In the current prospective cohort study, we described a novel application of metabolomics in identifying the serum metabolic signatures and assessing the severity of AR. The OPLS-DA models showed that there were obvious discriminators between patients with different severity of AR and health controls. 35 and 59 metabolites responsible for differentiating MAR and MSAR patients from health controls, respectively, were identified. And 30 metabolites were found to be responsible for discriminating MSAR patients from MAR patients. After analyzing the relationships between the major discriminative metabolites with clinical parameters of AR patients, we observed that sarcosine, sphingosine-1-phosphate, cytidine and linoleic acid levels were associated with the severity of AR. These results showed that the identified potential serum biomarkers might be useful for diagnosing AR and developing objective indication for evaluating the severity of AR. We will next discuss the most significant metabolites and related metabolic pathways which may help us to better understand the underlying pathogenesis of AR and monitor the disease severity.
Most importantly, arginine and proline metabolism pathway was significantly perturbed among the most affected pathways in AR patients. Arginase and proline metabolism is of particular importance in the nitric oxide synthesis and integrally links to cellular respiration, metabolism, and inflammation [22-23]. A recent publication detected arginase and proline metabolism significant perturbations in the serum of commuters following traffic pollution exposures, and the researchers considered that arginase and proline metabolism dysfunction strongly associated with oxidative stress and inflammation in the air pollution toxicity [24]. Yang et al [25] found that the levels of arginine and its downstream products, such as ornithine, citrulline, creatine, creatinine, hydroxyproline and sarcosine were higher in the serum of asthma patients than in health controls, and they held that arginine and proline metabolism was the most important pathway in the development of asthma. Consistent with the previous report, we also observed that the levels of sarcosine and creatinine were higher in the serum of AR patients than in the health controls, and the levels of sarcosine correlated positively with TNSS and VAS. Arginine is an essential amino acid related to endothelial function, inflammation, and airway hyperresponsiveness, and higher levels of arginine and its downstream products can regulate T cell function and promote its activity, and act a critical role in several inflammatory diseases, including asthma and AR [22-23, 26]. Therefore, we speculated that arginine and proline metabolism might be involved in the development of AR and sarcosine could roughly be related to the severity of AR.
Our results also provide evidence for the sphingolipids metabolism alteration is involved in occurrence and progression of AR. Sphingolipids are ubiquitous components of the cell membrane and play an important role in cell growth, inflammation and tissue remodeling [27-28]. Among the numerous sphingolipids, S1P has received the greatest attention in allergic diseases and autoimmune diseases, as it has been implicated in the modulation of a variety of cell responses such as immune cell proliferation, differentiation and regulation [27, 29]. A previous study reported that S1P up-regulated the cytokine production, such as IL-12, IL-23 and IL-27, in activated murine bone marrow derived dendritic cells, and it might serve as a novel therapeutic target in the treatment of several inflammatory diseases [30]. In another study, researchers found that the plasma levels of SIP were elevated in cystic fibrosis patients, and S1P levels correlated with routine laboratory parameters, lung function and clinical symptoms [31]. Kowal et al [29] analyzed targeted metabolites in the serum from 22 allergic asthma patients and 11 allergic rhinitis patients, and found that the sphingolipids metabolism was altered and the biosynthesis of S1P was augmented. In the present study, we observed that sphingolipids metabolism was disturbed and the S1P levels elevated in the AR patients, and the S1P levels were correlated positively with the disease severity of AR, which was in accordance with the results in the previous publications [29-30]. Our results support the hypothesis that alterations in the serum metabolites reflect the chronic activation of immune system in AR patients and that the disease severity is consistent with greater activation of the immune system. However, the mechanism underlying these manipulations has not been well clarified.
We firstly found that cytidine, identified form UHPLC-MS analysis, was associated with AR and it might be a novel marker and potential therapeutic target for AR. In our study, we observed that the levels of cytidine were elevated in the MAR and MSAR group, and the concentrations of cytidine were positively correlated with TNSS and VAS. Cytidine, a pyrimidine molecule, is considered as the precursor of the cytidine triphosphate (CTP), which is vital in the synthesis, interconversion and degradation of DNA, RNA and lipids [32-33]. Previous studies have found that abnormality of pyrimidine metabolism could influence the cell growth, development and differentiation, such as T cell and B cell [34]. A recent study demonstrated that interference of pyrimidine metabolism affected murine lymphocyte proliferation in vitro and attenuated the severity of experimental autoimmune arthritis [35]. Another study observed that the concentrations of 5,6-dihydorthymine were higher in the serum of current asthma patients compared with health controls, and the researchers believed that the alteration of pyrimidine metabolism might have relevance for asthma pathophysiology [36]. These events suggested that pyrimidine metabolism may play a role in the autoimmune diseases and allergic diseases. Therefore, we ultimately believed that cytidine was associated with AR, and that it might serve as a promising metablic biomarker for assessing the disease severity and evaluation of treatment.
Interestingly, we also found that fatty acids metabolism was dysregulated in all OPLS-DA models. In recent years, growing evidence suggests that fatty acids metabolism plays important roles in the modulation of immune responses in health and disease [37]. Most researchers hold that unsaturation fatty acids, especially polyunsaturated fatty acids exhibited potential protective effects on allergic inflammation, while saturation fatty acids promoted the inflammatory response [38]. In a recent animal experimental study, Lee et al [39] observed that oleic acid had anti-asthmatic effects such as the down-regulation of inflammatory cells and eosinophil in bronchial alveolar lavage fluid, IgE in serum. Several in vitro studies also demonstrated that unsaturation fatty acids could exert immunosuppressive effects on T cells, such as reducing its proliferation and activation in a dose-dependent manner [37, 40]. However, saturation fatty acids, such as palmitic acid, have been described as essential factors promoting T cell activation and cytokine secretion [41]. In addition, considerable evidence has shown that polyunsaturated fatty acids can modify mast cell functions and suppress its activation then reduce the production of cytokine or chemokine through receptors [42]. Therefore, we suppose that fatty acids metabolism may act an important role in the development of AR. In the current study, the concentrations of several unsaturated fatty acids (linoleic acid, arachidic acid and trans-vaccenic acid) were lower in the serum of MAR or MSAR patients in comparison with health controls, while the concentrations of palmitic acid were elevated. Moreover, the levels of linoleic acid were correlated negatively with TNSS and VAS. Our results were in line with most previous studies. However, further studies should be conducted to confirm these results and to clarify the underlying mechanism of AR subtypes.
There are also several limitations in our study which may affect the reported findings. First, the total sample sizes were relatively small, and a validation cohort study was needed to confirm the conclusions. Second, all the recruited participants were from single-center with the same ethnicity and region, which might limit the applicability of our findings. Third, only one biological sample (serum) was used in the present study, future studies should collect other biological samples, such as urine and nasal lavage fluid, to further verify whether the identified differential metabolites were associated with AR. Last, we did not compare serum metabolites between MAR and MSAR patients, but it does not mean that there are no differential metabolites. Future studies with larger sample sizes using targeted approaches will be important to support and extend our present findings.