Lactic streptococci, classically regarded as homolactic fermenters of glucose and lactose, became heterolactic when grown with limiting carbohydrate concentrations in a chemostat. At high dilution rates (D) with excess glucose present, about 95% of the fermented sugar was converted to L-lactate. However, as D was lowered and glucose became limiting, five of the six strains tested changed to a heterolactic fermentation such that at D = 0.1 h-1 as little as 1% of the glucose was converted to L-lactate. The products formed after this phenotypic change in fermentation pattern were formate, acetate, and ethanol. The level of lactate dehydrogenase, which is dependent upon ketohexose diphosphate for activity, decreased as fermentation became heterolactic with Streptococcus lactis ML3. Transfer of heterolactic cells from the chemostat to buffer containing glucose resulted in the nongrowing cells converting nearly 80% of the glucose to L-lactate, indicating that fine control of enzyme activity is an important factor in the fermentation change. These nongrowing cells metabolizing glucose had elevated (ca. twofold) intracellular fructose 1,6-diphosphate concentrations ([FDP]m.) compared with those in the glucose-limited heterolactic cells in the chemostat. [FDP]ui was monitored during the change in fermentation pattern observed in the chemostat when glucose became limiting. Cells converting 95 and 1% of the glucose to L-lactate contained 25 and 10 mM [FDP]in, respectively. It is suggested that factors involved in the change to heterolactic fermentation include both [FDP]in and the level of lactate dehydrogenase. Group N streptococci (Streptococcus cremoris, S. lactis, and S. diacetylactis) play a vital role in many commercial milk fermentations, in which their primary function is to convert lactose to lactic acid (18). Lactic streptococci are useful because they possess limited metabolic diversity and usually convert about 95% of the fermented sugar to L-lactate (23, 26). This homolactic fermentation of either lactose or glucose occurs in batch culture when organisms are grown anaerobically near pH 7 at 30°C. In contrast, heterolactic fermentation was observed during growth on galactose (26) and during growth on lactose of variants defective in either lactate dehydrogenase (LDH; 21) or the lactose phosphotransferase system and/or phospho-fB-D-galactosidase (7, 26), suggesting that these organisms have pathways which are not normally expressed. The alternative products reported include acetate, acetoin, C02, ethanol, formate, and glycerol. Unsuccessful attempts

No abstract

Bacteroides gingivalis W50 was grown in a chemostat under steady-state conditions at pH 7.5 ± 0.2 and a constant growth rate of 6.9 h for periods of up to 6 weeks (146 bacterial generations) in a complex medium. Hemin was capable of limiting the growth of cells up to a concentration of approximately 0.5 ,ug/ml since higher concentrations of hemin did not increase cell yields; cells grew in the absence of exogenously added vitamin Kl. Only a limited number of amino acids was metabolized during growth, but because none of these was totally depleted, the limiting nutrient under hemin excess conditions was probably a peptide. A range of fermentation products was produced under all conditions of growth; higher concentrations of cytotoxic metabolites such as propionate and butyrate were formed under hemin excess conditions, although more ammonia was released under hemin limitation. When viewed by electron microscopy, cells grown under hemin limitation appeared to be either coccobacillary or short rods and possessed few fimbriae per cell, but large numbers of extracellular vesicles could be seen both surrounding the cell surface and free in the environment. In contrast, cells grown under hemin excess conditions were more commonly coccus shaped and were more heavily fimbriated but had fewer extracellular vesicles. Marked differences were found in the susceptibility of mice to infection with cells grown under different concentrations of hemin. Cells transferred to media without any added hemin were avirulent, whereas those grown under conditions of hemin limitation (0.33 and 0.40 ,ug/ml) produced a 20 and 50% mortality in mice, respectively. In contrast cells grown under hemin excess always caused 100% mortality in mice, although this virulence was dose dependent. When virulent, the bacteria caused an extensive, spreading infection with necrosis of the skin and subcutaneous tissues. Collagen disintegration was seen histologically, implying a role for collagenase production in the pathogenicity of these bacteria.

Nine commonly isolated oral bacterial populations were inoculated into a glucose-limited and a glucose-excess (amino acid-limited) chemostat maintained at a constant pH 7.0 and a mean community generation time of 13.9 h. The bacterial populations were Streptococcus mutans ATCC 2-27351, Strep. sanguis NCTC 7865, Strep. mitior EF 186, Actinomyces viscosus WVU 627, Lactobacillus casei AC 413, Neisseria sp. A1078, Veillonella alkalescens ATCC 17745, Bacteroides intermedius T 588 and Fusobacterium nucleatum NCTC 10593. All nine populations became established in the glucose-limited chemostat although Strep. sanguis and Neisseria sp. were present only after a second and third inoculation, respectively. In contrast, even following repeated inoculations, Strep. mutans, B. intermedius and Neisseria sp. could not be maintained under glucose-excess conditions. A more extensive pattern of fermentation products and amino acid catabolism occurred under glucose-limited growth; this simultaneous utilization of mixed substrates also contributed to the higher yields (Y molar glucose) and greater species diversity of these communities. Microscopic and biochemical evidence suggested that cell-to-cell interactions and food chains were occurring among community members. To compare the reproductibility of this system, communities were established on three occasions under glucose-limitation and twice under glucose-excess conditions. The bacterial composition of the steady-state communities and their metabolic behaviour were similar when grown under identical conditions but varied in a consistent manner according to the nutrient responsible for limiting growth. Although a direct simulation of the oral cavity was not attempted, the results show that the chemostat could be used as an environmentally-related model to grow complex but reproducible communities of oral bacteria for long periods from a defined inoculum.

No abstract