key: cord-1023827-xr2kluw9
authors: Viciani, Elisa; Gaibani, Paolo; Castagnetti, Andrea; Liberatore, Andrea; Bartoletti, Michele; Viale, Pierluigi; Lazzarotto, Tiziana; Ambretti, Simone; Lewis, Russell; Cricca, Monica
title: Critically ill patients with COVID-19 show lung fungal dysbiosis with reduced microbial diversity in Candida spp colonized patients
date: 2022-02-09
journal: Int J Infect Dis
DOI: 10.1016/j.ijid.2022.02.011
sha: 6a7f15986f798a7b90654a644f6fce64e8e5b9d2
doc_id: 1023827
cord_uid: xr2kluw9

Objectives The COVID-19 pandemic has intensified interest in how the infection impacts the lung microbiome of critically ill patients and contributes to acute respiratory distress syndrome (ARDS). We aimed to characterize the lower respiratory tract mycobiome of COVID-19 critically ill patients in comparison to COVID-19-negative patients. Methods We performed an Internal transcribed spacer 2 (ITS2) profiling, with the Illumina MiSeq platform, on 26 respiratory specimens from COVID-19 positive patients as well as from 26 patients with non-COVID-19 pneumonia. Results COVID-19+ patients were more likely to be colonized with Candida spp. and ARDS was associated with lung dysbiosis characterised by a shift to Candida species colonisation and a decrease of fungal diversity. We also observed higher bacterial phylogenetic distance among taxa in COVID-19+ colonized patients. In COVID-19+ patients non-colonized with Candida spp, ITS2 amplicon sequencing revealed an increase of Ascomycota unassigned spp. and one Aspergillus spp positive specimen. Then, we found that corticosteroid therapy was frequently associated with positive Galactomannan cell wall component of Aspergillus spp among COVID-19+ patients. Conclusions Our study underpins that ARDS in COVID-19+ patients is associated with lung dysbiosis and that an increased density of Ascomycota unassigned spp. is present in patients not colonized with Candida spp.

Coronaviruses are important human and animal pathogens. At the end of 2019, a novel coronavirus was identified as the causative agent of a pneumonia outbreak in Wuhan, China that subsequently spread worldwide in a global pandemic. Globally there have been more than 240 million confirmed cases of COVID-19, including nearly 5 million deaths (https://www.who.int/). COVID-19 critically ill patients may develop acute respiratory distress syndrome (ARDS), requiring admission in intensive care unit (ICU) and mechanical ventilation, which predisposes them to bacterial and fungal superinfections , Lansbury et al. 2020 , Chong et al. 2021 . Candida is one of the most frequently isolated pathogens in ICU, affecting between 6% and 10% of patients (Zhang et al. 2020 , Koehler et al. 2019 . The estimated mortality rate attributed to invasive candidiasis is 19-40%, this mortality is even higher among ICU patients, approaching 70% (Kullberg et al. 2015 , Marra et al. 2011 . Recently, it has been reported that fungi are more frequently detected among SARS-CoV-2 positive patients, with Candida albicans the most frequently isolated yeast (Calderaro et al. 2021) . Furthermore, the wide use of antibiotics and corticosteroids along with the damage exerted by SARS CoV-2, may allow commensal yeast to invade internal organs causing deeply invasive infections (Arastehfar et al. 2020 , Talento et al. 2020 , Posteraro et al 2020 .

Despite the low number of studies on lung mycobiome, growing evidence indicates that the fungal microbiota is altered in critically-ill patients (Krause et al. 2016) . Fungi found in the human respiratory tract are predominantly from the Dikarya sub-kingdom which is composed of the phyla Ascomycota and Basidiomycota. In healthy individuals, the fungal burden is generally low and the mycobiome appears to be largely composed of environmental fungi or fungi disseminating from the oral cavity . By contrast, more stable fungal communities can colonize the lung when its physiology is altered. As an example, in most patients with cystic fibrosis, the fungal burden is increased, whereas alpha diversity is reduced and correlates with disease severity (Iliev et al. 2017) . Despite the speculated importance of lung dysbiosis in the genesis of both ventilator-associated pneumonia (VAP) and acute respiratory distress syndrome (ARDS), few studies have examined the lung mycobiome in these patient populations (Krause et al. 2016 . In the lung of COVID-19-positive patients fungal colonization/ infection represents a major concern although the clinical significance is debated (Peng et al. 2021) . Indeed, Candida albicans has been reported as the most frequently isolated yeast from the lung, while COVID-19associated pulmonary aspergillosis (CAPA) has been reported in a few centers (Bartoletti et al. 2020 , Permpalung et al. 2021 , Borman et al. 2020 . Despite this fact, no data concerning the lung mycobiome in COVID-19 ill patients with ARDS has been reported. The goal of this study is to analyze the composition of lung mycobiome in mechanically ventilated COVID-19-positive patients with ARDS.

