Download Lactic acid bacteria as a cell factory: rerouting of carbon metabolism

Enzyme and Microbial Technology 26 (2000) 840 – 848
www.elsevier.com/locate/enzmictec
Lactic acid bacteria as a cell factory: rerouting of carbon metabolism in
Lactococcus lactis by metabolic engineering夞
Michiel Kleerebezema,b,*, Pascal Holsb, Jeroen Hugenholtza,b
a
Wageningen Centre for Food Sciences, NIZO Food Research, P.O. Box 20, 6710 AB Ede, The Netherlands
b
Microbial Ingredients Section, NIZO Food Research, P.O. Box 20, 6710 AB Ede, The Netherlands
Received 30 April 1999; accepted 25 September 1999
Abstract
Lactic acid bacteria display a relatively simple metabolism wherein the sugar is converted mainly to lactic acid. The extensive knowledge
of metabolic pathways and the increasing information of the genes involved allows for the rerouting of natural metabolic pathways by
genetic and physiological engineering. We discuss several examples of metabolic engineering of Lactococcus lactis for the production of
important compounds, including diacetyl, alanine and exopolysaccharides. © 2000 Elsevier Science Inc. All rights reserved.
Keywords: Lactic acid bacteria; Lactococcus lactis; Metabolic engineering; Pyruvate metabolism; NICE system
1. Introduction
Lactic acid bacteria (LAB) are used world-wide in the
industrial manufacture of fermented food products. Their
most important application in this respect is undoubtedly in
the dairy industry, where these micro-organisms are used to
convert milk or milk-derived products to an enormous variety of fermented dairy products. However, next to milk
also other unprocessed food materials, like meat and vegetables, are subjected to fermentation by LAB on an industrial scale. In milk these bacteria encounter lactose as the
major carbon source, but they have the capacity to use a
number of other mono- and disaccharide substrates. The
major product of these fermentations is lactic acid, which
plays a crucial role in protection of the final fermented
product against spoilage. Besides this acidification that acts
as a natural preservative effect, LAB metabolism is essential
for development of desired product properties like flavor,
夞 This work was supported by several funding programs of the European community: Contracts BIOT-CT94-3055, BIOT-CT96-0498, BIOTCT96-5093, and ERBFMBICT-950438.
* Corresponding author. Tel.: ⫹31-(0)318-659629; Fax: ⫹31-(0)318650400.
E-mail address: [email protected] (M. Kleerebezem).
Current address Université Catholique de Louvain, Unité de Genetique,
5 Place Croix du Sud, Louvain-la-Neuve B-1348, Belgium.
shelf-life, and texture. The metabolic conversions that generate these end-products vary widely, depending on the
lactic acid bacterium. Metabolic engineering strategies that
involve inactivation of undesired genes and/or overexpression of existing or novel ones have been used to create
rerouting of the metabolic fluxes by changing the energy
metabolism or the concentrations of metabolic intermediates or of existing or completely new end-products. To
generate a rational approach for these strategies, it is essential to understand the pathways that are manipulated, their
fluxes, control factors, and the genes involved. Knowledge
of the traditional pathways recognized in LAB and the
characterization of new ones, in combination with physiological studies, has led to an extensive insight in LAB
metabolism. Proper appreciation of the application potential
of LAB as cell factories is related to several features that are
shared by these micro-organisms. In general they have a
long history of safe use and grow rapidly on media that are
derived from milk and contain lactose as the major carbon
source, which is converted to a number of interesting metabolic end-products during fermentation. Moreover, LAB
are applied world-wide as starter cultures in industrial food
fermentations, implying that LAB with modified metabolic
pathways cannot only be used in bioreactors but also for in
situ production of flavor, texture, or health promoting compounds in food products or even in the gastro-intestinal tract
of consumers of food. Several other features of LAB, like
0141-0229/00/$ – see front matter © 2000 Elsevier Science Inc. All rights reserved.
PII: S 0 1 4 1 - 0 2 2 9 ( 0 0 ) 0 0 1 8 0 - 0
M. Kleerebezem et al. / Enzyme and Microbial Technology 26 (2000) 840 – 848
841
their relatively simple metabolism, limited biosynthetic capacity, and apparent lack of gene multiplicity, illustrate their
advantages for metabolic engineering purposes compared to
other microorganisms like yeast or fungi.
We will review several strategies that have been used to
reach (efficient) rerouting of carbon metabolism in LAB.
The focus will be on the best studied LAB, Lactococcus
lactis. We will highlight some important genetic tools as
well as their application in some recently achieved examples of efficient rerouting of carbon metabolism. In addition,
several examples will be given of metabolic engineering
strategies that are currently investigated and could result in
efficient cell factory systems in the near future.
2. Genetic tools for LAB
Over the past decades, advances in the genetics of LAB
have resulted in the development of a large variety of
genetic techniques, including transformation procedures, integration, vector, and gene expression systems [1]. A large
number of genes and operons from LAB have been characterized and functionally analyzed. In addition, the determination of the entire genome sequence of L. lactis was recently reported to be completed [2], whereas genome
sequencing of several other LAB, including those belonging
to the genera Lactobacillus and Streptococcus has been
initiated. Moreover, several sustainable and safe selection
systems have been developed that allow application in the
food industry and therefore are designated “food-grade” [3].
These developments generate sophisticated tools and a mass
of genetic information that can be used by LAB as a cell
factory.
Several recent reviews give an excellent overview of
both controlled and constitutive gene expression systems
that are available for L. lactis and other LAB [1, 4 – 6].
