Multienzyme Complexes

BY DR. U. HENNING[Il
INSTITUT FUR GENETIK, UNIVERSITAT KoLN (GERMANY)

In 1948 D. E. Green[21 termed the mitochondria1 en- indolyl glycerophosphate

zymes the cyciophorase complex, in order, as he later indole + D-glyceraldehyde 3-phosphate (b)
commented[31 “ . . . to signify that the complex was a
unit of enzyme action and not a chance mixture of and the condensation
several hundred enzymes. . . . this complex was visu-
alized as aa organized mosaic of enzymes in which
each of the large number of component enzymes was The extensive investigations by Yanojiky and co-
uniquely located to permit efficient implementation of workers, which made this enzyme a component of one
consecutive reaction sequences”. of the best-known gene enzyme systems, also showed
These words contain the concept of the multienzyme that the synthetase consists of two sub-units, the proteins
complex: ordered association (not involving peptide A and B. The isolated crystallizable protein A (mole-
linkages) of various enzymes that catalyse successive cular weight about 30000 [81) consists of only one poly-
steps in a reaction sequence. The present paper deals with peptide chain c([91, while the protein B (A4= 117000)
recent findings regarding the properties of such com- consists of two chains ( p ~ [lo].
)
plexes, and the question whether the concept represents The physiological reaction (a) can take place only when
a biological principle is discussed. the proteins A and B are in contact with each other;
Predominantly protein-chemical aspects of th e association of indole does not occur as a free intermediate [61111.
enzymes were recently described by Sund a n d Weber [41, a n d
the “macromolecular organization of enzywe systems” by Isolated protein A is inactive in reactions (a) and fc),
Reed a n d Cox[sl. but can bring about reaction (b) (not requiring pyrid-
oxal phosphate, as expected) with about 1 % of the
activity exhibited in the presence of B [121. On the other
I. Multienzyme Complexes in a Narrower Sense hand, isolated B containing pyridoxal phosphate (2
moles of coenzyme per mole of protein) does not
1. Tryptophan Synthetase catalyse reactions (a) and (b), but catalyses the con-
densation (c) with 3 to 10 % of the activity exhibited in
A clear and relatively simple example is the tryptophan the presence of A [121. It was later found that protein B
synthetase from Escherichia coli. In the presence of can also act as an L-serine deaminase [I31
pyridoxal phosphate, the enzyme catalyses the last step
L-serine z2 pyruvate+ NH3 (d)
[Eq. (a)] in the reaction chain leading to the biosynthesis
of tryptophan : Addition of A causes almost complete inhibition of B
CHOH-CHOH-CH20@ + 7H2OH in reaction (d).
H2N- C H- C O2H There is strong evidence that the catalytically active
H centers of the respective complementary proteins are
I nd o l y l g l y c e r o p h o s p h a t e L-Serine not involved in the activation of A in reaction (b) and
(a) in that of B in reaction (c). Protein B which has lost its
HXC+O enzymatic activity as a result of mutation canactivate A
I
rt ~ C H Z - C H N H Z - C O +~ HHCOH in reaction (b) in the same way as unaltered protein B,
H 6HzO@ while the reverse is true of A which has been inactivated
by mutation. Moreover, the activity of B in reaction (c)
L-Tryptophan D- G l y c e r - is increased by a factor of about 20 in the presence of a
aldehyde 3 -phosphate
high concentration of NH4+ ions 1141. Finally, Hatunuka
It can also [6,71 bring about the cleavage [8] U. Henning, D. R . Helinski, F. C . Chao, and C . Yanofsky,
J. biol. Chemistry 237, 1523 (1962).
[l] Present address: Max-Planck-Institut fur Biologie, Tiibingen [9] B. C . Carlton and C . Yanofsky, J. biol. Chemistry 237, 1531
(Germany). (1962).
[2] D. E. Green, W . F. Loornis, and V . H . Auerbach, J. biol. [lo] D. A. Wilson and I . P. Crawford, Bacteriol. Proc. USA
Chemistry 172, 389 (1948). 1964, 92.
131 D. E. Green, Harvey Lectures 52, 177 (1958). [ l l ] C . Yanofsky and M. Rachmeler, Biochim. biophysica Acta
[4] H. Sund and K . Weber, Angew. Chem. 78, 217 (1966); An- 28, 640 (1958).
gew. Chem. internat. Edit. 5, 231 (1966). [12] C . Yanofsky, Bacteriol. Rev. 24, 221 (1960).
[5] L. J. Reed and D . J. Cox, Annu. Rev. Biochem. 35, in press. [13] I. P. Crawford and J. Ito, Proc. nat. Acad. Sci. USA 51, 390
[6] I. P . Crawford and C . Yanofsky, Proc. nat. Acad. Sci. USA (1964).
44, 1161 (1958). [14] M. Hatanaka, E. A . White, K . Horibata, and I. P . Crawford,
[7] I. P . Crawford, Biochim. biophysica Acta 45, 405 (1960). Arch. Biochem. Biophysics 97, 596 (1962).

