The type 1 insulin-like growth factor receptor (IGF-1R) is often overexpressed on tumor cells and is believed to play an important role in anchorage-independent proliferation. Additionally, cell culture studies have indicated that IGF-1R confers increased resistance to apoptosis caused by radiation or chemotherapeutic agents. Thus, inhibitors of the intracellular kinase domain of this receptor may have utility for the clinical treatment of cancer. As part of an effort to develop clinically useful inhibitors of IGF-1R kinase, a novel class of pyrrole-5-carboxaldehyde compounds was investigated. The compounds exhibited selectivity against the closely related insulin receptor kinase intrinsically and in cell-based assays. The inhibitors formed a reversible, covalent adduct at the kinase active site, and treatment of such adducts with sodium borohydride irreversibly inactivated the enzyme. Analysis of a tryptic digest of a covalently modified IGF-1R kinase fragment revealed that the active site Lys1003 had been reductively alkylated with the aldehyde inhibitor. Reductive alkylation of the insulin receptor kinase with one of these inhibitors led to a similarly inactivated enzyme which was examined by X-ray crystallography. The crystal structure confirmed the modification of the active site lysine side chain and revealed details of the key interactions between the inhibitor and enzyme.

The x-ray structure of the unactivated kinase domain of insulin-like growth factor-1 receptor (IGFRK-0P) is reported here at 2.7 Angstrom resolution. IGFRK-0P is composed of two lobes connected by a hinge region. The N-terminal lobe of the kinase is a twisted beta-sheet flanked by a single helix, and the C-terminal lobe comprises eight alpha-helices and four short beta-strands. The ATP binding pocket and the catalytic center reside at the interface of the two lobes. Despite the overall similarity to other receptor tyrosine kinases, three notable conformational modifications are observed: 1) this kinase adopts a more closed structure, with its two lobes rotated further toward each other; 2) the conformation of the proximal end of the activation loop (residues 1121-1129) is different; 3) the orientation of the nucleotide-binding loop is altered. Collectively, these alterations lead to a different ATP-binding pocket that might impact on inhibitor designs for IGFRK-0P. Two molecules of IGFRK-0P are seen in the asymmetric unit; they are associated as a dimer with their ATP binding clefts facing each other. The ordered N terminus of one monomer approaches the active site of the other, suggesting that the juxtamembrane region of one molecule could come into close proximity to the active site of the other.

Activation of phosphoinositide 3'-hydroxykinase (P13K) is required for mitogenic signal transduction by several growth factors and oncogenes. P13K is a heterodimer consisting of a p85 regulatory subunit and a p110 catalytic subunit. In the current study, we report the cloning and characterization of the p110 alpha catalytic subunit of human P13K. This clone is highly homologous (> 99% amino acid identity) to bovine brain p110 alpha, but contains 10 amino acid differences from the human p110 alpha sequence previously reported. Comparison of this sequence with known Ser/Thr kinases and p110 homologs highlighted several conserved residues within the putative kinase domain. Mutational analysis of these residues (Asp915, (Asp933 + Phe934)) yielded P13K mutants with virtually complete loss of phosphoinositide phosphorylating activity. Expression of the wild-type p110 alpha protein in CHO cells is sufficient to activate the serum response element derived from the promoter of c-fos, an immediate early gene product. In contrast, the catalytically impaired p110 alpha mutants as well as the p85 alpha subunit of P13K were inactive in the fos assay. These studies suggest that the mitogenic signal transduction pathway mediated by P13K is dependent upon the enzymatic activity of the p110 alpha subunit of P13K.

