Eur. J . Biochem. 150,47-52 (1985) l i ; FEBS 1985

Purification of and kinetic studies on a narrow specificity synaptosomal membrane pyroglutamate aminopeptidase from guinea-pig brain Brendan O'CONNOR and Gerard O'CUINN *

Department of Biochemistry, University College, Galway Department of Life Science, Regional Technical College, Galway

(Received January 7, 1985) - EJB 850004

A pyroglutamate aminopeptidase activity, distinct from that of cytoplasm, was released from a synaptosomal membrane preparation of guinea-pig brain by papain treatment. This activity was further purified 3560-fold relative to the homogenate with a yield of 17% by a procedure involving gel filtration chromatography, calcium phosphate cellulose chromatography and hydrophobic interaction chromatography on phenyl-Sepharose CL-4 B. The purified synaptosomal pyroglutamate aminopeptidase hydrolysed only thyroliberin, acid-thyroliberin, the luliberin N-terminal tripeptide (Glp-His-Trp) and, only slightly, Glp-His-Gly. No hydrolysis was observed with dipeptides containing N-terminal pyroglutamic acid (Glp) or with pyroglutamyl peptides containing more than three amino acids. A K, value of 40 pM was recorded when thyroliberin was used as substrate; however, luliberin was found to inhibit the hydrolysis of thyroliberin competitively with a Ki value of 20 pM.

Studies on soluble pyroglutamate aminopeptidases from a number of mammalian tissues [l - 51 and from bacteria [6 - 81 revealed that these enzymes are of low molecular mass, are inhibited by sulphydryl reagents and are capable of re- moving pyroglutamic acid from a wide range of peptides of different length and amino acid composition. A pyro- glutamate aminopeptidase activity has also been purified from serum of rat [9] and from serum of pig [lo, 111 which cleaves pyroglutamic acid from only thyroliberin or very closely related peptides. By contrast with the soluble cellular activity, the serum enzyme was of high molecular mass and was not inhibited by sulphydryl reagents. In this laboratory we have recently identified a guinea pig brain synaptosomal membrane bound pyroglutamate amino peptidase which, like the serum activities of rat and pig, has a substrate specificity restricted to thyroliberin or to closely related peptides [12]. This activity which can be released from synaptosomal membranes by treatment with papain has a relative molecular mass (230000) more comparable with that of the serum enzyme and is moreover unaffected by treatment with sulphydryl reagents. A neurotransmitter or neuromodulator role has been suggested for thyroliberin [13] and the presence of a thyroliberin-specific enzyme on synaptosomal membranes would appear to support that suggestion. In this paper the further purification of the papain-released synaptosomal aminopeptidase is described and this represents the first report of the purification of a particulate pyroglutamate aminopeptidase. Kinetic studies on the purified enzyme are also reported.

MATERIALS AND METHODS

Muter ials

Tris, the anorexogenic peptide (Glp-His-Gly), thyro- liberin, pyroglutamate aminopeptidase (bovine liver), pyro- glutamic acid and papain (papaya latex type 111) were ob- tained from Sigma Chemical Company (Poole, Dorset, Eng- land). Sephadex G-200, blue dextran and phenyl-Sepharose CL-4B were obtained from Pharmacia Fine Chemicals AB (Uppsala, Sweden).

Cellulose powder (CF I), cellulose phosphate P81 chro- matography paper, and Whatman no. 1 and no. 3 chroma- tography paper were obtained from Whatman (Maidstone, Kent, England). Ammonium persulphate, acrylamide, methy- lene bisacrylamide, tetramethylethylenediamine (TEMED), PAGE blue 83, imidazole, dipotassium hydrogen orthophos- phate, potassium dihydrogen orthophosphate and glycerol were obtained from B. D. H. (Poole, Dorset, England).

Glp-Pro and Glp-Ala were from Cyclo Chemical (Los Angeles, CA, USA).