Therefore, the development and characteristics of individual
gene expression systems will not be discussed here. Nevertheless, two recently developed systems seem worth mentioning because of their proven applicability as tools for
genetic engineering to facilitate efficient rerouting of metabolic fluxes. The first system was developed by the construction and screening of synthetic promoters that were
based on the L. lactis consensus promoter elements (i.e.
⫺35 and ⫺10 hexamers; [1], but in which the intervening
spacer region had been randomized [7]. This approach resulted in a large set of constitutive promoters that display a
wide range of expression efficiencies [7]. It seems likely
that these promoters (or a subset of these promoters) might
also be functional in other LAB and could thus provide a
general system for constitutive gene expression in these
bacteria, which is a very useful characteristic in certain
metabolic engineering strategies where gene expression is
preferentially maintained at a stable level. Although these
constitutive promoters hold great promise, no specific example of their application has been reported so far, which is
Fig. 1. Schematic representation of the nisin controlled expression (NICE)
system developed in L. lactis. A, lack of expression of the gene cloned
under control of the nisA promoter (P*), when no nisin in present in the
environment. B, response of the same cell (via NisK-NisR dependent
signal transduction) after induction by the addition of nisin to the growth
medium, resulting in ‘massive’ expression of the cloned gene (geneX)
probably due to their recent development. The second system is also of specific interest because it allows controlled
gene expression in a wide variety of lactic acid and other
Gram-positive bacteria [8,9]. This system has been designated the nisin controlled expression (NICE) system as it is
based on the autoregulatory characteristics of the antimicrobial peptide nisin that is produced by several strains of L.
lactis (Fig. 1). Nisin acts as extracellular peptide pheromone
that is involved in activation of its own biosynthesis [10,11].
Signal transduction occurs via a typical bacterial two component signal transduction machinery, consisting of a membrane anchored sensor protein (NisK) and a cytoplasmic
response regulator protein (NisR). Upon interaction with
nisin, NisK is likely to autophosphorylate, followed by
phosphotransfer to NisR, thereby generating the activated
form of the latter component that activates transcription of
the target genes (Fig. 1). Two nisin-responsive promoters
have been identified, preceding the nisA and nisF genes
[12], and coupling of such a promoter to a gene of interest
allows activation of expression of that gene by the addition
of the food-grade inducer nisin to the growth medium, or in
nisin containing milk products. A linear dose-response re-
842
M. Kleerebezem et al. / Enzyme and Microbial Technology 26 (2000) 840 – 848
Fig. 2. Schematic representation of the lactose utilising pathway used by L. lactis, leading to pyruvate. Known enzymes are indicated using abbreviations
based on established genetic nomenclature. Triose-P represents both glyceraldehyde-3P and dihydroxyacetone-P, from which the latter is converted to
glyceraldehyde-3P as indicated. Links to other figures in this article are indicated.
lationship between the amount of inducer (nisin) and the
level of nisA promoter-driven gene expression allows modulation of gene expression in a dynamic range of approximately 1000-fold. Moreover, due to the nisA promoter silence in the absence of nisin, this system allows the
controlled expression of lethal genes [13,14]. By introduction of the signal transduction system encoding genes,
nisRK, in other lactic acid bacteria, the NICE system was
functionally implemented in Lactococcus, Lactobacillus,
Leuconostoc, and Streptococcus [8,13]. This system has
been used for the efficient expression of homologous and
heterologous enzymes in L. lactis and has proven to be very
useful for metabolic engineering purposes (see below).
3. Sugar transport and glycolysis in L. lactis
L. lactis can use a variety of mono- and di-saccharides as
carbon source, including glucose, fructose, galactose, lactose, maltose, mannose, and cellobiose. Three different systems for sugar import are known in L. lactis. Primary
transport systems display a direct coupling of sugar import
to ATP hydrolysis via a transport specific ATPase [15]. In
secondary sugar transport systems the import of sugar is
coupled to transport of ions or other solutes [16]. Finally,
sugars that are imported via the phosphoenolpyruvate
(PEP)-dependent phosphotransferase system (PTS) are
phosphorylated during transport. This system transfers the
phosphate group from PEP, resulting in the formation of
pyruvate, to the incoming sugar via two cytoplasmic phosphocarrier proteins: enzyme I and HPr [17]. In lactic acid
bacteria, many sugars are imported via this PTS system,
which is an energetically efficient system for sugar transport
because sugar import is directly coupled to sugar phosphorylation at the expense of one ATP that would otherwise be
generated by PEP conversion to pyruvate by pyruvate kinase (pyk) (Fig. 2). Recently, the ptsHI operon of L. lactis,
encoding the general PTS components HPr and enzyme I,
respectively, has been cloned and characterized. HPr plays
an important role regulation of sugar utilization, both at the
level of transcription of genes involved in use of a specific
sugar and at the level of sugar transport. This phosphocarrier protein contains two residues that are target for phosphorylation. A histidine residue in the N-terminal part of the
protein, that is a target for enzyme I mediated phosphorylation and the resulting HPr-(His-P) primarily plays a role in
PTS sugar transport, but can also influence the use of nonPTS sugars [18,19]. The second HPr-residue that is a target
of phosphorylation is a conserved serine residue in the
C-terminal part of HPr. The HPr-(Ser-P), like HPr-(His-P),
plays a complex role in sugar use, where it can, (i) allosterically control PTS and non-PTS permease activity,
thereby regulating sugar import, (ii) control the activity of a
sugar-phosphate phosphatase, thereby playing a role in the
phenomenon called inducer exclusion and (iii) interact with
the catabolite control protein CcpA that negatively controls
the expression of a large number of genes or operons that
are subject to catabolite repression [18]. Moreover, it has
recently been shown that the lactococcal CcpA, besides its
role in catabolite repression, also regulates sugar metabolism by affecting the glycolytic flux through activation of
the las operon [20].
