Oral specimens as a tool for accurate metagenomic analysis: A pilot study

Full text

Original Article
Oral specimens as a tool for accurate metagenomic analysis: A pilot study
Manuel Eros Rodríguez-Fuentes
a,b,#
, Mario P
erez-Say
ans
a,b,#,
*, Gema Barbeito-Casti~
neiras
c
,
Alberto Molares-Vila
b,d
, Irene B. Prado-Pena
a
, Gisela C.V. Camolesi
a,b
, Rafael L
opez-L
opez
a,e
a
Faculty of Medicine and Dentistry, Universidade de Santiago de Compostela, c/ Entrerríos s/n, Santiago de Compostela 15782, Spain
b
Health Research Institute of Santiago de Compostela (IDIS, ORALRES Group), Santiago de Compostela, A Coru~
na 15706, Spain
c
Microbiology Unit, Complexo Hospitalario Universitario de Santiago de Compostela, Santiago de Compostela, A Coru~
na 15706, Spain
d
Health Research Institute of Santiago de Compostela (IDIS, RESMET Group, https://resmet.org), Santiago de Compostela, A Coru~
na 15706, Spain
e
Medical Oncology Unit, Complexo Hospitalario Universitario de Santiago de Compostela, Santiago de Compostela, A Coru~
na 15706, Spain
ARTICLE INFO
Article History:
Received 23 July 2024
Accepted 27 July 2024
Available online xxx
ABSTRACT
Objectives: Acute oral mucosal damage, as well as other inflammatory processes seem to be related to dysbio-
sis of the oral microbiome. The need to study changes in the oral microbiome led us to hypothesize what type
of sample would provide the most representative picture of the entire human oral microbiome.
Materials and methods: An observational, and cross-sectional study was carried out. Six healthy adult partici-
pants provided 3 different sample types each, that included saliva, oral rinse and mucosal biopsy tissue. We
performed 16S rRNA sequencing of the V3-V4 region of the 18 samples using Illumina MiSeq technology.
Results: Participants were 27 §6,3 years old. Bacterial alpha diversity was higher in oral rinse samples com-
pared to whole unstimulated saliva and oral mucosa tissue (p= 0,005). However, saliva specimens showed a
56 % relative abundance of identified species followed by a 30 % in oral rinse and only 1 % in tissue samples.
Conclusions: This study found differences on oral microbiome composition for each type of sample. Oral rinse
should be chosen when higher alpha diversity is needed, whereas whole unstimulated saliva should be more
appropriate for larger amount of bacterial DNA.
Clinical relevance: The results obtained demonstrate the importance of a correct choice of the optimal type of
oral sample for microbiome studies due to the differences found in its composition.
© 2024 The Author(s). Published by Elsevier Masson SAS. This is an open access article under the CC BY
license (http://creativecommons.org/licenses/by/4.0/)
Keywords:
Metagenomics
Microbiota
Mouth
Mouth mucosa
Saliva
1. Introduction
Acute oral mucosal damage is a frequently reported side effect in
patients treated with systemic antineoplastic agents [1-2]. Its overall
prevalence is 38,2 % [3], and its development is believed to have a
multifactorial origin [4], where oral microbiota of the host seems to
be a relevant component [5]. The study of changes in the oral micro-
biome composition may help to reveal certain biological mechanisms
that may be amenable to intervention in order to prevent or treat
this condition. This imbalance of the community structure of the
microorganisms within the oral cavity is known as dysbiosis and
appears to be closely related to local inflammatory processes [6]. The
term oral microbiota refers to all microbes inhabiting the oral cavity
including bacteria, fungi, archaea and viruses, meanwhile the oral
microbiome also comprises their genes [2,5,7]. In most cases, the
word microbiome is used imprecisely when meaning bacteriome,
since it is the most studied and known of the biomes [5]. Although
the oral microbiome is considered a single complex entity, evidence
shows that its composition can substantially differ among oral ana-
tomical locations within the same individual [8-9]. Multiple techni-
ques have been used throughout the years to identify
microorganisms; however Next Generation Sequencing has been
accepted as the most reliable technology [5,10-12]. Within this
method, shotgun metagenomics has proved to be useful when profil-
ing taxonomic composition [13−14]. Despite the technological
advances achieved so far, there are still difficulties in the taxonomic
identification of the different species of microorganisms. Previous
studies have analysed numerous types of samples from the oral cav-
ity that include supra- and subgingival dental plaque; tongue, palate
and retropharyngeal swabs, whole unstimulated saliva as well as oral
rinse and mucosa [9,15-16]. Logical thinking leads us to believe that
many of these samples, such as swabs and dental plaque, due to their
method of collection, provide us with information on subpopulations
of microorganisms. Nevertheless, acute oral mucosal damage onset
might occur at any region of the oral cavity [1]. Hence, we designed a
preliminary verification study to determine the optimal sample type
to allow the most accurate metagenomic analysis.
* Corresponding author.
E-mail address: mario.perez@usc.es (M. P
erez-Say
ans).
#
First and Second authors have participated equally.
https://doi.org/10.1016/j.jormas.2024.101991
2468-7855/© 2024 The Author(s). Published by Elsevier Masson SAS. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/)
ARTICLE IN PRESS
JID: JORMAS [m5G;August 1, 2024;22:59]
Please cite this article as: M.E. Rodríguez-Fuentes, M. P
erez-Say
ans, G. Barbeito-Casti~
neiras et al., Oral specimens as a tool for accurate
metagenomic analysis: A pilot study, Journal of Stomatology oral and Maxillofacial Surgery (2024), https://doi.org/10.1016/j.
jormas.2024.101991
J Stomatol Oral Maxillofac Surg 000 (2024) 101991
Available online at
ScienceDirect
www.sciencedirect.com

2. Material and methods
2.1. Study design and study population
This study was conceived as an observational, descriptive, and
cross-sectional pilot study. Every participant was recruited at the
Clinic of the Faculty of Dentistry by MPS. All the subjects were offered
to participate, and those that met eligibility criteria were voluntarily
enrolled. Inclusion criteria was as follows: healthy subjects who have
not been diagnosed with or are not suffering from systemic diseases
as well as not receiving treatments that alter the immune system
[13]. All of them were ≥18 years old and able to understand and
agreed to participate. On the other side, subjects could not enrol if
they were receiving anticoagulants or antiaggregant, had previous
allergic reaction to local anaesthetic drugs, were unable to compre-
hend the implications of the study, not willing to participate or
unable to give written consent. Furthermore, subjects were not eligi-
ble if they performed oral washout or brushed their teeth 12 h prior
to collecting the samples. Moreover, candidates with decrease overall
saliva rate were also excluded.
