Using animal models to decipher the role of polyamines in the pancreas and β Cell

Abstract

Pancreatic islet β cells are insulin-secreting cells that are responsible for sensing blood glucose levels and maintaining normoglycemia. Polyamines are vital to supporting the transcriptional and translational demands placed on secretory cells such as the β cell; however, recent evidence suggests the polyamine pathway may also be detrimental to normal β cell function. β cell dysfunction instigated by inflammation is common to both type 1 and type 2 diabetes, and notably, interventions that target the polyamine pathway may offer therapeutic benefits. The objective of this review is to synthesize the roles of polyamine metabolism in the pancreas and to frame this pathway as a link between nutrient status, immune activation, and β cell stress in diabetes. This review summarizes key findings from animal models used to study the polyamine pathway in pancreatic islet growth, development, and function. It also explores the prospects of polyamine inhibition to modify diabetes pathogenesis and improve β cell health.

Introduction

The polyamines (putrescine, spermidine, and spermine) are essential for the proper development, maintenance, and function of diverse cell types (Schibalski et al. 2024). High concentrations of polyamines have been reported in both exocrine and endocrine pancreatic cells, with the insulin-producing β cell showing the highest concentration among the islet cells (Hougaard and Larsson 1986; Hougaard et al. 1986). The pancreas is a vital organ that serves a dual purpose in digestion and hormone production (Karpińska and Czauderna 2022). Digestive functions of the exocrine pancreas revolve around producing and releasing digestive enzymes into the small intestine, breakdown of proteins and fats by proteases and lipases, respectively, and production of amylase, which breaks down starch and carbohydrates into sugar. Hormone functions of the endocrine pancreas are essential for the regulation of blood glucose by insulin by β cells and glucagon by α cells. Diabetes is characterized by the dysregulation of blood glucose levels, driven by the insufficient production of insulin by β cells (Prentki and Nolan 2006; Sims et al. 2020). Interventions that target the polyamine pathway hold promise for preserving β cell function in diabetes (Tersey et al. 2014; Sims et al. 2023), but it is important to understand the dual roles of polyamines that make these molecules beneficial or detrimental to metabolic health. In this review, we discuss the role of polyamine metabolism in pancreas development, β cell function, immune cell regulation, and the pathogenesis of both type 1 (T1D) and type 2 (T2D) diabetes development in the context of animal models. To do this, we performed primary literature searches using pubmed and google scholar with the search terms polyamines and/or hypusine along with pancreas development, islet, β cell, type 1 diabetes, or type 2 diabetes, along with additional searches as needed to flush out the review topics.

Polyamines and hypusine

Polyamines are small, positively charged molecules (aliphatic polycations) that are ubiquitous in living cells and indispensable for normal growth and survival. The principal human polyamines are putrescine, spermidine, and spermine. Their multiple amine groups enable electrostatic interactions with negatively charged biomolecules such as DNA, RNA, and phospholipids, stabilizing chromatin structure, modulating transcription, and influencing protein synthesis (van Dam et al. 2002). Because of these interactions, polyamines are central regulators of cell proliferation, differentiation, and stress responses. Their concentrations are tightly controlled within cells, reflecting the balance between biosynthesis, catabolism, transport, and excretion (Sagar et al. 2021; Holbert et al. 2024).

Polyamines

Polyamines originate from both exogenous (dietary uptake and gut microbiota) and endogenous (intracellular de novo synthesis and interconversion of other biomolecules) sources (Kulkarni et al. 2022a). The biosynthetic pathway of these polyamines starts with the production of ornithine from arginine by the enzyme arginase (ARG) (Fig. 1A). Arginine is derived from dietary protein intake, body protein breakdown, or endogenous de novo arginine production through the urea cycle (Luiking et al. 2012). Ornithine is decarboxylated by ornithine decarboxylase (ODC) to form putrescine in a rate-limiting reaction. In addition, ornithine is found in many foods, including fish, dairy products, and legumes (Ginguay et al. 2017). Ornithine is converted to spermidine by spermidine synthase (SPDS), which is further converted to spermine by spermine synthase (SPMS). For SPDS and SPMS to catalyze their respective reactions, they utilize aminopropyl groups derived from decarboxylated S-adenosylmethionine (dc-SAM). Methionine is converted to S-adenosylmethionine (SAM) and subsequently further converted to dc-SAM by S-adenosylmethionine decarboxylase (SAMDC) in a rate-limiting step (Kozbial and Mushegian 2005). Catabolic enzymes, including spermidine/spermine N¹-acetyltransferase (SAT) and polyamine oxidases (PAOX, SMOX), catalyze turnover by converting polyamines back to putrescine and simultaneously generate byproducts such as hydrogen peroxide and/or acrolein (Schibalski et al. 2024). This metabolic network is highly dynamic and sensitive to cellular energy status, growth signals, and environmental stress, ensuring that polyamine levels remain within narrow physiological limits.

Fig. 1.

A schematic of the polyamine/hypusine pathways. A The mammalian polyamine metabolism pathway including the cytoplasmic urea cycle (green box). Polyamines are identified by orange color. B The bacterial polyamine metabolism pathway. Bacterial specific proteins are identified by magenta box. C The hypusine pathway wherein hypusine is added to eIF5A. Images created in BioRender. ASL arginosuccinase; ARG arginase; ADC arginine decarboxylase; AUH arginine ureahydrolase; OTC ornithine transcarbamoylase; ASS arginosuccinate synthetase; ODC ornithine decarboxylase; SPDS spermidine synthase; SPMS spermine synthase; SAM S-adenosylmethionine; SAMDC S-adenosylmethionine decarboxylase; dc-SAM decarboxylated S-adenosylmethionine; SSAT spermine/spermidine N1-acetyltransferase; PAOX polyamine oxidase; 3-AP: 3-aminopropanol; 3-AAP: 3-acetoamidopropanal; NAD nicotinamide adenine dinucleotide; DHPS deoxyhypusine synthase; DOHH: deoxyhypusine hydroxylase; eIF5A^Lys unhypusinated eIF5A; eIF5A^dHyp: deoxyhypusinated eIF5A; eIF5A^Hyp hypusinated eIF5A

