|
Advertisement | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
J. Biol. Chem., Vol. 277, Issue 30, 27489-27493, July 26, 2002
From the Aaron Diamond AIDS Research Center, The Rockefeller
University, New York, New York 10016
Received for publication, March 29, 2002, and in revised form, May 8, 2002
We have previously shown that human
immunodeficiency virus-1 (HIV-1) integrase is an unstable
protein and a substrate for the N-end rule degradation pathway.
This degradation pathway shares its ubiquitin-conjugating enzyme, Rad6,
with the post-replication/translesion DNA repair pathway. Because DNA
repair is thought to play an essential role in HIV-1 integration, we
investigated whether other molecules of this DNA repair pathway could
interact with integrase. We observed that co-expression of human Rad18
induced the accumulation of an otherwise unstable form of HIV-1
integrase. This accumulation occurred even though hRAD18 possesses a
RING finger domain, a structure that is generally associated with E3
ubiquitin ligase function and protein degradation. Evidence for an
interaction between integrase and hRad18 was obtained through
reciprocal co-immunoprecipitation. Moreover we found that a
162-residue region of hRad18 (amino acids 65-226) was sufficient for
both integrase stabilization and interaction. Finally, we observed that
HIV-1 integrase co-localized with hRad18 in nuclear structures in a
subpopulation of co-transfected cells. Taken together, these findings
identify hRad18 as a novel interacting partner of HIV-1 integrase and
suggest a role for post-replication/translesion DNA repair in the
retroviral integration process.
Integrase (IN)1 is
essential for the completion of the retroviral life cycle. This viral
protein performs different tasks that include cooperation in reverse
transcription (1), nuclear import of the pre-integration
complex2 (2, 3), HIV-1
particle production (4), and integration of the viral DNA into the host
genome (5, 6). In order to mediate these very diverse processes,
integrase is believed to associate with a number of cellular factors
and to exploit their functions. One such instance is the filling of the
gaps left after the ligation of 3'-ends of the viral DNA to the 5'-ends
of the staggered cleavage in cellular DNA. The factors involved are
thought to be proteins specialized in the maintenance of the genome
integrity. It has recently been reported that a particular DNA repair
mechanism, the non-homologous end joining (NHEJ) repair pathway, has a
major impact on integration frequency (7) and that some of its
components interact with the pre-integration complex (8). NHEJ is
thought to act on the repair of the gaps (7, 9, 10), although other
evidence suggests that it plays a role in the formation of
non-integrated HIV-1 DNA circles (8). Regardless, failure to accomplish
either of these processes results in apoptosis, which is observed in
NHEJ-deficient cells upon infection with HIV-1 (7, 8). Components of
another DNA repair pathway, the base excision repair (BER), have been
successfully tested in vitro for their ability to repair
retroviral integration-dependent gaps (11).
Poly(ADP-ribose) polymerase-1 (PARP-1), an enzyme activated by DNA
strand breaks, has also been linked to the retroviral integration
process (12, 13). Its proposed role in chromatin decondensation would
facilitate access of the repair machinery to the integration site
(13).
We recently reported that integrase is a substrate for the N-end rule
(14) a distinct ubiquitin-proteasome pathway that defines the half-life
of target proteins according to their N-terminal residue (15).
