ACOD1 in immunometabolism and disease

Abstract

Immunometabolism plays a fundamental role in health and diseases and involves multiple genes and signals. Aconitate decarboxylase 1 (ACOD1; also known as IRG1) is emerging as a regulator of immunometabolism in inflammation and infection. Upregulation of ACOD1 expression occurs in activated immune cells (e.g., macrophages and monocytes) in response to pathogen infection (e.g., bacteria and viruses), pathogen-associated molecular pattern molecules (e.g., LPS), cytokines (e.g., TNF and IFNs), and damage-associated molecular patterns (e.g., monosodium urate). Mechanistically, several immune receptors (e.g., TLRs and IFNAR), adapter proteins (e.g., MYD88), ubiquitin ligases (e.g., A20), and transcription factors (e.g., NF-κB, IRFs, and STATs) form complex signal transduction networks to control ACOD1 expression in a context-dependent manner. Functionally, ACOD1 mediates itaconate production, oxidative stress, and antigen processing and plays dual roles in immunity and diseases. On the one hand, activation of the ACOD1 pathway may limit pathogen infection and promote embryo implantation. On the other hand, abnormal ACOD1 expression can lead to tumor progression, neurodegenerative disease, and immune paralysis. Further understanding of the function and regulation of ACOD1 is important for the application of ACOD1-based therapeutic strategies in disease.

Introduction

Pathogens are microorganisms, including viruses, bacteria, fungi, protozoa, and worms, that can cause various human diseases. Potential hosts have two lines of defense against microbial infections. The first line of defense is innate immunity (also known as natural immunity), which rapidly destroys invading pathogens, mainly through phagocytosis mediated by myeloid cells (e.g., neutrophils, monocytes, macrophages, and dendritic cells [DCs]).^1,2 The second line of defense is adaptive immunity (also known as acquired immunity), which relies on cytotoxic T cells or antigen-antibody responses mediated by lymphocytes (e.g., T cells and B cells)³. After successfully preventing an infection, the host will have an immune memory of the invading pathogen. If the same pathogen reappears, the host can fight the infection more effectively. In contrast, a dysfunctional immune system may cause inflammation and autoimmune diseases.

Immunometabolism describes changes in the interactions between immune and metabolic pathways during stress, which is a complex dynamic process involving many regulators or signals.^4,5 Among them, aconitate decarboxylase 1 (ACOD1, also known as immunoresponsive gene 1 [IRG1]), has attracted much attention as a multifunctional regulator of inflammation and infection.^6,7 ACOD1 expression is markedly upregulated by invading pathogens (e.g., Escherichia coli [E. coli], Mycobacterium tuberculosis [Mtb], and Zika virus [ZIKV]) and their derivatives (e.g., lipopolysaccharide [LPS] and CpG DNA).^8–12 In addition, cytokines (e.g., interferon beta [IFNB] and tumor necrosis factor [TNF]), hormones (e.g., progesterone and estradiol), and certain compounds (e.g., cycloheximide [CHX]) also trigger ACOD1 expression under some conditions.^8,13,14 Abnormal ACOD1 expression is associated with inflammatory responses, antibacterial processes, tumorigenesis, neurodegenerative changes, and embryo implantation.^15–19 Consequently, an impaired ACOD1 pathway affects metabolism-related innate immune signaling, leading to a variety of diseases.^20–22 In this review, we highlight the key features of the ACOD1 protein in immunometabolism and discuss its potential pathological roles in diseases. This knowledge can provide a new perspective for the development of effective treatments for ACOD1-related diseases.

Discovery of ACOD1

As stated above, the innate immune response is the first barrier against microorganisms, activating signaling pathways to trigger the secretion of cytokines or the expression of genes involved in immune responses.² This process relies on host pattern recognition receptors (PRRs) to recognize pathogen-associated molecular patterns (PAMPs), which are the components of microorganisms.^23,24 LPS is a typical PAMP in the outer membrane of gram-negative bacteria and has been widely used to study the mechanisms of innate immunity. In 1995, the Acod1 gene was first cloned as a 2.3-kb cDNA from a mouse macrophage cell line (RAW264.7 cells) stimulated with LPS (50 ng/ml).⁸ This original study showed that Acod1 transcripts could be detected as early as 1.5 h after stimulation, and the maximum level was observed at 4–6 h after LPS stimulation, with a 25-fold increase in induction.⁸ The human ACOD1 gene is located on chromosome 13q22.3 and is highly conserved among genes from other species, such as chimpanzees, dogs, cattle, mice, rats, chickens, zebrafish, Xenopus, and Aspergillus.^16,25–27 ACOD1 is thought to be an enzyme that is active in itaconate metabolism and is mainly found in mitochondria.^14,16,28 Recently, the crystal structure was analyzed, and eight active-site residues (Asp93, Thr97, His103, His159, Lys207, Lys272, His277, and Tyr318) were found to be critical for human ACOD1 function, whereas mutations in these residues could lead to a substantial loss of ACOD1 enzymatic activity and decreased itaconate production.²⁷ However, specific inhibitors targeting these ACOD1 sites have not yet been identified.

Induction of ACOD1

Infection with live pathogens can rapidly induce ACOD1 expression in vitro and in vivo (Table 1).¹⁶ For example, Acod1 mRNA expression is induced by Mtb infection in murine bone marrow-derived macrophages (BMDMs), with an approximately fivefold increase at 8 h after infection.⁹ At 48 h after whole-brain infeciton with ZIKV by intracranial administration, tissue homogenates were analyzed, and the expression of Acod1 mRNA was upregulated 16-fold compared to that of the control group.¹¹ These findings suggest that the upregulation of ACOD1 may be important even in the response to bacterial and viral infections.

Table 1.

