Arginine metabolic endotypes related to asthma severity

Aims Arginine metabolism via inducible nitric oxide synthase (iNOS) and arginase 2 (ARG2) is higher in asthmatics than in healthy individuals. We hypothesized that a sub-phenotype of asthma might be defined by the magnitude of arginine metabolism categorized on the basis of high and low fraction of exhaled nitric oxide (FENO). Methods To test this hypothesis, asthmatics (n = 52) were compared to healthy controls (n = 51) for levels of FENO, serum arginase activity, and airway epithelial expression of iNOS and ARG2 proteins, in relation to clinical parameters of asthma inflammation and airway reactivity. In parallel, bronchial epithelial cells were evaluated for metabolic effects of iNOS and ARG2 expression in vitro. Results Asthmatics with high FENO (≥ 35 ppb; 44% of asthmatics) had higher expression of iNOS (P = 0.04) and ARG2 (P = 0.05) in the airway, indicating FENO is a marker of the high arginine metabolic endotype. High FENO asthmatics had the lowest FEV1% (P < 0.001), FEV1/FVC (P = 0.0002) and PC20 (P < 0.001) as compared to low FENO asthmatics or healthy controls. Low FENO asthmatics had near normal iNOS and ARG2 expression (both P > 0.05), and significantly higher PC20 (P < 0.001) as compared to high FENO asthmatics. In vitro studies to evaluate metabolic effects showed that iNOS overexpression and iNOS+ARG2 co-expression in a human bronchial epithelial cell line led to greater reliance on glycolysis with higher rate of pyruvate going to lactate. Conclusions The high FENO phenotype represents a large portion of the asthma population, and is typified by greater arginine metabolism and more severe and reactive asthma.


Conclusions
The high F E NO phenotype represents a large portion of the asthma population, and is typified by greater arginine metabolism and more severe and reactive asthma. a1111111111 a1111111111 a1111111111 a1111111111 a1111111111

Introduction
Asthma is a chronic inflammation of the airways that is characterized by airway reactivity. Nitric oxide (NO), generated by inducible NO Synthase (iNOS; EC 1.14. 13.39) expressed in the airway epithelium, where it catalyzes the conversion of arginine to NO and citrulline, is typically higher in asthmatics than in healthy populations [1][2][3][4]. We [1,5,6] and others [7][8][9][10][11][12][13] found that levels of arginine and arginase 2 (ARG2; EC 3.5.3.1), which catabolizes arginine to ornithine and urea, are also higher in the airway of asthmatics as compared to healthy controls. Recently we reported that sustained NO production in asthma is dependent upon the cell autonomous citrulline-arginine-NO cycle [6]. ARG2, by delivering ornithine into the mitochondria, provides nitrogen for the citrulline-arginine-NO cycle in order to sustain the high NO production while also supporting cellular bioenergetics and the inflammatory state [6].
In this context, although NO is higher in asthmatic populations, there is a wide variability in fraction of exhaled NO (F E NO) in individuals with asthma; some have F E NO levels in a normal range. Little is known of the high and low F E NO endotypes, and the bioenergetic effects of high arginine metabolism on asthma. Guidelines for asthma care define a cut-point of 35 ppb F E NO for asthma control and assessment of inflammation [4,14]. Here, we hypothesized that a sub-phenotype of asthma could be defined by the magnitude of arginine metabolism identified on the basis of high and low F E NO levels.

Study population
The clinical characteristics of the study population are listed in Table 1. Some of the study subjects had participated in studies reported previously [5,6]. Asthma was verified based upon American Thoracic Society guidelines, which include positive methacholine challenge test and/or reversible airflow obstruction. The severity of asthma was classified as mild intermittent/persistent (mild), moderate persistent (moderate) and severe persistent (severe) based on National Asthma Education and Prevention Program (NAEPP) guidelines [15]. Healthy controls lacked cardiopulmonary symptoms and had normal spirometry and negative methacholine challenge. Exclusion criteria for both asthmatics and healthy controls included age less than 18 years, pregnancy, current smoking, smoking within the past year, or former smokers with ! 5 pack-year total history. Spirometry was performed with an automated spirometer, and F E NO was measured by an online method at a constant flow rate of 50 ml/second according to the standards published by the American Thoracic Society. A subgroup of participants underwent bronchoscopy for endobronchial brushing and for bronchoalveolar lavage (BAL). All studies were approved by the Cleveland Clinic Institutional Review Board (IRB # PPG8351). All subjects were recruited from Cleveland Clinic and gave written informed consent by signing a consent document approved by the Cleveland Clinic Institutional Review Board.

