expression of the two genes

Proc. NatL Acad. Sci. USA

Vol. 79, pp. 2803-2807, May 1982 Biochemistry

Biosynthesis of rat insulins I and II: Evidence for differential expression of the two genes (immunoelectrophoresis/electroblotting/photoaffinity linkage/growth hormone-tumor--bearing rats)

KEIJI KAKITA, STEPHEN GIDDINGS, AND M. ALAN PERMUTT* Department of Internal Medicine, Metabolism Division, Washington University School of Medicine, 660 South Euclid Avenue, St. Louis, Missouri 63110

Communicated by David M. Kipnis, January 22, 1982

ABSTRACT The purpose of these experiments was to determine the relative content and biosynthetic rate of insulins I and II under various experimental conditions. The two insulins were quantitated by polyacrylamide gel electrophoresis, electrotransfer to nitrocellulose paper, photoaffinity crosslinking, and immunodetection with anti-insulin antibody and '5I-labeled protein A. The ratio (mean ± SEM) of insulins, I/HI, was 1.2 ± 0.2 in Wistar-Furth rats fasted for 4 days, 1.6 ± 0.2 in normal rats, and 5.5 ± 0.8 in growth hormone-tumor-bearing hyperinsulinemic rats (P < 0.01). The increase in content of rat insulin I compared to II in the growth hormone-tumor-bearing animals was confirmed by radioimmunoassay of gel slices. To determine whether the difference in contents of rat insulins Iand El in the hyperinsulinemic rats was due to increased biosynthesis or a different turnover rate, isolated rat islets were incubated in [3H]leucine for 4 hr with 5.5 mM or 16.0 mM glucose in the incubation medium. Glucose stimulated insulin biosynthesis >8-fold. The ratio of synthesis of rat insulin I relative to II was 0.9 ± 0.1 at 5.5 mM glucose and 9.8 ± 3.3 (P < 0.01) at 16.0 mM glucose. Therefore, under conditions that stimulate insulin biosynthesis, there was a marked preferential synthesis of rat insulin I relative to H. These studies suggest that the two rat insulin genes are expressed independently and that, under stimulatory conditions, there is preferential expression of the rat insulin I gene. In rats (1, 2), mice (3, 4), and certain fish (5, 6), two different insulins have been found. The rat insulins have been shown to be encoded by two nonallelic genes separated by at least 9,000 base pairs of DNA (7, 8). The mRNAs are quite similar, approximately 93% homologous in the coding regions with only 34 of 439 nucleotides different (9, 10). Preproinsulins I and II, the initial translation products, differ by three amino acids in the preregion (7-11), two in the B chain, and two in the C peptide (see Fig. 1). Although the structures of the two genes, RNAs, and preproinsulins have been completely determined, little is known about the regulation of expression of the individual genes. In insulin isolated from Sprague-Dawley rat pancreas by Clark and Steiner (2), analysis by polyacrylamide gel electrophoresis indicated 58% insulin I and 42% insulin II. Biosynthesis of rat insulins I and II has been shown to occur in a ratio of 60:40 in several studies (2, 12-14). The present experiments were initiated in the course of the development of an immunoelectrophoretic method for studying pancreatic extracts of insulin. We noted, in a hyperinsulinemic rat, that the ratio of rat insulins, I/IL, appeared to be greater than 5:1, a value very different from that previously reported. We therefore decided to do a more complete analysis of the proportion of rat insulins I and II under various experimental The publication costs ofthis article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisenent" in accordance with 18 U. S. C. §1734 solely to indicate this fact.

