The Overall Synthesis of ~-5,6-Dihydroorotate by Multienzymatic Protein pyrl-3 from Hamster Cells KINETIC STUDIES, SUBSTRATE CHANNELING, AND THE EFFECTS OF INHIBITORS*

In mammals, a trifunctional protein (ME pyrl-3) syn- thesizes ~-5,6-dihydroorotate in three sequential reactions catalyzed by carbamyl phosphate synthetase (EC 2.7.2.9), aspartate transcarbamylase (EC 2.1.3.2), and dihydroorotase (EC 3.5.2.3). 14C-labeled HC03- has been used as a precursor for the synthesis of L-5,Cdihydro-orotate by purified ME pyrl-3, and when this product is conve7rted enzymatically to orotidine 5’-monophos-phate, the concentrations of the two intermediates of ME pyrl-3, carbamyl phosphate, and N-carbamyl-L-as-partate, reach steady state concentrations of approxi- mately 0.20 PM and 7.1 p ~ , respectively. At pH 7.4 in the presence of 0.1 lll~ 5-phosphoribosyl 1-pyrophosphate and 10% (v/v) glycerol, pure ME pyrl-3 has a Michaelis constant for HC03- of 0.61 lll~ and a maximal specific activity of 329 pmol of ~-5,6-dihydroorotate synthe- sized/min/pg, equivalent to a turnover number of 65.8 mol min-’ (mol of subunit)”. Consideration of the K,,, and V,, values of aspartate transcarbamylase and dihydroorotase determined un- maximal activity of carbamyl phosphate synthetase, endogenous as- partate transcarbamylase activity was used to convert carbamyl phosphate synthesized to N-carbamyl-L-aspartate et al., 1977), some of which was cyclized to L-5.6-dihydroorotate (see “Results”). Assay (v/v) MgCI2, L-glutamine, PRPP, phosphoenolpy- ruvate, milliunits pyruvate L-aspartate, Na[I4C]HCO3 Ci/mol) the assay was initiated by addition of appropriate dilutions of ME pyrl-3 to incubation mixtures. Samples of 5 pl were spotted onto PEI-C maintained at 4”C3 at 2,4, and 6 min and chromatograms were developed with 0.19 M LiCl, 0.1% (v/v) formic at 4’C3 and I4C-labeled N-carbamyl-L-aspartate and L- dihydroorotate were detected by autoradiography and quantitated by scintillation counting. Under these conditions of assay, [14C]carbamyl at relatively low concentrations (see “Re-sults”) and not quantitated. K,, value of assay mixtures for carbamyl phosphate synthetase made up as with appropriate concentrations of [I4C]HC03- (52.4 Ci/mol) standardized described The reaction was initiated by addition of 1.85 pg of hamster ME pyrl-3 and 5-pl window at 32.17%. The ratio of counts in the lower 3H window to counts in the upper I4C window when I4C was counted was 9.915 X IO-' and an inverse ratio of 4.356 X was obtained when 3H was counted indicating a minimal spillover of 3H counts into the I4C window. Using these values and the known specific activities of I4C (52.4 Ci/mol) and 3H (58.0 Ci/mol), a program was devised for a Hewlett-Packard HP-19C calculator which calculated the micromolar concentrations of all "C- and 3H-labeled intermediates. The contribution of endogenous N-[14C]carbamyl-~-aspartate to the synthesis of L-dihydroorotate (meas- ured as OMP) at various concentrations of exogenous N-carbamyl-~-[~H]aspartate determined and ~-5,6-dihydroorotate were determined at 37°C in 50 mM KHepes, pH 7.4, 10% (v/v) glycerol, and 2.0 mM dithiothreitol as described under “Experimental Procedures.” Initial reaction velocities were determined from a least squares fit to 3 time points for product formation using at least seven different substrate concentrations and each K, value was calculated from a least squares fit of the resultant data to a Lineweaver-Burk plot. Coefficients of determination so obtained were better than 0.99. using final concentrations of purified ME pyrl-3 of 12.9 and 24.6 pg/pl, respectively, in 50 m~ KHepes, pH 10% (v/v) glycerol, and 2.0 m~ dithiothreitol as described under “Experimental Procedures.” Initial reaction velocities were deter- mined from a least squares fit to 3 time points for product formation using at least six concentrations of carbamyl phosphate or N-carba- myl-L-aspartate and and K,,, values were calculated from a of the to

In mammals, the fist three reactions of the de novo pyrimidine pathway are catalyzed by a trifunctional protein, ME pyrl-3,' which contains the enzymatic activities carbamyl phosphate synthetase (EC 2.7.2.9), aspartate transcarbamylase (EC 2.1.3.2), and dihydroorotase (EC 3.5.2.3). Sequential operation of these three activities results in the synthesis of ~-5,6-dihydroorotate from MgATP, HCOs-, L-glutamine, and L-aspartate. The effect of changes in the substrate concentrations on the kinetics of each of the activities of ME pyrl-3 have been characterized individually (Tatibana and Shigesada, 1972b;Shoaf and Jones, 1973;Hoogenraad, 1974;Jones, 1979, 1980), but the overall synthesis of ~-5,6-dihydroorotate by this trifunctional protein, when its three enzymatic activities are functioning in concert, has not been studied. Chen and Jones (1979) presented evidence that in vivo carbamyl phosphate synthetase is usually the rate-limiting enzyme for the de nouo biosynthesis of UMP. This enzyme is subject to regulation by MgPRPP, a positive effector, and MgUTP, a negative effector , enabling regulation of the flux through the de novo pyrimidine pathway by modulation of carbamyl phosphate synthetase activity. Thus, characterization of the sequential synthesis of carbamyl phosphate, N-carbamyl-L-aspartate and ~-5,6-dihydroorotate by MEpyrl-3 and the effects of inhibitors upon the overall synthesis of ~-5,6-dihydroorotate is important to an understanding of the regulation of the de novo pathway in vivo.
Carbamyl phosphate synthetase and aspartate transcarbamylase exist as an enzyme complex in Neurospora crassa (Williams et aL, 1970), Saccharomyces cerevisiae (Lue and Kaplan, 1971) and in mammals  and Shoaf and Jones (1971) demonstrated that the mammalian protein also contains the third activity of the pathway, dihydroorotase. A large number of multifunctional proteins have now been characterized (Schmincke-Ott and Bisswanger, 1980) and one of the possible advantages conferred by such a n association of enzymes catalyzing sequential reactions is that an intermediate synthesized at the first catalytic site may be preferentially utilized as substrate by a subsequent site on the same protein molecule. Such "substrate channeling" (Davis, 1967) would allow synthesis of the final product of t h e multienzymatic protein without significant accumulation of the intermediate (s). Various degrees of substrate channeling have been demonstrated for the carbamyl phosphate synthetase-aspartate transcarbamylase of yeast (Lue and Kaplan, 1970) and for a number of other multienzymatic proteins Fall and West, 1971;Matchett, 1974;Welch and Gaertner, 1975;Heyde, 1979) including UMP synthase* (Traut and Jones, 1977), which catalyzes the fifth and sixth reactions of de nouo pyrimidine biosynthesis. More quantitative aspects of the general concept of channeling have been discussed by Duggleby et al. (1978).
We have developed an overall radioassay for the fist three reactions of de nouo pyrimidine biosynthesis3 which enables simultaneous measurement of the concentrations of carbamyl phosphate, N-carbamyl-L-aspartate, and the final product of ME pyrl-3, ~-5,6-dihydroorotate. A "coupling enzyme mixture" can be used to convert ~-5$-dihydroorotate to OMP to overcome the unfavorable equilibrium existing between ~-5,6dihydroorotate and N-carbamyl-L-aspartate in the physiological range of pH (Christopherson and Jones, 1979). Using this assay procedure, we find that carbamyl phosphate and Ncarbamyl-L-aspartate are maintained at low steady state concentrations due to partial channeling of these intermediates and also to the favorable kinetic characteristics ( Vmax/Km) of aspartate transcarbamylase and dihydroorotase relative to carbamyl phosphate synthetase. The effectiveness of PALA and L-cysteine as inhibitors of aspartate transcarbamylase and dihydroorotase, respectively, has been tested when ME pyrl-3 is synthesizing ~-5,6-dihydroorotate using HCOC as a precursor.

EXPERIMENTAL PROCEDURES
PALA was obtained from Mr. Leonard Kedda of the Drug Synthesis and Chemistry Branch of the National Cancer Institute and was 79.5% pure by weight. L-Cysteine (free base), 6-azauridine 5'-monophosphate, L-glutamine, L-aspartate, dipotassium ATP, dipotassium phosphoenolpyruvate, NAD', and PRPP were obtained from the Sigma Chemical Co. Acetaldehyde was from Mallinckrodt. All other chemicals were of the best grade commercially available. PEI-C chromatograms (20 X 20 cm) were manufactured by Machery, Nagel and Co. and obtained from Brinkmann Instruments, Inc. Chromatograms were washed (Christopherson et ab, 1978) and scored into 1.5cm strips, and the origin was marked with soft pencil 2 cm from the base.