Pathogens that induce ACOD1 expression

Pathogens	Host	Burden	Induction time	Fold change	Function	Animal model	Ref
Bacteria
Staphylococcus aureus	Blood from healthy human donors	1 × 106/ml	4 h	4	Sepsis biomarker	NA	12
Salmonella enterica serovar Typhimurium	NA	600 CFU		10–50	Antimicrobial activity	Zebrafish	16
Salmonella enterica serovar Enteritidis	Cecum Spleen	107 CFU 107 CFU	4 dpi 4 dpi	222 96.62	NA NA	S. Enteritidis-infected chickens	102–104
Escherichia coli	Blood from healthy human donors	4 ×103/ml	4 h	4	Sepsis biomarker	NA	12
Mycobacterium avium subspecies paratuberculosis	RAW264.7 cells J774A.1 cells BMDMs	NA	2–6 h	4	NA	NA	40
Mycobacterium smegmatis	RAW264.7 cells	NA	8 h	NA	NA	NA	40
Mycobacterium tuberculosis	WT, Tlr2−/−, Tlr4−/−, Myd88−/−, Tirap−/−, and Trif−/−	MOI 5	24 h	>4	NA	WT, Tlr2−/−/Tlr4−/−, and Myd88-/ C57BL/6 mice	9
	BMDMs	600 CFU	21 dpi	>3
	Lung
Borrelia burgdorferi spirochete	J774A.1 cells	NA	24 h	196.02	NA	NA	90
Burkholderia pseudomallei LPS mutant strain	RAW264.7 cells	MOI 2	4 h	NA	NA	NA	48
Legionella pneumophila	Lung BMDMs	1 × 106/40 µl in vivo	8 h	100 400	Antimicrobial activity	L. pneumophila-infected C57BL/6 J mice	28
Chlamydia pneumoniae	Lung ANA-1 cells	2.5 ×106 IFU MOI 10	3 dpi 8 h	20.1 9.1	NA	C. pneumoniae-infected C57BL/6 mice	46
Listeria monocytogenes	Spleen	0.1 × LD50	48 h	500	Mitochondrial localization	L. monocytogenes-infected C57BL/6 mice	14
Virus
Respiratory syncytial virus	A549 cells Lung	MOI 1 5 ×106 PFU/100 µl in vivo	36 h	6	Promote ROS production and immune lung injury	RSV-infected BALB/c mice	93
Zika virus	Primary neurons Brain	MOI 0.1 in vitro	24 h	>8	Antimicrobial activity	ZIKV-infected C57BL/6NJ and C57BL/6 J mice	11
		103 PFU/50 µl in vivo		>16
Marek’s disease virus	Spleen	103 PFU/200 µl	2–4 dpi	14	SNPs promote MDV susceptibility	MDV-infected chicks	89
Influenza A virus	Lung	2 ×103 FFU	48 h	8	NA	IAV-infected DBA/2 J mice	105
West Nile virus	Brain Primary neurons	102 PFU/50 µl in vivo	8 dpi	4	Antimicrobial activity	WNV-infected C57BL/6NJ mice	11
		MOI 0.001	24 h	>4
Vesicular stomatitis virus	PMs Lung Liver Spleen	MOI 0.1 5 × 107 PFU/g 5 × 107 PFU/g 5 × 107 PFU/g	8 h 24 h 24 h 24 h	>250 >100 >200 >400	Promote viral replication	VSV-infected C57BL/6 J mice	22
Herpes simplex virus 1	PMs Lung Liver Spleen	MOI 1 5 × 107 PFU/g 5 × 107 PFU/g 5 × 107 PFU/g	8 h 24 h 24 h 24 h	300 20 10 20	Promote viral replication	HSV-1–infected C57BL/6 J mice	22
Parasite
Toxoplasma gondii	Spleen Lung	20 cysts	12 dpi 12 dpi	2 10	Mitochondrial localization	T. gondii-infected C57BL/6 mice	14
Leishmania donovani	Splenic macrophages Spleen	5 Parasites per macrophage 1 ×106	28 dpi	>10 >100	NA	L. donovani-infected hamsters	32

Infection-mediated ACOD1 upregulation is related to the activation of Toll-like receptors (TLRs), a class of proteins that play a key role in innate immunity (Table 2). Many studies indicate that TLR4 is required for LPS (10 ng/ml to 10 µg/ml)- or lipid A (1 µg/ml)-induced ACOD1 expression in primary immune cells (e.g., human peripheral blood mononuclear cells [PBMCs], human PBMC-derived macrophages and DCs, murine peritoneal macrophages [PMs], murine BMDMs and murine bone marrow-derived dendritic cells) and cell lines (e.g., RAW264.7 cells, immortalized BMDMs, and THP1 [a human peripheral blood monocyte from acute monocytic leukemia] cells). In addition, TLR2 is responsible for lipoteichoic acid (LTA; 1 µg/ml)- or botulinum neurotoxin type A (BoNT/A; 1 nM or 5 nM)-induced Acod1 mRNA expression in ANA-1 cells, BMDMs, and RAW264.7 cells, whereas TLR9 facilitates CpG-DNA (1 µM)-induced Acod1 mRNA expression in BMDMs.^14,29 The effects of poly I:C, a TLR3 ligand, on ACOD1 expression are context-dependent. A high concentration (40 µg/ml, 24 h) of poly I:C induces Acod1 mRNA expression in BMDMs,³⁰ but a low concentration (1 µg/ml, 16 h) of poly I:C does not change ACOD1 expression in ANA-1 cells or BMDMs.¹⁴ As expected, TLR3 is required for high-concentration poly I:C-induced ACOD1 expression in macrophages. Although the administration of the TLR5 agonist KMRC011 in irradiated mice increases the expression of Acod1 mRNA in the small intestine,³¹ it has not been proven whether KMRC011 directly induces ACOD1 expression in cultured cells.³¹ It is also unclear whether other TLR members (e.g., TLR1, TLR6, TLR7, TLR8, TLR10, and TLR11) regulate ACOD1 expression in response to their corresponding ligands.

Table 2.

