Growth of T-Strain Mycoplasmas in Medium Without Added Urea: Effect of Trace Amounts of Urea and of a Urease Inhibitor

Urea is currently considered to be a requirement for the propagation of T-strain mycoplasmas. We report here the replication of T-strain 960 (ATCC 25023) in media prepared from dialyzed components with added putrescine and allantoin but without added urea, or in dialyzed medium containing small amounts of added urea. The least amount of urea which allowed growth in the medium without allantoin was above 10 mug/ml. The amount of urea estimated to contaminate the added allantoin or putrescine was 5 mug/ml or less, which is insufficient to support T-strain replication. T-strain 960 was grown in the presence of urea and the urease inhibitor acetohydroxamic acid AHA where the organisms multiplied at a slower rate in the presence of AHA than in its absence. Urea hydrolysis occurred with concomitant ammonia accumulation and pH increase in cultures with AHA added.     

Effects of nutritional characteristics of Streptococcus agalactiae on inhibition of growth by lactoperoxidase-thiocyanate-hydrogen peroxide in chemically defined culture medium.

Five cultures of Streptococcus agalactiae have an absolute requirement for L-cystine to grow in a chemically defined medium. The L-cystine could be replaced with cysteine, glutathione, or the disulfide form of glutathione. Dithiothreitol could not substitute for the sulfur-containing amino acids of glutathione; hence, the growth requirement appears to be truly nutritional. Growth was maximum with 4 to 5 mug of L-cystine per ml. If the concentration of L-cystine was no greater than 4 to 5 mug/ml, complete growth inhibition could be obtained by the addition of lactoperoxidase, thiocyanate, and H2O2. The growth inhibition, however, was nullified by additions of L-cystine 10-fold or more in excess of the concentration needed for maximum growth. During the aerobic degradation of glucose by cell suspensions, H2O2 accumulation could be shown with cultures 317 and 11-13, the only cultures the growth of which was inhibited without addition of exogenous H2O2. All of the cultures had varying degrees of peroxidase activity. The balance between H2O2 generation and peroxidase activity of the culture evidently determined whether growth could be inhibited with lactoperoxidase and thiocyanate without H2O2 addition. The growth yeilds per 0.5 mol of the disulfide forms (cystine and oxidized glutathione) were 1.5 and 1.9 times greater than that per 1 mol of the sulfhydryl forms (cysteine and glutathione).     

The effects of bovine serum albumin, bovine uterine fluid, and heat-treated bovine serum on in vitro mouse embryo development

Two-cell mouse embryos were cultured in Whitten's medium with one of three supplements: bovine serum albumin (WM + BSA), heat-treated bovine serum (WM + HTBS) or bovine uterine fluid (WM + BUF). Protein concentrations for cultures of WM + BSA were 50.2, 100.5, 251.2, 502.5, and 1005.0 mug/ml and for WM + HTBS were 70.4, 105.1, 269.0, 524.5 and 1193.9 mug/ml. Protein concentrations ranged from 56.9 to 739.1 mug/ml for 22 WM + BUF samples. Embryo development in all media was significantly correlated with the log total protein concentration. When compared to WM + BSA, development was not significantly inhibited or stimulated in any WM + BUF cultures or in WM with 70.4, 524.5 and 1193.9 mug/ml HTBS. Development was enhanced in WM with 105.1 and 269.0 mug/ml HTBS (P<0.05). The results suggest that at the protein concentrations used, culture media supplemented with BUF and BSA support similar mouse embryo development. Culture medium supplemented with HTBS supported embryo development more than medium with BSA. Uterine factors in the bovine capable of enhancing or inhibiting early embryo development were not detected.     

