RAPID DETECTION OF GERMINATING BACILLUS CEREUS CELLS USING FLUORESCENT IN SITU HYBRIDIZATION

RAPID DETECTION OF GERMINATING BACILLUS CEREUSCELLS USING FLUORESCENT IN SITU HYBRIDIZATION

CHRISTIAN LAFLAMME1,2, LOUIS GENDRON1,2, NATHALIE TURGEON1,2,GENEVIEVE FILION1,2, JIM HO3 and CAROLINE DUCHAINE1,2,4

1Institut Universitaire de cardiologie et de pneumologieHôpital Laval

2Département de biochimie et microbiologieFaculté des sciences et de génie

Université Laval,Québec, Québec, Canada

3Biological detection groupDefense R & D Canada Suffield

CFB SuffieldRalston, Alberta, Canada

Accepted for Publication July 21, 2008

ABSTRACT

Methods for the specific detection of Bacillus spores are needed in manysituations such as the recognition of food poisoning. This study presents anexperimental design in order to find the best combination of germinationconditions leading to a rapid and detectable fluorescent in situ hybridization(FISH) signal from Bacillus cereus spores present in pure cultures and milksamples.

B. cereus ATCC 14579 and HER 1414 were incubated in 20 differentgrowth media by using a combination of various germinants such as sugars,amino acids and dipicolinic acid. Also, three different germination factorswere tested: incubation temperature, inoculum concentration and a heat shocktreatment. Permeabilization procedure and hybridization time were optimizedon the best germination condition found. B. cereus-specific FISH probes werevalidated under the optimized condition and in detection of spiked B. cereusspores in 1% ultra heat-treated milk samples. FISH-labeled cells weredetected by using flow cytometry, and the results were confirmed by fluores-cence microscopy. The optimal condition allows the detection of B. cereus

4Corresponding author. Caroline Duchaine, Centre de recherche, Hôpital Laval, 2725 Chemin Ste-Foy,Ste-Foy (Quebec), Canada G1V 4G5. TEL: 418-656-8711 ext. 5837; FAX: 418-656-4509; EMAIL:[email protected]

Journal of Rapid Methods & Automation in Microbiology 17 (2009) 80–102. All Rights Reserved.© 2009, The Author(s)

Journal compilation © 2009, Wiley Periodicals, Inc.80

mailto:[email protected]

spores in less than 2 h. Overall, a ninefold reduction in total time for detectionwas achieved when comparing with previous works. Therefore, the permeabi-lization and hybridization optimizations mentioned in this study are majorimprovements for the detection time of B. cereus spores.

PRACTICAL APPLICATIONS

By using the optimized conditions of germination/outgrowth, permea-bilization and hybridization, the detection of 103 cfu/mL of Bacillus cereusspores using fluorescent in situ hybridization is possible within 2 h in milksample.

INTRODUCTION

Bacillus cereus is a common environmental bacteria that can be found insoil as well as a common contaminant of milk (Griffiths 1992; Eneroth et al.2001; Christiansson 2003). This endemic organism has been recognized asan agent of food spoilage and poisoning (Granum and Lund 1997). When theenvironmental conditions become unfavorable for the development of B.cereus vegetative form, those cells have the ability to enter in dormancy state,called spores (Setlow 1994). Spores can be produced in high numbers andsurvive for years, even decades (Driks 2002). The presence of B. cereus sporesis a serious problem in the food industry, as spores are heat-resistant and veryhydrophobic, and they can adhere to equipment surfaces (Koshikawa et al.1989; Faille et al. 2002). Their remarkable resistance allows them to survivefood processing and conservation methods. With the process of germinationand outgrow, the bacterial spores return to the metabolically active vegetativestate, which involves a rapid sequence of events that lead to the breakdown ofthe spores’ structure and the simultaneous loss of the spores’ resistance prop-erties (Moir 2006).

The fast and specific detection of B. cereus spores in the environment isa challenging task. Culture on Petri plates is often used, but this technique istime-consuming. Polymerase chain reaction (PCR) assay can also be used, butmay be affected by inhibition because of food components (Quarto and Chi-ronna 2005). Thus, enrichment step and sample preparation is needed beforePCR that increases the time to response (Fukushima et al. 2007; Park et al.2007; Perry et al. 2007). Fluorescent in situ hybridization (FISH) is used toidentify and characterize bacterial cells by using ribosomal RNA (rRNA) ashybridization targets for probes. FISH allows single cell detection of a specifictaxon and is suitable for complex environments (Amann et al. 1990). Com-

81DETECTION OF GERMINATING B. CEREUS CELLS USING FISH

pared with PCR, this technique has the advantage of providing the visualiza-tion of the cells in their environment and evaluating their proportion. Thetechnique is based upon a two-step sequence: cell permeabilization withproper fixative followed by hybridization under appropriate conditions witholigonucleotide probes. The FISH approach faces several constraints. Alongwith the cellular concentration of rRNA (probe target), the permeabilizationstep is another critical limitation (Oda et al. 2000; Wagner et al. 2003). In fact,FISH probes need to penetrate the cell to bind to their target.

A study published by the Kornberg group in the 1960s indicated that therRNA content of dormant spores of Bacillus, in terms of relative amount andphysicochemical property, are at the same level as log phase cells (Chambonet al. 1968). In order to be detectable by FISH, Bacillus spores would have tobe (1) strongly permeabilized or (2) induced to germinate/growth (i.e. return topermeable vegetative phase) and permeabilized. Indeed, germination and out-growth is facilitated by enzymes and energy reserves already present in Bacil-lus spores. Because the structure of the spore itself is the barrier, there is anadvantage in using natural biophysical means like germination as a startingpoint for a permeabilization treatment. In fact, the germination of spores isaccompanied by a loss of resistance properties. The inner membrane of thespore is supposed to be the major obstacle for the diffusion of small moleculesinto the spore’s core (Setlow 2003). The permeability of this membraneincreases by over 100-fold early in germination (Swerdlow et al. 1981). Fur-thermore and with respect to the first stages of spore germination, the largedepot of dipicolinic acid (DPA) is excreted from the spore’s core while wateris absorbed to reduce dehydration of the spore (Setlow 2003; Moir 2006). Thecortex hydrolysis that occurs later in germination but before outgrowth isneeded for complete core rehydration. Eventually the spore’s coat is degradedby an unresolved mechanism.

