ACETYLATION OF THE CHOLINE ANALOG HOMOCHOLINE BY MEMBRANE BOUND CHOLINE-0-ACETYLTRANSFERASE IN MOUSE FOREBRAIN

The choline analog homocholine is not acetylated ~vitro by choline-0-acetyltransferase (ChAT, EC 2.3. 1.6) which is solubilized by 100 mM sodium phosphate buffer washes of a crude vesicular fraction of mouse forebrain. However, both homocholine and choline are acetylated by a form of ChAT which is non-ionically associated with a subcellular fraction of mouse forebrain containing membrane associated organelles and occluded ACh (P4 ). Acetylation of homocholine by membrane associated ChAT is saturable. 4-(l-Naphthylvinyl) pyridine (NVP) inhibits the acetylation of both choline (60%) and homocholine (40%) by membrane associated ChAT but reduces the acetylation of only choline by soluble ChAT (76%). Choline and homocholine serve as competitive alternative substrates for the same membrane associated ChAT whereas homocholine only acts as a competitive inhibitor of choline acetylation by soluble ChAT. Acetylhomocholine competitively inhibits the acetylation of choline by both soluble and membrane associated ChAT more dramatically than does the natural end product ACh.

Presently not all studies support the classical model. In particular, the transfer of ACh from the cytoplasm to vesicles has not been demonstrated (Marchbanks and Israel, 1972;Suszkiw, 1976;Nelson et~., 1980). Also not all nerve ending ChAT is soluble. Some appears to be non-ionically associated with a crude vesicular fraction.
Rec~nt studies using the choline analog homocholine suggest that ACh released upon depolarization may not necessarily be synthesized by soluble ChAT. Homocholine is similar to choline in most all aspects of cholinergic metabolism. However, homocholine differs from choline in one very important respect -it is not acetylated by soluble ChAT (see lit. review), although it is acetylated by slices of rat cerebral cortex (Boksa and Collier, 1980). Therefore the objective of this study was to determine if an alternative form of ChAT, which is non-ionically associated with a crude vesicular fraction, can acetylate homocholine.

