Acta agriculturae Slovenica, 90(december 2007)2, 107–113. 
 
  
Agris category codes: L51 COBISS Code 1.01 
EXAMINATION OF AI-2 QUORUM SENSING SYSTEM IN Prevotella bryantii AND 
Prevotella ruminicola-LIKE STRAINS BY USING BIOLUMINISCENCE ASSAY 
Gregor GORENC
 a)
, Filip LUKÁŠ
 b)
 and Gorazd AVGUŠTIN
 c)
 
a)
 Univ. of Ljubljana, Biotechnical Fac., Zootechnical Dept., Chair for Microbiology and Microbial Biotechnology, 
Groblje 3, SI-1230 Domžale, Slovenia. 
b)
 Academy of Science of the Czech Republic, Institute of Animal Physiology and Genetics, Víde ňská 1083, 142 20 
Prague, Czech Republic. 
c)
 Same address as 
a)
, e-mail: gorazd.avgustin@bfro.uni-lj.si. 
Received November 12, 2007, accepted December 03, 2007. 
Delo je prispelo 12. novembra 2007, sprejeto 03. decembra 2007. 
ABSTRACT 
Rumen is an example of a complex microbial ecosystem where extensive intraspecies as well as 
interspecies cell-cell cross-communication is expected, but little is known about. Four ruminal 
Prevotella strains, Prevotella bryantii B
1
4, TC1-1 and TF1-3, and P. ruminicola-like strain 
223/M2/7A were examined for the presence of AI-2 type quorum sensing systems. Vibrio 
harveyi BB170 autoinducer bioassay was used in order to detect the production of AI-2 
autoinducer in examined strains. Optimization of autoinducer bioassay was made to test the 
influence of sampling time and the substrate used in chemically defined medium CD on reporter 
strain luminescence signal. We discovered that AI-2 type qurum sensing system is present in 
P. ruminicola-like strain 223/M2/7A, which induced more than 74% of reporter strain 
bioluminescence, regardless of the presence of glucose in chemically defined medium CD. Such 
induction of bioluminescence was not observed in P. bryantii strains, however. 
Key words: microbiology / rumen / Prevotella / quorum sensing / bioassay / V. harveyi BB170 
UGOTAVLJANJE SISTEMA ZA ZAZNAVANJE KRITI ČNE CELI ČNE GOSTOTE   
AI-2 PRI BAKTERIJI Prevotella bryantii IN SEVIH, KI SO PODOBNI BAKTERIJI 
Prevotella ruminicola, S TESTOM BIOLUMINISCENCE 
IZVLE ČEK 
Vamp prežvekovalcev predstavlja kompleksen mikrobni ekosistem, kjer pri čakujemo obsežne 
tako znotrajvrstne kot tudi medvrstne medceli čne komunikacije, vendar je o njih le malo 
znanega. Pri štirih vampnih sevih iz rodu Prevotella, t.j. Prevotella bryantii B
1
4, TC1-1 in TF1-3 
ter P. ruminicola-podobnem sevu 223/M2/7A smo preverili prisotnost sistemov za zaznavanje 
celi čne gostote tipa AI-2. Z uporabo biološkega testiranja avtoinduktorjev s testnim organizmom 
Vibrio harveyi BB170 smo ugotavljali produkcijo molekul AI-2 pri preu čevanih 
mikroorganizmih. S preverjanjem vpliva časa biološkega testiranja ter substrata, uporabljenega v 
kemijsko definiranem gojiš ču CD, na signal bioluminiscence reporterskega seva, smo 
optimizirali biološko testiranje avtoinduktorskih molekul. Ugotovili smo, da je AI-2 tip sistema 
za medceli čno komunikacijo prisoten pri P. ruminicola podobnem sevu 223/M2/7A, ki je 
induciral ve č kot 74 % bioluminiscence reporterskega seva, ne glede na prisotnost glukoze v 
kemijsko definiranem gojiš ču CD. Indukcije bioluminiscence pri sevih vrste P. bryantii nismo 
opazili. 
Klju čne besede: mikrobiologija / vamp / Prevotella / zaznavanje celi čne gostote / biološko testiranje / V. harveyi 
BB170 
http://aas.bf.uni-lj.si 

Acta agriculturae Slovenica, 90(december 2007)2. 
 
