Volume 2 number 1 1 November 1 975
Nucleic Acids Research
Specific binding of the first enzyme for histidine biosynthesis to the DNA of the histidine operon Marilyn Meyers, Francesco Blasi*, Carmelo B. Bruni*, Roger G. Deeley, John S. Kovach**, Mark Levinthal***, Kathleen P. Mullinix, Tikvah Vogel, Robert F. Goldberger
Laboratory of Biochemistry, National Cancer Institute, National Institutes of Health, Bethesda, MD 20014, USA Received 30 July 1975
SUMMARY Studies were done to examine direct binding of the first enzyu of the histidine biosynthetic pathway (phosphoribosyltransferase) to Plabeled 80dhis DNA and competition of this binding by unlabeled homologous DNA andTiT various preparations of unlabeled heterologous DNA, including that from a defective *80 bacteriophage carrying the histidine operon with a deletion of part of its operator region. Our findings show that phosphoribosyltransferase binds specifically to a site in or near the regulatory region of the histidine operon. The stability of the complex formed by interaction of the enzyme with the DNA was markedly decreased by the substrates of the enzyme and was slightly increased by the allosteric inhibitor, histidine. These findings are consistent with previous data that indicate that phosphoribosyltransferase plays a role in regulating expression of the histidine operon.
INTRODUCTION The pathway for the biosynthesis of histidine in Salmonella typhi-
murium, elucidated largely through the efforts of Ames and his colleagues
(1-5), consists of a series of ten enzymatic steps. The proteins that catalyze these steps are encoded in a single operon in the bacterial chromosome (6-9). The rate of histidine biosynthesis is regulated by two mechanisms:
feedback inhibition and repression (for reviews, see
10 and 11). Regulation by feedback inhibition involves inhibition of the first enzyme of the biosynthetic pathway, N-1-(5-phosphoribosyl)-
adenosine triphosphate: pyrophosphate phosphori bosyl transferase (E.C. 2.4.2.17, termed phosphoribosyltransferase hereafter) by the end product of the pathway, histidine (3, 12, 13). This inhibition is 2021
Nucleic Acids Research brought about by an allosteric transition (14., 15). Regulation by repression involves changes in the rate at which the histidine operon is expressed (1, 7, 16-19). For some time repression has been known to
require participation of aminoacylated histidine transfer ribonucleic acid (histidyl-tRNA) (20-27).
It has generally been assumed that histidyl-
tRNA acts as a corepressor, fulfilling its regulatory function as a complex formed by interaction with a regulatory protein, the aporepressor
(for reviews, see 10 and 11). Failure to isolate mutants in which a defect in the aporepressor renders the histidine operon constitutive has been taken as evidence that such a regulatory protein, if it exists at
all, may have an additional function vital for survival of the organism
under the conditions in which constitutive mutants were selected (11). In the past few years, evidence from several laboratories has led to the idea that phosphoribosyltransferase may function as a regulatory
protein for the histidine operon. This evidence may be summarized by the following essential points:
(a) When the allosteric site of the enzyme is altered either chemically (12) or by mutation (28), the pattern of repression is altered
(29). In fact, under certain conditions, alteration of the structure of the enzyme completely prevents repression of the histidine operon (30). (b) Mutants have been isolated in which a single mutation in the structural gene for the phosphoribosyltransferase renders the allosteric site of the enzyme completely insensitive to inhibition by histidine and also causes a 4-fold derepressed level of expression of the histidine operon; such constitutive mutations are trans recessive (31). Other constitutive mutants have also been identified as phosphoribosyltransferase mutants (32, 33; Meyers, M., Levinthal, M. and Goldberger, R.,
manuscript in preparation).
(c) Purified wild type phosphoribosyltransferase has been shown to 2022
Nucleic Acids Research bind histidyl-tRNA specifically and with high affinity (34-36).