Authentic HIV-1 integrase is the proteolytic product of the
Gag-Pol polyprotein and specifies an N-terminal phenylalanine, which
renders it particularly prone to N-end rule degradation, and therefore
makes it very unstable. In general, degradation of proteins by the
proteasome proceeds after the substrate has been covalently modified by
linkage of a ubiquitin chain. This type of protein tagging is
sequentially performed by enzymes or clusters of enzymes broadly named
E1, E2, and E3 (16). In yeast, where the N-end rule was first described
(17), the E2 molecule is Rad6 (18). Rad6 is involved in a number of
apparently diverse processes, such as sporulation, retrotransposition,
N-end rule protein degradation, and DNA repair (19). Rad6 is able to
bind directly to the E3 molecule of the N-end rule, Ubr-1 (20), and also associates with Rad18 when involved in DNA repair (21). The DNA
repair pathway in which Rad6 and Rad18 participate is known as DNA
post-replication (PRR)/translesion repair (19, 22). One of the primary
roles of the PRR/translesion pathway is to overcome structural
restraints such as DNA adducts or UV-induced thymine-dimers that
prevent the chromosomal DNA replication from being completed (19). The
precise mechanism of action of the Rad6-Rad18 complex has not yet been
elucidated. However, it is widely assumed that the Rad6-Rad18
heterodimer converges to the site of blocked DNA synthesis by means of
the single strand DNA binding ability of Rad18 (23, 24). Subsequently,
the stalled replicative machinery is removed, possibly by
Rad6-dependent ubiquitin/proteasome-mediated degradation,
and is replaced by protein complexes that include molecules such as
Mms2-Ubc13 and Rad5 or repair polymerases such as Pol The putative involvement of Rad6 in the degradation of integrase and
the similarity between the repair of the retroviral gaps with the
proposed post-replication/translesion DNA repair model prompted us to
investigate whether PRR/translesion molecules are implicated in
integrase function. We demonstrate here that hRad18 interacts with
HIV-1 integrase and that it protects integrase from an accelerated
degradation. Moreover, hRad18 causes a re-localization of integrase in
a subset of transfected cells so that both proteins co-localize in the
same sub-nuclear structures.
Plasmids--
HIV-1 integrase pEGFP*IRES-Ub-X-IN-CTE expression
constructs have been described previously (14). Human RAD18 cDNA
was generated by reverse transcription (RT)-PCR, using
GenBankTM accession number AB035274 as the reference
sequence, and was cloned into the mammalian expression vector pXJFLAG
(25), which includes an N-terminal FLAG tag for protein detection
purposes. Human RAD18 mutants were obtained by PCR-mediated deletions
of the wild-type full-length cDNA and cloned in pXJFLAG by
conventional molecular biology techniques. All DNA fragments obtained
by PCR and used for cloning purposes were sequenced by the Rockefeller University Protein and DNA Technology Center. For the numbering of
mutants, see Fig. 4A.
Western Blot and Immunoprecipitation--
Human embryonic kidney
(HEK) 293T cells, seeded at 2.5 × 105 cells/well in
6-well plates were transfected with 1 µg of each plasmid (a total of
2 µg/well DNA in case of co-transfection experiments) using the
DNA/calcium-phosphate co-precipitation procedure. 44 h after
transfection, cells were lysed in 200 µl/well lysis buffer (50 mM Tris-HCl, pH 7.3, 150 mM NaCl, 5 mM EDTA, 0.2% Triton X-100, and CompleteTM
protease inhibitors (Roche, Indianapolis, IN)). After a 30-min incubation at 4 °C, lysates were sonicated with a Sonic Dismembrator 550 (Fisher Scientific) and cleared by 10 min of centrifugation at
20,000 relative centrifugal force at 4 °C. For Western blot analysis, equal amounts of total protein were separated by
electrophoresis on polyacrylamide Nupage gels (Invitrogen), then
transferred to polyvinylidene difluoride membranes (Amersham
Biosciences). Membranes were incubated with 1 µg/ml anti-HIV-1
integrase mouse mAb12 for integrase detection (kindly provided by Dr.
A. M. Skalka, Fox Chase Cancer Center, Philadelphia, PA) (26) and with
1 µg/ml of anti-FLAG mouse mAb M2 for FLAG-hRad18. As a secondary
antibody, a sheep anti-mouse horseradish peroxidase-conjugated
antibody (Amersham Biosciences) was utilized. Proteins were detected
with the ECL-Plus chemiluminescence system (Amersham Biosciences). For
co-immunoprecipitation, cleared cell lysates from pooled wells were
incubated with either 1 µg/ml anti-HIV-1 integrase mouse mAb12 or
with 1 µg/ml anti-FLAG mouse mAb M2 (Sigma) at 4 °C for 1.5 h. Antibody-protein complexes were precipitated with protein A-Sepharose (Amersham Biosciences) and resolved on a 10%
polyacrylamide gel. Immunoblotting was then performed as described
above, using either anti-integrase mAb12 or anti-FLAG mAb M2 as primary antibodies.
