Expression of normal and mutant human pre-pro-insulins in Xenopus oocytes

Biochimie 70 (1988) 99-107 ©Soci6t6 de Chimie biologique/Elsevier, Paris

99

Expression of normal and mutant human pre-pro-insulins in Xenopus oocytes Kathleen I.J. SHENNAN and Kevin DOCHERTY*

Department of Medicine, University of Birmingham, Queen Elizabeth Hospital Birmingham B15 2TH, U.K. (Received 25-6-1987, accepted after revision 30-9-1987)

Summary - Conveniently situated Pstl sites were used to delete a major segment from the C-peptide coding region of a human pre-pro-insulin cDNA. The resultant mutant cDNA encoded a protein with the structure: pre-peptide B chain - A r g - A r g - G l u - A l a - G l u - A s p - L e u - G l n - L y s - A r g A chain. Normal and mutant human pre-pro-insulin cDNAs were used as templates for the synthesis of mRNA in a reaction catalysed by T7 RNA polymerase. The mRNAs were then microinjected into Xenopus oocytes to determine the effect of the deletion on the secretion of pro-insulin. When normal pre-pro-insulin mRNA was microinjected, pre-pro-insulin was processed to pro-insulin, which in turn was secreted into the media. When the mutant pre-pro-insulin mRNA was microinjected, however, mutant pro-insulin could be detected in the oocytes but at a much lower level than the normal pro-insulin. No mutant pro-insulin could be detected in the media. The stability of the mRNAs in the oocytes was investigated by microinjecting [32p]mRNA. 24 and 48 h after microinjection, the recovery of[33p]mRNA from the oocytes was 95 and 24% and 20 and 16% of that injected, for the normal and mutant mRNAs, respectively. In a cell-free translation system supplemented with dog pancreatic microsomal membranes, the pre-peptide was cleaved from the normal pre-proinsulin but not from nI J . rv ~ . n] [ .r~ ra v~ . .il lnl . 3ev ~l, lll lli .i n These ..... . . .~ .l .A. t.~. l k • "I, l ~ C-peptide plays an . . . . . the . . . . . .mHtant . . . . . . l l l * / I . ~ l l l J~' ll, O l f l l" , important role in the segregation of pro-insulin within and transport through the cellular secretory pathway. pre-pro-insulin / mutagenesis / sorting / pro-hormones

Introduction Almost all polypeptide hormones are synthesised as larger precursors, pre-pro-polypeptides, which undergo post-translational proteolysis during their transit through the cellular secretory apparatus [1, 2]. The first stage of processing occurs within the lumen of the endoplasmic reticulum where the pre-peptide is removed [3]. Pro-hormone conversion then occurs once the converting proteinases and

pro-hormones have been brought together for packaging in the trans-most cisternae of the Golgi apparatus and within the newly formed secretory granules [4, 5]. Our interest has been in the mechanisms involved in the intracellular sorting of proinsulin and its processing proteinases. Human pro-insulin consists of a 31 amino acid C-peptide linking the N-terminal 30 amino acid B chain and the C-terminal 21 amino acid A chain, and has the structure : B chain - Arg-Arg - C-pep-

*,4uthor to whom correspendence should be addressed. Abbreviations: SDS-PAGE: sodium dodecyl sulphate-polyacrylamide gel electrophoresis.

160

K.I.J. Shennan and K. Docherty

tide - L y s - A r g - A c h a i n . A l t h o u g h p r o - i n s u l i n

has been s h o w n to associate with intracellular m e m b r a n e s [6, 7], it is not clear w h e t h e r this process is involved in the sorting of pro-insulin into secretory vesicles. Xenopus oocytes represent a model system for studying pro-insulin sorting. W h e n microinjected with pre-pro-insulin e n r i c h e d m R N A , Xenopus oocytes will s y n t h e s i s e and secrete pro-insulin [8]. D N A m u t a g e n e s i s combined with the availability of m e t h o d s for synthesising pure m R N A in high yield from c D N A [9] provides an opportunity to introduce modifications into pre-pro-insulin m R N A and, w h e n microinjected into Xenopus oocytes, to test their effect on the secretion of pro-insulin. In this paper, we describe the effect of a major deletion in the C-peptide on the secretion of pro-insulin from Xenopus oocytes. A preliminary report on part of this work was previously published [10].

Materials and methods Animals Xenopus laevii were purchased from Xenopus Ltd., Redhill, Surrey, U.K.

