Donate Help Contact The AHA Sign In Home
American Heart Association
Circulation Research
Search: search_blue_button Advanced Search
« Previous Article | Table of Contents
Circulation Research. 2004;94:e46-e54
Published online before print March 4, 2004, doi: 10.1161/01.RES.0000124761.62846.DF
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
94/5/e46    most recent
01.RES.0000124761.62846.DFv1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Xu, Q.
Right arrow Articles by Duan, C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Xu, Q.
Right arrow Articles by Duan, C.
Related Collections
Right arrow Mechanism of atherosclerosis/growth factors
Right arrow Gene regulation
Right arrow Growth factors/cytokines
Right arrow Smooth muscle proliferation and differentiation
(Circulation Research. 2004;94:e46.)
© 2004 American Heart Association, Inc.

UltraRapid Communication

Evidence That IGF Binding Protein-5 Functions as a Ligand-Independent Transcriptional Regulator in Vascular Smooth Muscle Cells

Qijin Xu, Shenghua Li, Yang Zhao, Travis J. Maures, Ping Yin, Cunming Duan

From the Department of Molecular, Cellular, and Developmental Biology, University of Michigan, Ann Arbor, Mich.

Correspondence to Cunming Duan, Department of MCDB, University of Michigan, Natural Science Building, Room 3065B, Ann Arbor, MI 48109-1048. E-mail cduan{at}umich.edu


*    Abstract
up arrowTop
 *Abstract
down arrowIntroduction
 down arrowMaterials and Methods
 down arrowResults
 down arrowDiscussion
 down arrowReferences
 
Insulin-like growth factor binding protein (IGFBP)-5 is a conserved protein synthesized and secreted by vascular smooth muscle cells (VSMCs). IGFBP-5 binds to extracellular IGFs and modulates IGF actions in regulating VSMC proliferation, migration, and survival. IGFBP-5 also stimulates VSMC migration through an IGF-independent mechanism, but the molecular basis underlying this ligand-independent action is unknown. In this study, we show that endogenous IGFBP-5 or transiently expressed IGFBP-5-EGFP, but not IGFBP-4-EGFP, is localized in the nuclei of VSMCs. Using a series of IGFBP-4/5 chimeras and IGFBP-5 points mutants, we demonstrated that the IGFBP-5 C-domain is necessary and sufficient for its nuclear localization, and residues K206, K208, K217, and K218 are particularly critical. Intriguingly, inhibition of protein secretion abolishes IGFBP-5 nuclear localization, suggesting the nuclear IGFBP-5 is derived from the secreted protein. When added exogenously, 125I- or Cy3-labeled IGFBP-5 is capable of cellular entry and nuclear translocation. To identify potential transcriptional factor(s) that interact with IGFBP-5, a human aorta cDNA library was screened by a yeast two-hybrid screening strategy. Although this screen identified many extracellular and cytosolic proteins that are known to interact with IGFBP-5, no known transcription factors were found. Further motif analysis revealed that the IGFBP-5 N-domain contains a putative transactivation domain. When fused to GAL-4 DNA dinging domain and tested, the IGFBP-5 N-domain has strong transactivation activity. Mutation of the IGF binding domain or treatment of cells with IGF-I has little effect on transactivation activity. These results suggest that IGFBP-5 is localized in VSMC nucleus and possesses transcription-regulatory activity that is IGF independent. The full text of this article is available online at http://circres.ahajournals.org.


Key Words: insulin-like growth factor-1 • nuclear localization signal • vascular smooth muscle cells • transcription factors


*    Introduction
up arrowTop
 up arrowAbstract
 *Introduction
down arrowMaterials and Methods
 down arrowResults
 down arrowDiscussion
 down arrowReferences
 
There is an accumulating body of evidence that locally produced insulin-like growth factors (IGFs) may play a critical role in the development and progression of atherosclerotic plaques.1–5 Most, if not all, IGFs in the extracellular environments are bound to specific, high-affinity IGF binding proteins (IGFBPs). IGFBPs are a family of six secreted proteins that bind IGFs and modulate IGF distribution, stability, and biological activities.6,7 Previous studies have shown that mammalian VSMCs synthesize and secrete IGFBP-2, -3, -4, and -5.1,2,8 These IGFBPs can inhibit or potentiate IGF-I–induced VSMC proliferation and migration.9–14 Recently, we have shown that these locally produced IGFBPs also play an important role in determining whether VSMCs proliferate or migrate in responses to IGF-I stimulation and a key player in this paradigm is IGFBP-5.8,14 IGFBP-5 not only modulates IGF-I actions, but it also directly stimulates VSMC migration through an IGF-independent mechanism(s).14 The molecular basis underlying the ligand-independent action of IGFBP-5 is unknown.

Ligand-independent actions of IGFBP-5 on cell growth and differentiation have been documented in human osteoblast cells and breast cancer cells,15–19 although the underlying molecular mechanism(s) remains poorly understood. In human osteoblast cells, IGFBP-5 has been reported to bind to a putative cell surface receptor,20 but the molecular identity of this IGFBP-5 "receptor" is still unknown. Recognizing that multiple basic residues in the 201 to 218 region of IGFBP-5 are similar to the bipartite nuclear localization signal (NLS) found in viral and mammalian transcription factors, it was speculated that IGFBP-5 might also act in an "intracrine" fashion, perhaps serving as a cytosol-to-nuclear shuttle for its ligand.21 Indeed, when added to cultured human bone tumor and breast cancer cells, exogenous IGFBP-5 was shown to be capable of cellular and nuclear entry.22,23 Likewise, when fused to EGFP and transfected into CHO cells, a peptide corresponding to the NLS of IGFBP-5 (residues 201 to 218) targeted EGFP to the nucleus.24 However, it is not clear whether endogenous IGFBP-5 is present in the nucleus in VSMCs or any other cell types, nor is understood where is the nuclear IGFBP-5 is derived from. Moreover, the functional significance of IGFBP-5 in the nucleus and its role, if any, in mediating the IGF-independent actions of IGFBP-5 are unknown in VSMCs or any other cell types.