In addition to polyamine production via ODC, alternative pathways in a wide range of bacteria generate the three major polyamines and other microbial-derived polyamines, such as cadaverine (Fig. 1B). Arginine decarboxylase (ADC) converts arginine into agmatine and is a prevalent enzyme among major phyla of the gut microbiome (Burrell et al. 2010). Agmatine ureohydrolase (AUH) catabolizes agmatine into putrescine and urea. Interestingly, many commensal bacteria do not possess all the enzymes necessary to produce polyamines and rely on collaborative processes within the microbial community for polyamine synthesis (Kitada et al. 2018; Nakamura et al. 2019). Cadaverine is the only microbial polyamine derived from lysine which is accomplished through the activity of lysine decarboxylase (LDC). The known functions of microbial polyamines include their covalent linkage to peptidoglycan, regulation of biofilm formation, and incorporation into siderophores (Michael 2018). However, as gut bacteria are a major source of polyamines in the colon, these polyamines are also capable of entering the circulation and being used by host cells (Nakamura et al. 2021). Cadaverine has been shown to alter the polarization status of gut macrophages depending on the concentration and inflammatory conditions (Formiga et al. 2025). Recognizing the microbial contributions to the polyamine pool will be important towards understanding how polyamines influence downstream host processes.

Hypusine

Among the many cellular processes linked to polyamine metabolism, one of the most striking is the post-translational modification known as hypusination. Hypusine is an unusual amino acid found exclusively in the translation factor eIF5A, which in mammals, occurs as 2 isoforms, eIF5A1 and eIF5A2 (Smit-McBride et al. 1989; Pällmann et al. 2015). Hypusine is formed in a two-step reaction that exemplifies the integration between polyamine metabolism and protein synthesis (Fig. 1C) (Park and Wolff 2018). In the first step, deoxyhypusine synthase (DHPS) transfers an aminobutyl group from spermidine to a specific lysine residue on eIF5A (eIF5A^Lys), producing deoxyhypusinated eIF5A (eIF5A^dHyp). This intermediate is then hydroxylated by deoxyhypusine hydroxylase (DOHH) to generate the mature hypusinated eIF5A (eIF5A^Hyp) (Park and Wolff 2018). eIF5A^Hyp is essential for the translation of proteins containing proline-rich motifs and for sustaining normal rates of cell growth and proliferation (Gutierrez et al. 2013).

Because spermidine is the sole donor of the aminobutyl group required for hypusine formation, fluctuations in polyamine metabolism directly affect the efficiency of this modification (Park and Wolff 2018). This has been exemplified in a strain of S. cerevisiae with low spermidine content due to a deficiency in SAMDC. Nearly 50% of the remaining spermidine content was mobilized towards hypusine formation, whereas in a wildtype strain, only about 1.4% was directed towards hypusine (Chattopadhyay et al. 2008). When cellular polyamine levels fall, the production of eIF5A^Hyp decreases, disrupting protein synthesis and cell growth in metabolically active tissues such as the pancreas and spleen (Mastracci et al. 2015, 2020). This connection illustrates how polyamine availability couples the metabolic state of a cell to its capacity for protein translation, positioning hypusination as a key regulatory node between metabolism, growth, and disease progression (Sagar et al. 2021).

Genetic and biochemical tools for studying polyamines in animal models in vivo

In mice, whole-body knockout of Odc1, Eif5a1, or Dhps results in lethality between embryonic days E3.5 and E6.5 (Pendeville et al. 2001; Templin et al. 2011; Nishimura et al. 2012). To overcome this challenge, investigators have utilized various conditional and cell-specific knockout mouse models to extend the studies primarily through the use of Cre-LoxP mouse models that allow for cell specificity and/or tamoxifen-driven Cre, which allows for inducible gene deletion/activation after developmental time points (Kim et al. 2018). Besides mouse models, lower organisms such as drosophila and zebrafish have been extensively used in the field. In zebrafish, genes can be knocked down during early stages using morpholinos. Morpholinos act by binding to complementary mRNA sequences blocking the interactions of proteins with the mRNA and therefore preventing initiation complex read-through or modification of splicing of the mRNA (Moulton 2017). Alternatively, siRNA can also be used in both zebrafish and mouse models to knock down protein levels in the pancreas (Maier et al. 2010).

Additionally, there are numerous inhibitors of the enzymes in the polyamine metabolism pathway. The three rate-limiting steps in the polyamine/hypusine synthesis pathway have specific inhibitors that can be utilized in both in vivo and in vitro settings. ODC can be inhibited by alpha-difluromethylornithine (DFMO) (Bitonti et al. 1985), DHPS by N1-guanyl-1,7-diaminoheptane (Gc7) (Jakus et al. 1993), and SAMDC by ethylglyoxal bis(guanylhydrazone) (EGBG) (Shibata et al. 1998). SAT1 is the rate-limiting enzyme in polyamine catabolism and can be inhibited in vitro by diminazene aceturate (DA) (Neidhart et al. 2014). Non-rate limiting enzymes can also be inhibited by altering different polyamine levels. Alternatively, polyamines can be directly added to cell cultures, given by injection, or modulated/added through diet (Soda et al. 2009). The spermine analog N1,N11-diethylnorspermine (DENSpm) reduces polyamine production via inhibition of ODC and simultaneously accelerates polyamine degradation via activation of SAT1 (Porter and Bergeron 1988). Altogether, the breadth of genetic and chemical approaches has helped to overcome the challenges associated with studying polyamines in vivo and have been useful tools for understanding the roles of polyamines in pancreas and β cell biology.

Polyamines in the pancreas, islet, and β cell

Polyamine contribution to pancreas and islet development

The pancreas is made up of exocrine (acinar and duct) and endocrine (islet) components. In mammals, pancreas development begins with the specification of two separate pancreatic buds that subsequently merge to form a single organ. During fetal development, the mammalian pancreas forms from two independent bud protrusions arising from the dorsal and ventral regions of the foregut endoderm, from which both the endocrine and exocrine cells arise. The two pancreatic buds fuse to form the definitive pancreas and integrate with the common bile duct and duodenum (Slack 1995; Servitja and Ferrer 2004; Jeon et al. 2009). SOX9, PDX1, and GATA4 facilitate parenchymal growth of the pancreatic buds. Expression of NEUROG3 triggers pancreatic progenitor cells to undergo endocrine differentiation (Ehrhardt and Gomez 2025).

