Utility of homology models in the drug discovery process

Abstract

Advances in bioinformatics and protein modeling algorithms, in addition to the enormous increase in experimental protein structure information, have aided in the generation of databases that comprise homology models of a significant portion of known genomic protein sequences. Currently, 3D structure information can be generated for up to 56% of all known proteins. However, there is considerable controversy concerning the real value of homology models for drug design. This review provides an overview of the latest developments in this area and includes selected examples of successful applications of the homology modeling technique to pharmaceutically relevant questions. In addition, the strengths and limitations of the application of homology models during all phases of the drug discovery process are discussed.

Figure 1

The steps involved in the prediction of protein structure by homology modeling. Structure modeling of the bacterial transcriptional repressor CopR is shown [28]. Although the model is based on a low-sequence identity of only 13.8% between CopR and the P22 c2 repressor, several experimental methods support this homology model. Reproduced, with permission, from Ref. [84]. Abbreviation: CopR, plasmid copy control protein.

Figure 2

Relationship between target and template sequence identity and the information content of resulting homology models. Arrows indicate the methods that can be used to detect sequence similarity between target and template sequences. Applications of the homology models in drug discovery are listed to the right. The higher the sequence identity, the more accurate the resulting structure information. Homology models that are built on sequence identities above ∼50% can frequently be used for drug design purposes. Superimpositions of X-ray crystal structures of the ligand-binding domains of members of the nuclear receptor family are shown to the left. These X-ray structures illustrate the increase in structure deviation with a decreased sequence identity. The PR is red, the GR is green, the ERα is blue and the TRβ is cyan. Sequence identities: PR:GR, 54%; PR:ERα, 24%; and PR:TRβ, 15%. Abbreviations: ERα, estrogen receptor α GR, glucocorticoid receptor; PR, progesterone receptor; TRβ, thyroid receptor β.

Figure 3

Applications of homology models in the drug discovery process. The enormous amount of protein structure information currently available could not only support lead compound identification and optimization, but could also contribute to target identification and validation. Reproduced, with permission, from [84].

Figure 4

(a,b) Comparison of the two isotypes of the estrogen receptor. In the homology model, ERα (blue) and ERβ (green) ligand-binding pockets are shown in complex with the natural ligand of the ER, 17β-estradiol. The binding of 8β-VE₂, a highly potent and selective ERβ agonist, modeled into the ERβ ligand-binding niche is depicted to the right. Reproduced, with permission, from Ref. [31]. (c,d) A model of the antiprogestin RU 486 (Mifepristone) bound to hPR. A single amino acid mutation renders this compound inactive at the cPR and hamster PR. Steric clashes between RU 486 and cPR are shown on the right side. Abbreviations: ER, estrogen receptor; hER, human estrogen receptor; cPR, chicken progesterone receptor; hPR, human progesterone receptor; PR, progesterone receptor; RBA, relative binding affinity; 8β-VE₂, 8β-vinylestra-1,3,5(10)-triene-3,17β-diol.

Figure 5

Structure of AG7088. This compound is an inhibitor of HRV2 3C proteinase and, on the basis of a homology model of HRV2 3C proteinase, was suggested as a potential inhibitor of SARS-CoV M^pro. Abbreviations: CoV, coronavirus; HRV2, human rhinovirus type 2; SARS, severe acute respiratory syndrome.