Proc. Nat. Acad. Sci. USA Vol. 69, No. 1, pp. 258-263, January 1972 Neurotransmitter Synthesis by Neuroblastoma Clones (neuroblast differentiation/cell culture/choline acetyltransferase/acetylcholinesterase/ tyrosine hydroxylase/axons-dendrites) TAKEHIKO AMANO, ELLIOTT RICHELSON, AND MARSHALL NIRENBERG Laboratory of Biochemical Genetics, National Heart and Lung Institute, National Institutes of Health, Bethesda, Maryland 20014 Contributed by Marshall Nirenberg, November 15, 1971 ABSTRACT Neuroblastoma clones were examined for choline acetyltransferase (EC 2.3.1.6), tyrosine hydroxyl- ase (EC 1.14.3.a), acetylcholinesterase (EC 3.1.1.7), and also for neurite formation. One clone does not form axons or dendrites. Three types of clones were found with respect to neurotransmitter synthesis: cholinergic, adrenergic, and clones that do not synthesize acetylcholine or cate- chols, All clones contain acetylcholinesterase. These re- sults show that genes determining neurotransmitter species can be expressed in dividing cells, that the parental programs of gene expression are inherited, and that dividing cells can be programmed with respect to their ability to communicate with other cells. Elegant biological studies have vielded much information pertaining to the problem of how neural circuits form as the nervous system is assembled. However, virtually nothing is known about the molecular mechanisms for synapse forma- tion. The problem ultimately must be defined in terms of the genetic program for generating different cell types and the steps that determine the specificity of neurons in forming functional synapses. The neuroblastoma system established by Augusti-Tocco and Sato (1) provides an unusual opportunity to explore steps in neuron differentiation and function. The cells multiply rapidly in vitre, yet exhibit many properties characteristic of differentiated neurons (2-7). In this report, the properties of additional clones derived from the mouse neuroblastoma are described. Three cell types, cholinergic cells, adrenergic cells, and cells that do not synthesize acetylcholine or catechol- amines, were detected. METHODS AND MATERIALS Cells. Mouse neuroblastoma C-1300 cells were grown as described (7). Some clones were obtained in two stages: first, a well-isolated colony of cells in agar was picked and then cloned by isolation of a single cell with a stainless-steel cylinder. In other cases, cells were added to petri dishes con- taining broken coverslips; each glass shard with a single cell was then transferred to a separate dish. Chromosomes were analyzed by incubation of cells in logarithmic growth for 6-12 hr with 15-300 uM colcemide (N-desacetyl-N-methyl-colchicine) obtained from Ciba; chromosomes were spread by the method of Merchant, Kahn, and Murphy (8}. Choline Acetyltransferase (EC 2.3.1.6) Assay. Cell mono- layers were washed 3 times with an isotonic salt solution; then cells and protein were harvested by scraping and washing with 10 mM potassium phosphate buffer (pH 6.8)— 258 1mM EDTA (potassium salt). The recovered suspension was sonicated for 5 min at 3°C, divided into small portions, and stored in a vapor-phase liquid-nitrogen freezer. Choline acetyltransferase activity was assaved by a method modified (manuscript in preparation) from that of Schrier and Shuster (9). Each reaction contained the following components in a final volume of 0.05 ml, except where noted: 50 mM _ potassium phosphate buffer (pH 6.8), 200 mI NaCl, 1 mM EDTA (potassium salt), 2.6 mM _ choline iodide, 0.5% Triton X-100 (Packard), 2.2 mM [#C]lacetyl CoA (10 Ci/mol), 0.1 mM neostigmine methylsulfate, and 0-0.5 mg of homogenate protein. Each reaction was incubated at 37°C for 10 min; then 0.5 ml of H.O at 3°C was added and the diluted reaction and 2 subsequent 1.0-ml washes were passed through a 0.6 * 5 em column of Bio-Rad AG 1-X8 resin (CI~ form, 100-200 mesh). Each eluate was collected in a scintillation vial; 10 ml of scintillation solution [1000 ¢ Triton X-100-2 liters of toluene-165 ml Liquifluor (New England Nuclear Co.)] was added and radioactivity was determined. The counting efficieney for “4C was 80-90%. Duplicate or triplicate homogenates were prepared and each was assayed for choline