26 Oct 2016: Streata I, Weiner J 3rd, Iannaccone M, McEwen G, Ciontea MS, et al. (2016) Correction: The CARD9 Polymorphisms rs4077515, rs10870077 and rs10781499 Are Uncoupled from Susceptibility to and Severity of Pulmonary Tuberculosis. PLOS ONE 11(10): e0165853. https://doi.org/10.1371/journal.pone.0165853 View correction
Genetic variants in the CARD9 gene predispose to inflammatory disorders and chronic infectious diseases. Tuberculosis (TB), a chronic infectious disease affecting the lung, is lethal in Card9-deficient mice. We hypothesized that polymorphisms in the CARD9 gene influence TB progression and disease-associated lung damage in humans. We tested genotype distributions of the CARD9 polymorphisms rs4077515, rs10781499 and rs10870077 in TB patients and healthy subjects in a Caucasian cohort. SNPs were in linkage disequilibrium and none of the haplotypes was significantly enriched in the TB group. We determined total and differential leukocyte count, erythrocyte sedimentation rate and plasma abundance of cytokines and chemokines as markers for systemic inflammation and scored chest X-rays to assess lung involvement in TB subjects. Most disease parameters segregated independently of the CARD9 haplotypes. In contrast to multifactorial chronic inflammation, selected genetic variants in the CARD9 gene leave host responses apparently unaffected in TB, at least in the population analyzed here.
Citation: Streata I, Weiner J 3rd, Iannaconne M, McEwen G, Ciontea MS, Olaru M, et al. (2016) The CARD9 Polymorphisms rs4077515, rs10870077 and rs10781499 Are Uncoupled from Susceptibility to and Severity of Pulmonary Tuberculosis. PLoS ONE 11(9): e0163662. https://doi.org/10.1371/journal.pone.0163662
Editor: Selvakumar Subbian, Rutgers Biomedical and Health Sciences, UNITED STATES
Received: June 14, 2016; Accepted: September 12, 2016; Published: September 29, 2016
Copyright: © 2016 Streata 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.
Data Availability: All data are contained within the paper.
Funding: This work was supported by intramural funding from Max Planck Society. IS was supported by an EFIS/IL short-term fellowship. MSC, MO, and M. Ioana received funding from the research grant “Vitamin D deficiency and immunogenetic factors in relation to Tuberculosis (RID-TB)” through UEFISCDI—Young Research Teams, PN-II-RU-TE, 217EU/2013 and 163/01/10/2015. This work was also supported by the EU Seventh Framework Programme (FP7/2007-2013) through grant 305279. 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.
Caspase recruitment domain-containing protein 9 (CARD9) is a cytosolic adaptor abundant in myeloid cells which is essential for recognition of microbes [1, 2]. In experimental animals CARD9 protects against infection with the opportunistic fungus Candida albicans  and intracellular bacteria, such as Listeria monocytogenes , and is essential for control of tuberculosis (TB). During infection with the etiological agent of TB, Mycobacterium tuberculosis (Mtb), the adaptor integrates signals from various pattern recognition receptors [4, 5] and fine-tunes concentrations of pro-inflammatory (TNF-α; IL-6; IL-1) and regulatory (IL-10) cytokines. In line with its general propensity to regulate abundance of various TNF and IL-1 family members [6, 7], selected CARD9 gene polymorphisms predispose to inflammatory diseases in humans [8–10]. Polymorphisms in CARD9 have thus far not been studied in the context of human TB, a contagious disease causing extensive morbidity and mortality worldwide , with complex genetic components . We investigated several CARD9 single nucleotide polymorphisms (SNPs) and defined how this candidate gene influences infection phenotypes, notably progression to active TB, as well as the extent of pulmonary inflammation during active disease.
