Hypusinated eIF5A is expressed in the pancreas and spleen of individuals with type 1 and type 2 diabetes

Hypusinated eIF5A is expressed in the pancreas and spleen of individuals with type 1 and type 2... OPENACCESS The gene encoding eukaryotic initiation factor 5A (EIF5A) is found in diabetes-susceptibility Citation: Mastracci TL, Colvin SC, Padgett LR, loci in mouse and human. eIF5A is the only protein known to contain hypusine (hydroxypu- Mirmira RG (2020) Hypusinated eIF5A is expressed in the pancreas and spleen of individuals trescine lysine), a polyamine-derived amino acid formed post-translationally in a reaction with type 1 and type 2 diabetes. PLoS ONE 15(3): catalyzed by deoxyhypusine synthase (DHPS). Previous studies showed pharmacologic e0230627. https://doi.org/10.1371/journal. blockade of DHPS in type 1 diabetic NOD mice and type 2 diabetic db/db mice improved glu- pone.0230627 cose tolerance and preserved beta cell mass, which suggests that hypusinated eIF5A Editor: Cinzia Ciccacci, Unicamillus, Saint Camillus Hyp (eIF5A ) may play a role in diabetes pathogenesis by direct action on the beta cells and/or International University of Health Sciences, ITALY altering the adaptive or innate immune responses. To translate these findings to human, we Received: October 9, 2019 examined tissue from individuals with and without type 1 and type 2 diabetes to determine Accepted: March 4, 2020 Hyp Hyp the expression of eIF5A . We detected eIF5A in beta cells, exocrine cells and immune Hyp Published: March 24, 2020 cells; however, there was also unexpected enrichment of eIF5A in pancreatic polypep- Hyp tide-expressing PP cells. Interestingly, the presence of eIF5A co-expressing PP cells Copyright:© 2020 Mastracci et al. This is an open access article distributed under the terms of the was not enhanced with disease. These data identify new aspects of eIF5A biology and high- Creative Commons Attribution License, which light the need to examine human tissue to understand disease. permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability Statement: All relevant data are within the manuscript and its Supporting Introduction Information files. The mechanisms underlying the pathogeneses of type 1 diabetes (T1D) and type 2 diabetes Funding: TLM 5-CDA-2016-194-A-N Juvenile Diabetes Research Foundation https://www.jdrf. (T2D) involve the activation of systemic and local inflammatory pathways, leading to eventual org/ The funders had no role in study design, data dysfunction, de-differentiation and/or death of the beta cells in the pancreatic islet. Elucidating collection and analysis, decision to publish, or the molecular mechanisms driving the inflammatory response is applicable to the development preparation of the manuscript. RGM R01 DK60581 of therapies for both diseases. In addition, an urgent priority in T1D research is the discovery National Institute of Diabetes and Digestive and of biomarkers that can assist in the identification of individuals with pre-clinical disease so Kidney Diseases https://www.niddk.nih.gov/ The funders had no role in study design, data collection early preventative therapeutic interventions can be implemented. PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 1 / 17 PLOS ONE Hypusinated eIF5A in human diabetes and analysis, decision to publish, or preparation of Recently, our laboratories have been investigating the involvement of the hypusinated form the manuscript. of eukaryotic initiation factor 5A (eIF5A) in the development and progression of diabetes in mice. To date, eIF5A is the only known protein to contain hypusine (hydroxyputrescine Competing interests: The authors have declared that no competing interests exist. lysine) [1], which is a polyamine-derived amino acid. This post-translational modification, formed by the process of “hypusination” [2], is catalyzed through a multi-step reaction initi- ated by the rate-limiting enzyme deoxyhypusine synthase (DHPS) and uses the polyamine spermidine as a cofactor to modify the Lys50 of eIF5A [2]. Previous studies using human cell Hyp lines and yeast determined that eIF5A, the hypusinated form of eIF5A (eIF5A ) and DHPS are vital for cell viability and proliferation [3,4]. Evolutionarily, eIF5A is highly conserved including the amino acid sequence surrounding the hypusine residue, which suggests an important role for this modification [5]. Whereas studies across species have established that Hyp eIF5A efficiently binds the ribosome complex and facilitates mRNA translation [3,6,7], the Hyp exact function of eIF5A and eIF5A remains unknown. Interestingly, the gene encoding eIF5A is found in the Idd4 diabetes-susceptibility locus in Hyp non-obese diabetes (NOD) mice [8,9]. In prior studies, we showed that eIF5A is expressed in the pancreatic islets of mouse [10,11], is responsible for the translation of a subset of cyto- Hyp kine-induced transcripts in beta cells in mouse models of diabetes [12,13], and that eIF5A also appears to be required for the activation and proliferation of effector T helper cells [14]. Moreover, reducing the hypusination of eIF5A in NOD mice, a model of T1D, by pharmaco- logical inhibition of DHPS resulted in reduced insulitis, improved glucose tolerance and pre- served beta cell mass [14]. Similarly, pharmacological blockade of DHPS in db/db mice [15], a model of T2D improved glucose tolerance and enhanced beta cell mass [16]. Together these Hyp data suggest that eIF5A may play a role in the pathogenesis of diabetes such that altering Hyp the expression of eIF5A may improve diabetes outcomes long-term. Hyp To translate these findings to human, a greater understanding of eIF5A in the human Hyp pancreas and spleen is required. In particular, determining the expression pattern of eIF5A Hyp in human tissues and whether eIF5A -expressing cells stratify with characteristics of disease would be informative. In this study, we used human donor tissue samples from the Network of Pancreatic Organ Donors with Diabetes (nPOD) (www.jdrfnpod.org) to examine the expres- Hyp sion pattern of eIF5A in the human pancreas and spleen from individuals with T1D, T2D and non-diabetic controls. Materials and methods Pancreas and isolated islet cells from mouse All mice were purchased from the Jackson Laboratory and maintained under a protocol approved by the Indiana University School of Medicine Institutional Animal Care and Use Committee (OLAW Assurance Number: D16-00584 (A4091-01); USDA Certificate Number: 32-R-0025; Customer ID: 798; AAALAC Unit Number: 00083; Protocol Approval #19043). The approved method of euthanasia for mice was carbon dioxide inhalation overdose deliv- ered using a gas cylinder, flow meter/regulator, and induction chamber; 100% CO2 delivered at a rate such that 20–30% of the volume of the chamber is displaced per minute. This was fol- lowed by the secondary method of cervical dislocation. Human donor tissues were collected and provided to the investigator by the Network of Pancreatic Organ Donors with Diabetes (https://www.jdrfnpod.org). The study was approved by the University of Florida Institutional Review Board (IRB-1); approval #IRB201600029. Written consent was obtained. Total pancreas and isolated islets from wildtype C57BL/6 mice as well as acinar tissue and isolated islets from human donors (Table 1 details human donors) were subjected to immuno- Hyp blot analysis as previously described [17]. Rabbit anti-eIF5A ([13,18]; 1:1000), mouse anti- PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 2 / 17 PLOS ONE Hypusinated eIF5A in human diabetes Table 1. Human donor information for islet and acinar tissue preparations. Islet Preparation Acinar Preparation Unique identifier SAMN11578698 UNOS AGDB487 Donor Age (years) 57.0 24 Donor Sex (M/F) M M Donor BMI (kg/m ) 25.8 31.7 Donor HbA1c 5.7 not tested Origin/source of islets IIDP (Integrated Islet Distribution CORE (Center for Organ Recovery and Program) Education) Islet isolation center The Scharp-Lacy Research Institute Pittsburgh (AHN) Donor history of diabetes? Yes/ No No No If Yes, complete the next two lines if this information is available Diabetes duration (years) Not Applicable Not Applicable Glucose-lowering therapy at time Not Applicable Not Applicable of death https://doi.org/10.1371/journal.pone.0230627.t001 eIF5A (BD Biosciences; 1:2000) and guinea pig anti-insulin (DAKO; 1:5000) antibodies were used to confirm protein expression in the pancreas. Mice containing the RIP-cre allele (B6.CG-Tg(Ins2-cre)25Mgn/J) [19] were mated with Tomato tm14(CAG-tdTomato)Hze/J those containing the R26R allele (B6.Cg-Gt(ROSA)26Sor ) [20] to produce double transgenic animals wherein all the insulin-producing beta cells in the pancreas expressed a fluorescent (Tomato) reporter. Pancreatic islets were isolated from RIP-cre; Tomato Tomato R26R mice and R26R mice as previously described [21]. The isolated islets from all mice were pooled together and processed for fluorescence activated cell sorting (FACS), which facilitated the separation of islet cells into two populations: Tomato-positive beta cells and Tomato-negative non-beta cells (which included islet cells expressing glucagon, somatostatin, ghrelin, and pancreatic polypeptide). Pooled islets were washed with sterile PBS (Fisher Scien- tific) and incubated in Accutase cell detachment solution (Sigma) for 10 minutes at 37C with constant mixing (1000 rpm). Islet cells were removed from the Accutase solution by centrifu- gation (500 x g for 1 min) and resuspended in cold buffer containing 2% BSA,1uM EDTA, and equal parts PBS and HBSS (Fisher Scientific). The cells were filtered, collected and incubated with APC viability dye (Zombie NIR-IR dye; BioLegend) per the manufacturer’s recom- Tomato Tomato mended protocol. Single-cell suspensions from RIP-cre;R26R mice and R26R were then sorted using an iCyt Reflection with 100 μm nozzle at 23 psi. Dead cells (NIR-IR+) were excluded; Tomato(+) cells and Tomato(-) cells were collected into tubes containing sort buffer. Data were analyzed using FlowJo software (Tree Star). Lysates from the two populations of Hyp cells were subjected to immunoblot analysis. Rabbit anti-eIF5A and mouse anti-eIF5A anti- Hyp bodies were used as above, to evaluate the abundance of eIF5A in the beta cell and non- beta cell populations. Rabbit anti-Pdx1 (Chemicon; 1:1000) antibody was used to evaluate enrichment of the beta cell (tomato+) population. Mouse pancreas tissue and immunofluorescence analysis Pancreas tissue was harvested from wildtype C57BL/6 mice and fixed in 4% paraformaldehyde (Fisher Scientific), cryo-preserved using 30% sucrose, embedded in OCT (Fisher Scientific) and sectioned onto glass microscope slides. Methods previously described for pancreas preser- vation and immunofluorescence were followed [22]. Pancreas tissue sections (8 μm) were stained using the following primary antibodies: guinea pig anti-insulin (DAKO; 1:500), goat PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 3 / 17 PLOS ONE Hypusinated eIF5A in human diabetes Hyp anti-pancreatic polypeptide (abcam; 1:200), rabbit anti-eIF5A ([13,18]; 1:1000). Secondary antibodies including Alexa-488, Cy3, or Alexa-647 (Jackson Immunoresearch) were used, fol- lowed by DAPI (Sigma; 1:1000) to visualize nuclei. Images were acquired with a Zeiss 710 con- focal microscope. Human pancreas and spleen tissue Paraffin-embedded tissue sections were obtained from the nPOD consortium (www.jdrfnpod. org). A total of 10 nondiabetic donors, 4 donors with T2D, and 12 donors with T1D (6 autoan- tibody positive, 6 autoantibody negative) were included in this study (Tables 2 and 3). Infor- mation regarding donors’ demography, histology, and disease status were provided by nPOD. The autoantibody status was also determined by nPOD as previously described [23]. Immunofluorescence analysis of human tissues Immunofluorescent staining was performed as previously published [22] with modifications to account for the use of paraffin embedded tissue. Briefly, tissue sections were deparaffinized through graded ethanols (100%, 95%, 85%, 75%, 50%; Fisher Scientific) and then blocked using normal donkey serum (Sigma). Primary antibodies used included guinea pig anti-insulin (DAKO; 1:500), mouse anti-glucagon (Abcam; 1:500), rat anti-somatostatin (abcam; 1:200), goat anti-pancreatic polypeptide (abcam; 1:200), goat anti-ghrelin (Santa Cruz; 1:500), mouse anti-Pax5 (DAKO; 1:200), mouse anti-CD8 (Thermo Fisher; 1:500), mouse anti-CD4 (Leica; Hyp 1:500), rabbit anti-eIF5A ([13,18]; 1:1000). Secondary antibodies including Alexa-488, Cy3, or Alexa-647 (Jackson Immunoresearch) were used to visualize primary antibodies. DAPI (Sigma; 1:1000) was used to visualize