The Cystic Fibrosis Microbiome and Its Association with Incident Infections with Mycobacteroides (Mycobacterium) Abscessus

University of Calgary PRISM: University of Calgary's Digital Repository

Graduate Studies The Vault: Electronic Theses and Dissertations

2021-05-13 The Cystic Fibrosis Microbiome and its Association with Incident Infections with Mycobacteroides (Mycobacterium) abscessus

Bharadwaj, Lalit

Bharadwaj, L. (2021). The Cystic Fibrosis Microbiome and its Association with Incident Infections with Mycobacteroides (Mycobacterium) abscessus (Unpublished master's thesis). University of Calgary, Calgary, AB. http://hdl.handle.net/1880/113441 master thesis

University of Calgary graduate students retain copyright ownership and moral rights for their thesis. You may use this material in any way that is permitted by the Copyright Act or through licensing that has been assigned to the document. For uses that are not allowable under copyright legislation or licensing, you are required to seek permission. Downloaded from PRISM: https://prism.ucalgary.ca UNIVERSITY OF CALGARY

The Cystic Fibrosis Microbiome and its Association with Incident Infections with

Mycobacteroides (Mycobacterium) abscessus

Lalit Bharadwaj

A THESIS

SUBMITTED TO THE FACULTY OF GRADUATE STUDIES

IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE

DEGREE OF MASTER OF SCIENCE

GRADUATE PROGRAM IN MICROBIOLOGY AND INFECTIOUS DISEASES

CALGARY, ALBERTA

MAY, 2021

Abstract

Infection with M. abscessus complex (MABC) is increasingly detected within CF populations.

MABC infection has been associated with exaggerated lung function decline and poses significant treatment complexities. We performed a retrospective case-control study of twenty- one patients with MABC infection matching each to two randomly identified age (+/-2 yrs) and gender-matched uninfected controls. Total genomic DNA from sputum was extracted, amplified and Illumina MiSeq paired-end sequencing of the hypervariable V3 region of the 16S rRNA gene was performed. Demographics and dynamic variables of disease were recorded and compared between groups. 174 sputum samples (median 8, IQR (6-12)) from MABC cases and

42 control samples were assessed. The sputum microbiota from patients who would develop

MABC infection in the subsequent two years differed from controls (p=0.038, R2 = 2.5%,

PERMANOVA). In particular, sputum from MABC cases – prior to its identification – had higher alpha-diversity; Shannon diversity (p=0.023), Observed species (p=0.042) and lower P. aeruginosa relative and absolute abundance (p=0.035). We observed significant changes in community structure over time during potent antibacterial therapies, returning to baseline upon their discontinuation. These data suggest that sputum microbiome analysis and P. aeruginosa bioburden should be evaluated in multi-center studies as potential biomarkers to predict MABC infection and treatment response.

Keywords: cystic fibrosis, Non-tuberculous Mycobacterium, pulmonary microbiome, 16S

rRNA sequencing, sputum.

ii Acknowledgements

Several amazing people have been instrumental through the course of my master’s degree who all deserve a lot more gratitude than just an acknowledgement in my paper.

First and foremost, would be my supervisor, Dr. Michael Parkins. Dr. Parkins has not only been absolutely instrumental in guiding me through my research project but has always gone above and beyond in listening and helping me navigate issues in my life outside the lab and my struggles with my mental health. I can say without a doubt that it would not have been possible for me to complete this project without Dr. Parkins’ guidance, understanding, and support and I will forever be thankful.

I would also like to thank my co-supervisor, Dr. Douglas Storey, as well as my committee members, Dr. Michael Surette and Dr. Marie-Claire Arrieta, who have all given me so much through their expertise and guidance through my project. I find myself very lucky, although intimidated sometimes, to have such incredibly brilliant and renowned scientists helping me navigate my project.

I would also like to thank Dr. Lisa Gieg for taking time out of her busy schedule to serve as my examiner.

Members of the Parkins’ lab and the Southern Alberta Adult CF clinic have also been instrumental through my Master’s degree and have been like family to me. My appreciation is utmost for their continual support and guidance. Specifically, Dr. Nicole Acosta, Dr. Alya

Heirali, and Barbara Waddell who have always supported me and have always made themselves available for help. I am also thankful to Conrad Izydorczyk, Julianna Svishcuk,

Chloe Ligier, Jennifer Carpentero, Alexandra Mossman, Delia Lee, Taylor Woo, Bernice

Lee, and Benson Weyant for supporting and helping me through graduate education.

iii I would like to thank my family, who have loved and supported me and allowed for me to further my education through post-secondary; and my friends and teammates, who have given me additional courage and love, especially when my parents were not able to be there in person.

Lastly, I would like to thank the CF patients who have volunteered to participate in these studies and allow us the opportunity to further the scientific field.

iv Dedication

I dedicate this thesis to my parents, Sanjeev and Ratna Bharadwaj, who have faced lifelong adversities to allow me to have the best opportunities.

v Table of Contents

Abstract ...... ii

Acknowledgements ...... iii Dedication ...... v Table of Contents ...... vi List of Tables ...... viii List of Figures and Illustrations ...... ix List of Symbols, Abbreviations and Nomenclature ...... xi

Chapter One: Introduction ...... 1 1.1 An introduction to Cystic Fibrosis ...... 2 1.2 Genetics of Cystic Fibrosis ...... 3 1.3 Cystic Fibrosis Transmembrane Conductance Regulator ...... 4 1.4 Lung Disease Manifestations in CF ...... 6 1.5 Canonical CF Pathogens ...... 8 1.6 The CF Lung Microbiome ...... 11 1.7 Clinical Relevance of the CF Lung Microbiome ...... 13 1.8 Introduction to the Mycobacterium Genus ...... 15 1.8.1 An introduction to Mycobacteroides (mycobacterium) abscessus...... 17 1.9 Hypothesis...... 19 1.9.1 Project Aims ...... 20

Chapter Two: Methods and Materials ...... 21 2.1 Identification of MABC Cohort...... 22 2.2 Identification of Control Patient Cohort ...... 23 2.3 Patient Sputum Sample Identification ...... 24 2.4 Statistical analysis ...... 27 2.5 DNA Extraction...... 30 2.6 16S rRNA Gene Amplification ...... 32 2.7 DNA sequencing and analysis ...... 42 2.8 Quantitative PCR ...... 44 2.9 P. aeruginosa Inhibition Assay of MABC ...... 47

Chapter Three: Results ...... 48 3.1 Case and Control Patient Demographics...... 49 3.2 Pulmonary Function Linear Regression ...... 54 3.3 Lung Microbiome Diversity ...... 57 3.4 Microbiome structure through the natural history of infection ...... 62 3.5 Microbiome Comparison of MABC Case Patients to Controls ...... 66 3.6 Comparing Transient and Persistent Case Patient Sub-cohorts ...... 70 3.7 Quantitative PCR ...... 73 3.7.1 Absolute bacterial bioburden through the natural history of infection ...73 3.7.2 Absolute bacterial bioburden between case and control patients ...... 75 3.7.3 Quantifying MABC DNA in Sputum Samples ...... 77 3.7.4 Quantifying P. aeruginosa DNA in Sputum Samples ...... 82

vi 3.8 P. aeruginosa Inhibition Assay ...... 84

Chapter Four: Discussion...... 86 4.1 MABC Infections are a Major Cause for Concern in CF Patients ...... 87 4.2 Patients with MABC Present with a Lower BMI than Controls ...... 88 4.3 Persistent MABC Patients had Significantly Higher PPI use Compared to Transient Patients or Controls...... 89 4.4 Patients with MABC did not Differ in their Microbiome Diversity with New Infection ...... 90 4.5 Patients with MABC had significantly different microbiomes compared to control patients...... 92

CHAPTER 5: Summary ...... 95 5.2 Future Directions ...... 97 Further testing of the P. aeruginosa inhibition assay could be done with a possible testing with a semi-permeable polycarbonate membrane dividing P. aeruginosa from MABC could allow us to run a more realistic indirect competitive assay which would account for quorum sensing or direct competition for resources and could provide possible reasons to the negative correlation observed between MABC and P. aeruginosa. Additionally, a transcriptomic study involving P. aeruginosa and MABC co-cultures to assess any molecular mechanisms between the interactions...... 97 5.3 Study Summary and Conclusion ...... 98

References ...... 99

vii List of Tables

Table 1: Core members of CF pathogenesis. Data adapted from the 2019 CF Canada annual report...... 9

Table 2: The CF lung microbiome in health and disease and its supporting publication...... 14

Table 3: Case patient sputum samples categorized to natural history of disease stages according to time frame, AFB culture, and AFB smear status...... 26

Table 4. 16S rRNA V3 region forward and reverse barcoded primers...... 34

Table 5: Content configuration of each flask used to test P. aeruginosa inhibition of MABC. Three replicates of each flask was used for the assay...... 47

Table 6A: Baseline clinical demographics collected for case and control patients from the Calgary Adult CF clinic chart records...... 51

Table 6B: Mycobacteriology subspecies and coinfections collected for case and control patients from the Calgary Adult CF clinic chart records...... 52

Table 6C: On-going therapies and antibiotic collected for case and control patients from the Calgary Adult CF clinic chart records...... 53

viii List of Figures and Illustrations

Figure 1: CFTR mutation classes with phenotypic descriptions with common associated genetic mutations...... 5

Figure 2: Progressive CF lung disease starting from a defective CF gene and subsequent CFTR inhibition to the cyclic infection and inflammation phases exacerbating bronchiectasis...... 7

Figure 4: Lung function changes over time did not differ between cases and controls before or after infection...... 55

Figure 5: Lung function changes over time did not differ in transient vs persistent patients before or after infection...... 56

Figure 6: Phyla comparisons for 16S rDNA sequences observed in the microbiome of MABC case patients and their respective controls...... 60

Figure 7: Taxonomic composition of the top 20 ASVs at the genus level through infection natural history and matched controls...... 61

Figure 8: The natural history of the CF sputum microbiome through the course of MABC infection stages...... 65

Figure 9: The microbiome of sputum Pre-infection with MABC differs between cases and control patients...... 68

Figure 10: The microbiome of sputum At infection with MABC differs between cases and control patients...... 69

Figure 11: Pre-infection microbiome comparison between transient and persistent patients...... 71

Figure 12: At infection microbiome comparison between transient and persistent patients...... 72

Figure 13: Total bacterial bioburden stays relatively stable through the natural history of infection with exception of treatment intervention...... 74

Figure 14: Total bacterial bioburden does not change as a function of infection status...... 76

Figure 15: MABC bioburden changes through its Natural History of Infection...... 78

Figure 16: Z-N Smear positivity correlates with absolute abundance of MABC...... 80

Figure 17: A higher bioburden of MABC DNA correlated with a shorter time for MABC cultures to grow...... 81 ix Figure 18: Median P. aeruginosa DNA concentration was higher in control samples compared to at infection case samples...... 83

Figure 19: MABC growth in Middlebrook 7H9 broth infused with P. aeruginosa PA01 and PES supernatant...... 85

List of Symbols, Abbreviations and Nomenclature

Symbol Definition - Minus, Negative ˚C Degrees Celcius + Plus, Positive ABC ATP-Binding Cassette ABPA Allergic Bronchopulmonary Aspergillosis AFB Acid-Fast Bacillus AS Achromobacter species ASV Amplicon Sequence Variants ATP Adenosine Triphosphate ATS American Thoracic Society Bcc Burkholderia cepacia complex BMI Body Mass Index  Chi-squared CACFC Calgary Adult Cystic Fibrosis Clinic cAMP Cyclic Adenosine monophosphate CF Cystic Fibrosis CFLD Cystic Fibrosis related Liver Disease CFRD Cystic Fibrosis Related Diabetes CFTR Cystic Fibrosis Transmembrane Conductance Regulator DADA2 Divisive Amplicon Denoising Algorithm 2 DIOS Distal Intestinal Obstructive Syndrome DNA Deoxyribonucleic Acid FEV Forced Expiratory Volume FVC Forced Vital Capacity G force Gravitational Force GER Gastroesophageal Reflux GTDB Genome Taxonomy Database HI Haemophilus influenzae HIV Human Immunodeficiency Virus IDSA Infectious Diseases Society of America kg Kilograms LABA Long-acting Beta Agonist LB Luria-Bertani m meters MABC Mycobacteroides abscessus complex MAC Mycobacterium avium complex mL Milliliters MRSA Methicillin resistent S. aureus MSSA Methicillin sensitive S. aureus NTM Non-tuberculous Mycobacteria

xi ODI Observed Diversity Index OTU Operational Taxanomic Unit PA Pseudomonas aeruginosa PA01 P. aeruginosa lab strain PA01 PCoA Principle Coordinate Analysis PCR Polymerase Chain Reaction PERMANOVA Permutational Multivariate Analysis of Variance PES Prairie Epidemic Strain pH Potential of Hydrogen PPI Proton Pump Inhibitor QIIME Quantitative Insights into Microbial Ecology qPCR Quantitative Polymerase Chain Reaction rDNA Ribosomal DNA RDP Ribosomal Database Project rRNA Ribosomal RNA SA Staphylococcus aureus SABA Short-acting Beta Agonist SDI Shannon Diversity Index SL1P Short-read Library 16S rRNA gene Sequencing Pipeline SM Stenotrophomonas maltophilia UND Undefined V3 Variable 3 Region of the 16S rRNA gene Z-N Stain Ziel-Neelson Stain µL microlitre  Delta

xii 1

Chapter One: Introduction

1.1 An introduction to Cystic Fibrosis

Cystic fibrosis (CF) is the most common lethal genetic disorder observed in the

Caucasian population although it can affect individuals of all races and ethnicities.1

However, what was once a debilitating disease, causing death within the first few years of

life , has evolved into a manageable disease allowing patients to potentially survive into

their fifties and even beyond.2,3 Dysfunctional CFTR proteins manifest disease in various

organs throughout the body resulting in abnormal symptoms due to the aberrant chloride

transport across epithelial cell membranes.4 These complications are expressed in the

pancreatic, hepatic, gastro-intestinal, reproductive, and pulmonary systems.4,5

Manifestations of symptoms in the pancreatic system includes pancreatic insufficiency,

pancreatitis, and CF- related diabetes (CFRD).6–8 Gastro-intestinal deficiencies could

display distal intestinal obstructive syndrome (DIOS) and rectal prolapse whereas the

hepatic system could manifest CF-related liver disease (CFLD) and gallbladder disease.9–12

Reproductive malfunction sees infertility in more than 95% of men due to defects in the

sperm transportation whereas females are subject to a lesser degree of infertility and are

primarily caused by malnutrition and abnormal cervical mucus.13,14 Impairments to any of

these organs cause debilitating effects on the regular well-being of patients, however, none

more than the pulmonary system.5 Disease manifestation in the pulmonary system typically

involves the upper and lower respiratory tract, and the sinus.15,16 The respiratory tract

creates a hospitable environment for repeated pathogenic bacterial infections due to the

abnormally thick and difficult to clear mucosal layer which creates a cycle of inflammation,

bronchiectasis, and reinfection .17

1.2 Genetics of Cystic Fibrosis

CF is an autosomal recessive disorder caused by a mutation in the CFTR gene resulting

in a dysfunctional CFTR protein.18 The CFTR gene is located on the long arm of

chromosome 7 (7q31.2) comprising of 27 exons and spanning over 190 kb with a spliced

mRNA of 6.5 kb.18 Over 2000 reported mutation variants occur in the cystic fibrosis

transmembrane conductance regulator (CFTR) gene.1 Only approximately 400 mutation

variants, however, are known to be disease causing.19–21. Figure 1 displays some of the

common genetic mutations and their associated CFTR mutations and phenotype. The most

common CFTR mutation known is the deletion of the phenylalanine amino acid at the 508

position of the protein, also known as F508del. Approximately 50% of CF patients present

homozygous for this mutation while approximately 90% are heterozygous.22 Other common

CFTR alleles noted in CF patients are the G542X, G551D, and M1101K, which is regularly

seen in the closed communities of Hutterites.22,23

1.3 Cystic Fibrosis Transmembrane Conductance Regulator

The CFTR is a cAMP regulated chloride ion transport channel of the adenosine

triphosphate (ATP) – binding cassette (ABC) transporter family.24–27 Mutations in the CFTR

are classified into six major classes based on their molecular malfunction as seen in figure

1.28,29 These classifications are: i) defective protein synthesis due to lack of mRNA synthesis

or lack of protein synthesis, ii) improper trafficking of the protein to the membrane surface,

iii) impaired chloride channel regulation, iv) decreased conductance of chloride through the

channels at the membrane surface, v) reduction in the quantity of CFTR protein at the cell

surface, and vi) decreased stability of the CFTR protein at the cell surface.18,28 Phenotypic

severity decrease with increasing class number with the first three classes creating a lack of

functional CFTR proteins at the cell surface (more severe manifestations) and the last three

classes associated with a functionally inhibited CFTR (less severe manifestations).30–32

Figure 1: CFTR mutation classes with phenotypic descriptions with common associated genetic mutations.

1.4 Lung Disease Manifestations in CF

CF disease manifestation in the lungs, as seen in figure 2, occurs through repeated cycles

of infection and inflammation of the obstructed bronchi caused by the abnormal thickening

of the airway mucous, especially in the peripheral regions.33 The abnormally thickened

mucosal layer inhibit proper mucociliary function which allows for further aggregation of

mucous and decreased clearance of inhaled debris and/or microorganisms.34,35 Lack of

mucosal clearance allows for microorganisms – including opportunistic organisms - to persist

and establish chronic infections leading to cyclical bouts of inflammation resulting in further

secretion of thickened mucus and mucocialiary damage.1,36 These repeated cycles leads to the

permanent dilation of the bronchi known as bronchiectasis leading to chronic cough, sputum

production, recurrent respiratory infection, and acute pulmonary exacerbations.37–39 Ongoing

recruitment of the inflammatory cells progressively damages the airways furthering

bronchiectasis repeatedly until end-stage lung disease results in the death of the patient

unless a lifesaving lung transplant can be implemented.40–42

Figure 2: Progressive CF lung disease starting from a defective CF gene and subsequent CFTR inhibition to the cyclic infection and inflammation phases exacerbating bronchiectasis.

1.5 Canonical CF Pathogens

Many common CF pathogenic microorganisms noted in routine clinical tests are

opportunistic and exist naturally in the immediate environment and are capable of infecting

individuals with structural lung disease through this impairment in immune defense.

Although CF pulmonary infections can be caused by virus, fungi or bacteria, it’s the latter

that commonly cause problematic chronic infections.43 Canonical bacterial pathogens

regularly noted in CF patients include but are not limited to: Pseudomonas aeruginosa (PA),

Stenotrophomonas maltophilia (SM), Burkholderia cepacia complex (Bcc), Achromobacter

species (AS), non-tuberculous mycobacteria (NTM) such as Mycobacteroides abscessus

complex (MABC) and Mycobacterium avium complex (MAC).44 Primary pathogens such as

Haemophilus influenzae (HI), and Staphylococcus aureus (SA) (both methicillin sensitive

and methicillin resistant (MRSA) variants) are also very commonly noted in CF patients.

