Immunoglobulin GM and KM allotypes are associated with antibody responses to Pseudomonas aeruginosa antigens in chronically infected cystic fibrosis patients

Abstract

Background:

Chronic infection with Pseudomonas aeruginosa (P. aeruginosa) is a leading cause of death in patients with cystic fibrosis (CF). Immunobiology of P. aeruginosa infection is complex and not well understood. Chronically infected CF patients generate high levels of antibodies to P. aeruginosa, but this response does not lead to clinical improvement. Therefore, additional studies aimed at identification and understanding of the host factors that influence naturally occurring immune responses to P. aeruginosa are needed. In this investigation, we evaluated the contribution of immunoglobulin GM (γ marker) and KM (κ marker) allotypes to the antibody responses to P. aeruginosa lipopolysaccharide (LPS) O1, O6, O11, and alginate antigens and the broadly-conserved surface polysaccharide expressed by many microbial pathogens, poly-N-acetyl-D-glucosamine (PNAG), in 58 chronically infected CF patients.

Methods:

IgG1 markers GM 3 and 17 and IgG2 markers GM 23− and 23+ were determined by a pre-designed TaqMan® genotyping assay. The κ chain determinants KM 1 and 3 were characterized by PCR-RFLP. Antibodies to the LPS O antigens, alginate, and PNAG were measured by an ELISA.

Results:

Several significant associations were noted with KM alleles. Particular KM 1/3 genotypes were individually and epistatically (with GM 3/17) associated with the level of IgG antibodies to O1, O11, alginate, and PNAG antigens.

Conclusions:

Immunoglobulin GM and KM genotypes influence the magnitude of humoral immunity to LPS O, alginate, and PNAG antigens. These results, if confirmed in a larger study population, will be helpful in devising novel immunotherapeutic approaches against P. aeruginosa.

1. Introduction

Chronic infection with Pseudomonas aeruginosa (P. aeruginosa) is a leading cause of death in patients with cystic fibrosis (CF). These infections are difficult to treat due to their intrinsic and acquired resistance to many antibiotics. Therefore, alternative therapies for the treatment for P. aeruginosa infections are warranted. Identification and understanding of the host factors that influence naturally occurring immune responses to P. aeruginosa is an important prerequisite to successfully designing efficacious immunotherapeutic strategies against this pathogen. Our knowledge in this area is incomplete. High pathogen specific antibody responses, in general, are protective against the respective pathogens. For P. aeruginosa, however, this does not appear to be the case. Chronically infected CF patients generate high levels of antibodies to P. aeruginosa antigens, but this response does not lead to clinical improvement [1]. In fact, high levels of IgG2 antibodies to the O-antigen of P. aeruginosa have been shown to impair complement-mediated bacterial killing, resulting in failure to achieve bacterial eradication [2]. Therefore, a deeper understanding of the mechanisms underlying the host immunity to P. aeruginosa is imperative.

There are interindividual differences in the magnitude of antibody responses to P. aeruginosa antigens, but the host factors influencing these differences are incompletely understood. The aim of the present investigation was to determine whether GM and KM allotypes—genetic markers of IgG heavy and κ-type light chains, respectively—influence the level of IgG antibody responses to P. aeruginosa lipopolysaccharide (LPS) O antigens O1, O6, O11, and alginate, and to the conserved PNAG surface polysaccharide expressed by many other CF pathogens, such as Staphylococcus aureus, Burkholderia spp. and non-tuberculosis Mycobacteria.

Immunoglobulin GM and KM allotypes are encoded by immunoglobulin heavy chain G (IGHG) and immunoglobulin κ constant (IGKC) genes on chromosomes 14 and 2, respectively. They have been shown to be associated with susceptibility to several infectious, autoimmune, and malignant diseases and with immune responsiveness to a variety of self and nonself antigens [3–5]. GM alleles—expressed primarily on the Fc portion of the γ chain—influence the binding affinity between Fcγ and FcγR proteins, and contribute to the Fc-mediated effector functions, such as antibody-dependent cellular toxicity, making them excellent candidates for influencing the magnitude of antibody responsiveness to P. aeruginosa antigens [6,7]. Immunoglobulin KM alleles have been shown to be associated with immunity to many infectious pathogens, some relevant to the CF pathology, such as Haemophilus influenzae and Aspergillus fumigatus [8,9]. Both GM and KM allotypes could constitute receptors for antigenic epitopes on B-cell membrane bound IgG.