nuclei. Images were acquired with a Zeiss 710 confocal microscope. Drug treatments of cultured cells and mouse islets HEK293T cells (ATCC #CRL-3216) were maintained in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% FBS (Hyclone; Fisher Scientific), 1% Penicillin/Streptomy- cin and 2 mM L-Glutamine. HEK293T cells were grown to 60–70% confluency on coverslips in a 24-well plate and treated with 10 or 100 μM N1-Guanyl-1,7-diaminoheptane (GC7) [24] (or 10 mM acetic acid; vehicle) with 0.5 mM aminoguanidine for 16 hours. Cells were fixed with 4% paraformaldehyde in PBS and blocked for 30 min in 3% bovine serum albumin (BSA) in PBS followed by permeabilization with 0.2% Triton X-100 in BSA-PBS for 10 min. All anti- bodies were diluted in 3% BSA-PBS and were applied in sequential order. Cells were incubated Hyp with rabbit anti-eIF5A ([13,18], 1:1000) overnight at 4˚C followed by a one hour incubation with anti-rabbit 488 (Jackson Immunoresearch; 1:500) at room temperature. Coverslips were Table 2. Human donor pancreas and spleen tissue from T2D and matched controls. nPOD case # Sample name Age (years) Gender (male/female) Ethnicity BMI T2D (yes/no) C-peptide (ng/mL) HbA1c nPOD-6097 F1-control 43.1 female Caucasian 36.4 no 16.76 7.1 nPOD-6102 F2-control 45.1 female Caucasian 35.1 no 0.55 6.1 nPOD-6132 F1-T2D 55.8 female Hispanic 44.6 yes 0.8 9.1 nPOD-6109 F2-T2D 48.8 female Hispanic 32.5 yes <0.05 8 nPOD-6060 M1-control 24 male Caucasian 32.7 no 13.63 N/A nPOD-6091 M2-control 27.1 male Caucasian 35.6 no 7.71 6.3 nPOD-6114 M1-T2D 42.8 male Caucasian 31 yes 0.58 7.8 nPOD-6188 M2-T2D 36.1 male Hispanic 30.6 yes 3.45 7.2 https://doi.org/10.1371/journal.pone.0230627.t002 PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 4 / 17 PLOS ONE Hypusinated eIF5A in human diabetes Table 3. Human donor pancreas and spleen tissue from T1D and matched controls. nPOD case Sample name Age Gender (male/ Ethnicity BMI T1D (yes/ c-peptide Auto Antibodies Detected Duration of disease # (years) female) no) (years) nPOD-6179 1-control 21.8 female Caucasian 20.7 no 2.74 N/A N/A nPOD-6224 1-AAb- 21 female Caucasian 22.8 yes <0.05 negative 1.5 negative nPOD-6070 1-AAb- 22.6 female Caucasian 21.6 yes <0.05 mIAA+;1A-2A+ 7 positive nPOD-6034 2-control 32 female Caucasian 25.2 no 3.15 N/A N/A nPOD-6121 2-AAb- 33.9 female Caucasian 18 yes 0.24 negative 4 negative nPOD-6143 2-AAb- 32.6 female Caucasian 26.1 yes <0.05 1A-2A+;mIAA+ 7 positive nPOD-6229 3-control 31 female Caucasian 26.9 no 6.23 N/A N/A nPOD-6208 3-AAb- 32 female Caucasian 23.4 yes <0.05 negative 16 negative nPOD-6077 3-AAb- 32.9 female Caucasian 22 yes <0.05 mIAA+ 18 positive nPOD-6015 4-control 39 female Caucasian 32.2 no 1.99 N/A N/A nPOD-6038 4-AAb- 37.2 female Caucasian 30.9 yes 0.2 negative 20 negative nPOD-6054 4-AAb- 35.1 female Caucasian 30.4 yes <0.05 mIAA+ 30 positive nPOD-6055 5-control 27 male Caucasian 22.7 no 0.59 N/A N/A nPOD-6041 5-AAb- 26.3 male Caucasian 28.4 yes <0.05 negative 10 negative nPOD-6180 5-AAb- 27.1 male Caucasian 25.9 yes <0.05 GADA+;1A-2A+; ZnT8A+; 11 positive mIAA+ nPOD-6104 6-control 41 male Caucasian 20.5 no 20.55 N/A N/A nPOD-6173 6-AAb- 44.1 male Caucasian 23.9 yes <0.05 negative 15 negative nPOD-6141 6-AAb- 36.7 male Caucasian 26 yes <0.05 GADA+;1A-2A+; ZnT8A+; 28 positive mIAA+ https://doi.org/10.1371/journal.pone.0230627.t003 washed with PBS, and DAPI (Sigma; 1:1000) used to visualize nuclei. Coverslips were mounted and images acquired with a Zeiss 710 confocal microscope. Mouse islets were isolated as previously described [21] and subsequently cultured in RPMI 1640 media supplemented with 10% FBS (Hyclone; Fisher Scientific), and 1% Penicillin/Strep- tomycin. Islets were treated with 100 μM GC7 (or 10 mM acetic acid; vehicle) with 0.5 mM aminoguanidine for 72 hours; the duration of treatment to reduce hypusination was previously determined in [10]. Islet were fixed with 4% paraformaldehyde in PBS and blocked for 1 hour in 5% normal donkey serum (NDS) in PBS followed by permeabilization with 0.1% Triton X- 100 in NDS-PBS for 10 min. Islets were stained as described above using chambered slides; maximum intensity projection images were processed from Z-stack image collections acquired with a Zeiss 710 confocal microscope. Results Hyp Beta cell and non-beta cell distribution of eIF5A in mouse We previously developed and characterized a novel antibody that recognizes the unique amino acid hypusine, formed exclusively through posttranslational modification of the Lys50 Hyp residue of eIF5A (eIF5A ) [12,18]. In this study, we utilized this antibody to investigate the PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 5 / 17 PLOS ONE Hypusinated eIF5A in human diabetes Hyp Fig 1. Expression of hypusinated eIF5A (eIF5A ) in mouse and human pancreatic islets. (A) Western blot from mouse and human pancreas tissue and isolated pancreatic islets. (B) Western blot from FACS sorted mouse islet cell populations. (C, D) Representative immunofluorescence images of mouse tissue demonstrating robust expression of Hyp eIF5A in PP-expressing cells. https://doi.org/10.1371/journal.pone.0230627.g001 Hyp expression of eIF5A in mouse and human pancreas tissue and isolated islets as well as Hyp human spleen tissue, to characterize the expression pattern of eIF5A and determine if Hyp eIF5A -expressing cells stratify with characteristics of disease. To that end, we first con- Hyp firmed the presence of eIF5A in islets isolated from mouse and human pancreas as well as in mouse pancreas and human acinar (exocrine) tissue (Fig 1A, S1 Fig). PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 6 / 17 PLOS ONE Hypusinated eIF5A in human diabetes Tomato We next utilized the RIP-cre;R26R mouse model wherein the insulin-expressing cells were labeled with a lineage trace, thereby generating beta cells indelibly marked with fluores- Tomato cent reporter (Tomato) expression. Islet cells from RIP-cre;R26R and control animals were sorted by FACS, using the presence and absence of Tomato expression to separate cells into two populations: beta cells (Tomato-positive) and non-beta cells (Tomato-negative). The cell types represented in the “non-beta cell” sample included (ordered from largest population to smallest): glucagon-expressing alpha cells, somatostatin-expressing delta cells, pancreatic polypeptide-expressing PP cells, ghrelin-expressing epsilon cells, exocrine cells (a possible con- taminant from the process of islet isolation) and support cells including endothelial cells. A similar quantity of Tomato-positive beta cells (1.92x10 cells) and Tomato-negative non-beta cells (2.13x10 cells) were collected (S2 Fig). Subsequent western blot analysis identified that Hyp eIF5A was present in nearly identical abundance in both the beta cell (Tomato-positive) and non-beta cell (Tomato-negative) populations (Fig 1B, S3 Fig). The expression of Pdx1 con- firms the enrichment of beta cells in the Tomato positive cells; the lower level of Pdx1 expres- sion in the non-beta cell fraction can be attributed to the presence of somatostatin-expressing Hyp delta cells. These data demonstrate that eIF5A is expressed in both the beta cell and non- Hyp beta cell fractions; however, whether there is a differential expression of eIF5A in a specific non-beta cell type(s) cannot be clarified from these data. Therefore, to characterize the spatial Hyp distribution of eIF5A expression pattern in the islet, we performed co-immunofluorescence Hyp staining for eIF5A and islet hormones in mouse pancreas tissue. Whereas relatively weak Hyp immunostaining of eIF5A was found throughout the pancreas and islets, robust immunos- Hyp taining of eIF5A was found in the islet cell population that expressed pancreatic polypep- tide (Fig 1C and 1D). To confirm that the high expressing cell population was not an artifact and that our previ- Hyp ously published antibody [12,18] could detect expression of eIF5A by immunofluorescence, we treated HEK293T cells (human) or isolated pancreatic islets (mouse) with the DHPS inhibitor GC7 (N1-Guanyl-1,7-diaminoheptane) [24]. Cells and islets were then analyzed for Hyp eIF5A by immunofluorescence. Whereas the control HEK293T uniformly expressed Hyp Hyp eIF5A , the mouse islets contained cells with both weak and robust expression of eIF5A . Furthermore, following treatment with the inhibitor, we observed a reduction in expression of Hyp eIF5A in both the HEK293T cells and mouse islets compared with vehicle treated controls Hyp (S4 Fig), which verified the measurement of eIF5A expression using our antibody. Hyp eIF5A -expressing cells in the pancreas of human type 2 diabetes Hyp To characterize the expression pattern of eIF5A in the human pancreas, we utilized tissue samples from the Network of Pancreatic Organ Donors with Diabetes (nPOD). A cohort of tis- sues from donors with and without T2D were provided (Table 2). Both pancreas and spleen tissues were acquired from each donor; age, gender, ethnicity and BMI were matched where possible. Given the relatively small size of the cohort, quantitative evaluations were not possi- Hyp ble. Therefore, we evaluated the presence or absence of eIF5A , its cell-type expression pat- tern, and its expression correlation with disease. Hyp Pancreas tissue sections were co-immunostained with the eIF5A -specific antibody and antibodies that recognized the hormones expressed by each of the endocrine cell populations in the islet (insulin, glucagon, somatostatin, ghrelin and pancreatic polypeptide). Robust co- Hyp localization was not observed between eIF5A and insulin (Fig 2A and 2B), glucagon (Fig 2C and 2D), ghrelin (Fig 2E and 2F), or somatostatin (Fig 2G and 2H). However, as observed in the mouse pancreas, cells expressing pancreatic polypeptide were identified to co-express Hyp high levels of eIF5A in control pancreas tissue (Fig 3A). These cells also expressed PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 7 / 17 PLOS ONE Hypusinated eIF5A in human diabetes Hyp Fig 2. The expression pattern of eIF5A in T2D and control pancreatic tissue. In controls (matched for age, gender and BMI) and Hyp T2D pancreas, we evaluated the co-expression of eIF5A with all islet hormones and found no overlap with insulin (A,B), glucagon (C, D), ghrelin (E,F) or somatostatin (G,H). https://doi.org/10.1371/journal.pone.0230627.g002 chromograninA, which confirms their identity as neuroendocrine cells (Fig 3B). The co-locali- Hyp zation of eIF5A with pancreatic polypeptide in the PP-expressing cells was observed in pan- creas tissues from donors with T2D (Fig 3C and 3D) and non-diabetic controls, suggesting no Hyp differential expression related to disease status. Notably, whereas PP and eIF5A were expressed in the same cells, the expression pattern is suggestive of localization in different compartments (Fig 3E). Hyp Spleen tissue sections from the same donors were co-immunostained with eIF5A and markers of various cell types. In particular, Pax5-expressing B cells, CD4-expressing T cells, Hyp and CD8-expressing T cells were evaluated for co-expression of eIF5A . Whereas the expres- Hyp sion patterns observed suggest that most Pax5+ B cells expressed eIF5A , only a select group Hyp of eIF5A -expressing cells appear to co-expressed either CD4 or CD8 (Fig 4A–4C; S5–S7 Figs). No obvious differences in staining intensity or distribution were observed between sam- ples from T2D and controls (Fig 4D–4E). Hyp eIF5A -expressing cells in the pancreas of human type 1 diabetes Donor pancreas and spleen tissue from individuals with T1D were also acquired from nPOD Hyp and evaluated for the expression pattern of eIF5A . This cohort of samples included T1D donors that were autoantibody-positive and autoantibody-negative, with both short and long disease duration; non-diabetic controls were matched for age, gender, ethnicity and BMI (Table 3). Similar to the T2D/control samples, we identified cells co-expressing the hormone Hyp PP with high intensity eIF5A immunostaining (Fig 5A–5F); robust co-expression of eIF5A- Hyp Hyp with other islet hormones was not observed. Moreover, the eIF5A -expressing cells expressed ChromograninA (Fig 5G–5I), which again confirmed that these cells are neuroen- docrine in nature. Evaluation of spleen tissue for all T1D donors and controls revealed an identical pattern of expression to that observed in the T2D donors and controls. Specifically, Hyp the majority of eIF5A -expressing cells co-expressed Pax5 (Fig 6; S8–S10 Figs). PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 8 / 17 PLOS ONE Hypusinated eIF5A in human diabetes Hyp Fig 3. eIF5A is robustly expressed in the pancreatic polypeptide-expressing PP cells in the islet. (A-D) In both Hyp controls