L-Arginine Pathway in COPD Patients with Acute Exacerbation: A New Potential Biomarker

Abstract

Background: Acute exacerbation of chronic obstructive pulmonary disease (AECOPD) remains a major cause of mortality. Clinical criteria of AECOPD are subjective. Biomarkers for AECOPD may aid in the initiation of early treatment. Increased production of asymmetric and symmetric dimethylarginine (ADMA, SDMA) is related to hypoxia. In COPD, a rise in ADMA results in a shift of L-arginine breakdown, contributing to airway obstruction. We aimed to compare serum levels of ADMA, SDMA and L-arginine in patients with and without AECOPD.

Methods: L-arginine metabolites quantified by high-performance liquid chromatography in venous blood samples and partial capillary oxygen pressure were prospectively investigated in 32 patients with COPD, 12 with AECOPD and 30 healthy subjects.

Results: Both ADMA and SDMA were significantly higher in AECOPD compared to stable COPD (p = 0.004 and p < 0.001, respectively). Oxygen content in capillaries correlated with serum ADMA concentration. However, the concentration of L-arginine was not different between AECOPD and stable COPD. Both ADMA and SDMA separated AECOPD with high sensitivity and specificity (AUC: 0.81, p = 0.001; AUC: 0.91, p < 0.001, respectively). A cut-off value ≥0.57 for SDMA was an independent variable to confirm AECOPD in a regression model (OR: 1.632, p = 0.001). All markers were significantly higher in the sera of both patient groups compared to the controls (p < 0.05, respectively).

Conclusions: COPD is associated with elevated L-arginine, ADMA and SDMA serum levels. In patients with AECOPD, production of ADMA and SDMA are more pronounced presumably due to more severe hypoxic insult. Methylated arginine derivatives in the sera may help early recognition of AECOPD.

Keywords:

• asymmetric and symmetric dimethylarginine
• chronic obstructive pulmonary disease
• L-arginine

Introduction

COPD is an obstructive airway disease associated with chronic airway inflammation (1). Recent studies indicated an association of COPD with an abnormal inflammatory response of the lung leading to hypoxia, in addition to a systemic inflammation which coexist with vascular remodeling (2). Acute exacerbations aggravate these pathological processes (3). This highlights the importance of potential biomarkers for early recognition of patients with acute exacerbation.

The relationship between hypoxia and the L-arginine-NO pathway is dual: hypoxia not only induces nitric oxide (NO) overproduction (4), but also facilitates protein arginine methylation (4). NO is produced by the conversion of L-arginine into L-citrulline by nitric-oxide synthase (NOS) (5). Three NOS isoforms have been identified: endothelial NOS (eNOS), inducible NOS (iNOS) and neuronal NOS (nNOS). Whereas eNOS and nNOS are mostly ­calcium/calmodulin-dependent, iNOS can be induced in response to hypoxia (6) or inflammation (7). NOS activity can be reduced by the accumulation of endogenous inhibitors such as asymmetric dimethylarginine (ADMA), which has recently been recognized as being associated with lung disease (8–10). Due to the effect of hypoxia on protein arginine methylation, ADMA is generated.

Although ADMA causes endothelial dysfunction by inhibiting eNOS, it also increases hypoxic tolerance (11). Arginine methylation of cellular proteins is catalyzed by protein arginine methyltransferases (PRMT). Recently, increased lung ADMA levels and ADMA/L-arginine ratios were found due to PRMT2 over-expression under chronic hypoxic conditions. This suggests that structural and functional changes caused by hypoxia may be linked to ADMA metabolism (12). Recent data suggest that the metabolism of methylarginine by the pulmonary system significantly contributes to circulating ADMA and SDMA levels (13). Circulating ADMA levels have been assessed in a variety of systemic cardiovascular diseases, and are increased in conditions associated with hypoxia, ischemic stroke, renal failure, pulmonary hypertension, heart failure, and hypercholesterolemia (14). Despite these lines of evidence pointing towards ADMA and SDMA as potential systemic biomarkers of different clinical stage of COPD, this has not been tested.

