Research ArticleEndocrinologyTherapeutics Open Access | 10.1172/jci.insight.185299

Titli Nargis,1 Charanya Muralidharan,1 Jacob R. Enriquez,1 Jiayi E. Wang,1 Kerim B. Kaylan,1 Advaita Chakraborty,1 Sarida Pratuangtham,1 Kayla Figatner,1 Jennifer B. Nelson,1 Sarah C. May,1 Jerry L. Nadler,2 Matthew B. Boxer,3 David J. Maloney,3 Sarah A. Tersey,1 and Raghavendra G. Mirmira1

1Department of Medicine and the Kovler Diabetes Center, The University of Chicago, Chicago, Illinois, USA.

2Department of Pharmacology, New York Medical College, Valhalla, New York, USA.

3Veralox Therapeutics, Frederick, Maryland, USA.

Address correspondence to: Sarah A. Tersey, 900 E. 57th Street, KCBD 8152, Chicago, Illinois 60637, USA. Phone: 773.834.6928; Email: stersey@uchicago.edu. Or to: Raghavendra G Mirmira, 900 E. 57th Street, KCBD 8130, Chicago, Illinois 60637, USA. Phone: 773.702.2209; Email: mirmira@uchicago.edu.

Find articles by Nargis, T. in: JCI | PubMed | Google Scholar

1Department of Medicine and the Kovler Diabetes Center, The University of Chicago, Chicago, Illinois, USA.

2Department of Pharmacology, New York Medical College, Valhalla, New York, USA.

3Veralox Therapeutics, Frederick, Maryland, USA.

Address correspondence to: Sarah A. Tersey, 900 E. 57th Street, KCBD 8152, Chicago, Illinois 60637, USA. Phone: 773.834.6928; Email: stersey@uchicago.edu. Or to: Raghavendra G Mirmira, 900 E. 57th Street, KCBD 8130, Chicago, Illinois 60637, USA. Phone: 773.702.2209; Email: mirmira@uchicago.edu.

Find articles by Muralidharan, C. in: JCI | PubMed | Google Scholar |

1Department of Medicine and the Kovler Diabetes Center, The University of Chicago, Chicago, Illinois, USA.

2Department of Pharmacology, New York Medical College, Valhalla, New York, USA.

3Veralox Therapeutics, Frederick, Maryland, USA.

Address correspondence to: Sarah A. Tersey, 900 E. 57th Street, KCBD 8152, Chicago, Illinois 60637, USA. Phone: 773.834.6928; Email: stersey@uchicago.edu. Or to: Raghavendra G Mirmira, 900 E. 57th Street, KCBD 8130, Chicago, Illinois 60637, USA. Phone: 773.702.2209; Email: mirmira@uchicago.edu.

Find articles by Enriquez, J. in: JCI | PubMed | Google Scholar |

1Department of Medicine and the Kovler Diabetes Center, The University of Chicago, Chicago, Illinois, USA.

2Department of Pharmacology, New York Medical College, Valhalla, New York, USA.

3Veralox Therapeutics, Frederick, Maryland, USA.

Address correspondence to: Sarah A. Tersey, 900 E. 57th Street, KCBD 8152, Chicago, Illinois 60637, USA. Phone: 773.834.6928; Email: stersey@uchicago.edu. Or to: Raghavendra G Mirmira, 900 E. 57th Street, KCBD 8130, Chicago, Illinois 60637, USA. Phone: 773.702.2209; Email: mirmira@uchicago.edu.

Find articles by Wang, J. in: JCI | PubMed | Google Scholar

1Department of Medicine and the Kovler Diabetes Center, The University of Chicago, Chicago, Illinois, USA.

2Department of Pharmacology, New York Medical College, Valhalla, New York, USA.

3Veralox Therapeutics, Frederick, Maryland, USA.

Address correspondence to: Sarah A. Tersey, 900 E. 57th Street, KCBD 8152, Chicago, Illinois 60637, USA. Phone: 773.834.6928; Email: stersey@uchicago.edu. Or to: Raghavendra G Mirmira, 900 E. 57th Street, KCBD 8130, Chicago, Illinois 60637, USA. Phone: 773.702.2209; Email: mirmira@uchicago.edu.

Find articles by Kaylan, K. in: JCI | PubMed | Google Scholar

1Department of Medicine and the Kovler Diabetes Center, The University of Chicago, Chicago, Illinois, USA.

2Department of Pharmacology, New York Medical College, Valhalla, New York, USA.

3Veralox Therapeutics, Frederick, Maryland, USA.

Address correspondence to: Sarah A. Tersey, 900 E. 57th Street, KCBD 8152, Chicago, Illinois 60637, USA. Phone: 773.834.6928; Email: stersey@uchicago.edu. Or to: Raghavendra G Mirmira, 900 E. 57th Street, KCBD 8130, Chicago, Illinois 60637, USA. Phone: 773.702.2209; Email: mirmira@uchicago.edu.

