Date post: | 30-Oct-2014 |
Category: |
Documents |
Upload: | victoria-roseblack |
View: | 30 times |
Download: | 3 times |
Research review paper
Biosynthesis, characterisation, and design of bacterial
exopolysaccharides from lactic acid bacteria
Andrew Laws*, Yucheng Gu, Valerie Marshall
School of Applied Sciences, University of Huddersfield, Queensgate, Huddersfield HD1 3DH, UK
Abstract
Lactic acid bacteria (LAB) are characterised by their conversion of a large proportion of their
carbon feed, fermentable sugars, to lactic acid. However, in addition to lactic acid production, the LAB
are able to divert a small proportion of fermentable sugars towards the biosynthesis of
exopolysaccharides (EPSs) that are independent of the cell surface and cell wall material. These
microbial EPSs when suspended or dissolved in aqueous solution provide thickening and gelling
properties, and, as such, there is great interest in using EPSs from food grade microorganisms (such as
the LAB that are traditionally used for food fermentations) for use as thickening agents. The current
review includes a brief summary of the recent literature describing features of the biosynthetic
pathways leading to EPS production. Many aspects of EPS biosynthesis in LAB are still not fully
understood and a number of inferences are made regarding the similarity of the pathway to those
involved in the synthesis of other cell polysaccharides, e.g., cell wall components. The main body of
the review will cover practical aspects concerned with the isolation and characterisation of EPS
structures. In the last couple of years, a substantial number of structures have been published and a
summary of the common elements of these structures is included as is a suggestion for a system for
representing structures. A brief highlight of the attempts that are being made to design ‘tailor’-made
polysaccharides using genetic modification and control of metabolic flux is presented. D 2001 Elsevier
Science Inc. All rights reserved.
Keywords: Exopolysaccharides; Lactic acid bacteria; Biothickener; Oligosaccharide; Bacterial polysaccharide;
NMR
0734-9750/01/$ – see front matter D 2001 Elsevier Science Inc. All rights reserved.
PII: S0734 -9750 (01 )00084 -2
* Corresponding author. Tel.: +44-1484-472-668; fax: +44-1484-472-182.
E-mail address: [email protected] (A. Laws).
Biotechnology Advances 19 (2001) 597–625
1. Exopolysaccharides of lactic acid bacteria
Exopolysaccharides (EPSs) are long-chain polysaccharides that are secreted mainly by
bacteria and microalgae (Sutherland, 1972, 1977) into their surroundings during growth and
that are not permanently attached to the surface of the microbial cell. The physical
characteristics of EPSs are responsible for the slime-forming or mucoid trait of many
microorganisms. A second group of polysaccharides that are structurally similar but that
are permanently attached to the cell surface are classified as capsular polysaccharides
(Sutherland, 1985). The current review will focus exclusively on EPSs and capsular
polysaccharides will not be discussed here.
The ability of microorganisms to synthesise EPSs has both positive and negative attributes.
The food industry uses polysaccharides from plant and seaweeds for their thickening and
gelling properties. However, existing supplies are either not sufficiently reliable, are of
variable consistency, or the quantities available are not sufficient to match demand. In
response a number of polysaccharides of microbial origin have been developed including
xanthan from Xanthomanas campestris (Garcia-Ochoa et al., 2000) and gellan from
Sphingomonas paucimobilis (Banik et al., 2000; Giavasis et al., 2000; Sutherland, 1999).
These can be prepared in reliable quantities using conventional biotechnological processes.
However, the physical properties of these polymers are such that they are not suited to all
applications and there is a demand for novel materials that impart improved rheological
characteristics. One of the main drawbacks with using microbial polysaccharides in food
formulations is the requirement for them to be considered as a food additive.
Many lactic acid bacteria (LAB) are routinely used in food preparations for their
preservative effects: the acidification, resulting from sugar metabolism, restricts further
microbial contamination. LAB isolated from Scandinavian ropy fermented milk products
are known to produce EPSs (Forsen and Myllymaa, 1974; Forsen and Pakkila, 1979; Macura
and Townsley, 1984). These EPSs provide thickening properties, and are also considered to
improve the texture and mouthfeel of dairy products (Marshall and Rawson, 1999). LAB are
generally regarded as safe and EPSs isolated from LAB offer an alternative source of microbial
polysaccharides for wider use in food formulations. In addition to their use in food production,
there are a number of reports of possible health benefits of EPSs, and especially in regard to the
immunostimulating properties of bacterial EPSs (Oda et al., 1983; Schiffrin et al., 1995).
The negative attributes of EPS synthesis are associated with their spoilage properties. The
synthesis of EPS by LAB during wine (Denadra and Desaad, 1995; Lonvaud-Funel, 1999)
and cider (Duenas et al., 1995) production leads to products having undesirable rheological
properties. The formation of dental plaques is related to EPS synthesis by LAB (Loesche,
1986; Rozen, 2001). The EPSs from LAB are responsible for biofilm formation (Oliveria,
1992; Poulsen, 1999) that can lead to biofouling (Azeredo and Oliveira, 2000). The most
notable examples of biofouling are associated with biofilm formation in the equipment used
for the processing of dairy products. Biofilms cause a significant number of technical and
hygiene problems for the dairy industry (Poulsen, 1999).
Before trying to understand the biosynthesis of bacterial EPSs, it is important to realise that
bacteria synthesise a number of different classes of polysaccharide and that a significant
A. Laws et al. / Biotechnology Advances 19 (2001) 597–625598
proportion of these polysaccharides are used for the construction of components of the cell
wall. It is worth pointing out at this early stage that many of the mechanisms involved in the
synthesis of EPS are either known, or predicted, to be shared with those of other cell wall
components. The biosynthesis and the functionality of the main cell wall components have
recently been reviewed (Delcour et al., 1999).
Microbial EPSs can be divided into two groups: homopolysaccharides (e.g., cellulose,
dextran, pullulan, levan, and curdlan) and heteropolysaccharides (e.g., gellan and xanthan).
Homopolysaccharides are constructed from monosaccharides joined by either a single linkage
type (e.g., 1–2 or 1–4) or by a combination of a limited number of linkage types (e.g., 1–2
and 1–4). The current review will focus on recent work related to the synthesis and structural
characterisation of heteropolysaccharides from LAB. Heteropolysaccharides are constructed
from multiple copies of an oligosaccharide. The oligosaccharide can contain between three
and seven residues, it possesses a variety of two or more different types of monosaccharides
and frequently has a range of different linkage patterns. Current interest in aspects of EPS
synthesis is large, and consequently, there have been a number of reviews covering work in
this area (de Vuyst and Degeest, 1999; Ricciardi and Clementi, 2000; Cerning and Marshall,
1999; Sutherland, 1998; Cerning, 1990). In the current review, many aspects of the early
‘groundbreaking’ studies will not be covered and the author apologises to the many
researchers in the field whose efforts may appear to be treated without due reverence. This
article will attempt to review more practical aspects of current research and will try to paint a
picture of the current knowledge in the area. Readers interested in further development of the
subject are directed to the reviews listed above and particularly to that of Cerning (1990).
2. Biosynthesis of heteropolysaccharides by LAB
The biosynthesis of bacterial EPSs is complex and involves the concerted action of a large
number of gene products. The genes coding for the enzymes and regulatory proteins required
for EPS synthesis are of plasmid origin in the mesophilic LAB strains, e.g., Lactococcus and
chromosomally based in the thermophilic strains of Streptococcus and Lactobacilli. The EPS-
producing ability of LAB is regarded as being unstable. For mesophilic LAB strains, the
unstable nature of EPS synthesis is consistent with the genes for EPS synthesis being plasmid
bound. For the thermophilic LAB strains, it has been proposed that the loss of EPS-producing
character is due to deletions and rearrangement resulting from genetic instability. Identifica-
tion of EPS gene clusters and suggestions for the functional role/s of the gene products have
been reported for a number of food grade microorganisms: Streptococcus thermophilus Sfi 6
(Stingele et al., 1996), Streptococcus thermophilus Sfi 39 (Germond et al., 2001), and
Lactococcus lactis NIZO B40 (van Kranenburg et al., 1997) (Fig. 1). In addition to EPS-
specific gene products, the biosynthetic pathway relies on a number of the so-called
housekeeping enzymes such as those required for the preparation of sugar nucleotides.
The biosynthetic pathway can be broken down into four separate reaction sequences. These
are the reactions involved with sugar transport into the cytoplasm, the synthesis of sugar-1-
phosphates, activation of and coupling of sugars, and the processes involved in the export of
A. Laws et al. / Biotechnology Advances 19 (2001) 597–625 599
the EPS. These are schematically represented in Fig. 2. Each of these aforementioned
sequences must be evaluated in developing strategies for the engineering of EPS products.
2.1. Sugar transport into the cytoplasm
The movement of carbon feeds, mainly monosaccharides and disaccharides, from the
surrounding growth medium into the cytoplasm is a carefully regulated process. A number of
different proteins control the internalisation of sugars. Sugar transport and its possible
influence on the engineering of EPS structure has been reviewed (de Vos, 1996). The most
frequently encountered sugar transport machinery is that of the bacterial phosphoenolpyr-
uvate (PEP)–sugar phosphotransferase system (PTS) (Postma et al., 1993). The PEP–PTS
system contains a group of proteins that are responsible for binding, transmembrane transport,
and phosphorylation of a variety of sugar substrates. In addition, a number of other proteins
exist that regulate the activity of each of the latter processes.
The first group of proteins includes enzyme I, the histidine-containing phosphocarrier
protein HPr and a carbohydrate-specific permease enzyme complex (enzyme II) (Viana et al.,
2000). These enzymes act in sequence to provide phosphorylated sugars in the cytoplasm.