108 
INTRODUCTION 
Rumen is a complex ecosystem with mixed microbial community of high density and 
complexity, a semi-closed environment with many possible symbiotic and pathogenic 
conditions, where extensive possibilities, systems and routes of intraspecies as well as 
interspecies cell-cell communication could be found (Mitsumori et al., 2003). Many bacteria use 
autoinducer molecules to monitor their population density, and regulate their gene expression 
and population behavior accordingly, a system termed quorum sensing. The first discovered type 
of quorum sensing system, AI-1, used by Gram-negative proteobacteria, employs acylated 
homoserine lactones for intra-species communication (Miller and Bassler, 2001). Recently it was 
reported that AI-1 was also present in rumen contents, but surprisingly, not in any of the cultured 
rumen bacteria (Erickson et al., 2002). The second type of quorum sensing system, AI-2, is the 
most widely found quorum sensing system, used for inter-species communication. It employs 
short peptides, produced by LuxS homolog (Surette et al., 1999; Bassler and Losick, 2006). The 
presence of AI-2 in rumen was also detected. Possible quorum sensing was reported in several 
rumen species, results interestingly depending on the type of the medium used in autoinducer 
bioassay (Mitsumori et al., 2003).  
Different substrates used, especially glucose, can significantly affect the outcome of AI-2 
bioassay, even at low concentrations (DeKeersmaecker and Vanderleyden, 2003; Turovskiy and 
Chikindas, 2006). Modification and optimization of the protocol is important for the 
reproducibility and reliability of quorum sensing bioassay (Vilchez et al., 2007). We have tested 
the influence of sampling time and the substrate used in chemically defined medium CD on 
reporter strain luminescence signal, and in this paper we report on our examination on four 
ruminal Prevotella strains, Prevotella bryantii B
1
4, TC1-1 and TF1-3, and P. ruminicola-like 
strain 223/M2/7A for the presence of AI-2 type quorum sensing systems. 
Strictly anaerobic Gram-negative bacteria from the genus Prevotella represent one of the most 
abundant and thus major bacterial populations within ruminant gut. They may constitute up to 
70% of the culturable bacterial population from the ruminal samples (Van Gylswyk, 1990). The 
abundance of these bacteria in rumen fluid was confirmed also by several molecular enumeration 
studies (Wood et al, 1998; Tajima et al., 2001; Reilly et al., 2002; Stevenson and Weimer, 2007).  
MATERIAL AND METHODS 
Bacterial strains and culture conditions used 
Four strictly anaerobic ruminal strains from genus Prevotella were used in this study: 
P. bryantii B
1
4, P. bryantii TC1-1, P. bryantii TF1-3 and P. ruminicola-like strain 223/M2/7A. 
The origin of the strains was described previously (Avguštin et al., 1997). Strains were grown in 
chemically defined media CD, described by Scott and Dehority (1965). 0.4% (w/v) glucose, 
0.4% (w/v) cellobiose or 0.4% (w/v) of each glucose and cellobiose were used as a substrate. 
Bacterial strains were incubated in 10 ml of CD medium under anaerobic conditions at 39 °C for 
1–2 days until maximum growth. The cultures were then separated into bacterial cells and 
culture supernatants by centrifugation at 10.000 x g for 15 min at 4 °C. Cell-free supernatants 
were stored at –20 °C until use in the autoinducer bioassay. 
Two strains of V. harveyi (obtained from ATCC) were used in AI-2 autoinducer assay. 
V. harveyi BB170 (sensor 1
–
, sensor 2
+
) was used as the reporter strain and V. harveyi BB152 
(AI-1
–
, AI-2
+
) was used as the producer of AI-2 molecules. Strains were grown overnight in AB 
medium (Greenberg et al., 1979). Cell-free supernatant of V. harveyi BB152 was obtained in the 
same way as from Prevotella strains, and served as the positive control.  

Gorenc, G. et al. Examination of AI-2 quorum sensing system in Prevotella … bioluminiscence assay. 
 