acetvltransferase activity at 4 con- centrations of protein. The rate of reaction was proportional to enzyme concentration within the range 5-350 pmol of [!C acetylcholine formed per 10 min. Assay reproducibility with replicate homogenates was +15%. Each value reported is the average of values obtained with 2-3 homogenates. M4C-Labeled reaction products in column eluates were characterized by paper chromatography or electrophoresis. Reactions were modified so that the specific activity of the [4C]acety] CoA was 40-50 Ci/mol, and choline chloride rather than choline iodide was used. Solutions containing M4C-labeled products (25-30 yl of a column eluate); 0.2 nmol of unlabeled acetylcholine, and 0.2 wmol of unlabeled acetylearnitine were subjected to ascending paper chroma- tography for 16-24 hr with 1-propanol-0.1 N acetic acid 3:1. Chromatograms were dried and sprayed with the Dragendorf reagent (18) to visualize acetylcholine or acetylcarnitine. The chromatogram was cut into 1.0 x 0.5 cm segments, and the radioactivity of each was determined with a scintillation counter. Acetylcholinesterase (EC 8.1.1.7) Assay. The enzyme was assaved as described by Blume et al. (7). Tyrosine Hydroxylase (EC 1.14.8.a) Assay. Cell monolayers were washed, harvested, and sonicated as described above, Proc. Nat. Acad. Sct. USA 69 (1972) except that cells and protein were harvested in 0.1 M potas- sium phosphate buffer, pH 6.2. Tyrosine hydroxylase activity was assayed by a modifica- tion of the methods described by Nagatsu, Levitt, and Udenfriend (10), and by Shiman, Akino, and Kaufman (11). Each reaction contained the following components in a final volume of 0.05 ml: 0.1 M potassium phosphate buffer (pH 6.2), 0.6 mM L-[3,5-di *H]tyrosine (12 Ci/mol, from Amersham-Searle), 0.3 mM 6,7-dimethyl-2-amino-4-hydroxy- 5,6,7,8-tetrahydropteridine (Calbiochem), 0.25 mM NADPH (sodium salt), about 27 ug of sheep-liver dihydropteridine reductase protein (purified through the second ammonium sulfate-precipitation step of Kaufman (12)], and 0-0.5 mg of homogenate protein. Reactions were incubated at 34°C for 10 min, and were stopped by the addition of 0.5 ml of 0.17 N acetic acid at 3°C and assayed as described by Nagatsu et al. (10). 93% of the *H* released from t-[3,5"H |tyrosine was recovered in the column eluate; appropriate corrections were applied to reported values. An internal standard of *H OH then was added to each sample and radioactivity again was determined. The counting efficiency for 3H was about 30%. The rate of reaction was proportional to the concentration of homogenate protein for values reported. Duplicate homoge- nates were prepared and each was assayed for tyrosine hydroxylase at four protein concentrations; reproducibility was £25%,. Average values are reported. Protein was assayed by a modification of the method of Lowry (13). Characterization of the *#H-labeled product of the Tyrosine Hydroxylase Reaction. The tyrosine hydroxylase reaction contained 0.1 mM p-bromo-m-hydroxybenzyloxyamine, an inhibitor of aromatic L-amino acid decarboxylase, in addition to the components described above. 3H-Labeled products formed during incubation were adsorbed to alumina and separated from {*H tyrosine as described by Nagatsu et al. (14), except that ’H-labeled products were eluted with 0.2 N HCl. Recovery of 3,4-dihydroxypheny alanine was 78%. An appropriate correction was applied to values reported. The [Hleatecholamines then were characterized by paper and thin-layer chromatography with the following solvents: 1-butanol-glacial acetic acid-H.,O 12:38:58; methylethy]- ketone-formie acid-H,O 24:1:6; ethylacetate-glacial acetic acid-HLO 15:15:10; and 1-butanol-L N acetic acid-ethanol 35:10:10. Thin-layer chromatography was performed with Eastman Chromogram Sheet 6065 (20 X 20 cm). Spots were located after development by spraying with ethylenediamine ferricyanide solution (15) to locate vatechols, or with nin- hydrin to locate tyrosine. RESULTS Cell types The specific activities of tyrosine hydvosylase, choline acetyitransferase, and acetylcholinesterase found with homog- enates of the neuroblastoma tumor grown in vivo and different clonal cell lines derived from this tumor are shown in Table 