Growth phase-associated changes in the proteomeand transcriptome of Lactobacillus rhamnosus GG inindustrial-type whey mediummbt_275 746..766

Kati Laakso,1 Kerttu Koskenniemi,2

Johanna Koponen,3 Matti Kankainen,3 Anu Surakka,1

Tuomas Salusjärvi,1 Petri Auvinen,3 Kirsi Savijoki,4

Tuula A. Nyman,3 Nisse Kalkkinen,3

Soile Tynkkynen1 and Pekka Varmanen4*1Research and Development, Valio Ltd, Helsinki,Finland.2Department of Veterinary Biosciences, 3Institute ofBiotechnology and 4Department of Food andEnvironmental Sciences, University of Helsinki, Helsinki,Finland.

Summary

The growth phase during which probiotic bacteria areharvested and consumed can strongly influence theirperformance as health-promoting agents. In thisstudy, global transcriptomic and proteomic changeswere studied in the widely used probiotic Lactobacil-lus rhamnosus GG during growth in industrial-typewhey medium under strictly defined bioreactor con-ditions. The expression of 636 genes (P � 0.01) and116 proteins (P < 0.05) changed significantly overtime. Of the significantly differentially producedproteins, 61 were associated with alterations at thetranscript level. The most remarkable growth phase-dependent changes occurred during the transitionfrom the exponential to the stationary growth phaseand were associated with the shift from glucose fer-mentation to galactose utilization and the transitionfrom homolactic to mixed acid fermentation. Further-more, several genes encoding proteins proposed topromote the survival and persistence of L. rhamno-sus GG in the host and proteins that directly contrib-ute to human health showed temporal changes inexpression. Our results suggest that L. rhamnosusGG has a highly flexible and adaptable metabolismand that the growth stage during which bacterial cells

are harvested and consumed should be taken intoconsideration to gain the maximal benefit from probi-otic bacteria.

Introduction

Lactobacillus rhamnosus GG (ATCC 53103) is one of themost intensively studied probiotic strains worldwide. It hasmost of the characteristics generally desired for a goodprobiotic bacterium, including the ability to survivepassage through and to colonize the human gastrointes-tinal tract (GIT) (Goldin et al., 1992; Alander et al., 1997;1999). Adhesion capacity is a prerequisite for coloniza-tion, and L. rhamnosus GG has been shown to adhereefficiently to both intestinal mucus and epithelial cells(Jacobsen et al., 1999; Tuomola et al., 2000; Ouwehandet al., 2001). Consumption of L. rhamnosus GG promoteshuman health by reducing the risk of nosocomialrotavirus-related diarrhoea in infants (Szajewska et al.,2001) and shortening the duration of acute diarrhoea(Szajewska et al., 2007, recently reviewed by Guarinoet al., 2009). In addition to its influence on the GIT,L. rhamnosus GG have other beneficial effects, includingreducing upper respiratory tract infections among childrenin day care (Hatakka et al., 2001; Hojsak et al., 2010),decreasing the risk of developing atopic eczema (Kal-liomäki et al., 2001; 2003; 2007), alleviating the symp-toms of eczema (Majamaa and Isolauri, 1997; Isolauriet al., 2000) and decreasing the risk of dental caries inchildren (Näse et al., 2001). So far, the molecular mecha-nisms identified as being responsible for the health-promoting effects of lactobacilli mainly relate toimmunomodulation, intestinal epithelial barrier protectionand microbial balance maintenance (reviewed by Lebeeret al., 2008; 2010). The recently published genomesequence of L. rhamnosus GG (Kankainen et al., 2009)will boost mechanistic and functional genomic research ofthis well-known probiotic strain.

In industrial production, dairy strains, including probiot-ics, are commonly harvested during the late exponentialor stationary growth phase to ensure high cell numbers.However, the desired probiotic-associated factors are notnecessarily expressed at the highest levels during thestationary phase of growth. For example, cell surface

Received 2 March, 2011; revised 13 May, 2011; accepted 16 May,2011. *For correspondence. E-mail pekka.varmanen@helsinki.fi; Tel.(+358) 9 19157057; Fax (+358) 9 19158460. Conflict of interest state-ment: K.L., A.S., T.S. and S.T. are employed by Valio Ltd, whichproduces and markets L. rhamnosus GG strain.

Microbial Biotechnology (2011) 4(6), 746–766 doi:10.1111/j.1751-7915.2011.00275.x

© 2011 The AuthorsMicrobial Biotechnology © 2011 Society for Applied Microbiology and Blackwell Publishing Ltd

properties of L. rhamnosus GG have been shown tochange during growth, influencing the strain’s ability toadhere to epithelial cells (Deepika et al., 2009). Further-more, consumption of another widely recognized probioticbacterium, Lactobacillus plantarum, at different growthphases (mid-exponential or stationary) induced differenttranscriptional responses in human duodenal mucosa(van Baarlen et al., 2009). Hence, when aiming to opti-mize the health-promoting properties of probiotic bacteriain large-scale industrial cultivations, it is essential to deter-mine the most favorable growth conditions and phase forcell harvesting.

Functional genomics and proteomics analyses areuseful tools for identifying the molecular mechanismsbehind the health-promoting features of probiotic lacticacid bacteria. Transcriptomics provides detailed informa-tion about global changes in gene expression underparticular conditions. Furthermore, microarrays enablereliable measurement of the expression of all genes onthe array, regardless of the subcellular location of theircorresponding proteins. However, only changes in mRNAlevel can be measured with this tool, while most biologicalphenomena result from protein function. In addition,changes occurring at the post-transcriptional level (due toproteolysis, charged modification, etc.) can be investi-gated only with proteomic tools. Therefore, it is advisableto use both transcriptomic and proteomic approacheswhen linking genomic sequences to potential biologicalfunctions. Transcript or protein level approaches havebeen used to investigate global changes in other probioticlactic acid bacteria during growth transitions (Cohen et al.,2006; Koistinen et al., 2007; Azcarate-Peril et al., 2009).For example, transcriptomic analysis of Lactobacillus aci-dophilus cultured in skim milk indicated temporal expres-sion changes in genes involved in the proteolytic system,carbohydrate utilization, adhesion and cellular responsesto stress (Azcarate-Peril et al., 2009). In L. plantarum,protein expression profiles demonstrated differences inseveral anabolic and stress response pathways duringdifferent growth phases (Cohen et al., 2006). A compari-son of the proteomes of two L. plantarum strains duringgrowth demonstrated the growth phase-dependentexpression of proteins associated with energy metabo-lism, such as glycolysis, the phosphoketolase pathwayand ribose metabolism (Koistinen et al., 2007).

The global omics-level analyses conducted on theprobiotic L. rhamnosus GG so far include two proteomicstudies (Koskenniemi et al., 2009; Sánchez et al., 2009)and a study combining both proteomic and transcrip-tomic approaches (Koskenniemi et al., 2011). Sánchezand colleagues (2009) tested different extractionmethods for surface-associated proteins, and couldidentify 18 different possibly surface-associated proteinsof L. rhamnosus GG. Comparison of the proteomes of

L. rhamnosus GG grown in an industrial-type wheymedium (the same as used in this study) and in a labo-ratory medium demonstrated fundamental effects ofculture conditions on the protein production of L. rham-nosus GG (Koskenniemi et al., 2009). In a recent study,transcript and protein level responses of L. rhamnosusGG towards bile were shown to include diverse and spe-cific changes in general stress responses as well asalterations in cell envelope-related functions (Koskenni-emi et al., 2011). The present study involves comple-mentary approaches, i.e. DNA microarray analysescombined with two-dimensional difference gel electro-phoresis (2-D DIGE), to assess global changes in thetranscriptome and proteome of probiotic L. rhamnosusGG during growth in industrial-type whey medium understrictly controlled bioreactor conditions. In particular, thisstudy aimed to determine whether potential probiotic-associated factors were affected in a growth phase-dependent fashion during fermentation in whey.

Results and discussion

Expression profile of L. rhamnosus GG during growth inindustrial-type whey medium

In the present study, 2-D DIGE and whole-genome DNAmicroarrays were used to determine how global geneexpression in the widely used probiotic, L. rhamnosusGG, changed over time during growth in industrial-typewhey medium under strictly defined bioreactor conditions(Fig. 1). Overall, the transcript levels of 636 genes weresignificantly changed (� 2-fold difference, P � 0.01) at

Fig. 1. Growth curve of L. rhamnosus GG in whey medium at 37°Cat pH 5.8. Cell density values represent the mean of four individualpH-controlled bioreactor cultivations, and error bars represent thestandard deviation. RNA samples were taken at five different timepoints representing mid-exponential, late exponential, stationarytransition point, early stationary and late stationary phases, whileprotein samples were collected at four different time points (thesame as the RNA sampling phases, excluding the stationarytransition point). The time points at which the RNA (R) and protein(P) samples were taken are indicated by arrows.

Growth phase-associated changes in L. rhamnosus GG 747

© 2011 The AuthorsMicrobial Biotechnology © 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Microbial Biotechnology, 4, 746–766

one or more time points, representing 22% of the com-plete genome of L. rhamnosus GG (Fig. 2). This is com-parable to results from a previous study, where 21% of theORFs of the L. acidophilus genome were differentiallyexpressed during shift from exponential to stationarygrowth phase in skim milk (Azcarate-Peril et al., 2009).Expression profiles of the 636 significantly differentiallyexpressed genes were clustered into 16 groups (Fig. S1,Table S1). A clear difference could be seen between thetranscription profiles of the mid-exponential (4 h) and sta-

tionary growth phases (20 and 31 h), although only a fewgenes were down- (five genes) or upregulated (eightgenes) during the whole progression of growth (Fig. 2).Interestingly, clusters of orthologous groups (COGs) clas-sification of significantly differentially expressed genesshowed that the majority of genes regulated at time points12, 16, 20 and 31 h, compared with the 4 h time point,could not be included in any COG category (Fig. 3). At the20 h time point, Carbohydrate transport and metabolism(G), Energy production and conversion (C) and Transcrip-tion (K) seemed to be coordinately upregulated whencompared with 4 h, while most of the downregulatedgenes were involved in Amino acid transport and metabo-lism (E) or were categorized as ‘General function predic-tion only’ (R). At 31 h, most of the upregulated genes wereassigned to the G and C categories, and the majority ofdownregulated genes belonged to the Translation, riboso-mal structure and biogenesis (J) and Nucleotide transportand metabolism (F) categories.

Changes in protein production in L. rhamnosus GGduring growth were studied at 4, 12, 20 and 28 h. Usingthe 2-D DIGE technique, proteome composition at eachtime point was compared with the 20 h time point inthree separate DIGE experiments (Table S2). The 20 htime point was selected as the baseline because it wasassumed to involve the broadest selection of differentproteins. In total, 267 protein spots showing a statisti-cally significant difference (� 1.5-fold difference,P < 0.05) in abundance were detected and cut out fromthe gels, and 201 of these could be identified usingMALDI-TOF mass spectrometry and/or LC-MS/MS(Fig. S2, Tables 1–3 and Tables S3–S5). In seven cases,two proteins were identified from one spot. Conversely,30 proteins were differentially abundant and identified inmore than one experiment, and 26 proteins were foundto occur as different isoforms, i.e. they were identified intwo or more separate spots in the same experiment.Overall, the identified protein spots represented 116 dis-tinct gene products. The numbers of differentially abun-dant protein spots between the time points 4 versus20 h, 12 versus 20 h and 20 versus 28 h were 132, 46and 23 respectively. The most remarkable growth phase-dependent changes in protein abundance occurred inthe set of carbohydrate metabolism proteins (COG cat-egory G), consistent with the RNA results. The abun-dance of these proteins was continuously changingduring the fermentation process. For most other pro-teins, the differences were most obvious when compar-ing the mid-exponential and early stationary phases (4 hand 20 h). Altered abundances were detected amongothers in proteins from COG categories F (Nucleotidetransport and metabolism), E (Amino acid transport andmetabolism) and J (Translation, ribosomal structure andbiogenesis).

