Peptides,Vol. 13, pp. 395-400, 1992

0196-9781/92 $5.00 + .00 Copyright© 1992 PergamonPressLtd.

Printed in the USA.

Single-Step Isolation and Sequencing of Vasopressin and Oxytocin Precursors W I L L I A M G. N O R T H ,

E. F. O ' C O N N O R

A N D C. B. G O N Z ~ , L E Z

Department of Physiology, Dartmouth Medical School, Hanover, N H 03 755 R e c e i v e d 26 A u g u s t 1991 NORTH, W. G., E. F. O'CONNOR AND C. B. GONZALEZ. Single-stepisolation and sequencingofvasopressin and oxytocin precursors.PEPTIDES 13(2) 395-400, 1992.--The pre- and post-Golgi processing of preprovasopressin and prepro-oxytocin was evaluated by microsequencing for incorporated radiolabel. 35S-Cysteineand 3H-fucose were microinjected into rat supraoptic nuclei (SON), and proteins and peptides related to the biosynthesis of vasopressin (VP) and oxytocin (OT) were isolated at various times from the supraoptic nuclei and neural lobe by employing a one-step procedure of high performance liquid chromatography (HPLC). These proteins and peptides were recognized through their binding to specific antibodies against VP, OT, and rat neurophysins (RNPs), and by their binding to ConA-Sepharose. Two immunoreactive glycoproteins related to VP biosynthesis were recovered from the SON and both contained fucose and had a 3SS-cysteineplacement consistent with the location of the hormone sequence at the N-terminus. SDS-electrophoresis revealed the major protein form to be 21,000 daltons and the minor protein form to be 19,000 daltons. One nonglycosylated protein of 16,000 daltons related to oxytocin biosynthesis was recovered from the SON, and this protein also had a 3SS-cysteineplacement consistent with an N-terminal OT sequence. These data provide the first sequential evidence that prior to, or shortly after, packaging in the Golgi the preprohormones of VP and OT have lost their entire leader-peptide structures. Vasopressin

Oxytocin

Posttranslational processing

HPLC characterization

VASOPRESSIN and oxytocin neurons of the hypothalamic supraoptic nuclei (SON) and paraventricular nuclei (PVN) translate their peptide products as larger precursor proteins that are believed to undergo posttranslational modification between the rough and smooth endoplasmic reticulum (7,22,25,32). In the case of the vasopressin precursor, this modification includes the addition of a carbohydrate moiety, a process that culminates in the addition of fucose before the protein is packaged into neurosecretory vesicles within the Golgi apparatus (11,34). Intravesicular proteolytic enzymes then convert packaged protein into an active neuropeptide and an associated neurophysin (9,22). Although pre-Golgi processing is believed to remove N-terminal peptides from both precursors and generate, in each case, prohormones with the hormone sequence at the N-terminus, this proposition still lacks direct evidence. In the current study, we have sought to provide this evidence by performing microsequencing to locate 3SS-cysteine incorporated into the structure of hypothalamic proteins isolated at different times following the addition of this radiolabeled amino acid. The incorporation of 3H-fucose was also evaluated. Protein and peptide isolations were performed by employing a one-step procedure of high performance liquid chromatography.

Protein microsequencing

from Charles River Laboratories (Wilmington, MA). Animals received food and water ad lib for one week before surgery. They were anesthetized by intraperitoneal injection of Nembutal (50 mg/kg body wt.), and 20 #Ci of 3SS-cysteine (New England Nuclear; 1004 Ci/mmol; l0 mCi/ml) or 3H-fucose (NEN, l0 Ci/ mmol; l0 mCi/ml) was injected bilaterally into each supraoptic nucleus (SON) over a 10-min period by a previously described method (34). However, the needles employed in our studies were bevelled away from the center of the brain to promote lateral flow of the injected material. At times l0 min, l h, 3 h, 6 h, or 16 h following commencement of the injection, animals were killed by decapitation, their brains and neural lobes removed, and these tissues frozen on solid CO2. Brains were sliced while still frozen (-10°C), and the SON and median eminence were obtained using the procedure of Palkovits (23). Each tissue was sonicated in 100 ul of 0.1 M HCI and homogenates spun at 10,000 X g for 4 min. Homogenates were frozen at - 2 0 ° C until they could be further processed for isolation of radiolabeled proteins and peptides. Some extracts were subjected to high performance chromatography. Others were either treated with acetone at a final concentration of 95% and the protein separated by centrifugation at 10,000 X g for 4 min, or titrated with 0.5 M NaOH to pH 7 and then subjected to immunoprecipitation.