A report has demonstrated the feasibility of direct FISH on pure culturedspores (Fischer et al. 1995). The time required to obtain a FISH signal by usingthe protocol is 3 days and needs an extensive list of successive chemicals suchas 4% parformaldehyde, 50% ethanol and lysozyme. The assay was performedon a microscope slide. A recent study showed that it was possible to obtain aFISH signal from germinating pure cultured spores after 6–8 h of treatmentand handling using formaldehyde, lysozymes and different steps of drying andfixation of the samples on microscope slides. Additionally, a germination stepof at least 2 h was necessary prior to fixation and permeabilization of the cells(Regamey et al. 2000). Therefore, there is a need to reduce the time andimprove the conditions required to detect B. cereus spores by using FISH.Spore germination and growth induction followed by a nondestructive perme-abilization treatment (like ethanol or aldehyde) is a promising strategy in termsof execution time, cost efficiency and simplicity for environmental samples.

82 C. LAFLAMME ET AL.

Germination of Bacillus spores is induced by several nutrients calledgerminants (Paidhungat and Setlow 2002). These germinants are most oftensingle amino acids, sugars and purine nucleosides (Hornstra et al. 2005).These components are recognized by spores as signals for appropriate germi-nation conditions, but the mechanisms have not yet been elucidated (Paidhun-gat et al. 2001). Different combinations of germinants were used with B.cereus and among them, the addition of L-alanine and inosine was describedas one of the most optimal mix for germination (Clements and Moir 1998;Barlass et al. 2002; Hornstra et al. 2005). Other combinations, such as theaddition of calcium chloride (CaCl2) with L-alanine, produce synergisticeffects (Kamat et al. 1985). Another way to improve germination is by addingDPA in the culture media (Ragkousi et al. 2003; Setlow 2003). This supple-ment is a natural constituent of the spore cortex, which is released when thespores germinate and then causes other spores in the environment to germi-nate. Furthermore, the inoculum concentration and heat activation at 70C isknown to induce the germination (Keynan and Evenchik 1969; Caipo et al.2002; Setlow 2003).

The main objective of this study was to evaluate the germination condi-tions leading to a rapid and detectable FISH signal when starting from B.cereus spores. The current study aimed to answer two questions: (1) By usingthe most favorable germinating and growth conditions found, could B. cereuscells be specifically detected by FISH within a milk sample? (2) How quicklycan B. cereus spores be detected by FISH? Various germination conditionshave been tested, and the proposed strategy leads to specific detection of B.cereus spores by FISH in less than 2 h. The procedure was tested on spikedmilk samples with spores and allows for the specific detection of 103 colony-forming units (cfu) per milliliter in 2 h.

MATERIALS AND METHODS

Bacterial Strains

B. cereus ATCC 14579 and B. cereus HER 1414 strain (Ahmed et al.1995), encountered in food poisoning, were used. B. cereus HER 1414 wasprovided by the Felix d’Hérelle Reference Center for Bacterial Viruses.

Production and Purification of Spores

One aliquot (one loop) of each liquid culture, previously incubated intryptic soy broth (TSB) overnight at 37C, was used to inoculate a solidsporulation medium (nutrient agar [Difco, Sparks, MD] with 0.5% yeastextract, 7 ¥ 10-4 M CaCl2, 10-3 M MgCl2•6 H2O and 5 ¥ 10-5 M MnCl2•4 H2O,


final pH of 6.8), and it was incubated for a minimum 72 h at 37C (Holt andKrieg 1994). Colonies were harvested with a sterile cotton swab and trans-ferred in 1 mL of 0.22 mm filtered phosphate buffer saline (PBS). After twowashing steps by using PBS, the samples were gently deposited on top of aNaBr density gradient and centrifuged at 2,400 ¥ g at 25C for 45 min aspreviously described (Laflamme et al. 2004).

Spore Count and Purity Evaluation

The purity of the spore preparations was evaluated by phase contrastmicroscopy: a wet mount was used to estimate the degree of purity of thespores isolated from the NaBr density gradient. In addition, to confirm theabsence of debris and to quantify the spores, a fluorescent DNA marker, the 4′,6-diamidino-2-phenylindole dihydrochloride (DAPI), was used to stain thespores (Laflamme et al. 2005). According to the count obtained, spores wereresuspended at a concentration of 109 spores/mL. By using this technique, thespore preparations were at least 95% pure, bright phase spores without debrisor vegetative cells.

Growth Media and Experimental Design

A total of 20 different growth media were tested in this study (Table 1).An optimal concentration of L-alanine, inosine, CaCl2 and DPA for B. cereusgermination was chosen following a literature review (Raso et al. 1998; Paid-hungat et al. 2001; Ireland and Hanna 2002; Hornstra et al. 2006). Purified B.cereus spores of each strain were incubated at 37C under agitation at 150 rpmand at a concentration of 5 ¥ 106 spores/mL in each growth media listedin Table 1 (n = 3). Samples (1 mL) were taken after 0, 30, 60 and 120 minof incubation. The collected samples were pelleted by centrifugation at12,000 ¥ g for 2 min and resuspended in 50% ethanol-PBS for 18 h in order topermeabilize the cells and obtain a maximum of positive cells for FISH. Thismethod of permeabilization is frequently used in literature for gram-positivebacteria like Bacillus cells (Moter and Göbel 2000). The best growth media foreach strain were selected according to the best FISH signal in those media at37C after 120 min. The best FISH signal is defined as one that gave the highestpercentage of positive cells and fluorescence intensity. Moreover, the FISHresult must be visible under fluorescence microscopy. The selected media wereused for the rest of the experiment to test the influence of three germinationfactors, permeabilization and hybridization conditions. Finally, the best con-ditions found were validated on a milk sample spiked with B. cereus HER1414. Figure 1 summarizes the experimental design.


Germination Factors

Three factors known to increase the germination efficiency were tested:(1) incubation temperature (30C versus 37C); (2) inoculum concentration(from 5 ¥ 106 to 5 ¥ 108 spores/mL); and (3) a heat shock treatment. Thespores (5 ¥ 106 spores/mL) were inoculated in the optimal growth media andincubated for 30 min at 70C, under agitation in a hybridization incubator(Robbins Scientific, model 310 Robbins Scientific Corp., Sunnyvale, CA)prior to incubation at 37C under agitation at 150 rpm. To test the germinationfactors, 1 mL samples were taken after 0, 30, 60 and 120 min. The cells werepelleted by centrifugation at 12,000 ¥ g for 2 min and resuspended in 50%ethanol-PBS for 18 h before the FISH experiment.

