Blocking protein farnesyltransferase improves nuclear
blebbing in mouse fibroblasts with a targeted
Hutchinson–Gilford progeria syndrome mutation
Shao H. Yang*, Martin O. Bergo†, Julia I. Toth*, Xin Qiao*, Yan Hu*, Salemiz Sandoval*, Margarita Meta‡,
Pravin Bendale§, Michael H. Gelb§, Stephen G. Young*, and Loren G. Fong*¶
*Division of Cardiology, Department of Medicine, David Geffen School of Medicine, University of California, Los Angeles, CA 90095;†Department of
Internal Medicine, Bruna Stråket 16, Third Floor, Sahlgrenska University Hospital, SE-413 45 Go ¨teborg, Sweden;‡Musculoskeletal and Quantitative Research
Group, Department of Radiology, University of California, San Francisco, CA 94107; and§Departments of Chemistry and Biochemistry, University of
Washington, Seattle, WA 98195
Communicated by Daniel Steinberg, University of California at San Diego, La Jolla, CA, June 6, 2005 (received for review April 20, 2005)
Hutchinson–Gilford progeria syndrome (HGPS), a progeroid syn-
drome in children, is caused by mutations in LMNA (the gene for
prelamin A and lamin C) that result in the deletion of 50 aa within
prelamin A. In normal cells, prelamin A is a ‘‘CAAX protein’’ that is
farnesylated and then processed further to generate mature lamin
A, which is a structural protein of the nuclear lamina. The mutant
prelamin A in HGPS, which is commonly called progerin, retains the
CAAX motif that triggers farnesylation, but the 50-aa deletion
prevents the subsequent processing to mature lamin A. The pres-
ence of progerin adversely affects the integrity of the nuclear
lamina, resulting in misshapen nuclei and nuclear blebs. We hy-
pothesized that interfering with protein farnesylation would block
the targeting of progerin to the nuclear envelope, and we further
hypothesized that the mislocalization of progerin away from the
nuclear envelope would improve the nuclear blebbing phenotype.
To approach this hypothesis, we created a gene-targeted mouse
model of HGPS, generated genetically identical primary mouse
embryonic fibroblasts, and we then examined the effect of a
farnesyltransferase inhibitor on nuclear blebbing. The farnesyl-
transferase inhibitor mislocalized progerin away from the nuclear
envelope to the nucleoplasm, as determined by immunofluores-
ence microscopy, and resulted in a striking improvement in nuclear
blebbing (P < 0.0001 by ?2statistic). These studies suggest a
possible treatment strategy for HGPS.
aging ? lamin A?C ? laminopathy
notypes, including a wizened appearance of the skin, osteopo-
rosis, alopecia, and premature atherosclerosis (1). Children with
HGPS die at the mean age of 13, generally from myocardial
infarctions or strokes (1). This disease is caused by the accumu-
lation of a mutant form of prelamin A that cannot be processed
to mature lamin A (1). In normal cells, wild-type prelamin A is
virtually undetectable because it is fully converted to mature
lamin A, a structural protein of the nuclear lamina (2, 3). The
nuclear lamina is an intermediate filament meshwork adjacent to
the inner nuclear membrane that provides structural support for
the nucleus (2, 3).
Prelamin A contains a nuclear localization signal and ter-
minates with a CAAX motif (2), in which C is a cysteine, A
residues are usually aliphatic amino acids, and X can be one of
many different residues. CAAX motifs are also found on lamin
B1, lamin B2, the Ras family of proteins, and many other
cellular proteins. The CAAX motif triggers three sequential
enzymatic posttranslational modifications, beginning with pro-
tein prenylation. In the case of prelamin A, the first processing
step is carried out by protein farnesyltransferase (FTase) and
involves the addition of a 15-carbon farnesyl lipid to the thiol
group of the cysteine within the CAAX motif. Second, the last
utchinson–Gilford progeria syndrome (HGPS) is a prog-
eroid syndrome characterized by a host of aging-like phe-
3 aa of the protein (i.e., ?AAX) are removed by a prenylpro-
tein-specific endoprotease. For prelamin A, this step is likely
to be a redundant function of two endoplasmic reticulum
membrane endoproteases, Zmpste24 and Rce1 (4). Third, the
newly exposed farnesylcysteine is methylated by Icmt, a pre-
nylprotein-specific membrane methyltransferase of the endo-
plasmic reticulum (5). After these CAAX-box modifications
have been completed, prelamin A (in contrast to other CAAX
proteins) undergoes an additional processing step. The last 15
aa of the protein (including the farnesylcysteine methyl ester)
are clipped off by Zmpste24 and then degraded, leaving behind
mature lamin A (4, 6, 7).