2. Materials and methods

2.1. Study subjects

DNA and plasma samples were obtained from CF patients enrolled at the University of North Carolina, Chapel Hill. The sample of n=58 patients was 50% male and 50% female. At the time of blood draw, the mean ± SD age was 19.0 ± 7.1, and their forced expiratory volume (FEV1) percent predicted was 68.0 ± 23.4. The patients were homozygous for the allele encoding the amino acid phenylalanine at position 508 of the cystic fibrosis transmembrane conductance regulator protein. All patients were chronically infected with P. aeruginosa. Chronic infection was defined as 3 consecutive years of culture data, with at least one positive culture in 2 of the 3 years. The mean duration of P. aeruginosa positivity for these subjects was at least 9.2 years. The study was approved by the relevant Institutional Review Boards for human research.

2.2. Genotyping

IgG1 markers GM 3 and 17 (arginine to lysine), were determined by a pre-designed TaqMan® genotyping assay from Applied Biosystems Inc., employing the following primers and probes:

Forward primer: 5‘ CCCAGACCTACATCTGCAACGTGA-3‘

Reverse primer: 5‘ CTGCCCTGGACTGGGACTGCAT-3‘

Reporter 1 (GM 17-specific): VIC-CTCTCACCAACTTTCTTGT-NFQ

Reporter 2 (GM 3-specific): FAM-CTCTCACCAACTCTCTTGT-NFQ

IgG2 markers GM 23− and 23+ (valine to methionine), were also determined by a TaqMan® genotyping assay, employing the following primers and probes:

Forward primer: 5‘ CCCGAGGTCCAGTTCAACT-3‘

Reverse primer: 5‘ CGTGGCTTTGTCTTGGCATTATG-3‘

Reporter 1 (GM 23-specific): VIC-CACCTCCACGCCGTC-NFQ

Reporter 2 (GM 23+specific): FAM- CACCTCCATGCCGTC -NFQ

The κ chain determinants KM 1 and 3 were characterized by a previously described polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) technique [10]. Three alleles—KM 1, KM 1,2, and KM 3—segregate at the KM locus on chromosome 2. Over 98% of the people positive for KM 1 are also positive for KM 2 [11]. In this investigation, positivity for KM 1 includes both KM 1 and KM 1,2 alleles.

2.3. Antibody measurements

Antibodies to LPS O antigens O1, O6, O11, alginate, and PNAG were measured by an enzyme-linked immunosorbent assay (ELISA), using plasma from CF patients. Flat bottomed Immulon 4HBX plates and Fc-specific goat anti-mouse IgG-HRP conjugate were purchased from Thermo Fisher Scientific (Waltham, MA). Fc-specific anti-human IgG-HRP conjugate and monoclonal anti-human IgG1, IgG2, IgG3 and IgG4 antibodies were purchased from Millipore Sigma (St Louis, MO). All antigens except PNAG were dissolved in sensitizing buffer (0.04 M Phosphate buffer, pH 7.4). PNAG was first solubilized in 5 N HCl and neutralized with an equal volume of 5 M NaOH. It was further diluted with sensitizing buffer before coating.

Ninety-six well microtiter plates were incubated with target antigens (6 μg/ml for PNAG and 10 μg/ml for others) in 100 μl sensitizing buffer overnight at 4°C. Plates were washed with PBS (0.2 M phosphate, 0.15M NaCl; pH 7.4) containing 0.05% tween 20 and 1% skimmed milk and blocked by incubation at 37°C with PBS containing 0.05% tween 20 and 5% skimmed milk for 1 hr. Plates were then washed and incubated at 37°C with suitably diluted patient plasma for 1 hr. Plates were further washed and incubated at 37°C with anti-human IgG HRP conjugate for 30 min. Finally, plates were washed and incubated with HRP substrate hydrogen peroxide along with 3,3′,5,5′-Tetramethylbenzidine as chromogenic substrate in citrate phosphate buffer (pH 5.5). Reaction was stopped after 20 min by the addition of 100 μl of 2N HCl and the absorbance values at 450 nm were monitored in a BioTek ELISA reader.