Chemicals and reagents Plasmid pchil-19, containing a full length human pre-pro-insulin cDNA, was provided by Dr. G.I. ,-,~,.,t'^"Howard Hughes Medical Institute, University of Chicago; plasmid pT7 was purchased from United States Biochemicals through Strand Chemicals Ltd., Basingstoke, Hampshire, U.K.; i 7 RNA polymerase, nucleotide triphosphates, m7g (5')ppp(5')G cap structure analogue, bovine serum albumin (RNase ang DNase free) and RNase guard were from iZh:~rmacia, Milton Keynes, U.K.; calf intestinal phosphatase was from Boehringer, Lewes, Sussex, U.K.; RNase-free DNase (RQI DNase) was from P and S Biochemicals, Liverpool, U.K.; restriction enzymes were from Northumbria Biologicals, Cramlington, Northumberland, U.K. or from Bethesda Research Laboratories, Paisley, Scotland, U.K.; and dog pancreatic microsomes, L-[4,53Hlleucine (190 Ci/mmol), L-[2, 3, 4, 5, 6-3H]pheny lalanine (130 Ci/mmol) and uridine 5'-[a-32p]tri phosphate (800 Ci/mmol) from Amersham International, Amersham, Bucks, U.K.

DNA manipulations DNA manipulations were performed by standard procedures [11]. Large scale preparations of plasmids were by the alkali SDS method [12].

In vitro transcription In vitro transcriptions were performed in sterile 1.5 ml Eppendorf tubes in a final volume of 50 ~ 1, using a reaction mixture which contained 40 mM Tris-HC1 (pH 8.0), 15 mM MgCI2, 5 mM dithiothreitol, 1 mM each ofATP. UTP and CTP,0.1 mM GTP, 0.5 mM mTG(5')ppp(5')G, 0.25 mg/ml of bovine serum albumin, 200 units/ml of RNase guard, 40 l~g/ml of linearised DNA and 1400 units/ml ofT7 RNA polymerase. [32p]-Labeled transcripts were made by including ['~P]UTP (20 ~ Ci/ml) iri the reaction. After 60 min at 37°C, DNase (free of RNase) was added to a final concentration of 200 units/ml and the reaction mixture incubated for a further 10 mien at 37°C. The volume was adjusted to 200 ~ I with H20 and 5 ~1 of 0.5 M EDTA were added. The mixture was extracted once with phenol, once with phenol : chloroform:isoamyl alcohol (25:24:1) and once with chloroform:iso~,myl alcohol (24:1). RNA was precipitated using 0.I vol of 7 M ammonium acetate and 2.5 vol of absolute ethanol. Ethanol precipitation was repeated and the RNA resuspended at a concentration of 600/~g/ml in H20 and stored at -70°C. The yield of RNA was calculated from the extent of [32p]UTP incorporation.

Preparation and microinjection of Xenopus oocytes The removal of oocytes from female Xenopus laevis was performed under anaesthesia using 0.2% (w/v) m-aminobenzoate as described by Colman [13]. Oocytes were microinjected with 50 nl of mRNA (600 ~g/ml) and incubated overnight at 20°C in modified Barth's saline [13] to allow recovery from the puncture and to permit selection of healthy oocytes. Groups of 5 oocytes were then incubated at 20"C for the indicated time in a single well of a microtitre plate in 30 I~I of Barth's saline containing 5°/0 (v/v) dialysed foetal calf serum, 0.04°/0 (v/v) fungizone (Gibco), 0.4 units/ml of trasylol, and [3H]phenylalanine (1 mCi/ml). After incubation, the medium was removed and the eocytes washed 4 times in Barth's saline and homogenized in a glass on glass homogenizer (0.1 ml) in 20 mM Tris-HC1 (pH 7.6), 0.1 M NaCI, 1% (v/v) Triton X-100 and 1 mM phenylmethane suiphonyl fluoride. To determine the stability of the mRNA in the microinjected oocytes, [32P]-labeled 'mRNA was microinjected. Following incubation at 20"C for the indicated times, groups of 20 oocytes were washed twice in Barth's saline and homogenized in 1 ml of 10 mM Tris-HCl, pH 7.5, 1.5 mM MgCi2, 10 mM NaCI, l m g / m l of proteinase K and 2% (w/v) SDS. 'After 15 min at room temperature, the concentration of NaCI was adjusted to 0.3 M and the mixture extracted 3 times with phenol:chloroform:isoamyl alcohol (25:24:1) and 3 times with chloroform:isoamyl alcohol (24:1). RNA was precipitated with ethanol and re~uspended in 20~ l ofH~O prior to analysis by agarose gel electrophoresis.

