Introduction

The host innate immune response is paramount in defending against pathogens, with macrophages playing a central role in this process. Upon stimulation, macrophages undergo a series of activation states that dictates their functional fates. Classically activated (M1) pro-inflammatory macrophages exhibited a proinflammatory state and are typically induced by type I cytokine interferon-γ (IFNγ) and toll-like receptor (TLR) stimulation in the presence of microbial triggers. In contrast, alternative activated (M2) macrophages are induced by type II cytokines such as interleukin-4 (IL-4) and IL-13, promoting tissue remodeling and defending against parasite infection, especially helminths1. Distinct activation states are accompanied by differential metabolic phenotypes, both upon and following infection2. However, dysregulated macrophages activation can lead to compromised immune responses, resulting in prolonged infections and heightened susceptibility to pathogens. In conditions such as sepsis, an overwhelming cytokines storm can trigger septic shock and extensive tissue damage3,4,5. Conversely, overshooting activation of M1 macrophages could lead to immunosuppression and immune paralysis, rendering the immune system ineffective against subsequent microbial infections4,6,7. Furthermore, during the advanced stages of cancer, tumor-associated macrophages (TAMs) often adopt the M2 phenotype, promoting tumor growth, angiogenesis, and immunosuppression8,9. Macrophages also play a critical role in resolving inflammation, underscoring the need to unravel the intricate regulatory networks governing macrophage activation and function.

One of the crucial characteristics in M1 macrophages is the metabolic state shifts from oxidative phosphorylation (OXPHOS) to aerobic glycolysis10,11,12,13. M1 macrophages heavily consume glucose, generating substantial lactate while shifting mitochondria from ATP synthesis to ROS production10,11,12,14. This glycolytic shift is crucial for phagocytosis and production of inflammatory cytokines15,16. In contrast, M2 polarization relies on OXPHOS to ensure an adequate supply of ATP17,18. Concomitantly, fatty acid oxidation and glutamine metabolism pathways support the tricarboxylic acid (TCA) cycle in M2 macrophages19,20,21.

Metabolic pathways, including glucose, fatty acid, and amino acid metabolism, converge at acetyl-CoA, a precursor for the TCA cycle. In M1 macrophages, impaired isocitrate dehydrogenase (IDH) and succinate dehydrogenase (SDH) functions lead to increased succinate levels due to the the aspartate-argininosuccinate shunt (AAS), which connects the TCA cycle with the urea cycle16. The urea cycle is pivotal for macrophage NO production, while the AAS facilitates the connection between TCA cycle intermediates and the urea cycle. Within the AAS, glutamate oxaloacetate aminotransferase (GOT) utilizes pyridoxal phosphate as a cofactor to catalyze the mutual transformation between aspartate and oxaloacetate, as well as between α-ketoglutarate and glutamate16.

GOT exists in two isoforms, GOT1 and GOT2. Aminooxy acetic acid (AOAA), a GOT inhibitor, disrupts the malate-aspartate shuttle, reducing mitochondrial glycolytic substrate utilization22. AOAA preconditioning shifts M1 macrophages towards greater reliance on mitochondrial respiration, resembling M0 macrophages in terms of aerobic glycolysis16. Notably, AOAA inhibits M1 polarization, leading to reduced nitric oxide (NO) and inducible nitric oxide synthase (iNOS) levels, as well as suppressed IL-6 expression16. Cytokine and NO production are crucial for pathogen clearance during infections23,24. These findings suggest that AOAA-induced disruption of the aspartate-malate shuttle could perturb the balance between the Krebs cycle and NO cycle, potentially impacting to succinate dehydrogenase (SDH) function and overall cellular processes.

Here, our study shows that AOAA pre-treatment suppresses the differentiation of macrophages into the M1 phenotype, aligns with previous research16. However, our investigation reveals that genetic knockout of Got1 in macrophages does not exert any influence on the differentiation of pro-inflammatory macrophages. Meanwhile, we conducted experiments involving mice with myeloid cell-specific overexpression of Got1, and the results indicate that Got1 is not essential for macrophage inflammatory responses. Furthermore, whether Got1 is overexpressed or knocked out in myeloid cells, it has no impact on the differentiation of M2 macrophages induced by IL-4. Within LPS-induced tolerance assay, Got1 also exhibits a passive role, devoid of active involvement, as evident from the results. These findings present a contrasting perspective, challenging the previously assumed role of Got1 in the macrophage inflammatory response.

