All experiments were approved by the Administration Office of Laboratory Animals of South China Agricultural University. ICR female mice (6–8 weeks old) were housed with free food and water in a room with standard day/night cycle conditions. Piglets were purchased from Hunan New Wellful Co., Ltd. (Changsha, China). Antibodies against IL-1β (12426s), mammalian target of rapamycin (mTOR; 2972s), p-mTOR (5536s), and HIF-1α (14179s) were purchased from Cell Signaling Technology (Danvers, MA, USA). Antibodies against IκB kinase (IKK; Sc-7607), p-IKK (sc-293135), p-IκB (Sc-8404), and apoptosis-associated speck like protein containing a caspase recruitment domin (ASC; Sc-514414) were purchased from Santa Cruz Biotechnology, Inc. (Dallas, TX, USA). Antibodies against IκB (51066-1-AP), p65 (10745-1-AP), and β-actin (60008-1-Ig) were purchased from Proteintech (Rosemont, IL, USA). The antibody against p-p65 (bs-0982R) was purchased from Bioss (Beijing, China). Antibodies against Nod-like receptor protein 3 (NLRP3; ab214185) and caspase-1 (ab179515) were purchased from Abcam (Cambridge, UK).

Peritoneal macrophages were isolated as in our previous study (10). Mice were injected intraperitoneally with 2 ml of 4% sterile thioglycolate, and peritoneal macrophages were collected via peritoneal lavage with 5 ml cold phosphate-buffered saline (PBS). Peritoneal lavage was spun down for 5 min at 1,200 rpm and resuspended in Dulbecco’s modified Eagle’s medium (DMEM) containing 10% fetal bovine serum (FBS) and 1% penicillin–streptomycin. Non-adherent cells were removed 3 h later through washing twice with sterile PBS, whereas adherent cells were cultured in complete DMEM.

Peritoneal macrophages were stimulated with 1 μg/ml lipopolysaccharide (LPS) (L2630; Sigma, St. Louis, MO, USA) plus 20 ng/ml interferon-γ (IFN-γ) (315-05; PeproTech, Cranbury, NJ, USA) for 15 h to induce M1 polarization. In some experiments, Asp, asparagine, cytosine, or deoxycytidine was added to DMEM in the indicated dosages (0.1, 1, or 10 mM for Asp or asparagine and 100 μM for cytosine or deoxycytidine). PX-478, MCC-950, VX-765, and IKK16 were purchased from Selleck Chemicals (Houston, TX, USA). Dactinomycin and cycloheximide were purchased from MedChemExpress (Monmouth Junction, NJ, USA). The cells were treated with PX-478 (10 μM), MCC-950 (10 μM), VX-765 (20 μM), dactinomycin (2 μg/ml), or cycloheximide (10 μM) for 15 h when treated with LPS and IFN-γ in order to inhibit HIF-1α, NLRP3, and caspase-1 transcription or translation, respectively. The cells were treated with the IKK inhibitor IKK-16 (10 μM) for 20 min before treated with LPS and IFN-γ.

Six weeks old C57 female mice were infected by oral gavage with 200 μl of PBS containing approximately 1 × 10⁸ colony forming units (CFUs) of Citrobacter rodentium (DBS100). The colons of all mice were collected for further analysis at 10 days post-infection. The colons were homogenized in PBS and then serial diluted and inoculated on selective MacConkey agar. Bacteria were enumerated after growth overnight at 37°C.

A total of 44 six-week-old piglets were assigned into four groups: Con group (basal diet), 0.005% Asp group (basal diet with 0.005% Asp supplementation), 0.05% Asp group (basal diet with 0.05% Asp supplementation), and 0.5% Asp group (basal diet with 0.5% Asp supplementation). The piglets were kept individually in an environment-controlled facility with hard plastic slatted flooring and with free access to food and drinking water. Five piglets per group were sacrificed on day 70 and six piglets per group sacrificed on day 142. The tissues from the jejunum, ileum, and colon of piglets were collected immediately and stored at −80°C.