METHOD

Incorporation of Radiolabel and Isolation of Products

High Performance Liquid Chromatography (HPLC)

Weight-matched Long-Evans (LE) rats and Brattleboro homozygotes (250-270 g) were obtained from Blue Spruce Laboratories (Altamont, NY). Male Wistar rats (240-280 g) were

HPLC was performed on a Beckman unit, model 332 plus model 100A pump, an Hitachi model 155-40 variable wavelength spectrophotometer with flow cell, a model BD 41 dual

395

396 pen recorder, and a model C-RIA Shimadzu Integrator recorder. The column employed was a 120 × 5 mm column of Nucleosil (ODS-silica) from Phenomenex (Rancho Palos Verdes, CA). The solvent gradient used was that described by Swann (34). To generate this gradient, solvent A was an aqueous solution of 0.2 M NaH2PO4 adjusted to pH 2.1 with H3PO4, and solvent B was 95% aqueous acetonitrile. The flow rate for separations was 1 ml/min, and fractions of 1 ml were collected on an LKB 7000 fraction collector into tubes. Each tube contained 200 ~zl of 1 mg/ml bovine serum albumin (BSA) as this was found to prevent loss of radiolabeled peptides. An aliquot (100 ul) of each fraction was used to evaluate 35S or 3H content by liquid scintillation. Another aliquot of 100-500 #1 from some separations was utilized for radioimmunoassay evaluation following adjustment of the pH to 7 with 1.0 M NaOH, evaporation under nitrogen, and reconstitution to 500 #1 with distilled water. Reference standards for HPLC were synthetic VP (Calbiochem, La Jolla, CA), synthetic oxytocin (Vega Biochemicals, Tuscon, AZ), and purified preparations of vasopressin-associated rat neurophysin (VPRNP) and oxytocin-associated rat neurophysin (OT-RNP) obtained in this laboratory from rat pituitaries (18). For a few collections from chromatography, the BSA in fraction tubes was omitted and the proteins represented by peaks of 35S activity were precipitated by the addition of 50% trichloroacetic acid (TCA). Pellets were washed with ether to remove the TCA and redissolved in 100 ttl of 100 mM Trizma HC1, pH 7.5, containing 1 mM EDTA plus 0.1% Triton X-100, for electrophoresis in SDS-polyacrylamide gels. A portion of some albumin-containing fractions was used for immunoprecipitation.

lmmunoprecipitation Immunoglobulin G proteins purified from rabbit antisera on columns of protein A-agarose were used for immunoprecipitation. These antisera were to VP-RNP, to OT-RNP, and to VP. They were raised in these laboratories and the characteristics of those to neurophysins have been reported (20,22). Antisera to VP-RNP and OT-RNP were used separately and in combination as a 1:1 mixture. Anti-VP displayed a 100% cross-reaction with vasotocin and a 0.1% cross-reaction with oxytocin when used in radioimmunoassay at a dilution of 1:100,000. IgG preparations from protein A-agarose were dialyzed (4 × 200 vol. H20), lyophilized, and used in a volume of 50 mM phosphate buffer (pH 7.6) equivalent to the volume of antiserum from which they were isolated. Fractions from HPLC were first evaporated under nitrogen, and then reconstituted in 100 ul of 10 mM phosphate buffer (pH 7.6) containing 1 mM EDTA and 0.1% Triton X100. IgG (anti-VP or anti-RNP) solutions were added to each chromatographic sample as a 50 #1 aliquot and incubation was allowed t o proceed for 24 h at 4°C. Incubated samples were diluted with 50 mM phosphate buffer to 500 tA, 60 ul of protein A-agarose added as a 1:3 suspension, and the mixture agitated for 4 h in an orbital shaker at 4°C. Protein A complexes were then pelleted by centrifugation at 10,000 × g for 2 rain. Each pellet was washed twice with a solution comprising 0.5 M KCI, 50 mM phosphate (pH 7.6), 5 mM EDTA, and 0.25% Triton X-100, then twice with 10 mM phosphate (pH 7.2) in 15 mM NaCI. The treated pellet was then suspended in sample buffer for electrophoresis that contained 10% (v/v) of 2-mercaptoethanol and 4.6% (w/v) sodium dodecylsulphate (SDS), and heated to 100°C for 10 rain.

