Enzyme and Microbial Technology 52 (2013) 157– 162
Contents lists available at SciVerse ScienceDirect
Enzyme and Microbial Technology
jou rn al h om epage: www.elsev ier .com/ locate /emt
trengths and weaknesses in the determination of Saccharomyces cerevisiae celliability by ATP-based bioluminescence assay
ucia Paciello, Francesco Cristino Falco, Carmine Landi, Palma Parascandola ∗
epartment of Industrial Engineering, Università di Salerno, Via Ponte Don Melillo, 84084 Fisciano, Salerno, Italy
r t i c l e i n f o
rticle history:eceived 3 October 2012eceived in revised form0 December 2012ccepted 31 December 2012
eywords:
a b s t r a c t
Due to its sensitivity and speed of execution, detection of ATP by luciferin–luciferase reaction is a widelyspread system to highlight cell viability. The paper describes the methodology followed to successfullyrun the assay in the presence of yeast cells of two strains of the yeast Saccharomyces cerevisiae, BY4741and CEN.PK2-1C and emphasizes the importance of correctly determining the contact time between thelysing agent and the yeast cells. Once this was established, luciferin–luciferase reaction was exploited todetermine the maximum specific rate of growth, as well as cell viability in a series of routine tests. The
iabilityeastioprocess monitoringed-batch cultureTP based bioluminescence assay
results obtained in this preliminary study highlighted that using luciferin–luciferase can imply an over-estimation of maximum specific growth rate with respect to that determined by optical density and/orviable count. On the contrary, the bioluminescence assay gave the possibility to highlight, if employedtogether with viable count, physiological changes occurring in yeast cells as response to stressful envi-ronmental conditions such as those deriving from exposure of yeast cells to high temperature or those
ve co
espiro-fermentative metabolism depending on the operati
. Introduction
Detection of ATP by luciferin–luciferase reaction is aidely spread system to highlight cell viability [1,2]. The
uciferin–luciferase system from firefly has been first describedn 1963 by McElroy [3]. Luciferase (EC 1.13.12.7) catalyzes anxidative reaction (as shown in the scheme below), involvingellular ATP, firefly luciferin, a metallic cation [4,5], and molecularxygen, yielding an electronically excited oxyluciferin species [6].
uciferin+O2+ATPluciferase,Mg2+
−→ Oxyluciferin+AMP+PPi+CO2+hv
This excited species, returning to the ground state, emit visi-le light (hv), which is employed by the firefly in its reproductiveehaviour [7]. In addition to firefly, bioluminescence i.e. the emis-ion of visible light is observable in many other living organismsuch as various fish, fungi, and bacteria [8,9] and its applicationsnclude the estimation of bacterial content in clinical specimens10], foods [11] and drinking water [12], the use of firefly luciferaseo study the role of chaperones in protein folding [13] and theenes encoding luciferases to monitor transcriptional activities
14].
The reaction described above is very efficient because almostll the energy input into the reaction is transformed into
light [15] which is linearly related to the ATP concentrationand therefore directly proportional to the number of viablecells [16,17].
In the present work, cell viability of two strains of the yeastSaccharomyces cerevisiae, BY4741 and CEN.PK2-1C, has been deter-mined under different conditions, by employing the ATP-basedbioluminescence assay. These strains have been taken into con-sideration because they are commonly used as reference strainsby the yeast research community. Both of them bear four auxotro-phies and, according to our experience, exhibit the same kineticbehaviour when employed in both batch and fed-batch processes[18,19].
The kit used was one on the market (ViaLight® Plus - LonzaRockland, USA), based on Photinus pyralis (firefly) luciferase. Theprotocol suggested by the manufacturer has been amended inorder to adapt the method to the determination of cell viabilityin the yeast strains above mentioned, paying particular attentionto the time needed for the cell lysis and the stability of the lightsignal.
The method has been also investigated for its use in the eval-uation of exponential growth of S. cerevisiae. Further, the fireflyluciferin–luciferase system has been utilized together with othertwo techniques, the optical density and the viable count on agarplate, to (i) identify different sub-populations of BY4741 strain in a
broth culture after its exposure to high temperature and (ii) high-light peculiar physiological behaviours of CEN.PK2-1C strain cellsgrowing in a really complex environment such as that which arisesin a fed-batch reactor after many hours of run.
The strains used in this work were S. cerevisiae BY4741 (MATa ura3�0eu2�0 met15�0 his 3�1) and S. cerevisiae CEN.PK2-1C (MATa ura3-52 his3-�1eu2-3,112 trp1-289 MAL2-8C SUC2) both of them purchased at the EUROSCARFwww.uni-frankfurt.de/fb15/mikro/euroscarf).