109
Autoinducer bioassay 
The bioassay for detection of AI-2 molecules, reported by Surette and Bassler (1998) was 
used. Optimizations of the protocol were made as suggested by Vilchez et al. (2007), and 
DeKeersmaecker and Vanderleyden (2003). V. harveyi BB170 reporter strain was grown 
overnight at 30 °C in AB medium until optimum optical density of 1.0–1.1 at 600 nm was 
reached. The culture was then diluted 1:5000 in fresh AB medium to obtain 10
5
 CFU ml
–1
. 1 ml 
of the cell-free supernatant from tested ruminal strains, V. harveyi BB152 (positive control) or 
sterile CD medium (negative control) was added to 9 ml of diluted reporter strain, thoroughly 
mixed and then incubated at 30 °C with aeration (175 RPM). Within next 6 hours every hour a 
sample of bioassay mixture was retrieved, 5-fold diluted in 2% NaCl solution and the stimulation 
of light production in V. harveyi BB170 reporter strain was measured by luminometer (LKB 
Wallac 1251).  
RESULTS AND DISCUSSION 
As reported previously, AI-2 production is growth-medium dependent (Burgess et al., 2002). 
Media composition can greatly influence the growth and luminescence of V. harveyi reporter 
strain BB170, as samples of the tested cultures are added to reporter strain in relatively large 
portion, i.e. 10% of the total volume. The presence of residual glucose in medium even at low 
concentrations (2 mM) completely inhibits the luminiscence of reporter strain (DeKeersmaecker 
and Vanderleyden, 2003). Nevertheless, glucose enables the reporter strain to grow faster than 
glycerol, which is the only carbon source in AB medium. It was shown that the presence of 
glucose in AB medium at 1.25 g/l can promote up to 7-times faster growth, causing higher 
densities of reporter strain and self-response during the bioassay. Bioluminiscence should be 
detected in response to the added AI-2 only at lower cell densities (no larger than 10
6
–10
7
 CFU 
ml
–1
) of reporter strain, with the concentration of glucose below the level of inhibition 
(Turovskiy and Chikindas, 2006).  
As reporter strain produces and senses its own autoinducer molecules, the time of the assay is 
of crucial importance. The measurement at the time with the most pronounced difference 
between positive and negative controls is considered the most informative. Our experiments 
confirm (Figure 1) that this difference is maximal 5 hours after inoculation, when interference by 
endogenous AI-2 of the reporter is less than 3% of bioluminescence signal, regardless of the 
substrate used in chemically defined medium CD. The cell-free CD medium without added 
substrate was also submitted to bioassay. Compared to CD medium with added substrates, no 
significant differences of luminiscence in reporter strain were observed, therefore indicating that 
the concentration of residual glucose in growth medium was below the level of bioluminescence 
inhibition.  
All P. bryantii strains reached maximum growth rapidly, with optical density of more than 1.6 
at 600 nm after only 18 h of incubation, whereas P. ruminicola-like strain 223/M2/7A achieved 
maximum growth more slowly, after 40–64 h (data not shown). Only after population transitions 
into the stationary phase, the concentration of AI-2 emerges at its peak (Turovskiy and 
Chikindas, 2006). The level of AI-2 then remains relatively constant all throughout the stationary 
phase with less than 1% change during 27 hours after entering the stationary phase (Surette and 
Bassler, 1998). The time of collecting the tested cell-free culture samples is therefore of less 
importance, provided that cultures had reached the stationary phase. 
 

Acta agriculturae Slovenica, 90(december 2007)2. 
 
110 
   
0,4
0,6
0,8
1
1,2
1,4
1,6
23456
Time (h)
Biouluminiscence (mV)
Positive control
AB medium
CDM+glucose
CDM+cellobiose
CDM+glucose+cellobiose
CDM without substrate
 
 
Figure 1. Determination of optimal time for autoinducer bioassay, using different media and 
substrates. Data is shown in absolute values of luminometer measurements. Samples were retrieved 
every hour during 6 hours of observation. Bioluminescence of the positive control grows slowly during 
the first 4 hours, after which it starts to peak rapidly. Bioluminescence of the samples from chemically 
defined medium CD (CDM) with different added substrates remains at the same level during the first 4 
hours, and then starts to grow gradually due to the endogenously produced AI-2 of the reporter strain. 
The largest difference between bioluminescence measurements of positive control and the samples from 
chemically defined medium CD emerges 5 hours after inoculation. 
Slika 1. Dolo čanje optimalnega časa biološkega testiranja ob uporabi razli čnih gojiš č in 
substratov. Prikazane so absolutne vrednosti meritev z luminometrom. Vzorci so bili odvzeti vsako 
uro, skupaj 6 ur. Bioluminiscenca pozitivne kontrole po časi raste prve 4 ure, nato pa se za čne hitro 
pove čevati. Bioluminiscenca vzorcev iz kemijsko definiranih gojiš č CD (CDM) z dodanimi razli čnimi 
substrati prakti čno ne naraš ča do 4. ure, potem pa se postopoma pove čuje zaradi nastajanja endogenih 
avtoinduktorskih molekul tipa 2 reporterskega seva. Najve čja razlika med bioluminiscenco pozitivne 
kontrole in vzorcev iz kemijsko definiranih gojiš č CD se pojavi po 5 urah rasti. 
 
Using the V. harveyi BB170 bioassay, considering culture cell density and growth time, we 
have clearly demonstrated the presence of AI-2 quorum sensing system in P. ruminicola-like 
strain 223/M2/7A, but not in any of the P. bryantii strains. P. ruminicola-like strain 223/M2/7A 
induced more than 74% of reporter V. harveyi BB170 bioluminiscence, regardless of the 
substrate used in chemically defined medium CD (Figure 2).  
 

