www.fems-microbiology.org

FEMS Microbiology Reviews 29 (2005) 435–463

New insights in the molecular biology and physiologyof Streptococcus thermophilus revealed by comparative genomics

Pascal Hols a,*, Frederic Hancy a, Laetitia Fontaine a, Benoıt Grossiord a,b,Deborah Prozzi a, Nathalie Leblond-Bourget c, Bernard Decaris c,

Alexander Bolotin d, Christine Delorme d, S. Dusko Ehrlich d, Eric Guedon d,Veronique Monnet e, Pierre Renault d, Michiel Kleerebezem f

a Unite de Genetique, Institut des Sciences de la Vie, Universite catholique de Louvain, 5 Place Croix du Sud, 1348 Louvain-La-Neuve, Belgiumb Laboratoire de Microbiologie et de Biochimie Appliquees, ENITA de Bordeaux, BP201, 1 cours du General de Gaulle, 33170 Gradignan, France

c Laboratoire de Genetique et Microbiologie, UMR INRA 1128, IFR 110, Faculte des Sciences-Universite Henri Poincare Nancy 1,

BP239, 54506 Vandoeuvre-les-Nancy, Franced Genetique Microbienne, Centre de Jouy en Josas, Institut National de la Recherche Agronomique, 78352 Jouy en Josas, France

e Biochimie et Structure des Proteines, Centre de Jouy en Josas, Institut National de la Recherche Agronomique, 78352 Jouy en Josas, Francef Wageningen Centre for Food Sciences, P.O. Box 557, 6700 AN Wageningen, The Netherlands

Received 27 March 2005; accepted 28 April 2005

First published online 28 August 2005

Abstract

Streptococcus thermophilus is a major dairy starter used for the manufacture of yoghurt and cheese. The access to three genome

sequences, comparative genomics and multilocus sequencing analyses suggests that this species recently emerged and is still under-

going a process of regressive evolution towards a specialised bacterium for growth in milk. Notably, S. thermophilus has maintained

a well-developed nitrogen metabolism whereas its sugar catabolism has been subjected to a high level of degeneracy due to a paucity

of carbon sources in milk. Furthermore, while pathogenic streptococci are recognised for a high capacity to expose proteins at their

cell surface in order to achieve cell adhesion or to escape the host immune system, S. thermophilus has nearly lost this unique feature

as well as many virulence-related functions. Although gene decay is obvious in S. thermophilus genome evolution, numerous small

genomic islands, which were probably acquired by horizontal gene transfer, comprise important industrial phenotypic traits such as

polysaccharide biosynthesis, bacteriocin production, restriction–modification systems or oxygen tolerance.

� 2005 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved.

Keywords: Streptococci; Metabolic pathways; Pseudogenes; Genome decay; Milk

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 436

2. Genome plasticity and phylogenic position of S. thermophilus among streptococci . . . . . . . . . . . . . . . . . . . . . . . . . . . 437

0168

doi:1

* C

E

2.1. Comparison of LMG18311 and CNRZ1061 S. thermophilus genomes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437

-6445/$22.00 � 2005 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved.

0.1016/j.femsre.2005.04.008

orresponding author. Tel.: +32 10 47 88 96; fax: +32 10 47 31 09.

-mail address: hols@gene.ucl.ac.be (P. Hols).

436 P. Hols et al. / FEMS Microbiology Reviews 29 (2005) 435–463

2.2. Phylogenic position of S. thermophilus among the salivarius group using MLST . . . . . . . . . . . . . . . . . . . . . . . . 438

2.3. In silico analysis of horizontal gene transfer events. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438

3. Biosynthesis of the cell envelope and associated cellular processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439

3.1. Biosynthesis and degradation of peptidoglycan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439

3.2. Teichoic acids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 441

3.3. Biosynthesis of polysaccharides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 441

3.4. Protein export pathways and secretome. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 442

3.5. Global view of membrane proteins and solute transport systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443

4. Sugar metabolism, central carbon pathways and regulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 444

4.1. Sugar utilisation and transport systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 444

4.2. Central metabolism and pyruvate dissipating pathways. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 446

4.3. Regulation of galactose and lactose metabolism. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 448

5. Nitrogen metabolism and regulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 449

5.1. Amino acids biosynthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 449

5.2. Amino acids catabolism. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 451

5.3. Proteolytic system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 451

5.4. Amino acid and peptide transport systems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 452

5.5. Nitrogen metabolism regulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 452

5.6. Urea metabolism and regulation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 452

6. Oxygen metabolism and oxidative stress. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453

6.1. Oxidative stress tolerance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453

6.2. Regulators of the S. thermophilus defence to oxidative stress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 454

7. Two-component signal transduction and quorum sensing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 454

7.1. Two-component regulatory systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455

7.2. Competence for natural transformation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455

7.3. Regulation of bacteriocin production; the blp locus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455

8. Concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457

Appendix A. Supplementary data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 458

References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 458

1. Introduction

Streptococcus thermophilus is of major importance for

the food industry since it is massively used for the man-

ufacture of dairy products (annual market of around 40

billion USD) and it is considered as the second most

important industrial dairy starter after Lactococcus

(Lc.) lactis [1–3]. This bacterium belongs to the groupof the thermophilic lactic acid bacteria and is tradition-

ally used in combination with Lactobacillus delbrueckii

subsp. bulgaricus (Lb. bulgaricus) or Lb. helveticus for

the manufacture of yogurt and so-called hard ‘‘cooked’’

cheeses (e.g., emmental, gruyere, grana), at a relatively

high process temperature (45 �C) [2,4]. S. thermophilus

is always used together with Lb. bulgaricus for yogurt

making, which led to development of a complex symbi-otic relationship (‘‘proto-cooperation’’) between these

two partners sharing the same ecological niche [3,4]. S.

thermophilus is also used alone or in combination with

lactobacilli for the production of mozzarella and ched-

dar cheeses [2].

S. thermophilus is closely related to Lc. lactis, but it is

even more closely related to other streptococcal species

comprising several deadly human pathogens (e.g., S.

pneumoniae, S. pyogenes, S. agalactiae), which cause for

example pneumonia, bacterial sepsis or meningitis [5,6].

S. thermophilus is also related to S. mutans, the most

important pathogen in tooth decay. Nevertheless, S. ther-

mophilus is a ‘‘generally recognised as safe’’ (GRAS) spe-

cies and over 1021 live cells are ingested annually by the

human population. Recently, the complete genome se-

quence of two yogurt strains (LMG 18311 andCNRZ1066) and a third incomplete genome sequence

(strain LMD9) were made publicly available ([7],

http://genome.jgi-psf.org/draft_microbes/strth/strth.home.

html). The comparison of the S. thermophilus genomes

with published genomes of streptococcal pathogens

[8–13] highlights its relatedness to pathogenic species

but also reveals that the most important determinants

for pathogenicity are either absent or present as pseudo-genes, unless they encode basic cellular functions [7]. This

reinforced our view that the massive consumption of this

bacterium by humans entails no health risk. Comparative

genomics also revealed that evolution has shaped the S.

thermophilus genome mainly through loss-of-function

events, even if lateral gene transfer played an important

role, revealing that the dairy streptococcus has followed

a evolutionary path divergent to that of pathogenic

P. Hols et al. / FEMS Microbiology Reviews 29 (2005) 435–463 437

species due to its adaptation to the rather narrow and

well-defined ecological niche, milk.

Here, a defined number of topics is reviewed based on

an integrated view of the in silico investigation of the S.

thermophilus genome sequences, their comparison to

closely related species, combined with our currentknowledge of the molecular biology and physiology of

this species and its (streptococcal) close relatives.

2. Genome plasticity and phylogenic position of

S. thermophilus among streptococci

Viridans streptococci include 26 species divided intofive groups, (i) the mutans group, (ii) the anginosus

group, (iii) the sanguinus group, (iv) the mitis group,

and (v) the salivarius group, which includes S. salivarius,

S. vestibularis, and S. thermophilus [14]. Except for S.

thermophilus, all species of viridans streptococci are

encountered as commensal in the oral and gastrointesti-

nal cavities and genital tracts of mammals. S. salivarius

and S. vestibularis have been isolated from human oralcavity and are possibly associated with human infections

[15]. The three species of the salivarius group are very

closely related as shown by the comparison of their

16S RNA and sodA genes [16,17]. During several years,

S. thermophilus was classified as a S. salivarius subspe-

cies before regaining a full species status based on

DNA–DNA reassociation experiments [18]. In this sec-

tion, we report recent advances based on comparativegenomics and multilocus sequence typing analyses that

provide novel insights in the phylogenetic position and

evolutionary history of S. thermophilus as a separate

species.

2.1. Comparison of LMG18311 and CNRZ1061

S. thermophilus genomes

S. thermophilusCNRZ1066 and LMG18311 contain a

single circular chromosome of 1.8 Mb, encoding approx-

imately 1900 CDSs (Table 1 and Supplementary Fig. 1

Table 1

General features of S. thermophilus CNRZ1066 and LMG18311

genomes

Features CNRZ1066 LMG18311

Genome size (bp) 1,796,226 1,796,846

G + C content 39% 39%

Coding sequences 1915 1890

CDSs present in both genomes 1785 1785

Pseudogenes 182 180

Number of insertion–deletion regions

(indels) >50 bp

25 30

Total size of indels (bp) 72,180 71,692

Total size of indels <50 bp and single

nucleotide differences (bp)

3923 3923

Single nucleotide differences (SNPs) 2905 2905

Short (1–3) sequence shifts 362 362

for a circular map). Most ORFs (about 80%) of both

strains are orthologous to other streptococcal genes, indi-

cating that S. thermophilus and its pathogenic relatives

share a substantial part of their overall physiology and

metabolism. Remarkably, a relatively high proportion

of the S. thermophilus genes (approximately 10%) appearto be pseudogenes, due to frameshift, nonsense, deletion

or truncationmutation (Table 1). A nearly identical set of

pseudogenes is present in the two strains, which is pre-

sumably related to their dispensability during growth in

milk. Notably, many of these are involved in carbohy-

drate metabolism, transport and regulation (see Sections

4.1, 4.2, 3.5, and 7).While some redundancy of genes cod-

ing for key metabolic functions has been reported in thegenome of other LAB such as Lc. lactis or Lb. plantarum

[19,20] (e.g., NADH oxidase, lactate dehydrogenase,

pyruvate oxidase, fumarate reductase), this is rarely ob-

served in the genome of S. thermophilus. For example,

few functional gene redundancies are predicted in amino

acid biosynthesis, central metabolism and transcriptional

regulators (see below). Nevertheless, several functions

appear to be encoded bymore than one gene copy (see be-low), which most likely illustrates their importance in the

S. thermophilus lifestyle.

Comparison of the two genomes revealed around

3000 single nucleotide differences (0.15% polymor-

phism). This level of divergence could have arisen in

107 generations, according to the estimated natural

mutation rate [21]. The common ancestor of the two

strains may thus have lived 3000–30,000 years ago,assuming a growth rate between 1 and 10 divisions per

day. This roughly fits the duration of human dairy activ-

ity, believed to have begun about 7000 years ago [2]. The

two genomes differ by 170 single nucleotide shifts,

mostly in mononucleotide stretches (N > 3), and 42 re-

gions of sequence differences >50 bp (indels) that repre-

sent about 4% of genome length. The two strains have

>90% of CDSs in common (Table 1). The main differ-ences concern genes for extracellular polysaccharide bio-

synthesis (eps, rgp), bacteriocin synthesis and immunity,

a remnant prophage, and a locus known as ‘‘clustered

regularly interspaced short palindromic repeats’’, closely

linked to genes of unknown function (cas) [7].

An interesting difference concerns two widely con-

served recombination modulators sbcC and sbcD,

known to act together to dampen the efficiency ofrecombination; these genes are absent in pathogenic

streptococci but are present in the dairy streptococcus

and in the related dairy species, Lc. lactis. Consequently,

genome plasticity might be higher in the pathogens than

in the dairy species, in line with the more sedate life style

of the latter [22,23]. In addition, the streptococci lack

another widely conserved recombination modulator,

RecQ, and consequently appear to undergo rather fre-quent genomic inversions that occur symmetrically

around the origin–terminus axis [7].

438 P. Hols et al. / FEMS Microbiology Reviews 29 (2005) 435–463

2.2. Phylogenic position of S. thermophilus among the

salivarius group using MLST

Recently, the genetic diversity within the salivarius

group has been studied using Multilocus Sequence Typ-

ing (MLST), a method involving the identification ofnucleotide variations in housekeeping genes and allow-

ing the investigation of the population structure [24].

Twenty six S. thermophilus strains isolated from differ-

ent dairy products; twenty two S. salivarius and eight

S. vestibularis strains from the human oral cavity or hu-

man blood were included in these analyses. Seven house-

keeping genes (ilvC, pepO, pyrE, glcK, ddlA, thrS,

dnaE) were amplified and sequenced in the 56 strains(Delorme, C., Bolotin, A., Ehrlich, S.D, Renault, P.,

unpublished data). The analysis of the allelic profiles

of the S. thermophilus strains revealed no significant

cluster that allowed to correlate the sequence type with

the geographic origin or the type of products from

which the strains were isolated. Phylogenetic clustering

using the neighbour-joining method of each locus con-

firmed the status of distinct species for S. thermophilus

In analogy, S. thermophilus strains did not share com-

mon alleles with the two other species, and always clus-

tered in a distinct branch for each locus-specific tree as is

illustrated by the thrS locus tree (Fig. 1). This analysis

suggests that for the MLST loci used, no allele ex-

thrS14S

thrS15S

thrS6S

thrS5S

thrS4S

thrS17S

thrS16S

thrS3S

thrS8S

thrS7S

thrS1S

thrS2S

thrS2V

thrS3V

thrS11S

thrS12S

thrS4V

thrS1V

thrS1th

thrS4th

thrS2ththrS3th

0.01

Fig. 1. Phylogenetic relationship among alleles of thrS locus from

S. salivarius (S), S. vestibularis (V) and S. thermophilus (th) strains. The

tree was constructed by using the neighbor-joining method [191].

changes occur between S. thermophilus and the two

other streptococci. In contrast, phylogenetic clustering

mingles the S. salivarius and S. vestibularis strains, sug-

gesting frequent allelic exchanges between these two spe-

cies (Fig. 1). Interestingly, the sequence divergence

within the S. thermophilus MLST loci is very low, withan average of 0.19%, closely resembling the 0.15% poly-

morphism observed when comparing the two S. thermo-

philus genomes (see above). In contrast, S. salivarius and

S. vestibularis species displayed high divergence within

the MLST loci, with an average rate of 6.6% and

3.6%, respectively. The low polymorphism observed in

S. thermophilus by both MLST and comparative genom-

ics suggests that this species represents a recentlyemerged population with a clonal structure ([7], De-

lorme, C., Ehrlich, S.D., Renault, P., unpublished data),

which is in clear contrast with S. salivarius.

2.3. In silico analysis of horizontal gene transfer events

Although S. thermophilus displayed a low level of

polymorphism, lateral gene transfer (LGT) has contrib-uted to the shaping of the genome. A particularly inter-

esting case of LGT is a 17-kb region inserted in a

truncated pepD gene, encompassing a mosaic of frag-

ments exceeding 90% identity to Lb. bulgaricus and

Lc. lactisDNA sequences. A 3.6-kb fragment within this

region encodes a unique copy of metC required for bio-

synthesis of methionine, a rare amino acid in milk. This

metC gene displays 95% identity with a DNA fragmentof Lb. bulgaricus, suggesting a recent LGT event be-

tween these relatively distant species that are used in

association during yogurt manufacture, and implies that

ecological rather than phylogenetic proximity is a driver

for LGT [7].

Examination of the local and relative G + C content

of the S. thermophilus genome suggests additional

LGT events (Fig. 2 and Supplementary Table 1). Thedeviation in G + C content of 16 genomic regions (10

common to both genomes) exceeds three standard vari-

ations from the mean of the entire genome (Fig. 2). Of

these, 5 are G + C rich and encode ribosomal RNA

operons, unlikely to be transferred by LGT. In contrast,

the 10 regions with an exceptional low G + C content

may have been acquired by LGT, and include genes in-

volved in lantibiotic (lab locus), bacteriocin (blp locus),or exopolysaccharide (eps locus) biosynthesis, as well

as genes predicted to be involved in restriction–modifi-

cation, oxygen tolerance (Rgg regulators) and several

genes of unknown function (Fig. 2). A detailed analysis

of most of these regions and their importance for the

physiology of S. thermophilus is reported in the follow-

ing sections of this review. Another approach to detect

LGT is by examining the level of identity of genes ofan organism with sequences from the relevant dat-

abases. The 138 S. thermophilus genes displaying highest

20

25

30

35

40

45

50

55

60

0 500000 1000000 1500000 2000000

20

25

30

35

40

45

50

55

60

0 500000 1000000 1500000 2000000

1

6

23

45

L37 8 9

12

34 5

L1

L4

C1

10

6 C27 8 910

L2

LMG18311

CNRZ1066

M+3SD

M-3SD

M-2SD

M+2SD

M+3SD

M-3SD

M-2SD

M+2SD

Genome position (bp)

Genome position (bp)

G+C

co

nte

nt

(%)

G+C

co

nte

nt

(%)

Fig. 2. Variations of the G + C (600-bp window) content along

S. thermophilus genomes. Values corresponding to 2 and 3 standard

variations (SD) from the mean (M) are indicated by dotted lines.

Common regions in LMG18311 and CNRZ1066 genomes that

displayed a G + C content above or below 3 SD from M are indicated

by black numbers. 1–5, rDNA; 6, lantibiotic biosynthesis gene cluster

(labBCT); 7, transcriptional regulator MutR/Rgg and putative kinase

protein (stu/str0182-0183); 8, non-coding region; 9, gene cluster

including genes of unknown function and putative restriction–

modification genes (stu/str0706-0707-hsdS1-stu/str0709); 10, putative

bacteriocin locus containing two transcriptional regulators MutR/Rgg

(stu/str1947–1950), a putative peptide ABC exporter (stu/str1949), and

two unique genes of CNRZ1066 (pmrB, multidrug efflux protein and

mccB, bacteriocin biosynthesis protein). L1– L4 and C1–C2 are regions

of high divergent G + C content in LMG18311 and CNRZ1066,

respectively. L1, putative peptidoglycan hydrolase Cse (3 0 end of

stu0442); L2, peptide ABC exporter (stu0324) and hypothetical protein

(stu0325); L3, exopolysaccharide gene cluster (eps); L4, non coding

region; C1, locus containing genes coding for hypothetical proteins

and a putative transcriptional regulator (str0683–0690); C2, locus

containing genes coding for hypothetical proteins and a putative N-

acetylgalactosamine enolpyruvyl transferase (str1040–1042).

P. Hols et al. / FEMS Microbiology Reviews 29 (2005) 435–463 439

homology with non-streptococcal genes and thereforepotentially acquired by LGT are listed in Supplementary

Table 2. Some of these genes are positioned in regions of

highly deviating G + C content, supporting their acqui-

sition by LGT. Taken together, 226 genes might have

been acquired by S. thermophilus through LGT, as

judged by the two criteria above, representing almost

14 % of the total gene content of this species. However,

only 16 genes are revealed by both criteria, underliningboth the value of combination of multiple approaches

and the complexity of unequivocal detection of genomic

regions acquired by LGT. Notably, 47 of these 226

genes (21%) appear to be truncated and are presumably

inactive (Supplementary Tables 1 and 2). This observa-

tion contradicts the notion that acquisition is driven by

selection for a function that provides more efficient

growth or maintenance and could imply that some ofthese genes may have been acquired because of their

proximity to a function under selection and are under-

going the process of elimination.

Analysis of adjacent regions implies the involvement

of mobile genetic elements in LGT events. This is exem-

plified by the close genetic linkage between the lantibi-

otic biosynthesis genes and a phage-related integrase/

recombinase (stu0102), and the transposase homologue(IS3256 family from Lc. lactis) found to flank the region

predicted to encode bacteriocin production. Next to

their involvement in LGT, mobile genetic elements have

most likely played a major role in shaping the S. thermo-

philus genome. This is illustrated by the observation that

about 75% of the S. thermophilus IS elements is associ-

ated with changes in gene order in comparison with its

streptococcal relatives, which is exemplified for a partic-ular region in Supplementary Fig. 2.

3. Biosynthesis of the cell envelope and associated

cellular processes

The bacterial cell envelope is of major importance for

the interaction with the environment, cell–cell communi-cation and resistance to numerous stress conditions.

Notably, key cellular processes involve cell envelope

associated functions such as cell wall components

assembly (e.g., peptidoglycan, teichoic acid, polysaccha-

rides), protein secretion, cell-wall protein anchoring or

solute transport. The biology of the S. thermophilus cell

envelope can be considered as a ‘‘black box’’ with the

exception of few reports concerning key genes involvedin peptidoglycan assembly/degradation and a specific fo-

cus on biosynthesis of exopolysaccharides due to their

importance in industrial processes. Based on our current

knowledge of the molecular biology of the cell envelope

of pathogenic streptococci and related species, the in sil-

ico analysis of the S. thermophilus genome revealed

numerous novel insights such as a particular autolysin

complement, the maintenance of a high genetic diversityin polysaccharide biosynthesis, a strongly limited secre-

tome with a strikingly low number of cell-surface ex-

posed proteins, the presence of a TAT secretion

machinery and an impressively reduced capacity for su-

gar transport.