.2. ATP-based bioluminescence assay
To determine ATP by bioluminescence assay, ViaLight® Plus kit (Lonza Rock-and, Inc. Rockland, ME 04841 USA), containing Bactolyse® the agent for theeast cell-wall lysis and AMR® , the ATP Monitoring Reagent with the enzymeuciferase and the substrate luciferin in addition to ATPases inhibitors, wassed.
Bioluminescence assay was performed in polypropylene tubes according to theollowing procedure: 50 �l of Bactolyse® were added to 100 �l of a culture sam-le suitably diluted, and vigorously stirred to homogenize the reaction mixture (theample was analyzed in triplicate). The lysing agent was left in contact with the yeastell suspension for 20 min to be sure that yeast cell-lysis was completed, then 100 �lf AMR® were added to each one. The tubes were vigorously stirred to allow oxygeno be incorporated into the reaction mixture and incubated at room temperature.he first reading at the tube luminometer (Lumat LB 9507 - Berthold Technologies)as performed after 2 min incubation time, followed by a second reading after fur-
her 2 min. The average of the two readings corresponded to the intensity of theight produced by 100 �l sample. Each determination was made in triplicate. All thexperimental values were expressed as Relative Light Units per second and per mlRLUs ml−1) multiplying by ten (initial dilution) the reading obtained as describedbove.
.3. Determination of yeast cell number by viable count
Suitable dilutions of samples collected during shake-flask experiments, ther-al inactivation tests and fed-batch runs were spread over the surface of YPD agar
lates containing 2% glucose (Serva Feinbiochemica, Heidelberg, Germany), 2% bactoeptone, 1% yeast extract and 2% agar, being all % expressed as w/v. The plates were
ncubated for 48 h at 30 ◦C up to the formation of colonies visible to naked eye andach of them was assumed to be originated from an individual viable cell. All sam-les were analyzed in triplicate and the values of standard deviation obtained, variedetween 3 and 5%.
.4. Determination of yeast cell density by spectrophotometric detection
Yeast cell density was quantified by dividing optical density (O.D.) of cell sus-ensions at 590 nm (Hach Lange UV/Vis spectrophotometer) by a correlation factor2.30 and 2.45 O.D.590 per mg ml−1 for S. cerevisiae CEN.PK2-1C and BY4741 strains,espectively). The correlation factor corresponded to the slope of the calibrationurve obtained by plotting optical density of yeast cell suspensions as a function ofheir density evaluated on a dry weight basis [18]. Unless otherwise stated, mg ofiomass were always referred as dry weight. To achieve dry weight, culture samplesere centrifuged (4000 rpm, 3 min) and the pellet washed twice, resuspended withistilled water, and dried for 24 h at 105 ◦C until a constant weight [20]. Parallelamples varied about 3–5%.
.5. Shake-flask cultures
Yeast cells were grown in 500 ml shake flasks containing 100 ml of a definedineral medium prepared according to Verduyn et al. [21] and containing vita-ins and trace elements. The initial glucose concentration was 2% (w/v). Theedium was supplemented with 1% (w/v) casamino acids (BD BactoTM Casaminocids, BectonDickinson&Co., Sparks, MD 21152 USA) as auxotrophy-complementingmino acid source, and also uracil (7.5 ml of a 0.2%, w/v stock solution) andryptophan (0.8 ml of a 0.5%, w/v stock solution) for S. cerevisiae CEN.PK2-1Ctrain. The addition of tryptophan was necessary due to the low tryptophanontent in casamino acids (Becton–Dickinson technical handbook). The amount
f all the medium components above mentioned, was determined on the basisf the recommendations on the nutritional requests obtained from the litera-ure [22]. The shake flask inoculum (the aliquot of cell suspension needed toive the initial optical density of 0.2 at 590 nm (O.D.590)), was prepared from anxponential pre-culture containing the same medium of the shake flask culture.he pre-culture was inoculated with an aliquot of a frozen stock culture (80 ◦Cn 12.5%, v/v glycerol) to give a 0.2 initial O.D.590. Pre-cultures and subsequentultures were incubated at 30 ◦C and 220 rpm (Stuart Scientific S150 Orbital Incu-ator). Following this procedure, lag phase of yeast growing in shake flask waslways 2 h.