The farnesylation of prelamin A is important for its targeting
to the nuclear envelope (8–10). Each of the three CAAX motif
more hydrophobic, facilitating its association with the inner
nuclear membrane, where the protein is cleaved, releasing
mature lamin A (9, 11). In the absence of farnesylation (for
example, in mevinolin-treated cells), prelamin A accumulates in
the nucleoplasm and does not reach the nuclear envelope (9, 11).
In the setting of Zmpste24 deficiency, farnesyl prelamin A
accumulates at the nuclear envelope (6, 12) and adversely affects
the integrity of the nuclear envelope. The nuclei of Zmpste24-
deficient fibroblasts are misshapen, containing numerous nu-
clear blebs (6, 12).
HGPS is most commonly caused by a de novo point mutation
in exon 11 of LMNA (1). This mutation, which occurs in codon
608, activates a cryptic splice site and leads to the in-frame
deletion of 50 aa within prelamin A. This deletion leaves the
CAAX motif intact; hence, the mutant prelamin A (progerin) is
predicted to undergo farnesylation, release of the ?AAX, and
carboxyl methylation. However, the site for the second endo-
proteolytic cleavage step is eliminated by the deletion (1). Thus,
progerin cannot be processed to lamin A and likely retains a
farnesylcysteine methyl ester at its C terminus. Like Zmpste24-
deficient cells, HGPS fibroblasts contain misshapen nuclei with
numerous blebs of the nuclear envelope (1). In human HGPS
cells, the severity of nuclear blebbing is variable, depending in
part on the number of times the cells have been passaged (13).
A GFP-tagged progerin has been reported to accumulate along
the nuclear envelope of HeLa cells and cause nuclear shape
We hypothesized that the farnesylation of progerin targets
the protein to the nuclear envelope, where it might weaken the
ferase; FTI, FTase inhibitor; MEF, mouse embryonic fibroblast.
¶To whom correspondence should be addressed at: University of California, 675 Charles
East Young Drive South, MacDonald Medical Research Laboratory Building, Room 4770,
Los Angeles, CA 90095. E-mail: email@example.com.
© 2005 by The National Academy of Sciences of the USA
July 19, 2005 ?
vol. 102 ?
no. 29 ?
nuclear lamina and cause nuclear blebbing. We further hy-
pothesized that blocking farnesylation with an FTase inhibitor
(FTI) would mislocalize progerin away from the nuclear
envelope and reduce nuclear blebbing. Some researchers
might argue that the latter hypothesis is unattractive, given that
the FTI would also block the posttranslational processing of
lamin B1 and lamin B2, potentially weakening the lamina
further. However, we hypothesized that the salutary effects of
blocking farnesylation of progerin would ‘‘trump’’ any dele-
terious effects of the FTI on the lamina and lead to an overall
improvement in nuclear blebbing.
To test the impact of blocking the farnesylation of progerin
on nuclear shape, we reasoned that it would be helpful to
create gene-targeted ‘‘HGPS mice’’ that express progerin at
levels sufficient to cause nuclear blebbing. The existence of a
gene-targeted mouse model would make it possible to study
the impact of an FTI on nuclear shape in independent lines of
low-passage, genetically identical mouse embryonic fibroblasts
(MEFs). In this study, we adopted exactly this strategy, first by
generating a mouse model of HGPS and then by using primary
MEFs to define the impact of an FTI on the nuclear blebbing
Materials and Methods
Gene-Targeted Mouse Model of HGPS. The DNA fragments for the
arms of the gene-targeting vector were generated by long-range
PCR of genomic DNA from 129?Ola ES cells. A 6-kb 5?
fragment, which spanned from the end of intron 5 to sequences
1.9 kb downstream from the 3? UTR (encoded by exon 12) was
amplified with 5?-GGCTTCCTGGTCACTGGATA-3? and 5?-
GATCTGCCTGGAAGCTGAGT-3? and cloned into pCR2.1-
TOPO-XL (Invitrogen). Next, a 5-kb EcoRI fragment (spanning
from a polylinker EcoRI site in pCR2.1-TOPO-XL to an EcoRI
site 0.9 kb downstream from the 3? UTR) was cleaved from the
vector and cloned into pBSK (Stratagene). To create the 5? arm
of a sequence-replacement vector, this EcoRI fragment was
subjected to two sequential site-directed mutagenesis reactions
(QuikChange, Stratagene). First, intron 10 was deleted with the
and the reversed and complemented primer. Second, the last 150
nt of exon 11 and intron 11 were deleted with the primer
GCTCCCAGAACTGCAGCATCATGTAATCT-3? and the re-
versed and complemented primer.