Results

AOAA-induced cytokines production inhibition is independent of Got1

AOAA, a known inhibitor of GOT1, has been identified for its ability to curtail cytokines production in macrophages16. Here, we examine whether AOAA-induced cytokine reduction is mechanistically linked to GOT1 inhibition. To assess this, we generated Got1ΔLysM mice by breeding Got1-floxed mice with Lyz2-Cre mice, resulting in myeloid cell-specific Got1 depletion (Supplementary Fig. 1a, b). Genotype of mice was validated by genomic DNA PCR (Supplementary Fig. 1c). Confirmation of Got1 depletion efficiency was achieved through evaluating cellular protein and messenger RNA (mRNA) levels in bone marrow-derived macrophages (BMDMs) (Supplementary Fig. 1d, e). Subsequently, we explored whether AOAA treatment in BMDMs differentiated from Got1ΔLysM mice could reduce cytokines production (Fig. 1a). Remarkably, AOAA effectively suppressed IL-6 but not TNFα protein expression levels not only in BMDMs from Got1f/f mice but also in BMDMs from Got1ΔLysM mice (Fig. 1b, c). In addition, NO is generated by iNOS within pro-inflammatory macrophages, playing a crucial role as a notable marker of pro-inflammatory reactions25. AOAA dose-dependently reduced nitric oxide (NO) and inducible nitric oxide synthase (iNOS) levels16, in alignment in our data indicating a reduction in NO levels in both BMDMs from Got1f/f and Got1ΔLysM mice (Fig. 1d). As there are two isoforms of Got, we are intrigued by the possibility of Got2 expression being upregulated in response to AOAA, potentially serving as a compensatory mechanism against Got1’s inhibitory effects. However, the mRNA and protein levels of Got1 and Got2 remained unchanged upon AOAA treatment (Supplementary Fig. 2a–c). PF-04859989, another inhibitor reported to specifically targeted GOT126, was employed for comparison in our experiments. Our results indicate that PF-04859989, like AOAA, demonstrates comparable inhibitory effects on IL-6 and NO production, while having no impact on TNFα production (Supplementary Fig. 2d–f). Moreover, knockdown of Got2 by siRNA (sequences show in Supplementary Table 1) does not impact cytokine production upon LPS stimulation (Supplementary Fig. 3). Concurrently, we evaluated oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) levels in BMDMs derived from both Got1f/f and Got1ΔLysM mice. Additionally, our finding reveals that AOAA (Fig. 1e–h) or PF-04859989 (Supplementary Fig. 4a–h), not only suppresses OCR and ECAR in BMDMs from Got1f/f mice but also exerts similar effects in BMDMs from Got1ΔLysM mice. These results collectively underscore the pleiotropic effects of AOAA or PF-04859989, extending beyond the inhibition of GOT1 enzymatic activity. Their impact resonates significantly in modulating the production of inflammatory cytokines and NO not only in BMDMs from Got1f/f mice but also those from Got1ΔLysM mice.

Fig. 1: The inhibition of AOAA-induced cytokine production and metabolic states in BMDMs is independent of Got1.
figure 1

a Schematic of AOAA treatment in BMDMs differentiated from Got1f/f and Got1ΔLysM mice. b, c Supernatant IL-6, TNFα protein levels of BMDMs derived from Got1f/f and Got1ΔLysM mice. BMDMs were pre-treated with 10 mM AOAA and then stimulated with 10 mM AOAA and LPS (20 ng/mL) + IFNγ (100 ng/mL) for 24 h. Got1f/f mice, n = 3 biologically independent experiments, Got1ΔLysM mice, n = 3 biologically independent experiments. d NO production level in BMDMs derived from Got1f/f and Got1ΔLysM mice. Seahorse analysis of OCR in M0 macrophages (e) and M1 macrophages (f), ECAR in M0 macrophages (g) and M1 macrophages (h). ns, p > 0.05, not significant, **p < 0.01, ***p < 0.001, ****p < 0.0001, unpaired, p values were determined by two-tailed Student’s t test. Data are representative of three independent experiments (mean ± SD).

Full size image

The differentiation of pro-inflammatory macrophages is independent of Got1

To investigate the role of Got1 in pro-inflammatory macrophages, we therefore stimulated BMDMs with LPS or a combination of LPS and IFNγ to induce pro-inflammatory macrophages (Fig. 2a). Both LPS-stimulated and LPS + IFNγ-stimulated macrophages exhibited anticipated increases in inflammatory cytokines with significant elevation in mRNA (Il6, Tnfa, Il1b) and protein (IL-6, TNF-α) expression levels27. Surprisingly, Got1 deficiency in BMDMs from Got1ΔLysM mice did not influence the mRNA (Il6, Tnfa, Il1b) and protein (IL-6, TNF-α) production, when compared to Got1f/f mice (Fig. 2b–f). Furthermore, M1 macrophages primarily utilize glycolysis as their main metabolic pathway, generating significant levels of lactate28,29. However, the NO and lactate production remained unaffected by Got1 depletion (Fig. 2g, h).

