Nucleotide metabolic reprogramming: a “double-edged sword” in obesity-related inflammation—from metabolic adaptation to pathological imbalance

Abstract

Obesity has become a serious global public health challenge, characterized by chronic, sterile, low-grade inflammation with excessive NLRP3 inflammasome activation; however, the mechanism underlying the reduced activation threshold remains unclear. A recent study by Liu et al. demonstrated that obesity induces the phosphorylation and inactivation of SAMHD1, leading to a massive accumulation of cytosolic dNTPs. This causes excess dNTPs to enter mitochondria via the PNC1/2 transporters, bypassing the classical CMPK2 salvage synthesis pathway and triggering uncontrolled mtDNA synthesis and oxidative damage, ultimately resulting in the excessive activation of NLRP3 and an inflammatory response. Based on this finding, this paper presents a “double-edged sword” model of nucleotide metabolic reprogramming in obesity-related inflammation. Early reversible inactivation of SAMHD1 may represent a metabolic adaptive response that confers functional benefits to macrophages; however, once metabolic stress persists and causes nucleotide metabolic reprogramming to exceed a yet-to-be-defined threshold, an inflammatory positive feedback loop may be established. In terms of clinical translation, targeting PNC1/2 could specifically inhibit NLRP3 activation, which may provide new upstream intervention strategies for various aseptic inflammatory conditions such as gout; however, this approach also carries potential risks of mitochondrial toxicity and impaired anti-infective immunity. Future efforts should focus on macrophage-specific delivery, precision interventions tailored to disease stages, and precise anti-inflammatory strategies based on biomarkers such as circulating dNTPs, p-SAMHD1, and ox-mtDNA, thereby advancing the clinical translation of metabolism-related inflammation within safe parameters.

Introduction

Obesity, a complex metabolic disease resulting from pathological adipose tissue accumulation driven by prolonged positive energy balance, has become a major global public health challenge. Its pathological basis primarily involves endocrine and immune dysregulation resulting from chronic low-grade inflammation, which ultimately leads to systemic metabolic imbalance [1]. In recent years, the understanding of obesity has evolved from a simple metabolic disorder to a systemic state of low-grade inflammation state, termed“metaflammation.” This sterile inflammation is independent of pathogen infection but is a key driver of the onset and progression of type 2 diabetes and cardiovascular disease [2]. Among these mechanisms, the NOD-like receptor protein 3 (NLRP3) inflammasome, a sensor of metabolic stress in macrophages—plays a central role in initiating the inflammatory response. Numerous studies have demonstrated that interleukin-1 beta (IL-1β) levels are markedly elevated in the adipose tissue from obese individuals, and that genetic deletion of NLRP3 or its downstream targets substantially improves insulin resistance [3, 4]. However, why the NLRP3 inflammasome is in a hyperactivated state in obesity has long remained unresolved. Previous work have largely focused on the enhanced priming signals such as lipopolysaccharide (LPS) or activator signals such as adenosine triphosphate (ATP); however, these experimental findings do not explain why macrophages from obese patients continue to exhibit secrete markedly more IL-1β and display higher caspase-1 activation than those from normal individuals, under identical stimulation [5]. We therefore hypothesize that obesity may lower the intrinsic “activation threshold” of macrophages [6].

The recent study by Liu et al. provided new insight into this puzzle. Previous studies have shown that oxidized mitochondrial DNA (mtDNA) is a direct activator of NLRP3 [7], and that the increase in mtDNA in obese patients does not stem from the classical cytidine monophosphate kinase 2 (CMPK2)-dependent mitochondrial salvage synthesis pathway, but rather implicates to a long-neglected dNTP catabolic enzyme, SAM and HD domain–containing protein 1 (SAMHD1) [8]. This discovery brings the balance between nucleotide catabolism and anabolism to the forefront of obesity-related inflammation regulation; It is plausible that nucleotide metabolic reprogramming itself acts as a “double-edged sword,” supporting the adaptive functions of macrophages in a lipotoxic environment while also having the potential, upon crossing a certain threshold, to transform into a key effector that exacerbates pathological inflammation. Building on the findings of Liu et al., this article provides an in-depth analysis of this “double-edged sword” model and, from a perspective that spans metabolic adaptation to pathological imbalance, examine and reflect on the root causes of obesity-related inflammation and potential intervention strategies.