To create the gene-targeting vector, the mutant EcoRI frag-
ment was cloned into the polylinker EcoRI site of pKSloxPNT-
mod. The 3? arm, consisting of sequences immediately down-
stream of those in the 5? arm, was amplified with the primers
5?-GACAGCCACCTGGTCAGTTT-3? and 5?-GTAACTCTG-
GCTGCCCTCAA-3? and then cloned into pCR2.1-TOPO-XL.
To complete the gene-targeting vector, this fragment was cloned
into the polylinker AscI site of pKSloxPNTmod. The integrity of
the vector (?17 kb in length) was verified by DNA sequencing
and restriction mapping.
The vector was linearized with NotI and electroporated into
strain 129?Ola ES cells. To identify clones carrying the targeted
LmnaHGallele, we performed Southern blot analyses of EcoRI-
digested genomic DNA with a 348-bp 5?-flanking probe. The
probe was generated by PCR from mouse genomic DNA with
the following primers: 5?-CAAGGAGCTCGGATTCTGTC-3?
and 5?-GTCAGGGAAGAGTGCAGAGG-3?. The probe de-
tected a 10.4-kb band in the wild-type Lmna allele and a 9.3-kb
band in the LmnaHGallele. Genotyping was also performed on
genomic DNA by PCR with the following primers: 5?-
TGAGTACAACCTGCGCTCAC-3? and 5?-CAGACAGGAG-
GTGGCATGT-3?. The PCR fragment, spanning from exon 11
to exon 12, was 582 bp in the wild-type Lmna allele and 186 bp
in the LmnaHGallele.
Creating LmnaHGMEFs. Targeted 129?Ola ES cells were micro-
injected into C57BL?6 blastocysts to produce chimeric mice.
To generate MEFs, chimeras were bred with C57BL?6 fe-
males. Wild-type primary MEFs (Lmna?/?) and LmnaHG/?
MEFs were prepared from the same litter of day 13.5 embryos
and genotyped by PCR. Because the MEFs were generated
from embryos of chimera matings, they were genetically
identical (one 129?Ola chromosome and one C57BL?6 chro-
mosome). Cells homozygous for the LmnaHGmutation were
generated by subjecting LmnaHG/?cells to several months of
selection in increasing concentrations of G418 (14). The
LmnaHG/HGcells were euploid, but many of the cells were
differentiated, as determined by morphology.
FTIs. For all experiments, we used PB-43, which is a member of
the tetrahydroquinoline family of protein FTIs (15) that displays
potent inhibition of rat FTase in vitro (IC50? 1.7 nM). PB-43
readily crosses cell membranes, as demonstrated by its ability to
kill Plasmodium falciparum in human red blood cells (M.H.G.,
unpublished data). PB-43 was synthesized by described methods
(15) and shown to be pure by HPLC on a reverse-phase column.
The compound was dissolved in DMSO at a concentration of 10
mM and stored in aliquots at ?80°C.
