Search

Protein glycosylation is a diverse form of post-translational modification. Two to three consecutive O-linked N-acetylgalactosamines (Tn-antigens) are recognized by antibodies such as MLS128. MLS128 mAb inhibited cell growth and bound to a 110 kDa glycoprotein (GP) in LS180 and HT29 colon cancer cells. However, purification and identification of the 110 kDa GP was unsuccessful due to its low abundance. The present study used a highly sophisticated and sensitive mass spectrometry method to identify proteins immunoprecipitated with MLS128 and separated by two-dimensional gel electrophoresis. Three desmosome components were identified. Of these, desmocollin and desmoglein shared many similar characteristics, including molecular mass, pI, and potential Tn-antigen sites. Western blotting analyses of LS180 cell lysates revealed a common 110 kDa band recognized by MLS128 and anti-desmocollin, but not by anti-desmoglein. Immunofluorescence microscopy of LS180 cells revealed that desmocollin is membrane-bound, while desmoglein is primarily localized in the cytosol. Confocal microscopy demonstrated colocalization of the desmocollin-specific antibody with the MLS128 antibody on the cell membrane, suggesting that desmocollin may contain Tn-antigens recognized by MLS128. Treatment of LS180 cells with siRNA to knock down desmocollin expression or a desmocollin-specific antibody decreased cell viability, suggesting a critical role for this protein in cell growth and survival. N-glycosidase F digestion of the 110 kDa GP and desmocollin suggested that although both proteins contain N-glycosylation sites, they are not identical. These findings suggest that desmocollin colocalizes with the 110 kDa GP and that growth inhibition induced by the MLS128 antibody may be mediated through a mechanism that involves desmocollin.

The production of human single-chain variable-fragments (scFvs) against carbohydrate antigens by phage display technology is seemingly a logical strategy towards the development of antibody therapeutics, since carbohydrates are self-antigens. Panning and screening of phages displaying human scFvs using a variety of neoglycolipids presenting structurally-defined carbohydrates resulted in a number of candidate phage clones as judged by cautious evaluation of DNA sequences and specific binding to carbohydrate moieties of interest. ScFv proteins were expressed in prokaryotic or eukaryotic cells from the respective genes. The characterization of isolated scFvs gene products after establishing expression, production and purification of scFv protein in different expression systems demonstrated that the production of scFv-human IgG1 Fc conjugates were originally sufficient in the media of stably-transfected cells, but declined during early passages. Bacterial expression of soluble scFv proteins with binding activity suffered low yields, whereas overexpressed scFv proteins formed inclusion bodies, which required refolding. An insect cell expression system producing soluble and active scFv proteins was found to be cost- and time-effective. The best expression system and fine adjustments for the conditions to prepare active forms had to be determined for each scFv protein. The successful production of active scFv proteins seems to be dependent on their DNA and/or amino acid sequences.

T-antigen (Galβ1-3GalNAcα-1-Ser/Thr) is an oncofetal antigen that is commonly expressed as a carbohydrate determinant in many adenocarcinomas. Since it is associated with tumor progression and metastasis, production of recombinant antibodies specific for T-antigen could lead to the development of cancer diagnostics and therapeutics. Previously, we isolated and characterized 11 anti-T-antigen phage clones from a phage library displaying human single-chain antibodies (scFvs) and purified one scFv protein, 1G11. More recently, we purified and characterized 1E8 scFv protein using a Drosophila S2 expression system. In the current study, four anti-T-antigen scFv genes belonging to Groups 1-4 were purified from inclusion bodies expressed in Escherichia coli cells. Inclusion bodies isolated from E. coli cells were denatured in 3.5 M Gdn-HCl. Solubilized His-tagged scFv proteins were purified using Ni(2+)-Sepharose column chromatography in the presence of 3.5 M Gdn-HCl. Purified scFv proteins were refolded according to a previously published method of step-wise dialysis. Two anti-T-antigen scFv proteins, 1E6 and 1E8 that belong to Groups 1 and 2, respectively, were produced in sufficient amounts, thus allowing further characterization of their binding activity with T-antigen. Specificity and affinity constants determined using enzyme-linked immunosorbent assay (ELISA) and surface plasmon resonance (SPR), respectively, provided evidence that both 1E8 and 1E6 scFv proteins are T-antigen specific and suggested that 1E8 scFv protein has a higher affinity for T-antigen than 1E6 scFv protein.

An aqueous extract from toad skin, cinobufacini, has been known to possess anticancer ability. The present study examined effect of toad skin extract on activity of gelatinases including matrix metalloproteinases-2 and -9 which play an important role in invasion of carcinoma cells. Gelatinase activities derived from fetal serum albumin and culture medium of human breast carcinoma cell line MDA-MB-231 were significantly prevented in the presence of toad skin extract. The inhibitory activity was found in water-soluble fraction of the extract prepared by the Bligh & Dyer method but not in CHCl(3)-soluble lipid fraction. These results suggest that an aqueous extract from toad skin contains a water-soluble substance possessing a potent ability to prevent gelatinase activity. In conclusion, the water-soluble substance in toad skin extract cinobufacini may be able to regulate cancer cell migration accelerated by matrix metalloproteinases.

Anti-Tn-antigen monoclonal antibody MLS128 has affinity for three consecutive Tn-antigens (Tn3) more than Tn2. The major aim of this study was to isolate genes encoding MLS128 variable domains to produce a large quantity of recombinant MLS128 antibodies, in turn, allowing the conduct of studies on precise interactions between Tn3- or Tn2-epitopes and MLS128. This study describes cloning of the variable region genes of MLS128, construction of the variable region genes in single-chain variable fragments (scFv) and two scFvs conjugated with human IgG(1) hinge and Fc regions (scFv-Fc) types, and their respective expression in bacterial and mammalian cell. MLS128 scFv protein with the expected specificity and affinity was successfully prepared from inclusion bodies accumulating in Escherichia coli. Construction, expression and purification of two types of MLS128-scFv-Fc proteins with differing linker lengths in Chinese hamster ovary cells demonstrated that the purified scFv-Fc proteins had binding activity specific to the glycoprotein-expressing Tn-antigen clusters. These results revealed that VL and VH genes cloned from the hybridoma represent those of MLS128 and that recombinant antibodies produced from these genes should provide sufficient amounts of binding domains for use in 3D structural studies such as NMR and X-ray analysis.

The binding ability of anti-insulin-like growth factor Ι receptor (IGF-ΙR) single-chain variable fragments (scFvs) to IGF-IR was measured in the presence of plant lectins. Combinations of concanavalin A (Con A), wheat germ agglutinin (WGA), or peanut agglutinin (PNA) and 1H7 or 3B7 anti-IGF-ΙR scFv/phage antibodies that were previously produced and characterized were used. WGA inhibited binding of both scFvs proteins to the receptor. PNA slightly enhanced the binding of 1H7 scFv and phage antibody to the receptor. Con A led to enhancement of 3B7 scFv-binding but had no effect in a test of phage antibodies and determination of kinetic parameters. The effect of lectins differed for scFvs and phage antibodies, implying that affinity altered by lectins is dependent upon the molecular structure of the antibodies. Results indicated that animal lectins may affect the affinity of therapeutic antibodies targeting cell membrane receptors in vivo.