Find articles by Chakraborty, A. in: JCI | PubMed | Google Scholar

1Department of Medicine and the Kovler Diabetes Center, The University of Chicago, Chicago, Illinois, USA.

2Department of Pharmacology, New York Medical College, Valhalla, New York, USA.

3Veralox Therapeutics, Frederick, Maryland, USA.

Address correspondence to: Sarah A. Tersey, 900 E. 57th Street, KCBD 8152, Chicago, Illinois 60637, USA. Phone: 773.834.6928; Email: stersey@uchicago.edu. Or to: Raghavendra G Mirmira, 900 E. 57th Street, KCBD 8130, Chicago, Illinois 60637, USA. Phone: 773.702.2209; Email: mirmira@uchicago.edu.

Find articles by Pratuangtham, S. in: JCI | PubMed | Google Scholar

1Department of Medicine and the Kovler Diabetes Center, The University of Chicago, Chicago, Illinois, USA.

2Department of Pharmacology, New York Medical College, Valhalla, New York, USA.

3Veralox Therapeutics, Frederick, Maryland, USA.

Address correspondence to: Sarah A. Tersey, 900 E. 57th Street, KCBD 8152, Chicago, Illinois 60637, USA. Phone: 773.834.6928; Email: stersey@uchicago.edu. Or to: Raghavendra G Mirmira, 900 E. 57th Street, KCBD 8130, Chicago, Illinois 60637, USA. Phone: 773.702.2209; Email: mirmira@uchicago.edu.

Find articles by Figatner, K. in: JCI | PubMed | Google Scholar

1Department of Medicine and the Kovler Diabetes Center, The University of Chicago, Chicago, Illinois, USA.

2Department of Pharmacology, New York Medical College, Valhalla, New York, USA.

3Veralox Therapeutics, Frederick, Maryland, USA.

Address correspondence to: Sarah A. Tersey, 900 E. 57th Street, KCBD 8152, Chicago, Illinois 60637, USA. Phone: 773.834.6928; Email: stersey@uchicago.edu. Or to: Raghavendra G Mirmira, 900 E. 57th Street, KCBD 8130, Chicago, Illinois 60637, USA. Phone: 773.702.2209; Email: mirmira@uchicago.edu.

Find articles by Nelson, J. in: JCI | PubMed | Google Scholar

1Department of Medicine and the Kovler Diabetes Center, The University of Chicago, Chicago, Illinois, USA.

2Department of Pharmacology, New York Medical College, Valhalla, New York, USA.

3Veralox Therapeutics, Frederick, Maryland, USA.

Address correspondence to: Sarah A. Tersey, 900 E. 57th Street, KCBD 8152, Chicago, Illinois 60637, USA. Phone: 773.834.6928; Email: stersey@uchicago.edu. Or to: Raghavendra G Mirmira, 900 E. 57th Street, KCBD 8130, Chicago, Illinois 60637, USA. Phone: 773.702.2209; Email: mirmira@uchicago.edu.

Find articles by May, S. in: JCI | PubMed | Google Scholar

1Department of Medicine and the Kovler Diabetes Center, The University of Chicago, Chicago, Illinois, USA.

2Department of Pharmacology, New York Medical College, Valhalla, New York, USA.

3Veralox Therapeutics, Frederick, Maryland, USA.

Address correspondence to: Sarah A. Tersey, 900 E. 57th Street, KCBD 8152, Chicago, Illinois 60637, USA. Phone: 773.834.6928; Email: stersey@uchicago.edu. Or to: Raghavendra G Mirmira, 900 E. 57th Street, KCBD 8130, Chicago, Illinois 60637, USA. Phone: 773.702.2209; Email: mirmira@uchicago.edu.

Find articles by Nadler, J. in: JCI | PubMed | Google Scholar

1Department of Medicine and the Kovler Diabetes Center, The University of Chicago, Chicago, Illinois, USA.

2Department of Pharmacology, New York Medical College, Valhalla, New York, USA.

3Veralox Therapeutics, Frederick, Maryland, USA.

Address correspondence to: Sarah A. Tersey, 900 E. 57th Street, KCBD 8152, Chicago, Illinois 60637, USA. Phone: 773.834.6928; Email: stersey@uchicago.edu. Or to: Raghavendra G Mirmira, 900 E. 57th Street, KCBD 8130, Chicago, Illinois 60637, USA. Phone: 773.702.2209; Email: mirmira@uchicago.edu.

Find articles by Boxer, M. in: JCI | PubMed | Google Scholar

1Department of Medicine and the Kovler Diabetes Center, The University of Chicago, Chicago, Illinois, USA.

2Department of Pharmacology, New York Medical College, Valhalla, New York, USA.

3Veralox Therapeutics, Frederick, Maryland, USA.

