Archives of Oral Biology 73 (2017) 79–87

Changes in the oral ecosystem induced by the use of 8% argininetoothpaste

Jessica E. Koopmana, Michel A. Hoogenkampa, Mark J. Buijsa, Bernd W. Brandta,Bart J.F. Keijserb, Wim Crielaarda, Jacob M. ten Catea, Egija Zauraa,*aDepartment of Preventive Dentistry, Academic Centre for Dentistry Amsterdam, University of Amsterdam and VU University Amsterdam, Gustav Mahlerlaan3004, 1081 LA Amsterdam, The NetherlandsbResearch Group Microbiology and Systems Biology, TNO Earth, Life and Social Sciences, Utrechtseweg 48, 3704 HE Zeist, The Netherlands

A R T I C L E I N F O

Article history:Received 6 November 2015Received in revised form 2 September 2016Accepted 24 September 2016

Keywords:ArgininePrebioticOral ecologyMicrobiome

A B S T R A C T

Objective: Bacterial metabolism of arginine in the oral cavity has a pH-raising and thus, potential anti-caries effect. However, the influence of arginine on the oral microbial ecosystem remains largelyunresolved.Design: In this pilot study, nine healthy individuals used toothpaste containing 8% arginine for eightweeks. Saliva was collected to determine arginolytic potential and sucrose metabolic activity at theBaseline, Week 4, Week 8 and after a two weeks Wash-out period. To follow the effects on microbialecology, 16S rDNA sequencing on saliva and plaque samples at Baseline and Week 8 and metagenomesequencing on selected saliva samples of the same time-points was performed.Results: During the study period, the arginolytic potential of saliva increased, while the sucrosemetabolism in saliva decreased. These effects were reversed during the Wash-out period. Although a fewoperational taxonomic units (OTUs) in plaque changed in abundance during the study period, there wasno real shift in the plaque microbiome. In the saliva microbiome there was a significant compositionalshift, specifically the genus Veillonella had increased significantly in abundance at Week 8.Conclusion: Indeed, the presence of arginine in toothpaste affects the arginolytic capacity of saliva andreduces its sucrose metabolic activity. Additionally, it leads to a shift in the salivary microbiomecomposition towards a healthy ecology from a caries point of view. Therefore, arginine can be regarded asa genuine oral prebiotic.

ã 2016 Elsevier Ltd. All rights reserved.

Contents lists available at ScienceDirect

Archives of Oral Biology

journa l homepage: www.e lsev ier .com/ locate /aob

1. Introduction

Caries is an old and still common problem among humans(Grine, Gwinnett, & Oaks, 1990; Listl, Galloway, Mossey, &Marcenes, 2015; WHO, 2012). Severe caries can lead to excruciat-ing pain and tooth loss, which in turn can lead to an inability tochew or even eat. Logically, treatment and prevention of carieshave been a priority of the dental community for quite somedecades. Therefore it is now recognized that the lowered pH, as aresult of the acids formed by bacterial fermentation of sugars, leadsto demineralization of the enamel (Fejerskov, 1997; Peterson,Snesrud, Schork, & Bretz, 2011; ten Cate, 2013). In contrast, the

* Corresponding author at: Gustav Mahlerlaan 3004, 1081 LA Amsterdam, TheNetherlands.

E-mail address: e.zaura@acta.nl (E. Zaura).

http://dx.doi.org/10.1016/j.archoralbio.2016.09.0080003-9969/ã 2016 Elsevier Ltd. All rights reserved.

bacterial metabolism of arginine has a pH-raising effect (Huang,Exterkate, & ten Cate, 2012).

The human oral cavity is an ecosystem, and like all ecosystems,the oral ecosystem is a combination of many different components(e.g. bacteria, fungi, metabolic compounds, host cells, salivaryconstituents) and in one way or another, this system retains abalance. The introduction of a specific compound might shift thebalance of an ecosystem towards a different microbiologicalcomposition and functional potential. For instance, the frequentintake of sucrose facilitates a favorable environment for fermen-tative bacteria, leading to a cariogenic environment (Bradshaw &Lynch, 2013; Raner et al., 2014). In contrast, the bacterialmetabolism of arginine elevates the pH in the oral cavity. ThepH-raising effect of arginine is, amongst others, facilitated by thebacterial arginine deiminase system (ADS). This ADS metabolizesarginine, forming ammonia in the process and the subsequentprotonation of ammonia into ammonium causes the pH to rise

80 J.E. Koopman et al. / Archives of Oral Biology 73 (2017) 79–87

(Burne & Marquis, 2000; Casiano-Colon & Marquis, 1988; Marquis,Bender, Murray, & Wong, 1987).

Compounds that have a positive effect on human healththrough the selective stimulation of growth or activity of specificmicrobes are called prebiotics (Mugambi, Young, & Blaauw, 2014).Some research has been done on the effect of prebiotics on theecosystems of the human gastro-intestinal tract and skin (Nole,Yim, & Keri, 2014; Orel & Kamhi Trop, 2014; Sanders et al., 2014),while knowledge on the use of prebiotics in the oral cavity is verylimited.

Arginine naturally occurs in a variety of foods and has recentlybeen added to oral care products, for its pH-raising effect and anti-caries potential (Burne & Marquis, 2000; Kraivaphan et al., 2013;Nascimento et al., 2013; Sharif, Iram, & Brunton, 2013; Wu et al.,2009). Yet, how arginine affects the total oral ecosystem remainslargely unknown.

Therefore, the aim was to enhance knowledge on the effect ofarginine on the human oral ecosystem and to shed light on thequestion if this potential prebiotic is active on a functional as wellas compositional level.

