Statins upregulate PCSK9, the gene encoding the proprotein convertase neural apoptosis-regulated convertase-1 implicated in familial hypercholesterolemia.

Abstract

Neural apoptosis-regulated convertase (NARC)-1 is the newest member of the proprotein convertase family implicated in the cleavage of a variety of protein precursors. The NARC-1 gene, PCSK9, has been identified recently as the third locus implicated in autosomal dominant hypercholesterolemia (ADH). The 2 other known genes implicated in ADH encode the low-density lipoprotein receptor and apolipoprotein B. As an approach toward the elucidation of the physiological role(s) of NARC-1, we studied its transcriptional regulation.
Using quantitative RT-PCR, we assessed NARC-1 regulation under conditions known to regulate genes involved in cholesterol metabolism in HepG2 cells and in human primary hepatocytes. We found that NARC-1 expression was strongly induced by statins in a dose-dependent manner and that this induction was efficiently reversed by mevalonate. NARC-1 mRNA level was increased by cholesterol depletion but insensitive to liver X receptor activation. Human, mouse, and rat PCSK9 promoters contain 2 typical conserved motifs for cholesterol regulation: a sterol regulatory element (SRE) and an Sp1 site.
PCSK9 regulation is typical of that of the genes implicated in lipoprotein metabolism. In vivo, PCSK9 is probably a target of SRE-binding protein (SREBP)-2.

Full text

Page 1

Statins Upregulate PCSK9, the Gene Encoding the
Proprotein Convertase Neural Apoptosis-Regulated
Convertase-1 Implicated in Familial Hypercholesterolemia
Genevie `ve Dubuc, Ann Chamberland, Hanny Wassef, Jean Davignon, Nabil G. Seidah,
Lise Bernier, Annik Prat
Objective—Neural apoptosis-regulated convertase (NARC)-1 is the newest member of the proprotein convertase family
implicated in the cleavage of a variety of protein precursors. The NARC-1 gene, PCSK9, has been identified recently
as the third locus implicated in autosomal dominant hypercholesterolemia (ADH). The 2 other known genes implicated
in ADH encode the low-density lipoprotein receptor and apolipoprotein B. As an approach toward the elucidation of the
physiological role(s) of NARC-1, we studied its transcriptional regulation.
Methods and Results—Using quantitative RT-PCR, we assessed NARC-1 regulation under conditions known to regulate
genes involved in cholesterol metabolism in HepG2 cells and in human primary hepatocytes. We found that NARC-1
expression was strongly induced by statins in a dose-dependent manner and that this induction was efficiently reversed
by mevalonate. NARC-1 mRNA level was increased by cholesterol depletion but insensitive to liver X receptor
activation. Human, mouse, and rat PCSK9 promoters contain 2 typical conserved motifs for cholesterol regulation: a
sterol regulatory element (SRE) and an Sp1 site.
Conclusions—PCSK9 regulation is typical of that of the genes implicated in lipoprotein metabolism. In vivo, PCSK9 is
probably a target of SRE-binding protein (SREBP)-2. (Arterioscler Thromb Vasc Biol. 2004;24:1454-1459.)
Key Words: cholesterol ? QPCR ? SRE ? HepG2 ? primary hepatocytes
N
ily of subtilases. Its acronym reflects the fact that its mRNA
was upregulated when apoptosis was induced in neuronal
primary cultures and that it is similar to 8 other subtilase-like
proteinases, called proprotein convertases (PCs; Millenium
Pharmaceuticals, patent No. WO 01/57081 A2). PCs are
involved in the processing (and activation) of precursors of
hormones, receptors, surface glycoproteins, etc, which transit
through the secretory pathway.1–3Seven of them, PC1/3,
PC2, furin, PC4, PACE4, PC5/6, and PC7/LPC, recognize
basic sites and belong to the kexin subfamily. The eighth
member, SKI-1/S1P,4,5is classified in the pyrolysin subfam-
ily of subtilases. It has been involved in the processing of
endoplasmic reticulum (ER)–anchored transcription factors
such as sterol regulatory element (SRE)-binding proteins
(SREBPs) and thus plays a key role in cholesterol homeosta-
sis.4,6When cellular cholesterol is low, SREBPs are relocated
from the ER to the Golgi apparatus, where SKI-1/S1P cleaves
in their luminal loop. A second cleavage by the metallopro-
tease S2P in their first transmembrane domain liberates the
cytosolic N-terminal region that goes to the nucleus and
eural apoptosis-regulated convertase (NARC)-1 is a
serine proteinase belonging to the proteinase K subfam-
activates target genes. SREBP-1c, the isoform that is domi-
nant in liver, regulates the lipogenic pathway, whereas
SREBP-2 preferentially activates genes of cholesterol
metabolism.7
See page 1337
NARC-1 is highly expressed in embryonic liver.8It then
decreases in the adult liver but significantly increases after
hepatectomy.8The transcript is also detected transiently in
specific areas such as the telencephalon, skin, kidney, intes-
tine, and cerebellum. It has been hypothesized that NARC-1
may be expressed preferentially in progenitor cells and play a
role in hepatic and neuronal differentiation.8In human, the
NARC-1 gene, PCSK9, is ?22-kb long and comprises 12
exons encoding a 692-aa protein. Located on chromosome
1p32, PCSK9 was identified recently as the third locus
involvedinautosomal dominant
(ADH),9characterized by high levels of low-density lipopro-
tein (LDL) cholesterol, xanthomas, and a high frequency of
coronary artery diseases. The majority of familial hypercho-
lesterolemia cases are attributable to mutations in the genes
encoding the LDL receptor (LDLR) and apolipoprotein B
hypercholesterolemia
Received April 28, 2004; accepted May 18, 2004.
From the Laboratory of Hyperlipidemia and Atherosclerosis Research Group (G.D., H.W., J.D., L.B.) and the Laboratory of Biochemical
Neuroendocrinology (A.C., N.G.S., A.P.), Clinical Research Institute of Montreal, Quebec, Canada.
Correspondence to Annik Prat, Laboratory of Biochemical Neuroendocrinology, Clinical Research Institute of Montreal, 110 W Pine Ave, Montreal,
Quebec, Canada H2W 1R7. E-mail prata@ircm.qc.ca
© 2004 American Heart Association, Inc.
Arterioscler Thromb Vasc Biol. is available at http://www.atvbaha.orgDOI: 10.1161/01.ATV.0000134621.14315.43
1454

