0099-2240/92/113709-06$02.00/0
Copyright © 1992,AmericanSociety forMicrobiology
Flocculence
of
Saccharomyces cerevisiae Cells Is Induced by
Nutrient
Limitation,
with
Cell
Surface
Hydrophobicity
as a
Major
Determinant
GERRIT SMIT,* MARIKA H. STRAVER, BEN J. J. LUGTENBERG,
ANDJAN W. KIJNE
InstituteofMolecularPlantSciences, Leiden University,
Nonnensteeg
3,
2311 VJLeiden,
TheNetherlandsReceived22 May1992/Accepted 17August 1992
Initiationofflocculation
ability
ofSaccharomyces cerevisiaeMPY1 cells was observed at the moment the cells stop dividing because of nitrogen limitation. A shift in concentration of the limiting nutrient resulted in a corresponding shift in cell division and initiation of flocculence. Other limitations also led to initiation of flocculence, with magnesium limitation as the exception. Magnesium-limited S. cerevisiae cells did not flocculateat anystageof growth. Cell surfacehydrophobicity
was found to be strongly correlated with the ability of the yeastcells to flocculate.Hydrophobicity sharply
increased at the end of the logarithmic growth phase, shortly before initiation of flocculationability.
Treatments of cells which resulted in a decrease inhydrophobicityalso yieldedadecrease in flocculation
ability.
Similarly, the presence ofpolycations
increased bothhydrophobicityand theabilitytoflocculate. Magnesium-limited cells were found to be stronglyaffected in cell surface hydrophobicity. A proteinaceous cell surface factor(s) was identified as a flocculin. This heat-stable component had a strong emulsiflying activity, and appears to be involved in both cell surfacehydrophobicityand inflocculation
ability
of theyeastcells. Flocculation of yeast (Saccharomyces cerevisiae) cells has been defined as "the phenomenon wherein yeast cellsadhere inclumps and sediment rapidlyfrom the medium in which they aresuspended" (21).Yeast cellsgain the ability
toflocculateattheendof thefermentationprocess(2, 8, 12,
14),whichmakes thiscell adhesionprocessofconsiderable
interest in brewing industry (5, 23). However, the causal
mechanism of initiationofflocculation ability is not known. Data on determinants of flocculation are scattered and
conflicting in the literature and concern different strains,
treatments, and growth conditions. Initially, flocculation
wasreported tobea processbasedsolely on ionic
interac-tions, with
Ca2"
ions actingas bridges between the yeast cells (13). A requirement forCa2"
inflocculation of yeast cells isgenerallyreported (see reference5forareview), but in some cases magnesium andmanganese ions may act as substitutes (11, 22). Since flocs can be dispersed in the presenceof sugars,mostnotablymannose(9, 11,17),itwas suggested that most likely a lectin-sugar interaction is in-volved in flocculation. Stratford and Keenan (24-26) pro-vided evidence that agitationwas required forinitiation offlocculation.Theseresults indicate that
physicochemical
cell surface interactionsmight also be involved in flocculation.Beavan andBelk(3) reportedacorrelationbetween floccu-lation and electrophoretic mobility of yeast cells under certain conditions, whereas others reported a correlation
between
hydrophobicity
and flocculation for some yeast strains (1, 8). In most cases, however, the data are notconclusive and do not cover the
possible
role ofspecific
interactions besides
nonspecific
adhesion. For instance,Kamada and Murata(8) foundflocculationnot tobecation
dependent, incontrast to mostreports, andAmoryetal.
(1)
only used one method to measure hydrophobicity and*
Corresponding
author.showed the results from only two measuring points during
growthof the yeast cells.
Taken together, the scattered results andthe differences between reports dealing with various factors that affect flocculationof different yeaststrains underdifferent growth andtestconditions hamperpropercomparison oftheresults.
The molecular basis of flocculation is still poorly
under-stood, despite the importance of this process in industrial
processes. The present study was initiated to determine characteristicsofflocculationunderstandardized conditions and, subsequently,toidentifyyeastcellsurface components involved inflocculation. Throughout this study, one strain
and defined
growth
conditionswereused. Nutrient limitation appearedtotriggeranincrease incell surface hydrophobic-ity and, concomitantly, flocculation ability. The resultsstronglysuggestthat both
Ca2"-dependent
sugarbindingandphysicochemical
interactions, mostnotablycell surfacehy-drophobicity,
are involved in the flocculation process of yeastcells.Furthermore,aproteinaceoussurfacecompound
with emulsifying activitywas identified as a flocculin, in-volved incell surfacehydrophobicity andflocculation.MATERIAIS
AND METHODSYeast strain and culture conditions. S. cerevisiae MPY1
used in these studies is a bottom-fermentation
brewery
isolate. The yeast cellsweremaintainedat4°Con wortagarslopes. The standard medium used contains
(per
liter ofdeionized water): maltose, 80.0 g;
KH2PO4,
3.0 g;MgCl2
6H20,
0.6 g;K2S04,
0.4 g;L-alanine,
0.107 g;L-arginine,0.126 g;
L-asparagine,
0.083 g;L-isoleucine,
0.078 g; DL-leucine, 0.158 g;L-lysine,0.103 g;L-proline,
0.369 g; L-serine, 0.069 g; L-threonine,0.063 g;L-tyrosine,
0.100 g; L-valine, 0.150 g;Ca2+-pantothenate,
1.0 mg;myoinositol,
25.0 mg; nicotinicacid, 0.2 mg;
biotin,
0.05 mg;EDTA,
30 mg;ZnCl2,
2.2 mg;CaCl2.
