Influence of Glutamate on Growth, Sporulation, and Spore Properties of Bacillus cereus ATCC 14579 in Defined Medium

A chemically defined medium in combination with an airlift fermentorsystem was used to study the growth and sporulation of Bacilluscereus ATCC 14579. The medium contained six amino acids andlactate as the main carbon sources. The amino acids were depletedduring exponential growth, while lactate was metabolized mainlyduring stationary phase. Two concentrations of glutamate wereused: high (20 mM; YLHG) and low (2.5 mM; YLLG). Under bothconditions, sporulation was complete and synchronous. Sporulationstarted and was completed while significant amounts of carbonand nitrogen sources were still present in the medium, indicatingthat starvation was not the trigger for sporulation. Analysisof amino acids and NH₄⁺ in the culture supernatant showed thatmost of the nitrogen assimilated by the bacteria was taken upduring sporulation. The consumption of glutamate depended onthe initial concentration; in YLLG, all of the glutamate wasused early during exponential growth, while in YLHG, almostall of the glutamate was used during sporulation. In YLLG, butnot in YLHG, NH₄⁺ was taken up by the cells during sporulation.The total amount of nitrogen used by the bacteria in YLLG wasless than that used by the bacteria in YLHG, although a significantamount of NH₄⁺ was present in the medium throughout sporulation.Despite these differences, growth and temporal expression ofkey sigma factors involved in sporulation were parallel, indicatingthat the genetic time frames of sporulation were similar underboth conditions. Nevertheless, in YLHG, dipicolinic acid productionstarted later and the spores were released from the mother cellsmuch later than in YLLG. Notably, spores had a higher heat resistancewhen obtained after growth in YLHG than when obtained aftergrowth in YLLG, and the spores germinated more rapidly and completelyin response to inosine, L-alanine, and a combination of thesetwo germinants.

INTRODUCTION

Top

Introduction

Bacillus cereus is a gram-positive, facultative anaerobic rod-shapedbacterium able to form spores. It is a ubiquitous bacteriumfound in soil and in many raw and processed foods, such as rice,milk and dairy products, spices, and vegetables (8, 12, 20,44). Many strains of B. cereus are able to produce toxins andcause distinct types of food poisoning (19, 31). Concerns overB. cereus contamination have increased over the past few yearsbecause of the rapidly expanding market of chilled foods thatmay be pasteurized but still contain viable spores (8, 20, 34).Spores from B. cereus can germinate and outgrow during storage,even at low temperatures (8, 11, 20). To address this increasingproblem, major efforts focus on determining the causes of sporeresistance and the mechanisms of germination.

It has been well established that bacterial spore propertiesare affected by the conditions during sporulation (1, 17, 18,33, 41). In most studies, spores are routinely produced fromfortified agar or rich liquid media, which results in heterogeneoussporulation conditions for the individual cells. This preventscareful analysis of the metabolism during growth and sporulation.Recent studies describing the effect of sporulation conditionson spore properties involved modulation of sporulation temperature(1, 18, 33, 41, 42) or compared spores produced from differentmedia (9, 32).

Studies employing defined conditions and media to link substrateuse with sporogenesis are relatively rare; moreover, the effectof carbon sources on sporulation has not been studied systematicallyin recent decades. The most common carbon source used in sporulationmedia is glucose, but in natural conditions, B. cereus may encounterrather different substrates, such as lactate. We have previouslyshown that B. cereus is able to metabolize lactate (14), whichis formed in the natural environment by the fermentation ofa variety of naturally occurring polymers, such as lactose indairy or plant sugars in silage. Indeed, silage is a known sourceof B. cereus contamination of milk (47). Furthermore, in contrastto glucose, lactate does not cause catabolite repression orrepression of the tricarboxylic acid (TCA) cycle, which is requiredfor activation of the Spo0A phosphorelay, which in turn leadsto sporulation (25). Therefore, we investigated the growth andsporulation of B. cereus ATCC 14579, which has been characterizedat the genome level (26), on a chemically defined medium withlactate instead of glucose as the main carbon source. Secondly,because glutamate has been reported to have a large impact onsporulation as well as spore properties of bacilli (7, 10, 29),we used two different concentrations of glutamate: low (2.5mM; YLLG) and high (20 mM; YLHG).