3.1. Biosynthesis and degradation of peptidoglycan

Peptidoglycan (PG) is one of the major constituents

of the cell envelope of Gram-positive bacteria. The PG

440 P. Hols et al. / FEMS Microbiology Reviews 29 (2005) 435–463

layer plays a key role in cell shape determination and

resistance to osmotic pressure. It is a molecular mesh

of glycan strands made of repeating disaccharide units

(N-acetyl-glucosamine and N-acetyl-muramic acid)

cross-linked by peptide bridges [25]. The structure of

the peptide bridges in the PG of S. thermophilus hasbeen established [26]. The pentapeptides involved in

the cross-linking of glycan strands contain successively

L-alanine, D-glutamate, L-lysine and the dipeptide D-ala-

nine–D-alanine. A peptide bridge made of two or three

L-alanine residues is linked to the e-amino group of L-ly-

sine. These amino acids form cross-bridges between L-

Lys3 and D-Ala4 of two adjacent pentapeptides. The

same PG structure has been identified in Enterococcus

faecalis, S. pyogenes, and S. disagalactiae [26]. The com-

plete set of genes involved in the biosynthesis of the PG

precursor (lipid II) has been identified in the genome,

including three putative UDP-MurNAc-pentapeptide:

L-alanine ligases (MurM, MurN, and MurO) responsi-

ble for the synthesis of cytoplasmic branched precursors

such as recently described in E. faecalis [27]. Modifica-

tions of PG glycan strands are frequently observed suchas the presence of non-acetylated glucosamine residues

due to the activity of an N-acetyl glucosamine deacetyl-

ase (PgdA) [25], which was shown to increase resistance

to cell wall degradation by lysozyme and contribute to

virulence in S. pneumoniae [28]. Interestingly, the S. ther-

mophilus pgdA ortholog appears to be a pseudogene,

suggesting that the lack of this PG modification is of

minor importance for growth in milk.The incorporation of new PG precursors in the exist-

ing mesh and PG maturation are performed by penicillin

binding proteins (PBPs) [25]. The S. thermophilus gen-

ome encodes five high molecular weight PBPs, including

three class A PBPs (bifunctional glycosyltransferase/

transpeptidase) and two class B PBPs (transpeptidase).

In addition, two intact genes and two pseudogenes

encoding low molecular weight PBPs were identified inthe genome (DD-carboxypeptidases). Recently, an inter-

action of PBP2 (class B) with the cell-shape determining

complex composed of MreBCD, and RodA has been

proposed for Escherichia coli [29]. MreB has been desig-

nated as the prokaryotic actin homologue involved in

cytoskeleton formation [29]. The MreBCD, RodA,

PBP2 macromolecular complex appears to play a key

role in directing the PG synthesis machinery for longitu-dinal cell wall biosynthesis in rod-shaped bacteria

[29,30]. Strikingly, ovoid bacterial cells (some strepto-

coccal and staphylococcal species), including S. thermo-

philus, have an mreCD operon devoid of mreB. S.

thermophilus mreD, pbp2b, rodA, and pbp2b-rodA mu-

tants were constructed. Their cells displayed spherical in-

stead of ovoid cell-shape and a reduced cell-size and were

arranged in long chains that were curled rather than lin-ear [31–33]. These data support a key role of these genes

in maintenance of the cell shape of S. thermophilus as has

been found in rod-shaped bacteria and this suggests a

role of these genes in the cell segregation process [32,33].

Incorporation of new peptidoglycan precursors in the

nascent PG, external PG desquamation, and cell separa-

tion is performed by PG hydrolases (autolysins) [25].

The autolysin complement of S. thermophilus was estab-lished on the basis of known PG hydrolase families and

putative PG hydrolases recently discovered in strepto-

cocci [34–37]. Three complete genes (mur1, mur3, and

stu0699) and two pseudogenes (w mur2, w stu1046)

encoding putative PG hydrolases were identified in the

S. thermophilus genome. The three full-length PG hydro-

lases displayed similar features. They contain a signal-

peptidase-I specific N-terminal signal sequence, followedby a PG hydrolase catalytic domain, but lack the classi-

cal repeat domains found in numerous autolysins of

Gram-positive bacteria [37]. Both Mur1 and Mur3 con-

tain a lysozyme/amidase (PFAM01832) catalytic do-

main, while the product of stu0699 contains a partial

Acm domain (COG3757) and a highly positively

charged C-terminal domain. Previously, the Mur1 autol-

ysin was shown to be extracellular in a cell-associatedform in S. thermophilus CNRZ302 [38]. Whereas the

purified enzyme displayed PG hydrolase activity, this

activity could not be detected in crude cell extracts of

S. thermophilus by zymography and its functional role

was not established by gene knockout [38]. Besides these

autolysins, three predicted extracellular proteins (PcsB,

Cse, Cbp1) contain a CHAP domain (PFAM05257),

which has been proposed to hydrolyse PG containingc-glutamyl [39]. The S. thermophilus PcsB modular

organisation displays similarity to its orthologues in

pathogenic streptococci and related species (e.g., S.

pneumoniae PcsB, Lc. lactis USP45). The recognizable

domains in these proteins encompass an N-terminal type

I signal peptide, followed by a large domain (�200 res-

idues) of unknown function, a spacer of high interspe-

cies variability, and the CHAP domain (�120 residues)([35] and references therein). Recently, Ng and collabo-

rators showed that a reduced expression of pcsB in S.

pneumoniae resulted in the formation of long chains of

cells, which is a phenotype often associated with cell wall

hydrolysis deficiency [37]. This cell separation defect is

accompanied with an excess of PG synthesis at nearly

every division septum and cell equator. The authors sug-

gest that PcsB acts as a critical PG hydrolase that bal-ances cell wall synthesis and participates in cell

separation [35]. The modular organisation of the S. ther-

mophilus Cse proteins appears similar to that of PcsB,

except that Cse contains a lysM (PFAM01476) domain

for cell wall attachment instead of the �200 residue do-

main of unknown function identified in PcsB. Two

hypothetical proteins showing a similar lysM-CHAP

modular organisation domain were identified in S. mu-

tans and S. intermedius [34]. Recently, cse deletion deriv-

atives of S. thermophilus LMG18311 and CNRZ368

P. Hols et al. / FEMS Microbiology Reviews 29 (2005) 435–463 441

were constructed [34]. The main difference with the wild

type was a growth-phase independent long chain pheno-

type, supporting a function of the S. thermophilus Cse as

a PG hydrolase involved in cell separation. However,

PG hydrolase activity could not be detected by zymog-

raphy using embedded S. thermophilus cells [34]. More-over, the S. thermophilus CbpD1 protein is orthologous

to the choline binding CbpD protein of S. pneumoniae

[36] but lacks the C-terminal choline binding motifs

for attachment on teichoic acids. Besides the CHAP-

containing proteins, S. thermophilus contains a single

member (stu0757) of the putative PG hydrolase family

containing a PECACE domain (PEptidoglycan CArbo-

hydrate Cleavage Enzyme), and displaying similarity tolytic transglycosylases (e.g., E. coli Stl70) [36]. Finally,

the orthologs of LrgA (stu1524) and LrgB (stu1523),

which are putative modulators of PG hydrolase activity

in Staphylococcus (St. aureus) were also identified in the

S. thermophilus genome sequence [37].

Overall, S. thermophilus shows a unique PG hydrolase

complement since predicted PG muramidases/glucos-

aminidases/amidases lack cell wall attachment domainsclassically found in many PG hydrolases from Gram-

positive bacteria involved in cell separation (e.g., Lc. lac-

tis AcmA, S. pneumoniae LytA, LytB, LytC). Based on

preliminary results obtained for the S. thermophilus

Cse-deficient strain, the cell separation process mediated

by PG hydrolases seems to involve at least one secreted

member of the CHAP protein family in this species.

3.2. Teichoic acids

Teichoic acids are the second major component of the

cell walls of many Gram-positive bacteria [25]. These an-

ionic cell wall polymers are generally made of glycerol-P

or ribitol-P repeating units anchored on the PG (wall

teichoic acids, WTA) or attached to the cytoplasmic

membrane (lipoteichoic acids, LTA). Polyglycerolphos-phate LTAs, which are substituted with D-alanine, are

found in many streptococci, including S. pyogenes, S.

agalactiae, S. mutans, S. vestibularis, and S. salivarius

[40–42]. However, S. pneumoniae produces unique tei-

choic acids (WTA and LTA) consisting of repeats of a

tetrasaccharide linked to ribitol-5P and substituted by

phosphocholine residues [40]. To the best of our knowl-

edge, the structure of teichoic acids from S. thermophilus

remains to be established. Careful examination of the S.

thermophilus genome failed to reveal any obvious gene

cluster involved in teichoic acid biosynthesis such as

the tag or tar gene clusters found in B. subtilis [43].

Moreover, S. thermophilus lacks all the lic genes in-

volved in phosphocholine substitution of teichoic acids

in S. pneumoniae [7], which is in good agreement with

the absence of predicted extracellular proteins contain-ing choline-binding motifs in S. thermophilus. However,

S. thermophilus contains a complete dlt operon, which is

responsible for D-alanylation of teichoic acids in other

Gram-positive organisms [44].

3.3. Biosynthesis of polysaccharides

Extracellular polysaccharides are produced by a vari-ety of bacteria and are present as capsular polysaccha-

rides (CPS and LPS) bound to the cell surface, or are

released into the growth medium (EPS). These polymers

can consist of a single type of sugar (homopolysaccha-

rides) or are regular, repeating units consisting of differ-

ent sugars (heteropolysaccharides). Most strains of S.

thermophilus synthesise heteropolymer EPS [45], but

some S. thermophilus strains are also encapsulated [46].The production of EPS during milk fermentation

provides a desirable ‘‘ropy’’ or viscous texture to the fer-

mented product [47,48], which contributes to mouth-

feel, texture, and taste perception typically associated

with certain fermented dairy products [49]. In addition,

production of EPS could contribute to maintenance of

texture properties and avoid syneresis in products with

reduced fat levels such as yogurt, sour cream or cheeses.Most S. thermophilus EPS are highly variable and

predominantly composed of galactose, glucose, and

rhamnose, but N-polymers containing acetyl-galactos-

amine, fucose, and acetylated galactose moieties have

also been reported [50]. A large array of different CPS

is produced by pathogenic streptococci, such as S. pneu-

moniae, where CPS variation is an important determi-

nant of the more than 90 different serotypes describedfor this species.

At least 12 distinct EPS gene clusters have been se-

quenced in S. thermophilus to date (for a review see

[51], [52]). The eps gene cluster in the genome of S. ther-

mophilus CNRZ1066 [7] appears to be identical to a pre-

viously described cluster from strain Sfi6 [53]. More than

60 different S. thermophilus eps gene clusters may exist,

as suggested by restriction fragment length polymor-phism analysis [52]. Generally, eps clusters are consid-

ered to be highly diverse. Mobile genetic elements play

a role in this diversity, as is illustrated by the almost

identical sequences encountered in streptococci, but also

in more distant species like Lc. lactis and several lacto-

bacilli [52]. The modular gene organisation in eps gene

clusters is conserved and the biosynthesis of EPS is pro-

posed to occur via a common molecular mechanism.EPS biosynthesis and genetics have been reviewed re-

cently and will only be discussed shortly here [47,51].

The genes encoded at the 5 0 end of the clusters display

the highest level of overall conservation, ranging from

95% to 100%, and have been reported to be involved

in polymer chain length determination, and regulation

of EPS production. Several studies with S. pneumoniae

indicate that the reversible phosphorylation state oftyrosine residues in the C-terminal end of CpsD plays

a critical role in regulation of CPS production [54,55].

442 P. Hols et al. / FEMS Microbiology Reviews 29 (2005) 435–463

The CpsD phosphorylation state is proposed to be di-

rectly modulated by the autokinase activity of CpsD it-

self, as well as the phosphatase activity of CpsB.

However, the exact role of these factors in CPS produc-

tion appears to be strain dependent, as is illustrated by

the finding that cpsB mutation leads to different pheno-types in different strains [56]. Generally, a gene encoding

a glycosyl-1P transferase or priming glycosyltransferase

is present downstream of the regulatory genes, which in

its turn is followed by a variable number of additional

glycosyltransferase genes that are required for the

sequential addition of sugar moieties to the lipid carrier

undecapronylphosphate to form the lipid-linked repeat-

ing units. Finally, repeating unit transporter and poly-merase functions are usually encoded close to the 3 0

end of the gene cluster. However, although it was ini-

tially proposed that the 3 0 end of the eps clusters was de-

fined by the inversely oriented orf14.9, recent work

suggests that several eps related genes may still be found

downstream of this gene in some clusters [51,52]. Next

to the genes encoding EPS biosynthesis machinery itself,

several eps gene clusters contain genes involved in theproduction of sugar nucleotide precursors that are re-

quired as substrates for one or more of the eps glyco-

syltransferases [52]. The finding of such genes within

eps gene clusters provides functional redundancy in S.

thermophilus, since similar genes are present in the gen-

ome, suggesting that these enzyme activities could repre-

sent limiting steps in EPS biosynthesis. In analogy,

overexpression of the UDP glucose pyrophosphorylaseencoding galU, in combination with phosphoglucomu-

tase deficiency (pgmA inactivation) led to a 2-fold

increased EPS yield. Similarly, combination of overex-

pression of Leloir pathway enzymes in a Gal+ mutant

and pgmA inactivation increased EPS yield [57].

Overall, eps clusters in S. thermophilus are usually

around 15–20 kb in length, and are generally flanked

by the deoD and bglH genes. Since different genes arefound to flank cps clusters in S. pneumoniae (dexB and

aliA), the conservation of eps gene clusters among strep-

tococci appears to be restricted to the genes encom-

passed by those clusters. Finally, the EPS producing

capacity in S. thermophilus has been reported to be an

unstable characteristic [58]. This might be explained by

the frequent presence of IS elements directly flanking

or even within eps gene clusters, which is corroboratedby the finding that loss of EPS production in a sponta-

neous mutant of S. thermophilus Sfi39 was associated

with IS905 transposition into the epsF* gene [59]. More-

over, the occurrence of several copies of the same IS ele-

ment within the same cluster has been encountered and

is likely to render a genetically highly unstable situation

[52]. These features probably also correspond with the

high degree of eps gene cluster diversity encounteredamong S. thermophilus strains (see above), exemplifying

its underlying molecular mechanism. However, since no

clear biological function could be associated with EPS

production in S. thermophilus to date [51], the evolution-

ary advantage driving eps gene cluster diversity in this

species remains to be established. In contrast, it is well

established that EPS/CPS production by commensal or

pathogenic streptococci is involved in interactions withthe host and required for evasion of the host�s immune

system [60]. Although it is very likely that S. thermophi-

lus originated from a strain related to the commensal

species S. salivarius, eps gene cluster diversity is not

likely explained by a polyphylogenetic origin of S. ther-

mophilus. This is underlined by the finding that most

variable eps genes seem to be provided by bacteria of

close ecological proximity such as Lc. lactis, suggestingthat the diversity driving selection occurs in milk.

There are a number of reports of LAB that synthesise

mixtures of EPSs that may have different structures and/

or different molecular weights [50]. Interestingly, the S.

thermophilus genome contains a second gene cluster pre-

dicted to be involved in polysaccharide production (Sup-

plementary Fig. 3). Similar clusters are present in other

streptococci, and in particular in S. mutans where thesegenes allow the synthesis of a rhamnose–glucose poly-

saccharide (RGP), which plays a role in its pathogenic-

ity. Six conserved genes (rgpA through rgpF) are

involved in the assembly of the rhamnose–glucose poly-

saccharide (RGP), while two or more variable genes lo-

cated downstream (rgpH and rgpI) are required for

glucose side-chain coupling, control the frequency of

branching in S. mutans [61], and are held responsiblefor RGP-related serotype diversity among strains [62].

The S. thermophilus rgpAF homologues share high iden-

tity levels with their S. mutans counterparts. However, in

contrast to S. mutans, the variable genes (rgpH and rgpI

like) are located upstream of the rgpAF cluster in S.

thermophilus rather than downstream (this variation in

genetic organisation is visualised in Supplementary

Fig. 3). It remains to be established whether S. thermo-

philus can actually produce RGP, and what its func-

tional role could be.

3.4. Protein export pathways and secretome

3.4.1. Secretion and processing machinery

Components of the secretion machinery found in S.

thermophilus include the signal recognition particle pro-teins (Ffh and FtsY), the general trigger factor chaper-

one (RopA, prolyl isomerase), and the components of

the Sec translocase (SecA, SecE, SecG, SecY, YajC).

The SecDF counterparts are absent, as in the genomes

of other streptococci, Lc. lactis and various Lactobacilli.

Two orthologs (Stu1810, Stu0245) of the E. coli YidC

protein were identified, which could interact with the

Sec translocase and play a stimulatory role in the inser-tion of membrane proteins [63]. Strikingly, two compo-

nents (TatA, TatC) of the twin arginine translocation

P. Hols et al. / FEMS Microbiology Reviews 29 (2005) 435–463 443

(TAT) pathway were identified in the S. thermophilus

genome. This pathway is dedicated to the secretion of

folded proteins that generally contain cofactors/pros-

thetic groups (e.g., FeS or NiFe centres, hemes) and ap-

pears to be absent in other streptococci, Lc. lactis and

various Lactobacilli [64]. In addition, the S. thermophi-

lus genome is predicted to encode two signal peptidases

I (SipA, SipB), a single signal peptidase II (LspA) and a

prolipoprotein diacylglyceryl transferase (Lgt) for the

cleaveage of prolipoproteins and their coupling to mem-

brane lipids, respectively, two putative signal peptidases

of unclear function (Sip, PilD), a single PrsA/PrtM pep-

tidylprolyl isomerase (lipoprotein) involved in the fold-

ing of exported proteins and a single HtrA serineprotease homologue, potentially involved in the degra-

dation of misfolded secreted proteins. Besides the TAT

translocase, all of these components are also present in

the pathogenic streptococci (Supplementary Table 3).

3.4.2. Secreted and cell-surface exposed proteins

The secretome of S. thermophilus LMG18311 and

four additional genomes of pathogenic streptococci(S. pyogenes M1, S. pneumoniae TIGR4, S. agalactiae

2603V/R, S. mutants UA159) were analysed in silico

(Supplementary Table 3). A total of 89 proteins (exclud-

ing pseudogene products) of S. thermophilus LMG18311

contain a predicted signal peptide cleaved by SPase I.

Approximately 50 of these proteins are predicted to be

secreted in the external medium or cell-wall attached,

indicating that S. thermophilus encodes the lowestamount of putative secreted proteins known among

streptococci. This group of proteins includes 9 proteins

dedicated to PG assembly/degradation (see above), 7

proteins involved in transport processes, 1 compe-

tence-associated nuclease, 1 deoxyribonuclease, 1 serine

protease (HtrA), 2 putative hydrolases and around 20

hypothetical proteins. Extensive searches revealed only

two proteins with a LysM-like cell-wall attached motif(the putative Cse autolysin and the hypothetical protein

Stu2005), while the secretomes of pathogenic strepto-

cocci contained 15–35 cell-wall attached proteins. The

latter group included 6–23 proteins predicted to be an-

chored to the peptidoglycan in a sortase-dependent

manner [65]. Notably, the genome of S. thermophilus

LMG18311 encodes at least 14 pseudogenes ortholo-

gous to genes encoding cell-wall attached proteins inother streptococci, including three protein fragments

containing the typical LPXTG, indicative of sortase-

dependent anchoring. In analogy, both the LMG18311

and CNRZ1066 genomes contained a sortase-resem-

bling pseudogene (w strA). In contrast, the incomplete

S. thermophilus LMD9 genome appears to encode an in-

tact sortase as well as three LPXTG-containing proteins,

including the previously identified cell-wall proteasePrtS ([66], see Section 5.3), a protein with highest

homology with mucus binding proteins described for

various lactobacilli [67], and the cyclo-nucleotide phos-

phodiesterase CpdB. Intriguingly, adherence of S. ther-

mophilus strains to inert surfaces appears to be

correlated with the presence of an intact srtA gene

(Bolotin, A., unpublished data).

Cell surface exposure of proteins can also be achievedby the presence of a single, non-cleaved N- or C-termi-

nal transmembrane-anchor sequence or by membrane

anchoring as a lipoprotein (SPase II dependent). The

LMG18311 genome encodes 55 proteins with an N-ter-

minal anchor sequence, which is by far the smallest

number of proteins belonging to this category among

streptococci. S. thermophilus LMG18311 encodes 24

predicted lipoproteins, including 15 ABC-transporterassociated substrate-binding proteins, which does not

deviate much from the number of proteins included in

this category in pathogenic streptococci (ranging from

25 to 32). Since surface exposed proteins of pathogenic

streptococci are required for adhesion to mucosal sur-

faces and evasion of host defences, they can be consid-

ered as key-factors for virulence [5]. The absence of

these genes (or their presence as pseudogenes in some in-stances) corroborates the food-grade and �safe� status ofS. thermophilus [7].

The finding of the TAT secretion machinery in the

three S. thermophilus genomes is highly intriguing. A

single candidate TAT-exported protein (Stu1023) con-

taining the [S/T]-R-R-X-F-L-K consensus motif (TIGR-

FAM1409) could be identified in the S. thermophilus

genome, which could function as an iron dependent per-oxidase (DyP type family). This function could play a

role in oxidative stress tolerance since an ISS1 insertion

mutant obtained in strain CNRZ368 showed an in-

creased resistance to oxidising agents [68]. Notably,

the stu1023 gene appears to be genetically linked to

the TAT encoding genes (tatA; stu1020 and tatC;

stu1021) and this locus also encompasses two genes

(stu1022, stu1024) encoding proteins predicted to be in-volved in iron transport. This genetic organisation is un-

iquely found in S. thermophilus. Notably, Stu1023

displays a high level of identity (78%) with Spr1179 en-

coded by S. pneumoniae R6. However, Spr1179 lacks a

TAT signal sequence.