conditions and to examine their relative rates of synthesis. Our analysis confirmed that both the steady-state content and synthesis of rat insulin I relative to II was at a ratio of approximately 60:40 in the control animal but there was a marked increase of content and synthetic rate of insulin I relative to II in a hyperinsulinemic rat model. METHODS Extraction of Insulin. Female Wistar-Furth rats (GIBCO) either normal or bearing the subcutaneously transplanted pituitary tumor MtTW15 (15, 16) of 4 weeks' duration-or Sprague-Dawley rats (male and female), body weight 180-200 g(Eldridge Laboratory Animals, St. Louis, MO) werefed Purina rat chow ad lib until the time of the experiments. (The Wistar-Furth rats were kindly supplied by Ida Mariz and William Daughaday.) Pancreatic tissue was homogenized in 70% acid/ alcohol with a Polytron homogenizer at setting 4 for 1 min, extracted at 4°C for 20 hr, and centrifuged at 2,000 x g for 30 min. Insulin was partially purified essentially by the method of Davoren as described (17). The supernate was decanted, the tissue was reextracted, and extracts were combined, adjusted to pH 8.3 with ammonium hydroxide, and centrifuged at 2,000 x g for 20 min (4°C). The supernate was adjusted to pH 5.3 with HC1, after the addition of 0.025 ml of 2 M ammonium acetate solution per ml of supernate. To the extract were added slowly 15 ml of cold 95% ethanol and 50 ml of diethyl ether per 10 ml of the extract, and the mixture was kept at 4°C for 20 hr. The precipitate was collected after centrifugation (2,000 X g, 30 min at 40C), dried under reduced pressure, and dissolved in an adequate volume of water. Immunoreactive insulin was measured in a double-antibody radioimmunoassay as described (17). Protein determinations were performed by the method of Lowry et al (18). Polyacrylamide Gel Electrophoresis. Electrophoresis at pH 8.9 in'12.5% slab gels was performed according to the method of Davis (19) as described (17). After fixation in 50% isopropanol/20% trichloroacetic acid for 25 min, the gels were stained in 0.1% Coomassie blte (G250) in 50% methanol/7% acetic acid for 1 hr and destained in 10% acetic acid. Immunoelectrophoretic Detection of Insulin. An immunoelectrophoretic method (to be described elsewhere) for analysis of insulin has been devised by modification of several methods (20-22). The method is essentially as follows. After electrophoresis, electrophoretic transfer was performed onto presoaked nitrocellulose filters (0.45 ,um, Schleicher & Schuell) in 25 mM Tris/192 mM glycine, pH 8.3, at 4-5 V and 70 mA for 4 hr at room temperature. After transfer, the filter was soaked in 4% bovine serum albumin in 10 mM phosphate buffer (pH 7.4) for Abbreviation: GHTB, growth hormone-tumor-bearing. * To whom reprint requests should be addressed.

2803

Proc. Natl. Acad. Sci.- USA 79 (1982)

Biochemistry: Kakita et al.

2804

48 hr. Incorporation of [3H]leucine into total protein and newly synthesized proinsulin and insulin was measured by using guinea pig anti-insulin antibody as described (24). The rest of the sample solution was lyophilized, dissolved in H20, and analyzed by polyacrylamide gel electrophoresis (pH 8.9) at 200 V for 2.5 hr at room temperature. After electrophoresis, the gel was stained, destained, and subjected to fluorography as described' (17). RESULTS Pancreatic Content of Rat Insulins I and II in Fasted, Fed, and Growth Hormone-Tumor-Bearing Rats. Rat insulins I and II are readily separable by polyacrylamide gel electrophoresis at pH 8.9 because rat insulin I contains an additional basic residue at position B29 (Fig. 1B). Growth hormone-tumor-bearing (GHTB) rats have a 4-fold increase in pancreatic insulin content and marked hyperinsulinemia (refs. 15 and 16; Table 1). Extracts of pancreas from control and GHTB rats were electrophoresed on polyacrylamide slab gels and stained with Coomassie blue (Fig. 2A). The mobilities of rat insulins I and II were determined by slicing the gels, extracting the insulin, and measuring it in a radioimmunoassay (see below). There was a marked increase ofinsulin I relative to II in the GHTB rats compared to control Wistar-Furth or Sprague-Dawley rats. Densitometric tracings of the stained gels are shown in Fig. 2C. Analysis of multiple samples indicated a mean (± SEM) ratio of I/I1 of 1.0 ± 0.1 in control rats and 7.7 ± 1.0 (P < 0.01) in GHTB rats (Table 1). Because we could not be certain that the two rat insulins stained equally well with Coomassie blue, further analysis by immunoelectrophoresis and radioimmunoassay was performed.