Purified
phosphoribosyltransferase from a mutant strain in which the enzyme is
insensitive to inhibition by histidine does not bind to histidyl-tRNA
(35, 37). (d) In a purified transcription system, using DNA from a defective transducing bacteriophage carrying the histidine operon of E. coli,
phosphoribosyltransferase from Salmonella typhimurium blocks transcription of the histidine operon specifically (38; Di Nocera, P. P., Avita-
bile, A., and Blasi, F., unpublished data). Analysis of these pieces of evidence makes it clear that it is of interest to examine the interaction of phosphoribosyltransferase with
the DNA of the histidine operon. Using purified phosphoribosyltransferase and 32 P-labeled DNA isolated from a defective *80 transducing bacterio-
phage carrying the histidine operon of Salmonella typhimurium, we demonstrate that the enzyme binds with a high degree of specificity to histidine
operon DNA.
In addition, we present evidence that the enzyme binds to
the regulatory region of the operon.
MATERIALS AND METHODS Purification of phosphoribosyltransferase. Wild type phosphoribosyltransferase was purified from the constitutive strain of Salmonella typhimurium, hisTl504, by the method of Parsons and Koshland (39). The purified enzyme showed a single band on SDS polyacrylamide gel electrophoresis at a
protein load of 100 p'g. The enzyme was stored in a liquid nitrogen freezer as a precipitate formed in anmmonium sulfate. Just prior to use, it was dissolved in 0.05 M Tris-HCl buffer, pH 7.5, containing glycerol (50%, v/v) and
EDTA (5 x
104 M).
Appropriate dilutions of the enzyme for binding experi-
ments were made immediately prior to the experiment in the same buffer.
The
concentration of glycerol in the binding reaction mixtures was kept constant at 2.5%
(v/v) regardless of the amount of enzyme solution added. 2023
Nucleic Acids Research Bacterial strains. The strains used
were
TA1940, constructed by Smith and Tong (40).
the double lysogens, TA1933 and
Strain TA1933 is E. coli K12
deleted for the entire his-gnd region (his-6607) StrR carrying B80h.imwnm c1857 susS7, *80dhis+ inmm c1857 susS7 (referred to hereafter TA1940 is E. coli his-6607 StrR carrying *80h immx c1857 susS7 (referred to hereafter
0c
*80dhis); strain
c1857 susS7, *80dhis01242
inX
*80dhis01242).
In order to be certain that strain TA1940
Growth of bacteria. ing the
as
as
mutation, his01242, single colony isolates
were
was carry-
tested for resis-
tance to the compounds, 3-amino-1,2,4-triazole and 1,2,4-triazole-3-alanine
(TRA). At appropriate concentrations, these compounds inhibit the growth of any
strain in which the histidine
operon
is not constitutively expressed
(41, 42). Preparations of unlabeled phage
carried out by
were
a
modification
of the method of G. Smith and J. Lewis (personal comnunication) in
a
New
Brunswick Magnaferm Fermenter in 10 liter batches. The medium contained 160
g
of Tryptone, 50
were grown
at 320 to
g
an
of NaCl, and 100
g
of yeast extract. The bacteria
O.D.650 of 0.50. The phage
were
raising the temperature to 420 for 20 min. The culture to grow for 3 hrs at 370.
centrifuge, using
a
The cells were harvested in
induced by was
a
then allowed
Beckman J-21
continuous flow rotor (JCFx), and suspended in 40 ml
of 0.1 M Tris-HCl, pH 7.5, containing MgS04 7H20 (0.1 M) and gelatin
(0.1%). The cells
were
lysed by the addition of 2 ml of chloroform and
incubation at 370 for 10 min. The bacterial DNA
was
then digested with
pancreatic DNA'ase (Worthington Biochemical Corp., 0.3 ig/ml final concentration) at 370 for 15 min. The suspension 10,000 was
rpm
then
for 10 min and the sediment discarded.
run on a
was
centrifuged at
The supernatant fluid
cesium chloride block gradient containing 5 ml of 50%
CsCl (bottom), 10 ml of 45% CsCl (middle) and 10 ml of 38% CsCl (top) for 45 min at 35,000 rpm in 2024
a
Type 35 rotor in
a Beckman
L5-65 centri-
Nucleic Acids Research fuge. The phage band, at the interface of the middle and lower layers, was removed, and the refractive index adjusted to n=1.3800. An isopicnic
banding was done at 28,000 rpm for approximately 20 hrs, using a type 65Ti rotor in a Beckman L5-65 centrifuge.