and T2D pancreas, co-expression of eIF5A with pancreatic polypeptide (PP) and chromograninA (ChgA) Hyp was observed. (E) An expression pattern of eIF5A in the PP cells suggestive of localization to the ER was observed in cells in both controls and T2D pancreas. https://doi.org/10.1371/journal.pone.0230627.g003 Discussion Previous data from mouse models identified that pharmacological modulation of the hypusi- nation of eIF5A enhanced beta cell mass and improved glucose tolerance in mouse models of Hyp both T1D and T2D [14,16], thereby suggesting an important role for eIF5A in the setting of Hyp diabetes. However, to translate these findings to human, a greater understanding of eIF5A in the human pancreas and spleen would be required. This study represents the first descrip- Hyp tion of eIF5A expression in human organs from donors with and without diabetes. PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 9 / 17 PLOS ONE Hypusinated eIF5A in human diabetes Hyp Hyp Fig 4. eIF5A expression pattern in the spleen of control and T2D. eIF5A is expressed in immune cells in the spleen. We Hyp evaluated expression of eIF5A in Pax5+ B cells, CD4+ T cells and CD8+ T cells in the spleens of donors with T2D and controls Hyp Hyp matched for age, gender and BMI. Most eIF5A + cells co-expressed Pax5+ (A,D); however, a select group of eIF5A + cells expressed either CD4+ (B,E) or CD8+ (C,F). https://doi.org/10.1371/journal.pone.0230627.g004 Hyp Importantly, our results reveal a heretofore unappreciated enrichment of eIF5A in subsets of endocrine cells in the pancreas and immune cells in the spleen. Interestingly, the presence Hyp of eIF5A co-expressing cells was not obviously enhanced in diseased tissue; however, larger cohorts are required where tissue can be sampled from across whole organs in order to pre- cisely quantitate the presence of these cells and definitively determine correlation with disease. Further investigation is also required as to the relative abundance of the deoxyhypusine and hypusine forms of eIF5A in both the normal and diseased setting. Currently, published work suggests that the deoxyhypusine form of eIF5A is transient and reversible, and therefore not as abundant due to its rapid modification by the enzyme DOHH (deoxyhypusine hydroxylase) to the hypusinated form of eIF5A during the process of hypusine biosynthesis [25]. Our antibody is specific for the modified forms of eIF5A over the unmodified forms [18], and the form most detectable by our antibody due to its high prevalence is the more stable hypusinated form of Hyp eIF5A (eIF5A ). Further investigation and tool development will be required to determine and understand the relative abundance of the deoxyhypusine and hypusine forms of eIF5A in both the normal and diseased settings. Hyp Our findings in the pancreas demonstrate that eIF5A is expressed in both the exocrine and endocrine compartments in mouse and human. Previous reports have also shown expres- Hyp sion of eIF5A in mouse islets [10,11]. However, our immunoblot of sorted mouse islet cells Hyp further defined that eIF5A expression in the islet can be found in both the beta cell and non-beta cell populations. The non-beta cell populations encompass multiple hormone- PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 10 / 17 PLOS ONE Hypusinated eIF5A in human diabetes Hyp Fig 5. Expression of eIF5A in the T1D pancreas. (A-F) Identical to the pattern identified in T2D and control tissues, high Hyp expression of eIF5A is observed in PP cells in the T1D, both auto-antibody positive (aAb+) and auto-antibody negative(aAb-), pancreas and controls (matched for age, gender, ethnicity and BMI). (G-I) In all cases, these cells express the endocrine cell marker ChromograninA (ChgA). https://doi.org/10.1371/journal.pone.0230627.g005 expressing cell types, and our immunostaining analysis of mouse tissue clarified that the most Hyp robust expression of eIF5A is in the PP cell population. Given the over-representation of PP cells in the uncinate region of the pancreas [26], we analyzed tissue sections that contained Hyp the uncinate region and found that eIF5A is robustly co-expressed in PP cells of human islets. Despite evidence that PP cells have a critical secretory function in the brain-gut axis [27] and may serve as a regulator of intra-islet secretion [28], the role of PP cells in the context of diabetes has received little attention. From a developmental perspective, PP cells are predomi- nantly derived from the ghrelin-expressing cell lineage found in the embryonic pancreas [29]; Hyp however, the function of eIF5A in the PP cell population postnatally or any function for Hyp eIF5A in the development of PP cells has yet to be elucidated. Interestingly, expression analysis of 12-lipoxygenase, a factor known to promote inflammation in the setting of diabetes, PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 11 / 17 PLOS ONE Hypusinated eIF5A in human diabetes Hyp Fig 6. Expression of eIF5A in spleen tissue from donors with T1D and matched control donors. We examined spleen tissue from persons with autoantibody positive (AAb+) and auto-antibody negative (AAb-) T1D, and corresponding controls matched for age, gender, ethnicity, and BMI. As observed in T2D and matched control spleen Hyp Hyp tissue, most eIF5A -expressing cells were Pax5+ (A-C); however, some eIF5A + cells expressed either CD4+ (D-F) or CD8+ (G-I). https://doi.org/10.1371/journal.pone.0230627.g006 is also increased in the PP-expressing cell population in pancreas tissue from human donors (collected through nPOD; [30]). Clearly, a greater understanding is required for the role of PP cells in the pathogenesis of diabetes. Given that much of the published and ongoing work on hypusine biosynthesis in mice has Hyp Hyp studied eIF5A in the context of diabetes, we had hypothesized that eIF5A expression would be identified predominantly in the insulin-producing beta cell population. Our western Hyp Hyp blot analysis did reveal eIF5A expression in human islets. Moreover, we observed eIF5A expression in a purified population of beta cells (Tomato+) from mouse islets. Interestingly, Hyp the quantitative nature of western blots indicates that the expression of eIF5A must be lower in the purified beta cells compared with non-beta cells given that PP cells comprise only a small portion of the Tomato(-) non-beta cell fraction whereas the Tomato(+) fraction is Hyp composed exclusively of beta cells, and we see near equivalent expression of eIF5A in both sorted populations. This finding is consistent with the immunofluorescence data, wherein we PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 12 / 17 PLOS ONE Hypusinated eIF5A in human diabetes Hyp identified robust expression of eIF5A in PP-expressing cells. The lack of observable co- Hyp expression of eIF5A and hormones in all cell types in the islet was unexpected; however, we must consider the possibility that this may be due to the limitations in detection of low protein expression by immunofluorescence. Therefore, and considering all data together, our observa- Hyp tions indicate the presence of eIF5A in both beta cells and non-beta cells, with particularly high expression in one small non-beta cell populations, the PP cells. Our previous finding that pharmacological inhibition of eIF5A hypusination (using the drug GC7; N1-Guanyl-1,7-diaminoheptane) in NOD mice improved glucose tolerance and preserved beta cell mass [14]. These improvements were also accompanied by reductions in insulitis, which led us to question whether the improvements in beta cell function were due to a direct effect of DHPS inhibition in beta cells, or an indirect effect related to DHPS inhibition Hyp in infiltrating immune cells. Our work and that of others suggest a role for eIF5A and DHPS in promoting T cell and B cell proliferation [14,31,32], which was the basis for our Hyp hypothesis that perhaps eIF5A is differentially expressed in immune cells in individuals with diabetes compared with controls. However, identical expression patterns were noted in Hyp all spleen tissue evaluated. The identical expression patterns of eIF5A between healthy and disease in both the immune cell populations and islet cell populations could also suggest that it Hyp is not the abundance of eIF5A that is critical for promotion of disease. Rather, the presence Hyp of eIF5A facilitating the translation of different mRNAs in the disease setting compared with the healthy setting could drive pathogenesis. Given our recent findings that deletion of Dhps in adult mouse beta cells results in reduced diet-induced beta cell proliferation and sub- sequent glucose intolerance due to altered translation of cyclinD2 [10], we are now investigat- Hyp ing the impact of eIF5A on mRNA translation in other diabetes-related cell populations. Supporting information S1 Fig. Source images for western blots of mouse and human pancreas and islets. Hyp (A) Immunoblot images show expression of eIF5A , insulin and total eIF5A in cell lysates from mouse whole pancreas and isolated islets. The top portions of these blots were proved with antibodies not related to this study. (B) Total protein expression as visualized by Pon- Hyp ceauS staining. (C) Immunoblot images show expression of eIF5A , insulin and total eIF5A in cell lysate from human exocrine tissue and isolated islets. One blot was probed twice, and this second antibody was not related to this study. (D) Total protein expression as visualized by PonceauS staining. The dotted box shows the lanes where the exocrine and islet samples were run; the other samples are unrelated to this study. (PDF) S2 Fig. Collection of beta cell and non-beta cell populations by FACS. (A) Islets isolated from multiple RIP-cre;R26RTomato mice were pooled together and processed for fluorescence activated cell sorting (FACS). Islet cells were sorted into two populations: Tomato-positive beta cells (R4), and Tomato-negative non-beta cells (R3, islet cells expressing glucagon, somatostatin, ghrelin, and pancreatic polypeptide). (PDF) S3 Fig. Source images for western blots of FACS sorted mouse islet cell populations. Hyp (A) Immunoblot for expression of Pdx1 and eIF5A in cell lysates from Tomato-negative non-beta cells and Tomato-positive beta cells. (B) Immunoblot for expression of total eIF5A. (C) Total protein expression as visualized by PonceauS staining. (PDF) PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 13 / 17 PLOS ONE Hypusinated eIF5A in human diabetes S4 Fig. Evaluation of eIF5AHyp expression following treatment with DHPS inhibitor. HEK293T cells were treated with the DHPS inhibitor GC7 (N1-Guanyl-1,7,diaminoheptane) and analyzed for eIF5AHyp expression by immunofluorescence. (A) Control HEK293T cells Hyp uniformly expressed eIF5A . (B,C) Treatment with GC7 resulted in reduced expression of Hyp eIF5A . (D) A secondary antibody control was also performed to confirm that the observed signal was not an artifact. Mouse pancreatic islets were also treated with GC7 and analyzed for Hyp eIF5A expression by immunofluorescence. (E) Control mouse islets contained cells with Hyp both weak and robust expression of eIF5A . (F, G) Islets treated with GC7 showed a reduc- Hyp tion in expression of eIF5A . Images are 20X. Inset images are higher magnification of the areas outlined with white boxes. (PDF) Hyp S5 Fig. eIF5A expression pattern in the Pax5-expressing cell population in spleen tissue Hyp of control and T2D. We evaluated the expression of eIF5A in Pax5-expressing B cells in the spleens of donors with T2D and controls matched for age, gender and BMI. The fluores- cent channels have been separated to better display the expression patterns of the Pax5-expres- Hyp sing B cells (A, B), eIF5A -expressing cells (C, D), and the overlap between the Pax5- Hyp expressing and eIF5A -expressing populations (E, F). All images are 20X. (PDF) Hyp S6 Fig. eIF5A expression pattern in the CD4-expressing cell population in spleen tissue Hyp of control and T2D. We evaluated the expression of eIF5A in CD4-expressing T cells in the spleens of donors with T2D and controls matched for age, gender and BMI. The fluores- cent channels have been separated to better display the expression patterns of the CD4-expres- Hyp sing T cells (A, B), eIF5A -expressing cells (C, D), and the minimal overlap between the Hyp CD4-expressing and eIF5A -expressing populations (E, F). All images are 20X. (PDF) Hyp S7 Fig. eIF5A expression pattern in the CD8-expressing T cell population in spleen tis- Hyp sue of control and T2D. We evaluated the expression of eIF5A in CD8-expressing T cells in the spleens of donors with T2D and controls matched for age, gender and BMI. The