Acid-thyroliberin, luliberin, neurotensin, bombesin, ele- doisin, bradykinin-potentiating peptide B, Glp-NH2, Glp- Val, Glp-Asn-Gly, Glp-Lys-Trp-Ala-Pro, Glp-Gly-Pro-Trp- Met-Glu, and bradykinin-potentiating peptide 9a were from ' Bachem Feinchemikalien (Bubendorf, Switzerland). Histidyl- proline diketopiperazine, Glp-His and Glp-His-Trp were obtained from U. C. B. Bioproducts (Brussels, Belgium). [Pro- 2,3,4,5-3H]Thyroliberin (100 Ci/mmol) was purchased from New England Nuclear.

Correspondence to G. OCuinn, Department of Life Sciences, Regional Technical College, Galway, Ireland

Abbreviation. The abbreviations used follow the recommendation of the IUPAC-IUB Commission on Biochemical nomenclature (1983) [Eur. J . Biochem. 138, 9-37 (1984)l. Glp, pyroglutamine acid i.e. 5-pyrrolidone-2 carboxylic acid.

Enzyme. Pyroglutamate aminopeptidase (EC 3.4.19.3).

Determination of pyroglutanzate uminopeptiduse activity

Pyroglutamate aminopeptidase was measured using [Pr~-~H]thyoliberin as substrate by a modification of the method of Bauer and Kleinkauf [4]. [Pr~-~H]Thyroliberin was diluted with unlabelled thyroliberin to give a specific activity of 0.125 Ci/mmol. 10 pl of each sample to be assayed was

48

added to 10 pl of diluted [Pr~-~H]thyroIiberin and 10 p1 of potassium phosphate buffer pH 7.4 at 37°C. The reaction was terminated by addition of 10 pl methanol and 10-pl aliquots were spotted onto prewashed cellulose phosphate P81 paper which was then developed using 1.0 M acetic acid. The start segments containing the highly basic His-ProNH2 were cut out and placed into scintillation vials to which were added 1 ml of 2 M ammonia and 10 min later 10 ml of toluene- based scintillation cocktail containing 33% Triton X-100. The conversion of thyroliberin could be measured by counting the vials in a scintillation spectrometer.

Protein determination

Protein was determined by the method of Lowry et al. [14] as modified by Hartree [15] and using bovine serum albumin as standard.

Preparation of synaptosomal membrane fraction

Synaptosomal membranes were prepared as described before [I21 except that sucrose was buffered by 100 mM potassium phosphate pH 7.4 rather than by imidazole.

Release of the pyroglutamate aminopeptidase from synaptosomal membrane

Synaptosomal membranes were centrifuged from the su- crose in which they were isolated, resuspended in 20mM potassium phosphate buffer pH 7.4 and incubated with papain (3.08 units/lO mg membrane protein) at 30°C for 60 min. The solubilised pyroglutamate aminopeptidase was separated from particulate matter by centrifugation at 100000 x g for 75 min.

top of a calcium-phosphate -cellulose column (1 8 x 0.7 cm) preequilibrated with 20 mM potassium phosphate buffer pH 7.4 containing 0.15 M KCI. The column was washed with 10 ml of the same buffer and the enzyme activity was elued with a linear phosphate gradient established between 15 ml of 20 mM potassium phosphate pH 7.4 containing 0.15 M KCl and 15 ml of 200 mM potassium phosphate pH 7.4 containing 0.15 M KCl at a flow rate of 10.8 ml/h. l-ml fractions were collected and assayed for pyroglutamate aminopeptidase as described above.

