Microparticles synthesized from itaconate polyesters enable metabolite delivery and elicit immunoregulatory outcomes

Abstract

Metabolites directly influence immune cell function. Consequently, altering metabolite availability can regulate immunity and represents a window of opportunity for therapeutic manipulation. Itaconate (ITA) – a tricarboxylic acid cycle derivative generated intracellularly – has been shown to exert anti-inflammatory effects in models of autoimmunity. However, efficacy is dependent on systemic, frequent, high-dose treatments due to ITA’s poor membrane permeability and rapid clearance from circulation. Thus, there is a need for metabolite delivery strategies that overcome these challenges. Here, we synthesized ITA-based polyesters that enable microparticle (MP) assembly entirely from the polymeric metabolite (pITA). Treatment with biocompatible pITA MPs, which hydrolyze to release unmodified ITA, reduced inflammatory cytokine secretion (e.g., IL-6, IL-1β) by both macrophages and dendritic cells. These data suggest MPs effectively deliver metabolite to the intracellular compartment for immunomodulation in an accessible form. Furthermore, pITA MPs inhibited polarization of inflammatory T cells (T_H17) and promoted regulatory T cells (T_reg), linking the effect of MPs on innate signaling to adaptive immunity. In a preclinical mouse model of multiple sclerosis (MS), pITA MPs reduced circulating serum IL-6 – a cytokine that drives inflammation in several autoimmune diseases. Together, this work advances a biomaterials-based strategy that overcomes hurdles to metabolite delivery and expands opportunities for intervention during autoimmunity and inflammation.

INTRODUCTION

Immune cells rely on distinct signaling and metabolic pathways to promote their survival, lineage commitment, and function [1, 2]. Itaconate (ITA) is a metabolite derived intracellularly from the decarboxylation of the tricarboxylic acid (TCA) cycle intermediate, cis-aconitate [3], and is emerging as a key regulator of inflammation through diverse mechanisms [4]. In response to inflammatory cues, ITA rapidly accumulates in healthy macrophages and controls cell metabolism through direct inhibition of succinate dehydrogenase (SDH) [5, 6]. In addition, ITA acts as a complex signaling molecule that controls inflammation through transcriptional activity (e.g., IL-6) [7], protein modifications [8–11], inflammasome activation (e.g., IL-1β) [12, 13], and epigenetic mechanisms [14]. Accordingly, modulating the availability of intracellular ITA is emerging as a highly relevant therapeutic strategy for redirecting dysregulated immune responses.

Autoimmune diseases, which affect millions of individuals worldwide, are heterogeneous disorders with no cure. While their etiology is not fully understood, strong evidence links autoimmune pathogenesis to inflammation [15]. In multiple sclerosis (MS), for example, elevated IL-6 – an inflammatory cytokine that promotes differentiation of autoreactive T cells – polarizes T_H17s that mistakenly attack myelin and lead to neurodegeneration [16–18]. Thus, the association between disease pathology and excessive activation of inflammatory pathways make autoimmune diseases an attractive target for metabolite-based therapies that reduce inflammation. This rationale is evidenced by the success of dimethyl fumarate (DMF)(Tecfidera^™) – a methyl ester derivative of the TCA cycle metabolite, fumarate – that is prescribed to individuals with both psoriasis and MS [19]. However, dose-limiting side effects include lymphopenia, gastrointestinal symptoms, and in some cases – severe opportunistic infection [20, 21]. In contrast, targeted delivery of ITA could enable more pathway-specific immune modulation with potentially improved tolerability and reduced off-target effects [22]. Although therapeutic applications for ITA are still in the preclinical research phase, results are promising: ITA regulates experimental autoimmune encephalomyelitis (EAE) [23], psoriasis [7], systemic lupus erythematosus (SLE) [24], and type 1 diabetes [25]. Notably however, low molecular weight metabolites are susceptible to rapid clearance and degradation in circulation when administered systemically [26]. Consequently, efficacy in preclinical models of autoimmune disease is dependent on systemic, frequent, high-dose treatments [7, 23, 27] that pose potential hurdles for future translation to patients.

ITA faces an additional, notable obstacle for efficient administration – a requirement for intracellular delivery [3, 28]. The metabolite is endogenously synthesized and due to its polarity, exhibits low membrane permeability when administered extracellularly [29]. For this reason, derivatives of ITA such as 4-octyl ITA (4-OI) and dimethyl ITA (DMI) that are highly membrane-permeable at lower doses have been relied on to elucidate ITA’s complex mechanisms of action [6, 7, 9, 10]. Excitingly, ITA mimetics (SYX-1042) are currently under clinical development for Phase I trials (Sitryx Therapeutics). Still, studies suggest that derivatives are not directly metabolized to ITA [30] or exhibit off-target effects attributed instead to their relatively higher magnitude of electrophilicity [7, 29]. Thus, there is significant interest in novel strategies that deliver unmodified ITA directly to the intracellular compartment to enable recapitulation of endogenous ITA signaling.

Biomaterials-based delivery could overcome obstacles to ITA administration through protection from degradation, intracellular targeting, and controlled release [31–33]. Toward this goal, lipid-nanoparticles [34] and “carrier-free” particles assembled from polymeric metabolites such as α-ketoglutarate [35–37], succinate [38], and ITA [39, 40] have been investigated. Furthermore, synthesizing particles in the micron-sized range takes advantage of the phagocytic behavior of innate cells, which are specialized to engulf larger particulates for pathogen clearance [41], and allows for passive targeting of cells that are significant sources of inflammatory cytokines. Nonetheless, achieving sufficient uptake to elicit anti-inflammatory effects can be challenging. A better understanding of the role of bulk material molecular weight and particle morphology, and their relation to cellular uptake and ITA release, could further strengthen metabolite delivery strategies.

To advance approaches to ITA delivery for controlling inflammation, we developed a tunable biomaterials platform that delivers unmodified ITA via biodegradable polyesters that form micron-sized particles. Particles enable efficient uptake by innate immune cells and facilitate intracellular metabolite release. This system suppressed inflammatory cytokine production, promotes regulatory immune responses and achieves dose-sparing effects compared to soluble ITA. Overall, these findings expand a platform for delivery of unmodified ITA for modulating both innate and adaptive immunity and enables mechanistic studies and therapeutic investigation.

RESULTS

Itaconate polyesters are hydrolyzable and form MPs

To overcome challenges of exogenous metabolite administration, we generated polyesters with ITA covalently bonded in the backbone to facilitate assembly of particles containing ITA. Polyesters were synthesized through the bulk condensation of dimethyl itaconate (DMI) and either 1,8 octanediol, 1,10 decanediol, or 1,12-dodecanediol using tin(II) ethylhexanoate (Sn(oct)₂) as a catalyst and 4-methoxyphenol (MeHQ) as a radical scavenger at 130°C under nitrogen flow (Fig. 1A). Only 1,12-dodecanediol generated the solid product required for subsequent synthesis of MPs, which eliminated the other materials from further study (Fig. S1A). Due to increased inter- and intramolecular interactions and chain entanglement, higher molecular weight materials are more difficult to hydrolyze than lower molecular weight or oligomeric counterparts [42]. Thus, we hypothesized that the bulk polyesters would hydrolyze to enable controlled release of ITA over time, and that the molecular weight could have a significant impact on release kinetics. To test this, two conditions were used to generate ITA polyesters with low (pITA) and high molecular weights (pITA-XL) (Fig. 1A, Fig. S1B). Polyester synthesis was confirmed via ¹H nuclear magnetic resonance spectroscopy (NMR), evidenced by the presence of ester resonances appearing at approximately 4.1 ppm (Fig. 1B). Resonances identified as methyl ester or hydroxyl end groups were also observed and were more pronounced in the pITA material, suggesting a lower degree of polymerization. Size exclusion chromatography (SEC) revealed the bulk pITA contained primarily oligomer populations, with pITA-XL exhibiting a molecular weight roughly five times higher than pITA (Fig. 1C). To establish reproducibility of polyester synthesis, two additional batches of pITA were generated with similar molecular weights and curve shapes among all three batches, validating the synthetic method (Fig. S1C).