3.5. Global view of membrane proteins and solute

transport systems

A global prediction of membrane proteins using

TMHMMV2.0(http://www.cbs.dtu.dk/services/TMHMM)

was achieved for S. thermophilus LMG18311 (pseudo-

genes excluded). 326 membrane proteins containing at

least one TMH were identified, representing approxi-

mately 21% of total proteins, thereby resembling the rela-

tive amount of this class of proteins encountered inpathogenic streptococci (ranging from 20% to 24%) (Sup-

plementary Table 3). Of these 326 proteins, 220 are

444 P. Hols et al. / FEMS Microbiology Reviews 29 (2005) 435–463

predicted to contain more than 2 TMHs and 117 proteins

were clearly annotated as transport system (TPS) compo-

nents. Detailed classification of TPSs was performed by

TCDB classification (http://tcdb.ucsd.edu/index.php)

(Supplementary Tables 4A–C) and compared with simi-

lar classifications available for pathogenic streptoco-cci at TransporterDB (http://www.membranetransport.

org) (Supplementary Table 3). In S. thermophilus

LMG18311, 95 complete TPSs were identified (encom-

passing 195 proteins), including 48 ATP-binding cassette

(ABC) transporters (47%), 29 secondary transporters

(35%), 7 ion channels (9%), 6 F- or P-type ATPase (7%),

and 2 sugar phosphotransferase systems (PTS, 2%). Sim-

ilarly to other Gram-positive bacteria that lack an elec-tron transport chain, ATP-dependent transporters

appeared to be over-represented (54%) [69]. Within the

largest group, the ABC transporters, 30 were classified

as importers and 18 as exporters. Of the importers, 11

are predicted to import amino acids or peptides, while

the specificity ofmost of the exporters is unknown. In gen-

eral, the comparative analysis of the transporter content

among streptococci revealed thatS. thermophilus displaysthe lowest capacity for sugar uptake (see also below). Fur-

thermore, in S. thermophilus themain functional category

of transporters contains a relatively high level of pseudo-

genes (20%), with the ABC transporters (14 pseudoTPSs)

and PTS systems (4 pseudoTPSs) as the most affected

subclasses.

4. Sugar metabolism, central carbon pathways and

regulation

The main role of S. thermophilus in dairy fermenta-

tions is the rapid conversion of lactose into lactate but

also the production of other compounds that contribute

to flavour and texture. Five different sugars are fer-

mented by S. thermophilus: lactose, sucrose, glucose,galactose, and fructose [70–72]. The two last sugars

are only fermented by a limited number of strains. S.

thermophilus is highly adapted to grow on lactose as

carbon source and detailed investigations of lactose

transport, regulation, and metabolism are available

[71,73–77]. Similarly, the diversity of fermentation

end-products is rather limited. Besides L-lactate, the

main fermentation product, low levels of formate, acet-oin, diacetyl, acetaldehyde, and acetate have been de-

tected as additional end-products [78–80]. Here we

present a genome-based reinvestigation of carbon

metabolism in this species.

4.1. Sugar utilisation and transport systems

The pathogenic streptococci display a certain capac-ity to degrade complex polymers such as starch or glu-

cans (dextrans). Specific enzymes, such as dextranase

and mutanase, are produced by oral streptococci (S. mu-

tans, S. salivarius) and are involved in the remodelling of

glucan structures in dental plaque [81,82]. Other en-

zymes such as levansucrase are directly involved in the

synthesis of fructans from sucrose and contribute to bio-

film formation [83,84]. In the complete genomes of 4pathogenic streptococci (S. pneumoniae, S. pyogenes,

S. agalactiae, and S. mutans), genes coding dextranases,

amylases and pullulanases were identified for each spe-

cies. The capacity to degrade, remodel or synthesise

these complex polymers is severely reduced in S. thermo-

philus (Fig. 3). The genome contains three pseudogenes

(w dexT, w pulA, w sacB) encoding remnants of a dex-

tranase, a pullulanase and a levansucrase. However, acomplete amylase gene was identified (amyL), which is

in agreement with the finding that this species displays

low amylolytic activity [85].

Sucrose and fructose are the only sugars that are

transported by phosphoenolpyruvate-dependent phos-

photransferase systems (PTS) in S. thermophilus, but

with a lower efficiency than lactose [71]. Glucose is a

non-PTS sugar in S. thermophilus and a poor substratefor growth [86]. Remarkably, S. thermophilus is highly

adapted to growth on lactose, which is transported by

LacS. Once inside the cell, lactose is hydrolysed into glu-

cose and galactose by b-glalactosidase (LacZ). While all

strains utilise the glucose moiety of lactose, galactose is

generally not metabolised and is expelled in the external

medium by LacS, which functions as a lactose/galactose

antiporter (Fig. 3) [74]. The absence of galactose utilisa-tion does not result from lack of genetic information

since the genes coding for the Leloir pathway (galK-

TEM) are present in the genome of numerous strains

unable to metabolise galactose (Fig. 3) [71]. The analysis

of spontaneous Gal+ mutants and the genetic comple-

mentation by the galactokinase gene (galK) from S. sal-

ivarius have clearly established that the low

galactokinase activity is the major limiting step in galac-tose utilisation in S. thermophilus [72,87].

In the two S. thermophilus genomes, 7 loci code for

PTS systems (ptsG, fruA, bglP, treP, manLMN, celB,

scrA), 2 loci for sugar ABC (malEF, stu/str0809-811)

transporters and 2 loci for sugar porters (lacS, stu/

str0855) (Fig. 3). Strikingly, both genomes contain many

pseudogenes in loci dedicated to sugar utilisation and

uptake: w ptsG (glucose PTS), fruR fruB w fruA (fruc-tose PTS), w bglP (b-glucosides PTS), w treR w treP wtreA (trehalose PTS), w scrR scrB scrA scrK (sucrose

PTS), malQ malR w malE w malF (maltose/matodextrin

ABC transporter). Three of these PTS systems (w ptsG,

w bglP, and w fruA) were also shown to be inactive in 8

additional strains of S. thermophilus with the exception

of a single intact copy of fruA in one strain, while these

systems appeared intact in four S. salivarius strains[7]. Additionally, although the manLMN PTS system

(mannose, fructose, glucose PTS transporter) of

[manMNL] scrA [ψ ptsG][ψ fruA][ψ bglP][ψ treP]

celB (IIC)[ψ malEF] Sugar ABC ?GlucosePorter ? lacS

PTS ABC PorterABCPTS

[malQ][ψ bglAH][ψ treA]

sucrose glucose lactose

lactose

galactose

galactoseGlu-6P GluSuc-6P

Fru

glcKscrB

Fru-6P scrK

? ?

pgi

FBP

pfk

fba

gapAPgk

gpmA, [ψ B], Ceno

PEP

Pyruvate

tpi

Fru-1P

fruB

ppcOxo-glutarate

aa

lacZ

icd, citB, gltA

OAC

ptsK, ptsIH

Glu-1P

pyk

Xylulose-5P

Ribulose-5P

Ribose-5P

PRPP

Pyr, Pur, folate

tkt rpe

rpiA

prsA1, A2

Man-6P

pmi1, [ψ 2]

?

[galM][galK][galT][galE1]

UDP-sugarsEPS, RPS, TA

pgmA

galUgalE2, [ψ 3]rmlABCD

α-amylase (amyL) [amylopullulanase (ψ pulA)] [dextranase (ψ dexT)][levansucrase (ψ stu0354)]

Ac-P ccpA,

[fruR], [galR],[ψ treR], [ψ scrR]

Glucosamine-6P

UDP-N-acetyl-glucosamine

nagAglmMgcaD

EPS, PG

glmS, nagB

tkt

Fig. 3. Sugar metabolism and central carbon pathways in S. thermophilus. Plain arrows correspond to potential active enzymatic reactions catalysed

by the corresponding genes products encoded by the S. thermophilus genome. Dotted arrows correspond to enzymatic reactions from incomplete/

inactive pathways in S. thermophilus. Genes in between brackets are either shown to be inactive, present as pseudogene (w or -tr), or intact but

included in operons containing pseudogenes. Complex carbohydrates metabolism (green): amyL, a-amylase; dexT-tr, 1,6-a-glucanhydrolase(dextranase); pulA-tr, alkaline amylopullulanase; stu0354-tr, Levansucrase precursor (b-D-fructofuranosyltransferase). Sugar uptake (blue, light blueindicates non functional transport systems): ABC-MSP1/MSP2/NBP (stu0809–stu0810), carbohydrate ABC uptake transporter; bglP-tr, b-glucosidePTS system component IIABC; celB, similar to cellobiose PTS system component IIC; fruA-tr, fructose PTS system component IIABC; lacS, lactose

permease; malE-tr malF-tr, maltose/maltodextrin ABC uptake transporter; manLMN, mannose PTS system IIABCD; ptsG-tr, glucose PTS enzyme

component IIABC; ptsH, phosphocarrier protein HPr; ptsI, phosphoenolpyruvate:sugar phosphotransferase system enzyme I; treP-tr, trehalose PTS

trehalose component IIBC; scrA, sucrose PTS component II; stu0855, glucose/ribose porter (GRP) family protein. Sugar catabolism (red): bglA-tr, 6-

phospho-b-glucosidase; bglH-tr, 6-phospho-b-glucosidase; fruB, fructose-1-phosphate kinase; galE1, UDP-glucose 4-epimerase; galK, galactokinase;

galM, aldose 1-epimerase; galT, galactose-1-phosphate uridylyltransferase; glcK, glucose kinase; lacZ, b-galactosidase; malQ, 4-a-glucanotransferase;pmi1 pmi2-tr, mannose-6-phosphate isomerase; scrB, sucrose-6-phosphate hydrolase; scrK, fructokinase; treA-tr, trehalose-6-phosphate hydrolase.

Glycolysis (black): eno, 2-phosphoglycerate dehydratase, enolase; fba, fructose-bisphosphate aldolase; gapA, glyceraldehyde-3-phosphate

dehydrogenase; gpmA, gpmB-tr, gpmC phosphoglycerate mutase; pfk, 6-phosphofructokinase; pgi, glucose-6-phosphate isomerase; pgk, phospho-

glycerate kinase; pyk, pyruvate kinase; tpi, triosephosphate isomerase. Biosynthesis of nucleotide sugars and nucleotide aminosugars (orange): galE2,

galE3-tr, UDP-glucose 4-epimerase; galU, UDP-glucose pyrophosphorylase; pgmA, phosphoglucomutase; rmlA, glucose-1-phosphate thymidyl

transferase; rmlB, dTDP-glucose-4,6-dehydratase; rmlC, dTDP-4-dehydrorhamnose 3,5-epimerase; rmlD, dTDP-4-keto-L-rhamnose reductase);

gcaD, UDP-N-acetylglucosamine pyrophosphorylase; glmM, phosphoglucosamine mutase; glmS, glucosamine-fructose-6-phosphate aminotrans-

ferase; nagA, N-acetylglucosamine-6-phosphate deacetylase; nagB, glucosamine-6-phosphate isomerase. Pentose phosphate pathway (purple): prsA1,

prsA2, ribose-phosphate pyrophosphokinase; rpe, ribulose-phosphate 3-epimerase; rpiA, ribose 5-phosphate isomerase; tkt, transketolase. 2-

Oxoglutarate biosynthesis (brown): citB, aconitate hydratase; gltA, citrate synthase; icd, isocitrate dehydrogenase; ppc, phosphoenolpyruvate

carboxylase. Regulation of sugar metabolism (grey): ccpA, catabolite control protein A; fruR, fructose operon transcriptional repressor; galR,

galactose operon repressor; malR, maltose operon transcriptional repressor, putative; scrR-tr, sucrose regulon regulatory protein; treR-tr, trehalose

operon transcriptional repressor GntR family. ABC, ATP-binding cassette transporters; Ac-P, acetyl phosphate; EPS, exopolysaccharides, FBP,

fructose-1, 6-biphosphate; Fru, fructose; Glu, glucose; Man, mannose, OAC, oxaloacetate; PEP, phosphoenolpyruvate; PG, peptidoglycan; PRPP, 5-

phosphoribosyldiphosphate; PTS, sugar phosphotransferase systems; Pur, purines; Pyr, pyrimidines; RPS, rhamnose polysaccharides; Suc, sucrose;

TA, teichoic acids.

P. Hols et al. / FEMS Microbiology Reviews 29 (2005) 435–463 445

S. thermophilus appears intact and is expressed, it was

recently shown to be unable to transport glucose and

mannose, raising doubts about its functionality [88]. Be-

sides the sucrose PTS (scrA) and lacS, three loci are can-didates for glucose uptake: celB (incomplete PTS,

similar to cellobiose PTS orphan-IIC), stu/str0855 (a

putative ribose/glucose porter of GRP family) and stu/

str0809-811 (a putative sugar ABC transporter). The

high level of pseudogenes and the presence of non-func-tional genes in sugar utilisation clearly illustrate the

446 P. Hols et al. / FEMS Microbiology Reviews 29 (2005) 435–463

regressive evolution of S. thermophilus towards a spec-

ialised microbe dedicated to lactose utilisation. In

contrast, the genomes of pathogenic streptococci (S.

pneumoniae, S. pyogenes, S. agalactiae, and S. mutans)

revealed a high flexibility towards sugar metabolism

(13–20 PTS systems and 5–7 sugar ABC transporters),which is proposed to contribute to their pathogenic po-

tential [5,12]. Several pathogenic streptococci also con-

tain genes for citrate, malate, gluconate transport and

utilisation, which are not present in S. thermophilus.

4.2. Central metabolism and pyruvate dissipating

pathways

A global reconstruction of central metabolism in S.

thermophilus is depicted in Fig. 3. The entry of glucose

and sucrose-6P in the glycolytic pathway probably in-

volves glucokinase (GlcK) and sucrose-6P hydrolase/

fructokinase (ScrB, ScrK), respectively. Analogous with

pathogenic streptococci, S. thermophilus contains a gene

encoding a putative transketolase (Tkt), interconnecting

glycolysis and the pentose phosphate pathway leading to5-phosphoribosyl diphosphate (PRPP), the precursor

for de novo synthesis of purines, pyrimidines and folate.

Complete pathways for de novo synthesis of these com-

pounds were identified in the genome. Additionally, a

metabolic pathway is present for the conversion of phos-

phoenolpyruvate into 2-oxo-glutarate, a key precursor

in the biosynthesis of numerous amino acids (see Section

5.1). Among streptococci, this pathway, which involvesa putative phosphoenolpyruvate carboxylase (Ppc) and

three genes of the TCA cycle (gltA, citB, icd), is only

present in S. thermophilus and S. mutans. The intercon-

nection between glucose-6P and glucose-1P, the precur-

sor of nucleotide sugars (UDP-glucose, UDP-galactose,

and TDP-rhamnose) involved in polysaccharides bio-

synthesis (EPS and RPG), was shown to involve a single

phosphoglucomutase enzyme (PgmA) in S. thermophilus

LY03 [89]. The synthesis of UDP-glucose from glucose-

1P takes place via the UDP-glucose pyrophosphorylase

(GalU). Notably, three copies of galE (galE1 [gal oper-

on], galE2, and w galE3), predicted to encode two func-

tional UDP-glucose 4-epimerases that are responsible

for the interconversion between UDP-glucose and

UDP-galactose, are present. Finally, a pathway leading

to the production of UDP-N-acetyl glucosamine, a pre-cursor for both peptidoglycan and EPS [90], was

identified.

S. thermophilus is an obligate homolactic bacterium

and produces L-lactate as the major fermentation end-

product from sugar metabolism (>95%) in a range of

growth conditions. However, the production of low

amounts of alternative fermentation end-products, such

as a-acetolactate (a-AL), acetoin, acetaldehyde, formate,and acetate has been reported [78–80,91]. Aerobic sugar

metabolism of S. thermophilus STH450 was shown to re-

sult in the production of substantial amounts of CO2,

acetoin, and trace amounts of a-AL and diacetyl [79].

The production of these compounds by strain STH450

correlated with 8-fold increased water-forming NADH

oxidase activity in this strain compared to S. thermophilus

ATCC19258. Teraguchi et al. [79] also reported the ab-sence of pyruvate dehydrogenase, pyruvate oxidase,

and peroxidase activities. Growth of S. thermophilus in

milk results in the production of formate [91], which

plays an important role in the stimulation of Lb. bulgar-

icus growth [92] and probably results from pyruvate-

formate lyase activity (Pfl). Recently, it was shown that

the Pfl protein from S. thermophilus LMG18311 and four

other strains was strongly induced during growth in milkunder anerobic conditions, and represented one of the ten

most abundant proteins under these conditions (Derzelle,

S., Bolotin, A.,Mistou,M.-Y. andRul, F., personal com-

munication). Although acetaldehyde production by S.

thermophilus has been reported to involve threonine

catabolism via the bifunctional serine hydroxymethyl

transferase/threonine aldolase enzyme (GlyA) [93], the

production of this compound from sugar metabolismvia an unknown pathway has also been reported [78].

The number of pyruvate dissipating enzymes identified

in the S. thermophilus genome is extremely limited in

comparison to other streptococci and Lc. lactis. Only

three enzymes for pyruvate dissipation were detected:

the L-lactate dehydrogenase (Ldh), the pyruvate-formate

lyase (Pfl) and its activating enzyme (PflA) and the a-AL

synthase (Als). Overall, two alternative pyruvate dissi-pating routes besides lactate dehydrogenase were identi-

fied: the Pfl/phosphotransacetylase (Pta)/acetate kinase

(AckA) pathway for formate and acetate production

and the Als/a-AL decarboxylase (AldB) route for acetoin

production (Fig. 4(a)). The absence of genes coding for

the pyruvate dehydrogenase complex was confirmed.

However, an acetoin dehydogenase complex (AcoABCL)

sharing high similarity with compounds E2 and E3 of thepyruvate dehydrogenase complex was identified in the

genome. In other species, such as B. subtilis, this complex

is involved in acetoin utilisation when the sugar source is

completely exhausted [94]. Notably, two alcohol/acetal-

dehyde dehydrogenase genes (w adhA and w adhE) and

the diactyl/acetoin reductase gene (w butA) are pseudo-

genes in both genomes sequences, which explains the

absence of ethanol and 2–3 butanediol production byS. thermophilus. Similarly to what was reported above

in relation to PTS systems, the adhE and butA are pseu-

dogenes in numerous S. thermophilus strains and are in-

tact in the closely related S. salivarius [7].

In order to evaluate the functionality of the pyruvate

dissipating pathways and their implication in the forma-

tion of fermentation products, the inactivation of the

unique ldh gene was attempted. Several unsuccessful ap-proaches led to the conclusion that inactivation of ldh is

probably lethal due to the absence of an alternative

α-AL

pfl, pflA

Pyruvate

ldh

ackA

L-Lactate Acetoin [2,3-BD]

aldB

pta

[ψ butA]

[ψ adhA][ψ adhE]

als

DiacetylAc-coA

Ac-P

Acetate

Acetaldehyde

[Ethanol] Formate[ψ butA]

acoABCL

[ψ acuB ]

[ψ adhA][ψ adhE]adhB

-O2 +O2

Pyruvate

ldh

L-Lactate

Acetaldehyde

Ethanol

adhB

pfl, pflA

pdcZm

O2 H2O

noxSt

α-ALPyruvate

ldh

L-Lactate Acetoin

aldB

als

Diacetyl

pfl, pflA

O2 H2O

noxSt

(a)

(b) (c)

Acetaldehyde+A c-coA

(pta)ackA

Ac-coA/Ac-P

Acetate

str/stu1263

Acetoin

als

aldB

Fig. 4. Pyruvate metabolism and modulation of the NADH/NAD+ balance in S. thermophilus. (a) Pyruvate dissipating pathways. Fermentation end-

products from S. thermophilus, which are detected in aerobic (+O2) or anaerobic conditions (�O2), are boxed. End-products in between brackets were

not detected either due to incomplete pathways or to pathways containing pseudogenes. Black arrows correspond to potential active enzymatic

reactions catalysed by the corresponding genes products encoded by the S. thermophilus genome. Thick and thin black arrows correspond to NADH-

dependent and NADH-independent enzymatic reactions, respectively. Small dotted arrows correspond to missing enzymatic reactions (pseudogenes).

Striped arrow corresponds to the chemical oxidative decarboxylation of a-acetolactate into diacetyl. Genes in between brackets are pseudogenes (wor -tr). AckA, acetate kinase; AcoABCL, acetoin dehydrogenase complex; acuB-tr, acetoin utilisation protein; adhA-tr, alcohol dehydrogenase; adhB,

alcohol dehydrogenase, zinc-containing; adhE-tr, alcohol–acetaldehyde dehydrogenase; aldB, a-acetolactate decarboxylase; als, acetolactate

synthase; butA-tr, acetoin reductase; ldh, L-lactate dehydrogenase; pfl, pyruvate-formate lyase; pflA, pyruvate-formate lyase activating enzyme; pta,

phosphotransacetylase. a-AL, a-acetolactate; Ac-P, acetyl-phosphate; Ac-coA, acetyl-coenzyme A; 2,3-BD, 2,3-butanediol. (b) Modulation of the

NADH/NAD+ balance in LMG18311 resulting in the production of a-AL, acetoin and diacetyl in aerobic conditions. Fermentation end-products

resulting from the overexpression of the S. thermophilus nox gene (NADH oxidase (H2O-forming)) in an aldB (a-acetolactate decarboxylase) mutant

strain of LMG18311, indicated by the road block in the aldB arrow, are boxed. Arrow description, genes symbols, and abbreviations are as in panel

a. (c). Modulation of the NADH/NAD+ balance in LMG18311 resulting in the production of acetaldehyde, ethanol and acetate in aerobic

conditions. Fermentation end-products resulting from the co-overexpression of the S. thermophilus nox gene and the pdc gene (pyruvate

decarboxylase) from Z. mobilis are boxed. str/stu1263, aldehyde dehydrogenase. Arrow description, genes symbols, and abbreviations are as in

panel a.

