Download pdf - Elevated prevalence of azole resistant Aspergillus fumigatus · 27 Introduction 28 Aspergillus fumigatus is a ubiquitous ascomycete fungus with a pan-global population distribution

Elevated prevalence of azole resistant Aspergillus fumigatus in urban versus rural environments in the 1

United Kingdom 2

Thomas R Sewell1, Yuyi Zhang1, Amelie P Brackin, Jennifer MG Shelton, Johanna Rhodes and Matthew C 3

Fisher 4

MRC Centre for Global Infectious Disease Analysis, Department of Infectious Disease Epidemiology, 5

Imperial College London, London, UK 6

1These authors contributed equally to this work 7

.CC-BY-NC-ND 4.0 International licensecertified by peer review) is the author/funder. It is made available under aThe copyright holder for this preprint (which was notthis version posted April 5, 2019. . https://doi.org/10.1101/598961doi: bioRxiv preprint

https://doi.org/10.1101/598961

http://creativecommons.org/licenses/by-nc-nd/4.0/

Abstract 8

Azole resistance in the opportunistic pathogen Aspergillus fumigatus is increasing, dominated primarily 9

by two environmentally-associated resistance alleles: TR34/L98H and TR46/Y121F/T289A. Using an 10

environmental sampling strategy across the South of England we assess the prevalence of azole resistant 11

A. fumigatus (ARAf) in soil samples collected in both urban and rural locations. We characterise the 12

susceptibility profiles of the resistant isolates to three medical azoles, identify the underlying genetic 13

basis of resistance and investigate their genetic relationships. ARAf was detected in 6.7% of the soil 14

samples, with a higher prevalence in urban (13.8%) compared to rural (1.1%) locations. Nineteen 15

isolates were confirmed to exhibit clinical breakpoints for resistance to at least one of three medical 16

azoles, with 18 isolates exhibiting resistance to itraconazole, four to voriconazole, with two also showing 17

additional elevated minimum inhibitory concentration to posaconazole. Thirteen of the resistant isolates 18

harboured the TR34/L98H resistance allele and six isolates carried TR46/Y121F/T289A allele. The 19 azole-19

resistant isolates were spread across five csp1 genetic subtypes, t01, t02, t04B, t09 and t18 with t02 the 20

predominant subtype. Our study demonstrates that ARAf can be easily isolated in the South of England, 21

especially in urban city centres, which appear to play an important role in the epidemiology of 22

environmentally-linked drug resistant A. fumigatus. 23

24

Keywords: Aspergillus fumigatus, triazoles, antifungal resistance, fungal pathogen, fungal epidemiology, 25

multidrug resistance, environment 26


https://doi.org/10.1101/598961


Introduction 27

Aspergillus fumigatus is a ubiquitous ascomycete fungus with a pan-global population distribution and a 28

primary ecological niche of decaying vegetation and soil (1). This fungus is the most prevalent species 29

among ~250 described Aspergilli, partly due to its ability to survive and grow in a wide range of 30

conditions, but also through the large-scale dispersal of airborne conidia (1–3). A. fumigatus is also an 31

opportunistic pathogen and is commonly responsible for aspergillosis, a spectrum of clinical syndromes 32

caused by Aspergillus spp. that affects millions of individuals worldwide (4). Invasive aspergillosis (IA), 33

the most severe form of the disease, can lead to serious and even fatal illness in immunocompromised 34

individuals, with a mortality rate of 40 - 90% (5, 6). 35

36

Triazole antifungals are used for the treatment and prophylaxis of Aspergillus spp. infections, although 37

resistance has emerged, often conferred by the presence of mutations in lanosterol 14 alpha 38

demethylase (erg11, syn. CYP51), which is a key component of the ergosterol biosynthetic pathway and 39

target for azole antifungals (7, 8). Recently, azole-resistant A. fumigatus (ARAf) has emerged 40

environmentally, where selection is thought to be driven by the broad application of agricultural azole 41

fungicides; structurally similar to their medical counterparts and indistinguishable in their mode of 42

action (9, 10). Environmentally sourced ARAf are typically found with a tandem repeat mutation in the 43

cyp51A promoter region and linked single nucleotide polymorphisms in the coding region, with common 44

examples being TR34/L98H and TR46/Y121F/T289A (11, 12). 45

46

Studies have shown that ARAf isolates harbouring resistance-associated cyp51A variants are globally 47

distributed, and are often found alongside wild-type A. fumigatus in a diverse set of environmental 48

substrates, including: agricultural soil (13, 14), flower beds (13, 15–17) and timber mills (18). Moreover, 49

clinical cases of aspergillosis with ARAf isolates harbouring either TR34/L98H or TR46/Y121F/T289A 50


https://doi.org/10.1101/598961


continue to emerge (19, 20), with one study specifically linking a fatal case of aspergillosis to a 51

genotypically indistinguishable isolate sourced from the patients own home (21). Retrospective studies 52

of patients with invasive aspergillosis (IA) and infected with azole-resistant genotypes of A. fumigatus 53

show an excess mortality of 25% at day 90 when compared against patients with wild-type-infections 54

