Search

More than 20 years ago, it was initially predicted that hormones and growth factors might promote cell growth by binding and thereby inactivating tumor suppressors, as exemplified by the proposed complex formation between insulin and retinoblastoma protein (RB). This mainly intracellular/nuclear growth-regulatory circuit was termed "the nucleocrine pathway" and the physical interaction between insulin and RB was subsequently proven through several methods, primarily by immunofluorescence and co-immunoprecipitation. Meanwhile, additional nucleocrine pairs have emerged through further experimental studies, specifically the FGF1-p53 and angiogenin-p53 heterodimers. Moreover, first experimental clues have been obtained as to the intranuclear presence of the previously surmised heterodimer between the EGF precursor and the p130 tumor suppressor. In addition, RB-binding motifs have recently been discovered in interleukin-6 (IL-6) and cellular apoptosis susceptibility (CAS) protein. These findings point to a more general significance of the nucleocrine pathway in cell growth regulation and as a particularly useful target in cancer therapy.

About twenty years ago, the search for a peptide mimetic of the crucial 928-amino acid human retinoblastoma tumor suppressor protein (RB) has led to the identification of a potential active site in RB spanning 6 amino acids. Subsequent studies revealed that this hexapeptide needs to be coupled to a cellular internalization sequence in order to be biologically active. The prototype molecule resulting therefrom has been the synthetic 22-amino acid peptide MCR-4 that was proven through several investigations to exert both in vitro and in vivo anti-cancer activity both resembling and going beyond the antiproliferative effects of its template RB. This was followed by the elaboration of a distinct series of MCR peptides designed to also interfere with target structures located on the surface of cells and known to be involved in triggering cell proliferation such as the insulin receptor. For this second generation group, the prototype peptide has been another 22-amino acid peptide coined MCR-14 and equally demonstrated to display in vitro and in vivo anti-tumor effects. Finally, a miniaturization of the structure-function relationships intrinsic to the already small MCR-4 peptide has led to the development of the third generation 17-amino acid peptide MCR-15 which in turn was also shown to inhibit in vitro and in vivo malignant cell reproduction. Given more recent conceptual advances, it is warranted to claim that MCR peptides are likely candidate therapeutics not only for the treatment of neoplasias, but also of viral and bacterial infections, type 2 diabetes as well as hypertension.

Tumor suppressors constitute the body's primary defense line against malignant transformation. Since Theodor Boveri's initial insight one century ago, a huge amount of knowledge on these molecules has been generated. However, the final step of application of this profound understanding in the clinical setting, i.e., the treatment of cancer patients with tumor suppressors and their derivatives, is still ahead. Nevertheless, the important success achieved with similar biomimetic approaches in the therapy of other diseases suggests that tumor suppressor-based antineoplastic interventions should be accomplished soon as they may be equally rewarding.

In this brief survey, I look back upon a quarter of a century of my personal experience with synthetic peptides. Thereby, I focus on major steps in the design and experimental exploration of peptides that I have derived from the retinoblastoma tumor suppressor protein (RB). Along this way, both Merrifield's solid phase peptide synthesis method and collaborations with established investigators in the peptide and cancer research fields have played an important role.

In this survey, the initial insights on the sub- and transcellular process of oncoprotein metastasis (OPM) are linked to recent observations and advances in related fields. The six proteins described here, i.e. insulin, osteopontin, interleukin-6, anterior gradient-2 protein, cellular apoptosis susceptibility protein and hepatoma-derived growth factor, as well as distinct peptide fragments thereof might henceforth serve as pivotal biomarkers for OPM in clinical chemistry, molecular morphology, pathology and oncology and, as a result, guide as potential targets future structure-based interventions in cancer treatment.

Insulin is one of the major metabolic hormones regulating glucose homeostasis in the organism and a key growth factor for normal and neoplastic cells. Work conducted primarily over the past 3 decades has unravelled the presence of insulin in human breast cancer tissues and, more recently, in human non-small cell lung carcinomas (NSCLC). These findings have suggested that intracellular insulin is involved in the development of these highly prevalent human tumors. A potential mechanism for such involvement is insulin's binding and inactivation of the retinoblastoma tumor suppressor protein (RB) which in turn is likely controlled by insulin-degrading enzyme (IDE). This model and its supporting data are collectively covered in this survey in order to provide further insight into insulin-driven oncogenesis and its reversal through future anticancer therapeutics.