SAMHD1 inactivation—the shift from “braking” to “accelerating”

The physiological role of SAMHD1: the “gatekeeper” of the dNTP pool

SAMHD1 was initially identified as a HIV-1 restriction factor in human myeloid cells [9]. It was subsequently confirmed to be a highly conserved deoxynucleoside triphosphatase (dNTPase), that hydrolyzes the four dNTPs—dATP, dGTP, dCTP, and dTTP—into deoxynucleosides and triphosphate, thereby maintaining low cytoplasmic dNTP level [10]. This dNTPase activity is particularly important for non-proliferating, terminally differentiated cells such as macrophages, which, although they do not require large dNTP pools for nuclear DNA replication, must rely on this function to preserve the homeostasis of the nucleotide pool and prevent aberrant DNA synthesis driven by excess dNTPs [11]. The dNTPase activity of SAMHD1 depends on homotetramer assembly; this process not only requires dGTP as an allosteric activator but is also tightly controlled by post-translational modifications [12]. Phosphorylation at site T592 (human)/T603 (mouse) is an important modification regulating SAMHD1 function. Although the specific impact of this phosphorylation on overall dNTP hydrolase activity remains controversial [13], it has been demonstrated to alter the enzyme’s substrate specificity, primarily by selectively reducing the hydrolysis rate of dCTP [14]. Furthermore, current evidence suggests that this phosphorylation promotes tetramer dissociation by enhancing dynamic movement, leading to a significant decrease in enzyme activity or even inactivation [13, 15].

How obesity inactivates SAMHD1: phosphorylation and tetramer disruption

Liu et al. demonstrated that obesity inactivates SAMHD1 by increasing phosphorylation and reducing tetramer formation. Elevated phosphorylation levels and a substantial reduction in tetramer formation were observed in both monocyte-derived macrophages from obese patients and adipose tissue-associated macrophages from high-fat diet-fed mice. Of note, total SAMHD1 protein levels remained unchanged in these cells, indicatingthat obesity likely inactivates SAMHD1 rather than by downregulating its expression, but through specific post-translational modifications [8]. Earlier work has demonstrated that the phosphorylation status of SAMHD1 is governed by the opposing activities of cyclin-dependent kinase 2 (CDK2) and protein phosphatase 2 A (PP2A) [16, 17]. However, it remains unclear how the lipotoxic environment triggers this phosphorylation process. One possible explanation is that contribute to this process obesity-associated oxidative stress and altered adipokine profiles. For instance, dietary fat can upregulate CDK2 [18]; simultaneously, adiponectin, which activates PP2A, is reduced in obesity [19]. Additionally, reactive oxygen species (ROS) can promote PP2A inactivation while enhancing the phosphorylation and activation of CDK2 [20, 21]. These putative upstream regulatory mechanisms warrant further investigation.

SAMHD1 as a “metabolic-inflammatory checkpoint”

The traditional view of SAMHD1 as merely an antiviral restriction factor clearly may underestimates its role in aseptic inflammation. The work by Liu et al. resefines SAMHD1 as a key molecular checkpoint linking nucleotide metabolism to innate immunity. Under physiological conditions, SAMHD1 effectively limits excessive synthesis by maintaining a “low nucleotide flux” in macrophages, thereby keeping NLRP3 inflammasome activation within the moderate range required for defense. Once SAMHD1 is inactivated, this “molecular brake” is released, and macrophages immediately enter a state of “high nucleotide flux.” These findings indicate that metabolic checkpoints extend beyond energy metabolism to include nucleotide metabolism [8].Indeed, DNTP pool homeostasis directly determines the rate and fidelity of DNA synthesis, and mtDNA synthesis is closely linked to the production of mitochondrial ROS and the release of damage-associated molecular patterns (DAMPs) [22]. Based on this, we propose that the active state of SAMHD1 may function as a “metabolic-inflammatory checkpoint” that modulates the inflammatory sensitivity of macrophages.

The “adaptive cost” of SAMHD1 inactivation from an evolutionary perspective

SAMHD1 exhibits a high degree of evolutionary conservation, with homologous genes present from zebrafish to humans [23]. Liu et al. also demonstrated that samhd1 deletion in zebrafish leads to excessive NLRP3 inflammasome activation, suggesting that SAMHD1 may serve as an ancient “anti-inflammatory defense” mechanism that prevents unnecessary inflammatory responses triggered by fluctuations in the dNTP pool [8]. As for why the lipotoxic microenvironment actively suppresses this protective mechanism, one plausible hypothesis is that, in the early stages of obesity, transient and reversible inactivation of SAMHD1 may confer a “metabolic dividend”to macrophages (Figure 1).

