A Novel Polymorphism in Secretory Phospholipase A^sub 2^-IID Is Associated with Body Weight Loss in Chronic Obstructive Pulmonary Disease

Chronic obstructive pulmonary disease (COPD) is a major cause of morbidity and mortality worldwide. In a recent review of the global burden of human illness, COPD ranked 12th as a cause of lost quantity and quality of life and was projected to rank fifth by 2020 (1, 2). This life-threatening disease is characterized by significant chronic inflammation not only in the pulmonary compartment but also in the systemic circulation (3-7). In particular, the presence of a chronic and systemic inflammatory response has an important influence on patient survival, because unexplained weight loss (muscle wasting and adipose tissue depletion), the characteristic feature in advanced COPD, has been reported to be linked to the systemic inflammation (3-11). This unexplained weight loss is clinically relevant because it limits patients' physical performance, jeopardizes their quality of life, and is related to their poor prognosis, independent of the severity of airflow obstruction (12, 13). The main cause for the presence of chronic and systemic inflammation in patients with COPD still remains to be elucidated.

Recent studies have demonstrated the relationship between metabolic derangement and elevated levels of proinflammatory mediators, such as tumor necrosis factor a (TNF-a), a pleiotropic cytokine, in the systemic circulation of patients with COPD (3-7). TNF-a (cachexin) is associated with accelerated metabolism and protein turnover as well as with chronic wasting diseases in other cachexie patients, resulting in the loss of skeletal muscle protein and adipose tissue (3-7, 14). However, because chronic smoking is necessary but not sufficient to cause COPD, not all patients with COPD exhibit weight loss during the course of their disease (15). In fact, unexplained weight loss is particularly prevalent in patients with severe COPD, occurring in approximately 50% of these patients, but this condition can also be seen in approximately 10 to 15% of patients with mild to moderate disease (15). Similarly, it is notable that the systemic inflammatory profile, potentially related to the underlying mechanism of weight loss, varies in each individual patient with COPD, and that a wide distribution of the levels of the proinflammatory mediators is reported to be observed in patients with COPD (3-11). Although this may be related to severity (15) or phenotype of disease (16), a genetic component similar to that suggested as an explanation for the development of COPD in only a proportion of smokers (17, 18) cannot be excluded. The genes and polymorphisms that may predispose to this weight loss process still remain unknown.

Phospholipases A^sub 2^ (PLA^sub 2^s) are enzymes responsible for mobilization of fatty acids, including arachidonic acid, from phospholipids (19). PLA^sub 2^ enzymes are classified as high-molecular-weight cytosolic PLA^sub 2^s (cPLA^sub 2^s) and low-molecular-weight secretory PLA^sub 2^s (sPLA^sub 2^s) (19). There is increasing evidence that large quantities of sPLA^sub 2^s are released in the plasma of patients with systemic inflammatory diseases, such as septic shock and extensive burns (20, 21), and autoimmune disorders, such as rheumatoid arthritis and inflammatory bowel diseases (22, 23). Initially, these extracellular enzymes were considered to play an important role in inflammation by releasing arachidonic acid, which is subsequently converted to proinflammatory prostaglandins and leukotrienes. However, recent studies have suggested that the proinflammatory effects of sPLA^sub 2^ are not limited to the initiation of arachidonic acid metabolism. Rather, several diseases associated with high levels of extracellular sPLA^sub 2^ are characterized by a significant increase in plasma or tissue concentrations of proinflammatory cytokines, such as TNF-a and interleukin 1ß (24). In addition, sPLA^sub 2^ proteins are reported to induce degranulation and production of proinflammatory cytokines from a variety of cells involved in inflammatory and immune responses in an autocrine manner (19, 24). These effects are exerted by mechanisms that are independent of the enzymatic activity and are mediated by the interaction of sPLA^sub 2^s with specific or promiscuous membrane receptors (24). It is therefore considered that proinflammatory cytokines and sPLA^sub 2^s potentiate each other's synthesis, thereby creating an amplification loop for propagation of inflammatory responses.

