Biomaterial mediated simultaneous delivery of spermine and alpha ketoglutarate modulate metabolism and innate immune cell phenotype in sepsis mouse models

Abstract

Although different metabolic pathways have been associated with distinct macrophage phenotypes, the field of utilizing metabolites to modulate macrophage phenotype is in a nascent stage. In this report, we developed microparticles based on polymerization of alpha-ketoglutarate (a Krebs cycle metabolite), with or without encapsulation of spermine (a polyamine metabolite), to modulate cell phenotype that are critical for resolution of inflammation. Poly(alpha-ketoglutarate) microparticles encapsulated and released spermine (spermine(encap)paKG MPs) in vitro, which was accelerated in an acidic environment. When delivered to bone marrow-derived-macrophages, spermine(encap)paKG MPs induced a complex phenotypic profile outside of the typical M1/M2 paradigm, with distinct effects in the presence or absence of the pro-inflammatory stimulus lipopolysaccharide. Of particular interest was the increase in expression of CD163, which has been linked to anti-inflammatory responses in sepsis. Therefore, we systemically administered spermine(encap)paKG MPs to two different murine models of sepsis using acute or chronic injection of LPS. Macrophages and neutrophils in the liver and spleen of animals treated with spermine(encap)paKG MPs increased expression of CD163, concomitant with normalizing of glycolysis and oxidative phosphorylation, in both models. Overall, these results show that spermine(encap)paKG MPs modulate macrophage phenotype in vitro and in vivo, with potential applications in inflammation-associated diseases.

1. Introduction

Macrophages play a central role in homeostasis, disease, and the response to infection or injury. They accomplish these diverse functions by dramatically shifting phenotype in response to environmental cues. A spectrum of macrophage phenotypes have been described, ranging from pro-inflammatory “M1-like” macrophages that combat infection and initiate wound healing processes to multiple “M2-like” phenotypes that are critical mediators of wound healing resolution, anti-parasitic responses, and allergy [1]. Because of their key role in so many disease and injury settings, an active area of research is modulation of macrophage phenotype for therapeutic purposes [2, 3]. Many strategies focus on strategies to deliver drugs or cytokines to macrophages. However, many small molecule drugs fail to activate macrophages to physiologically desirable phenotypes, and manufacturing and translation hurdles limit the potential of cytokine delivery systems. There is a need for novel biologically inspired strategies to control macrophage phenotype.

In recent years, it has been shown that different phenotypes of macrophages exhibit distinct metabolic profiles [4]. For example, M1 macrophages stimulated with lipopolysaccharide (LPS) with or without the pro-inflammatory cytokine interferon-gamma (IFNg) primarily utilize glycolysis, whereas M2-like macrophages stimulated with interleukin-4 (IL4) with or without IL13 primarily utilize oxidative phosphorylation [5, 6] for their energy needs [7, 8]. Furthermore, it has been demonstrated that if glycolysis is blocked in M1-like macrophages inflammation can be reduced in mouse models of sepsis [9]. Additionally, the metabolites within the OXPHOS and glycolysis pathways themselves can act as signaling molecules and modulate the function of immune cells. For example, we have demonstrated that aKG can directly modulate the function of dendritic cells and T cells. Moreover, metabolites such as glutamate and aKG have been shown to be essential in maintaining M2-like alternatively activated macrophage phenotype [10]. But macrophage phenotype is much more complex than the two more commonly studied phenotypes of M1-like and M2-like; the true extent of their diversity, and how metabolism regulates them, is not known.

One metabolic pathway that plays a key role in macrophage polarization is alpha-ketoglutaric acid (aKG) (Figure 1a). M1-like macrophages have a broken Krebs cycle, which leads to the accumulation of metabolites such as succinate, which then acts as a pro-inflammatory metabolite [11–13]. On the other hand, M2-like macrophages have an intact Krebs cycle and are heavily dependent on the oxidative phosphorylation for ATP generation [14, 15]. Furthermore, it has been shown that glutamine supports M2-like phenotype by generating aKG (downstream of glutamine) and feeding this metabolite into the Krebs cycle [16]. Therefore, delivery of aKG may be a powerful tool to further support M2-like macrophage phenotype. Thus, in this study we utilized microparticles composed of poly(aKG) (Figure 1a, 1b) which degrade into aKG byproducts [7]. We have previously demonstrated that paKG MPs modulate dendritic cell metabolism, a cell that has similar metabolic needs as macrophages in activated state [7].

Figure 1:

TCA cycle targeted delivery of spermine and aKG through delivery of spermine encapsulated poly(alpha ketoglutarate)-based microparticles (paKG) (a) to modulate function of inflammatory macrophages and neutrophils in chronic and acute sepsis (b). LPS = lipopolysaccharide.

Another critical metabolic pathway that regulates macrophage function is polyamine biosynthesis, which has been shown to modulate mitochondrial respiration and activation of macrophages (Figure 1c) [17–20]. Interestingly, lower polyamine levels are associated with inflammatory diseases[21, 22], and thus increasing the amount of polyamines within the inflamed tissue may be a strategy for targeting inflammation associated diseases. Spermine, a type of polyamine, is a small molecule with molecular weight of 202.34 g/mol, and as such can rapidly diffuse away from the injection site [23]. Therefore, modulating the local environment of the tissue by injections of spermine is challenging. Therefore, we reasoned that controlled release of spermine might be a more effective strategy to modulate macrophage behavior. Notably, spermine has two primary amine groups with approximate pKa of 10, so these molecules are protonated at physiological pH of 7.4. Because paKG polymers have carboxylic acid terminal moieties and are thus (deprotonated at physiological pH).

Herein, we tested if paKG MPs can be utilized to encapsulate and release spermine in a sustained manner. Then, we thoroughly evaluated the effects on primary murine bone marrow derived macrophages in vitro. To test if the spermine-encapsulated MPs could modulate macrophage phenotype in vivo, as a proof-of-concept, two sepsis mouse models were utilized. Sepsis is an inflammatory disease where the body generates amplified inflammatory responses leading to life-threatening conditions [24, 25]. In fact, sepsis has very high mortality rate (>15%) and currently tested treatments (e.g. cytokine blockers, inhibitors of toll-like receptors) have not led to success in clinic. Sepsis is propagated by inflammatory innate immune cells, which leads to damaging inflammation in large organs (e.g. liver, kidneys, brain) and eventually their failure [25]. Moreover, in response to infection or systemic inflammation, the immune system also responds by upregulating several immunosuppressive pathways. Overall, the metabolism of immune cells is dysfunctional in sepsis [26, 27], and therefore, modulating inflammatory cell metabolism and phenotype can be highly beneficial in sepsis (Figure 1d).

2. Results

2.1. Spermine modulates bone marrow derived macrophage phenotype

To determine if spermine by itself can modulate the function of macrophages, spermine was added across a range of doses to primary murine bone marrow derived macrophages (BMDMs) for 24 hours and the effect on macrophage phenotype was characterized by flow cytometry for expression of CD80, CD86, CD163, and CD206 (Figure 2, and representative flow plot analysis in Figure S1). This analysis was conducted in the presence and absence of LPS as a pro-inflammatory stimulus [28]. Interestingly, it was observed that the expression of CD80, a costimulatory molecule for T cell activation, was decreased to undetectable levels with doses of spermine greater than 6.25 μg/ml, regardless of the presence or absence of LPS in macrophages (Figure 2a). In contrast, another main T cell costimulatory molecule, CD86, increased with increasing doses of spermine in the absence of LPS, with no significant effects in the presence of LPS. Taken together these data suggest that spermine might be able to skew the phenotype of macrophages toward low CD80 and unaltered CD86 in the presence of inflammation.

