Functional Analysis of D-Alanylation of Lipoteichoic Acid in the Probiotic Strain Lactobacillus rhamnosus GG

ApPLIED ANDENVIRONMENTAL MICROBIOLOGY,

June

2007, p.3595-3604 0099-2240/07/$08.00

+

0 doi:10.1128/AEM.02083-06

Copyright © 2007, American Society for Microbiology.All Rights Reserved.

Vol. 73, No. 11

Functional Analysis of D-Alanylation of Lipoteichoic Acid ^In the Probiotic Strain Lactobacillus rhamnosus GG'V

Monica Perea V61ez,1 Tine L. A. Verhoeven,' Christian Draing.f Sonja Von Aulock.i Markus Pfitzenrnaier," Armin Geyer," lVG Lambrichts," Corinne Grangette,"

^I

Bruno Pot," J08 Vanderleyden,' and Sigrid C. J. De Keersmaecker! "

Centre of Microbial and Plant Genetics, Katholieke Universiteit Leuven, Kasteelpark Arenberg 20, 3001 Leuven, Belgium

I;

Department of Biochemical Pharmacology, University of Konstanz, D-78457 Konstanz; Germany'; Institute for Organic Chemistry, Philipps-Uni versitiit Marburg, D-35043 Marburg , Germany:'; Laboratory of Histology, BIOMED Researcli Institute,

University Hasselt, Agoralaan Gebouw D, 3590 Diepenbeek, B elgium "; and Bacteriologic des Ecosystemes.

Institut Pasteur Lille, 59019 Lille Cedex, France"

Received 4 September 2006/Accepted 24 March 2007

Lipoteicholc acid (LTA) is a macroarnphiphile molecule which performs several functions in gram-positive bacteria, such as maintenance of cell wall homeostasis. n-Alanylation of LTA requires the proteins encoded by the dlt operon, and this process is directly related to the charge properties of this polymer strongly contributing to its function. The insertional inactivation of dUD of the probiotlc strain Lactobacillus rhamnosus GG (ATCC 53103) resulted in the complete absence of n-alanyl esters in the LTA as confirmed by nuclear magnetic resonance analysis. This was reflected in modifications of the bacterial cell surface properties. The

dltD

strain showed 2.4-fold-increased cell length, a low survival capacity in response to gastric juice challenge,

~1O

increased sensitivity to human beta-defensin-Z, an increased rate of autolysis, an increased capacity to initiate growth in the presence of an anionic detergent, and a decreased capacity to initiate growth in the presence of cationic pep tides compared to wild-type results. However, in vitro experiments revealed no major differences for adhesion to human intestinal epithelial cells, biofilm formation, and immunomodulation. These properties are considered to be important for probiotics. The role of the dlt operon in lactobacilli is discussed in view of these results.

The cell wall of gram-positive bacteria constitutes a protec- tive barrier essential for survival, shape, and integrity. Proteins and teichoic acids (TAs) composed of wall teichoic acid (WTA) and/or lipoteichoic acid (LTA) are associated with this peptidoglycan-containing wall (33, 40). WTA is covalently linked to the peptidoglycan, whereas LTA is a macroamphi- phile molecule with its glycolipid moiety anchored to the mem- brane and its polyglycerophosphate (Gro-P) chain extending into the wall (40). Together, TAs are the most abundant poly- anions of gram-positive bacteria and represent up to 50% of the cell wall dry weight (24). Glycosyl substitutions in WTA and D -alanyl ester (u-Ala ester) substitutions in LTA are directly related to the charge properties of these polymers and strongly contribute to their function, as evidenced by the phenotypes of strains with genetic changes altering the glycosyl substitution and n-alanylation process (17, 33, 45, 57) .

The n-Ala ester substitution of LTA requires four proteins encoded by the dlt operon. Two of these proteins are the n-alanyl carrier protein ligase (Del, encoded by dltA), which activates n-alanine by use of ATP, and the n-alanyl carrier protein (Dcp), which is encoded by dlrC. DltB is a putative transmembrane protein predicted to be involved in the passage of the activated n-alanyl-Dcp complex across the glycerol phos-

;0Corresponding author. Mailingaddress: Centre of Microbial and Plant Genetics,K.U.Leuven, KasteelparkArcnberg20,3001 Leuven, Belgium. Phone: 32 16 321631. Fax: 32 16 321%6. E-mail: sigrid .dcke c rsm aecke rq!'b iw.k ule uve n .be .

"Published ahead of print on 13 April 2007.

3595

phate backbone of LTA. Finally, the DltD membrane protein facilitates the binding of Dcp for ligation with n-Ala and ad- ditionally has thioesterase activity for removing mischarged o-alanyl carrier proteins (13, 40). Inactivation of genes within this operon in various gram-positive bacteria results in the decrease or complete absence

of n-Ala

esters from

LTA (1, 18,

35, 46, 47). n-Ala-delicicnt mutants are found to exhibit

Cl

variety of phenotypic changes that could be attributed to the resulting charge modification of their cell surface. For in- stance, alterations of cell morphology associated with defects in septum formation (9,41 ,45), variations in modulation of the activity of autolysins (6, J 9,55,58), differences in cation bind- ing to the cell envelope required for enzyme function

(33),

alterations of the electromechanical properties

of

the

cell wall

(40), altered resistance to antimicrobial cationic pcptides

(H.

40), modified adhesion, epithelial cell invasion, and virulence (1, 35) , effects

on

biofilm formation (7, IX, 23), and alte rations in immune response (14, 22, 38, 39, 51) have been reported.

While it is apparent that the o-Ala esters

of LTA

play an essential role in the physiology and properties of the cell sur- face of gram-positive bacteria, the genotype-phenotype rela- tion of the dlt operon is complex and appears to be species dependent.

By construction of a dltl) knockout mutant, this study

aimed

to determine first the role

of

the dltl) gene in the

u-alunylation of

the LTA

in

the

probiotic

strain Lactobacillus thamnosus GO. Probiotic bacteria are defined as "live microurgunisms which, when administered in adequ a te amount s. confer a health benefit on the host"

(20).

L. rlutmuosus GG is a well- First publ. in: Applied and environmental microbiology 73 (2007), 11, pp. 3595-3604

Konstanzer Online-Publikations-System (KOPS) URN: http://nbn-resolving.de/urn:nbn:de:bsz:352-opus-79816

URL: http://kops.ub.uni-konstanz.de/volltexte/2009/79816/

35lJ6

TABLE

I. Bacterial strains and plasrnids used in this study

Strain orplasmid

Strains

E. coli

DF£S(x

S. entcricascrovar Typhimuriurn

SLl344

L.rhuinnosusGG

Wild type CMPG5540 CMPG5541

Plasmids

peRII-TOPO

pF!\./530

I

rLABI301 pMD5057 pCMPG5221 pCMP(j5222 pCMPG5223 pCMPG522lJ pCMPG5227 pCMPG5lJ01

Relevant genotype or description

F-<1>80dlacZ6.MI5 6.(IacZYA-argF)UI69deoRrecA1endAI hsdR17(r_{K -} m_{K - )} supE44A-thi-Igyl:4

96

relA1

,\)'1 hisG'11.11.;

virulent: Srn'

Human isolate

dltl)

knockout mutant of

L. rhumnosus GG; dltD::terR

CMPG5540 complemented by pCMPG5227

Cloning vector, ampicillin and kanamycin resistance Cloning vector, pLJC18 derivative. erythromycin resistance

E.coli-Lactobacillus shuttle vector, ampicillinand

erythromycin resistance

Tetracycline

resistance plasmid from Lactobacillus

plantatum 5057

peRII-TOPO vector containing the 2,682-bp dlti: gene and Banking regions

from L. r!z<WIIIO,I"lI'\"

GG (Pro-230 and Pro-23

I )

pFAJ5301 containing an HindlII-EcoRV fragment (dltD gene and flanking regions)

Suicide vector to knock out the L rhamuosus GG dltl) gene through insertion

of [erR

gene from pMD5057 in the BbsI-NcoI sill'

pCMPG5YOI containing the functional

dltl)

gene driven by the

L.rhamnosus GG IdhL promoter

PLAB1301 containing the functional dltl) gene driven by the ldhl: promoter peRIl·TOPO vector containing the ldlil. promoter (176 bp)

from L. rhamnosus

GO

amplified using primers

Pro-I27

and Pro-128

Source or reference

Gibw·BRL

26

ATC(: 531lJ3 (54)

This

study This study

Invitrogen

Unpublished results 28

11 This study This study This study This study This study Unpublished results

studied probiotic strain (4,21). For this reason, the second aim of this study was to investigate the role of the n-alanylation of the L.

,.!lanlIWSIIS

GG LT'A in some morphological, membrane charge property, and prohiotic characteristics. such as adhcr- cnce to human intestinal epithelial cells, hiofilm formation, resistance to gastric juice challenge, and imrnunomodulation of human intestinal epithelial cells and peripheral blood mono- nuclear cells.