Address correspondence to: Sarah A. Tersey, 900 E. 57th Street, KCBD 8152, Chicago, Illinois 60637, USA. Phone: 773.834.6928; Email: stersey@uchicago.edu. Or to: Raghavendra G Mirmira, 900 E. 57th Street, KCBD 8130, Chicago, Illinois 60637, USA. Phone: 773.702.2209; Email: mirmira@uchicago.edu.

Find articles by Maloney, D. in: JCI | PubMed | Google Scholar

1Department of Medicine and the Kovler Diabetes Center, The University of Chicago, Chicago, Illinois, USA.

2Department of Pharmacology, New York Medical College, Valhalla, New York, USA.

3Veralox Therapeutics, Frederick, Maryland, USA.

Address correspondence to: Sarah A. Tersey, 900 E. 57th Street, KCBD 8152, Chicago, Illinois 60637, USA. Phone: 773.834.6928; Email: stersey@uchicago.edu. Or to: Raghavendra G Mirmira, 900 E. 57th Street, KCBD 8130, Chicago, Illinois 60637, USA. Phone: 773.702.2209; Email: mirmira@uchicago.edu.

Find articles by Tersey, S. in: JCI | PubMed | Google Scholar |

1Department of Medicine and the Kovler Diabetes Center, The University of Chicago, Chicago, Illinois, USA.

2Department of Pharmacology, New York Medical College, Valhalla, New York, USA.

3Veralox Therapeutics, Frederick, Maryland, USA.

Address correspondence to: Sarah A. Tersey, 900 E. 57th Street, KCBD 8152, Chicago, Illinois 60637, USA. Phone: 773.834.6928; Email: stersey@uchicago.edu. Or to: Raghavendra G Mirmira, 900 E. 57th Street, KCBD 8130, Chicago, Illinois 60637, USA. Phone: 773.702.2209; Email: mirmira@uchicago.edu.

Find articles by Mirmira, R. in: JCI | PubMed | Google Scholar |

Published November 12, 2024 - More info

Published in Volume 9, Issue 24 on December 20, 2024
JCI Insight. 2024;9(24):e185299. https://doi.org/10.1172/jci.insight.185299.
© 2024 Nargis et al. This work is licensed under the Creative Commons Attribution 4.0 International License. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. Published November 12, 2024 - Version history
Received: July 29, 2024; Accepted: October 31, 2024 View PDF Abstract

Type 1 diabetes (T1D) is characterized by the autoimmune destruction of insulin-producing β cells and involves an interplay between β cells and cells of the innate and adaptive immune systems. We investigated the therapeutic potential of targeting 12-lipoxygenase (12-LOX), an enzyme implicated in inflammatory pathways in β cells and macrophages, using a mouse model in which the endogenous mouse Alox15 gene is replaced by the human ALOX12 gene. Our finding demonstrated that VLX-1005, a potent 12-LOX inhibitor, effectively delayed the onset of autoimmune diabetes in human gene replacement non-obese diabetic mice. By spatial proteomics analysis, VLX-1005 treatment resulted in marked reductions in infiltrating T and B cells and macrophages, with accompanying increases in immune checkpoint molecule PD-L1, suggesting a shift toward an immunosuppressive microenvironment. RNA sequencing analysis of isolated islets and polarized proinflammatory macrophages revealed significant alteration of cytokine-responsive pathways and a reduction in IFN response after VLX-1005 treatment. Our studies demonstrated that the ALOX12 human replacement gene mouse provides a platform for the preclinical evaluation of LOX inhibitors and supports VLX-1005 as an inhibitor of human 12-LOX that engages the enzymatic target and alters the inflammatory phenotypes of islets and macrophages to promote the delay of autoimmune diabetes.

Graphical Abstract Introduction

The pathogenesis of type 1 diabetes (T1D) involves a complex interplay between multiple cell types within the pancreatic islet, including innate immune cells (macrophages, dendritic cells), insulin-producing cells (β cells), and adaptive immune cells (T cells, B cells) (1). Although the disease has traditionally been viewed as arising from a primary defect in immune tolerance, an emerging perspective posits that environmental factors (such as viruses or other systemic inflammatory disorders) may aggravate an interaction between macrophages and β cells, facilitating oxidative and endoplasmic reticulum (ER) stress pathways in β cells (2–4). These pathways facilitate the generation of β cell neoepitopes that then trigger adaptive autoimmunity (5, 6). Disease-modifying therapies — those that alter disease pathogenesis rather than correcting the underlying disease phenotypes — have largely focused on the adaptive immune system and seen some successes in clinical trials. For example, an anti-CD3 monoclonal antibody (teplizumab) that targets activated T cells has been shown to delay the onset of T1D by up to 2 years in individuals at high risk for the disease (7). Given the increasing appreciation of innate immune cells and β cells in early T1D pathogenesis, the identification of drugs targeting these cell types raises the possibility that combination therapeutic approaches may provide more durable outcomes.

