Intracellular Localization of Enzymes of Arginine Metabolism iu Neurospora*

SUMMARY The physical basis for the confinement of carbamyl phos-phate, ornithine, and arginine to specific anabolic fates has been investigated. Cell extracts of Neurosjora grown in minimal medium were fractionated by differential centrifugation to determine the subcellular localization of the relevant enzymes. (organellar)

and arginine to specific anabolic fates has been investigated.
Cell extracts of Neurosjora grown in minimal medium were fractionated by differential centrifugation to determine the subcellular localization of the relevant enzymes.
Carbamyl-P synthase A (arginine-specific; EC 2.7.2.5), ornithine acetyltransferase, and ornithine carbamyltransferase (EC 2.1.2.2) were found to be associated with a particulate (organellar) fraction. All three enzymes co-sediment with mitochondria on linear sorbitol gradients, and appear to be components of the mitochondrial matrix. The mitochondrial membrane limits the ability of exogenous substrates to be metabolized by these enzymes in vitro. Arginine synthesis from citrulline, putrescine synthesis from ornithine, and the catabolism of ornithine and arginine are carried out by cytoplasmic enzymes.
The catabolic enzymes are found at significant levels even during growth in minimal medium, but little catabolism occurs in vivo. The results indicate that the anabolic and catabolic pathways of ornithine metabolism are, in part, separated by the mitochondrial membrane.
However, the mechanisms responsible for confining carbamyl-P and ornithine to arginine synthesis and arginine to protein synthesis are more complex than the simple physical separation of the anabolic and catabolic enzymes.
The ability to catabolize exogenous amino acids is a widespread feature of metabolism.
Where exogenous amino acids are lacking, however, cellular economy requires mechanisms to prevent catabolism of endogenous amino acids. In most procaryotes, the pool sizes of many endogenous amino acids are often maintained at low levels, and cat.abolic enzymes remain uninduced. In eucaryotic cells, however, endogenous amino acids frequently accumulate despite the presence of significant levels of catabolic enzymes. These large pools arc not catabolized (1). * I). Significant levels of the first two catabolic enzymes arc found in mycclia growing in minimal media (a), and supplementation of the medium with arginine induces only a 4-fold increase in their activity.
Moreover, cultures grown in minimal medium contain large pools of both ornithine (12 rnnl in cell water) and arginine (8 mM in cell water) (2).
Despite the potential for significant catabolism of endogenous ornithine and arginine, lit,&, if any, catabolism occurs in vivo. Two observations support this conclusion.
First, urease-less mut.ants do not accumulate urea in exponential cultures ( Fig. 1) (3). Second, mutants blocked in proline biosynthesis prior to glutamic-y-semialdehyde formation do not form proline from the cndogenous pools of ornithine and arginine (Fig. 1). However, exogenous proline, ornithine, or arginine support optimal growth (4). Despite the absence of catabolism during growth in min-ma1 medium, exogenous ornithine is readily catabolized even before significant change in catabolic enzyme activity or increase in the size of the intracellular pool (5, 6). The mechanism which confines endogenous ornithine to the arginine biosynthetic pathway may also confine carbamyl-P ( Fig. 1). Mutants lacking the pyrimidine-specific source of this compound cannot use carbamyl-P formed by the arginine-specific enzyme (7,8). This "channeling" can be overcome under tollditions which result in an expansion of the intracellular carbamyl-I' pool (7,8).
It has been suggested that channeling of the intermediates may result from compartmentation of the anabolic enzymes and intermediates within an organelle (7)(8)(9).
Histochemical localization of ornithine carbamyltransferase ( Fig. 1) within the mitochondrion supports this view (9). To fully explore this possibility, we have determined the subcellular localization of enzymes involved in arginine and ornithine metabolism.
