Arginine depletion-induced autophagy and metabolic dysregulation are involved in the disease severity of hand, foot, and mouth disease

ABSTRACT

Amino acid metabolism provides significant insight into the development and prevention of many viral diseases. Therefore, the present study aimed to compare the amino acid profiles of hand, foot, and mouth disease (HFMD) patients with those of healthy individuals and to further reveal the molecular mechanisms of HFMD severity. Using UPLC-MS/MS, we determined the plasma amino acid expression profiles of pediatric patients with HFMD (mild, n = 42; severe, n = 43) and healthy controls (n = 25). Brain tissues from CVA6-infected mice were examined using untargeted metabolomics. Several amino acids were significantly different between the three groups. Pathway analysis revealed that arginine, proline, and tryptophan metabolism are implicated in the pathogenesis of HFMD. A similar arginine depletion was observed in the brain tissues of CVA6-infected mice. Importantly, L-arginine supplementation improved the survival rate of CVA6-infected mice, inhibited virus multiplication, and reduced pathological autophagy associated with mTOR-autophagy pathway in the brain. Collectively, arginine, as the hub amino acid metabolite of the mammalian target of rapamycin (mTOR) signaling pathway affecting autophagy, plays an important role in the pathogenesis of severe HFMD. L-arginine supplementation may serve as a potential therapeutic option for critical patients with HFMD.

Introduction

Hand, foot, and mouth disease (HFMD) is a prevalent viral disease caused by enteroviruses (EVs) that typically affects infants and young children [1]. Enterovirus A71 (EVA71) and Coxsackievirus (CV) A16 are always considered as the most common pathogens. However, an increasing number of HFMD outbreaks have been reported due to other EVs such as CVA6 and CVA10 [1]. Generally, this common childhood infection is self-limiting, but a minority of pediatric cases can develop serious complications [1]. Furthermore, recent studies have demonstrated that some patients with severe illness may be at risk for future neurodevelopmental delays and serious mental health issues such as mental retardation, depression, and anxiety [2,3]. Therefore, HFMD severity remains a major public health issue across the Asia-Pacific region and beyond [1]. Immune escape from EVs, immune imbalance, and excessive systemic inflammatory responses contribute to HFMD severity [1]. However, the intrinsic mechanisms involved in the severity are still poorly understood.

Small-molecule amino acids (AAs) are essential for the viral life cycle as they provide building blocks for assembling viral nucleic acids, capsid proteins, and membranes. Amino acid metabolism critically involved in enterovirus infection pathogenesis, facilitating diverse biochemical pathways that include alternative energy generation and the provision of precursors for the biosynthesis of proteins, lipids, and nucleotides [4]. Evidence suggests that that glutathione and its related metabolites, as well as several amino acids, such as glutamate and aspartate, changed significantly in accordance with the infectious dose of EV-A71-infection [5,6]. Recently, the researchers have revealed that the disturbances in arginine/ornithine metabolism was a pivotal factor in the initiation of cytokine storms observed in severe HFMD cases, and spermine effectively mitigated the inflammatory injury phenotype observed in murine models with severe HFMD [7]. Arginine, a non-essential amino acid, is pivotal in the protein synthesis and the urea cycle, serving as a precursor for a range of bioactive molecules, including glutamate, citrulline, and nitric oxide [8]. Moreover, arginine concentration is implicated in the activation of mTORC1, which is a key regulator of autophagy [9]. Consequently, targeting amino acid metabolism and its associated signaling pathways may represent innovative therapeutic avenues for the modulation of infection and immune responses. The current study is the first comprehensive AA analysis of the plasma of patients with HFMD. Through in vivo experiments, we identified unanticipated alterations in AA expression profile, particularly dysregulated arginine. We further demonstrated that L-arginine supplementation curbs the severity of HFMD by altering the mechanistic target of the rapamycin (mTOR)-autophagy pathway.