*    Materials and Methods
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 *Materials and Methods
down arrowResults
 down arrowDiscussion
 down arrowReferences
 
Materials
All chemicals and reagents were purchased from Fisher Scientific unless noted otherwise. The IGFBP-5 antibodies were purchased from Upstate Biotechnology Inc and Diagnostic Systems Laboratories, Inc. The C-Fos and IGF-IR antibodies were purchased from Santa Cruz Biotech, and Akt antibody was obtained from Cell Signaling Inc. Fetal bovine serum (FBS), cell culture media, antibiotics, and trypsin were purchased from GIBCO/BRL. Peptides were synthesized at the Protein Structure Facility, University of Michigan. Peptide BP5(201-218) corresponds to IGFBP-5 amino acid residues Arg201-Arg218 and peptide BP5(131-141) peptide includes IGFBP-5 amino acid residues Ala131-Thr141. They were purified by HPLC to >90% purity and analyzed by mass spectroscopy before use.

Cell Culture, Subcellular Fractionation, Immunological Assays, and Ligand Blot
Porcine VSMCs were isolated and cultured as reported previously.25 Human embryonic kidney (HEK) 293, A7R5, and CHO-K1 cells were obtained from ATCC and were cultured as recommended. For immunostaining, cells were grown on 8-chamber slides (Falcon). After permeabilization and blocking in 3% BSA/1% Triton X-100/PBS, they were incubated with the primary antibody at 4°C overnight. After washing, the cells were incubated with appropriate secondary antibodies and microphotographs were taken with a Nikon EC600 Fluorescence Microscope. Cells were fractionated following a previously published method.26 These fractions were processed to immunoprecipitation, immunoblot, and ligand blot analysis as reported previously.14

Plasmid Construction
For the construction of IGFBP-EGFP expression plasmids, DNA fragments corresponding to the signal peptide plus N-domains (containing residues -21 to 79 of IGFBP-4 and residues -20 to 80 of IGFBP-5), L-domains (containing residues 80 to 150 and 81 to 169 of IGFBP-4 and -5, respectively), C-domains (containing residues 151 to 237 and 170 to 252 of IGFBP-4 and -5), NL-domains, and LC-domains of human IGFBP-4 and -5 were generated by PCR amplification as described previously.27 DNA fragments encoding various IGFBP-5 point mutants were generated by PCR using mutant DNA provided by Dr D.R. Clemmons, University of North Carolina at Chapel Hill. The fragments were subcloned into the HindIII/KpnI or XhoI/KpnI sites of pEGFP-N1 vector (Clontech). The DNAs were transfected into cells using LipofectAMINE method.14 The transfected cells were washed and photographed. To produce Gla4-DBD and IGFBP fusion proteins, DNA fragments corresponding to the N-domain of human and zebrafish IGFBP-5 and -1 were generated by PCR and subcloned into the EcoRV/XbaI sites of pBIND vector (Promega) in frame. All plasmids were confirmed by DNA sequencing.

Yeast Two-Hybrid Screen
The screen was performed as reported recently.27 Briefly, the Matchmaker two-hybrid system 3 (CLONTECH) was used to identify candidate proteins that interact with IGFBP-5. The bait, pGBKT7-IGFBP-5, generated by inserting full-length human IGFBP-5 cDNA into the NcoI and BamH1 sites of the pGBKT7, was used to screen a human aorta cDNA library constructed in the pACT2 vector (CLONTECH). Positive clones were identified, and they were retested twice under high stringency.

Internalization Assays
The internalization of 125I-labeled IGFBP-5 was performed following previously published method.15,26 Peptide BP5(201-218) was labeled with activated Cy3 dye (Amersham Pharmacia Biotech) following the manufacturer’s protocol. The labeled peptide was purified by a Sephadex G-25 column and was eluted with 1x PBS (pH7.4). The collected fractions were analyzed by spectrophotometer at wavelengths of 280 and 552 nm and the labeled peptide was quantified using a standard curve of the pure peptide. The calculated dye/peptide ratio was about 0.4. For the internalization assay, cells were grown in the 4-well or 8-well chamber slides (Nalge Nubc Intl). After washing with serum-free medium (SFM) containing 0.1% BSA three times, cells were incubated with Cy3-labeled BP5(201-218) (2.5 µg/mL) or Cy3-IgG ( 8.0 µg/mL) in SFM containing 0.1% BSA for 4 hours at 37°C. After incubation, the cells were washed twice with 2 mol/L NaCl in 1x PBS followed by further washings with 1x PBS. The cells were then fixed with 4% paraformaldehyde, stained with DAPI, and examined under a fluorescence microscope. The internalization of IGFBP-5-EGFP was performed under similar conditions.

One-Hybrid Transcription Assay
The transcription activation activity of IGFBP-5 N domain was determined following Kabuta et al.28 Briefly, cells were cotransfected with pG5-luc and each pBIND-IGFBP fusion construct. Twenty-four hours after transfection, cells were washed and lysed. The lysates were measured for firefly and Renilla luciferase activities using the Dual-Luciferase Reporter assay system (Promega). The result was expressed as fold change over the pBIND empty vector group. Transfection efficiency was normalized by Renilla luciferase activity.

Statistical Analysis
Values are mean±SE. Differences among groups were analyzed by one-way analysis of variance followed by Fisher’s protected least significance difference test using Stat-View (Abacus Concept, Inc).