Compartmentation of ornithine and argininc has also been demonstrated, both in vitro and in vivo, and will be reported elsewhere (10,11 was obtained from Worthington. Preparation of Cell Extract--The preparation of the exponentially growing cultures of N. cra.ssa has been previously described (13). The culture (1000 ml) was harvested when it had reached 1 mg, dry wt, per ml by pouring it through two layers of cheesecloth.
Subsequent steps were carried out at 4" unless otherwise indicated.
The wet mycelial pad was rinsed with cold distilled water and compacted by centrifugation at 500 X g for 5 min. The mycclia were then washed three times with buffered sorbitol (0.1 M citrate, brought to pH 5.8 with KzHPO+ 1 mM EDTA, 0.6 $1 sorbitol, and 0.14 M 2-mercaptoethanol).
The washed cells were extracted and fractionat.ed as outlined in Fig. 2.
The washed pad, representing approximately 1 g, dry wt, was suspended in 50 ml of buffered sorbitol.
After the addition of 1 ml of glusulase, the suspension was incubated with gentle agitation for 30 min at 30" in a shaking water bath.
The material was then centrifuged at 500 x g for 5 min. The pellet was washed twice with 1 M sorbitol by gentle resuspension and centrifugation.
The washed pellet was suspended in 30 ml of fractionation buffer (10 mM potassium phosphate, pH 7.5, 1 mM EDTA, 1 M sorbitol) and the cells were lysed by homogenization in a Teflonglass homogenizer.
The Teflon pestle was attached to a motor and rotated at 1600 rpm.
Initial breakage was accomplished by 6 strokes in the homogenizer.
Unbroken cells and cell debris were removed by centrifugation at 600 x g for 10 min. The result.ing pellet was re-extracted in 15 ml of fractionation buffer by 3 strokes in the homogenizer.
After ccntrifugation, the combined supcrnatants were rccentrifugcd at 600 X g for 10 min. The resulting crude extract was then fractionated as shown in Fig. 2. Material which was not released into the crude estract was solubilized by treating the 600 X g pellet with 5"j0 Triton X-l 00.
Because of the instability of ~~arbamyl-P synthase A, its activity was determined using material desalted by passage through columns of Scphadex G-25.
Density gradients of 30 to 60G10 (w/w) sorbitol were used. Centrifugation was done in a Spinco SW 39L rotor at 4" at 35,000 rpm for 60 min. Fractions were collected in drops after piercing the bottom of the centrifuge tube. Ihxyme Assays-Unless otherwise noted, all enzyme activities were determined at 37". All assays were done in duplicate and correction was made for nonenzymatic activity.
All assays included zero time controls and were checked for substrate and cofactor dependence, linearit,y with time, and dependence on enzyme concentration.
One unit of activity is defined as the formation of 1 pmole of product or the utilization of 1 pmole of substrate in 1 min. Protein was determined by the method of Lowry et al. (14).
Ornithine acetyltransfernse was assayed by the method of nhes (15). Carbamyl-I' synthase A (arginine-specific) was determined by inhibiting the pyrimidine-specific enzyme with UTP. The reaction mixture (0.5 ml) contained 100 mM Tris-acetate, pH 8, 6 mM L-glutamine, 7.5 ml1 potassium [r4C]bicarbonate (specific activity of 0.267 PC1 per pmole), 12 mM MgCls, 12 m&r ATP, 1.0 rnM UTP, and 0.1 ml of enzyme fractions treated with Sephadex. After incubation for 30 min at 25", the reaction was stopped and the carbamyl-I' was converted to urea by boiling the reaction mixtures for 10 min after adding 0.2 ml of 1.5 RI NH&l.
The mixtures were then cooled and acidified by the addition of 0.1 ml of 1 s HCl.
After treatment at 100" for 5 min, the tubes were cooled, covered with filter paper wetted with saturated KOII, and placed in a closed container overnight to remove volatile radioactivity.
The O&ml mixtures were then brought to approximately pH 5.6 by the addition of 0.2 ml of 1 M (sodium)3 citrate, pH 5.6, which contained 5 pmoles of urea as carrier. Urease (0.5 mg) was then added, and evolved radioactive CO2 (from carbamyl-P converted to urea) was trapped and counted as described by Morris et al. (16).