Fig. 1. Itaconate polyesters are hydrolyzable and form MPs.

A) Synthetic route to generate ITA polyesters. B) ¹H NMR spectra with resonances identified in A. C) SEC differential refractive indices to quantify molecular weight and dispersity. D) Release of ITA quantified by LC-MS from bulk pITA over time as a function of pH. E) Release of ITA over time quantified by LC-MS from bulk pITA and pITA-XL at pH 5.6. Error bars represent mean ± SD. SEM images of (F) pITA MPs and (G) pITA-XL MPs. Scale bar = 2μm.

To establish release of ITA over time, we digested bulk pITA and pITA-XL under hydrolyzing conditions at neutral and low pH (5.6), that simulates physiological conditions in the acidic endocytic compartment of cells through which phagocytosed material traffic. Liquid chromatography-mass spectrometry (LC-MS) revealed the digestion of bulk pITA released both ITA and monomethyl itaconate (mmITA), evidenced by two distinct elution profiles (Fig S1D). As expected, a larger amount of ITA was released over time in acidic environments when compared to neutral pH (Fig. 1D). At early time points (≤24h), no significant difference was observed in the release profile of bulk pITA and pITA-XL polyesters digested in a pH of 5.6 (Fig. 1E). However, pITA released over 50ng ITA per mg of bulk material compared with 35ng ITA per mg of pITA-XL after 48h. These data confirm that while both ITA polyesters hydrolyze to release unmodified ITA, metabolite release is affected by molecular weight.

Having confirmed release of unmodified ITA from bulk polyesters, we next generated micron-sized particles to passively target innate cells for uptake. Using a single emulsion synthesis technique, MPs assembled from both pITA and pITA-XL generated uncharged particles measuring approximately 3-4μm in size (Fig. S1E, Supp. Table 1). Importantly, ¹H NMR (Fig. S1F) and SEC (Fig. S1G) confirmed polyesters remained intact after MP synthesis. Despite identical assembly methods and particle size similarity, scanning electron microscopy (SEM) revealed differences in the gross morphology of MPs formed using pITA (Fig. 1F) or pITA-XL (Fig. 1G) bulk polyesters. Taken together, these data indicate that molecular weight of the bulk polyester impacts both ITA release kinetics and MP morphology.

Biocompatible MPs are readily internalized by innate cells

A persistent hurdle to metabolite delivery is the need for transport into the intracellular compartment. We hypothesized that MPs assembled from pITA could overcome this obstacle by taking advantage of phagocytic behavior of innate cells. To evaluate uptake, we treated bone marrow-derived macrophages (BMDMs) and dendritic cells (BMDCs) with MPs labeled with a fluorescent dye (DiO). Consistent with recently published reports [39], a dose-escalation study indicated that no cytotoxicity was observed following treatment with ≤10ug pITA or pITA-XL MP/100,000 cells, compared with untreated cells (Fig. 2A, Fig. S2A–C). The 10ug treatment dose was therefore utilized during in vitro studies moving forward. Because 1,12-dodecanediol independently does not form MPs using a single emulsion synthesis, dose-matched empty PLGA MPs were included as a control for polymer particle treatment. BMDCs readily engulfed fluorescently-labeled MPs (Fig. 2B–C, Fig. S2D); uptake of pITA MPs by both BMDCs and BMDMs occurred in a dose-dependent manner (Fig. S2E). Confocal microscopy further confirmed internalization and the spatial relationship between cargo and cell structures (Fig. 2D). However, uptake of pITA-XL MPs by both BMDCs and BMDMs was reduced compared with pITA MPs (Fig. 2C, Fig. S2F–G). These data reveal that bulk polyester molecular weights impact interactions of pITA MPs with innate cells.

Fig. 2. Biocompatible pITA MPs are engulfed by innate cells.

A) Viability of BMDCs following 16h pre-treatment with 10ug of pITA MPs or empty PLGA MPs as a control. B) Uptake of DiO+ pITA MPs and pITA-XL MPs by BMDCs, 6h after treatment as measured by flow cytometry. C) Quantification of MP uptake by BMDCs, 6h and 16h after treatment. D) Representative maximum projections from confocal microscopy showing BMDC uptake of DiO+ pITA MPs 6h after treatment. Scale bar = 10μm. Fluorescent signals from CellMask, DiO, and merged channels are shown insets. Scale bar = 1μm. E) DiO+ pITA MP engulfment by BMDCs following pre-treatment with uptake inhibitors (Nys: nystatin, Chlor: chlorpromazine HCl, CytoD: cytochalasin D, CytoB: cytochalasin B, 4°C). Error bars represent mean ± SD. *p<0.05, **p<0.01, ***p<0.001, **** p<0.0001 by one-way ANOVA using Tukey-Kramer multiple comparisons test. N=3 per experiment; data is representative of two or more experiments (n=2-3).

To further investigate the mechanisms through which cells engulf pITA MPs, we performed uptake studies in the presence of caveolin- or clathrin-mediated endocytosis inhibitors (nystatin and chlorpromazine HCl, respectively), or inhibitors of phagocytosis (cytochalasin D) and macropinocytosis (cytochalasin B). Cytochalasins, but not nystatin or chlorpromazine, significantly inhibited pITA MP engulfment by BMDCs (Fig. 2E). Uptake inhibitors had no significant effect on cell viability (Fig. S2H). These findings indicate that pITA MPs are primarily internalized through phagocytosis, ensuring the material enters cells through a progressively acidifying intracellular environment that facilitates hydrolysis and intracellular ITA release.

IL-6 secretion by DCs and macrophages is reduced by pITA MP treatment

One pathway through which ITA exerts anti-inflammatory effects is through reduced IL-6 production, a cytokine commonly triggered by innate inflammatory cues [7]. Our data demonstrates that MPs are engulfed by innate cells (Fig. 2B–D), and that unmodified ITA is released in acidic environments over time (Fig. 1D). Thus, we hypothesized that by using MPs to deliver ITA to the intracellular compartment, we could eliminate the need for high dose treatment – a requirement to achieve IL-6 regulation by soluble ITA. First, we tested whether pITA MPs regulates IL-6 secretion by innate cells in a challenging scenario where inflammatory signals are present at the time of treatment. BMDCs and BMDMs were treated with 10ug pITA MPs or pITA-XL MPs administered simultaneously with lipopolysaccharide (LPS) that induces significant IL-6 production. pITA MPs reduced secretion of the inflammatory cytokine by BMDCS, but not BMDMs, compared with untreated cells (Fig. S3A). Importantly, MPs did not stimulate IL-6 production in the absence of LPS stimulation suggesting no inherent immunogenicity (Fig. S3B). Next, we reasoned that pre-treatment of innate cells with MPs would allow time for uptake and intracellular ITA release and therefore could more effectively regulate IL-6 production. BMDCs and BMDMs were pre-treated for 16h with 10ug pITA or 10ug pITA-XL MPs, or soluble ITA. High dose, soluble ITA (5mM) was chosen in concordance with mechanistic work in the literature; low dose, soluble ITA (0.15mM) estimates a dose match to in vitro pITA MP treatment that assumes complete hydrolysis of the polyester backbone for ITA release. Pretreatment with pITA MPs significantly reduced secretion of IL-6 in response to LPS stimulation by BMDCs (Fig. 3A) and BMDMs – albeit to a lesser extent (Fig. S3C). Importantly, no effect on IL-6 was observed in cells treated with soluble, low-dose ITA controls. These data support our hypothesis that delivery of ITA via MPs overcomes the need for high dose treatments.