Fig. 2. Venn diagrams showing the numbers of significantlydifferentially transcribed genes (� 2-fold difference, P � 0.01)during growth in whey. Time points at 12 h, 16 h, 20 h and 31 hwere compared with the time point at 4 h (4 h, mid-exponential;12 h, late exponential; 16 h, stationary transition point; 20 h, earlystationary; and 31 h, late stationary phase).

748 K. Laakso et al.

© 2011 The AuthorsMicrobial Biotechnology © 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Microbial Biotechnology, 4, 746–766

In conclusion, the microarray analyses coupled withproteomics revealed that the transcription of 636 genesand the production of 116 proteins were altered inL. rhamnosus GG during the progression of growth fromthe mid-exponential to the late stationary growth phases.Of these changes, a total of 61 could be confirmed bygene expression analysis at both the transcript andprotein level.

Genes and proteins involved in central metabolicpathways are modulated in a growth-dependent manner

Carbohydrate transport and metabolism. The wheymedium used in this study contains glucose and galactosederived from hydrolysed lactose. In L. rhamnosus GG,there are two different pathways for the catabolismof the galactose moiety of lactose. The Leloir pathway(proteins encoded by the galKETRM operon and pgm)converts galactose to glucose-6-phosphate, and thetagatose-6-phosphate pathway (proteins encoded bythe lacCDBAR operon) metabolizes galactose toglyceraldehyde-3-phosphate and dihydroxyacetone phos-phate (Kankainen et al., 2009). Genes encoding enzymesfrom both the Leloir pathway and the tagatose-6-phosphate pathway showed a significant increase inexpression at the mRNAlevel when the culture shifted fromthe exponential to the stationary phase of growth (Fig. 4A,Table S1). The same expression pattern was also evidentat the protein level, as galactose utilization enzymes weremore abundant in the early stationary growth phase (20 h)than in the exponential growth phase (4 or 12 h) (Tables 1and 2). Furthermore, chemical analyses of the culturesshowed that glucose was no longer detectable (detectionlimit 0.05%) in the early stationary phase of growth whilegalactose was still found in high amounts (1.0%). There-fore, increased expression of genes involved in galactoseutilization at the stationary-phase transition point wasexpected because L. rhamnosus GG first metabolizes thereadily fermented glucose moiety; only after that does itmetabolize the less easily exploited galactose portion ofthe hydrolysed lactose in the whey medium. Applying aproteomic approach, Cohen and colleagues (2006) alsodetected upregulation of the Leloir pathway in L. plantarumWCFS1 during stationary growth in the laboratory MRSmedium, which possibly contained trace amounts of galac-tose. Our transcript results also indicated that genesencoding components of the phosphotransferase system(PTS), currently annotated as galactitol-specific(LGG_00343, LGG_00345–00346), were significantlyupregulated upon entry into the stationary growth phaseand clustered together with lacCDBAR genes (cluster 2 inFig. S1). Increased expression of the IIA component(LGG_00345) of that PTS system was also detected at theprotein level (Tables 1 and 2). These results suggest that

Fig. 3. COGs classification of significantly differentially expressedgenes at 12, 16, 20 and 31 h compared with 4 h. Black barsindicate upregulated genes, while grey bars representdownregulated genes. COG functional categories are as follows:C, Energy production and conversion; D, Cell cycle control, celldivision, chromosome partitioning; E, Amino acid transport andmetabolism; F, Nucleotide transport and metabolism; G,Carbohydrate transport and metabolism; H, Coenzyme transportand metabolism; I, Lipid transport and metabolism; J, Translation,ribosomal structure and biogenesis; K, Transcription; L, Replication,recombination and repair; M, Cell wall/membrane/envelopebiogenesis; N, Cell motility; O, Post-translational modification,protein turnover, chaperones; P, Inorganic ion transport andmetabolism; Q, Secondary metabolites biosynthesis, transport andcatabolism; R, General function prediction only; S, Functionunknown; T, Signal transduction mechanisms; U, Intracellulartrafficking, secretion and vesicular transport; V, Defencemechanisms.

Growth phase-associated changes in L. rhamnosus GG 749

© 2011 The AuthorsMicrobial Biotechnology © 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Microbial Biotechnology, 4, 746–766

Tab

le1.

Pro

tein

sw

ithdi

fferin

gab

unda

nce

betw

een

the

mid

-exp

onen

tial(

4h)

and

early

stat

iona

ry(2

0h)

grow

thph

ases

and

tran

scrip

tion

diffe

renc

esin

the

corr

espo

ndin

gge

nes.

Locu

sID

Gen

ena

me

Fun

ctio

n

Pro

teom

ic(f

old

chan

ge)

Tran

scrip

tom

ic(L

og2R

)20

h/4

hS

pot

No.

20h/

4h

Am

ino

acid

met

abol

ism

LGG

_005

68cy

sKC

yste

ine

synt

hase

11.

90.

518

LGG

_011

58pe

pDD

ipep

tidas

eA

2-2

–0.5

07LG

G_0

1201

sufS

Cys

tein

ede

sulfu

rase

31.

6–0

.281

LGG

_014

78pe

pOE

ndop

eptid

ase

O4

1.6

1.78

8LG

G_0

2639

–A

min

otra

nsfe

rase

51.

73.

741

LGG

_027

08se

lAP

yrid

oxal

phos

phat

e-de

pend

ent

enzy

me

6a2.

53.

039

Car

bohy

drat

em

etab

olis

mG

alac

tose

met

abol

ism

LGG

_003

41la

cCTa

gato

se-6

-pho

spha

teki

nase

73.

47.

214

LGG

_006

53ga

lKG

alac

toki

nase

82.

52.

755

LGG

_006

64la

cCTa

gato

se-6

-pho

spha

teki

nase

97.

96.

897

LGG

_006

65la

cDTa

gato

se1,

6-di

phos

phat

eal

dola

se10

4.4

6.91

0LG

G_0

0666

lacB

Gal

acto

se-6

-pho

spha

teis

omer

ase

subu

nit

B11

,12

23.9

6.92

3LG

G_0

0667

lacA

Gal

acto

se-6

-pho

spha

teis

omer

ase

subu

nitA

13,1

431

.96.

988

LGG

_006

68la

cRLa

ctos

eph

osph

otra

nsfe

rase

syst

emre

pres

sor

1515

7.05

5LG

G_0

1062

galU

UT

P-g

luco

se-1

-pho

spha

teur

idyl

yltr

ansf

eras

e16

,17

2.1

0.87

2LG

G_0

2575

lacD

Taga

tose

1,6-

diph

osph

ate

aldo

lase

184

-0.0

7G

lyco

lysi

s/gl

ucon

eoge

nesi

sLG

G_0

0524

fba

Fru

ctos

e-bi

spho

spha

teal

dola

se19

-1.5

-1.9

7LG

G_0

0921

pgm

Pho

spho

gluc

omut

ase

201.

51.

686

LGG

_009

33ga

pAG

lyce

rald

ehyd

e-3-

phos

phat

ede

hydr

ogen

ase

21–2

52.

90.

644

LGG

_009

33ga

pAG

lyce

rald

ehyd

e-3-

phos

phat

ede

hydr

ogen

ase

26,2

7-1

.80.

644

LGG

_009

34pg

kP

hosp

hogl

ycer

ate

kina

se28

,29,

30a

3.1

0.28

6LG

G_0

0935

tpiA

Trio

seph

osph

ate

isom

eras

e31

2.2

0.14

7LG

G_0

0935

tpiA

Trio

seph

osph

ate

isom

eras

e32

-1.9

0.14

7LG

G_0

0936

eno

Eno

lase

332

0.67

2LG

G_0

1375

pyk

Pyr

uvat

eki

nase

34–3

64.

2-2

.004

Oth

ers

LGG

_000

52ed

a2-

Deh

ydro

-3-d

eoxy

phos

phog

luco

nate

aldo

lase

/4-h

ydro

xy-2

-oxo

glut

arat

eal

dola

se37

-1.7

0.64

LGG

_003

21de

oCD

eoxy

ribos

e-ph

osph

ate

aldo

lase

381.

90.

859

LGG

_003

42sr

lDS

orbi

tol-6

-pho

spha

te2-

dehy

drog

enas

e39

,40

53.6

7.87

7LG

G_0

0757

adhE

Ald

ehyd

e-al

coho

ldeh

ydro

gena

se41

–44

185.

86LG

G_0

1322

pdhC

Pyr

uvat

ede

hydr

ogen

ase

com

plex

E2

com

pone

nt,

dihy

drol

ipoy

llysi

ne-r

esid

ueac

etyl

tran

sfer

ase

45-1

.60.

796

LGG

_013

60fr

uB1-

Pho

spho

fruc

toki

nase

46-3

-1.6

59LG

G_0

1421

pflB

For

mat

eac

etyl

tran

sfer

ase

47–4

94.

74.

288

LGG

_019

14ci

tFC

itrat

ely

ase,

alph

asu

buni

t,C

itrat

eC

oA-t

rans

fera

se50

2.8

-2.6

73LG

G_0

2025

glgD

Glu

cose

-1-p

hosp

hate

aden

ylyl

tran

sfer

ase

regu

lato

rysu

buni

t51

1.9

3.58

8

750 K. Laakso et al.

© 2011 The AuthorsMicrobial Biotechnology © 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Microbial Biotechnology, 4, 746–766

Tab

le1.

cont

.

Locu

sID

Gen

ena

me

Fun

ctio

n

Pro

teom

ic(f

old

chan

ge)

Tran

scrip

tom

ic(L

og2R

)20

h/4

hS

pot

No.

20h/

4h

Lipi

dm

etab

olis

mLG

G_0

0509

dhaK

Dih

ydro

xyac

eton

eki

nase

52,5

33.

41.

109

LGG

_021

13fa

bZ(3

R)-

hydr

oxym

yris

toyl

-[ac

yl-c

arrie

r-pr

otei

n]de

hydr

atas

e54

-5.7

-6.3

21LG

G_0

2115

fabF

3-O

xoac

yl-[

acyl

-car

rier-

prot

ein]

synt

hase

II55

,56

-2.8

-5.9

69LG

G_0

2118

fabK

Eno

yl-(

Acy

l-car

rier-

prot

ein)

redu

ctas

eII

57-1

.9-5

.285

Mem

bran

etr

ansp

ort

LGG

_003

45ga

tAP

TS

syst

em,

gala

ctito

l-spe

cific

IIAco

mpo

nent

585.

57.

069

LGG

_009

51m

alE

AB

Ctr

ansp

orte

r,su

gar

tran

spor

ter

perip

lasm

icco

mpo

nent

59a

1.9

-0.7

4LG

G_0

1940

oppF

AB

Ctr

ansp

orte

r,ol

igop

eptid

etr

ansp

orte

rA

TP

ase

com

pone

nt60

3.2

2.51

LGG

_019

41op

pDA

BC

tran

spor

ter,

olig

opep

tide

tran

spor

ter

AT

Pas

eco

mpo

nent

612.

32.

733

LGG

_028

38m

anA

PT

Ssy

stem

,m

anno

se-s

peci

ficIIA

Bco

mpo

nent

62,6

31.

81.

199

Met

abol

ism

ofco

fact

ors

and

vita

min

sLG

G_0

2764

entB

Isoc

horis

mat

ase

fam

ilypr

otei

n64

4.7

3.53

8LG

G_0

2765

pncB

Nic

otin

ate

phos

phor

ibos

yltr

ansf

eras

e65

4.7

3.70

7LG

G_0

1258

nifS

Am

inot

rans

fera

secl

ass

V66

-3.2

-0.5

46N

ucle

otid

em

etab

olis

mP

urin

em

etab

olis

mLG

G_0

0249

guaB

Inos

ine-

5′-m

onop

hosp

hate

dehy

drog

enas

e67

-2.1

-1.3

49LG

G_0

1803

purD

Pho

spho

ribos

ylam

ine-

glyc

ine

ligas

e30

a1.

7-1

.771

LGG

_018

05pu

rHB

ifunc

tiona

lpur

ine

bios

ynth

esis

prot

ein

purH

68,6

92.

3-1

.698

LGG

_018

07pu

rMP

hosp

horib

osyl

form

ylgl

ycin

amid

ine

cycl

o-lig

ase

70a

2-1

.574

LGG

_018

08pu

rFA

mid

opho

spho

ribos

yltr

ansf

eras

e71

,72

2.8

-1.6

25LG

G_0

1809

purL

Pho

spho

ribos

ylfo

rmyl

glyc

inam

idin

esy

ntha

seII

733.