Figure 2. Spermine skew macrophage toward an anti-inflammatory phenotype in a dose dependent manner.

(a) Flow cytometry analysis of macrophage surface markers (CD80, CD86, CD206, CD163) shows that soluble spermine influences expression dose-dependently. One way Anova with Tukey post hoc. (b, c) Volcano plots showing differential expressed genes (DEGs) (shown as red dots; cut-off: unadjusted p-value <0.05) of soluble spermine treated macrophages compared to non-treated (NT) macrophages without further stimulation (M0) (b) or stimulated with LPS (c). Labeled dots represent the DEGs with cut-off: Log2(fold change) 0.5/−0.5 and unadjusted p-value <0.05. (d,e) Hierarchical clustering of samples and genes in the heatmap represents Z-scores of DEGs (p>0.05) of either M0 (d) or LPS treated (e) macrophages. (d) The DEGs with a cut-off: Log2(fold change) 0.5/−0.5 and unadjusted p-value <0.05 of M0 and LPS treated macrophages are plotted as normalized counts and categorized depending on macrophage marker involvement and function, n=6, Student’s t-test. (f) Genes that were significantly modified by the treatment in the presence or absence of LPS are shown, One way Anova with Tukey post hoc.

Next, we investigated if spermine can modulate expression of CD206, a mannose receptor, and CD163, a scavenger receptor, both of which are known markers for M2-like macrophages [29, 30]. In the absence of LPS, spermine caused an increasing trend in the frequency of macrophages expressing CD163 in a dose dependent manner, whereas in the presence of LPS, this expression was unaltered (Figure 2a, S16). In contrast, spermine decreased the frequency of macrophages expressing CD206, in both the presence and absence of LPS. Thus, these data suggested that spermine by itself can influence macrophage phenotype in the presence of LPS, especially at 3.125 μg/mL concentration and above.

To further understand the effect of spermine on genetic changes, differential gene expression analysis was performed on bone marrow derived macrophages after 24 hours incubation with spermine at 3.125 μg/mL. Spermine-treated macrophages modestly modulated gene expression in comparison to untreated (M0) macrophages (Figure 2b). Different genes were modulated in the presence of LPS, and the effect sizes were larger (Figure 2c). Hierarchical clustering showed clear separation between M0 and spermine-treated macrophages in the absence of LPS (Figure 2d), and a lack of separation in the presence of LPS (Figure 2e). These results suggest that although in the absence of LPS, spermine modulated modest number of genes, these genes were consistently modulated across samples, whereas the results were more heterogeneous in the presence of LPS. Importantly, in the absence of LPS, spermine led to increased expression of genes associated with angiogenesis (Jag1) and led to decrease in the immune signaling associated gene Asprv1 (Figure 2f). In the presence of LPS, spermine led to increase in M2-like macrophage markers Arg1, Cbr2, and MMP9, angiogenesis-related gene Pecam1, and immune signaling-related genes TREM1 and TLR5 (Figure 2f). It was also determined that these different concentrations of spermine did not induce cell death in BMDMs as determined by cell-permeable dye assay (Figure S3).

2.2. Spermine can be efficiently encapsulated and released from paKG MPs in a sustained manner

Having determined that spermine can modulate macrophage phenotype in the presence and absence of LPS, we next set out to develop a formulation for its controlled release to mitigate delivery challenges resulting from its small molecular weight and high diffusion rate from injection sites. paKG polymers were utilized to generate microparticles (MPs) encapsulating spermine using an emulsion-evaporation process. The size of the paKG encapsulated spermine MPs (spermine(encap)paKG) was 2.7±0.1 μm, determined using dynamic light scattering (Figure 3a). Moreover, scanning electron microscopy showed that these particles were spherical in nature and had smooth morphology (Figure 3b). To have a depot effect paKG MPs should be able to release spermine in a sustained manner, and therefore release kinetics of spermine from paKG MPs was tested. Additionally, when phagocytes such as macrophages infiltrate the injection site, they may be able to phagocytose these particles, and degrade them in their endosome. The pH of endosomes can reduce to as much as 5 [31–33]. Therefore, the release of spermine from paKG MPs was measured at physiological extracellular pH of 7.4 and intracellular endosome pH of 5 (Figure 3c). It was observed that there were similar amounts of burst release at pH 5 and at pH 7.4, and spermine was released at a faster rate at pH 5 than at pH 7.4 (Figure 3c). The faster release of spermine from spermine (encap)paKG MPs can be attributed to faster degradation of the ester bonds of the paKG MPs. Overall, paKG MPs were able to release spermine in a sustained manner.

Figure 3: paKG MPs encapsulate and release spermine in a sustained manner.

(a) Dynamic light scattering data for size determination (representative of n=3). (b) The scanning electron microscope images. Scale bar = 5 μm. (c) The release kinetics at pH 7.4 and pH 5.0. (d) The zeta potential of paKG and spermine encapsulated paKG MPs. (n=3 for PaKG MPs; n=6 for spermine(encap)PaKG MPs). Data depicted as mean ± std err.

It was observed that paKG MPs could encapsulate spermine with 20±1% loading capacity and 90% encapsulation efficiency. This efficiency is considered high for encapsulation of a small molecule. In order to understand this high encapsulation efficiency, zeta potential changes in the paKG MPs before and after encapsulation was determined. The zeta potential of the paKG MPs was −35 mV and it changed to +3.2 mV for spermine (encap)paKG MPs (Figure 3d), suggesting that protonated spermine can effectively attach to the de-protonated carboxylic acid groups of paKG MPs. Also, spermine(encap)paKG MPs degrade in the presence of cell culture media as observed by decreasing particle size over 7 days (Figure S3c). This result suggests that paKG MPs can be utilized to encapsulate small molecules that are protonated.

2.3. Spermine encapsulated paKG MPs modulates macrophage phenotype in vitro

Spermine and aKG when released intracellularly may modulate the signaling pathways and thus ultimately modulate the function of macrophages. To test if macrophages can phagocytose spermine(encap)paKG MPs, BMDM were cultured in the presence of rhodamine loaded spermine(encap)paKG MPs, and confocal imaging was utilized to image the cells. It was determined that macrophages could indeed phagocytose these particles (Figure S3b).

Next, to test if spermine encapsulated in paKG MPs can induce phenotypic changes in macrophages, BMDM were cultured with different concentrations of spermine (encap)paKG MPs and characterized by flow cytometry and NanoString after 24 hrs. At a concentration of 1 μg/mL spermine(encap)paKG MPs decreased the frequency of macrophages expressing CD80+CD86+ by ~2-fold as compared to the control of no treatment (Figure S4a). At 10, 25 and 50 μg/mL, spermine(encap)paKG MPs increased the population of CD80+CD86+ macrophages. Interestingly, spermine (encap)paKG MPs increased the frequency of CD163+CD206+ macrophages in a dose dependent manner, and at 25 μg/mL, spermine (encap)paKG MPs increased the frequency of CD163+CD206+ population ~8-fold higher than the control of no treatment (Figure S4b). These data suggest that the spermine released from paKG MPs, along with the degradation products of paKG (aKG and 1,10 decanediol) together increase the frequency of = CD163+CD206+ macrophages. Furthermore, the addition of spermine (encap)paKG MPs did not decrease cell viability as compared to the no treatment control (Figure S5). In applications such as diabetic ulcers where the inflammation is driven by macrophages [34], and tissue regeneration is required that can be driven by CD163+ macrophages these spermine (encap)paKG MPs can be useful.

