INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Chronic obstructive pulmonary disease (COPD) is characterized by expiratory airflow obstruction and hyperinflation due to inflammation of peripheral airways and loss of lung elastic recoil. Tobacco smoking is the major risk factor for COPD, however, only approximately 15% of smokers develop clinically relevant airflow obstruction (1). This variation in response to cigarette smoke in combination with familial aggregation of chronic obstructive respiratory disease (2) suggests that there is a genetic component to the development of COPD.

Numerous candidate genes could be risk factors for COPD (8, 9). Because airway obstruction is due to both loss of lung elastic recoil and inflammatory narrowing of peripheral airways, genetic polymorphisms that affect either process could be involved. One candidate gene that has been implicated in COPD is Gc-globulin, also known as vitamin D binding protein. The Caucasian population expresses three major isotypes of this serum protein named 1S, 1F, and 2 with allele frequencies of 0.56, 0.16, and 0.28, respectively (10). The differences among isoforms are the result of two separate point mutations, each of which results in a single amino acid substitution.

Whether different isoforms of Gc-globulin alter the risk of developing COPD remains controversial. Kueppers and coworkers (11) found that the homozygous Gc2 genotype was protective against COPD (relative risk [RR] = 0.2). Although Kauffmann and colleagues (12) were unable to replicate this result, a more recent study by Horne and coworkers (13) found that the presence of one or more Gc2 alleles decreased the relative risk of developing COPD (RR = 0.7) whereas homozygous 1F-1F individuals had a significantly increased risk of developing the disease (RR = 4.9).

We have used a case-control approach to examine whether the Gc2 allele, in either the heterozygous or homozygous state, is associated with protection against development of COPD in a cohort of smokers undergoing resection of a lung cancer. In addition, we sought an association between Gc2 alleles and measurements of lung elastic recoil or upstream resistance. Finally, we examined whether the protection provided by the Gc2 allele was associated with a decrease in this isoform's ability to enhance the C5a-mediated chemotaxis of neutrophils. Our results show that all isoforms of Gc-globulin equally enhance the chemotactic ability of C5a and therefore the protective effect of the Gc2 allele cannot be attributed to this mechanism.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subject Group

Subjects for the study were recruited from 890 unrelated patients admitted to St. Paul's Hospital to undergo lobar or lung resection for a peripheral lung cancer. Prior to surgery, all patients gave informed consent and completed an interviewer-administered questionnaire regarding smoking history, occupational exposure to dusts or fumes, and respiratory symptoms. All of the subjects were of Caucasian ancestry. Lung function measurements performed on each patient included subdivisions of lung volume measured in a pressure-compensated body plethysmograph and maximal expiratory flow and volume. Values of forced expiratory volume in one second (FEV₁), forced vital capacity (FVC), and FEV₁/FVC ratio were calculated. Values for FEV₁ were expressed as percent predicted using the equations of Crapo (14). In approximately 60% of the subjects a pressure volume curve of the lung was also obtained, from which the maximal static recoil pressure (PL_max) was derived as previously described (15). In these subjects upstream airway resistance (RUS) was calculated by comparison of lung recoil pressure with airflow at isovolume (50% FVC) as previously described (16).

Patients in whom the lung lesion was obstructing a segmental or larger bronchus were not included in the study as this may influence lung function. All patients had postoperative examination of the resected specimen and lung histopathologic abnormalities. Any patient who had functional or pathologic evidence of a process other than those associated with COPD (such as significant obstructive pneumonitis) was excluded. These criteria excluded 358 patients from the study group. A further 28 patients were excluded because they were nonsmokers.

On the basis of lung function tests the remaining 504 patients were divided into those with and without significant airway obstruction. Individuals with FEV₁ < 80% predicted and FEV₁/FVC < 70% were classified as having COPD and placed into the obstructed category; those with FEV₁ > 85% predicted and FEV₁/FVC > 75% were classified as nonobstructed. There were 219 patients classified as obstructed, and 73 nonobstructed control subjects. This classification clearly differentiates individuals whose lung function was substantially affected by smoking from those whose lung function was not. There were 212 subjects with intermediate levels of lung function who were excluded from the study. We were able to extract DNA from, and genotype, 64 of the nonobstructed subjects. Seventy-five COPD patients were chosen at random to form the obstructed group.

Genotyping

DNA for genotyping was extracted from fixed or frozen lung tissue or blood using standard techniques (17). DNA was isolated from paraffin-embedded tissue blocks as previously described (18).

