Optimization of Spore and Antifungal Lipopeptide Production During the Solid-state Fermentation ...

17 Pages·2007·0.24 MB·English

ApplBiochemBiotechnol(2007)143:63–79
DOI10.1007/s12010-007-0036-1
Optimization of Spore and Antifungal Lipopeptide
Production During the Solid-state Fermentation
of Bacillus subtilis
Scott W. Pryor&Donna M. Gibson&Anthony G. Hay&
James M. Gossett&Larry P. Walker
Received:9October2006/Accepted:9March2007/
Publishedonline:17April2007
#HumanaPressInc.2007
Abstract Bacillus subtilis strain TrigoCor 1448 was grown on wheat middlings in
0.5-l solid-state fermentation (SSF) bioreactors for the production of an antifungal
biologicalcontrolagent.Totalantifungalactivitywasquantifiedusinga96-wellmicroplate
bioassay against the plant pathogen Fusarium oxysporum f. sp. melonis. The experimental
designforprocessoptimizationconsistedofa26−1fractionalfactorialdesignfollowedbya
centralcompositeface-centereddesign.InitialSSFparametersincludedintheoptimization
were aeration, fermentation length, pH buffering, peptone addition, nitrate addition, and
incubator temperature. Central composite face-centered design parameters included
incubator temperature, aeration rate, and initial moisture content (MC). Optimized
fermentation conditions were determined with response surface models fitted for both
sporeconcentrationandactivityofbiologicalcontrolproductextracts.Modelsshowedthat
activity measurements and spore production were most sensitive to substrate MC with
highest levels of each response variable occurring at maximum moisture levels. Whereas
maximum antifungal activity was seen in a limited area of the design space, spore
:
S.W.Pryor L.P.Walker(*)
DepartmentofBiologicalandEnvironmentalEngineering,CornellUniversity,
Riley-RobbHall,Ithaca,NY14853,USA
e-mail:[email protected]
D.M.Gibson
USDAARSPlantProtectionResearchUnit,
Ithaca,NY14853,USA
A.G.Hay
DepartmentofMicrobiology,CornellUniversity,
Ithaca,NY14853,USA
J.M.Gossett
SchoolofCivilandEnvironmentalEngineering,CornellUniversity,
Ithaca,NY14853,USA
Presentaddress:
S.W.Pryor
AgriculturalandBiosystemsEngineeringDepartment,NorthDakotaStateUniversity,
Fargo,ND58105,USA
64 ApplBiochemBiotechnol(2007)143:63–79
production was fairly robust with near maximum levels occurring over a wider range of
fermentation conditions. Optimization resulted in a 55% increase in inhibition and a 40%
increase in spore production over nonoptimized conditions.
Keywords Solid-statefermentation.Optimization.B. subtilis.Lipopeptides.Spores.
Biocontrol
Introduction
The biological control agent (BCA), Bacillus subtilis, produces environmentally resistant
sporesandisknowntoproduceaspectrumofantifungallipopeptides[1–3],yetitisunclear
howeachofthesepropertiescontributestotheoverallefficacyofthefinalbiologicalcontrol
product (BCP). In vitro production of lipopeptides during fermentation could have an
immediateeffectonthepathogenpopulationpostapplication;invitroproductionofbacterial
sporescouldallowforpotentialinsitulipopeptideproduction;orthecombinationofbothin
vitro and in situlipopeptideproduction could benecessary for maximum efficacy.
FurthercomplicatingthedevelopmentofaneffectiveB.subtilisBCPisthefactthatthe
organism produces a variety of antifungal lipopeptides such as iturin, surfactin, and
fengycin [4–6] that are expressed at different levels under different environmental
conditions. Published data detailing the effect of fermentation conditions on lipopeptide
production come from a combination of both liquid- and solid-state fermentation (SSF)
processes[5,7–9].ThesestudieshaveshownmaximumiturinAandsurfactinproductionat
25 and 37 °C, respectively, but little data are available for temperature dependency in the
production of other compounds [5, 9, 10]. Aeration has been shown to be an important
factor in SSF production of iturin A [10], whereas surfactin production is increased under
anaerobic, nitrate-limited conditions in liquid fermentation [8]. Other factors such as pH,
peptone addition, and shaking rate have been shown to significantly affect B. subtilis
lipopeptide production in liquid fermentation [9].
