Circulating Insulin: the Proinsulin-Like Properties

of "Big" Insulin in Patients without Islet Cell Tumors

BARRYM. SHERMAN,Pi-mauP GORDEN,JESSE RoTH, andPIEmR FREYCHET

From the Diabetes Section, Clinical Endocrinology Branch, National Instituteof Arthritis and Metabolic Diseases, National Institutes of Health,Bethesda, Maryland 20014

A B S T R A C T When plasma is filtered on SephadexG-50, insulin immunoreactivity is recovered in two peaks."Big" insulin, the higher molecular weight component,and "little" insulin, the lower molecular weight com-ponent, have elution volumes that correspond to those ofproinsulin-'I and insulin-'I respectively. When plasmawas extracted with acid ethanol and filtered in 1.0 Macetic acid, the patterns and proportions of "big" and"little" insulin were indistinguishable from those ob-tained by filtration of whole plasma in neutral buffer.When "big" insulin was isolated from plasma and mixedwith a tracer of porcine proinsulin-'I, trypsin convertedthe "big" insulin immunoreactivity to the gel filtrationpattern of "little" insulin in the same way that it con-verted the proinsulin radioactivity. More than 90% ofboth "big" insulin and--proinsulin were converted at op-timal trypsin concentrations. Our present guinea piganti-insulin serum failed to distinguish "big" from"little" but a porcine proinsulin anti-serum, under ap-propriate conditions of assay, reacted strongly with "big"insulin but not at all with "little." When tested on iso-lated fat cells, "little" insulin had the same bioactivity asporcine insulin, whereas "big" insulin had the samelow activity as porcine proinsulin. These studies sug-gest that "big" insulin represents either single-chainproinsulin and/or a proinsulin intermediate that hassimilar low bioactivity.

Portions of this work were presented at the AmericanFederation For Clinical Research Eastern Section Meeting,December 1969, and at the National Meeting, May 1970(1,2).

Dr. Freychet's permanent address is Groupe Parvis U-55(I.N.S.E.R.M.), Hotel Dieu, 1 Place du Paris Notre-Dame, Paris 4, France.

Received for publication 25 May 1970 and in revised form23 November 1970.

INTRODUCTIONInsulin is synthesized as a single-chain polypeptide, pro-insulin, most of which is converted to insulin within thefl-cell (3-5). When pancreatic extracts or commercialinsulins are filtered on Sephadex G-50, proinsulin andother insulin-like substances are recovered in a peak thatis eluted ahead of insulin. Each component of this earlierpeak contains the complete insulin molecule as part ofits structure, yet each has distinctive chemical, immuno-chemical, and biological characteristics (5).

When plasma is filtered on Sephadex G-50, a portionof the insulin immunoreactivity elutes in a region com-parable to proinsulin-"big" insulin (6, 7). The majorityof the insulin immunoreactivity is recovered in the regioncomparable to insulin-"little" insulin. Both componentscome from the pancreas and some of the physiologicalconditions that produce fluctuations in each of the com-ponents have been described (6-8).

In the present study we have characterized furtherthe "big" and "little" insulins. Their low concentrationsin plasma and the lack of proinsulin and proinsulin in-termediates from human pancreas limited our study to acomparison of "big" and "little" insulin in human plasmawith proinsulin and insulin from porcine pancreas.

METHODSHuman serum albumin (HSA) (Pentex, crystallized) wasiodinated with 'I (Union Carbide) at 2 ,uCi//Ag and freedof radioactive contaminants by batch adsorption with Dowex1 X 10. Porcine insulin and porcine proinsulin' were iodi-nated with 'I (Union Carbide) and 'I at specific activitiesof 200-300 ,uCi/,ig and freed of radioactive contaminants byadsorption to cellulose (9).

Unless otherwise noted, studies were performed at 4'Cand the diluent was 0.05 M Veronal buffer, pH 8.6, con-taining HSA (2.5 mg/ml) and rabbit plasma Cohn fraction

'Gifts of Dr. Ronald Chance and the Eli Lilly Company.