1 hr at 409C. Photolinkage was performed by adding enough (N-hydroxysuccinimidyl-p-azidobenzoate (Pierce Chemical, Rockford, IL) in dimethyl sulfoxide (Sigma) to give a 1:50 dilution and then incubating the mixture for 2 min in a darkened room at 40C. The paper was photolyzed for 10 min with longwave ultraviolet light (Black Ray B 100A; Ultra-violet Products, San Gabriel, CA) at a distance of 15 cm from the surface of the paper. Tris HCl (20 mM; pH 7.5) was then added to prevent further chemical reaction. The paper was washed twice with 10 ml of 10 mM phosphate buffer (pH 7.4) and incubated with 1:10,000 final dilution of guinea pig anti-insulin antiserum in the same phosphate buffer for 3 hr at 370C. The photolinking was performed again after a washing with buffer. The paper was washed twice and incubated with 10 uCi (1 Ci = 3.7 X 1010 becquerels) of "2I-labeled protein A (Pharmacia) per gel [iodination as described (22)] for 1 hr at room temperature. The paper was soaked in 1% Triton X-100 for 1 hr at room temperature, dried, and exposed to Kodak X-Omat x-ray film for 15-20 hr at -800C. The absorbance of the stained gels or film image was measured by using a Joyce-Loebl microdensitometer (MKIIIC) with a wedge range absorbance of 0.79-3.76. The absorbance of the same film in Coomassie blue-stained gels was measured by using a soft laser densitometer (BioMed). Biosynthetic Study. Rat islets were isolated by collagenase digestion (23) and were incubated (25 islets per tube) in Krebs' bicarbonate buffer containing 5.5 mM or 16.0 mM glucose and 1.2 mCi of [3H]leucine, for 4 hr at 37C in 95% 02/5% CO2. Islets were washed with 1.5 ml ofHanks' Hepes buffer (25 mM, pH 7.4) and centrifuged, the supernate was removed, and the islets were extracted with 200. ,l of 70% acid/alcohol at 4°C for Base pairs x 10-2

A

0 1V5

5 co

UT

6

5

4

7

8

(119)

IUT IVS I (119)

5'cap

Pro

c

C

B

rVs 2

B

3

2

1

A

'UT'

A

UT

31

I

3

(499)

B-chain

I phe val lys gin his leu cys gly pro his leuwvol glu ala leu tyr leu vol cys gly glu arg gly phe phe tyr thr pro lys ser ser

I

met

C-peptide I (arg arg) glb vol glu asp pro gin vol pro gin leu glu leu gly gly gly-pro glu ala gly asp lku gin thr leu ala leu glu val ala arg gin (ys arg)

gly

ala

A-chain I gly ile val asp gin- cys cys thr ser ile cys

ser

leu tyr gin leu glu asn tyr cys asn

FIG. 1. Structure of the two rat insulin genes (7, 8), mRNAs (9, 10), and proinsulins (6). (A) Diagram of insulin I and II genes from the 5' to the 3' end, including the intervening sequences (IVS 1 and 2) which do not appear in the mature mRNA. cap, the 7-methyl-guanosine on the 5' end of the mRNA;,UT, untranslated portion of the mRNA; pre, prepeptide; B, C, and A, B chain, C peptide, and A chain, respectively. (B) Primary sequence-of rat proinsulins I and II along with the four basic amino acids that are removed in conversion to insulin and C peptide. The net charge on the two proinsulins isthesame because a glutamic acid and a lysine in rat proinsulin I are not in rat proinsulin II. On the other hand, rat insulin I has one more positive charge than rat insulin II.