Excellent separation of 080
from *8Odhis was achieved, as judged by optical density and by titering
plaque forming units and the ability to transduce a his recipient to
his+. Each of the bands was collected, without contamination by the other.
The phages were dialyzed exhaustively against 0.01 M Tris-HCl, pH 7.5, containing NaCl (0.1 M) and EDTA (0.001 M). The DNA was extracted by the addition of phenol and SDS, according to the method of Thomas and
Abelson (43). The DNA was then dialyzed against 0.01 M EDTA, pH 8.0,
followed by 0.01 M Tris-HCl, pH 7.4. DNA concentrations were measured spectrophotometrically, using an extinction coefficient of 0.02 cm2/j/g. The 32P-labeled DNA was obtained by growing the strains in 200 ml of Tryptone-yeast extract-NaCl medium that had been depleted of phosphate ions by a modification of the method of Lin and Riggs (44) as follows: 1 ml of 1.0 M MgSO4 and 0.6 ml of concentrated NH40H were added per 100 ml of
medium; the mixture was stirred for several hours and then filtered. The
filtrate was adjusted to pH 7.4 and autoclaved. The growth and isolation procedures were the same as for the unlabeled phage except that 10 mCi of
32P-(H3P04) were added inmediately after the induction period at 420. The phage bands obtained following isopicnic centrifugation were collected using an Isco Density Gradient Fractionator so that complete separation of 080 from O80dhis was achieved.
The specific activity of the DNA obtained by
this procedure was approximately 10,000 cpm/og.
Filter binding assay. The assay for studying the binding of
32p
labeled DNA to phosphoribosyltransferase was based on that used by Riggs et al. (45) for examining lac repressor-operator interactions.
The assay 2025
Nucleic Acids Research depends on the ability of the enzyme to retahn 32P-labeled DNA on a nitrocellulose filter to measure the amount of DNA-phosphoribosyltransferase complex present in the solution being filtered. The filters used were
either Schleicher & Schuell B6 or Millipore HAWP 0.25 P. The filters were pretreated according to the method of Smolarsky and Tal (46).
used were as folldws:
The buffers
(1) Buffer I: 0.066 M potassium cacodylate, pH
6.5, containing MgCl2 (0.010 M) and dimethylsulfoxide (10%, v/v); and Buffer II: 0.033 M potassium cacodylate, pH 6.5, containing MgCl2
(0.005 M) and dimethylsulfoxide (5%, v/v). Dimethylsulfoxide was added to the buffers to reduce nonspecific binding of DNA to the filters (45). The precise conditions for the experiments are described in the legends to the figures and tables. RESULTS
Binding of phosphoribosyltransferase to 32P-O80dhis DNA. Using a fixed level of 32P-080dhis DNA, and varying the amount of enzyme, we found that binding of phosphoribosyltransferase was linear and reached saturation at a
level at which approximately 80% of the DNA was bound to enzyme and therefore retained on the filter. The results of an experiment of this type are
shown in Fig. 1. The addition of bovine serum albumin had no effect on this
interaction. In order to determine to what extent the binding of phosphoribosyl-
transferase to
32P-080dhis
DNA was specific, competition with unlabeled
heterologous or homologous DNA was performed as shown in Fig. 2. Using a fixed level of
32P-080dhis DNA and
an amount of enzyme that was just
barely limiting, increasing amounts of unlabeled DNA from various sources were added.