Previous investigations on proteasomal preparations containing insulin-degrading enzyme (IDE; EC 3.4.24.56) have invariably yielded a co-purifying protein with a molecular weight of about 110kDa. We have now found both in MCF-7 breast cancer and HepG2 hepatoma cells that this associated molecule is the retinoblastoma tumor suppressor protein (RB). Interestingly, the amount of RB in this protein complex seemed to be lower in HepG2 vs. MCF-7 cells, indicating a higher (cytoplasmic) protein turnover in the former vs. the latter cells. Moreover, immunofluorescence showed increased nuclear localization of RB in HepG2 vs. MCF-7 cells. Beyond these subtle differences between these distinct tumor cell types, our present study more generally suggests an interplay between RB and IDE within the proteasome that may have important growth-regulatory consequences.

The past two decades have witnessed an increasing appreciation of the role of the tumor microenvironment, of genetic and epigenetic alterations in normal cells adjacent to tumors and of the migration of normal cells with aberrant intrinsic properties in cancer pathophysiology. Aside from these insights, a novel concept termed "oncoprotein metastasis" (OPM) has recently been advanced and proposed to reflect protein-based neoplastic phenomena that might occur even before any modifications relating to the morphology, location or (epi)genetic outfit of cells during the malignant process. Here, evidence is presented that supports the OPM perception and thus should contribute not only to further rethink the definition of a normal cell, but also the treatment of cancer disease in the years to come.

Insulin has been known for a long time to influence the growth and differentiation of normal and transformed cells. In order to delineate the role of insulin specifically in non-small cell lung cancer (NSCLC), we have now searched by immunohistochemistry (IHC) for the presence of insulin in NSCLC samples. Among the 112 samples we studied, 30 were found to contain insulin, which was detected in the form of intracytoplasmic granula. Moreover, its expression significantly correlated with (a) the morphological/histopathological subtype of NSCLC, being more frequent in adenocarcinomas; (b) the grade of tumor differentiation, displaying an increase in low-grade carcinomas; (c) tumor size, occurring predominantly in smaller tumors; (d) the presence of phosphorylated, activated insulin receptor; (e) the median patient age, being present in relatively younger individuals. Furthermore and interestingly, surrounding atypical adenomatous hyperplastic areas and normal alveolar pneumocytes scored insulin-positive in some of the insulin-negative tumors. In addition, PCR exploration for insulin transcripts in some samples positive for immunoreactive insulin was negative, indicating a possibly exogenous origin for the intracellular insulin in our NSCLC cohort. Taken together, our data suggest that an intracellular insulin activity is important for the progression of low-grade human lung adenocarcinomas.

Tumor suppressor proteins are the main cellular barrier opposing neoplastic transformation. Among these host molecules, retinoblastoma protein (RB) plays a central role. A novel insight is now advanced to suggest that various inhibitors of insulin and insulin-like growth factor (IGF) signalling such as the putative tumor suppressor insulin-degrading enzyme (IDE) as well as the anti-oncogenic proteins PTEN and insulin-like growth factor-binding protein 7 (IGFBP-7) serve the common goal of ensuring that RB remains active. Since, moreover, IDE and IGFBP-7 each potentially achieves RB protection through preventing both binding and inactivation of RB by insulin, IGF-1 or IGF-2, the present perception also vindicates the importance of previous findings on the physical interaction of any of these growth factors with RB for cell fate. Notably, the therapeutic counterpart of this natural principle for maintaining or restoring RB function through insulin/IGF neutralization is the innovative class of anti-cancer agents termed MCR peptides and developed over the past decade.