Twenty-six COVID-19 positive (COVID-19+) patients and 26 COVID-19 negative were enrolled in this study, all the patients were recovered at IRCCS Sant'Orsola Malpighi University Hospital, Bologna, Italy, from March to April 2020. The study was conducted in accordance with the Declaration of Helsinki. Samples were coded and analysis was performed with anonymized database. Informed consent for study participation was obtained from each patient. The study was approved by the local IRB (Comitato Etico characteristics are presented in Table 1 .

The presence of SARS-CoV-2 was detected by reverse transcriptase (RT)-PCR assay.

Briefly, detection of SARS-CoV-2 was performed by real time RT-PCR following the World Health Organization and/ or Centers for Disease Control and Prevention protocols in a QuantStudio S5 Real-time PCR system (ThermoFisher). All COVID-19+ patients were managed in a dedicated COVID-19 ICU and underwent mechanical ventilation (Table 1) .

The control group included 26 patients admitted to hospital presenting clinical and radiological findings of pneumonia and with PCR performed on nasopharyngeal swab or BAL negative for SARS-CoV-2. Radiological findings of the control group were consistent with interstitial pneumonia in 9 (35%) patients, or patchy multifocal infiltrates in 8 (30%) cases or other findings in the remaining ones (35%). In our population, a total of 8/26 COVID-19 associated pulmonary aspergillosis (CAPA) were classified as probable, according to recent ECMM/ISHAM consensus criteria (Koehler et al. 2020) . Direct microscopy or biopsy, which are the microbiological criteria to classify CAPA as proven,

were not performed in our study. Six out of 26 COVID-19patients were mechanically ventilated and 3 were admitted in ICU (Table 1) .

BALs were cultured for the isolation of fungal and bacterial pathogens at 32°C for 5 days for filamentous fungi and at 37°C for 2 days for yeasts and bacterial pathogens. In particular, selective media such as CHROMagar Candida and Sabouraud-dextrose (SAB) chloramphenicol agar (Vacutest Kima, Italy) as well as non-selective media such as SAB were used for in vitro fungal growth. Selective media such as Agar Herellea, Agar saltmannitol, agar chocolate haemophilus and non-selective media such as sheep blood agar were used for bacterial growth. The level of galactomannans (GMs), which are fungal antigens, was measured with a sandwich enzyme-linked immunosorbent assay (ELISA; Platelia Aspergillus; Bio-Rad Laboratories) in BAL specimens following manufacturer's instructions.

DNA extraction from 1 ml volume of respiratory material (diluted 1:1 with dithiothreitol) and

Real time PCR for Aspergillus spp targeting rDNA 18S (ELITe MGB kit, Elitech Group, Italy) were performed on ELITe InGenius automated platform. DNA was eluted in 100 µl volume and the DNA copy number was expressed as copies/ml. Negative and positive controls were used in each run as well as standard 10-fold dilutions of Aspergillus DNA.

The lower respiratory mycobiome was characterized both in COVID-19 critically ill patients as well as in non-infected patients. The lower respiratory bacterial microbiome of these samples has been previously analysed by Gaibani et al. (Gaibani et al. 2021) . Total microbial DNA was extracted from BAL samples using the QIAamp 96 PowerFecal QIAcube HT kit on the QIAcube HT instrument (QIAGEN, Hilden, Germany) following the manufacturer's instructions. A bead-beating step with Lysing Matrix E (MP Biomedicals) was performed on a FastPrep24 bead-beater (MP Biomedicals, Irvine, CA) at 6.0 movements per second for 40 seconds, before total DNA extraction. Negative controls were PCR-grade water which underwent library preparation steps and Next Generation Sequencing (NGS). DNA was quantified using the Qubit™ 4 Fluorometer (Fisher Scientific). Internal transcribed spacer 2 (ITS2) was amplified using the primer set ITS3: 5'-GCATCGATGAAGAACGCAGC -3' and ITS4: 5'-TCCTCCGCTTATTGATATGC -3' (White et al. 1990 ). PCR products were purified with a magnetic bead-based clean-up system (Agencourt AMPure XP; Beckman Coulter, Brea, CA). Indexed libraries were prepared by limited-cycle PCR using Nextera technology and further cleaned up with AMPure XP magnetic beads (Beckman Coulter).