MATERIALS AND METHODS

amplified from plasmid pMD5057 ofL. plantarum50S7(11)by useof primers 1'1'0-221 andjJro-222.Theresultingsuicide vector,pCMPG5223,was electropo- rated into1..rluimnosusGG (15).Trunsformants were selected for resistance tu )(Jp.g)ml of tetracycline. Confirmation ofDNArecombination\V,tSperformed hy peRusing primers Pru-2(1! and Pru-231and Southern hyhridizution lISill)! a dltl) probe synthesized with primers Pm-26! and 1'1'0..262. The dtti) mutant was designated CMPG55411 andwasfurtheranalyzed,

Complemenratiun or the knockout dltl) mutant. Fur the constructionIll"the complemented strain, tbe functional dltl) gene was cloned under the control of

the L.rlWlIlIlO.I'lI.1" GG ldhl: promoter in pCMPG5lJUI. yielding pCMPG52211,

Subsequently, an Eclll136-NotJ fragment containing the Iunctionul dltl) genc and theldhl.promoter was subcloncd intopLABI30\ (2H),This construct was

.. UnderlinedSl·ljUCl1C..·s represent restrictionsitl'sindicated parcnthcticallv,

TABLE 2. Primer sequences used in this study

Pro-12X

(Noel) ...GT~":.t..I_6IQGATATCATCCTrTCTTATGTGC

ATGC

Prn-127 (SacI)

CTQ!.~Gc.rr:CIT(;TCACAGGATrCACAAGT

CTTGC

Pn<!21 (Ecok l) ....

(i.L\..tUTCGAGATTCCITrACAAATATGCTC

TrAC

1'1'0-222 (EctlRI) ....

C~L6bTn:Cj'rrc·GGAATA(j(jTL'\·['/\C'TAGA

CAAAAG

Pn,-230

GCCGG'rrACAG'ITTGTrCGCGG

Pro-231 ACAGGAGGCACAACA·rGGCAAAATC

Pro-261

TCCA'T·GTCGGCCAACG'rGCT

Pro-262 A'lTGGCT·TCCCGCCATTCCC

Pnl-443 ,

TCAACrn

v r G T r A A ( i A T G C i A T

Pro-444 GG ATAA'lTITCCfGCrrG AA

Pro-445 TfTGTCGAAATI'GCG"n'ACT

Pro-446 TCGAAACACCGATGAACTTT

Bacterial strains, plasrnids, media, and growth conditlons. The bacterial xuuius and plasmid» used inIhisstudy an: listed in TubleI.Lr!1lIf1I1IOSlI,\'GG was routinely grown innonshaken MHS medium (Difco) at:'7"C.Escherichiu coli cc·lIsweregrown in l.uria-Bertani (1.13) m..xliumwith aeration at"J7'lC When required, antibiotics were used at the following conccntrauons: 10~Lglmltetra- cvcliuc,lOllp,g/llll .nnpicillin, and:'i "..Will I(L.r!WIIIllo'\'1I.IUti) or lOOfLgirnl(E, coli; ervthromycin.

DNA manipulntious, Routine molecular biology techniques wen: employed as desciihcd before (49). Primer scqucnces used in this study arc listed in Table 2 (Furogcntcc),Enzymes formolecularhilllngywere purchased lrum New En- gland Biolahsand used accordingto the instructions uf thesuppliers. Plasmid DNA from l:..coli wasprepared using ()IAGEN Miniprep kits. Chrornosomal DNA from L I'IlIJIIIlIO\lIJGO was isolated as previously described (15),

('nllslt'ul'liollllnd aualysls of theL.rllOt1IIIOSU,\'<iGdUDmutant (CMPG5S4IH.

I'hccomplete genome sequenceof the tllt operon of Lactobacillus,.1l1l1ll1W.WI.l'

ATCC ^74(l1) is published in the NCBI database under accession number

1\1"1¹)2553(1.\),Based un this sequence, a fragment ul' 2,6K2hpcontaining tileL.

rIWIll!UI,\/II (iCidltl) gene and its 77Y·hp upstream and 635-hpdownstream

rl:,~IlIUSwaxarnpl"il~dusIng.prunersI'r11-230andPro-231andclonedin

pen,n-

1'01'(l. yil'ldiug pCMPCj5221, Subsequently, the dltl) gene with its flunking

Il'glllllSwas suhcloncd itS a l lindl l l-Iicuk V fragment in pFAJ5"J() I. resulting

III pC·I'vlI'{i5222. Tu inactivate the dltl) gene. a Bbsl-Ncol fragment from pt ·[\11'(i~~.:.~ \\,1~r"'plilCl:d hvllwtctr.uyrlinercsisuincc _~·assc·tICIt"(I'vl)previously

Bacterial primer Sequence (5' III3')"

3597

designated pCMPG5227 and introduced in CMPG5540 by elcctroporation as previously described (15). Transformants were selected for resistance to 10 f.l.glml of tetracycline and to 5 ug/rnl of erythromycin. The complemented strain, CMPG554-t, was confirmed by PCR using primers Pro-127, Pro-2n2,andPro-261 and included in all assays performed.

Cloning and analysisofthe L. rhamnosus GG dlt operon. Concomitantly, in ordertosequence the additional genes within the dlt operon ofL.rhamnosus GG, primers Pro-4-43, Pro-444, Pro-445, and Pro-446 were designed after selec- tion on the DNA sequence of the (/1/ operon ofL.rluunnosusATCC 7469.The PCR:; were carried out withPIx(Invitrogen),a DNA polymerase with proof- reading activity.The PCR fragments were cloned in pCRIl-TOPO and se- quencedbythe chain terminationdideoxynucleosidetriphosphate method (50) witha BigDyeTerminator V3.1 cyclesequencing kit, using an ABI PRISM 3100-Avant genetic analyzer (Applied Biosysterns, Lennik, Belgium).Datahases were screened fur similaritiesbyusing BLAST(1,3), andalignmentof overlap- ping fragments was performed with Vector NT[ Advance 10 ContigExpress software (Inforrnax, Oxford, United Kingdom).

LTA purifientlnn, For the LTA isolation, 11.83 g of lyophilizedL.r!UlnWOSUS

GOcells and 22.12gofthe dltD mutant (CMPG5540) cells were extracted using butanol and hydrophobic interaction chromatography as previously described (38), with minor modifications. Briefly, bacterial cells were disrupted [or 15 min by sonication and resuspended in 50 ml of 0.1 M citrate buffer (pH 4.7). The bacteriallysate was mixed while being stirred with an equal volume of n-butanol (Merck.Darmstadr,Germany) for20ruinat room temperature. The suspension was thencentrifugedat8,J(j() x 8 for 4() min,resulting in atwo-phase system,The lower aqueous phase. was lyophilized and subsequently resuspended in 35 ml of chromatography start buffer (IS'!';!I/-propanol inIIIM ammonium acetate: pH 4.7) followed by GO min of ceutrlfugution at 26,900 X gand sterilization by Illtration (0.2. urn). The supernatant was subjectedto hydrophobic interaction chromatography on an ocryl-Sepharose column (2.5 by 11 cm) using a linear gradient of from 15% to 60% n-propanol in 0.1 M ammonium acetate (pH 4.7).

LTA-containing fractionswere identified by their phosphatecontentbased on the formation of phosphornolybdenum blue from phosphate (52). The endotoxin contamination of the LTA preparations was assessed by the kinetic Limulus amoebocyte lysate assay (Charles River, Charleston,

sq.

LTA structure analysis by NMR spectroscopy, Nuclear magnetic resonance (NMR) spectra of the LTA samples were determined with a Bruker Avance 600 MHzspectromeierwith Cl 5-111111probe at .lOO K. Spectra wereobtalned using Di)solutions and 3-(tl"imethylsilyl)3,3,2,2-wtradeuteropropionic acid sodium salt (d4-TSPAjas an iruernal standard forIHNMR (81-1' 0.00 ppm) and acetone for chemical shifts of I.'C(oe,.10.02ppm), Assignments were taken from two- dimensional hornonuelcar douhlc-quantum-Iiltcrcd correlationspCClTUSCOPY,lo- tal correlated spectrnscopy, and rotational nuclear Overhauser effect spectres- copy experiments using a water suppression technique and two-dimensional heteronuclear single-quantumcorrelation

eH

to LlC) spectra. In the total cor- related spectroscopy experiments the mixing times were 100I11S,and the rota- tional nuclear Overhauser effect spectroscopy experiments were performed with 2.00 msofspinlocktime. Data acquisition and processing were done using standard Bruker software.The average number of repeating units in the poly- glycerophosphate backbone, the percentage of substitution, and the chain length of the fatty adds in the membrane anchor were calculated directly from the integrals of the proton spectra.

Transmlssion electron microscopy, Bacteria were grown overnight ([6 h at 37"C). Uncoatcd grids were used as a probe to adsorb bacterial cells. The grids were placed on a drop nf bacreriul suspension for 15 s, incubated ill 0.25%

phosphotungstenic acid (pH == 7) for 30 s, and washed three times, and excess liquid was drained. The hacteria were observed \vitb a Philips EM 208S trans- mission dCC·tl'llll microscope al 80 kY. Images were digitalized using an SIS image alUllysis system.

Analysis of lIutolysis. Triton X-lOO was used to study induced autolysis under nongrowing conditions as previously described (48), with minor modifications.