P. Hols et al. / FEMS Microbiology Reviews 29 (2005) 435–463 447

route from pyruvate for NAD+ regeneration as shown

by the reconstruction of pyruvate metabolism in S. ther-

mophilus (Hols, P., Van Der Kaai, H., unpublished data)(Fig. 4(a)). Knockout of ldh of S. mutans was also

impossible and a thermosensistive mutant could only

be rescued by complementation with the alcohol dehy-

drogenase gene from Zymomonas mobilis, which allows

NAD+ regeneration [95]. An alternative strategy to eval-

uate pyruvate metabolism is to modulate the NADH/

NAD+ ratio such as described for Lc. lactis by the over-

production of the water-forming NADH oxidase

[96,97]. The constitutive overexpression of the nox gene

from S. thermophilus in strain LMG18311 led to the pro-

duction of a mixture of lactate and acetoin in aerobicconditions (Hols, P., unpublished data) (Fig. 4(b)).

Additionally, Nox overproduction in a a-AL decarbox-

ylase (AldB) deficient mutant resulted in an extracellular

production of a-AL and diacetyl in addition to lactate

(Hols, P., unpublished data) (Fig. 4(b)). In the presence

of oxygen, the link between pyruvate and acetaldehyde

or acetyl-coA is missing in S. thermophilus. This link

was restored by the co-overexpression of the pyruvate

448 P. Hols et al. / FEMS Microbiology Reviews 29 (2005) 435–463

decarboxylase gene (pdc) from Z. mobilis and the nox

gene [98]. In this case, sugar fermentation resulted in

the production of a mixture of acetaldehyde, ethanol

and acetate in addition to lactate and acetoin (Hols,

P., unpublished data) (Fig. 4(c)). The production of eth-

anol could be explained by the presence of an alcoholdehydrogenase gene (adhB) that converts acetaldehyde

into ethanol but does not use acetyl-CoA as substrate.

The production of acetate under these conditions was

unexpected. One possible route is the conversion of acet-

aldehyde into acetyl-P or acetyl-CoA by an aldehyde

dehydrogenase (stu1263) (Fig. 4(c)). Taken together,

these observations confirmed the functionality of several

pyruvate dissipating pathways identified by in silicometabolic reconstruction and illustrate the potential of

the development of new S. thermophilus starters with im-

proved production of key aroma compounds (e.g., acet-

aldehyde, diacetyl).

4.3. Regulation of galactose and lactose metabolism

In view of its importance for fermentation in milk,lactose metabolism of S. thermophilus has extensively

been studied. The genes involved in galactose (Leloir en-

zymes) and lactose (transport and hydrolysis) metabo-

lism are genetically linked and form a gene cluster

(galKTEMlacSZ), which is highly conserved among dif-

ferent strains of this species [7,71,99]. Moreover, this gal

lac gene cluster organisation is conserved in S. salivarius

[70], reflecting the close evolutionary relationship be-tween these bacteria. Upstream of the S. thermophilus

galK gene, a divergently oriented transcriptional regula-

tor encoding gene, galR, is found. GalR was shown to

positively regulate the transcription of the two operons

encompassed by the gal-lac cluster, galKTEM and

lacSZ, while it negatively regulated its own expression

[72]. In contrast, the GalR homologue of S. mutans,

was shown to repress expression of the galKTE genes,indicating that different streptococci employ variable

modes of regulation of gal-gene expression using highly

homologous regulatory factors [72,100]. Upon galactose

or lactose-mediated induction, the structural galKTE

genes of S. thermophilus appeared to be transcribed at

a very low level, accounting for the galactose negative

phenotype displayed by many strains of this species.

Interestingly, spontaneous galactose fermenting mutantsof S. thermophilus CNRZ302 could readily be obtained

and were shown to express galKTE at a much higher le-

vel as compared to the wild type. Molecular analysis re-

vealed that this was due to specific promoter-up

mutations within the galK promoter region [72], which

was corroborated by the finding of specific galK pro-

moter-up mutations in isolates of S. thermophilus that

can ferment galactose [71].In addition to positive regulation by GalR, the lacS

promoter of S. thermophilus is also controlled by the

catabolite control protein A, CcpA, which is in agree-

ment with the presence of a CcpA responsive element

(cre) in the lacS promoter. CcpA has been studied in

many Gram-positive bacteria where it mediates catabo-

lite repression when cells are grown on PTS carbon

sources, of which glucose is generally the most preferred(for a review see [101]). As stated above, and in contrast

to many other Gram-positive bacteria, S. thermophilus

prefers lactose over glucose and transports this disac-

charide via the non-PTS transporter LacS. Functional

disruption of ccpA in S. thermophilus CNRZ302 relieved

lacSZ repression when cells were grown on lactose [77].

In addition to regulating the levels of LacS and LacZ,

CcpA also regulated the expression levels of three key-enzymes of the glycolytic pathway in S. thermophilus.

The expression of lactate dehydrogenase (ldh), phospho-

fructokinase and pyruvate kinase (pfk pyk operon) genes

appeared to be induced in a ccpA-dependent manner

when cells were grown on lactose [77,102]. CcpA medi-

ated control of the same glycolytic enzymes has also

been reported for several other bacteria, including S. bo-

vis and Lc. lactis [103,104], suggesting that global catab-olite control of glycolytic flux is a conserved regulation

mechanism in LAB. At the physiological level, the ccpA

mutant displayed a strongly reduced growth rate on lac-

tose, while at the metabolite level, the mutant displayed

a strongly increased rate of lactose uptake and hydroly-

sis, massive expulsion of both galactose and glucose,

and reduced lactate formation rates [77]. It was con-

cluded that CcpA-control of the lacS promoter plays akey role in fine-tuning of lactose uptake and hydrolysis

rates with overall glycolytic capacity.

Carbon catabolite control by CcpA is closely linked

to the metabolic and energy status of the cell via modu-

lation of the histidine-containing phosphocarrier protein

(HPr) phosphorylation status. Differential phosphoryla-

tion of HPr at a histidine (HPr-His-P) and/or a serine

(HPr-Ser-P) residue represents a key-determinant in car-bon metabolism in many bacteria and is primarily af-

fected by the intracellular metabolites that reflect the

metabolic and energy state of the cell, including the gly-

colytic intermediate fructose biphosphate (FBP) and

ATP (for a review see [101,102,105]). In S. thermophilus

HPr-Ser-P is the dominant HPr form during exponential

growth in the presence of excess lactose [75]. Under

these conditions, HPr-Ser-P forms a complex withCcpA, which exerts carbon catabolite control including

the above mentioned repression of lacSZ and induction

of ldh and pyk pfk, thereby fine tuning lactose import

and glycolytic flux to optimise growth and energy pro-

duction [75,77,102]. Upon lactose limitation, a transi-

tion of the HPr phosphorylation state takes place,

leading to HPr-His-P as the dominant form of HPr

[75], which leads to a decreased glycolytic flux and in-creased expression of lacSZ due to relieved CcpA-med-

iated carbon catabolite control [77,102]. While in many

P. Hols et al. / FEMS Microbiology Reviews 29 (2005) 435–463 449

bacteria HPr-His-P function appears to be limited to its

role in the phosphorelay involved in PTS sugar trans-

port (for a review see [101]), this HPr species has an

additional function in S. thermophilus: it can transfer

its phosphoryl group to the PTS-like IIA domain of

LacS, which has been shown to stimulate the galac-tose–lactose exchange rate of this transporter [73–

75,101]. Thus, HPr has a dual role in regulation of

lactose metabolism in S. thermophilus. Overall, these mod-

ulations of lactose metabolism in S. thermophilus allow

the effective adaptation of its central carbon metabolism

and transport capacities to the availability of lactose in

the growth medium. The lactose metabolism character-

istics of S. salivarius are strikingly different from thoseseen in S. thermophilus. S. salivarius also transports lac-

tose via a IIA-domain containing LacS-transporter.

However, S. salivarius LacS phosphorylation not only

depends on HPr-His-P but can also be generated by

HPr-Ser-P-His-P, a doubly phosphorylated form of

HPr present in large amounts in rapidly growing S. sal-

ivarius cells [106,107]. Moreover, S. salivarius utilises

both the galactose and glucose moieties of lactose, andthus does not expel galactose into the medium during

growth on lactose like S. thermophilus. The impact of

IIA-domain LacS phosphorylation in S. salivarius is less

clear, since a phosphorylation defective mutant does not

display significant differences in terms of its lactose and

galactose metabolism as compared to the wild type

[107].

Preliminary DNA microarray experiments compar-ing S. thermophilus wild type and ccpA mutant cells sug-

gest that the expression of a relatively large number of

transcriptional regulators is affected, including several

that are predicted to regulate various metal transport

systems (Hols, P., Guedon, E., unpublished data), sug-

gesting major secondary responses under CcpA control.

In addition, ccpA mutation affected the expression of

several genes involved in nitrogen and co-factor metab-olism (Hols, P., Guedon, E., unpublished data), which

is in good agreement with similar studies in other

Gram-positive bacteria that suggest that the ccpA regu-

lon exceeds carbon metabolism and includes various

other metabolic pathways and secondary responses

[108–110].

5. Nitrogen metabolism and regulation

LAB are generally recognised as auxotroph for sev-

eral amino acids. Since milk is poor in free amino acids

and short peptides, optimal growth of LAB in milk re-

quires hydrolysis of caseins, internalisation and degra-

dation of the resulting peptides or de novo amino acid

biosynthesis. Consequently, their capacity to metabolisenitrogen is of the highest importance for the efficacy of

the acidification process in milk. Genomic analysis of

S. thermophilus reveals that it maintained the capacity

to use caseins and to synthesise several amino acids

throughout evolution, and both aspects reveal this spe-

cies� adaptation to milk.

5.1. Amino acids biosynthesis

S. thermophilus, like most LAB, requires an exoge-

nous supply of amino acids for growth. However, it

seems to be less demanding than several other LAB such

as Lc. lactis and dairy lactobacilli. Single-amino-acid

omission analysis revealed that the number and the type

of essential amino acids required for growth is strain-

dependent [111–113]. While some S. thermophilus strainsexhibited no absolute amino acid requirement, others

appeared auxotrophic for at least four amino acids.

Multiple-omission experiments of amino acids have

shown that depletion of both glutamate and glutamine,

or of the two sulphur-containing amino acids cysteine

and methionine, completely abolishes growth of S. ther-

mophilus, indicating that the corresponding synthesis

pathways are not fully functional [112]. In addition,while not essential for growth, amino acids like leucine,

valine, methionine, and cysteine have been shown to

stimulate growth [111–114].

In silico prediction of the amino acid biosynthesis

pathways of LMG18311 and CNRZ1066 was performed

(Fig. 5). Both strains seem to contain all the genes cod-

ing for the enzymes required for the biosynthesis of all

amino acids with the exception of histidine. This aminoacid is synthesised from 5-phosphoribosyl diphosphate

(PRPP) via a well-conserved pathway involving 11 enzy-

matic reactions. Only hisK, coding for a putative histidi-

nol-phosphatase that shares 26% of identity with HisK

of Lc. lactis, is present in both genomes. This observa-

tion is in apparent contradiction with the finding that

most S. thermophilus strains are histidine prototrophs

[112]. However, additional experiments have confirmedthe in silico prediction of histidine auxotrophy for

LMG18311 and CNRZ1066 (Delorme, C., Guedon,

E., unpublished data). Interestingly, a complete his gene

cluster is present in the partial sequence of the LMD9

genome and comparative analysis of the corresponding

region in CNRZ1066 and LMG18311 suggests that the

his gene cluster was deleted in a common ancestor of

the latter two strains. The three genome sequences lacka glutamate synthase gene, which could explain the lack

of growth when both glutamate and glutamine were

omitted, as well as the glutamate auxotrophy of some

S. thermophilus strains [112,113].

The relatively high conservation of functional amino

acid biosynthetic genes in S. thermophilus might reflect

the importance of amino acid synthesis for growth in

milk, which is in agreement with the low level of pseudo-genes associated with amino acids biosynthesis. Only

alaD, dapE, ilvD2, and yhcE involved in the biosynthesis

CysteineLysine

Glucose

Glyceraldehyde-3-P

Glycerate-3-P

SerineGlycineserC

PEP

Pyruvate

Erythrose-4-P

Chorismate Tryptophan

Acetyl-CoA

Oxaloacetate

Homoserine Aspartatesemialdehyde

homthrBthrC

2-Oxo-butanoate

ilvA

Cysteine

Methionine

Ribose-5-P Histidine

Threonine

Isoleucine

ilvNilvCilvDilvE

Methionine

Tyrosine Phenylalanine

ValineLeucine

ilvE

ilvNilvCilvD

ilvEleuB

leuCleuA

Aspartate

Aspargine

asnA

lysCasd

Arginine

Alanine

glyA

aspC, araT

aroG1G2BHDEKAC

trpGDFCABaroH

2-Oxoglutarate

araT, aspC,gdhA

Glutamine

Glutamate

Proline

glnANH3

+

Arginine

proBproAproC

argJBCDJOrnithine

argFargGargH

dapAdapBdapD

dapE-tr

lysA

metAcysDmetEF, mmuM

metEF, mmuMcysM2metB

metEF,mmuM

cysM1cysE2, cysE1

metC

stu0353

argHargG

pheAaspC/araT

tyrAaspC/araT

serB

serA

hisK

Alanine

AT

AT

alaD-tr

Fig. 5. Amino acids biosynthetic pathways in S. thermophilus. This proposition of metabolic building of amino acids biosynthetic pathways of S.

thermophilus was deduced from gene products encoded by the S. thermophilus LMG18311 genome and combined data from KEGG database (http://

www.genome.jp), from the annotation of the S. thermophilus LMG18311 genome sequence (http://www.ncbi.nlm.nih.gov/genomes) and from

knowledge of amino acid biosynthetic pathways in other bacteria. Black arrows correspond to potential active enzymatic reactions catalysed by the

corresponding genes products encoded by the S. thermophilus genome. Grey arrows correspond to potential inactive enzymatic reactions due to the

absence of the corresponding gene products in S. thermophilus. AlaD-tr, alanine dehydrogenase; araT, aromatic amino acid aminotransferase; argB,

acetylglutamate kinase; argC, N-acetyl-c-glutamyl-phosphate reductase; argD, acetylornithine aminotransferase; argF, ornithine carbamoyltrans-

ferase; argG, argininosuccinate synthase; argH, argininosuccinate lyase; argJ, ornithine acetyltransferase/amino-acid acetyltransferase; aroA,

3-phosphoshikimate 1-carboxyvinyltransferase; aroB, 3-dehydroquinate synthase; aroD, 3-dehydroquinate dehydratase; aroE, shikimate 5-

dehydrogenase; aroF, chorismate synthase; aroG1, phospho-2-dehydro-3-deoxyheptonate aldolase; aroG2, phospho-2-dehydro-3-deoxyheptonate

aldolase; aroH, chorismate mutase; aroK, shikimate kinase; asd, aspartate-semialdehyde dehydrogenase; asnA, asparagine synthetase A; aspC1,

aspartate aminotransferase; aspC2, aspartate aminotransferase, putative; aspC3, aspartate aminotransferase, putative; AT, undefined aminotran-

ferase; bcaT, branched chain amino acid aminotransferase; cysD, O-acetylhomoserine sulfhydrylase; cysE1, serine acetyltransferase; cysE2, serine

acetyltransferase; cysM1, cysteine synthase; cysM2, cystathionine b-synthase; dapA, dihydrodipicolinate synthase; dapB, dihydrodipicolinate

reductase; dapD, 2,3,4,5-tetrahydropyridine-2-carboxylate N-succinyltransferase, putative; dapE-tr, succinyl-diaminopimelic descuccinlyasadipepti-

dase, truncated; gdhA, NADP-specific glutamate dehydrogenase; glnA, glutamine synthetase; glyA, serine hydroxymethyltransferase; hisK, histidinol-

phosphatase, putative; hom, homoserine dehydrogenase; ilvA, threonine deaminase; ilvB, acetolactate synthase, large subunit; ilvC, ketol-acid

reductoisomerase; ilvD1, dihydroxy-acid dehydratase; ilvD2-tr, dihydroxy-acid dehydratase, truncated; ilvN, acetolactate synthase, small subunit;

leuA, 2-isopropylmalate synthase; leuB, 3-isopropylmalate dehydrogenase; leuC, 3-isopropylmalate dehydratase large subunit; leuD, 3-isopropylm-

alate dehydratase small subunit; lysA, diaminopimelate decarboxylase; lysC, aspartate kinase; metA, homoserine O-succinyltransferase; metC,

cystathionine b-lyase; metB, cystathionine c-lyase; metE, 5-methyl tetrahydropteroyltriglutamate/homocysteine methyltransferase; metF, 5,10-

methylenetetrahydrofolate reductase; mmuM, homocysteine S-methyltransferase (S-methylmethionine); pheA, prephenate dehydratase; proA, c-glutamyl phosphate reductase; proB, c-glutamyl kinase; proC, pyrroline carboxylate reductase; serA, D-3-phosphoglycerate dehydrogenase; serB,

phosphoserine phosphatase; serC, phosphoserine aminotransferase; thrB, homoserine kinase; thrC, threonine synthase; trpA, tryptophan synthase,

alpha subunit; trpB, tryptophan synthase, b subunit; trpC, indole-3-glycerol phosphate synthase; trpD, anthranilate phosphoribosyltransferase; trpE,

anthranilate synthase component I; trpF, N-(5 0-phosphoribosyl)-anthranilate isomerase; trpG, anthranilate synthase component II; tyrA, prephenate

dehydrogenase.

450 P. Hols et al. / FEMS Microbiology Reviews 29 (2005) 435–463

of alanine, lysine, branched chain and sulphur aminoacids, respectively, are pseudogenes in strain

LMG18311, while CNRZ1066 lacks both these genes

and aspC3, encoding aspartate aminotransferase. The

product of yhcE displays similarity to the vitamin B12-

independent 5-methyltetrahydropteroyltriglutamate-

homocysteine S-methyltransferase, and its orthologue

in Lc. lactis was shown to be involved in the synthesisof cysteine from methionine [115]. In all three S. thermo-

philus genome sequences, yhcE is truncated by a con-

served stop codon, indicating that the mutation

occurred in a common ancestor of the three strains. This

gene inactivation may explain the auxotrophy for at

least one of the two sulphur amino acids.

P. Hols et al. / FEMS Microbiology Reviews 29 (2005) 435–463 451

Although S. thermophilus has lost numerous genes

unessential for growth in milk, it also acquired new

functions. Notably, all three genome sequences contain

the cysM2 metC cysE2 gene cluster, which shares 95%

identity with a cluster from Lb. bulgaricus and encodes

enzymes that play roles in cysteine and methionine bio-synthesis (Fig. 5) [7]. Moreover, two copies of ilvD

(ilvD1, wilvD2) encoding a dihydroxyacid dehydratase

involved in branched-chain amino acid (BCAA) biosyn-

thesis are present in S. thermophilus, but only the ilvD1

copy is intact. IlvD1 displays up to 91% identity with

IlvD proteins from streptococci, while IlvD2 shows the

highest level of identity (72%) with its orthologue from

St. aureus. Furthermore, w ilvD2 is located betweenstu1875 and ilvBNC in S. thermophilus while it is absent

from orthologous stu1875 ilvBNC gene clusters in other

streptococci. Although the absence of functionality of wilvD2 remains to be established, these observations

strongly suggest that this gene was acquired by LGT.

To conclude, all potential genes involved in synthe-

sis of most amino acids were identified in both genome

sequences. However, only the functionality of genes in-volved in BCAA and proline biosynthesis was experi-

mentally investigated in S. thermophilus [111,116].

Nevertheless, since most of S. thermophilus strains

are able to grow with a mixture of glutamine, methio-

nine, leucine, isoleucine, valine and histidine as sole

sources of amino acids, it is likely that the biosynthetic

pathways of other amino acids are fully functional

[112].

5.2. Amino acids catabolism

The catabolism of amino acids plays an important

role in providing precursors for the biosynthesis of ami-

no acids, nucleotides, and vitamins, generating energy in

nutrient-limited environments and increasing intracellu-

lar pH. In addition, amino acids catabolism has a spe-cific impact on the formation of a large number of key

aroma compounds in fermented dairy products. Catab-

olism of amino acids is commonly initiated by a trans-

amination step, requiring the presence of an a-ketoacid as the amino group acceptor, to produce a-ketoacids. The S. thermophilus genome contains aminotrans-

ferase genes similar to those identified in Lc. lactis such

as three aspartate aminotransferases (aspC1, aspC2,aspC3), one aromatic aminotransferase (araT) and one

BCAA aminotransferase (bcaT) as well as glyA encoded

threonine aldolase activity [78,93]. The glutamate dehy-

drogenase activity, which produces a-keto-glutaratefrom amino acid transamination, is widespread among

S. thermophilus strains, but its level of activity is

strain-dependent and often higher than in other LAB

([117]; Yvon, M., personnal communication). Addition-ally, S. thermophilus was shown to produce leucine,

phenylalanine, and methionine-derived aroma com-

pounds [117]. However, besides the involvement of the

glutamate dehydrogenase and of the specific amino-

transferases, the enzymes involved in the subsequent

reactions leading to production of aroma compounds

remain uncharacterised.