The lipoxygenases (LOXs) encompass a family of enzymes involved in lipid metabolism that facilitates the oxygenation of polyunsaturated fatty acids to form eicosanoids, some of which are proinflammatory in nature (8). In the mouse, 12/15-LOX is encoded by the Alox15 gene and is the primary active LOX present in macrophages and β cells and produces the proinflammatory eicosanoid 12-hydroxyeicosatetraenoic acid (12-HETE) as a principal product from the substrate arachidonic acid (9). Whole-body deletion of Alox15 on the autoimmune non-obese diabetic (NOD) mouse background results in almost complete protection against diabetes (10). Deletion of Alox15 in either the innate immune myeloid cells (2) or in β cells (11) recapitulates the autoimmune diabetes protection seen in the whole-body deletion, emphasizing both the early role of these cell types in T1D and the importance of the 12/15-LOX pathway in disease pathogenesis. In these cell-specific deletion models, islets exhibit marked reductions in invading pathogenic T cells (insulitis), a finding reflecting the disease-modifying response. The molecular events tied to disease protection ostensibly emanate from reductions in oxidative and ER stress (and the resultant reduction in neoepitope formation and presentation) as well as from enhanced display of PD-L1 (an immunosuppressive checkpoint ligand) on the surface of myeloid cells and β cells (2, 11).

In humans, the relevant LOX enzyme that produces 12-HETE is 12-LOX, encoded by the ALOX12 gene. Like the mouse 12/15-LOX, human 12-LOX is present in residual insulin-positive cells in donors with T1D or in autoantibody-positive donors at risk for T1D (12) — a finding consistent with a potential role in promoting β cell sensitivity to autoimmunity. A major challenge to using mice as a platform to test inhibitors is that human 12-LOX exhibits structurally distinct characteristics from mouse 12/15-LOX, thereby necessitating the development of different inhibitors that cannot be tested for efficacy in mice (13–15). Previously, VLX-1005 (also known as ML355) was described as a potent and selective inhibitor of human 12-LOX, while also displaying a favorable half-maximal inhibitory concentration (IC50) and pharmacokinetic properties (16). VLX-1005 was shown to protect human islets in vitro against dysfunction caused by proinflammatory cytokines (17), but the lack of appropriate in vivo model systems has made it challenging to pharmacologically validate VLX-1005 as a therapeutic target in autoimmune diabetes. To address this challenge, we developed new mouse strains in which the mouse Alox15 gene is replaced by the human ALOX12 gene while retaining the mouse gene’s upstream control elements. This human gene replacement platform was leveraged to test whether and how pharmacologic inhibition of human 12-LOX with VLX-1005 modifies disease progression in autoimmune T1D.

Results

Generation and validation of the hALOX12 gene replacement mouse model. To establish a platform to test potential inhibitors of human 12-LOX in vivo, we generated a mouse model in which the endogenous mouse Alox15 gene is replaced by the human ALOX12 gene (Figure 1A). This model leaves the mouse upstream regulatory region intact to ensure that the expression of ALOX12 recapitulates the expression of Alox15. These mice (henceforth referred to as hALOX12 mice) were introgressed onto the C57BL/6J (also known as B6) mouse background using a speed congenics approach and bred to homozygosity. Microsatellite genotyping showed that the mice were 100% congenic on the C57BL/6J background (Supplemental Table 1; supplemental material available online with this article; https://doi.org/10.1172/jci.insight.185299DS1). To confirm the successful deletion of mouse Alox15 and replacement with human ALOX12, we performed standard genotyping (Supplemental Figure 1A). Additionally, we isolated tissues (kidney, spleen, lung, fat, liver, islets, peritoneal macrophages, and bone marrow–derived macrophages [BMDMs]) from WT C57BL/6J and B6.hALOX12 mice and subjected them to gene expression analysis for Alox15 and ALOX12. As anticipated, WT tissues expressed mouse Alox15 and did not express human ALOX12; conversely, B6.hALOX12 mouse tissues expressed ALOX12 but not Alox15 (Table 1).

Figure 1

12-LOX inhibition protects against streptozotocin-induced diabetes. C57BL/6J and B6.hALOX12 male mice (n = 4–7 per group as indicated) were treated with 30 mg/kg i.p. or p.o. VLX-1005 and multiple low-dose streptozotocin (STZ). (A) Schematic of the generation of hALOX12 mice by replacing mouse Alox15 with human ALOX12. (B) Chemical structure of VLX-1005. (C) Random-fed blood glucose values in vehicle-treated male C57BL/6J and B6.hALOX12 mice after STZ. (D) GTT of vehicle-treated male C57BL/6J and B6.hALOX12 mice on day 4 after STZ treatment. (E) Random-fed blood glucose values in VLX-1005–treated male C57BL/6J and B6.hALOX12 mice after STZ. (F) GTT of VLX-1005–treated male C57BL/6J and B6.hALOX12 mice on day 4 after STZ treatment. (G) AUC of C57BL/6J and B6.hALOX12 on day 4 after STZ-treatment (1-way ANOVA). (H) Random-fed blood glucose values in male vehicle- or VLX-1005–treated (p.o.) B6.hALOX12 mice after STZ. (I) GTT of male vehicle- or VLX-1005–treated (p.o.) B6.hALOX12 mice on day 4 after STZ treatment. (J) AUC of B6.hALOX12 on day 4 after STZ treatment. (K) Pancreata stained for insulin (left panel) and β cell mass measurement (right panel) from male B6.hALOX12 mice on day 26 after STZ treatment. Scale bars: 500 μm. Data are presented as mean ± SEM and statistical significance was determined by a 2-tailed Student’s t test (all except G) or 1-way ANOVA with Tukey’s post hoc test (G).