Briefly, cells were grown overnight, harvested by centrifugation (4,000

x

gfor 20 min ut 4°C), washed three times with equal volumes of phosphate buffer (pH 6.5;

IOmM), and resuspended (optical density at 600 om rODh ll o ] of -1.6) in phosphate buller (pH 6.5; 200 mM) containing Cl.05% (vollvol) Triton X-lOO.

'111e cell slIspension was incubated at 37°C untler agitlltion, and autolysis was rnnnitmed by examining the decrease of OD_{w ll}in time.

G.·owthin the prcsence of the cationic pC[ltides nisin and polymyxin B. Nisin and polymyxin B were purchused from Sigma-Aldrich. Overnight cultures con- taining ahoul !OS to10"CFU/m[ of1..r!Wl/l/lOSIISGG wild-lype, (/ltD mulant,and complemcnted strain cells were diluted [S,OOO-fold in M RS medium containing either nisin at a concentration hetween ll.2 and 2.5 f.l.glml or polymyxin 13 at a concentration hetween 8 and 1,OO() ,...g/ml. Bacteria were grown at 37"C with

continuous shaking in ordertoavoid cell aggregation, and the ODw lJwas mea- sured automatically each 30 min during at least SO h using a Bioscreen C instrument (Labsysrems Ltd. Oy).Concomitantly, growth ofL.rhumnosus GG.

the dltD mutant, and the complemented strain in the presence of an anionic detergent (sodium dodecyl sulfate [SDS]at 0.01G;;;,and 0.015/1-0)was assessed.

Eachrunwas performed at least in triplicate.

Human beta-defensin sensitivity assay. L.r!UlI1I110SIISGG wild-typeand dltl) mutant cells were tested for their sensitivity to3ug/rnlhuman beta-defensin I (hBDI) and hBD2. After 3 h of incubation at 37°C, bacterial viability was measured by plating serialdilutions on MR5agar as previouslydescribed (15). The experiment was performed in triplicate.

Survival in slmulated gastric juice. Simulated gastric juice was prepared and survival tests were performed as previously described (10). The percentages of survival were calculated by comparing the cell numbers before and after additiontosimulatedgastric juice at 0, 30, 60. and gO min.The experiment was performed in triplicate.

In vitro adhesion assay to Caco-2 and HT-29 human intestinal epithelial cell lines. Caco-Z and HT-29 cells were purchased from ATCC (Rockville, MD).

Cells were routinely grown in 75-cm²culture Ilasksunderconditionsof 37'"'(:.:'i':;;

CO;;:, and 90/-HJ relative humidity.For both cell lines, Dulbecco modified Eagle medium (DMEM)/F·12 (GibcoBRL) (1:1) supplemented with 10% fetal bovine serum (FBS; HyClone) but without antibiotics was used as the culture medium.

Cells were passaged~vCI)'3days (at 70 to 80';'0confluence) ata split ratio of 1 to 7. For adhesion experiments, epithelial cells were plated at a densiry of40,OO() cells/crrr'ill 12-well plates(Cellstar). Confluencewas reached within.'IO.:tdays after seeding, and rnonolayers were used for the experiments 15 days aftl;:l seeding.Adherence of L. rh(1II11/0SUS GG wild-type,dltl) mutant, and comple- mented strain cells to the epithelial cells was examined by adding 1.5 ml of DMEM containing 10⁹CFU/ml.After incubation at 37°C for 45 min, epithelial cells were washed two times with prewarrned phosphatc-bullercd saline (1'135).

Subsequently, 100 j.l.l of 1X trypsin-EDTA (Invitrogen) was added to eachw~1I and Incubated for IQmin at 37°C. Finally, 900 p.1 of PUS was added and mixed.

and serial dilutions were plated out. Plates were incubated at 37°C for72h.

Adhesion ofL.rhamnosusGG wild-type, dltl) mutant, and complemented strain cells was tested in triplicate in three independent experiments.

In vitro biofiIm assay. In vitro biofilm formation was determined as previously described (16, 36), with minor modifications. Briefly, biofilm formation was assayedon polystyrene pegs hanging into microtitcr plate wells.The pcgs wer ...

placed in wells containing a bacterial concerurationof 3 x ^J(J"1CFU in 2(JIJ p.l AOAC medium(Ditco) (36)andincubated inanaerobicjarsfor 72h al37"C, Medium was changed every day. Wells containing sterile AOAC medium andL.

rhainuosusGG wild-type cells were included as negative and positive controls, respectively.Each experiment was performed in triplicate.

lmmunomodulation of HT-29 cell line. HT-29 cells were maintained as de- scribed above.L.r!1II1111l0SlIS GG wild-type anddltl)mutant cells weregrown overnight, centrifuged at 4,tlOO xgand 4^{Q C}for 20rnin,and washed with cold PBS. Immunomodulatiou of HT-29 cells was examined by adding 1.5 1111 of DMEM without FBS containing f x 10⁶ CFU/ml of either1..rhamnosus GG wild-type or dltD mutant cells. Salmonella cnterica serovar Typhimuriurn SLl344 cells in a concentration of5X 10" CFU/ml and interlcukin-! beta(11.-113) (Sigma-Aldrich) (10 ng/ml) were used as positive controls. and DMEM without FBS was used as a negative control. After incubation at 37"C for .1.5 h (5^r:i.CO~

and9WIi,humidity), epithelial cells were washed two limes with prewarmed fiBS.

Subsequently, 200 1.1.1 of PBSwasadded to each well and RI\'A extractionwas performed by using a High Pure RNA isolation kit (Roche Molecular Biuchcrni- culs) and following the manutacturer's instructions. Alter isolation, RNA integ- rity was analyzed lIsing an Agilent RNA (iOn kil and 2100 hi(lallillyzer expel'!

software. Cytokine gc:ne cxpression was determincd hy real-lime reverSL'trans- crjptas~peR using primers and probes as prcviollsly descrihcd (43).

Cytol<ine induction in periphcnlll.llood mO/lonuclellr cells(PBMC).CVlokine induction using a 24-h culture of hacterial cells, and using Streptococcus.~f1rd(}lIii LMG17843 andE.coliTOl cells asint~rnalcontrols, was pcrformed as previ- ously describecl(22).

Nucleotide sequcnce accession /lumber. The tilt sequence of1..rhallll/lI.l'II.1'GG has been deposited in the NCBI database undcr GenBank accession numhel DQ906101.

RESULTS

Analysis of the dlt operon of L.

rlza11l1l0S11S

GG. Sincl' thl'

genome sequence of L.

rhamnosus

GG is unavailable. a ho-

mology strategy based on the

till

sequence of

L. rllet/lIl/osus

35<.)1:\

r

⁶⁵

i60

1"S5

1 50 a

r⁴⁵

13

r ~~ ~

!'30

g

125 ~

;20 ~

~1 5 ~

!10

r

^S

,0

--dltD

mutant

9 14 19 24 29 34 39 44 49 54 59 64 69 Fraction No.

(B) 10 9·

~ ^B

15 7

~

^6·

2

tIl

..cc.

IIIo s:0..

The molecular structure determined by NMR spcctroscopy revealed that LTA

from L. rIIlJIIllIOS/lS

GG wild-type cells

IS

constituted or Gro-P with n-alunyl esters as unique detectable substitucnts

(74(~t,

o-Ala.Gro-P), LTA from the L.

rlWlJlfIOSLlS

GG

wild type is formed

by

average chains or

⁵⁰

glycerophos- phate residues (Fig . .:1). For the

dllD

mutant. two fractions were eluted from the chromatography column as previously men- tioned. The major peak (fractions

41

to

46)

and the second peak (fractions 48 to 55) contain LTA molecules with average lengths of only 29 and 7 glycerophosphate residucs, respec- tively. D-Alanyl ester substitution in the dlt I) mutant was

com-

pletely abolished in both peaks. The analysis of the

glycolipic

moiety in the ... ild type reveals

an

average Iuuy acid chain of

('1-1'

with one double bond per fatty acid (two double bonos in every rncmhrane anchor). The Iauy acid chain in the

dlt l)

mutant

wasOil

average two carbon atoms longer

(C\I.)

in

com-

parison to that seen in the parental strain. Likewise, as in the

5.8 kb

FIG. I. Genetic organization or the

L.rluimnosus

GG

dlt

operon. A schematic representation of the

dlt operon (-S.Hkb)

of

L.rltamnosus

GCi as revealed by

sequencing

and BLAST analysis

is

presented. Open reading frames corresponding to

dltA, dltl), tlltC,

and

dltl)

are indi- cated with arrows. The putative promoter region

(5' end),

represented in the (lgure by a flag, contains a-I

(J"1'

AITAA" region a nd a - 35

"TGGTTT"

region separated

by 19

bp. The potential ribosome bind- ing site

"GGGGG"

is located 1'1 bp upstream of the putative

ATGstart codon

of

cIlrA

(not shown).

A

second potential ribosomal binding site.

"A AAG A(J C," was found <}

bp upstream of the putative start codun of

the

elite

gene (nUl shown).

e1llE

overlaps the stop

codon

of

cilIA

by

I

bp, anti

(1I,f)

overlaps the stop codon of

dl,e by

4 hp.

c/lle

and

dllB

arc separated

by

a

69.\)p intergcnic

region.

