109 



have to be between the in vivo and in vitro action, distribution 

 being random in vitro. We can circumvent this difficulty in two 

 ways. If a drug exerts its biological action by acting on £* then it 

 will have to have an active nucleus (conjugated double bonds, 

 NO2 or SCN groups, iodine, etc. ) , by which it exerts this action. 

 Our drug in question, if it acts by acting on £*, will have to pro- 

 duce an effect on £* in vitro at the same concentration as related 

 compounds which have no specific affinities, lacking the specific 

 structural details to which specific affinity is due, but having the 

 same active nucleus. To quote an example: if thyroxine owes its 

 biological action to its iodine, then we can expect it to act on £* 

 in vitro in the same concentrations as does free iodide or other 

 organic compounds of iodine. The other way to circumvent this 

 difficulty would be to look out for the less specific symptoms of 

 the drug in question (for no drug action is entirely specific) . If E* 

 has a general importance for cell life then we can expect any drug 

 acting on £* to affect all cells to some extent. To produce such 

 unspecific symptoms with a specific drug, we have to administer it 

 in higher doses than are needed to produce its specific action. We 

 can expect that the drug will affect £* also in vitro in the concen- 

 tration in which it produced its unspecific action in vivo. To illus- 

 trate this point: chlorpromazine (Chapter 16) has a specific effect 

 on certain nerv'ous centers but in higher doses it affects the whole 

 basal metabolism. If its pharmacological activity is due to its action 

 on £*, then we can expect it to affect E* in vitro in the same con- 

 centration as it affected the basal metabolism in vivo. 



Unfortunately, when comparing active concentrations in vitro 

 and in vivo, we will also have difficulty in making final statements 

 about the concentrations in our test tube with its frozen contents. 

 Water has a tendency to crystallize out in pure condition, leaving 

 dissolved substances behind in higher concentration. So, however 

 rapid our cooling may be, there is always a chance that the actual 

 concentrations were higher than corresponds to a random distri- 

 bution. 