Treatment of Cells with the FTI and Western Blot Analyses. Adherent
early-passage MEFs in six-well tissue culture plates were
incubated with the vehicle control (DMSO) or the indicated
concentrations of PB-43 diluted in culture medium at 37°C for
48 h. The cells were washed with PBS, and urea-soluble
extracts were prepared as described in ref. 11. Cell pellets
solubilized with SDS-containing buffers were also prepared
and yielded results indistinguishable from those with urea
extraction. Proteins were size-separated on 4–12% gradient
polyacrylamide Bis-Tris gels (Invitrogen) and then electro-
phoretically transferred to nitrocellulose membranes for West-
ern blotting. The following antibody dilutions were used: 1:400
anti-lamin A?C goat IgG (sc-6215, Santa Cruz Biotechnology),
1:400 anti-lamin B (sc-6217, Santa Cruz Biotechnology),
1:6,000 anti-mouse prelamin A rabbit antiserum (12, 13), 1:500
anti-Hdj-2 mouse IgG (LabVision, Fremont, CA), 1:1,000
anti-actin goat IgG (sc-1616, Santa Cruz Biotechnology),
1:6,000 horseradish peroxidase (HRP)-labeled anti-goat IgG
(sc-2020, Santa Cruz Biotechnology), 1:4,000 HRP-labeled
anti-mouse IgG (Amersham Biosciences), and 1:6,000
HRP-labeled anti-rabbit IgG (Amersham Biosciences). Anti-
body binding was detected with the ECL Plus chemilumines-
cence system (Amersham Biosciences) and exposure to x-ray
Immunofluoresence Microscopy. Primary cells of different geno-
types were grown on coverslips, fixed in 3% paraformaldehyde,
permeabilized with 0.2% Triton X-100, and blocked with BSA
(12). Cells were incubated for 60 min with antibodies against
lamin A (sc-20680) or lamin B (sc-6217) (Santa Cruz Biotech-
nology). After washing, cells were stained with species-specific
Cy3-conjugated secondary antibodies (Jackson Immuno-
Research) and DAPI to visualize DNA. Images were obtained
on an Axiovert 40CFL microscope (Zeiss) with a ?63?1.25
oil-immersion objective and processed with AXOVISION software
(version 4.2; Zeiss). Nuclear shape abnormalities were scored by
two independent observers, who were blinded to genotype or
www.pnas.org?cgi?doi?10.1073?pnas.0504641102Yang et al.
We created a mouse model of HGPS that expressed large
amounts of progerin. To accomplish that goal, we created a
mutant Lmna allele, LmnaHG, that exclusively yields progerin
(Fig. 1A). We deleted intron 10 of Lmna, thereby eliminating
lamin C synthesis; we also deleted the last 150 nt of exon 11 and
intron 11, which results in the synthesis of progerin (and
precludes wild-type prelamin A synthesis). Two targeted ES
cell clones (from 192 G418- and FIAU-resistant clones) were
identified by Southern blotting (Fig. 1B) and used to create 22
high-percentage chimeric mice. Those mice were bred with
C57BL?6 mice to produce the LmnaHG/?mice for this study
(Fig. 1C). In addition, LmnaHG/HGcell lines were created from
LmnaHG/?ES cells by high-G418 selection (Fig. 1D). As
predicted, primary MEFs from LmnaHG/?embryos produced
progerin, along with lamin A and lamin C from the wild-type
Lmna allele (Figs. 1D and 3), whereas the LmnaHG/HGcells
yielded only progerin (Fig. 1D). Lmna expression in wild-type
ES cells is low (2, 3, 16). However, the LmnaHG/HGcells, which
had undergone ?2 months of high G418 selection and ap-
peared to be differentiated, expressed progerin at levels
comparable with those in MEFs (Fig. 1D). Not surprisingly,
the nuclei of many LmnaHG/?MEFs were misshapen and
contained large blebs (Fig. 2).
We sought to define the impact of an FTI, PB-43, on nuclear
shape. To document that this compound was effective in
blocking farnesylation, we treated LmnaHG/?MEFs with the
PB-43 FTI and then examined the electrophoretic migration of
Hdj-2, a farnesylated protein, with Western blot analyses of
SDS?PAGE gels (Fig. 3). In FTI-treated MEFs, nearly all of
the Hdj-2 migrated slowly and almost none migrated normally,
indicating that the FTI had been effective in blocking protein
We also performed Western blot analyses of wild-type and
LmnaHG/?cell extracts with an antibody specific for the C
terminus of prelamin A (Fig. 3). The prelamin A-specific
antibody does not normally detect prelamin A in cells because
prelamin A is farnesylated and rapidly converted to mature
lamin A. However, in the presence of the FTI, prelamin A
processing is blocked, and prelamin A is easily detectable in
both wild-type and LmnaHG/?MEFs (Fig. 3). Western blot
analyses with an antibody against the N-terminal portion of
lamin A detected mature lamin A in wild-type MEFs and in
LmnaHG/?MEFs but detected the slightly larger prelamin A
(and virtually no mature lamin A) in MEFs that had been
treated with the FTI. The FTI did not yield clear-cut or
consistent alterations in the amount of either progerin or lamin
B1 in LmnaHG/?MEFs (Fig. 3).