The 3' end

of the operon

reveals

a

putative

terminator

^l)l)

bp downstream or the

e1IID

stop

codon,

represented in the figure by a stem-loop.

It

is a 12-bp inverted repeal followed by a series of T rcsidues

(TTTATTIT).

ATCC 7469 (AFI92553) (13) was used to isolate the dlt genes within the dlt operon of

L.

rhamnosus GG cells. Based on this strategy, a continuous 5.H-kb sequence of

L.

rhumnosus Gc:.:;

genomic DNA was determined . The analysis of the

sequence

revealed four upcn reading Irurncs (dlut dhls, dlt;', and dill), as depicted ill Fig. I. Additionally, the in silico-translated pro- tein sequences encoded by the L. rliamnosus GG dlt operon, i.e., those corresponding to DIlA protein (506 amino acids [aul), DltB protein (405 aa), DltC protein (81 aa), and DltD protein

(423

aa), were used for screening databases using BLAST (2, 3). Homologics of these

L

rhamnosus GG OIt protein sequences to the Dlt protein sequences of 1...

rhamno- sus

ATCC 7469 and 1...

plantarum

WCFSl are summarized in Table 3.

Finally, the tlltl) gene of L. rhumnosus GG which encodes a putative protein of

423

aa with a theoretical mass and pl of

47.9K7

kDa and

1).57,

respectively. was successfully knocked out by tile insertion u! the tetracycline cassette Irom

L. /11011- tarum

5U57. Tile correct insertion was continued by peR and Southern hybridization as described in Materials and Methods (data

not shown).

LTA purification and structure analysis. After purification, dillcrcnt

L'I'Avcnnruining

fractions were identified. Fur 1...

rhatnnosus

GG wild-type cells (Fig. 2A), fractions 41 to 49 were pooled, resulting in 14.Kl mg L'TA. For the dlt.I) mutant cells (Fig. 2B), the phosphate determination revealed fractions 41 to

4(1 (represented

hy peak

1)

and fractions

4H to 8H (repre- Sl.'llh:d hy

peak

2)

with ( ),) () mg

LTA

in peak

I

and 'J.2(i

mg

LTA ill peak 2, respectively . Endotoxin contamination of all I T.\ extractions showed lipopolysaccharide

conuuniruuion

he- low

0.1)

endotoxin units/mg LTA, Le.. less than

60

pg lipopoly-

saccharide

per

mg

LTA.

TABLE J. Homology analysis of the translated

l..rhumnosus

GCi Dlt proteins

(;:;,Homology(Gcnlsankuccession no.) Strain

DltA DltB

one

DIID

/..,.11'111111(1'\/1'\ATCC

7469

I..plunuuum WCFS I

1)1)(AA r0l)20 I)

62 (NP

_71'15546)

95

(AArOl)202) fi4

(NP_78554h)

lOO

(AArOl}2llJ)

h4 (NP_785544)

99 (A/\ Fm204 )

51 (NP_7S55·B)

I~ Jll

3599

dltl) mutant LGG wild type

R

=

^D-Ala,OH

R =OH HO

l:- L ^{o t _{· o-~tl}

^{A 0}

^"OH~ _":\0 ^{~O ~} _{t:Jl: :}

HO~~O

OH

FIG. 3. Schematic representation of the structure of the LTA from the L. rhamnosus GG wild type determined from NMR spectroscopy analysis. The average numberofre pe a ting units in the polyglycerophosphate backbone, the average percentageof substitution, and the averag e chain length of the fatty acids in the membrane anchor were calculated directly from the integrals of the proton spectra. RI and R²•

Iauy

acid chains.n-Alaester substitution of the wild type (R ;::D-Ala ,OH) is replaced

by

OH groups in the dltI) mutan t.

wild type , in the dltD mutan t one double bond per fatty acid was present (data not shown).

Increase in cell length and defects in septum formation assessed by transmission electron microscopy. Comparison of transmission electron micrographs of the L. rhamnosus GG wild-type and dltD mutant cells showed an increase in the length of the mutant of about 2A-fold in comparison to the

parental strain results (Fig. 4A). Additionally, morphological alterations at the level of the septum and defective cell se pa- ration in the dLtD mutan t were observed (Fig. 4B).

Analysis of autolysis. As shown in Fig. 5, the dltl) mutant lyses to a greater extent than the parental strain . An OD!>(l() decrease of 0.2 units fo r the dltl) mutant corresponding to

_lOH

CFU/ml was observed during the first hour or incuba-

(A) L. rhamnosus GG wild type dUD mutant

(B) L. rhamnosus GC wild type dltD mutant

FIG. 4. Effect of D-alanylation on cell length andsepturn formation.

(A)

Ultrastructural analysis of thecell morphology of L. rluunnosus

l

iej wild-type and dltD mutant cells grown overnight in

MRS

medium and visualized with transmission electron microscopyat 80 kV.Thece ll length ofthe complemented strain wasresto re d to wild-type values. (B)Defect inseptum morphology.The arrowshows the altered cell wall around the

seprum

in the"!ID mutant.

+LGG IVIO.Ol'~·"

oLGGwl0.D15a,~

- CI'.1FG!·1540(].D1!)~""

2 1.6

.

c E 1.2

.

0

...

0LO

ci

^0.8

0 .. ..,

0.4

- 0 -

0 10 20

30 40 50 60 70 80 90 100

Time (hours)

FIG. 7. Effe ct of SOS on the grow

th of theL. rhuntnosus(j(j

wild type and the

d/ID

mut

an t.

Valu

es ob tained for lhl' gro wth or

the

1..

rlutmnosus

GO wild

type (cross

and lines) versu

s

that or the

dltl)

mutant (CMPG5540) (triangles and squares) in medium containing

eithe r (UlI% o

r

().()I5'~~!

SDS represent the

ave rage results orIhIL'l'

inde pe nde ntru ns.

J60()

2 1.6₁

r.: ^{- - - - -}

E

-

e _1.2

-

0

- - ^-

0 t.O 0

0.8

0

0.4 ---LGGwt

..--.- CMffi5540 0

0 2 3 4 5 6

Time (hours)

Fl G, 5.

Effects of the

mutati

on

of the

d/ID

gene on autolysis. Au- tolysis of the L

rliamnosu s

GG wild type (squares) and the

d/ID

mutant (CMPG5540) (circles) was followed in time by incubation of

the cells

with

Cl

lysis inducer

(T rito n X-lOO).

The value s reported in the

exper-

iment did not dillcr by more than 5

%.

Therefore, single datum points are pre

sented

in the figure without standard deviation bars

.

ODc,

o(l

values co mparable to

wild-typevalueswereobtained for the comple- me ntedstrain (CMPG55 41).

120 - . - -- - - -- - --- - - -

o

of L. rhumnosits GG and the dltl) mutan t to

grow

ill medium containing low concentrations of a strong

anionic

dctc

rgc

lit (SDS) was evaluated. In this case, the dltl) mutant was shown to be less affect ed by the action

of

SDS than the parent

a l

strain, reaching a higher optical density in the stationary phase (Fig.

7). The sen sitivity of the dltl) mutant to the tested catio nic peptides ancl SDS was restored to wild-type levels when the mutant was complemented with the dltl) gene (d ata not shown) .

Sensitivity of L.

rhamnosus

GG wild type and

dUD

mutant cells to human beta-defensins. Wt.: investigated the cllecr o

f

hBDJ and hBD2 on L.

r!W/llIIOSII.I'

GCi wild-type a nd dltl) mutant cells (Fig. 8). While hBDl showed no cflect on L.

rhamnosus OG wild-type or dltD mutant cells after ::I h of incuba tio n, the cells were sen sitive to hBD2. The dltl ) mutunt was on ave rage -SOC;0 more sensitive to cationic hBD 2 than the parental strain (Fig. 8).

Survival in simulated gastric juice. The dltl) mutant showed a strong reduction in the capacity to survive the in vitro gastric juice challenge, resulting in complete loss of viability after ::Ill min of incubation (Fig. 9) . Add itionally . the capacity to initia te

LGG

wt

CMPG5540

FIG . H

. Sen

sit ivity

of the

f..rhuntnosus

(j

G

\....·ild type

(wt )and the

.un: ^mu

^tant

to hUIll'an bet u-de len sin s. Bacte rial Vi ;ll;ilitv

Ill'L rliam-

1/0 .\/1.\ GG wild- type: a

nd dltl) mutan t eL'11s alter

~^Itof con tuct with

.1 ug/ml hBD I (stippled white ba

rs) o

r hBD 2 (stippled gray bar

s)

was determined

by

plate counting. The values represent

peln~lltagt:s^(If

surviva

l

(11' the

initia l numbers orI..rluunnosus

cells.

' i l l

hu

manhl:I ~I'

de

fensins

we re added to the controls (bl

ack bars) ,

Th

e e

rror

hars indica testa nda rd deviations of the results

of three

inde pe nde nt mea- sureme nts.

.."'1

70 80 60

o Crv1PG55401.5pghnl ... LGGwl 1.511C1.111~

--_..._-...__..._...