Hamster ME pyrl-3 purified by the procedure of Coleman et al. (1977) was a kind gift from Dr. George Stark of Stanford University. Additional ME pyrl-3 was purified by the same procedure from a mutant hamster cell line (165-23) obtained from Dr. Stark which overproduces ME pyrl-3 by more than 100-fold. The crude dialyzed extract of mouse Ehrlich ascites tumor cells used to demonstrate carbamyl phosphate phosphatase activity was prepared in a solution of 20 mM Tris-HC1, pH 7.4, 2.0 mM dithiothreitol as described by Traut (1980). The following enzyme preparations were obtained from the Sigma Chemical Co.: dihydroorotate dehydrogenase (1.5 units/ mg) from Clostridium oroticum, horse liver alcohol dehydrogenase (1.9 units/mg), yeast orotate phosphoribosyltransferase (182 units/ mg), and yeast inorganic pyrophosphatase (9.5 units/mg). Rabbit muscle pyruvate kinase (200 units/mg) was from Boehringer Mannheim. For determination of protein concentrations, the protein was fvst precipitated with 5% (w/v) trichloroacetic acid to remove interfering compounds and then assayed by a modification of the Lowry procedure described by Hartree (1972 To measure the decreasing concentration of [14C]HC03-in incubation mixtures, the assay for carbamyl phosphate synthetase activity was carried out as described below using K[I4C]HC03 (0.1 Ci/mol) except that MEpyrl-3 was omitted. Samples of 5 pI were transferred to scintillation vials containing 500 pl of 10 m~ Tris-HCI, pH 9.2, and the vials were immediately sealed with a rubber serum cap holding a plastic center well with a paper wick which had been soaked in 2 M NaOH (Prabhakararao and Jones, 1975 give an aspartate transcarbamylase activity of 4.2 pmol/min under the standard conditions of assay described below. The assay mixtures were incubated in sealed tubes at 28'C and 5-pl samples were spotted onto PEI-C chromatograms at 50, 6 0 , and 70 min. The 14C-labeled N-carbamyl-L-aspartate may be separated from ~-["C]aspartate and ~-['~C]dihydroorotate by chromatography with 0.19 M LiCI, 0.75% (v/v) formic acid at room temperature3 and ',Clabeled N-carbamyl-L-aspartate and ~-['~C]dihydroorotate were cut out and quantitated by scintillation counting? The concentration of carbamyl phosphate was calculated by the method of least squares from the slope of a plot of microliters of carbamyl phosphate versus counts per min in the products. Standardized carbamyl phosphate solutions were stored as small aliquots at -70°C and used once.
Assay of Carbamyl Phosphate Synthetase-To determine the maximal activity of carbamyl phosphate synthetase, endogenous aspartate transcarbamylase activity was used to convert carbamyl phosphate synthesized to N-carbamyl-L-aspartate (Coleman et al., 1977), some of which was cyclized to L-5.6-dihydroorotate (see "Results"). Assay mixtures contained in a total volume of 25 pl 50 m~ KHepes, pH 7.4, 10% (v/v) glycerol, 2.0 m~ dithiothreitol, 10 m~ ATP, 15 mM MgCI2, 4.0 mM L-glutamine, 0.1 m~ PRPP, 10 mM phosphoenolpyruvate, 75 milliunits of pyruvate kinase, 5.0 mM L-aspartate, 3.26 mM Na[I4C]HCO3 (52.4 Ci/mol) and the assay was initiated by addition of appropriate dilutions of ME pyrl-3 to incubation mixtures. Samples of 5 pl were spotted onto PEI-C maintained at 4"C3 at 2,4, and 6 min and chromatograms were developed with 0.19 M LiCl, 0.1% (v/v) formic acid at 4'C3 and I4C-labeled N-carbamyl-L-aspartate and Ldihydroorotate were detected by autoradiography and quantitated by scintillation counting. Under these conditions of assay, [14C]carbamyl phosphate was present at relatively low concentrations (see "Results") and was not quantitated. For determination of the K,, value of HC03-, assay mixtures for carbamyl phosphate synthetase were made up as described above with appropriate concentrations of [I4C]HC03-(52.4 Ci/mol) standardized as described above. The reaction was initiated by addition of 1.85 pg of hamster ME pyrl-3 and 5-pl samples were spotted onto PEI-C chromatograms maintained at 4°C at 1,2, and 3 min. Isolation and quantitation of I4C-labeled N-carbamyl-L-aspartate and ~-5,6dihydroorotate were as described above.
Carbamyl phosphate synthetase activity was calculated by the method of least squares from plots of product formed versus time. The specific activity of carbamyl phosphate synthetase was calculated by the method of least squares from the slope of a plot of picomoles/ min of product synthesized versus the concentration of ME pyrl-3. These calculation procedures were also used for determination of aspartate transcarbamylase and dihydroorotase activities.
Assay of Aspartate Transcarbamylase-To determine the maximal activity of aspartate transcarbamylase, a modification of the procedure of Bethell et al. (1968) was used. Assay mixtures contained in a total volume of 80 pl 50 m~ KHepes, pH 7.4, 10% (v/v) glycerol, 2.0 m~ dithiothreitol, 2.0 mM ['4C]carbamyl phosphate (2.0 Ci/mol), and 5.0 mM L-aspartate. The assay was initiated by addition of four different dilutions of ME pyrl-3 to incubation mixtures and 20-p1 samples were transferred into 180 p1 of 0.37 M HC10, in a Biovial (Beckman Instruments Inc.) at 2, 4, and 6 min. Samples were processed and the specific activity of aspartate transcarbamylase was calculated as described above for carbamyl phosphate synthetase.
For determination of the K, value of carbamyl phosphate, assay mixtures for aspartate transcarbamylase contained in a total volume of 6.2 rnk 50 mM KHepes, pH 7.4, 10% (v/v) glycerol, 2.0 m~ dithiothreitol, 5.0 mM L-aspartate, appropriate concentrations of [I4C]carbamyl phosphate (10.4 Ci/mol) standardized directly by chromatography as described above, and the reaction was initiated by addition of 23 ng of hamster MEpyrl-3. Samples of 2.0 ml of the assay mixture were transferred into 58.0 pl of 11.7 M HC1O4 in scintillation vials at 2, 4, and 6 min. The samples were heated in a boiling water bath for 10 min and allowed to stand overnight prior to addition of 18.0 ml of toluene/Triton X-100 scintillant for counting.
Assay of Dihydroorotase-To determined the maximal activity of dihydroorotase, the procedure of Christopherson et al. (1978) was used. Assay mixtures contained in a total volume of 25 pl: 50 m M KHepes, pH 7.4, 10% (v/v) glycerol, 2.0 m~ dithiothreitol, 5.0 m~ N -[14C]carbamyl-~-aspartate (10 Ci/mol) and the reaction was initiated by addition of 4 different dilutions of ME pyrl-3 to incubation mixtures. Samples of 5 pl were spotted onto PEI-C at 2, 4, and 6 min and the ~-['~C]dihydroorotate synthesized was isolated by chromatography and quantitated as described by Christopherson et al. (1978).