Page 2

(apoB), the main component of LDL particles.10By genetic
analyses of several French families, 2 exonic NARC-1
mutations, S127R and F216L, were associated with haplo-
types segregating with the disease.9This work was confirmed
recently by the identification of a new PCSK9 mutation,
D374Y, in a large Utah kindred11and 2 Japanese poly-
morphisms, intron 1/C(-161)T and I474V,12all associated
with abnormally high levels of LDL-cholesterol.
The NARC-1 substrate(s) and physiological function(s) are
still unknown. Thus, we decided to check for NARC-1
involvement in cholesterol homeostasis by studying its reg-
ulation under various conditions known to modulate the
expression of genes involved in cholesterol metabolism.
Quantitative RT-PCR (QPCR)
NARC-1 mRNA levels in the human hepatic cell line HepG2,
or human primary hepatocytes. Here we report that statins,
which inhibit 3-hydroxy-3-methylglutaryl (HMG)-coenzyme
A (CoA) reductase,15,16a key enzyme in cholesterol synthe-
sis, increased NARC-1 expression, most likely through
SREBP-2 activation. In contrast, NARC-1 expression was not
affected by liver X receptor (LXR) stimulation. In agreement,
the NARC-1 promoter shows the typical association of
conserved Sp1 and SRE-1 sites but no LXR response
element.
13,14was used to measure
Methods
Materials
Lovastatin and simvastatin were kindly provided by Merck Frosst
(Montreal, Canada). Cerivastatin, atorvastatin, and pitavastatin were
a generous gift from Bayer (Toronto, Canada), Parke-Davis (Ann
Harbor, MI) and Kowa (Tokyo, Japan), respectively. Stock solutions
containing 10 mmol/L cerivastatin and atorvastatin (both in water),
lovastatin and simvastatin (both in ethanol), and pitavastatin (in
dimethyl sulfoxide) were stored at ?20°C until use. Compactin
(mevastatin), mevalonolactone, 25-hydroxycholesterol, 22(R)-
hydroxycholesterol, 9-cis-retinoic acid, and cholesterol were pur-
chased from Sigma.
Cell Culture
The human hepatoma HepG2 cells were grown in DMEM containing
100 ?mol/L nonessential amino acids, 100 U/mL penicillin, 100
?g/mL streptomycin, and 10% FBS or lipoprotein-deficient serum
(LPDS), in a humidified atmosphere (5% CO2, 37°C). All culture
reagents were from GIBCO/BRL (Invitrogen; Grand Island, NY).
LPDS was prepared by ultracentrifugation as described17and stored
at ?20°C. Typically, 106cells per well were seeded in 6-well plates
and grown to 70% to 80% confluence. Primary hepatocytes were
isolated from patients undergoing hepatectomy and were a generous
gift from Dr Marc Bilodeau with agreement of the institutional
review committee of St-Luc Hospital (Montreal, Canada) and in-
formed consent of the patients. These cells were grown in William’s
E medium with the same additives. Before seeding, the plates were
coated with a thin layer of rat tail collagen type-1 (BD Biosciences)
at 5 ?g/cm2.
Secreted ApoB-100
After the required incubation, the medium was collected and centri-
fuged 10 minutes at 1500 rpm to remove cell debris. Secreted