C-reactive protein (CRP) is increased in most forms of tissue damage, inflammation, and/or infection (15). CRP predicts death and cardiovascular complications in several diseases including inflammatory diseases (16), sepsis (17), pulmonary diseases (18), coronary heart diseases (19) and ischemic stroke (20). The interplay between ADMA and inflammation is a relevant issue because both factors, ADMA/SDMA and CRP, have been implicated in endothelial dysfunction in humans (21–23). Patients with COPD have elevated CRP levels compared to controls (24), and CRP was extensively evaluated as potential diagnostic and predictive biomarker in patients with AECOPD. However the wide range of variation in elevated CRP level was probably due either to different causes of AECOPD (bacterial vs. non-bacterial) or other influential factors such as inhaled corticosteroids (25).

Identifying novel biomarkers for recognition of patients with AECOPD seems to be highly valuable. To our knowledge, there is no data on comparing serum L-arginine, ADMA and SDMA levels between patients with stable COPD and AECOPD. We therefore aimed to compare the serum level of the precursor molecule L-arginine and its metabolites (ADMA and SDMA) as markers of vascular changes due to hypoxia and potential markers of acute inflammation/infection, and the acute-phase reactant hsCRP as a classic infection marker in patients with/without AECOPD.

Material and Methods

The study protocol was approved by the Regional Local Ethics Committee at University of Pecs, Faculty of Medicine (approval number: 3950, issued on 28th January, 2011) and informed consent was obtained from each patient. Anonymous blood donors served as age-matched healthy controls (n = 30).

Subjects

A total of 44 patients with COPD were recruited into this prospective study. Patients with COPD were enrolled from Department of Pulmonology at University of Pecs, Hungary. COPD was defined according to the Global initiative for chronic Obstructive Lung Disease (GOLD) guidelines (26). Exclusion criteria were chronic renal failure (eGFR < 50 and/or creatinine >120 μmol/l) and decline to participate in the study.

Severity of COPD was classified as follows: I = mild, FEV₁ ≥ 80% predicted (n = 3); II = moderate, 50% ≤ FEV₁ < 80% predicted (n = 23); III = severe, 30% ≤ FEV₁ < 50% predicted (n = 9); IV = very severe, FEV₁ < 30% predicted or FEV₁ < 50% predicted plus chronic respiratory failure (CRF) (n = 9). CRF (n = 16) was defined when PaO2 ≤ 55mmHg, or oxygen saturation ≤ 88% was measured twice over a three week period.

Exacerbations of COPD (AECOPD, n = 12) were defined as “an event in the natural course of the disease characterized by a change in the patient's baseline dyspnea, cough and/or sputum that is beyond normal day-to-day variations, acute onset, which may warrant a change in regular medication” (1). In this subgroup, three patients were classified into class II and III respectively, and six patients into class IV according to the above-mentioned criteria.

The medical history concerning previous coronary artery disease, myocardial infarction, diabetes mellitus, hypertension, dyslipidaemia, stroke, current cigarette smoking and orally taken medication was obtained from all patients. A history of hypertension was defined as the use of antihypertensive drugs or blood pressure more than 140/90 mmHg on at least two separate occasions. Diabetes mellitus was defined as the use of anti-diabetic drugs or a fasting plasma glucose value ≥7.0 mmol/L.

Blood sampling and determination of arginine derivatives and hsCRP

Partial oxygen pressure (capillary pO₂) was analyzed from capillary blood taken from the earlobe. Fasting venous blood samples were drawn and serum samples were frozen within 60 minutes and stored at -70°C until analysis. Serum high-sensitivity C-reactive protein (hsCRP) was measured by automated fluorescence immunoassay (BRAHMS Kryptor, Hennigsdorf, Germany). The amino-acid content of the blood serum samples was retrieved by solid-phase extraction (SPE) method and was quantified by high-performance liquid chromatography (HPLC) after derivatization. SPE of the analytes was performed as previously described (27, 28). Arginine and homoarginine were detected at λ_ex = 337 nm, λ_em = 520 nm, and λ_em = 454 nm was used for asymmetric dimethylarginine (ADMA) and symmetric dimethylarginine (SDMA).

