 |
|
|
| |
Resistance to UK-427,857 does not result in cross-resistance to other entry inhibitors
|
| |
Reported by Jules Levin
There were three posters today on the 3 CCR5 inhibitors in clinical development: Pfizer's UK-427,857, Schering's '690', and GSK's '140'. Each poster is interesting. Here is the first.
Note from Jules Levin: at the Quebec Resistance Workshop there was a panel discussion today about cross-resistance between the three CCR5 inhibitors. Pfizer researchers reported in this poster below NOT finding cross-resistance, but Schering researchers report in their poster that cross-resistance was found, although they studied CCR5 inhibitors but not specifically the GSK & Pfizer drugs. The Pfizer poster will follow my next report. Can these drugs be used together in combination? This is a question that should be addressed by research? Can these drugs be used sequentially? This remains to be determined as the Pfizer & Schering data suggest conflict on this question. These two questions were the subject of much discussion in the afternoon panel discussion & at the reception tonight following the panel discussion. As we are in the process of developing a new class of HIV drugs, these types of questions and more will be the subject of a new exciting era in treatment. Remember the early days of protease inhibitors & the controversies surrounding cross-resistance.
"Maraviroc (MVC, UK-427,857)-resistant HIV-1 variants, selected by serial passage, are sensitive to CCR5 antagonists (GW873140, Schering-C, Schering-D) and T-20 (enfuvirtide)"
M Westby, J Mori, C Smith-Burchnell, M Lewis, M Mosley, F Perruccio, R Mansfield, P Dorr, M Perros
Pfizer Global R&D, Sandwich, UK
AUTHOR CONCLUSIONS:
MVC resistance
--MVCres clones showed changes in the V3 loop suggesting these mutations are important for MVC resistance.
--The amino acid mutation A315T in the V3 loop plays an important role in conferring MVC resistance in CC1/85, since phenotypic susceptibility of the MVCres variant (CC1/85 clone 14) to MVC was associated with a reversion to the wild type residue at this position during preparation of a second virus stock.
Resistance to MVC does not result in cross-resistance to other entry inhibitors
--MVC did not select for HIV-1 variants cross resistant to other entry inhibitors in vitro, notably other CCR5 antagonists.
--We propose that SCH-C, SCH-D and GW873140 bind to similar regions of CCR5 to MVC, but hold the receptor in different conformations that prevent entry of MVCres variants.
BACKGROUND
--Maraviroc (MVC, UK-427,857) is a CCR5 antagonist in Phase 2b/3 clinical development for the treatment of HIV-1 infection.
--CCR5 antagonists inhibit virus entry but differ from the licensed fusion inhibitor, enfuvirtide (ENF,T-20), in that they bind to the host cell and not the virus envelope.
--The selection of viruses with resistance to MVC and retained CCR5 tropism by sequential passage of the primary isolates CC1/85 (subtype B) and RU570 (subtype G) through peripheral blood lymphocytes (PBLs) in the presence of MVC for up to 20 weeks has been reported previously.1
--To further understand the mechanisms of MVC resistance and the potential for cross resistance:
- biological clones were generated from these MVC-resistant (MVCres) viruses by limiting dilution and were tested against MVC and other inhibitors of HIV entry
- the binding of CCR5 antagonists to CCR5 was modeled.
METHODS
Biological clones were obtained from MVCres viruses by limiting dilution in PBLs in the presence of MVC.
Biological clones were:
- gp160-sequenced
- tested for phenotypic susceptibility to MVC, other CCR5 antagonists and the fusion inhibitor
ENF in replication-competent antiviral assays in PBLs.
--Binding of compounds to CCR5 was modeled using published information.
RESULTS
Sequencing of gp160 to identify biological clones
--Ten potential viral clones from the CC1/85 MVCres pool were analysed.
--Nine potential viral clones from the RU570 MVCres pool were analysed.
--All biological "clones" analysed had fewer "X"s in the gp160 sequence than the original stock of virus, validating the limiting dilution approach.
--Viruses used in further experiments were those with no "X"s in the gp160 sequence.
--Amino acid changes were identified in the V3 loop region of MVCres viruses.
--Amino acid changes associated with resistance to MVC were strain-specific and were different to those reported for other CCR5 antagonists.3
--No ENF resistance-associated amino acid substitutions were identified in the heptad repeat (HR) 1 domain involved in the interaction with HR2 (gp41 amino acids 36 to 45).
Drug susceptibility of MVCres clones in PBLs:
lack of cross-resistance to CCR5 antagonists and ENF
--Initial experiments in PBLs confirmed that the four MVCres clones were all phenotypically resistant to MVC (not shown).
--Fresh stocks of biological clones from each MVCres virus were tested for susceptibility to MVC, Schering-C (SCH-C), Schering-D (SCH-D), GW873140, ENF and a protease inhibitor control
(saquinavir, SQV) (Figure 2; data summarized in Table 2; see tables at end of this report).
--Three of the four MVCres clones remained fully resistant to MVC.
--All four MVCres clones remained sensitive to CCR5 antagonists, ENF and SQV, with <5 fold increases in IC50 relative to their respective start clones.
--The sensitivity of the MVCres clones to the CCR5 antagonists confirms that these viruses remain CCR5 tropic.
Further characterization of the CC1/85 MVCres clone 14:
sensitivity to MVC in PBLs is associated with a single amino acid change in the V3 loop
--In contrast to its original stock, CC1/85 MVCres clone 14 was sensitive to MVC (Figure 2).
--The Env region of the virus stock of CC1/85 clone 14 used in the PBL experiments described above was re-sequenced.
--A mixture of amino acid residues at a single position within the V3 loop (T315T/A) was identified, in contrast to the sequence of the original virus stock analysed.
--This result underscores the importance of the A315T mutation in the context of the MVCres CC1/85 clone 14 gp160 envelope, in conferring a resistant phenotype.
Molecular modeling
--Molecular modeling showed that all the CCR5 antagonists tested bound in a similar pocket of CCR5 within the trans-membrane region.
--All compounds occupy a slightly different molecular space within the binding pocket, especially in the region extending towards the 2nd extracellular loop (ECL-2).
--Differences in occupancy of the CCR5 binding pocket may explain the lack of cross-resistance seen in vitro.
REFERENCES
1 Westby M et al.Antivir Ther 2004; 9: S10.
2 Palczewski K et al. Science 2005; 289: 739-745.
3 Kuhmann SE et al. J Virol 2004; 78: 2790-807.
(data is representaive from n=2)
| |
| |
| |
| |
|
|
|
|
|
