High-Throughput Humanized Mouse Models for Evaluation of HIV-1 Therapeutics and Pathogenesis.

Mice cannot be used as a model to evaluate HIV-1 therapeutics because they do not become infected by HIV-1 due to structural differences between several human and mouse proteins required for HIV-1 replication. This has limited their use for in vivo assessment of anti-HIV-1 therapeutics and the mechanism by which cofactors, such as illicit drug use accelerate HIV-1 replication and disease course in substance abusers. Here, we describe the development and application of two in vivo humanized mouse models that are highly sensitive and useful models for the in vivo evaluation of candidate anti-HIV therapeutics. The first model, hu-spl-PBMC-NSG mice, uses NOD-SCID IL2rγ(-/-) (NSG) mice intrasplenically injected with human peripheral blood mononuclear cells (PBMC) which develop productive splenic HIV-1 infection after intrasplenic inoculation with a replication-competent HIV-1 expressing Renilla reniformis luciferase (HIV-LucR) and enables investigators to use bioluminescence to visualize and quantitate the temporal effects of therapeutics on HIV-1 infection. The second model, hCD4/R5/cT1 mice, consists of transgenic mice carrying human CD4, CCR5 and cyclin T1 genes, which enables murine CD4-expressing cells to support HIV-1 entry, Tat-mediated LTR transcription and consequently develop productive infection. The hCD4/R5/cT1 mice develop disseminated infection of tissues including the spleen, small intestine, lymph nodes and lungs after intravenous injection with HIV-1-LucR. Because these mice can be infected with HIV-LucR expressing transmitted/founder and clade A/E and C Envs, these mouse models can also be used to evaluate the in vivo efficacy of broadly neutralizing antibodies and antibodies induced by candidate HIV-1 vaccines. Furthermore, because hCD4/R5/cT1 mice can be infected by vaginal inoculation with replication-competent HIV-1 expressing NanoLuc (HIV-nLucR)-, this mouse model can be used to evaluate the mechanisms by which substance abuse and other factors enhance mucosal transmission of HIV-1.

Overcoming the genotoxicity of a pyrrolidine substituted arylindenopyrimidine as a potent dual adenosine A(2A)/A(1) antagonist by minimizing bioactivation to an iminium ion reactive intermediate.

2-Amino-4-phenyl-8-pyrrolidin-1-ylmethyl-indeno[1,2-d]pyrimidin-5-one (1) is a novel and potent selective dual A(2A)/A(1) adenosine receptor antagonist from the arylindenopyrimidine series that was determined to be genotoxic in both the Ames and Mouse Lymphoma L5178Y assays only following metabolic activation. Compound 1 was identified as a frame-shift mutagen in Salmonella typhimurium tester strain TA1537 as indicated by a significant dose-dependent increase in revertant colonies as compared to the vehicle control. The metabolic activation-dependent irreversible covalent binding of radioactivity to DNA, recovery of 1 and its enamine metabolite from acid hydrolysis of covalently modified DNA, and protection of covalent binding to DNA by both cyanide ion and methoxylamine suggest that the frame-shift mutation in TA1537 strain involved covalent binding instead of simple intercalation to DNA. Compound 1 was bioactivated to endocyclic iminium ion, aldehyde, epoxide, and α,ß-unsaturated keto reactive intermediates from the detection of cyano, oxime, and glutathione conjugates by data-dependent high resolution accurate mass measurements. Collision-induced dissociation of these conjugates provided evidence for bioactivation of the pyrrolidine ring of 1. The epoxide and α,ß-unsaturated keto reactive intermediates were unlikely to cause the genotoxicity of 1 because the formation of their glutathione adducts did not ameliorate the binding of compound related material to DNA. Instead, the endocyclic iminium ions and amino aldehydes were likely candidates responsible for genotoxicity based on, first, the protection afforded by both cyanide ion and methoxylamine, which reduced the potential to form covalent adducts with DNA, and, second, analogues of 1 designed with low probability to form these reactive intermediates were not genotoxic. It was concluded that 1 also had the potential to be mutagenic in humans based on observing the endocyclic iminium ion following incubation with a human liver S9 preparation and the commensurate detection of DNA adducts. An understanding of this genotoxicity mechanism supported an evidence-based approach to selectively modify the structure of 1 which resulted in analogues being synthesized that were devoid of a genotoxic liability. In addition, potency and selectivity against both adenosine A(2A) and A(1) receptors were maintained.