*    Results
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 up arrowMaterials and Methods
 *Results
down arrowDiscussion
 down arrowReferences
 
IGFBP-5 Is a Nuclear Protein in VSMCs
To determine whether endogenous IGFBP-5 is present in the nuclei of VSMCs, primarily cultured porcine VSMCs were subjected to immunocytochemical staining using a polyclonal and a monoclonal IGFBP-5 antibody. These antibodies do not cross-react with other IGFBPs.10,14 As shown in Figure 1A, IGFBP-5 was detected in the nuclei by both antibodies (Figures 1Aa and 1Ac). Preabsorption of the IGFBP-5 antibodies by excess IGFBP-5 abolished the nuclear IGFBP-5 signal (Figure 1Ad), confirming the authenticity of the antibody staining. Figure 1Ad shows an overexposed view in order to appreciate the complete lack of nuclear signal. As a control, VSMCs were stained with a c-Fos (a nuclear protein) antibody and an IGF-IR (a transmembrane receptor kinase) antibody (Figures 1Ae and 1Af). Next, cells were fractionated into cytosolic, membrane, and nuclear fractions. These fractions were precipitated by the IGFBP-5 antibody and analyzed by ligand blot using 125I-IGF-I. A 30-kDa protein in the nuclear and membrane fractions was precipitated by the IGFBP-5 antibody, and this protein was capable of 125I-IGF-I binding (Figure 1B). These fractions were also subjected to immunoblot using antibodies against IGF-IR, Akt, and c-Fos. IGF-IR was detected in the membrane fraction, whereas Akt was seen predominantly in the cytoplasmic faction (Figure 1B). c-Fos, on the other hand, was exclusively detected in the nuclear fraction, suggesting that these fractions were relatively clean. We further investigated the nuclear localization of IGFBP-5 by transiently expressing an IGFBP-5-EGFP fusion protein and observing the EGFP signal in living cells. EGFP and IGFBP-4-EGFP were used as controls. EGFP is a small protein that can move between the nuclear and cytoplasmic compartments. When transfected to VSMCs, EGFP was detected in both the cytoplasm and nucleus (Figures 1Ca and 1Cb). The IGFBP-5-EGFP signal, on the other hand, was seen predominantly in the nucleus (Figures 1Cc and 1Cd). In comparison, the IGFBP-4-EGFP was not seen in the nucleus (Figures 1Ce and 1Cf). Similar results were obtained in A7R5 cells, a transformed cell line derived from rabbit aorta (Figure 1Cg and 1Ch). The easily transfectable A7R5 cells were therefore used in subsequent experiments. Data from five independent experiments indicated 83±2% of the IGFBP-5-EGFP–transfected cells showed exclusive nuclear presence, whereas none of the IGFBP-4-EGFP–transfected cells showed nuclear signal.



View larger version (29K):
[in this window]
[in a new window]
  		Figure 1. IGFBP-5 is not only a secreted protein but is also localized in the nucleus. A, Immunofluorescence staining of VSMCs. Immunostaining of VSMCs using a polyclonal IGFBP-5 antibody (a), a monoclonal IGFBP-5 antibody (c), the polyclonal IGFBP-5 antibody after it was preabsorbed by excess IGFBP-5 protein (d), a polyclonal c-Fos antibody (e), and a polyclonal IGF-IR antibody (f). b is the corresponding DAPI staining of a. B, Subcellular localization of IGFBP-5; cytosolic (C), membrane (M), and nuclear fractions (N) were precipitated by a polyclonal IGFBP-5 antibody and analyzed by ligand blot using 125I-IGF-I. The same fractions were analyzed by immunoblot using antibodies against IGF-IR, Akt, and c-Fos. C, Nuclear localization of IGFBP-5-EGFP fusion protein in living cells. pEGFP-N1 (a), pEGFP-IGFBP-5 (c), or pEGFP-IGFBP-4 (e) was transiently expressed in VSMCs and visualized. pEGFP-IGFBP-5 was also expressed in A7R5 cells (g). b, d, f, and h are the corresponding DAPI staining.

Several Basic Residues in IGFBP-5 C-Domain Are Critical for IGFBP-5 Nuclear Localization
IGFBP-5 and IGFBP-4 share similar domain structure, substantial sequence identity, and the ability to bind IGFs with high affinities, but only IGFBP-5 is localized in the nucleus. We therefore used a strategy based on IGFBP-4/5 chimeras to map the unique region in IGFBP-5 responsible for its nuclear localization. Six IGFBP chimeras, namely, IGFBP-545, -454, -554, -445, -544, and -455, were generated by exchanging the N-, L-, and C-domain of the two IGFBPs (Figure 2A). Ligand blot analysis revealed that all six chimera proteins retained IGF binding ability (data not shown). The six IGFBP chimeras were fused to EGFP and introduced into A7R5 cells by transient transfection. As shown in Figure 2B, the three chimeras lacking the IGFBP-5 C-domain, namely IGFBP-544, -454, and -554, exhibited little or no nuclear presence (0%, 3.6±1.2%, and 4.8±1.2%, respectively). In contrast, IGFBP-455, -445, and -545, all containing the IGFBP-5 C-domain, were predominantly nuclear (88±8.4%, 76±3.6%, and 65±1.2%, respectively), indicating that the IGFBP-5 C-domain is necessary and sufficient for its nuclear localization.



View larger version (28K):
[in this window]
[in a new window]
  		Figure 2. Several basic residues in IGFBP-5 C-domain are critical for IGFBP-5 nuclear localization. A, Schematic diagram showing the structure of various IGFBP 4/5 chimeras. B, Nuclear localization of the IGFBP-EGFP chimeras. Values are mean±SE of 2 separate experiments. In each experiments, more than 300 cells were counted. C, Nuclear presence of various mutant IGFBP-5-EGFP. Values are mean±SE of 3 separate experiments.

The human IGFBP-5 C-domain contains a putative NLS motif between residues 201 to 218. This sequence is perfectly conserved in all known vertebrate IGFBP-5 homologues ranging from human to zebrafish, but is absent in IGFBP-4. Mutation of K217/R218 into 217A/218A reduced the nuclear presence by 72% (Figure 2C). Mutation of two additional basic residues in this half, ie, changing K211/R214/K217/R218 into 211N/214A/217A/218A, resulted in an 82% reduction. Changing R201/K202 into A201/N202 caused a modest 17% decrease. Further mutation of K206/K208 into N206/N208 on the A201/N202 background resulted in a 66% reduction (Figure 2C). These data suggest that K206, K208, K217, and K218 are critical. To prove that the particular sequence in the NLS motif of IGFBP-5, but not the charge changes introduced by these mutations, is critical, four basic residues in the L-domain, K133/R136/K138/K139, were changed into N133/A136/A138/A139. This mutant acted as the wild-type protein (Figure 2C).