8-1

.653

LGG

_018

12pu

rCP

hosp

horib

osyl

amin

oim

idaz

ole-

succ

inoc

arbo

xam

ide

synt

hase

742.

5-1

.466

LGG

_018

13pu

rKP

hosp

horib

osyl

amin

oim

idaz

ole

carb

oxyl

ase,

AT

Pas

esu

buni

t6a ,

75,7

62.

8-1

.539

LGG

_019

68gu

aAG

MP

synt

hase

771.

8-1

.916

LGG

_024

66ad

kA

deny

late

kina

se78

-1.7

-1.8

65P

yrim

idin

em

etab

olis

mLG

G_0

1175

upp

Ura

cilp

hosp

horib

osyl

tran

sfer

ase

79-1

.6-1

.009

LGG

_014

56ca

rBC

arba

moy

l-pho

spha

tesy

ntha

se,

larg

esu

buni

t80

,81

-2.3

-3.8

19LG

G_0

1458

pyrC

Dih

ydro

orot

ase

82-1

.9-3

.269

LGG

_014

59py

rBA

spar

tate

carb

amoy

ltran

sfer

ase

83-1

.6-3

.143

LGG

_014

61py

rRP

yrim

idin

eop

eron

regu

lato

rypr

otei

n,bi

func

tiona

lpro

tein

pyrR

84-2

.2-2

.752

LGG

_025

46py

rGC

TP

synt

hase

85-2

-0.9

64O

ther

sLG

G_0

1474

nrdE

Rib

onuc

leos

ide-

diph

osph

ate

redu

ctas

e,al

pha

subu

nit

86,8

7-2

.3-1

.339

Pep

tidog

lyca

nbi

osyn

thes

isLG

G_0

0254

dacA

D-a

lany

l-D-a

lani

neca

rbox

ypep

tidas

e59

a1.

90.

149

LGG

_006

62–

Bet

a-la

ctam

ase

clas

sC

-rel

ated

peni

cilli

n-bi

ndin

gpr

otei

n88

7.3

6.84

4LG

G_0

1282

mur

DU

DP

-N-a

cety

lmur

amoy

lala

nine

-D-g

luta

mat

elig

ase

89-1

.7-0

.545

LGG

_017

68m

urC

UD

P-N

-ace

tylm

uram

ate-

L-al

anin

elig

ase

90-1

.9-1

.239

Growth phase-associated changes in L. rhamnosus GG 751

© 2011 The AuthorsMicrobial Biotechnology © 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Microbial Biotechnology, 4, 746–766

Tab

le1.

cont

.

Locu

sID

Gen

ena

me

Fun

ctio

n

Pro

teom

ic(f

old

chan

ge)

Tran

scrip

tom

ic(L

og2R

)20

h/4

hS

pot

No.

20h/

4h

Pro

tein

synt

hesi

sLG

G_0

1628

rpsB

SS

U/3

0Srib

osom

alpr

otei

nS

2P70

a2

-1.4

08LG

G_0

1628

rpsB

SS

U/3

0Srib

osom

alpr

otei

nS

2P91

,92

-2.1

-1.4

08LG

G_0

1690

rplU

LSU

/50S

ribos

omal

prot

ein

L21P

93-2

.5-0

.991

LGG

_024

93fu

sAP

rote

intr

ansl

atio

nel

onga

tion

fact

orG

(EF

-G)

94-1

.7-1

.288

Str

ess LG

G_0

1367

clpB

AT

P-d

epen

dent

chap

eron

eC

lpB

951.

70.

783

LGG

_016

04dn

aKC

hape

rone

prot

ein

dnaK

96-1

.5-0

.027

LGG

_021

51us

pU

nive

rsal

stre

sspr

otei

n,U

spA

fam

ily97

5.6

-0.3

64LG

G_0

2239

groE

L60

kDa

chap

eron

inG

RO

EL

98,9

92.

30.

058

LGG

_024

99cl

pCA

TP

-dep

ende

ntC

lppr

otea

seA

TP

-bin

ding

subu

nit

100

4.5

0.25

LGG

_028

06ht

rAS

erin

epr

otea

se10

1-1

.6-0

.454

Tran

scrip

tion

LGG

_024

61rp

oAD

NA

-dire

cted

RN

Apo

lym

eras

e,su

buni

tal

pha

102

-1.8

-1.5

36LG

G_0

2498

rpoB

DN

A-d

irect

edR

NA

poly

mer

ase,

beta

chai

n10

3,10

4-2

.5-0

.533

Tran

slat

ion

LGG

_008

48le

uSLe

ucyl

-tR

NA

synt

heta

se10

5-2

.4-0

.541

LGG

_012

61va

lSV

alyl

-tR

NA

synt

heta

se10

6-2

.1-0

.427

LGG

_017

86ar

gSA

rgin

yl-t

RN

Asy

nthe

tase

107

-1.8

-0.8

27LG

G_0

2332

gltX

Glu

tam

yl-t

RN

Asy

nthe

tase

108

-1.6

-1.5

53LG

G_0

2584

met

GM

ethi

onyl

-tR

NA

synt

heta

se/p

rote

inse

cret

ion

chap

eron

in,

Csa

A10

9,11

0-2

-0.5

86M

isce

llane

ous

LGG

_002

26–

Pyr

idox

ine

5′-p

hosp

hate

oxid

ase

Vre

late

dfa

vin-

nucl

eotid

e-bi

ndin

gpr

otei

n11

1-3

.4-3

.048

LGG

_004

91–

NA

DH

pero

xida

se11

24.

1-0

.197

LGG

_006

15ye

aEA

ldo/

keto

redu

ctas

e11

32.

10.

153

LGG

_006

34–

Dyp

-typ

epe

roxi

dase

fam

ilypr

otei

n11

42.

52.

248

LGG

_006

63–

Con

serv

edpr

otei

n11

54.

96.

808

LGG

_007

40gp

hH

ydro

lase

,ha

loac

idde

halo

gena

se-li

kefa

mily

116

-1.8

-0.4

95LG

G_0

0744

dkgA

Ald

o/ke

tore

duct

ase

117

1.6

0.71

LGG

_008

99se

cAP

rote

intr

ansl

ocas

esu

buni

tse

cA11

8-2

.1-0

.076

LGG

_012

65m

reB

Rod

shap

e-de

term

inin

gpr

otei

nM

reB

119

1.6

-0.3

31LG

G_0

1311

ykqC

Met

allo

-bet

a-la

ctam

ase

supe

rfam

ilypr

otei

n12

0-2

0.49

3LG

G_0

1327

typA

GT

P-b

indi

ngpr

otei

nTy

pA12

1-1

.8-2

.590

LGG

_013

95–

Con

serv

edpr

otei

n12

23.

43.

504

LGG

_014

33–

Nitr

ored

ucta

se12

32.

31.

344

LGG

_014

65fh

sF

orm

ate-

tetr

ahyd

rofo

late

ligas

e12

41.

9-1

.209

LGG

_014

68–

Nitr

ored

ucta

sefa

mily

prot

ein

125,

126

-3.6

-3.8

52LG

G_0

1821

ptsH

Pho

spho

carr

ier

prot

ein

HP

r12

72.

20.

884

LGG

_018

37m

vaS

Hyd

roxy

met

hylg

luta

ryl-C

oAsy

ntha

se12

8-1

.50.

071

LGG

_020

50gl

fU

DP

-gal

acto

pyra

nose

mut

ase

129

-1.7

-1.7

11LG

G_0

2098

–C

onse

rved

prot

ein

130

2.4

-0.5

26LG

G_0

2415

cueO

Mul

ticop

per

oxid

ase

131

3.9

1.73

LGG

_026

30yg

hZA

ldo/

keto

redu

ctas

e(o

xido

redu

ctas

e)13

22.

61.

3

a.Tw

opr

otei

nsw

ere

iden

tified

from

thes

esp

ots.

Itre

mai

nsun

clea

rw

hich

ofth

ese

prot

eins

was

mor

eab

unda

nt.

752 K. Laakso et al.

© 2011 The AuthorsMicrobial Biotechnology © 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Microbial Biotechnology, 4, 746–766

this particular PTS is involved in the transportation ofgalactose, and revision of its present annotation could beconsidered. It is noteworthy that current annotations ofPTS transporters do not necessarily indicate their true rolein the metabolism of particular sugars because the speci-

ficities of many lactobacilli PTS transporters are incorrectlyannotated (Francl et al., 2010), and current computationalmethods are unreliable for the prediction of substratespecificity. Therefore, further experiments are needed tocharacterize the PTS transporter specificities of L. rham-

Table 2. Proteins with differing abundance between the late exponential (12 h) and early stationary (20 h) growth phases and transcriptiondifferences in the corresponding genes.

Locus IDGenename Function

Proteomic(fold change) Transcriptomic

(Log2R)20 h/12 hSpot No. 20 h/12 h

Carbohydrate metabolismGalactose metabolism

LGG_00653 galK Galactokinase 133,134 4 4.176LGG_00654 galE UDP-glucose 4-epimerase 135 3.1 3.817LGG_00655 galT Galactose-1-phosphate uridylyltransferase 136 3.7 3.831LGG_00664 lacC Tagatose-6-phosphate kinase 137 5.3 6.928LGG_00665 lacD Tagatose 1,6-diphosphate aldolase 138 2.3 6.461LGG_00666 lacB Galactose-6-phosphate isomerase subunit B 139,140 30.1 6.675LGG_00667 lacA Galactose-6-phosphate isomerase subunit A 141 26.8 6.561LGG_00668 lacR Lactose phosphotransferase system repressor 142 15.9 6.699LGG_02575 lacD Tagatose 1,6-diphosphate aldolase 143,144 2.1 -1.145

Glycolysis/gluconeogenesisLGG_00921 pgm Phosphoglucomutase 145 1.9 1.331LGG_00933 gapA Glyceraldehyde-3-phosphate dehydrogenase 146,147 1.6 0.239LGG_00933 gapA Glyceraldehyde-3-phosphate dehydrogenase 148 -2.9 0.239LGG_01375 pyk Pyruvate kinase 149,150 2 -2.596LGG_01375 pyk Pyruvate kinase 151 -1.8 -2.596

OthersLGG_00342 srlD Sorbitol-6-phosphate 2-dehydrogenase 152,153 26.1 6.56LGG_00373 rbsK Ribokinase 154 2.1 4.074LGG_00757 adhE Aldehyde-alcohol dehydrogenase 155–157 5.6 2.279LGG_01421 pflB Formate acetyltransferase 158 2.2 3.134LGG_02523 ldh L-lactate dehydrogenase 159a -1.6 -3.868

Membrane transportLGG_00345 gatA PTS system, galactitol-specific IIA component 160 13 6.286

Metabolism of cofactors and vitaminsLGG_02764 entB Isochorismatase family protein 161 2.6 1.32LGG_02765 pncB Nicotinate phosphoribosyltransferase 162,163 3.3 1.623

Nucleotide metabolismPurine metabolism

LGG_00249 guaB Inosine-5′-monophosphate dehydrogenase 164 -1.6 0.419LGG_01808 purF Amidophosphoribosyltransferase 165 1.7 -2.035

Peptidoglycan biosynthesisLGG_00662 – Beta-lactamase class C-related penicillin-binding

protein166 5.2 6.994

StressLGG_01823 clpE ATP-dependent clp protease ATP-binding subunit 167 1.8 0.722LGG_02151 usp Universal stress protein, UspA family 168 2.3 -2.413LGG_02499 clpC ATP-dependent Clp protease ATP-binding subunit 169 1.9 0.399

TranslationLGG_00767 alaS Alanyl-tRNA synthetase 170 -1.5 -0.758LGG_01019 gatA Aspartyl/glutamyl-tRNA(Asn/Gln)

amidotransferase subunit A171 -1.7 -0.603

MiscellaneousLGG_00226 – Pyridoxine 5′-phosphate oxidase V-related

favin-nucleotide-binding protein172 -2.1 -1.605

LGG_00491 – NADH peroxidase 173 1.9 -2.539LGG_00634 – Dyp-type peroxidase family protein 174 1.5 1.042LGG_00763 ytqI Phosphoesterase, DHH family protein 159a -1.6 -0.088LGG_01395 – Conserved protein 175 2.4 3.379LGG_01465 fhs Formate-tetrahydrofolate ligase 176 -1.5 -1.411LGG_01468 – Nitroreductase family protein 177 -1.6 -1.985LGG_02630 yghZ Aldo/keto reductase (oxidoreductase) 178 1.6 -0.024

a. Two proteins were identified from this spot. It remains unclear which of these proteins was more abundant.