fluores- cent channels have been separated to better display the expression patterns of the CD8-expres- Hyp sing T cells (A, B), eIF5A -expressing cells (C, D), and the minimal overlap between the Hyp CD8-expressing and eIF5A -expressing populations (E, F). All images are 20X. (PDF) Hyp S8 Fig. eIF5A expression pattern in the Pax5-expressing B cell population in spleen tis- Hyp sue of control and T1D. We evaluated the expression of eIF5A in Pax5-expressing B cells in the spleens of donors with auto-antibody positive (AAb+) and auto-antibody negative (AAb-) T1D, and corresponding controls matched for age, gender, ethnicity, and BMI. The fluorescent channels have been separated to better display the expression patterns of the Pax5- Hyp expressing B cells (A—C), eIF5A -expressing cells (D—F), and the overlap between the Hyp Pax5-expressing and eIF5A -expressing populations (G—I). All images are 20X. (PDF) Hyp S9 Fig. eIF5A expression pattern in the CD4-expressing T cell population in spleen tis- Hyp sue of control and T1D. We evaluated the expression of eIF5A in CD4-expressing T cells in the spleens of donors with auto-antibody positive (AAb+) and auto-antibody negative (AAb-) T1D, and corresponding controls matched for age, gender, ethnicity, and BMI. The fluorescent channels have been separated to better display the expression patterns of the Hyp CD4-expressing T cells (A—C), eIF5A -expressing cells (D—F), and the minimal overlap PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 14 / 17 PLOS ONE Hypusinated eIF5A in human diabetes Hyp between the CD4-expressing and eIF5A -expressing populations (G—I). All images are 20X. (PDF) Hyp S10 Fig. eIF5A expression pattern in the CD8-expressing T cell population in spleen tis- Hyp sue of control and T1D. We evaluated the expression of eIF5A in CD8-expressing T cells in the spleens of donors with auto-antibody positive (AAb+) and auto-antibody negative (AAb-) T1D, and corresponding controls matched for age, gender, ethnicity, and BMI. The fluorescent channels have been separated to better display the expression patterns of the Hyp CD8-expressing T cells (A—C), eIF5A -expressing cells (D—F), and the minimal overlap Hyp between the CD8-expressing and eIF5A -expressing populations (G—I). All images are 20X. (PDF) Acknowledgments The authors wish to thank Dr. David Morris and the Flow Cytometry Core Facility at Indiana University School of Medicine for assistance with FACS. Human pancreatic islets were pro- vided by the NIDDK-funded Integrated Islet Distribution Program (IIDP) at City of Hope, NIH Grant # 2UC4DK098085. Human donor acinar tissue was provided by Dr. Rita Bottino at the Center for Organ Recovery and Education (CORE), Pittsburg PA. This research was also performed with the support of the Network for Pancreatic Organ donors with Diabetes (nPOD; RRID:SCR_014641), a collaborative type 1 diabetes research project sponsored by JDRF (nPOD: 5-SRA-2018-557-Q-R) and The Leona M. & Harry B. Helmsley Charitable Trust (Grant#2018PG-T1D053). Organ Procurement Organizations (OPO) partnering with nPOD to provide research resources are listed at http://www.jdrfnpod.org//for-partners/ npod-partners/. This manuscript was released as a preprint at bioRxiv https://www.biorxiv. org/content/10.1101/745919v1. Author Contributions Conceptualization: Teresa L. Mastracci, Raghavendra G. Mirmira. Data curation: Teresa L. Mastracci, Leah R. Padgett. Formal analysis: Teresa L. Mastracci, Stephanie C. Colvin, Leah R. Padgett, Raghavendra G. Mirmira. Funding acquisition: Teresa L. Mastracci, Raghavendra G. Mirmira. Investigation: Teresa L. Mastracci, Stephanie C. Colvin, Leah R. Padgett, Raghavendra G. Mirmira. Methodology: Teresa L. Mastracci, Stephanie C. Colvin, Leah R. Padgett. Project administration: Teresa L. Mastracci. Resources: Teresa L. Mastracci, Raghavendra G. Mirmira. Software: Teresa L. Mastracci. Supervision: Teresa L. Mastracci, Raghavendra G. Mirmira. Validation: Teresa L. Mastracci. Visualization: Teresa L. Mastracci, Leah R. Padgett. PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 15 / 17 PLOS ONE Hypusinated eIF5A in human diabetes Writing – original draft: Teresa L. Mastracci, Raghavendra G. Mirmira. Writing – review & editing: Teresa L. Mastracci, Stephanie C. Colvin, Leah R. Padgett, Ragha- vendra G. Mirmira. References 1. Park MH, Cooper HL, Folk JE. Identification of hypusine, an unusual amino acid, in a protein from human lymphocytes and of spermidine as its biosynthetic precursor. Proc Natl Acad Sci U S A. 1981; 78: 2869–2873. https://doi.org/10.1073/pnas.78.5.2869 PMID: 6789324 2. Park MH. The post-translational synthesis of a polyamine-derived amino acid, hypusine, in the eukary- otic translation initiation factor 5A (eIF5A). J Biochem. 2006; 139: 161–9. https://doi.org/10.1093/jb/ mvj034 PMID: 16452303 3. Saini P, Eyler DE, Green R, Dever TE. Hypusine-containing protein eIF5A promotes translation elonga- tion. Nature. 2009; 459: 118–21. https://doi.org/10.1038/nature08034 PMID: 19424157 4. Park MH, Wolff EC, Lee YB, Folk JE. Antiproliferative effects of inhibitors of deoxyhypusine synthase. Inhibition of growth of Chinese hamster ovary cells by guanyl diamines. J Biol Chem. 1994; 269: 27827–32. PMID: 7961711 5. Wolff EC, Kang KR, Kim YS, Park MH. Posttranslational synthesis of hypusine: evolutionary progres- sion and specificity of the hypusine modification. Amino Acids. 2007; 33: 341–50. https://doi.org/10. 1007/s00726-007-0525-0 PMID: 17476569 6. Schuller AP, Wu CC-C, Dever TE, Buskirk AR, Green R. eIF5A Functions Globally in Translation Elon- gation and Termination. Mol Cell. 2017; 66: 194–205.e5. https://doi.org/10.1016/j.molcel.2017.03.003 PMID: 28392174 7. Gutierrez E, Shin BS, Woolstenhulme CJ, Kim JR, Saini P, Buskirk AR, et al. eIF5A promotes transla- tion of polyproline motifs. Mol Cell. 2013; 51: 35–45. https://doi.org/10.1016/j.molcel.2013.04.021 PMID: 23727016 8. McDuffie M, Maybee NA, Keller SR, Stevens BK, Garmey JC, Morris MA, et al. Nonobese diabetic (NOD) mice congenic for a targeted deletion of 12/15-lipoxygenase are protected from autoimmune dia- betes. Diabetes. 2008; 57: 199–208. https://doi.org/10.2337/db07-0830 PMID: 17940120 9. Serreze DV, Chapman HD, Varnum DS, Gerling I, Leiter EH, Shultz LD. Initiation of autoimmune diabe- tes in NOD/Lt mice is MHC class I-dependent. J Immunol Baltim Md 1950. 1997; 158: 3978–3986. 10. Levasseur EM, Yamada K, Piñeros AR, Wu W, Syed F, Orr KS, et al. Hypusine biosynthesis inβ cells links polyamine metabolism to facultative cellular proliferation to maintain glucose homeostasis. Sci Sig- nal. 2019; 12. https://doi.org/10.1126/scisignal.aax0715 PMID: 31796630 11. Tersey SA, Colvin SC, Maier B, Mirmira RG. Protective effects of polyamine depletion in mouse models of type 1 diabetes: implications for therapy. Amino Acids. 2014; 46: 633–42. https://doi.org/10.1007/ s00726-013-1560-7 PMID: 23846959 12. Maier B, Ogihara T, Trace AP, Tersey SA, Robbins RD, Chakrabarti SK, et al. The unique hypusine modification of eIF5A promotes islet beta cell inflammation and dysfunction in mice. J Clin Invest. 2010; 120: 2156–70. https://doi.org/10.1172/JCI38924 PMID: 20501948 13. Maier B, Tersey SA, Mirmira RG. Hypusine: a new target for therapeutic intervention in diabetic inflam- mation. Discov Med. 2010; 10: 18–23. PMID: 20670594 14. Colvin SC, Maier B, Morris DL, Tersey SA, Mirmira RG. Deoxyhypusine synthase promotes differentia- tion and proliferation of T helper type 1 (Th1) cells in autoimmune diabetes. J Biol Chem. 2013; 288: 36226–35. https://doi.org/10.1074/jbc.M113.473942 PMID: 24196968 15. Coleman DL. Obese and diabetes: two mutant genes causing diabetes-obesity syndromes in mice. Dia- betologia. 1978; 14: 141–148. https://doi.org/10.1007/bf00429772 PMID: 350680 16. Robbins RD, Tersey SA, Ogihara T, Gupta D, Farb TB, Ficorilli J, et al. Inhibition of deoxyhypusine synthase enhances islet {beta} cell function and survival in the setting of endoplasmic reticulum stress and type 2 diabetes. J Biol Chem. 285: 39943–52. https://doi.org/10.1074/jbc.M110.170142 PMID: 17. Hatanaka M, Anderson-Baucum E, Lakhter A, Kono T, Maier B, Tersey SA, et al. Chronic high fat feed- ing restricts islet mRNA translation initiation independently of ER stress via DNA damage and p53 acti- vation. Sci Rep. 2017; 7. https://doi.org/10.1038/s41598-017-03869-5 PMID: 28630491 18. Nishiki Y, Farb TB, Friedrich J, Bokvist K, Mirmira RG, Maier B. Characterization of a novel polyclonal anti-hypusine antibody. Springerplus. 2013; 2: 421. https://doi.org/10.1186/2193-1801-2-421 PMID: PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 16 / 17 PLOS ONE Hypusinated eIF5A in human diabetes 19. Postic C, Shiota M, Niswender KD, Jetton TL, Chen Y, Moates JM, et al. Dual roles for glucokinase in glucose homeostasis as determined by liver and pancreatic beta cell-specific gene knock-outs using Cre recombinase. J Biol Chem. 1999; 274: 305–15. https://doi.org/10.1074/jbc.274.1.305 PMID: 20. Madisen L, Zwingman TA, Sunkin SM, Oh SW, Zariwala HA, Gu H, et al. A robust and high-throughput Cre reporting and characterization system for the whole mouse brain. Nat Neurosci. 2010; 13: 133–40. https://doi.org/10.1038/nn.2467 PMID: 20023653 21. Stull ND, Breite A, McCarthy R, Tersey SA, Mirmira RG. Mouse Islet of Langerhans Isolation using a Combination of Purified Collagenase and Neutral Protease. J Vis Exp JoVE. 2012 [cited 30 Mar 2018]. https://doi.org/10.3791/4137 PMID: 22987198 22. Mastracci TL, Anderson KR, Papizan JB, Sussel L. Regulation of Neurod1 contributes to the lineage potential of Neurogenin3+ endocrine precursor cells in the pancreas. PLoS Genet. 2013; 9: e1003278. https://doi.org/10.1371/journal.pgen.1003278 PMID: 23408910 23. Campbell-Thompson M, Wasserfall C, Kaddis J, Albanese-O’Neill A, Staeva T, Nierras C, et al. Net- work for Pancreatic Organ Donors with Diabetes (nPOD): developing a tissue biobank for type 1 diabe- tes. Diabetes Metab Res Rev. 2012; 28: 608–617. https://doi.org/10.1002/dmrr.2316 PMID: 22585677 24. Jakus J, Wolff EC, Park MH, Folk JE. Features of the spermidine-binding site of deoxyhypusine synthase as derived from inhibition studies. Effective inhibition by bis- and mono-guanylated diamines and polyamines. J Biol Chem. 1993; 268: 13151–13159. PMID: 8514754 25. Park MH, Nishimura K, Zanelli CF, Valentini SR. Functional significance of eIF5A and its hypusine mod- ification in eukaryotes. Amino Acids. 2010; 38: 491–500. https://doi.org/10.1007/s00726-009-0408-7 PMID: 19997760 26. Wang X, Zielinski MC, Misawa R, Wen P, Wang T-Y, Wang C-Z, et al. Quantitative Analysis of Pancre- atic Polypeptide Cell Distribution in the Human Pancreas. PLoS ONE. 2013; 8. https://doi.org/10.1371/ journal.pone.0055501 PMID: 23383206 27. Holzer P, Reichmann F, Farzi A. Neuropeptide Y, peptide YY and pancreatic polypeptide in the gut- brain axis. Neuropeptides. 2012; 46: 261–274. https://doi.org/10.1016/j.npep.2012.08.005 PMID: 28. Brereton MF, Vergari E, Zhang Q, Clark A. Alpha-, Delta- and PP-cells: Are They the Architectural Cor- nerstones of Islet Structure and Co-ordination? J Histochem Cytochem Off J Histochem Soc. 2015; 63: 575–591. https://doi.org/10.1369/0022155415583535 PMID: 26216135 29. Arnes L, Hill JT, Gross S, Magnuson MA, Sussel L. Ghrelin expression in the mouse pancreas defines a unique multipotent progenitor population. PloS One. 2012; 7: e52026. https://doi.org/10.1371/journal. pone.0052026 PMID: 23251675 30. Grzesik WJ, Nadler JL, Machida Y, Nadler JL, Imai Y, Morris MA. Expression pattern of 12-lipoxygen- ase in human islets with type 1 diabetes and type 2 diabetes. J Clin Endocrinol Metab. 2015; 100: E387–395. https://doi.org/10.1210/jc.2014-3630 PMID: 25532042 31. Bevec D, Jaksche H, Oft M, Wo ¨ hl T, Himmelspach M, Pacher A, et al. Inhibition of HIV-1 replication in lymphocytes by mutants of the Rev cofactor eIF-5A. Science. 1996; 271: 1858–1860. https://doi.org/10. 1126/science.271.5257.1858 PMID: 8596953 32. Schlee M, Krug T, Gires O, Zeidler R, Hammerschmidt W, Mailhammer R, et al. Identification of Epstein-Barr virus (EBV) nuclear antigen 2 (EBNA2) target proteins by proteome analysis: activation of EBNA2 in conditionally immortalized B cells reflects early events after infection of primary B cells by EBV. J Virol. 2004; 78: 3941–3952. https://doi.org/10.1128/JVI.78.8.3941-3952.2004 PMID: 15047810 PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 17 / 17 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png PLoS ONE Public Library of Science (PLoS) Journal

Hypusinated eIF5A is expressed in the pancreas and spleen of individuals with type 1 and type 2 diabetes

PLoS ONE, Volume 15 (3) – Mar 24, 2020

Loading next page...