Angew. Chem. internat. Edit. / Vol. 5(1966) 1 No. 9 785

 and Cruwford[lsl have isolated a mutant of E. coli It is now known[5.21] that this overall reaction (e)
producing a defective protein B, which by itself is proceeds viu the steps (f) to (kl.
inactive in reaction (c); in the presence of A, however,
the activities of all three tryptophan synthetase reactions R-CO-COzH + TPP-El *
are observed.
R-CHOH-TPP-El + C02 (f )
The two proteins are only loosely associated[161. L- __
Serine and pyridoxal phosphate are essential to the
complex formation between A and B. Centrifugation R-CHOH-TPP-El + *
of a mixture of A and B in a sucrose density gradient uE 2
leads to practically complete separation of the com-
ponents A (2.7 S) and B (5.1 S), while the associated
A-B complex (6.4 S) appears after incubation of the
mixture with r-serine and coenzyme [161. The completely
associated A-B complex, tryptophan synthetase, con- SH S-Co-R + CoA-SH # SH SH + R-CO-SCoA (h)
tains two moles of A to one mole of B, and its probable u E2 u E 2
polypeptide-chain formula is therefore c(2p2 [16,*71.
The enzymatic activities of the components change, by SH SH + FAD-E3,,. * s-s + FAD-E3,,d. (i)
several orders of magnitude in some cases, on complex u E2 W E2
formation. The activation by NH4+ and the behavior
of the various proteins which have become inactive as a FAD-E3,,d. + NAD’ # FAD-E30x. + NADH + H+ (k)
result of mutation provide very strong evidence that
complex formation per se is responsible for the change The best-known case is that of the pyruvate dehydro-
in enzymatic activity. Association presumably involves genase of E. coli, which has been isolated as a homo-
changes in the conformation of the proteins, for which geneous protein ( M = 4.8~106)[221. Electron micro-
the active site of the complementary component is not graphsr231 of the complex show (Fig. 1) a polyhedral
directly responsible. structure with a diameter of 300 to 350 A.
It is interesting to compare the tryptophan synthetase of
E. coli (or other enterobacteria) with that of Neurospora.
The neurospora enzyme[l*,191(A4 = 135000) so far has
not been dissociated into enzymatically active suo-units.
On reduction with mercaptoethanol in 5 M guanidinium
chloride, the protein dissociates into four sub-units,
possibly made up of two identical pairs [191. This result
led D.M. Bonner [201 to assume that the two tryptophan
synthetases represent two different stages of evolution.
It is conceivable that the tryptophan synthetase reaction
initially proceeded in separate steps, with indole as a
free intermediate. With the development of the asso-
ciation of the two proteins, indole is no longer liberated; Fig. la). Pyruvate dehydrogenase complex from E. coli. Magnification
the binding of the two proteins may have become factor 260000. b) Lipoate reductase transacetylase component of the
“tighter” during the furrher evolution. enzyme complex, same magnification. The tetrad structure of the com-
ponent is visible in the center of the complex (la). In these projections,
the diameter of the complex is about 300 A and that of the components
is about 120 A. Both samples were obtained by negative staining with
phosphotungstic acid.
2. Pyruvate and a-Ketoglutarate Dehydrogenases