(22). 55

56

Despite a generally good understanding of ARAf global prevalence, very few studies have investigated 57

resistance in the United Kingdom or focused sampling strategies across a diverse set of substrates. In 58

this study, we aim to determine the prevalence of ARAf in soil collected from a wide range of sites 59

across the UK's southernmost region. Sampling locations include ancient woodlands, agricultural fields, 60

tourist attractions and densely populated city centres. We show that the UK has similar prevalence rates 61

to other countries and that both TR34/L98H and TR46/Y121F/T289A can be regularly isolated from 62

urbanised locations, especially flower beds in close proximity to city centre hospitals. 63


https://doi.org/10.1101/598961


Materials and Methods 64

Environmental sampling 65

Samples were collected from 16 sites across South England between May and July 2018. Locations were 66

selected to include a range of habitat types and included remote forested regions, urban city centre, 67

agricultural and flower fields. At each sample site, dry surface soil was loosened and collected into a 5ml 68

Eppendorf tube (Eppendorf AG, Hamburg, Germany). All tubes were labelled with site description, 69

longitude and latitude coordinates. All samples were stored at 4OC until processing. 70

71

Recovering of A. fumigatus and screening for azole resistance 72

To isolate A. fumigatus from environmental samples 2 g of each sample was suspended in 8 ml of sterile 73

distilled water with 0.85% NaCl (SIGMA, Sigma-Aldrich Chemie Gmbh, Steinheim, Germany) plus 0.01% 74

Tween 20 (SIGMA, Sigma-Aldrich Chemie Gmbh, Steinheim, Germany) and vortexed vigorously for 1 75

min. After 1 min settling, 200 µl of supernatant was added to two control plates, containing sabouraud 76

dextrose agar (CM0041, Oxoid Ltd, Basingstoke, Hants, UK) supplemented with 16 g/ml penicillin G 77

sodium salt (SIGMA, Sigma-Aldrich Chemie Gmbh, Steinheim, Germany) and 16 g/ml streptomycin 78

sulfate salt (SIGMA, Sigma-Aldrich Chemie Gmbh, Steinheim, Germany), and two azole-containing plates 79

containing sabouraud dextrose agar supplemented with 16 g/ml penicillin G sodium salt, 16 g/ml 80

streptomycin sulfate salt and 8 g/ml tebuconazole (PESTANAL analysis standard, Sigma-Aldrich AG 81

Industriestrasse, Switzerland). These plates were incubated at 42OC and examined for growth after 72 82

hours. A. fumigatus isolates were identified by observation of their macro and microscopic morphology. 83

84

In vitro susceptibility testing - Minimal inhibitory concentration (MIC) measurement 85

Seven wild-type isolates and 19 suspected azole-resistant isolates were tested for antifungal drug 86

susceptibility against three medical azoles, itraconazole (ITC), voriconazole (VRC) and posaconazole 87


https://doi.org/10.1101/598961


(PSC) using the European Committee Antimicrobial Susceptibility Testing (EUCAST) microdilution 88

method using MICRONAUT-AM EUCAST MIC (Minimum Inhibitory Concentration) plates (Merlin 89

Diagnostika GmbH, Bornheim, Germany). The reference wild-type clinical A. fumigatus isolates Af293 90

and an environmental azole-resistant isolate BUU09 (TR34/L98H) were used as control strains. All isolates 91

were grown for 72 hours at 37 OC before washing with 8 to 10 ml of 0.01% Tween 20. The resulting 92

conidial suspension was filtered through Whatman 1 filter paper (Whatman plc, GE Healthcare Life 93

Sciences, UK) to remove hyphae and debris. The filtered conidial suspension was adjusted to 0.5 94

McFarland with phosphate-buffered saline (PBS pH7.4 (1X), gibco, Life Technologies Limited, Paisely, UK) 95

using a BioTek ELx808 spectrophotometer (BioTek Instruments Ltd, Winooski, USA). The adjusted 96

conidial suspension (0.5 ml) was added to 9.5 ml of the MICRONAUT-RPMI medium (MICRONAUT-RPMI 97

Medium + MOPS + Glucose, Merlin Diagnostika GmbH, Bornheim, Germany) and 100 µl of the mixture 98

was then used to populate the MICRONAUT-AM test plate. All plates were examined after 48-hour 99

incubation at 37OC in a humid chamber. The MIC endpoint was read visually and determined as the 100

lowest concentrations of the azoles yielding no visible growth. The results of all susceptibility tests were 101

interpreted using EUCAST clinical breakpoints [91]. The isolates were regarded as susceptible when the 102

MIC was ≤ 1 mg/l for ITC and VRC, and ≤ 0.125 mg/l for POS; and resistant when the MIC was > 2mg/l for 103

ITC and VRC, and > 0.25 mg/l for POS. Isolates with MIC values in between the susceptible and resistant 104

breakpoint were considered to be intermediately susceptible. 105

106

DNA extraction for sequence analysis 107

Genomic DNA was extracted using a modified MasterPure Yeast DNA Purification (Lucigen Corporation, 108

Cambridge, UK) protocol which included an additional bead-beating treatment to enhance DNA yield. 109

110


https://doi.org/10.1101/598961


Cultured A. fumigatus isolates were washed in 10 ml of 0.05% Tween 20; 1.8 ml of the conidial 111

suspension was transferred into a 2 ml Eppendorf tube and centrifuged at 8000 rpm for 5 mins. The 112

condial pellet was mixed with 300 µl of Yeast Cell Lysis Solution and 1 µl of 5 g/l RNase A and transferred 113

into a new 2 ml Eppendorf tube containing 1.0 mm diameter Zirconia/Silica Bead (Thistle Scientific, 114

Glasgow, UK). The tube was subjected to bead beating for 3 x 45 seconds at 30 m/s in a TissueLyser II 115