For the bactericidal activity assays, 1 × 10⁶ cells/well macrophages were incubated with 1 × 10⁷ ETEC (Escherichia coli strain SEC470) for 3 h on a six-well chamber slide. After infection, the macrophages were washed with cold PBS to stop additional bacterial uptake, and the extracellular bacteria were completely killed with 300 μg/ml gentamicin for 30 min. After the clearance of gentamicin with PBS, 500 μl 1% Triton X-100 was used for cell lysis, and 100 μl diluted lysate was inoculated on Luria–Bertani (LB) agar plates and incubated at 37°C in a CO₂ incubator overnight to count and calculate the number of CFUs.

RNA was isolated from peritoneal macrophages using the EZ-press RNA Purification Kit (B0004D) purchased from EZBioscience^® (Roseville, MN, USA). RNA was quantified and 500 ng of RNA was reverse transcribed using the Color Reverse Transcription Kit (A0010CGQ; EZBioscience^®). Real-time PCR was performed using 2× Color SYBR Green qPCR Master Mix (A0012-R2, EZBioscience^®) on the QuantStudio 6 Real-Time PCR System (Thermo Fisher Scientific, Waltham, MA, USA). Fold change was assessed using the 2^−ΔΔCt method with β-actin. The primers used in this study are shown in Supplementary Table S1.

Western blot was performed according to our previous studies (10, 25). The cells were lysed in ice-cold RIPA for 10 min. The bicinchoninic acid (BCA) protein assay kit (Beyotime Biotechnology, Shanghai, China) was used to quantify the protein concentration. Then, equal amounts of protein were subjected to 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). The proteins were transferred onto polyvinylidene difluoride (PVDF) membranes (Bio-Rad, Hercules, CA, USA), and blocked with 5% bovine serum albumin (BSA) Tris/Tween-buffered saline buffer (TBST) for 1 h. Then, the proteins were incubated with primary and secondary antibodies and visualized using a chemiluminescent reagent. The signal intensity was digitally quantified and normalized to actin protein abundance.

IL-1β and tumor necrosis factor alpha (TNF-α) on supernatants obtained from cell cultures were quantified by ELISA (#KE10003 and #KE10002; Proteintech) based on protocol from the manufacturer.

The intracellular and supernatant levels of the metabolites of macrophages were quantified through liquid chromatography–mass spectrometry (LC-MS), as in our previous study (25). Approximately 1 × 10⁷ cells/well macrophages were plated in six-well plates, treated as indicated, and collected for analysis. Briefly, the cells were washed with cold PBS and collected with a cell scraper (Thermo Fisher Scientific). Then, 1 ml extract solution (50% methanol acetonitrile) was added into each sample. The samples were homogenized for 2 min, sonicated at 35 Hz in ice water for 10 min, and centrifuged at 14,500 rpm for 15 min at 4°C. The supernatant was dried in a vacuum concentrator at 50°C, and re-dissolved with 200 μl 50% methanol water. Subsequently, the samples were homogenized for 2 min, sonicated at 35 Hz in ice water for 10 min, and centrifuged at 14,500 rpm for 15 min at 4°C. Then, the supernatant was transferred into a fresh glass vial for LC-MS analysis.

All data were analyzed and prepared with GraphPad Prism v.8.0. All data displayed in graphs are expressed as the mean ± SEM. A t-test was used to analyze the data with a Gaussian distribution and had equal variance. Unpaired t-test with Welch’s correction was used to analyze the data with a Gaussian distribution but without equal variance. A non-parametric test (Mann–Whitney U test) was used to analyze the data that were not normally distributed. One-way ANOVA and Dunnett’s multiple comparisons were used to analyze the data that have more than two groups, with a Gaussian distribution, and had equal variance. Kruskal–Wallis and Dunn’s multiple comparisons tests were used to analyze the data that were not normally distributed. The D’Agostino–Pearson omnibus normality test and the Kolmogorov–Smirnov test were used to analyze the Gaussian distribution of the data. The homogeneity of variance test or the Brown–Forsythe test was used to analyze the variance of data. A significant difference was defined as p < 0.05.