To date, 10 sPLA^sub 2^ isoforms (IB, IIA, IIC, IID, IIE, IIF, III, V, X, and XII) have been identified in mammals (25). These isoforms have a highly conserved catalytic site, Ca^sup 2+^-binding loop, and common molecular mass of 14 to 19 kD except for group III isoforms (25). Among these sPLA^sub 2^ isoforms, sPLA^sub 2^-IIA, sPLA^sub 2^-IIC, sPLA^sub 2^-IID, sPLA^sub 2^-IIE, sPLA^sub 2^-IIF, and sPLA^sub 2^-V are clustered on the same chromosome locus (1p34-p36); these are often referred to as the group II subfamily sPLA^sub 2^s (25). A biological feature of the group II subfamily sPLA^sub 2^s is that almost all isoforms are associated with an inflammatory and immune process except for sPLA^sub 2^-IIC (pseudogene in humans) (25).

Taking these backgrounds into account, we exclusively focused on the proinflammatory properties of sPLA^sub 2^s, with the chronic and systemic inflammation as a potential cause of body weight loss in patients with COPD. In this study, we hypothesized that the individual susceptibility to bodyweight loss in patients with COPD may be attributed to the genetic variance of the sPLA^sub 2^ genes, especially those of the group II subfamily sPLA^sub 2^s. We found 12 single nucleotide polymorphisms (SNPs) around these genes by searching the nucleic acid database. Thus, the purpose of this study was to identify SNPs of the group II subfamily sPLA^sub 2^s that may render susceptibility to body weight loss in patients with COPD, and thereby to contribute in predicting its occurrence during the course of the disease.

METHODS

Study Population

Of the Japanese patients with COPD who consulted the outpatient clinic of the University of Yamagata between 2002 and 2003, we studied 276 male patients whose genomic DNA was available, after obtaining written, informed consent for genotyping. These patients with COPD were diagnosed according to the criteria established by the American Thoracic Society (26). Their irreversible chronic airflow obstruction was confirmed by spirogram. The patients had been clinically stable for at least 3 mo and lacked clinical signs of exacerbation. Patients who were associated with conditions known to affect body weight, such as chronic heart failure, autoimmune disorders, and malignant diseases, were strictly excluded. Our patients with COPD included 19 patients with diabetes mellitus and 33 patients with gastrointestinal tract disease, such as gastric ulcer or gastritis, all of whom were receiving appropriate therapy for their concomitant diseases. None of the patients was receiving nutritional support therapy. After an overnight fast, all subjects had anthropometric measurements. The study protocol was approved by the local ethics committee of the Yamagata University School of Medicine, and the written, informed consent was obtained from all subjects before participating in this study.

Pulmonary Function Test

FVC and FEV^sub 1^ were measured with standard spirometric techniques (CHESTAC-25 part II EX; Chest Corp., Tokyo, Japan). The highest value from at least three spirometric maneuvers was used. Reference values were those proposed by Japanese Society of Chest Diseases (27). Arterial blood gas was analyzed with the subject breathing with or without supplemental oxygen in the sitting position (280 Blood Gas System; Ciba Corning Diagnostics Corp., Medfield, MA).

Candidate SNP Selection and Polymorphism Cenotyping

The genes and SNPs of the group II subfamily sPLA^sub 2^s used for the test are shown in Table 1. Their chromosomal locations are illustrated in Figure 1. SNPs of the group II subfamily sPLA^sub 2^s were extracted from the dbSNP (http://www.ncbi.nlm.nih.gov/SNP/), JSNP (http://snp.ims. u-tokyo.ac.jp), and Applied Biosystems genotyping databases (http:// myscience.appliedbiosystems.com/genotype/). Altogether, 12 SNPs were derived from five genes, sPLA^sub 2^-IIE, sPLA^sub 2^-IIA, sPLA^sub 2^-V, sPLA^sub 2^-IID, and sPLA^sub 2^-IIF, and all were tested for the association with the degree of body weight loss in COPD. The genotyping was conducted with a fluorogenic polymerase chain reaction. The alleles and the genotype frequencies of the tested SNPs were determined and combined with the clinical data to conduct statistical analysis.