Reagents that induce ACOD1 expression

Compound	Target	Host cell	Induction time	Concentration	Fold change	Function	Animal Model	Ref
Infection
LPS	TLR4	RAW264.7 cells	4–6 h 6–9 h 6 h	50 ng/ml 10 ng/ml (055:B5) 10 ng/ml or 5 µg/ml	25 >400 NA	NA	NA NA NA	8, 50, 40, 53
		ANA-1 cells	16 h	100 ng/ml (055:B5)	50	NA	NA	14
		J774A.1 cells	6 h	5 µg/ml	NA	NA	NA	40
		BMDMs	6 h 6 h 16 h 24 h	5 µg/ml 10 ng/ml 100 ng/ml (055:B5) 100 ng/ml	NA NA 500 >50	NA Promote itaconate production NA Promote itaconate production	NA NA NA NA	40, 7, 14, 30
		A20−/− BMDMs	1 h 6 h	10 µg/ml (0111:B4)	3 1.44	NA	NA	53
		WT, Myd88−/−, Ifnar−/−, and Stat−/− murine DCs	4 h	100 ng/ml (055:B5)	NA	NA	NA	10
		Tlr4-overexpressing ovine blood macrophages	1 h	1 µg/ml	3	Anti-inflammatory activity	NA	45
		Human monocyte-derived macrophages	3 h	100 ng/ml (0111:B4)	125	Promote itaconate production	LPS-infected C57BL/6 mice	41
		Human PBMC-derived monocytes, macrophages, and DCs	6–9 h	10 ng/ml (055:B5)	>400	NA	NA	50
		Human PBMC-derived monocytes	4 h	10 ng/ml (055:B5)	25	NA	NA	15
Lipid A	TLR4	WT and Myd88−/− PMs	4 h 8 h	1 µg/ml	NA	NA	NA	44
IFNB	IFNAR	ANA-1 cells BMDMs	16 h 16 h 24 h	10 ng/ml 10 ng/ml 1000 U/ml	9 3 >4	NA NA Promote itaconate production	NA	14, 30
IFNG	IFNGR	ANA-1 cells	16 h	10 ng/ml	33.91	NA	NA	14
		Human PBMC-derived monocytes, macrophages, and DCs	6–9 h	10 ng/ml	>400	NA	NA	50
TNF	TNFR	ANA-1 cells	16 h	10 ng/ml	5.89	NA	NA	14
		Human PBMC-derived monocytes, macrophages, and DCs	6–9 h	50 ng/ml	>400	NA	NA	50
		A20−/− BMDMs	1 h 6 h	1000 U/ml	2.45 2.27	NA	NA	53
LTA	TLR2	ANA-1 cells BMDMs	16 h	1 µg/ml	9 50	NA	NA	14
BoNT/A	TLR2	RAW264.7 cells	4 h 2 h	1 nM 5 nM	28.04 319.27	NA	NA	29
Poly I:C	TLR3	BMDMs	24 h	40 µg/ml	>50	NA	WT and Ifnar1−/− C57BL/6 mice	30
CpG1668 DNA	TLR9	ANA-1 cells	16 h	1 µM	60	NA	NA	14
CpG DNA	TLR9	WT, Ifnar−/−, and Stat−/− murine DCs	4 h	0.1 µM	NA	NA	NA	10
KMRC011	TLR5	Mouse small intestine	Day 5 post TBI	0.2 mg/kg	>2	NA	Total-body irradiation (TBI) C57BL6/N mice
IL1B	IL1R	ANA-1 BMDMs	16 h	10 ng/ml	10 3	NA	NA	14
Pregnancy
Pregnancy		Uteri	Day 4 pregnancy	NA	> 75	Facilitate embryo implantation	CD1 female mice	13
Progesterone	PR	Uteri	Day 4 pregnancy	40 mg/kg 1 mg Additional 50 ng estradiol	>100 >50 >75	Facilitate embryo implantation	Ovariectomized CD1 female mice	13, 18
LIF		Uteri	12 h	5 µg 50 ng/ml	3.7 40	Facilitate embryo implantation	WT and Lif−/− MF1 female mice	33, 34
Other
CHX		RAW264.7 cells	4 h	50 µM	NA	NA	NA	8
CoPP		RAW264.7 cells	8–16 h	10 µM	8	Regulate A20 expression Anti-inflammatory activity	LPS-infected C57BL6 mice	38
Hemin		RAW264.7 cells	8–16 h	10 µM	6.5	Regulate A20 expression Anti-inflammatory activity	LPS-infected C57BL6 mice	38
CORM-2		RAW264.7 cells	8–16 h	20 µM	4.5	Regulate A20 expression Anti-inflammatory activity	LPS-infected C57BL6 mice	38
Blast overpressure		BV-2 cells		Pulse duration of 12 ms	3.7	NA	NA	39
MSU		Air pouch membrane	9 h	2 mg/ml	19.8	NA	BALB/c mice	37

In addition to TLR ligands, certain inflammatory cytokines are strong inducers of ACOD1 expression (Table 2). Acod1 mRNA is highly upregulated in ANA-1 cells following treatment with IFNB (10 ng/ml), IFNG (10 ng/ml), and TNF (10 ng/ml or 50 ng/ml).^14,28,32 IFNB (1000 U/ml) enhances LPS-induced Acod1 mRNA expression in BMDMs, and this process is blocked by the depletion of the type I interferon receptor (Ifnar), indicating that IFNAR is required for IFNB/LPS-induced ACOD1 expression.³⁰ Similar to IFNB, interleukin-1β (IL1B; 10 ng/ml) also enhances LPS-induced Acod1 mRNA expression.¹⁴ Thus, there is a synergistic effect between PAMPs (e.g., LPS) and cytokines (e.g., IFNB and IL1B) that mediates the production of ACOD1 during inflammation.

In addition to infection, the upregulation of ACOD1 has also been observed during embryo implantation, which is a highly organized process that involves interactions between a receptive uterus and a competent blastocyst. In the early stage of pregnancy, progesterone (P₄),^13,18 estradiol (E₂),¹³ and leukemia inhibitory factor (LIF)^33–35 play vital roles in embryo implantation. Microarray analysis^18,36 and northern blot analysis¹³ identified Acod1 as a P₄ response gene in the uterine epithelium. Although the mechanism of costimulation is unclear, P₄-induced Acod1 mRNA expression can be further enhanced by administering E2 in luminal epithelial cells.¹³ Intrauterine administration of recombinant LIF also induces a 3.7-fold increase in Acod1 mRNA expression.³³ Therefore, the upregulation of ACOD1 is implicated in P₄-, E₂-, and LIF-mediated implantation responses.