Growth phase-associated changes in L. rhamnosus GG 753

© 2011 The AuthorsMicrobial Biotechnology © 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Microbial Biotechnology, 4, 746–766

nosus GG; for example, this could be done by studyingtranscript expression profiles in response to differentcarbohydrates.

The abundance of several glycolytic enzymes and thetranscription of the corresponding genes changed in agrowth phase-dependent fashion. Genes coding for6-phosphofructokinase (pfkA), fructose-bisphosphatealdolase (fba), phosphoglyceromutase (gpmA) and pyru-vate kinase (pyk) were significantly downregulated duringthe stationary-phase transition (Fig. 4A, Table S1). At theprotein level, Fba was more abundant in the mid-exponential growth phase than in the early stationaryphase, consistent with the mRNA results (Table 1). Noclear trend in the abundance changes of glycolytic proteinswas seen during growth, however. Many glycolytic proteins(GapA, Pgk, TpiA, Eno, Pyk) were detected in severalparallel protein spots (two to seven spots) on 2-D gels, andin some of these spot sets (GapA, TpiA, Pyk), both adecrease and an increase in abundance were detected(Tables 1 and 2). For instance, five glyceraldehyde-3-phosphate dehydrogenase (GapA) isoforms were more

abundant during the early stationary growth phase (com-pared with 4 h) while two other isoforms were more abun-dant during the mid-exponential growth phase (comparedwith 20 h). Post-translational protein modifications such asphosphorylation (Cozzone, 1998) and methionine formy-lation (Bandow et al., 2003) may alter the migration ofproteins in 2-D gels, which could also explain the appear-ance of parallel spots in this case. In Lactococcus lactis,glycolysis has been shown to be regulated at the transcrip-tional level only to a lesser degree, and post-translationalmodifications might be more important for regulation (dis-cussed in Soufi et al., 2008). In Lactococcus, Escherichiacoli and Bacillus, several glycolytic proteins have beenshown to be phosphorylated (Eymann et al., 2007; Maceket al., 2007; 2008; Soufi et al., 2008), and characterizationof the phosphoproteome suggests conservation of thistype of modification of glycolytic enzymes in L. rhamnosusGG (J. Koponen, K. Laakso, K. Koskenniemi, M. Kan-kainen, K. Savijoki, T.A. Nyman, S. Tynkkynen, N. Kalkki-nen and P. Varmanen, in preparation). Transcripts of genescoding for GapA, Pgk, TpiA and Eno were only slightly

Table 3. Proteins with differing abundance between the early stationary (20 h) and late stationary (28 h) growth phases and transcriptiondifferences in the corresponding genes.

Locus IDGenename Function

Proteomic(fold change) Transcriptomic

(Log2R)31 h/20 hSpot No. 28 h/20 h

Amino acid metabolismLGG_02639 – Aminotransferase 179 2.1 0.824

Carbohydrate metabolismGalactose metabolism

LGG_00341 lacC Tagatose-6-phosphate kinase 180 2.1 -1.548LGG_00664 lacC Tagatose-6-phosphate kinase 181 2.3 -2.407LGG_00665 lacD Tagatose 1,6-diphosphate aldolase 182 2 -2.313LGG_00666 lacB Galactose-6-phosphate isomerase subunit B 183,184 2.6 -2.027LGG_00667 lacA Galactose-6-phosphate isomerase subunit A 185,186 2 -2.163LGG_00668 lacR Lactose phosphotransferase system repressor 187 2.3 -2.18

Glycolysis/gluconeogenesisLGG_00921 pgm Phosphoglucomutase 188 1.6 1.085LGG_01375 pyk Pyruvate kinase 189 -1.5 -0.329

OthersLGG_00342 srlD Sorbitol-6-phosphate 2-dehydrogenase 190,191 2.4 -2.053LGG_00757 adhE Aldehyde-alcohol dehydrogenase 192,193 2.5 -0.466LGG_01421 pflB Formate acetyltransferase 194 1.5 0.176LGG_01876 pck Phosphoenolpyruvate carboxykinase (ATP) 195 1.8 0.693LGG_02693 xylB Alcohol dehydrogenase 196a 1.5 1.601

Membrane transportLGG_02421 mtsB ABC transporter, ATPase component 197a -1.5 -0.438

Peptidoglycan biosynthesisLGG_00662 – Beta-lactamase class C-related penicillin-binding protein 198 2.6 -2.265

Nucleotide metabolismPyrimidine metabolism

LGG_01625 pyrH Uridylate kinase 197a -1.5 0.296Miscellaneous

LGG_01016 ligA NAD-dependent DNA ligase 199 -1.6 -0.432LGG_01061 mvaK Phosphomevalonate kinase 200 1.6 -0.477LGG_01395 – Conserved protein 201 2.5 0.926LGG_02124 yqhD Iron-containing alcohol dehydrogenase 196a 1.5 0.64

a. Two proteins were identified from these spots. It remains unclear which of these proteins was more abundant.

754 K. Laakso et al.

© 2011 The AuthorsMicrobial Biotechnology © 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Microbial Biotechnology, 4, 746–766

upregulated in the early stationary growth phase comparedwith the exponential growth phase in our study, suggestingthat they are constitutively expressed (Fig. 4A, Tables 1and 2).

Dihydroxyacetone kinase (DhaK), which catalyses thephosphorylation of dihydroxyacetone, was more abun-dant in early stationary-phase cells than in exponential-phase cells at both the transcript and protein levels(Fig. 4A, Table 1). Dihydroxyacetone phosphate ischannelled to glycolysis via the activity of triosephos-phate isomerase (TpiA), which catalyses the formationof glyceraldehyde-3-phosphate from dihydroxyacetonephosphate. Deoxyribose-phosphate aldolase (DeoC,LGG_00321), which converts deoxyribose-5-phosphateto glyceraldehyde-3-phosphate (a glycolytic intermedi-ate), and its concomitant gene were upregulated uponentry into the stationary phase of growth (Fig. 4A,Table 1). One of the most highly differentially expressed

proteins and genes was sorbitol-6-phosphate2-dehydrogenase (SrlD, LGG_00342), which channelssorbitol-6-phosphate to glycolysis via the conversion ofsorbitol-6-phosphate to fructose-6-phosphate (Fig. 4A,Tables 1–3). The expression of SrlD increased over time,especially when the culture shifted from the exponentialgrowth to the stationary phase. Transcription of the genesencoding the glucitol/sorbitol-PTS (LGG_02717–02719)was induced when the culture shifted from exponentialgrowth to stationary growth (Table S1). These resultssuggest that L. rhamnosus GG starts to use alternativeenergy sources other than the previously mentionedgalactose, namely sorbitol, dihydroxyacetone anddeoxyribose-5-phosphate, at the beginning of the station-ary phase after the exhaustion of glucose. In Lactobacilluscasei, the sorbitol operon is repressed by glucose (Alcán-tara et al., 2008), and this seems to also apply forL. rhamnosus GG because the upregulation of SrlD

Fig. 4. Gene expression patterns of selectedgenes and operons during growth in wheycoding for (A) carbohydrate and pyruvatemetabolic proteins, (B) proteinases,peptidases and amino acid transporters and(C) probiotic-associated factors. Geneexpression changes (averages of threebiological replicates) between paired timepoints are represented colorimetrically, withdark red indicating an expression ratio of 5.0and dark blue indicating an expression ratioof -5.0 on a log2 scale. The first four lanesshow results of comparison of time points 12,16, 20 and 31 h to the time point 4 hcorresponding to the mid-exponential phase.Statistically significant changes(P-value � 0.01) are marked with asterisks.

Growth phase-associated changes in L. rhamnosus GG 755

© 2011 The AuthorsMicrobial Biotechnology © 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Microbial Biotechnology, 4, 746–766

occurs when glucose has been used up (according tochemical analyses, detection limit 0.05%).

In summary, during the transition to the stationarygrowth phase, the utilization of glucose as the maincarbon and energy source was replaced in L. rhamnosusGG by alternative pathways for carbohydrate metabolism,such as galactose utilization, using the tagatose-6-phosphate and the Leloir pathways.

Energy production and conversion (pyruvate metabo-lism). Under normal glucose fermentation conditions,lactic acid bacteria reduce pyruvate to lactate by aNAD+-dependent lactate dehydrogenase. In addition,lactic acid bacteria may have alternative means of uti-lizing pyruvate, including the pyruvate-formate lyasesystem, the pyruvate dehydrogenase pathway, thepyruvate oxidase pathway and the diacetyl/acetoinpathway (Axelsson, 2004). Genes encoding pyruvate-formate lyase (LGG_01421–01422), aldehyde-alcoholdehydrogenase (LGG_00757), pyruvate dehydrogenase(LGG_01320–01323) and enzymes of the pyruvateoxidase pathway (LGG_01836, pyruvate oxidase andLGG_02183, acetate kinase) were induced at the

stationary-phase transition point (Fig. 4A, Table S1). ThePflB subunit of pyruvate-formate lyase and aldehyde-alcohol dehydrogenase (AdhE) were also more abun-dant in stationary-phase cells than in exponential-phasecells at the protein level, supporting the transcript results(Table 1–3). However, proteomic analyses indicated thatpyruvate dehydrogenase (PdhC) was more abundant inthe exponential growth phase than in the stationarygrowth phase, in contrast with the transcript results(Table 1). A gene encoding lactate dehydrogenase(LGG_00606) clustered together with the pyruvate-formate lyase and acetate kinase genes, and its expres-sion was significantly increased over time (cluster 4 inFig. S1). Expression of another gene coding for a lactatedehydrogenase (LGG_02523) was, however, decreasedupon entry into the stationary growth phase, and thesame tendency was seen at the protein level (12 hversus 20 h) (Table 2). Thus, our results suggest thatL. rhamnosus GG changes from homolactic fermentationto a mixed acid fermentation when the culture reachesthe stationary phase, and the end-products formed, inaddition to lactate, are formate, acetate and ethanol.This is probably due to the shift detected from glucose

Fig. 4. cont.

756 K. Laakso et al.

© 2011 The AuthorsMicrobial Biotechnology © 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Microbial Biotechnology, 4, 746–766

to galactose utilization in L. rhamnosus GG at thestationary-phase transition point. The same trend hasbeen observed in Lactococcus strains, which shift fromhomolactic to mixed acid fermentation when glucose islimited or during galactose fermentation (reviewed byNeves et al., 2005). In L. plantarum cells, more lactatedehydrogenase was produced in the exponential than inthe stationary growth phase (Cohen et al., 2006), sup-porting our conclusions.

Nucleotide transport and metabolism. The abundance ofpyrimidine biosynthetic gene cluster (pyr) transcripts wasreduced upon entry into the stationary phase of growth,and the downregulation of the pyrimidine biosynthesispathway was also detected at the protein level (Table 1and Table S1). In addition, the expression of several othergenes and gene products involved in nucleotide metabo-lism were decreased over time in whey. This is probablydue to protracted growth and, thus, reduced requirementsfor pyrimidine nucleotides. However, in the purine biosyn-

thesis pathway, no clear downregulation was observed. Atthe mRNA level, the purine biosynthesis operon (purBCD-FHKLM and guaAB) was expressed at lower levels in thestationary growth phase than in the exponential growthphase, while the PurCDFHKLM and GuaA proteins weremore abundant at 20 h (early stationary) than at 4 h (mid-exponential) (Table 1 and Table S1). The whey mediumused here does not supply L. rhamnosus GG cells withpurines. It has previously been shown that in L. rhamno-sus GG cells grown in whey medium, during the stationaryphase, the abundance of Pur proteins is elevated com-pared with cells grown in medium rich in purines (Kosk-enniemi et al., 2009). Therefore, our results suggest thatunder the growth conditions described here, purine bio-synthesis in L. rhamnosus GG cells remains active duringthe switch from the exponential to the stationary phase.Furthermore, the identified growth phase-associatedchanges in Pur protein abundance were not associatedwith changes in the corresponding transcript levels, sug-gesting either the involvement of post-transcriptional

Fig. 4. cont.