l Technology 52 (2013) 157– 162
2.6. Thermal inactivation of yeast cells and monitoring of residual viability
Cells of the S. cerevisiae BY4741 strain were cultivated as described above in theSection 2.5. After 24 h incubation, when yeast population had already entered thestationary phase, a suitable amount of broth culture was withdrawn and washedtwice. The pellet obtained by centrifugation (3 min, 4000 rpm) was re-suspendedin the same volume of saline (NaCl 0.9%, w/v), and divided into aliquots of 3 mleach, contained in glass tubes. These were placed in a thermostatic shaker bathand incubated at 53 ◦C. At defined time intervals, the tubes were collected, cooledand suitably processed to be tested, immediately or after having been kept at roomtemperature for 24 h, for both ATP-based bioluminescence assay and viable count onagar plate. Before initiating the thermal inactivation test, an aliquot of broth culturewas processed for determination of the RLUs ml−1 in addition to both total count ina Burker chamber and viable count on agar plate to be aware of the level of viabilityof the initial yeast population.
For kd determination, it was assumed that the death reaction followed the firstorder kinetics. The CFU ml−1 values, obtained by viable count, were plotted vs time.The exponent of the regression curve corresponded to kd value.
2.7. Fed-batch culture
Fed-batch culture was carried out with cells of S. cerevisiae and CEN.PK2-1Cstrain proliferating at 30 ◦C in the stirred fermenter, a 2.0 l final working volumeBioflo® 110 (New Brunswick Scientific). The fermenter initially contained 1 l of thedefined mineral medium above mentioned (Section 2.5) and was inoculated to givean initial O.D.590 of 0.04. Fed-batch culture started, after 15 h of batch phase, by sup-plying the fermenter with, an exponentially increasing feed to allow the CEN.PK2-1Ccells to proliferate with a constant specific growth rate of 0.08 h−1, a value suffi-ciently lower than the critical one to avoid over-flow metabolism [23]. The feedingsolution contained glucose (50%, w/v), salts, trace elements, glutamic acid, vitamins,and casamino acids, the concentration of which was calculated according to Pacielloet al. [18] and Porro et al. [24], taking into account the value of biomass yield for thegiven amino acid in aerobic conditions [25]. Oxygen was supplied by air sparging(DOT 30%, v/v air saturation). The culture pH was maintained at 5.0 by automaticaddition of 2 N KOH during batch phase and 10% (v/v) NH4OH during exponentialphase. The foam level in the fermenter was controlled by the automatic addition ofthe antifoam B (Sigma–Aldrich) (dil. 1:10).
3. Results
3.1. Standardization of Bactolyse®- AMR® system
Before routinely using the bioluminescence test to determineyeast viability, two important parameters were investigated, (i) theoptimum contact time of yeast cells with Bactolyse® that is the timerequired for the cell-walls to be completely disrupted by the lysingagent, and (ii) the time dependence of the intensity of the light sig-nal after addition of AMR®, the ATP monitoring reagent containingboth luciferin and luciferase.
To this purpose, a series of four light-signal monitoring tests,20 min each, were carried out with cells of S. cerevisiae CEN.PK2-1Cstrain, collected after 24 h incubation in shake flasks. The sam-ples were processed as previously described (see Section 2.2)except for the contact time with Bactolyse® which was allowed toreact with yeast cells for different times, namely 10, 15, 20 and25 min (Fig. 1). Once the contact time was over, the first detec-tion of the light signal was made within 4 min since the additionof AMR®, then readings continued until the end of the monitoringtest.
The optimum of contact time, was experimentally establishedand corresponded to 20 min because the difference in RLUs ml−1
between the last two experiments of 20 and 25 min contact time,was considered negligible being 1.13%.