Small molecular compounds also have the ability to induce ACOD1 expression in immune cells (Table 2). As a protein synthesis inhibitor, CHX (50 µM) alone induces Acod1 mRNA expression in RAW264.7 cells at 4 h, suggesting that an interaction between protein synthesis and gene transcription may determine ACOD1 expression.⁸ Unlike cytokines, CHX fails to enhance LPS-induced Acod1 mRNA expression in RAW264.7 cells. Certain damage-associated molecular patterns, such as monosodium urate (MSU) or uric acid, can induce Acod1 expression in murine air pouch membranes,³⁷ indicating a potential pathological role of ACOD1 in gout. Other reagents or stressors, including the CO inducer CO-releasing molecule (CORM-2),³⁸ HO-1 inducer hemin or cobalt protoporphyrin IX (CoPP),³⁸ and even stimulated explosive overpressure treatments, have been shown to induce ACOD1 expression.³⁹ These findings further highlight ACOD1 as a stress-related inducible protein associated with inflammation, infection, and tissue damage.

Although ACOD1 expression can be elevated in most activated immune cells, the extent and peak time of the increase in ACOD1 will vary under different circumstances. The dynamic changes in Acod1 mRNA were first observed in LPS-stimulated macrophages, such as RAW264.7 cells, J774A.1 cells, and BMDMs. Acod1 mRNA can usually be detected after 2 h and reaches maximum expression levels ~6 h after LPS (10 ng/ml to 5 µg/ml) treatment.^6,8,40 Unlike mouse macrophages, different kinetics of ACOD1 mRNA are observed in LPS-stimulated human PBMCs, and the expression levels peak is at 18 h.²⁵ PAMPs, such as LPS, often induce higher ACOD1 expression than infection by live bacteria.⁴⁰ Compared to individual stimulation in ANA-1 cells and BMDMs, certain cytokines (e.g., TNF and IFNG) produce a synergistic effect on Acod1 mRNA induction in the absence of LPS.¹⁴ The degree of ACOD1 induction during live pathogen infection also varies from species to species. Details are listed in Table 1.

Several amino acid metabolic pathways may affect ACOD1 expression in response to inflammatory stimuli. Branched-chain aminotransferase 1 (BCAT1) has been identified as a positive regulator of ACOD1 expression in activated macrophages.⁴¹ The loss of BCAT1 inhibits LPS-induced ACOD1 expression and reduces oxygen consumption and glycolysis, thereby preventing the infiltration and activation of macrophages.⁴¹ Since BCAT1 is the enzyme responsible for reversible transamination of some amino acids, the mechanism by which BCAT1 regulates ACOD1 may be related to amino acid metabolism. Indoleamine 2,3-dioxygenase 1 (IDO1) is a rate-limiting metabolic enzyme that converts the essential amino acid tryptophan into kynurenines. The expression of Acod1 mRNA is upregulated in primary microglial cells from Ido1^−/− mice compared to wild-type (WT) mice,⁴² suggesting that tryptophan metabolism has a potential role in regulating ACOD1 expression. However, whether the altered ACOD1 levels in turn regulate the metabolism of these amino acids is still unclear.⁴³

Notably, ACOD1 is generally considered to be an inducible gene, and the baseline expression of ACOD1 is usually very low under normal conditions. However, ACOD1 can be detected in unstimulated human fetal PBMCs but not in human fetal tissue samples.²⁵ It is unclear whether this difference is due to experimental detection methods or physiological stress.

Modulation of ACOD1

TLRs

TLRs are the most important PRRs in host immune cells for recognizing PAMPs. As discussed above, TLR2, TLR3, TLR4, TLR5, and TLR9 have been shown to mediate ACOD1 expression in response to LPS, LTA, BoNT/A, CpG-DNA, poly I:C, or KMRC011 (Fig. 1).^{8,10,14,29–31} Depletion of Tlr4 or Tlr2 blocks ACOD1 expression in activated immune cells,^29,44 whereas overexpression of Tlr4 in ovine macrophages promotes Acod1 mRNA expression.⁴⁵ While TLR3- or TLR5-mediated Acod1 upregulation is observed under certain conditions, detailed signal transduction has not been reported. As a key adapter molecule, myeloid differentiation primary response 88 (MYD88), is recruited to the cytoplasmic portion of the TLR and triggers a downstream signal cascade.² CpG DNA-induced Acod1 mRNA expression is diminished in Myd88^−/− DCs, suggesting that TLR9-mediated ACOD1 upregulation is MYD88-dependent.¹⁰ Of note, Acod1 is one of the most upregulated genes in C. pneumoniae = Chlamydia pneumoniae infection, but this upregulation is not observed in Myd88^−/− macrophages, suggesting that MYD88 is essential in mediating ACOD1 expression.⁴⁶ Interestingly, MYD88 seems to not always be important for LPS-induced Acod1 mRNA expression in murine macrophages and DCs,^10,44 indicating a context-dependent role of MYD88 in the regulation of LPS-induced ACOD1 expression. In addition, Mtb infection-mediated Acod1 mRNA expression in BMDMs is independent of TLR2 and TLR4, as well as their downstream adapter proteins, such as MYD88, TIR domain-containing adapter protein (TIRAP), and toll-like receptor adaptor molecule 1.⁹ Possible signaling pathways that induce ACOD1 expression in MYD88-deficient cells may involve the activation of interferon regulatory factor 3 (IRF3),⁴⁴ nuclear factor-κB (NF-κB),⁴⁴ IFNAR,⁹ and signal transducer and activator of transcription 1 (STAT1).⁹ Thus, TLR-signaling pathway-mediated ACOD1 induction is intricate and largely depends on the type of host cell and stimulation.

Fig. 1.