Libraries were pooled at equimolar concentrations (4 nM), denatured, and diluted to 5 pM before loading onto the MiSeq flow cell. Sequencing on Illumina MiSeq platform was performed by using a 2 × 250 bp paired end protocol, according to the manufacturer's instructions (Illumina, San Diego, CA).

Paired-end sequenced reads of samples were analysed combining PANDAseq2 and QIIME2 version 2018.6 (Bolyen et al. 2018) . The Divisive Amplicon Denoising Algorithm 2 (DADA2) (Hall et al. 2018) plugin was used to remove noise, chimeras, and to generate ASVs (Amplicon Sequence Variants). Quality filtering, and clustering were performed using VSEARCH (Rognes et al. 2016) . High-quality reads were classified taxonomically using the UNITE reference database version 7.2 (UNITE Community (2017): UNITE QIIME release.

Version 01.12.2017. UNITE Community. https://doi.org/10.15156/BIO/587481). Samples that had less than 1,000 reads after Illumina MiSeq sequencing were discarded. The bacterial abundance data were imported into R (version 3.6.1) on Rstudio v1.1.456 where all statistical analyses were performed using R package phyloseq (McMurdie et al. 2016) . By using the decontam R package at 1% and 5% stringency (Davis et al. 2018 ) on our negative controls (ultrapure water samples which underwent the whole library preparation processes) we assessed the absence of detectable contaminant ASVs. We obtained 258 fungal taxa after quality filtering and the removal of unidentified fungal phyla. Sample sequencing reads were not rarefied to avoid introducing unwanted bias, since samples reached the total ASV number asymptote at around 800 reads, even when having higher sequencing read depths (Supplementary Fig S1) . The differences in alpha diversity were evaluated, based on the data distribution of metrics, using ANOVA and Tukey's HSD (honestly significant difference) tests for normally distributed data or Wilcoxon-Mann-Whitney with Holm-Bonferroni correction method for non-normally distributed data. To check that the non application of rarefaction would have not led us to inexact conclusions, alpha diversity analysis was repeated after rarefying to the lowest sequencing depth, and results matched the ones obtained without rarefaction (Supplementary Fig S2) . To compare microbial composition between samples, beta diversity was measured by calculating the Bray-Curtis distance matrix.

Principal coordinates analysis (PCoA) was applied on the distance matrices to generate bidimensional plots in R. Dispersion of the PCoA clusters was compared using the betadisper function in R vegan package (Anderson et al. 2013) . The permutational analysis of variance (PERMANOVA) test, calculated using the function adonis in the vegan package (Oksanen et al. 2014) , was performed to determine whether there was a significant separation between different sample groups. The plots were graphed using ggplot2 R packages (Wickham et al. 2019) .

Dissimilarity percentage (SIMPER) analysis function (https://github.com/asteinberger9/seq_scripts) based on R packages vegan and dplyr was used to determine the contribution of individual taxa driving the average dissimilarities between groups. A p-value < 0.05 after False Discovery Rate (FDR) correction was considered as statistically significant. Linear discriminant analysis (LDA) effect size (LEfSE) algorithm (Segata et al. 2011) , a tool which is hosted on the Galaxy web application at https://huttenhower.sph.harvard.edu/galaxy/, was also used to discover bacterial biomarkers associated to COVID-19 patients. The differences in abundance were regarded as significant when the logarithmic LDA score was higher than 2.

Given the fact that the microbial diversity of the mycobiome has not been well described in 

This study was performed in Italy during the COVID-19 major pandemic wave of March- Table S1) .

Comparisons on fungal microbial composition using ITS2 amplicon sequencing in BAL samples from COVID-19+ and COVID-19-patients were performed, but no significant differences were found between these groups. We then focused on the COVID-19+ individuals, where we detected microbial composition differences between the patients who were colonised by Candida spp. or not. The six COVID-19+ and Candida spp. colonized individuals were sex, age and treatment matched with the rest of the COVID-19+ samples ( Curtis dissimilarities = 0.002) (Fig.2) and a significantly higher relative abundance of Candida spp. (SIMPER, p-value FDR-corrected = 0.008) with respect to non-colonized patients. In contrast, non-colonized patients exhibited increased density of Ascomycota unclassified spp. (LDA Lefse > 2.00) and a negative association with Candida spp. (Fig. 3) .