Biochemistry: Kaldta et aL

Proc. Nad Acad. Sci. USA 79 (1982)

2805

Table 1. Insulin content and insulin I and II distribution in fasted normal, normal, and GHTB rats Wistar-Furth Sprague-Dawley Normal GHTB Normal Fasted normal 5.5 ± 0.4 (3) 4.9 ± 0.5 (6) 4.6 ± 0.3 (14) Protein content,* mg/g of pancreas 32.7 ± 3.0 (6) 130.9 ± 9.3t (14) 9.8 ± 0.6t (3) MIR content, gg/g of pancreas 6.7 + 0.4 (6) 28.5 ± 1.6t (14) 1.8 ± 0.2t (3) IRI concentration, ug/mg of protein Insulin distribution Coomassie blue staining: I, % 49.2 ±_3.4 (3) 88.3 ± 1.3 (3) 54.8 ± 1.8 (3) 50.8 ± 3.4 (3) 11.8 ± 1.3 (3) 45.2 ± 1.8 (3) II, % 7.7 ± 1.3t (3) 1.2 ± 0.1 (3) 1.0 ± 0.1 (3) I/It Immunoelectrophoresis: 61.5 ± 2.0 (3) 81.8 ± 2.0 (13) 56.9 ± 2.1 (6) 52.5 ± 3.3 (5) I,% 38.5 2.0 (3) 18.2 ± 2.0 (13) 43.1 2.1 (6) 47.5 3.3 (5) II, % 1.4 ± 0.1 (6) I/I 5.5 ± 0.8t* (13) 1.2 ± 0.2t (5) 1.6 ± 0.2 (3) Radioimmunoassay: 52.6 ± 3.1(3) 84.6 ± 2.1 (5) I,% 47.4 ± 3.1 (3) 15.4 ± 2.1 (5) II, % I/II 1.1 ± 0.1 (3) 6.2 ± 1.3 (5) IRI, immunoreactive insulin determined by radioimmunoassay (17). The number of rat pancreases evaluated for each condition is indicated in parentheses. Statistical analysis was by Student's nonpaired t test. Results are shown as mean + SEM. * Acid/alcohol extractable protein. tP < 0.01 for difference from Wistar-Furth normal. tP < 0.05 for difference from Wistar-Furth normal.

Biosynthesis of Rat Insulins I and II. To determine whether the difference in content of rat insulins I and II in the GHTB rats was due to preferential synthesis, rather than to a difference in the rate of turnover, isolated rat islets were incubated for 4 hr in the presence of [3H]leucine. After incubation, the islets were extracted, and incorporation of 3H into total islet protein and into immunoprecipitable proinsulin and insulin was determined. Increasing the glucose in the medium from 5.5 mM to 16.0 mM increased proinsulin and insulin synthesis about 8-fold (Table 2). The proportion of the total protein synthesis accounted for by proinsulin and insulin increased from approximately 10% to 30%. To determine what fraction of the newly synthesized insulin was accounted for by insulins I and II, [3H]leucine-labeled extracts were electrophoresed on polyacrylamide slab gels and the radioautographs were subjected to densitometric analysis (Table 2). At 5.5 mM glucose, the I/

After electrophoresis of pancreatic extracts, the proteins were transferred to nitrocellulose paper, covalently bound, and then treated with anti-insulin antibody. Binding of anti-insulin antibody to insulins I and II was detected with '"I-labeled-protein A. A radioautograph of the gel from Fig. 2A is shown in Fig. 2B. Densitometric tracings of radioautographs from fasted normal, normal, and GHTB rats yielded insulin I/II ratios of 1.2 ± 0.2 in the fasted normal rat, 1.6 ± 0.2 in the normal rat, and 5.5 ± 0.8 in the GHTB rat (Table 1). These ratios determined by immunoelectrophoresis were confirmed by electrophoresis of pancreatic extracts, slicing the gel and extracting insulin, and measuring insulins I and II by standard radioimmunoassay. A summary of the insulin content determined by immunoelectrophoretic analysis is shown in Fig. 3. Although both rat insulins I and II increased in the GHTB rats, there was a marked increase in insulin I content relative to insulin II.

A

B 1

2

3

C 1

2

3

InsI

Ins Ht W.i..

-.

I'

a'

.*-Ins ItIo-

FIG. 2. Electrophoresis of pancreatic extracts on polyacrylamide gels. (A) Coomassie blue-staining patterns of electrophoresed pancreatic extracts. Lanes: 1, Wistar-Furth GHTB rat; 2, normal Wistar-Furth rat; 3, normal SpragueDawley rat. (B) Immunoelectrophoretic detection of insulins I and II. The gel from A was treated with anti-insulin antibody and then with'mI-labeled protein A, and a radioautograph was obtained. Ins., insulin. (C) Densitometric analysis of the stained gels in A.