The sequence is initiated when a phosphate group is transferred from PEP to enzyme I;
enzyme I subsequently phosphorylates a histidine residue of HPr to yield HPr(His-P) (Postma
et al., 1993). At the same time, a protein of the enzyme II complex binds the carbohydrate
(Sliz et al., 1996, 1997). The carbohydrate-specific enzymes (II) transport sugars across the
membrane and catalyse the transfer of the phosphate group from HPr(His-P) onto the sugar.
In the PEP–PTS transport system of L. lactis, three Type II enzymes are required for lactose
transport: enzymes IIA–C (Wang et al., 2000). Enzymes IIA and IIB are located in the
cytoplasm and enzyme IIC acts as a membrane channel. On first inspection, the requirement
for a number of enzymes for the transport of a phosphate group from PEP to a sugar appears
over elaborate; however, the complexity of the process is required by the need for tight
regulation of nutrient acquisition.
A second set of proteins is responsible for regulation of nutrient acquisition. These
regulatory proteins do not only activate and repress PEP–PTS systems, they also regulate
non-PTS uptake systems (Poolman et al., 1997; Stulke and Hillen, 1998; Gunnewijk and
Fig. 1. Organisation of the EPS gene cluster of (A) L. lactis ssp. cremoris NIZO B40 (plasmid localised) (Stingele
et al., 1999) and (B) S. thermophilus Sfi6 (chromosomally encoded) (van Kranenburg et al., 1997). The proposed
function of the different gene products is indicated.
A. Laws et al. / Biotechnology Advances 19 (2001) 597–625600
Fig. 2. (A) Schematic representation of a number of possible pathways for sugar transport and metabolism in LAB. (B) Schematic representation of a
possible pathway for EPS biosynthesis in L. lactis NIZO B40 starting from glucose-6-phosphate (the site of polymerisation of the repeat unit has not been
established and may occur on either face of the membrane). Known enzymes are indicated using either accepted nomenclature for the enzyme or by reference
to their encoding genes and using established genetic nomenclature. The figure is a highly adapted one from those presented by (Kleerebezem, 2000).
A.Lawset
al./Biotech
nologyAdvances
19(2001)597–625
601
Poolman, 2000a,b; Gunnewijk et al., 1999). Most notable of the regulators is the binary
complex formed between the catabolite control protein CcpA and the PTS protein HPr (van
den Bogaard et al., 2000). The histidine-containing phosphocarrier protein HPr has two
phosphorylation sites and His(Ser-P) has been shown to regulate sugar transport; CcpA in
combination with HPr(Ser-P) binds to the cis-acting DNA sequence termed the catabolite-
responsive element (cre). Jones et al. (1997) demonstrated that CcpA only binds phosphory-
lated HPr(Ser-P), which suggests that catabolite repression is linked to active sugar transport.
Alternative, non-PEP–PTS, transport systems exist for the import of sugars, e.g., primary
and secondary transport systems. A number of LAB do not have active PEP–PTS transport
system for all the sugars that they are able to internalise. In the absence of a sugar-specific
PEP–PTS transport system there is a requirement for active transport. In S. mutans, multiple
sugars are transported by a primary transport system (Tao et al., 1993; Sutcliffe et al., 1993).
In S. thermophilus, sugar transport is mainly by a secondary transport mechanism coded for
by lacS (Friesen et al., 2000a,b). lacS is able to import lactose in symport with protons or,
alternatively, lacS can function as a lactose–galactose antiport system.
2.2. Synthesis of sugar-1-phosphates
Once inside the cytoplasm, the fate of the carbon feed is determined by the state of
phosphorylation of the sugar: sugar-6-phosphates are consumed in catabolic pathways
whereas sugar-1-phosphates can participate in polysaccharide synthesis. As the majority of
sugars are transported into the cytoplasm by PEP–PTS systems, which generate sugar-6-
phosphates, a number of authors have pointed to the possible key role that phosphoglu-
comutases (PGMs) may play in diverting flux between the two pathways (Sjoberg and
Hahn-Hagerdal, 1989; Degeest and de Vuyst, 2000).
Early work on PGMs focused on the role they play in the catabolism of sugar-1-
phosphates (Qian et al., 1994). In the metabolism of maltose, a sugar that is transported
via a permease in L. lactis, the disaccharide is initially converted into one molecule of
glucose and a molecule of b-glucose-1-phosphate. If the b-glucose-1-phosphate is to be used
in the catabolic pathway, it must subsequently be converted to glucose-6-phosphate and this
reaction is catalysed by b-PGM (Qian et al., 1997). In a similar manner, it has been suggested
that a-PGMs may play exactly the opposite role in EPS biosynthesis, i.e., in diverting the
glucose-6-phosphate to a-glucose-1-phosphates and EPS biosynthesis (Sjoberg and Hahn-
Hagerdal, 1989; Degeest and de Vuyst, 2000). It has recently been demonstrated that for the
synthesis of EPS in L. lactis grown on glucose, a proportion of the carbon feed must be
converted to glucose-6-phosphate for EPS synthesis (Ramos et al., 2001).
A number of alternative pathways leading to a-glucose-1-phosphate have been suggested
(see bottom right-hand side portion of Fig. 2 and the discussion below). The mode of
synthesis of glucose-1-phosphate is dependent on a number of variables: most notably, the
carbon source on which the culture is grown and the transport system used for sugar import. It
has been proposed (Grobben et al., 1996) that for L. bulgaricus grown on fructose that
fructose is imported via a PEP–fructose PTS that specifically yields fructose-1-phosphate.
Fructose-1-phosphate is converted, via fructose-1,6-bisphosphate, to fructose-6-phosphate, to
A. Laws et al. / Biotechnology Advances 19 (2001) 597–625602
glucose-6-phosphate, and finally to glucose-1-phosphate. A much shorter pathway to glucose-
1-phosphate is available for L. bulgaricus grown on glucose; glucose is internalised via a
PEP–glucose PTS yielding glucose-6-phosphate that is transformed directly into glucose-1-
phosphate (Grobben et al., 1996). When L. lactis is grown on lactose (de Vos and Vaughan,
1994), lactose is imported by a lactose-specific phosphotransferase transport system providing
internal lactose-6-phosphate. Lactose-6-phosphate is subsequently hydrolysed by a phospho-
b-galactosidase to generate galactose-6-phosphate and glucose. A glucokinase is required for
the synthesis of a-glucose-6-phosphate. In contrast, galactose-negative S. thermophilus
internalises lactose by coupling a lactose permease within an antiport secondary transport
system (Poolman, 1993). In the latter system, it is necessary to use a combination of a
glycosidase and kinase to generate glucose-6-phosphate.
Given the variety of routes leading to glucose-1-phosphate, it might be expected that a
pathway should be available for the synthesis and subsequent utilisation of galactose-1-
phosphate. This is indeed the case and utilises the enzymes of the Leloir pathway. The enzymes
of the Leloir pathway convert galactose to a-galactose-1-phosphate (GalK), a-galactose-1-phosphate to UDP-galactose (GalT) and UDP-galactose to UDP-glucose (GalE) (see Fig. 3). It
is worth noting that for galactose-grown cells, the presence of UDP-glucose pyrophosphorylase
and a supply of UTP would allow the reaction to be considered as catalytic in UDP-glucose.
The combined action of the enzymes could supply both UDP-glucose and UDP-galactose
without a requirement for a PGM. However, it is well known that a number of dairy strains of S.
thermophilus lack an available galactokinase; though, under appropriate conditions, mutants
that are able to ferment galactose are selected indicating that the galactokinase genes are present
but are not transcribed (Thomas and Crow, 1984; Hutkins et al., 1985).
2.3. Synthesis of sugar nucleotides and polymerisation into the EPS repeat unit
It is instructive, at this point in the reaction sequence, to split the genes coding for the
proteins required for EPS biosynthesis into two groups: genes required for the synthesis of
sugar nucleotides and EPS-specific genes. These two groups are physically separated in the
genome, and in the case of L. lactis, are further removed from each other as the genes for EPS
biosynthesis (upper part of Fig. 2) are extrachromosomal. The first group consists of the
Fig. 3. Conversion of galactose to UDP-galactose using enzymes of the Leloir pathway. Enzymes involved in the
conversions are indicated.
A. Laws et al. / Biotechnology Advances 19 (2001) 597–625 603
genes coding for enzymes and proteins required for the synthesis of sugar nucleotides from
which the repeat unit is constructed. Sugar nucleotides are needed for the synthesis of a range
of polysaccharides and are not specific to EPS biosynthesis and, as such, are frequently
termed ‘housekeeping’ enzymes (see middle of Fig. 2).
The sugar nucleotides required for the construction of the majority of EPS structures are
UDP-glucose, UDP-galactose, and dTDP-rhamnose: the precursors of the repeat unit. The
genes coding for the enzymes needed for the synthesis of the sugar nucleotides from glucose-
1-phosphate (galU, galE, rfbA, rfbB, rfbC, and rfbD) have been identified and cloned from L.
lactis strain MG1363 (Boels et al., 1988; Reeves, 1993; Kleerebezem et al., 1999). The
function of the respective gene products is illustrated in the middle section of Fig. 2. The first
enzyme in the sequence is GalU, a UDP-glucose pyrophosphorylase. It has been reported
(Kleerebezem et al., 1999) that the intracellular levels of UDP-glucose is determined by the
activity of the enzyme GalU; functional overexpression of the lactococcal galU gene results in
much larger UDP-glucose levels in L. lactis. The production of UDP-Gal is believed to be
derived principally from UDP-Glc through the action of GalE that catalyses the interconver-
sion of the two UDP-sugars. The requirement for GalE for EPS biosynthesis in L. lactis NIZO
B40 was demonstrated by Kleerebezem et al. (1999): a galE mutant was not able to synthesise
EPS when grown on glucose but EPS was produced when the mutant was grown on galactose.