(Qiagen, Hilden, Germany) and placed on ice for 2 mins before repeating. The lysate was transferred to a 116

new 1.5 Eppendorf tube and was twice centrifuged at 13000 rpm for 2 mins to pellet debris. The 117

supernatant was transferred to a 1.5 Eppendorf tube and incubated at 65OC for 15 mins before placing 118

on ice for another 30 mins. After incubation, 150 µl of MPC Protein Precipitation Reagent was added to 119

the sample and mixed by pulse vortexing for 10 seconds. For further pelleting of cellular debris, the 120

mixture was centrifuged at 13000 rpm for 2 mins and the supernatant was transferred to a clean 1.5 121

Eppendorf tube. DNA was precipitated by adding 500 µl of isopropanol (SIGMA, Sigma-Aldrich Chemie 122

Gmbh, Steinheim, Germany) to the supernatant before 4OC centrifugation at 14000 rpm for 10 mins. The 123

supernatant was removed, and the remaining DNA pellet washed with 500 µl of 70% ethanol (SIGMA, 124

Sigma-Aldrich Chemie Gmbh, Steinheim, Germany) before centrifugation at 13000 rpm for 2 mins. After 125

removing the ethanol, the DNA pellet was air-dried for approximately 10-15 mins in a biosafety cabinet. 126

Finally, 40 µl of nuclease-free water was added to the DNA pellet and left to resuspend at 4OC for 24 127

hours. The DNA was either used immediately or stored at -20OC until required. DNA yield was quantified 128

using a Qubit 2.0 Fluorometer (Invitrogen by Thermo Fisher Scientific corporation, Massachusetts, USA) 129

and the Qubit dsDNA BR (Broad-Range) Assay Kit (Invitrogen) according to the products guidelines. 130

131

PCR amplification 132

PCR was used to amplify part of the cyp51A gene and promoter, the csp1 region and beta-tubulin as 133

previously described [45], [80]. Amplification of cyp51A was performed using the L98HR primer (5′-134


https://doi.org/10.1101/598961


TTCGGTGAATCGCGCAGATAGTCC-3′) and TR34R primer (5′-AGCAAGGGAGAAGGAAAGAAGCACT-3′) 135

(Invitrogen) at 100 nM, 20.125 µl nuclease-free water (Qiagen, Hilden, Germany), 2.5 µl PCR Buffer 136

(Qiagen PCR Buffer containing 15mM MgCl2, Qiagen, Hilden, Germany), 0.5 µl dNTP mix (dNTP Set, PCR 137

Grade, 10mM each, Qiagen, Hilden, Germany), 0.125 µl Taq DNA Polymerase (5 units/L, Qiagen, Hilden, 138

Germany) and 1 µl of DNA template. Amplification conditions were: a 2 mins denaturation step at 94OC, 139

followed by 35 cycles of 15 seconds at 94OC, 30 seconds at 57OC, and 30 seconds at 72OC. A final 140

elongation step for 2 mins at 72OC was followed after the last cycle. Amplification of csp1 was 141

performed using the CSP1F primer (5′-TTGGGTGGCATTGTGCCAA-3′) and CSP1R primer (5′-142

GAGCATGACAACCCAGATACCA-3′) at 100 nM, 20.125 µl nuclease-free water, 2.5 µl PCR Buffer, 0.5 µl 143

dNTP mix, 0.125 µl Taq DNA Polymerase and 1 µl of DNA template. Amplification conditions were: a 5 144

mins denaturation step at 94OC, followed by 35 cycles of 15 seconds at 94OC, 30 seconds at 55OC, and 30 145

seconds at 68OC. A final elongation step for 2 mins at 68OC was followed after the last cycle. 146

Amplification of beta-tubulin was performed using the Bt2A_F primer (5′- 147

GGTAACCAAATCGGTGCTGCTTTC - 3′) and Bt2A-R primer (5′ - ACCCTCAGTGTAGTGACCCTTGGC - 3′) at 148

100 nM, 20.125 µl nuclease-free water, 2.5 µl PCR Buffer, 0.5 µl dNTP mix, 0.125 µl Taq DNA Polymerase 149

and 1 µl of DNA template. Amplification conditions were: a 5 mins denaturation step at 95OC, followed 150

by 35 cycles of 30 seconds at 95OC, 30 seconds at 55OC, and 1 min at 72OC. A final elongation step for 5 151

mins at 72OC was followed after the last cycle. 152

153

PCR product visualisation, purification and sequencing 154

Both cyp51A gene and csp1 gene PCR products were visualised on a 2% agarose gel using agarose gel 155

electrophoresis. 100 ml of 2% agarose solution was prepared by mixing 2 g of agarose powder with 100 156

ml of TBE buffer. 10 µl of SafeView Nucleic Acid Stain was added for visualisation of the PCR product in 157

the gel. Before loading the samples and molecular weight ladder (Quick-load Purple 50 bp DNA Ladder 158


https://doi.org/10.1101/598961


and Quick-load Purple 1 kb DNA Ladder, New England Biolabs Inc, UK) to the gel, 1 µl of Gel Loading Dye 159

was mixed with 5 µl of each sample. Finally, the gel was run at 120 to 130 V for approximately 45 to 60 160

mins. The DNA fragments were visualised using a G:Box Gel Image Analysis System (Syngene UK, 161

Cambridge, UK). The PCR products were then purified using ExoSAP-IT PCR Product Cleanup Reagent 162

(ThermoFisher Scientific, Massachusetts, USA) following the product guidelines. Briefly, 7.5 µl of each 163

post-PCR reaction product was mixed with 3 µl of the reagent. The mixture was incubated for 15 mins at 164