Fig. 1.

Nucleotide metabolic reprogramming: a new perspective on the mechanisms of obesity-related inflammation and translational research. SAMHD1 SAM and HD domain–containing protein 1, PNC1/2 pyrimidine nucleotide carrier 1/2, CMPK2 cytidine monophosphate kinase 2, mtDNA mitochondrial DNA, NLRP3 NOD-like receptor protein 3

Cytosolic dNTPs entering mitochondria—how does the bypass pathway circumvent CMPK2?

Conventional understanding: CMPK2 is the rate-limiting enzyme in mtDNA synthesis

In macrophages under physiological conditions, the supply of dNTPs within mitochondria depends primarily on the “mitochondrial dNTP salvage synthesis pathway” mediated by CMPK2 [24]. As mitochondrial enzyme, CMPK2 phosphorylates dNMPs entering from the cytoplasm to form dNDPs, which are subsequently converted to dNTPs by nucleoside diphosphate kinase (NDPK). This pathway is relatively slow but controlled and sustainable [25]. A study by Zhong et al. (2018) demonstrated that, under normal nutritional conditions, CMPK2 is essential for NLRP3 inflammasome activation [26].

The groundbreaking finding of this study: a bypass pathway that circumvents CMPK2

An important finding of the study by Liu et al. is that, in the context of obesity or SAMHD1 deficiency, CMPK2 is no longer a rate-limiting enzyme for mtDNA synthesis. They demonstrated that CMPK2 knockdown in Samhd1^+/+ bone marrow–derived macrophages (BMDMs) did not affect mtDNA biosynthesis, oxidized mtDNA (ox-mtDNA) production, or NLRP3 inflammasome activation. In contrast, in wild-type cells, CMPK2 knockdown significantly suppressed these parameters. This suggests the existence of a CMPK2-independent, more efficient mtDNA synthesis pathway. The core mechanism of this bypass involves the direct entry of cytoplasmically accumulated dNTPs into the mitochondrial matrix directly via mitochondrial nucleotide transporters. Liu et al. further identified pyrimidine nucleotide carrier 1/2 (PNC1/2) as the key transporter in this process. Knocking down PNC1/2 not only significantly reduced dNTP levels mitochondrial Samhd1^−/− BMDMs but also completely reversed the hyperactivated state of the NLRP3 inflammasome [8].

A critical perspective: the bypass pathway is a “high-efficiency but high-risk” strategy

From a kinetic standpoint, direct transport of dNTPs into mitochondria via PNC1/2 is considerably more efficient than the stepwise salvage synthesis that depends on CMPK2. When macrophages face energy-intensive conditions—such as active phagocytosis and apoptotic cells—this bypass may provide “instant fuel” to enhance mitochondrial output. Yet, this gain in efficiency is offset by the loss of the “rate-limiting” protective mechanism intrinsic to the CMPK2 pathway. With persistent SAMHD1 inactivation, cytoplasmic dNTPs accumulate, the mitochondrial dNTPs pool becomes chronically overloaded, mtDNA synthesis escapes normal regulation. Such efficiency–risk trade-offs are common in biology; for example, glycolysis rapidly provides ATP but predisposes to lactic acidosis, whereas aerobic oxidation is highly efficient but heavily dependent on intact mitochondrial function [27]. We propose that the CMPK2 pathway represents a“robust”metabolic mode, while the dNTP direct transport bypass represents a“sprint”metabolic mode. A lipotoxic microenvironment may force macrophages into a sustained “sprint” state, ultimately triggering metabolic collapse and an inflammatory burst.

Substrate specificity and network effects of PNC1/2: a detail requiring clarification

From a mechanistic perspective, PNC1 and PNC2 are mitochondrial pyrimidine nucleotide transporters with high selectivity for pyrimidine dNTPs; to date, there is no direct evidence that they transport dATP [28]. This substrate specificity raises a key scientific question: as observed by Liu et al., why does PNC1/2 knockdown reduce mitochondrial dATP levels? As the authors discuss, this is an indirect effect, possibly resulting from mtDNA replication arrest, metabolic reprogramming, or stress-induced changes in transporter activity [8]. This observation also points to the possible existence of an as-yet-unidentified mitochondrial transporter for purine dNTPs that acts in concert with PNC1/2 to maintain mitochondrial nucleotide pool homeostasis (Figure 1).