Phosphorylation of phosphoinositides by phosphoinositide 3'-hydroxykinase (PI3K) occurs at a lipid/water interface. We have determined that highly purified recombinant human P13K binds tightly to vesicle interfaces composed primarily of phosphatidylinositol (PI) or 1,2-dimyristoyl-sn-glycero-3-phosphomethanol (DMPM). The rate of desorption of PI3K from the vesicle interface is slow and does not significantly affect the observed product formation kinetics. Observations which demonstrate that PI3K is tightly bound to the vesicle lipid/water interface include the following: (1) product formation plateaus rapidly, even in the presence of active enzyme and excess substrate; (2) total product formation is proportional to the amount of PI3K; (3) initial product formation rates are unaffected by bulk lipid concentration but are dependent on the interfacial substrate concentration; and (4) PI3K partitions with lipid vesicles in sedimentation gradients. This enzymatic profile has been referred to as catalysis in the "scooting" mode (Berg et al., 1991). A kinetic analysis of PI3K catalysis in the scooting mode is presented. The interfacial Km,app for PI was determined to be approximately 6.0 mol % in PI/DMPM vesicles. The ratio of specificity constants (kcat/Km) for PI, phosphatidylinositol 4-monophosphate (PIP), and phosphatidylinositol 4,5-diphosphate (PIP2) utilization was determined to be near unity. These results provide a rigorous enzymological framework for the kinetic analysis of PI3K inhibitors.

The human papillomavirus E7 gene can transform murine fibroblasts and cooperate with other viral oncogenes in transforming primary cell cultures. One biochemical property associated with the E7 protein is binding to the retinoblastoma tumor suppressor gene product (pRB). Biochemical properties associated with pRB include binding to viral transforming proteins (E1A, large T, and E7), binding to cellular proteins (E2F and Myc), and binding to DNA. The mechanism by which E7 stimulates cell growth is uncertain. However, E7 binding to pRB inhibits binding of cellular proteins to pRB and appears to block the growth-suppressive activity of pRB. We have found that E7 also inhibits binding of pRB to DNA. A 60-kDa version of pRB (pRB60) produced in reticulocyte translation reactions or in bacteria bound quantitatively to DNA-cellulose. Recombinant E7 protein used at a 1:1 or 10:1 molar ratio with pRB60 blocked 50 or greater than 95% of pRB60 DNA-binding activity, respectively. A mutant E7 protein (E7-Ala-24) with reduced pRB60-binding activity exhibited a parallel reduction in its blocking of pRB60 binding to DNA. An E7(20-29) peptide that blocks binding of E7 protein to pRB60 restored the DNA-binding activity of pRB60 in the presence of E7. Peptide E7(2-32) did not block pRB60 binding to DNA, while peptide E7(20-57) and an E7 fragment containing residues 1 to 60 partially blocked DNA binding. E7 species containing residues 3 to 75 were fully effective at blocking pRB60 binding to DNA. These studies indicate that E7 protein specifically blocks pRB60 binding to DNA and suggest that the E7 region responsible for this property lies between residues 32 and 75. The functional significance of these observations is unclear. However, we have found that a point mutation in pRB60 that impairs DNA-binding activity also blocks the ability of pRB60 to inhibit cell growth. This correlation suggests that the DNA-binding activity of retinoblastoma proteins contributes to their biological properties.

The retinoblastoma gene product (pRB) participates in regulating mammalian cell replication. The mechanism responsible for pRB's growth regulatory activity is uncertain. However, pRB is known to bind viral transforming proteins including the papilloma virus E7 protein, cellular proteins, and DNA. pRB contains a critical domain termed the "binding pocket" which is required for binding activities. This binding pocket contains 8 cysteine residues. A naturally occurring mutation affecting one of these cysteines is known to eliminate pRB's protein and DNA binding activities. To investigate the cysteine residues in pRB's binding pocket, each residue was mutated to alanine, phenylalanine, or serine. These mutant genes were used to prepare pRBs harboring specific amino acid substitutions. Individual mutations at positions 407, 553, 666, and 706 depressed pRB binding to E7 protein, DNA, and a conformation-specific anti-pRB antibody, XZ133. Combinations of these inhibitory mutations exhibited additive inhibitory effects on pRB's binding properties. Mutations at positions 438, 489, 590, 712, and 853 did not affect pRB binding to E7 protein, DNA, or the XZ133 antibody. Combination of these five neutral mutations yielded a pRB species with full E7 protein, DNA, and XZ133 binding activities. These studies indicate that the cysteine residues at positions 407, 553, 666, and 706 contribute to the E7 protein and DNA binding properties of pRB and appear to do so by maintaining pRB's normal conformation.