Typing Method

Analysis of genetic polymorphisms SNP genotyping was performed by TaqMan allelic discrimination assay (28). Reagents were purchased from Applied Biosystems (Foster City, CA). TaqMan probes were designed and synthesized by Applied Biosystems, and distinguish the SNPs at the end of a polymerase chain reaction. One allelic probe was labeled with the fluorescent FAM dye and the other with the fluorescent VIC dye. Polymerase chain reaction was performed by TaqMan Universal Master Mix without UNG (Applied Biosystems) with polymerase chain reaction primers at concentrations of 900 nM and TaqMan MGB probes at concentrations of 200 nM. Reactions were performed in 384-well formats in a total reaction volume of 3 µl using 3.0 ng of genomic DNA. The plates were then placed in a GeneAmp PCR System 9700 (Applied Biosystems) and heated at 95°C for 10 min, followed by 40 cycles at 92°C for 15 s and at 60°C for 1 min, with a final soak at 25°C. The plates were read by the Prism 7900HT instrument (Applied Biosystems) where the fluorescence intensity in each well of the plate was read. Fluorescence data files from each plate were analyzed by the SDS 2.0 allele calling software (Applied Biosystems). Several data (signal intensity) were eliminated to preserve the reliability of the assay system (missing data are guaranteed to be less than 5%).

Statistical Analysis

The Hardy-Weinberg equilibrium of alleles at the individual loci was evaluated using a ?^sup 2^ test (with p > 0.05). The association between the genotypes and the degree of body weight loss in COPD (the decrease of BMI in patients with COPD) as well as other clinical parameters were evaluated using the Fisher's exact test, logistic regression analysis, and analysis of covariance (ANCOVA) (29). We applied both dominant and recessive genetic models in these statistical analyses. Logistic regression analysis and ANCOVA were adjusted for age, smoking index, and lung functional index (FEV^sub 1^ % predicted value), because age, lung functional index, and smoking index can be possible confounding variables in our patients with COPD and we need to compensate the effects of these confounding variables in multivariate analysis. In particular, ANCOVA can be applied in the case when the dependent variable is a numeric variable (BMI in this study) and the independent variables are categoric and nominal (SNP genotypes in this study), with some confounding variables existing. However, if there are some interactions between independent variables (SNP genotypes) and confounding variables, the ANCOVA test cannot be applied. In this study, we confirmed that there were no significant interactions between SNP genotypes and confounding variables (interactive p value > 0.05). To compare the clinical values of the three different genotypic groups, a single-factor analysis of variance test was used with post hoc correction (Fisher's protected test). Results were expressed as mean ± SD. A p value less than 0.05 was considered significant. All data analyses were performed using SPSS version 12.0.1J (SPSS, Inc., Chicago, IL).

RESULTS

Fisher's Exact Test and Logistic Regression Analysis

To test our hypothesis, we first chose our patients with COPD (n = 200) whose BMI was at the bottom 100 (phenotype A group: BMI, 17.13 ± 1.29 kg/m^sup 2^ [range, 12.49-18.98]; age; 73.8 ± 7.1 yr; smoking index, 1,117 ± 573; FEV^sub 1^, 40.18 ± 20.37 % predicted; n = 100), and whose BMI was at the top 100 (phenotype a group: BMI, 23.83 ± 1.98 kg/m^sup 2^ [range, 21.63 - 32.74]; age, 73.4 ± 6.7 yr; smoking index, 1,089 ± 551; FEV^sub 1^, 53.62 ± 23.80 % predicted; n = 100). Our analyses of Fisher's exact test (odds ratio, 2.36; 95% confidence interval, 1.34-4.18; p = 0.004) and logistic regression (odds ratio, 2.10; 95% confidence interval, 1.13-3.90; p = 0.019) showed that SNP10 (National Center for Biotechnology Information SNP reference: rs584367) in the sPLA^sub 2^-IID gene was significantly associated with the degree of body weight loss in patients with COPD in the recessive model (Table 2). The analyses also showed that mutant allele (thymine [T]) of SNP10 in the sPLA^sub 2^-IID gene increased the risk of body weight loss in COPD (phenotype A group) more than twice than that of the nonmutant allele (cytosine [C], phenotype a group; p < 0.05). In addition, this nucleic acid change (from C to T) resulted in the amino acid alteration (from glycine [Gly] to serine [Ser]) in the sPLA^sub 2^-IID protein (Gly80Ser; Table 1). Logistic regression analysis also indicated that the lung functional index, evaluated by FEV^sub 1^ % predicted, was the significant risk factor in phenotype A group, in which a 1% decrease of FEV^sub 1^ % predicted led to a 1.03 times increase of the risk of the degree of body weight loss in COPD (phenotype A group; p < 0.001, data not shown). The Hardy-Weinberg equilibrium of SNP10 was 0.497 and the p value (?^sup 2^ test) was 0.481. The allele frequency for SNP10-C and SNP10-T was 0.6825 and 0.3175, respectively. None of the other SNPs gave positive results except for SNP11 in the sPLA^sub 2^-IID gene, which is juxtaposed to SNP10 and is in strong linkage disequilibrium.

