THE JOURNAL OF Bm~oorcar. CHED~TRY
Vol. 241, No. 4, Isaue of February 25, 1966
Printed in U.S.A.
Association of the a and & Subunits of the Tryptophan
Synthetase of Escherichia coZi*
(Received for publication, September 20, 1965)
THOMAS E. CREIGHTON~ AND CHARLES YANOFSKY
From the Department of Biological Sciences,Stanford University, Stanford, California 9.@05
SUMMARY composed of two nonidentical and readily separable protein
subunits, formerly designated as the A and B proteins (10).
The association of the (Y and & subunits of tryptophan
synthetase has been studied by means of sucrose gradient The separated A protein has been shown to be a single poly-
centrifugation, Sephadex gel filtration, and enzymatic ac- peptide chain (ll), hereafter called the (Y subunit, while the
tivity measurements. The results indicate that the fully separated B protein is believed to be composed of two identical
associated enzyme (cr&) has a sedimentation coefficient of polypeptide chains (12) and will hereafter be designated the
Downloaded from www.jbc.org by guest, on October 22, 2011
6.4s and is in relatively rapid equilibrium with the free sub- /12 subunit. Fully associated tryptophan synthetase consists
units. Mass law and kinetic considerations are consistent of one pZ and two (IL subunits, LX&* (13). The tryptophan syn-
with the reversible binding of individual (Y subunits to two thetase complex catalyzes the following reactions (10, 14)
identical and independent sites on the /z& subunit, with each pyridoxal-P
combined site contributing the same enzymatic activity. Indole + n-serine + n-tryptophan (1)
Pyridoxal phosphate and serine together markedly increase
the association of the two subunits. Prior incubation of the Indoleglycerol-P * indole + glyceraldehyde-3-P (2)
two subunits with serine and pyridoxal-P apparently increases
their association when they are sedimented in sucrose gradi- pyridoxal-P
Indoleglycerol-P + n-serine ______f
ents in the absence of serine and pyridoxal-P. Apparent (3)
association constants for the subunits have been measured L-tryptophan + glyceraldehyde-3-P
enzymatically under various conditions and found to range
The physiologically significant reaction is probably Reaction 3,
from 4 X lo6 to 2.6 X IO9 ~-l. Under the two extreme
conditions of afhnity of the cz and /3z subunits, apparent rate which is believed to be a distinct reaction, not merely the sum of
the first two (10, 15). A unique property of the tryptophan
constants for association have been estimated to be 2 X 104
and 6 X lo5 se& 116-l. Likewise, the dissociation rate synthetase complex is that each subunit alone has trace catalytic
constants determined were 4.8 X 10e3 and 1.8 X 1O-4 se&. activity in one of the reactions. The a! subunit can catalyze
Reaction 2, and the & subunit can carry out Reaction 1. How-
Mutations abolishing the enzymatic activity of the /3*
subunit can selectively alter the stimulatory effect of pyri- ever, for maximum activity in Reactions 1 or 2 under normal
doxal-P and serine on the ability of mutant pZ subunit to conditions, and for any activity in Reaction 3, the two subunits
combine with normal cr subunits, without affecting its basal must be in physical contact (10). Thus in this case association
aEnity for the Q! subunits. of the two subunits is directly responsible for enzyme activity.
Enzymatic studies with mutant proteins have indicated that
mutant cz subunits will activate the normal pZ subunit in Reaction
1, and most mutant pZ subunits will activate the normal LYsub-
unit in Reaction 2. However, enzyme complexes formed either
with a mutant (Y subunit or a mutant /3z subunit and the normal
In recent years our understanding of the nature and signifi- second subunit will not catalyze Reaction 3 (16). Apparently
cance of the subunit structure of proteins has increased con- the normal subunits must interact to form an effective surface
siderably. Numerous examples of proteins composed of sub- for the catalysis of the third, physiologically essential, reaction
units are now known (l), and the physiological significance of (17). In view of the unusual properties of the tryptophan syn-
their quaternary structure is increasingly becoming evident thetase complex, the present study of factors affecting the as-
(2-9). sociation of the two subunits was undertaken.
The tryptophan synthetase (n-serine hydro-lyase (adding
indole), EC 4.2.1.20) of Escherichiacoli has been shown to be EXPERIMENTAL PROCEDURE
* This investigation was supported by grants from the United Organimzs-The mutant strains used in this study were iso-
States Public Health Service and the National Science Founda- lated by penicillin selection following ultraviolet irradiation of
tion.
t Predoctoral trainee of the United States Public Health 1 M. Goldberg, T. E. Creighton, R. L. Baldwin, and C. Yanof-
Service. sky, manuscript in preparation.
Issue of February 25, 1966 T. E. Creightm and C. Yanofsky 981
the K-12 strain of E. coli. All the strains employed form high mixtures also contained 0.18 M NaCl. The spectrophotometric
levels of the enzymes of the tryptophan pathway when grown assay for the conversion of InGP to indole or tryptophan was
on limiting levels of indole or tryptophan. Mutant T-3 is adapted from the procedure described by Crawford (14). The
blocked in the synthesis of anthranilic acid and forms high l.O-ml reaction mixture (InGP to tryptophan) contained tryp-
levels of the wild-type a! and pz subunits (18). Mutant B-8 tophan synthetase, 0.1 mM InGP, 60 mM nn-serine, 0.04 mM
forms high levels of the wild-type a! subunit and traces of an pyridoxal-P, 180 mM NaCl, 12 mM sodium arsenate, 1.0 IIIM
altered @Z subunit (19). Mutant A-2 forms high levels of the DPN, 0.1 M Tris-HCl buffer, pH 7.8, and excess crystalline
wild-type pz subunit, but no detectable (Y subunit (16). Mutant glyceraldehyde 3-phosphate dehydrogenase. For the con-
T-3 has been used as a source of the intact tryptophan synthetase version of InGP to indole, the serine, pyridoxal-P, and NaCl
complex, while mutants B-8 and A-2 have been used as sources were omitted. DPN reduction was followed spectrophotometri-
of the individual a! and pz subunits, respectively. Mutant A-3 tally at 340 rnp in a Gilford recording spectrophotometer. One
forms a normal /3~ subunit and an altered (Ysubunit (a3) which is unit of activity in any of the reactions is defined as that activity
active in Reaction 1 with the normal @Zsubunit but is inactive in which leads to the disappearance of 0.1 pmole of substrate or the
Reactions 2 and 3 with or without the & subunit (16). Mutants formation of 0.1 pmole of product in 20 min at 37”. One unit
B-l and B-13 form normal a! subunits and altered & subunits of each subunit is that amount of protein which gives 1 unit of
((@l)z and @13)2, respectively) which are active in Reaction 2 activity in the presence of an excess of the other subunit. All
with the normal (Y subunit and inactive in Reactions 1 and 3 enzymatic units reported in this paper are units in the indole to
with or without the cr subunit (16). tryptophan reaction. In all quantitative experiments the sub-
Preparation of E&o&+-Cultures were grown in minimal units used were assayed in the standard indole to tryptophan
medium (20) supplemented with 0.2yo glucose and either indole reaction so that results could be expressed in molarity by the
(2 pg per ml) or tryptophan (4 pg per ml). Cells were harvested use of the molecular weight of the a! subunit (30,000) (19) and
Downloaded from www.jbc.org by guest, on October 22, 2011
by centrifugation, washed once with 0.80/, NaCl, suspended in its maximum specific activity in this reaction under standard
0.1 M Tris-HCl buffer, pH 7.8, and disrupted in a lo-Kc Raytheon conditions, 5000 units per mg.
sonic oscillator. The resulting extract was then centrifuged at Sucrose Gradient Centrijugution-The experimental procedures
105,000 x g for 60 min. The supernatant solution was generally outlined in detail by Martin and Ames (27) for the handling of
treated with ammonium sulfate to 40% of saturation, with the sucrose gradients were followed explicitly. All gradients were
exception of B-8 extracts, which were precipitated with am- of 5 to 20’% sucrose in 0.05 M Tris buffer at pH 7.8. Sucrose
monium sulfate at 60% of saturation. The precipitates were solutions containing additional compounds were prepared just
dissolved in 0.1 M Tris-HCl buffer, pH 7.8, and were stored at before use. Unless otherwise stated, enzyme preparations were
-15”. Unless otherwise stated, the tryptophan synthetase mixed in the cold immediately prior to layering on the gradients.
