Pone 0265891

PLOS ONE
RESEARCH ARTICLE
An optimized approach for processing

of frozen lung and lavage samples for
microbiome studies
Rosana Wiscovitch-Russo ID1*, Harinder Singh1, Lauren M. Oldfield1¤, Alexey
V. Fedulov2‡, Norberto Gonzalez-Juarbe1‡
1 Department of Infectious Diseases and Genomic Medicine, J. Craig Venter Institute, Rockville, Maryland,
United States of America, 2 Department of Surgery, Division of Surgical Research, Rhode Island Hospital,
Alpert Medical School of Brown University, Providence, Rhode Island, United States of America
a1111111111
a1111111111 ¤ Current address: Replay Therapeutics, San Diego, California, United States of America
a1111111111 ‡ AVF and NGJ also contributed equally to this work.
a1111111111 * [email protected]
a1111111111
Abstract
The respiratory tract has a resident microbiome with low biomass and limited diversity. This
OPEN ACCESS
results in difficulties with sample preparation for sequencing due to uneven bacteria-to-host
Citation: Wiscovitch-Russo R, Singh H, Oldfield DNA ratio, especially for small tissue samples such as mouse lungs. We compared effec-
LM, Fedulov AV, Gonzalez-Juarbe N (2022) An
tiveness of current procedures used for DNA extraction in microbiome studies. Bronchoal-
optimized approach for processing of frozen lung
and lavage samples for microbiome studies. PLoS veolar lavage fluid (BALF) and lung tissue samples were collected to test different forms of
ONE 17(4): e0265891. https://doi.org/10.1371/ sample pre-treatment and extraction methods to increase bacterial DNA yield and optimize
journal.pone.0265891
library preparation. DNA extraction using a pre-treatment method of mechanical lysis (lung
Editor: Franck Carbonero, Washington State tissue) and one-step centrifugation (BALF) increased DNA yield and bacterial content of
University - Spokane, UNITED STATES
samples. In contrast, a significant increase of environmental contamination was detected
Received: November 9, 2021 after phenol chloroform isoamyl alcohol (PCI) extraction and nested PCR. While PCI has
Accepted: March 9, 2022 been a standard procedure used in microbiome studies, our data suggests that it is not effi-
cient for DNA extraction of frozen low biomass samples. Finally, a DNA Enrichment kit was
Published: April 5, 2022
tested and found to improve the 16S copy number of lung tissue with a minor shift in micro-
Copyright: © 2022 Wiscovitch-Russo et al. This is
bial composition. Overall, we present a standardized method to provide high yielding DNA
an open access article distributed under the terms
of the Creative Commons Attribution License, and improve sequencing coverage of low microbial biomass frozen samples with minimal
which permits unrestricted use, distribution, and contamination.
reproduction in any medium, provided the original
author and source are credited.
Data Availability Statement: To comply with

PLOS ONE’s data availability statement, we made
the samples metadata file available as part of the
supporting information (S1 Data). In addition, the 1. Introduction
raw sequence files have been deposited in the
NCBI database under the BioProject
The lower respiratory tract (LRT) was once thought to be a sterile environment [1–3]. Initial
PRJNA804973. studies used culture-based methods, microscopy, and biochemical assays to isolate and identify
microorganisms from the lung environment. However, the microbes isolated were mostly
Funding: N.G-J and A.F. were supported by
National Institutes of Health award R01ES030227.
associated with pathogenic infections (e.g., S. pneumoniae, H. influenzae, P. aeruginosa, and S.
https://www.nih.gov. The funders had and will not aureus among others) or contamination during sample collection [1–3]. Recent developments
have a role in study design, data collection and in culture-independent methods rely on 16S ribosomal RNA (16S rRNA) and next-generation
PLOS ONE | https://doi.org/10.1371/journal.pone.0265891 April 5, 2022 1 / 17

PLOS ONE Optimized processing of frozen lung and lavage for microbiome studies
analysis, decision to publish, or preparation of the sequencing (NGS) aimed towards identification of viable but non-culturable bacteria [4]. The
manuscript. Human Microbiome Project (HMP) mainly used 16S rRNA based sequencing to study the
Competing interests: The authors have declared oral, nasal, vaginal, gut, and skin microbial flora and improved the understanding of host-
that no competing interests exist. microbe interactions [5, 6]. While a better understanding of human microbial communities of
multiple host sites is currently available, the microbial composition of lung flora and how the
environment alters such is still to be elucidated.
Pioneering studies of the human lung microbiome used diverse methods for sample collec-
tion and processing (S1 Table). Due to tissue accessibility and ethical concerns, most human
studies assessed the microbiome of the upper respiratory tract (URT) by collecting swab sam-
ples from the nasopharynx or pharynx [7, 8]. Although they are topologically distinct environ-
ments, recent studies have shown that URT is a close representation of the LRT housing
similar microbiota but in distinct logarithmic quantities [9, 10]. It has been shown that the
human lung has a resident microbiome of bacteria, fungi and viruses derived from nasopha-
ryngeal commensals [11]. Importantly, compared to other niches (e.g., gut), the lung micro-
biome is limited in diversity and has a low microbial biomass (estimated 2.2 x 103 bacterial
genomes per cm2 of lung tissue), leading to difficulties in its study [11, 12]. Most human stud-
ies also present additional variables such as pre-existing health conditions (e.g. obesity, smok-
ing, antibiotics, asthma or cancer), the difference in age groups and geographical locations
which can influence the baseline microbiome of an individual [13, 14]. Of note, the host envi-
ronment (rural vs. urban) and lifestyle (diet, smoking, etc.) highly influences the lung micro-
biome [8]. These intriguing effects of lung microbiome interactions with environment
produce a lot of interest and are being actively studied. However, lack of consistent sample iso-
lation and preparation procedures hampers our ability to compare studies [15, 16]. Animal
models serve as strong tools to dissect microbiome changes specific to environmental expo-
sures and the interaction between host and microbes in a controlled manner.
In addition to limited microbial diversity, low microbial biomass may reduce the detection
of bacteria in mouse lung samples using 16S amplicon-based sequencing [17]. Most published
studies using mouse models have shown some variability in lung microbiome profiles due to
the use of different methods for sample processing, gene amplification and bioinformatic anal-
ysis [18–21]. Therefore, variability of normal lung microbiome of barrier laboratory mice
includes not only the effects of strain, housing conditions and vendor [22, 23], but full contri-
bution of sample pre-treatment is poorly studied [24]. It has been suggested that Phenol Chlo-
roform Isoamyl alcohol (PCI) method is suitable for the low-biomass lung tissue investigation
[25], while others have pointed out it leads to artefactual results in low-biomass specimens and
is the most labor-intensive [26]. These two reports published last year emphasize the lack of
cohesion in the matter. Here we tested different forms of processing for low-microbial-bio-
mass frozen lung samples and bronchoalveolar lavage fluid (BALF) aiming to increase DNA
yield and improve library preparation towards an optimized standardized method for
sequencing of lung microbiomes.
2. Methods
2.1 Mice and sample collection
Six to nine -week-old female BALB/c mice were purchased from Jackson Laboratory (Bar Har-
bor, ME) or Charles River Laboratory (Wilmington, MA). Experiment was conducted in
accordance with the Institutional Animal Care and Use Committee approved by the Brown
University (Rhode Island Hospital IACUC #504718). Mice were housed in barrier conditions.
Harvesting of lung tissue and/or bronchoalveolar lavage was performed under sterile condi-
tions as described in [27, 28]. Briefly, mice were euthanized by intraperitoneal injection of