DespitetheavailableresearchonB.subtilislipopeptideproduction[7,9,11–16],littlehas
been done to examine different combinations of factors in SSF and their impact on overall
antifungal activityandspore production.Someworkshaveexaminedproductionofspecific
lipopeptides without measuring direct antifungal activity [7, 9]. Given the potential for
synergistic or antagonistic relationships between lipopeptide classes in the BCP, antifungal
activity maynot bea direct functionofindividual ortotallipopeptide concentrations.
A factorial experimental design was chosen to explore the effects of fermentation
conditionsonBCPantifungalactivityandsporeconcentration.Factorsandfactorlevelsfor
theexperimentaldesignweredefinedbasedonpreliminaryexperimentsandareviewofthe
literature. Production of both iturin A and fengycin has been shown to be highest after
exponentialgrowthphase,whereassurfactinproductionishighestduringexponentialgrowth
inliquidfermentation[9, 17]. Thus, fermentation time levels of 48 and 72 h were chosen
based on preliminary studies evaluating growth profiles. Levels of aeration and nitrate
supplementation were based on the work of Davis et al. [8] who reported that surfactin
production from a B. subtilis culture greatly increased in anaerobic cultures after nitrate
depletion. Jacques et al. [9] found that peptone, temperature, and pH all had significant
effects on lipopeptide production from B. subtilis during liquid fermentation; peptone had
the greatest effect of all factors tested with the maximum level of 4 g/l producing the
highest levels. Temperature has also been shown to have a significant effect on SSF
ApplBiochemBiotechnol(2007)143:63–79 65
production of lipopeptides by B. subtilis with maximum production of iturin A and
surfactin occurring at 25 and 37 °C, respectively [5]. Preliminary studies indicated that
4.3gN,N-bis[2-hydroxyethyl]-2-aminoethanesulfonicacid(BES)/l(100mM)providedthe
greatest pH buffering without adversely affecting cell growth.
This paper describes the use of a two-stage optimization consisting of a fractional
factorial design coupled with a central composite face-centered (CCF) design to model B.
subtilis SSF. Optimized fermentation conditions were determined with response surface
models fitted for both antifungal activity of the BCP and final spore concentration. This
approachallowedustoquantitativelyassesstheeffectsofmultiplefermentationconditions
on producing a BCP with a high concentration of antifungal lipopeptides and B. subtilis
spores to maximize in situ biological control.
Materials and Methods
Experimental Design
First-stage Optimization
A 26−1 fractional factorial experimental design was used to quantify the effects of (A)
fermentation time, (B) aeration rate, (C) nitrate level, (D) peptone level, (E) incubator
temperature, and (F) pH buffering (see Table 1). Biological control product spore
concentrations and extract inhibition measurements were both used as separate response
variables.Thefractionalfactorialdesignwasconstructedwitheachfactorsetateitherahighor
alowlevel,codedas+1and−1,respectively,asshowninTable1.Thedesign(seeTable2)
wascreatedbyconfoundingthesixthdesignfactor(F)withthefive-factorinteractioneffectof
the remaining five factors (A–E). Factor effects, defined as the mean difference between
treatments at the high and low levels of each factor or interaction, were calculated for each
main factor and all two-factor interactions in the experimental design [18]. All three-factor
interactionswereassumedtobenegligible andusedtoestimateexperimentalerror[18–20].
Second-state Optimization
ThesecondstageofprocessoptimizationwascomposedofaCCFdesignforthreefactors.
Central composite design is a common follow-up design to a fractional factorial design in
optimizationstudies[18,19,21].Itisusedtoestimatequadraticandinteractioneffectsbut
requires fewer runs than a full 3n factorial. The choice of factors included in the second
stage design is discussed in the “Results” section.
