606 R. J. BRITTEN AND F. T. MCCLURE 
different characteristics provides qualitative agreement with conclusions deriving 
from experiments on exchange and osmotic shock. 
The removal of the pool by osmotic shock (items 16, 17 and 18) is not an obvious 
prediction of the carrier model. The simplest explanation, from the point of view 
of this model, is that the sites themselves are temporarily affected during a transient 
period of distension of the cell. The implication is that the macromolecules on which 
the sites are located are temporarily distorted in such a way that the affinity of the 
site for amino acid is drastically reduced. This could result from a direct change in 
the hydration of the macromolecule itself or could result from a mechanical coupling 
of the macromolecule to major cell structures. It must be pointed out that a semi- 
permeable membrane is not necessary in order that osmotic phenomena occur. An 
ion-exchange column (Dowex-50, 2° cross-linked) will undergo striking volume 
changes when sucrose solutions of different concentrations are passed over it. 
From this point of view, some interesting speculation about item(s) 1g (and 10) 
may be indulged in. The size of a very large pool, which is non-specific, is roughly 
proportional to the osmotic strength of the medium. Thus, the maximum pool is 
somehow related to the osmotic balance of the cell although even the largest amino 
acid pools account for only a small fraction of the total osmotically-active material 
that may be released from the cell. It may be suggested that there exist non-specific 
associations between the amino acids and.the dense protoplasm (25° dry material) 
of the bacterial cell. The maximum quantity of amino acid in such an association 
might decrease with increasing hydration of the protoplasm. No such associations 
are observed in relatively dilute protein solutions or in disrupted suspensions of cells. 
However, a carrier or energy donor present in the living cell might be a necessary 
condition in the formation of such an association. 
The fact that different pool compounds are removed in varying degrees by a 
given osmotic shock (item 18) probably reflects the differences in sensitivity of site 
complexes. 
Finally, it is difficult to leave the discussion of this model without some speculation 
on the nature of the carrier. The properties required of the carrier, for its function 
in this model, are that it be a large enough molecule to form a stereospecific association 
with an amino acid and, on the other hand, that it be small enough to diffuse with 
some freedom within the cell. Further, it must have a high affinity for free amino 
acid and be able to give up amino acid freely to form a site complex. In turn, un- 
occupied carriers must be able to accept amino acids from the site complexes. 
At least two possible candidates for the carrier are known at present. The lipid— 
amino acid complex observed by HENDLER!? in the hen oviduct has been observed 
in £. colt. The quantity and rapidity of labeling of such complexes in tracer ex- 
periments are consistent with the possibility of their function as carriers in the sense 
of this model. However, the molecular weight is unknown. The S-RNA-—amino acid 
complex discovered by HoaGLAND also occurs in EF. coli in very small quantities. 
However, there is no indication™ that the rate of turnover of the amino acid in this 
complex is fast enough to carry out the function of the carrier. 
A crude lower limit on the amount of carrier present in the cell can be set from its 
turnover number and rate of diffusion. At the maximum rate of proline pool forma- 
tion, there are 40 000 molecules entering the pool per second per cell. The time 
required for a small molecule such as proline to diffuse Iw is approx. I msec. The 
References p. 609 