components studied were in such extracts. In certain cases, the The volume of enzyme preparation applied to the gradient was
Q and pz subunits were purified by the procedures of Henning always 0.1 ml. After centrifugation for 10 to 14 hours3 at
et al. (19), and of Hatanaka et al. (21) and Crawford and Ito (22), 39,000 rpm in a SW-39 rotor in a refrigerated model L Spinco
respectively. centrifuge, B-drop fractions were collected by a device similar to
Reagents-With the exception of InGP,z all reagents were that described by Martin and Ames (27). Each gradient yielded
obtained commercially and used without further purification. an average of 380 drops. The total contents of each fraction
InGP was synthesized enzymatically by the procedure of Smith were usually assayed; alternate fractions were assayed for a!
and Yanofsky (23). subunit and,& subunit activity. Normal (Y and 8~ subunits and
Enzyme Assays-The details of the procedures used for measur- mutant cr subunits were assayed in the indole to tryptophan
ing the enzymatic activities of the (Y and pz subunits have been reaction (Reaction l), and mutant ,& subunits were assayed in
described previously (24). All assays were carried out in a total the InGP to indole reaction (Reaction 2) in the presence of 1 M
volume of 1.0 ml at 37”. The conversion of InGP to indole NH%OH.HCl neutralized with NaOH.4 Each assay mixture
(Reaction 2) was followed by measurement of the amount of contained an excess of the normal second subunit. The inde-
indole formed from 0.1 to 0.3 pmole of InGP in the presence of pendently sedimenting o( subunit peak was assumed to have an
0.1 M Tris-HCl buffer, pH 7.8, in 20 min (24), by reaction with szo, u) value of 2.7s (19), and the other sedimentation coefficients
Ehrlich’s indole reagent (25). The enzymatic conversion of were calculated with this value as a basis. Control experiments
InGP to tryptophan (Reaction 3) was carried out in the presence
with catalase (27) and InGP synthetasel have established the
of 14C-InGP (prepared from 14C-indole), 60 mM nn-serine, 0.04
validity of this s value under the conditions employed.
mM pyridoxal-P, and 0.1 M Tris-HCl buffer, pH 7.8. After
Sepha&r Gel Filtration-Sephadex G-200 columns (0.6 x
incubation for 30 min, the remaining r4C-InGP was oxidized to
i4C-indole-3-aldehyde by metaperiodate at pH 5 and was ex- 36 cm) at 4” were used for gel filtration. Samples of 0.2 ml were
tracted with ethyl acetate (26). The radioactivity of both the layered on the column and eluted with solutions containing 0.05
r4C-tryptophan) and the ethyl M Tris-HCl buffer, pH 7.8. Fractions of about 0.7 ml were
aqueous layer (representing
acetate layer (representing residual 14C-InGP) was measured in collected.
a Packard liquid scintillation spectrometer. The indole to
3 The times of centrifugation are given as the time from start-
tryptophan reaction (Reaction 1) was carried out with a mixture ing to turning off of the centrifuge. They have not been corrected
of enzyme, 0.4 mM indole, 60 mM nn-serine, 0.04 InM pyridoxal-P, for the actual speed of the rotor. This was consistently measured
and 0.1 M Tris-HCl buffer, pH 7.8, and the disappearance of and did not vary significantly.
indole was followed (24). Routine indole to tryptophan assay 4 These conditions markedly stimulate the activity of the
associated (Yand 62 subunits in catalyzing Reaction 2 (I. P. Craw-
2 The abbreviation used is: InGP, indoleglycerol phosphate. ford, personal communication).
982 Tryptophan Synthetase Subunits Vol. 241, No. 4
RESULTS same ~20,u) previously reported for a complex of the a and @J
Sucrose Gradient Centrifugation-A typical sucrose gradient subunits (28). The excess cy subunit present appeared at the
sedimentation pattern of the E. coli tryptophan synthetase in position expected for the free a subunit, whereas all of, the fi2
the presence of Tris-HCl buffer (Fig. 1) shows a single symmetri- subunit sedimented in the 6.4s peak. In separate experiments
cal peak of (Y subunit activity and a similar peak of /I2 subunit with mixtures containing equal units of c11 and /I2 subunits, only
activity. The distances sedimented by the (Y and PQ subunit the 6.4s peak was observed. When a mixture containing signifi-
are consistent with their previously reported sZo,w values of cantly more units of the /lz subunit than the a subunit was sedi-
2.7s and 5.1S, respectively (19, 28). Each of the two subunits mented under similar conditions, all of the cx subunit sedimented
sedimented identically, whether the other subunit was present or with an apparent s value of 5.6 to 5.8S, while most of the /32
not. An extract of mutant T-3 containing both cr and pZ sub- subunit sedimented as the 5.1s species.’ The 6.4s peak has
units behaved the same as a mixture of the two subunits from been shown to be the cvZp2 complex and the 5.6 to 5.8s peak the
B-8 and A-2 extracts. These results indicate that any interac- ap2 complex.1 In view of these characteristics, all further sedi-
tion between the two subunits was insufficient under the condi- mentation studies were performed with mixtures containing cr
tions used to affect significantly their sedimentation character- subunit activity equal to or in excess of that of the pZ subunit.
istics. Under these conditions, complete association will be defined as
Ejfect of Pyridoxal-P and Serine-When mixtures of the two that state in which all the & subunit activity is in the 6.4s peak,
subunits containing an excess of the cr subunit were sedimented as in Fig. 2A. Any conditions of sedimentation producing peaks
in sucrose gradients containing 0.02 mM pyridoxal-P and 0.15 with s values less than 6.4S, but greater than those of the free
mM n-serine, a new, more rapidly sedimenting peak appeared, subunits, will be considered to be conditions favoring inter-
containing approximately equal levels of a! and & subunit en- mediate, but not complete, association of the a! and /32 subunits.
zymatic activities (Fig. 24). The distance sedimented by this In control experiments in which the two subunits were sedi-
Downloaded from www.jbc.org by guest, on October 22, 2011
peak indicates a sedimentation coefficient of 6.4S, which is the mented separately, no effect of pyridoxal-P and serine on their
sedimentation was observed. Of particular importance is the
fact that the a subunit did not form dimers. Thus, the 6.4s
J
5 peak would appear to be a specific complex formed by association
. of (Y and p2 subunits.
> 60 1 2i7 5i“i”
Identical results were observed with a T-3
extract containing both subunits and with a mixture of (Y and &
c
5 subunits from B-8 and A-2 extracts. Furthermore, purified (Y
i= subunit and the a! subunit in crude extracts were identical in
240 their sedimentation behavior with the & subunit.
8 Sucrose gradient experiments with o and p2 subunits in the
rn 20 presence of pyridoxal-P alone (0.02 mM) indicated a more limited
!z degree of interaction between the two subunits (Fig. 2B). Al-
5 --w though the extent of association produced by pyridoxal-P was
60 40 20 somewhat variable, a definite effect was consistently observed.
