Download Nuclear Localization of the Parafibromin Tumor Suppressor Protein

Nuclear Localization of the Parafibromin Tumor Suppressor
Protein Implicated in the Hyperparathyroidism-Jaw Tumor
Syndrome Enhances Its Proapoptotic Function
Ling Lin,1 Meggan Czapiga,2 Lylia Nini,1 Jian-Hua Zhang,1 and William F. Simonds1
1
Metabolic Diseases Branch, National Institute of Diabetes and Digestive and Kidney Diseases
and 2Research Technologies Branch, National Institute of Allergy and
Infectious Diseases, NIH, Bethesda, Maryland
Abstract
Introduction
Parafibromin is a tumor suppressor protein encoded by
HRPT2, a gene recently implicated in the hereditary
hyperparathyroidism-jaw tumor syndrome, parathyroid
cancer, and a subset of kindreds with familial isolated
hyperparathyroidism. Human parafibromin binds to
RNA polymerase II as part of a PAF1 transcriptional
regulatory complex. The mechanism by which loss
of parafibromin function can lead to neoplastic
transformation is poorly understood. Because the
subcellular localization of parafibromin is likely to
be critical for its function with the nuclear PAF1
complex, we sought to experimentally define the
nuclear localization signal (NLS) of parafibromin and
examine its potential role in parafibromin function.
Using site-directed mutagenesis, we define a dominant
bipartite NLS and a secondary NLS, both in the
NH2-terminal region of parafibromin whose combined
mutation nearly abolishes nuclear targeting. The
NLS-mutant parafibromin is significantly impaired in
its association with endogenous Paf1 and Leo1. We
further report that overexpression of wild-type but
not NLS-mutant parafibromin induces apoptosis in
transfected cells. Inhibition of endogenous parafibromin
expression by RNA interference inhibits the basal
rate of apoptosis and apoptosis resulting from DNA
damage induced by camptothecin, a topoisomerase I
inhibitor. These experiments identify for the first time a
proapoptotic activity of endogenous parafibromin likely
to be important in its role as a tumor suppressor and
show a functional role for the NLS of parafibromin in this
activity. (Mol Cancer Res 2007;5(2):183 – 93)
The tumor suppressor gene HRPT2 was recently identified
by positional candidate cloning (1). Mutation of HRPT2 in the
germ line confers susceptibility to the hyperparathyroidismjaw tumor syndrome (HPT-JT), an autosomal-dominant syndrome with high but incomplete penetrance (2). The major
features are primary hyperparathyroidism (90%) including 15%
of all affected by HPT-JT with parathyroid cancer, jaw tumors
(30%), bilateral renal cysts (10%), and less commonly solid
renal tumors (2-7). Germline inactivating HRPT2 mutation has
also been reported in a minority of kindreds with familial
isolated hyperparathyroidism (1, 7-9) and, unexpectedly, in up
to 30% of patients with apparently sporadic parathyroid cancer
(10, 11).
HRPT2 encodes parafibromin, a 531-amino-acid putative
tumor suppressor protein. The COOH-terminal region of
parafibromin shows sequence homology to Cdc73p, a budding
yeast protein component of the RNA polymerase II – associated
Paf1 complex. Indeed, recent evidence suggests that in humans
parafibromin interacts with RNA polymerase II via a human
PAF1 complex, whose other protein components include
human Paf1, CTR9, Leo1 (12-14), and the WD40-repeat
protein Ski8 (14). More recently, a domain in the NH2-terminal
region of parafibromin has been reported to interact with
h-catenin and mediate a functional interaction between the Wnt
signaling and PAF1 transcriptional regulatory complexes (15).
The subcellular localization of endogenous parafibromin
has been described as nuclear in HeLa cells (13) and both
nuclear and cytoplasmic in human parathyroid tissue (16) and
fibro-osseous jaw tumors (17). The subcellular localization of
parafibromin in the nucleus is likely to be critical for its
function with the PAF1 complex, and several potential nuclear
localization signals (NLS) in its primary structure have been
proposed (13, 18). The aim of this study was to experimentally
define the NLS of parafibromin and examine its potential role
in parafibromin function.
Received 5/10/06; revised 11/17/06; accepted 12/12/06.
Grant support: Intramural Research Program of the National Institute of
Diabetes and Digestive and Kidney Diseases (W.F. Simonds) and National
Institute of Allergy and Infectious Diseases (M. Czapiga), NIH.
The costs of publication of this article were defrayed in part by the payment of
page charges. This article must therefore be hereby marked advertisement in
accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Requests for reprints: William F. Simonds, Metabolic Diseases Branch/National
Institutes of Diabetes, Digestive and Kidney Diseases, NIH, Room 8C-101,
Building 10, 10 Center Drive, MSC 1752, Bethesda, MD 20892-1752. Phone:
301-496-9299; Fax: 301-402-0374. E-mail: [email protected]
Copyright D 2007 American Association for Cancer Research.
doi:10.1158/1541-7786.MCR-06-0129
Results
To facilitate the identification of a functional NLS on human
parafibromin, an expression construct was generated in which
the tumor suppressor protein was fused in frame to green
fluorescent protein (GFP). This NH2-terminally GFP-tagged
parafibromin construct (GFP-HRPT2) was expressed and
strongly localized to the cell nucleus of transfected HeLa cells
as seen in laser confocal microscopic imaging of cells
counterstained with Hoechst 33342 DNA-intercalating dye
Mol Cancer Res 2007;5(2). February 2007
Downloaded from mcr.aacrjournals.org on August 11, 2017. © 2007 American Association for Cancer
Research.
183
184 Lin et al.
(Fig. 1A, compare top two). Similar nuclear localization of
GFP-HRPT2 was found in transfected HEK-293 kidney cells
and U2-OS osteosarcoma epithelial cells upon confocal analysis
(Fig. 1A, bottom two). Analysis of transfected HeLa cells by
subcellular fractionation and immunoblotting also showed the
strong nuclear localization of GFP-HRPT2 in contrast to the
mainly cytoplasmic distribution of the GFP control (Fig. 1B).