Argininosuccinate synthetase (EC 6.3.4.5) was assayed as the disappearance of the substrate, citrulline, in a reaction mixture similar to that described by Wampler and Fairley (18). The modifications included the use of a l-ml reaction volume containing 1 mM citrulline, 2 rnM ATP, and the addition of an ATP-regenerating system (1.5 mM P-enolpyruvate and 30 pg of pyruvate kinase).
Argininosuccinate lyase (EC 4.3.2.1) was assayed in a coupled reaction containing activated bovine liver arginase. Activation of arginase was accomplished by incubating it at a concentration of 5 mg per ml for 5 min at 55" in a buffer consisting of 10 rnM Tris-HCl, pH 7.5, and 1 mat MnCL.
The l-ml reaction mixtures contained 50 mM potassium phosphate, pH 7.5, 1 mM potassium argininosuccinate, 0.5 mM MnC12, 0.5 mg of the activated bovine liver arginase, and appropriate amounts of the dialyzed cell fractions.
After incubation for 30 min, the reaction was terminated with 1 ml of 20% trichloroacetic acid. After centrifugation, ornithine in the supernatant was determined by the method of Chinard (19).

RESULTS
Cell Fractionation- Fig.  2 indicates the procedure employed to isolate the cell fractions.
The glusulase washes and the 600 X g pellet were analyzed for enzyme content.
This was done to determine how well the crude extract represented the total cell content (21). The washes contained only minor amounts of protein and of the enzymes examined.
Because the 600 X g pellet was enriched in mitochondrial enzyme activities, a typical crude extract contained 80% of the total activity of soluble enzymes, but only 50% of the succinate dehydrogenase.
The nonrepresentative nature of the crude extract does not affect the conclusions about enzyme localization, however, and therefore this problem has been ignored in t,he presentation of the results. No attempt has been made to isolate a nuclear fraction, because most of the nuclei break during cell lysis and the remainder are lost in the 600 X g pellet.
The relat,ively high speed used to s:edimant the particulate fraction was designed to separate all intact organellcs from the soluble components.
Much slower speeds (e.g. 8000 X g) have been found to be adequate for sedimenting mitochondrial ac-tis5ks.. amined among the cell fractions is shown in Fractionation was performed as shown in Fig. 2, and enzyme activities were determined as described under "Materials and Methods." The activity found in a sample of the crude extract was taken as 100%. The results are expressed as the percentage of the enzyme units in the crude extract which were recovered in the specific subcellular fraction. enzymes found associated with the particulate fraction are ornithine acetyltransferase, carbamyl-P synthase A, and ornithine carbamyltransferase.
The small amount (3%) of the argininosuccinate lyase activity in the particulate fraction has not been observed in subsequent experiments.
The measurement of succinate dchydrogenase activity was used to monitor mitochondrial distribution. The particulate carbamyl-P synthase activity was insensitive to inhibition by UTP, whereas the soluble activity was inhibited more than 95y0,. This observation indicates that the particulate activity is that of the argininespecific enzyme.
Enzymes found largely in the soluble fraction are ornithine decarboxylase, argininosuccinate synthetase, argininosuccinate lyase, arginase, and ornithine aminotransferase.
No physical distinction could be observed between the intracellular site of arginine synthesis (argininosuccinate lyase) and catabolism (arginase).
However, enzymes of ornithine and carbamyl-P metabolism are found both in the cytosol and in the particulate fractions, and would thus appear to be in different compartments of the cell.  C, carbamyl-P synthase A (arginine specific).
Centrifugation and enzyme assays were performed as indicated under "Materials and Methods." positions in the gradients was observed if centrifugation was continued for 18 hours. This density (approximately 1.19 g per cm3) is the same as that determined by Luck (22) for Neurospora mitochondria.