FRACTION NUMBER Serine alone (15 mM) did not detectably promote association of
FIG. 1. Sedimentation pattern of the tryptophan synthetase the two subunits. In other experiments with 0.02 mM pyri-
subunits in a 5 to 20% sucrose gradient containing 0.05 M Tris-HCl, doxal-P in the gradient solutions, the concentration of n-serine
pH 7.8. The vert& arrows- represent positions expected for was varied from 0.15 to 15 mM. In each case the (Y and /32 sub-
species with the indicated sedimentation coefficients. Time of
centrifugation: 12 hours. (Y subunit activity, 0; pi subunit ac- units sedimented together with an apparent sedimentation co-
tivity, 0. efficient of 6.4S, as in Fig. 2A. With the n-serine concentration
lowered to 0.015 mM in the presence of pyridoxal-P, intermediate
association was evident. These results indicate that pyridoxal-P
1
B and serine together at certain minimum concentrations cause
+ ‘i7 “i’ “I” the sedimentation of tryptophan synthetase in the completely
\ 40.
> associated form.
I- To investigate the specificity of n-serine in promoting complex
formation, the cx and /3z subunits were sedimented in the presence
of pyridoxal-P and either n-serine, n-threonine, or n-cysteine.
n-Serine was without effect, while 15 mM n-threonine, with pyri-
doxal-P, gave a very slight increase in the apparent association
of the cx and & subunits over that observed with pyridoxal-P
alone. At the same concentration, L-cysteine, with pyridoxal-P,
led to an apparent independent sedimentation of the two sub-
20 60 40 20 units, completely reversing combination stimulated by pyridoxal-
FRACTION NUMBER P alone. This inhibitory effect of cysteine is probably a result
FIG. 2. Sedimentation pattern of the tryptophan synthetase of its effectively competing with the enzyme for pyridoxal-P by
subunits from B-8 and A-2 extracts, respectively, in a sucrose forming a stable thiazolium derivative (29).
gradient containing 0.05 M Tris-HCl, pH 7.8, and (A) 0.02 mM Preincubatkn with Pyridaxal-P and Serine-To investigate
pyridoxal-P and 0.15 mM L-serine, and (B) 0.02 mM pyridoxal-P.
Times of centrifugation: (A) 13 hours, 31 min; (B) 11 hours, 39 further the effect of pyridoxal-P and serine on the sedimentation
min. (Y subunit activity, 0 ; pz subunit activity, l . of the QI and p2 subunits, the two subunits, each at concentrations
Issue of February 25, 1966 T. E. Creightm and C. Yanofsky 983
of 200 to 600 units per ml, were incubated together at 37” with and n-serine concentrations during the 37” incubation to 2 /.LM
pyridoxal-P (0.02 mM) and n-serine (15 mM) for 15 min, cooled and 1.5 mM, respectively, caused a more limited association of
to 0”, and immediately layered on a sucrose gradient containing the two subunits during sedimentation. Unincubated mixtures
only Tris buffer. In the previously described experiments the containing both subunits, pyridoxal-P, and serine gave a sedimen-
supplements were contained only in the gradient solutions. The tation pattern indicating less association (Fig. 3C), demonstrating
resulting sedimentation pattern showed that the LYand /3~ sub- that under the conditions used incubation was essential for com-
units had sedimented together as the 6.4s complex (Fig. 3~4). plex formation.
Incubation of the two subunits at 37” with only pyridoxal-P gave Following these observations, the role of each of the four com-
a sedimentation pattern indicating intermediate association (Fig. ponents (a and /3~ subunits, pyridoxal-P, and serine) in promoting
3B). Incubation at the same temperature with only serine or complex formation following incubation was investigated. The
without supplements did not cause significant association of the desired components were mixed, incubated at 37” for 15 min, and
two subunits during sedimentation. Decreasing the pyridoxal-P then cooled to 0”. The remaining components were then added
so that in all cases the cr and 82 subunits, pyridoxal-P, and serine
s
. B
2.7 5.1 6.4
were layered on the sucrose gradient. The only difference be-
tween the mixtures was in the components which were present
during the incubation at 37”. In each experiment the controls
z 60 of the cr and pz subunits with pyridoxal-P and serine, incubated
6
i= and not incubated, were run simultaneously. The resulting
240 sedimentation patterns in all experiments were clearly like either
i3 the incubated or unincubated controls (Fig. 3, A and C). The
pro results are given in Table I. They indicate that the (Y and PZ
Downloaded from www.jbc.org by guest, on October 22, 2011
subunits and pyridoxal-P must be present together during the
5 incubation, and serine must be added either during or after the
60 40 60 40 incubation, in order to observe the fully associated complex.
Effect of InGP, Indole, and Tryptophan on Sedimenta&-The
C
2.7 5.1 6.4 presence of InGP (0.1 mM) or indole (0.2 mM) in the sucrose gra-
dient in the absence of pyridoxal-P and serine did not produce
any apparent association of the tryptophan synthetase subunits.
With added pyridoxal-P (0.02 mM) and n-serine (15 mM), com-
plete association of the subunits was observed in both cases. To
test more critically for any effect on association of the 01 and /3~!
subunits, the compounds were added to sucrose gradients con-
taining pyridoxal-P (0.02 mM) and a low concentration of n-serine
(0.02 mM), conditions which were found to produce intermediate
association of the two subunits (Fig. 4A). InGP (0.1 mM) caused
FRACTION NUMBER
no alteration in this sedimentation pattern, but indole (0.2 mM)
FIG. 3. Effect of preincubation on the sedimentation patterns consistently caused an apparent decreased association of the (Y
of the tryptophan synthetase subunits from B-S and A-2 extracts, and & subunits (Fig. 4B). This inhibition could be reversed
respectively, in sucrose gradients containing only 0.05 M Tris-
HCl, pH 7.8. The two subunits in A and B were preincubated t o- somewhat by the addition of InGP (0.13 mM) (Fig. 4C). n-Tryp-
gether at 37” for 15 min in (A) 0.02 m&r pyridoxal-P and 15 mM L- tophan (1.0 mM) caused no alteration of the sedimentation pat-
serine; (B) 0.02 mM pyridoxal-P; C is as in A, except without terns obtained with pyridoxal-P (0.02 mM) and n-serine (15 mM
incubation at 37”. Time of centrifugation: 12 hours. (Y subunit
activity, 0 ; p? subunit activity, 0. or 0.02 mM). Thus, of the compounds tested only indole had
TABLE I
Effect of pretreatment on sedimentation pattern of a and 82 subunits
- -
Components added after preincubation Association observed Example of
Experiment Components present during preincubation L
sediientation pattern
-.
a, @z,* serine, pyridoxal-P Complete Fig. 3A
OL,82, pyridoxal-P Intermediate Fig. 3B
a, Pn, serine None detectable Fig. 1
Serine, pyridoxal-P ff, 82 Intermediate Fig. 3C
a, 82 Serine, pyridoxal-P Intermediate Fig. 3C
OL,serine, pyridoxal-P 82 Intermediate Fig. 3C
~2, serine, pyridoxal-P CY Intermediate Fig. 3C
(Y, serine, pyridoxal-P Intermediate Fig. 3C
Mixture
&, serine, pyridoxal-P
a, &, serine Pyridoxal-P Intermediate Fig. 3C
(Y,~2, pyridoxal-P Serine Complete Fig. 3A
- -
* cz = a subunit; pz = & subunit.
