Zbl. Bakt. 277, 161-169 (1992) © Gustav Fischer Verlag, Stuttgart/New York
Optimization of Culture Conditions for Toxin Production of Type G Clostridium botulinum MARIA C. CALLERI D E MILA N , LUI S S. MA YORGA , and OLGA N . PUIG DE C E N T O R BI* Catedra de Bacteriolog ia II and Caredra de Bioesradfstica, Univerdad Nacional de San Luis, 5700 San Luis, Argentina
With 1 Figure' Received Jun e 3, 1991 . Accepted in revised form January 10, 1992
Summary Culture conditions were optimized for toxin production of Clostridium botulinum type G, the last toxigenic serotype described. Six factorial experiment s were performed to assess the effect of medium composition (nutrients, metal ions, sterile soil, pH ), incubation conditions (time, temperature and anerobiosis) and associated microorganisms (Bacillus subtilis, Lactobacillus plantarum) on the toxinogen esis of C. botulinum type G. A significant (4 to 10 fold) improvement of toxin production was obtained by using an optimized medium (3% proteose - peptone , 0.5% trypticase, 1.1% glucose, 0.5 % yeast extract, adjusted to pH 8.0) and incubation of the culture for 12 days at 26 °C in a nitrogen atmosphere.
Zusammenfassung Es wurden die Kulturb edingungen fur den als am wenigsten toxigen beschriebenen Seroryp G von Clostridium botulinum zum Zweck der Toxinbildung optim iert. Es wurden 6 Versuche mit untersch iedlichen Faktoren durchgefiihrt, urn die Auswirkun gen der Nahrbodenzusammensetzung (Na hrstoffe, Met allionen, steriler Boden, pH) auf die Toxinbildung von C. botulinum Typ G zu beurteilen. Durch die Verwendun g eines optimierten Nahrbodens (3% Proteose-Pepton, 0,5% Tryptica se, 1,1% Glucose, 0,5% Hefeextr akt , eingestellt auf pH 8,0) und Inkub ation der Kultur iiber 12 Tage bei 26 °C in einer Stickstoffatmosphare wurde eine signifikanre (4- lOfache) Verbesserung der Tox inbildung erzielt.
Clostridium botulinum type G, the sero type described last, wa s isolated for the first time from a soil sample in Argent ina (6). More recently, Sonnabend et al. (12), in Switzerland, identified C. bo tulinum type G and its assoc iated neurotoxin in necropsy specimens in cases of unexpl ained death in ad ults and infants . These findings, although
* Corr esponding auth or J 1 Zbl. Rake 277/2
162
M. C. Calleri de Milan, L. S.Mayorga, and O. N. Puig de Centorbi
not conclusive, suggest that C. botulinum type G may be related to human disease. On the other hand, its toxic effect on animals has been well characterized (3). When C. botulinum type G is cultured in standard media used for toxin production of other serotypes, it proved to be the least toxigenic one. This fact has hampered the purification of type G botulinal toxin and the production of antisera required for identification and neutralization in human therpay. The aim of the authors has been to optimize the conditions for toxin production for this serotype. Since several factors in the medium composition and incubation conditions can affect toxin production, factorial designs were used to test a large number of different culture conditions using a minimal number of assays.
Materials and Methods Organisms. C. botulinum type G strains 117 and 89 were grown from lyophilized samples maintained in chopped meat medium (CM). In some experiments, C. botulinum was grown in association with Bacillus subtilis or Lactobacillus plantarum. Nutrients. Trypticase (BBL), proteose peptone (Oxoid), yeast extract (Oxoid), glucose, lactose and glycerol were tested in different concentrations and combinations. The effect of several cations (Fe 2+, Co2 +, Cu 2 + and Ca2 +) was also assessed. Incubation conditions. Several temperatures, anaerobic conditions and incubations times were tested. Anaerobiosis was obtained by a) the anaerobic jar (Oxoid) and b) the gas substitution technique. Gas subtitution was carried out in a vacuum desiccator flushed with nitrogen or with a propane butane mixture. Iron fibre acidified with acetic acid was used as a reducing agent. Under all conditions, methylene blue was used as redox indicator. Toxin production. As a routine, 20 ml of medium to be tested were autoclaved at 120°C for 20 min, cooled and inoculated with two drops of a 24 h culture of C. botulinum type G, strain 89, grown in chopped meat medium. After incubation under the appropriate conditions, the medium was centrifuged at 10000 x g for 10 min at 4°C to separate the bacteria. The approximate toxin content in the supernatant was obtained by i.p. injection 0.5 ml of 5 or 10 fold dilutions of the centrifuged medium into mice. The titre was empirically calculated taking into account the deaths counted up to 96 h after the injection, the time elapsed till death, and the onset of symptoms in the surviving mice. Quantitative titrations were performed for final comparison of the optimized conditions with the conditions used for toxin production in previous studies (7). In these experiments, samples of cultures as described above, or using the cellophane tubing method (13) were diluted at log, base. Mice were injected i.p. with 0.5 ml of the dilutions (six animals per dilution) and the titre was calculated by using the Reed and Muench method (8). When required, samples were activated with trypsin before titration using the method of Duff et al. (5). Experimental designs and statistical methods (4). Six experiments using factorial designs were performed. All possible combinations of up to five factors tested at two or three levels were assayed for toxin production. The mean square of the error was estimated from duplicate assays and from interactions of an order of three or higher. Only rarely, these high order interactions were significant. For each experiment, the error variance was estimated with at least 5 degrees of freedom.