,

~~~:: 1

50

.r

10 2.5

. /7 ^..

. ^..

. .

. ^..

1

J "

I :

! "

0.5

! . "

' ..J.,

· .L ·

o~~. o

20

tion ill comp arison

to

the OD/lOO value obtained for the wild type.

Effects of the cationic peptides nisin and polymyxin Band the anionic detergent SDS on the growth of L. rhamnosus GG wild-type and dltl) mutant cells. The capacity of L. rhamnosu s GG and the

dltl)

mutant cells to grow in medium supple- mented with the cationic peptides nisin and polymyxin B was evaluated. In general, a minimum of -ID h of lag-phase retar- dation was observed for the mutant grown in the presence of nisin at

differen t

conc entration s

compared to

the wild-type

result s.However, the final

cell

densities reached were sim i la r

for the two strains (Fi g. ( i). For polymyxin B, the dltl) mutant was found to have -·-lO h of lag-phase retardation even with a concent ration of polymyxin B WO times lower than tha t used for the wild type. Again, no differences were observed regard- ing final cell density (dat a not shown). Additionally, the ability

30 40

Time (hours)

FIG. (1. EIl'L't: 1

llf tile cationic pep tide nisin on the

growthofl..

rluunnusu »(j(j wild- type

and

dltl) mu ta nt ce lls. Nil

differenc es in

grow th inr

vl RS medium between the

L.rhumnosusGG

wild type

a~ld tll<.' .lhl)mutant wcr« observed.Therefore.average growth values till'

hptll',Ird il" ;lll'Il'I, ll· ...

c ntcd

by'l[uilrl'~

(contrul} . Value» obtained for lite grllwth uf the

I..rltumnosus

GCi wild type ( triangles)

vers us

th

atIll'

(he

~(IIt1) ^muumr (C

MPCi5540)

(circles) in medium containing

1.5 lI!!

iml

nisin represent the average results of three

inde pe nde nt runs.

El',.

e:

§

o

c:i

3llU I

73

, . - - - r

72

'#-

71 co 70

'ID

69 .c"0 68·

(\l

.~ ⁶⁷

ro

66 iii 65

ex:

64

63 · ! - - - -

LGG

wt

CMPG5540

FIG. 10. Adhesion

of L. rhamnosus

GG and the dltl) mutant

to

human intestinal epithelial cells. The

levelsof bacterin _il1iti~dl~addl'd

to the Caco-2 cells were set

to IOO'i"...

Data shown are

re

lat ive

10

this percentage. Similar percentages of adhesion were observed lor

I

he HT-29 and Caco-2 cell lines. Graphs show the values obtaine

d for

Caco-2 cells only.

80

100

--CMFG5540

20

20

80

120 T·-·-- --·--·--·--····.··-···--·..- - -··--·-·- -·- - ·..-·-··..-._..._--.

I

--.10-LGG

wt

40 60

Time (minutes)

FIG. 9. Comparison

of

survival of

L.rhamnasus

GG wild-type (wt) and

(U,D

mutant cells in simulated gastric juice. Percentages of survival

of L. rhamnosus

GG (triangles) compared to the

dltD

mutant (CMPG5540) (squares) were calculated

by

comparing the cell num- bers before and after addition to simulated gastric juice

at 30-min

intervals. Recovery of acid toler

ance

was restored to wild-type values in the complemented strain.

'Co (jj

<ll Cl -0

2

<ll

"5

^Q,l 60 E.~

'(ij .~ 40

.s

growth under conditions of low pH (3.5, 4.0, and 4.5) was evaluated. Results confirmed that the acid tolerance of the dltD mutant was strongly diminished compared to that of the wild type (data not shown) .

In vitro adhesion to Caco-2 and HT-29 human intestinal epithelial cell lines and biofilm formation. No significant dif- fer ences in adhesion to human intestinal epithelial cells (Fig.

la) and capacity to form biofilm in vitro were observed be- tween the wild type and the dltl) mutant (data not shown).

Cytokine induction by HT-29 cells and PBMCs after

L.

rhamnosus GG and dltD challenge. An increase in cytokine induction in HT-29 was observed after either bacterial chal- lenge (L. rhamnosus 00 wild type, dltl) mutant, and S. enterica serovar Typhirnurium SL1344) or cytokinc stimulation (IL-lf3) compared to the values obtained using the negative control (data not shown). However, no significant differences in cyto- kine induction were observed for the mutant and the parental strain challenge (Fig. 11

A).

As expected, the proinflarnmatory cytokines tested (lL-8 and tumor necrosis fac to r alpha [TNF- a.]) were highly induced by th e positive controls (IL-lf3 and S.

en/erica serovar Typhimurium SL1344) in comparison to the low levels detected for the negative control (medium) and for L.

rhamnosus

GG or the dlil) mutant. Lack of o-Ala residues in the LTA of the

L.

rhamnosus GO dltl) mutant did not result in significant differences in the levels of IL-lO, IL-12, gamma interferon, and TNF-ex released from PBMC in comparison to the wild-type strain results (Fig. lIB). The gram-negative con- trol strain (E.

coli)

induced more lL- 10 than the gram-positive bacteria

(L.

rhamnosus GG and S . gordoniiy. as previously reported (25) (data not shown).

DISCUSSION

Inactivation of the dltl) gene in L. rhamnosus GG has a strong impact on LTA composition, resulting in a complete absence of D-alanyl ester content. This is in contrast to other lactic acid bacteria so far a nalyzed with respect to the dlt operon. A reduct ion in o-alanylation of LTA was previously reported

1'01'

Lactococcus lactis (fivefold lower for the dlt I) mutant) and Lactobacillus plantarum NCIMB8t-:26 (8- to 40-

fold lower for the dltli mutant) (45,55). A complete absence or D-alanylation of LT A has been reported for the dlt,« mutants or

a number of gram-positive pathogens

(I, 18,

35. 46,

47 ).

Recently, the func tion of the n-Ala ester substitution in LTA has been the subject of investigation for several genera or gm m-positive bacteria. Previous results with o-Alu ester-defi- cient mutants have shown changes in cell morphology. in so me cases associated with defective cell separation

(41).

For

1...

rhamnosus GG, electron micrographs showed an increase in cell length for the {UrD mutant compared to the wild type. This observation can probably be related to defects in the septal region. For another L. rhamnosus strain (the ATCC 7469 dltl) mutant), an increase in cell length was also reported compared to parental strain results. However, in this case no obvious alteration at the septum was reported (13). In a recent publi- cation, elongated

L.

plantatum

NCIMB8826

cells were ob - served as a consequence of the mutation in the dltb gene (45) . Taking all of these data together, it is clear that for all the bacteria studied thus far, n-alanylation of LTA plays

all

im- portant role in determining cell shape and cell scpuuion

In addition to the role of n-alanylation in determining mor- phology, this process also allows gram-positive bacteria to modulate surface charge. For exa mp le, LTA app ears to pla y a crucial role in the control of autolysin activity (55), and n-Alu ester content seems to determine the number or anionic sites on LTA for autolysin binding (58) . In accordance with these data, it was expected that the

L.

rhatnnosus GG dltl) mutant would show an increased rate of autolysis in the presence of Triton X-lOO as a consequence of the complete ubscnce Ill' n-Ala es te r residues in the LTA. Similar results have been reported for the

L.

lactis MG 1363 dltl) mutant (55). In addi- tion to the role i n autolysis. autolysins arc also involved in l·I.:11 division

unci

separation (32) . Therefore. the

ohscrv.uiou

,If elongated L.

rhamnosus

GG dltl) mutant cells having dl'kclS in septum formation and showing increased uutoly six is sugg.l.'stL:d to be the result or changes in electromechanical propert ies of the cell wall.

A second clear example of the role of n-alanylation in the modification of the surface charge is the correlation between