LITERATURE REVIEW
Historical Acetylcholine (ACh) is a neurotransmitter at autonomic ganglia, the neuromuscular junction, at parasympathetic and some sympathetic neuroeffector junctions, and a putative neurotransmitter in the central nervous system. Katz and his associates (Fatt and Katz, 1951, 1953 del Castillo and Katz, 1954) presented evidence that neurotransmitter released spontaneously from nerve endings of the neuromuscular junction (ACh) in multi-molecular quanta generates post-synaptic miniature endplate potentials (m.e.p.p.s.). Nerve stimulation increases the frequcy of m.e.p.p.s. Later, several investigators (e.g. Palade and Palay, 1954;deRobertis andBennett, 1954, 1955) reported that electron microscopy revealed the presence of large numbers of spherical organelles ("synaptic vesicles 11 ) contained within nerve endings. These synaptic vesicles were considered to be the morphological correlates of the multi-molecular quanta reported earlier. The demonstration of quantal release of neurotransmitter, and the observation of synaptic vesicles in nerve endings led to the formulation of the vesicle hypothesis (first described by del Castillo and Katz, 1957). According to this hypothesis, ACh which is stored in synaptic vesicles is released into the synaptic cleft upon nerve stimulation, or depolarization, by a Ca++ dependent exocytotic process. Birks and Macintosh I (1961) demonstrated that the sympathetic gangkion contains several ~ pools of ACh. One pool, known as "stationary" ACh (approximately 15% of the total ACh content of the ganglion) was thought to be contained within nerve axons and cell bodies, and remains unaltered upon nerve stimulation. The other pool (remaining 85% of total ACh), termed "depot" ACh, was thought to be present in nerve endings, and was at least partially releasable upon depolarization. These authors noted that this latter pool seemed to be made up of two subfractions, one of which is smaller than the other and more readily releasable upon depolarization. In 1962, Gray and Whittaker isolated nerve ending particles with synaptic vesicles contained within them (synaptosomes) from homogenates of guinea-pig brain. Hebb and Whittaker (1958) also presented evidence suggesting that the synaptosomes contained "bound" ACh, which corresponds to the depot ACh of Birks and Macintosh (1961).
In 1964, Whittaker et~· reported that hypo-osmotic rupture of synaptosomes released approximately 50% of the bound ACh (S 3 fraction, cytoplasmic or labile-bound) while the other 50% remained associated with membrane fragments (P 3 fraction, vesicular or stable-bound). The stable-bound ACh was bimodally distributed in subcellular fractions of hypo-osmotically ruptured synaptosomes spun through a discontinuous sucrose gradient: one fraction (0) was enriched in synaptic vesicles and the other (H) contained large membrane fragments, then thought to be only nondisrupted synaptosomes. Several investigators using j.!:!_ vivo (Chakrin ~t ~., 1972) or~ vitro (Richter and Marchbanks, 197la,b;Molenaar et~., 1973) techniques, reported that after labelling the subcellular pools of ACh to different specific activities using the precursor labelled choline, the specific activity of the ACh subsequently released was considerably higher than that con-tained in either the cytoplasmic or vesicle-bound stores. These results indicate that newly synthesized ACh is preferentially released in response to nerve stimulation, and also suggest that a small subpool of vesicles may turn over its ACh content faster than the main vesicle population. Barker~! ..!l_. (1972) then demonstrated that a tail vein injection of labelled choline into the guinea pig led to labelling of ACh in the different subcellular fractions of ruptured synaptosomes of the cortex to different specific activities, also suggesting that different pools of ACh turn over at different rates. Fraction H (containing the large neuronal membrane fragments and some membrane associated vesicles) had the highest specific activity, and fraction D (containing monooisperse synaptic vesicles) had the lowest specific activity of ACh. Soluble (cytoplasmic, or labile-bound) ACh had an intermediate specific activity.
In the early 1970s, most investigators agreed that most if not all of the enzyme which synthesizes ACh, choline-0-acetyltransferase (ChAT) is soluble within the nerve ending (see review on the subcellular location of ChAT). Therefore, based on the experimental evidence summarized thus far, and the evidence that most all ChAT is cytoplasmic, the classical model of the cholinergic nerve ending was developed. According to this model, all nerve ending ACh is synthesized by soluble ChAT. ACh is then transported into vesicles where it is stored, until depolarization stimulates the release of ACh by exocytosis. As mentioned earlier, not all studies support this model. Although attempted by several investigators, using vesicle fractions (Marchbanks and Israel, 1972;Nelson et a]_., 1980) or vesicle ghosts (Suszkiw, 1977) transfer of ACh from cytop1asm into vesicles above diffusional levels has never been demonstrated.
Another alternative hypothesis presented, from studies using Torpedo electroplaque (see review by Israel, Dunant, and Manaranche, 1979), was that ACh is synthesized by cytoplasmic ChAT, and may be released directly from the cytoplasm in response to depolarization. Ca++ regulates gates in the neuronal membrane through which ACh is thought to be released during depolarization. In this case the vesicles serve only as storage sites for ACh.
The development of freeze-fracture techniques has provided evidence that ACh is released from nerve ending vesicles by exocytosis. Reese (1973) andHeuser et~-(1974) have observed clusters of "dimples" in the presynaptic membrane in a regular array around "active zones." Thin sections of similar tissues reveal the presence of numerous vesicles clustered near electron-dense regions in the neuronal membrane postulated to be the active-zones of release. These investigators also found that nerve stimulation increases the number of dimples adjacent to the active zones. The dimples are thought to be vesicles in the process of merging with and collapsing into the neuronal membrane during exocytosis.
Recently Katz and Mi1edi (1977) and Gorio et ~l· (1978) have presented studies suggestng that ACh may be re1eased from nerve endings by more than one process, as follows: 1) Ca++ dependent depolarized release (generating electrophysiological end plate potentials); 2) Ca++ dependent spontaneous re1ease (generating m.e.p.p.s.); and 3) Ca++ independent spontaneous release (generating background "noise," thought to account for approximately 90% of spontaneous release, Katz and Miledi, 1977). Carroll and Aspry (1980) have also presented evidence that these different forms of ACh release may originate from different subcellular pools of ACh within nerve endings. The functional significance of these several forms of release remains to be elucidated.