Nuclear IGFBP-5 Is Derived From Secreted Protein
To show that the IGFBP-5-EGFP fusion protein is secreted and retains IGF binding property and to determine whether the nuclear IGFBP-5 is derived from the secreted protein, stable CHO-K1 cell lines expressing IGFBP-5-EGFP were obtained. Analysis of the media conditioned by the transfected cells indicated the presence of a secreted protein at the predicted size (55 kDa). We conclude this to be IGFBP-5-EGFP because it bound 125I-IGF-I and was immunoreactive to a GFP antibody (Figure 3A). When brefeldin A was added, IGFBP-5-EGFP disappeared from the conditioned media (Figure 3B), confirming that the IGFBP-5-EGFP detected was originated from secretion. Intriguingly, the inhibition of IGFBP-5-EGFP secretion abolished the nuclear localization of IGFBP-5-EGFP, whereas it increased its cytoplasmic presence (Figures 3Ca and 3Cc). This suggests that the nuclear IGFBP-5 is not simply imported from cytoplasm but derived from the secreted protein.



View larger version (64K):
[in this window]
[in a new window]
  		Figure 3. Nuclear IGFBP-5-EGFP is derived from the secreted protein. A, IGFBP-5-EGFP is secreted and capable of IGF binding. Serum-free media were conditioned by stable CHO-K1 clones expressing IGFBP-5-EGFP (lanes 4, 5, and 6), wild-type (lane 1), or mock cells (lanes 2 and 3). They were subjected to ligand blot using 125I-IGF-I (a) and immunoblot using a GFP antibody (b). Secreted IGFBP-5-EGFP is indicated by an arrow. B, Inhibition of IGFBP-5-EGFP secretion by brefeldin A. CHO-K1 cells stably expressing IGFBP-5-EGFP were treated without (lane 1) or with 1.25, 2.5, 5, and 10 µg/mL of brefeldin A (lanes 2 to 5) for 30 minutes, and the media conditioned for 24 hours were collected and analyzed by ligand blot. C, Inhibition of protein secretion abolishes the nuclear localization of IGFBP-5-EGFP. After transfecting with the IGFBP-5-EGFP construct, cells were treated without (a) or with (c) brefeldin A (5 µg/mL) and stained 24 hours later with a GFP antibody. b and d are the corresponding DAPI staining.

To test whether the secreted IGFBP-5 could enter into the cells and subsequently into the nucleus, 125I-iodine labeled human IGFBP-5 was added to cultured VSMCs. 125I-IGFBP-5 was rapidly internalized by VSMCs and reached maximal levels 4 hours after the addition (Figure 4A). In the presence of excess amounts of unlabeled IGFBP-5, but not IGFBP-4, the internalization was significantly reduced (Figure 4B), suggesting that this is a specific event. Likewise, a peptide corresponding to the NLS region of IGFBP-5, peptide BP5(201-218), was translocated into the nucleus with a similar time course (Figures 4Ca and 4Ce). The unclear uptake was abolished when excess amount of unlabeled BP5(201-218) was added (Figures 4Cb and 4Cf). Addition of BP5(131-141), another basic peptide corresponding to residues 131 to 141 of human IGFBP-5, had no such effect (Figures 4Cc and 4Cg). To show that full-length IGFBP-5 is capable of nuclear entry, recombinant IGFBP-5-EGFP was generated. When added to cultured cells, the GFP signal was detected in the nucleus after 6 hours (Figures 4Da and 4Db).



View larger version (28K):
[in this window]
[in a new window]
  		Figure 4. Exogenous IGFBP-5 can enter the cell and nucleus. A, Internalization of 125I-IGFBP-5 by VSMCs. Confluent VSMCs were exposed to 125I-IGFBP-5 at the time indicated. Surface-bound and internalized radioactivity was measured and their ratio is shown. Values are means of 2 experiments. B, Specific 125I-IGFBP-5 uptake by VSMCs. Confluent VSMCs were exposed to 125I-IGFBP-5 and excess amount of unlabeled IGFBP-5 or IGFBP-4 for 4 hours. Internalized radioactivity was measured. Values are means of 2 experiments. C, Nuclear translocation of exogenous IGFBP-5. CHO-K1 cells were incubated with Cy3-labeled BP5(201-218) in the absence (a) or presence of excess amount of unlabeled BP5(201-218) (b) or unlabeled BP5(131-141) (c). e, f, and g are the corresponding DAPI staining. D, Nuclear translocation of exogenous IGFBP-5-EGFP. CHO-K1 cells were incubated with recombinant IGFBP-5-EGFP (1 µg/mL). Cells were washed and photographed after 4 hours of incubation (a). b is the DAPI staining.

IGFBP-5 N-Domain Possesses Transactivation Activity That Is Evolutionarily Conserved
The nuclear localization of IGFBP-5 raised the possibility that it may regulate gene expression directly or indirectly. To identify potential IGFBP-5 interacting proteins, we recently screened a human aorta cDNA library using a yeast two-hybrid screening strategy. Over 60 positive clones were obtained after screening 4.3 million colonies using human IGFBP-5 as bait. As reported in a recent article, 28 of these encode polypeptides corresponding to fibronectin.27 Others encoded extracellular and cytosolic proteins such as fibrin and importin, which are known to interact with IGFBP-5.6,7 No known transcription factors, however, were found by this approach.

By motif analysis, we found that the IGFBP-5 N-domain contains a conserved proline-rich sequence, which is a typical feature of transactivation domains of transcription factors. To determine whether the IGFBP-5 N-domain has any transactivation activity, human IGFBP-5 N-domain was fused to the Gal4-DNA binding domain (DBD) and introduced to VSMCs together with a Gal4 reporter plasmid. As shown in Figure 5A, IGFBP-5 N-domain caused a highly significant increase in Gal-4 reporter gene transcription over the Gal4-DBD control (Figure 5A). An even greater activity (up to 33-fold increase) was observed when tested in human HEK293 cells (Figure 5B). This was probably due to the higher transfection efficiency in HEK293 cells. The transactivation activity is unique to IGFBP-5 N-domain and is evolutionarily conserved. Human IGFBP-1 N-domain had no such activity. Likewise, the N-domain of zebrafish IGFBP-5, but not zebrafish IGFBP-1, caused a significant increase when tested in HEK293 cells although the activity of the fish protein was considerably weaker in these human cells (Figure 5C).