2H20,
18 mg;MnCl2.
2H20, 1.0
3709
on January 20, 2017 by WALAEUS LIBRARY/BIN 299
http://aem.asm.org/
APPL.ENVIRON. MICROBIOL.
mg; FeCl2. 4H20, 4.3 mg; Na2MoO4 2H20, 0.2 mg;
CuCl2. 2H20,
0.4 mg;CoCl2. 6H20,
0.3 mg; H3BO3, 1.0mg; andKI, 1.0 mg. In carbon-limitedmedium,theamount
of maltose in the medium was reduced to 16.0 g/liter.
Zn2+-limited
medium contains 44 ,ug ofZnCl2
per literinstead of 2.2 mg,andMg2+-limitedmedium contains 6 ,ug of
MgCl2. 6H20perliter instead of 0.6 mg.
Yeast cells were cultivated at 28°C in 100-ml cotton-plugged Erlenmeyer flasks containing 50 ml of standard medium
(180 rpm).
Yeast cells for isolation of flocculinswere grownin 2-literErlenmeyerflasks
containing
1.0 liter of standard medium. Growth was measured by the A620value
(after
properdilutions)
with a Novaspec IIspectro-photometer
(Pharmacia-LKB, Uppsala,
Sweden), by
direct cellcounting
with ahemacytometer,
andby dry weight
determinationof cells.Direct cell
countings
weredone in the presence of 10 mM EDTA, which preventsaggregation
of the cells. For dryweight determinations,
three culturesamples of 1.0 ml each were filtered over a
0.45-,um-pore-size membrane filter(Sartorius GmbH,
Gottingen, Germany)
and washed threetimes with deionized waterbeforedrying
at 80°Cfor 24 h.
Flocculation assay. The flocculation assay is essentially
based on the assay described by Miki et al.
(11),
with anumber ofmodifications. Briefly, after determination of the
A620 value of the culture, yeast cells were harvested by
centrifugation,
washed, andresuspended
in a 50 mM Naacetate-0.1%
CaCl2
buffer(pH
4.5) (flocculation buffer)
to afinalA620 value of 2.5. After 30 min of acclimatization at roomtemperature,650 p.1 of cellsuspensionwasaddedtoa
1.0-mlcuvette.With this
sample volume,
thelight
beam ofaNovaspecIIspectrophotometermonitors the
optical density
slightly
below the surface of the cellsuspension.
The cellsuspension
was whirlmixed for 20 s with a Vortex Genie apparatus at maximum speed; this was followed by five inversions of the cuvette.Immediately thereafter,
theset-tling profiles
werespectrophotometrically
determined(see
Fig.
1).
The maximal decrease inoptical density
perminutewaschosen as ameasureofflocculation
ability
of thecells.Very
strong flocculationresulting
in a decrease inoptical
density higher
than0.5/min
could not be measuredaccu-rately
and isrepresented
as >0.5. The influence ofadding
salts and monosaccharides was tested after these com-pounds were added to the flocculation buffer just beforewhirlmixing.
Thevariability
of thistestis about10%.Treatments of yeast cells. Yeast cells were harvested by
centrifugation
at anA620 value of 5.0 andsuspended in 10 mM Tris-HCl buffer (pH 7.5). Proteinase K and pronase(both
fromSigma)
wereaddedtothe cells inafinalconcen-tration of 0.1 and 1.0
mg.
ml-1, respectively, and thenincubatedfor 60 minat 28°Cundergentle agitation.
Subse-quently,
the yeast cells were harvested by centrifugation, washed threetimeswith deionized water, and resuspended in flocculation buffer to a finalA620 value of 2.5. Controlswereincubated for 60 min inTris-HClbufferonly.
Yeast cellsweresheared for 30 min in 50 mM Na-acetate buffer
(pH 4.5)
with a Sorvall Omnimixer (DuPont Instru-ments,Newton, Conn.)at asettingof 6.0.Subsequently, the cellswereharvested, washed, andresuspended in floccula-tion buffer.Determination ofphysicochemicalcell surface characteris-tics. Cell surface hydrophobicity of yeast cells was deter-mined bythree methods: (i) interaction of yeast cells with
hexadecane; (ii) adhesion of yeast cellstopolystyrene; and
(iii)
contactangle determinationsofwaterdroplets on a layerofyeastcells. Determination of the interaction of yeast cells
with hexadecane was done essentially as described previ-ously for bacteria (18). Two volumes of yeast cells,
sus-pended in flocculation buffer to anA620value of 2.5,were
whirlmixed for 20s with 1 volume of hexadecane(Sigma). After 30 min at room temperature, the interaction was
determinedastheamountofemulsification of the hexadec-anein thewaterphase.