1. Values obtained with mouse L-cells, a fibroblastic cell line, and mouse brain are also given for comparative purposes. The specific activity of choline acetyltransferase from tumor was about 2% that of mouse brain. Acetylcholine synthesis was detected with most cell extracts, ineluding L-cells, and other estublished cell lines not shown here: however, the rate of acetylcholine synthesis was <2% that of mouse brain Neurotransmitter Synthesis by Neuroblastoma 259 (1-10 pmol of acetylcholine formed per min per mg of pro- tein). The neuroblastoma-tumor and mouse-brain specific activities of tyrosine hydroxylase were <3% that of adrenal medulla. Of the 21 clones derived from neuroblastoma C-1300 that were assayed for tyrosine hydroxylase and choline acetyl transferase, 12 clones were inactive with respect to both enzymes; 6 clones were cholinergic with high choline acetyl- transferase activity; and 1 clone was adrenergic, with tyrosine hydroxylase specific activity about 200-fold higher than that of brain. Two cell lines were found with low activities of both tyro- sine hydroxylase and choline acetyltransferase, but further evi- dence is needed to distinguish between a cholinergic-adren- ergic cell type and a mixture of adrenergic cells and cholin- ergic cells. No cells were found with high activities of both tyrosine hydroxylase and choline acetyltransferase. These re- sults show that three, possibly four, classes of neuroblastoma cells with respect to neurohormone synthesis can be derived from neuroblastoma C-1800. The specific activity of acetylcholinesterase was high with all neuroblastoma clones tested. In addition, electrically excitable cells were found with each cholinergic, adrenergic, and inactive neuroblastoma clone tested. Homogenates were prepared from stationary-phase cells rather than from logarithmically growing cells because neuroblastoma C-1300 tyrosine hydroxylase, choline acetyl- transferase and acetylcholinesterase activities respond to regulatory mechanisms, and are 30-, 4-, and 25-fold higher, respectively, in nondividing than in logarithmically multi- plying cells (7, 16, and unpublished data). Genetic heterogeneity was examined by determination of the modal number of chromosomes per cell. The mouse neuroblastoma arose spontaneously in 1940; a modal value of 66-70 chromosomes was found by Levan in 1956 (17). Modal values of 59 and 118 were found with cholinergic clones, 104 and 192 with adrenergic clones, and 100-120 with in- active clones. It is clear that neuroblastoma cells contain more chromosomes than does a diploid mouse cell, and that clones differ from one another in their chromosome content. Neuroblastoma clones without tyrosine hydroxylase or choline acetyltransferase may synthesize other transmitter- like compounds. Histamine was detected in uncloned neuro- blastoma cells that had been subcultured frequently at a concentration of 17 pmol/mg protein (unpublished data), a value similar to that of rat brain. However, glutamic acid decarboxylase was not detected with neuroblastoma extracts (unpublished data). Subclones A cholinergic neuroblastoma clone and two inactive clones were recloned, and the properties of the sublines then were studied (Table 2). 23 of the 26 subclones derived from cholinergic clone NS-20 contained active choline acetyl- transferase, and closely resembled the parent-cell type; however, 3 clones were found with relatively low activities of choline acetyltransferase. Three sublines derived from inactive clone N-4, and 5 sublines derived from inactive clone N-18, were similar to the parent cells. However, 2 subclones of N-18 contained higher activities of choline acetyltransferase than the parent clone. These results show that most, but not all, subclones resemble the parental type. Clone N-1, with both tyrosine hydroxylase and choline 260 Cell Biology: Amano eé al. Proc. Nat. Acad. Sct. CSA 69 (1972) Taste l. Types of neuroblastoma C-1300 clones Choline Acetyl- Modal number Tyrosine acetyl- cholin- of Source of homogenate hydroxylase transferase esterase chromosomes —————-pmol product formed per min per mg of protein ————~ Mouse L cells, clone A9 2 440 Mouse brain 5 590 69, 000 Neuroblastoma tumor, 71 vivo 0 10* 130,000 