Other organelles of Neurospora which might contain these enzyme activities are glyoxysomes and vacuoles.
Attempts to detect these organelles in the gradients shown in Fig. 3 wcrc unsuccessful.
The enzymes characteristic of these organelles were undetectable in the vigorously aerated cultures used here, and are evidently repressed.
However, the reported densities of these organelles (1.22 g per cm3 for glyoxysomes and greater than 1.35 g per cm3 for vacuoles) are sufficiently greater than mitochondria that they would have been distinguishable from the latter in the density gradients (23,24). Some yeast microbodies have been found to band at the same density as mitochondria, but to be distinguishable during rate sedimentation (21). Sedimentation rates of the particulate arginine enzymes were therefore compared with succinate dehgdrogcnase.
Each enzyme appears to co-sediment with succiuate dehydrogenase.
Thus, there is no evidence that ornithinc acctyltransfcrase, carbamyl-I' synthase 8, and ornithine carbamyltransferase are associated with a nonmitochondrial organelle.

Solubilixalion
of Xitochondrial Enxymes-Two possibilities could explain the association of the three particulate arginine enzymes with mitochondria.
First, a common property of these enzymes (i.e. ionic charge) leads to their association with mitochondria without their being integral components of the mitochondrion.
Second, the enzymes may be integrated into the membranes or the matrix of the mitochondrion.
To distinguish Samples of a washed particulate fraction were suspended in 5 ml of a solution containing 10 mM pot.assium phosphate, pH 7.5, 1 mM EDTA, 1 M sorbitol (except the "No sorbitol"sample) and the indicated additions.
Each sample was kept at 0" for 2 hours. During this time they received the indicated treatments.
The percentage of the ornithine carbamyltransferase, ornithine acetyltransferase, and succinate dehydrogenase which no longer sedimented at 15,000 X g in 20 min was determined as described under "Materials and Methods." Enzyme solubilized among these possibilities, the stability of the association of the enzymes with the particulate fraction was tested. Table II shows the effect of various treatments on the retention of ornithine carbamyltransferase and ornithine acetyltransferase by the particulate fraction.
Because of its instability, carbamyl-P synthase A was not similarly investigated.
As expected, both Triton X-100 and sonication solubilized both enzymes.
However, KC1 is essentially without effect, and the absence of sorbitol released only a portion of the enzyme activities.
Dialysis overnight under these conditions, however, solubilized both enzymes, without affecting the sedimentation of succinate dehydrogenase. This behavior distinguishes these enzymes from succinate dehydrogenase, an enzyme integrated into the inner membrane of the mitochondrion.
Since only the more drastic treatments fully solubilized the enzyme activities, their mitochondrial association appears strong. The inability of KC1 to solubilize the activities suggests that the association is not simply an ionic interaction.
Combined with the distinction in behavior described above, the fact that recoveries of the arginine enzymes in the particulate fraction are always less than those for succinate dehydrogenase suggests that these enzymes are associated with the mitochondrial matrix. Whatever the nature of the mitochondrial localization, the similarity in the behavior of ornithine carbamyltransferase and ornithine acetyltransferase suggests that they are in the same mitochondrial compartment.

Effect of Tntact Mitochondrial Structure on Observed Enzyme
Activity-The effect of maintaining intact mitochondrial structure on the apparent activity of ornithine carbamyltransferase and ornithine acetyltransferase is shown in Table III. While decreasing the osmotic strength results in increasing activity, additional activity is observed in the presence of Triton X-100 (the latter has no effect on soluble enzyme activity).
These results indicate that an intact mitochondrial membrane serves as a permeability barrier to at least one of the substrates of each reaction.
No reliable method has been found to examine the ability of the individual substrates to penetrate the mitochondrial membrane.
Similarly, there is no means of assessing whether the nonlatent activity observed at the highest sorbitol concentration is the result of partial permeability of the 1ml11brane to the substrates, or damage to the mitochondrial membranc during isolation of the organ&.