Adhesion of yeast cellsto polystyrene wasmeasured as
follows: yeast cellswereharvested andsuspendedin 50 mM Na-acetate buffer (pH 4.5) to an A620 value of 2.5. Ten milliliters ofthissuspensionwastransferred intoastandard
petri plate (Greiner) and incubated for 2 hat room
temper-ature.Subsequently, the supernatant fluidwasremoved and the plate was rinsed five times under a gentle stream of deionizedwater. Adhesion of the cellsto thepetridishwas
then monitored with alight microscope.
Thecontact angle ofwater on a layerof yeast cellswas
determined essentially bythe method of Van Loosdrechtet
al. (27). Briefly,10.0 ml of yeast culturewasharvested, and the cellsweresuspendedinphosphate-bufferedsaline(PBS; NaCl, 8.5 g/liter; KH2PO4, 0.272g/liter; Na2HPO4 2H20, 1,424 g/liter; pH7.2) andcollected on a 0.20 pum-pore-size
Sartorius membrane filter. The number of cells usedyielded a film of cells with a thickness of approximately 50 cell
layers, asjudgedfrom the cellsize.With chitosan pretreat-ment, a 100-fold-diluted PBS was used, which prevents dissociation of chitosan from the cell surface. The PBS concentration itself didnotaffectthecontactangle,astested for untreated controls inwhich PBS was used in different concentrations up to 10 times the standard concentration. Filters were mounted on glass slides and air dried in the presence ofsilicagelforatleast3 h. No changein contact angle occurred after 3 h ofdrying, which is in accordance with results from Busscher et al. (4). Contactangles were
measureddirectly by usingamicroscopewithagoniometric
eye piece (Smit Science Control Systems, Leiden, The
Netherlands). Eachreportedcontactangleis themeanofat
least 10independentmeasurements.
Electrophoretic mobility, as a measure for the negative
surface charge of the yeast cells, was determined as de-scribed by Van Loosdrecht et al.
(28), using
a Doppler velocimeter with a ZetaSizer (Malvern Instruments,Mal-vein,
England). Flocculation buffer without CaCl2 (50mM Na acetatebuffer, pH 4.5)was used formobilitystudies.Isolationofflocculins.S. cerevisiaecellsweregrownin 1.0 liter of standard mediumto anA620 of 5.0to 6.0, harvested
by centrifugation,andsuspendedin 50 ml of 10 mMTris-HCl
buffer(pH 7.5).After shearinginanSorvall Omnimixer for 15 min at a setting of 6.0, the cells were removed by
centrifugation at 9,000 x g for 10 min, and the resulting supernatant fluidwasusedasthe cellsurface-derived prep-aration. Thispreparationwasfurtherpurified by
ultrafiltra-tionby theuseof membrane filters with nominal cutoffs of
10, 100, and300 kDa(Amicon Co., Danvers, Mass.). Floc-culin fractionswereheat treated by incubation of the sam-plesat100°Cfor15 min.Enzymatic digestionwasdone with 0.1 mg of proteinase K. ml-1 for 30 h at 28°C. The
pro-longedincubation time with protease was done to inactivate the enzyme itselfaswell.
RESULTS
Flocculation characteristics of S. cerevisiaeMPY1. Floccu-lation could be quickly andconveniently quantified by the decrease inturbidity ofayeast cell suspension in floccula-tion buffer withtime, which reflects the cell settling profiles
3710 SMIT ET AL.
on January 20, 2017 by WALAEUS LIBRARY/BIN 299
http://aem.asm.org/
A62o 1.0-I 0.8 - 0.6- 0.4-0 1 2 3 4 5 Time(min)
FIG. 1. Flocculation ofS. cerevisiaeMPY1cells in the presence of7 mMCaCl2 ( ) and 1 mM EDTA (---) as quantified by measuring the decreaseinA620 of a cell suspension. The decrease in thepresenceofEDTAobservedafter 3 to 4 min is due to settling of (nonflocculated)cells.
(Fig. 1). In the absence of flocculation, the turbidity drops slowly, due to settling of the cells after several minutes. Agitation of the yeast cells (by whirlmixing) was found to be essential for initiation offlocculation.
Flocculation occurred in the presence of Ca2, ions, whereas noflocculationwasobserved in the absence of
Ca2+
orinthe presence of EDTAorEGTA(Fig. 1). Mg2+,
Sr2+,
Cu2+, Mn2+, Fe3+, K+, andNa+ were not ableto replace
the Ca2+ ions, indicating that Ca2+ is specifically required
forflocculation ofMPY1cells.Flocculation was found to be
strongly dependentonthe pH of the flocculation buffer, with
anoptimumatpH4.5(Fig. 2).Noflocculationwasobserved above pH 6.0 or at pH 3.0. Flocculation is also highly sensitivetothe presenceof certain saccharides during
floc-culation(Table 1). Especially, mannoseand
a-methyl-man-nopyranosidewere found tobe very effective inhibitors of flocculationatconcentrations as lowas25 mM.