The maximum of light intensity was achieved more rapidlythe longer was the contact time between yeast cell suspensionand Bactolyse®. Indeed, in the first monitoring test (10 min con-tact time), maximum intensity of the emitted light was achievedafter 11 min (Fig. 1) since the addition of AMR®, whereas in the
third monitoring test (that of optimum contact time) (Fig. 1), it wasachieved immediately after AMR® addition (data not shown), thenthe intensity of the light gradually diminished, and, at the end ofthe monitoring test, reduced by only 2% of the initial value. Taking
L. Paciello et al. / Enzyme and Microbial Technology 52 (2013) 157– 162 159
Fig. 1. 3D representation of the tests performed to identify the optimum contact-t ®
mi
imaww(
aclbl
3m
Btr
pbcOarti(rviBs
cdTde(t
RLUs ml-1= 7.07 ·106 e0.565 t
R² = 0.99 3
CFU ml-1 = 2.80 ·106 e0.501 t
R² = 0.99 9
OD = 0.166 e0.515 t
R² = 0.99 9
0
2
4
6
8
10
0
20
40
60
80
100
120
140
160
180
0 2 4 6 8 10
OD
590
( RLU
sm
l-1, C
FU m
l-1 )·
10-6 B
RLUs ml-1= 6.07 ·105 e0.513 t
R² = 0.99 1
CFU ml-1 = 2.95 ·106 e0.460 t
R² = 0.99 4
OD = 9.12 ·10-2 e0.470 t
R² = 0.99 9
0
1
2
3
4
5
0
20
40
60
80
100
120
0 2 4 6 8 10
OD
590
( RLU
sm
l-1, C
FU m
l-1 )·
10-6 A
0
1
2
3
0 2 4 6 8 10
(RLU
sm
g-1) ·
10-7
Time [h]
C
Fig. 2. Determination of maximum specific growth rate: S. cerevisiae BY4741 strain(A) and S. cerevisiae CEN.PK2-1C strain (B). Exponential growth monitored throughbioluminescence assay (full and empty triangles), optical density (full and empty
ime between S. cerevisiae CEN.PK2-1C cells and Bactolyse : four bioluminescenceonitoring tests of 20 min each (detection time) are set up by progressively increas-
ng the contact time (t* = 10, 15, 20, 25 min) with the lysing agent.
nto consideration that during the experiments all the measure-ents of the emitted light were generally made in the first 4 min
fter AMR® addition, the observed decrease in the intensity of lightas considered negligible. Fully comparable results were obtainedhen the experimental protocol was repeated with BY4741 strain
data not shown).The firefly luciferin–luciferase system ensured the linearity in
measuring range of three orders of magnitudes of ATP con-entrations (1 × 10−9 to 1 × 10−6 M) (data not shown). The veryow ATP concentration in correspondence of which ATP coulde still detectable was indicative of a high sensitivity of the
uciferin–luciferase assay for this substrate.
.2. ATP-based bioluminescence assay and the determination ofaximum specific yeast growth rate
Shake flask experiments, carried out with either S. cerevisiaeY4741 or S. cerevisiae CEN.PK2-1C strains, were set up to evaluatehe possibility to determine the maximum value of specific growthate, �max, by bioluminescence.
At defined time intervals during the exponential phase, sam-les were withdrawn from the flasks to monitor the yeast growthy bioluminescence as well as through optical density and viableount. The time-course of the RLUs ml−1, the CFU ml−1 and the.D.590, is reported in Fig. 2A and B, for S. cerevisiae BY4741nd S. cerevisiae CEN.PK2-1C respectively. The good exponentialegression of the experimental data (Fig. 2A–B) coming from thehree different techniques employed, accounted for their reliabil-ty and reproducibility. �max values evaluated from RLUs ml−1
0.51 and 0.56 h−1 in the case of BY4741 and CEN.PK2-1C strains,espectively) resulted always higher than those obtained by bothiable count and optical density which were, in turn, quite sim-lar (0.46 and 0.47 h−1 respectively, in the case of S. cerevisiaeY4741, 0.51 and 0.52 h−1, respectively, for S. cerevisiae CEN.PK2-1Ctrain).
To investigate this behaviour, samples of the two strains wereollected during exponential growth, and also processed for theetermination of cell density by dry weight (see Section 2.4).he values of RLUs mg−1 obtained by dividing RLUs ml−1 by cell
ensity were plotted against time (Fig. 2C). These values, asxpected (Fig. 2C), wandered an approximately constant valueexcept for the first point, presumably still not adapted to the cul-ivation conditions), only for a short time interval (i.e. until the
rhombus), and viable count on agar plate (full and empty squares). Time-course ofRLUs mg−1 (C) for S. cerevisiae BY4741 strain (full symbols) and S. cerevisiae CEN.PK2-1C strain (empty symbols).
fifth-sixth hour of incubation). After this time, the RLUs mg−1 val-ues suddenly increased causing the overestimation of the �max
value.
3.3. ATP-based bioluminescence assay to determine viability ofyeast cells undergoing high temperature treatment
Bioluminescence assay has been exploited also to study theeffect of high temperature on S. cerevisiae cell viability. To this pur-pose, stationary yeast cells belonging to the S. cerevisiae BY4741
strain, were processed as described in Section 2.6, then they wereplaced in a thermostatic shaker bath and exposed to a temperatureof 53 ◦C for 3 h. Samples were collected at fixed time intervals forthe determination of both the RLUs ml−1 emitted and CFU ml−1 on
160 L. Paciello et al. / Enzyme and Microbia
Fig. 3. Thermal inactivation of S. cerevisiae BY4741 cells after exposure to 53 ◦C:viable count (rhombus) or bioluminescence (triangles) of samples assayed soonafter collection (full symbols) or after having been kept at room temperature for24 h (empty symbols) (A). Yeast sub-populations arising after having been kept thesamples at room temperature for 24 h (B): viable cells preserving intact the abilityto reproduce (light grey), “resilient” cells which have irreversibly lost the ability torl
apo
stTwgi
boadv
oRcilmdiag
−1
eproduce but keep their structure (dark grey), dead cells which have irreversiblyost their both structural and metabolic integrity (black).
gar plate. These determinations were repeated on the same sam-les after they have been stored at room temperature for 24 h inrder to highlight any non-permanent change.