Notably, even in the presence of LPS at 50 μg/mL and 100 μg/mL spermine(encap)paKG MPs significantly decreased the frequency of CD80+CD86+ in macrophages (Figure S4). Importantly, spermine (encap)paKG MPs was able to increase CD163+CD206+ frequency 1.5 to 2-fold as compared to the control of no treatment and paKG MPs at the concentration of 25 μg/mL (Figure S4b). Overall, these data demonstrated that spermine (encap)paKG MPs were able to increase the frequency of CD163+CD206+ macrophages but not CD80+CD86+ macrophages.

The geometric mean fluorescent intensity (gMFI) of CD80, CD86, CD163 and CD206 were also analyzed on macrophages treated with spermine(encap)paKG MPs and paKG MPs with no treatment (NT) used as a control, to understand the effect of the formulation on number of these molecules per cell. In the absence of LPS, the geometric mean fluorescence intensity (gMFI) of CD80, CD86 and CD206 were not affected by the formulations as compared to controls (Figure 4a). Spermine(encap)paKG MPs and paKG MPs increased CD163 expression in a dose dependent manner, from 25 μg/mL and above (Figure 4a). As compared to LPS treatment alone, CD80 gMFI was decreased at 100 μg/mL spermine(encap)paKG MPs and 100 μg/mL paKG MPs, in the presence of LPS. CD86 gMFI was increased in 1 μg/mL paKG MPs, 25 μg/mL spermine(encap)paKG MPs and 25 μg/mL paKG MPs, and 50 μg/mL spermine(encap)paKG MPs and 50 μg/mL paKG MPs. CD206 gMFI was decreased in 50 μg/mL paKG MPs, and 100 μg/mL spermine(encap)paKG MPs and 100 μg/mL paKG MPs. Notably, both spermine(encap)paKG MPs and paKG MPs increased gMFI of CD163 in a dose dependent manner in macrophages in the presence of LPS, with significant differences as compared to no treatment in 25 μg/mL spermine(encap)paKG MPs, 50 μg/mL spermine(encap)paKG MPs and paKG MPs, and 100 μg/mL spermine(encap)paKG MPs and paKG MPs.

Figure 4: Spermine(encap)paKG MPs affect the bone-marrow-derived macrophages phenotype.

(a) paKG and Spermine encapsulated microparticles influence BMDMs surface marker expression in a dose-dependent manner; one way Anova with Tukey post hoc (b,f) Distribution of macrophage treatment groups in each cluster (identified via FlowSOM) of unstimulated (M0) (b) and LPS treated macrophages (f), n=3, Two-way ANOVA with Tukey post hoc analysis, *p<0.05, **p<0.01, ***p<0.001. (c,g) Percentage distribution of all cells included in the experiment into the clusters of M0 (c) and LPS treated macrophages (g) (d,h) Z-scores of the MFI intensities for each tested surface marker in each cluster for M0 (d) and LPS treatment (h). (e,i) Pie graphs showing the distribution of cells and bar graphs showing MFI Z-score for each marker in clusters that were identified as significantly different between treatment groups in b for M0 macrophages (e) and e for LPS treated macrophages (i).

To further understand the sub-populations of macrophages in the M1 to M2 spectrum, FlowSOM clustering was performed on the flow cytometry data obtained after treating the BMDMs with 50 μg/mL spermine(encap)paKG MPs and paKG MPs, with no treatment as control. 15 distinct population of macrophages were identified (Figure 4b). Cluster numbers 10, 14, 9, 5, 7, and 8 constituted approximately 90% of all cells (Figure 4c, S6a) and were utilized to assess the major phenotype of the cells. The majority of changes occurred in clusters 10 and 14. Cluster 14 was the only cluster that was decreased upon treatment with spermine(encap)paKG MPs, and this cluster was characterized by low expression of CD163 and CD86 and intermediate expression of CD80 and CD206 (Figure 4d, 4e). Cluster 10, which was increased by MP treatment, was characterized by high expression of CD206 and CD80 and moderately low expression of CD163 and CD86. Together, these results agree with analysis of each marker individually, which showed increased CD163 and CD86 with MP treatment.

Clusters 9, 5, and 8 were also enriched in cells treated with either blank or spermine-encapsulated MPs. Clusters 9 and 8 showed similar expression profiles, which was low expression of all 4 markers analyzed, suggesting a phenotype that was not well described by the selected markers. Finally, cluster 5 was enriched in cells treated with MPs, and this cluster was characterized by elevated CD86, low CD206, and moderate expression of CD163 and CD80. There were no significant differences between blank paKG MPs and spermine(encap) paKG MPs groups across any of the clusters, suggesting that the effects of the blank MPs may have been stronger than the encapsulated spermine in 24 hours of treatment.

Cluster analysis of the flow cytometry data obtained from LPS stimulated BMDMs with no treatment or treatment with 50 μg/mL blank or spermine-encapsulated paKG MPs led to the identification of 12 unique clusters (Figure 4f). Clusters 7, 10, 8, and 5 constituted approximately 79% of all cells (Figure 4g, S6b) and were utilized to assess the major phenotype of the cells. The largest effects were seen in cluster 7, which was decreased with MP treatment, and this cluster was characterized by high levels of CD80 and CD206 and low levels of CD163 and CD86. These results are in agreement with single marker analysis, which showed increased expression of CD86 and CD163 with MP treatment. Clusters 10, 8, and 5 were increased with paKG MP treatment. Cluster 10 showed high expression of CD86 and CD80; cluster 8 showed low expression of CD206 and CD163; and cluster 5 showed low expression of CD80 and CD86. Taken together, these results show substantial heterogeneity among macrophages treated with MPs, but without major differences between blank or spermine-encapsulated MPs.

To further understand the effect of spermine(encap)paKG MPs on macrophage phenotype, NanoString analysis was used to assess gene expression profiles of macrophages treated with 50 μg/mL blank or spermine(encap)paKG MPs in comparison to no treatment, again with separate analyses in the presence or absence of LPs (Figure 5). Both spermine(encap)paKG MPs and paKG MPs strongly decreased the expression of several genes when compared to no treatment (Figure 5a). Furthermore, in the presence of LPS, both spermine(encap)paKG MPs and paKG MPs strongly increased the expression of several genes when compared to LPS alone macrophages (Figure 5a). Hierarchical clustering analysis suggested that in the absence of LPS, both spermine(encap)paKG MPs and paKG MPs modulated expression of genes associated with metabolism, M1, M2, angiogenesis, fibroblast activation and antigen presentation (Figure 5b). In the presence of LPS, all treatment groups clustered separately (Figure 5c).

Figure 5: Spermine(encap)paKG MPs influence the bone-marrow-derived macrophages at the mRNA level.