Gorenc, G. et al. Examination of AI-2 quorum sensing system in Prevotella … bioluminiscence assay. 
 
111
         
74,65
0,29
91,30
3,35
6,39
-1,14
-0,69
0,38
-10,00
10,00
30,00
50,00
70,00
90,00
B14 TC1-1 223/M2/7A TF1-3
RLU
glc+cb
cb
 
 
Figure 2. Average light induction in AI-2 bioassays of ruminal strains from the genus 
Prevotella. Bioluminiscence values are shown in relative light units (% induction of bioluminiscence 
in positive control). Black bars denote samples from chemically defined medium CD with added 
glucose and cellobiose as a substrate. White bars denote samples from chemically defined medium CD 
with added cellobiose as a substrate. 
Slika 2. Povpre čne vrednosti indukcije svetlobe pri biološkem testiranju avtoinduktorskih 
molekul tipa 2 pri sevih rodu Prevotella iz vampa. Vrednosti bioluminiscence so prikazane v 
relativnih enotah svetilnosti (delež indukcije bioluminiscence pri pozitivni kontroli v %). Črni stolpci 
ozna čujejo vzorce iz kemijsko definiranega gojiš ča CD z dodano glukozo in celobiozo. Beli stolpci 
ozna čujejo vzorce iz kemijsko definiranega gojiš ča CD z dodano celobiozo. 
 
Partial substitution of glucose in some cases increases, in other oppositely decreases the 
induction of bioluminiscence in reporter strain, but never changes the outcome of bioassay. The 
increase in light production could be attributed to significant increase in growth rates of reporter 
strain due to addition of easily accessible substrate, as decrease could be caused by inhibition 
effect of glucose on bioluminiscence.  
In P. bryantii B
1
4, P. bryantii TC1-1 and P. bryantii TF1-3 stimulation of bioluminiscence, 
indicating the presence of quorum sensing system, was not observed. Despite P. bryantii strain 
TF1-3 induced significantly more light production in reporter strain than other P. bryantii strains, 
its 3.35–6.39% induction was below 10% induction threshold, sufficient for confirmation of AI-2 
molecules (Bassler et al., 1997).  
We cannot affirm that the P. bryantii strains that didn't induce significant bioluminiscence of 
reporter strain do not possess quorum sensing systems of any type. In addition to the 
abovementioned factors that notably influence the autoinducer bioassay, the difference in 
autoinducer molecule structure between bacteria from the genus Prevotella and V. harveyi, two 
phylogenetically relatively distant bacteria taxons, could cause the inability for successful 
interspecies communication and consecutive failure to detect it by using the V. harveyi 
autoindicer bioassay. Alternatively, one could employ the chemical detection of AI-2 by HPLC 

Acta agriculturae Slovenica, 90(december 2007)2. 
 