Fig. 2: Got1 deficiency fails to influence the differentiation of pro-inflammatory macrophages.
figure 2

a Schematic of pro-inflammatory macrophages differentiation. b–d Il6, Tnfa and Il1b mRNA levels in BMDMs stimulated with LPS (100 ng/mL) for 4 h or LPS (20 ng/mL) + IFNγ (100 ng/mL) for 24 h. Got1f/f mice, n = 3 biologically independent experiments; Got1ΔLysM mice, n = 3 biologically independent experiments. e, f IL-6, TNFα protein levels in the supernatant determined by ELISA. g NO production level in BMDMs stimulated with LPS (100 ng/mL) or LPS (20 ng/mL) + IFNγ (100 ng/mL) for 24 h. Got1f/f mice, n = 3 biologically independent experiments; Got1ΔLysM mice, n = 3 biologically independent experiments. h Lactate production levels in BMDMs stimulated with LPS (100 ng/mL) or LPS (20 ng/mL) + IFNγ (100 ng/mL) for 24 h. Got1f/f mice, n = 3 biologically independent experiments; Got1ΔLysM mice, n = 3 biologically independent experiments. ns, p > 0.05, not significant, unpaired, two-tailed Student’s t test. Data are representative of three independent experiments (mean ± SD).

Full size image

To validate our findings, we isolated peritoneal macrophages (PMs), recognized as LyzM+ cells within the peritoneal cavity30,31, from both Got1f/f or Got1ΔLysM mice and repeated the same experiments as described in BMDMs (Supplementary Fig. 5a). We intraperitoneally injected 3% thioglycolate per mouse to stimulate the generation of an ample population of peritoneal macrophages (PMs) in mice, and subsequently harvested the PMs after a period of 3-4 days. Consistent with BMDMs, the mRNA (Il6, Tnfa, Il1b) and protein (IL-6, TNF-α) levels upon stimulation with LPS or LPS + IFNγ were comparable in both Got1f/f or Got1ΔLysM mice (Supplementary Fig. 5b–i). Since LPS is a TLR4 agonist, we investigated whether stimulation with an intact pathogen using a non-TLR4 agonist would yield different results. To address this, PMs were stimulated with Heat-killed fungal pathogen C. albicans (HKCA) while there was still no difference between Got1f/f or Got1ΔLysM mice (Supplementary Fig. 5b–d).

The AAS is pivotal in M1 macrophages10,32. GOT1, an AAS enzyme, catalyzes the interconversion between aspartate and oxaloacetate, as well as α-ketoglutarate and glutamate (Fig. 3a)33,34. The knockout of GOT1 resulted in a significant accumulation of aspartate. However, this effect was not observed in the AOAA-treated group35. Aspartate metabolism is crucial in M1 macrophages, as evidenced by their depletion of aspartate and its metabolic products32. Nonetheless, it remains unclear whether aspartate metabolism is associated with macrophage differentiation. Interestingly, we observed that pre-treatment with aspartate resulted in a significant increase in Il1b mRNA levels when inducing pro-inflammatory macrophages (Fig. 3b, c). Given the importance of aspartate metabolism, we hypothesized the accumulation of Got1 could potentially modify the functionality of M1 macrophages. We then employed the distinctive attributes of the Rosa26 locus to facilitate the construction of transgenic mouse model with Got136. As shown in Supplementary Fig. 6a, b, we crossbreed Got1stop/+ mice with Lyz2-Cre mice to obtain mice with myeloid cell-specific overexpression of Got1. Genotyping was performed by extracting genomic DNA from mouse tail tissue (Supplementary Fig. 6c). Then GFP reporter was detected by flow cytometry (Supplementary Fig. 6d). Cellular mRNA was extracted in BMDMs to assess the expression level Got1 (Supplementary Fig. 6e). BMDMs derived from Got1stop/+ or Got1stop/+; Lyz2-Cre mice were then stimulated with HKCA, LPS or a combination of LPS and IFNγ to induce pro-inflammatory macrophages. The mRNA (Il6, Tnfa, Il1b) and protein (IL-6, TNF-α) levels remained unaffected by Got1 overexpression in response to various stimulations (Fig. 3d–k).

Fig. 3: Got1 overexpression fails to influence the differentiation of pro-inflammatory macrophages.
figure 3