mtDNA oxidation and NLRP3 hyperactivation—from “energy support” to “danger signal”

Physiological significance of new mtDNA synthesis: supporting the renewal of the mitochondrial respiratory chain

The 13 core structural subunits of the electron transport chain encoded by mtDNA form the basis for mitochondrial oxidative phosphorylation capacity [29]. Upon macrophages activation, mtDNA synthesis is upregulated to meet the increased ATP demand. Under physiological conditions, this serves a protective role, as macrophages would struggle to perform functions such as phagocytosis and antigen presentation without sufficient newly synthesized mtDNA [26]. However, research by Liu et al. revealed another aspect: in SAMHD1-deficient cells, EdU labeling experiments showed a significant increase in mtDNA biosynthesis. One interpretation is that this excessive synthesis reflects a metabolic hypersensitivity response rather than purely pathological damage [8].

Why is newly synthesized mtDNA more susceptible to oxidation? — from the perspective of TFAM packaging

Is newly synthesized mtDNA as susceptible to oxidation as mature mtDNA? The answer is no. Mature mtDNA is tightly packaged by transcription factor A, mitochondria (TFAM) into highly ordered nucleoid-like structures; this physical shielding protects it from attack by mitochondrial ROS [30]. In contrast, nascent mtDNA remains unpackaged for an extended period, during which it is vulnerable to oxidative modification by mitochondrial ROS, yielding large amounts of 8-hydroxy-2′-deoxyguanosine (8-OH-dG) [26]. Liu et al. confirmed that, in Samhd1^−/− BMDMs, ox-mtDNA levels were significantly elevated in both mitochondria and the cytoplasm [8].

From mitochondria to cytoplasm: the mPTP/VDAC-dependent release pathway

Oxidatively modified mtDNA must cross the mitochondrial membrane and be released into the cytoplasm to bind to and activate the NLRP3 inflammasome [7]. Liu et al. demonstrated that this release depends on the mitochondrial permeability transition pore (mPTP) and the voltage-dependent anion channel (VDAC). In Samhd1^−/− cells, inhibition of mPTP or VDAC blocks the leakage of ox-mtDNA into the cytoplasm and reverses NLRP3 hyperactivation. These observations link mtDNA metabolism, oxidative damage, and altered mitochondrial membrane permeability into a coherent pathogenic cascade [8].

The core of the “double-edged sword”: differences in concentration and spatiotemporal context

Low to moderate levels of mtDNA de novo synthesis and oxidative modification may fall within the physiological range needed to maintain immune surveillance and stress responses [8]. In the early stages of infection or tissue injury, moderate NLRP3 activation contributes to pathogen clearance and tissue repair [31]. However, when persistent dNTP overload drives uncontrolled mtDNA synthesis and massive ox-mtDNA release, NLRP3 shifts from a defender to a driver of inflammatory damage. This captures the essence of the proposed double-edged sword model: the same metabolic pathway can serve adaptive or pathological functions, depending on intensity, duration, and spatial distribution. However, the threshold ox-mtDNA concentration that triggers an inflammatory response, and whether a clear “safe window” separates physiological from pathological states, remain undefined. These questions await more refined kinetic models and single-cell analyses (Figure 1).

Metabolic adaptation vs. pathological imbalance—the existence of a tipping point and the evidence gap

A plausible hypothesis: SAMHD1 inactivation in early obesity may offer a “metabolic dividend”

Although Liu et al. did not present dynamic data from early obesity, previous work indicates that during the initial stages of obesity, macrophages extensively infiltrate adipose tissue and the liver to clear leaked lipids and apoptotic adipocytes [32]. These functions demand a high energy supply and membrane remodeling capacity. Partial SAMHD1 inactivation could mildly elevate cytoplasmic dNTP, which would then replenish the mitochondrial dNTP pool via the bypass pathways, boosting mtDNA synthesis and respiratory chain efficiency to support heightened macrophage activity [8]. This suggests that nucleotide metabolic reprogramming may initially serve as a metabolic adaptation that helps macrophages maintain clearance functions in the obese microenvironment. However, this adaptive interpretation remains hypothetical, at this stage, as direct longitudinal evidence tracking the transition from adaptation to pathology is lacking. Whether a transient, reversible phase of SAMHD1 inactivation exists in early obesity—and if so, whether it confers benefits without progressing to chronic inflammation—requires formal validation through time-series studies in preclinical models. Notably, the adaptive phase described here aligns with the broader concept of metaflammation, wherein macrophages first undergo adaptive changes in response to altered adipose tissue cues, but evolve into a maladaptive, pro-inflammatory state under sustained metabolic stress. The nucleotide metabolic reprogramming described by Liu et al. may represent the specific molecular manifestation of this general immunometabolic principle [33].