Fatal-Plus, chest wall was rinsed with alcohol, the trachea was accessed via a cutaneous cut,
soft tissues were separated blunt and a sterile angiocath catheter made of fluorinated ethylene
propylene was inserted into trachea. Lavage was performed with sterile phosphate buffered
saline (PBS) without Ca or Mg (Lonza), five repeat instillations of 0.5 mL. After lavage the
chest wall was opened and the lungs were excised, washed in PBS and frozen.
2.2 Decontamination of workstation for sample processing and DNA

extraction
Isolation of low microbial biomass from murine samples was done in a Class II Biosafety cabi-
net treated with Eliminase (DeconLabs, Cat# 1101) and 70% ethanol then exposed to UV light
for 20–30 minutes to ensure sterile conditions. Additional suggestions for controlling contam-
ination during sample processing:
2.2.1 Frequently change gloves. During the procedure it is recommended to change
gloves frequently. Change gloves after disinfecting the workstation and when solution spills on
to gloves.
2.2.2 Include controls during sample extraction. Include a water extraction (negative
control) for each batch of samples processed to trace contamination back to the source when
processing through the bioinformatic pipeline (Method 2.7). Extraction negative and cell
mock community extraction (positive control) should be included in amplification and
sequencing of all samples. ZymoBIOMIC Standards (Cat# D6300 and Cat# D6310) were used
as extraction positive controls.
2.2.3 Keep tube caps closed when not in use. Maintain all sample tubes closed at all
times. Only open one tube at a time to discard or dispense the solution. Maintain distance
between tubes while processing on the rack.
2.3 Phenol chloroform isoamyl alcohol DNA extraction

To pellet cells present in BALF samples, 1000 μL of the total sample volume was centrifuged
(8,000xg for 15 minutes at 4˚C), then resuspended in 500 μL of lysis buffer (20 mM Tris-Cl, pH
8.0, 2 mM EDTA, 1.2% Triton X-100). Separately, lung samples (left lobe) were submerged in
500 μL of lysis buffer. All samples, both tissue and BALF, were then incubated at 75˚C for 10 min-
utes. After cooling at room temperature, samples were transferred into a Matrix B Lysing tube
and further lysed in a BioSpec Mini-Beadbeater for 45 seconds. The obtained lysate was then
treated with 60 μL of lysozyme (Sigma Aldrich, Cat# L6876-10G) and 5 μL of Linker RNase A
(ThermoFisher, Cat# 12091039) and incubated at 37˚C for 60 minutes. Subsequently, 100 μL 10%
Sodium Dodecyl Sulfate (SDS) and 42 μL Proteinase K (QIAGEN, Cat# 19131) were added and
incubated overnight at 55˚C. After completing lysis protocol, we proceeded to isolate the DNA.
Samples were treated twice with equal volumes of phenol chloroform isoamyl alcohol and cleaned
through ethanol precipitation [29]. DNA was eluted in 50 μL of Tris-EDTA (TE) buffer and DNA
concentration was measured using a Qubit 1X dsDNA HS Assay (ThermoFisher, Cat# Q33231).
2.4 Increasing extraction efficiency through sample pre-treatments

To extract DNA from 500–1500 μL of BALF and lung tissue (left lobe) samples were processed
using the DNeasy PowerSoil Pro Kit (QIAGEN, Cat# 47016). To examine extraction efficiency,
the step 2 of the of PowerSoil protocol was modified (Fig 1). Briefly:
2.4.1 Mechanical cell lysis. First, 800 μL of CD1 lysis buffer provided by the kit was added
to the collected BALF and lung tissue in PowerBead Pro tubes. Then, as described in the
DNeasy PowerSoil Pro Kit protocol, disruption of the cell membrane was done by physical
lysis via a Vortex-Genie 2 with a 24 sample adaptor at maximum speed for 15 minutes. In

Fig 1. Workflow for processing BALF and lung samples for microbiome data. Initially we compared two DNA
extraction methods (experimental kit-based and PCI) and three forms of sample pre-treatment method. A small subset
of samples was used to compare the efficiency of Microbial Enrichment kit. The V4 16S region was selected for
preparing two nested (25x25 and 35x25 PCR cycles) and one non-nested (35 PCR cycles) libraries preparation methods.
https://doi.org/10.1371/journal.pone.0265891.g001
addition to that mentioned in the protocol, we treated the samples either by using a BioSpec
Mini-Beadbeater at maximum speed for 45 seconds or using the QIAGEN PowerLyzer at 3000
rpm for 30 seconds.
2.4.2 Enzymatic cell lysis. Further disruption of the cell membrane was done through
enzymatic cell lysis using Proteinase K. Here, lung tissue samples were lysed in the provided

PowerBead Pro tubes with 200 μL of TE buffer and 100 μL 10% SDS using the BioSpec Mini-
Beadbeater at max speed for 45 seconds. Then, 40 μL of proteinase K was added and incubated
overnight at 55˚C.
2.4.3 Cell concentration. An additional set of BALF samples were processed to concen-
trate biological material and/or to reduce host cell presence. To do the latter we used either
one or two-step centrifugation processes:
2.4.3.1 One-spin. One-step centrifugation: samples were centrifuged once at 5,000xg for 5
minutes at 4˚C to pellet all the cells (mammalian and bacterial).
2.4.3.2 Two-spin. Two-step centrifugation, (i) samples were centrifuged first at 300xg for 10
minutes at 4˚C to pellet host cells, (ii) then supernatant was transferred into a new 1.5mL
Eppendorf and centrifuged at 5,000xg for 5 minutes at 4˚C to collect remaining bacterial cells.
The supernatant was collected and stored at -80˚C for future use. The final pellets were
resuspended in 800 μL of CD1 buffer and transferred to the provided PowerBead Pro tubes.
The samples were then mechanically lysed using BioSpec Mini-Beadbeater at max speed for 45
seconds as described in 2.4.1.
For all lysed samples (see above), steps 3–17 of the PowerSoil protocol were carried out
without further modifications except for the final step when DNA was eluted in 50 μL of C6
buffer (10 mM Tris) instead of 100 μL. DNA concentration was measured through Qubit 1X
dsDNA HS Assay and Nanodrop ND-1000 Spectrophotometer.
2.5 Microbiome DNA enrichment kit

Purified gDNA extracted using the PowerSoil Kit (Methods 2.4.1 and 2.4.3.1) was split into
two aliquots for the purpose of comparing the efficiency before and after treatment with the
NEBNext Microbiome DNA Enrichment Kit (Cat# E2612S). This kit aims to increase bacterial
DNA by selectively targeting and reducing host methylated DNA [30]. NEBNext Microbiome
DNA Enrichment Kit was carried out according to manufacturer protocol. Briefly, 25 μL of
BALF and 3–5 μL of lung gDNA was used during the procedure. Due to the high DNA con-
centration of lung tissue, aliquoted lung gDNA sample was diluted 1:10 in qPCR grade water,
then 1 μL was used to match the DNA to bead ratio optimized in the manufacturer’s protocol.
Target DNA was purified through ethanol precipitation and eluted in 25 μL of 1X TE buffer as
the final step of the manufacturer’s protocol.
2.6 V4 16S amplification and sequencing

