Local Synthesis and Developmental Regulation of Avian Vitreal Insulin-Like Growth Factor-Binding Proteins: A Model for Independent Regulation in Extravascular and Vascular Compartments TIMOTHY ROBERT

J. SCHOEN, J. WALDBILLIG

DAVID

C. BEEBE,

DAVID

R. CLEMMONS,

GERALD

J. CHADER,

AND

Laboratory of Retinal Cell and Molecular Biology, National Eye Institute, National Institutes of Health (T.J.S., G.J.C., R.J. W.), Bethesda, Maryland 20892; the Department of Anatomy and Cell Biology, Uniformed Services University of the Health Sciences (T.J.S., D.C.B.), Bethesda, Maryland 20814; and the Division of Endocrinology, University of North Carolina School of Medicine (D.R.C.), Chapel Hill, North Carolina 27599 ABSTRACT The expression and regulation of insulin-like growth factor-binding proteins (IGFBPs) in developing avian vitreous humor and serum were compared. Vitreal IGF-I-binding activity was highest on embryonic day 6 [E-6; bound/free ratio (B/F), 0.22 f 0.019/50 J), decreased lofold between E-6 and E-19, and then remained stable through the remainder of embryonic development. In contrast, serum IGF-I binding increased 2-fold over this period, from a B/F of 0.380 f 0.056 (E-6) to a B/F of 0.89 f 0.18 (E-19). After hatching, serum IGF-I-binding activity continued to increase through posthatching week 12, while vitreal IGF-I binding increased only slightly and then remained constant. Although IGF-II binding in the vitreous humor and serum is 2to 3-fold higher than that of IGF-I, the same pattern of developmental regulation was observed as with IGF-I. Western ligand blots revealed a vitreal24-kilodalton (kDa) IGFBP that was absent from both embry-

onic and adult sera. Likewise, posthatching serum was found to contain a 70-kDa IGFBP absent in vitreous humor. Deglycosylation of vitreal and serum IGFBPs followed by Western ligand blotting revealed unique glycosylation patterns for vitreal and serum IGFBPs. One of the IGFBPs that is differentially glycosylated in vitreous and serum is a 33-kDa IGFBP that is precipitated with human IGFBP-2 antiserum. Northern blot analysis revealed the presence of IGFBP-2 mRNA in several embryonic ocular tissues as well as liver. The observations that vitreal and serum IGFBP levels are independently regulated during development and that IGFBPs from these two compartments have different molecular weights and glycosylation patterns suggest that the vitreal IGFBPs are not derived from serum. The presence of IGFBP-2 mRNA in ocular tissue surrounding the vitreal chamber supports the view that certain vitreal IGFBPs may be synthesized locally. (Endocrinology 131: 2846-2854, 1992)

E

proposed (19). The IGFs are normally found complexed to specific, high affinity binding proteins (IGFBPs), six of which have been cloned and sequenced(20). IGFBPs are present during embryogenesis and exhibit both tissue- and stage-specific regulation (21-23). Since they are able to modulate IGF activity by either facilitating (24-25) or inhibiting (26-28) ligandreceptor interactions, IGFBPs may play a role in the regulation of IGF-mediated developmental processes. The vitreous humor, a specialized ocular extracellular fluid protected from serum by the blood-ocular barrier (29, 30), has been shown to contain several IGFBPs (31-33). While the origin of vitreal IGFBPs remainsto be determined, mRNA for one of the major vitreal IGFBPs, IGFBP-2, has been identified in the retina (34) and ciliary body (35), suggesting a local ocular synthesis. To determine whether vitreal IGFBPs form part of an autonomous ocular IGFBP system, vitreal and serum IGFBPs were examined for similaritiesand differences in their pattern of developmental regulation, structure, and glycosylation. To assessthe possibility that vitreal IGFBPs are synthesized locally in ocular tissue surrounding the vitreal chamber, embryonic ocular tissueswere examined for the presenceof IGFBP-2 mRNA.

involves a seriesof precisely coordinated, tightly regulated events that, upon completion, give rise to the mature organism.Among the factors proposed to be involved in regulating growth and differentiation are insulin-like growth factor-I and -11 (IGF-I and IGF-II) (for reviews, see Refs. 1 and 2). Evidence supporting a role for IGF-I in early development comesfrom studiesshowing the presence of IGF-I (3-5), IGF-I mRNA (6, 7), and IGF-I receptors (8, 9) during embryogenesis. IGF-I has also been shown to be an important component for the in vitro differentiation of embryonic lens (lo), neuronal (1 l), and muscle (12) tissue. IGF-II is also thought to play a role in fetal and postnatal development and exhibits a number of in vitro growthpromoting and metabolic effects that are mediated through both IGF-I and IGF-II receptors (for reviews, see Refs. 1315). As with IGF-I, IGF-II is differentially expressed across development, suggestinga stage- and tissue-specificrole for this factor during embryonic development (16-18). A local role for IGF-II in the regulation of ocular size has also been MBRYOGENESIS

Received June 24, 1992. Address requests for reprints to: Robert J. Waldbillig, Building 6, National Institutes of Health, Bethesda, Maryland

Room 304, 20892.

2846

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 18:43 For personal use only. No other uses without permission. . All rights reserved.

DEVELOPMENTAL

REGULATION

Materials and Methods White Leghorn chicken eggs were purchased from Truslow Farms (Chestertown, MD). Monoiodinated (HPLC-purified) recombinant human IGF-I (2000 Ci/mmol), IGF-II (2000 Ci/mmol), and prestained mol wt markers were obtained from Amersham (Arlington Heights, IL). Unlabeled (HPLC-purified) recombinant human IGF-I and IGF-II were purchased from Amgen Biologicals (Thousand Oaks, CA). BSA (fraction V; insulin-free) was purchased from Armour Pharmaceutical Co. (Kankakee, IL). N-Glycanase [peptide-N”-(N-acetyl-glucosaminyl)], asparagine amidase, and recombinant 0-Glycanase (endo a-N-acetylgalactosaminidase) were purchased from Genzyme (Boston, MA). Affinitypurified type X neuraminidase was purchased from Sigma Chemical Co. (St. Louis, MO). Nitrocellulose paper (0.2 pm) was purchased from Schleicher and Schuell (Keene, NH).