Subcellular location of ChAT
The subcellular location of ChAT was the subject of controversy for a number of years. Hebb and her coworkers (Hebb and Smallman, 1956;Hebb and Whittaker, 1958) demonstrated that ChAT in brain homogenates is partially 11 soluble 11 and partially "particle bound." The 6 soluble ChAT is that which originates from the disrupted nerve axons and cell bodies, and the particle bound ChAT (which represents the majority of ChAT in brain) is that which is present in synaptosomes.
Hypo-osmotic rupture of synaptosomes releases their contents and different groups of investigators isolated ChAT in different subcellular fractions. Whittaker and his coworkers (Whittaker, Michaelson and Kirkland, 1964) recovered ChAT in a soluble cytoplasmic fraction of ruptured synaptosomes, which is removed from the occluded ACh stores (vesicle fractions). deRobertis et al. (1963) found that ChAT was associated with a subcellular fraction which did contain occluded ACh.
Initially the difference was attributed simply to species variations (Tucek, 1966 a,b). However, the work of Fonnum (1967Fonnum ( , 1968) appeared to resolve the situation. He found that the subcellular location from which ChAT was recovered depended on the pH and ionic strength of the suspending medium. Positively charged ChAT binds to the negatively charged membranes more readily at a low pH and low ionic strength, and can be removed by washing the nerve ending fractions with a high ionic strength buffer. Also he found that rat and rabbit brain enzymes (which deRobertis studied) were much more resistant to these effects than guinea-pig (which Whittaker used) and pigeon enzymes.
Fonnum concluded (1968) that at physiological pH and ionic strength "much if not all of the [ChAT] must be in a free form, 11 and this concept has been generally accepted until recently.
Several investigators have also attempted to establish the subcellular location of ChAT using electron microscopic histochemical techniques. Kasa (1970Kasa ( , 1975, utilizing Pb capture of Coenzyme A, found the reaction product in the cytoplasm, and adsorbed to mitochondria and vesicles, but not within vesicles. However, Fiegenson and Barrnett (1977) using different histochemical techniques (utilizing Mn++ and ferrocyanide) found reaction product within the vesicles, and the appearance of the product was sensitive to inhibition by the ChAT inhibitor 4(1-Naphthyvinyl)pyridine (NVP). Therefore, based on cytochemical techniques, the subcellular location of ChAT was still somewhat controversial.
ChAT from mammalian tissue can be separated into several molecular forms on the basis of isoelectric point (e.g. White and Wu, 1973,c).
It is not presently clear which form (or forms) is (are) functionally significant in neurotransmission. Krell and Goldberg (1975) reported that the ChAT inhibitor NVP, when injected into mouse or guinea-pig, inhibited soluble ChAT but there was no significant alteration in steady state brain ACh levels. ~ven at a neutral pH and high ionic strength, a certain amount (20-40%) of nerve ending ChAT remains insoluble. Hebb (1972) has suggested that "Even if reduced to 10%, the enzyme remaining bound, if strategically located, could be responsible for the production of as much ACh as is present in the storage system, the 'depot' ACh of Birks and Macintosh 11 (1961).
Recently Smith and Carroll (1980) havademonstrated that a crude vesicular fraction (P 3 ) of mouse forebrain, prepared from hypo-osmo-. tically ruptured synaptosomes, retains approximately 15% of synaptosomal ChAT activity, after washing the fraction twice with 100 mM sodium phosphate buffer at a pH of 7.4. ChAT which is solubilized by the high ionic strength buffer washes appears to differ from the bound ChAT, in that it is more sensitive to inhibition by ACh and NVP, and is also more dependent on Na+. The fraction (P 3 ) which contains membrane bound ChAT also contains stable bound ACh (protected from hydrolysis by AChE in the absence of a cholinesterase inhibitor). Also ACh synthesized by membrane bound ChAT is partially protected from hydrolysis when excess AChE is added to the incubation medium, whereas ACh synthesized by soluble ChAT is not. This membrane bound form of ChAT may be functionally significant in neurotransmission.