Purified gDNA from PCI method, PowerSoil Pro and Microbiome Enrichment Kits was
used to amplify the hypervariable V4 region of the 16S rRNA gene comparing nested and
non-nested PCR library preparation methods. Taxonomic profiling of the mice lung
microbiome was based on the V4 16S rRNA hypervariable region because of its taxonomic
resolution and accurate assignment [31–33]. The primers used were 515-533F forward
(GTGCCAGCMGCCGCGGTAA) and 806-787R reverse (GGACTACHVGGGTWTCTAAT) [53].
The nested libraries were first prepared by amplifying the target region using non-bar-
coded primers at 25 or 35 cycles, then the second round of amplification used the barcoded
primers at 25 cycles. The nested and non-nested libraries used the Illumina MiSeq adapter-
and barcode-ligated dual-index primers [31]. PCR was performed using ThermoFisher
Platinum Taq DNA Polymerase (Life Technologies, Cat# 10966–026); 94˚C for 5 minutes,
94˚C for 30 seconds, 55˚C for 30 seconds, 72˚C for 30 seconds for 25–35 cycles, 72˚C for 7
minutes. Libraries were purified using SPRIselect beads (Beckman, Cat# B23318) for size
selection of 300 bp fragments then quantified using ThermoFisher SYBRGold (Cat#
S11494) in a TECAN microplate reader. Afterwards libraries were normalized, pooled (3.5

pM), and sequenced on Illumina MiSeq system using a V2 500 (2x250 bp) cycles chemistry
kit supplemented with 20% PhiX.
2.7 Bioinformatic pipeline

We filtered poor quality reads (>Q30), trimmed primers and removed reads of <220 bp
prior to analysis. Assigned taxonomy using in-house pipeline supported by Mothur [34]
and Uparse [35]. Operational taxonomic units (OTUs) were based on SILVA 16S rRNA
database version 123 at 97% sequence similarity [36]. We also used R statistical environ-
ment to generate plots using PhyloSeq package [37] and ggplot 2 packages to assess the
microbial composition [38].
3. Results
BALF and lung samples underwent differential extractions, pre-treatments, and library prepa-
ration prior to 16S sequencing (Fig 1, workflow). A total of 50 samples were processed using
two forms of DNA isolation methods: an experimental DNA isolation method (Method 2.4)
that has a relatively short workflow, or the traditional and more laborious phenol chloroform
isoamyl (PCI) DNA extraction method (Method 2.3). To increase DNA yield from these sam-
ples, we examined different experimental pre-treatment methods by modifying steps from an
extraction kit protocol to test the lysis efficacy of various bead beating equipment (Method ver-
sion 2.4.1). We also tested additional disruption of the cell membrane by enzymatic lysis
(Method version 2.4.2) and concentrating the cells prior to mechanical lysis (Method version
2.4.3). Moreover, in separate experiments we tested the NEBNext Microbiome DNA Enrich-
ment Kit aiming to improve bacterial-to-host DNA ratio and enhance primer specificity dur-
ing library preparation (Method 2.5). For samples with low microbial biomass, we amplified
the V4 16S region under nested and non-nested PCR conditions (Method 2.6). Finally, all sam-
ples underwent 16S sequencing and the microbiome was profiled using an in-house bioinfor-
matics pipeline and R environment (Method 2.7).
3.1 Nested-PCR promotes DNA contamination from environmental

organisms
Negative controls (see Method 2.2.2) were included in batch extractions. This is done as
part of our bioinformatic pipeline, where negative controls are examined first to identify
and remove contamination. The latter assures the observation of the true microbial commu-
nity in a sample and aims to exclude microbial contamination. We tested 3 PCR conditions
for all the samples: PCR-1 (one round of 35 cycles), PCR-2 (round 1 of 35 cycles, round 2 of
25 cycles), PCR-3 (round 1 of 25 cycles, round 2 of 25 cycles). Negative control samples
showed a significant (p < 0.01) increase of reads upon nested (PCR-2) when compared to
the non-nested PCR (PCR-1) (S1A Fig). To further analyze the biomass of 16S RNA copy
number, we plotted the read count for each genus after nested or non-nested PCR condi-
tions. The data showed that while an increase in reads was obtained, there was a higher
amount of DNA contamination potentially from the kit’s reagents (e.g. Pseudomonas and
Shewanella sp.) or trace amount of cross contamination from loci other than lungs (e.g. Lac-
tobacillus sp. and Enterobacteriaceae) (S1B Fig) [39, 40]. We suggest this should be taken
into consideration when working with samples that are naturally low-biomass and are easily
susceptible to contamination. Hence, we only used the non-nested approach to evaluate
pre-treatments and extraction methods, which we detail next.

3.2 Evaluating the effects of physical or chemical treatments to improve

DNA content in frozen low microbial biomass samples
Pre-treatment methods that yield a higher DNA and bacterial content improves the chances
for better representation of the lung microbial community [41]. We tested the effectiveness of
two different mechanical and chemical DNA extraction methods.
3.2.1 BALF samples. BALF samples had the lowest DNA concentration (Fig 2A) and read
count (Fig 2B) compared to lung tissue samples (Fig 3A and 3B). The total DNA content (total
DNA extracted by the volume of the sample processed) was the lowest for PCI extracted sam-
ples and higher for two-step centrifugation followed by bead-based mechanical cell lysis (Fig
2C). Similar outcomes were observed for estimated bacterial content per volume (μL) of sam-
ples processed using different forms of pre-treatments and methods (Fig 2D). Pre-treatment
by one-step centrifugation resulted in the highest DNA concentration (average of 3.7 ng/μL)
and bacterial content (average of 3825 reads) of the BALF samples. Overall, PCI was the least
effective method for extracting DNA from BALF samples, the mean bacterial DNA content
was estimated at 301 reads associated to contamination potentially from the reagents (S2 Fig).
In contrast, the pre-treatment of BALF samples with one-step centrifugation followed by
bead-based mechanical lysis (Method 2.4.3.1) increased the DNA concentration (Fig 2A) and
significantly increased 16S read count when compared to other extraction methods (Fig 2B).
Fig 2. DNA content and estimated biomass of BALF samples. A) DNA concentration (ng/μL) of BALF samples, B) total read count is estimated based on the 16S rRNA
gene copy number after quality filtering of the data, C) estimated DNA content dividing total DNA extracted by volume of the sample processed, and D) estimated
bacterial DNA content dividing read count by volume of the sample processed. Significance (p-value) between groups is indicated by stars (� <0.05 and �� < 0.01). Based
on overall data, most effective pre-treatment method for BALF samples was a one Method centrifugation followed by mechanical cell lysis. Two Method centrifugation
was less effective.

Thus, one-step centrifugation followed by bead-based mechanical lysis of BALF samples was
the more efficient and high yielding DNA extraction method and thus produced a higher 16S
copy number library.
3.2.2. Lung tissue samples. Lung tissue samples naturally had higher total DNA concen-
tration because of the high amount of host DNA (Fig 3A). Furthermore, host DNA did not
inhibit the specificity of the V4 16S primers (Fig 3B). The mean DNA concentration of the
lung tissue sample per pre-treatment method was as follows: Vortex (609 ng/μL) > Mini-Bead-
beater (520 ng/μL) > PowerLyzer (475 ng/μL) > Proteinase K (371 ng/μL) > PCI (15 ng/μL).
While the mean 16S read count was as follows: PCI (14,176 reads) > PowerLyzer (10,781
reads) > Vortex (8,899 reads) > Mini-Beadbeater (1,193 reads) > Proteinase K (960 reads).
Mechanical cell lysis of lung tissue using the Mini-Beadbeater resulted in improved total DNA
content per weight (mg) (Fig 3C). However, PowerLyzer and Vortex-Genie 2 produced a
higher estimated bacterial content per weight (mg) than the Mini-Beadbeater (Fig 3D), sug-
gesting they are more efficient in extracting high yielding DNA and 16S copy number for lung
tissue samples. PCI method resulted in a low DNA concentration of lung tissue samples (Fig
3A) and bacterial reads obtained from PCI were substantially associated with contamination
potentially from the reagents used during extraction [28] (S2 Fig). We further validated the
effectiveness of the commercial kit method on the positive controls (ZymoBIOMICS Mock
Community), which effectively lysed representative bacterial species resulting in similar
Fig 3. DNA content and estimated biomass of lung tissue samples. A) DNA concentration (ng/μL) of lung tissue samples, B) total read count is estimated based on the
16S rRNA gene copy number after quality filtering of the data, C) estimated DNA content dividing total DNA extracted by weight of the sample processed, and D)
estimated bacterial DNA content dividing read count by weight of the sample processed. Significance (p-value) between groups is indicated by stars (� <0.05, �� < 0.01,
��
< 0.001, and �� < 0.0001). Overall, most mechanically lysed tissue samples resulted in high DNA yield with exception of the PCI treated samples.