Cell culture
BET1A cells, a human bronchial cell line transformed by SV40 T antigen, were cultured in serum-free LHC-9 medium (Thermo Fisher Scientific) on pre-coated plates.

Radioisotope studies of glucose metabolism in cells
The rate of oxidation of glucose to CO 2 and the rate of glycolysis (glucose to lactate) were measured by incubating 1−2×10 6 BET1A cells in LHC9 medium containing 6 mM glucose with [ 14 C]-glucose in an atmosphere of 95% O 2 and 5% CO 2 in an airtight Erlenmeyer flask [6]. Cells were incubated for 3 hours. The CO 2 generated was flushed by adding sodium bicarbonate and sulfuric acid into the medium. CO 2 trapped in hyamine was counted in a scintillation counter. Lactate was separated by ion exchange chromatography and radioactivity in the isolated organic acids, mostly lactate, was measured using a scintillation counter. The rate of conversion of glucose to lactate was calculated using precursor-product relationship.

Statistics
Data are shown as mean ± SEM. All statistical comparisons were performed using the Student's t-test, Wilcoxon nonparametric analyses or ANOVA as appropriate. Relationships between groups were analyzed using Multivariate Pairwise Correlation. The level of significance for P was chosen at 0.05. All data were analyzed with statistical program JMP Pro 10 (SAS Institute, Cary, NC).

Clinical population
The study population included 51 healthy controls and 52 individuals with asthma. The clinical characteristics of the study subjects are displayed in Table 1. Asthmatics had airway reactivity to methacholine and mild to moderate airflow limitation as measured by forced expiratory volume in 1 second (FEV 1 ) and the ratio of FEV 1 to forced vital capacity (FVC)(FEV 1 % predicted, Control 96 ± 1, Asthma 82 ± 2, P < 0.001; FEV 1 /FVC, Control 0.80 ± 0.01, Asthma 0.74 ± 0.01, P = 0.0002). Asthmatics were divided into mild, moderate and severe asthma according to FEV 1 (FEV 1 % predicted, mild, ! 80, n = 27, moderate, > 60 but < 80, n = 18, severe, 60, n = 3) [15]. Individuals with severe asthma had the highest body mass index (BMI)(ANOVA P = 0.03) and worst lung function (P < 0.001) as compared with mild and moderate asthmatics (Table 1). Asthma was stable and controlled by inhaled corticosteroids without recent exacerbations and without recent systemic corticosteroids use (Table 1). Not all individuals underwent all experimental studies; the numbers of subjects assessed are provided with each experiment and result. Medications were withheld for one day prior to testing.

