Register for Mechanobiology 2016!

Transcriptional co-activator p75 binds and tethers the Myc-interacting protein JPO2 to chromatin
G. N. Maertens, P. Cherepanov, A. Engelman


Transcriptional co-activator p75 is implicated in human cancer, autoimmunity and replication of human immunodeficiency virus type 1 (HIV-1) as a dominant integrase-interacting protein. Although characterized as chromatin associated, the normal biological role(s) of p75 remains fairly unclear. To gain insight into p75 function, we have characterized its cellular binding partners and report that JPO2, a recently identified Myc-binding protein, associates with p75 in vitro and in vivo. The pseudo HEAT repeat analogous topology (PHAT) domain of p75, which mediates its interaction with integrase, also mediates the interaction with JPO2, and recombinant integrase protein competes with JPO2 protein for binding to p75 in vitro. JPO2 binds p75 through a 61-residue (amino acids 58-119) region that is distinct from its Myc-interacting domain. In cells, JPO2 and p75 co-localize throughout the cell cycle, and both proteins concentrate on condensed chromosomes during mitosis. Strikingly, the association of JPO2 with chromatin strictly depends upon p75, similar to that of ectopically expressed integrase. Also similar to its effect on integrase, p75 stabilizes intracellular steady-state levels of JPO2 protein. Our results suggest a role for p75 in the Myc regulatory network, and indicate that p75 is a general adaptor protein tethering divergent factors to chromatin through its versatile integrase-binding domain.


Transcriptional co-activator p75 (also referred to as lens epithelium-derived growth factor, LEDGF) and its mRNA splice variant, p52, were originally identified through co-fractionation with the transcriptional co-activator PC4 from HeLa cell extracts (Ge et al., 1998a). In vitro, p75 and p52 each displayed transcriptional co-activator activity (Ge et al., 1998a). p75 is a nuclear protein that displays a characteristic heterogeneous distribution pattern in interphase cells and associates intimately with condensed chromosomes during mitosis (Cherepanov et al., 2003; Maertens et al., 2003; Nishizawa et al., 2001; Ochs et al., 2000). p75 is the main interaction partner of human immunodeficiency virus type 1 (HIV-1) integrase (IN) in human cells (Cherepanov et al., 2003; Maertens et al., 2003; Turlure et al., 2004). Residues 347-429 constitute the integrase-binding domain (IBD) of p75 (Cherepanov et al., 2004; Emiliani et al., 2005; Vanegas et al., 2005). The three-dimensional structure of the IBD was solved by nuclear magnetic resonance (NMR) spectroscopy (Cherepanov et al., 2005b) and as part of a complex with the catalytic core domain (CCD) of HIV-1 IN (Cherepanov et al., 2005a). The IBD, which is primarily α-helical, contains a pair of HEAT repeat-like elements and was therefore characterized as a pseudo HEAT repeat analogous topology (PHAT) domain (Cherepanov et al., 2005b).

When expressed in human cells, HIV-1 IN accumulates in nuclei (Petit et al., 1999; Pluymers et al., 1999) and binds tightly to chromosomes during mitosis (Cherepanov et al., 2000; Devroe et al., 2003). Strikingly, these properties strictly depend on the interaction with endogenous p75 protein. Thus, transient or stable knockdown of p75 resulted in the re-localization of HIV-1 IN to the cytoplasm and caused IN to dissociate from chromatin (Llano et al., 2004b; Maertens et al., 2003). Concordantly, a mutant of HIV-1 IN defective for binding to p75 did not associate with chromatin (Emiliani et al., 2005). Through its interaction with HIV-1 IN, p75 influences the intracellular dynamic properties of the viral protein (Maertens et al., 2005) and protects it from degradation through the ubiquitin-proteasome pathway (Llano et al., 2004a). Furthermore, knockdown of p75 resulted in significant reduction of HIV-1 integration into transcription units (Ciuffi et al., 2005), suggesting that the co-factor directly tethers lentiviral pre-integration complexes to active genes during integration (Bushman et al., 2005; Engelman, 2005; Van Maele et al., 2006).

In addition to its proposed role in viral replication, p75 affects normal cellular and physiological processes, and was reported to protect cells from apoptosis under oxidative stress (Fatma et al., 2001; Matsui et al., 2001; Sharma et al., 2000; Shinohara et al., 2002; Singh et al., 2001). Autoantibodies against p75 were detected in patients with certain autoimmune diseases, and increased expression levels of p75 were observed in patients suffering from prostate cancer (Daniels et al., 2005; Dellavance et al., 2005; Ganapathy et al., 2003; Ogawa et al., 2004; Okamoto et al., 2004). Additionally, NUP98-LEDGF fusions occur in patients with acute and chronic myeloid leukemia (Ahuja et al., 2000; Grand et al., 2005; Hussey et al., 2001; Morerio et al., 2005).

Although p75 is implicated in a variety of human ailments, the molecular mechanisms of p75 function(s) are for the most part unknown. Mechanistic insight can be gleaned by determining the `interactome' of a given protein (Rual et al., 2005), and in this study we sought to determine the intracellular binding partners of human p75. Our results identify JPO2, also known as RAM2, which was recently reported to repress transcription of the gene encoding monoamine oxidase (MAO) A through binding to Sp1 sites in the core promoter (Chen et al., 2005). JPO2, independently isolated from a yeast two-hybrid screen for Myc interactors, was shown to have transforming activity (Huang et al., 2005). The protein is highly similar to JPO1, another Myc-interacting protein (Huang et al., 2005) encoded by a Myc responsive gene (Prescott et al., 2001). A 454-residue protein, JPO2, contains a PEST region within its N-terminal region (residues 29-54), a C-terminal RING-finger-like zinc-binding motif (residues 349-425), a putative leucine zipper (LZ; residues 213-235), and a putative nuclear localization signal (NLS; residues 301-318) (Chen et al., 2005). We characterized the interaction between JPO2 and p75, and demonstrate that p75 dictates the intracellular distribution of JPO2. On the basis of these results, we suggest that p75 is a general tethering factor involved in linking HIV-1 IN, endogenous JPO2 and perhaps other members of the transcription machinery to chromatin.