Participants’data were collected and registered in a coded data-
base to avoid their identification through personal information and
to ensure privacy. Also, three different samples were taken from each
subject following a strict order to minimize possible interferences in
the specimens: First whole unstimulated saliva, second oral rinse,
and lastly an oral mucosa biopsy. In order to reduce risks and unnec-
essary invasive techniques such as the extraction of oral mucosa
biopsy in healthy people, we chose to select candidates who required
extraction of a third molar for non-infectious causes.
The present research was approved by the Galician Ethics Com-
mittee of Clinical Research (Ref. No. 2020/552) and conducted in
accordance with the guidelines for Good Clinical Practice and the
Declaration of Helsinki. All participants signed the informed consent
form.
2.2. Specimen’s collection
Each subject agreed to provide three different biological speci-
mens for the study. Unstimulated whole saliva was the first sample
obtained from patients by leaning their heads forward and letting
saliva flow towards a sterile empty Eppendorf sample container until
reaching a volume of 5 mL. Next, participants were asked to rinse
mouth with 5 mL of 0,9 % sodium chloride (NaCl) solution for 1 min
[16], ensuring they reached every part of the oral cavity with it and
finally spitting it into a sterile empty sample container without any
other medium. Finally, a biopsy of the oral mucosa was aseptically
extracted by an odontologist who infiltrated local anaesthesia in the
chosen area of the participant’s oral cavity, a 5 £5 mm tissue sample
was removed with a scalpel and the resulting incision was sutured
with silk to stop bleeding and prevent future infection. The selected
fragment corresponded to tissue adjacent to the gingiva of the third
molar. These tissue samples were collected into a sterile sample con-
tainer with 0,9 % sodium chloride to avoid drying out of the speci-
men. All samples were stored frozen at 80 °C from collection, as
subjects were enrolled, until all samples were collected for simulta-
neous processing.
2.3. DNA extraction, quantification and purity assessment
The isolation of DNA from all samples was carried out by using
MaxwellÒRSC PureFood GMO and Authentication Kit (Cat.# AS1600,
Promega Corporation, Madison, WI) adhering to the manufacturer’s
instructions. Liquid samples such as whole saliva and oral rinse were
centrifuged prior to DNA extraction treatment to increase the con-
centration of genetic material. Once supernatant was discarded, sam-
ples were re-suspended in lysis buffer (1 ml of CTAB), subsequently
samples were processed following the same protocol as tissue sam-
ples. For solid samples, 150 mg of tissue were placed into a conical
tube and 1 ml of CTAB Buffer was added followed by homogenization
with a disposable pestle for 40 s. Resulting liquid was then trans-
ferred into a Lysing Matrix E bead beating tube and vortexed for 30 s.
Next, samples were taken to 95 °C for 10 min and cooled down dur-
ing 1 min at room temperature and vortexed once again. Then, 40ml
of Proteinase K and 20ml of RNase A were added and mixed using a
vortex followed by incubation at 70 °C for 10 min. Consecutively, RSC
cartridges were prepared and placed in the rack according to manu-
facturer. Elution tubes with samples were filled with 100ml of Elution
Buffers and placed in cartridge to, finally, run MaxwellÒRSC with the
PureFood GMO and Authentication Protocol. Extracted DNA samples
were stored at 20 °C until further processing.
DNA concentrations were measured using the QuantiFluorÒONE
dsDNA System (Promega Corporation, Madison, WI, USA) on a
QuantusÒfluorometer (Promega Corporation). In addition, purity
was also evaluated as previously described [17]. DNA samples were
frozen at 20 °C until further processing.
2.4. Library construction and 16S rRNA gene sequencing
Library generation was carried out through amplification of the V3
- V4 region of the 16S ribosomal RNA gene. Sequencing was per-
formed on an Illumina MiSeq (Illumina, San Diego, CA) using the
MiSeq v3 reagent kit. All samples were randomized and normalized
to 5 ng/mL with PCR-grade water. Amplicons of approximately
460 bp were generated using primers from Integrated DNA Technolo-
gies, Inc. flanked by Illumina overhang adapters (Forward overhang:
50TCGTCGGCAGCGTCAGATGTGTATAAGAGACAG; Reverse overhang:
50GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAG.). The first PCR was
carried out using a 25 mL reaction consisting of 5 ng/mL template
DNA from 2X KAPA HiFi HotStart ReadyMix (KAPA Biosystems, Wil-
mington, MA, USA). The thermal cycler was set as follows: 95 °C
(3:00) + [95 °C(0:30) + 55 °C(0:30) + 72 °C(0:30)] x 25 cycles + 72 °C
(5:00) and holding at 4 °C. For the second round of PCR, we added
Illumina sequencing adapters as well as dual-index barcodes were
added in a 50 mL PCR reaction from 5 mL of amplicon PCR product for
8 cycles using 2X KAPA HiFi HotStart ReadyMix using the same
parameters [17].
Database used was Silva version 138_99. Illumina sequencing raw
data are available in the BioStudies database (https://www.ebi.ac.uk/
biostudies/) under accession number S-BSST1150.
2.5. Microbiome and statistical analysis
The sequencing data generated from the Illumina MiSeq
sequencer was analyzed adhering to the methodology described in
the literature [17−24] using the following software: BaseSpace
Sequence Hub version: 7.11.0; Python version: 3.8.13; QIIME 2 ver-
sion: 2022.2.0 [19]; q2cli version: 2022.2.0; FastQC v0.11.9 [20]; fastp
0.23.2 [21]; cutadapt 3.7 [22]; multiqc, version 1.12 [24].