Macrophages reprogram their metabolism to instruct their effector functions, such as producing mediators and cytokines (26). With LC-MS, our previous study showed a remarkable difference in the cellular metabolism between resting macrophages (M0) and macrophages stimulated with LPS plus IFN-γ (27). We reanalyzed the differential metabolites and found that most of them were enriched in glycolysis, tricarboxylic acid (TCA) cycle, and amino acid metabolism (Figure 1A). Notably, half of them could be incorporated into Asp and nucleotide metabolism (Figure 1B), including Asp, asparagine, purines, and pyrimidines (Figures 1C, D, F, G). Particularly, the levels of the majority of these metabolites in M1 macrophages were decreased compared to those in M0 macrophages, except for guanosine and adenosine (Figures 1C, D). Besides, the gene expressions of the rate-limiting enzymes in purine and pyrimidine synthesis were also decreased, in addition to Cmpk2 and Impdh1 (Figure 1E). Also, the messenger RNA (mRNA) expressions of the genes related to Asp synthesis and transportation were changed (Supplementary Figure S1). Collectively, M1 macrophage polarization shows remarkable change in cellular metabolism, especially Asp metabolism.

Asp is a key substrate in nucleotide and protein synthesis, but cellular Asp availability is largely dependent on mitochondrial respiration (24). However, M1 macrophages are characterized as a dysfunctional electron transport chain (28), which might result in insufficient Asp metabolism and inhibit the synthesis of pro-inflammatory cytokines. Thus, we then explored the influence of Asp on the polarization of M1 macrophages. Our results showed that exogenous Asp supplementation promoted the secretion of IL-1β and TNF-α (Figure 2A and Supplementary Figure S2A). However, Asp had little effect on the mRNA expressions of IL-1β and TNF-α (Supplementary Figure S2B). As IL-1β has been shown to be directly involved in the production of multiple pro-inflammatory mediators (29), and is specifically secreted by inflammatory macrophages (30), we thus focused on the role of Asp in the IL-1β production of M1 macrophages in the subsequent study.

Nuclear factor kappa-B (NF-κB) and the NLRP3 inflammasome are two key pathways involved in M1 polarization (31, 32). Moreover, mTOR and HIF-1α signaling are also associated with macrophage polarization (10, 33, 34). Thus, we analyzed the activation of these pathways. Aspartate increased the protein levels of pro-IL-1β (31 kDa) and IL-1β (17 kDa) (Figures 2B, C). As expected, Asp significantly promoted the activation of mTOR, HIF-1α, and NLRP3 inflammasome (including NLRP3, ASC, and caspase-1), while it had little effect on the NF-κB pathway (Figures 2B, D). These results indicated that Asp promotes M1 polarization through the activation of these pathways. To validate this hypothesis, PX-478, MCC-950, and VX-765 were used to inhibit HIF-1α, NLRP3, and caspase-1 in Asp-treated M1 macrophages, respectively. These inhibitors significantly blocked the effect of Asp on IL-1β secretion (Figure 2E). Collectively, Asp-enhanced M1 macrophage polarization is largely dependent on the activation of HIF-1α and NLRP3 inflammasome.

Given that the metabolic output is closely associated with macrophage polarization (35), M1 macrophage polarization showed remarkable change in Asp metabolism (Figure 1). Thus, we asked whether Asp facilitates M1 macrophage polarization through altering intracellular metabolism. Here, we performed metabolomics to assay the intracellular metabolites in macrophages cultured with Asp. The cellular metabolites of M1 macrophages cultured with Asp-supplemented media were obviously different from those in M1 macrophages cultured with normal medium (Figure 3A and Supplementary Figures S2C, D). Similarly, the differential metabolites were also enriched in Asp metabolism, including Asp, asparagine, and nucleotides (Figures 3B, C). However, Asp had little effect on the mRNA expressions of the enzymes for nucleotide metabolism (Supplementary Figure S2E). Intriguingly, Asp significantly affected the amino acid metabolism and the aminoacyl-tRNA biosynthesis (Figures 3D, E and Supplementary Figure S2F). Aminoacyl-tRNA biosynthesis is an essential biological process responsible for charging amino acids to their cognate transfer RNAs (tRNAs) in order to provide the substrates for global protein synthesis (36), suggesting that Asp promotes the secretion of IL-1β through translation. Thus, we added dactinomycin and cycloheximide to inhibit transcription and translation, respectively. Both dactinomycin and cycloheximide were sufficient to block the Asp-mediated IL-1β secretion (Figure 3F). Collectively, these results suggest that Asp promotes the production of IL-1β through metabolic remodeling.

