Insertion as a Resistance Mechanism Against Integrase Inhibitors in Several Retroviruses

To the Editor—We read with great interest a recent report by Le Hingrat and colleagues [1] of a new mechanism of human immunodeficiency virus (HIV)-2 resistance against integrase inhibitors through the development of 5 amino acid insertions within integrase coding sequences. Integrase inhibitors are currently the leading class for antiretroviral treatment initiation [2]. Insertions in the integrase of HIV-2 conferred various levels of resistance against integrase strand transfer inhibitors [1]. Insertions had been previously described in the context of HIV-1 drug resistance against protease and reverse transcriptase, but not integrase inhibitors [3]. To the best of our knowledge, there has not been any report of integrase insertion in HIV-1, which suggests that this type of genetic alteration is specific to HIV-2. Altogether, the findings by Le Hingrat and colleagues are very significant.

Intriguingly, this paper was not the first to report insertions in the integrase coding sequence of retroviruses (Table 1). The first report was made in a Simian Immunodeficiency Virus (rhesus macaque) strain 251 (SIVmac251)-infected rhesus macaque unsuccessfully treated with a long-acting nano-suspension of cabotegravir, an integrase strand transfer inhibitor currently in advanced phases of development [4]. The insertion consisted of 5 amino acids after position 232 in the integrase of SIVmac251, and could not be introduced in either SIVmac239 or HIV-1 without severely impairing fitness [4]. More recently, we also reported the frequent emergence of various 5 or 6 amino acid insertions in the SIVmac251 integrase in rhesus macaques treated with dolutegravir monotherapy [5]. Integrase insertions in this setting were heterogeneous in sequence, but all contained 2 positively charged amino acids, and all were found in the immediate vicinity of arginine (R) 231, a residue that was shown by cryogenic electron microscopy (cryo-EM) to be the only one outside of integrase catalytic domain to interact with host target DNA strongly [6]. This specific residue is also involved in interactions with viral DNA. Based on structural data [6], it is thus expected that insertions that extend the length of the loop carrying residue R231, while at the same time increasing the local, positive charge, may bestow flexibility in DNA binding that could confer resistance against integrase strand transfer inhibitors. In agreement with this hypothesis, all insertions reported to date also contained the small amino acid glycine (G), which could contribute to loop flexibility. Importantly, in our study, insertions could be found in circulating strains as well as proviruses, indicating that both replicative capacity and integration were conserved [5]. The replicative competence of retroviruses carrying integrase insertions is now confirmed by Le Hingrat et al’s report of evolution from 2 to 5 amino acid insertions in 1 patient [1].

Table 1.

Summary of Integrase Insertions Associated With Resistance Against Integrase Strand Transfer Inhibitors

Amino Acid Insertion	Position (-1)	Organism	Reference
GK	231	HIV-2	Le Hingrat et al [1]
SREGK/K
YREGR
GYKGK
GYRGR
YREGR	232	SIVmac251	Andrews et al [4]
EGRGIK	228	SIVmac251	Van Rompay et al [5]
EGRYK	228
SREGK	234
EGRYK	235
GIKGR	235

Positively charged amino acids (at neutral pH) are indicated in bold.

Abbreviations: HIV, human immunodeficiency virus; SIVmac251, simian immunodeficiency virus (rhesus macaque) strain 251.

Finally, phylogenetic studies have concluded that HIV-2 is very closely related to the SIV strains found in Sooty mangabeys [7]. This genetic proximity may help to explain the evolutionary convergence in resistance via insertion between HIV-2 and SIVmac. For structural reasons [6], HIV-1 is less likely to become resistant via this pathway. However, Le Hingrat et al’s results indicate that patients living with HIV-2 may need special clinical attention when using integrase inhibitors. Future studies will have to investigate the potential for the transmission of such resistant strains, a question that could be critical given the relatively high prevalence of insertions in Le Hingrat et al’s report, compared to reverse transcriptase (RT) and protease (PR) insertions [3].

Note

Potential conflicts of interest. All authors: No reported conflicts of interest. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.