2.6. Bioinformatic quality control measures
FastQC was carried out with default flags. Cutadapt was used to
remove forward TCGTCGGCAGCGTCAGATGTGTATAAGAGACAGCC-
TACGGGNGGCWGCAG and reverse GTCTCGTGGGCTCGGAGATGTG-
TATAAGAGACAGGACTACHVGGGTATCTAATCC primers from
sequences to analyze, following suggestions from Illumina 16S Meta-
genomics Sequencing Workflow (https://support.illumina.com/con-
tent/dam/illumina-marketing/documents/products/other/16s-meta-
genomics-faq-1270−2014−003.pdf), and rest of default flags.
Fastp was carried out with cutadapt output files and following
flags (the rest are by default): trim_poly_x (enable polyX trimming in
30ends); cut_front −cut_tail (cutting extremes 50and 30of sequence,
ARTICLE IN PRESS
JID: JORMAS [m5G;August 1, 2024;22:59]
M.E. Rodríguez-Fuentes, M. P
erez-Say
ans, G. Barbeito-Casti~
neiras et al. Journal of Stomatology oral and Maxillofacial Surgery 00 (2024) 101991
2

drop the bases in the window if its mean quality <threshold, stop
otherwise); cut_mean_quality=15 (the mean quality requirement
option shared by cut_front and cut_tail); length_required=220 (reads
shorter will be discarded); qualified_quality_phred=15 (means phred
quality >=Q15 is qualified); unqualified_percent_limit=40 (40 % of
bases are allowed to be unqualified). Alpha-rarefaction, from Qiime 2
tool, was used with flags by default.
3. Results
3.1. Characteristics of the study cohort
This pilot study included 6 participants (4 female and 2 male) with
a mean age of 27 §6,3 years. Other sociodemographic data of the
subjects are shown in Table 1. All of them were healthy subjects with
caries, however none of them suffered from periodontitis, had
implants, prosthetic teeth or were diagnosed with any medical condi-
tion of the gastrointestinal tract. Only 2 (33,33 %) of them were cur-
rent smokers (of 5 cigarettes daily) at that time. In addition, 4
(66,67 %) out of the 6 participants had received systemic antibiotics a
mean of 578,75 (§475,27) days prior to enrolment, therefore this is
not expected to alter the results of the oral microbiome analysis.
Antibiotic treatments received by the participants are also displayed
on Table 1.
3.2. Sequence reads and composition of the salivary microbiome
Each participant provided 3 samples, which means that the study
comprised 18 specimens (Supplementary Table 1). The quality con-
trol analysis performed demonstrates that nearly all reads of the
samples were in the range between 30 (0.1 % reading errors) and 40
(0.01 % reading errors) on the Phred Score Scale (Fig. 1a and Fig. 1b).
3.3. Quantitative metagenomic data
The mean of total reads in saliva samples was
265,452 §507,833,57 while in oral rinse was 209,772 §382,954,80
and in tissue specimens 57,936,17§27,466,13 as detailed in Table 2.
Saliva was the type of sample where a higher mean number of
species were identified (638,00 §559,99), followed by oral rinse
(618,17 §426,2) and lastly by tissue samples (404,67 §125,42).
According to phylum level, Firmicutes were the most abundant
microorganism throughout all the specimens (55,57% in saliva,
55,25 % in oral rinse and 34,35 % in oral mucosa). Bacteroidetes were
the second most abundant phylum (18,25 %) in saliva samples fol-
lowed by Actinobacteria (13,4 %). However, Actinobacteria (15,54 %)
and Bacteroidetes (13,08 %) rank second and third in terms of oral
rinse abundance. Whereas 12,68 % of oral mucosa bacteria were
unclassified at phylum level and 14,71 % were Fusobacteria (14,71 %)
as shown in Supplementary Figure 1.
Focusing on the lowest taxonomic category, illustrated in Supple-
mentary Figure 2, we found that in saliva samples the highest per-
centage were unclassified species (35,86 %) followed by Rothia
mucilaginosa (5,81 %), Granulicatella adiacens (5,09 %), Prevotella mela-
ninogenica (4,43 %) and Actinomyces odontolyticus (3,42 %). In oral
rinse 39,23 % were unclassified, 5,36 % Rothia mucilaginosa, 3,65 %
Actinomyces odontolyticus, 2,96 % Gemella sanguinis and 2,73 % Strep-
tococcus salivarius. On the other hand, 43,76 % of species in tissue
samples were unclassified, 2,65 % Rothia mucilaginosa, 0,79 % Granuli-
catella adiacens, 0,54 % Porphyromonas pasteri and 0,36 % Prevotella
melaninogenica.
The data for the other classification levels are detailed in Table 2.
In addition, graphical representations have been made, considering
the five major components in each type of sample as well as their
respective percentages in the other samples, at the class (Supplemen-
tary Figure 3), order (Supplementary Figure 4), family (Supplemen-
tary Figure 5) and genus levels (Supplementary Figure 6).
3.4. OTUs analyses of all samples were adequate
First, bacterial species count revealed that half of the samples (9
out of 18) contained between 0 and 300 Operational Taxonomic Units
(OTUs). Three samples contained more than 1200 OTUs, 2 samples
had between 600 and 900 OTUs and only 1 sample had 900−1200
OTUs as shown in Supplementary Figure 7. Data analysis led us to
exclude 3 samples (S2, S3 and OR1) from the study because OTUs fea-
tures could not be identified in them (Supplementary Figure 8).
In order to evaluate sequencing depth, a rarefaction plot [23,25]
was built. Fig. 2 reflects that, although not all the types of specimens
Table 1
Socio-demographic and clinical cohort data.