ANCOVA and Genotypic Characterization of SNP10

To confirm the results of Fisher's exact test and logistic regression analysis, we next analyzed all the patients with COPD (n = 276). The ANCOVA test also indicated the significant difference between the values of BMI (log transformed) of the mutant group (CT + TT) and that of the nonmutant group (CC) in the recessive model of SNP10 (mean [SE], 1.293 [0.005] vs. 1.317 [0.006]; p = 0.003; Table 3). The whole patients enrolled were stratified into the four classes of BMI usually used (Figure 2a) (12), and were also stratified in the six classes of the adjusted BMI by ANCOVA (Figure 2b). Carriage of the minor allele (T) significantly differentiated the distribution of BMI (Fisher's exact test: p = 0.00329) and the adjusted BMI by ANCOVA (Fisher's exact test: p = 2.33 × 10^sup -7^) of the mutant group (CT + TT) from that of the nonmutant group (CC). This association between the phenotype (i.e., body weight loss in COPD) and the genetic variant resulted in an excess of genotypes carrying the minor allele (T) in the lower class of BMI (Figures 2a and 2b). The BMI average value of the mutant "Ser" group (Gly/Ser + Ser/Ser: n = 148) was 1.2 (kg/m^sup 2^) less than that of the other nonmutant group (Gly/ Gly: n = 128; p = 0.0017; Figure 3a). The significant difference in the adjusted BMI by ANCOVA between the two groups (Gly/Ser + Ser/Ser: n = 143; Gly/Gly: n = 118) showed more pronounced p value (p < 0.0001; Figure 3b). The other diagnostic parameters for the severity of COPD according to the three genotypes of SNP10 in the sPLA^sub 2^-IID gene (CC: Gly/Gly; CT: Gly/Ser; and TT: Ser/Ser) are summarized in Table 4. There were no significant differences in the levels of these parameters including lung function among the three groups. The average value of BMI as well as adjusted BMI by ANCOVA of the heterogeneous combination "Gly/Ser" were lower than those of homogeneous combination "Ser/Ser." These results may be caused by the small number of homogeneous samples that have relatively large deviations in our analyses. These results are also illustrated in Figures 4a and 4b, respectively.

DISCUSSION

The aim of this study was to clarify whether SNPs exclusively of the group II subfamily sPLA^sub 2^s are associated with the individual susceptibility to body weight loss in patients with COPD, which is the clinical hallmark of poor prognosis in these patients (12, 13). Body weight loss in patients with COPD has been considered to be attributed, at least in part, to the presence of the chronic and systemic inflammation observed in these patients (3-7). On the other hand, biophysiologic properties of the group II subfamily sPLA^sub 2^s are involved in the progression of the proinflammatory effects and the immune processes (19, 24). Because SNP10 is the mutant site that causes the amino acid change in the sPLA^sub 2^-IID gene (Table 1), and the Fisher's exact test, the logistic regression analysis, and the ANCOVA showed the statistical significance between SNP10 and the reduction of BMI in patients with COPD in our analyses, it is conceivable that SNP10 may be the susceptible SNP associated with the decrease of BMI in patients with COPD. It is likely that sPLA^sub 2^-IID is one of the susceptibility genes that contributes to body weight loss in patients with COPD, because the adjacent SNP11 also showed a low p value in the recessive model in our study, although it did not reach statistical significance (Tables 2 and 3).