apoB-100 levels in the medium were measured by a sandwich
ELISA as described,18using a goat human apoB-48/100 antibody
(Biodesign; Saco, ME).
Intracellular Lipids
Intracellular lipids were extracted as described19. Total cholesterol
and triglycerides were determined by enzymatic methods (Roche
Diagnostics). The remaining cells were digested in 2 mL 0.1N
NaOH, and total proteins were quantified.20
RNA Preparation and cDNA Synthesis
Cells were washed 3? with PBS and directly incubated with Trizol
reagent (Life Technologies). Total RNA was extracted according to
the recommendations of the manufacturer and resuspended in ?30
?L of water. Isolated RNA integrity was electrophoretically verified
by ethidium bromide staining and optical density (OD), with an
OD260/OD280average absorption ratio of 1.8 to 2.0.
Typically, 250 to 600 ng of total RNA were used for cDNA
synthesis in a total volume of 20 ?L using SuperScript II reverse
transcriptase, 25 ?g/mL oligo(dT)12–18, 0.5 mmol/L 2?-
deoxynucleoside 5?-triphosphates, and 40 U of RNaseOUT, all
products from Life Technologies, and used according to the recom-
mendations of the manufacturer.
Quantitative RT-PCR
In a typical experiment, each cDNA sample was submitted to 2
polymerase chain reaction (PCR) amplifications: one for the normal-
izing ribosomal protein S14 gene and the other for the gene of
interest, each in triplicate. Each reaction was in a final volume of 25
?L using the QuantiTec SYBR green PCR master mix from Qiagen,
cDNA dilutions that gave threshold cycle (Ct) values for both
amplifications, and primers for S14 or the chosen target gene
(Table). All primers (Life Technologies) were designed using Prim-
er3 software to produce amplicons that overlap exonic splicing
junctions to avoid genomic DNA amplification. Oligonucleotide Cts
were optimized for each amplification. The PCR program comprised
a polymerase activation step (15 minutes at 95°C) followed by 40
cycles of 30 seconds at 94°C, 30 seconds at 58°C and 30 seconds at
72°C. The Mx4000 system from Stratagene was used to perform and
analyze the QPCR reactions, using S14 amplifications as normalizers
and control samples as calibrators. Excel software was used for SD
and Student test calculations. The data shown correspond to repre-
sentative experiments.
Results
NARC-1 Is Upregulated by Statins
Because NARC-1 seemed to be involved in cholesterol
metabolism and that patients responded well to statin treat-
ment, we studied the effect of 5 different statins on NARC-1
Primers for QPCR
Assessed mRNA Forward PrimerReverse Primer
human S14
human NARC-1
human LDLR
human SKI-1
human SREBP-1
human SREBP-2
GGCAGACCGAGATGAATCCTCA
ATCCACGCTTCCTGCTGC
AGGAGACGTGCTTGTCTGTC
CAAGTGAGGCCCTTGTCC
ATCGACTACATTCGCTTTCT
AGAATGTCCTTCTGATGTCC
CAGGTCCAGGGGTCTTGGTCC
CACGGTCACCTGCTCCTG
CTGAGCCGTTGTCGCAGT
CTCCCAGGAAGGCAAAGAC
CAGATCCTTCAGAGATTTGC
GGAGAGTCTGGCTCATCTT
Dubuc et al NARC-1 Gene, PCSK9, Is Upregulated by Statins
1455