(a) Volcano plots showing differential expressed genes (DEGs) (shown as red dots; cut-off Log2(fold change) of 0.5/−0.5 and unadjusted p-value <0.05) of either 50 μg/ml blank or spermine encapsulated microparticles (MP) treated macrophages compared to non-treated (NT) macrophages without further stimulation (M0) or stimulated with LPS. Labeled dots represent the top 10 DEGs. (b,c) Hierarchical clustering of samples and genes in the heatmap represent Z-scores of DEGs of either M0 (b) or LPS treated (c) macrophages. (d) The top 10 DEGs of M0 and LPS treated macrophages identified in (a) are plotted as normalized counts and categorized depending on macrophage marker involvement and function, n=5–6, t-test, *p<0.05, **p<0.01, ***p<0.001

Differentially expressed genes were then plotted individually to further assess differences between groups. Both paKG MPs and spermine (encap)paKG MPs decreased Timp1, Timp3, Ccn2, Bgn, and Fn1 genes, which related to regulation of the extracellular matrix (ECM) (Figure 5d). In the presence of LPS, there was no significant differences observed between the groups for these genes, although the Mmp8 gene, which encodes a potent matrix-degrading enzyme, was significantly decreased in macrophages treated with paKG MPs both in the presence and absence of LPS as compared to the control. In addition to ECM related genes, paKG MPs and spermine (encap)paKG MPs decreased fgf7 as compared to no treatment control, and paKG MPs decreased tgfb3 as compared to no treatment control. Interestingly, the proliferation gene Csf1, was decreased in both MP groups as compared to no treatment in the absence of LPS, however this gene was increased in the MP-treated macrophages in the presence of LPS. The adhesion genes Cd61 and Ly69 were increased in both MP groups as compared to the no treatment group. Also, Ly69, was increased in both MP groups in the presence of LPS. The inflammatory pathway genes pdl1, Il1a, Il6, CD86 and Tnf were increased in the MP groups compared to the control both in the presence and absence of LPS. However, the inflammatory pathway gene Tlr8, was decreased in the MP groups compared to the control both in the presence and absence of LPS. Antigen presentation genes H2-Ab1 and H2-Eb1 were decreased significantly in MP treated macrophages compared to no treatment. As far as recruitment genes were concerned, the genes Ccl7, Ccl2, Cxcl12, were all increased in macrophages treated with MPs in the presence of LPS, as compared to the LPS control alone. On the other hand, the migration gene Ccr2 was decreased significantly in the macrophages treated with spermine (encap)paKG MPs in the presence of LPS as compared to LPS alone group. Furthermore, paKG MPs and spermine (encap)paKG MPs in the presence of LPS increased M2 gene Klf4, as compared to LPS alone group. The lipid metabolism associated gene Cst7, was significantly increased in the macrophages treated with paKG MPs and spermine (encap)paKG MPs in the presence of LPS as compared to LPS alone group. Overall, these data suggest that both MP formulations had substantial effects on macrophage phenotype, with notable effects in the presence of LPS including decreased ECM related genes, increased genes associated with both M1 and M2 phenotypes, increased lipid metabolism, and increased recruitment related genes.

2.4. Spermine(encap)paKG MPs modulates macrophage and neutrophil phenotype in the spleen and liver in acute sepsis

Sepsis and septic shock represent a heterogeneous spectrum of complex biology and pathophysiology [35]. Novel therapies, outside of antibiotics, fluid resuscitation, and basic supportive care are utilized in the clinic, which are not very effective [36]. In this study, we utilized a murine sepsis model, which consisted of injection of purified lipopolysaccharide (20 mg/kg of LPS), intraperitoneally [37, 38]. LPS is a single, yet potent component of the complex pathogen associated molecular patterns (PAMPs) released by gram negative organisms. To test if the spermine loaded paKG MPs can reduce the indicators of acute sepsis, formulations (spermine(encap)paKG MPs, spermine only, PBS) were introduced intravenously via retro-orbital injections, 30 min. after LPS injections (Figure 6a). This hard to treat model was chosen to represent treatment of sepsis immune responses [39].

Figure 6: Spermine(encap)paKG MPs generate anti-inflammatory cellular phenotype in acute sepsis.

(a) Schematic of the study design for acute sepsis in mice. (b) Glycolytic capacity of liver cells (n=4; One-way ANOVA). (c) Maximal respiration in liver cells (n=4; One-way ANOVA). (d) gMFI of CD163 in macrophage in spleen (n=4; One-way ANOVA). (e) gMFI of CD206 in macrophages in spleen (n=4; One-way ANOVA). (f) gMFI of CD163 in neutrophils in spleen (n=4; One-way ANOVA). (g) gMFI of CD163 in macrophages in liver (n=4; One-way ANOVA). (h) gMFI of CD163 in neutrophils in liver (n=4; One-way ANOVA). Data depicted as mean ± std err.

To test if the spermine(encap)paKG MPs are delivered in the different organs of mice, spermine(encap)paKG MPs were generated encapsulating a hydrophobic near-infra red dye IR780. These particles were then injected intravenously in mice and then the mice were sacrificed after 4 hours (similar time frame as acute sepsis). Different organs from the mice were harvested and imaged using IVIS imaging. It was observed that the spermine(encap)paKG MPs were able to accumulate in brain, heart, spleen, kidneys, liver and lungs of these mice (Figure S15). Since intravenously injected particles are known to accumulate in the liver and spleen, and the cells from these organs can contribute to organ failure, all the cells of these organs were isolated and extracellular flux analyses were performed using Seahorse extracellular flux assays to identify glycolysis and mitochondrial respiration (oxidative phosphorylation) of cells. This sepsis model significantly decreased the glycolytic capacity of cells in the liver (Figure 6b). Treatment with soluble spermine increased the glycolytic levels compared to the PBS control, and the spermine(encap)paKG MPs further increased the glycolytic capacity higher than all the conditions tested. Since glycolytic capacity is an indication of the cells’ capability to respond to stress, a higher glycolytic capacity provided by spermine(encap)paKG MPs (Figure 6b). Furthermore, sepsis also significantly decreased maximal respiration of cells in the liver (Figure 6c). There was an increasing trend in maximal respiration of cells in liver upon spermine(encap)paKG MPs, treatments, however these differences were not significant.

Next, upon analysis of macrophages (F4/80+ cells) in the spleen, significant changes were observed in populations across treatment groups (Figure 6d, S7). The gMFI of CD80 in spleen macrophages in the acute sepsis was significantly lower in mice that did not have sepsis, treated with soluble spermine, and treated with spermine(encap)paKG MPs as compared to untreated mice with sepsis (Figure S7a). The gMFI of CD163 in macrophages in the acute sepsis was significantly increased in mice treated with spermine(encap)paKG MPs as compared to all the conditions (Figure 6d). The gMFI of CD206 in spleen macrophages in the acute sepsis was significantly increased in mice treated with spermine(encap)paKG MPs as compared to all the conditions except untreated mice with sepsis (Figure 6e).

Since other innate immune cells such as neutrophils are also involved in immune responses toward sepsis and might be affected by the formulations [40–42], the ability of spermine(encap)paKG MPs to modulate neutrophil population was also studied (Figure S8). In case of neutrophils, soluble spermine and spermine(encap)paKG MPs significantly decreased CD80 gMFI on neutrophils as compared to untreated mice with sepsis in spleen (Figure S9a). Furthermore, gMFI of CD163 in neutrophils in the acute sepsis in spleen was significantly increased in mice treated with spermine(encap)paKG MPs as compared to all the conditions (Figure 6f). The gMFI of CD206 in neutrophils in the acute sepsis was significantly increased in mice treated with spermine(encap)paKG MPs as compared to all the conditions except untreated mice with sepsis (Figure S9b). Additionally, paKG MPs by themselves induced significantly higher gMFI for CD163 and CD206 on neutrophils as compared to PBS treated mice in acute sepsis. Moreover, gMFI for CD80 was not significantly decreased by paKG MPs in neutrophils as compared to PBS treatment (Figure S14).