Permeabilization

By using the best growth medium and germination conditions determinedin steps 1 and 2, three standard fixation/permeabilization treatments for FISHwere tested: 1% formaldehyde, 4% paraformaldehyde and 50% ethanol-PBS(Amann et al. 1990; Moter and Göbel 2000). Fixation time intervals of 15,30 min, 1, 2, 4, 8 and 18 h were tested for each of them.

TABLE 1.GROWTH MEDIA TESTED

Media Composition

1 Tris–HCl 10 mM (pH 8)2 Tris–HCl 10 mM (pH 8) + L-alanine 100 mM3 Tris–HCl 10 mM (pH 8) + inosine 10 mM4 Tris–HCl 10 mM (pH 8) + L-alanine 100 mM + inosine 10 mM5 Tris–HCl 10 mM (pH 8) + L-alanine 100 mM + CaCl2 10 mM6 PBS7 PBS + L-alanine 100 mM8 PBS + inosine 10 mM9 PBS + L-alanine 100 mM + inosine 10 mM

10 PBS + L-alanine 100 mM + CaCl2 10 mM11 TSB12 TSB + L-alanine 100 mM13 TSB + inosine 10 mM14 TSB + L-alanine 100 mM + inosine 10 mM15 TSB + CaCl2 10 mM16 TSB + L-alanine 100 mM + CaCl2 10 mM17 TSB + DPA 60 mM18 TSB + L-alanine 100 mM + DPA 60 mM19 TSB + L-alanine 100 mM + CaCl2 10 mM + DPA 60 mM20 TSB + CaCl2 10 mM + DPA 60 mM

CaCl2, calcium chloride; DPA, dipicolinic acid; PBS, phosphate buffer saline; TSB, tryptic soy broth.


FISH

Universal FISH probe designed for eubacteria 16S rRNA, EUB338 (5′-GCT GCC TCC CGT AGG AGT-3′) and the antisense sequence notEUB338(Amann et al. 1990), both labeled with FITC, were used to test steps 1–4 of theexperimental design (Fig. 1) (Invitrogen Life Technologies, Frederick, MD).These probes were chosen to set up the rapid conditions as their target is wellcharacterized and their use in FISH protocols was thoroughly described andvalidated (Amann et al. 1990; Fischer et al. 1995; Fuchs et al. 1998; Wagneret al. 2003). In addition, a B. cereus-specific probe, pB394 (5′-ATG CGG TTCAAA ATG TTA TCC GG-3′) and its corresponding antisense notpB394 wereused to validate the best conditions obtained at steps 5 and 6 of the experi-

Step 1: Growth media

Permeabilization : 50 %Ethanol/PBS-18 hrs

Step 3:Permeabilization treatment

and minimal time of permeabilization

Step 6:Validate the optimal condition in environmental models

Step 4:Minimal hybridization time

Step 2:Influence of germination factor

on optimal condition found

Experimental design

Step 5:Validate the optimal condition using specific probe

FIG. 1. EXPERIMENTAL DESIGN OF THE STUDY


mental design (Fig. 1). The probes were labeled with Alexa 488 (IntegratedDNA Technology, Coralville, IA). The pB394 probe targeted a uniquesequence present in the 16S rDNA of B. cereus (Liu et al. 2001). Prior to eachFISH reaction, the hybridization buffer was preheated at 46C. The reactiontubes contained 95 mL of hybridization buffer (0.9 M, NaCl, 0.01% SDS,20 mM Tris–HCl pH 7.6 with 6.5 ng/mL of probe) and 5 mL of germinatingspore suspension. Each sample was hybridized for 2 h at 46C. FISH reactionswere performed by using a thermal cycler (DNA engine, DYAD, MJ Research,BIO-RAD, Waltham, MA) for temperature accuracy.

Flow Cytometry

Flow cytometry analyses were performed by using an EPICS XL-MCLflow cytometer (Beckman-Coulter, Miami, FL) with the acquisition softwareEXPO 32 (version 1.1c). The flow cytometer was equipped with an air-cooled15 mw argon laser as a light source. Fluorescence signals were collectedthrough a 525 nm bandpass filter (FL1). The acquisition of fluorescence datawere gated by forward angle light scatter and side scatter while the data ratewas set at less than 500 events/s. Samples were allowed to run approximately1 min before the acquisition of a minimum of 5,000 events. The percentageof positive cells as well as the mean fluorescence intensity were calculatedand normalized between samples by using a standard method by subtractingthe corresponding antisense probe results. The flow cytometer performanceand stability over time was periodically controlled by using flow set beads(Beckman-Coulter) in order to evaluate the possible fluorescence drift becauseof the instrument. No variation was observed during the period of experimen-tations (data not shown). Flow cytometry was the instrument used to determinethe percentage of FISH positive cells and the mean fluorescence that is anindication of the FISH signal intensity.

Fluorescence Microscopy

The fluorescence microscopy analyses were performed by using a NikonEclipse 6600 with three sets of spectral filters (UV-2A, FITC and B-2A). Themicroscope was connected to a system of picture capture (Retiga 1300, QIm-aging). The software Simple PCI version 5.1 (Campix, Inc., Cranberry Town,PA) was used for picture capture and image analysis. Fluorescence microscopywas used at each step of the experimental design in order to confirm the signalobtained with flow cytometry and for the environmental sample analysis.

Specific Detection of B. cereus HER 1414 in a Milk Sample

The number of colony-forming units (cfu) of freshly isolated B. cereusHER 1414 spores was determined by plate counts using tryptic soy agar.


Various quantities of spores (106–101 cfu) were inoculated directly in 1 mL of1% ultra heat-treated milk. In order to remove the major part of fat present inmilk, 60 mL of 25% sodium citrate (Sigma-Aldrich, Oakville, ON, Canada)was put into the 1 mL of spiked milk and shaken for 5 min at 200 rpm (Lucoreet al. 2000). The tubes were centrifuged at 15,000 ¥ g for 5 min. The parts ofthe cream that adhered to the wall of the tube were removed and withdrawn byusing a sterile cotton swab. The tubes were then emptied by inversion while thepellets remained at the bottom. The residual cream was scrapped and removedfrom the tube, and the pellet was resuspended by using 1 mL of TSB +L-alanine + inosine (medium 14, Table 1). The tubes were incubated for 1 h at37C under agitation at 150 rpm. The tubes were centrifuged at 12,000 rpm for2 min, and the pellets were resuspended in 1 mL of 4% paraformaldehyde for15 min. Then, they were centrifuged at 12,000 rpm for 2 min, and the FISHprotocol was performed for 15 min by using pB394 and notpB394 probes.