The Journal of Clinical Investigation Volume 50 1971 849

'.5-

* 1.0-

EEzz 0

Z 0.5-

SEPHADEXG500.05 M AMMONIUMCARBONATE -3000

0

-2000 c

_UsL"

an

02000

-o

FIGURE 1 Preparation of "big" and "little" insulin by gel filtration. Plasma(250 ml) was filtered on an 8 X 100 cm column of Sephadex G-50 "fine."Albumin-2'I and 'I, added to the plasma, marked the void and salt volumeswhich are denoted by the arrows.

II (Pentex, 0.1 mg/ml). For analytic discrimination of "big"and "little" insulin, 1- to 2-ml samples were applied to a1 X 50 cm column of Sephadex G-50 that had been equili-brated in diluent (6). 1 ml fractions were collected; 0.8ml aliquots were assayed for insulin. To prepare largequantities of "big" and "little" insulin, 250 ml of plasmawas filtered on an 8 X 100 cm column of Sephadex G-50that had been equilibrated in 0.05 M (NH4)2COG, pH 9.0(Fig. 1, Table I). Fractions of 30 ml each were collected;1 ml aliquots were lyophilized, reconstituted in diluent, andassayed for insulin. The fractions corresponding to "big"insulin and "little" insulin were pooled, lyophilized, andstored at - 20'C until used. The digestion with trypsinis described in the legends to the figures. For most studies

TABLE ICharacteristics of Donors of Plasma for "Big"

Insulin Isolation

Standard GTTblood glucose

IdealSubject Sex Age body wt. 0 hr 2 hr

yr % mg/100 mgF. I. E. F 20 100 84 105S. C. H. M 19 100 94 110B. E. L. F 16 180 90 158M. I. T. F 54 130 214 396C. W. N. M 43 180 105 197

Volunteers were fed a high carbohydrate diet for 3 days afterwhich a standard 100 g oral tolerance test was performed; the0 and 2 hr values for blood glucose in this test are shown. Toobtain plasma with high concentration of "big" insulin, thevolunteers were fasted for 40 hr, given 100 g of glucose orallyand 2 hr later were subjected to plasmaphoresis. In patientM. I. T. the fast preceding the plasmaphoresis was only 12 hr.

of conversion by trypsin, "big" insulin was used within afew days of its isolation.

For immunoassay, porcine insulin-WI, unlabeled porcineinsulin, and guinea pig anti-porcine insulin serum wereused unless otherwise indicated. Bound and free hormonewere separated by the addition of rabbit anti-guinea piggammaglobulin (6).

Biological activity was measured as the conversion ofglucose-1-"C to "CO2 by isolated fat cells (10, 11). Cellswere isolated from the epididymal fat of young rats by ex-posure to collagenase. Each incubation flask in a total volumeof 2 ml of Krebs-Ringer bicarbonate buffer contained fatcells (15 mg/ml), glucose (0.15 /umoles/ml), glucose-U-`C (2.2 X 10' dpm/ml, New England Nuclear), and albumin(20 mg/ml, bovine Fraction V, Armour), adjusted to pH7.4 under an atmosphere of 95% O0-5% C02. After shakingfor 2 hr at 37'C, the CO was trapped in Hyamine andcounted in a liquid scintillation spectrometer.

RESULTS

Gel filtration. When plasma was filtered on SephadexG-50 in (NH4)2COS, "big" insulin was separated almostcompletely from "little" insulin and from the plasmaproteins (Fig. 1). When the isolated "big" insulin wasfiltered on Sephadex G-50, the immunoreactivity wasrecovered in a single peak; when "big" insulin wasmixed with a tracer of porcine proinsulin-'I and fil-tered, the peaks of immunoreactivity and radioactivitywere coincident. Furthermore, when plasma was ex-tracted by a method used for the isolation of insulinand proinsulin from pancreas (12) (acid ethanol, pHadjustments, isoelectric precipitation with ether) andfiltered on Sephadex G-50, the location and relative

2Gift of Doctors A. Kagan and S. M. Glick.

850 B. M. Sherman, P. Gorden, 1. Roth, and P. Freychet

l)roportions of 'big" and 'little" were the same as thoseobtained by filtration of whole p)lasnla.