Fig. 1. Sampling scheme of the study. The n represents the number of subjects contributaken on two consecutive days. The data from these samples were averaged for subsequeconsecutive days was used for further analysis. Saliva samples for the biochemical as(Baseline), four weeks (Week 4) and eight weeks (Week 8) after the Baseline, and after a tThe saliva and plaque samples to be used for DNA analysis were taken at the Baseline

2. Materials & methods

2.1. Pilot study design and sampling

Nine healthy volunteers with no overt caries, who had givenwritten informed consent, participated in this pilot study. Thestudy was approved by the review board of the local MedicalEthical Committee (VU University Medical Center, referencenumber 2010/147).

In the two weeks prior to the first of a total of four samplingmoments (Fig. 1), the participants used a control toothpaste(1.45 mg g�1

fluoride, Prodent, Sara Lee Household & Bodycare,Exton, PA, USA). Sampling moments consisted of two consecutivedays. Samples were taken on each day, and the data of theseduplicate samples were averaged. After the first samples weretaken (Baseline), the subjects were instructed to brush their teethusing toothpaste containing 8% arginine (1.45 mg g�1

fluoride,Colgate-Palmolive, New York, NY, USA). Subsequent samples weretaken four weeks (Week 4) and eight weeks (Week 8) after theBaseline. After Week 8, use of the 8% arginine toothpaste wasstopped and the control toothpaste was reintroduced. The last

ting to the samples from different visits. Duplicate indicates that the samples werent statistical analysis. Single indicates that only one of the samples taken at the twosays were taken before the participants started using the 8% arginine toothpastewo week wash-out period that followed the finish of the 8% arginine toothpaste use.

visit and Week 8.

J.E. Koopman et al. / Archives of Oral Biology 73 (2017) 79–87 81

samples, used for the biochemical assays, were taken two weeksafter discontinuation of the 8% arginine toothpaste (Wash-out). Onsampling days, the subjects were asked to refrain from oral hygienein the morning and not to eat and drink 2 h prior to sampling, toobtain overnight plaque.

At each sampling moment both, stimulated saliva samples andsupragingival plaque samples were obtained. Stimulated salivawas collected by chewing on gum base (Wrigley, Chicago, Il, USA)for 10 min while expectorating the saliva into a sterile container.During expectoration the tube was kept on ice until furtherprocessing. On arrival to the lab, the saliva was vortexed vigorouslyand distributed into 2 ml portions. Supragingival plaque wascollected, by a dentist, from the buccal surface of the upper firstmolar using a sterile microbrush (Microbrush International,Grafton, WI, USA). The tip of the microbrush was placed inRNAProtect solution (Qiagen, Hilden, Germany). All samples werestored at �80 �C.

2.2. Biochemical assays

To determine either the arginolytic potential or the saccha-rolytic activity of the salivary pellet, each stimulated saliva samplewas washed and concentrated five times in either 10 mM tris-maleate buffer or buffered peptone water, respectively. Theconcentrate was pipetted, in duplicate, into the wells of a 96-wells plate and the remaining cell suspension was stored at �20 �Cfor subsequent protein determination (Bradford, 1976).

To determine the arginolytic potential, an arginine metabolismassay was performed, described in detail elsewhere (Hoogenkamp& ten Cate, 2014). In short, arginine was added to the concentratedsamples to a final concentration of 50 mM and incubated for 3 h at37� C. Aliquots were taken immediately after addition of thearginine and after 3 h. All samples were heat inactivated,centrifuged and the supernatants were stored at -20� C untilfurther analysis. The ammonium produced was analyzed spectro-photometrically, using a UV-method, wherein the formed ammo-nium reacts in the presence of 2-oxoglutarate, glutamatedehydrogenase and NADH into NAD+ (da Fonseca-Wollheim,Bergmeyer, & Gutmann, 1974).

The saccharolytic potential was determined in a similar way byadding sucrose to the concentrated samples to a final concentra-tion of 0.5% (w/v). Samples were taken and processed similar to thearginolytic potential assay and the lactate produced was deter-mined by another UV-method. This method was based on thereduction of NAD+ into NADH, in the presence of the lactic acidproduced in the assay and lactate dehydrogenase (Gutmann &Wahlefeld, 1974).

2.3. 16S rRNA gene amplicon sequencing

The saliva and plaque samples, taken in duplicate at theBaseline and Week 8, were used for amplicon sequencing. The DNAwas isolated as described previously (Zaura, Keijser, Huse, &Crielaard, 2009). An amplicon library based on the partial 16S rRNAgene was prepared using degenerate primers annealing to regionsV5–V7. Sequencing was performed using 454 GS-FLX Titaniumchemistry (Roche, Branford, CT, USA). The sequence data have beensubmitted to NCBI’s SRA database under accession numberPRJNA288541.

2.4. Amplicon sequencing data analysis

Quantitative Insights Into Microbial Ecology (QIIME) version1.5.0 (Caporaso et al., 2010) was used to filter the raw amplicondata. Reads with a length <200 or >1000 bp, a mean quality score<30, a homopolymer sequence >6 nt, N > 0, >1 barcode error, >1

error in the forward or >2 errors in the reverse primer wereremoved from the dataset. The reads were clustered intooperational taxonomic units (OTUs) (Kraneveld et al., 2012). Thesamples were randomly subsampled at 2060 reads per samplebefore further analysis.

2.5. Shotgun sequencing of saliva samples

Baseline and Week 8 saliva samples from four subjects who hadthe highest arginolytic potential after eight weeks of using thearginine containing toothpaste were selected for metagenomicsshotgun sequencing. These were the same samples that were usedfor amplicon sequencing. The DNA to be used for shotgunsequencing was isolated as described previously (Koopmanet al., 2015). The shotgun sequencing was performed using 454GS-FLX Titanium chemistry. The sequence data have beensubmitted to the Genbank database under accession numberPRJNA288541.