37OC in order to degrade the remaining nucleotides and primers in the PCR product, followed by 165

incubation of 15 mins at 80OC to inactivate the reagent. The treated products were then stored at -20OC 166

until required. The treated PCR products and sequencing primers, L98HR, TR34R, CSP1F and CSP1R were 167

prepared for sequencing following the DNA Sanger sequencing sample submission guidelines provided 168

by GENEWIZ (GENEWIZ UK LTD, Takeley, UK). They were then sent away for Sanger sequencing by 169

GENEWIZ UK laboratory. 170

171

Sanger sequencing of cyp51A mutations, csp1 types and beta-tubulin 172

Sanger sequencing results from all selected isolates were trimmed and assembled using CLC Main 173

Workbench 8.0.1 software (QIAGEN Bioinformatics, Hilden, Germany), with alignment stringency set to 174

medium. The forward and reverse reads were merged into a consensus sequence using the assemble 175

sequences function. The consensus cyp51A sequences of each environmental isolates, and the known 176

TR34/L98H isolate BUU09, were aligned against the referenced wild-type isolate, A. fumigatus Af293. The 177

isolates that exhibited the same 34-bp tandem repeat as the isolate BUU09 were confirmed to harbor 178

the TR34/L98H allele. The isolates that had a 46-bp tandem repeat were compared with A. fumigatus 179

isolate VRJ056 (GenBank accession number: MF070884.1), a known clinical isolate harboring the 180

TR46/Y121F/T289A mutation, to confirm their resistant mechanism. The csp1 types of each 181

environmental isolates and two control isolates were assigned according to the csp1 typing 182


https://doi.org/10.1101/598961


nomenclature described previously (23), which was based on manually inspecting the repeat regions in 183

each consensus csp1 sequences. 184


https://doi.org/10.1101/598961


Results 185

Environmental A. fumigatus isolates 186

A total of 178 soil samples were collected across Southern England, including soil and compost samples 187

obtained from public gardens, parks, cemeteries, and flower beds outside hospitals. Samples were 188

collected in Central London (n = 64), Bath (n = 8), flower beds around Stonehenge (n = 8), remote 189

forested regions in the New Forest National Park (n = 46), a lavender farm in Surry (n = 13) and farmland 190

in Cambridgeshire (n = 39). Of the 178 soil samples, 131 (74%) were positive for A. fumigatus type 191

growth on control plates, with a varied recovery rate amongst different sampling sites (Table 1). The 192

highest recovery rate was from Central London (93.8%) and the lowest from Cambridgeshire (35.9%). 193

Eleven A. fumigatus positive soil samples yielded a total of 83 A. fumigatus type isolates that were able 194

to grow on the azole-containing plates. Nine of these soil samples were collected from urban locations 195

in London (n=8) and Bath (n=1). Among the eight samples from London, four were collected from flower 196

beds outside Royal Free Hospital and The Whittington Hospital, and another two samples were obtained 197

from compost in Waterlow Park, 500 meters away from The Whittington Hospital. Nineteen putatively 198

resistant isolates were selected for further MIC testing and sequence analysis, ensuring representation 199

from each azole positive soil samples. Seven control isolates were randomly selected for further 200

analyses, representing each sampling site. All 26 isolates were identified as A. fumigatus by sequencing 201

of the beta-tubulin gene. 202

203

MIC measurement 204

All 19 azole-tolerant isolates were confirmed to be resistant to at least one of the three medical azoles 205

tested (Table 2). Eleven of these exhibited resistance to itraconazole only and were susceptible to both 206

voriconazole and posaconazole but with elevated MICs. Six isolates were resistant to voriconazole and 207

four of them were also resistant to itraconazole, with the remaining two having intermediate resistance 208


https://doi.org/10.1101/598961


against itraconazole. Isolate L2731 was cross-resistant against itraconazole and posaconazole, whereas 209

isolate L3131 was resistant to itraconazole and had intermediate resistance toward posaconazole. 210

Interestingly, isolates from the same soil sample yielded different susceptibility results. Among the 211

seven azole-susceptible isolates, six of them were confirmed to be susceptible to all three medical 212

azoles. One isolate (F311), exhibited resistance to itraconazole and elevated MIC values for voriconazole 213

and posaconazole, 0.25 g/ml and 0.125 g/ml, respectively. 214

215

Molecular determination of the azole resistance mechanisms 216

The molecular mechanisms of azole resistance were determined by sequencing part of the cyp51A gene 217

and its promoter region. Of the 19 azole-resistant isolates, 13 were found to be harbouring the 218

TR34/L98H allele and six the TR46/Y121F/T289A allele (Table 2). All six azole susceptible isolates had the 219

wild-type cyp51A allele. Both the negative and positive control isolates (Af293 and BUU09) carried the 220

wild-type and TR34/L98H allele respectively. 221

222

Molecular characterisation of A. fumigatus 223

Sequence typing of csp1 yielded 9 different csp1 types (Table 2). The six azole-susceptible wild-type 224

environmental isolates were distributed over csp1 subtypes t03 (1/6), t04A (3/6), t05 (1/6) and t08 (1/6). 225

The 20 azole-resistant isolates were spread across five csp1 subtypes, t01 (2/20), t02 (14/20), t04B 226

