Lactococcus lactis strains from the subsp. cremoris are described as more sensitive to osmotic stress than subsp. lactis strains. We examined the relation between osmotic tolerance and the activity of the betaine transporter BusA among 34 strains of L. lactis. The cremoris strains that showed reduced growth at high osmolality failed to accumulate betaine. The nature of the defect was found to vary among cremoris strains: lack of the busA encoding region, absence of synthesis or synthesis of an inactive form of BusA. The results suggest that the selection of strains well fitted to the dairy production lead to the loss of an otherwise efficient adaptation mechanism.
Lactococcus lactis is a mesophilic lactic acid bacterium used in dairy fermentation and cheese making. The subspecies lactis strains, mainly used as fast acidifying starters, can grow at up to 40°C, at pH 9.2, or at 4% NaCl (w/v), whereas the cremoris strains, which are slow-growing and generally used as flavouring strains, cannot support such conditions. This different ability to support environmental stresses was historically used to differentiate L. lactis strains at the subspecies level. The lactis strains can be isolated from various environments outside dairy products or raw milk, such as plant material or animal surfaces, while the cremoris strains seem to be restricted to dairy environments [3,13].
Osmoprotectants are small organic solutes accumulated at high intracellular concentrations to cope with the detrimental effects of high external osmolality . Betaine is used as the main osmoprotective compound by many living organisms, and bacteria possess osmoregulated transport systems dedicated to the accumulation of this quaternary amine . We recently characterised the unique betaine transport system BusA of L. lactis subsp. cremoris NCDO 763 . In that strain, the disruption of the busA operon led to an impaired growth in chemically defined media containing betaine  or in M17 when the osmolality was raised above 0.3 M (unpublished data).
Consequently, we put forward the hypothesis that the betaine accumulation through BusA activity is a strong determinant of the level of osmotolerance among L. lactis strains. Here, we report that the capacity to accumulate betaine is extremely variable among lactococci strains and that the lack or a low activity of BusA is associated with an osmosensitive phenotype.
2Materials and methods
2.1Bacterial strains and media
The 34 strains of L. lactis used (Table 1) were obtained from the CNRZ collection (Jouy-en-Josas, France). Bacteria were grown on M17Glc (M17+0.5% glucose (w/v)) . The osmolality was raised by the addition of 20% NaCl (w/v) in M17.
Table 1. L. lactis strains used in this study
NCDO 604, type strain
traditional brie, France
traditional brie, France
traditional brie, France
Mung Bean, China
Lactobacillus xylosus, type strain
United Dairies, UK
NCDO 712 derivative
DRI, New Zealand
NCDO 712 derivative
UD 712, Australia, 1955
plasmid-free NCDO 712 derivative
cremoris NCDO 607, type strain
cheese starter, New Zealand
phage-resistant AM1 derivative
UCD, cheese starter
cheese starter, New Zealand
2.2Polymerase chain reaction (PCR)
The busA operon was detected by PCR on fresh colonies (M17Glc agar, overnight, 30°C), collected with a sterile pipette tip and resuspended in 20 μl of PCR buffer. Reactions were performed on a DNA Thermal cycler 2400 (Applied Biosystems) using Taq Polymerase (Appligene) with degenerated primers (forward primer (EIFV-1): 5′-GARATHTTYGTXATHATGGG; reverse primer (3R-MDEA): 5′-GCMSWRAAMGCYTCRTCCAT (Eurogentec)). The amplified 410-kb band corresponds to a conserved region in the genes encoding the ATP-binding proteins (residues 55–191 of BusAA (GenBank: AF139575)) of bacterial betaine ABC-transporters.
2.3DNA extraction, enzymatic modification and Southern blotting
DNA from L. lactis was extracted as described previously , digested (5 μg) with EcoRI (Eurogentec), and analysed by Southern hybridisation  with a whole busA 3.4-kb PCR probe labelled with the ECL kit (Amersham). The probe was obtained by PCR amplification using the primers (YHAT-1): 5′-GTTGTTGCAATTTTACAGAATGAAG (forward) and (BUS-4071): 5′-TCACTGAGATTTTCTTAGTTAACTC (reverse) and the chromosomal DNA of NCDO 763 as the template.
2.4Immunological detection of BusAA and BusAB proteins on total cell extracts
Cells grown in M17Glc at 30°C (OD650=1) were submitted to an osmotic upshock by addition of 2% or 4% NaCl (w/v) for 1 h, harvested and washed in NaPO4 buffer (200 mM, pH 6.0) with equivalent amounts of NaCl. Total protein extracts were prepared by SDS-lysis of protoplasts after lysozyme treatment (10 mg ml−1, 37°C, 15 min), separated by SDS–PAGE, and blotted for Western blots hybridisations as described previously . Antisera were prepared from rabbits (Eurogentec, Belgium) immunised with BusAA and BusAB proteins of L. lactis subsp. lactis IL 1403, provided by T. van der Heide and B. Poolman .