Substrates for ChAT
ChAT catalyzes the synthesis of ACh from choline and acetyl CoA.
An exogenous supply of choline is required to maintain ACh synthesis over a period of time since choline is not synthesized in the brain.
The concentration of choline present in the extracellular space is maintained by choline from the plasma, and also choline resulting from the hydrolysis of ACh by acetylcholineserase present in the synaptic cleft. Choline may also be partially supplied by the breakdown of membrane lipids. Uptake of choline into nerve terminals from the synaptic cleft is thought to occur by one of two transport processes: low-affinity transport, with a Kt of 50-100 uM, and a high-affinity transport, with a Kt of 1-5 uM. The latter has been postulated to be specific for cholinergic nerve endings (Kuhar et~., 1973), and also to be coupled to ACh synthesis (e.g. Barker and Mittag, 1975) (see reviews by Kuhar and Murrin, 1978;and Jope, 1980). The origin of acetyl CoA for ACh synthesis in nerve endings is presently unknown. It is synthesized within mitochondria; but whether acetyl CoA for ACh synthesis is generated at another location in the nerve ending, remains to be elucidated. An external source of acetate groups is required to maintain ACh synthesis. Several investigators have found glucose and pyruvate to be good precursors, both~ vivo (e.g. Tucek and Cheng , 1974) and~ vitro (e.g. Lefresne ~! ~., 1973; . Balfour and Hebb (1952) found that acetone extract powders of brain tissue are able to synthesize ACh from chol~ ine and either citrate or acetate. Tucek (1967 a,b,c) has reported that citrate-ATP lyase (EC 4. 1.3.8) and acetyl CoA synthase (EC 6.2. 1. 1) (which synthesize acetyl CoA from citrate and acetate, respectively) are at least partially extramitochondrial enzymes. Alternatively, Sollenburg and Sorbo (1970), and Sterling ·~dO'Neill (1978) have suggested that citrate may simply mediate the transport of the acetyl moiety from mitochondria to the site of synthesis of ACh. Gibson and Shimada (1980) report that citrate may not be the only intermediate in acetyl CoA metaboli~m in brain tissue, but that acetate and carnitine may also be involved. They also present evidence that acetyl CoA involved in ACh synthesis may be distinguished from acetyl CoA involved in lipid synthesis, by using (-) hydroxycitrate and n-butylmalonate, both inhibitors of citrate metabolism. This agrees with a previous study by Tucek and Cheng (1974) which reports that when rats were injected with [l,s-14 c] citrate, the lipid stores in the brain became labelled, whereas ACh did not. However this last study only examined whole brain ACh, and did not distinguish between the different subcellular pools of ACh.
Another alternative, proposed by Tucek (1967, c) is that acetyl CoA may itself be .transported out of mitochondria during mitochondrial contractions.

Kinetic mechanism of ChAT
Several different mechanisms have been proposed by different groups of investigators on how ChAT catalyzes the synthesis of ACh.
Kinetic studies have often led to different conclusions for several reasons. ChAT has not yet been isolated in pure form and often enzyme preparations may be contaminated with other proteins which interfere with the assay. Also ChAT decreases in specific activity the more it is purified (Hebb, 1972). Finally a variety of different species have been used in kinetic studies. In general, most investigators believe that ChAT functions by the Theorell-Chance mechanism (e.g. White and Wu, 1973, b;Morris et~..!_., 1971). According to this mechanism of enzyme activity both the reactants (choline and acetyl CoA) bind and the products (ACh and CoA) dissociate from the enzyme in a particular order. However, Hersh and Peet (1977) have suggested that not all experimental evidence uniquely supports this type of mechanism.
These authors employed dead-end inhibitors and alternate substates, and suggest that ChAT functions by a sequential random mechanism, in which the substrates bind and the products dissociate in random order. l'l They indicate that product inhibition studies (in which ACh competitively inhibits acetyl CoA but noncompetitively inhibits choline) can be also explained by the sequential random mechanism. Also, the deuterium isotope effect observed by Currier and Mautner (1974) (interpreted as general base catalysis) is not consistent with the Theorel l -Chance mechanism.
Several investigators have attempted to classify ChAT as a sulfhydryl enzyme (i.e., the enzyme requires a free sulfhydryl group in its active site) utilizing various sulfhydryl agents (e.g. Reisburg, 1957;Mannervik and Sorbo, 1970). In general most investigators now agree that ChAT contains "nonessential" sulfhydryl groups, and that inhibition of ChAT by sulfhydryl reagents may be caused by conformational changes in the enzyme, and not by direct binding to the active site (e.g. White and Cavallito, 1970; see also review by Hebb, 1972).
Ionic requirements of ChAT h-ave also been studied by several groups of investigators. Rossier ~! 2-}_. (1977) demonstrated that only Cl activates ChAT solubilized from rat brain. However, Hersh (1980) found that all monovalent anions activate ChAT to the same extent.  and Smith and Carroll (1980) have found that ChAT is dependent on Na+ for activity. · Additionally, Hersh and his coworkers (1978, a,b,c) demonstrated that enzyme activity and mechanism · may be regulated by ionic strength, and that divalent cations

