You are here

Cocci GWAS Project Summary

Background

Valley Fever is typically an infection of the lungs caused by the fungi Coccidioides immitis and Coccidioides posadasii. The incidence of coccidioidomycosis (CM), or infection with Coccidioides, has dramatically increased over the last 20 years. This is particularly true in the Southwest of the United States, where people often breathe fungal spores that arise from the soil. Reasons for increased infection rates are thought to include population growth and construction in these endemic regions, an increase in the number of people whose immune systems are compromised due to infection or treatment with drugs required for organ transplants, climate change, as well as improved testing practices and greater physician awareness. 

Mild CM most commonly presents itself with flu-like symptoms and rashes which can last weeks to months1. Individuals with compromised immune systems, specifically substantial suppression of the immune cells known as T cells, can develop severe pulmonary and disseminated disease2. Infection that remains localized to the lungs is referred to as pulmonary disease, but when the infection spreads out of the lungs into other parts of the body it represents a more serious condition referred to as a disseminated disease, or disseminated CM. In nature, Coccidioides spp. exists as mold and lives in dust and soil. When the contaminated soil or dust is disturbed by human activity, animals, or weather, the Coccidiodies spores are released into the air. Airborne spores are taken up by breathing and settle in the lungs. Once in the moist and warm environment of the lung, spores transform into spherules, which divide and become filled with smaller spores, called endospores. When the spherules get large enough, they rupture and release these endospores, which can spread and disseminate to surrounding tissue. The cycle then repeats itself as these endospores develop into new spherules3

Different ethnic groups have been described to vary in their susceptibility to developing disseminated CM after initial infection with Coccidioides. For example, evidence suggests that African-American and Filipino patients suffer the disseminated disease at a greater rate than other ethnicities4,5. The suggestion that race plays a role in the clinical expression of the disease is still a source of debate amongst the scientific community and any genetic mechanisms responsible for these differences have yet to be fully elucidated. If our genetic makeup influences our ability to limit the spread of infection, finding which DNA differences cause these variances could provide clues to how the body successfully fights infection, and provide opportunities to boost the body’s ability to do this. Further, if we are able to identify the specific genetic risk factors that correlate with the development of disseminated infection, physicians could perform genetic screenings to identify high-risk patients and provide them with preemptive antifungal therapy prior to developing disseminated disease.

The genome, made up of DNA, contains all of the information needed for us humans to develop and grow. Genome-wide association studies (GWAS) allow us to look for inherited differences that are more common between people who share a particular trait, for example, height or susceptibility to certain diseases, compared to those who do not share the trait. Although some traits and diseases are controlled by a single gene, the majority are influenced by contribution from several, or even many, different genes. To find evidence of genes that contribute to specific traits, GWAS typically compare genome information from large numbers of people who have a particular disease (referred to as “cases”) looking for DNA sequences that are common among these samples, and are different from DNA sequences seen in large numbers of people who lack the trait, but are as much like the cases as possible (referred to as “controls”). The DNA sequence data from each group, cases versus controls, are analyzed to see if there are specific genomic differences that tend to be associated with the disease. 

Methods

Two separate GWAS approaches were taken to look for genetic differences that could be responsible for the observed differences between the different patient populations we are studying. The first method, known as genotyping, scans for differences at a set of positions across the genome, which includes both the genes that encode our proteins and the larger amount of DNA that does not. The second method, known as exome sequencing, allows us to compare the entire sequence of the portion of the genome that codes for proteins.  

For this study, DNA from patients with either pulmonary or disseminated CM were genotyped and exome sequenced to look for DNA differences that are associated with one condition or the other. All patients were at least 18 years old, had no evidence of immunosuppression, and had proven or probable pulmonary coccidioidomycosis according to established diagnostic criteria. Of these patients, a subset demonstrated disseminated disease, i.e., they showed evidence of coccidioidal infection outside of the thorax by biopsy/aspiration, had radiographic imaging, and show positive coccidioidal serology. Our criteria for including patients with the pulmonary disease were that they must not require ongoing antifungal treatment or show evidence of active CM (in skin test positive patients), show no evidence of extrapulmonary dissemination, and have no evidence of ongoing pulmonary infection (pulmonary nodules are accepted) beyond six months from diagnosis.

