Immobilization-based isolation of capsule-negative mutants of Streptococcus pneumoniae.

The capsular polysaccharide (CPS) is the most important identified virulence factor of Streptococcus pneumoniae, a human pathogen of the upper respiratory tract. One limitation in studies of S. pneumoniae surface virulence factors is the lack of a reliable procedure for isolation of capsule-negative mutants of clinical strains. This paper presents an approach, based on the immobilization of pneumococci in semi-liquid (0.04 % agar) medium, to easily distinguish and select for non-capsulated mutants. A clinical S. pneumoniae type 37 strain was used as a model to show that CPS production results in bacterial immobilization in semi-liquid agar medium and restricts cell sedimentation. Descendants of CPS(-) mutants sedimented faster under these conditions and therefore could be separated from immobilized parental cells. The CPS(-) phenotype of the obtained mutants was confirmed by both immunoagglutination and immunostaining experiments using specific type 37 capsular antibodies. Complementation of immobilization with the cloned tts gene, encoding type 37 CPS synthase, confirmed that faster sedimentation of mutants was specifically due to loss of the capsule. DNA sequence determination of three independent mutants revealed a point mutation, a 46 nt deletion and a heptanucleotide duplication in the tts gene. Immobilization of strains producing other CPSs (type 2, 3 and 6) also resulted in the appearance of CPS(-) mutants, thus showing that immobilization-based isolation is not restricted to type 37 pneumococci. Bacterial growth in semi-liquid medium proved to be a useful model system to identify the genetic consequences of immobilization. The results indicate that immobilization due to CPS may impose selective pressure against capsule production and thus contribute to capsule plasticity.

Respiration capacity and consequences in Lactococcus lactis.

We recently reported that the well-studied fermenting bacterium Lactococcus lactis could grow via a respirative metabolism in the presence of oxygen when a heme source is present. Respiration induces profound changes in L. lactis metabolism, and improvement of oxygen tolerance and long-term survival. Compared to usual fermentation conditions, biomass is approximately doubled by the end of growth, acid production is reduced, and large amounts of normally minor end products accumulate. Lactococci grown via respiration survive markedly better after long-term storage than fermenting cells. We suggest that growth and survival of lactococci are optimal under respiration-permissive conditions, and not under fermentation conditions as previously supposed. Our results reveal the uniqueness of the L. lactis respiration model. The well-studied 'aerobic' bacteria express multiple terminal cytochrome oxidases, which assure respiration all throughout growth; they also synthesize their own heme. In contrast, the L. lactis cydAB genes encode a single cytochrome oxidase (bd), and heme must be provided. Furthermore, cydAB genes mediate respiration only late in growth. Thus, lactococci exit the lag phase via fermentation even if heme is present, and start respiration in late exponential phase. Our results suggest that the spectacularly improved survival is in part due to reduced intracellular oxidation during respiration. We predict that lactococcal relatives like the Enterococci, and some Lactobacilli, which have reported respiration potential, will display improved survival under respiration-permissive conditions.

Positive role of peptidoglycan breaks in lactococcal biofilm formation.

Bacterial attachment to solid matrices depends on adhesive molecules present on the cell surface. Here we establish a positive correlation between peptidoglycan (PG) breaks, rather than particular molecules, and biofilm-forming capacity in the Gram-positive bacterium Lactococcus lactis. The L. lactis acmA strain, which is defective in PG hydrolase, adhered less efficiently than the wild-type (wt) strain to different solid surfaces and was unable to form biofilms. These phenotypes were abolished by addition of lysozyme, a PG hydrolytic enzyme. Thus, the presence of PG breaks introduced by PG hydrolase, and not the AcmA protein itself, appears to be responsible for biofilm formation. Two different genetic screens confirmed the importance of PG breaks in L. lactis biofilm formation. Using the chain-forming ability of the acmA strain as a phenotypic indicator of PG integrity, we selected for insertional mutants generating short chains. Five independent mutants were all mapped to ponA, which encodes the PG synthesis enzyme PBP1A. Double acmA ponA mutants displayed increased adhesion and biofilm-forming capacity. Direct selection for strains with increased biofilm-forming capacity resulted in the isolation of another five mutations in ponA. Based on these results, we conclude that PG breaks are important for both adhesion and biofilm formation in L. lactis.