Growth phase-associated changes in L. rhamnosus GG 757

© 2011 The AuthorsMicrobial Biotechnology © 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Microbial Biotechnology, 4, 746–766

regulation mechanisms or that these proteins have a longhalf-life.

Lipid transport and metabolism. Fatty acids are constitu-ents of phospholipids and glycolipids found in the cyto-plasmic membrane of bacteria. Our results revealed thatfatty acid biosynthetic genes (LGG_02110–02122) werestrongly repressed at the stationary-phase transition point(Table S1). Similarly, the abundance of the proteins,FabZ, FabF and FabK, which are involved in the biosyn-thesis of long-chain saturated fatty acids, were decreasedover time (Table 1). Furthermore, the level of transcriptsencoding cyclopropane-fatty-acyl-phospholipid synthase(cfa, LGG_02109), which participates in the cyclopropa-nation of fatty acids, was reduced upon entry into thestationary phase of growth (Table S1). Fatty acids areenergetically the most expensive membrane lipid compo-nents to produce, and their production is tightly regulatedto match the growth rate of bacterial cells (Zhang andRock, 2009). Therefore, our observations probably resultfrom the deceleration of the growth rate in the stationaryphase. Applying a proteome-level approach, Cohen andcolleagues (2006) observed similar downregulation ofseveral enzymes involved in lipid metabolism in L. plan-tarum during the shift from the exponential to the station-ary phase of growth.

Proteolytic systems, amino acid transportand metabolism

Lactobacillus rhamnosus GG has a limited capacity tosynthesize amino acids; therefore, it requires exogenousamino acids and peptides for growth (Kankainen et al.,2009). In this study, several components of the proteolyticsystem were identified as exhibiting growth phase-dependent induction at both the transcript and the proteinlevel, including proteinases, which degrade proteins intooligopeptides, peptide transporter systems involved in oli-gopeptide uptake, and peptidases, which degrade oli-gopeptides into shorter peptides and amino acids(reviewed by Savijoki et al., 2006). The whey mediumused in this study contains free amino acids and oligopep-tides derived from hydrolysed casein and whey proteinand probably also traces of non-hydrolysed proteins.During growth in whey medium, L. rhamnosus GG firstconsumes the accessible free amino acids and, after that,oligopeptides, which require only minimal processing toliberate the free amino acids. Expression of genes encod-ing cell-envelope proteinase PrtP, maturation protein PrtMand proteinase PrtR increased during growth of L. rham-nosus GG in whey medium until the early stationarygrowth phase, suggesting that the non-hydrolysed or par-tially hydrolysed proteins in whey medium may be utilizedafter the exhaustion of other amino acid sources (Fig. 4B,

Table S1). In protein gels, no differentially produced pro-teinases were detected. However, it is well known thathigh-molecular-weight proteins, hydrophobic proteins andproteins with extreme pI values are under-represented on2-D gels possibly explaining the lack of identification ofPrtP (206.6 kDa), PrtR (155.8 kDa) and PrtM (pI value10.3). It has previously been shown that whey-grownL. rhamnosus GG is unable to hydrolyse intact casein(Kankainen et al., 2009), and the role of the proteinases inthe utilization of intact or partially hydrolysed whey pro-teins remains to be elucidated. Consistent with our find-ings, in L. acidophilus cultivated in milk, the expression ofgenes encoding PrtP and PrtM was increased over time(Azcarate-Peril et al., 2009).

The genome of L. rhamnosus GG holds three completeoligopeptide ABC transporter operons (LGG_01652–01656, LGG_01940–01945 excluding LGG_01942, andLGG_02062–02066) (Kankainen et al., 2009). Each con-sists of five genes encoding oligopeptide-binding protein(OppA), two integral membrane proteins (OppB andOppC) and two ATP-binding proteins (OppD and OppF).Of these operons, LGG_01940–01945 was found to bedifferentially expressed at both the mRNA and proteinlevels. Expression of the whole operon at the mRNA levelwas increased until the early stationary growth phase(Fig. 4B, Table S1), and the ATPase components, OppDand OppF, were more abundant in the early stationarygrowth phase than in the mid-exponential growth phaseat the protein level, supporting the transcript results(Table 1). LGG_02062–02066 transcripts were moreabundant in the mid-exponential growth phase (4 h) thanat the 12, 16 and 31 h time points but less abundant at 4 hthan in the early stationary growth phase (20 h) (Fig. 4B,Table S1). Transcription of these genes was thereforetransient, with a short peak in the early stationary growthphase. Differential production of the corresponding pro-teins was not observed possibly because the foldchanges were below threshold value of 1.5. In addition tothe three oligopeptide-binding proteins (OppA) encodedwithin the complete Opp operons, four additional oppAgenes are present in the L. rhamnosus GG genome (Kan-kainen et al., 2009). Of these, three oppA genes weresignificantly differentially expressed in a growth phase-dependent manner (Fig. 4B, Table S1). Based on theseobservations, it could be speculated that different oli-gopeptide transport systems may have different specifici-ties, and while one Opp transporter system is active in theexponential growth phase, another might be active in thestationary phase of growth. Similar results were obtainedin a microarray study of L. acidophilus grown in milk,which showed that different Opp genes were expressed atdifferent growth stages (Azcarate-Peril et al., 2009).

At the mRNA level, the genes encoding tripeptidasePepT (LGG_00358), proline-specific aminopeptidase

758 K. Laakso et al.

© 2011 The AuthorsMicrobial Biotechnology © 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Microbial Biotechnology, 4, 746–766

PepX (LGG_01695), dipeptidase PepD (LGG_01972),aminopeptidases PepC2 (LGG_02346) and AmpS(LGG_01050) were significantly differentially expressed atdifferent growth phases (Fig. 4B, Table S1). The genecoding for PepT clustered together with the OppA-encoding gene (LGG_00359) and was maximallyexpressed in L. rhamnosus GG at the early stationaryphase of the growth curve (cluster 10 in Fig. S1). Thetranscript encoding PepX clustered together with theORFs LGG_02062–02066 (the Opp transporter system),and its expression was transiently elevated at 20 h (cluster6 in Fig. S1). The expression levels of pepD, pepC2 andampS, as well as genes involved in polyamine transport(potABCD, LGG_00969–00972), were decayed duringgrowth (Fig. 4B, Table S1). Proteomic results indicatedthat endopeptidase PepO (LGG_01478) was producedat higher levels in the early stationary cultures than inthe mid-exponential cultures while dipeptidase PepD(LGG_01158) was produced more in mid-exponential cellscompared with early stationary cells (Table 1). The samepattern was seen at the mRNA level, although it was notstatistically significant (Fig. 4B). These findings suggestthat different peptidases are utilized specifically in differentgrowth phases, probably depending on their specificitiesfor the particular peptides available at each stage.

General stress responses

Nutrient limitations activate a variety of bacterial stressresponses (reviewed by De Angelis and Gobbetti, 2004).Changes in environmental conditions also rapidly triggerstress-related proteolytic systems to cope with the accu-mulation of irreparably damaged proteins (Savijoki et al.,2006). Not surprisingly, the abundance of several stressproteins and the transcription of genes encoding stressresponsive proteins was elevated when the cells reachedthe stationary growth phase. At the mRNA level, tran-scripts encoding Clp protease subunits, ClpL and ClpE,endopeptidase subunits, HslV (ClpQ) and HslU (ClpY),and heat shock proteins, Hsp1 and Hsp3, were elevated inthe stationary phase of growth. However, genes encodinguniversal stress protein (usp) and 33 kDa chaperonin(hslO) showed their highest expression in the exponentialgrowth phase (Table S1). At the protein level, production ofthe universal stress protein, the chaperone protein GroELand the stress-inducible Clp protease complex subunitsClpC and ClpB was increased over time while the chap-erone protein DnaK and the stress-related serine proteaseHtrA were more abundant in the mid-exponential growthphase (4 h) than in the early stationary growth phase(20 h) (Table 1). The corresponding genes, groEL, clpC,clpB, dnaK and htrA, were not significantly differentiallyexpressed. As chaperone proteins are required for denovo protein folding, the expression of chaperones and

their corresponding genes in the exponential growthphase, when essential nutrients are still sufficiently avail-able, could be a response to the high protein synthesisrate. Because growth phase-associated changes in stressproteins were detectable at either the mRNA or proteinlevel, but not at both levels, it is tempting to speculate thatpost-transcriptional regulation mechanisms are involved inthe adaptation of L. rhamnosus GG to the stationarygrowth phase conditions. Recently, it has been demon-strated that most of the stress-related genes showedincreasing levels of expression in the stationary growthphase in L. acidophilus (Azcarate-Peril et al., 2009). Fur-thermore, in a proteomic study of L. plantarum WCFS1, ithas been observed that some stress proteins were highlyabundant in the late exponential and early stationarygrowth phases (ClpL and Hsp3, among others) while somewere highly abundant during the exponential growth phase(ClpP, among others) (Cohen et al., 2006). It should benoted that the earlier studies describe bacterial growthphase responses in flask culture conditions where gradualacidification of the cultures causes additional stress.These growth conditions were different from those of thepresent study where pH-controlled conditions were used.

Expression of genes mediating potential probiotic-associated traits changes over time in whey

Lactobacillus rhamnosus GG colonizes the human intes-tine and adheres efficiently to mucus and epithelial cells(Alander et al., 1997; 1999; Laparra and Sanz, 2009). Cellsurface factors, such as cell surface proteins, lipoteichoicacids (LTAs) and exopolysaccharides (EPSs), mediateadherence, biofilm formation and colonization by lactoba-cilli (reviewed by Lebeer et al., 2008). Recently, MabA(modulator of adhesion and biofilm) encoded byLGG_01865 was shown to contribute to the adhesion ofL. rhamnosus GG to the host cells (Vélez et al., 2010).Cell surface-exposed pilus, encoded by a gene clusterlocated in a genomic island unique to L. rhamnosusGG, was suggested to play a key role in the adhe-sion of L. rhamnosus GG to human intestinal mucus(Kankainen et al., 2009). Our transcriptomic resultsshowed that strain-specific pilus-encoding genes(spaCBA, LGG_00442–00444) were active in the expo-nential phase of growth, and their expression was clearlyreduced when the culture transitioned to the stationarygrowth phase (Fig. 4C, Table S1). In contrast, growthphase dependence in the expression pattern of mabAwas not observed. However, several other genes encod-ing proteins predicted to contain an adhesion domain,such as LGG_02337 (inlJ) and LGG_02274 (prtP), wereupregulated in L. rhamnosus GG cultures in the stationarygrowth phase compared with the exponential growthphase (Fig. 4C, Table S1). No changes in the correspond-

Growth phase-associated changes in L. rhamnosus GG 759

© 2011 The AuthorsMicrobial Biotechnology © 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Microbial Biotechnology, 4, 746–766

ing proteins were observed, which may be due to thehydrophobic nature or the low abundance of these pro-teins. Furthermore, pilus proteins may occur at het-eropolymeric high-molecular-weight complexes (Mandliket al., 2008) that are not detectable using 2-D DIGE. InL. lactis, cell wall-anchored PrtP has been shown to havepositive effect on cell adhesion to solid surfaces (Habi-mana et al., 2007); however, its effect on adhesion tohuman cells or mucus remains to be elucidated. Pili ofGram-positive pathogens have been extensively investi-gated and have been shown to promote bacterial adhe-sion to host cells, play a role in biofilm formation andevoke host immune responses (reviewed by Mandliket al., 2008). In L. rhamnosus GG, it has also been dem-onstrated that pili play an important role in adhesion toCaco-2 epithelial cells and in biofilm formation (Lebeeret al., 2011). In a virulent Streptococcus pneumoniaestrain, pilus genes were found to be expressed in a growthphase-dependent manner, and the regulation of pilusgenes was shown to be mediated by a two-componentsystem (Song et al., 2009). In our study, the expressionpattern of genes encoding a two-component system(LGG_02387–02388, cluster 14 in Fig. S1) was similar tothat of the pilus-encoding genes (cluster 13 in Fig. S1,Table S1), but further studies are needed to establishwhether this particular two-component system plays arole in the regulation of the pilus locus genes in L. rham-nosus GG or whether it is regulating one of the variousother functions altered during exponential growth.