Before initiating the test of thermal inactivation an aliquot oftationary yeast cell suspension was assayed for RLUs ml−1 emit-ed, total count in a Burker chamber and viable count on agar plate.his procedure was intended to highlight that total and viable countere quite comparable (data not shown), i.e. cell death was negli-
ible in the stationary yeast population used to start the thermalnactivation test.
Fig. 3A shows that the data of CFU ml−1 (full and empty rhom-us) coming from the samples assayed immediately (full rhombus),verlapped those obtained from the samples kept at room temper-ture for 24 h (empty rhombus). In fact, in both cases, CFU ml−1
ecreased according to the first order kinetics with very similar kdalues, being them 3.15 h−1 and 3.17 h−1.
The same overlap was not observable in the case of detectionf the RLUs ml−1 (Fig. 3A full and empty triangles). Indeed, theLUs ml−1 of the samples tested immediately after having beenollected from the bath, (Fig. 3A full triangles) were not hardlynfluenced by the high temperature because the intensity of theight signal did not diminish significantly, at least in the first twenty
inutes. Only after this time interval, the RLUs ml−1 began to fall−1
own. The RLUs ml decreased according to the first order kinet-
cs only when the samples were assayed after having been keptt room temperature for 24 h (kd = 0.72 h−1) (Fig. 3A empty trian-les). So it was evident that in the early phase of the exposure to
l Technology 52 (2013) 157– 162
the high temperature, i.e. until 20 min incubation at 53 ◦C, the ATPcontent of yeast cells at the moment of collection was higher thanthat of the same cells kept at room temperature for 24 h. After 1 hof exposure at 53 ◦C onwards, the situation reversed and the sam-ples which were kept at room temperature for 24 h, exhibited thehighest ATP level.
In the first 20 min of exposure at 53 ◦C, the maintenance ofRLUs ml−1 at the levels near to the initial ones could be ascribed toadditional cellular request for ATP synthesis to make possible theactivation of the pathways that yeast cells bring up in the effort tosurvive to heat shock [26]. On the contrary, since 1 h onwards, ther-mal death of yeast cells takes over, presumably due to the enhancedpermeability of mitochondrial membrane [27–29], responsible, inits turn, for the decrease in ATP production. In these circumstances,it is possible that some cells which have been defined as “resilient”,underwent only a slight mitochondrial swelling, which could becompletely reversible, causing the restoration of basal levels of ATPper cell, if not the ability to reproduce.
Relying on these observations and considering that all the cellsof the initial population were viable, the ratio between RLUs ml−1
and total count corresponded to the RLUs emitted per cell. So theentire yeast cell population subjected to thermal inactivation over a3 h test, and assayed after standing 24 h at room temperature, couldbe distributed (Fig. 3B) into three yeast sub-populations; one con-sisting of living cells that had preserved the ability to reproduce(evaluated by viable count), a second population formed by thecells that had irreversibly lost the ability to reproduce (evaluatedthrough the difference between the number of cells per ml obtaineddividing the RLUs ml−1 by the RLUs emitted per cell and those pro-vided by viable count), and a third one, formed by dead cells i.e.the cells that had irreversibly lost their structural and metabolicintegrity (evaluated by the difference between the total count andthe number of cells obtained by the bioluminescence assay at eachfixed collection time).
3.4. ATP-based bioluminescence assay as an useful tool todescribe the behaviour of yeast cells proliferating in a fed-batchreactor
The standardized bioluminescence assay has also been utilized,together with more traditional techniques, spectrophotometry andviable count, as a tool to monitor metabolic fluctuations which canoccur in S. cerevisiae, a glucose sensitive yeast, during prolongedoperations in fed-batch.
For this purpose, cells of the S. cerevisiae CEN.PK2-1C strain werecultivated in a fermenter working in semi-continuous mode and ata constant � value of 0.08 h−1 in order to promote a respiratorymetabolism in the yeast culture and so achieve high biomass yield.To this aim an exponentially increasing feeding profile was built up[30].