The latter result implies that, in the absence of galactose, UDP-galactose required for EPS
synthesis is derived solely from UDP-glucose. Details of the nature of the enzymes needed to
produce dTDP-rhamnose were established in Gram-negative bacteria where rhamnose is a key
constituent of the O antigens of lipopolysaccharides (Reeves, 1993). Four enzymes, RfbA,
RfbB, RfbC, and RfbD, convert a-glucose-1-phosphate initially to dTDP-glucose then to
4-keto-6-deoxymannose and finally to dTDP-rhamnose.
The next stage in EPS biosynthesis uses the EPS-specific enzymes. The first gene clusters
coding for production of secreted EPSs to be identified and characterised were those from S.
thermophilus Sfi 6 (Stingele et al., 1996) and for L. lactis NIZO B40 (van Kranenburg et al.,
1997). In S. thermophilus Sfi 6, the gene cluster is located on the chromosome, is 14.5 kb in size,
and contains 13 genes. In contrast, the L. lactis gene cluster is located on a 40-kb plasmid, the
EPS gene cluster is 12 kb in size and contains 14 genes. The organisation of the gene clusters is
similar for both organisms and consists of four separate domains (see Fig. 1). A central core
coding for the glycosyl-transferases is flanked at the ends by genes coding for proteins having a
strong homology with enzymes used for polymerisation and export. A regulatory domain is
present at the start of the gene cluster. Stingele et al. (1999) have demonstrated that the gene
cluster contains EPS-specific enzymes by insertion of the cluster into the non-EPS-producing
heterologous host L. lactis MG1363 and demonstrating EPS synthesis.
In vitro experiments using [14C]-labelled sugar nucleotides have provided evidence that
the monosaccharide repeat unit is assembled on a lipid carrier, which is attached to the
cytoplasmic membrane (van Kranenburg et al., 1997, 1999). In O antigen synthesis, the
repeat unit construction occurs on the inside surface of the cytoplasmic membrane (see top of
Fig. 2). For L. lactis NIZO B40, the first sugar to be attached to the membrane-anchored
phosphorylated lipid is glucose. Significantly, the [14C] glucose is attached through its
anomeric carbon by a pyro-diester link. The mode of action of the ‘priming’ glycosyl-
A. Laws et al. / Biotechnology Advances 19 (2001) 597–625604
transferase, a UDP-glucose transferase for L. lactis NIZO B40, requires donation of the sugar
and a phosphate group to the lipid. In L. lactis NIZO B40, the second sugar in the repeat unit
is glucose and this is added, with a b-glycosidic link, through the attachment of the anomeric
carbon of a-UDP-glucose to the 4-hydroxy group of the lipid-bound glucose. This process
requires two EPS gene products, EpsE and EpsF. The final backbone residue, a b(1–4)-linked-galactose, is derived from UDP-galactose and the coupling is catalysed by a single
gene product EpsG. van Kranenburg et al. (1999) suggest that the branches, a rhamnose and
a phosphogalactose, are added to the 2- and 3-positions of the repeat unit in successive
glycosyl/phosphoglycosyl-transferase-catalysed steps generating the required repeat unit
joined by a diphosphate link to the lipid. Whether the sugar is incorporated as an a- or
b-glycoside is dictated by the catalytic mechanism of the glycosyl-transferases. These
enzymes have different catalytic domains, and some may catalyse insertion such that there is
retention of the anomeric configuration whilst others will result in inversion of the
configuration (Saxena et al., 1995).
The assembly of the repeat unit on a lipid carrier is a process that is used for the synthesis
of excreted polysaccharides, for cell wall peptidoglycans and for cell surface oligosaccharides
and polysaccharides. There is evidence to suggest that the various oligosaccharide and
polysaccharide syntheses use the same building blocks (sugar nucleotides) and scaffolding
(lipid carrier). The latter may account for the close relationships between rates of EPS
synthesis and cell growth that have been observed by a number of authors (Garcia-Garibay
and Marshall, 1991; Cerning et al., 1992).
2.4. EPS polymerisation and export to the surrounding medium
Details relating to the events leading to the polymerisation of the repeat unit and its export
from the plasma membrane through the peptidoglycan layers of Gram-positive bacteria are
very scarce. Gonzalez et al. (1998) have studied the genes involved in polymerisation and
export of the EPS succinoglycan in Sinorhizobium melilot, formerly Rhizobium meliloti. These
authors concluded that the subunits are constructed on an undecaprenol lipid carrier on the
cytoplasmic face of the plasma membrane, which are polymerised in a block fashion. Details
of the mechanism by which blocks are polymerised in EPS biosynthesis are not known. In O
antigen synthesis, three gene products are required for polymerisation and export (Whitfield
and Valvano, 1993). The gene products code for a protein that catalyses the movement of the
lipid-bound material form the cytoplasmic face of the membrane to the periplasmic face
(flippase or translocase), a protein that catalyses the polymerisation of the blocks (polymerase)
and a protein responsible for controlling polymer chain length.
By analogy, a simple model for EPS polymerisation and export requires the action of a
flippase to translocate the lipid-bound repeat units, a polymerase to catalyse the coupling of
repeat units and finally an enzyme to catalyse the detachment of the lipid-bound polymer and
that will control chain length. Evidence for the identification of the proteins responsible for
the catalytic functions described above come from comparison of sequences of genes and
from hydrophobicity cluster analysis of the derived amino acid sequences of similar proteins.
In succinoglycan synthesis, there is strong evidence to suggest that the gene product from
A. Laws et al. / Biotechnology Advances 19 (2001) 597–625 605
exoQ is the polymerase (Gonzalez et al., 1998). ExoQ proteins show sequence and
topological similarities with the O antigen polymerase gene (Wzy(Rfc)) product (Whitfield
and Valvano, 1993). Becker et al. (1995) have suggested that the ExoP product of R. meliloti
is involved in chain length determination.
In L. lactis NIZO B40, the genes epsA and epsB show homology with the R. meliloti ExoP
genes and the gene products are proposed to be involved in chain length determination (one
protein may detach the polymer whilst the second may regulate the activity of the first). EpsK
and EpsI are homologous to the flippase and polymerases in Salmonella or Shigella O antigen
synthesis (Morona et al., 1994; Liu et al., 1996).
3. EPS structure and yield
In determining the structure and yields of the EPSs, care should be taken in the choice of
techniques used for the preparation, isolation, and characterisation of the EPS. Early EPS
literature inevitably contains many examples where the presence of contaminants, both low-
molecular mass material and high-molecular mass polysaccharides, are included in calcu-
lations of monomer compositions and of yield. Many concluding statements regarding EPSs
having variable composition, dependent on the carbon feed, may be false and are likely to
simply be a measure of the extent to which foreign polysaccharides have been isolated with the
EPS. Particular caution should be taken when the quantities of the EPS recovered are small and
are of an order of a few milligrams. At the same time, it is worth noting that many of the
physical and chemical processes involved in the isolation and characterisation of polysac-
charides can seriously influence structure and yield of the EPS (see discussion below). There
are a number of experimental practices that can be adopted in order to avoid some of the simple
errors that are frequently reported. This review will briefly cover those areas of concern.
3.1. Preparation of EPSs
The starting point for EPS production is the preparation of a culture inoculum and this is
the first point at which contaminants may be added to the system. Master cultures are
frequently prepared from broths that contain polysaccharides. Degeest and de Vuyst (1999)
and Marshall et al. (2001b) generate inocula for use in fermentations using a minimum of two
subculturing steps. The subculturing steps are necessary to remove, via dilution, unwanted
high-molecular mass material. Many of the reports of presence of trace amounts of sugars
other than rhamnose, galactose, glucose, and N-acetylglucosamine in EPS compositions are
most likely false: the sugars being derived from polysaccharides originally present in the
starting inoculum culture broth.
Fermentation conditions greatly influence EPS production. There are a number of reports
suggesting that for a variety of LAB strains, mainly mesophilic strains, that EPS production is
greatest under conditions that are not optimal for growth (Cerning et al., 1992; Gassem et al.,
1995; Gamar et al., 1997; Schellhass, 1983; Mozzi et al., 1995). The latter result is consistent
with both EPS and cell wall biosynthesis pathways utilising the same membrane lipid.
A. Laws et al. / Biotechnology Advances 19 (2001) 597–625606
Optimum temperatures for EPS synthesis have been determined for a number of LAB.
Cerning et al. (1992) have shown that the optimum temperature for EPS synthesis is 25 �C for
L. lactis. For S. thermophilus temperatures less than 37 �C are optimal for EPS biosynthesis.
EPS production in Lb. rhamnosus is high between 20 and 25 �C whilst growth is optimal
between 30 and 37 �C (Gamar et al., 1997). For Lb. delbrueckii ssp. bulgaricus RR, EPS
synthesis is increased at temperatures above (45 �C) and below (30 �C) that at which
optimum growth is observed (Schellhass, 1983). In contrast, for Lb. acidophilus and Lb. casei
EPS production is at a maximum at the optimum growth temperature (Mozzi et al., 1995;
Petry et al., 2000).
The literature relating to the influence of pH and medium composition on EPS production
has recently been reviewed (Ricciardi and Clementi, 2000; Cerning and Marshall, 1999) and
only a brief summary will be presented here. In studies of the influence of pH fermentations
are performed either using controlled pH or the pH is allowed to fall. Optimum pH values
vary from one species of LAB to another and, for a specific species, the pH optimum is
specific to that organism.