Assay of the Overall Reaction Catalyzed by ME pyrl-3-To determine the overall activity of MEpyrl-3, a procedure described in detail elsewhere3 was used. Briefly, assay mixtures contained in a total volume of 50 d: 50 mM KHepes, pH 7.4, 10% (v/v) glycerol, 2.0 mM dithiothreitol, 10 mM ATP, 15 m~ MgC12, 3.26 m~ Na['4C]HC03 (52.4 Ci/mol), 4.0 mM L-glutamine, 0.2 m~ PRPP, 10 mM phosphoenolpyruvate, 150 milliunits of pyruvate kinase, 5.0 mM aspartate^ and the assay was initiated by addition of 0.93 pg of hamster ME pyrl-3. When required, the final product of MEpyrl-3, L-5,6-dihydroorotate, was converted enzymatically to OMP by inclusion of the following "coupling enzyme mixture3" in the total assay volume of 50 pl 10 mM phosphoenolpyruvate, 150 milliunits of rabbit pyruvate kinase, 10 m~ NAD', 63 milliunits of dihydroorotate dehydrogenase, 10 mM acetaldehyde, 5.0 milliunits of equine alcohol dehydrogenase, 1.0 mM PRPP, 300 milliunits of yeast orotate phosphoribosyltransferase, 50 PM 6-azauridine 5'-monophosphate, and 20 milliunits of yeast inorganic pyrophosphatase. Samples of 5 1-11 were spotted into PEI-C chromatograms maintained at 4°C. Radiolabeled carbamyl phosphate, N-carbamyl-L-aspartate, ~-5,6-dihydroorotate, orotate, and OMP were separated by ascending chromatography at 4°C using 0.19 M LiCl, 0.1% (v/v) formic acid as developing solvents.' Vials of concentrated NaOH were included in the chromatography tank to trap residual I4CO2. Detection of 14C-labeled intermediates by autoradiography and their quantitation by scintillation counting have been described elsewhere." Competition between Carbamyl Phosphate Synthesized by ME pyrl-3 and Exogenous Carbamyl Phosphate for the Aspartate Transcarbamylase Reaction-To study possible substrate channeling of carbamyl phosphate by ME pyrl-3, endogenous carbamyl phosphate synthesized from HCO3-of a particular radiolabel (*) was competed with exogenous carbamyl phosphate of a different radiolabel (*) according to the following scheme: A second set of incubation mixtures of opposite radiolabeling contained unlabeled 3.26 m~ HC03-as precursor for endogenous carbamyl phosphate and ["Clcarbamyl phosphate (10.4 Ci/mol) at the same concentrations used in the first set of incubation mixtures. Reactions were initiated by addition of 56 ng of hamster ME pyrl-3 to give a total volume of 50 pl and 5 4 samples were spotted onto PEI-C chromatograms maintained at 4"C3 at 2, 4, 6, and 8 min. Carbamyl phosphate, N-carbamyl-L-aspartate, and L-dihydroorotate were separated by Chromatography for 18 cm with 0.5 M LiC1, 0.1% (v/v) Fig. 4A and Christopherson and Jones, 1979), it was not included in calculations for determination of the contributions of endogenous and exogenous carbamyl phosphate to the aspartate transcarbamylase reaction. The Na[I4C]HC03 used in the first set of incubations was standardized as described above and solutions of unlabeled NaHC03 used in the second set were prepared on the same day from the solid. The unlabeled carbamyl phosphate K+ ion was present as a counterion of a number of assay components giving a total concentration of approximately 90 m~. used in the first set of incubations and the ['4C]carbamyl phosphate used in the second set were standardized concurrently against L-["Claspartate as described above. Because conditions for both sets of incubations were identical except for the radiolabeling, the contribution of endogenous carbamyl phosphate to the aspartate transCubaphosphate could be determined.

mylase reaction at various concentrations of exogenous carbamyl
Competition between N-Carbamyl-L -aspartateSynthesized by ME pyrl-3 and Exogenous N-Carbamyl-L -aspartate for the Dihydroorotase Reaction-To determine if N-carbamyl-L-aspartate synthesized by ME pyrl-3 is utilized preferentially compared with exogenous N- For clarity, ~-5,6-dihydroorotate and orotate have been omitted from the scheme above, since the coupling enzyme mixture converted most of the ~-5,6-dihydroorotate synthesized to OMP.3 Incubation mixtures contained all the components described above for assay of the overall reaction including 3.26 mM [I4C]HC03-(52.4 Ci/mol) and various concentrations of N-carbamyl-~-[~H]aspartate (0 to 1.0 mM, 58.0 Ci/ mol). Reactions were initiated by addition of 93 ng of hamster ME pyrl-3 to give a total volume of 25 pl and 5-pl samples were spotted onto PEI-C chromatograpms maintained at 4"C, at 3, 6 , and 9 min. The radiolabeled intermediates were separated by chromatography with 0.19 M LiCl, 0.1% (v/v) formic acid at 4"C3 Triplicate samples of standard I4C-and 3H-labeled N-carbamyl-L-aspartate (25,500 dpm) were also chromatographed and processed along with the experimental samples.
I4C-labeled intermediates were detected on PEI-C by autoradiography at -70°C.3 For detection of tritiated compounds, chromatograms were saturated with a solution of 7% (w/v) 2,5-diphenyloxazole in diethyl ether, dried, and the positions of the spots were determined by fluorography at -70°C using Kodak XR-5 X-Omat R f i l m (Randerath, 1970). Spots of [14C]carbamyl phosphate were cut out and counted immediately3 and spots of I4C-and 3H-labeled N-carbamyl-L-aspartate, ~-5,6-dihydroorotate, orotate, and OMP were cut out and eluted from the PEI-C in scintillation vials containing 1.5 ml of 0.5 M NaCl, 1.0 M HCl. After gentle agitation for 1 h, 13.5 ml of toluene/ Triton X-100 scintillant was added directly to each vial and the samples were counted using a Beckman LS-100C liquid scintillation system with the gain control set at 250 and counting windows set at 0 to 70 for 3H and 250 to lo00 for I4C. Under these conditions, standard N-[14C]carbamyl-~-aspartate counted with an efficiency of 53.97% in the upper window after subtraction of background, and standard Ncarbamyl-~-[~H]aspartate counted in the lower window at 32.17%. The ratio of counts in the lower 3H window to counts in the upper I4C window when I4C was counted was 9.915 X IO-' and an inverse ratio of 4.356 X was obtained when 3H was counted indicating a minimal spillover of 3H counts into the I4C window. Using these values and the known specific activities of I4C (52.4 Ci/mol) and 3H (58.0 Ci/mol), a program was devised for a Hewlett-Packard HP-19C calculator which calculated the micromolar concentrations of all "Cand 3H-labeled intermediates. The contribution of endogenous N-[14C]carbamyl-~-aspartate to the synthesis of L-dihydroorotate (measured as OMP) at various concentrations of exogenous N-carbamyl-~-[~H]aspartate was determined from these values.

RESULTS
Many of the experiments described in this paper utilize HC03as a precursor for the synthesis of carbamyl phosphate and subsequent intermediates by ME pyrl-3. Fig. 1A5 shows Thus to obtain the K , value of ME pyrl-3 for HC03-where concentrations of HC03-were subsaturating, sampling times of 1, 2, and 3 min were used for determination of initial reaction velocities to minimize the loss of HC03-. Using this precaution and ['4C]HC03-standardized as described under "Experimental Procedures," a K , value of hamster carbamyl phosphate synthetase for HC03-of 610 p~ was obtained in the presence of 10 m~ ATP, 15 mM MgC12, and 100 PM PRPP (Fig. lb). The K , value obtained for HC03was not changed significantly in the absence of PRPP or when 500 PM MgUTP was substituted for PRPP in assay mixtures containing 50 mM ATP, 55 m~ MgCl2 to ensure saturation of carbamyl phosphate synthetase with this substrate. The K,,, value of 610 PM obtained for HC03-is considerably lower than values reported previously from mouse spleen of about 10 mM at pH 7.4 ) and 9.5 m~ at pH 7.2 (Tatibana and Shigesada, 1972a), but is comparable to a value of 1.2 mM obtained for carbamyl phosphate synthetase I1 from Escherichia coli (Anderson and Meister, 1966).
The apparent Michaelis constants of s0.5 values of MEpyrl-3 for substrates are summarized in Table I '' Similar experiments using sealed plastic tubes (1.5 m l ) indicated that 14C02 was evolved at the same rate for 20 min until equilibrium was established between I4CO2 in the incubation mixture and in the air space above it. Further loss of 14C02 then proceeded at a much slower rate. 5,6-dihydroorotate were determined under the standard conditions for assay of the overall activity described under "Experimental Procedures." Values for the other substrates were obtained for N-carbamyl-L-aspartate (92 p~) and L-dihydroorotate (3.1 p~) for dihydroorotase of mouse or hamster ME pyrl-3 (Table I). The K,,, value of hamster aspartate transcarbamylase for carbamyl phosphate of 3.7 p~ (Table I) Table I) than their value of 1.7 m~ from rat liver. However, this difference may not be significant because values obtained in the presence of PRPP and UTP from both sources were comparable. Thus, the only major discrepancy between the apparent K , or S0.s values obtained for the three activities of ME p y r l -3 from different rodent sources is the K , value for HC03-of 610 p~ obtained here with pure hamster ME p y r l -3 (Table I) compared with a value of approximately 9.5 mM obtained with partially purified ME pyrl-3 from mouse spleen Tatibana and Shigesada, 1972a). This discrepancy may be attributed in part to evolution of I4CO2 prior to and during assays ( Fig. 1A) and/or the possible partial The K,, values for HC03-, carbamyl phosphate, N-carbamyl-Laspartate, and ~-5,6-dihydroorotate were determined at 37°C in 50 mM KHepes, pH 7.4, 10% (v/v) glycerol, and 2.0 mM dithiothreitol as described under "Experimental Procedures." Initial reaction velocities were determined from a least squares fit to 3 time points for product formation using at least seven different substrate concentrations and each K , value was calculated from a least squares fit of the resultant data to a Lineweaver-Burk plot. Coefficients of determination so obtained were better than 0.99.   Table 111) agrees well with previous values (Shoaf and Jones, 1973;Hoogenraad, 1974).

Enzymatic
' Christopherson and Jones, unpublished experiment. ' See Table I11 and . . denaturation of the carbamyl phosphate synthetase from these wild type cells during purification.