Immunoblotting for lamin B and tubulin validated the quality of
the nuclear and cytoplasmic fractions, respectively (Fig. 1B).
Nuclear localization of GFP-tagged parafibromin was previously described when the HRPT2 gene product was known
only as c1orf28 before its implication in HPT-JT (19).
To identify the NLS conferring nuclear localization to GFPHRPT2, a series of NH2-terminal and COOH-terminal truncation mutants of parafibromin were prepared as GFP fusion
proteins, and their subcellular distribution was determined by
laser confocal microscopy (Fig. 2A-C). The first GFP fusion
mutants analyzed consisted of an NH2-terminal parafibromin
fragment ending at residue 235 (GFP-HRPT2-B) and a COOHterminal fragment encompassing residues 230 to 531 (GFPHRPT2-C; Fig. 2A-C). The latter fragment of parafibromin
contains the region from residues 233 to 525 homologous to
cdc73p, the yeast RNA polymerase II accessory factor
component of the Paf1 complex (20, 21). Because yeast cdc73p
FIGURE 1. Human parafibromin-GFP fusion proteins
are localized to the nucleus in
multiple human cell lines. A.
The nuclear localization of
GFP-fused parafibromin is
revealed by confocal laser microscopy in HeLa, HEK-293,
and U2-OS cells. Images
corresponding to GFP fluorescence, nuclear staining by
Hoechst dye, differential interference contrast, and the
merged images of fluorescence with Hoechst staining
of the same fields are indicated.
The expression of pEGFP-C1
(expressing GFP alone) is
shown as a control. B. Subcellular localization of EGFPHRPT2 as indicated by cell
fractionation and Western blot
analysis. HeLa cells were
transiently transfected with
pEGFP-C1 vector or EGFPHRPT2 . Twenty-four hours
after transfection, cells were
fractionated, and the protein
expression in whole-cell extract (W ), nuclear (N), and
cytoplasmic (C) fractions were
analyzed by Western blot with
anti-GFP antibody, anti-Lamin
B antibody, and anti-a-tubulin
antibody. IB, immunoblot.
Mol Cancer Res 2007;5(2). February 2007
Downloaded from mcr.aacrjournals.org on August 11, 2017. © 2007 American Association for Cancer
Research.
Parafibromin Nuclear Localization and Apoptosis
FIGURE 2. Mapping of the functional NLS on human parafibromin by deletion mutational analysis. A. Schematic diagrams of the wild-type GFP-tagged
HRPT2 and 10 deletion constructs assessed for NLS function. NH2 or COOH terminus of each fragment, with numbers referring to the positions in wild-type
un-fused parafibromin. Right, presence (+) or absence ( ) of green fluorescence observed in the nucleus (Nu ) or cytoplasm (Cy ) for each constructs.
B. Protein expression of the full-length GFP-HRPT2 and its NH2 or COOH terminus truncation constructs. HeLa cells were transiently transfected with each
constructs as in (A), and protein expression was confirmed by Western blotting employing anti-GFP antibody. C. Laser confocal microscopic analysis of the
subcellular localization of wild-type EGFP-HRPT2 and its deletion mutants illustrated in (A). Representative micrographs of the fluorescence analysis. Right,
differential interference contrast is shown. D. Schematic diagram of parafibromin showing that NLS activity is restricted to the region between amino acids
115 and 160 (red box) based on the fine mapping as in (A), (B), and (C). The putative bipartite NLS sequence from residues 125 to 139 of human
parafibromin is shown using the single-letter amino acid code, with spacer amino acids in lower case.
and the other yeast Paf1 complex components are nuclear
proteins (22), and because human parafibromin (hCdc73) has
been identified as a component of a human PAF1 complex (1214), it was expected that GFP-HRPT2-C would exhibit nuclear
localization like the full-length parafibromin fusion (GFPHRPT2-A). Surprisingly, it was the NH2-terminal fusion GFPHRPT2-B, and not the COOH-terminal cdc73-homologous
parafibromin fusion GFP-HRPT2-C, that showed nuclear
localization like the parental full-length fusion (Fig. 2A-C).
To define the NH2-terminal boundary of the NLS, a series of
NH2-terminally truncated mutants starting with residues
ranging from 75 to 200 were constructed as GFP fusion
proteins (GFP-HRPT2 mutants D-H, Fig. 2A-C). GFP fusion
mutants D and E showed strong nuclear localization and
began with parafibromin residues 75 and 115, respectively
(Fig. 2A-C). In contrast, GFP mutants F to H, whose
parafibromin sequences began at residue 155 or higher, showed
substantial cytoplasmic expression, unlike the full-length
parafibromin fusion protein GFP-HRPT2-A (Fig. 2A-C). These
results suggested the NH2-terminal boundary of the putative
NLS was between residues 115 and 155 of parafibromin.
Definition of the COOH-terminal boundary of the NLS
involved analysis of a series of three COOH-terminally
truncated GFP fusion mutants that ended with parafibromin
residues 160, 116, and 75 (GFP-HRPT2 mutants I-K,
respectively; Fig. 2A-C). Among these three constructs, only
fusion mutant GFP-HRPT2-I showed strong nuclear localization like the full-length parental parafibromin fusion, suggesting that the putative NLS was NH2 terminal to residue 160
(Fig. 2A-C).
Mol Cancer Res 2007;5(2). February 2007
Downloaded from mcr.aacrjournals.org on August 11, 2017. © 2007 American Association for Cancer
Research.