Identification of JPO2 as a novel p75-interacting protein

To identify cellular binding partners of p75, HEK 293T (commonly known as 293T) cells were engineered to stably express a hemagglutinin (HA)-tagged form of the protein. Preliminary experiments revealed that the resulting 293T-HAp75 cell line expressed HA-p75 at a level similar to that of endogenous p75 (data not shown). Nuclear extracts prepared from 293T-HAp75 and parental 293T cells were incubated with HA-affinity resin. Proteins eluted from extensively washed beads were analyzed by SDS-PAGE and staining with silver. Protein bands excised from gels were digested with trypsin, and the resulting peptides were identified by liquid chromatography-tandem mass spectrometry (LC-MS/MS) sequencing. Two criteria were established to exclude nonspecific p75-interacting proteins, such that only the following proteins were taken into account (1) peptides originating from 293T-HAp75 samples that were completely absent from control 293T cell samples and (2) proteins yielding more than two unambiguous peptides. Although several dozen proteins were identified, only two satisfied these criteria. One, which migrated at about 68 kDa, was a proteolytic fragment of p75 that produced 15 specific peptides. The second protein, the putative transcription factor JPO2 (gene symbol CDCA7L; UniGene ID Hs.520245), was found in the same gel slice and gave a total of five peptides covering approximately 14% of the protein sequence (Table 1). Persuasively, JPO2-derived peptides were detected in two independent immunoprecipitation experiments.

View this table:
Table 1.

JPO2 tryptic peptides identified following IP and LC-MS/MS sequencing

Reciprocal co-immunoprecipitation experiments were performed to confirm that JPO2 is a bona fide interactor of p75. A Flag-tagged version of JPO2 was initially used to facilitate these analyses. Flag-tagged Mob2, a known interactor of serine/threonine kinases NDR1 and NDR2 (Devroe et al., 2004), served as a negative control. Flag-JPO2 was co-immunoprecipitated with HA-p75 using anti-HA antibody, and vice versa (Fig. 1A). Importantly, Flag-Mob2 was not co-immunoprecipitated by the anti-HA antibody, and HA-p75 was not recovered from Flag-Mob2 containing cell extracts using the anti-Flag antibody (Fig. 1A). These results establish that the noted co-immunoprecipitates were independent of the utilized affinity tags. Furthermore, Flag-JPO2 was recovered with endogenous p75 in immunoprecipitations using monoclonal anti-p75/p52 or anti-Flag antibodies (Fig. 1B). To enable the analysis of endogenous protein, rabbits were inoculated with a peptide derived from the C-terminal region of JPO2. Although resultant sera failed to immunoprecipitate JPO2, the endogenous protein was detectable by western blotting. Importantly, endogenous JPO2 was co-immunoprecipitated with endogenous p75 using monoclonal anti-p75 or p52 antibody (Fig. 1C).

We noted that HA-p75 dramatically increased the steady-state level of Flag-JPO2 protein in co-transfected cells. When 293T cells were transfected with 1 μg of the Flag-JPO2 expression vector and various amounts of pBHA-p75 (0-3 μg; the total amount of transfected DNA was kept constant), Flag-JPO2 protein levels clearly paralleled those of HA-p75 (Fig. 1D, lanes 1-4). Importantly, Flag-Mob protein levels were insensitive to the level of co-expressed HA-p75 (Fig. 1D, lanes 5-8). Since Flag-JPO2 and Flag-Mob2 were each expressed from the cytomegalovirus immediate early promoter, the observed increase in Flag-JPO2 levels was not caused by an effect of p75 on transcription. This result further corroborates the direct interaction between p75 and JPO2, and strongly indicates that p75 increases the stability of JPO2 through direct protein binding.

p75 is a nuclear protein that displays a characteristic heterogeneous distribution pattern in interphase, and concentrates on condensed chromosomes during mitosis (Cherepanov et al., 2003; Maertens et al., 2003; Nishizawa et al., 2001; Ochs et al., 2000). To explore the interaction between p75 and JPO2 further, the cellular distribution of ectopically expressed protein was visualized by laser-scanning microscopy. The intracellular distribution of HcRed1-JPO2 (data not shown) or EGFP-tagged JPO2 (Fig. 2A,B) was highly reminiscent of that of p75. When co-expressed, EGFP-JPO2 and HcRed1-p75 co-localized in cell nuclei (Fig. 2B). To evaluate the specificity of the interaction further, two different p75 loss-of-function mutants were used. Mutation of Lys150 to Ala disrupts the nuclear localization signal (NLS) of the protein and precludes active nuclear import (Maertens et al., 2004). Thus, ectopically expressed EGFP-GST-p75K150A fusion protein was excluded from cell nuclei at 24 hours post-transfection (Maertens et al., 2004). Strikingly, co-expression of EGFP-GST-p75K150A mislocalized HcRed1-JPO2 to the cytoplasm (Fig. 2C), both highlighting the protein-protein interaction and suggesting that p75 might be involved in the nuclear import of JPO2. The isolated C-terminal domain of p75 (residues 326-530) lacks the characteristic nuclear distribution of wild-type p75 (Maertens et al., 2004). As HcRed1-JPO2 co-localized with EGFP-p75326-530 (Fig. 2D), we inferred that JPO2 interacts with the C-terminal region of p75.