Subjects S1 S2 S3 S4 S5 S6 TOTAL
Mean (§SD) %
Gender F M F F M F 66,67 (F), 33,33 (M)
Age 29 35 30 18 21 29 27(§6,29)
Marital status (single) X X X X X X 100
Occupation Student X X X 50
Unemployed X 16,67
Worker X X 33,33
Education Primary X X X X 66,67
College X X 33,33
Smoker X X 33,33
Numb. Cig/day 5 5 5 (§0)
Caries X X X X X X X 100
Periodontitis 0
implants 0
Prosthetic dental pieces 0
Active drug treatment X X 33,33
Type of drug treatment Allergy shots Anxiolytics; oral
contraceptive
Gastrointestinal (GI) Tract Medical Conditions 0
Systemic antibiotic use X X X X 66,67
Time from antibiotics (days) 383 252 1285 395 578,75 (§475,27)
Type of antibiotic Amoxicillin trihy-
drate, potassium
clavulanate
Amoxicillin Levofloxacin Amoxicillin
ARTICLE IN PRESS
JID: JORMAS [m5G;August 1, 2024;22:59]
M.E. Rodríguez-Fuentes, M. P
erez-Say
ans, G. Barbeito-Casti~
neiras et al. Journal of Stomatology oral and Maxillofacial Surgery 00 (2024) 101991
3

have the same number of reads, each one of them reached a plateau,
that indicates an adequate sequencing depth. Thus, it was observed
that from 30 reads on, the quality of the samples was comparable.
Fig. 3 shows the relative frequency of OTUs distribution on each
sample. Regarding tissue samples, DNA from 4 out of 6 samples could
not be assigned to any specific bacterial species and the remaining
two contained less than 10 % of DNA from known bacterial species.
Only 1 out of 4 saliva samples had 100 % relative frequency of known
bacterial species, 2 samples contained 50−60 % of known bacterial
species’DNA and 1 sample contained purely unassigned bacterial
DNA. Three out of 5 oral rinse samples were optimal (100 % known
OTUs), one contained >35 % and only 1 of the samples had 100 %
unassigned bacterial DNA.
3.5. Alpha diversity was higher in oral rinse samples
The complete phylogenetic analysis of all processed samples
resulted in the circular cladogram displayed in Fig. 4a Subsequently,
a Krona chart [26] was constructed for each type of specimen, aiming
to exhibit the differences of the microbiome composition among
them (Fig. 4b shows saliva analysis, Fig. 4c illustrates composition of
oral rinse samples, and Fig. 4d represents tissue sample data). The
information obtained manifests that about 44 % of bacteria in the
saliva samples belongs to unassigned species, whereas in oral rinse
samples this percentage is 70 %, reaching 99 % in tissue biopsy speci-
mens.
We next examined alpha diversity by using Simpson’s Index [27]
which showed a p-value=0,005. Thus, a statistically significant differ-
ence was found between the alpha diversity of oral rinse versus
unstimulated saliva and tissue biopsy (Fig. 5). Herein, bacterial alpha
diversity is higher in oral rinse samples.
4. Discussion
Despite the large number of studies published on the oral micro-
biome to date, there is still no clear evidence to determine the opti-
mal type of oral sample for an adequate analysis of alterations in its
composition, and this is where the novelty of this study lies.
Considering the results obtained, we ought to expect low percen-
tages of known bacterial species (<1 %) in oral mucosal tissue sam-
ples. Beforehand, this seems consistent with current knowledge,
since it is known that most of the germs of the oral flora are simply
Fig. 1. Quality control analysis of the sequencing reads. a. FastQC: Mean Quality Score. The analysis shows that almost all sequencing reads were between 30 (0,1 % reading error)
and 40 (0,01 % reading error) on the Phred Score Scale according to the number of base pairs. b. FastQC: Per Sequence Quality Scores. The analysis shows that almost all sequencing
reads were between 30 (0,1 % reading error) and 40 (0,01 %) on the Phred Score Scale according to the sequencing read counts.
ARTICLE IN PRESS
JID: JORMAS [m5G;August 1, 2024;22:59]
M.E. Rodríguez-Fuentes, M. P
erez-Say
ans, G. Barbeito-Casti~
neiras et al. Journal of Stomatology oral and Maxillofacial Surgery 00 (2024) 101991
4

found in the oral biofilm lining the mucous membranes, therefore the
rest of the sample tissue should presumably be bacteria free thanks to
the protective function of the oral epithelium [28]. Thus, most DNA
extracted from the sample should belong to the host and not to com-
mensal microorganisms [29].
On the other hand, liquid samples like saliva and oral rinse have
turned out to be quite different from each other. Saliva showed a
56 % relative abundance of identified species versus a 30 % found in
oral rinse. In addition to this finding, it has been spotted that oral
rinse statistically differed from saliva and tissue in terms of alpha
Table 2
Quantitative sequencing data from germs that constituted the five most abundant for each type of sample. Raw data obtained
prior to processing for statistical quality purposes.