Role of C-peptide in pro-insulin secretion Agarose gel electrophoresis ['-'P]-Labeled RNA was incubated for 15 min at 50"C in 10 mM NaH2PO4 (pH 6.8), 50% (v/v) dimethyl sulphoxide and 0.13 M glyoxal in a final volume of 16/d. Following addition of 4 ~! ofl0 mM NaH2PO4 (pH 6.8), 4 mg/ml of bromophenol blue and 50% (v/v) glycerol, the samples were applied to a 1.4% (w/v) agarose gel. Electrophoresis was at 60 V for 4 h in 10 mM NaH2PO4 (pH 6.8) with buffer circulation from cathode to anode. The agarose gel was then dried and autoradiography performed using Fuji RX film.

Immunoprecipitations Immunoprecipitations were performed using a monocional antibody raised against intact human pro-insulin which cross-reacts strongly with insulin [14]. This antibody was kindly provided by Professor C.N. Hales, Department of Clinical Biochemistry, University of Cambridge, U.K. Samples were diluted to 200 u I in 0.1 M Tris-HC1 (pH 7.6), 0.25% (w/v) bovine serum albumin, 0.05 M NaCI and 0.1% (v/v) Triton X-100. Cellulose bound monoclonai antibody (10/~i) and carrier cellulose (5/tl) were added and the sample incubated overnight at 4°C. The antigen-antibody complex was centrifuged at 12 000 × g for 30 s at room temperature in an Eppendorf centrifuge, washed 3 times in immunoprecipitation buffer, and analysed by SDS-PAGE.

SDS-PA GE Samples wer,, heated for 5 min at 100"C in 50 u l of 62.5 mM Tris-HC1 (pH 6.8), 2% (w/v) SDS, 0.25 M sucrose, 5% (v/v) #-mercaptoethanoi and 0.03% (w/v) bromophenol blue, and subjected to electrop~horesis in SDS-PAGE gels containing i5% (w/v) acrylamide and 0.09% (w/v) N,N'-methylenebisacrylamide, using the discontinuous buffer system of Laemmli [15]. Gels were fixed for 30 rain in 10% (w/v) trichloroacetic acid, 25% (v/v) isopropanol and prepared for fluorography [16]. Mr calibration was performed with a mixture of [~4C]-labeled proteins (Amersham) which included phosphorylase b, bovine serum albumin, ovalbumin, cart;onic anhydrase and lysosyme.

Cell-free translation Wheat germ extract was purchased from BRL, Paisley, Scotland and used according to the manufacturers instructions. Following incubation at 25°C for 60 min with [3H]leucine (300 u Ci/ml) in a volume of 30 l, 2 u l of the reaction mixture were added to SDS sample buffer and analysed by SDS-PAGE and fluorogra~hy. Nuclease-treated rabbit reticulocytes were prepared and used as described [l 1]. mRNA (6 t~g/ml) was incubated at 28"C for 90 min in a reaction mixture containing 400 u Ci/ml of [3Hlleucine in a final volume of 26 u I. Samples were immunoprecipitated and subjected to SDS-PAGE and fluorography.