Materials and methods

Ethics approval and consent to participate

Written informed consent was obtained from the parents or guardians of study participants, including HCs. All procedures were performed according to the guidelines of the Life Science Ethics Review Committee of Zhengzhou University (ethical approval number: ZZUIRB2023–180). Studies involving human participants adhere to the Declaration of Helsinki (https://www.wma.net/policies-post/wma-declaration-of-helsinki-ethical-principles-for-medical-research-involving-human-subjects/). All animal experimental procedures were approved by the Life Science Ethics Review Committee of Zhengzhou University (ethical approval number: ZZUIRB2023–156). Studies involving laboratory animals follow the ARRIVE guidelines (Figshare: ARRIVE checklist, https://doi.org/10.6084/m9.figshare.27253218).

Study participants

This study included pediatric patients with HFMD who were hospitalized at the First Affiliated Hospital of Xinxiang Medical University between April and September 2017. The diagnosis and classification of cases followed the Guidelines for the Diagnosis and Treatment of Hand, Foot, and Mouth Disease (2010 Edition), issued by the China Ministry of Health [10]. The healthy controls (HCs) were determined to exhibit normal clinical indicators and were tested as pan-enterovirus-negative through both qRT-PCR and serological IgM antibody assays. Given that various infections, including dengue virus infection, tuberculosis, and sepsis, can disrupt amino acid metabolism, we excluded participants with such conditions prior to recruitment. Additionally, individuals with immune disorders, nephrotic syndrome, or those undergoing immunosuppressive therapy were also excluded from the study. HCs were recruited from Zhengzhou Central Hospital Affiliated to Zhengzhou University. Ultimately, eligible 42 mild, 43 severe, and 25 healthy individuals were included in this study (Supplemental Figure S1A).

Laboratory assessments and data processing

Plasma samples were collected from the participants using EDTA and sodium citrate anticoagulants. After centrifugation at 3000 rpm for 10 min (4 °C), the supernatant was aspirated, aliquoted into 500 μL centrifuge tubes, and stored at −80 °C for later use. AAs in plasma samples were measured via pre-column derivatization according to a previous report with minor modifications [11]. All AA standards were obtained from Sigma-Aldrich (St. Louis, MO, USA) and Steraloids Inc. (Newport, RI, USA) (Supplemental Figure S1B and S2A). Data were analyzed using the Bio Deep Platform (http://www.biodeep.cn).

Animal infection experiments

Specific pathogen free (SPF) grade ICR mice were purchased from the Medical Animal Center in Zhengzhou University, Henan, China. Animals were raised in stainless steel cages in the Medical Animal Center located in the College of Public Health of Zhengzhou University on a 12 h light/dark cycle and allowed free access to food and water. Ten-day-old ICR mice were inoculated intraperitoneally (i.p.) with a lethal dose of CVA6 strain (NCBI Accession number: OM179765) [12]. At 3 or 5 days post-infection (dpi), infected mice (n = 6 per group) and control mice were euthanized. The brains were removed to conduct untargeted metabolomics. Brain tissues were isolated under sterile conditions. The samples were washed with pre-cooled PBS to remove surface blood. Subsequently, the liquid on the surface of the tissues was absorbed using sterile absorbent paper. These samples were then rapidly frozen in liquid nitrogen and placed into a centrifugal tube for future use. All procedures were completed within 5 min. The raw data were uploaded at www.ebi.ac.uk/metabolights/MTBLS11341. L-arginine supplementation was performed at the appropriate dose (200 mg/kg) [13,14]. At 5 dpi, the blood was separated and assessed by the Arginine content assay kit (Solarbio) based on the manufacturer’s procedure. The viral loads of infected brain tissues were measured as described in previous study [12]. As for the animal experiment with EVA71 infection, the age of mice and dose of challenge were optimized based on our previous study [15]. Briefly, 10 days old mice were all died after being injected with a dose of 4.5 × 10⁷ TCID₅₀ EVA71 (NCBI Accession number: OP806304) via i.p. The clinical scores were recorded as described in our previous study [16]. Briefly, clinical scores were evaluated using the following criteria: 0, healthy; 1, lethargy and inactivity; 2, ataxia; 3, loss of weight; 4, hind limb paralysis; 5, dying or dead.