Materials and Methods
NCBI SNP database was used to select the CARD9 SNPs previously associated with chronic inflammation (rs4077515 [9, 10, 13, 14]; rs10870077 [15, 16]; rs10781499 [15, 17, 18]). rs4077515, a nonsynonymous SNP located in codon 2 of CARD9 gene, leads to an amino acid substitution at position 12 (p.Ser12Asn) and appears”benign” for protein function . The rs10781499 is a synonymous SNP located in the third exon of the gene with undefined functional effects , while rs10870077 is an intronic variant that might modify CARD9 expression . DNA was isolated and purified from blood samples collected on EDTA from 716 Caucasian subjects (342 patients diagnosed with active pulmonary TB and 374 healthy controls). The Ethics Committee of University of Medicine and Pharmacy of Craiova, Romania, approved this study (Ethical approval No.03/07.02.2011) and all subjects provided written informed consent. All subjects were HIV- and type II diabetes-negative. Clinical characteristics of subjects included in the study, both TB and healthy controls, are presented in Table 1. Demographic data, total and differential leukocyte counts, erythrocyte sedimentation rate (ESR) and body-mass index (BMI) were recorded at TB diagnosis, prior to initiation of the chemotherapy. The algorithm employed to establish the TB diagnosis followed the recommendations of the WHO . Patients with clinical signs indicative of active disease, TB suggestive chest radiographic findings and Mtb-positive sputum samples were classified as TB cases.
Genomic DNA was prepared from peripheral blood leukocytes according to the standard protocol of Wizard® Genomic DNA Purification Kit (Promega, Madison, WI). Genotyping for the CARD9 polymorphisms was performed with the predesigned TaqMan single nucleotide assays (C_25956930_20, C_32127979_20 and C_25957125_10) on the ViiA7 Real-Time Polymerase Chain Reaction system (all from Applied Biosystems, Foster City, CA).
Plasma cytokine and chemokine measurements
Cytokine and chemokine concentrations were measured in all available plasma samples. Specimens were collected prior to chemotherapy against TB, upon confirmation of active disease. Abundances of immune mediators were determine by multiplex magnetic bead immunoassay (Luminex assay, R&D Systems), with a Bio Plex microbead analyzer (Bio-Plex® 200 Systems, Bio-Rad Laboratories, Inc.) according to the manufacturer’s protocol.
Statistical analyses were performed with the R programming language, version 3.1.1. Association between haplotypes and disease status (TB vs. non TB) was tested using Fisher's exact test. For machine learning, we used the random forest algorithm as implemented in the randomForest R package . Associations between haplotypes and continuous variables (quantitative locus trait analysis) were tested using analysis of variance. Binomial logistic regression was used to test the effect of continuous variables on categorical outcomes (BMI and age as predictors for the TB vs. control outcome). Local linear polynomial regression was performed using the loess function in R. PCA was performed with the prcomp function in R and visualized with the pca3d package . Multiple correspondence analysis for calculation of the disease severity score was carried out with the mca function from the MASS R package . For post-hoc ANOVA tests, the Tukey Honest Significance Differences test was used.
Results and Discussion
Recent genome-wide association studies and meta-analyses indicate that selected CARD9 SNPs result in predisposition to chronic inflammatory disorders [8–10, 25]. Based on our previous study in murine TB  and considering that TB is characterized by non-resolved inflammation, we hypothesized that similar genetic variants could influence progression of latent to active TB. To test our hypothesis, genotype distributions of three CARD9 polymorphisms, -201C/T (rs4077515), -311G/C (rs10870077), and -126C/T (rs10781499) were tested in a Caucasian cohort of Romanian ethnicity. Vaccination is mandatory in this TB-endemic region and thus we anticipated that subjects enrolled in our study had been immunized as infants with the vaccine M. bovis Bacille Calmette-Guérin (BCG). The analyzed SNPs showed linkage disequilibrium and haplotypes were equally distributed in both TB and healthy groups (Table 2). Our findings are in line with reports nullifying a link between CARD9 rs4077515 genotypes and susceptibility to candidemia and recurrent vulvovaginal candidiasis [13, 19]. However, a lack of association between CARD9 variants and active TB diverges from associations reported for several chronic inflammatory diseases, including Crohn’s disease [8, 10, 16] and ankylosing spondylitis . Despite the chronic feature of adult pulmonary TB, preexisting anti-mycobacterial immunity in the population studied could mask effects of CARD9 polymorphisms on TB progression.
