Figures
Abstract
Introduction
Human beta-defensins are key components of human innate immunity to a variety of pathogens, including Staphylococcus aureus. The aim of the present study was to investigate a potential association between gene variations in DEFB1 and DEFB103/DEFB4 and the development of S. aureus bacteremia (SAB) employing a case-control design.
Methods
Cases were unique patients with documented SAB, identified with the National S. aureus Bacteremia Register, a comprehensive dataset of all episodes of community associated-SABs (CA-SAB) occurring in children (≤20 yrs) in Denmark from 1990 to 2006. Controls were age-matched healthy individuals with no history of SAB. DNA obtained from cases and controls using the Danish Newborn Screening Biobank were genotyped for functional polymorphisms of DEFB1 by Sanger sequencing and copy number variation of the DEFB103 and DEFB4 genes using Pyrosequencing-based Paralogue Ratio Test (P-PRT).
Results
193 ethnic Danish SAB cases with 382 age-matched controls were used for this study. S. aureus isolates represented a variety of bacterial (i.e., different spa types) types similar to SAB isolates in general. DEFB1 minor allele frequencies of rs11362 (cases vs. controls 0.47/0.44), rs1800972 (0.21/0.24), and rs1799946 (0.32/0.33) were not significantly different in cases compared with controls. Also, DEFB4/DEFB103 gene copy numbers (means 4.83/4.92) were not significantly different in cases compared with controls.
Citation: Fode P, Larsen AR, Feenstra B, Jespersgaard C, Skov RL, Stegger M, et al. (2012) Genetic Variability in Beta-Defensins Is Not Associated with Susceptibility to Staphylococcus aureus Bacteremia. PLoS ONE 7(2): e32315. https://doi.org/10.1371/journal.pone.0032315
Editor: Michael Otto, National Institutes of Health, United States of America
Received: September 8, 2011; Accepted: January 26, 2012; Published: February 22, 2012
Copyright: © 2012 Fode et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: The work was supported by Danish Medical Research Council; Idella Foundation; and The Lundbeck Foundation. Dr. Fowler was supported by NIH grant R01-AI068804. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Staphylococcus aureus is a leading cause of bacteremia and endocarditis in the industrialized world. In 2008, the incidence of S. aureus bacteremia (SAB) in Denmark was 25.6 cases per 100,000 inhabitants [1]. Approximately 20% of these cases had a fatal outcome [1].
A growing body of evidence indicates that host genetic factors are involved in susceptibility to a variety of bacterial pathogens, including meningococci [2], Mycobacterium tuberculosis [3], and leprosy [4], [5]. Similar evidence also exists for genetic susceptibility to S. aureus. For example, higher rates of invasive S. aureus infection have been described in ethnically distinct populations, including Australian Aborigines, New Zealand Maori, and Canadian Native Americans [6]. Using a murine S. aureus sepsis model, Ahn et al. recently demonstrated that inbred A/J mice were highly susceptible to S. aureus infection as compared to C57BL6, and that this susceptibility was associated with regions on A/J chromosomes 8, 11, and 18, chromosomes that included defensin genes and several other innate immune genes [7]. Additionally, investigators at Erasmus University recently suggested that variability within key innate immunity genes were associated with persistent S. aureus carriage [8]. Despite these findings, however, the precise genetic determinants for susceptibility to S. aureus infection in humans are unknown.
Defensins are small cationic peptides with antimicrobial activity [9]–[12] and shown to be important players in the innate immune system. The majority of the defensin genes including DEFB1, DEFB4 and DEFB103 encoding human beta-defensins 1, 2 and −3, respectively [13], are located on chromosome 8p23.1, a region of great complexity with both functional SNPs (DEFB1) and copy number variation (CNV) of a wide range of beta-defensin genes including DEFB4 and DEFB103. DEFB1 is apparently constitutively expressed [14], but functional single nucleotide polymorphisms (SNPs) in DEFB1 have been associated with susceptibility to severe sepsis [15]. Furthermore, human beta-defensin 3 (HBD3) has a strong activity against S. aureus [16]. The genomic copy number (GCN) for DEFB4/DEFB103 varies from 2 to 12 copies per diploid genome [13]. Interestingly, an increased GCN of DEFB4/DEFB103 is associated with exacerbations of psoriasis, an inflammatory skin disease characterized by a notable absence of S. aureus colonization [17].