In animal tissues, as well as in many fungi and bacteria, The complex can be dissociated, under mild conditions,
the principal route for the degradation of the two a-keto into three proteins 1241, to which the activities of reac-
acids is by oxidative decarboxylation: tions (f) to (k) could be assigned with the aid of model
reactions : the thiamine pyrophosphate(TPP)-dependent
R-COCOzH + COASH+ N A D + + pyruvate decarboxylase (El), the lipoate reductase
R-COSCoA + COz + N A D H + H + (e)
transacetylase (E2) containing lipoic acid, and the
1151 M . Hatanaka and 1. P. Crawford, Bacteriol. Proc. USA flavoprotein dihydrolipoate dehydrogenase (E 3). The
1965, 88.
[16] T. Greighton and C. Yanofsky, J. biol. Chemistry 241, 980
proteins re-associate spontaneously in vitro 1241;E l does
(1966).
1171 D . Wilson and I. P. Crawford, J. biol. Chemistry 240, 4018 [21] I. C . Gunsalus in W. D . McEIroy and B. Glass: The Mecha-
(1965). nism of Enzyme Action. Johns Hopkins Press, Baltimore 1964,
[18] W . C . Mohler and S. R. Suskind, Biochim. biophysica Acta p. 545.
43, 288 (1960). [22] M . Koike, L. J. Reed, and W . R. Carroll, J. biol. Chemistry
[19] M . Carsiotis, E. Apella, P. Provost, J. Germershausen, and 235, 1924 (1960).
S. R . Suskind, Biochem. biophysic. Res. Commun. 18, 877(1965). 1231 H . Fernandez-Moran, L. J. Reed, M . Koike, and C . R.
1201 D . M . Bonner, J. A . DeMoss, and S. E. Mills in V. Bryson W i l h s , Science (Washington) 145, 930 (1964).
and H . J. Vogel: Evolving Genes and Proteins. Academic Press, [24] M . Koike, L. J. Reed, and W. R . Carroll, J. biol. Chemistry
New York 1965, p. 305. 238, 30 (63).

786 Angew. Chem. internat. Edit. / Vol. 5 (1966) 1 No. 9

not associate with E3, but E 2 binds both E l and E3. A similar situation is found with the a-ketoglutarate
One molecule of the complex contains about twelve dehydrogenase of E. coli ( M = 2.4~106)[2*,*41.This
molecules of the decarboxylase ( M == 1.83~105),one also could be reversibly dissociated into cr-ketoglutarate
molecule of E 2 ( M = 1.6x106), and about six molecules decarboxylase, lipoate reductase transsuccinylase, and
of the flavoprotein ( M = 1.12~105)[51.The lipoate dihydrolipoate dehydrogenase. The flavoproteins of the
reductase transacetylase is probably centrally situated two x-ketoacid dehydrogenases are functionally inter-
(cf. Fig. l), and surrounded by the two other enzymes. changeable, that of the pyruvate dehydrogenase being
The lipoate reductase transacetylase has also been bound by the lipoate reductase transsuccinylase and