Expression of ACOD1 in infection and immunity. a–c Many inflammatory stimuli (e.g., PAMPs, IFNs, TNF, and IL1B) can induce ACOD1 expression by activating receptors (e.g., TLRs, IFNAR, TNFR, and IL1R) and transcription factors (such as NF-κB, IRFs, and STATs) in a context-dependent manner

IRFs

The IRF family, which consists of nine members in mammals, plays an important role in the regulation of immune responses by producing type I IFN and interferon-stimulated genes (ISGs).⁴⁷ ACOD1 is generally regarded as an ISG in activated immune cells (Fig. 1). Pathogen infection (e.g., by Legionella pneumophila [L. pneumophila]²⁸ and B. pseudomallei = Burkholderia pseudomallei⁴⁸) leads to the production of type I and type II IFN and the subsequent expression of hundreds of ISGs, including ACOD1. In contrast, the loss of Ifnar limits inducible Acod1 mRNA expression in macrophages during Mtb⁹ or L. pneumophila infection²⁸ or LPS stimulation.³⁰ However, LPS or CpG DNA induces Acod1 mRNA expression in DCs in an IFNB-independent manner,¹⁰ although TLR4 and TLR9 are required for Acod1 mRNA upregulation in DCs.¹⁰ Inhibition of IRF9 expression by miR-93 reduces ACOD1 induction in a model of peripheral artery disease (Fig. 1a),⁴⁹ whereas IRF3 is required for IFNB-induced Acod1 mRNA expression in ovine macrophages.⁴⁵ In addition, a bioinformatics analysis suggested that IRF1 may regulate ACOD1 expression, but experimental evidence is still lacking.⁵⁰ In summary, IFN induces ACOD1 expression mainly in an IFN-dependent manner.

A20

A20, also known as TNFα-inducible protein 3, is a cytokine-inducible protein that controls inflammatory responses by negatively regulating NF-κB signaling.^51,52 Some studies have showed that LPS-, CORM-2- or hemin-mediated Acod1 mRNA upregulation is required for the subsequent upregulation of A20 in various types of macrophages.^20,38 Consistently, overexpression of ACOD1 in LPS-tolerized RAW264.7 cells restores LPS-induced A20 expression.²⁰ Genetic silencing of A20 eliminates ACOD1-dependent inhibition of proinflammatory cytokine (e.g., TNF, IL6, and IFNB) production, suggesting that A20 functions as a downstream signal of the ACOD1 pathway.^20,38 Another proteomics study showed that ACOD1 expression is upregulated in A20^−/− BMDMs in the absence or presence of LPS or TNF compared to that of WT cells, suggesting that A20 functions as a negative regulator of ACOD1 expression in macrophages.⁵³ Thus, there may be a feedback modulation loop between A20 and ACOD1 expression, but this mechanism requires further study.

PKC

The protein kinase C (PKC) family of protein kinases regulates the activity of other proteins by phosphorylating the hydroxyl groups on the serine/threonine side chains of these target proteins.⁵⁴ PKC has been identified as a positive regulator of LPS-induced ACOD1 expression through the phosphorylation of unknown substrates (Fig. 1a).⁸ Phorbol myristate acetate, an activator of the PKC pathway, can further augment LPS-induced Acod1 mRNA expression in macrophages. Consistently, pharmacological inhibition of PKC using H7 blocks LPS-induced Acod1 mRNA expression in RAW264.7 cells.⁸ In addition, genistein abolishes LPS-induced Acod1 mRNA expression by preventing tyrosine phosphorylation and PKC δ activation.^8,55 Unlike PKC, the protein kinase A (PKA) family of protein kinases has anti-inflammatory effects that rely on cellular levels of cyclic AMP (cAMP).⁵⁶ In contrast, the PKA agonist cAMP has no effect on inducible Acod1 mRNA expression.⁸ Taken together, these findings indicate that PKC but not PKA is required for LPS-induced ACOD1 expression in macrophages.

STATs

Seven mammalian STAT family members have been identified (STAT1, STAT2, STAT3, STAT4, STAT5A, STAT5B, and STAT6), which play multiple roles in orchestrating a cytokine-induced immune response during infection and cell death.⁵⁷ Among them, STAT3 and STAT1 play important roles in regulating ACOD1 expression in different situations (Fig. 1b). Activation of the STAT3 and glucocorticoid receptor (GR) signaling pathways can synergistically induce Acod1 mRNA expression in macrophage lineage cells from a zebrafish model.¹⁶ In addition to STATs, the expression of ACOD1 by GR signaling requires the activation of CCAAT enhancer-binding protein beta, a transcription factor that contains a basic leucine zipper domain (Fig. 1b). In addition, the lack of Stat1 limits Mtb infection-induced Acod1 mRNA expression in BMDMs.⁹ Therefore, STAT is also required for ACOD1 induction under certain conditions.

Other

Posttranscriptional mechanisms are also implicated in the regulation of ACOD1 gene expression, including events that may cause Acod1 mRNA degradation or prevent translation initiation.⁵⁸ For example, Acod1 mRNA in murine macrophages is increasingly stabilized by LPS treatment but not by infection with Mycobacterium avium subspecies paratuberculosis.⁴⁰ In addition, a miRNA is defined as a 21–25 nucleotide single-stranded RNA that regulates mRNA translation or stability.⁵⁹ MiR-378 was found to directly target and downregulate ACOD1 expression in glioma cells (Fig. 1b).⁶⁰ Taken together, these findings indicate that Acod1 mRNA expression is regulated at multiple levels. It remains unknown whether protein degradation is involved in the modulation of ACOD1 levels in activated immune cells.

Functions of ACOD1

Immune cells undergo extensive metabolic reprogramming under inflammatory conditions. Activated macrophages usually switch from oxidative phosphorylation (OXPHOS) to glycolysis, along with increasing glucose consumption or reactive oxygen species (ROS) production for host defense.^61,62 This metabolic reprogramming requires the expression of certain metabolic enzymes during macrophage activation. ACOD1 participates in immune responses through multiple pathways, including mediating the production of itaconate (an immune metabolite) and mitochondrial reactive oxygen species (mROS). In addition to metabolic functions, nonmetabolic functions are also important for ACOD1-mediated immune regulation. In the following sections, we discuss the main functions of ACOD1 in infection and immunity.

Itaconate metabolism

ACOD1 is a member of the bacterial 2-citrate citrate dehydratase family of proteins, whose enzymes are encoded by prpD.⁶³ A metabolomics analysis of LPS-activated macrophages identified ACOD1 as an enzyme that catalyzes itaconate production through cis-itaconate decarboxylation.^6,7 Increased itaconate production and secretion are observed in LPS- and IFNG-activated macrophages, whereas Acod1 depletion can prevent these processes.⁶⁴ Moreover, ACOD1-mediated itaconate production has been shown to inhibit LPS-induced IL6, IL12, IL18, IL1B, and type I IFN production in BMDMs.^21,30 Itaconate regulates the immune response through multiple mechanisms (Fig. 2a),^64,65 as described below.