I)-AUt

ester content and the action of cationic antimicrobial peptides (34 , 40). For this reason, the capacity or ^L.

r!IiIIllIl(}SlIS

GG wild-type and dltI) mutant cells tu initiate growtl:

ill

till'

36(J2

(A) H1'-29 cell line cytokine induction

700 .- - --..-_.._ _- _..- --.- ---..- - -.--- - --- --- -..---.--- - ..- -.-..-- .

___a -WIi

LGGwt

CMPG5540

SL1344

IL-1 b

(B) PBIVIC cytokine induction

160!J , . . . - - - , 1200

800 400

• LGGwl ClCMPG5MO

SO()()()

.10000·

e

^30DOO

~ ₂₀₀₀₀

ioooo

o·

IL·12

IL·10 _{TNFa INFy}

FIG. 1

I.

Cytokinc response of human HT-29 intestinal cells and

PBMe

to stimulation with L. rhamnosus GG and dltl) mutant. (A) Bars represent the averages of triplicate results of three independent reverse transcriptase peR experiments. Black bars, IL-H; stippled white bars, IL-15;

stippled light gray' bars, transforming growth factor 1'; stippled dark gray bars, TNF-n.

(B)

Results represent the cytokine responses of at least six individual donors as determined by enzyme-linked imrnunosorbent assays. Black bars,

L.rhamnosus GO wild type; stippled white bars, dltl) mutant.

presence of two cationic peptides was evaluated. As expected, a

significant

difference between the dltl) mutant

and

the pa- rental strain was observed. On the other hand, it can be spec- ulated that negatively charged compounds might be repelled to a greater extent in bacteria with a lower degree of n-alanyla- tion.

Our

results indeed show that inactivation of the dltl) gene in

L.rhumnosus

GG increases the capacity of the bacterium to grow in a medium containing SI)S, a potent anionic detergent.

Human bcta-delensins are cationic host dclense pep tides expressed

by

epithelial cells.

It

has been reported that consti- tut ivcly expressed hHD 1 can mediate epithelial interactions with the commcnsal flora whereas I1BD2 may participate in the host defcnsc response to enteric microbes that can breach the epithelial barrier (15), We have previously demonstrated that L. rhumnosus GC; is not sensitive to hBDI, whereas it is very sensitive to h13D2 (15). Consequently, considering these data together with the charge-mediated mode of action of this an- timicrohial peptide, it can be speculated that the increased sensitivity

ut

the L.

rhumnosus

GG

dltl)

mutant to 11BD2 is

likelv thl..'

result

ut

an

increased

net negative charge

of

the bactcnal ccll envelope as a consequence 01 the modification in the

II-Ala

content of the LTA (40).

In general, probiotic bacteria should survive gastric transit in order to confer beneficial effects

Oil

the host. A mutation

ill

the dltl) gene of L. rhamnosus

GG

causes a dramatic decrease in acid tolerance, This particular phenotype is interesting, taking into account that cell wall components, soluble factors. and genomic

DNA

from L.

rhanutosus GG

have been

found

tu have a strong immunostimulatory capacity (27, ::16). Conse- quently, after gastric transit,

the

dill) mutant

could

still exert immunological benefits as a result 01 release 01 cell wall com- ponents

and DNA.

After colonization of the gasuointestinal tract, in vivo biofilm formation

on

the intact intestinal mucosa represents for the host an additional part or the mucosal barrier

(31, 37,

44). In the present study we found no differences either in adhesion to the tested human intestinal epithelial cells

or

in biolilrn formation between L. rhamnosus GG wild-type and dltl) mutant cclis,

'The capacity of probiotic bacteria to stimulate or regulate the mucosal immune system and therefore maintain the gut immunological barrier has been widely studied

(:l,

22, 42, :'3 J.

Furthermore, the role of n-ulanylation of the LT/\. in the anti-

inflammatory properties of the probiotic strain L.

plunturum

NCIMB8826 was recently described

(12).

The

imrnunomodu-

lation of an L.

plantatum

dltb mutant was significantly different from that of the parental strain in the in vivo and in vitro model systems studied (IL-IO/IL-12 ratio,

1.1

for wild type and

160.8

for the clltB mutant) (22). In the case of L. rhamnosus GG, cytokine stimulation of human intestinal epithelial cells and peripheral blood mononuclear cells was not significantly al- tered by the lack of o-Ala ester substitution in the dltD mutant of L.

,./Wt1I1IOS11S

GG (IL-lO/IL-12 ratio,

51.09

for the wild type and 4.33 for the dlil) mutant), although the level or n-Ala content of the L. rhamnosus GG wild type is

1.8

times higher than that of the n-Ala content of the LTA from L. plantarum, These results are in agreement with the findings of B. Pot and

C.

Grangette that a dltD mutant of Lactococcus lactis MG1363 did not yield a considerable increase in anti-inflammatory po- tential either (unpublished data).

In addition to the differences in LTA o-Ala substitution results observed, the cell wall of L. rhamnosus contains only one type of TA, LTA (29), in contrast to L.

plantarum,

which contains two types of TAs, LTA and WTA. Although it has been reported that WTA from L.

plantarum

carries n-Ala and glucose residues in a strain-dependent ratio (40), no reports about the substitution of WTA in the particular strain NCIMB8826 have been published. However, it can be specu- lated that the degree of n-alanylarion of WTA from L. plan-

tarum

NCIMB8826 is affected by the mutation in the £lItE gene, since the D-AJa ester substitucnts of WTA are derived from those of LTA (40). Additionally, the LTA from the L. planta- nun NCIMB8826 {lltE mutant was shown to contain a large amount of glucose substitutions, whereas glucose substitutions were nearly undetectable in the LTA from the NCIMB8826 wild type (22). These differences in the LTA (and WTA) of L.

plantarum and L. rhatnnosus GG strains might well offer an explanation for the different immunological responses in chal- lenges with the two different species.

Moreover, other structural features of the LTA need to be considered as key factors for immune stimulation. For in- stance, the LTA glycolipid anchor and the length of the Gro-P backbone, as they have been reported to have immunostimu- latory potential in other gram-positive bacteria (14,38), should be taken into account. Interestingly, the fatty acid chains of the glycolipid anchor of the L. rharnnosus GG dltl) mutant showed an average increase in length of two carbon atoms compared to the length of fatty acid chains of the lipid anchor in the wild type. Additionally, the polyglycerophosphate chains, contain- ing an average of

50

Gro-P residues in the L. rhamnosus GG wild type, were reduced to averages of 29 Gro-P and 7 Gro-P residues for the major and minor peaks 01' the dltD mutant, respectively. In contrast, the Gro-P backbone of the

L.

plan-

tarum

cl/lE mutant increased threefold in length compared to the backbone in the wild type. It can therefore be concluded from both studies that altering the n-Ala substitution of LTA, by either

^Cl

dltl) CL. rhamnosus GG) or a dltli (L.

plantarums

mutation. also affects the other building blocks (glycolipid an- chor and/or polyglycerophosphate chains) of LTA. This is of interest and needs to be taken into account for future studies of the immunomodularory properties of LTA,

Conclusively, our study showed the importance of DltD of the probiotic strain Lactobacillus rlwnmoslls GO in the biosyn- thesis of the LTA. Lack or D-alanylation of the LTA all'ecls

^Cl