185
186 Lin et al.
Taken together, analysis of the NH2-terminal and COOHterminal parafibromin truncation mutant GFP fusion constructs
suggested that the putative NLS was located between residues
115 and 160 (Fig. 2D). Within this region, there was a sequence
from residues 125 to 139 with characteristics strongly suggestive
of a bipartite NLS consisting of two clusters of
basic amino acids separated by a short stretch of ‘‘spacer’’
amino acids (refs. 23, 24; Fig. 2D).
To test the importance of this putative bipartite NLS for
parafibromin nuclear localization, the basic residues on either
side of the spacer amino acids were mutated either alone or in
combination, and the effect on subcellular localization was
assessed by laser confocal microscopy. These mutations were
introduced in the context of the full-length GFP parafibromin
fusion construct GFP-HRPT2 (Fig. 3A and B). Single mutation
of basic residues in either the NH2-terminal cluster, such as
K125A or R126A, or in the COOH-terminal basic cluster, such
as K136A, K137A, or R139A, had little effect on the strong
nuclear localization of GFP-HRPT2 (Fig. 3A). In contrast,
multiple mutation of residues in the COOH-terminal basic
cluster (K136A/K137A/R139A) or else paired single mutations
involving both NH2-terminal and COOH-terminal basic clusters
(R126A/K137A and R126A/R139A) were sufficient to disrupt
nuclear localization of the GFP-HRPT2 fusion (Fig. 3A and B).
These data strongly suggest that the bipartite cluster of basic
amino acids spanning residues 125 to 139 functions as a
dominant NLS in the human tumor suppressor parafibromin.
Other potential bipartite NLS in parafibromin have been
previously noted, however, at residues 76 to 92 and 393 to
409 (13, 18). Because mutants of the dominant NLS identified
here, such as R126A/K137A and R126A/R139A, still retained
significant nuclear localization (Fig. 3A-C), we hypothesized
that secondary potential NLS, such as the clusters of basic
amino acids at residues 76 to 92 and 393 to 409, might
contribute to nuclear localization, especially upon inactivation
of the dominant NLS.
To test the importance of putative secondary NLS to the
nuclear localization of parafibromin, we mutated basic amino
acids present within residues 76 to 92 (N-cluster) or residues
393 to 409 (C-cluster) either alone or in combination with the
R126A/R139A mutations of the dominant NLS in the context
of the GFP-parafibromin fusion construct (Fig. 4A). Mutation
of either the N-cluster or C-cluster alone did not impair the
strong nuclear localization of GFP-HRPT2 (Fig. 4A; data not
shown), suggesting that neither basic amino acid cluster plays
a dominant role as a NLS for parafibromin. This finding was
consistent with the experiments presented above employing
NH2-terminal and COOH-terminal GFP-parafibromin truncation mutants that mapped the dominant NLS between residues
115 and 160 (Fig. 2A-D). Combining the C-cluster mutation
with the R126A/R139A NLS mutations resulted in a GFP
fusion with mixed cytoplasmic and nuclear localization no
different than the R126A/R139A NLS mutant alone (Fig. 4A;
data not shown). In contrast, combining the N-cluster mutation
with the R126A/R139A NLS mutations nearly abolished
nuclear localization of the resulting GFP-parafibromin fusion
protein altogether (Fig. 4A and B). This finding suggests that
the cluster of NH2-terminal basic amino acids between residues
76 and 92 comprises a secondary NLS that can function if the
dominant NLS is impaired. This combined R126A/R139A/Ncluster parafibromin mutant with severely impaired nuclear
localization was therefore used along with the wild-type protein
to assess the importance of parafibromin nuclear localization to
its function.
Interactions with the other components of the PAF1
transcriptional regulatory complex are likely to require the
nuclear localization of parafibromin; thus, we compared the
ability of wild-type parafibromin and the R126A/R139A/
N-cluster mutant to interact with endogenous Paf1 and Leo1
in transfected HeLa cells. Endogenous Leo1 and Paf1 readily
coimmunoprecipitated with transfected epitope-tagged wildtype parafibromin using AU5 monoclonal antibody but not with
control mouse IgG (Fig. 4C), confirming previous reports
(12-14). In contrast, when the AU5-tagged R126A/R139A/
N-cluster mutant was immunoprecipitated, little or no endogenous Leo1 or Paf1 was coimmunoprecipitated (Fig. 4C).
Quantification and averaging of immunoblots taken from
multiple experiments like that shown in Fig. 4C showed a
highly significant f80% reduction of Leo1 and Paf1 that
coimmunoprecipitated with the parafibromin mutant compared
with the wild-type (Fig. 4D). Consistent with the lack of
interaction evidenced by the coimmunoprecipitation studies,
overexpression of the R126A/R139A/N-cluster parafibromin
mutant did not alter the nuclear localization of endogenous
Leo1 in immunofluorescence studies (data not shown). Because
the binding domain on parafibromin for Leo1 and Paf1 has
been mapped to residues 228 to 531 (12) and residues 200 to
531 (13), respectively, our findings suggest that disruption of
the NLS on parafibromin significantly reduces its functional
interaction with components of the PAF1 complex by impairing
its nuclear targeting.
During the course of the laser confocal microscopy studies,
morphologic changes, including nuclear fragmentation
(Fig. 5A, middle) and nuclear condensation (Fig. 5A, bottom),
were often seen in cells transfected with the GFP-wild type
parafibromin fusion construct but not in the GFP-only transfected cells (Fig. 5A, top). Chromatin fragmentation and
nuclear condensation are morphologic changes characteristic
of apoptosis. Because parafibromin functions as a tumor
suppressor, and because other tumor suppressors like p53 have
proapoptotic properties (25), we examined other indices of
apoptosis in cells transfected with wild-type and the R126A/
R139A/N-cluster parafibromin mutant using p53 as a positive
control. Transfection of wild-type parafibromin significantly
increased caspase-3 activity and elevated cytoplasmic concentrations of histone-associated oligonucleosomes, processes
indicative of programmed cell death (Fig. 5B and C). In
contrast, cells transfected with the R126A/R139A/N-cluster
mutant parafibromin showed no difference from the empty
vector – transfected control cells in apoptotic activity (Fig. 5B
and C).