MATERIALS AND METHODS

Top

Materials and Methods

Strains, medium, and fermentation.
B. cereus strain ATCC 14579 was obtained from the American TypeCulture Collection and used throughout this study. It was cultivatedin a chemically defined medium described previously (Table 1)(14). Fermentation conditions and inoculation were as describedbefore (14). The spores were harvested and washed as describedpreviously and were stored in 10 mM KPO₄, pH 7, with 0.1% Tween80 to prevent clumping and attachment of spores to plastic laboratoryequipment. The addition of Tween 80 had no effect on spore heatresistance.

View this table:
[in this window]
[in a new window]
TABLE 1. Medium compositions of YLHG and YLLG

Analytical procedures.
Concentrations of D- and L-lactic acid and glucose in culturesupernatant were determined by the UV method with enzymaticbioanalysis kits from Boehringer Mannheim (Darmstadt, Germany).Phase-contrast microscopy was performed with an Axioskop microscopefrom Zeiss. Amino acids in culture supernatant, cells, and sporeswere determined with the PicoTaq high-performance liquid chromatographymethod (5). The amino acids and NH₄⁺ from culture supernatantwere derivatized with the AccQ-Fluor kit from Waters (Etten-Leur,The Netherlands) according to manufacturer's instructions. Forcell and spore analyses, samples were autoclaved and acid hydrolyzedprior to derivatization.

Dipicolinic acid (DPA) was measured with a fluorescent assaymodified from Hindle and Hall (23). Spores or cells were suspendedin Tris-NaCl-Tw buffer (Tris-HCl [10 mM, pH 7.5] buffer with10 mM NaCl and 0.1% Tween 80), destroyed by autoclaving (121°Cfor 15 min), a treatment known to release all DPA (27), andpelleted at maximum speed in an Eppendorf centrifuge for 2 min.The supernatant was assayed for DPA as follows: 80 µlof supernatant was used in a final volume of 200 µl Tris-NaCl-Twbuffer with freshly added TbCl₃ at a final concentration of100 µM, and the fluorescence was measured with the Safiremicroplate reader from Tecan (Salzburg, Austria) in combinationwith XFLUOR4 software, version 4.40. Standard curves of DPAwith this method showed straight lines (R² > 0.99) up to150 µM of DPA with this Tb³⁺ concentration, with a detectionlimit of about 0.5 µM. The settings of the machine wereas follows: the excitation wavelength was 272 nm, the emissionwavelength was 547 nm, the bandwidth excitation and emissionwavelengths were 7.5 nm, the lag time was 10 µs, and theintegration time was 2,000 µs.

Spore properties.
For the heat resistance assay, 100 µl spores suspendedat an A₆₀₀ of 0.7 to 1.0 in Tris-NaCl-Tw buffer was sealed inthe tip of a Pasteur pipette, placed in an oil bath calibratedto the designated temperature, and cooled after a set time inice-cold water. Because of the small diameters of the Pasteurpipette tips and the small volume of spore suspension, we assumethat the heating and cooling of the spore suspensions were instantaneous.Serial dilutions were made in Tris-NaCl-Tw buffer and platedonto diluted nutrient broth (2.6 g/liter; Difco) solidifiedwith 1.5% agar. Colonies were counted after overnight incubationat 30°C.

The density of the whole spore was measured with a Percoll gradientaccording to the method described by Tisa et al. (48). Sporeswere concentrated in 0.15 M NaCl, added to 90% (vol/vol) Percoll(Amersham Pharmacia Biotech) solution with 0.15 M NaCl, andcentrifuged for 20 min at 30,000 x g. Whole-spore densitieswere derived by comparison of the positions of the spores andof density marker beads (Amersham Pharmacia Biotech) in theself-established gradient. The spore core density was determinedwith Nycodenz (Nygaard, Oslo, Norway) density gradients (43)according to the method described by Lindsay et al. (30) asfollows: spores were permeabilized with a decoating solutioncontaining 0.1 M dithiothreitol, 0.1 M NaCl, and 0.5% sodiumdodecyl sulfate according to the method described by Vary (51).Permeabilized spores were equilibrated in 30% Nycodenz (1.159g/ml). Continuous gradients were made by letting discontinuousgradients diffuse overnight at 5°C (43) or for a few hoursat room temperature. After centrifugation in a swing-out rotorfor 45 min at 3,700 rpm in a Centaur 2 centrifuge, the sporesformed a band at a certain position. The refractive index, whichwas linearly correlated with the density, from samples takenjust above and below the band was measured with a precisionrefractometer, and the average yielded the spore core density.