Conl Zc siwull of "big"' to 'li/Il''. \Without trypsil,inctilhation at 370 (i for' 3 lhr )ro(llced no challCge ill 'i,''iinsulin (Fig. 2 A), while trvpsin 20-200 zg/,l"il trans-formed 90%c of "big" insulin to "little" insulin (Fig.2 B and C). A 10-fold greater concentration of trypsin(2 mg/mIl) caused insulin degradation (Fig. 2 D). Incontrast another "big" insulin preparation failed totransform comlpletely even wN-ith 2 imig/ml of tryPsin(Fig. 3).

Because the concentration of trypsin required forcomplete conversion varied among the individual )repa-rations of "big" insulin and because "big" inlsulin re-quired 20-200 times higher concentrations of trypsinthan are required for conversion of pure porcine pro-insulin (5), proinsulin-'I was mixed with "big" insulin

3.Or

S.C.H. A

2.0H

No Trypsin

I.F-z0

z

U)

z

0+

c

2.0F

1.0-

U

Trypsin

200 19 /ml

°§-il-. .--v 1 ...

Fi(it-RF, 2 Conversion of "big" to 'little" instThe lyophilized "big" insulin (10-20 m~ug) froject S. C. H. was dissolved in 1 ml of 0.04pH 9.4 and mixed with TPCK (L-1-tosylaethyl chloromethyl ketone) trypsin (Worthinical Corp.) at 0-200 ,ug/ml. After 3 hr atexcess of soybean trypsin inhibitor (Sigmawas added and the entire reaction mixture1 X 50 cm column of Sephadex G-50 in dilumtions were collected and assayed for insulin.

I

2.0 k B.E.L. A

aNo Trypsin

I.0-

z0

U

C-)

LL.

Ez

z

C2.0k

.0FB

Trypsin

20,ELg/mI

Trypsin200Lg/mi

I

B

Trypsi n20,ug /mln

.

r--

D

Trypsin

2000/Lg/mi

FIGURlE 3 Conversion of "big" to "little" insulin by trypsin.Isolated "big" insulin (10-20 mug) frorm obese subjectB. E. L. was exposed to trypsin and after 3 hr filtered onSephadex G-50. See legenid to Fig. 2.

and their try-psin sensitivities compared. With each ofthree 'b)ig" insulin preparations (from two normal sub-jects and one obese diabetic patient) at each concen-tration of trypsin, the degree of conversion of "big" in-

D sulin was the same as that of proinstilin-l'I (Figs. 4and 5). The only difference in the conversion of "big"insulin immunoreactivity and proinsulin-'1 radioactivitywvas that at relatively low trypsin concentrations, a

Trypsin small peak of immunoreactivity appeared w-hich was2000,g/ml intermediate in position (Fig. 4: S.C.H. with 0.5 ,ug

2OOOFglmI trypsin, F.I.E. with 5.0 jug trvpsin) and which disap-peared at higher concentrations of enzyme. At optimaltrypsin concentrations more than 90%' of the "big" in-sulin and proinsulin-'I converted (Figs. 4 and 5).Because we suspected that some contamination withplasma proteins in the "big" insulin preparations might

ilin by trvl)sill. be the cause of the relative resistance to trvpsin, "big")m normnal sub-m Tori subuffe insulin was purified further by a second fltration on

Amido-2-phebyf- Sephadex G-50. In this case, less than 1 jug-/ml of trv-p-gton Biochem- sin conxverted more than 80%' of the "big" insulin37°C a 10-fold (Table II), which is comparable to results obtainedChemical Co.) ) with pure proinsulin (5).

flte nlred - Atanenza-me concentratioIls, recoveries of totalent. 1 ml frc- Atal nye ocradioactivity and of immunoreactivity were 80-120%;,~

"Big" Insulin: Proinsulin-Like Properties 851

reductions in "big" insulin immunoreactivity were ac-counted for fully by increases in "little" insulin. Thissuggests that "big" insulin, on a molar basis, reactedwith this antiserum about as well as "little" insulin.

Immunologic characterization. With the guinea piganti-porcine insulin serum that was used in these stud-ies, porcine proinsulin on a molar basis was 25% lessreactive and human insulin was 40% less reactive thanporcine insulin (Figs. 6 and 7). However, the curvesof porcine insulin, human insulin, and porcine proin-

sulin, tested over a 100-fold range of concentrations,had the same shape and could be superimposed (Figs.6 and 7). Presumably the quantitative immunologic dis-tinction made by this antiserum was due to a constantdifference in affinity of each insulin for the entire anti-body population; without an absolute measurement ofconcentration, based on their weights, we would havefound no distinction between porcine insulin, humaninsulin, and porcine proinsulin. "Little" insulin wastested over a 20- to 30-fold range of dilution and "big"

2

0

FLE.

