ABSTRACT

   Antimicrobial resistance is a global concern in chronic respiratory
   diseases, including chronic obstructive pulmonary disease (COPD). The
   collection of antibiotic resistance genes or resistome in human airways
   may underlie the resistance. COPD is heterogeneous, and understanding
   the airway resistome in relation to patient phenotype and endotype may
   inform precision antibiotic therapy. Here, we characterized the airway
   resistome for 94 COPD participants at stable disease. Among all
   demographic and clinical factors, patient inflammatory endotype was
   associated with the airway resistome. There were distinct resistome
   profiles between patients with neutrophilic or eosinophilic
   inflammation, two primary inflammatory endotypes in COPD. For
   neutrophil-predominant COPD, the resistome was dominated by multidrug
   resistance genes. For eosinophil-predominant COPD, the resistome was
   diverse, with an increased portion of patients showing a macrolide-high
   resistome. The differential antimicrobial resistance pattern was
   validated by sputum culture and in vitro antimicrobial susceptibility
   testing. Ralstonia and Pseudomonas were the top contributors to the
   neutrophil-associated resistome, whereas Campylobacter and
   Aggregatibacter contributed most to the eosinophil-associated
   resistome. Multiomic analyses revealed specific host pathways and
   inflammatory mediators associated with the resistome. The arachidonic
   acid metabolic pathway and matrix metallopeptidase 8 (MMP-8) exhibited
   the strongest associations with the neutrophil-associated resistome,
   whereas the eosinophil chemotaxis pathway and interleukin-13 (IL-13)
   showed the greatest associations with the eosinophil-associated
   resistome. These results highlight a previously unrecognized link
   between inflammation and the airway resistome and suggest the need for
   considering patient inflammatory subtype in decision-making about
   antibiotic use in COPD and broader chronic respiratory diseases.

   IMPORTANCE Antibiotics are commonly prescribed for both acute and
   long-term prophylactic treatment in chronic airway disorders, such as
   chronic obstructive pulmonary disease (COPD), and the rapid growth of
   antibiotic resistance is alarming globally. The airway harbors a
   diverse collection of microorganisms known as microbiota, which serve
   as a reservoir for antibiotic resistance genes or the resistome. A
   comprehensive understanding of the airway resistome in relation to
   patient clinical and biological factors may help inform decisions to
   select appropriate antibiotics for clinical therapies. By deep
   multiomic profiling and in vitro phenotypic testing, we showed that
   inflammatory endotype, the underlying pattern of airway inflammation,
   was most strongly associated with the airway resistome in COPD
   patients. There were distinct resistome profiles between
   neutrophil-predominant and eosinophil-predominant COPD that were
   associated with different bacterial species, host pathways, and
   inflammatory markers, highlighting the need of considering patient
   inflammatory status in COPD antibiotic management.

   KEYWORDS: airway microbiome, antibiotic resistance, COPD, inflammation,
   macrolides

OBSERVATION

   Chronic obstructive pulmonary disease (COPD) is one of the most
   prevalent respiratory diseases and is characterized by impaired lung
   function and persistent airway inflammation ([56]1). COPD is
   heterogeneous, and there is a growing consensus for COPD management
   based on patient phenotype, endotype, and treatable traits ([57]2,
   [58]3). Neutrophilic and eosinophilic inflammation are two major
   inflammatory endotypes with distinct pathogenesis, airway microbiota,
   and therapeutic options ([59]4). Antimicrobial resistance is a concern
   in COPD since antibiotics are required for acute and long-term
   prophylactic therapy ([60]5, [61]6). The COPD airway harbors a diverse
   microbiota that serve as a reservoir for antimicrobial resistance genes
   (ARGs) known as the resistome. As the airway resistome may underlie
   antibiotic resistance, understanding the resistome in relation to
   disease phenotype may help inform precision antibiotic therapy. Despite
   emerging efforts to unravel the airway resistome in chronic respiratory
   diseases ([62]7, [63]8), it remains unclear how the resistome is
   related to clinical characteristics and host pathophysiology, limiting
   its applicability to clinical practice. One study has characterized the
   COPD resistome using quantitative PCR and reported a limited
   association between ARGs and clinical parameters ([64]8). We
   hypothesize that the COPD airway resistome is inherently shaped by
   airway microbiota and is associated with patient inflammatory endotype.
   We report the use of deep metagenomic sequencing in conjunction with
   host multiomic characterization to assess the clinical and biological
   factors associated with the airway resistome in COPD.