compositions as the one suggested by the ZymoBIOMICS 16S rRNA gene abundance (S3 Fig).
Overall, the most effective extraction method for frozen lung tissue samples was mechanical
cell lysis using bead-based equipment (i.e., PowerLyzer or Vortex-Genie 2).
3.3 Evaluating bacterial diversity of distinct sample types

At phylum level, we observed that BALF and lung tissue samples had similar microbial pro-
files, even if each was processed using different forms of pre-treatment methods. Bacteroidetes,
Firmicutes and Proteobacteria were abundant among all samples (Fig 4A). We used two types
of lung tissue samples, unlavaged and lavaged. In addition, we compared these lung samples
with BALF. According to alpha diversity measures, species richness is more abundant in the
BALF than in the lavaged lung samples (p < 0.05) (Fig 4B). Surprisingly, lavaged lungs had
only slightly lower diversity than unlavaged lung samples (Fig 4B). Overall, beta diversity plots
Fig 4. Diversity analysis of BALF and lung tissue samples. A) Bacterial composition of lung and BALF samples. Overall,
samples had a higher abundance of Proteobacteria and Firmicutes bacteria. B) Alpha diversity measures shows that BALF
samples have a higher species richness compared to both lavaged and unlavaged lung samples. Observed significant (p < 0.05)
difference in species richness between BALF and lavaged lung samples. C) Beta diversity plot shows admixture of sample types.
D) Statistical summary of ADONIS test shows the variance and significance when comparing mice age and extraction pre-
treatment methods.

demonstrated similar microbial composition in the three sample types (Fig 4C). Permutation
test suggests that sample pre-treatment significantly (p-value 0.002) influenced the microbial
communities (Fig 4D). Samples were sequenced on the same MiSeq lane, thus reducing bias
introduced by the difference in sequencing depth. Overall, BALF samples showed the most
efficient way to test the pulmonary microbiome.
3.4 Microbiome DNA enrichment kit improves recovery of microbial DNA

from frozen lung tissue samples
To improve 16S library prep of low microbial biomass samples, NEBNext Microbiome DNA
Enrichment Kit was used to increase the bacterial to host DNA ratio. This kit depletes host
DNA by targeting and reducing methylated host DNA from the sample [42]. Select samples
were divided into two DNA aliquots to compare the effect of the NEBNext kit. No significant
change of the 16S copy number was observed in the enriched BALF samples (Fig 5A). How-
ever, significant (p < 0.05) improvement was observed for the microbiome enriched lung tis-
sue samples (Fig 5B). When comparing microbial composition, differences in bacterial profiles
were observed at genus level between non-treated and NEBNext treated samples. These differ-
ences in genus level profiles were more prominent in the lung tissue than in the BALF, primar-
ily when using diluted DNA (Fig 5C and 5D). The NEBNext treated BALF samples closely
resembled the microbial composition of the untreated BALF samples at the genus level (Fig
5D). Bacteria identified in NEBNext treated tissue resembled the profile observed in BALF at
phylum level (Figs 4A and S4C). Taken together, our data shows that the NEBNext increases
Fig 5. Comparing the efficiency before and after treatment with the NEBNext microbiome DNA Enrichment Kit. A) No significant improvements observed when
comparing microbiome enriched (ME) and non-treated (NT) BALF samples total read count. B) However, significant improvements observed when comparing DME
(1:10 diluted sample then treated with ME kit), ME, and NT lung samples total read count. Significance (p-value) between groups is indicated by stars (� <0.05 and �� <
0.01). Both C) BALF and D) lung samples showed difference in community composition (genus level) between NT and ME samples, particularly diluted lung samples
treated with microbiome enrichment kit.

bacterial DNA yields of lung tissue samples. However, slight differences in recovered microbial
profiles were observed between NEBNext treated and non-treated samples.
4. Discussion
The purpose of this study was to develop an optimal method of sample homogenization, DNA
extraction and library preparation for low microbial biomass samples. Culture-based methods
estimate that murine lungs harbor about 103–105 colony forming units (CFU) per gram of
lung tissue [43]. Having a low microbial density and at the same time a complex tissue struc-
ture, lung sample pre-treatment methods are necessary to effectively lyse all cell types and
increase the DNA yield. The estimated DNA and bacterial content is used to determine effec-
tive pre-treatment methods for the samples. Here we aimed to define the most efficient
method to ensure isolation of high yield DNA from low microbial biomass samples to effec-
tively study the pulmonary microbiome.
We first tested two DNA extraction methods, an experimental DNA isolation method with
a short workflow and the traditional PCI. Currently, PCI is the most common and cost-effi-
cient method for DNA isolation used in microbiome studies. The method claims to deliver
high yielding and quality DNA [25]. Here we tested PCI isolation method to examine the latter
claim and compare to an extraction kit to establish a standard protocol for DNA extraction of
low microbial biomass samples. Overall, PCI isolation method produced a lower DNA yield
resulting in a poor representation of the pulmonary microbial communities after 16S sequenc-
ing. Of note, in the PCI method a high level of bacterial DNA was isolated from the contami-
nated reagents causing to suppress the 16S signal of the lung microbial community. Similar
observations have been shown in studies targeting the placental or sputum microbiome, where
reagent contamination had an effect on the microbial profiles [26, 44]. PCI method was also
found not ideal for DNA isolation of low biomass samples due to multiple treatments and
washing steps which resulted in DNA loss [45]. In addition, trace amounts of phenol may be
carried over and potentially interfere with the downstream processing of the sample [46, 47].
In sum, PCI isolation method is not optimal for processing low biomass samples such as those
targeting lung microbiome analysis.
Compared to the PCI method, the experimental method using a combination of bead-
based lysis and an extraction kit, resulted in minimal sample loss and contamination. In con-
trast to other commercial extraction kits, QIAGEN PowerSoil Pro was used since it can be
applied to a broad range of sample types (in addition to soil or stool samples) and customizable
protocol. Our experimental approach that further customized and improved the kit’s ability to
extract high quality DNA. Other extraction kits, like the Molzym Ultra-Deep Microbiome
Prep, QIAamp DNA Microbiome and Zymo HostZERO microbial DNA kits (among others),
have been used as an alternative to process samples containing a low amount of microbial bio-
mass with variable results [42, 48]. These kits selectively lyse host cells and enzymatically
(mostly using Dnase) deplete host DNA assuring high yielding bacterial DNA [42, 49]. How-
ever, the efficiency of such kits may be tied to the use of fresh samples. Freeze-thawing cycles
compromise the integrity of bacterial cell membrane [50] and thus expose bacterial DNA can
be degraded along with the host DNA in the enzymatic depletion step. The samples used in the
presented study were previously frozen and the bacterial cell wall integrity has been compro-
mised through freeze-thaw cycles, by using the pre-treatments and kit extraction method we
were able to effectively isolate high yield DNA in a manner independent of membrane integ-
rity. Thus, our experimental approach provides an ideal methodology for groups assessing
microbial communities in samples that have undergone freezing and long-term storage. Taken
together, the experimental DNA isolation method was found to provide a fast and reliable