To judge the impact of the FTI on nuclear blebbing in
LmnaHG/?MEFs, we examined both Lmna?/?and LmnaHG/?
MEFs in a blinded fashion by immunofluorescence microscopy.
LmnaHG/?MEFs had more nuclear blebs than Lmna?/?cells
(P ? 0.0001, ?2test) (Fig. 4). The FTI did not affect nuclear
shape in Lmna?/?MEFs. However, FTI treatment of the
LmnaHG/?MEFs reduced the frequency of nuclear blebbing
(P ? 0.0001). This result was consistent in several independently
isolated LmnaHG/?cell lines and in several independent exper-
LmnaHG/?MEFs was not different, as determined by the ?2
allele in mouse ES cells, with EcoRI-cleaved genomic DNA and the indicated 5?
flanking probe. (C) PCR identification of the LmnaHGallele. Results with
wild-type MEFs (WT), heterozygous MEFs (LmnaHG/?), homozygous ES cells
(LmnaHG/HG), and heterozygous MEFs (LmnaHG/?) are shown. (D) Western
blotting identification of progerin with a lamin A?C-specific monoclonal
antibody. Wild-type cells, LmnaHG/?MEFs, and LmnaHG/HGcells are shown. On
SDS?PAGE gels, the electrophoretic migration of progerin in mouse LmnaHG/?
MEFs and human HGPS fibroblasts was identical (data not shown).
blasts. (A–E) Immunostaining showing nuclear blebs in LmnaHG/?MEFs. Blebs
stained for lamin B1 (red).
Immunofluorescence microscopy of wild-type and LmnaHG/?fibro-
Cells were grown in the presence and absence of an FTI (10 ?M PB-43), and
for prelamin A (specific for the extreme C terminus), lamin A?C (binds to both
lamin A and lamin C), lamin B1, Hdj-2, and actin.
Yang et al.PNAS ?
July 19, 2005 ?
vol. 102 ?
no. 29 ?
statistic, from the frequency of blebbing in FTI-treated or
Because LmnaHG/?MEFs synthesize lamin A and lamin C
in addition to progerin, it was impossible to use immuno-
fluorescence microscopy of LmnaHG/?MEFs to define the
location of progerin within those cells. However, the intracel-
lular location of progerin could be examined in experiments
with LmnaHG/HGcells, which synthesize exclusively progerin.
In those cells, progerin was located mainly at the nuclear
envelope (Fig. 5 A and B). However, in the presence of the FTI,
virtually all of the progerin was mislocalized to intensely
staining aggregates within the nucleoplasm, and none of it was
detectable at the nuclear envelope (Fig. 5 C and D).
FTIs were developed initially as anticancer agents (17). The
concept was straightforward: to eliminate the farnesyl lipid
from mutationally activated Ras proteins, thereby mislocalizing
these signaling proteins away from the plasma membrane, where
they ‘‘cause trouble’’ by stimulating uncontrolled cell division. In
this study, we assessed an analogous concept with HGPS, which
is a disease in which another farnesylated protein causes trouble
in the cell. The synthesis of progerin leads to misshapen nuclei
and frequent nuclear blebs (1). We hypothesized that blocking
farnesylation would mislocalize progerin away from the nuclear
envelope and ameliorate the nuclear blebbing phenotype. We
generated a gene-targeted mouse model of HGPS, created
genetically identical primary LmnaHG/?MEFs as well as
LmnaHG/HGcells, and tested the impact of an FTI on both
nuclear blebbing and progerin localization. In untreated cells,
progerin was located mainly along the nuclear envelope,
whereas in FTI-treated cells the protein was mislocalized to
the nucleoplasm. In the LmnaHG/?MEFs, the FTI reduced
nuclear blebbing to a baseline level observed in untreated
wild-type cells. These studies visualize progerin localization
independently of lamin A and lamin C, and they show that
mislocalization of progerin is associated with a change in
nuclear shape. Interestingly, the FTI did not adversely affect
nuclear shape in wild-type fibroblasts.
Nuclear blebbing is the principal cellular phenotype in
HGPS (13, 18), and the FTI clearly ameliorates this phenotype.
