Spatial distribution of microbial communities in the
cystic fibrosis lung
Dana Willner1,2,3, Matthew R Haynes1, Mike Furlan1, Robert Schmieder4, Yan Wei Lim1,
Paul B Rainey5,6, Forest Rohwer1,7and Douglas Conrad8,9
1Department of Biology, San Diego State University, San Diego, CA, USA;2Australian Centre for
Ecogenomics, University of Queensland, Brisbane, Queensland, Australia;3Diamantina Institute, University
of Queensland, Brisbane, Queensland, Australia;4Computational Science Research Center, San Diego State
University, San Diego, CA, USA;5New Zealand Institute for Advanced Study and Allan Wilson Centre for
Molecular Ecology & Evolution, Massey University, Auckland, New Zealand;6Max Planck Institute
for Evolutionary Biology, Plo ¨n, Germany;7Center for Microbial Sciences, San Diego, CA, USA;8Department
of Medicine, University of California San Diego, La Jolla, CA, USA and9San Diego VA Healthcare System,
San Diego, CA, USA
Cystic fibrosis (CF) is a common fatal genetic disorder with mortality most often resulting from
microbial infections of the lungs. Culture-independent studies of CF-associated microbial
communities have indicated that microbial diversity in the CF airways is much higher than
suggested by culturing alone. However, these studies have relied on indirect methods to sample the
CF lung such as expectorated sputum and bronchoalveolar lavage (BAL). Here, we characterize the
diversity of microbial communities in tissue sections from anatomically distinct regions of the CF
lung using barcoded 16S amplicon pyrosequencing. Microbial communities differed significantly
between different areas of the lungs, and few taxa were common to microbial communities in all
anatomical regions surveyed. Our results indicate that CF lung infections are not only polymicrobial,
but also spatially heterogeneous suggesting that treatment regimes tailored to dominant
populations in sputum or BAL samples may be ineffective against infections in some areas of
The ISME Journal (2012) 6, 471–474; doi:10.1038/ismej.2011.104; published online 28 July 2011
Subject Category: microbial population and community ecology
Keywords: cystic fibrosis; amplicon sequencing; microbial communities
Cystic fibrosis (CF) is one of the most common
heritable genetic diseases in Caucasian populations,
affecting over 70000 individuals worldwide (Cystic
Fibrosis Foundation, 2011). Most mortality in CF
is due to microbial infections of the lungs, which
lead to eventual respiratory failure (Cystic Fibrosis
Staphylococcus aureus and Burkholderia cepacia
are common pathogens cultured from adult CF
patients, but recent studies have demonstrated that
microbial communities in the CF airways are highly
diverse (Armougom et al., 2009; Rogers et al., 2009;
Cox et al., 2010; Guss et al., 2011). Culture-based
analysis of bronchoalveolar lavage (BAL) and au-
topsy samples indicates that microbial populations
vary significantly in different spatial regions of the
CF lung (Smith et al., 1998; Gutierrez et al., 2001;
Gilchrist et al., 2011). Erb-Downward et al. (2011)
recently demonstrated spatial heterogeneity in mi-
crobial communities from the ex-planted lung tissue
of COPD patients using culture-independent meth-
ods, but no similar study has been conducted in CF
patients. Here, we present the first culture-indepen-
dent characterization of microbial communities
isolated directly from CF lung tissue.
To characterize microbial communities, we ob-
tained lung tissue sections from the ex-planted
lungs of a transplant patient, as well as tissue
sections and a tracheal mucus plug from a deceased
CF patient following autopsy. Patient clinical data
are presented in the Supplementary Materials and
methods. Microbial communities were profiled by
high-throughput barcoded amplicon sequencing of
the 16S rDNA gene and were analyzed using the
methods) (Caporaso et al., 2010).
Microbial community richness was greater in
the ex-plant lungs than in the postmortem lungs.
At 97% similarity, there were 498 operational
Received 28 February 2011; revised 14 June 2011; accepted 17
June 2011; published online 28 July 2011
Correspondence: D Willner, School of Chemistry and Molecular
Biosciences, University of Queensland, Australian Centre for
Ecogenomics, St. Luica, Queensland 4072 Australia.
The ISME Journal (2012) 6, 471–474
& 2012 International Society for Microbial Ecology All rights reserved 1751-7362/12
taxonomic units (OTUs) corresponding to 20 phyla
in the ex-plant lungs, as compared with 106 OTUs
comprising nine phyla in the postmortem lungs
(Figure 1; Supplementary Tables 1 and 2). Rarefac-
tion curves and Chao1 estimates indicated that
expected OTU richness by lobe was an order of
magnitude higher in the ex-plant lungs (Supple-
mentary Figure 1; Supplementary Table 1). Both sets
of lungs exhibited pathology characteristic of late-
stage CF respiratory disease, but the ex-plant lungs
maintained some areas of normal tissue, while the
postmortem lungs showed near-complete destruc-
tion of the peripheral airways in all areas (Supple-
mentary Materials and methods). This may be
correlated with reduced richness as previous studies
of microbial diversity in CF and COPD have demon-
strated a negative relationship between microbial
diversity and disease severity (Erb-Downward et al.,
2011; van der Gast et al., 2011).