Metabolomics analysis

PCA, PLS-DA, and OPLS-DA were performed to visualize differences in metabolome profiles. Open databases, including the Human Metabolome Database, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway database, and MetaboAnalyst, were used to identify metabolic pathways. A p-value of less than 0.05 was deemed statistically significant, and the differential metabolites identified through multidimensional statistical analysis were subsequently validated. Variables with a Variable Importance in Projection (VIP) greater than 1 for multidimensional statistics, along with a p-value less than 0.05 for unidimensional statistics and a fold change of ≥1.5 or ≤0.667 between the two groups, were classified as significantly altered metabolites (or differentials). The MetaboAnalyst platform, in conjunction with the online Kyoto Encyclopedia of Genes and Genomes (KEGG) classifications, was employed for pathway analysis of these differential metabolites. The differential metabolites were mapped to their corresponding biochemical pathways. Pathway significance was determined based on both the total number of metabolites that map to a given pathway and their respective significances; a pathway was considered perturbed if it contained at least two significant metabolites (i.e. hits), an impact value of ≥0.10, and a raw p-value of less than 0.10. Raw p-values were calculated based on the number of hits relative to the total number of compounds within each pathway. An impact value equal to or exceeding 0.10 indicates that the altered pathway has a substantial effect. The MetaboAnalyst software package was used to enrich the differential metabolites (VIP >1, p < 0.05) under the established comparison strategy. The igraph software package was used to complete the plot after inputting the p-value and FC value of these differential metabolites in the enrichment results.

Histopathology

At 5 dpi, mice in the three groups were anesthetized, and their brains were removed and fixed in 4% paraformaldehyde at room temperature. Subsequently, the brains, which had been embedded in paraffin, were sectioned into 5 μm slices and stained using Nissl’s stain as well as hematoxylin and eosin (H&E).  The tissue sections were taken by dedicated slicer made in Leica Biosystems (Leica RM2125 RTS). Viral VP1 protein levels were determined by immunohistochemical (IHC) staining, as described previously [17]. For immunofluorescence staining, technical services were provided by Servicebio Biotech Co. Ltd. (Wuhan, China). To evaluate the degree of pathological lesions, two pathologists blinded to the treatment conditions obtained pathological scores using a modified scoring system [18].

Western blotting analysis

At 5 dpi, total proteins in the brain tissues were extracted using a protein extraction kit (Beyotime Institute of Biotechnology) with protease and phosphatase inhibitors. Total protein from each sample (10–20 μg) was separated using SDS-PAGE and transferred to PVDF membranes (Merck-Millipore, Germany). The bands were analyzed using ImageJ software. The following commercially available antibodies were used: anti-phospho-mTOR (Ser2448), anti-p62, anti-LC3B (Cell Signaling Technology, Inc.), and anti-β-actin (Proteintech Group, Inc.).

Transmission electron microscopy (TEM)

The brainstems of mice were removed and immersed in a special electron microscope fixator (2.5% glutaraldehyde, pH 7.0–7.5) for at least 2 h. Sample pre-treatment and transmission electron microscopy were carried out based on our previous study [19]. Electron microscopy is the most traditional and common method for monitoring autophagy. Autophagosomes are characterized by their double membranes and contain uncompact cytoplasmic material, including organelles such as mitochondria and ribosomes. In contrast, autolysosomes are vesicles limited by single or double membranes that contain densely compacted amorphous or multilamellar contents [20,21]. The number of autophagic vacuoles (including both autophagosomes and autolysosomes) per cell body was quantified using electron micrograph images at a direct magnification of 6000× across various treatment conditions. A minimum of 40 electron micrographs per treatment condition were randomly captured and analyzed, with numerical values expressed as mean ± SD per cell. The number of autophagosomes per cell, and autophagosomes per cell was calculated by two pathologists blinded to the treatment conditions [21].