Vitreous

humor and serum preparation

Embryonic chickens were decapitated, and the heads were rinsed with PBS to remove any contaminating blood. The eyes were enucleated, and the anterior portion (cornea, lens, and iris) of the eye was removed. The vitreous body was removed from the posterior eye cup using fine forceps, stored on ice during collection, and pooled. Pooled vitreous was centrifuged (4 C; 12,000 X g; 15 min), and the supernatant (vitreous humor) was stored at -20 C. Because the vitreous body of older birds (e.g. 3-12 weeks old) contained a distinctive gel phase component, the vitreous bodies were sonicated (Ultrasonics model W225 sonicator, Farmington, NY) for 15 set before centrifugation. Serum was collected from embryos by inserting a pulled glass capillary pipette into extraembryonic blood vessels or directly into the atria of the heart. Adult animals were killed by decapitation and exsanguinated. The blood was allowed to clot at room temperature for 1 h before centrifuging at 12,000 x g for 15 min, after which the serum was collected and stored at -20 C. The number of animals used for pooling vitreous humor and serum varied with developmental stage. Material from approximately 60 chick embryos [embryonic days 6-19 (E-6 through E-19)] was pooled. For the second and seventh day posthatching time points, samples from approximately 12 chickens were pooled. Finally, for animals that were 212 weeks old, 3 chickens were pooled/stage. At each developmental stage a minimum of at least 2 different pools of samples were used in each of the binding assays and cross-linking and Western ligand blot experiments.

IGF binding

assay

Fifty microliters of sample were incubated with 100 ~1 [‘251]IGF-II or [“‘IIIGF-I and 50 ~1 unlabeled IGF-I, or with unlabeled IGF-II or calcium-free Krebs-Ringer phosphate (KRP) buffer containing 0.1% BSA (pH 8.0). Stock radiolabeled IG’F (IGF-I and -11) was in KRP-containing 3.0% BSA and 3.0 ma/ml bacitracin. The final concentrations of radiolabeled and unlabeled IGF were 0.037 and 171 nM, respectively. The duration, temperature, pH, and protein concentration used during the binding assay (pH 8.0; 23 C; 3 h) were based on previous optimization work using bovine and embryonic chick (day 15) vitreous humor. The binding reaction was terminated by adding 100 ~1 ice-cold bovine yglobulin (3.0 mg/ml), then 300 ~1 cold polyethylene glycol (25%; Polyethylene Glycol-8000). After centrifugation (2000 X g; 4 C; 15 min) and aspiration, the surface of the pellet was washed with 300 ~1 12.5% cold polyethylene glycol. This assay termination procedure precipitates apuroximatelv 10% of the free IGF-I and aooroximatelv 15% of the free iGF-II. These values were excluded from’ihe analysis. The bound IGF was counted in a LKB -r-counter (Rockville, MD). Nonspecific binding was defined as the binding persisting in the presence of a 4600.fold excess of unlabeled homologous ligand. Nonspecific binding was subtracted from total binding to yield specific binding. The ratio of specifically bound IGF to free IGF (B/F) was expressed per 50.~1 volume of vitreous humor or serum. To assure binding assay linearity, all samples were diluted such that the total binding was 10% or less. Vitreal and serum IGF-I- and IGF-II-binding levels were assayed on E-6, E-9, E-12, E-15, and E-19. In addition, binding assays were con-

OF AVIAN

VITREAL

IGFBPs

ducted on samples obtained 2 and 7 days and 2, 3, 6, and 12 weeks after hatching. At each developmental stage, the levels of IGF-I and IGF-II binding were assayed in at least two independently prepared sample pools. The level of binding in each pool was assayed two to four times in duplicate.

Affinity cross-linking

with r”“I]IGF-I

and -II

[?]IGF-I or -11 was cross-linked to vitreal and serum IGFBPs by incubating samples (vitreous humor undiluted and serum diluted 18) with [‘Z51]IGF (final [‘?]IGF concentration, 0.23 nM) for 3 h at 23 C. To reduce interference in the radiographs due to the nonspecific binding of [‘251]IGF to BSA, radiolabeled IGF was prepared in KRP (pH 8.0) with either no BSA or 0.1% BSA. Cross-linking was achieved by the addition of disuccinimidyl suberate (DSS; final concentration, 0.36 mg/ml in dimethylsulfoxide) to the incubation mixture for 30 min at 4 C. The cross-linking reaction was terminated by adding a stopping solution containing Tris and EDTA (final concentrations: Tris, 20.0 mM; EDTA, 2.0 mM; pH 6.8). The samples were diluted with reducing sample buffer containing 2.0% sodium dodecyl sulfate (SDS), 40.0 rnM HEPES (pH 7.0), 10.0% glycerol, 5.0% P-mercaptoethanol, and 0.1 mg/ml bromphenol blue and heated to 100 C for 10 min. Fifty microliters of the diluted sample were subjected to 12.0% SDS-polyacrylamide gel electrophoresis (SDS-PAGE) using Tris-glycine running buffer. Mol wt were determined using p&stained mol wt markers (Amersham) in reducing SDS-PAGE samole buffer. Slab eels (thickness, 1.5 mm) were run at 7.5 mAmp (constani current) for 12:18 ‘h and then at 35.6 mamp for 2-3 h. The cross-linked [‘2sI]IGF-IGFBP complexes were visualized using Kodak X-AR5 film (Eastman Kodak, Rochester, NY) and a DuPontCronex intensifying screen (Wilmington, DE) at -70 C. Cross-linking studies were conducted on samples from each of the developmental stages that had been examined with the binding assay. At each developmental stage, cross-linking experiments were conducted with samples from at least two independently prepared pools of vitreous humor and serum. Cross-linked samples were run on duplicate gels, with consecutive developmental stages on juxtaposed lanes. When the number of developmental stages to be run exceeded the capacity of the gel, the remaining stages were run on a second gel. The intergel variation was monitored by including many of the same samples on both gels.