Published activities of carbamyl phosphate synthetase, aspartate transcarbamylase and dihydroorotase have related to purification of ME pyrl-3 and assays have consequently been designed for convenience and maximal sensitivity Coleman et al., 1977). To facilitate analysis of the overall reaction catalyzed by ME pyrl-3, the maximal enzymatic activities of hamster ME pyrl-3 were determined individually at pH 7.4 in 10% (v/v) glycerol, 2.0 m~ dithiothreitol (Table 11). Dihydroorotase was assayed in the biosynthetic direction using N-[14C]carbamyl-~-aspartate as substrate (Christopherson et al., 1978) and with consideration of the unfavorable equilibrium ratio of N-carbamyl-L-aspartate to L-5,6-dihydroorotate of 16.6 to 1 existing at pH 7.4 (Christopherson and Jones, 1980). Plots of activity versus the concentration of pure ME pyrl-3 were linear for carbamyl phosphate synthetase and aspartate transcarbamylase after an initial lag at low protein concentration and the plot for dihydroorotase passed close to the origin. Extrapolated intercepts on the abscissa for protein concentration obtained from a least squares fit to the linear portions of the plots were: carbamyl phosphate synthetase, 2.02 pg/ml, aspartate transcarbamylase, 0.61 pg/ml; dihydroorotase, 0.094 pg/ml. The maximal specific activities presented in Table I1 are the slopes obtained from the linear portions of these plots. The maximal specific activity of aspartate transcarbamylase from a 100,OoO X g supernatant of a crude homogenate of mouse Ehrlich ascites tumor cells prepared by the procedure of Shoaf and Jones (1973) was 18.1 pmol/min/pg of protein. Assuming the turnover numbers of hamster and mouse aspartate transcarbamylase to be the same, ME pyrl-3 therefore constitutes 0.1% by weight of the total soluble cellular protein in these cells.
To analyze the three reactions catalyzed by ME pyrl-3 under optimal conditions at pH 7.4, 10% (v/v) glycerol and 100 p~ P R P P (Ishida et al., 1977) and 90 m~ K' ion  were included in assay mixtures to give maximal carbamyl phosphate synthetase activity. To further approximate conditions prevailing in vivo, a "coupling enzyme mixture" was included in assays to regenerate ATP from ADP, which inhibits carbamyl phosphate synthetase (Shoaf and Jones, 1973), and to convert ~-5,6-dihydroorotate to OMP, preventing accumulation of N-carbamyl-L-aspartate due to the unfavorable equilibrium. The overall reaction scheme is shown in Fig. 2. Because the equilibria between N-carbamyl-L-aspartate and ~-5$-dihydroorotate; ~-5,6-dihydroorotate and orotate; and orotate and OMP all favor the fist men- Hamster ME pyrl-3 (1.85 mg of protein/ml) was purified by the procedure of Coleman et al. (1977) and assays were performed in 50 mM KHepes, pH 7.4, 10% (v/v) glycerol, and 2.0 m~ dithioothreitol. The maximal specific activities were calculated by the method of least squares from the linear portion of a plot of activity versus protein concentration. Carbamyl phosphate synthetase was assayed in the presence of 100 ~L M PRPP and 5.0 mM free Mgz+ and the values obtained were adjusted to VmaX using the concentration of HCOs" in the assay of 3.26 mM and a K, for HCOl-of 610 p~ (Table 1). Dihydroorotase was assayed in the biosynthetic direction. Details of the assavs are described under "ExDerimental Procedures." bamyl phosphate synthetase by ME pyrl-3 tioned intermediate (Christopherson and Jones, 1979;Krakow and Vennesland, 1961;Traut and Jones, 1977) and orotate is a competitive inhibitor of dihydroorotase , it was necessary to hydrolyze the pyrophosphate produced in the synthesis of OMP (Fig. 2) to convert most of the L-5,6-dihydroorotate synthesized by ME pyrl-3 to OMP.3 Control incubation mixtures containing the coupling enzyme mixture defined under "Experimental Procedures" contained low levels of aspartate transcarbamylase and carbamyl phosphate phosphatase activities which consumed ['%]carbamyl phosphate at a total rate of 0.019 pM/miII when the concentration of carbamyl phosphate was 1 PM. This mixture also contained low levels of dihydroorotase activity which consumed 6 pM N-['%]carbamyl-L-aspartate at a rate of 0.13 @min.
Thus, the rates of consumption of carbamyl phosphate and N-carbamyl+aspartate at concentrations prevailing in assay mixtures by these contaminating activities were low compared with their rates of synthesis by ME pyrl-3 (5 pM/min, see Fig. 4). Fig. 3 shows a plot of the overall biosynthetic activity of ME pyrl-3 versus the concentration of pure protein, assayed in the absence and presence of the coupling enzyme mixture.
The activity values obtained are very similar indicating that this mixture has no inhibitory effect upon ME pyrl-3 at the concentrations used. The slope of the linear portion of the plot gave a maximal specific activity for carbamyl phosphate synthetase (after correlation for saturation by 3.26 mru HCO,-) of 329 pmol/min/~g of ME pyrl-3 (Table II), equivalent to a turnover number of 65.8 mol mm-' (mol of subunit)-'.
Without the coupling enzyme mixture, N-carbamyl-L-aspartate is the major product synthesized by ME pyrl-3 from HC03- (Fig. 4A) accumulating to a concentration of 100 pM after 30 mm. The concentration of carbamyl phosphate reaches a steady state level of 0.72 m and L-5,6-dihydroorotate increases to a concentration of 4.5 w after 30 min (Fig.  4A). The low levels of carbamyl phosphate observed confirm the finding of Coleman et al. (1977) that endogenous aspartate transcarbamylase activity is sufficient to convert virtually all carbamyl phosphate to N-carbamyk-aspartate when carba-my1 phosphate synthetase is assayed. The ratio of L-5,6-dihydroorotate to N-carbamyl+aspartate (R) decreases to a steady state value of 0.046 which is lower than the value obtained at equilibrium at pH 7.4 of 0.060 (Christopherson and Jones, 1979). Fig. 4B demonstrates that inclusion of the coupling enzyme mixture efficiently removes L-5,6-dihydroorotate by its con-   (Fig. 2); X, CAP + CA-asp + OMP; 0, CA-asp + version to OMP3 (Fig. 2). This prevents a significant backreaction to N-carbamyl-L-aspartate which reaches a maximal concentration of 7.13 p~ after 5 min and then decreases slowly. Carbamyl phosphate reaches a steady state concentration of 0.20 p~ and the amount of ~-5,6-dihydroorotate, the majority of which is converted to OMP3 (Fig. 2), was equivalent to 80 p~ after 20 min. Carbamyl phosphate synthetase activity, calculated by summing the concentrations of all intermediates between carbamyl phosphate and OMP (Fig. 2) is constant and maximal during the first 5 min at a rate of 4.99 pM/min (Fig. 4B). Aspartate transcarbamylase activity, calculated in a similar manner is 4.96 pM/min during the first 5 min (Fig.  4B) and dihydroorotate is synthesized at a rate of 4.07 p~/ min after an initial lag. This lag is better illustrated in Fig. 4C which shows that the net synthesis of N-carbamyl-L-aspartate (and therefore carbamyl phosphate also) proceeds with no detectable lag after addition of MEpyrl-3, while extrapolation of the time course for net synthesis of ~-5,6-dihydroorotate (most of which is converted to OMP, Fig. 2) 093 p~, Fig. 3). If carbamyl phosphate were channeled to the aspartate transcarbamylase site and were sequestered in some way to M E pyrl-3, an increase in the concentration of active sites would result in a corresponding increase in the level of carbamyl phosphate. To test this possibility, the levels of carbamyl phosphate were measured as a function of the concentration of M E pyrl-3, with and without the coupling enzyme mixture (Fig. 5A). In the presence of the coupling enzyme mixture, the steady state level of carbamyl phosphate was rapidly reached and varied little as the concentration of M E pyrl-3 was varied from 5.5 nM to 220 rn (Fig. 5A). By contrast, in the absence of the coupling enzymes (except for phosphoenolpyruvate and pyruvate kinase), the level of carbamyl phosphate reached in apparent steady state only after prolonged incubation. The values plotted in Fig. 5A  60, and 65 min. The incubation mixtures contained 7.5 mM ['4C]HC03-to compensate for evolution of 14C02 (Fig. 1A) and the assay containing 220 mu M E pyrl-3 subunits had accumulated 545 +M N-carbamyl-L-aspartate and 21.5 p~ L-5,6-dihydroorotate after 65 min of incubation. In the absence of coupling enzymes, the apparent steady state concentration of carbamyl phosphate shows a strong dependence upon the concentration of MEpyrl-3 (Fig. 5A) and the curve intersects on the ordinate at 0.20 p~ with the curve for the steady state concentration of carbamyl phosphate in the presence of coupling enzymes.