The next obvious questions are as follows. Will the nuclear
shape abnormalities in LmnaHG/?cells be accompanied by any
disease phenotypes at the ‘‘whole-animal’’ level? And, if so,
would an FTI also ameliorate those disease phenotypes,
including perhaps the atherosclerotic disease that claims the
lives of most humans with the disease? With regard to the first
question, the LmnaHG/?mice clearly exhibit unequivocal
disease phenotypes (S.H.Y., L.G.F., and S.G.Y., unpublished
data). By 4 months of age, all of the LmnaHG/?mice (n ? 22)
exhibit growth retardation and?or bone disease, akin to ho-
mozygous Zmpste24-deficient mice (4). Detailed radiographic
and pathologic analyses (e.g., assessing bone density and
defining atherosclerosis susceptibility in inbred backgrounds)
will take many months to complete. However, the existence of
tractable phenotypes in our heterozygous gene-targeted mice
makes us confident that this model could be very useful for
assessing therapeutic strategies.
To properly assess the impact of FTIs on disease phenotypes
in mice, we first need to identify methods for the long-term
FTI (10 ?M PB-43). A and B represent two independent experiments. Each black circle shows the frequency of nuclear blebbing with an independently isolated
fibroblast cell line. Bars indicate the mean frequency of blebbing. The number of cells with nuclear blebs and the total number of cells examined are recorded
within each bar. In both experiments, the FTI did not change the frequency of blebbing in Lmna?/?MEFs, as determined by the ?2test. LmnaHG/?MEFs exhibited
in FTI-treated LmnaHG/?MEFs was not different from the frequency of blebbing in the treated or untreated Lmna?/?MEFs. Very similar results, with identical
levels of statistical significance, were obtained when the microscopic slides were scored by a second blinded observer.
Bar graph showing increased frequency of nuclear blebbing in LmnaHG/?MEFs and a reduction in nuclear blebbing in LmnaHG/?MEFs treated with an
untreated and FTI-treated LmnaHG/HGcells. DNA was visualized with DAPI
(blue), and progerin was visualized with an antibody against lamin A (red). (A
and B) Untreated LmnaHG/HGcells, showing progerin along the nuclear enve-
lope. Misshapen nuclei were common (arrow). (C and D) FTI-treated
LmnaHG/HGcells, revealing intensely staining progerin aggregates (arrow-
heads) in the nucleoplasm.
Immunofluoresence images showing the distribution of progerin in
www.pnas.org?cgi?doi?10.1073?pnas.0504641102 Yang et al.
oral delivery of the drugs and document that the drug is
efficacious in blocking protein farnesylation in multiple tis-
sues. Again, we anticipate that optimization of FTI delivery
schemes will require considerable experimentation. In the
meantime, we can only speculate about whether FTIs might be
useful for treating the disease phenotypes in the LmnaHG/?
mice. Optimists would contend that it makes intuitive sense
(i.e., in line with Occam’s razor) that improvements in ‘‘whole-
animal’’ disease phenotypes would parallel improvements in
cellular phenotypes. Optimists would also point out that the
nuclear blebbing phenotypes in cultured fibroblasts and the
whole-animal disease phenotypes were clearly associated in a
recent study of Zmpste24-deficient mice (12). Zmpste24-
deficient fibroblasts, which accumulate wild-type farnesyl pre-
lamin A, manifest nuclear blebbing (6, 12), and Zmpste24-
deficient mice develop a variety of progeria-like phenotypes (4,
12). When prelamin A synthesis was reduced by 50% (by
introducing a single copy of a Lmna knockout allele), the
nuclear blebbing in fibroblasts and the disease phenotypes in
mice were eliminated (12). That study was also intriguing
because it proved that dramatic improvements in progeria-like
disease phenotypes can occur with a 50% reduction in the
amount of farnesyl prelamin A in the cell. Thus, one might
easily imagine that an FTI could ameliorate the disease
phenotypes even with incomplete inhibition of farnesylation
and incomplete mislocalization of progerin (i.e., without push-
ing FTI doses to levels associated with side effects). However,
drug toxicity may not be much of an issue because FTIs have
been generally well tolerated in humans and in mice (19, 20).
Genetic studies also support the idea that it should be possible
to give FTIs safely on a long-term basis (21). Completely
inactivating FTase in adult mice by using Cre?loxP approaches
did not cause histopathological abnormalities or any notice-
able disease phenotypes (21).
There are also reasons to be pessimistic about the proposition
that an FTI would be a panacea for HGPS. Although an FTI
mislocalizes progerin and improves nuclear blebbing, one could
imagine that nonfarnesylated progerin could still be toxic in vivo.