As expected, when unlabeled
080dhis DNA
was used as the
competitor, the counts bound were reduced to zero. When unlabeled *80 DNA or unlabeled chicken blood DNA
(Cal-biochem)
was used, however, very
little competition was observed, even at a 100-fold excess of unlabeled 2026
Nucleic Acids Research
0
a0
2
z 0 z
0
o
0
*, .25
.5
.75
1.0
G ENZYME ADDED (ig)
Figure 1: Saturation curve with a fixed amount of 32P-080dhis DNA and increasing amounts of phosphoribosyltransferase. The tota'FFeaction was 0.4 ml, containing 0.2 ml of Buffer I and 0.112 ig of yplume 'P-.80dhis DNA. The reaction was started by addition of enzyme and was incubated at 0° for 5 min -- more than twice the time required for the reaction to reach equilibrium. Duplicate aliquots of 0.15 ml from each reaction were filtered, and the filters were washed twice with 0.2 ml of Buffer II. The cpm of a reaction in which no enzyme was added were subtracted from each point to allow for the background level of DNA retained on the filter in the absence of protein. The filters were dissolved in Instabray (Yorktown Research) and counted in a Beckman LS-355 spectrometer. The counts bound represent approximately 80% of the input counts, after correction for the blank. The blank value was no higher than 20% of the input counts.
DNA over labeled DNA. The same results were obtained with several
different preparations of homologous and heterologous DNA. Although unlabeled homologous DNA always competed 100% with the labeled
080dhis
DNA, the amount of competition by unlabeled heterologous DNA was variable. The variation appeared to be dependent upon the preparation of phospho-
ribosyltransferase employed.
Freshly prepared enzyme displayed the
highest degree of specificity, as illustrated by the experiment shown in Fig. 2, whereas enzyme that had been stored for several months displayed a much lower degree of specificity, showing as much as 50% competition
with a 60-fold excess of heterologous unlabeled DNA.
From these experi2027
Nucleic Acids Research 0 0
80
z 2460 -\ D
0
202
4
6
8
UNLABELED DNA ADDED (pg)
Figure 2: Competition with unlabeled homologous and heterologous DNA. Increasing amounts of unlabeled DNA were added to reaction mixtures containing a fixed amount of '4P-080dhis DNA (0.112 pg) and a fixed amount of phosphoribosyltransferase (0.2 ig). The reaction was started by addition of enzyme, as described in the legend for Fig. 1. After incubation for 5 min at 0°, duplicate aliquots were removed, filtered, and counted, as described in the legend for Fig. 1. The counts bound are plotted as a percentage of the amount bound in the tbsence of unlabeled DNA. The unlabeled DNA preparations were: 80dhisO DNA (-I); *80 DNA (0-O); chicken blood DNA (A -); and *80dhis01242 DNA
ments, we conclude that some of the binding of phosphoribosyltransferase to
32P-480dhis DNA is specific for the bacterial genes carried in the phage genome. Phosphoribosyltransferase binding to the regulatory region of the his operon DNA.
Competition experiments like those described above were done
using unlabeled DNA of the phage carrying the histidine operon from which a portion of the operator region (his01242) had been deleted. The data presented in Fig. 2 show that this unlabeled Oc DNA did not compete with
32pP80dhis DNA for binding
to phosphoribosyltransferase.
Several prepa-
rations of Oc DNA gave the same results. Table 1 shows the results of another set of experiments. 32P-480dhis DNA was incubated with phosphoribosyltransferase at QO for 10 min in order 2028
Nucleic Acids Research TABLE 1 Effect of 080 DNA,
W80dhisO+ DNA,
and
W8OdhisOC
DNA on Binding of 32P-*80dhisO0
DNA to Phosphoribosyltransferase* Additions
% Bound
.01 M Tris-HCl pH 7.4
74
Chicken Blood
73
080 DNA
67
*80dhisO1242 DNA
85
W8OdhisO DNA
4
*The total reaction volume was 3.0 ml, cootaining 1.5 ml of Buffer I, 0.55 jg of phosphoribosyltransferase, 0.84 jig of 3'P-+80dhis DNA, and 1.25 ml of H 0. After incubation at 00 for 10 min, a 20-folWixcess of unlabelled DNA wJs added and duplicate aliquots were removed and filtered 6 min later. Each value was corrected for a blank value determined by filtering aliquots prior to the addition of enzyme. to allow complex formation to be completed.