2.6. Shotgun sequencing data analysis

Reads were quality filtered using PRINSEQ (Schmieder &Edwards, 2011); sequences with a minimal length of <70 bp,mean quality score <20, percentage of ambiguous bases >10 or anentropy value <70, were removed from the dataset, as were exactduplicates and (reverse complement) 50 and 30 end duplicates.Sequences of human origin were removed using BMTagger(Rotmistrovsky & Agarwala, 2011). HUMAnN v0.99b (Abubuckeret al., 2012) was used to generate pathway summaries accordingto the KEGG Orthology (Kanehisa & Goto, 2000; Kanehisa et al.,2014).

2.7. Statistical analysis

The data from the duplicate samples were averaged and usedfor further statistical analysis. The non-parametric Friedmantest was performed to determine if there were significantdifferences in either the arginolytic potential or the saccha-rolytic activity between any of the four visits (Baseline, Week 4,Week 8 and Wash-out). Subsequently, the Wilcoxon SignedRanks test was performed to determine if there were significantdifferences between the time-points in a pairwise manner. Bothtests were performed using SPSS v21 (IBM Corp, Armonk, NY,USA). False Discovery Rate (FDR), according to Benjamini andHochberg in R code, was used to correct the P-values formultiple comparisons. The statistical significance for the testswas set at P < 0.05.

To visualize similarity of the microbial profiles based on the 16SrDNA of the plaque and saliva samples at the two different visits(Baseline and Week 8), non-metric multidimensional scaling(nmMDS) plots, based on the Bray-Curtis coefficient (stress <0.2),were made in PAST v 3.0 (Hammer, Harper, & Ryan, 2001). The one-way permutational multivariate analysis of variance (PERMA-NOVA) was used to test the difference between the Baseline andWeek 8 visit per sample type in PAST. The Shannon Diversity Indexwas calculated using PAST. The Wilcoxon Signed Ranks test in SPSSwas used to determine if there was a significant difference inShannon Diversity Index between visits. Linear DiscriminantAnalysis Effect Size (LEfSe) was used to determine which taxaand which functions were differentially abundant between theBaseline and Week 8 (Segata et al., 2011). The default settings wereused (factorial Kruskal-Wallis test among classes: a = 0.05,threshold on the logarithmic LDA score for discriminative features:2.0 and the strategy for multi-class analysis was set to be all-against-all).

Fig. 3. Sucrose metabolism in saliva. Sucrose metabolism was measured as theamount of lactate produced by the salivary pellet. Friedman test for K-relatedsamples: P = 0.01. Statistical significance (P < 0.05) was determined using theWilcoxon Signed Ranks test, corrected for multiple comparisons using FalseDiscovery Rate (FDR). The boxes represent the median and interquartile range (IQR),outliers more than 1.5 � IQR are depicted by �, and more than 3 � IQR by $. n = 9.

82 J.E. Koopman et al. / Archives of Oral Biology 73 (2017) 79–87

3. Results

3.1. The effect of 8% arginine toothpaste on the arginolytic potentialand saccharolytic activity of saliva

The duplicate saliva samples taken at all four time-points(Baseline, Week 4, Week 8 and Wash-out) from all nineparticipants were used for the biochemical assays.

In the arginolytic assay, the ammonium produced wasrepresentative for the arginolytic capacity of the salivary micro-biome. The arginolytic potential increased significantly from theBaseline to the visit at Week 4 and again from Week 4 to Week 8(Fig. 2). On the other hand, the arginolytic potential of the salivarypellet decreased significantly from Week 8 to the end of the Wash-out period (Week 10). In the sucrose metabolism assay, the lactateproduced indicated the sucrose metabolic potential of the salivarymicrobiome. The saccharolytic activity decreased significantlyfrom the Baseline to Week 8, while it increased again significantlyat the Wash-out visit (Fig. 3).

3.2. The effect of 8% arginine toothpaste on the plaque and salivarymicrobial composition

Duplicate plaque and saliva samples from all nine subjects,taken at the Baseline and after eight weeks of using the argininecontaining toothpaste (Week 8), were used for 16S rDNA ampliconsequencing.

Of the 1,010,385 raw input sequences, 451,514 with an averageof 6271 reads per sample (SD 2440, range 2063–13089) remainedafter filtering and the removal of chimeric sequences. Thesubsampling threshold was set at 2060 reads. The remainingsubset of the plaque samples consisted of 362 OTUs with anaverage of 101 OTUs per sample (SD 19, range 58–136), while thesaliva samples comprised a total of 415 OTUs with an average of103 OTUs per sample (SD 19, range 67–148).

Fig. 2. Arginolytic potential of saliva. The arginolytic potential of saliva wasmeasured as the amount of ammonia produced by the salivary pellet. Friedman testfor K-related samples: P = 0.0005. Statistical significance (P < 0.05) was determinedusing the Wilcoxon Signed Ranks test, corrected for multiple comparisons by FalseDiscovery Rate (FDR). The boxes represent the median and interquartile range (IQR),outliers more than 1.5 � IQR are depicted by �, and more than 3 � IQR by $. n = 9.

The non-metric multidimensional scaling (nmMDS) plot didnot reveal a strong divide in the composition of the plaquemicrobiome at the Baseline or at Week 8 (Fig. 4), in contrast to thenmMDS plot based on the salivary microbiome (Fig. 5). One-wayPERMANOVA indicated a significant difference in the saliva derivedmicrobial profiles between the two visits (P = 0.007, F = 2.76),contrary to the plaque derived microbial profile (P = 0.692,F = 0.68). There was no significant change in Shannon diversityobserved for both sample types after the use of the argininecontaining toothpaste (Supplementary Figs. 1 and 2).

Fig. 4. Non-metric multidimensional scaling plot based on 16S rDNA derived fromplaque, calculated using the three-dimensional Bray-Curtis similarity index. Stress:0.0.118. One-way PERMANOVA: P = 0.692, F = 0.68. Baseline: �, week 8: *. n = 9.

