Bioelectronics

Intermolecular electron transfer may play a major role
   in biological regulation, defense, and cancer.

Albert Szent-Gy6rgyi

Many years ago, working with J. A.
McLaughlin, 1 came to the conclusion
that. in animal tissues, cell division may
be controlled by two antagonistic sub-
stances, an inhibitor and a promotor
(I). In plants, growth is known to he
controlled by such antagonists, which
have been isolated and identified (2).
My arduous efforts to isolate the inhibi-
tor of animal tissues failed, and this
made it necessary to look more deeply
into the problem.

Present-day biology is dominated by
the molecular outlook-the view that
living systems are built of isolated small
units, molecules, and that in order to
understand life we only have to know
these molecules, the rest will take care
of itself.

Joseph Weiss discovered in 1942 (3)
that in certain molecular complexes an
electron can go spontaneously from one
molecule (the donor) to another (the
acceptor), a reaction hc called "charge
transfer." Weiss worked with complexes
formed by strongly oxidizing and re-
ducing agents. Later, attention was giv-
en to charge transfer in which the cn-
ergy of light moves electrons from one
molecule to another. This was called a
"weak transfer" to distinguish it from
the "strong" transfer studied by Weiss,
in which the transfer was spontaneous.
R. S. Mulliken cleared up the quantum
mechanics of these reactions (4) and
systematized them. He preferred the
name "DA [donor-acceptor] interac-
tions" to "charge transfer."

Though in several instances DA in-
teractions between biological substances
have been produced in vitro, the idea
of charge transfer found no real place
in biology. Strong charge transfer could
play no role because the presence of
--- -- .--.        ___-
The .~utbm is afliliated with the tnstitute for
Muscle Research of the  Marine Biological
Laboratory, Woods Hole, Massachusetts.

988

strong oxidizing agents is incompatible
with life, and we have no light in our
body to move electrons (except in the
eye and skin). So charge transfer re-
mained, for the hiologist, more or less
a chemical curiosity.

Using the method of electron spin
resonance (5). I could show that even
molecules with low reactivity, which
play a major role as metabolites or
hormones, can give off a whole elec-
tron, forming a free radical; this sug-
gested that charge transfer may be one
of the most common and fundamental
biological reactions. Such considcra-
tions led to the study of the nature of
the various donor and acceptor atomic
groups.

Donor and Acceptor Groups

The cell has a rich source of trans-
ferable electrons in its nitrogen, sulfur,
and oxygen atoms, which all have pairs
of "lone" electrons-electrons which do
not take part in bonding and are thus
available for transfer. Not so with ac-
ceptors. The cell is poor in these. I
could find one acceptor group only,
CO, the carbonyl. This is a "ketoid
acceptor" which, as shown by Mulliken,
can accept in its double hond an addi-
tional electron, acting as a "7i accep-
tor." As an acceptor, CO is very weak.
However, if its acceptor ability is due
to its double bond, then it should be
possible to boost. this ability by insert-
ing into the molecule another double
link in the <u-/3 position. This produces
good acceptors, but no substance of
this composition is known to play a role
in biology.

Another way to extend the double
bonding and therehy increase the accep-
tor ability is to introduce a second CO
on the neighhoring carbon atom, in the

n position. If we also link a hydrogen
atom to this atom, and give the mole-
cule, as HCO, the character of an alde-
hyde, then we also lend the molecule
a greater chemical reactivity, aldehydes
being, on the whole, more reactive than
ketones. So the question was: Could
keto-aldchydes play a major role in
biology as acceptors?

L. G. Egyiid found indication of the
presence of a keto-aldehyde in our
growth-retarding preparations (6). The
simplest n-keto-aldehyde is methylgly-
oxal (pyruvic aldehyde) (Fig. I); this
Iact seemed most exciting (7) because,
as far as we know, all cells contain a
very powerful enzymic system for the
conversion of ,I-keto-aldehydes into the
corresponding unreactive oxyacids-for
converting, for instance, methylglyoxal
into lactic acid. This enzymic system,
called the "glyoxalase," occupied the
attention of several of the most out-
standing biochemists in the first half
of this century, but the interest later
faded out, for no glyoxal derivative
could he found on the main metabolic
pathways, nor could such a substance
he isolated from tissues under normal
conditions. And what is the use of an
enzyme without a substrate?