These pathogens are regularly present in environmental reservoirs such as water, soil,

animals, and food products from where they are known to infect.43 CF patient-to-patient

transmission have been noted due to improper hospital or home isolation, prevention, and

control measures.43

Table 1: Core members of CF pathogenesis. Data adapted from the 2019 CF Canada annual report. Pathogen Prevalence Relevance in CF disease

in Canada

(2019)

S. aureus 53% Most common pathogen in children and adolescents – infection

associated with bronchial inflammation and decreasing

pulmonary function.45

P. aeruginosa 38% Previously the most common infection in adults with over 50%

prevalence and commonly transmitted between patients.46,47

B. cepacia 4% Frequent in end-stage lung disease, associated often with poor

complex prognosis and requirement for lung transplantation. Patient to

patient transmission commonly observed with some species.48

S. maltophilia 14% Increasingly prevalent pathogen with inherent resistance to a

wide range of broad-spectrum antibiotics and common patient to

patient transmission.49,50

H. influenzae 11% Common infections in young children with multiple

simultaneous strains while 10-15% infection observed in adults

with one dominating strain.51,52

MRSA 6% Significant decline in pulmonary function values and longer

hospital stays compared to controls.53,54

NTM 6% Increasingly detected CF pathogen frequenting older age, greater

pulmonary function patients with higher MSSA and lower P.

aeruginosa infections.55,56 MAC and MABC are the most

prevalent of NTM infections with infections associated with

worsening lung function and prolonged IV antibiotics57–59

Achromobacter 5% Increasingly prevalent pathogen with chronic infections leading

Species to impaired pulmonary function and higher frequency of

hospitalization.56,60

1.6 The CF Lung Microbiome

Contrary to previous beliefs of the sterility of the lower airways, advancement in the next

generation sequencing technologies have allowed for a greater depth of analysis of the

polymicrobial community that in fact inhabits the lungs as with various other bodily niche,

albeit at a lower relative biomass.61–63 A vast array of bacterial, viral, and eukaryote species

have been discovered to inhabit the lower airway cavity and create a diverse and complex

society of microorganisms.64 Several airway environment factors influence the community

dynamics that are observed including oxygen concentrations, blood flow, and pH and are

ever changing through the lifespan of an individual.65

The natural aging process of CF patients combined with repeated infections and clinical

interventions shape an ever-evolving lower airway microbiome.66–68 Traditional culture-

based protocols of identification using selective media and aerobic growth conditions only

identified a small breadth of the overall organisms that were present.69 Analysis of the 16S

ribosomal RNA sequence diversity has allowed for the identification of numerous anaerobic

and aerobic bacteria that inhabit the lower airways but were previously under recognized.70

The 16S rRNA gene contains conserved regions within species along with polymorphisms in

hypervariable regions of the gene that allows differentiation between species to establish

phylogenic identities.71 Through the increased use of next generation sequencing technology,

recent studies have been able to identify 20 to over 1000 unique taxa present in a single CF

sputum sample.68 The airways in itself vastly differ in their compositional diversity and

abundance between the lower airways and oropharynx.72

Figure 3: Phylogenetic tree displaying the phyla and genera of bacteria isolated from CF lung sputum based on 16S rRNA sequence differences against the Greengenes 2013 database with at least 97% identity. Operational taxonomic unit (OTU) representing greater than 1% of the entire collection are represented with black dots with size correlating to abundance.73

1.7 Clinical Relevance of the CF Lung Microbiome

Several recent studies have effectively allowed for the consensus that individuals with an

advanced lung disease are limited in their lung microbiome diversity and are often dominated

by a few organisms.74–76 Some postulate that crucial interventions for a range of conditions

including CF will only be completely comprehended by understanding the ecological and

evolutionary relationships within the microbial community as well as with the host.77,78 One

such example is Cuthbertson et al. (2020), which explores the percent predicted forced

expiratory volume in 1 second (%FEV1) in comparison to the microbial diversity present in

74 the CF airways. %FEV1 is a useful measure of clinical health for individuals with lung

disease and allow for the clinicians to make critical decisions for treatment.79–81 Cuthbertson

found significant correlation between decreased diversity and increased dominance with a

reduction in lung function as well as a strong negative correlation between microbial

diversity and dominance.74

Additional research such as Acosta et al. (2018) provided association of specific

pathogens alongside overall diversity towards progression to end-stage lung disease.82

Acosta et al. demonstrated that a low microbiome diversity dominated by PA has a higher

association with progression to end-stage lung disease after five years than did a microbiome

dominated by Streptococcus.82 Further microbiome associations have been made by Heirali

et al. (2017) correlating lack of response to inhaled aztreonam lysine with higher OTU

presence of Fusobacterium, Staphylococcus, Prevotella, and Bacteroides.83 Although current

medical practice has yet to recognize the use of the lung microbiome as an effective clinical

decision making tool, continual research such as these of the CF airways may lead to

solidified associations between the microbiome and canonical infections.82,84

Table 2: The CF lung microbiome in health and disease and its supporting publication. Factor Change Strength of data Citations

Pulmonary Microbiome diversity remains Strong Stressmann et al. (2011) exacerbations stable regardless of clinical Zhao et al. (2012) status. Fodor et al. (2012) Carmody et al. (2013) Price et al. (2013) Carmody et al. (2015) Acute Immediate decrease in Strong Tunney et al. (2011) antibiotics richness with rapid reversal to Stressmann et al. (2012) norm but overall microbiome Fodor et al. (2012) diversity stays stable. Zhao et al. (2012) Smith et al. (2014) Heirali et al. (2019) Heirali et al. (2020) Age Microbiome diversity Strong Cox et al. (2010) increases during first decade Zhao et al. (2012) of life then decrease through Stressmann et al. (2012) adolescents and adulthood. Stokell et al. (2015) Baseline lung Poor lung function correlates Strong Cox et al. (2010) function with low microbiome Fodor et al. (2012) community diversity. Carmody et al. (2013) Coburn et al. (2015) Acosta et al. (2017) Genotype Microbiome association noted Weak Cox et al. (2010) in patients with F508 mutations compared to rare mutations.

1.8 Introduction to the Mycobacterium Genus

The genus Mycobacterium includes over 180 species belonging to the phylum

Actinobacteria.85,86 Members of the genus are Gram positive, rod shaped and Acid Fast.

Many are free living organisms and can be readily isolated from environmental niches such

as soil, water, animals, food products, and metalworking fluids.87,88 Mycobacteria is

subdivided into the quartet of M. tuberculosis complex, M. leprae complex, slow growing

non-tuberculous Mycobacterium (NTM) and rapid growing NTM (7 days is used to

distinguish slow from rapid growers).85 M. tuberculosis complex and M. leprae complex

species cause historically important human diseases; tuberculosis and leprosy,

respectively.87,89 NTM species, on the other hand, are opportunistic pathogens preying on

individuals with pre-existing conditions such as lung disease, immunodeficiencies or who are

receiving immune-modulating treatments.90

Tuberculosis has a global disease burden in excess of 10 million people each year with

approximately 1.4 million deaths and is second only to HIV as an infectious cause of

death.89,91,92 Leprosy affects approximately 200,000 people mainly in the tropical countries

and causes major deformities in the infected individuals often leading to an early mortality.89

M. tuberculosis is transmitted through the inhalation of aerosolized particles of the bacteria

from the aerosols and droplets from infected individual. In contrast, the exact method of

transmission of M. leprae complex is unknown, however, it is postulated to be in a similar

matter.93,94 The remaining mycobacteria species are classified under the NTM label and

opportunistically cause disease in various organs but very often manifest as pulmonary

infections.90 NTM is subdivided into slow-growing species that take 7 to 42 days to develop

16 colonies in vitro and include M. avium complex and M. kansasii, and rapid-growing species that form colonies in under 7 days and include M. abscessus complex.

1.8.1 An introduction to Mycobacteroides (mycobacterium) abscessus

MABC often manifest in patients with predisposed immune conditions and infects the respiratory tract, skin, or soft-tissues.90,95 Advancement in isolation techniques and identification strategies have allowed for increasing detection of MABC, which accounts for approximately 15% of all NTM infection within North American CF Patients.55,95,96 MABC infections are more prevalent in the European cohorts where infections can be up to 70% of total NTM observed.97–99 Additionally, recent evidence suggests potential patient-to-patient transmission within CF communities of MABC which was previously unrecognized.88,100

MABC is a particularly difficult organism to treat once infection has progressed requiring a barrage of toxic antibiotics delivered over a prolonged course. During the induction phase, treatment of MABC is conducted using a combination therapy of intravenous amikacin, alongside one or more of tigecycline, imipenem, and/or cefoxitin, supplemented with oral clarithromycin or azithromycin for one to 3 months. 66–68 These therapies are particularly difficult to tolerate and are associated with very high toxicity as complications. This is followed by a 12-month continuation phase comprised of the oral azithromycin or clarithromycin alongside with inhaled amikacin and 2 to 3 additional orally bioavailable antibiotics from a list of minocycline, clofazimine, moxifloxacin, and/or linezolid.102,103

Azithromycin is a semi-synthetic macrolide with action against the 50S subunit of bacterial ribosome inhibiting protein synthesis.104 Amikacin, tigecycline and minocycline are members of the aminoglycoside, glycylcycline, and tetracycline classes, respectively. These all have activity against the 30S bacterial ribosome halting peptide synthesis – albeit by different mechanisms.105–107 Imipenem and cefoxitin are both beta-lactam that inhibits the synthesis of bacterial cell walls.108 Moxifloxacin is a fluoroquinolone that binds to

18 topoisomerase II and IV and inhibits DNA replication and transcription.109 Clofazimine is a riminophenazine originally developed for the treatment of tuberculosis and a component of multidrug combination therapy for leprosy. Its mechanism of action is a combination of membrane disruption and the generation of reactive oxygen species.110 Lastly, linezolid belong to a group of oxazolidinones that inhibit protein synthesis by binding the P site of the

50S ribosomal subunit.111

A joint committee of the American Thoracic Society and Infectious Disease Society of

America (ATS/IDSA) was formed in 2007 to provide guidelines for clinical use for NTM infections adopted by CF specialists and are regularly updated as such in June, 2020.102

These guidelines consider a set of clinical, radiological, and microbiological characteristics for diagnosing NTM disease. Under these guidelines as it pertains to CF, a patient must present two or more positive sputum cultures of the same NTM species or one positive culture from a bronchoalveolar lavage or wash to satisfy the microbiology criteria.

Radiological criteria are difficult to assess in CF patients as signs pertaining to a MABC infection are common irrespective of infection source and could include inflammatory nodules, tree-in-bud opacities, and cavitation. Clinical criteria looks at worsening respiratory symptoms and decreasing pulmonary functions despite of treatment of common CF pathogens.102

In 2016, a consensus recommendation for the management of NTM infection in individuals with CF was created by the US Cystic Fibrosis Foundation and the European

Cystic Fibrosis Society to help better identify and treat NTM infections with specific focus to

CF patients.103 The guidelines for treatment built on the 2007 ATS/IDSA criteria with adaptation based on patient variables, putative risks, and expectation of outcome.103

1.9 Hypothesis

I hypothesize that the diversity and composition of the community of bacteria residing in

the CF lower airways will influence the susceptibility of the host to invasion by MABC

infection. Thus, my hypotheses are:

1. The microbiome of patients with incident infections of MABC will have fundamental

differences from patients who do not experience the infection. Patients who do not

experience incident infections may thus be protected by colonizing lower airway

communities that resist invasion.

2. Among those patients who do develop incident infections, two results are possible.

Clearance of the pathogen, whereby the patient avoids the negative sequelae of

chronic infection or persistence of infection whereby adverse outcomes ensue.

Differences in community structure may yield information on organisms that assist in

pathogen clearance. We hypothesize that airway microbiome will enable distinction

between these two outcomes.

1.9.1 Project Aims i) To determine how the microbiome of patients with each incident infection

evolves over time – encompassing the period of time prior to infection, through

initial infection prior to treatment, both induction and consolidation therapies and

finally resolution/persistence.

ii) To determine if the microbiome may serve as a novel biomarker to identify

patients at risk of infection by determining if the clinical demographics and

microbial communities of patients with incident infections differ from control

patients matched for age and gender.

iii) To determine if specific factors within the microbiome associate with treatment

outcomes.

Chapter Two: Methods and Materials

2.1 Identification of MABC Cohort

Patients for our cohort study were identified from a retrospective review of the

microbiology clinical records with matching sputum samples from the Calgary Adult Cystic

Fibrosis Clinic to which the Parkins laboratory has access. Clinical demographic and

dynamic variables of disease were obtained from medical records. Comprehensive patient

records accompanying the sputum samples included: demographics such as nutritional status

(body mass index (BMI) (kg/m2)), age, gender, co-morbidities, cftr genetic information;

infection details; infection history; treatments received; treatment outcomes; antibiotic

resistance data; and key spirometic values including percent predicted forced vital capacity

(%FVC), and percent predicted forced expiratory volume in one second (%FEV1). Cases

were classified as incident (occurring during the time of follow-up within the CACFC) or

pre-existing (occurring prior to transition to CACFC). Case patient that were selected had at

least one positive MABC culture (Middlebrook 7H10) and were tested by Alberta Health

Services Laboratory where the pathogen samples were also stored. Classification of

chronicity was conducted using the criteria established by the ATS/IDSA.112 Patients

deemed persistent with an MABC infection were seen with at least half of their cultures

positive with a minimum of three samples in the year following initial infection. Patients

with less than half of their cultures positive were hence deemed transient.

2.2 Identification of Control Patient Cohort

Control patients were identified from the CACFC cohort and were only selected if there

was no known previous recorded case of MABC in their respiratory secretions. Two controls

were selected for each patient that were age matched to within two years of the initial

infection date. The control patients were also matched to their respective case’s sex and time

period. For each control patient, one samples was selected. For samples to be analyzed we

ensured that they were collected earlier than four weeks after a change in antibacterial

therapy or pulmonary exacerbation to avoid confounding effects. Patient demographic data

was similarly collected for each control patient such as nutritional status (body mass index

(BMI) (kg/m2)), age, gender, co-morbidities, cftr genetic information; infection details;

infection history; treatments received; treatment outcomes; antibiotic resistance data; and key

spirometic values including %FVC and %FEV1.

2.3 Patient Sputum Sample Identification

Sputum samples from attendees are meticulously stored within the Calgary Adult CF

Clinic Biobank housed in multiple -80˚Celsius freezers to ensure optimal conservation to the

original state of the samples. Sputum samples have been collected and stored since 1998 and

are continuously updated with new arriving samples. As of 2019, the biobank contains over

18,000 samples from over 275 adult patients who have prospectively consented towards the

collection and storage of sputum samples for research purposes. Most sputum samples were

homogenized and stored into vials or cryogenic tubes immediately after clinic visit by the

patients. Samples are also received from the clinical microbiology, documented, and placed

into long-term storage in -80˚Celsius freezers. Inclusion criteria included identification of

Mycobacteroides abscessus complex within acid-fast bacillus (AFB) sputum culture during

follow up within the CF clinic; Patients were excluded if their diagnosis of MABC predated

infection and no positive cultures were obtained during the period of follow up at the

CACFC. Cases and control patient samples were only selected if they were not influenced

immediately by antibiotic exposure (within two weeks) not relevant to the treatment

associated with MABC. Samples were classified into stages of disease and were as follows:

1. Pre-infection – samples preceding the infection date up to two years unless insufficient

sputum samples were available or were impacted by antibiotic intervention for other

infections.

2. At-infection – samples dated on the initial clinical positive.

3. Induction therapy – initiation of therapy course with numerous potent IV antibiotics to

eradicate majority of MABC.

4. Consolidation therapy – continuation of several IV and oral antibiotics following the

induction therapy phase to eradicate any remaining MABC.

5. Maintenance therapy – continuation of one or more antibiotics as a complement to

induction and/or consolidation therapy.

6. Intensification therapy – reversion to induction therapy with other antibiotics or a higher

dosage.

7. Clearance – samples that have cleared MABC infection either due to the antibiotic

therapy or through spontaneous measures.

8. Chronic and off treatment – samples that have persistent MABC infections and have been

taken off antibiotic therapy.

Samples associated with the antibiotic therapy were grouped into a single “at treatment” stage. The natural history of each patient was reviewed and detailed in the medical record.

Sputum samples were specifically sought from a number of categories and scored according to table 3.

Table 3: Case patient sputum samples categorized to natural history of disease stages according to time frame, AFB culture, and AFB smear status. Natural Time Frame Last AFB culture Culture positive/days to AFB smear status History of conducted positive Disease Pre-infection Two years Ziehl Neelsen Samples collected in the Samples were before initial Stain to assess for time period before a assessed by Ziehl- infection date the presence of clinical MABC positive Neelsen Stain for (unless bacteria. Samples was observed bacterial bioburden. insufficient ranged positive Samples ranged number from latest clinic from smear negative samples visit to 36 months to a smear 3+ available) prior with several patients never having had one before At infection First clinical Positive samples Positive samples ranged Samples ranged positive date ranged from from 1 – 7 days to grow. from smear negative to initiation latest clinic visit Patients with negative to a smear 3+ of treatment to a year prior samples either grew again within the year or never grew again

At Treatment All stages Positive samples Positive samples ranged Samples ranged involved from ranged from from 1 – 7 days to grow. from smear negative induction of latest clinic visit Patients with negative to a smear 3+ therapy, to 1.5 months samples either grew consolidation, prior again within the year or maintenance, never grew again and intensification

Post After Positive samples Positive Samples were Samples ranged Treatment treatment ranged from either chronic and from smear negative phase with latest clinic visit cultures ranged from 1 to a smear 3+ either to 2 months prior day to over 7 days to (persistent patients) clearance of grow or were cleared and MABC or did not grow MABC chronic within the next year designation

2.4 Statistical analysis

Patient and control demographic data were analyzed using 2-sided fisher’s exact and

Chi-squared (2) tests for distinct values while Wilcoxon rank-sum was used to analyze

continuous values. Fisher’s exact was used to test significant differences between the

proportions of one variable and whether they are different among values of the other

variable.113 Fisher’s exact test is more reliable for a small number of samples being compared

and assumes that compared values are not related.114 Fisher’s exact also relies on the

assumption that the null hypothesis is true.115 To calculate Fisher’s exact test, a 2 x 2

contingency table being analyzed is compared to all possibilities of that 2 x 2 table fitting

within a marginal tool. The equation below allows to equate the p-values for each possible 2

x 2 table. To calculate the one-sided Fisher’s exact, the probability of the 2 x 2 proportions

in question and all probabilities more extreme are summed. For a two-sided Fisher’s exact

value, a mid-p method can be used where the p-values more extreme than the observed are

added to one half of the observed p-value.116 If the resulting p-value is less than 0.05, a

statistical significance can be assumed and the null hypothesis rejected.

(푎 + 푏)! (푐 + 푑)! (푎 + 푐)! (푏 + 푑)! 푝 = (푎 + 푏 + 푐 + 푑)! 푎! 푏! 푐! 푑!

The 2 test for independence is used when comparing two variables of discrete values

when the population size is relatively large. The 2 equation relies on the difference between

the observed value and the expected value, granted the variables are completely independent

of each other.113 The 2 test depends on a null hypothesis which can be failed to be rejected

or rejected and an alternative hypothesis that is accepted. A 2 distribution is employed

alongside the degrees of freedom (number of variables minus 1) to extrapolate the p-value.

