nelson: screening methods in microbiology 313 



in "fermentor flasks", inoculated from a previously grown "vegeta- 

 tive flask", stirred and aerated on rotary shakers at constant tempera- 

 ture. The misnomer "fermentation" is industrial jargon for a highly 

 aerobic process. The activity of the antibiotic is measured by bioassay 

 which has a readily determinable error, usually less than 5 per cent. 

 Problems of bio-assay are discussed in a recent article (68). 



An apparent increase in yield may be due to the production of 

 related antibiotics, hence all new strains must be tested for the pres- 

 ence of other components. The major problem in shake flask testing 

 has been the large variation between separate runs under apparently 

 identical conditions. Attention to many details is necessary to reduce 

 this variation below 30 per cent when determined by an analysis of 

 variance between and within runs. While different antibiotic fer- 

 mentations will each be found to have their own peculiarities, the 

 major factor causing variation between runs has been the different 

 states of the vegetative growth. The age and extent of vegetative 

 growth can be determined and controlled on pilot plant and produc- 

 tion scales but not in shake flasks, where smaller volumes of media 

 and larger numbers of tests are required. Several inoculations of 

 vegetative growth at different times may be used to compensate for 

 different ages of growth. 



Any claims of increased antibiotic yield should be accompanied 

 by a measure of reliability based on repetitive determinations in 

 separate experiments, preferentially using pilot plant scale equip- 

 ment. Use of sequential analysis as an aid in making decisions has 

 been described (91). An increase in yield may occur on "scaling-up" 

 the fermentation from the smaller laboratory equipment through 

 the pilot plant. Since the same culture and media are used in both 

 shake flasks and fermentor tanks, the fortuitous increase in yield is an 

 indication that the medium can support more antibiotic production 

 than is found in shake flasks. The limiting factor in laboratory scale 

 fermentations may be the physical characteristics of the equipment, 

 often giving insufficient aeration and agitation. Strains selected for 

 higher yield in shake flasks may possess a different response to these 

 physical limitations and not scale-up to larger equipment. 



Purification of Culture 



A new antibiotic-producing culture is usually "unstable", pro- 

 ducing colonies with different morphology and variegated sectors 