The combination of strong activity against S. aureus, functional DEFB1 SNPs and high variability in DEFB4/DEFB103 GCN makes these genes promising candidates for human susceptibility to S. aureus infections. In the present study we used a unique cohort of all pediatric patients with SAB in Denmark over a 16-year period to evaluate potential associations between either functional SNPs of DEFB1 or low GCN of DEFB4/DEFB103 and pediatric SAB.
Materials and Methods
Ethics Statement
The present study was approved by the Regional Science Ethics Committee of Copenhagen (2007-0104), the Danish Data Protection Agency (2008-54-0458), and The Danish Newborn Screening Biobank's review board a similar, but more independent review board than an Institutional Review Board. The parents are informed verbally and by homepage (www.ssi.dk/nyfoedte) about the use of The Danish Newborn Screening Biobank and can opt out at any time [18].
Identification of cases and controls
The present study included all Danish SAB pediatric patients in the period 1990–2006. For the purposes of this study, pediatric patients were defined as ≤20 years of age at the time of their SAB episode. SAB acquisition was classified as Hospital associated (HA)-SAB, Community associated (CA)-SAB or unknown, based on the time of positive SAB sampling after hospital admission. A SAB episode was registered as CA-SAB if diagnosed in less than 48 h after hospital admission based on information retrieved prospectively from discharge summaries. These data were retrieved from The Danish S. aureus Bacteremia Database at SSI, containing all Danish SAB isolates and corresponding bacteriological and patient data since 1958. Patients were included if they were born after 1982, were of Danish origin, and had CA-SAB. Danish origin was defined as individuals where both of the individual's parents were born in Denmark.
For each case two Danish individuals (same definition as for cases) from The Danish Newborn Screening Biobank who according to the Danish National Patient Registry had not been diagnosed with SAB or other severe bacterial infections served as controls. Samples for these control samples were adjacent to the case samples in the biobank regardless of gender. Two controls suffered from diabetes and two had died from cancer.
Bacteriological data
All S. aureus isolates were retrieved from The Danish S. aureus Bacteremia Repository and were spa-typed according to standard practise and assigned to clonal complexes (CC) [19].
DNA extraction
DNA was extracted from 3 mm filter paper blood samples using Generation DNA Elution solution and DNA purification solution (Qiagen, Valencia, CA) as described by Baker et al. [20]. Extraction was performed in 96-well format with each well containing a dried blood sample. Washing buffer (PBS, 0.5% Tween20) was added to each well and the samples were shaken at 1000 rpm for 45 min at room temperature before the supernatant was removed. The washing procedure was repeated. Purification solution was added and the supernatant was removed. The procedure was repeated. An elution solution was then added and after 5 min the supernatant was removed. Sterile water was added, and the plate was placed at −20°C for 15 min before being heated to 99°C for 15 min.
DEFB1 SNP analysis
Genotyping of three DEFB1 5′UTR SNPs (rs11362 [−20G/A]; rs1800972 [−44 C/G] and rs1799946 [−52 G/A]) was performed using Sanger DNA sequencing after obtaining a 318 bp fragment (Forward primer: 5′-CTC CCT TCA GTT CCG T-3′ and reverse primer 5′-CTT GTT CCT CGT CCC TT-3′). Hardy-Weinberg equilibrium was seen in cases and controls separately and combined for all three SNPs.