It was surprising to fiid that the absence of coupling enzymes affected the steady state level of carbamyl phosphate because this system (Fig. 2) was designed to investigate the steady state concentrations of N-carbamyl-L-aspartate when ~-5,6-dihydroorotate is efficiently removed. The values for the steady state concentration of N-carbamyl-L-aspartate as a function of the concentration of M E pyrl-3 (Fig. 5B) were obtained from the same experiment where carbamyl phosphate levels were measured in the presence of coupling enzymes (Fig. 5A). The curve obtained (Fig. 5B) is biphasic, having a linear upper portion which extrapolates to 5.0 PM at zero MEpyrl-3 concentration. The steady state concentration of N-carbamyl-L-aspartate does not show a strong dependence upon M E pyrl-3 concentration for the upper linear portion of the curve, increasing from 6.75 p~ at 55 nM M E pyrl-3 subunits to 12.1 pM at 220 nM M E pyrl-3.
To quantitatively inter-relate the maximal velocities of the individual activities of M E pyrl-3 (Table 11) with the time course obtained for the overall reaction in the presence of coupling enzymes (Fig. 4B) and the steady state levels of carbamyl phosphate (Fig. 5A) and N-carbamyl-L-aspartate (Fig. 5B), the kinetic constants of aspartate transcarbamylase and dihydroorotase in the presence of the substrates for carbamyl phosphate synthetase were determined ( Table 111).
The apparent K , of aspartate transcarbamylase for carbamyl phosphate increases from 3.65 PM (see also Table I phate with endogenous unlabeled carbamyl phosphate. There is also a substantial increase in the Vmax (Table 111). Values obtained in the absence of L-glutamate were similar. The remarkable increase in the apparent K , for carbamyl phosphate may be due to competitive inhibition with added anions or, more specifically, with ATP. Shoaf and Jones (1973) showed that 10 mM inorganic phosphate is an effective inhibitor of aspartate transcarbamylase and 150 mM KC1 increases the SOa for L-aspartate (Table I) approximately 9-fold.' The substrates for carbamyl phosphate synthetase also increase the apparent K,,, of dihydroorotase for N-carbamyl-L-aspartate from 92.1 IJM to 263 IJM when 100 PM PALA is included in the incubation mixture; the VmaX is almost unaffected by these additions (Table 111). Values obtained in the absence of PALA were similar. The increase observed in the apparent K , for N-carbamyl-L-aspartate may be attributed to competitive inhibition by anions because addition of 150 mM KC1 increased this value from 92.1 PM to 308 IJM and the v,,, remained unchanged. It seems likely from the data obtained that the changes observed in the kinetic constants of aspartate transcarbamylase and dihydroorotase (Table 111) may be attributed to competitive inhibition by added anions and that occupancy of other active sites of ME pyrl-3 does not induce these changes. The data of Table I11 are utilized under "Discussion'' to quantitatively demonstrate that the kinetics of the overall reaction (Fig. 4B)  endogenously synthesized intermediate or from the exogenous equivalent according to their effective concentrations at the active site of the subsequent enzymatic activity. Such experiments have been performed for the intermediates, carbamyl phosphate and N-carbamyl-L-aspartate, of ME pyrl-3 as described under "Experimental Procedures" (Schemes 1 and 2). The rate of synthesis of product ( u ) from endogenous intermediate ( A ) in the presence of various concentrations of exogenous A is given by the expression: where V is the maximal activity and KmA is the apparent Michaelis constant for A under the conditions of assay of the catalytic site utilizing A as a substrate ( and D have been calculated from experimentally determined concentrations of endogenous and exogenous A prevailing during the period of the assay, as described under "Appendix" appearing in miniprint. It has been assumed for this analysis that the rate of synthesis of endogenous A is unaffected by exogenous A and that the rate of consumption of endogenous and exogenous A is constant during the assay period. Equation 3 and expressions for V, A , and D are derived under "Appendix." To investigate possible channeling of carbamyl phosphate between the carbamyl phosphate synthetase and aspartate transcarbamylase sites of ME pyrl-3, endogenous [I4C]carba-my1 phosphate synthesized from ['4C]HC03-was competed with unlabeled carbamyl phosphate for conversion to N-carbamyl-L-aspartate. A parallel experiment was carried out concurrently under identical conditions using unlabeled HCOsas precursor for endogenous carbamyl phosphate and exogenous ['4C]carbamyl phosphate as shown in Scheme 1. With a final concentration of ME pyrl-3 of 1.1 ng/pl, the rate of consumption of exogenous carbamyl phosphate was approximately constant for the 4-min assay period and the rate of appearance of endogenous carbamyl phosphate was constant over this period when the average concentration of exogenous carbamyl phosphate was greater than 124 IJM (Fig. 6A). In the

TABLE I11
Effects of the substrates for carbamylphosphate synthetase upon the kinetic constants of aspartate transcarbamylase and dihydroorotase Aspartate transcarbamylase and dihydroorotase activities were assayed at 37'C using final concentrations of purified hamster ME pyrl-3 of 12.9 and 24.6 pg/pl, respectively, in 50 m~ KHepes, pH 7.4, 10% (v/v) glycerol, and 2.0 m~ dithiothreitol as described under "Experimental Procedures." Initial reaction velocities were determined from a least squares fit to 3 time points for product formation using at least six concentrations of carbamyl phosphate or N-carbamyl-L-aspartate and VmaX and apparent K,,, values were calculated from a least squares fit of the resultant data to Lineweaver-Burk plots. Coefficients of determination so obtained were better than 0.99.

Condition
Apparent K,,, The Overall Synthesis of L-5,6-Dihydroorotate by ME pyrl-3 absence of exogenous carbamyl phosphate, the endogenous carbamyl phosphate rapidly reached a steady state concentration within the 4-min assay period (see Fig. 4B). Fig. 6A shows that increasing concentrations of exogenous carbamyl phosphate rapidly dilute the endogenous carbamyl phosphate at the aspartate transcarbamylase site resulting in a very marked reduction in the rate of synthesis of N-carbamyl-Laspartate derived from HCOS-. There is also a corresponding rapid increase in the concentration of endogenous carbamyl phosphate present in the assay medium but this increase is not equal to the decreased concentration of N-carbamyl-Laspartate, indicating that exogenous carbamyl phosphate has inhibited carbamyl phosphate synthetase activity (upper curue, Fig. 6A). This inhibition of carbamyl phosphate synthetase by exogenous carbamyl phosphate means that the decreased rate of synthesis of endogenous N-carbamyl-L-aspartate (Fig. 6A) cannot be analyzed by using Equation 3. However, the data clearly show that carbamyl phosphate is not tightly channeled by ME pyrl-3 although maintenance of some synthesis of N-carbamyl-L-aspartate at a high concentration of exogenous carbamyl phosphate of 2.6 mM (Fig. 6A) suggests partial channeling of carbamyl phosphate. This result is at variance with the data of Lue and Kaplan (1970) who found total channeling of carbamyl phosphate to a concentration of exogenous carbamyl phosphate of 1 mu by the bifunctional carbamyl phosphate synthetase-aspartate transcarbamylase from yeast. Their data could be reconciled with our findings if a major portion of the exogenous carbamyl phosphate was rapidly consumed within the assay period by excess aspartate transcarbamylase activity. However, the bifunctional yeast protein may have different channeling properties 1 , FIG. 6. Effect of exogenous carbamyl phosphate or N-carbamyl-L-aspartate upon utilization of the endogenous equivalent as a substrate by  A, effect of exogenous carbamyl phosphate upon conversion of endogenous carbamyl phosphate to Ncarbamyl-L-aspartate during a I-min incubation period. 0, endogenous carbamyl phosphate; 13, endogenous N-carbamyl-L-aspartate; A, endogenous carbamyl phosphate + N-carbamyl-L-aspartate. B, effect of exogenous N-carbamyl-L-[3H]aspartate upon the conversion of endogenous N-['4C]carbamyl-~-aspartate to L-5,6-["Cldihydroorotate during a 3-min incubation period. 0, endogenous carbamyl phosphate + N-carbamyl-L-aspartate; 0, endogenous L-5,6-dihydroorotate converted to OMP (Fig. 2) for conversion to L-5,6-dihydroorotate ( Fig. 6B) according to Scheme 2. The L-5,6-dihydroorotate was converted to OMP using the coupling enzyme mixture (Fig. 2) and concentrations of 3H-labeled orotate and OMP were calculated using a specific radioactivity one-third the value of the N-carbamyl-L-["Hlaspartate used (58.0 Ci/mol). This corrected for the loss of 2 atoms of "H as [3H]ethanol and [:'H]water when L-5,6-["Hldihydroorotate was converted to ["Hlorotate ( Fig. 2). Using a final concentration of ME pyrl-3 of 3.7 ng/pl, the rate of appearance of endogenous N-[%]carbamyl-L-aspartate and the rate of consumption of exogenous N-carbamyl-L-["Hlaspartate were approximately constant during the 3-min assay period. If higher concentrations of ME pyrl-3 or longer sampling times were used, endogenous N-[%]carbamyl-L-aspartate approached a steady state concentration (see Fig. 4B) and exogenous Ncarbamyl-L-["Hlaspartate was greatly depleted, preventing analysis by the relatively simple procedure defined by Equation 3. The total rate of synthesis of L-['%]dihydroorotate from endogenous N-['4C]carbamyl-L-aspartate is only moderately decreased by concentrations of exogenous N-carbamyl-L-["Hlaspartate, averaged during the 3-min assay period, of up to 510 pM (Fig. 6B) [A] and D calculated from the measured concentration of l"C and "H-labeled N-carbamyl-L-aspartate (see "Appendix") and values for Km* of 100,200,300, and 400 pM (Fig. 6B). These values were used because the apparent K,,, for N-carbamyl-L-aspartate increases from 92.1 pM to 263 PM in the presence of substrates for carbamyl phosphate synthetase (Table III). The theoretical curves for K,,,A = 300 pM, which would best approximate the predicted operation of the system if endogenous and exogenous N-carbamyl-L-aspartate mixed freely prior to binding at the dehydroorotase site, do not coincide with the experimental values obtained (Fig. 6B). Considerably more total L-5,6-dihydroorotate is actually synthesized from endogenous N-carbamyl-L-aspartate which consequently accumulates less than would be predicted. These data are consistent with partial channeling of N-carbamyl-L-aspartate between the aspartate transcarbamylase and dihydrooorotase sites of ME pyrl-3.