Western ligand blots Vitreal and serum samples were subjected to Western ligand blotting following a modification of a previously published technique (36). Briefly, samples were subjected to nonreducing SDS-gel electrophoresis, transferred to nitrocellulose, and probed with either [‘251]IGF-I or [‘?I IGF-II. Because of its high protein content, the serum was diluted I:8 before the addition of 3-fold concentrated SDS-PAGE sample buffer. Protein was transferred from the gel to 0.2-pm nitrocellulose using either a Bio-Rad Trans-Blot (semidry; Richmond, CA) transfer system (30 min; 60 V), or a Novex (immersion) transfer system (120 min; 25 V). The semidry buffer consisted of Tris-HCI (25 mM) and glycine (192 mM) in 20% methanol. The immersion system transfer buffer contained 10 mM sodium borate and 40 mM boric acid in 20% methanol. The two transfer systems gave equivalent results. After transfer, the blot was incubated for 30 min at 4 C in a Tris-buffered saline (TBS) solution [Tris (10.0 mM), NaCl (150.0 mM), and sodium azide (0.5 mg/ml), pH 7.41 with 3.0% (vol/vol) Nonidet P-40. Subsequently, the blot was blocked by incubation (2 h; 4 C) in TBS with 5.0% BSA. Binding of [‘251]IGF-I or -11 (final concentration, 17.0 PM) to the blot was conducted for 16-20 h at 4 C on a rocking platform in TBS with 1.0% BSA and 0.1% Tween-20. To terminate binding, the incubation solution was removed, and the blot was washed twice for 20 min on a rocker platform at 4 C with TBS buffer containing 0.1% Tween-20, followed by two washes for 20 min in TBS buffer without Tween-20. The Western ligand blot experiments were repeated at least three times using at least two independently prepared pools of vitreous and serum samples. Samples were run on duplicate gels, with consecutive developmental stages on juxtaposed lanes, as described above. The affinities of individual vitreal and serum IGFBPs for IGF-I and II were assayed by incubating individual lanes of a transfer blot in [““I]

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 18:43 For personal use only. No other uses without permission. . All rights reserved.

DEVELOPMENTAL IGF-II

REGULATION

with

various concentrations of unlabeled IGF-I and -II (IGF-II and heterologous competition-inhibition studies). To determine whether changes in IGF-binding activity across development represent changes in binding affinity or binding capacity, the competitioninhibition studies were conducted at two developmental stages markedly differing in binding activity. In the case of the vitreous, these two stages examined were E-6 and E-19. For the serum, samples from l- and 3week posthatching chicks were examined. IGFBPs were visualized by autoradiography, as described above.

OF AVIAN Northern

VITREAL

IGFBPs

Endo. 1992 Vol 131. No 6

blotting

homologous

Zmmunoprecipitation IGFBP-2 antiserum was collected from New Zealand rabbits that were previously immunized with 200 pg of a 20-amino acid synthetic multiple antigenic peptide (Applied Biosystems, Foster City, CA). The peptide sequence was derived from a region near the carboxyl-terminal sequence of human IGFBP-2 (single letter amino acid code: PECHLFYNEQQEARGVHTGR). This antiserum, designated h-C20-IGFBP-2, precipitates a 30- to 34.kilodalton (kDa) IGFBP present in extracellular fluids from a number of different species, including human, monkey bovine, rabbit, and chicken. Extracellular fluids from which a specific 30- to 34-kDa IGFBP has been precipitated include serum, vitreous humor, aqueous humor, cerebrospinal fluid, and amniotic fluid, The antiserum has also been shown to precipitate a 30. to 34.kDa IGFBP from conditioned medium from retinal pigment epithelial cells and retinoblastoma cells. The antiserum is specific, as larger (e.g. 46 kDa) and smaller (e.g. 28 kDa) IGFBPs are not immunoprecipitated. In this study immunoprecipitation experiments were conducted by incubating 450 ~1 chick vitreous or serum dil(l:8) samples [2 days posthatching (PH-DZ)] overnight at 25 C with 50 ~1 of either h-C20IGFBP-2 (final dilution, 1:5) or preimmune serum (final dilution, 1:5). After this incubation, a protein-A-Sepharose suspension (Pharmacia, Piscataway, NJ) was added, and the mixture was incubated for an additional 2 h. The protein-A-antibody-antigen mixture was transferred to an ice bath for 60 min and then centrifuged for 5 min at 12,000 x g. The supernatant was aspirated, and the pellet was washed three times with TBS. The immmunoprecipitated IGFBPs were disassociated from the protein-A-antibody complex by boiling the pellet for 5 min in nonreducing sample buffer. After centrifugation at 12,000 X g for 5 min, the supernatant was electrophoresed, transferred to nitrocellulose, and probed with [‘ZSI]IGF-II, as previously described.

Deglycosylation Samples of chick vitreous and serum [3 weeks posthatching (PHW3)] were treated with various deglycosylating enzymes, either singly or in combination. The enzymes employed were 1) N-Glycanase alone, 2) neuraminidase alone, and 3) 0-Glycanase with neuraminidase. The purpose for including a condition with both neuraminidase and OGlycanase is that the activity of 0-Glycanase has been shown to be inhibited by the presence of sialic acid side-chains on the core oligosaccharide (Genzyme). Experimentally, equal amounts (-20 pg) of vitreal and serum protein (vitreous humor, 20 ~1; serum, 2.5 ~1) were incubated overnight (37 C) with 4.4 mU 0-Glycanase, 2.2 U N-Glycanase, or 370 mU neuraminidase. Control experiments omitted the enzymes. After the incubation, the reaction was stopped by adding nonreducing sample buffer and boiling for 10 min. This material was then subjected to Western ligand blotting, as described above. The presence of glycosylated side-chains on IGFBPs was detected by an enzyme-induced reduction

in mol

wt.

RNA isolation Total RNA was isolated from E-19 ocular tissue and liver, using a guanididium hydrochloride extraction technique (RNAzol, Cinna/Biotecx Laboratories, Friendswood, TX). I’oly(A)’ RNA was isolated using oligo-DT (Fast Track, In Vitrogen, San Diego, CA). RNA quantitation was determined spectrophotometrically at 260 nM (Beckman model DU30, Palo Alto, CA).

One microgram of poly(A)+ RNA from chicken cornea, ciliary body, retina, and liver was electrophoresed on a 1% formaldehyde agarose gel and vacuum transferred to a nylon membrane (Genescreen, DuPont, Boston, MA). Northern blots were probed using a 212-basepair ?‘labeled IGFBP-2-specific polymerase chain reaction (PCR) product obtained from E-19 cornea cDNA. The PCR product was sequenced and found to be 90% identical to rat IGFBP-2 (data not shown). The PCR labeling procedure has been previously described (37). The specific activity of the probe was approximately 5 x lo9 cpm/Hg, and the probe concentration during the incubation was 1 X 10h cpm/ml. Blots were hybridized at 42 C in a 50% formamide solution containing 1 x Denhardt’s solution, 5 X SSC, 1.0% SDS, and 10% dextran sulfate for 24 h. After hybridization, blots were washed twice at room temperature for 5 min in 1 X SSC-1.0% SDS and then twice at 55 C in 0.1 x SSC-0.01% SDS before autoradiographic exposure. To determine the relative amount of poly(A)’ loaded in each lane, IGFBP-2 Northern blots were stripped and probed with a 32P-labeled chicken p-actin oligomer (5’AGCTTCTCCTTGATGTCACGCACAATTTC3’). The oligonucleotide was prepared on an Applied Biosystems model 392 DNA synthesizer (Foster City, CA), deprotected by incubation in NH,OH at 55 C for 18 h, dried in a Speed-Vat (Savant, Hicksville, NY), and resuspended in distilled HZO, and the 260/280 OD was determined. The actin oligomer was end-labeled with [y-?‘]deoxy-CTP, using standard technique (38).