Because the stress associated with the overexpression of
heterologous protein might conceivably trigger programmed
cell death in transfected cells, we sought evidence that
endogenous parafibromin could promote apoptosis under
certain cellular conditions. To this end, apoptosis was
monitored after RNA interference to reduce endogenous
levels of parafibromin in HeLa cells treated with small
Mol Cancer Res 2007;5(2). February 2007
Downloaded from mcr.aacrjournals.org on August 11, 2017. © 2007 American Association for Cancer
Research.
Parafibromin Nuclear Localization and Apoptosis
FIGURE 3. Nuclear localization of
the full-length HRPT2 is dependent on
a bipartite NLS. A. Diagram of the fulllength HRPT2 (top ). Asterisks, positively charged amino acid residues in
the putative bipartite NLS of HRPT2 .
The amino acid residues replaced by
alanine in the indicated NLS point
mutants of HRPT2 are below the
native sequence (A ). Dots, unchanged amino acids. Right, presence
(+) or absence ( ) of green fluorescence observed in the nucleus or
cytoplasm for each constructs. B.
Representative laser confocal micrographs of EGFP-HRPT2 (R126A/
R139A) and EGFP-HRPT2 (R126A/
K137A) mutants in comparison with
the wild-type EGFP-HRPT2 . Third
column, differential interference contrast is shown. C. Subcellular localization of double point mutant EGFPHRPT2 (R126A/R139A) as indicated
by Western blotting. HeLa cells were
transiently transfected with EGFPHRPT2 (R126A/R139A) and analyzed
as in Fig. 1B legend.
Mol Cancer Res 2007;5(2). February 2007
Downloaded from mcr.aacrjournals.org on August 11, 2017. © 2007 American Association for Cancer
Research.
187
188 Lin et al.
interfering RNA (siRNA) targeting two different regions of
human parafibromin mRNA (Fig. 6A and B). The level of
endogenous parafibromin protein expression as well as the
cytoplasmic concentration of histone-associated oligonucleosomes as an index of apoptosis was monitored in response to
treatment with control or anti-HRPT2 siRNA (Fig. 6A and
B). Both basal apoptosis and that resulting from DNA
damage caused by treatment of the cells with two different
concentrations of the topoisomerase I inhibitor camptothecin
was determined. Both parafibromin-directed siRNA con-
Mol Cancer Res 2007;5(2). February 2007
Downloaded from mcr.aacrjournals.org on August 11, 2017. © 2007 American Association for Cancer
Research.
Parafibromin Nuclear Localization and Apoptosis
structs, but not the control siRNA construct, inhibited the
expression of endogenous HeLa parafibromin protein and
significantly reduced basal and camptothecin-induced apoptosis in HeLa cells (Fig. 6A and B).
Discussion
Disease-associated HRPT2 mutation uniformly predicts loss
of parafibromin expression or function (1, 7-10), supporting the
view that parafibromin is a tumor suppressor protein. Although
the mechanism by which loss of parafibromin function
promotes neoplasia is unknown, a deeper understanding of its
cell biology and molecular interactions is likely to provide
important clues. The subcellular localization of endogenous
parafibromin has been reported to be nuclear (13) or as both
nuclear and cytoplasmic (16, 17). Its nuclear targeting is likely
to be critical for its function with the PAF1 complex (12-14). In
the present work, we experimentally define a dominant bipartite
NLS comprising residues 125 to 139 of parafibromin distinct
from previously predicted NLS at residues 76 to 92 and 393 to
409 (13, 18). We show, however, that the cluster of basic amino
acids within residues 76 to 92 can function as a secondary NLS
when the dominant NLS is inactivated by mutation. It is
significant that the dominant and secondary NLS sequences we
identify in the NH2-terminal region of parafibromin, preceding
the COOH-terminal cdc73-homologous domain, are lost by
5¶-truncation or frameshift HRPT2 mutations that frequently
occur in human parathyroid tumors (1, 7-10). Our findings
support the recent results of Hahn and Marsh who identified the
same bipartite NLS in parafibromin using a similar experimental approach and whose report appeared while this work was
being completed (26).
The identification and functional characterization of the NLS
in human parafibromin provides one clue to its mechanism of
action as a tumor suppressor. The nuclear localization of
parafibromin is expected to be important for its PAF1 complexdependent functions, and we were indeed able to show a
significant loss of Paf1 and Leo1 association by the NLSmutant parafibromin in immunoprecipitation assays. The NLS
identified here is evolutionarily conserved among metazoan
parafibromin homologues (26). Interestingly, this bipartite NLS
maps to the NH2-terminal domain of parafibromin present in
the cdc73 homologues of metazoans but lacking in yeast cdc73.
Because yeast cdc73 functions in the nucleus where its role in
the PAF1 complex was first defined (21), its mechanism of
nuclear localization must differ from that of parafibromin. This
raises the possibility that the evolution of the non-cdc73
homologous NH2-terminal extension of parafibromin and its
metazoan homologues may have been driven, at least in part, by
the need for an alternative mechanism of nuclear targeting.
Using several different assays, we show here that wild-type
but not NLS-mutant parafibromin promotes apoptosis in
transfected cells. We previously showed that transfected
parafibromin strongly inhibited cell growth in proliferation
assays (16). We extend these observations now with the use of
RNA interference to inhibit the expression of endogenous
parafibromin and show for the first time that parafibromin may
play a role in the initiation and/or execution of apoptosis in
response to DNA damage and perhaps other cytotoxic stimuli.
It is interesting to note that an approximate 50% reduction in
parafibromin protein expression was sufficient to significantly
inhibit HeLa cell apoptosis in our assay (Fig. 6). This may
represent a phenotype of parafibromin haploinsufficiency in
this process, although both the apoptosis and immunoblotting
assays measure populations of, and not individual, cells.