Statistical analysis

Data were evaluated using SPSS software package (Version 19.0, SPSS Inc, Chicago, USA). Categorical data were summarized by means of absolute and relative frequencies (counts and percentages). Quantitative data were presented as mean and 95% confidence interval, as well as mean ± SD. The Kolmogorov–Smirnov test was applied to check for normal distribution. Due to normally distributed data, parametric methods (chi-square test for categorical data, Student-t test for continuous data) were used for demographic and clinical data. Non-parametric method (Mann–Whitney test) was used for hsCRP, L-arginine, ADMA and SDMA. Multiple logistic regression analysis was applied to confirm independent predictors. Cut-off value was calculated by a ROC analysis. Correlation analysis was performed calculating Spearman's correlation coefficient (r). A p-value < 0.05 was considered statistically significant.

Results

Demographic and clinical data are summarized in Table 1.

Table 1. Demographic and clinical data

	Healthy N = 30	AECOPD N = 12	Stable COPD N = 32	p-value
Age, years	51.1 ± 3.4	64.9 ± 9.2	59.3 ± 9.6	0.10
Male (%)	15 (50)	6 (50)	14(44)	0.71
Smoking (%)	6 (20)	4 (33)	8(25)	0.53
CAD (%)	NA	6 (50)	7(22)	0.06
Hypertension (%)	NA	10 (83)	18 (56)	0.17
Diabetes (%)	NA	3 (25)	7 (22)	0.51
Stroke (%)	NA	2 (17)	4 (13)	0.72
ICS therapy	NA	10 (83)	20 (63)	0.19
CRF (%)	NA	7 (58)	9 (28)	0.06
LTOT (%)	NA	5 (42)	8 (25)	0.12
Number of AE in previous year	NA	6 (50)	7 (22)	0.05
CBG pO2, mmHg	NA	52.2 ± 7.2	60.5 ± 11.5	0.02
SpO2%	NA	86 ± 5	89 ± 9	0.45
Creatinine, μmol/l	82.7 ± 13.5	81.2 ± 30.4	73.9 ± 27.6	0.53

Values are presented as mean ± SD or as numbers (relative percentages), respectively (chi-square test and one-way ANOVA test); CAD: coronary artery disease, COPD: chronic obstructive pulmonary disease, AECOPD: COPD with acute exacerbation, ICS: inhaled ­corticosteroid therapy, CRF: chronic respiratory failure (if SpO₂ < 90%) on admission, LTOT: long-term oxygen therapy, AE: acute exacerbation episode, CBG pO₂: partial pressure of oxygen in capillary blood gas, SpO₂: saturation of oxygen measured by pulse oximetry.

COPD vs. healthy subjects

Levels of L-arginine were significantly higher in recruited patients (n = 44) compared to healthy subjects (p < 0.05). Concentration of ADMA and SDMA were also significantly higher in patients compared to healthy controls (p < 0.001, respectively). Serum level of hsCRP was also significantly higher in the COPD group (p < 0.05) (see Table 2).

Table 2. Serum concentration of the L-arginine pathway metabolites and hsCRP in patients with COPD

	COPD N = 44	Healthy N = 30	p-value
L-arginine, μmol/l	109.5 (96.4–123.8)	84.1 (78.9–91.0)	<0.05
ADMA, μmol/l	0.67 (0.59–0.78)	0.41 (0.38–0.46)	<0.001
SDMA, μmol/l	0.54 (0.43–0.65)	0.36 (0.33–0.39)	<0.001
hsCRP, mg/l	7.8 (4.1–22.7)	3.4 (2.2–7.5)	<0.05
	AECOPD N = 12	Stable COPD N = 32
L-arginine, μmol/l	119.9 (102.7–144.9)	104.7 (91.5–117.9)	0.07
ADMA, μmol/l	0.83 (0.68–0.95)	0.63 (0.57–0.72)	<0.001
SDMA, μmol/l	0.68 (0.60–0.98)	0.46 (0.41–0.56)	<0.001
hsCRP, mg/l	11.9 (4.5–33.8)	7.5 (4.1–21.3)	0.53

COPD: chronic obstructive pulmonary disease, AECOPD: COPD with acute exacerbation, Stable COPD: COPD without acute exacerbation, ADMA: asymmetric dimethylarginine, SDMA: symmetric dimethylarginine, hsCRP: high-sensitivity C-reactive protein. Data are shown as median (25th–75th percentiles). Mann–Whitney U-test.