Anti-mannotriose (Man3) antibodies were previously isolated from a Keio phage library displaying human single chain variable fragments (scFvs) using a neoglycolipid, Man3- dipalmitoylphosphatidylethanolamine. Of three genes constructed, the 5A3 clone was expressed in mouse myeloma NS0 cells as a conjugate with human IgG1 Fc (scFv-Fc) and characterized (Sakai, K., Shimizu, Y., Chiba, T., Matsumoto-Takasaki, A., Kusada, Y., Zhang, W., Nakata, M., Kojima, N., Toma, K., Takayanagi, A., Shimizu, N., and Fujita-Yamaguchi, Y. (2007) Biochemistry 46, 253–262; Zhang, W., Matsumoto-Takasaki, A., Kusada, Y., Sakaue, H., Sakai, K., Nakata, M., and Fujita-Yamaguchi, Y. (2007) Biochemistry 46, 263–270). Similarly, anti-Lex phages were screened from the same library with lacto-N-fucopentaose III (LNFPIII; Lex)-dipalmitoylphosphatidylethanolamine. Of five phage clones isolated, two scFv genes were constructed to express scFv-Fc proteins in NS0 cells. As was experienced with anti-Man3 scFv-Fc clones, only one anti-LNFPIII clone, 1F12, was successfully produced and purified as an scFv-Fc protein. Although anti-LNFPIII 1F12 and anti-Man3 5A3 scFv-Fc proteins were secreted into media, a decline in scFv-Fc production was observed with both stable clones during early passages. Transient expression of anti-LNFPIII and anti-Man3 scFv-Fc genes in COS-7 cells and subsequent analyses of scFv-Fc protein expression revealed accumulation of translated proteins in the endoplasmic reticulum for scFv-Fc proteins derived from clones that did not survive as stable clones. This report describes the following: (i) isolation of anti-LNFPIII scFv genes; (ii) purification of anti-LNFPIII scFv-Fc proteins from stably and transiently expressed cells; and (iii) extracellular or intracellular localization of two anti-LNFPIII and three anti-Man3 scFv-Fc proteins. The results suggest that expression of anti-Man3 and other anti-carbohydrate antibodies in mammalian cells is disadvantageous for cell growth.

The Tn-antigen, GalNAcalpha-Ser/Thr, is a tumour-associated carbohydrate antigen that may provide a sensitive and specific marker for pre-clinical detection of carcinoma and a target for cancer therapies. We recently reported that MLS128 monoclonal antibody treatment significantly inhibited colon and breast cancer cell growth. On the basis of our observations, the present study aimed to produce human anti-Tn-antigen antibodies with specificity similar to that of MLS128 monoclonal antibody, which recognizes a structure of three consecutive Tn-antigens (Tn3). Six phage clones displaying human single-chain variable fragments (scFvs) were isolated from a newly constructed phage library by panning and screening with a synthetic Tn3-peptide. Deduced amino-acid sequences of six anti-Tn3 scFvs exhibited a high degree of homology. Of those, anti-Tn3 4E10 and 4G2 scFv proteins were successfully purified from phage-infected Escherichia coli to near homogeneity. Surface plasmon resonance analyses revealed a K(D) of purified scFv proteins for Tn3 on an order of 10(-7) M, which is high for carbohydrate-specific monovalent antibodies. Further analyses suggested that both scFv proteins also bind to Tn2 and cultured colon and breast cancer cells. These results demonstrated the potential for use of these scFvs in developing antibody therapeutics targeting colon and breast cancer.

Anti-influenza hemagglutinin (HA) Fabs were isolated from a phage display library using purified HA of influenza virus A/Puerto Rico/8/34 (PR8; H1N1) as an antigen. Four Fab clones displaying a 25-50-fold higher binding signal to PR8 HA than the control were selected for further analysis and comparison with anti-PR8 monoclonal antibody (mAb). All four Fabs and mAb recognized the PR8 HA under non-reducing conditions but rarely bound to reduced PR8 HA. Inhibition of influenza virus infection on MDCK cells was observed with Fab1 and mAb in a dose-dependent manner while Fab3 and 4 exhibited only a partial inhibitory effect. Moreover, Fab1 clone and mAb exhibited cross-reactivity with the A/Peking/262/95 (A/Peking; H1N1) strain. The inhibitory effects of mAb on both influenza strains were more potent than Fab1, which is likely attributed to its higher affinity for the antigen. SPR analyses, in fact, revealed that Fab1 and mAb have K(D) of 1.5 x 10(-8) and 3.2 x 10(-9)M, respectively. These results strongly suggest that phage library-derived Fabs can be readily prepared and that such HA-specific Fabs with inhibitory action on influenza infection may be used to treat influenza patients.

Recombinant antibody consisting of the single-chain variable fragment (scFv) of 1H7 monoclonal antibody against insulin-like growth factor-I receptor (IGF-IR) and human IgG(1) Fc domain, scFv-Fc, has been found to exhibit inhibitory effects on breast cancer growth in vitro and in vivo [Li et al. (2000) Cancer Immunol. Immunother. 49, 243; Sachdev et al. (2003) Cancer Res. 63, 627]. Various types of scFvs from hybridomas producing 1H7 or 3B7 mAb were constructed using conventional phage display technology to further characterize the specificity and affinity of anti-IGF-IR mAbs. Binding studies performed using either phage antibodies or soluble scFv proteins to IGF-IR or insulin receptor (IR) and IGF-IR pre-incubated with mAbs suggested that (i) 1H7 and 3B7 bind to IGF-IR but do not bind to its structurally related IR, (ii) either the VL-VH or VH-VL sequence order does not apparently affect specificity for IGF-IR and (iii) 1H7 and 3B7 bind the independent epitopes, located in or near the N-terminal (440-514) and C-terminal (62-184) domains of the alpha subunit, respectively. This study not only revealed new information on binding regions for two anti-IGF-IR mAbs, but also provided the scFv genes as tools for further manipulation of the affinity or development of new IGF-IR-targeted cancer therapeutics.

Affinity chromatography is an efficient method to isolate proteins by taking advantage of their affinities for specific molecules such as substrates, inhibitors, antigens, ligands, antibodies, and other interacting molecules, including subunits. Nowadays, we take the effectiveness and excellence of this technology for granted. This essay will mainly cover the use of affinity chromatography based on my experience.

Insulin-like growth factor (IGF)-II plays an important role in fetal growth and development. IGFs are potent mitogens for a variety of cancer cells. A paracrine/autocrine role of IGF-II in the growth of breast and prostate cancer cells has been suggested. To test the role of IGF-II in cancer cell growth, hammerhead ribozymes targeted to human IGF-II RNA were constructed. Single (R)- and double (RR)-ribozymes were catalytically active in vitro whereas mutant ribozymes (M or MM) did not cleave IGF-II RNA. RR was more active than R. In human prostate cancer PC-3 cells, both R and RR similarly suppressed IGF-II messenger RNA (mRNA) levels (approximately 40%) compared with the level in parental or M-expressing PC-3 cells. Polymerase II and III promoter-driven R similarly suppressed IGF-II mRNA levels. Suppression of IGF-II mRNA levels by R was associated with suppression of IGF-II protein levels. R- (or RR-) expressing PC-3 cells did not grow under serum-starved conditions and showed prolonged doubling times in the presence of 10% FCS compared with those of parental or M-expressing cells. These results substantiated that IGF-II plays a critical role in prostate cancer cell growth.

Receptors for insulin and insulin-like growth factor-I (IR and IGFIR) consisting of the alpha2beta2 structure are protein tyrosine kinases (PTKs). Carboxyl-terminal (CT) domains of their beta subunits are structurally diverse while the PTK domains share the highest homology. Interactions between CT and PTK domains of IR and IGFIR were studied by means of PTK activity, fluorescence energy transfer or surface plasmon resonance using BIAcore. We present evidence that IGFIR CT directly interacts with both IGFIR and IR. Although binding to both receptors, stimulation of PTK activity only occurs with IR but not IGFIR.

