ABSTRACTWe have previously shown that the addition of the raltegravir/elvitegavir (RAL/EVG) primary resistance mutation N155H to the R263K dolutegravir (DTG) resistance mutation partially compensated for the fitness cost imposed by R263K while also slightly increasing DTG resistancein vitro(K. Anstett, T. Mesplede, M. Oliveira, V. Cutillas, and M. A. Wainberg, J Virol89:4681–4684, 2015, doi:10.1128/JVI.03485-14). Since many patients failing RAL/EVG are given DTG as part of rescue therapy, and given that the N155H substitution often is found in combination with other compensatory resistance mutations in such individuals, we investigated the effects of multiple such substitutions within integrase (IN) on each of integrase function, HIV-1 infectivity, and levels of drug resistance. To this end, each of the L74M, E92Q, T97A, E157Q, and G163R substitutions were introduced into NL4.3 subtype B HIV-1 vectors harboring N155H and R263K in tandem [termed NL4.3_{IN(N155H/R263K)}]. Relevant recombinant integrase enzymes also were expressed, and purified and biochemical assays of strand transfer efficiency as well as viral infectivity and drug resistance studies were performed. We found that the addition of T97A, E157Q, or G163R somewhat improved the affinity of IN_N155H/R263Kfor its target DNA substrate, while the presence of L74M or E92Q had a negative effect on this process. However, viral infectivity was significantly decreased from that of NL4.3_{IN(N155H/R263K)}after the addition of each tertiary mutation, and no increases in levels of DTG resistance were observed. This work shows that the compensatory mutations that evolve after N155H under continued DTG or RAL/EVG pressure in patients are unable to improve either enzyme efficiency or viral infectivity in an N155H/R263K background.IMPORTANCEIn contrast to other drugs, dolutegravir has not selected for resistance in HIV-positive individuals when used in first-line therapy. We had previously shown that HIV containing the primary raltegravir/elvitegravir resistance substitution N155H could select for R263K under dolutegravir pressure and that this virus was fit and displayed low-level resistance to dolutegravir (Anstett et al., J Virol89:4681–4684). Therefore, the current study aimed to uncover whether accessory mutations that appear after N155H in response to raltegravir/elvitegravir were compatible with N155H and R263K. We found, however, that the addition of a third mutation negatively impacted both the enzyme and the virus in terms of activity and infectivity without large shifts in integrase inhibitor resistance. Thus, it is unlikely that these substitutions would be selected under dolutegravir pressure. These data support the hypothesis that primary resistance against DTG cannot evolve through RAL/EVG resistance pathways and that the selection of R263K leads HIV into an evolutionary dead-end.