Biochemistry: Kakita et at

2806

Proc. NatL Acad. Sci. USA 79 (1982) 18 N

° 16 _ x

*20

E 14

40

0.

cr

co

12

4

_

0

E 15-

c

o 10 r._

CI

0

o

10

CL

8

o C

I-

5-

._j 0

4

I

A

B

C

A

B

C

2

r

FIG. 3. (Left) Insulin content (rat insulins I and II) in fasted (bar A), fed (bar B), and GHTB (bar C) Wistar-Furth rats as detected by the immunoelectrophoretic method (from data in Table 1). (Right) The ratio of rat insulins, I/II, in fasted (bar A), fed (bar B), and GHTB rats (bar C). II ratio of [3H]leucine incorporation was 0.9 0.1; at 16 mM glucose it increased to 9.8 3.3 (P < 0.01). Preferential synthesis of insulin I at 16 mM glucose was also observed in islets from normal rats. The data are summarized in Fig. 4. Therefore, under conditions that stimulate insulin biosynthesis, there was a marked preferential synthesis of rat insulin I relative to II. ±

±

DISCUSSION The results of this systematic evaluation ofthe proportion of the two insulin gene products under various experimental conditions confirm previous observations in control rats that the content and basal rate of synthesis of both rat insulins occurred in an approximately 60:40 ratio (2, 12-14). By contrast, in the hyperinsulinemic GHTB rats, there was a small change in the insulin II content compared to the control animal, but there was an almost 6-fold increase in the content of insulin I. Biosynthetic studies suggested that these differences in the content of rat insulins I and II under hyperinsulinemic conditions were due to a relative increase in synthesis of rat insulin I. At 5.5 mM glucose the ratio I/IH was approximately 60:40, whereas at 16 mM glucose it was nearly 90:10; The mechanism by which biosynthesis of rat insulin I is increased relative to that of insulin II is unknown. Although

5.5

5.5 16.0 Glucose, mM

16.0

FIG. 4. [3HlLeucine incorporation into immunoprecipitable proinsulin and insulin by islets from normal.(Left) and GHTB (Right) rats after 4-hr incubation in either 5.5 or 16 mM glucose. Pancreatic extracts were electrophoresed and incorporation into rat insulins I and II was determined by densitometric tracings of the autoradiographs (see Table 2).

growth hormone could have a direct effect on pancreatic islets, the diabetogenic action of growth hormone (15, 16), which includes insulin resistance, may be responsible for the changes noted in this study. It has been demonstrated (24-30) that glucose is a potent regulator of insulin biosynthesis in the rat. An effect of glucose at the translational level has been shown to occur immediately after addition of glucose to incubation medium, leading to an increase in the number of polysomes active in protein synthesis by enhancing initiation of all islet mRNA (29). Another component of glucose-stimulated insulin synthesis may involve new mRNA synthesis (24, 30). Rats fasted for 3-4 days have an 80% decrease in proinsulin mRNA levels (25). Refeeding restores proinsulin mRNA levels to values greater than in fed controls within 24 hr. Glucose given intraperitoneally to starved rats in amounts sufficient to increase plasma glucose but not provide caloric replacement also increased proinsulin mRNA levels more than 3-fold relative to that in starved animals (31). These data suggest that glucose is an important modulator of the rate of insulin biosynthesis, through changes in proinsulin mRNA levels.