5.3. Proteolytic system

S. thermophilus contains a proteolytic system similar

to that found in other milk LAB and is composed of

(i) an extracellular cell-anchored protease capable of

casein hydrolysis, (ii) a set of amino acid and peptide

transport systems required for import, (iii) a set of intra-

cellular peptidases involved in the hydrolysis of casein-derived peptides and various house-keeping processes.

Although S. thermophilus is potentially equipped with

more than 20 proteolytic enzymes, its proteolytic system

remains poorly characterised and is largely inferred

from the well-studied species Lc. lactis.

The extracellular protease PrtS of S. thermophilus is a

sortase cell-wall-anchored serine proteinase of the sub-

tilisin family [66,118,119]. PrtS is present in only a fewstrains of S. thermophilus and its presence has been asso-

ciated with rapid growth and acidification rates in milk

[66,118,119]. PrtS is essential for the optimal growth of

S. thermophilus when present alone in milk. However,

when co-cultivated with a proteinase-positive Lb. bul-

garicus strain, S. thermophilus is capable to grow opti-

mally using peptides released by Lb. bulgaricus [120].

This cooperation could explain the absence of PrtS innumerous S. thermophilus strains. The prtS gene is pres-

ent in the LMD9 genome but absent in the two other

genomes [7]. PrtS is similar to both C5a/Csp peptidases

from pathogenic streptococci and to cell envelope prote-

ases from LAB. PrtS and other proteases of streptococ-

cal origin differ from cell envelope proteases of other

LAB by the absence of the B-domain involved in

auto-proteolysis [66,121]. The PrtS maturation processis unknown but the putative peptidylprolyl isomerase

PrsA displays low level identity with the Lc. lactis PrtM

maturation protein [122] and could be involved in this

process.

Although 15 different peptidase activities have been

detected after chromatographic separation of S. thermo-

philus intracellular proteins [123], a limited set of endo-

and aminopeptidases has been characterised to date(PepC, PepN, PepO, PepS, and PepX) (Supplementary

Table 5 and references therein). All of these peptidases

have orthologues in other streptococci as well as Lc. lac-

tis, except PepS, which is found in streptococci but is ab-

sent in Lc. lactis. Additionally, the S. thermophilus

genome contains intact genes coding for peptidases nec-

essary for basic cellular processes (SPases, HtrA, Clp

proteases, carboxypeptidases) as well as a range of pep-tidases homologous to those present in other LAB and

Gram-positive bacteria (Supplementary Table 5). The

452 P. Hols et al. / FEMS Microbiology Reviews 29 (2005) 435–463

latter group includes di- and tripeptidases (e.g.), a puta-

tive membrane-associated pheromone maturation pepti-

dase (Eep) such as described for E. faecalis and several

proteases or glycoprotein endopeptidases.

5.4. Amino acid and peptide transport systems

Experimental data on S. thermophilus amino acid or

peptide transport systems is very limited. The only sys-

tem characterised in detail is the oligopeptide ABC

transport system from an industrial strain that was

shown to be essential for growth in milk [124,125]. This

system was called Ami since its proteins show highest

identity (62–86%) with the Ami system from S. pneumo-

niae [125,126]. The Ami transporter is composed of

three oligopeptide binding proteins (AmiA1, 2, 3), two

membrane proteins that form the permease (AmiC, D)

and two ATPase (AmiE, F) that provides energy to

the system. The amiA1 and amiCDEF genes are organ-

ised in an operon while amiA2 and amiA3 are unlinked

and surrounded by remnants of IS elements. The orga-

nisation of ami genes in the LMG18311 and CNRZ1066genomes is slightly different since amiA1 is a pseudogene

and the other amiA genes (amiA2, amiA3, and w amiA4)

are differently localised in the two chromosomes. Since

the three AmiA oligopeptide binding proteins from the

industrial strain were all functional and displayed differ-

ent but overlapping specificities, the absence of a func-

tional amiA1 paralogue in LMG18311 and CNRZ1066

should only have a minor impact on growth in milk.Although no amino acid and di-tripeptide transport

systems have been characterised in detail in S. thermo-

philus, growth experiments have shown that amino acids

as well as mixtures of very short peptides have a stimu-

latory effect on growth [112]. In addition, only prelimin-

ary experimental evidence for transport of glutamate

[127,128] and BCAA is available for this species [129].

The LMG18311 and CNRZ1066 genomes contain aproton-dependent di-tripeptide transporter, homolo-

gous to DtpT from Lc. lactis [7]. Unlike the situation

in Lc. lactis, no ATP-dependent di-tripeptide transport

system (Dpp, [130]) appears to co-exist with DtpT in

S. thermophilus. However, dpp-like gene fragments ap-

pear to be present in both S. thermophilus genomes (woptABCD). In silico analysis of amino acid transport

system-encoding genes indicated the presence of severalsystems that are predicted to be complete and func-

tional, such as six symporters, a branched-chain amino

acid ABC transporter (Liv), and several glutamine

ABC transporters (Supplementary Tables 4A–C). Over-

all, the main transport systems for peptides and amino

acids could be identified in the genome sequence, and

some of the corresponding genes appeared to be sub-

jected to breaking, shuffling and duplication processes(Supplementary Tables 4A–C).

5.5. Nitrogen metabolism regulation

Regulation of the nitrogen metabolism in LAB is

poorly documented and most of the data were obtained

in Lc. lactis. In this bacterium, amino acid biosynthetic

operons coding for histidine, tryptophane and BCAAare under the transcriptional control of an attenuation

mechanism [131–133]. In silico analysis of upstream re-

gions of amino acids biosynthetic genes and operons

of S. thermophilus revealed that they display many fea-

tures typical for transcription attenuation control, sug-

gesting that this regulation mechanism is widespread

in S. thermophilus.

The amino acid biosynthesis of sulphur amino acids,arginine, and glutamine is regulated by dedicated tran-

scriptional regulators in Lc. lactis. In this species, FhuR

globally controls sulphur amino acid metabolism [115],

ArgR and AhrC control arginine synthesis and the argi-

nine deaminase pathway, respectively [134], and GlnR

controls glutamate and glutamine synthesis [135]. Final-

ly, the expression of genes coding for proteins involved

in nitrogen supply in Lc. lactis (PrtP, Opp, Dpp, PepC,and PepN) is negatively controlled by CodY in response

to intracellular BCAA level [136,137]. Analysis of the

LMG18311 and CNRZ1066 genomes indicates that

homologues of CodY (Stu1635), ArgR (Stu0048), GlnR

(Stu0177), AhrC (Stu1214), and three homologues of

FhuR (Stu0520, Stu0452, and Stu0872) are present in

S. thermophilus. DNA binding sites for these regulators

have been proposed in Lc. lactis [115,134,135] and asearch for their occurrence in the S. thermophilus gen-

ome revealed the presence of similar motifs upstream

of the relevant genes, suggesting conservation of lacto-

coccal nitrogen regulatory mechanisms in S. thermophi-

lus. Although the proteolytic system of S. thermophilus

displays some specific features, it also shares many enzy-

matic factors with the system found in Lc. lactis such as

PepC, PepN, PrtP/S, and Opp. Therefore, it is temptingto speculate that S. thermophilus CodY also controls the

proteolytic system of this bacterium as shown in Lc. lac-

tis [136–138]. Preliminary data on prtS regulation in

S. thermophilus support such a role for CodY [139]. Ta-

ken together, very limited experimental data on nitrogen

metabolism regulation in S. thermophilus are available,

but in silico analyses suggest regulatory patterns similar

to those found in Lc. lactis.

5.6. Urea metabolism and regulation

S. thermophilus is the only LAB starter displaying sig-

nificant urease activity. Conversion of urea into ammo-

nia during dairy processes involving S. thermophilusmay

have some impact on flavour and acidification profiles

due to natural variations of the urea level of milk. Theurease gene clusters from S. thermophilus and S. saliva-

rius show a similar organisation (ureIABCDEFGMQO)

P. Hols et al. / FEMS Microbiology Reviews 29 (2005) 435–463 453

[7,140]. The structural proteins are encoded by ureABC,

accessory proteins by ureEFGD, urease permease by ureI

and nickel uptake system by ureMQO. The ure gene

cluster is transcribed from a vegetative promoter up-

stream of ureI in both species ([140,141], Anba, J., San-

chez, N., Ehrlich, S.D., and Renault, P., unpublisheddata). Transcriptional regulation of the S. salivarius ure-

ase operon is rather complex: it is activated by both low

pH and sugar excess and involves components of the

PTS as well as one or more unidentified transcriptional

regulator [141–144]. Although the functional role of the

S. thermophilus urease has not been completely assessed,

it was proposed to protect against acidic challenge and

provide an additional nitrogen source in S. salivarius

[145]. Although it is tempting to propose that urease

plays a similar role in S. thermophilus, de-acidification

of milk by urea degradation occurs at a relatively high

pH that is not associated with a significant loss of viabil-

ity, suggesting that urease may not provide an important

selective advantage during fermentation. Also, S. ther-

mophilus ure mutants display a similar fitness level as

the wild-type strain during growth in milk [146,147]and addition of the urease inhibitor flurofamide does

not reduce sugar consumption by S. thermophilus

[148]. Growth experiments in milk supplemented with

several amino acids or in chemically defined media show

that S. thermophilus ureC mutants display reduced

growth relative to the wild type, supporting a metabolic

role for urease rather than a function in acid stress resis-

tance. Moreover, approximately 15% of S. thermophilus

strains are urease-negative, corroborating that urease

plays an important but not essential role in this species

[149].

6. Oxygen metabolism and oxidative stress

During dairy processes, S. thermophilus is exposed toreactive O2 species (ROS) and to their potential deleteri-

ous effects [150] that may affect growth, fermentative

capabilities and viability and consequently may have

repercussions on texture and flavour of the final prod-

uct. Although streptococci are not able to respire by

using oxygen as an electron acceptor, S. thermophilus

can grow in aerobic conditions and survive in the pres-

ence of low concentrations of ROS [151]. Survival ofS. thermophilus is attributed to the ROS-activated

induction of an adaptive oxidative stress response

[151]. Valuable information on S. thermophilus oxidative

response, including identification of enzymes involved in

this response, was obtained experimentally using IS

mutagenesis [33,68] and in silico by the genome sequence

analysis [7]. Here, we describe how S. thermophilus can

cope with oxidative stress in both maintaining an appro-priate intracellular redox environment and detoxifying

ROS with antioxidant enzymes.

6.1. Oxidative stress tolerance

Two major systems are responsible for the bacterial

intracellular low redox potential and the maintenance

of proteins in their reduced state: the glutathione

(GSH/Gr/Grx) and the thioredoxin (TRX/TrxR) sys-tems (for a review see [152]). The tripeptide glutathione

(c-GluCysGly, GSH) plays a role in protecting cells

against oxygen toxicity. In contrast to most Gram-

positive bacteria, many Streptococcus strains contain

GSH. Some streptococci such as S. agalactiae ATCC

12927 or S. pyogenes ATCC 8668 are able to synthe-

sise GSH while others, including S. mutans ATCC

33402, import it from their growth medium [153]. Innumerous GSH-containing species, the tripeptide is

synthesised by the consecutive action of c-glutamate-

cysteine ligase (encoded by gshA) and GSH synthetase

(gshB). In Streptococcus agalactiae, the gshA gene has

recently been demonstrated to encode a bifunctional c-glutamylcysteine synthetase-GSH synthetase responsi-

ble for GSH synthesis [154]. One orthologue of the

S. agalactiae gshA gene (stu/str1413) is observed inS. thermophilus LMG18311 and CNRZ1066 genomes.

Additionally, they encode a GSH reductase gene

(gor; GR) known to reduce glutathione disulfide

(GSSG) formed upon oxidation, at the expense of

NADPH. In S. mutans, disruption of the gor gene pre-

vented growth in the presence of diamide [155]. GR

activity in S. thermophilus and S. mutans appeared to

be increased at high O2 level [155] suggesting thatexpression of gor is responsive to the oxygen

concentration.

Thioredoxin and glutaredoxin systems are responsi-

ble for the reduction of intracellular disulfides in vivo.

Both systems play an established protective role against

oxidative stress (for a review see [152]). The S. thermo-

philus genome encodes two thioredoxin (trxA1, trxA2)

and thioredoxin reductase (trxB1, trxB2) genes, as wellas a single glutaredoxin gene (nrdH), suggesting a role

for these systems in the oxidative stress defence.

One of the first protective lines of defence against oxi-

dative stress is the reduction of the intracellular O2 level

(for a review see [150]). Growth of S. mutans in the pres-

ence of O2 correlated with the induction of NADH oxi-

dase and superoxide dismutase [156]. Two distinct

NADH oxidases catalyse either the two electron reduc-tion of O2 to H2O2 (Nox1) or the four-electron reduc-

tion of O2 to H2O (Nox2), thereby providing an

enzymatic defence against oxidative stress [156]. Inacti-

vation of nox1 in S. pyogenes and S. pneumoniae led

to reduced oxidative stress tolerance [157], while Nox1

of S. pneumoniae plays an additional role in oxidative

stress defence via regulation of competence, which al-

lows the acquisition of DNA as a source of nucleotidesand DNA fragments for the repair of O2-induced

damage to the chromosome [158]. In the genome of

454 P. Hols et al. / FEMS Microbiology Reviews 29 (2005) 435–463

S. thermophilus, nox1 is absent and a single H2O-form-

ing NADH oxidase is found.

Superoxide dismutases (Sod) are essential enzymes

for the maintenance of a low intracellular level of

ROS as they convert superoxide anions to molecular

oxygen and hydrogen peroxide. The latter is subse-quently metabolised by catalases and/or peroxidases.

In most species, including S. thermophilus, this activity

is performed by an Mn-superoxide dismutase (SodA)

[159]. Other classes of Sod exist in streptococci,

including an atypical Fe/Mn-Sod in S. pneumoniae

[160]. SodA mutants of streptococci are viable but ex-

tremely susceptible to oxidative stress. S. thermophilus

sodA is expressed in a growth-phase-dependent man-ner, increasing 3- to 4-fold upon entry into stationary

phase [161]. In contrast to some other streptococci in

which sodA expression is induced by oxidative stress

[160], SodA activity in S. thermophilus is not regulated

by O2 [161].

The specific mechanisms by which S. thermophilus

adapts to peroxide stress remain unknown since it lacks

enzymes known to participate in peroxide defence. Asa consequence of its inability to synthesise heme, S.

thermophilus lacks catalase activity [162], but it is also

deprived of other well-known peroxidase encoding

genes. Nevertheless, the S. thermophilus aerotolerance

implies a hydrogen peroxide defence mechanism, which

might involve the thiol-peroxidase encoding psaD gene

[163] or the dpr gene that has been shown to comple-

ment nox1 and ahpC deficiencies in S. mutans [164].Synthesis of Dpr in S. thermophilus has been shown

to be induced during cold shock, suggesting that its

role expands beyond oxidative stress tolerance [165].

The relatively low abundance of detoxifying enzymes

in S. thermophilus as compared to pathogenic strepto-

cocci is in agreement with its lack of pathogenicity,

since these functions are required to survive the

ROS-stress that is part of the host�s inflammatoryresponse.

6.2. Regulators of the S. thermophilus defence to

oxidative stress

The expression of many antioxidant activities is in-

duced by increased ROS levels. To date, no oxidative

stress-specific regulators have been described in strepto-cocci. However, several regulators are known that mod-

ulate the interplay between oxidative stress response,

iron homeostasis, competence and virulence. Iron plays

an essential role in regulatory protein that sense redox

changes in the presence of O2 (for a review see [166]).

Accordingly, increased superoxide radical sensitivity

was observed in S. thermophilus CNRZ368 mutants that

lacked ossH (encoding an iron ABC transporter), sufD,or iscU (both involved in intracellular iron balance) [33].

Thereby, control of iron homeostasis is coupled to ROS

protection in streptococci and could involve the Fur-like

regulator, annotated as fur.

In addition to specific responses to individual types of

stress, general stress response pathways are present in all

bacteria. Streptococci lack the alternative, stress re-

sponse sigma factor SigB (for a review see [167]) andregulation of the general stress response could involve

the CovRS two-component regulatory system and/or

Rgg-like regulators. CovRS has been shown to influence

transcription of �15% of all S. pyogenes genes, includ-

ing those involved in stress adaptation [168]. S. thermo-

philus contains a covRS-like system (annotated as rr01/

hk01, see below). The S. pyogenes Rgg regulator coordi-

nates expression of secreted virulence factors, metabolicenzymes, and general stress response proteins [169]. Its

inactivation revealed its involvement in regulation of

proteins involved in tolerance to oxidative and thermal

stresses in this species [169]. Remarkably, the S. thermo-

philus genome contains 7 rgg-like genes of which one ap-

pears to be truncated (stu1950, also named rrgC in S.

thermophilus CNRZ368). Disruption of either rggC or

rggA in strain CNRZ368 resulted in increased toleranceto oxidative stress ([68]; Fernandez, A. personal commu-

nication), supporting their involvement in S. thermophi-

lus oxidative stress response. Interestingly, both loci

display a highly divergent GC content and could have

been acquired by LGT. Full appreciation of the role

of Rgg-like regulators and their redundancy in S. ther-

mophilus physiology requires the identification of the

target genes of these regulators. Interestingly, Rgg wasshown to affect CovRS expression and vice-versa, indi-

cating that Rgg participates in a global regulatory net-

work including CovRS [168,169].

7. Two-component signal transduction and quorum

sensing

Signal transduction and corresponding gene regula-

tion is generally considered as the main mechanism of

bacterial adaptation to environmental changes. Re-

cently, an exhaustive search for regulatory proteins en-

coded in 145 complete and draft prokaryotic genomes

was reported, in which the identification of regulator

proteins was based on in silico domain analysis. The

authors have classified all of the regulatory proteins asmembers of one-component systems (OCSs, e.g., tran-

scriptional regulators) or two-component systems

(TCSs) [170]. Based on such a classification, S. thermo-

philus contains 81 OCSs and 10 TCSs [170]. However,

the proportion of pseudogenes in this functional class

(15%) is relatively high. Here, we focus on the TCSs of

S. thermophilus and functions that are potentially quo-

rum-sensing (QS) regulated, such as competence or bac-teriocin production.

P. Hols et al. / FEMS Microbiology Reviews 29 (2005) 435–463 455

7.1. Two-component regulatory systems

TCSs are often used in bacteria to sense and respond

to their environment. They play a key role in important

physiological characteristics such as virulence, natural

competence, bacteriocin production, biofilm formation,stress response, and many other adaptive responses

[171]. They typically consist of a sensor or histidine pro-

tein kinase (HK, hk) and an effector or response regula-

tor (RR, rr); the genes encoding these proteins usually

are genetically linked and transcribed as an operon

[172]. Nine of the TCSs encoded by the S. thermophilus

genome display this typical HK and RR linked genetic

organisation, while one orphan HK and one orphanRR can be identified that are not genetically linked

to their cognate partners. The S. thermophilus TCSs

were numbered according their occurrence in the

LMG18311 and CNRZ1066 genomes. Four of the 20

TCS genes in S. thermophilus (rr03, hk03, hk10, and

rr11) are pseudogenes, thereby encompassing one com-

plete TCS (TCS03) and the two orphan genes (HK10

and RR11). Interestingly, the non-functional TCS03 ofS. thermophilus is highly similar to the CiaH/CiaR

TCS of S. pneumoniae that is involved in regulation of

virulence and natural competence development [173].

Overall, S. thermophilus appears to encode eight com-

plete and potentially functional TCSs. All available

S. thermophilus genome sequences (LMG18311,

CNRZ1066, and LMD9) encode these TCSs and these

systems are virtually identical in the three strains.The TCS01 of S. thermophilus displays significant

homology with the covRS system described for other

streptococci that might be involved in regulation of gen-

eral stress responses (see above). The S. thermophilus

TCS02 displays a particularly low GC content (below

32% G + C) and shares the highest similarity with a

E. faecalis TCS of unknown function [174]. No ortho-

logues of TCS02 were found in other streptococci, sug-gesting that this system was recently acquired by S.

thermophilus, probably via LGT (see above). TCS04 dis-

plays high homology with SpaK/SpaR, NisK/NisR, and

DvnK/DvnR of Bacillus subtilis, Lc. lactis, and Carno-

bacterium divergens, respectively, that are involved in

the regulation of the production of bacteriocins in these

species (for a review see [175]). However, no genes with a

potential role in bacteriocin production were found inthe vicinity of the rr04 hk04 genes in S. thermophilus.

The S. thermophilus TCS05 is similar to a TCS that is

shared by various Gram-positive bacteria (e.g., B. sub-

tilis YycG/YycF, S. pneumoniae MicB/MicA or VicK/

VicR, Lc. lactis TCS-C) and has been demonstrated to

be essential for viability in these species [176–180].

TCS06 and TCS07 of S. thermophilus are similar to each

other and resemble the S. pneumoniae TCS11 [178] andLc. lactis TCS-G [179], of which the functions have

not yet been elucidated. The S. thermophilus TCS08 is

similar to S. pneumoniae TCS03 [178] and Lc. lactis

TCS-D that are proposed to be involved in response

to salt or osmotic stress [179]. Finally, TCS09 of S. ther-

mophilus (see Section 7.3), resembles the S. pneumoniae

TCSs BlpH/BlpR and ComD/ComE, involved in the

regulation of bacteriocin biosynthesis and competencefor natural transformation, respectively.