We further investigated whether distinct CARD9 variants influence inflammatory responses in active TB. Plasma abundances of TB-relevant cytokines and chemokines were determined in a sub-cohort of TB cases. Individuals carrying the TT.GG.AA haplotype produced significantly lower levels of IFN-ɣ, CCL3, and CCL5 compared with the other two major haplotypes (Fig 1). IFN-ɣ is critical for defense against TB , yet high concentrations could indicate more severe pathology . Because the CC chemokines orchestrate the dynamics of leukocytes, and reduced abundance reflects limited inflammation, we therefore expected an effect of the genotypes on clinical manifestation of TB. This presumption was not supported by principal component analysis (PCA) integrating clinical covariates (Fig 2). The clinical covariates included BMI and parameters allowing evaluation of emergency hematopoiesis, such as whole blood cell (WBC) counts and frequencies of neutrophils and lymphocytes, as well as ESR, a marker of systemic inflammation. No correlation between the major CARD9 haplotypes and principal components was observed, thus revealing similar systemic inflammation irrespective of the gene variant. These findings could be due to variable expression of the CARD9 gene in distinct haplotypes. Alternatively, tested SNPs regulate expression quantitative trait loci (eQTL) in trans. A cis eQTL effect of the tested SNPs is unlikely, as genes neighboring CARD9 are not directly associated with immune functions.
Abundances (pg/ml) of selected soluble immune mediators for the three major haplotypes (rs4077515, rs10870077, rs10781499) analyzed. Each point corresponds to a single sample. Horizontal bars show the p-value in post-hoc Tukey Honest Significance Differences test, calculated after an initial ANOVA on log-transformed data.
PCA of samples from TB patients applied to clinical covariates (age, BMI, WBC, neutrophils, lymphocytes and ESR) and colored by haplotype.
In order to assess how distinct CARD9 SNPs affect disease severity we first elaborated a severity score (RS1). RS1 served as a disease proxy (Fig 3A) and was built on sputum microbiology and chest X-ray scores (Table 3). RS1 showed a direct correlation with leukocyte count, including WB and neutrophil counts, and ESR values, and furthermore it was inversely correlated with the BMI (Fig 3B). Indeed, neutrophilia represents a marker of poor prognosis in TB patients  and heightened neutrophil frequencies correlate with severity of pulmonary TB . These observations support our newly defined disease proxy parameter. However, patients carrying various CARD9 haplotypes presented similar disease severity scores (Fig 3C). Although Card9 mutant mice  develop severe inflammation during TB, in the case of humans, inflammatory mediators and disease severity did not differ profoundly in patients carrying distinct CARD9 variants. The disparity between mouse and man is unlikely to rely on the variability of outcomes measured. In both mice and humans we measured several inflammatory parameters to monitor systemic (sera/plasma levels of cytokines) and lung inflammation. Indeed, in humans we employed RS1 to overcome inaccessibility of the lung tissue and assess the lung involvement whereas in mice we evaluated lung changes by histopathologic examination of the tissue. RS1 was built on lung radiographic changes, correlated very well with systemic inflammation markers, which increase upon exacerbated lung damage , and thus emerged as a valuable surrogate of pathology, as used in mice.
(a) Disease severity score RS1 calculated using multifactor correspondence analysis (MCA) based on the clinical scores obtained from parameters monitored during sputum microscopy and chest X-ray; (b) Local polynomial regression on four clinical covariates (BMI, WBC, neutrophils and ESR) and severity score RS1 derived from MCA for samples from TB patients. Each point corresponds to a single sample. Light colored area, 95% prediction confidence interval; dark colored area, 95% estimation confidence interval; (c) TB severity score among patients carrying various CARD9 haplotypes.
The phenotype reported in mice largely relies on expression of Card9 in neutrophils for IL-10 release, whereas human neutrophils are unable to produce this regulatory cytokine due to an inactive chromatin configuration at the IL10 locus . This difference could explain discrepancies observed between mice and humans regarding CARD9 control of inflammation. Moreover, the observed discrepancy between Card9 knock-out mice and severity of TB in humans can rely on loss-of-function in the murine system, while in humans the SNPs could result in lower abundance of or modifications in the CARD9 protein structure leading to signaling defects. We did not determine the expression level of the CARD9 gene or abundances of the encoded protein in blood specimens of variable haplotypes. Of note, the CARD9 gene is under significant cis regulation  and presumably single SNPs minimally affect protein structure.