101

Results Construction o f a mutant pre-pro-insulin eDNA containing a major deletion in the Cpeptide coding region A full length h u m a n pre-pro-insulin cDNA, isolated as a 511 bp fragment from plasmid pchil-19 [17] was ligated into the EcoRI site of plasmid pT7 and a recombinant with the prepeptide coding region adjacent to the T7 promoter was identified and designated pT7hppI-1. A major segment was deleted from the Cpeptide coding region of pT7hppI-1 using conveniently situated PstI sites (Fig. 1). Digestion of pT7hppI-1 with PstI generated an approx. 3 kb fragment consisting of the plasmid, the prepeptide, the B chain of insulin and a small part of the C-peptide coding regions of the eDNA; a 207 bp fragment consisting ofpart of the C-peptide coding region, the A chain coding region and part of the 3' untranslated region; and 3 small fragments of 48 pb and 27 bp from the Cpeptide coding region and 75 bp from the 3' untranslated region which included part of the pT7 polylinker. The 3 kb and 207 bp fragments were purified by electroelution from an agarose gel and, following treatment of the 3 kb fragment with calf intestinal phosphatase, the two fragments were annealed and ligated. The orientation of the 207 bp fragment was checked using the unique Pvuil site in the A chain coding region of the eDNA and the two Pvull sites of plasmid pT7. A recombinant with the structure shown in Fig. 1 was identified and designated pT7hppl-3. This manipulation resulted in joining two thirds of the codon for Leu 37to one third of the codon for Leu 62. The resultant protein had the structure" pre-peptide-B chain - A r g - A r g Glu-Ala-Glu-Asp-Leu-Gln-Lys-Arg - A chain; i.e, h u m a n pre-des-(38-62)pro-insulin. Plasmids pT7hppI-I and pT7hppl-3 were linearised by digestion with BamHI and HindIII, respectively, and used in an in vitro transcription reaction catalysed by T7 RNA polymerase to synthesise normal and mutant pre-pro-insulin mRNA. Both m R N A s contained a short sequence of 6 nucleotides preceding the transcription start site, the coding sequences of the normal and mutant pre-proinsulins and a 3' section containing the polyadenylation signal preceding a run of 41 A residues. The m u t a n t m R N A lacked 75 nucleo-

K.I.J. Sherman and K. Docherty

102

EcoRl

[

T~

~

& ~

Pstl PstI

Hlna III

I 68bp

/

•

/

Illl][[I

I

II

/

1, mS 27b.

07bp j:oex

hP

___/-o- \

'vutI

~

11[

y }kb and 207bp fragment Phosphatase t r e a t ] k b fraoment

~/

II I

•

I T0o Hind III

Fig. I. Construction o f t h e mutant pre-pro-insulin cDNA (pT7hppl-3). P, B, and A represent the pre-peptide, B chain and A chain coding regions of human pre-pro-insulin. The C-peptide coding region is denoted in vertical stripes. The thin line denotes the plasmid pT7-1 while the dark thickened line represents the T7 promoter and the 5' non-coding and the 3' non. . . . . . s rc~;ons U I t l l ~ pre-pro-insulin cDNA. The restriction enzyme sites used in the construction and characterization of the recombinant plasmid are shown.

tides from the C-peptide coding region and 48 nucleotides at the 3' end.

Translation of normal and mutant pre-proinsulin mRNAs in microinjected X e n o p u s oocytes To study the effect of the deletion on pro-insulin secretion, Xenopus oocytes were microinjected with mRNA generated from pT7hppl-1 or pT7hppl-3. The oocytes were then incubated in media containing [3H]phenylalanine, and proteins immunoprecipitated with the antipro-insulin antibody at various time intervals and analyse~l by SDS-PAGE. Phenylalanine was chosen for these experiments, since there were equal numbers of this amino acid in the

normal and mutant pro-insulins. In oocytes microinjected with the normal pre-pro-insulin mRNA, a major 9 kDa radiolabeled protein was observed (Fig. 2). The higher molecular weight proteins in Fig. 2 were present in water injected oocytes and were also brought down with nonimmune serum. They represent incomplete washing of the immunoprecipitates. By comparison with the pre-pro-insulin marker (track M 1) which was synthesised in a cell-free translation system, the 9 kDa protein was identified as pro-insulin. Pro-insulin levels increased within the oocytes during a 12 h incubation period and decreased during a subsequent 12 h incubation. This decrease coincided with an increase in pro-insulin in the media. The accumulation of pro-insulin in the media was not due to leakage of proteins from the oocytes,

Role of C-peptide #1 pro-insulin secretion

103

Fig. 2. Synthesis of normal and mutant pre-pro-insulins in microinjected Xenopus oocytes. Xenopus oocytes were microinjected with transcripts generated from pT7hppl-1 or pT7hppl-3 and incubated in media containing [~H]phenylalanine. At the indicated times, samples were immunoprecipitated using an anti-pro-insulin monoclonal antibody and analysed by SDSPAGE and fluorography. The track marked M contains molecular weight protein markers. [3H]Leucine-labeled proteins synthesised in a wheat germ cell-free system, using pTThppl-1 (M1) or pT7hppl-3 (M3) mRNA, were run as markers for the primary translation products.