M. Kleerebezem et al. / Enzyme and Microbial Technology 26 (2000) 840 – 848
843
Fig. 3. Schematic representation of the lactococcal pyruvate metabolism. Enzymes are indicated by their encoding genes, using established genetic
nomenclature. NADH and ATP producing/consuming characteristics of all individual steps are indicated. The redox rescue pathway that seems to be used
by LDH-deficient strains (see text for further explanation), is indicated with a thick-lined arrow. The heterologously introduced alanine dehydrogenase
catalized conversion (see text for further explanation) is boxed. Finally, the reaction catalized by the NADH-oxidase is shown in the right-hand corner.
In dairy fermentations lactose is the main carbon source
encountered by L. lactis and the pathway used to convert
this sugar is relatively well understood (Fig. 2). In many
lactococcal strains, the genes involved in lactose use (lac
operon) are plasmid encoded, and the best described is the
lactose mini-plasmid pMG820 [21,22]. This plasmid contains genes involved in lactose-PTS transport (lacEF), lactose-6P hydrolysis (lacG), and the tagatose-6P pathway
(lacABCD) (Fig. 2). Expression of the lac-operon genes is
controlled by the transcriptional repressor encoded by lacR
and repression of the lac gene expression can be relieved by
the inducer tagatose-6P, which is an intermediate of the
tagatose-6P pathway [23; Fig. 2]. An elegant example of
engineering of lactose utilisation in L. lactis was described
in 1985 by Thompson et al. By classical mutagenesis, a L.
lactis double mutant was constructed that lacked both the
glucokinase (glk; Fig. 2) activity and the mannose PTS
system involved in glucose import. As a result, when this
strain was grown on lactose, the galactose-6P moiety of this
disaccharide can be converted via the tagatose-6P pathway,
whereas the glucose moiety remains unphosphorylated and
is secreted. The lack of a mannose PTS in these cells
prevents the subsequent uptake of the expelled glucose. The
overall result is an apparent quantitative conversion of lactose in glucose by these cells, which shows that galactose
and glucose metabolism in lactose grown L. lactis can be
completely separated [24].
Many of the genes encoding glycolytic enzymes in L.
lactis have been cloned and characterized. Moreover, a
predictive metabolic flux model of lactococcal glycolysis
was recently developed on basis of in vivo NMR measurements, allowing qualitative prediction of metabolic shifts
upon changes in the environmental conditions [25]. Interestingly, the transcription of the las operon [26], encoding
three key-enzymes involved in glycolysis and production of
lactate, phosphofructo-kinase (pfk), pyruvate-kinase (pyk)
and lactate dehydrogenase (ldh) (Fig. 2), has recently been
shown to be activated by CcpA [20]. By inactivation of the
lactococcal ccpA gene, strains were generated that displayed
a strongly reduced expression level of the las operon genes,
which resulted in a shift from homo-lactate to mixed acid
fermentation [20]. This study is an elegant example of the
possibility of metabolic engineering by manipulation at the
level of global regulation, and the availability of the keyregulatory components that exert this control in L. lactis
(HPr, CcpA and Enzyme I) [19,20] will probably allow
further fine-tuning of metabolic engineering strategies in the
future.
4. Engineering of the pyruvate metabolism
Pyruvate is the key metabolic intermediate; it is the
endproduct of the substrate level phosphorylation reactions
of the glycolytic pathway and is also produced through the
phosphorylation cascade involved in PTS-mediated sugar
import (Fig. 2). Pyruvate can be converted to various endproducts by L. lactis, including acetate, ethanol, diacetyl,
acetoin, and 2,3-butanediol (Fig. 3). However, the main
endproduct is evidently lactate, which is formed by the
lactate dehydrogenase reaction. This reaction acts as main
electron sink under anaerobic conditions, leading to homolactic fermentation. However, a shift toward mixed acid
fermentation has been observed under certain fermentation
844
M. Kleerebezem et al. / Enzyme and Microbial Technology 26 (2000) 840 – 848
conditions, such as carbohydrate limitation [27], diminished
rates of sugar metabolism [28], and aerobic conditions [29].
Several genetic engineering strategies have been used to
create mutant lactococcal strains that display a significant
shift in their metabolism and that produce either one of the
endogenous endproducts more efficiently or produce an
endproduct that is normally not produced by L. lactis. Here
we will describe examples of efficient rerouting of pyruvate
metabolism toward the endproducts diacetyl and alanine.
5. Engineering diacetyl production
One of the endproducts that can be generated from pyruvate is diacetyl (Fig. 3), which is an important flavor compound in dairy products generating the typical butter-aroma.
This compound is naturally produced by L. lactis, especially
by L. lactis subsp. lactis biovar. diacetylactis, from citrate in
co-metabolic fermentation with lactose [30]. During this
fermentation the intermediate ␣-acetolactate is accumulated, which is chemically (nonenzymatically) converted to
diacetyl through an oxidative decarboxylation reaction (Fig.
3). Based on its aroma value, efficient diacetyl (or its precursor ␣-acetolactate) production from lactose rather than
from citrate by L. lactis has been one of the targets of
various metabolic engineering strategies.