AECOPD vs. stable COPD

Both ADMA and SDMA were significantly higher (p < 0.001, respectively) in AECOPD compared to stable COPD patients (Figures 1 and 2); however, neither the levels of precursor molecule L-arginine ­(Figure 3) nor hsCRP showed significant difference (see Table 2). Based on a ROC analysis, both ADMA (cut-off ≥ 0.69 μmol/l, AUC: 0.81, p = 0.001) with a sensitivity of 75% and specificity of 73%, and particularly SDMA (cut-off ≥0.57 μmol/l, AUC: 0.91, p < 0.001) with a sensitivity of 92% and specificity of 86%, differentiated patients with AECOPD from stable COPD. SDMA was found as an independent variable in a multiple regression model including age, gender, co-morbidities, serum creatinine and L-arginine to verify AECOPD (OR: 1.632, 95% CI:0.75–2.5, p = 0.001). Using a ROC analysis, a cut-off value of ≥0.57 μmol/l for SDMA discriminated AECOPD with a sensitivity of 92% and specificity of 86% (AUC: 0.91, p < 0.001). Capillary pO₂ was significantly lower in patients with AECOPD compared to stable COPD (p = 0.01) (Table 1). We also found a significant negative correlation between partial capillary O₂ pressure reflecting severity of hypoxia and serum ADMA level (p < 0.05) (Figure 4).

Figure 1. Serum ADMA levels in COPD patients with and without acute exacerbation compared to healthy subjects. Serum concentration of ADMA (μmol/l) is shown in COPD patients with acute exacerbation (AECOPD), stable COPD and in healthy subjects (HS). Data are presented as mean and 95% confidence interval.

Figure 2. Serum SDMA levels in COPD patients with and without acute exacerbation compared to healthy subjects. Serum concentration of SDMA (μmol/l) is shown in COPD patients with acute exacerbation (AECOPD), stable COPD and in healthy subjects (HS). Data are presented as mean and 95% confidence interval.

Figure 3. Serum L-arginine levels in COPD patients with and without acute exacerbation compared to healthy subjects. Serum concentration of L-arginine (μmol/l) is shown in COPD patients with acute exacerbation (AECOPD), stable COPD and in healthy subjects (HS). Data are presented as mean and 95% confidence interval.

Figure 4. Correlation between partial capillary oxygen pressure and ADMA in patients with COPD. Correlation between partial capillary oxygen pressure (mmHg) and serum concentration of ADMA (μmol/l) is shown in COPD patients. (Pearson correlation, r = -0.34, p < 0.05, scatter plot diagram).

COPD patients with chronic respiratory failure (CRF)

Levels of L-arginine or its derivatives (ADMA, SDMA) in patients with CRF (n = 16) were not different from the levels in patients without CRF. Nevertheless, concentration of hsCRP was significantly higher in COPD patients with CRF (mean: 25.2, 95% CI: 5.9–44.4 vs. 6.5, 2.9–10.0, p = 0.008). CRF was not associated with AECOPD.

Discussion

To our knowledge this is the first study to demonstrate that methylarginine derivates (ADMA and SDMA) and their precursor L-arginine, are increased in the blood of patients with COPD compared to healthy subjects. We also demonstrated that the levels of SDMA and ADMA are further elevated in patients with AECOPD, and provide surrogate information in the diagnosis of COPD exacerbation with high sensitivity and specificity.

Our data obtained by analyzing systemic circulation extend observations of other studies demonstrating elevated ADMA levels in airway samples (i.e., exhaled breath condensate and induced sputum supernatant) (29,30). Although blood ADMA levels were also compared among patients with COPD, obstructive sleep apnea and overlap syndrome, healthy subjects were not included in the previous analyses (31).