The expression of genes involved in D-alanylation ofLTAs (dlt operon, LGG_00777–00780, clusters 13 and 16in Fig. S1) and most of the genes encoding enzymesrequired for EPS biosynthesis (LGG_02044–02047 andLGG_02049–02054, clusters 14 and 16 in Fig. S1) wasreduced over time in whey (Fig. 4C, Table S1). Consistentwith our findings, previous studies have demonstratedthat EPS production is growth rate-dependent (Welmanet al., 2003; 2006). In L. rhamnosus GG, EPS plays anegative role in adhesion and biofilm formation, possiblyby shielding adhesion factors such as pili (Lebeer et al.,2009). However, EPS appears to play an important rolein the protection against host defence mechanismsinvolving the innate immune system (Lebeer et al., 2010).Because the balance between adhesion properties andresistance against the host immune system seems to beof the utmost importance for the adaptation of L. rhamno-sus GG in the intestine, expression of pili and EPS bio-synthetic genes are likely to be regulated accordingly. Theproducts of the dlt operon are involved in the D-alanylationof secondary cell wall polymer LTA. D-alanyl ester substi-tutions for LTA add positive charges to otherwise nega-tively charged LTAs and, in this way, directly affect cellsurface charges (Neuhaus and Baddiley, 2003). Lebeerand colleagues (2007) have shown that D-alanyl-

substituted LTAs serve a modulatory function in the adhe-sion of L. rhamnosus GG to Caco-2 cells and in biofilmformation and activate immune responses in the host.Claes and colleagues (2010) recently showed that dele-tion of dltD resulted in a significant increase in the probi-otic efficacy of L. rhamnosus GG in a mouse model ofcolitis. Therefore, it is suggested that D-alanylation ofLTAs may play a role in the survival and persistence aswell as in the probiotic effects of L. rhamnosus GG in thehost, and our results indicate that the expression of thisprobiotic-associated factor is growth phase-dependent.

In addition to adhesion factors such as pili and InlJ,contact between L. rhamnosus GG and host cells canalso be mediated by secreted, small soluble proteins. Forinstance, secreted, low-molecular-weight proteins fromL. rhamnosus GG have been shown to promote in vitrointestinal epithelial cell homeostasis through certain sig-nalling pathways (Yan and Polk, 2002; Tao et al., 2006;Yan et al., 2007; Seth et al., 2008). Expression of a genecoding for the NLP/P60 family secreted protein, p40(LGG_00031), was reduced during growth (Fig. 4C,Table S1). In contrast, the genes LGG_00504 andLGG_00661, coding for secreted proteins of low molecu-lar mass (< 10 kDa), were strongly induced in stationary-phase cells (Fig. 4C, Table S1). Interestingly, a genelocated next to one of these small-sized soluble proteins(LGG_00503) and annotated as a myosin-cross-reactiveantigen orthologue was also highly expressed in the sta-tionary growth phase compared with the exponentialgrowth phase. None of these responses was detected atthe protein level, probably because cell surface-exposedand secreted proteins are usually hydrophobic and areprecipitated within the first dimension once they havemigrated to their isoelectric point.

Concluding remarks

Consumption of probiotic bacteria harvested at differentphases of growth has been shown to cause profoundlydifferent mucosal responses in human (van Baarlenet al., 2009). Accordingly, we assume that the expres-sion status of the genes during the industrial cultivationof a probiotic reflects its activity in the GIT. We investi-gated how gene expression, at both the transcript andprotein levels, in the probiotic bacterium, L. rhamnosusGG, changed over time during growth in industrial-typewhey medium under controlled bioreactor conditions. Inparticular, we aimed to determine whether any traits thatare associated with the probiotic activity of L. rhamnosusGG were affected during the fermentation of whey.Expression of 636 genes, many implicated in theprobiotic-linked actions of L. rhamnosus GG, and 116proteins were modulated in a growth-dependent mannerand half of the changes in protein abundance were

760 K. Laakso et al.

© 2011 The AuthorsMicrobial Biotechnology © 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Microbial Biotechnology, 4, 746–766

associated with changes in transcript levels. Most of theobserved growth phase-dependent changes at both themRNA and protein level appeared during the shift fromthe exponential growth phase to the stationary growthphase. In particular, genes and proteins involved in thegalactose utilization pathways were among the highestinduced upon entry into stationary phase. Many tran-scripts and proteins involved in central metabolic path-ways such as carbohydrate, nucleotide, lipid andpyruvate metabolism were differentially expressedduring growth in whey (Fig. 5). Differential expression ofdistinct carbohydrate utilization and energy productionpathways and several distinct PTS transportersobserved in response to changes in nutritional environ-ment (e.g. exhaustion of the preferred carbon source inthe medium) indicate that L. rhamnosus GG has a flex-ible and adaptable metabolism. This adaptability mightprovide a competitive advantage for L. rhamnosus GG inthe host’s GIT. Furthermore, expression of many genesencoding adhesion factors and secreted proteins withlow molecular masses were increased over time duringthe growth of L. rhamnosus GG in whey medium(Fig. 5), with the highest expression in the stationaryphase. The expression patterns of genes encodingmucus-binding pili and secreted protein p40 differedfrom this general pattern (Fig. 5), having the highestexpression values in the exponential phase. Like nutri-tional versatility, these adhesion and probiotic-

associated factors provide a competitive advantage forL. rhamnosus GG by promoting its survival and persis-tence in the GIT. At present, the number of probiotic-associated factors of L. rhamnosus GG that has beenidentified is still limited which makes it impossible todraw conclusions about the optimal harvesting phase forthis probiotic bacterium. However, as soon as the keymediators of probiotic traits have been identified, theresults presented here together with future studiesinvolving growth conditions mimicking those in the GITwill be instrumental for these purposes.

Experimental procedures

Growth conditions and chemical analyses

Lactobacillus rhamnosus GG (ATCC 53103) was maintainedin a laboratory culture collection as a glycerol stock at -70°Cand propagated at 37°C in MRS medium (Labema). Appro-priate dilutions of a culture grown for 17 h were plated onMRS agar and incubated anaerobically at 37°C for 3 days.Single colonies from these plates were resuspended in MRSmedium. One individual colony represented one biologicalreplicate. Cells were passaged through two sequential sub-cultures in MRS medium at 37°C for 12 h and one subculturein industrial-type whey medium containing 5% hydrolysedwhey, 0.6% casein hydrolysate and 0.0015% MnSO4·H2O at37°C for 17 h. Four Biostat Q fermentation vessels (B. BraunBiotech International) containing 750 ml of whey mediumwere inoculated at 2% (v/v) with four individual cultures of

Fig. 5. A model for physiological changes in L. rhamnosus GG during the progression of growth. Increased and decreased expression isindicated by plus symbol ‘+’ and minus symbol ‘-’ respectively.

Growth phase-associated changes in L. rhamnosus GG 761

© 2011 The AuthorsMicrobial Biotechnology © 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Microbial Biotechnology, 4, 746–766

L. rhamnosus GG. The cultures were grown at 37°C andstirred constantly (150 r.p.m. min-1). The pH was maintainedat 5.8 by automatic titration with 5% (v/v) ammonia. Tomonitor growth, samples were taken at regular intervals fromeach vessel, and their optical density at 600 nm was mea-sured. The amount of glucose, galactose and lactose in wheymedium during fermentation was determined by enzymaticmethods based on ISO 5765-1: 2002 (IDF 79-1: 2002) andISO 5765-2: 2002 (IDF 79-2: 2002) standards. Samples forRNA extraction were collected at the mid-exponential (4 h),late exponential (12 h), stationary transition point (16 h), earlystationary (20 h) and late stationary (31 h) phases (Fig. 1).Protein samples were collected at the mid-exponential (4 h),late exponential (12 h), early stationary (20 h) and late sta-tionary (28 h) phases (Fig. 1). Three parallel RNA samplesand four parallel protein samples from each sampling pointwere included in the analyses.

Transcriptomic methods

Experimental design, RNA methods, cDNA synthesis, labelingand hybridization. mRNA samples from three independentbiological replicates, i.e. from three fermentations, for eachtime point were hybridized, each sample independently, tomicroarrays according to an ANOVA design where all possibledirect pair-wise comparisons within biological replicates wereconducted. A total of 30 hybridizations were performed. Dye-swaps were carried out over biological replicate sets.

Depending on the growth stage, 1–4 ml of culture wasadded to 2–8 ml of RNAprotect Bacteria reagent (Qiagen),and the samples were processed according to the manufac-turer’s instructions. Cell pellets were stored at -70°C for laterRNA extraction. RNA was extracted and purified as describedpreviously (Koskenniemi et al., 2011).

For each sample, 5 mg of RNA was reverse transcribed tocDNA using the SuperScript Indirect cDNA Labelling System(Invitrogen) and fluorescently labelled with Cy3 or Cy5 mono-reactive dyes (Amersham Biosciences) according to previ-ously published protocols (Koskenniemi et al., 2011). Thelabelling efficiency was quantified using a NanoDropND-1000 spectrophotometer (NanoDrop Technologies), andlabelled cDNA samples were hybridized to microarraysaccording to Agilent’s procedure, ‘Two-Color Microarray-Based Gene Expression Analysis’ (http://www.agilent.com).

Scanning, image analysis and data analyses. Microarrayswere scanned at 5 mm resolution with a GenePix 4200 ALscanner (Axon Instruments). The fluorescence signal inten-sities of each feature were measured and addressed togenomic ORFs using GenePix® Pro 6.0 software (AxonInstruments/Molecular Devices Corp.) with default param-eters. The computational alignments were further improvedmanually, and features that could not be verified wereflagged.

This study was a part of a larger experiment, comprisinga total of 66 two-color microarrays, which were analysed asan entity. After pre-processing, samples involved in this par-ticular study were separated and analysed independentlyfrom the remaining data sets. All analyses were performedwith the Bioconductor package for statistical analysis usingthe R programming language (Gentleman et al., 2004). The

data were pre-processed with limma (Smyth and Speed,2003). The foreground and background median intensityestimates were corrected for the background using thenormexp-function (offset set to 50) (Ritchie et al., 2007),normalized within arrays using loess (100 iterations, suspi-cious spots and probe sequences matching multiple hits orborders of the coding regions were down weight to zero,and intergenic probe sequences were down weight to 0.1)(Smyth and Speed, 2003) and normalized between arraysusing quantile normalization (Yang and Thorne, 2003).Gene expression ratios were obtained by taking theaverage over the log2-transformed expression ratios of fea-tures describing the same gene and matching a singlegenetic coding locus. Expression ratios between conditionswere calculated using linear models implemented in limma(Smyth, 2004). Gene expression ratios were calculated for2798 genes out of the total 2944 genes predicted to beencoded by the genome (95% coverage). After pre-processing, samples involved in this study were separatedfrom the remaining samples, and the statistical significanceof the expression ratio of a gene between two conditionswas analysed using the paired t-test method implementedin CyberT (the Bayesian prior estimate of within-treatmentvariance was set to five, and the window size was set to101) (Baldi and Long, 2001). P-values were adjusted formultiple hypothesis correction using the Bonferroni methodand the number of performed t-tests in total (61 556). Sig-nificantly differentially expressed genes with at least atwofold change in expression ratio between paired timepoints were clustered in MeV version 4.3 (Saeed et al.,2006) by using the K-means clustering algorithm and aEuclidean distance metric. Data for the classification ofCOGs were obtained from Kankainen and colleagues(2009). The statistical significance of groups was assessedusing the hypergeometric distribution, and the results werecorrected with the Bonferroni method. Microarray platformand data details are available at the Gene ExpressionOmnibus (GEO platform GPL10580 and series GSE28903).