An obvious target for engineering is the lactate dehydrogenase (LDH), and an LDH deficient strain was created by
ldh gene disruption. The mutant displayed a mixed acid
fermentation under anaerobic conditions, producing high
amounts of formate and ethanol. In contrast, upon aeration
a dramatic shift in pyruvate metabolism was observed, leading to increased acetate production and acetoin as the main
endproduct [31]. An alternative approach is to increase the
amount of enzyme involved in conversion of pyruvate to
␣-acetolactate. Two enzymes are known in L. lactis that
catalyse this reaction: ␣-acetolactate synthase (ALS; encoded by the als gene; Fig. 3) [32] and acetohydroxy acid
synthase (ILVBN; encoded by the ilvBN gene; Fig. 3) [33].
Although the als geneproduct acts as the catabolic synthase,
the ilvBN geneproduct is an anabolic synthase that is involved in branched chain amino acid biosynthesis. Overexpression of the ilvBN genes in L. lactis grown under aerobic
conditions resulted in a 3- to 4-fold increased flux toward
acetoin [34], suggesting that the ILVBN activity level was
probably low and indicating that this approach could possibly be improved. A 100-fold increase in ALS enzyme
level was achieved by cloning the als gene on a multi-copy
plasmid, resulting in a 40% rerouting efficiency of the
pyruvate pool towards acetoin under aerobic conditions
[31]. Moreover, overexpression of ALS in a LDH-deficient
background led to the production of very high amounts of
acetoin (and relatively low amounts of ␣-acetolactate), accounting for 85% of the pyruvate converted. The approaches described here have been more or less successful
in generating lactococcal strains that effectively produce
acetoin from lactose or glucose, but have failed to create
strains that produce diacetyl efficiently. This is due to the
endogenous ␣-acetolactate decarboxylase (ALDB; encoded
by the aldB gene; Fig. 3) activity that efficiently converts
␣-acetolactate to acetoin. Inactivation of the aldB gene
resulted in strains that produce low levels of diacetyl under
aerobic conditions, and the diacetyl production level could
be increased by overexpression of the ilvBN genes in this
strain [35]. However, only low absolute levels of diacetyl
production could be achieved, and these cells still produced
relatively high levels of acetoin, which is due to the acetoin
diacetyl reductase activity (ADR; Fig. 3). Interestingly, an
elegant approach to obtain ALDB deficient natural L. lactis
variants has recently been described [36]. Taken together,
these results suggest that a strain in which all three mutations are combined (LDH- and ALDB-deficiency and ALS
or ILVBN overproduction) could be an effective diacetyl
producer. However, presently no such L. lactis strains have
been described, which is possibly explained by putative
lethality or poor viability of lactococcal cells that are deficient in both LDH and ALDB.
6. Efficient diacetyl production by cofactor engineering
The intracellular redox balance, most importantly reflected by the NADH/NAD⫹ ratio, plays a predominant role
in control of the fermentation pattern displayed by L. lactis
[29,37]. Pyruvate reduction by LDH effectively regenerates
NAD⫹ and thereby effectively transfers reduction equivalents formed during glycolysis to the metabolic endproduct.
Interestingly, and illustrative for the energetically unfavorable pathways that will be used by these cells to maintain
their redox balance, it has recently been shown that the rate
of sugar fermentation by LDH-deficient L. lactis strains can
be increased by co-metabolisation of acetate. The acetate
consumed under these conditions is exclusively converted
into ethanol at the expense of NADH and ATP, thereby
creating a redox rescue pathway for these cells that lack an
effective reductive pathway [38]. The stringent redox balance control of fermentation patterns has recently been
exploited in a metabolic engineering approach designated
cofactor engineering [39]. L. lactis strains were created in
which the redox balance could directly be controlled, which
was achieved by exploiting the NICE system for controlled
expression of the S. mutans nox gene. Overexpression in a
dynamic range of this water-forming NADH-oxidase
(NOX) encoding gene in combination with specific aeration
levels allowed effective control of the NADH/NAD⫹ balance [39]. These induced variations in NOX activity resulted in inhibition of all NADH-dependent pathways and a
rerouting of the pyruvate pool towards oxidative or NADHindependent pathways. In effect, the pyruvate metabolism
was shifted from homolactic toward mixed acid fermentation, and at maximal NOX-induction level combined with
boosting of NOX activity by addition of FAD to the growth
M. Kleerebezem et al. / Enzyme and Microbial Technology 26 (2000) 840 – 848
medium, an almost complete replacement of lactate by
acetate and acetoin (and some diacetyl) could be achieved
[39].
Furthermore, by overproduction of NOX in an ALDBdeficient lactococcal background, the pyruvate metabolism
could be efficiently rerouted toward ␣-acetolactate and diacetyl. These cells appeared to convert more than 80% of
the available pyruvate pool to this butter aroma compound
and its precursor [40]. Regarding the central control role of
the redox balance on lactococcal metabolism, this approach
of cofactor engineering in combination with inactivation or
overexpression of other enzymatic steps will probably have
a strong impact on future metabolic engineering strategies
that aim at rerouting toward metabolic endproducts other
than lactate. Moreover, cofactor engineering could readily
be applied in other lactic acid bacteria because the expression system used (NICE) can be functionally transferred to
other bacteria by an easy to handle two-plasmid system [8].