The formation of a tipping point: positive feedback and system collapse

Under prolonged high-fat diet feeding, sustained fatty acid loading, endoplasmic reticulum stress, and oxidative stress together may maintain SAMHD1 in a phosphorylated state through continuous phosphatase inhibition or kinase activation. One possible explanation is that with long-term SAMHD1 inactivation, cytoplasmic dNTP shift from a transient, mild elevation to a persistent, pathological elevation [11]. Consequently, dNTPs into mitochondria would becomes dysregulated, and mtDNA synthesis exceeds the packaging capacity of TFAM. Newly synthesized, unprotected mtDNA is then exposed to mitochondrial ROS, and the accumulating ox-mtDNA may surpass the NLRP3 activation threshold, potentially establishing a positive feedback loop that sustains inflammatory tissue damage [26, 34]. Although this model is biologically plausible and consistent with existing cross-sectional data, its direct in vivo confirmation remains an important goal. Moreover, the ensuing IL-1β secretion recruits additional immune cells, further amplifying tissue injury and metabolic disruption. This positive feedback mechanism offers a plausible explanation for the difficulty in spontaneously resolving obesity-related chronic inflammation [8]. Liu et al. revealed the molecular basis of this feedback loop; whether bidirectional regulation exists between SAMHD1 and oxidative stress, however, warrants further investigation.

Key missing evidence and emerging supportive observations

Although the double-edged sword model is conceptually appealing, several important gaps remain. First, dynamic time-series data are lacking. The animal model used by Liu et al. was based on established obesity; longitudinal studies monitoring SAMHD1 phosphorylation, dNTP levels, mtDNA synthesis, and NLRP3 activity at different time points during the pre-obesity phase are still needed. Consequently, the markers that signal the transition from adaptation to pathological damage have not been defined.

Second, although Liu et al. did not include perform dNTP titration experiments, their data nonetheless point to a quantitative relationship. Compared with wild-type controls, dNTP levels in Samhd1⁻/⁻ cells were significantly elevated, and this elevation was positively correlated with increased mtDNA synthesis, ox-mtDNA accumulation, and NLRP3 activation. This observation is consistent with a concentration-dependent threshold at which metabolic support gives way to inflammatory pathology, although it does not provide direct evidence of such a threshold. Confirming this threshold definitively will require chemogenetic tools that allow dNTP levels to be titrated across a continuous gradient in tightly controlled cellular environments [8, 35].

Third, sensitivity to SAMHD1 phosphorylation likely differs among individuals. Factors such as sex, age, gut microbiota composition, and SAMHD1 polymorphisms may influence this threshold, potentially explaining why some obese individuals develop severe nonalcoholic steatohepatitis while others remain metabolically healthy [36].

Fourth, SAMHD1 phosphorylation may mediate a progressive regulatory mechanism rather than a binary switch. Phosphorylation at T592 alters substrate specificity by selectively reducing dCTP hydrolytic while preserving activity toward other dNTPs; however, its effect on overall dNTPase activity remains debated [14]. SAMHD1 inactivation in obesity may be gradual: partial phosphorylation could initially fine-tune the dNTP pool to meet metabolic demands, whereas sustained phosphorylation exceeding a critical threshold may abolish dNTPase function and drive pathological nucleotide accumulation [37].

Finally, not all macrophages enter the high nucleotide flux state simultaneously at the single-cell level. It is unclear whether a susceptible macrophage subpopulation first crosses the threshold and subsequently recruits surrounding cells via paracrine signaling. Recent single-cell studies indicate that macrophage phenotypes evolve temporally during obesity progression. In diet-induced obese mice, adipose tissue macrophages were predominantly consisted of the pro-inflammatory Mac2 and Mac3 subsets early in obesity; by week 8 of high-fat feeding, transitional Mac4 macrophages became dominant; and by week 14, the Mac5 subset of lipid-associated macrophages (LAMs) predominated [38]. Although that study did not directly measure SAMHD1 phosphorylation or dNTP flux, it confirmed that macrophages do exhibit phasic phenotypic shifts during obesity progression, providing indirectly supports for the notion that nucleotide metabolic reprogramming gradually evolve transitions from an adaptive to a maladaptive state. Thus, single-cell metabolomics or spatial transcriptomics may help resolve these questions.