In Fig. 4, yeast proliferation was monitored by determinationof total RLUs and cell density, the latter spectrophotometricallyobtained (see Section 2.4) during the fed-batch process. The totalRLUs increased with a monotonous trend up to 25 h of feeding, thenthe total RLUs suddenly dropped. Conversely the data regardingthe increasing yeast biomass obtained by optical density, followeda monotonous trend over the entire time-course of exponentialfeeding (38 h).
The really interesting thing in the first 25 h of run was that therates of increase in total RLUs and total biomass were comparable(see the values of the exponent in the corresponding kinetic expres-sion in Fig. 4) and they approached the � value imposed when
the exponential feeding profile had been designed (0.08 h ). Thisbehaviour, together with the high biomass yield (0.5), experimen-tally determined (data not shown), accounted for a fully oxidativemetabolism in the first 25 h of feeding.
L. Paciello et al. / Enzyme and Microbia
RLUs = 6.18 ·1011e0.0811 t
R² = 0.98 4
X = 11 .8e0.0761 t
R² = 0.96 9
0
10
20
30
40
50
0
1
2
3
4
5
0 10 20 30 40
X [g
]
RLU
s·10-1
2
time [h]
Fig. 4. Fed-batch reactor carried out with S. cerevisiae CEN.PK2-1C strain: yeastpr(
cTitotfcd
4
icdclbtrcoocycbphbvrtmmuo
adct
roliferation monitored by bioluminescence and data reported as total RLUs (fullhombus) or spectrophotometrically and data reported as total biomass (grams d.w.)empty rhombus).
The factors that cause the sudden drop of RLUs after 25 h of run,ould be the onset of cell death and/or a metabolic fluctuation.he results obtained by viable count highlighted that the grow-ng biomass was not involved in death phenomena (kd = 0) duringhe entire fed-batch run (data not shown). Instead, the appearancef ethanol detected after 25 h of feed (data not shown), supportedhe hypothesis of a metabolic shift towards a mixed respiratory-ermentative metabolism so it seems obvious to assume that theontent of ATP per cell was gradually decreasing causing the RLUrop.
. Discussion
ATP, as the energy carrier of all cells, is best suited to correlatets amount with cell viability. Indeed, within minutes after death,ells lose the ability to synthesize ATP, and endogenous ATPasesestroy any remaining ATP. For this reason the commercialized kitsontain inhibitors of endogenous ATPases, in addition to luciferin,uciferase and other reagents necessary to measure ATP by using aioluminescent reaction.However, the cell viability assay based onhe detection of ATP, should be accurately investigated before beingoutinely employed, especially as regards the contact time betweenells and the lysing agent because contact time depends on the typef microorganism employed as shown in this work by using cellsf two strains, CEN.PK2-1C and BY4741 of the conventional yeast S.erevisiae. The choice fell on these strains for a preliminary study oneast cell viability determined through the ATP-based biolumines-ence assay because they have been considered as a good referenceeing commonly used in the research laboratories for heterologousrotein production as well [18,19,31]. In the present work, theyave been systematically tested for bioluminescence after havingeen kept in contact with the lysing agent for different time inter-als and then processed with AMR® (ATP Monitoring Reagent). Theesults obtained highlighted that the time required to completehe cell-wall digestion was much larger than that suggested by the
anufacturer (10 min) so that it would be preferable that the biolu-inescence assay was drawn up depending on the microorganism
sed. In the case of yeast, the cell-wall is more resistant to chemicalr physical disintegration than that of bacteria.
The intensity of the light detected through the bioluminescence
ssay is directly proportional to the concentration of ATP. The ATPetected, in turn, should be directly proportional to the number ofells growing under the same metabolic conditions. In this case,he average content of ATP per cell is constant. Actually, using
l Technology 52 (2013) 157– 162 161
bioluminescence to monitor the increase in cellular mass duringexponential growth can be misleading due to an overestimation ofthe maximum specific growth rate (�max) value. The assay is so sen-sitive to changes in ATP concentration that even a little increase inATP can be detected by luciferin–luciferase system. This techniqueshould be correctly employed to evaluate �max when the determi-nations are carried out during fully fermentative yeast metabolism.Similarly, a great attention must be paid when, after the exposure ofyeast cells to high temperatures, the bioluminescence assay is usedto evaluate the decrease in viable cells which is requested for thecalculation of the specific death rate (kd). Indeed, under stressfulenvironmental conditions, such as those arising by heat shock, ATPsynthesis could increase in response to the activation of metabolicpathways needed to cope with these adverse conditions.