units having M -
reversibly dissociated; it dissociates at pH 2.6 into sub-
7x104[s]. The enzyme contains one
mole of lipoic acid per 3 . 5 ~ 1 0 4g of protein, and may
vice versa; no “hybrid” binding of the decarboxylases
takes place [27J.
It is possible that the two flavoproteins are identical[5,z*l. If
this were the case, it would raise a n interesting problem in
therefore consist, not of about 24, but of about 48 poly-
connection with the regulation of the synthesis of the di-
peptide chains having M = 3.5~104.The composition of hydrolipoate dehydrogenasL, since the syntheses of the
the decarboxylase is not yet known; that of the flavo- pyruvate and cc-ketoglutarate dehydrogenases are not
protein will be discussed below. The lipoic acid is regulated by the same mechanism [*91.
covalently bound to the E-aniino group of a lysine It is not yet known whether the structure of the enzyme
residuer251. The dithiolan ring of the coenzyme must complex in the cell is the same as that in which it is
react with (a) the hydroxyethyl TPP of the decarboxy- obtained after purification. However, since the inter-
lase, (b) the flavin-adenine dinucleotide (FAD) of the mediates of the overall reaction (e) remain bound to the
dihydrolipoate dehydrogenase, and (c) the coenzyme A. enzyme, some form of association of the enzymes is
Since the intermediates of the overall reaction remain necessary. It is conceivable that the large complex
bound to the enzyme, there appear to be steric diffi- isolated is an artifact resulting from the aggregation of
culties opposing the interaction of one and the same smaller complexes that were bound to other structures
lipoic acid residue with all the prosthetic groups men- in the cell. There are several arguments against this
tioned. The lipoic acid:FAD ratio in the complex is view; e.g., mutants in which the decarboxylase com-
about 4 : l . The lipoic acid linkage forms a “flexible ponent has lost its enzymatic activity were found to
arm” (about 14 8, long), and the coenzyme may be able contain a “partial” complex consisting of the lipoate
to rotate about the x-C atom of the lysine residue in the reductase transacetylase and the flavoprotein [301. This
enzyme complex 124,261. It was shown that all the lipoic partial complex appears to be the same as that obtained
acid residues can takepart in the oxidation:ofpyruvate. It by the removal of the decarboxylase from the pyruvate
was therefore suggested [241 that different lipoic acids dehydrogenase complex in vitro.
could participate in the various reaction steps (Fig. 2), Concerning the cr-keto acid dehydrogenases from animal
which indeed would be a remarkable type of electron cells in which the enzymes are components of mito-
transport in a protein molecule. chondria, the reader is referred to the review by Reed
and Cox 151.

3. Fatty Acid Synthetase

A similar organization of structure and function is found

in the synthesis of unbranched saturated fatty acids.
Investigations-by Lynen et al. [3*1 have shown that the
overall reaction (I)

CHI CO-SCoA +
n H02C-CH2-CO-SCoAS 2n NADPH +
2rr Hi- + CH~(-CH~-CH~-)CO-SCOAf n COz
n CoASH 2n NADP + +
n H2O ( n = 7 for palmityl-CoA) (I)

in yeast is catalysed by a homogeneous protein ( M =

2.3 x106), the principal products being palmityl and
stearyl CoA. The net reaction takes place on the fatty
n
rA53721 CoASH Acetyl-CoA acid synthetase in the following steps. The acetyl
Fig. 2. Scheme of the hypothetical interaction between the coenzymes
transfer [Eq. (m)] is the initiation reaction.
of the pyruvate dehydrogenase [24]. The lipoic acid residue pivots about
the a-C atom of a lysine residue. HS, HS\
CH3-CO-SCoA + E # E + CoASH (m)
H i CH3 - C O - s’
[27] B. B. Mukherjee, J . Matthews, D . L. Horney, and L. J. Reed,
J. biol. Chemistry 240, PC 2268 (1965).
[28] U . Henning, G. Dennert, R . Hertel, and S . W. Sliipp, Cold
Spring Harbor Symp. Quant. Biol. 31, in press.
[25] H . Nawa, W. T . Brady, M . Koike, and L. J . Reed, J. Amer. [29] U . Henning and G . Deppe, unpublished work.
chem. SOC.82, 896 (1960). [30] U . Henning, C. Herz, and K . Szolyvuy, Z. Vererbungslehre
[261 R . M . Bock and R . S. Criddle, quoted i n D . E. Green, C o n - Y5, 236 (1964).
parat. Biochem. Physiol. 4 , 81 (1962). [31] F. L.~/7en,Angew. Chen. 77, 929 (1965).