Table 2. Biosynthesis of insulins I and II in pancreatic islets from normal and GHTB rats Glucose,

~Incorporation of

[3 Hlleucine*

"IRI %

of total Insulin I, Insulin II, Incorporation of [3H~leucine* Rats Total. proteint IRI Ratio, I/II protein % % Normal 5.5 1.6 ± 0.1$ 199.6 ± 80.2 10.9 61.0 ± 1.7t 39.0 ± 1.7* 1,834.3 ± 54.6 16.0 8.4 ± 1.6* 31.5 89.3 ± 1.6* 10.7 ± 1.6* (n = 3) 5,445.6 ± 187.5 1,717.1 ± 397.8 5.5 GHTB 231.8 ± 54.6 8.5 47.3 ± 1.9* 52.7 ± 1.9 0.9 ± 0.1* 2,733.0 ± 250.3 16.0 9.8 ± 3.3* 29.2 90.7 ± 3.3* 9.3 ± 3.3 (n = 3) 6,766.2 ± 519.1 1,978.0 ± 309.2 Islets were incubated in either 5.5 mM or 16.0 mM glucose for 4 hr (370C), and [3Hlleucine incorporation into insulins I and II was determined by electrophoresis on polyacrylamide slab gels at pH 8.9, autoradiography, and densitometric analysis. Results are shown as mean ± SEM. IRI,

Glucose, mM

immunoreactive insulin.

* Shown as cpm per islet. t Acid/alcohol-soluble, trichloroacetic acid-precipitable protein. t For difference between 5 mM and 16.0 mM, P < 0.01.

Biochemistry: Kakita et aL In the present studies, the GHTB rats contained approximately twice as much proinsulin mRNA as did control animals (data not shown), suggesting that the differential synthesis of rat insulin I relative to II might be due to different amounts of proinsulin I mRNA. Use of specific probes for distinguishing between the mRNAs for the two rat proinsulins would answer this question. The data presented here are also compatible with an enhanced stability or preferential translation of rat proinsulin I mRNA relative to II. In a rat insulinoma adapted to cell culture, the two mRNAs are equal, but there is increased synthesis of insulin I relative to II due to preferential translation of rat insulin I mRNA (B. Cordell and H. Goodman, personal communication). The present studies suggest that the two rat insulin genes may function independently, and they offer the potential opportunity to explore the relationship between gene structure and expression. The nucleotide sequences of the two rat insulin genes are completely known. Lomedico et aL (8) have shown that the two rat insulin genes form a heteroduplex for 500 bases 5' to the coding region of the gene, suggesting marked similarities in this potential promoter region. Although the promoter regions appear to be homologous, other studies have shown that a single base pair change can markedly alter RNA transcription rates (32). It is also interesting to speculate about the introns in the preproinsulin genes and their possible effect on gene expression (33). An intervening sequence of 119 base pairs is located in the 5' untranslated region preceding the prehormone sequence of both rat insulin genes (Fig. 1). The rat insulin II gene contains an additional 499-base-pair intervening sequence in the coding region between the codons for the sixth and seventh amino acids of the C peptide. Because the human (34) and rat II and chicken (35) preproinsulin genes all contain two introns in approximately the same locations, it has been suggested that the two rat insulin genes have evolved by a recent gene duplication followed by the loss of the second intron in the rat insulin I gene (34, 35). If the second intron were important, perhaps the expression of the rat insulin I gene might be impaired. Just the opposite seems to be the case. The fact that an additional 499 bases of RNA have to be synthesized in the precursor to rat proinsulin II mRNA and then removed may alter the rate of gene expression. The authors thank Keith O'Connell for technical assistance, Pam Helms for assistance with preparation of the manuscript, and Peter Rotwein for review and helpful discussions. This work was supported by National Institutes of Health Grants AM-16746 and AM-24950. M.A.P. is a recipient of U.S. Public Health Service Career Development Award AM-00033. 1. Smith, L. F. (1966) Am. J. Med. 40, 662-666. 2. Clark, J. L. & Steiner, D. F. (1969) Proc. NatL Acad. Sci. USA 62, 278-285. 3. Bunzli, H. F., Glatthaar, B., Kunz, P., Mulhaupt, E. & Humbel, R. E. (1972) Hoppe-Seyler's Z. PhysioL Chem. 353s, 451-458.