Fig. 2.

Function of ACOD1 in infection and immunity. a ACOD1-mediated itaconate production leads to NFE2L2 activation, SDH inhibition, isocitrate lyase inhibition, and prenylation induction. b ACOD1-mediated ROS production contributes to antigen processing, pathogen killing, and A20 modulation. c ACOD1 mediates the utilization of fatty acids in OXPHOS to produce mROS for bacterial killing

Cysteine alkylation

Nuclear factor erythroid 2-like 2 (NFE2L2, also known as NRF2) is a master transcription factor in the response to oxidative stress and inflammation. NFE2L2 expression is reduced under normal conditions, which is due to protein degradation mediated by Kelch-like ECH-associated protein 1 (KEAP1), a component of the cullin 3-based E3 ubiquitin ligase complex.⁶⁶ Itaconate is an anti-inflammatory metabolite that can inactivate KEAP1 by alkylating certain cysteine residues (cys-151, -257, -288, -273 and -297), resulting in NFE2L2 accumulation and activation.^30,67 In contrast, this process is not observed in LPS-activated Acod1^−/− macrophages.⁶⁷ NFE2L2 activation promotes the expression of heme oxygenase 1 and the production of glutathione, thereby enhancing the anti-inflammatory and antioxidant capacity of immune cells.⁶⁸ Indeed, itaconate limits the production of LPS-induced cytokines (e.g., IL1B and TNF) and IRGs in an NFE2L2-dependent manner.³⁰ Consequently, itaconate-mediated activation of the NFE2L2 pathway prevents endotoxemia and liver ischemia-reperfusion injury in mice.^30,69 Whether ACOD1 overexpression promotes NFE2L2 activation requires further characterization.

Itaconate can also inactivate glyceraldehyde 3-phosphate dehydrogenase (GAPDH) by directly alkylating the cysteine residue at position 22 (cys-22), thereby reducing IL1B, iNOS, and TNF production by limiting aerobic glycolysis.⁷⁰ In contrast, the loss of Acod1 in macrophages promotes LPS-induced inflammation in part by enhancing GAPDH-dependent glycolysis. Overexpression of cys-22 mutant GAPDH results in a failure to reverse the inhibitory effect of itaconate.⁷¹ Taken together, these findings indicate that itaconate inhibits inflammation in part by cysteine alkylation of inflammatory or metabolic regulators.

SDH inhibition

Succinate dehydrogenase (SDH), also known as succinate:quinone oxidoreductase, is a key component of the citrate cycle and electron transport chain, contributing to mROS production and inflammatory conditions.⁷² SDH inactivation leads to the accumulation of succinic acid, thereby promoting hypoxia-inducible factor 1 subunit alpha-mediated IL1B production and the inflammatory response by inhibiting prolyl hydroxylase enzymes.⁷³ Itaconate directly inhibits the enzymatic activity of SDH in vivo and in vitro by competitively blocking the active site of SDH.^21,74 Both ACOD1 activation and itaconate production have been shown to increase succinate accumulation following LPS treatment.^7,21 In contrast, the production of itaconate and succinate is blocked in Acod1^−/− BMDMs after LPS and IFNG stimulation^7,21 In addition to bacterial infections, ACOD1-mediated itaconate inhibits SDH activity in ZIKV-infected neurons in vivo and in vitro.¹¹ Itaconate-mediated inhibition of the oxygen consumption rate occurs only in the presence of succinic acid and rotenone, and this process is abolished in Acod1-deficient macrophages,^7,21,75 further highlighting that itaconate controls inflammation in a succinate-dependent manner. Notably, the excessive inhibition of SDH by ACOD1 and itaconate may cause immune tolerance.¹⁵ In contrast, the recovery of SDH expression by β-glucan can offset the immune paralysis caused by ACOD1 and itaconate, which in turn contributes to the complete tricarboxylic acid (TCA) cycle and enhanced immune response.¹⁵ Therefore, proper itaconate levels are necessary for immune regulation.

Isocitrate lyase inhibition

The glyoxylate shunt is a two-step metabolic pathway that serves as an alternative to the TCA cycle and generates energy for bacterial growth and survival.⁷⁶ Apart from host cells, itaconate can be used as an antimicrobial metabolite by inhibiting the activity of isocitrate lyase, a key enzyme of the glyoxylate shunt.⁶

Prenylation induction

Prenylation is a site-specific lipid modification of proteins that plays a key role in the viral life cycle.^77,78 During vesicular stomatitis virus (VSV) infection, ACOD1 promotes viral entry and overgrowth by producing itaconate, which can increase the expression of enzymes related to geranylgeranyl diphosphate (GGPP) synthesis and promote GGPP-mediated protein prenylation, rather than in an IFN-dependent manner. VSV replication is inhibited when itaconate production is downregulated.^22,79 These findings indicate dual roles of itaconate in innate immunity, in a context-dependent manner.

Oxidative stress

ROS play a wide range of roles in immune defense, antibacterial effects, and signal transduction.⁸⁰ In addition to directly killing pathogens by causing oxidative damage, ROS can also promote the expression of genes involved in immune regulation.⁸¹ ACOD1-mediated ROS production has been observed in LPS-untolerized and LPS-tolerized macrophages,^20,38 which promotes the upregulation of A20 expression (Fig. 2b). ACOD1-mediated A20 expression may form a negative feedback loop to regulate NFκB signals, thereby limiting the production of proinflammatory cytokines³⁸ or causing immune tolerance.²⁰ Furthermore, ACOD1-mediated ROS production contributes to STAT1/3 pathway activation and subsequent cytokine production (Fig. 2b).⁸² These findings indicate dual roles of ACOD1-mediated oxidative stress in shaping the immune response in a downstream signal-dependent manner.