(2/20), t09 (1/20) and t18 (1/20). All the TR34/L98H and TR46/Y121F/T289A isolates predominantly 227

belonged to csp1 subtype t02. The 13 isolates with the TR34/L98H allele were distributed over three csp1 228

subtypes t01 (1/13), t02 (10/13) and t04B (2/13). The isolates carrying TR46/Y121F/T289A allele were 229

grouped into four csp1 subtypes: t01 (1/6), t02 (3/6), t09 (1/6) and t18 (1/6). The control isolate Af293, 230

which was isolated from a patient with invasive aspergillosis belonged to csp1 subtype t06 and the azole 231


https://doi.org/10.1101/598961


resistant control isolate BUU09 belonged to the t04B csp1 subtype, which agreed with previously 232

published results (24). 233

234


https://doi.org/10.1101/598961


Discussion 235

Given that ARAf continues to emerge globally, it is imperative that environmental monitoring, 236

particularly in previously unsampled locations, is sustained. Here we present environmental sampling of 237

London and the south east of England, home to approximately 15 million people living both urban and 238

rural lifestyles. We show that azole-resistant A. fumigatus was detected in 6.7% (12/178) of the soil 239

samples collected, and in 9.2% (12/131) of the soil samples positive for A. fumigatus growth. The 240

prevalence of ARAf in the environment was 13.8% in urban areas (10/72), which included Greater 241

London and Bath. In contrast, zero resistant isolates were found in soil samples collected from 242

agricultural land and just two resistant isolates were found in rural samples collected from non-243

agricultural land (1.1%). We also report the UK's first environmental ARAf isolates confirmed to be 244

carrying the TR46/Y121F/T239A resistance allele, alongside an expanded distribution of the TR34/L98H 245

allele. 246

247

Prevalence of ARAf in the UK was found to be lower than that of other European countries and 248

Colombia (15, 17, 25–29), but higher than most Asian countries (24, 30–33), with the exception of India 249

(29, 34). Our findings are also in close agreement with a recent UK-based environmental prevalence 250

study in Wales, where ARAf was detected in 4.5% (30/671) of soil samples, with resistance 251

predominantly found in urban city locations (13). However, they also describe a notable contrast to the 252

only other UK based study, which found zero ARAf in urban locales but four resistant isolates from 253

agricultural sites (1.7%) (14). 254

255

Our findings, and that of the Welsh study (13), appear to contradict the hypothesis that UK ARAf is 256

driven by the environmental application of azoles in arable agriculture (35, 36). Indeed, of the 53 257

samples collected directly on or surrounding agricultural land, zero azole-tolerant isolates were 258


https://doi.org/10.1101/598961


identified. Rather, we found that the prevalence of resistance was higher in urban city centres, 259

specifically flower beds and gardens, a finding that lends itself more readily to the hypothesis that the 260

expanding range of ARAf stems more from the distribution and cultivation of horticultural crops, such as 261

flowers, ornamentals and vegetables (15–17) 262

263

One particularly concerning discovery was the repeat isolation of ARAf genotypes – TR34/L98H and 264

TR46/Y121F/T289A – from flower beds surrounding city centre hospitals. Owing to the opportunistic 265

nature of A. fumigatus, its ability to cause debilitating illnesses in immunocompromised patients and the 266

elevated mortality that is associated with IA caused by azole-resistant genotypes (22), the gardened 267

areas around hospitals could be considered high risk locations if ARAf is present. Concern over the use of 268

azole treated flower bulbs in hospital environments has been raised previously (15, 16, 37), and 269

although we do not link any cases of azole-resistant aspergillosis to the isolates found in this study, our 270

findings do add to a worrying trend of ARAf populated soil sampled from flower beds (15, 17). 271

272

The continued emergence of ARAf in urban city centres fits with the observation that most isolates 273

belonged to a single csp1 subtype (t02), a pattern consistent with the selective sweep of drug resistant 274

genotypes (38). However, despite this obvious pattern of selection, city centres still harboured greater 275

diversity than rural locations. All csp1 subtypes identified during this study were found in Central 276

London, suggesting an anthropological driver facilitating the migration of many A. fumigatus genotypes 277

into a densely populated metropolis. Consequently, urban city centres, where A. fumigatus diversity is 278

high, could facilitate a ‘melting pot’, where resistance alleles inadvertently brought into the city are 279

given the opportunity to introgress onto novel genetic backgrounds via recombination, increasing the 280

diversity potential of ARAf. Indeed, in London alone, TR34/L98H was found on two csp1 subtypes and 281

TR46/Y121F/T298A on four CSP subtypes. 282


https://doi.org/10.1101/598961


283

Ultimately, this study highlights the importance of urban environments in the epidemiology of ARAf. We 284

have shown that although ARAf appears to be environmental by origin, urban city centres, that are 285

densely populated, are of high importance when mitigating strategies are considered. The use of azole 286

treated bulbs in the environment around hospitals should be reconsidered and the continuous global 287

monitoring of ARAf is imperative. 288

289


https://doi.org/10.1101/598961


Acknowledgments 290

The authors would like to thank Ali Abdolrasouli, Darius Armstrong-James and Andrew 291

Scourfield for their helpful discussions whilst analysing the data. We also acknowledge joint 292

Centre funding from the UK Medical Research Council and Department for International 293

Development. Funding: T.R.S., A.P.B., J.R. and M.C.F. were supported by the Natural 294

Environmental Research Council (NERC; NE/P001165/1) and all authors were supported by the 295

Medical Research Council (MRC; MR/R015600/1). 296

297


https://doi.org/10.1101/598961


References 298 299 1. Mullins J, Harvey R, Seaton A. 1976. Sources and incidence of airborne Aspergillus fumigatus 300