Proc. Natd Acad. Sci. USA 79 (1982)

2807

4. Markussen, J. (1971) lnt. J. Pept. Protein Res. 3, 149-155. 5. Yamamoto, M., Kotaki, A., Okuyama, T. & Satake, K. (1960)J. Biochem. 48, 84-92. 6. Humbel, R. E., Bosshard, H. R. & Zahn, H. (1972) in Handbook of Physiology-Endocrinology, ed. Geiger, S. R. (Waverly, Baltimore), Vol. 1, pp. 111-132. 7. Cordell, B., Bell, G., Tischer, E., DeNoto, F. M., Ullrich, A., Pictet, R., Rutter, W. J. & Goodman, H. M. (1979) Cell 18, 533-543. 8. Lomedico, P., Rosenthal, N., Efstratiadis, A., Gilbert, W., Kolodner, R. & Tizard, R. (1979) Cell 18, 545-558. 9. Ullrich, A., Shine, J., Chirgwin, J., Pictet, R., Tischer, E., Rutter, W. J. & Goodman, H. M. (1977) Science 196, 1313-1319. 10. Villa-Komaroff, L., Efstratiadis, A., Broome, S., Lomedico, P., Tizard, R., Naber, S. P., Chick, W. L. & Gilbert, W. (1978) Proc. Natl Acad. Sci. USA 75, 3727-3731. 11. Chan, S. J., Keim, P. & Steiner, D. F. (1976) Proc. Natl Acad. Sci. USA 73, 1964-1968. 12. Tanese, T., Lazarus, N. R., Devrim, S. & Recant, L. (1970) J. Clin. Invest. 49, 1394-1404. 13. Rall, L. B., Pictet, R. L. & Rutter, W. J. (1979) Endocrinology 105, 835-841. 14. Itoh, N., Nose, K. & Okamoto, H. (1979) Eur. J. Biochem. 97, 1-9. 15. Martin, J. M., Akerblom, H. K. & Garay, G. (1968) Diabetes 17, 661-667. 16. Peake, G. T., McKeel, D. W., Mariz, I. K., Jarett, L. & Daughaday, W. H. (1969) Diabetes 18, 619-624. 17. Albert, W. G. & Permutt, M. A. (1979) J. Biol Chem. 254, 2483-2492. 18. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. (1951)J. Biol Chem. 193, 265-275. 19. Davis, B. J. (1964) Ann. N.Y. Acad. Sci. 121, 404-427. 20. Renart, J., Reiser, J. & Stark, G. R. (1979) Proc. Natl Acad. Sci. USA 76, 3116-3120. 21. Towbin, H., Staehelin, T. & Gordin, J. (1979) Proc. Natl Acad. Sci. USA 76, 4350-4354. 22. Reiser, J. & Wardale, J. (1981) Eur. J. Biochem. 114, 569-575. 23. Lacy, P. E. & Kostianovsky, M. (1967) Diabetes 16, 35-39. 24. Permutt, M. A. & Kipnis, D. M. (1972) J. Biol Chem. 247, 1194-1199. 25. Giddings, S. J., Chirgwin, J. & Permutt, M. A. (1981)1. Clin. Invest. 67, 952-960. 26. Howell, S. L. & Taylor, K. W. (1967) Biochem.J. 102, 922-927. 27. Lin, B. G. & Haist, R. E. (1969) Can. J. Physiot Pharmacol 47, 791-801. 28. Morris, G. E. & Korner, A. (1970) Biochim. Biophys. Acta 208, 404 413. 29. Permutt, M. A. (1974) J. Biol, Chem. 248, 2738-2742. 30. Kaelin, D., Renold, A. E. & Sharp, G. W. G. (1978) Diabetologia 14, 320-335. 31. Giddings, S., Chirgwin, J. & Permutt, M. A. (1982) Diabetes, in press. 32. Corden, J., Wasylyk, B., Buchwalder, A., Sassone-Corsi, P., Kedinger, C. & Chambon, P. (1980) Science 209, 1406-1414. 33. Gruss, P., Efstratiadis, A., Karathanasis, S., Konig, M. & Khoury, G. (1981) Proc. Natl Acad. Sci. USA 78, 6091-6095. 34. Bell, G. I., Pictet, R. L., Rutter, W. J., Cordell, B., Tischer, E. & Goodman, H. M. (1980) Nature (London) 284, 26-32. 35. Perler, F., Efstratiadis, A., Lomedico, P., Gilbert, W., Kolodner, R. & Dodgson, J. (1980) Cell 20, 555-568.