Poster Presentation 9th Australasian Virology Society Meeting 2017

Mapping resistance to CCR5 antagonist maraviroc in clinical HIV-1 strains (#129)

Jacqueline Flynn 1 , Paula Ellenberg 1 , Renee Duncan 2 , Anne Ellett 2 , Jingling Zhou 1 , Jasminka Sterjovski 3 , Kieran Cashin 1 , Katharina Borm 2 , Lachlan Gray 2 , Marilyn Lewis 4 , Becky Jubb 4 , Mike Westby 4 , Benhur Lee 5 , Sharon Lewin 3 6 , Melissa Churchill 1 , Michael Roche 1 3 6 , Paul Gorry 1
  1. RMIT University, Melbourne, Victoria, Australia
  2. Burnet Institute, Melbourne, Victoria, Australia
  3. Melbourne University, Melbourne, Victoria, Australia
  4. Pfizer Global Research and Development, Sandwich, UK
  5. Icahn School of Medicine at Mount Sinai, New York, USA
  6. The Peter Doherty Institute for Infection and Immunity, , Melbourne, Victoria, Australia

Maraviroc (MVC) is the only CCR5 antagonist licensed for use as an anti-HIV-1 therapeutic and is being tested as a Pre-exposure prophylaxis prevention strategy. MVC resistance can occur, however the mechanisms behind its development are unclear. Elucidating these mechanisms will assist in the design of improved CCR5 antagonists for use in prevention and therapy.

 

Envs were cloned from plasma of pre-treatment (n=36) and treatment failure (n=41) samples from 8 subjects who experienced phenotypically verified, CCR5-tropic MVC resistance (phase III MOTIVATE clinical trial). MVC resistance was characterised by measuring the maximal percent inhibition (MPI) in two cell lines. The role of the V3 loop in clinical resistance was examined by swapping the V3 loops from one representative baseline MVC sensitive Env and one representative failure time point MVC resistant Env for each subject. Additionally cross resistance to other CCR5 antagonists was examined.

 

MVC resistance Envs showed a variation in the magnitude of MVC resistance, measured by reductions in MPI, which was more pronounced in 293-Affinofile cells (35/41 resistance MPI mean 40+/-33%) compared to NP2 CD4 CCR5 (24/41 MPI mean 58+/-31%), despite both cells having similar CCR5 levels. V3 loop analysis revealed amino acid determinants of MVC resistance localized to the V3 Env region and were strain-specific. There was minimal cross-resistance to other CCR5 antagonists by MVC-resistant strains.

 

Our study revealed that 293-Affinofile cells are highly sensitive for detecting and measuring MVC resistance through a mechanism that is CCR5-dependent yet independent of CCR5 expression levels. Resistance mutants were strain-specific, suggesting that sequence based diagnostics and prognostics will need to be more sophisticated than position scoring for managing resistance in subjects taking MVC. Finally, the lack of cross-resistance to other CCR5 antagonists suggests that recognition of the MVC-modified form of CCR5 does not necessarily lead to recognition of other antagonist-modified forms of CCR5.