(Fres). Clin Exp Allergy 6:209–217. 301

2. Samson RA, Hong S, Peterson SW, Frisvad JC, Varga J. 2007. Polyphasic taxonomy of Aspergillus 302

section Fumigati and its teleomorph Neosartorya. Stud Mycol 59:147–203. 303

3. Kwon-Chung KJ, Sugui JA. 2013. Aspergillus fumigatus-What makes the species a ubiquitous 304

human fungal pathogen? PLoS Pathog 9:1–4. 305

4. Brown GD, Denning DW, Gow NAR, Levitz SM, Netea MG, White TC. 2012. Hidden killers: Human 306

fungal infections. Sci Transl Med 4:165rv13. 307

5. Dagenais TRTT, Keller NP. 2009. Pathogenesis of Aspergillus fumigatus in invasive aspergillosis. 308

Clin Microbiol Rev 22:447–465. 309

6. Kosmidis C, Denning DW. 2015. The clinical spectrum of pulmonary aspergillosis. Thorax 70:270–310

277. 311

7. Alcazar-Fuoli L, Mellado E. 2012. Ergosterol biosynthesis in Aspergillus fumigatus: Its relevance as 312

an antifungal target and role in antifungal drug resistance. Front Microbiol 3:1–6. 313

8. Parker JE, Warrilow AGS, Price CL, Mullins JGL, Kelly DE, Kelly SL. 2014. Resistance to antifungals 314

that target CYP51. J Chem Biol 7:143–161. 315

9. Fisher MC, Hawkins NJ, Sanglard D, Gurr SJ. 2018. Worldwide emergence of resistance to 316

antifungal drugs challenges human health and food security. Science 360:739–742. 317

10. Verweij PE, Snelders E, Kema GH, Mellado E, Melchers WJ. 2009. Azole resistance in Aspergillus 318

fumigatus: a side-effect of environmental fungicide use? Lancet Infect Dis 9:789–795. 319

11. Mellado E, Garcia-Effron G, Alcázar-Fuoli L, Melchers WJG, Verweij PE, Cuenca-Estrella M, 320

Rodríguez-Tudela JL. 2007. A new Aspergillus fumigatus resistance mechanism conferring in vitro 321

cross-resistance to azole antifungals involves a combination of cyp51A alterations. Antimicrob 322


https://doi.org/10.1101/598961


Agents Chemother 51:1897–904. 323

12. Van Der Linden JWM, Camps SMT, Kampinga GA, Arends JPA, Debets-Ossenkopp YJ, Haas PJA, 324

Rijnders BJA, Kuijper EJ, Van Tiel FH, Varga J, Karawajczyk A, Zoll J, Melchers WJG, Verweij PE. 325

2013. Aspergillosis due to voriconazole highly resistant Aspergillus fumigatus and recovery of 326

genetically related resistant isolates from domiciles. Clin Infect Dis 57:513–520. 327

13. Tsitsopoulou A, Posso R, Vale L, Bebb S, Johnson E, White PL. 2018. Determination of the 328

prevalence of triazole resistance in environmental Aspergillus fumigatus strains isolated in South 329

Wales, UK. Front Microbiol 9:1395. 330

14. Bromley MJ, van Muijlwijk G, Fraczek MG, Robson G, Verweij PE, Denning DW, Bowyer P. 2014. 331

Occurrence of azole-resistant species of Aspergillus in the UK environment. J Glob Antimicrob 332

Resist 2:276–279. 333

15. Snelders E, Huis In 't Veld RA, Rijs AJ, Kema GH, Melchers WJ, Verweij PE. 2009. Possible 334

environmental origin of resistance of Aspergillus fumigatus to medical triazoles. Appl Environ 335

Microbiol 75:4053–4057. 336

16. Dunne K, Hagen F, Pomeroy N, Meis JF, Rogers TR. 2017. Intercountry transfer of triazole-337

resistant Aspergillus fumigatus on plant bulbs. Clin Infect Dis 65:147–149. 338

17. Alvarez-Moreno C, Lavergne R-A, Hagen F, Morio F, Meis JF, Le Pape P. 2017. Azole-resistant 339

Aspergillus fumigatus harboring TR34/L98H, TR46/Y121F/T289A and TR53 mutations related to 340

flower fields in Colombia. Sci Rep 7:45631. 341

18. Jeanvoine A, Rocchi S, Reboux G, Crini N, Crini G, Millon L. 2017. Azole-resistant Aspergillus 342

fumigatus in sawmills of Eastern France. J Appl Microbiol 123:172–184. 343

19. Abdolrasouli A, Scourfield A, Rhodes J, Shah A, Elborn JS, Fisher MC, Schelenz S, Armstrong-James 344

D. 2018. High prevalence of triazole resistance in clinical Aspergillus fumigatus isolates in a 345

specialist cardiothoracic centre. Int J Antimicrob Agents 52:637–642. 346


https://doi.org/10.1101/598961


20. Wiederhold NP, Gil VG, Gutierrez F, Lindner JR, Albataineh MT, McCarthy DI, Sanders C, Fan H, 347

Fothergill AW, Sutton DA. 2016. First detection of TR34/L98H and TR46/Y121F/T289A Cyp51 348

mutations in Aspergillus fumigatus isolates in the United States. J Clin Microbiol 54:168–171. 349