T-antigen (Galβ1-3GalNAcα-1-Ser/Thr), also known as Thomsen-Friedenreich antigen (TF antigen), is an oncofetal antigen commonly found in cancerous tissues. Availability of anti-T-antigen human antibodies could lead to the development of cancer diagnostics and therapeutics. Four groups of single-chain variable fragment (scFv) genes were previously isolated from a phage library (Matsumoto-Takasaki et al. (2009) Isolation and characterization of anti-T-antigen single chain antibodies from a phage library. BioSci Trends 3:87-95.). Here, four anti-T-antigen scFv genes belonging to Group 1-4 were expressed and produced in a Drosophila S2 cell expression system. ELISA and surface plasmon resonance (SPR) analyses confirmed the binding activity of 1E8 scFv protein to various T-antigen presenting conjugates. NMR experiments provided evidence of the folded nature of the 1E8 scFv protein. ScFv-ligand contact was identified by STD NMR, indicating that the galactose unit of T-antigen at the non-reducing end was primarily recognized by 1E8 scFv. This thus provides direct evidence of T-antigen specificity.

Tn-antigen (GalNAcα-Ser/Thr) is one of the most common aberrations associated with cancer progression and metastasis, and thus is an excellent target for development of cancer diagnostics and therapeutics. MLS128 monoclonal antibody (mAb), derived from a mouse immunized with human colon carcinoma cells, was reported to bind to two or three consecutive Tn-antigens (Tn2 or Tn3) with one-order higher affinity for Tn3 than for Tn2. Our recent studies demonstrated that MLS128 significantly inhibits breast and colon cancer cell growth. Molecular cloning of the variable regions of heavy (VH) and light (VL) chains revealed that the VH sequence of MLS128 shared 97% nucleotide sequence identity with the VH of 83D4 mAb, derived from breast cancer-immunized mice, which has a similar affinity for Tn2/Tn3. MLS128 single-chain antibodies (scFv) and scFv-Fc were constructed to confirm the affinity for synthetic Tn2/Tn3 peptides. Thermodynamic studies on MLS128 binding to Tn2/Tn3 revealed its unique nature of temperature-dependent binding.

MLS128 is an anti-carbohydrate monoclonal antibody (mAb) that binds three or two consecutive Tn-antigens. MLS128 bound 110-210 kDa glycoproteins (GPs) and inhibited the growth of LS180 and HT29 colon and MCF-7 breast cancer cells. One possible mechanism of MLS128's inhibition of growth may be via insulin-like growth factor-I receptor (IGF-IR) down-regulation (Morita et al. BioScience Trends. 2009; 3:32-37). The current study examined the role of IGF-IR signaling in the growth of colon cancer cells and its possible interaction with MLS128-induced inhibition of cell growth in LS180, LS174T, and HT29 human colon cancer cells treated with MLS128 or anti-IGF-IR 1H7. Both MLS128 and 1H7 treatment significantly inhibited the growth of colon cancer cells. All three colon cancer cell lines expressed IGF-IR. Their growth was in part IGF-I dependent, but inhibition by MLS128 was independent of IGF-IR signaling. All of the colon cancer cell lines expressed an 110kDa GP for MLS128 binding, but MCF-7 cells expressed MLS128-detectable bands with higher molecular masses. 1H7 treatments caused down-regulation of IGF-IR but did not affect 110kDa GP levels. MLS128 treatments resulted in partial disappearance of the 110kDa band but did not affect IGF-IR levels. Western blotting analyses of colon and breast cancer cell lysates revealed that colon and breast cancer cells differed significantly in patterns of expression of growth-related molecules while colon cancer cells were similar but distinctive. In conclusion, MLS128 inhibited the growth of colon cancer cells by binding to the 110kDa GP receptor. Inhibition of growth by MLS128 did not appear to affect IGF-IR signaling and instead only affected other growth signaling pathways.

Our previous study using a deletion mutant indicated that the COOH-terminal (CT) domain of the insulin receptor plays important roles in both catalytic efficiency and stability of the receptor kinase (Yan et al., J. Biol. Chem., 268 [1993] 22444). In this study, we purified the CT domain of 98 amino acids from bacterial cells over-expressing the CT domain and examined its effect on insulin and IGF-I receptor protein kinases. The purified CT domain stimulated the kinase activities of purified insulin receptor-transmembrane/cytoplasmic domain (IRTMTPK) and its CT domain-deletion mutant (IRTMTPK delta CT), 3.3-fold and 2.3-fold, respectively, while it was less effective in stimulating the kinase activity of purified IGF-I receptor transmembrane/cytoplasmic domain (IGFIRTMTPK) (1.4-fold). When the effect of the CT domain on autophosphorylation was examined, a marked increase in autophosphorylation was observed only with IRTMTPK delta CT. These results suggest that the CT domain specifically interacts with the insulin receptor cytoplasmic domain, thereby activating the kinase or autophosphorylation activity.

Tn-antigens are tumour-associated carbohydrate antigens that are involved in metastatic processes and are associated with a poor prognosis. MLS128 monoclonal antibody recognizes the structures of two or three consecutive Tn-antigens (Tn2 or Tn3). Since MLS128 treatment inhibits colon and breast cancer cell growth [Morita, N., Yajima, Y., Asanuma, H., Nakada, H., and Fujita-Yamaguchi, Y. (2009) Inhibition of cancer cell growth by anti-Tn monoclonal antibody MLS128. Biosci. Trends 3, 32-37.], understanding the interaction between MLS128 and Tn-clusters may allow us to the development of novel cancer therapeutics. Although MLS128 was previously reported to have specificity for Tn3 rather than Tn2, similar levels of Tn2/Tn3 binding were unexpectedly observed at 37°C. Thus, thermodynamic analyses were performed via surface plasmon resonance (SPR) using synthetic Tn2- and Tn3-peptides at 10, 15, 20, 25 and 30°C. SPR results revealed that MLS128's association constants for both antigens were highly temperature dependent. Below 25°C MLS128's association constant for Tn3-peptide was clearly higher than that for Tn2-peptide. At 30°C, however, the association constant for Tn2-peptide was higher than that for Tn3-peptide. This reversal of affinity is due to the sharp increase in K(d) for Tn3. These results were confirmed by NMR, which directly measured MLS128-Tn binding in solution. This study suggested that thermodynamic control plays a critical role in the interaction between MLS128/Tn2 and MLS128/Tn3.

We have previously expressed the human insulin receptor beta subunit domain containing transmembrane and cytoplasmic domains (IRTMTPK) in insect cells, and showed that the purified IRTMTPK was highly active (Li, S. L., Yan, P.-F., Pax, I. B., and Fujita-Yamaguchi, Y. (1992) Biochemistry 31, 12455-12462). To investigate the role of COOH-terminal and acidic domains of the insulin receptor kinase, we have expressed deletion mutants IRTMTPK delta CT (delta 76 amino acids) and IRTMTPK delta Acid (delta 19 amino acids). Both enzymes were purified by a one-step method using the same immunoaffinity column as used for IRTMTPK. While Km and Vmax for prephosphorylated IRTMTPK and delta Acid mutant enzyme determined using poly(Glu, Tyr)(4:1) were similar, catalytic efficiency of the delta CT mutant enzyme was significantly lower than those of IRTMTPK and delta Acid mutant enzyme as judged by Km and Vmax. Experiments for thermostability and susceptibility to proteases revealed that Tm of delta CT mutant enzyme was 3.5 degrees C lower than that of IRTMTPK enzyme (= 33.3 degrees C) and that delta CT mutant enzyme was digested by either trypsin or Lys-C into a 28,000 core domain much faster than IRTMTPK. Activation of delta CT mutant enzyme by polylysine was less significant than that of IRTMTPK and delta Acid mutant enzyme, approximately 4-versus approximately 17-fold. These studies suggested that the COOH-terminal domain plays important roles in both catalytic efficiency and stability of the insulin receptor kinase, and that the acidic domain by itself is not responsible for kinase activation by polylysine.