One aspect of EPS production that has only recently received attention is the degradation
of EPS that is observed after extended fermentations. Glycosyl-hydrolase activities, that
reduce the molecular mass, have been observed in fermentations of Lb. rhamnosus R (Pham
et al., 2000). A rapid reduction in molecular mass would require an endoglucanase to be
active in the cell supernatant. The influence that the pH of the medium has on EPS yield will
be dependent on the pH optimum of the glycosyl-hydrolases. The pH optimum for EPS
production will be that pH at which the opposing effects of production and degradation are
balanced. The effect of medium composition on EPS production is very marked. The
influence of medium components on EPS yields has received a great deal of attention and
has recently been reviewed (Ricciardi and Clementi, 2000). In our laboratories, the most
convenient medium for production of EPS is either unsupplemented skim milk or skim milk
supplemented with small proportions of casein hydrolysate. Attempts to isolate and character-
ise EPS from broth media or media to which peptone or yeast have been added as sources of
added nitrogen have been problematic. Polymannans present in broth, yeast, and peptone are
extracted with the EPS providing enhanced yields and monomer ratios that include high
percentages of mannose.
Only a limited number of studies of the kinetics of EPS production have been carried out
(Gassem et al., 1997a,b; Bouzar et al., 1996; de Vuyst et al., 1998; Kimmel and Roberts,
1998; Gancel and Novel, 1994a,b). Production of EPS appears to occur during the
logarithmic phase and, for some LAB, continues into the stationary phase.
3.2. Isolation of polysaccharides
Isolation of polysaccharide is required for their characterisation, and the procedure should
not alter the chemical and physical properties of the polysaccharide. Two different isolation
procedures are in common use. The only substantial difference between the two methods is
the way in which milk proteins, caseins, are removed. The first method to be reported is due
to Cerning et al. (1986, 1988) and utilises pronase to hydrolyse caseins. In the second
A. Laws et al. / Biotechnology Advances 19 (2001) 597–625 607
procedure, cell material and milk caseins are precipitated from the culture medium on the
addition of trichloroacetic acid (Garcia-Garibay and Marshall, 1991). These extraction
procedures have not been optimised and a number of workers use slight variants:
particularly in regard to the final concentration of added trichloroacetic acid that varies
between 4% and 20%.
There are a number of reports of LAB that synthesise mixtures of EPSs. The EPSs can
have different structures: Duenas-Chasco et al. (1998) have supplied strong evidence for a
strain that secretes two homopolysaccharides having different structures. There is also the
possibility of recovering EPS samples that have identical structure but different molecular
masses (Degeest and de Vuyst, 1999). The literature available describing the ability of LAB
to secrete mixtures of heteropolysaccharides of different structure should be read carefully.
Frequently, the only evidence used to support such mixtures comes from analysis of the
monomer composition of samples isolated by fractionation. Great care should be taken when
interpreting such data. This is particularly true when small samples are being analysed (see
discussion above). The preparation and isolation procedures may well provide contaminating
polysaccharides that are either originally present in the medium, have been added as yeast or
peptone supplements or that have been removed from cell walls either as a result of
mechanical disruption or lysis during fermentations.
The results of monomer composition ratios that are determined in chemically defined
media are more reliable (Petry et al., 2000). Grobben et al. (1997) have used a defined
medium to grow Lb. delbrueckii ssp. bulgaricus NCFB 2772. When grown on glucose or
fructose, two EPSs are isolated (Grobben et al., 1996): high-molecular mass polysaccharide
and a low-molecular mass polysaccharide. The monomer composition and linkages reported
for the large molecular mass polysaccharides derived from glucose and fructose are, to a first
approximation ( ± 5%), the same (Grobben et al., 1997). However, the low-molecular mass
polysaccharides, which are produced in small amounts, do appear to have a different
monomer composition.
As stated earlier, evidence confirming that a single LAB culture produces two EPSs that
have different repeat unit structures is only available for Lactobacillus spp. G-77 (Duenas-
Chasco et al., 1998) an organism that secretes two homopolysaccharides. Both monomer
composition and NMR spectral data are available for each of the EPSs. NMR evidence to
confirm that a single LAB culture produces more than one heteropolysaccharide repeat unit is
currently not available. To date there have been no complete structural studies published that
show that LAB are able to secrete more than one heteropolysaccharide that have different
structures. On a number of occasions we have observed NMR spectra that indicate that more
than one type of EPS is present; however, by the use of careful controls (performing
fermentations and isolation procedures without bacteria), we have demonstrated that the
spectra contain resonances that are characteristic of the media (M17) used to culture the
bacteria (S. thermophilus).
In contrast, there is substantial evidence for the synchronous production of EPSs having
the same structure but of different molecular mass. Degeest and de Vuyst (1999) reported the
production of a high-molecular mass (1.8� 106) and a low-molecular mass EPS (4.1�105)
by S. thermophilus LY03. The production of two polysaccharides by Lb. rhamnosus has
A. Laws et al. / Biotechnology Advances 19 (2001) 597–625608
recently been reported (Pham et al., 2000); however, the low-molecular mass material is
believed to be generated by the glycosyl-hydrolase-catalysed hydrolysis of high-molecular
mass products.
3.3. Characterisation of EPSs
Before a polysaccharide can be considered to be fully characterised, it is necessary to
determine information about the molecular mass of the material, to identify the composition
and absolute configuration of the monomers, and finally, to determine the linkage pattern of
the monomers.
A number of methods for determining the molecular mass of polysaccharides are available.
Historically, retention times, as determined by size exclusion chromatography using refractive
index detection, have been used to determine molecular mass (Cerning et al., 1986).
However, the majority of EPS samples have very large molecular masses and elute close
to the exclusion limits of the available stationary phases and very much in excess of the
highest molecular mass standards used to calibrate the columns. Combination detectors
(Williams et al., 1992; Tinland et al., 1988), using light scattering for measurement of
molecular mass (Wyatt, 1993), provide a more accurate average molecular mass.
A number of different methods are available for determining the monomer composition of
EPS samples. Methanolysis and per-trimethylsilylation provides samples that can be analysed
by GLC. A simpler method, requiring acid hydrolysis followed by monomer detection using
high-pressure anion exchange chromatography with pulsed ampometric detection, has
recently been introduced (Cataldi et al., 2000; Hanko and Rohrer, 2000). Absolute config-
urations of monomers are determined by preparation of their per-trimethylsilyl (� )butyl
glycosides using the procedures described by Gerwig et al. (1978, 1979).
The linkage pattern of the monomers is determined using a combination of ‘methylation’
analysis (Stellner et al., 1973; Sweet et al., 1975) and NMR spectroscopy. There have been a
very large number of reviews covering the use of NMR spectroscopy in carbohydrate
chemistry, and the most recent by Duus et al. (2000) lists the reviews published between 1992
and 1999. Many of these have collated reference spectroscopic data, chemical shifts, and
coupling constants, and are extremely useful for assigning spectra of complex carbohydrates.
A review describing the application of 2D-NMR in determining the primary structure of
bacterial EPSs has recently been published (Leeflang et al., 2000).
In summary, 1D-1H and 13C spectra are recorded for solutions of EPS samples in
deuterium oxide and spectra are recorded at temperatures of 70 �C or above: the high
sample temperature reduces the viscosity of the medium and shifts the residual proton solvent
signal up-field and away from important resonances. Vliegenthart et al. (1983) have
suggested that spectra should be viewed as having ‘structural reporter signals’ and a ‘bulk
region’ (3–4 ppm). In proton spectra the low-field reporter region includes resonances from
the anomeric and ring atoms shifted out of the bulk region as a consequence of glycosylation.
There is also a high-field reporter region including resonances from ring substituents such as
acyl, alkyl, and acetal, and ring substitutions such as N-acetylamino and H6 of 6-deoxy
sugars. Homonuclear 2D spectra are used to assign protons resonances to individual rings
A. Laws et al. / Biotechnology Advances 19 (2001) 597–625 609
Table 1
Structures of the EPS repeat unit isolated from the genus Lactococci (Lactococcus lactis)
Group Origin Author Structure
L-3A B891 van Casteren et al.
(2000a,b)
L-3B cremoris
H414
Gruter et al. (1992)
L-3C SBT0495 Nakajima et al. (1992)
B40 van Kranenburg et al.
(1997)
cremoris
ARH74
Yang et al. (1999)
L-5A B39 van Casteren et al.
(2000a)
(Ac)0.5#6
b-D-Galp-(1! 4)-b-D-Glcp1
#6
! 4)-b-D-Galp-(1! 4)-b-D-Glcp-(1! 4)-a-D-Glcp-(1!
b-D-Galp-(1! 3)-b-D-Glcp1
#3
! 4)-b-D-Galp-(1! 3)-b-D-Galp-(1! 4)-a-D-Galp-(1!
a-D-Galp-1-phosphate#3
! 4)-b-D-Glcp-(1! 4)-b-D-Glcp-(1! 4)-b-D-Galp-(t!2
"1
a-L-Rhap
b-D-Galp-(1! 4)-b-D-Glcp1
#4
! 2)-a-L-Rhap-(1! 2)a-D-Galp-(1! 3)-a-D-Glcp-(1! 3)-a-D-Galp-(1! 3)-a-L-Rhap-(1!
A.Lawset
al./Biotech
nologyAdvances
19(2001)597–625
610
(COSY and TOCSY) and heteronuclear 2D spectra, 13C–1H, are used to assign carbons
(HMQC, HMBC) and to obtain linkage information (HMBC). Further information about
linkage, using nonscalar coupling, is available from ROESY spectra.
A number of techniques that have recently been applied to carbohydrate analysis but that
have not yet been routinely applied to the analysis of EPSs are worth highlighting. Vincent
and Zwahlen (2000) have reported the use of spin diffusion to provide information relating to
linkage and the technique, which requires shaped pulses, is complementary to the hetero-
nuclear HMBC experiments. Navarini et al. (2001) have reported the use of deuterium-
induced differential isotope shifts in the determination of linkage type. The latter technique is
particularly valuable when 2D experiments and shaped pulses are not available. A method for
undertaking linkage analysis by direct inspection of the NMR of oligosaccharides that have
been per-acetylated using 13C (CO)-labelled acetic anhydride has been reported (Bendiak,
1999a,b). The latter experiment, if applicable to polysaccharides, may remove the need to
undertake classical methylation analysis.