Results r’“Z]ZGF-Z

and -ZZ binding

assay

The upper panel of Fig. 1 shows the level (mean + SEM) of vitreal and serum[‘251]IGF-I-binding activity at various stages of development (note the interrupted ordinate in Fig. 1). Vitreal IGF-I binding was highest on E-6 and decreasedlofold between E-6 and E-19 (mean f SEM B/F: E-6, 0.22 + 0.019; E-19,0.022 + 0.003). In contrast, serum IGF-I binding increased more than 2-fold during the sameperiod, from a B/F of 0.380 + 0.056 on E-6 to a B/F of 0.89 + 0.18 on E19. On posthatching day 2 (PH-D2), there was a 2-fold increase in IGF-I binding in both the vitreous humor and serum. In the interval between PH-D2 and posthatching week 12 (PH-W12), IGF-I binding in the vitreous humor remained relatively stable, while serum IGF-I binding again increased approximately 2-fold (PH-D2 serum B/F, 2.4 + 0.35; PH-W12 serum B/F, 4.1 f 0.14). The lower panel of Fig. 1 shows that IGF-II binding in the vitreous and serum had a developmental pattern very similar to that observed for IGF-I binding. However, at all developmental stages, IGF-II binding was 2- to 3-fold higher than IGF-I binding. Affinity

cross-linking

To examine the structural basisof IGF binding, vitreal and serum IGFBPs from various developmental stages were cross-linked to [‘251]IGF-I and [‘251]IGF-II. Figure 2 shows the cross-linked vitreal and serum [1251]IGF-IGFBPsobserved in PH-D2 chicks under reducing SDS-PAGE conditions. The two lefthand lanes show that IGF-I affinity-cross-linked vitreal and serumIGFBPS appear assingleradiographic bands. It can be seen, however, that the vitreal IGFBPs have a slightly higher mol wt and a broader radiographic profile than their serum counterparts (vitreous, 40-47; serum, 4244 kDa). The two righthand lanes of Fig. 2 show that when

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 18:43 For personal use only. No other uses without permission. . All rights reserved.

DEVELOPMENTAL

IGF-I

BINDING CHICKEN

ACTIVITY VITREOUS

REGULATION

IN DEVELOPING AND SERUM

OF AVIAN

VITREAL

IGFBPs

2849

IGF AFFINITY CROSSLINKING TO VITREAL AND SERUM IGF-BPS

kDa 92-

? 2 E

5

= z

s

=

66-

4s

IGF-II BINDING ACTIVITY CHICKEN VITREOUS

12 -

IN DEVELOPING AND SERUM

10 9-

m

8-

4 5 z

7 6-

= 4 m 0 E 2

1.35 2.4 .30 t 20

t-l

-

-31

21-

-

-21

FIG. 2. IGF-I (left two lanes) and IGF-II (right two lanes) affinity cross-linking to IGFBPs in vitreous humor and serum of PH-D2 chicks. Fifty microliters of vitreous humor or 6 pl serum were cross-linked to either [iz51]IGF-I or [‘*“I]IGF-II using DSS. Samples were subjected to electrophoresis under reducing conditions on a 12.5% polyacrylamide gel and visualized by autoradiography.

11 -

iz

31-

VITREOUS

geneity observed in IGF-I and -11affinity cross-linked vitreal and serum IGFBPs shown in Fig. 2 (PH-D2 chicks) are also observed at the other developmental stages(data not shown). The developmental trends in the levels of vitreal and serum IGF-I and -II-binding activity (observed in binding assays) were also observed in affinity-cross-linked samples(data not shown).

/ 175

rz5111GF-II

05 t

l-l 6rLrE9 EMBRYONIC IDAYSI

1 AGE

7 1 23456

I

I

I

POST-HATCH IWEEKSI

I 7

I 8 AGE

I 9

I I I 10 11 12

FIG. 1. [‘2SI]IGF-I (upperpanel) and [‘251]IGF-II (lowerpanel) specific binding (B/F) in developing chicken vitreous humor (0) and serum (0). Binding is expressed per 50 ~1 fluid. Each point represents the mean f SEM for a minimum of two assays, conducted in duplicate. The number of chickens used at each time point varied with age, as described in Materials and Methods.

IGF-II, instead of IGF-I, is used for affinity cross-linking, there are again differences in vitreal and serum IGFBPs. One of these differences is that IGF-II affinity-cross-linked serum IGFBPs appear as two distinct bands (39 and 43 kDa), while the cross-linked vitreal IGFBPs appear asa single broad band (43-47 kDa). A second difference is that vitreal IGFBPs appear slightly larger than serum IGFBPs. The microhetero-

Western &and

blots

Vitreous humor and serum sampleswere electrophoresed under nonreducing conditions, blotted to nitrocellulose, and probed with [‘251]IGF-II. The developmental pattern of vitreal and serum IGFBPs did not vary when either [‘251]IGF-I or [‘251]IGF-II was used. However, since the [‘251]IGF-II signal to noise ratio was superior to that of IGF-I (data not shown), subsequentblots were probed with [‘251]IGF-II. In general, Western ligand blots of vitreous and serum showed the same developmental trends in IGF-binding activity that were observed in the binding assay and crosslinking studies (data not shown). However, ligand blots revealed a complex population of vitreal and serum IGFBPs not visualized with affinity cross-linking techniques. Western ligand blots showed a compartment-specific developmental pattern, with unique IGFBPs appearing and disappearing at specific developmental stages.Figure 3 shows a representative [‘251]IGF-II Western ligand blot of vitreal and serum

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 18:43 For personal use only. No other uses without permission. . All rights reserved.

DEVELOPMENTAL

2850

REGULATION

1251GF-II LIGAND BLOT EMBRYONIC CHICKEN 6 DAY

OF

AVIAN

VITREAL

1 WEEK

15 DAY

-

-

1992 No 6

1251GF-lILIGAND BLOT POST-HATCHING CHICKEN 3 WEEK

2

46-

Endo. Voll31.