since the cytoplasmic protein globin was present in cells but not in the media of oocytes microinjected in parallel with globin mRNA (data not shown). When Xenopus oocytes were microinjected with mutant pre-pro-insulin, mRNA generated from pT7hppl-3, a faint immunoprecipitable protein was detected, which by comparison with track M3 was identified as mutant proinsulin (Fig. 2). Although equal quantities of mRNA were microinjected, there was much less of the mutant pro-insulin than normal proinsulin. There was no detectable mutant proinsulin in the media. Fig. 3 shows the result of a

24 h incubation in [3H]phenylalanine ofoocytes microinjected with the normal or mutant mRNAs. The mutant pre-pro-insulin is clearly processed to mutant pro-insulin but there is much less of this than the normal pro-insulin. In addition, the mutant pro-insulin is not secreted into the media. These results were not due to differences in the ability of the antibody to recognise the pro-insulins, since similar results were observed when total media protein was run in the SDS-gel (data not shown). Moreover, the antibody brought down equal amounts (as a percentage oftrichloroacetic acid precipitate radioactivity) of normal and mu-

104

K.I.J. Sherman and K. Docherty

.28S

.18S

92

69

q6 31~

14.~

Fig. 4. Stability of normal and mutant pre-nro-insulin mRNAs in microinjected Xenopus oocytes. [~-'P]-Labeled transcripts generated from pT7hppl-I (tracks 1-3) or pT7hppl-3 (tracks 4-6) were microinjected into Xenopus oocytes and the RNA extracted and analysed by agarose gel electrophoresis and autoradiography, after incubation ofthe oocytes for 15 min (tracks I and 4), 24 h (tracks 2 and 5) and 48 h (tracks 3 and 6). The uninjected normal (track 7) and mutant (track 8) mRNAs are also shown. 28S and 18S ribosomal RNA size markers were run on the same gel in parallel and visualised by ultraviolet irradiation of an ethidium bromide stained section of the gel.

Stabifity of pre-l~ro-insulin mRNas in ,,,,,.,

Fig. 3. Synthesis of normal and mutant pre-pro-insulins in microinjected Xenopus oocytes. Xenopus oocytes were microinjected with transcripts generated from pT7hppl-I or pT7hppl-3 and incubated in media containing [3H]phenylalanine for 24 h. Oocyte homogenates and media were immunoprecipitated, using an anti-pro-insulin monocional antibody and analysed by S D S - P A G E and fluorography.

tant pre-pro-insulin synthesised in a cell-free system. It is also noteworthy that the translational efficiencies of the mRNAs were similar in wheat germ (compare tracks M1 and M3 of Figs. 2 and 3) and rabbit reticulocyte cell-free trans.,ation systems.

u,~,uc,~c.

A~llUlaV, S oocytes

One explanation for the reduced level of the mutant pro-insulin in the oocytes was that the mutant mRNA was less stable than the normal. To test this, [32p]-labeled normal and mutant mRNAs were microinjected into the Xenopus oocytes and RNA extracted and analysed by agarose gel electrophoresis (Fig. 4). The intensity of each band was quantified by scanning laser densitometry. After 24 and 48 h incubations, the recovery of [32P]-labeled RNA was 95 and 24 and 20 and 16% of that recovered immediately after microinjection (15 min) for the normal and mutant mRNAs, respectively. Thus, the reduced level of mutant pro-insulin could, to a large extent, be explained by differences in the stability of the microinjected normal and mutant mRNAs. Fig. 4 also shows the [32p]-labeled mRNAs prior to microinjection. The transcription reaction produced single mRNA species with the expected size difference.

Role of C-peptide in pro-insulin secretion

105

sin and chymotrypsin) in the presence or absence of Trito]z X-100 to show that the normal pro-insulin was segregated within the lumen of the endoplasmic reticulum (data not shown).

66 Discussion

45 •i ~

¸

14

Fig. 5. Cell-free translation of pT7hppl-I and pT7hppl-3 transcripts, pT7hppl-l (tracks l and 2) and pT7hppl-3 (tracks 3 and 4) transcripts were translated in a rabbit reticulocyte cell-free system in the absence (tracks l and 3) or provence

/'lrnek~ 9 ~nd

4) ~f d .........

t~...~; .......

I

membranes. Samples were immunoprecipitated using an anti-pro-insulin monocional antibody and analysed by SDS-PAGE and fluorography.

