[ Pobierz całość w formacie PDF ]
.The assay consistedin the microscopic examination of vaginal smears obtained from castratedfemale mice (and later, rats) after the animals had been inoculated withknown quantities of the material under test (Figure 10.3).A positive resultwas associated with the disappearance of white blood cells and the appear-ance of epithelial cells.One mouse or rat unit was the minimum dose thatinduced a full change in histology.15 In other words, the unit for the follicle-hormone, later estrogens, was defined as the volume of preparation capableof inducing the differentiation of the vaginal epithelium in rodents.Standardization is often described as a means for reducing variability inorder to achieve a reasonable homogeneity of products and the replicationof scientific or medical results.Within the context of biological laborato-ries, this kind of total control was a distant goal whose realization requiredvast amounts of work and the mobilization of huge material resources.Incidentally, this movement would also result in the disappearance ofan important source of experimental innovation, rendering productionprocesses less imaginative.Laboratory standardization is therefore usuallylimited and highly local.In the context of early hormone production, bio-logical tests were the means of a more thorough, although less ambitioushandling of variability.The problematic sources of variation targeted inthe physiological literature were numerous: the animals (rats did not reactlike mice; strains could be sensitive or resistant); the modes of administra-tion (a single inoculation did not produce the same effect as fractionated 184 The Visible IndustrialistFigure 10.2 Register of urine testing at Scheringinoculations, and subcutaneous injections did not produce the same resultsas intravenous ones); the timing of the assay (mice did not react well indry, cold weather); the number of animals used, etc.Thus, the physiolo-gists focused on three aspects to achieve standardization: a) they employed Jean-Paul Gaudillière 185Figure 10.3 Mouse assay for estrogens ca.1925 normal bodies (selecting strains of animals that behaved normally andcould be bred in controlled ways); b) uniformity of practices (e.g.definingthe number of smears to be taken and the number of inoculations); c) estab-lishing the statistical significance of the results by using large numbers (e.g.establishing reference curves using hundreds of animals).A good assay wastherefore less a matter of purity and composition, than a matter of stability,specificity and most importantly biological significance.In the research stage, industrial standardization could follow the sameroutes, but when it came to issues of development, quality control andsurveillance of production standard experimental systems were replacedby standard test systems.One obvious difference between the two derivedfrom the difference in scale and productivity.An industrial assay needed toplace the emphasis on being rapid, easy to learn, and cost-effective rather 186 The Visible Industrialistthan being biologically meaningful since it was to be implemented on amuch larger scale than a research procedure, which, while it had to berepeatable, could afford to be maximally comprehensive and informative.These constraints can be illustrated by the fate of the chicken crest assay forthe male hormone that was developed by Schering s test laboratory.Schering s procedure used the growth of the crest of a castrated chicken tomeasure the potency of male hormones and was developed in collaborationwith the biochemists working at the KWIB mentioned above.The standardprotocol proposed by Butenandt and his colleagues emphasized a form of pre-cision based on the strict control of gestures and materials.16 In the early 1930s,the test was defined in the following terms: Chickens from the white Leghornrace are castrated by the age of 8 10 weeks.Repeated measurements of thesurface area of the crest will certify that in the next 8 10 weeks this sexualtrait remains stable.Only the animals displaying such stability may be usedfor the experiments.The first phase is the inoculation of a control solutionwith known hormone content in order to check the animal s responsiveness.The animals not reacting according to the norm or showing a crest, whichdoes not return to its initial size after the end of the test will be discarded.(& )Measurement of hormone potency is the growth of the crest.This is determinedin the following way.Photographs of the crest are shot before the test inocula-tion, and after three and four days.In order to obtain these two photographs,one must place a piece of sensitive photographic paper behind the crest.Thescene is briefly illuminated.The surface of the crest is then measured with aplanimeter.The chicken unit is defined as the dosage which, in a solution of1cm3 given in two successive days, induces an increase by an average of 20 percent of the surface area of the crest as measured in three different birds
[ Pobierz całość w formacie PDF ]