Nutrient limitation and initiation offlocculation. It is not
knownatwhichmomentduring growthflocculationability is initiated, nor what is the causal factor(s) in this process. Therefore,weexaminedinitiation of flocculationabilityina
defined medium, which enabled us to accurately vary the
composition of the growthmedium. Spontaneous
floccula-tionwas notobserved in the standardgrowthmediumused, unlessextraCaCl2wasaddedtothe medium. This indicates
thatalthoughthecells donotflocculate in the mediumitself,
which is apparently due to low Ca2+ conditions, they do become flocculent. To test whether a positive correlation exists between flocculence and growth limitation of yeast
cells, the flocculationabilityof MPY1cellswasdetermined
E 0.40 E 8 0.30 c 0.20 .2 ' 0.10' LL nnn.
TABLE 1. Influenceof presenceof variousmonosaccharides on flocculation of S. cerevisiae MPY1cellsa
Sugaradded Concn(mM)requiredto
abolish flocculation oa-D-Mannose... 25
a-Methyl-mannopyranoside
... 25 D-Glucose... 100 D-Trehalose... 100 Maltose... 1502-Deoxy-D-glucose
... 150a-Methyl-glucopyranoside
... 200 D-Lyxose... 250 a-L-Rhamnose... >250 D-Xylose... >250 D-Ribose... >250 D-Galactose... >250a
Monosaccharides
were added to the cells in theflocculation bufferjustbeforewhirlmixing.
during growthin standard defined medium, with appropriate changes in nutrient limitation.
During growth in batch culture,apositivecorrelation was observed betweenthe dryweight of the cells and the optical
density.However, these parameters and thenumber of cells per ml did not correlate at anA620 of more than 3.5. This result can be clearly visualized when the number of cells
during culture growthisplottedagainsttheoptical densityin
aso-calledn/Aplot(Fig. 3;seealsoreference10).AtanA620
of 3.5, the cell number in standard medium reached its
maximal value, whereas the optical density subsequently increased. These results indicate that already at an A620 of
3.5,thecellsaregrowth limited. Subsequently,theA620and thedryweightof the cells still increases, possiblyduetoan
increase in cell volume.Significantly,the stop incell division coincides with the initiation of flocculation ability as shown in Fig. 3.
Byaddition of various componentstostandardmedium,it wasfound that thenitrogen source was thegrowth-limiting
factor. Addition of either glutamate or an extra amount of the standard mixture of amino acids (the N source of the
medium) resultedinanupwardshift of the maximum number of yeast cells, the ultimate dry weight, and the ultimate
optical density. Moreover, a decrease of the amount of
0 S
01
c 4 - 5 7 8 9 3 4 5 6 7 8 9 pHFIG. 2. Influence of thepHof theflocculation bufferon floccu-lation of S. cerevisiae MPY1cells.InthepHrangeof 3.0to6.0,a 50 mM Na acetate-1% CaCl2 buffer was used, and a 10 mM
Tris-HCI-1% CaCl2 bufferwas used, and a 10 mM Tris-HCl-1%
CaCl2bufferwasusedforpH7.0to9.0.
cl 0 Cu 0 LL 0 1 2 3 4 5 6 7 8 9 A620nm
FIG. 3. Relationshipbetween theopticaldensity(A620),thecell number, and the ability toflocculate of S. cerevisiae MPY1 cells growingin standard medium. Note that themoment atwhich the cellsbecomeflocculent coincideswithcelldivision stop.
on January 20, 2017 by WALAEUS LIBRARY/BIN 299
http://aem.asm.org/
APPL.ENVIRON. MICROBIOL. TABLE 2. Influenceof the initialconcentrationofaminoacids
in, andaddition of glutamate to, the growth mediumonthe finalopticaldensity,final cellnumber,and finaldry
weight ofS. cerevisiaeMPY1 cells andoptical densityatwhich the cells become flocculent
Content Additional . . ...
of
acdia
amino combmnednitrogen
FinalAina
Fia Fnl Iittonf drywt cell no. flocculation %))nitrogen'
A6
(g/liter)
(108/ml) (620) 50 None 6.0 6.3 0.70 1.9 80 None 7.7 7.4 0.75 2.9 100 None 9.5 7.8 1.20 3.5 120 None 11.0 8.6 1.45 3.7 140 None 11.6 9.1 1.50 4.1 100 20(glutamate) 11.8 8.0 1.70 4.0aPercentageofthe amount present in standardmedium.
amino acid mixture added to the medium resulted in a
downward shift of cell number, optical density, and dry weightreached atthe stationary phaseofgrowth (Table 2).
Flocculationabilityof the yeast cellswasdeterminedduring
growth, and in all these cases, initiation of flocculationwas
shifted accordingly (Table 2).