For the quantitation of IgG subclasses, antigen-coated plates were incubated with patient plasma and then with suitably diluted mouse monoclonal anti-human IgG-Fc antibodies. Plates were further incubated with anti-mouse IgG HRP conjugate and assay was continued as before. Wells containing plasma diluent buffer alone were used as blank. Absorbance values of blank wells were subtracted from sample wells. The levels of total IgG, as well as IgG subclass-specific, antibodies to particular antigens were expressed as arbitrary units per mL (AU/mL), after multiplying with the dilution factor.

2.4. Statistical analysis

GM and KM genotype determinations were made blinded with respect to the status of antibodies to the target antigens, and the results were provided to an independent biostatistician (PJN) who conducted the analyses. To determine the significance of the association between GM and KM genotypes and antibody responses, parametric tests (t-test and ANOVA) were used when the antibody levels or their log-transformed values were normally distributed, and non-parametric tests (Wilcoxon rank sum and Kruskal-Wallis) were used otherwise. Genetic epistasis was also investigated by ANOVA and Kruskal-Wallis. All tests were two-tailed, and the statistical significance was defined as p < 0.05. The p values were not adjusted for multiple comparisons, as this was largely a hypothesis-generating study. Cohen’s d effect sizes were also calculated to provide a standardized measure of the magnitude of the difference in antibody responses [12]. Sex differences were also explored. Analyses were conducted using SAS v9.4 (SAS Institute, Cary, NC).

3. Results

Table 1 presents the mean levels of antibodies to the P. aeruginosa and the PNAG antigens in relation to GM and KM genotypes in chronically infected CF patients. The levels of IgG4 antibodies to the alginate antigen of P. aeruginosa were slightly but not significantly higher in patients homozygous for the GM 3 allele compared to heterozygotes and GM 17 homozygotes (mean ± SD: 36.5±13.8 vs. 29.8±8.9 AU/mL, p = 0.062). The GM 3/17 genotypes were not significantly associated with antibody responses to any other P. aeruginosa antigens. The alleles at the GM 23 locus did not influence antibody responsiveness to the analyzed P. aeruginosa antigens (data not shown).

Table 1.

Mean levels (AU/mL) of antibodies to P. aeruginosa antigens in relation to GM and KM genotypes in chronically infected CF patients.