Studies using homocholine
Recent results obtained with the choline analog homocholine (3-trimethylamino propan-1-ol) support the possibility that the activity of soluble ChAT may not be essential to the formation of transmitter released during depolarization (Collier et~., 1977). The structure of homocholine is shown in Fig. 1. Homocholine is similar to choline in many aspects of cholinergic metabolism. It is transported into rat brain synaptosomes (Collier et ~_l., 1977) and el ectroplaques of Torpedo (van Schwarzenfeld ~!. ~., 1979; Luqmani et 2.l·, 1980) and is acetylated and released as acetylhomocholine (AHCh) from the superior cervical ganglion (Collier et 2.l·, 1977), guinea-pig cortex (von Schwarzenfeld, 1979), Torpedo electroplaques (Luqmani et~., 1980) ++ and minces of mouse forebrain (Carroll and Aspry, 1980) by a Ca dependent process during depolarization. A crude vesicular (P 3 ) fraction of mouse forebrain, depleted of its acetylcholine content independently of the cytoplasm (S 3 ) by incubation of mouse forebrain minces in a lithium-high potassium Krebs medium, can also be refilled with either newly synthesized AHCh or new ly synthesized ACh formed from their respective precursors homocholine or choline. This refilling does not occur by transfer of the acetylaterl product from the s 3 fraction since the ratio of newly synthesized to preformed transmitter in the P 3 fraction exceeds that of the s 3 fraction (Carroll and Nelson, 1978;Nelson et~., 1980). Homocholine, however, differs from choline in one very important respect. Homocholine is not acetylated by soluble ChAT (Burgen et~., 1956;Dauterman and Mehrotra, 1963;Currier and Mautner, 1974;Barker and Mittag, 1975;Collier~.!~., 1977). Therefore the objective of this investigation was to determine if homocholine is acetylated by the membrane associated form of ChAT present in the nerve endings of mouse forebrain.

MATERIALS AND METHODS
Choline and 3-heptanone (practical grade) were obtained from Eastman Company (Rochester, NY) and the 3-heptanone was redistilled

Enzyme preparations
Male CD-1 albino mice were killed by cervical dislocation, the forebrains quickly removed and placed into ice-cold Normal Krebs (composition in mM; NaCl, 117; KCl, 3.5; MgS0 4 , 1.2; NaHC0 3 , 28.0; KH 2 Po 4 , 1.2; CaC1 2 , 2.5; Glucose, 11.0). They were blotted dry, weighed, minced, and incubated for 15 min in Normal Krebs at 38°C. Minces were then washed twice with ice-cold 0.32 M sucrose, and subcellular fractions prepared according to the method of Gray and Whittaker (1962), as modified by Collier et~· (1972) and Salehmoghaddam and Collier (1976). Minces were homogenized in 5.0 ml of 0.32 M sucrose at 840 rpm with 8 up and down strokes. This homogenate was spun at 1000 g for 10 min, and the supernatant spun at 17,000 g for 15 min, to yield a P 2 (synaptosomal) pellet. Synaptosomes were hypo-osmotically ruptured by homogenizing the P 2 pellet in 2.0 ml of water at 400 rpm using 8 up and down strokes. This homogenate was spun at 100,000 g for 60 min to yield a crude vesicular fraction (P 3 ) and an s 3 fraction believed to contain soluble cytoplasmic contents. The s 3 fraction was discarded. The P 3 fraction, which contains occluded ACh, was then washed twice with 1.0 ml of 100 mM sodium phosphate buffer (pH 7.4) in order to remove ionically bound ChAT (Fonnum, 1968). Approximately 703 of the ChAT activity recovered from the hypo-osmotically ruptured P 2 fraction was present in the 100 mM sodium phosphate buffer wash of the P 3 fraction whereas 15% remained in the washed P 3 (Smith and Carroll, 1980). The washed P 3 fraction was then homogenized in 1.0 ml of water and layered on top of a discontinuous sucrose gradient containing 2 ml each of 0.4 M and 0. 6 M sucrose and spun at 192,000 g for 30 min. The pellet was homogenized in 500 ul of water and stored at -20°C until determinat i on of ChAT activity. The 0.4 M sucrnse layer was pelleted at 100,000 g for 60 min. This fraction was suspended in 200 ul of water for enzyme assay. The 0.6 M fraction was also pelleted at 100,000 g for 60 min and suspended in 500 ul of water for enzyme assay. P 4 , 0.4 Mand 0.6 M fractions were never kept for more than 7 days, and the solubilized enzyme never more than 2 days, since it was noted that enzyme activity declined after these times even though stored at -20°C. In one set of experiments, the P 4 fraction was washed with 1.0 ml of 100 mM sodium phosphate buffer, centrifuged at 100,000 g for 60 min and the ChAT activity determined in the wash fraction as well as the washed P 4 fraction.