workflow that allows for streamlined sample processing, isolate high yielding DNA and is ideal
for low biomass frozen samples.
Having a low microbial density, sample pre-treatment methods are necessary to increase
DNA yield of mice lung samples. According to the overall yield (DNA concentration and read
count), lung tissue samples were efficiently processed by mechanical lysis by using either
PowerLyzer or Vortex-Genie 2 equipment. Overall, the PowerLyzer or Vortex-Genie 2 are
equally effective at lysing tissue cells because of the unique motion and horizontal tube posi-
tioning of the equipment. DNA isolation from BALF samples was successful using one-step
centrifugation followed by mechanical lysis of the cells. Previously published data had shown a
two-step centrifugation process to initially collect heavy mammalian cells at a low speed, fol-
lowed by a second centrifugation step at a higher speed to only collect bacterial cells had an
impact on microbiome profiles of BALF [18]. In contrast, in our hands, one-step centrifuga-
tion of BALF was more efficient in increasing DNA yield, since both host cell-attached and
free-living microbe DNA was isolated. Of note, many bacterial species (e.g. Salmonella, Yersi-
nia, Pseudomonas, Klebsiella, and Haemophilus sp.) have been shown to strongly adhere to
host cells [50]. Moreover, discarding host cells prior to isolating DNA for a microbiome study
deprives of any information regarding intracellular microbes (e.g. Chlamydia and Mycoplasma
sp.), which can be essential in lung research [51]. Furthermore, pelleting the cells does not
affect the reaction volume or the downstream chemistry of the kit. We suggest a one-step cen-
trifugation protocol as a more representative of the overall pulmonary microbial community.
For library preparation, we experimented with nested PCR at different cycle numbers and
observed higher level of contamination. Similar to our findings, Drengenes et al. showed that dur-
ing library preparation for human pulmonary samples, nested PCR increased contamination
causing a decrease in 16S signal [52]. In our study, contamination was more evident in the nega-
tive controls, nested PCR more than doubled the amplification of DNA contamination compared
to the non-nested negative controls. In fact, other studies (metagenomics and 16S-based analysis)
have stressed the importance and use of negative and other extraction controls to distinguish the
source of contamination in low microbial biomass samples [44, 53, 54]. Besides nested PCR,
other library preparation methods are recommended to increase 16S coverage of low microbial
biomass samples with minimal amplification of environmental contaminants. For instance,
slightly increasing PCR cycle number (from 35 to 40 cycles) or doubling library preparation of a
sample can improve the 16S copy number [18, 54]. Whichever method is decided, the use of posi-
tive and negative controls during DNA extraction and library preparation is key for tracking con-
tamination to the source, which can be later removed using diverse bioinformatic tools.
Considering the excess amount of host nucleic acid after DNA extraction from lung tissue
and BALF, we expected loss of specificity and reduced annealing of the16S rRNA primers dur-
ing library preparation. We aimed to improve host-to-bacteria DNA ratio using an enrichment
kit to determine if the representation of the pulmonary microbial community in our frozen
samples could be further improved. A subset of the samples was processed using the NEBNext
Microbiome DNA Enrichment kit to deplete mammalian host derived DNA. Initially, we
attempted to measure the efficiency of the kit using a SYBR Green qPCR approach to assess
the host-bacteria DNA ratio by targeting host glyceraldehyde 3-phosphate dehydrogenase
(Gapdh) and 16S rRNA genes. The assay was unsuccessful due to the low sensitivity of 16S
SYBR Green qPCR assay (S5 Fig) which has a detection limit estimated at 100 bacterial cells
[55, 56]. The latter results suggest SYBR Green may not be sensitive enough to determine the
quality of the isolated DNA. In contrast, the MiSeq sequencing data provided information on
the 16S rRNA copy number between NEBNext treated and non-treated samples. Host deple-
tion was found more successful on the lung samples due to the high amount of tissue and
immune cell types [11, 12]. In BALF samples, the overall reduced presence of host cells did not

alter recovered 16S copy numbers. Therefore, it is unnecessary to treat BALF samples or other
low host DNA samples (like other lavages or swabs) with the NEBNext kit, reducing cost of
sample processing [30–57]. When analyzing 16S data, we observed slight differences in com-
munity composition between NEBNext treated and non-treated lung tissue samples. Similar
to our findings, other studies have shown changes in microbial profiles when treating human
sino-nasal swab, saliva or blood samples with the enrichment kit [30–57]. The commercial kit,
marketed as targeting and reducing host DNA, could potentially be skewing the microbiome
data by removing methylated bacterial DNA in the process [58]. Additionally, further diluting
the sample, resuspended NEBNext treated bacterial DNA in a larger volume compared to
input volume of gDNA, could be further skewing the microbiome data. However, further
investigation is required for these claims. Importantly, microbial composition of BALF and
NEBNext treated lung tissue was similar, suggesting that BALF may be the most efficient way
to assess the pulmonary microbiome.
Overall, our work suggests that the optimal processing of murine lung tissue for micro-
biome analysis is by mechanically lysis the sample using bead-based equipment (PowerLyzer
or Vortex-Genie 2), extraction kit following manufacturer’s instructions post-lysis steps, and if
possible, treating the samples with NEBNext Microbiome DNA Enrichment Kit. For BALF
samples we observed increased DNA recovery by centrifugating the samples (once at 5,000xg)
followed by mechanical lysis and extraction kit following manufacturer’s instructions post-
lysis steps. For all sample types, the microbial profile was successfully represented with mini-
mal contamination using a library preparation method with 35 PCR cycles (non-nested PCR).
5. Conclusion
We have developed a customized protocol to efficiently extract and 16S-sequence mice BALF
and lung tissue samples to study the pulmonary microbiome. Alternately, the protocol can also
be applied to other tissue and lavage samples with low microbial density. The protocol can be
used to study the effects of the environment or disease state between host and microbes.
Supporting information
S1 Table. Brief description of cited article’s methodology and findings.
(PDF)
S1 Fig. Total read count (16S rRNA copy number) of extraction negative controls. A) Data
shows that nested PCR more than doubles amplification of contamination. Significant
(p < 0.01) difference observed when comparing nested (35x25) and non-nested PCR. B)
Nested samples exponentially increased contamination in all negative control. Thus, nested
PCR samples were discarded form the analysis and from further processing.
(PDF)
S2 Fig. Total read count (16S rRNA copy number) from PCI extracted samples. Most sam-
ples had a high level of contamination potentially from water and reagents by detecting a
higher abundance of Pseudomonas and Shewanella sp. overshadowing the lung microbiome
signal in most samples.
(PDF)
S3 Fig. Examine extraction efficiency of pre-treatment methods on ZymoBIOMICS micro-
bial community standards. A) Mock Community Standard I (Cat# D6300) and B) Mock Com-
munity Standard II (Cat# D6310), all pre-treatment methods give a good representation of the
microbial mock community as suggested by ZymoBIOMICS 16S theorical microbial composition.
(PDF)