Development of mouse embryos in media containing lectins

Lectins known to stimulate mitosis in cultured cells were evaluated for effects on development of mouse embryos in vitro. Two-cell mouse embryos were cultured in one of the following treatments: Whitten's medium as the control medium; Whitten's medium with 1, 10 or 100 mug/ml concanavalin A; Whitten's medium with 1, 10 or 100 mug/ml leucoagglutinin; Whitten's medium with 1, 10 or 100 mug/ml phytohemagglutinin; Whitten's medium with 1, 10 or 100 mug/ml pokeweed-mitogen; and Whitten's medium with 1, 10 or 100 mug/ml wheat germ agglutinin. Development to the morula stage was blocked in media with 100 mug/ml concanavalin A and 10 and 100 mug/ml wheat germ agglutinin, whereas blastocyst formation was blocked in all pokeweed-mitogen supplemented media. Embryos incubated in 10 and 100 mug/ml wheat germ agglutinin underwent premature cavitation or vacuolation at 24 to 48 h of culture. More embryos formed blastocysts in media with 1 and 100 mug/ml phytohemagglutinin and 10 mug/ml leucoagglutinin than in Whitten's medium (P<0.05). The percentage of embryos hatching was greatest in 1 mug/ml phytohemagglutinin (P<0.05), but it was the same in Whitten's medium, 1 mug/ml concanavalin A and 1 mug/ml leucoagglutinin (P>0.05). Cell division was not stimulated by the lectins; however, it was significantly suppressed in media with 10 and 100 mug/ml concanavalin A, 100 mug/ml phytohemagglutinin, 1, 10 and 100 mug/ml pokeweed-mitogen, and 10 and 100 mug/ml wheat germ agglutinin. Solubility of the zona pellucida in sodium isothicyanate (NaSCN) was reduced in 100 mug/ml phytohemagglutinin, 100 mug/ml leucoagglutinin and 1 mug/ml wheat germ agglutinin media (P<0.05) when compared to Whitten's medium and may have accounted for the reduced hatching observed in these treatments. Development of isolated blastomeres into blastocysts was reduced in media with 1 mug/ml wheat germ agglutinin, 1 mug/ml concanavalin A, and 10 and 100 mug/ml leucoagglutinin (P<0.05) but was similar in media with 1 mug/ml leucoagglutinin and 1, 10 and 100 mug/ml phytohemagglutinin when compared to Whitten's medium (P>0.05). The extent of embryo development in media with lectins depended upon the degree of cytotoxicity and potential biochemical modifications induced in the zona pellucida. Greatest embryo development took place in medium with 1 mug/ml phytohemagglutinin; however, the mechanism was not that of stimulation of cell division or a change in zona pellucida solubility.     

In Vitro Effect of Rifampin on Mycobacteria

Rifampin inhibited 20 strains of Mycobacterium tuberculosis in concentrations of 0.005 to 0.02 mug/ml in 7H-9 broth with Tween 80 and killed all or nearly all of the inoculum in four to eight times greater concentrations. In the same medium without Tween 80, as well as on 7H-10 agar, about 16 to 64 times these amounts were required to produce the same effect. Rifampin was also active against M. kansasii and some of the nonchromogenic mycobacteria. The incidence of mycobacterial cells resistant to rifampin within the cultures studied was in the range of one to four per 10(8) to 10(9) colony-forming units with concentrations of 4 to 125 mug of rifampin per ml. Only one of the Battey cultures and that of M. fortuitum yielded cells resistant to rifampin at 125 mug/ml but not at 500 mug/ml. The same strains yielded more than double that number of organisms resistant to streptomycin and up to 100 times more organisms resistant to isoniazid. All three drugs stopped the growth or reduced the mycobacterial population in growing cultures after contact for 24 to 48 hr. Complete inhibition of growth was produced by rifampin at 1.0 mug/ml in an average of 6 days and by streptomycin at 5.0 mug/ml in 3 days. After an average contact of 10.7 days with rifampin, five of seven strains resumed growth and all strains began regrowth after exposure to streptomycin for 9.4 days. The marked susceptibility of M. tuberculosis and of atypical mycobacteria to rifampin in vitro and the relatively low incidence of resistant mutants suggests that this agent may have clinical usefulness in the treatment of tuberculosis and some other mycobacterioses.     

