ROUND TABLE DISCUSSION 767 
passively; diffusion will make a contribution, and indeed a stoichiometric binding of a portion of a 
solute may well occur in addition to an uphill transport. 
Chemical mediation of the entrance of solute molecules into cells seems to be extremely varie- 
gated. It is not limited to metabolically useful solutes. For instance, one cannot imagine that the 
human red blood cell needs its enormous capacity to admit glycerol. The cell boundary may 
present such a wide spectrum of mediating sites that the chemist can hardly synthesize a hydro- 
philic molecule that will not somehow interact with any of these sites. Why some of these mediating 
groups can release the molecule at a higher chemical potential than they acquired it, and into a 
different phase, is the interesting question here. 
To an earlier comment I want to reply that none of us here, I think, looks on the membrane in a 
completely morphologic sense. We are thinking in the physicochemical sense of a boundary 
between two phases, and no assumption is made about the thickness of the membrane phase; 
nor do we assume that the dissociation from a transport site (or series of such sites) occurs at a 
given distance from the association reaction. How deep the penetration goes before another event 
intervenes may be merely a statistical matter. In short we look on the membrane in a functional 
sense, and let the morphologist try to find a structural basis for the function that we observe. 
We are also trying to take into account possible general changes in activity coefficients arising 
from the unusual character of the cytoplasmic environment. In this connection the largest 
distribution ratios are interesting, because they demand the most improbable diminutions of the 
activity coefficient. When conscientious efforts are made to measure the activity coefficients of 
ordinary solutes in broken cell preparations, no great changes are seen from those applying outside 
the cell. One may, of course, say that as soon as we touch the cell the vital state within has been 
entirely changed, so that no such observations are valid. But consider that for many solutes the 
activity coefficients would need to fall from values or nearly unity to perhaps 0.01 or 0.001 within 
the cell. Without ignoring this uncertainty, we need to consider how large it can plausibly be. 
By selecting our solute and the level used we can get extremely high distribution ratios—and I see 
no reason why in exploring this matter—we should limit ourselves to amino acids of biological 
occurrence and to physiological levels. By selecting favorable models we can require that a general 
fall in activity coefficients must be enormous to account for the extent of accumulation by cells, 
just as we can also set the requirement for binding sites improbably high by achieving high 
gradients. Once model solutes have been shown to set such extreme demands on the binding and 
inactivation hypotheses, we need to show that these models are appropriate models and that the 
conclusions may be applhed to the ordinary amino acids at ordinary levels. 
LajtuHa: There are some experiments of ours mentioned in more detail elsewhere in this volume, 
which I interprete as transport into what I call a free solution of amino acids. I find active trans- 
port a more reasonable explanation of these experiments than the binding of amino acids. In the 
higher organisms (rats 77 vivo) one can produce and maintain high plasma levels by continuous 
infusion. Upon subsequent intraspinal or intracerebral injection, brain levels also increase but still 
stay below the elevated plasma levels. During the experiment the brain level decreases in time 
against a concentration gradient of the high plasma levels. In this case amino acids are actively 
transported from an organ into the plasma, unless we suppose that most of the amino acid in the 
plasma is bound. In some of these experiments plasma levels were increased more than 10~15 fold. 
The transport process is fairly specific in that under the same experimental circumstances 
leucine is transported out of the brain until a plasma to brain ratio of 11-15 fold is reached, 
whereas lysine transport stops at a plasma to brain ratio of 5—6. No transport of phenylalanine is 
found. Leucine transport, present in adults, could not be shown in newborns in similar experi- 
ments; we interpreted this as showing that the transport system for this amino acid is not fully 
developed in newborns. 
These experiments show transport from the brain into a pool which most likely consists of free 
amino acids. 
Ho.LpeEN: I wonder whether Dr. REINER’s formulation can be made consistent with the apparent 
dependence of some swelling phenomena on metabolism. 
REINER: It seems to me that you could make an argument for the potential difference between 
the two phases being the barrier in the extreme case. If you calculate conventionally a distribution 
ratio, for example, between ether and water as two adjacent phases, there will be certain sub- 
stances that will distribute themselves mostly in the water and very little in the ether. The 
barrier is their solubility in the ether relative to their solubility in water. Although a cell is a much 
more complicated system than this, I think that the same fundamental principles hold. Obviously 
there is some evidence for special molecular layers at the surface of some cells. So you probably 
have not a simple jump of potential energy but whole series of gradations of energy. There may 
also be a potential barrier, as I drew it originally, extending not over a couple of Angstroms but 
over several molecular layers. The principle, however, would be the same. You can keep a solute 
out, without the intervention of various mechanisms such as have been discussed this past week, 
merely by a difference in solubility, which could mean a difference in the forces that attract or 
References p. 777 