As polyamine biosynthesis is known to be critical for cellular proliferation and essential for the earliest stages of embryonic development and polyamine levels are high in the pancreas (Hougaard and Larsson 1986; Hougaard et al. 1986), studies utilizing zebrafish and mouse models have delineated a role in early pancreas development (see Table 1 and 2). Using zebrafish as a model organism for pancreas development, biochemical inhibition of ODC with DFMO resulted in a 46% shorter pancreas at 72 h post fertilization in DFMO-treated embryos (Mastracci et al. 2015). The phenotype was due to a blunting of the differentiation and proliferation of exocrine acinar cells and not a reduction in the progenitor cell population. Additionally, it was observed that the total endocrine and the α cell populations were unchanged; however, DFMO-treated zebrafish had a reduced number of β cells. Spermidine rescued the phenotype observed with DFMO treatment but did not affect pancreas growth on its own (Mastracci et al. 2015). Inhibition of DHPS, the rate-limiting enzyme for the hypusination of eIF5A, with Gc7 or dhps knockdown using morpholinos, led to a shortened pancreas and a reduced number of β cells, phenocopying the results seen with ODC inhibition (Mastracci et al. 2015; Anderson et al. 2025). However, knockdown of eif5a1/2 using morpholinos did not reduce pancreas size. In zebrafish, the gene most similar to mammalian Eif5a1 was duplicated to form eif5a1 and eif5a2. Remarkably, the combined effect of dhps and eif5a1/2 morpholinos rescued pancreas length (Anderson et al. 2025), which seems to indicate an independent and underappreciated role for unhypusinated eIF5A^Lys in controlling pancreas growth.

Table 1.

Zebrafish Models

Genetic/Inhibitor Manipulation	Disease model	Major findings	Reference
ODC Inhibition—DFMO	pancreas development	shortened pancreas, reduction in β cell number	(Mastracci et al. 2015)
DHPS inhibition—Gc7	pancreas development	shortened pancreas, reduction in β cell number	(Mastracci et al. 2015)
dhps morpholino	pancreas development	shortened pancreas, reduction in β cell number	(Mastracci et al. 2015; Anderson et al. 2025)
eif5a morpholino	pancreas development	normal pancreas size	(Anderson et al. 2025)
dhps and eif5a morpholino	pancreas development	normal pancreas size	(Anderson et al. 2025)
Spermidine supplementation	N/A	suppressed ROS acclimatation, decreased inflammation	(Jeong et al. 2018)

Table 2.

Mouse Models

Strain	Genetic/Inhibitor manipulation	Disease model	Major findings	Reference
C57BL/6	Dhps pancreas knockout	pancreas development	reduced acinar tissue death by 6 weeks of age	(Padgett et al. 2021)
C57BL/6	Dhps β cell knockout	pancreas development	adult hypoglycemia, reduced insulin positive cells	(Connors et al. 2023)
NOD	ODC inhibition—DFMO	T1D	delayed onset, increased Treg cells	(Tersey et al. 2014; Hammoud et al. 2024)
NOD	DHPS inhibition—Gc7	T1D	delayed onset, decreased Th1 cells	(Colvin et al. 2013)
NOD	ODC and SST inhibition—DENSpm	T1D	no protection, increased death	(Hammoud et al. 2024)
NOD	Spermidine supplementation	T1D	increased diabetes incidence	(Karacay et al. 2022)
C57BL/6	DHPS inhibition—Gc7	STZ-induced diabetes	reduction in hyperglycemia, preserved β cell mass	(Maier et al. 2010)
C57BL/6	Eif5a siRNA knockdown	STZ-induced diabetes	reduction in hyperglycemia, preserved β cell mass	(Maier et al. 2010)
C57Bl/6	Odc1 β cell knockout	STZ-induced diabetes	reduction in hyperglycemia, preserved β cell mass	(Sims et al. 2023)
C57BL/6 (high fat diet)	Dhps β cell knockout	Obesity glucose intolerance	reduced β cell mass, increased hyperglycemia	(Levasseur et al. 2019; Anderson et al. 2025)
C57BL/6 (high fat diet)	Eif5a β cell knockout	Obesity glucose intolerance	normal β cell mass, normal glycemia	(Anderson et al. 2025)
C57BL/6 (high fat diet)	ODC inhibition—DFMO	Obesity glucose intolerance	reduced β cell mass, increased hyperglycemia	(Levasseur et al. 2019)
C57BL/6	Sat1 overexpression	N/A	increased β cell mass, impaired glucose tolerance	(Cerrada-Gimenez et al. 2012)
C57BL/6	Sat1 whole body knockout	aging	aged-induced insulin resistance	(Niiranen et al. 2006)

Deletion of Dhps in pancreas progenitor cells led to a significant reduction in body growth, hypoglycemia, and death by 6 weeks of age in both male and female knockout mice (Padgett et al. 2021). These knockout mice had significantly reduced pancreas size with reductions in the acinar tissue compartment, however, both insulin- and glucagon-positive islets were identified and unchanged. This was due to a block in mRNA translation elongation, leading to decreased levels of digestive enzymes or proteins involved in cellular development (Padgett et al. 2021). When Dhps is deleted in β cells, the development of the mice as well as the islet composition and β cell mass was unchanged compared to control littermates (Connors et al. 2023). However, by 5–6 weeks of age, the mice developed hypoglycemia with a significant reduction in insulin-positive β cells. This reduction in the number of β cells appeared to result from a failure to translate proteins critical to β cell health and function (Connors et al. 2023). Together, these studies highlight the roles of polyamines and hypusine in regulating pancreatic cell growth during development and β cell function in the adult, using both zebrafish and mice as model systems.