Microbial communities in both sets of lungs were
dominated by a small number of highly abundant
taxa. One OTU classified as Comomonadaceae at
abundance as a percentage of the total population is indicated by color. Phyla are indicated on the left, and a subset of OTU
classifications is indicated on the right. A complete list of OTU classification appears in Supplementary Table 2. Lung regions are
abbreviated as RUL for right upper lobe, RML for right middle lobe, RLL for right lower lobe, LUL for left upper lobe, Ling for left lingula
and LLL for left lower lobe. LLLA indicates the left lower lobe anterior and LLLP indicates the left lower lobe posterior. Trach indicates
the tracheal sample.
Relative abundance of microbial OTUs in the ex-plant and postmortem lungs. Each row represents a different OTU, and the
Microbes in the CF lung
D Willner et al
The ISME Journal
the family level comprised B68% of total bacterial
diversity in each lobe in the ex-plant lungs (Figure 1;
Supplementary Table 2). Comomonadaceae have
previously been identified in CF respiratory sam-
ples, but are not commonly recognized as CF patho-
gens (Coenye et al., 2002; Guss et al., 2011). OTUs
corresponding to Flavobacteriaceae sp. (14–23%),
Diaphorobacter sp. (3–5%) and Novosphingobium
sp. (2–4%) were also ubiquitous. In the postmortem
lungs, Pseudomonas aeruginosa (BLASTn, 100%
identity, e-valueo1?10–180) was the most abundant
taxon, comprising between 51–94% of sequences in
each lobe, and 99% of the tracheal sample. Other
OTUs were ubiquitous but found in lower (o3%)
abundances, including phylotypes classified as
Lactobacillus delbruckii, Enterococcus sp., Cloaci-
bacterium sp. and Diaphorobacter sp.
Microbial communities in the ex-plant lungs were
more similar to each other than those in the
postmortem lungs when communities were com-
pared using the weighted Unifrac distance metric
(Figure 2; Supplementary Table 3). Distances were
greater on average between lobes of the postmortem
lungs than between lobes of the ex-plant lungs,
reflecting more divergent communities (Supplemen-
tary Table 3). Less than 1% of OTUs were shared by
all lobes in the postmortem lungs, while B60% of
OTUs were identified in only one lobe (Supplemen-
tary Table 2). Several unique OTUs corresponded to
potential CF pathogens, such as Xanthomonas (left
upper lobe only) and Chyrseobacterium (Harrison,
2007; LiPuma, 2010). In the ex-plant lungs, B3% of
OTUs were shared by all lobes, but most OTUs
which were not universally shared occurred in low
abundance (o1%). These may represent transient
CF microbial populations as opposed to core CF
microbiota as described in van der Gast et al. (2011).
Average linkage (UMPGA) clustering based on
Unifrac distances produced two main clusters
corresponding to the ex-plant and postmortem
lungs and several smaller clusters within each set
of lungs (Figure 2). In the ex-plant lungs, microbial
communities formed two significantly different
clusters (Bonferroni’s-corrected P-valueo0.05), as
determined by XIPE, which compares community
distributions using non-parametric statistics (Rodri-
guez-Brito et al., 2006). One cluster contained the
RUL and Lingula and the other contained all other
lobes. This clustering was largely due to abundance
differences in dominant OTUs, including Flavo-
bacteriaceae spp. In the postmortem lungs, commu-
nities formed two clusters with the LLL as an
outgroup (Figure 2). The divergence of the LLL
was driven by a high abundance of chloroplast
sequences, which may represent plant DNA intro-
duced by aspiration (Bousbia et al., 2010). Aspira-
tion may have occurred while the patient was
intubated 3 weeks before death, as Bousbia et al.
(2010) demonstrated the persistence of plant DNA in
the lungs for several weeks in patients undergoing
mechanical ventilation. Clustering of tracheal, RML
and RLL communities may indicate gravity-depen-
dent deposition of aspirated tracheal mucus, as
tracheal contents would be expected to settle in the
RLL and/or RML (Bittar, 2002). Despite the observed
clustering, the tracheal community was significantly
distance metric, and clustered using average linkage. Node labels indicate jackknife support values, and the bar below the tree represents
a weighted Unifrac distance of 0.01. Shaded boxes indicate communities that were not significantly different from each other, but were
significantly different from all other communities as determined by XIPE.