Page 3

expression. HepG2 cells were treated with 1 ?mol/L ceriv-
astatin, atorvastatin, lovastatin, simvastatin, or pitavastatin.
Expression of the LDLR gene, which is known to be
upregulated by statins, was used as a positive control. Statin
treatment significantly increased PCSK9 expression by a
factor of ?3 for cerivastatin and 1.5 for the other statins
(Figure 1A). This increase was confirmed when HepG2 cells
were treated with increasing Cts of statins. The dose-
dependent response induced by atorvastatin is shown in
Figure 1B. PCSK9 expression was upregulated ?7.5-fold by
10 ?mol/L atorvastatin, whereas under the same conditions,
LDLR expression increased by only 2.5-fold. By inhibiting
HMG-CoA reductase, an enzyme of the cholesterol synthesis
pathway, statins induce a cellular depletion in cholesterol. In
contrast to NARC-1 and LDLR, SKI-1 that cleaves SREBPs
on cholesterol depletion was not regulated at the transcrip-
tional level by statin treatment (Figure 1B).
NARC-1 Is Upregulated by Cholesterol Depletion
To verify whether NARC-1 mRNA upregulation was attrib-
utable to inhibition of cholesterol synthesis or to another
effect of statins, we quantified PCSK9 expression in the
presence or absence of sterols, a mixture of cholesterol and
25-hydroxycholesterol (Figure 2). In both HepG2 cells and
human hepatocytes in primary culture, NARC-1 mRNA level
was upregulated, albeit to a higher extent in HepG2 cells (5
versus a 2-fold increase in primary hepatocytes). In HepG2
cells, LDLR and SREBP-2 mRNA levels were also upregu-
lated (2-fold), whereas that of SREBP-1 remained unaffected,
as reported previously.21,22As expected, in the absence of
exogenous sterols, the intracellular cholesterol and the levels
of apoB in the medium were reduced, whereas intracellular
triglycerides were increased (Figure 2, inset). These data
suggested that statin-induced upregulation of NARC-1 was
mediated by the cholesterol-lowering effect of statins.
Statin-Induced Upregulation of NARC-1 Was
Reversed by Addition of Mevalonate
The above hypothesis was verified by treating HepG2 cells or
primary hepatocytes with atorvastatin in the absence or
presence of mevalonate (Figure 3). Mevalonate was expected
to prevent the cholesterol depletion caused by statin because
it is the product of the reaction catalyzed by HMG-CoA
reductase. Analysis of NARC-1 and LDLR expression levels
showed that addition of mevalonate efficiently reversed
NARC-1 and LDLR upregulations. As in Figure 2, NARC-1
upregulation was higher than that of LDLR (2-fold). Inter-
estingly, NARC-1 downregulation by addition of mevalonate
was also more drastic than that of LDLR. It is important to
note that in HepG2 cells and hepatocytes at 1 ?mol/L statin,
mevalonate lowers NARC-1 (but not LDLR) level below that
of the control, suggesting that PCSK9 is regulated more
tightly by cholesterol than the LDLR gene.
NARC-1 Is Not Affected by LXR Induction
Although SREBP-2 is the prominent factor that regulates
cholesterol synthesis and uptake, the transcription factor LXR
plays a key role in cholesterol elimination.23LXR is a nuclear
hormone receptor that binds oxysterols and activates its target
genes, such as CYP7A encoding the rate-limiting enzyme in
the conversion of cholesterol to bile acids by dimerizing with
retinoid X receptor (RXR).24PCSK9 regulation was assessed
in the absence or presence of 22(R)-hydroxycholesterol, 1 of
Figure 1. NARC-1 is upregulated by statins in HepG2 cells.
HepG2 cells were treated (A) for 24 hours with 1 ?mol/L of cer-
ivastatin, atorvastatin, simvastatin, or pitavastatin (3 to 4 mRNA
samples were analyzed in duplicates) and (B) for 48 hours with
atorvastatin at various Cts (2 independent experiments, each
with 4 mRNA samples in duplicate). Gene expression levels
were quantified by QPCR. Bars represent the mean?SD, and
asterisks represent P values according to the Student test
(*P?0.05; **P?0.01; ***P?0.001).
Figure 2. NARC-1 is upregulated by cholesterol depletion.
Human primary hepatocytes or HepG2 cells were incubated for
18 hours in 5% LPDS, 50 ?mol/L mevastatin, and 50 ?mol/L
mevalonolactone in absence (?sterols) or presence (?sterols) of
1 ?g/mL 25-hydroxycholesterol and 10 ?g/mL cholesterol.
Gene-specific expression was measured by QPCR. Two to 3
cDNAs were analyzed in triplicate. Bars represent the
mean?SD, and 3 asterisks represent a P value of ?0.001 (Stu-
dent test). Values for intracellular cholesterol (chol.) and triglyc-
erides (TGs) and extracellular apoB are given in micrograms per
milligram of cellular proteins (inset).
1456Arterioscler Thromb Vasc Biol.
August 2004