Fig. 1.

Transcriptional co-activator p75 interacts with JPO2. (A) Reciprocal co-immunoprecipitations of Flag-JPO2 and HA-p75. 293T cells co-transfected with HA-p75 and Flag-JPO2 or Flag-Mob2 expression constructs were lysed 24 hours post-transfection. Proteins immunoprecipitated (IP) with anti-Flag (left panel) or anti-HA (right panel) antibodies were detected by western blotting (WB). The migration positions of molecular mass standards are indicated to the left of each panel. (B) Flag-JPO2 complexes with endogenous p75. Lysates of 293T cells transfected with Flag-JPO2 were immunoprecipitated with anti-Flag or anti-p75/p52 antibodies. (C) Interaction between endogenous JPO2 and p75 proteins. Extracts of non-transfected 293T cells were immunoprecipitated with monoclonal anti-p75/p52 antibody, and JPO2 was detected using rabbit anti-JPO2 antibodies. The asterisk indicates a nonspecific protein detected during western blotting. (D) Overexpression of p75 post-transcriptionally augments the steady-state level of Flag-JPO2 protein. 293T cells were transfected with variable amounts of pBHA-p75 (lanes 2 and 6, 1 μg; lanes 3 and 7, 2 μg; lanes 4 and 8, 3 μg) with Flag-JPO2 (lanes 1-4) or Flag-Mob2 (lanes 5-8). Lanes 1 and 5, pBHA-p75 was omitted from transfection. The total amount of DNA in each transfection was adjusted to 4 μg using pcDNA6.V5HisB (Invitrogen). Cells were harvested 24 hours post-transfection, and 15 μg of total cellular protein was analyzed by SDS-PAGE and western blotting.

Fig. 2.

Co-localization of JPO2 with wild-type and NLS-deficient p75. (A) EGFP-JPO2 accumulates in nuclei with a heterogeneous distribution pattern. (B) EGFP-JPO2 and HcRed1-p75 co-localize in cell nuclei. (C) EGFP-GST-p75K150A (K150A) sequesters HcRed1-JPO2 to the cell cytoplasm. (D) Co-localization of EGFP-p75326-530 (p75/Ct) and HcRed1-JPO2. Images were acquired 24 hours post-transfection. Bars, 10 μm.

Fig. 3.

HIV-1 IN and JPO2 compete for binding to the p75 IBD. (A) The p75 IBD is necessary and sufficient for the interaction with JPO2. Lanes 1 and 2; input levels of BSA and 35S-labeled JPO2, respectively. Lanes 3-9: GST, GST-p75 or the indicated p75 deletion mutant pre-bound to GS beads was incubated with 35S-labeled JPO2. BSA (10 μg) was included in each reaction as an internal specificity control. Bound proteins separated by SDS-PAGE were stained with Coomassie Blue (top panel); JPO2 was detected by autoradiography (bottom panel). (B) HIV-1 IN competes with JPO2 for binding to p75 IBD. Lanes 1-4: input BSA, IN-His6, p75 and 35S-labeled JPO2, respectively. Lanes 5-10: His6-tagged HIV-1 IN was incubated with p75 and/or JPO2 in the presence of Ni-NTA beads and BSA, and bound proteins fractionated by SDS-PAGE were detected by staining with Coomassie Blue (top panel) or autoradiography (bottom panel). Migration positions of molecular mass markers, IN-His6, p75, BSA and JPO2 are indicated.

The p75 IBD mediates the interaction with JPO2

To characterize the interaction further, recombinant p75 and JPO2 proteins were analyzed using in vitro pull-down assays. Attempts to express recombinant JPO2 in bacteria failed, presumably as a result of extreme protein instability under these conditions. Thus, recombinant JPO2 protein was synthesized using in vitro transcription and translation. Full-length GST-p75 efficiently recovered in-vitro-translated 35S-labeled JPO2 protein from solution (Fig. 3A, lane 2). By contrast, GST displayed no affinity for JPO2 under these conditions (Fig. 3A, lane 3). To map the region of p75 involved in the interaction, deletion mutant proteins were analyzed in the pull-down assay. The C-terminal portion of p75 (GST-p75326-530; Fig. 3A, lane 6), but not its N-terminal 325 residues (GST-p751-325, lane 5) quantitatively recovered JPO2, confirming that the determinant for the interaction with JPO2 is located within the 205 C-terminal residues of p75. Furthermore, residues 347-429, previously identified as the minimal region of p75 able to pull-down HIV-1 IN (Cherepanov et al., 2004), efficiently recovered JPO2 (Fig. 3A, lane 8). The N-terminal α helix of the IBD comprises p75 residues 348-362 (Cherepanov et al., 2005b), and removing IBD residues 347-365 was previously shown to disrupt its interaction with HIV-1 IN (Cherepanov et al., 2004). Strikingly, this deletion likewise ablated the IBD-JPO2 interaction (Fig. 3A, lane 9). We therefore conclude that the intact p75 IBD is necessary and sufficient for the interaction with JPO2 protein in vitro.