Fig. 3. MP-enabled ITA delivery decreases IL-6 secretion and inflammatory T cell polarization.

A) BMDCs pre-treated for 16h with 10ug pITA MPs, pITA-XL MPs, or soluble ITA, followed by LPS (100ng/mL) stimulation for 24h before supernatants were collected. IL-6 secretion was measured by ELISAs. B) IL-6 secretion by BMDCs pre-treated with 10ug pITA-XL MPs for 16h or 36h before LPS stimulation. Supernatants were collected 24h after LPS administration. C) IL-6 secretion by BMDCs left untreated, or pretreated with uptake inhibitors (Chlor: chlorpromazine HCl, CytoD: cytochalasin D, CytoB: cytochalasin B). D) Representative flow cytometry plots illustrating IL-17a+ (T_H17) and Foxp3+ (T_reg) among CD4+ T cells polarized under inflammatory conditions in the presence of APCs left untreated (top), or following pre-treatment with 10ug pITA MPs (bottom). E) Quantification of the frequency of IL-17a+ CD4 T cells polarized in the presence of indicated treatments. F) Quantification of IL-17a+ T cells with and without the addition of exogenous IL-6 to APC:T cell co-cultures. G) IL-17a production by CD4 T cells stimulated with aCD3+aCD28 under T_H17 polarizing conditions, in the absence of APCs. H) Foxp3 expression in CD4 T cells during T_H17 (left bars) and T_H0 (right bars) polarization conditions in APC:T cell co-cultures. N=3-6 per experiment; data is representative of two or more experiments (n=2-3). Error bars represent mean ± SD. *p<0.05, **p<0.01, ***p<0.001, **** p<0.0001 by Student’s t test or one-way ANOVA using Tukey-Kramer multiple comparisons test.

Unlike pITA MPs, pre-treatment of BMDCs or BMDMs with pITA-XL MPs did not significantly reduce IL-6 secretion (Fig. 3A, Fig. S3C). To test whether this could be due to slower kinetics of pITA-XL MP uptake (Fig. 2C), we treated BMDCs with pITA-XL MPs for a total of 36h before LPS stimulation. However, longer pre-treatment had no additional effect (Fig. 3B), suggesting that MPs must be efficiently engulfed to reduce IL-6 secretion. To further determine if effects on IL-6 secretion by pITA MPs were dependent on internalization, we again treated BMDCs with uptake inhibitors before pITA MP administration and LPS stimulation. Inhibition of phagocytosis by cytochalasins (e.g., CytoD, CytoB) partially rescued IL-6 secretion (Fig. 3C). Taken together, these data suggest that regulation of IL-6 by pITA MPs is dependent on delivery of the metabolite to the intracellular compartment of cells.

Decreased IL-6 secretion skews CD4 T cell polarization away from inflammatory and toward regulatory subsets

IL-6 plays a critical role in promoting inflammatory T cell (T_H17) polarization that mediates attack of tissues in several autoimmune diseases, notably MS/EAE [43]. Given pITA MPs reduced IL-6 production by innate cells, we hypothesized that treatment could further alter adaptive immunity by skewing T cell polarization away from inflammatory phenotypes. First, antigen presenting cells (APCs) were treated for 16h with pITA MPs or soluble ITA controls, followed by LPS stimulation as above that drives IL-6 production. 24h later, naïve CD4 T cells were added to APC cultures alongside exogenous cues to support T_H17 polarization (TGFβ, aIL-12, aIL-4). To examine the relationship between pITA MP-mediated inhibition of IL-6 secretion by APCs and inflammatory T cell polarization, no additional exogenous IL-6 was added to co cultures. Thus, the only source of IL-6, which is required for T_H17 polarization, were APCs. pITA MPs and high dose soluble ITA similarly and significantly reduced IL-17a production among CD4 T cells, while low dose soluble ITA had no effect (Fig. 3D–E Fig. S4). To directly test whether reduced T_H17 polarization was due to reduced IL-6 secretion by APCs pre-treated with pITA MPs, we added exogenous IL-6 back to co-cultures and assessed whether T_H17 polarization could be rescued. Indeed, exogenous IL-6 restored IL-17a production by CD4 T cells cultured with APCs treated with pITA MPs (Fig. 3F), directly linking modulation of innate cell function by pITA MPs to adaptive immune responses. To further rule out that pITA MPs exert effects on T cells independent of APC functions, we treated naïve CD4 T cells with pITA MPs in the absence of APCs under T_H17 polarizing conditions (i.e., IL-6, TGFβ, aIL-12, aIL-4). Under these conditions, pITA MPs also did not significantly inhibit IL-17a production (Fig. 3G). Together, these data support our hypothesis that effects of pITA MPs on T_H17 polarization are at least partially due to altered IL-6 production by innate cells.

Lastly, we examined the effect of pITA MPs on regulatory T cells (T_reg) that play a critical role in controlling self-reactive, inflammatory T cells during autoimmunity [16]. Excitingly, pITA MPs also increased the frequency of Foxp3+ T cells (T_reg) despite T_H17 polarizing conditions, and in non-polarizing (T_H0) cultures (Fig. 3H). Thus, in addition to modulating innate cell signaling, pITA MPs exhibit the capacity to regulate inflammatory T cell responses.

Circulating serum IL-6 is reduced by pITA MP treatment of mice during EAE

Having shown pITA MPs control IL-6 secretion by innate cells in vitro (Fig. 3A), we next asked whether MP-based delivery of ITA could regulate IL-6 in an in vivo setting. First, we quantified biodistribution and pITA MP uptake in mice to test whether APCs engage with MPs in vivo. pITA MPs labeled with fluorescent dye (DiD) were injected subcutaneously (s.c.). 24h after treatment, particle uptake and distribution in tissues was assessed via flow cytometry (Fig. S5A) and the In Vivo Imaging System (IVIS). Flow cytometry confirmed that particle-derived fluorescence was associated with macrophages (F4/80+) and dendritic cells (CD11c+) (Fig. 4A–B, Fig. S5B–C) – but not T cells that are non-phagocytic – in the cervical (draining) lymph nodes (LNs) and spleen (Fig. S5C–D). Overt off-target delivery to kidneys or the liver was not observed, although particles were minimally detected via IVIS (Fig. S5E–F).

Fig. 4. Circulating serum IL-6 is reduced by pITA MP treatment in EAE.