Similar to the spleen, there were significant changes in the macrophage and neutrophil populations in the liver, where most of the particles are expected to localize. There was no significant difference observed between the groups for gMFI of CD80 and CD206 in macrophages in liver (Figure S10a, S1b). Spermine(encap)paKG increased gMFI of CD163 in macrophages as compared to all the other groups in liver (Figure 6g). Similar to macrophages, there was no significant difference observed between the groups for gMFI of CD80 and CD206 in neutrophils (Figure S10c, S10d). It was also observed that spermine(encap)paKG MPs were able to significantly increase gMFI of CD163 in neutrophils in liver of mice as compared to all the other groups (Figure 6h).

2.5. Spermine releasing paKG MPs modulates macrophage and neutrophil phenotype in liver and spleen in chronic sepsis model

Sepsis is a highly heterogeneous disease [43], and even after initial resolution or treatment of acute sepsis, there are fluctuations observed in the inflammatory responses, and a low early sepsis inflammatory response may become highly aggressive after few days in patients [44, 45]. Due to this heterogeneity in time-based changes in inflammatory profile, it is important to study chronic sepsis just as much as acute sepsis [46, 47]. Therefore, in this study we utilized a chronic sepsis mouse model, where low level of LPS (3 mg/kg) was injected intraperitoneally [48], and after treatment with different formulations the immune cell profile was investigated at end of 7 days in mice (Figure 7a).

Figure 7: Spermine(encap)paKG MPs modulate neutrophil and macrophage phenotype in chronic sepsis.

(a) Schematic of the study design for chronic sepsis in mice. (b) Glycolytic capacity of liver cells (n=4; One-way ANOVA). (c) Maximal respiration in liver cells (n=4; One-way ANOVA). (d) gMFI of CD163 in macrophage in spleen (n=4; One-way ANOVA). (e) gMFI of CD163 in neutrophils in spleen (n=4; One-way ANOVA). (f) gMFI of CD80 in macrophage in liver (n=4; One-way ANOVA). (g) gMFI of CD163 in macrophage in liver (n=4; One-way ANOVA). (h) gMFI of CD80 in neutrophils in liver (n=4; One-way ANOVA). (i) gMFI of CD163 in neutrophils in liver (n=4; One-way ANOVA). Data depicted as mean ± std err.

It was observed that after 7 days of sepsis induction and treatment, there was no difference in the glycolytic capacity of liver cells (Figure 7b). Moreover, even after 7 days of sepsis induction and treatment, PBS treated mice had decreased maximal respiration of cells in the liver as compared to no sepsis induction control (Figure 7c). Spermine(encap)paKG MPs increased the maximal respiration of liver cells, which was statistically similar to the no sepsis induction condition. Similar to acute sepsis, significant changes in the macrophage and neutrophil population were seen in the spleen and liver across treatment groups. Specifically, in the spleen of mice suffering from chronic sepsis, there was a significant increase in gMFI of CD163 in macrophages as compared to all the other groups (Figure 7d). Furthermore, gMFI of CD80 and CD206 was not significantly different between the groups in spleen in chronic sepsis mice (Figure S11). In case of neutrophils, in the spleen of mice there was a significant increase in gMFI of CD163 as compared to all the other groups, except untreated mice with sepsis (PBS) (Figure 7e). There was no significant differences between the groups as far as gMFI of CD80 on neutrophils is concerned (Figure S12a). Also, there was significant upregulation of gMFI CD206 in spleen neutrophils as compared to mice that had no sepsis (Figure S12b). In the liver of mice with chronic sepsis, the macrophages had lower CD80 gMFI as compared to untreated mice with sepsis in spermine(encap)paKG MPs, and this CD80 gMFI between spermine(encap)paKG MPs and no sepsis control (no induction) was not significantly different from each other (Figure 7f). In case of CD163 and CD206 gMFI on macrophages in liver, spermine(encap)paKG MPs were significantly higher than no sepsis control but it was not significantly different than other groups (Figure 7g, S13). Moreover, there was no significant differences between the groups for gMFI of CD80, CD163, and CD206 on neutrophils in the liver in the chronic sepsis (Figure 7h, 7i, S14).

To assess the health of the liver, alanine transaminase (ALT) levels in the serum of mice was tested at the end of the study. It was observed that there was no significant differences between no sepsis induction and spermine(encap)paKG MPs, whereas these two groups were significantly lower as compared to mice treated with spermine or PBS (Figure S16). These data suggest that spermine(encap)paKG MPs might protect mice from liver damage.

3. Discussion

The field of metabolite delivery for modulating the metabolism of immune cells and their phenotype is exciting as it provides the ability to utilize synthetic molecules to affect disease outcomes. In this manuscript, we demonstrated that paKG MPs can be utilized to deliver spermine and aKG, metabolites that can affect macrophage metabolism and their phenotype. These particles modulated the surface protein expression on macrophages in vitro and macrophages and neutrophils in vivo that shape the innate immune responses in inflammatory conditions. Moreover, these formulations also led to modulation of genes at the mRNA level and develop a complex phenotype of macrophages. This simultaneous delivery of spermine and aKG can also lead to modulation of metabolism in cells of the liver and spleen in mouse sepsis model, which is significant as the metabolism of cells in these organs in sepsis is dysfunctional. Nanoparticles delivering metabolites have been shown to modulate immune responses in sepsis mouse models [49]. In this study, microparticles of spermine(encap)paKG were utilized for delivery of spermine and aKG, however, nanoparticles can also be utilized for delivery of these molecules. Specifically, evaporation-emulsion technique, which was utilized to generate microparticles, same process can be utilized to generate nanoparticles as well. Also, these spermine(encap)paKG nanoparticles might have a differential biodistribution and may accumulate in more organs than the microparticles. Moreover, the spermine(encap)paKG MPs vary from size of 1–4 μm, which is in the phagocytosable range for macrophages (Figure S3b). However, this polydispersity can be prohibitive for use in humans, and therefore, for translation purposes the particle generation will need to be further optimized in future studies to get a narrow polydispersity.

Notably, to support an anti-inflammatory phenotype of innate cells like macrophages and dendritic cells, mitochondrial function needs to be supported. A class of metabolites called polyamines (e.g. spermine, spermidine, putrescine) are known to support mitochondrial respiration [50]. Recent studies suggest that hypusine (a rare amino acid) addition to eIF5a (a transcription factor) led to increase in mitochondrial respiration [51]. Polyamines are known to upregulate hypusination of eIF5a and lead to upregulated mitochondrial respiration. However, polyamines, such as spermine and spermidine, are only effective with relatively high dose at 3–6 mM (dose high enough to induce side effect both in vitro and in vivo) [52, 53]. Our results demonstrated that soluble spermine actively modulated macrophage phenotype both at protein (CD80, CD86, CD163 and CD206) and at RNA level. Interestingly, spermine increased CD86 gMFI in macrophages and decreased CD80 gMFI in the presence of LPS. Although CD86 is often associated with inflammation, it has been shown that CD86 and CD80 differentially regulate IL-10 production in T cells in inflammatory disorders (e.g. sepsis patients), with lower expressions of CD86 and IL-10 associated with septic patients [54, 55]. Moreover, patients that did not survive sepsis, have higher levels of CD80 [54, 56]. However, both CD80 and CD86 can be predictive of sepsis survival in human patients, and for this study we focused on analyzing CD80 marker on innate immune cells in vivo [57]. Nonetheless, since these markers by themselves might not be sufficient to understand macrophage polarization and therefore analyzing several parameters simultaneously at mRNA level might be useful.