Should the inhibitor of growth prove
to be a dicarbonyl, like methylglyoxal
or a compound thereof, then I had an
excuse for not having been able to iso-
late it, for the isolation of the expected
trace amounts of such a very reactive
substance would be very difficult indeed.

To investigate this matter, Egyiid syn-
thesized a greater number of different
n-keto-aldehydes and studied their ac-
tion on ceil division in bacteria (8)
and other cells. At a low concentration
they all inhibited cell division reversibly,
in a specific way (9), inhibiting protein
synthesis (10) on the ribosomal level
(II).

All this suggested that cell division
may hc regulated by DA interactions
and that the DA balance may he an
important parameter of cell life.

One of the most active donor groups
iq the sulfhydryl (SH), and SH is known
to have an important function in cell
division. So the regulation may actually
rest on a DA interaction of a kcto-al-
dehyde and SH. Egyiid found that the
inhibition, induced by methylglyoxal,
could be released instantaneously by the
addition of an equivalent quantity of
cysteine or other SH-containing sub-
stances (IZ), a finding which strongly
supported the assumption that the in-

SCIENCE. VOL. 161


Ii
I

H-C-H
I
  c=o
  [
H-C=0

Fig. 1. Structure of methylglyoxal.

hihitory action of the keto-aldehyde was
actually due to its interaction with SH
groups. These groups are known to
play an important role in many biolog-
ical reactions, but the SH groups in-
volved in cell division appeared to he
especially reactive, and to open the
way to a specific inhibition of prolifera-
tion.

Regulatory Systems

If one double bond, introduced in
the N-P position, increases the acceptor
ahility of a molecule, then the introduc-
tion of a more widely conjugated sys-
tem of double bonds must do so even
more strongly. Such a system is found
in aromatic molecules. Accordingly,
two CO groups, in an aromatic molc-
cule, must make a very good acceptor.
Aromatic quinones actually are known
to be strong oxidizing agents. They arc
much too strong to he compatible with
life. If not compatible with life, they
could be used by an organism for kill-
ing,  in self-defense, invading micro-
organisms. This thought led mc back
to one of the earliest pieces of biochem-
ical research I did as a beginner. As is
generally known, many plants discolor

Fig. 2. A banana which, on the previous
day, had been dipped in chloroform for 1
minute.

6 SEPTEMBER 1968

if they are damaged; about half of all
plants do so. If you drop your pear or
apple, the next day you find a brown
patch on it. This reaction is very sensi-
tive; the slightest damage may induce
such a change. Figure 2 shows a ba-
nana which, on the previous day, had
been dipped up to its middle in chloro-
form for 1 minute. The underlying
mechanism of the pigment formation
is as follows. The intact plant contains,
side by side, a polyphenol and a poly-
phenoloxidase, which can oxidize the
phenol into a quinone (Fig. 3). The
two are separated and cannot interact.
This is a most subtle situation; the
slightest damage will disturb it, releas-
ing the oxidase which, then, will oxi-
dize the phenol, eventually producing
the dark color. The biological meaning
of this is simple: the reaction serves as
a defense and has a great survival value.
Suppose a bacterium penetrates the
plant. By the damage it induces it re-
leases the phenolosidase from its hond-
age; the enzyme oxidizes phenols to
quinones, and the quinones kill the hac-
terium. Nature has set, cunningly, a
trap and makes the invading micro-
organisms commit suicide by activating
the cnzymic system. The quinone has
other functions, too. It "tans" the pro-
teins. If the plant is damaged by a cut.
the tanned proteins form a protective
film over the cut, closing the wound.
There is thus a regulatory system in the
plant which functions in such a way
that damage activates it, whereupon it
corrects the damage. There are many
such mechanisms: they seem to repre-
sent one of nature's widely applied
principles. Take. for instance, sunlight.
Ultraviolet radiation damages our skin,
the damage releases the "tyrosinase,"
the tyrosinase produces pigments. and
the pigments protect us against sunlight.
In blood coagulation, damage to blood
and blood vessels releases an enzymic
system which produces fibrin, which
plugs up the damaged vascular system.
If damage is deep, it is desirable that
the damaged cell be eliminated. Such
damage releases cathepsin, a proteolytic
enzyme, which digests and eliminates
the damaged cell.