Conversely, the ATP-based bioluminescence assay revealed tobe a useful tool to quantify sub-populations in a heterogeneousyeast population. Indeed, combining bioluminescence with bothviable and total cell count, it was possible to identify a new yeastsub-population which was defined as “resilient” because capableto survive not to reproduce in addition to the sub-populations ofsurvivors capable to reproduce and that of the dead cells. The seg-regation of a new sub-population of yeast is probably ascribable tothe ability of yeast cells to withstand the mitochondrial swellingwhich depends on enhanced mitochondrial membrane permeabil-ity occurring during thermal cell death. As far as this, it is wellknown that mitochondria play a key role in the mechanism ofboth apoptotic and necrotic yeast cell death [29] and that the mito-chondrial permeability transition (MPT), determines the swellingof mitochondria and the decrease in ATP production due to theuncoupling of oxidative phosphorylation [26,27,29].
Similarly, the bioluminescence assay used during fed-batchoperations, permitted in the short term and in combination withoptical density method, the monitoring of physiological changesthat can occur during the time-course of such a bioprocess.
5. Conclusions
In the light of the results obtained, while remaining unques-tionable the importance of the ATP-based cell viability test beingfast, relatively simple with respect to other methods and the idealchoice to detect microbial contamination, its sensitivity can be adisadvantage. In fact when we need to convert the ATP assayedinto viable microbial mass, the occurrence of metabolic oscilla-tions which cause fluctuations of the ATP content per cells, shouldbe taken into consideration. On the contrary, the bioluminescenceassay, if appropriately and carefully employed with other tech-niques, reveals potentialities which could be exploited to identifyand quantify yeast sub-populations to be used for example in seg-regated models or gain information on the physiological state of ayeast population in a stressful environment.
To conclude we would like to stress that choosing a cell viabilityassay among the different available options, can be a challengingtask. Understanding of what each assay is measuring as an end-point, and how the measurement correlates with cell viability arethe most important points because they allow to evaluate the lim-itations of the assay and its potentialities.
Acknowledgements
This work was supported by the University of Salerno fundsto Palma Parascandola in the framework of the research projects
“Proteine eterologhe da lievito: lo stress ambientale e la fisiolo-gia dell’ospite durante la produzione in sistemi fed-batch aerati”(FARB-2009) and “Studio della produttività in fermentatore di ceppidi lievito da impiegare nella produzione di biomasse e/o come ospiti
1 crobia
p(sA
R
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
62 L. Paciello et al. / Enzyme and Mi
er l’espressione di proteine eterologhe di interesse per l’uomo”FARB-2010). Special thanks go to Dr. Carolina Florio whose usefuluggestions and criticism contributed to the improvement of theTP-based bioluminescence assay.
eferences
[1] Kangas L, Grönroos M, Nieminen AL. Bioluminescence of cellular ATP:a new method for evaluating cytotoxic agents in vitro. Medical Biology1984;62:338–43.
[2] Crouch SPM, Kozlowski R, Slater KJ, Fletcher J. The use of ATP bioluminescenceas a measure of cell proliferation and cytotoxicity. Journal of ImmunologicalMethods 1993;160:81–8.
[3] McElroy WD. In: Colowick SP, Kaplan NO, editors. Methods in enzymology VI,vol. 6. New York: Academic Press; 1963. p. 775–7.
[4] McElroy WD. The energy source for bioluminescence in an isolated system.Proceedings of the National Academy of Sciences of the United States of America1947;33:342–5.
[5] McElroy WD. Properties of the reactions utilizing adenosinetriphosphate forbioluminescence. Journal of Biological Chemistry 1951;191:547–57.
[6] De Luca M. Firefly luciferase. Advances in Enzymology and Related Areas ofMolecular Biology 1976;44:37–68.
[7] Case JF, Strause LG. Neurally controlled luminescent systems. In: Herring PJ,editor. Bioluminescence in action. New York: Academic Press; 1978. p. 331–66.
[8] Garcìa-Campana AM, Baeyens WRG. Chemiluminescence. In: Analytical Chem-istry. New York: Marcel Dekker Publishers; 2001. pp. 529–549.
[9] Marques SM, Esteves da Silva JCG. Firefly bioluminescence: a mechanisticapproach of luciferase catalyzed reactions. IUBMB Life 2009;61:6–17.
10] Thore A, Ansehn S, Lundin A, Bergman S. Detection of bacteriurìa by luciferaseassay of adenosine triphosphate. Journal of Clinical Microbiology 1975;1:1–8.
11] Sharpe AN, Woodrow MN, Jackson AK. Adenosine triphosphate (ATP) lev-els in food contaminated by bacteria. Journal of Applied Bacteriology1970;33:758–67.