Statistical analyses

The median with interquartile range (IQR), mean with standard deviation, and proportion (%) were calculated and analyzed using GraphPad Prism software 8.3. The Mann-Whitney U test or Student t-test was used to compare any two data sets, and the Kruskal-Wallis test or one-way ANOVA was used for comparisons of more than two groups. The χ² test was used for categorical variables. Spearman’s rank correlation test was used to analyze the correlations between the two variables. Survival rate was estimated using the Kaplan-Meier method, and any differences in survival rates were evaluated using a stratified log-rank test. The potential predictive ability of differential AAs for severe illness was tested using receiver operating characteristic (ROC) curves. Leave-one-out analysis of cross-validation was performed to verify the predictive ability. For all analyses, p value < 0.05 was considered significant.

Results

The AA expression profile varies with the severity of HFMD.

We determined the plasma AA profiles in 85 children with HFMD and 25 hCs. The demographic and clinical characteristics of the study participants are presented in Supplemental Table S1. As illustrated in the heatmap presented in Supplemental Figure S2B, three primary clusters of distinct metabolites were identified. Additionally, it was observed that a significant number of amino acid metabolic pathways exhibit differential expression across various types of plasma samples. PCA and PLS-DA analyses of the obtained AA profiles could not clearly distinguish different participants (Supplemental Figure S2C and SD). Based on the obtained AA profiles, the OPLS-DA model demonstrates superior performance compared to PCA and PLS-DA analyses in differentiating between patients with HFMD and HCs (Supplemental Figure S2E). Although it was unable to distinctly separate severe cases from mild cases, there were notable differences in the distribution of certain individuals within the two groups. This suggests that specific AA metabolites may be linked to clinical deterioration in HFMD patients. Correlation plots revealed apparent positive correlations among different AAs of both mild and severe groups, suggesting they had coordinated variation after infection (Supplemental Figure S2F). Taken together, our results provide a possible that the key single biomarkers or biomarker combinations involved in the pathogenic pathways might be found from the amino acid metabolite data.

Precise plasma AA concentrations of participations measured by targeted metabolomics

We exhibited the AA metabolites concentrations (with statistical difference between groups) in Figure 1a. The AA metabolites whose concentrations have no statistical difference were shown in this Supplementary Figure S3. Compared with HCs, the plasma concentrations of alanine were lower in both mild and severe patients. In addition, compared to HCs, severe patients had enriched taurine (23.21%) and depleted proline (21.65%), citrulline (17.62%), tryptophan (16.12%). Aminoadipic acid, arginine, proline, kynurenine, and isoleucine were also depleted in severe cases compared to mild cases. The degrees of decline in aminoadipic acid (29.77%) and arginine (19.09%) were the top two. Our results also showed that the tyrosine/phenylalanine ratio slightly decreased in patients with severe disease (Figure 1b). We observed a significant reduction in the valine/glycine ratio in severe patients compared to either HCs or mild patients (Figure 1c). Correspondingly, the Pearson’s correlation coefficient between glycine and valine plasma concentrations indicated an imbalance in valine/glycine metabolism (Figure 1d–f). Together, these results indicate that pediatric patients with HFMD exhibit perturbed expression profiles of AAs, especially those with severe illness.

Figure 1.

Precise plasma AA concentrations in the plasma of study participants. (a) The concentrations of AA metabolites with statistical difference between groups. The changes of these metabolites were also summed up and displayed at the bottom. Dot plot depicting mean for the tyrosine/phenylalanine concentration ratio (b) and the valine/glycine concentration ratio (c), and scatterplots between tyrosine (y-axis) and phenylalanine (x-axis) in each healthy control (d), mild case (e) and severe case (f). *p < 0.05, **p < 0.01, ***p < 0.001, ns: no statistical difference.