21. Dauchy C, Bautin N, Nseir S, Reboux G, Wintjens R, Le Rouzic O, Sendid B, Viscogliosi E, Le Pape P, 350

Arendrup MC, Gosset P, Fry S, Fréalle E. 2018. Emergence of Aspergillus fumigatus azole 351

resistance in azole-naïve patients with chronic obstructive pulmonary disease and their homes. 352

Indoor Air 28:298–306. 353

22. Lestrade PP, Bentvelsen RG, Schauwvlieghe AFAD, Schalekamp S, van der Velden WJFM, Kuiper 354

EJ, van Paassen J, van der Hoven B, van der Lee HA, Melchers WJG, de Haan AF, van der Hoeven 355

HL, Rijnders BJA, van der Beek MT, Verweij PE. 2018. Voriconazole resistance and mortality in 356

invasive Aspergillosis: A multicenter retrospective cohort study. Clin Infect Dis. 357

23. Klaassen CHW, de Valk HA, Balajee SA, Meis JFGM. 2009. Utility of CSP typing to sub-type clinical 358

Aspergillus fumigatus isolates and proposal for a new CSP type nomenclature. J Microbiol 359

Methods 77:292–296. 360

24. Tangwattanachuleeporn M, Minarin N, Saichan S, Sermsri P, Mitkornburee R, Groß U, 361

Chindamporn A, Bader O. 2016. Prevalence of azole-resistant Aspergillus fumigatus in the 362

environment of Thailand. Med Mycol 55:myw090. 363

25. Bader O, Tünnermann J, Dudakova A, Tangwattanachuleeporn M, Weig M, Groß U. 2015. 364

Environmental Isolates of Azole-Resistant Aspergillus fumigatus in Germany. Antimicrob Agents 365

Chemother 59:4356–4359. 366

26. Riat A, Plojoux J, Gindro K, Schrenzel J, Sanglard D. 2018. Azole resistance of environmental and 367

clinical Aspergillus fumigatus isolates from Switzerland. Antimicrob Agents Chemother 368

62:e02088-17. 369

27. Mortensen KL, Mellado E, Lass-Flörl C, Rodriguez-Tudela JL, Johansen HK, Arendrup MC. 2010. 370


https://doi.org/10.1101/598961


Environmental study of azole-resistant Aspergillus fumigatus and other aspergilli in Austria, 371

Denmark, and Spain. Antimicrob Agents Chemother 54:4545–9. 372

28. Prigitano A, Venier V, Cogliati M, De Lorenzis G, Esposto MC, Tortorano AM. 2014. Azole-resistant 373

Aspergillus fumigatus in the environment of northern Italy, May 2011 to June 2012. Euro Surveill 374

19:20747. 375

29. Sharma C, Hagen F, Moroti R, Meis JF, Chowdhary A. 2015. Triazole-resistant Aspergillus 376

fumigatus harbouring G54 mutation: Is it de novo or environmentally acquired? J Glob 377

Antimicrob Resist 3:69–74. 378

30. Chowdhary A, Sharma C, van den Boom M, Yntema JB, Hagen F, Verweij PE, Meis JF. 2014. Multi-379

azole-resistant Aspergillus fumigatus in the environment in Tanzania. J Antimicrob Chemother 380

69:2979–83. 381

31. Badali H, Vaezi A, Haghani I, Yazdanparast SA, Hedayati MT, Mousavi B, Ansari S, Hagen F, Meis 382

JF, Chowdhary A. 2013. Environmental study of azole-resistant Aspergillus fumigatus with 383

TR34/L98H mutations in the cyp51A gene in Iran. Mycoses 56:659–663. 384

32. Ahmad S, Khan Z, Hagen F, Meis JF. 2014. Occurrence of triazole-resistant Aspergillus fumigatus 385

with TR34/L98H mutations in outdoor and hospital environment in Kuwait. Environ Res 133:20–386

26. 387

33. Ahmad S, Joseph L, Hagen F, Meis JF, Khan Z. 2015. Concomitant occurrence of itraconazole-388

resistant and -susceptible strains of Aspergillus fumigatus in routine cultures. J Antimicrob 389

Chemother 70:412–5. 390

34. Chowdhary A, Sharma C, Kathuria S, Hagen F, Meis JF. 2014. Azole-resistant Aspergillus fumigatus 391

with the environmental TR46/Y121F/T289A mutation in India. J Antimicrob Chemother 69:555–392

557. 393

35. Berger S, Chazli Y El, Babu AF, Coste AT. 2017. Azole resistance in Aspergillus fumigatus: A 394


https://doi.org/10.1101/598961


consequence of antifungal use in agriculture? Front Microbiol 8:1–6. 395

36. Chowdhary A, Kathuria S, Xu J, Meis JF. 2013. Emergence of azole-resistant Aspergillus fumigatus 396

strains due to agricultural azole use creates an increasing threat to human health. PLoS Pathog 397

9:3–7. 398

37. Chowdhary A, Kathuria S, Xu J, Sharma C, Sundar G, Singh PK, Gaur SN, Hagen F, Klaassen CH, 399

Meis JF. 2012. Clonal expansion and emergence of environmental multiple-triazole-resistant 400

Aspergillus fumigatus strains carrying the TR34/L98H mutations in the cyp51A Gene in India. PLoS 401

One 7:e52871. 402

38. Abdolrasouli A, Rhodes J, Beale M, Hagen F, Rogers TR, Chowdhary A, Meis JF, Fisher MC. 2015. 403

Genomic context of azole-resistance mutations in Aspergillus fumigatus using whole-genome 404

sequencing. MBio 6:1–11. 405

406


https://doi.org/10.1101/598961


Figure legends 407

408

Fig 1: Distribution of azole susceptible and resistant Aspergillus fumigatus isolates collected in London 409