 
/lp/public-library-of-science-plos-journal/hypusinated-eif5a-is-expressed-in-the-pancreas-and-spleen-of-P9HkSBS1qu
Publisher
Public Library of Science (PLoS) Journal
Copyright
Copyright: © 2020 Mastracci et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability: All relevant data are within the manuscript and its Supporting Information files. Funding: TLM 5-CDA-2016-194-A-N Juvenile Diabetes Research Foundation https://www.jdrf.org/ The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. RGM R01 DK60581 National Institute of Diabetes and Digestive and Kidney Diseases https://www.niddk.nih.gov/ The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interests: The authors have declared that no competing interests exist.
eISSN
1932-6203
DOI
10.1371/journal.pone.0230627
Publisher site
See Article on Publisher Site

Abstract

OPENACCESS The gene encoding eukaryotic initiation factor 5A (EIF5A) is found in diabetes-susceptibility Citation: Mastracci TL, Colvin SC, Padgett LR, loci in mouse and human. eIF5A is the only protein known to contain hypusine (hydroxypu- Mirmira RG (2020) Hypusinated eIF5A is expressed in the pancreas and spleen of individuals trescine lysine), a polyamine-derived amino acid formed post-translationally in a reaction with type 1 and type 2 diabetes. PLoS ONE 15(3): catalyzed by deoxyhypusine synthase (DHPS). Previous studies showed pharmacologic e0230627. https://doi.org/10.1371/journal. blockade of DHPS in type 1 diabetic NOD mice and type 2 diabetic db/db mice improved glu- pone.0230627 cose tolerance and preserved beta cell mass, which suggests that hypusinated eIF5A Editor: Cinzia Ciccacci, Unicamillus, Saint Camillus Hyp (eIF5A ) may play a role in diabetes pathogenesis by direct action on the beta cells and/or International University of Health Sciences, ITALY altering the adaptive or innate immune responses. To translate these findings to human, we Received: October 9, 2019 examined tissue from individuals with and without type 1 and type 2 diabetes to determine Accepted: March 4, 2020 Hyp Hyp the expression of eIF5A . We detected eIF5A in beta cells, exocrine cells and immune Hyp Published: March 24, 2020 cells; however, there was also unexpected enrichment of eIF5A in pancreatic polypep- Hyp tide-expressing PP cells. Interestingly, the presence of eIF5A co-expressing PP cells Copyright:© 2020 Mastracci et al. This is an open access article distributed under the terms of the was not enhanced with disease. These data identify new aspects of eIF5A biology and high- Creative Commons Attribution License, which light the need to examine human tissue to understand disease. permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability Statement: All relevant data are within the manuscript and its Supporting Introduction Information files. The mechanisms underlying the pathogeneses of type 1 diabetes (T1D) and type 2 diabetes Funding: TLM 5-CDA-2016-194-A-N Juvenile Diabetes Research Foundation https://www.jdrf. (T2D) involve the activation of systemic and local inflammatory pathways, leading to eventual org/ The funders had no role in study design, data dysfunction, de-differentiation and/or death of the beta cells in the pancreatic islet. Elucidating collection and analysis, decision to publish, or the molecular mechanisms driving the inflammatory response is applicable to the development preparation of the manuscript. RGM R01 DK60581 of therapies for both diseases. In addition, an urgent priority in T1D research is the discovery National Institute of Diabetes and Digestive and of biomarkers that can assist in the identification of individuals with pre-clinical disease so Kidney Diseases https://www.niddk.nih.gov/ The funders had no role in study design, data collection early preventative therapeutic interventions can be implemented. PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 1 / 17 PLOS ONE Hypusinated eIF5A in human diabetes and analysis, decision to publish, or preparation of Recently, our laboratories have been investigating the involvement of the hypusinated form the manuscript. of eukaryotic initiation factor 5A (eIF5A) in the development and progression of diabetes in mice. To date, eIF5A is the only known protein to contain hypusine (hydroxyputrescine Competing interests: The authors have declared that no competing interests exist. lysine) [1], which is a polyamine-derived amino acid. This post-translational modification, formed by the process of “hypusination” [2], is catalyzed through a multi-step reaction initi- ated by the rate-limiting enzyme deoxyhypusine synthase (DHPS) and uses the polyamine spermidine as a cofactor to modify the Lys50 of eIF5A [2]. Previous studies using human cell Hyp lines and yeast determined that eIF5A, the hypusinated form of eIF5A (eIF5A ) and DHPS are vital for cell viability and proliferation [3,4]. Evolutionarily, eIF5A is highly conserved including the amino acid sequence surrounding the hypusine residue, which suggests an important role for this modification [5]. Whereas studies across species have established that Hyp eIF5A efficiently binds the ribosome complex and facilitates mRNA translation [3,6,7], the Hyp exact function of eIF5A and eIF5A remains unknown. Interestingly, the gene encoding eIF5A is found in the Idd4 diabetes-susceptibility locus in Hyp non-obese diabetes (NOD) mice [8,9]. In prior studies, we showed that eIF5A is expressed in the pancreatic islets of mouse [10,11], is responsible for the translation of a subset of cyto- Hyp kine-induced transcripts in beta cells in mouse models of diabetes [12,13], and that eIF5A also appears to be required for the activation and proliferation of effector T helper cells [14]. Moreover, reducing the hypusination of eIF5A in NOD mice, a model of T1D, by pharmaco- logical inhibition of DHPS resulted in reduced insulitis, improved glucose tolerance and pre- served beta cell mass [14]. Similarly, pharmacological blockade of DHPS in db/db mice [15], a model of T2D improved glucose tolerance and enhanced beta cell mass [16]. Together these Hyp data suggest that eIF5A may play a role in the pathogenesis of diabetes such that altering Hyp the expression of eIF5A may improve diabetes outcomes long-term. Hyp To translate these findings to human, a greater understanding of eIF5A in the human Hyp pancreas and spleen is required. In particular, determining the expression pattern of eIF5A Hyp in human tissues and whether eIF5A -expressing cells stratify with characteristics of disease would be informative. In this study, we used human donor tissue samples from the Network of Pancreatic Organ Donors with Diabetes (nPOD) (www.jdrfnpod.org) to examine the expres- Hyp sion pattern of eIF5A in the human pancreas and spleen from individuals with T1D, T2D and non-diabetic controls. Materials and methods Pancreas and isolated islet cells from mouse All mice were purchased from the Jackson Laboratory and maintained under a protocol approved by the Indiana University School of Medicine Institutional Animal Care and Use Committee (OLAW Assurance Number: D16-00584 (A4091-01); USDA Certificate Number: 32-R-0025; Customer ID: 798; AAALAC Unit Number: 00083; Protocol Approval #19043). The approved method of euthanasia for mice was carbon dioxide inhalation overdose deliv- ered using a gas cylinder, flow meter/regulator, and induction chamber; 100% CO2 delivered at a rate such that 20–30% of the volume of the chamber is displaced per minute. This was fol- lowed by the secondary method of cervical dislocation. Human donor tissues were collected and provided to the investigator by the Network of Pancreatic Organ Donors with Diabetes (https://www.jdrfnpod.org). The study was approved by the University of Florida Institutional Review Board (IRB-1); approval #IRB201600029. Written consent was obtained. Total pancreas and isolated islets from wildtype C57BL/6 mice as well as acinar tissue and isolated islets from human donors (Table 1 details human donors) were subjected to immuno- Hyp blot analysis as previously described [17]. Rabbit anti-eIF5A ([13,18]; 1:1000), mouse anti- PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 2 / 17 PLOS ONE Hypusinated eIF5A in human diabetes Table 1. Human donor information for islet and acinar tissue preparations. Islet Preparation Acinar Preparation Unique identifier SAMN11578698 UNOS AGDB487 Donor Age (years) 57.0 24 Donor Sex (M/F) M M Donor BMI (kg/m ) 25.8 31.7 Donor HbA1c 5.7 not tested Origin/source of islets IIDP (Integrated Islet Distribution CORE (Center for Organ Recovery and Program) Education) Islet isolation center The Scharp-Lacy Research Institute Pittsburgh (AHN) Donor history of diabetes? Yes/ No No No If Yes, complete the next two lines if this information is available Diabetes duration (years) Not Applicable Not Applicable Glucose-lowering therapy at time Not Applicable Not Applicable of death https://doi.org/10.1371/journal.pone.0230627.t001 eIF5A (BD Biosciences; 1:2000) and guinea pig anti-insulin (DAKO; 1:5000) antibodies were used to confirm protein expression in the pancreas. Mice containing the RIP-cre allele (B6.CG-Tg(Ins2-cre)25Mgn/J) [19] were mated with Tomato tm14(CAG-tdTomato)Hze/J those containing the R26R allele (B6.Cg-Gt(ROSA)26Sor ) [20] to produce double transgenic animals wherein all the insulin-producing beta cells in the pancreas expressed a fluorescent (Tomato) reporter. Pancreatic islets were isolated from RIP-cre; Tomato Tomato R26R mice and R26R mice as previously described [21]. The isolated islets from all mice were pooled together and processed for fluorescence activated cell sorting (FACS), which facilitated the separation of islet cells into two populations: Tomato-positive beta cells and Tomato-negative non-beta cells (which included islet cells expressing glucagon, somatostatin, ghrelin, and pancreatic polypeptide). Pooled islets were washed with sterile PBS (Fisher Scien- tific) and incubated in Accutase cell detachment solution (Sigma) for 10 minutes at 37C with constant mixing (1000 rpm). Islet cells were removed from the Accutase solution by centrifu- gation (500 x g for 1 min) and resuspended in cold buffer containing 2% BSA,1uM EDTA, and equal parts PBS and HBSS (Fisher Scientific). The cells were filtered, collected and incubated with APC viability dye (Zombie NIR-IR dye; BioLegend) per the manufacturer’s recom- Tomato Tomato mended protocol. Single-cell suspensions from RIP-cre;R26R mice and R26R were then sorted using an iCyt Reflection with 100 μm nozzle at 23 psi. Dead cells (NIR-IR+) were excluded; Tomato(+) cells and Tomato(-) cells were collected into tubes containing sort buffer. Data were analyzed using FlowJo software (Tree Star). Lysates from the two populations of Hyp cells were subjected to immunoblot analysis. Rabbit anti-eIF5A and mouse anti-eIF5A anti- Hyp bodies were used as above, to evaluate the abundance of eIF5A in the beta cell and non- beta cell populations. Rabbit anti-Pdx1 (Chemicon; 1:1000) antibody was used to evaluate enrichment of the beta cell (tomato+) population. Mouse pancreas tissue and immunofluorescence analysis Pancreas tissue was harvested from wildtype C57BL/6 mice and fixed in 4% paraformaldehyde (Fisher Scientific), cryo-preserved using 30% sucrose, embedded in OCT (Fisher Scientific) and sectioned onto glass microscope slides. Methods previously described for pancreas preser- vation and immunofluorescence were followed [22]. Pancreas tissue sections (8 μm) were stained using the following primary antibodies: guinea pig anti-insulin (DAKO; 1:500), goat PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 3 / 17 PLOS ONE Hypusinated eIF5A in human diabetes Hyp anti-pancreatic polypeptide (abcam; 1:200), rabbit anti-eIF5A ([13,18]; 1:1000). Secondary antibodies including Alexa-488, Cy3, or Alexa-647 (Jackson Immunoresearch) were used, fol- lowed by DAPI (Sigma; 1:1000) to visualize nuclei. Images were acquired with a Zeiss 710 con- focal microscope. Human pancreas and spleen tissue Paraffin-embedded tissue sections were obtained from the nPOD consortium (www.jdrfnpod. org). A total of 10 nondiabetic donors, 4 donors with T2D, and 12 donors with T1D (6 autoan- tibody positive, 6 autoantibody negative) were included in this study (Tables 2 and 3). Infor- mation regarding donors’ demography, histology, and disease status were provided by nPOD. The autoantibody status was also determined by nPOD as previously described [23]. Immunofluorescence analysis of human tissues Immunofluorescent staining was performed as previously published [22] with modifications to account for the use of paraffin embedded tissue. Briefly, tissue sections were deparaffinized through graded ethanols (100%, 95%, 85%, 75%, 50%; Fisher Scientific) and then blocked using normal donkey serum (Sigma). Primary antibodies used included guinea pig anti-insulin (DAKO; 