2.5Initial rate of betaine uptake assays
Cells were grown, osmotically upshocked with 2% NaCl and harvested as described above. Cell wash, resuspension and transport assays were performed as described . In brief, after release of accumulated osmolytes by osmotic downshock, cells (0.05 mg of protein per ml) were resuspended in 50 mM MES-NaOH (pH 6.5) containing 0.5% (w/v) glucose at 30°C. Uptake was initiated by the addition of NaCl (2% final concentration) and 20 μM [1-14C]betaine (2.5 mCi mmol−1). Aliquots were filtered at time intervals of 1 min for 4 min. The radioactivity associated with the filters was measured by scintillation counting.
2.6Quantification of cytoplasmic betaine accumulation
Betaine was quantified as previously described  by high performance liquid chromatography analysis on cytoplasmic extracts obtained by hypoosmotic downshock of cells adsorbed on nitrocellulose filters.
3.1Growth at high osmolality of L. lactis strains
The capacity to grow in the presence of increasing salt concentrations (Fig. 1) was measured for 34 strains of L. lactis, belonging to the two subspecies, lactis and cremoris (Table 1), representative of the genetic diversity of the species . The subsp. lactis strains (Fig. 1A) were exhibiting canonical phenotypes, except for three non-dairy lactis strains (CNRZ 145, NCDO 2110 and NCDO 2125) that were able to grow at 6.5% NaCl (1.1 M). Concerning the cremoris group, the growth aptitude at high osmolality was found to be more heterogeneous (Fig. 1B). At 1% NaCl, the cremoris strains reached a final cell density comparable to that found for lactis strains, except for three fastidious strains (ML1, SK11 and C13). At 2% NaCl, the development of six strains was strongly inhibited (AM1, IL 182, HP, Z8, CNRZ 118 and US3). At 4% NaCl, the four strains, NCDO 712, C2, NCDO 763 and MG 1363, which belong to the same genetic lineage, presented phenotypic characters typical of the lactis subspecies [4,5,8].
3.2Accumulation of the osmoprotectant betaine
M17 is complemented with yeast extract, which is a potent source of osmoprotective molecules such as betaine . We measured cytoplasmic amounts of betaine in cells cultivated in M17 at various salt concentrations to determine whether betaine accumulation could be correlated with the growth capacity (Fig. 2). We focused our attention on betaine, because this molecule was the sole among others tested (taurine, carnitine, ectoine and proline) to act as an osmoprotectant for L. lactis (, unpublished data). For lactis strains (Fig. 2A), the cytoplasmic betaine concentration increased with the osmolality of the medium from 0 to 4% NaCl. Betaine was accumulated up to 4300 nmol mg−1 protein at 6.5% NaCl for CNRZ 145, but the amount reached an upper limit for NCDO 2110 and even diminished for the NCDO 2125 strain, if compared to the betaine measured at 4% NaCl. The higher osmotolerance of these strains is therefore hardly related to increased betaine accumulation.
Concerning the cremoris strains, three groups can be defined (Fig. 2B). NCDO 712 and its derivatives accumulated amounts of betaine equivalent to the lactis strains at 1, 2 and 4% NaCl. E8, AM2, WG2 and KH (Fig. 2B) accumulated betaine at 1% NaCl, but displayed a deficit at 2% (average: 463 (SD103) nmol betaine mg−1 protein versus 863 (SD112) for NCDO 712 and derivatives). The strains AM1, IL 182, HP, Z8, CNRZ 118, US3, ML1, SK11, C13 which did not develop at 2% NaCl, were found to be totally deficient in betaine accumulation when grown at 1% NaCl (Fig. 2B). We observed that this regrouping was in accordance with the growth capacities under osmotic constraint.
3.3Detection of the busA locus among L. lactis strains
Betaine uptake in L. lactis subsp. cremoris NCDO 763 is due to the activity of a unique ABC-transporter encoded by the busA operon , and the recently published genome sequence of the L. lactis subsp. lactis IL 1403 did not reveal additional osmodependent transport systems . A PCR assay was used to detect the busAA gene in the 34 strains. A unique 410-bp PCR product was detected in 32 reactions. The noticeable exceptions were the HP and the Z8 cremoris strains, for which no specific amplification product could be obtained (Fig. 3). Randomly amplified polymorphic DNA analysis indicated that these two strains were closely related .