Tryptophan Synthetase Subunits Vol. 241, No. 4
Therefore, no explanation can be given for the differential effect
A B
J.
2.7 5.1 6.4 2.7 5.1 6.4 of NaCl.
J 4 I 5 1 4 SephadexGel Filtration-Sephadex gel filtration of mixtures
6.2 5.8 of B-8 and A-2 extracts gave results similar to those obtained by
sucrose gradient centrifugation. The presence of serine and
pyridoxal-P together led to migration of the a! and pZ subunits
as the fully associated complex, while 0.1 M NaCl did not. How-
& ever, preincubation of the a! and & subunits with pyridoxal-P
and serine, as in Table I, did not result in the migration of the
subunits as a complex.
60 40 AssociationConstant Measurements--The usefulness of sucrose
gradient centrifugation for analysis of the factors affecting asso-
C
2.7 5.1 6.4 ciation of the subunits of tryptophan synthetase is evident, but
& this technique has given only qualitative data on subunit asso-
' 6.2 ciation. As mentioned previously, tryptophan synthetase
&
catalyzes three primary enzymatic reactions, and the two sub-
units must be in physical contact for maximal activity in each of
these reactions (10). Under the usual assay conditions, the (II
subunit alone has 1% of the activity of the associated subunits
in the InGP to indole reaction (Reaction 2), and the & subunit
has 3% of the activity of the complex in the indole to tryptophan
60 40
Downloaded from www.jbc.org by guest, on October 22, 2011
reaction (Reaction 1). It should be possible, therefore, to meas-
FRACTION NUMBER ure association quantitatively by determining the enzymatic
FIG. 4. The effect of indole on the sedimentation pattern of the activity of a known mixture of subunits, but this requires knowl-
tryptophan synthetaae subunits from B-S and A-2 extracts, re- edge of the species present in the mixture and their enzymatic
spectively, and its reversal by InGP. Each sucrose gradient con- activities. We have carried out sedimentation velocity studies
tained 0.05 M Tris-HCl, pH 7.8, 0.02 mM pyridoxal-P, and 0.02
in the analytical ultracentrifuge with high concentrations of the
mM n-serine. In addition, B contained 0.2 mM indole, and C con-
tained 0.2 mM indole and 0.13 mM InGP. Time of centrifugation: purified (Y subunit and have been unable to detect aggregation
12 hours, 30 min. (Ysubunit activity, 0 ; 82 subunit activity, l . in the presence or absence of InGP, serine, and pyridoxal-P.
The finding that the s values observed in sucrose gradient studies
any observable influence on the sedimentation of the enzyme with very low concentrations (as low as 4 units per ml) of the PZ
complex. subunit were identical with those found with the purified protein
Effect of NaCl on Sedimentation-The addition of NaCl to at high concentrations suggests that this dimer does not reversi-
assay mixtures increases the tryptophan synthetase activity of bly dissociate to form significant amounts of the free @ polypep-
crude extracts (30). In view of the results obtained in the pres- tide chain at the concentrations and conditions normally
ent study, it was thought that this stimulation could be due to employed in enzymatic activity determinations. The associa-
increased association of the two subunits. When the effect of tion reaction between the (Y and & subunits can thus be described
NaCl on the sedimentation of the 01 and pz subunits was investi- as
gated, it was found that at NaCl concentrations of 0.04 to 0.1 M,
K1 K2
in the absence of serine and pyridoxal phosphate, the subunits Cr+lszF=+ a82 ; a& + LyF===+ a& (1)
from T-3 extracts sedimented in the 6.4s associated form. NaCl The enzymatic activity found in Reaction 1 upon addition of a
concentrations below 0.04 M produced correspondingly less asso- constant small amount of (Y subunit to varying amounts of /3r
ciation. Surprisingly, 0.1 M NaCl had no effect on the sedimen- subunit is shown in Fig. 5. If increasing amounts of c& complex
tation pattern observed with a mixture of the (Y and & subunits
derived from B-8 and A-2 extracts, respectively, in that each
subunit sedimented in the apparently unassociated form. In all
the previously described experiments, identical results were ob-
tained with the a! and pZ subunits from T-3 extracts and from B-8
and A-2 extracts. Neither precipitation of the proteins of the
T-3 extract with ammonium sulfate nor dialysis of the extract
against 0.1 M Tris buffer, pH 7.8, significantly altered the sedi-
mentation pattern of the a! and PZ subunits in 0.1 M NaCl, indi-
cating that no readily removable small molecules were involved
in their association. However, there was an indication that the
extent of association found in the presence of NaCl decreased
Y .J
with the age of the T-3 extract. 0 5 IO 15
The presence of NaCl in the sucrose gradient solutions was UNITS OF a, SUBUNIT ADDED
only found to increase association of the two subunits from B-8
FIG. 5. Saturation of 01 subunit by 82 subunit in Reaction 1 in
and A-2 extracts when pyridoxal-P was also present, but these the presence of 0.18 M NaCl. Activity was corrected for intrinsic
conditions produced an intermediate degree of association. activity of 02 subunit present in each mixture.
Issue of February 25, 1966 T. E. Creightcmand C. Yanofsky
were formed as the concentration of flZ subunit was increased, as
is inferred from the sedimentation studies,’ such a saturation
curve would indicate that the c& complex is equally as active
as the a&~ complex per mole of (Y subunit bound (i.e. a& has
twice the absolute activity of c&). Similar saturation curves
for the addition of (Y subunit to a constant amount of pZ subunit
have been presented (lo), and these observations give no indica-
tion of any significant difference in affinity of the p2 subunit for
the first and second (2 subunits. The finding in sedimentation
studies’ of predominantly (~82 complex with equal levels of (Yand
pz subunit activities and a majority of c& complex with excess
levels of & subunit activity supports this interpretation. In the
remainder of this paper we shall treat the association of the (Y
and & subunits as a case in which there are two identical and
independent (Y subunit-binding sites on the &X subunit, with the
enzymatic activity of mixtures of the two subunits a measure of
such sites occupied by the a subunit.
Such a system of identical and independent combining sites
may be treated kinetically as the interaction of a single site of the
(Ysubunit (a) with an individual site on the pZ subunit (b), taking
the two separate sites on the ,& subunit to be equivalent (31), UNITS OF a SUBUNIT ADDED
Downloaded from www.jbc.org by guest, on October 22, 2011
kl ,L$SUBUNlT / Q SUBUNIT = 1.48
a+b e ab
kz
FIG. 6. The catalytic activity at varying concentrations of a
where ab represents a catalytic site formed by combination of two mixture of LYand & subunits from B-8 and A-2 extracts, respec-
tively. Activity was measured in Reaction 1 in the absence of
nonidentical subunits. Similarly, an association constant, K,, NaCl. The straight line represents the activity expected with
can be defined as complete association of the a and 82 subunits. The curved line
is the line calculated with Equation 4 for “KA” = 0.21 (units per
ml)-‘, the average of the “KA” values calculated at each of the
experimental points.