Hydrophobic interact ion chromatography on phenyl-Sepharose CL-4 B

The active fractions from the calcium-phosphate - cellulose column were pooled and dialysed against 100 mM potassium phosphate buffer pH 7.4 containing 0.15 M KCl for 8 h at 4°C. The dialysed sample was then applied to a phenyl-Sepharose column (3.0 x 0.7 cm) which had been previously equilibrated with the same buffer. The column was then washed with six column volumes of 100 mM potassium phosphate pH 7.4 containing 0.15 M KC1 and then with six column volumes of 30% glycerol in 100 mM potassium phosphate pH 7.4 containing 0.10 M KCl at a flow rate of 10 ml/h. The bound pyroglutamate aminopeptidase was then eluted by administration of six column volumes of 60% glyc- erol in 100 mM potassium phosphate pH 7.4 containing 0.05 M KCI at a similar flow rate; 0.6-ml fractions were collected and assayed for pyroglutamate aminopeptidase as before. The active fractions from the phenyl-Sepharose column were pooled and diluted with an equal volume of 100 mM potassium phosphate buffer pH 7.4 and stored in 0.3-ml aliquots at - 20‘ C under which conditions it was stable for six months.

Ge1,filtration on Sephadex G-200 High- voltage electrophoresis

The supernatant obtained by centrifugation at 100000 x g for 75 min was dialysed against distilled water for 4 h, lyophilised and reconstituted in 1 ml of 100 mM potassium phosphate pH 7.4 containing 0.1 M NaCI. This sample was applied to a Sephadex G-200 column (38 x 1.9 cm) pre- equilibrated with the same buffer and eluted with this buffer at a flow rate of 4ml/h. l-ml fractions were collected and assayed for pyroglutamate aminopeptidase as described above.

Stability oj’solubilised pyroglutamate aminopeptidase to frgezing and thawing in various buffers

The active fractions from gel filtration were pooled and dialysed against distilled water for 4 h. The sample was sub- divided into three aliquots and these aliquots were dialysed overnight against 100 mM potassium phosphate pH 7.4, 100 mM Tris/HCl pH 7.4 and 100 mM imidazole/HCl pH 7.4 respectively. Each aliquot was frozen at - 20°C for 16 h, rethawed and assayed for pyroglutamate aminopeptidase activity.

Chromatography on calcium-phosphate - cellulose

The substrate specificity of the purified pyroglutamate aminopeptidase was tested by incubation of the enzyme (50 pl) with 50 pl of a 2 mM solution of each peptide to be tested. High-voltage electrophoresis was used to separate the peptides tested from their degradation products. The samples were applied to Whatman no. 3 paper and a constant voltage (35 V/ cm) was applied for 1 h using the buffer system pyridine/acetic acid/water (1 : 10:89, v/v/v) pH 3.5. After electrophoresis the cleavage products where identified with Pauly’s reagent or by chlorination. Standards were also subjected to electrophoresis at the same time and used to identify cleavage products. Where degradation products were commercially unvailable the peptides to be tested were incubated with a commercially available bovine liver broad specificity pyroglutamate aminopeptidase (partially purified by filtration on Sephadex G-75) and the products obtained were used to identify the products from the action of synaptosomal pyroglutamate aminopeptidase on these peptides.

Kinetic studies on the purifiedpyroglutamate aminopeptidase

A stock dilution of [Pr~-~H]thyroliberin (0.5 pCi/4 nmol in 10 pl) was further diluted with distilled water to give a range of thvroliberin concentrations for use in kinetic studies.

The active fractions from gel filtration were pooled and IO-iI aliquots of each dilution was added to 20 p1 of 100 mM dialysed firstly against distilled water for 4 h and subsequently potassium phosphate buffer pH 7.4 containing peptide in- against 20 mM potassium phosphate pH 7.4 containing hibitor. In each case the reaction was initiated by addition 0.15 M KCI overnight. The sample was then applied to the of 10 p1 of purified pyroglutamate aminopeptidase. Separate

49

Fract ion number

Fig. 1 . Gelfiltration on Sephadex G-200. (0-0) Pyroglutamate aminopeptidase activity; (0 . . . . . 0) protein; (A ~ ~ ~ ~ A) blue dextran (detected by absorbance at 620 nm and used to determine void volume ( Vo). 1 .O-ml fractions were collected

blanks were used for each concentration of [ Pt-~-~H]thyro- liberin used. Because His-ProNH2 is reported to cyclise spontaneously and non-enzymatically at a constant rate of 1.5% min [4] to His-Pro diketopiperazine, which does not remain at the start segment, the radioactivity detected in the start segment was adjusted mathematically to take account of the cyclisation process.