Unlabeled DNA (20-fold excess)
that had been sheared by passage through a No. 27 needle (five times) was added and the amount of complex remaining was determined by filtering the
incubation mixtures after 6 min. The results showed that when 80dhisO DNA was added only 4% of the complex remained intact; whereas, when 080 DNA or
f80dhis01242
DNA was added 67% or 85%, respectively, of the complex
remained intact. We conclude that the enzyme binds to a sequence of the histidine operon DNA at least part of which has been deleted from his01242. Effect of histidine, phosphoribosylpyrophosphate, and ATP on the
stability of the phosphoribosyltransferase-DNA complex. The first step in the biosynthesis of histidine is the conversion of phosphoribosylpyrophosphate and ATP to phosphoribosyl-ATP via the phosphoribosyltransferase (3). 2029
Nucleic Acids Research The enzyme, as previously discussed, is allosteric and subject to feedback
inhibition by histidine (3, 12, 13).
It was, therefore, of interest to
examine the effects of these ligands on the complex formed between the enzyme and 80dhis DNA.
The experiments were done as follows: 32P-+80dhis
DNA was incubated with enzyme at QO for 10 min in order to allow complex
formation to be completed. Unlabeled +80 DNA (20-fold excess) that had been sheared by passage through a No. 27 needle (five times) was added alone or together with either histidine (final concentration 5 x 10 4 M) or ATP and
phosphoribosylpyrophosphate (final concentration 5 x 10 4 M each). The amount of complex remaining was determined by filtering the incubation
mixtures after 6 min. The results (Table 2) showed that in the presence of the substrates of phosphoribosyltransferase only 11% of the complex remained
intact, whereas histidine may have slightly increased the amount of complex remaining intact.
TABLE 2 Effect of Ligands for Phosphoribosyltransferase on Binding of
32P-+80dhis
DNA to the Enzyme*
Additions
% Bound
None
63
Histidine
80
ATP and phosphoribosylpyrophosphate
11
*The total reaction volume was 3.3 ml., containing 1,65 ml. of Buffer I, 0.6 .g of phosphoribosyltransferase and 0.93 ig of P-+80dhis DNA. After incubation at QO for 10 min, unlabelled *80 DNA (20-fold Xxcs) was added either alone, with histidine (final concentration 5 x 10~ M), o0 with ATP and phosphoribosylpyrophosphate (final concentration 5 x 10- M each). Duplicate aliquots were removed and filtered 6 min later. Each value was corrected for a blank value determined by filtering aliquots prior to the addition of enzyme. 2030
Nucleic Acids Research DISCUSSION Though much has been learned through the elegant studies of Ames and Hartman and their colleagues about the pathway for histidine biosynthesis,
the genetics of this system, and regulation of expression of the histidine operon (for review, see 10), the molecular mechanism of the repression process is still not fully understood.
It is clear that repression of the
histidine operon requires participation of the corepressor, histidyl-tRNA
(20-27). However, no regulatory protein that functions as a classical aporepressor has been identified (for review, see 11).
Phosphoribosyltransferase has been the subject of intensive inves-
tigation in a number of laboratories for more than a decade. A large number of studies has demonstrated that this allosteric, hexameric enzyme
regulates the flow of substrates into the pathway for histidine biosynthesis in vivo through the mechanism of feedback inhibition (3, 12, 13, 15, 28, 35,
47, 48).
In addition, the enzyme has been implicated as a regulatory pro-
tein that modifies expression of the histidine operon. Previous studies have shown that under certain conditions, alteration of this enzyme prevents
repression from occurring in vivo (29, 30); that the enzyme interacts specifically and with high affinity with the aminoacylated form of histidine tRNA in vitro (34-36); that certain trans recessive mutations in the structural gene for the enzyme result in expression of the histidine operon that is
somewhat derepressed and is unresponsive to the availability of histidine
(31); and that the enzyme inhibits transcription of the E. coli histidine operon specifically in vitro (38; Di Nocera, P. P., Avitabile, A. and Blasi, F., unpublished data). The availability of a defective *80 transducing bacteriophage carrying the histidineoperon of Salmonella typhimurium (40,
49) made it possible for us to further investigate the role of phosphoribosyltransferase as a regulatory protein by studying the physical interaction between the enzyme and histidine operon DNA in vitro. 2031
Nucleic Acids Research Our initial studies demonstrate that phosphoribosyltransferase binds to
32P-.80dhis DNA. Using a fixed amount of DNA and varying the amount of enzyme, we found that binding of phosphoribosyltransferase was linear and reached saturation at a level at which approximately 80% of the DNA was bound (Fig. 1).
