Current Pharmaceutical Design - Volume 8, Issue 24, 2002

The cloning and mass production of recombinant cytokine proteins opened a new world of treatment possibilities. While some cytokines, including several haematopoietic factors and interferons, are now used routinely in the clinic, there are still many problems with side effects. These are due to the many different activities of cytokines on different cell populations. In some cases, activities responsible for side effects have been attributed to discreet areas of the proteins and “structure driven design” can be used to generate novel proteins with better clinical profile. In other cases, structural alterations can enhance activity by increasing serum half-life. This review summarizes the structures of cytokines and their receptor complexes deposited in the Protein Data Base (PDB) and introduces the other articles in this issue on structure-driven design of cytokines for therapy. Cytokines fall into only a few structural classifications. Most of the growth regulatory cytokines including serum factors, growth hormone, haematopoietic growth factors, colony stimulating factors, erythropoietin, IL-3, IL-2 and interferons, are four or five helix bundles. Factors which primarily induce inflammatory responses, including TNF, lymphotoxin and IL-1, form β-barrel structures that resemble the FGF family. Chemokines and factors that regulate multicellular responses, such as macrophage migration, neutrophil invasion and chemotaxis, have similar structures, classified as α +β. One biological paradox is that many cytokines that vary greatly in function have a similar structure and share receptors. However, homologous cytokines may differ considerably in their mode of interaction with a shared receptor. A few structures for the extracellular regions of cytokine receptors are known, in several cases complexed with their biological target. These structures, coupled with structural alignment of families, indicate areas that control binding to receptors, as opposed to specific areas responsible for the specific activities of this diverse group of proteins. Methods to use cytokine structure to derive better therapeutics are summarized.

The Type I interferons (IFN-α / ß) exhibit pleiotropic biological activities. Notably, the different IFN subtypes activate the same cell surface receptor complex to mediate variable responses. Accumulating evidence suggests that distinct differences in critical amino acid residues among the different IFN-αs and IFN-ß determine the nature of the ligand-receptor interaction and the subsequent responses. This review focuses on IFN-receptor interactions, the key residues involved in this interaction and the potential for targeted modifications of the ligand to enhance bioactivity.

Derivatization of protein-based therapeutics with polyethylene glycol (pegylation) can often improve pharmacokinetic and pharmacodynamic properties of the proteins and thereby, improve efficacy and minimize dosing frequency. This review will provide an overview of pegylation technology and pegylated protein-based drugs being used or investigated clinically. The novel therapeutic, PEG Intron, formed by attaching a 12-kDa mono-methoxy polyethylene glycol (PEG) to the interferon alpha-2b protein, will be discussed in detail in terms of its structure, biological activities, pharmacokinetic properties, and clinical efficacy for the treatment of chronic hepatitis C. Detailed physicochemical and biological characterization studies of PEG Intron revealed its composition of pegylated positional isomers and the specific anti-viral activity associated with each of them. Pegylation of Intron A at pH 6.5 results in a mixture of ≥ 95% mono-pegylated isoforms with the predominant species (approximately 50%) derivatized to the His34 residue with the remaining positional isomers pegylated at various lysines, the N-terminal cysteine, as well as serine, tyrosine, and another histidine residue. The anti-viral activity for each pegylated isomer showed that the highest specific activity (37%) was associated with the His34-pegylated isomer. Though pegylation decreases the specific activity of the interferon alpha-2b protein in vitro, the potency of PEG Intron was comparable to the Intron A standard at both the molecular and cellular level. The substituted IFN had an enhanced pharmacokinetic profile in both animal and human studies, and, when combined with ribavirin, was very effective in reducing hepatitis C viral load and maintaining sustained viral suppression in patients.

Aberrant expression of chemokines and their receptors play causative roles in the pathophysiology of numerous autoimmune and inflammatory disease processes. Moreover, an integral step in HIV infection involves binding to chemokine receptors, and hence chemokines are intimately linked to HIV-related diseases. Therefore, chemokines and their receptors are excellent targets for developing drugs that are more specific and may be of benefit in the management of disease. Knowledge of the chemokine and chemokine receptor structures, and an understanding of the structural basis of their function are essential for structure-aided design of receptor decoys. Chemokine ligands bind their receptors with nanomolar (nM) affinity, and successful design of a small molecule antagonist should bind the receptor with similar high affinity and specificity. Chemokines bind receptors that belong to the 7-transmembrane class on leukocytes, and highly negatively charged proteoglycans that are present on the cell surface. Structure-function studies have identified regions in both the ligand and the receptor that mediate binding and activation. Structures of numerous chemokines have been solved though very little is known regarding receptor structures. This review will summarize the current knowledge on the structures, structure-function, and the efficacy of chemokine derivatives and functional domain peptides as antagonists, and discuss strategies for exploiting this information for designing decoys for inflammatory, autoimmune, and HIV-related diseases.