Thepositivecorrelation betweenastopincell division and initiation of flocculation ability was not found only for N
limitation, but forothernutrientlimitations aswell. Chang-ing themediumcompositionin suchawaythat thelimiting
factorwas notthe N source but another
nutrient,
e.g.,the carbon source orZn2+,
againyieldeda positivecorrelation between the moment that the cells stop dividing and the initiation of flocculation ability. These results led us to hypothesize that flocculation ability is induced at the mo-mentcelldivision stops because of limitation foranutrient. One limitation tested, namely,Mg2+
limitation, differed fromthe other limitations mentioned above in that it resulted in yeast cells which were unable to flocculate during any stageofgrowth (Table 3).Physicochemical surface characteristics of yeast cells and TABLE 3. Influence of cultureconditions,shearing, and
treatmentwith proteaseorchitosanonhydrophobicity andflocculation ability ofS. cerevisiae MPY1cellsa
Hydrophobicity
Growth Optical Flocculation
medium' density Treatment Contact
Hexa-
ability(A620) angle decane (AA620/Min) (degrees) Std 2.3 None 53 2 +++ 0 Std 2.3 Chitosand 65 ±4 +++ 0.28 Std 5.9 None 68 + 3 +++ 0.36 Std 5.0 Chitosan 78 + 2 +++ >0.5c Std 5.0 Shearing 32 ± 0.03 Std 5.0 Protease <10 - 0
Mg2+
limited 3.0 None 33 + 2 ± 0 Mg2+limited 3.0 Chitosan 44+ 2 +++ 0.25aS. cerevisiae cells were harvested bycentrifugation andsuspended in
either50 mMsodiumacetatebuffer(pH 4.5) (shearing)or 10mMTris-HCl buffer (pH 7.5) (protease treatment). For details about treatments, see Materials and Methods.
bStd,standardmedium.Mg2+-limitedmedium contains 1% of theMgCl2 concentration of standard medium. The turbidity of 3.0 represents cells harvestedattheend of thegrowthphase.
cFlocculationcould even beobservedin the absence ofwhirlmixing. dChitosanwasused inafinalconcentrationof1,ug/ml.
1.5 S 01 0 1.2 0.9 0.6 0.3 0 -80 0 0
.00i0
0 A I *R~~~~~~~6 , y < ^*A*
70 AlAr I A 560 0 2 3 4 6 8 50 o 1 2 3 4 5 6 7 8 9 I c a) 0) c cJ A620nmFIG. 4. Relationship between the optical density (A620), cell number,andcellsurfacehydrophobicity. Thenumberofyeastcells (0) and the contact angle (A) areplotted against the A620 of the culture. Notethat the increase inhydrophobicity precedes the cell division stopaswellasinitiation of flocculationability (Fig. 3).
flocculation.We analyzed whether cellsurface hydrophobic-ity and surface charge are important in the flocculation
process of S. cerevisiae MPY1 cells. Hydrophobicity and electrophoretic mobility of yeast cells were determined during growth in standard medium, during growth under
Mg2+ limitation, and after various treatments of the yeast cells. Flocculent yeast cellsappearedtoshowahighcontact anglewithwater(Fig.
4;
Table3)and toadherestronglyto polystyrene(datanotshown).Whenthe interaction between flocculent yeast cells andhexadecanewas studied by light microscopy, the yeast cellsappearedtoformamonolayerofcellsaround each hexadecanedroplet,resulting information
of an emulsion of hexadecane in water (data not shown). These results demonstrate that flocculent yeast cells are highly hydrophobic.
Treatment of flocculent yeast cells with proteolytic
en-zymes resulted in a strong decrease in hydrophobicity as judged fromstronglydecreasedvaluesofcontactangles and from the abolishment of emulsifying activity of the yeast cells (Table 3). Such treatment also abolished flocculation
abilityof the cells(Table 3).Addition of either monosaccha-rides (e.g., mannose) in concentrations up to 1,000 mM or EDTA (25 mM), which both inhibit flocculation, did not affect the cell surfacehydrophobicityof yeastcells asjudged from hexadecane interactions (data not shown). Also, changes in the pH of the buffer in which the interaction with hexadecane is determined, in the range of pH 2.0 to 10.0, had
nosignificant effecton hydrophobicity of the yeast cells. Hydrophobicity of the yeast cells was found to increase
rapidlyatthe end ofthe logarithmic phase during growth in standard medium (Fig. 4). Significantly, this increase in hydrophobicity shortly precedes the moment at which the cells stop dividing and become flocculent (Fig. 3). In con-trast, growth of the yeast cells in
Mg2+-limited
medium resulted, in addition to absence of flocculence, in strongly reduced cell surface hydrophobicityasjudged from the lowcontactangle between water andMg2+-limitedcells, in poor adhesion abilitytopolystyrene, and in lack of emulsifying activity (Table 3).
No significant differences in the electrophoretic mobility of yeast cellswereobserved under all conditions tested (data
notshown). This indicates that the negative surface charge of the cells is most likely not significantly affected during growth and by the various treatments used.
3712 SMIT ET AL.
on January 20, 2017 by WALAEUS LIBRARY/BIN 299
http://aem.asm.org/
TABLE 4. Influence of growth conditions and treatments on emulsifying and flocculation-stimulating activity of cell
surface-derivedpreparations of S. cerevisiae MPY1 Growth Treatment ofcell Flocculation- Emulsifying medium' surface-derived stimulating activity
prepn activity'
Standard None + +++
Standard Heath + +++
Standard Proteaseb -
-Mg2+limited None
ICellsurface-derivedpreparationswereobtainedfrom yeastcellsgrown in
standardmedium andMg2+-limited medium and harvestedatthe momentcell division stopped.
bSee Materials and Methods.
cStimulation offlocculationwastested by addition ofcellsurface prepa-rationstoflocculentyeastcellsgrownin standardmedium andharvestedat an A620of4.0.