Association between L-arginine and hypoxia

We found that serum level of the precursor molecule L-arginine was elevated in the peripheral blood in patients with COPD. In experimental model of airway hyper reactivity, the exogenous administration of L-arginine resulted in a protective effect, though it was different depending on airway level and pre-treatment duration (32). In addition, the total length of vessels within the gas exchange region of the hypoxic lungs was significantly increased after L-arginine supplementation (32). Taking that into consideration, an elevated serum L-arginine level might be an adaptive response to chronic hypoxia associated with local and systemic vascular remodeling. As COPD is accompanied by altered gas-exchange which causes sustained systemic hypoxemia because of an increased diffusion distance, poor ventilation, or loss of alveolar structures, the beneficial effect of L-arginine can also be important in prevention of systemic hypoxic insults (33,34).

In accordance with this, a recent study reported that endothelial cell dysfunction induced by hypoxia due to smoking was attenuated by the addition of L-arginine or by overexpression of DDAH-2 (35). However, the effect of L-arginine on the NO-pathway under chronic hypoxia remains controversial. Interestingly, the long-term oral L-arginine administration has failed to improve the hypoxia-induced pulmonary vascular remodeling and right ventricular hypertrophy (36), presumably because the intracellular arginine is not increased during L-arginine supplementation (5). Several studies have shown that L-arginine stimulates NO-overproduction in oxidative stress, inflammation and ischemia (37). On the contrary, a recent in vitro study reported a dose-dependent inhibition of DDAH activity by L-arginine, which competes with ADMA for binding to DDAH resulting in decreased NO production (38).

Link between L-arginine and ADMA in pulmonary diseases

In addition to L-arginine, ADMA was also significantly higher in the serum of patients with COPD in our cohort. L-arginine can be metabolized by NOS and arginase, and arginase is the first step in polyamine production necessary for cellular proliferation. ADMA inhibits NOS, but it does not inhibit arginase activity, which result in a greater L-arginine bioavailability to arginase and therefore increases viable pulmonary endothelial cell number (39). Although endothelial dysfunction and increased risk of various cardio- and cerebrovascular incidents are associated with high ADMA-concentrations, several studies point towards a potential beneficial effect of ADMA, mainly in the context of angioproliferative diseases such as cancer and pulmonary fibrosis (32).

It has also been reported that ADMA infusion results in significantly enhanced lung resistance and decreased dynamic compliance in response to methacholine in an animal model (40). These aphysiologic changes were associated with significantly increased lung ­collagen content. Furthermore, a significant decrease of nitrite in lung fluid was accompanied by elevated lung fluid urea and arginase activity in lung homogenates (40). These changes were reversed after completion of ADMA administration. In addition to this, arginase activity and collagen formation was stimulated by treatment with ADMA in mouse lung fibroblasts (40). These data support the hypothesis that ADMA may play a role in airway diseases through the inhibition of NOS and enhancement of arginase activity (40).

Superimposed hypoxic events in AECOPD and ADMA/SDMA production

We observed that levels of ADMA, and particularly SDMA, were higher in patients with AECOPD compared to stable COPD. This indirectly suggested that acute hypoxia may contribute to elevation of these molecules in the serum. Therefore, we correlated serum levels of ADMA and SDMA with oxygen content in capillaries. This indicated a direct correlation between the severity of an acute hypoxic episode in patients with COPD and their serum ADMA-concentration. We may hypothesize that such an elevation of ADMA in acute hypoxia may be beneficial. Animal data suggests that acute hypoxic conditioning increases hypoxic tolerance (12). Furthermore, inhibition of eNOS, but not iNOS, by endogenous ADMA also mediates enhancement of hypoxic tolerance (12). Several data support that ADMA exerts its inhibitory effect in a concentration-dependent manner on different isoforms of NOS (39,40). This protective effect of ADMA may be exerted via a dose-dependent inhibition of inflammation induced iNOS activity, preventing the deleterious consequences that such hypoxic insults provoke in the lung (41).