0.5 54

.0O

104.5

No Trypsin

. 0-50

).5

--- --0

0z0

0

cc-

4.EzJ

(nz

2.O0

I.c

0O

0

z0

cab

r en

0

z.

E

z

z

Un

2

1500

Trypsin10 sg/ml

P' P

2

1000

500C

M.I.T.

5 ffi Trypsin_5 IOJ~agl /mI

1.5

1.0

~~~~4

-n0

K

-u0zc-f1-z

U%

z0

'I0

w

La.CP

z

1500

1000

500

o0 Immunoreactivity

FIGURE 4 Trypsinization of a mixture of "big" insulin and proinsulin-"'I. "Big" insulin (10-20 mug) was mixed with proinsulin-'I (5000 cpm, 0.075 mgg) and incubated with trypsin.After 3 hr trypsin inhibitor was added and the mixture filtered on Sephadex. (See legend,Fig. 2). Radioactivity and immunoreactivity were determined on each fraction and plottedas a function of the column volume. Recovery of immunoreactivity and radioactivity rangedfrom 80 to 120%.

852 B. M. Sherman, P. Gorden, J. Roth, and P. Freychet

S.C.H.

-a0zUn

z-

1-u

U Radioactivity

iOO

100

I

loor

80

O

>I-

z0z

hlzLd

CLJ

60

40

201

AProinsulin- 125I'1 e~~~~~~~~BigInsulin

..RI.E.-S.C.H.

\\ /~~~~~--6lo'~'

0 10 20 30 40 50 " 200TRYPSINNLg /ml

FiGuRE 5 Comparison of "big" insulin and proinsulin-''Isensitivity to trypsin. From the experiments in Fig. 4 thepercentages of unconverted "big" insulin and proinsulin-'Iare plotted as a function of the trypsin concentration foreach of the three plasmas studied.

insulin was tested over a 30- to 40-fold range; bothwere indistinguishable from porcine insulin (Figs. 8and 9). This experiment did not exclude a substantialquantitative difference between "big" and "little" in-sulin and porcine insulin that was undetected becausewe lacked a weighed quantity of purified plasma "big"or "little" insulin.

"Big" and "little" insulin were readily distinguishedby the use of labeled porcine proinsulin with a guineapig anti-porcine proinsulin serum' (Fig. 10) that hadbeen preincubated with an excess of porcine insulin toblock the insulin-specific antibody sites (13). Underthese conditions "big" insulin and proinsulin competedfairly strongly whereas "little" insulin and porcine in-sulin failed to bind. Note, however, that in this systemporcine proinsulin was bound more strongly than washuman "big" insulin. This difference presumably re-flects the 50% difference in amino acid composition

'Gift of Dr. M. RooS l-

TABLE I IEffect'.of Repeated Filtration of "Big" Insulin

on Its Sensitivity to Trypsin

Insulin recovered after trypsinand final filtration

"Big"+

Trypsin "Big" "Little" "Little" Conversion

0 3.0 0.4 3.40 90.1 0.90 2.44 3.34 730.3 0.45 3.20 3.65 880.5 0.53 2.95 3.48 85

Plasma was filtered over Sephadex G-50. The effluent frac-tions corresponding to "big" insulin were pooled, concen-trated, and refiltered on Sephadex G-50, and the "big" insulinwas again pooled and concentrated. Aliquots were exposedto trypsin for 3 hr at 370C, the reaction mixture was filteredon a 1 X 50 cm column of Sephadex G-50, and each fractionassayed for insulin. See legend to Fig. 2.

between porcine and human connecting peptides (14-16).

Biological activity. In a series of experiments, withinsulin at 25 ,eU/m;Lg, we found that our purified por-cine proinsulin had a relative biological activity of1.72 ±0.10 ,uU/mgug, corresponding on a molar basis toactivity 11% that of the insulin. When isolated plasmacomponents from an obese subject (C.W.N.) were bio-assayed, the "little" insulin was indistinguishable fromporcine insulin and the "big" insulin was indistinguish-able from porcine proinsulin (Fig. 11). In ,other ex-periments, by use of another porcine proinsulin whichbioassayed at 0.35-0.50 AU/m.g, the "little" insulin iso-lated from a normal volunteer (S.C.H.) had 19-22 thUof biological activity per mig of immunoreactivity and"big" insulin from the same sample had 0.25-0.40 uU/mtg. Note that if a correction had been applied for thelower immunoreactivity of the human materials, thespecific biological activity of "big" and "little" insulinwould have been moderately lower.