A n g e w . C h e m . internat. E d i t . Vol. S(1966) No. 9 787

The five subsequent reactions lead to lengthening of the can be separated into enzyme fractions with different
chain; they begin with the malonyl transfer Eq, (n)]. functions [341.

Malonyl transfer COzH

y02H Condensation
CHz- CO- ?* C H3 -CO-CHs-CO-S\'"
E *
CH3-CO-s' H? cu2
First reduction
CH3-CO-CH2-CO-S: CH~H,-CHOH-CH,-CO*S:
E f NADPH 4. H+ ;P E f NADP' (p)
Hs' Hs'

Second reduction

'
E + NADP' [r)
Hs' HS

The SH group which accepts the malonyl residue is then Details of bacterial and animaI systems cannot be discussed
liberated by the butyryt transfer Eq. (s)]. here. The electron micrographs, and the fact that the syn-
thetase from yeast can be dissociated into inactive sub-units
only under fairly drastic conditions strongly indicate that the
isolated enzyme is not an artifact produced during purifi-
cation.

Finally, when the chain Iength becomes c16 to CIS,the

stearyl or palmityl CoA is formed.

The existence of the functionally different SH groups

(HS*-, HS-) has been demonstrated[3*1. As with the
other enzyme complexes, all intermediates remain (Pteparafion a5 in Fig. I ,
Fig. 3. Fatty acid synthotasc from yeast 1331%
magnification factor 250000).
enzyme-bound.
The seven reaction steps (m>to (s) apparently correspond 4, Other Multienzyme Complexes
to seven sub-units in the synthetase; however, disso-
ciation of the enzyme into enzymatically active sub-
units has not been achieved. An electron micrograph [331 The examples described undoubtedly are not special
of the synthetase (Fig. 3) shows its ordered structure in cases of little general importance. There is evidence that
particles having a diameter of 200 to 250 A. As in the a number of other reaction sequences are catalysed by a
case of tryptophan synthetase, the enzyme system from functional complex. This is true of the four enzymes
E. calf which carries out the fatty acid synthesis easily that synthesizevaline from pyruvate and isoleucine from
pyruvate and cr-ketobutyrate in Neimspora [35~361, and
[3t] F. Lynen in M. Seia: Xew Perspzctives in Biology. B.B.A also of the synthesis of isoleucine in Sahonella [371.
Library Vol. 4, Elsevier Publishing Comp., Amsterdam-London- - . ~
New York 1964, p. 132. 1351 K. Xirirani, S. Narise, A. Bergquist, and R . P. Wagner, Bio-
[33] A. Hagen and P. H.Hofickneider in M. Titlbaclt: Electron chim. biophysica Acta IOU, 432 (1965).
Microscopy 1964 (Proc. 3rd European Regional Conference). I361 R . P. Wagner, A. Bergquisr, and T. Barbee, Biochim. bio-
Publishing House Czechoslovak Acad. Sci., Prague 1964, physics Acta 100,444 (1965).
Vol. 13, p. 69. [371 C. S. Cronenweit and R. P. Wagner, Proc. nat. Acad. Sci,
{341 P. R. Vagefos, Annu. Rev. Biochem. 33, 139 (1964). USA 54, 1643 (1965).