Immunofluorescence--
For subcellular localization studies of
hRad18, cells were plated and transfected on human fibronectin
(Invitrogen) coated coverslips. Cells were first fixed for 30 min on ice with phosphate-buffered saline with 3% formaldehyde,
diluted from a 20% electron microscopy (EM) grade stock
solution (Tousimis, Rockville, MD), and then permeabilized with 3%
formaldehyde, 0.2% Triton X-100 in phosphate-buffered saline for 15 min on ice. For co-localization purposes, transfected cells were
permeabilized with CSK buffer (100 mM KCl, 300 mM sucrose, 10 mM PIPES, pH 6.8, 3 mM MgCl2, 0.5% Triton X-100 and
CompleteTM protease inhibitors) for 10 min on ice and then
fixed with phosphate-buffered saline, 3% formaldehyde for 20 min on
ice. Cells treated in either way were probed for the presence of HIV-1
integrase and FLAG-tagged hRad18 with anti-integrase mAb 6G5 (kindly
provided by Dr. D. E. Helland, University of Bergen, Norway) (27),
1:500 dilution of culture supernatant, and anti-FLAG (Sigma) rabbit
polyclonal antibody (5 µg/ml), respectively. The fluorescent
secondary antibodies used were a goat anti-mouse Alexa Fluor® 647 (Molecular Probes, Eugene, OR) at 10 µg/ml and a
fluorescein-conjugated goat anti-mouse, (Vector Laboratories,
Burlingame, CA) at 1:500 dilution. DNA was stained with 0.5 µM Hoechst 33258. Images were obtained and deconvolved through DeltaVision restoration microscopy system using Softworx software (Applied Precision, Issaquah, Washington).
An in silico amino acid sequence analysis (28) revealed
that hRad18 contains a specific type of zinc finger motif known as RING
finger. This domain has been proposed to be a distinguishing feature of
molecules with E3 ubiquitin ligase activity (29-32). The RING finger
domain is also required for the degradative activity of Ubr1, the N-end
rule E3 ligase in yeast (33).
Unexpectedly, Western blot analysis showed that three different forms
of integrase, two inherently unstable (Phe-IN and Arg-IN) (14), as well
as a stable form (Met-IN) significantly accumulated in the presence of
co-transfected hRad18 (Fig.
1B). This indicated that
hRad18 did not increase degradation of integrase. Rather, this
experiment showed that hRad18-mediated stabilization of integrase was
largely independent from the N-end rule, since all three forms of
integrase were similarly affected. No variation in the amounts of EGFP
were detected in the presence or absence of hRad18 (data not shown)
ruling out a transcriptional effect on the integrase expression vectors
as the cause of the accumulation.
To determine whether the hRad18-induced accumulation of integrase might
depend on an interaction between the two molecules, an unstable form of
integrase (Phe-IN) was transfected into HEK 293T cells together with
FLAG-hRad18 or with an empty vector as control. Following lysis and
sonication protein complexes were precipitated either with an
anti-integrase or with an anti-FLAG antibody and analyzed by reciprocal
immunoblotting. Integrase could be precipitated with an anti-FLAG
antibody and hRad18 with an anti-integrase antibody (Fig.
2), which demonstrated an interaction between the two proteins.
Human Rad18 has been localized to the nucleus (34), which is compatible
with its DNA repair function. Indeed, when hRad18 was transfected in
HEK 293T cells it was strictly confined to the nuclear compartment
(Fig. 3A). However, hRad18
distribution was not always homogeneous, and we consistently observed
that it displayed a structured nuclear pattern in a subpopulation of cells. These different distributions conceivably depend on the different physiologic states of the cells.