number of cell morphology and surface properties but does not

affect important probiotic characteristics, including immune- modulation properties. On the other hand, it is very likely that the dltD mutant of L. rhamnosus GO will more easily lyse after administration than will the wild type. Therefore, as a fol- low-up study it will be interesting to compare

imrnunomodu-

lation effects in vivo.

ACKNOWLEDGMENTS

1\11. Perea Velez holds a Ph.D .

grant from ihe Irue rtncultv

Ct\lIrlt'il

for Development Cooperation of

K. U.

Leuven (IRO-16JU2). S. De Keersrnaecker is a postdoctoral research associate of the Belgian Fund for Scientific Research (FWO-Vlaanderen)

.

Additionally. this work was financially supported by the K. U. Leuven Research Council (100/

03/(05) and the Flemish Institute for the Promotion of Innovation by Science

and Technology (IWT- Vlaanderen,

Brussel

s, Belgium)

through projects STWW-OOOJ 62 and

SBO-040073.

We thank Chantal Mathieu and

Lutgart

Overbergh for providing

LIS

with primers and probes and necessary equipment for the reverse transcriptase

PCR

analysis at the Laboratory for Experimental Medi- cine and Endocrinology

(LEGENDO,

Laboratory for Experimental Transplantation, University Hospital Gasthuisberg, K. U. Leuven), P.

Augustijns for providing the Caco-Z cells used in this study. V. Dennin for help with the PBMC analysis, D. Valckx and W. Cockx for technical assistance, and

M.

Fauvurr for his valuable graphical ussistuncc.

REFERENCES

I. Abnchin,Eo,C. Poyart, E. Pellcgrini, E. Milohunic, F.Fiedler, 1'. Berchc, and 1'.Trieu-Cuot. 2002. Formation of u-ulunyl-lipoteichoic acidls requiredrill'

adhesion and virulence ofListeria monocytogenes.Mill. Micrubiol.43:1-·14.

2.Altschul, S. F., W. Gish, W. Miller, E. W. Myers, and D..J.Liprnan. I(J90. Basic local alignment search tool.J.Mol. Biol.21S:403-41lJ.

3.Altschul, S. F., T. L. Madden, A. A. Sehaffer, J. H. Zhung, Z. Zhang, W.

Miller, and O•.I.Llpman, 1997. Gapped BLAST and PSJ·BLAST: Clnew generation of protein database search programs. Nucleic Acids Res, 25:

3389-3402.

4. Alvnrez-Olmos, M. 1., and R. A. Oberhelman. 20Dl.Prohiotic agents and infectious diseases:a modern perspective011 a traditional the rapy.Clin. Infect.Dis. 32:1567-l576.

5. Baken, K. A.,.I.Ezendam, E. R. Gremrner, A. de Klerk,,J.L.Pennings, B.

Matthee, A. A. Peijnenburg, nnd H. van Loveren. 200().Evaluation of im- munornodularion lw Lactobacilluscasei Shirutu:immune Iuncuou, .uuuim- munity and gene expression.lnt../. Food Microhiol.1I2:8·-/K.

6. Bierbuum, G., and H. G. Sahl, /1)S:;.JI1du~'1ion\IfuuIl1lysis\II'Sl:ljlhl11Il'lIl'L' j

by the basic peptide antibiotics Pep 5 and nisinand their influenceon the activity of autolytic enzymes. Arch. Microhiol. 141:249-·254.

7. Boyd, D.A., D.G. Cvitkovltch, A. S. Blciwcis, M. Y. Kiriukhin, D.V.

Debabov, F.C. Neuhaus, and 1. R. Hamilton. 200D. Defectsin u-alanyl- lipoteichoic acid synthesis inStreptococcus mutans results in acid sensuivity.

J. Bacterial. 182:6055-6065 .

8. Cao, M.• and J. D. Helmann, 2004.TheBacillus subtilisextrucytuplnsrnic- functionlr xfactor regulates modification ofthe cell envelope and resistance to

cationic

untimicrohial peptides, J, Bacteriol.186:1136-·11411.

9.Clernans, D. L., P. E. Knlenhrnnder, D. V. Debuhnv, Q. Zhllng, R. D.

Lunsford, H. Sakone, C..I.Whittaker, M. P. Heaton,and F.C.Neuhaus, 1999. Insertionul inactivation

or

genes responsible for the n-ulanvlatiun of llporeichoic acid inStreptococcus gordoniiDU (Challis) affects inrrugcncric coaggregations. Infect. Immun.67:2464-2474.

10.Corcoran,

u

M., C. Stanton, G. F. Fitzgeruld. unci R. P. Ross.2[JlJ:'.Survival of prohiotic lactobacilli in acidic environments is enhanced in th<.'IH<:S~'IH':c"I metubolizubleSUg'HS.Appl. Environ. Microhiul.71:~(JW-· _'(J(i7

11. Danlelsen, M. 20(J2.ChuracterizationIll'lllCtctrucychnc n.:siSl<lIlc·c·1'101:'111111 pMD5057fromI.((ClOlJIlCi!tI/.I'pluntutum50S7 reveals a cnmpositc structure Plasmid 4H:lJK-1 lB.

12.Dehabov, D. V., M. P. Heaton, Q. Zhang, K. D. StCWUI·t, R.11.Lumbulot, und F. C. Neuhaus. 19%.The n-alunyl carrier protein in Luctuhacillu» casei:

cloning. scquencing, and expression ofcl/le. .J.Bucu..riol,17H:JllhlJ·-.IK7h.

U. Debabov, D. V., M. Y. Kiriukhin, nnd F. C. Neuhaus, :W(J(J. Biosynthesis ut Iipoteichoic add

in

Lactobacillus tltamnosus:

mic of

Dltl)illn-alunylatinn ..I, Bacteriol. 182:2H55-2864.

14.Deiningcr, S., A. Stadelmaier, S. von Aulock, S. Moruth, H..ItSchmldr, nnd T. Hartung. 2003.Definitionof structural prerequisites for lipotcichoic acid- inducible cytukine inductionbysynthetic derivatives..J.hnmuuol. 1711:413-1· 4138.

15.Dc Kecrsmaccker, S.

c.,

K. Brackcn, T. L. Verhocvcn,1\1.Perca Veil'/.. S.

Lcbcer,.I. Vanderlc)'dell, !lnd P. Bols. 2006.Flnwc.::Yllll1l<.'lrictC...till!!otgrcc'lI fluorescent protein-tagged!.m:tohaciltll.l·,."1111I'10.1'/1.1(i(jI'mrl'l>pOn,eIIIdc·

felll>ins.Appl. Environ.Microbiol. 72:4lJ2:l--I9JO.

3604

16. Dc Keersmnecker,S.

c.,

C.Varszegl.N. van Boxel,L.W. Hahcl, K. Metzger, R.Duniels,K.Marchal,D.De Vos,and.I.Vanderleydcn.21J05.Chemical synthesisof (S)-4.5-d ihydroxy-2.3·pcnlanedio[\e. ahucter iulsignal molecu le precursor,and validariun of itsactivityinSalmo n cll«tvphiuu uiu ru..I.Biol.

Che m,2111l:195(13-·1951lS.

17. DclcOIII" .I.,T. Fcruin, M.Dcghorain, EoPalumbo, nnd P.1Iols. 1999. The biosynthesis and functionality of the cell-wall of lactic acid bacteria. Antonie Leeuwe nhoek76:15~)-1H4.

IH. Fnbretti, F., C.Theilacker, L. Hnldassarri, Z. Kaczynski,A. Kropec, O.

Hoist.andJ.Huebner.2006. Alanine esters ofentcrucoccallipoteicho ic acid playit role in hiolilmIormat io nand resistance to antimicrobial peptidcs.

Infect. Imrnun.74:4164-4171.

IlJ. Fischcr, W., P. Rosel, and H. U. Koch. 1981. Effect of alanine ester substi- tution and other structural features of lipoteichoic acids on their inhibitory activity against aurolysins01'Staphylococcus aureus..I.Bacterial. 146:467- 47:'i.

2(). Food and Agriculture Organization of the United Nations and World Heulth Orgunlzauon. 200I.EvaluationIll'health and nutritional propertie sof pow- der milk andlive lact icacidbacteria.Food and Agriculture Organizationof

IIll' United 1\alilllls and World Health Organization expert consultation reportFoodand Agric utrureOrganiznti on01'the United Nation s andWorld I leulthUrganization,Rome,ltuly.

21. Gorbach,S.L.:WOO.Prubioticsandgustrolntestinnlhealth.Am. J.Gastro- crucrol.95:S2-S4.

22. Grangette, C.• S. Nutten,E.Pulumbo, S. Morath,C.Hennnnn,.I.Dewulf, B.

Pot. T. Hurtung, P. Hots, nnd A. Mercenier. 2.00S. Enhanceduntiinflamma - tory capacity of a Luctoba cillusplantarum mutant synthesizingmodified rcicholc acids.1'I'IlC.Natl,Acud. Sci. USA 102:lD.12.1-10326 .

23. Gross, M., S.EoCramtnn, F. Gotz, and A. Peschel,200 I.Key role ofteichoic acidnetcharge inStaphylococcus aureuscolnnizatinnofartificial surfaces.

Infect.lmmun. l>9:3423-342b.

24. 1IIIIlC()l'k, I.C.2001.Teichoicacids of Gram-positive bacteria, p. 79-92.III M. Sussrnan (ed.),Molecular medicalmicrobiology.Academic Press, San Diego,CA

25.lIcssle. C.,L.A. Hansnn, and A.EoWold. 1999.Lactobacilli from human gastruintesunal mucosa arc strong stimulators of (L-12 prodllL·tion. Clin.

Exp. Illlrnuno!. 116:276-2H2.

2CJ. Hlliseth, S. K., und B, A. Stocker. 1981. Aromatic·dependentSl//lIlolle/(a lyphimuriu/Ilarc non-virulenl and elTeclivc as live vaccincs.Nalllre 291:2.18-·

239.

27,lIiev. 1.D.,H.KitazllwlI.T.Shimosato, S.Katoh, H.Morita, F.He, M.