112 
or GC, but at the present these methods are rarely used due to the instability and low 
concentrations of AI-2 molecules in biological samples (Vilchez et al., 2007). 
CONCLUSIONS 
Our experiments confirmed that the signal of V. harveyi reporter strain is minimal 5 hours 
after inoculation, when interference by endogenous AI-2 of the reporter is less than 3% of 
bioluminescence signal, regardless of the presence of glucose in chemically defined medium CD. 
As it was expected, using V. harveyi BB170 bioassay, considering culture cell density and 
growth time, we have demonstrated the presence of AI-2 quorum sensing system in ruminal 
Prevotella strains. P. ruminicola-like strain 223/M2/7A induced more than 74% of reporter 
V. harveyi BB170 bioluminescence, regardless of the presence of glucose in chemically defined 
medium CD, however, such induction of bioluminescence was not observed in P. bryantii 
strains. 
REFERENCES 
Avguštin, G./ Wallace, J.R./ Flint, H.J. Phenotypic diversity among rumen isolates of Prevotella ruminicola, 
proposal for redefinition of Prevotella ruminicola and the creation of Prevotella brevis sp. nov., Prevotella 
bryantii sp. nov., and Prevotella albensis sp. nov. International Journal of Systematic Bacteriology, 47(1997), 
284–288. 
Avguštin, G./ Ramšak, A./ Peterka, M. Systematic and evolution of ruminal species of the genus Prevotella. Folia 
Microbiologica, 46(2001), 40–44. 
Bassler, B.L./ Greenberg, E.P./ Stevens, A.M. Cross-Species Induction of Luminiscence in the Quorum-Sensing 
Bacterium Vibrio harveyi. Journal of Bacteriology, 179(1997), 4043–4045. 
Bassler, B.L./ Losick, R. Bacterially speaking. Cell, 125(2006), 237–246. 
Burgess, N.A./ Kirke, D.F./ Williams, P./ Winzer, K./ Hardie, K.R./ Meyers, N.L./ Aduse-Opoku, J./ Curtis, M.A./ 
Camara, M. LuxS-dependent quorum sensing in Porphyromonas gingivalis modulates protease and 
haemagglutinin activities but is not essential for virulence. Microbiology, 148(2002), 763–772. 
DeKeersmaecker, S.C./ Vandrleyden, J. Constraints on detection of autoinducer-2 (AI-2) signalling molecules using 
Vibrio harveyi as a reporter. Microbiology Comment, 149(2003), 1953–1956. 
Erickson, D.L./ Nsereko, V.L./ Morgavi, D.P./ Selinger, L.B./ Rode, L.M./ Beauchemin, K.A. Evidence of quorum 
sensing in the rumen ecosystem: detection of N-acyl homoserine lactone autoinducers in ruminal contents. 
Canadian Journal of Microbiology, 48(2002), 374–378. 
Greenberg, E.P./ Hastings, J.W./ Ulitzur, S. Induction of luciferase synthesis in Beneckea harveyi by other marine 
bacteria. Archives of Microbiology, 120(1797), 87–91. 
Miller, M.B./ Bassler, B.L. Quorum Sensing in Bacteria. Annual Review of Microbiology, 55(2001), 165–199. 
Mitsumori, M./ Xu, L./ Kajikawa, H./ Kurihara, M./ Tajima, K./ Hai, J./ Takenaka, A. Possible quorum sensing in 
the rumen microbial community: detection of quorum-sensing signal molecules from rumen bacteria. FEMS 
Microbiology Letters, 219(2003), 47–52. 
Reilly, K./ Carruthers, V.R./ Attwood, G.T. Design and use of 16S ribosomal DNA-directed primers in competitive 
PCRs to enumerate proteolytic bacteria in the rumen. Microbial Ecology, 43(2002), 259–270. 
Scott, H.W./ Dehority, B.A. Vitamine requirement of several cellulolitic bacteria. Journal of Bacteriology, 89(1965), 
1169–1175. 
Stevenson, D.M./ Weimer, P.J. Dominance of Prevotella and low abundance of classical ruminal bacterial species in 
the bovine rumen revealed by relative quantification real-time PCR. Applied Microbiology and Biotechnology, 
75(2007), 165–174. 
Surette, M.G./ Bassler, B.L. Quorum sensing in Escherichia coli and Salmonella typhimurium. Proceedings of the 
National Academy of Sciences of the United States of America, 95(1998), 7046–7050. 
Surette, M.G./ Miller, M.B./ Bassler, B.L. Quorum sensing in Escherichi coli, Salmonella typhimurium, and Vibrio 
harveyi: a new family of genes responsible for autoinducer production. Proceedings of the National Academy of 
Sciences of the United States of America, 96(1999), 1639–1644. 
Tajima, K./ Aminov, R.I./ Nagamine, T./ Matsui, H./ Nakamura, M./ Benno, Y. Diet-dependent shifts in the 
bacterial population of the rumen revealed with real time PCR. Applied and environmental microbiology, 
67(2001), 2766–2774. 

Gorenc, G. et al. Examination of AI-2 quorum sensing system in Prevotella … bioluminiscence assay. 
 
113
Turovskiy, Y./ Chikindas, M.L. Autoinducer-2 bioassay is a qualitative, not quantitative method influenced by 
glucose. Journal of Microbiological Methods, 66(2006), 497–503. 
Van Gylswyk, N.O. Enumeration and presumptive identification of some functional groups of bacteria in the rumen 
of dairy cows fed grass silage-based diets. FEMS Microbiology Ecology, 73(1990), 243–254. 
Vilchez, R./ Lemme, A./ Thiel, V./ Schulz, S./ Sztajer, H./ Wagner-Döbler, I. Analysing traces of autoinducer-2 
requires standardization of the Vibrio harveyi bioassay. Analytical and Bioanalytical Chemistry, 387(2007), 
489–496. 
Wood, J./ Scott, K.P./ Avguštin, G./ Newbold, C.J./ Flint, H.J. Estimation of the relative abundance of different 
Bacteroides and Prevotella ribotypes in gut samples by restriction enzyme profiling of PCR-amplified 16S 
rRNA gene sequences. Applied and Environmental Microbiology, 64(1998), 3683–3689.