From saturation curves such as the one shown in Fig. 5, it is evi- TABLE II
dent that the total concentration of a and b combining sites may Apparent
._ association constants of 01.and 62 subunits
be determined by measuring the enzymatic activity present with
a saturating concentration of the other subunit, as in the normal Addition to
Enzymatic reaction reaction mixture “KA ”
assays for a! and fiZ subunit activities. An apparent association
at-1
constant (“KA”) can then be defined on the basis of enzymatic
activity as pyridoxal-P
1. Indole + L-serine j 3.2 X lo7
-Lb L-tryptophan 0.18 M NaCl 8.0 x 108
(4)
‘lKA” = (A,, - AoJ(Ab - B,a) 2. InGP + indole + glyceraldehyde- 4 x 108
3-P 0.10 M NaCl 4 x 106
where &, is the enzymatic activity found with a mixture of the pyridoxal-P
3. InGP + L-serine p 5.4 x 108
cx and flz subunits, and A, and & are the enzymatic activities of L-tryptophan + glyceraldehyde- 0.10 M NaCl 2.6 X lo8
the same mixture found in the presence of added excess /Iz and (Y 3-P
subunits, respectively (i.e. the total concentration of a and b
combining sites). In this treatment it is assumed that essen-
tially all of the a! and pZ subunits exist as either free subunits or (Y and /3z subunits at several concentrations. The results of such
in an enzymatically active complex. Furthermore, the slight an experiment are shown in Fig. 6 and are consistent with the
activity of the unassociated subunits is ignored in determining equilibrium reaction shown in Equation 2, with “Kg” as defined
A ab. “KA” is an apparent association constant because the in Equation 4. This experiment was repeated in each of the
equilibria of the subunits with low molecular weight substances three enzymatic reactions in the presence and absence of NaCl,
such as the substrates and the coenzyme are not considered. with the use of a! and pZ subunits from B-8 and A-2 extracts,
Precautions were taken, however, to insure that the subunits respectively, and the results converted to molarity (Table II).
were saturated with such substances. It was of course realized The measured “KA” is the apparent intrinsic association constant
that flZ subunits with and without bound pyridoxal-P probably of the cr and & subunit combining sites. The values of “K1” and
behaved as very different species. Nevertheless, it was of inter- “K;’ of Equation 1 are then 2 x “KA” and i x “KA,” respec-
est to determine the apparent association constants in the differ- tively (32). The larger association constants observed for
ent reactions so that the effect of various substances on complex Reaction 3 as compared to those found for Reaction 2 are pre-
formation could be assessed. sumably due to the presence of pyridoxal-P and serine in the
The procedure used to estimate “KA” was to determine the reaction mixtures of Reaction 3. As was shown previously, the
ensvmatic activity of a mixture of approximately equal units of association of the two subunits during sucrose gradient centrifu-
Tryptophan Synthetase Subunits Vol. 241, No. 4
gation is favored by these additions. Since InGP was found to sucrose gradient containing only Tris buffer. Thus, in these
have no effect on the association of the two subunits in the sedi- tests, this mutant Q! subunit behaved like normal a~ subunit.
mentation studies, while indole was found to inhibit association Previous studies (16, 17) also suggested that most mutant (Y
to a certain extent, the somewhat smaller “Ka” found in the subunits combined normally with the pz subunit.
indole to tryptophan reaction (Reaction 1) compared to the Extracts of two mutant strains, B-l and B-13, which produce
InGP to tryptophan reaction (Reaction 3) may be due to the normal (Y subunits and mutationally altered, enzymatically in-
inhibition of association by indole. It is also to be noted that active fiZ subunits ((@l)~ and (/313)2, respectively) were used to
NaCl increased the association constants only in those reactions study the association of mutant & subunits with normal (Y sub-
where serine and pyridoxal-P were also present, in agreement unit. Upon sedimentation in the presence of pyridoxal-F and
with the sedimentation studies of the two subunits from B-8 and serine, the a! and (01)~ subunits showed a limited degree of asso-
A-2 extracts. ciation, the @l)s subunit sedimenting with an apparent s value
If the effect of incubation of the (Y and pZ subunits with pyri- of 5.7s. The (Y and (/?13)~ subunits showed no association upon
doxal-P and serine seen in the sedimentation studies were to sedimentation under the same conditions. However, both mu-
increase association, the measurements of “Ka” for Reactions tant & subunits seemed to be quite labile under the conditions
1 and 3 could be in error, since during the course of these reac- employed. Only a very small amount of the (/313)2 subunit
tions the two subunits are incubated in the presence of pyridoxal- (which had sedimented with an apparent s value of 5.18) was
P and serine. This might have the effect of placing the value detected following centrifugation. The association constants
measured somewhere between the initial and final association of the two subunits from these two strains were measured in the
constants. To test this possibility, a mixture of a and & sub- 1nGP to indole reaction (Reaction 2), and the effect of added
units was incubated with pyridoxal-P and serine under the condi- pyridoxal-P and serine on the association constants was deter-
tions used in the sedimentation studies. The association con- mined (Table III). The components from both mutant strains
stant in the indole to tryptophan reaction for these two subunits gave “KA” values which were not different from the values ob-
Downloaded from www.jbc.org by guest, on October 22, 2011
was then measured and compared with that of an identical mix- tained with normal a and /3~ subunits in the InGP to indole
ture of a! and /3~ subunits which had not been incubated. The reaction. With added pyridoxal-P and serine, the subunits from
results were identical for the two mixtures. It must be con- mutant B-13 showed no detectable change in “KA,” while those
cluded, therefore, either that the incubation in the presence of from mutant B-l were apparently more firmly associated. In
serine and pyridoxal-P produces no significant change in the both cases the enzymatic activity of the associated subunits was
association constant of the two subunits, or that any change that significantly increased. Of course, the effect of pyridoxal-P and
is produced occurs within a length of time very small compared to serine on the association of normal a! and p? subunits cannot be
the length of the assay incubation period. In neither case were determined in this reaction, as their presence would lead to the
the measured association constants of Table II in error because conversion of the InGP to tryptophan. However, for compari-
of an effect of incubation. son the “KA” for the normal components in the presence of
Studies with Mutant Subunits-It was desired to determine IndP, serine, and pyridoxal-P is 5 x 108 M-I. Thus the enzy-
what effect mutations altering the enzymatic activity of the or matic data on association is qualitatively consistent with the
or pZ subunit have on the ability of the mutationally altered sub- sedimentation findings, suggesting that in the presence of pyri-
unit to associate with the other component. One mutant a! doxal-P and serine neither of the mutant p2 subunits enters into
subunit, ar3 from strain A-3, was selected for testing. Upon complex formation with the (Y subunit as well as the normal pZ
sedimentation in a sucrose gradient containing only Tris buffer subunit. These results indicate that mutational alteration of the
the purified a3 subunit behaved the same as the normal or sub- flZ subunit can selectively prevent the increased association of the
unit and showed no significant interaction with normal fiz sub- tryptophan synthetase components normally caused by pyri-
unit, In the presence of pyridoxal-P and serine, the (r3 subunit doxal-P and serine, without affecting their association in the
sedimented with the normal 0~ subunit as the 6.48 species. The absence of pyridoxal-P and serine. Previous studies (33, 34)
same result was found if the two subunits were incubated at 37” have also indicated that mutations may alter the complex-form-
with pyridoxal-P and serine for 15 min, prior to layering on a ing ability of the & subunit.
Kinetic Studies-The enzymatic conversion of InGP to indole
TABLE III or tryptophan can be followed spectrophotometrically by cou-
Apparent association constants of normal 01 and mutant pling the reaction to the glyceraldehyde 3-phosphate dehydro-
~2 subunits genase reaction (14). This procedure is convenient for kinetic
-
studies and was used to follow the association reaction of the
Relative
Source of tryptophan Addition to reaction catalytic subunits of tryptophan synthetase. It had previously been
synthetase mixture rate of “KA”
COllpl~X’
found that there was at most a l- to 2-min lag in attaining maxi-
_ mal rates in the indole to tryptophan reaction (Reaction 1) after
Al-1 mixing the a! and /32 subunits to start the reaction (10). This
B-8 and A-2 (normal 1.0 4 x 108 result has been confirmed in Reactions 2 and 3, as shown in Fig.