We have expressed, purified, and characterized the insulin receptor protein tyrosine kinase (PTK) retaining the transmembrane and downstream domains. The proteins expressed in insect cells using a baculovirus expression system were identified as membrane-bound by immunofluorescence staining and biochemical characterization. One-step purification by immunoaffinity chromatography from Triton X-100 cell extracts resulted in a approximately 360-fold increase in the specific kinase activity with a yield of approximately 50%. An appMr = approximately 60,000 protein was the major component identified by both silver staining of the purified enzyme and immunostaining of the crude extracts after separation by sodium dodecyl sulfate-polyacrylamide gel electrophoresis under reducing conditions. Using nondenaturing conditions, the molecular weight was estimated to be approximately 250,000 and approximately 500,000 by glycerol gradient centrifugation and gel permeation chromatography, respectively, suggesting that oligomers of the beta-subunit domains such as tetramers and octamers are formed. The basal PTK activity of this enzyme was much higher than those of previously reported soluble-form insulin receptor PTKs expressed in insect cells or the native receptor. Km and Vmax for two substrates, src-related peptide and poly(Glu, Tyr) (4:1), were 2.4 mM and 2.5 mumol min-1 mg-1 and 0.26 mM and 1.2 mumol min-1 mg-1, respectively. Specific activities measured under two previously reported conditions using histone H2B as a substrate were 100 or 135 nmol min-1 mg-1, in contrast to those of soluble PTKs which were reported to be 20 or 70 nmol min-1 mg-1, respectively. The purified enzyme was autophosphorylated at Tyr residues. Autophosphorylation activated the enzyme approximately 3-fold.(ABSTRACT TRUNCATED AT 250 WORDS)

Several groups including us reported that basic proteins and polycations activate the insulin receptor tyrosine-specific protein kinase (TPK) in vitro. However, some inconsistency has become obvious in the observations. The most intriguing was the brief description by Morrison et al. [(1989) J. Biol. Chem. 264, 9994-10001] that polylysine had no effect on the IGF-I receptor TPK despite its 84% identity to the insulin receptor TPK. In the present study, we used highly purified IGF-I and insulin receptor TPKs in an effort to solve the discrepancies noted in the recent publications and to reveal the mechanism by which polycations stimulate the receptor TPKs. We report that the IGF-I receptor TPK is stimulated by polycations and basic proteins in a manner similar to their effects on the insulin receptor TPK. When effects of polylysine and polyarginine on both receptor TPKs were closely compared, subtle qualitative differences were found: Polylysine stimulated autophosphorylation and exogenous substrate phosphorylation activities of both insulin receptor TPK and IGF-I receptor TPK similarly. In contrast, another polycation, polyarginine, affected both TPKs in a manner quite different from polylysine: Polyarginine stimulated insulin receptor autophosphorylation to a greater extent than polylysine did while it had a very small effect on the IGF-I receptor autophosphorylation as well as the exogenous substrate phosphorylation activities of the two receptor TPKs. We have further extended the studies to include the domains of natural proteins which contain a polylysine-like sequence.(ABSTRACT TRUNCATED AT 250 WORDS)

In a study that was reported from this laboratory, the mitogenic potency of an apparent mol wt (appMr) of 15,000 precursor form of human insulin-like growth factor-II (hIGF-II) was shown to be greater than that of completely processed hIGF-II for human fetal-derived fibroblasts, and both were more potent than rIGF-I. Since it is generally acknowledged that the stimulation of cell replication by the IGFs is mediated by IGF-I receptors, we undertook to determine whether differences between the receptors' affinity for the two Mr forms of hIGF-II and recombinant IGF-I (rIGF-I) or between its efficiency to couple specific growth factor occupancy to the activation of protein kinase could explain the greater replicating potential of appMr 15,000 hIGF-II. Equilibrium dissociation, i.e. Kd, and inhibition, i.e. Ki, constants were determined by measuring the ability of rIGF-I, hIGF-II, appMr 15,000 hIGF-II, insulin, and the antireceptor monoclonal antibody alpha IR-3 to compete with 125I-labeled rIGF-I and hIGF-II for binding to purified preparations of IGF-I receptors prepared from an enriched source of fetal membrane, i.e. human term placenta. The results of these experiments established that 1) hIGF-II and appMr 15,000 hIGF-II bind to the IGF-I receptor with the same affinity as rIGF-I, e.g. with Kd and Ki values between 0.03-0.07 nM; 2) the total binding capacity, i.e. Ro, for IGF-I binding was not statistically different from the Ro calculated for IGF-II binding; and 3) the statistical analysis of 12 data sets from the competitive binding experiments for goodness of fit indicated that a 1-site model for IGF-I and -II binding was a better fit of the data than a 2-site model. Measurements of the stimulation of IGF-I receptor autophosphorylation at low ligand concentrations established that appMr 15,000 hIGF-II and hIGF-II were more effective than rIGF-I in coupling receptor occupancy to the activation of its protein kinase. At saturating ligand concentrations, the 3 had similar potencies. The original preparation of appMr 15,000 hIGF-II contains a mixture of forms with acidic isoelectric points (pIs) and was more potent than Mr 7,500 IGF-II in stimulating receptor autophosphorylation. These results are consistent with the relative potencies of this preparation, hIGF-II, and rIGF-I in stimulating the replication of 12-week-old fetal dermal fibroblasts.(ABSTRACT TRUNCATED AT 400 WORDS)

Five mutants of recombinant insulin-like growth factor-II (rIGF-II) that bound with high affinity to either the IGF-II/cation-independent mannose 6-phosphate (IGF-II/CIM6-P) or the IGF-I receptor were prepared by site-directed mutagenic procedures, expressed as fusion proteins in the larva of Bombyx mori or Escherichia coli, purified to homogeneity, renatured, and characterized in terms of their receptor binding affinities and specificities as well as their biological activities. Class I mutants in which Phe26, Tyr27, and Val43 were substituted with Ser, Leu, and Leu, respectively, bound to enriched preparations of rat placental IGF-II/CIM6-P receptors with apparent equilibrium dissociation constants (Kd(app)) that were only slightly greater, i.e. 0.10, 0.05, and 0.06 nM, than that of rIGF-II (0.04 nM) or hIGF-II (0.03 nM). In contrast, replacing Phe26 with Ser resulted in 5- and 20-fold decreases in the affinities of this mutant for highly purified human placental IGF-I and insulin receptors, respectively. The affinities of the two other Class I mutants, [Leu27]- and [Leu43]rIGF-IIs, for these two receptors were reduced 80- to 220-fold. The affinities of Class II mutants, i.e. [Thr48,Ser49,Ile50]- and [Arg54,Arg55] rIGF-IIs, for IGF-I receptors were as potent as rIGF-II; however, they bound very poorly or not at all to the IGF-II/CIM6-P receptor. In the binding study of those mutant rIGF-IIs, IGF-II was observed to have an unexpectedly high affinity for pure human placental insulin receptor preparations. For example, the affinities of hIGF-II, rIGF-II, and two Class II rIGF-II mutants for the insulin receptor were only 3-, 9-, and 5-fold less, respectively, than that of porcine insulin. In two biological assay systems, i.e. the stimulation of DNA synthesis in Balb/c 3T3 cells and glycogen synthesis in HepG2 cells, the Kd(app) of the rIGF-II mutants for the IGF-I receptor but not the IGF-II/CIM6-P receptor correlated with their abilities to produce biological responses.