Statistical Analyses

The statistical analyses were completed by using the Statistical AnalyticalSoftware. The results were expressed as mean values � standard error of themean. The data were analyzed by using a three-way ANOVA. All reportedP values were declared significant at P < 0.05.

RESULTS

Step 1: Selection of the Best Growth Media

Purified spores of B. cereus ATCC 14579 and HER 1414 were incubatedin 20 different growth media (Table 1). Samples of cells were taken after 0, 30,60 and 120 min, fixed and FISH was completed for each sample by using anEUB338 probe and a notEUB338 probe (n = 3). Each sample was analyzed byusing flow cytometry and fluorescence microscopy. For B. cereus ATCC 14579and HER 1414, the growth media 1–10 and 19 yielded negative results after120 min of incubation. In order to simplify Fig. 2, negative results wereomitted. In addition, the results obtained at time intervals of 0 and 30 min wereomitted in Fig. 2 as the results were consistently too low to be detected. At60 min, only medium 14 provided a statistical difference with that obtainedfrom medium 12 (P = 0.048). At 120 min, media 11, 12, 14, 15 and 16represented the best growth media for the percentage of positive results(Fig. 2A). The mean fluorescence is an indication of the averaged intensity offluorescence emitted by the entire cell population analyzed. At 60 min, therewas no significant difference between each medium. However, at 120 min,media 14, 15 and 18 led to similar results and were better than the other media(Fig. 2B).


At 60 min, the media 11, 13, 14, 15 and 16 constituted the best growthmedia in terms of percentage of positive cells. At 120 min, media 11, 12, 14,17 and 20 were the five best media (Fig. 2C).

At 60 min, media 11, 14, 15, 16 and 20 gave the best results. At 120 min,the media 11, 12, 14 16 and 17 gave the best result (Fig. 2D). The FISH resultswere validated by fluorescence microscopy (data not shown). After 60 minof incubation, positive cells can be observed. Control experiments by usingthe antisense probe notEUB338 were still negative under fluorescencemicroscopy.

Following the experimental design in Fig. 1, step 1 was performed toselect the medium that gave the best results in terms of percentage of positivecells and mean fluorescence intensity in flow cytometry. Moreover, the FISHresults had to be validated in fluorescence microscopy. For the percentage ofpositive cells and mean fluorescence obtained with B. cereus ATCC 14579(Fig. 2A,B), at 60 and 120 min, the media 14 and 15 were the best accordingto ANOVA statistics. For HER 1414, by using the same criteria, (Fig. 2C,D),media 11 and 14 were the best. Therefore and in accordance with these results,the best growth medium for the rest of the experiment was medium 14. In factand based upon the two strains studied, only this medium was significantly

0

20

40

60

80

100

0

20

40

60

80

100

M-11 M-12 M-13 M-14 M-15 M-16 M-17 M-18 M-20

Media

M-11 M-12 M-13 M-14 M-15 M-16 M-17 M-18 M-20

Media

M-11 M-12 M-13 M-14 M-15 M-16 M-17 M-18 M-20

M-11 M-12 M-13 M-14 M-15 M-16 M-17 M-18 M-20

Media

Media

Perc

en

tag

e o

f F

ISH

po

sit

ive

s c

ell

sP

erc

en

tag

e o

f F

ISH

po

sit

ives c

ell

s

60 min120 min

60 min120 min

60 min120 min

60 min120 min

A

C D

0

10

20

30

40

50

0

10

20

30

40

50

Mean

flu

ore

sc

en

ce

Mean

flu

ore

scen

ce

B

FIG. 2. FLOW CYTOMETRY RESULT FOR BACILLUS CEREUS ATCC 14579 (A, B) ANDHER 1414 (C, D) FOR FLUORESCENT IN SITU HYBRIDIZATION (FISH) REACTION BY

USING AN EUB338 PROBE IN SELECTED MEDIA (M)(A, C) The percentage of positive cells at 60 and 120 min. (B, D) Mean fluorescence intensity obtained

at 60 and 120 min.


equal or better in terms of the percentage of positive cells and mean fluores-cence intensity.

Step 2: Influence of Germination Factors

None of the germination factors tested gave a significant increase in thesignal obtained after 120 min of incubation (data not shown). Therefore, thefurther experiments were performed in medium 14 incubated at 37C, by usinga starting inoculum of 5 ¥ 106 spores/mL and without heat shock treatment asused in step 1.

Step 3: Optimal and Minimal Time of Permeabilization Treatments

Permeabilization with 50% ethanol, 4% paraformaldehyde and 1% form-aldehyde were compared. The two B. cereus strains were incubated in growthmedium 14 for 60 and 120 min as positive results can be observed using flowcytometry and fluorescence microscopy at those incubation periods. Figure 3Apresents the percentage of positive cells for the different permeabilizationtreatments after 60 and 120 min of incubation in growth medium 14 obtainedfor B. cereus ATCC 14579. Treatment with 50% ethanol required longerincubation time to be optimal (percentage of positive cells and mean fluores-cence) compared with 4% paraformaldehyde and 1% formaldehyde thathad almost reached their maximum after 15 min of incubation. The resultsobtained for the B. cereus HER 1414 were similar to those obtained forB. cereus ATCC 14579 (data not shown).

Using fluorescence microscopy, results obtained after 15 min of incuba-tion in 1% formaldehyde and 4% paraformaldehyde were still the same, butslightly better in terms of the quality of images in fluorescence microscopyfor 4% paraformaldehyde (data not shown). A permeabilization treatment of15 min in 4% paraformaldehyde was chosen for the rest of the experiments.

Step 4: Minimal Hybridization Time

The hybridization time used thus far was 2 h with the universal probe.Figure 4 presents the results for B. cereus ATCC 14579 and B. cereus HER1414 for the optimization of hybridization time. Comparable results wereobtained for the two strains for the percentage of positive cells (Fig. 4A) andfluorescence intensity (Fig. 4B). Reducing hybridization to 15 min and 30 mingave fewer positive cells and less intensity of fluorescence than with 1 and 2 hof hybridization. Equivalent results were obtained for a hybridization of 1 or2 h. A FISH signal was obtained in fluorescence microscopy with all theincubation periods (data not shown). Even if the results were slightly better


with 1 h of incubation, hybridization time of 15 min was kept for the rest ofthe experiment to reduce the overall time of the protocol to achieve a fastdetection.