Type of sample Saliva (Mean §SD) Oral rinse (Mean §SD) Tissue (Mean §SD)
Number of Species Identified 638,00 §559,99 618,17 §426,2 404,67 §125,42
Total reads 265,452 §507,833,57 209,772 §382,954,80 57,936,17§27,466,13
% of total Kingdom reads 99,95 §0,02 99,93 §0,13 87,64 §10,51
% of total Phylum reads 99,69 §0,1 99,65 §0,08 87,32 §10,4
% of total Class reads 98,41 §2,46 99,43 §0,12 87,10 §10,31
% of total Order reads 99,33 §0,15 99,33 §0,14 87,03 §10,33
% of total Family reads 98,99 §0,24 98,97 §0,19 86,34 §10,99
% of total Genus reads 98,24 §0,49 98,14 §0,56 85,22 §11,28
% of total Species reads 64,14 §4,24 60,77 §8,79 56,25 §20,27
% of Taxonomic_bacteria 99,95§0,02 99,91 §0,05 87,61 §10,49
% of Taxonomic unclassified 0,05 §0,02 0,09 §0,05 12,36 §10,51
% of Taxonomic Archaea 0,00 0,00 0,14
% of Phylum Firmicutes 55,57 §5,27 55,25 §12,33 34,35 §20,11
% of Phylum Bacteroidetes 18,25 §5,52 13,08 §6,13 10,97 §7,19
% of Phylum Fusobacteria 5,30 §2,63 5,91 §6,03 14,71 §8,55
% of Phylum Proteobacteria 4,79 §2,83 7,65 §5,22 8,84 §7,4
% of Phylum Actinobacteria 13,40 §5,48 15,51 §7,09 11,14 §4,68
% of Phylum Candidatus Saccharibacteria 1,40 §0,99 1,30 §0,75 3,75 §4,54
% of Phylum Absconditabacteria (SR1) 0,32 0,19 §0,21
% of Phylum Unclassified 0,31 §0,1 0,36 §0,08 12,68 §10,4
% of Phylum Spirochaetes 0,67 §0,73 0,78 §0,58 2,92 §3,56
% of Class unclassified 1,82 §1,35 2,09 §0,79 19,88 §10,81
% of Class Bacilli 43,95 §4,93 44,01 §11,95 26,83 §24,13
% of Class Bacteroidia 16,61 §5,48 11,19 §5,16 8,89 §6,13
% of Class Fusobacteria 5,30 §2,63 6,83 §6,25 14,71 §8,55
% of Class Betaproteobacteria 2,24 §1,05 4,46 §2,46 3,82 §3,88
% of Class Clostridia 4,49 §1,85 4,16 §2,14 7,42 §5,01
% of Class Actinobacteria 13,42 §5,5 15,51 §7,09 11,14 §4,68
% of Class Negativicutes 6,20 §4,63 6,83 §4,97 5,77 §3,38
% of Class Gammaproteobacteria 4,08 §1,48 3,58 §2,4 14,57
% of Order Lactobacillales 39,11 §3,78 37,91 §11,69 21,27 §24,36
% of Order Bacteroidales 16,61 §5,48 11,19 §5,16 8,21 §6,33
% of Order Actinomycetales 12,36 §5,75 14,22 §6,87 8,24 §5,79
% of Order Selnomonadales 6,20 §4,63 5,69 §4,97 4,81 §3,38
% of Order Fusobacteriales 5,30 §2,63 5,70 §6,25 14,71 §8,55
% of Order Bacillales 2,11 §0,32 2,19 §2,45
% of Order Clostridiales 3,89 §1,97 3,15 §2,41 6,15 §4,99
% of Family Streptococcaceae 33,19 §6,34 33,86 §11,37 19,08 §20,89
% of Family Prevotellaceae 11,43 §5,05 7,96 §3,82 2,13 §0,57
% of Family Micrococcaceae 6,15 §5,67 7,47 §5,83 14,17
% of Family Unclassified 14,85 §10,37
% of Family Carnobacteriaceae 5,58 §2,66 2,58 §1,1
% of Family Veillonellaceae 5,4 §4,6 5,69 §4,97 4,81 §3,38
% of Family Bacillales Incertae Sedis XI 3,74 §2,3 5,31 §3,82
% of Family Fusobacteriaceae 1,77 §0,3 2,81 §6,22 9,38 §5,91
% of Family Porphyromonadaceae 4,93 §2,06 2,34 §0,67 5,01 §3,04
% of Genus Streptococcus 33,16 §6,40 33,81 §11,36 15,19 §15,50
% of Genus Prevotella 10,67 §4,40 7,61 §3,71 2,04 §0,67
% of Genus Granulicatella 5,54 §2,65 2,55 §1,08
% of Genus Fusobacteria 1,76 §0,3 2,80 §6,2 9,36 §5,9
% of Genus Veillonella 4,86 §3,77 4,40 §4,25
% of Genus Actinomyces 4,39 §1,05 5,19 §1,96 2,14 §1,27
% of Genus Rothia 3,87 §5,57 6,04 §4,72 2,34
% of Genus Gemella 4,06 §2,57 6,01 §3,54
% of Genus Unclassified 14,78 §11,28
% of Genus Leptotrichia 4,3 §4,11
% of Species Unclassified 35,86 §4,24 39,23 §8,79 43,76 §20,27
% of Species Granulicatella adiacens 5,09 §2,58 2,28 §0,92 0,79 §1,26
%ofPrevotella melaninogenica 4,43 §2,05 2,22 §1,41 0,36
% of Species Porphyromonas pasteri 2,99 §0,57 1.07 §1,42 0,54
% of Species Actinomyces odontolyticus 3,42 §0,48 3,65 §1,88
% of Species Gemella sanguinis 2,35 §2,15 2,96 §3,33 0,12
% of Species Gemella haemolysans 2,15 §0,83 1,61 §0,57
% of Species Streptococcus salivarius 1,89 §5,61 2,73 §6,1
% Species Rothia mucilaginosa 5,81 §5,43 5,36 §4,51 2,65 §8,11
ARTICLE IN PRESS
JID: JORMAS [m5G;August 1, 2024;22:59]
M.E. Rodríguez-Fuentes, M. P
erez-Say
ans, G. Barbeito-Casti~
neiras et al. Journal of Stomatology oral and Maxillofacial Surgery 00 (2024) 101991
5

diversity, and therefore a higher microbial diversity was acknowl-
edged. One of the hypotheses generated about this topic consists of
the possibility that, the sample acquisition technique used for oral
rinse may drag bacteria inhabiting different anatomical locations of
the oral cavity. This phenomenon may not happen when obtaining
unstimulated saliva as it simply flows, without stimulation, from the
mouth into the sample collection container by gravity. In addition,
the higher adiversity found in oral rinse within our study, compared
to other specimens, also match previous investigations where oral
rinse diversity was superior to heterogeneity detected with swabs
collected from lesion and non-lesion sites [30]. Additionally, previous
research has concluded that sample handling and processing can sig-
nificantly affect the relative abundance of the identified microorgan-
isms [31].
With regard to oral microbiome composition and given that our
results showed superiority of saliva and oral rinse samples against
mucosal tissue specimens, we compared our raw data with previ-
ously published studies. The main bacterial phyla identified in saliva
samples were Firmicutes (55,57 %) similar to previous studies where
authors identified 50−60 % [31]. The amount of Bacteroidetes differed
in their relative abundance from our study (18,25 %) compared to
<10 % in previous studies [32]. Actinobacteria relative abundance was
slightly superior in our results (13,4 %) compared to a relative abun-
dance close to 10 % in the literature [32]. Fusobacteria relative abun-
dance was also comparable to previous results (5,30 %). Another
study from the field identified the top five abundant genera in saliva
samples to be Prevotella, Neisseria, Streptococcus, Haemophilus and
Rothia [15] while in our results Streptococcus, Prevotella, Granulica-
tella, Veillonela and Actinomyces lead the ranking. In the case of oral
rinse samples researchers found Streptococcus, Prevotella, Neisseria,
Haemophilus and Rothia to constitute the majority of the genera [15],
whereas we determined Streptococcus, Prevotella, Rothia, Gemella and
Actinomyces to be the most prevalent genera in oral rinse samples.