S4 Fig. Diversity analysis of treated and non-treated samples with the NEBNext Micro-
biome DNA Enrichment Kit. Abbreviation for non-treated (NT), microbiome enriched (ME)
and diluted sample then treated with ME kit (DME). A) Alpha diversity measures shows that
BALF samples, treated samples have a lower species diversity comparted to the non-treated
samples. B) Alpha diversity measures shows that lung tissue samples, slight increase species
richness was observed for treated lung samples. C) Bacterial composition of lung and BALF
samples. Overall, samples had a higher abundance of Proteobacteria and Firmicutes bacteria.
(PDF)
S5 Fig. Comparing 16S/GAPDH primers and controls for Sybr Green qPCR assay. Both
16S and GAPDH genes had a low primer specificity in the qPCR assay. For 16S qPCR assay,
cutoff was set to 100 copies. GAPDH cutoff was set to 1000 copies.
(PDF)
S1 Data.
(XLSX)
Acknowledgments
We would like to thank Vikash C. Tripathi for handling the mice. In addition, we would also
like to thank Claire Kuelbs from the JCVI Sequencing facility, for library preparation and
sequencing our samples.
Author Contributions
Conceptualization: Rosana Wiscovitch-Russo, Harinder Singh.
Data curation: Rosana Wiscovitch-Russo.
Formal analysis: Rosana Wiscovitch-Russo, Lauren M. Oldfield, Alexey V. Fedulov.
Funding acquisition: Lauren M. Oldfield, Alexey V. Fedulov, Norberto Gonzalez-Juarbe.
Investigation: Rosana Wiscovitch-Russo, Lauren M. Oldfield, Alexey V. Fedulov, Norberto
Gonzalez-Juarbe.
Methodology: Rosana Wiscovitch-Russo.
Visualization: Rosana Wiscovitch-Russo, Harinder Singh.
Writing – original draft: Rosana Wiscovitch-Russo, Norberto Gonzalez-Juarbe.
Writing – review & editing: Rosana Wiscovitch-Russo, Harinder Singh, Lauren M. Oldfield,
Alexey V. Fedulov, Norberto Gonzalez-Juarbe.
References
1. Baughman RP, Thorpe JE, Staneck J, Rashkin M, Frame PT. Use of the protected specimen brush in
patients with endotracheal or tracheostomy tubes. Chest. 1987; 91(2):233–6. https://doi.org/10.1378/
chest.91.2.233 PMID: 3802934
2. Jordan GW, Wong GA, Hoeprich PD. Bacteriology of the lower respiratory tract as determined by fiber-
optic bronchoscopy and transtracheal aspiration. J Infect Dis. 1976; 134(5):428–35. https://doi.org/10.
1093/infdis/134.5.428 PMID: 993617
3. Thorpe JE, Baughman RP, Frame PT, Wesseler TA, Staneck JL. Bronchoalveolar lavage for diagnos-
ing acute bacterial pneumonia. J Infect Dis. 1987 May 1; 155(5):855–61. https://doi.org/10.1093/infdis/
155.5.855 PMID: 3559289

4. Mongodin EF, Emerson JB, Nelson KE. Microbial Metagenomics. Genome Biol. 2005; 6, 347. https://
doi.org/10.1186/gb-2005-6-10-347 PMID: 16207365
5. Human Microbiome Project Consortium. Structure, function and diversity of the healthy human micro-
biome. Nature 2012; 486. https://doi.org/10.1038/nature11234 PMID: 22699609
6. Integrative H. M. P. Research Network Consortium. The Integrative Human Microbiome Project. Nature
2019; 569, 641–48. https://doi.org/10.1038/s41586-019-1238-8 PMID: 31142853
7. Barcik W, Boutin RC, Sokolowska M, Finlay BB. The role of lung and gut microbiota in the pathology of
asthma. Immunity 2020; 52, 241–255. https://doi.org/10.1016/j.immuni.2020.01.007 PMID: 32075727
8. Hufnagl K, Pali-Schöll I, Roth-Walter F, Jensen-Jarolim E. Dysbiosis of the gut and lung microbiome
has a role in asthma. Sem. Immun 2020; 42, 75–93.
9. Charlson ES, Bittinger K, Haas AR, Fitzgerald AS, Frank I, Yadav A, et al. Topographical continuity of
bacterial populations in the healthy human respiratory tract. Am J Respir Crit Care Med. 2011; 184
(8):957–6. https://doi.org/10.1164/rccm.201104-0655OC PMID: 21680950
10. Dickson RP, Erb-Downward JR, Freeman CM, McCloskey L, Falkowski NR, Huffnagle GB, et al. Bacte-
rial topography of the healthy human lower respiratory tract. MBio. 2017; 8(1):e02287–16 https://doi.
org/10.1128/mBio.02287-16 PMID: 28196961
11. Huffnagle GB, Dickson RP, Lukacs NW. The Respiratory Tract Microbiome and Lung Inflammation: A
Two-Way Street. Mucosal Immunol. 2017; 10, 299–306. https://doi.org/10.1038/mi.2016.108 PMID:
27966551
12. Hilty M, Burke C, Pedro H, Cardenas P, Bush A, Bossley C, et al. Disordered microbial communities in
asthmatic airways. PloS one. 2010; 5(1):e8578. https://doi.org/10.1371/journal.pone.0008578 PMID:
20052417
13. Durack J, Boushey HA, Lynch SV. Airway Microbiota and the Implications of Dysbiosis in Asthma. Curr
Allergy Asthma Rep. 2016, 16, 52. https://doi.org/10.1007/s11882-016-0631-8 PMID: 27393699
14. Kumpitsch C, Koskinen K, Schöpf V, Moissl-Eichinger C. The Microbiome of the Upper Respiratory
Tract in Health and Disease. BMC Biol. 2019, 17, 87. https://doi.org/10.1186/s12915-019-0703-z
PMID: 31699101
15. Allaband C, McDonald D, Vázquez-Baeza Y, Minich JJ, Tripathi A, Brenner DA, et al. Microbiome 101:
studying, analyzing, and interpreting gut microbiome data for clinicians. Clinical Gastroenterology and
Hepatology. 2019; 17(2):218–30. https://doi.org/10.1016/j.cgh.2018.09.017 PMID: 30240894
16. Greathouse KL, Sinha R, Vogtmann E. DNA extraction for human microbiome studies: the issue of
standardization. Genome Biol 2019, 20, 1–4. https://doi.org/10.1186/s13059-018-1612-0 PMID:
30606230
17. Carney SM, Clemente JC, Cox MJ, Dickson RP, Huang YJ, Kitsios GD, et al. Methods in lung micro-
biome research. Am J Respir Cell Mol Biol. 2020; 62(3):283–99. https://doi.org/10.1165/rcmb.2019-
0273TR PMID: 31661299
18. Barfod KK, Roggenbuck M, Hansen LH, Schjørring S, Larsen ST, Sørensen SJ, et al. The murine lung
microbiome in relation to the intestinal and vaginal bacterial communities. BMC microbiology. 2013; 13
(1):1–2
19. Kostric M, Milger K, Krauss-Etschmann S, Engel M, Vestergaard G, Schloter M, et al. Development of a
Stable Lung Microbiome in Healthy Neonatal Mice. Microb Ecol. 2018; 75. https://doi.org/10.1007/
s00248-017-1068-x PMID: 28905200
20. Singh N, Vats A, Sharma A, Arora A, Kumar A. The Development of Lower Respiratory Tract Micro-
biome in Mice. Microbiome 2017, 5, 61. https://doi.org/10.1186/s40168-017-0277-3 PMID: 28637485
21. Yun Y, Srinivas G, Kuenzel S, Linnenbrink M, Alnahas S, Bruce K.D, et al. Environmentally Determined
Differences in the Murine Lung Microbiota and Their Relation to Alveolar Architecture. PLoS One 2014;
9, e113466. https://doi.org/10.1371/journal.pone.0113466 PMID: 25470730
22. Ericsson AC, Davis JW, Spollen W, Bivens N, Givan S, Hagan CE, et al. Effects of vendor and genetic
background on the composition of the fecal microbiota of inbred mice. PloS one. 2015; 10(2):e0116704
https://doi.org/10.1371/journal.pone.0116704 PMID: 25675094
23. Yang J, Park J, Park S, Baek I, Chun J. Introducing murine microbiome database (MMDB): a curated
database with taxonomic profiling of the healthy mouse gastrointestinal microbiome. Microorganisms.
2019; 7(11):480 https://doi.org/10.3390/microorganisms7110480 PMID: 31652812
24. Dumont-Leblond N, Veillette M, Racine C, Joubert P, Duchaine C. Development of a robust protocol for
the characterization of the pulmonary microbiota. Commun Biol 2021, 4, 1–9. https://doi.org/10.1038/
s42003-020-01566-0 PMID: 33398033
25. Pérez-Brocal V, Magne F, Ruiz-Ruiz S, Ponce CA, Bustamante R, San Martin V, et al. Optimized DNA
extraction and purification method for characterization of bacterial and fungal communities in lung tissue
samples. Scientific reports. 2020; 10(1):1–5. https://doi.org/10.1038/s41598-019-56847-4 PMID: 31913322