Antibody	N	Mean	SD	(95% CI)	N	Mean	SD	(95% CI)	Effect Size	p-value
	Genotype GM 3/3				Genotypes GM 3/17, 17/17				d
Anti-O1 whole IgG	29	363.5	204.4	(285.7, 441.2)	29	397.9	236	(308.2, 487.7)	−0.16	NS
Anti-O1 IgG1	29	54.8	40.2	(39.5, 70.1)	29	59.5	35	(46.2, 72.8)	−0.12	NS
Anti-O1 IgG2	29	318.1	198.7	(242.5, 393.6)	29	329.3	216.6	(247.0, 411.7)	−0.05	NS
Anti-O1 IgG3	29	46.2	38.4	(31.6, 60.8)	29	52.8	42	(36.9, 68.8)	−0.17	NS
Anti-O1 IgG4	29	61.7	42.6	(45.5, 77.9)	29	68.2	42.5	(52.0, 84.3)	−0.15	NS
Anti-alginate whole IgG	29	367	143.7	(312.3, 421.6)	29	316.3	118.4	(271.3, 361.4)	0.38	NS
Anti-alginate IgG1	29	30.6	11.7	(26.1, 35.0)	29	28.8	12	(24.2, 33.4)	0.15	NS
Anti-alginate IgG2	29	474.3	167.4	(410.7, 538.0)	29	394.4	172.8	(328.7, 460.1)	0.47	NS
Anti-alginate IgG3	29	24.4	14.8	(18.8, 30.0)	29	20.8	10.2	(16.9, 24.7)	0.29	NS
Anti-alginate IgG4	29	36.5	13.8	(31.2, 41.7)	29	29.8	8.9	(26.4, 33.2)	0.57	0.062
Anti-O6 whole IgG	29	23.4	17.5	(16.7, 30.0)	29	24.4	16.7	(18.0, 30.8)	−0.06	NS
Anti-O11 whole IgG	29	23.6	15.7	(17.6, 29.6)	29	28.2	14.9	(22.6, 33.9)	−0.3	NS
Anti-PNAG whole IgG	29	20.9	21.5	(12.8, 29.1)	29	19	19.4	(11.6, 26.4)	0.1	NS
	Genotype KM 3/3				Genotypes KM 1/3, 1/1
Anti-O1 whole IgG	51	402.2	222.7	(339.6, 464.9)	7	223.9	107.8	(124.1, 323.6)	0.84	0.045
Anti-O1 IgG1	51	60.7	38.1	(50.0, 71.4)	7	31.5	18.3	(14.5, 48.4)	0.8	0.015
Anti-O1 IgG2	51	341.5	210.3	(282.4, 400.7)	7	193.9	112.2	(90.2, 297.7)	0.73	NS
Anti-O1 IgG3	51	53.4	41.1	(41.9, 65.0)	7	20.9	10.2	(11.5, 30.3)	0.84	0.04
Anti-O1 IgG4	51	69.6	42.8	(57.6, 81.6)	7	31.1	15.7	(16.6, 45.6)	0.95	0.014
Anti-alginate whole IgG	51	345.9	128.9	(309.6, 382.1)	7	311	168.5	(155.1, 466.8)	0.26	NS
Anti-alginate IgG1	51	30.3	10.84	(27.2, 33.3)	7	25.4	17.8	(9.0, 41.8)	0.41	NS
Anti-alginate IgG2	51	442	166.5	(395.2, 488.8)	7	378.6	224.6	(170.9, 586.3)	0.37	NS
Anti-alginate IgG3	51	23.8	12.9	(20.2, 27.5)	7	13.7	5.6	(8.5, 18.9)	0.82	0.017
Anti-alginate IgG4	51	34.3	12	(30.9, 37.7)	7	24.7	8.3	(17.0, 32.4)	0.82	0.026
Anti-O6 whole IgG	51	24.6	17	(19.8, 29.4)	7	18.5	16.8	(2.9, 34.0)	0.36	NS
Anti-O11 whole IgG	51	27.2	15.6	(22.8, 31.6)	7	16.6	9.7	(7.6, 25.5)	0.7	0.044
Anti-PNAG whole IgG	51	21.4	21.3	(15.4, 27.4)	7	9.4	3.5	(6.2, 12.6)	0.6	0.038

Several significant associations of moderately large magnitude were noted that involved the KM genotype. For the O1 antigen of P. aeruginosa, whole IgG as well as IgG1, IgG3, and IgG4 subclass-specific antibody levels were significantly higher in KM 3 homozygotes compared to the heterozygotes and KM 1/1 homozygotes: whole IgG (mean ± SD: 402.2±222.7 vs. 223.9±107.8 AU/mL, d = 0.84, p = 0.045); IgG1 (mean ± SD: 60.7±38.1 vs. 31.5±18.3 AU/mL, d = 0.80, p = 0.015); IgG3 (mean ± SD: 53.4±41.1 vs. 20.9±10.2 AU/mL, d = 0.84, p = 0.040); IgG4 (mean ± SD: 69.6±42.8 vs. 31.1±15.7 AU/mL, d = 0.95, p = 0.014). The levels of IgG3 and IgG4 antibodies to the alginate antigen were higher in patients homozygous for the KM 3 allele compared to heterozygotes and KM 1 homozygotes: IgG3 (mean ± SD: 23.8±12.9 vs. 13.7±5.6 AU/mL, d = 0.82, p = 0.017); IgG4 (mean ± SD: 34.3±12.0 vs. 24.7±8.3 AU/mL, d = 0.82, p = 0.026). The levels of whole IgG antibodies to O11 and PNAG antigens were higher in patients homozygous for the KM 3 allele compared to heterozygotes and KM 1 homozygotes: O11 (mean ± SD: 27.2±15.6 vs. 16.6±9.7 AU/mL, d = 0.70, p = 0.044); PNAG (mean ± SD: 21.4±21.3 vs. 9.4±3.5 AU/mL, d = 0.60, p = 0.038). Among these significant findings, the observed effect sizes were generally similar between males and females; however, for anti-O1 IgG1, the effect size was over 2-fold larger for females compared with males (d=1.10 vs. 0.49), and for anti-O11 whole IgG, the effect size was over 3.5-fold larger for females (d=1.03 vs. 0.29). These sex differences should be viewed with caution, given that they rely on rather small sample sizes (n=29 males and n=29 females).