Living matter has an inherent drive
to proliferate. In monocellular orga-
nisms it may he simply the quantity of
food available which sets limits to
growth, hut, in multicellular organisms.
cell division had to he subjected to strict
regulation in the interest of the whole
organism. A brake had to be put on,

   H                H
Fig. 3. Oxidation of a diphenol to an
o-quinone.

hut put on loosely, for, if I cut myself,
my cells have to start multiplying at
short notice to fill the gap and heal the
wound. If it is methylglyoxal which acts
as a brake, then, in the resting cells,
the glyoxalase must be kept separated
from the methylglyoxal. Cell division
would be induced by liberation of the
glyoxalase. The glyoxalase, by decom-
posing the keto-sldehyde, would release
the brake and start up cell division
which, then, would continue until the
gap was filled-until the wound was
healed-whereupon, with balances re-
stored, the glyoxalase would he bound
again, and the system would return to
its initial resting state. If we suppose
that the damage induced by the cut
liberates the glyoxalase. then the whole
system comes into line with other regu-
latory systems in which a damage in-
duces the changes which lead to its
correction.

A Tentative New Theory of Cancer

One could ask what would happen
if a cell lost its ability to hind its own
glyoxalase? Then it would have to go
on multiplying senselessly and endlessly,
behaving like a cancer cell. As far as
we know. the only differcncc between
a normal ccl1 and a cancer cell is the
fact that the latter divides when no pro-
liferation is needed. All this leads, ten-
tatively, to a new theory of cancer:
a cancer cell ic a cell which has lost
its ability to bind its own glyoxalasc.
Whether this theory is right or wrong
remains to he demonstrated. It recom-
mends itself by its clarity. its simplicity,
and its ability to explain why such a
great variety of noxious influences can
lead to the same end, cancer. The the-
ory also has the earmark of a good
theory: it can he proved or disproved.
What may lend it additional value is
the fact that it suggests various ways of
seeking a therapy for cancer. It is rc-
grettablc that, owing to cuts in the
budget, this research will have to he
discontinued. killing now having prccc-
dencc over healing.

9R9


References and Notes

1. A. Szent-Gytirgyi, Science 149, 34 (1965):
I,,,, oducrio,, to a Submolecular Biology (Aca-
demic Press. New York, 1960). p. 124.
2. 1. van Overbcek, Sci. Amer. 219. 75 (1968).
3. .I. Weiss, J. Chem. Sot. 245, IS (1942).
4. R. S. Mulliken. J. Plrys. Cheer. 56, 801
(1952); .I. Amer. Chew. Sot. 74, 811 (1952);
Rrc. Tram. Chbn. 75, 845 (1956); J. Chim.
Phys. 61, 20 (1964).

5. A. Szent-GyBrgyi, Proc. Nat. Acad. Sci. U.S.
58, 2012 (1967).

6. L. G. Egyiid, ibid. 54. 200 (1965).
7. A. Szent-Gyiirgyi, L. G. Egyiid. J. A. Mc
Laughlin, Scier~ce 155, 539 (1967).
8. L. G. Egriid and A. Szent-Gysrgyi, Proo.
Not. Acad. Sci. U.S. 55, 3R8 (1966); L. G.
E&d, Currents Mod. Biol. 1, 14 (196,).
9. The cancerostatic action of substituted keto-
aldehydes has been noted by various authors
[see F. A. French and B. L. Freelander,
Cancer Ha-. 18, 172 (1958)I. The use of SH
gro"~S for attack on cancer has been advo-
cated by F. E. Knock [see F. E. Knock,

Anllcarxw Agents (Thomas, Springfield, III.,
1967)l.

10. L. G. Egyiid and A. Szent-GyGrgyi, Proc.
Nat. Acad. SC!. U.S. 56, 201 (1966); H.
Otwka and L. G. Egyiid, Cnr~cer Res. 27,
  1498 11967).

11. H. Otsuka and L. G. E&d, Currrr~ts Mod.
  Aiol. 2, 106 (1968).

12. L. G. Egyiid, ibid.. in press.
13. The experimental work discussed in this
  article was supported by grant GMlU383
  from the National Institutes of Health.