Proteomic methods

Protein extraction and CyDye labeling. Cells were harvestedfrom 1–5 ml of samples by centrifuging at +4°C and washedtwice with ice-cold 50 mM Tris-HCl pH 8 (Sigma-Aldrich).Bacterial cells were broken by bead beating (Koskenniemiet al., 2009). The protein samples were processed using aClean-up kit (GE Healthcare) and dissolved in 25–50 ml of abuffer containing 7 M urea (Sigma-Aldrich), 2 M thiourea(Sigma-Aldrich), 4% 3-[(3-Cholamidopropyl)dimethylam-monio]-1-propanesulfonate (Sigma-Aldrich) and 30 mM Tris(Bio-Rad). Protein concentrations in the samples were deter-mined using a 2-D Quant Kit (GE Healthcare).

Protein samples were adjusted to pH 8.5 by adding 2 MTris. The samples were then labelled using Cy2, Cy3 and Cy5dyes (CyDye DIGE Fluor minimal dyes; GE Healthcare),according to the Ettan™ 2-D DIGE protocol and Koskenniemiand colleagues (2009). The experiment was divided intothree parts; each time point (4 h, mid-exponential; 12 h, lateexponential; 28 h, late stationary growth state) was sepa-rately compared with the early stationary-phase (20 h) timepoint. Labelling was performed reciprocally so that both the

762 K. Laakso et al.

© 2011 The AuthorsMicrobial Biotechnology © 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Microbial Biotechnology, 4, 746–766

early stationary-phase samples and the other time pointsamples (4 h, 12 h or 28 h) were labelled with Cy3 and Cy5 toaccount for any preferential protein labelling by the CyDyes(labelling chart as Table S2). Cy2 was used for the internalpooled standard, consisting of equal amounts of eachsample.

2-DE and DeCyder analyses. The labelled proteins wereseparated by isoelectric focusing using immobilized pH gra-dient strips (24 cm, pH 3–10 non-linear, Bio-Rad) and aProtean IEF Cell (Bio-Rad) for 80 000 Vh (Koskenniemiet al., 2009). The strips were then equilibrated and loadedonto 12% acrylamide gels, and the gels were subjected toelectrophoresis in an Ettan™ DALTsix Electrophoresis Unit(GE Healthcare) (Koskenniemi et al., 2009). The gels wereimaged using an FLA-5100 laser scanner (Fujifilm), and theimages were cropped to identical sizes by removing areasextraneous to the proteins spots with ImageQuant™ TL 7.0software (GE Healthcare). After scanning, the gels werefixed in 30% ethanol and 0.5% acetic acid for a minimum of60 min and then silver stained (O’Connell and Stults, 1997).Image and statistical analyses for the cropped 2-D DIGEgels were performed using DeCyder™ 2D 7.0 software (GEHealthcare) (Koskenniemi et al., 2009). Approximately 700–800 separate protein spots were detected on each gel.Protein spots were picked and identified if they demon-strated at least a 1.5-fold difference in average spot volumeratio (average ratio � 1.5 or � -1.5) between different timepoints in at least three out of four separate biological rep-licates. Statistical significance was calculated using the Stu-dent’s t-test; a P-value value less than 0.05 was consideredsignificant.

Protein identification by mass spectrometry. MS-compatiblesilver staining (O’Connell and Stults, 1997) was performed tovisualize the protein spots for identification. Protein spots ofinterest were digested in-gel with trypsin and the peptidesrecovered as previously described (Koskenniemi et al.,2009). The resulting peptides were analysed by peptide-mass fingerprinting (PMF) or by fragment ion analysis withLC-ESI-MS/MS as described in detail previously (Koskenni-emi et al., 2011).

The PMF spectra were processed with FlexAnalysisversion 3.0 (Bruker Daltonik). The PMF and LC-ESI-MS/MSdata were searched with the local Mascot version 2.2 (MatrixScience) against the in-house database of the publishedORF set of L. rhamnosus GG, which contains 2944 proteinentries (Kankainen et al., 2009), using the Biotools 3.0(Bruker Daltonik) and ProteinPilot 2.0.1 (Applied Biosystem)interfaces respectively. The search criteria for both Mascotsearches were as follows: trypsin digestion with one missedcleavage allowed, carbamidomethyl modification of cysteineas a fixed modification and oxidation of methionine as avariable modification. For the PMF spectra the maximumpeptide mass tolerance was �50 ppm. The protonated mol-ecule ion ‘MH+’ and ‘monoisotopic’ were defined for the peakmass data input. For the LC-ESI-MS/MS spectra, both themaximum precursor ion mass tolerance and the MS/MS frag-ment ion mass tolerance were 0.2 Da, and a peptide chargestate of +1, +2 or +3 was used. A successful identification wasreported when a significant match (P � 0.05) was obtained.

In addition, to consider the LC-ESI-MS/MS identification reli-able, a minimum of two peptides with an ions score of at least40 was required.

Acknowledgements

Elina Ahola-Iivarinen, Hanna Jefremoff, Saija Laakso andEeva-Marja Turkki are acknowledged for their technicalassistance. This work was supported financially by theAcademy of Finland (Grants 210740 and 117746), theFinnish Funding Agency for Technology and Innovations(Grant 201/08) and by an ABS Graduate School scholarshipto K.K.

References

Alander, M., Korpela, R., Saxelin, M., Vilpponen-Salmela, T.,Mattila-Sandholm, T., and von Wright, A. (1997) Recoveryof Lactobacillus rhamnosus GG from human colonic biop-sies. Lett Appl Microbiol 24: 361–364.

Alander, M., Satokari, R., Korpela, R., Saxelin, M., Vilpponen-Salmela, T., Mattila-Sandholm, T., and von Wright, A.(1999) Persistence of colonization of human colonicmucosa by a probiotic strain, Lactobacillus rhamnosus GG,after oral consumption. Appl Environ Microbiol 65: 351–354.

Alcántara, C., Sarmiento-Rubiano, L.A., Monedero, V., Deut-scher, J., Pérez-Martínez, G., and Yebra, M.J. (2008)Regulation of Lactobacillus casei sorbitol utilization genesrequires DNA-binding transcriptional activator GutR andthe conserved protein GutM. Appl Environ Microbiol 74:5731–5740.

Axelsson, L. (2004) Lactic acid bacteria: classificationand physiology. In Lactic Acid Bacteria: Microbiologicaland Functional Aspects. Salminen, S., von Wright, A., andOuwehand, A. (eds). New York, USA: Marcel Dekker,pp. 1–66.

Azcarate-Peril, M.A., Tallon, R., and Klaenhammer, T.R.(2009) Temporal gene expression and probiotic attributesof Lactobacillus acidophilus during growth in milk. J DairySci 92: 870–886.

van Baarlen, P., Troost, F.J., van Hemert, S., van der Meer,C., de Vos, W.M., de Groot, P.J., et al. (2009) DifferentialNF-kB pathways induction by Lactobacillus plantarum inthe duodenum of healthy humans correlating with immunetolerance. Proc Natl Acad Sci USA 106: 2371–2376.

Baldi, P., and Long, A.D. (2001) A Bayesian framework for theanalysis of microarray expression data: regularized t-testand statistical inferences of gene changes. Bioinformatics17: 509–519.

Bandow, J.E., Becher, D., Büttner, K., Hochgräfe, F.,Freiberg, C., Brötz, H., and Hecker, M. (2003) The role ofpeptide deformylase in protein biosynthesis: a proteomicstudy. Proteomics 3: 299–306.

Claes, I.J.J., Lebeer, S., Shen, C., Verhoeven, T.L.A., Dilis-sen, E., De Hertogh, G., et al. (2010) Impact of lipoteichoicacid modification on the performance of the probiotic Lac-tobacillus rhamnosus GG in experimental cholitis. Clin ExpImmunol 162: 306–314.

Cohen, D.P.A., Renes, J., Bouwman, F.G., Zoetendal, E.G.,Mariman, E., de Vos, W.M., and Vaughan, E.E. (2006)Proteomic analysis of log to stationary growth phase

Growth phase-associated changes in L. rhamnosus GG 763

© 2011 The AuthorsMicrobial Biotechnology © 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Microbial Biotechnology, 4, 746–766

Lactobacillus plantarum cells and a 2-DE database. Pro-teomics 6: 6485–6493.

Cozzone, A.J. (1998) Post-translational modification ofproteins by reversible phosphorylation in prokaryotes.Biochimie 80: 43–48.

De Angelis, M., and Gobbetti, M. (2004) Environmental stressresponses in Lactobacillus: a review. Proteomics 4: 106–122.

Deepika, G., Green, R.J., Frazier, R.A., and Charalampopou-los, D. (2009) Effect of growth time on the surface andadhesion properties of Lactobacillus rhamnosus GG.J Appl Microbiol 107: 1230–1240.

Eymann, C., Becher, D., Bernhardt, J., Gronau, K., Klutzny,A., and Hecker, M. (2007) Dynamics of protein phospho-rylation on Ser/Thr/Tyr in Bacillus subtilis. Proteomics 7:3509–3526.

Francl, A.L., Thongaram, T., and Miller, M.J. (2010) The PTStransporters of Lactobacillus gasseri ATCC 33323. BMCMicrobiol 10: 77.

Gentleman, R.C., Carey, V.J., Bates, D.M., Bolstad, B., Det-tling, M., Dudoit, S., et al. (2004) Bioconductor: open soft-ware development for computational biology andbioinformatics. Genome Biol 5: R80.

Goldin, B.R., Gorbach, S.L., Saxelin, M., Barakat, S., Gualt-ieri, L., and Salminen, S. (1992) Survival of Lactobacillusspecies (strain GG) in human gastrointestinal tract. Dig DisSci 37: 121–128.

Guarino, A., Lo Vecchio, A., and Canani, R.B. (2009) Probi-otics as prevention and treatment for diarrhea. Curr OpinGastroenterol 25: 18–23.

Habimana, O., Le Goff, C., Juillard, V., Bellon-Fontaine,M.-N., Buist, G., Kulakauskas, S., and Briandet, R. (2007)Positive role of cell wall anchored proteinase PrtP in adhe-sion of lactococci. BMC Microbiol 7: 36.

Hatakka, K., Savilahti, E., Pönkä, A., Meurman, J.H., Poussa,T., Näse, L., et al. (2001) Effect of long term consumptionof probiotic milk on infections in children attending daycare centres: double blind, randomised trial. BMJ 322: 1–5.

Hojsak, I., Snovak, N., Abdovic, S., Szajewska, H., Mišak, Z.,and Kolacek, S. (2010) Lactobacillus GG in the preventionof gastrointestinal and respiratory tract infections in chil-dren who attend day care centers: a randomized, double-blind, placebo-controlled trial. Clin Nutr 29: 312–316.

Isolauri, E., Arvola, T., Sütas, Y., Moilanen, E., and Salminen,S. (2000) Probiotics in the management of atopic eczema.Clin Exp Allergy 30: 1604–1610.

Jacobsen, C.N., Rosenfeldt Nielsen, V., Hayford, A.E.,Møller, P.L., Michaelsen, K.F., Pærregaard, A., et al. (1999)Screening of probiotic activities of forty-seven strains ofLactobacillus spp. by in vitro techniques and evaluation ofthe colonization ability of five selected strains in humans.Appl Environ Microbiol 65: 4949–4956.

Kalliomäki, M., Salminen, S., Arvilommi, H., Kero, P., Koski-nen, P., and Isolauri, E. (2001) Probiotics in primary pre-vention of atopic disease: a randomised placebo-controlledtrial. Lancet 357: 1076–1079.

Kalliomäki, M., Salminen, S., Poussa, T., Arvilommi, H., andIsolauri, E. (2003) Probiotics and prevention of atopicdisease: 4-year follow-up of a randomised placebo-controlled trial. Lancet 361: 1869–1871.

Kalliomäki, M., Salminen, S., Poussa, T., and Isolauri, E.(2007) Probiotics during the first 7 years of life: acumulative risk reduction of eczema in a randomized,placebo-controlled trial. J Allergy Clin Immunol 119: 1019–1021.