Enzyme assays
ChAT activity in the various fraction was determined in duplicate or triplicate by the method of McCamen and Hunt (1965) as modified by Fonnum (1969). Aliquots of 2 ul of the tissue fractions or the pooled sodium phosphate buffer washes of the P 3 fraction were incubated for varying time periods (activity was linear for at least 30 min for all substrate concentrations used) at 39°C, a temperature which activates brain ChAT activity (Hebb, 1975), with 20 ul of a buffer substrate consisting of the following components: NaCl (234 mM Chromatography Tissue aliquots were incubated as;previously described except that acetylated products were extracted with 5 mg/ml TPB/3-hept. 100 ul of the organic phase was mixed with 50 ul lN HCl and aliquots of the HCl pooled and dried. Two ul of water was added to the dried extract and the sample was streaked on a strip of Whatman #3 paper . Acety~lated products were separated by descending paper chromatography in a solvent system of n-butanol: ethanol:lN acetic acid: water (8:2:1:3) and identified by comparison with [ 14 c] ACh and [ 14 c] AHCh standards run simultaneously. Labelled products were dissolved from 1 cm strips in l ml of water and counted in Aquasol II.

Electron microscopy
Pellets obtained from the subcellular fractionation were fixed in 1% formaldehyde in 50mM cacodylate buffer, and post-fixed in 1% Oso 4 in Millonig's phosphate buffer. They were then dehydrated in ethanol and embedded in Spurr's resin according to standard procedures (Spurr, 1969). Ultra-thin sections were stained with lead citrate and uranyl acetate, and examined under an RCA EMU-3F electron microscope .

RESULTS
Recovery of P 3 ChAT activity following sucrose density gradient centrifugation.
Aliquots of the subcellular fractions of mouse forebrain were assayed for ChAT activity using 10 mM choline. ChAT which is nonionically bound to the P 3 fraction, i.e., ChAT which remains associated with that fraction after two washes with 100 mM sodium phosphate buffer (pH 7.4), formed 429.7_2: 71.4 nmol/g wet wt/hr of acetylated product (n=4). Approximately 81% of the ChAT activity associated with the washed P 3 fraction was recovered from the 3 fractions of the sucrose density gradient: the 0.4 M sucrose fraction formed 9.0_:2.5 runol/g wet wt/hr (n=24) and electron microscopy revealed that this fraction is enriched with spherical organelles resembling vesicles; the 0.6 M sucrose fraction formed 48.5_:14.9 nmol/g wet wt/hr (n=6); and the P 4 fraction formed 289_2:61 nmol/g wet wt/hr (n=18). Electron micrographs of this fraction revealed that it contains large membrane fragments, which have spherical organelles associated with some fragments, similar to those shown in a recent report (von Schwarzenfeld, 1978). In the experiment in which the P 4 fraction was homogenized in 1.0 ml of 100 mM sodium phosphate buffer (pH 7.4), centrifuged, and its ChAT activity compared with an unwashed P 4 fraction, the ChAT activity was identical (353.4_:37.4 vs. 362.4_2:14.9 nmol/g wet wt/hr, n=3). ChAT activity in the 100 mM sodium phosphate buffer wash of the P 4 fraction was not detectable. The P 4 fraction, which also contains occluded ACh (Carroll and Benishin,1~79) most closely reflects the metabolic activity of the P 3 , a conclusion supported by a recent publication in which Torpedo electroplaques were studied (Luqmani ~! ~., 1980). Therefore, this fraction was selected for the subsequent kinetic analyses.
Kinetics of homocholine acetylation by the P 4 fraction.
The kinetic parameters for P 4 ChAT activity as a function of added homocholine were determined using substrate concentrations ranging -5 -2 from 10 to 10 M. Each point represents the average velocity, corrected for acetylation of endogenous substrates, for 15 forebrains. Figure 2 indicate that the apparent Km value for homocholine acetylation is 49 uM, and the apparent Vmax for homocholine acetylation is 70 nmol/g wet wt/hr. Boiling tissue prior to incubation with homocholine (10 mM, n=6) reduced the formation of acetylated products to the level of tissue-free blanks.