7.2. Competence for natural transformation

Competence for natural transformation is a common

trait among streptococci [181,182]. However, S. thermo-

philus is considered as a non-competent organism. Com-

petence is generally tightly regulated and involves twosets of genes: the early and late competence genes

[181,183,184]. Streptococcal competence development

has been most extensively studied in S. pneumoniae,

but appears to involve similar components in other nat-

urally competent streptococci. In S. pneumoniae, the

early competence genes encode a TCS (ComD/ComE),

its cognate induction factor (competence stimulating

peptide, CSP; ComC), and the components of an ABCtransporter dedicated to CSP export and maturation

(ComA/ComB) [181]. Upon CSP induction, the

ComD-E TCS activates transcription of the competence

sigma factor ComX (two comX copies are present in S.

pneumoniae), which is required for the expression of the

late competence genes, including those encoding the

DNA uptake machinery [181]. Despite its non-compe-

tent reputation, the S. thermophilus genome appears toencode all late competence genes, which is also observed

in several other bacterial genomes of non-competent

bacteria such as Lc. lactis [183]. In addition, the S. ther-

mophilus genome contains a comX-like gene and encodes

a typical peptide pheromone dependent TCS (TCS09;

see above) that is similar to the competence-control loci

of S. mutans (ComD/ComE; [185]) and S. pyogenes

(SilB/SilA; [186]). However, preliminary experimentalwork on the S. thermophilus TCS09 locus suggests its

involvement in bacteriocin production rather than com-

petence (see below; blp locus). Taken together, the

apparent non-competent state of S. thermophilus could

be either due to the lack of dedicated competence-con-

trol loci, or else competence is controlled via an alterna-

tive as yet unknown mechanism in this species.

7.3. Regulation of bacteriocin production; the blp locus

Over the last two decades, the production of antimi-

crobial peptide or bacteriocin production by LAB has

received considerable attention, which is related to their

potential use as natural preservatives in food products

(for a review see [187]). Nevertheless, a relatively limited

amount of information is available on antimicrobialpeptide production by S. thermophilus. To the best

of our knowledge, only 7 bacteriocins produced by

456 P. Hols et al. / FEMS Microbiology Reviews 29 (2005) 435–463

S. thermophilus have been characterised to date ([188]

and references therein). Interestingly, the S. thermophi-

lus genome sequences available all contain two loci that

are predicted to be involved in bacteriocin production.

The first of these loci is designated lab (for lantibiotic)

and contains genes that share similarity with genes gen-erally found in loci involved in lantibiotic production

(for a review see [189]). The S. thermophilus lab locus

is most likely acquired by LGT (see above). However,

the extremely small size of the predicted mature portion

of the structural lantibiotic gene (labA), which is only 9

residues long, raises doubts as to whether this locus is

actually involved in the production of a functional lan-

tibiotic. Moreover, the canonical ABC-transporterencoding genes that are involved in lantibiotic transport

also appear to be missing from the S. thermophilus lab

locus. The second of these loci displays the typical char-

acteristics of a Class II bacteriocin locus and strongly

resembles a part of the blp (bacteriocin-like peptide) lo-

cus in S. pneumoniae [190] and is therefore designated

the S. thermophilus blp locus. The universally conserved

region of these streptococcal blp loci encompasses genesencoding an ABC-transporter (blpA) and an accessory

blpB

blpR

blpH

blpS

blpC

dg

blpT

com

A

com

B

blpU

dg

blpL

blpU

dg

dg

ISSpn1

com

A

blpB

∗

blpY

ISSth1orfB orfA’

blpA

blpB

#

blpC

dg

blpR

blpH

blpK

dg

’blp

X

orf6

ISSth1orfB orfA

blpU

dg

blpA

blpB

#

blpC

dg

blpR

blpH

orf5

orf3

blpK

dg#

blpG

F

orf4

orf4

orf2

blpD

dg

orf1

orf6

orf3

orf5

blpA

blpB

blpC

dg

orf7

orf8

blpR

blpH

blpU

dg

blpE

dg

blpF

dg

blpG

1kb

(a)

(b)

(c)

(d)

(e)

orf2

orf1

Fig. 6. Genomic organisation of the blp loci in S. thermophilus LMD9 (a), LM

S. pneumoniae TIGR4 (sp0041 to sp0043; sp0524 to sp0546) (d), and S. m

smu.1917) (e). Genes encoding peptides of similar predicted functions are rep

protein (orange), induction factor (yellow), response regulator (middle g

hydrophobic peptide of unknown function (black), hydrophilic peptide of un

peptides with a leader of the double-glycine type are indicated with dg (su

nonsense mutations leading to premature translation termination compared t

open reading frame of blpB contains an internal sequence deletion compared

and orf8 open reading frames in LMG18311, compared to LMD9 (missense

protein (blpB) involved in bacteriocin transport and

processing, and a typical quorum sensing regulatory

module composed of a two-component system (blpH

and blpR; TCS09 in S. thermophilus) and the corre-

sponding bacteriocin-like inducer peptide precursor

(blpC) that contains the typical double-Gly cleavage site(Fig. 6; for a review see [175]). However, the S. pneumo-

niae blpA gene appears to contain several nonsense

mutations, while the blpB gene of S. thermophilus

LMG18311 and CNRZ1066 also contains such muta-

tions, suggesting that these systems are not functional.

Only S. thermophilus LMD9 seems to contain functional

blpA and blpB genes. Interestingly, these streptococci

display a high variation with regard to the bacteriocinencoding complement associated with these general blp

factors (Fig. 6). Microarray technology revealed that

addition of the mature BlpC in the culture medium of

S. pneumoniae TIGR4 induces transcription of genes

encoding small bacteriocin-like peptides with the typical

double-Gly leader sequence and some putative corre-

sponding immunity genes (blpI, blpJ, blpK, blpM, blpN,

blpO, blpP, blpU) [190]. However, to the best of ourknowledge, Blp-associated bacteriocin production by

orf1

blpK

dg

blpO

dg

blpL

blpM

dg

blpN

dg

orf2

orf3

orf4

blpX

blpY

blpZ

blpA

#

blpI

dg

blpJ

dg

IS1381orfA orfB

blpA

#

blpB

∗

blpC

dg

com

E

com

D

blpJ

blpU

dg

blpU

dg

orf4

orf6

orf7

blpO

dg

com

Cd

g

blpL

orf1

0or

f9

blpQ

blpX

orf9

blpX

blpQ

orf1

0

orf5

orf3

orf8

G18311 (stu1691 to stu1673) (b), CNRZ1066 (str1691 to str1673) (c),

utans UA159 (smu.286 to smu.287; smu.789; smu.925; smu.1889c to

resented by colored arrows: ABC-transporter (red), transport accessory

rey), histidine kinase (dark grey), bacteriocin-like peptide (green),

known function (white), and insertion sequence (blue). Genes encoding

perscript). blpB#, blpA# and blpK# indicates that these genes contain

o the blpB, blpA, and blpK genes, respectively. blpB* indicates that the

to the blpB gene of S. pneumoniae. blpGF indicates the fusion of blpG

mutation).

P. Hols et al. / FEMS Microbiology Reviews 29 (2005) 435–463 457

S. pneumoniae has never been established, which is in

agreement with the non-functional blpA gene. The blp

loci of the S. thermophilus strains also contain genes

encoding bacteriocin-like peptides; e.g., blpD, blpU,

blpE, blpF in LMD9, blpU, blpK in LMG18311, and

blpK in CNRZ1066.In order to validate the role of the S. thermophilus

TCS09 regulatory module in the regulation of blp-bacte-

riocin production, the corresponding inducer peptide

precursor BlpC was overproduced in the three S. ther-

mophilus strains. The S. thermophilus blpC genes are al-

most identical and only differ with regard to the last

amino acid of the mature region (downstream the dou-

ble-Gly motif; Ala as C-terminal residue in LMD9,and Val in the LMG18311 and CNRZ1066 strains).

These strain-specific blpC genes were cloned in expres-

sion vectors and introduced in LMD9, LMG18311,

and CNRZ1066. Bacteriocin detection by agar-well dif-

fusion assays revealed that both blpC expression vectors

induced production of active antimicrobial compounds

in LMD9 that inhibit growth of both LMG18311 and

CNRZ1066, while, as expected, LMD9 was resistantto this antimicrobial activity. In contrast, blpC overex-

pression did not induce bacteriocin production in either

strain LMG18311 or CNRZ1066. Moreover, Northern

blot analyses showed blpC overexpression dependent

transcription of blpD, blpU, blpE, and blpF in S. thermo-

philus LMD9, while induction of blpU and blpK tran-

scription could not be detected in LMG18311 or

CNRZ1066 irrespective of blpC overexpression. Thislack of blpC overexpression mediated bacteriocin induc-

tion in the latter two S. thermophilus strains is in good

agreement with their non-functional blpB gene. Taken

together, these preliminary results establish functional-

ity of the blp-bacteriocin locus in S. thermophilus

LMD9 and support the proposed role in its regulation

of TCS09 (blpH and blpR) (Fontaine, L., Grossiord,

B., Hols; P., unpublished data).

8. Concluding remarks

The access to three genome sequences of S. thermo-

philus has allowed a better understanding of the evolu-

tionary path followed by this species that belongs to a

genus that encompasses many harmful pathogenic spe-cies. The S. thermophilus species, which is closely related

to S. salivarius found in the human oral cavity, is rela-

tively coherent and homogenous with a low level of

nucleotide polymorphism, suggesting that this species

has recently emerged. However, the presence of numer-

ous pseudogenes also suggests an ongoing regressive

evolution process towards a specialised bacterium dedi-

cated to growth in milk. The detailed in silico investiga-tion of its cellular metabolism and its regulation as

reported here illustrates that evolution has shaped the

S. thermophilus genome by selection for optimal growth

in this well-defined ecological niche. Notably, S. thermo-

philus has maintained a well-developed nitrogen metab-

olism whereas its sugar catabolism has strongly

degenerated. Additionally, S. thermophilus shares its

ecological niche with other LAB such as Lb. bulgaricus,resulting in specific metabolic cooperation, which is

either revealed by the maintenance of dedicated path-

ways (e.g., folate and formate production) or by the loss

of key metabolic functions provided by the symbiotic

partner (e.g., casein hydrolysis). Furthermore, S. ther-

mophilus has lost many virulence related functions com-

mon among pathogenic streptococci that play important

roles in cell adhesion, host invasion or escape from itsimmune system. Although gene decay is obvious in the

S. thermophilus genome, numerous small genomic is-

lands seem to have been acquired by LGT. These re-

gions encode a number of important industrial

phenotypic traits such as polysaccharide biosynthesis

(eps, rgp), bacteriocin production (blp, lab), restric-

tion–modification systems or oxygen tolerance (TAT

pathway, Rgg regulators). Overall, its restricted ecolog-ical niche and its corresponding adaptive evolution can

most likely explain the very rapid growth of S. thermo-

philus in milk.

The wealth of genomic information will not only aid

our understanding of the molecular biology and physiol-

ogy of S. thermophilus, but will also facilitate selection

of appropriate S. thermophilus starters by the food

industry, as well as enhance our insight in pathogenicprocesses involving streptococci. Moreover, the preli-

minary metabolic engineering studies described here

provide a first glimpse of what future mutagenesis and

selection efforts might contribute to a next generation

of S. thermophilus starter cultures providing enhanced

flavour characteristics to fermentation products. The

development of post-sequencing genomics technology

and its application will undoubtedly accelerate our func-tional understanding of S. thermophilus. The prelimin-

ary results obtained by DNA microarray-based

transcriptome profiling of the S. thermophilus wild-type

and its ccpA derivative, convincingly show that molecu-

lar biology research related to this LAB is entering a

new dimension, which provides exciting opportunities

to both academic and industrial research.

Acknowledgements

The S. thermophilus LMG18311 chromsome se-

quence was supported by funding from the Wallon

Region (Bioval No. 981/3866, 3845 and First Europe

No. EPH3310300R0082) and FNRS (Grant No.

2.4586.02). P. Hols is Research Associate at FNRS. L.Fontaine holds a fellowship of the ‘‘Fonds pour la

458 P. Hols et al. / FEMS Microbiology Reviews 29 (2005) 435–463

Formation a la Recherche dans l�Industrie et dans

l�Agriculture’’ (FRIA).

Appendix A. Supplementary data

Supplementary data associated with this article can

be found, in the online version at doi:10.1016/

j.femsre.2005.04.008.

References

[1] Chausson, F. and Maurisson, E. (2002) L�economie Laitiere En

Chiffres. Centre National Inter-Professionnel de l�Economie

Laitiere, Paris, France.

[2] Fox, P.F. (1993) Cheese: Chemistry, Physics and Microbiology.

Chapman & Hall, London, UK.

[3] Rasic, J.L. and Kurmann, J.A. (1978). Fermented Fresk Milk

Products. Yogurt: Scientific Grounds, Technology, Manufacture

and Preparations, vol. 1. Technical Dairy Publishing House,

Copenhagen, DK.

[4] Tamime, A.Y. and Deeth, H.C. (1980) Yogurt: technology and

biochemistry. J. Food Protect. 43, 939–977.

[5] Mitchell, T.J. (2003) The pathogenesis of streptococcal infec-

tions: from tooth decay to meningitis. Nat. Rev. Microbiol. 1,

219–230.

[6] Tettelin, H. (2004) Streptococcal genomes provide food for

thought. Nat. Biotechnol. 22, 1523–1524.

[7] Bolotin, A., Quinquis, B., Renault, P., Sorokin, A., Ehrlich,

S.D., Kulakauskas, S., Lapidus, A., Goltsman, E., Mazur, M.,

Pusch, G.D., Fonstein, M., Overbeek, R., Kyprides, N., Purn-

elle, B., Prozzi, D., Ngui, K., Masuy, D., Hancy, F., Burteau, S.,

Boutry, M., Delcour, J., Goffeau, A. and Hols, P. (2004)

Complete sequence and comparative genome analysis of the

dairy bacterium Streptococcus thermophilus. Nat. Biotechnol. 22,

1554–1558.

[8] Ajdic, D., McShan, W.M., McLaughlin, R.E., Savic, G., Chang,

J., Carson, M.B., Primeaux, C., Tian, R., Kenton, S., Jia, H.,

Lin, S., Qian, Y., Li, S., Zhu, H., Najar, F., Lai, H., White, J.,

Roe, B.A. and Ferretti, J.J. (2002) Genome sequence of

Streptococcus mutans UA159, a cariogenic dental pathogen.

Proc. Natl. Acad. Sci. USA 99, 14434–14439.

[9] Ferretti, J.J., McShan, W.M., Ajdic, D., Savic, D.J., Savic, G.,

Lyon, K., Primeaux, C., Sezate, S., Suvorov, A.N., Kenton, S.,

Lai, H.S., Lin, S.P., Qian, Y., Jia, H.G., Najar, F.Z., Ren, Q.,

Zhu, H., Song, L., White, J., Yuan, X., Clifton, S.W., Roe, B.A.

and McLaughlin, R. (2001) Complete genome sequence of an

M1 strain of Streptococcus pyogenes. Proc. Natl. Acad. Sci. USA

98, 4658–4663.

[10] Glaser, P., Rusniok, C., Buchrieser, C., Chevalier, F., Frangeul,

L., Msadek, T., Zouine, M., Couve, E., Lalioui, L., Poyart, C.,

Trieu-Cuot, P. and Kunst, F. (2002) Genome sequence of

Streptococcus agalactiae, a pathogen causing invasive neonatal

disease. Mol. Microbiol. 45, 1499–1513.

[11] Hoskins, J., Alborn Jr., W.E., Arnold, J., Blaszczak, L.C.,

Burgett, S., DeHoff, B.S., Estrem, S.T., Fritz, L., Fu, D.J.,

Fuller, W., Geringer, C., Gilmour, R., Glass, J.S., Khoja, H.,

Kraft, A.R., Lagace, R.E., Leblanc, D.J., Lee, L.N., Lefkowitz,

E.J., Lu, J., Matsushima, P., McAhren, S.M., McHenney, M.,

McLeaster, K., Mundy, C.W., Nicas, T.I., Norris, F.H., O�Gara,

M., Peery, R.B., Robertson, G.T., Rockey, P., Sun, P.M.,

Winkler, M.E., Yang, Y., Young-Bellido, M., Zhao, G., Zook,

C.A., Baltz, R.H., Jaskunas, S.R., Rosteck Jr., P.R., Skatrud,

P.L. and Glass, J.I. (2001) Genome of the bacterium Strepto-

coccus pneumoniae strain R6. J. Bacteriol. 183, 5709–5717.

[12] Tettelin, H., Nelson, K.E., Paulsen, I.T., Eisen, J.A., Read,

T.D., Peterson, S., Heidelberg, J., DeBoy, R.T., Haft, D.H.,

Dodson, R.J., Durkin, A.S., Gwinn, M., Kolonay, J.F.,

Nelson, W.C., Peterson, J.D., Umayam, L.A., White, O.,

Salzberg, S.L., Lewis, M.R., Radune, D., Holtzapple, E.,

Khouri, H., Wolf, A.M., Utterback, T.R., Hansen, C.L.,

McDonald, L.A., Feldblyum, T.V., Angiuoli, S., Dickinson,

T., Hickey, E.K., Holt, I.E., Loftus, B.J., Yang, F., Smith,

H.O., Venter, J.C., Dougherty, B.A., Morrison, D.A., Hol-

lingshead, S.K. and Fraser, C.M. (2001) Complete genome

sequence of a virulent isolate of Streptococcus pneumoniae.

Science 293, 498–506.

[13] Tettelin, H., Masignani, V., Cieslewicz, M.J., Eisen, J.A.,

Peterson, S., Wessels, M.R., Paulsen, I.T., Nelson, K.E.,

Margarit, I., Read, T.D., Madoff, L.C., Wolf, A.M., Beanan,

M.J., Brinkac, L.M., Daugherty, S.C., DeBoy, R.T., Durkin,

A.S., Kolonay, J.F., Madupu, R., Lewis, M.R., Radune, D.,

Fedorova, N.B., Scanlan, D., Khouri, H., Mulligan, S., Carty,

H.A., Cline, R.T., Van Aken, S.E., Gill, J., Scarselli, M., Mora,

M., Iacobini, E.T., Brettoni, C., Galli, G., Mariani, M., Vegni,

F., Maione, D., Rinaudo, D., Rappuoli, R., Telford, J.L.,

Kasper, D.L., Grandi, G. and Fraser, C.M. (2002) Complete

genome sequence and comparative genomic analysis of an

emerging human pathogen, serotype V Streptococcus agalactiae.

Proc. Natl. Acad. Sci. USA 99, 12391–12396.

[14] Facklam, R. (2002) What happened to the streptococci: overview

of taxonomic and nomenclature changes. Clin. Microbiol. Rev.

15, 613–630.

[15] Idigoras, P., Valiente, A., Iglesias, L., Trieu-Cout, P. and Poyart,

C. (2001) Meningitis due to Streptococcus salivarius. J. Clin.

Microbiol. 39, 3017.

[16] Kawamura, Y., Hou, X.G., Sultana, F., Miura, H. and Ezaki, T.

(1995) Determination of 16S rRNA sequences of Streptococcus

mitis and Streptococcus gordonii and phylogenetic relationships

among members of the genus Streptococcus. Int. J. Syst.

Bacteriol. 45, 406–408.

[17] Poyart, C., Quesne, G., Coulon, S., Berche, P. and Trieu-Cuot,

P. (1998) Identification of streptococci to species level by

sequencing the gene encoding the manganese-dependent super-

oxide dismutase. J. Clin. Microbiol. 36, 41–47.

[18] Schleifer, K.H., Ehrmann, M., Krusch, U. and Neve, H. (1991)

Revival of the Species Streptococcus thermophilus (Ex Orla-

Jensen, 1919). Nom Rev. Syst. Appl. Microbiol. 14, 386–388.

[19] Bolotin, A., Wincker, P., Mauger, S., Jaillon, O., Malarme, K.,

Weissenbach, J., Ehrlich, S.D. and Sorokin, A. (2001) The

complete genome sequence of the lactic acid bacterium Lacto-

coccus lactis ssp. lactis IL1403. Genome Res. 11, 731–753.

[20] Kleerebezem, M., Boekhorst, J., van Kranenburg, R., Molenaar,

D., Kuipers, O.P., Leer, R., Tarchini, R., Peters, S.A., Sand-

brink, H.M., Fiers, M.W., Stiekema, W., Lankhorst, R.M.,

Bron, P.A., Hoffer, S.M., Groot, M.N., Kerkhoven, R., de Vries,

M., Ursing, B., de Vos, W.M. and Siezen, R.J. (2003) Complete

genome sequence of Lactobacillus plantarum WCFS1. Proc.

Natl. Acad. Sci. USA 100, 1990–1995.

[21] Ochman, H., Elwyn, S. and Moran, N.A. (1999) Calibrating

bacterial evolution. Proc. Natl. Acad. Sci. USA 96, 12638–

12643.

[22] Dobrindt, U. and Hacker, J. (2001) Whole genome plasticity in

pathogenic bacteria. Curr. Opin. Microbiol. 4, 550–557.

[23] Giraud, A., Matic, I., Tenaillon, O., Clara, A., Radman, M.,

Fons, M. and Taddei, F. (2001) Costs and benefits of high

mutation rates: adaptive evolution of bacteria in the mouse gut.

Science 291, 2606–2608.

[24] Enright, M.C. and Spratt, B.G. (1999) Multilocus sequence

typing. Trends Microbiol. 7, 482–487.