788 Angew. Cliem. internat. Edit. Vol. 5 (1966) 1 No. 9

Three enzymes involved in the biosynthesis of trypto- myoglobin c41-431, which Frieden “1 aptly described as
phan appear to act as a functional unit in Neurosporu [381, “perfect” enzymes, since they permit the study of the
as do three enzymes involved in the biosynthesis of enzyme-substrate interaction without the formation of
histidine in Neurosporur391. In the last case, since the interfering products. The four heme groups, and hence
three enzymes in question correspond, not to conse- the four active sites, corresponding to the four
cutive reaction steps, but to the second, third, and tenth polypeptide chains of hemoglobin, are situated in
steps, it is reasonable to suspect that all relevant en- “pockets” in each of the two c( and p chains on the
zymes in vivo are associated in a functional unit. surface of the molecule, without being in contact with
The features common to all these examples are readily one another. The tertiary structure (and also the position
recognized, and have been characterized by Lynen [321 2f the heme) of myoglobin, which consists of only one
as “strict compartmentalization in the smallest possible polypeptide chain, is extremely similar to the tertiary
space”. If enzymes that operate in sequence are also structure of each of the hemoglobin chains, and in this
physically adjacent, the activity of the overall reaction sense myoglobin may be regarded as a naturally-
sequence presumably may be greater than if these en- occurring sub-unit of hemoglobin. The activities of all
zymes were separated. In the latter case the action would the polypeptide chains are the same. In contrast to
be optimal only when the concentrations of diffusing myoglobin, however, the binding of 0 2 by hemoglobin
intermediates reach certain values. Moreover, reaction leads to a heme-heme interaction, i.e. a co-operation of
chains in which the intermediates are bound to the the binding sites C44-461. The rate of reaction of a heme
enzyme avoid competition between different enzymes with 0 2 depends on the state of loading of the other
for the same intermediate. hemes with 0 2 , and increases with the amount of 0 2
The specific activities of the fatty acid synthetase from yeast already bound. Thus the association of different sub-
and of the pyruvate dehydrogenase are high; one synthetase units leads (in a wider sense) to an increase in efficiency.
complex can catalyse the conversion of 2 . 4 ~lo3 molecules of The question arises whether the fact that a protein
malonyl CoA per minute into fatty acid[32J,and one pyruvate
molecule consists of several su b-units, i.e. the presence
dehydrogenase complex at 30°C can produce at least
1 . 2 5 ~ 1 0 4molecules of acetyl CoA per minute from pyruvate. of a quarternary structure, is enougn to constitute a
multienzyme complex. It-may be assumed that this is
One might conclude that the concept of the multi-
enzyme complex represents an example of the biological
not the case, but that the cell also makes use of the
association to produce an active site. This appears to be
principle of “increasing the efficiency”. However, this
the situation with the dihydrolipoate dehydrogenase
view probably is still too narrow. In higher organisms
[cf.reactions (i) and (k)], as M a s s e y ’ s investigations 147,483
at least, the large enzyme complexes discussed are