Candida spp. colonized COVID-19+ patients show a higher bacterial phylogenetic distance among taxa (p-value, FDR-corrected = 0.049) (Fig.4) . However, no bacterial genera were specifically associated with Candida colonization.

Among COVID-19+ patients, the LDA LEfSE analysis found a significant (LDA > 2) increase in Ascomycota unclassified spp. in Candida spp. uncolonized patients, moreover of the COVID-19-patients was positive (Fig. 5) . Furthermore, the NGS extracts were further analysed by a target-specific PCR assay able to detect different species of Aspergillus genus.

In this way, we confirmed the positivity of case #7 (Fig. 5) and we found 1 positive case among COVID-19-patients (#1).

Considering that a main concern in COVID-19+ patients with ARDS was pulmonary aspergillosis (CAPA), we further analysed BAL specimens, collected during routine diagnostic procedures, for the presence of Aspergillus spp. by PCR. In total, we found 6 positive cases, including case #7 (Fig. 5) . These specimens were collected from 0 to 6 days after those used for NGS analysis. Interestingly, among the six COVID-19+ patients, two

were colonized by C. glabrata (Fig. 5 ).

The main finding of this study was that COVID-19 infection was associated with lung dysbiosis characterised by a shift to Candida species colonisation and a decrease in fungal diversity. We also observed higher bacterial phylogenetic distance among taxa in colonized patients.

Until now, few studies have evaluated the lung mycobiota using high-throughput sequencing (Soltani et al. 2021) . While in healthy individuals the fungal burden is generally low and lung mycobiota appears to be largely composed of environmental fungi such as Cladosporium, Aspergillus, Penicillium and yeasts belonging to the two main phyla Ascomycota and Basidiomycota, more stable fungal communities colonize the lung when its physiology is altered (Krüger et al. 2019) . As an example, alpha diversity is reduced and fungal burden is increased in patients with cystic fibrosis (Linnane et al. 2021) . Once established, the dysbiotic fungal communities seem to persist even in the presence of antibiotic or immunosuppressant therapy (Iliev et al. 2017) . Similarly, we did not find a specific association with Candida colonization of the respiratory tract and concurrent antibiotic therapy in COVID-19+ patients. However, patients with Candida colonization did demonstrate higher bacterial phylogenetic distance among taxa, which implies a higher ecological differentiation among bacterial metabolic functions. In our study, bacterial superinfections, as determined by culture, were significantly higher in COVID-19 positive patients versus negative ones, while no significant difference was present for the administration of broad-spectrum antibacterial therapy. In the end, we can't exclude that bacterial superinfections could have influenced the microbial niche and consequently the comparison of COVID-19+ vs COVID-19-in terms of Candida colonization. No significant difference was present in COVID-19+ versus COVID-19-patients in the time from symptoms onset to BALs collection as well as from hospitalization to BALs collections. So, the two population were homogeneous in their clinical characteristics with the exception of mechanical ventilation and ICU admission. We can't exclude that this fact might have had some impact on our findings when comparing COVID-19+ and COVID-19-populations.

COVID-19-associated pulmonary aspergillosis (CAPA) must be considered a serious and potentially life-threatening complication in patients with severe COVID-19 receiving immunosuppressive treatment (Machado et al. 2021 ). In fact, many studies reported an increase in COVID-19 associated pulmonary aspergillosis (CAPA) in intubated patients with COVID-19-related ARDS (Bartoletti et al. 2020 , Dewi et al. 2021 . In this study, we found that corticosteroid therapy was frequently associated with subsequent positive GM tests. ITS2

amplicon sequencing revealed an increase of Ascomycota unclassified spp. in COVID-19+ patients non-colonized with Candida spp., nevertheless only 1 patient was positive for Aspergillus spp. Considering Aspergillus spp. PCR results, on PCR dedicated DNA extracts collected from 1 to 6 days later than those for NGS, we detected an additional 6 positive cases in COVID-19+ patients that were not positive by NGS.