Table 1

RNA expression levels of human ALOX12 and mouse Alox15 normalized to mouse Actb from various isolated tissues of C57BL/6J and B6.hALOX12 mice

Because lipoxygenases are known to affect metabolic function, we next performed metabolic characterization to determine whether/how the replacement of Alox15 with ALOX12 altered metabolic phenotypes. We found no significant differences in body weight, lean mass, fat mass, random-fed blood glucose levels, or glucose tolerance between WT C57BL/6J and B6.hALOX12 mice (Supplemental Figure 1, B–F). Moreover, islet ultrastructure (relative immunostaining patterns of α cells and β cells) and composition (α cell mass and β cell mass) were indistinguishable between 10-week-old WT and B6.hALOX12 mice (Supplemental Figure 1, G–I). Taken together, these data suggest that the successful replacement of Alox15 with human ALOX12 did not alter gross metabolic or islet phenotypes.

Effects of VLX-1005 against STZ-induced diabetes are specific to B6.hALOX12 mice. Prior studies demonstrated that whole-body deletion of mouse Alox15 protects against diabetes induced by the chemical streptozotocin (STZ) (18). To test whether the human 12-LOX inhibitor VLX-1005 (14) (Figure 1B) phenocopies deletion of the enzyme in our human gene replacement mice, we employed a similar STZ diabetes induction protocol. STZ is a β cell toxin that induces low-grade inflammation, macrophage influx into islets, and eventual diabetes in mice after 5 daily low-dose i.p. injections (55 mg/kg) (19). Eight-week-old male WT C57BL/6J and B6.hALOX12 mice were injected i.p. daily with vehicle or 30 mg/kg VLX-1005 in the peri-STZ treatment period (for the 5 days before, during, and after STZ). STZ-injected C57BL/6J and B6.hALOX12 mice receiving vehicle became overtly hyperglycemic within 10 days of starting STZ treatment and displayed equivalent glucose intolerance by glucose tolerance test (GTT) (Figure 1, C and D). Upon receiving VLX-1005, however, B6.hALOX12 mice showed complete protection from STZ-induced diabetes, whereas WT C57BL/6J mice became overtly hyperglycemic (Figure 1E); GTTs at the end of the study confirmed improved glucose tolerance in B6.hALOX12 mice compared with WT C57BL/6J mice (Figure 1, F and G). These data indicate a specific effect of the drug in preventing hyperglycemia in B6.hALOX12 mice and support the effectiveness of the hALOX12 platform for interrogating VLX-1005 action.

Pharmacokinetics of oral VLX-1005 and its effects on STZ-induced diabetes. Given that the oral route is the preferred route for systemic drug delivery in humans, we next asked whether oral administration (p.o.) of VLX-1005 provides adequate exposure in mice. We performed pharmacokinetic analysis following a single p.o. administration (as a suspension in 0.5% methylcellulose) of VLX-1005 spray-dried dispersion at a dose of 30 mg/kg in C57BL/6J mice, followed by serial analysis of VLX-1005 levels by liquid chromatography–tandem mass spectrometry (LC-MS/MS). The pharmacokinetic profile of p.o. administered VLX-1005 in mice showed a mean half-life (t1/2) of 3.24 ± 0.07 hours and a consistent tmax of 0.250 hours across all mice. The Cmax was 13,300 ± 624 ng/mL, with moderate variability in AUC (15,029 ± 3177 h•ng/mL). These parameters, particularly the low variability in tmax and Cmax, support the feasibility of once-daily dosing for maintaining therapeutic levels over a 24-hour period (Table 2). We next tested the effects of p.o. administration of VLX-1005 on the low-dose STZ model, with VLX-1005 (at 30 mg/kg) given 3 days prior to the start of STZ, during STZ, and for 3 days following STZ treatment. Similar to i.p. delivery, p.o. administration of VLX-1005 in B6.hALOX12 mice resulted in lower random-fed blood glucose levels (Figure 1H) and significantly improved glucose tolerance (Figure 1, I and J) compared with vehicle — although this effect was not as robust as with i.p. delivery of the drug. Consistent with improved glucose homeostasis, p.o. VLX-1005–treated mice exhibited greater β cell mass at the end of the study compared with vehicle-treated mice (Figure 1K). Collectively, these data suggest that a single daily p.o. delivery of VLX-1005 (at 30 mg/kg) achieves plasma levels with therapeutic efficacy.