Step 5: Validating the Optimal Condition Using Specific Probes

The next step was to validate the optimal conditions determined by usingthe specific probes for B. cereus pB394. Figure 5 shows the results for B.cereus ATCC 14579 and HER 1414 obtained with the pB394 probe (gray bar)compared with the EUB38 probe (black bar) for the percentage of positive

0

20

40

60

80

100

60min-

EtOH50%

60min-

Para4%

60min-

Form1%

120min-

EtOH50%

120min-

Para4%

120min-

Form1%

Treatment

60min-

EtOH50%

60min-

Para4%

60min-

Form1%

120min-

EtOH50%

120min-

Para4%

120min-

Form1%

Treatment

Perc

en

tag

e o

f F

ISH

po

sit

ve c

ells

15 min30 min60 min120 min4 h8 h18 h

0

5

10

15

20

25

30

35

Mean

flu

ore

scen

ce

A

B

15 min30 min60 min120 min4 h8 h18 h

FIG. 3. PERMEABILIZATION TREATMENT AND MINIMAL TIME OF PERMEABILIZATIONFROM THE BACILLUS CEREUS ATCC 14579 SPORE

The spores were incubated in culture medium 14 for 60 and 120 min before permeabilization with 50%ethanol (EtOH 50%), 4% paraformaldehyde (Para 4%) or 1% formaldehyde (Form 1%) for 15, 30,60, 120 min, 4, 8 and 18 h. (A) The percentage of positive cells obtained for fluorescent in situhybridization (FISH) reaction by using EUB338 probe. (B) Mean fluorescence intensity obtained for

FISH reaction by using an EUB338 probe.


cells and mean fluorescence intensity. Comparable results were obtained, byflow cytometry, for the percentage of positive cells (Fig. 5A), but differenceswere observed for the mean fluorescence intensity between the two stains ofB. cereus (Fig. 5B).

Figure 6 shows representative results with B. cereus ATCC 14579obtained in fluorescence microscopy comparing the optimal conditions.

0

20

40

60

80

100

15 min 30 min 1 hr 2 hrs

Hybridization time

15 min 30 min 1 hr 2 hrs

Hybridization time

Pe

rce

nta

ge

of

po

sit

ive

ce

lls

0

0.5

1

1.5

2

2.5

3

Me

an

flu

ore

sc

en

ce

A

B

FIG. 4. INFLUENCE OF HYBRIDIZATION TIME ON FLUORESCENT IN SITUHYBRIDIZATION (FISH) SIGNAL FOR BACILLUS CEREUS HER 1414 (BLACK BAR) AND

B. CEREUS ATCC 14579 (GRAY BAR)The spores were incubated for 60 min in culture medium 14, permeabilized in 4% paraformaldehydefor 15 min and hybridized for 15, 30 min, 1 and 2 h. (A) The percentage of positive cells obtained forFISH reaction by using an EUB338 probe. (B) Mean fluorescence intensity obtained for FISH reaction

by using an EUB338 probe.


Visible difference is obtained in the number of positive cells when hybridiza-tion time was decreased from 2 h (Fig. 6A) to 15 min (Fig. 6B) by usingEUB338 probes. Also, the percentage of positive cells observed by fluores-cence microscopy with the pB394 probe was lower compared with theEUB338 probe at 15 min of hybridization (Fig. 6C). Control experiments

0

5

10

15

20

25

30

35

40

45

50

B.cereus ATCC 14579

Perc

en

tag

e o

f p

osit

ive c

ells

A

B. cereus HER 1414

0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

Mean

flu

ore

scen

ce

B

B. cereus ATCC 14579 B. cereus HER 1414

FIG. 5. VALIDATION OF THE OPTIMAL CONDITION BY USING THE BACILLUSCEREUS-SPECIFIC PROBE PB394 (GRAY BAR) COMPARED WITH EUB338 PROBE(BLACK BAR) FROM B. CEREUS ATCC 14579 AND B. CEREUS HER 1414 SPORE

The spores were incubated for 60 min in culture medium 14, permeabilized in 4% paraformaldehydefor 15 min and hybridized for 15 min. (A) The percentage of positive cells obtained for fluorescentin situ hybridization (FISH) reaction by using a pB394 and EUB338 probe. (B) Mean fluorescence

intensity obtained for FISH reaction by using a pB394 and EUB338 probe.


A

B

C

FIG. 6. PHOTOMICROGRAPH OF THE BACILLUS CEREUS ATCC 14579 CELLSINCUBATED FOR 60 MIN IN CULTURE MEDIUM 14 (TRYPTIC SOY

BROTH + L-ALANINE + INOSINE)(A) The cells were permeabilized with 4% paraformaldehyde for 15 min followed by fluorescent in situhybridization (FISH) reaction for 2 h by using the EUB338 probe. (B) The cells were permeabilizedwith 4% paraformaldehyde for 15 min followed by FISH reaction for 15 min by using the EUB338probes. (C) The cells were permeabilized for 15 min by using 4% paraformaldehyde, and FISH

protocol was performed for 15 min by using the pB394 probe.


using antisense probe notEUB338 and notpB394 were still negative underfluorescence microscopy.

Step 6: Validating the Optimal Conditions in Spiked Milk Samples

Various concentrations of viable B. cereus HER 1414 spores were inocu-lated in sterile milk samples. Specific detection of 103 cfu/mL B. cereus sporesby using pB394 probes was possible within 2 h using the optimized protocol.The milk clarification protocol using sodium citrate did not affect FISHdetection. The antisense probe notpB394 gave no signal. Figure 7 shows a

A

B

FIG. 7. PHOTOMICROGRAPH OF BACILLUS CEREUS HER 1414 DETECTED FROM A MILKSAMPLE BY USING A PB394 PROBE

The milk sample was inoculated at a concentration of 103 cfu/mL, clarified by using sodium citrate andgrowth for 1 h in tryptic soy broth + L-alanine + inosine. The cells were permeabilized for 15 min byusing 4% paraformaldehyde, and fluorescent in situ hybridization protocol was performed for 15 min

by using a pB394 probe (A) and a notpB394 probe (B).


representative result. B. cereus ATCC 14579 gave the same result as HER 1414(data not shown).