Nonfarnesylated progerin is still a structurally abnormal protein,
and it is sobering to remember that even single amino acid
substitutions in mature lamin A and lamin C (neither of which
is farnesylated) can cause a host of different genetic diseases
(e.g., several forms of muscular dystrophy, cardiomyopathy,
partial lipodystrophy, and mandibuloacral dysplasia) (2, 3). Note
genetic diseases (22, 23). Moreover, one could argue that
improved nuclear blebbing in cultured cells might not be an
accurate indicator of disease phenotypes at the whole-animal
level, for the simple reason that some LMNA missense mutations
have been reported to cause human disease without affecting
certainly point out that HGPS is a dominant disease, caused by
a single mutant chromosome. Treatment with an FTI interferes
with the biogenesis of mature lamin A from the normal LMNA
allele. To our knowledge, the short or long-term consequences
of reducing lamin A biogenesis in humans are not known,
although a single LMNA null allele, causing a 50% reduction
in both lamin A and lamin C synthesis, causes muscular
amounts of progerin, and our strategy was successful. Although it
mutation identified in humans with HGPS (24), we did not follow
this approach because we worried that that the ‘‘point mutation’’
approach would not yield sufficient amounts of progerin to elicit
phenotypes in the mouse. In humans with HGPS, most of the
transcripts from the mutant LMNA allele actually yield wild-type
about this type of an allele for mouse experimentation because
experience with mouse Lmna mutations has suggested that mice
lamin proteins required to elicit disease phenotypes. For example,
dystrophy, whereas two copies of the H222P allele are required to
elicit a muscle phenotype in mice (25). Similarly, humans heterozy-
gous for a LMNA nonsense mutation develop muscular dystrophy
(24), whereas mice with a single Lmna knockout allele are normal
(26). Thus, we worried that the point mutation approach in the
mouse might not yield sufficient levels of progerin to elicit pheno-
types. Accordingly, we chose a gene-targeting strategy that would
guarantee high levels of progerin expression. As it turned out, we
a single copy of the mutant allele.
Progerin is clearly the ‘‘culprit molecule’’ in HGPS (1, 13, 18),
and studies raise the hope that FTIs might ultimately prove to be
useful for treating HGPS. The FTI strategy appears to be well
suited for testing in children because the drugs have been studied
extensively, are generally well tolerated, and can be given orally.
However, FTIs are not the only potential hope for HGPS.
Recent studies have indicated that the nuclear blebbing pheno-
type in HGPS fibroblasts can be ameliorated with morpholino
antisense reagents (18) or by expressing short hairpin RNA
constructs (RNA interference) (Junko Oshima, personal com-
the production of progerin transcripts, so the finding of reduced
nuclear blebbing is probably not particularly surprising. Never-
theless, those studies were very important because they provided
hope to patients affected by HGPS and because they will
stimulate interest in overcoming the practical and pharmaco-
logical obstacles to delivering RNAi and morpholino oligonu-
cleotides to humans.
We thank Dr. Luanne Peters for counting chromosomes in the
LmnaHG/HGES cells. This work was supported in part by National
Institutes of Health Grants AI054384 (to M.H.G.), R01 CA099506, and
R01 AR050200 and a grant from the Progeria Research Foundation (to
1. Eriksson, M., Brown, W. T., Gordon, L. B., Glynn, M. W., Singer, J., Scott, L.,
Erdos, M. R., Robbins, C. M., Moses, T. Y., Berglund, P., et al. (2003) Nature
2. Mounkes, L. C., Burke, B. & Stewart, C. L. (2001) Trends Cardiovasc. Med. 11,
3. Burke, B. & Stewart, C. L. (2002) Nat. Rev. Mol. Cell Biol. 3, 575–585.
4. Bergo, M. O., Gavino, B., Ross, J., Schmidt, W. K., Hong, C., Kendall, L. V.,
Mohr, A., Meta, M., Genant, H., Jiang, Y., et al. (2002) Proc. Natl. Acad. Sci.
USA 99, 13049–13054.