SDS-Polyacrylamide Gel Electrophoresis The molecular weight of radiolabeled proteins was assessed by SDS-polyacrylamide gel electrophoresis employing a Bio-Rad

NORTH, O'CONNOR AND GONZ,~LEZ slab-gel apparatus. A separating gel of 15% acrylamide and a stacking gel of 4.75% acrylamide were used according to the method of Laemeli (15). Molecular weight markers were ~Clabeled proteins from Amersham International and proteins were visualized by fluorography (1).

Protein Sequencing 35S-Labeled proteins immunoreactive with antibodies to rat neurophysins were isolated as peaks of radioactivity from a HPLC separation performed on the SON extracts from 16 LE rats (killed one h after injection of 35S-cysteine). A 3H-labeled peptide was also isolated as a single radioactive peak from the HPLC eluant of four rat neural lobes 16 h following injection of label. Solutions were dried under vacuum, extensively dialyzed against distilled water, and the material retained by the membrane lyophilized. Bovine apomyoglobin (200 tzg, 12 nmol) was mixed with each protein in a 500 #1 solution of Trizma HC1 at pH 8.0 containing 8 M urea, 0.2% EDTA, and 0.7 mmol 2mercaptoethanol. Each solution was placed under an atmosphere of N2 for 4 h at room temperature. Iodoacetamide (1.44 mmol) in 100 #1 of 1 M NaOH was added, and the mixture left for a further 40 rain. Finally, 20 #1 (0.7 mmol) of 2-mercaptoethanol was added and the product was subjected to gel filtration on a column (100 × 1.0 cm) of Sephadex G-25 with 0.1 M acetic acid as eluant. Columns were preequilibrated with 30 ml of 0.1 M acetic acid containing 1 rag/m1 BSA, followed by 150 ml of 0.1 M acetic acid. The protein fraction containing S-alkylated product was collected and dried under vacuum in a Brinkman Sample Concentrator. Residues were reconstituted in 400 ul of 50% aqueous acetic acid, added to an 890C Beckman Protein Sequencer, and automated Edman degradation was performed employing a 0.5 M Quadral Program (19). PTC-amino acids from sequencing were dissolved in 1 M HCI and counted in a Packard Liquid Scintillation Counter. RESULTS

Radioactive Protein and Peptide Components in Neural Lobes Typical profiles for the HPLC elution of hormones and neurophysins from crude acid extracts of neural lobes from LE rats are shown in Fig. I. The same profiles were obtained for rats of the Wistar strain. Figure la represents the absorbance profile at 210 nm, Fig. l b the profile of radioactivity at 6 h after injection of 35S-cysteine into the SON, and Fig. lc the profile of radioactivity at 16 h after injection of 3H-fucose into the SON. Peaks of absorbance and radioactivity were identified as AVP, OT, and RNPs by performing RIA for each substance on aliquots of fractions from chromatography, and from the elution of purified and synthetic forms of these substances subjected to the same chromatographic gradient. The HPLC gradient was seen to clearly resolve AVP, OT, VP-RNP, OT-RNP, and VP-GL (vasopressin-associated glycopeptide) as peaks of absorbance and peaks of radioactivity. OT-RN1 r, a major metabolic form of OT-RNP (18), coeluted with OT-RNP (Fig. la,b,c). At 2 h, 3 h, 6 h, and 16 h following injection of 3SS-cysteine into the SON, the ratios of labeling for each hormone and its associated neurophysin in neural lobes were: 4.8, 5.8, 6.5, and 6.5 for VP-RNP/VP; and 8.3, 7.8, 7.4, and 7.4 for OT-RNPs/OT. Combining RIA data for VP-RNP with counts for incorporation of 35S-cysteine into this protein, the specific activities of VP-RNP in neural lobes at 2 h, 3 h, and 6 h postinjection were 13, 59, and 519 mCi/mmol. In addition to the neurophysins and hormones, one 35S-cysteine-1abeled peak eluted just prior to OT-RNP in the gradient. This material was

VASOPRESSIN A N D O X Y T O C I N P R E C U R S O R S

04

(o)

-

m i n o r labeled protein peaks eluted at 55 and 57 m i n of the gradient (Fig. 2a,b,c). At 16 h following injection of 3H-fucose into the SON, there was found a single peak of this fucose incorporated into a glycopeptide of neural lobe that eluted at 55 m i n of the gradient (Fig. lc). Microsequencing of this glycopeptide demonstrated that all of the radiolabeled fucose was located at position 6 from the N-terminus.