View larger version (14K):
[in this window]
[in a new window]
  		Figure 5. IGFBP-5 N-domain has transactivation activity. A, N-domain of human IGFBP-5 were fused to the Gal4-DBD and introduced into porcine VSMCs together with a Gal4 reporter plasmid by transient transfection. Transfection efficiency was normalized by Renilla luciferase activity, n=3. *P<0.05 compared with the pBIND group. B, N-domain of human IGFBP-5 and IGFBP-1 were fused to the Gal4-DBD. They were introduced into HEK293 cells together with a Gal4 reporter plasmid. Transfection efficiency was normalized by Renilla luciferase activity, n=5. **P<0.01 compared with the pBIND group. Expression of various Gal4 fusion proteins was determined by immunoblot using a Gal4 antibody. C, N-domain of zebrafish IGFBP-5 and IGFBP-1 was fused to the Gal4-DBD and their transcriptional activities determined as described in B, n=7. **P<0.01 compared with the pBIND group.

Nuclear Activity of IGFBP-5 Is Ligand Independent
Because the IGFBP-5 N-domain also contains the high-affinity IGF binding site,29,30 we wondered whether the transactivation activity of IGFBP-5 is ligand dependent. Nuclear uptake of IGF-I has been reported in other cell types.31 We therefore examined the effect of IGF-I treatment. The addition of IGF-I (200 ng/mL) did not alter the transactivation activity (Figure 6A). Next, the activity of mini-IGFBP-5, which comprises residues 40 to 92 of human IGFBP-5, was determined. Mini-IGFBP-5 contains the high-affinity IGF binding site and can bind to IGFs at nanomolar concentrations.29 As shown in Figure 6B, mini-IGFBP-5 had no transactivation activity. In contrast, a construct covering residues 1 to 39 of human IGFBP-5 N-domain (BP5N1-39) had significant transactivation activity (Figure 6B). We further compared the activity of the N-domain of a non-IGF binding IGFBP-5 mutant and wild-type IGFBP-5. This mutant (K68N/P69Q/L70Q/L73Q/L74Q) has a 1000-fold reduced affinity for IGF-I.30 As shown in Figure 6C, it had similar transactivation activity as the wild-type IGFBP-5 (Figure 6C). Taken together, these data indicate that the transactivation activity of IGFBP-5 is ligand independent and the transactivation domain is distinct from the IGF binding domain.



View larger version (14K):
[in this window]
[in a new window]
  		Figure 6. Transactivation activity of IGFBP-5 is IGF independent. A, IGF-I treatment has no effect on the transactivation activity of IGFBP-5. HEK293 cells transfected with the Gal4-DBD-IGFBP-5 N-domain fusion protein and the Gal4 reporter plasmid were treated with or without IGF-I (200 ng/mL). Transcriptional activities determined after 2 hours, n=3. **P<0.01 compared with the pBIND group. B, Transactivation activity can be separated from the IGF binding domain. Regions covering residues 40 to 92 (miniBP-5) and 1 to 39 (BP5N1-39) of human IGFBP-5 N-domain were fused to the Gal4-DBD and their transcriptional activities were determined as described earlier, n=4. **P<0.01 compared with the pBIND group. #P<0.05 compared with the miniBP-5 group. C, Mutation in IGFBP-5 ligand binding domain has no effect on transactivation activity. DNA encoding the N-domain of the non-IGF binding mutant IGFBP-5 (LBD) was fused to the Gal4-DBD and its transcriptional activity was compared with that of the wild-type protein (hBP5N), n=2. **P<0.01 compared with the pBIND.


*    Discussion
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 up arrowMaterials and Methods
 up arrowResults
 *Discussion
down arrowReferences
 
In this study, we have demonstrated that IGFBP-5 is not only secreted but also localized in the nucleus in VSMCs. The nuclear IGFBP-5 has a molecular mass similar to that of the secreted protein and is capable of IGF binding. We have provided several lines of evidence indicating that the nuclear IGFBP-5 is derived from the secreted protein. Our study further reveals the presence of an evolutionarily conserved, functional transactivation domain in IGFBP-5 N-domain. This novel activity of IGFBP-5 is IGF independent and can be physically separated from the IGF binding site. The nuclear localization and nuclear action of IGFBP-5 provides a molecular basis for several prominent IGFBP-5 activities that are IGF independent.

Although the presence of a putative NLS in IGFBP-5 C-domain was previously suggested,21 and exogenous IGFBP-5 was found to be capable of nuclear uptake by cultured human breast and bone tumor cells22,23; to our knowledge, the present study is the first report demonstrating that endogenous IGFBP-5 is localized in VSMC nucleus. This conclusion was supported by several independent experimental approaches, including (1) colocalizing IGFBP-5 with DNA by immunocytochemistry using two independent antibodies, (2) cell fractionation followed with immunoprecipitation and ligand blot analysis, and (3) transfecting VSMCs with IGFBP-5-EGFP fusion constructs. Using a series of IGFBP-4/5 chimeras and IGFBP-5 points mutants, we show that the IGFBP-5 C-domain is both necessary and sufficient for its nuclear localization, and residues K206, K208, K217, and K218 are particularly critical. This is in overall agreement with a previous study in which K206/R207/K208 and R214/G215/R216/K217/R218 were found to be important.22 However, there are key differences between the two studies. In the study by Schedlich et al,22 a peptide corresponding to 201 to 218 of IGFBP-5 was fused to EGFP. The resulted fusion protein is 27 kDa and as such is capable of passive diffusions into the nucleus.32 More importantly, this 27-kDa fusion protein is not a functional IGFBP. The EGFP fusion protein used in our study was generated by fusing EGFP to the C-terminus of the entire coding region of human IGFBP-5. The resulted fusion protein is {approx}55 kDa and therefore cannot enter the nucleus by diffusion. Furthermore, our IGFBP-5-EGFP is secreted and capable of IGF binding.