Arginine metabolism alteration in severe paediatric patients with HFMD

Metabolic pathway analysis of altered AAs was performed using MetaboAnalyst 3.0, an online tool [22]. As shown in Figure 2a–d, the most significantly enriched metabolic pathways were identified, including arginine and proline metabolism, tryptophan metabolism, taurine, and hypo-taurine metabolism. Arginine and proline metabolism and arginine biosynthesis were the most dysregulated pathways during the course of HFMD (Figure 2a–d). These dysregulated metabolic pathways contained most of the differential AAs (Figure 2e), emphasizing a strong association between metabolic imbalance and disease severity. Together, we speculated that altered arginine metabolism could potentially promote the viral pathogenic process and worsen patient outcomes.

Figure 2.

Arginine metabolism in severe paediatric patients with HFMD. Under different comparison schemes, (a) HCs vs mild, (b) HCs vs severe, (c) HCs vs all HFMD patients, (d) mild vs severe, scatter plots of the most relevant metabolic pathways from KEGG library arranged by adjusted p values on the y-axis, and pathway impact values on the x-axis. (e) Schematic map of the arginine and taurine metabolism (left), tryptophan metabolism (right).

The potential of plasma AAs as possible indicators of the disease severity of HFMD

Compared to mild cases, five AAs with significant changes were found in severe cases, including aminoadipic acid (Figure 3a), proline (Figure 3b), isoleucine (Figure 3c), kynurenine (Figure 3d), and arginine (Figure 3e). A generally accepted approach suggests an area under the ROC curve (AUC) that is greater than 0.60 but less than 0.75 reflects possibly helpful discrimination [23]. Here, we utilized ROC curve based on single index feature to evaluate the performance of discriminating severe and mild patients. In the ROC analysis, these five AAs had an AUC above 0.6, which indicated the potential predictive value of severe illness. Based on these five AAs, we used binary logistic regression to create a ROC joint curve, and the AUC was 0.799 (Figure 3f). In the Bayesian model and artificial neural network analysis, these combinations of differential AAs contributed significantly to distinguishing patients with severe HFMD (Figure 3g,h). The corresponding AUCs are 0.811 and 0.927, respectively. Notably, the classifier for the distinction based on random forest analysis achieved nearly perfect discrimination (Figure 3I). However, the holdout cross-validation analysis showed that the AUC based on logistic regression was the highest (0.713) (Figure 3f). Overall, these five differential plasma AAs have potential applications in determining HFMD prognosis.

Figure 3.

The potential of plasma AAs as indicators for severe outcomes related to HFMD. The diagnosis power of aminoadipic acid (a), proline (b), isoleucine (c), kynurenine (d), and arginine (e) were evaluated by ROC analysis. The Logistic Regression model (f), bayesian model (g), artificial neural networks (h), random forest model (i), were applied to enhance the diagnosis power of plasma AAs. In each panel, the left is the ROC of diagnostic efficiency and the right is the ROC of the cross validation.

Involvement of arginine depletion and metabolic dysregulation in the development of HFMD severity in a mouse model

Given that CVA6 has emerged as a major cause of HFMD [24], to verify these results, further experiments were conducted using a CVA6 infection mouse model. The score plot analysis successfully discriminated between different groups of mice based on the altered metabolites (Figure 4a–c). Some differential metabolites with statistical differences are presented in the scree plot (Figure 4d). The heatmap depicted an obvious increase and decrease in metabolites in the mouse brain (Figure 4e). For further comparison, the altered metabolites were presented and analyzed in separate panels. As shown in Figure 4f, the levels of several metabolites (a, L-Arginine; b, L-Proline; c, γ-Aminobutyric acid; d, Aminoadipic acid; e, L-Glutamic acid; f, L-Methionine; g, L-Tyrosine; h, L-Aspartic acid) in both infected groups were significantly reduced. Many of the abnormally depleted AAs, such as arginine, were consistent with the results found in patients with HFMD (Figure 1a). Notably, enrichment of the arginine biosynthesis pathway was in accordance with that observed in the study population (Figure 2d, 4g). The mTOR signaling pathway was specifically enriched, which was further confirmed by the correlation network analysis (Figure 4h). Therefore, we hypothesized that arginine in the brain as a hub metabolite of the mTOR signaling pathway plays an important role in the development of HFMD severity.