Spore germination was measured by the drop in A₆₀₀ of sporesuspensions at 30°C by using a Safire microplate readerfrom Tecan (Salzburg, Austria) in combination with XFLUOR4 software,version 4.40. Spores were suspended at an A₆₀₀ of 0.7 to 1.0in Tris-NaCl-Tw buffer, and after the addition of the germinants,the A₆₀₀ was followed automatically with intermittent shakingto prevent settling of the spores.

RT-PCR.
Real-time reverse transcription (RT)-PCR was used to monitorthe expression of the genes for key sigma factors (sigA, sigB,sigF, and sigG) and performed as described earlier (14). Thelevel of expression was calibrated by using the positive control,sigA expression, as a reference. All the samples in this communicationwere handled in exactly the same way to enable a good comparisonof the gene expressions in the two conditions we tested. Real-timeRT-PCRs were carried out in duplicate, and the duplicates differedless than 5% in all cases.

RESULTS AND DISCUSSION

Top

Results and Discussion

Carbon and nitrogen use during growth and sporulation.
B. cereus was able to grow and sporulate in the chemically definedmedium devoid of glucose. With lactate as the main carbon source,the exponential growth phases of B. cereus with low and highglutamate, YLLG and YLHG, respectively, were identical (Fig.1A). The A₆₀₀ curves and pH curves ran exactly parallel up tothe point where the glutamate in YLLG is used up (Fig. 1B),while during growth in YLHG, glutamate was barely consumed atthis time. From the moment of glutamate depletion in YLLG, theabsorbance and pH differed between cultures grown in YLLG orYLHG, although lactic acid was metabolized in a parallel mannerfor both media (Fig. 1B). During the growth of culture in YLHG,the extra rise in pH coincided with glutamate consumption, indicatingthat it was a result of glutamate metabolism.

View larger version (25K):
[in this window]
[in a new window]
FIG. 1. Growth and sporulation of B. cereus cultures in YLHG (closed symbols) and YLLG (open symbols). (A) Circles, A₆₀₀; squares, pHs. (B) Triangles, L-lactate; inverted triangles, D-lactate; diamonds, DPA; circles, glutamate. The dashed line indicates the moment of upregulation of sigG transcription.

The higher A₆₀₀ that was reached during the growth of culturein YLHG suggests that many more cells are present in this culturethan in the culture grown in YLLG. However, neither plate countsnor direct microscopic observations confirmed that this wasthe case, although plate counts were difficult because the cellsformed large clusters at the end of exponential growth, as weobserved before (14). It is possible that the A₆₀₀ is influencedby the pH, because we noticed that in the medium that we useda high pH (above 8.5) results in some precipitation, presumablyof phosphate with calcium or magnesium. Furthermore, a darkbrown color developed in the supernatant during sporulationin YLHG but not in YLLG, further raising the A₆₀₀. This makesit plausible that although the final A₆₀₀ reached with the culturegrown in YLHG was higher, the actual cell density was quitesimilar to that of the culture grown in YLLG. Indeed, the totalbiomasses, as measured by nitrogen contents, were similar forthe two cultures at the end of sporulation, both cultures producedsimilar amounts of DPA, sporulation frequency was >99% inboth cases, and the cleaned spores had similar DPA contents(Table 2). This confirmed that, indeed, equal amounts of cellswere present in the cultures in YLLG and YLHG.

View this table:
[in this window]
[in a new window]
TABLE 2. Properties of spores from YLLG and YLHG^a

We initially expected the maximum growth rate of B. cereus inthe medium with lactate to be lower than that of B. cereus inthe same medium with the much more favorable substrate, glucose,but the rates were identical (about 0.45 h^–1). However,the lag phase was much longer without glucose. This indicatesthat the exponential growth rate is primarily determined bythe amino acids in the medium. Indeed, during exponential growth,the bacteria metabolized the amino acids present in the medium,while lactate was used upon entry into stationary phase (Fig.1 and 2). The metabolism of a 10 mM concentration of amino acidsduring exponential growth was accompanied by a 10-mmol risein NH₄⁺ in the supernatant of both cultures, indicating thatB. cereus used the amino acids as a carbon source and fuel,while most of the lactic acid was metabolized after the cellshad entered stationary phase. These findings confirm the anticipatedpreference of B. cereus for amino acids, as derived from genomeanalysis studies (26). In the presence of either glucose orlactate, the amino acids were depleted in the same order: threoninewas first, then valine, histidine, and leucine, and methioninewas the last (data not shown). The moment of glutamate consumptiondepended on the initial concentration; B. cereus metabolizedglutamate earlier in YLLG than in YLHG, and during cultivationin YLLG, glutamate was depleted early, after threonine. In YLHG,the concentration of glutamate started to decrease 3 h later,and glutamate was the last amino acid to be metabolized.