(푂 − 퐸)2 2 = ∑ 퐸

The Wilcoxon Rank-Sum test is used to calculate probabilities of two independent variables consisting of continuous values.117 Wilcoxon Rank-Sum is a non-parametric test that assumes as the null hypothesis that the competing observation median have the same continuous distribution. This test is used in replacement of a t-test when two populations have unequal spread and are not normally distributed. The test uses a summed rank of the values present in each variable compared to the median of the extremes to assess if there are significant differences between the variables to reject the null hypothesis. The Wilcoxon

Rank-Sum test is very similar to the Mann-Whitney test and often produces the same results.

The following formula is used for each of the groups with the lower resulting U being used to determine the p-value based off a table of critical values.

푛 (푛 + 1) 푈 = 푛 푛 + 1 1 − 푅 1 1 2 2 1

푛 (푛 + 1) 푈 = 푛 푛 + 2 2 − 푅 2 1 2 2 2

The Kruskal-Wallis test was used to calculate probability of non-parametric continuous dependent variables by categorical independent variable of two or more groups.118 Similar to

Mann-Whitney and Wilcoxon Rank-Sum, Kruskal-Wallis test assumes an abnormal distribution with a varying spread. The null hypothesis assumes that groups are from identical populations with the Kruskal-Wallis test aimed to reject the null hypothesis and accept the alternative hypothesis that at least one group is different from the others.

Permutational multivariate analysis of variance (PERMANOVA) was used to compare the beta-diversity Bray-Curtis dissimilarities matrix and test the null hypothesis that the

29 centroids and dispersion of the groups are equally distributed for all groups.119

PERMANOVA assumes that all objects in the data set are exchangeable, independent, and have similar multivariate dispersion. PERMANOVA analysis was done in R with stratification using 999 permutations to normalize dataset proportionally for the categories being analyzed.119 Differences in taxa was assessed using the DESeq2 package with a

“Wald” test and “parametric” fit type.120 Lastly, multiple comparisons of p-values of relative abundance for each taxa was calculated using two-tailed Mann-Whitney U test on PRISM and corrected using Benjamini-Hochberg for multiple comparisons.

These statistical analysis were conducted on STATA v.13.0 (StataCorp, Texas, USA),

PRISM v8.2.1 (GraphPad, California, USA), and/or R v3.5.0 (R Core Team, 2018).

2.5 DNA Extraction

Total DNA extraction was conducted using a previously published and validated method

from the McMaster Genome Facility (Hamilton, ON) as well as other microbiome

publications from the Parkins laboratory (Heirali, 2017; Acosta, 2018; Heirali, 2019).82,83,121

The process included extreme care in ensuring samples remain as near to the state of sample

collection. This includes using dry ice to transfer and store the samples until the time of

extraction. All exactions were conducted in the biosafety cabinet after a thorough cleaning

with 70% ethanol and UV light eradication of any contaminants. Contaminations during the

extraction process could lead to false identification of microbiota resulting in data that is

unreliable and unreproducible. Proper personal protective equipment is also worn to ensure

minimal impact of contaminations and are cleaned between samples using 70% ethanol. A

negative control was included with every extraction set to ensure lack of cross contamination

and verified using PCR and gel electrophoresis. Select negative controls (6) were also

sequenced and qPCR for presence of bacterial 16S rDNA.

Genomic DNA extraction was performed as previously described.82,83,121,122 Briefly,

frozen sputum samples were thawed and homogenized using a 18mm syringe needle. 300

µL of each sample was mechanically lysed by bead beating with 800 µL NaPO4 (200mM)

and 100 µL of guanidine thiocyanate-ethylenediaminetetraacetic acid followed by enzymatic

lysis using 50 µL of lysozyme (100mg/mL) (Sigma-Aldrich, #L6876-10G), 50 µL

mutanolysin (10 U/µL) (Sigma-Aldrich, #M9901-50KU), and 10 µL RNase A (10 mg/mL)

(Qiagen, #19101) and incubated for 90-minutes at 37˚C. A secondary lysis step was

performed next using 25 µL sodium dodecyl sulfate (25%), 25 µL proteinase K (Sigma-

Aldrich, #P2308-1G), and 62.5 µL NaCl (5 M) and incubated again for 90-minutes at 65˚C.

The samples were centrifuged at 13,500 X g for 5 minutes. DNA was then isolated using phenol-chloroform-isoamyl (Sigma-Aldrich, #P3803- 400mL) and purified through a DNA column (Cedarlane laboratories, #D4034) using a DNA binding buffer and DNA wash buffer

(DNA clean and concentrator-25, Cedarlane laboratories, #D4034).122 Quantification of the eluted DNA was conducted using a NanoDrop-1000 (Thermo Scientific, Waltham, MA).

2.6 16S rRNA Gene Amplification

100 ng of the extracted sample DNA was amplified for the V3 region of the 16S rRNA

gene using modified 341F and 518R 16S rRNA primers for adaptation to Illumina Miseq

(San Diego, CA).122–124 Extraction blanks and its consequent PCR reactions were included to

ensure no contamination had occurred during the process. Extracted DNA from sputum

samples are analysed by the amplification of the V3 hypervariable region of the 16S rRNA

gene. The V3 region is used due to its small number of total nucleotides (65 base pairs)

containing the maximum number of single nucleotides polymorphisms between nucleotides

456 and 479.125 The amplification was done in a polymerase chain reaction (PCR) machine

using modified forward and reverse primers using the methodology developed by Bartram et

al.126 The DNA acquired from the DNA extraction was run through gel electrophoresis to

confirm bands at approximately 300 base pairs, the length of the V3 region. Once ensured

that the amplified DNA is correct, the PCR samples were sent to the McMaster Genome

Facility in Hamilton, ON to be sequenced using their Illumina MiSeq technology.

For each sample, a master mix was created with the following: 5µL of 10x PCR buffer

(Life Technologies, #10342020), 1.5µL of 50mM MgCl2 (Life Technologies, #10342020), 1µL of

10mM dNTPs (New England Biolabs, #N0447L), 5µL of 1µM V3F primer (5 pmoles), 5µL of

1µM V3R primer (5 pmoles), 0.25µL of Taq polymerase (Life Technologies, #10342020),

and 27.25µL of ddH2O to a total of 45µL per sample. Then add 5µL of the extracted DNA

sample normalized to 100ng/50µL. Split each sample tube into thirds (~17µL in each tube).

Place sample tubes in PCR machine and program as follows:

• 94˚C for 2 mins

• 30 cycles of:

o 94˚C for 30 secs

o 50˚C for 30 secs

o 72˚C for 30 secs

• 72˚C for 10 mins

• 4˚C hold

Table 4. 16S rRNA V3 region forward and reverse barcoded primers.

Primer Adapter Sequence

caagcagaagacggcatacgagatCGTGATgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_1F P7 GAGGCAGCAG

caagcagaagacggcatacgagatACATCGgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_2F P7 GAGGCAGCAG

caagcagaagacggcatacgagatGCCTAAgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_3F P7 GAGGCAGCAG

caagcagaagacggcatacgagatTGGTCAgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_4F P7 GAGGCAGCAG

caagcagaagacggcatacgagatCACTGTgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_5F P7 GAGGCAGCAG

caagcagaagacggcatacgagatATTGGCgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_6F P7 GAGGCAGCAG

caagcagaagacggcatacgagatGATCTGgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_7F P7 GAGGCAGCAG

caagcagaagacggcatacgagatTCAAGTgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_8F P7 GAGGCAGCAG

caagcagaagacggcatacgagatCTGATCgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_9F P7 GAGGCAGCAG

caagcagaagacggcatacgagatAAGCTAgtgactggagttcagacgtgtgctcttccgatctCCTACG V3_10F GGAGGCAGCAG P7 caagcagaagacggcatacgagatGTAGCCgtgactggagttcagacgtgtgctcttccgatctCCTACG V3_11F P7 GGAGGCAGCAG

Primer Adapter Sequence

caagcagaagacggcatacgagatTACAAGgtgactggagttcagacgtgtgctcttccgatctCCTACG GGAGGCAGCAG V3_12F P7

caagcagaagacggcatacgagatCGTACTgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_13F P7 GAGGCAGCAG

caagcagaagacggcatacgagatGACTGAgtgactggagttcagacgtgtgctcttccgatctCCTACG V3_14F P7 GGAGGCAGCAG

caagcagaagacggcatacgagatGCTCAAgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_15F P7 GAGGCAGCAG

caagcagaagacggcatacgagatTCGCTTgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_16F P7 GAGGCAGCAG caagcagaagacggcatacgagatTGAGGAgtgactggagttcagacgtgtgctcttccgatctCCTACG V3_17F P7 GGAGGCAGCAG

caagcagaagacggcatacgagatACAACCgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_18F P7 GAGGCAGCAG

caagcagaagacggcatacgagatACCTCAgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_19F P7 GAGGCAGCAG

caagcagaagacggcatacgagatACGGTAgtgactggagttcagacgtgtgctcttccgatctCCTACG V3_20F P7 GGAGGCAGCAG

caagcagaagacggcatacgagatAGTTGGgtgactggagttcagacgtgtgctcttccgatctCCTACG V3_21F P7 GGAGGCAGCAG

caagcagaagacggcatacgagatCTCTCTgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_22F P7 GAGGCAGCAG

caagcagaagacggcatacgagatCAAGTGgtgactggagttcagacgtgtgctcttccgatctCCTACG V3_23F P7 GGAGGCAGCAG

caagcagaagacggcatacgagatCCTTGAgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_24F P7 GAGGCAGCAG

Primer Adapter Sequence

caagcagaagacggcatacgagatACCACTgtgactggagttcagacgtgtgctcttccgatctCCTACGG GAGGCAGCAG V3_25F P7

caagcagaagacggcatacgagatAGTGTCgtgactggagttcagacgtgtgctcttccgatctCCTACGG GAGGCAGCAG V3_26F P7

caagcagaagacggcatacgagatAGAAGGgtgactggagttcagacgtgtgctcttccgatctCCTACG GGAGGCAGCAG V3_27F P7

caagcagaagacggcatacgagatTTATCCgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_28F P7 GAGGCAGCAG

caagcagaagacggcatacgagatTTAAGGgtgactggagttcagacgtgtgctcttccgatctCCTACG GGAGGCAGCAG V3_29F P7

caagcagaagacggcatacgagatTTCTTGgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_30F P7 GAGGCAGCAG

caagcagaagacggcatacgagatTTCAACgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_31F P7 GAGGCAGCAG

caagcagaagacggcatacgagatTTGTGAgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_32F P7 GAGGCAGCAG

caagcagaagacggcatacgagatTTGACTgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_33F P7 GAGGCAGCAG

caagcagaagacggcatacgagatTATTCGgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_34F P7 GAGGCAGCAG

caagcagaagacggcatacgagatTATAGCgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_35F P7 GAGGCAGCAG

caagcagaagacggcatacgagatTAACTCgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_36F P7 GAGGCAGCAG

Primer Adapter Sequence

caagcagaagacggcatacgagatTACCAAgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_37F P7 GAGGCAGCAG

caagcagaagacggcatacgagatTACGTTgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_38F P7 GAGGCAGCAG

caagcagaagacggcatacgagatTAGTACgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_39F P7 GAGGCAGCAG

caagcagaagacggcatacgagatTAGATGgtgactggagttcagacgtgtgctcttccgatctCCTACG V3_40F P7 GGAGGCAGCAG

caagcagaagacggcatacgagatTCTACAgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_41F GAGGCAGCAG

P7 caagcagaagacggcatacgagatTCTGATgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_42F P7 GAGGCAGCAG

caagcagaagacggcatacgagatTCATGTgtgactggagttcagacgtgtgctcttccgatctCCTACGG GAGGCAGCAG V3_43F P7

caagcagaagacggcatacgagatTGTCTAgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_44F P7 GAGGCAGCAG

caagcagaagacggcatacgagatATTCTCgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_45F P7 GAGGCAGCAG

caagcagaagacggcatacgagatATTGAGgtgactggagttcagacgtgtgctcttccgatctCCTACG V3_46F P7 GGAGGCAGCAG

caagcagaagacggcatacgagatATACCTgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_47F P7 GAGGCAGCAG

caagcagaagacggcatacgagatATGCAAgtgactggagttcagacgtgtgctcttccgatctCCTACG V3_48F P7 GGAGGCAGCAG

caagcagaagacggcatacgagatAATCCAgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_49F P7 GAGGCAGCAG

Primer Adapter Sequence

caagcagaagacggcatacgagatAATGGTgtgactggagttcagacgtgtgctcttccgatctCCTACG V3_50F P7 GGAGGCAGCAG

caagcagaagacggcatacgagatAACTAGgtgactggagttcagacgtgtgctcttccgatctCCTACG V3_51F P7 GGAGGCAGCAG

caagcagaagacggcatacgagatAACACTgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_52F P7 GAGGCAGCAG

caagcagaagacggcatacgagatAAGAGAgtgactggagttcagacgtgtgctcttccgatctCCTACG V3_53F P7 GGAGGCAGCAG

caagcagaagacggcatacgagatACTTACgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_54F P7 GAGGCAGCAG

caagcagaagacggcatacgagatACATTGgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_55F P7 GAGGCAGCAG

caagcagaagacggcatacgagatACGAATgtgactggagttcagacgtgtgctcttccgatctCCTACG V3_56F GGAGGCAGCAG P7 caagcagaagacggcatacgagatAGTCATgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_57F GAGGCAGCAG P7

caagcagaagacggcatacgagatAGAAGTgtgactggagttcagacgtgtgctcttccgatctCCTACG P7 GGAGGCAGCAG V3_58F caagcagaagacggcatacgagatCTTATGgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_59F P7 GAGGCAGCAG

caagcagaagacggcatacgagatCTAGAAgtgactggagttcagacgtgtgctcttccgatctCCTACG GGAGGCAGCAG V3_60F P7

caagcagaagacggcatacgagatCATCTTgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_61F P7 GAGGCAGCAG

caagcagaagacggcatacgagatCACATAgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_62F P7 GAGGCAGCAG

Primer Adapter Sequence

caagcagaagacggcatacgagatCCAATTgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_63F P7 GAGGCAGCAG

caagcagaagacggcatacgagatCGATTAgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_64F P7 GAGGCAGCAG

caagcagaagacggcatacgagatGTTAGTgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_65F P7 GAGGCAGCAG

caagcagaagacggcatacgagatGTAACAgtgactggagttcagacgtgtgctcttccgatctCCTACG V3_66F P7 GGAGGCAGCAG

caagcagaagacggcatacgagatGTGTATgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_67F P7 GAGGCAGCAG

caagcagaagacggcatacgagatGATAAGgtgactggagttcagacgtgtgctcttccgatctCCTACG V3_68F P7 GGAGGCAGCAG

caagcagaagacggcatacgagatGAATCTgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_69F P7 GAGGCAGCAG

caagcagaagacggcatacgagatTTCCGTgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_70F P7 GAGGCAGCAG

caagcagaagacggcatacgagatTTCGCAgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_71F P7 GAGGCAGCAG

caagcagaagacggcatacgagatTTGGTCgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_72F GAGGCAGCAG

P7 caagcagaagacggcatacgagatTGACAGgtgactggagttcagacgtgtgctcttccgatctCCTACG V3_73F P7 GGAGGCAGCAG

caagcagaagacggcatacgagatATCTGCgtgactggagttcagacgtgtgctcttccgatctCCTACGG GAGGCAGCAG V3_74F P7

caagcagaagacggcatacgagatACACGAgtgactggagttcagacgtgtgctcttccgatctCCTACG V3_75F P7 GGAGGCAGCAG

Primer Adapter Sequence

caagcagaagacggcatacgagatAGGTTCgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_76F P7 GAGGCAGCAG

caagcagaagacggcatacgagatCATGACgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_77F P7 GAGGCAGCAG

caagcagaagacggcatacgagatGCTATCgtgactggagttcagacgtgtgctcttccgatctCCTACGG GAGGCAGCAG V3_78F P7

caagcagaagacggcatacgagatGGACTTgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_79F P7 GAGGCAGCAG

caagcagaagacggcatacgagatGGCAATgtgactggagttcagacgtgtgctcttccgatctCCTACG V3_80F P7 GGAGGCAGCAG

caagcagaagacggcatacgagatTCTCGGgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_81F P7 GAGGCAGCAG

caagcagaagacggcatacgagatTCAGCGgtgactggagttcagacgtgtgctcttccgatctCCTACG V3_82F P7 GGAGGCAGCAG

caagcagaagacggcatacgagatTGTGCCgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_83F P7 GAGGCAGCAG

caagcagaagacggcatacgagatTGCACGgtgactggagttcagacgtgtgctcttccgatctCCTACG V3_84F P7 GGAGGCAGCAG

caagcagaagacggcatacgagatAAGGCCgtgactggagttcagacgtgtgctcttccgatctCCTACG V3_85F P7 GGAGGCAGCAG

caagcagaagacggcatacgagatACCAGGgtgactggagttcagacgtgtgctcttccgatctCCTACG V3_86F P7 GGAGGCAGCAG

caagcagaagacggcatacgagatAGCCTGgtgactggagttcagacgtgtgctcttccgatctCCTACG V3_87F GGAGGCAGCAG P7 caagcagaagacggcatacgagatAGCGACgtgactggagttcagacgtgtgctcttccgatctCCTACG V3_88F GGAGGCAGCAG P7

Primer Adapter Sequence

caagcagaagacggcatacgagatCTACGCgtgactggagttcagacgtgtgctcttccgatctCCTACGG P7 GAGGCAGCAG V3_89F caagcagaagacggcatacgagatCTCCAGgtgactggagttcagacgtgtgctcttccgatctCCTACGG V3_90F P7 GAGGCAGCAG

caagcagaagacggcatacgagatCCGTAGgtgactggagttcagacgtgtgctcttccgatctCCTACG GGAGGCAGCAG V3_91F P7

caagcagaagacggcatacgagatCGGTGTgtgactggagttcagacgtgtgctcttccgatctCCTACG V3_92F P7 GGAGGCAGCAG

caagcagaagacggcatacgagatCGGAACgtgactggagttcagacgtgtgctcttccgatctCCTACG V3_93F P7 GGAGGCAGCAG

caagcagaagacggcatacgagatGTGCTGgtgactggagttcagacgtgtgctcttccgatctCCTACG V3_94F P7 GGAGGCAGCAG

caagcagaagacggcatacgagatGAACGGgtgactggagttcagacgtgtgctcttccgatctCCTACG V3_95F P7 GGAGGCAGCAG

caagcagaagacggcatacgagatGGATGCgtgactggagttcagacgtgtgctcttccgatctCCTACG V3_96F P7 GGAGGCAGCAG

2.7 DNA sequencing and analysis

Amplified DNA was sequenced using Illumina MiSeq at the McMaster Genome Facility

in Hamilton, ON to generate raw paired-end fastq sequence files. R (v.3.5) and RStudio

(v.1.2.5001) was used to analyse data. Within RStudio, Cutadapt (v.2.5) was used to trim

primers and Divisive Amplicon Denoising Algorithm (DADA2; v1.8.0) was used to identify

the amplicon sequence variants (ASV).127,128 Briefly, DADA2 workflow begins by filtering

forward and reverse reads by removing sequences with unknown nucleotides or more than

two expected errors and trimming to 130 nucleotides.129 Denoising of the database were

conducted using a Poisson error model which is conducted prior to merging reads.127,130 Pair

reads were merged using a mergePairs function after dereplication of identical reads were

conducted. Chimeras were identified by alignment to all more abundant sequence

combinations from left and right and removed if the sequences match exactly.127 All

singleton ASVs were readily removed in the DADA2 pipeline.127 Resulting ASVs were

annotated using the IDTAXA algorithm from the DECIPHER package referencing the most

recently updated SILVA v132 database.131,132

Microbiota genus present at above 1% abundance was visualized using alpha and beta-

diversity functions. Alpha diversity measures were conducted using Observed diversity

index (ODI), which displays the total richness (number of taxonomic groups) of bacteria

present in the samples; and Shannon diversity index (SDI), which displays the total richness

and evenness (distribution of abundance of the taxonomic groups) of the bacteria present in

each sample.133 Wilcoxon Rank-Sum tests were performed to analyze statistical differences

between groups.