Growth factor-toxin fusion proteins may also be used to regulate immune responses by eliminating subsets of lymphocytes or mononuclear cells that express specific cytokine receptors. A variety of genetic constructions employing several types of microbial-expression plasmids have been used successfully to produce growth factor-toxin fusion proteins. In particular, the optimal transcriptional promoters, ribosome-binding sites, and distances between the promoter and translational start site for each genetic construct are often different for different growth factor-toxin fusion proteins. The major obstacle to bringing several growth factor-toxin fusion proteins to clinical trials is the production and purification of sufficient quantities of biologically active protein. One of the most critical requirements for successful large-scale isolation of a recombinant fusion protein from bacteria is proper formation of disulfide bonds. An appropriate refolding step is generally essential for recovery of biologically active foreign proteins from bacteria.

Transforming growth factor alpha (TGF alpha)-Pseudomonas exotoxin 40 (PE40) is a chimeric protein consisting of an N-terminal TGF alpha domain fused to a C-terminal 40-kDa segment of the Pseudomonas exotoxin A protein. TGF alpha-PE40 exhibits the receptor-binding activity of TGF alpha and the cell-killing activity of PE40. These properties make TGF alpha-PE40 an effective cytotoxic agent for cells that possess epidermal growth factor receptors (EGFR). However, the utility of this protein as an anticancer agent has been unclear because many normal tissues express EGFR and may be damaged by exposure to TGF alpha-PE40. To address this issue, we injected nude mice with a lethal inoculum of either A431 or HT29 human tumor cells that possess EGFR or with Chinese hamster ovary (CHO) tumor cells that lack EGFR. Animals were treated with a derivative of TGF alpha-PE40 in which the cysteine residues are replaced by alanine, termed "TGF alpha-PE40 delta cys," or with saline once a day for 5 days. Mice bearing EGFR+ tumor cells lived significantly (P less than 0.001) longer when treated with TGF alpha-PE40 delta cys compared with saline-treated controls (median survival: A431 cells, 51.5 vs. 25.5 days; HT29 cells, 101 vs. 47.5 days). TGF alpha-PE40 delta cys did not prolong the survival of mice bearing tumor cells that lack EGFR (median survival: CHO cells, 15.5 vs. 19.5 days). The only toxicity to normal tissues was mild periportal hepatic necrosis. These studies indicate that a therapeutic window exists in vivo for the use of some growth factor-toxin fusion proteins as anticancer agents.

The chemical, physical, and biological properties of left-handed DNA have been studied using polymers, restriction fragments, and recombinant plasmids with established sequences. Our goal in the past several years has been to determine the existence of left-handed DNA in biological systems as part of our evaluation of the involvement of DNA microheterogeneity in gene regulation. General aspects of the involvement of DNA structure in gene expression have been reviewed recently (Wells et al. 1980b).

Studies on DNA polymers, restriction fragments, and recombinant plasmids have revealed the following: A) A family of left-handed DNA conformations exists for (dC-dG)n.(dC-dG)n. The observation of a particular conformation is dependent on the salt, the salt concentration and dehydrating agent. B) In sodium acetate solutions, (dC-dG)n.(dC-dG)n forms left-handed, psi(+)-condensed structures as detected by Raman spectroscopy and circular dichroism. C) (dT-dG)n.(dC-dA)n undergoes a right-to-left-handed transition only when reacted with AAF and at high salt concentrations. D) Transitions observed for polymer DNAs also are observed for restriction fragments containing both (dC-dG).(dC-dG) and (dT-dG).(dC-dA) sequences, but the transitions in the fragments generally require higher salt concentrations than observed for the polymers. E) Studies with recombinant plasmids containing (dC-dG) sequences from 10 to 58 bp in length demonstrate that left-handed Z-DNA segments can exist contiguous to B-DNA segments. F) Negative supercoil density (sigma less than or equal to -0.072) is sufficient to convert the (dC-dG) regions in those plasmids into left-handed structures under physiological ionic conditions (200 mM NaCl). G) The favorable free energy contribution of methylation in stabilizing the Z form in fragments and plasmids is approximately offset by the unfavorable free energy contributions of the B/Z junctions. H) Sl and BAL 31 nucleases recognize aberrant structural features at the confluence of the B and Z regions. I) Detailed mapping of Sl nuclease cleavage on supercoiled plasmids shows that the nuclease sensitive regions extend over at least five to ten bp. J) Even though the (dT-dG)n.(dC-dA)n polymer requires base modification and high salt conditions to undergo the R----L transition, supercoiling (sigma less than or equal to -0.07) can supply enough energy to allow a plasmid containing the intervening sequence of a human fetal globin gene with (dT-dG).(dC-dA) sequences to undergo a R----L transition.