Polyacrylumide gel electrophoresis

Electrophoresis was conducted on 5 - 27% gradient polyacrylamide slab gels using 89 mM Tris, 82 mM boric acid, 2.5 mM EDTA, 0.025% sodium azide pH 8.3 as both gel and electrophoresis buffer. Samples (25 - 50 pl), containing 15 - 25 pg of purified enzyme, 10% sucrose and 0.05% bromophenol blue, were applied to gels and electrophoresis was carried out in a Pharmacia GE-2/4 L 5 electrophoresis unit at a constant voltage (150 Vjcm) until the bromophenol blue marker reached the end of the gel. The gels were stained for 12 h with 0.2% PAGE blue in methanol/acetic acid/water (3 : 1 : 6, v/v/v) and destained in methanol/acetic acid/water (3 : 1 : 6, v/v/v).

RESULTS

Pyroglutamate aminopeptidase was released from synaptosomal membranes by papain treatment, separated from particulate matter by centrifugation, dialysed against distilled water, lyophilised, reconstituted in 100 mM potas- sium phosphate buffer pH 7.4 containing 0.1 M NaCl and applied to a Sephadex G-200 column. The enzyme eluted shortly after the void volume and well clear of the main protein peak (Fig. 1). Fractions 42 - 56 were pooled, dialysed against distilled water and subsequently against 20 mM potassium phosphate pH 7.4 containing 0.15 M KCI. In earlier experiments where isolation of the enzyme was conducted in imidazole buffer considerable losses in enzyme activity were experienced on freezing the enzyme to - 20°C at this stage. When the enzyme recovered from gel filtration was dialysed into potassium phosphate, Tris/HCl, and imidaz- ole buffers respectively, each at pH 7.4, frozen to - 20"C, and rethawed, it was found that the activity was best conserved in a phosphate buffer and that most unacceptable losses

Table 1. Storage and stability of partially purified papain-solubilised synupiosomal pyroglutamate arninopeptidiw

Buffer system in which enzyme was frozen at - 20°C for 16h and rethawed

Activity after freezing and thawing

% original

98 62 4

containing 50% glycerol 84

100 mM potassium phosphate pH 7.4 100 mM Tris/HCl pH 7.4

100 mM imidazole/HCl pH 7.4 100 mM imidazole/HCl pH 7.4

N

0 -

0 5 10 15 20 25 30 35 LO Fraction nurnbei

Fig. 2. Chromatography on calcium-phosphate - cellulose. (0-0) Pyroglutamate aminopeptidase activity; (n - - ~ ~ 0) protein; ( A-A) phosphate concentration. 1 .O-ml fractions were collected

occurred in imidazole buffer (Table 1). It was found that inclusion of 50% glycerol could prevent most of the harmful effect of freezing in imidazole buffer.

The enzyme activity recovered from gel filtration was applied to a calcium-phosphate - cellulose column and eluted with a linear phosphate gradient. The enzyme activity eluted from the column as a single peak (Fig. 2 ) and the fractions containing activity (18 - 24) were pooled, dialysed against 100 mM potassium phosphate pH 7.4 containing 0.15 M KC1 and added to a phenyl-Sepharose column. The enzyme bound tightly to the column and was eluted with 100 mM potassium phosphate buffer containing 60% glycerol and 0.05 M KCI (Fig. 3). Fractions containing activity (14- 16) were pooled, diluted with an equal volume of 100 mM potassium phosphate buffer pH 7.4 and stored at - 20°C in 0.3-ml ali- quots until further use. Under these conditions the enzyme was highly stable for six months. The purified enzyme was also stable at 15"C, losing only 5% of its activity in 7 days. One band was seen when the purified enzyme was run on gradient polyacrylamide gels. Table 2 gives a summary of the purification procedure and shows an overall purification of 3558-fold relative to the homogenate with a final yield of 17%.