Spatial clustering of microbial communities in the CF lungs. Communities were compared using the weighted Unifrac
Microbes in the CF lung
D Willner et al
The ISME Journal
different from the RML and RLL, containing seven Download full-text
OTUs not found in the lungs, including Prevotella
spp., a common constituent of oral microbiota.
Our results indicate that microbial communities
in the cystic fibrosis lung are spatially hetero-
geneous. The use of autopsy samples is a caveat
to this study, as microbial community composition
may be affected by aspiration at death. Despite
this limitation, direct sampling of lung tissue
provides a more specific characterization of micro-
biota than other sampling methods, which may
misrepresent the true nature of CF lung disease due
to a lack of spatial resolution. Spatial heterogene-
ities may extend further than can be resolved by
16S sequencing, including regional differences in
microbial biomass, antibiotic resistance and strain
level differences in highly abundant and ubiquitous
populations. Treatment strategies for cystic fibrosis,
which rely on antibiotics targeted at microbes of
specific species with known antibiotic resistance
may only be effective against pathogenic microbes
in particular regions or microenvironments of the
Armougom F, Bittar F, Stremler N, Rolain JM, Robert C,
Dubus JC et al. (2009). Microbial diversity in the
sputum of a cystic fibrosis patient studied with 16S
rDNA pyrosequencing. Eur J Clin Microbiol Infect Dis
Bittar EE. (2002). Pulmonary Biol health Dis. Springer.
Bousbia S, Papazian L, La Scola B, Raoult D. (2010).
Detection of plant DNA in the bronchoalveolar lavage
of patients with ventilator-associated pneumonia.
PLoS One 5: e11298.
Caporaso JG, Kuczynski J, Stombaugh J, Bittinger K,
Bushman FD, Costello EK et al. (2010). QIIME allows
analysis of high-throughput community sequencing
data. Nat Meth 7: 335–336.
Coenye T, Goris J, Spilker T, Vandamme P, LiPuma JJ.
(2002). Characterization of unusual bacteria isolated
from respiratory secretions of cystic fibrosis patients
and description of inquilinus limosus gen. nov., sp.
nov. J Clin Microbiol 40: 2062–2069.
Cox MJ, Cox MJ, Allgaier M, Taylor B, Baek MS, Huang YJ
et al. (2010). Airway microbiota and pathogen abun-
dance in age-stratified cystic fibrosis patients. PLoS
One 5: e11044.
Cystic Fibrosis Foundation Patient Registry 2008 Annual
Cystic Fibrosis Foundation (2011) Available at: http://
www.cff.org (accessed 20 February 2011).
Erb-Downward JR, Thompson DL, Han MK, Freeman CM,
McCloskey L, Schmidt LA et al. (2011). Analysis of the
lung microbiome in the ‘Healthy’ smoker and in
COPD. PLoS One 6: e16384.
Gilchrist FJ, Salamat S, Clayton S, Peach J, Alexander J,
Lenney W et al. (2011). Bronchoalveolar lavage in
children with cystic fibrosis: how many lobes should
be sampled? Arch Dis Child 96: 215–217. Available
(accessed 7 December 2010).
Guss AM, Roeselers G, Newton IL, Young CR, Klepac-Ceraj
V, Lory S et al. (2011). Phylogenetic and metabolic
diversity of bacteria associated with cystic fibrosis.
ISME J 5: 20–29. Available at: http://www.ncbi.nlm.
nih.gov/pubmed/20631810 (accessed 19 September
Gutierrez JP, Grimwood K, Armstrong DS, Carlin JB,
Carzino R, Olinsky A et al. (2001). Interlobar differ-
ences in bronchoalveolar lavage fluid from children
with cystic fibrosis. Eur Respir J 17: 281–286.
Harrison F. (2007). Microbial ecology of the cystic fibrosis
lung. Microbiology (Reading, Engl) 153: 917–923.
LiPuma JJ. (2010). The changing microbial epidemio-
logy in cystic fibrosis. Clin Microbiol Rev 23: 299–323.
Rodriguez-Brito B, Rohwer F, Edwards RA. (2006). An
application of statistics to comparative metagenomics.
BMC Bioinformatics 7: 162.
Rogers GB, Carroll MP, Bruce KD. (2009). Studying
approaches. J Med Microbiol 58: 1401–1418.
Smith D, Smith E, Pitt T, Stableforth D. (1998). Regional
microbiology of the cystic fibrosis lung: a post-mortem
study in adults. J Infect 37: 41–43.
van der Gast CJ, Walker AW, Stressmann FA, Rogers GB,
Scott P, Daniels TW et al. (2011). Partitioning core and
satellite taxa from within cystic fibrosis lung bacterial
communities. ISME J 5: 780–791.
Supplementary Information accompanies the paper on The ISME Journal website (http://www.nature.com/ismej)
Microbes in the CF lung
D Willner et al
The ISME Journal