I1l1soda, and T. SllitO.lOllS.Strong il1lJlJunostimulationinmurinc immune cells by l.ocw!Jacillu.\· r1Wtnllo.ms GO DNA containing novel oligode- ()xynucleotide pattern. Cell Microbio!. 7:403-414.

2X.•Iosson, K.,T.Schcirlinck,F.Michiels,C.Pllltlceuw,P.Stanssens. H. .loos, P. Dhaese, M. Zabeau, und.l. Muhillon. 19l:l9. Characterization of a Gram- positivc broad-host-rallgeplasmid isolatedfroml.acwbaci//II.I'llilgal'dii.Plus- mid 21:9-20.

2LJ.Kelemcn, 1\1. V., and.l. BlIddilcy.Il)(jl.Structure of theintracd lular glycerol tcichllk: add fl'llml.o('w!>acilllI.l·caseiATCC 746l).Biochem .

.r,

80:246-254.

~II, Kleen~lH'lclll. M...I.Bocl<hol'st, R. Van Knmenhurg. D. MlJlcnallr. O. P.

KuipCl's, R. Lecl', R. TlIl'chini,S. A.Peters. H.M.SlIndbrink, M.\\1.Ficrs, W.Stickcmll.R. M.Klein Lankhllrsl,1'. A. Broil, S. M.110Jl'el',M. N. Niel'op Gnlllt. R. Kel·khoven.M.Dc Vi'ics.

n.

Ursing,W. M.lie VIIS.and It.l. Siczen.

2()(J3.C\lIl1pkll'g,~nllllll'sequcllcl' III!.1/('(()lltIcillu.l' plalllllrt/1/1WCFS I. Proc.

r-.:a ll. '\~·;ld,S~'i,['SA Ill(l: Il) l)l )..,I1)1)5,

JI.Kleessen,

n.,

B.Sykunl, H•.I.Zunft. und1\1.BlllUt. ILJl)7.rmects ofinulin and laCLOse⁰¹¹fccal micruflora, minubial aClivity, and bowel Iwbit in ddcrly wnstipated persons.Am.J.Clin.NutI'. l>S: ]397-1402.

32. Koch. A.L.200!. Autolysis control hypotheses for tolerance to wall antibi·

otics. Antimicrob. Agl:l1ts Chcmnlht:r. 45:2671-2675.

.\3. Knprh'njak,T.,V. Mlnkllr. L.Swnnsou,H.Foul'llicr. A.Pcschel, null.1.P.

\\'eiss. 2ll()ll. Cation-inuuced transcriptional regulation of the,illoperon of Swphylococc/(.I'tlW·,'I/,I'•./. Bactcriol. 1811:3622.-3630.

34, KOI'lIl·s. M., A. lIalfmunn,I.l"cdtkc,M. lIcintz, A. Pcschel, W. Vollmer. R.

II1Ikenbcck, and R.BI'udmcl',20ll(,. A functionaltilloperon,encoding pm·

h'ill~ H:qllil~d 101illl'lIrp"liltiun ul'u-alanin~ in I~'ichllk ileitis in gra m- pnsitivc hacteria, wlllers resistance to ~'i1tinnicantimicrobial pcptides ill St/'('I,loc:occ/lsIJIIl'1I1ll0/licu'.1.Baclcrin!'J1l!l:5797-..5H05.

.15.Kristillll. S. A.,V. J)utlu,C. Wcidcllnlllicl·.1{,Kllllsul. 1. Fl.'dtkc, A. Peschel.

R. L. Gllllu. und V. Nizet.J(J(J5. II-AlanylilLion,If[eichnic acids prolllotes group i\S!rl!!J/1lt'O" " II,I'antimicrobialpeptide n:sistance,ll~utl'llphilsurvival.

amI epithelial~'ellinvasioll.J.Bacterin !.J87:h7IlJ-(,72S.

Jh. L,'bccr, S.,S. C. Dc Kcersllllleckcr, T. L. Vel'hocvell. A.A. Fudda, K. 1\:1111'-

dull.and.I.Vanderleyden.2tHJ7.Functionalanalysisoflu.xSinthe pro h io tic stra inLuctobacillusrhamuosusG(j reveals acentra lmet uhulicrule impor- tum forgrowthand biolilmlurmatiun..I.Bacteri o l.ISI):KCJtl..·H7 I, 37. Mucfarlane, G. '1'.,EoFurrie, and S.Mncfurlnne. 2(J(J 4,Hal'l~' ri a lmilieuand

mucosalbacteriain ulcerativecolitis.Nova r tis Found.Syrup,263:57..·h4.

3K.Moruth, S., A.Gcyer, andT. Hartung,20l1l.Structure-fu nctionrchuion ship of cytokine inductionbylipoteichoicadd fromStaphylococcusaureus.J.Exp, Med.193:393-397.

39. Murath, S.,A.Stndetrnaier, A.Geyer, H.H.Schmidt,and T. Hurtling.2002, Syntheticlipo teic ho ic acidfromStaphylococcusaurcusisapotent stimulusof cytokine release..I.Exp.Mcd,195:1635-1640.

40.Neuhaus, F. C., and .I.Buddiley. 2003. A continuum of anionic charge: str uctures and functions of o-alanyl-teichoic acids in gram-positivebacteria.

Microbial. Mol.Biol. Rev.67:686-723.

41. Ntumere, A. S., D.J.Taron, andF. C.Neuhaus.1987.Assembly ofn-alanyl- lipotelchnicacid inLactobacilluscasei:rnurunts deficientin theu-ul anylester conte nt01 this amphiphlle. J. Bact erio l.169:1702-1711.

42. O'Hnra,A.M., P. O'Regun, A.Fanning,C. O'Muhony,.I.Macsharry, A.

Lyons,.I.Bienenstock, L.O'Mahony, and I;'.Shanahan.2lJ06.Functional modulntion of human intestinal epithelialcel lresponses byUijidll!Jf/l'/"/'iIl/1/

infaut isand Luctobueillnssalivarius.lrnmun ology118:202-21;'i.

43.Overhergh, L., A. Glulicrtl,D.Vulc kx R. Derallunne, It lluuillon,andC.

Mathieu. 2lJ03.The use ofreal-time reverse trunscrlptase peR for the quantillcutionofcytokinegene expression.J.Biomol.Tech .14:33-43.

44.Pulestrant,D.,Z. E. Holzknecht,B.H.CII1Iins,W. Parker, S. E. Miller, and R. R. Bollingcl·. 2004. Microbialbiofllmsinthe gut: visualizat ionby electr on microscopyand by acridine orange staining.Ultrastruct, Path ol,28:23-27.

45. Palumbo, E., M. Dcghoraln, P. S. Coccnncelll, M. Kleerebezem, A. Geyer, T.

Hartung, S,Morath, andI). Hols, 2006. n-Alanylester depletion oftetchoic acidsin Lactobacillus pluntarumresult s inamajormodiflcat ionofIipo- teichoic acid compositionand cell wall perforations at theseptum mediated by the AC1l12 autolysin.J. Bactcriol.1811:3709-37 15.

46. Peschcl,A., M. oue,R. W..Il1ck,H.KalbachcJ',G• .lung,and F.Gntz. 19()LJ.

Inuctivation ofthedlt operuninS/aphylococcusaurcusconfersscnsitivityIu defensins,proregrins, and otherantimicro bial pep tides ..I.Biol, Chern. 274:

X405-·H410.

47,Pu.yurt,

c.,

E. Pcllegrini, M. Man:ellu, M. Baptistll. F. .Iauhcl·t. M. t'. Lamy.

und P. 'I'riell-Cuol. 2003. Attclllwted virulcnce ofSlrcpw('()l,:CIISagalunitle deficient inD-alanyl-lipotcichoic acidis duI.' tuanincreased susccplihilityto dd<:.:nsinsand phagocytic cells.Mol. Minohiul.49: IIdS·-ltl~5.

4H.Raychlludhuri,D.,and A, N. Chuttcrjec. IlJH5. US!;of rcsistant 1lI11t illll s III

study the interactionof TritollX-]()()withSlaphylol'Occusalln'us..I.Bacte- ria!. 164:1337-1349.

4\J.Snmbrook• ./.,E.F.Fritsch, aud T. Mnniutis. 1989. M()I~cularcloning: a lahoratory manual, 2nd cd.Cold Spring Harbor Laboratory,Cold Spring I-(arbor,NY.

50.Snugcr, F., S. Nickleu. and A.R. Coulsou. 1977. DNA sel.juenciug with chain-terminatinginhibitors. Proc. Nail. Acat.!.Sci. USA 74:5463-5467. 51. Schar-ZllmmaJ'ctti,P., !lud J. Ubbink.2003.·11lC cell wall of lacticad J

bacteria:surface constituents and mucromolccularconformatiol\s. Billphys.

J.85:4076-4092.

52. Schnitgel',H.,K. I)upcnberg,E.Gausc.R.Czok.T. Buchel" undH.Adalll.

19S1J. Chro ma w gTil p hie phosphathaltiger Metaholite dnes ll1enschlichen LeberpunklUts.Binchem.Z.332:167- lli5.

53.Shcil, B••.I. Mncsharry.1..O'ClIJllIghall,A.O'Rillrdlln,A.\'Vaters•.

1.

Mor-

\:1111••1.K. CulJins,L.O'Mahllny,and F. Shllnahan. 2110(),Rok ulillte rlc ukill (IL-1O) in probiolic-mcd iat...d immune modulation:an assessmc lltill wild- type alld IL-IO klllJl.:k-nul mice. Clill.Exp,Imlllullol. 144:27J..2:m,

54. Shcnvo()d, L., and M.D.G()rhach.19%.Thediscovery01'l.uc/"hal'illwGC;, Nutr. Today 31:2S-4S.

55. Steen, A., E. Palumbo. M. I)eghorain, P. S. Cocconeelli, .I. f)clcour, O.P.

KuiJlcrs,.1.Kok, G. Huist, and P. Hols. 2005. Autolysis ofl.actm·occ/l.\·I(/clis is increased upon o-ulaninc depletion of peptidoglycan and Iipoteichoic acids..J.Bacterio!'187: 114-124.

56, Tuo. Y.• K. A. Drabik, T. S. WU'ypll. M. \Y. Musch,.I.C.Alvcrd~'.O. Schnec- wind, E, B. Chung, lInd E. O. Pctl'Of.:W06. Sllluble factms frnml.a,'(o/laci!lll.l· GO activate rV1A1'Ks and indut.:e,'yloIHIJLe,·ti vc Ill'at shllck PlOll.'illsin in- testinal epithelial cells.Am ..I.Physiol. Cell Physiol.290:CIIlIH-CIlJ30.

57.Volkman, B.F., Q. Zhllng, D. V.Dchllb()~'.EoRh'era,G.Co Krcshcck,llnd F.Co Ncuhaus.~(llJ l. l3iusylll hesis

,. 1'

ll-alilnyl.lipOlCid,ok·ad d:lh~'teniary stru,·tllI'C of ap()-ll-alany l~'arrierprotein.Biochl:mist ry411:7%4..7' 172.

S~. Weckc•.I.,M.Pel'l'!-:o, llnd W. Fischer. 1')%.Il-/\I alllllclk'privilllllnul /tu · ci//I/.I·.I'1I/Hili.I'tt.:idllli... Hdds is withoutelk..., on (.'ellgl'!lw th,mdIIlIHf'hlllllg~'

butalkcls lh,'allh llyli...at:tivity.M icroh. Drug Rcsis t.2: 12J-12^l).