At the time of writing, this review the complete structural analysis, for polymers that
have been described both as being secreted EPSs and as heteropolysaccharides, have been
reported for 6 lactococci (Table 1), 21 streptococci (Table 2), and 19 lactobacilli (Table 3).
A small but significant number of LAB secrete the same EPS and in total 25 unique
structures have been reported. Each genus, Lactobacillus, Streptococcus, and Lactococcus
can be divided into groups according to the EPS they synthesise (see Tables 1, 2, and 3,
respectively). A convenient method for assigning EPSs from the different taxa into groups
is based, in the first instance, solely on the number of monosaccharides present in the
backbone of the oligosaccharide repeat unit. Each grouping in the table has been coded
using initials that identify the genus (S for Streptococcus, L for Lactococcus, and LB for
Lactobacilli). For lactobacilli, species are identified by an additional initial, e.g., H for
helveticus. The initial is followed by a number that reflects the number of residues in the
main chain of the repeat unit. Similarities between the EPS structures can only be easily
identified by visual inspection if a common mode of writing structures is adopted. The
latter requires the adoption of a set of rules for presenting structures. In the author’s
laboratory, the following rules have been developed and are loosely derived from the
guidelines presented by the International Union of Pure and Applied Chemistry (IUPAC)
for the nomenclature of polysaccharides composed of more than one kind of residue
(2-Carb-39.7):
1. The repeat unit is drawn such that the sugar residues of the back bone (principal chain)
is on the horizontal axis and the residue of highest priority is on the right-hand side with
the anomeric carbon at the end (see below for suggestions for assigning priorities to
sugar residues). The principal chain is drawn using standard IUPAC abbreviated
nomenclature for naming residues and for identifying linkage patterns.
2. If a single branch is attached to a sugar residue it is drawn above the principal chain.
3. If more than one branch is attached to a sugar residue the highest priority branch is
drawn above the main chain and the lower priority branch is placed below (see below
for suggestions for assigning priorities to branches).
A. Laws et al. / Biotechnology Advances 19 (2001) 597–625 611
Table 2
Structures of the EPS repeat unit isolated from the genus Streptococcus
Family Origin Author Structure
S-3A CNCMI 733,
734, 735
Doco et al.
(1990)
IMDO 01,
02, 03
and NCFB
859
Marshall et al.
(2001a,b)
EU 21 Laws
(unpublished)
Sfi 6 Stingele et al.
(1999)
S-3B Sfi 39 Lemoine et al.
(1997)
SY 89 Marshall et al.
(2001a,b)
SY 102
CH101 Laws
(unpublished)
S-3C Sc136 Vincent et al.
(2001)
a-D-Galp1
#6
! 3)-a-D-GalpNAc-(1! 3)-b-D-Galp-(1! 3)-b-D-Glcp-(1!
b-D-Galp1
#6
! 3)-b-D-Galf-(1! 3)-a-D-Glcp-(1! 3)-b-D-Glcp-(1!
b-D-Galf-(1! 6)-b-D-Glcp-(1! 6)-b-D-GlcpNAc1
#3
! 4)-b-D-Glcp-(1! 4)-a-D-Glcp-(1! 4)-b-D-Galp-(1!
A.Lawset
al./Biotech
nologyAdvances
19(2001)597–625
612
S-5A Sfi 12 Lemoine et al.
(1997)
S-5B OR901 Bubb et al.
(1997)
Rs and
Sts
Faber et al.
(1998)
S-5C S3 Faber et al.
(2001b)
S-6A EU20 Marshall et al.
(2001a)
b-D-Galp-(1! 6)-b-D-Galp1
#4
! 2)-a-L-Rhap-(1! 2)-a-D-Galp-(1! 3)-a-D-Galp-(1! 3)-a-D-Galp-(1! 3)-a-L-Rhap-(1!
Ac0.4#2
b-D-Galf1
#6
! 3)-a-L-Rhap-(1! 2)-a-L-Rhap-(1! 2)-a-D-Galp-(1! 3)-b-D-Galp-(1! 3)-a-D-Galp-(1!
a-L-Rhap1
#2
! 4)-b-D-Glcp-(1! 6)-a-D-Galf-(1! 6)-b-D-Glcp-(1! 6)-b-D-Galp-(1! 6)-a-D-Galp-(1! 3)-b-L-Rhap-(1!
b-D-Galp1
#4
! 2)-a-L-Rhap-(1! 2)-a-D-Galp-(1! 3)-a-D-Glcp-(1! 3)-a-D-Galp-(1! 3)-a-L-Rhap-(1!A.Lawset
al./Biotech
nologyAdvances
19(2001)597–625
613
Table 3
Structures of the EPS repeat unit isolated from the genus Lactobacillus
Group Origin Authors Structure
Lactobacillus
helveticus
LB-H-3A K16 Yang et al.
(2000)
LB-H-4A TN-4 Yamamoto
et al. (1995)
Lh-59 Stingele
et al. (1997)
LB-H-4B TY1-2 Yamamoto
et al. (1994)
LB-H-5A 2091 Staaf et al.
(1996)
b-D-Glcp1
#2
b-D-Galp-(1! 4)-b-D-Glcp1
#6
! 4)-b-D-Galp-(1! 4)-b-D-Glcp-(1! 4)-a-D-Glcp-(1!
b-D-Galp-(1! 4)-b-D-Glcp1
#3
! 3)-a-D-Galp-(1! 3)-a-D-Glcp-(1! 3)-b-D-Glcp-(1! 5)-b-D-Galf-(1!
b-D-Galp-(1! 4)-b-D-Glcp1
#6
! 6)-b-D-Glcp-(1! 3)-b-D-Glcp-(1! 6)-a-D-GalpNAc-(1! 3)-b-D-Galp-(1!4
"a-D-Galp0.8
b-D-Galp1
#6
! 4)-b-D-Glcp-(1! 6)-b-D-Glcp-(1! 6)-b-D-Galp-(1! 4)-a-D-Galp-(1! 3)-b-D-Galp-(1!
A.Lawset
al./Biotech
nologyAdvances
19(2001)597–625
614
LB-H-5B Lb 161 Staaf et al.
(2000)
LB-H-5C 766 Robijn
et al. (1995)
Lactobacillus
LB-A-4A Lb.
acidophilus
LMG 9433
Robijn et al.
(1996a)
LB-D-3A Lb.
delbrueckii
ssp.
bulgaricus
291
Faber et al.
(2001a,b)
LB-D-4A Lb.
delbrueckii
ssp.
bulgaricus
RR
Gruter et al.
(1993)
EU03,
EU24,
EU25
Marshall
et al.
(2001a,b)
(continued on next page)
b-D-Glcp b-D-Glcp1 1
# #3 3
! 3)-b-D-Glcp-(1! 4)-a-D-Glcp-(1! 4)-b-D-Galp-(1! 3)-a-D-Galp-(1! 2)-a-D-Glcp-(1!
b-D-Galf1
#3
! 6)-a-D-Glcp-(1! 6)-a-D-Galp-(1! 6)-a-D-Glcp-(1! 3)-b-D-Glcp-(1! 4)-b-D-Glcp-(1!
b-D-GlcpNAc1
#3
! 4)-b-D-Glcp-(1! 4)-b-D-GlcpA-(1! 6)-a-D-Glcp-(1! 4)-b-D-Galp-(1!
b-D-Galp-(1! 4)-b-DGlcp1
#6
! 4)-b-D-Galp-(1! 4)-b-D-Glcp-(1! 4)-a-D-Glcp-(1!
b-D-Galp a-L-Rhap b-D-Galp1 1 1
# # #4 3 3
! 3)-b-D-Glcp-(1! 3)-b-D-Galp-(1! 4)-a-D-Galp-(1! 2)-a-D-Galp-(1!
A.Lawset
al./Biotech
nologyAdvances
19(2001)597–625
615
LB-P-4A Lb.
paracasei
34-1
Robijn et al.
(1996b)
LB-P-4B Lb.
paracasei
van Calsteren
(2001)
two strains
LB-R-6A Lb.
rhamnosus
RW 9595
M and R
van Calsteren
et al.
Table 3. (continued )
Group Origin Authors Structure
sn-glycerol-3-phosphate
#3
! 6)-b-D-Galp-(1! 6)-b-D-Galp-(1! 3)-b-D-GalpNAc-(1! 4)-b-D-Galp-(1!
a-D-Galp1
#6
! 6)-a-D-Galp-(1! 3)-b-L-Rhap-(1! 4)-b-D-Glcp-(1! 4)-b-D-GlcpNAc-(1!3
"1
a-L-Rhap
HOOC 4
R a-D-GalpH3C 6 1
#2
! 3)-a-L-Rhap-(1! 3)-a-L-Rhap-(1! 2)-a-D-Glcp-(1! 3)-a-L-Rhap-(1! 3)-b-D-Glcp-(1! 3)-a-L-Rhap-(1!
A.Lawset
al./Biotech
nologyAdvances
19(2001)597–625
616
LB-R-5A Lb.
rhamnosus
C83
Vanhaverbeke
et al. (1998)
LB-R-5B Lb.
rhamnosus
GG
Landersjo
et al. (2001)
LB-S-3A sake 0–1 Robijn et al.