Figure 4.

Involvement of arginine depletion and metabolic dysregulation in the development of HFMD severity in a mouse model. Non-targeted metabolomics results showed score plot of PCA (a), PLS-DA (b), OPLS-DA (c) analysis. (d) Scree plot of metabolites. The horizontal axis (M/Z) represents the ratio of the mass number to the charge number of electric particles. (e) Heatmap visualization and clustering analysis of differential metabolites. (f) The normalized intensity levels of the down-regulated metabolites. (h) Network analysis of differential metabolites. n = 6 per group. **p < 0.01, ***p < 0.001, ****p < 0.0001.

L-arginine supplementation could reduce viral replication and pathological autophagy associated with HFMD

Firstly, we assessed the efficacy of L-arginine supplementation based on a mouse model of EV71 infection (Supplemental Figure S4A). Our results showed that these infected mice rapidly displayed clinical manifestations, such as limb weakness and paralysis, and all died within 9 dpi (Supplemental Figure S4B-D). However, a third of these infected mice treated with L-arginine were rescued from fatal outcomes although they also exhibited distinct clinical symptoms. To further evaluate the benefits of arginine supplementation, based on a CVA6 infection mouse model, an L-arginine supplementation experiment was also performed as described in Figure 5a. CVA6–infected mice with L-arginine supplementation exhibited a slight loss of body weight (Figure 5b), relatively lower clinical scores (Figure 5c), and improved survival rates (Figure 5d). We also measured the serum L-arginine levels from the three groups and found it was elevated significantly along with L-arginine supplementation in CVA6 infected mice at 5 dpi (Figure 5e). We also found that L-arginine supplementation alleviated CVA6-induced brain damage (Figure 5f,g), which was in accordance with clinical observations. The viral loads in the brains of CVA6 infected mice were also significantly decreased after supplementing L-arginine (Figure 5h). Further experiments showed that L-arginine supplementation reduced CVA6-induced autophagy by increasing the expression of p-mTOR and p62 in brain cells from mice infected with CVA6 (Figure 5i,j). Transmission electron microscopy (TEM) is the main tool used to observe the physiological processes of autophagy. As shown in Figure 6a, CVA6 infection led to an increase in the number of autophagosomes and lysosomes, whereas L-arginine supplementation inhibited pathological autophagy in brain cells from mice with CVA6 infection (Figure 6b,c). Collectively, the notable beneficial effect of L-arginine supplementation is demonstrated to be associated with the reduced viral replication and pathological autophagy.

Figure 5.

In vivo experiments to assess the benefits of L-arginine supplementation on HFMD severity. (a) The experimental design of L-arginine intervention. Body weight (b), mean clinical scores (c) and survival rates (d) of mice were monitored and recorded daily (n = 8 per group). (e) The serum L-arginine levels of three groups after supplementing L-arginine at 5 dpi (n = 6 per group). (f) Pathological changes of brain slices and the distribution of VP1 antigens (×400). Histopathological score (g) and the viral loads (h) in the brains of mice (n = 3–4 per group). (i) Colocalization of LC3B and VP1 in brain slices was observed by immunofluorescence assay. The white arrows indicated cells with two proteins colocalization. (j) The expression levels of p-mTOR, p62, LC3B. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 6.

The effect of L-arginine supplementation on pathological autophagy. (a) N: nucleus. Lysosomes (red arrows): monolayer organelles with high electron density. Autophagosome (green arrows): vesicles with double-membrane structures engulfing cytoplasmic material. The scale bar has been indicated in the figures. The number of autophagy lysosome (b) and autophagosome (c) (n = 3 per group). *p < 0.05, **p < 0.01, ***p < 0.001.

Discussion

This study examined the serum AA profiles of HFMD patients with different severities, compared them to the profiles of HCs, and identified new targets for both diagnosis and treatment. The subsequent L-arginine supplementation experiment provided more evidence that AAs could affect disease progression and outcome by affecting autophagy signaling pathways.