Kankainen, M., Paulin, L., Tynkkynen, S., von Ossowski, I.,Reunanen, J., Partanen, P., et al. (2009) Comparativegenomic analysis of Lactobacillus rhamnosus GG revealspili containing a human-mucus binding protein. Proc NatlAcad Sci USA 106: 17193–17198.

Koistinen, K.M., Plumed-Ferrer, C., Lehesranta, S.J., Kären-lampi, S.O., and von Wright, A. (2007) Comparison ofgrowth-phase-dependent cytosolic proteomes of two Lac-tobacillus plantarum strains used in food and feed fermen-tations. FEMS Microbiol Lett 273: 12–21.

Koskenniemi, K., Koponen, J., Kankainen, M., Savijoki, K.,Tynkkynen, S., de Vos, W.M., et al. (2009) Proteome analy-sis of Lactobacillus rhamnosus GG using 2-D DIGE andmass spectrometry shows differential protein production inlaboratory and industrial-type growth media. J ProteomeRes 8: 4993–5007.

Koskenniemi, K., Laakso, K., Koponen, J., Kankainen, M.,Greco, D., Auvinen, P., et al. (2011) Proteomics and tran-scriptomics characterization of bile stress response in pro-biotic Lactobacillus rhamnosus GG. Mol Cell Proteomics10: M110.002741.

Laparra, J.M., and Sanz, Y. (2009) Comparison of in vitromodels to study bacterial adhesion to the intestinal epithe-lium. Lett Appl Microbiol 49: 695–701.

Lebeer, S., Verhoeven, T.L.A., Perea Vélez, M., Vanderley-den, J., and De Keersmaecker, S.C.J. (2007) Impact ofenvironmental and genetic factors on biofilm formation bythe probiotic strain Lactobacillus rhamnosus GG. ApplEnviron Microbiol 73: 6768–6775.

Lebeer, S., Vanderleyden, J., and De Keersmaecker, S.C.J.(2008) Genes and molecules of lactobacilli supporting pro-biotic action. Microbiol Mol Biol Rev 72: 728–764.

Lebeer, S., Verhoeven, T.L.A., Francius, G., Schoofs, G.,Lambrichts, I., Dufrêne, Y., et al. (2009) Identification of agene cluster for the biosynthesis of a long, galactose-richexopolysaccharide in Lactobacillus rhamnosus GG andfunctional analysis of the priming glycosyltransferase. ApplEnviron Microbiol 75: 3554–3563.

Lebeer, S., Vanderleyden, J., and De Keersmaecker, S.C.J.(2010) Host interactions of probiotic bacterial surface mol-ecules: comparison with commensals and pathogens. NatRev Microbiol 8: 171–184.

Lebeer, S., Claes, I.J.J., Verhoeven, T.L.A., Vanderleyden, J.,and De Keersmaecker, S.C.J. (2011) Exopolysaccharidesof Lactobacillus rhamnosus GG form a protective shieldagainst innate immune factors in the intestine. Microb Bio-technol 4: 368–374.

Macek, B., Mijakovic, I., Olsen, J.V., Gnad, F., Kumar, C.,Jensen, P.R., and Mann, M. (2007) The serine/threonine/tyrosine phosphoproteome of the model bacterium Bacillussubtilis. Mol Cell Proteomics 6: 697–707.

Macek, B., Gnad, F., Soufi, B., Kumar, C., Olsen, J.V., Mija-kovic, I., and Mann, M. (2008) Phosphoproteome analysisof E. coli reveals evolutionary conservation of bacterialSer/Thr/Tyr phosphorylation. Mol Cell Proteomics 7: 299–307.

764 K. Laakso et al.

© 2011 The AuthorsMicrobial Biotechnology © 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Microbial Biotechnology, 4, 746–766

Majamaa, H., and Isolauri, E. (1997) Probiotics: a novelapproach in the management of food allergy. J Allergy ClinImmunol 99: 179–185.

Mandlik, A., Swierczynski, A., Das, A., and Ton-That, H.(2008) Pili in Gram-positive bacteria: assembly, involve-ment in colonization and biofilm development. TrendsMicrobiol 16: 33–40.

Näse, L., Hatakka, K., Savilahti, E., Saxelin, M., Pönkä, A.,Poussa, T., et al. (2001) Effect of long-term consumption ofa probiotic bacterium, Lactobacillus rhamnosus GG, in milkon dental caries and caries risk in children. Caries Res 35:412–420.

Neuhaus, F.C., and Baddiley, J. (2003) A continuum ofanionic charge: structures and functions of D-alanyl-teichoic acids in gram-positive bacteria. Microbiol Mol BiolRev 67: 686–723.

Neves, A.R., Wietske, A.P., Kok, J., Kuipers, O.P., andSantos, H. (2005) Overview on sugar metabolism and itscontrol in Lactococcus lactis – the input from in vivo NMR.FEMS Microbiol Rev 29: 531–554.

O’Connell, K.L., and Stults, J.T. (1997) Identification ofmouse liver proteins on two-dimensional electrophoresisgels by matrix-assisted laser desorption/ionization massspectrometry of in situ enzymatic digests. Electrophoresis18: 349–359.

Ouwehand, A.C., Tuomola, E.M., Tölkkö, S., and Salminen,S. (2001) Assessment of adhesion properties of novel pro-biotic strains to human intestinal mucus. Int J Food Micro-biol 64: 119–126.

Ritchie, M.E., Silver, J., Oshlack, A., Holmes, M., Diyagama,D., Holloway, A., and Smyth, G.K. (2007) A comparison ofbackground correction methods for two-colour microar-rays. Bioinformatics 23: 2700–2707.

Saeed, A.I., Bhagabati, N.K., Braisted, J.C., Liang, W.,Sharov, V., Howe, E.A., et al. (2006) TM4 microarray soft-ware suite. Methods Enzymol 411: 134–193.

Sánchez, B., Bressollier, P., Chaignepain, S., Schmitter,J.-M., and Urdaci, M.C. (2009) Identification of surface-associated proteins in the probiotic bacterium Lactobacillusrhamnosus GG. Int Dairy J 19: 85–88.

Savijoki, K., Ingmer, H., and Varmanen, P. (2006) Proteolyticsystems of lactic acid bacteria. Appl Microbiol Biotechnol71: 394–406.

Seth, A., Yan, F., Polk, D.B., and Rao, R.K. (2008) Probioticsameliorate the hydrogen peroxide-induced epithelialbarrier disruption by a PKC- and MAP kinase-dependentmechanism. Am J Physiol Gastrointest Liver Physiol 294:G1060–G1069.

Smyth, G.K. (2004) Linear models and empirical bayesmethods for assessing differential expression inmicroarray experiments. Stat Appl Genet Mol Biol 3: Article3.

Smyth, G.K., and Speed, T. (2003) Normalization of cDNAmicroarray data. Methods 31: 265–273.

Song, X.-M., Connor, W., Hokamp, K., Babiuk, L.A., andPotter, A.A. (2009) The growth phase-dependent regula-tion of the pilus locus genes by two-component systemTCS08 in Streptococcus pneumoniae. Microb Pathog 46:28–35.

Soufi, B., Gnad, F., Jensen, P.R., Petranovic, D., Mann, M.,Mijakovic, I., and Macek, B. (2008) The Ser/Thr/Tyr

phosphoproteome of Lactococcus lactis IL1403 revealsmultiply phosphorylated proteins. Proteomics 8: 3486–3493.

Szajewska, H., Kotowska, M., Mrukowicz, J.Z., Armanska,M., and Mikolajczyk, W. (2001) Efficacy of LactobacillusGG in prevention of nosocomial diarrhea in infants.J Pediatr 138: 361–365.

Szajewska, H., Skórka, A., Ruszczynski, M., andGieruszczak-Białek, D. (2007) Meta-analysis: Lactobacil-lus rhamnosus GG for treating acute diarrhoea in children.Aliment Pharmacol Ther 25: 871–881.

Tao, Y., Drabik, K.A., Waypa, T.S., Musch, M.W., Alverdy,J.C., Schneewind, O., et al. (2006) Soluble factors fromLactobacillus GG activate MAPKs and induce cytoprotec-tive heat shock proteins in intestinal epithelial cells. Am JPhysiol Cell Physiol 290: C1018–C1030.

Tuomola, E.M., Ouwehand, A.C., and Salminen, S.J.(2000) Chemical, physical and enzymatic pre-treat-ments of probiotic lactobacilli alter their adhesion to humanintestinal mucus glycoproteins. Int J Food Microbiol 60:75–81.

Vélez, M.P., Petrova, M.I., Lebeer, S., Verhoeven, T.L., Claes,I., Lambrichts, I., et al. (2010) Characterization of MabA, amodulator of Lactobacillus rhamnosus GG adhesion andbiofilm formation. FEMS Immunol Med Microbiol 59: 386–398.

Welman, A., Maddox, I., and Archer, R. (2003) Exopolysac-charide and extracellular metabolite production by Lacto-bacillus delbrueckii subsp. bulgaricus, grown on lactose incontinuous culture. Biotechnol Lett 25: 1515–1520.

Welman, A.D., Maddox, I.S., and Archer, R.H. (2006)Metabolism associated with raised metabolic flux to sugarnucleotide precursors of exopolysaccharides in Lactobacil-lus delbrueckii subsp. bulgaricus. J Ind Microbiol Biotech-nol 33: 391–400.

Yan, F., and Polk, D.B. (2002) Probiotic bacterium preventscytokine-induced apoptosis in intestinal epithelial cells.J Biol Chem 277: 50959–50965.

Yan, F., Cao, H., Cover, T.L., Whitehead, R., Washington,M.K., and Polk, D.B. (2007) Soluble proteins produced byprobiotic bacteria regulate intestinal epithelial cell survivaland growth. Gastroenterology 132: 562–575.

Yang, Y.H., and Thorne, N.P. (2003) Normalization for two-color cDNA microarray data. In Science and Statistics: AFestschrift for Terry Speed. Goldstein, D.R. (ed.). Beach-wood, OH, USA: Institute of Mathematical Statistics,pp. 403–418.

Zhang, Y.-M., and Rock, C.O. (2009) Transcriptional regula-tion in bacterial membrane lipid synthesis. J Lipid Res 50:S115–S119.

Supporting information

Additional Supporting Information may be found in the onlineversion of this article:

Fig. S1. K-means clustering of transcriptomic expressionprofiles. Significant changes in gene expression are repre-sented as log2 intensity ratio values between paired timepoints. Genes were grouped into 16 clusters using a Euclid-ean distance metric, and the analysis included 636 genes in

Growth phase-associated changes in L. rhamnosus GG 765

© 2011 The AuthorsMicrobial Biotechnology © 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Microbial Biotechnology, 4, 746–766

total. The purple expression profile in each cluster representsthe cluster median.Fig. S2. Representative 2-D DIGE images of proteinsextracted from different growth phases: (A) 4 h and 20 h,(B) 12 h and 20 h and (C) 20 h and 28 h. The total amountof protein used for CyDye labelling was 105 mg. The num-bered protein spots were cut from silver-stained 2-D DIGEgels and subjected to MS or MS/MS identification. Proteinspots that were more abundant at time point 20 h areshown in red and those more abundant at time points 4 h,12 h or 28 h are shown in green. The protein spots with nodifference in abundance between the treatments areyellow.Table S1. Log2 ratio values of significantly (P � 0.01) differ-entially expressed genes with at least a twofold change in theexpression ratio between paired time points. Clusters refer toFig. S1.

Table S2. Set-up of the DIGE experiments. Letters A–D referto biological replicate samples. Gels 1–4, 5–8 and 9–12 eachform a separate DIGE experiment.Table S3. Proteins with differing abundance between themid-exponential (4 h) and early stationary (20 h) growthphases.Table S4. Proteins with differing abundance between the lateexponential (12 h) and early stationary (20 h) growth phases.Table S5. Proteins with differing abundance between theearly stationary (20 h) and late stationary (28 h) growthphases.

Please note: Wiley-Blackwell are not responsible for thecontent or functionality of any supporting materials suppliedby the authors. Any queries (other than missing material)should be directed to the corresponding author for thearticle.

766 K. Laakso et al.

© 2011 The AuthorsMicrobial Biotechnology © 2011 Society for Applied Microbiology and Blackwell Publishing Ltd, Microbial Biotechnology, 4, 746–766