Table 2

Plasma concentration vs. time profile for VLX-1005 after 30 mg/kg p.o. in C57BL/6J and NOD.ShiLt/J mice.

VLX-1005 treatment reduces β cell inflammation in NOD.hALOX12 mice. The NOD mouse model is a model of T1D that recapitulates many of the immune and β cell features of the disease (20). We, therefore, asked whether pharmacologic inhibition of 12-LOX using p.o. administered VLX-1005 protects against spontaneous diabetes development in the NOD mouse model. To address this question, we introgressed humanized hALOX12 mice onto the NOD background using a speed congenics approach. Genome scanning of microsatellites was performed to confirm that mice were 100% congenic on the NOD mouse background (NOD.hALOX12 mice) (Supplemental Table 1). We next measured human 12-LOX protein levels in the NOD.hALOX12 mice. Similar to the gene profile we observed in B6.hALOX12 mice compared with WT C57BL/6J mice, WT NOD tissues robustly expressed mouse 12/15-LOX (the protein encoded by mouse Alox15) and little or no human ALOX12; conversely, NOD.hALOX12 mice tissues robustly expressed human 12-LOX and minimal levels of 12/15-LOX (Supplemental Figure 1, J and K). Consistent with their congenic nature, female NOD.hALOX12 mice exhibited islet pathology similar to that of NOD mice at the (prediabetic) age of 10 weeks, with evidence of T and B cell infiltration of islets (Supplemental Figure 1L) and indistinguishable insulitis score (Supplemental Figure 1M), suggesting that replacement of Alox15 with human ALOX12 did not alter the islet pathology of the disease. Pharmacokinetics of p.o. administered VLX-1005 (30 mg/kg) on the NOD background were similar to those seen in C57BL/6J mice (Table 2), suggesting that the NOD background does not affect drug absorption or clearance.

To assess the effect of VLX-1005 administration on products of 12-LOX activity in NOD.hALOX12 mice, we administered 30 mg/kg VLX-1005 (or vehicle) p.o. to female NOD.hALOX12 mice for 1 week during the prediabetic phase (8 weeks of age) and harvested serum. Lipidomics analysis (by LC-MS/MS) was performed for a series of 12-LOX products resulting from different fatty acid substrates (Figure 2A). Notably, levels of 12-HETE (from arachidonic acid), 13-HODE (from linoleic acid), and 14-HDHA and 17-HDHA (from docosahexaenoic acid) were all significantly reduced (Figure 2B). Levels of 12-HEPE (from eicosapentaenoic acid) were not significantly changed (Figure 2B), suggesting minimal involvement of eicosapentaenoic acid metabolism in NOD mice. Lipids within the pathway that are processed by enzymes other than 12-LOX were not statistically significantly altered (Supplemental Table 2). These data are collectively consistent with the expected 12-LOX engagement by VLX-1005.

Figure 2

VLX-1005 decreased islet inflammation in NOD.hALOX12 female mice. Six-week-old female prediabetic NOD.hALOX12 mice were treated p.o. with 30 mg/kg VLX-1005 for 4 weeks prior to tissue analysis. (A) Schematic representation of 12-lipoxygenase products. (B) Serum lipidomics results of 12-lipoxygenase products as indicated (n = 4–5). (C) Schematic representation of mouse treatment paradigm. (D) Pancreata from mice stained for CD3 (magenta), B220 (teal), insulin (white), and nuclei (blue). Scale bars: 50 μm. (E) Average insulitis score; each dot represents an individual mouse (n = 4–5). (F) Heatmap of identified proteins in the insulitic area (left) and insulin-positive area (right). (G) Pancreata of mice stained and quantified for CD3 (brown, top panels: arrows indicate positive CD3 staining within the islet), F4/80 (brown, middle panels: arrows indicate positive F4/80 staining within the islet), or MAC2 (brown, bottom panels: arrows indicate positive MAC2 staining within the islet) and nuclei (blue). Each dot represents an individual mouse (n = 4–5). Scale bars: 50 μm. Data are presented as mean ± SEM and statistical significance was determined by a 2-tailed Student’s t test in all cases.