A) Representative flow cytometry plots of particle uptake by macrophages in cervical LNs. B) Quantification of DiD+ macrophages (left) and DCs (right) in cervical LNs, 24h after s.c. treatment with 2mg DiD+ pITA MPs. C) Timeline for treatment and analysis of EAE mice. Tx = treatment. D) Circulating serum IL-6 in EAE mice over time, measured by ELISA. E) Serum IL-6 in mice on day 6 post EAE induction. Samples were pooled for analysis (N=5/group). Error bars represent mean ± SD. F) Clinical score in untreated controls and EAE mice treated with pITA MPs (s.c.) or soluble ITA (i.p.). (n=8 mice/group). Error bars represent mean ± SEM. Data is representative of two experiments (n=2). *p<0.05, **p<0.01 by one-way ANOVA using Tukey–Kramer multiple comparisons test or Steel–Dwass test (clinical scores).

To test whether pITA MPs regulate systemic IL-6 in the context of an inflammatory disease, we utilized an autoimmune model (EAE) as a test bed to assess MP-mediated regulation of the inflammatory cytokine [17, 44]. Mice were treated s.c. on the day of disease induction with 2mg pITA MPs, followed by three re-treatments administered every five days until peak disease (day 15) (Fig. 4C). This dosing schedule was chosen with the goal of establishing a MP depot that would release ITA over time, informed by our data showing that unmodified ITA underwent controlled release over a period of 48h (Fig. 1D). A second cohort of EAE mice received 2mg soluble ITA via intraperitoneal (i.p.) injection every other day as a high dose benchmark control, consistent with reports in the literature [23]. Examination of circulating IL-6 over time revealed that pITA MPs reduced serum IL-6 compared with untreated EAE mice during disease onset (Fig. 4D), when IL-6 has been shown to be significantly elevated in this model [45]. Notably, the reduction occurred after just two s.c. treatments and was statistically indistinguishable from mice treated i.p. with high dose, soluble ITA (Fig. 4E). Although the specific MP dosing regimen tested here only nominally impacted paralysis (measured by clinical score) (Fig. 4F, Fig. S6A–D), the biomaterials-enabled delivery of ITA facilitated by polymeric MPs exhibited dose-sparing effects and modulated systemic IL-6, a key driver of pathology in MS and other autoimmune diseases.

pITA MPs regulate IL-1β secretion by innate cells

In addition to regulating IL-6, ITA has been shown to exert control over other inflammatory pathways with potential relevance in autoimmune and chronic inflammatory diseases [46]. One such pathway is inflammasome activation that results in maturation of IL-1β [12, 13], a pro-inflammatory cytokine with pleiotropic effects [47]. Excessive IL-1β has been reported in patients with rheumatoid arthritis [48] and systemic lupus erythematosus [49, 50]. Therefore, we next investigated whether pITA MP treatment also regulates IL-1β. To test this, BMDMs and BMDCs were pre-treated with 10ug pITA MPs or soluble controls and stimulated using adenosine triphosphate (ATP) and LPS that that trigger inflammasome activation. pITA MPs and high dose soluble ITA similarly decreased IL-1β secretion compared with untreated cells (Fig. 5A. Fig. S7). However, low dose, soluble ITA had no effect. These findings parallel results from IL-6 studies supporting our hypothesis that release of ITA from MPs directly to the intracellular compartment overcomes the need for high dose, soluble treatments to regulate diverse immunological outcomes.

Fig. 5. Itaconate polyester delivery reduced IL-1β secretion and differentially regulate IFNβ production by innate cells.

A) BMDMs pre-treated for 16h under indicated conditions, followed by LPS+ATP stimulation for 24h before supernatants were collected for analysis of IL-1β secretion, measured via ELISA. B) Pre-treated BMDMs stimulated with poly(I:C) for 24h. IFNβ secretion was assayed via ELISA. C) Quantification of Gclm, Nqo1, and Hmox1 gene expression analysis in BMDMs following 16h pre-treatment and 24h macrophage polarization, expressed as relative to stimulated, untreated cells (LPS only). D) IFNβ secretion by BMDCs pretreated for 16h as indicated, followed by poly(I:C) stimulation for 24h. N=3-4, data are representative of two or more experiments (n=2-3). Treatments = 10ug pAA or pITA MPs, 0.25mM 4-OI or DMI. Error bars represent mean ± SD. *p<0.05, **p<0.01, ***p<0/001 by one-way ANOVA using Tukey-Kramer multiple comparisons test.

Metabolite release from pITA MPs mimics effects of high dose, soluble ITA

Studies exploring the anti-inflammatory effects of ITA on macrophages in pre-clinical applications have relied heavily on the use of derivatized ITA (i.e., 4-OI, DMI) that are highly membrane permeable. However, it is now widely appreciated that 4-OI and DMI exhibit differing effects [29, 51]. Notably, type I interferons (e.g., IFNβ) that are important for antiviral immunity are inhibited by derivatives, but not ITA [52, 53]. These findings further motivate development of strategies that permit delivery of the unmodified metabolite that more closely mimics endogenous ITA. To test the hypothesis that intracellular release of ITA via MPs recapitulates effects of unmodified, soluble ITA on type I interferons, we administered synthetic, double-stranded RNA (i.e., polyinosinic:polycytidylic acid (poly(I:C)) that triggers IFNβ secretion by innate cells. As an additional control for particle uptake, we pre-treated cells with MPs synthesized from adipic acid polyesters (pAA) that are similar in size and morphology to pITA MPs (Fig. S8A–C). 4-OI and DMI both strongly inhibited IFNβ production by BMDMs, while pAA MPs and low dose, soluble ITA had no effect. High dose soluble ITA also did not affect IFNβ, as expected. pITA MPs reduced IFNβ secretion, albeit not significantly when compared with soluble ITA (Fig. 5B). Thus, in BMDMs, pITA MPs elicit an intermediate outcome on IFNβ secretion compared with high dose, soluble ITA and its derivatives.

Next, we investigated the possible mechanisms through which pITA MPs inhibit IFNβ. Although the pathways controlling type I IFNs are complex [54, 55], electrophilic molecules have been shown to act as potent modulators of IFNβ [7, 29, 56, 57]. Thus, we hypothesized that pITA MPs induce a weakly electrophilic response that correlates with its immunoregulatory activity of IFNβ. To test this, we examined transcriptional markers of electrophilic stress. Consistent with their magnitude of electrophilicity (DMI > 4-OI > ITA), 4-OI and DMI significantly induced electrophilic stress response genes – Gclm, Nqo1, and Hmox1 – relative to untreated BMDMs (Fig. 5C, Fig. S9A–C). pITA MPs (10ug) drove some expression of these genes compared with high dose, soluble ITA, albeit to a lesser degree compared with derivatives. While pAA MPs also induced limited expression of Hmox1, release of adipic acid is not expected to exert effects due to the absence of the α,β unsaturation in the backbone [57]. Taken together, these data suggest that while MP-based delivery of ITA approximates effects of high dose, soluble ITA, pITA MP treatment induced some electrophilic stress in BMDMs associated with regulation of type I interferons.

Intriguingly, treatment of BMDCs with derivatives, high dose, soluble ITA, and 10ug pITA MPs all significantly reduced IFNβ secretion. (Fig. 5D). These findings are in contrast with effects of ITA and pITA MPs on BMDMs, where IFNβ was less sensitive to the polyester treatment (Fig. 5B). These data reveal that release of ITA from pITA MPs exerts complex, nuanced effects on inflammatory signaling pathways and potentially different regulatory functions across innate immune cell types. Nonetheless, IFNβ secretion by BMDCs was unaffected by pAA MP or low-dose, soluble ITA controls, emphasizing that pITA MPs enable metabolite delivery and regulation of immunomodulatory outcomes.