   Induced sputum from 99 stable COPD participants was collected in the
   First Affiliated Hospital of Guangzhou Medical University (n = 72) and
   Shenzhen People’s Hospital (n = 27) in China (ethics approval reference
   numbers 2017-22 and KY-LL-2020294-01) and subjected to deep metagenomic
   sequencing (targeted >30 Gb of data per sample) with reagent controls
   ([65]9). Five participants with missing records of long-term antibiotic
   use information were excluded from the analyses. All remaining 94
   participants were free of antibiotics over the preceding 1 month, and
   none had long-term macrolide use. Participants were classified into
   neutrophilic, eosinophilic, and mixed- and pauci-granulocytic endotypes
   based on whether sputum neutrophil and eosinophil percentages were
   above or below established criteria (77.7% for neutrophil and 3% for
   eosinophil; Table S1 in the supplemental material) ([66]10). A total of
   307 ARGs across 18 antibiotic classes were obtained from metagenomic
   reads using ARGs-OAP (v2.0) ([67]11) after quality filtering and host
   read removal using Sunbeam (supplemental methods) ([68]12).
   Unsupervised clustering revealed four subtypes of the resistome, each
   with a distinct dominant antibiotic class (β-lactam-high,
   tetracycline-high, macrolide-lincosamide-streptogramin [MLS]-high and
   multidrug-high; [69]Fig. 1a). A significant association was found
   between the resistome and microbiota composition using Procrustes
   analysis based on Bray-Curtis dissimilarity indices (M^2 = 0.19,
   P < 0.001; Fig. S1). Among all demographic and clinical features, the
   inflammatory endotype had a statistically significant, albeit modest,
   association with the resistome (Adonis, R^2 = 0.087, P = 0.043;
   [70]Fig. 1b and [71]c and Table S2). There was a reduced ARG diversity
   for participants in the neutrophilic subtype, whereas the total ARG
   abundance was higher for participants in the neutrophilic and
   mixed-granulocytic subtypes ([72]Fig. 1d). The majority of participants
   (62.5%) in the neutrophilic subtype possessed a multidrug-high
   resistome, whereas participants in the eosinophilic subtype had a
   higher likelihood of harboring an MLS-high resistome ([73]Fig. 1d).
   Despite a relatively greater proportion of multidrug-high resistomes in
   Guangzhou participants than in Shenzhen participants, the high
   representation of multidrug-high resistomes in neutrophilic COPD
   patients and MLS-high resistomes in eosinophilic COPD patients were
   observed in both sites (Fig. S2). For 33 participants with extra sputum
   available, their sputum was subjected to bacterial culture,
   matrix-assisted laser desorption ionization–time of flight mass
   spectrometry (MALDI-TOF MS) identification, and antimicrobial
   susceptibility testing in the clinical laboratory. In 8 of 13
   neutrophilic participants, the bacteria cultured from their sputum
   exhibited resistance to multiple antibiotic classes, while bacterial
   isolates from 3 of 5 eosinophilic participants showed specific
   resistance to macrolides ([74]Fig. 1d and Table S3). The results were
   partially in agreement with an endotype-related antimicrobial
   resistance pattern, although they should be interpreted with caution
   given the small sample size.

FIG 1.