DEFB4/DEFB103 Copy number determination
Gene copy number was determined using the Pyrosequencing-based Paralogue Ratio Test (P-PRT). The method is based on the Paralogue Ratio Test (PRT) described previously [21], [22]. In brief, the DEFB103 region on chromosome 8 and an identified paralogue gene (HSPD21 on chromosome 21) with only two copies per genome were PCR amplified using one set of primers. The resulting PCR amplicons differed at 10 positions. One of the positions where the amplicons differed was used to quantify the two chromosome regions against each other by pyrosequencing across it. The P-PRT method used has recently been described [23].
Primers for the pyrosequencing assay were designed using the PSQ assay design software version 1.0.6 (Qiagen, Hilden, Germany). The following sequence was analysed with the position that varied between chromosomes 8 and 21 marked in bold with underline: KATGCYAT. For the PCR, 20 ng of template DNA in a total volume of 50 µl using a forward primer (5′-GAGGTCACTGTGATCAAAGAT-3′) and a reverse primer (5′-Biotin- AACCTTCAGCACAGCTACTC-3′) was used. Pyrosequencing was carried out using 40 µL of the PCR product and a sequencing primer (5′-AGGTCACTGTGATCAAAGAT-3′) on the PSQ 96 MA Pyrosequencer according to manufacturer's recommendations (Qiagen).
The relative percentages of the two variants were calculated by the Pyrosequencing software (Qiagen) and used for the gene copy number determination. Positive controls from Coriell Cell Repositories with known copy number were included in each run (NA07048: 4 copies; NA10846: 5 copies; NA10847: 7 copies and NA10861 3 copies) and used to generate a correction curve by linear regression. Corrected copy number estimates were calculated for each sample using this run-specific regression equation. A “No template control” was included in each run.
Statistical analysis
In the DEFB1 SNP analyses, we tested for association between SNP genotype and case-control status using an additive genetic model. We also tested for possible non-additive effects using dominant, recessive, and full genotype models. Association tests were performed using PLINK [24]. Mean gene copy numbers of cases and controls were compared using unpaired t-test with the Welch correction. If one of the groups did not have a Gaussian distribution, a Mann–Whitney test was performed. Copy number analyses were performed using R (http://www.r-project.org).
Power calculations
In the design phase of the study, we used published association results between higher copy number for β–defensin genes and risk of psoriasis [25] to estimate power for various sample sizes. The effect sizes seen in the study by Hollox et al [25] ranged from 0.25 to 0.4 in mean copy number difference between cases and controls, and the standard deviations of copy numbers in cases and controls ranged between 1.1 and 1.3. Using an unpaired t-test with an significance level of α = 0.05 and assuming equal variances in cases and controls, the power estimates based on these effect sizes ranged from 0.58 (true difference 0.25, standard deviation 1.3) to 0.98 (true difference 0.4, standard deviation 1.1) for 193 cases and 382 controls (see Figure 1 for power curves). Using the largest observed standard deviations in our study (1.17 in cases), the power was 0.67 for a true effect size of 0.25, and for an effect size of 0.40 the power was 0.97. Power calculations were performed using R.
True difference in mean copy number is shown on the x-axis, and the power to detect such differences is shown on the y-axis. Curves are shown for three different copy number standard deviations: 1.1 (solid), 1.2 (dashed), and 1.3 (dotted).
Results
A total of 1,674 patients ≤20 years were diagnosed with SAB in Denmark from 1990 to 2006. Of these patients, 464 had CA-SAB, 882 had HA-SAB, and for 328 patients the origin of infection remained unknown. Among the CA-SAB pediatric cases, 193 patients met inclusion criteria being born after 1982 and had an available blood sample in The Danish Newborn Screening Biobank for DNA extraction. These 193 patients constitute the cases for the present study.
Next, we identified control subjects. A total of two controls per sample were selected for the study, but four had to be excluded due to lack of DNA in the samples resulting in 382 controls. The cases included 130 males (67%) and 63 females (33%). Among the controls 229 (60%) were males and 153 (40%) were females. There was no significant difference in gender distribution between cases and controls (χ2 = 2.69, P = 0.10). The gender distribution among the included cases and controls was also similar to HA-SAB in patients ≤20 years of age and to SAB patients in general.
