Current Topics in Medicinal Chemistry - Volume 4, Issue 14, 2004

A number of studies appearing in the mid-1990s employing antibody reagents and gene ablation technologies offered promise for a role of VLA-4 in a number of inflammatory diseases including asthma, rheumatoid arthritis, multiple sclerosis, inflammatory bowel disease, type I diabetes and atherosclosis. A brief summary and leading references are provided in reference [1]. These findings prompted a number of firms, among them Wyeth, Elan (then Athena Neurosciences), Merck, Biogen and Roche to initiate small molecule programs taking a number of different approaches that are detailed in this volume. Each of them has lead to the identification of important lead series providing either a clinical candidate or a clear path to one and highlights a number of issues, which have slowed progress in this area. All lead compounds incorporate carboxylic acids leading to generally poor oral absorption and particularly vexing has been the almost universally observed rapid hepatic clearance of unchanged drug via an acid transporter, limiting the duration of exposure. The animal disease model data reported herein suggest that there hurdles may be surmountable. Activated integrins are high molecular weight heterodimeric glycoproteins extending stalk-like from the cell surface. The ligand bindng domain is near the N-terminus some 100 Å from the cell membrane. Although it was widely known that affinity is subject to modulation through intracellular signaling (inside out) and ligand binding induces cytoskeletal responses (outside in signaling) the mechanism of these effects was unknown. The availability of the first x-ray structures of an integrin [2,3] combined with the creative work carried out in the Springer laboratories summarized in the Shimaoka and Springer paper provide important insights on these mechanisms and provide clues to further optimization of lead molecules. Finally, the recent reports in the New England Journal of Medicine on phase II clinical trials of the humanized anti-alpha4 integrin monoclonal antibody, natralizumab, in multiple sclerosis [4] and Crohn's disease [5] provide the first validation of the alpha4 integrins as relevant targets for human disease and increase our hopes that small molecule approaches such as those illustrated in this volume will lead to import new medicines in the coming years.

Through a directed screening of a combinatorial library containing carboxylic acids, N-sulfonylated dipeptides were identified as leads in the Merck Research Laboratories VLA-4 antagonist program. Further optimization quickly identified subnanomolar compounds with varying degrees of specificity over the related integrin α4β7. Various metabolic liabilities were identified and addressed. However, the pharmacokinetic properties of nearly all compounds in this class were unacceptable. Other leads were identified with apparent good oral bioavailability, but these were generally associated with very high plasma protein binding and a loss of potency. The mechanism of high plasma clearance was identified in the rat as the organic acid transporter, mrp-2. Compounds were identified that were not substrates of mrp-2, but they still suffered from poor oral bioavailability. Finally, a shift in strategy to identifying VLA-4 antagonists that would be suitable as candidates for inhalation therapy resulted in the preparation of compounds with exception tight binding properties. These compounds were superior to BIO-1211 in the ovalbumin-sensitized mouse model of eosinophil trafficking to the lung. One particular compound had an exceptionally long off-rate with a KD ≤ 2 pM. The evolution of the structure activity relationships in our laboratories and strategies for improving potencies and pharmacokinetic profiles are the subject of this review.

The identification of orally active, small molecule antagonists of the α4β1 integrin, VLA-4, could lead to therapeutic agents with utility in a number of clinical settings, including asthma, multiple sclerosis and IBD. Starting from CDR3 sequences conserved among neutralizing α4 antibodies, peptides were identified that antagonized VLA-4 mediated adhesion in vitro. Through a series of structural modifications, these peptides evolved into small molecules that exhibited high potency and selectivity for VLA-4 in cell adhesion assays. Finally, through the optimization of physical and pharmacokinetic properties, compounds were identified that exhibited oral activity in animal models of asthma and multiple sclerosis.

The β2 integrins are validated therapeutic targets for inflammatory disorders. Two distinct mechanistic classes of small molecule inhibitors, termed α I allosteric and α / β I-like allosteric antagonist, have recently been developed. The α I allosteric antagonists bind underneath the C-terminal helix of the I domain and stabilize the I domain in the inactive closed conformation. By contrast, the α / β I-like allosteric antagonists bind to the b2 I-like domain MIDAS and disrupt conformational signal transmission between the I and the I-like domain, leaving the I domain in a default inactive form. Furthermore, the two classes of the antagonists have opposite effects on integrin conformation; the a I allosteric antagonists stabilize the bent conformation, whereas the α / β I-like allosteric antagonists induce the extended conformation with inactive I domain. The small molecule antagonists to the β2 integrin highlight the importance of the structural linkages within and between integrin domains for transmission of the conformational signals and regulation of the overall conformation.

Asthma, a chronic inflammatory disease of the airways, is a significant burden on our healthcare system. There is high unmet need for treatments directed towards the underlying causes of the disease. The cell surface integrin VLA-4 (very late antigen-4; α4β1; CD49d / CD29) plays an important role in the trafficking of white blood cells to sites of inflammation and represents an exciting target for the development of novel anti-inflammatory drugs for the treatment of asthma. Here, we review our efforts to use rational design to identify potent, selective inhibitors of VLA-4. We describe the discovery of a series of potent VLA-4 inhibitors through the addition of a novel N-terminal organic cap to a tetrapeptide VLA-4 binding motif 4-((N'-2-methylphenyl)uriedo)phenylacetyl-Leu-Asp-Val-Pro ; Kd = 70 pM), and rationalize their structure-activity relationships using 3D-QSAR. Also, we show our rational peptidomimetic design strategy using “template hopping” from the gpIIb / IIIa integrin antagonist field, and also a novel virtual screening strategy. Two series have been developed, one that has high selectivity for the activated over the non-activated state of the receptor, and the other which is non-selective inhibiting both activated and non-activated VLA-4. Both series are highly selective for VLA-4 versus against other integrin family members. These inhibitors show promise in the treatment of asthma, based upon efficacy in a sheep model of asthma, where they inhibit both the early and late-phase responses to asthma and also block hypersensitivity.

Starting with a cyclic peptide of moderate potency as a VLA-4 antagonist, highly potent and conformationally defined cyclic peptides were developed incorporating a constrained tyrosine and an achiral Asp-Pro spacer. N-Acyl phenylalanine derivatives were also discovered to have VLA-4 antagonist activity. During the course of development of this series, we found that the active acylphenylalanines mimic the pharmacophores present in the cyclic peptides and hypothesized that they bind to the same site on VLA-4. This insight guided our optimization strategy. Based on the emerging SAR, as well as insights from the recent X-ray crystal structure of the integrin αvβ3 bound to a RGD containing cyclic peptide, we propose a binding model for these compounds.