View larger version (21K):
[in this window]
[in a new window]
FIG. 2. Nitrogen sources during growth and sporulation of B. cereus cultures in YLHG (A) and YLLG (B). Symbols: circles, A₆₀₀; squares, total nitrogen; diamonds, nitrogen in glutamate; inverted triangles, nitrogen in amino acids other than glutamate; triangles, NH₄. The dashed line indicates the moment of upregulation of sigG transcription.

Because the metabolism of amino acids during exponential growthresulted in a nearly equimolar increase of NH₄⁺ in the supernatant,the total amount of nitrogen assimilated by B. cereus was verylow during this growth phase. Indeed, our measurements clearlyindicated that most of the nitrogen assimilated by B. cereusis taken up during sporulation (Fig. 2), underscoring the highnitrogen requirement of sporulating cells. The total concentrationsof nitrogen initially present were 40 mM in YLHG and 22.5 mMin YLLG, a significant part of which (10 mM) consisted of NH₄⁺,which was routinely added to the media as (NH₄)₂SO₄ (Table 1).When cultivated in YLHG, B. cereus used the bulk of the glutamatein the stationary and sporulation phase, after the other aminoacids had disappeared. During this phase, the NH₄⁺ concentrationin the supernatant remained 20 mM, indicating that the nitrogenin the glutamate was completely assimilated. When cultivatedin YLLG however, B. cereus assimilated NH₄⁺ during the stationaryand sporulation phase.

At the end of sporulation, a total of 20 mM of nitrogen haddisappeared from the culture supernatant in YLHG, of which wecould trace back 16 mM as cell-bound nitrogen (Table 2). Theremaining 4 mM might be present in the brown pigment that wasformed in the culture supernatant during sporulation, as a brownpigment produced by B. subtilis during sporulation was shownto contain a significant proportion of nitrogen (3). The culturegrown in YLLG had assimilated 15 mM of nitrogen in total atthe end of sporulation, all of which we could trace back ascell-bound nitrogen (Table 2).

In both media tested, 99% of the cells produced a phase-brightspore and the concentrations of DPA reached about 115 µM,after which the oxygen consumption of the cultures decreaseddramatically (data not shown). DPA production and phase brighteningof the forespores are indicative of the final phases of sporulation(16). Notably, DPA production and the release of the sporesfrom the mother cells started much earlier after cultivationin YLLG than in YLHG. It is generally assumed that sporulationis a response to starvation (45). In our experiments, however,significant amounts of carbon (D-lactate and glutamate) andnitrogen (NH₄⁺ and glutamate) were present when the cells startedto sporulate. Even after sporulation was completed, a significantamount of these C and N sources remained. Nutrient limitationtherefore seems an unlikely trigger for sporulation under theconditions we employed. From B. subtilis, it is known that inaddition to starvation, quorum sensing can be involved in theinitiation of sporulation (46). In silico analysis of the B.cereus ATCC 14579 genome with the ERGO genome and discoverysystem (36) revealed the presence of the gram-positive quorum-sensingand competence II pathway which is also present in B. subtilis.Thus, although additional experiments in which cell densityis carefully manipulated are needed to provide definite evidence,it is tempting to speculate that in our experiments, cell densityrather than nutrient limitation may have been the factor thattriggered the cells to sporulate. Because extra glutamate didnot result in extra biomass, the cell densities in YLHG andYLLG were similar at all times. Indeed, sporulation as judgedby the rise in sigF and sigG transcription started at the samemoment under both conditions (Fig. 3) (see below).

View larger version (24K):
[in this window]
[in a new window]
FIG. 3. Sigma factor expression levels (calculated as described in the text) of B. cereus bacteria grown and sporulated in YLHG (closed symbols) and YLLG (open symbols). (A) Expression levels of sigA (circles) and sigB (squares). (B) Expression levels of sporulation sigma factors sigF (triangles) and sigG (squares). Diamonds, DPA concentrations. The dashed line indicates the moment of upregulation of sigG transcription.