Results First experiment
Six experiments using factorial designs were performed to optimize the conditions for toxin production of C. botulinum type G. The first experiment was designed to assess the requirements of nutrients for toxinogenesis. The effects of the nutrients were
Toxin Production for Type G C. botulinum
163
expressed as quotients between the LDso/ml obtained at one level of a factor (arbitrarily called "high") and the LDso/ml measured at the other level of the same factor (arbitrarily called "low"). Interactions were measured by the different effect observed for one factor when assessed at two different levels of another factor. For example, Table 1 shows that by adding 2 g/100 ml of proteose peptone, the toxin production was increased by a factor of 3.7 and by adding trypticase, the LDsolml was increased by a factor of 5.9. However, by adding both proteose peptone and trypticase, the total effect was 3.7 x 5.9 x 0.2 = 4.4. This indicates that the presence of one kind of protein hydrolyzate interfered with the stimulatory effect of the other. The presence of glucose and lactose were not important for toxin production. Yeast extract had an inhibitory effect at high concentration.
Table 1. Effect of nutrients on the toxin production by type G C. botulinum. Five nutrients were assayed at two different concentrations as indicated in the table (2s experimental design). Other incubation conditions were: pH, 7.0; incubation period, 10 days; incubation temperature, 32°C; anaerobiosis, propane-butane mixture Factors
Trypticase Proteose-peptone Yeast extract Lactose Glucose
Levels (%, w/v) Low
High
0 0 0.5 0 0
3 2
1 1 1
Effects* Main effect 5.9 3.7 0.4 ns ns Interactions 0.2
TrypticaseProteose peptone
* The effects are expressed as a quotient of the toxin production at the "high" level of a
factor divided by the toxin production at the "low" level of the same factor. Only significant interactions between two factors are shown (p < 0.05). Higher order interactions were pooled with the error term. ns, non significant effect (p > 0.05).
Second experiment As a result of the first experiment it was decided to test two media, one rich in proteose peptone and the other rich in trypticase. Glucose was included since it does not have any negative effect and yeast extract was added at low concentration. The addition of two different aminoacid preparations was also assessed. Inocula to seed the media were prepared from a logarithmic-phase culture that was replicated in CM Finally, the media were adjusted medium and incubated for 24 h either 32°C or to two different pH values. The results of these experiments are shown in Table 2. Medium B (rich in proteose peptone) gave a better toxin production than medium C. The addition of peptone hydrolysis products did not improve the toxin production and the pH did not have any significant effect. Inocula grown at 37°C gave better results.
3rc.
164
M. C. Calleri de Milan , L. S. Mayo rga, and O . N. Puig de Centorbi
Table 2. Toxinogenesis of type G C. botulinum in different culture media and under different culture conditions. Five factors were assayed at two levels as indicated in the ta ble (25 experimental design). Ot her incubation conditions were: incu bation period, 10 days; incubation temperature, 32 °C j anae robiosis, gas substitution using a propane-butane mixture . See Tab le 1 for more deta ils Levels
Facto rs
Medium" Inocule temp . * ,. Pr.-pep . aa (%) * " * Tryp aa (% j *" Medium pH
Effects
Low
High
C 32 °C 2 2 7.2
B
3rC 5 5 7.6
Main effect 7.3 3.4 0.22 ns ns Int eractions ns
* Medium B: 0.5 g % trypticase, 3.0 g % proteose-peptone, 1.1 g % glucose, 0.5 g % yeast extract. Medium C: 3.5g % trypt icase, 0.5g % protepse-peprone, 0.7g % glucose, 0.7 g % yeast extract. ** The inocula to seed the media were grow n in CM for 24 h at either 32 °C or 3r c. » * * Aminoaci ds were ob tained from prote ose-peptone of trypt icase by acid hydrol ysis (14). Th e preparat ions were neutralized with NaOH before adding th em to the media.
Table 3. Effect of incubation conditions and metal ions on the toxinogenesis of type G C. botulinum. Five factor s were tested at two levels (25 factor ial design). Other incubation condition were: medium, B (see Tab le 2); incubation per iod, 10 days Levels
Factors
Anaero biosis., Incuba t. temp . 'f" Med ium pH Fe2 + (mg/l 00 ml)" »» Co 2 + and Cu 2 + .:. •, .:.• ,
Effects
Low
High
propane 32 °C 7.6 0.08 0.00
Jar 37 °C 8.0 0.20 0.02 and
M ain effect 3.16 0.09 ns ns ns
0.01 Anaero biosis-pH
Inter actions 4.7
* Anaerob ic conditions were obtaine d by propane-butane mixture (pro pane) or by the jar method (jar). * * Toxin production was assessed at two temperatures of incubation. ** * Iron was tested at the concentration existing in th e medium (determined by atomic absorptio n) or after addition of FeS04 up to 0.20 mg/lO O m!. * * * . Th e effect of these catio ns wa s tested by adding CoCh and CUS04 to final concentration s of 0.02 and 0.01 mg/100 ml, respectively.