DISCUSSION

ITA is an immunoregulatory metabolite with significant therapeutic promise [7, 23, 24, 58], but its potential for clinical translation is limited by poor cellular uptake and rapid clearance. Here, we use an MP assembled entirely from ITA polyesters that delivers unmodified ITA intracellularly to control inflammation and promote regulatory immune responses, highlighting a promising strategy for ITA-based immunotherapies.

Prior studies [35–39] have demonstrated the promise of biomaterials-based metabolite delivery, primarily in macrophage-focused models; these works establish feasibility in several contexts. Here, we couple tunable materials design with expanded mechanistic interrogation to advance this biomaterials-based approach for ITA delivery both in vitro and in vivo. We investigate features of ITA polyesters and show molecular weight shapes release kinetics and cellular uptake of MPs that influence immunoregulatory outcomes – an important consideration for materials design. Furthermore, we show pITA MPs regulate inflammasome activation and directly compare effects of 4-OI and DMI (ITA derivatives), with soluble, underivatized ITA and polyester-derived ITA. This comparison, not previously examined in the context of IFNβ, is particularly important given its sensitivity to electrophilic compounds and suggest that an aliphatic, polyester-based approach to delivery of ITA recapitulates some, but not all effects, of soluble, unmodified ITA. Differential regulation of inflammatory signaling was also observed between DCs and macrophages, which has not yet been investigated in the context of this novel class of biomaterials. Together, these findings show that the immunological effects of ITA polyesters are potentially shaped by delivery modality, cellular target, and immune context, and advance the goal of metabolite delivery in therapeutic settings.

While recent reports indicate that exogenously administered ITA can indeed accumulate inside of cells [29, 59], high dose treatment is required in both cell culture and in vivo. To address delivery limitations, we synthesized ITA polyesters to generate “carrier-free” particles, eliminating the need for encapsulation approaches (e.g., poly(lactic-co-glycolic acid)-based materials). At this time, we and others have not directly quantified unmodified ITA delivered to the intracellular compartment by this biomaterials-based strategy to metabolite delivery – a limitation of this work. Instead, we directly compared 10ug pITA MP treatments to two soluble ITA doses (5mM and 0.15mM). The high dose (5mM) was selected in based on prior mechanistic studies [10, 13, 29], while the low dose ITA (0.15mM) serves as a “theoretical dose-match” to pITA MP treatment. This estimate assumes complete hydrolysis of 10ug of pITA MPs, which would release a maximum of 3.9μg of ITA (=0.15mM in the well). Importantly however, our LC-MS data indicate that ITA polyesters release 40ng ITA/mg of bulk material after 24h at a pH representative of the endocytic compartment, suggesting that release from pITA MPs may be far more limited in physiological conditions. Assuming particle release kinetics are similar to bulk material, this corresponds to 0.4ng of ITA released making the concentration in the well ≤20nM. This dose is over three orders of magnitude lower than the low-dose soluble treatment and still exerts immunomodulatory effects, consistent with a dose-sparing effect afforded by this biomaterials-based approach. Therefore, the comparable absence of immune modulation by low-dose, soluble treatment used as a surrogate dose-match strongly supports the delivery of ITA to the intracellular compartment through engulfment and hydrolysis, which were examined.

Tunable features of biomaterials, such as particle size, greatly influence cellular uptake and have been investigated in the context of metabolite delivery [38, 39]. Given polymer properties offer a unique opportunity to customize polyester-based biodegradable materials [42], we investigated the effects of low (pITA) and high (pITA-XL) molecular weight on polyester degradation and MP production. To our knowledge, this has not yet been explored for a biomaterials-based approach to ITA delivery. Interestingly, degradation of bulk pITA-XL differed from pITA. We hypothesized that the larger, more entangled polymer chains of pITA-XL helped reduced the rate of ITA release over time, and therefore could offer a longer therapeutic window. However, DCs and macrophages more readily engulfed MPs assembled from low molecular pITA, despite similar size and charge of the particles. Whether this is due to MP topography [60], polyester molecular weight, or both is yet to be determined. Nonetheless, these findings highlight the importance of considering polyester properties that could be further tuned for particle assembly [61].

It is now widely appreciated that unmodified ITA and its derivatives exhibit divergent mechanisms of action, described during regulation of type I interferons [6, 52, 53, 62, 63]. These findings further motivate our aim to advance strategies for intracellular delivery of unmodified ITA that could impart a greater degree of selectivity on immune regulation – a persistent goal of novel therapeutics for treatment of autoimmune diseases. In support of this, pITA MP treatment approximated the effects of high-dose, soluble ITA on IFNβ secretion by BMDMs – consistent with hydrolysis of pITA that releases unmodified itaconic acid. While monomethyl itaconate was also released from pITA MPs, quantification was challenging due to unresolved elution of 1-methyl itaconate, 4-methyl itaconate, and possible mesaconate isomers. The relative magnitude of electrophilicity of monomethyl ITA is less than DMI and 4-OI, but release of these isomers could explain the intermediate effect of pITA MP treatment on IFNβ production and expression of electrophilic response-related genes [4, 64]. Importantly, our data also show that in contrast to BMDMs, pITA MPs significantly inhibited IFNβ secretion by BMDCs. This may reflect a divergent role for ITA-mediated immunoregulation in DCs vs. macrophages. Thus, it will be of utmost importance to expand investigation of effects of pITA MPs across culture conditions and inflammatory cell types from both mouse and human sources. Lastly, it will also be of interest to examine whether pITA MPs differentially regulate cytokines with bona fide anti-inflammatory activity, such as IL-10 or IL-4, produced by innate and/or innate lymphoid cells.

Soluble ITA treatment has been shown to control inflammation in models of autoimmunity [7, 23, 27], infection [53, 65] sepsis [58], atherosclerosis [66], and reperfusion injury [26]. Consistent with these findings, studies monitoring ITA in circulation of mice document rapid catabolism and clearance in vivo [67]. Acute applications aside, efficacy consistently required high-dose, daily (or frequent) administration that may not be feasible in all clinical settings. Our in vitro data comparing pITA MP treatment with low-dose, soluble ITA suggest that dose-sparing effects, and therefore fewer treatments, might be achieved by a particle-based approach to metabolite administration. In support of this, we show that pITA MP administration successfully reduced circulating IL-6 early during disease onset, consistent with kinetics of the serum cytokine in this model [45]. To our knowledge, this specific in vivo modulation by ITA-based materials has not yet been reported. Early reduction of IL-6 was achieved by two subcutaneous (s.c.) pITA MP treatments compared with three intraperitoneal (i.p.) soluble injections. Our LC-MS data suggest that the maximum amount of ITA released from in vivo MP treatments would be 100ng per injection over a 48h period. On the other hand, a single dose of 2mg soluble ITA could be orders of magnitude higher (up to 2,000 times), even when conservatively estimating that only 10% of soluble ITA crosses the plasma membrane of relevant cells. Although this initial report did not optimize pharmacokinetic profiles for efficacy during EAE, this result is likely dose-related – supported by recent data showing that ITA-MP-mediated regulation of BMDM phenotypes was dependent on the degree of particle uptake [39]. Nonetheless, this is the first report of pITA MP-based treatment in vivo, highlighting the potential for biomaterials to deliver unmodified ITA via controlled release to the intracellular compartment of innate cells to regulate immunity. This strategy could change the treatment landscape of metabolite-based therapies for chronic, inflammatory autoimmune diseases – a revolution that has already been validated in cancer therapeutics (i.e., doxorubicin, paclitaxel).