Polymerase chain reaction (PCR) primers were designed from the published Gc-globulin gene sequence (19). For amplification of exon X which contains both point mutations, we designed a novel upstream oligonucleotide primer: (5'TAA TGA GCA AAT GAA AGA AG3') to complement the downstream primer used by Braun and colleagues (20), (5'ACC TCC TCT TTA CTG TGA TT3'). The primers produce an amplified region of 388 base pairs (bp).

Amplification of DNA isolated from blood and frozen lung tissue was accomplished by adding 100 ng of genomic DNA to a 20-µl reaction mixture containing 1 unit of Taq DNA polymerase (Gibco BRL, Grand Island, NY), 1.5 mM MgCl₂, 1 µM of each primer, and 200 µM each of deoxyguanosine triphosphate (dGTP), deoxycytidine triphosphate (dCTP), deoxythymidine triphosphate (dTTP), and deoxyadenosine triphosphate (dATP) (Pharmacia, Baie D'Urfé, Québec). Amplification conditions were 30 cycles of 94° C for 30 s, 56° C for 30 s, and 72° C for 30 s. For DNA isolated from paraffin blocks all reagent concentrations remained the same, however the volume of the reaction mixture was increased to 30 µl, 0.3 µg of bovine serum albumin (BSA) (Pharmacia) was added, the annealing temperature was lowered to 54.5° C and the number of cycles was raised to 40.

The PCR product was genotyped by restriction fragment length polymorphism analysis. Digestion was accomplished by separately incubating the divided PCR product at 37° C overnight with both the Hae III and Sty I restriction enzymes (New England Biolabs, Mississauga, Ontario). The Hae III enzyme produces cut bands of 295 and 93 bp if the 1S allele is present. The Sty I enzyme cuts the Gc2 allele producing bands of 304 and 84 bp. The 1F variant remains uncut in both enzyme digests. The digested DNA fragments were then resolved on a 2.0% agarose gel with 0.2 µg/ml of ethidium bromide (Figure 1).

View larger version (58K): [in this window] [in a new window]   Figure 1.   Restriction fragment length polymorphism analysis of Gc-globulin genotypes. All six genotypes are shown. H = cut by Hae III restriction enzyme. S = cut by Sty I restriction enzyme. A cut 295 base pair (bp) lower band in lane H of the sample indicates the presence of a Gc1S allele; whereas a noncut (388 bp) upper band may either be a Gc1F or Gc2 allele. A cut band (304 bp) in lane S indicates a Gc2 allele. Sample 1 shows only uncut (388 bp) bands, demonstrating that the individual contains neither a Gc1S nor a Gc2 allele and is therefore of genotype 1F-1F.

Chemotaxis

Neutrophils were obtained by standard procedures (21) from healthy volunteers. Briefly, polymorphonuclear cells (PMNs) were pelleted with red blood cells (RBCs) using the ficoll method, after which RBC lysis was performed. After purification, neutrophils were diluted to 1.6 × 10⁶/ml in Hanks' buffer (Stemcell Technologies Inc., Vancouver, BC). Neutrophils were incubated with a 1/1,000 dilution of serum (obtained from individuals homozygous for one of the three Gc-globulin genotypes) for 30 min at 37° C to facilitate Gc-globulin interaction (22). Chemotaxis assays were then performed using the modified 48-well chemotaxis chamber (Neuroprobe, Cabin John, MD) and 5.0-µm pore size polycarbonate filters (Poretics, Livermore, CA) (similar to Kew and coworkers [22]). To the lower chambers Hanks' buffer and varying concentrations of C5a (0.01 to 10 nM) were added in either duplicate or triplicate, while 47 µl of the neutrophil suspensions were added to the upper wells. Migration occurred for 30 min at 37° C in 5% CO₂. Nonmigrated neutrophils were then removed and the membrane was fixed using 100% methanol. Neutrophils were stained with Diff-Quick (Starplex Scientific, Etobicoke, Ontario) and the filters were mounted. Migrated neutrophils were counted in three high-powered fields and the mean value for the buffer control was subtracted from each sample. Data are reported as mean number of migrated neutrophils.