VP-NP

OT-NPI

0.3 -

¢P-GI

E

397

o

Radioactive Protein Components of the SON 0.2 -

o

HPLC analyses of supraoptic nuclear extracts taken 1 h after injection of 35S-cysteine to LE and Wistar rats showed incorporation of isotope into at least five proteins. The pattern of incorporation was identical for the two rat strains, and is illus-

g

0,~

-

L I

I

I

I

(o) I

h

1

,

I

.?

l0 20 30 40 50 60 70 80 time (rain.)

x E

(b)

29

VP-NP

2: I

VP 7

OT-NP

6

o

5

/•60

30 40 50 time (rain)

20

30 40 50 time (rain)

60

70

80

(b)

5O

-

1::: u

10 20

/

4

~40 ~30

?

420

_j,o

, ]1 d/

'

Io 20 30 40

LL_

50 60 70 80

time (rain.) (C)

VP-GL

I0

(c)

,? o

E ¢L

70

80

]

20 i3"

x

60

12 I0

20 30 40 50 60 ?0 80 time (rain.}

i

o

E

FIG. I. HPLC profiles of rat neural lobe extracts 16 h following the injection of 35S-cysteineor 3H-fucose into the supraoptic nuclei: (a) absorbance at 210 nm; (b) 35S radioactivity; and (c) 3H radioactivity. Separation was performed on a 120 × 5 mm column of Nucleosil (5 tz) and the gradient (shown in b) employed 0.2 M phosphate, pH 2.1, and 90% aqueous acetonitrile. a protein of mol.wt. 30,000, was not immunoreactively related to the neurophysins and hormones, was a glycoprotein because it was reversibly b o u n d to ConA-Sepharose, and was identified as a labeled protein in the SON. The peak appeared at 2 h after injection and slightly increased at 3, 6, and 16 h, making its slope of appearance nonparallel to that of neurophysins. Two

10

20

50

40

50

60

70

80

time (rain)

FIG. 2. HPLC profiles of rat neural lobe extracts at (a) 2 h, (b) 3 h, and (c) 6 h following injection of 3SS-cysteine into the supraoptic nuclei. Separation conditions were those given in Fig. I. Each profile represents the total activity of an equivalent of one neural lobe.

398

NORTH, O'CONNOR AND GONZALEZ (0) 8

OT-NP VP-NP VP-GL

4

: ~o

b 10 20 30 40

o

50

60 70 80

time (rain) 3

I

I0

I

2~0 30 40 50 time (min)

60

70

80

FIG. 4. HPLC profile of ConA-bound material from extracts of rat supraoptic nuclei removed 1 h following injection of 3SS-cysteine. The acid-soluble material shown was bound to a column of ConA-Sepharose at pH 7.0 and eluted with 0.5 M c~-methylmannoside. Separation conditions for HPLC were those given in Fig. 1.

(b)

,'~

5

o

4 5

,b

b

2

i

50 40

50 time (rain)

60

I

I

70

80

FIG. 3. HPLC profiles of extracts from rat supraoptic nuclei removed 1 h following injection of 35S-cysteine. (a) Long-Evans rats, and (b) homozygous Brattleboro rats. Separation conditions were those given in Fig. 1. Radioactive peaks are labeled 1, 2, 3, and 4 in order of their elution. trated for LE rats in Fig. 3a. In order of their elution, most of the radiolabeled proteins of peaks 2, 3, and 4 were immunoprecipitated with antibodies to RNPs and had respective mol.wt. by SDS-electrophoresis of 15,000, 19,000, and 21,000 daltons. Peak 1 was found to comprise two proteins of 19,000 and 16,000 daltons (Fig. 5). Of these four radioactive peaks in the eluate, the major peak of incorporation was, by far, peak 4. Both peaks 3 and 4 were conspicuously absent in extracts from DI rats (Fig. 3b). The proteins in peak 3 and 4 and either one, or both, proteins in peak 1 were glycoproteins because they were bound to ConASepharose and eluted with a-methyl-D-mannoside (Fig. 4). One h after the injection of 3H-fucose into the SON, the isotope was found incorporated into proteins of peak 3 and peak 4 (according to SDS-electrophoretic analysis) and these proteins appeared to be identical to those immunoreactive proteins into which 3sScysteine became incorporated, having mol.wt, of 19,000 and 21,000 daltons. They both fit the criteria for being the provasopressins or proneurophysins described in earlier studies (21,34). Almost all of the radiolabeled protein in peak 4 and about 40% of that in peak 3 was immunoreactive with antibodies to VP. Sequential analysis for positioning of 35S-cysteine residues in S-alkylated protein(s) of peak 4 is shown in Fig. 6a. Edman degradation showed 3SS-labeled carboxymethyl cysteine at positions 1, 6, 22, and 25, with an estimated percentage repetitive yield for this microsequencing of 97%. The labeled material of peak 3, from sequential analysis, appeared to comprise more than one protein. Radioactive PTC-amino acids were recovered