In this study we provide evidence that CARD9 genetic variants in rs4077515, rs10781499 and rs10870077 do not affect progression to active TB in a cohort from a TB-endemic region with an ongoing BCG vaccination program. In mice, vaccination has been found to protect the Card9-deficient animals against lethality due to TB . Genetic studies in BCG-naive humans are required before conclusions can be made on the development of active TB, irrespective of the CARD9 haplotypes. Moreover, our study was undertaken in a single cohort. Investigations in additional populations and larger cohorts will need to overcome such limitations. We have guided our study by reports centered on chronic inflammation, and tested selected SNPs. As the effects of additional CARD9 SNPs on progression to active TB cannot be excluded, a fine-mapping of the CARD9 gene could help elucidate whether other gene variants imprint on susceptibility to TB in humans.
While the selected CARD9 SNPs predispose to chronic inflammation in multifactorial inflammatory diseases [10, 14, 25, 31], their roles in human infections driven by specific microbes, such as Mtb, as shown here, and Candida spp. [13, 19], appears limited. Nonetheless, more than 15 missense and nonsense CARD9 point mutations (Q295X; Q289X; R70W; R35Q or R101L are some of the rare variants identified) have been associated with development of fungal diseases [32–36]. Such mutations result either in synthesis of a nonfunctional protein or lack of the CARD9 protein. In contrast to loss-of-function mutations, CARD9 SNPs including rs4077515, leave cytokine production unchanged in leukocytes stimulated with fungi. The SNP thus minimally affects the function of the adaptor protein and is uncoupled from susceptibility to candidemia . Unlike previous reports in inflammatory polyfactorial diseases, selected SNPs do not affect the course of TB or fungal infection. Additional studies investigating supplementary CARD9 SNPs and effects of CARD9 variants in unrelated infections will permit insights into divergent roles of this adaptor molecule in chronic inflammation of microbial and non-microbial origin.
- Conceptualization: AD.
- Formal analysis: IS JW M. Iannaconne GM.
- Funding acquisition: AD IS M. Ioana SHEK.
- Investigation: IS JW M. Iannaconne MSC MO RC M. Ioana.
- Supervision: AD M. Ioana SHEK.
- Writing – original draft: IS JW AD SHEK.
- Writing – review & editing: M. Iannaconne GM MSC MO RC M. Ioana.
- 1. Gross O, Gewies A, Finger K, Schafer M, Sparwasser T, Peschel C, et al. Card9 controls a non-TLR signalling pathway for innate anti-fungal immunity. Nature. 2006;442(7103):651–6. Epub 2006/07/25. pmid:16862125.
- 2. Roth S, Ruland J. Caspase recruitment domain-containing protein 9 signaling in innate immunity and inflammation. Trends Immunol. 2013;34(6):243–50. Epub 2013/03/26. pmid:23523010.
- 3. Hsu YM, Zhang Y, You Y, Wang D, Li H, Duramad O, et al. The adaptor protein CARD9 is required for innate immune responses to intracellular pathogens. Nat Immunol. 2007;8(2):198–205. Epub 2006/12/26. pmid:17187069.
- 4. Dorhoi A, Desel C, Yeremeev V, Pradl L, Brinkmann V, Mollenkopf HJ, et al. The adaptor molecule CARD9 is essential for tuberculosis control. J Exp Med. 2010;207(4):777–92. Epub 2010/03/31. pmid:20351059; PubMed Central PMCID: PMCPmc2856020.
- 5. Wilson GJ, Marakalala MJ, Hoving JC, van Laarhoven A, Drummond RA, Kerscher B, et al. The C-type lectin receptor CLECSF8/CLEC4D is a key component of anti-mycobacterial immunity. Cell Host Microbe. 2015;17(2):252–9. Epub 2015/02/13. pmid:25674984; PubMed Central PMCID: PMCPmc4334100.
- 6. Poeck H, Bscheider M, Gross O, Finger K, Roth S, Rebsamen M, et al. Recognition of RNA virus by RIG-I results in activation of CARD9 and inflammasome signaling for interleukin 1 beta production. Nat Immunol. 2010;11(1):63–9. Epub 2009/11/17. pmid:19915568.
- 7. Roth S, Rottach A, Lotz-Havla AS, Laux V, Muschaweckh A, Gersting SW, et al. Rad50-CARD9 interactions link cytosolic DNA sensing to IL-1beta production. Nat Immunol. 2014;15(6):538–45. Epub 2014/04/30. pmid:24777530; PubMed Central PMCID: PMCPmc4309842.