A radioimmunoassay for the human insulin-like growth factor-I (IGF-I) receptor was developed using a rabbit polyclonal antibody to the human IGF-I receptor and a highly purified IGF-I receptor. The purified receptor was radiolabeled with 125I-Bolton-Hunter reagent. Over 18% of the radiolabeled receptor was immunoprecipitated with the polyclonal antireceptor antibody. Purified IGF-I receptor concentrations as low as 5 ng/0.5 mL inhibited the radiolabeled IGF-I receptor binding. Purified insulin receptor weakly inhibited this binding, while the ligand IGF-I did not show inhibition. The radioimmunoassay was applicable to the measurements of IGF-I receptors in the Triton X-100 extracts of various tissues and cells. Breast cancer tissues and cells showed detectable IGF-I receptors, which correlated with IGF-I ligand binding. Receptor content was measurable in placenta and IM-9 cells, but receptor content was not measurable in liver and muscle extracts.

Human placental insulin receptor contains 47 Cys per an alpha beta dimer. Most of the 94 Cys in an intact alpha 2 beta 2 receptor are expected to form interchain or intrachain disulfide bonds, since there appears to be only one free cysteine residue in each beta subunit. In order to gain more insight into the three-dimensional organization of the insulin receptor, we have used limited trypsin digestion, SDS-PAGE, and protein microsequencing. The present study revealed the following; major tryptic cleavages occurred at alpha 164, alpha 270, alpha 582, and beta 1115, generating Mr 175,000, 130,000, 100,000, 70,000, and 55,000 disulfide-linked complexes. Under reducing conditions, tryptic fragments of Mr values = 30,000, 70,000, 20,000, 55,000, and 20,000 were identified to be alpha(1-164), alpha(165-582), alpha(165-270), alpha(271-582), and alpha(583-C-terminal), respectively. The major beta subunit tryptic fragment of Mr = 55,000 was assumed to have beta(724-1115) or beta(N-terminal-392). The Mr 175,000 complex appeared to contain two alpha(1-164) and two alpha(165-582), whereas the Mr 70,000 complex contained alpha(583-C-terminal) and beta(724-1115). Tryptic cleavage at alpha 582 apparently produced one Mr 175,000 and two Mr 70,000 complexes, suggesting that the alpha(583-C-terminal) domain interacts with the extracellular domain of the beta subunit by disulfide bonds. Tryptic cleavage at alpha 270 resulting in a formation of one Mr 100,000 complex consisting of two alpha(1-270) and two Mr 130,000 complexes consisting of alpha(271-C-terminal) and beta(724-1115) suggests that Cys residues involved with disulfide bonds between the two alpha subunits are located in the alpha(1-270) domain. The identification of the Mr 55,000 complex consisting of small tryptic fragments between alpha(122-270) indicates that 40 Cys residues in the two alpha(122-270) domains are inter- and intramolecularly associated by disulfide bonds. The alpha(1-121) domain does not appear to be linked to any other domains by disulfide bonds. These results are consistent with the structural model that the N-terminal domains of alpha subunits (122-270) are disulfide-linked together while the C-terminal domain (583-C-terminal) of the alpha subunit is linked to the N-terminal domain of the beta subunit by disulfide bonds.

To gain insight into the regulation of low Km cAMP phosphodiesterases (PDE) by insulin in human tissues, PDEs in human placenta were studied. Human placenta contained cAMP PDEs in particulate and cytosolic fractions. More than 99% of the total activity was localized in the cytosolic fraction. The cytosolic fraction exhibited at least four cyclic nucleotide PDEs when fractionated by DEAE-cellulose chromatography. The first form was a calmodulin-activated PDE which hydrolyzed both cGMP and cAMP. The second form was a high affinity cAMP PDE with a nonlinear kinetic characteristic, but was not inhibited by either cGMP or cilostamide (either compound is known to specifically inhibit rat insulin-sensitive cAMP PDE). The third form was a low Km cAMP PDE, but was only modestly sensitive to inhibition by cGMP or cilostamide. The fourth form was a cAMP PDE which showed high sensitivity to inhibition by cGMP or cilostamide. The IC50 values of the fourth form were comparable to those of rat adipose insulin-sensitive PDE. However, its Km for cAMP was 2 microM, which is about 10 times higher than that of the rat enzyme. Insulin treatment on placenta tissues stimulated at least two PDEs, the third and fourth forms. To our knowledge, this is the first report to describe insulin-sensitive cAMP PDEs in the cytosolic fraction of human placenta.

Phage-display technology is probably the best available strategy to produce antibodies directed against various carbohydrate moieties since conventional hybridoma technologies have yielded mostly low-affinity antibodies against a limited number of carbohydrate antigens. Because of difficulties in immobilization of carbohydrate antigens onto plastic plates, however, the same procedures used for protein antigens cannot be readily applied. We adapted phage-display technology to generate human single chain antibodies (scFvs) using neoglycolipids as antigens. This study describes the isolation and characterization of phage-displayed antibodies (phage Abs) that recognized nonreducing terminal mannose residues. We first constructed a phage Ab library with a large repertoire using CDR shuffling and VL/VH shuffling methods with unique vector constructs. The library was subjected to four rounds of panning against neoglycolipids synthesized from mannotriose (Man3) and dipalmitoylphosphatidylethanolamine (DPPE) by reductive amination. Of 672 clones screened by enzyme-linked immunosorbent assay (ELISA) using Man3-DPPE as an antigen, 25 positive clones encoding scFvs with unique amino acid sequences were isolated as candidates for phage Abs against Man3 residues. TLC-overlay assays and surface plasmon resonance analyses revealed that selected phage Abs bound to neoglycolipids bearing mannose residues at nonreducing termini. In addition, binding of the phage Ab to RNase B carrying high mannose type oligosaccharides but not to fetuin carrying complex type and O-linked oligosaccharides was confirmed. Furthermore, first round characterization of scFvs expressed from respective phages indicated good affinity and specificity for nonreducing terminal mannose residues. These results demonstrated the usefulness of this strategy in constructing human scFv against various carbohydrate antigens. Further studies on the purification and characterization of these scFvs are presented in an accompanying paper in this issue.

Immunological memory induced by nasal immunization with adjuvant-combined influenza vaccine was analyzed in different ages and strains of mice. The memory activities were assessed by secondary nasal-wash IgA and serum IgG antibody (Ab) responses and protection against challenge infection with a lethal dose of influenza virus. Mice were primed with 0.1 microg of vaccine and boosted with 0.1 or 1.0 microg vaccine 1 (short-term memory)- or 17 (long-term memory)-months later. Influenza-specific short-term memory responses in young adult BALB/c mice (2-month-old) were significantly higher than those of long-term memory activities in mice boosted at 19 months of age. However, those influenza-specific long-term memory responses provided protective immunity against influenza virus challenge and were higher than short-term memory in aged mice primed at 18-month-old and boosted 1 month later. These results show that the age at which initial nasal immunization is given is critically important in order to induce protective immunity in aged mice. Similar findings were noted in the C3H mouse strain; however, C57BL/6 mice failed to induce influenza-specific immune responses in both young adult and aged mice. These results indicate that low doses of cholera toxin B subunit (supplemented with 0.2% of hole toxin) combined nasal vaccine may required further improvement in order to provide protective immunity in human use.

RNA- or antibody-based strategies may provide novel strategies for inhibiting the expression of specific gene products that may be involved in prostate cancer cell growth and progression. Prostate cancer cells could escape hormonal control by constitutively expressing growth factors such as insulin-like growth factors (IGFs). This proposal is to test the hypothesis that IGF-II/IGF receptor signaling plays an important role in prostate cancer growth and progression. To prove the hypothesis, our newly constructed IGF-II ribozymes and single-chain antibodies against the IGF-I receptor (alpha IGFIR scFvs) are being used. Specific Aim 1 is to test our hypothesis using already available PC-3 prostate cancer cell transfectants and control cells. Effects of intracellular expression of IGF-II ribozyme or alpha IGFIR scFv on cell growth, anticancer-drug induced apoptosis, and tumorigenesis will be investigated in cell culture and in athymic mice. Specific Aim 2 is to take alternative approaches towards testing our hypothesis and developing novel gene and/or immunotherapy approaches for prostate cancer treatment. RNA- or antibody-based strategies targeting growth factors in prostate cancer have never been evaluated. The proposed studies should thus lay the foundation for the development of novel strategies for prostate cancer treatment.