Polyamines in islet and β cell mass

Islet mass is dynamically regulated by a balance between β cell proliferation, differentiation/neogenesis, and death. Polyamines are critical to cellular proliferation, positioning these metabolites at a vital nexus whereby dietary and developmental cues integrate to shape islet mass. A window in early life has been observed in both humans and rodents, during which β cells undergo a proliferative burst that establishes most of the islet mass that persists into adulthood (Finegood et al. 1995; Meier et al. 2008). This rapid increase in β cells peaks within the first 2 years of life in infants (Gregg et al. 2012) and coincides with major developmental milestones, such as the transition from breastfeeding/formula feeding to solid foods and the diversification of the gut microbiota. New studies suggest that microbial products induce β cell proliferation and the timing of their exposure influences the susceptibility to both autoimmune and diet-induced diabetes in mice (Hill et al. 2016, 2022, 2025). It is tempting to speculate that the polyamines derived from the infant diet or from commensal bacteria are also involved in this early life spike in β cell proliferation, but as of now these studies are lacking. Despite the aforementioned roles of polyamines in pancreas and islet development, the roles of polyamines in cell proliferation are seemingly less applicable to mature β cells. In both humans and rodents, β cells show very slow replication rates that also decline in adulthood (Teta et al. 2005; Gregg et al. 2012). Moreover, the generation of β cells seems to be driven by duplication from other adult β cells rather than differentiation from stem cells, which poses a challenge to the regeneration of functional islet mass in diabetes (Dor et al. 2004). It is important to note however that β cells can exhibit facultative proliferation as an early adaptive response to metabolic stressors. Mice that are fed a high fat diet exhibit islet hypertrophy and are characterized by an increase in proliferative β cells to compensate for the increased demands for insulin (Stamateris et al. 2013; Mosser et al. 2015). Non-obese diabetic (NOD) mice, a model of autoimmune diabetes, also display an increase in β cell proliferation prior to the onset of autoimmune diabetes, but these mice ultimately do not have gains in islet mass due to β cell destruction (Sreenan et al. 1999).

In a zebrafish model of β cell ablation and proliferation, in which nifurpirinol (NFP) induces cellular ablation of β cells in genetically altered zebrafish, ODC inhibition through DFMO dosing led to an increase in insulin-positive cellular proliferation (Robertson et al. 2020). While the molecular mechanisms driving facultative proliferation in β cells are still being elucidated, novel mouse models have helped illuminate the role of polyamines. Through the development and use of a transgenic mouse with a tamoxifen-inducible β cell-specific knockout of Dhps, it was observed that β cell-specific Dhps deletion impaired glucose tolerance under high fat diet conditions and inhibited translation of the cell cycle regulator cyclin D2 (Fig. 2)(Levasseur et al. 2019). Decreased levels of Ki67 staining were observed in the islets of these mice after 1 week of high fat diet feeding, and polyamine depletion with DFMO also mitigated the effects of a high fat diet on β cell proliferation (Levasseur et al. 2019). In addition to polyamines, growth factor signaling through the mTORC1 axis is also known to promote β cell proliferation and affect β cell mass through the upregulation of the cyclin D proteins (Balcazar et al. 2009). Protein kinase C-ζ (PKC-ζ) is an activator of mTORC1 in β cells (Velazquez-Garcia et al. 2011) and mice who harbor β cells with a kinase-dead PKC-ζ, and are fed a high fat diet, do not exhibit compensatory β cell hypertrophy (Lakshmipathi et al. 2016). Islets from both the β cell-specific Dhps knockout mouse and the kinase-dead PKC-ζ mouse have lower levels of eIF5A^Hyp when fed a high fat diet; however, phosphorylation of PKC-ζ in the islets of β cell-specific Dhps knockout mice remains intact, which positions DHPS downstream of PKC-ζ (Levasseur et al. 2019). It is possible that the decline in eIF5A^Hyp in kinase-dead PKC-ζ β cells could also be attributable to a role for mTORC1 in regulating AMD1 stability (Zabala-Letona et al. 2017), but this has not yet been explored in the context of facultative adaptation in β cells. Overall, the β cell-specific knockout of Dhps mouse was an important first step towards understanding the β cell-intrinsic mechanisms by which polyamines and the downstream DHPS/hypusine axis regulate β cell proliferation. However, the inducible knockout of total eIF5A in β cells, when compared to the Dhps-Δβ mice, did not impair the effects of a high fat diet to increase islet mass (Anderson et al. 2025). Taken together, the findings from these mouse models and from the previously mentioned zebrafish models suggest there could be an antagonistic effect by the accumulation of eIF5A^Lys on β cell replication or the potential for alternative targets of DHPS activity that would promote replication. However, the studies so far have only been in the context of an adaptation to a high fat diet. Backcrossing these mice onto the NOD background as a model for T1D will be valuable for understanding how the autoimmune process relates to β cell proliferation and islet mass.

Fig. 2.

Diagram of the role of polyamines in the molecular process of adaptive β cell proliferation. A high fat diet promotes a cycle of stress and dysfunction in metabolic tissues that contributes to an adaptive proliferative response in β cells. Components of the polyamine pathway, ODC and DHPS, are downstream targets of molecular processes known to be involved in cellular proliferation such as the PI3K/mTOR axis. The requirement for DHPS but not eIF5A in β cell proliferation suggests that alternative targets of DHPS activity or unhypusinated eIF5A^Lys could have roles in regulating the proliferation observed in early T2D pathogenesis. Image created in BioRender. PI3K phosphoinositide 3-kinase; PKCζ protein kinase C ζ; mTORC1 mechanistic target of rapamycin complex 1; ODC: ornithine decarboxylase; DHPS deoxyhypusine synthase; eIF5A^Lys50 unhypusinated eIF5A; eIF5A^Hyp hypusinated eIF5A