There are several plausible sources/routes for the nuclear IGFBP-5. First, an intracellular isoform(s) of IGFBP-5 may exist, and it is translocated into the nucleus under the guidance of the NLS. Such a mechanism has been documented for fibroblast growth factor (FGF)-2 and parathyroid hormone related peptide (PTHrP).33,34 In the case of FGF-2, alternative initiation of translation of FGF2 mRNA occurs at three CUG codons downstream of the conventional AUG start codon.33 These CUG-initiated variants of FGF-2s were shown to remain in the cytoplasm and translocate to the nucleus. Retrograde trafficking from ER to the nucleus is another plausible mechanism of translocating the intracellular IGFBP-5 from cytoplasm to the nuclear compartment. It has been suggested that misfolded proteins that would classically undergo degradation in the ER can be back-translocated from the ER to the cytoplasm.35 Finally, the secreted IGFBP-5 may reenter the cell by an endocytosis-dependent pathway. We favor the last model for following reasons: (1) the endogenous IGFBP-5 in the nucleus is the mature protein capable of IGF binding; (2) IGFBP-5-EGFP is secreted as well as localized in the nucleus; (3) exogenously added, functional IGFBP-5 is capable of cell entry and nuclear translocation with relatively rapid kinetics; and (4) inhibition of protein secretion by BFA abolishes the nuclear presence of IGFBP-5-EGFP, whereas it increases its cytoplasmic content. This conclusion is in good agreement with a recent report that IGFBP-3, a closely related member of the IGFBP family, can be internalized through transferrin receptor–mediated and caveolin-mediated endocytic pathways in 22RV1 and PC3 prostate cancer cells.36 The cellular receptor(s) mediated IGFBP-5 internalization is unknown at present. IGFBP-5 has been shown to interact with heparan sulfate (HS) proteoglycans located on VSMC surface, and this interaction is important for the ligand-independent action of IGFBP-5 on VSMC migration.14 Cell surface HS proteoglycans are known to act as cellular receptors for several viruses and viral protein, including HIV tat protein, adeno-associated virus type 2 virions, foot-and-mouth disease type O virus, HSV types 1 and 2, and dengue virus.37 They are also involved in the internalization and nuclear localization of a number of growth factors and cytokines such as fibroblast growth factor-2.38 A recent study indicated that HS proteoglycans themselves can be targeted into the nucleus of cultured primary corneal fibroblasts and this process is regulated by fibronectin.39 Toward this end, it is worthy noting that IGFBP-5 also interacts fibronectin and this binding is independent from IGF binding.27 This binding is mediated by IGFBP-5 C domain. This is consistent with the observation made in this study that IGFBP-5 C domain is critical for its internalization and nuclear localization. It will be of interest to determine whether the binding of IGFBP-5 with HS proteoglycans and/or fibronectin plays any role in IGFBP-5 internalization.

Another novel finding made in this study is that IGFBP-5 N-domain contains a functional transactivation domain. When fused to a heterogeneous DNA binding domain, the IGFBP-5 N-domain has strong transactivation activity. This activity is evolutionarily conserved and is not found in IGFBP-1 N-domain. Furthermore, this activity of IGFBP-5 is not affected by IGF-I binding nor by mutations in the ligand binding site. In addition, we were able to partially separate the transactivation domain from the IGF binding domain. These findings suggest that IGFBP-5 possesses transcription-regulating activity that is ligand-independent and may function as a transcriptional regulator or coregulator in VSMC nucleus in an IGF-independent fashion. It should be pointed out, however, that our results only prove that IGFBP-5 N-domain has the ability to activate the GAL-4 reporter gene when fused to GAL-4-DBD. More studies are needed to determine whether endogenous IGFBP-5 has similar activity and to identify its target genes.

Our findings suggest a novel pathway by which the secreted IGFBP-5 reenters cells and targets to the nuclear compartment where it may exert ligand-independent actions on gene expression. The concept that a secreted protein may enter the nucleus and exert nuclear actions is highly unusual but not without precedents. An increasing number of growth factors/cytokines with classical roles as extracellular signaling molecules and their cognate receptors are now found in the nucleus, where some of them are reported to act in an "intracrine" fashion.40,41 Recently, it was reported that secreted IGFBP-3, a closely related member of the IGFBP family, is internalized by prostate cancer cells through both transferrin receptor and caveolin-mediated endocytic pathways.36,42 It was reported that IGFBP-3 can interact with retinoid X receptor-{alpha}, and this interaction resulted in the modulation of the transcriptional activity of retinoid X receptor-{alpha}.43 Despite previous reports on the nuclear presence of IGFBPs, their nuclear functions and the underlying mode of actions have not been vigorously explored. Our findings that IGFBP-5 is present in the nucleus of VSMCs and possesses ligand-independent transactivation activity have raised the possibility that IGFBP-5 may function as a hormone binding protein as well as a transcriptional regulator depending on its location. More studies are needed to identify the target genes of IGFBP-5 and to elucidate the molecular events that trigger its cell entry and nuclear translocation in VSMCs.


*    Acknowledgments
 
This study was supported in part by NIH Grant RO1HL60679 and National Science Foundation Grant IBN 0110864 to C.D. We thank Dr David R. Clemmons, University of North Carolina at Chapel Hill, for providing the IGFBP-5 antibody and mutant IGFBP-5 cDNAs as PCR templates, Dr Shin-Ichiro Takahashi, University of Tokyo, for his help with the one-hybrid transcriptional assay, and Dr Antony Wood, University of Michigan, for critical reading of this manuscript.


*    Footnotes
 
Original received January 22, 2004; revision received February 18, 2004; accepted February 20, 2004.