Insulin-like growth factors (IGF) I and II are potent mitogens for a variety of cancer cells. The proliferative and anti-apoptotic actions of IGF are mediated by the IGF-I receptor (IGF-IR), to which both IGF-I and IGF-II bind with high affinity. To investigate the mitogenic and anti-apoptotic activities of IGF-IR and to achieve better inhibition of IGF-IR function, single-chain antibodies against human IGF-IR (alphaIGF-IR scFvs) were constructed and expressed. IgG cDNA encoding variable regions of light and heavy chains (VL and VH) from mouse IgG were cloned from a hybridoma producing the 1H7 alphaIGF-IR monoclonal antibody [Li et al., Biochem Biophys Res Commun 196: 92-98 (1993)]. The splice-overlap extension polymerase chain reaction was used to assemble a gene encoding the alphaIGF-IR scFv, including the N-terminal signal peptide, VL, linker peptide, VH, and C-terminal DYKD tag. Two types of soluble alphaIGF-IR scFvs, a prototype alphaIGF-IR scFv and its alternative type alphaIGF-IR scFv-Fc, were constructed and expressed in murine myeloma cells. alphaIGF-IR scFv-Fc, containing the human IgG1 Fc domain, was stably expressed in NS0 myeloma cells, using a glutamine synthase selection system, and purified from the conditioned medium of stable clones by protein-A--agarose chromatography. Levels of alphaIGF-IR scFv-Fc expression ranged from 40 mg/l to 100 mg/l conditioned medium. Sodium dodecyl sulfate/polyacrylamide gel electrophoresis analysis under reducing and nonreducing conditions indicated that alphaIGF-IR scFv-Fc is a dimeric antibody. alphaIGF-IR scFv-Fc retained general characteristics of the parental 1H7 monoclonal antibody except that its binding affinity for IGF-IR was estimated to be approximately 10(8) M(-1), which was one-order of magnitude lower than that of 1H7 monoclonal antibody. Injection of alphaIGF-IR scFv-Fc (500 microg/mouse, twice a week) significantly suppressed MCF-7 tumor growth in athymic mice. These results suggest that the alphaIGF-IR scFv-Fc is a first-generation recombinant alphaIGF-IR for the potential development of future alphaIGF-IR therapeutics.

The release of insulin-like growth factor binding proteins (IGFBPs) and their regulation in human glomerular endothelial cells (GENC) was characterised. GENC produce IGFBP-4, IGFBP-2 and IGFBP-3 and express mRNA for IGFBP-2 to IGFBP-5. Due to the fact that IGF-I and TGF-beta1 modulate glomerular hypertrophy, their action on IGFBP release and GENC growth was studied. IGF-I increased IGFBP-3, IGFBP-2 and decreased IGFBP-4, while TGF-beta1 decreased IGFBP-3 and apparently increased IGFBP-4. All of the IGFBPs, except the TGF-beta1-regulated IGFBP-4, were modulated at mRNA level. IGF-I stimulated GENC proliferation, while TGF-beta1 inhibited their growth. It was demonstrated that an IGFBP-3 antibody reduced GENC proliferation. However, rhIGFBP-3 alone had no effect on GENC, but after 48 h pre-incubation the IGF-I stimulated GENC growth was increased, suggesting that IGFBP-3 could modulate the IGF-I induced GENC proliferation. It was concluded that the stimulatory IGFBP-3 and the inhibitory IGFBP-4 could regulate GENC growth, although the IGFBP-3 seems to have a predominant effect in this control.

Protein microsequencing of human placental IGF-I receptors purified by immunoaffinity chromatography using IGF-I receptor specific monoclonal antibody revealed amino acid sequences of both IGF-I and insulin receptors. Since this finding indicated the presence of IGF-I/insulin receptor hybrids, hybrid receptors were further purified by immunoaffinity chromatography using insulin receptor specific monoclonal antibody. The molecular size of the nonreduced hybrid receptor was approximately 350K, indicating that the IGF-I and insulin receptor alpha beta halves were disulfide-linked. The ratio of IGF/insulin binding activity of purified hybrid receptors was approximately 25 when measured using tracer amounts of radioactive ligands. 125I-IGF binding to these receptors was inhibited by IGF-I and insulin with IC50s of approximately 2 and approximately 1000 nM, respectively. 125I-Insulin binding to these receptors was similarly inhibited by IGF-I and insulin with IC50 of approximately 3 nM. Autophosphorylation and kinase activities of the hybrid receptor were stimulated by IGF-I more effectively than insulin in a dose-dependent manner. Thus, the present studies indicate that hybrid receptors purified from human placenta have the functional properties of an IGF-I receptor.

Peptide hormone/growth factor receptors are membrane glycoproteins. The receptor specifically binds its ligand, which approaches the cell from the outside, transduces the signal to the inside, and, consequently initiates biological programs specific to the receptor-ligand complex. In the case of insulin and IGF-I, the ligands not only share amino acid sequence homology (Rinderknecht and Humbel, 1978; Blundell et ai., 1978), but also the receptors are structurally and functionally similar (Jacobs et ai., 1983; Rechler and Nissley, 1985; Fujita-Yamaguchi et al., 1986). Recent progress made in purification and subsequent cloning of both receptors revealed ~60% identity in their overall amino acid sequence (Ullrich et al., 1985; Ebina et ai., 1985; Ullrich et ai., 1986). The receptors for insulin and IGF-I consist of α and β subunits, linked in a β-α-α-β form by disulfide bonds. The α subunit of Mr = ~125,000 is extracellular and responsible for ligand binding, whereas the β subunit of Mr = ~90,000 is a transmembrane protein carrying a tyrosine-specific protein kinase in its cytoplasmic domain. When the specific ligand binds to the extracellular domain, the β subunit kinase is activated, and consequently, phosphorylates itself (autophosphorylation) as well as intracellular components (Rees-Jones and Taylor, 1985; White et al., 1985; Bernier et al., 1987).

Since the studies on tyrosine phosphorylation of calmodulin by the insulin receptor kinase in vitro suggested that protamine and poly(L-lysine) may activate phosphorylation of the receptor beta subunit [Sacks & McDonald (1988) J. Biol. Chem. 263, 2377-2383], we examined the effects of a variety of basic polycations/proteins and polyamines on insulin receptor kinase activity. The insulin receptor purified from human placental membranes was incubated with each basic polycation/protein or polyamine and assayed for tyrosine-specific protein kinase activity by measuring 32P incorporation into the src-related peptide. At a concentration of 1 microM, poly(L-lysine) and poly(L-ornithine) markedly stimulated kinase activity, whereas poly(L-arginine) and histones H1 and H2B inhibited insulin receptor kinase. In contrast, at a concentration of 1 mM, three polyamines (spermine, spermidine and putrescine) did not alter kinase activity. Poly(L-lysine) and poly(L-ornithine) stimulated the insulin receptor kinase by 5-10-fold at concentrations of 0.1-1 microM. Protamine sulphate also showed a significant stimulatory effect at a concentration of 100 microM. Preincubation of the receptor with poly(L-lysine) or poly(L-ornithine) for 20-60 min resulted in maximal kinase activation. Poly(L-lysine), the most effective activator of the receptor kinase, was used to characterize further the mechanisms of the kinase activation. Poly(L-lysine) activates the insulin receptor kinase by increasing the Vmax. without changing the Km. Poly(L-lysine) markedly stimulates the kinase activity of insulin receptor preparations that have lost both basal kinase activity and the ability to be stimulated by insulin. Insulin and poly(L-lysine) also differed in their ability to stimulate the kinase activity of prephosphorylated receptors. Prephosphorylation of the receptors did not affect the stimulation of the kinase by insulin. In contrast, prephosphorylation of receptors resulted in a markedly enhanced ability of poly(L-lysine) to stimulate kinase activity. These studies suggest that the mechanisms by which poly(L-lysine) and insulin activate the kinase are different. In conjunction with other additional evidence, it is suggested that poly(L-lysine) interacts directly with the beta-subunit of the receptor, thereby activating the receptor kinase.