Polyamines in islet function

The primary function of islet β cells is to secrete insulin in response to a rise in blood glucose levels. Early studies using ¹⁴C-labelled putrescine revealed a link between polyamines and glucose-stimulated insulin secretion (GSIS). The incorporation of ¹⁴C-labelled putrescine into islet proteins after a glucose challenge suggested that polyamines are substrates for calcium-dependent transglutaminase activity during insulin release (Bungay et al. 1984). In subsequent cytochemical studies, high levels of polyamines were detected in the pancreas of rats and mice, and notably, β cells showed a high ratio of spermine to spermidine (Hougaard and Larsson 1986). The high polyamine content in β cells suggests a role for polyamines in mRNA expression and protein translation, as β cells must meet secretory demands to produce insulin and maintain normoglycemia. Early studies also showed that polyamines are localized within the secretory granules of β cells. This subcellular localization of polyamines in β cells further supports the notion that polyamines are involved in the regulation of insulin storage and secretion, but it also raises the question of how different intracellular polyamine stores affect β cell function, since nuclear polyamines were unaffected by DFMO treatment (Sjoholm et al. 1990). Nevertheless, numerous studies have provided evidence linking polyamines to insulin secretion in cultured islets and have revealed modulation of polyamine content in response to glucose conditions (Sjoholm 1993). It was observed that in both mouse and rat islets cultured under high (16.7 mM) glucose concentrations for 48–72 h, there is an increase in spermidine and spermine content compared to islets cultured in low (3.3 mM) glucose conditions (Welsh and Sjöholm 1988; Sjoholm et al. 1990). The increase in islet polyamine content parallels the increases in glucose-stimulated insulin mRNA synthesis, which are both lowered by a combined treatment containing DFMO and EGBG (Welsh 1990). Glucose-stimulated insulin biosynthesis, total protein synthesis, and insulin release were also lowered by this combination of polyamine pathway inhibitors which further illustrates the role for polyamines in controlling transcription and translation in β cells (Welsh and Sjöholm 1988). Interestingly, the production and secretion of insulin by both rat and mouse islets were most sensitive to the levels of spermine, which was also determined from the combined inhibition by DFMO and EGBG (Welsh and Sjöholm 1988; Welsh 1990; Sjoholm et al. 1990). Consistent with these observations, the ratio of spermine to spermidine is high in adult rat islets, whereas spermidine is much higher in fetal islets (Sjoholm et al. 1990; Sjoholm 1993). The rapid response to glucose stimulation by mature islets could therefore depend on the relative availability of individual polyamines, and that polyamine flux in β cells could be a determinant of glucose homeostasis.

In studies utilizing transgenic mice that overexpress SAT1 under the control of a metallothionein I promoter, the islet polyamine pools were depleted of spermidine and spermine. In contrast, putrescine and N¹-acetyl spermidine were increased (Cerrada-Gimenez et al. 2012). Although this model does not exclusively increase SAT1 activity in the pancreas, the islets of these mice displayed an increase in β cell area, a decrease in insulin production, and impaired glucose tolerance (Cerrada-Gimenez et al. 2012). In contrast, mice deficient in SAT1 showed increased spermidine and spermine levels but eventually developed insulin resistance with age (Niiranen et al. 2006). Overall, both SAT1 mouse models demonstrate that perturbations in polyamine flux can impair islet function, with substantial implications for the decline in polyamines known to occur with aging.

In the T2D-prone ob/ob mouse, it was observed that polyamine content decreased in the islets of ob/ob mice as they aged from 2 to 8 months and was also lower compared to the islets from 2-month-old lean mice (Sjöholm et al. 2001). A limitation to the study, however, is the absence of a lean, 8-month-old control group, which begs the question of whether normal islets resist age-related polyamine decline. In a separate study, polyamine content was seemingly unchanged in the pancreas of 3-, 10-, and 26-week-old female mice, but the individual contributions of polyamines from exocrine pancreas and islets were not discerned (Nishimura et al. 2006). Altogether, polyamines play quintessential roles in regulating islet function by supporting insulin production and secretion. Additional work with β cell-specific deletions of polyamine pathway genes, along with more sensitive measurement techniques such as mass spectrometry and novel biosensors (Sharma et al. 2025), will be instrumental in clarifying how polyamines and their metabolic flux ultimately regulate glucose homeostasis.

Polyamines in the oxidative and ER stress responses of β cells

Pancreatic β cells are highly susceptible to oxidative and endoplasmic reticulum (ER) stress due to the high demand to produce insulin and are poorly equipped to deal with oxidative stress, given their low expression of antioxidant proteins (Delmastro and Piganelli 2011). As such, persistent reactive oxygen species (ROS) in β cells can lead to protein misfolding and triggering of the unfolded protein response (UPR). The activation of the UPR relieves ER stress by inducing protein chaperones, but chronic UPR activation can lead to a maladaptive response that ultimately triggers apoptosis (Kulkarni et al. 2022b). Given this interplay between oxidative stress, ER stress, and β cell dysfunction, β cell stress is relevant to the pathogenesis of both T1D and T2D. In the context of T1D, stressed β cells produce excess neoantigens and neoepitopes. These β cell neoantigens and neoepitopes contribute to enhanced immunogenicity, resulting in the accumulation of autoimmune B cell and T cell responses that lead to the destruction of pancreatic islet β cells (James et al. 2018; Piganelli et al. 2021). Alternatively, for both T1D and T2D, excessive ROS can activate signaling pathways such as mitogen-activated protein kinase (MAPK), c-Jun N-terminal kinase (JNK), and nuclear factor-kB (NF-kB), which increase inflammation and induce β cell apoptosis (Kulkarni et al. 2022b). The direct inhibition of oxidative stress or ER stress pathways is protective in mouse models of T1D (Muralidharan et al. 2024; Nargis et al. 2024) and T2D (Kaylan et al. 2025). These studies support the notion that ameliorating oxidative and ER stress in β cells will be vital to preventing and treating diabetes.

