Publication Date


Document Type

Doctoral Dissertation

Academic Program

Biochemistry and Molecular Pharmacology


Program in Biochemistry and Molecular Pharmacology

First Thesis Advisor

Celia A. Schiffer, Ph.D.


HIV Protease, Drug Resistance, Viral, Substrate Specificity


In the 30 years since the Center for Disease Control's Morbidity and Mortality Weekly Report published the first mention of what later was determined to be AIDS (Acquired immunodeficiency syndrome) and HIV (Human immunodeficiency virus) recognized as the causative pathogen, much has been done to understand this disease’s pathogenesis, development of drugs and emergence of drug resistance under selective drug therapy. Highly Active Antiretroviral Therapy (HAART), a combination of drugs that includes HIV-1 reverse transcriptase, protease, and more recently, integrase and entry inhibitors, have helped stabilize the HIV prevalence at extraordinarily high levels. Despite the recent stabilization of this global epidemic, its dimensions remain staggering with estimated (33-36 million) people living with HIV-AIDS in 2007 alone. This is because the available drugs against AIDS provide treatment for infected individuals, but HIV evolves rapidly under drug pressure and develops resistant strains, rendering the therapy ineffective. Therefore, a better understanding underlying the molecular mechanisms of viral infection and evolution is required to tackle drug resistance and develop improved drugs and treatment regimens.

HIV-1 protease is an important target for developing anti-HIV drugs. However, resistant mutations rapidly emerge within the active site of the protease and greatly reduce its affinity for the protease inhibitors. Frequently, these active site drug resistant mutations co-occur with secondary/ non-active site/ associated or compensatory mutations distal to the active site. The role of these accessory mutations is often suggested to be in maintaining viral fitness and stability of protease. Many of the non-active site drug resistant mutations are clustered in the hydrophobic core in each monomer of the protease. Molecular dynamic simulation studies suggest that the hydrophobic core residues facilitate the conformational changes that occur in protease upon ligand binding. There is a complex interdependence and interplay between the inherent adaptability, drug resistant mutations and substrate recognition by the protease. Protease is inherently dynamic and has wide substrate specificity. The PI (protease inhibitor) resistant mutations, perhaps, modulate this dynamics and bring about changes in molecular recognition, such that, in resistant proteases, the substrates are recognized specifically over the PIs for the same binding site. In this thesis research, I have investigated these three complementary phenomena in concert.

Chapter II examines the importance of hydrophobic core dynamics in modulating protease function. The hydrophobic core in the WT protease is intrinsically flexible and undergoes conformational changes required for protease to bind its substrates. This study investigated if dynamics is important for protease function by engineering restricted vs. flexible hydrophobic core region in each monomer of the protease, using disulfide chemistry. Under oxidizing conditions, disulfide bond established cross-link at the interface of putative moving domains in each monomer, thereby, restricting motion in this region. Upon reduction of the disulfide bond, the constraining influence was reversed and flexibility returned to near WT. The disulfide cross-linked protease showed significant loss of function when tested in functional cleavage assay. Two protease variants (G16C/L38C) and (R14C/E65C) were engineered and examined for changes in structure and enzymatic activity under oxidizing and reducing conditions. (R14C/E65C) was engineered as an internal control variant, such that cysteines were engineered between putative non-moving domains. Structurally, both the variants were very similar with no structural perturbations under oxidizing or reducing conditions. While significant loss in function was observed for (G16C/L38C) only under oxidizing conditions, (R14C/E65C) did not show any loss of function under oxidizing or reduced conditions, as expected. Successful regain of function for cross-linked (G16C/L38C) was obtained upon reversible reduction of the disulfide bond. Taken together, these data demonstrate that the hydrophobic core dynamics modulates protease function and support the hypothesis that the distal drug resistant mutations, possibly causing drug resistance by modulating hydrophobic core dynamics via long range structural perturbations. Since protease recognizes and cleaves more than 10 substrates at different rates, our further interest is to investigate if there is a differential loss of activity for some specific substrates over the others, and whether the order of polypeptide cleavage is somehow affected by restricted core mobility. In order to better answer these questions it is essential to understand: what determines the substrate binding specificity in protease? A two-pronged approach was applied to address this question as described in chapter III and IV respectively.

In chapter III, I investigated the determinants of substrate specificity in HIV-1 protease by using computational positive design and engineered specificity-designed asymmetric protease (Pr3, A28S/D30F/G48R) that would preferentially bind to one of its natural substrates, RT-RH over two other substrates, p2-NC and CA-p2, respectively. The designed protease was expressed, purified and analyzed for changes in structure and function relative to WT. Kinetic studies on Pr3 showed that the specificity of Pr3 for RT-RH was increased significantly compared to the wild-type (WT), as predicted by the positive design. ITC (Isothermal Titration Calorimetry) studies confirmed the kinetic data on RT-RH. Crystal structural of substrate complexes of WT protease and Pr3 variant with RT-RH, CA-p2 and p2-NC were further obtained and analyzed. The structural analysis, however, only partially confirmed to the positive design due to the inherent structural pliability of the protease. Overall, this study supports the positive computational design approach as an invaluable tool in facilitating our understanding of complex proteins such as HIV 1 protease and also proposes the integration of internal protein flexibility in the design algorithms to make the in-silico designs more robust and dependable.

Chapter IV probed the substrate specificity determining factors in HIV-1protease system by focusing on the substrate sequences. Previous studies have demonstrated that three N-terminal residues immediate to the scissile bond (P1-P3) are important in determining recognition specificity. This work investigated the structural basis of substrate binding to the protease. Catalytically active WT protease was crystallized with decameric polypeptides corresponding to five of the natural cleavage sites of protease. The structural analyses of these complexes revealed distinct P side product bound in all the structures, demonstrating the higher binding affinity of N terminal substrate for protease.

This thesis research successfully establishes that intrinsic hydrophobic core flexibility modulates function in HIV-1 protease and proposes a potential mechanism to explain the role of non-active site mutations in conferring drug resistance in protease. Additionally, the work on specificity designed and N terminal product bound protease complexes advances our understanding of substrate recognition in HIV protease.



Rights and Permissions

Copyright is held by the author, with all rights reserved.