Arginine metabolic endotypes related to clinical asthma phenotypes
Previously, we [1-3, 5, 6] and others [7][8][9][10][11][12] reported that asthmatics had higher levels of serum arginase activity and had higher F E NO. Using untargeted analysis, we observed a nitric oxide-related metabolomic endotype in asthma [3]. We also reported that the high rate of NO production was dependent upon induction of ARG2 and upon cell autonomous arginine-NOcitrulline cycle [6]. To examine whether arginine metabolic endotypes were informative of clinical asthma phenotypes, F E NO levels, serum arginase activity, and the expression of iNOS and ARG2 in freshly obtained airway epithelium were evaluated in asthmatics and healthy controls. As reported previously [1][2][3][4][5], asthmatics had significantly higher F E NO and greater serum arginase activity when compared with healthy controls [F E NO ppb, Control 19 ± 1, n = 51, Asthma 43 ± 5, n = 52, P < 0.001; Arginase activity μmol/ml/h, Control 0.30 ± 0.07, n = 3, Asthma 0.52 ± 0.08, n = 14, P = 0.05](S1 Table). Both iNOS and ARG2 proteins were generally expressed at higher levels in the airway epithelium of asthmatics compared with healthy controls [iNOS protein expression relative to cytokeratin expression, control 1.1 ± 0.2, n = 5, asthma 23.9 ± 13.8, n = 10, P = 0.01; ARG2 protein expression relative to cytokeratin expression, control 1.0 ± 0.1, n = 5, asthma 3 ± 1.3, n = 9, P = 0.03](S1 Table). Severe asthmatics had the highest F E NO levels (ANOVA P = 0.003), but similar expression of iNOS and ARG2 proteins in the airway epithelium (both ANOVA P > 0.05) as compared to individuals with mild and moderate asthma (S1 Table). Similar to previously reported suppressive effects of corticosteroids on iNOS/NO [1], asthmatics using inhaled corticosteroids tended to have lower F E NO levels than asthmatics not on corticosteroids (P = 0.17)(S2 Table). However, serum arginase activity and airway expression of iNOS and ARG2 proteins were similar between asthmatics on corticosteroids and those not on corticosteroids (all P > 0.05)(S2 Table). As previously reported [4,5], F E NO was inversely related to lung function in asthma  Table 2). ARG2 expression and arginase activity were not related to lung function in asthmatics (all P > 0.05)( Table 2), which is consistent with our prior studies [5,6]. F E NO levels were inversely related to lung function in individuals grouped by mild or severe asthma (P < 0.05), but not in individuals with moderate asthma (P > 0.05)(S3 Table). In asthmatics on corticosteroids, F E NO levels were inversely related to lung function (P < 0.05). IgE levels were positively correlated with the expression of iNOS in the airway epithelium in asthmatics not on corticosteroids (P = 0.01)(S4 Table).
Asthmatics had higher levels of F E NO than controls (P < 0.001), but 56% of asthmatics actually had F E NO within the normal range. When we stratified asthmatics by high and low F E NO [high F E NO, ! 35 ppb, n = 23, low F E NO, < 35 ppb, n = 29] [4,14], asthmatics with high F E NO had higher expression of iNOS (P = 0.04) and ARG2 (P = 0.05) in the airway epithelium (Table 3) (Fig 1). These data suggest that F E NO can be used as a marker of high arginine metabolism in asthmatic airways. Serum arginase activity, although higher in asthmatics, was not significantly different in asthmatics with high or low F E NO (P = 0.12)( Table 3) (Fig 1). With all three severe asthmatics stratified to high F E NO phenotype, high F E NO asthmatics tended to be more severe than low F E NO asthmatics (Asthma severity, mild/moderate/severe, low F E NO, 17/11/0, high F E NO, 10/7/3, P = 0.06)( Table 3). Because arginine metabolism influences cellular bioenergetics [6,20], we measured the concentration of lactate in the bronchoalveolar lavage (BAL). Asthmatics with high F E NO had higher levels of lactate in the BAL (ANOVA P = 0.02), lower FEV 1 % predicted (ANOVA P < 0.001), lower FEV 1 /FVC (ANOVA P = 0.0002) and lower PC 20 (ANOVA P < 0.001) as compared to low F E NO asthmatics or healthy controls. Importantly, asthmatics with low FeNO, whose airway expression of iNOS and ARG2 were similar to healthy controls (all P > 0.05), had significantly higher PC 20 (P < 0.001) and lower IgE (P = 0.04) as compared to high F E NO asthmatics (Table 3) (Fig 1). There was no difference in proportion of asthmatics receiving corticosteroids or not on corticosteroids between high F E NO and low F E NO groups (P > 0.05)( Table 3).
These findings indicate that high arginine metabolism in asthma is associated with the high F E NO phenotype, which is characterized clinically by greater airflow obstruction and airway reactivity and more severe asthma.

Greater reliance on glycolysis in bronchial epithelial cells with iNOS overexpression
In our recent study [6], we reported that greater arginine flux through ARG2 can support cellular mitochondrial bioenergetics. In order to understand the impact of increased arginine metabolism via high expression of iNOS or high co-expression of iNOS and ARG2 in asthmatic bronchial epithelial cells, we analyzed immortalized human bronchial epithelial cells (BET1A) in vitro. BET1A cells were transiently transfected with iNOS expression vector, ARG2 expression vector, co-transfected with iNOS vector and ARG2 vector (iNOS+ARG2), or control vector. iNOS expression is not detectable in BET1A cells at baseline culture [21]. When transfected with iNOS vector alone or co-transfected with iNOS and ARG2 vectors, high-level expression of iNOS and ARG2 were observed (Fig 2A). High nitrate levels accumulated in the culture medium of cells transfected with iNOS vector or with iNOS+ARG2 vectors, and this was suppressed by N-nitroarginine methyl ester (L-NAME), a non-selective inhibitor of NO synthase, confirming iNOS activity [nitrate μM, vector alone 2.6 ± 0.1, n = 26, iNOS 52.5 ± 3.3, n = 15, iNOS+L-NAME 3.6 ± 0.2, n = 4, iNOS+ARG2 41.0 ± 3.3, n = 22, iNOS +ARG2+L-NAME 5.6 ± 0.9, n = 9, ANOVA P < 0.001].