Recently, Rosenberg and coworkers (7) reported that adhesion of microorganisms to polystyrene could be en-hanced in the presence ofpolycations, such as chitosan and
poly-L-lysine.We tested theeffectof treatment of yeast cells
with these polycationsand compared these treated cells with
control cells for hydrophobicity, cell surface charge, and flocculation ability. Low-molecular-weight poly-L-lysine
was used, since high-molecular-weight poly-L-lysine was foundto causelysis of yeast cells(6).Addition of chitosan (1
,ug/ml) orpoly-L-lysine (1
mg/ml)
significantly increased the hydrophobicity of the cells(Table 3), whereas this treatment didnotsignificantlyaffect electrophoretic mobility(data notshown). Polycation-treated yeastcells were not only more
hydrophobic, but also showed a stronger ability to floccu-late. Chitosan treatment resulted even in spontaneous floc-culation in the absence ofwhirlmixing. Moreover,
Mg2+-limited yeast cells which normally do not flocculate and exhibitastronglyreduced cell surfacehydrophobicitywere partially restored for these characteristics after treatment
withpolycations (Table 3). Furthermore,nonflocculent cells harvested at low
A620
values and treated with chitosanshowedanincrease incontactangleto avalue comparable withthatofflocculent cells(65and 68degrees,respectively)
and concomitantly became flocculent (Table 3). Taken to-gether, these results clearly show that hydrophobicity of yeast cells is positively correlated with their ability to
flocculate. In all cases, flocculation appeared to be
Ca2+
dependentandmannosesensitive.
Isolation and preliminary characterization of yeast floccu-lin. A flocculin was experimentally defined as a surface
componentof yeastcellsabletoinhibitorenhance
(depend-ingonitbeing monovalentormultivalent)flocculation of S.
cerevisiae cells when added during the flocculation assay.
After treatment of the yeast cells by shearing forces, we found thatthe cell surface
hydrophobicity
as well as theirabilitytoflocculatewasstronglyreduced
(Table 3).
Cellsdidnotlooseviabilityafterthistreatment.The
resulting
surface-derived preparation, obtained after this treatment, was
tested for the presence offlocculins. Addition of this
prep-arationtountreated cells in the flocculation assay resulted in
an enhancement offlocculation in a
Ca2+-dependent
andmannose-sensitiveway, indicating thatthis fraction indeed contained one or more flocculins (Table
4).
The flocculinpreparation could even induce flocculation when added to
(nonflocculent)
Mg2e-limited
cells (data notshown).
Fur-thermore, this fraction had a high
emulsifying activity,
similar to intact MPY1 cells(data
notshown).
A cellsurface-derivedpreparationisolated from cells grown under
Mg2e-limiting
conditions neither stimulated flocculation norcontainedsignificant emulsifying activity (Table 4).
Treatment withproteinase K abolished both the
floccula-tion-stimulating activity and the emulsifying activity
com-pletely, whereas heat treatment didnotsignificantlyaffectits
activity(Table 4). Ultrafiltrationoftheflocculinpreparation
revealed the molecular massof the flocculation-stimulating and
emulsifying
component to be higher than 300 kDa. However, afterheattreatmentof thepreparation, nearlyallactivitypassed througha100-kDa-cutoff membrane butwas
retained by a 10-kDa-cutoffmembrane, indicating that the >300-kDa flocculin fraction is a complex which can be reduced in sizeby heat treatmentwithout loss ofactivity.
Takentogether,these resultsstronglyindicate thata hydro-phobiccell surfacecompoundis involved in yeast
floccula-tion.
DISCUSSION
Characterization of flocculation mechanism of S. cerevisiae MPY1. Flocculation of MPY1 cells(i) requires agitation, (ii) requires calcium, (iii) is highly sensitive to mannose and
mannose derivatives, and (iv) is pH dependent. These
re-sultsindicate thata
Ca2"-dependent
lectin-sugarinteraction is involved in flocculation of MPY1 cells.Flocculence is initiatedby nutrient limitation. In thedefined standard mediumused, the component whichfirstbecomes limited is thenitrogensource. Atthemomentthe cells stop
dividing, because ofthis limitation, the abilityto flocculate increases rapidly (Fig. 3). An increase or decrease of the initial amountofnitrogen in the medium results inashift in the moment at which the cells stop dividingand
concomi-tantlybecomeflocculent
(Table
2).This correlation betweengrowth limitation of the cells and initiation of
ability
toflocculatewas also observed for othernutrient limitations,
with
Mg2e
limitationbeinganexception(Table 3).Addition ofMgCl2during
the flocculation assayor addition ofextraCaCl2 during growth did not restore the ability of Mg2+-limited cells to flocculate (data not
shown), indicating
thatMg2+
is an essential nutrientrequired
forsynthesis
of aflocculin or involved in a moregeneral
feature,
e.g.,mem-brane stability (see also references 15 and
16).
Previous reports from our laboratory on attachment of Rhizobium bacteria have also shown thatgrowth
limitation acts as atrigger
for initiation ofcelladhesionphenomena
(10, 19, 20).
Thus,limitation-induced flocculation
ability
mightillustratea more
general
feature ofmicroorganisms.