Statistical Methods

The relationship between Gc-globulin alleles and COPD was assessed using logistic regression. Potential confounders examined were age, sex, and smoking. Smoking was measured by pack-years: the number of years smoked times the number of packs of 20 cigarettes smoked per day. Gc-globulin was considered as a categorical variable with three groups: (1) 1F-1F, 1S-1S, 1S-1F; (2) 2-1S, 2-1F; and (3) 2-2.

The associations between Gc-globulin and inflammatory airway narrowing (RUS) and elastic recoil (PL_max% predicted) were assessed by analysis of variance (ANOVA). Tests for trend were made using linear contrasts. Preliminary analysis showed no difference in either PL_max% predicted and RUS between patients with and without COPD, thus the estimated mean differences between Gc-globulin groups are not biased by the retrospective sampling scheme used.

Differences in chemotactic rates among neutrophils incubated with differing isoforms were assessed by analysis of variance.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patient Characteristics

Selection of lung cancer patients as a study group resulted in a population with high levels of exposure to cigarette smoke. Although the subjects were not matched, there was no significant difference with respect to smoking history between the obstructed and nonobstructed groups. Pack-years of exposure (mean ± SD) were 52 ± 30 for the obstructed group and 48 ± 27 for the nonobstructed group. The characteristics and lung function results of each genotypic group are summarized in Table 1. There were no significant differences among groups for age or the number of cigarettes smoked, however a significant difference in the percentage of COPD patients within the genotypic classes was observed. There were no significant differences in lung function, age, sex, and smoking history between those patients who had pressure volume curves available and those who did not (Table 2).

Genotyping and Lung Function

Gc-globulin allele frequencies for the patients were: Gc1S 0.53, Gc1F0.19, and Gc20.28. These values are similar to allelic distributions reported from large population studies (10).

Gc-genotype was found to be a significant predictor of COPD (p = 0.04). The odds ratio for COPD for Group 2 compared with Group 1 was 1.01 (95% CI = 0.49 to 2.1). The odds ratio for Group 3 compared with Group 1 was 0.17 (95% CI = 0.03 to 0.83). Thus the 2-2 genotype (Group 3) is significantly protective against development of COPD. The odds ratios were unchanged after adjustment for the potential confounders (age, sex, and cigarette smoking: none of which were associated with the presence of COPD in the study).

There were no significant differences in the measurement of lung elastic recoil (PL_max) or upstream airway resistance (RUS) as a function of genotypic class (Table 3).

Chemotaxis

Neutrophils incubated for 30 min with serum of any Gc-globulin isotype demonstrated significantly enhanced chemotaxis in response to C5a compared with neutrophils incubated in the absence of serum (p < 0.05). However, there was no difference in chemotaxis regardless of which isoform was present with the neutrophils (Figure 2).

View larger version (34K): [in this window] [in a new window]   Figure 2.   Number of neutrophils migrated (mean ± SD) after a 30-min incubation with either serum homozygous for the various Gc-globulin isoforms, or without serum.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Our work, using a novel method of Gcg genotyping, has demonstrated that in the homozygous form the Gc2 allele is protective against pulmonary deterioration due to chronic cigarette smoke. However, the protective effect could not be attributed to differences in enhancement by Gc-globulin of neutrophil chemotaxis in response to C5a. These results confirm and extend findings from previous association studies (11, 13). Kueppers and coworkers compared the prevalence of Gc-globulin isotypes in 114 COPD patients and 114 control subjects matched for age, sex, occupation, and smoking history (11). The results of their analysis showed a significantly lower frequency of the homozygous Gc2 phenotype among the patient population (RR = 0.2). This result was confirmed by a subsequent study (13) in which it was found that, in comparison with the general population (n = 413), the prevalence of Gc2 homozygotes and heterozygotes was reduced in a COPD group (n = 104) (relative risks for Gc2-2, Gc2-1S, and Gc2-1F were 0.8, 0.7, and 0.5, respectively). In addition, in the latter study the investigators found that the homozygous Gc1F phenotype was a significant risk factor for airway obstruction (relative risk of 4.8). A study by Kauffmann and coworkers (12), using a different experimental design, failed to confirm the associations. In their comparison, between a group of heavy smokers (n = 45) with preserved lung function (mean FEV₁, 3.80 L) and a group of nonsmokers (n = 43) with low lung function (mean FEV₁, 2.56 L), they found no difference in Gc-globulin phenotypic frequencies between the two groups.