The association of ornithine acctyltransferase, carbamyl-I' synthase A, and ornithinc carbamyltransferase with the mitechondrion identifies this organellc as the site of ornithinc and carbamyl-P synthesis, and their utilization to form citrulline. Because argininosuccinate synthetase and argininosuccinatc lyase are soluble, citrulline evidently leaves the mitochondrion and is converted to arginine in the cytoplasm.
Similar conclusions have been reached in studies with rat liver cells (25). The catabolic enzymes appear to be exclusively cytoplasmic.
The localization of ornithinc acetyltransferasc and ornithine carbamyltransferase within the mitochondrion raises the possibility that other early enzymes of the argininc biosynthetic pathway are confined to this organcllc.
Of particular intcrcst is the inquiry into whether the enzyme which is feedback-sensitive to arginine is mitochondrial or cytoplasmic. The localization of the relevant enzymes and their sensitivity to feedback iiihibition are now b&g investigated.
The question most clearly posed by these results is the location of the large ornithine pool in the cell. The synthesis and USC of ornithinc in the argininc pathway takes place in the mitochon drion.
It is tempting to think that the ornithinc pool is mitochoiidrial.
-k similar view of carbamyll metabolism in the argininc pathmay has been presented previously (8). A test of this possibility, in the case of ornithine, shows it to bc untrue: the oriiithine pool is largely in a nonmitochondrial "vesicle" (lo), and this pool is evidently not a direct intermediate in the synthesis of citrullinc.
Therefore, although most of the ornithine made in the niitochondrion is used there, only the fraction which escapes the mitoclionclrio~~ accumulates iii the cell. In the course of its passage t,hrough the cytoplasm from mitochondrion to resiclc, some of the ornithinc can be used for polyamine biosynthesis by ornithine decarbosylnse (Fig. I), a cytoplasmic en zymc with high affinity for ornit~hilic (I<, of approximately 0.4 mlr).l The fact, that it is not significant,ly catabolizctl 1)~ orn-1 R. L. Weiss, unpublished observations. 5407 thinc aminotransferase (cytoplasmic) may bc csplained by the low affinity (K, of 2 mM) of this enzyme for ornithine (26).
l'crhaps the most intriguing possibility which arises from separation of anabolic and catabolic pathways of ornithinc metabolism is the prevention of energetically wasteful recycling of catabolically derived ornithine (Fig. I). In yeast, where there is as yet no evidence of enzyme localization, this function is accomplished by inactivation of oruithinc carbamyltransferase by arginase in the presence of argininc and ornithine (27,28). Two cxperirnental observations suggest that the mitochondrial membranc might serve this function in hJeurospora.
Second, a mutation affecting ornithine carbamyltransferase has been isolated which is unable to use csogenous ornithine to satisfy the requirements for growth, but is prototrophic if identical levels of the endogenous source of ornithine arc available (29,30). The mitochondrial membrane in Neurospora may minimize the amount of catabolic ornithine which re-enters the biosynthetic pathway.
Hecause the synthesis and catabolism of arginine and its utilization for protein synthesis occur in the cytoplasm, we might expect the argininc pool to be readily available for catabolism. Despite the fact that intracellular arginine (8 m&r in cell water) is more than adequate to satisfy arginase (&, = 5 IYlM),2 no detectable catabolism is observed (3,11). The paradox is rcsolved by the finding that the pool of arginine is confined to the cellular vesicle in which we find ornithine (10).
Argininc metabolism in hTeurospora has proven to be a fruitful sy item for investigation of the role of subcellular organization in regulation (7)(8)(9)(10)(11).
A complete understanding requires identification and characterization of the membrane systems which separate intracellular enzymes and intermediates. Localization of the pools of ornithinc and argiiiinc has proven to bc possible and amenable to further analysis (10). These investigations are continuing in an attempt to define the role and importance of compartmcntation in regulating amino acid metabolism in Neurospora.