In addition to its main effects described above, the KM 1/3 genotype appears to have epistatically (with GM 3/17) contributed to the interindividual differences in the level of IgG4 antibodies to alginate. Thus, CF patients lacking the GM 17 and KM 1 alleles had the highest, and those positive for these alleles had the lowest, levels of anti-alginate IgG4 antibodies (mean ± SD: 38.7±13.7 vs. 22.2±11.7AU/mL, p = 0.015, Fig. 1).

Fig. 1.

Interaction between GM 3/17 and KM 1/3 genotypes on anti-alginate IgG4 antibody levels. The boxplots represent medians (solid lines) and means (diamond markers), along with interquartile ranges (box lengths) and outlying values (circles). The ANOVA model indicated that the mean antibody levels were significantly different (p=0.015) across the 4 groups.

4. Discussion

We found a slightly higher level of IgG4 antibodies to the alginate antigen of P. aeruginosa in subjects homozygous for the GM 3 allele. In contrast, an earlier study found an association of GM 3 and GM 23 alleles with IgG3 antibodies to alginate [13]. The two studies, however, are not comparable. In addition to the differences in the sample size and the study population (Danish vs. American), the studies differed in allotyping methodology (serological vs. molecular).

The levels of IgG subclass-specific (except IgG2) as well whole IgG antibodies to the O1 antigen were higher in the KM 3 homozygotes. The levels of whole IgG antibodies to O11 and the PNAG antigen were also higher in KM 3 homozygotes. The influence of KM and GM allotypes on antibody responses to O and PNAG antigens has not been previously investigated. The KM homozygotes also had higher levels of IgG3 and IgG4 antibodies to alginate. In the previous study [13], higher levels of IgG2 antibodies to alginate were associated with the KM 3 allele. In addition to the main effects of KM alleles on antibody responses to alginate, we also found evidence for epistasis, defined as the modulation of the action of a gene by a gene at another (usually unlinked) locus. Thus, GM 17 encoded on chromosome 14 may interact with KM 1 encoded on chromosome 2 to influence the magnitude of IgG4 antibody responses to alginate. Because of relatively small sample sizes in some of our subgroups, such associations should be verified in the context of a larger study.

Mechanistically, GM and KM allotypes could influence antibody responsiveness to the P. aeruginosa antigens and PNAG through the B-cell mediated antigen processing/presentation pathway, by being part of the recognition structure for these epitopes on the B-cell membrane-bound IgG. Perhaps the B-cell membrane bound IgG expressing certain GM and KM specificities constitute a lower/higher affinity receptor for the critical epitopes of O, alginate, and PNAG antigens, resulting in weak/strong humoral response to them. GM allotypes could also cause conformational changes in the antigen-binding site in the immunoglobulin variable regions associated with antibody specificity to these antigens. The GM 3 and GM 17 alleles are expressed in the CH1 region of γ1 chain. In murine studies, amino acid sequence polymorphism in this region has been shown to modulate the kinetic competence of antigen binding sites [14]. Significant interaction between GM 17 and KM 1 alleles on the IgG4 antibody responses to alginate may be a reflection of preferential association of heavy and light chains of particular genotypes in the synthesis of an antibody molecule. Nonrandom pairing of heavy and light chains has been shown in experimental animals [15,16]. Possibly, only γ chains expressing the GM 17 allele and κ chains expressing the KM 1 allele might form a paratope with the necessary quaternary structure for an effective recognition of the alginate epitopes.

It might be relevant to point out that in addition to P. aeruginosa, immunity to several other bacterial pathogens relevant to the CF pathology, such as Haemophilus influenzae, Mycobacterium, Moraxella catarrhalis, and Aspergillus fumigatus, has been shown to be associated with GM and KM allotypes [9,17–20]. It would be of interest to investigate the contribution of GM and KM genes to the antibody responses to these pathogens in patients with CF.

In sum, we have presented evidence for the involvement of immunoglobulin GM and KM genes in antibody responses to LPS O, alginate, and PNAG antigens in patients with CF. It is important to note that there may be other genetic (e.g. FcγR and HLA) and non-genetic confounders relevant to CF pathology that may contribute to the interindividual differences in antibody responsiveness to P. aeruginosa, but this study is underpowered to evaluate such putative covariates. Additional studies, using a larger and independent study population, are warranted.