The kinetics of hydrolysis of thyroliberin by the purified particulate pyroglutamate aminopeptidase was examined. As shown in Fig. 4 the hydrolysis of thyroliberin was found to obey normal Michaelis-Menten kinetics and K , and V values of 40 pM and 1.11 nmol miti-' ml-' respectively were obtained. The effect of the presence of a variety of peptides containing N-terminal pyroglutamyl residues on the hydroly- sis of thyroliberin by the purified enzyme was also examined

50

0.15

0.1c I

z 0 Y

I

-

0.05

C Fraction number

Fig. 3. Hydrophobic interaction chromatography on phenyl-Sepharose CL-4 B. (0-0) Pyroglutamate aminopeptidase activity; (0. . . '0) protein; (- - - - - -) KCI; (******) Glycerol. 0.6-ml fractions were collected

L 12 1 0 2 0 50 100 - -

€10 -

E a - c .- -

L

-10 1020 50 100 200 l / [Thyroliberinl ImM).'

Fig. 4. Inhibition qf the purified synaptosornal pyroglutarnate aminopeptidase by luliberin ( A ) or by Glp-His ( B ) . Lineweaver-Burke plots of data obtained by varying the concentration of thyroliberin (0.005-0.100mM) in the absence of inhibitor (0-0) and in the presence of 0.1 mM luliberin (A-A) or of 1 mM Glp-His (.--.I

Table 2. Purification ojguinea-pig brain synaptosomal pyroglutamate aminopeptidase One unit of pyroglutamate aminopeptidase activity is the amount which catalyses the hydrolysis of 1 nmol thyroliberin min-' at 3 7 T

Step Total Total Specific Purification Yield activity protein activity factor

units mg units mg-' %

1. Homogenate 74.27 816.00 0.091 1 100 2. Synaptosomal membrane preparation 40.37 71.76 0.563 6.2 54

4. Chromatography on Sephadex G-200 35.40 1.92 18.440 203.5 47.5 5. Chromatography on calcium-phosphate -cellulose 30.1 1 0.486 62.00 681.3 40.5

3. Papain-solubilised supernatant 37.11 13.50 2.749 30.2 50

6. Chromatography on phenyl-Sepharose 12.95 0.040 323.75 3558.0 17.0

Table 3. Activity of synaptosornal pyroglutamate aminopeptidase towards thyroliberin and other peptides C = competitive, NC = non-competitive, n. d. = not determined

Trivial name of peptide Sequence of N-terminal Number of Whether Type of Ki pentapeptide amino acids hydrolysed Inhibition

1. Luliberin 2. Thyroliberin 3. Neurotensin 4. Luliberin(1 -3) 5. Acid thyroliberin 6. 7. Eledoisin 8. Bradykinin-potentiating peptide 5a 9.

10. Anorexogenic peptide 11. 12. 13. 14. 15. Bradykinin-potentiating peptide 9a 16. 17. 18. Bombesin 19. Bradykinin-potentiating peptide B

Glp-His-Trp-Ser-Tyr- Glp-His-Pro-NH2 Glp-Leu-Tyr-Glu- Asn- Glp-His-Trp Glp-His-Pro Glp- Asn-Gly Glp-Pro-Ser-Lys-Asp- Glp-Lys-Trp-Ala-Pro His-Pro diketopiperazine Glp-His-Gly Glp-Gly-Pro-Trp-Met- Glp-Val Glp- Ala Glp-His Glp-Trp-Pro- Ala-Pro- Glp-Pro

Glp-Gln-Arg-Leu-GI y- Glp-Gly-Leu-Pro-Pro-

Glp-NHZ

10 3

13 3 3 3

11 5 2 3 6 2 2 2 9 2 1

14 11

C c C C C C C C C C C C C NC C n. d. n.d. n.d. n.d.

mM

0.020 0.042 0.085 0.170 0.400 0.750 0.880 1.050 1.45 1.65 1 . I5 2.05 3.85 4.50 6.61 n. d. n.d. n.d. n.d.