DISCUSSION

In the present report, our optimized protocol leads to a FISH signal fromB. cereus spores in less than 2 h including treatment and handling. The con-dition retained was: incubation at 37C for 1 h in TSB + L-alanine 100 mM andinosine 10 mM, 15 min permeabilization in 4% paraformaldehyde followed by15 min of hybridization. A maximum of 30 min of manipulation between thesesteps is necessary: washings, transfers, addition of probes, preparations forflow cytometry or fluorescence microscopy.

This study shows that activation/germination and outgrowth of B. cereusspores facilitates fluorescence in situ hybridization protocols by reducing timeto signal when compared with the direct spores’ permeabilization. The FISHsignal intensity depends on the number and accessibility of probe targets onthe rRNA structure. Therefore, an enriched medium such as TSB reached thisgoal as vegetative cells reach a maximum of rRNA during the logarithmicgrowth phase. However, one study showed that an rRNA concentration forBacillus megaterium spores are at the same level as log phase cells (Chambonet al. 1968). Considering this information, permeabilization of the spores andaccessibility of the targets is probably more an issue here than the number oftargets (rRNA) for FISH on Bacillus sp. spores. Previous authors have shownthat it is possible to do FISH directly on spores without germination andoutgrowth (Fischer et al. 1995). However, the proposed protocol needs a totalof 3 days of incubation and the use of successive combinations of chemicalsthat can limit the application on environmental samples. Regamey et al.showed that 6–8 h of treatment and handling are required to obtain a FISHsignal by using spore germination (Regamey et al. 2000). In the presence ofgerminants like L-alanine and inosine, those small molecules penetrate theouter layers, coat and cortex of the spore to bind to specific receptors locatedon the inner membrane (Setlow 2003; Moir 2006). This initiates the germina-tion of spores that is an irreversible phenomenon. The presence of thosegerminant in the culture media improves the time to detection performedby Regamey et al.

A study shows that a mixture of L-alanine and inosine efficiently stimu-lates the germination of adhered spores, resulting in 3.2 decimal logarithmsof germination (Hornstra et al. 2007). The result obtained in this study isin accordance with the observations of Hornstra et al. that the mixture ofL-alanine and inosine is the strongest combination of germinants observed forB. cereus. We experimented minimal media composed of buffer (Tris, PBS)


plus L-alanine and inosine (Table 1). After 2 h of incubation in those buffer-based media, no positive results were observed with FISH in flow cytometryand fluorescence microscopy. This observation demonstrated that germination/outgrowth of B. cereus spore is necessary to achieve a FISH signal by using theconditions of this study. This is not surprising as complete germination isnecessary to remove the different layers that shield the spore. Those buffer-based media might be suitable for germination induction, but not for support-ing growth. In fact, the germination induction of Bacillus spores is a rapidphenomenon, and amino acid synthesis occurs after 2–5 min after the sporesare put in a buffer containing only the germinating agents (Garrick-Silversmithand Torriani 1973; Welkos et al. 2004). However, as the objective was toobtain a rapid FISH signal, those media were not suitable for our purpose.

The permeabilization treatments tested were standard in FISH studies.The envelope of B. cereus cells differs greatly depending on the growth stageof the cell. The challenge of the permeabilization procedure is to enable theFISH probe to enter the cell in order to reach its target without disrupting thecell’s morphology or lose its cytoplasmic content. There are no universal cellpermeabilization protocols for FISH. This study shows that 50% ethanol needsmore incubation time (minimum 8 h) to be fully efficient compared with 1%formaldehyde and 4% paraformaldehyde. The latter two are similar in terms ofefficiency (15 min of treatment is sufficient) and mechanisms of action. In fact,ethanol is a precipitating agent, and formaldehyde and paraformaldehyde arecross-linking agents (Moter and Göbel 2000). Thus, a cross-linking agent ismore suitable with vegetative B. cereus cells for fast and efficient permeabi-lization. This result is crucial for a rapid in situ detection protocol.

A reduction of the time of hybridization was also evaluated. The resultsshowed that a hybridization of 1 h versus 2 h gave the same FISH result.Reducing the hybridization to 30 min and 15 min slightly decreases the FISHsignal in flow cytometry. This result was confirmed by fluorescence micros-copy. However, a good and acceptable detection was reached after 15 min ofhybridization by using both flow cytometry and fluorescence microscopy. It isdifficult to draw a line when a signal is considered too weak, but after 15 minof hybridization by using the optimized conditions, the FISH signal is clearlydetectable.

The B. cereus-specific probe pB394 in this study was originally validatedin a microchip system of detection (Liu et al. 2001). This probe binds avariable region in the 16S rRNA. When used with the optimized conditionsdetermined in this study, the pB394 shows comparable results for the detectionwith flow cytometry in terms of the percentage of positive cells comparedwith the universal probe for eubacteria EUB338, but presents a reductionin the fluorescence intensity. When the results of pB394 are observed byfluorescence microscopy, the percentage of positive cells and the fluorescence


intensity are weaker when compared with those of the universal probeEUB338. The difference in the percentage of positive cells obtained betweenflow cytometry (pB394 similar to EUB338) and the fluorescence microscopy(pB394 result in less positive cells number than EUB338) can be explained bythe fact that flow cytometry is equipped with a photomultiplier that renders thisinstrument more sensitive to low signal compared with fluorescence micro-scopy. Because the fluorescence intensity is weaker with pB394 than withEUB338, the fluorescence microscopy does not allow the observation ofpositive cells with the lowest fluorescence intensity. This difference in theintensity of fluorescence between the universal and specific probes can beexplained, in part, by the accessibility of the probe to the target. In a previousstudy conducted by Fuchs et al., the accessibility of FISH probes that targetedthe 16S rRNA was performed in flow cytometry by using the mean fluores-cence intensity (Fuchs et al. 1998). The three-dimensional structure of theribosome compromises the access of some targets for the probes, thus reducingthe signal. The pB394 probe targeted the position 162–185 of the 16S rRNAaccording to Escherichia coli numbering (Brosius et al. 1981). At this posi-tion, the study of Fuchs et al. has shown that this target gave low–moderatefluorescence signal (category IV of VI) whereas a category I gave the strongestsignal. In comparison, EUB338 probe targeted a position overlapping catego-ries II and III.