DISCUSSIONThe characterization of "big" insulin was limited by theminute amounts of materials, their impure state, andthe unavailability of purified human proinsulin. Ourstudy was therefore largely limited to a comparison of"big" and "little" insulin from human plasma to puri-fied insulin and proinsulin from porcine pancreas.

From its position on gel filtration, its sensitivity totrypsin, and its strong reactivity with insulin- and pro-insulin-specific antisera, we can conclude that the "big"

"Big" Insulin: Proinsulin-Like Properties 853

5 i.o 1.5HORMONECONCENTRATION,mj4g /ml

2:0

FIGuRE 6 Immunologic reactivity of porcine insulin and human pancre-

atic insulin. Bound/free (B/F) of porcine insulin-WI is plotted as afunction of the concentrations of porcine insulin (0 0) and ofhuman insulin (0-0). To compare the shapes of the curves, thescale of human insulin concentration was reduced (not shown) to givethe best fit on the porcine insulin curve and the points of human insulin(U) were replotted accordingly.

RK>,,/

o \\<~~PROINSUL//J,es

Z-0.

Q8- PORPKPROINSULIN--

0.2-

PORK: IN L/NRK<

1.0 2.0 5.0 10.0HORMONECONCENTRATION,meLg/mI

FIGURE 7 Immunologic reactivity of porcine insulin and

porcine proinsulin. Bound/free of porcine insulin-MI was

plotted as a function of the concentration of unlabeled por-

cine insulin (0-0) and proinsulin ( ). The pro-

insulin curve was adjusted to account for its * greater

molecular weight (- ). The scale of porcine proinsulin

was reduced to give the best fit on the porcine insulin

curve and the points replotted accordingly (-).

insulin component contains both insulin and a majorportion of the connecting segment and thus representseither single-chain proinsulin or one of the two-chainproinsulin intermediates that still retains most of theconnecting segment. Similar conclusions were reachedby Melani, Rubenstein, and Steiner (17) and by Laza-rus, Tanese, Gutman, and Recant (18), from theirstudies, though they lacked experiments to distinguishproinsulin from one or more of its intermediates. Sincewe found that the biological activity of "big" insulinwas so low, relative to its immunological reactivity, we

suggest that the "big" insulin is either one-chain pro-

insulin or a two-chain proinsulin intermediate of very

low biological activity, possibly similar to the porcineproinsulin intermediate in which the 54-55 bond in theconnecting segment is split (split proinsulin).' We at-tempted to distinguish further between the one-chainand two-chain forms by subjecting "big" insulin to re-

duction and reoxidation and to acrylamide-gel electro-phoresis (5), but we were unsuccessful in adapting

'Chance, R. E. Personal communication.

854 B. M. Sherman, P. Gorden, J. Roth, and P. Freychet

I'

z

z

OwO0.

* PORCINEINSULINVLTTLE INSULIN S.C. H.A LITTLE INSUL IN F. lIE.

PORK INSULIN, mp~g/ml

100 200 300LITTLE INSULIN SCH,kLI

100I

200LITTLE INSULIN FIE, pJ

400 500

400300

FIGURE 8 Immunologic reactivity of "little" insulin and porcine insulin.Multiple aliquots of isolated "little" insulin were assayed. The bound/free of porcine insulin-"'3I is plotted as a function of the concentrationof unlabeled porcine insulin and of the volumes of the "little" insulinconcentrates.

these techniques to nanogram quantities of material. Insome of our samples at relatively low trypsin concentra-tions, some of the "big" insulin but not the proinsulin-I was found in a region between that of proinsulin