Gene expression.
The differences we found in the nitrogen metabolism and timingof DPA formation prompted us to check for differences in thegenetic programs of sporulation in YLLG and YLHG. Therefore,we measured the expression of genes for four key sigma factorsinvolved in stress response, growth, and sporulation: sigB,sigA, and sigF and sigG, respectively (Fig. 3). The productof sigA, sigma factor ${sigma}$ ^A, is the primary sigma factor. It regulatesmacromolecular synthesis, plays a key role in the housekeepingfunctions of the cell (22), and is expressed by B. cereus ata high level throughout growth and sporulation (14); therefore,we used sigA expression as a reference and a positive control.During cultivation in YLLG and YLHG, sigB transcription levelswere high compared to the levels of transcription during growthwith glucose (14) and showed no clear peak upon entry into thestationary phase (Fig. 3A). It was reported that sigB transcriptionis increased upon entry in stationary phase (14, 22) and whenthe cell encounters stress (50), directly affecting cell andcarbon metabolisms (40). Our present findings indicate thatB. cereus experiences more stress when grown in a medium withlactate instead of glucose and that the sigB peak observed before(14) was caused by a shift from glucose metabolism to othercarbon sources rather than by changes in growth rate upon entryinto the stationary phase.

During the cultivations in YLLG and YLHG, the expression patternsof sigF and sigG (Fig. 3B) were comparable to those on glucose(14) in that we observed a high peak of sigF transcription,followed by a sigG peak. However, with lactate as a carbon source,sigF expression showed a sharp peak, while during sporulationwith glucose, we observed a broad peak (14). The product ofsigF, sporulation sigma factor ${sigma}$ ^F, is produced in the predivisionalcell (37) well before the first irreversible step of sporulation(15). After septation, when the cells have become committedto sporulate, ${sigma}$ ^F is released from its anti-sigma factor to becomeactive in the forespore only (37). The gene encoding ${sigma}$ ^G, sigG,is under the control of ${sigma}$ ^F and expressed in the forespore compartmentjust after the appearance of the sporulation septum (15). Thus,an increase of sigG transcription indicates ${sigma}$ ^F activity. ${sigma}$ ^G playsa role in the final stages of sporulation, regulating the synthesisof SASP and Ger proteins (2, 22, 24). Interestingly, the timesat which the maximum sigF and sigG expressions occurred werealmost the same under both conditions, with only a slight delayof 30 min for the YLHG cultivation. This suggested that therewas no considerable difference in the programming of sporulation,judging by the temporal expression of the key sigma factorswe measured. Nevertheless, B. cereus started to produce DPAand phase-bright spores in YLLG more than 3 h earlier than inYLHG, indicating that the length of the sporulation processwas increased by glutamate. In contrast, Kennedy et al. (29)reported that lower glutamate concentrations led to a longersporulation time. These authors used different media and a differentstrain, however, which may explain this difference. DPA synthesisis regulated by ${sigma}$ ^K (16), a sporulation-specific sigma factorexpressed by the mother cell compartment in response to ${sigma}$ ^G activityin the forespore during the final stages of sporulation (37).It is possible that although sigG expressions are parallel inYLLG and YLHG, the actual ${sigma}$ ^G activity in YLHG is delayed, orthat ${sigma}$ ^G activity is also parallel but ${sigma}$ ^K activity in the mothercell is delayed. Either way, our findings suggest that thereis a certain degree of plasticity in the tightly regulated sporulationprogram.

Spore properties.
Important spore properties are germinability and heat resistance.Therefore, we determined the resistance of the spores from YLHGand YLLG to 95°C, and their germination response to twowell-known germinants for B. cereus: L-alanine and inosine.Spores from YLHG were more resistant to 95°C than thosefrom YLLG (Table 2). Because heat resistance has been associatedwith the core water content (4, 38), we measured the core density,which is linearly correlated with the water content (38). Thewhole-spore density and the DPA content of spores from YLLGwere rather similar to those of spores from YLHG, but the coredensity was somewhat higher (Table 2). This was in contrastto our expectations, because core density is generally believedto be positively correlated with heat resistance. Indeed, aswas stated by Atrih and Foster (1), complex cooperative phenomenaare involved in the assembly of spores, and a multicomponentsystem is the basis for spore heat resistance.