(Y and normal &) 7. The short lag observed in Curve 2 presumably represents the
B-l 1.0 4 x 106 time required for the association of the subunits to form an enzy-
B-l Serine, pyridoxal-Pt 1.6 2 x 10’ matically active complex. Since the limiting subunit in each
B-13 1.0 4 x 106
B-13 Serine, pyridoxal-Pt 1.2 case could be calculated to be 90% or more in the associated form
4 x 106
under the conditions used (with the use of the appropriate “KA”
* InGP + indole + glyceraldehyde-3-P. of Table II), the association reactions were taken to be essentially
t 30 mM and 0.04 mM, respectively. irreversible (35). The association rate constants (ki of Equa-
Issue of February 25, 1966 T. E. Creightm and C. Yanofsky 987
tion 2) could then be estimated (35) to be 2 x lo* see-1 M+ in 1.0
Reaction 2 and 6 X lo6 see-l M-’ in Reaction 3. In both cases 0
-0
the rapidity of the association reaction made it impossible to
analyze the shape of the curves. Thus the presence of pyridoxal- 0.5 3
P, serine, and NaCl increases the association-rate constant by a L
factor of 30. Of course, these rate constants, and those below, I
4
are subject to the limitations discussed in relation to association
constant measurements.
Determination of the rate of dissociation of the combined sub-
-6-
units cannot be accomplished in such a straightforward manner, 0 20 40
but it is possible to obtain an estimate of dissociation with the TIME (minutes)
use of competition of mutant and normal ar subunits for a limit- FIG. 9. A, kinetics of generation of DPNH at 30” by a glyc-
ing amount of pa subunit. The complex formed between a eraldehyde a-phosphate dehydrogenase system coupled to
mutationally altered (11subunit, such as the (~3 subunit from Reaction 3 catalyzed by tryptophan synthetase. For Curve 1,
mutant A-3, and a normal /3~ subunit, is inactive in Reactions 2 purified ~2 subunit was added to a mixture of purified a3 and
and 3 (16). The a3 subunit had previously been shown to com- normal 01 subunits, and the other components of the reaction
mixture (see “Materials and Methods”) then added to start the
bine with the & subunit as readily as normal ~11 subunit (16), and reaction.‘ For Curve 2, the 82 subunit &as added to the c~3 sub-
to be indistinguishable from the normal subunit in sucrose gradi- unit, followed by the reaction mixture. No production of DPNH
ent sedimentation studies (described earlier in this paper). was observed until normal (Ysubunit was added at zero time. In
Further evidence showing this point was obtained in competition both cases, 62 units of 8~ subunit, 130 units of (~3 subunit, and 11
units of Al subunit were present. B, analysis of data of (A) by
experiments with the normal a! subunit. Normal /3~ subunit was plotting (1 - r), when r is the ratio of the slopes of Curve 1 to that
added to a mixture of purified a3 and normal cx subunits, and of Curve 2 at the same DPNH concentration, on a logarithm
Downloaded from www.jbc.org by guest, on October 22, 2011
activity in Reaction 3 was followed spectrophotometrically. scale against time (see “Appendix”) to obtain Curve PI. Curve 4
Linear reaction kinetics was observed over a considerable period, is the result of a similar experiment in Reaction 2, in which the
concentration of each of the subunits was 20.fold greater than
indicating that the mixture was in equilibrium. The findings
those used in A.
with various ratios of mutant to normal (Y subunits are plotted
in Fig. 8. The results indicate that the or3 subunit combines
followed spectrophotometrically upon addition of a small amount
with the normal & subunit to the extent expected for a subunit
with the same affinity as the normal ar subunit for the /3z subunit. of normal (Y subunit. The equilibration reaction between sub-
units can be described as
To measure the rate of dissociation of the mutant (Yand normal
pZ complex, the a3 subunit was equilibrated with a limiting a + a*b e a* + ab (5)
amount of pz subunit, and catalysis of either Reaction 2 or 3 was
where a* represents the combining site of the mutant a! subunit
and a and 6 are as defined previously. The rate obtained was
1.0 compared with that obtained with an identical mixture which
8
was at equilibrium, in which the flZ subunit was added to a previ-
-g.o3 ously prepared mixture of the ar3 and normal (II subunits. The
‘I
a results obtained by measuring activity in Reaction 3 are shown
E in Fig. 9A. Purified subunits were used in each case, and there
3 R
0.5 was no indication of any contaminating enzymes capable of oxi-
2.02 I 2
dizing DPNH. Experiments in which equilibration was accom-
E plished in a mixture containing only serine, pyridoxal-P, NaCl,
?,I and buffer indicated that the substances added to permit glyc-
5 eraldehyde-3-P determination had no effect on the dissociation
k of the a*b complex. The data should ideally yield a straight line
E
0 2 4 6 0 when plotted as in Fig. 9B (see “Appendix”). The values for
TIME (minutes) the dissociation rate constant (kz of Equation 2) for Reactions
2 and 3 that were estimated from these data (see “Appendix”)
were 4.8 X 1W3 see-r and 1.8 X 10-4 set-I, respectively. These
FIG. 7 (left). Kinetics of generation of DPNH by a glyceral- values correspond to half-lives of 2.4 and 64 min, respectively.
dehvde 3-nhosnhate dehvdrogenase system coupled to Reaction 3
catalyzed-by &yptopha”n synthetase. Curve i, a mixture of 30 Thus, the tryptophan synthetase complex is readily dissociable,
units of purified CI subunit and 11.4 units of 82 subunit from an and serine, pyridoxal-P, and NaCl, when present together, in-
A-2 extract was added to the reaction mixture (see “Experimental crease its half-life by a factor of about 27. It is also to be noted
Procedure”) to start the reaction; Curve 2, the (Y subunit was that the ratios kr:& obtained for Reactions 3 and 2 are 3 X log
added to the reaction mixture containing the 82 subunit to start M? and 4 x lo6 M?, respectively, in good agreement with the
the reaction.
FIG. 8 (G&t). Competition of mutationally inactive or3 sub- apparent association constants measured previously (2.6 X log
unit and normal CY subunit for limiting 02 subunit. The ratio of ~-1 and 4 x lo6 ~-1, respectively).
activitv CR1 with normal CLsubunit and ~2 subunit to that with (r3
added is plotted against the fraction of i subunit that was mutant. DISCUSSION
Activity was measured spectrophotometrically in Reaction 3.
The solid line represents the results expected for mutant and The results of this and other investigations’ (16) are consistent
normal OLsubunits with identical affinities for PZ subunit. with the interpretation that the formation of the tryptophan
988 Tryptophan Synthetase Subunits Vol. 241, No. 4
synthetase complex involves the reversible binding of individual No attempt has been made to correlate the association con-
(Y subunits to two identical and independent sites on a /3z sub- stants measured enzymatically and the association constants
unit. Furthermore, it would appear that all a subunits bound that would be predicted to be governing the migration of the a!
to the /3~ subunit are equally active enzymatically and that the and 0~ subunits in sedimentation experiments. Such a compari-
increased enzymatic activity of a mixture of (Y and pz subunits son would be of interest, since it must be realized that any physi-
over that exhibited by the individual subunits is a measure of the cal aggregation would be observed during migration, while only
a! subunits bound to pZ subunits. This interpretation is consist- enzymatically active complexes would be measured in the enzy-
ent with the known characteristics of the two subunits; the QI matic measurements of association. However, the excellent
subunit is a single polypeptide chain (11) and the & subunit is a qualitative correlation found here between the two different
dimer with 2 bound molecules of pyridoxal-P (13), and therefore methods of examining association suggests that under the condi-
presumably two catalytic sites. tions employed every aggregate observed by migration was either
The characteristics of the tryptophan synthetase made it pos- an enzymatically active complex or an intermediate in the forma-
sible to measure quantitatively the kinetic parameters of the tion or dissociation of such an active complex.