   [75]FIG 1
   [76]Open in a new tab

   The inflammatory endotype is associated with the airway resistome in
   COPD. (a) Unbiased clustering was performed on the ARG profile of all
   COPD participants using Ward’s method, which revealed four resistome
   subtypes. The ARG profile, inflammatory endotype, site,
   postbronchodilator FEV1, ICS usage, antibiotic usage in the preceding
   6-month period (6m Abx), and exacerbation frequency (Freq Exac) for
   each participant are shown. (b) Permutational multivariate analysis of
   variance using Adonis shows the association between the resistome and
   participant demographic and clinical features. (c) Principal-component
   analysis on the resistomes of all participants, colored by the
   inflammatory endotypes. (d) Box plots show the diversity and total
   abundance of ARGs in participants with different inflammatory
   endotypes. Histograms show the distribution of resistome subtypes and
   the resistance pattern in antimicrobial susceptibility test in
   participants with different inflammatory endotypes. Neu, neutrophils;
   Eos, eosinophils; Mixed, mixed granulocytes; Pauci, pauci granulocytes;
   pred, prediction; Abund, abundance; FEV1, forced expiratory volume in
   the first second; ICS, inhaled corticosteroids; FVC, forced vital
   capacity; LABA, long-acting beta-agonists; BMI, body mass index; GOLD,
   global initiative for chronic obstructive lung disease; mMRC, modified
   medical research council; CAT, COPD assessment test; LAMA, long-acting
   muscarinic antagonist.

   Examining the association of individual ARGs with sputum neutrophil and
   eosinophil percentages further revealed an endotype-specific
   correlation pattern. The abundance of 22 and 20 ARGs was correlated
   with sputum neutrophil and eosinophil percentages, respectively,
   whereas no ARGs were correlated with both (false discovery rate
   [FDR]-adjusted P value of <0.05; [77]Fig. 2a). Neutrophil-associated
   ARGs were mostly of multidrug resistance, including mdtBCDHL, emrB, and
   mexI (Table S4). Eosinophil-associated ARGs were related to a diverse
   range of antibiotic classes, among which the macB and lnuA genes of MLS
   resistance were most significant (Table S4). Ralstonia spp. and
   Pseudomonas spp. were the top contributors to the neutrophil-associated
   ARGs, whereas species in Campylobacter and Aggregatibacter, previously
   found to be associated with sputum eosinophilia ([78]13), contributed
   most to the eosinophil-associated ARGs ([79]Fig. 2b).

FIG 2.

   [80]FIG 2
   [81]Open in a new tab

   The association between antibiotic resistance genes (ARGs) and
   inflammatory endotype, host transcriptome, and inflammatory mediators
   in COPD. (a) The association between ARGs and sputum neutrophil and
   eosinophil percentages. Each dot represents an ARG colored by
   antibiotic class. The x and y axes represent the directionality of
   correlation multiplied by the −log[10] of the FDR-adjusted P value. The
   ordination space is divided into colored area for association with
   neutrophils only, eosinophils only, both, and neither (FDR P < 0.05).
   (b) Bubble chart illustrating the relative contribution of bacterial
   species to the ARGs significantly correlated with sputum neutrophils or
   eosinophils. The average contribution of each species to the
   neutrophil- or eosinophil-associated ARGs is indicated beside the
   chart. (c) Heat maps illustrating the correlation of neutrophil- and
   eosinophil-associated ARGs with sputum host transcriptome modules and
   inflammatory mediators. The −log[10] P values for the associations of
   modules and mediators with the resistome profiles in the canonical
   correspondence analysis for participants of neutrophilic or
   eosinophilic endotypes are shown beside the heat maps. The top module
   and mediator in the association are highlighted in red. Aver. contrib,
   average contribution; assoc, association.