Polyamine metabolism may have context-dependent roles in regulating the UPR and in response to oxidative and ER stress. Spermidine and spermine protect cells against hydrogen peroxide-induced oxidative stress through a mechanism distinct from that of classical antioxidants such as glutathione (Rider et al. 2007; Smirnova et al. 2018). Spermine prevents mitochondrial swelling and a decrease of mitochondrial potential, and polyamines bound to DNA protect from ROS-induced damage (Muscari et al. 1995). And spermine induces the UPR through splicing of the bZIP60 transcript in Arabidopsis (Sagor et al. 2015), which could act as a compensatory mechanism to prevent further activation of ER stress and apoptosis. Food supplementation with spermidine confirmed resistance to H₂O₂ and paraquat, an inducer of superoxide anion production in Drosophila melanogaster (Minois et al. 2012) and decreased oxidative stress markers in aged mice (Eisenberg et al. 2009). In a zebrafish model, spermidine supplementation significantly suppressed ROS production and, consequently, alleviated lipopolysaccharide-triggered inflammation (Jeong et al. 2018). These studies illustrate how polyamines and their supplementation to model organisms can protect against oxidative stress and induce the UPR, but there are situations in which polyamines contribute to cellular stress. Polyamine catabolism via SMOX and SSAT enhances the turnover of spermine and spermidine, leading to overproduction of H₂O₂ and acrolein, a byproduct that also causes oxidative stress, suggesting that an imbalance in polyamine levels results in oxidative stress. When spermidine was given to NOD mice there was an increase in diabetes incidence (Karacay et al. 2022). Additionally, when β cells or human islets treated with proinflammatory cytokines to mimic the inflammation seen during T1D initiation were also treated with DFMO to block ODC activity, there was a decrease in pathways associated with oxidative and ER stress (Sims et al. 2023; Hammoud et al. 2024). Whereas polyamines are associated with antioxidant activity, the interconversion of spermidine and spermine within β cells could nonetheless generate ROS (Murray Stewart et al. 2018) and contribute to the cascade of events that will be fatal to the β cell. Together, these led to the hypothesis that either different cell types and/or different disease models respond differently to polyamine levels during oxidative and/or ER stress. Therefore, it is important to consider the contexts and approaches in which polyamines are being added or depleted as the perturbation could either lower or promote cell stress.

Polyamines in inflammation

Inflammation is a key driver of β cell dysfunction in both T1D and T2D. Multiple factors can provoke inflammation in β cells, such as proinflammatory cytokines, free fatty acids, elevated glucose, and reactive oxygen species (ROS). eIF5A^Hyp has been shown to promote the translation of mRNAs associated with cellular stress and inflammation. One such mRNA is Nos2 (encoding iNOS), a key enzyme that is upregulated during cytokine-induced β cell dysfunction (Fig. 3)(Maier et al. 2010). Studies in mouse islets demonstrated that siRNA-mediated knockdown of eIF5A blunts Nos2 translation after cytokine treatment. Moreover, eIF5A^Hyp was shown to be required for the nuclear export of Nos2 transcripts. These studies helped rationalize targeting polyamines and hypusination to attenuate inflammation in β cells.

Fig. 3.

A diagram of the role of polyamines in the molecular process of β cell inflammation. Polyamines have both direct and indirect roles in driving β cell inflammation and T1D pathogenesis. Polyamine catabolism in β cells contributes to the accumulation of reactive oxygen species (ROS) which can lead to β cell dysfunction. Alternatively, polyamines can lead to the generation of hypusine which is required for the known functions of eIF5A^Hyp involving the export and translation of mRNAs associated with inflammation. The culmination of inflammatory signaling along with ROS and ER stress can elicit the formation of neoantigens and neoepitopes that enhance the immunogenicity of β cells and drive autoimmunity. Polyamines could also be involved in regulating the effector functions and proliferation of immune cells that infiltrate the islet. Image created in BioRender. SPD spermidine; SPM spermine; SMOX spermine oxidase; SSAT spermidine/spermine N1-acetyltransferase; PAOX polyamine oxidase; iNOS inducible nitric oxide synthase; ROS reactive oxygen species

In addition to their role in β cells, polyamines are essential regulators of innate and adaptive immunity. Macrophage polarization between M1 and M2 fates is known to be influenced by polyamines (Latour et al. 2020). Other studies demonstrated that Gc7-mediated inhibition of DHPS activity impaired mitochondrial gene translation, leading to reduced oxidative phosphorylation and favoring polarization toward an M2-like state. In contrast to these findings, a study showed that macrophages deficient in Dhps exhibit impaired polarization towards a classical M1-like state, with lower levels of pro-inflammatory proteins (Anderson-Baucum et al. 2021). In the context of a high fat diet, mice harboring Dhps-deficient macrophages showed improved glucose tolerance and reduced pro-inflammatory macrophage infiltration into adipose tissue. The discrepancy could lie with potential off-target effects concerning Gc7 treatment versus a macrophage-specific knockout.

T cells undergo metabolic adaptations to support their proliferation after activation, which strongly implicates a role for polyamine metabolism in T cell responses. CD4 + T cells deficient in Odc1, Dohh, or Dhps exhibit delayed proliferation and dysregulated differentiation into T helper cell subsets, linked to perturbed histone acetylation and TCA cycle metabolism (Puleston et al. 2021). A role for polyamine metabolism in regulating T helper cell differentiation was further supported by in silico modeling of single-cell RNA-sequencing data with metabolic flux balance analysis (Wagner et al. 2021). Predictions from the model associated Th17 pathogenicity with polyamines. In vitro validation experiments demonstrated that DFMO inhibited ODC and restricted Th17 differentiation in a putrescine-dependent manner. Using ATAC-seq, it was found that DFMO treatment altered the open chromatin landscape, favoring Treg-specific regions while restricting Th17-specific regions. The effects of ODC inhibition on T cells in vitro are consistent with the observed effects of DFMO in NOD mice, as DFMO treatment shifted the balance of T cells towards Tregs and away from Th17 cells in the pancreatic lymph nodes (Tersey et al. 2014). It is also possible that polyamines are involved in the dialogue between β cells and immune cells. Using untargeted metabolomics, studies identified a conserved metabolite secretome across apoptotic primary lymphocytes and macrophages, as well as cell lines (Medina et al. 2020). Their unbiased profiling identified a cocktail of metabolites that could attenuate inflammation, including spermidine. Along this line of research, macrophages were shown to increase polyamine content after engulfing apoptotic cells through a process known as efferocytosis (McCubbrey et al. 2022). The increase in polyamines suppressed IL-1β and IL-6, which suggests a mechanism by which a damaged tissue environment could bias macrophages towards a pro-reparative M2-like phenotype. Finally, β cells have also been shown to communicate with immune cells through the production of extracellular vesicles (EVs), which contain cargo such as miRNAs and the immune checkpoint protein PD-L1 (Rao et al. 2025). Given the high polyamine content of β cells, it would be interesting to determine whether β cells supply polyamines to recipient immune cells via EVs and whether β cell-derived polyamines influence immune cell proliferation or function.