Page 4

the most potent oxysterols25for LXR activation (Figure 4).
Because LXR also plays a role in fatty acid metabolism
through SREBP-1 upregulation,26we measured both
SREBP-1 and SREBP-2 expression levels as positive and
negative controls, respectively. As expected, SREBP-2
mRNA remained stable whereas SREBP-1 mRNA was up-
regulated 3-fold in the presence of 22(R)-hydroxycholesterol,
an effect comparable to the 2.5-fold increase found in HepG2
cells stimulated with a synthetic LXR agonist26or in mice fed
the same agonist.27Under these conditions, NARC-1 expres-
sion level was not upregulated.
Comparative Analysis of Human, Mouse, and Rat
PCSK9 Promoters
Regulation of NARC-1 mRNA levels by cholesterol was
strongly in favor of a SREBP-2–mediated effect. The latter
activates cholesterol biosynthetic genes by binding to SREs
exhibiting adjacent sites for Sp1 or nuclear factor-Y (NF-Y)
cofactors.28,29Human, mouse, and rat promoters were ana-
lyzed using the MatInspector software, and the identified
consensus binding motifs for SREBPs, Sp1, and NF-Y are
represented schematically in Figure 5. Both mouse and rat
sequences exhibit, in addition to an ATG codon aligned with
that of the human sequence, an upstream ATG that extents
their open reading frame of 13 and 55 codons, respectively.
Whether these upstream ATGs are bona fide translation
initiation sites remains to be determined. Only 2 sites,
separated by ?75 bp, were conserved perfectly in the PCSK9
proximal promoter of the 3 species: an SRE (ATCACGC-
CAC) at ?337, ?227, and ?218, and an Sp1 site (GGG-
GAGGGC) at ?430, ?320, and ?313 in human, mouse, and
rat sequences, respectively. In the LDLR promoter, the
orientation of the SRE-1 (?159; ATCACCCCAC) and the
most important Sp1 site (?144; GGGGAGGAG) is inverted
Figure 4. NARC-1 is insensitive to LXR activation. HepG2 cells
were washed twice in PBS for 1 hour and incubated for 24 hours
in DMEM containing 2 mg/mL BSA with either vehicle (CTL) or 2.5
?g/mL 22-hydroxycholesterol and 10 ?mol/L 9-cis-retinoic acid
(RA), which are LXR and RXR39activators, respectively. Six cDNAs
were analyzed in triplicate. Bars represent the mean?SD, and 3
asterisks represent a P value of ?0.001 (Student test).
Figure 5. Comparative analysis of human, mouse, and rat
NARC-1 promoters. The Mat Inspector software 2.2 was used
to detect SRE/E-box, Sp1, and NF-Y consensus motifs. No
information was available on transcription or translation initiation
sites. Position ?1 was fixed arbitrarily to the nucleotide preced-
ing the conserved ATG codon, which is the first in human and
the second in mouse and rat sequences. The shift between the
human and mouse/rat numbering is attributable to important
gaps (?110 positions of the human sequence) in mouse and rat
sequences in the proximal region. Density of dots in the thick
dotted line correlates with sequence identity. Open arrows rep-
resent SRE and Sp1 site orientations. The question mark and
the asterisks indicate regions that may or may not be translated
and variable nucleotide positions, respectively.
Figure 3. The statin-induced upregulation of NARC-1 is
reversed by addition of mevalonate. After a 24-hour preincuba-
tion in 10% LPDS, HepG2 cells were incubated in the same
medium for 48 hours in absence (CTL) or presence of 1 or
10 ?mol/L atorvastatin with (?) or without (?) 2.5 mmol/L me-
valonolactone (mevalonate). Because of statin toxicity, primary
hepatocytes (inset) were treated with only 1 ?mol/L atorvastatin
for 24 hours after a 12-hour preincubation. Gene-specific
expression was measured by QPCR. Two to 4 cDNAs were an-
alyzed in triplicate. Bars represent the mean?SD, and asterisks
represent P values according to the Student test (*P?0.05;
**P?0.01; ***P?0.001).
Dubuc et al NARC-1 Gene, PCSK9, Is Upregulated by Statins
1457