IGFBPs

46

z

97-

-

-97

69-

-

-69

46-

-

-46 -30

21-

r

21-

-

- 21

FIG. 3. IGF-II Western ligand blots of vitreous humor and serum from chicks at two embryonic stages (E-6 and E-15) of development. Fifty microliters of vitreous humor or 6 ~1 serum were subjected to electrophoresis on a 12.5% gel, transferred to a nitrocellulose membrane, probed with [‘251]IGF-II, and visualized by autoradiography.

at two stages of embryonic development. On E-6 (left two lanes of Fig. 3), the vitreous exhibited a 28-kDa band not detected in the serum even with extended radiographic exposures. There were also differences in vitreal and serum IGFBPs on E-15 (right two lanes of Fig. 3). For example, E15 vitreous contained a small IGFBP (24 kDa) not detected in the serum even with extended radiographic exposures. Likewise, E-15 serum contained 42- and 28-kDa IGFBPs not present in the vitreous humor. Western ligand blots of vitreous and serum at earlier and later embryonic stages also exhibited differences in IGFBPs (data not shown). Only the 33-kDa IGFBP appeared in vitreous and serum at all embryonic stages. Western ligand blots of PH-Wl and PH-W3 vitreous and serum also showed differences in the pattern of IGFBPs (Fig. 4). PH-1W serum contained a 70-kDa band not observed in the vitreous humor. In addition, the intensity of the serum 70- and 42-kDa IGFBP markedly increased between PH-1W and PH-3W. In contrast, at PH-W3, there was only a trace appearance of the 42-kDa band in the vitreous humor. Another difference in vitreal and serum IGFBPs was that the vitreal 28-kDa IGFBP increased significantly between PH1W and PH-3W, while the serum analog only slightly increased in this period. Similar differences were observed at PH-W6 and PH-W12 (data not shown). IGF-II Western ligand blots used in homologous and heterologous competition/inhibition studies revealed that at two embryonic (E-6 and E-19) and two post-hatching (PH-1W and PH-3W) stages, vitreal and serum IGFBPs had a higher affinity for IGF-II than IGF-I (data not shown). [‘251]IGF-II binding to vitreal and serum ligand blots was 50% inhibited

IGFBPs

FIG. 4. IGF-II

Western ligand blotting of vitreous and serum from chicks at two posthatching stages (PH-Wl and PH-W3). Fifty microliters of vitreous humor or 6 ~1 serum were subjected to ligand blotting, as described in Fig. 3.

by approximately 0.1 nM unlabeled IGF-II or 1.0 nM unlabeled IGF-I. The apparent affinities of individual vitreal or serum IGFBPs did not appreciably vary across the four developmental stagesexamined. Immunoprecipitation

Although vitreous and serum exhibit different patterns of IGFBP developmental regulation and contain IGFBPs of different sizes, a 33-kDa IGFBP band is common to both compartments throughout development. The characteristics of this IGFBP were, therefore, further examined using an antiserum (h-C20-IGFBP-2) produced against a 20-amino acid sequencenear the carboxyl-terminus of human IGFBP2. Vitreous humor and serum from PH-D2 chicks were immunoprecipitated with IGFBP-2 antiserum and examined by the Western ligand blot technique (Fig. 5). Both vitreal and serum 33-kDa IGFBPs were precipitated by IGFBP-2 antiserum. The immunoprecipitation was specific, since nonimmune rabbit serum was ineffective in precipitating these IGFBPs (see lanes labeled N.S.). Further evidence that the immunoprecipitation of vitreal and serum 33-kDa IGFBPs is specific is the finding that the 28-kDa IGFBP present in these fluids was not immunoprecipitated by this antiserum. The faint 31-kDa band observed in the immunoprecipitated vitreal lane represents a rabbit IGFBP that contaminates the immunoprecipitate. The contaminating rabbit serum IGFBP was also observed when the immunoprecipitation studies were carried out in the absenceof vitreous and serum (data not shown). Deglycosylation

To examine further the similarities and differences between the vitreal and serum 33-kDa IGFBPs, their glycosy-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 18:43 For personal use only. No other uses without permission. . All rights reserved.

DEVELOPMENTAL

IGF-BP-2

lmmunoprecipitation

Vitreous

46-

REGULATION

Serum

’

- 46

30-

- 30

;;I

- 21 -14

-

FIG. 5. Immunoprecipitation and IGF-II Western ligand blotting of vitreal and serum IGFBPs from PH-D2 chicks. Fifty microliters of undiluted IGFBP-2 antiserum (h-C20-IGFBP-2) were incubated with 450 ~1 vitreous humor or serum (1:8), precipitated with protein-A, and subjected to [‘251]IGF-II ligand blotting, as described in Materials and Methods. Total, Nonimmunoprecipitated vitreous humor or serum; N.S., nonspecific, nonimmune rabbit serum immunoprecipitated; Specific, h-C20-IGFBP-2 antiserum immunoprecipitated.

GLYCOSYLATION Vitreous Humor

kDa 69 -

PAllERN Serum

J” ”

46 -

_, M

30-

“Ir

21-

kDa -69 -46 -30 - 21

FIG. 6. 0- and N-linked deglycosylation of PH-W3 chick vitreous humor and serum. Twenty microliters of vitreous humor or 2.5 ~1 serum were treated with N-Glycanase (N-Gly), neuraminidase (Neura), or neuraminidase and 0-Glycanase (0-Gly) and then subjected to Western ligand blotting, as described in Materials and Methods. Similar patterns were observed in two independent experiments.

lation patterns were examined. Samples of vitreous humor and serum from PH-W3 were treated with various deglycosylating enzymes: N-Glycanase alone (cleaves N-linked oligosaccharides),neuraminidase alone (cleaves sialic acid residues), and 0-Glycanase plus neuraminidase (cleaves Olinked oligosaccharides). Samples were then examined by the Western ligand blot technique. Figure 6 shows that the 33-kDa IGFBP found in the