Results in
Ability of NVP to inhibit choline and homocholine acetylation by the P 4 fraction.
Several reports have suggested that carnitine acetyltransferase, which is present in cholinergic preparations, may acetylate cholinergic neurotransmitter precursors (White and Wu, 1973a;Tucek et~., 1978). To determine if choline and/or homocholine are primarily acetylated by P 4 ChAT rather than carnitine acetyltransferase, aliquots of the P 4 fraction were incubated either in the presence or absence of NVP, which inhibits ChAT but not carnitine acetyltransferase (White and Wu, 1973a). Results shown in Fig. 3 indicate that NVP 21 (4.6 uM) significantly reduces acetylation of choline (60%) and homocholine (40%), but stimulates acetylation of carnitine (18%) by the P 4 fraction.
Jnability of NVP to reduce acetylation of homocholine by solubilized ChAT.
Results presented in Table 1 indicate that NVP (4.6 uM) signifi- When the Vmax equals V'max the curves will merge at the Y axis and not be straight lines. If substrate B is not metabolized by the same enzyme as A but competes with A for binding to the same enzyme, then a series of straight lines is generated which yield similar Vmaxs.
If two enzymes. are involved in the metabolism of the two substrates, the curves obtained will not intersect at one point. For example, the correlation coefficients obtained for the least squares regression lines do not change with increasing homocholine concentration and polynomial regression analysis indicates that the points are best fit by straight lines rather than second degree polynomials. Regression coefficients increase with increasing homocholine concentration and the Vmaxs for choline acetylation do not change with increasing homocholine concentration suggestingthat homocholine competitively inhibits the binding of choline to soluble ChAT but is not being acetylated.

Ability of ACh and AHCh to inhibit acetylation of choline by solubilized and P 4 ChAT
ACh competitively inhibits choline acetylation by ChAT (Kaita and Goldberg, 1969;White and Wu, 1973 b,c;Smith and Carroll, 1980). To determine if AHCh also inhibits acetylation of choline by soluble and/or P 4 ChAT, aliquots of the sodium phosphate buffer wash or P 4 fraction were incubated in the presence or absence of varying concentrations of AHCh and compared with similar concentrations of ACh. To insure that the added inhibitor did not interfere with the extraction of labelled acetylated product into the TPB/3-hept., ACh or AHCh was added to some samples following incubation with the buffer substrate. Their respective effects on the amounts of acetylated products extracted into TPB/3-hept. were determined to be negligible at the inhibitor concentration used.
ACh, as expected, inhibits acetylation of choline by ChAT in the sodium phosphate buffer wash and the P 4 (Figure 7). AHCh also inhibits both soluble and P 4 ChAT activity in a dose · dependent manner, and is a more potent inhibitor of both enzyme forms than is ACh. Dixon plots of ACh inhibition of P 4 ChAT and AHCh inhibition of P 4 and soluble ChAT indicate that inhibition is competitive in all 3 cases, as illustrated in Figures 8, 9, and 10. Plots of l/v of AHCh or ACH concentrati on for the different concentrations of choline (1.1 and 4.3 mM) intersect in tht fourth quadrant, which is consistent with competitive inhibition, and yield Kis of 63mM, lOmM, and 200mM, respectively.