and insulin. Proinsulin with a 54-55 split, when treatedwith trypsin, would yield such an intermediate (des-nonapeptide proinsulin),' whereas single-chain proinsu-lin would not yield a component of intermediate mo-

bility. It should be noted that porcine proinsulin andsplit proinsulin have biological activities in vivo andin vitro, immunological reactivities, and gel filtrationpatterns that are about the same. Human proinsulin,though it has a 50% difference in the composition ofits connecting segment, has a leucyl-alanyl bond in a

position similar to that in the porcine connecting seg-

ment. Finally it is still unclear whether split proinsulinoccurs in vivo or whether its appearance represents some

proteolytic event during collection and processing.

Melani, Rubenstein, and Steiner (17) found that theproducts of tryptic conversion of circulating proinsulinhad enhanced immunological reactivity, presumably be-

cause their anti-porcine insulin serum was more re-active with human insulin than with human proinsulin.We failed to find any difference in immunological re-

activity after conversion because our antiserum did notdistinguish "big" from "little." A previous antiserumthat we had used, the supply of which is now exhausted,did distinguish "big" from "little" (6) and presumablywould have detected an increase in immunoreactivityafter trypsin treatment of "big."

The "big" insulins obtained from the normal, obese,and diabetic subjects were indistinguishable on gel fil-tration, and as to sensitivity to trypsin and immunologi-cal reactivity. Thus far we have no evidence for theexistence of an abnormal form of proinsulin in thesedisorders. Our studies were limited to "big" insulinobtained 2 hr after oral glucose administration frompatients without islet cell tumors. Further studies with"big" insulin obtained under other physiological circum-stances and in other disordered states such as tumors are

necessary before we can generalize on the identity of"big" insulin.

"Big" Insulin: Proinsulin-Like Properties 855

H

-iz

C)z

.0

0

* Porcine Insulin° Big Intulin B.EL.& Big Insulin SKHI|

PORK INSULIN mrpg/ml. I .

10 20 30BIG INSULIN S.C.H. gI

L̂ I-.

10 20 30 40 50BIG INSULIN B.EL.L LI

40

* Porcine InsulinA Big Insulin S.C.HI1o Big Insulin F.LE..

PORK INSULIN nug/ml.

( 10 20 30BIG INSULIN FIE. j

I

40

lV 20BIG INSULIN S.C.H. iI

oPorcine Insulin*Big Insulin MIT

PORCINE INSULIN mpg10 20

BIG INSULIN MIT. pIl30 40

FIGURE 9 Immunologic reactivity of "big" insulin and porcine in-sulin. Bound/free of porcine insulin--I is plotted as a function ofthe concentrations of unlabeled porcine insulin and of the volume ofthe "big" insulin concentrates.

856 B. M. Sherman, P. Gorden, J. Roth, and P. Freychet

- H

z

z

0.

Lo

z

D

z

0.0IujGo1

LL

50

-iz

Uoz

0

0.o0Qlt

IA

u

ACKNOWLEDGMENTS

We are grateful to C. M. Hendricks for superior technicalassistance and to Dr. J. Gliemann for some of the studiesin isolated fat cells on plasma components from patientS.C.H.

Dr. Freychet is the recipient of a U. S. Public HealthService International Postdoctoral Research Fellowship (1F05TW01512-01Y.

REFERENCES

1. Sherman, B., P. Gorden, and J. Roth. 1969. Character-ization of "big" insulin. Clin. Res. 17: 613. (Abstr.)

2. Sherman, B., P. Gorden, J. Roth, and J. Gliemann.1970. "Big" insulin: circulating proinsulin. Clin. Res.18: 372. (Abstr.)

3. Steiner, D. F., and P. E. Oyer. 1967. The biosynthesisof insulin and a probably precursor of insulin by ahuman islet cell adenoma. Proc. Nat. Acad. Sci. U.S.A.57: 473.

4. Steiner, D. F., D. Cunningham, L. Spigelman, and B.Aten. 1967. Insulin biosynthesis: evidence for a pre-cursor. Science (Washington). 157: 697.