It should be noted that the specific binding of phosphori-
bosyltransferase to the DNA of the histidine operon reported here occurs in the absence of added histidine-tRNA. This finding is consistent with the results of Blasi et al. (38) which demonstrated that phosphoribosyltransferase
inhibits transcription of the histidine operon in vitro in the absence of added histidyl-tRNA.
In order to investigate the question of whether the binding of phos-
phoribosyltransferase to 480dhis DNA was specific, we studied the ability of
heterologous DNA to compete with the
his+
DNA in the binding reaction. We
found that chicken blood DNA did not compete at all. A more stringent test,
however, was the use of *80 DNA, in which most of the sequences are identical with *80dhis DNA. The fact that *80 DNA was not able to compete (Fig. 2 and Table 1) demonstrates that the enzyme interacted with that portion of
48Odhis not present in the wild type phage, namely the bacterial DNA carried in its genome. To further define the site of interaction of phosphoribosyltrans-
ferase, we used, as competitor in the binding reaction, DNA from a phage
carrying the histidine operon from which a portion of the his operator had been deleted (his01242). The finding that this his01242 DNA was not able to compete with his DNA (Fig. 2 and Table 1) demonstrates that the nucleotide sequence to which the enzyme specifically binds includes part or all of that sequence deleted in his01242. An alternative interpretation is
that the phosphoribosyltransferase binds to a segment of DNA near the portion deleted in his01242, and that the deletion results in an altered
conformation of this segment that prevents interaction with the enzyme. 2032
Nucleic Acids Research Thus, the DNA-binding site could be the operator-promoter region, a
leader sequence like that found in the tryptophan operon (50), or the structural gene specifying the phosphoribosyltransferase itself, as has been proposed for the first enzyme for tryptophan biosynthesis by Somerville and Stetson (51).
The stability of the phosphoribosyltransferase-DNA complex was also
investigated in the presence of various ligands for the enzyme. The stability of the complex was significantly decreased by the presence of the substrates for the enzyme, ATP and phosphoribosylpyrophosphate, and may have
been slightly increased by the presence of the allosteric inhibitor for the enzyme, histidine (Table 2).
It is interesting to note that the substrates
were previously found to destabilize the complex formed between phosphori-
bosyltransferase and histidyl-tRNA, while histidine was found to have no effect on the stability of the enzyme-tRNA complex (35).
If phosphori-
bosyltransferase acts as a negative regulator for the histidine operon, then one might expect, on the basis of physiological considerations, that under conditions in which increased expression of the histidine operon was required,
phosphoribosyltransferase would be most active as an enzyme and least active as a regulator; conversely, under conditions in which decreased expression
of the operon was required, phosphoribosyltransferase would be least active as an enzyme and most active as a regulator.
Our findings on the effects of
the ligands on the enzyme-DNA complex are in keeping with such a model,
though they do not permit us to specify the model in greater detail. Nor do these considerations eliminate the possibility that the enzyme also may act as a positive regulator for the histidine operon.
The finding that the
substrates of phosphoribosyltransferase influence the interaction between the enzyme and DNA indicates that the form of the enzyme active with respect to DNA binding must also be active with respect to substrate binding.
known from the work of Parsons and Lipsky (personal
It is
communication) that 2033
Nucleic Acids Research phosphoribosyltransferase can exist, under physiological conditions, in a number of different aggregation states, all of which bind the substrates.
Any of these could therefore be candidates for the regulatory form of the enzyme.