66-70T Cholinergic clones NS-20Y 920 47,000 NS-20 0 490 48,000 59 NS-26 1 437 47,000 116 NS-25 1 125 55,000 59 NS-18 1 76 80, 000 59 NS-21 0 A2 36,000 61 NS-16 0 43 67,000 60 Adrenergic clones N1-106 60 a* 179, 000 104 Nik-115 980 0.1 256, 000 192 Inactive clones N-1A-103 0 2 19,000 101 N-3 9 4 174,000 N-4 4 0.1 46,000 105 N-7 6 0 99 , 000 N-8 2 2 42,000 N-9 11 1 53, 000 N-10 1 3 55,000 N-11 0 5* 404, 000 N-12 0 4 346, 000 N-13 0 6 130,000 N-18 2 2 105, 000 110 13 other clones 5* Uneertain (clonal homogeneity not established) N-1 70 22 23,000 106 N-5 25 13 151,000 109 * Estimate based on determination of C-labeled products in column eluates. t Data of Levan, cited by Hauschka (17). acetyltransferase, was recloned, and the enzyme activities of the sublines were determined (Table 3). Unlike the other neuro- blastoma clones, N-1 cells do not extend axons or dendrites. The cell population was homogeneous initially, consisting of cells that were attached well to the surface of a petri dish an TaBLE 2. Subclones of cholinergic and inactive clones Choline acetyl- Tyrosine Cell line transferase hydroxylase pmol of product formed per min per mg of protein Cholinergic clone NS-20 100-750* 23 Subelones 100-930* 3 Subclones 25-55* Inactive clone N-4 0.1 5 3 Subclones 1 0 Inactive clone N-18 2 2 5 Subclones 5 2 2 Subclones 16 1 * Estimate based on determination of 4C- labeled products in column eluates. usually formed short spikes (<30 um), but were entirely devoid of neurites 100-2000 ym in length. However, after the 10-20th subculture, some relatively large cells were observed with long, branched neurites. N-1 then was recloned and 19 subclones and 10 colonie= were examined. Two N-1 subclones were found that were without neurites, tyrosine hydroxylase, or choline acety!- transferase; however, acetylcholinesterase was present and cells with electrically active membranes were found when one clone was examined. Fifteen adrenergic clones with neurites were found. Each clone studied contained tyrosine hydroxy- lase and acetylcholinesterase activities, but cells were almo-t devoid of choline acetyltransferase activity. One clone (N1-106) with neurites contained 104 chromosomes per cell; other cell lines contained about 200 chromosomes per cell, Several nonadrenergic clones with neurites were found: clone NIE-113 contained 205 chromosomes per cell. These results show that the N-1 cell population is heterogeneous and contains adrenergic and nonadrenergic cells. No cholin- ergic or adrenergic- cholinergic cell type was found. Clonal morphology Four types of neuroblastoma clones, incubated in the absence of serum to stimulate neurite extension, are shown in Fig. 1: proc. Nat. Acad. Sci. USA 69 (1972) Neurotransmitter Synthesis by Neuroblastoma 261 Tape 3. Azon-minus and adrenergic subclones of neuroblastoma clone N-1 Choline Acetyl- Modal number Tyrosine acetyl- cholin- of Subclones hydroxylase transferase esterase chromosomes -—————-pmol of product formed per min per mg of protein Parent clone N-1f 70 22 23, 000 106 Axon-dendrite minus subclones N1A-103t 0 2 19,000 101 N1A-1047 1 5 Adrenergic subclones N 1-106 60 5* 179,000 104 NLE-115 980 O.L 236, 000 192 NLE-124 350 0.9 175, 000 207 NLE-125 330 1.3 98, 000 N1E-fi6 225 0.5 109, 000 202 N Le-122 215 0.4 40, 000 N 1-110 150 N 1LE-126 122 0.3 73,000 202 NLE-114 93 0.1 74, 000 Nile-112 63 N 114-128 60 NL}E-127 60 NLE-123 40 NIE-111 32 NILE-118 17 Minus tyrosine hydroxylase subclones NIE-113 6 205 N1iE-117 1 * Estimate based on determination of C-labeled products in column eluates. ; All subelones were positive for neurites, except those marked f. Fig. 1. Axon-dendrite formation by (A) cholinergic clone NS-20; (B) adrenergic clone NIE; (C) inactive clone N-18; and (D) inactive clone N1A-103, which does not form axons or dendrites. Cells were incubated in growth medium without serum for five days to stimulate neurite formation. The scale shown in A applies to all panels, and corresponds to 10 um. 262 Cell Biology: Amano et al. Proc. Nat. Acad. Sct. USA 69 (1972) Tass 4. Neuroblastoma C-1300 clones Acetylcholin- esterase and Choline Modal number excitable acetyl- Tyrosine of Cell type membranes Neurites transferase hydroxylase chromosomes Axon-minus + _ _ _ 101 Inactive + + - - 110 Cholinergic + + + - 59, 116 Adrenergic + + - + 104, 200 cholinergic, adrenergic, inactive with neurites, and the clone lacking neurites. Cells from each clone adhered well to the surface of the petri dish, but only the axon-minus line (N1A-103) was devoid of long neurites. Axon-minus cells and N-18 cells that form long neurites were mixed and cultivated in the same flasks for more than a week ; however, no influence of one cell type upon the other was detected. Product identification Neuroblastoma N-1E tyrosine hydroxylase activity was dependent upon a pteridine cofactor (6,7-dimethy]-2-amino- 4-hydroxy-5,6,7,8-tetrahydropteridine). Tritium release from L-[3,5-*H Jtyrosine agreed well (within 5%) with PH]di- hydroxyphenylalanine formation. The labeled product of the tyrosine hydroxylase reaction formed in the presence of an inhibitor of aromatic amino-acid decarboxylase was character- ized by paper and thin-layer chromatography with four solvents. Between 86 and 99% of the applied tritiated product was identical in chromatographic mobility with authentic dihydroxyphenylalanine. Clone N-1 cells, in the absence of an inhibitor of aromatic amino-acid decarboxylase, synthesize 3,4-dihydroxy phenylalanine, 3,4-dihydroxyphenylethylamine, norepinephrine, 3,4-dihydroxyphenylacetic acid, —3,4- dihydroxyphenylethylglycol, 3-methoxytyrosine, 3-methoxy- tyramine, and 3-methoxy-4-hydroxyphenylacetic acid (un- published data). Catechols formed by clone C-1300 have been characterized also by Schubert, Humphreys, Baroni, and Cohn (2), and by Anagnoste, Goldstein, and Broome (4). The '4C-labeled products of the choline acetyltransferase reaction were eluted from the column and routinely character- ized by paper chromatography. 84-99% of the labeled product formed with homogenates of cholinergic cells was identical in chromatographic mobility to authentic acetyl- choline. However, only 2-25% of the labeled material formed with adrenergic or inactive homogenates was acetylcholine. The major contaminant was identified as [*C acetyl carnitine (unpublished data). DISCUSSION Clones of neuroblastoma C-1300 were examined for two enzymes required for neurotransmitter synthesis, choline acetyltransferase and tyrosine hydroxylase, catalyzing acetyl- choline formation and the first step in norepinephrine syn- thesis, respectively. A summary of data is shown in Table 4. Three types of clones were found: (a) clones that form little or no acetylcholine or catechols; (6) cholinergic clones; and (c) adrenergic clones. One additional clone was found with relatively low cholinergic and adrenergic activities, but we do not know whether this is a fourth cell type or a mixture of adrenergic and cholinergic cells. However, no cells were found that actively synthesize both acetylcholine and cate- chols. Tumors of cholinergic neurons have not been reported before. They probably afflict man, but may not have been recognized for lack of a diagnostic test. Since few mammalian cells other than neurons have high activities of choline acetyltransferase, the enzyme and acetylcholine can be used as specific diagnostie markers for cholinergic neuroblastomas. Since acetylcholinesterase was found with all neuroblastoma cell types, the enzyme apparently is not a specific marker of cholinergic neurons. At least six types of neurons or corresponding tumors that arise from the neural crest can be distinguished on the basis of transmitter synthesis: those synthesizing acetylcholine: 3,4-dihydroxy phenylalanine, 3,4-dihydroxyphenylethylamine, norepinephrine, epinephrine, and sensory neurons that do not synthesize these compounds. The three types of neura- blastoma thus exhibit properties expected of neural-crest neurons. Although neuroblastoma stem cells capable of giving rise to cholinergic, adrenergic, and inactive cells were not found, it seems likely that such cells normally exist. The available information suggests that there are relatively few kinds of universal neurotransmitters in the nervou- system of both vertebrates and invertebrates. Most neurons probably are greatly restricted with respect to the number of kinds of neurotransmitters that can be synthesized. We find cholinergic, adrenergic, and