5. Steiner, D. F., J. T. Clark, C. E. Nolan, A. H. Ruben-stein, E. Margoliash, B. Aten, and P. E. Oyer. 1969.Proinsulin and the biosynthesis of insulin. Recent Progr.Hormone Res. 25: 207.

6. Roth, J., P. Gorden, and I. Pastan. 1968. "Big" insulin:a new component of plasma insulin detected by immuno-assay. Proc. Nat. Acad. Sci. U.S.A. 61: 138.

0.8r

,, 0.6

z

(n2 0.425cra.

z5O 0.2a.ULm

Little Insulinf -. ,]1cIS=.

Porcine IF

Big Insulin,

nsulin

Porcine Proinsulinu f ,

5 10 15 " 50

m~ug EQUIVILANTS PORCINE INSULIN100

FIGURE 10 Immunologic reactivity with anti-proinsulinserum. Guinea pig anti-proinsulin serum was preincubatedwith an excess of porcine insulin. Bound/free of porcine pro-insulin-1~"I was plotted as a function of the concentrationsof porcine insulin. The scale of porcine proinsulin wasappropriately reduced to account for its higher molecularweight.

0 I.0 22i5 - '"Little" Immunoreactivity - mlng/mlo i0 2.5 5-0 II_ _ _

2 5 Pork Insulinmfg/ml

I

OL LO

0 2 5

0 1.0 2o---a Bigl' Immunoreoctivity

m1.sg/ml

5

Pork10 25 Proinsulin

m/.g /m I

FIGURE 11 Biological activity in isolated fat cells. Plasma from an obesesubject (C. W. N.) was filtered on Sephadex and the fractions correspondingto "big" and to "little" were pooled and lyophilized. Each pool was filteredagain (1.5 X 90 cm column) and the effluent fractions immunoassayed for in-sulin (see insets at the right). The fractions denoted by the shaded area were

pooled and lyophilized. For assay they were dissolved in Krebs-Ringer phos-phate buffer, pH 7.4, which contained albumin 2.5 mg/ml to yield "big" and"little" with concentrations of immunoreactivity of 175 and 100 mog/ml, re-

spectively, assayed against the porcine insulin standard. In the figure at theleft, counts per minute of 14CO2 produced is plotted as a function of the con-

centration of hormone. Each point is the mean +SEM of three determinationsin a single assay. The proinsulin scale has been adjusted so that insulin andproinsulin correspond on a molar basis.

"Big" Insulin: Proinsulin-Like Properties 857

40

eV0

00

x

a.0

20 F

7. Gorden, P., and J. Roth. 1969. Circulating insulins. Arch.Intern. Med. 123: 237.

8. Gorden, P., and J. Roth. 1969. Plasma insulin: fluctua-tions in the "big" insulin component in man after glu-cose and other stimuli. J. Clin. Invest. 48: 2225.

9. Yalow, R. S., and S. A. Berson. 1960. Immunoassay ofendogenous plasma insulin in man. J. Clin. Invest. 39:1157.

10 Gliemann, J. 1966. Assay of insulin-like activity by theisolated fat cell method. Diabetologia. 3: 382.

11. Rodbell, M. 1964. Metabolism- of isolated fat cells. Ef-fects of hormones on glucose metabolism and lipolysis.J. Biol. Chem. 239: 375.

12. Davoren, P. R. 1962. The isolation of insulin from asingle cat pancreas. Biochim. Biophys. Acta. 63: 150.

13. Yip, C. C., /and J. Logothetopoulos. 1969. A specific

anti-proinsulin serum and the presence of proinsulinin calf serum Proc. Nat. Acad. Sci. U.S.A. 62: 415.

14. Chance, R. E., R. M. Ellis, and W. W. Bromer. 1968.Porcine proinsulin: characterization and amino acid se-quence. Science (Washington). 161: 165.

15. Oyer, P. E., S. Cho, and D. F. Steiner. 1970. Isolationand structure of human proinsulin C-peptide. Fed. Proc.29: 533.

16. Nolan, C., E. Margoliash, and D. F. Steiner. 1968.Bovine proinsulin. Fed. Proc. 28: 343. (Abstr.)

17. Melani, F., A. H. Rubenstein, and D. F. Steiner. 1970.Human serum proinsulin. J. Clin. Invest 49: 497.

18. Lazarus, N. R., T. Tanese, R. Gutman, and L. Recant.1970. Synthesis and release of proinsulin and insulin byhuman insulinoma tissue. J. Clin. Endocrinol. 30: 273.

858 B. M. Sherman, P. Gorden, J. Roth, and P. Freychet