Future research will focus on initiation of flocculation
during
fermentation inwort.Thisis necessarytoaddress thequestion
whether flocculation inwort isprimarily
initiatedby the same mechanism
and/or by
flocculation-promoting
compounds
releasedduring
fermentation,
asproposed
in a number of reports(see
reference5 for areview).
Positivecorrelation betweencell surface
hydrophobicity
and flocculence. MPY1 cells grown in standard medium werehydrophobic.
Several observations support thehypothesis
that cell surfacehydrophobicity
is amajor
determinant in flocculation of yeast cells:(i)
hydrophobicity
of MPY1cellssignificantly
increasesshortly
before initiation offloccula-tion
(Fig.
4); (ii) growth
ofMPY1 cells underMg2+-limiting
conditions resulted ina concomitant decrease in
hydropho-bicityand flocculence
(Table
3);
(iii)
treatmentof nonfloccu-lentcells,
either harvestedatlowoptical
densitiesorgrown underMg2e-limiting
conditions,
withpolycations
renders these cells both morehydrophobic
as well as flocculenton January 20, 2017 by WALAEUS LIBRARY/BIN 299
http://aem.asm.org/
APPL.ENVIRON. MICROBIOL. (Table 3); (iv)treatmentof flocculent cells withproteasesor
withshearingforces resultedinastrong decreaseboth in cell
surfacehydrophobicityand inabilitytoflocculate(Table3).
Electrophoretic mobility of MPY1 cells did not change
significantly under all conditions tested (Table 3).
Appar-ently,
cell surface charge is not correlated with any ofthedifferencesfound in flocculence of MPY1 cells.However,it
is important to recognize the possible role of(nonspecific) electrostatic
repulsion
inflocculation,
since without thisrepulsion, selective cell-cell adhesioncannotfunction. Itisimportanttonotice thatmonosaccharides and
Ca2"
aswell as variations in pH had no effect on cell surface
hydrophobicity.
Besides theprofoundeffect ofhydrophobic-ity
onflocculationability,involvement ofaCa2+-dependentlectin-sugar binding
is essential for flocculation. The strong correlation between flocculence and cell surfacehydropho-bicity
indicates that theregulation
of flocculationmightbecontrolled
by
theexpression
of this surface characteristic.Identification ofa
hydrophobic
surfaceproteinas afloccu-lin.Shearingof MPY1 cellsresulted inpoorlyflocculentcells
with reduced cell surface
hydrophobicity (Table 3).
The cell surface-derived fraction obtained after such a treatmentpossessed
bothemulsifying
andflocculation-stimulating
ac-tivity, indicating
that this fraction containsaflocculin.Boththe
emulsifying
aswellastheflocculation-stimulating
activ-ity
were found to be protease sensitive(Table 4).
This corroborates the results from protease treatment onwhole yeast cells(Table 3). Furthermore,
a cell surface-derivedpreparation
fromMg2+-limited
cells showedneitherfloccu-lation-stimulating
noremulsifying activity (Table 4).
Bothemulsifying activity
andflocculation-stimulating activity
of the cell surface-derivedpreparation
were found in ahigh-molecular-weight complex,
which after heattreatmentcould bedegraded
into smallerfragments.
Theseresults stronglyindicate that a
heat-stable, proteinaceous
cell surfacecom-ponent is involved in cell surface
hydrophobicity
and floc-culationability
ofS. cerevisiae MPY1cells, adding
weighttothe
hypothesis
that cell surfacehydrophobicity
is a majorfactor in flocculation. Future research will focus on
purifi-cation, characterization,
andregulation
of thisemulsifier,
theputative
Ca2+-dependent
lectin andits receptor.ACKNOWLEDGMENTS
We thank HuubReynaardsforhelpfuldiscussionsandassistance in contact angle determinations and Mark van Loosdrecht for assistance withmeasurementsofelectrophoretic mobility.
Theinvestigations were supportedby agrant from the Nether-landsMinistryofEconomical Affairs.
G.S. and M.H.S.have madeequalcontributionstothiswork. REFERENCES
1. Amory,D.E.,P. G. Rouxhet,andJ.P. Dufour.1988. Floccu-lence ofbreweryyeasts and their surfaceproperties:chemical
composition, electrostatic chargeandhydrophobicity. J. Inst. Brew.94:79-84.
2. Amri,M.A.,R.Bonaly,B.Duteutre,and M.Moll.1982.Yeast flocculation: influenceof nutritional factorsoncell wall compo-sition. J. Gen. Microbiol. 128:2001-2009.
3. Beavan,M.J.,and D. M.Belk.1979.Changesinelectrophoretic
mobilityandlyticenzymeactivity associated withdevelopment of flocculating ability in Saccharomyces cerevisiae. Can. J. Microbiol. 25:888-895.
4. Busscher, H.J.,A. H.Weerkamp,H.C. van der Mei, A. W. J. vanPelt,H.P. deJong,andJ.Arends. 1984.Measurementofthe surface freeenergy ofbacterialcell surfaces andits relevance for adhesion.Appl. Environ.Microbiol. 48:980-983.
5. Calleja,G. B. 1987.Cellaggregation,p. 164-237.In A.H. Rose
andJ. S. Harrison(ed.), Theyeast,2nded., vol. 2. Academic Press, Inc., (London), Ltd., London.