The receptors for insulin and insulin-like growth factor (IGF)-I carry intrinsic tyrosine-specific protein kinases (TPK) in their cytoplasmic domains that show 84% homology. Our previous studies using tyrosine-containing synthetic polymers (Arch. Biochem. Biophys. 260, 416, 1988) revealed subtle differences between the two receptor TPKs. In the present study, low molecular weight kinase inhibitors were used to compare the two receptor TPKs purified from human placenta. Staurosporine was the most potent inhibitor of both receptor TPKs among the three inhibitors tested. It was 100 times more inhibitory to insulin receptor TPK (ED50 = 61nM) than IGF-I receptor TPK (ED50 = 6.2 microM). Apigenin and kaempferol showed approximately the same inhibitory potency toward both TPKs with a range of 10 approximately 1000 microM. Staurosporine is thus an excellent tool to biochemically characterize the two receptor TPKs as well as to selectively inhibit insulin-activated TPK in intact cells.

A newly developed immunoprecipitation assay, with 125I-labeled highly purified human placental insulin-like growth factor I (IGF-I) receptor, was used to search for IGF-I-receptor antibodies in human sera. Eleven of 141 patient sera tested (7.8%) immunoprecipitated labeled IGF-I receptor. Immunoprecipitation was comparable with sera and IgG prepared from these sera. Seven of the 11 sera (3 of 31 with rheumatic disorders, 3 of 48 with non-insulin-dependent diabetes, and 1 of 52 with insulin-dependent diabetes) failed to inhibit IGF-I binding to human placental membranes and thus contained non-binding-inhibitory IGF-I-receptor antibodies. Their pathophysiological function remained uncertain. The remaining 4 sera (2 of 3 with type B severe insulin resistance, 1 of 7 with polycystic ovary syndrome (PCO), and 1 of 31 with rheumatic disorders) inhibited IGF-I binding. Plasma IGF-I concentrations were elevated (663 and 802 ng/ml, respectively) in 2 patients (1 with PCO and another with systemic lupus erythematosus) with binding-inhibitory IGF-I-receptor antibodies, suggesting IGF-I resistance that was probably mediated by the IGF-I-receptor antibodies. In conclusion, we identified two species of human IGF-I-receptor antibodies. Sera from 7 of 141 patients tested contained IgG autoantibodies that bound to the IGF-I receptor at a locus different from the IGF-I binding site and did not inhibit IGF-I binding. Sera from 4 of 141 patients contained antibodies that bound to the IGF-I receptor at or near the IGF-I binding site, inhibited IGF-I binding, and probably caused IGF-I resistance.

Summary Insulin was found to stimulate the phosphorylation of the 21,000-dalton protein encoded by the ras oncogene of Harvey murine sarcoma virus in membrane fraction both in vivo and in vitro . When the human ras proteins expressed in E. coli were reconstituted with purified human insulin receptor, GTPase activity of normal or its mutated oncogenic ras protein was not stimulated by the addition of insulin. Likewise, tyrosine kinase activity or insulin binding capaeity of the receptor was not influenced when assayed in the presence of the ras proteins. These results suggest that ras proteins may be coupled with the insulin receptor system through some unidentified membrane factors.

Various lipids were tested as substrates for the insulin receptor kinase using either receptor partially purified from rat hepatoma cells by wheat-germ-agglutinin-Sepharose chromatography or receptor purified from human placenta by insulin-Sepharose affinity chromatography. Phosphatidylinositol was phosphorylated to phosphatidylinositol 4-phosphate by the partially purified insulin receptor. In contrast, phosphatidylinositol 4-phosphate and diacylglycerol were not phosphorylated. In some, but not all preparations of partially purified insulin receptor, the phosphatidylinositol kinase activity was stimulated by insulin (mean effect 33%). Phosphatidylinositol kinase activity was retained in insulin receptor purified to homogeneity. Insulin regulation of the phosphatidylinositol kinase was lost in the purified receptor; however, dithiothreitol stimulated both autophosphorylation of the purified receptor and phosphatidylinositol kinase activity in parallel about threefold. (Glu80Tyr20)n, a polymeric substrate specific to tyrosine kinases, inhibited the phosphatidylinositol kinase activity of the purified receptor by greater than 90% and inhibited receptor autophosphorylation by 67%. Immunoprecipitation by specific anti-receptor antibodies depleted by greater than 90% the phosphatidylinositol kinase activity in the supernatant of the purified receptor and the phosphatidylinositol kinase activity was recovered in the precipitate in parallel with receptor autophosphorylation activity. These characteristics of the phosphatidylinositol kinase activity of the purified insulin receptor and its metal ion preference paralleled those of the receptor tyrosine kinase activity and differed from bulk phosphatidylinositol kinase activity in cell extracts, which was not significantly inhibited by (Glu80Tyr20)n, stimulated by dithiothreitol or depleted by immunoprecipitation with anti-(insulin receptor) antibody. These results suggest that the insulin receptor is associated with a phosphatidylinositol kinase activity; however, this activity is not well regulated by insulin. This kinase appears to be distinct from the major phosphatidylinositol kinase(s) of cells. Its relationship to insulin action needs further study.

Insulin-like growth factor (IGF) I receptor was purified from Triton X-100-solubilized human placental membranes by wheat germ agglutinin-Sepharose chromatography followed by immunoaffinity chromatography using alpha IR-3, a monoclonal antibody directed against the IGF-I receptor. Purification of 3200-fold and 2800-fold was achieved from wheat germ agglutinin-Sepharose eluates with regard to IGF-I binding and kinase activities. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis under reducing conditions revealed two major protein bands corresponding to the alpha and beta subunits of the receptor, which accounted for at least 90% of the protein content. The purified receptor bound 10-20 micrograms of IGF-I/mg of protein and was more than 95% free of contamination by insulin receptor. It sedimented in glycerol gradients as a single species with a sedimentation coefficient of 13.7 S and gave three protein bands with Mr = approximately 300,000 on sodium dodecyl sulfate-polyacrylamide gel electrophoresis under nonreducing conditions, indicating that alpha 2 beta 2 is an intact form of the IGF-I receptor. The purified receptor, when incubated with [gamma-32P] ATP, became phosphorylated at tyrosine residues of its beta subunit. This was stimulated 3-fold by IGF-I. It also had IGF-I-stimulated tyrosine kinase activity (5264 pmol of 32P incorporated/min/mg of protein) toward a synthetic peptide corresponding to the autophosphorylation site of pp60src. These data strongly suggest that it is a tyrosine-specific protein kinase.