a Model of GOT1 in malate-aspartate shuttle. b Schematic of Aspartate treatment in wildtype BMDMs stimulated with LPS (100 ng/mL) or LPS (20 ng/mL) + IFNγ (100 ng/mL). c Il1b mRNA levels in BMDMs stimulated with LPS (100 ng/mL) for 4 h or LPS (20 ng/mL) + IFNγ (100 ng/mL) for 24 h. n = 3 biologically independent experiments. d–f Il6, Tnfa and Il1b mRNA levels in BMDMs stimulated with LPS (20 ng/mL) + IFNγ (100 ng/mL) for 24 h. Got1stop/+ mice, n = 3 biologically independent experiments; Got1stop/+; Lyz2-Cre mice, n = 3 biologically independent experiments. g–i Il6, Tnfa and Il1b mRNA levels in BMDMs stimulated with LPS (100 ng/mL) or HKCA (HKCA: BMDM = 1:1) for 4 h. Got1stop/+ mice, n = 3 biologically independent experiments; Got1stop/+; Lyz2-Cre mice, n = 3 biologically independent experiments. j, k IL-6, TNFα protein levels in the supernatant determined by ELISA. BMDMs were stimulated with 100 ng/mL LPS, LPS (20 ng/mL) + IFNγ (100 ng/mL) or HKCA (HKCA: BMDM = 1:1) for 24 h. Got1f/f mice, n = 3 biologically independent experiments; Got1ΔLysM mice, n = 3 biologically independent experiments. ns, p > 0.05, not significant, *p < 0.05, **p < 0.01, unpaired, two-tailed Student’s t test. Data are representative of three independent experiments (mean ± SD).

Full size image

Taken together, our comprehensive approach utilizing Got1 conditional knockout and Got1 overexpression models demonstrates unequivocally that Got1 is dispensable for the differentiation of pro-inflammatory macrophages.

Got1 depletion decreases ROS production in macrophages without affecting pathogen defense

We next explored the potential impact of Got1 on reactive oxygen species (ROS) production in macrophages, a critical component of the innate immune response against invading pathogens37,38. Our experimental setup (Fig. 4a) included the use of 2′,7′-Dichlorodihydrofluorescein diacetate (H2DCFDA) as an indicator of ROS generation, with the ROS inhibitor N-acetyl-l-cysteine (NAC) as a negative control39,40. BMDMs derived from Got1ΔLysM mice exhibited lower DCF-DA fluorescence levels indicating compromised ROS production capacity due to Got1 depletion (Fig. 4b). Nevertheless, no alterations in ROS production were detected in BMDMs derived from Got1stop/+ or Got1stop/+; Lyz2-Cre mice (Supplementary Fig. 6f, g). Intriguingly, this reduction in ROS levels did not affect the phagocytosis (Fig. 4c) or bactericidal abilities of BMDMs against Staphylococcus aureus (S. aureus) and Escherichia coli (E. coli) (Fig. 4d). Furthermore, we pretreated BMDMs with either AOAA (Fig. 4b) or PF-04859989 (Supplementary Fig. 7). Interestingly, our results showed that regardless of the presence of inhibitors, the deletion of Got1 effectively suppressed ROS production. Similarly, PMs from Got1ΔLysM mice displayed lower DCF-DA fluorescence levels without any alteration in phagocytosis activity (Supplementary Fig. 5j, k). Next, we performed pathogens infection in vivo to test the role of Got1-mediated ROS production in Got1f/f and Got1ΔLysM mice (Fig. 4e). It is foreseeable that there is no difference in survival between Got1f/f and Got1ΔLysM mice following infections with S. aureus or Candida albicans (C. albicans) (Fig. 4f, g). Together, our findings reveal that Got1 depletion decreases macrophage ROS production without impacting the ability to combat pathogens effectively.

Fig. 4: Got1 deficiency decreases ROS production in BMDMs without affecting pathogen defense.
figure 4

a Schematic of ROS detection in macrophages. b ROS production in BMDMs. Prior to ROS detection, BMDMs were pre-treated with AOAA for 24 h, then treated with HKCA (HKCA: BMDM = 1:1) for 1 h. Got1f/f mice, n = 3 biologically independent experiments; Got1ΔLysM mice, n = 3 biologically independent experiments. c Phagocytosis of BMDMs. FITC labeled-HKCA was utilized for phagocytosis assay with BMDMs. d CFU Count of S. aureus and E. coli in BMDMs-In-Vitro-Killing Assay. e Schematic of infection design. f, g Percentage of mice survival. ns, p > 0.05, not significant, *p < 0.05, **p < 0.01, unpaired, p values were determined by two-tailed Student’s t test in b–d and log-rank test was used for survival comparison in f and g. Data are representative of three independent experiments (mean ± SD).