Insulin is a hormone crucial to metabolism and an essential growth factor for normal and neoplastic tissues. We have now determined insulin in extracts of 23 primary breast cancer specimens and of non-neoplastic breast tissues by a chemiluminescent immunoassay. Remarkably, insulin was measured only in grade 3 tumors, whereas grade 2 carcinomas and the normal mammary gland were each insulin-negative. We also performed immunohistochemistry for insulin-degrading enzyme (IDE), a cytoplasmic zinc metalloprotease belonging to the inverzincin family and participating in insulin cleavage. IDE was detected in most insulin-positive grade 3 carcinomas, indicating that it might be dysfunctional in these anaplastic tumors. IDE was equally present in the insulin-negative grade 2 carcinomas. Moreover, five grade 3 carcinomas and one grade 2 carcinoma displayed a loss of heterozygosity in the 10q chromosomal region harboring the IDE gene, but, despite these alterations, IDE was detected immunohistochemically, indicating a retention of the second allele. Compared to the expression of IDE in 92% of the tumors examined, only 57% of 21 normal breast specimens stained positively for IDE. In contrast to this increase in IDE-positive epithelial cells in breast cancer vs. normal breast, additional immunohistochemical analysis of 17 node-positive breast carcinomas and corresponding tumor-bearing lymph nodes showed that IDE expression decreases from primary tumor to lymph node metastasis. Altogether, this study represents the first demonstration of IDE in normal and neoplastic human mammary tissues. Our present report should also provide an experimental starting point towards exploring a potential role of IDE in the control of tumor progression.

Previous studies have shown that MCR peptides possessing the retinoblastoma protein (RB) fragment LFYKKV as the active site are able to inhibit the proliferation of human non-small cell lung cancer (NSCLC) cells in vitro and in vivo. The goal of the present study was to test these compounds against human small cell lung cancer (SCLC) in vivo since this tumor is notoriously resistant to conventional therapy or, respectively, is characterized by rapid relapse after initially successful treatment. We herein report that the MCR peptides MCR-4 and MCR-14 display potent antiproliferative activity against RB-negative H82 SCLC xenograft tumors in nude mice, whereas the chemotherapeutic agent VP-16 tested in parallel in a clinically relevant dose had no anti-tumor effect. These encouraging results warrant accelerating the introduction of MCR peptides into clinical trials in patients with RB-negative tumors such as SCLC in the near future.

Peptides containing retinoblastoma protein (RB) fragment 649-654 (LFYKKV) were tested for their ability to block the proliferation of RB-negative and RB-positive human non-small cell lung cancer (NSCLC) cells. These peptides potently restrained the growth of both types of tumor cells, as measured by metabolic (MTT) and cellular viability (trypan blue exclusion) assays. As such, and remarkably, the peptides were able to overcome the resistance of RB-positive cells usually observed with RB gene or protein replacement therapy. Compared to the overall performance of conventional chemotherapy tested in parallel, the peptides were more cytotoxic against RB-negative neoplastic cells and equipotent toward RB-positive tumor cells, yet less toxic toward normal human cells. Thus, these new molecules hold great promise to evolve into an efficient therapy for human lung cancer, a common malignancy still defying treatment and holding a poor prognosis, as well as for other human neoplasias., (Copyright 2000 Academic Press.)

Considerable evidence exists that lung cancer cell lines produce large amounts of insulin-like growth factor-binding proteins (IGFBPs). In addition, these cells are subject to an autocrine or paracrine growth control by insulin-like growth factors (IGFs). We now demonstrate by immunocytochemistry with IGFBP-3 antibodies that nuclei of a lung cancer cell line distinctly immunostain for IGFBP-3. This finding led us to investigate in more detail the localization of this protein that, to date, had only been known to occur extracellularly. Ligand blotting revealed that purified nuclear extracts contain a 43,000-Da IGFBP which can bind [I125]IGF-I. By Western blot this protein was identified as IGFBP-3. Thus, our data are consistent with the results of a previous structural study predicting a nuclear localization for IGFBP-3. Moreover, our findings raise the possibility that nuclear IGFBP-3 is functional and involved in the pathogenesis of lung cancer.

Prion diseases are transmissible, neurodegenerative disorders associated with as yet incompletely defined isoforms of a cellular protein termed prion protein (PrP). We have now identified in PrP structural information compatible with nucleotide- and nucleic acid-binding. As such, PrP contains a putative nicotinamide adenine dinucleotide (NADH)-binding site. Moreover, the PrP octarepeats reveal homology to the nucleic acid-binding and strand-annealing octarepeats of mammalian heterogeneous ribonucleoprotein (RNP) A1. Therefore, PrP may have NADH-dependent oxidoreductase activity as well as A1-like functions such as nucleic acid annealing and splicing. Moreover, we propose that infectious prions are propagated through a dynamic molecular symbiosis between a ribozyme-like nucleic acid and a conformational isomer of the RNP-like prion protein. Thus, our model has important implications for the understanding and treatment of prion diseases.