inactive neuroblastoma cell types, but have not detected cells capable of synthesizing both acetylcholine and catechols at rapid rates. It seems likely that the expression of a gene required for the sythesi= of one neurotransmitter may restrict the expression of genex for alternate neurotransmitters. For example, a product derived from the choline acetyltransferase gene directly or indirectly might restrict the expression of genes for tyrosine hydroxylase, and vice versa. Simultaneous expression of acetyltransferase and tyrosine hydroxylase genes might inactivate both genes, or might result in a balanced state of mutual inhibition; i.e., a cell with low cholinergic and adren- ergic activities. Whether a product of one neuron affects the expression of genes for transmitter svnthesis of neighboring neurons is a problem for future study. Analysis of sublines derived from clonal cells demot- strates that most, but not all, resemble the parent type. The mechanisms underlying altered gene expression are nol known. Gene expression may be reversible or mutations may affect structural or regulatory genes. In any event, the loss of a step that commits a cell to one developmental pattern may enable the cell or its descendants to differentiate along an alternate pathway. Thousands of cell generations have elapsed since tumor C-1300 originated, and extensive genetic heterogeneity in the proc. Nat. Acad, Sei. USA 69 (1972) vell population is expected. By a relatively simple experi- mental approach, it may be possible to obtain lines that express genes that are characteristic of many kinds of neurons jrom a single neural tumor cell; i.e, with two sequential lective procedures: first, selection for dedifferentiated cells might yield cell populations enriched in stem cells; then lection for differentiated cells might yield cells that follow alternate programs of differentiation. We conclude that genes determining the species of neuro- transmitter synthesized can be expressed in dividing cells, aud that the parental programs of gene expression are ittherited and perpetuated for hundreds of cell generations. Similarly, normal neuroblasts may be destined with respect to trans- mitter synthesis and may generate different types of clonal cell populations that are programmed with regard to their ability to establish functional communication with other cell types before synapses are formed. We thank Miss Eve Cutler for excellent technical assistance. 1, Augusti-Tocco, G. & Sato, G. (1969) Proc. Nat. Acad. Sct. USA, 64, 311-315. 2. Schubert, D., Humphreys, 8., Baroni, C. & Cohn, M. (1969) Proc. Nat. Acad. Sci. USA, 64, 316-323. 3. Olmsted, J. B., Carlson, K., Klebe, R., Ruddle, F. & Rosen- baum, J. (1970) Proc. Nat. Acad. Set. USA, 65, 129-136. 4. Anagnoste, B. F., Goldstein, M. & Broome, J. (1970) Pharmacologist, 12, 269. we 11. 12. 13. 14. 16. 17. 18. Neurotransmitter Synthesis by Neuroblastoma 263 Nelson, P., Ruffner, W. & Nirenberg, M. (1969) Proc. Nat. Acad. Sci. USA, 64, 1004-1010. Seeds, N. W., Gilman, A. G., Amano, T. & Nirenberg, M. W. (1970) Proc. Nat. Acad. Sci. USA, 66, 160-167. Blume, A., Gilbert, F., Wilson, 8., Farber, J., Rosenberg, R. & Nirenberg, M. (1970) Proc. Nat. Acad. Sei. USA, 67, 786-792. Merchant, D. J., Kahn, R. H. & Murphy, W. H., (1960) Handbook of Cell and Organ Culture (Burgess Publishing Co., Minneapolis, Minn.), pp. 198-200. Schrier, B. K. and Shuster, L. (1967) J. Neurochem. 14, 977- 985. Nagatsu, T., Levitt, M. & Udenfriend, S. (1964) Anal, Biochem. 9, 122-126. Shiman, R., Akino, M. & Kaufman, 8. (1971) J. Biol. Chem., 246, 1330-1340. Kaufman, S. (1962) in Methods in Enzymology, Vol. VY, ed. Colowick, 8. P. & Kaplan, N. O. (Academic Press, New York) p. 802. Lowry, O. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. (1951) J. Biol. Chem., 193, 265-275. Nagatsu, T., Levitt, M. & Udenfriend, 8. (1964) J. Biol. Chem., 239, 2910-2917. Schneider, F. H. & Gillis, C. N. (1965) Biochem. Pharm. 14, 623-626. Amano, T., Richelson, E. & Nirenberg, M. (1971) Fed. Proc. 30, 1085. Hauschka, T. S., Kvedar, B. J., Grinnell, S. T. & Amos, D. B. (1956) Ann. N.Y. Acad. Sci. 63, 683-705. Bregoff, H. M., Roberts, E. & Delwiche, C. C. (1953) J. Biol. Chem. 205, 565-574.