Full size image

The differentiation of M2 macrophages is independent of Got1

M2 macrophages rely on an intact tricarboxylic acid (TCA) cycle and are distinct from their pro-inflammatory counterparts13,15,16,41. α-ketoglutarate (αKG) plays a crucial role in the alternative (M2) activation of macrophages21. Previous metabolomic studies have indicated that the deficiency of GOT1 results in a significant reduction in intracellular αKG levels (Xu et al.). Motivated by these findings, we investigated the role of Got1 in IL-4-induced M2 macrophage differentiation. We assessed mRNA samples from BMDMs derived from Got1f/f and Got1ΔLysM mice treated with IL-4 for 24 h as the schematic shown in Fig. 5a. The mRNA expression levels of M2 markers, including Chil3, Arg1, and Retnla were evaluated42,43. Remarkably, Got1 deficiency in BMDMs did not influence M2 macrophage differentiation, as evidenced by the unchanged mRNA expression of these marker (Fig. 5b–d). Consistently, PMs from Got1ΔLysM mice exhibited no differences in M2 marker expression upon IL-4 stimulation (Fig. 5e–g). Therefore, we pretreated BMDMs or PMs derived from Got1f/f and Got1ΔLysM mice with AOAA (Fig. 5b–g) or PF-04859989 (Supplementary Fig. 8a–f), followed by IL-4 induction to drive BMDM or PMs towards M2 differentiation. Interestingly, our results demonstrate that both AOAA and PF-04859989 exert inhibitory effects on M2 differentiation in BMDMs or PMs, although AOAA did not exhibit inhibition of Chil3 and Arg1 in BMDMs (Fig. 5b, c). Moreover, overexpression of Got1 in macrophages, as observed in Got1stop/+ or Got1stop/+; Lyz2-Cre mice, did not affect the mRNA expression levels of M2 markers (Fig. 5h–j). These data collectively establish that the differentiation of M2 macrophages is independent of Got1.

Fig. 5: The differentiation of M2 macrophages is independent of Got1.
figure 5

a Schematic of the differentiation of M2 macrophages. b–d Chil3, Arg1, and Retnla mRNA levels in BMDMs, derived from Got1f/f and Got1ΔLysM mice, pretreated with 10 mM AOAA for 24 h, then treated with IL-4 (20 ng/mL) for 24 h. n = 3 biologically independent experiments. e–g Chil3, Arg1, and Retnla mRNA levels in PMs, derived from Got1f/f and Got1ΔLysM mice, pretreated with 10 mM AOAA for 24 h, then treated with IL-4 (20 ng/mL) for 24 h. n = 3 biologically independent experiments. h–j Chil3, Arg1, and Retnla mRNA levels in BMDMs, derived from Got1stop/+ and Got1stop/+; Lyz2-Cre mice, treated with IL-4 (20 ng/mL) for 24 h. n = 3 biologically independent experiments. ns, p > 0.05, not significant, *p < 0.05, **p < 0.01, ***p < 0.001, unpaired, p values were determined by two-tailed Student’s t test. Data are representative of three independent experiments (mean ± SD).

Full size image

LPS induced immune tolerance in macrophages is independent of Got1

Sepsis-induced immunosuppression carries risks for adverse outcomes7,44. Metabolic reprogramming towards aerobic glycolysis, known as the Warburg effect, is observed in human peripheral blood monocytes following LPS stimulation, often associated with mitochondrial damage45,46. To understand whether Got1 is involved in LPS-induced immune tolerance, we employed an in vitro model using BMDMs derived from Got1ΔLysM or Got1stop/+; Lyz2-Cre mice (Fig. 6a). Remarkably, neither Got1 knockout nor overexpression in BMDMs alters the protein (IL-6, TNF-α) levels (Fig. 6b, c, f, g), lactate levels (Fig. 6d) and NO levels (Fig. 6e) of LPS-induced immune tolerance. Furthermore, we pretreated BMDMs with AOAA (Fig. 6b–e) or PF-04859989 (Supplementary Fig. 9a–d). The results reveal that both AOAA and PF-04859989 exhibit inhibitory effects on the production of IL-6 by BMDMs upon the single LPS stimulation, with no impact on TNFα production. Additionally, we unexpectedly observed that pretreatment with AOAA slightly enhances the production of IL-6 and TNFα upon the LPS retreatment (Fig. 6b, c). These results firmly establish that Got1 does not play a role in the inflammatory response underlying LPS-induced immune paralysis in macrophages.

Fig. 6: LPS induced immune tolerance in macrophages is independent of Got1.
figure 6

a Schematic of immune tolerance induced with LPS. b, c IL-6, TNFα protein levels in the supernatant of Got1f/f and Got1ΔLysM mice derived BMDMs were determined by ELISA. Before LPS retreatment, BMDMs were pretreated with 10 mM AOAA for 24 h, n = 3 biologically independent experiments. d Lactate production level in BMDMs stimulated with LPS (100 ng/mL) retreatment for 24 h. e NO production level in BMDMs stimulated with LPS (100 ng/mL) retreatment for 24 h. f, g IL-6, TNFα protein levels in the supernatant of Got1stop/+ and Got1stop/+; Lyz2-Cre mice derived BMDMs were determined by ELISA in. n = 3 biologically independent experiments. ns, p > 0.05, not significant, *p < 0.05, **p < 0.01, ***p < 0.001, unpaired, p values were determined by two-tailed Student’s t test. Data are representative of three independent experiments (mean ± SD).

Full size image

Discussion

The intricate interplay between macrophage activation and metabolic reprogramming constitutes a pivotal axis governing their dynamic engagement in immune responses. Within this intricate network, the role of Got1 is less explored and warrants further investigation. Our study endeavors to discern the multifaceted implications of Got1 in macrophage polarization and function.