A recombinant human IL-2 analog (rIL-2, Proleukin) is currently being evaluated for clinical benefit in HIV infected patients. It is approved for therapy of patients with metastatic melanoma and renal cell carcinoma. Treatment of cancer patients with rIL-2 results in durable responses but is associated with life-threatening toxicity, which limits its use to patients in relatively good health. Antitumor efficacy associated with rIL-2 therapy are hypothesized to be mediated by distinct types of cells that express structurally different forms of the IL-2 receptor. This hypothesis suggests that it might be possible to engineer an IL-2 variant addressing the risks associated with the therapeutic use of IL-2. In this article, we review the clinical experience with IL-2 and its analogs, the evidence that different IL-2 receptors may dissociate efficacy and toxicity, and describe the generation of a novel IL-2 variant with the potential for a superior therapeutic index.

Granulocyte-macrophage colony stimulating factor (GM-CSF) activity has been linked to pro-inflammatory effects in autoimmune syndromes, such as rheumatoid arthritis. Thus GM-CSF mimetics with antagonist activity might play a therapeutic role in these diseases. The human GM-CSF core structure consists of a four α-helix bundle, and GM-CSF activity is controlled by its binding to a two-subunit receptor. A number of residues located on the B and C helices of GM-CSF are postulated to interact with the alpha chain of the GM-CSF receptor (GM-CSFR). Several approaches have been successfully utilized to develop peptide mimetics of this site, including peptides from the native sequence, a peptide derived from a recombinant antibody (rAb) light chain which mimicked GM-CSF receptor binding activity, and structurally guided de novo design. Analysis of the rAb light chain had suggested mimicry of GM-CSF with residues mostly contributed by the CDR I region. Key residues involved in CDR I peptide / GM-CSFR binding were identified by truncation and alteration of individual residues, while the structural elements required to antagonize the biological action of GM-CSF were separately tested in binding and inhibitory activity assays of multiple cyclic analogues. A peptide designed to retain the loop conformation of the CDR I region of the rAb light chain competed with GM-CSF for both antibody and receptor binding, but the role of specific residues in antibody versus receptor binding differed markedly. These studies suggest that structural analysis of peptide mimetics can reveal differences in receptor and antibody binding, perhaps including key interactions that impact binding kinetics. Peptide mimetics of other four-helix bundle cytokines are reviewed, including helical and reverse turn mimetics of helical structures. Use of peptide mimetics coupled with structural and kinetic analysis provides a powerful approach to identifying important receptor-ligand interactions, which implications for rational design of novel therapeutics.

Neurotrophins (NTFs) are a family of polypeptide growth factors that control the apoptotic death or survival, growth, and differentiation of neurons. NTFs also regulate several other cell populations such as lymphoid, epithelial, oligoglia, and mast cells. Disregulation of the NTFs or their receptors plays a key role (etiological or upstream) in certain human pathologies. Hyperactivity may lead to inflammatory pain, or some forms of cancer by autocrine / paracrine growth. Loss of activity may lead to neurodegeneration, neuropathic pain, or some forms of cancer by absence of differentiation. Consequently the NTFs and their receptors are important therapeutic targets, and pharmacological modulation may have applications ranging from treatment of chronic or acute neurodegeneration, some forms of cancer, and chronic pain (with agonists), and some forms of cancer or acute pain (with antagonists).

We used a novel rational design approach in the development of novel immunomodulatory peptides, in particular RDP58, with at least two primary biological activities, inhibition of TNF production and upregulation of heme oxygenase-1 (HO-1) activity. The design strategy used a variety of topological and shape descriptors in combination with an analysis of molecular dynamics trajectories to identify potential drug candidates. The process was initiated using a panel of 19 lead molecules derived from a functionallyimportant region of HLA I, on the premise that the peptides might modulate immune responses by blocking HLA / T lymphocyte interactions. Each of these 19 peptides was tested in vivo for potential cardiac allograft protection in a mouse model. Outcome of graft survival was the primary data available to describe nine of the peptides as active and ten as inactive. A virtual combinatorial library of 279,936 peptides was then generated, based on physical properties associated with efficacy in the original learning set of 19 peptides - in particular, the distribution of lipophilic residues. The library was screened using descriptors that fell into 2 categories:static (physico-chemical) and dynamic (conformational). The screenings identified 5 peptides with theoretic potential efficacy. These were synthesized and tested for activity in vitro and in vivo. Besides prolonging graft survival in rodents, RDP58 was found to inhibit TNF and increase HO-1 activity, these biological activities were tested exhaustively in various in vivo disease models. RDP58 demonstrated convincing potential to alleviate disease symptoms in many of the experimental diseases.