Cell-free translation of the normal and mutant pre-pro-insulin mRNAs The effect of the deletion on the segregation of the mutant pro-insulin within the secretory apparatus was studied in a cell-free translation system. In the absence of dog pancreatic microsomal membranes, the immunoprecipitable translation products encoded by pT7hppl-1 and pT7hppl-3 were of the expected size of normal pre-pro-insulin and the deletion matant pre-pro-insulin (Fig. 5). In the presence of dog pancreatic membranes, the normal prepro-insulin was processed to pro-insulin but the mutant pre-pro-insulin was not processed. These experiments were also performed with microsomal membranes and proteinases (tryp-

In order to elucidate the role of C-peptide in pro-insulin secretion, a mutant human pre-proinsulin cDNA was constructed which contained a major deletion in the C-peptide coding region. The effect of this deletion on the secretion of pro-insulin from microinjected Xenopus oocytes and on the proteolytic removal of the pre-peptide by dog pancreatic microsomal membranes was studied. Although the Xenopus oocyte does not contain a regulated secretory pathway [18] similar to that in pancreatic B cells [19], it nevertheless provides a suitable model system for investigating the effect of structural alterations within a secreted peptide on its ability to become segregated within the secretory pathway and to pass efficiently through the Golgi apparatus into secretory vesicles [20]. The synthesis of pre-pro-insulin in microinjected Xenopus oocytes confirmed the findings of Rappoport [8] that Xenopus oocytes were capable of processing pre-pro,insulin to proinsulin and secreting pro-insulin into the media. As expected, there was no processing of pro-insulin to insulin, despite recent studies [21] which have shown that Xenopus oocytes are capable of processing pro-albumin to albumin at pairs of basic amino acids. Our principal observation was that the amount of the mutant pro-insulin in the oocytes was considerably less than that of the normal pro-insulin, and that the mutant proinsulin was inefficiently secreted into the media. This was to a large extent explained by the differences in the stability of the microinjected mRNAs. The synthetic mRNAs shared structural similarities; both contained a monomethyl cap structure and a 41 residue poly A track, features which increase the stability and translational efficiency of mRNAs within oocytes [22]. The major difference was in the size of the transcripts, 539 and 416 nucleotides, respectively, and it was this which appeared to affect their stability. We have, however, shown similar differences in the secretion of the normal and mutant pro-insulins in transfected

106

K.I.J. Shennan and K. Docherty

monkey kidney cells (Shakur, Shennan and Docherty, manuscript in preparation). In this case, a polycistronic mRNA was produced [23] in which the deletion represented a small percentage of the total size of the mRNA. This suggested that processes in addition to the differences in stability of the mRNAs were responsible for the reduced levels of the mutant pro-insulin in the oocytes and heterologous cells. The cell free translation studies demonstrated that the mutant pre-pro-insulin was inefficiently segregated within the endoplasmic reticulum. The reduced level of pro-insulin in the oocytes may therefore have also resulted from inefficient segregation of newly synthesised material into the secretory system with resultant degradation by cytoplasmic proteinases. Some mutant pre-pro-insulin was processed to mutant pro-insulin within the oocytes, but the fact that this was not secreted suggested that degradation was also occurring, either within the endoplasmic reticulum or at some stage distal to the endoplasmic reticulum, i.e, by proteinases within the secretory pathway or through miscompartmentation into the lysosomal system. Mutant insulin precursors, including a mutant similar to that described here, have been expressed in yeast [24, 25]. In this case, the mutant des-(38-62)-pro-insulin was converted into insulin. There was no substantial difference in the quantity of insulin secreted from yeast transfected with the normal or mutant insulin precursors. In these experiments, however, the mutant pro-insulin was used as a fusion protein containing the first 85 amino acids of yeast mating factor a I connected to the amino terminus of the B chain. This construct encoded 142 amino acids, whereas the products of pT7hppI-I and pT7hppl-3 encoded 110 and 85 amino acids, respectively. The observation that the longer 142 amino acid mutant insulin precursor was efficiently transported through the yeast secretory pathway, whereas the 85 amino acid mutant pre-pro-insulin reported here was inefficiently processed ~o mutant proinsulin by dog pancreatic microsomal membranes, suggested that there was a minimum length (>85 amino acids) for ttte efficient translocation of pre-pro-insulin into the endoplasmic reticulum. Although the first 38 amino acids of pre-pro-insulin contain all the information required for binding to the endoplasmic reticulum membrane, and for translocation and pre-