Toxin Production for Type G C. botulinum
165
Third experiment In this experiment, the addition of some salts and incubation conditions were assessed. Table 3 shows that an addition of metal ions was not required. Toxin production, however, was strongly affected by the temperature of incubation and the method used to generate anaerobic conditions. The pH by itself did not affect toxinogenesis, but it showed an important interaction with anaerobiosis. A more alcaline pH stimulated toxin production when the anaerobiosis was obtained by the "jar" method.
Fourth experiment After the previous three experiments, it was decided to compare the optimized medium (medium B) with the medium originally used for toxin production (medium A) (7). Three different systems of achieving anerobic conditions were also tested. Table 4 shows that medium B gave better results than medium A and that cultures incubated in a nitrogen atmosphere or in the anaerobic jar produced a higher toxin yield than those incubated in a propane-butane atmosphere.
Table 4. Toxinogenesis of type G C. botulinum in two media using three different methods to generate anaerobiosis (3 x 2 experimental design). Other incubation conditions were: incubation period, 10 days; incubation temperature, 32°C Levels
Factors
Anaerobiosis * Medium"*
Effects
Low
High
propane propane A
jar
N2 B
Main effect 3.0 2.0 2.3 Interactions ns
* Anaerobic was obtained by gas substitution with a propane-butane mixture (propane)
or with nitrogen (N2 ) or by the jar method (jar).
** Medium A is the medium used for toxin production for other C. botulinum serotypes (3.0g % trypticase, 2.0g % proteose-peptone, 1.0g % glucose, 1.0g % yeast extract, pH 7.0). Medium B was the same as described in Table 2, adjusted to pH 8.0.
Fifth experiment In this experiment, we observed that a lower temperature stimulated toxin production. A slightly higher production was obtained after 12 days of incubation. The presence of an associated microorganism like B. subtilis also favoured toxin production. However, association with 1. plantarum completely inhibited toxin production (results not shown). The addition of calcium or sterile soil did not induce any effect (Table 5).
166
M. C. Calleri de Milan, L. S. Mayorga, and O. N. Puig de Centorbi
Table 5. Effect of five factors assayed at two levels on the toxinogenesis of C. botulinum type G (25 experimental design). Other experimental conditions were: medium B (as described in Table 2) adjusted to pH 8.0; anaerob iosis in jar Levels
Factors
Incub. period (days) Incubat. temp. B. subtilis" CaC0 3 (g %) soil (g % )**
Effects
Low
High
6 26°C
12
32 °C present 0.2 1.0
0 0
Main effect 1.9 0.02 2.8 ns ns Interactions ns
* B. subtilis was grown for 24 h in CM and two drops of a 1:1000 dilution were added to the media. * * Garden soil was dried at 37"C, sieved, and added to the media.
Sixth exp eriment This experiment suggested that the op timal temperature of incubation was 2 6°C since a lower temperature inhibited toxin production. The addition of glycerol had a negative effect. Longer incubation did not improve toxin production and two methods to generate anaerobic conditions (i.e. jar and nitrogen) were equally suitable for toxin production. In this experiment, a medium containing high concentrations of both trypticase and proteose-peptone was compared with the relatively nutrient-poor med ium we have selected in the second experiment. Toxin production was slightly higher in the latter medium (Table 6).
Table 6. Effect of five factors assayed at two levels on the toxinogenesis of type G C. botulinum (2 5 experimental design) Levels
Factors
Incub. temp. Incubat. period (days) Anaerobiosis Glycerol (g % ) Medium"
Effects
Low
High
23 °C 10 jar 0 A
26 °C 15 N2 1 B
Main effect 9.3 ns ns 0.17 1.8 Interactions ns
»
Medium A, as described in Table 4; medium B, as described in Table 2, adjusted to pH 8.0.
Toxin Production for Type G C. botulinum
167
Conventional conditions versus optimized conditions The conditions selected for the previous experiments were tested against the conditions previously used, to assess whether any improvement had been achieved. Ta ble 7 shows that the optimized conditions yielded a higher production than the original ones.
Table 7. Production of type G botulinal roxin under conventional and optimized conditions Strain
Title (LDsolml)
Culture
Opt.lConv.
Conventional * * Optimized G89
Total
1.10 x lO z
1.12 X 10 3
10.2
4
G89 G89
Dialysis Dialysis"
5.00 x 10 2.77 x lOs
1.78 x 105 1.60 X 106
3.5 5.8
G1l7 Gl17
Dialysis Dialysis"
1.72 x 105 1.60 x 106
2.26 X 105 6.70 X 106
1.4 4.2
* Toxin was activated with trypsin (5) before titration. ** Conventional condition s: bacteria were grown in medium A (as described in Table 4) for 10 days at 32 °C in a propane-butane atmosphere. Optimized condition s: bacteria were grown in medium B (as described in Table 2), adjusted to pH 8.0, for 12 days at 26 °C in an N z atmo sphere.
1,4
•
1.2
1.0 ?: .;;; r:::
Q>
"0
0,8
•
~
ii
0
0,6
o o
0,4
0,2
0 0
3
6
9
12
Days of incubation
Fig. 1. Bacterial density of C. botulinum type G grown under optimized conditions (0) or conventional conditions (e) . Bacterial density was estimated from the optical density at 580 nm of total cultures. After 15 days, the toxin concentration was 1120 and 110 DLsolml for optimized conditions and conventional conditions, respectively.