In case of mRNA, spermine significantly modulated several pathways including M2 macrophage, angiogenesis and autoimmunity associated genes. This data is interesting because in tissue regeneration applications such as wound healing, elevated M2 to M1 ratios have been shown to affect tissue wound healing rates [58, 59]. From the Nanostring analysis it was determine that spermine upregulated M2-like phenotype in macrophages in the presence of LPS. Therefore, the strategy of delivering spermine through paKG MPs was tested in a sepsis model of systemic inflammation. Furthermore, these data also suggest that in the presence of inflammation spermine can be utilized in inflammatory diseases such as dry eye disease [60, 61] or acetaminophen-induced liver toxicity [62], where macrophages play an important role in tissue destruction. It has been previously demonstrated that spermine can modulate the metabolism of macrophages and their activation profile as well through hypusination [20], which is also supported by our data. However, our data demonstrated that spermine not only modulates pro-inflammatory, but also anti-inflammatory phenotype in these macrophages. Therefore, the use of spermine by itself may not be sufficient if anti-inflammatory macrophage phenotype is desired. Notably, M1-like macrophage gene expression was not affected by spermine(encap)paKG MPs as compared to the controls. This might depend on the levels of spermine released from the particles intracellularly in the macrophages. Since these assays were run for 24 hours, the effect might not be as pronounce, and in future studies longer assays will need to be run to understand long-term effect of these particles on gene analyses of M1-like macrophage associated genes.

Notably, when macrophages are polarized with IL-10 a phenotype called M2c is generated, which is characterized by upregulated CD163 and secretion of ECM degrading matrix metalloprotease (MMP)-7 and −8, suggesting distinct effects on ECM remodeling [63]. It was observed that paKG MPs significantly downregulated MMP8, but spermine(encap)paKG MPs did not modify this gene. Moreover, both paKG MPs and spermine(encap)paKG MPs significantly downregulated Timp1 and Timp3, which are associated with M1-like macrophages, thus suggesting support of M2-like macrophage phenotype from these particles [64]. Overall, modulation of these ECM-associated genes might support a M2-like macrophage phenotype by spermine(encap)paKG MPs.

In addition to spermine, aKG is another interesting metabolite that supports mitochondrial function. We have previously shown that immunosuppressive Krebs cycle metabolite, alpha-ketoglutarate (aKG[65]) can be formulated into polymers and used as a delivery vehicle, and can fuel the Krebs cycle and can regulate cell function, metabolism and growth [7]. In fact, aKG controls amino acid production, functions as a reactive oxygen species (ROS) scavenger, regulates G protein function (therefore acting as a signaling molecule), and given that aKG is a homologue of glutamine, it impedes protein catabolism, enhances nitrogen retention, and has been shown by us and others to be an immunomodulatory molecule [66, 67]. Furthermore, simultaneous delivery of spermine and aKG can have beneficial effects on mitochondrial respiration and macrophage phenotype. Therefore, in this study spermine encapsulated paKG MPs were generated, which were in the phagocytosable range. Moreover, we were also able to demonstrate that these MPs encapsulated and released spermine in a sustained manner, which might be helpful in providing these metabolites to resolve diseases where long-term responses are desired. The spermine(encap)paKG MPs can act as a depot for sustained release of aKG and spermine simultaneously, which can have beneficial effects in enhancing local alternatively activated macrophage population. Moreover, spermine might lead to similar decrease in inflammatory markers in other antigen presenting cells, as we observe in this study in macrophages. Additionally, spermine has been shown to polarize T cells toward expressing Foxp3 (a transcription factor[68] involved in tissue regeneration) and is shown to be decreased in autoimmune diseases such as multiple sclerosis [69]. Thus, combinatorial delivery of aKG and spermine might polarize both innate and adaptive T cells toward a regenerative phenotype.

This study also demonstrated that paKG MPs and spermine(encap)paKG MPs were able to significantly increase gMFI of CD163 in macrophages in a dose dependent manner as compared to no treatment group. Moreover, clustering analysis demonstrated that several distinct macrophage populations were generated with variable expression of CD80, CD86, CD163 and CD206. These data suggested that there is a heterogenous macrophage population generated when aKG or aKG and spermine delivery is performed. Nonetheless, both MPs increased the cell population with CD163 gMFI and led to higher frequency of cells with CD163+CD206+ macrophages, which might be useful in application where inflammation mediated tissue destruction is observed. In fact, in inflammatory diseases such as sepsis there is a severe dysfunction in macrophage metabolism, and leads to mixed activation phenotype in humans and in mice [70]. Therefore, a formulation that stabilizes the metabolism of macrophages can then help in generating an appropriate response to the disease. Notably, we found that Klf4 gene was upregulated in particle treated macrophages in the presence of LPS, which can be useful for treatment of inflammatory diseases like sepsis [71]. However, it is important to note that in addition to M2-like macrophage associated genes, several M1-like macrophage associated genes such as TLRs were upregulated in the presence of the particles, which might play an important role in tissue reorganization. Additionally, imbalance of spermine and aKG is utilized by cancerous cells to survive and proliferate [72]. There is a possibility that delivery of spermine and aKG for extended period may allow for cancerous cells to survive and due to immunosuppression allow for the tumors to form. To mitigate this the size or chemistry of the particles can be modified such that the particles have a transient effect on the immune system [73, 74]. This strategy will be useful in diseases such as acute sepsis where transient effect is desirable.

In both acute and chronic sepsis, spermine(encap)paKG MPs increased the mitochondrial metabolism of cells in the liver as compared to the control, and toward normalization of no sepsis control. These data suggest that delivery of spermine and aKG simultaneously was able to modulate the metabolism of cells even in the presence of sepsis. This data is important because the metabolism of cells in the liver during sepsis is dysfunctional [70]. Overall, the data suggested that spermine(encap)paKG MPs might provide metabolic fitness toward homeostasis to the cells in the liver. In addition to metabolism, spermine(encap)paKG MPs generated macrophages with higher CD163 gMFI in macrophages present in the spleen and liver in both acute and chronic sepsis, which correlated well with the in vitro results. In addition to macrophages, neutrophils also play an important role in propagating sepsis [75]. Neutrophils expressing CD163 are known to be immunosuppressive, and it is highly beneficial that spermine(encap)paKG MPs were able to increase the neutrophil population with higher CD163 gMFI in a major organ like liver, which is susceptible to failure and is a major cause of mortality and morbidity due to sepsis [76, 77]. These data are highly encouraging as these suggest that the formulation was able to reduce systemic activation of innate immune response, in both acute and chronic sepsis. Overall, these data suggest that paKG MPs that are able to release spermine and aKG in a sustained manner, and were able to modulate the innate immune cell phenotype in a acute and chronic sepsis model. Interestingly, others have demonstrated that spermine by itself can reduce cytokine production in sepsis,[78] however, in this study 0.5 to 1 mM doses were used, which can be themselves toxic. Moreover, although in our study, the change in macrophage phenotype was dominated by paKG MP, it is expected that in vivo with longer release times, both paKG MPs and spermine play an important role in macrophage phenotype modulation.