have shown in the case of the pig’s heart enzyme.
localized in the mitochondria. The same is true of the
above mentioned enzymes of the biosyntheses of valine The enzyme consists of two probably identical polypeptide
and isoleucine in Neurospora [35,361; the isoleucine sys- chains, each with M = 50000; 2 moles of FAD and 4 moles
of cystine units (as well as 1 2 cysteine units) are present per
tem in Salnzonellu [371 occurs in the membrane fraction. mole of enzyme. The intermediate is the reduced enzyme
It appears reasonable to combine enzyme systems [reaction (i)], in which, in contrast to several other de-
physically when they are bound to large cell structures, hydrogenases containing FAD, the flavin coenzyme takes
even if they are not also components or ‘‘supply’’ en- up only one electron. The difference spectrum of the resulting
zymes of further processes (e.g. the respiratory chain). reduced enzyme is similar to that of the semiquinoid form of
FMN which occurs as a charge-transfer complex in an
Another interesting speculative aspect L39al is that the equilibrium mixture of FMN and FMMH2. This form
structure-bound enzymes could greatly increase their (“red” enzyme) has been obtained by the action of NADH
surface if the substrate can be retained by the structure or dihydrolipoic acid on stoichiometric quantities of the
(e.g. membrane) and can migrate along the latter to dehydrogenase. However, spectrophotometric titration of
this last reaction with either of the substrates shows that the
the enzyme.
red form of the protein corresponds to the uptake of two
electrons per FAD. This discrepancy was resolved when a
disulfide was discovered as a second prosthetic group of the
11. Extension of the Concept: Homofunctional and enzyme, and it was found by amperometric titration that
Heterofunctional Enzyme Complexes the addition of NADH leads to the formation of a dithiol.
When the NAD formed in a mixture of NADH and enzyme
is removed with an NADase (which does not attack NADH),
With regard to the enzyme protein, the definition of the enzyme is completely reduced and becomes colorless;
the multienzyme complex as described in the introduc- four electrons are taken up per FAD. The colorless enzyme
tion is really teleological, and therefore is too specialized. can react with NAD to give again the red form.
This can be seen particularly well with hemoglobin and The dihydrolipoicdehydrogenase in the oxidized yellow state
is unusually stable towards urea (in 6.5 M urea at 0 “C only
[38] J. A . De Moss and J . Wegman, Proc. nat. Acad. Sci. USA
54, 241 (1965). [44] R. Benesch and R. E. Benesch, J. molecular Biol. 6, 489
(1963).
[39] A . Ahmed, M. E. Case, and N . H. Giles, Brookhaven Symp.
Biol. 17, 53 (64). [45] H. Muirhead and M . F. Perutz, Cold Spring Harbor Symp.
Quant. Biol. 28, 451 (1963).
[39aJ M. Delbriick, personal communication.
[46] R . E. Benesch, R . Benesch, and G . McDuf, Proc. nat. Acad.
[40] C. Frieden, Brookhaven Symp. Biol. 17, 98 (1964). Sci. USA 54, 535 (1965).
[41] J. C. Kendrew, Angew. Chem. 75, 595 (1963).
[47] V . Massey, T. Hofmann, and G . Palmer, J. biol. Chemistry
[42] M . F. Ptrutz, Angew. Chem. 75,589 (1963). 237, 3820 (1962).
[43] G . Braunitrer, K . Hilse, W. Rudloff, and U . Hilschmann, [481 V . Massey in P . D . Boyer, H. L . Lardy, and K . Myrback:
Advances Protein Chern. 19, 1 (1964). Thc Enzymes. Academic Prsss, New York, 1963, Vol. 7, p. 275.