168
M. C. Calleri de Milan, L. S.Mayorga, and O. N. Puig de Centorbi
This was true not only for total cultures but also for culture in dialysis tubing and for two different strains of C. botulinum type G. The difference in toxin production was enhanced by trypsin activation of the toxins, suggesting that a higher proportion of inactive toxin was produced under optimized conditions. Bacterial growth as assessed by optical density was similar in both conditions. However, after day 5, optical density decreased under the optimized conditions while under the original conditions there was no change up to day 15 (Fig. 1). Discussion Production of botulinal toxins is of general interest for research and clinical purposes. Unfortunately, the latest isolated serotype, type G, produces a low amount of toxin when grown in media optimized for toxin production by other serotypes. We have made an effort to optimize incubation conditions for toxin production by this serotype. Since there is a large number of variables that can affect toxin production in C. botulinum and other microoganisms we have used factorial designs to assess several factors simultaneously. Despite the limited number of mice used to estimated toxin production, the factorial designs showed clearly the effect of relevant factors on the toxinogenesis. The more important factors were the temperature incubation and the method to generate anaerobiosis. We have optimized the temperature for toxin production which is in the range of 23°C to 37°C. The best yield was obtained at 26 "C. Our results agree with the observations of other authors who studied the influence of temperature and time of incubation on toxin production by C. botulinum type G (10, 11). A medium relatively poor in nutrients gave better results than the original one. On the other hand, the addition of several elements that favoured the toxinogenesis of other microorganisms, e.g. metal ions (9), calcium (15), glycerol, did not improve toxin production by C. botulinum type G. The fact that bacterial yield, as assessed by optical density, decreases under the optimized conditions suggests that, at least in part, the higher toxin production may be related to the autolysis process (1, 2). The conditions selected increased the toxin yield not only in the system where they were optimized (i.e., test tube cultures using strain 89) but also in cellophane tubing culture using strain 117. A significant improvement on the toxin production has been achieved by using optimized conditions. However, the titres obtained were still not comparable to those observed for other more toxigenic serotypes of C. botulinum. We speculate that other factors, probably related with the genetic information of type G are affecting toxin production. Incubation conditions may be unable to suppress those limiting factors for toxinogenesis. At present our aim is to select clones that produce higher titres of type G botulinal toxin. Acknowledgements. Thanks are due to Mr. Alfredo R. Villegas for his efficienttechnical assistance. This work has been supported by funds from Consejo Nacional de Investigaciones Cientfficas y Tecnicas, Republica Argentina.
References 1. Bonventre, P. F. and L. L. Kempe: Physiology of toxin production by Clostridium botulinum types A and B. J. Bact. 7 (1959) 372-374
Toxin Production for Type G C. botulinum
169
2. Boroff, D. A.: Studies of toxins of Clostridium botulinum. III. Relation to autolysis to toxin production. J. Bact. 70 (1955) 363-367 3. Ciccarelli, A. S., D. N. Whaley, L. M. Mc Croskey, D. F. Gimenez, V. R. Dowel, and Ch. L. Hatheway: Cultural and physiological characteristics of Clostridium botulinum type G and susceptibility of certain animals to its toxin. Appl. Environ. Microbiol. 34 (1977) 843-848 4. Cochran, W. G. and G. M. Cox: Disefios experimentales. Editorial Trillas, Mexico (1980) 5. Duff,]. T., G. G. Wright, and A. Yarinsky: Activation of Clostridium botulinum type E by trypsin. J. Bact. 72 (1956) 455-460 6. Gimenez, D. F. and A. S. Ciccarelli: Another type of Clostridium botulinum. Zbl. Bakt., I. Abt. Orig. 215 (1970) 221-224 7. Puig de Centorbi, O. N., A. S. Ciccarelli, and D. F. Gimenez: Toxoides de toxina botulinicatipo G. Rev. Arg. Microbiol. 12 (1980) 10517 8. Reed, L. ]. and H. Muench: A simple method of estimation fifty percent endpoints. Amer. J. Hyg. 27 (1938) 493-497 9. Siegel, L. S.: Effect of iron on growth and toxin production by Clostridium botulinum type A. Curro Microbiol. 6 (1981) 127-130 10. Solomon, H. M. and D. A. Kautter: Sporulation and toxin production by Clostridium botulinum type G. J. Food Prot. 42 (1979) 965-967 11. Solomon, H. M., D. A. Kautter, and R. K. Lynt: Effect of low temperatures on growth of nonproteolytic Clostridium botulinum types Band F and proteolytic type G in crabmeat and broth. ]. Food Prot. 45 (1982) 516-518 12. Sonnabend, 0., W. Sonnabend, R. Heinzle, T. Sigrist, R. Dirnhofer, and U. Krech: Isolation of Clostridium botulinum type G and identification of type G botulinal toxin in humans: report of five sudden unexpected deaths. J. Infect. Dis. 143 (1981) 22-27 13. Sterne, M. and L. M. Wentzell: A new method for the large-scale production of hightitre botulinum formol-toxoid types C and D. J. Immunol. 65 (1950) 175-183 14. Suttie, ]. W.: Fundamentos de Bioquimica, 2a. ed. Editorial Interamericana, Mexico (1979) 15. Whitmer, M. E. and E. A. Johnson: Development of improved defined media for Clostridium botulinum serotypes A, Band E. Appl. Environ. Microbiol. 54 (1988) 753-759
Dra. Olga Puig de Centorbi, Chair of Bacteriologi II, Universidad Nacional de San Luis, 5700 San Luis, Argentina
Optimization of Culture Conditions for Toxin Production of Type G Clostridium botulinum MARIA C. CALLERI D E MILA N , LUI S S. MA YORGA , and OLGA N . PUIG DE C E N T O R BI* Catedra de Bacteriolog ia II and Caredra de Bioesradfstica, Univerdad Nacional de San Luis, 5700 San Luis, Argentina
With 1 Figure' Received Jun e 3, 1991 . Accepted in revised form January 10, 1992
Summary Culture conditions were optimized for toxin production of Clostridium botulinum type G, the last toxigenic serotype described. Six factorial experiment s were performed to assess the effect of medium composition (nutrients, metal ions, sterile soil, pH ), incubation conditions (time, temperature and anerobiosis) and associated microorganisms (Bacillus subtilis, Lactobacillus plantarum) on the toxinogen esis of C. botulinum type G. A significant (4 to 10 fold) improvement of toxin production was obtained by using an optimized medium (3% proteose - peptone , 0.5% trypticase, 1.1% glucose, 0.5 % yeast extract, adjusted to pH 8.0) and incubation of the culture for 12 days at 26 °C in a nitrogen atmosphere.