The genomic organisation of the busA operon of 10 strains was analysed by Southern hybridisation using a 3-kb probe, overlapping the entire encoding sequence of busA (Fig. 4). The experiment revealed two hybridisation profiles, characteristic of each subspecies. The size of the EcoRI fragments was in good agreement with those expected from the NCDO 763 (0.9, 1.3 and >10 kb) and IL 1403 (2.8 and 3.1 kb) busA nucleotide sequences. The lack of a detectable signal for the HP strain confirmed the absence of busA in that strain (Fig. 4, lane 1).
3.4Consequences of osmotic upshock on the expression of BusA and the betaine uptake activity
The expression of the betaine transporter was investigated by immunodetection of the BusAA and BusAB proteins in a subset of eight strains, after an osmotic upshock at 2 or 4% NaCl (Fig. 5). Except for CNRZ 118, all strains contained higher amounts of BusA proteins when the salt concentration raised from 0 to 2%. This result indicated that the osmodependent transcriptional activation of busA observed for NCDO 763  was reflected at the level of protein synthesis. The osmotolerant strain NCDO 2110 (Fig. 5D) exhibited increasing amounts of BusA protein at 4%, which may account for enhanced betaine accumulation and osmotolerance. NCDO 763, IL 1403 and CNRZ 145 (Fig. 5A–C, respectively) showed reduced synthesis of BusA at 4% compared to 2%. However, this was sufficient to support betaine accumulation and growth at this osmolality. The AM2 strain held a similar pattern of expression (Fig. 5F). Interestingly, the osmosensitive C13 and AM1 cremoris strains (Fig. 5E,G) were also able to induce the synthesis of BusA, while these strains did not accumulate betaine. In the case of the CNRZ 118 strain, no protein was detected by the immunodetection experiment (Fig. 5H).
We measured the betaine uptake capacity of five cremoris strains activated by a 1-h osmotic upshock at 2% NaCl (Fig. 6). When compared to the transport activity for NCDO 763, the results showed a 15-fold, 10-fold and four-fold lower activity for CNRZ 118, C13 and AM1, respectively. For the AM2 strain, which displays an intermediary phenotype: growth and betaine accumulation at 2% NaCl, the reduction in the betaine uptake capacity was only two-fold.
In this work, we found that all the subsp. lactis strains tested were proficient for betaine accumulation, whereas the cremoris strains at the exception of the NCDO 7112 lineage were deficient in betaine utilisation, though at various degrees.
Some environmental isolates of L. lactis subsp. lactis can grow at 6.5% NaCl, a concentration otherwise inhibitory to Lactococcus reference strains. These resistant strains do not necessarily display an enhanced betaine accumulation capacity. This supports the hypothesis that their growth advantage is related to other, undefined mechanisms of adaptation that deserve further investigations. The tolerance of these strains suggests that, with appropriate sampling and screening methods, non-dairy environments can be a source of stress-tolerant isolates of Lactococci or other lactic acid bacteria.
The four cremoris strains (E8, AM2, WG2, KH) which grow at 2% NaCl, but do not develop at 4%, possess an active betaine transport system allowing a significant cytoplasmic betaine accumulation. However, when tested in the AM2 strain, the betaine uptake activity was twice lower than that found in the osmotolerant NCDO 763 strain.
The osmosensitive phenotype of cremoris strains, for which no measurable amount of accumulated betaine can be detected after growth (Fig. 2B), was associated with either a very low activity of the BusA system (C13, AM1), a low expression (CNRZ 118) or the absence of the operon (HP, Z8). The absence of the busA operon can be explained either by (i) recent acquisition by busA-deficient strains, or (ii) the deletion of genetic material. The taxonomic position of HP and Z8 strains within L. lactis makes the latter hypothesis much more likely .
The reason of the low activity of the BusA transporter in AM1 and C13 or its low level of synthesis in CNRZ 118 is not known. The sequencing of the busA region in these osmosensitive strains could provide an explanation.
The various origins of the loss of function in cremoris strains are interesting, because the recognition of such genes can furnish interesting clues to understand the evolution of prokaryotic genomes . Additional genome regions are probably subjected to evolution since the two L. lactis subspecies present many differences towards stress tolerance . A relaxation of selection pressure during the thorough cultivation of cremoris strains in milk – a rich and stable medium – might have resulted in the loss of important functions for the survival outside the dairy environment.
We thank P. Taillez for providing the strains of the CNRZ collection. We thank B. Poolman and T. van der Heide for the generous gift of purified BusAA and BusAB proteins. D.O. was the recipient of a fellowship from the Ministère de l'Education Nationale de la Recherche et de la Technologie.