The sPLA^sub 2^-IID protein consists of 125 amino acids (M^sub r^ = 14,500) preceded by a 20-residue prepeptide and is most similar to sPLA^sub 2^-IIA with respect to the number and positions of cysteine residues with an overall identity of 48% (30). sPLA^sub 2^-IID is constitutively expressed in the immune and digestive organs in humans and is upregulated by systemic proinflammatory stimuli in some restricted tissues, such as lung, thymus, and spleen of mice, suggesting its functional role in the progression of the inflammatory process (30, 31). In fact, searching the nucleic acid database reveals the presence of the TATA box and the binding motifs for AP-1 and nuclear factor-?B in the putative 5'-flanking promoter region of the human sPLA^sub 2^-IID gene, consistent with its proinflammatory, signal-associated inducible nature. The SNP type of SNP10 (National Center for Biotechnology Information SNP reference: rs584367) is a missense mutation, and causes amino acid change in sPLA^sub 2^-IID protein (Gly80Ser). Although the location of this amino acid polymorphism is not in the catalytic domain nor in the Ca^sup 2+^-binding loop, which are highly conserved in all the group II subfamily sPLA^sub 2^s, it is next to cysteine residue, which is completely preserved among the group II subfamily sPLA^sub 2^s (30, 32). It is tempting to speculate that a single amino acid change leads to the structural alteration of the sPLA^sub 2^-IID protein, and consequently influences the functional properties of this molecule in some degree.

COPD is a disease that develops and progresses for many years (33, 34). This study made only a one-time measurement of body weight as a marker of body weight loss, because it has been demonstrated that BMI at one-time measurement has a similar predictability for the mortality in patients with COPD, compared with the changes in body weight (12). Indeed, the relationship between the genotypes and body weight loss, and not low body weight, is clearly demonstrated in Figure 2.

The fact that there was no statistical difference in BMI between the two homogeneous groups (Table 4: Gly/Gly vs. Ser/Ser) may raise a question as to the relevance between this polymorphism and clinical findings. This result may be due to the fact that the body weight loss phenotype in COPD is influenced by many other confounding factors that cannot be ignored. In this study, age, smoking index, and lung function index (FEV^sub 1^ % predicted) showed significant interactions with BMI as expected. When these confounding factors were compensated by ANCOVA (29), the adjusted BMI can still differentiate the mutant group (CT + TT) from the nonmutant group (CC), which was even clearer than without adjustment (Figures 2b and 2a). Figure 2b implies that patients with COPD consist of two distinct groups that are normally distributed with respect to body weight loss. The same holds true for Table 4 (simple BMI vs. BMI by ANCOVA) and Figures 4a (simple BMI) versus Figure 4b (adjusted BMI by ANCOVA), where the adjusted BMI by ANCOVA, not simple BMI, demonstrated the significant difference between the two homogeneous groups (Gly/Gly vs. Ser/ Ser) and a typical "dominant-negative" genetic model.

Although the statistics show significant relationship between SNP10 genotypes of sPLA^sub 2^-IID and BMI (Table 4), small differences in BMI among the three groups may also raise a question with regard to its clinical relevance. However, because of the reasons mentioned above, we have to compare the mean values of the adjusted BMI by ANCOVA instead of those of the simple BMI among the three groups. For example, the differences between the mean values of the Gly/Gly and Gly/Ser groups, and the Gly/Gly and Ser/Ser groups, are 1.22 and 0.92 kg/m^sup 2^, respectively. Because the mean value of height among the three genotypic groups was not different (1.59 m, the average values of the whole patients with COPD studied), the body weight reduction can be calculated as 3.08 and 2.33 kg for the Gly/Ser group and for the Ser/Ser group, compared with the Gly/Gly group, respectively. The mean values of body weight of the Gly/Ser group and the Ser/Ser group are 50.6 and 50.8 kg, respectively. Therefore, as a single gene involvement, we do think that there is clinical relevance of the small difference in BMI (adjusted BMI by ANCOVA) associated with the sPLA^sub 2^-IID gene polymorphism.