Polyamines in diabetes

The pathogenesis of both T1D and T2D results from insufficient insulin production by pancreatic islet β cells (Weir and Bonner-Weir 2013). In 2019, 463 million people (9.3% of the world’s population) were estimated to have diabetes, and an additional 374 million (7.5% of the world’s population) were estimated to have prediabetes. By 2045, that number is predicted to rise to 700 million (10.9%) with diabetes and 548 million (8.6%) with prediabetes (Saeedi et al. 2019). The scientific literature over the past 50 years has increasingly pointed to a role for dysregulated cellular metabolism in multiple cell types, particularly β cells, hepatocytes, muscle cells, and immune cells in the development of diabetes (Hameed et al. 2015). The availability of genetic tools, in particular, has enabled the discovery of a crucial role for polyamine metabolism as a pathogenic factor in diabetes development.

Inhibition of polyamine metabolism in T1D pathogenesis

T1D emanates from the autoimmune destruction of the pancreatic islet β cell as a consequence of a complex interplay of genetic, environmental, immune, and pancreatic cell-intrinsic factors that contribute to the pathogenesis of T1D. These factors lead to the activation of both innate and adaptive autoimmunity, and progressive β cell destruction, ultimately resulting in insulin deficiency and dysglycemia (Atkinson et al. 2011). The established model of T1D is one in which β cells are progressively destroyed by T-cell-mediated autoimmunity. However, there is increasing evidence that β cells initiate their own demise. Recent scientific efforts have added to this model with the belief that there is a specific role of the β cell in the early inflammatory processes, which feeds forward into both β cell dysfunction and autoimmunity (Roep et al. 2021; Atkinson and Mirmira 2023).

Individuals with T1D exhibit increased levels of eIF5A^Hyp within the pancreatic islet (Mastracci et al. 2020), suggesting that increased polyamine levels may play a role in the pathogenesis of T1D, and inhibiting polyamine metabolism and/or eIF5A^Hyp might be a mechanism of alleviating β cell dysfunction. To this end, it was demonstrated that injecting mice with siRNA targeting Eif5a protects against multiple low-dose streptozotocin (STZ)-induced diabetes, a model that modestly stimulates macrophage infiltration, similar to T1D (Maier et al. 2010). Additionally, another study showed that reducing eIF5A^Hyp by inhibiting DHPS with Gc7 delayed the onset of autoimmune diabetes in female NOD mice (Colvin et al. 2013). Upstream of hypusine, ODC has also been inhibited using DFMO or conditionally knocked out in β cells and protection has been seen in either NOD mice or STZ-induced diabetes (Tersey et al. 2014; Sims et al. 2023). Alternatively, when spermidine was given to NOD mice, there was an increase in diabetes incidence (Karacay et al. 2022). Taken together, these findings show that excessive polyamines lead to an increase in autoimmune diabetes and that disrupting various enzymes in the polyamine/hypusine pathway leads to the delay of autoimmune diabetes in the NOD mouse model.

A phase 1b clinical trial (ClinicalTrials.gov: NCT02384889) demonstrated that higher DFMO doses conferred metabolic benefits by preserving β cell function. Individuals diagnosed with recent-onset T1D were treated with DFMO to inhibit polyamine synthesis in a dose-escalating toxicity study. Polyamine metabolism was monitored by urinary putrescine levels, a downstream product of ODC activity, and individuals treated with higher DFMO doses exhibited lower urinary putrescine levels, confirming effective ODC inhibition. After 3 months of treatment, participants receiving higher doses of DFMO showed significantly higher C-peptide area under the curve (AUC) levels than those receiving placebo, indicating improved or maintained β cell function with no significant changes in circulating immune cell levels (Sims et al. 2023). This study showcased that DFMO treatment enhances β cell function and suggests the potential that when combined with immunomodulatory agents, may further increase β cell survival in individuals with T1D. Collectively, these studies provide critical insight into the mechanisms of the polyamine pathway and its contribution to the pathogenesis of T1D and inhibiting polyamine metabolism may offer new perspectives for therapeutic interventions.

Polyamines in T2D

T2D is a chronic metabolic disorder characterized by insulin resistance and the progressive dysfunction of pancreatic β cells (Freeman et al. 2025). In individuals with T2D, peripheral tissues such as skeletal muscle, adipose tissue, and the liver become less responsive to insulin, leading to impaired glucose uptake and increased hepatic glucose production. This insulin resistance is often driven by a genetic predisposition, combined with environmental and lifestyle factors, particularly obesity and nutrient excess. Visceral adiposity promotes low-grade, chronic inflammation through the release of pro-inflammatory cytokines like TNF-α and IL-6, which disrupt insulin signaling pathways. Additionally, elevated levels of circulating free fatty acids, especially saturated species like palmitate, contribute to lipotoxicity and metabolic stress in both insulin-sensitive tissues and β cells. In response to increasing insulin demands, β cells initially undergo hypertrophy and hyperplasia, temporarily maintaining glycemic control. However, prolonged metabolic strain leads to β cell dysfunction and eventual failure (Hameed et al. 2015).

As previously mentioned, polyamines are essential for cellular growth, proliferation, and protein synthesis, but human studies and mouse models suggest dysregulation of polyamine metabolism is associated with the pathogenesis of T2D. High polyamine levels, specifically putrescine and spermine, positively correlate with T2D (Fernandez-Garcia et al. 2019), and serum spermine levels are significantly higher in obese children compared to non-obese controls (Codoñer-Franch et al. 2011). Mouse models of T2D support the notion that polyamine imbalances dysregulate islet function. Within mouse islets, the ratio of spermine is downregulated, and spermidine is unchanged in a model of obesity-induced diabetes (Sjöholm et al. 2001). And despite polyamine metabolism already being increased under high fat diet conditions, the administration of spermidine or spermine to high fat diet-fed obese mice resulted in a reduction in body weight, improved glucose tolerance, and reduced insulin resistance (Sadasivan et al. 2014; Fernández et al. 2017).