(1996a,b)
! 6)-a-D-Galp-(1! 6)-a-D-Glcp-(1! 3)-b-D-Galf-(1! 3)-a-D-Glcp-(1! 2)-b-D-Galf-(1!
b-D-Galf1
#6
! 3)-a-L-Rhap-(1! 3)-a-D-Galp-(1! 3)-b-D-Galf-(1! 3)-b-D-Galp-(1! 4)-a-D-Glcf NAc-(1!
sn-glycerol-3-phosphate! 4)-a-L-Rhap(Ac)0.85 1
# #2 3
! 3)-b-L-Rhap-(1! 4)-b-D-Glcp-(1! 4)-a-D-Glcp-(1!6
"1
b-D-Glcp
A.Lawset
al./Biotech
nologyAdvances
19(2001)597–625
617
The different sugar residues of the principal chain are assigned priorities as follows:
1. The sugar residue (including its attendant substituents and branches) of highest mass has
highest priority.
2. For sugar residues that have equal priority, after application of Rule 1, furanose sugars
are assigned priority over pyranose.
3. For sugar residues that have equal priority, after application of Rules 1 and 2, the residue
having linkages (locants) closest to the anomeric carbon has highest priority.
The different branches are assigned priorities as follows:
1. The branch of highest mass has highest priority.
2. If two branches are of equal mass then the branch linked closest to the anomeric carbon
has highest priority.
Whilst it is appreciated that the guidelines presented above will not allow definitive
structures to be drawn for all the different possible combinations of EPS structures they are
extremely useful for observing similarities between the different EPS structures and for
attempting to group structures into ‘families’.
The structures presented in the tables in this article are drawn using the guidelines.
Inspection of the different repeat unit structures (Tables 1–3), allows a number of conclusions
to be drawn regarding the frequency at which structural elements are observed:
� Taking both furanose and pyranose rings into account, the monosaccharide present in
highest frequency is galactose followed closely by glucose and then, at a much lower
frequency, rhamnose.� GalNAc, GlcNAc, and GlcA are observed in a small number of structures; as are the
phospho-diester, acetyl-ester, phosphate-ester, and pyruvate-acetal substitutions.� Without exception glucose and galactose have D-absolute configuration whereas
rhamnose has L-absolute configuration.� For glucose and galactose, there is a small preference for the b-anomer; however, for
rhamnose there is a definite preference for a-rhamnose: the ratio of the two anomers is
approximately 4a to 1b.� The only sugar that is found to adopt a furanose ring is galactose.
The frequency with which the different linkages, anomeric configuration, and linkage type,
are found is worthy of comment and, again, a number of conclusions can be drawn:
� A large number of branches are terminated by the addition of galactose to the branch,
over 75%.� b-D-Gal is preferentially attached either as a branch terminus or linked to the chain via
its 3-, 4-, or 6-hydroxyl group. In contrast, a-D-Gal shows a strong preference for
attachment through its 3-hydroxyl group.
A. Laws et al. / Biotechnology Advances 19 (2001) 597–625618
� b-D-Glc is preferentially attached via either its 3-, 4-, or 6-hydroxyl group whilst a-D-Glc is preferentially attached via either its 3- or 6-hydroxyl group.
� Rhamnose is frequently used as a branch junction.� In bonding with sugars of the main chain b-Rha is very similar to a-D-Gal and prefers
attachment via either the 2- or 3-hydroxyl group.
In viewing the structures represented in Tables 1–3, it is clear that several of the structures
possess backbones that are very closely related. The EPSs isolated from two distantly related
species of Lactobacillus have identical structures: the main chain for the EPS isolated from
species helveticusK16 (group LB-H-3A) is the same as that derived from species L. lactisB891
(group L-3A). The main chain for the EPS isolated from species S. thermophilus Sfi 12 (group
S-5A) is the same as that derived from species L. lactis B39 (group L-5A). A number of EPSs
have main chains that differ only in the substitution of single sugar residues, e.g., the EPSs from
S. thermophilus groups S-5A and S-5B differ by substitution of a glucose to galactose.
Marshall et al. (2001a,b) have recently reported the genetic typing of groups of strains that
produce the same EPS structure. Strains from S. thermophilus group S-3A (IMDO1, IMDO2,
IMDO3, NCFB859, and EU21) and strains from Lactobacilli group LB-D-4A (EU03, EU24,
and EU25) were genetically typed using restriction endonuclease analysis of total DNA and
random amplification of polymorphic DNA. A number of S. thermophilus strains producing
different EPS structures were included in the analysis for comparison (group S-3B SY89 and
SY102). The profiles showed that those LAB secreting the same EPS were also grouped by
the genetic typing.
4. EPS engineering and design
The commercial exploitation of EPSs, as materials for enhancing the texture and mouthfeel
of food, requires the synthesis of EPS having desirable physical properties and for the EPS to
be available in sufficient quantities to match demand. Given that in most cases the desired
contribution of the EPS is to add texture, i.e., to provide thickening properties, it is necessary
to understand how structural components such as linkage, monomer type, substituents, and
molecular mass influence intrinsic viscosity.
In aqueous solutions, most EPSs can be described as random coil polymers. Navarini et al.
(2001) have described experiments investigating the rheological properties of solutions of the
EPS from S. thermophilus SFi 20 and have characterised the EPS as being composed of
random coils as opposed to ordered or rigid polymers. Tuinier et al. (1999a) have reported
results of multiangle light scattering experiments that are consistent with solutions of EPS
derived from L. lactis ssp. cremoris being composed of random coils. As the name suggests,
random coil polymers have no fixed shape and have randomly fluctuating tertiary structure. It
has also been demonstrated that the viscosity of solutions of random coil EPS polymers is
related to the concentration of the EPS and the specific volume of the polysaccharide in
solution (Tuinier et al., 1999b). High specific volumes occur for EPSs having large chain
lengths and for EPSs that are composed of linkages that are ‘stiff’. The general features
A. Laws et al. / Biotechnology Advances 19 (2001) 597–625 619
relating to the flexibility of different linkages has been described (Rees, 1977; Lapasin and
Pricl, 1995) b(1–4) links impart stiffness compared to b(1–3) and b(1–2), a-linkages are
usually more flexible than b-linkages. It is worth noting at this point that the main chains
derived from strains of S. thermophilus having large repeat units appear to have been
constructed with a view to providing maximum flexibility, a(1–2) and a(1–3) links dominate.
The influence that chain stiffness has on the specific volume can be seen by
comparison of the EPSs from L. lactis NIZO B891 and L. lactis NIZO B40. An EPS
from L. lactis NIZO B891 that has an average molar mass of 2.4� 106 has a smaller
radius of gyration than that measured for an EPS of average molecular mass of 1.4� 106
derived from the L. lactis NIZO B40 (Tuinier, 1999). The structures of the two EPSs have
been described. The backbone of L. lactis NIZO B40 is composed entirely of b(1–4)linked residues and will have a rigid structure and a high specific volume. In contrast, the
back bone of L. lactis NIZO B891 has both an a-linked residue and a (1–6) link that will
impart a degree of flexibility to the backbone and hence the EPS would be expected to
have a smaller radius of gyration. The latter correlation is extremely crude: the structures
of the two EPSs are too varied and the influence that the flexibility that the main chain
might have on the radius of gyration cannot be proven with certainty. More work is
required in this area; knowledge of the structures of a larger number of EPSs will permit
studies of the effect that systematic variation of structural components, such as branches
and substituents, has on the radius of gyration of the EPS.
In considering intrinsic viscosity of solutions of EPSs in isolation, if the intention is to
increase their intrinsic viscosity, then an ideal polysaccharide would be one that has a high
molar mass and a rigid structure. In real systems, it is necessary to consider the influence that
other media components, e.g., proteins and salts, have on the conformation and distribution
of EPS.
4.1. Engineering of EPS structures
One of the main problems encountered in the use of EPSs in texture enhancement is related
to the quantity of material available: production levels are very variable ranging from a few
milligrams to approaching 1 g/l for high producers grown on supplemented media. Given that
the biosynthetic pathways leading to EPS production and for sugar metabolism are known,
given that many of the factors involved in controlling and in the regulation of flux through the
pathway are understood, opportunities arise for engineering of EPSs. Engineering has focused
on increased production of EPS and the production of ‘designer’ polysaccharides.
4.2. Production of designer EPSs
Almost all of the work directed at the engineering of EPS structures has focused on the
genes coding for the glycosyl-transferases. van Kranenburg et al. (1999) have reported the
production of a nonpolar disruption in the gene coding for a priming glycosyl-transferase,
the epsD gene of L. lactis NIZO B40, which results in loss of EPS production. They also
report that a homologous insertion of an epsD gene into the mutated L. lactis NIZO B40
A. Laws et al. / Biotechnology Advances 19 (2001) 597–625620
results in recovered EPS production. Heterologous complementation with the gene
encoding for the primary glucosyltransferase involved in the synthesis of a capsular
polysaccharide of S. pneumonie serotype 14 (cps14E gene) restored EPS synthesis (van
Kranenburg et al., 1999). However, homologous or heterologous insertion of genes coding
for priming galactosyl-transferases resulted in failure to produce EPS synthesis. The failure
to observe EPS synthesis when complementation is with a galactosyl-transferase suggests
that the activities of the transferases, involved in synthesis of the oligosaccharide repeat
unit, is restricted to specific lipid-bound acceptor sugars.
In developing the use of heterologous gene insertion to effect EPS biosynthesis, Stingele
et al. (1999) have expressed the entire gene cluster from S. thermophilus SFi 6 into a non-EPS
producer L. lactis MG1363 and observed limited production of EPS. More importantly,
the structure of the EPS was different to that generated by SFi 6. An a 1! 6 linked branching
a-D-Galp was absent and a main chain a-D-galactose replaced the a-D-N-acetylgalactosamine.
It is worth noting at this point that heterologous exchange of EPS gene clusters from one Gram-
negative bacterium to another Gram-negative bacterium has also been performed. Pollock et al.