OF AVIAN

VITREAL

IGFBPs

2851

vitreous and serum is differentially glycosylated in the two compartments. Specifically, the apparent mol wt of the serum, but not the vitreal, 33-kDa IGFBP was reduced after deglycosylation with neuraminidase alone or neuraminidase added to 0-Glycanase. The apparent mol wt of the desialated 33-kDa IGFBP was approximately 28 kDa (seelane labeled Neura). When 0-Glycanase was added with the neuraminidase (see lane labeled 0-Gly), the apparent mol wt was slightly, but reliably, further reduced. Attempts to determine whether the vitreal and serum 33-kDa IGFBPs differ in Nlinked glycosylation patterns were complicated by the superimposition of the 42-kDa IGFBP over the 33-kDa position after deglycosylation. In addition to vitreal-serum differences in glycosylation of the 33-kDa IGFBP, these two compartments differed in glycosylation of the 42-kDa IGFBP. In serum, N-Glycanase decreasedthe size of the 42-kDa IGFBP and increased its binding activity. The vitreal42-kDa IGFBP was also reduced in size by N-Glycanase, but there was no indication that its binding activity was increased. N-Glycanase-induced changesin binding activity of the serum 42-kDa IGFBP were observed in two independent experiments. Northern

blots

Since IGFBP-2 is the major BP in vitreous humor and is present at all developmental stages,ocular tissueswere examined for IGFBP-2 mRNA. Northern blots containing poly(A)’ mRNA from E-19 retina, ciliary body, cornea, and liver were probed with a 212-basepair 32P-labeledchicken IGFBP-2 PCR product. The PCR product used to probe the Northern blots was synthesized from E-19 cornea cDNA, sequenced,and found to be 90% identical to rat IGFBP-2 (data not shown). Autoradiographs revealed a transcript approximately 1.8 kilobases in all of the ocular tissues as well as in liver (Fig. 7). Densitometric analysis of the autoradiographs showed that the cornea had the largest amount of IGFBP-2 message,followed by liver, retina, and ciliary body. The differences in IGFBP-2 mRNA abundance persistedwhen the densitometric data were normalized for the amount of P-actin mRNA in each lane. Discussion The results of this study demonstrate the presence of specific, high affinity IGFBPs in embryonic and posthatching chick vitreous humor and serum. Importantly, IGFBPs in the two compartments exhibit differential developmental regulation. IGF binding in the vitreous humor is highest on E-6, decreaseson E-9, and then remains relatively stablethroughout the remainder of embryonic and posthatching development. In contrast, serum IGF binding steadily increases throughout embryonic and posthatching development. The developmental variations in the level of IGF binding are probably not the result of changing endogenous IGF-I and -11levels, since similar developmental trends were observed by Western ligand blotting, a technique that removes endogenous IGF. [‘251]IGF-II Western ligand blot competition/inhibition studies indicate that the affinity for IGF-II does not

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 18:43 For personal use only. No other uses without permission. . All rights reserved.

2852

DEVELOPMENTAL IGFBP-2 Northern Blot Analysis Chicken Ocular Tissues

1.8 Kb-

-1.8

REGULATION of

Kb

FIG. 7. IGFBP-2 Northern blot analysis of mRNA from E-19 cornea, ciliary body, retina, and liver. One microgram of poly(A)+ RNA from each tissue was electrophoresed on a 1.0% formaldehyde-agarose gel, transferred to a nylon membrane, and probed for IGFBP-2 mRNA, as described in Materials and Methods.

vary across development or different IGFBPs (data not shown). Thus, it appearsthat the developmental variation in binding activity is the result of changesin the relative abundance of vitreal and serum IGFBPs. The maintenance of relatively constant levels of vitreal IGF-binding activity during stagesof embryonic eye growth and vitreal volume expansion suggeststhat IGFBPs are actively being added to the vitreous humor in a way that offsets the dilution that would otherwise occur. In contrast to reports of decreasesin vitreal IGFBPs in neonatal rats (32), the posthatching chicken exhibits stable vitreal IGFBP levels throughout adulthood, suggestingeither a lack of turnover or, more likely, an equilibrium between synthesis and degradation. In distinction to the multiple bands observed in Western ligand blots, affinity cross-linking studiesdemonstrate single, albeit broad, bands. The difference between cross-linking and ligand blot techniques suggeststhat someIGFBPs either are not cross-linkable or are differentially saturated with nonexchangeable endogenous ligand. Although affinity cross-linking studies may not reveal all IGFBPs, the finding that the cross-linked vitreal and serum IGFBPs exhibit differences in microheterogeneity suggests that vitreal and serum IGFBPs are structurally distinct. Further evidence that the vitreal and serum IGFBP systems are independent of each other comes from Western ligand blot studiesshowing that the vitreous contains a low mol wt IGFBP (24 kDa) not detected in the serum at any stage. Additionally, ligand blots show that posthatching serum contains a high mol wt IGFBP (70 kDa) not present in the vitreous, The presence of a blood-ocular barrier makes it

OF AVIAN

VITREAL

IGFBPs

Endo. Voll31.

1992 No 6

unlikely that the low mol wt vitreal IGFBPs are proteolytic fragments of larger serum IGFBPs. The pattern of serum IGFBPs observed on Western ligand blots of early posthatching chickens resemblesthat found by Armstrong et al. (39). However, their study did not detect the 70-kDa serum IGFBP reported here. This apparent discrepancy may be related to strain differences or, possibly, differences in the transfer efficiency of the IGFBPs. The finding that serum IGFBP levels steadily increasethroughout posthatching stagesof development is compatible with other observations in both chickens (40) and postnatal mammals (41). The large posthatching increase in the serum 42-kDa IGFBP may be the result of the influences of GH, which has been shown to increase in serum after hatching (42). The similarity between the chick 42-kDa IGFBP and mammalian IGFBP-3 in terms of mol wt (43), glycosylation (44), and possible GH sensitivity (45) suggestthat this IGFBP may be the avian analog of mammalian IGFBP-3. The finding that IGFBP-2 antiserum immunoprecipitates both the vitreal and serum 33-kDa IGFBPs indicates that the two IGFBPs are probably structurally similar to each other and to human IGFBP-2. Although the vitreous humor and serum both contain IGFBP-2, the fact that the serum form is sialated while the vitreal form is not is further evidence for an independent vitreal IGFBP system. It should be pointed out that it is not clear how the vitreal and serum forms of IGFBP-2 can have the sameapparent mol wt while differing in glycosylation. Possibly, differences in secondary structure causeIGFBPs of different sizes to migrate at the same electrophoretic position. Another possibility, is that the vitreal and serum IGFBP-2 represent products of alternative transcription and/or different posttranslational modification. In this scheme, the serum IGFBP-2 would have a smaller protein core, but would have glycosylated side-chainsthat offset the decreasedprotein size. Another glycosylation difference between vitreal and serum IGFBPs is seen in the 42-kDa IGFBP. Here, N-deglycosylation results in an increased binding activity of the serum, but not the vitreal, 42-kDa IGFBP. The increase in the binding activity of the serum 42-kDa IGFBP after Ndeglycosylation is reminiscent of similar findings with insulin (46) and PRL receptors (47). Consistent with the idea that the eye contains an independent vitreal IGFBP system is the finding that IGFBP-2 mRNA is present in ocular tissueslining the vitreal chamber. The present study showing IGFBP-2 mRNA in neural ectodermal tissuessuch as the cornea and retina agreeswith a previous study demonstrating IGFBP-2 mRNA in the developing rat central nervous system by in situ hybridization (48). Further evidence for the local synthesis of vitreal IGFBPs is that in adult mammals, IGFBP-2 mRNA has been demonstrated in retina (34) aswell ascornea and ciliary body (35). It has also been shown that cultured retinal pigmented epithelial cellssynthesize and releaseIGFBPs (32, 49). Taken together, the presence of IGFBP-2 mRNA in developing ocular tissues of the chicken and the fact that vitreal and serum IGFBPs are differentially regulated during development and have distinct structural and glycosylation charac-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 18:43 For personal use only. No other uses without permission. . All rights reserved.