Aspartate is the key precursor for the synthesis of nucleotides and asparagine (22). Moreover, this study revealed that Asp metabolites were decreased in M1 macrophages (Figure 1B), while supplementation of Asp reversed this (Figure 3G). Therefore, we explored whether Asp regulates macrophage polarization through its derivatives. Strikingly, the targeted LC-MS analysis showed that intracellular nucleoside was decreased (Figure 1D), but the supernatant cytosine and deoxycytidine were increased in M1 macrophages (Supplementary Figure S3A). In addition, Asp also increased the supernatant cytosine and deoxycytidine (Supplementary Figure S3B). Then, we determined whether cytosine and deoxycytidine could contribute to the polarization of M1 macrophages. However, cytosine or deoxycytidine was not able to increase the abundance of IL-1β and caspase-1 and the phosphorylation of IKK (Supplementary Figures S3C, D).

We next explored the role of asparagine in M1 macrophages. Our results showed that 1 mM asparagine was sufficient to promote the production of IL-1β and TNF-α in M1 macrophages (Figures 4A–C and Supplementary Figures S4A, B). To determine whether asparagine promotes the production of IL-1β in M1 macrophages by the same mechanism as Asp, we examined the activation of the associated signaling pathways. Interestingly, asparagine promoted the activation of NF-κB, HIF-1α, and NLRP3 inflammasome (Figures 4D, E). Furthermore, IKK16 (an inhibitor of IκB) and VX 765 (an inhibitor of HIF-1α) significantly reduced the production of IL-1β in M1 macrophages cultured with asparagine (Figure 4F). Taken together, these data suggest that Asp promotes M1 macrophage polarization possibly through asparagine.

Considering that Asp is the precursor of asparagine, and asparagine supplementation increased the supernatant and intracellular levels of Asp and asparagine (Figure 5A), we investigated whether asparagine also affects the polarization of macrophages through intracellular metabolic remodeling. Similar to the effects of Asp, asparagine modulated the nucleotide and amino acid metabolism in M1 macrophages and induced the activation of aminoacyl-tRNA biosynthesis (Figures 5B–F and Supplementary Figures S5A–D). Also, the inhibition of transcription or translation significantly blocked the effect of asparagine on the production of IL-1β in macrophages (Figure 5G). These data support the hypothesis that asparagine modulates nucleotide and amino acid metabolism to support M1 macrophage polarization.

To identify whether Asp and asparagine affect the bactericidal activity of macrophages, we assessed their effects on the bactericidal activity of macrophages against ETEC. The number of engulfed bacteria in the macrophages was determined at 3 h post-infection. Aspartate or asparagine enhanced the bactericidal activity of macrophages (Figure 6A). Moreover, C. rodentium has been shown to induce the activation of the NLRP3 inflammasome in macrophages, and IL-1β plays a critical role in the defense against C. rodentium infection (37, 38). To identify the role of Asp in the clearance of C. rodentium, we gave mice drinking water supplemented with Asp at a dosage of 0.5 mg/ml for 2 weeks before they were orally infected with 10⁸ CFU/mouse. Although the body weight and the length of the colon were similar between the two groups (Supplementary Figures S6A, B), a lower bacterial burden was observed in the colon of mice receiving Asp (Figure 6B). Aspartate increased the baseline level of IL-1β in the colon, but no significant difference was observed between the two groups (Figure 6C). The possible explanation is that the colonic IL-1β analyzed is derived from a variety of immune cells [e.g., neutrophils (39) and macrophages (10)]; the effects of Asp on these cells are inconsistent. These results indicate that Asp promotes the bactericidal activity of macrophages in the context of bacterial infection.

Infants are susceptible to pathogen infection due to the immature development of their gastrointestinal tract and the immature function of their underdeveloped immune system, leading to diarrhea and even death (40). Thus, it is important to maintain the intestinal immune balance of a newborn host. Considering that macrophage polarization plays an important role in maintaining intestinal immune balance (41), we wanted to uncover whether Asp could alter the intestinal immune status of a newborn host. The piglet is similar to a human in terms of anatomy and physiology, and it is a more excellent animal model for experiments than rodents (42). Therefore, weaned piglets were used as the models to investigate the role of Asp in intestinal immune responses. As expected, diet supplementation with Asp enhanced the intestinal immune responses, especially the IL-1β expression in the jejunum, in a dosage- and time-dependent manner (Figures 6D, E and Supplementary Figures S6C, D). Taken together, Asp confers protection against intestinal bacterial infections by enhancing the macrophage-mediated inflammatory responses in vivo.