Fig. 5. Non-metric multidimensional scaling plot based on 16S rDNA derived fromsaliva, calculated using the three-dimensional Bray-Curtis similarity index. Stress:0.121. One-way PERMANOVA: P = 0.007, F = 2.76. Baseline: �, week 8: *. n = 9.

J.E. Koopman et al. / Archives of Oral Biology 73 (2017) 79–87 83

Descending from the total microbial profile to the separate taxa,Streptococcus was the most abundant genus in the plaquemicrobiome (Fig. 6) as well as in the salivary microbiome(Fig. 7). Their level of abundance remained similar between bothvisits. The relative proportions of Veillonella and Fusobacteriumwere significantly different between the Baseline and Week 8 insaliva (Fig. 7 and Supplementary Fig. 3).

To determine which OTUs were significantly different inabundance between the two visits, linear discriminant analysiseffect size (LEfSe) (Segata et al., 2011) was used. In the plaquesamples, only Candidate division TM7 was differentially abundantbetween the two visits and was associated with the Baseline (LDAscore 2.67). The overall abundance of Candidate division TM7 was

Fig. 6. Relative abundance of the predominant genera in the plaque microbiome. The brespectively. n = 9.

<1% (Baseline: 0.049%. Week 8: 0.003%) (Fig. 6). At the OTU level,seven OTUs were differentially abundant in plaque between thetwo visits. Both OTU184 (Candidate division TM7) and OTU148(Dialister) were associated with the Baseline, while OTU516(Treponema), OTU102 and OTU530 (both Prevotella) and OTU150and OTU 423 (both Eubacterium) were associated with Week 8(Fig. 8).

In the salivary microbiome, twelve genera were differentiallyabundant between the two visits, with Veillonella being the onlygenus associated with Week 8 (Supplementary Fig. 4). At the OTUlevel, 47 OTUs were differentially abundant between the two visits(Fig. 9). The OTU429 (Porphyromonas) had the strongest associa-tion with the Baseline. Of the seven OTUs that were associated withWeek 8, the association of OTU472 (Veillonella) was the strongest.

3.3. The effect of 8% arginine toothpaste on the functional potential ofthe salivary microbiome

Shotgun sequencing was performed on Baseline and Week 8saliva samples of four of the subjects who had the highestarginolytic potential after using the arginine containing toothpastefor eight weeks.

After filtering the 1,408,122 shotgun reads using PRINSEQ(Schmieder & Edwards, 2011), 1,181,472 reads with an averagelength of 323 bp remained. The average number of reads persample was 147684 (SD 46030, range 81,199–218,604).

The base excision repair (ko03410) and the lysosome (ko04142)KEGG pathways were the most abundant pathways in the Baselinesamples compared to the samples taken at Week 8 (SupplementaryFig. 5). A total of six pathways were more abundant in the samplestaken at Week 8 compared to the samples taken at the Baselinevisit. Of these six pathways, the carbon fixation pathway inprokaryotes (ko00720) explained the greatest differences betweenthe two time-points, followed by the carbon fixation in photosyn-thetic organisms (ko00710) pathway, the pyrimidine metabolism(ko00240) pathway, the spliceosome (ko03040) pathway, thebenzoate degradation (ko00362) pathway and the tropane,piperidine and pyridine alkaloid biosynthesis (ko00960) pathway.

ars and error bars represent the mean relative abundance and standard deviation,

Fig. 7. Relative abundance of the predominant genera in the salivary microbiome. The bars and error bars represent the mean relative abundance and standard deviation,respectively. The * indicates that these genera were significantly differentially abundant between the two visits according to LEfSe. n = 9.

Fig. 8. Relative abundance of differentially abundant OTUs in plaque. (A) The relative abundance of the OTUs that were found to be significantly differentially abundant byLEfSe between the two visits, displayed per individual per visit. (B) The histogram of the differentially abundant OTUs between the two visits, identified through lineardiscriminant analysis effect size score. The white bars represent the OTUs that are associated with the Baseline and the black bars represent the OTUs that are associated withWeek 8. n = 9.

84 J.E. Koopman et al. / Archives of Oral Biology 73 (2017) 79–87

4. Discussion

The findings of this pilot study indicate that using toothpastethat contains 8% arginine, affects the composition of the oralmicrobiome. In addition, it enhances the arginolytic potential ofthe salivary pellet, while the sucrose metabolic activity isrepressed.

Indeed, we found the capacity of the salivary pellet tometabolize arginine to have increased during the time the subjectsused toothpaste containing 8% arginine. However, this activity wasreduced after the subjects stopped using the toothpaste. From ourresults it is clear that the capacity to metabolize arginine is presentin the oral cavity and is adjusted to the amount of substratepresent. In contrast to the arginolytic capacity, the capacity toconvert sucrose into lactate became lower during the study, yetbecame higher again after the use of arginine containingtoothpaste was discontinued.

During the course of this study, the abundance of the genusVeillonella increased. Veillonella species are lactate fermenters(Rogosa, 1964, 1984) and might have used the lactate formed in the

sucrose metabolism assays. That would mean that the capacity toferment lactate increased, instead of a decrease in the capacity toconvert sucrose into lactate. However, since we lack themicrobiological data of the last time-point, we can only speculate.

Nonetheless, members of this genus have an optimum growthpH of 6.5–8 (Rogosa, 1964, 1984). Hence, a raised pH of theecosystem creates a favorable environment for these commensals.Interestingly, Veillonella was the only genus that was differentiallyabundant in saliva at Week 8, in contrast to the Baseline visit whereeleven genera were differentially abundant.