Beta-diversity measures the variability in taxonomy between samples. Principle coordinate analysis (PCoA) plots were employed to visualize the community wide differences of samples relying on Bray-Curtis dissimilarity test.134 PERMANOVA analysis was used to analyse community-wide differences at the ASV level in PCoA plots.135

2.8 Quantitative PCR

Quantitative PCR (qPCR) was performed to establish the absolute 16S rDNA quantity

present in each sputum samples.136 Additionally, qPCR analysis was also used to determine

the quantity of P. aeruginosa and MABC DNA with the extracted samples.

Quantifying the absolute 16S rDNA present in each sputum sample was done by creating

a standard curve of extracted DNA from a pure culture of P. aeruginosa PA01 strain at a 1:5

serial dilution of a 25ng/µL stock. Extraction of DNA was conducted using the same DNA

extraction method as sputum samples as described previously in the DNA extraction section.

Each 10µL reaction contained 5µL of TaqMan Fast Advanced Master Mix (Thermo

Fisher), 0.5µL of TaqMan gene expression probe for pan-bacterial detection of 16S rRNA

(Ba04230899_s1; Thermo Fisher Scientific), 2.5µL of the template DNA from sputum

extraction, and 2µL of double distilled water. The qPCR reaction program was set with the

following parameters: 50˚C for 2mins followed by 95˚ for 2mins then 40 cycles of 95˚C for

1secs and 60˚C for 20secs. Mathematical correction for the copies of 16S rRNA operon

needs to be conducted since the P. aeruginosa standard being used has four copies of the 16S

rRNA operon.137 The following equations are used:

16S gene copies [DNA]ng 1g 1mol bp DNA 6.023x1023bp 4 copy 16S rRNA = ( ) ( ) ( ) ( ) ( ) (volume of template) µL 10003ng 660g DNA 1 mol bp PA genome size bp

16푆 gene copies [DNA extraction]ng (( ) ( ) (50µL)) Copies 25ng µL 1000µL = ( ) mL 300µL sputum 1mL

The abundance of P. aeruginosa DNA in sputum samples was conducted using the absolute quantification method. A standard curve was created of extracted DNA from a pure culture of P. aeruginosa PA01 strain at a 1:5 serial dilution of a 1ng/µL stock. DNA extracted from Stenotrophomonas maltophilia was used as the non-template control. The P. aeruginosa and S. maltophilia were both extracted using the same DNA extraction method as previously described in the DNA extraction section. Primers for the amplification were created as described by Qin et al. (2003)138. Duplicates of each sample were conducted with the mean quantity calculated based off the standard curve being used. Duplicates with discrepancies large discrepancies in the quantity values or other errors were repeated.

Each 10µL reaction well contained 5µL of SsoAdvanced Universal Inhibitor-Tolerant

SYBR Green Supermix (Bio-Rad), 1µL of forward and reverse primer mix (final concentration of 150nM), and 4µL of the template DNA from sputum extraction. The qPCR reaction program was set with the following parameters: initial denaturation at 95˚C for

10mins followed by 40 cycles of 95˚C for 15secs, 68˚C for 1min then ending with a melting curve analysis of one cycle with the parameters of 95˚C for 15secs, 60˚C for 1min, 95˚C for

15secs, and 60˚C for 15secs.

The abundance of MABC in sputum samples was also conducted using the absolute quantification method. Primers and probe for the amplification were created as described in

Rocchetti et al. (2016)139. Extraction of a pure culture of MABC was used for the standard curve at a 1:5 serial dilution of a 1ng/µL stock. DNA extraction from P. aeruginosa was used as the non-template control. All extractions were conducted as per the methodology previously described under DNA extraction. Duplicates of each sample were conducted with

46 the mean quantity calculated based off the standard curve being used. Duplicates with large discrepancies in the quantity values or other errors were repeated.

Each 10µL reaction contained 5µL of TaqMan Fast Advanced Master Mix (Thermo Fisher),

1µL of forward and reverse primer mix (final concentration of 900nM), 1µL of probe (final concentration of 400nM), and 3µL of the template DNA from sputum extraction. The qPCR reaction program was set with the following parameters: 50˚C for 2mins followed by 95˚ for

2mins then 40 cycles of 95˚C for 1secs and 60˚C for 20secs.

2.9 P. aeruginosa Inhibition Assay of MABC

Microbiome analysis exhibited a significant negative correlation between the presence of

Pseudomonas and MABC. To test whether these differences could be due to any excreted

factors that P. aeruginosa produced that negatively impacted MABC, an inhibition assay was

designed using P. aeruginosa supernatant. Supernatant mixture from two PA strains (P.

aeruginosa PA01 lab strain and Prairie Epidemic strain (PES)) were collected from cultures

grown in Luria-Bertani broth (LB) and filter sterilized. MABC grown in Middlebrook 7H9

broth for 5 days was incubated with both supernatant at different dilution ratios of 1:5 and

1:20. Table 5 displays the content configuration that was constructed to test MABC

susceptibility with varying dilutions of P. aeruginosa supernatant and grown for 5 days. For

each day, including initial incubation (day zero), 1:10 serial dilution of 10µL from each flask

was conducted and plated on Middlebrook 7H10. The plates were allowed 4 days to grow

cultures which were then used to calculate the CFU/mL.

Table 5: Content configuration of each flask used to test P. aeruginosa inhibition of MABC. Three replicates of each flask was used for the assay. Negative Control Control PA01 Flask Control 1:5 1:20 PA01 1:5 1:20 PES 1:5 PES 1:20 M7H9 broth 11 11 11 11 11 11 11 PA supernatant 0 0 0 4 1 4 1 LB broth 0 4 1 0 0 0 0 Bacteria 0 5 5 5 5 5 5 0.9% NaCl 9 0 3 0 3 0 3 Total 20 20 20 20 20 20 20 Dilution 0 0.2 0.05 0.2 0.05 0.2 0.05

Chapter Three: Results

3.1 Case and Control Patient Demographics

This study cohort employed patients from the Calgary Adult CF clinic and were divided

based on acquisition of MABC. Case patients were identified based on chart review and had

acquired the bacteria - and were positive based on ATS/IDSA criteria. Control patients did

not have any positive history for positive cultures for MABC as identified through clinical

review of medical records and clinical microbiology testing (Middlebrook 7H10, done

annually) for dates preceding their used samples. Twenty-one CF patients with MABC

infections were identified meeting inclusion and exclusion criteria (median age 21.8 years,

IQR 18.5 – 33.1) contributing 174 sputum samples (median 8 samples/patient, IQR 6-12).

This cohort was matched to two gender, age (±2 years) and birth cohort compliant controls.

Each control patient was assessed at a single time point – consistent with a time period

corresponding to infection in the cases – for a total of 42 control samples. Furthermore, case

patients were defined as persistent according to 50% or greater positive mycobacteriology

tests with a minimum of three tests in the year following initial infection.

Case and control patients did not differ based on gender, age, genetics, comorbidities, or

coinfections, (Tables 6A, 6B, and 6C). The only parameter differences noted between case

and control patients were related to patient body mass index (BMI), where case patients had a

lower BMI (suggesting worse nutritional status). Baseline lung function at the time of

infection also did not differ between cases and controls (Table 6A). In general, differences

were not observed between cases and controls with respect to medications patients were

receiving at the time of MABC identification. Median baseline lung function as measured by

percent predicted forced expiratory volume in one second (%FEV1) or by percent predicted

50 total forced vital capacity (%FEV) of case patients did not differ significantly from controls although tended to be slightly higher.

To understand the natural history of MABC infection we compared outcomes of those with MABC who developed persistent infection (persistent) with those who ultimately cleared infection (transient). In general, these groups did not differ with respect to demographics, co-morbidities, or co-infections (Table 6A, 6B, and 6C). In general, differences were not observed in the medications that patients were receiving at the time of first MABC infection – with one exception. The only significant difference in the medication patients received was the use of proton pump inhibitors (PPI) were observed with persistent patients more likely to be on PPI compared to transient patients as well as control patients (Table 6C).

MABC subspecies M. abscessus abscessus (17/21, 81%) dominated the case patients with M. bolletii (2/21, 10%) being a smaller factor. No M. massiliense subspecies were observed although two patient mycobacteriology tests resulted undifferentiated subspecies.

All but one case patient was infected with at least one other co-infecting bacterial species at the time of infection with MABC with eight case patients co-infected with multiple other bacterial species (Table 6B). In the control cohort, all but two patients were infected with at least one bacterial species with 18 control patients exhibiting multiple co-infections.

Table 6A: Baseline clinical demographics collected for case and control patients from the Calgary Adult CF clinic chart records. p values Case Transient Persistent Control Transient Persistent Demographics Total vs (n=21) (n=9) (n=12) (n=42) vs vs Controls Persistent Controls 21.8 19.8 (18.6- 23.4 (18.3- 23.0 (18.4- Age 0.89 0.92 0.9 (18.5-33.1) 26.2) 33.8) 33.2) Sex – Male 12 (57%) 7 (78%) 5 (42%) 24 (57%) 0.18 0.51 1 F508 12 (57%) 5 (56%) 7 (58%) 23 (55%) 1 1 1 homozygous 19.5 (18- 19.5 (18.4- 19.4 (17.5- 21.5 (19.3- BMI, kg/m2 0.89 0.26 0.04 21) 20.7) 23.7) 24) CFRD 1 (5%) 1 (11%) 0 (0%) 7 (17%) 0.43 0.33 0.25 CFLD 5 (24%) 3 (33%) 2 (17%) 10 (24%) 0.61 0.71 1 ABPA 1 (5%) 0 (0) 1 (8%) 3 (7%) 1 1 1 62.5 (50.4- %FEV1 72 (51-88) 72 (40-87) 74.5 (54-93) 0.72 0.45 0.57 84) 91 (76- 91.5 (78.3- 89.5 (72- %FVC 91 (74-106) 0.29 0.43 0.31 106) 109) 102)

Abbreviations: BMI: body mass index; CFRD: CF related diabetes, CFLD: CF liver disease; ABPA: allergic bronchopulmonary aspergillosis; FEV: forced expiratory volume; FVC: forced vital capacity. P values are obtained using 2-sided Fisher's exact or Wilcoxon rank-sum. Data are presented as n (%) or median (interquartile range).

Table 6B: Mycobacteriology subspecies and coinfections collected for case and control patients from the Calgary Adult CF clinic chart records. p values Case Transient Persistent Control Transient Persistent Infections Total vs (n=21) (n=9) (n=12) (n=42) vs vs Controls Persistent Controls M. abscessus 17 6 (67%) 11 (92%) NA 0.27 NA NA abscessus (81%) 2 M. bolletii 1 (11%) 1 (8%) NA 1 NA NA (10%) M. massiliense 0 (0%) 0 (0%) 0 (0%) NA NA NA NA MABC 2 2 (22%) 0 (0%) NA 0.17 NA NA undifferentiated (10%) Pseudomonas 10 5 (56%) 5 (42%) 21 (50%) 0.67 0.75 1 aeruginosa (48%) 12 MSSA 5 (56%) 7 (58%) 22 (52%) 1 0.75 0.79 (57%) Burkholderia 0 (0%) 0 (0%) 0 (0%) 0 (0%) UND UND UND cenocepacia Burkholderia 3 2 (22%) 1 (8%) 1 (2%) 0.55 0.4 0.1 multivorans (14%) 2 MRSA 1 (11%) 1 (8%) 5 (12%) 1 1 1 (10%) Stenotrophomonas 2 0 (0%) 2 (17%) 1 (2%) 0.49 0.12 0.26 species (10%) 2 Aspergillus 0 (0%) 2 (17%) 4 (10%) 0.49 0.6 1 (10%) Achromobacter 0 (0%) 0 (0%) 0 (0%) 3 (7%) UND 1 0.54 species Abbreviations: MSSA: methicillin-sensitive Staphylococcus aureus; MRSA: methicillin-resistant Staphylococcus aureus; UND: undefined. P values are obtained using 2-sided Fisher's exact or Wilcoxon rank-sum. Data are presented as n (%) or median (interquartile range).

Table 6C: On-going therapies and antibiotic collected for case and control patients from the Calgary Adult CF clinic chart records. p values Case Transient Persistent Control Transient Persistent Therapies Total vs (n=21) (n=9) (n=12) (n=42) vs vs Controls Persistent Controls Nutritional 2 (10%) 1 (11%) 1 (8%) 4 (10%) 1 1 1 supplements Pancreatic 38 20 (95%) 8 (89%) 12 (100%) 0.43 0.56 0.66 enzymes (90%) 21 41 CF vitamins 9 (100%) 12 (100%) UND 1 1 (100%) (98%) 39 Vitamin D 19 (90%) 8 (88%) 11 (92%) 1 1 1 (93%) Ranitidine 1 (5%) 1 (11%) 0 (0%) 8 (19%) 0.43 0.18 0.25 18 PPI 13 (62%) 3 (33%) 10 (83%) 0.03 0.02 0.19 (43%) 24 SABA 14 (67%) 4 (44%) 10 (83%) 0.16 0.17 0.59 (57%) 33 LABA 18 (86%) 8 (89%) 10 (83%) 1 1 0.74 (79%) Inhaled 21 11 (52%) 5 (56%) 6 (50%) 1 1 1 corticosteroids (50%) 17 DNase 11 (52%) 4 (44%) 7 (58%) 0.67 0.33 0.43 (40%) Inhaled 15 5 (24%) 2 (22%) 3 (25%) 1 0.73 0.4 hypertonic saline (36%) Inhaled 18 6 (29%) 2 (22%) 4 (33%) 0.66 0.74 0.41 Tobramycin (43%) Inhaled Colistin 1 (5%) 1 (11%) 0 (0%) 0 (0%) 0.43 UND 0.33 Inhaled 3 (14%) 2 (22%) 1 (8%) 3 (7%) 0.55 1 0.39 Aztreonam 16 Azithromycin 4 (19%) 1 (11%) 3 (25%) 0.6 0.51 0.16 (38%) Chronic oxygen 2 (10%) 2 (22%) 0 (0%) 4 (10%) 0.17 0.56 1 Abbreviations: PPI: proton pump inhibitor; SABA: short-acting beta agonist; LABA: long-acting beta agonist; UND: undefined. P values are obtained using 2-sided Fisher's exact or Wilcoxon rank-sum. Data are presented as n (%) or median (interquartile range).

3.2 Pulmonary Function Linear Regression

Extensive patient pulmonary function data was collected for all available clinic visits

associated with sputum samples dating 2 years prior to initial infection as well as 2 years

following (additional dates were collected for patients if insufficient data was present). To

understand rates of change of lung function over time we used linear regression. Linear

regression was extrapolated from the percent forced expiratory volume during the first

second (%FEV1) values and presented in terms of case patients compared to their controls as

displayed in Figure 4. Percent FEV1 was compared for two-years prior to clinical infection

date as well as two years after. Significant differences between any of the groups was not

noted however case patients did have a near significance of p=0.089 suggesting a trend

towards higher lung function change with clinical intervention. Within the case patient

cohort, no significance was noted between the before and after infection samples for transient

and persistent patients as seen in Figure 5.

r s

R V

E a

e n

i L

-10

n n io n io n ct io ct io fe ct fe ct In fe In fe e In e In or r or r ef fte ef fte B A B A

Case Patients Control Patients

Figure 4: Lung function changes over time did not differ between cases and controls before or after infection. Changes in the percent predicted FEV1 values as measured over 2 years before and after initial clinical infection as determined through linear regression. Two-tailed Mann Whitney analysis was conducted to analyze the difference between the cohorts and no significance was noted between any groups.

n o

s r

e e

R V 0

E a

F e

n i

-10

n n io n io n ct io ct io fe ct fe ct In fe In fe e In e In or r or r ef fte ef fte B A B A

Transient Patients Persistent Patients

Figure 5: Lung function changes over time did not differ in transient vs persistent patients before or after infection. Changes in the percent predicted FEV1 values as measured over 2 years before and after initial clinical infection date determined through linear regression. Two- tailed Mann Whitney analysis was conducted to analyze the difference between the cohort and no significant differences were noted between any groups.

3.3 Lung Microbiome Diversity

Two different variations of 16S rDNA gene sequencing analysis pathway was used to

classify the microbiome present in the samples sequenced: the SL1P pathway and DADA2

pathway. Initially, the short-read library 16S rRNA gene sequencing pipeline (SL1P) was

used to sequence the raw reads into operational taxonomic units (OTUs) and in conjunction

with Greengenes2013 furthered to assign genomic taxonomy. The SL1P pathway was

served to combine pre-existing tools into a computational pipeline to produce reproducible

sequence data.140 The SL1P pathway automates accurate processing of paired-end amplicon

reads by first cutting PCR primers using cudadapt, then PANDAseq to align the forward and

reverse reads, quality trim is conducted by sickle, and USEARCH to remove chimeric

sequences.128,141,142 Following, there are a variety of eight different choices to designate

operational taxonomy units (OTU) with CD-HIT v3.1.1 being used for this data set. Next,

RDP classifier is used against the Greengenes 2013 data base to assign taxonomy and are

accurate to the genus level.143,144 OTUs are typically clustered based on a 97% similarity

threshold between sequences.

Divisive Amplicon Denoising Algorithm (DADA2) was the second of the pipelines that

was used for the processing for raw 16S rDNA gene sequences.127 DADA2 models and

corrects Illumina-sequenced amplicon errors and can resolve difference of as little as one

nucleotide to assign sequence variants. Differences in DADA2 from SL1P arises with the

denoising step occurring after merging paired end reads so as to allow for error model to

differentiate sequences into amplicon sequence variants (ASV). DADA2 filters fastq files to

trim sequences to a specified length. Next is dereplication to output unique sequences and

58 their abundances. Followed by the denoising algorithm and removal of chimeras. Finally, forward and reverse reads are merged together if they exactly overlap. Any number of taxonomy assignment databases can be used from here. Specifically, the databases Silva,

RDP, and GTDB were used to assign taxonomy in this study. Silva is the more widely used database for microbiome analysis pertaining to health sciences compared to RDP and is regularly updated with the last update coming in August 2020. GTDB is similarly updated, however, is new and is not as commonly used in the health science and medical field .

Greengenes was last updated in May 2013 and is generally considered outdated although some still use it. OTUs that were observed in the microbiome of patients using the SL1P –

Greengenes 2013 pathway was much lower than was observed in DADA2 – Silva v138 pathway. The SL1P – Greengenes pathway observed 8,455,210 different OTUs whereas the

DADA2 – Silva pathway showed 12,732,176 different ASVs. Figure 6 shows the major phyla that are observed with each of the pathways with a clearly greater observed sequence variations in the DADA2 – Silva pathway.