from positions 1, 6, 10, 13, 22, 24, and 25. The amount of 3SSlabeled PTC-amino acid recovered from positions 1 and 6 indicated a 96% repetitive yield of sequencing. However, the values for radiolabel at positions 10 and 13 indicated a 20% repetitive yield, while that from positions 21 and 24 suggested a 190% yield. Assessment of repetitive yields from positions 22 and 25 were obscured by the high counts from positions 21 and 24. These findings suggest the presence of three labeled proteins in peak 3: one protein with labeling at positions 1 and 6 (and possibly 22 and 25 like that protein of peak 4); one protein (a major form) with labeling at positions 21 and 24 from the N-terminus; and a third protein (a minor component) with labeling at positions 10 and 13 from the N-terminus. The labeled material from peak 2 of HPLC profiles (Fig. 3a, b) was not bound to ConA-Sepharose, was precipitated by antibodies to RNPs, and had a mol.wt, on SDS-electrophoresis of 15,000 daltons (Fig. 5). It was a dominant form in profiles representing SON extracts from DI rats at 1 h following injection of 35S-cysteine (Fig. 3b), and was therefore believed to contain a protein associated with the biosynthesis of oxytocin. Sequential analysis for the positioning of 3sS-label in S-alkylated protein(s) of peak 2 (Fig. 6b) revealed radioactive PTCamino acids liberated from positions 1, 6, 22, and 25 with an estimated 95% repetitive yield for microsequencing.

I

2

5

4

5 Mr x I0 - 3

--21 - - 19 -15 --12 FIG. 5. SDS-polyacrylamide gel electrophoresis of 35S-radiolabeledproteins from HPLC. Proteins in aliquots of peaks 1, 2, 3, 4 (Fig. 3a), and a region from the trailing profile of peak 4 (4t) were precipitated with TCA, redissolved in SDS sample buffer and loaded onto 15% polyacrylamide gels. Lanes 1, 2, 3, 4, and 5 represent peaks 1, 2, 3, 4, and 4t, respectively. Molecular weight values obtained by reference to a standard mixture of t4C-labeledproteins (Sigma Chemical Co.).

VASOPRESSIN AND OXYTOCIN PRECURSORS

3

,2L

(o)

3

(b)

2

E

O.

5 I

t t i ~ i I 5 10 I5 20 2530

1520 530 cycle

FIG. 6. Automated Edman degradation of 3SS-cysteine-labeledproteins of HPLC-isolated (a) peak 4 and (b) peak 2 (see Fig. 3a) showing the location of incorporated 35S-cysteine.Extracts from the supraoptic nuclei of 16 rats killed 1 h after injection of 35S-cysteinewere subjected to HPLC-isolated proteins S-alkylated, and sequencing was performed on an 890C Beckman sequencer. PTC-amino acids were counted for 3sS activity.