Depending on the type of study we are planning to carry out, one
sample is likely to be more appropriate than the other. In cases where
we care about having a larger number of identified OTUs we would
opt for a saliva sample, whereas if we are interested in analyzing
more diverse populations of the oral biofilm, oral rinse would be the
best choice. Knowing the best practice for oral bacteriome analysis
and deciding the optimal specimen is fundamental when designing
Fig. 2. Rarefaction curve plot. Simpsons Index analysis shows that from 30 reads on, the quality of the samples was comparable.
Fig. 3. Stacked bar plots represent the relative abundance at species level for each sample. Three samples were excluded due to DNA concentration insufficiency.
ARTICLE IN PRESS
JID: JORMAS [m5G;August 1, 2024;22:59]
M.E. Rodríguez-Fuentes, M. P
erez-Say
ans, G. Barbeito-Casti~
neiras et al. Journal of Stomatology oral and Maxillofacial Surgery 00 (2024) 101991
6

an oral microbiome research study [33]. Our results support the
importance of how selecting one sample type or another will reveal
very diverse information and so results and conclusions would also
substantially differ. Therefore, all studies should consider the rele-
vance of the type of specimen when designing an investigation proto-
col that aim to study changes in the oral microbiome composition
associated with local or systemic pathologies [16,34-36], as well as
undesired effects associated with treatments [37].
The characteristics of this study, inherent to its conception and
design, imply various limitations, therefore, the results should be
taken with caution. We are aware that 6 subjects are a very small
sample size. However, our aim is to expand this study in the near
Fig. 4. Phylogenetic analysis. a. Cladogram: Complete phylogenetic tree obtained from all the samples analyzed of the study population. b. Krona diagram of saliva samples at genus
level. 44 % of the genetic material could not be assigned to any specific taxa. c. Krona diagram of oral rinse samples at genus level. 70 % of the genetic material could not be assigned
to any specific taxa. d. Krona diagram of oral mucosal tissue samples at domain level. 99 % of the genetic material could not be assigned to any specific taxa.
ARTICLE IN PRESS
JID: JORMAS [m5G;August 1, 2024;22:59]
M.E. Rodríguez-Fuentes, M. P
erez-Say
ans, G. Barbeito-Casti~
neiras et al. Journal of Stomatology oral and Maxillofacial Surgery 00 (2024) 101991
7

future by recruiting a larger number of participants, so that more reli-
able conclusions can be drawn. This will also allow us to carry out a
statistical analysis of clinical and socio-demographic characteristics
of the patients that might be correlated with differences in the popu-
lations of microorganisms of the oral flora.
Having to discard 3 samples, for not being able to identify genetic
characteristics attributable to bacterial taxonomic units, makes us
think that some inaccuracy may have been occurred during samples’
processing. Despite the quality analysis performed in terms of the
reading depth, misleading in sample handling or treatment may have
resulted in the loss of genetic information from specimens that sub-
sequently had to be excluded from the study. Furthermore, these
errors may have been accentuated because of the limited experience,
given that these are unconventional techniques in a hospital environ-
ment due to the lack of clinical usefulness. Also, it is remarkable that
there is a high percentage of genetic material that could not be
assigned to any known bacterial species among the samples in which
genetic features were identified. This fits with the fact that in our
results we miss some of the main members of the oral microbiome
such as Streptococcus mitis species [38]orNeisseria and Haemophilus
genus [15]. This makes us wonder whether there are still technical
limitations that may be attributable to an unoptimized or debugged
protocol from the commercial companies, the specificity of the
reagents and primers [13,18,31], or either a general lack of knowl-
edge about the species that comprise the human oral microbiome
[7,36]. The knowledge about the oral microbiome is increasing by
leaps and bounds in recent years and, considering that our research
was conducted in 2020, we can anticipate some barriers. For instance,
recent investigations showed a superiority of V1-V2 sequencing in
oral samples versus our method that aimed the V3-V4 region [38].
Besides the limitations of this study, we should also state that the
specificity of the technology used in our study is higher than other
methods that were traditionally used for bacterial identification such
as culturing [7]. Nevertheless, the possibility that our observations
were biased due to methodological limitations cannot be excluded.
The need to determine the optimal specimen for the oral micro-
biome analysis, justifies this research. This pilot study provides a
basis for future research on the human oral microbiome. The differ-
ences found in the results obtained about microbiome on each sam-
ple type support the relevance of choosing an appropriate sample
prior to carrying out an investigation of the oral microbiome.
Depending on the objectives of the study, one sample should be
more appropriate than another as a source of microbial DNA. Thus,
oral rinse would be the best option in cases were a wider alpha
diversity is needed, whereas whole unstimulated saliva would be
more convenient in situations where a larger amount of bacterial
genetic material is required.
Ethics approval statement
The present research was approved by the Galician Ethics Com-
mittee of Clinical Research (Ref. No. 2020/552) and conducted in
accordance with the guidelines for Good Clinical Practice and the
Declaration of Helsinki. All participants signed the informed consent
form.
Financial interests
The authors declare they have no financial interests.
Funding
Molares-Vila. A. is supported by a fellowship co-funded by the State
Investigation Agency from the Spanish Ministry of Science and Innova-
tion (MCIN/AEI/10.13039/501100011033) through the procedure file
“PTA2021-019927-I”and by the European Social Fund+. No other
funds, grants or support was received for conducting this study.
Data availability
The data that support the findings of this study are not openly avail-
able due to reasons of sensitivity and are available from the corre-
sponding author upon reasonable request. Data are located in
controlled access data storage in the Microbiology Unit of the
Complexo Hospitalario Universitario de Santiago de Compostela.