1:500), mouse anti-glucagon (Abcam; 1:500), rat anti-somatostatin (abcam; 1:200), goat anti-pancreatic polypeptide (abcam; 1:200), goat anti-ghrelin (Santa Cruz; 1:500), mouse anti-Pax5 (DAKO; 1:200), mouse anti-CD8 (Thermo Fisher; 1:500), mouse anti-CD4 (Leica; Hyp 1:500), rabbit anti-eIF5A ([13,18]; 1:1000). Secondary antibodies including Alexa-488, Cy3, or Alexa-647 (Jackson Immunoresearch) were used to visualize primary antibodies. DAPI (Sigma; 1:1000) was used to visualize nuclei. Images were acquired with a Zeiss 710 confocal microscope. Drug treatments of cultured cells and mouse islets HEK293T cells (ATCC #CRL-3216) were maintained in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% FBS (Hyclone; Fisher Scientific), 1% Penicillin/Streptomy- cin and 2 mM L-Glutamine. HEK293T cells were grown to 60–70% confluency on coverslips in a 24-well plate and treated with 10 or 100 μM N1-Guanyl-1,7-diaminoheptane (GC7) [24] (or 10 mM acetic acid; vehicle) with 0.5 mM aminoguanidine for 16 hours. Cells were fixed with 4% paraformaldehyde in PBS and blocked for 30 min in 3% bovine serum albumin (BSA) in PBS followed by permeabilization with 0.2% Triton X-100 in BSA-PBS for 10 min. All anti- bodies were diluted in 3% BSA-PBS and were applied in sequential order. Cells were incubated Hyp with rabbit anti-eIF5A ([13,18], 1:1000) overnight at 4˚C followed by a one hour incubation with anti-rabbit 488 (Jackson Immunoresearch; 1:500) at room temperature. Coverslips were Table 2. Human donor pancreas and spleen tissue from T2D and matched controls. nPOD case # Sample name Age (years) Gender (male/female) Ethnicity BMI T2D (yes/no) C-peptide (ng/mL) HbA1c nPOD-6097 F1-control 43.1 female Caucasian 36.4 no 16.76 7.1 nPOD-6102 F2-control 45.1 female Caucasian 35.1 no 0.55 6.1 nPOD-6132 F1-T2D 55.8 female Hispanic 44.6 yes 0.8 9.1 nPOD-6109 F2-T2D 48.8 female Hispanic 32.5 yes <0.05 8 nPOD-6060 M1-control 24 male Caucasian 32.7 no 13.63 N/A nPOD-6091 M2-control 27.1 male Caucasian 35.6 no 7.71 6.3 nPOD-6114 M1-T2D 42.8 male Caucasian 31 yes 0.58 7.8 nPOD-6188 M2-T2D 36.1 male Hispanic 30.6 yes 3.45 7.2 https://doi.org/10.1371/journal.pone.0230627.t002 PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 4 / 17 PLOS ONE Hypusinated eIF5A in human diabetes Table 3. Human donor pancreas and spleen tissue from T1D and matched controls. nPOD case Sample name Age Gender (male/ Ethnicity BMI T1D (yes/ c-peptide Auto Antibodies Detected Duration of disease # (years) female) no) (years) nPOD-6179 1-control 21.8 female Caucasian 20.7 no 2.74 N/A N/A nPOD-6224 1-AAb- 21 female Caucasian 22.8 yes <0.05 negative 1.5 negative nPOD-6070 1-AAb- 22.6 female Caucasian 21.6 yes <0.05 mIAA+;1A-2A+ 7 positive nPOD-6034 2-control 32 female Caucasian 25.2 no 3.15 N/A N/A nPOD-6121 2-AAb- 33.9 female Caucasian 18 yes 0.24 negative 4 negative nPOD-6143 2-AAb- 32.6 female Caucasian 26.1 yes <0.05 1A-2A+;mIAA+ 7 positive nPOD-6229 3-control 31 female Caucasian 26.9 no 6.23 N/A N/A nPOD-6208 3-AAb- 32 female Caucasian 23.4 yes <0.05 negative 16 negative nPOD-6077 3-AAb- 32.9 female Caucasian 22 yes <0.05 mIAA+ 18 positive nPOD-6015 4-control 39 female Caucasian 32.2 no 1.99 N/A N/A nPOD-6038 4-AAb- 37.2 female Caucasian 30.9 yes 0.2 negative 20 negative nPOD-6054 4-AAb- 35.1 female Caucasian 30.4 yes <0.05 mIAA+ 30 positive nPOD-6055 5-control 27 male Caucasian 22.7 no 0.59 N/A N/A nPOD-6041 5-AAb- 26.3 male Caucasian 28.4 yes <0.05 negative 10 negative nPOD-6180 5-AAb- 27.1 male Caucasian 25.9 yes <0.05 GADA+;1A-2A+; ZnT8A+; 11 positive mIAA+ nPOD-6104 6-control 41 male Caucasian 20.5 no 20.55 N/A N/A nPOD-6173 6-AAb- 44.1 male Caucasian 23.9 yes <0.05 negative 15 negative nPOD-6141 6-AAb- 36.7 male Caucasian 26 yes <0.05 GADA+;1A-2A+; ZnT8A+; 28 positive mIAA+ https://doi.org/10.1371/journal.pone.0230627.t003 washed with PBS, and DAPI (Sigma; 1:1000) used to visualize nuclei. Coverslips were mounted and images acquired with a Zeiss 710 confocal microscope. Mouse islets were isolated as previously described [21] and subsequently cultured in RPMI 1640 media supplemented with 10% FBS (Hyclone; Fisher Scientific), and 1% Penicillin/Strep- tomycin. Islets were treated with 100 μM GC7 (or 10 mM acetic acid; vehicle) with 0.5 mM aminoguanidine for 72 hours; the duration of treatment to reduce hypusination was previously determined in [10]. Islet were fixed with 4% paraformaldehyde in PBS and blocked for 1 hour in 5% normal donkey serum (NDS) in PBS followed by permeabilization with 0.1% Triton X- 100 in NDS-PBS for 10 min. Islets were stained as described above using chambered slides; maximum intensity projection images were processed from Z-stack image collections acquired with a Zeiss 710 confocal microscope. Results Hyp Beta cell and non-beta cell distribution of eIF5A in mouse We previously developed and characterized a novel antibody that recognizes the unique amino acid hypusine, formed exclusively through posttranslational modification of the Lys50 Hyp residue of eIF5A (eIF5A ) [12,18]. In this study, we utilized this antibody to investigate the PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 5 / 17 PLOS ONE Hypusinated eIF5A in human diabetes Hyp Fig 1. Expression of hypusinated eIF5A (eIF5A ) in mouse and human pancreatic islets. (A) Western blot from mouse and human pancreas tissue and isolated pancreatic islets. (B) Western blot from FACS sorted mouse islet cell populations. (C, D) Representative immunofluorescence images of mouse tissue demonstrating robust expression of Hyp eIF5A in PP-expressing cells. https://doi.org/10.1371/journal.pone.0230627.g001 Hyp expression of eIF5A in mouse and human pancreas tissue and isolated islets as well as Hyp human spleen tissue, to characterize the expression pattern of eIF5A and determine if Hyp eIF5A -expressing cells stratify with characteristics of disease. To that end, we first con- Hyp firmed the presence of eIF5A in islets isolated from mouse and human pancreas as well as in mouse pancreas and human acinar (exocrine) tissue (Fig 1A, S1 Fig). PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 6 / 17 PLOS ONE Hypusinated eIF5A in human diabetes Tomato We next utilized the RIP-cre;R26R mouse model wherein the insulin-expressing cells were labeled with a lineage trace, thereby generating beta cells indelibly marked with fluores- Tomato cent reporter (Tomato) expression. Islet cells from RIP-cre;R26R and control animals were sorted by FACS, using the presence and absence of Tomato expression to separate cells into two populations: beta cells (Tomato-positive) and non-beta cells (Tomato-negative). The cell types represented in the “non-beta cell” sample included (ordered from largest population to smallest): glucagon-expressing alpha cells, somatostatin-expressing delta cells, pancreatic polypeptide-expressing PP cells, ghrelin-expressing epsilon cells, exocrine cells (a possible con- taminant from the process of islet isolation) and support cells including endothelial cells. A similar quantity of Tomato-positive beta cells (1.92x10 cells) and Tomato-negative non-beta cells (2.13x10 cells) were collected (S2 Fig). Subsequent western blot analysis identified that Hyp eIF5A was present in nearly identical abundance in both the beta cell (Tomato-positive) and non-beta cell (Tomato-negative) populations (Fig 1B, S3 Fig). The expression of Pdx1 con- firms the enrichment of beta cells in the Tomato positive cells; the lower level of Pdx1 expres- sion in the non-beta cell fraction can be attributed to the presence of somatostatin-expressing Hyp delta cells. These data demonstrate that eIF5A is expressed in both the beta cell and non- Hyp beta cell fractions; however, whether there is a differential expression of eIF5A in a specific non-beta cell type(s) cannot be clarified from these data. Therefore, to characterize the spatial Hyp distribution of eIF5A expression pattern in the islet, we performed co-immunofluorescence Hyp staining for eIF5A and islet hormones in mouse pancreas tissue. Whereas relatively weak Hyp immunostaining of eIF5A was found throughout the pancreas and islets, robust immunos- Hyp taining of eIF5A was found in the islet cell population that expressed pancreatic polypep- tide (Fig 1C and 1D). To confirm that the high expressing cell population was not an artifact and that our previ- Hyp ously published antibody [12,18] could detect expression of eIF5A by immunofluorescence, we treated HEK293T cells (human) or isolated pancreatic islets (mouse) with the DHPS inhibitor GC7 (N1-Guanyl-1,7-diaminoheptane) [24]. Cells and islets were then analyzed for Hyp eIF5A by immunofluorescence. Whereas the control HEK293T uniformly expressed Hyp Hyp eIF5A , the mouse islets contained cells with both weak and robust expression of eIF5A . Furthermore, following treatment with the inhibitor, we observed a reduction in expression of Hyp eIF5A in both the HEK293T cells and mouse islets compared with vehicle treated controls Hyp (S4 Fig), which verified the measurement of eIF5A expression using our antibody. Hyp eIF5A -expressing cells in the pancreas of human type 2 diabetes Hyp To characterize the expression pattern of eIF5A in the human pancreas, we utilized tissue samples from the Network of Pancreatic Organ Donors with Diabetes (nPOD). A cohort of tis- sues from donors with and without T2D were provided (Table 2). Both pancreas and spleen tissues were acquired from each donor; age, gender, ethnicity and BMI were matched where possible. Given the relatively small size of the cohort, quantitative evaluations were not possi- Hyp ble. Therefore, we evaluated the presence or absence of eIF5A , its cell-type expression pat- tern, and its expression correlation with disease. Hyp Pancreas tissue sections were co-immunostained with the eIF5A -specific antibody and antibodies that recognized the hormones expressed by each of the endocrine cell populations in the islet (insulin, glucagon, somatostatin, ghrelin and pancreatic polypeptide). Robust co- Hyp localization was not observed between eIF5A and insulin (Fig 2A and 2B), glucagon (Fig 2C and 2D), ghrelin (Fig 2E and 2F), or somatostatin (Fig 2G and 2H). However, as observed in the mouse pancreas, cells expressing pancreatic polypeptide were identified to co-express Hyp high levels of eIF5A in control pancreas tissue (Fig 3A). These cells also expressed PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 7 / 17 PLOS ONE Hypusinated eIF5A in human diabetes Hyp Fig 2. The expression pattern of eIF5A in T2D and control pancreatic tissue. In controls (matched for age, gender and BMI) and Hyp T2D pancreas, we evaluated the co-expression of eIF5A with all islet hormones and found no overlap with insulin (A,B), glucagon (C, D), ghrelin (E,F) or somatostatin (G,H). https://doi.org/10.1371/journal.pone.0230627.g002 chromograninA, which confirms their identity as neuroendocrine cells (Fig 3B). The co-locali- Hyp zation of eIF5A with pancreatic polypeptide in the PP-expressing cells was observed in pan- creas tissues from donors with T2D (Fig 3C and 3D) and non-diabetic controls, suggesting no Hyp differential expression related to disease status. Notably, whereas PP and eIF5A were expressed in the same cells, the expression pattern is suggestive of localization in different compartments (Fig 3E). Hyp Spleen tissue sections from the same donors were co-immunostained with eIF5A and markers of various cell types. In particular, Pax5-expressing B cells, CD4-expressing T cells, Hyp and CD8-expressing T cells were evaluated for co-expression of eIF5A . Whereas the expres- Hyp sion patterns observed suggest that most Pax5+ B cells expressed eIF5A , only a select group Hyp of eIF5A -expressing cells appear to co-expressed either CD4 or CD8 (Fig 4A–4C; S5–S7 Figs). No obvious differences in staining intensity or distribution were observed between sam- ples from T2D and controls (Fig 4D–4E). Hyp eIF5A -expressing cells in the pancreas of human type 1 diabetes Donor pancreas and spleen tissue from individuals with T1D were also acquired from nPOD Hyp and evaluated for the expression pattern of eIF5A . This cohort of samples included T1D donors that were autoantibody-positive and autoantibody-negative, with both short and long disease duration; non-diabetic controls were matched for age, gender, ethnicity and BMI (Table 3). Similar to the T2D/control samples, we identified cells co-expressing the hormone Hyp PP with high intensity eIF5A immunostaining (Fig 5A–5F); robust co-expression of eIF5A- Hyp Hyp with other islet hormones was not observed. Moreover, the eIF5A -expressing cells expressed ChromograninA (Fig 5G–5I), which again confirmed that these cells are neuroen- docrine in nature. Evaluation of spleen tissue for all T1D donors and controls revealed an identical pattern of expression to that observed in the T2D donors and controls. Specifically, Hyp