- 8. Franke A, McGovern DP, Barrett JC, Wang K, Radford-Smith GL, Ahmad T, et al. Genome-wide meta-analysis increases to 71 the number of confirmed Crohn's disease susceptibility loci. Nat Genet. 2010;42(12):1118–25. Epub 2010/11/26. pmid:21102463; PubMed Central PMCID: PMCPmc3299551.
- 9. McGovern DP, Gardet A, Torkvist L, Goyette P, Essers J, Taylor KD, et al. Genome-wide association identifies multiple ulcerative colitis susceptibility loci. Nat Genet. 2010;42(4):332–7. Epub 2010/03/17. pmid:20228799; PubMed Central PMCID: PMCPmc3087600.
- 10. Rivas MA, Beaudoin M, Gardet A, Stevens C, Sharma Y, Zhang CK, et al. Deep resequencing of GWAS loci identifies independent rare variants associated with inflammatory bowel disease. Nat Genet. 2011;43(11):1066–73. Epub 2011/10/11. pmid:21983784; PubMed Central PMCID: PMCPmc3378381.
- 11. Zumla A, Raviglione M, Hafner R, von Reyn CF. Tuberculosis. N Engl J Med. 2013;368(8):745–55. Epub 2013/02/22. pmid:23425167.
- 12. Thye T, Meyer CG. Human genetic variability and susceptibility to pulmonary TB. European Respiratory Monograph 58: Tuberculosis. 2014;58:38.
- 13. Rosentul DC, Delsing CE, Jaeger M, Plantinga TS, Oosting M, Costantini I, et al. Gene polymorphisms in pattern recognition receptors and susceptibility to idiopathic recurrent vulvovaginal candidiasis. Front Microbiol. 2014;5:483. Epub 2014/10/09. pmid:25295030; PubMed Central PMCID: PMCPmc4172055.
- 14. Janse M, Lamberts LE, Franke L, Raychaudhuri S, Ellinghaus E, Muri Boberg K, et al. Three ulcerative colitis susceptibility loci are associated with primary sclerosing cholangitis and indicate a role for IL2, REL, and CARD9. Hepatology. 2011;53(6):1977–85. Epub 2011/03/23. pmid:21425313; PubMed Central PMCID: PMCPmc3121050.
- 15. Pointon JJ, Harvey D, Karaderi T, Appleton LH, Farrar C, Stone MA, et al. Elucidating the chromosome 9 association with AS; CARD9 is a candidate gene. Genes Immun. 2010;11(6):490–6. Epub 2010/05/14. pmid:20463747; PubMed Central PMCID: PMCPmc2933507.
- 16. Zhernakova A, Festen EM, Franke L, Trynka G, van Diemen CC, Monsuur AJ, et al. Genetic analysis of innate immunity in Crohn's disease and ulcerative colitis identifies two susceptibility loci harboring CARD9 and IL18RAP. Am J Hum Genet. 2008;82(5):1202–10. Epub 2008/04/29. pmid:18439550; PubMed Central PMCID: PMCPmc2427314.
- 17. Jostins L, Ripke S, Weersma RK, Duerr RH, McGovern DP, Hui KY, et al. Host-microbe interactions have shaped the genetic architecture of inflammatory bowel disease. Nature. 2012;491(7422):119–24. Epub 2012/11/07. pmid:23128233; PubMed Central PMCID: PMCPmc3491803.
- 18. Anderson CA, Boucher G, Lees CW, Franke A, D'Amato M, Taylor KD, et al. Meta-analysis identifies 29 additional ulcerative colitis risk loci, increasing the number of confirmed associations to 47. Nat Genet. 2011;43(3):246–52. Epub 2011/02/08. pmid:21297633; PubMed Central PMCID: PMCPmc3084597.
- 19. Rosentul DC, Plantinga TS, Oosting M, Scott WK, Velez Edwards DR, Smith PB, et al. Genetic variation in the dectin-1/CARD9 recognition pathway and susceptibility to candidemia. J Infect Dis. 2011;204(7):1138–45. Epub 2011/09/02. pmid:21881131; PubMed Central PMCID: PMCPmc3164426.
- 20. NCBI. SNP linked to Gene (geneID:64170) 2016. Available from: http://www.ncbi.nlm.nih.gov/SNP/snp_ref.cgi?locusId=64170.
- 21. Organization WH. Global tuberculosis report 2015: World Health Organization; 2015.