Table1 FactorlevelsinfractionalfactorialdesignforBCPoptimization.
Level Time(h) Airflow(lpm) Nitrate Peptone Incubator pHbuffer
(g/l) (g/l) temperature(°C) (gBES/l)
−1 48 0 0 0 22 0
+1 72 0.4 0.8 4 30 4.3
66 ApplBiochemBiotechnol(2007)143:63–79
Table2 First-levelfractionalfactorialdesignstructureforoptimizationofBCPproduction.
Treatment Time Airflow Nitrate Factorpeptone Incubatortemperature pHbuffer
A B C D E F=ABCDE
1 −1 −1 −1 −1 −1 −1
2 1 −1 −1 −1 −1 1
3 −1 1 −1 −1 −1 1
4 1 1 −1 −1 −1 −1
5 −1 −1 1 −1 −1 1
6 1 −1 1 −1 −1 −1
7 −1 1 1 −1 −1 −1
8 1 1 1 −1 −1 1
9 −1 −1 −1 1 −1 1
10 1 −1 −1 1 −1 −1
11 −1 1 −1 1 −1 −1
12 1 1 −1 1 −1 1
13 −1 −1 1 1 −1 −1
14 1 −1 1 1 −1 1
15 −1 1 1 1 −1 1
16 1 1 1 1 −1 −1
17 −1 −1 −1 −1 1 1
18 1 −1 −1 −1 1 −1
19 −1 1 −1 −1 1 −1
20 1 1 −1 −1 1 1
21 −1 −1 1 −1 1 −1
22 1 −1 1 −1 1 1
23 −1 1 1 −1 1 1
24 1 1 1 −1 1 −1
25 −1 −1 −1 1 1 −1
26 1 −1 −1 1 1 1
27 −1 1 −1 1 1 1
28 1 1 −1 1 1 −1
29 −1 −1 1 1 1 1
30 1 −1 1 1 1 −1
31 −1 1 1 1 1 −1
32 1 1 1 1 1 1
SSF System
Forced aeration batch microreactors used in this study were the same as described
previously ([22]. Reactors were constructed of 0.5-l polypropylene bottles (Nalgene,
Rochester, NY, USA). Each reactor was fitted with a 3.2-cm-high PVC pipe covered with
nylon screentosupport thesubstrate bed.Thereactor workingvolume wasapproximately
0.4 l. Reactor lids had two pinhole openings with attached rubber septa (Chemglass,
Vineland, NJ, USA), allowing thermocouples to be placed in the bed while maintaining a
tightairseal.Boththermocoupleswereinsertedintothecenterofthebed.Inletandexhaust
fittings were glued in the reactor bottom and lid, respectively, for reactor aeration. Each
reactorhada0.2-μmairfilter(PallCorp.,EastHills,NY,USA)fittedjustbeforethereactor
inlet to eliminate microbial contamination from the incoming air stream.
Substrate for all experiments was prepared separately for each reactor according to
experimentaldesignspecifications.Onehundredtwenty-fivegramsofwheatmiddlingsata
ApplBiochemBiotechnol(2007)143:63–79 67
moisture content (MC) of 8.7% dry basis (d.b.) were added to individual autoclave bags,
mixedwith77.5mlofdistilledwater,andautoclavedtwiceat121°Cfor25minwith24h
betweensterilizations.Theremainingwaterneededtobringthematerialuptotherequired
initial MC (1–1.7 g/g d.b.) was added directly before reactor loading. A portion of the
additionalwaterwasmixedwithrequiredsupplements,asnotedintheexperimentaldesign,
and was sterilized separately before mixing with substrate. The remaining 40 ml of sterile
water was mixed with 2.5 ml of inoculum culture for more uniform inoculum distribution
throughoutthesubstrate.InoculumofB.subtilisstrainTrigoCor1448,ATCC202152[23]
waspreparedasdescribedpreviously[22].Substratewasmixedthoroughlywiththediluted
inoculumbeforebeingloadedintosterilizedbioreactors.Allmixingofmaterialwasdonein
individual autoclave bags in a sterile laminar flow hood.