26. Sui HY, Weil AA, Nuwagira E, Qadri F, Ryan ET, Mezzari MP, et al. Impact of DNA extraction method
on variation in human and built environment microbial community and functional profiles assessed by
shotgun metagenomics sequencing. Frontiers microbiol. 2020; 11:953. https://doi.org/10.3389/fmicb.
2020.00953 PMID: 32528434
27. Fedulov AV, Silverman E, Xiang Y, Leme A, Kobzik L. Immunostimulatory CpG oligonucleotides abro-
gate allergic susceptibility in a murine model of maternal asthma transmission. J Immunol 2005, 175,
4292–4300. https://doi.org/10.4049/jimmunol.175.7.4292 PMID: 16177069
28. Fedulov AV, Leme A, Yang Z, Dahl M, Lim R, Mariani TJ, et al. Pulmonary exposure to particles during
pregnancy causes increased neonatal asthma susceptibility. Am J Respir Cell Mol Biol 2008, 38, 57–
67. https://doi.org/10.1165/rcmb.2007-0124OC PMID: 17656681
29. Freire M, Moustafa A, Harkins DM, Torralba MG, Zhang Y, Leong P, et al. Longitudinal study of oral
microbiome variation in twins. Scientific reports. 2020; 10(1):1–0. https://doi.org/10.1038/s41598-019-
56847-4 PMID: 31913322
30. Feehery GR, Yigit E, Oyola SO, Langhorst BW, Schmidt VT, Stewart FJ, et al. A method for selectively
enriching microbial DNA from contaminating vertebrate host DNA. PloS one. 2013; 8(10):e76096.
https://doi.org/10.1371/journal.pone.0076096 PMID: 24204593
31. Kozich JJ, Westcott SL, Baxter NT, Highlander SK, Schloss PD. Development of a Dual-Index
Sequencing Strategy and Curation Pipeline for Analyzing Amplicon Sequence Data on the Miseq Illu-
mina Sequencing Platform. Appl Environ Microbiol. 2013; 79, 5112–20. https://doi.org/10.1128/AEM.
01043-13 PMID: 23793624
32. Youssef N, Sheik C.S, Krumholz LR, Najar FZ, Roe BA, Elshahed MS. Comparison of Species Rich-
ness Estimates Obtained Using Nearly Complete Fragments and Simulated Pyrosequencing-Gener-
ated Fragments in 16s Rrna Gene-Based Environmental Surveys. Appl Environ Microbiol. 2009; 75,
5227–36. https://doi.org/10.1128/AEM.00592-09 PMID: 19561178
33. Wang Q, Garrity GM, Tiedje JM, Cole JR. Naive Bayesian Classifier for Rapid Assignment of Rrna
Sequences into the New Bacterial Taxonomy. Appl Environ Microbiol. 2007; 73, 5261–7. https://doi.
org/10.1128/AEM.00062-07 PMID: 17586664
34. Schloss PD, Westcott SL, Ryabin T, Hall JR, Hartmann M, Hollister EB, et al. Introducing mothur: open-
source, platform-independent, community-supported software for describing and comparing microbial
communities. Appl Environ Microbiol. 2009; 75(23):7537–41. https://doi.org/10.1128/AEM.01541-09
PMID: 19801464
35. Edgar RC. Uparse: Highly Accurate Otu Sequences from Microbial Amplicon Reads. Nat Methods
2013; 10, 996–8. https://doi.org/10.1038/nmeth.2604 PMID: 23955772
36. Quast C, Pruesse E, Yilmaz P, Gerken J, Schweer T, Yarza P, et al. The SILVA ribosomal RNA gene
database project: improved data processing and web-based tools. Nucleic Acids Res. 2012; 41(D1):
D590–6. https://doi.org/10.1093/nar/gks1219 PMID: 23193283
37. McMurdie PJ, Holmes S. Phyloseq: A Bioconductor Package for Handling and Analysis of High-
Throughput Phylogenetic Sequence Data. Pac Symp Biocomput. 2012; 235–46. PMID: 22174279
38. Singh H, Miyamoto S, Darshi M, Torralba MG, Kwon K, Sharma K, et al. Gut microbial changes in dia-
betic db/db mice and recovery of microbial diversity upon pirfenidone treatment. Microorganisms. 2020;
8(9):1347. https://doi.org/10.3390/microorganisms8091347 PMID: 32899353
39. Glassing A, Dowd SE, Galandiuk S, Davis B, Chiodini RJ. Inherent bacterial DNA contamination of
extraction and sequencing reagents may affect interpretation of microbiota in low bacterial biomass
samples. Gut pathogens. 2016; 8(1):1–2. https://doi.org/10.1186/s13099-016-0103-7 PMID: 27239228
40. Salter SJ, Cox MJ, Turek EM, Calus ST, Cookson WO, Moffatt MF, et al. Reagent and laboratory con-
tamination can critically impact sequence-based microbiome analyses. BMC biol. 2014; 12(1):1–2.
https://doi.org/10.1186/s12915-014-0087-z PMID: 25387460
41. Vishnivetskaya TA, Layton AC, Lau MC, Chauhan A, Cheng KR, Meyers AJ, et al. Commercial DNA
extraction kits impact observed microbial community composition in permafrost samples. FEMS micro-
biol ecol. 2014; 87(1):217–30. https://doi.org/10.1111/1574-6941.12219 PMID: 24102625
42. Heravi FS, Zakrzewski M, Vickery K, Hu H. Host DNA depletion efficiency of microbiome DNA enrich-
ment methods in infected tissue samples. J Microbiol Methods 2020; 170, 105856. https://doi.org/10.
1016/j.mimet.2020.105856 PMID: 32007505
43. Remot A, Descamps D, Noordine ML, Boukadiri A, Mathieu E, Robert V, et al. Bacteria isolated from
lung modulate asthma susceptibility in mice. ISME J. 2017; 11(5):1061–74. https://doi.org/10.1038/
ismej.2016.181 PMID: 28045458
44. Benny PA, Al-Akwaa FM, Dirkx C, Schlueter RJ, Wolfgruber TK, Chern IY, et al. Placentas delivered by
pre-pregnant obese women have reduced abundance and diversity in the microbiome. FASEB J. 2021;
35(4):e21524. https://doi.org/10.1096/fj.202002184RR PMID: 33742690

