JOURNAL TRANSCRIPT

---

Dominant Inheritance of Large Molecular Weight Immunoreactive Glucagon JERRY P. PALMER, PHILLIP L. WERNER, JAMES W. BENSON, and JOHN W. ENSINCK, Department of Medicine and the Diabetes

Research Center, University of Washington, Seattle, Washington 98144

A B S T R A C T Plasma from some individuals contains substances which are reactive with glucagon antiserum, are larger than 3,500-dalton glucagon, and have been proposed as possible precursors of glucagon. We have evaluated three generations of a kindred in which 9 of 15 members evaluated had elevated plasma levels of large molecular weight immunoreactive glucagon (L-IRG) with an average concentration of 822 pg/ml. The distribution of individuals with elevated L-IRG levels in this pedigree is consistent with autosomal dominant inheritance. Gel filtration of plasma revealed that all affected family members had excessive amounts of two L-IRG peaks, one with a molecular weight of approximately 9,000 daltons and another in the 10,000 to 20,000-dalton range. Oral glucose tolerance tests were nondiabetic and elicited a fall in L-IRG levels, whereas L-IRG concentrations rose dramatically during the infusion of arginine. These L-IRG species may be precursors of 3,500-dalton glucagon and may be elevated in this kindred because of an inherited defect in either their synthesis or degradation.

INTRODUCTION Human plasma contains more than one species of immunoreactive somatotropin, gastrin, parathyroid This work was presented in part at the 36th Annual Meeting of the American Diabetes Association and at the International Symposium on Glucagon and published as an abstract (1976. Diabetes. 25[Suppl. 1]: 326) and a preliminary report (1976. Metab. Clin. Exp. 25[Suppl. 1]: 1483) in connection with these meetings, respectively. Dr. Werner was the recipient of a fellowship award from the American Diabetes Association, Washington Affiliate. His present address is Emory University School of Medicine, Grady Hospital, Diabetes Unit, Atlanta, Ga. 30303. Dr Benson's present address is The Mason Clinic, Seattle, Wash. 98101. Received for publication 15 April 1977 and in revised form 4 November 1977.

hormone, insulin, and glucagon which have been distinguished by differences in molecular weight (1, 2). The precise biologic significance of many of these compounds is not frilly understood, whereas some, such as proinsulini, are known to be biosynthetic precursors of the smaller molecular weight, biologically more active species. Plasma levels of proinsulin increase after glucose administration and are elevated in patients with several clinical conditions, most notable being patients with insulinomas (3). Recently, Gabbay et al. have described a family with hyperproinsulinemia inherited in an autosomal dominant pattern (4). Pancreatic vein catheterization has revealed that the majority of the immunoreactive glucagon (IRG)l secreted by pancreatic alpha cells in the dog has a molecular weight of approximately 3,500 daltons (5). Species of IRG larger and smaller than 3,500-dalton glucagon have been described in human plasma (6-8). Noe and Bauer have found a polypeptide of approximately 9,000 daltons which, by pulse-chase experiment, appears to be an intermediate in glucagon biosynthesis (9). Therefore, the 9,000-dalton IRG species observed in human plasma may represent a precursor form of 3,500-dalton glucagon. We have evaluated a family with excessive plasma levels of large molecular weight IRG (L-IRG) but normal levels of 3,500-dalton IRG, in which the pattern in the pedigree is consistent with autosomal dominant inheritance. The L-IRG in affected members consists of species with mol wt 9,000 and approximately 10,000-20,000 daltons, which are similar to the molecular weight of the putative precursors of 3,500-dalton glucagon which have been observed in human and animal in vitro pancreas studies (9, 10). In addition, plasma levels of these L-IRG species IAbbreviations used in this paper: IRG, immunoreactive glucagon; L-IRG, large molecular weight immunoreactive glucagon; S-IRG, small molecular weight immunoreactive glucagon.