Haploinsufficiency of tumor suppressor genes may contribute
to neoplastic progression in some tissues (27).
Proapoptotic and antiproliferative properties are common to
many tumor suppressor proteins, and the inactivation of
parafibromin in human parathyroid and other tumors is likely
to promote neoplasia at least in part because of such lossof-function. It is interesting to note that Yart et al. found
parafibromin and other PAF1 complex components tightly
bound to the unconventional prefoldin RPB5 interactor
(URI) scaffolding protein (13). Recent data suggest that the
homologue of URI in Caenorhabditis elegans may participate
in the suppression of endogenous DNA damage and play a
role in the maintenance of DNA integrity (28). It is tempting
to speculate that parafibromin might function on one hand to
help suppress endogenous DNA damage through its interaction with the URI complex and on the other hand to promote
apoptosis under circumstances of overwhelming genotoxic
stress.
Materials and Methods
Cell Culture and Transfection
Human cervical carcinoma cell line HeLa, human embryonic kidney cell line HEK293, human osteosarcoma cell line
U2-OS, and mouse NIH/3T3 cells were from the American
Type Culture Collection (Manassas, VA). All of the cell lines
FIGURE 4. Mutation of both the dominant and a secondary NLS in the NH2-terminal region abolishes the nuclear localization of parafibromin and its
interaction with PAF1 components. A. Diagram of the full-length HRPT2 (top ) with the dominant NLS (residues 125-139) and putative secondary NLSs in the
NH2-terminal region (N-cluster, residues 76-92) and COOH-terminal region (C-cluster, residues 393-409) in red. Mutation of the N-cluster encompassed
R76A/R77S/R91A/K92A and of the C-cluster encompassed the K393A/K394S/R407A/R408S/K409A mutations. Bottom left, Protein expression level of the
indicated constructs. Bottom right, presence (+) or absence ( ) of green fluorescence observed in the nucleus or cytoplasm for each constructs.
B. Representative laser confocal micrographs of EGFP-HRPT2 (R126A/R139A/N-cluster mutant) in comparison with the wild-type EGFP-HRPT2 . Third
column of confocal images shows differential interference contrast. C. The amount of endogenous Paf1 and Leo1 interacting with parafibromin is diminished
by the R126A/R139A/N-cluster mutations. Twenty-four hours after transfection with pcDNA3, AU5-HRPT2 , or the mutant AU5-HRPT2 (R126A/R139A/Ncluster), HeLa cell extracts were subjected to immunoprecipitation as described in Materials and Methods. Expression of parafibromin, Leo1, and Paf1 in the
immunoprecipitates was analyzed by immunoblotting (top ). Expression of transfected AU5-parafibromin and endogenous levels of Leo1 and Paf1 in the
whole-cell extracts (bottom ). D. Quantitative analysis of endogenous Leo1 protein (top ) or Paf1 protein (bottom ) in AU5 immunoprecipitates of HeLa cells
transiently transfected with AU5-HRPT2 or AU5-HRPT2 (R126A/R139A/N-cluster mutations). The intensity of AU5-reactive bands in the immunoprecipitates
(such as those in C) was estimated by densitometric scanning. The ratios were derived from the intensities of Leo1 or Paf1 protein versus those of AU5parafibromin in the same immunoprecipitates. Columns, average ratio from triplicate experiments; bars, SD. **, P < 0.01, significance of differences
evaluated by Student’s t test.
Mol Cancer Res 2007;5(2). February 2007
Downloaded from mcr.aacrjournals.org on August 11, 2017. © 2007 American Association for Cancer
Research.
189
190 Lin et al.
FIGURE 5. Induction of apoptosis by the overexpression of parafibromin. A. Overexpression of
parafibromin in NIH/3T3 cells causes an increase in
the number of apoptotic cells. NIH/3T3 cells were
transfected with pEGFP or wild-type pEGFP-HRPT2
and examined by fluorescence microscopy. Nuclear
morphology was analyzed after counterstaining with
Hoechst 33342. Many cells expressing wild-type GFPHRPT2 but not unfused GFP exhibited nuclear
fragmentation (middle ) or condensation (bottom )
as indicated by the arrowheads. Fold-induction of
caspase-3 activity (B) and DNA fragmentation (C)
resulted from overexpression of wild-type or mutant
parafibromin. NIH/3T3 cells were transiently transfected with pcDNA3 vector only, AU5-HRPT2 , AU5HRPT2 (R126A/R139A/N-cluster mutations), or
FLAG-p53. Forty-eight hours after transfection, caspase-3 activity or cytosolic histone-associated DNA
fragmentation were assayed as described in Materials
and Methods. Results are shown as fold induction (B)
or the activity level (C) compared with cells transfected with pcDNA3 vector only. Representative of two
or three independent experiments with similar results.
*, P < 0.05, significance of differences evaluated by
Student’s t test.
used were maintained at 37jC in DMEM supplemented with
10% fetal bovine serum, 2 mmol/L glutamine, and 100 units/mL
penicillin and streptomycin (Invitrogen, Carlsbad, CA). The
cells were transfected at 70% confluence 20 to 24 h after
seeding on dishes using the LipofectAMINE 2000 (Invitrogen)
method according to the manufacturer’s instructions.
Plasmid Constructs
Plasmid AU5-HRPT2 expressing the full-length human
parafibromin protein tagged with double AU5 epitope was
described previously (16). For construction of the NH2terminally GFP-tagged HRPT2 plasmid (EGFP-HRPT2), residues 2 to 531 of full-length human HRPT2 were amplified by
PCR using oligonucleotide forward primer 5¶-GAGTCGACGCGGACGTGCTTAGCGTCCTGCGA-3¶ (the SalI site
is underlined) and reverse primer 5¶-GAGTCGACTCAGAATCTCAAGTGCGATTTATGCTT-3¶ (the SalI site is underlined) using pAU5-HRPT2 DNA as the template. The amplified
DNA was digested with SalI and then inserted into the same site
of pEGFP-C1 (Clontech, Mountain View, CA). Deletion
constructs of GFP-tagged HRPT2 (EGFP-HRPT2-B-K) were
generated by PCR amplification using wild-type construct as
template and appropriate primer pairs, including the artificial
SalI restriction enzyme site. The protein coding regions of all of
the cDNA constructs were verified by DNA sequencing. The
FLAG-p53 construct was a generous gift from Dr. Akira
Nakagawara (Chiba University, Chiba, Japan).