4. Conclusion

In conclusion, we report that paKG MPs are suitable for releasing spermine in a sustained manner, and that these spermine(encap)paKG MPs modulated macrophage phenotype both at protein and mRNA level in vitro. Our findings can be extended to encapsulation and release of other polyamines such as Putrescine and Spermidine, which also play an important role in control of immune cell function. We were also able to demonstrate that spermine(encap)paKG MPs modulated the CD80, CD206, CD163 markers in both acute and chronic sepsis (macrophages, and neutrophils) and were able to bring metabolic homeostasis in these two sepsis models. This formulation thus can be very useful in modulating innate immune responses in inflammation associated diseases.

Materials and Methods

paKG synthesis

The paKG polymer was synthesized using previously reported methodologies [7]. Briefly, a 1:1 mole ratio of ketoglutaric acid and 1,10-decanediol was combined in a 100 mL round-bottom flask. The contents were stirred at 130 °C for 2 hours and were placed under a vacuum. Using methanol, the paKG polymer was then precipitated and washed. The remaining methanol was then dried using a rotary evaporator.

paKG MP-spermine synthesis

Microparticles were generated using previously reported double water-oil-water emulsion techniques [7, 79]. Briefly, 4 mL of chloroform (VWR, Radnor, PA) was utilized to dissolve 200 mg of the paKG polymer. Then 400 μL of either DIH₂O (for paKG MPs) or 100 mg of spermine in 400 μL of either DIH₂O (for paKG(spermine) MPs) was added and sonicated for 30 seconds at an amplitude of 70%. This mixture was then combined with 30 mL of 2% polyvinyl alcohol (PVA) (Acros Organics, Fairlawn, NJ) in DIH₂O and homogenized at 7,000 rpm with an industrial homogenizer for 2 min. The generated emulsion was then added to 80 mL of 1% PVA and stirred at 400 rpm for 3 hours to evaporate chloroform. The synthesized particles were centrifuged at 2000 × Gs for 5 min (Eppendorf, Hauppauge, NY) The supernatant was removed, and particles were resuspended in DIH₂O. This procedure was repeated 3 times to remove any remaining PVA. The formed particles were then resuspended in 5 mL of DIH₂O, placed in −80 °C for 2 hours, and lyophilized for 48 hours. The microparticles were stored at −20 °C until further use.

Microparticle imaging

Images of the microparticles were acquired using the NOVA 200 - FEI at Erying Materials Center at Arizona State University.

Release kinetics of spermine

Release kinetics of aKG and spermine was assessed by incubating 5 mg of the microparticles in 1 mL of 1x phosphate buffered saline (PBS) with a pH 7.4 or 5. The release kinetics was completed in triplicates. Each sample was placed on a rotisserie rotator at 37 °C and were centrifuged at 2000 × Gs for 5 min. The supernatant of 800 μL was carefully obtained and placed in a new 1.5 mL eppendorf tube (Eppendorf, Hauppauge, NY) once a day for 10 days and was stored in −80°C until use. The displaced 800 μL of supernatant from the original eppendorf tubes was replaced with 800 μL of new buffer and was placed back on the rotisserie rotator at 37 °C.

The release of spermine from the microparticles was analyzed by performing high-performance liquid chromatography (HPLC – Agilent Infinity II 1260, Santa Clara, CA) using a C-18 column (Gemini^® 3 mm NX-C-18 100 Å LC column 100×4.6 mm), with isocratic mobile phase of 98% acetonitrile and 2% DIH₂O for 3 min (elution time of 1.25 to 1.75 min) with detection at 210 nm.

Additionally, to determine if spermine(encap)paKG MPs degrade in the presence of cell culture media, 10 mg of spermine(encap)paKG MPs were incubated in 1 mL of BMDM culture media (see below for recipe) for 0 days, or 4 hrs, or 3 days or 7 days. The size of the particles was then determined using dynamic light scattering.

BMDM culture

BMDMs were generated from isolated hematopoietic stem cells (HSC) originating from the bone marrow of C57BL/6j mice in accordance with the institutional animal care and use committee (IACUC) of Arizona State University for the approved protocol 19–688R. The femur and tibia were extracted from C57BL/6j mice and placed in a wash medium containing DMEM/F-12 with L-glutamine (1:1) with 1% penicillin-streptomycin (VWR, Radnor, PA). A homogenous cell suspension was achieved by flushing the femur and tibia with 10 mL of the wash media. This suspension was then centrifuged at 300 × Gs for 5 mins and the supernatant was removed. Red blood cells within the cell pellet were lysed by adding 3 mL of 1x red blood cell (RBC) lysis buffer for 3 mins at 4°C to the cell pellet. The cell suspension was centrifuged at 300 × Gs for 5 mins and was resuspended in DMEM/F-12 with L-glutamine (VWR, Radnor, PA), 10% fetal bovine serum, 1% sodium pyruvate (VWR, Radnor, PA), 1% non-essential amino acids (VWR, Radnor, PA), 1% penicillin–streptomycin (VWR, Radnor, PA) and 30% macrophage colony-stimulating factor (M-CSF; M-CSF was obtained by culturing a L929 fibroblast cell line in DMEM/F-12 with L-glutamine (VWR, Radnor, PA), 10% fetal bovine serum, and 1% penicillin–streptomycin (VWR, Radnor, PA) until 70% confluency was reached.

The HSC suspension was then relocated to a tissue culture-treated T-75 flask (denoted as day 0) and incubated in 37 °C, 5% CO₂ incubator. After 48 hours, the cells within the flask were obtained, centrifuged, re-suspended in fresh media and placed in 6-well low attachment plates (VWR, Radnor, PA) for 10 days. Half or whole media was changed every other day or as needed. On day 9, the cells were transferred from the low attachment plates and seeded on 96 well round bottom tissue culture-treated polystyrene plates. On day 10, the cells were treated with either soluble aKG (0.05 mg/mL), soluble diol (0.05 mg/mL), soluble spermine (3.125 – 200 μg/mL), paKG MPs (0.01 mg/mL) or spermine(encap)paKG MPs (0.001–0.1 mg/mL). Lipopolysaccharide (LPS - 10 μg/mL) was utilized to assess how each group responded to inflammation. No treatment and LPS alone were used as a control.

For Nanostring analysis, BMDMs were cultured for 6 days after isolation from bone marrow in RPMI-1640 (ThermoFisher, Waltham, MA) supplemented with 10% FBS, 1% penicillin-streptomycin and 20 μg/ml murine M-CSF (Peprotech, Cranbury, NJ). Media was changes every 2 or 3 days. On day 6, cells were treated with 50 μg/ml microparticles or 3.125 μg/ml soluble spermine and in the absence of or with LPS for 24h.