Moreover, the genera that showed differential abundancebetween the two visits were different in the saliva and plaquederived-samples. Candidate division TM7 as well as OTU184(Candidate division TM7) were both more abundant at the Baselinein plaque. Until now, this still quite elusive member of the oralecosystem has been related to periodontitis and gingivitis (Duran-Pinedo et al., 2014; Huang et al., 2011). In contrast to Candidatedivision TM7, which seemingly did not favor the conditionsintroduced by arginine metabolism, were the OTUs identified asmembers of the genera Eubacterium, Prevotella and Treponema,which became more abundant in plaque at Week 8. Likely, these

Fig. 9. Relative abundance of differentially abundant OTUs in saliva. (A) The relative abundance of the OTUs that were found to be significantly differentially abundant byLEfSe between the two visits, displayed per individual per visit. (B) The histogram of the differentially abundant OTUs between the two visits, identified through lineardiscriminant analysis effect size score. The white bars represent the OTUs that are associated with the Baseline and the black bars represent the OTUs that are associated withWeek 8. n = 9.

J.E. Koopman et al. / Archives of Oral Biology 73 (2017) 79–87 85

OTUs represent species that favor the conditions brought about bythe arginine metabolism. Some members of these genera areassociated with an elevated pH, as well as periodontal disease(Downes et al., 2001; Marsh, 1994; Nadkarni et al., 2012; Norris,

Paster, Moter, & Göbel, 2006; Wade, 2006; Zhu et al., 2013).However, Li et al. did not find an association between the use of 8%arginine containing toothpaste and an increase in gingivitis (Liet al., 2012).

86 J.E. Koopman et al. / Archives of Oral Biology 73 (2017) 79–87

A number of OTUs identified as Prevotella and Treponema werealso present in the salivary microbiome, mostly high in abundanceat the Baseline. However, OTU140 and OTU208 (both Prevotella)were more abundant at Week 8 in the salivary microbiome, yetthese particular OTUs were not significantly abundant in theplaque microbiome. Interestingly, there are members of the genusPrevotella, which are associated with caries (Nadkarni et al., 2004;Torlakovic et al., 2012; Yang et al., 2012). It is possible that the taxawhich become lower in abundance in time are those associatedwith a lower pH, as would be the desired effect of arginine as aprebiotic. Still, the taxonomic resolution is too low to distinguishbetween species. Nevertheless, these results do emphasize thatsaliva and plaque are different niches within the same ecosystem(Human Microbiome Project Consortium, 2012; Yamanaka et al.,2012; Zaura et al., 2009). It is likely that a compound such asarginine has a different influence on a different habitat. Indeed, itwas observed that the use of toothpaste containing 1.5% argininedid induce the activity of the arginine deiminase system in dentalplaque and not in saliva (Nascimento et al., 2014). In our study weonly looked at the arginolytic potential of saliva and not of plaque.

Additionally, in this study we observed the functional potentialof saliva through metagenome shotgun sequencing. There wasenrichment of the tropane, piperidine and pyridine alkaloidbiosynthesis pathway (ko00960) at Week 8. Alkaloids aresecondary metabolites, mostly isolated from plants, which havea range of functions within human society, e.g. medication orstimulants (Facchini, 2001; Koleva, van Beek, Soffers, Dusemund, &Rietjens, 2012). Interestingly, arginine is a precursor in the tropanealkaloid pathway, where arginine is metabolized into putrescinethrough arginine decarboxylase (ADC) (Cui et al., 2015; Fuell,Elliott, Hanfrey, Franceschetti, & Michael, 2010; Hashimoto &Yamada, 1994). ADC is detected in higher plants and bacteria,generally not in mammals or fungi (Fuell et al., 2010; Rossi, Marina,& Pieckenstain, 2014). In contrast, the lysosome pathway(ko04142) was reduced at Week 8. Lysosomes are organelles inanimal cells; hence there is a possibility that part of the host DNAwas not removed using the stringent filters. However, the origins ofthe lysosome mechanism are thought to come from primitiveprotozoans and are strongly conserved in amoeba as well asmammalian cells (Cosson & Soldati, 2008). Therefore, lysosomepathway genes might originate from oral protozoa (Kutisova et al.,2005; Wantland & Lauer, 1970). Although this does not clarify whythese genes were reduced at Week 8, it does suggest that oralprotozoa might be an interesting topic for future research.

On a different note, we found OTU309 (Streptococcus), to bemore abundant in saliva after the use of arginine toothpaste. Thearginine deiminase system (ADS) is present in a number ofStreptococcus species (Casiano-Colon & Marquis, 1988; Huang,Schulte, Burne, & Nascimento, 2015; Nascimento et al., 2014), yetwe did not observe significant enrichment of ADS orthologs in thesalivary metagenome.

Why we did not observe enrichment of the ADS after eightweeks of 8% arginine toothpaste use remains largely unexplained,as do the observed changes in functional potential, although it isvery likely that the low sample size plays a part.

In order to determine the differences in abundance of OTUs,genera and KEGG pathways between the Baseline and Week 8, weused the biomarker selection tool LEfSe. This method is developedfor comparisons of independent multivariate datasets. To ourknowledge, there are to date no pairwise multivariate biomarkerselection tools available. Since the samples in our comparisonswere paired, the use of LEfSe might have led to an increase in bothtype I and type II errors. The need for the development of pairwisemultivariate statistical tools is highly urgent to avoid such errors infuture studies.

Overall, the small sample size and lack of a control group are themain drawbacks of this pilot study. Yet, instead of using a separatecontrol group, the data that were obtained at the Baseline andWash-out visits were used as control data. Because the changesobserved in this study agree with previous studies, it is reasonableto attribute these changes to the arginine toothpaste usage.

Nonetheless, a placebo-controlled randomized longitudinalclinical study including a higher number of participants andsamples should provide more conclusive answers, including theanalysis of the microbiota after cessation of treatment.