the majority of eIF5A -expressing cells co-expressed Pax5 (Fig 6; S8–S10 Figs). PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 8 / 17 PLOS ONE Hypusinated eIF5A in human diabetes Hyp Fig 3. eIF5A is robustly expressed in the pancreatic polypeptide-expressing PP cells in the islet. (A-D) In both Hyp controls and T2D pancreas, co-expression of eIF5A with pancreatic polypeptide (PP) and chromograninA (ChgA) Hyp was observed. (E) An expression pattern of eIF5A in the PP cells suggestive of localization to the ER was observed in cells in both controls and T2D pancreas. https://doi.org/10.1371/journal.pone.0230627.g003 Discussion Previous data from mouse models identified that pharmacological modulation of the hypusi- nation of eIF5A enhanced beta cell mass and improved glucose tolerance in mouse models of Hyp both T1D and T2D [14,16], thereby suggesting an important role for eIF5A in the setting of Hyp diabetes. However, to translate these findings to human, a greater understanding of eIF5A in the human pancreas and spleen would be required. This study represents the first descrip- Hyp tion of eIF5A expression in human organs from donors with and without diabetes. PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 9 / 17 PLOS ONE Hypusinated eIF5A in human diabetes Hyp Hyp Fig 4. eIF5A expression pattern in the spleen of control and T2D. eIF5A is expressed in immune cells in the spleen. We Hyp evaluated expression of eIF5A in Pax5+ B cells, CD4+ T cells and CD8+ T cells in the spleens of donors with T2D and controls Hyp Hyp matched for age, gender and BMI. Most eIF5A + cells co-expressed Pax5+ (A,D); however, a select group of eIF5A + cells expressed either CD4+ (B,E) or CD8+ (C,F). https://doi.org/10.1371/journal.pone.0230627.g004 Hyp Importantly, our results reveal a heretofore unappreciated enrichment of eIF5A in subsets of endocrine cells in the pancreas and immune cells in the spleen. Interestingly, the presence Hyp of eIF5A co-expressing cells was not obviously enhanced in diseased tissue; however, larger cohorts are required where tissue can be sampled from across whole organs in order to pre- cisely quantitate the presence of these cells and definitively determine correlation with disease. Further investigation is also required as to the relative abundance of the deoxyhypusine and hypusine forms of eIF5A in both the normal and diseased setting. Currently, published work suggests that the deoxyhypusine form of eIF5A is transient and reversible, and therefore not as abundant due to its rapid modification by the enzyme DOHH (deoxyhypusine hydroxylase) to the hypusinated form of eIF5A during the process of hypusine biosynthesis [25]. Our antibody is specific for the modified forms of eIF5A over the unmodified forms [18], and the form most detectable by our antibody due to its high prevalence is the more stable hypusinated form of Hyp eIF5A (eIF5A ). Further investigation and tool development will be required to determine and understand the relative abundance of the deoxyhypusine and hypusine forms of eIF5A in both the normal and diseased settings. Hyp Our findings in the pancreas demonstrate that eIF5A is expressed in both the exocrine and endocrine compartments in mouse and human. Previous reports have also shown expres- Hyp sion of eIF5A in mouse islets [10,11]. However, our immunoblot of sorted mouse islet cells Hyp further defined that eIF5A expression in the islet can be found in both the beta cell and non-beta cell populations. The non-beta cell populations encompass multiple hormone- PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 10 / 17 PLOS ONE Hypusinated eIF5A in human diabetes Hyp Fig 5. Expression of eIF5A in the T1D pancreas. (A-F) Identical to the pattern identified in T2D and control tissues, high Hyp expression of eIF5A is observed in PP cells in the T1D, both auto-antibody positive (aAb+) and auto-antibody negative(aAb-), pancreas and controls (matched for age, gender, ethnicity and BMI). (G-I) In all cases, these cells express the endocrine cell marker ChromograninA (ChgA). https://doi.org/10.1371/journal.pone.0230627.g005 expressing cell types, and our immunostaining analysis of mouse tissue clarified that the most Hyp robust expression of eIF5A is in the PP cell population. Given the over-representation of PP cells in the uncinate region of the pancreas [26], we analyzed tissue sections that contained Hyp the uncinate region and found that eIF5A is robustly co-expressed in PP cells of human islets. Despite evidence that PP cells have a critical secretory function in the brain-gut axis [27] and may serve as a regulator of intra-islet secretion [28], the role of PP cells in the context of diabetes has received little attention. From a developmental perspective, PP cells are predomi- nantly derived from the ghrelin-expressing cell lineage found in the embryonic pancreas [29]; Hyp however, the function of eIF5A in the PP cell population postnatally or any function for Hyp eIF5A in the development of PP cells has yet to be elucidated. Interestingly, expression analysis of 12-lipoxygenase, a factor known to promote inflammation in the setting of diabetes, PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 11 / 17 PLOS ONE Hypusinated eIF5A in human diabetes Hyp Fig 6. Expression of eIF5A in spleen tissue from donors with T1D and matched control donors. We examined spleen tissue from persons with autoantibody positive (AAb+) and auto-antibody negative (AAb-) T1D, and corresponding controls matched for age, gender, ethnicity, and BMI. As observed in T2D and matched control spleen Hyp Hyp tissue, most eIF5A -expressing cells were Pax5+ (A-C); however, some eIF5A + cells expressed either CD4+ (D-F) or CD8+ (G-I). https://doi.org/10.1371/journal.pone.0230627.g006 is also increased in the PP-expressing cell population in pancreas tissue from human donors (collected through nPOD; [30]). Clearly, a greater understanding is required for the role of PP cells in the pathogenesis of diabetes. Given that much of the published and ongoing work on hypusine biosynthesis in mice has Hyp Hyp studied eIF5A in the context of diabetes, we had hypothesized that eIF5A expression would be identified predominantly in the insulin-producing beta cell population. Our western Hyp Hyp blot analysis did reveal eIF5A expression in human islets. Moreover, we observed eIF5A expression in a purified population of beta cells (Tomato+) from mouse islets. Interestingly, Hyp the quantitative nature of western blots indicates that the expression of eIF5A must be lower in the purified beta cells compared with non-beta cells given that PP cells comprise only a small portion of the Tomato(-) non-beta cell fraction whereas the Tomato(+) fraction is Hyp composed exclusively of beta cells, and we see near equivalent expression of eIF5A in both sorted populations. This finding is consistent with the immunofluorescence data, wherein we PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 12 / 17 PLOS ONE Hypusinated eIF5A in human diabetes Hyp identified robust expression of eIF5A in PP-expressing cells. The lack of observable co- Hyp expression of eIF5A and hormones in all cell types in the islet was unexpected; however, we must consider the possibility that this may be due to the limitations in detection of low protein expression by immunofluorescence. Therefore, and considering all data together, our observa- Hyp tions indicate the presence of eIF5A in both beta cells and non-beta cells, with particularly high expression in one small non-beta cell populations, the PP cells. Our previous finding that pharmacological inhibition of eIF5A hypusination (using the drug GC7; N1-Guanyl-1,7-diaminoheptane) in NOD mice improved glucose tolerance and preserved beta cell mass [14]. These improvements were also accompanied by reductions in insulitis, which led us to question whether the improvements in beta cell function were due to a direct effect of DHPS inhibition in beta cells, or an indirect effect related to DHPS inhibition Hyp in infiltrating immune cells. Our work and that of others suggest a role for eIF5A and DHPS in promoting T cell and B cell proliferation [14,31,32], which was the basis for our Hyp hypothesis that perhaps eIF5A is differentially expressed in immune cells in individuals with diabetes compared with controls. However, identical expression patterns were noted in Hyp all spleen tissue evaluated. The identical expression patterns of eIF5A between healthy and disease in both the immune cell populations and islet cell populations could also suggest that it Hyp is not the abundance of eIF5A that is critical for promotion of disease. Rather, the presence Hyp of eIF5A facilitating the translation of different mRNAs in the disease setting compared with the healthy setting could drive pathogenesis. Given our recent findings that deletion of Dhps in adult mouse beta cells results in reduced diet-induced beta cell proliferation and sub- sequent glucose intolerance due to altered translation of cyclinD2 [10], we are now investigat- Hyp ing the impact of eIF5A on mRNA translation in other diabetes-related cell populations. Supporting information S1 Fig. Source images for western blots of mouse and human pancreas and islets. Hyp (A) Immunoblot images show expression of eIF5A , insulin and total eIF5A in cell lysates from mouse whole pancreas and isolated islets. The top portions of these blots were proved with antibodies not related to this study. (B) Total protein expression as visualized by Pon- Hyp ceauS staining. (C) Immunoblot images show expression of eIF5A , insulin and total eIF5A in cell lysate from human exocrine tissue and isolated islets. One blot was probed twice, and this second antibody was not related to this study. (D) Total protein expression as visualized by PonceauS staining. The dotted box shows the lanes where the exocrine and islet samples were run; the other samples are unrelated to this study. (PDF) S2 Fig. Collection of beta cell and non-beta cell populations by FACS. (A) Islets isolated from multiple RIP-cre;R26RTomato mice were pooled together and processed for fluorescence activated cell sorting (FACS). Islet cells were sorted into two populations: Tomato-positive beta cells (R4), and Tomato-negative non-beta cells (R3, islet cells expressing glucagon, somatostatin, ghrelin, and pancreatic polypeptide). (PDF) S3 Fig. Source images for western blots of FACS sorted mouse islet cell populations. Hyp (A) Immunoblot for expression of Pdx1 and eIF5A in cell lysates from Tomato-negative non-beta cells and Tomato-positive beta cells. (B) Immunoblot for expression of total eIF5A. (C) Total protein expression as visualized by PonceauS staining. (PDF) PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 13 / 17 PLOS ONE Hypusinated eIF5A in human diabetes S4 Fig. Evaluation of eIF5AHyp expression following treatment with DHPS inhibitor. HEK293T cells were treated with the DHPS inhibitor GC7 (N1-Guanyl-1,7,diaminoheptane) and analyzed for eIF5AHyp expression by immunofluorescence. (A) Control HEK293T cells Hyp uniformly expressed eIF5A . (B,C) Treatment with GC7 resulted in reduced expression of Hyp eIF5A . (D) A secondary antibody control was also performed to confirm that the observed signal was not an artifact. Mouse pancreatic islets were also treated with GC7 and analyzed for Hyp eIF5A expression by immunofluorescence. (E) Control mouse islets contained cells with Hyp both weak and robust expression of eIF5A . (F, G) Islets treated with GC7 showed a reduc- Hyp tion in expression of eIF5A . Images are 20X. Inset images are higher magnification of the areas outlined with white boxes. (PDF) Hyp S5 Fig. eIF5A expression pattern in the Pax5-expressing cell population in spleen tissue Hyp of control and T2D. We evaluated the expression of eIF5A in Pax5-expressing B cells in the spleens of donors with T2D and controls matched for age, gender and BMI. The fluores- cent channels have been separated to better display the expression patterns of the Pax5-expres- Hyp sing B cells (A, B), eIF5A -expressing cells (C, D), and the overlap between the Pax5- Hyp expressing and eIF5A -expressing populations (E, F). All images are 20X. (PDF) Hyp S6 Fig. eIF5A expression pattern in the CD4-expressing cell population in spleen tissue Hyp of control and T2D. We evaluated the expression of eIF5A in CD4-expressing T cells in the spleens of donors with T2D and controls matched for age, gender and BMI. The fluores- cent channels have been separated to better display the expression patterns of the CD4-expres- Hyp sing T cells (A, B), eIF5A -expressing cells (C, D), and the minimal overlap between the Hyp CD4-expressing and eIF5A -expressing populations (E, F). All images are 20X. (PDF) Hyp S7 Fig. eIF5A expression pattern in the CD8-expressing T cell population in spleen tis- Hyp sue of