As a byproduct of OXPHOS, mROS aid in macrophage-mediated bactericidal clearance.⁸³ Indeed, the induction of ACOD1 promotes the utilization of fatty acids in OXPHOS, thereby increasing mROS production and subsequently enhancing the antibacterial activity of macrophages (Fig. 2c).¹⁶ In contrast, the loss of Acod1 limits the production of mROS and inhibits bacterial killing capacity.¹⁶ It is worth noting that mROS-independent signals are also involved in ACOD1-mediated bactericidal clearance. For example, ACOD1 restricts the growth of L. pneumophila in alveolar macrophages through a TLR-dependent but not mROS-dependent mechanism.²⁸ Given that ROS are an important cause of cell death, it is unclear whether ACOD1 regulates the death of host cells, especially by pyroptosis, which is a type of inflammatory cell death in immune cells.⁸⁴

Antigen processing

T-cell recognition by antigen-presenting cells depends on the expression of a series of peptides that bind to the major histocompatibility complex class I (MHC I) and class II (MHC II) molecules.⁸⁵ In particular, MHC I can recognize and bind viral proteins, present these proteins to cytotoxic T lymphocytes, and cause apoptosis in virally infected cells.⁸⁶ It has been shown that ACOD1 upregulation is the bridge linking innate immunity and adaptive immunity. LPS- or IFNB-induced ACOD1 expression can enhance MHC I function by promoting the expression of transporter associated with antigen processing 1 (TAP1) and proteasome subunit beta 9 (PSMB9), which are transporter proteins associated with antigen processing (Fig. 2b).^82,87 Furthermore, ACOD1-induced activation of MHC I exerts its antiviral effects through antigen presentation, although the exact mechanism is unknown.⁸² Recently, eight missense polymorphisms have been identified that are closely related to the active center of human ACOD1. Six of these polymorphisms are rare and may have no enzymatic activity.²⁷ These findings also suggest that ACOD1 may affect susceptibility to certain diseases, such as hepatitis B virus infection in humans and Marek’s disease virus (MDV) infection in chickens, through its own nucleotide mutations.^88,89 Single nucleotide polymorphism (SNP) analysis confirmed that two SNPs of the ACOD1 gene (rs17470171 and rs17385627) enhance the expression of ACOD1, which may activate the immune response to hepatitis B vaccination.⁸⁸ Two other SNPs of ACOD1 were found at the transcriptional binding site, which is related to sensitivity to Marek’s disease (MD), while homozygous genotypes increase resistance to MD.⁸⁹ The genetic variation of ACOD1 in other diseases merits further investigation.

Pathological role of ACOD1

Infectious diseases

In response to infection, host cells trigger the activation of immune pathways, which leads to the production of inflammatory mediators (e.g., cytokines, chemokines, and interferons) and maintains a balance between pro- and anti-inflammatory states. In this context, many studies have identified ACOD1 as one of the most upregulated genes under various conditions associated with infection (Table 1). For example, gene expression profiling revealed that infection with live B. burgdorferi can cause a 196.2-fold induction of Acod1 expression in J774A.1 macrophages,⁹⁰ and stimulation with IFNG and TNF leads to a 73-fold upregulation of Acod1 expression in ANA-1 macrophages.¹⁴ These findings suggest that ACOD1 may function as an important alarm signal in the immune response to infection.

Several studies have shown that Acod1 prevents infection in mice. Compared with WT mice, Acod1 global knockout mice generally have poorer survival, inflammatory cytokine bursts, uncontrolled neutrophil recruitment, and immune-mediated tissue damage.^11,28,91 Acod1 global knockout mice also exhibit higher mortality and viral burden in the brain than WT mice when infected with ZIKV and West Nile virus (WNV).¹¹ Moreover, myeloid cell-specific Acod1 knockout mice are sensitive to Mtb infection and have an increased Mtb burden in the lungs.⁹¹ In vitro studies in Acod1-deficient BMDMs show enhanced replication of Lactobacillus halophilus.²⁸ In primary neurons, Acod1 knockdown results in increased ZIKV and WNV replication,^11,92 while Acod1 overexpression reduces the production of multiple viruses, including WNV, St. Louis encephalitis virus (SLEV), and mouse hepatitis virus (MHV).⁹² These in vitro and in vivo studies support the anti-infective effects of ACOD1 under multiple conditions.

However, ACOD1 does not always play a protective role in a complex inflammatory environment and sometimes leads to increased pathogen burden and tissue damage. Acod1 mRNA is highly upregulated in splenic macrophages of V. Leishmania-infected mice, which aids parasite growth and survival.³² ACOD1 may also contribute to Borrelia-induced immunopathology, as ACOD1 expression can be induced by Borrelia burgdorferi infection and suppressed by the anti-inflammatory cytokine IL10.⁹⁰ In addition, ACOD1 expression contributes to mouse respiratory syncytial virus (RSV) infection, thereby promoting the production of proinflammatory cytokines and ROS, leading to extensive lung damage.⁹³ Moreover, ACOD1-mediated itaconate production promotes the replication of VSV, while Acod1^−/− mice show a reduction in inflammatory cell infiltration in the lung during VSV infection.²² The increase in ACOD1 and itaconate production in macrophages leads to the induction of M1-like polarization, leading to impaired angiogenesis, arteriogenesis, and perfusion in peripheral artery disease.⁴⁹ Furthermore, the immunosuppressive effects of ACOD1 on endotoxin-resistant macrophages and monocytes may lead to immune paralysis and life-threatening secondary infection.^15,20 Collectively, these studies suggest that ACOD1 plays dual roles in infection and may elicit different immune responses.

Cancer

Although many studies have focused on the role of ACOD1 in immune cells, especially macrophages, some studies have noted that dysregulated ACOD1 expression may promote tumorigenesis by modulating antitumor immunity. Although considerable progress has been made in surgery and conventional chemotherapy, ovarian cancer continues to have a high mortality rate in women. In the early and advanced stages of ovarian cancer, strong expression of ACOD1 was observed in malignant cells, which may limit the antitumor effect of infiltrating CD8⁺ T cells and IgY cell-containing cells in the tumor microenvironment (TME) through unknown mechanisms.⁹⁴ Tumor-associated macrophages (TAMs) are key cells that produce an immunosuppressive TME. A recent study showed that ACOD1-mediated itaconate production is an immunometabolic component of crosstalk between tumors and TAMs within the TME.⁹⁵ Cancer cells can increase ACOD1 expression in TAMs, leading to itaconate production, oxidative stress, and subsequent protumor macrophage polarization in the TME, ultimately promoting tumor progression in melanoma and ovarian cancer.⁹⁵ ACOD1 plays an oncogene-like role in gliomas and can promote the growth and invasion of cancer cells, resulting in a poor prognosis.^17,60 In contrast, intrauterine bisphenol A exposure inhibits Acod1 mRNA expression, which may promote breast cancer cell growth in mice.⁹⁶ These findings also suggest that ACOD1 plays dual roles in tumor biology, but the direct role of ACOD1 in cancer cells remains to be fully elucidated.