- 22. Liaw A, Wiener M. Classification and regression by randomForest. R news. 2002;2(3):18–22.
- 23. Weiner J. pca3d: Three Dimensional PCA Plots. R package version 0.8. 2015.
- 24. Venables WN, Ripley BD. Modern Applied Statistics with S.. Fourth Edition. ed. New York.: Springer,; 2002.
- 25. Beaudoin M, Goyette P, Boucher G, Lo KS, Rivas MA, Stevens C, et al. Deep resfuencing of GWAS loci identifies rare variants in CARD9, IL23R and RNF186 that are associated with ulcerative colitis. PLoS Genet. 2013;9(9):e1003723. Epub 2013/09/27. pmid:24068945; PubMed Central PMCID: PMCPmc3772057.
- 26. Flynn JL, Chan J. Immunology of tuberculosis. Annu Rev Immunol. 2001;19:93–129. Epub 2001/03/13. pmid:11244032.
- 27. Hasan Z, Jamil B, Khan J, Ali R, Khan MA, Nasir N, et al. Relationship between circulating levels of IFN-gamma, IL-10, CXCL9 and CCL2 in pulmonary and extrapulmonary tuberculosis is dependent on disease severity. Scand J Immunol. 2009;69(3):259–67. Epub 2009/03/14. pmid:19281538.
- 28. Abakay O, Abakay A, Sen HS, Tanrikulu AC. The relationship between inflammatory marker levels and pulmonary tuberculosis severity. Inflammation. 2015;38(2):691–6. Epub 2014/07/17. pmid:25028104.
- 29. Gopal R, Monin L, Torres D, Slight S, Mehra S, McKenna KC, et al. S100A8/A9 proteins mediate neutrophilic inflammation and lung pathology during tuberculosis. Am J Respir Crit Care Med. 2013;188(9):1137–46. Epub 2013/09/21. pmid:24047412; PubMed Central PMCID: PMCPmc3863739.
- 30. Tamassia N, Zimmermann M, Castellucci M, Ostuni R, Bruderek K, Schilling B, et al. Cutting edge: An inactive chromatin configuration at the IL-10 locus in human neutrophils. J Immunol. 2013;190(5):1921–5. Epub 2013/01/29. pmid:23355741.
- 31. Burghardt KM, Avinashi V, Kosar C, Xu W, Wales PW, Avitzur Y, et al. A CARD9 polymorphism is associated with decreased likelihood of persistent conjugated hyperbilirubinemia in intestinal failure. PLoS One. 2014;9(1):e85915. Epub 2014/01/28. pmid:24465786; PubMed Central PMCID: PMCPmc3897546.
- 32. Glocker EO, Hennigs A, Nabavi M, Schaffer AA, Woellner C, Salzer U, et al. A homozygous CARD9 mutation in a family with susceptibility to fungal infections. N Engl J Med. 2009;361(18):1727–35. Epub 2009/10/30. pmid:19864672; PubMed Central PMCID: PMCPmc2793117.
- 33. Lanternier F, Mahdaviani SA, Barbati E, Chaussade H, Koumar Y, Levy R, et al. Inherited CARD9 deficiency in otherwise healthy children and adults with Candida species-induced meningoencephalitis, colitis, or both. J Allergy Clin Immunol. 2015;135(6):1558–68.e2. Epub 2015/02/24. pmid:25702837.
- 34. Lanternier F, Barbati E, Meinzer U, Liu L, Pedergnana V, Migaud M, et al. Inherited CARD9 deficiency in 2 unrelated patients with invasive Exophiala infection. J Infect Dis. 2015;211(8):1241–50. Epub 2014/07/25. pmid:25057046; PubMed Central PMCID: PMCPmc4447834.
- 35. Grumach AS, de Queiroz-Telles F, Migaud M, Lanternier F, Filho NR, Palma SM, et al. A homozygous CARD9 mutation in a Brazilian patient with deep dermatophytosis. J Clin Immunol. 2015;35(5):486–90. Epub 2015/06/06. pmid:26044242.
- 36. Drummond RA, Lionakis MS. Mechanistic Insights into the Role of C-Type Lectin Receptor/CARD9 Signaling in Human Antifungal Immunity. Front Cell Infect Microbiol. 2016;6:39. Epub 2016/04/20. pmid:27092298; PubMed Central PMCID: PMCPmc4820464.