Research Article: Nucleosome DNA unwrapping does not affect prototype foamy virus integration efficiency or site selection

Date Published: March 13, 2019

Publisher: Public Library of Science

Author(s): Randi M. Mackler, Nathan D. Jones, Anne M. Gardner, Miguel A. Lopez, Cecil J. Howard, Richard Fishel, Kristine E. Yoder, Michael Nevels.


Eukaryotic DNA binding proteins must access genomic DNA that is packaged into chromatin in vivo. During a productive infection, retroviral integrases (IN) must similarly interact with chromatin to integrate the viral cDNA genome. Here we examine the role of nucleosome DNA unwrapping in the retroviral integrase search for a target site. These studies utilized PFV intasomes that are comprised of a tetramer of PFV IN with two oligomers mimicking the viral cDNA ends. Modified recombinant human histones were used to generate nucleosomes with increased unwrapping rates at different DNA regions. These modifications included the acetylmimetic H3(K56Q) and the chemically engineered H4(K77ac, K79ac). While transcription factors and DNA damage sensors may search nucleosome bound DNA during transient unwrapping, PFV intasome mediated integration appears to be unaffected by increased nucleosome unwrapping. These studies suggest PFV intasomes do not utilize nucleosome unwrapping to search nucleosome targets.

Partial Text

Eukaryotic biology is dependent on proteins interacting with DNA in the context of chromatin. An enduring mystery in retrovirology is the criteria used by the viral protein integrase (IN) to choose an integration site in host chromatin. Integration of a reverse transcribed viral complementary DNA (cDNA) into the host genome is required for replication [1]. Integration site selection is not random and appears to be influenced by multiple factors including the port of nuclear entry, chromatin features, local DNA sequence and, in some cases, host cofactors for integration [2–8]. Over twenty years ago, human immunodeficiency virus 1 (HIV-1) and murine leukemia virus (MLV) integration were found to favor DNA wrapped in nucleosomes in vitro and in vivo [9–12]. Since those initial studies, host proteins that serve as integration cofactors have been identified, including LEDGF/p75 for HIV-1 and BET proteins for MLV [2, 7, 13, 14]. Although host cofactors of integration had not been described at the time of the earliest chromatin studies, several key observations of retroviral integration with nucleosomes were made. Not surprisingly, integration sites were overwhelmingly confined to the exposed DNA regions of nucleosomes [15, 16]. Combined with studies of intrinsically and physically bent DNA, these observations are consistent with the conclusion that bent or distorted DNA is a preferred target for retroviral integration [9, 10, 15].

We have used recombinant human histones with specific PTMs that increase NPS unwrapping to dissect the mechanism of PFV intasome target search. We also tested PFV integration to a naturally occurring NPS derived from HeLa cells, but found that under physiological conditions the D02 NPS is not stably positioned relative to the histone octamer. In contrast, the synthetic 601 NPS was designed to remain stable at or below physiologically relevant ionic concentrations [22]. Although the 601 NPS has enhanced stability compared to natural NPS sequences, 601 displays predictable nucleosome dynamics when present in murine hepatocytes in vivo [54]. As previously shown with other retroviral integrases, PFV intasomes showed a preference for exposed DNA helices and significantly distorted regions of the NPS DNA. We identified four major sites in 601 nucleosomes that exhibited a cluster of 2–5 integration events. These observations suggest a limited search that is associated with integration events at these particular exposed helices of nucleosome DNA. The use of specific histone acetylation PTMs or mimetics here suggests that increased unwrapping rate or decreased nucleosome stability have no effect on PFV integration efficiency or target site selection.