7. Homo-L-alanine fermentation by modified L. lactis
Most metabolic engineering strategies that have been
applied in L. lactis were aimed at increasing the production
of one of the natural lactococcal metabolic endproducts. As
an alternative, introduction of a heterologous pathway in L.
lactis could lead to lactococcal variants that produce interesting new products from lactose (or another carbonsource). An example of this type of approach is the overexpression of the Zymomonas mobilis genes encoding
pyruvate decarboxylase (pdc) and alcohol dehydrogenase
(adh). Expression of this ethanol forming pathway in LDHdeficient L. lactis cells led to a significant amount of ethanol
production (2%) (Hugenholtz, unpublished results). Recently, extremely efficient rerouting of the pyruvate flux
was achieved by overexpression of L-alanine dehydrogenase (ALAD), encoded by the alaD gene derived from
Bacillus sphearicus (Fig. 3) [41]. The alaD gene was cloned
under control of the nisA promoter and the encoded enzyme
accounted for approximately 30 – 40% of total soluble proteins under maximally induced conditions. In wild-type
lactococcal cells this already resulted in a 30 – 40% rerouting efficiency from lactate toward alanine when resting cells
were used as glucose-utilizing biocatalists in the presence of
an excess of ammonium, which is required for the pyruvate
to alanine conversion by ALAD. Moreover, introduction of
this system in LDH-deficient lactococcal cells and incubation of the resulting cells under the appropriate conditions,
resulted in a complete conversion of the pyruvate pool into
alanine. The alanine produced consisted of a mixture of both
stereo-isomers (D- and L-alanine), and in all cases D-alanine amounted to approximately 10 –15% of the total alanine produced. This partial racemisation of the L-alanine
produced is due to the alanine racemase (ALR) activity
encoded by the alr gene of L. lactis. In order to create a
stereospecific L-alanine producing lactococcal variant, the
845
alr gene was disrupted in the LDH-deficient strain and it
could be shown that the resulting strain indeed produced
stereospecific L-alanine as the sole endproduct of its fermentation [41]. In conclusion, this strategy of metabolic
engineering has led to a complete conversion of L. lactis
from the naturally occurring homo-L-lactate fermentation to
a homo-L-alanine fermentation. Similar as described for the
redox engineering strategy, the alanine producing capacity
can readily be transferred to other LAB using the twoplasmid NICE-transfer system [8]. This extremely successful engineering strategy illustrates the potential of simple
rerouting pathways in creating LAB that produce large
amounts of completely new compounds. In this case the
compound produced, L-alanine, is used as a food-sweetener
and in pharmaceutical applications, where it is added in
combination with other L-amino acids to standard infusions
in preoperative and postoperative nutrition therapy [42].
The system using resting lactococcal cells as biocatalists for
the quantitative conversion of a cheap carbon source (i.e.
sucrose, whey, starch) into L-alanine clearly has potential as
a new industrial process for the production of this compound as a fine chemical. Another interesting application of
alanine producing lactic acid bacteria could be the in situ
production of this sweetener in fermented dairy products
like buttermilk, yogurt, or cheese. The potential of this latter
application of the described system is currently being evaluated by several European laboratories, in a concerted research program.
8. Complex pathway engineering
A major challenge in current and future engineering
approaches is the modulation of complex pathways, leading
toward interesting endproducts. A good example of such a
product is given by the polysaccharides that are naturally
produced by some so-called ropy strains of LAB. Recent
years have shown increasing interest in the production of
exopolysaccharides (EPS) by LAB, caused by their relevance in dairy-product properties like texture and mouthfeel. Based on the food-grade status of LAB, these EPSs
could readily be applied as food additives, where they can
contribute to several product properties (i.e. viscosity, emulsion-stability). Moreover, it has been suggested that they
can be active as prebiotic [43], cholesterol lowering neutraceutical [44] or immunomodulant [45,46]. Over the last
years, the chemical structure, the molecular biology and
genetics of the biosynthesis of various repetitive heteropolysaccharide EPSs produced by several LAB have been studied. In EPS producing L. lactis, the EPS specific eps genes
are encoded on large plasmids (⬎20 kb) that can be transferred by conjugation, which concomitantly results in transfer of EPS production [47,48]. L. lactis strain NIZO B40
produces a phosphopolysaccharide with a known structure
(Fig 4B;34)[␣-L-Rhap-(132)][␣-D-Galp-1-PO4-3]-␤-DGalp-(134)-␤-D-Glcp-(134)-␤-D-Glcp(13) [47,49]. The
846
M. Kleerebezem et al. / Enzyme and Microbial Technology 26 (2000) 840 – 848
Fig. 4. A, schematic representation of the genetic organization of the plasmid encoded L. lactis NIZO B40 eps genecluster, epsRXABCDEFGHIJKL, flanked
by an IS element and orfY. B, schematic representation of NIZO B40 EPS biosynthesis starting from glucose-6P (branching point from Fig. 2). Primary
products indicated are the nucleotide sugars that are subsequently used in stepwise B40 repeating unit biosynthesis and after polymerization and export result
in formation of B40-EPS. Enzymes are indicated by their encoding genes using established genetic nomenclature.
12 kb NIZO B40 eps gene cluster contains 14 coordinately
expressed genes, epsRXABCDEFGHIJKL (Fig. 4A), and is
localized on a 40 kb plasmid [47]. Recently, in vitro experiments have shown that the biosynthesis of B40-polysaccharide backbone is initiated by the linkage of a glucose
from UDP-glucose to the lipid carrier by priming-glucosyltransferase EpsD [47,50]. Subsequently, the addition of a
second glucose from UDP-glucose to the lipid-linked glucose involves the combined activity of EpsE and EpsF, and
finally, the addition of the third backbone-sugar moiety,
galactose, from UDP-galactose to this carrier bound cellobiose involves the activity of EpsG (Fig. 4B) [50]. So far,
the rhamnosyl- and phospho-galactosyl-transferase activities involved in coupling of the side-chain sugar moieties
have not been experimentally established but have been
proposed to be encoded by epsH and epsJ (Fig. 4B) [50].