Our observation that ADMA and SDMA are elevated in the acute exacerbation of COPD may also reflect that this pathway is not only linked to hypoxia, but also to inflammation: proinflammatory pathways are activated in COPD and also in acute exacerbation (42). although ADMA may play a beneficial role by improving hypoxic tolerance, it also promotes inflammation by inducing oxidative and nitrosative stress in lung epithelial cells, which can be deleterious in COPD (12,43). ADMA increases the generation of the downstream proinflammatory mediators TNF-α and IL-8, and it activates the NF-κβ pathway and the binding of monocytes to endothelial cells (34). During the acute phase of infection, a substantial increase of SDMA was observed parallel to inflammatory/infection markers, while ADMA concentration was similar to healthy subjects (23).

Interestingly, the resolution of acute inflammation secondary to infectious process was associated with a clear-cut rise in ADMA levels coincident with a marked decline in biomarkers of inflammation and nitrosative stress, while SDMA levels remained unchanged throughout (23). Cytokines like IL-1 (44), IL-6 (45) and procalcitonin (46) all amplify the expression of iNOS, as well as NO-production. Pro-inflammatory cytokines in vitro have a dual effect on the ADMA-iNOS system (47). IL-1 stimulates iNOS and activates parallel DDAH resulting in decreased ADMA levels, with a net effect of NO overproduction (47). The effect of pro-inflammatory cytokines on SDMA levels has still not been clarified. In a recent study, SDMA serum levels were significantly elevated in patients with sepsis and closely correlated with biomarkers of inflammation (48).

Limitation of C-reactive protein in diagnostics of AECOPD

In our cohort, hsCRP did not distinguish patients with AECOPD from the stable COPD group. CRP is increased in response to a number of infectious and inflammatory conditions, but is not specific to either COPD or acute exacerbation of COPD (49). CRP was not able to discriminate infectious from noninfectious causes of AECOPD (49), but serves as a useful monitoring marker indicating resolution of the primary insult and inflammation. We have recently shown the time-dependent differential role of hsCRP in risk assessment related to ischemic stroke: the early increase in hsCRP level reflects atherosclerosis and therefore a risk for ischemic stroke, while its late elevation in the subacute phase of stroke was a risk factor for post-stroke infection, presumably reflecting the initial subclinical phase of infection (20). A similar situation may characterize COPD, where CRP is elevated due to its association with infection, and depends on the time of investigation. There is also a great need for longitudinal studies that could prove the role of hsCRP and L-arginine metabolites (ADMA, SDMA) as predictors of AECOPD, similar to stroke (50).

Importantly, COPD is a special clinical entity, where chronic hypoxia, chronic lung inflammation and systemic inflammation coexist, leading to vascular inflammation and acute exacerbations even aggravate these pathological processes (2,3). Therefore, our findings suggest that ADMA and particularly SDMA are not only determinants of endothelial dysfunction, but also biomarkers of AECOPD. Longitudinal studies are required to determine the role of methylated arginine derivatives as prospective biomarkers that may aid in prediction of acute exacerbation and its outcome in COPD.

Abbreviations
AECOPD	=	COPD with acute exacerbation
ADMA	=	asymmetric dimethylarginine
AUC	=	area under the curve
COPD	=	chronic obstructive pulmonary disease
CRP	=	C-reactive protein
DDAH	=	dimethylarginine dimethylaminohydrolase
HPLC	=	high-performance liquid chromatography
IL	=	interleukin
PCT	=	procalcitonin
ROC	=	receiver operating characteristic
NO	=	nitric oxide
NOS	=	nitric oxide synthase
PRMT	=	protein arginine methyltransferase
SDMA	=	symmetric dimethylarginine

Declaration of Interest Statement

Author contributions: TM had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. LN and SK contributed substantially to study design, measurement of the biomarkers and interpretation of the results. IR, VS, BI contributed substantially to recruitment of patients, blood sampling and study design. TM, LB, IR, ZI and IH contributed substantially to data analysis and interpretation, and the writing of the manuscript. TM takes responsibility for the content of the manuscript, including the data and analysis.

Acknowledgments

The authors thank Elise Steen for copyediting the revised manuscript.

Funding

This study was supported by a grant from University of Pecs AOK-KA- 34039/10-17 to T. Molnar.