In summary, we conclude that arginine added to toothpastedoes bring about changes in the oral ecosystem, compositional aswell as functional, and serves as a prebiotic.

Conflicts of interest

The authors declare that no competing interests exist. Theinvolvement of the Colgate-Palmolive Company in this study byfinancial support had no role in study design, data collection andanalysis, decision to publish, or preparation of the manuscript.

Funding

This work was supported by the Colgate-Palmolive Company.

Acknowledgements

We are grateful for the volunteers willing to donate saliva andplaque. We like to thank Marja Jakobs of the AMC Amsterdam forher advice and assistance with the shotgun sequencing.

Appendix A. Supplementary data

Supplementary data associated with this article can be found,in the online version, at http://dx.doi.org/10.1016/j.archoralbio.2016.09.008.

References

Abubucker, S., Segata, N., Goll, J., Schubert, A. M., Izard, J., Cantarel, B. L., et al. (2012).Metabolic reconstruction for metagenomic data and its application to thehuman microbiome. PLoS Computational Biology, 8(6), e1002358. http://dx.doi.org/10.1371/journal.pcbi.1002358.

Bradford, M. M. (1976). A rapid and sensitive method for the quantitation ofmicrogram quantities of protein utilizing the principle of protein-dye binding.Analytical Biochemistry, 72, 248–254.

Bradshaw, D. J., & Lynch, R. J. (2013). Diet and the microbial aetiology of dentalcaries: New paradigms. International Dental Journal, 63(Suppl. 2), 64–72. http://dx.doi.org/10.1111/idj.12082.

Burne, R. A., & Marquis, R. E. (2000). Alkali production by oral bacteria andprotection against dental caries. FEMS Microbiology Letters, 193(1), 1–6.

Caporaso, J. G., Kuczynski, J., Stombaugh, J., Bittinger, K., Bushman, F. D., Costello, E.K., et al. (2010). QIIME allows analysis of high-throughput communitysequencing data. Nature Methods, 7(5), 335–336. http://dx.doi.org/10.1038/nmeth.f.303.

Casiano-Colon, A., & Marquis, R. E. (1988). Role of the arginine deiminase system inprotecting oral bacteria and an enzymatic basis for acid tolerance. Applied andEnvironmental Microbiology, 54(6), 1318–1324.

Cosson, P., & Soldati, T. (2008). Eat, kill or die: When amoeba meets bacteria. CurrentOpinion in Microbiology, 11(3), 271–276. http://dx.doi.org/10.1016/j.mib.2008.05.005.

Cui, L., Huang, F., Zhang, D., Lin, Y., Liao, P., Zong, J., et al. (2015). Transcriptomeexploration for further understanding of the tropane alkaloids biosynthesis inAnisodus acutangulus. Molecular Genetics and Genomics, 290(4), 1367–1377.http://dx.doi.org/10.1007/s00438-015-1005-y.

da Fonseca-Wollheim, F., Bergmeyer, H. U., & Gutmann, I. (1974). 2nd ed. In H. U.Bergmeyer (Ed.), New York and London: Verlag Chemie Weinheim andAcademic Press.

Downes, J., Munson, M. A., Spratt, D. A., Kononen, E., Tarkka, E., Jousimies-Somer, H.,et al. (2001). Characterisation of Eubacterium-like strains isolated from oralinfections. Journal of Medical Microbiology, 50(11), 947–951.

Duran-Pinedo, A. E., Chen, T., Teles, R., Starr, J. R., Wang, X., Krishnan, K., et al. (2014).Community-wide transcriptome of the oral microbiome in subjects with and

J.E. Koopman et al. / Archives of Oral Biology 73 (2017) 79–87 87

without periodontitis. ISME Journal, 8(8), 1659–1672. http://dx.doi.org/10.1038/ismej.2014.23.

Facchini, P. J. (2001). Alkaloid biosynthesis in plants: Biochemistry, cell biology,molecular regulation, and metabolic engineering applications. Annual Review ofPlant Physiology and Plant Molecular Biology, 52, 29–66. http://dx.doi.org/10.1146/annurev.arplant.52.1.29.

Fejerskov, O. (1997). Concepts of dental caries and their consequences forunderstanding the disease. Community Dentistry and Oral Epidemiology, 25(1),5–12.

Fuell, C., Elliott, K. A., Hanfrey, C. C., Franceschetti, M., & Michael, A. J. (2010).Polyamine biosynthetic diversity in plants and algae. Plant Physiology andBiochemistry, 48(7), 513–520. http://dx.doi.org/10.1016/j.plaphy.2010.02.008.

Grine, F. E., Gwinnett, A. J., & Oaks, J. H. (1990). Early hominid dental pathology:Interproximal caries in 1.5 million-year-old Paranthropus robustus fromSwartkrans. Archives of Oral Biology, 35(5), 381–386.

Gutmann, I., & Wahlefeld, A. (1974). L-(+)-Lactate determination with lactatedehydrogenase and NAD. Methods of Enzymatic Analysis, 3, 1464–1468.

Hammer, Ø., Harper, D. A. T., & Ryan, P. D. (2001). PAST: Paleontological statisticssoftware package for education and data analysis. Palaeontologia Electronica, 4(1), 9.

Hashimoto, T., & Yamada, Y. (1994). Alkaloid biogenesis: Molecular aspects. AnnualReview of Plant Biology, 45(1), 257–285.

Hoogenkamp, M. A., & ten Cate, J. M. (2014). Determination of arginine catabolismby salivary pellet. MethodsX, 1(0), 1–5. http://dx.doi.org/10.1016/j.mex.2014.01.001.

Huang, S., Yang, F., Zeng, X., Chen, J., Li, R., Wen, T., et al. (2011). Preliminarycharacterization of the oral microbiota of Chinese adults with and withoutgingivitis. BMC Oral Health, 11, 33. http://dx.doi.org/10.1186/1472-6831-11-33.