The perturbations within the TCA cycle within M1 macrophages, characterized by dual disruptions at the IDH and SDH nodes, underpin citrate and itaconic acid accumulation, concomitant with heightened succinate levels. This sequential perturbation amplifies the aspartate-argininosuccinate shunt, consequently invigorating the urea cycle and facilitating NO production10,16,47. Of note, Abhishek K. Jha et al. demonstrated that treating macrophages with AOAA can inhibit the pro-inflammatory phenotype of M1 macrophages by suppressing GOT116, leading to reduced IL-6 and NO production. Correspondingly, our findings substantiate AOAA-mediated suppression of M1 differentiation, culminating in concordant diminution of IL-6 protein levels and NO release within BMDMs following LPS stimulation. This concurrence underscores the regulatory role of the AASS pathway in the inflammatory response, albeit in another way (Fig. 1b–d).

AOAA has gained widespread recognition as a selective GOT1 inhibitor in various contexts16,34,48,49,50,51. Nonetheless, our investigation transcends the conventional confines of AOAA’s influence, shedding light on its potential multifaceted effects that extend beyond GOT1 inhibition. Notably, the study by Xu et al. in 2017 unveils the pivotal role of Got1 in contributing to the dynamic equilibrium between Th17 and Treg subsets49. Given that Got1 is critical in the conversion of glutamate to α-ketoglutarate (α-KG), AOAA’s modulation of GOT1 activity attains paramount significance. Crucially, inhibition of GOT1 by AOAA counters the suppressive impact exerted by 2-hydroxyglutarate (2-HG) on Foxp3 expression, potentially facilitating Treg differentiation49. Conversely, the further investigation by Xu et al. broadens the implications of AOAA, transcending its traditional role as a specific GOT1 inhibitor35. Their meticulous exploration, utilizing Got1f/f; CD4-Cre mice, introduces intriguing complexities by revealing divergent outcomes in T cells exposed to AOAA treatment. This intriguing finding propels an in-depth exploration into AOAA’s comprehensive mechanisms of action, raising pertinent questions regarding the scope of its specificity as a GOT1 inhibitor. Our own data, utilizing both AOAA and PF-04859989 to inhibit GOT1 activity in both Got1f/f and Got1ΔLysM BMDMs, reveal that both inhibitors effectively inhibit cytokine production regardless of Got1 expression. This suggests a pleiotropic effect of current GOT1 inhibitors. Therefore, further research is warranted to develop more specific GOT1 inhibitors for studying GOT1 effects using inhibitors accurately.

Crucially, the viability of BMDMs remained unaltered upon AOAA treatment16. This intervention induced significant reduction in cytokine and NO production within M1 macrophages, a trend consistent across our observations in our Got1f/f and Got1ΔLysM mice (Fig. 1b–d). This intriguing outcome hints at the potential participation of the AASS pathway in this modulation, albeit leaning more towards an outcome rather than a causal factor. The precise cellular target of AOAA’s influence in macrophages prompts intricate considerations. Our study primarily observed a decrease in the generation of reactive oxygen species (ROS) in macrophages, yet this phenomenon appeared inconsequential to the overall antimicrobial defense exhibited by the mice. Such duality underscores an intricate compensatory interplay between Got1 and Got2. Notably, we gauged the mRNA expression of Got2 to ascertain any compensatory response in the absence of Got1 in macrophages, revealing no overt alterations (Supplementary Fig. 10). Moreover, we knocked down Got2 in BMDMs using siRNA. The experimental results were not surprising: knocking down Got2 did not affect the production of inflammatory factors in LPS-stimulated macrophages (Supplementary Fig. 3). Furthermore, our observations underscore the limited involvement of Got1 in IL-4-induced M2 macrophages (Fig. 5) and its role in LPS-induced immune tolerance (Fig. 6).

In summary, our investigations collectively unravel that the anti-inflammatory effects elicited by AOAA in macrophages manifest independently of Got1. Consequently, the intrinsic implication of Got1 in shaping the broader panorama of macrophage inflammatory responses appears to be of a lesser magnitude.

Methods

Generation of animals

The female or male mice used in our study were all C57BL/6J strain, aged between 6 and 12 weeks. These mice were housed in a controlled environment of the Laboratory Animal Center at Xiamen University, a facility that adheres to SPF-level standards and maintains a regulated day-night light cycle. Got1f/f and Lyz2-Cre mice were procured from Gempharmatech Co., Ltd, whereas Got1stop/+ mice were generated by Xiamen University Animal Center. To ensure experimental consistency, all mice included in our study were carefully selected to be sex and age-matched, originating from the same litters. All mouse experiments were conducted in accordance with the guidelines and regulations set forth by the Xiamen University Laboratory Animal Center and were approved by the Institutional Animal Care and Use Committee. We have complied with all relevant ethical regulations for animal use.