Reactorswereplacedinarefrigeratedincubator(BiocoldScientific,Fenton,MO,USA)to
control ambient temperatures throughout fermentation. After 48 or 72 h of fermentation,
fermentation material was dried to approximately 11–18% (d.b.) by aerating with approx-
imately 4 lpm of dry air for 2–3 days. Reactor temperatures immediately dropped to
approximately15°C,effectivelyslowingmicrobialgrowthtonearminimum.FinaldriedBCP
washomogenizedinaCuisinartMini-PrepPlus(Stamford,CT,USA)andstoredat−20°C.
Microbial and Biochemical Assays
Finalsporeconcentrationforeachreactorwascalculatedbydilutionplatecountusingsterile
0.1% sodium pyrophosphate (SPP) for all dilutions. Approximately 0.5 g of dried,
homogenized BCPwasplaced in a 50-ml vialwith20mlofsterileSPP, vortexed twicefor
1mintodislodgecellsfromthesolidmatrix,andseriallydilutedinsterile1.5-mlmicrofuge
tubes. The first dilution was placed in a water bath at 75 °C for 10 min to eliminate viable
vegetativecellsbeforeplating.Quadruplicate10-μlspotswereplacedonnutrientyeastextract
agarplatesandincubatedat25°Cfor15–20h.Colonieswerecountedforallreplicatespotsand
themeanwasusedtocalculatetheconcentrationofB.subtilis spores per dry gram of BCP.
Biologicalcontrolproductpreparationswereextractedwithmethanoltoobtainmaterials
containing antifungal lipopeptides for use in inhibition bioassays. Biological control
productsamples(5drygram)wereplacedin40-mlglassvialswith25mlofmethanoland
allowed to steep for approximately 16 h. Material was mixed several times by hand, then
vacuum-filtered through two 70-mm Whatman no. 1 filter papers; the final volume was
broughtto30mlwithmethanol.A15-mlsampleofthefinalvolume(equivalentto2.5dry
gram of BCP) was passed through a 0.2-μm polyethersulfone membrane filter (Whatman,
Inc, Clifton, NJ, USA) and then mixed with 15 ml of distilled water in preparation for a
solid-phase extraction cleanup step [24].
A 96-well microplate bioassay was used to measure the inhibition of BCP extracts
against Fusarium oxysporum f. sp. melonis [22]. Extract inhibition was tested by adding
100μlofpotatodextrosebroth,10μlofBCPextractorsolventcontrol,and100μlofaF.
oxysporum spore suspension (104 spores/ml) to each well. Fungal cultures were cultivated
for1–2weeksonquarterstrengthpotatodextroseagarplates.Plateswerethenfloodedwith
10–15mlofsteriledistilledwater,gentlyscrapedtoformasporesuspension,anddilutedto
1×104spores/mlforuseinbioassays.Biologicalcontrolproductextractsweredilutedwith
methanol tothe specified concentration (mg BCP/ml) and alltreatments were measured in
triplicate. Plates were read using a Synergy HT plate reader (Bio-Tek Instruments,
Winooski, VT, USA) at an absorbance of 620 nm before and after incubation for 48 h at
25 °C. Percent inhibition was determined by comparing the absorbance increase of the
wellscontainingBCPextractwithwellscontainingcontrolsolventusedforextractdilution.
68 ApplBiochemBiotechnol(2007)143:63–79
Data Collection
Temperaturesweremonitoredwithtwotype-Tthermocouplesinthecenterofeachreactor.