Mol Cancer Res 2007;5(2). February 2007
Downloaded from mcr.aacrjournals.org on August 11, 2017. © 2007 American Association for Cancer
Research.
Parafibromin Nuclear Localization and Apoptosis
Confocal Imaging Analysis
Cells seeded on glass coverslips in 12-well plates were
transfected with GFP-tagged expression constructs. Twenty-four
hours after transfection, the cells were fixed with 2% formaldehyde
in PBS for 15 min at room temperature and then stained with
Hoechst dye for visualization of the nuclei. The coverslips were
mounted with PermaFluor (Immunon, Pittsburgh, PA). Confocal
images were collected on a Leica SP2-UV 405 confocal microscope (Leica Microsystems, Exton, PA) using a 63 oil immersion
objective (numerical aperture = 1.4). Fluorochromes were excited
using a 405-nm diode laser for Hoechst 33342 and a 488-nm laser
for GFP. To avoid possible crosstalk, the two wavelengths were
collected separately and later merged. Differential interference
contrast (DIC) images were collected simultaneously with the
fluorescence images using the transmitted light detector. Images
were processed using Leica TCS-SP software (version 2.5.1347)
and Adobe Photoshop CS2 (Adobe Systems).
Subcellular Fractionation
HeLa cells transiently transfected with pEGFP C1, wild-type,
or mutant EGFP-HRPT2 were fractionated into cytosolic and
nuclear fraction using NE-PER Nuclear and Cytoplasmic
Extraction Reagents (78833; Pierce Biotechnology, Rockford,
IL) according to the manufacturer’s instructions. Equal protein
amounts of the nuclear and cytoplasmic fractions along with
whole-cell extract were subjected to immunoblot analysis. The
immunoblots were also probed with monoclonal antibodies
specific for Lamin B or a-tubulin to indicate the validity of the
fractionation method.
Western Blot Analysis and Immunoprecipitation
Transfected HeLa cells were washed and collected by
centrifugation. The cell pellets were suspended in lysis buffer
[25 mmol/L Tris-HCl (pH 7.5), 137 mmol/L NaCl, 2.7 mmol/L
KCl, 1% Triton X-100, and 1 mmol/L phenylmethylsulfonyl
fluoride] and were incubated on ice for 30 min and sonicated.
After centrifugation at 13,000 g at 4jC for 5 min, the resulting
soluble supernatant was used as the whole-cell extract. The
protein concentrations were determined by the Bradford protein
assay (Bio-Rad, Hercules, CA) using bovine serum albumin as a
standard. The proteins separated by SDS-PAGE were transferred
by electrophoresis from the gels onto a polyvinylidene difluoride
or nitrocellulose membrane (Invitrogen). The membranes were
blotted with appropriately diluted anti-GFP antibody (sc-9996;
Santa Cruz Biotechnology, Inc., Santa Cruz, CA), anti-Lamin B
antibody (NA12, EMD Biosciences, San Diego, CA), anti-atubulin antibody (sc-8035, Santa Cruz Biotechnology), anti-hactin antibody (A5316, Sigma, St. Louis, MO), anti-AU5
antibody (MMS-135R, Covance Research Products, Denver,
PA), polyclonal anti-Leo1 antibody (A300-175A; BETHYL
Laboratories, Inc., Montgomery, TX), polyclonal anti-Paf1
antibody (A300-172A; BETHYL Laboratories), and polyclonal
anti-parafibromin antibody [GRAPE antibody, raised in rabbits
against a maleimide-activated mcKLH (Pierce Biotechnology)
conjugate of the synthetic peptide CGRAPEQRPAPNAAPVDPTLRTKQPIPAAYNRYDQERFK-amide encompassing
residues 262-299 of human/mouse parafibromin] and the
appropriate horseradish peroxidase – conjugated secondary antibodies. Protein bands were visualized using chemiluminescence
substrates (enhanced chemiluminescence; GE Healthcare
FIGURE 6. Parafibromin is involved in the
apoptotic response to DNA damage in HeLa cells.
A. Inhibition of parafibromin expression by RNA
interference using HRPT2 -specific siRNA blocks
both basal apoptosis and that induced by DNAdamaging reagents. Cultured HeLa cells were
treated with control siRNA or one of two HRPT2 specific siRNA constructs as described in Materials and Methods. Fifteen hours before assay of
cytosolic histone-associated DNA fragmentation,
cells were treated with 0, 2, or 5 Amol/L of the
topoisomerase I inhibitor and DNA-damaging
agent camptothecin (CPT) as indicated. Representative of three independent experiments with
similar results. *, P <0.05; **, P < 0.01; ***, P <
0.001, significance of differences evaluated by
Student’s t test. B. The expression of endogenous
parafibromin in the HeLa cells following treatment
with control or parafibromin-targeting siRNAs is
shown by immunoblotting as is the expression of
h-actin in the same samples as a loading control.
Mol Cancer Res 2007;5(2). February 2007
Downloaded from mcr.aacrjournals.org on August 11, 2017. © 2007 American Association for Cancer
Research.
191
192 Lin et al.
Biosciences, Piscataway, NJ). For immunoprecipitation, transfected HeLa cells were lysed in lysis buffer containing 50
mmol/L Tris-HCl (pH 7.5), 120 mmol/L NaCl, 0.5% NP40, 1
mmol/L phenylmethylsulfonyl fluoride, and 1 protease
inhibitor mixture (Roche Applied Science, Indianapolis, IN).