Examining the stoichiometry of the interaction between phosphoribosyltransferase and *80dhis DNA, it is clear that only about 2% of the enzyme is
capable of binding DNA. We have considered the possibility that the DNAbinding activity is due to a contaminant of the enzyme preparation, but
have rejected this possibility for the following reasons: First, the procedure by which the enzyme was purified for the studies reported here
depends upon the fact that the enzyme displays an unusual and specific
change in solubility in ammonium sulfate solutions, depending upon the presence of its allosteric inhibitor, histidine (39); second, phosphori-
bosyltransferase, purified by an entirely different procedure, involving chromatography on DEAE-cellulose and hydroxyapatite (15), also displayed
specific binding to *80dhis DNA (unpublished results); and third, the protein involved in the specific binding to *80dhis DNA must have sites for
binding the substrates of phosphoribosyltransferase, since these substrates have a dramatic effect on the stability of the protein-DNA complex.
Consistent with the idea that only a portion of the phosphoribosyltransferase is active with respect to specific DNA binding is the suggestion of Parsons and Lipsky (personal conmnunication) that the species of the enzyme that is active in regulation is the dimer that is present in small amounts under physiological conditions.
In keeping with this idea are the
facts that the substrates of phosphoribosyltransferase both favor aggrega-
tion of the enzyme (diminishing the proportion of dimers) and destabilize the enzyme-DNA complex.
*Present address: Institute of General Pathology, via Sergio Pansini, 1-80131 Naples, Italy 2034
University of Naples 2nd Medical School,
Nucleic Acids Research
**Present address: Department of Medicine, Columbia University College of Physicians and Surgeons, New York, NY 10032, USA ***Present address: Department of Biological Sciences, IN 47907, USA
Purdue University, Lafayette,
REFERENCES 1. Ames, B. N., Garry, B. and Herzenberg, L. A. (1960). J. Gen. Microbiol. 22, 369-378. 2. Moyed, H. S. and Magasanik, B. (1960). J. Biol. Chem. 235, 149-153. 3. Ames, B. N., Martin, R. G. and Garry, B. (1961). J. Biol. Chem. 236, 20192026. 4. Smith, D. W. E. and Ames, B. N. (1964). J. Biol. Chem. 239, 1848-1855. 5. Smith, D. W. E. and Ames, B. N. (1q65). J. Biol. Chem. 240, 3056-3063. 6. Hartman, P. E., Hartman, Z. and S6rman, D. (1960). J. Gen. Microbiol. 22, 354-368. 7. Ames, B. N. and Hartman, P. E. (1963). Cold Spring Harbor S Quant. Biol. 28, 349-356. 8. Loper, J. C., Grabnar, M., Stahl, R. C., Hartman, Z. and Hartman, P. E. (1964). Brookhaven . Biol. 17, 15-52. 9. Hartman, P. E., Hartman, Z., Stahl, R. C. and Ames, B. N. (1971). Advan. Genet. 16, 1-34. 10. Brenner, M. and Ames, B. N. (1971). In Metab. Pathways, Vol. 5 (Vogel, Henry J., ed.), p. 349, Academic Press, New York. 11. Goldberger, R. F. and Kovach, J. S. (1972). Curr. . Cell. Regul. 5, 285-308. 12. Moyed, H. S. (1961). J. Biol. Chem. 236, 2261-2267. 13. Martin, R. G. (1963). J.Bi3. Chem.28, 257-268. 14. Blasi, F., Aloj, S. M. and Goldberger, R. F. (1971). Biochemistry, 10, 1409-1417. 