Patient DNA was purified from blood or from sputum samples by the labs of our collaborators, Drs. George Thompson (UC Davis School of Medicine) and John Galgiani (University of Arizona Health Sciences). Genome-wide association (GWAS) analysis was carried out to look for candidate loci associated with pulmonary versus disseminated disease, taking into account the population structure of the samples. Single nucleotide or insertion/deletion variants were identified from whole-exome sequences (WES) using the Picard/BWA/GATK pipeline. 

Results

Table 1 shows the number of samples analyzed from patients with pulmonary versus disseminated disease, and patient ethnicity, where known. In all, we worked with 507 samples, including 134 samples from patients with disseminated disease and 373 samples from patients with pulmonary disease. Of these, 505 samples were genotyped using the Multi-Ethnic Global Array from Illumina Inc. In addition, we were able to generate whole exome sequence from 498 patient samples. 




Table 1. Pulmonary versus Disseminated Cases of Coccidiomycosis for GWAS, Sorted by Ethnicity 

  Pulmonary Cases Disseminated Cases
Ethnicity    
Asian 8 5
Black/African Amer 16 64
Caucasian/White 40 15
Filipino 0 3
Hispanic/Latino 34 14
Indian 2 1
Mexican Amer 103 9
Pacific Islander 0 1
Samoan 0 3
Vietnamese 1 0
Unknown 169 17
More than one race 0 2
     
Total 373 134
 

No significant associations were detected that differed between samples from patients with pulmonary versus disseminated disease; that is, no particular DNA sequences were found to be significantly enriched in patients with disseminated disease compared to patients with pulmonary disease. The ability to detect genetic association between specific sequences and genetically determined traits, is influenced by several factors, including how many patient samples are available to compare, how many different genes contribute to the trait and how strong their contributions are. When the number of genes is small and the contribution of each gene is great, smaller numbers of patient samples are needed to detect an association. When more genes are involved, or the contribution from each gene is more modest, larger numbers of patient samples must be examined. While we were not able to detect any associated within this study it does not mean that subsequent studies would not find this connection. Our study suggests significantly more samples should be analyzed in further studies.

Whole exome sequences were generated from 498 samples and were aligned to reference sequences to identify positions where the sequences differed from the reference. These data are being analyzed to determine if any variants are associated with pulmonary, versus disseminated, disease. 

References

  1. Jin J. Valley Fever (Coccidioidomycosis). JAMA. 2013;310(22):2470. doi:10.1001/jama.2013.280726
  2. Einstein, H. E., and R. H. Johnson. “Coccidioidomycosis: New Aspects of Epidemiology and Therapy.” Clinical Infectious Diseases, vol. 16, no. 3, 1993, pp. 349–356., doi:10.1093/clind/16.3.349.
  3. “Sources of Valley Fever (Coccidioidomycosis).” Centers for Disease Control and Prevention, Centers for Disease Control and Prevention, 2 Jan. 2019, www.cdc.gov/fungal/diseases/coccidioidomycosis/causes.html.
  4. Rosenstein NE, Emery KW, Werner SB, Kao A, Johnson R, Rogers D, et al. Risk factors for severe pulmonary and disseminated coccidioidomycosis: Kern County, California, 1995-1996. Clin Infect Dis Off Publ Infect Dis Soc Am. 2001;32: 708–715. doi:10.1086/319203
  5. Thompson GR. Pulmonary coccidioidomycosis. Semin Respir Crit Care Med. 2011;32: 754–763. doi:10.1055/s-0031-1295723