Several studies have shown that HFMD patients exhibit metabolic abnormalities compared to healthy individuals [25–27]. The present study reached similar conclusions from the perspective of AA expression profiles. Disturbances in the AA expression profile were closely associated with clinical deterioration. Further analysis showed a decrease of approximately 25% in alanine levels in both mild and severe cases, suggesting a likely increase in alanine catabolism. Since alanine may serve as an ammonia nitrogen carrier between glutamatergic neurons and surrounding astrocytes, its depletion could affect astrocyte function [28]. In patients with severe HFMD, taurine levels were enriched, whereas citrulline levels were depleted. We hypothesized that the elevated levels of taurine might be linked to the activation of astrocytes in the brain [29]. Additionally, taurine is found at high concentrations in leukocytes, which is believed to affect immune cell function [30]. Increased levels of taurine may be a contributing factor to altered immune responses in patients with HFMD. Maintaining an adequate citrulline balance is crucial for clearing ammonia via the urea cycle. Consequently, decreased citrulline levels may reflect abnormal metabolic activity in the urea cycle [31]. Additionally, as citrulline acts as an innate immune signaling metabolite, the depletion of cellular citrulline activates inflammatory macrophages and triggers the immune response [32]. Compared to mild cases, the levels of aminoadipic acid and arginine in severe cases experienced the greatest decrease. Aminoadipic acid is an intermediate metabolite of lysine and tryptophan, and the reduction in adipic acid correlates with lower levels of lysine and tryptophan. The tryptophan degradation pathway results in the emergence of kynurenine, which is a vital molecule linked to inflammatory reactions and immune responses [33]. Heightened pro-inflammatory cytokines are liable to accelerate the degradation of tryptophan. It is imperative to identify high-risk patients who may develop severe HFMD at an early stage. In the present study, AAs profiles also had potential diagnostic separation power to differentiate severe HFMD patients from HCs and mild patients. Although mild and severe cases cannot be clearly distinguished from the overall AA metabolic profiles, differences in some amino acid metabolites are also evident. We can distinguish mild and severe cases based on these differential AA metabolites. Moreover, the combined strategy may outperform single index schemes.

Arginine and proline metabolism and arginine biosynthesis were the most dysregulated pathways during HFMD. In a CVA6 infection mouse model, the mTOR signaling pathway was specifically enriched in the brains of infected mice. Moderate autophagy can exert antiviral functions during diverse viral infections by selectively targeting viral particles or components to lysosomes for degradation [34]. As a central component of autonomous innate immunity, autophagy defends individual cells against enterovirus infection [35–37]. Nevertheless, excessive formation and accumulation of autophagosomes are beneficial for the replication of viral strains in nerve cells. CVA6 can trigger cell death through autophagy, thereby contributing to the pathogenesis of CVA6 strains [38]. The downregulation of autophagy has been identified as a potential strategy to inhibit enterovirus infection [39,40]. L-arginine is considered semi-essential in mammals, and the requirements exceed the production capacity of the organism during infection. L-arginine metabolism is a complex biological process. Beyond its direct impact on autophagy via concentration-dependent mechanisms, it also serves as a substrate for several key enzymes. Consequently, L-arginine can simultaneously affect immune functions, intraluminal metabolism, intestinal microbiota composition, and microbial pathogenesis [41]. Therefore, L-arginine may directly inhibit enterovirus replication either through direct modulation of host immune responses or via secondary metabolites [41,42]. Reduced viral replication, in turn, could potentially lead to decreased autophagy. Collectively, enterovirus infection and autophagy connect with each other, influence with each other and make cause and effect with each other. However, the mechanisms underlying the observed improvements in survival and clinical outcomes in infected mice following L-arginine supplementation remain inconclusive. While the exact mechanisms are yet to be fully elucidated, it is evident that L-arginine exerts beneficial effects. Further research is required to unravel the core mechanisms involved.