association of the two subunits. Although the results obtained Sucrose gradient sedimentation, Sephadex gel filtration, appar-
were in agreement with the interpretation mentioned, it was not ent association constant measurements, and association and
possible to rule out relatively small differences in the affinity of dissociation rate constant measurements have indicated that
cr for the two /3~ combining sites or in the enzymatic activity of pyridoxal-P and serine are the most effective of the compounds
the two PZ subunit combining sites. The quantitative measure- tested in favoring the association of the tryptophan synthetase
ments presented here should be considered with these limitations subunits. It is of interest that the effect of these two compounds
in mind. on the “KA” of the two subunits is a consequence of an approxi-
The theoretical considerations of Bethune and Kegeles (36, 37) mately equal effect on the association and dissociation rate con-
on the zone migration of interacting species in rapid equilibrium stants. Of the two compounds, the cofactor pyridoxal-P seems
Downloaded from www.jbc.org by guest, on October 22, 2011
have not dealt explicitly with a complex analogous to tryptophan to play a greater role, as serine appears to be effective only in its
synthetase, but their general conclusions should be applicable to presence. Such a role for pyridoxal-P is not unique to trypto-
the present study. With any subunit complex, when the sub- phan synthetase, as this coenzyme is also involved in the associa-
unit affinity is above a certain level, migration of the subunits tion of the subunits of glycogen phosphorylase a (39). In other
as an aggregate should be observed. Furthermore, Bethune and cases as well, enzymatic cofactors influence association of enzyme
Kegeles (37) showed that if an excess of one of the subunits were subunits (for example, see Reference 40). Indole and cysteine
present, the excess would migrate in the unassociated form. were the only compounds found to have any inhibitory effect
That the tryptophan synthetase 6.48 peak is the fully associated on association. The significance of the cysteine effect is un-
complex (cr$L) is suggested by the symmetry of the (Y and pZ known, but of interest, since cysteine has recently been shown to
subunit peaks comprising it and by the presence of equal units replace serine to a small extent in Reactions 1 and 3 catalyzed
of (Y and & subunit enzymatic activity within the peak, as would by tryptophan synthetase (22). The absence of any effect of
be expected for the fully associated enzyme.1 This interpreta- InGP on association during sedimentation makes it likely that
tion of the 6.48 peak is also supported by the finding that only the apparent association constant measured in the InGP to
the 6.4s complex is observed when much higher subunit concen- indole reaction is a measure of the basal affinity of the (II and pZ
trations are used’ (28) or when conditions are imposed that in- subunits in forming an enzymatically active complex, and the
crease complex formation (for example, the addition of NaCl to “KA” measured in the InGP to tryptophan reaction probably
pyridoxal-P and serine). Bethune and Kegeles (36, 37) also represents the affinity of the two subunits in the presence of
predicted that subunit affinities less than those required for com- pyridoxal-P and serine.
plete association would produce migration of each of the subunits Of the two tryptophan synthetase subunits, the p2 subunit
at a rate intermediate between that of the free subunit and the would appear to have the more important role in the association
aggregate. Although in the present study the (Y and p2 subunits response to small molecular weight substances. Since the sepa-
were observed to migrate at intermediate rates, analyses of these rated /32 subunit binds 2 molecules of pyridoxal-P (13) and can
situations would be complicated by the existence of two possible catalyze Reaction 1 involving pyridoxal-P and serine, it is prob-
aggregates, ap, and a,/%. Accordingly, no detailed analyses of able that it possesses the catalytic binding sites for the two com-
such instances of intermediate migration were attempted. How- pounds found to be most effective in promoting association of
ever, the use of equal or excess units of the (ILsubunit (relative to the a! and /3z subunits. Furthermore, the effect of mutational
the pz subunit) insured that complete association of the two alterations of the /3~ subunit on the response of the a! and p2 sub-
subunits would always produce the 6.48 peak. If the affinity of units to pyridoxal-P and serine, observed here and elsewhere (33),
subunits for each other were below a certain level, their interac- suggests a critical role for the PZ subunit in responding to pyri-
tion would be expected to have no observable effect on their doxal-P and serine.
migration (1). In the case of tryptophan synthetase, no appar- The explanation of the effect of incubation of the cr and PZ
ent interaction between the QI and & subunits was observed in subunits with pyridoxal-P and serine on the stability of the com-
the presence of buffer and InGP alone, although the enzymatic plex during sucrose gradient centrifugation is not known. Fur-
studies indicated that there was an effective, though small, inter- thermore, no explanation can be given for the difference in sedi-
action between the two subunits under similar conditions. The mentation behavior in the presence of NaCl of the a! and &
chromatographic behavior of the LY and /3~ subunits (10) has subunits synthesized together in strain T-3 and those synthe-
previously been concluded to agree with expectations for an sized separately in strains B-8 and A-2. The fact that the two
enzyme complex (38). subunits from T-3 were effectively incubated together during
Issue of February 25, 1966 T. E. Creightm and C. Yanofsky 989
growth leads one to suspect that the differences in the subunits can be taken to be constant with time
are related to the effect of incubation with pyridoxal-P and serine
d(b)
mentioned above. However, it has not been possible to repro- - = 0 = kd(ab) + (a*b)l - h(b)[(a) + (a*)1 (3)
at
duce the behavior of the T-3 enzyme with the two subunits from
separate extracts, although the possibility exists that the required assuming the mutant and normal Q! subunits to have identical
conditions have not been found. rate constants for their association (ki) and dissociation (k2) with
The question naturally arises as to what significance, if any, the 02 subunit. Solving Equation 3 for (b) and substituting this
these findings have for the regulation of tryptophan biosynthesis value in Equation 2 yields
in E. coli. The role of pyridoxal-P and serine in association of
the cy.and /?z subunits bears certain similarities to the roles of d&a
-= Aa& - (Aa + &d-&b
kz (4)
certain cofactors and substrates as allosteric effecters in other dt A, + Aa* - Aa
systems (4-6, 41). However, it is not known as yet whether
pyridoxal-P and serine are bound at sites other than the active where AF,, A,, and A,* are the total concentrations of &, (Y, and
site of the enzyme complex. It is conceivable that the cell could a3 subunit enzymatic activities (i.e. combining sites) added to
regulate tryptophan biosynthesis by controlling the association the mixture, and &a is the enzymatic activity in Reaction 2 or
of the two subunits of tryptophan synthetase. However, since 3 of the mixture due to ab complex. & is taken to be the sum
this enzyme catalyzes the terminal reaction in the tryptophan (ab) + (a*b), since the concentration of free & combining sites
biosynthetic sequence, feedback inhibition of the enzyme cata- (b) could be calculated (wit.h the appropriate “Ka” values of
lyzing the initial reaction (42) and repression (43) would both be Table II of the text) to be a negligible part of Ab. Integration of
expected to be more economical. Nevertheless, it is of interest Equation 4 after separation of variables and setting A,b = 0 at
to consider whether association of the tryptophan synthetase t = 0, yields
subunits could be of any physiological significance. Extrapola-
Downloaded from www.jbc.org by guest, on October 22, 2011
A,* = C(1 - e--DW) (5)
tion of results in vitro to predict behavior in tivo can be very
dangerous, and special care must be taken to consider the differ- where
ences in concentrations of the various components in tivo and
in vitro. Wild-type E. coli grown under conditions of tryptophan c = Aa Ab
Aa + Aa*
repression yields about 100 units of both (Y and /% subunit activity
per-ml of packed cell volume, which would be a minimum esti- and
mate of the concentration in viva of these subunits. With the
smallest association constant measured in vitro (in Reaction 2), Aa + Aa*
D=
about 60% of these subunits would be expected to be associated. Aa + Aa* - Aa
If the association constants have any validity in vivo, the cell Since the catalytic rate of the equilibrated control is propor-
could achieve, at most, limited control of the degree of association tional to C, the ratio (r) of the enzymatic rate of the equilibrating
of the tryptophan synthetase subunits. Such a limited control mixture at a given time to that of the control mixture at the same
could be of significance if serine or pyridoxal-P or both play a DPNH concentration will be
role in coordinating different amino acid biosynthetic pathways.