Zusammenfassung Es wurden die Kulturb edingungen fur den als am wenigsten toxigen beschriebenen Seroryp G von Clostridium botulinum zum Zweck der Toxinbildung optim iert. Es wurden 6 Versuche mit untersch iedlichen Faktoren durchgefiihrt, urn die Auswirkun gen der Nahrbodenzusammensetzung (Na hrstoffe, Met allionen, steriler Boden, pH) auf die Toxinbildung von C. botulinum Typ G zu beurteilen. Durch die Verwendun g eines optimierten Nahrbodens (3% Proteose-Pepton, 0,5% Tryptica se, 1,1% Glucose, 0,5% Hefeextr akt , eingestellt auf pH 8,0) und Inkub ation der Kultur iiber 12 Tage bei 26 °C in einer Stickstoffatmosphare wurde eine signifikanre (4- lOfache) Verbesserung der Tox inbildung erzielt.
Clostridium botulinum type G, the sero type described last, wa s isolated for the first time from a soil sample in Argent ina (6). More recently, Sonnabend et al. (12), in Switzerland, identified C. bo tulinum type G and its assoc iated neurotoxin in necropsy specimens in cases of unexpl ained death in ad ults and infants . These findings, although
* Corr esponding auth or J 1 Zbl. Rake 277/2
162
M. C. Calleri de Milan, L. S.Mayorga, and O. N. Puig de Centorbi
not conclusive, suggest that C. botulinum type G may be related to human disease. On the other hand, its toxic effect on animals has been well characterized (3). When C. botulinum type G is cultured in standard media used for toxin production of other serotypes, it proved to be the least toxigenic one. This fact has hampered the purification of type G botulinal toxin and the production of antisera required for identification and neutralization in human therpay. The aim of the authors has been to optimize the conditions for toxin production for this serotype. Since several factors in the medium composition and incubation conditions can affect toxin production, factorial designs were used to test a large number of different culture conditions using a minimal number of assays.
Materials and Methods Organisms. C. botulinum type G strains 117 and 89 were grown from lyophilized samples maintained in chopped meat medium (CM). In some experiments, C. botulinum was grown in association with Bacillus subtilis or Lactobacillus plantarum. Nutrients. Trypticase (BBL), proteose peptone (Oxoid), yeast extract (Oxoid), glucose, lactose and glycerol were tested in different concentrations and combinations. The effect of several cations (Fe 2+, Co2 +, Cu 2 + and Ca2 +) was also assessed. Incubation conditions. Several temperatures, anaerobic conditions and incubations times were tested. Anaerobiosis was obtained by a) the anaerobic jar (Oxoid) and b) the gas substitution technique. Gas subtitution was carried out in a vacuum desiccator flushed with nitrogen or with a propane butane mixture. Iron fibre acidified with acetic acid was used as a reducing agent. Under all conditions, methylene blue was used as redox indicator. Toxin production. As a routine, 20 ml of medium to be tested were autoclaved at 120°C for 20 min, cooled and inoculated with two drops of a 24 h culture of C. botulinum type G, strain 89, grown in chopped meat medium. After incubation under the appropriate conditions, the medium was centrifuged at 10000 x g for 10 min at 4°C to separate the bacteria. The approximate toxin content in the supernatant was obtained by i.p. injection 0.5 ml of 5 or 10 fold dilutions of the centrifuged medium into mice. The titre was empirically calculated taking into account the deaths counted up to 96 h after the injection, the time elapsed till death, and the onset of symptoms in the surviving mice. Quantitative titrations were performed for final comparison of the optimized conditions with the conditions used for toxin production in previous studies (7). In these experiments, samples of cultures as described above, or using the cellophane tubing method (13) were diluted at log, base. Mice were injected i.p. with 0.5 ml of the dilutions (six animals per dilution) and the titre was calculated by using the Reed and Muench method (8). When required, samples were activated with trypsin before titration using the method of Duff et al. (5). Experimental designs and statistical methods (4). Six experiments using factorial designs were performed. All possible combinations of up to five factors tested at two or three levels were assayed for toxin production. The mean square of the error was estimated from duplicate assays and from interactions of an order of three or higher. Only rarely, these high order interactions were significant. For each experiment, the error variance was estimated with at least 5 degrees of freedom.