Huang, X., Exterkate, R. A. M., & ten Cate, J. M. (2012). Factors associated with alkaliproduction from arginine in dental biofilms. Journal of Dental Research, 91(12),1130–1134. http://dx.doi.org/10.1177/0022034512461652.

Huang, X., Schulte, R. M., Burne, R. A., & Nascimento, M. M. (2015). Characterizationof the arginolytic microflora provides insights into pH homeostasis in humanoral biofilms. Caries Research, 49(2), 165–176. http://dx.doi.org/10.1159/000365296.

Human Microbiome Project Consortium (2012). Structure, function and diversity ofthe healthy human microbiome. Nature, 486(7402), 207–214. http://dx.doi.org/10.1038/nature11234.

Kanehisa, M., & Goto, S. (2000). KEGG: Kyoto encyclopedia of genes and genomes.Nucleic Acids Research, 28(1), 27–30.

Kanehisa, M., Goto, S., Sato, Y., Kawashima, M., Furumichi, M., & Tanabe, M. (2014).Data, information, knowledge and principle: Back to metabolism in KEGG.Nucleic Acids Research, 42, D199–D205. http://dx.doi.org/10.1093/nar/gkt1076Database issue.

Koleva, I. I., van Beek, T. A., Soffers, A. E., Dusemund, B., & Rietjens, I. M. (2012).Alkaloids in the human food chain-natural occurrence and possible adverseeffects. Molecular Nutrition and Food Research, 56(1), 30–52. http://dx.doi.org/10.1002/mnfr.201100165.

Koopman, J. E., Roling, W. F., Buijs, M. J., Sissons, C. H., ten Cate, J. M., Keijser, B. J., et al.(2015). Stability and resilience of oral microcosms toward acidification andCandida outgrowth by arginine supplementation. Microbial Ecology, 69(2), 422–433. http://dx.doi.org/10.1007/s00248-014-0535-x.

Kraivaphan, P., Amornchat, C., Triratana, T., Mateo, L. R., Ellwood, R., Cummins, D., etal. (2013). Two-year caries clinical study of the efficacy of novel dentifricescontaining 1.5% arginine, an insoluble calcium compound and 1,450 ppmfluoride. Caries Research, 47(6), 582–590. http://dx.doi.org/10.1159/000353183.

Kraneveld, E. A., Buijs, M. J., Bonder, M. J., Visser, M., Keijser, B. J., Crielaard, W., et al.(2012). The relation between oral Candida load and bacterial microbiomeprofiles in Dutch older adults. PLoS One, 7(8), e42770. http://dx.doi.org/10.1371/journal.pone.0042770.

Kutisova, K., Kulda, J., Cepicka, I., Flegr, J., Koudela, B., Teras, J., et al. (2005).Tetratrichomonads from the oral cavity and respiratory tract of humans.Parasitology, 131(Pt. 3), 309–319.

Li, Y., Lee, S., Stephens, J., Mateo, L. R., Zhang, Y. P., & DeVizio, W. (2012). Comparisonof efficacy of an arginine-calcium carbonate-MFP toothpaste to a calciumcarbonate-MFP toothpaste in controlling supragingival calculus formation andgingivitis: A 6-month clinical study. American Journal of Dentistry, 25(1), 21–25.

Listl, S., Galloway, J., Mossey, P. A., & Marcenes, W. (2015). Global economic impact ofdental diseases. Journal of Dental Research, 94(10), 1355–1361. http://dx.doi.org/10.1177/0022034515602879.

Marquis, R. E., Bender, G. R., Murray, D. R., & Wong, A. (1987). Arginine deiminasesystem and bacterial adaptation to acid environments. Applied EnvironmentalMicrobiology, 53(1), 198–200.

Marsh, P. D. (1994). Microbial ecology of dental plaque and its significance in healthand disease. Advances in Dental Research, 8(2), 263–271.

Mugambi, M. N., Young, T., & Blaauw, R. (2014). Application of evidence onprobiotics, prebiotics and synbiotics by food industry: A descriptive study. BMCResearch Notes, 7, 754. http://dx.doi.org/10.1186/1756-0500-7-754.

Nadkarni, M. A., Caldon, C. E., Chhour, K. L., Fisher, I. P., Martin, F. E., Jacques, N. A., etal. (2004). Carious dentine provides a habitat for a complex array of novelPrevotella-like bacteria. Journal of Clinical Microbiology, 42(11), 5238–5244.http://dx.doi.org/10.1128/JCM.42.11.5238-5244.2004.

Nadkarni, M. A., Browne, G. V., Chhour, K. L., Byun, R., Nguyen, K. A., Chapple, C. C., etal. (2012). Pattern of distribution of Prevotella species/phylotypes associatedwith healthy gingiva and periodontal disease. European Journal of ClinicalMicrobiology and Infectious Diseases, 31(11), 2989–2999. http://dx.doi.org/10.1007/s10096-012-1651-5.

Nascimento, M. M., Liu, Y., Kalra, R., Perry, S., Adewumi, A., Xu, X., et al. (2013). Oralarginine metabolism may decrease the risk for dental caries in children. Journalof Dental Research, 92(7), 604–608. http://dx.doi.org/10.1177/0022034513487907.

Nascimento, M. M., Browngardt, C., Xiaohui, X., Klepac-Ceraj, V., Paster, B. J., & Burne,R. A. (2014). The effect of arginine on oral biofilm communities. Molecular OralMicrobiology, 29(1), 45–54. http://dx.doi.org/10.1111/omi.12044.

Nole, K. L. B., Yim, E., & Keri, J. E. (2014). Probiotics and prebiotics in dermatology.Journal of the American Academy of Dermatology, 71(4), 814–821.