Page 5

compared with that observed in PCSK9 promoter. As sym-
bolized by the dotted line in Figure 5, PCSK9 SRE and the
Sp1 site are comprised in an ?200-nucleotide conserved area
of the promoter (basically no gaps and 90% identity between
human and mouse sequences). In the proximal region, aside
from important gaps in mouse and rat sequences (equivalent
to 37% of the human sequence), human and mouse promoters
share only 66% identity. In the distal region, the identity is
58%. Two (human) and 1 (mouse and rat) other SREs with
adjacent Sp1 and NF-Y sites were present further upstream in
the PCSK9 promoter (data not shown) and may also contrib-
ute to gene regulation by sterols. Consistent with the absence
of a significant increase of NARC-1 mRNA expression by
LXR activation (Figure 4), no conserved LXR response
element was detected in the PCSK9 promoter.
Discussion
Our results showed that the NARC-1 gene, PCSK9, involved
in familial hypercholesterolemia,9,11,12is regulated as a typi-
cal cholesterogenic gene. We showed for the first time that
NARC-1 mRNA expression was upregulated by statins (Fig-
ure 1) and cholesterol depletion (Figure 2), ?2-fold more
than that of LDLR. The statin-induced upregulation of
PCSK9 was reversed quantitatively by addition of meval-
onate (Figure 3). This indicated that the effect of statins was
attributable to the inhibition of HMG-CoA reductase and not
to the other effect(s) of pleiotropic statins.16,30Therefore,
NARC-1 upregulation was most likely a result of the choles-
terol-lowering effect of the drug. This is the first study of
NARC-1 regulation in human cells. The data obtained in
HepG2 cells were similar to that observed in human hepato-
cytes in primary culture (Figures 2 and 3), thereby validating
the studies in HepG2 cells.
Our in vitro results are in agreement with a recent study by
Breslow et al, who identified PCSK9 as a putative family
member of the genes involved in cholesterol homeostasis
using a DNA microarray approach.31They found an ?2-fold
decrease in NARC-1 mRNA levels in livers of mice fed a
cholesterol-rich diet. In contrast, PCSK9 was highly upregu-
lated in SREBP-2 transgenic mice.
Interestingly, they also showed a slight increase (1.6-fold)
in NARC-1 transcripts in liver from mice treated with an
LXR agonist.31The latter observation could not be confirmed
in our study, a discrepancy that may reflect species-specific
regulations or in vivo versus in vitro variations. In addition,
genes for which expression is repressed by an excess of
cholesterol and upregulated by SREBP-2 are usually not LXR
targets. Finally, another DNA microarray study confirmed the
above results and showed that in SREBP cleavage-activating
protein?/? liver, in which SREBP activation cannot take
place, NARC-1 messengers were reduced.32
Human, mouse, and rat NARC-1 promoters share 2 con-
served sites for transcription factor binding: Sp1 and SRE.
The NARC-1 SRE differs from the classical LDLR SRE-1
(ATCACCCCAC) by 1 transversion at position 6 (C to G),
shown not to affect SREBP-2 binding.33It has been suggested
that in LDLR promoter, the respective directional binding of
SREBP-2 and Sp1 to their sites, which are head to head in
repeats 2 and 3, allows the interaction of SREBP-2 with the
N-terminal region of Sp1 and facilitates Sp1 recruitment28