Increasing knowledge on EPS structures, combined with the
availability of individual biosynthetic components from eps
geneclusters could allow directed EPS-engineering approaches aiming at changing its chemical structure (sugarcomposition and/or sugar-linkages) or other relevant fea-
tures like its chain-length. An essential first step in this
direction has been realized by the demonstration that a
B40-EpsD deletion mutant could functionally be complemented, not only by the homologous EpsD gene itself, but
also by heterologous priming glucosyltransferase encoding
genes [51]. Recently, other examples that hint at the possibilities for production of heterologous or new EPS structures in L. lactis have been reported. Expression of the
genes involved in type 3 capsular polysaccharide (CPUs)
production in S. pneumoniae resulted in high level production of the type 3 polysaccharide by L. lactis [52]. The entire
S. thermophilus Sfi6 eps gene cluster was expressed in L.
lactis, resulting in production of low but significant amounts
of EPS of a different structure as compared with the EPS
produced by the original S. thermophilus strain [53]. These
different approaches offer good perspective for future eps
engineering in L. lactis (or in other LAB), and show that this
host could be used for the expression of oligo- or polysaccharides that have certain desirable characteristics.
Besides the specific eps genes, EPS production also requires a number of so-called housekeeping genes that are
M. Kleerebezem et al. / Enzyme and Microbial Technology 26 (2000) 840 – 848
involved in the biosynthesis of the EPS building blocks, the
nucleotide-sugars. Possibly, the relatively low EPS production levels displayed by LAB could be increased by increasing the metabolic flux toward these nucleotide-sugars (Fig.
4B). Therefore, the physiology of nucleotide sugar biosynthesis in relation to EPS production levels should be evaluated to eventually use metabolic engineering as a means to
increase EPS production by LAB. Currently, the role of the
genes that encode potentially important enzymes in these
metabolic pathways are being investigated (i.e., galE, galU,
and rfbABCD; Fig. 4B). Moreover, this metabolic engineering approach will be supported by the simultaneous development of a kinetic flux model that describes the contribution of the individual enzymes to the level of the various
nucleotide-sugars generated. Taken together, these strategies could eventually generate a rational for improvement of
the production levels of a complex endproduct like EPS,
which is crucial when these EPS producing LAB are to be
used as food-additive producer organisms. In contrast, the
requirements for application of modified LAB that produce
a certain desirable oligo- or polysaccharide in situ are determined more by their functionality (e.g. role as neutraceutical; probiotic, or vaccine, generating texture or mouthfeel), than by their production level. It is expected that the
engineering of EPS and its biosynthesis pathway in LAB
will develop into an important objective in future metabolic
engineering research, especially since the in situ production
of certain complex polysaccharides that generate a health
benefit to the fermented products seems a very attractive
application.
9. Conclusion
The examples described above illustrate the potential of
lactic acid bacteria as cell factories to produce desirable
products during fermentation. It is clear that the metabolic
flexibility of these bacteria is larger than could be expected
on basis of the simple fermentation pattern they display
naturally. One possible application of engineered LAB is
their use as real cell factories that displays an efficient
conversion of the carbon source into a certain desirable
endproduct that is subsequently separated from the producing cells and used as additive in food or other applications.
The other, and maybe more important application of engineered LAB as producers of certain products, is their potential for in situ production (engineered starter cultures) of
a compound that contributes to the functional characteristics
(e.g. taste, texture, nutritional value, health benefits) of the
fermented product.
We have described a few examples of metabolic engineering strategies used in LAB. Future engineering of LAB
requires the identification of new high-value endproducts
that can be targeted by engineering strategies leading to, for
example, the tailor-made oligo and polysaccharides mentioned, or to compounds like vitamins, antioxidants, or
847
compatible solutes. In most cases these compounds are
generated by a complex biosynthetic pathway, the engineering of which will be an important challenge for future
metabolic research in LAB.
Acknowledgments
We thank Dr. Roland Siezen and Dr. Dick van den Berg
for critically reading the manuscript.
References
[1] de Vos WM. Gene expression systems for lactic acid bacteria. Curr
Opin Microbiol 1999;2:289 –295.
[2] Bolotin A, Mauger S, Malarme K, Sorokin A, Ehrlich DS. Lactococcus lactis IL1403 diagnostic genomics. Proceedings 5th ASM on the
genetics and molecular biology of streptococci, enterococci, and
lactococci. 1998. p. 10 –1.
[3] de Vos WM. Safe and sustainable systems for food-grade fermentations by genetically modified lactic acid bacteria. Int Dairy J 1999;
9:3–10.
[4] Djordjevic GM, Klaenhammer TR. Inducible gene expression systems in Lactococcus lactis Mol Biotechnol 1998;9:127–39.
[5] Kuipers OP, de Ruyter, PGGA, Kleerebezem M, de Vos WM. Controlled overproduction of proteins by lactic acid bacteria. Trends
Biotechnol 1997;15:135– 40.
[6] Kuipers OP, de Ruyter PGGA, Kleerebezem M, de Vos WM. Quorum sensing-controlled gene expression in lactic acid bacteria. J Biotech 1998;64:15–21.
[7] Jensen PR, Hammer K. The sequence of spacers between the consensus sequences modulates the strengths of prokaryotic promoters.
Appl Environ Microbiol 1998;63:82–7.
[8] Kleerebezem M, Beerthuyzen MM, Vaughan EE, de Vos, WM,
Kuipers OP. Controlled gene expression systems for lactic acid bacteria: transferable nisin inducible expression cassettes for Lactococcus, Leuconostoc, and Lactobacillus spp. Appl Environ Microbiol
1997;63:4581– 4.