(1997) have placed those genes responsible for xanthan gum formation in X. campestris into
bacteria from the genus Sphingomonas and observed xanthan gum synthesis.
4.3. Increased production of EPS
Increased production of EPS can be brought about by either genetic manipulation or by
control of microbial physiology; rerouting carbon metabolism towards EPS biosynthesis. de
Vuyst and Degeest (1999) have recently reviewed the influence that microbial physiology has
on EPS production and, as such, this topic will not be covered here. In principle, genetic
manipulation can be used to modify the activities of any of the enzymes involved in the
biosynthetic pathway. In order to optimise EPS production, a detailed knowledge of the flux
of metabolites in each of the enzyme-catalysed transformations is required. A number of
workers are currently investigating if the flux of carbon feed to the sugar nucleotides can be
altered either by deletion of genes coding for key enzymes, e.g., the PGMs or by functional
overexpression of the UDP-glucophosphorylases and/or uridyl-transferases.
In conclusion, LAB bacteria provide scope as cell factories for the synthesis of novel
polysaccharides for use in the food processing industry. Further work in determining the
mechanisms responsible for the control and regulation of EPS biosynthesis, both at the
level of genes and of proteins, is required before EPS production can be optimised. Work
is currently in progress in these areas and it will not be long before new ‘designer’
polysaccharides are being reported. The synthesis of a ‘designer’ polysaccharide has
recently been reported (Colquhoun et al., 2001), they inactivated the aceP gene of
Acetobacter xylinum: the aceP gene codes for a glucosyl-transferase required in the
construction of a branching side chain. The EPS synthesised by the genetically modified
bacteria had a truncated branch. The results complement those of Stingele et al. (1999) and
suggest that the transferase enzymes constructing the main chain must either act before the
branch is added or, alternatively, the transferase enzymes do not recognise the presence of
the branch. They also indicate that the polymerase and export mechanisms are able to work
A. Laws et al. / Biotechnology Advances 19 (2001) 597–625 621
with oligosaccharides and polysaccharides having modified structures. These results are
extremely encouraging and provide strong impetus for further work to be undertaken in
this area.
References
Azeredo J, Oliveira R. Biofouling 2000;16:17–27.
Banik RM, Kanari B, Upadhyay SN. World J Microbiol Biotechnol 2000;16:407–14.
Becker A, Niehaus K, Puhler A. Mol Microbiol 1995;16:191–203.
Bendiak B. Carbohydr Res 1999a;315:206–21.
Bendiak B. Carbohydr Res 1999b;321:139.
Boels IC, Kleerebezem M, Hugenholtz J, de Vos WM. In: Proceedings 5th ASM on the Genetics and Molecular
Biology of Streptococci, Enterococci and Lactococci, 1988;66.
Bouzar F, Cerning J, Desmazeaud M. J Dairy Sci 1996;79:205–11.
Bubb WA, Urashima T, Fujiwara R, Shinnai T, Ariga H. Carbohydr Res 1997;301:41–50.
Cataldi TRI, Campa C, De Benedetto GE. Fresenius J Anal Chem 2000;368:739–58.
Cerning J. FEMS Microbiol Rev 1990;87:113–30.
Cerning J, Marshall VM. Recent Res Dev Microbiol 1999;3:195–209.
Cerning J, Bouillanne C, Desmazeaud MJ, Landon M. Biotechnol Lett 1986;8:625–8.
Cerning J, Bouillanne C, Desmazeaud MJ, Landon M. Biotechnol Lett 1988;10:255–60.
Cerning J, Bouillanne C, Landon M, Desmazeaud M. J Dairy Sci 1992;75:692–9.
Colquhoun IJ, Jay AJ, Eagles J,Morris VJ, Edwards KJ, Griffin AM, GassonMJ. Carbohydr Res 2001;330:325–33.
Degeest B, de Vuyst L. Appl Environ Microbiol 1999;65:2863–70.
Degeest B, de Vuyst L. Appl Environ Microbiol 2000;66:3519–27.
Delcour J, Ferain T, Deghorain M, Palumbo E, Hols P. Antonie van Leeuwenhoek 1999;76:159–84.
Denadra MCM, Desaad AMS. Int J Food Microbiol 1995;27:101–6.
de Vos WM. Antonie van Leeuwenhoek 1996;70:223–42.
de Vos WM, Vaughan. Genetics of lactulose utilisation in Lactic acid bacteria. FEMS Microbiol Rev 1994;
15:217–317.
de Vuyst L, Degeest B. FEMS Microbiol Rev 1999;23:153–77.
de Vuyst L, Vanderveken F, Van de Ven S, Degeest B. J Appl Microbiol 1998;84:1059–68.
Doco T, Wieruszeski JM, Fournet B, Carcano D, Ramos P, Loones A. Carbohydr Res 1990;198:313–21.
Duenas M, Irastorza A, Fernandez K, Bilbao A. J Food Prot 1995;58:76–80.
Duenas-Chasco MT, Rodriguez-Carvajal MA, Tejero-Mateo P, Espartero JL, Irastorza-Iribas A, Gil-Serrano AM.
Carbohydr Res 1998;307:125–33.
Duus JO, Gotfredsen CH, Bock K. Chem Rev 2000;100:4589–614.
Faber EJ, Zoon P, Kamerling JP, Vliegenthart JFG. Carbohydr Res 1998;310:269–76.
Faber EJ, Kamerling JP, Vliegenthart JFG. Carbohydr Res 2001a;331:183–94.
Faber EJ, van den Haak MJ, Kamerling JP, Vliegenthart JFG. Carbohydr Res 2001b;331:173–82.
Forsen R, Myllymaa R. Acta Univ Ouluensis A23 Biochem 1974;6:1–18.
Forsen R, Pakkila M. Acta Univ Ouluensis A78 Biochem 1979;22:1–13.
Friesen RHE, Knol J, Poolman B. J Biol Chem 2000a;275:33527–35.
Friesen RHE, Knol J, Poolman B. J Biol Chem 2000b;275:40658.
Gamar L, Blondeau K, Simonet JM. J Appl Microbiol 1997;83:281–7.
Gancel F, Novel G. J Dairy Sci 1994a;77:685–8.
Gancel F, Novel G. J Dairy Sci 1994b;77:689–95.
Garcia-Garibay M, Marshall VME. J Appl Bacteriol 1991;70:325–8.
Garcia-Ochoa F, Santos VE, Casas JA, Gomez E. Biotechnol Adv 2000;18:549–79.
A. Laws et al. / Biotechnology Advances 19 (2001) 597–625622
Gassem MA, Schmidt KA, Frank JF. J Food Sci 1997a;62:171–3.
Gassem MA, Sims KA, Frank JF. Food Sci Technol (Lebensm-Wiss Technol) 1997b;30:273–8.
Germond JE, Delley M, D’Amico N, Vincent SJF. Heterologous expression and characterization of the exopo-
lysaccharide from streptococcus thermophilius Sfi39. Eur J Biochem 2001;268(19):5149–56.
Gerwig GJ, Kamerling JP, Vliegenthart JFG. Carbohydr Res 1978;62:349–57.
Gerwig GJ, Kamerling JP, Vliegenthart JFG. Carbohydr Res 1979;77:1–7.
Giavasis I, Harvey LM, McNeil B. Crit Rev Biotechnol 2000;20:177–211.
Gonzalez JE, Semino CE, Wang LX, Castellano-Torres LE, Walker GC. Proc Natl Acad Sci USA 1998;95:
13477–82.
Grobben GJ, Smith MR, Sikkema J, Debont JAM. Appl Microbiol Biotechnol 1996;46:279–84.
Grobben GJ, van Casteren WHM, Schols HA, Oosterveld A, Sala G, Smith MR, Sikkema J, Debont JAM. Appl
Microbiol Biotechnol 1997;48:516–21.
Gruter M, Leeflang BR, Kuiper J, Kamerling JP, Vliegenthart JFG. Carbohydr Res 1992;231:273–91.
Gruter M, Leeflang BR, Kuiper J, Kamerling JP, Vliegenthart JFG. Carbohydr Res 1993;239:209–26.
Gunnewijk MGW, Poolman B. J Biol Chem 2000a;275:34080–5.
Gunnewijk MGW, Poolman B. J Biol Chem 2000b;275:34073–9.
Gunnewijk MGW, Postma PW, Poolman B. J Bacteriol 1999;181:632–41.
Hanko VP, Rohrer JS. Anal Biochem 2000;283:192–9.
Hutkins R, Morris HA, McKay LL. Appl Environ Microbiol 1985;50:777–80.
Jones BE, Ryu R, Yang ZH, Cave MD, Pogoda JM, Otaya M, Barnes PF. Am J Respir Crit Care Med 1997;156:
1270–3.
Kimmel SA, Roberts RF. Int J Food Microbiol 1998;40:87–92.
Kleerebezem M, van Kranenburg R, Tuinier R, Boels IC, Zoon P, Looijesteijn E, Hugenholtz J, de Vos WM.
Antonie van Leeuwenhoek 1999;76:357–65.
Kleerebezem M, Hols P, Hugenholtz J. Lactic acid bacteria as a cell factory: rerouting of carbon metabolism in
Lactococcus lactis by metabolic engineering. Enzyme Microb Tech 2000;26(9–10):840–8.
Lapasin R, Pricl S. Rheology of industrial polysaccharides: theory and applications. Sheffield, UK: Academic
Press, Blackie, 1995.
Leeflang BR, Faber EJ, Erbel P, Vliegenthart JFG. J Biotechnol 2000;77:115–22.
Lemoine J, Chirat F, Wieruszeski JM, Strecker G, Favre N, Neeser JR. Appl Environ Microbiol 1997;63:3512–8.