The pathophysiologic relevance between body weight loss and the amino acid polymorphism (SNP10) in the sPLA^sub 2^-IID protein in patients with COPD demonstrated in this study is difficult to explain, but several speculations warrant further discussion. First, the ability to degranulate or synthesize the proinflammatory cytokines with specific or promiscuous membrane receptors of the inflammatory cells may be modified or even enhanced by this amino acid change (from Gly to Ser) in sPLA^sub 2^-IID protein. The persistent and systemic hyperinflammatory status leads to the body weight loss in patients with COPD whose genotype is the heterogeneous combination "Gly/Ser" or the homogeneous combination "Ser/Ser" (Table 4 and Figure 4). second, one possible function of sPLA^sub 2^-IID protein is the digestion of phospholipids in nutrition (30). The functional alteration of this property of the sPLA^sub 2^-IID protein due to the amino acid polymorphism (SNP10) may influence the efficacy of digestion of phospholipids in patients with COPD; that is, the patients with COPD whose genotype is the heterogeneous combination "Gly/Ser" or the homogeneous combination "Ser/Ser" may exhibit less efficacy of digestion of phospholipids and are consequently afflicted with body weight loss despite nutritional support therapy (Table 4 and Figure 4).

In summary, we exclusively focused on chronic and systemic inflammation, and specifically on the proinflammatory properties of the group II subfamily sPLA^sub 2^s, as a potential cause of body weight loss in patients with COPD who had 12 SNPs extracted from the nucleic acid database. We tested our hypothesis that the individual susceptibility to body weight loss in patients with COPD is attributed to the genetic variance of the group II subfamily sPLA^sub 2^s that stems from these SNPs, by using the genomic DNA from 276 male patients with COPD. Because the Fisher's exact test, logistic regression analysis, and ANCOVA showed the statistical significance between SNP10 (National Center for Biotechnology Information SNP reference: rs584367) and the reduction of BMI in patients with COPD, we conclude that the SNP10 may be the susceptible SNP associated with the decrease of BMI in patients with COPD. We can also suggest that sPLA^sub 2^-IID may be one of the susceptibility genes that contribute to body weight loss in patients with COPD, because SNP10 is the mutant site that causes the amino acid change in the SPLA^sub 2^-IID gene. The functional analysis with respect to the amino acid polymorphism (SNP10) of the SPLA^sub 2^-IID protein will be necessary, and its related mechanism that can explain the individual susceptibility to body weight loss in patients with COPD must be clarified. Nevertheless, we believe this study provides important information for the better understanding of the unexplained weight loss in patients with COPD, which may ultimately lead to creating novel therapeutic strategies in the future.

Conflict of Interest Statement: None of the authors have a financial relationship with a commercial entity that has an interest in the subject of this manuscript.

Acknowledgment: The authors thank Masayuki Saito, Ph.D. (HuBit Genomix, Inc., Tokyo, Japan), for his helpful discussion.

In addition to the authors, the following investigators and Japanese institutions participated in the study: Shinoda General Hospital, Yamagata: H. Atsumi; Okitama General Hospital, Kawanishi: M. Inage and S. Kato; Japan Sea Hospital, Sakata: H. Saito; Saisei Hospital, Yamagata: H. Takeda; Shinjo Prefectural Hospital, Shinjo: Y. Katagiri and K. Otake; Yonezawa City Hospital, Yonezawa: H. Yuki; Sagae City Hospital, Sagae: M. Sato; Saiseikan Hospital, Yamagata: J. Sato and K. Iwabuchi; Kahoku Prefectural Hospital, Kahoku: H. Suzuki; Tohoku Central Hospital, Yamagata: T. Sayama and K. Shida; Kaminoyama Nagaoka Clinic, Kaminoyama: M. Nagaoka; Yakuwa Clinic, Murayama: N. Yakuwa; Sasai Clinic, Yonezawa: Y. Sasai.