Understanding the “adaptation-dysregulation” dialectic from an evolutionary perspective

From an evolutionary perspective, the plasticity of metabolic pathway represents a means of adapting to environmental stressors. Throughout human history, food scarcity has been the norm, while overnutrition is a recent phenomenon, emerging only in the last century. The phosphorylation-dependent regulation of SAMHD1 may have evolved as a response to infection or tissue injury, temporarily boosting macrophage function without inducing chronic inflammation. However, when this acute stress response is hijacked by sustained overnutrition, it becomes a classic example of maladaptation. This phenomenon extends beyond metabolic diseases: insulin resistance may acutely prioritize glucose delivery to the brain during energy scarcity, yet chronically leads to diabetes [39]; similarly, adipose tissue expansion initially serves as a safe energy depot, but excessive expansion causes lipotoxicity [40]. Thus, nucleotide metabolic reprogramming may also involve adaptive imbalance mechanisms (Figure 1).

Translational implications—potential and challenges of targeting mitochondrial dNTP transport

From mechanism to intervention: PNC1/2 as potential drugable targets

An notable finding by Liu et al. was that knocking down PNC1/2 to block the cytoplasmic dNTPs into mitochondria reverses excessive activation induced by SAMHD1 deficiency or obesity. Importantly, this intervention did not alter TNF-α expression, suggesting that it may act specifically on the NLRP3 pathway rather than causing broad immunosuppression [8]. Thus, interfering with mitochondrial dNTP transport may represent a therapeutic strategy to alleviate NLRP3-driven inflammation. However, the current evidence is limited to in vitro macrophage models and SAMHD1 knockout mice; whether PNC1/2 inhibition provide therapeutic benefits in human obesity-related inflammatory conditions, and its safety profile, require further investigation. Targeting PNC1/2 as an upstream node may theoretically offer advantages over direct NLRP3 inhibitors, because PNC1/2 may only become a rate-limiting only under nucleotide metabolic stress. Nevertheless, this advantage remains theoretical and demands rigorous clinical evaluation; direct NLRP3 inhibitors such as MCC950 have already entered clinical trials, whereas PNC1/2-targeting are still at the preclinical stage, and their safety profiles has not been established [41]. Moreover, future translational efforts should prioritize macrophage-specific delivery, for instance by using nanoparticles to deliver siRNA or small molecule to macrophages; or by employing the Lyz2-Cre system to generate macrophage-specific conditional knockouts [42].

Potential risks: “another double-edged sword” on the path to translation

Although targeting PNC1/2 has shown potential therapeutic prospects in preclinical models, strategies that interfere with mitochondrial dNTP transport also carry a double-edged effect, warranting careful evaluation of safety alongside efficacy.

First, the tissue distribution of PNC1/2 raises safety concerns for systemic inhibition. PNC1/2 is expressed not only in macrophages but also in neurons, cardiomyocytes, and hepatocytes, where it contributes to mtDNA replication [28, 43, 44]. Studies have shown that PNC1/2 inhibition leads to mitochondrial dysfunction and ROS-dependent phenotypes, and dysregulation of these proteins has been linked to various metabolic and mitochondrial diseases [45, 46]. These observations suggest that systemic PNC1/2 inhibition may result in considerable off-target toxicity in tissues with under high oxidative stress.

Second, the clinical use of nucleoside analogue reverse transcriptase inhibitors offers relevant insights into the safety of interventions that target mitochondrial nucleotide transport. Drugs such as zidovudine inhibit mitochondrial adenine nucleotide transporters, leading to mtDNA depletion and inducing mitochondrial toxicity that includes cardiomyopathy, neuropathy, and lactic acidosis [47, 48]. Although PNC1 and PNC2 are distinct transporters, these findings underscore the sensitivity of mitochondrial nucleotide homeostasis to pharmacological interventionsand the risk of unpredictable tissue-specific toxicity.

Third, PNC1 and PNC2 contribute to the maintenance of mitochondrial DNA replication and regulating ROS production. Their knockdown inhibits mtDNA and ROS production; While this effect is beneficial in the setting of NLRP3 hyperactivation, it raises the concern that PNC1/2 inhibition may inadvertently suppress physiological mitochondrial function in tissues that rely on sustained mtDNA turnover, such as neurons and cardiomyocytes [28, 49].