When expressed alone, integrase usually exhibited a fine punctate nuclear pattern in 293T cells. However, further immunofluorescence experiments revealed that in a subset of cells co-transfected with hRad18, integrase re-localized within the nucleus to larger structures that coincided with the distribution of hRad18 (Fig. 3B). The permeabilization step used before fixation allowed us to visualize only the complexes that were detergent-resistant and thus more likely to have a relevant role in vivo. To determine which region of hRad18 was responsible for stabilization
of integrase, we performed a deletion mutagenesis analysis. Human Rad18
has at least three putative functional domains, a RING finger, a zinc
finger, and a SAP domain (Fig.
4A). We constructed a total of
eight deletion mutants (Fig. 4B), each of them harboring an
N-terminal FLAG tag for antibody recognition. Experiments were carried
out by co-transfecting each of the deletion mutants of hRad18 with the
physiologic, unstable form of integrase (Phe-IN). Cell lysates were
analyzed by Western blotting for stabilization of integrase and by
co-immunoprecipitation to detect binding. All of the hRad18 mutants,
except for mutant 1-76, were expressed at levels comparable to that of
wild-type hRad18 (Fig. 4B, lower panel). Fig.
4B shows that a number of the mutants were able to stabilize
integrase to various degrees (upper panel), but that only
one of these mutants (65-226) exhibited a detectable level of binding
after our immunoprecipitation protocol (middle panel). The
portion of hRad18 encoded by this mutant encompasses the region C-terminal of the RING finger and the putative zinc finger.
Interestingly, longer exposure times revealed a minor accumulation of
integrase also in two mutants lacking the zinc finger motif, namely
239-495 and to a lesser degree 285-495 (data not shown). This
stabilization, though less in magnitude, could be evidence for a second
region of interaction with integrase positioned in the C terminus of hRad18. This observation could in part explain why the interaction between integrase and mutant 65-226 is not as effective as with full-length hRad18.
Weaker interactions due to steric interference with improperly folded
domains could be the reason why integrase did not co-precipitate with
mutants encoding fragments larger than 65-226. Taken together these
results show that a relatively small region within hRad18 encompassing amino acids 65-226 is sufficient for both accumulation of
and binding to integrase.
Repair of host chromosomal DNA at the insertion site is pivotal for productive retroviral integration. In this work we tested the hypothesis that components of the DNA postreplication/translesion repair pathway could play a role in HIV-1 integrase biology. We describe the interaction between HIV-1 integrase and hRad 18, whose Saccharomyces cerevisae homologue is an essential member of the PRR/translesion pathway. This interaction results in an increased stabilization of integrase, which in its natural form is a particularly unstable protein (14). In addition to the accumulation of integrase, we observed through reciprocal co-immunoprecipitation an association of hRad18 with integrase. These results suggest that the stability of integrase is tightly regulated by its interaction with hRad18. A parallel can be drawn between the possible role of hRad18 in HIV-1 integration and that of the protein MuB in the transposition of bacteriophage Mu. Indeed, Mu transposition is a prototype model for retroviral integration. The Mu bacteriophage requires multiple transpositions into the bacterial chromosome to replicate. This complex reaction is orchestrated by its transposase, MuA, in concert with both host and viral factors. After recombination is completed, the nucleoprotein complex (transpososome) is remodeled by the bacterial chaperone molecule ClpX (35, 36) that specifically recognizes MuA and mediates its unfolding and eventually its destabilization (37). The susceptibility of HIV-1 integrase to rapid degradation by the ubiquitin/proteasome pathway N-end rule (14) and a recent report indicating that in vitro an excess of integrase inhibits closure of the DNA single strand intervals (11) seems to suggest that indeed the integration complex (intasome) needs to be disassembled for integration to be completed. The phage protein MuB has been