DISCUSSION
In agreement with the results of others (Burgen et i.!_., 1956;Dauterman and Mehrotra, 1963;Currier and Mautner, 1974;Collier et i.!_., 1977), the choline analog homocholine does not appear to be acetylated by soluble ChAT. The ChAT inhibitor NVP (Krell and Goldberg, 1975) does not reduce its acetylation at a concentration which reduces choline acetylation by approxi~ately 75%. It does,however, compete with choline for binding to soluble ChAT since its presence alters the kinetic plots obtained for choline acetylation in a mamersuggestive of competitive binding (Cha, 1968).
Homocholine, like choline, is acetvlated by a form of ChAT which is nonionically associated with a fraction of mouse forebrain nerve endings that contains both occluded ACh and subcellular organelles associated with neuronal membranes (P 4 ). Acetylation of homocholine by this fraction is saturable and inhibited by the ChAT inhibitor NVP.
Homocholine alters the kinetic plots obtained for P 4 ChAT as a function of choline in a manner which suggests that both it and choline are acetylated by the same ChAT in the P 4 fraction. Choline is a superior substrate to homocholine for acetylation by ChAT in the P 4 since the Vmax value obtained for choline by the method of Lineweaver and Burk (1934) (295 ! 86 nmol/g wet wt/hr, n=3) exceeds the V'max values obtained for homocholine using either the method of Lineweaver and Burk (70 nmol/g wet wt/hr, n=l5) or that of Cha (79 + 23 nmol/g wet wt/hr, n=3) by almost 4-fold.
Several results obtained in the present study and a previous one (Smith and Carroll, 1980) indicate that membrane associated ChAT does not represent soluble ChAT which has become artificially associated with membranes during tissue preparation or soluble ChAT contained in nondisrupted synaptosomes. When the P 4 fraction is washed with 100 mM sodium phosphate buffer following sucrose density gradient centrifugation rather than suspended in distilled water, the ChAT activity determined in the washed and unwashed P 4 fractions is identical. This result indicates that suspension of the P 3 fraction in distilled water prior to sucrose density gradient centrifugation does not cause soluble ChAT to bind ionically to membranes recovered in the P 4 fraction.
Addition of excess acetylcholinesterase during the initial brain homogenization hydrolyzes at least 90% of the ACh formed by s 3 without Results presented here appear to differ from those of Boksa and Collier (1980) who concluded that homocholine is probably a substrate for ChAT based on~ situ evidence, but were unable to demonstrate appreciable acetylation of homocholine ~vitro relative to choline.
Their investigation, in contrast to this experimental protocol, encouraged the association of soluble nerve ending ChAT to membranes by suspending ruptured synaptosomes in a low osmotic strength medium.
Then they compared the ability of membrane associated ChAT to acetylate choline and homocholine and were unable to detect appreciable acetylation of homocholine relative to choline. The ability to detect acetylation of homocholine by membrane bound ChAT may be due to the fact that nerve ending ChAT was separated into soluble (acetylates choline but not homocholine, and represents at least 70% of nerve end-; ng ChAT) and membrane bound ChAT ( acetyl ates cha H ne better than homochol ine and only represents 15% of nerve ending ChAT).
Both AHCh and ACh compete with choline for binding to ChAT, as illustrated by the Dixon plots. The product AHCh is more potent than ACh in inhibiting ChAT solubilized from the P 3 fraction and ChAT nomionically associated with the P 4 fraction. As Cha et al. (1975) have suggested, the larger chain analog may become "wedged" in the catalytic site and act as a more potent inhibitor than the natural end product.   (6) 65 + 24 (12) 123 + 8 ( 6) ~;·:C-~ aAliquots of subcellular fractions were incubated with either 10 mM choline or 10 mM homocholine, or 1.2 mM carnitine for 30 min at 39°C, and amounts of acetylated products determined following correction for acetylation of endogenous substrates. Results are expressed as nmol of acetylated product formed/g wet wt/hr. bResults significantly different from control at P< 0.05 (t -test, unpaired Fig. 3 and 4. In all three experiments, the K and Vmax for choline were determined from line A (0 Homocholine). V'max for homocholine was calcu~ated from the Y-coordinate of the point of intersection (Vmaxlv = V'max • Cha, 1968). S, the choline concentration at which homocholine appears to have no effect, was determined from the X-coordinate of the point of intersection (K /S). The theoretical value for S was determined from a plot of the theoretical equation m m m where Km· and V were the means from these three experiments (146 uM and 295 nmol/g wet wt/hr, respectively), aW8xK'm and V'max obtained from Fig. 1 (49 uM and 70 umol/g wet wt/hr, respectively).   indicates that NVP treatment differs from contra 1 at P < 0. 05 (t-test, unpaired).