CONCLUSION

Biomaterials assembled from hydrolyzable ITA polyesters provide a biocompatible, scalable approach to ITA-based therapy. Through uptake and controlled release, pITA MPs suppress inflammatory cytokine production by innate cells that in turn alter adaptive immune responses. This strategy overcomes challenges of systemic soluble ITA delivery, including rapid degradation and clearance. In addition, we found immunoregulatory effects closely mimicked endogenous ITA compared with derivatives, making this a viable approach to preservation of ITA’s intrinsic immunoregulatory activity. Together, these findings advance a platform for intracellular delivery of unmodified ITA that enables mechanistic interrogation and supports further investigation as a therapeutic strategy.

METHODS

Animals.

6- to 12-week-old C57BL/6J female and male mice were used in all experiments. Mice were purchased from The Jackson Laboratory. All animal studies were fully compliant with local, state, and federal guidelines per the Association for Assessment and Accreditation of Laboratory Animal Care expectations for animal care and use/ethics. Studies were approved and carried out under the supervision of the University of Maryland Institutional Animal Care and Use Committee in compliance with all local, state, and federal guidelines.

Synthesis of Polymeric Metabolites.

All polymerizations were carried out under nitrogen flow. Methanol (Pharmco, ACS grade) chloroform (Fisher, ACS grade), tetrahydrofuran (Fisher, HPLC) and dichloromethane (Sigma Aldrich, ACS grade) were used as received. Chloroform-d₁ was purchased from Cambridge Isotopes and used as received. Itaconic acid (≥99%), (98%), tin (II) ethylhexanoate (92.5-100%), 4-methoxyphenol (99%), and dimethyl adipate (≥99%) were purchased from Sigma Aldrich and used as received. 1,12-dodecanediol (>99%) was purchased from TCI Chemicals and used as received. Dimethyl itaconate (DMI) was purchased and used as received (Sigma Aldrich, 99%), or synthesized and distilled. To synthesize dimethyl itaconate (DMI), 160mL of methanol, itaconic acid (0.40 mol, 52.04 g), sulfuric acid (0.32 mol, 11.77 g) was added to a 500mL round bottom flask. The reaction was stirred and refluxed at 80°C for 16 hours. Approximately half of the methanol was removed under reduced pressure, and the reaction quenched by slow addition of a saturated solution of sodium carbonate. When fully quenched (pH 7), the product was extracted with ethyl acetate. The excess solvent was removed under reduced pressure, and the product isolated by distillation (70°C, 2 mmHg) as a colorless liquid (41.90 g, 66%).

To an appropriately sized round bottom flask, dimethyl ester and 1,12-dodecanediol were added in a 1:1 molar ratio. The catalyst tin(II) ethylhexanoate (Sn(oct)₂) was added at 20 mmol per mol of ester. The radical scavenger 4-methoxyphenol (MeHQ) was added at 12.2 mmol per mol of ester. A Dean Stark trap was utilized to facilitate removal of condensed methanol from the reaction flask. The reaction was stirred at 130°C and quenched through precipitation into ice cold methanol, stirring rapidly. A white solid precipitate resulted, which was allowed to settle for a minimum of one hour. Methanol was decanted or removed by pipette, and the crude product dried under reduced pressure. The product was redissolved in minimal chloroform, and reprecipitated in ice cold methanol to remove residual starting materials. Excess methanol was decanted or removed by pipette, and the final product was dried in vacuo at room temperature for a minimum of 16 hours. For synthesis of poly(dodecyl itaconate) “pITA,”dimethyl itaconate (31.63 mmol, 5.00 g) and 1,12-dodecanediol (31.63 mmol, 6.40 g) were added to a flask containing Sn(oct)₂ (0.63 mmol, 256 mg) and MeHQ (0.39 mmol, 48 mg). The reaction was stirred at 130°C for 18 hours. The dried purified product was a white waxy solid (8.26 g, 88%). For synthesis of poly(dodecyl itaconate) – high molecular weight “pITA-XL” – dimethyl itaconate (31.63 mmol, 5.00 g) and 1,12-dodecanediol (31.63 mmol, 6.40 g) were added to a flask containing Sn(oct)₂ (0.63 mmol, 256 mg) and MeHQ (0.39 mmol, 48 mg). The reaction was stirred at 130°C for 68 hours. The pressure was then reduced to 30mmHg for 6 hours. After double precipitation in ice cold methanol, the dried purified product was a white solid (9.12 g, 97%). For synthesis of poly(dodecyl adipate), dimethyl adipate (31.63 mmol, 5.51 g) and 1,12-dodecanediol (31.63 mmol, 6.40 g) were added to a flask containing Sn(oct)₂ (0.63 mmol, 256 mg) and MeHQ (0.39 mmol, 48 mg). The reaction was stirred at 130°C for 18 hours. The dried purified product was a white solid (9.40 g, 95%).

Nuclear Magnetic Resonance (NMR).

¹H NMR spectroscopy was performed on a Bruker AV III 400 MHz spectrometer. Spectra were referenced to residual solvent peaks (¹H: 7.26 ppm for chloroform-d1). Samples were dissolved in chloroform-d1 at approximately 20 mg/mL.

Size Exclusion Chromatography (SEC).

SEC was performed on a Viscotek GPCMax equipped with two columns (Agilent PLgel 5μm) in a column oven and differential refractometer (Viscotek TDA 302) maintained at 40°C. Tetrahydrofuran (HPLC grade) was used as the eluent with a flow rate of 1 mL/min. Polystyrene standards (from Agilent Technologies, 1.21 – 215 kDa) were used for calibration. For SEC sample preparation, approximately 8 mg of material was dissolved in 2 mL of THF.

Liquid Chromatography-Mass Spectrometry (LC-MS).

LC-MS was performed on a Bruker Maxis-II Q-TOF mass spectrometer equipped with an ESI ion source coupled with a Waters Acquity I-Class Plus LC system. A Waters Atlantis Premier BEH C18 AX column was used for LC separation. For the elution, the mobile phase composed of 0.1% aqueous formic acid solution (A) and 0.1% formic acid in acetonitrile (B) and was used at a mixture of 2% A and 98% B at a flow rate of 0.250 mL/min to being. After 11 minutes, the mobile phase was changed to 90% A and 10% B for 5 minutes, before returning to the initial mixture ratio for the remainder of the test. The Maxis-II settings were as follows: endplate offset voltage = 500 V, capillary voltage = 4500 V, dry temperature = 220 °C, dry gas flow rate = 5.0 L/min, nebulizer gas pressure = 1.0 bar. The mass range was from 80 to 400 Da and the measurement was performed in the negative mode.

An aqueous solution of pH 5.6 was synthesized using formic acid, and an aqueous solution of ammonia acetate was used to create a weak buffer and maintain a pH of 7, as MilliQ water is known to acidify over time when exposed to carbon dioxide in air. Samples of bulk polyester were suspended at approximately 100mg/mL in pH 5.6 or pH 7 solutions for either 18, 24, or 48 hours. After the determined amount of time, the samples were centrifuged at 5000 g for 5 minutes, and the supernatant removed. The supernatant was filtered through a 0.22 μm PVDF filter prior to analysis. Standards of itaconic acid between 0.25 and 10 ppm were created for the calibration curves. Itaconic acid eluted at 3 minutes with an m/z of 129.00, while monomethyl itaconate eluted at 1.2 minutes with an m/z of 143.02.