Ambient and outlet oxygen (O ) concentrations were measured with two separate O
2 2
sensors (NeuwGhent Technology, LaGrangeville, NY, USA). One sensor was dedicated to
reading ambient O concentrations whereas the second was connected to a multiposition
2
actuator valve (Valco Instruments Co., Inc., Houston, TX, USA) linked to multiple
microreactors. The data acquisition system was developed using a Dell OptiPlex GX1p
desktop computer. A CIO-DAS08 analog and digital I/O board and two CIO-EXP32 data
acquisition board (Measurement Computing- Middleboro, MA, USA) were used to record
reactortemperatureandoutletO concentrationdataandtocontrolthemultipositionvalve.
2
ALabVIEW(NationalInstruments,Austin,TX,USA)programwaswrittentomonitorand
record temperatures from up to 18 microreactors with two separate thermocouples. Mean
reactordatawerecontinuallyuploadedtoareal-timechartforprocessmonitoringanda2-min
runningaverageofeachthermocouplewaswrittentoafilein5-minintervals.Theprogram
alsomonitoredandrecordeddatafromtwoseparateO sensors.Thedataacquisitionprogram
2
controlled the multiposition valve so that the sensor sampled exhaust gas from a different
reactor every 4 min until all reactors were sampled. After a 3-min purge time after each
sample,a1-minrunningaverageofO concentrationwaswrittentoafileforlateranalysis.
2
Data Modeling
The CCF design was structured to explore the role of aeration rate, MC, and incubator
temperature at three different levels. Main factor effects, quadratic effects, and two-factor
interactions were all discernable with this design. Five replicate runs at the design center
were completed to measure experimental error in the system. Tables 3 and 4 show the
codedfactorlevelsandcentralcompositedesigntable,respectively.TheCCFdesignresults
wereusedtodevelopamultiplelinearregressionmodelforthesystem.Forthethree-factor
design, the regression had ten component parameters (β0–9): one overall mean, three
individual main factor effects, three quadratic factor effects, and three two-factor
interactions [21]. The model has the form
y¼b þb x þb x þb x þb x2þb x2þb x2þb x x þb x x þb x x ð1Þ
0 1 1 2 2 3 3 4 1 5 2 6 3 7 1 2 8 1 3 9 2 3
where
y response variable of percent inhibition or spore concentration
x aeration level (lpm)
1
x substrate initial d.b. MC (g/g)
2
x incubator temperature (°C)
3
β0–9 model parameters
The method of least squares was used to estimate values for β0–9.
Table3 CodeforfactorlevelsincentralcompositedesignforoptimizationofBCPproduction.
Level Aeration(lpm) Initialmoisturecontent(g/gd.b.) Incubatortemperature(°C)
−1 0.1lpm 1.00 19
0 0.8lpm 1.35 23
1 1.5lpm 1.70 27
ApplBiochemBiotechnol(2007)143:63–79 69
Table4 Centralcompositeface-
centereddesignschemeforopti- Treatment Aeration Moisture Incubatortemperature
mizationofBCPproduction.
1 −1 −1 −1
2 −1 −1 1
3 −1 1 −1
4 −1 1 1
5 1 −1 −1
6 1 −1 1
7 1 1 −1
8 1 1 1
9 −1 0 0
10 1 0 0
11 0 −1 0
12 0 1 0
13 0 0 −1
14 0 0 1
15 0 0 0
16 0 0 0
17 0 0 0
18 0 0 0
19 0 0 0
Spore concentrations and bioassay inhibition data were used for response surface
modeling. A MATLAB program was written to calculate model parameters for Eq. 1 for
each response variable. The MATLAB program was also used to determine theoretical
stationary points and local maxima within the design space, create analysis of variance
(ANOVA) tables for models, and plot response surface model results.
Model coefficients were tested for significance using SPSS software (SPSS, Inc.,
Chicago,IL,USA)andparameterswhosecoefficientshadapvaluegreaterthan0.05were
sequentiallyremovedtodeterminetheeffectoftheparameteronthemodelR2andadjusted
R2 (R2 ) values. Parameters were permanently removed from the model when it was seen
adj
that their inclusion had a negative impact on R2 values.
adj
Model coefficients (β0–9) using scaled unit variables were found using the method of
leastsquaresandconvertedforusewithstandardunitsforeachfactoraccordingtoTable3.