found to exert part of its function before the destabilization of the transpososome. MuB has been reported to stabilize the transposition complex by binding the transposase MuA in a region that overlaps with the binding site recognized by ClpX for disassembly and remodeling (38). Therefore, it is suggested that MuB has a modulatory activity, allowing the remodeling of the transpososome to proceed exclusively during the transition to the replicative stage (38). By analogy we propose that hRad18 may regulate the stability of integrase to guarantee a productive insertion of the viral genome into the host DNA. A ubiquitin ligase function of hRad18 favoring the degradation of integrase cannot be excluded, and one could interpret the stabilization effect as a titration of the endogenous factors necessary for an efficient proteolytic degradation due to over-expression of hRad18 itself. However, the possibility that hRad18 titrates the human homologues of Rad6 is unlikely since it has been reported that the RING finger of hRad18 is necessary for hRad6A/B-hRad18 binding (34). Our deletion analysis showed that the RING finger motif does not play an essential role in the interaction between hRad18 and HIV-1 integrase, ruling out a direct competition between HIV-1 integrase and hRad6A/B for hRad18 binding. Indeed, the 162-amino acid peptide that encompasses a putative zinc finger seems to be sufficient to both stabilize and associate with integrase. The re-localization of integrase and its co-localization with hRad18 in a subset of cells suggests an additional function for this association. Human Rad18 contains a putative SAP-box (39), a domain recently recognized to mediate the binding of certain proteins to specific A/T-rich DNA regions known as the scaffold attachment regions (SAR) (40). Interestingly, PARP-1, Ku antigens, and HMG-I/Y, which are involved in retroviral integration (7, 12, 13, 41), have all been found to be SAR-binding proteins (42-44). An intriguing possibility is that the molecules relevant for HIV-1 integration cluster together, perhaps in the vicinity of SARs, achieving in this way the coordination required for these complex reactions. A role for the molecules of the PRR/translesion pathway in retroviral
integration does not rule out other DNA repair mechanisms such as NHEJ
or BER (7, 11). Examples for an overlap between DNA repair pathways
that can function on the same damaged site have been recently reported
in yeast. For instance, the activity of the Srs2 helicase is the main
switch for the interaction between RAD6-RAD18 and the RAD52 epistasis
(19). Thus, it is conceivable that different DNA repair pathways
complement each other to complete the sequence of events that leads to
a productive HIV-1 integration.
We thank Cecilia Cheng-Mayer, David Ho, and Paul Bieniasz for insightful discussions, Inês Chen and Viviana Simon for critically reading the manuscript, Amanda Brown for plasmids, Yen Shing Ng for technical assistance, Gregor Cardué for help with graphics, and Anna Marie Skalka and Dag E. Helland for monoclonal antibodies.
* This work was supported in part by Columbia-Rockefeller Center for AIDS Research Grant P30 AI 42848 (to L. C. F. M.) from the NIAID, National Institutes of Health, by internal funding from the Aaron Diamond AIDS Research Center, and National Institutes of Health Grant 1R01 AI47054-01 (to M. A. M).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
§ To whom correspondence should be addressed. Tel.: 212-448-5060; Fax: 212-448-5159; E-mail: mmuesing@adarc.org.
Published, JBC Papers in Press, May 16, 2002, DOI 10.1074/jbc.M203061200
2 D. R. Kaufman and M. A. Muesing, submitted manuscript.
The abbreviations used are: IN, integrase; PRR, post-replication DNA repair; HIV-1, human immunodeficiency virus-1; NHEJ, non-homologous end joining; BER, base excision repair; mAb, monoclonal antibody; HEK, human embryonic kidney; PIPES, piperazine-N,N'-bis(2-ethanesulfonic acid); IRES, internal ribosome entry sequence; Ub, ubiquitin; CTE, constitutive transport element; pol, polymerase; EGFP, enhanced green fluorescent protein.
Copyright © 2002 by The American Society for Biochemistry and Molecular Biology, Inc. This article has been cited by other articles:
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Advertisement | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||