The germination of spores results in an immediate loss of the spores’resistance that makes the permeabilization more efficient. Our findings openthe possibility of including sporulating bacteria in overall taxonomic distri-bution assessment studies and in detecting specific agents with the FISHapproach. This study shows that a brief exposure of Bacillus spores to TSBplus L-alanine and inosine makes them more permeable to be fixed with 4%paraformaldehyde and enables probe binding for standard FISH protocols.With this new rapid germination method and optimized FISH protocol,B. cereus spores present in a milk sample can be detected in 2 h.

ACKNOWLEDGMENTS

These studies were undertaken under a Department of National Defense(DND) contract W7702-00R802/001/EDM: Dr. J. Ho was the scientificauthority acting for DND. The authors are thankful to Mr. Luc Trudel and Dr.Benjamin Nehmé for their insightful review of the manuscript and thank SergeSimard for statistical analysis. Christian Laflamme received a Natural Sciencesand Engineering Research Council of Canada (NSERC)/Institut de rechercheRobert-Sauvé en santé et sécurité du travail (IRSST) studentship. Dr. C.Duchaine acknowledges IRSST/Canadian Institutes of Health Research


(CIHR) and Fonds de la Recherche en Santé du Québec (FRSQ) Junior 2Scholarships and a NSERC Discovery Grant.

REFERENCES

AHMED, R., SANKAR-MISTRY, P., JACKSON, S., ACKERMANN, H.-W.and KASATIYA, S.S. 1995. Bacillus cereus phage typing as an epide-miological tool in outbreaks of food poisoning. J. Clin. Microbiol. 33,636–640.

AMANN, R.I., BINDER, B.J., OLSON, R.J., CHISHOLM, S.W.,DEVEREUX, R. and STAHL, D.A. 1990. Combination of 16S rRNA-targeted oligonucleotide probes with flow cytometry for analyzing mixedmicrobial populations. Appl. Environ. Microbiol. 56, 1919–1925.

BARLASS, P.J., HOUSTON, C.W., CLEMENT, M.O. and MOIR, A. 2002.Germination of Bacillus cereus spores in response to L-alanine and toinosine: The roles of gerL and gerQ operons. Microbiology 148, 2089–2095.

BROSIUS, J., DULL, T.J., SLEETER, D.D. and NOLLER, H.F. 1981. Geneorganization and primary structure of a ribosomal RNA operon fromEscherichia coli. J. Mol. Biol. 148, 107–127.

CAIPO, M.L., DUFFY, S., ZHAO, L. and SCHAFFNER, D.W. 2002. Bacillusmegaterium spore germination is influenced by inoculum size. J. Appl.Microbiol. 92, 879–884.

CHAMBON, P., DEUTSCHER, M.P. and KORNBERG, A. 1968. Biochemi-cal studies of bacterial sporulation and germination. X. Ribosomes andnucleic acid of vegetative cells and spore of Bacillus megaterium. J. Biol.Chem. 243, 5110–5116.

CHRISTIANSSON, A. 2003. Bacillus cereus. In Encyclopedia of DairySciences (H. Roginsky, J.W. Fuquay and P.F. Fox, eds.) pp. 123–128,Academic Press, Amsterdam, The Netherlands.

CLEMENTS, M.O. and MOIR, A. 1998. Role of the gerI operon of Bacilluscereus 569 in the response of spores to germinants. J. Bacteriol 180,6729–6735.

DRIKS, A. 2002. Maximum shields: The assembly and function of thebacterial coat. Trends Microbiol. 10, 251–254.

ENEROTH, A., SVENSSON, B., MOLLIN, G. and CHRISTIANSSON, A.2001. Contamination of pasteurized milk by Bacillus cereus in the fillingmachine. J. Dairy Res. 68, 186–196.

FAILLE, C., JULLIEN, C., FONTAINE, F., BELLON-FONTAINE, M.N.,SLOMIANNY, C. and BENEZECH, T. 2002. Adhesion of Bacillus


spores and Escherichia coli cells to inert surfaces: Role of surface hydro-phobicity. Can. J. Microbiol. 48, 728–738.

FISCHER, K., HAHN, D., HÖNERLAGE, W., SCHÖNHOLZER, F. andZEYER, J. 1995. In situ detection of spore and vegetative cells ofBacillus megaterium in soil by whole cell hybridization. System Appl.Microbiol. 18, 265–273.

FUCHS, B.M., WALLNER, G., BEISKER, W., SCHWIPPL, I., LUDWIG, W.and AMANN, R. 1998. Flow cytometric analysis of the in situ accessi-bility of Escherichia coli 16S rRNA for fluorescently labelled oligonucle-otide probes. Appl. Environ. Microbiol. 64, 4973–4982.

FUKUSHIMA, H., KATSUBE, K., HATA, Y., KISHI, R. and FUJIWARA, S.2007. Rapid separation and concentration of food-borne pathogens infood samples prior to quantification by viable-cell counting and real-timePCR. Appl. Environ. Microbiol. 73, 92–100.

GARRICK-SILVERSMITH, L. and TORRIANI, A. 1973. Macromolecularsyntheses during germination and outgrowth of Bacillus subtilis spores.J. Bacteriol 114, 507–516.

GRANUM, P.E. and LUND, T. 1997. Bacillus cereus and its food poisoningtoxins. FEMS Microbiol. Lett. 157, 223–228.

GRIFFITHS, M.W. 1992. Bacillus cereus in liquid milk and other milk prod-ucts. Bull. Int. Dairy Fed. 275, 36–39.

HOLT, J.G. and KRIEG, N.R. 1994. Growth, enrichment and isolation. InMethods for General and Molecular Bacteriology (P. Gerhardt, R.G.E.Murray, W. Wood and N.R. Krieg, eds.) pp. 179–215, American Societyfor Microbiology, Washington, DC.

HORNSTRA, L.M., DE VRIE, Y.P., DE VOS, W.M., ABEE, T. and WELL-BENNIK, M.H. 2005. gerR, a novel ger operon involved in L-alanineand iosine-initiated germination of Bacillus cereus ATCC 14579. Appl.Environ. Microbiol. 71, 774–781.

HORNSTRA, L.M., DE VRIES, Y.P., WELLS-BENNIK, M.H.J., DE VOS,W.M. and ABEE, T. 2006. Characterization of germination receptors ofBacillus cereus ATCC 14579. Appl. Environ. Microbiol. 72, 44–53.