[9] Eichenbaum Z, Federle MJ, Marra D, de Vos, WM, Kuipers OP,
Kleerebezem M, Scott JR. Use of the lactococcal nisA promoter to
regulate gene expression in Gram-positive bacteria: comparison of
induction level and promoter strength. Appl Environ Microbiol 1998;
64:2763–9.
[10] Kuipers OP, Beerthuyzen MM, de Ruyter, PGGA, Luesink EJ, de
Vos WM. Autoregulation of nisin biosynthesis in Lactococcus lactis
by signal transduction. J Biol Chem 1995;270:27299 –27304.
[11] Kleerebezem M, Quadri LEN, Kuipers OP, de Vos WM. Quorum
sensing by peptide pheromones and two-component signal-transduction systems in gram-positive bacteria. Mol Microbiol 1997;24:895–904.
[12] de Ruyter PGGA, Kuipers OP, Beerthuyzen MM, van Alen-Boerrigter I, de Vos WM. Functional analysis of promoters in the nisin
gene cluster of Lactococcus lactis. J Bacteriol 1996;178:3434 –9.
[13] de Ruyter PGGA, Kuipers OP, de Vos WM. Controlled gene expression systems for Lactococcus lactis with the food-grade inducer nisin.
Appl Environ Microbiol 1996;62:3662–7.
[14] de Ruyter PGGA, Kuipers OP, de Vos WM. Food-grade controlled
lysis of Lactococcus lactis for accelerated cheese ripening. Nat Biotechnol 1997;15:976 –9.
[15] Fath MJ, Kolter R. ABC exporter. Microbiol Rev 1993;57:995–1017.
[16] Poolman B. Energy transduction in lactic acid bacteria. FEMS Microbiol Rev 1993;12:125– 48.
[17] Postma PW, Lengeler JW, Jacobsen GR. Phosphoenolpyruvate dependent carbohydrate phosphotransferase systems of bacteria. Microbiol Rev 1993;57:543–94.
848
M. Kleerebezem et al. / Enzyme and Microbial Technology 26 (2000) 840 – 848
[18] Deutscher J, Kuster E, Bergstedt U, Charrier V, Hillen W. Proteinkinase dependent HPr/CcpA interaction links glycolytic activity to
carbon catabolite repression in Gram-positive bacteria. Mol Microbiol 1995;15:1049 –53.
[19] Luesink EJ, Beumer CM, Kuipers OP, de Vos WM. Molecular
characterization of the Lactococcus lactis ptsHI operon and analysis
of the regulatory role of HPr. J Bacteriol 1999;181:764 –71.
[20] Luesink EJ, van Herpen RE, Grossiord BP, Kuipers OP, de Vos WM.
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 1998;30:789 –98.
[21] De Vos WM, Boerrigter I, van Rooijen RJ, Reiche B, Hengstenberg
W. Characterization of the lactose-specific enzymes of the phosphotransferase system in Lactococcus lactis. J Biol Chem 1990;265:
22554 – 60.
[22] Van Rooijen RJ, van Schalkwijk S, de Vos WM. Molecular cloning,
characterization, and nucleotide sequence of the tagatose 6-phosphate
pathway genecluster of the lactose operon of Lactococcus lactis.
J Biol Chem 1991;266:7176 – 81.
[23] Van Rooijen RJ, Dechering KJ, Wilmink CMJ, de Vos WM. Lysines
72, 80, 213, and aspartic acid 210 of the Lactococcus lactis LacR
repressor are involved in the response to the inducer tagatose-6phosphate leading to induction of lac operon expression. Protein Eng
1993;6:208 –15.
[24] Thompson J, Chassy BM, Egan W. Lactose metabolism in Streptococcus lactis: studies with a mutant lacking glucokinase and mannose-transferase activities. J Bacteriol 1985;162:217–33.
[25] Neves AR, Ramos A, Nunes MC, Kleerebezem M, Hugenholtz J, de
Vos WM, Almeida J, Santos H. In vivo NMR studies of glycolytic
kinetics in Lactococcus lactis. Biotechnol Bioeng 1999;64:200 –12.
[26] Llanos RM, Marin CJ, Hillier AJ, Davidson BE. Identification of a
novel operon in Lactococcus lactis encoding enzymes for lactic acid
synthesis: phosphofructokinase, pyruvate kinase and lactate dehydrogenase. J Bacteriol 1993;175:254 –5.
[27] Thomas TD, Ellwood DC, Longyear VMC. Change from homo- to
heterolactic fermentation by Streptococcus lactis resulting from glucose limitation in anaerobic chemostat cultures. J Bacteriol 1979;138:
109 –17.
[28] Cocaign-Bousquet M, Garrigues C, Loubiere P, Lindley N. Physiology of pyruvate metabolism in Lactococcus lactis. Antonie van Leeuwenhoek 1997;70:253– 67.
[29] Lopez de Felipe F, Starrenburg MJC, Hugenholtz J. The role of
NADH-oxidation in acetoin and diacetyl production from glucose in
Lactococcus lactis. FEMS Microbiol Lett 1997;156:15–9.
[30] Hugenholtz J. Citrate metabolism in lactic acid bacteria. FEMS Microbiol Rev 1993;12:165–78.
[31] Platteeuw C, Hugenholtz J, Starrenburg MJC, van Alen-Boerrigter I,
de Vos WM. Metabolic engineering of Lactococcus lactis: influence
of the overproduction of ␣-acetolactate synthase in strains deficient in
lactate dehydrogenase as a function of culture conditions. Appl Environ Microbiol 1995;61:3967–71.
[32] Marugg JD, Goelling D, Stahl U, et al. Identification and characterization of the ␣-acetolactate synthase gene from Lactococcus lactis
subsp. liacetylactis. Appl Environ Microbiol 1994;60:1390 – 4.