DISCUSSION The single-step HPLC isolation procedure plus sequential analysis employed in this study for radiolabeled proteins from rat supraoptic nuclei and rat neural lobe has provided the first structural information on the early events that occur in the processing of vasopressin and oxytocin precursors. It is well known from other studies (31) that a specific VP mRNA in rat hypothalamus translates a protein of 168 amino acids. This protein comprises sequences for VP, VP-RNP, an N-terminal peptide component of 23 amino acids, and a C-terminal peptide component of 39 amino acids with a glycosylation site. In processing, the protein is predicted to first lose the N-terminal peptide of 23 amino acids and acquire a carbohydrate side-chain, a process culminating in the addition of fucose. The resulting glycoprotein, provasopressin, is packaged into neurosecretory granules and is enzymatically converted within these granules to VP, VP-RNP, and a glycopeptide, through the intermediate formation of proneurophysin, a form of the glycoprotein that lacks the VP sequence and three additional amino acids (22). A specific OT mRNA translates a protein of 125 amino acids that includes the sequences of OT, OT-RNP, a 19 amino acid component at the N-terminus, and an arginine at the C-terminus. This smaller precursor lacks a glycosylation site. It is predicted to first lose the 19 amino acid component, and then be packaged into neurosecretory granules where enzymatic generation of OT and OTRNP takes place (12). Our data demonstrates that during the first hour, and as early as 10 minutes, two fucosylated glycoprotein precursors of vasopressin are generated in supraoptic neurons, and both of these have the vasopressin sequence at the N-terminus. Such a structure is dictated by the placement ofcysteine residues at positions 1, 6, 23, and 25, and demonstrates that indeed the entire 23 amino acid N-terminal sequence of the translated protein is rapidly lost in early processing. This finding supports the elegant work of Blobel and coworkers (32) on the selective action of a signal peptidase isolated from both eukaryotic and prokaryotic sources. By far, the major form of the two glycoproteins (provasopressins) has a molecular weight apparent (mol.wt.app) of 21,000 daltons, while the other is 19,000 daltons. The presence of fucose in these glycoproteins showed that addition and modification of the carbohydrate moiety was complete, an event synonymous with their arrival in the Golgi apparatus.

399 The 19,000 dalton provasopressin is apparently coeluted from HPLC with two other proteins. We have deduced that one of these has the placement ofcysteine at positions 10 and 13 from the N-terminus, and we believe this to be a form of neurophysin with an intact N-terminal structure (21). The third coeluted protein appears to have cysteine residues first appearing at positions 21 and 24 from the N-terminus. This protein is at present an anomaly, because even an intact translated sequence would have cysteines at positions 20 and 26. All three of these coeluting proteins and the predominant 21,000 dalton forms of provasopressin are absent in profiles from Brattleboro homozygotes, which points to their origin in vasopressin neurons. The 21,000 dalton and 19,000 dalton forms ofprovasopressin were reported in an earlier study by Swann and coworkers (34), but the 19,000 dalton protein was proposed from that study to be the major form at 1-h postinjection of tracer. However, the results of the present study indicate their findings were distorted by a large loss of provasopressin (especially the 21,000 dalton form) occurring in the additional gel filtration step that preceded HPLC analyses. Such a conclusion is supported by the much higher levels of label incorporated into proteins and peptides from both LE and Wistar rats found in HPLC profiles when the gel filtration step is omitted. A single form of oxytocin precursor was found following HPLC separation 1 h after radiolabel injection. Since this nonglycosylated protein had cysteine residues at positions 1, 6, 23, and 25, and was immunoprecipitated with antibodies to neurophysins, we believe it is pro-oxytocin with the oxytocin moiety at the N-terminus. This protein was also prominent in the hypothalamic extracts from Brattleboro homozygous rats. It demonstrates that the 19 amino acid peptide moeity is rapidly removed from the N-terminus of the protein translated by OT mRNA. It is not known if the pro-oxytocin isolated by HPLC still had an intact N-terminal arginine. The HPLC profiles obtained from neural lobes demonstrate that the single step isolation used here provides a clear resolution of vasopressin, oxytocin, oxytocin-related neurophysins, vasopressin-associated neurophysins, and vasopressin-associated glycopeptide(s). Fucose in vasopressin-associated glycopeptide was located at residue 6 from the N-terminus, showing the carbohydrate moiety was attached at this position as is predicted from the studies of Smyth and Massey (33). Because there are seven times the number ofcysteine residues in each neurophysin as is present in each hormone, a radioactive ratio of 7.0 for each neurophysin to its respective hormone would be expected in neural lobes following 3SS-cysteine injection if an equivalent number of moles of neurophysin and hormone were generated. The range of values from 4.8 to 6.5 for VP-RNP'VP and 7.4 to 8.4 for OT-RNP:OT is in good agreement with the formation of an equal number of moles of neurophysin and hormone by both vasopressin and oxytocin neurons. The lower specific activities of 35S-cysteine-labeled VP-RNP at times up to 16 h postinjection indicate that this labeled protein represented only a small percentage (