Neurological disorders

Microglia are the most abundant macrophages in the central nervous system (CNS) and play a dynamic role in maintaining immune homeostasis in the brain. Activation of microglia from a priming state to a destructive state contributes to the neuropathy of CNS diseases.⁹⁷ Recent studies have reported that upregulated ACOD1 is involved in the initiation and activation of microglia. Mice primed and reinfected with the intracranial parasite Toxoplasma gondii (T. gondii) had high gene expression of Acod1 and Ccr9 in microglia, thereby increasing microglial resistance to apoptosis and subsequent neurotoxicity.¹⁹ In contrast, inhibiting Acod1 in parasite-infected mice reduces parasitic DNA levels in brain tissue and shifts microglia into a protective phenotype.¹⁹ The production of Acod1-deficient microglia is triggered in lymphocytic choriomeningitis virus (LCMV)-infected mice, which promotes neuronal regeneration. This finding suggests that increased ACOD1 expression can be considered a marker of neurotoxic microglia. Elevated Acod1 mRNA is also observed in microglia in high-fat diet mice or mice infected with neurotrophic pathogens (e.g., T. gondii, HIV vaccine, recombinant A3A capsid protein of Theiler’s murine encephalomyelitis virus, and β-amyloid protein fragments). This change in ACOD1 expression is associated with neurotoxic microglial activation and chronic neuroinflammation, which causes progressive neuronal loss and ultimately leads to cognitive decline and neurodegenerative diseases, such as Alzheimer’s disease and dementia.^19,98 Although it is not clear how ACOD1 specifically regulates the immune function of microglia, ACOD1 may exert its enzymatic effects by altering microglial metabolism. The precise contributions of ACOD1 in other tissue-resident macrophages remain unclear.

Embryo implantation

Embryo implantation takes place in the first trimester of pregnancy, and this stage is related to the production of a large number of inflammatory molecules by epithelial cells and immune cells.⁹⁹ This proinflammatory condition is regulated by ovarian steroid hormones, estradiol, and progesterone, which help and protect early pregnancy. Regulation of the expression of certain inflammatory molecules is essential for obtaining and developing uterine receptivity, which is one of the key factors for successful implantation. ACOD1 has been shown to be related to implantation. Acod1 mRNA expression in the luminal epithelium has been shown to occur in a specific manner at the implantation stage, increase on day 3 after pregnancy, peak at day 4, and abruptly decrease at day 5.¹⁸ In mice, day 4 after pregnancy is the time at which the blastocyst disappears from the recipient endometrium.¹⁰⁰ The administration of progesterone for 24 h in ovariectomized mice can induce expression of Acod1 mRNA. This increase in Acod1 expression can be reversed by pharmacologically (e.g., with PR486) or genetically inhibiting the progesterone receptor (PR).^18,36 A similar inhibition of Acod1 expression is observed when mice are treated with PKC inhibitors during implantation, while E₂ and P₄ synergistically and independently increase PKC activity on day 4 of pregnancy.¹³ These findings indicate that the PR pathway is essential for P-responsive ACOD1 expression in the uterine epithelium by regulating PKC activity. Triggered by estrogen, the level of the endometrial cytokine LIF, which is also a well-known modulator of successful implantation, increases and peaks on day 4.¹⁰¹ In pseudopregnant Lif-deficient mice, an intrauterine Lif injection promotes Acod1 expression in the uterus through Stat3 activation.^33,34 Therefore, the injection of anti-Lif antibodies or Stat3 peptide inhibitors into the uterine cavity of pregnant mice on day 3 prevents the induction of Acod1 and subsequent embryo implantation.^34,35 In addition, the protective role of Acod1 in the implantation process has been further demonstrated by using Acod1-related antisense oligodeoxynucleotides on the lumen on day 2.¹⁸ These findings highlight the need for ACOD1 in embryo implantation, despite the complex molecular network of the implantation process.

Conclusions and perspectives

ACOD1 has emerged as an important modulator with different functions in innate and adaptive immunity. Activation of the ACOD1 pathway may limit pathogen infection and promote embryo implantation. Moreover, abnormal ACOD1 expression can cause tumor progression, neurodegeneration, and immune paralysis. Although the dual roles of ACOD1 in immunity are mainly related to ACOD1-mediated itaconate metabolism, oxidative stress, and antigen processing, more research is needed to clarify the roles and regulatory mechanisms of ACOD1 in both immune and nonimmune cells. In addition to the broad role of ACOD1 in innate immunity, ACOD1 may regulate adaptive immunity by promoting MHC I expression⁸². However, how to express ACOD1 in specialized antigen-presenting cells to regulate adaptive immunity in diseases requires further investigation.

The ACOD1-itaconate axis is becoming a target of disease treatment. In particular, the development of metabolically stable itaconate derivatives may have great potential in the treatment of inflammatory diseases. In contrast, for patients with peritoneal tumors, inhibition of ACOD1 may be beneficial because ACOD1-expressing TAMs mediate tumor growth.⁹⁵ With the discovery of its crystal structure, technical modifications of the ACOD1 protein to interfere with or improve its activity may be a promising therapeutic strategy. Some agents have been found to inhibit the induction of ACOD1 under certain conditions, such as dexamethasone,⁸ ethylene glycol tetraacetic acid,⁸ and β-glucan.¹⁵ Further research needs to not only distinguish the overlapping and unique functions of ACOD1 and itaconic acid in immunity but also evaluate the pathological consequences of drugs targeting ACOD1 on disease.

Competing interests

The authors declare no competing interests.