Generation and manipulation of BMDMs

For BMDMs differentiation, bone marrow cells were obtained from the femur and tibia of 6-12-week-old C57BL/6J mice and cultured in complete DMEM medium containing 10% FBS and 40 ng/μL M-CSF (Novoprotein, CD34). On day 3, half of the culture volume was replenished with complete medium containing 40 ng/μL M-CSF, and the cells were cultured until day 7 to yield BMDMs.

For LPS-induced pro-inflammatory macrophages polarization, 105 BMDMs per well were seeded in 96-well-plate, and then stimulated with 100 ng/mL LPS (Lipopolysaccharides from E. coli, Invivogen # tlrl-pb5lps). After 4 h of LPS stimulation, cells were lysed for mRNA extraction. Culture supernatants were collected 24 h post-LPS stimulation to assess cytokine production by ELISA.

For M1 macrophage polarization, 105 BMDMs per well were seeded in a 96-well-plate and treated with 20 ng/mL LPS + 100 ng/mL IFNγ (Novoprotein, CM41). Cell lysis and culture supernatants were collected 24 h after LPS + IFNγ stimulation.

For M2 macrophage polarization, 105 BMDMs per well were seeded in a 96-well plate and treated with 20 ng/mL IL-4 (Novoprotein, CK74). Cell lysis was collected 24 h after IL-4 treatment.

Regarding HKCAs treatment, 105 BMDMs per well were seeded in a 96- well-plate, and HKCAs were added at a comparable quantity to the BMDM cells. Cell lysis was collected 4 h after HKCAs stimulation for mRNA extraction, and culture supernatants were collected 24 h after HKCAs stimulation for ELISA analysis.

Isolation and cultivation of peritoneal macrophages

To obtain peritoneal macrophages, we intraperitoneally injected 3 mL 3% thioglycolate per mouse to stimulate the generation of an ample population of PMs in mice. After a period of 3-4 days, we injected 4 mL of PBS containing 1% FBS into the mouse peritoneal cavity for harvesting the PMs. After gently massaging the abdomen and incubating on a shaker for 5 min, we collected peritoneal fluid using a syringe, repeating the process twice. The fluid was then centrifuged at 1500 g for 5 min at 4 °C, and the supernatant was discarded. If necessary, red blood cells were lysed with ACK buffer. For ACK lysis (1.5 M NH4Cl, 100 mM KHCO3, 10 mM EDTA-2Na), 3 mL of ACK buffer was added to each sample and incubated for 3 minutes at room temperature, followed by termination with PBS in volumes greater than twice that of ACK. After centrifugation, peritoneal cells were plated in cell culture dishes. Following a 6–8-h incubation, non-adherent cells were removed by PBS washing, leaving adherent peritoneal macrophages. These macrophages were subsequently detached, counted, and seeded in plates for subsequent experiments.

Real-time RT-PCR

The mRNA extraction from cells was performed as describe in reference20. BMDMs were treated as the requirement, followed by treatment with 200 μL of RNA lysis buffer. Magnetic beads conjugated with oligo-dT18 were added for mRNA extraction. Subsequently, the extracted mRNA underwent reverse transcription using a mixture of dNTPs (Beyotime, D7366), oligo-dT, RNA transcriptase (Accurate Biology, AG11605) and RNase inhibitor (Accurate Biology, AG11608) to yield cDNA. The obtained cDNA was employed for real-time PCR analysis with 2× SYBR (Accurate Biology, AG11701). The relative mRNA expression levels were assessed utilizing the 2-ΔCT method with Ct values, employing B2m as the endogenous reference gene. The primer sequences are available in Supplementary Table 2.

Western blot

The cultured medium of BMDMs was removed and cells were washed with PBS. BMDMs were then lysed by RIPA buffer containing a protease inhibitor cocktail. After that, 5× SDS-loading buffer was added and cell lysate samples were boiled at 95 °C for 5 min. Subsequently, 10 μL of the sample was carefully loaded into the wells of an SDS-PAGE gel for protein separation according to their molecular weights. The separated proteins were then transferred onto a Polyvinylidene fluoride (PVDF) membrane. The PVDF membrane was meticulously blocked by immersion in a solution of 5% non-fat milk for 1 h at room temperature. Following blocking, the PVDF membrane was incubated overnight at 4°C with the primary antibody against GOT1 (1:1000 dilution, Abclonal A11363), GOT2 (1:1000 dilution, Abclonal A19245) and β-Actin (1:1000 dilution, Abcam ab8226). The subsequent day, the primary antibody was carefully removed, and the PVDF membrane was gently washed three times with PBST (Phosphate Buffered Saline with 1% Tween-20). The PVDF membrane was then subjected to incubation at room temperature for 1 h with the secondary antibody (1:5000 dilution, SAB L3012), followed by another three washes with PBST. For protein visualization, HyperSignal ECL kit (4A Biotech Co. Ltd, 4AW012-500) was used and Chemiluminescent Imaging and Analysis System (ChampChemiTM Professional+) was used for images analyzation.