To assess the effect of VLX-1005 administration on immune cell phenotypes in NOD.hALOX12 mice, we administered 30 mg/kg VLX-1005 (or vehicle) p.o. to female NOD.hALOX12 mice for 4 weeks during the prediabetic phase (6–10 weeks of age) and harvested pancreas, pancreatic lymph nodes (pLNs), and spleen (Figure 2C). Pancreas pathology showed reduced T and B cell infiltration and that the extent of insulitis (by insulitis scoring) was significantly reduced in VLX-1005–treated NOD.hALOX12 mice compared with vehicle-treated mice (Figure 2, D and E). To specifically interrogate the nature of immune cells within the insulitic region, we performed spatial tissue-based proteomics (NanoString). We used insulin immunostaining and nuclei staining to identify β cells and the surrounding insulitic regions. Prevalidated antibodies in the GeoMx mouse immune panel were used to probe for immune cell subtypes in the peri-islet insulitic region and within the islet. NOD.hALOX12 mice exhibited a notable reduction in myeloid population subtypes in both insulitic and islet areas, including macrophages (F4/80+CD11b+) and dendritic cells (CD11c+) (Figure 2F and Supplemental Figure 2, A and B). This reduction in myeloid cell populations was accompanied by a decrease in T and B cells populations, including CD4+, CD3+, CD8+, and CD19+ cells (Figure 2F and Supplemental Figure 2, A and B). Immunohistochemistry of pancreas sections confirmed the reductions in both T cells (CD3+), macrophages (F4/80+), and activated macrophages (Mac2+) following p.o. VLX-1005 treatment (Figure 2G). A notable observation in spatial proteomics was the increased levels of the immune checkpoint ligand PD-L1 (Figure 2F). Enhanced PD-L1/PD-1 interactions shift T cells to less aggressive, more regulatory phenotypes (21). To interrogate this possibility, we performed immune profiling by flow cytometry of pLNs from mice treated with p.o. VLX-1005 or vehicle. pLNs are key sites in the initial priming of autoreactive T cells in NOD mice (22). Treatment with p.o. VLX-1005 led to an increase in CD4+Foxp3+ regulatory T cells (Tregs) in the pLNs (Supplemental Figure 2C). This effect on Tregs was specific for the pLNs since no changes in Tregs were observed in the spleen after VLX-1005 treatment (Supplemental Figure 2D).

Orally administered VLX-1005 reduces autoimmune diabetes incidence in female and male NOD.hALOX12 mice. Because 4 weeks of p.o. VLX-1005 dosing led to improvements in insulitis and reductions in infiltrating T and B cells, we next asked whether these changes lead to prevention or delay of subsequent diabetes development in NOD.hALOX12 mice. Both female and male mice were administered VLX-1005 via daily oral gavage (30 mg/kg) or vehicle for 4 weeks during the prediabetic phase (6–10 weeks of age). Mice were followed for diabetes development (blood glucose ≥250 mg/dL on 2 consecutive days) until 25 weeks of age (Figure 3A). At 25 weeks of age, 60% of female mice and 75% of male mice receiving VLX-1005 were protected from diabetes development compared with 25% of female and 50% of male mice receiving vehicle (Figure 3, B and C). Whereas the preceding studies demonstrate that 12-LOX inhibition with p.o. VLX-1005 delays the development of diabetes, they do not address whether administration of the drug might reverse established diabetes or mitigate hyperglycemia. We allowed female NOD.hALOX12 mice to develop diabetes (defined as 2 consecutive random-fed blood glucose measurements ≥250 mg/dL), and then administered VLX-1005 or vehicle for up to 6 weeks via daily oral gavage or until the mice exhibited signs of physical deterioration from hyperglycemia (loss in body weight, dishevelment) (Figure 3D). Notably, we did not observe a reversal in diabetes, but did see relative reductions in blood glucose levels in mice treated with VLX-1005 compared with vehicle (Figure 3, E and F).

Figure 3

VLX-1005 treatment delays autoimmune diabetes onset in female and male NOD.hALOX12 mice. NOD.hALOX12 mice (n = 20 per group) were treated during the prediabetic stage from 6 to 10 weeks of age or at the time of diabetes development (n = 11–12 per group). (A) Schematic representation of diabetes prevention experimental design. (B) Diabetes incidence in female NOD.hALOX12 mice. (C) Diabetes incidence in male NOD.hALOX12 mice. (D) Schematic representation of diabetes reversal experimental design. (E) Random-fed blood glucose levels in each female mouse. (F) Average random-fed blood glucose levels of female mice. Data are presented as mean ± SEM and statistical significance was determined by a Mantel-Cox log-rank test.

