atoms incorporated per 100 ev ab- 
sorbed. For purposes of computation 
it is assumed that all of the tritium 
radiation is absorbed by the sample. 
The results of several exposures and 
the values of Gy for both the crude 
mixture and the purified parent are 
shown in Table 1. Comparable G 
values for recoil labeling are smaller by 
a factor of 10°, which indicates that 
radiation damage for the same amount 
of labeling can be expected to be 
considerably smaller when the gas-ex- 
posure technique is used. As an ex- 
ample, it has been reported that ribo- 
nuclease and lysozyme can be labeled 
without great loss of enzymatic activity 
(6). 
Position of label. Wilzbach has in- 
vestigated the positions assumed by the 
label in toluene. The method is to 
separate derivative fractions and de- 
termine the amount of activity that 
remains with each one. From the 
data one can determine the fraction of 
the total toluene activity that is as- 
sociated with each of the hydrogen 
positions in the toluene molecule. 
The results of these investigations are 
shown in Table 2. 
Tritium Radioautography 
Conventional Labeling 
When the special nature of the 
labeled molecule does not require the 
special techniques of gas exposure or 
tritium recoil, simpler methods produce 
compounds of high specific activity. 
These include reduction, exchange, hy- 
drolysis, and synthesis by means of in- 
termediates. 
Reduction. This is the method of 
choice if a precursor is available that 
has a double bond or a reducible group 
such as a carbonyl radical. Activity 
as great as 58 c/millimole is possible by 
this method, although one could not 
expect to keep such a compound under 
control. However, curies per mil- 
limole are possible even after dilution 
with carrier hydrogen. 
Finding a suitable solvent is a 
problem. In polar solvents exchange 
between the tritium and solvent hy- 
drogen reduces the eventual activity by 
a large factor. Modification of the 
compound before the reduction and 
subsequent restoration to the desired 
form often simplifies the problem. 
Simple exchange. This method is 
limited by the fact that labels that go 
Spectacular print* shows distribution of tetracycline in a mouse sacrificed 
20 min after intravenous injection. 
figure is light. 
high concentration. 
Concentration is high where this 
Brain at right shows low, while kidney at upper left shows 
This exposure required 2.5 months. 
Short range of tritium betas produces resolution that is unusually high 
for radioautographs. 
frequently distinguishable. 
In microscopic enlargements individual cells are 
At the tritium symposium Walter L. Hughes gave figures to indicate 
the specific activity required in thymadine to label a mammalian cell: 
one locus requires 5 silver grains, which, at 5% efficiency, require 100 
disintegrations, which might occur during a 1-month exposure. 
~10~ curies per locus. 
This is 
Taking into account the composition of the aver- 
age cell, one finds that one needs 0.2 curies per mole of thymadine. 
* Torsten André, Farmakologiska Avdelningen, Kungl. Veterinérhégskolan (1956) 
published as supplementum 142 to Acta Radiologica. 
in easily come out easily; if a tracer is 
easily prepared, it is likely to be un- 
stable, and vice versa. The method is 
useful if the exchange-labeled molecule 
is quickly modified to a stable form 
after the initial labeling. 
Catalyzed reactions. Acid-cata- 
lyzed reactions are useful for aromatic 
compounds. The usefulness is limited 
by the tendency toward sulfonation. 
Other types of molecules, notably 
acetone and fatty acids, can be labeled 
by base- and metal-catalyzed reactions, 
respectively. 
Hydrolysis. The following equa- 
tions illustrate labeling by hydrolysis. 
As applied to benzene derivatives this 
method has the advantage of provid- 
ing specific labeling. Reactions with 
the Grignard reagent, illustrated by 
the first two equations, have rather 
general applicability. 
RMgX + H}0 — RH? + H8OMgX 
SOC], + H30 > SO. + 2H°Cl 
H°Cl + RMgX—RH! + MgXCl 
CaC2 + 2H}0 > 
H®C=CH? + Ca(OH?), 
KCN + 2H30 — H°COOK + NH3 
Cl;CCHO + NaOH? > 
Cl;CH’ + HCOONa 
TRITIUM DETECTION 
The current “‘rediscovery”’ of tritium 
as a tracer arises in part from develop- 
ment of new sensitive methods for 
detecting it. The beta particles have 
a maximum energy of only 18 kev, 
which means that few of them will 
penetrate even the thinnest window. 
Even more important, almost any 
tritium sample, no matter how thin, 
will absorb most of its own beta energy. 
Two counting techniques conquer 
the difficulty effectively—liquid scintil- 
lation and gas counting. 
The more powerful of the two is 
liquid scintillation counting, which has 
been widely developed in the past few 
years (6-12). The sample is dissolved 
or suspended in the scintillating liquid 
or gel. Thus there is a minimum of 
self-absorption and essentially no bar- 
rier between sample and detector. 
Beta energy is converted immediately 
into light, and the light can transverse 
the scintillating substance without the 
absorption experienced by beta rays. 
Such a system will recognize as little 
as 25 wuc of tritium. In a recent ex- 
tension of the liquid-scintillator tech- 
nique, leaf patterns have been squeezed 
187 
