractions was subjected to quantitative RT-PCR for Nos2, Actb, Nfkb1, and Gapdh mRNAs. Data are expressed as the ratio of mRNA in the cytoplasmic fraction to mRNA in the nuclear fraction (n = 3). *P < 0.05.

Figure 8. Nucleocytoplasmic shuttling of eIF5A in response to cytokines.

(A) INS-1 β cells were exposed to vehicle (untreated), GC7 overnight, or leptomycin B for 3 hours and then to 4-hour cytokine treatment (or not) as indicated. Cells were then fixed and stained for eIF5A and visualized by fluorescence microscopy at 488 nm. Representative images are shown. Original magnification, ×630. The cytoplasmic-to-nuclear ratios of eIF5A staining (graphs) were quantitated from cytokine-treated cells and untreated cells exposed to each inhibitor by measurement of spatial pixel intensity. (B) INS-1 cells were transfected with expression vectors encoding GFP fusions of either eIF5A or eIF5A (K50A) mutant and then visualized by fluorescence microscopy (488 nm). Representative images are shown. Original magnification, ×630. Quantitation of cytoplasmic-to-nuclear ratios is shown in the graphs. For the ratio graphs, a minimum of 10 cells from 3 different experiments were quantitated. *P < 0.05.

Figure 9. eIF5A specifically interacts with Nos2 mRNA in INS-1 β cells.

(A) Quantitative RT-PCR from INS-1 cells exposed to vehicle or GC7 overnight and then to 4-hour vehicle or cytokine treatment as indicated. Data were normalized to Actb mRNA and are expressed as fold induction relative to no treatment (n = 3). (B) RIP assay from INS-1 cells. INS-1 cells were exposed to vehicle or GC7 overnight as indicated and then to 4-hour cytokine treatment. Then, they were subjected to RIP assays using either the eIF5A antibody or an isotype-matched control antibody (FLAG-M2). The bar graph shows quantitative RT-PCR for the genes indicated (n = 3). Agarose gel electrophoresis for the genes indicated is shown. Data are expressed as percent recovery relative to input mRNA.

Figure 10. Inhibition of hypusination protects against low-dose STZ-induced hyperglycemia.

GC7 or control saline was administered to male C57BL/6J mice by either daily intraperitoneal injection or subcutaneous implanted osmotic pumps. Then, mice underwent 5 consecutive injections of low-dose STZ, as detailed in Figure 1A. (A and B) IPGTTs at day 7 in mice administered saline or GC7 via (A) intraperitoneal injection (n = 6–8) or (B) osmotic pump (n = 4). Data are significantly different (*P < 0.05) between all 3 groups in A and for STZ saline only in B. (C) Blood insulin levels at 0 and 30 minutes during the GTT shown in A (n = 4). *P < 0.05. (D) β cell mass in mice from A (n = 3 mice per group). (E) Representative images of islets from fixed and stained pancreata from mice in A that were stained for iNOS (red) and counterstained with hematoxylin (blue). Original magnification, ×630. (F) Serum levels of the indicated cytokines in mice from A and from STZ-treated C57BL/6J mice that were intraperitoneally injected with IL-1Ra (see Figure 1B). n = 3–4 animals per group.

Figure 11. Model for eIF5A control of Nos2 translation.

Signaling from cytokine receptors triggers the nuclear translocation of NF-κB, which activates transcription of the Nos2 gene. Nos2 transcripts are shuttled out of the nucleus in a CRM1- and eIF5A-dependent manner, then delivered to ribosomes, where translation occurs to form iNOS. Nitric oxide produced by iNOS leads to suppression of ATP generation and to the eventual inhibition of insulin release. The figure is designed to be descriptive of possible events that are occurring based on data from this and other studies, but it is not meant to be exhaustive or explicit, as many known factors involved in cytokine signaling, eIF5A action, and insulin release have been omitted for clarity. Hyp, hypusine.