The most recent update of the Silva database with the species extension shows the presence of the “mycobacterium abscessus” taxonomy although with a mere 1931 reads present for the ASV. Other databases are accurate to the genus level but are only able to decipher the “Order” level of the sequence associated with MABC. Silva v138 and GTDB

2020 displayed Corynebacteriales while RDP v18 displayed the synonym Mycobacteriales for the associated “Order”. Only four of the 21 patients (21 samples) displayed any ASVs through the 16S rRNA microbiome analysis with ASV counts ranging from five to 855. The higher end of this range was associated with a chronically infected patient that displayed a

Ziehl-Neelsen smear 3+ and only took one day to grow with a qPCR quantitative mean of

0.044ng/µl. The reason behind the low number of ASVs observed through the microbiome analysis are unknown but could be attributed to insufficient extraction of MABC 16S rDNA from sputum, difficulty in amplification of MABC 16 rDNA, or improper sequencing analysis ability.

Figure 7 displays the relative abundance of the top 20 ASV compositions for each patient sample through the course of their MABC infection achieved by the DADA2 pipeline using the SILVA v138 taxonomy database as reference. Patients are divided based on persistence of MABC as described by previous criteria established. Most patients displayed a diverse array of bacterial compositions in their microbiome which shift towards a dominating genus during antibiotic exposures but reverted back to the pre-antibiotic norm. Most patients also appeared to have no dominant genus prior to the initial clinical infection.

A B 6×106 6×106

4×106 4×106

T A O

2×106 2×106

0 0 s s a a a s s ia ia ia e e ri ri ri te te r r r t t e e e u e te te te u e t t t c d c c c ic id c c c i i a a a o a a a m ro b b b rm r b b b ir e o o o i te o o o F t e n s F c te in s c t ti u a o t u a ro c F B r c F B P A P A Phyla Phyla

Figure 6: Phyla comparisons for 16S rDNA sequences observed in the microbiome of MABC case patients and their respective controls. (A) Phyla output from the SL1P pathway using Greengenes 2013 database. (B) Phyla output from the DADA2 pathway using Silva v.132 pathway.

Figure 7: Taxonomic composition of the top 20 ASVs at the genus level through infection natural history and matched controls. Patients included in the top panel are diagnosed ultimately with persistent infections and bottom panel are those patients who are defined as having transient infections. Patient sex is designated at the top of each column with either an M or F below which is the study specific patient ID letter. Genus not in the top 20 are not represented in the figure and appear as white above each bar up to a total relative abundance of 1.00. Samples were also classified into infection stages as; A – before infection; B – at infection before treatment; C – induction of therapy; D – consolidation therapy; E – maintenance therapy; F – intensification of therapy; G – clearance; H – Chronic and off-treatment. Red fonts for letters indicate antibiotic therapy during those days. Numbers below each patient’s panel represent the days from initial MABC positive.

3.4 Microbiome structure through the natural history of infection

Relative abundance of the top 20 genera present in the microbiome suggested a consistent

microbiome composition with a dramatic shift during antibiotic treatment and a rapid shift

back to status norm upon its discontinuation. To assess patient microbiome impact on

incident infections, it was important to understand changes in the microbial community

through the natural history of infection. Patient samples were divided into subsets based on

the progression of disease and included the stages of pre-infection (n=38), at infection

diagnosis (n=37), during treatment (n=79), and post-treatment (n=20).

Pre-infection stage accounted for all sputum samples that were collected in the two years

prior to initial clinical identification of MABC growth in sputum samples. These samples

could (unknowingly) be positive prior to their clinical identification due to AFB testing

occurring only once per year. At infection stage encompassed the samples at the time of

initial infection diagnosis and until initiation of treatment (if undertaken). At treatment

samples included all the stages of the treatment course that a patient may be prescribed and

include the initiation, continuation and/or maintenance, and re-intensification stages (during

subsequent exacerbations). Post-treatment samples encompassed for all patient sputum

samples represented after treatment was stopped regardless of whether the patient cleared the

infection or were declared persistent (including any patients on chronic suppressive therapy).

Reductions in alpha diversity were noted through different treatment stages due to

selective antibacterial therapies. Specifically, as seen in figure 8A, significant ODI

differences are noted between pre-infection and at treatment stages (84.5 vs 40.0, p=3.5x10-9)

and between at infection stage and at treatment stage (88.0 vs 40.0, p=9.5x10-10). At

treatment stage also significantly different from the post-treatment stages (66.5 vs 40.0,

63 p=0.002). Similarly, as seen in figure 8B, significant difference in the SDI was also noted for pre-infection samples compared to at treatment samples (2.52 vs 1.72, p=1.1x10-4) as well as at infection samples compared to at treatment samples (2.29 vs 1.72, p=1.2x10-4). A near significant difference between samples at the treatment stage compared to post treatment samples was also observed (1.72 vs 2.39, p=0.057).

Significance was calculated using Wilcoxon rank-sum statistical analysis rarefied with sample variations between groups. These values suggest a decrease in microbial richness and evenness during the treatment phase of disease natural history from the norm.

Subsequent recovery/rebound of the microbiome in the post treatment stage is clearly noted in terms of richness of bacteria present with a trend towards community evenness suggesting that post therapy microbiome regenerates in terms of biomass from the microbes that were not killed off.

Figure 8C explores the beta-diversity through clustering of microbial communities using

PCoA based on Bray-Curtis dissimilarities displayed clustering of microbial communities based on disease progression (p=0.001, R2 = 5.7%, PERMANOVA test). Significant dissimilarities were observed between pre-infection and at infection diagnosis samples compared to during treatment and post treatment samples. Pre-infection samples did not significantly differ in clustering to at infection samples and similarly at treatment samples did not differ significantly in clustering than post-treatment samples.

Figure 8: The natural history of the CF sputum microbiome through the course of MABC infection stages. (A)Observed diversity index changes over time in response to disease stage and antibiotic intervention. Box plots represent median values with top and bottom bars displaying the interquartile range. Wilcoxon rank-sum statistical analysis was conducted between the groups to obtain the p values and rarefied to accommodate variation in sample numbers between groups. (B) Shannon diversity index changes over time in response to disease stage and antibiotic intervention. (C) PCoA of Bray-Curtis dissimilarities analysis of the disease natural history. PERMANOVA analysis was conducted to analyze statistical significance and showed significant clustering (p = 0.001, R2 = 7.3%) based on infection stage.

3.5 Microbiome Comparison of MABC Case Patients to Controls

One of the most important goals of our project was to determine if and how individuals

with MABC infection differed from those without. To ascertain this, we sought to compare

the community structure of case patient’s sputum before clinically recognized as MABC

infection to age, sex, and birth cohort-matched controls. Specifically, microbiome

characteristics prior to and at the time of the MABC incident infection were of interest for

comparison to non-infected controls. Microbial differences in these samples could allude to

why a patient would develop an incident infection in the future. Case patient samples prior

to infection (n=38) and at the time of infection (n=37) were compared to control patient

samples (n=42). Significant beta-diversity clustering was observed between pre-infection

case patients compared to their control counterpart as seen in figure 9B (p=0.048, R2 = 2.5%,

PERMANOVA test) suggesting that fundamental community structures differed between

groups. Samples from patients with MABC were more diverse at baseline than those of

controls. Figure 9A displayed the alpha diversity analysis confirming microbial diversity

differences between the cohorts. In the years prior to an MABC infection, the expectorated

sputum of case patients had a significantly higher ODI richness (84.5 vs 71.0, p=0.037) and

SDI richness and evenness (2.52 vs 2.02, p=0.022) compared to matched controls leading up

to an infection. Further, the presence of individual ASVs were studied (of the top 20 ASVs)

and was observed that, in particular, the relative abundance of Pseudomonas was

significantly higher in the control patients compared to the case patients as seen in Figure 9C.

Significant dissimilarities were equally prevalent in the at infection diagnosis stage

compared to the matched controls. Alpha diversity analysis displayed a significantly higher

ODI and SDI of case samples (ODI: 88 vs 71.0, p=0.014; SDI: 2.28 vs 2.02, p=0.012) as

67 seen in figure 10A and clustering was observed at during this stage prior to the onset of treatment (p=0.004, R2 = 3.6%, PERMANOVA test) as per Figure 10B. Similar to pre- infection ASV differences, the presence of Pseudomonas was observed to be significantly higher in the expectorated sputum of control patients compared to control patients as seen in

Figure 10C (0.0038 v 0.0011, p=0.0067. Ten of the 21 MABC patients were co-infected with P. aeruginosa compared to 21 of 42 controls were co-infected with P. aeruginosa.

Transient and persistent patients each had five chronically infected P. aeruginosa patients.

Figure 9: The microbiome of sputum Pre-infection with MABC differs between cases and control patients. (A) Alpha diversity analysis displaying median ODI and SDI with interquartile range between case and control patients in the samples prior to initial infection of MABC. Wilcoxon rank-sum statistical analysis was conducted between groups to obtain p values rarefied to accommodate sample variations. (B) PCoA of Bray- Curtis dissimilarities analysis of case vs control patient cohorts in the samples prior to infection. PERMANOVA analysis was conducted to analyze statistical significance and showed significant clustering (p = 0.038, R2 = 2.5%). (C) 2-fold log abundance of Pseudomonas genus between the control and case patients. Two-sided Mann-Whitney test was conducted to analyze significance between the groups.

Figure 10: The microbiome of sputum At infection with MABC differs between cases and control patients . (A) Alpha diversity analysis displaying median ODI and SDI with interquartile range between case and control patients in the samples at the time of initial MABC infection prior to treatment onset. Wilcoxon rank- sum statistical analysis was conducted between groups to obtain p values rarefied to accommodate sample variations. (B) PCoA of Bray-Curtis dissimilarities analysis of case vs control patient cohorts in the samples at infection. PERMANOVA analysis was conducted to analyze statistical significance and showed significant clustering (p = 0.004, R2 = 3.6%). (C) Log abundance of Pseudomonas genus between the control and case patients. Two-sided Mann-Whitney test was conducted to analyze significance between the groups.

3.6 Comparing Transient and Persistent Case Patient Sub-cohorts

MABC infection progression towards either a persistent long-term infection that might

have consequence or rapidly be cleared without significant impact was an important factor to

consider - and to understand if the microbiome could play a role in the differentiation of

these outcomes. Case patients were divided into sub-cohorts based on disease progression in

the year following initial clinical infection date and as per ATS/ISDA criteria. Case patients

were identified as transient (n=9) or persistent (n=12) patients based on the aforementioned

criteria of at least 50% positive MABC samples per year with a minimum of 3 samples

tested. When we assessed patient samples based on persistence of MABC infections,

samples associated with transient infections (n=17) did not differ significantly from

persistent infection samples (n=21) based on alpha or beta diversity matrix as seen in Figure

11A and B. Additionally, no differences in the relative abundance of genera present were

noted. Similarly, no differences in alpha diversity were noted for the SDI of cases at the time

of infection comparing transient (n=13) and persistent (n=24) patient samples. When we

assessed ODI we did observe that differences did exist – and that transient patients had less

diversity (p=0.026) but not when the samples were rarefied (p=0.08) to accommodate the

difference in number of samples between transient and persistent patients as can be seen in

figure 12A. Figure 12B displays the insignificant beta diversity differences present between

the sub-cohorts suggesting no evidence of microbiome diversity clustering due to patients

developing a persistent infection compared to those that are transiently infected.

Figure 11: Pre-infection microbiome comparison between transient and persistent patients. (A) Alpha diversity analysis displaying median ODI and SDI with interquartile range between case and control patients in the samples prior to initial infection of MABC. No significance was noted for either the ODI (p=0.15) nor the SDI (p=0.71). Wilcoxon rank-sum statistical analysis was conducted between groups to obtain p values rarefied to accommodate sample variations. (B) PCoA of Bray-Curtis dissimilarities analysis of transient vs persistent patient cohorts in the samples prior to infection. PERMANOVA analysis was conducted to analyze statistical significance and showed no significant clustering (p = 0.50, R2 = 2.5%).

Figure 12: At infection microbiome comparison between transient and persistent patients. (A) Alpha diversity analysis displaying median ODI and SDI with interquartile range between case and control patients in the samples prior to initial infection of MABC. Significance was noted for the ODI (p=0.026) but not for the SDI (p=0.65). Wilcoxon rank-sum statistical analysis was conducted between groups to obtain p values rarefied to accommodate sample variations where the ODI p value was 0.080 and hence insignificant. (B) PCoA of Bray-Curtis dissimilarities analysis of transient vs persistent patient cohorts in the samples prior to infection. PERMANOVA analysis was conducted to analyze statistical significance and showed no significant clustering (p = 0.51, R2 = 2.4%).

3.7 Quantitative PCR

Using real-time qPCR techniques, we sought to find the absolute 16S rDNA quantity

present in each sputum sample was calculated as well as the quantity of P. aeruginosa and

MABC DNA that were presented within these extracted sputum samples.

3.7.1 Absolute bacterial bioburden through the natural history of infection

Quantitative PCR was used to obtain the absolute 16S rDNA concentration present in all

sputum samples. Comparison of the natural history of infection stages were conducted to

help understand differences in total bacterial burden and displayed in Figure13. Total

bacterial load for each sample was measured as total copies of 16S rDNA per milliliter of

sputum. Median 16S rDNA copies/mL of samples were consistent through pre-infection

(1.77 x 1012), at infection (1.47 x 1012), and post treatment (1.58 x 1012) with a slight decline

during at treatment stage (5.84 x 1011). Significant differences, however, were only observed

between during treatment and post treatment stages (p= 0.03) where the total bacterial load

increased at the completion of suppressive antibacterial therapy. Near significance is

observed between pre-infection and at infection stages compared to at treatment stage

(p=0.069 and p=0.079, respectively). Significant differences between pre-infection samples

and post treatment samples was not observed (p=0.48)

Figure 13: Total bacterial bioburden stays relatively stable through the natural history of infection with exception of treatment intervention. Median value for the total bacterial load through the natural infection history stages remained consistent with the following value for each stage present: pre-infection (n=38) was 1.77 x 1012 16S rDNA/mL, at infection (n=37) was 1.47 x 1012 16S rDNA/mL, during treatment (n=79) was 5.84 x 1011, and post treatment (n=20) was 1.58 x 1012). Two-sided Mann-Whitney test was conducted to analyze significance between the groups.

3.7.2 Absolute bacterial bioburden between case and control patients

The bioburden of case patients versus control patients was assessed to see if the potential microbiome differences noted could be as a result of the underlying quantity of total microbiota present in the sputum sample. As seen in figure 14A, there are no significant differences in the median bioburden of case patients as measured by 16S rRNA in all pre- infection samples (1.1 x 1012, n=174) compared to control patients (1.6 x 1012 n=42).

Similarly, control patient bioburden did not differ from the sub-cohort of pre-infection samples (1.8 x 1012, n=38) and at infection samples (1.5 x 1012, n=37) in figure 14B and figure 14C, respectively. Lack of significant difference in absolute bioburden of bacteria in sputum between the cohorts allows for the speculation of any microbiome differences observed above through the alpha and beta diversities to be associated with factors leading towards development of an incident infection with MABC rather than a simple difference in the available microbiota.

A B C 1014 1014 1014

1013 1013 1013

L L

12 12 L 12

10 m 10 10

p p

o 11 o

1011 1011 o 10

D D

r 10 r 10 10

r 10

10 10

g o

9 g 9 9 o L 10

10 o 10

L L

8 108 108 10

7 107 107 10 Case Patients Control Patients Pre-Infection Patients Control Patients At Infection Patients Control Patients

Figure 14: Total bacterial bioburden does not change as a function of infection status. (A) All case patients (n=174) compared to controls (n=42), p = 0.51. (B) Pre-infection case patients (n=38) compared to control patients (n=42), p = 0.93). (C) At infection case patient (n=37) compared to control patients (n=42), p=0.77). Two-tailed Mann-Whitney U analysis was conducted for all comparisons.

3.7.3 Quantifying MABC DNA in Sputum Samples

RT-qPCR was also performed on all sputum samples for the presence of MABC DNA through the detection of the internal transcriber spacer gene using a probe. All extracted sputum sample (174 case, and 42 control samples) were subjected to the test. In case patient individuals, we noted all patients had at least one MABC DNA as detected by qPCR. We noted one control positive for MABC DNA in one time point – but at low levels.

Figure 15 displays the mean MABC DNA concentration through the natural history progression of disease. Amongst infected patients, the presence of MABC is noted highest at the time of initial infection diagnosis and is significantly higher than pre-infection, during treatment and post treatment. This holds true as samples during the clinical at infection stage would contain the most MABC DNA. Prior to a clinical positive stage, fewer samples would contain MABC DNA (as these cases were clinically diagnosed as yet) and those that do would not have concentrations as high as seen during at infection stages. During treatment,

MABC DNA concentration would be expected to reduce due to potent antibacterial selective pressure. Post treatment samples would only contain MABC DNA if treatment was unsuccessful, and patients develop chronic infections.

Figure 15: MABC bioburden changes through its Natural History of Infection. Natural infection history as a function of mean abundance of MABC DNA present per microliter. Significant values between stages were calculated using two-tailed Mann-Whitney U test and only significant comparisons are displayed. Mean values with interquartile range is displayed as using the median is not feasible for a logarithmic figure where many values are zero resulting in the median being zero and not graphable.

With all but one MABC qPCR signals being associated with case samples, there was an obvious significance in the median MABC abundance for case patient prior to and at the time of an infection compared to controls (p =<0.0001). MABC qPCR values were compared with clinical values for MABC grown on Middlebrook 7H10 (ProvLab, Alberta, Canada) to assess whether a correlation could exist with physical examination of sputum (via Ziel-

Neelson (Z-N) stain to assess burden of organisms as measured by smear negative, or positive at 1+, 2+ or 3+ - as reported by the clinical microbiology laboratory) and days taken to form colonies as seen in figures 16 and 17, respectively. A positive correlation was noted between MABC DNA concentration compared to increasing burden of organisms as indicated by Z-N staining of sputum. Most samples were smear negative which correlates to a low concentration of acid-fast bacilli (AFB) present. Increasing number of smear positive denotes increasing infectiousness of patients. Samples with a 3+ smear positivity AFB test have a significantly higher MABC DNA concentration than smear negative and 1+ smear positive samples. Similarly, samples that have a higher MABC DNA concentration correlated with less time to grow colonies (and therefore a higher initial inoculum). A significant difference is noted between samples that grew within 1-3 days compared to those that took longer than 7 days to grow.

Figure 16: Z-N Smear positivity correlates with absolute abundance of MABC. Median MABC DNA concentration correlated with Z-N staining performed in the clinical microbiology laboratory. Smear negative (n=54) had a median value of 1.64 x 10-5, Smear 1 (n=8) had a median value of 6.57 x 10-5, Smear 3 (n=3) had a median value of 1.57 x 10-3. The top and bottom of the box plot represent the interquartile ranges. Significant values were calculated using a two-tailed Mann-Whitney U test. Each patient (letter) is represented by a different colour.