Flow cytometry

Flow cytometry (FACS) staining buffer was prepared by generating 1% bovine serum albumin (VWR, Radnor, PA), 2 mM Na2EDTA (VWR, Radnor, PA) and 0.1% sodium nitrate (VWR, Radnor, PA). Antibodies used for staining were purchased and used as is (Tonbo Biosciences, Thermo Scientific, BD biosciences, Invitrogen). Flow cytometry was performed by following the manufacturer’s recommendation for antibody staining and guidelines set by ASU flow cytometry core using Attune NXT Flow cytometer (ThermoFisher Scientific, Waltham, MA, USA). Specifically, antibody cocktail was generated in FACS buffer fresh for every staining cycle. The cells either isolated from tissue or cultured in vitro were then resuspended in the antibody cocktail, and incubated in the dark at 4 °C for 40 min. The cells were then centrifuged at 300 xGs for 5 min, and washed 3 times using centrifuge-wash step and 1X FACS buffer. Finally, the cells were resuspended in FACS buffer and the data was acquired using flow cytometry. The data obtained was then analyzed using FlowJo software. Following are the reagents and antibodies used in the study:

	Target	Fluorophore	Company	Catalog #	Clone
1	CD86	SB600	Thermo	63-0862-82	GL1
2	CD80	PE-Cy5	Invitrogen	15-0801-82	16-10A1
3	CD16/CD32:Fc Block	NA	Tonbo	70-0161-M001	2.4G2
4	F4/80	BV702	Invitrogen	67-4801-80	BM8
5	Comp beads	NA	Invitrogen	01-2222-42	NA
6	L/D	eF780
7	CD163	SB436	ThermoFisher	62-1631-82	TNKUPJ
8	CD206	PECy7	ThermoFisher	25-2061-82	MR6F3
9	Ly6G	PE	ThermoFisher	12-5931-82	RB6-8C5
10	Ly6c	APC	ThermoFisher	17-5932-82	HK1.4

For single cell analysis, the obtained flow cytometry data was analyzed with tSNE dimensionality reduction, Phenograph, and FlowSOM clustering algorithms using FlowJo. Cells from each sample were downsampled and then concatenated together. A Phenograph algorithm was used to find an appropriate number of clusters. Concatenated cells were then clustered with the FlowSOM method with 16 clusters for untreated (M0) macrophages and 12 clusters for LPS treated macrophages and a SOM grid size 10 by 10. tSNE reduction was performed with the Barnes-Hut method using 1000 iterations and perplexity of 30. The MFI values from each cluster were converted into a z-score and plotted.

Gene expression analysis via Nanostring

Total RNA was extracted from cells using the RNAqueous^™-Micro Total RNA Isolation Kit (Thermo Fisher) according to the manufacturer’s instructions. Nanostring (NanoString, Seattle) multiplex gene expression analysis was performed using 100 ng RNA and a custom-designed panel of 259 genes related to macrophage phenotypes, angiogenesis, recruitment, migration, immune response, ECM, fibrosis, metabolism, phagocytosis and cellular functions. Nanostring data was normalized to the internal positive and negative control. Differentially expressed genes were identified using the NanoSolver advanced software (NanoString). Heatmaps were generated in R with heatmap.3 function using the row Z-scores of log(2)-transformed normalized data. Outliers were identified visual by scatter plots and statistical by Z-score >3 (resulted in one outlier each for M0 and LPS paKG blank particles).

Sepsis models and treatment of mice

All the animal experiments were approved by ASU IACUC under protocol # 22–1884R.

Acute Sepsis model

In this study, C57BL/6J mice were injected intraperitoneally with 20 mg/kg of lipopolysaccharide (LPS) in 50 μL sterile phosphate buffered saline (1X PBS) solution [37, 38]. On the same day, microparticles (2 mg/mouse) were intravenously (retro-orbital administration) administered in 100 μL sterile PBS solution. Alternatively, mice were be injected with the relevant control groups such as PBS alone and sol. spermine alone (2 mg/mouse). All mice were closely monitored for 30 mins post procedure (hunching, dyspnea, cyanosis and/or loss of balance). Next, mice were then sacrificed, 4 hours after LPS administration before mice suffer from sepsis shock for immunological studies using CO₂ asphyxiation followed by cervical dislocation. Organs were harvested and meshed in a 70 μm filter, centrifuged washed with 1X PBS (300 xGs) to obtain single cells suspension for immunological and extracellular flux assay experiments. The onset of the disease was determined by clinical signs of sepsis including reduced motor activity, lethargy, shivering, and hunch posture.

Chronic Sepsis model

In this study, C57BL/6J were injected intraperitoneally with 3 mg/kg of lipopolysaccharide (LPS) in 50 μL sterile PBS solution [48]. Microparticles (2 mg/mouse) or control monomer treatments were injected intravenously (retro-orbital administration) in 100 μL sterile phosphate buffered saline solution, respectively [38]. Only one injection of the treatment (or control) was performed on Day 0. All mice were closely monitored for 30 mins post procedure for clinical signs of sepsis including reduced motor activity, lethargy, shivering, and hunch posture. Thereafter, hourly monitoring (hunching, dyspnea, cyanosis and/or loss of balance) was conducted till 4 hours. Mice were then sacrificed 7 days after LPS administration for immunological studies before mice suffer from sepsis shock using CO₂ asphyxiation followed by cervical dislocation. Organs were harvested and meshed in a 70 μm filter, centrifuged washed with 1X PBS (300 xGs) to obtain single cells suspension for immunological and extracellular flux assay experiments.

Extracellular Flux Assays

Glycolysis and oxidative phosphorylation were measured with the Seahorse Extracellular Flux XF-96) analyzer (Seahorse Bioscience, North Billerica, MA). Cells were seeded in Seahorse XF-96 plates at a density of 200,000 cells per well. The cells were resuspended in unbuffered DMEM in the absence of glucose. Sequential injections were performed with D-glucose (10 mM), oligomycin (1 mM), and 2-deoxyglucose (100 mmol/L). The extracellular acidification rates (ECAR) after the injection of D-glucose was a measure of glycolysis, and the ECAR after the injection of oligomycin represented maximal glycolytic capacity. Non-glycolytic activity was quantified by the measure of ECAR after the injection of 2-deoxyglucose. Samples were analyzed with 10 technical replicates.

Oxygen consumption rate (OCR) was measured using Seahorse Extracellular Flux XF-96 analyzer (Seahorse Bioscience, North Billerica, MA). Briefly, 200,000 cells/well were seeded in Seahorse XF-96 plates. For OCR, media was changed to unbuffered DMEM containing 2 mM glutamine, 1 mM pyruvate, and 10 mM glucose following sequential injections of oligomycin (2 mM), 7 Carbonyl cyanide-4 (trifluoromethoxy) phenylhydrazone (FCCP) (1 mM), and antimycin/rotenone (1 mM). The OCR after the injection of oligomycin was a measure of ATP-linked respiration and the OCR after the injection of FCCP represented maximal respiratory capacity. Basal respiration was quantified by measuring OCR prior to the injection of oligomycin. All samples were analyzed with 6 technical replicates.

Statistics

Statistical analysis calculations were carried out using GraphPad Prism software 9.0. Comparisons between multiple treatment groups were performed using one-way ANOVA (LSD-Fisher test), or unpaired two-tailed t-test and p-values < 0.05 was considered statistically significant. All data is expressed in the form of mean ± standard error unless otherwise specified. The data that was significant is shown in the figures.

Data Availability

The raw/processed data required to reproduce these findings cannot be shared at this time as the data also forms part of an ongoing study.