Discussion
The presence of high levels of iNOS, which produces NO and citrulline from arginine, and of arginase, which converts arginine to ornithine and urea, suggests that some asthma phenotypes are in a state of very high arginine catabolism. Important to identification of this high arginine metabolic sub-phenotype, we found that asthmatics with high F E NO had higher expression of iNOS and higher expression of ARG2 in the airway epithelium than the low F E NO group. As compared to low F E NO asthmatics, high F E NO asthmatics had a more severe clinical phenotype, which confirms prior report of high F E NO being an at-risk phenotype for reactivity and exacerbations of asthma [4]. The high F E NO group had higher airway lactate levels, suggesting other metabolic changes may be present in this group. The in vitro studies of immortalized bronchial epithelial cells showed that high co-expression of iNOS and ARG2 shifted cellular metabolism to greater oxidative metabolism of glucose and higher rate of ATP production. Altogether, the high F E NO group is characterized by greater arginine metabolism and greater oxidative glucose metabolism, which is associated with a more severe clinical phenotype.
Inhaled corticosteroids are considered to be the most effective medications for asthma control. Previously we reported suppressive effects of corticosteroids on iNOS induction [1]. We [14,23] and others [24][25][26][27] have shown that F E NO levels decrease when asthmatics are treated with corticosteroids. Here, asthmatics receiving inhaled corticosteroids tended to have lower F E NO levels than asthmatics not on corticosteroids, however, there was no difference in proportion of asthmatics receiving corticosteroids or not on corticosteroids between high F E NO and low F E NO groups, suggesting that high F E NO and low F E NO groups are phenotypes which represent arginine metabolic endotypes and not only a reflection of corticosteroid use.
The greater reliance of iNOS-expressing cells on glycolysis has been reported previously. NO binds to several targets and inhibits their functions within the mitochondrial respiratory chain (e.g., complexes I, III and IV) [28][29][30][31]. Inhibition of the respiratory chain by NO consequently decreases oxygen consumption and cellular respiration and results in an increase in the rate of aerobic glycolysis and an increase in the production of pyruvate and lactate [31,32]. Here, iNOS transfected cells had greater lactate production, significantly reduced coupling efficiency and spare respiratory capacity, most likely as a result of the inhibition of cellular respiration by NO. Expression of ARG2 reversed suppressive effects of iNOS/NO on glucose oxidative metabolism, which suggests that ARG2 arginine metabolism may protect against some of the adverse effects of NO on mitochondrial respiratory function.
Many studies support a link between metabolism and asthma. Prior studies have identified metabolic changes in the airway in the murine model of asthma [33]. Furthermore, mitochondria numbers and oxygen consumption in airway smooth muscle of asthmatics are greater than in healthy controls [34]. Platelets from asthmatic individuals have less reliance on glycolysis and greater tricarboxylic acid (TCA) cycle turnover [35]. Greater arginine flux through ARG2 in the mitochondria has recently been shown to drive TCA cycle and cellular respiration [6]. Metabolism of arginine regulates T cell function and fate, with higher levels of arginine increasing T cell survival and anti-tumor responses [20]. Thus, this work suggests that identification of the high NO sub-phenotype of asthma may enable strategies to target metabolic pathways for personalized asthma care.
Arginine bioavailability may impact nitric oxide production and potentially F E NO. We previously reported an increase in whole body metabolism of arginine in asthmatics as compared to healthy controls using stable isotope-labeled tracers [6]. Extracellular arginine levels do not reflect arginine metabolism or its cellular compartmentalization. Arginine is taken into cells by cationic amino acid transporter (CAT) proteins, and arginine and other urea cycle amino acids are highly compartmentalized in tissues and within intracellular pools, and do not equilibrate rapidly with extracellular pools. A recent study showed that intracellular arginine is critical in the control of glycolysis and mitochondrial activity [20], and intracellular arginine flux and mitochondrial arginine metabolism is a critical determinant of cell bioenergetics and function [6,20]. In the present study, we assessed airway epithelial enzyme levels and airway exhaled NO. However, we did not evaluate plasma arginine and arginine/ornithine ratios, or intracellular arginine.
Taken together, this study shows that the high F E NO phenotype represents a greater arginine metabolism endotype that is clinically characterized by greater airflow obstruction and airway reactivity. This new understanding is important to plan metabolic interventions and provide therapeutic options separate from corticosteroids and other anti-inflammatories.
Supporting information S1 Table. Aginine metabolic endotype of asthmatics based on asthma severity.