Previous studies have demonstrated the presence of the insulin receptor in the cell nucleus. Recently, it was shown that the insulin receptor also exhibits nuclear tyrosine kinase activity. In the present investigation, I have searched for structural correlates to a nuclear localization of the insulin receptor as well as to other potential nuclear actions of this molecule. Interestingly, this analysis yielded that the insulin receptor (alpha-subunit) contains a bipartite nuclear localization signal (consistent with the preceding experimental data), several zinc finger-like motifs and an RGG box. These findings have intriguing implications with regard to a presumable role of the insulin receptor (alpha-subunit) as a gene regulatory molecule.

Environmental challenges appear to elicit similar patterns of cellular responses such as positive autoregulation and autoamplification whether one considers the generation of antibodies with identical antigen specificity or the accumulation of host-protective transcription factors. Therefore, I analyzed the structure of immunoglobulins (Ig) for motifs commonly found in transcription factors. Specifically, the well-known abundance and periodic location of cysteine residues in immunoglobulin chains prompted me to check antibody constant regions for the presence of putative metal-binding domains and zinc finger-like sequences. The constant regions of Ig light and heavy chains were found to harbor one or several copies, respectively, of a short cysteine- and histidine-containing sequence. Moreover, all four IgG subclasses were detected to comprise zinc finger-like motifs in their heavy chain constant and hinge domains. Yet another finding is the occurrence of several sequences of the form serine-proline-X-X and/or threonine-proline-X-X in the hinge sections of IgA and IgG3. These results suggest that antibody constant regions, as a fragment and/or embedded in a full-length immunoglobulin chain, may complex metal, thus acquiring conformations conducive to dimerization and nucleic acid binding. As such, my study provides a putative structural basis for the known requirement of divalent metal cations, particularly of zinc ions, for a normal immune response, and warrants further investigations, both theoretical and experimental, into the potential of antibody constant regions for metal binding and gene regulation. Moreover, future testing of the proposed zinc finger peptides from Ig constant domains should yield information relevant to zinc finger design with potentially wide applications in research and clinical medicine.(ABSTRACT TRUNCATED AT 250 WORDS)

Peptides corresponding to retinoblastoma protein (RB) fragment 649-654 (LFYKKV) were tested for their ability to recognize the LXCXE sequence motif in human papilloma virus type 16E7 protein (HPV-16E7) encompassing E7 residues 21-26 (DLYCYE) and an identical motif in human insulin comprising insulin B-chain residues 16-21 (YLVCGE), respectively. Interaction between these complementary peptide sequences was observed by several approaches, including direct and competitive ELISA as well as affinity chromatography. Moreover, we demonstrated that immobilized RB649-654 displays specific recognition properties towards full-length insulin. Hence, this study provides a first experimental support for the previously anticipated complex formation between insulin and RB.

The present study reports structural similarities between viral oncoproteins, growth factors belonging to the insulin family, members of the steroid/thyroid receptor superfamily, a D-type cyclin, the Elf-1 transcription factor and Bcl oncoproteins in regions that have been shown or proposed to mediate complex formation of these proteins with the tumor suppressor retinoblastoma protein (RB). This relationship predicts a common intracellular pathway for mitogenic signals and molecules promoting cell survival. Conversely, the structural evidence described here suggests that RB may play a central role both at the boundary between negative and positive cell growth regulation as well as in developmental decisions between cell death and cell survival.

The present study proposes a novel type of synthetic chimaeric polypeptides potentially useful in the therapy of various diseases. The prototype peptide termed 'synthetic inducible biological response amplifier' (SIBRA) would comprise a ligand-binding site, a DNA-binding region, a trans-activating domain as well as strings of residues ensuring bioavailability and targeting to specific compartments such as the cell nucleus. These domains would be selected from cellular proteins, artificially tailored to a SIBRA and further modified towards a molecule with both in vivo and intracellular activity. Since proposed to resemble a host molecule with autoregulatory properties, a SIBRA would be activated upon exposure to a defined environmental stimulus and amplify host responses appropriate for this stimulus. Proteins would accumulate that share functional domains with the administered SIBRA and have a positive autoregulatory capacity. The latter may involve the interaction of the induced protein with the promoter of its gene resulting in a direct positive autoregulatory loop or require the induction of intermediary proteins that eventually upregulate the production of SIBRA-like host proteins. Since the ligand-binding site of a SIBRA is rationally designed to target a pathogenic protein, SIBRAs could be regarded as the product of an artificial acceleration and refinement of strategies intrinsic to the immune system.