P. Hols et al. / FEMS Microbiology Reviews 29 (2005) 435–463 459

[25] Delcour, J., Ferain, T., Deghorain, M., Palumbo, E. and Hols,

P. (1999) The biosynthesis and functionality of the cell-wall of

lactic acid bacteria. Anton. Leeuw. 76, 159–184.

[26] Schleifer, K.H. and Kandler, O. (1972) Peptidoglycan types of

bacterial cell walls and their taxonomic implications. Bacteriol.

Rev. 36, 407–477.

[27] Bouhss, A., Josseaume, N., Severin, A., Tabei, K., Hugonnet,

J.E., Shlaes, D., Mengin-Lecreulx, D., Van Heijenoort, J. and

Arthur, M. (2002) Synthesis of the L-alanyl-L-alanine cross-

bridge of Enterococcus faecalis peptidoglycan. J. Biol. Chem.

277, 45935–45941.

[28] Vollmer, W. and Tomasz, A. (2000) The pgdA gene encodes for a

peptidoglycan N-acetylglucosamine deacetylase in Streptococcus

pneumoniae. J. Biol. Chem. 275, 20496–20501.

[29] Kruse, T., Bork-Jensen, J. and Gerdes, K. (2005) The morpho-

genetic MreBCD proteins of Escherichia coli form an essential

membrane-bound complex. Mol. Microbiol. 55, 78–89.

[30] Lee, J.C. and Stewart, G.C. (2003) Essential nature of the mreC

determinant of Bacillus subtilis. J. Bacteriol. 185, 4490–4498.

[31] Stingele, F. and Mollet, B. (1996) Disruption of the gene

encoding penicillin-binding protein 2b (pbp2b) causes altered cell

morphology and cease in exopolysaccharide production in

Streptococcus thermophilus Sfi6. Mol. Microbiol. 22, 357–366.

[32] Thibessard, A., Fernandez, A., Gintz, B., Leblond-Bourget, N.

and Decaris, B. (2002) Effects of rodA and pbp2b disruption on

cell morphology and oxidative stress response of Streptococcus

thermophilus CNRZ368. J. Bacteriol. 184, 2821–2826.

[33] Thibessard, A., Borges, F., Fernandez, A., Gintz, B., Decaris, B.

and Leblond-Bourget, N. (2004) Identification of Streptococcus

thermophilus CNRZ368 genes involved in defense against

superoxide stress. Appl. Environ. Microbiol. 70, 2220–2229.

[34] Borges, F., Severine, L., Thibessard, A., Fernandez, A., Brigitte,

G., Hols, P., Decaris, B. and Leblond-Bourget, N. (2005) cse, a

chimeric and variable gene, encodes an extracellular protein

involved in cellular segregation in Streptococcus thermophilus. J.

Bacteriol. 187, 2737–2746.

[35] Ng, W.L., Kazmierczak, K.M. and Winkler, M.E. (2004)

Defective cell wall synthesis in Streptococcus pneumoniae R6

depleted for the essential PcsB putative murein hydrolase or the

VicR (YycF) response regulator. Mol. Microbiol. 53, 1161–1175.

[36] Pagliero, E., Dideberg, O., Vernet, T. and Di Guilmi, A.M.

(2005) The PECACE domain: a new family of enzymes with

potential peptidoglycan cleavage activity in Gram-positive bac-

teria. BMC Genomics 6, 19.

[37] Smith, T.J., Blackman, S.A. and Foster, S.J. (2000) Autolysins

of Bacillus subtilis: multiple enzymes with multiple functions.

Microbiology 146, 249–262.

[38] Husson-Kao, C., Mengaud, J., Benbadis, L. and Chapot-

Chartier, M.P. (2000) Mur1, a Streptococcus thermophilus

peptidoglycan hydrolase devoid of a specific cell wall binding

domain. FEMS Microbiol. Lett. 187, 69–76.

[39] Bateman, A. and Rawlings, N.D. (2003) The CHAP domain: a

large family of amidases including GSP amidase and peptido-

glycan hydrolases. Trends Biochem. Sci. 28, 234–237.

[40] Fischer, W. (1988) Physiology of lipoteichoic acids in bacteria.

Adv. Microb. Physiol. 29, 233–302.

[41] Fischer, W., Mannsfeld, T. and Hagen, G. (1990) On the basic

structure of poly(glycerophosphate) lipoteichoic acids. Biochem.

Cell Biol. 68, 33–43.

[42] Hogg, S.D., Whiley, R.A. and De Soet, J.J. (1997) Occurrence of

lipoteichoic acid in oral streptococci. Int. J. Syst. Bacteriol. 47,

62–66.

[43] Lazarevic, V., Abellan, F.X., Moller, S.B., Karamata, D. and

Mauel, C. (2002) Comparison of ribitol and glycerol teichoic

acid genes in Bacillus subtilis W23 and 168: identical function,

similar divergent organization, but different regulation. Micro-

biology 148, 815–824.

[44] Neuhaus, F.C. and Baddiley, J. (2003) A continuum of anionic

charge: structures and functions of D-alanyl-teichoic acids in

gram-positive bacteria. Microbiol. Mol. Biol. Rev. 67, 686–723.

[45] Cerning, J. (1995) Production of exopolysaccharides by lactic

acid bacteria and dairy propionibacteria. Lait 75, 463–472.

[46] Hassan, A.N., Frank, J.F., Schmidt, K.A. and Shalab, S.I.

(1996) Rheological properties of yogurt made with encapsulated

nonropy lactic cultures. J. Dairy Sci. 79, 2091–2097.

[47] De Vuyst, L., De Vin, F., Vaningelgem, F. and Degeest, B.

(2001) Recent developments in the biosynthesis and applications

of heteropolysaccharides from lactic acid bacteria. Int. Dairy J.

11, 687–707.

[48] Laws, A.P. and Marshall, V.M. (2001) The relevance of

exopolysaccharides to the rheological properties in milk fer-

mented with ropy strains of lactic acid bacteria. Int. Dairy J. 11,

709–721.

[49] Duboc, P. and Mollet, B. (2001) Applications of exopolysac-

charides in the dairy industry. Int. Dairy J. 11, 759–768.

[50] Laws, A., Gu, Y.C. and Marshall, V. (2001) Biosynthesis,

characterisation, and design of bacterial exopolysaccharides

from lactic acid bacteria. Biotechnol. Adv. 19, 597–625.

[51] Broadbent, J.R., McMahon, D.J., Welker, D.L., Oberg, C.J. and

Moineau, S. (2003) Biochemistry, genetics, and applications of

exopolysaccharide production in Streptococcus thermophilus: a

review. J. Dairy Sci. 86, 407–423.

[52] Rallu, F., Taillez, P., Ehrlich, S.D. and Renault, P. (2002)

Common scheme of evolution between eps clusters of the food

bacteria Streptococcus thermophilus and cps clusters of the

pathogenic streptococci. In: Proceedings of the 6th American

Society of Microbiology Conference on Streptococcal Genetics,

Asheville, NC. p. 112..

[53] Stingele, F., Neeser, J.R. and Mollet, B. (1996) Identification and

characterization of the eps (Exopolysaccharide) gene cluster

from Streptococcus thermophilus Sfi6. J. Bacteriol. 178, 1680–

1690.

[54] Bender, M.H. and Yother, J. (2001) CpsB is a modulator of

capsule-associated tyrosine kinase activity in Streptococcus

pneumoniae. J. Biol. Chem. 276, 47966–47974.

[55] Morona, J.K., Morona, R., Miller, D.C. and Paton, J.C. (2003)

Mutational analysis of the carboxy-terminal (YGX)4 repeat

domain of CpsD, an autophosphorylating tyrosine kinase

required for capsule biosynthesis in Streptococcus pneumoniae.

J. Bacteriol. 185, 3009–3019.

[56] Bender, M.H., Cartee, R.T. and Yother, J. (2003) Positive

correlation between tyrosine phosphorylation of CpsD and

capsular polysaccharide production in Streptococcus pneumo-

niae. J. Bacteriol. 185, 6057–6066.

[57] Levander, F., Svensson, M. and Radstrom, P. (2002) Enhanced

exopolysaccharide production by metabolic engineering of

Streptococcus thermophilus. Appl. Environ. Microbiol. 68, 784–

790.

[58] Cerning, J. (1990) Exocellular polysaccharides produced by

lactic acid bacteria. FEMS Microbiol. Rev. 7, 113–130.

[59] Germond, J.E., Delley, M., D�Amico, N. and Vincent, S.J.

(2001) Heterologous expression and characterization of the

exopolysaccharide from Streptococcus thermophilus Sfi39. Eur. J.

Biochem. 268, 5149–5156.

[60] Krinos, C.M., Coyne, M.J., Weinacht, K.G., Tzianabos, A.O.,

Kasper, D.L. and Comstock, L.E. (2001) Extensive surface

diversity of a commensal microorganism by multiple DNA

inversions. Nature 414, 555–558.

[61] Yamashita, Y., Tsukioka, Y., Tomihisa, K., Nakano, Y. and

Koga, T. (1998) Genes involved in cell wall localization and side

chain formation of rhamnose–glucose polysaccharide in Strep-

tococcus mutans. J. Bacteriol. 180, 5803–5807.

[62] Shibata, Y., Ozaki, K., Seki, M., Kawato, T., Tanaka, H.,

Nakano, Y. and Yamashita, Y. (2003) Analysis of loci required

460 P. Hols et al. / FEMS Microbiology Reviews 29 (2005) 435–463

for determination of serotype antigenicity in Streptococcus

mutans and its clinical utilization. J. Clin. Microbiol. 41, 4107–

4112.

[63] Dalbey, R.E. and Chen, M. (2004) Sec-translocase mediated

membrane protein biogenesis. Biochim. Biophys. Acta 1694, 37–

53.

[64] Dilks, K., Rose, R.W., Hartmann, E. and Pohlschroder, M.

(2003) Prokaryotic utilization of the twin-arginine translocation

pathway: a genomic survey. J. Bacteriol. 185, 1478–1483.

[65] Comfort, D. and Clubb, R.T. (2004) A comparative genome

analysis identifies distinct sorting pathways in gram-positive

bacteria. Infect. Immun. 72, 2710–2722.

[66] Fernandez-Espla, M.D., Garault, P., Monnet, V. and Rul, F.

(2000) Streptococcus thermophilus cell wall-anchored proteinase:

release, purification, and biochemical and genetic characteriza-

tion. Appl. Environ. Microbiol. 66, 4772–4778.

[67] Roos, S. and Jonsson, H. (2002) A high-molecular-mass cell-

surface protein from Lactobacillus reuteri 1063 adheres to mucus

components. Microbiology 148, 433–442.

[68] Fernandez, A., Thibessard, A., Borges, F., Gintz, B., Decaris, B.

and Leblond-Bourget, N. (2004) Characterization of oxidative

stress-resistant mutants of Streptococcus thermophilus

CNRZ368. Arch. Microbiol. 182, 364–372.

[69] Paulsen, I.T., Nguyen, L., Sliwinski, M.K., Rabus, R. and Saier

Jr., M.H. (2000) Microbial genome analyses: comparative

transport capabilities in eighteen prokaryotes. J. Mol. Biol.

301, 75–100.

[70] Vaillancourt, K., Moineau, S., Frenette, M., Lessard, C. and

Vadeboncoeur, C. (2002) Galactose and lactose genes from

the galactose-positive bacterium Streptococcus salivarius and

the phylogenetically related galactose-negative bacterium

Streptococcus thermophilus: organization, sequence, transcrip-

tion, and activity of the gal gene products. J. Bacteriol. 184,

785–793.

[71] van den Bogaard, P.T., Hols, P., Kuipers, O.P., Kleerebezem, M.

and de Vos, W.M. (2004) Sugar utilisation and conservation of

the gal-lac gene cluster in Streptococcus thermophilus. Syst. Appl.

Microbiol. 27, 10–17.

[72] Vaughan, E.E., van den Bogaard, P.T., Catzeddu, P., Kuipers,

O.P. and de Vos, W.M. (2001) Activation of silent gal genes in

the lac-gal regulon of Streptococcus thermophilus. J. Bacteriol.

183, 1184–1194.

[73] Gunnewijk, M.G., Postma, P.W. and Poolman, B. (1999)

Phosphorylation and functional properties of the IIA domain

of the lactose transport protein of Streptococcus thermophilus. J.

Bacteriol. 181, 632–641.

[74] Gunnewijk, M.G. and Poolman, B. (2000) HPr(His approxi-

mately P)-mediated phosphorylation differently affects counter-

flow and proton motive force-driven uptake via the lactose

transport protein of Streptococcus thermophilus. J. Biol. Chem.

275, 34080–34085.

[75] Gunnewijk, M.G. and Poolman, B. (2000) Phosphorylation

state of HPr determines the level of expression and the

extent of phosphorylation of the lactose transport protein

of Streptococcus thermophilus. J. Biol. Chem. 275, 34073–

34079.

[76] Gunnewijk, M.G.W., van den Bogaard, P.T.C., Veenhoff, L.M.,

Heuberger, E.H., de Vos, W.M., Kleerebezem, M., Kuipers, O.P.

and Poolman, B. (2001) Hierarchical control versus autoregula-

tion of carbohydrate utilization in bacteria. J. Mol. Microbiol.

Biotechnol. 3, 401–413.

[77] van den Bogaard, P.T., Kleerebezem, M., Kuipers, O.P. and de

Vos, W.M. (2000) Control of lactose transport, beta-galactosi-

dase activity, and glycolysis by CcpA in Streptococcus thermo-

philus: evidence for carbon catabolite repression by a non-

phosphoenolpyruvate-dependent phosphotransferase system

sugar. J. Bacteriol. 182, 5982–5989.

[78] Ott, A., Germond, J.E. and Chaintreau, A. (2000) Origin of

acetaldehyde during milk fermentation using (13)C-labeled

precursors. J. Agric. Food Chem. 48, 1512–1517.

[79] Teraguchi, S., Ono, J., Kiyosawa, I. and Okonogi, S. (1987)

Oxygen uptake activity and aerobic metabolism of Streptococcus

thermophilus STH450. J. Dairy Sci. 70, 514–523.

[80] Tinson, W., Hillier, A.J. and Jago, G.R. (1982) Metabolism of

Streptococcus thermophilus. 4. Utilisation of pyruvate and

citrate. Aust. J. Dairy Technol. 37, 22–25.

[81] Hayacibara, M.F., Koo, H., Vacca-Smith, A.M., Kopec, L.K.,

Scott-Anne, K., Cury, J.A. and Bowen, W.H. (2004) The

influence of mutanase and dextranase on the production and

structure of glucans synthesized by streptococcal glucosyltransfe-

rases. Carbohydr. Res. 339, 2127–2137.

[82] Lawman, P. and Bleiweis, A.S. (1991) Molecular cloning of the

extracellular endodextranase of Streptococcus salivarius. J. Bac-

teriol. 173, 7423–7428.

[83] Eifuku, H., Yoshimitsu-Narita, A., Sato, S., Yakushiji, T. and

Inoue, M. (1989) Production and partial characterization of the

extracellular polysaccharides from oral Streptococcus salivarius.

Carbohydr. Res. 194, 247–260.

[84] Rathsam, C., Giffard, P.M. and Jacques, N.A. (1993) The cell-

bound fructosyltransferase of Streptococcus salivarius: the car-

boxyl terminus specifies attachment in a Streptococcus gordonii

model system. J. Bacteriol. 175, 4520–4527.

[85] Slos, P., Bourquin, J.C., Lemoine, Y. and Mercenier, A. (1991)

Isolation and characterization of chromosomal promoters of

Streptococcus salivarius subsp. thermophilus. Appl. Environ.

Microbiol. 57, 1333–1339.

[86] Poolman, B. (1993) Energy transduction in lactic acid bacteria.

FEMS Microbiol. Rev. 12, 125–147.

[87] Vaillancourt, K., Lemay, J.D., Lamoureux, M., Frenette, M.,

Moineau, S. and Vadeboncoeur, C. (2004) Characterization of a

galactokinase-positive recombinant strain of Streptococcus ther-

mophilus. Appl. Environ. Microbiol. 70, 4596–4603.

[88] Cochu, A., Vadeboncoeur, C., Moineau, S. and Frenette, M.

(2003) Genetic and biochemical characterization of the phos-

phoenolpyruvate:glucose/mannose phosphotransferase system of

Streptococcus thermophilus. Appl. Environ. Microbiol. 69, 5423–

5432.

[89] Levander, F. and Radstrom, P. (2001) Requirement for phos-

phoglucomutase in exopolysaccharide biosynthesis in glucose-

and lactose-utilizing Streptococcus thermophilus. Appl. Environ.

Microbiol. 67, 2734–2738.

[90] Degeest, B., Vaningelgem, F., Laws, A.P. and De Vuyst, L.

(2001) UDP-N-acetylglucosamine 4-epimerase activity indicates

the presence of N-acetylgalactosamine in exopolysaccharides of

Streptococcus thermophilus strains. Appl. Environ. Microbiol.

67, 3976–3984.

[91] Perez, P.F., Deantoni, G.L. and Anon, M.C. (1991) Formate

production by Streptococcus thermophilus cultures. J. Dairy Sci.

74, 2850–2854.

[92] Juillard, V., Spinnler, H.E., Desmazeaud, M.J. and Boquien,

C.Y. (1987) Cooperation and inhibition between lactic-acid

bacteria used in dairy-industry. Lait 67, 149–172.

[93] Chaves, A.C., Fernandez, M., Lerayer, A.L., Mierau, I.,

Kleerebezem, M. and Hugenholtz, J. (2002) Metabolic engineer-

ing of acetaldehyde production by Streptococcus thermophilus.

Appl. Environ. Microbiol. 68, 5656–5662.

[94] Ali, N.O., Bignon, J., Rapoport, G. and Debarbouille, M. (2001)

Regulation of the acetoin catabolic pathway is controlled by

sigma L in Bacillus subtilis. J. Bacteriol. 183, 2497–2504.

[95] Hillman, J.D., Chen, A. and Snoep, J.L. (1996) Genetic and

physiological analysis of the lethal effect of L-(+)-lactate dehy-

drogenase deficiency in Streptococcus mutans: complementation

by alcohol dehydrogenase from Zymomonas mobilis. Infect.

Immun. 64, 4319–4323.

P. Hols et al. / FEMS Microbiology Reviews 29 (2005) 435–463 461

[96] Hugenholtz, J., Kleerebezem, M., Starrenburg, M., Delcour, J.,

de Vos, W. and Hols, P. (2000) Lactococcus lactis as a cell

factory for high-level diacetyl production. Appl. Environ.

Microbiol. 66, 4112–4114.

[97] de Felipe, Lopez, Kleerebezem, M., de Vos, W.M. and Huge-

nholtz, J. (1998) Cofactor engineering: a novel approach to

metabolic engineering in Lactococcus lactis by controlled

expression of NADH oxidase. J. Bacteriol. 180, 3804–3808.

[98] Bongers, R.S., Hoefnagel, M.H. and Kleerebezem, M. (2005)

High-level acetaldehyde production in Lactococcus lactis by

metabolic engineering 1. Appl. Environ. Microbiol. 71, 1109–

1113.

[99] Ercolini, D., Fusco, V., Blaiotta, G. and Coppola, S. (2005)

Sequence heterogeneity in the lacSZ operon of Streptococcus

thermophilus and its use in PCR systems for strain differentia-

tion. Res. Microbiol. 156, 161–172.

[100] Ajdic, D. and Ferretti, J.J. (1998) Transcriptional regulation of

the Streptococcus mutans gal operon by the GalR repressor. J.

Bacteriol. 180, 5727–5732.

[101] Titgemeyer, F. and Hillen, W. (2002) Global control of sugar

metabolism: a gram-positive solution. Anton. Leeuwen. 82, 59–

71.

[102] van den Bogaard, P.T.C., Crispie, F., Hols, P., Kuipers, O.P.,

Kleerebezem, M., and de Vos, W.M. (2002) Modulation of

glycolysis by lactose availability in Streptococcus thermophilus.

In: Catabolite Control of Sugar Metabolism in Streptococcus

thermophilus (van den Bogaard, P.T.C., Ed.). pp. 87-103. Thesis

Wageningen University, Publishers Ponsen en Looijen BV,

Wageningen, Nl..

[103] Asanuma, N., Yoshii, T. and Hino, T. (2004) Molecular

characterization of CcpA and involvement of this protein in

transcriptional regulation of lactate dehydrogenase and pyruvate

formate-lyase in the ruminal bacterium Streptococcus bovis.

Appl. Environ. Microbiol. 70, 5244–5251.

[104] Luesink, E.J., van Herpen, R.E., Grossiord, B.P., Kuipers, O.P.

and de Vos, W.M. (1998) Transcriptional activation of the

glycolytic las operon and catabolite repression of the gal operon

in Lactococcus lactis are mediated by the catabolite control

protein CcpA. Mol. Microbiol. 30, 789–798.

[105] van den Bogaard, P.T.C., Kleerebezem, M., Kuipers, P. and de

Vos, W.M. (2002) Determination of glycolytic intermediates in

Streptococcus thermophilus and their effect on metabolite control

of sugar utilization. In: Catabolite Control of Sugar Metabolism

in Streptococcus thermophilus (van den Bogaard, P.T.C., Ed.).

pp. 73–86. Thesis Wageningen University, Publishers Ponsen en

Looijen BV, Wageningen, Nl..

[106] Cochu, A., Roy, D., Vaillancourt, K., Lemay, J.D., Casabon, I.,

Frenette, M., Moineau, S. and Vadeboncoeur, C. (2005) The

doubly phosphorylated form of HPr, HPr(Ser-P)(His � P), is

abundant in exponentially growing cells of Streptococcus ther-

mophilus and phosphorylates the lactose transporter LacS as

efficiently as HPr(His � P). Appl. Environ. Microbiol. 71, 1364–

1372.