Microparticle Assembly.

Particles were assembled using a single emulsion, oil-in-water (o/w) method. Polyesters were dissolved in 1 mL of dichlormethane (DCM) at a concentration of 50 mg/mL and homogenized in 10mL of 2% (w/v) polyvinyl alcohol (PVA) (Millipore Sigma) solution for 3 minutes at ≥16,000 rpm (IKA T25 Digital Ultra Turrax). The primary emulsion was added to 40 mL of 1% (w/v) polyvinyl alcohol solution as a stabilizer. After overnight stirring to evaporate remaining organic solvent, the particle/PVA mixture was filtered through a 40 μm mesh. The filtered mixture was centrifuged at 5000 g for 5 minutes, PVA was decanted, and the particles were resuspended in MilliQ water (18.2 MΩ). This process was repeated twice to wash the MPs. Particles were resuspended in a known volume of MilliQ water and 10% of that volume was removed and lyophilized to determine yield. Particles were stored in either their stock solution, at 2mg/mL in PBS at 4°C, or lyophilized. To fluorescently label particles, carbocyanine dyes (e.g., DiO, DiD) (ThermoFisher Scientific) were added to the initial dichloromethane solution prior to homogenization, and the same procedure was followed.

Particle Sizing (Laser Diffraction/Scattering).

MPs were suspended in MilliQ water (18.2 MΩ) at a concentration to reach the threshold of detection by the Horiba Partica LA-950V2 Laser Scattering Particle Size Analyzer. Light sources were one red solid state 5 mW laser diode (650 nm) and one blue solid state 3 mW LED (405 nm)

Zeta Potential.

MPs were suspended in 0.1X phosphate buffered saline diluted with MilliQ water (18.2 MΩ) at a concentration of approximately 0.5 mg/mL. Zeta potential was measured using a Brookhaven Nanobrook Omni Particle Sizer and Zeta Potential Analyzer at room temperature, with the charge reported as the average of 30 measurements.

Scanning Electron Microscopy (SEM).

SEM was performed on a Hitachi SU-70 Schottky field emission gun (FEG) SEM. To prepare samples, MPs were dispersed in water and dropped onto clean silicon wafers and allowed to air dry at room temperature. The silicon wafers with dried particles were sputter coated with 3 nm gold/palladium prior to imaging (Safematic Compact Coating Unit-010). Images were captured in secondary electron mode at an accelerating voltage of 3kV.

Cell Culture.

To culture BMDCs and BMDMs, femurs were harvested from 8-12-week-old C57BL/6J mice. Bone marrow was flushed using hydraulic extrusion in cold PBS and filtered through a 70um strainer. Red blood cells were lysed in ACK lysis buffer (ThermoFisher Scientific) and cell suspensions were washed before counting. 1x10⁶ cells/mL were resuspended in complete RPMI 1640 (RPMI-1640 supplemented with 10% fetal bovine serum (FBS), 2mM L-glutamine, 1mM sodium pyruvate, 0.1mM non-essential amino acids, 55μM β-mercaptoethanol, 100U Pen/Strep,) and cultured in 10mL/100mm dish in the presence of 20ng/mL of granulocyte macrophage-colony stimulating factor (GM-CSF) (BioLegend) for differentiation of BMDCs or 40ng/mL of macrophage-colony stimulating factor (M-CSF) (BioLegend) for BMDMs. On day 3, an equal volume of freshly supplemented complete RPMI was added to dishes. On day 6, a 1:1 media exchange with freshly supplemented complete RPMI was performed. Differentiated BMDCs and BMDMs were used on day 7-9 after start of culture. For assays, 100,000 BMDC or BMDMs were seeded in 96-well flat bottom tissue culture plates for 1.5hrs, followed by treatment with 10ug (unless otherwise indicated) MPs for 6-16hrs before 100ng/mL LPS (for IL-6 secretion), 100ng/mL LPS (3hr) followed by 5mM ATP (for IL-1β secretion), or 20ug/mL high molecular weight poly(I:C) (Invivogen) (for IFNβ secretion). 5mM and 0.15mM soluble ITA solutions were prepared by dilution from a 615mM ITA standard in MilliQ water (18.2 MΩ). The 0.15mM concentration was selected as a “dose matched” soluble equivalent to 10μg of pITA MPs. Assuming a 1:1 molar monomer ratio, ITA constitutes 39% of each pITA repeat unit. This corresponds to a maximum of 3.9μg ITA in 10μg MP dose that equates to 150μM per well. In uptake inhibition studies, cells were pretreated for 15-30min with the following at indicated concentrations: cytochalasin B (2.5ug/mL), cytochalasin D (2.5ug/mL), chlorpromazine HCl (5ug/mL), nystatin (10ug/mL).

Microscopy.

Glass bottom chamber slides (Ibidi) were coated with 0.01 % w/v poly-L-lysine solution (Millipore Sigma) for 30 min before drying overnight. 400,000 BMDCs were plated per well in 400μL of culture media with 1 μg/mL LPS. Imaging slide was cultured for 6hrs with pITA MP treatments at 37 °C followed by a gentle PBS wash. To stain cell membranes, cells were flooded with a 2000X dilution of CellMask Plasma Membrane Stain (ThermoFisher Scientific) for 5 min at 37 °C, followed by a wash. Samples were fixed for 5min in 2 % paraformaldehyde (Electron Microscopy Sciences) at 37 °C, washed three times, and stained with 300 nM DAPI stain for 5 min at 37°C before final resuspension in PBS. Confocal images were collected using a Zeiss LSM 980 confocal system fitted with an Airyscan 2 module, mounted on an inverted Zeiss Axio Observed Z1 microscope with an oil immersion Plan-Apochromat 63x/1.40 differential interferences contrast (DIC) objective lens (Carl Zeiss, Inc.). Images were acquired in Airyscan super-resolution (SR) mode using a 32-channel GaAsP-PMT Airyscan detector. Zeiss ZEN 2.3 (black) software was used for collection and post processing of the images. Airyscan processing was done using an auto-strength 3D method. Post-processing of images was performed in ImageJ 1.53c. Fluorescence intensity ranges were adjusted using linear scaling and for display purposes only.

Enzyme-linked Immunosorbent Assay (ELISA).

For analysis of cytokine secretion, cell culture supernatants were harvested 12-36hrs after treatments and stored at −80°C until ready for use. ELISAs were performed according to the manufacturer’s instructions (BioLegend, Deluxe Set Mouse). Samples were diluted 1:2 (IFNβ), 1:50 (IL-1β), or 1:500 (IL-6). Serum samples were undiluted. For analysis of serum cytokines, serum was pooled from up to 2 mice per sample. Serum from the same animals were pooled for each time point in EAE studies. Absorbance was read on a Tecan Spark Multimode Microplate Reader at 450nm, with background subtraction of absorbance at 570nm. Standards and samples were run in triplicate, unless indicated. A standard curve for sample quantification was generated by plotting mean absorbance against known standard concentrations.

Flow Cytometry.