ELISA

105 cells per well were seeded in a 96-well-plate. Pro-inflammatory macrophage differentiation was induced by adding either 100 ng/mL LPS or 105 HKCA for 24 h. For M1 macrophage polarization, a combination of 20 ng/mL LPS and 100 ng/mL IFNγ was applied for 24 h. After this incubation, culture supernatants were collected from each well for cytokine level assessment. IL-6 (Invitrogen; 88-7064-88) and TNFα (Invitrogen; 88-7324-88) levels were quantified following the respective ELISA kit instructions.

Lactate and NO

For lactate level detection, the collection of cell culture supernatant followed the same procedure as the preparation before conducting the ELISA assay. Subsequently, the cell culture supernatant was diluted with PBS and mixed with a mixture containing lactate oxidase (Sigma L0638), Amplex Red (Alfachem 119171-73-2), and HRP in a 1:1 ratio. The mixture was incubated for approximately 10–20 min at room temperature. Fluorescence values were then measured at an excitation wavelength of 528 nm and an emission wavelength of 590 nm.

For NO level detection, the collection of cell culture supernatant followed the same procedure as the preparation before conducting the ELISA assay. In the cell culture, NO production results in the formation of nitrite ions through reactions with oxygen and water. These nitrite ions further react with sulfanilamide to create a diazonium salt, which can react with N-1-naphthylethylenediamine dihydrochloride. We prepared a mixture by combining Solution A (ethylenediamine dihydrochloride) and Solution B (sulfanilamide) in a 1:1 ratio. Subsequently, we mixed this mixture with the cell culture medium in a 1:1 ratio to facilitate the reaction. The level of nitrite ions serves as an indicator of NO production. Finally, we measured the absorbance of the reaction solution at 540 nm.

ROS

Prior to utilization, BMDMs were washed with either PBS or FBS-free DMEM twice. For Got1ΔLysM mice, 5 ×105 BMDMs were seeded per well in 96-well-U-shaped-bottom-plate and 5 μM DCF-DA working solution was added in each well, followed by incubation at 37 °C in the dark for 30 min. Subsequently, the cells were washed twice with PBS, resuspended by FACS buffer. The Fluorescence intensity was measured by CytoFLEX Flow Cytometer. For Got1stop/+; Lyz2-Cre mice, 2.5 ×104 BMDMs were seeded per well in 96-well cell culture plate. After 1.5 h, each well received an amount of HKCA double the number of cells. Phagocytosis occurred for 30 min, following which HKCA was discarded. Subsequently, 200 uL of reaction solution, comprising 5 uM Amplex Red and 0.1 U/mL HRP, was added to each well. The reaction was carried out at room temperature in a light-avoidant environment, and fluorescence measurements were taken.

Pathogen in vitro killing assay and in vivo infection

S. aureus, E. coli and C. albicans were cultivated to an optical density (OD600) of approximately 0.8 following overnight incubation at 37 °C, 220 rpm for 12 h. For S. aureus and E. coli, we calibrated an OD600 reading of 0.5 to correspond to 2.5 ×108 CFU. For C. albicans, CFU were determined using a serial dilution method, followed by manual counting under a microscope.

For in vitro killing assay, 2 ×105 BMDMs per well were seeded in 24-well-plate. Next, S. aureus was added into BMDMs at multiplicity of infection (MOI) of 1, while E. coli was added into BMDMs at MOI of 10. The cultures were carried out without antibiotics. After a 30-min incubation for phagocytosis, extracellular bacteria were removed by washing with PBS containing 400 ng/mL gentamicin. Then, we supplemented them with DMEM complete medium containing 400 ng/mL gentamicin and allowed the BMDMs to continue incubating for an additional 2 h. After this period, we washed the cells with PBS, lysed them with sterile water, and released the intracellular bacteria. The cell lysates were then stepwise diluted using PBS and subsequently plated and cultured onto LB agar plates at 37 °C overnight.

In our in-vivo infection experiments, S. aureus and C. albicans were cultivated as above. S. aureus infection was administered intraperitoneally (i.p.) with 5000 CFU per mouse. C. albicans infection was administered intravenously (i.v.) with 2 ×105 CFU per mouse. The survival of the mice was meticulously monitored over a 7-day period.

Statistics and reproducibility

All statistical samples were carried out in accordance with the descriptions provided in the main text. All data were analyzed with the two-tailed Student’s t test method by GraphPad Prism software. All quantified data are represented as mean ± standard deviation (SD). “*” indicates P-value less than 0.05 and “**” indicates P-value less than 0.01, “***” indicates P-value less than 0.001, “****” indicates P-value less than 0.0001, while “ns” indicates P-value more than 0.05 and have no significant difference.