Distribution of spa types and antibiogram
All isolates were susceptible to methicilin. The diversity of the S. aureus isolates was determined by spa typing to elucidate if the distribution was biased, e.g. whether specific types of S. aureus were overrepresented in CA-SAB. We did, however, not find any such bias as the distribution of phage groups were similar in both the selected cases and other SAB patient cases both within the same period and age group as well as among the total number of cases retrieved from the SAB register (Table 1). A total of 113 different spa types were found among 192 isolates that were assigned to 18 known CC groups. The remaining 31 isolates were either missing (n = 7) or had atypical spa repeats that could not be assigned. The most prevalent CC groups were: CC45 (26.3%), CC30 (20.4%), CC15 (12.9%), CC509 (6.5%), CC121 (6.5%), and CC8 (4.8%).
DEFB1 SNP analysis
Patients and controls were genotyped for three DEFB1 promoter SNPs, rs11362 [−20G/A]; rs1800972 [−44 C/G] and rs1799946 [−52 G/A]. Results of the genotyping are given in Table 2. The minor allele frequencies of the three SNPs did not differ between SAB cases and controls (P>0.05 for all three SNPs). Analyses with dominant, recessive and full genotype models as well as haplotype analyses also did not show any differences between cases and controls (data not shown).
DEFB4/DEFB103 Copy number determination
Gene copy number for DEFB4/DEFB103 varied from 2 to 9 in both CA-SAB cases and controls with comparable frequency distributions for the two groups (Figure 2). Mean copy numbers (and standard deviations) for cases and controls were 4.83 (1.17) and 4.92 (1.12). There was no significant difference in copy number between cases and controls (P = 0.33, t-test). However, a Shapiro-Wilks test showed deviation from normality in the copy number distribution for the controls (P = 0.002). We therefore also applied a Mann-Whitney test, but again we found no significant difference in copy number between cases and controls (P = 0.37).
Discussion
Using a large, Caucasian Danish, and clinically well-described bacteremia registry of children with CA-SAB, we found no association between the DEFB1 SNPs (rs11362 [−20G/A]; rs1800972 [−44 C/G] and rs1799946 [−52 G/A]) or the DEFB4/DEFB103 GCN and susceptibility to CA-SAB [23]. The CA-SAB isolates in our registry were of diverse types indicating that no specific type was dominant among CA-SABs.
We have established a large well-characterized nested CA-SAB case-control study (193 cases and 382 controls), which has allowed us to study genetic association in an ethnic homogenous population. This is among the largest CA-SAB cohorts for genetic association studies, and our study had good statistical power to detect case-control differences in mean copy number larger than 0.25. It is, however, possible that weaker genetic associations exist that will not be detected with a cohort of this size. The GCN was determined by P-PRT, which we previously have shown to be a reliable method [23], but like other GCN assays is associated with errors. However, in the present study the relative differences between cases and controls rather than exact GCN are the principal aim where we assume similar error rates in cases and controls.
Defensin genes have been the focus of the present susceptibility study as the functional consequences of such mutations appear to be associated with reduced expression and presumably reduced activity of the resulting peptides. A promoter polymorphism (rs11362) of DEFB1 has previously been shown to be associated with sepsis [15]. However, we did not see any association between either the DEFB1 polymorphism or other closely positioned polymorphisms and SAB in a study population comparable in size (193 vs. 211) to that of Chen and co-workers [15]. The genetic association previously found could thus be the result of infection by other types of bacteria than SAB or the difference could be due to ethnic differences between the two studies or the fact that our study is focused only on S. aureus bacteremia cases in contrast to the broader study of Chen and co-workers who did not distinguish between different bacterial infections.