The relationship between the structure of the insulin receptor and its kinase activity was studied on the purified receptor treated with different concentrations of dithiothreitol. An enhanced autophosphorylation of the beta subunit (Mr, 90,000) was observed on NaDodSO4/PAGE under reducing conditions when the receptor was treated with 0.1-0.75 mM dithiothreitol in the presence of 1 microM insulin. Since we have previously observed (unpublished data) that incubation of the purified receptor with 1 mM dithiothreitol completely reduced an intact form of the receptor, alpha 2 beta 2, to free alpha subunit (Mr, 125,000) and beta subunit, the population of disulfide-linked complexes of the receptor after the dithiothreitol treatment was analyzed by NaDodSO4/PAGE under nonreducing conditions. The same receptor preparations were assayed for tyrosine kinase activity by using an exogenous substrate. Treatment of the receptor with dithiothreitol significantly enhanced both basal and insulin-dependent kinase activity. The kinase activity was enhanced 12- to 37-fold at concentrations of 0.5-0.75 mM dithiothreitol in the presence of 1 microM insulin. The amount of alpha 2 beta 2, alpha beta, and beta forms in each dithiothreitol-treated receptor preparation was quantified and compared with its kinase activity. These studies clearly indicate a correlation between the appearance of an alpha beta form and an increase in kinase activity. Therefore, we conclude that the alpha beta form of the insulin receptor exhibits much higher kinase activity than the intact receptor in the alpha 2 beta 2 form.

Added N alpha-p-tosyl-l-arginine methyl ester or N alpha-benzoyl-l-arginine ethyl ester inhibited the stimulation by insulin of phosphorylation of the 95,000 dalton subunit of the insulin receptor both in a partially purified insulin receptor fraction from rat adipocytes and in a highly purified insulin receptor preparation from human placenta. N-alpha-p-tosyl-l-lysine chloromethyl ketone, N alpha-p-tosyl-l-lysine methyl ester, or N-acetyl-l-phenylalanine ethyl ester were much less potent, while N-benzoyl-1-alanine methyl ester was without effect. Inhibition of the phosphorylation by the arginine analogues did not require preincubation of the insulin receptor with inhibitors in the presence of insulin prior to phosphorylation. Inhibition by N alpha-p-tosyl-l-arginine methyl ester was decreased by preincubation of the receptor fraction with cold ATP and MnCl2. These results suggest that N alpha-p-tosyl-l-arginine methyl ester inhibits an initial ATP and Mn2+ dependent reaction in insulin-stimulated phosphorylation process.

An antibody was raised against a synthetic peptide corresponding to the carboxyl-terminal amino acids of the human insulin receptor (Anti-R beta C). Immunoprecipitation of the human insulin receptor and immunoblotting to the beta-subunit by Anti-R beta C could be inhibited by competition with the corresponding peptide. However, even at saturating concentrations, anti-R beta C could not completely immunoprecipitate or immunodeplete insulin receptors compared to a human autoantibody (anti-R B2). Using receptor labeled directly by 125I, evidence of multiple forms of the beta-subunit was found. When the receptor could be immunoprecipitated by anti-R beta C, the beta-subunit migrated with an apparent mol wt (MW) of 96,000 (at or above the phosphorylase b MW marker). However, in preparations where anti-R beta C was not able to immunoprecipitate the insulin receptor, the beta-subunit migrated at a significantly lower MW of 91,000 (below phosphorylase b), as detected by immunoprecipitation with Anti-R B2. Intermediate forms could also be detected. Phosphorylation of partially purified insulin receptor did not affect is ability to be immunoprecipitated by anti-R beta C, although insulin-stimulated phosphorylation increased the apparent MW of the beta-subunit. However, insulin receptor that was phosphorylated in solubilized extracts of whole cells had a beta-subunit that migrated at lower MW and was not immunoprecipitated by anti-R beta C. One possible explanation for this is that the beta-subunit may be degraded during preparation. When the MW of insulin receptor that has been purified to homogeneity from human placenta is compared to our data, it is clear that many of these insulin receptor preparations contain lower MW beta-subunits. These results must be taken into account when the sites of phosphorylation and kinase activity of purified insulin receptor preparations are studied.

Since we observed that after purification the receptor kinase activity is rapidly lost under conditions where insulin binding function seems to be preserved, we have studied the cause(s) of receptor kinase inactivation. Highly purified placental insulin receptor preparations were analyzed by NaDodSO4/PAGE followed by silver staining or immunostaining using domain-specific antibodies raised against synthetic peptides corresponding to the amino acid sequences of the beta subunit. These studies revealed the intact 90-kDa beta subunit is degraded first to an 88-kDa form and then to a 50-kDa beta 1-subunit form by proteolysis even after purification when stored at 4 degrees C. The 88-kDa beta subunit, which lacks the carboxyl-terminal approximately equal to 2-kDa portion exhibits almost no autophosphorylation activity, nor does insulin stimulate autophosphorylation. The loss of kinase activity as measured by phosphorylation of the src-related peptide is correlated with the loss of the intact 90-kDa beta subunit. Degradation of the beta subunit to the 50-kDa form seems to be facilitated by the removal of the approximately equal to 2-kDa peptide. Present studies thus suggest that only the intact form of the beta subunit has full kinase activity in an insulin-dependent manner and that other forms, such as the 88-kDa beta subunit show little kinase activity. The inactivation appears to arise from a conformational change of the 90-kDa form, which makes it susceptible to proteolysis at the carboxyl-terminal end. These results imply that the carboxyl-terminal of the beta subunit is important for the manifestation of the tyrosine kinase activity of the insulin receptor.

The receptors for insulin and insulin-like growth factor (IGF) I are structurally similar transmembrane proteins. Ligand binding to the extracellular domain of the receptor stimulates its cytoplasmic tyrosine protein kinase which phosphorylates its own beta subunit as well as exogenous substrates. It is believed, from several lines of evidence, that tyrosine-specific protein kinases are mediating some or all of the actions of insulin (or IGF-I). In order to gain insights into the substrate specificity of the structurally related insulin and IGF-I receptor kinases, we have studied the action of highly purified receptors isolated from human placental membranes. Present studies using selected tyrosine-containing polymers have revealed: (i) Polymers such as (Y,A,E)n and (Y-A-E)n inhibit beta subunit autophosphorylation and exogenous substrate phosphorylation by autophosphorylated receptors. (ii) Insulin receptor kinase is at least 10 times more sensitive to these inhibitors than IGF-I receptor kinase. (iii) (Y-A-E)n is approximately 8 times more potent an inhibitor than (Y,A,E)n toward both receptors. (iv) While (E4,Y1)n and (E6,A3,Y1)n are good substrates for both receptor kinases, the ratio of phosphate incorporation into the former to the latter is characteristically high (approximately 4) for the IGF-I receptor and low (approximately 1) for the insulin receptor. These results imply that the substrate specificity and enzymatic action of these two receptor kinases are distinct.

We have investigated the in vitro tyrosine phosphorylation of the HRAS and KRAS proteins by human placental insulin receptor kinase. Purified HRAS proteins are not phosphorylated by purified insulin receptor kinase. Since the tyrosine phosphorylation of calmodulin by the insulin receptor kinase in vitro requires cofactors such as protamine and poly(L-lysine), we examined the possibility that poly(L-lysine) may also potentiate the interaction between RAS proteins and the insulin receptor. We found that purified HRAS proteins are indeed phosphorylated by purified insulin receptor kinase in the presence of poly(L-lysine). In contrast, the KRAS protein, which carries an extremely basic domain (residues 172-182, Lys-Asp-Glu-Lys6-Ser-Arg), is phosphorylated by the receptor kinase without the addition of basic proteins. We then determined whether the KRAS basic domain peptide plays a role similar to that of poly(L-lysine) and found that both the HRAS protein and calmodulin are phosphorylated by the receptor kinase in the presence of the KRAS basic domain peptide. Further examination of the role of poly(L-lysine) in potentiating tyrosine phosphorylation of the HRAS protein and calmodulin by purified insulin receptor kinase indicates that poly(L-lysine) affects the conformation of these protein substrates as well as that of the receptor kinase domain. These studies suggest that polylysine-like basic proteins or domains are required to establish the interaction between insulin receptor kinase and its substrate.