PALA is a transition state analog and inhibitor of aspartate transcarbamylase (Swyryd et al., 1974) with a Ki of 3 x lo-"' M (Kempe et al., 1976) and it was therefore of interest to investigate the effect of this potent inhibitor upon the overall reaction catalyzed by ME pyrl-3.
The time course for synthesis of pyrimidine intermediates from ['%]HCO,-in the presence of 0.80 pM PALA (Fig. 7A), is very different from that obtained in the absence of PALA (Fig. 4B). The rate of synthesis of total N-carbamyl-L-aspartate is initially low due to inhibition of aspartate transcarbamylase, but increases as carbamyl phosphate accumulates in the incubation medium.  (Fig. 2); X, carbamyl phosphate + Ncarbamyl-L-aspartate + OMP; O, N-carbamyl-L-aspartate + OMP. B, dependence of the steady state concentration of carbamyl phosphate required for restoration of aspartate transcarbamylase activity upon the concentration of PALA. Conditions of assay were the same as for Fig. 4 except that assay mixtures for B were incubated for 40 min and contained 6.5 mM ["CIHCOR-.
After carbamyl phosphate has reached an apparent steady state concentration of 18.9 p~, the total rate of synthesis of N-carbamyl-L-aspartate approaches the total biosynthetic activity of ME pyrl-3, as indicated by the dashed line (Fig. 7A).
The concentration of N-carbamyl-L-aspartate of 4.6 p~ after 20 min is lower than that of Fig. 4B due to its initially reduced rate of synthesis and NCO-, the degradation product of carbamyl phosphate at pH 7.4 (Allen and Jones, 1964), has accumulated to a concentration of 3.0 p~. The data of Fig. 7A indicate that inhibition of the overall biosynthetic activity of ME pyrl-3 by PALA can be overcome by accumulated carbamyl phosphate. To further investigate this effect, experiments similar to that of Fig. 7A with different concentrations of PALA were performed using a 40-min incubation time and a higher concentration of [14C]HC03of 6.5 m~ (to compensate for the additional loss of 14C02, Fig. 1A). Under these conditions, the total rate of synthesis of N-carbamyl-L-aspartate again approached the total biosynthetic activity of ME pyrl-3 as the concentration of carbamyl phosphate reached a steady state. The results of this experiment performed at a concentration of ME pyrl-3 subunits of 0.093 p~ (Fig. 7B) indicate that the accumulated steady state concentration of carbamyl phosphate required to overcome the block by PALA increases in direct proportion with the concentration of PALA after an initial lag. Blockade of ME pyrl-3 by concentrations of PALA higher than 2.4 p~ (Fig. 7B) could have been tested using incubation times longer than 40 min, but significant evolution of I4CO2 and possible inactivation of ME pyrl-3 were considered to be limiting factors. The logical extension of the results of Fig. 7B is to ask whether carbamyl phosphate could accumulate in intact growing cells in response to similar intracellular concentrations of PALA, giving those cells resistance to the inhibitory effects of PALA. In this laboratory, the regulation of de novo pyrimidine nucleotide biosynthesis in vivo is under study in mouse Ehrlich ascites tumor cells. It was therefore of interest to determine whether these cells could maintain high intracellular concentrations of carbamyl phosphate. Fig. 8A is a control experiment showing the chemical decomposition of carbamyl phosphate to NCOat 37"C, pH 7.4 in 10% (v/v) glycerol using our recently developed assay pr~cedure.~ The line of best fit passing through the experimental points for carbamyl phosphate concentration is a theoretical curve for first order decomposition with a half-time of 37.6 rnin, a value close to that obtained by Allen and Jones (1964) of 42.0 min at 37"C, pH 7.16. NCOincreases in a corresponding manner from an intercept at zero time of 22 p~, indicating that the preparation of carbamyl phosphate used for this particular experiment was not pure. The total concentration of carbamyl phosphate plus NCOdecreases slightly over the prolonged period of the incubation indicating some evolution of I4CO2. When a crude dialyzed extract of mouse Ehrlich ascites tumor cells was added to a final concentration of 0.74 pg of protein/pl, the rate of decomposition of carbamyl phosphate was more rapid (Fig.  8B), NCOdid not accumulate to the high levels of Fig. 8A and the total I4C in the assay mixture decreased with time indicating considerable evolution of 14C02. The data of Fig.  8B indicate that these cells contain a phosphatase activity capable of hydrolyzing carbamyl phosphate to carbamate (NHzCOO-) plus phosphate. NHzCOO-; decomposes rapidly to NH, plus COz whereas NCOis relatively stable having a half-time for decomposition of 345 min at 37"C, pH 6.0 (Allen and Jones, 1964). This difference in stability of NCOand NHZCOOexplains why the total 14C decreases rapidly in the presence of cell extract (Fig. 8B) due to evolution of 14C02 while the total 14C is relatively stable in the control (Fig. 8A).
The time course for synthesis of pyrimidine intermediates proceeded in a manner similar to that of Fig. 4B until 5 min when L-cysteine was added after which the total rate of synthesis of ~-5,6-dihydroorotate, as indicated by the slope of the time course, decreased to almost zero at 25 min. After 5 mi n, the concentration of N-carbamyl-L-aspartate increased in a corresponding manner and the overall rate of synthesis of pyrimidine intermediates was unaffected. It is concluded that although N-carbamyl-L-aspartate is partially channeled to the dihydroorotase site (Fig. 6B), the local concentration of this intermediate is not sufficient to protect dihydroorotase against inactivation by L-cysteine.

DISCUSSION
The three enzymatic activities of ME pyrl-3 are contained in a single oligomeric molecule (Coleman et al., 1977) and the data of Fig. 6, A and B indicate partial channeling of both the intermediates, carbamyl phosphate and N-carbamyl-L-aspartate. It is therefore of interest to quantitatively inter-relate the kinetic constants for the individual activities of ME pyrl-3 obtained under the conditions of the overall assay (Tables   I1 and 111) with the time course for appearance of pyrimidine intermediates during this assay (Fig. 4B). The predicted rates of the aspartate transcarbamylase and dihydroorotase reactions can be calculated by substituting the steady state concentrations of carbamyl phosphate (0.20 p~, Fig. 4B) and Ncarbamyl-L-aspartate (7.13 p~, Fig. 4B) into appropriate forms of the Michaelis-Menten equation incorporating the maximal specific activities (Table 11), the ratios of the V,,, values in the presence and absence of the substrates for carbamyl phosphate synthetase (Table 111), and the apparent K , values for carbamyl phosphate and N-carbamyl-L-aspartate in the presence of these substrates (Table 111). For aspartate transcarbamylase, the rate predicted is 2.29 pM/min compared with a rate of 4.96 pM/min actually observed (Fig.  4B). The 2.2-fold higher rate observed compared with that predicted can be explained if the local concentration of car-bamyl phosphate at the aspartate transcarbamylase site is 0.44 p~ rather than the average value of 0.20 p~, observed in the assay mixture (Fig. 4B). Similarly for dihydroorotase, the rate predicted from a steady state concentration of N-carbamyl-L-aspartate of 7.13 p~ is 1.48 pM/min compared with that observed of 4.07 pM/min (Fig. 4B). The 2.8-fold higher rate actually observed could be explained if the local concentration of N-carbamyl-L-aspartate is 20.6 p~ while the average concentration in the assay mixture is 7.13 p~ (Fig. 4B).