It is intriguing that we now have the possibility to address the issue of genetic association to a specific type of bacterial infection. Addressing other candidate genes based on information from murine studies and in silico mapping [7] may give further insight in the genetic susceptibility to S. aureus bacteremia.
Acknowledgments
We would like to thank John Armour (University of Nottingham, UK) for providing us with controls for DEFB103 copy number determination. We also appreciate the expert technical assistance of Britta Hansen and Elvira Chapka. We would also like to thank Lis Vestergaard, Bent Nørgaard-Pedersen and David Hougaard from the Danish Newborn Screening Biobank for providing the filter paper blood samples.
Danish SAB Study Group Consortium: Andreas Petersen1, Allan Garlik Jensen3, Henrik Westh4, Henrik Carl Schønheyder5§, Niels Frimodt-Møller6, Peter Skinhøj7, Thomas Benfield8
§ Lead author for Danish SAB Study Group Consortium
1 Department of Microbiological Surveillance and Research, Statens Serum Institut, Denmark
2 Pfizer, Ballerup, Denmark
3 Department of Clinical Microbiology, Hvidovre University Hospital, Denmark
4 Departments of Clinical Microbiology, Aalborg Hospital, Aarhus University Hospital, Denmark
5 Department of Microbiology, Hvidovre University Hospital
6Department of Orthopaedics and Internal Medicine, Copenhagen University Hospital, Rigshospitalet, Denmark
7Department of Infectious Diseases, Hvidovre University Hospital, Denmark
Author Contributions
Conceived and designed the experiments: PF ARL RLS MS VGF PSA. Performed the experiments: PF ARL MS CEJ PSA. Analyzed the data: PF BF VF RLS ARL MS PSA. Contributed reagents/materials/analysis tools: PF CEJ PSA. Wrote the paper: PF VF RLS ARL PSA. Contributed with sample collection and data analysis of bacteremia cases: CEJ Danish SAB Study Group Consortium.
References
- 1.
Skov R, Larsen A, Benfield T, Schønheyder H, Westh H (2008) Annual Staphylococcus aureus bacteremia report 2008: 1–12 Statens Serum Institut Website. Available: http://www.ssi.dk/Forskning/Infektioner/Systemiske%20infektioner/~/media/Indhold/DK%20-%20dansk/Forskning/Infektioner/SAB%202008.ashx Accessed 2012 Jan 25.
- 2. Davila S, Wright VJ, Khor CC, Sim KS, Binder A, et al. (2010) Genome-wide association study identifies variants in the CFH region associated with host susceptibility to meningococcal disease. Nat Genet 42: 772–776.
- 3. Thye T, Vannberg FO, Wong SH, Owusu-Dabo E, Osei I, et al. (2010) Genome-wide association analyses identifies a susceptibility locus for tuberculosis on chromosome 18q11.2. Nat Genet 42: 739–741.
- 4. Zhang F-R, Huang W, Chen S-M, Sun L-D, Liu H, et al. (2009) Genomewide association study of leprosy. medicine N Engl J Med 361: 2609–2618.
- 5. Zhang F, Liu H, Chen S, Low H, Sun L, et al. (2011) Identification of two new loci at IL23R and RAB32 that influence susceptibility to leprosy. Nat Genet 43: 1247–1251.
- 6. Hill PC, Birch M, Chambers S, Drinkovic D, Ellis-Pegler RB, et al. (2001) Prospective study of 424 cases of Staphylococcus aureus bacteraemia: determination of factors affecting incidence and mortality. Int Med J 31: 97–103.
- 7. Ahn S-H, Deshmukh H, Johnson N, Cowell LG, Rude TH, et al. (2010) Two genes on A/J chromosome 18 are associated with susceptibility to Staphylococcus aureus infection by combined microarray and QTL analyses. PLoS Pathog 6: e1001088.
- 8. Ruimy R, Angebault C, Djossou F, Dupont C, Epelboin L, et al. (2010) Are host genetics the predominant determinant of persistent nasal Staphylococcus aureus carriage in humans? J Infect Dis 202: 924–934.