DEVELOPMENTAL

REGULATION

teristics strongly suggest the presence of an autonomous ocular IGFBP system. Moreover, the finding that in both vitreous and serum, individual IGFBPs have unique patterns of developmental regulation suggeststhat individual IGFBPs may have separatedevelopmental functions. Further studies are needed to determine the factor(s) responsiblefor regulating IGFBP levels in the vitreous humor during embryonic development. Becauseof the easeof obtaining tissueand the ability to manipulate the embryonic chick’s environment, this model may be valuable in studying the role of both IGFs and IGFBPs in the development of ocular and nonocular tissues. References 1.

Nissley MM, Yang YW-H, Brown AL, Romanus JA, Adams SO, Kiess W, Nissley SP 1988 Insulin-like growth factors in fetal growth. In: Bercu Hormone. Plenum

2.

biology

endocrinology.

in developing

chick

embryo.

Spaventi R, Antica M, Pavelic K 1990 Insulin growth factor 108:491-495

5.

to molecular

J Exp Zoo1 [Suppl]

Gaspard KJ, Klitgaard HM, Wondergem R 1981 Somatomedin thyroid hormones Med 16624-27

4.

of Growth

De Pablo F, Serrano J, Girbau M, Alemany J, Scavo L, Lesniak MA 1990 Insulin and insulin-like growth factor I action in chick embryo: from 4:187-191

3.

BB (ed) Basic and Clinical Aspects Press, New York, pp 233-249

I (IGF-I)

in early mouse

embryogenesis.

Robcis HL, Caldes T, De Pablo F 1991 Insulin-like

and Proc Sot Exp Biol and

insulin-like Development

growth factorI serum levels show a midembryogenesis peak in chicken that is absent in growth-retarded embryos cultured PX ova. Endocrinology 128:1895-1901 6. Han VKM, D’Ercole J, Lund PK 1987 Cellular localization of somatomedin (insulin-like growth factor) messenger RNA in the human fetus. Science 236:193-197 7. Serrano J, Shuldiner AR, Roberts CT, LeRoith D, DePablo F 1990 The insulin-like growth factor I (IGF-I) gene is expressed in chick embryos during early organogenesis. Endocrinology 127:1547-1549 8. Girbau M, Bassas L, Alemany J, DePablo F 1989 in situ autoradiography and ligand dependent tyrosine kinase activity reveal insulin receptors and insulin-like growth factor-l receptors in pre-pancreatic chicken embryos. Proc Nat1 Acad Sci USA 86:5868-5872 9. Sara VR, Hall K, Misaki M Fryklund L, Christensen N, Wetterberg L 1983 Ontogenesis of somatomedin and insulin receptors in human fetus. J Clin Invest 71:1084-1094 10. Beebe DC, Silver H, Belcher KS, Van Wyk JJ, Svoboda ME, Zelenka PS 1987 Lentropin, a protein that controls lens fiber formation, is related functionally and immunologically to the insulin-like growth factors, Proc Nat1 Acad Sci USA 84:2327-2330 11. Svrzic D, Schubert D 1990 Insulin-like growth factor 1 supports embryonic nerve cell survival. Biochem Biophys Res Commun 172:54-60 12. Ewton D, Florini JR 1981 Effects of the somatomedins and insulin on myoblast differentiation in-vitro. Dev Biol 86:31-39 13. Rechler MM, Nisslev SP 1985 The nature and regulation of the receptors for insulin-like growth factors. Annu Rev fhysiol47:425442 14. Humbel RE 1990 Insulin-like growth factors I and II. Eur J Biochem 190:445-462 15. Sara VH, Hall K 1991 Insulin-like growth factors and their binding proteins. Physiol Rev 70:591-614 16. Gray A, Tam AW, Dull TJ, Hayflick J, Pintar J, Cavanee WK, Koufos A, Ullrich A 1987 Tissue-specific and developmentally regulated transcription of the insulin-like growth factor 2 gene. DNA 6:283-295 17. Stylianopoulou F, Efstratiadis A, Herbert J, Pintar J 1988 Pattern of the insulin-like growth factor II gene expression during rat

OF AVIAN embryogenesis. 18.

VITREAL

IGFBPs

Development

2853

103:497-506

Brice AL, Cheetham JE, Bolton VN, Hill NCW, Schofield PN 1989 Temporal changes in the expression of the insulin-like growth factor II gene associated with tissue maturation in the human fetus. Development 106:543-554

19.

Cuthbertson RA, Beck F, Senior PV, Haralambidis J, Penschow JD, Coughlan JP 1989 Insulin-like growth factor II may play a local

20.

Shimasaki

21.