A general treatment of the relationship between kinetic constants and the steady state concentrations of intermediates in coupled enzyme assays has been reviewed by Rudolph et al. (1979). Such an analysis is also applicable to multifunctional proteins catalyzing successive reactions although the steady state concentrations of intermediates predicted at the active sites may be higher than the average concentration prevailing in the bulk solvent due to the physical association of these sites. The following kinetic properties of ME pyrl-3 under the conditions of assay of Fig. 4B are consistent with the assumptions necessary for this type of analysis. The concentrations of MgATP and L-glutamine are saturating for carbamyl phosphate synthetase (Table I) and the concentration of HC03-is 5.3-fold higher than the K , value of 610 p~ (Table I). Thus, because only a small fraction of these substrates are utilized (Fig. 4B), the carbamyl phosphate synthetase reaction can be considered an irreversible, zero order step where carbamyl phosphate is continuously removed by aspartate transcarbamylase. The zero order rate constant for this reaction is simply the initial rate calculated from the maximal specific activity for carbamyl phosphate synthetase (329 pmol/min/pg of protein, Table 11) multiplied by the concentration of ME pyrl-3 used for Fig. 4B (18.5 X pg/pl) and a factor to correct for the saturation of the enzyme with 3.26 IIIM HC03-z k l = 329 X 18.5 X IO-3 X 0.610 + 3.26 =

pM/min
This value is in excellent agreement with the actual rate of carbamyl phosphate synthetase reaction observed (Fig. 4B) of 4.99 pM/min.
The aspartate transcarbamylase reaction is irreversible, saturated for L-aspartate and f i t order with respect to the concentration of carbamyl phosphate since the apparent K,,, under these conditions (48.7 p~, Table 111) is 244-fold higher than the steady state concentration of carbamyl phosphate observed (0.20 p~, Fig. 4B). The first order rate constant for this reaction is the ratio V,../K, for aspartate transcarbamylase calculated from the maximal specific activity (18,600 pmol/min/pg of protein, Table 11), the concentration of ME pyrl-3 (18.5 X pg of protein/pl), the ratio of maximal activity in the presence and absence of carbamyl phosphate synthetase substrates (390:240, Table 111) and the apparent K, for carbamyl phosphate under these conditions (48.7 p~, x (E) = 11.5 min" The dihydroorotase reaction is assumed to be irreversible because the coupling enzyme mixture removes ~-5,g-dihydroorotate as it is synthesized and first order with respect to Ncarbamyl-L-aspartate because the apparent K , for N-carbamyl-L-aspartate under these conditions (263 p~, Table 111) is 37-fold higher than the steady state concentration of N-carbamyl-L-aspartate of 7.13 p~ (Fig. 4B). The f i s t order rate of ~-5,6-Dihydroorotate by ME pyrl-3 constant ( VmaX/K,,,)  From the values of the rate constants k,, kp, and k3 as they apply to the experiment of Fig. 4B, steady state concentrations of carbamyl phosphate and N-carbamyl-L-aspartate can be predicted using the following equations:  Tables I1 and I11 and the data of Fig. 4B for synthesis of ~-5,g-dihydroorotate from HC0.7-by ME pyrl-3 demonstrate by two independent methods that the local concentrations of carbamyl phosphate and N-carbamyl-L-aspartate at the active sites of aspartate transcarbamylase and dihydroorotase, respectively, are approximately 2.2-fold and 3.1-fold the average concentrations prevailing in the assay medium. An alternative kinetic explanation for rates of the aspartate transcarbamylase and dihydroorotase reactions higher than those predicted from the steady state concentration of intermediates (Fig. 4B) is that when ME pyrl-3 is catalyzing the overall synthesis of L-5,6-dihydroorotate from HC03-, there is activation of these two activities. This explanation seems unlikely since kinetic constants for aspartate transcarbamylase were obtained in the presence of MgATP, HC03-, and L-glutamate to occupy the f i s t active site without synthesis of unlabeled carbamyl phosphate, and N-carbamyl-L-aspartate synthesized was partially cyclized to ~-5,6-dihydroorotate by the third site (Table 111).
Similarly, kinetic constants for dihydroorotase were obtained in the presence of MgATP, HC03-, and L-glutamine to occupy the first site and PALA which binds at the second site without producing N-carbamyl-L-aspartate to dilute the N-['4C]carbamyl-L-aspartate for the dihydroorotase assay (Table 111).
High local concentrations of carbamyl phosphate and Ncarbamyl-L-aspartate at the active sites of aspartate transcarbamylase and dihydroorotase could be due to the close proximity of active sites catalyzing successive reactions or to physical containment of the intermediates by partially buried active sites. The latter possibility is particularly applicable to carbamyl phosphate whose concentration of 0.2 PM is in the same order of magnitude as the concentration of carbamyl phosphate synthetase sites (0.093 pM, Fig. 4B). However, the steady state concentrations of carbamyl phosphate (measured in the presence of the coupling enzyme mixture (Fig. 5A)) and N-carbamyl-L-aspartate (at concentrations of ME pyrl-3 above 10 ng/p1 (Fig. 5 B ) ) did not increase in proportion to the concentration of ME pyrl-3. This is consistent with the proposal that these intermediates are not sequestered or contained during transit between the first and second sites and second and third sites, respectively. There is a strong dependence of the steady state concentration of N-carbamyl-L-aspartate upon ME pyrl-3 at lower concentrations which can be attributed to a decrease in the ratio of the rate constants k l / k~ (see above). As discussed under "Results," carbamyl phosphate synthetase (Fig. 3) and, to a lesser extent, aspartate transcarbamylase, are sensitive to dilution below concentrations of ME pyrl-3 of 10 ng/p1 (Fig. 3), while dihydroorotase activity remains stable under these conditions. Thus, at low concentrations of ME pyrl-3, k , decreases after allowance is made for the protein concentration while k3 remains constant resulting in a decrease in [CA-asp], (Equation 5, Fig. 5B). The very small increase observed for [CAP], in the presence of coupling enzymes, the marked increase in [CAP], in their absence (Fig. 5A), and the moderate increase in [CA-asp], above ME pyrl-3 concentrations of 10 ng/p1 (Fig. 5B) can be attributed to various degrees of product inhibition by accumulated N-carbamyl-L-aspartate and ~-5,6-dihydroorotase, respectively.
The two independent kinetic arguments presented above for high local concentrations of carbamyl phosphate and Ncarbamyl-L-aspartate at the active sites of aspartate transcarbamylase and dihydroorotase, respectively, and evidence that these two intermediates are not physically contained (Fig. 5,   A and B ) indicates that the partial channeling observed (Fig.  6, A and B ) can be attributed to close proximity of the active sites on trifunctional ME pyrl-3. This channeling probably occurs for purely kinetic reasons. The three active sites are located on a single particle in solution and therefore exist within concentration gradients of the two intermediates and are able to operate at effectively higher concentrations than those existing in the bulk medium.
No significant transient (lag) time could be detected for aspartate transcarbamylase activity (Fig. 4C) and therefore the transient time for synthesis of ~-5,6-dihydroorotase from HC03-of 1.3 min (Fig. 4C) can be attributed to the transient for dihydroorotase. Because the maximal rates of the second and third activities are well in excess of the operating rate of the fist activity (Table IT) Easterby (1973) and further discussed by Welch (1977) can be used to calculate the predicted transient times.
where T~ is the transient time for the ith enzyme of a sequence of reactions and K, is the first order rate constant for that reaction. Substituting the value for kp obtained above for aspartate transcarbamylase into Equation 6, a maximal transient time of 5.2 s is obtained, a value below the limits of detection of our experiment (Fig. 4C). Similar substitution of 123 into Equation 6 gives a transient time of 4.7 min for dihydroorotase. The total transient time for the synthesis of ~-5,g-dihydroorotase from HC03-is simply the sum of the individual transient times for aspartate transcarbamylase and dihydroorotase (Easterby, 1973). Since the value of T for aspartate transcarbamylase predicted (5.2 s) and obtained experimentally (Fig. 4C) is low, the total transient time for ME pyrl-3 of 1.3 min can be compared directly with that calculated above for dihydroorotase of 4.7 min. The discrepancy between these two values can be attributed to substrate channeling of N-carbamyl-L-aspartate. Welch and Gaertner (1975) found a considerable reduction in the transient time from that predicted for the aromatic complex of Neurospora crassa which they interpreted in terms of the containment of intermediates within or on the surface of the aggregate. To account for the approximately 3.1-fold higher local concentration of N-carbamyl-L-aspartate at the active site of dihydroorotase (see above), the first order rate constant KI should be multiplied by this factor giving a value for 7 3 of approximately 1.5 min (Equation 6). The transient time for dihydroorotase may be predicted from another equation formulated by Easterby  (Fig. 4C) is the time required for the local concentration of N-carbamyl-L-aspartate to reach approximately 20 p~ at the third active site while the average concentration in the assay medium reaches 7.13 p~ (Fig. 4B). Thus, the channeling of N-carbamyl+ aspartate predicted by two kinetic procedures and demonstrated by a competition experiment (Fig. 6B) is also supported by a reduction in the predicted transient time from 4.7 min to 1.3 min (Fig. 4C).