In conclusion, our study demonstrates that lung fungal dysbiosis is more severe in critically ill patients with COVID-19 infection. In particular, Candida spp. colonization was accompanied by a decreased diversity (richness and evenness of fungal taxa) of fungi in respiratory tract overall, while in non-Candida colonized patients, the mycobiome was characterized by a higher density of unclassified fungi from the Ascomycota phylum. CAPA is a major concern for COVID-19+ patients, and by NGS we found one case positive for Further studies are needed to confirm our results on this interesting topic.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. No animal studies are presented in this manuscript. NGS BAL: next generation sequencing on NGS dedicated extracted specimens PCR BAL: polymerase chain reaction on NGS dedicated extracted specimens Follow-up BAL PCR: polymerase chain reaction on PCR dedicated extracted specimens collected from 0 to 6 days after those for NGS 

ANOSIM, and the Mantel test in the face of heterogeneous dispersions: What null hypothesis are you testing?

An Underestimated Complication in the Absence of Immunological Predispositions? J Fungi (Basel)

Epidemiology of invasive pulmonary aspergillosis among COVID-19 intubated patients: a prospective study

Bacterial and fungal superinfections in critically ill patients with COVID-19

Associated Invasive Aspergillosis: Data from the UK National Mycology Reference Laboratory

Detection of SARS-CoV-2 and Other Infectious Agents in Lower Respiratory Tract Samples Belonging to Patients Admitted to Intensive Care Units of a Tertiary-Care Hospital, Located in an Epidemic Area, during the Italian Lockdown. Microorganisms

State-of-the-art review of secondary pulmonary infections in patients with COVID-19 pneumonia

Simple statistical identification and removal of contaminant sequences in marker-gene and metagenomics data

Invasive pulmonary aspergillosis associated with viral pneumonitis

The Lung Microbiome: New Principles for Respiratory Bacteriology in Health and Disease

Homeostasis and its disruption in the lung microbiome

Reconsidering ventilator-associated pneumonia from a new dimension of the lung microbiome

Respiratory microbiome in mechanically ventilated patients: a narrative review

The lower respiratory tract microbiome of critically ill patients with COVID-19. Sci Rep

16S rRNA Gene Analysis with QIIME2

Fungal dysbiosis: immunity and interactions at mucosal barriers

Defining and managing COVID-19-associated pulmonary aspergillosis: the 2020 ECMM/ISHAM consensus criteria for research and clinical guidance

Morbidity and mortality of candidaemia in Europe: an epidemiologic meta-analysis

Zollner-Schwetz I. Characterisation of Candida within the Mycobiome/Microbiome of the Lower Respiratory Tract of ICU Patients

Fungal-Bacterial Interactions in Health and Disease. Pathogens

Invasive Candidiasis

Co-infections in people with COVID-19: a systematic review and meta-analysis

The Lung Microbiome in Young Children with Cystic Fibrosis: A Prospective Cohort Study. Microorganisms

COVID-19 Study Group. Invasive pulmonary aspergillosis in the COVID-19 era: An expected new entity

Brazilian SCOPE Study Group. Nosocomial bloodstream infections in Brazilian hospitals: analysis of 2,563 cases from a prospective nationwide surveillance study

phyloseq: An R Package for Reproducible Interactive Analysis and Graphics of Microbiome Census Data

Bioconductor workflow for microbiome data analysis: from raw reads to community analyses

Vegan: Community Ecology Package. R Package Version 2.2-0

Candida colonization in mechanically ventilated patients

Fungal co-infection in COVID-19 patients: evidence from a systematic review and meta-analysis

COVID-19 Associated Pulmonary Aspergillosis in Mechanically Ventilated Patients

Pan-Echinocandin-Resistant Candida glabrata Bloodstream Infection Complicating COVID-19: A Fatal Case Report. J Fungi (Basel)

VSEARCH: A versatile open source tool for metagenomics

Metagenomic Biomarker Discovery and Explanation

The Role of Bacterial and Fungal Human Respiratory Microbiota in COVID-19 Patients

Fungal Infections Complicating COVID-19: With the Rain Comes the Spores. J Fungi (Basel)

Impact of Candida spp. isolation in the respiratory tract in patients with intensive care unit-acquired pneumonia

Amplification and Direct Sequencing of

Fungal Ribosomal RNA Genes for Phylogenetics

ggplot2: elegant graphics for data analysis

Risk of invasive candidiasis with prolonged duration of ICU stay: a systematic review and meta-analysis