45. van Heesch S, Mokry M, Boskova V, Junker W, Mehon R, Toonen P, et al. Systematic biases in DNA
copy number originate from isolation procedures. Genome biol. 2013; 14(4):1–9. https://doi.org/10.
1186/gb-2013-14-4-r33 PMID: 23618369
46. McOrist AL, Jackson M, Bird AR. A Comparison of Five Methods for Extraction of Bacterial DNA from
Human Faecal Samples. J Microbiol Methods 2002; 50, 131–9. https://doi.org/10.1016/s0167-7012
(02)00018-0 PMID: 11997164
47. Toni LS, Garcia AM, Jeffrey DA, Jiang X, Stauffer BL, Miyamoto SD, et al. Optimization of phenol-chlo-
roform RNA extraction. MethodsX. 2018; 5:599–608. https://doi.org/10.1016/j.mex.2018.05.011 PMID:
29984193
48. Bjerre RD, Hugerth LW, Boulund F, Seifert M, Johansen JD, Engstrand L. Effects of Sampling Strategy
and DNA Extraction on Human Skin Microbiome Investigations. Sci Rep. 2019; 9, 17287. https://doi.
org/10.1038/s41598-019-53599-z PMID: 31754146
49. Marotz CA, Sanders JG, Zuniga C, Zaramela LS, Knight R, Zengler K. Improving saliva shotgun meta-
genomics by chemical host DNA depletion. Microbiome. 2018; 6(1):1–9. https://doi.org/10.1186/
s40168-017-0383-2 PMID: 29291746
50. Krachler AM, Orth K. Targeting the Bacteria-Host Interface: Strategies in Anti-Adhesion Therapy. Viru-
lence. 2013; 4, 284–94. https://doi.org/10.4161/viru.24606 PMID: 23799663
51. Fol M, Włodarczyk M, Druszczyńska M. Host Epigenetics in Intracellular Pathogen Infections. Int J Mol
Sci. 2020; 21, 4573. https://doi.org/10.3390/ijms21134573 PMID: 32605029
52. Drengenes C, Eagan TM, Haaland I, Wiker HG, Nielsen R. Exploring protocol bias in airway microbiome
studies: One versus two PCR steps and 16S rRNA gene region V3 V4 versus V4. BMC genomics.
2021; 22, 1–15. https://doi.org/10.1186/s12864-020-07350-y PMID: 33388042
53. Leiby JS, McCormick K, Sherrill-Mix S, Clarke EL, Kessler LR, Taylor LJ, et al. Lack of detection of a
human placenta microbiome in samples from preterm and term deliveries. Microbiome. 2018; 6(1):1–1.
https://doi.org/10.1186/s40168-017-0383-2 PMID: 29291746
54. Dickson RP, Erb-Downward JR, Falkowski NR, Hunter EM, Ashley SL, Huffnagle GB. The lung micro-
biota of healthy mice are highly variable, cluster by environment, and reflect variation in baseline lung
innate immunity. Am J Respir Crit Care Med. 2018; 198(4):497–508. https://doi.org/10.1164/rccm.
201711-2180OC PMID: 29533677
55. Maeda H, Fujimoto C, Haruki Y, Maeda T, Kokeguchi S, Petelin M, et al. Quantitative real-time PCR
using TaqMan and SYBR Green for Actinobacillus actinomycetemcomitans, Porphyromonas gingivalis,
Prevotella intermedia, tetQ gene and total bacteria. FEMS Immunol Med Microbiol. 2003; 39(1):81–6.
https://doi.org/10.1016/S0928-8244(03)00224-4 PMID: 14557000
56. Rinttilä T, Kassinen A, Malinen E, Krogius L, Palva A. Development of an Extensive Set of 16s rDNA-
Targeted Primers for Quantification of Pathogenic and Indigenous Bacteria in Faecal Samples by Real-
Time PCR. J Appl Microbiol. 2004; 97, 1166–77. https://doi.org/10.1111/j.1365-2672.2004.02409.x
PMID: 15546407
57. Mackenzie BW, Waite DW, Biswas K, Douglas RG, Taylor MW. Assessment of microbial DNA enrich-
ment techniques from sino-nasal swab samples for metagenomics. Rhinol Online. 2018; 14,160–93.
58. Sanchez-Romero MA, Casadesus J. The Bacterial Epigenome. Nat Rev Microbiol. 2020; 18, 7–20.
https://doi.org/10.1038/s41579-019-0286-2 PMID: 31728064

Pone 0265891

Uploaded by

Copyright:

Available Formats

Pone 0265891

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Pone 0265891

Uploaded by

Copyright:

Available Formats

PLOS ONE

An optimized approach for processing

Data Availability Statement: To comply with

PLOS ONE | https://doi.org/10.1371/journal.pone.0265891 April 5, 2022 1 / 17

PLOS ONE | https://doi.org/10.1371/journal.pone.0265891 April 5, 2022 2 / 17

2.2 Decontamination of workstation for sample processing and DNA

2.3 Phenol chloroform isoamyl alcohol DNA extraction

2.4 Increasing extraction efficiency through sample pre-treatments

PLOS ONE | https://doi.org/10.1371/journal.pone.0265891 April 5, 2022 3 / 17

PLOS ONE | https://doi.org/10.1371/journal.pone.0265891 April 5, 2022 4 / 17

2.5 Microbiome DNA enrichment kit

2.6 V4 16S amplification and sequencing

PLOS ONE | https://doi.org/10.1371/journal.pone.0265891 April 5, 2022 5 / 17

2.7 Bioinformatic pipeline

3.1 Nested-PCR promotes DNA contamination from environmental

PLOS ONE | https://doi.org/10.1371/journal.pone.0265891 April 5, 2022 6 / 17

3.2 Evaluating the effects of physical or chemical treatments to improve

PLOS ONE | https://doi.org/10.1371/journal.pone.0265891 April 5, 2022 7 / 17

PLOS ONE | https://doi.org/10.1371/journal.pone.0265891 April 5, 2022 8 / 17

3.3 Evaluating bacterial diversity of distinct sample types

PLOS ONE | https://doi.org/10.1371/journal.pone.0265891 April 5, 2022 9 / 17

3.4 Microbiome DNA enrichment kit improves recovery of microbial DNA

PLOS ONE | https://doi.org/10.1371/journal.pone.0265891 April 5, 2022 10 / 17

PLOS ONE | https://doi.org/10.1371/journal.pone.0265891 April 5, 2022 11 / 17

PLOS ONE | https://doi.org/10.1371/journal.pone.0265891 April 5, 2022 12 / 17

PLOS ONE | https://doi.org/10.1371/journal.pone.0265891 April 5, 2022 13 / 17

PLOS ONE | https://doi.org/10.1371/journal.pone.0265891 April 5, 2022 14 / 17

PLOS ONE | https://doi.org/10.1371/journal.pone.0265891 April 5, 2022 15 / 17

PLOS ONE | https://doi.org/10.1371/journal.pone.0265891 April 5, 2022 16 / 17

PLOS ONE | https://doi.org/10.1371/journal.pone.0265891 April 5, 2022 17 / 17

You might also like