Figure 17: A higher bioburden of MABC DNA correlated with a shorter time for MABC cultures to grow. Median MABC DNA concentration correlated with days to form colony data. Days 1 to 3 (n=9) had a median value of 4.38 x 10-4, days 4 to 6 (n=24) had a median value of 8.87 x 10-5, days greater than 7 (n=12) had a median value of 2.74 x 10-5. The top and bottom of the box plot represent the interquartile ranges. Significant value was calculated using a two-tailed Mann-Whitney U test. Each patient (letter) is represented by a different colour.

3.7.4 Quantifying P. aeruginosa DNA in Sputum Samples

Differences in the relative abundance of Pseudomonas was observed in the patient samples in the years prior to initial infection as well as samples at the time of infection as compared to control patient samples. Further RT-qPCR testing was performed with primer specific to P. aeruginosa to confirm the differences between the cohorts observed through the microbiome analysis as well as quantify to the species level, which is not seen through the 16S rRNA sequencing analysis. Specifically, the at infection relative abundance difference of Pseudomonas observed through the 16S rRNA sequencing analysis displayed approximately three and a half fold higher in the control patients compared to the case patients (0.003826 ng/µL vs 0.001065 ng/µL, p = 0.0067). Similar differences are observed for at infection samples compared to controls for RT-qPCR analysis as seen in figure 18.

Case patients at the time of initial infection had significantly lower P. aeruginosa abundance

(4.24 x 10-6, n=37) compared to the control patients (1.04 x 10-4, n=42), p= 0.0006. This assessment allows to infer a potential negative correlation of P. aeruginosa to infections with

MABC prior to the occurrence of an infection and confirms our results from the 16S rRNA microbiome analysis.

100

10-1

10-2

10-3

)

µ /

n -4

10(

A N

D 10-5

-6 10

10-7

10-8 Case Patients Control Patients

Figure 18: Median P. aeruginosa DNA concentration was higher in control samples compared to at infection case samples. Median P. aeruginosa DNA concentration in case patients at the time of infection was 4.24 x 10-6 ng/µL (n=37) compared to control patients at 1.05 x 10-4 ng/µL (n=42). Two-tailed Mann-Whitney U statistical analysis revealed a significant p value of 0.0006.

3.8 P. aeruginosa Inhibition Assay

Several reasons could influence this correlation, one of which could be the presence of P.

aeruginosa providing potential shielding from MABC colonization through secreted

inhibitory factors. As this could be assessed with relative ease – we sought to determine if in

fact this might be the case through an inhibition assay. Figure 19A and B shows results from

the experiment designed to assess the impact of the extracted supernatant of the lab strain of

P. aeruginosa (PA01) as well as the Prairie Epidemic Strain (PES) derived from individuals

with CF in separate dilutions of 1:5 and 1:20 on the growth of MABC over 5 days.

Differences noted between the control samples compared to the PA01 or PES strains for

either dilution factors were not significant. All of the groups followed a logarithmic growth

curve reaching their maximum by day 3. Although no significant differences were noted

between the variables, the control population had slightly higher colony forming units per

milliliter than its P. aeruginosa strain counterparts. This, however, was not deemed enough

to formulate any correlation regarding P. aeruginosa growth medium supernatant containing

secreted molecules with a mechanism to inhibit MABC growth and thus, the hypothesis for

was refuted.

A B

Figure 19: MABC growth in Middlebrook 7H9 broth infused with P. aeruginosa PA01 and PES supernatant. (A) Dilution at a ratio of 1:5. (B) Dilution at a ratio of 1:20. Significance was not noted between the variants based on Kruskal-Wallis test for either dilutions. Negative control flasks reported no growths in any replicates. A logarithmic growth trend is noted for all variables with non-supernatant controls having slightly higher CFU/mL at the end of day 5.

Chapter Four: Discussion

4.1 MABC Infections are a Major Cause for Concern in CF Patients

Increasing incidence and prevalence of NTM infections within CF populations have been noted in the last 15 years with a considerable portion infected with MABC. Factors such as improved laboratory practice, increased patient survival age, increased environmental exposure, numerous antibiotic exposure, as well as person-to-person transmission have been posited as being contributory.145–148 Patients with prolonged MABC infection have been shown to have a greater rate of decline in pulmonary function (%FEV1) compared to the classical pathogen P. aeruginosa

(-2.22% per year vs. -0.95% per year).59 MABC infected patients are also ineligible for lung transplant in many centers due to poor clinical outcomes post-transplant.149–151 MABC infection also incurs a prolonged treatment course with several different classes of antibiotics due to several inherent antimicrobial resistance mechanisms present.152,153 These reasons present an utmost importance towards understanding factors that could predispose CF patients to MABC infections.

4.2 Patients with MABC Present with a Lower BMI than Controls

Twenty-one case patients for the microbiome study were selected from the Calgary Adult

CF Clinic that presented with a positive MABC culture meeting the ATS/IDSA criteria.

Control patients (n=42) on the contrary did not have any history of MABC as identified

through the clinical reviews of medical records and clinical microbiology testing done

annually. Comparisons of the patient demographics between the case and control cohorts

yielded no significant differences in age, genetics, or lung function values at the time of

initial MABC infection.68,154 These factors could present a pre-emptive bias for any microbial

difference that would be noted between the case patient and control patient cohorts.

Significant demographic difference between the cohort was only observed for patient

BMI where control patients presented with a higher BMI compared to their case counterparts

(21.5 kg/m2 vs 19.5 kg/m2, p=0.04). Direct associations between diet and gut microbiome

dysbiosis have been reported on previously but have only had indirect implications to the

lung microbiome.155–158 Multicenter studies conducted on the prevalence of NTM have not

found BMI to be a significant factor dissociating between patients that develop MABC

infections compared to those that do not develop an infection.55 These significance were

corrected for using Benjamini-Hochberg correction for multiple comparisons however these

may still only be due to chance alone. Further multicenter studies focused on MABC

correlation with BMI could allow for a better understanding of any possible correlations.

4.3 Persistent MABC Patients had Significantly Higher PPI use Compared to Transient

Patients or Controls.

Case patients that developed a persistent infection was defined as 50% or greater positive

mycobacteriology tests with a minimum of three tests in the year following initial infection.

This guideline is an adaptation from the Leeds criteria used for P. aeruginosa to assess

persistent infections as a universally applied criteria for NTM infections does not exist for

CF.159 Transiently infected and persistently infected patients did not significantly differ by

any demographic parameters nor with their co-infections. Only the therapeutic use of PPI

was observed to be significantly higher in persistent patients compared to transient patients

as well as compared to control patients. PPI is used in the management of the vastly

prevalent gastroesophageal reflux disease (GERD) observed in CF patients.160 Studies have

correlated GERD in CF to poor pulmonary outcomes and an increased frequency of

hospitalization.161–163 Aggressive PPI therapy recommended to manage GERD have been

linked to the development of community-acquired pneumonia in CF patients due to

inhibition of several leukocyte functions.164,165 The association for persistent MABC patients

having higher prescribed PPI is unclear but could be a consequence of hindered immune cell

response leading to easier persistent colonization of MABC. This factor, although significant,

could be a result of chance alone and should be subject to additional research with a larger

cohort size in a multicenter study.

4.4 Patients with MABC did not Differ in their Microbiome Diversity with New Infection

Patients with MABC infection had their sputum samples characterized and subdivided

into categories of pre-infection, at infection, at treatment, and post treatment. Lung

microbiome richness and evenness was assessed using alpha-diversity measures which

showed no significant changes between samples prior to an infection compared to samples at

the time of infection. Significant beta-diversity clustering was also not observed between

samples prior to an infection compared to samples at the time of infection. This suggests that

community diversity prior to an infection may not suggest as a conclusive biomarker for

patients that will develop an incident MABC infection. These findings are consistent with

previous literature which found no significant changes in total bacterial density with

exacerbations.75,166–168

With antibiotic intervention, however, a reduction in the microbiome diversity is noted in

the samples associated with during treatment stages compared to samples from pre-infection

and at infection stages. These differences are seen in the alpha-diversity measures within

samples as well as in the beta-diversity clustering between samples. Subsequent recovery of

the microbiome towards pre-treatment diversity is also observed. These trends are

previously described in several studies that found acute antibiotics to cause an immediate

albeit temporary change in the CF microbiome structure that quickly return to baseline

community structure post therapy.83,154

Quantitative PCR analysis of the natural history of infection was done to identify the total

bacterial bioburden present in the subcategories. Here we found no significant differences in

the pre-infection, at infection, and at treatment samples for total bacterial bioburden and a

slight significance post treatment. This lack of variability in the total bacterial bioburden

91 allows for any bacterial diversity observed to be attribuetd with changes in the relative abundance of microbial constituents rather than an overall increase in the amount of bacteria present.

4.5 Patients with MABC had significantly different microbiomes compared to control patients.

The lung microbial community structure of case patients was compared to age, sex, and

birth cohort-matched control, where significant differences in microbiome diversity was

observed. Specifically, case patient samples prior to an infection as well as at the time of

infection presented significantly different microbiome structure from control patients. Case

samples presented significantly higher microbiome richness and evenness than control

samples. Further, significant beta-diversity clustering between samples was also observed

for case patients compared to control patients. This suggests an aversion from the classical

definition of lower diversity relaying to increased disease by a dominating bacteria.74–76

Quantitative PCR analysis comparing the total bacterial bioburden as measured by 16S

rRNA between case patients and controls displayed no significant differences between the

cohorts allowing for associations in diversity analysis to be credited to the microbiome alone.

Quantitative PCR analysis of all case and control samples was also conducted to assess

for burden of MABC. Twenty seven of 37 samples associated with the at “infection stage”

presented the highest mean quantity of MABC signals as expected. However, 15 of 38

samples associated with pre-infection sample (as per clinical classification) also had a

positive MABC signal - suggesting missed positives through the traditional AFB method of

testing. Accordingly, our strict definitions of pre and at-infection based on Z-N stain and

culture were insensitive to identifying paitnets with new infection. Twenty nine of 79 at

treatment samples continue to present a MABC signal through the treatment stages and two

samples of 20 continue to present a MABC signal post treatment. Successful qPCR analysis

of CF sputum for the presence of MABC has not been explored previously and could be

93 extremely valuable to allow for more frequent testing of MABC leading to earlier diagnosis and immediate treatment.

Significant differences in the individual ASVs present between control and case samples prior to an infection and at the time of infection showcased higher Pseudomonas in control patients compared to case patients. This is consistent with suggesting that non-classical pathogens occupy airway in patients that are not already dominated by a classical pathogen.

These factors are consistent with a previously described multi-center study on the prevalence of NTM in CF by Olivier et al. (2003). Olivier et al. observed patient that were positive for

NTM infections had significantly lower rates of P. aeruginosa chronic infection compared to patients negative for NTM.169 Similarly, Esther Jr. et al (2010) also found a significant negative association between NTM and P. aeruginosa.170

We confirmed these differences in P. aeruginosa between the cohorts using qPCR – but not with culture status alone. Similarly, to the microbiome outcome, we found control patients with a significantly higher P. aeruginosa quantity compared to case patients at the time of infection further suggesting a potential shielding of MABC infection by a pre- existing P. aeruginosa infection. This may suggest that it is not the mere presence or absence of P. aeruginosa that matters but rather the amount.

We sought to explore if P. aeruginosa inhibition of MABC was occurring through a secreted inhibitory factor, which we found to not be the case in our experiment. These factors include phenazines, proteases, elastases, hemolysin, and rhamnolipids that are excreted by P. aeruginosa to contribute to their virulence.171 Other interaction factors between P. aeruginosa and MABC could play a role in suppressing the MABC growth in an environment pre-infected with P. aeruginosa including the ability of Pseudomonas species to

94 simply overgrow NTM in small fractions of AFB cultures.172,173 Other inhibition methods employed by P. aeruginosa include the formation of biofilms under nutrient limiting conditions.174,175 We were able to rule out any excreted molecules presented that may persist long term in solution to inhibit growth of MABC. The use of a polycarbonate membrane could be useful to assess indirect inhibition between P. aeruginosa and MABC where a constant release of exoproducts could be assessed.

CHAPTER 5: Summary

5.1 Limitations Several limiting factors in this study are notable including the relatively small number of

patients included due to MABC prevalence being low. Additionally, the research is only a

single-center study thus making generalizations and associations difficult. Multicenter

studies in addition to including a range of geographic and treatment practices will enable

greater number of patients and samples to be accrued in order to assess more subtle

differences. Furthermore, sequencing the V3 region of the 16S rRNA region can only give

certain accuracy up to the genus level leaving out associations towards species changes in the

cohort. Additionally, extracted sputum would contain DNA from live and dead rememnants

of the microbiome and could influence the sequencing accuracy. Metagenomic sequencing is

an approach that can move beyond this limitation, however, has not been widely deployed in

CF owing to the extent of oropharyngeal contamination of sputum samples by the oral

flora.176

As well, sputum samples from patient are dynamic, changing day to day, and largely

contain organisms from the upper respiratory tract, although studies of induced sputum

samples have shown comparable bacterial correlations to BAL.177

5.2 Future Directions

Further exploration of the microbiome association with MABC infection should be done

in a multicentered study involving several different geographic locations to best understand

the surrounding epidemiology associated with the infection. Additionally, whole genome

sequencing of MABC could be done within our study cohort to identify potential

transmission occurrences.

Further testing of the P. aeruginosa inhibition assay could be done with a possible testing

with a semi-permeable polycarbonate membrane dividing P. aeruginosa from MABC could

allow us to run a more realistic indirect competitive assay which would account for quorum

sensing or direct competition for resources and could provide possible reasons to the negative

correlation observed between MABC and P. aeruginosa. Additionally, a transcriptomic

study involving P. aeruginosa and MABC co-cultures to assess any molecular mechanisms

between the interactions.

5.3 Study Summary and Conclusion

Significant differences were observed in the diversity of the microbial community of CF

sputum in patients that develop MABC infections compared to non-infected controls. These

observations are seen in the samples leading up to the initial infection date as well as on the

date itself. Patients who acquire MABC have a distinctively different community structure

with a relative defiance of Pseudomonas genus suggesting a negative correlation with

MABC infection itself and possibly even a protective role from infections with MABC.

Between persistent and transient-infected patients microbiome community structure was not

sufficient different to distinguish these two groups. These suggest further investigation

should be conducted in a multi-center study involving more patients to create a more

solidified association.

References

1. Cutting GR. Cystic fibrosis genetics: from molecular understanding to clinical application.

Nat Rev Genet. 2015;16(1):45-56. doi:10.1038/nrg3849.Cystic

2. Cystic Fibrosis Canada. The Canadian CF Registry 2016 Annual Data Report. 2017:44.

http://www.cysticfibrosis.ca/uploads/2016 Registry Annual Data Report.pdf.

3. Keogh RH, Szczesniak R, Taylor-Robinson D, Bilton D. Up-to-date and projected

estimates of survival for people with cystic fibrosis using baseline characteristics: A

longitudinal study using UK patient registry data. J Cyst Fibros. 2018;17(2):218-227.

doi:10.1016/j.jcf.2017.11.019

4. Rowntree RK, Harris A. The phenotypic consequences of CFTR mutations. Ann Hum

Genet. 2003. doi:10.1046/j.1469-1809.2003.00028.x

5. Derichs N. Targeting a genetic defect: Cystic fibrosis transmembrane conductance

regulator modulators in cystic fibrosis. Eur Respir Rev. 2013;22(127):58-65.

doi:10.1183/09059180.00008412

6. Bronstein MN, Sokol RJ, Abman SH, et al. Pancreatic insufficiency, growth, and nutrition

in infants identified by newborn screening as having cystic fibrosis. J Pediatr.

1992;120(4 Pt 1):533-540. doi:10.1016/s0022-3476(05)82478-3

7. De Boeck K, Weren M, Proesmans M, Kerem E. Pancreatitis among patients with cystic

fibrosis: correlation with pancreatic status and genotype. Pediatrics. 2005;115(4):e463-9.

doi:10.1542/peds.2004-1764

8. Kelly A, Moran A. Update on cystic fibrosis-related diabetes. J Cyst Fibros Off J Eur

Cyst Fibros Soc. 2013;12(4):318-331. doi:10.1016/j.jcf.2013.02.008

9. Colombo C, Ellemunter H, Houwen R, Munck A, Taylor C, Wilschanski M. Guidelines

100

for the diagnosis and management of distal intestinal obstruction syndrome in cystic

fibrosis patients. J Cyst Fibros Off J Eur Cyst Fibros Soc. 2011;10 Suppl 2:S24-8.

doi:10.1016/S1569-1993(11)60005-2

10. El-Chammas KI, Rumman N, Goh VL, Quintero D, Goday PS. Rectal prolapse and cystic

fibrosis. J Pediatr Gastroenterol Nutr. 2015;60(1):110-112.

doi:10.1097/MPG.0000000000000546

11. Koh C, Sakiani S, Surana P, et al. Adult-onset cystic fibrosis liver disease: Diagnosis and

characterization of an underappreciated entity. Hepatology. 2017;66(2):591-601.

doi:10.1002/hep.29217

12. Williams SG, Westaby D, Tanner MS, Mowat AP. Liver and biliary problems in cystic

fibrosis. Br Med Bull. 1992;48(4):877-892. doi:10.1093/oxfordjournals.bmb.a072583

13. Taussig LM, Lobeck CC, di Sant’Agnese PA, Ackerman DR, Kattwinkel J. Fertility in

males with cystic fibrosis. N Engl J Med. 1972;287(12):586-589.

doi:10.1056/NEJM197209212871204

14. Sueblinvong V, Whittaker LA. Fertility and pregnancy: common concerns of the aging

cystic fibrosis population. Clin Chest Med. 2007;28(2):433-443.

doi:10.1016/j.ccm.2007.02.009

15. Gibson RL, Burns JL, Ramsey BW. Pathophysiology and management of pulmonary

infections in cystic fibrosis. Am J Respir Crit Care Med. 2003;168(8):918-951.

doi:10.1164/rccm.200304-505SO

16. Yung MW, Gould J, Upton GJG. Nasal polyposis in children with cystic fibrosis: a long-

term follow-up study. Ann Otol Rhinol Laryngol. 2002;111(12 Pt 1):1081-1086.

doi:10.1177/000348940211101204

101

17. Wine JJ. The genesis of cystic fibrosis lung disease. J Clin Invest. 1999;103(3):309-312.

doi:10.1172/JCI6222

18. Fanen P, Wohlhuter-Haddad A, Hinzpeter A. Genetics of cystic fibrosis: CFTR mutation

classifications toward genotype-based CF therapies. Int J Biochem Cell Biol. 2014;52:94-

102. doi:10.1016/j.biocel.2014.02.023

19. Tsui LC. The cystic fibrosis transmembrane conductance regulator gene. Am J Respir Crit

Care Med. 1995;151(3 Pt 2):S47-53. doi:10.1164/ajrccm/151.3_Pt_2.S47

20. Tsui L-C, Dorfman R. The cystic fibrosis gene: a molecular genetic perspective. Cold

Spring Harb Perspect Med. 2013;3(2):a009472. doi:10.1101/cshperspect.a009472

21. Dechecchi MC, Tamanini A, Cabrini G. Molecular basis of cystic fibrosis: from bench to

bedside. Ann Transl Med. 2018;6(17):334. doi:10.21037/atm.2018.06.48

22. Moskowitz SM, Chmiel JF, Sternen DL, et al. Clinical practice and genetic counseling for

cystic fibrosis and CFTR-related disorders. Genet Med. 2008;10(12):851-868.

doi:10.1097/GIM.0b013e31818e55a2

23. Zielenski J, Fujiwara TM, Markiewicz D, et al. Identification of the M1101K mutation in

the cystic fibrosis transmembrane conductance regulator (CFTR) gene and complete

detection of cystic fibrosis mutations in the Hutterite population. Am J Hum Genet.