- 9. Singh , Jia H, Wiles K, Hesselberth J, Liu L, et al. (1998) Production of beta-defensins by human airway epithelia. Proc Natl Acad Sci USA 95: 14961–14966.
- 10. Zhao C, Wang I, Lehrer RI (1996) Widespread expression of beta-defensin hBD-1 in human secretory glands and epithelial cells. FEBS Lett 396: 319–322.
- 11. Valore EV, Park CH, Quayle AJ, Wiles KR, Mccray PB, et al. (1998) Human beta-Defensin-1: An Antimicrobial Peptide of Urogenital Tissues. J Clin Invest 101: 1633–1642.
- 12. Morrison G, Kilanowski F, Davidson D, Dorin J (2002) Characterization of the Mouse Beta Defensin 1, Defb1, Mutant Mouse Model. J Bacteriol 70: 3053–3060. 13.
- 13. Hollox EJ, Armour JaL, Barber JCK (2003) Extensive normal copy number variation of a beta-defensin antimicrobial-gene cluster. Am J Hum Genet 73: 591–600.
- 14. Harder J, Meyer-Hoffert U, Wehkamp K, Schwichtenberg L, Schröder J-M (2004) Differential Gene Induction of Human b-Defensins (hBD-1, −2, −3, and −4) in Keratinocytes Is Inhibited by Retinoic Acid. J Invest Dermatol 123: 522–529.
- 15. Chen Q-X, Lv C, Huang L-X, Cheng B-L, Xie G-H, et al. (2007) Genomic variations within DEFB1 are associated with the susceptibility to and the fatal outcome of severe sepsis in Chinese Han population. Genes Immun 8: 439–443.
- 16. Harder J, Bartels J, Christophers E, Schroder JM (2001) Isolation and characterization of human beta -defensin-3, a novel human inducible peptide antibiotic. J Biol Chem 276: 5707–5713.
- 17. Christophers E, Henseler T (1987) Contrasting disease patterns in psoriasis and atopic dermatitis. Arch Dermatol Res 279:: SupplS48–S51.
- 18. Nørgaard-Pedersen B, Hougaard DM (2007) Storage policies and use of the Danish Newborn Screening Biobank. Inherit Metabol Dis 30: 530–536.
- 19. Shopsin B, Gomez M, Montgomery SO, Smith DH, Waddington M, et al. (1999) Evaluation of protein A gene polymorphic region DNA sequencing for typing of Staphylococcus aureus strains. J Clin Microbiol 37: 3556–3563.
- 20. Baker MW, Grossman WJ, Laessig RH, Hoffman GL, Brokopp CD, et al. (2009) Development of a routine newborn screening protocol for severe combined immunodeficiency. J Allergy Clin Immun 124: 522–527.
- 21. Armour JaL, Palla R, Zeeuwen PLJM, den Heijer M, Schalkwijk J, et al. (2007) Accurate, high-throughput typing of copy number variation using paralogue ratios from dispersed repeats. Nucleic Acids Res 35: e19.
- 22. Aldhous MC, Abu Bakar S, Prescott NJ, Palla R, Soo K, et al. (2010) Measurement methods and accuracy in copy number variation: failure to replicate associations of beta-defensin copy number with Crohn's disease. Hum Mol Genet 19: 4930–4938.
- 23. Fode P, Jespersgaard C, Hardwick RJ, Bogle H, Theisen M, et al. (2011) Determination of beta-defensin genomic copy number in different populations: a comparison of three methods. PLOS one 6: e16768.
- 24. Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MaR, et al. (2007) PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet 81: 559–575.
- 25. Hollox EJ, Huffmeier U, Zeeuwen PLJM, Palla R, Rodijk-olthuis D, et al. (2008) Psoriasis is associated with increased beta-defensin genomic copy number. Nat Genet 40: 23–25.