[33] Godon JJ, Delorme C, Bardowski J, Chopin MC, Ehrlich SD, Renault
P. Gene inactivation in Lactococcus lactis branched chain amino acid
biosynthesis. J Bacteriol 1993;175:4383–90.
[34] Benson KH, Godon JJ, Renault P, Griffin HG, Gasson MJ. Effect of
ilvBN-encoded ␣-acetolactate synthase expression on diacetyl production in Lactococcus lactis. Appl Microbiol Biotechnol 1996;45:
107–11.
[35] Swindell SM, Benson KH, Griffin HG, Renault P, Ehrlich SD, Gasson MJ. Genetic manipulation of the pathway for diacetyl metabolism
in Lactococcus lactis. Appl Environ Microbiol 1996;62:2641–3.
[36] Curic M, Stuer-Lauridsen B, Renault P, Nilsson D. A general method
for selection of alpha-acetolactate decarboxylase-deficient Lactococ-
[37]
[38]
[39]
[40]
[41]
[42]
[43]
[44]
[45]
[46]
[47]
[48]
[49]
[50]
[51]
[52]
[53]
cus lactis mutants to improve diacetyl formation. Appl Environ Microbiol 1999;65:1202– 6.
Garrigues C, Loubiere P, Lindley ND, Cocaign-Bouquet M. Control
of the shift from homolactic acid to mixed acid fermentation in
Lactococcus lactis: predominant role of NADH/NAD⫹ ratio. J Bacteriol 1997;179:5282–7.
Hols P, Ramos A, Hugenholtz J, Delcour J, de Vos WM, Santos H,
Kleerebezem M. Acetate utilization in Lactococcus lactis deficient in
lactate dehydrogenase: a rescue pathway for maintaining redox balance. J Bacteriol 1999;181:5521–26.
Lopez de Felipe F, Kleerebezem M, de Vos WM, Hugenholtz J.
Cofactor engineering: a novel approach to metabolic engineering in
Lactococcus lactis by controlled overexpression of NADH oxidase. J
Bacteriol 1998;180:3804 – 8.
Hugenholtz J, Hols P, Starrenburg MJC, de Vos WM, Kleerebezem
M. Metabolic engineering of Lactococcus lactis leading to high diacetyl production. Submitted for publication.
Hols P, Kleerebezem M, Schank AN, Ferain T, Hugenholtz J, Delcour J, de Vos WM. (1998) Conversion of Lactococcus lactis from
homolactic to homoalanine fermentation through metabolic engineering. Nat Biotechnol 1999;17:588 –92.
Leuchtenberger W. Amino acids-technical production and use. Products of primary metabolism. Biotechnolog In: Rhem H-J, Reed G,
Puhler A, Stadler PJW. editors VCH verlagsgesellschaft mbH, Germany. 1996. p. 465–502.
Gibson GR, Roberfroid MB. Dietary modulation of the human colonic microbiota: introducing the concept of prebiotics. J Nutr 1995;
124:1401–12.
Nakajima H, Suzuki Y, Kaizu H, Hirota T. Cholesterol-lowering
activity of ropy fermented milk. J Food Sci 1992;57:1327–9.
Kitazawa H, Yamaguchi T, Miura M, Saito T, Itoh H. B-cell mitogen
produced by slime-forming, encapsulated Lactococcus lactis subspecies cremoris isolated from ropy sour milk, viili. J Dairy Sci 1993;
76:1514 –9.
Hosono A, Lee J, Ametani A, Natsume M, Hirayama M, Adachi T,
Kaminogawa S. Characterization of a water-soluble polysaccharide
fraction with immunopotentiating activity from Bifidobacterium adolescentis M101– 4. Biosci Biotech Biochem 1997;61:312– 6.
van Kranenburg R, Marugg JD, van Swam II, Willem NJ, de Vos
WM. Molecular characterization of plasmid-encoded eps gene cluster
essential for exopolysaccharide biosynthesis in Lactococcus lactis.
Mol Microbiol 1997;24:387–97.
van Kranenburg R, de Vos WM. Characterization of multiple regions
involved in replication and mobilization of plasmid pNZ4000 coding
for exopolysaccharide production in Lactococcus lactis. J Bacteriol
1998;180:5285–90.
Nakajima H, Toyoda S, Toba T, et al. A novel phosphopolysaccharide from slime-forming Lactococcus lactis subspecies cremoris
SBT0495. J Dairy Sci 1990;73:1472–7.
van Kranenburg R, van Swam II, Marugg JD, Kleerebezem M, de
Vos WM. Exopolysaccharide biosynthesis in Lactococcus lactis
NIZO B40: functional analysis of the glycosyltransferase genes involved in synthesis of the polysaccharide backbone. J Bacteriol 1999;
181:338 – 40.
van Kranenburg R, Vos HR, van Swam II, Kleerebezem M, de Vos
WM. Functional analysis of glycosyltransferase genes from Lactococcus and other Gram-positive cocci: complementation, expression
and diversity. J Bacteriol 1999;181:6347–53.
Gilbert C, Robinson K, LePage RWF, Wells JM. Heterologous biosynthesis of pneumococcal type 3 capsule in Lactococcus lactis and
immunogenicity studies. Proceedings 5th ASM on the genetics and
molecular biology of streptococci, enterococci and lactococci. 1998.
p. 67.
Stingele F. Exopolysaccharide production and engineering in dairy
Streptococcus and Lactococcus. Proceedings 5th ASM on the genetics and molecular biology of streptococci, enterococci and lactococci.
1998. p. 31–2.