Defense Date


Document Type


Degree Name

Master of Science


Pharmaceutical Sciences

First Advisor

Yan Zhang


Chemokine receptor CCR5 (CCR5) is a G-protein coupled receptor (GPCR) predominantly expressed on leukocytes, or white blood cells.1–3 During inflammation, the body releases chemokines that bind to receptors such as CCR5 and attract leukocytes to the area of inflammation, leading to an immunological response.1 CCR5 is also an important receptor in the human immunodeficiency virus's (HIV-1) invasion of host cells, as CCR5 acts as a co-receptor that facilitates HIV-1 viral entry.4,5 The continued destruction of leukocytes as a result of HIV-1 viral entry produces a disease state called acquired immunodeficiency syndrome (AIDS).5 Of note, this receptor is also expressed on the glial cells of the central nervous system (CNS).6,7 The mu-opioid receptor (MOR) is also a GPCR is predominantly expressed in the central nervous system.8–10 It binds to signal molecules such as endorphins and produces analgesic effects upon activation.9 The protein binds to morphine and morphine derivatives, which are extracts from the opium poppy plant.10 Besides the analgesic effects produced from MOR activation, morphine and its derivatives are also highly addictive and can result in drug dependence.11 Like CCR5, MOR is also expressed on the glial cells of the CNS.8 The accelerated progression of AIDS-like symptoms, in particular HIV-associated neurocognitive disorders (HANDS), has been observed in opiate-addicted patients.6,7,12–14 It has been discovered that opiate-addicted patients who have AIDS are susceptible to higher levels of HIV-1 viral proliferation and a greater level of CNS host cell destruction.12–14 This is because the activation of MOR by opiates appears to increase the expression of CCR5 on glial cells and may alter CCR5's conformational state to one more susceptible to HIV-1 binding.15 Then, entry and subsequent destruction of glial cells by HIV-1 leads to the release of neurotoxic HIV-1 proteins that destroy primary neuronal cells.15 A bivalent ligand targeting the putative CCR5-MOR heterodimer was proposed to probe the interaction between the two proteins and act as a potential therapeutic ligand to combat neuroAIDS.16 A bivalent ligand attaching maraviroc, a CCR5 antagonist, with naltrexone, a non-selective opioid receptor antagonist, was synthesized and tested in vitro.16 The initial bivalent ligand was separated by a 21-atom spacer (Figure 1), the length of which was dictated by modeling studies and other bivalent ligands.17 The spacer was attached to the 6-position of 6β-naltrexamine, a modified variation of naltrexone replacing the 6 position ketone with an amino group, and the 4'-position of the maraviroc phenyl ring.16 These positions were chosen based on separate modeling studies of naltrexone and maraviroc docked in homology models of MOR and CCR5, respectively.18,19 From these studies, it appeared as if these positions were optimal given that they faced outward from their respective binding pockets and hence could tolerate spacer attachment.18,19 However, based on these modeling studies there was also room for structure optimization of the bivalent molecule.17–19 The original 21-atom spacer was subjected to numerous structural changes by our laboratory in an effort to increase CCR5 and MOR binding. The first type of structural modification included changing maraviroc's point-of-attachment to the spacer from the 4'- to the 3'-position. Based on results from calcium mobilization functional assays involving CCR5-transfected human acute lymphoblastic leukemia (MOLT-4) cells, the activity of the bivalent molecule decreased from an IC50 of 126.0 ± 28.0 to 1340.0 ± 110.0 when the point-of-attachment was changed to the 3'-position. Thus, the 4'-position was kept in future structural studies. After this, additional structural modification was pursued in the form of changing spacer length. We synthesized two additional bivalent ligands, i.e., 19-atom and 23-atom bivalents with their controls. It is important to note that each of these molecules had a separate synthetic route starting with a specified diamine spacer. For the 19-atom bivalent molecule and its controls, the starting material was 1,5-diaminopentane. For the 23-atom bivalent molecule and its controls, the starting material was a 1,9-diaminononane molecule. Once these molecules were synthesized, in vitro biological testing was conducted. The bivalent molecules and 6β-naltrexamine controls were subjected to a competitive radioligand binding assay involving hMOR membranes and then to a calcium mobilization functional assay involving hMOR-transfected chinese hamster ovarian (CHO) cells (Figure 2). The affinities from the radioligand binding assay were similar in order-of-magnitude to other modified opioid antagonists and had a fold-decrease in affinity relative to naltrexone ranging from 1.1 to 9.3. The IC50 values from the calcium mobilization assay were similar in order-of-magnitude to other modified opioid antagonists and had a fold-decrease in activity relative to naltrexone ranging from 7.6 to 32. Thus, it was concluded that spacer attachment to the 6-position of 6β-naltrexamine was tolerated in MOR-binding. The bivalent molecules and maraviroc controls were then tested in calcium mobilization assays involving CCR5-transfed MOLT-4 cells to assess CCR5 activity. Unlike in the hMOR-CHO calcium mobilization assay, the activity of the bivalent molecules for CCR5 was not similar in magnitude to the receptor's antagonist, maraviroc. The 23-atom bivalent and 19-atom bivalent had fold-decreases in activity relative to maraviroc of 1,100 and 250, respectively. Recently, the co-crystal structure of maraviroc bound to CCR5 was published in Science.20 Contrary to our previous understanding, it appeared as if modifications to the phenyl ring in maraviroc were not tolerated.20 After the biological testing, conformational analysis on the 19-, 21- and 23-atom bivalent compounds using Confort conformational modeling software was conducted. This was done to observe the possible viable conformations of each molecule. It was hypothesized that 1) the molecules could adopt viable conformations for binding to two different receptors simultaneously and that 2) the molecules could adopt similar conformations relative to each other. The first hypothesis was proposed to assess the realism of the project's design strategy whereas the second was to analyze whether significant conformational differences could account for differences in binding activity between the three molecules. Results from this experiment showed that the molecules all adopted viable conformations for binding to two receptors simultaneously and that the conformational differences between the three molecules were negligible enough to conclude that significant differences in binding were not because of conformational differences. In conclusion, our laboratory synthesized a set of bivalent compounds to probe the CCR5-MOR heterodimer and tested such compounds in vitro. While spacer modifications to the 6-position of naltrexone were tolerated in hMOR competitive radioligand binding assays and in hMOR-CHO calcium mobilization assays, spacer modifications to maraviroc's 4'-position on the phenyl ring were not tolerated very well in CCR5-MOLT-4 calcium mobilization assays. Therefore, future design strategies might focus on changing the spacer's point-of-attachment to the maraviroc molecule.


© The Author

Is Part Of

VCU University Archives

Is Part Of

VCU Theses and Dissertations

Date of Submission