6. DeNobel, J. G., F. M.Klis, T. Munnik, J. Priem, and H. van der Ende. 1990. An assay ofrelative cell wallporosity in Saccha-romycescerevisiae, Kluyveromyces lactis, and Schizosaccha-romycespombe.Yeast6:483-490.
7. Goldberg, S., R. J. Doyle, and M. Rosenberg. 1990. Mechanism of enhancement of microbial cell hydrophobicity by cationic
polymers.J.Bacteriol. 172:5650-5654.
8. Kamada, K., and M. Murata. 1984. On the mechanism of brewer'syeast flocculation.Agric. Biol. Chem. 48:2423-2433. 9. Kihn, J. C.,C. L.Masy,M. M.Mestdagh,and P.G. Rouxhet.
1988. Yeast flocculation: factorsaffectingthemeasurementof flocculence. Can. J. Microbiol. 34:779-781.
10. Kline, J. W., G. Smit, C. L. Diaz, and B. J. J. Lugtenberg. 1988. Lectin-enhanced accumulation of manganese-limited Rhizo-biumleguminosarum cells on pea roothairtips. J. Bacteriol. 170:2994-3000.
11. Miki,B. L.A.,N.H.Poon,A. P.James,and V. L.Seligy. 1982. Possible mechanism for flocculationinteractionsgoverned bygene
flolinSaccharomycescerevisiae. J. Bacteriol. 150:878-889. 12. Mild,B. L.A.,N. H.Poon,A. P.James,and V. L.Seligy. 1982.
Repressionandinduction offlocculation interactions in Saccha-romycescerevisiae.J. Bacteriol. 150:890-899.
13. Mill, P. J. 1964. The effect ofnitrogenous substanceson the time offlocculation inSaccharomycescerevisiae. J. Gen. Mi-crobiol.35:53-60.
14. Nagarajan, L.,andS. Umesh-Kumar.1990.Antigenic studieson flocculating brewer's yeast, Saccharomyces cerevisiae NCYC227. J. Gen. Microbiol. 136:1747-1751.
15. Nishihara, H. 1977. Effect of chemical modification of cell surfacecomponents ofabrewer'syeastonfloc-formingability. Arch. Microbiol.115:19-23.
16. Nishihara,H.1982.Flocculationofcell walls ofbrewer's yeast and effects of metal ions, protein-denaturants and enzyme treatments.Arch. Microbiol. 131:112-115.
17. Nishihara, H., and T. Toraya. 1987. Essential roles of cell surfaceproteinandcarbohydratecomponentsinflocculation of abrewer'syeast.Agric. Biol. Chem. 51:2721-2726.
18. Rosenberg, M.,D.Gutnick,andE.Rosenberg.1980.Adherence of bacteria to hydrocarbons: a simple methodfor measuring cell-surfacehydrophobicity.FEMS Microbiol.Lett.9:29-33. 19. Smit, G., J.W.Kijne,andB.J. J. Lugtenberg.1986.Correlation
betweenextracellularfibrilsandattachment ofRhizobium legu-minosarumtopearoothairtips.J. Bacteriol. 168:821-827. 20. Smit, G.,T.J. J. Logman,M.E. T. I.Boermgter, J. W.Klne,
and B.J. J.Lugtenberg. 1989. Purification andpartial charac-terization of the
Ca2"-dependent
adhesin fromRhizobium legu-minosarum biovar viciae, which mediates the first step in attachment of Rhizobiaceae cells to plant root hair tips. J. Bacteriol.171:4054-4062.21. Stewart,G. G. 1975.Yeastflocculation.Practicalimplications andexperimental findings.Brew. Dig.50:42-62.
22. Stewart,G.G.,and T. E.Goring. 1976. Effectofsome monov-alent and divalent metal ions on the flocculation of brewers yeaststrains.J.Inst. Brew. 82:341-342.
23. Stewart, G. G., and I. Russell. 1986. The relevance of the flocculation properties ofyeast in today's brewery industry. Eur. Brew.Conv.Mon. 12:53-68.
24. Stratford, M.,and M. H.J.Keenan.1987. Yeastflocculation: kineticsandcollisiontheory. Yeast 3:201-206.
25. Stratford, M., and M. H.J.Keenan. 1988. Yeastflocculation:
quantification.Yeast4:107-115.
26. Stratford, M.,and M. H.J.Keenan.1988. Yeastflocculation:a
dynamic equilibrium.Yeast4:199-208.
27. VanLoosdrecht,M.C.M., J. Lyklema,W.Norde, G. Schraa, andA.J.B. Zehnder. 1987. The roleof bacterial cell surface
hydrophobicityinadhesion.Appl. Environ.Microbiol. 53:1893-1897.
28. VanLoosdrecht,M. C.M., J.Lyklema,W. Norde, G. Schraa, and A.J.B.Zehnder. 1987.Electrokinetic potentialand
hydro-phobicity as a measurement to predict the initial steps of bacterial adhesion.Appl.Environ.Microbiol.53:1898-1901. 3714 SMIT ET AL.
on January 20, 2017 by WALAEUS LIBRARY/BIN 299
http://aem.asm.org/