We next tested whether JPO2, HIV-1 IN and p75 formed a ternary complex in vitro. As previously demonstrated (Cherepanov et al., 2005b; Maertens et al., 2003), His6-tagged IN efficiently pulled-down non-tagged p75 (Fig. 3B, lane 5). Although the addition of in-vitro-translated JPO2 did not interfere with the IN-p75 interaction, JPO2 was not recovered under these conditions (Fig. 3B, lane 6). These results suggest that HIV-1 IN and JPO2 bind to the p75 IBD in a mutually exclusive manner.

We further investigated which residues within p75 mediate the interaction with JPO2. Initially, a set of 27 point mutants of p75347-471, previously evaluated for their interaction with HIV-1 IN (Cherepanov et al., 2005b), were tested for their ability to pull-down 35S-labeled JPO2. Under these conditions, mutants K360A, I365A, D366N, D369A, V370A and F406A displayed reduced affinity for JPO2 (data not shown). These mutations were next incorporated into full-length HA-p75 and the resulting proteins were tested for interaction with Flag-JPO2 in co-transfected 293T cells (Fig. S1A, supplementary material). The E379A mutation, which did not disrupt the JPO2-p75347-471 interaction in vitro, was included as a control. Although none of the tested point mutations completely abrogated the interaction, substituting Ala for Phe406, Val370, Ile365 or Lys360 reduced the apparent affinity. Notably, these residues are located at or near the interhelical loop regions of the IBD (Fig. S1B, supplementary material), further implicating the IN-binding surface of p75 in the interaction with JPO2.

Characterization of the p75-binding domain in JPO2

To determine the region of JPO2 that interacts with p75, JPO2 deletion mutants generated by in vitro transcription and translation were tested for binding to the p75 IBD by GST pull-down. Full-length, wild-type JPO2 was specifically and quantitatively pulled-down by GST-p75347-471 (Fig. 4A, compare lanes 1, 7 and 13). Removal of up to 322 C-terminal amino acids from JPO2 did not affect its binding to p75 (Fig. 4A, lanes 2-6, 8-12, 23, 24, 29 and 30). By contrast, truncating 76 N-terminal residues from JPO2 reduced binding (Fig. 4A, compare lanes 16 and 20), and removal of 21 additional N-terminal residues obliterated the interaction (Fig. 4A, lanes 17 and 21). Thus, JPO2 residues 77-98 play an important role in its interaction with p75. Further analyses revealed that a 61-residue fragment encompassing JPO2 amino acids 58-119 was sufficient for the interaction with p75 (Fig. 4A, lanes 27 and 33). In vitro transcription and translation failed to yield shorter fragment derivatives, presumably owing to protein folding and/or stability issues, precluding analyses of JPO258-119 sub-fragments by GST pull-down. Of note, neither full-length JPO2 nor any of the JPO2 deletion mutants were recovered when using GST in place of GST-p75347-471 (lane 13 and data not shown), confirming that the results of these GST pull-downs reflected a specific interaction between the p75 IBD and JPO2.

Fig. 4.

Identification of the p75-binding domain in JPO2. (A) JPO2 (wt) or its deletion mutants (indicated) produced by in vitro transcription and translation were incubated with GST or GST-p75347-471 pre-bound to GS beads in the presence of BSA. Eluted proteins fractionated by SDS-PAGE were visualized by autoradiography (top panels) or staining with Coomassie Blue (bottom panels). Migration positions of molecular mass markers, GST and GST-p75347-471are indicated. None of the JPO2 deletions mutants bound to GST alone (data not shown). (B) Schematic representation of wild-type JPO2 and deletion mutants used in panel A. Predicted α helices and β strands are indicated below the full-length sequence as filled and hatched boxes, respectively. Binding affinities of deletion mutant proteins relative to wild-type protein are indicated (+, +/- or -).

Primary sequence analysis of JPO2 predicted a PEST region (residues 29-54), a putative LZ (residues 213-235), NLS (residues 301-318) and a RING-finger-like Zn-binding domain (residues 349-425) (Chen et al., 2005). Secondary structure predictions obtained using the NSP@ Web Server [ (Combet et al., 2000)] suggested that approximately 41% and 6% of JPO2 residues are in α-helices and β-strands, respectively (Fig. 4B). The minimal p75-binding fragment, JPO258-119, is predicted to contain both α-helical and β-strand elements (Fig. 5), suggesting it could fold as an independent structural domain. Approximately 31% of the residues within a slightly larger JPO258-128 fragment are identical across a sequence alignment of vertebrate orthologs (overall 30.1% sequence identity) (Fig. 5).

Fig. 5.

Conservation of the p75-binding domain in JPO2 among mammalian and avian species. Sequence alignment of the p75-binding domain (human residues 58-128) among known JPO2 orthologs. Amino acids identical in all species are highlighted in red, and sequence identity in at least 70% of the species as well as conservative amino acid substitutions are in yellow. Predicted secondary structural elements (α-helices, wavy lines; β-strands, arrows) are shown at the bottom; the minimal binding region identified in Fig. 4 is indicated. Numbering on the top of the alignment corresponds to the human sequence. Accession numbers for JPO2 orthologs are: chimpanzee (Pan troglodytes), XP 527681; mouse (Mus musculus), NP 666152; rat (Rattus norvegicus), NP 001030125; cow (Bos taurus), XP 592513; dog (Canis familiaris), XP 539464; chicken (Gallus gallus), NP 001026153. The alignment was created using ESPript (Gouet et al., 1999).