It is also possible that some unknown compound causes dissocia- r = 1 - e--l)kzt (6)
tion of the enzyme and regulation is achieved by balancing the or
effect of this compound against the effect of pyridoxal-P and
serine. A more definitive estimate of the physiological signifi- Dktt = - ln(1 - r) (7)
cance of the factors affecting complex formation observed here
Therefore, the slope of the line given by plotting In (1 - r)
must await measurements of the association of the tryptophan
versus t will be -Dh$.
synthetase subunits in vivo.
Acknozuledgrnent-The authors are indebted to Drs. S. J.
APPENDIX
Singer and R. L. Baldwin for their valuable suggestions.
Kinetics of Dissociation of a*b Complex-The rate of dissocia-
tion of the a*b complex was followed as in Fig. 9A of the text by REFERENCES
the approach to equilibrium of the mixture 1. REITHEL, F. J., Advan. Protein Chem., 18, 123 (1963).
2. MUIRHEAD, H., AND PERUTZ, M. F., Cold Spring Harbor Symp.
a + a*b = a* + a6 (1) &ant. Biol., 28, 451 (1963).
3. WYMAN, J., Cold Spring Harbor Symp. Quant. Biol., 28, 483
upon adding a small amount of normal a: subunit to a mixture (1963).
of & subunit and excess inactive 013 subunit. The designations 4. GERHART, J. C., AND SCHACHMAN, H. K., Biochemistry, 4,
in Equation 1 are identical with those of Equation 5 of the text. 1054 (1965).
5. CHANGEUX, J., Cold Spring Harbor Symp. Quant. Biol., 28,
The rate equation for the formation of ab is 497 (1963).
6. FREUNDLICH, M., AND UMBARGER, H. E., Cold Spring Harbor
d(ab)
- = kl(a)(b) - kz(ab) Symp. Quant. Biol., 28.505 (1963).
dt 7. DATTA, P., GEST, H., AND SEGAL, H. L., Proc. Natl. Acad. Sci.
U. S., 61, 125 (1964).
Since only a small amount of (II subunit was added to start the 8. CAHN, R. D., KAPLAN, N. O., LEVINE, L., AND ZWILLING, E.,
reaction, the concentration of free flZ subunit combining sites (b) Science, 136, 962 (1962).
990 Tryptophun Synthetase Subunits Vol. 241, No. 4
9. SCHLESINGER, M. J., AND LEVINTHAL, C., J. Mol. Biol., 7, 1 27. MARTIN, R. G., AND AIIIES, B. N., J. Biol. Chem., 236, 1372
(1963). (1961).
10. CRAWFORD, I. P., AND YANOFSBY, C., Proc. Natl. Acad. Sci. 28. CRAWFORD, I. P., Bacteriological Proc., 119 (1962).
U. S., 44, 1161 (1958). 29. BUELL, M. V., AND HANSEN, R. E., J. Am. Chem. Sot., 82,
11. CARLTON, B. C., AND YANOFSKY, C., J. Biol. Chem., 237, 1531 6042 (1960).
(1962). 30. YANOFSKY, C., AND STADLER, J., Proc. Natl. Acad. Sci. U. S.,
12. WILSON, D. A., AND CRAWFORD, I. P., Bacteriological Proc., 44, 245 (1958).
92 (1964). 31. TANFORD, C., Physical chemistry of macromolecules, John Wiley
13. WILSON, D., AND CRAWFORD, I. P., J. Biol. Chem., 240, 4081 and Sons, Inc., New York, 1961, p. 533.
(1965). 32. EDSALL, J. T., AND WYMAN, J., Biophysical chemistry, Vol. I,
14. CRAWFORD, I. P., Biochim. Biophys. Acta, 46,405 (1960). Academic Press, Inc., New York, 1958, p. 624.
15. YANOFSKY. C.. AND RACHMELER. ,
M.. Biochim.
I
Biowhus. ” Acta.
‘ 33. GIBSON, F., GIBSON, M. I., AND YANOFSKY, C., J. Gen. Micro-
28, 640 (i958j. biol., a4, ‘301 (196i).
16. YANOFSKY, C., AND CRAWFORD, I. P., Proc. Natl. Acad. Sci. 34. CRAWFORD, I. P., AND JOHNSON, L. M., Genetics, 48, 725 (1963).
U. S., 46, 1016 (1959). 35. BRAY, H. G., AND WHITE, K., Kinetics and thermodynamics in
17. YANOFSKY, C., HELINSKI, D. R., AND MALING, B., Cold Spring biochemistry, Academic Press, Inc., New York, 1967, p. 162.
Harbor Symp. &ant. Biol., 26, 11 (1961). 36. BETHUNE, J. L., AND KEGELES, G., J. Phys. Chem., 66, 433
18. YANOFSKY, C., J. Biol. Chem., 224,783 (1957). (1961).
19. HENNING, U., HELINSKI, D. R., CHAO, F. C., AND YANOFSKY, 37. BETHUNE, J. L., AND KEGELES, G., J. Phys. Chem., 66. 1755
(1961).
C.. J. Biol. Chem.. 237. 1523 (1962).
38. NICHOL, L. W., BETHUNE, J. L., KEGELES, G., AND HESS, E.
20. VOGEL, H. J., AND’BONNER, b. i., Microbial Genet. Bull.,
L., in H. NEURATH (Editor), The Proteins, VoZ. II, Academic
13, 43 (1956).
Press, Inc., New York, 1964, p. 305.
21. HATANAKA, M., WHITE, E. A., HORIBATA, K., AND CRAWFORD, 39. ILLINGWORTH, B., JANSZ, H. S., BROWN, D. H., AND CORI, C.
I. P., Arch. Biochem. Biophys., 97, 596 (1962). F., Proc. Natl. Acad. Sci. U. S., 44, 1180 (1958).
22. CRAWFORD, I. P., AND ITO, J., Proc. Natl. Acad. Sci. U. S., 40. CHILSON, 0. P., KITTO, G. B., AND KAPLAN, N. O., Proc. Natl.
61; 390 (1964). Acod. Sci. U. S., 63, 1006 (1965).
Downloaded from www.jbc.org by guest, on October 22, 2011
23. SMITH, O., AND YANOFSKY, C., Methods Enzymol., 6,590 (1963). 41. FRIEDEN, C., J. Biol. Chem., 238. 3286 (1963).
24. SMITH, O., AND YANOFSKY, C., Methods Enzymol., 6,794 (1962). 42. MOYED, H. S., J. Biol. Chem., 236, 1098 (1960).
25. YANOFSKY, C., Methods Enzymol., 2.233 (1956). 43. MATSUSHIRO, A., SATO, K., ITO, J., KIDA, S., AND IMAMOTO,
26. YANOFSKY, C., J. Biol. Chem., 223, 171 (1956). F., J. Mol. Biol., 11, 54 (1965).