51

(Table 3). Each of these peptides competitively inhibited the hydrolysis of thyroliberin whether they were substrates for the enzyme or not. The one exception was Glp-His which inhibited the enzyme in a non-competitive manner (Fig. 4B). It is interesting to note that Glp-His is the common sequence at the N-terminus of both thyroliberin and luliberin and has a Ki value of 4.5 mM. The Ki values for the remaining peptides derived from Lineweaver-Burk plots ranged from 20 pM for luliberin (Fig. 4A) to 6.67 mM for bradykinin-potentiating peptide 9 a. Interestingly, His-Pro diketopiperazine, a cyclisa- tion product of His-ProNH2 which is formed by the action of the enzyme on thyroliberin, is also a competitive inhibitor with a Ki of 1.45 mM.

The results presented in Table 3 also show that hydrolysis yielding pyroglutamic acid occurred only when tripeptides, which contained pyroglutamic acid in the N-terminal position and histidine in the penultimate position, were presented to the purified enzyme. Biologically active peptides such as luliberin and neurotensin were not hydrolysed.

DISCUSSION

In a previous paper we showed the presence of a thyroliberin-specific pyroglutamate aminopeptidase in syn- aptosoinal membranes [I 21. These synaptosomal membrane preparations showed enrichment of 5’-ribonucleotidase and ouabain-sensitive Na+/K+-ATPase activities (markers for synaptosomal membranes) relative to crude extracts, these membrane preparations were prepared from a synaptosomal fraction which showed enrichment in occluded lactate dehy- drogenase activity as well as in 5‘-ribonucleotidase and in ouabain-sensitive Na/K+ -ATPase activities. We now show that a papain-released preparation of this enzyme could be purified by gel filtration chromatography, calcium-phos- phate - cellulose chromatography and hydrophobic inter- action chromatography. Because of the instability of the enzyme in iniidazole buffer during freezing, phosphate buffer was used throughout the purification. Of the limited range of pyroglutamate-containing peptides tested, only tripeptides containing histidine in the central position were hydrolysed. Pyroglutamate-containing dipeptides including pyroglutamyl histidine were not hydrolysed nor were pyroglutamate- containing peptides with more than three amino acids. Although this study was limited by the commercial availability of pyroglutamate-containing peptides, the results suggest that the enzyme specificity is restricted to tripeptides containing N-terminal pyroglutamate and with histidine in the penulti- mate position. The possibility cannot be ruled out, however, that the enzyme may also hydrolyse tetrapeptides or pentapeptides which commence with N-terminal pyroglutamyl histidyl residues. Interesting similarities have previously been noted between the present enzyme [I21 and a pyroglutamate aminopeptidase activity of rat serum [9] and of porcine serum [lo, 111 in respect of substrate specificity, molecular mass and susceptibility to inhibitors. A previous study [16] has also indicated that the synaptosomal pyroglutamate aminopeptidase and the porcine serum pyroglutamate aminopeptidase [lo] both display optimum activity at pH 7.4 and are both similarly affected by a number of metal ions. Kinetic studies using a range of peptides in- dicated that, in general, naturally occurring biologically active peptides containing N-terminal pyroglutamic acid displayed the lowest Ki values while pyroglutamic-acid-containing dipeptides showed the highest Ki values. With the exception