p75 tethers JPO2 to chromatin

We and others have previously demonstrated that p75 fully accounts for the interaction of ectopically expressed HIV-1 IN with chromatin (Emiliani et al., 2005; Maertens et al., 2003; Vanegas et al., 2005). Given the striking similarities between various aspects of the p75-IN and p75-JPO2 interactions (Figs 2, 3), we asked whether p75 might exert a similar chromatin-tethering effect on another IBD-interacting protein. Hereto, RNA interference was used to monitor the effects of p75 silencing on the cellular distribution of JPO2 protein. EGFP-JPO2 was expressed in 293T-si1340/1428 cells that stably express a pair of short hairpin (sh) interfering RNAs to knockdown p75, or in control 293T-siScram cells containing scrambled shRNA with no effect on the levels of endogenous p75 (Ciuffi et al., 2005; Llano et al., 2004b). Although EGFP-JPO2 accumulated in the nuclei of 293T-siScram cells, it displayed diffuse, pan-cellular distribution in 293T-si1340/1428 cells (Fig. 6). This observation was independent of EGFP-JPO2 expression levels, since both low- and high-expressing cells displayed the same distribution (Fig. 6). Furthermore, whereas EGFP-JPO2 associated specifically with condensed chromosomes in mitotic 293T-siScram cells, it distributed diffusely throughout mitotic 293T-si1340/1428 cells (Fig. 6). Thus, p75 tethers JPO2 to chromosomes, as it does for ectopically expressed HIV-1 IN.

Fig. 6.

p75 tethers JPO2 to chromosomes. 293T-si1340/1428 cells depleted for endogenous p75 or control 293T-siScram cells were transfected with the EGFP-JPO2 expression construct. At 48 hours post-transfection, fixed cells were immunostained for p75 (red). EGFP-JPO2 was detected by epifluorescence (green) and DNA by staining with DRAQ5 (blue). Bars, 10 μm.


Transcriptional co-activator p75 has been implicated in various human diseases such as autoimmunity (Daniels et al., 2005; Dellavance et al., 2005; Ganapathy et al., 2003; Okamoto et al., 2004), cancer (Ahuja et al., 2000; Daniels et al., 2005; Grand et al., 2005; Hussey et al., 2001; Morerio et al., 2005) and AIDS (Cherepanov et al., 2003; Cherepanov et al., 2005a; Ciuffi et al., 2005; Emiliani et al., 2005; Llano et al., 2004a; Llano et al., 2004b; Maertens et al., 2003; Vandegraaff et al., 2006; Vandekerckhove et al., 2006). Although their exact cellular function(s) is yet to be established, the literature has linked p75 and its less-abundant alternative splice form p52 to transcription regulation and splicing (Fatma et al., 2001; Ge et al., 1998a; Ge et al., 1998b; Singh et al., 2001). p75, a 530-residue protein, contains almost the entire p52 sequence (333 residues) except for eight C-terminal residues. Oddly, it was p52 that was the more potent general transcriptional co-activator in vitro (Ge et al., 1998a). Reportedly, p52, but not p75, interacted with the ASF/SF2 splicing factor and displayed activity in in vitro splicing assays (Ge et al., 1998b). It seems reasonable to assume that the disparate in vitro activities of p75 and p52 are a result of differences in the way these proteins interact with the transcription and splicing machineries. We propose that the PHAT IBD, a protein-interaction domain within the C-terminal region of p75 that is absent from p52, affects regulation of protein function and thus in large part accounts for the observed differences.

Using various complementary techniques, we establish here that JPO2 is a genuine interactor of p75. Indeed, the proteins were recovered in reciprocal co-immunoprecipitation experiments (Fig. 1) and co-localized in human cells (Fig. 2). Persuasively, EGFP-JPO2 re-localized to the cytoplasm upon co-expression of an NLS-deficient form of p75. The similarities between the functional consequences of JPO2 and HIV-1 IN association with p75 are remarkable. Both proteins share the same interaction domain within the C-terminal region of p75. HIV-1 IN (Cherepanov et al., 2005a; Cherepanov et al., 2005b) and JPO2 (Fig. S1, supplementary material) bind to the interhelical loop regions of the IBD. p75 influenced the intracellular stability of both HIV-1 IN (Llano et al., 2004a; Maertens et al., 2003) and JPO2 (Fig. 1D). Depletion of endogenous p75 caused dramatic redistribution of ectopically expressed HIV-1 IN (Llano et al., 2004b; Maertens et al., 2003) and JPO2 (Fig. 6) from the nucleus to the cytoplasm, implying that p75 dictates the intracellular distribution of both interaction partners. Finally, association of both interactors with chromosomes is strictly dependent on p75 (Emiliani et al., 2005; Llano et al., 2004b; Maertens et al., 2003) (Fig. 6).

The structure at the dimer interface of the IN CCD defines the principal p75-recognition element in HIV-1 (Cherepanov et al., 2005a; Maertens et al., 2003). The interhelical loop regions of the IBD can probably bind to divergent lentiviral INs by recognizing the backbone conformation of the viral proteins (Cherepanov et al., 2005a). Secondary structural elements predicted within the p75-binding fragment of JPO2 (Fig. 5) suggest that this region is involved in a stable tertiary structure. Although it will be important to characterize the structural basis of the JPO2-p75 interaction further, thus far, the absence of recognizable sequence similarity between JPO2 and HIV-1 IN underscores the striking capacity of the IBD to recognize highly divergent proteins specifically.

On the basis of the intracellular re-localization of JPO2 upon co-expression of NLS-deficient mutants of p75 or knockdown of endogenous p75, we tentatively speculate that p75 is involved in the nuclear import of JPO2. Further experiments will be required to ascertain whether JPO2 gains access to the nucleus through p75. Alternatively, the latter might indirectly contribute to the nuclear accumulation of JPO2 by trapping and stabilizing the protein within the nucleus. It would be of interest to determine if the predicted JPO2 NLS (residues 301-318) (Chen et al., 2005) is functional.

The C-terminal half of JPO2 is highly similar to another Myc-interacting protein, JPO1 (Huang et al., 2005). Only the C-terminal part of the p75-binding domain in JPO2 is present in JPO1, and the overlapping sequences fail to reveal recognizable homology (Fig. S2, supplementary material). Whereas both JPO1 and JPO2 appear to be present in mammalian and avian species, after exhaustive sequence database searches we could identify only a single JPO1/2-like protein (JPOL) in Xenopus laevis (Fig. S2, supplementary material) or Xenopus tropicalis (data not shown). We speculate that avian and mammalian JPO1 and JPO2 are paralogs evolved from a common ancestor, related to an amphibian JPOL precursor.

JPO1, a product of a Myc-responsive gene, is implicated in neoplastic transformation (Prescott et al., 2001) and is frequently over-expressed in human tumors (Huang et al., 2005; Osthus et al., 2005). Interestingly, JPO2 is a Myc-interacting protein that binds Myc through a Leu-rich motif (residues 213-235) (Huang et al., 2005). This motif is conserved in JPO1 and, concordantly, JPO1 was reported to bind Myc (Huang et al., 2005). Both JPO1 and JPO2 induced colony formation in vitro and contributed to Myc-mediated transformation (Huang et al., 2005; Prescott et al., 2001). JPO2 was recently reported to repress the activity of the human MAO A promoter (Chen et al., 2005) and might therefore contribute to deregulation of MAO levels associated with cancers (Pietrangeli and Mondovi, 2004). In our hands, endogenous Myc robustly co-immunoprecipitated with Flag-JPO2 (data not shown), confirming the report by Penn and colleagues (Huang et al., 2005). Since JPO2 binds to p75 and Myc through distinct, non-overlapping sequences, it seems that a ternary complex of these three proteins might exist in cells. Our preliminary experiments indicated that the ternary complex does form, as detectable levels of Myc were co-immunoprecipitated with endogenous p75 (data not shown). Further research will be necessary to answer the question of whether p75, JPO2 and Myc function in concert during cellular transformation.

p75 is a modular protein that contains a pair of small structural domains (the N-terminal PWWP domain and the PHAT IBD) connected by flexible regions (Cherepanov et al., 2004). Association of p75 with chromatin is dependent on a central Arg-Lys region that contains the NLS and AT-hook-like elements of the protein and, to a lesser extent, the PWWP domain (Turlure et al., 2006). The IBD does not bind to DNA in vitro (Turlure et al., 2006) and is therefore free for interaction with other proteins. Our data indicate that the PHAT IBD is a versatile domain that interacts with, and recruits, divergent factors to chromatin. Combinatorial networks of mammalian transcription factors that primarily rely on protein-protein recognition rather than solely on DNA signals are not a new concept. For example, activation versus repression through the ETS family of transcription factors depends on the identity of specific protein-binding partners (reviewed by Verger and Duterque-Coquillaud, 2002).

Given that p75 appears to affect the human transcriptome globally, regulating the expression of about 10% of genes (Ciuffi et al., 2005), it is perhaps not accidental that HIV-1 has evolved such that it uses p75 to dock its integration machinery to chromosomal targets. Although more work is required to ascertain fully where p75 fits into the jigsaw puzzle of gene expression, our results reveal an important function for the protein, laying groundwork for future research and hopefully its exploitation in controlling oncogenesis and viral infections.

Materials and Methods

Plasmid constructs

Plasmids pBHA-p75, pECMV-FlagMob2, pGG-p75(K150) and pEGFP-p75/Ct were described elsewhere (Cherepanov et al., 2004; Devroe et al., 2004; Maertens et al., 2004). Mutations were introduced into pBHA-p75 using the QuickChange protocol (Stratagene). To make pGM-Flag, the following oligonucleotides containing the Kozak consensus sequence and encoding a Flag-tag were annealed and ligated to BamHI/HindIII-digested pcDNA6V5HisB (Invitrogen): AE2178 (5′-AGCTTGCAGACACCATGGATTACAAGGATGACGATGACAAGG-3′) and AE2179 (5′-GATCCCTTGTCATCGTCATCCTTGTAATCCATGGTGTCTGCA-3′). The JPO2 open-reading frame (ORF) was amplified from cDNA isolated from HeLa cells using primers AE2250 (5′-GCAGGATCCATGGAGTTGGCGACTCGCTACC-3′) and AE2251 (5′-AGCGAATTCTCAATTGTCTTCTACCAGCTCC-3′). BamHI/EcoRI-digested PCR product was ligated to BamHI/EcoRI-cut pGM-Flag to give pGM-Flag-Ram2. Similarly, PCR-amplified JPO2 deletions were ligated to BamHI/EcoRI-digested pGM-Flag using: JPO2/1-367, AE2250 and AE2493 (5′-CGCGAATTCTCACACTGTCTTGGTGTCGATGGTC-3′); JPO2/1-346, AE2250 and AE2486 (5-CTCGAATTCTCAATCATAGATTTTATCTCG-3′); JPO2/1-301, AE2250 and AE2488 (5′-CTCGAATTCTCATCGGAAGCTGTAAAACTCTTCCG-3′); JPO2/1-257, AE2250 and AE2490 (5′-CGCGAATTCTCACCTCACTGTCTTCTTCCT-3′); JPO2/1-237, AE2250 and AE2491 (5′-CGCGAATTCTCACGAGTTCAATTCCGCC-3′); JPO2/58-454, AE2251 and AE2492 (5′-CGCGGATCCGTGCGCTTTCATTCCAAATAC-3′); JPO2/58-367, AE2492 and AE2493; JPO2/77-346, AE2486 and AE2479 (5′-CGCGGATCCACTGACTCAGAGACTGAGG-3′); JPO2/98-454, AE2251 and AE2480 (5′-CGCGGATCCCCAGAAGTAATGGTCGTGG-3′); JPO2/1-182, AE2250 and AE2539 (5′-CGCGAATTCTCATCTTTCAAGAATTGTTTTTTTCTCATTCTG-3′); JPO2/1-132, AE2250 and AE2540 (5′-CGCGAATTCTCATCTTCTAGGGGTAGCCTTATCTTCTTCTTCATC-3′); JPO2/77-182, AE2479 and AEAE2539; JPO2/77-132, AE2479 and AE2540.

To make pEGFP-Ram2, the JPO2 ORF, PCR amplified using primers AE2251 and AE2278 (5′-GCAAGATCTCTATGGAGTTGGCGACTCGCTACC-3′), was digested with BglII and EcoRI, and ligated to BglII/EcoRI-digested pEGFP-C2 (Clontech). Plasmid pHcRed1-Ram2 and deletion mutants were constructed by ligating the BamHI/EcoRI restriction fragment from pGM-FlagRam2 to BglII/EcoRI-digested pHcRed1-C1 (Clontech). All plasmid constructs were verified by sequencing. Plasmids for expressing full-length p75 and deletion mutants fused to glutathione S-transferase (GST) were previously described (Cherepanov et al., 2004). Plasmids pCP-Nat75 and pKB-IN6H were used to express non-tagged full-length p75 and HIV-1 IN with a C-terminal His6-tag, respectively, in bacteria (Maertens et al., 2003).

Cell culture, transfection, immunostaining and confocal microscopy

293T and HeLa cells were cultured at 37°C in a humidified, 5% CO2 atmosphere in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum (FBS), 100 IU/ml penicillin and 100 μg/ml streptomycin. The 293T-si1340/1428 and 293T-siScram cell lines (Llano et al., 2004b), kindly provided by E. Poeschla (Mayo Clinic College of Medicine, Rochester, MN), were maintained in selection media containing 200 μg/ml hygromycin B-30 μg/ml puromycin and 200 μg/ml hygromycin B, respectively. Cells were transfected in 6-well dishes using Lipofectamine-2000 (Invitrogen) according to the manufacturer's recommendations.

For immunofluorescence, cells were plated in 8-well LabTek chambered cover glass cuvettes (Nunc). At 24-36 hours post-transfection, cells were fixed in 4% formaldehyde in phosphate-buffered saline (PBS) for 5 minutes. After extensive washing in PBS, cells were further permeabilized in ice-cold methanol for 5 minutes. After washing in PBS, cells were incubated in blocking solution (PBS, 10% FBS, 5 mM NH4Cl). Expression of p75 was detected using monoclonal anti-p75/p52 antibody (BD Biosciences) diluted 1:200 in blocking solution and Alexa Fluor 568-conjugated goat anti-mouse antibody (Molecular Probes) diluted 1:300. DRAQ5 (Alexis Biochemicals) was used at 1 μM to visualize DNA. Fluorescent images were acquired using a Nikon C1 laser confocal microscope at 1024×1024 resolution with a 60× water immersion objective. Enhanced green fluorescent protein (EGFP) was excited at 488 nm, Alexa Fluor 568 and HcRed1 at 543 nm, and DRAQ5 at 633 nm.

Derivation of a p75-expressing stable cell line

To obtain a cell line stably expressing influenza hemagglutinin (HA)-tagged p75, pBHA-p75 linearized by digestion with AhdI was introduced into 293T cells using Lipofectamine-2000. At 24 hours post-transfection, cells were trypsinized and plated in 10 cm dishes in the presence of 10 μg/ml blasticidin (Invitrogen). A single-cell clone (hereafter referred to as 293T-HAp75) out of 40 screened was selected because it homogenously expressed HA-p75 within the cell population at levels similar to those of endogenous p75, as determined by immunofluorescence and western blotting.

Immunoprecipitation (IP) and western blotting

For large-scale IPs to identify interaction partners of p75, 293T-HAp75 and parental 293T cells were grown to confluence in 13 flasks (175 cm2 each). Cells harvested by trypsinization were washed twice in 50 ml of cold PBS. All further operations were done on ice or at 4°C. The cells were lysed for 5 minutes in 6 ml of 100mCSK buffer [100 mM NaCl, 10 mM Hepes pH 6.82, 1 mM dithiothreitol (DTT), 1 mM MgCl2, 10% sucrose (w/v)] supplemented to contain 0.5% Nonidet P-40 (NP-40), 0.1 mM phenylmethylsulfonyl fluoride (PMSF), and complete EDTA-free protease inhibitor mix (Roche). Nuclei washed three times in 100mCSK buffer were extracted in 4 ml of 450mCSK buffer (same as 100mCSK, but containing 360 mM NaCl and supplemented with 50 mM NaF, 40 mM β-glycerophosphate, 0.1 mM Na-o-vanadate). After removing debris by centrifugation at 20,630 g for 20 minutes, the soluble nuclear extract was supplemented to contain 5 mM EDTA and 50 μl protein G agarose (Roche). After 30 minutes, NP-40 was added to a final concentration of 0.1% to the precleared supernatant, and the extract was incubated with 75 μl of HA.11 affinity matrix (Covance) for 4 hours. Beads collected by gentle centrifugation were washed twice in 450mCSK supplemented with 0.1% NP-40 and three times with 100mCSK/0.1% NP-40. Bound proteins were eluted in 1 ml of 6 M ammonium thiocyanate for 1 hour, followed by precipitation with 10% (w/v) trichloroacetic acid overnight. Precipitated proteins washed in acetone were re-dissolved in 40 μl sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) loading buffer, separated on an 11% SDS-PAGE gel, and detected with silver staining (Shevchenko et al., 1996). Proteins were identified by trypsin digestion in gel slices and liquid chromatography-tandem mass spectrometry (LC-MS/MS).

For small-scale IPs, 293T cells plated at 1.4×106 cells per well in a 6-well dish the day before transfection were transfected with 4 μg pGM-FlagRam2 or pECMV-FlagMob2 using Lipofectamine-2000. In co-transfection experiments, 2 μg of each plasmid was used. Cells harvested at 24 hours post-transfection were lysed in 300 μl 450mCSK supplemented with 0.5% NP-40 and protease inhibitors. For IP, 100 μl of pre-cleared extract was incubated with 25 μg of monoclonal anti-p75/p52 and 15 μl of protein G agarose, 15 μl of anti-Flag M2 conjugated agarose (Sigma), or 15 μl of anti-HA.11 affinity matrix. Beads collected by centrifugation were washed once in extraction buffer, and then three times in 100mCSK/0.5% NP-40. Bound proteins were eluted in 60 μl SDS-PAGE sample buffer containing 50 mM DTT. Eluted proteins were separated on 4-20% SDS-PAGE gels and electroblotted onto polyvinyledene difluoride membranes. Proteins were detected using rabbit anti-JPO2 antibodies raised against a C-terminal peptide (CYLESLQKELVEDN) (Bethyl Laboratories) at 1:10,000 dilution, monoclonal anti-p75/p52, anti-Flag M2, anti-β-actin (Sigma), or monoclonal anti-HA antibody 16B12 (Covance) at 1:2,000 dilution. Horseradish peroxidase (HRP)-conjugated goat anti-rabbit and goat anti-mouse secondary antibodies (DakoCytomation) were used at a 1:10,000 and 1:30,000, respectively. Signals were developed with the ECL+ chemiluminescent kit (BioRad).

In vitro pull-down assays

Flag-tagged full-length and JPO2 deletion mutants were expressed in vitro using the TNT T7 coupled reticulocyte lysate system (Promega) according to the manufacturer's instructions. The templates were prepared by linearizing pGM-Flag-Ram2 or various deletion constructs with XhoI. In vitro transcription and translation reactions (50 μl) contained 1 μg plasmid, 30 μCi [35S]Met (GE Healthcare) and 25 μl reticulocyte lysate.

For GST pull-down assays, 20 μl of glutathione sepharose (GS) (GE Healthcare) slurry was incubated overnight at 4°C with 5 μg of GST or GST-fusion protein in 200 mM NaCl, 5 mM DTT, 25 mM Tris-HCl, pH 7.4. Beads collected by gentle centrifugation were washed in GST pull-down buffer (GPDB: 150 mM NaCl, 5 mM MgCl2, 5 mM DTT, 0.1% NP-40, 25 mM Tris-HCl, pH 7.4). Pull-down reactions contained 10 μl pre-bound GS beads (settled volume), 10 μg bovine serum albumin (BSA) and 2 μl of 35S-labeled JPO2 in 200 μl GPDB. Following 3 hours at 4°C, beads were collected by centrifugation and washed three times with GPDB. Bound proteins eluted in SDS-PAGE sample buffer containing 50 mM DTT were separated in SDS-PAGE gels and detected by staining with Coomassie-R250 and autoradiography.

For His6-tag pull-down assays, 20 μl of Ni-nitrilotriacetic acid (NTA) beads (Qiagen) were incubated with 10 μg IN-His6, 5 μg LEDGF/p75, 10 μg BSA, and/or 2 μl of in-vitro-translated JPO2 protein in 300 μl His6-tag pull-down buffer (HPDB: 0.4 M NaCl, 2 mM MgCl2, 25 mM imidazole, 25 mM Tris-HCl, pH 7.4, 0.1% NP-40). Reactions were rocked at 4°C for 4 hours. After gentle centrifugation, the beads were washed in three changes of HPDB. Bound proteins eluted in 30 μl of SDS-PAGE sample containing 200 mM imidazole and 50 mM DTT were separated in SDS-PAGE gels and detected by staining with Coomassie-R250 and autoradiography.


We are grateful to E. Poeschla (Mayo Clinic College of Medicine, Rochester, MN) for providing 293T-si1340/1428 and 293T-siScram cell lines, and to E. Devroe and P. A. Silver (Dana-Farber Cancer Institute) for pECMV-FlagMob2 DNA. Confocal microscopy and LC-MS/MS were carried out at the Nikon Imaging Center and the Taplin Biological Mass Spectrometry Facility, respectively (Harvard Medical School). This work was supported by NIH grants AI39394 and AI52014. Core facilities were supported by a Center for AIDS Research grant (AI60354) and the Dana-Farber Cancer Institute/Harvard Cancer Center.


  • Supplementary material available online at

  • * Present address: Molecular Oncology Laboratory, Cancer Research UK, London Research Institute, 44 Lincoln's Inn Fields, London, WC2A 3PX, UK

  • Present address: Imperial College London, St Mary's Campus, Norfolk Place, London, W2 1PG, UK

  • Accepted March 24, 2006.


View Abstract