Angew. Chem. internat. Edit. 1 Vol. 5 (1966) No. 9 789

about 30 % of the enzyme activity is lost after one month); it is correct ts classify such enzymes separately, since
the red form of the enzyme also retains its activity in urea. the transition to proteins of the tryptophan synthetase
In the presence of excess NADH or dihydrolipoic acid in
6.5 M urea, the spectrum is initially that of the semi-reduced
type may be continuous. (Incidentally, it would not be
enzyme, but this soon gives way t o that of the fully reduced easy to give a good account for the biological evolution
protein. The completely reduced enzyme is unstable in the of enzymes with “mixed”centers of activity, as dihydroli-
urea solution. The FADH2 (or FAD after the admission of poate dehydrogenase).
air) becomes separated from the protein and dialysable; the
protein becomes insoluble when the urea concentration is From the examples described, it appears that a protein
decreased to less than 2 M, and the molecular weight of the may oe regarded with certainty as a multienzyme com-
resulting reduced dehydrogenase falls to half that of the plex if the centers of activity for various steps in an
native protein. overall reaction chain are situated on different sub-
units. As far as the protein is concerned, however, there
is no fundamental difference between such complexes
and enzymes in which similar active centers are also
present in their sub-units and are not formed only by
the combination of the sub-units in such a manner
that two or more sub-units contribute in forming an
active site. If a classification is desired, it appears more
reasonable to distinguish between homofunctional and
jI sI s j
I
heterofunctional multienzyme complexes, depending on
H02C whether the sub-units and active centers are identical
II II ALa-NH2
Em S- S or different.
Fig. 4. Scheme of the dihydrolipoate dehydrogenase [471. Model
The so-called allosteric effect [56,571, for which asso-
representation of the steps of the enzyme reduction as described in the ciation is evidently a prerequisite, fits in easily with the
text. “heterofunction” on association. An example is the
heme-heme interaction in hemoglobin described earlier.
Figure 4 shows the model derived by Massey from these Another extreme case was reported in the aspartate
and other data. Both polypeptide chains appear to be transcarbamylase ( A X ) of E. coli [ S f i I .
involved, with the disulfide bridges, in the catalysis. The enzyme initiates the biosynthesis of cytidine triphosphate
The active sites literally are situated between the poly- +
(CTP) with the reaction: carbamyl phosphate aspartate +
peptide chains, since at least cystine residues of both carbamyl aspartate. It is regulated by feed-back, its activity
chains act as components of one active site. However, being inhibited by CTP. The ATC ( M M 3x 105) consists of
enzymatically inactive regulatory sub-units (A4w 3x 104) and
it is not known whether other amino acid residues
enzymatically active sub-units ( M w 9x104); only the first
belonging to one active site originate from both chains. type binds CTP. The enzymatically active sub-unit, when
The example described is not intended to show impli- dissociated from the regulating sub-unit, is not inhibited by
CTP.
citly that enzymes of this type must necessarily consist
It may be concluded that multienzyme complexes as
of identical polypeptide chains. An association of
defined initially do not occupy a special position in
different polypeptide chains can gi ,e almost the same
enzyme structures. It remains to be seen whether the
catalytic activity, as is well known from various forms
concept of homofunctional and heterofunctional com-
of lactic dehydrogenase 149,501. These can consist either
plexes provides a better insight, since the organization
of four identical chains (the so-called isozymes I and V,
of enzymes is more elaborate than has been described
consisting of chains H and M) or mixtures of integral
here. We have not considered systems which act at the
molar quantities of chains H and M (isozymes TI, Ill,
level of large cell structures, e.g. the vectorial enzymes
and 1V: 1H2M, 2H2M, lH3M). Holbrookr511 recently
of the transport system on membranes or the mechano-
reported results that led him to conclude that the active
chemical systems of contractile or other mobile struc-
sites in this case also are composed of parts of adjacent
tures.
sub-units. A similar possibility exists for aldolase, which
has only one active site, but consists of at least three The author is grateful to Prof: L. J. Reed and Dr. R. M .
(or six) polypeptide chains r52-551. Oliver (Fig. I) and to Prof. F. Lynen (Fig. 2) for the use
However, the direct participation of several sub-units of illustrations. The author’s own investigations were
of an enzyme molecule in one active site is hypothetical. supported by the Deutsche Forschirngsgemeinschajt.
If this hypothesis is true, it remains to be seen whether Received: June Ist, 1966 [A 537 I E ]
German version: Angew. Chem. 78, 865 (1966)
Translated by Express Translation Service, London
[49] N . 0. Kaplan, Brookhaven Symp. Biol. 17, 131 (1964).
[50] T . Wieland and G. Pfleiderer, Angew. Chem. 74, 261 (1962); [55] E. Grazi, T . Cheng, and B. L. Horecker, Biochem. biophysic.
Angew. Chem. internat. Edit. 1, 169 (1962). Res. Commun. 7, 250 (1962).
[511 J. J. Holbrock, Biochem. Z. 344, 141 (1966). [56] J. Monod, J . P. Changeux, and F. Jacob, J. molecular. Biol.
[52] L. F. Hass, Biochemistry 3, 535 (1964). 6, 306 (1963).
[53] J. A. Winstead and F. Wold, J. biol. Chemistry 239, 4212 [57] J. Monod, J . Wyrnan, and J. P. Changeux, J. molecular.
(1964). Biol. 12, 88 (1965).
[54] E. W. Westhead, L . Butler, and P . D. Boyer, Biochemistry 2, [58] J. C. Gerharf, Brookhaven Symp. Quant. Biol. 17, 222
927 (1963). (1964).

790 Angew. Chem. internat. Edit. 1 Vol. 5 (1966) No. 9