[107] Lessard, C., Cochu, A., Lemay, J.D., Roy, D., Vaillancourt, K.,

Frenette, M., Moineau, S. and Vadeboncoeur, C. (2003)

Phosphorylation of Streptococcus salivarius lactose permease

(LacS) by HPr(His � P) and HPr(Ser-P)(His � P) and effects on

growth. J. Bacteriol. 185, 6764–6772.

[108] Moreno, M.S., Schneider, B.L., Maile, R.R., Weyler, W. and

Saier Jr., M.H. (2001) Catabolite repression mediated by the

CcpA protein in Bacillus subtilis: novel modes of regulation

revealed by whole-genome analyses. Mol. Microbiol. 39, 1366–

1381.

[109] Stanley, N.R., Britton, R.A., Grossman, A.D. and Lazazzera,

B.A. (2003) Identification of catabolite repression as a physio-

logical regulator of biofilm formation by Bacillus subtilis by use

of DNA microarrays. J. Bacteriol. 185, 1951–1957.

[110] Yoshida, K., Kobayashi, K., Miwa, Y., Kang, C.M., Matsu-

naga, M., Yamaguchi, H., Tojo, S., Yamamoto, M., Nishi, R.,

Ogasawara, N., Nakayama, T. and Fujita, Y. (2001) Combined

transcriptome and proteome analysis as a powerful approach to

study genes under glucose repression in Bacillus subtilis. Nucleic

Acids Res. 29, 683–692.

[111] Garault, P., Letort, C., Juillard, V. and Monnet, V. (2000)

Branched-chain amino acid biosynthesis is essential for optimal

growth of Streptococcus thermophilus in milk. Appl. Environ.

Microbiol. 66, 5128–5133.

[112] Letort, C. and Juillard, V. (2001) Development of a minimal

chemically-defined medium for the exponential growth of

Streptococcus thermophilus. J. Appl. Microbiol. 91, 1023–

1029.

[113] Neviani, E., Giraffa, G., Brizzi, A. and Carminati, D. (1995)

Amino acid requirements and peptidase activities of Streptococ-

cus salivarius subsp. thermophilus. J. Appl. Bacteriol. 79, 302–

307.

[114] Monnet, C., Nardi, M., Hols, P., Gulea, M., Corrieu, G. and

Monnet, V. (2003) Regulation of branched-chain amino acid

biosynthesis by alpha-acetolactate decarboxylase in Streptococ-

cus thermophilus. Lett. Appl. Microbiol. 36, 399–405.

[115] Sperandio, B., Polard, P., Ehrlich, S.D., Renault, P. and

Guedon, E. (2005) Sulfur amino acids metabolism and its

control in Lactococcus lactis IL1403. J. Bacteriol. 187, 3762–

3778.

[116] Limauro, D., Falciatore, A., Basso, A.L., Forlani, G. and De

Felice, M. (1996) Proline biosynthesis in Streptococcus thermo-

philus: characterization of the proBA operon and its products.

Microbiology 142, 3275–3282.

[117] Helinck, S., Le Bars, D., Moreau, D. and Yvon, M. (2004)

Ability of thermophilic lactic acid bacteria to produce aroma

compounds from amino acids. Appl. Environ. Microbiol. 70,

3855–3861.

[118] Shahbal, S., Hemme, D. and Desmazeaud, M. (1991) High cell

wall-associated proteinase activity of some Streptococcus ther-

mophilus strains (H-strains) correlated with a high acidification

rate in milk. Lait 71, 351–357.

[119] Shahbal, S., Hemme, D. and Renault, P. (1993) Characterization

of a cell envelope-associated proteinase activity from Strepto-

coccus thermophilus H-strains. Appl. Environ. Microbiol. 59,

177–182.

[120] Courtin, P., Monnet, V. and Rul, F. (2002) Cell-wall proteinases

PrtS and PrtB have a different role in Streptococcus thermophilus/

Lactobacillus bulgaricus mixed cultures in milk. Microbiology

148, 3413–3421.

[121] Siezen, R.J. (1999) Multi-domain, cell-envelope proteinases of

lactic acid bacteria. Anton. Leeuwen. 76, 139–155.

[122] Haandrikman, A.J., Kok, J., Laan, H., Soemitro, S., Ledeboer,

A.M., Konings, W.N. and Venema, G. (1989) Identification of a

gene required for maturation of an extracellular lactococcal

serine proteinase. J. Bacteriol. 171, 2789–2794.

[123] Rul, F. and Monnet, V. (1997) Presence of additional peptidases

in Streptococcus thermophilus CNRZ 302 compared to Lacto-

coccus lactis. J. Appl. Microbiol. 82, 695–704.

[124] Garault, P., Le Bars, D., Besset, C. and Monnet, V. (2002) Three

oligopeptide-binding proteins are involved in the oligopeptide

transport of Streptococcus thermophilus. J. Biol. Chem. 277, 32–

39.

[125] Monnet, V. (2003) Bacterial oligopeptide-binding proteins. Cell

Mol. Life Sci. 60, 2100–2114.

[126] Alloing, G., de Philip, P. and Claverys, J.P. (1994) Three highly

homologous membrane-bound lipoproteins participate in oligo-

peptide transport by the Ami system of the gram-positive

Streptococcus pneumoniae. J. Mol. Biol. 241, 44–58.

[127] Benateya, A., Bracquart, P., Le Deaut, J.Y. and Linden, G.

(1986) Croissance et prise de l�acide glutamique par les bacteries

462 P. Hols et al. / FEMS Microbiology Reviews 29 (2005) 435–463

du genre Streptococcus en fonction de la source d�energie. Lait66, 289–303.

[128] Bracquart, P., Le Deaut, J.Y. and Linden, G. (1989) Uptake of

glutamic acid by Streptococcus salivarius subsp. thermophilus

CNRZ 302. J. Dairy Res. 56, 107–116.

[129] Akpemado, K.M. and Bracquart, P. (1983) Uptake of branched-

chain amino acids by Streptococcus thermophilus. Appl. Environ.

Microbiol. 45, 136–140.

[130] Sanz, Y., Lanfermeijer, F.C., Renault, P., Bolotin, A., Konings,

W.N. and Poolman, B. (2001) Genetic and functional charac-

terization of dpp genes encoding a dipeptide transport system in

Lactococcus lactis. Arch. Microbiol. 175, 334–343.

[131] Delorme, C., Ehrlich, S.D. and Renault, P. (1999) Regulation of

expression of the Lactococcus lactis histidine operon. J. Bacte-

riol. 181, 2026–2037.

[132] Godon, J.J., Chopin, M.C. and Ehrlich, S.D. (1992) Branched-

chain amino acid biosynthesis genes in Lactococcus lactis subsp.

lactis. J. Bacteriol. 174, 6580–6589.

[133] van de, Guchte, M., Ehrlich, D.S. and Chopin, A. (1998)

tRNATrp as a key element of antitermination in the Lactococcus

lactis trp operon. Mol. Microbiol. 29, 61–74.

[134] Larsen, R., Buist, G., Kuipers, O.P. and Kok, J. (2004) ArgR

and AhrC are both required for regulation of arginine metab-

olism in Lactococcus lactis. J. Bacteriol. 186, 1147–1157.

[135] Guedon, E., Jamet, E. and Renault, P. (2002) Gene regulation in

Lactococcus lactis: the gap between predicted and characterized

regulators. Anton. Leeuwen. 82, 93–112.

[136] Guedon, E., Serror, P., Ehrlich, S.D., Renault, P. and Delorme,

C. (2001) Pleiotropic transcriptional repressor CodY senses the

intracellular pool of branched-chain amino acids in Lactococcus

lactis. Mol. Microbiol. 40, 1227–1239.

[137] Petranovic, D., Guedon, E., Sperandio, B., Delorme, C., Ehrlich,

D. and Renault, P. (2004) Intracellular effectors regulating the

activity of the Lactococcus lactis CodY pleiotropic transcription

regulator. Mol. Microbiol. 53, 613–621.

[138] Guedon, E., Renault, P., Ehrlich, S.D. and Delorme, C. (2001)

Transcriptional pattern of genes coding for the proteolytic

system of Lactococcus lactis and evidence for coordinated

regulation of key enzymes by peptide supply. J. Bacteriol. 183,

3614–3622.

[139] Letort, C., Nardi, M., Garault, P., Monnet, V. and Juillard, V.

(2002) Casein utilization by Streptococcus thermophilus results in

a diauxic growth in milk. Appl. Environ. Microbiol. 68, 3162–

3165.

[140] Chen, Y.Y. and Burne, R.A. (2003) Identification and charac-

terization of the nickel uptake system for urease biogenesis in

Streptococcus salivarius 57.I. J. Bacteriol. 185, 6773–6779.

[141] Chen, Y.Y., Weaver, C.A., Mendelsohn, D.R. and Burne, R.A.

(1998) Transcriptional regulation of the Streptococcus salivarius

57.I urease operon. J. Bacteriol. 180, 5769–5775.

[142] Chen, Y.Y., Hall, T.H. and Burne, R.A. (1998) Streptococcus

salivarius urease expression: involvement of the phosphoenol-

pyruvate:sugar phosphotransferase system. FEMS Microbiol.

Lett. 165, 117–122.

[143] Chen, Y.Y., Betzenhauser, M.J. and Burne, R.A. (2002) cis-

Acting elements that regulate the low-pH-inducible urease

operon of Streptococcus salivarius. Microbiology 148, 3599–

3608.

[144] Weaver, C.A., Chen, Y.Y. and Burne, R.A. (2000) Inactivation

of the ptsI gene encoding enzyme I of the sugar phosphotrans-

ferase system of Streptococcus salivarius: effects on growth and

urease expression. Microbiology 146, 1179–1185.

[145] Chen, Y.Y., Weaver, C.A. and Burne, R.A. (2000) Dual

functions of Streptococcus salivarius urease. J. Bacteriol. 182,

4667–4669.

[146] Monnet, C., Pernoud, S., Sepulchre, A., Fremaux, C. and

Corrieu, G. (2004) Selection and properties of Streptococcus

thermophilus mutants deficient in urease. J. Dairy Sci. 87, 1634–

1640.

[147] Mora, D., Maguin, E., Masiero, M., Parini, C., Ricci, G.,

Manachini, P.L. and Daffonchio, D. (2004) Characterization of

urease genes cluster of Streptococcus thermophilus. J. Appl.

Microbiol. 96, 209–219.

[148] Pernoud, S., Fremaux, C., Sepulchre, A., Corrieu, G. and

Monnet, C. (2004) Effect of the metabolism of urea on the

acidifying activity of Streptococcus thermophilus. J. Dairy Sci. 87,

550–555.

[149] Mora, D., Fortina, M.G., Parini, C., Ricci, G., Gatti, M.,

Giraffa, G. and Manachini, P.L. (2002) Genetic diversity and

technological properties of Streptococcus thermophilus strains

isolated from dairy products. J. Appl. Microbiol. 93, 278–287.

[150] Farr, S.B. and Kogoma, T. (1991) Oxidative stress responses in

Escherichia coli and Salmonella typhimurium. Microbiol. Rev. 55,

561–585.

[151] Thibessard, A., Fernandez, A., Gintz, B., Leblond-Bourget, N.

and Decaris, B. (2001) Hydrogen peroxide effects on Strepto-

coccus thermophilus CNRZ368 cell viability. Res. Microbiol. 152,

593–596.

[152] Carmel-Harel, O. and Storz, G. (2000) Roles of the glutathione-

and thioredoxin-dependent reduction systems in the Escherichia

coli and saccharomyces cerevisiae responses to oxidative stress.

Annu. Rev. Microbiol. 54, 439–461.

[153] Sherrill, C. and Fahey, R.C. (1998) Import and metabolism of

glutathione by Streptococcus mutans. J. Bacteriol. 180, 1454–

1459.

[154] Janowiak, B.E. and Griffith, O.W. (2005) Glutathione synthesis

in Streptococcus agalactiae. One protein accounts for gamma-

glutamylcysteine synthetase and glutathione synthetase activi-

ties. J. Biol. Chem. 280, 11829–11839.

[155] Yamamoto, Y., Kamio, Y. and Higuchi, M. (1999) Cloning,

nucleotide sequence, and disruption of Streptococcus mutans

glutathione reductase gene (gor). Biosci. Biotechnol. Biochem.

63, 1056–1062.

[156] Higuchi, M., Yamamoto, Y., Poole, L.B., Shimada, M., Sato,

Y., Takahashi, N. and Kamio, Y. (1999) Functions of two types

of NADH oxidases in energy metabolism and oxidative stress of

Streptococcus mutans. J. Bacteriol. 181, 5940–5947.

[157] Gibson, C.M., Mallett, T.C., Claiborne, A. and Caparon, M.G.

(2000) Contribution of NADH oxidase to aerobic metabolism of

Streptococcus pyogenes. J. Bacteriol. 182, 448–455.

[158] Auzat, I., Chapuy-Regaud, S., Le Bras, G., Dos Santos, D.,

Ogunniyi, A.D., Thomas I, Le, Garel, J.R., Paton, J.C. and

Trombe, M.C. (1999) The NADH oxidase of Streptococcus

pneumoniae: its involvement in competence and virulence. Mol.

Microbiol. 34, 1018–1028.

[159] Poyart, C., Pellegrini, E., Gaillot, O., Boumaila, C., Baptista, M.

and Trieu-Cuot, P. (2001) Contribution of Mn-cofactored

superoxide dismutase (SodA) to the virulence of Streptococcus

agalactiae. Infect. Immun. 69, 5098–5106.

[160] Yesilkaya, H., Kadioglu, A., Gingles, N., Alexander, J.E.,

Mitchell, T.J. and Andrew, P.W. (2000) Role of manganese-

containing superoxide dismutase in oxidative stress and virulence

of Streptococcus pneumoniae. Infect. Immun. 68, 2819–2826.

[161] Chang, S.K. and Hassan, H.M. (1997) Characterization of

superoxide dismutase in Streptococcus thermophilus. Appl.

Environ. Microbiol. 63, 3732–3735.

[162] Condon, S. (1987) Responses of lactic acid bacteria to oxygen.

FEMS Microbiol. Rev. 46, 269–280.

[163] McAllister, L.J., Tseng, H.J., Ogunniyi, A.D., Jennings, M.P.,

McEwan, A.G. and Paton, J.C. (2004) Molecular analysis of the

psa permease complex of Streptococcus pneumoniae. Mol.

Microbiol. 53, 889–901.

[164] Yamamoto, Y., Poole, L.B., Hantgan, R.R. and Kamio, Y.

(2002) An iron-binding protein, Dpr, from Streptococcus mutans

P. Hols et al. / FEMS Microbiology Reviews 29 (2005) 435–463 463

prevents iron-dependent hydroxyl radical formation in vitro. J.

Bacteriol. 184, 2931–2939.

[165] Nicodeme, M., Perrin, C., Hols, P., Bracquart, P. and Gaillard,

J.L. (2004) Identification of an iron-binding protein of the Dps

family expressed by Streptococcus thermophilus. Curr. Microbiol.

48, 51–56.

[166] Touati, D. (2000) Iron and oxidative stress in bacteria. Arch.

Biochem. Biophys. 373, 1–6.

[167] Hecker, M. and Volker, U. (1998) Non-specific, general and

multiple stress resistance of growth-restricted Bacillus subtilis

cells by the expression of the sigmaB regulon. Mol. Microbiol.

29, 1129–1136.

[168] Graham, M.R., Smoot, L.M., Migliaccio, C.A., Virtaneva, K.,

Sturdevant, D.E., Porcella, S.F., Federle, M.J., Adams, G.J.,

Scott, J.R. and Musser, J.M. (2002) Virulence control in group A

Streptococcus by a two-component gene regulatory system:

global expression profiling and in vivo infection modeling. Proc.

Natl. Acad. Sci. USA 99, 13855–13860.

[169] Chaussee, M.A., Callegari, E.A. and Chaussee, M.S. (2004) Rgg

regulates growth phase-dependent expression of proteins asso-

ciated with secondary metabolism and stress in Streptococcus

pyogenes. J. Bacteriol. 186, 7091–7099.

[170] Ulrich, L.E., Koonin, E.V. and Zhulin, I.B. (2005) One-

component systems dominate signal transduction in prokary-

otes. Trends Microbiol. 13, 52–56.

[171] Hoch, J.A. (2000) Two-component and phosphorelay signal

transduction. Curr. Opin. Microbiol. 3, 165–170.

[172] Stock, A.M., Robinson, V.L. and Goudreau, P.N. (2000) Two-

component signal transduction. Annu. Rev. Biochem. 69, 183–

215.

[173] Dagkessamanskaia, A., Moscoso, M., Henard, V., Guiral, S.,

Overweg, K., Reuter, M., Martin, B., Wells, J. and Claverys, J.P.

(2004) Interconnection of competence, stress andCiaR regulons in

Streptococcus pneumoniae: competence triggers stationary phase

autolysis of ciaRmutant cells. Mol. Microbiol. 51, 1071–1086.

[174] Hancock, L.E. and Perego, M. (2004) Systematic inactivation

and phenotypic characterization of two-component signal trans-

duction systems of Enterococcus faecalis V583. J. Bacteriol. 186,

7951–7958.

[175] Kleerebezem, M. and Quadri, L.E. (2001) Peptide pheromone-

dependent regulation of antimicrobial peptide production in

Gram-positive bacteria: a case of multicellular behavior. Pep-

tides 22, 1579–1596.

[176] Echenique, J.R. and Trombe, M.C. (2001) Competence repres-

sion under oxygen limitation through the two-component

MicAB signal-transducing system in Streptococcus pneumoniae

and involvement of the PAS domain of MicB. J. Bacteriol. 183,

4599–4608.

[177] Fabret, C. and Hoch, J.A. (1998) A two-component signal

transduction system essential for growth of Bacillus subtilis:

implications for anti-infective therapy. J. Bacteriol. 180, 6375–

6383.

[178] Lange, R., Wagner, C., de Saizieu, A., Flint, N., Molnos, J.,

Stieger, M., Caspers, P., Kamber, M., Keck, W. and Amrein,

K.E. (1999) Domain organization and molecular characteriza-

tion of 13 two-component systems identified by genome

sequencing of Streptococcus pneumoniae. Gene 237, 223–234.

[179] O�Connell-Motherway, M., van Sinderen, D., Morel-Deville, F.,

Fitzgerald, G.F., Ehrlich, S.D. and Morel, P. (2000) Six putative

two-component regulatory systems isolated from Lactococcus

lactis subsp. cremoris MG1363. Microbiology 146, 935–947.

[180] Wagner, C., Saizieu, Ad A., Schonfeld, H.J., Kamber, M.,

Lange, R., Thompson, C.J. and Page, M.G. (2002) Genetic

analysis and functional characterization of the Streptococcus

pneumoniae vic operon. Infect. Immun. 70, 6121–6128.

[181] Claverys, J.P. and Havarstein, L.S. (2002) Extracellular-peptide

control of competence for genetic transformation in Streptococ-

cus pneumoniae. Front. Biosci. 7, 1798–1814.

[182] Cvitkovitch, D.G. (2001) Genetic competence and transforma-

tion in oral streptococci. Crit. Rev. Oral Biol. Med. 12, 217–243.

[183] Claverys, J.P. and Martin, B. (2003) Bacterial competence genes:

signatures of active transformation, or only remnants?. Trends

Microbiol. 11, 161–165.

[184] Peterson, S.N., Sung, C.K., Cline, R., Desai, B.V., Snesrud,

E.C., Luo, P., Walling, J., Li, H., Mintz, M., Tsegaye, G., Burr,

P.C., Do, Y., Ahn, S., Gilbert, J., Fleischmann, R.D. and

Morrison, D.A. (2004) Identification of competence pheromone

responsive genes in Streptococcus pneumoniae by use of DNA

microarrays. Mol. Microbiol. 51, 1051–1070.

[185] Li, Y.H., Lau, P.C., Lee, J.H., Ellen, R.P. and Cvitkovitch, D.G.

(2001) Natural genetic transformation of Streptococcus mutans

growing in biofilms. J. Bacteriol. 183, 897–908.

[186] Hidalgo-Grass, C., Ravins, M., Dan-Goor, M., Jaffe, J., Moses,

A.E. and Hanski, E. (2002) A locus of group A Streptococcus

involved in invasive disease and DNA transfer. Mol. Microbiol.

46, 87–99.

[187] O�Sullivan, L., Ross, R.P. and Hill, C. (2002) Potential of

bacteriocin-producing lactic acid bacteria for improvements in

food safety and quality. Biochimie 84, 593–604.

[188] Mathot, A.G., Beliard, E. and Thuault, D. (2003) Streptococcus

thermophilus 580 produces a bacteriocin potentially suitable for

inhibition of Clostridium tyrobutyricum in hard cheese. J. Dairy

Sci. 86, 3068–3074.

[189] Siezen, R.J., Kuipers, O.P. and de Vos, W.M. (1996) Compar-

ison of lantibiotic gene clusters and encoded proteins. Anton.

Leeuwen. 69, 171–184.

[190] de Saizieu, A., Gardes, C., Flint, N., Wagner, C., Kamber, M.,

Mitchell, T.J., Keck, W., Amrein, K.E. and Lange, R. (2000)

Microarray-based identification of a novel Streptococcus pneu-

moniae regulon controlled by an autoinduced peptide. J.

Bacteriol. 182, 4696–4703.

[191] Saitou, N. and Nei, M. (1987) The neighbor-joining method: a

new method for reconstructing phylogenetic trees. Mol. Biol.

Evol. 4, 406–425.