(b) and southern UK counties (a). Isolates haboring cyp51a allele TR34/L98H are depicted in red, 410

TR46/Y121F/T289A in blue and wild-type in green. 411

412


https://doi.org/10.1101/598961


Author Contributions 413

T.R.S., Y.Z. and M.C.F. conceived and designed the study. T.R.S., Y.Z., A.P.B. and J.M.G.S. 414

collected the data. T.R.S., Y.Z. and A.P.B. analyzed the data. T.R.S. and Y.J. wrote the 415

manuscript. T.R.S., Y.Z., A.P.B., J.M.G.S., J.R., and M.C.F. discussed the results and commented 416

on the manuscript417


https://doi.org/10.1101/598961


Tables 418 419 Table 1. Soil sampling areas and A. fumigatus recovery rate, with azole resistance 420 determined by MIC and cyp51A sequencing. 421

Sampling site Samples collected

(n)

Number of samples with A.

fumigatus growth (%)

Number of samples

recovered azole-resistant A. fumigatus a

Prevalence of azole-

resistant A. fumigatus

London Overall 64 60 (93.8%) 9 14.0%

Kensal Green Cemetery 5 5 (100%) 0 0% Queen Mary's Rose Garden & Regents Park

8 7 (87.5%) 0 0%

The Whittington Hospital & Waterlow Park

5 5 (100%) 3 60.0%

Hampstead Heath 3 1 (33.3%) 0 0% Royal Free Hospital 5 5 (100%) 3 60.0% Green Park & Hyde Park 9 9 (100%) 1 11.1% Charing Cross Hospital & Margravine Cemetery

8 7 (87.5%) 0 0%

Brompton Cemetery 6 6 (100%) 1 16.7% Elstree Open Space 8 8 (100%) 1 12.5% Flower beds in urban city 7 7 (100%) 0 0%

Bath city centre, Somerset Flower beds 8 4 (50%) 1 12.5%

Stonehenge, Wiltshire Flower beds and farm 8 6 (75%) 1 12.5%

New Forest National Park, Hampshire Remote forest 46 37 (80.4%) 1 2.2%

Surrey Lavender farm 13 10 (76.9%) 0 0%

Cambridgeshire Wheat farm 30 12 (40%) 0 0%

Open space surrounded by farm 9 2 (22.2%) 0 0%

COMBINED TOTAL 178 131 (73.6%) 12 6.7%

a. Resistance determined by both MIC testing and sequencing analysis. 422 423


https://doi.org/10.1101/598961


Table 2. Characteristics of azole-resistant A. fumigatus isolates and the control isolates. 424

Soil sample

Sampling site Isolate MIC (g/ml)a cyp51A mutation

csp1 type

ITC VRC POS

L10 Waterlow Park, London

L1031 2 8 0.0625 TR46/Y121F/T239A t02 L1032 4 4 0.125 TR46/Y121F/T239A t02 L1041 4 8 0.125 TR46/Y121F/T239A t02 L1043 4 0.25 0.125 TR34/L98H t02

L27 Waterlow Park, London

L2731 4 1 0.5 TR34/L98H t02

L2741 4 8 0.125 TR46/Y121F/T239A t09

L28 Whittington Hospital, London

L2831 4 8 0.125 TR46/Y121F/T239A t01

L2841 4 0.25 0.125 unknownb t02

L12 Royal Free Hospital, London

L1231 4 0.5 0.125 TR34/L98H t02

L1241 4 0.25 0.125 TR34/L98H t02


L2931 4 0.125 0.0625 TR34/L98H t01


L3131 4 1 0.25 TR34/L98H t02 L3141 4 0.25 0.125 TR34/L98H t02

L19 Hyde Park, London L1931 2 4 0.0625 TR46/Y121F/T239A t18

L41 Brompton Cemetery, London

L4131 4 0.25 0.125 TR34/L98H t02

BS1 Bath city center BS131 4 0.25 0.125 TR34/L98H t02

BS9

Stonehenge, Wiltshire

BS941 4 0.125 0.0625 TR34/L98H t02 BS942 4 0.125 0.125 TR34/L98H t02

NF63 New Forest, Hampshire

NF6341 4 0.25 0.125 TR34/L98H t04B

F3 Elstree Open Space, London

F311 4 0.25 0.125 TR34/L98H t04B

F44 Lavender farm, Surrey

F4411 0.125

0.031

0.016

wild type t05

F55 Wheat farm, Cambridge

F5511 0.25 0.063 0.031 wild type t04A


L1111 0.25 0.031 0.016 wild type t04A

L42 Brompton Cemetery, London

L4211 0.125 0.031 0.031 wild type t03

BS10 Stonehenge, Wiltshire

BS1011 0.125 0.063 0.031 wild type t04A

NF42 New Forest, Hampshire

NF4211 0.125 0.063 0.016 wild type t08

wild-type control Af293 0.25 0.063 0.016 wild-type t06 azole-resistant control BUU09 4 0.25 0.125 TR34/L98H t04B

a. ITC, itraconazole; VRC, voriconazole; POS, posaconazole. 425 b. Isolate L2841 amplified poorly using the primers in this study. 426


https://doi.org/10.1101/598961


50.5

51.0

51.5

52.0

−3 −2 −1 0

Allele

51.50

51.55

Allele

Allele

TR34/L98H

TR46/Y121F/T239A

Wild type


https://doi.org/10.1101/598961