   These findings led us to further test the association between ARGs and
   host gene expression and inflammatory mediators. We obtained host
   transcriptomes from induced sputum for all participants by mRNA
   sequencing using an Illumina NovaSeq. A panel of sputum inflammatory
   mediators was characterized for 43 participants, with extra sputum
   samples available using antibody microarray, as described previously
   (RayBiotech, Norcross, USA) ([82]14). Genes in the transcriptome were
   subject to coabundance clustering using a weighted correlation network
   analysis and pathway enrichment analysis ([83]15) to obtain
   coexpression modules with functional annotation. For modules
   significantly enriched for marker genes for macrophages, neutrophils,
   and eosinophils, ([84]16), their expression levels were highly
   correlated with sputum differential cell count of the corresponding
   cell type (Fig. S3), suggesting the ability of the host transcriptome
   in sputum to reflect patient immune and inflammatory status. Thirteen
   modules and 10 sputum mediators were significantly correlated with the
   neutrophil-associated ARGs, among which M379 enriched with arachidonic
   acid metabolism and matrix metallopeptidase 8 (MMP-8) exhibited the
   strongest associations with overall resistome profiles in
   neutrophilic-predominant COPD participants (canonical correspondence
   analysis, P < 0.05; [85]Fig. 2c and Table S5). Ten modules and 6
   mediators were significantly correlated with eosinophil-associated
   ARGs, with M226 enriched with eosinophil chemotaxis, and IL-13 showing
   greatest associations with the resistome in eosinophilic-predominant
   participants ([86]Fig. 2c and Table S5).

   Taken together, these results uncover a previously unrecognized link
   between inflammatory endotype and the airway resistome in COPD. For
   neutrophilic-predominant COPD, the resistome was dominated by multidrug
   resistance genes from respiratory pathogens. For
   eosinophilic-predominant COPD, the resistome was diverse, with an
   increased portion of patients showing an MLS-high resistome. Our
   results suggest that the varied resistome across inflammatory endotypes
   could be a result of the underlying differences in the resident
   microbiota, whereas prior antibiotic history may be of less relevance,
   based on current data and literature ([87]7, [88]8). These findings
   have potential clinical implications. A multidrug-high resistome driven
   by pathogenic taxa in neutrophilic COPD patients at the stable state
   may be a risk factor for antibiotic treatment failure during an
   exacerbation. Eosinophilic COPD is potentially associated with less
   frequent bacterial infection ([89]17), but it nevertheless harbors a
   diverse range of ARGs, with a greater abundance of those being
   macrolide resistant. Macrolides are increasingly used as prophylactic
   antibiotics for the management of frequent exacerbation in COPD
   ([90]18). The ARGs of macrolide resistance can be of concern, as
   macrolides may be used for eosinophil-high COPD patients who exacerbate
   despite inhaled corticosteroid-containing therapy. Multiomic analysis
   indicates that differential host pathways and sputum mediators (MMP-8
   and IL-13) are associated with the resistome in patients with
   neutrophilic or eosinophilic inflammation, suggesting endotype-specific
   host-resistome interactions. These results raise intriguing
   possibilities that merit further investigation, such as whether these
   mediators could be markers that inform antibiotic selection and whether
   modification of host pathways may alter airway microbiota and
   antibiotic resistance.

   The limitations of this study include its relatively small number of
   participants and cross-sectional nature. Despite having a validation
   cohort, the generalizability of the findings to patients with different
   demographic backgrounds and antibiotic history needs to be further
   assessed. In addition, it is important to assess the baseline resistome
   in healthy individuals in comparison with COPD patients. A previous
   study identified both shared and unique ARGs in healthy individuals and
   patients across a broad range of respiratory diseases, suggesting a
   complex airway resistome with both a core and discriminatory set of
   elements ([91]7). Our results highlight the potential need for
   considering patient inflammatory endotype in decision-making about
   antibiotic use in COPD and the broader chronic respiratory diseases.
   Longitudinal surveys on the dynamics of the COPD airway resistome and
   its association with antibiotic treatment outcome are warranted to test
   the clinical applicability of resistome assessment in antibiotic
   management in COPD.

Data availability.

   The raw multiomic sequencing data have been deposited in the Chinese
   National GenBank Nucleotide Sequence Archive ([92]CNP0001954). The
   computer code has been deposited at
   [93]https://github.com/wangzlab/COPD_Resistome.

Supplementary Material

   Reviewer comments
   [94]reviewer-comments.pdf^ (536.1KB, pdf)

ACKNOWLEDGMENTS