Alternating (dC-dG)n regions in DNA restriction fragments and recombinant plasmids were methylated at the 5 position of the cytosine residues by the HhaI methylase. Methylation lowers the concentration of NaCl or MgCl2 necessary to cause the B-Z conformational transition in these sequences. Ionic strengths higher than physiological conditions are required to form the Z conformation when the methylated (dC-dG)n tract is contiguous with regions that do not form Z structures, in contrast to the results with the DNA polymer poly(m5dC-dG) . poly(m5dC-dG). In supercoiled plasmids containing (dC-dG)n sequences, methylation reduces the number of negative supercoils necessary to stabilize the Z conformation. Calculations of the observed free energy contributions of the B-Z junction and cytosine methylation suggest that two junctions offset the favorable effect of methylation on the Z conformation in (dC-dG)n sequences (about 29 base-pairs in length). Studies with individual methylated topoisomers demonstrate that increasing Na+ concentration up to approximately 0.2 M inhibits the formation of the Z conformation in the (m5dC-dG)n region of supercoiled plasmids. The results suggest that methylation may serve as a triggering mechanism for Z DNA formation in supercoiled DNAs.

CD spectroscopy on the double-stranded strictly alternating dinucleotide polymer (dT-dG)n . (dC-dA)n partially modified by N-acetoxy-N-acetyl-2-aminofluorene suggests a left-handed conformation in concentrated NaCl solutions. Modification of the (dT-dG)n . (dC-dA)n polymer with acetylaminofluorene is required to promote formation of the left-handed helix since high salt concentrations and several other ionic conditions, which cause a similar transition for (dG-dC)n . (dG-dC)n, are ineffective. Furthermore, substitution of dC with 5-methyl dC in (dT-dG)n . (dC-dA)n does not facilitate formation of a left-handed helix, also in contrast to results found for (dG-dC)n . (dG-dC)n. A 62-base pair tract of almost perfectly alternating (dT-dG)n . (dC-dA)n from the 3'-side of the mouse kappa immunoglobulin gene modified with acetylaminofluorene undergoes the salt-induced transition to a left-handed helix when studied within a 140-base pair restriction fragment. High NaCl concentrations alone will not cause the transition for this 62-base pair tract in this fragment nor in the recombinant plasmid pRW777, which contains this fragment.

Recombinant pBR322 derivatives were constructed containing tracts of (dC-dG) sequences which are 58, 32, 26, and 10 base pairs (bp) in length in conjunction with a 95-bp fragment containing the Escherichia coli lac operator-promoter. The biological properties of these plasmids were unusual since deletions in the (dC-dG) regions, but not in the pBR322 nor the lac segments, were frequently observed; segments of (dC-dG) longer than approximately 50 bp were not stable but suffered deletions. Segments of approximately 30 bp or shorter were stable in most cases. The (dC-dG) tracts seemed to enhance recA-mediated recombination when they were of suitable length and were cloned into certain sites in the recombinants. Also the (dC-dG)-containing plasmids were less supercoiled (by 6-12 turns) than expected, relative to control plasmids, after isolation from E. coli hosts. These recombinant plasmids were used in the following paper to evaluate the properties of segments containing the unorthodox left-handed conformations.