*    References
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 up arrowMaterials and Methods
 up arrowResults
 up arrowDiscussion
 *References
 

  1. Clemmons DR. Growth factors and atherosclerosis. Diabetes Rev. 1997; 5: 353–364.
  2. Bayes-Genis A, Conover CA, Schwartz RS. The insulin-like growth factor axis: a review of atherosclerosis and restenosis. Circ Res. 2000; 86: 125–130.[Abstract/Free Full Text]
  3. Zhu B, Zhao G, Witte DP, Hui DY, Fagin JA. Targeted overexpression of IGF-I in smooth muscle cells of transgenic mice enhances neointimal formation through increased proliferation and cell migration after intraarterial injury. Endocrinology. 2001; 142: 3598–3606.[Abstract/Free Full Text]
  4. Zaina S, Pettersson L, Ahrén B, Brånén L, Hassan AB, Lindhol M, Mattsson R, Thyberg J, Nilsson J. Insulin-like growth factor II plays a central role in atherosclerosis in a mouse model. J Biol Chem. 2002; 277: 4505–4511.[Abstract/Free Full Text]
  5. Frystyk J, Ledet T, Moller N, Flyvbjerg A, Orskov H. Cardiovascular diseases and insulin-like growth factor I. Circulation. 2002; 106: 893–895.[Free Full Text]
  6. Clemmons DR. Use of mutagenesis to probe IGF-binding protein structure/function relationships. Endocr Rev. 2001; 22: 800–817.[Abstract/Free Full Text]
  7. Firth SM, Baxter RC. Cellular actions of the insulin-like growth factor binding proteins. Endocr Rev. 2002; 23: 824–854.[Abstract/Free Full Text]
  8. Duan C. Specifying the cellular responses to insulin-like growth factor (IGF) signals: role of IGF binding proteins. J Endocrinol. 2002; 175: 41–54.[Abstract]
  9. Gockerman A, Prevette T, Jones JI, Clemmons DR. Insulin-like growth factor binding protein-2 inhibits smooth muscle cell migration response to IGF-I. Endocrinology. 1995; 136: 4168–4173.[Abstract]
  10. Duan C, Clemmons DR. Differential expression and opposing biological effects of insulin-like growth factor binding protein-4 and -5 in vascular smooth muscle cells. J Biol Chem. 1998; 273: 16836–16842.[Abstract/Free Full Text]
  11. Zheng B, Duan C, Clemmons DR. The effect of extracellular matrix proteins on porcine smooth muscle cell insulin-like growth factor binding protein-5 synthesis and responsiveness to IGF-I. J Biol Chem. 1998; 273: 8994–9000.[Abstract/Free Full Text]
  12. Patel VA, Zhang QJ, Siddle K, Soos MA, Goddard M, Weissberg PL, Bennett MR. Defect in insulin-like growth factor-1 survival mechanism in atherosclerotic plaque-derived vascular smooth muscle cells is mediated by reduced surface binding and signaling. Circ Res. 2001; 88: 895–902.[Abstract/Free Full Text]
  13. Anwar A, Zahid AA, Scheidegger KJ, Brink M, Delafontaine P. Tumor necrosis factor-{alpha} regulates insulin-like growth factor-1 and insulin-like growth factor binding protein-3 expression in vascular smooth muscle. Circulation. 2002; 105: 1220–1225.[Abstract/Free Full Text]
  14. Hsieh T, Gordon R, Clemmons DR, Busby WH Jr, Duan C. Regulation of the mitogenic and chemotactic responses of vascular smooth muscle cells to insulin-like growth factor (IGF)-I by local IGF binding proteins. J Biol Chem. 2003; 278: 42886–42892.[Abstract/Free Full Text]
  15. Andress DL. Heparin modulates the binding of insulin-like growth factor (IGF) binding protein-5 to a membrane protein in osteoblastic cells. J Biol Chem. 1995; 270: 28289–28296.[Abstract/Free Full Text]
  16. Mohan S, Nakao Y, Honda Y, Landale E, Leser U, Dony C, Lang K, Baylink DJ. Studies on the mechanisms by which insulin-like growth factor (IGF) binding protein-4 (IGFBP-4) and IGFBP-5 modulate IGF actions in bone cells. J Biol Chem. 1995; 270: 20424–20431.[Abstract/Free Full Text]
  17. Abrass CK, Berfield AK, Andress DL. Heparin binding domain of insulin-like growth factor binding protein-5 stimulates mesangial cell migration. Am J Physiol. 1997; 273: F899–F906.[Medline] [Order article via Infotrieve]
  18. Miyakoshi N, Richman C, Kasukawa Y, Linkhart TA, Baylink DJ, Mohan S. Evidence that IGF-binding protein-5 functions as a growth factor. J Clin Invest. 2001; 107: 73–81.[CrossRef][Medline] [Order article via Infotrieve]
  19. McCaig C, Perks CM, Holly JMP. Intrinsic actions of IGFBP-3 and IGFBP-5 on Hs578T breast cancer epithelial cells: inhibition or accentuation of attachment and survival is dependent upon the presence of fibronectin. J Cell Sci. 2002; 115: 4293–4303.[Abstract/Free Full Text]
  20. Andress DL. Insulin-like growth factor-binding protein-5 (IGFBP-5) stimulates phosphorylation of the IGFBP-5 receptor. Am J Physiol. 1998; 274: E744–E750.[Medline] [Order article via Infotrieve]
  21. Radulescu R. From insulin, retinoblastoma protein and the insulin receptor to a new model on growth factor specificity: the nucleocrine pathway. J Endocrinol. 1995; 146: 365–368.[Medline] [Order article via Infotrieve]
  22. Schedlich LJ, Young TF, Firth SM, Baxter RC. Insulin-like growth factor-binding protein (IGFBP)-3 and IGFBP-5 share a common nuclear transport pathway in T47D human breast carcinoma cells. J Biol Chem. 1998; 273: 18347–18352.[Abstract/Free Full Text]
  23. Amaar YG, Thompson GR, Linkhart TA, Chen S-T, Baylink DJ, Mohan S. Insulin-like growth factor-binding protein 5 (IGFBP-5) interacts with a four and a half LIM protein 2 (FHL2). J Biol Chem. 2002; 277: 12053–12060.[Abstract/Free Full Text]
  24. Schedlich LJ, Le Page SL, Firth SM, Briggs LJ, Jans DA, Baxter RC. Nuclear import of insulin-like growth factor-binding protein-3 and-5 is mediated by the importin ß subunit. J Biol Chem. 2000; 275: 23462–23470.[Abstract/Free Full Text]
  25. Duan C, Bauchat JR, Hsieh T. Phosphatidylinositol 3-kinase is required for IGF-I-induced vascular smooth muscle cell proliferation and migration. Circ Res. 2000; 86: 15–23.[Abstract/Free Full Text]
