478 
HEMATOLOGY 
hibit the greatest variations; it clotted bovine 
fibrinogens ten times faster than chicken fibrin- 
ogen. With horse thrombin, the variation of 
speed was only about twofold. Thus, it would 
not make much difference for the horse if we re- 
placed horse fibrinogen by chicken fibrinogen or 
any of the others. In contrast, such a substitu- 
tion of fibrinogen in man could lead to serious 
complications. The rate of fibrinogen-thrombin 
interaction would become too fast or too slow 
depending on which fibrinogen is substituted 
for the human fibrinogen. For horse, an altera- 
tion (brought about by mutation) in the fibrino- 
gen molecules would hardly affect the rate of 
clotting; but for humans, a similar alteration 
could mean a great difference in the speed of 
clotting. Apparently, human thrombin magni- 
fies the differences which for horse thrombin ap- 
pear insignificant. 
The basic point that emerged from these ex- 
periments is that all the thrombins and fibrino- 
gens tested were different. During evolution, 
both fibrinogen and thrombin underwent modi- 
fications. 
I shall now attempt to correlate the data in 
Table II with those in Table I in order to see if 
some light can be shed on how evolution man- 
aged the fibrinogen-thrombin interaction. 
In the arguments, actual rates of clotting are 
not involved. I am utilizing only the order in 
which the various fibrinogens can be arranged 
(Table II) according to the speed of their 
clotting brought about by a given thrombin. 
Setting our sight on the fibrinogen first, the 
question naturally arises which portion of the 
fibrinogen molecule is responsible for the varia- 
tion in the speed of clotting. Let us look first at 
the peptides which are released during clotting. 
Naturally we have to look for that portion of 
the peptides where the variations is the greatest 
— where the amino acid residues show the great- 
est variation. Since, in these experiments, the 
initial rates were measured, we have to examine 
peptide A which comes off during clotting ear- 
lier than peptide B.^^ In other words, it is pep- 
tide A that can be expected to have an infiuence 
on the initial rate. 
A look at Table I shows the similarities of 
these peptides. All end at the C-terminal in ar- 
ginine. This is, of course, due to the trypsin-like 
specificity of thrombin liberating these peptides 
by cleaving the bond between arginine and gly- 
cine. If we number the positions in peptide A 
starting with arginine, we can see that up to po- 
sition 6, the residues in all the peptides are the 
same. The consistent presence of glutamic acid 
in position 6 is understandable. The main role 
of these peptides is to keep the fibrinogen mole- 
cules apart with their repelling negative 
charges. We expect this attribute of the pep- 
tides to be conserved during evolution. The 
greait similarity of these peptides shows that 
they are really homologous and probably go 
back to a common ancestor. 
Beyond position 11, the variations are becom- 
ing greater probably because up to this point, 
the essential requirements (charge in the 
proper position, complementariness to throm- 
bin) have been fulfilled. The variation is the 
greatest at position 13. Thus we may conclude 
that the amino acid residue in this position has 
an effect on the rate of the cleavage of the bond 
13 residues away. 
This is not an entirely unexpected finding. 
When we used synthetic peptides containing a 
bond susceptible to thrombin action, we found 
that elongation of the peptide 3 to 4 residues be- 
yond the bond split had a noticeable effect on 
the rate of splitting. For example, phe-gly-arg- 
amide (NH3 is split off by thrombin) is hy- 
droyzed faster than phe-arg-amide.^^ The 
action of thrombin apparently depends on what 
amino acid residue happens to be 13 residues 
away. 
A corollary to this conclusion is that if we 
have identical residues at this point, the cataly- 
tic action of thrombin should be closely identi- 
cal. Inspection of Table II reveals that in 6 out 
of 10 cases investigated, goat and sheep turn up 
together in the table. Six out of 10 thrombins 
clot goat and sheep fibrinogen with nearly the 
same rate. Thus, for 6 diflPerent thrombins, goat 
and sheep fibrinogen appear indistinguishable. 
This means that peptide A of goat and sheep fi- 
brinogens must be similar at position 13. It is 
seen from Table I that both goat and sheep pep- 
tides contain valine at position 13. 
Let us now examine the four instances in 
Table II where the thrombins are not strictly 
discriminating. With these four thrombins, the 