Metabolism of Deoxycorticosterone by Human Fecal Flora

21-Dehydroxylation, a feature of metabolism of corticoids in humans, was observed in mixed cultures of fecal flora of normal individuals on a Western diet. The model substrate, 11-deoxycorticosterone (DOC), was metabolized to 3alpha-21-dihydroxy-5beta-pregnan-20-one (THDOC), 3alpha-hydroxy-5beta-pregnan20-one (pregnanolone), and to two unidentified structures, metabolites X and Y. DOC was not metabolized in all media supporting growth of fecal flora. Conversion required an initial pH between 6.0 and 8.0. 21-Dehydroxylation occurred within 4 days of incubation in media inoculated with 10-minus 1 to 10-minus 7 fecal suspensions. In higher dilutions, containing obligatory anaerobes only, DOC was converted to metabolite X and sometimes also to metabolite Y. The yield of pregnanolone was related to the promptness with which the specimen was processed, to the presence of cysteine in the medium, and to the concentrationof substrate (optimum, 16 to 64 mug of DOC per ml). The yield of THDOC was related to the delay in the processing of the specimen, the concentration of substrate (maximum at 256 mug/ml), and aeration of the culture. Pure cultures of aerobic organism of fecal origin either failed to metabolize DOC or converted it to metabolite Y. Pure cultures of fecal anaerobes converted DOC to metabolite X and sometimes also to metabolite Y. Neither THDOC nor pregnanolone was produced by pure cultures.     

Interaction of staphylococcal enterotoxin B with cell cultures of human embryonic intestine

Schaeffer, W. I. (Massachusetts Institute of Technology, Cambridge), J. Gabliks, and R. Calitis. Interaction of staphylococcal enterotoxin B with cell cultures of human embryonic intestine. J. Bacteriol. 91:21-26. 1966.-The cytotoxic effect of staphylococcal enterotoxin B upon human embryonic intestine cell cultures is characterized by retraction of cells from the monolayer. This is followed by clumping of the retracted cells to form clear areas in the monolayer and finally by sloughing of the clumps from the glass surface. The 50% effective dose of the toxin, determined by protein analysis of the cultures used in titration studies, was found to be between 40 and 60 mug/ml. The cytotoxic property of the enterotoxin was completely neutralized by 3.9 x 10(-5) ml of specific antitoxin per mug of toxin. The cytotoxicity was found to be slightly enhanced by 2.2 g of bicarbonate per liter of Eagle's basal medium (Earle's salt solution level), the absence of serum, the absence of penicillin and streptomycin, and the presence of 2.8 mmoles of calcium in the medium. The cytotoxicity was profoundly influenced by the age of the culture. No cytotoxicity was evident until after 2 days of growth had taken place, when the cell number was approximately 4.0 x 10(5) cells per culture.     

Effects of steroid hormones in fetal bovine serum on plating ang cloning of human cells in vitro.

Fetal bovine sera from each of three different commercial sources were tested for their ability to support cloning of human fibroblastoid cells in vitro. Cloning efficiencies varied according to serum source. Serum (10 samples) from company A did not support growth, while sera (10 samples) from companies B and C provided adequate to excellent conditions for cloning and growth. Cells from neonatal foreskin or embryonic lung responded to each serum similarly. Bovine serum albumin type H7 from company C supported cell growth in media without serum. Sera containing 1.0 ng per ml or more of progesterone inhibited growth, whereas sera containing less than 1.0 ng per ml supported cloning and growth. In the low progesterone sera, the concentration of 17-beta-estradiol exceeded 100 pg per ml. Growth supporting sera could be made non-supportive by adding 0.1 mug per ml of progesterone. The addition to non-supporative sera of 0.1 mug per ml of 17-beta-estradiol or hydrocortisone made these sera supportive of cell growth. Addition of estrogen or hydrocortisone to a culture medium that inhibits growth, with subsequent reversal of the inhibitory effect, implies that these hormones competitively regulate growth of responsive cells in vitro.     

Follicle stimulating hormone stimulation of 125-I-human chorionic gonadotropin binding in porcine granulosa cell cultures.

In order to see if FSH acts directly upon the granulosa cell to stimulate hCG binding, granulosa cells harvested from small 1-2 mm porcine follicles were grown in 250 ml flasks in chemically defined media containing 0.05 mug/ml highly purified human FSH for 2, 4, and 6 days. The defined medium consisted of culture medium 199 plus 0.4% bovine serum albumin, 0.2% lactalbumin hydrolysate and 10 munit/ml insulin. The cultures were harvested by scraping with a rubber policeman and incubated with 0.1 mug/ml 131-I- or 125-I-hCG. Binding expressed as cpm/culture or per mg protein yielded similar results. In five separate experiments addition of FSH stimulated hCG binding two- to fourfold above control cultures. In a typical experiment after 2 days of culture, the specific binding of control cultures to hCG was 962 plus or minus 45 cpm/culture (-x plus or minus SE; n = 3) and the binding in cultures grown in the presence of 0.05 mug/ml FSH was 3933 plus or minus 1787 (n = 3; P less than 0.01). Granulosa cells harvested from large (8-12 mm) follicles grown under similar conditions bound 29,669 plus or minus 948 cpm/culture (n = 4). These data demonstrate that FSH may have a direct stimulatory role upon induction of granulosa cell LH-hCG receptors in vitro.     