The PI3K/Akt/mTOR pathway plays a vital role in the regulation of autophagy, and suppressing mTOR activity leads to the initiation of autophagy and formation of autophagosomes [43]. The activity of mTOR complex 1 (mTORC1) is influenced by a variety of signals such as AAs [44]. In the presence of arginine, the direct binding of this AA to the cellular arginine sensor for mTORC1 (CASTOR1) induces the activation of mTORC1 [45]. The elevated expression levels of p-mTOR and p62, as well as reduced levels of LC3B, resulted from L-arginine supplementation. Furthermore, L-arginine treatment decreased the number of autophagosomes in the present study. Autophagosomes are the intermediate products of the autophagic process, so a decrease in autophagosomes may arise due to diminished autophagosome production [46]. Therefore, L-arginine supplementation could alleviate nerve damage and improve the survival rate of infected mice, at least partially, by inhibiting autophagy.

Conclusion

Arginine, a hub AA metabolite of the mTOR signaling pathway that affects autophagy, plays an important role in the pathogenic mechanism of severe HFMD. L-arginine supplementation may serve as a potential therapeutic option for critical patients with HFMD.

Abbreviations

HFMD

Hand, foot, and mouth disease

CVA6

Coxsackievirus A6

mTOR

mechanistic target of rapamycin

EVA71

Enterovirus A71

AA

amino acid

i.p.

intraperitoneal

dpi

days post infection

IHC

immunohistochemistry

IQR

interquartile range

HC

healthy control

PCA

principal component analysis

PLS-DA

partial least squares discriminant analysis

OPLS-DA

orthogonal projections to latent structure-discriminant analysis

KEGG

Kyoto Encyclopedia of Genes and Genomes

ROC

receiver operating characteristic

AUC

area under the ROC curve

mTORC1

mTOR complex 1

CASTOR1

cellular arginine sensor for mTORC1

Funding Statement

This work was supported by the National Natural Science Foundation of China [No. 82002147, No. 82372229, No. 82273695, and No. 82073618], China Postdoctoral Science Foundation [No. 2019M662543], and the Open Research Fund of the National Health Commission Key Laboratory of Birth Defects Prevention & Henan Key Laboratory of Population Defects Prevention [No. ZD202301], the Open Project of the Henan Province Engineering Research Center of Diagnosis and Treatment of Pediatric Infection and Critical Care [NO. ERC202302], supported by an Open Grant from the Pingyuan Laboratory [NO. 2023PY-OP-0202]. The funders had no role in the study design, data collection and analysis, decision to publish, or manuscript preparation. No author received a salary from any of the funding agencies.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Authors’ contributions

Yuefei Jin, Wangquan Ji, and Liang Zhang contributed equally to the study. Yuefei Jin: Software, Data curation, formal analysis, funding acquisition, Investigation, Methodology, Writing, review, and editing. Wangquan Ji: Data curation, formal analysis, Investigation, Validation, Methodology, Writing of the original draft. Liang Zhang: Formal analysis, Investigation, Writing of original draft. Dejian Dang: Software, Visualization. Bingqing Yu: Investigation, Software. Xiaolong Zhang: Visualization. Yuxiang Zhang: Investigation, Validation. Jiaqi Li: Software, Investigation. Yaodong Zhang: Validation and Visualization. Rongxin Yang: Formal analysis, Methodology, Resources. Haiyan Yang: Methodology, Visualization. Shuaiyin Chen: Funding Acquisition and Methodology. Fang Wang: Conceptualization, Project administration, writing review, and editing. Guangcai Duan: Conceptualization, Funding acquisition, project administration, Supervision, and editing. All authors have read and approved the final version of the manuscript.

Data Availability Statement

The raw data that support the findings of this study are openly available in Figshare under the https://doi.org/10.6084/m9.figshare.27253218. The data generated during the mouse brain tissue metabolism study is available at the “MetaboLights,” unique persistent identifier MTBLS11341 (www.ebi.ac.uk/metabolights/MTBLS11341).

Consent for publication

All participants including HCs approved the submission of the manuscript to this journal.

Supplemental data

Supplemental data for this article can be accessed online at https://doi.org/10.1080/21505594.2024.2440541