1993;52(3):609-615.

24. Zegarra-Moran O, Galietta LJ V. CFTR pharmacology. Cell Mol Life Sci. 2017;74(1):117-

128. doi:10.1007/s00018-016-2392-x

25. Wang Y, Wrennall JA, Cai Z, Li H, Sheppard DN. Understanding how cystic fibrosis

mutations disrupt CFTR function: From single molecules to animal models. Int J Biochem

Cell Biol. 2014;52:47-57. doi:https://doi.org/10.1016/j.biocel.2014.04.001

102

26. Meng X, Clews J, Kargas V, Wang X, Ford RC. The cystic fibrosis transmembrane

conductance regulator (CFTR) and its stability. Cell Mol Life Sci. 2017;74(1):23-38.

doi:10.1007/s00018-016-2386-8

27. Elborn S, Vallieres E. Cystic fibrosis gene mutations: evaluation and assessment of

disease severity. Adv Genomics Genet. 2014;4:161. doi:10.2147/AGG.S53768

28. Boeck D, Amaral M. Classifi cation of CFTR mutation classes. WwwThelancetCom.

2016;4(August):37-38. doi:10.1016/S2213-2600(16)30188-6

29. Lopes-Pacheco M. CFTR Modulators: Shedding Light on Precision Medicine for Cystic

Fibrosis. Front Pharmacol. 2016;7:275. doi:10.3389/fphar.2016.00275

30. Welsh MJ, Smith AE. Molecular mechanisms of CFTR chloride channel dysfunction in

cystic fibrosis. Cell. 1993;73(7):1251-1254. doi:10.1016/0092-8674(93)90353-R

31. Elborn JS. Cystic fibrosis. Lancet. 2016;388(10059):2519-2531. doi:10.1016/S0140-

6736(16)00576-6

32. McKenney D, Pouliot KL, Wang Y, et al. Broadly protective vaccine for Staphylococcus

aureus based on an in vivo-expressed antigen. Science (80- ). 1999;284(5419):1523-1527.

doi:10.1126/science.284.5419.1523

33. Tiddens HA, Paré PD, Hogg JC, Hop WC, Lambert R, de Jongste JC. Cartilaginous

airway dimensions and airflow obstruction in human lungs. Am J Respir Crit Care Med.

1995;152(1):260-266. doi:10.1164/ajrccm.152.1.7599833

34. Bennett WD, Olivier KN, Zeman KL, Hohneker KW, Boucher RC, Knowles MR. Effect

of uridine 5’-triphosphate plus amiloride on mucociliary clearance in adult cystic fibrosis.

Am J Respir Crit Care Med. 1996;153(6):1796-1801. doi:10.1164/ajrccm.153.6.8665037

35. Donaldson SH, Bennett WD, Zeman KL, Knowles MR, Tarran R, Boucher RC. Mucus

103

Clearance and Lung Function in Cystic Fibrosis with Hypertonic Saline. N Engl J Med.

2006;354(3):241-250. doi:10.1056/NEJMoa043891

36. Worlitzsch D, Tarran R, Ulrich M, et al. Effects of reduced mucus oxygen concentration

in airway Pseudomonas infections of cystic fibrosis patients. J Clin Invest.

2002;109(3):317-325. doi:10.1172/JCI0213870

37. Gibson GJ, Loddenkemper R, Lundbäck B, Sibille Y. Respiratory health and disease in

Europe: The new European Lung White Book. Eur Respir J. 2013;42(3):559-563.

doi:10.1183/09031936.00105513

38. Pasteur MC, Bilton D, Hill AT. British Thoracic Society guideline for non-

CFbronchiectasis. Thorax. 2010;65(Suppl 1):i1-i58. doi:10.1136/thx.2010.136119

39. Redondo M, Keyt HRD, Chalmers JD. Educational aims. Breathe. 2016;12(3):222-235.

doi:10.4324/9780203390740_chapter_1

40. Chalmers JD, Smith MP, McHugh BJ, Doherty C, Govan JR, Hill AT. Short- and long-

term antibiotic treatment reduces airway and systemic inflammation in non-cystic fibrosis

bronchiectasis. Am J Respir Crit Care Med. 2012;186(7):657-665.

doi:10.1164/rccm.201203-0487OC

41. Cole PJ. Inflammation: a two-edged sword--the model of bronchiectasis. Eur J Respir Dis

Suppl. 1986;147:6-15.

42. Chalmers JD, Hill AT. Mechanisms of immune dysfunction and bacterial persistence in

non-cystic fibrosis bronchiectasis. Mol Immunol. 2013;55(1):27-34.

doi:10.1016/J.MOLIMM.2012.09.011

43. LiPuma JJ. The changing microbial epidemiology in cystic fibrosis. Clin Microbiol Rev.

2010;23(2):299-323. doi:10.1128/CMR.00068-09

104

44. Mahenthiralingam E. Emerging cystic fibrosis pathogens and the microbiome. Paediatr

Respir Rev. 2014;15(S1):13-15. doi:10.1016/j.prrv.2014.04.006

45. Rosenfeld M, Gibson RL, McNamara S, et al. Early pulmonary infection, inflammation,

and clinical outcomes in infants with cystic fibrosis. Pediatr Pulmonol. 2001;32(5):356-

366. doi:10.1002/ppul.1144

46. Razvi S, Quittell L, Sewall A, Quinton H, Marshall B, Saiman L. Respiratory

microbiology of patients with cystic fibrosis in the United States, 1995 to 2005. Chest.

2009;136(6):1554-1560. doi:10.1378/chest.09-0132

47. Tümmler B, Koopmann U, Grothues D, Weissbrodt H, Steinkamp G, von der Hardt H.

Nosocomial acquisition of Pseudomonas aeruginosa by cystic fibrosis patients. J Clin

Microbiol. 1991;29(6):1265-1267. doi:10.1128/JCM.29.6.1265-1267.1991

48. Govan JRW, Brown AR, Jones AM. Evolving epidemiology of Pseudomonas aeruginosa

and the Burkholderia cepacia complex in cystic fibrosis lung infection. Future Microbiol.

2007;2(2):153-164. doi:10.2217/17460913.2.2.153

49. Krzewinski JW, Nguyen CD, Foster JM, Burns JL. Use of random amplified polymorphic

DNA PCR to examine epidemiology of Stenotrophomonas maltophilia and

Achromobacter (Alcaligenes) xylosoxidans from patients with cystic fibrosis. J Clin

Microbiol. 2001;39(10):3597-3602. doi:10.1128/JCM.39.10.3597-3602.2001

50. Elting LS, Khardori N, Bodey GP, Fainstein V. Nosocomial infection caused by

Xanthomonas maltophilia: a case-control study of predisposing factors. Infect Control

Hosp Epidemiol. 1990;11(3):134-138. doi:10.1086/646136

51. Murphy TF, Sethi S, Klingman KL, Brueggemann AB, Doern G V. Simultaneous

respiratory tract colonization by multiple strains of nontypeable haemophilus influenzae

105

in chronic obstructive pulmonary disease: implications for antibiotic therapy. J Infect Dis.

1999;180(2):404-409. doi:10.1086/314870

52. Longo D, Jameson J, Kasper D, Fauci A, Loscalzo J, Hauser S. Harrison’s Principles of

Internal Medicine. McGraw-Hill Education / Medical; 2015.

53. Nadesalingam K, Conway SP, Denton M. Risk factors for acquisition of methicillin-

resistant Staphylococcus aureus (MRSA) by patients with cystic fibrosis. J Cyst Fibros

Off J Eur Cyst Fibros Soc. 2005;4(1):49-52. doi:10.1016/j.jcf.2004.09.002

54. Cox DW, Kelly C, Rush R, O’Sullivan N, Canny G, Linnane B. The impact of MRSA

infection in the airways of children with cystic fibrosis; a case-control study. Ir Med J.

2011;104(10):305-308.

55. Olivier KN, Weber DJ, Wallace RJJ, et al. Nontuberculous mycobacteria. I: multicenter

prevalence study in cystic fibrosis. Am J Respir Crit Care Med. 2003;167(6):828-834.

doi:10.1164/rccm.200207-678OC

56. Marshall B, Elbert A, Petren K, et al. Cystic fibrosis foundation patient registry 2014

annual data report. Bethesda, Md Cyst Fibros Found. 2015;1:49.

57. Park IK, Olivier KN. Nontuberculous mycobacteria in cystic fibrosis and non-cystic

fibrosis bronchiectasis. Semin Respir Crit Care Med. 2015;36(2):217-224. doi:10.1055/s-

0035-1546751

58. Samra Z, Kaufman L, Pitlik S, Shalit I, Bishara J. Emergence of Mycobacterium simiae in

respiratory specimens. Scand J Infect Dis. 2005;37(11-12):838-841.

doi:10.1080/00365540500264712

59. Qvist T, Taylor-Robinson D, Waldmann E, et al. Comparing the harmful effects of

nontuberculous mycobacteria and Gram negative bacteria on lung function in patients

106

with cystic fibrosis. J Cyst Fibros. 2016;15(3):380-385. doi:10.1016/j.jcf.2015.09.007

60. Firmida MC, Pereira RH V, Silva EASR, Marques EA, Lopes AJ. Clinical impact of

Achromobacter xylosoxidans colonization/infection in patients with cystic fibrosis.

Brazilian J Med Biol Res = Rev Bras Pesqui medicas e Biol. 2016;49(4):e5097.

doi:10.1590/1414-431X20155097

61. Behjati S, Tarpey PS. What is next generation sequencing? Arch Dis Child Educ Pract Ed.

2013;98(6):236-238. doi:10.1136/archdischild-2013-304340

62. Moffatt MF, Cookson WO. The lung microbiome in health and disease. Clin Med.

2017;17(6):525-529. doi:10.7861/clinmedicine.17-6-525

63. Chang D, Dela Cruz CS, Sharma L. Challenges in understanding lung microbiome: It is

NOT like the gut microbiome. Respirology. 2020;25(3):244-245. doi:10.1111/resp.13759

64. Nguyen LDN, Viscogliosi E, Delhaes L. The lung mycobiome: an emerging field of the

human respiratory microbiome. Front Microbiol. 2015;6:89.

doi:10.3389/fmicb.2015.00089

65. Huang YJ, Charlson ES, Collman RG, Colombini-Hatch S, Martinez FD, Senior RM. The

role of the lung microbiome in health and disease. A National Heart, Lung, and Blood

Institute workshop report. Am J Respir Crit Care Med. 2013;187(12):1382-1387.

doi:10.1164/rccm.201303-0488WS

66. LiPuma J. The new microbiology of cystic fibrosis: it takes a community. Thorax.

2012;67(10):851-852. doi:10.1136/thoraxjnl-2012-202018

67. Surette MG. The cystic fibrosis lung microbiome. Ann Am Thorac Soc. 2014;11 Suppl

1:S61-5. doi:10.1513/AnnalsATS.201306-159MG

68. Cox MJ, Allgaier M, Taylor B, et al. Airway microbiota and pathogen abundance in age-

107

stratified cystic fibrosis patients. PLoS One. 2010;5(6):e11044.

doi:10.1371/journal.pone.0011044

69. Mahboubi MA, Carmody LA, Foster BK, Kalikin LM, VanDevanter DR, LiPuma JJ.

Culture-Based and Culture-Independent Bacteriologic Analysis of Cystic Fibrosis

Respiratory Specimens. J Clin Microbiol. 2016;54(3):613-619. doi:10.1128/JCM.02299-

70. Muhlebach MS, Zorn BT, Esther CR, et al. Initial acquisition and succession of the cystic

fibrosis lung microbiome is associated with disease progression in infants and preschool

children. PLoS Pathog. 2018;14(1):e1006798. doi:10.1371/journal.ppat.1006798

71. Huang YJ, LiPuma JJ. The Microbiome in Cystic Fibrosis. Clin Chest Med.

2016;37(1):59-67. doi:10.1016/j.ccm.2015.10.003

72. Rogers GB, Skelton S, Serisier DJ, van der Gast CJ, Bruce KD. Determining cystic

fibrosis-affected lung microbiology: comparison of spontaneous and serially induced

sputum samples by use of terminal restriction fragment length polymorphism profiling. J

Clin Microbiol. 2010;48(1):78-86. doi:10.1128/JCM.01324-09

73. Sibley CD, Grinwis ME, Field TR, et al. Culture Enriched Molecular Profiling of the

Cystic Fibrosis Airway Microbiome. PLoS One. 2011;6(7):e22702.

https://doi.org/10.1371/journal.pone.0022702.

74. Cuthbertson L, Walker AW, Oliver AE, et al. Lung function and microbiota diversity in

cystic fibrosis. Microbiome. 2020;8(1):45. doi:10.1186/s40168-020-00810-3

75. Carmody LA, Zhao J, Schloss PD, et al. Changes in cystic fibrosis airway microbiota at

pulmonary exacerbation. Ann Am Thorac Soc. 2013;10(3):179-187.

doi:10.1513/AnnalsATS.201211-107OC

108

76. Klepac-Ceraj V, Lemon KP, Martin TR, et al. Relationship between cystic fibrosis

respiratory tract bacterial communities and age, genotype, antibiotics and Pseudomonas

aeruginosa. Environ Microbiol. 2010;12(5):1293-1303. doi:10.1111/j.1462-

2920.2010.02173.x

77. Proctor L. Priorities for the next 10 years of human microbiome research. Nature.

2019;569(7758):623-625. doi:10.1038/d41586-019-01654-0

78. Einarsson GG, Zhao J, LiPuma JJ, Downey DG, Tunney MM, Elborn JS. Community

analysis and co-occurrence patterns in airway microbial communities during health and

disease. ERJ open Res. 2019;5(3). doi:10.1183/23120541.00128-2017

79. Taylor-Robinson D, Whitehead M, Diderichsen F, et al. Understanding the natural

progression in %FEV1 decline in patients with cystic fibrosis: a longitudinal study.

Thorax. 2012;67(10):860-866. doi:10.1136/thoraxjnl-2011-200953

80. Rosenbluth DB, Wilson K, Ferkol T, Schuster DP. Lung function decline in cystic fibrosis

patients and timing for lung transplantation referral. Chest. 2004;126(2):412-419.

doi:10.1378/chest.126.2.412

81. Kerem E, Reisman J, Corey M, Canny GJ, Levison H. Prediction of mortality in patients

with cystic fibrosis. N Engl J Med. 1992;326(18):1187-1191.

doi:10.1056/NEJM199204303261804

82. Acosta N, Heirali A, Somayaji R, et al. Sputum microbiota is predictive of long-term

clinical outcomes in young adults with cystic fibrosis. Thorax. 2018;73(11):1016-1025.

doi:10.1136/thoraxjnl-2018-211510

83. Heirali AA, Workentine ML, Acosta N, et al. The effects of inhaled aztreonam on the

cystic fibrosis lung microbiome. Microbiome. 2017;5(1):51. doi:10.1186/s40168-017-

109

0265-7

84. Rogers GB, Bruce KD, Hoffman LR. How can the cystic fibrosis respiratory microbiome

influence our clinical decision-making? Curr Opin Pulm Med. 2017;23(6):536-543.

doi:10.1097/MCP.0000000000000419

85. Brown-Elliott BA, Griffith DE, Wallace RJJ. Newly described or emerging human species

of nontuberculous mycobacteria. Infect Dis Clin North Am. 2002;16(1):187-220.

86. RUNYON EH. Anonymous mycobacteria in pulmonary disease. Med Clin North Am.

1959;43(1):273-290.

87. Gupta RS, Lo B, Son J. Phylogenomics and Comparative Genomic Studies Robustly

Support Division of the Genus Mycobacterium into an Emended Genus Mycobacterium

and Four Novel Genera. Front Microbiol. 2018;9:67. doi:10.3389/fmicb.2018.00067

88. Falkinham JO 3rd. Surrounded by mycobacteria: nontuberculous mycobacteria in the

human environment. J Appl Microbiol. 2009;107(2):356-367. doi:10.1111/j.1365-

2672.2009.04161.x

89. Malhotra S, Vedithi SC, Blundell TL. Decoding the similarities and differences among

mycobacterial species. PLoS Negl Trop Dis. 2017;11(8):e0005883-e0005883.

doi:10.1371/journal.pntd.0005883

90. Larsson L-O, Polverino E, Hoefsloot W, et al. Pulmonary disease by non-tuberculous

mycobacteria - clinical management, unmet needs and future perspectives. Expert Rev

Respir Med. 2017;11(12):977-989. doi:10.1080/17476348.2017.1386563

91. Riley L. Tuberculosis: Natural history, microbiology, and pathogenesis. Uptodate. 2018.

https://www.uptodate.com/contents/tuberculosis-natural-history-microbiology-and-

pathogenesis?search=tuberculosis&source=search_result&selectedTitle=6~150&usage_ty

110

pe=default&display_rank=6.

92. Vynnycky E, Fine PEM. The natural history of tuberculosis: the implications of age-

dependent risks of disease and the role of reinfection. Epidemiol Infect. 1997;119(2):183–

201.

93. Dheda K, Barry CE 3rd, Maartens G. Tuberculosis. Lancet (London, England).

2016;387(10024):1211-1226. doi:10.1016/S0140-6736(15)00151-8

94. Suzuki K, Udono T, Fujisawa M, Tanigawa K, Idani G, Ishii N. Infection during Infancy

and Long Incubation Period of Leprosy Suggested in a Case of a Chimpanzee Used for

Medical Research. J Clin Microbiol. 2010;48(9):3432 LP - 3434.

http://jcm.asm.org/content/48/9/3432.abstract.

95. Mougari F, Guglielmetti L, Raskine L, Sermet-Gaudelus I, Veziris N, Cambau E.

Infections caused by Mycobacterium abscessus: epidemiology, diagnostic tools and

treatment. Expert Rev Anti Infect Ther. 2016;14(12):1139-1154.

doi:10.1080/14787210.2016.1238304

96. Roux A-L, Catherinot E, Ripoll F, et al. Multicenter study of prevalence of

nontuberculous mycobacteria in patients with cystic fibrosis in france. J Clin Microbiol.

2009;47(12):4124-4128. doi:10.1128/JCM.01257-09

97. Pierre-Audigier C, Ferroni A, Sermet-Gaudelus I, et al. Age-related prevalence and

distribution of nontuberculous mycobacterial species among patients with cystic fibrosis.

J Clin Microbiol. 2005;43(7):3467-3470. doi:10.1128/JCM.43.7.3467-3470.2005

98. Jönsson BE, Gilljam M, Lindblad A, Ridell M, Wold AE, Welinder-Olsson C. Molecular

epidemiology of Mycobacterium abscessus, with focus on cystic fibrosis. J Clin

Microbiol. 2007;45(5):1497-1504. doi:10.1128/JCM.02592-06

111

99. Sermet-Gaudelus I, Le Bourgeois M, Pierre-Audigier C, et al. Mycobacterium abscessus

and children with cystic fibrosis. Emerg Infect Dis. 2003;9(12):1587-1591.

doi:10.3201/eid0912.020774

100. Bryant JM, Grogono DM, Rodriguez-Rincon D, et al. Emergence and spread of a human-

transmissible multidrug-resistant nontuberculous mycobacterium. Science.