In brief, cell lysates were precleared by incubation with
protein G-Sepharose 4 Fast Flow (GE Healthcare Biosciences)
for 30 min at 4jC with gentle rotation and then incubated
with either monoclonal anti-AU5 antibody or control mouse
IgG for 2 h at 4jC followed by incubation with protein GSepharose 4 Fast Flow for 1 h. The washed immunoprecipitates were subjected to Western blotting.
24 h. At that time, control (Stealth RNAi Negative Control,
Invitrogen) or two siRNAs targeting human parafibromin
(sequences of siRNA1 and siRNA2 were 5¶-TGAACAGATTAGGTCTTTGTCTGAA-3¶ and 5¶-GGGCACTGCAATTAGTGTTACAGTA-3¶, respectively) were transfected by
using LipofectAMINE 2000 reagent (Invitrogen). The DNAdamaging reagent camptothecin (CPT; C9911, Sigma) was
added to cell culture medium 36 h after transfection at a
concentration of 0, 2, or 5 Amol/L for 15 h. Cells were collected
for immunoblotting or for the measurement of apoptotic
induction using the Cell Death Detection ELISA kit (Roche
Applied Science) as described above.
Site-Directed Mutagenesis
The amino acid residues of the dominant NLS of HRPT2
(K125, R126, K136, K137, and R139) were mutated to alanine
individually employing QuikChange II site-directed mutagenesis kit (Stratagene, La Jolla, CA) using Pfu high-fidelity
polymerase and EGFP-HRPT2 or AU5-HRPT2 plasmid as
template. The double (K125A/R126A, R126A/K137A, and
R126A/R139A), triple (K136A/K137A/R139A), N-cluster
(R76A/R77S/R91A/K92A), or C-cluster (K393A/K394S/
R407A/R408S/K409A) mutants were constructed by subjecting
the appropriate mutant template to a subsequent round or
rounds of QuikChange mutagenesis, according to the manufacturer’s directions. The resulting vector DNA incorporating
the desired mutations was then transformed into XL-1 Blue
supercompetent cells. The selected colonies were then chosen
for DNA isolation and purification. The sequence of all mutant
constructs derived from EGFP-HRPT2 or AU5-HRPT2 was
confirmed by DNA sequencing.
Acknowledgments
Apoptotic Analysis
For assessing apoptotic alterations in cellular morphology,
NIH/3T3 cells grown on coverslips were transiently transfected
with either pEGFP or pEGFP-HRPT2 (WT). Forty-eight hours
after transfection, cells were fixed with 3.7% formaldehyde, and
the DNA was visualized by incubating with 1 mmol/L Hoechst
33342 dye (Invitrogen). Cells expressing GFP were analyzed
using a confocal laser scan microscope as described above. For
more quantitative analysis of DNA fragmentation induced by
overexpression of the HRPT2 gene product, NIH/3T3 cells were
acutely transfected with pcDNA3, AU5-HRPT2 (WT), or AU5HRPT2 mutant by electroporation (Amaxa Nucleofector Kit R).
At least 6,000 cells were seeded per well of 96-well plate in
triplicate. Forty-eight hours after transfection, the release of
mononucleosomes and oligonucleosomes into the cytoplasm
was analyzed by a Cell Death Detection ELISA (Roche Applied
Science) kit according to the manufacturer’s instructions. For the
measurement of caspase-3 activity, NIH cells were transfected as
for detecting DNA fragments. Forty-eight hours after transfection, the activity of caspase-3 was assayed in triplicate using a
Colorimetric CaspACE Assay System (Promega, Madison, WI)
using a synthetic tetrapeptide, Asp-Glu-Val-Asp (DEVD),
coupled to the colorimetric molecule p-nitroanilide as substrate.
RNA Interference
For endogenous parafibromin gene silencing by RNA
interference, HeLa cells were cultured as described above for
We thank Geoffrey Woodard, Sunita Agarwal, and Stephen Marx for
encouragement and helpful discussion; Dr. Owen M. Schwartz for sharing his
expertise in confocal microscopy; and Dr. Akira Nakagawara for providing us
with the FLAG-p53 construct.
References
1. Carpten JD, Robbins CM, Villablanca A, et al. HRPT2, encoding
parafibromin, is mutated in hyperparathyroidism-jaw tumor syndrome. Nat Genet
2002;32:676 – 80.
2. Jackson CE, Norum RA, Boyd SB, et al. Hereditary hyperparathyroidism and
multiple ossifying jaw fibromas: a clinically and genetically distinct syndrome.
Surgery 1990;108:1006 – 12.
3. Mallette LE, Malini S, Rappaport MP, Kirkland JL. Familial cystic parathyroid
adenomatosis. Ann Intern Med 1987;107:54 – 60.
4. Teh BT, Farnebo F, Kristoffersson U, et al. Autosomal dominant primary
hyperparathyroidism and jaw tumor syndrome associated with renal hamartomas
and cystic kidney disease: linkage to 1q21 – 32 and loss of the wild type allele in
renal hamartomas. J Clin Endocrinol Metab 1996;81:4204 – 11.
5. Teh BT, Farnebo F, Twigg S, et al. Familial isolated hyperparathyroidism maps
to the hyperparathyroidism-jaw tumor locus in 1q21 – 32 in a subset of families.
J Clin Endocrinol Metab 1998;83:2114 – 20.
6. Simonds WF, James-Newton LA, Agarwal SK, et al. Familial isolated
hyperparathyroidism: clinical and genetic characteristics of 36 kindreds. Medicine
(Baltimore) 2002;81:1 – 26.