Our observation that steady state concentrations of carbamyl phosphate accumulated in the presence of PALA can overcome the blockade of this potent inhibitor (Fig. 7, A and   B ) has kinetic interest and implications for possible cancer chemotherapy. For simple competitive inhibition, the data of  where [CAPJ and [CAP], are the local steady state concentrations of carbamyl phosphate at the aspartate transcarbamylase site in the presence and absence of PALA, respectively, and K,PALA is the inhibition constant for PALA. However, Equation 8 describes a straight line and Fig. 7B exhibits curvature at low PALA concentrations. Since PALA is a reversible tight binding inhibitor and the concentration of aspartate transcarbamylase sites of 0.093 pM is of a similar magnitude to the concentrations of PALA where the effect is observed (0 to 0.4 p~, Fig. 7B), the curvature of Fig. 7B may be attributed to significant depletion of free PALA concentrations by formation of enzyme.inhibitor complex. To fully characterize this effect and derive an equation with predictive value, a detailed analysis similar to those described by Williams and Momson (1979) should be performed. Carbamyl phosphate is unstable, having a half-time for decomposition under these conditions of 37.6 min or a first order rate constant for decomposition of 0.0184 min" (Fig. 8A). The rate of decomposition of carbamyl phosphate accumulated in response to PALA inhibition will equal its rate of synthesis, (5.14 pM/min, Fig. 4B) when a concentration of 278 p~ is reached. In practice, this concentration of carbamyl phosphate would not be attained because leakage through the PALA block would lower its net rate of synthesis.
The physiological signifkance of these findings, therefore, depends upon the cellular concentration of ME pyrl-3 and its capacity for carbamyl phosphate synthesis. Fig. 7B suggests that cells containing high levels of MEpyrl-3 that are able to maintain high intracellular concentrations of carbamyl phosphate would be resistant to quite high cellular concentrations of PALA. Jayaram et al. (1979) have found intracellular concentrations of PALA to vary between 1 and 100 p~ in tumor cells. By extrapolation of the linear portion of Fig. 7B to 100 p~ PALA, a very approximate steady state concentration of 3.0 m~ carbamyl phosphate is obtained which would require a synthetic rate of 55 p~/ m i n to balance its rate of decomposition (see above, Fig. 8A), plus the rate necessary for de nouo biosynthesis of pyrimidine nucleotides. The blockade by PALA would result in depletion of cellular UTP and likely elevation of the PRPP concentration, resulting in maximal activation of carbamyl phosphate synthetase  similar to that obtained in our in vitro assays in the presence of 100 p~ PRPP. A concentration of 3.0 m~ carbamyl phosphate would approximately halve the carbamyl phosphate synthetase activity (Fig. 6A), presumably due to product inhibition. Using the specific activity of carbamyl phosphate synthetase of 329 pmol/min/pg of ME pyrl-3 (Table  111) and considering the factors mentioned above, a cellular concentration of ME pyrl-3 subunits of approximately 1.7 PM would be required to maintain a rate of carbamyl phosphate synthesis of 55 pM/min.
The intracellular levels of ME pyrl-3 in tumor cells have been correlated with their susceptibility to PALA. Tumors sensitive to PALA have low levels of ME pyrl-3 (Johnson et al., 1978) while stable mutants resistant to PALA contain up to 100 times the wild type levels of MEpyrl-3 (Kempe et al., 1976). ,As mentioned by Kempe et al. (1976) and demonstrated in Fig. 7B, high levels of carbamyl phosphate synthesized by these cells in the presence of inhibitory concentrations of PALA would decrease the effectiveness of PALA by competition. The question as to whether cells resistant to high concentrations of PALA can sustain intracellular carbamyl phosphate concentrations of the order of 3 mM remains unanswered. Such intracellular concentrations are not inconceivable since Escherichia coli K12, which has low levels of OMP decarboxylase (Womack and O'Donovan, 1978), maintains a cellular concentration of carbamyl phosphate of approximately 0.84 mM during growth in minimal medium (Christopherson and Finch, 1978). In addition, Cohen et al. (1980) found carbamyl phosphate concentrations in the matrix of isolated respiring rat liver mitochondria of 3 m~ which increased to 15 m~ in the absence of L-ornithine. They also found that high concentrations of carbamyl phosphate coincided with inhibition of carbamyl synthetase I activity, an observation that we have made here with carbamyl phosphate synthetase I1 of MEpyrl-3 ( Fig. 6 A ) . Fig. 8B demonstrates that mouse Ehrlich ascites tumor cells contain a phosphatase activity capable of degrading carbamyl phosphate to phosphate plus carbamate. Several nonspecifk phosphatases have been characterized which are capable of hydrolyzing carbamyl phosphate (Diederich et al., 1971;Herzfeld and Knox, 1972;Lueck et al., 1972), but the cellular location and kinetic properties of these enzymes under physiological conditions will determine whether accumulated carbamyl phosphate is hydrolyzed enzymatically in intact cells. Indeed, cells able to hydrolyze accumulated carbamyl phosphate may be sensitive to PALA while cells lacking such phosphatases may have more resistance. Work in progress in this laboratory will determine whether accumulation of cellular carbamyl phosphate is a significant mechanism for PALA resistance.
The time-dependent and ultimately total inhibition of dihydroorotase activity by L-cysteine when ME pyrl-3 is synthesizing ~-5,6-dihydroorotate from HCO3- (Fig. 9) indicates that substrate protection against this inactivation  by endogenous N-carbamyl-L-aspartate partially channeled to the dihydroorotase site is not of ~-5,6-Dihydroorotate by ME pyrl-3 sufficient to prevent this inhibition. Although L-cysteine is not of interest as an inhibitor of pyrimidine biosynthesis in intact cells, the inhibition observed ( Fig. 9) and previous detailed studies with inhibitors  does allow design of potential inhibitors of high potency. As postulated previously, the hydrolysis by dihydroorotase of the "peptide-like" bond of ~-5,6-dihydroorotate may resemble the catalytic mechanism of carboxypeptidase A  where a zinc atom interacts with the carbonyl group of the peptide bond to be cleaved. Ondetti et al. (1979) have developed potent inhibitors of carboxypeptidases A and B which combine thiol groups capable of coordinating specifically with the zinc atoms and functional groups able to interact with other areas of the active sites of these enzymes. A similar inhibitor has been developed by Nishino and Powers (1979) for thermolysin. This approach could be used for designing potent inhibitors of dihydroorotase combining the structure of ~-5,6-dihydroorotate with an "SH or =S group in position 4 of the pyrimidine ring to interact with the putative zinc of this enzyme. Such an inhibitor could fit the active site and coordinate to the zinc atom giving rapid inactivation of dihydroorotase rather than the time-dependent inactivation shown in Fig. 9.
Data presented in this paper that ME p y r l -3 partially channels the two intermediates carbamyl phosphate and Ncarbamyl-L-aspartate due to close proximity of the active sites of this oligomeric trifunctional protein. This channeling can be accounted for by simple substrate kinetics if the local concentration of carbamyl phosphate at the active site of aspartate transcarbamylase is 2.2-fold higher, and the concentration of N-carbamyl-L-aspartate at the dihydroorotase site is 3.1-fold higher, than their average concentrations in the assay medium. The low steady state concentrations of carbamyl phosphate and N-carbamyl-L-aspartate prevailing during synthesis of ~-5,6-dihydroorotate from HC03-by ME p y r l -3 (Fig. 4 B ) are primarily due to the favorable ratios of the rate constants kl/kp and k l / k 3 (Tables I1 and 111) as discussed above. Maintenance of low concentrations of these two intermediates would have the general advantages of conservation of the solvent capacity of the cytoplasm and minimization of chemical side reactions (Atkinson, 1977) and the specific advantages of preventing futile cycling of carbamyl phosphate and possible toxic side effects of N-carbamyl-L-aspartate . Coordinate expression of the fiist three enzymes of de nouo pyrimidine biosynthesis as a single polypeptide chain enables the favorable ratios of rate constants of the three activities to be maintained under all conditions of cellular growth, thus minimizing the accumulation of these intermediates. The major selective advantage to higher animals of having the trifunctional ME p y r l -3 instead of three distinct and separable enzymes would seem to be that these three activities are expressed coordinately, substrate channeling occurs but appears to be of secondary importance and simply a consequence of having three catalytic activities on a single particle in solution.