control and T2D. We evaluated the expression of eIF5A in CD8-expressing T cells in the spleens of donors with T2D and controls matched for age, gender and BMI. The fluores- cent channels have been separated to better display the expression patterns of the CD8-expres- Hyp sing T cells (A, B), eIF5A -expressing cells (C, D), and the minimal overlap between the Hyp CD8-expressing and eIF5A -expressing populations (E, F). All images are 20X. (PDF) Hyp S8 Fig. eIF5A expression pattern in the Pax5-expressing B cell population in spleen tis- Hyp sue of control and T1D. We evaluated the expression of eIF5A in Pax5-expressing B cells in the spleens of donors with auto-antibody positive (AAb+) and auto-antibody negative (AAb-) T1D, and corresponding controls matched for age, gender, ethnicity, and BMI. The fluorescent channels have been separated to better display the expression patterns of the Pax5- Hyp expressing B cells (A—C), eIF5A -expressing cells (D—F), and the overlap between the Hyp Pax5-expressing and eIF5A -expressing populations (G—I). All images are 20X. (PDF) Hyp S9 Fig. eIF5A expression pattern in the CD4-expressing T cell population in spleen tis- Hyp sue of control and T1D. We evaluated the expression of eIF5A in CD4-expressing T cells in the spleens of donors with auto-antibody positive (AAb+) and auto-antibody negative (AAb-) T1D, and corresponding controls matched for age, gender, ethnicity, and BMI. The fluorescent channels have been separated to better display the expression patterns of the Hyp CD4-expressing T cells (A—C), eIF5A -expressing cells (D—F), and the minimal overlap PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 14 / 17 PLOS ONE Hypusinated eIF5A in human diabetes Hyp between the CD4-expressing and eIF5A -expressing populations (G—I). All images are 20X. (PDF) Hyp S10 Fig. eIF5A expression pattern in the CD8-expressing T cell population in spleen tis- Hyp sue of control and T1D. We evaluated the expression of eIF5A in CD8-expressing T cells in the spleens of donors with auto-antibody positive (AAb+) and auto-antibody negative (AAb-) T1D, and corresponding controls matched for age, gender, ethnicity, and BMI. The fluorescent channels have been separated to better display the expression patterns of the Hyp CD8-expressing T cells (A—C), eIF5A -expressing cells (D—F), and the minimal overlap Hyp between the CD8-expressing and eIF5A -expressing populations (G—I). All images are 20X. (PDF) Acknowledgments The authors wish to thank Dr. David Morris and the Flow Cytometry Core Facility at Indiana University School of Medicine for assistance with FACS. Human pancreatic islets were pro- vided by the NIDDK-funded Integrated Islet Distribution Program (IIDP) at City of Hope, NIH Grant # 2UC4DK098085. Human donor acinar tissue was provided by Dr. Rita Bottino at the Center for Organ Recovery and Education (CORE), Pittsburg PA. This research was also performed with the support of the Network for Pancreatic Organ donors with Diabetes (nPOD; RRID:SCR_014641), a collaborative type 1 diabetes research project sponsored by JDRF (nPOD: 5-SRA-2018-557-Q-R) and The Leona M. & Harry B. Helmsley Charitable Trust (Grant#2018PG-T1D053). Organ Procurement Organizations (OPO) partnering with nPOD to provide research resources are listed at http://www.jdrfnpod.org//for-partners/ npod-partners/. This manuscript was released as a preprint at bioRxiv https://www.biorxiv. org/content/10.1101/745919v1. Author Contributions Conceptualization: Teresa L. Mastracci, Raghavendra G. Mirmira. Data curation: Teresa L. Mastracci, Leah R. Padgett. Formal analysis: Teresa L. Mastracci, Stephanie C. Colvin, Leah R. Padgett, Raghavendra G. Mirmira. Funding acquisition: Teresa L. Mastracci, Raghavendra G. Mirmira. Investigation: Teresa L. Mastracci, Stephanie C. Colvin, Leah R. Padgett, Raghavendra G. Mirmira. Methodology: Teresa L. Mastracci, Stephanie C. Colvin, Leah R. Padgett. Project administration: Teresa L. Mastracci. Resources: Teresa L. Mastracci, Raghavendra G. Mirmira. Software: Teresa L. Mastracci. Supervision: Teresa L. Mastracci, Raghavendra G. Mirmira. Validation: Teresa L. Mastracci. Visualization: Teresa L. Mastracci, Leah R. Padgett. PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 15 / 17 PLOS ONE Hypusinated eIF5A in human diabetes Writing – original draft: Teresa L. Mastracci, Raghavendra G. Mirmira. Writing – review & editing: Teresa L. Mastracci, Stephanie C. Colvin, Leah R. Padgett, Ragha- vendra G. Mirmira. References 1. Park MH, Cooper HL, Folk JE. Identification of hypusine, an unusual amino acid, in a protein from human lymphocytes and of spermidine as its biosynthetic precursor. Proc Natl Acad Sci U S A. 1981; 78: 2869–2873. https://doi.org/10.1073/pnas.78.5.2869 PMID: 6789324 2. Park MH. The post-translational synthesis of a polyamine-derived amino acid, hypusine, in the eukary- otic translation initiation factor 5A (eIF5A). J Biochem. 2006; 139: 161–9. https://doi.org/10.1093/jb/ mvj034 PMID: 16452303 3. Saini P, Eyler DE, Green R, Dever TE. Hypusine-containing protein eIF5A promotes translation elonga- tion. Nature. 2009; 459: 118–21. https://doi.org/10.1038/nature08034 PMID: 19424157 4. Park MH, Wolff EC, Lee YB, Folk JE. Antiproliferative effects of inhibitors of deoxyhypusine synthase. Inhibition of growth of Chinese hamster ovary cells by guanyl diamines. J Biol Chem. 1994; 269: 27827–32. PMID: 7961711 5. Wolff EC, Kang KR, Kim YS, Park MH. Posttranslational synthesis of hypusine: evolutionary progres- sion and specificity of the hypusine modification. Amino Acids. 2007; 33: 341–50. https://doi.org/10. 1007/s00726-007-0525-0 PMID: 17476569 6. Schuller AP, Wu CC-C, Dever TE, Buskirk AR, Green R. eIF5A Functions Globally in Translation Elon- gation and Termination. Mol Cell. 2017; 66: 194–205.e5. https://doi.org/10.1016/j.molcel.2017.03.003 PMID: 28392174 7. Gutierrez E, Shin BS, Woolstenhulme CJ, Kim JR, Saini P, Buskirk AR, et al. eIF5A promotes transla- tion of polyproline motifs. Mol Cell. 2013; 51: 35–45. https://doi.org/10.1016/j.molcel.2013.04.021 PMID: 23727016 8. McDuffie M, Maybee NA, Keller SR, Stevens BK, Garmey JC, Morris MA, et al. Nonobese diabetic (NOD) mice congenic for a targeted deletion of 12/15-lipoxygenase are protected from autoimmune dia- betes. Diabetes. 2008; 57: 199–208. https://doi.org/10.2337/db07-0830 PMID: 17940120 9. Serreze DV, Chapman HD, Varnum DS, Gerling I, Leiter EH, Shultz LD. Initiation of autoimmune diabe- tes in NOD/Lt mice is MHC class I-dependent. J Immunol Baltim Md 1950. 1997; 158: 3978–3986. 10. Levasseur EM, Yamada K, Piñeros AR, Wu W, Syed F, Orr KS, et al. Hypusine biosynthesis inβ cells links polyamine metabolism to facultative cellular proliferation to maintain glucose homeostasis. Sci Sig- nal. 2019; 12. https://doi.org/10.1126/scisignal.aax0715 PMID: 31796630 11. Tersey SA, Colvin SC, Maier B, Mirmira RG. Protective effects of polyamine depletion in mouse models of type 1 diabetes: implications for therapy. Amino Acids. 2014; 46: 633–42. https://doi.org/10.1007/ s00726-013-1560-7 PMID: 23846959 12. Maier B, Ogihara T, Trace AP, Tersey SA, Robbins RD, Chakrabarti SK, et al. The unique hypusine modification of eIF5A promotes islet beta cell inflammation and dysfunction in mice. J Clin Invest. 2010; 120: 2156–70. https://doi.org/10.1172/JCI38924 PMID: 20501948 13. Maier B, Tersey SA, Mirmira RG. Hypusine: a new target for therapeutic intervention in diabetic inflam- mation. Discov Med. 2010; 10: 18–23. PMID: 20670594 14. Colvin SC, Maier B, Morris DL, Tersey SA, Mirmira RG. Deoxyhypusine synthase promotes differentia- tion and proliferation of T helper type 1 (Th1) cells in autoimmune diabetes. J Biol Chem. 2013; 288: 36226–35. https://doi.org/10.1074/jbc.M113.473942 PMID: 24196968 15. Coleman DL. Obese and diabetes: two mutant genes causing diabetes-obesity syndromes in mice. Dia- betologia. 1978; 14: 141–148. https://doi.org/10.1007/bf00429772 PMID: 350680 16. Robbins RD, Tersey SA, Ogihara T, Gupta D, Farb TB, Ficorilli J, et al. Inhibition of deoxyhypusine synthase enhances islet {beta} cell function and survival in the setting of endoplasmic reticulum stress and type 2 diabetes. J Biol Chem. 285: 39943–52. https://doi.org/10.1074/jbc.M110.170142 PMID: 17. Hatanaka M, Anderson-Baucum E, Lakhter A, Kono T, Maier B, Tersey SA, et al. Chronic high fat feed- ing restricts islet mRNA translation initiation independently of ER stress via DNA damage and p53 acti- vation. Sci Rep. 2017; 7. https://doi.org/10.1038/s41598-017-03869-5 PMID: 28630491 18. Nishiki Y, Farb TB, Friedrich J, Bokvist K, Mirmira RG, Maier B. Characterization of a novel polyclonal anti-hypusine antibody. Springerplus. 2013; 2: 421. https://doi.org/10.1186/2193-1801-2-421 PMID: PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 16 / 17 PLOS ONE Hypusinated eIF5A in human diabetes 19. Postic C, Shiota M, Niswender KD, Jetton TL, Chen Y, Moates JM, et al. Dual roles for glucokinase in glucose homeostasis as determined by liver and pancreatic beta cell-specific gene knock-outs using Cre recombinase. J Biol Chem. 1999; 274: 305–15. https://doi.org/10.1074/jbc.274.1.305 PMID: 20. Madisen L, Zwingman TA, Sunkin SM, Oh SW, Zariwala HA, Gu H, et al. A robust and high-throughput Cre reporting and characterization system for the whole mouse brain. Nat Neurosci. 2010; 13: 133–40. https://doi.org/10.1038/nn.2467 PMID: 20023653 21. Stull ND, Breite A, McCarthy R, Tersey SA, Mirmira RG. Mouse Islet of Langerhans Isolation using a Combination of Purified Collagenase and Neutral Protease. J Vis Exp JoVE. 2012 [cited 30 Mar 2018]. https://doi.org/10.3791/4137 PMID: 22987198 22. Mastracci TL, Anderson KR, Papizan JB, Sussel L. Regulation of Neurod1 contributes to the lineage potential of Neurogenin3+ endocrine precursor cells in the pancreas. PLoS Genet. 2013; 9: e1003278. https://doi.org/10.1371/journal.pgen.1003278 PMID: 23408910 23. Campbell-Thompson M, Wasserfall C, Kaddis J, Albanese-O’Neill A, Staeva T, Nierras C, et al. Net- work for Pancreatic Organ Donors with Diabetes (nPOD): developing a tissue biobank for type 1 diabe- tes. Diabetes Metab Res Rev. 2012; 28: 608–617. https://doi.org/10.1002/dmrr.2316 PMID: 22585677 24. Jakus J, Wolff EC, Park MH, Folk JE. Features of the spermidine-binding site of deoxyhypusine synthase as derived from inhibition studies. Effective inhibition by bis- and mono-guanylated diamines and polyamines. J Biol Chem. 1993; 268: 13151–13159. PMID: 8514754 25. Park MH, Nishimura K, Zanelli CF, Valentini SR. Functional significance of eIF5A and its hypusine mod- ification in eukaryotes. Amino Acids. 2010; 38: 491–500. https://doi.org/10.1007/s00726-009-0408-7 PMID: 19997760 26. Wang X, Zielinski MC, Misawa R, Wen P, Wang T-Y, Wang C-Z, et al. Quantitative Analysis of Pancre- atic Polypeptide Cell Distribution in the Human Pancreas. PLoS ONE. 2013; 8. https://doi.org/10.1371/ journal.pone.0055501 PMID: 23383206 27. Holzer P, Reichmann F, Farzi A. Neuropeptide Y, peptide YY and pancreatic polypeptide in the gut- brain axis. Neuropeptides. 2012; 46: 261–274. https://doi.org/10.1016/j.npep.2012.08.005 PMID: 28. Brereton MF, Vergari E, Zhang Q, Clark A. Alpha-, Delta- and PP-cells: Are They the Architectural Cor- nerstones of Islet Structure and Co-ordination? J Histochem Cytochem Off J Histochem Soc. 2015; 63: 575–591. https://doi.org/10.1369/0022155415583535 PMID: 26216135 29. Arnes L, Hill JT, Gross S, Magnuson MA, Sussel L. Ghrelin expression in the mouse pancreas defines a unique multipotent progenitor population. PloS One. 2012; 7: e52026. https://doi.org/10.1371/journal. pone.0052026 PMID: 23251675 30. Grzesik WJ, Nadler JL, Machida Y, Nadler JL, Imai Y, Morris MA. Expression pattern of 12-lipoxygen- ase in human islets with type 1 diabetes and type 2 diabetes. J Clin Endocrinol Metab. 2015; 100: E387–395. https://doi.org/10.1210/jc.2014-3630 PMID: 25532042 31. Bevec D, Jaksche H, Oft M, Wo ¨ hl T, Himmelspach M, Pacher A, et al. Inhibition of HIV-1 replication in lymphocytes by mutants of the Rev cofactor eIF-5A. Science. 1996; 271: 1858–1860. https://doi.org/10. 1126/science.271.5257.1858 PMID: 8596953 32. Schlee M, Krug T, Gires O, Zeidler R, Hammerschmidt W, Mailhammer R, et al. Identification of Epstein-Barr virus (EBV) nuclear antigen 2 (EBNA2) target proteins by proteome analysis: activation of EBNA2 in conditionally immortalized B cells reflects early events after infection of primary B cells by EBV. J Virol. 2004; 78: 3941–3952. https://doi.org/10.1128/JVI.78.8.3941-3952.2004 PMID: 15047810 PLOS ONE | https://doi.org/10.1371/journal.pone.0230627 March 24, 2020 17 / 17

Journal

PLoS ONEPublic Library of Science (PLoS) Journal

Published: Mar 24, 2020

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create folders to
organize your research

Export folders, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off