There are several factors that distinguish the present study from the previous publications. This study was the first to use DNA-based methods to identify the subjects' genotypes, and this allowed us to use a well-characterized patient cohort that previously could not be studied for these polymorphisms. We used a novel study design which drew the cases and controls from the same group of patients and therefore minimized the chances of a systematic bias in the selection of the subjects. This study was the first to investigate different aspects of COPD pathogenesis, i.e., loss of elastic recoil and peripheral airway narrowing. We have presented the first investigation of a possible mechanism for the involvement of Gc-globulin in the susceptibility to COPD. Investigations of this type are needed to determine whether the Gc2 allele is responsible for the protective effect against COPD or is merely in linkage disequilibrium with the true causal allele.

Although we found preserved pulmonary function among 2-2 individuals, we were unable to verify results obtained by Horne and coworkers (13) that suggested an increased risk of COPD among 1F-1F individuals. This genotype is present in less than 4% of the population, and as we only identified three 1F-1F individuals (one of whom was obstructed) we had insufficient numbers to test this hypothesis. Also, we found that heterozygosity for the Gc2 allele was not associated with preserved lung function. Therefore, it is possible that the Gc-globulin gene has a recessive mode of action, and is only protective in the homozygous state.

The confirmation of previous studies showing that Gc2 offers protection from COPD suggests that this isoform could play a biological role in maintaining the structural integrity of the lung when chronically insulted by cigarette smoke. It is known that loss of lung function in COPD may result from two pathophysiological mechanisms. First, release of extracellular proteases may cause the destruction of the lung parenchyma, increasing the size of airspaces and leading to emphysema. Second, inflammation of the peripheral airways associated with edema, mucous hypersecretion, and fibrosis leads to the narrowing of these airways and an increase in their resistance. Both processes decrease the maximal expiratory flow from the lung. To affect these processes a gene product could alter either the intensity or the consequence of the inflammatory response to cigarette smoke. However, we were unable to demonstrate that Gc2-2 individuals had a lower upstream resistance or better preserved lung elastic recoil.

The data from this study suggest that the Gc2 allele is protective against the development of COPD. However, there are alternative interpretations of these results. Unidentified population stratification could give rise to a spurious association, even though all the subjects were of Caucasian ancestry. Alternatively, the Gc2 allele may not affect susceptibility to COPD, but could be in linkage disequilibrium with the true susceptibility allele in a closely linked gene. The Gc-globulin gene is located on chromosome 4 and is closely linked to the genes for albumin and -fetoprotein. There are no obvious mechanisms for the involvement of these genes in the pathogenesis of COPD. However, there may be another gene in disequilibrium with Gc-globulin which could affect susceptibility to COPD.

Gc-globulin has two different biological functions in relation to inflammation, both of which could be involved in the development of COPD. First, Gc-globulin has been shown to interact directly with neutrophils to increase their chemotactic rates to the C5a peptide produced during the activation of the complement cascade (22). Second, Gc-globulin is known to undergo conversion to a potent macrophage activating factor (MAF) (26, 27). Because current hypotheses suggest that neutrophils and macrophages damage the lung by the release of toxic free radicals and proteases, variability among Gc-globulin isoforms, with respect to these functions, could play a part in determining the degree of damage to the lung parenchyma.

Because we demonstrated that, in vitro, the ability of serum from homozygous individuals with the different Gc-globulin isoforms showed no significant difference in the ability to attract neutrophils, we suggest that an alternate explanation must underlie our result.

Besides increasing neutrophil chemotaxis, Gc-globulin increases the activation of macrophages at the sites of inflammation. The conversion of Gc-globulin to a MAF is accomplished by the removal of specific glycosylated moieties from the protein. The moieties are cleaved by glycosidases present on both B and T cells that are activated by the presence of lysophospholipids (released as cells die at the site of inflammation). Interestingly, the glycoside side chain structure varies among genotypes. Both the Gc1S and Gc1F isoforms contain identical side chain structure, while the Gc2 isoform has alternate sugar moieties. The Gc2 protein is found in two different forms. The less prevalent form (10%) displays a shorter side chain that need only be modified by T-cells to become a MAF. However, at least 90% of the Gc2 protein molecules undergo no glycosylation and are incapable of being converted to a MAF. The low glycosylation level in homozygous Gc2 individuals could result in less MAF being produced and, therefore, less pulmonary inflammation.

Although the biological mechanism remains unknown, this study both provides additional evidence for a protective effect of Gc2 isoforms in chronic cigarette smokers and demonstrates a novel, effective method of Gcg genotyping.