  26. Li S, Liu X, Min L, Ascoli M. Mutations of the second extracellular loop of the human lutropin receptor emphasize the importance of receptor activation and de-emphasize the importance of receptor phosphorylation in agonist-induced internalization. J Biol Chem. 2001; 276: 7968–7973.[Abstract/Free Full Text]
  27. Xu Q, Yan B, Li S, Duan C. Fibronectin binds insulin-like growth factor binding protein (IGFBP)-5 and abolishes its ligand-dependent action on cell migration. J Biol Chem. 2004; 279: 4269–4277.[Abstract/Free Full Text]
  28. Kabuta T, Hakuno F, Asano T, Takahashi S-I. Insulin receptor substrate-3 functions as transcriptional activator in the nucleus. J Biol Chem. 2002; 277: 6846–6851.[Abstract/Free Full Text]
  29. Kalus W, Zweckstetter M, Renner C, Sanchez Y, Georgescu J, Grol M, Demuth D, Schumacher R, Dony C, Lang K, Holak TA. Structure of the IGF-binding domain of the insulin-like growth factor-binding protein-5 (IGFBP-5): implications for IGF and IGF-I receptor interactions. EMBO J. 1998; 17: 6558–6572.[CrossRef][Medline] [Order article via Infotrieve]
  30. Imai Y, Moralez A, Andag U, Clarke JB, Busby WH Jr, Clemmons DR. Substitutions for hydrophobic amino acids in the N-terminal domains of IGFBP-3 and -5 markedly reduce IGF-I binding and alter their biologic actions. J Biol Chem. 2000; 275: 18188–18194.[Abstract/Free Full Text]
  31. Li WL, Fawcett J, Widmer HR, Fielder PJ, Rabkin R, Keller GA. Nuclear transport of insulin-like growth factor-I and insulin-like growth factor binding protein-3 in opossum kidney cells. Endocrinology. 1997; 138: 1763–1766.[Abstract/Free Full Text]
  32. Shulga N, Roberts P, Gu Z, Spitz L, Tabb MM, Nomura M, Goldfarb DS. In vivo nuclear transport kinetics in Saccharomyces cerevisiae: a role for heat shock protein 70 during targeting and translocation. J Cell Biol. 1996; 135: 329–339.[Abstract/Free Full Text]
  33. Vagner S, Touriol C, Galy B, Audigier S, Gensac MC, Amalric F, Bayard F, Prats H, Prats AC. Translation of CUG- but not AUG-initiated forms of human fibroblast growth factor 2 is activated in transformed and stressed cells. J Cell Biol. 1996; 135: 1391–1402.[Abstract/Free Full Text]
  34. Fiaschi-Taesch NM, Stewart AF. Parathyroid hormone-related protein as an intracrine factor: trafficking mechanisms and functional consequences. Endocrinology. 2003; 144: 407–411.[Abstract/Free Full Text]
  35. Kopito RR. ER quality control: the cytoplasmic connection. Cell. 1997; 88: 427–430.[CrossRef][Medline] [Order article via Infotrieve]
  36. Lee K-W, Liu B, Ma L, Li H, Bang P, Koeffler HP, Cohen P. Cellular internalization of insulin-like growth factor binding protin-3 (IGFBP-3): distinct endocytic pathways facilitate reuptake and nuclear localization. J Biol Chem. 2004; 279: 469–476.[Abstract/Free Full Text]
  37. Belting M. Heparan sulfate proteoglycan as a plasma membrane carrier. Trends Biochem Sci. 2003; 28: 145–151.[CrossRef][Medline] [Order article via Infotrieve]
  38. Hsia E, Richardson TP, Nugent MA. Nuclear localization of basic fibroblast growth factor is mediated by heparan sulfate proteoglycans through protein kinase C signaling. J Cell Biochem. 2003; 88: 1214–1225.[CrossRef][Medline] [Order article via Infotrieve]
  39. Richardson TP, Trinkaus-Randall V, Nugent MA. Regulation of heparan sulfate proteoglycan nuclear localization by fibronectin. J Cell Sci. 2001; 114: 1613–1623.[Abstract]
  40. Jans DA, Hassan G. Nuclear targeting by growth factors, cytokines, and their receptors: a role in signaling? Bioessays. 1998; 20: 400–411.[CrossRef][Medline] [Order article via Infotrieve]
  41. Olsnes S, Klingenberg O, Widocha A. Transport of exogenous growth factors and cytokines to the cytosol and to the nucleus. Physiol Rev. 2003; 83: 163–182.[Abstract/Free Full Text]
  42. Singh B, Charkowicz D, Mascarenhas D. IGF-independent effects mediated by a C-terminal metal-binding domain of insulin-like growth factor binding protein-3. J Biol Chem. 2004; 279: 477–487.[Abstract/Free Full Text]
  43. Liu B, Lee HY, Weinzimer SA, Powell DR, Clifford JL, Kurie JM, Cohen P. Direct functional interactions between insulin-like growth factor-binding protein-3 and retinoid X receptor-{alpha} regulate transcriptional signaling and apoptosis. J Biol Chem. 2000; 275: 33607–33613.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 J Mol EndocrinolHome page
D J Flint, M Boutinaud, C B A Whitelaw, G J Allan, and A F Kolb
Prolactin inhibits cell loss and decreases matrix metalloproteinase expression in the involuting mouse mammary gland but fails to prevent cell loss in the mammary glands of mice expressing IGFBP-5 as a mammary transgene.
J. Mol. Endocrinol., June 1, 2006; 36(3): 435 - 448.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
P. Yin, Q. Xu, and C. Duan
Paradoxical Actions of Endogenous and Exogenous Insulin-like Growth Factor-binding Protein-5 Revealed by RNA Interference Analysis
J. Biol. Chem., July 30, 2004; 279(31): 32660 - 32666.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
94/5/e46    most recent
01.RES.0000124761.62846.DFv1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Xu, Q.
Right arrow Articles by Duan, C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Xu, Q.
Right arrow Articles by Duan, C.
Related Collections
Right arrow Mechanism of atherosclerosis/growth factors
Right arrow Gene regulation
Right arrow Growth factors/cytokines
Right arrow Smooth muscle proliferation and differentiation

Circulation Research Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search
Copyright © 2004 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited.