Docstoc

strategene

Document Sample
strategene Powered By Docstoc
					      JBC Papers in Press. Published on March 14, 2002 as Manuscript M201799200




                 Association of the human SUMO-1 protease
                        SENP2 with the nuclear pore




                                                                                           Downloaded from www.jbc.org by on January 16, 2009
                             Jun Hang and Mary Dasso*




Laboratory of Gene Regulation and Development, NICHD/NIH, Building 18T, Room 106, 18
Library Drive, Bethesda MD 20892-5431.

*Corresponding author: Tel: 301-402-1005; Fax: 301-402-1323; email: mdasso@helix.nih.gov

Running title: SENP2 localizes to nuclear pores


                                                  1
SUMO-1 is a small ubiquitin-like protein that becomes covalently conjugated to other

proteins. A family of proteases catalyzes deconjugation of SUMO-1-containing species.

Members of this family also process newly-synthesized SUMO-1 into its conjugatable form.

To understand these enzymes better, we have examined the localization and behavior of the

human SUMO-1 protease SENP2. Here we show that SENP2 associates with the nuclear

face of nuclear pores, and that this association requires protein sequences near the N-

terminus of SENP2. We also show that SENP2 binds to Nup153, a nucleoporin that is

localized to nucleoplasmic face of the pore. Nup153 binding requires the same domain of

SENP2 that mediates its targeting in vivo. Removal of the Nup153-interacting region of




                                                                                               Downloaded from www.jbc.org by on January 16, 2009
SENP2 results in a significant change in the spectrum of SUMO-1 conjugates within the

cell. Our results suggest that association with the pore plays an important negative role in

the regulation of SENP2, perhaps by restricting its activity to a subset of the conjugated

proteins within the nucleus.




                                             2
       SUMO-1 is a ubiquitin-like protein. SUMO-1 can be covalently conjugated to other

proteins through an isopeptide linkage in a manner similar to ubiquitin (1). The SUMO-1

conjugation pathway utilizes proteins that both show sequence similarity to analogous enzymes

in the ubiquitin pathway and utilize similar biochemical mechanisms (1). A large and growing

number of SUMO-1 conjugation substrates have been reported in vertebrates (1). Notably, the

profile of SUMO-1 conjugated proteins changes substantially in response to altered cellular

conditions (for instance, (2)), suggesting that there are mechanisms to control the specificity of

conjugation and/or deconjugation of SUMO-1 differentially between distinct substrates.




                                                                                                       Downloaded from www.jbc.org by on January 16, 2009
       Unlike enzymes of the SUMO-1 conjugation pathway, enzymes involved in SUMO

processing and deconjugation are not closely related by sequence to their ubiquitin counterparts.

Rather, known SUMO proteases share sequence homology in their catalytic domains that is more

nearly conserved to viral proteases (3). Two SUMO proteases have been described in budding

yeast, Ulp1p (3) and Ulp2p/Smt4p (4,5). Ulp1p is concentrated near the nuclear periphery (5)

and interacts with nuclear pore components in two-hybrid assays (6), while Ulp2p is localized

throughout the nucleus (5). ULP1 is an essential gene, and temperature sensitive Ulp1p mutants

arrest at the G2/M transition of the cell cycle. Ulp2 is not essential but it is required for normal

meiotic development, for regulation of spindle checkpoint arrest and for chromatin condensation

of rDNA during mitosis (4, 5). Interestingly, these proteins do not appear to act in a simple

complimentary manner, since ulp1-ts ulp2 null double mutants grows better than ulp2 single

mutants under a variety of conditions (5).

       In mammals, database searches find at least seven members of the SUMO protease

family (7), some of which have now been confirmed to act as SUMO proteases in vitro (8-11).




                                                 3
Outside of their conserved catalytic domain, these proteases possess non-conserved N-terminal

extensions of varying lengths and relatively short non-conserved C-terminal sequences. SENP2

was discovered both through its homology to other SUMO proteases (7), and also through its

interactions with murine Axin, a regulator of the Wnt signaling pathway (12). When

overexpressed in tissue culture cells or under in vitro conditions, the murine SENP2 homologue

(Smt3IP2) cleaves conjugates of SUMO-1, SUMO-2 and SUMO-3 (11).

       Here we show that full-length human SENP2 associates with nuclear pores in a manner

similar to Ulp1 in yeast. This association occurs exclusively with the nuclear face of the pore and

requires sequences near the N-terminus of SENP2. We also show that SENP2 binds specifically




                                                                                                      Downloaded from www.jbc.org by on January 16, 2009
to Nup153, a nucleoporin localized to nucleoplasmic face of the nuclear pore, and that this

association requires the same domain of SENP2 that mediates its targeting in vivo. Remarkably,

a mutant SENP2 protein that is unable to bind Nup153 is significantly more effective in

promoting deconjugation of SUMO-1-conjugated species, indicating that localization of SENP2

to the nuclear pore plays an important role in spatially restricting the activity of this enzyme.




                                                  4
                             EXPERIMENTAL PROCEDURES

Antibodies - Affinity-purified antibodies against Xenopus Nup153 were the kind gift of K.

Ullman (Huntsman Cancer Institute, Salt Lake City). Affinity-purified antibodies against

Xenopus Nup98 were the kind gift of M. Powers (Emory University, Atlanta). The mouse

monoclonal antibody mAb414 against nucleoporins, mouse monoclonal antibody against

RanGAP1, and anti-V5 antibody were purchased from BAbCO (Richmond, CA), Zymed

Laboratories, Inc. (South San Francisco, CA), and Strategene (La Jolla, CA) respectively. Mouse

monoclonal anti-FLAG antibody and rabbit polyclonal antibody against green fluorescence

protein (GFP) were purchased from Clontech, Inc. (Palo Alto, CA). Rhodamine- and Alexa 488-




                                                                                                  Downloaded from www.jbc.org by on January 16, 2009
conjugated secondary antibodies were purchased from Molecular Probes, Inc (Eugene, OR).



DNA Constructs and protein expression - The SENP2 cDNA was amplified from the human

universal QUICK-CloneTM cDNA (CLONTECH). The sequence of the encoded protein was

identical to human cDNAs that have been reported by other laboratories (GenBank accession

numbers NM021627, AF151697). The encoded protein has previously been designated as

SENP2, and we will use this nomenclature throughout this report. We fused the SENP2 coding

sequence in-frame to the 3' end of the green fluorescent protein (GFP) coding sequence by

insertion between the EcoRI and SalI sites of pEGFP-C2 (CLONTECH). Similarly, we prepared

a vector encoding a version of SENP2 with an N-terminal FLAG tag by insertion of the SENP2

coding region between EcoRI and SalI sites of pCMV-Tag2B (Strategene). Truncation mutants

were generated by PCR using Pfu DNA polymerase (Strategene), and subcloned into the same

vectors.




                                              5
       The SENP2 truncation mutants were also subcloned into pGEX4T-1 between EcoRI and

SalI for expression of glutathione S-transferase (GST) fusion proteins in bacteria. Expression of

the recombinant GST-SENP2 fusion proteins was induced with 0.05 mM IPTG at room

temperature for 4 h, and purified according to the manufacture's instruction (Amersham

Pharmacia Biotech, Piscataway, NJ). A cDNA encoding the mature form SUMO-1 (amino acids

1-97) was subcloned into pcDNA4/HisMax C (Invitrogen) between the BamH I and Not I

restriction sites, allowing expression with six histidine and Xpress epitope tags at its N-terminus.

A human RanGAP1 cDNA was amplified from the pET-RanGAP1 plasmid provided kindly by

Volker Gerke (University of Muenster, Muenster, Germany). The cDNA was cloned into




                                                                                                       Downloaded from www.jbc.org by on January 16, 2009
pcDNA 3.1/V5-His-TOPO TA cloning vector (Invitrogen), allowing its expression with V5

epitope and six histidine tags at its C-terminus. A form of RanGAP1 that cannot be modified by

SUMO-1 was generated by PCR using Pfu DNA polymerase to produce a point mutation that

substituted lysine 524 with an arginine residue.



Cell Culture, Transfection and Immunofluorescence - COS7 and HeLa cells were cultured in

DMEM (GIBCO BRL, Gaithersburg, MD) containing 10% fetal bovine serum, 100 µg/ml

penicillin, 100 µg/ml streptomycin and 2 mM glutamine (Biofluids, Rockville, MD). In all cases,

the cells were transfected using Effectene transfection reagent (Qiagene Inc., Valencia, CA). To

measure the effect of SENP2 expression on the overall profile of conjugated substrates (Figure

5A), COS7 cells were grown in 6-well plates and transfected with vectors expressing His-

Xpress-tagged SUMO-1 and eGFP-SENP2, as indicated. To measure the effect of SENP2

expression specifically on the conjugation of RanGAP1 conjugation (Figure 5B), COS7 cells

were transfected with vectors expressing eGFP or eGFP-SENP2, as well as vectors expressing




                                                   6
V5-tagged wild type or unconjugatable mutant human RanGAP1. In both Figure 5A and 5B, the

cells were washed twice with cold PBS 24 h after transfection and suspended in 125 µl boiling

2×SDS sample buffer, followed by brief sonication. The samples were analyzed by Western

blotting as described below.

       HeLa cells were grown for immunofluorescence on glass coverslips. In all figures except

Figure 1B, the cells were fixed for 5 min with 4% formaldehyde plus 2% sucrose in KB buffer

(10 mM TrisCl, pH7.7, 150 mM NaCl, 0.1% bovine serum albumin), and permeabilized with

0.1% Triton X-100 in KB buffer. In Figure 1B, HeLa cells were fixed with 3% formaldehyde for

20 minutes at room temperature, and cells were alternatively permeabilized with Triton X-100 as




                                                                                                  Downloaded from www.jbc.org by on January 16, 2009
described above or permeabilized with 0.004% digitonin at 4oC for 15 minutes. In all cases, the

coverslips were incubated for 1 h at room temperature in primary and secondary antibodies that

had been diluted in KB buffer with 2% normal horse serum. Hoechst 33258 was added to the

secondary antibody incubation to stain the DNA. After each incubation, the cells were rinsed 5

times for 2 minutes in KB buffer with 2% normal horse serum. The coverslips were mounted in

Vectashield (Vector Laboratories, Burlingame, CA). Images were captured on a Zeiss Axioskop

microscopy with a Hamamatsu Orca II digital CCD camera (Carl Zeiss, Inc., Thornwood, NY)

and Openlab software (Improvision, Inc., Lexington, MA).



Western Blotting - Proteins were resolved in 4-20% SDS-PAGE gel and transferred to PVDF

membrane. The blots were blocked in 5% non-fat dried milk in PBS (Biofluids) containing

0.15% Tween-20 (PBS-T) at room temperature for 1 h, incubated in primary and HRP-labeled

secondary antibodies for 1 h each, thoroughly rinsed with PBS-T after each incubation, and

detected using ECL Plus reagents (Amersham Pharmacia Biotech).




                                              7
GST Pull-down Assays - Xenopus interphase egg extracts were prepared exactly as described

elsewhere (13). 15 µg purified recombinant GST-SENP2 fusion proteins were incubated with 15

µl egg extract plus 500 µl buffer B (20 mM TrisCl, pH 8.0, 50 mM NaCl, 0.5 mM DTT, 2.5 mM

MgCl2, 0.1% Triton X-100, 10% glycerol) for 5 hours at 4 oC. 20 µl glutathoine agarose beads

that had previously been equilibrated with buffer B were added to the reaction and incubated

overnight at 4 oC. The beads were washed 4 times with 1.5 ml buffer B, and the bound proteins

were eluted using 30 µl of SDS sample buffer. The products were subjected to Western blotting

using the indicated antibodies.




                                                                                                Downloaded from www.jbc.org by on January 16, 2009




                                             8
                                          RESULTS

SENP2 is localized to the nucleoplasmic side of the nuclear pore – To better understand the

function of individual SUMO proteases, we cloned the human homologue of SENP2 by PCR,

and performed experiments to confirm that a recombinant fragment of the human SENP2 protein

(amino acids 178-590, containing the putative catalytic region) had activity in assays for SUMO-

1 processing and isopeptide cleavage of bonds between SUMO-1 and RanGAP1 (data not

shown). Our results were very similar to those previously described for in vitro analysis of

mouse SENP2 (11). To pursue a better understanding of SENP2 in vivo, we subcloned the

SENP2 open reading frame into plasmid vectors for the expression of fusion protein encoding




                                                                                                   Downloaded from www.jbc.org by on January 16, 2009
SENP2 tagged with either green fluorescent protein (eGFP-SENP2) or a Flag epitope (Flag-

SENP2).

       To determine the localization of SENP2, we examined the localization of a fusion

between GFP and the full SENP2 coding region (eGFP-SENP2) in HeLa cells. When expressed

at low levels (Figure 1A), this protein co-localized with the nuclear envelope, whereas eGFP

alone was diffusely distributed. At higher levels of expression, we found the fusion protein not

only at the nuclear envelope but also within inclusions in the nucleus (data not shown). eGFP-

SENP2 distribution overlapped with immunofluorescent staining using the monoclonal antibody

414 (mAb414), which recognizes a family of FxFG-containing nuclear pore proteins (Figure

1A). Notably, the distribution of eGFP-SENP2 was more restricted than mAb414 staining and

appeared to be found primarily on the nuclear side of the nuclear envelope (Figure 1A, lower

panel). The Flag-SENP2 localized similarly (data not shown), suggesting that the pattern of

localization was independent of the fusion moiety. In addition, polyclonal rabbit antibodies

directed against SENP2 showed strong staining of the nuclear envelope in immunofluorescence




                                               9
experiments (data not shown), suggesting that the endogenous SENP2 protein is also localized to

the nuclear envelope. Together, these observations demonstrate unambiguously that SENP2 is

associated with the nuclear envelope, and strongly indicate that SENP2 is resident on the nuclear

face of the nuclear pore.

       To test more directly whether SENP2 is restricted to the nuclear side of the pore, we

examined eGFP-SENP2 accessibility to anti-eGFP antibody staining under different

permeabilization conditions (Figure 1B). When cells were permeabilized with digitonin, which

disrupts the plasma membrane but leaves the nuclear envelope intact (14), we observed no

staining with anti-eGFP antibodies. However, it was clear that the fusion protein remained was




                                                                                                    Downloaded from www.jbc.org by on January 16, 2009
still present because the GFP emission was observed on the nuclear envelope. By contrast,

permeabilization with a detergent that disrupts the nuclear envelope (triton X-100) allowed

staining, indicating that the fusion protein could be recognized well by anti-eGFP antibodies.

These observations show that SENP2 is localized to the nucleoplasmic side of the nuclear

envelope.



The SENP2 N-terminus is necessary and sufficient for nuclear envelope targeting - We wished to

determine which sequences within SENP2 are required for its correct targeting to the nuclear

envelope. To do this, we made a series of deletions in the SENP2 fusion protein, encompassing

both the N- and C-termini (Figure 2A). Deletion of as few as 30 amino acids from the N-

terminus of SENP2 disrupted its association to the pore (Figure 2B). By contrast, fusion proteins

that were extensively deleted from the C-terminus, including one that retained only 70 amino

acids of the N-terminus of SENP2 (eGFP-SENP2(1-70)), were able to localize correctly to the

nuclear envelope (Figure 3). A similar deletion analysis using Flag-SENP2 fusion proteins




                                               10
provided essentially identical results (data not shown), confirming that this finding was

independent of the fusion epitope used. These observations show that the sequences within the

N-terminal 70 amino acids of SENP2 are both necessary and sufficient for its correct localization

at the nuclear pore.



The N-terminus of SENP2 interacts with Nup153 - In order to test whether SENP2 binds to

nuclear pore proteins, we utilized interphase Xenopus egg extracts (13) as a source of

unassembled, soluble nuclear pore components. We incubated the egg extracts with recombinant

glutathione-S transferase (GST) fusion proteins that encoded different regions of the SENP2




                                                                                                    Downloaded from www.jbc.org by on January 16, 2009
protein. After incubation, we purified the fusion proteins by affinity chromatography and

examined whether the samples specifically retained extract proteins that could be recognized in

Western blotting assays by mAb414 (Figure 4, upper panel).

       Fusion proteins encoding the N-terminus of SENP2 (e.g., GST-SENP2(1-170) and GST-

SENP2(1-70)) specifically retained a mAb414-reactive band that migrated with a mobility

corresponding to 180 kD on gel electrophoresis. Previous characterization of mAb414-reactive

proteins in Xenopus egg extracts (15), suggested that the 180 kD band was likely to be the

nucleoporin Nup153 (16). To confirm this identification, we subjected the same samples to

Western blotting using antibodies against the human Nup153 protein that also recognize the

Xenopus Nup153 homologue (17). This analysis showed that Nup153 associated with GST-

SENP2(1-170) and GST-SENP2(1-70) (Figure 4, lanes 5,6), but did not associate with GST or

with fusion proteins containing any other region of SENP2 (Figure 4).

       Several additional observations suggested that the retention of Nup153 was highly

specific. First, no other mAb414-reactive bands were specifically retained in association with




                                               11
GST or any of the GST fusion proteins. Second, Western blots using antibodies directed against

a GLFG nucleoporin associated with the nucleoplasmic side of the pore (18), Nup98, did not

recognize any proteins associated with SENP2 (Figure 4, lower panel). Taken together, our

findings strongly suggest that SENP2 interacts with Nup153 in Xenopus egg extracts. This

finding is consistent with the localization of SENP2 determined in Figures 1 and 2, since Nup153

has been demonstrated to be a component of the nuclear basket in vertebrates (19). The

specificity of these interactions was independently confirmed by directed two-hybrid analysis, in

which strong interactions were observed between Nup153 and either full length SENP2 or the N-

terminal domain of SENP2 (data not shown). By contrast, no specific interactions were observed




                                                                                                     Downloaded from www.jbc.org by on January 16, 2009
between Nup153 and SENP2 sequences outside of the first 100 amino acids.



SENP2 targeting regulates its activity against cellular substrates - In order to determine whether

SENP2 localization has any role in regulation of its activity, we transfected COS7 cells with

wild-type eGFP-SENP2, a mutant lacking the N-terminal 70 amino acids of SENP2 (eGFP-

SENP2(71-590)), and a mutant in which a critical cysteine in the predicted active site of the

enzyme was changed to serine (eGFP-SENP2-C/S). eGFP-SENP2-C/S correctly targeted to the

nuclear pore in a manner that was indistinguishable from the wild-type protein (data not shown).

To monitor the pattern of SUMO-1 conjugation within the transfected cells, we simultaneously

transfected with a lower concentration of a plasmid expressing tagged SUMO-1 protein (His-

Xpress-SUMO-1) and assayed the conjugation of the tagged SUMO-1 by Western blotting

(Figure 5).

       We observed that His-Xpress-SUMO-1 became conjugated to a variety of proteins when

it was co-transfected with a vector expressing eGFP, most easily observed on Western blots as a




                                               12
high molecular weight smear. This spectrum was slightly attenuated when the eGFP expression

vector was replaced with a vector expressing wild-type SENP2. By contrast, when eGFP-

SENP2-C/S was co-expressed with the tagged SUMO-1 it dramatically increased the level of

SUMO-1 conjugation within transfected cells, indicating that it was not only enzymatically

inactive but also functioned in a dominant fashion to disrupt deconjugation by endogenous

SENP2. We do not currently know the mechanism whereby eGFP-SENP2-C/S inhibited the

endogenous protein.

       Most remarkably, expression of eGFP-SENP2(71-590) caused the loss of almost all

conjugated forms of His-Xpress-SUMO-1, indicating that SENP2 was much more effective in




                                                                                                   Downloaded from www.jbc.org by on January 16, 2009
deconjugation of nuclear proteins when it was no longer tethered to the nuclear pore. These

results suggest that association with the pore may play an important negative role in the

regulation of SENP2, restricting its activity to a subset of the conjugated proteins within the

nucleus, and that allowing the protein to access other locations within the nucleus promotes

inappropriate deconjugation of SUMO-1 species that are not normally substrates for this enzyme.

       Since SENP2 is resident at the nuclear pore, we also examined the capacity of

overexpressed SENP2 to alter the conjugation status of RanGAP1 in vivo. RanGAP1 is localized

in the cytosol, where SUMO-1 conjugation targets it to the nuclear pore through association with

a large nucleoporin, RanBP2 (20, 21). In this experiment, we co-transfected a plasmid expressing

V5-tagged RanGAP1 with the plasmid expressing eGFP-SENP2 (Figure 5B). As a control, we

performed similar experiments with a mutant version of RanGAP1 lacking the single lysine

residue that becomes modified by SUMO-1 conjugation (RanGAP1-K524R; (22)). eGFP-SENP2

expression did not substantially alter the pattern of RanGAP1 modification. Moreover, under

conditions of moderate or even massive eGFP-SENP2 overexpression, RanGAP1 staining at the




                                              13
nuclear pore was retained (data not shown), further arguing that its SUMO-1 modification status

is not regulated by SENP2, despite the fact that RanGAP1 can be deconjugated from SUMO-1 in

vitro by SENP2 ((11); data not shown). These observations suggest that the primary substrate of

SENP2 at the nuclear pore is unlikely to be RanGAP1.




                                                                                                  Downloaded from www.jbc.org by on January 16, 2009




                                              14
                                          DISCUSSION

       SUMO proteases share homology in their catalytic domains, but otherwise have widely

divergent primary sequences in their N- and C-terminal domains. We have found that human

SENP2 localizes to the nucleoplasmic face of nuclear pores through sequences in its N-terminal

domain. We further find that this domain of SENP2 interacts with Nup153, a nucleoporin on the

nucleoplasmic face of the nuclear pore. Together, our observations suggest that at least one

function of the divergent regions in SUMO proteases is their proper localization within the

nucleus. Notably, elimination of the N-terminal domain of SENP2 not only allows it to re-

localize to other parts of the nucleus but also increases its capacity to deconjugate SUMO-1




                                                                                                      Downloaded from www.jbc.org by on January 16, 2009
conjugated species within the cell. These observations suggest that targeting SENP2 to the

nuclear pore is a mechanism to sequester it from SUMO-1 conjugated proteins in the nuclear

interior. Alternatively, the N-terminal domain may have a role in negatively regulating SENP2

through additional mechanisms.

       Our findings suggest that interactions between Nup153 and SENP2 may be responsible

for the localization of SENP2 at the nuclear pore (Figure 4). Nup153 is associated with the

basket structure on the nucleoplasmic side of the pore (19), and it has been implicated in multiple

aspects of nuclear transport and pore structure (17, 23). The association between Nup153 and

SENP2 is therefore entirely consistent with immunofluorescence data showing that SENP2 was

found only on the nucleoplasmic side of the nuclear envelope (Figure 1). Ulp1p, a protease for

budding yeast SUMO-1 (Smt3), is found at the nuclear pore (5), and has been reported to bind

Nup42p (6). Like Nup153, Nup42p is an FG-repeat containing nucleoporin (24). These

observations may suggest a conserved role for SUMO-1 deconjugation in the regulation of pore

activity. However, there are likely to be some differences between vertebrates and yeast in the




                                                15
details of this function, since Nup42p is localized on the cytosolic face of the nuclear envelope in

yeast (24).

       Although SENP2 was localized to the nuclear pore, we have not yet found any role for

SENP2 in nuclear trafficking. Overexpression of wild type or mutant versions of SENP2 did not

significantly alter nuclear import or export of a GFP-labeled chimeric model substrate consisting

of HIV-1 Rev and a hormone-inducible nuclear localization sequence (Rev-GR-GFP; (25)) (data

not shown). While such negative results do not exclude the possibility that some nuclear

transport pathways are controlled by SENP2, they suggest that modulation of SENP2 activity

does not grossly alter Nup153’s role in pore structure or nuclear import.




                                                                                                       Downloaded from www.jbc.org by on January 16, 2009
       Interestingly, Pichler et al. (26) have demonstrated that RanBP2 is an E3 enzyme for

SUMO-1 that can catalyze both its own hyperconjugation to SUMO-1 and the conjugation of

Sp100 in vitro. Pichler et al. have proposed a model wherein RanBP2 may couple nuclear import

with the conjugation of a subset of SUMO-1 substrates (26). It would be attractive to speculate

that SENP2 might have a role in transport-linked deconjugation of the same subset of SUMO-1

conjugated proteins. Notably, translocation through the nuclear pore is not essential for their

efficient conjugation of Sp100 by RanBP2 (26). Furthermore, nonconjugatable forms of Sp100

show no defects in nuclear localization in vivo (27), suggesting that SUMO-1 modification

cannot be essential for its nuclear import. These observations further indicate that pore-

associated SUMO-1 conjugation and deconjugation activities are unlikely to be involved in

nuclear transport per se. An alternative idea of their function might be that they specifically mark

newly-imported proteins, perhaps to direct their localization after nuclear entry or to regulate

their activity before their assembly into macromolecular complexes within the nucleus (28).




                                                16
       SENP2 is closely related to a rat SUMO-1 protease, Axam, which has been reported as an

Axin-binding protein (11, 12). Axam antagonizes the binding of Dvl-1 to Axin and suppress

GSK-3β-dependent phosphorylation in the Axin complex, thereby enhancing β-catenin

degradation (12). Enhancement of β-catenin degradation does not require SUMO protease

activity (11), suggesting that SENP2 may have multiple, independent functions. Earlier reports

on Axam also indicated that Axam and SENP2 (SMT3IP2) are localized to the cytosol (11, 12).

We believe that the difference between these observations and ours was caused by the fact that

both reports visualized the localization of N-terminally truncated forms of Axam or SENP2,




                                                                                                       Downloaded from www.jbc.org by on January 16, 2009
missing 72 or 42 amino acids, respectively. We suspect that expression levels and cell type-

specific differences may also contribute to the differences between our observations and those

reports, since we did not observe localization of truncated forms of SENP2 in the cytosol. We

cannot currently explain the relationship between the APC/β-catenin pathway and SENP2, but it

will be of interest to examine whether SENP2 or other SUMO-1 pathway enzymes regulate

nuclear translocation or sub-nuclear localization of any of the proteins of the APC/β-catenin

pathway.

       In summary, we have shown that full length SENP2 localized to the nuclear face of the

nuclear pore, and have presented that this localization is likely to be achieved through interaction

with Nup153. Appropriate localization to the pore plays an important role for the correct

regulation of SENP2, since its mis-localization leads to the inappropriate deconjugation of many

SUMO-1 conjugated species. It will be of interest in the future to determine which SUMO-1

conjugated species are normally the targets of SENP2 activity.




                                                17
ACKNOWLEDGEMENTS - This work was supported by HSFP Research Grant RG0229/1999-M

and by NICHD intramural funds (Project number #Z01 HD 01902-05). We thank Jomon Joseph

and Tadashi Anan for their help in generating constructs expressing tagged human RanGAP1

and SUMO-1. We thank Shyh-Han Tan for help with confocal microscopy. Finally, we thank

Alexei Arnaoutouv, Yoshiaki Azuma, Byrn Booth Quimby and Shyh-Han Tan for critical

reading of this manuscript.




                                                                                           Downloaded from www.jbc.org by on January 16, 2009




                                          18
                                         REFERENCES

1. Melchior, F. (2000) Annu. Rev. Cell Dev. Biol. 16, 591-626

2. Azuma, Y., Tan, S. H., Cavenagh, M. M., Ainsztein, A. M., Saitoh, H., and Dasso, M. (2001)

   FASEB J. 15, 1825-7

3. Li, S. J., and Hochstrasser, M. (1999) Nature 398, 246-51

4. Strunnikov, A. V., Aravind, L., and Koonin, E. V. (2001) Genetics 158, 95-107

5. Li, S. J., and Hochstrasser, M. (2000) Mol. Cell Biol. 20, 2367-77

6. Takahashi, Y., Mizoi, J., Toh, E. A., and Kikuchi, Y. (2000) J. Biochem. (Tokyo) 128, 723-5

7. Yeh, E. T., Gong, L., and Kamitani, T. (2000) Gene 248, 1-14




                                                                                                   Downloaded from www.jbc.org by on January 16, 2009
8. Gong, L., Millas, S., Maul, G. G., and Yeh, E. T. (2000) J. Biol. Chem. 275, 3355-9

9. Kim, K. I., Baek, S. H., Jeon, Y. J., Nishimori, S., Suzuki, T., Uchida, S., Shimbara, N.,

   Saitoh, H., Tanaka, K., and Chung, C. H. (2000) J. Biol. Chem. 275, 14102-6

10. Nishida, T., Tanaka, H., and Yasuda, H. (2000) Eur. J. Biochem. 267, 6423-7

11. Nishida, T., Kaneko, F., Kitagawa, M., and Yasuda, H. (2001) J. Biol. Chem. 276, 39060-6

12. Kadoya, T., Kishida, S., Fukui, A., Hinoi, T., Michiue, T., Asashima, M., and Kikuchi, A.

    (2000) J. Biol. Chem. 275, 37030-7

13. Smythe, C., and Newport, J. W. (1991) Methods Cell Biol. 35, 449-68

14. Adam, S. A., Marr, R. S., and Gerace, L. (1990) J. Cell Biol. 111, 807-16

15. Meier, E., Miller, B. R., and Forbes, D. J. (1995) J. Cell Biol. 129, 1459-72

16. Sukegawa, J., and Blobel, G. (1993) Cell 72, 29-38

17. Ullman, K. S., Shah, S., Powers, M. A., and Forbes, D. J. (1999) Mol. Biol. Cell 10, 649-64

18. Powers, M. A., Forbes, D. J., Dahlberg, J. E., and Lund, E. (1997) J. Cell Biol. 136, 241-50




                                                19
19. Pante, N., Bastos, R., McMorrow, I., Burke, B., and Aebi, U. (1994) J. Cell Bio.l 126, 603-

    17

20. Matunis, M. J., Coutavas, E., and Blobel, G. (1996) J. Cell Biol. 135, 1457-70

21. Mahajan, R., Delphin, C., Guan, T., Gerace, L., and Melchior, F. (1997) Cell 88, 97-107

22. Matunis, M. J., Wu, J., and Blobel, G. (1998) J. Cell Biol. 140, 499-509

23. Walther, T. C., Fornerod, M., Pickersgill, H., Goldberg, M., Allen, T. D., and Mattaj, I. W.

    (2001) EMBO J. 20, 5703-14

24. Vasu, S. K., and Forbes, D. J. (2001) Curr. Opin. Cell Biol. 13, 363-75

25. Love, D. C., Sweitzer, T. D., and Hanover, J. A. (1998) Proc. Natl. Acad. Sci. U. S. A. 95,




                                                                                                   Downloaded from www.jbc.org by on January 16, 2009
    10608-13

26. Pichler, A., Gast, A., Seeler, J. S., Dejean, A., and Melchior, F. (2002) Cell 108, 109-120

27. Sternsdorf, T., Jensen, K., Reich, B., and Will, H. (1999) J. Biol. Chem. 274, 12555-66

28. Senger, B., Simos, G., Bischoff, F. R., Podtelejnikov, A., Mann, M., and Hurt, E. (1998)

    EMBO J. 17, 2196-207.




                                                20
Figure Legends



FIG. 1. SENP2 localizes to the nuclear side of the pore during interphase. (A) HeLa cells

were transiently transfected with eGFP-SENP2, fixed and stained for immunofluorescence using

the monoclonal antibody mAb414, which recognizes components of the nuclear pore. The cells

were also stained using Hoechst 33258 DNA dye. The cells were examined by confocal laser

microscopy. SENP2-eGFP is shown in green, mAb414 staining is shown in red, and DNA is

shown in blue in the right column only. The bottom panels show an enlargement of partial area

of a transfected HeLa cell. (B) HeLa cells were transiently transfected with eGFP-SENP2, and




                                                                                                 Downloaded from www.jbc.org by on January 16, 2009
stained for immunofluorescence using rabbit polyclonal antibodies against the eGFP moiety. The

upper panels shows cells that were permeabilized with digitonin before incubation with primary

antibodies, the lower panels show cells that were permeabilized with Triton X-100.



FIG. 2. Sequences at the N-terminus of SENP2 are required for localization to the nuclear

pore. (A) Schematic representation of eGFP deletion constructs. Association to the pore and

binding to hNup153 data are summarized on the right. N/D indicates binding or association has

not been determined for this construct. The hatched box indicates the conserved isopeptidase

domain. (B) eGFP fusion proteins with the indicated N-terminal deletions of SENP2, were

transfected into HeLa cells. Localization of the fusion proteins was visualized by eGFP

fluorescence (green, left column). DNA stained with Hoechst 33258 dye is shown in the center

column (blue), and merged images are shown in the right column. Similar results were observed

using indirect immunofluorescence with anti-flag antibodies against Flag-tagged SENP2

proteins.




                                              21
FIG. 3. Sequences at the N-terminus of SENP2 are sufficient for localization to the nuclear

pore. eGFP fusion proteins with the indicated segments of SENP2, were transfected into HeLa

cells. Localization of the fusion proteins was visualized by eGFP fluorescence (green, left

column). DNA stained with Hoechst 33258 dye is shown in the center column (blue), and

merged images are shown in the right column. Similar results were observed using indirect

immunofluorescence with anti-flag antibodies against Flag-tagged SENP2 proteins.



FIG. 4. The N-terminus of SENP2 interacts with Nup153. The indicated GST-SENP2 fusion




                                                                                                      Downloaded from www.jbc.org by on January 16, 2009
proteins were incubated with Xenopus egg extracts. Resultant complexes were purified by

affinity to glutathione Sepharose. The purified proteins were subjected to Western blotting

analysis using (upper panel) mAb414, (middle panel) anti-Nup153, and (lower panel) anti-

Nup98 antibodies. In each panel, the lane order is as follows: starting material (lane 1); lane2-6,

GST pull-down products using GST (lane 2), GST-SENP2(101-590) (lane 3), GST-SENP2(201-

590) (lane 4), GST-SENP2(1-170) (lane 5), GST-SENP2(1-70) (lane 6).



FIG. 5. Removal of the Nup153 binding region alters SENP2 activity in vivo. (A) Removal of

the N-terminal targeting domains results in hyperactivity of SENP2 against most SUMO-1

conjugated species. Lane 1-4: EGFP, eGFP-SENP2, eGFP-SENP2(71-590), and eGFP-SENP2-

C/S were co-transfected into COS7 cells with His-Xpress-SUMO-1 (10:1) respectively. Lane 5:

The COS7 cells were transfected with eGFP-SENP2 in absent of His-Xpress-SUMO-1. The total

cell lysates were subjected to Western blotting with antibodies against the Xpress tag (upper

panel) or against the GFP moiety of the eGFP-SENP2 fusion proteins (lower panel). (B)




                                                22
Overexpression of SENP2 does not alter SUMO-1 conjugation of RanGAP1. Plasmids

expressing eGFP-SENP2 was co-transfected into COS7 cells with constructs expressing wild

type or mutant V5-tagged RanGAP1 (10:1), as indicated. The total cell lysates were subjected to

Western blotting with antibodies against the V5 epitope (upper panel) or against the GFP moiety

of the eGFP-SENP2 fusion proteins (lower panel). The unmodified form of RanGAP1 is

identified in the control lane where a RanGAP1 mutant lacking the SUMO-1 modification site

(RanGAP1-K524R) is expressed.




                                                                                                  Downloaded from www.jbc.org by on January 16, 2009




                                              23
         Downloaded from www.jbc.org by on January 16, 2009
merge
mAb414
GFP




                                                              Hang et al., Figure 1A
Downloaded from www.jbc.org by on January 16, 2009
              merge
              DNA
              anti-GFP
              GFP
                     TX 100 digitonin




                                                     Hang et al., Figure 1B
                                                                                     NPC association
                                                                                                       Nup153 binding
                                Downloaded from www.jbc.org by on January 16, 2009
              SENP2-(1-590)                                                            +               N/D
              SENP2-(1-470)                                                            +               N/D
              SENP2-(1-370)                                                            +               N/D
              SENP2-(1-270)                                                            +               N/D
              SENP2-(1-170)                                                            +                +
              SENP2-(1-70)                                                             +                +
              SENP2-(31-590)                                                           -               N/D
              SENP2-(51-590)                                                           -               N/D
              SENP2-(71-590)                                                           -               N/D
              SENP2-(101-590)                                                        N/D                  -
              SENP2-(201-590)                                                        N/D                  -


Hang et al., Figure 2A
                   Downloaded from www.jbc.org by on January 16, 2009
merge
DNA
GFP




                                                                        Hang et al., Figure 2B
        (31-590)      (51-590)                            (71-590)
         SENP2         SENP2                               SENP2
                         GFP   DNA   merge
               SENP2
               (1-470)



               SENP2
               (1-370)




                                             Downloaded from www.jbc.org by on January 16, 2009
               SENP2
               (1-270)



               SENP2
               (1-170)



               SENP2
                (1-70)




Hang et al., Figure 3
                           1 2 3 4   5 6
                RanBP2
                Nup214
                 Nup153
                                           WB: mAb414




                                                          Downloaded from www.jbc.org by on January 16, 2009
                   Nup60



                 Nup153                    WB: α-Nup153

                   Nup98                   WB: α-Nup98




Hang et al., Figure 4
               kDa        1   2   3   4   5




                                                             Downloaded from www.jbc.org by on January 16, 2009
               199                            WB: α-Xpress

               133


                87




                                              WB: α-GFP




Hang et al., Figure 5 A
     eGFP                +   +   -   -
     eGFP-Senp2          -   -   +   +
     V5-RanGAP1          -   +   -   +
     V5-RanGAP1 K524R    +   -   +   -
      SUMO-RanGAP1
                                         WB: α-V5
               RanGAP1




                                                     Downloaded from www.jbc.org by on January 16, 2009
                                         WB: α-GFP




Hang et al., Figure 5B

				
DOCUMENT INFO
Shared By:
Tags: strategene
Stats:
views:49
posted:1/16/2009
language:English
pages:31