5. Dai, Q., Choy, E., Chiu, V., Romano, J., Slivka, S. R., Steitz, S. A., Michaelis,
S. & Philips, M. R. (1998) J. Biol. Chem. 273, 15030–15034.
6. Penda ´s, A. M., Zhou, Z., Cadin ˜anos, J., Freije, J. M. P., Wang, J., Hultenby,
K., Astudillo, A., Wernerson, A., Rodrı ´guez, F., Tryggvason, K. & Lope ´z-Otı ´n,
C. (2002) Nat. Genet. 31, 94–99.
7. Corrigan, D. P., Kuszczak, D., Rusinol, A. E., Thewke, D. P., Hrycyna, C. A.,
Michaelis, S. & Sinensky, M. S. (2005) Biochem. J. 387, 129–138.
8. Hennekes, H. & Nigg, E. A. (1994) J. Cell Sci. 107, 1019–1029.
9. Lutz, R. J., Trujillo, M. A., Denham, K. S., Wenger, L. & Sinensky, M. (1992)
Proc. Natl. Acad. Sci. USA 89, 3000–3004.
10. Izumi, M., Vaughan, O. A., Hutchison, C. J. & Gilbert, D. M. (2000) Mol. Biol.
Cell 11, 4323–4337.
11. Dalton, M. & Sinensky, M. (1995) Methods Enzymol. 250, 134–148.
12. Fong, L. G., Ng, J. K., Meta, M., Cote, N., Yang, S. H., Stewart, C. L., Sullivan,
T., Burghardt, A., Majumdar, S., Reue, K., et al. (2004) Proc. Natl. Acad. Sci.
USA 101, 18111–18116.
13. Goldman, R. D., Shumaker, D. K., Erdos, M. R., Eriksson, M., Goldman, A. E.,
Gordon, L. B., Gruenbaum, Y., Khuon, S., Mendez, M., Varga, R. & Collins,
F. S. (2004) Proc. Natl. Acad. Sci. USA 101, 8963–8968.
Yang et al.PNAS ?
July 19, 2005 ?
vol. 102 ?
no. 29 ?
14. Mortensen, R. M., Conner, D. A., Chao, S., Geisterfer-Lowrance, A. A. T. & Download full-text
Seidman, J. G. (1992) Mol. Cell. Biol. 12, 2391–2395.
15. Nallan, L., Bauer, K. D., Bendale, P., Rivas, K., Yokoyama, K., Horne ´y, C. P.,
Chem. 48, 3704–3713.
16. Raharjo, W. H., Enarson, P., Sullivan, T., Stewart, C. L. & Burke, B. (2001)
J. Cell Sci. 114, 4447–4457.
17. Reiss, Y., Goldstein, J. L., Seabra, M. C., Casey, P. J. & Brown, M. S. (1990)
Cell 62, 81–88.
18. Scaffidi, P. & Misteli, T. (2005) Nat. Med. 11, 440–445.
19. Caraglia, M., D’Alessandro, A. M., Marra, M., Giuberti, G., Vitale, G.,
Viscomi, C., Colao, A., Prete, S. D., Tagliaferri, P., Tassone, P., et al. (2004)
Oncogene 23, 6900–6913.
20. Rogers, M. J. (2003) Curr. Pharm. Des. 9, 2643–2658.
21. Mijimolle, N., Velasco, J., Dubus, P., Guerra, C., Weinbaum, C. A., Casey, P. J.,
Campuzano, V. & Barbacid, M. (2005) Cancer Cell 7, 313–324.
22. Muchir, A., Medioni, J., Laluc, M., Massart, C., Arimura, T., van der Kooi,
A. J., Desguerre, I., Mayer, M., Ferrer, X., Briault, S., et al. (2004) Muscle Nerve
23. Csoka, A. B., Cao, H., Sammak, P. J., Constantinescu, D., Schatten, G. P. &
Hegele, R. A. (2004) J. Med. Genet. 41, 304–308.
24. Bonne, G., Di Barletta, M. R., Varnous, S., Be ´cane, H.-M., Hammouda, E.-H.,
Merlini, L., Muntoni, F., Greenberg, C. R., Gary, F., Urtizberea, J.-A., et al.
(1999) Nat. Genet. 21, 285–288.
25. Arimura, T., Helbling-Leclerc, A., Massart, C., Varnous, S., Niel, F., Lacene,
Genet. 14, 155–169.
26. Sullivan, T., Escalante-Alcalde, D., Bhatt, H., Anver, M., Bhat, N.,
Nagashima, K., Stewart, C. L. & Burke, B. (1999) J. Cell Biol. 147,
www.pnas.org?cgi?doi?10.1073?pnas.0504641102Yang et al.