Polyamines exert their effects not only through direct interactions with DNA and RNA to regulate transcription and translation but also via the hypusine modification of eIF5A, which is necessary for the translation of proteins involved in stress responses, inflammation, and cell survival. Dysfunction of this polyamine–hypusine circuit contributes to β cell stress and death in T2D, particularly under conditions of acute overnutrition and inflammation. β cell knockout of Dhps led to decreased eIF5A^Hyp in β cells, which in turn reduced β cell proliferation, decreased β cell mass, and ultimately increased hyperglycemia within 4 weeks of high fat diet treatment (Levasseur et al. 2019; Anderson et al. 2025). However, when eIF5A was deleted from the β cell, the mice did not exhibit the same phenotype (Anderson et al. 2025); suggesting that eIF5A^Lys exhibits its own deleterious effects on the β cell in addition to the lack of eIF5A^Hyp. Alternatively, knockout of Dhps in the macrophage of high fat diet-treated mice led to a reduction of pro-inflammatory macrophages in adipose tissue, resulting in improved glucose tolerance and reduced insulin resistance (Anderson-Baucum et al. 2021).

In addition to endogenous sources, polyamines are produced by the gut microbiota and obtained through dietary intake. Commensal bacteria capable of synthesizing polyamines contribute to circulating levels (Nakamura et al. 2021), which can impact β cell health and systemic glucose regulation. A wide range of bacterial isolates from humans have been shown to produce polyamines which include Bifidobacteria, Clostridia, Enterococcus, Lactobacillus, and Bacteroides (Pugin et al. 2017; Kitada et al. 2018). Dysbiosis of the gut microbiota associated with T2D often reduces the abundance of polyamine-producing microbes, potentially exacerbating polyamine deficiency in pancreatic β cells and peripheral tissues (Bui et al. 2022). Conversely, dietary supplementation or microbiota-targeted interventions that increase polyamine availability have been shown in preclinical models to improve glucose tolerance, enhance insulin sensitivity, reduce inflammation, and protect β cell mass, highlighting the importance of host–microbiota interactions in polyamine-mediated metabolic regulation (Ma et al. 2022; Bui et al. 2022). However, as stated earlier, the contexts by which polyamines are perturbed could yield protective or detrimental effects on different cell types and tissues, so simply increasing polyamine output from the microbiota may not be the most generalizable strategy for diabetes. Longitudinal microbiota profiling in The Environmental Determinants of Diabetes in the Young (TEDDY) study identified an increase in the L-arginine, putrescine, and 4-aminobutanoate super pathway among the T1D controls cohort-wide (Vatanen et al. 2018) which could link microbial fermentation with altered polyamine metabolism in the gut microbiota. In a human cohort of adults with T1D, a 6-week intervention with a biotherapy aimed at raising microbial fermentation products also induced metabolic shifts associated with a decrease in arginine and ornithine metabolism (Tillett et al. 2025). Microbiota transplants from the people who responded to the biotherapy into germ-free NOD mice recapitulated the metabolic shifts observed in the human responders and delayed diabetes incidence in the NOD mice (Tillett et al. 2025). Along with ODC inhibition with oral DFMO delaying diabetes in NOD mice (Tersey et al. 2014), the observation of decreased arginine and ornithine metabolism in controls from the TEDDY cohort and in the responders to biotherapy could indicate that a reduction in polyamine metabolism in the intestines contributes towards promoting immune regulation in the pancreatic islets. High levels of microbial-derived cadaverine in the context of intestinal inflammation drives gut macrophages towards a proinflammatory state (de Oliveira Formiga et al. 2025). Islet-resident macrophages support the growth and function of β cells (Grosjean et al. 2025) and they possess a steady-state inflammatory gene signature that suggests they are poised to respond to pathogens (Ferris et al. 2017). It could be possible that fluctuations in polyamine metabolism in the islets, and polyamines circulating from the gut microbiota, are influencing the inflammatory tone of islet-resident macrophages to either exacerbate islet inflammation or maintain homeostasis. Collectively, these findings underscore that maintaining proper polyamine metabolism, both through endogenous synthesis and microbial contributions, is essential for β cell function and systemic metabolic homeostasis, and that inhibition or dysregulation of this network contributes directly to both T1D and T2D pathogenesis.

Current limitations and future directions

Pharmacological inhibitors have provided valuable insights into the roles of polyamines in preclinical models of diabetes, with DFMO paving the way for polyamine-depletion as a β cell-preserving therapy for T1D. Yet the systemic effects of DFMO and other inhibitors make it difficult to pinpoint the precise mechanisms of polyamine manipulation in vivo. It is conceivable that systemic polyamine depletion, whether via dietary or pharmacological approaches, is also affecting microbial-derived polyamines, warranting further investigation with antibiotics or germ-free models to tease out the contributions of polyamines from the host and the microbiota to islet biology and diabetic contexts. The continued development of animal models with inducible gene knockouts will provide valuable tools for the field to study the polyamine pathway in the pancreas and islets systematically. In addition to pancreatic exocrine and endocrine cells, macrophages and lymphocytes will be interesting candidates for cell-specific knockouts that can inform the roles of polyamines and immune cells in the pancreas and in diabetes pathogenesis. Along with further animal models, human studies are necessary to confirm animal model findings to help inform the next generation of polyamine-targeted therapeutics for diabetes.

Conclusions

Polyamines are integral to numerous cellular processes that are vital to the normal development, growth, and function of pancreatic islets and β cells. However, because polyamines have diverse functions that also affect β cell stress, inflammation, and immune cells, polyamine imbalances could be a central feature of diabetes pathogenesis. In this review, we have discussed how polyamine metabolism integrates nutrient status, cellular stress, and immune activation in the pancreas and in the pathogenesis of diabetes. It is important to note that modification of polyamine/hypusine pathway influences different stages of diabetes in very specific and unique ways in such that inhibiting polyamine production is beneficial in blocking both β cell stress and immune response but is determinantal to β cell development and proliferation. Deciphering these cell specific and potentially timing specific influences of polyamines to the pancreas development and disease will pave the way for potential future clinical therapeutics.

Author contributions

A.A.P., L.K., L.J.C, and S.A.T. wrote the main manuscript. A.A.P., R.G.M., and S.A.T. edited the manuscript. A.A.P and S.A.T. prepared figures. All Authors reviewed the manuscript.

Funding

This work was supported in part by National Institutes of Health grants R01 DK124906 (SAT), R01 DK124906 (RGM), Diabetes Research Connection (AAP), and the Katen Scholars program (LJC). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Conflict of interest

The authors declare no competing interests.