(Figure 5). Because of the mirror image of SRE and Sp1 site
disposition in the PCSK9 promoter, it is possible that the
same interaction takes place. In vitro studies7demonstrated
that SREBP-2 binds efficiently to classical SRE, whereas
SREBP-1c shows little binding. The absence of a conserved
E-box, which is known to be preferred by SREBP-1c,7is also
in favor of NARC-1 regulation by SREBP-2. NARC-1 is
particularly abundant in liver and small intestine.8Further
studies will define whether the sterol regulation of the PCSK9
requires tissue-specific factors, such as hepatocyte nuclear
factor-4, which was shown to be essential for SREBP-2
activation of sterol ?8-isomerase.34
The absence of an LXR-mediated upregulation of NARC-1
is not in favor of its implication in cholesterol catabolism.
Our data rather suggest that the enzyme is implicated in
cholesterol biosynthesis or uptake. The fact that patients
harboring a mutated PCSK9 have high plasma cholesterol
levels reinforces a putative role of NARC-1 in LDL uptake.
The autoprocessing site of NARC-1 has been identified
recently35(Benjannet et al, submitted) and further studies that
will better define NARC-1 cleavage specificity should help in
identifying NARC-1 substrate(s). The dominant character of
PCSK9 mutations could be attributable either to a dominant-
negative or gene-dosage effect. The S127R9and D374Y11
mutations that have been associated with ADH partially and
totally abrogated NARC-1 autocatalytic zymogen processing,
respectively (Benjannet et al, submitted), supporting a corre-
lation with the enzyme activity. However, we cannot exclude
that these mutations generated dominant-negative forms of
the enzyme that acquired novel deleterious properties, which
may have no relation to the normal physiological function(s)
of NARC-1, as reported for superoxide dismutase 1.36Nev-
ertheless, because the enzyme belongs to the PC family, it is
tempting to hypothesize that an as yet unknown substrate(s)
activated by NARC-1 is an essential actor in the cholesterol
pathway. Identification of NARC-1 substrates may lead to
elucidation of disease mechanism(s), and these substrates
may constitute targets for new strategies to limit elevation of
LDL particles and prevent morbidity and mortality from
premature atherosclerosis.
In addition, and as reflected by its acronym, NARC-1 was
shown to be upregulated by apoptosis induction in primary
culture of neurons. Caspase 3, which has a pivotal role in
apoptosis, was reported to generate active SREBPs.37,38By
using reporter genes under control of the LDLR SRE and Sp1
sites, Higgins and Ioannou38showed that the 2 elements
mediated a sterol-independent upregulation of the reporter
genes very early in apoptosis induction. The physiological
relevance of stimulation of SREBP targets in apoptotic cells
remains unclear. It will be important to verify whether the
identified PCSK9 SRE and Sp1 site also mediate the in-
creased expression of NARC-1 observed during apoptosis
induction. To date, we cannot exclude that NARC-1 may play
an important role in cholesterol homeostasis and apoptosis,
both of which may be related functionally.
Acknowledgments
G.D. is a recipient of a Clinical Research Institute of Montreal
studentship award. This work was supported by the Canadian
1458Arterioscler Thromb Vasc Biol.
August 2004