Retinoblastoma protein (RB) is a tumor suppressor gene product involved in embryogenesis and cell cycle progression. One of the major mechanisms leading to RB dysfunction is complex formation with viral oncoproteins using the common RB binding motif Leu X Cys X Glu (LXCXE) which has also been identified in cellular ligands, e.g., RBP-1 and RBP-2. p107, a cellular protein with RB sequence homology, has been shown to bind to the same viral oncoproteins associating with RB and is therefore thought to contribute to cell cycle regulation. It has recently been suggested that insulin stimulates gene transcription through direct association with an, as yet, unidentified intracellular transcription factor. Due to the central roles of RB and p107 in coupling external growth signals with the progression of the cell cycle clock, we have hypothesized that these two proteins might be candidates for mediating the effects of insulin on DNA. We report here the identification of a region in the B-chain of human insulin that has the sequence LXCXE. Based on this finding we predict that the insulin B-chain may interact with RB and/or p107. Since we have also identified sequences hydropathically related to LXCXE in insulin-like growth factor I (IGF-I) and II (IGF-II), but not in relaxin, nerve growth factor, epidermal growth factor, glucagon or beta-endorphin, we further propose that both IGF-I and -II may assemble with RB and/or p107, too. Moreover, binding sites on RB and p107 identical with those suggested for viral oncoproteins and cellular ligands are predicted for insulin/IGF-I/IGF-II by using the hydropathic complementarity approach.(ABSTRACT TRUNCATED AT 250 WORDS)

Hydropathic anticomplementarity of amino acids specifies that peptides translated from complementary DNA strands may acquire amphiphilic conformations and bind to each other. This concept has been coined 'Molecular Recognition Theory' (MRT) or 'complementary peptide theory'. Inactivation of retinoblastoma protein (RB), a tumor suppressor gene product, has been shown to be involved in the pathogenesis of many tumors and to be due to either mutation of the RB gene, hyperphosphorylation or complex formation with viral oncoproteins. The viral oncoproteins share a common RB binding motif with cellular ligands. The exact site on RB associating with this common RB binding motif of viral oncoproteins and cellular ligands has not been identified yet. This study is the first to predict putative binding sites on RB and p107, a cellular protein with RB sequence homology, respectively, by using the hydropathic complementarity approach. These sites are residues 649-654 of RB and 657-662 of p107. Moreover, this paper proposes a structure for a potential antineoplastic agent based on the amino acid sequence of the predicted RB binding site. The data presented herein should have important implications both for the understanding of cancer pathophysiology and for the drug design of antineoplastic compounds.

The mucosal immune system plays an important role in blocking the penetration of invasive organisms into various mucosal surfaces. Evidence now suggests neuroendocrine peptide hormones have immunomodulatory properties, including the ability to alter mucosal immunity. The potential for opioid compounds and corticotropic hormone (ACTH) to modulate mucosal immune function was investigated. We have found beta-endorphin, ACTH, and naltrindole (delta-class opioid receptor antagonist) to significantly suppress concanavalin A-stimulated Peyer's patch lymphocyte immunoglobulin production of IgA, IgG, and IgM isotypes. Oxymorphindole, a delta class opioid receptor agonist, significantly decreased IgM but not IgA or IgG production by the mitogen-stimulated Peyer's patch lymphocytes. Both oxymorphindole and naltrindole modestly reduced interleukin-2 receptor expression of concanavalin A- (Con A)-stimulated splenic and Peyer's patch lymphocytes. Neither compound appreciably affected immunoglobulin production by lipopolysaccharide-stimulated Peyer's patch lymphocytes. Collectively, these results indicate stress-related peptides such as ACTH and opioids may be involved in the regulation of immunoglobulin synthesis by Peyer's patch lymphocytes.