Norris, S. J., Paster, B. J., Moter, A., & Göbel, U. B. (2006). The genus treponema. In M.Dworkin, S. Falkow, E. Rosenberg, K.-H. Schleifer, & E. Stackebrandt (Eds.), Theprokaryotes (pp. 211–234).New York: Springer.

Orel, R., & Kamhi Trop, T. (2014). Intestinal microbiota, probiotics and prebiotics ininflammatory bowel disease. World Journal of Gastroenterology, 20(33), 11505–11524. http://dx.doi.org/10.3748/wjg.v20.i33.11505.

Peterson, S. N., Snesrud, E., Schork, N. J., & Bretz, W. A. (2011). Dental cariespathogenicity: A genomic and metagenomic perspective. International DentalJournal, 61(Suppl. 1), 11–22. http://dx.doi.org/10.1111/j.1875-595X.2011.00025.x.

Raner, E., Lindqvist, L., Johansson, S., Hassan, H., Carlen, A., Suksu-art, N., et al.(2014). pH and bacterial profile of dental plaque in children and adults of a lowcaries population. Anaerobe, 27, 64–70. http://dx.doi.org/10.1016/j.anaerobe.2014.03.010.

Rogosa, M. (1964). The genus Veillonella. I. General cultural, ecological, andbiochemical considerations. Journal of Bacteriology, 87, 162–170.

Rogosa, M. (1984). Family 1 Veillonellaceae. In J. G. Holt (Ed.), Baltimore, MD:Williams & Wilkins.

Rossi, F. R., Marina, M., & Pieckenstain, F. L. (2014). Role of arginine decarboxylase(ADC) in Arabidopsis thaliana defence against the pathogenic bacteriumPseudomonas viridiflava. Plant Biology (Stuttg), 17(4), 831–839. http://dx.doi.org/10.1111/plb.12289.

Rotmistrovsky, K., & Agarwala, R. (2011). BMTagger: Best match tagger for removinghuman reads from metagenomics datasets. . ftp://ftp.ncbi.nlm.nih.gov/pub/agarwala/bmtagger/.

Sanders, M. E., Lenoir-Wijnkoop, I., Salminen, S., Merenstein, D. J., Gibson, G. R.,Petschow, B. W., et al. (2014). Probiotics and prebiotics: Prospects for publichealth and nutritional recommendations. Annals of the New York Academy ofSciences, 1309, 19–29. http://dx.doi.org/10.1111/nyas.12377.

Schmieder, R., & Edwards, R. (2011). Quality control and preprocessing ofmetagenomic datasets. Bioinformatics, 27(6), 863–864. http://dx.doi.org/10.1093/bioinformatics/btr026.

Segata, N., Izard, J., Waldron, L., Gevers, D., Miropolsky, L., Garrett, W. S., et al. (2011).Metagenomic biomarker discovery and explanation. Genome Biology, 12(6), R60.http://dx.doi.org/10.1186/gb-2011-12-6-r60.

Sharif, M. O., Iram, S., & Brunton, P. A. (2013). Effectiveness of arginine-containingtoothpastes in treating dentine hypersensitivity: A systematic review. Journal ofDentistry, 41(6), 483–492. http://dx.doi.org/10.1016/j.jdent.2013.01.009.

ten Cate, J. M. (2013). Contemporary perspective on the use of fluoride products incaries prevention. British Dental Journal, 214(4), 161–167. http://dx.doi.org/10.1038/sj.bdj.2013.162.

Torlakovic, L., Klepac-Ceraj, V., Ogaard, B., Cotton, S. L., Paster, B. J., & Olsen, I. (2012).Microbial community succession on developing lesions on human enamel.Journal of Oral Microbiology, 4, . http://dx.doi.org/10.3402/jom.v4i0.16125.

WHO (2012). World health organization: Oral health fact sheet N(318). WHO.Wade, W. G. (2006). The genus Eubacterium and related genera. In M. Dworkin, S.

Falkow, E. Rosenberg, K.-H. Schleifer, & E. Stackebrandt (Eds.), The prokaryotes(pp. 823–835).US: Springer.

Wantland, W. W., & Lauer, D. (1970). Correlation of some oral hygiene variables withage, sex, and incidence of oral protozoa. Journal of Dental Research, 49(2), 293–297.

Wu, G., Bazer, F. W., Davis, T. A., Kim, S. W., Li, P., Marc Rhoads, J., et al. (2009).Arginine metabolism and nutrition in growth, health and disease. Amino Acids,37(1), 153–168. http://dx.doi.org/10.1007/s00726-008-0210-y.

Yamanaka, W., Takeshita, T., Shibata, Y., Matsuo, K., Eshima, N., Yokoyama, T., et al.(2012). Compositional stability of a salivary bacterial population againstsupragingival microbiota shift following periodontal therapy. PLoS One, 7(8),e42806. http://dx.doi.org/10.1371/journal.pone.0042806.

Yang, F., Zeng, X., Ning, K., Liu, K. L., Lo, C. C., Wang, W., et al. (2012). Salivamicrobiomes distinguish caries-active from healthy human populations. ISMEJournal, 6(1), 1–10. http://dx.doi.org/10.1038/ismej.2011.71.

Zaura, E., Keijser, B. J., Huse, S. M., & Crielaard, W. (2009). Defining the healthy coremicrobiome of oral microbial communities. BMC Microbiology, 9, 259. http://dx.doi.org/10.1186/1471-2180-9-259.

Zhu, Y., Dashper, S. G., Chen, Y. Y., Crawford, S., Slakeski, N., & Reynolds, E. C. (2013).Porphyromonas gingivalis and Treponema denticola synergistic polymicrobialbiofilm development. PLoS One, 8(8), e71727. http://dx.doi.org/10.1371/journal.pone.0071727.