Liu D, Cole RA, Reeves PR. J Bacteriol 1996;178:2102–7.
Loesche WJ. Microbiol Rev 1986;50:353–80.
Lonvaud-Funel A. Antonie Van Leeuwenhoek 1999;76:317–31.
Macura D, Townsley PM. J Dairy Sci 1984;67:735–44.
Marshall VM, Rawson HL. Int J Food Sci Technol 1999;34:137–43.
Marshall VM, Dunn H, Elvin M, McLay N, Gu Y, Laws AP. Carbohydr Res 2001a;331:413–22.
Marshall VM, Laws AP, Gu Y, Levander F, Radstrom P, De Vuyst L, Degeest B, Vaningelem F, Dunn H, Elvin M.
Lett Appl Microbiol 2001b;32:433–7.
Morona R, Mavris M, Fallarino A, Manning PA. J Bacteriol 1994;176:733–47.
Mozzi F, Degiori GS, Oliver G, Devaldez GF. Milchwissenschaft (Milk Sci Int) 1995;50:307–9.
Nakajima H, Hirota T, Toba T, Itoh T, Adachi S. Carbohydr Res 1992;224:245–53.
Navarini L, Abatangelo A, Bertocchi C, Conti E, Bosco M, Picotti F. Int J Biol Macromol 2001;28:219–26.
Oda M, Hasegawa H, Komatsu S, Kambe M, Tsuchiya F. Agric Biol Chem (Tokyo) 1983;47:1623–5.
Oliveria DR. Biofilms—Science and Technology (Melo LF Bott TR Fletcher, M and Capdeville beds) Dordrecht:
Kluwer Academic Publishers, 1992.
Petry S, Furlan S, Crepeau MJ, Cerning J, Desmazeaud M. Appl Environ Microbiol 2000;66:3427–31.
Pham PL, Dupont I, Roy D, Lapointe G, Cerning J. Appl Environ Microbiol 2000;66:2302–10.
Pollock TJ, Mikolajczak M, Yamazaki M, Thorne L, Armentrout RW. Production of xanthan gum by Sphingo-
monas bacteria carrying genes from Xanthomonas campestri. J Ind Microbiol Biot 1997;19(2):92–7.
Poolman B. Energy transduction in lactic acid bacteria. FEMS Microbiol Revs 1993;12:125–8
A. Laws et al. / Biotechnology Advances 19 (2001) 597–625 623
Poolman B, Gunnewijk MGW, Knol J, Postma PW, Veenhoff LM. FASEB J 1997;11:1321.
Postma PW, Lengeler JW, Jacobson GR. Microbiol Rev 1993;57:543–94.
Poulsen LV. Food Sci Technol (Lebensm-Wiss Technol) 1999;32:321–6.
Qian N, Stanley GA, Hahn-Hagerdal B, Radstrom P. J Bacteriol 1994;176:5304–11.
Qian N, Stanley GA, Bunte A, Radstrom P. Microbiology (UK) 1997;143:855–65.
Ramos A, Boels IC, de Vos WM, Santos H. Appl Environ Microbiol 2001;67:33–41.
Rees DA. Polysaccahride shapes. Outline studies in biology. London: Chapman & Hall, 1977.
Reeves P. Trends Genet 1993;9:17–22.
Ricciardi A, Clementi F. Ital J Food Sci 2000;12:23–45.
Robijn GW, Thomas JR, Haas H, Vandenberg DJC, Kamerling JP, Vliegenthart JFG. Carbohydr Res 1995;276:
137–54.
Robijn GW, Gallego RG, Vandenberg DJC, Haas H, Kamerling JP, Vliegenthart JFG. Carbohydr Res 1996a;288:
203–18.
Robijn GW, Wienk HLJ, Vandenberg DJC, Haas H, Kamerling JP, Vliegenthart JFG. Carbohydr Res 1996b;285:
129–39.
Rozen R, Bachrach G, Gedalia J, Steinberg D. The role of fructans on dental biofilm formation by Streptococcus
sobrinus, Streptococcus mutans, Streptococcus gordonii and Actinomyces viscous. FEMS Microbiol Lett
2001;195(2):205–10.
Saxena IM, Brown RM, Fevre M, Geremia RA, Henrissat B. Multidomain architecture of beta-glycosol trans-
ferases-implications for mechanisms of action. J Bacteriol 1995;177(6):1419–24.
Schellhass SM. PhD thesis, University of Minnesota, USA, 1983.
Schiffrin EJ, Rochat F, Linkamster H, Aeschlimann JM, Donnethughes A. J Dairy Sci 1995;78:491–7.
Sjoberg A, Hahn-Hagerdal B. Appl Environ Microbiol 1989;55:1549–54.
Sliz P, Schorter KH, de Vos WM, Pai EF. Acta Crystallogr, Sect D: Biol Crystallogr 1996;52:1199–201.
Sliz P, Engelmann R, Hengstenberg W, Pai EF. Structure 1997;5:775–88.
Staaf M, Yang ZN, Huttunen E, Widmalm G. Carbohydr Res 2000;326:113–9.
Staff M, Widmalm G, Yang ZN, Huttunen E. Carbohydr Res 1996;291:155–64.
Stellner K, Saito H, Hakomori SI. Arch Biochem Biophys 1973;155:464–72.
Stingele F, Neeser JR, Mollet B. J Bacteriol 1996;178:1680–90.
Stingele F, Lemoine J, Neeser JR. Carbohydr Res 1997;302:197–202.
Stingele F, Vincent SJF, Faber EJ, Newell JW, Kamerling JP, Neeser JR. Mol Microbiol 1999;32:1287–95.
Stulke J, Hillen W. Naturwissenschaften 1998;85:583–92.
Sutcliffe IC, Tao L, Ferretti JJ, Russell RRB. J Bacteriol 1993;175:1853–5.
Sutherland IW. Adv Microbiol Physiol 1972;8:143–213.
Sutherland IW. Microbial EPS synthesis in surface carbohydrate of the prokaryotic cell. London: Academic
Press, 1977.
Sutherland IW. Annu Rev Microbiol 1985;39:243–70.
Sutherland IW. Trends Biotechnol 1998;16:41–6.
Sutherland IW. Microbial polysaccharide products. Biotechnol Genet Eng Revs 1999;16:217–29.
Sweet DP, Shapiro RH, Albersheim P. Carbohydr Res 1975;40:217–25.
Tao L, Sutcliffe IC, Russell RRB, Ferretti JJ. J Dent Res 1993;72:1386–90.
Thomas TD, Crow VL. Appl Environ Microbiol 1984;48:186–91.
Tinland B, Mazet J, Rinaudo M. Makromol Chem, Rapid Commun 1988;9:69–73.
Tuinier R. PhD thesis, Wageningen Agricultural University, Wageningen, 1999.
Tuinier R, Zoon P, Olieman C, Stuart MAC, Fleer GJ, de Kruif CG. Biopolymers 1999a;49:1–9.
Tuinier R, Zoon P, Stuart MAC, Fleer GJ, de Kruif CG. Biopolymers 1999b;50:641–6.
van Calsteren MR. In: de Vuyst L, Degeest B, editors. First International Symposium on Exopolysaccharides from
Lactic Acid Bacteria. 2001;13.
van Casteren WHM, de Waard P, Dijkema C, Schols HA, Voragen AGJ. Carbohydr Res 2000a;327:411–22.
van Casteren WHM, Dijkema C, Schols HA, Beldman G, Voragen AGJ. Carbohydr Res 2000b;324:170–81.
A. Laws et al. / Biotechnology Advances 19 (2001) 597–625624
van den Bogaard PTC, Kleerebezem M, Kuipers OP, de Vos WM. J Bacteriol 2000;182:5982–9.
Vanhaverbeke C, Bosso C, Colin-Morel P, Gey C, Gamar-Nourani L, Blondeau K, Simonet JM, Heyraud A.
Carbohydr Res 1998;314:211–20.
van Kranenburg R, Marugg JD, van Swam II, Willem NJ, de Vos WM. Mol Microbiol 1997;24:387–97.
van Kranenburg R, van Swam II, Marugg JD, Kleerebezem M, de Vos WM. J Bacteriol 1999;181:338–40.
Viana R, Monedero V, Dossonnet V, Vadeboncoeur C, Perez-Martinez G, Deutscher J. Mol Microbiol 2000;36:
570–84.
Vincent SJF, Zwahlen C. J Am Chem Soc 2000;122:8307–8.
Vincent SJF, Faber EJ, Neeser JR, Stingele F, Kamerling JP. Structure and properties of the expolysaccharide
produced by Streptococcus macedonicus Sc 136 glycobiology 2001;11(2):131–9.
Vliegenthart JFG, Dorland L, Vanhalbeek H. Adv Carbohydr Chem Biochem 1983;41:209–374.
Wang GS, Louis JM, Sondej M, Seok YJ, Peterkofsky A, Clore GM. EMBO J 2000;19:5635–49.
Whitfield C, Valvano MA. Adv Microb Physiol 1993;35:135–246.
Williams DL, Pretus HA, Browder IW. J Liq Chromatogr 1992;15:2297–309.
Wyatt PJ. Abstr Pap — Am Chem Soc 1993;206:40-CARB.
Yamamoto Y, Murosaki S, Yamauchi R, Kato K, Sone Y. Carbohydr Res 1994;261:67–78.
Yamamoto Y, Nunome T, Yamauchi R, Kato K, Sone Y. Carbohydr Res 1995;275:319–32.
Yang ZN, Huttunen E, Staaf M, Widmalm G, Tenhu H. Int Dairy J 1999;9:631–8.
Yang ZN, Staaf M, Huttunen E, Widmalm G. Carbohydr Res 2000;329:465–9.
A. Laws et al. / Biotechnology Advances 19 (2001) 597–625 625