Orlowski CC, Brown AL, Guck TO, Yang YWH, Tseng LYH, Rechler MM 1990 Tissue, developmental, and metabolic regulation

role in the regulation

of ocular

size. Development

107:123-130 and molecular characterization of three novel insulin-like growth factor binding proteins (IGFBP-4, 5 and 6). In: Spencer EM (ed) Modern Concepts of Insulin-Like Growth Factors. Elsevier, New York, vol 1:343-358

S, Zhang HP, Ling N 1991 Isolation

of messenger factor-binding

ribonucleic acid encoding a rat insulin-like growth protein, Endocrinology 126:644-652 22. Wood TL, Brown AL, Rechler MM, Pintar JE 1990 The expression pattern of an insulin-like growth factor (IGF)-binding protein gene is distinct from IGF-II in the midgestational rat embryo. Mel Endocrinol 4:1257-1263 23. Lee CY, Bazer FW, Etherton TD, Simmen FA 1991 Ontogeny of insulin-like growth factors (IGF-I and IGF-II) and IGF binding proteins in porcine serum during fetal and postnatal developmen; Endocrinologv 128:2336-2344 24. Elgin RG, Busby WH, Clemmons DR 1987 An insulin-like growth factor (IGF) binding protein enhances the biologic response to IGFI. Proc Nat1 Acad Sci USA 84:3254-3258 25. Blum WF, Jenne EW, Reppin JF, Kietzmann K, Ranke MB, Bierich JR 1989 Insulin-like growth factor I (IGF-I)-binding protein complex is a better mitogen than free IGF-I. Endocrinology 125:766-771 26. Rutanen EM, Pekonen F, Makinen T 1988 Soluble 34K binding protein inhibits the binding of insulin-like growth factor I to its cell receptors in human secretory phase endometrium: evidence for Autocrine/Paracrine regulation of growth factor action. J Clin Endocrinol Metab 66: 173-l 79 27. Burch WM, Correa J, Shively JE, Powell DR 1990 The 25-kilodalton insulin-like growth factor (IGF)-binding protein inhibits both basal and IGF-I-mediated growth of chick embryo pelvic cartilage in-vitro. J Clin Endocrinol Metab 77:173-180 . _ 28. Gouinath R, Walton PE, Etherton TD 1989 An acid-stable insulinlike’ growth.factor (IGFj-binding protein from pig serum inhibits binding of IGF-I and IGF-II to vascular endothelial cells. J Endocrinology 120:231-236 29. Cunha-Vaz J 1979 The blood ocular barriers. Surv Ophthalmol 23~279-296 30. Latker CH, Beebe DC 1984 Developmental changes in the bloodocular barriers in chicken embryos. Exp Eye Res 39:401-414 31. Grant M, Russell B, Fitzgerald C, Merimee TJ 1986 Insulin-like growth factors in vitreous, Studies in control and diabetic subjects with neovascularization. Diabetes 35:416-420 32. Ocrant I, Fay CT, Parmelee JT 1991 Expression of insulin and insulin-like growth factor receptors and binding proteins by retinal pigment epithelium. Exp Eye Res 52:581-589 33. Waldbillig RJ, Pfeffer B, Schoen TJ, Adler A, Shen-Orr Z, Scavo L, LeRoith D, Chader GJ 1991 Evidence for an insulin-like growth factor-I autocrine-paracrine system in the retinal photoreceptorpigment epithelial cell complex. J Neurochem 57:1522-1533 34. Agarwal N, Hsieh CL, Sills D, Swaroop M, Desai 8, Francke U, Swaroop A 1991 Sequence analysis, expression and chromosomal localization of a gene isolated from a subtracted human retina cDNA library, that encodes an insulin-like growth factor-binding protein (IGF-BP). Exp Eye Res 52:549-561 35. Arnold DR, Schoen TJ, Jones E, Chader GJ, Waldbillig 1991 High levels of IGF-binding protein (IGFBP) in cornea, aqueous and vitreous and synthesis of IGFBP-2 in the ocular iris/ciliary body complex. Invest Ophth Vis Sci 32:755 (Abstract) 36. Hossenlopp P, Seurin D, Segovia-Quinson B, Hardoun S, Binoux M 1986 Analysis of serum insulin-like growth factor binding proteins using Western blotting: use of the method for titration of the binding proteins and competitive binding studies. Anal Biochem 154:138-143 37. Schowalter DB, Sommer SS 1989 The generation of radiolabeled

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 18:43 For personal use only. No other uses without permission. . All rights reserved.

2854

38.

39.

40.

41.

42.

43.

DEVELOPMENTAL

REGULATION

DNA and RNA probes with polymerase chain reaction. Anal Biochem 177:90-94 Sambrook J, Fritsch EF, Maniatis T 1989 Molecular Cloning-A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor Armstrong DG, McKay CO, Morel1 DJ, Goddard C 1989 Insulinlike growth factor-I binding proteins in serum from the domestic fowl. J Endocrinol 120:373-378 Lee I’D, Peacock A, Roessler MK, Hester J, Reeves JT 1989 Insulinlike growth factor (IGF)-I and IGF-binding activity in normal and fast growing chickens. Life Sci 45:2465-2470 Hardouin S, Gourmelen M, Noguiez P, Seurin D, Roghan M, Bout YL, Povoa G, Merimee TJ, Hossenlopp P, Binoux M 1989 Molecular forms of serum insulin-like growth factor (IGF)-binding proteins in man: relationships with growth hormone and IGFs and physiological significance. J Clin Endocrinol Metab 69:1291-1301 Kikuchi K, Buonomo FC, Kajimoto Y, Rotwein P 1991 Expression of insulin-like growth factor-I during chicken development, Endocrinology 128:1323-1328 Conover CA, Ronk M, Lombana F, Powell DR 1990 Structural and biological characterization of bovine insulin-like growth factor

OF AVIAN

VITREAL

IGFBPs

Endo. Vol131.

1992 No 6

binding protein-3. Endocrinology 127:2795-2803 44. Baxter RC, Martin JL 1989 Structure of the Mr 140,000 growth hormone-dependent insulin-like growth factor binding protein complex: determination by reconstitution and affinity-labeling. Proc Nat1 Acad Sci USA 86:6898-6902 45. Baxter RC, Martin JL 1986 Radioimmunoassay of growth hormonedependent insulin like growth factor binding protein in human plasma. J Clin Invest 78:1504-1512 46. Podskalny JM, Rouiller DG, McElduff A, Gorden P 1986 Glucose starvation alters insulin but not IGF-I binding to Chinese hamster ovary (CHO) cells. Biochem Biophys Res Commun 140:821-826 47. Lascols 0, Capeau J, Cherqui G, Caron M, Bachimont J, Picard J 1989 Glycosylation characteristics of the mouse liver lactogenic receptor. Mol Cell Endocrinol 65:145-155 48. Wood TL, Brown AL, Rechler MM, Pintar JE 1990 The expression pattern of an insulin-like growth factor (IGF)-binding protein gene is distinct from IGF-II in the midgestation rat embryo. Mol Endocrino1 4:1257-1263 49. Waldbillig RJ, Schoen TJ, Chader GJ, Pfeffer B 1992 Monkey retinal pigment epithelial cells in-vitro synthesize, secrete and degrade insulin-like growth factor binding proteins. J Cell Physiol 150:76-83

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 November 2015. at 18:43 For personal use only. No other uses without permission. . All rights reserved.