Orally administered VLX-1005 reduces islet death and oxidative stress in NOD.hALOX12 mice. 12-LOX is primarily present in islets and macrophages, and deletion of the mouse gene (Alox15) in either tissue separately was previously shown to reduce diabetes incidence. We, therefore, first asked how treatment with VLX-1005 affects islet cell phenotypes. We first analyzed isolated islets from female NOD.hALOX12 mice treated with vehicle or VLX-1005 by RNA sequencing to identify how islet gene expression might be altered. Principal component analysis of transcriptomics revealed that islets from vehicle- and VLX-1005–treated NOD.hALOX12 mice clustered separately, suggesting an effect of VLX-1005 treatment on gene expression (Figure 4A). Pairwise comparison of gene expression using a false discovery rate (FDR) of less than 0.05 and fold-change (FC) of greater than 2 yielded only 189 differentially expressed genes. Instead, a P-value cutoff of 0.05 and FC greater than 2 revealed changes in 709 genes between vehicle- and VLX-1005–treated NOD.hALOX12 mice (Figure 4B, volcano plot, and Supplemental Table 3 for full sequencing results). Gene Ontology (GO) pathway analysis showed significantly altered pathways related to DNA replication (e.g., Anp32b, Skp1a, Itfg2, Dmrt1i), inflammation (NF-κB activity; e.g., Elf1, Trim75, Rnase1, Lmo1, Bcl3, Ptgis, Commd1, Lrrc14, Foxp3), and G protein–coupled receptor signaling (e.g., Gpr89, Glp2r), among others (Figure 4C). These pathways suggest responses that may be related to changes in cellular survival in response to VLX-1005. We, therefore, immunostained pancreatic sections for markers of cell death and proliferation in the islet. VLX-1005–treated NOD.hALOX12 mice exhibited decreased islet cell death, as measured by reduced terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) and H2A histone family member X (H2A.X) staining compared with vehicle-treated mice (Figure 4D). Additionally, VLX-1005–treated mice demonstrated decreased β cell proliferation, as measured by proliferating cell nuclear antigen (PCNA) immunostaining (Figure 4D); reduced PCNA immunostaining was also consistent with the reduction in Ki-67 observed in spatial proteomics of the insulin+ area (Figure 2F and Supplemental Figure 2B). We interpret the reduction in β cell proliferation as a consequence of reduced β cell apoptosis. The change in NF-κB signaling led us to investigate whether markers of oxidative stress were affected since inflammation, oxidative stress, and β cell survival are closely linked (23, 24). We performed immunostaining for the oxidative stress marker, 4-hydroxynonenal (4-HNE), and observed reduced immunostaining in mice treated with VLX-1005 compared with placebo (Figure 4E). Consistent with this observation, β cells from VLX-1005–treated animals also displayed an increase in levels of the antioxidant enzyme GPx1 (Figure 4E). Collectively, these data are consistent with prior observed effects of reduced inflammation, oxidative stress, and β cell death in β cell–specific deletion of mouse Alox15 in NOD mice (11).

Figure 4

VLX-1005 decreased β cell death, proliferation, and oxidative stress in female NOD.hALOX12 mice. Pancreata or islets were harvested from 10-week-old prediabetic female NOD.hALOX12 mice after 4 weeks of treatment with vehicle or VLX-1005 (n = 3–4 per group). (A) Principal component analysis plot of RNA sequencing results from isolated islets of vehicle- or VLX-1005–treated mice. (B) Volcano plot of differentially expressed genes. (C) Gene ontology pathway analysis of differentially expressed genes. (D) Pancreata from mice stained and quantified for TUNEL (brown, left panels: black arrow indicates positive TUNEL staining within the islet), H2A.X (brown, middle panels: black arrowheads indicate positive H2A.X staining within the islet), or PCNA (magenta, right panels: white arrowheads indicate positive PCNA staining within the islet), insulin (green) and nuclei (blue). Each dot represents an individual mouse (n = 4–5). Scale bars: 50 μm. (E) Pancreata from mice stained and quantified for 4-HNE (magenta, left panels), or GPx1 (magenta, right panels), and insulin (green) and nuclei (blue). Each dot represents an individual mouse (n = 4). Scale bars: 50 μm. Data are presented as mean ± SEM and statistical significance was determined by a 2-tailed Student’s t test.

VLX-1005 alters the proinflammatory macrophage phenotype. Whereas the preceding findings are consistent with improved β cell survival following p.o. VLX-1005 administration, these studies do not rule out the possibility that the drug directly modifies the phenotype of infiltrating macrophages, which could secondarily affect β cells. Because bulk islet transcriptomics analysis does not resolve gene expression events associated with specific cell types, we isolated BMDMs from female NOD.hALOX12 mice and then performed RNA sequencing in the presence or absence of VLX-1005. BMDMs were unpolarized (M0) or polarized to an M1-like state (with lipopolysaccharide and IFN-γ) to mimic the inflammatory state that might be observed during T1D pathogenesis. During polarization, BMDMs were treated with vehicle or VLX-1005 (Figure 5A). Principal component analysis of transcriptomics revealed that M0 macrophages treated with VLX-1005 co-clustered with vehicle-treated M0 macrophages, suggesting a minimal transcriptional effect of the drug on unpolarized cells (Figure 5B). Consistent with this interpretation, only 1% of genes (159 out of 15,888) were significantly altered with VLX-1005 treatment (when using criteria FC > 2 and P < 0.05) (Supplemental Table 4 for full sequencing results). Upon polarization to the M1-like state, a clear rightward shift in the principal component analysis plot was observed with both vehicle- and VLX-1005–treated BMDMs and a notable separation was seen between vehicle and VLX-1005 treatment (Figure 5B); this finding suggests that the impact of 12-LOX inhibition is more prominent upon a shift to a proinflammatory state of macrophages.