7. Simonds WF, Robbins CM, Agarwal SK, Hendy GN, Carpten JD, Marx SJ.
Familial isolated hyperparathyroidism is rarely caused by germline mutation in
HRPT2, the gene for the hyperparathyroidism-jaw tumor syndrome. J Clin
Endocrinol Metab 2004;89:96 – 102.
8. Howell VM, Haven CJ, Kahnoski K, et al. HRPT2 mutations are associated
with malignancy in sporadic parathyroid tumours. J Med Genet 2003;40:657 – 63.
9. Villablanca A, Calender A, Forsberg L, et al. Germline and de novo mutations
in the HRPT2 tumour suppressor gene in familial isolated hyperparathyroidism
(FIHP). J Med Genet 2004;41:e32.
10. Shattuck TM, Valimaki S, Obara T, et al. Somatic and germ-line mutations of
the HRPT2 gene in sporadic parathyroid carcinoma. N Engl J Med 2003;349:
1722 – 9.
11. Cetani F, Pardi E, Borsari S, et al. Genetic analyses of the HRPT2 gene in
primary hyperparathyroidism: germline and somatic mutations in familial and
sporadic parathyroid tumors. J Clin Endocrinol Metab 2004;89:5583 – 91.
12. Rozenblatt-Rosen O, Hughes CM, Nannepaga SJ, et al. The parafibromin
tumor suppressor protein is part of a human Paf1 complex. Mol Cell Biol 2005;
25:612 – 20.
13. Yart A, Gstaiger M, Wirbelauer C, et al. The HRPT2 tumor suppressor gene
product parafibromin associates with human PAF1 and RNA polymerase II. Mol
Cell Biol 2005;25:5052 – 60.
14. Zhu B, Mandal SS, Pham AD, et al. The human PAF complex coordinates transcription with events downstream of RNA synthesis. Genes Dev 2005;19:1668 – 73.
15. Mosimann C, Hausmann G, Basler K. Parafibromin/Hyrax activates Wnt/Wg
target gene transcription by direct association with beta-catenin/Armadillo. Cell
2006;125:327 – 41.
16. Woodard GE, Lin L, Zhang JH, Agarwal SK, Marx SJ, Simonds WF.
Parafibromin, product of the hyperparathyroidism-jaw tumor syndrome gene
HRPT2, regulates cyclin D1/PRAD1 expression. Oncogene 2005;24:1272 – 6.
17. Pimenta FJ, Gontijo Silveira LF, et al. HRPT2 gene alterations in ossifying
fibroma of the jaws. Oral Oncol 2006;42:735 – 9.
Mol Cancer Res 2007;5(2). February 2007
Downloaded from mcr.aacrjournals.org on August 11, 2017. © 2007 American Association for Cancer
Research.
Parafibromin Nuclear Localization and Apoptosis
18. Bradley KJ, Hobbs MR, Buley ID, et al. Uterine tumours are a phenotypic
manifestation of the hyperparathyroidism-jaw tumour syndrome. J Intern Med
2005;257:18 – 26.
23. Robbins J, Dilworth SM, Laskey RA, Dingwall C. Two interdependent basic
domains in nucleoplasmin nuclear targeting sequence: identification of a class of
bipartite nuclear targeting sequence. Cell 1991;64:615 – 23.
19. Cronshaw JM, Krutchinsky AN, Zhang W, Chait BT, Matunis MJ.
Proteomic analysis of the mammalian nuclear pore complex. J Cell Biol 2002;
158:915 – 27.
24. Dingwall C, Laskey RA. Nuclear targeting sequences: a consensus? Trends
Biochem Sci 1991;16:478 – 81.
20. Wade PA, Werel W, Fentzke RC, et al. A novel collection of accessory factors
associated with yeast RNA polymerase II. Protein Expr Purif 1996;8:85 – 90.
21. Shi X, Chang M, Wolf AJ, et al. Cdc73p and Paf1p are found in a novel RNA
polymerase II-containing complex distinct from the Srbp-containing holoenzyme.
Mol Cell Biol 1997;17:1160 – 9.
22. Porter SE, Penheiter KL, Jaehning JA. Separation of the Saccharomyces
cerevisiae Paf1 complex from RNA polymerase II results in changes in its
subnuclear localization. Eukaryot Cell 2005;4:209 – 20.
25. Lowe SW, Cepero E, Evan G. Intrinsic tumour suppression. Nature 2004;
432:307 – 15.
26. Hahn MA, Marsh DJ. Identification of a functional bipartite nuclear
localization signal in the tumor suppressor parafibromin. Oncogene 2005;24:
6241 – 8.
27. Payne SR, Kemp CJ. Tumor suppressor genetics. Carcinogenesis 2005;26:
2031 – 45.
28. Parusel CT, Kritikou EA, Hengartner MO, Krek W, Gotta M. URI-1 is
required for DNA stability in C. elegans . Development 2006;133:621 – 9.
Mol Cancer Res 2007;5(2). February 2007
Downloaded from mcr.aacrjournals.org on August 11, 2017. © 2007 American Association for Cancer
Research.
193
Nuclear Localization of the Parafibromin Tumor Suppressor
Protein Implicated in the Hyperparathyroidism-Jaw Tumor
Syndrome Enhances Its Proapoptotic Function
Ling Lin, Meggan Czapiga, Lylia Nini, et al.
Mol Cancer Res 2007;5:183-193.
Updated version
Cited articles
Citing articles
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://mcr.aacrjournals.org/content/5/2/183
This article cites 28 articles, 9 of which you can access for free at:
http://mcr.aacrjournals.org/content/5/2/183.full.html#ref-list-1
This article has been cited by 3 HighWire-hosted articles. Access the articles at:
/content/5/2/183.full.html#related-urls
Sign up to receive free email-alerts related to this article or journal.
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Department at [email protected].
To request permission to re-use all or part of this article, contact the AACR Publications
Department at [email protected].
Downloaded from mcr.aacrjournals.org on August 11, 2017. © 2007 American Association for Cancer
Research.