768 Editor: E. HEINZ 
repel the solute itself, and a difference in the activity of the available water. This reminds me that 
Dr. CHRISTENSEN’s remarks about the possible values of activity coefficients may give a key to some 
of the species differences that seem to be observed between animal and bacterial cells. For example, 
there is no question that E. coli is not isoosmotic with frog Ringer solution, as anybody knows who 
has worked on the stabilization of EF. coli protoplasts. In other words, in some bacterial cells the 
osmolarity is clearly very much higher than it is in animal cells, and it may be, although I can’t 
produce figures for it at the moment, that the activity coefficient of water is a great deal lower and 
may be down to the point that Dr. CHRISTENSEN wasimplying would never be reached. It would have 
to be painstakingly calculated in terms of analyses that I don’t think have been made yet. 
I would like to add, since Dr. BRITTEN referred to it one of the experiments with protoplast. If you 
make penicillin protoplasts of E. coli and lyse them thoroughly by homogenization you get some- 
thing that looks like a badly rumpled protoplast, but not like a membrane. Some people lyse 
protoplasts and publish nice photographs and say: “These are photographs of cell membranes.” 
I think they are photographs of cell ghosts, at least in the cases I have worked with. But at all 
event, if you lyse protoplasts of E. coli in this fashion, after having induced the formation of 
f-galactosidase to the maximum capacity of the protoplast system to form it, and if you then assay 
it by means of ONPG, you find that unless you treat the protoplast with toluene, you get only 
about 25% of the /-galactosidase activity that has been developed. In other words, you could 
speak here of some kind of osmotic barrier, that survives the rather thorough osmotic and mechani- 
cal disruption of the intact protoplast. I feel that this is partial evidence, that the osmotic effects 
are not all at the surface of the protoplast. 
CHRISTENSEN: I wonder if I may shift your attention elsewhere for a moment, before I have to 
leave this discussion. I refer you toa line of evidence that Dr. HErnz has pioneered, namely the driving 
of exchange processes (HEINZ AND WaLsu?:’). Experiments we are now making have emphasized 
the value of this approach, namely by comparing the effect of an amino acid inside the cell with 
its effectiveness outside the cell, in driving the counterflow of another amino acid. We find that 
valine, for example, is almost exactly as effective in driving the counter movement of other 
amino acids, whether the valine is inside and the other amino acid outside or whether they are 
in the reversed positions. If you make a profile of the response of five selected amino acids this 
profile of migration in response to valine is essentially the same whatever the relative initial 
position of the two amino acids under study, inside or outside. Essentially the same result is 
obtained down the line, using each of the amino acids to provide the driving force for exchange. 
The behavior certainly leads one to feel that he is dealing with the same mass-action situation 
at the two faces of the membrane, that the amino acid reactant is the same in the two cases, and 
that the measured levels are approximately correct. 
I hope that Dr. Heinz will continue the discussion of this behavior. 
Hernz: I think Dr. CHRISTENSEN is referring to what we call the preloading effect, or the counter- 
flow-effect as it is called by others. This effect, which we interpreted as due to a partial exchange 
diffusion has already been discussed earlier in this meeting. To review it briefly, we found many 
years ago that the uptake of some labeled amino acids by Ehrlich cells is greatly enhanced after 
preincubation of these cells with the same—or a related—amino acid in the unlabeled form’:*. In 
other words, the influx of the labeled amino acid, at constant extracellular level, rises almost 
proportionally with the intracellular concentration (tvansconcentration) of the amino acid with 
which the cell has been preloaded. Normally after adding the label the cellular radioactivity rises 
asymptotically towards the steady state value. After preloading, however, the cellular radio- 
activity, apart from rising more rapidly, may temporarily exceed the final steady state value. 
Furthermore we found evidence that the increment in influx, as it is caused by the preloading, is 
almost stoichiometrically equivalent to the carrier bound exit (counterflow) of the cellular amino 
acid. Such an effect obviously corresponds to the definition of exchange diffusion, as originally 
given by Ussinc. As already mentioned there is also evidence that this exchange diffusion refers 
to the same mechanism as the active transport of the amino acid, and for this reason we cal] it 
partial exchange diffusion. 
The preloading effect appeared to us as strong evidence of an active transport process mediated 
by a mobile carrier. The details of this argumentation have been reported elsewhere’: §. The phe- 
nomenon seemed also to suggest that the major part of the accumulated amino acids is in a free 
state, or at least that binding to an intracellular, non-mobile site, if it occurs, is not responsible 
for the accumulation process. I have to admit, however, that this evidence, though very suggestive, 
is not entirely conclusive. It is known that e.g. activated complexes may exchange their ligands 
much faster than they are formed de novo. So Horzer recently reported, that labeled acetaldehyde 
is bound to a thiaminpyrophosphate much faster if this coenzyme had been preincubated with 
unlabeled acetaldehyde. Such kind of mechanism should, of course, be considered in explaining 
the accumulation of amino acids and in particular the preloading effect. The fact that the forma- 
tion of an activated complex requires the expenditure of energy would agree with the high sensi- 
tivity of the accumulation process towards metabolic inhibitors and anoxia. It should be kept in 
References p. 777 