The Journal of Clinical Investigation Volume 61 March 1978-763-769

763

were decreased during hyperglycemia and elevated by arginine, implying that they may be secretory products of alpha cells. This kindred may represent familial

hyperproglucagonemia. METHODS 15 individuals in three generations of this kindred, ranging in age from 8 to 71 yr old, were evaluated. All were healthy, had weights ranging from 100 to 148% of ideal (11), and had no personal or family history of diabetes or endocrine neoplasia. None had known renal impairment. Subjects were informed of the nature, purpose, and potential risks before their voluntary participation. Blood was withdrawn and arginine injected by scalp vein needles in the subjects' antecubital veins. Patency of these lines was maintained by the slow infusion of normal saline. The oral glucose tolerance test (100 g), arginine infusions (500 mg/kg per 30 min), and arginine pulses (5 g/30 s) were performed after an overnight fast with a 3-day high carbohydrate diet preceding the oral glucose tolerance test. Each plasma sample was assayed for IRG (30K) level in both the native state and after treatment with acetone. The 3,500-dalton glucagon is soluble in acetone with 75% recovery (12). The 160,000-dalton species is virtually excluded by precipitation, and we have found that less than 10% of the 9,000-dalton and 10,000- to 20,000-dalton species is extracted into acetone. Therefore, the IRG level in the acetone extract of plasma represents primarily the smaller molecular weight glucagon, which we have designated S-IRG. Since IRG in native plasma measures all IRG species, the difference between IRG in unextracted plasma and S-IRG allows quantification of the larger molecular weight glucagon species. Glucose was measured by an AutoAnalyzer adapta-

tion of the ferrocyanide or neocuproine reduction method and insulin by radioimmunoassay (13). 1-ml samples of plasma were fractionated on a 1.5 x 90-cm Bio-Gel P-10 column (BioRad Laboratories, Richmond, Calif.) in 0.05 M Veronal buffer plus 0.25% bovine serum albumin at pH 8.6. The columnls were standardized by applying proinsulin, insuilini, and glucagon to the columns and measuring by radioimmunoassay the eluates for these substances. 400-,ul aliquots of the 1-ml fractions were assayed for IRG using antiserum 30K. Recoveries were calculated as the sum of the IRG measured in the eluate fractions divided by the total IRG in the respective native plasma.

RESULTS

The pedigree of this family is shown in Fig. 1. All affected members of this kindred had S-IRG levels within our normal range, which was defined as the mean fasting S-IRG +2 SD in 23 nondiabetic control subjects (21-101 pg/ml). The nmean f:asting level of LIRG plus 2 SD (62 + 70 = 132 pg/ml) of the 23 nondiabetic control volunteers was taken as the upper limit of normal; therefore family members with L-IRG values exceeding 132 pg/ml were considered affected. Plasma for L-IRG and S-IRG of affected individuals was obtained after an overnight fast except for J. S., whose specimen was drawn at random. Of the 15 individuals evaluated, 9 were affected with L-IRG values ranging from 246 to 3,618 pg/ml (mean±SD = 822+1,059). Five blood relatives of the propositus had normal L-IRG levels. 3 of the 15 family members

W. & ITUS

FIGURE 1 Inheritance pattern of excessive large molecular weight IRG in three generations of a family. Affected individuals are shaded, nonaffected are clear, and a question mark (?) represents individuals not tested. Numbers represent L-IRG/S-IRG. Letters below each box represent the individual's initials.

764

J. P. Palmer, P. L.

Wernter, J. W. Benson, and J. W. Ensinick

were relatives by marriage, had normal L-IRG levels, aindl thereby served as internal controls. Individuals in all three generationis evaluated and both men and women were affected. Fig. 2 illustrates the gel-filtration profile of plasma IRG from J. A. Q. and depicts the pattern which charlacterizes this familv. All affected members had excessive levels of L-IRG distributed in two peaks: one with a molecular weight similar to proinsulin, 9,000 daltons (peak A), and another slightly larger weight species eluiting earlier from the Bio-Gel P-10 columni (peak B). We have estimated the molecular weight of peak B to be approximately 10,000-20,000 daltons because of its position between the void volume of the column (20,000 daltons) and the proinstulin marker. Some but not all affected subjects had a peak which was excluded from the gel (peak C). By gel filtratioin on Sephadex G-100 and G-200 columns (Pharmacia Fine Chemicals, Inc., Piscataway, N. J.), we and others have found this peak of IRG to have a molectular weight of about 160,000 daltons (6, 7). There was considerable variability between subjects, with somiie individuals having no detectable void volume peak (W. H. S., W. L. S., H. P. S.), while in other subjects it comprised 48% and 62% of the total IRG (R. H. S. and W. S., respectively). Similarly, there was a wide range for peak A (11-72%) and for peak B (18-76%). Because of the limits of assay sensitivity we did not always detect a distinct peak in the area corresponding to 3,500-dalton IRG. Six affected family members (W. H. S., W. L S., H. P. S., J. A. Q., W. S., R. H. S.) had oral glucose tolerance tests which were all nondiabetic (Fig. 3). NMean S-IRG concentrations fell from a fasting level of 80 pg/ml to a nadir at 3 h of 59 pg/ml. Mean L-IRG levels decreased 200 pg/ml during the oral glucose tolerance test, and when each individual's values were compared with his (her) fasting level, the decrements in levels were significant (P < 0.05) at each hour after adimiinistration of the glucose. Basal IRI and the IRI Vo