2016;354(6313):751-757. doi:10.1126/science.aaf8156

101. Lopeman RC, Harrison J, Rathbone DL, Desai M, Lambert PA, Cox JAG. Effect of

Amoxicillin in combination with Imipenem-Relebactam against Mycobacterium

abscessus. Sci Rep. 2020;10(1):928. doi:10.1038/s41598-020-57844-8

102. Griffith DE, Aksamit T, Brown-elliott BA, et al. American Thoracic Society Documents

An Official ATS / IDSA Statement : Diagnosis , Treatment , and Prevention of

Nontuberculous Mycobacterial Diseases. 2007. doi:10.1164/rccm.200604-571ST

103. Floto RA, Olivier KN, Saiman L, et al. US Cystic Fibrosis Foundation and European

Cystic Fibrosis Society consensus recommendations for the management of non-

tuberculous mycobacteria in individuals with cystic fibrosis. Thorax. 2016;71 Suppl 1:i1-

22. doi:10.1136/thoraxjnl-2015-207360

104. Rapp RP, McCraney SA, Goodman NL, Shaddick DJ. New macrolide antibiotics:

usefulness in infections caused by mycobacteria other than Mycobacterium tuberculosis.

Ann Pharmacother. 1994;28(11):1255-1263. doi:10.1177/106002809402801109

105. Zhanel GG, Homenuik K, Nichol K, et al. The glycylcyclines: a comparative review with

the tetracyclines. Drugs. 2004;64(1):63-88. doi:10.2165/00003495-200464010-00005

106. Mingeot-Leclercq MP, Glupczynski Y, Tulkens PM. Aminoglycosides: activity and

resistance. Antimicrob Agents Chemother. 1999;43(4):727-737.

112

107. Schnappinger D, Hillen W. Tetracyclines: antibiotic action, uptake, and resistance

mechanisms. Arch Microbiol. 1996;165(6):359-369.

108. Papp-Wallace KM, Endimiani A, Taracila MA, Bonomo RA. Carbapenems: past, present,

and future. Antimicrob Agents Chemother. 2011;55(11):4943-4960.

doi:10.1128/AAC.00296-11

109. Drlica K, Zhao X. DNA gyrase, topoisomerase IV, and the 4-quinolones. Microbiol Mol

Biol Rev. 1997;61(3):377-392.

110. Cholo MC, Mothiba MT, Fourie B, Anderson R. Mechanisms of action and therapeutic

efficacies of the lipophilic antimycobacterial agents clofazimine and bedaquiline. J

Antimicrob Chemother. 2017;72(2):338-353. doi:10.1093/jac/dkw426

111. Bozdogan B, Appelbaum PC. Oxazolidinones: activity, mode of action, and mechanism of

resistance. Int J Antimicrob Agents. 2004;23(2):113-119.

doi:10.1016/j.ijantimicag.2003.11.003

112. Griffith DE, Aksamit T, Brown-Elliott BA, et al. An official ATS/IDSA statement:

diagnosis, treatment, and prevention of nontuberculous mycobacterial diseases. Am J

Respir Crit Care Med. 2007;175(4):367-416. doi:10.1164/rccm.200604-571ST

113. Campbell I. Chi‐squared and Fisher–Irwin tests of two‐by‐two tables with small sample

recommendations. Stat Med. 2007;26(19):3661-3675.

114. Freeman J, Julious S. The Analysis of Categorical Data. 16(1). Scope; 2007.

115. Armitage P, Berry G, Matthews JN. Statistical Methods in Medical Research. 4th Ed.

Oxford: Blackwell Publishing; 2002.

116. Swinscow T, Campbell M. Statistics at Square One. 10 Ed. London: BMJ Books; 2002.

117. Wu P, Han Y, Chen T, Tu XM. Causal inference for Mann–Whitney–Wilcoxon rank sum

113

and other nonparametric statistics. Stat Med. 2014;33(8):1261-1271.

118. Ruxton GD, Beauchamp G. Some suggestions about appropriate use of the Kruskal–

Wallis test. Anim Behav. 2008;76(3):1083-1087.

doi:https://doi.org/10.1016/j.anbehav.2008.04.011

119. Anderson MJ. Permutational Multivariate Analysis of Variance (PERMANOVA). In:

Wiley StatsRef: Statistics Reference Online. American Cancer Society; 2017:1-15.

doi:10.1002/9781118445112.stat07841

120. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for

RNA-seq data with DESeq2. Genome Biol. 2014;15(12):550. doi:10.1186/s13059-014-

0550-8

121. Heirali AA, Acosta N, Storey DG, et al. The effects of cycled inhaled aztreonam on the

cystic fibrosis (CF) lung microbiome. J Cyst Fibros. March 2019.

doi:10.1016/j.jcf.2019.02.010

122. Whelan FJ, Verschoor CP, Stearns JC, et al. The Loss of Topography in the Microbial

Communities of the Upper Respiratory Tract in the Elderly. Ann Am Thorac Soc.

2014;11(4):513-521. doi:10.1513/AnnalsATS.201310-351OC

123. Bartram AK, Lynch MDJ, Stearns JC, Moreno-Hagelsieb G, Neufeld JD. Generation of

Multimillion-Sequence 16S rRNA Gene Libraries from Complex Microbial Communities

by Assembling Paired-End Illumina Reads. Appl Environ Microbiol. 2011;77(11):3846

LP - 3852. doi:10.1128/AEM.02772-10

124. Muyzer G, de Waal EC, Uitterlinden AG. Profiling of complex microbial populations by

denaturing gradient gel electrophoresis analysis of polymerase chain reaction-amplified

genes coding for 16S rRNA. Appl Environ Microbiol. 1993;59(3):695 LP - 700.

114

http://aem.asm.org/content/59/3/695.abstract.

125. Chakravorty S, Helb D, Burday M, Connell N, Alland D. A detailed analysis of 16S

ribosomal RNA gene segments for the diagnosis of pathogenic bacteria. J Microbiol

Methods. 2007;69(2):330-339. doi:10.1016/j.mimet.2007.02.005

126. Bartram AK, Lynch MDJ, Stearns JC, Moreno-Hagelsieb G, Neufeld JD. Generation of

multimillion-sequence 16S rRNA gene libraries from complex microbial communities by

assembling paired-end Illumina reads. Appl Environ Microbiol. 2011;77(11):3846-3852.

doi:10.1128/AEM.02772-10

127. Callahan BJ, McMurdie PJ, Rosen MJ, Han AW, Johnson AJA, Holmes SP. DADA2:

High-resolution sample inference from Illumina amplicon data. Nat Methods.

2016;13(7):581-583. doi:10.1038/nmeth.3869

128. Martin M. Cutadapt removes adapter sequences from high-throughput sequencing reads.

EMBnet.journal; Vol 17, No 1 Next Gener Seq Data Anal - 1014806/ej171200 . May

2011. http://journal.embnet.org/index.php/embnetjournal/article/view/200.

129. Edgar RC, Flyvbjerg H. Error filtering, pair assembly and error correction for next-

generation sequencing reads. Bioinformatics. 2015;31(21):3476-3482.

doi:10.1093/bioinformatics/btv401

130. Mukherjee C, Beall CJ, Griffen AL, Leys EJ. High-resolution ISR amplicon sequencing

reveals personalized oral microbiome. Microbiome. 2018;6(1):153. doi:10.1186/s40168-

018-0535-z

131. Wright E. Using DECIPHER v2.0 to analyze big biological sequence data in R. R Journ.

2016;8:352-359.

132. Quast C, Pruesse E, Yilmaz P, et al. The SILVA ribosomal RNA gene database project:

115

improved data processing and web-based tools. Nucleic Acids Res. 2013;41(Database

issue):D590-D596. doi:10.1093/nar/gks1219

133. Willis AD. Rarefaction, Alpha Diversity, and Statistics. Front Microbiol. 2019;10:2407.

doi:10.3389/fmicb.2019.02407

134. McMurdie PJ, Holmes S. phyloseq: an R package for reproducible interactive analysis and

graphics of microbiome census data. PLoS One. 2013;8(4):e61217.

doi:10.1371/journal.pone.0061217

135. Oksanen J, Blanchet FG, Kindt R, et al. vegan: Community Ecology Package. R package

version 1.17-2. Software. 2011.

136. Nadkarni MA, Martin FE, Jacques NA, Hunter N. Determination of bacterial load by real-

time PCR using a broad-range ( universal ) probe and primers set. 2017;(2002):257-266.

137. Farrelly V, Rainey FA, Stackebrandt E. Effect of genome size and rrn gene copy number

on PCR amplification of 16S rRNA genes from a mixture of bacterial species. Appl

Environ Microbiol. 1995;61(7):2798-2801. doi:10.1128/AEM.61.7.2798-2801.1995

138. Qin X, Emerson J, Stapp J, Stapp L, Abe P, Burns JL. Use of real-time PCR with multiple

targets to identify Pseudomonas aeruginosa and other nonfermenting gram-negative

bacilli from patients with cystic fibrosis. J Clin Microbiol. 2003;41(9):4312-4317.

doi:10.1128/jcm.41.9.4312-4317.2003

139. Rocchetti TT, Silbert S, Gostnell A, Kubasek C, Campos Pignatari AC, Widen R.

Detection of Mycobacterium chelonae, Mycobacterium abscessus Group, and

Mycobacterium fortuitum Complex by a Multiplex Real-Time PCR Directly from Clinical

Samples Using the BD MAX System. J Mol Diagnostics. 2017;19(2):295-302.

doi:10.1016/j.jmoldx.2016.10.004

116

140. Whelan F, Surette M. A comprehensive evaluation of the sl1p pipeline for 16S rRNA gene

sequencing analysis. Microbiome. 2017;5. doi:10.1186/s40168-017-0314-2

141. Ye Y. Identification and quantification of abundant species from pyrosequences of 16S

rRNA by consensus alignment. In: 2010 IEEE International Conference on

Bioinformatics and Biomedicine (BIBM). IEEE; 2010:153-157.

142. Masella AP, Bartram AK, Truszkowski JM, Brown DG, Neufeld JD. PANDAseq: paired-

end assembler for illumina sequences. BMC Bioinformatics. 2012;13(1):1-7.

143. Wang Q, Garrity GM, Tiedje JM, Cole JR. Naive Bayesian classifier for rapid assignment

of rRNA sequences into the new bacterial taxonomy. Appl Environ Microbiol.

2007;73(16):5261-5267.

144. DeSantis TZ, Hugenholtz P, Larsen N, et al. Greengenes, a chimera-checked 16S rRNA

gene database and workbench compatible with ARB. Appl Environ Microbiol.

2006;72(7):5069-5072.

145. Bryant JM, Grogono DM, Greaves D, et al. Whole-genome sequencing to identify

transmission of Mycobacterium abscessus between patients with cystic fibrosis: a

retrospective cohort study. Lancet (London, England). 2013;381(9877):1551-1560.

doi:10.1016/S0140-6736(13)60632-7

146. Saiman L, Siegel J. Infection control in cystic fibrosis. Clin Microbiol Rev. 2004;17(1):57-

71. doi:10.1128/cmr.17.1.57-71.2004

147. Thomson R, Tolson C, Sidjabat H, Huygens F, Hargreaves M. Mycobacterium abscessus

isolated from municipal water - a potential source of human infection. BMC Infect Dis.

2013;13:241. doi:10.1186/1471-2334-13-241

148. Feazel LM, Baumgartner LK, Peterson KL, Frank DN, Harris JK, Pace NR. Opportunistic

117

pathogens enriched in showerhead biofilms. Proc Natl Acad Sci. 2009;106(38):16393-

16399. doi:10.1073/pnas.0908446106

149. Watkins RR, Lemonovich TL. Evaluation of infections in the lung transplant patient. Curr

Opin Infect Dis. 2012;25(2):193-198. doi:10.1097/QCO.0b013e32834c54bc

150. Luong M-L, Morrissey O, Husain S. Assessment of infection risks prior to lung

transplantation. Curr Opin Infect Dis. 2010;23(6):578-583.

doi:10.1097/QCO.0b013e32833f9f93

151. Gilljam M, Scherstén H, Silverborn M, Jönsson B, Ericsson Hollsing A. Lung

transplantation in patients with cystic fibrosis and Mycobacterium abscessus infection. J

Cyst Fibros Off J Eur Cyst Fibros Soc. 2010;9(4):272-276. doi:10.1016/j.jcf.2010.03.008

152. Leung JM, Olivier KN. Nontuberculous mycobacteria: the changing epidemiology and

treatment challenges in cystic fibrosis. Curr Opin Pulm Med. 2013;19(6):662-669.

doi:10.1097/MCP.0b013e328365ab33

153. Nessar R, Cambau E, Reyrat JM, Murray A, Gicquel B. Mycobacterium abscessus: a new

antibiotic nightmare. J Antimicrob Chemother. 2012;67(4):810-818.

doi:10.1093/jac/dkr578

154. Caverly LJ, Zhao J, LiPuma JJ. Cystic fibrosis lung microbiome: Opportunities to

reconsider management of airway infection. Pediatr Pulmonol. 2015;50(S40):S31-S38.

doi:https://doi.org/10.1002/ppul.23243

155. Simpson HL, Campbell BJ. Review article: dietary fibre-microbiota interactions. Aliment

Pharmacol Ther. 2015;42(2):158-179. doi:10.1111/apt.13248

156. Xu Z, Knight R. Dietary effects on human gut microbiome diversity. Br J Nutr. 2015;113

Suppl(Suppl 0):S1-5. doi:10.1017/S0007114514004127

118

157. Shukla SD, Budden KF, Neal R, Hansbro PM. Microbiome effects on immunity, health

and disease in the lung. Clin Transl Immunol. 2017;6(3):e133. doi:10.1038/cti.2017.6

158. Trompette A, Gollwitzer ES, Yadava K, et al. Gut microbiota metabolism of dietary fiber

influences allergic airway disease and hematopoiesis. Nat Med. 2014;20(2):159-166.

doi:10.1038/nm.3444

159. Taccetti G, Campana S, Festini F, Mascherini M, Doring G. Early eradication therapy

against Pseudomonas aeruginosa in cystic fibrosis patients. Eur Respir J. 2005;26(3):458-

461. doi:10.1183/09031936.05.00009605

160. Robinson NB, DiMango E. Prevalence of gastroesophageal reflux in cystic fibrosis and

implications for lung disease. Ann Am Thorac Soc. 2014;11(6):964-968.

doi:10.1513/AnnalsATS.201401-044FR

161. Stringer DA, Sprigg A, Juodis E, et al. The association of cystic fibrosis, gastroesophageal

reflux, and reduced pulmonary function. Can Assoc Radiol J = J l’Association Can des

Radiol. 1988;39(2):100-102.

162. Button BM, Roberts S, Kotsimbos TC, et al. Gastroesophageal reflux (symptomatic and

silent): a potentially significant problem in patients with cystic fibrosis before and after

lung transplantation. J Hear lung Transplant Off Publ Int Soc Hear Transplant.

2005;24(10):1522-1529. doi:10.1016/j.healun.2004.11.312

163. Ayoub F, Lascano J, Morelli G. Proton Pump Inhibitor Use Is Associated With an

Increased Frequency of Hospitalization in Patients With Cystic Fibrosis. Gastroenterol

Res. 2017;10(5):288-293. doi:10.14740/gr917w

164. Laheij RJF, Sturkenboom MCJM, Hassing R-J, Dieleman J, Stricker BHC, Jansen JBMJ.

Risk of community-acquired pneumonia and use of gastric acid-suppressive drugs. JAMA.

119

2004;292(16):1955-1960. doi:10.1001/jama.292.16.1955

165. de Jager CPC, Wever PC, Gemen EFA, et al. Proton pump inhibitor therapy predisposes

to community-acquired Streptococcus pneumoniae pneumonia. Aliment Pharmacol Ther.

2012;36(10):941-949. doi:10.1111/apt.12069

166. Carmody LA, Zhao J, Kalikin LM, et al. The daily dynamics of cystic fibrosis airway

microbiota during clinical stability and at exacerbation. Microbiome. 2015;3:12.

doi:10.1186/s40168-015-0074-9

167. Price KE, Hampton TH, Gifford AH, et al. Unique microbial communities persist in

individual cystic fibrosis patients throughout a clinical exacerbation. Microbiome.

2013;1(1):27. doi:10.1186/2049-2618-1-27

168. Stressmann FA, Rogers GB, Marsh P, et al. Does bacterial density in cystic fibrosis

sputum increase prior to pulmonary exacerbation? J Cyst Fibros Off J Eur Cyst Fibros

Soc. 2011;10(5):357-365. doi:10.1016/j.jcf.2011.05.002

169. Olivier KN, Weber DJ, Wallace RJ, et al. Nontuberculous Mycobacteria. Am J Respir Crit

Care Med. 2003;167(6):828-834. doi:10.1164/rccm.200207-678OC

170. Esther CRJ, Esserman DA, Gilligan P, Kerr A, Noone PG. Chronic Mycobacterium

abscessus infection and lung function decline in cystic fibrosis. J Cyst Fibros.

2010;9(2):117-123. doi:10.1016/j.jcf.2009.12.001

171. Kaur J, Pethani BP, Kumar S, et al. Pseudomonas aeruginosa inhibits the growth of

Scedosporium aurantiacum, an opportunistic fungal pathogen isolated from the lungs of

cystic fibrosis patients. Front Microbiol. 2015;6:866. doi:10.3389/fmicb.2015.00866

172. Whittier S, Olivier K, Gilligan P, Knowles M, Della-Latta P. Proficiency testing of

clinical microbiology laboratories using modified decontamination procedures for

120

detection of nontuberculous mycobacteria in sputum samples from cystic fibrosis patients.

The Nontuberculous Mycobacteria in Cystic Fibrosis Study Gro. J Clin Microbiol.

1997;35(10):2706-2708. doi:10.1128/JCM.35.10.2706-2708.1997

173. Jordan PW, Stanley T, Donnelly FM, et al. Atypical mycobacterial infection in patients

with cystic fibrosis: update on clinical microbiology methods. Lett Appl Microbiol.

2007;44(5):459-466. doi:10.1111/j.1472-765X.2007.02130.x

174. Hogan DA, Kolter R. Pseudomonas-Candida Interactions: An Ecological Role for

Virulence Factors. Science (80- ). 2002;296(5576):2229-2232.

doi:10.1126/science.1070784

175. Manavathu EK, Vager DL, Vazquez JA. Development and antimicrobial susceptibility

studies of in vitro monomicrobial and polymicrobial biofilm models with Aspergillus

fumigatus and Pseudomonas aeruginosa. BMC Microbiol. 2014;14(1):53.

doi:10.1186/1471-2180-14-53

176. Lim YW, Evangelista JS, Schmieder R, et al. Clinical Insights from Metagenomic

Analysis of Sputum Samples from Patients with Cystic Fibrosis. Forbes BA, ed. J Clin

Microbiol. 2014;52(2):425-437. doi:10.1128/JCM.02204-13

177. Blau H, Linnane B, Carzino R, et al. Induced sputum compared to bronchoalveolar lavage

in young, non-expectorating cystic fibrosis children. J Cyst Fibros. 2014;13(1):106-110.

doi:10.1016/j.jcf.2013.05.013