Declaration of competing interest
None
CRediT authorship contribution statement
Manuel Eros Rodríguez-Fuentes: Conceptualization, Data cura-
tion, Investigation, Methodology, Writing −original draft, Writing −
review & editing. Mario P
erez-Say
ans: Conceptualization, Investiga-
tion, Methodology, Project administration, Resources, Writing −
Fig. 5. Alpha diversity analysis: Box plot representing the Simpson’s index for each type of sample. Box plots depict median (central horizontal line), 1st quartile and 3rd quartile,
and outliers. Statistically significant difference between the alpha diversity of oral rinse and saliva samples (p-value 0,005). Oral rinse showed a superior alpha diversity.
ARTICLE IN PRESS
JID: JORMAS [m5G;August 1, 2024;22:59]
M.E. Rodríguez-Fuentes, M. P
erez-Say
ans, G. Barbeito-Casti~
neiras et al. Journal of Stomatology oral and Maxillofacial Surgery 00 (2024) 101991
8

review & editing. Gema Barbeito-Casti~
neiras: Methodology, Valida-
tion, Writing −review & editing. Alberto Molares-Vila: Data cura-
tion, Formal analysis, Methodology, Software, Visualization, Funding
acquisition. Irene B. Prado-Pena: Conceptualization, Methodology.
Gisela C.V. Camolesi: Data curation, Investigation, Methodology.
Rafael L
opez-L
opez: Conceptualization, Funding acquisition, Resour-
ces, Supervision.
Supplementary materials
Supplementary material associated with this article can be found,
in the online version, at doi:10.1016/j.jormas.2024.101991.
References
[1] Ji L, Hao S, Wang J, Zou J, Wang Y. Roles of toll-like receptors in radiotherapy- and
chemotherapy-induced oral mucositis: a concise review. Front Cell Infect Micro-
biol 2022;12:831387 Published 2022 Jun 2. doi: 10.3389/fcimb.2022.831387.
[2] Zagury-Orly I, Khaouam N, Noujaim J, Desrosiers MY, Maniakas A. The effect of
radiation and chemoradiation therapy on the head and neck mucosal micro-
biome: a Review. Front Oncol 2021;11:784457 Published 2021 Dec 2. doi:
10.3389/fonc.2021.784457.
[3] Rodríguez-Fuentes ME, P
erez-Say
ans M, Carreras-Presas CM, Marichalar-Mendia
X, Bag
an-Deb
on L, L
opez-L
opez R. Prevalence of acute oral mucosal damage sec-
ondary to the use of systemic antineoplastics: a systematic review and meta-
analysis. Oral Surg Oral Med Oral Pathol Oral Radiol 2023;135(3):385–95. doi:
10.1016/j.oooo.2022.11.016.
[4] Shetty SS, Maruthi M, Dhara V, et al. Oral mucositis: current knowledge and
future directions. Dis Mon 2022;68(5):101300. doi: 10.1016/j.
disamonth.2021.101300.
[5] Girija ASS, Ganesh PS. Functional biomes beyond the bacteriome in the oral eco-
system. Jpn Dent Sci Rev 2022;58:217–26. doi: 10.1016/j.jdsr.2022.05.002.
[6] Elias-Oliveira J, Leite JA, Pereira 
IS, et al. NLR and intestinal dysbiosis-associated
inflammatory illness: drivers or dampers? Front Immunol 2020;11:1810. Pub-
lished 2020 Aug 11. doi: 10.3389/fimmu.2020.01810.
[7] Zhang JS, Chu CH, Yu OY. Oral Microbiome and Dental Caries Development. Dent J
(Basel) 2022;10(10):184. Published 2022 Sep 30. doi: 10.3390/dj10100184.
[8] Hong BY, Sobue T, Choquette L, et al. Chemotherapy-induced oral mucositis is
associated with detrimental bacterial dysbiosis. Microbiome 2019;7(1):66. Pub-
lished 2019 Apr 25. doi: 10.1186/s40168-019-0679-5.
[9] Lamont RJ, Koo H, Hajishengallis G. The oral microbiota: dynamic communities
and host interactions. Nat Rev Microbiol 2018;16(12):745–59. doi: 10.1038/
s41579-018-0089-x.
[10] Mougeot JC, Stevens CB, Morton DS, Brennan MT, Mougeot FB. Oral Microbiome
and Cancer Therapy-Induced Oral Mucositis. J Natl Cancer Inst Monogr
2019;2019(53) lgz002. doi: 10.1093/jncimonographs/lgz002.
[11] Saito S, Aoki Y, Tamahara T, et al. Oral microbiome analysis in prospective genome
cohort studies of the Tohoku Medical Megabank Project. Front Cell Infect Micro-
biol 2021;10:604596 Published 2021 Jan 29. doi: 10.3389/fcimb.2020.604596.
[12] Verma D, Garg PK. Dubey AK. Insights into the human oral microbiome. Arch
Microbiol 2018;200(4):525–40. doi: 10.1007/s00203-018-1505-3.
[13] Knight R, Vrbanac A, Taylor BC, et al. Best practices for analysing microbiomes.
Nat Rev Microbiol 2018;16(7):410–22. doi: 10.1038/s41579-018-0029-9.
[14] Quince C, Walker AW, Simpson JT, Loman NJ, Segata N. Shotgun metagenomics,
from sampling to analysis [published correction appears. Nat Biotechnol 2017;35
(9):833–44. doi: 10.1038/nbt.3935.
[15] Caselli E, Fabbri C, D’Accolti M, et al. Defining the oral microbiome by whole-
genome sequencing and resistome analysis: the complexity of the healthy pic-
ture. BMC Microbiol 2020;20(1):120. Published 2020 May 18. doi: 10.1186/
s12866-020-01801-y.
[16] Zhu XX, Yang XJ, Chao YL, et al. The potential effect of oral microbiota in the pre-
diction of mucositis during radiotherapy for nasopharyngeal carcinoma. EBioMe-
dicine 2017;18:23–31. doi: 10.1016/j.ebiom.2017.02.002.
[17] McGaughey KD, Yilmaz-Swenson T, Elsayed NM, et al. Comparative evaluation of
a new magnetic bead-based DNA extraction method from fecal samples for
downstream next-generation 16S rRNA gene sequencing [published correction
appears in PLoS One. 2019 Feb 19;14(2):e0212712]. PLoS ONE 2018;13(8):
e0202858 Published 2018 Aug 23. doi: 10.1371/journal.pone.0202858.