peptide processing [26], the total length of the precursor also appears to be important. Thus, our results lend support to the hypothesis [27] that the major role of the C-peptide is to enlarge the pre-pro-insulin molecule to a suitable length for transport into the endoplasmic reticulum. Further studies are in progress on the preparation of pre-pro-insulin mutants of varying lengths to test their effect on the segregation of pro-insulin within the endoplasmic reticulum, and more subtle mutations are being introduced into the C-peptide region to define sequences involved in the efficient transport of pro-insulin through the secretory apparatus of Xenopus oocytes and transfected B cell lines. Such studies should provide valuable information on the mechanisms whereby pro-hormones and their converting proteinases are packaged into secretory vesicles.

Acknowledgments This work was supported by the Medical Research Council and the West Midlands Regional Health Authority. We are grateful to A. Colman for advice on the oocyte microinjections.

References |.

l - ] p r h o~r.t.

. . . . . .

!~ ,~, lv l .h, l.ov r ,l~! ~ ,a.

~ .

t~ llOp ~u ~9 ~ I (~ ."- La, , l l .2J u ,f i

1_") z--~.

2. Docherty K. & Steiner D.F. (1982) Annu. Rev. PhysioL 44, 625-638 3. Kreil G. (]982) Annu. Rev. Biochem. 50, 3]7-348 4. Steiner D.F., San Segundo B., Chan S.J. & Docherty K. (]984) in'Endocrinolo~, (Labrie F. & Proulx L., eds.), Elsevier-North Holland, New York, pp. 387-392 5. Orci L., Ravazzola M., Amherdt M., Madsen O., Vassail J.-D. & Perrelet A. (1985) Cell 42, 671-681 6. Noe B.D. & Moran M.N. (1984) J. Cell Biol. 99, 418424 7. Orci L., Ravazzola M. & Perrelet A. (1984) Proc. Natl, Acad. Sci. USA 81, 6743-6746 8. Rappoport T.A. (198 l) Eur. J. Biochem. 115, 665-669 9. Kreig P.A. & Melton D.A. (]984) Nucleic Acids Res. 12, 7057-7070 10. Docherty K., Phillips I.D. & Sherman K.I.J. (1986) Biochem. Soc. Trans. 14, 860 I I. Maniatis T., Fritsch E.F. & Sambrook J.. (1983) in: Molecular Cloning: A Laboratory Manual Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y., pp. 545 12. Birnboirn H.C. & Doly J. (1979) Nucleic Acids Res. 7, 1513-1523 13. Colman A. (1984) in "Transcription and Translation, A. Practical Approach (Haines B.D. & Higgins S., eds.), IRL Press, Oxford, pp. 271-302 14. Gray I.P., Siddle K., Frank B.H. & Hales C.N. (1987) Diabetes 36, 684-688

R o l e o f C - p e p t i d e in p r o - i n s u l i n secretion 15. Laemmli U.K. (1970) Nature (London) 227, 680-685 16. Bonner W.M. & Laskey R.A. (1974) Eur. J. Biochem. 46, 83-88 17. Bell G.I., Pictet R., Rutter W.J.. Cordell B., Tischer E. & Goodman H.M. (1980) Nature (London) 284, 26-32 18. Colman A. & Morser J. (1979)Cell 17, 517-526 19. Rhodes C.J. & Halban P.A. (1987) J. Cell Biol. 105, 145-153 20. Errington D.M., Bathurst I.C. & Carrell R.W. (1985) Eur. J. Biochem. 153, 361-365 21. Foreman R.C. & Judah J.D. (1987) FEBS Lett. 219, 75-78

i07

22. Drummond D.R., Armstrong J. & Colman A. (1985t Nucleic Acids Res. 30, 7375-7394 23. Kaufman R.J. (1985) Proc. NatL Acad. Sci. USA 82, 689-693 24. Thim L., Hansen M.T., Norris K., Hoegh 1., Boel E., Forstrom J., Ammerer G. & Fill N.P. (1986) Proc. NatL Acad. Sci. USA 83, 6766-6770 25. Thim L., Mogens T.H. & Sorensen A.R. (1987) FEBS Lett. 212, 307-312 26. Eskridge E.M. & Shields D. (1986) J. Cell Biol. 103, 2263-2272 27. Steiner D.F. (1984) Harvey Lect. 78, 191-228