HORNSTRA, L.M., DE LEEUW, P.L., MOEZELAAR, R., WOLBERT, E.J.,DE VRIES, Y.P. and ABEE, T. 2007. Germination of Bacillus cereusspores adhered to stainless steel. Int. J. Food Microbiol. 116, 367–371.

IRELAND, J.A.W. and HANNA, P.C. 2002. Amino acid- and purineribonucleoside-induced germination of Bacillus anthracis Dsterneendospore: gerS mediates responses to aromatic ring structures. J.Bacteriol 184, 1296–1303.

KAMAT, A.S., LEWIS, N.F. and PRADHAN, D.S. 1985. Mechanism of Ca2+

and dipicolinic acid requirement for L-alanine induced germinationBacillus cereus BIS-59 spores. Microbios 44, 33–44.


KEYNAN, A. and EVENCHIK, Z. 1969. Activation. In The Bacterial Spore(G.W. Gould and A. Hurst, eds.) pp. 359–396, Academic Press, NewYork.

KOSHIKAWA, T., YAMAZAKI, M., YOSHIMI, M., OGAWA, S.,YAMADA, A., WATABE, K. and TORII, M. 1989. Surface hydropho-bicity of spores of Bacillus spp. J. Gen. Microbiol. 135, 2717–2722.

LAFLAMME, C., LAVIGNE, S., HO, J. and DUCHAINE, C. 2004. Assess-ment of bacterial endospore viability with fluorescent dyes. J. Appl.Microbiol. 96, 684–692.

LAFLAMME, C., HO, J., VEILLETTE, M., DE LATRÉMOILLE, M.C.,VERREAULT, D., MÉRIAUX, A. and DUCHAINE, C. 2005. Flowcytometry analysis of germinating Bacillus spore, using membranepotential dye. Arch. Microbiol. 183, 107–112.

LIU, W.T., MIRZABEKOV, A.D. and STAHL, D.A. 2001. Optimization of anoligonucleotide microchip for microbial identification studies: A non-equilibrium dissociation approach. Environ. Microbiol. 3, 619–629.

LUCORE, L.A., CULLISON, M.A. and JAYKUS, L.A. 2000. Immobilizationwith metal hydroxides as a mean to concentrate food-borne bacteria fordetection by cultural and molecular methods. Appl. Environ. Microbiol.66, 1769–1776.

MOIR, A. 2006. How do spore germinate? J. Appl. Microbiol. 101, 526–530.MOTER, A. and GÖBEL, U.B. 2000. Fluorescence in situ hybridization

(FISH) for direct visualization of microorganisms. J. Microbiol. Methods41, 85–112.

ODA, Y., SLAGMAN, S.J., MEIJER, W.G., FORNEY, L.J. andGOTTSCHAL, J.C. 2000. Influence of growth rate and starvation onfluorescent in situ hybridization of Rhodopseudomonas palustris. FEMSMicrobiol. Ecol. 32, 205–213.

PAIDHUNGAT, M. and SETLOW, P. 2002. Spore germination and out-growth. In Bacillus subtilis and Its Relatives : From Genes to Cells (J.A.Hosh, R. Losick and A.L. Sonenshein, eds.) pp. 537–548, AmericanSociety for Microbiology, Washington, DC.

PAIDHUNGAT, M., RAGKOUSKI, K. and SETLOW, P. 2001. Geneticrequirements for induction of germination of spores of Bacillus subtilisby Ca2+-dipicolinate. J. Bacteriol 183, 4886–4893.

PARK, S.H., KIM, H.J., KIM, J.H., KIM, T.W. and KIM, H.Y. 2007. Simul-taneous detection and identification of Bacillus cereus group bacteriausing multiplex PCR. J. Microbiol. Biotechnol. 17, 1177–1182.

PERRY, L., HEARD, P., KANE, M., KIM, H., SAVIKHIN, S., DOMIN-GUEZ, W. and APPLEGATE, B. 2007. Application of multiplex poly-merase chain reaction to the detection of pathogens in food. J RapidMethods Autom. Microbiol. 15, 176–198.


QUARTO, M. and CHIRONNA, M. 2005. Hepatitis A: Sources in food andrisk for health. In Review in Food and Nutrition Toxicity. Vol. 2 (V.R.Preedy and R.R. Watson, eds.) pp. 91–126, CRC Press, London, U.K.

RAGKOUSI, K., EICHENBERGER, P., VAN OOJI, C. and SETLOW, P.2003. Identification of a new gene essential for germination of Bacillussubtilis spores with Ca2+-dipicolinate. J. Bacteriol. 185, 2315–2329.

RASO, J., MARCELA GONGORA-NIETO, M., BARBOSA-CANOVAS,G.V. and SWANSON, B.G. 1998. Influence of several environmentalfactors of germination and inactivation of Bacillus cereus by high hydro-static pressure. Int. J. Food Microbiol. 44, 125–132.

REGAMEY, A., HARRY, E.J. and WAKE, R.G. 2000. Mid-cell Z ring assem-bly in the absence of entry into the elongation phase of the round ofreplication in bacteria: Co-ordinating chromosome replication with celldivision. Mol. Microbiol. 38, 423–434.

SETLOW, P. 1994. Mechanisms which contribute to the long-term survival ofspores of Bacillus species. J. Appl. Bacteriol. Suppl. 76, 49S–60S.

SETLOW, P. 2003. Spore germination. Curr. Opin. Microbiol. 6, 550–556.SWERDLOW, B.M., SETLOW, B. and SETLOW, P. 1981. Level of H+ and

other monovalent cations in dormant and germinating spores of Bacillusmegaterium. J. Bacteriol. 148, 20–29.

WAGNER, M., HORN, M. and DAIMS, H. 2003. Fluorescence in situ hybrid-ization for the identification and characterization of prokaryotes. Curr.Opin. Microbiol. 6, 302–309.

WELKOS, S.L., COTE, C.K., REA, K.M. and GIBBS, P.H. 2004. A microtiterfluorometric assay to detect the germination of Bacillus anthracis sporesand the germination inhibitory effects of antibodies. J. Microbiol.Methods 56, 253–265.


Documents

RAPID DETECTION OF GERMINATING BACILLUS CEREUS CELLS USING FLUORESCENT IN SITU HYBRIDIZATION