Coefficientsforusewithstandardunitvariableswerefoundbyconvertingeachfactorlevel,
x , in Eq. 1 to standard units using the following equations:
1,2,3
x ðlpmÞ¼0:7x þ0:8 ð2Þ
1 1
x ðg=g d:b:Þ¼0:35x þ1:35 ð3Þ
2 2
x ð(cid:2)CÞ¼4x þ23 ð4Þ
3 3
Results
First Stage of Optimization
Typical reactor temperature profiles are shown in Fig. 1. Temperature profiles were very
similarforreactorswiththesameincubatortemperaturelevelandaerationlevelregardless
70 ApplBiochemBiotechnol(2007)143:63–79
Fig.1 Typicalbioreactortemper- 50
atureprofilesforgivenincubator Incubator Aeration
temperaturesandaerationlevels e (˚C) . 4405 33220022˚˚˚˚CCCC AANNeeoorrnnaaaatteeeeddrraatteedd
ur
at
per 35
m
e
T
or 30
ct
a
e
R 25
20
0 10 20 30 40 50 60 70
Time (hrs)
of all other factor levels. Mean maximum temperatures for all aerated reactors in 30 and
22 °C incubators were 49.8±2.2 and 47.9±2.8 °C, respectively. As expected, the
nonaeratedreactorsdidnotexhibitsignificantautoheatingwhereasaeratedreactorsshowed
reactor temperature increases of approximately 20 °C above incubator temperature.
Increased convective heat losses were unable to significantly reduce maximum temper-
atures in aerated reactors in the low-temperature incubator. Incubator temperature did,
however,influencelagtimesasillustratedinFig.1.Lagtimesincreasedfromlessthan6h
for the 30-°C incubator reactors to greater than 15 h for the 22-°C incubator reactors.
Despite the high temperatures reached in the reactors in the low-temperature incubator,
temperatures dropped more rapidly after exponential growth phase.
60% nonaerated, low temperature
aerated, low temperature
50% nonaerated, high temperature
aerated, high temperature
40%
30%
n
o
biti 20%
hi
n
I
10%
0%
-10%
-20%
12345678901234567890123456789012
11111111112222222222333
Treatment
Fig.2 Inhibitiondataat20mgBCP/mlforfractionalfactorialdesignforBCPoptimizationfortreatments
listedinTable2(errorbarsrepresentsamplestandarddeviation)
ApplBiochemBiotechnol(2007)143:63–79 71
TheresultantBCPextractsweretestedforgrowthinhibitionofF.oxysporumf.sp.melonis
at concentrations between 2.5 and 160 mg BCP/ml. Aerated reactors had the highest
inhibitionlevelsbetween20and40mgBCP/mlanddeclinedathigherconcentrations(data
not shown). Extracts from nonaerated reactors showed little or no inhibition at lower
concentrations and less than 25% inhibition even at the highest concentrations tested.
Inhibitionvaluesat20mgBCP/mlwerechosenforfurtherstatisticalanalysis;dataresolution
was greatest at this concentration and lower and higher concentrations had a poor response
and were saturated at maximum inhibition, respectively. Inhibition (at 20 mg BCP/ml) and
spore concentration results for the fractional factorial design are shown in Figs. 2 and 3,
respectively.Inhibitionvarianceisbaseduponreplicatewellsfromasinglesampleextraction
whereas spore count variance is assumed to be 5% of the measured value based upon
preliminarystudies ofreplicate counts fromhomogenized contentsofa singlereactor.
Data from Figs. 2 and 3 have been formatted to highlight specific differences between
reactorsatdifferentaerationlevelsandincubatortemperatures.Althoughmostfactortrends
are difficult to distinguish from a cursory check of the data, extracts from all aerated
reactorsshowednoticeablyhigherinhibitionlevelsthanthosefromnonaeratedreactors.In
addition, extracts from reactors in a lower temperature incubator showed slightly greater
inhibition than those in a higher temperature incubator.