[18] Kim YT, Jeong J, Mun S, Yun K, Han K, Jeong SN. Comparison of the oral microbial
composition between healthy individuals and periodontitis patients in different
oral sampling sites using 16S metagenome profiling. J Periodontal Implant Sci
2022;52(5):394–410. doi: 10.5051/jpis.2200680034.
[19] Bolyen E, Rideout JR, Dillon MR, et al. Reproducible, interactive, scalable and
extensible microbiome data science using QIIME 2. Nat Biotechnol 2019;37:852–
7. doi: 10.1038/s41587-019-0209-9.
[20] Andrews S. (2010). FastQC: a quality control tool for high throughput sequence
data. Available online at: https://www.bioinformatics.babraham.ac.uk/projects/
fastqc
[21] Chen S, Zhou Y, Chen Y, Gu J. fastp: an ultra-fast all-in-one FASTQ preprocessor.
Bioinformatics 2018;34(17):i884–90. doi: 10.1093/bioinformatics/bty560.
[22] Martin M. Cutadapt removes adapter sequences from high-throughput sequenc-
ing reads. EMBnet.journal 2011;17(1):10–2. doi: 10.14806/ej.17.1.200.
[23] Weiss S, Xu ZZ, Peddada S, Amir A, Bittinger K, Gonzalez A, Lozupone C, Zaneveld
JR, V
azquez-Baeza Y, Birmingham A, Hyde ER, Knight R. Normalization and micro-
bial differential abundance strategies depend upon data characteristics. Micro-
biome 2017;5(1):27. doi: 10.1186/s40168-017-0237-y.
[24] Ewels P, Magnusson M, Lundin S, K€
aller M. MultiQC: summarize analysis results
for multiple tools and samples in a single report. Bioinformatics 2016;32
(19):3047–8. doi: 10.1093/bioinformatics/btw354.
[25] Ondov BD, Bergman NH, Phillippy AM. Interactive metagenomic visualization in a
Web browser. BMC Bioinformatics 2011;12:385. Published 2011 Sep 30. doi:
10.1186/1471-2105-12-385.
[26] Ramette A. Multivariate analyses in microbial ecology. FEMS Microbiol Ecol
2007;62(2):142–60. doi: 10.1111/j.1574-6941.2007.00375.x.
[27] Brizuela M, Winters R. Histology, oral mucosa. Treasure Island (FL): StatPearls;
2023 https://www.ncbi.nlm.nih.gov/books/NBK572115/ Accessed 8 May 2023.
[28] Bruggeling CE, Garza DR, Achouiti S, Mes W, Dutilh BE, Boleij A. Optimized bacte-
rial DNA isolation method for microbiome analysis of human tissues. Microbiolo-
gyopen 2021;10(3):e1191. doi: 10.1002/mbo3.1191.
[29] Shitozawa Y, Haro K, Ogawa M, Miyawaki A, Saito M, Fukuda K. Differences in the
microbiota of oral rinse, lesion, and normal site samples from patients with
mucosal abnormalities on the tongue. Sci Rep 2022;12(1):16839. Published 2022
Oct 7. doi: 10.1038/s41598-022-21031-8.
[30] Liu F, Lu H, Dong B, et al. Systematic evaluation of the viable microbiome in the
human oral and gut samples with spike-in gram+/- bacteria. mSystems 2023;8
(2):e0073822. doi: 10.1128/msystems.00738-22.
[31] Nearing JT, DeClercq V, Van Limbergen J, Langille MGI. Assessing the variation
within the oral microbiome of healthy adults. mSphere 2020;5(5):e00451.
-20Published 2020 Sep 30. doi: 10.1128/mSphere.00451-20.
[32] Gopinath D, Menon RK, Wie CC, et al. Differences in the bacteriome of swab,
saliva, and tissue biopsies in oral cancer. Sci Rep 2021;11(1):1181. Published
2021 Jan 13. doi: 10.1038/s41598-020-80859-0.
[33] D’Souza G, Clemens G, Troy T, et al. Evaluating the utility and prevalence of HPV
biomarkers in oral rinses and serology for HPV-related oropharyngeal cancer.
Cancer Prev Res (Phila) 2019;12(10):689–700. doi: 10.1158/1940-6207.CAPR-19-
0185.
[34] Soffritti I, D’Accolti M, Fabbri C, et al. Oral microbiome dysbiosis is associated with
symptoms severity and local immune/inflammatory response in COVID-19
patients: a cross-sectional study. Front Microbiol 2021;12:687513 Published
2021 Jun 23. doi: 10.3389/fmicb.2021.687513.
[35] Wright M.L., Podnar J., Longoria K.D., et al. Comparison of commercial DNA
extraction kits for whole metagenome sequencing of human oral, vaginal, and
rectal microbiome samples. Preprint. bioRxiv. 2023; 2023.02.01.526597. Pub-
lished 2023 May 4. doi:10.1101/2023.02.01.526597
[36] Rodríguez-Fuentes ME, P
erez-Say
ans M, Chauca-Baja~
na LA, Barbeito-Casti~
neiras
G, Molino-Bernal ML, L
opez-L
opez R. Oral microbiome and systemic antineoplas-
tics in cancer treatment: a systematic review. Med Oral Patol Oral Cir Bucal
2022;27(3):e248–56 Published 2022 May 1. doi: 10.4317/medoral.25121.
[37] Okahashi N, Nakata M, Kuwata H, Kawabata S. Oral mitis group streptococci: a
silent majority in our oral cavity. Microbiol Immunol 2022;66(12):539–51. doi:
10.1111/1348-0421.13028.
[38] Na HS, Song Y, Yu Y, Chung J. Comparative analysis of primers used for 16S rRNA
gene sequencing in oral microbiome studies. Methods Protoc 2023;6(4):71. Pub-
lished 2023 Aug 6. doi: 10.3390/mps6040071.
ARTICLE IN PRESS
JID: JORMAS [m5G;August 1, 2024;22:59]
M.E. Rodríguez-Fuentes, M. P
erez-Say
ans, G. Barbeito-Casti~
neiras et al. Journal of Stomatology oral and Maxillofacial Surgery 00 (2024) 101991
9