of pyroglutamyl histidine, which displayed non-competitive inhibition, all other peptides tested inhibited the hydrolysis of thyroliberin in a competitive manner. In a previous study [12] we showed that the synaptosomal membrane-bound pyroglutamate aminopeptidase activity displayed the same sensitivity to inhibitors and substrate specificity as pyroglutamate aminopeptidase preparations solubilised from the membrane by either papain or Triton X-100, suggesting that these releasing agents do not significantly alter the cata- lytic properties of the enzyme. The one difference in substrate specificity between the membrane-bound pyroglutamate aminopeptidase and the pyroglutamate aminopeptidase activity solubilised by either Triton X-100 or papain and partially purified on Sephadex G-200 arose in the case of luliberin. I t was noted that while luliberin was not a sub- strate for a partially purified synaptosomal membrane pyro- glutamate aminopeptidase preparation, incubation of luliberin with synaptosomal membrane fractions produced a number of degradation products including pyroglutamic acid. In view of the fact that Glp-His-Trp, i.e. luliberin(1- 3), is a substrate for the present enzyme, it is possible that release of pyroglutamic acid from luliberin by synaptosomal mem- branes is mediated by the action of synaptosomal pyroglut- amate aminopeptidase on luliberin degradation products. Most interesting was the observation that luliberin, while not a substrate for the synaptosomal pyroglutamate aminopeptid- ase, had the lowest Ki value recorded (Ki = 20 pM) lower than the enzymes K,,, for thyroliberin (K, = 40 pM). This result closely mirrors that achieved with the porcine serum pyroglutamate aminopeptidase [ll] and is predicted by a study of thyroliberin degradation by rat hypothalamus [I 71. These observations strengthen the suggestion that the synaptosomal membrane pyroglutamate aminopeptidase and the serum pyroglutamate aminopeptidase may be derived from the same gene. The restriction of the specificity of the synaptosomal membrane enzyme to thyroliberin and very closely related peptides indicates a special function for thyroliberin in the synaptosomal area. While luliberin is not hydrolysed by the enzyme its affinity for the enzyme is perhaps indicative of a modulatory role for luliberin in this function.

We wish to achknowledge helpful discussions with Dr John Donlon and Dr Tadhg Griffin. We thank the Medical Research Coun- cil of Ireland for awarding a grant-in-aid to Gerard O’Cuinn and the Department of Education (Dublin, Ireland) for awarding a research maintenance grant to Brendan O’Connor. We also wish to thank Patricia McLaughlin for typing the manuscript.

REFERENCES 1. Szewczuk, A. & Kwiatkowska, J. (1970) Eur. J . Biochem. I S , 92-

2. Armentrout, R. W. (1979) Biochim. Biophys. Acta 191,756-759. 3. Mudge, A. & Fellow, R. E. (1973) Endocrinology 93,1428 - 1434. 4. Bauer, K. & Kleinkauf, H. (1980) Eur. J . Biochem. 106, 107-

5. Browne, P. & O’Cuinn, G. (1983) Eur. J . Biochem. 137, 75 - 87. 6. Armentrout, R. W. (1969) Arch. Biochem. Biophys. 132, 80-90. 7. Fujiwara, K., Kobayashi, R. & Tsuru, D. (1979) Biochim. Bio-

8. Fujiwara, K., Kobayashi, R. & Tsuru, D. (1981) Biochim. Bio-

9. Taylor, W. L. & Dixon, J. E. (1978) J . Biol. Chem. 253, 6934-

96.

117.

phys. Acts 570, 140- 148.

phys. Acta 655, 10- 16.

6940. 10. Bauer, K. & Novak, P. (1979) Eur. J . Biochem. 99, 239-246. 11. Bauer, K., Novak, P. & Kleinkauf, H. (1981) Eur. J. Biochem.

118, 173-176.

52

12. O’Connor, B. & O’Cuinn, G. (1984) Eur. J . Biochem. 144,271 -

13. Guillemin, R. (1978) Science (Wash. DC) 202, 390-402. 14. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J.

15. Hartree, E. F. (1972) Anal. Biochem. 48,422-427. 16. O’Connor, B. & O’Cuinn, G. (1985) Biochem. Soc. Trans., in the

17. Griffiths, E. C., Kelly, J. A., White, N. & Jeffcoate, S. L. (1980)

278. press.

Acta. Endocrinol. 93, 385 - 391. (1951) J . Biol. Chern. 193, 265-275.