The urea requirement and urease production of some human species of T-mycoplasmas.

Seven species of human T-mycoplasmas that grow in Fraction A and 20 mug urea/ml died when the urea was omitted. Two species would not grow in Fraction A broth containing 10 mug/urea/ml. The other five strains grew in broth containing 10 mug urea/ml and were adapted by serial passage in broth containing decreasing concentrations of urea to grow in broth containing 2.5 mug/ml urea, but not in broth containing 1.25 mug/ml. Therefore the minimal urea requirement is not the same for the growth of all strains of T-mycoplasmas. In exponential phase broth cultures, urease was detected only intracellularly, none being found in the medium.     

Streptomycin resistance in chloramphenicol-producing strains of Streptomyces species 3022a.

Resistance to low (5 mug/ml) concentrations of streptomycin in agar media was not inherited by all of the surviving population. Outgrowth of cultures in liquid media supplemented with the antibiotic depended upon inoculum size. Antibiotic titers in the supplemented cultures decreased during incubation, and an inactive radioactive product was detected when [14C] streptomycin was used. This low-level resistance is, therefore, attributed to enzymic inactivation of the antibiotic. Growth 10 mug/ml or higher concentrations of streptomycin on agar media was due to selection of resistant variants present in the parent strain. A range of such variants existed, decreasing in frequency as their degree of resistance increased. Examination of one that was resistant to moderate concentrations of streptomycin, (25 mug/ml) and a second that was resistant to high (100 mug/ml) concentrations of streptomycin suggested that both possed ribosomes which had lower affinity for the antibiotic than those of the parent strain, and that tolerance to high levels of streptomycin was due to a resistant ribosomal system for protein biosynthesis.     

Optimal Combination and Concentration of Antibiotics in Media for Isolation of Pathogenic Fungi and Nocardia asteroides

Strains of Blastomyces dermatitidis, Sporothrix schenckii, Histoplasma capsulatum, Cryptococcus neoformans, Nocardia asteroides, and Coccidioides immitis were tested for in vitro susceptibility to polymyxin, gentamicin, kanamycin, chloramphenicol, and neomycin at concentrations of 1, 2, 4, 8, and 16 mug/ml. Polymyxin was the most inhibitory and gentamicin was the least inhibitory of the five antibiotics. Two Histoplasma mycelial strains were partially inhibited by 2 and 8 mug of gentamicin per ml and showed at least a 2+ growth at the higher antibiotic concentration. Kanamycin and neomycin produced significant inhibition of N. asteroides but otherwise were noninhibitory. A combination of chloramphenicol and kanamycin, each at 16 mug/ml, and gentamicin, at 4 mug/ml, was noninhibitory to the strains tested except for N. asteroides. Chloramphenicol at 16 mug/ml was not inhibitory for N. asteroides. The results suggest that the optimal antibiotic combination to use in the isolation of fungi and higher bacteria is chloramphenicol, 16 mug/ml, and gentamicin, 4 mug/ml. Addition of sheep blood (5%) had no effect on antibiotic susceptibility of the organisms studied.     

Effect of Isomeric cis-Octadecenoic Acids on the Growth of Leptospira interrogans Serotype patoc

Leptospira interrogans serotype patoc exhibited an increasing growth response when cultivated in media containing from 50 to 250 mug of sodium oleate per ml. Leptospiral growth in the presence of 250 mug of sodium oleate per ml was as good as that in the basal medium which contained 700 mug of oleic acid (in Tween 80) per ml. When positional isomers of oleic acid (9-octadecenoic acid) were present at a concentration of 200 mug/ml, the 2- and 8-isomers were not readily utilized, whereas the 3-, 4-, 6-, 11-, 15-, and 16-isomers gave a growth response equivalent to that of oleic acid, i.e., the 9-isomer. The 5-, 7-, 10-, 12-, 13-, 14-, and 17-isomers of octadecenoic acid induced growth responses which differed in magnitude but were intermediate to those of 2-18:1 and 3-18:1. When 200 mug of either 2- or 3-octadecenoic acid per ml was added in addition to 200 mug of 9-18:1 alone; 400 mug of 9-18:1 alone per ml inhibited growth of this organism. The growth response of leptospira to octadecenoic acids differed from that of mammalian cells, suggesting the presence of different enzymes in the two systems for the utilization of these substrates.     