We measured the germinability of the spores without a priorheat activation step, in order to prevent any possible interferenceby compounds released from the spores themselves (39). Furthermore,spores from YLLG tended to germinate spontaneously after a heatshock for 15 min at 75°C, which reflects their heat sensitivity.Clear differences in germination speed and extent between sporesfrom YLHG and YLLG were observed, as can be seen in Fig. 4.The response to L-alanine in spores from YLLG was severely limited,while these spores responded faster to inosine but much slowerstill than spores from YLHG. The most powerful germinant knownfor B. cereus is a combination of L-alanine and inosine (13).Although the difference was less pronounced, even with a combinationof L-alanine and inosine, spores from YLHG germinated quickerand more completely than spores from YLLG. Thus, with everygerminant tested, spores from YLHG germinated faster and toa greater extent than spores from YLLG. It is known that sporesfrom different media may have different responses to L-alaninebecause of differences in the alanine racemase content (A. Moir,personal communication), an enzyme that is present in the exosporiumof B. cereus (49) and converts the germinant L-alanine intothe germination inhibitor D-alanine. Spores produced from YLLGmay have higher alanine racemase contents than those from YLHG.In addition to containing alanine racemase, the exosporium ofB. cereus contains nucleoside hydrolase, an enzyme that couldinhibit inosine-induced germination (49). The presence of higherlevels of nucleoside hydrolase in spores from YLLG could thereforeexplain the differences in inosine-induced germination thatwe observed. We tried germination in the presence of D-cycloserine,a known alanine racemase inhibitor (6), but this resulted inan overall lower response in all cases. Experiments with otheralanine racemase inhibitors (e.g., O-carbamyl-D-serine; 28)and nucleoside hydrolase inhibitors could bring more clarityto those issues.

View larger version (19K):
[in this window]
[in a new window]
FIG. 4. Germination measured by the decrease in A₆₀₀ of B. cereus spores from YLHG (closed symbols) and YLLG (open symbols). A₆₀₀ at time zero was 0.8 to 1.0 in all cases; curves are averages from four independent experiments. The following germinants were used: L-alanine, 10 mM (circles); inosine, 5 mM (triangles); and inosine in combination with L-alanine, 0.5 mM (each) (squares).

Clearly, the glutamate concentration affected the sporulationprocess and spore properties. Indeed, glutamate has been reportedto have a large impact on sporulation as well as spore propertiesof bacilli. In Bacillus thuringiensis, a close relative of B.cereus (21), glutamate appears to regulate TCA cycle activityas well as influence spore heat resistance and DPA content (35).B. cereus has been reported to metabolize glutamate primarilyas an energy source during the time of sporogenesis (10). Severalstudies have attempted to determine the fate of the carbon skeletonof glutamate, and it was concluded that glutamate used duringsporulation is respired almost completely to CO₂, largely viathe TCA cycle (10, 35). Our results confirm that glutamate isused largely during sporulation when present in a high concentration,but an interesting outcome is that glutamate metabolism duringsporulation does not result in an increase of NH₄⁺ in the medium.This implies that the glutamate served as an nitrogen donorduring the sporulation process, not only as a source of energyand reducing power, as was stated by Charba and Nakata (10).Indeed, in YLLG, NH₄⁺ was taken up by the cells during sporulation,but D-lactate, which can serve as a source of reducing potentialand energy, partly remained in the supernatant.

In conclusion, we have found that B. cereus is able to formspores on a chemically defined medium without glucose but withlactate as a main carbon source. Sporulation was not inducedby nutrient limitation. Glutamate delayed the final stages ofsporulation but not the initiation. Clearly, the glutamate concentrationinfluences such key spore properties as heat resistance andgermination. Our defined approach to sporulation and spore propertiesis highly suitable for gene expression analysis employing DNAmicroarrays, which we are developing at this moment. Futureexperiments coupling defined sporulation conditions with fulltranscriptome analysis and spore properties will allow us tocarefully dissect the molecular basis of such important sporeproperties as heat resistance and germination.

FOOTNOTES

* Corresponding author. Mailing address: Bornsesteeg 59, P.O. Box 17, 6700 AA Wageningen, The Netherlands. Phone: 31-317-475109. Fax: 31-317-475347. E-mail: Ynte.deVries{at}wur.nl.

REFERENCES

Top