Fourth, the conclusion that PNC1/2 inhibition leaves normal immune function unaffected requires caution. Although Liu et al. showed that PNC1/2 knockdown did not significantly alter TNF-α expression, this observation alone is insufficient to confirm that anti-infectious immunity remains intact. The NLRP3 inflammasome participates in the host’s defense against a range of pathogens; any inhibition of this pathway, whether direct or indirect, may increase infection risk [50, 51]. Therefore, systematic infection challenge studies in PNC1/2-deficient are needed to define the safety boundaries of this strategy [52].

Fifth, as a downstream transcriptional target of the insulin-like growth factor 1 (IGF-1) signaling pathway, PNC1 is involved in cell proliferation, growth, and mitochondrial homeostasis. This raises notable safety concerns: targeted inhibition of PNC1 may interfere with IGF-1-mediated tissue repair and regeneration. Particularly in metabolic diseases that depend on continuous tissue remodeling to maintain homeostasis, such an intervention could impair tissue repair, leading to adverse systemic metabolic consequences [53].

Finally, the potential to interfere with metabolic adaptation and the timing of intervention must be considered. In early obesity, partial SAMHD1 inactivation may represent an adaptive response that helps maintain metabolic homeostasis. Inhibiting PNC1/2 during this phase could compromise the capacity of macrophages to clear lipids and apoptotic cells, thereby paradoxically accelerating metabolic dysfunction [46]. Thus, PNC1/2 inhibition strategies are best viewed as therapeutic interventions targeting established inflammatory states, rather than as preventive treatments. Such stage-specific approaches require accurate determination of the intervention window based on disease staging to avoid premature intervention during the adaptive phase [35]. The hypothesis and its translational potential still need to be validated through rigorously designed preclinical studies.

Broader indications: exploring applications beyond obesity to other aseptic inflammatory conditions

Given that the NLRP3 inflammasome is implicated in the pathogenesis of multiple aseptic inflammatory diseases, strategies targeting mitochondrial dNTP transport may have broader translational relevance; however, this remains remains speculative and requires experimental validation. Gout is a chronic aseptic inflammatory disease triggered by the deposition of monosodium urate (MSU) crystals resulting from hyperuricemia. MSU crystals are the primary trigger of gout; they induce acute joint inflammation by activating the NLRP3 inflammasome [54]. Does the activation of the NLRP3 inflammasome by MSU crystals also depend on dNTP metabolic reprogramming? This is a scientific question that remains to be investigated.

Atherosclerosis is a chronic aseptic inflammatory process initiated by the deposition of lipoproteins on the arterial intima, followed by oxidative modification, leading to foam cell formation, plaque progression, and ultimately vascular lesions resulting from events following plaque rupture. Previous studies have shown that macrophages within atherosclerotic plaques are in a state of chronic metabolic stress [35]. Determining whether these lesional macrophages exhibit elevated SAMHD1 phosphorylation levels and aberrant dNTP metabolism could shed light on the persistence of sterile inflammation within plaques.

Alzheimer’s disease (AD) is a neurodegenerative disorder characterized by progressive cognitive decline, with senile plaques formed by β-amyloid deposition and neurofibrillary tangles composed of hyperphosphorylated tau as core pathological features. Neuroinflammation driven by microglia surrounding Aβ plaques contributes to AD progression [55]. Whether this neuroinflammatory response is accompanied by uncontrolled mtDNA synthesis and oxidative damage—which in turn activates the NLRP3 inflammasome that triggers represents a area of research at the intersection of immunology and metabolism.

Aicardi-Goutières syndrome (AGS) is a hereditary autoinflammatory encephalopathy caused by mutations in nucleic acid metabolism genes, such as SAMHD1, leading to abnormal accumulation of endogenous dNTPs and excessive type I interferon pathway activation. Liu et al. suggested that neuroinflammation in AGS may be driven by excessive NLRP3 activation resulting from SAMHD1 deficiency [56]. It is therefore plausible that PNC1/2 inhibitors may serve as novel drug candidates for treating this type of rare hereditary inflammatory disease.