Effect of Nitrofurans and Chlortetracycline on Microorganisms Associated with Shrimp

Nitrofuran AF-2 displayed greater inhibitory effect than did nitrofuran Z when a mixed bacterial culture, including several proteolytic bacteria, isolated from shrimp was subjected to these compounds in vitro. Nitrofuran Z exhibited greater bactericidal properties than did chlortetracycline in all cultures used. Only 10 mug of nitrofuran AF-2 per ml was sufficient to inhibit the growth of mixed bacteria in nutrient broth, whereas 50 mug of nitrofuran Z per ml was necessary to accomplish the same inhibition. A 50-mug amount of chlortetracycline per ml displayed about the same inhibitory effect as either 10 mug of AF-2 per ml or 20 mug of Z per ml. The isolated proteolytic bacteria showed greater suppression of growth when subjected to AF-2 than when subjected to Z; however, both nitrofurans were effective in preventing growth. The addition of either 1 mug of AF-2 per ml or 5 mug of Z per ml to nutrient broth inhibited the growth of Achromobacter aquarmarinus, whereas chlortetracycline was less effective, requiring about 20 mug to suppress growth to the same degree.     

Enterotoxin B Synthesis by Replicating and Nonreplicating Cells of Staphylococcus aureus

Although 95% of the enterotoxin B produced by Staphylococcus aureus appears during the latter part of the exponential phase of growth, growth per se is not necessary for toxin synthesis. A procedure is described whereby a concentrated suspension (at least 6 x 10(10) cells per ml) of a 16-hr culture of S. aureus was found to be capable of producing toxin, without replication, when air and glucose were present. This technique allows the growth requirement to be separated from toxin formation. Although higher (100 mug/ml) concentrations of toxin appeared in the medium when nitrogen was present, lower levels (30 mug/ml) were produced in the absence of N-Z-amine A. Toxin production proceeded without any net increase in deoxyribonucleic acid, ribonucleic acid, or protein. Chloramphenicol did not inhibit toxin formation in a nitrogen-free medium. The optimal pH for toxin production in a nitrogen-free medium was 8.0 to 8.5; for synthesis in a medium where nitrogen was available, the optimal pH was 7.0 to 7.5. Increasing the rate of aeration increased toxin release during growth, but decreased the amount of toxin subsequently produced when the bacteria were resuspended. These results suggest the presence of a precursor pool in the cells collected after 16 hr of growth.     

Rifampin-Blood-Agar as a Selective Medium for the Isolation of Certain Anaerobic Bacteria

A selective medium which allows detection of relatively small numbers of Fusobacterium varium in fecal specimens is described. Blood-agar containing 50 mug of rifampin per ml inhibits the growth of many species of Bacteriodes and of F. fusi-forme/nucleatum but allows good growth of F. varium and most strains of F. mortiferum. Quantitative cultures of 11 fecal specimens were done on rifampin and other selective and nonselective media. F. varium was recovered in counts of 10(6) and 10(7) per gram from two specimens on rifampin only. A third specimen yielded 10(10)F. varium on several media, including rifampin. Some Eubacterium and Clostridium species also grew on rifampin, and these ordinarily were distinguished from the Fusobacterium by colony morphology. This medium is of value in fecal flora studies and should be useful with other kinds of specimens where mixtures of organisms are common.     