15. Whitfield, H. J., Jr. (1971). J. Biol. Chem. 246, 899-908. 16. Ames, B. N. and Garry, B. (1959T. Proc. Nat. cad. Sci., U.S.A. 45, 1453-1461. 17. Ames, B. N. and Hartman, P. E. (1962). In The Molecular Basis of Neoplasia, pp. 322-345, Univ. of Texas Press, Austi 18. Ames, B. N., Hartman, P. E. and Jacob, F. (1963). J. Mol. Biol. 7, 23-42. 19. Venetianer, P. (1969). J. Mol. Biol. 45, 375-384. 20. Schlesinger, S. and Magasanik, B.1T964T. J. Mol. Biol. 9, 670-682. 21. Roth, J. R. and Hartman, P. E. (1965). Yirology, 27, 297-307. 22. Roth, J. R. and Ames, B. N. (1966). J. Mol. Biol.T2, 325-334. 23. Roth, J. R., Ant6n, D. N. and Hartman, P. E. (1966). J. Mol. Biol. 22, 305-323. 24. Roth, J. R. and Sanderson, K. E. (1966). Genetics, 53, 971-976. 25. Silbert, D. F., Fink, G. R. and Ames, B. N. (1966). J. Mol. Biol. 22, 335-347. 26. Anton, D. N. (1968). J. Mol. Biol. 33, 533-546. 27. Lewis, J. A. and Ames, B.N. (197). J. Mol. Biol. 66, 131-142. 28. Sheppard, D. E. (1964). Genetics, 50, 611x3.29. Kovach, J. S., Berberich. M. A.; Venetianer, P. and Goldberger, R. F. (1968). J. Bacteriol. 97, 1283-1290. 30. Kovach, J. S.Phang,J. M., Ference, M. and Goldberger, R. F. (1969). Proc. Nat. Acad. Sci., U.S.A. 63, 481-488. 31. Kovach,. S., Balesteros,-A.6., Meyers, M., Soria, M. and Goldberger, R. F. (1973). J. Bacteriol. 114, 351-356. 2035
Nucleic Acids Research 32. Patthy, L. & D4nes, G. (1970). Biochim. pBiophys. Acta, 5, 147-157. 33. Rothman-Denes, L. and Martin, R. G. (1971). J. BacterioT. 106, 227-231. 34. Kovach, J. S., Phang, J. M., Blasi, F., Barton, R. W., BallesterosOlmo, A. and Goldberger, R. F. (1970). J. Bacteriol. 104, 787-792. 35. Blasi, F., Barton, R. W., Kovach, J. S. and Goldbger, R. F. (1971). J. Bacteriol. 106, 508-513. 36. Vogel, T., Meyers, M., Kovach, J. S. and Goldberger, R. F. (1972). J. Bacteriol. 112, 126-130. 37. Smith, 0., Meyers, M. and Vogel, T., Deeley, R. G. and Goldberger, R. F. (1974). Nucleic Acids Res. 1, 881-888. 38. Blasi, F., Bruni, C., AvitabTle, A., Deeley, R. G., Goldberger, R. F. and Meyers, M. (1973). Proc. Nat. Acad. Sci., U.S.A. 70, 2692-2696. 39. Parsons, S. M. and Koshland, D.T., Jr. (T74). J.7BiqT. Chem. 249, 4104-4109. 40. Smith, G. R. and Tong, B. (1974). J. Bacteriol. 120, 1223-1226. 41. Hilton, J. L., Kearney, P. S. and Aiiies, B. N. (19353. Arch. Biochem. Biophys. 112, 544-547. 42. Levin, A. P. and Hartman, P. E. (1963). J. Bacteriol. 86, 820-828. 43. Thomas, C. A. and Abelson, J. (1966). In Procedures inFNucleic Acid Research (Cantoni, G. L. and Davies, D. R., eds.), pp. 53T-561, Harper and Row, New York. 44. Lin, S. and Riggs, A. D. (1972). J. Mol. Biol. 72, 671-690. 45. Riggs, A. D., Suzuki, H. and Bourgeois, S. (1970} J. Mol. Biol. 48, 67-83. 46. Smolarsky, M. and Tal, M. (1970). Biochim. Biophys. Acta, 199, 447-452. 47. Atkinson, D. E., Hathaway, J. A. and Smith, CC. (1965). J. Biol. Chem. 240, 2682-2690. 48. Bell, R. M., Parsons, S. M., Dubravac, S. A., Redfield, A. G. and Koshland, D. E., Jr. (1974). J. Biol. Chem. 249, 4110-4118. 49. Voll, M. J. (1972). J. Bacteriol. 109, 74F750. 50. Jackson, E. N. and Yanofsky, C. (1973. J. Mol. Biol. 76, 89-101. 51. Somerville, R. L. and Stetson, H. (1974). Moec. Gen. Genet. 131, 247261.
2036