Tabulated in Table 5 are factor effects for the first-stage fractional factorial design.
Fermentation time and aeration both had significant positive effects on the inhibitory
quality of the BCP extract, whereas the addition of nitrate and peptone had significant
negative effects. The interaction effects of aeration with both nitrate and peptone
(interactions BC and BD, respectively) were both significant and negative. These
interactionsindicatethatthenegativeeffectofeachofthesefactorsisgreaterinmagnitude
in aerated reactors. As the aeration effect was strong and positive, this provided further
evidence that neither nitrate nor peptone should be included in further studies as they
negativelyimpacttheproductionofinhibitorycompounds.Thestrongnegativeinteraction
between aeration and temperature also showed that the effect of temperature was
1.0E+12 nonaerated, low temperature
aerated, low temperature
nonaerated, high temperature
1.0E+11
aerated, high temperature
g)
y
dr
u/ 1.0E+10
cf
n (
o
ati 1.0E+09
ntr
e
c
n
o
C 1.0E+08
e
or
p
S
1.0E+07
1.0E+06
12345678901234567890123456789012
11111111112222222222333
Treatment
Fig.3 SporeconcentrationdataforfractionalfactorialdesignofBCPoptimizationfortreatmentslistedin
Table2(errorbarsrepresentsamplestandarddeviation)
72 ApplBiochemBiotechnol(2007)143:63–79
Table 5 Main factor effects and two-factor interactions from fractional factorial design using microplate
inhibitionandsporeconcentrationdata.
Factor Effectinhibitionat20mgBCP/ml Effectlog(sporecount)
(Time)A 0.039a 0.03
(Airflow)B 0.313a 2.69a
(Nitrate)C −0.012 0.12
(Peptone)D −0.055a −0.41a
(Temp)E −0.060a −0.19
(pH)F 0.028 −0.02
AB 0.031 0.03
AC 0.009 −0.15
AD −0.022 −0.18
AE −0.023 −0.15
AF −0.009 −0.26
BC −0.032b −0.14
BD −0.032b 0.30b
BE −0.058a −0.11
BF −0.000 0.11
CD −0.028 −0.01
CE −0.028 0.03
CF 0.005 0.16
DE −0.010 0.29
DF 0.022 −0.00
EF 0.002 −0.11
Standarddeviation 0.019 0.18
aIndicatessignificanceatthe95%confidencelevel
bIndicatessignificanceatthe90%confidencelevel
exacerbated at high aeration rates. Incubator temperature had a small negative effect in
nonaerated reactors and a sizeable negative effect in aerated reactors, providing further
evidence that incubator temperature is an especially important factor in the production of
bioactive compounds by B. subtilis in small-scale aerobic cultures.
Factor effects for spore production were similar to those for the inhibition data in that
aerationandpeptoneadditionshowedsignificantpositiveandnegativeeffects,respectively.
However, unlike the inhibition data, spore counts did not show significant effects from
fermentation time or incubator temperature.
Although fermentation time did show a positive effect on inhibition levels, the
significant impact of cooler ambient temperature on lag time makesit difficult to compare
thereactors onthis basis.The48-hstoptime for reactorsinthe cooland warmincubators
occurred at 15 and 22 h after peak temperature, respectively. A follow-up experiment was
conducted to explore fermentation time relative to the time of peak temperature. Four
reactorswererunat22and30°Candareactorwaskilledat 0,8,16,and40hafterpeak
temperature. Reactor contents were dried and extracted as described earlier.
ResultsofinhibitionassaysforthoseextractsareshowninFig.4andrevealthatproduction
of bioactive compounds in reactors with different incubator temperatures followed similar
trends. Quadratic equations were fitted through the data for the 22- and 30-°C reactors, as
shown in Fig.4. Maximum inhibition was predicted at 24 h past peak temperature for both
datasets.Furtherstudiesdidnotincludefermentationtimeasadesignfactor;regularaeration
was stoppedat 24h pastpeak temperaturefor alltreatments.