Influence of Macromolecular Biosynthesis on Cellular Autolysis in Streptococcus faecalis

The addition of several different antibiotics to growing cultures of Streptococcus faecalis, ATCC 9790, was found to inhibit autolysis of cells in sodium phosphate buffer. When added to exponential-phase cultures, mitomycin C (0.4 mug/ml) or phenethyl alcohol (3 mg/ml) inhibited deoxyribonucleic acid synthesis, but did not appreciably affect the rate of cellular autolysis. Addition of chloramphenicol (10 mug/ml), tetracycline (0.5 mug/ml), puromycin (25 mug/ml), or 5-azacytidine (5 mug/ml) to exponential-phase cultures inhibited protein synthesis and profoundly decreased the rate of cellular autolysis. Actinomycin D (0.075 mug/ml) and rifampin (0.01 mug/ml), both inhibitors of ribonucleic acid (RNA) synthesis, also reduced the rate of cellular autolysis. However, the inhibitory effect of actinomycin D and rifampin on cellular autolysis was more closely correlated with their concomitant secondary inhibition of protein synthesis than with the more severe inhibition of RNA synthesis. The dose-dependent inhibition of protein synthesis by 5-azacytidine was quickly diluted out of a growing culture. Reversal of inhibition was accompanied by a disproportionately rapid increase in the ability of cells to autolyze. Thus, inhibition of the ability of cells to autolyze can be most closely related to inhibition of protein synthesis. Furthermore, the rapidity of the response of cellular autolysis to inhibitors of protein synthesis suggests that regulation is exerted at the level of autolytic enzyme activity and not enzyme synthesis.     

Induction of Prodigiosin Biosynthesis after Shift-Down in Temperature of Nonproliferating Cells of Serratia marcescens

Nonpigmented bacteria obtained by growth of Serratia marcescens at 38 C synthesized prodigiosin at 25 C if certain individual amino acids were added to cultures of nonproliferating cells. In order of effectiveness, the amino acids were: DL-histidine, L-proline, L-hydroxyproline, DL-alanine, L-alanine, DL-aspartic acid, D-alanine, DL-proline, L-serine, L-ornithine, L-glutamic acid, and D-proline. DL-Histidine at its optimal concentration (20 mg/ml) induced formation of prodigiosin (198 mug of prodigiosin per mg of bacterial protein) after incubation of cultures for 54 hr. Lower concentrations (10 mg/ml) of the other amino acids usually were optimum but less prodigiosin was synthesized, and the maximal amount of pigment occurred between 36 and 48 hr. DL-Methionine was not effective alone but at a low concentration (40 mug/ml) enhanced and accelerated biosynthesis of prodigiosin in the presence of other suitable amino acids. Addition of 2 mg of L-proline per ml at 0 hr induced formation of only 30 mug of prodigiosin after incubation for 42 hr, but addition at 36 hr of 5 mg more of L-proline per ml increased synthesis to 120 mug at 42 hr. Again, DL-methionine markedly augmented prodigiosin biosynthesis in these cultures. Synthesis of prodigiosin ceased if cultures were shifted from 25 to 38 C. Prodigiosin biosynthesis by the nonproliferating cells was maximum when cultures were aerated, the amount of bacterial protein was about 2.0 mg/ml, and amino acids were added at 0 hr. Bacteria synthesized prodigiosin most efficiently when they were harvested from aerated cultures grown at 38 C for 24 hr in a complete medium in a fermentor.     

Interactive effects of solutes, potassium sorbate and incubation temperature on growth, heat resistance and tolerance to freezing of Zygosaccharomyces rouxii.

The interactive effects of solutes, potassium sorbate and incubation temperature on growth, heat resistance and tolerance to freezing of Zygosaccharomyces rouxii were investigated. Growth rates in media supplemented with glucose, sucrose or NaCl to aw 0.93 were more rapid than in unsupplemented media (aw 0.99). Although growth in unsupplemented medium was lower at 35 degrees C, incubation at 21 degrees C or 35 degrees C had little effect on growth in media supplemented with glucose and sucrose. The addition of 300 micrograms potassium sorbate/ml to media resulted in reduced growth rates, particularly at 35 degrees C. Heat resistance of Z. rouxii was substantially greater in cultures previously incubated at 35 degrees C than in cultures incubated at 21 degrees C in media both with and without 300 micrograms potassium sorbate/ml. Zygosaccharomyces rouxii was tolerant to freezing at -18 degrees C for up to 120 d in all test media supplemented with glucose, sucrose or NaCl. The addition of 300 micrograms potassium sorbate/ml to sucrose-supplemented media resulted in increased resistance to freezing in cultures previously incubated at 21 degrees C. Sensitivity to freezing increased when cultures were incubated at 21 degrees C in media not supplemented with solutes. Glucose and sucrose provided the best protection against inactivation by heating and freezing, regardless of the presence of potassium sorbate in growth media.