Proinsulin

Insulin

Glucagon

125I.J

I00 85

i

i

i

I

90

~,80-

t

70-

b

60-

%

5 40

:j-s;~~~. a

r

*

% -100-

% -200'1

14-300100

t 75 : 50k3. 25O0

0

i

2

3

HOURS

FIGURE 3 Oral glucose toleranice tests (100 g) in six affectedl familv members. V'alues are mean±SENI except for L-IRG graph, which is depicted as change from basal. An asterisk (*) denotes values significanitlx- (P < 0.05) differenit from fasting v-aluie by paired comparison analysis.

30r

15

20-]

|10-,'"'', 20

40

, 60 FRACTION NUMBER

80

FIGURE IRG concentrations in eluates from gel filtration oni Bio-Gel P-1() of J. A. Q.'s plasma (recovery = 100%). Peaks A and B (solid line) were observed in all affected family members, whereas peak C, (dashed line) was observed in

on)lx' sonie memlbers.

responses to the 100 g of glucose were normiial when comiipared with values obtained by others (14). The elution profiles of L-IRG from WA. L. S. in the fasting state and 3-h after oral glucose are showvn in Fig. 4. Both peaks A anid B were depressed after the administrationi of glucose. The S-IRG and L-IRG responses to the arginine infusion in four affected famil- members (H. P. S., J. A. Q., XVT. H. S., R. H. S.) and in seven noniFain ilial "Big" Glucagoni

765

Proinsulin Insulin

Vo

Proinsulin Insulin

Glucagon

Glucagon

150 g

tR251

I

O0_

36

96

76 56 FRACTION NUMBER

FIGURE 4 IRG concentrations in eluates (Bio-Gel P-10) from plasmnca of W. L. S. fasting (solid line) aind 3 h after 100 g of oral glucose load (dashed line). Fasting recovery = 65% 3-h recovery = 70%.

diabetic controls are shown in Fig. 5. L-IRG in the conitrol volunteers rose only slightly during the arginiine infusion, achievinig significanit (P < 0.05) elevationi al)ove basal at intermittenit time points (see 20004e, ""

1600-

I

!1200-

800400

200-

0-i

JL

1-1/

L

I

+---, I

I

I-000 t0

3000t 200-

FRACTION NUMBER

FIGURE 6 IRG concentrations in eluates (Bio-Gel P-10) from plasmna of H. P. S. fasting (solid line) and 90 min after arginine stimulation (dashed line). Fasting recovery = 45%; postarginine recovery = 51%.

Fig. 5). In contrast, the family members demonstrated marked inereases of L-IRG concentrations with mean levels inereasing from 506 pg/ml to a maximum of 1,727 pg/ml and then slowly declining. S-IRG rose briskly in 1oth controls and family members and fell promptly after infusion to near basal values by 90 min. Mean S-IRG levels during the arginine infusion were slightly higher in the family members than in the control subjects but achieved statistical significance only intermittently. The gel filtration profile of plasma IRG from H. P. S. in the fasting state and at 90 min of the arginine infusion is shown in Fig. 6. Both peaks A and B were increased, and a small peak corresponding to the 3,500-dalton glucagon marker was observed after arginiine. The bolus administration of arginine in W. H. S., H. P. S., and D. Q. resulted in acuite elevation of both S-IRG and L-IRG levels. At 2 min after arginine administration, mean S-IRG and L-IRG concentrations had risen 98 and 226 pg/ml, respectively. Serumil creatinine levels were measured in J. A. Q., D. Q., WV. H. S., R. H. S., and W. S., and all were normlal. DISCUSSION

O300-

* * -0-t-;;-; 90 60 30 Milutes FIGURE 5 L-IRG and S-IR G levels in seven controls (solid line) anid four afffected family mnenmbers (dashed line) during anl arginine infusion. The asterisk (*) in the L-IRG graph designates values signifi cany different (P < 0.0)5) from flastinlg values in the control .subljects. The astrisk (*) in the S-IRG graph inidicates tinmes vvbhen S-IRG in asmilyk members; was significantlN differenit ( P