For viability testing, BMDCs or BMDMs were harvested using cold 2mM EDTA in PBS 24hrs after LPS stimulation. Cells were stained with LiveDead Near-IR Viability Dye (ThermoFisher Scientific) in buffer (1% FBS in PBS), washed twice and fixed with 2% paraformaldehyde in buffer before analysis. For uptake assays, live, DiO+ or DiD+ cells were quantified at 6h or 16h post MP treatment. For intracellular staining, cells were stained with viability dye as described and surface markers at 4C for 20min in buffer, followed by washing and fixation for 30 min at 4°C (Foxp3 Fix/Perm kit, eBiosciences). Intracellular stains were performed in Perm/Wash for 40 minutes at 4°C. Cells were washed twice and resuspended in buffer before analysis. Data were acquired on a Beckman Coulter CytoFlex and analyzed using FlowJo software (BD Biosciences).

T Cell Polarization.

To prepare APCs, spleens from C57BL6/J mice were mechanically disrupted through a 70μM filter. Red bloods cells were lysed using ACK lysis buffer (ThermoFisher Scientific). To deplete T cells from splenocyte suspensions, cells were stained with anti-CD4-FITC (clone RM4-5) and anti-CD8-FITC (clone 53-6.7) antibodies for 15 minutes at 4°C in MACS buffer (1mM EDTA, 0.05%BSA in PBS), followed by depletion with anti-FITC magnetic microbeads according to manufacturer’s instructions (Miltenyi, anti-FITC Microbeads). Cell suspensions were then loaded onto a magnetic column and flow-through collected as T-cell depleted, untouched APCs. APCs were cultured at 5 million/mL and pre-treated as indicated. 12-16 hours later, naïve CD4+ T cells were isolated from spleens of C57BL6/J mice using naïve isolation kits (Miltenyi), according to manufacturer’s instructions. T cells (1mil/mL), anti-CD3 (1ug/mL) and anti-CD28 (3ug/mL) were added to APC cultures, with cytokines and blocking antibodies dependent on polarizing conditions (T_H0: anti-IL4, anti-IL6, anti-IL-12; T_H17: anti-IL4, anti-IL-12, TGFβ; T_reg: anti-IL4, anti-IL-12, IL-2, TGFβ). Unless otherwise noted, no recombinant IL-6 was added to TH17 polarization conditions. After four days of co-culture, cells were restimulated using phorbol myristate acetate (PMA)/ionomycin (50ng/mL/1ug/mL) for four hours before staining for analysis using flow cytometry.

Real-time Quantitative Polymerase Chain Reaction (RT-qPCR).

RNA was isolated using the Quick-RNA Microprep Kit (Zymo Research) from BMDMs and lysed in RNA lysis buffer (Zymo Research), per the manufacturer’s instructions including DNA degradation with DNase I. RNA was diluted to 20ng/μL in RT-qPCR grade water (Thermo Fisher Scientific). cDNA was reverse transcribed using the High-Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific). The qPCR reaction mix was prepared using TaqMan Gene Expression Assay probes in TaqMan Gene Expression Master Mix (Thermo Fisher Scientific). TaqMan probes included: actin beta (Actb); 18s rRNA (18s); heme oxygenase 1 (Hmox1); NAD(P)H quinone dehydrogenase 1 (Nqo1), glutamate-cysteine ligase modifier subunit (Gclm). qPCR was performed in a MicroAmp Optical 384-well reaction plate with optical adhesive film on a QuantStudio 7 Flex Real-Time PCR System (Applied Biosystems, MA, USA). Samples were normalized to Actb and 18s RNA, and data are expressed as relative to stimulated but untreated conditions using the 2−ΔΔCT method.

In Vivo Imaging System (IVIS).

10-week-old female C57BL/6J mice (Jackson Laboratory) were injected s.c. in the skin over the neck with pITA-DiD MPs. Imaging of tissues surgically excised 24h after treatment was performed using an IVIS Spectrum in vivo imaging system (PerkinElmer). Region of interest (ROI) values are reported as average radiant efficiency (p/s/cm²/sr)/{μW/cm²]) (= radiant efficiency/ROI surface area). ROIs were kept the same size for each organ (ex vivo). Image analysis was performed using Living Image software, and quantitative analysis using doing ROI analysis of radiant efficiency.

Experimental Autoimmune Encephalomyelitis (EAE).

EAE was induced in 10-week-old female C57BL/6J mice (Jackson Laboratory), as previously described [68, 69]. Briefly, Complete Freund’s Adjuvant (CFA) was produced using 4mg/mL heat-inactivated tuberculosis (hiTB, nonviable desiccated Mycobacterium tuberculosis H37 Ra) in Incomplete Freund’s Adjuvant, combined with 2mg/mL peptide (MOG35–55) was prepared in sterile PBS. Solutions were mixed at a 50:50 ratio (v/v) and emulsified for 35sec using a probe sonicator at 87% amplitude (Qsonica, CL-18 1/8 in. probe). For disease induction, mice were anesthetized under isoflurane and injected at two sites in the subcutaneous space along the midline, totaling 200uL of emulsion per mouse. EAE mice received 50ng of pertussis toxin (List Biological Laboratories) in sterile PBS i.p. two and 24hrs after emulsions. For treatments, mice were injected s.c. with 50uL of pITA MPs resuspended at 40mg/mL (2mg total) into the tented skin over the neck (i.e., scruff) every five days, beginning on at the start of disease induction (day 0) for a total of four treatments (8mg over the course of disease). Alternatively, mice received 100uL of soluble ITA (i.p.) resuspended at 20mg/mL (2mg total) every other day, beginning at the start of disease induction (day 0). To neutralize pH, the ITA solution was adjusted using 1M NaOH before administration. Mice were monitored daily for weight and paralysis, which was assessed using a clinical score rubric: 0, no symptoms; 0.5, partial tail paralysis; 1, paralyzed (limp) tail; 1.5, decrease in hind limb stability; 2, hind limb weakness and altered gait; 2.5, partial hind limb paralysis; 3, hind limb paralysis; 3.5, hind limb paralysis and trunk weakness; 4, hind limb paralysis and partial front limb paralysis; 4.5, hind and front limb paralysis; and 5, moribund. Humane end points include a score of 4 for two consecutive days or ≥ 4.5. Water (hydrogel) and food were placed on the cage floor level to accommodate paralysis during disease course. Disease incidence (symptomatic) was defined as the first day a mouse exhibited a clinical score >0; paralysis incidence was defined as the first day a mouse exhibited a clinical score ≥3.

Statistics.

Statistical calculations were performed using Prism (GraphPad). Sample sizes (number of experimental replicates (N) and the number of biological replicates (n)) are indicated in the figure legends. Error bars in all panels represent the mean ± standard deviation (unless otherwise noted). p ≤ 0.05 was considered significant; levels of significance were defined as *p<0.05, **p<0.01, ***p<0.001, and ****p<0.0001. For comparisons of more than two groups, a one-way ANOVA with Tukey’s post hoc test was performed to compare across all groups; a Dunnett’s test was performed to compare each group to a control. For comparisons of two groups, two-tailed, unpaired Welch’s t tests were performed. For conditions where number of replicates differed between groups, Brown-Forsythe analyses were performed to confirm insignificant variance before comparisons. For EAE, Wilcoxon test for Kaplan–Meier analyses, Steel–Dwass (clinical scores), and two-way ANOVA with Tukey–Kramer post hoc test (weight loss) tests were used for comparisons at individual time points.