Biomarkers and the prospects for precision anti-inflammation

Beyond potential therapeutic strategies, this study also offers insights into the development of diagnostic and monitoring tools for metabolic inflammation. With the identification of specific biomolecular markers, anti-inflammatory treatment could move from a broad, traditional approach toward precision anti-inflammation. However, the clinical utility of these markers still requires validation through large-scale cohort studies. First, circulating dNTPs. As a direct product of mitochondrial stress, dNTP levels in plasma or serum may serve as a dynamic biomarker reflecting activation status of NLRP3 in obese patients. This application, however, this depends on high-sensitivity liquid chromatography–mass spectrometry (LC-MS) and large clinical cohorts to establish diagnostic thresholds [8]. Second is the phosphorylation status of SAMHD1 in monocytes. By using flow cytometry to detect p-SAMHD1 levels in peripheral blood monocytes, may enable the development of an early warning method to predict an individual’s future risk of metabolic diseases such as diabetes. Thereby providing new immunometabolic indicators for the early screening of metabolic diseases [8]. Third is the level of ox-mtDNA. As a downstream product of mitochondrial oxidative damage and NLRP3 inflammasome activation, quantifying ox-mtDNA in plasma or serum may provide a biomarker for assessing the efficacy of anti-inflammatory treatments [57] (Figure 1).

Limitations

The present study has several limitations. First, the core mechanism linking SAMHD1 phosphorylation to NLRP3 hyperactivation has been examined primarily in in vitro macrophage models and gene-knockout mice; in vivo interventions studies and exploration in other immune cells or metabolically active tissues are still lacking; Second, the hypothesis that SAMHD1 inactivation in early obesity represents a “metabolic adaptation” and “tipping point” still requires longitudinal time-series data to elucidate the dynamic changes of relevant molecules during disease progression; Third, the physiological and pathological thresholds for dNTP accumulation have not been quantitatively defined, and no concentration-response analysis has yet identified the critical point at which metabolic support gives way to inflammatory pathology, are therefore needed chemogenetic tools capable of titrating dNTP concentrations; Fourth, the clinical translation of PNC1/2 inhibition faces safety risks, including off-target mitochondrial toxicity and weakened anti-infective immunity, which call for systematic evaluation; Fifth, remains absent direct experimental evidence linking this pathway to other aseptic inflammatory conditions such as gout; the regulatory roles of interindividual factors—such as sex, age, and genetic polymorphisms—on the pathway, as well as the heterogeneity of macrophage subsets, remain unclear and require further analysis using single-cell metabolomics and spatial transcriptomics.

Conclusion

The study by Liu et al., represents an important advance in understanding obesity-driven inflammation.It reveals a pathway linking SAMHD1 inactivation-induced dNTP accumulation, PNC1/2-mediated mitochondrial import, uncontrolled mtDNA synthesis, and NLRP3 inflammasome activation triggered by ox-mtDNA, thereby establishing a coherent molecular chain. An important implication of this work is that metabolic reprogramming is not inherently pathological; whether it leads to inflammatory damage or supports tissue repair likely depends on the precision of spatiotemporal control and the intensity thresholds of nucleotide flux. Based on this, the proposed double-edged sword model offers a framework for understanding these dynamics: in early obesity, the reversible inactivation of SAMHD1 may represent a metabolic adaptation that helps macrophages in maintaining essential functions within a nutrient-rich environment; however, when prolonged metabolic stress forces this reprogramming past a critical threshold, the adaptive response may evolves into an uncontrolled inflammatory positive feedback loop. Precisely identifying this tipping point from adaptation to imbalance and elucidating the underlying molecular switches remains a central challenge for future research in future research in this field. Furthermore, targeting the mitochondrial dNTP transporters PNC1/2 could specifically inhibit NLRP3 inflammasome activation, offering a potential intervention strategy for various sterile inflammatory diseases; however, this strategy is itself a double-edged sword. Future studies should prioritize macrophage-specific delivery, precise intervention at defined disease stages, and targeted anti-inflammatory approaches based on biomarkers such as circulating dNTPs, SAMHD1 phosphorylation status, and ox-mtDNA, in order to advance clinical translation within safe boundaries.

Author contributions

Zijing Li and Maoyuan Wang conceived and designed this review. Zijing Li was responsible for literature search, manuscript drafting, and the creation of mechanism diagrams. All authors reviewed and revised the manuscript. All authors approved the final draft.

Funding

National Natural Science Foundation of China(82560457), National Natural Science Foundation of China (82060420), Bureau of Science and Technology of Ganzhou Municipality (2023LNS37155).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

This research did not require ethical approval or participant consent.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.