12] Chappelle W, Levin GV. Use of the firefly bioluminescence reaction for rapiddetection and counting of bacteria. Biochemical Medicine 1968;2:41–52.
13] Szabo A, Langer T, Schröder H, Flanagan J, Bukau B, Hartl FU. The ATP hydrolysis-dependent reaction cycle of the Escherichia coli Hsp70 system—DnaK, DnaJ, andGrpE. Proceedings of the National Academy of Science 1994;91(22):10345–9.
14] Pazzagli M, Devine JH, Peterson DO, Baldwin TO. Use of bacterial and firefly
luciferases as reporter genes in DEAE/dextranmediated transfection of mam-malian cells. Analytical Biochemistry 1992;204:315–23.
15] Suzuki N, Goto T. Study on firefly bioluminescence. II. Identification of oxy-luciferin as a product in the bioluminescence of firefly lanterns and in thechemiluminescence of the firefly luciferin. Tetrahedron 1972;28:4075–82.
[
l Technology 52 (2013) 157– 162
16] Petty RD, Sutherland LA, Hunter EM, Cree IA. Comparison of MTT and ATP-basedassays for the measurement of viable cell number. Journal of Bioluminescenceand Chemiluminescence 1995;10:29–34.
17] Stanley PE. Extraction of adenosine triphosphate from microbial and somaticcells. Methods in Enzymology 1986;133:14–22.
18] Paciello L, Romano V, de Alteriis E, Zueco J, Parascandola P. Performance of theauxotrophic Saccharomyces cerevisiae BY4741 as host for human Interleukin-1� production in an aerated fed-batch reactor. Microbial Cell Factories2009;8:70–82.
19] Landi C, Paciello L, de Alteriis E, Brambilla L, Parascandola P. Effect of aux-otrophies on yeast performance in aerated fed-batch reactor. Biochemical andBiophysical Research Communications 2011;414(3):604–11.
20] Paciello L, Andrès I, Zueco J, Bianchi MM, de Alteriis E, Parascandola P. Expres-sion of human interleukin-1� in Saccharomyces cerevisiae using PIR4 as fusionpartner and production in aerated fed-batch reactor. Annals of Microbiology2010;60:719–28.
21] Verduyn C, Postma E, Scheffers A, van Dijken J. Effect of benzoic acid onmetabolic fluxes in yeasts: a continuous-culture study on the regulation ofrespiration and alcoholic fermentation. Yeast 1992;8:501–17.
22] Hahn-Hägerdal B, Karhumaa K, Larsson CU, Gorwa-Grauslund M, Görgens J, vanZyl WH. Role of cultivation media in the development of yeast strains for largescale industrial use. Microbial Cell Factories 2005;4:31–46.
23] Enfors SO. Baker’s yeast. In: Ratledge C, Kristiansen B, editors. Basic biotech-nology. U.K.: Cambridge University Press; 2001. p. 377–89.
24] Porro D, Martegani E, Tura A, Ranzi BM. Development of a pH controlled fed-batch system for budding yeast. Research in Microbiology 1991;142:535–9.
25] Pronk JT. Auxotrophic yeast strains in fundamental and applied research.Applied and Environment Microbiology 2002;68:2095–100.
26] Trott A, Morano KA. The yeast response to heat shock. In: Hohmann S, MagerPWH, editors. Yeast stress responses, vol. 1. Heidelberg, Germany: Springer-Verlag; 2003. p. 71–119.
27] Tsujimoto Y, Shimizu S. Role of the mitochondrial membrane permeabilitytransition in cell death. Apoptosis 2007;12:835–40.
28] Dressler C, Beuthan J, Mueller G, Zabarylo U, Minet O. Fluorescence imaging ofheat-stress induced mitochondrial long-term depolarization in breast cancercells. Journal of Fluorescence 2006;16:689–95.
29] Azzolin L, von Stockum S, Basso E, Petronilli V, Forte MA, Bernardi P. Themitochondrial permeability transition from yeast to mammals. FEBS Letters2010;584:2504–9.
30] Paciello L, Romano F, de Alteriis E, Parascandola P, Romano V. Glucoamylaseby recombinant Kluyveromyces lactis cells: production and modelling of a fed
batch bioreactor. Bioprocess and Biosystems Engineering 2010;33:525–32.
31] van Dijken JP, Bauer J, Brambilla L, Duboc P, Francois JM, Gancedo C, et al.An interlaboratory comparison of physiological and genetic properties offour Saccharomyces cerevisiae strains. Enzyme and Microbial Technology2000;26:706–14.