Results First experiment
Six experiments using factorial designs were performed to optimize the conditions for toxin production of C. botulinum type G. The first experiment was designed to assess the requirements of nutrients for toxinogenesis. The effects of the nutrients were
Toxin Production for Type G C. botulinum
163
expressed as quotients between the LDso/ml obtained at one level of a factor (arbitrarily called "high") and the LDso/ml measured at the other level of the same factor (arbitrarily called "low"). Interactions were measured by the different effect observed for one factor when assessed at two different levels of another factor. For example, Table 1 shows that by adding 2 g/100 ml of proteose peptone, the toxin production was increased by a factor of 3.7 and by adding trypticase, the LDsolml was increased by a factor of 5.9. However, by adding both proteose peptone and trypticase, the total effect was 3.7 x 5.9 x 0.2 = 4.4. This indicates that the presence of one kind of protein hydrolyzate interfered with the stimulatory effect of the other. The presence of glucose and lactose were not important for toxin production. Yeast extract had an inhibitory effect at high concentration.
Table 1. Effect of nutrients on the toxin production by type G C. botulinum. Five nutrients were assayed at two different concentrations as indicated in the table (2s experimental design). Other incubation conditions were: pH, 7.0; incubation period, 10 days; incubation temperature, 32°C; anaerobiosis, propane-butane mixture Factors
Trypticase Proteose-peptone Yeast extract Lactose Glucose
Levels (%, w/v) Low
High
0 0 0.5 0 0
3 2
1 1 1
Effects* Main effect 5.9 3.7 0.4 ns ns Interactions 0.2
TrypticaseProteose peptone
* The effects are expressed as a quotient of the toxin production at the "high" level of a
factor divided by the toxin production at the "low" level of the same factor. Only significant interactions between two factors are shown (p < 0.05). Higher order interactions were pooled with the error term. ns, non significant effect (p > 0.05).
Second experiment As a result of the first experiment it was decided to test two media, one rich in proteose peptone and the other rich in trypticase. Glucose was included since it does not have any negative effect and yeast extract was added at low concentration. The addition of two different aminoacid preparations was also assessed. Inocula to seed the media were prepared from a logarithmic-phase culture that was replicated in CM Finally, the media were adjusted medium and incubated for 24 h either 32°C or to two different pH values. The results of these experiments are shown in Table 2. Medium B (rich in proteose peptone) gave a better toxin production than medium C. The addition of peptone hydrolysis products did not improve the toxin production and the pH did not have any significant effect. Inocula grown at 37°C gave better results.
3rc.
164
M. C. Calleri de Milan , L. S. Mayo rga, and O . N. Puig de Centorbi
Table 2. Toxinogenesis of type G C. botulinum in different culture media and under different culture conditions. Five factors were assayed at two levels as indicated in the ta ble (25 experimental design). Ot her incubation conditions were: incu bation period, 10 days; incubation temperature, 32 °C j anae robiosis, gas substitution using a propane-butane mixture . See Tab le 1 for more deta ils Levels
Facto rs
Medium" Inocule temp . * ,. Pr.-pep . aa (%) * " * Tryp aa (% j *" Medium pH
Effects
Low
High
C 32 °C 2 2 7.2
B
3rC 5 5 7.6
Main effect 7.3 3.4 0.22 ns ns Int eractions ns
* Medium B: 0.5 g % trypticase, 3.0 g % proteose-peptone, 1.1 g % glucose, 0.5 g % yeast extract. Medium C: 3.5g % trypt icase, 0.5g % protepse-peprone, 0.7g % glucose, 0.7 g % yeast extract. ** The inocula to seed the media were grow n in CM for 24 h at either 32 °C or 3r c. » * * Aminoaci ds were ob tained from prote ose-peptone of trypt icase by acid hydrol ysis (14). Th e preparat ions were neutralized with NaOH before adding th em to the media.
Table 3. Effect of incubation conditions and metal ions on the toxinogenesis of type G C. botulinum. Five factor s were tested at two levels (25 factor ial design). Other incubation condition were: medium, B (see Tab le 2); incubation per iod, 10 days Levels
Factors
Anaero biosis., Incuba t. temp . 'f" Med ium pH Fe2 + (mg/l 00 ml)" »» Co 2 + and Cu 2 + .:. •, .:.• ,
Effects
Low
High
propane 32 °C 7.6 0.08 0.00
Jar 37 °C 8.0 0.20 0.02 and
M ain effect 3.16 0.09 ns ns ns
0.01 Anaero biosis-pH
Inter actions 4.7
* Anaerob ic conditions were obtaine d by propane-butane mixture (pro pane) or by the jar method (jar). * * Toxin production was assessed at two temperatures of incubation. ** * Iron was tested at the concentration existing in th e medium (determined by atomic absorptio n) or after addition of FeS04 up to 0.20 mg/lO O m!. * * * . Th e effect of these catio ns wa s tested by adding CoCh and CUS04 to final concentration s of 0.02 and 0.01 mg/100 ml, respectively.
Toxin Production for Type G C. botulinum
165
Third experiment In this experiment, the addition of some salts and incubation conditions were assessed. Table 3 shows that an addition of metal ions was not required. Toxin production, however, was strongly affected by the temperature of incubation and the method used to generate anaerobic conditions. The pH by itself did not affect toxinogenesis, but it showed an important interaction with anaerobiosis. A more alcaline pH stimulated toxin production when the anaerobiosis was obtained by the "jar" method.
Fourth experiment After the previous three experiments, it was decided to compare the optimized medium (medium B) with the medium originally used for toxin production (medium A) (7). Three different systems of achieving anerobic conditions were also tested. Table 4 shows that medium B gave better results than medium A and that cultures incubated in a nitrogen atmosphere or in the anaerobic jar produced a higher toxin yield than those incubated in a propane-butane atmosphere.
Table 4. Toxinogenesis of type G C. botulinum in two media using three different methods to generate anaerobiosis (3 x 2 experimental design). Other incubation conditions were: incubation period, 10 days; incubation temperature, 32°C Levels
Factors
Anaerobiosis * Medium"*
Effects
Low
High
propane propane A
jar
N2 B
Main effect 3.0 2.0 2.3 Interactions ns
* Anaerobic was obtained by gas substitution with a propane-butane mixture (propane)
or with nitrogen (N2 ) or by the jar method (jar).
** Medium A is the medium used for toxin production for other C. botulinum serotypes (3.0g % trypticase, 2.0g % proteose-peptone, 1.0g % glucose, 1.0g % yeast extract, pH 7.0). Medium B was the same as described in Table 2, adjusted to pH 8.0.
Fifth experiment In this experiment, we observed that a lower temperature stimulated toxin production. A slightly higher production was obtained after 12 days of incubation. The presence of an associated microorganism like B. subtilis also favoured toxin production. However, association with 1. plantarum completely inhibited toxin production (results not shown). The addition of calcium or sterile soil did not induce any effect (Table 5).
166
M. C. Calleri de Milan, L. S. Mayorga, and O. N. Puig de Centorbi
Table 5. Effect of five factors assayed at two levels on the toxinogenesis of C. botulinum type G (25 experimental design). Other experimental conditions were: medium B (as described in Table 2) adjusted to pH 8.0; anaerob iosis in jar Levels
Factors
Incub. period (days) Incubat. temp. B. subtilis" CaC0 3 (g %) soil (g % )**
Effects
Low
High
6 26°C
12
32 °C present 0.2 1.0
0 0
Main effect 1.9 0.02 2.8 ns ns Interactions ns
* B. subtilis was grown for 24 h in CM and two drops of a 1:1000 dilution were added to the media. * * Garden soil was dried at 37"C, sieved, and added to the media.
Sixth exp eriment This experiment suggested that the op timal temperature of incubation was 2 6°C since a lower temperature inhibited toxin production. The addition of glycerol had a negative effect. Longer incubation did not improve toxin production and two methods to generate anaerobic conditions (i.e. jar and nitrogen) were equally suitable for toxin production. In this experiment, a medium containing high concentrations of both trypticase and proteose-peptone was compared with the relatively nutrient-poor med ium we have selected in the second experiment. Toxin production was slightly higher in the latter medium (Table 6).
Table 6. Effect of five factors assayed at two levels on the toxinogenesis of type G C. botulinum (2 5 experimental design) Levels
Factors
Incub. temp. Incubat. period (days) Anaerobiosis Glycerol (g % ) Medium"
Effects
Low
High
23 °C 10 jar 0 A
26 °C 15 N2 1 B
Main effect 9.3 ns ns 0.17 1.8 Interactions ns
»
Medium A, as described in Table 4; medium B, as described in Table 2, adjusted to pH 8.0.
Toxin Production for Type G C. botulinum
167
Conventional conditions versus optimized conditions The conditions selected for the previous experiments were tested against the conditions previously used, to assess whether any improvement had been achieved. Ta ble 7 shows that the optimized conditions yielded a higher production than the original ones.
Table 7. Production of type G botulinal roxin under conventional and optimized conditions Strain
Title (LDsolml)
Culture
Opt.lConv.
Conventional * * Optimized G89
Total
1.10 x lO z
1.12 X 10 3
10.2
4
G89 G89
Dialysis Dialysis"
5.00 x 10 2.77 x lOs
1.78 x 105 1.60 X 106
3.5 5.8
G1l7 Gl17
Dialysis Dialysis"
1.72 x 105 1.60 x 106
2.26 X 105 6.70 X 106
1.4 4.2
* Toxin was activated with trypsin (5) before titration. ** Conventional condition s: bacteria were grown in medium A (as described in Table 4) for 10 days at 32 °C in a propane-butane atmosphere. Optimized condition s: bacteria were grown in medium B (as described in Table 2), adjusted to pH 8.0, for 12 days at 26 °C in an N z atmo sphere.
1,4
•
1.2
1.0 ?: .;;; r:::
Q>
"0
0,8
•
~
ii
0
0,6
o o
0,4
0,2
0 0
3
6
9
12
Days of incubation
Fig. 1. Bacterial density of C. botulinum type G grown under optimized conditions (0) or conventional conditions (e) . Bacterial density was estimated from the optical density at 580 nm of total cultures. After 15 days, the toxin concentration was 1120 and 110 DLsolml for optimized conditions and conventional conditions, respectively.
168
M. C. Calleri de Milan, L. S.Mayorga, and O. N. Puig de Centorbi
This was true not only for total cultures but also for culture in dialysis tubing and for two different strains of C. botulinum type G. The difference in toxin production was enhanced by trypsin activation of the toxins, suggesting that a higher proportion of inactive toxin was produced under optimized conditions. Bacterial growth as assessed by optical density was similar in both conditions. However, after day 5, optical density decreased under the optimized conditions while under the original conditions there was no change up to day 15 (Fig. 1). Discussion Production of botulinal toxins is of general interest for research and clinical purposes. Unfortunately, the latest isolated serotype, type G, produces a low amount of toxin when grown in media optimized for toxin production by other serotypes. We have made an effort to optimize incubation conditions for toxin production by this serotype. Since there is a large number of variables that can affect toxin production in C. botulinum and other microoganisms we have used factorial designs to assess several factors simultaneously. Despite the limited number of mice used to estimated toxin production, the factorial designs showed clearly the effect of relevant factors on the toxinogenesis. The more important factors were the temperature incubation and the method to generate anaerobiosis. We have optimized the temperature for toxin production which is in the range of 23°C to 37°C. The best yield was obtained at 26 "C. Our results agree with the observations of other authors who studied the influence of temperature and time of incubation on toxin production by C. botulinum type G (10, 11). A medium relatively poor in nutrients gave better results than the original one. On the other hand, the addition of several elements that favoured the toxinogenesis of other microorganisms, e.g. metal ions (9), calcium (15), glycerol, did not improve toxin production by C. botulinum type G. The fact that bacterial yield, as assessed by optical density, decreases under the optimized conditions suggests that, at least in part, the higher toxin production may be related to the autolysis process (1, 2). The conditions selected increased the toxin yield not only in the system where they were optimized (i.e., test tube cultures using strain 89) but also in cellophane tubing culture using strain 117. A significant improvement on the toxin production has been achieved by using optimized conditions. However, the titres obtained were still not comparable to those observed for other more toxigenic serotypes of C. botulinum. We speculate that other factors, probably related with the genetic information of type G are affecting toxin production. Incubation conditions may be unable to suppress those limiting factors for toxinogenesis. At present our aim is to select clones that produce higher titres of type G botulinal toxin. Acknowledgements. Thanks are due to Mr. Alfredo R. Villegas for his efficienttechnical assistance. This work has been supported by funds from Consejo Nacional de Investigaciones Cientfficas y Tecnicas, Republica Argentina.
References 1. Bonventre, P. F. and L. L. Kempe: Physiology of toxin production by Clostridium botulinum types A and B. J. Bact. 7 (1959) 372-374
Toxin Production for Type G C. botulinum
169
2. Boroff, D. A.: Studies of toxins of Clostridium botulinum. III. Relation to autolysis to toxin production. J. Bact. 70 (1955) 363-367 3. Ciccarelli, A. S., D. N. Whaley, L. M. Mc Croskey, D. F. Gimenez, V. R. Dowel, and Ch. L. Hatheway: Cultural and physiological characteristics of Clostridium botulinum type G and susceptibility of certain animals to its toxin. Appl. Environ. Microbiol. 34 (1977) 843-848 4. Cochran, W. G. and G. M. Cox: Disefios experimentales. Editorial Trillas, Mexico (1980) 5. Duff,]. T., G. G. Wright, and A. Yarinsky: Activation of Clostridium botulinum type E by trypsin. J. Bact. 72 (1956) 455-460 6. Gimenez, D. F. and A. S. Ciccarelli: Another type of Clostridium botulinum. Zbl. Bakt., I. Abt. Orig. 215 (1970) 221-224 7. Puig de Centorbi, O. N., A. S. Ciccarelli, and D. F. Gimenez: Toxoides de toxina botulinicatipo G. Rev. Arg. Microbiol. 12 (1980) 10517 8. Reed, L. ]. and H. Muench: A simple method of estimation fifty percent endpoints. Amer. J. Hyg. 27 (1938) 493-497 9. Siegel, L. S.: Effect of iron on growth and toxin production by Clostridium botulinum type A. Curro Microbiol. 6 (1981) 127-130 10. Solomon, H. M. and D. A. Kautter: Sporulation and toxin production by Clostridium botulinum type G. J. Food Prot. 42 (1979) 965-967 11. Solomon, H. M., D. A. Kautter, and R. K. Lynt: Effect of low temperatures on growth of nonproteolytic Clostridium botulinum types Band F and proteolytic type G in crabmeat and broth. ]. Food Prot. 45 (1982) 516-518 12. Sonnabend, 0., W. Sonnabend, R. Heinzle, T. Sigrist, R. Dirnhofer, and U. Krech: Isolation of Clostridium botulinum type G and identification of type G botulinal toxin in humans: report of five sudden unexpected deaths. J. Infect. Dis. 143 (1981) 22-27 13. Sterne, M. and L. M. Wentzell: A new method for the large-scale production of hightitre botulinum formol-toxoid types C and D. J. Immunol. 65 (1950) 175-183 14. Suttie, ]. W.: Fundamentos de Bioquimica, 2a. ed. Editorial Interamericana, Mexico (1979) 15. Whitmer, M. E. and E. A. Johnson: Development of improved defined media for Clostridium botulinum serotypes A, Band E. Appl. Environ. Microbiol. 54 (1988) 753-759
Dra. Olga Puig de Centorbi, Chair of Bacteriologi II, Universidad Nacional de San Luis, 5700 San Luis, Argentina
