| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
- and basic FGF-dependent pathway in muscle cells
* Department of Internal Medicine, General Internal Medicine,
Department of Neurology,
Department of Plastic Surgery, and
Department of Pharmacology, Medical University of Innsbruck, Innsbruck, Austria;
|| The Laboratory of Microvascular Research and Vascular Tissue Engineering, New York University Medical Center, New York, New York, USA; and
¶ Division of Cardiovascular Research and
Vascular Medicine, St. Elizabeth's Medical Center, Tufts University School of Medicine, Boston, Massachusetts, USA
1Correspondence: Department of Internal Medicine, University of Innsbruck, Anichstr. 35, 6020 Innsbruck, Austria. E-mail: rudolf.kirchmair{at}uibk.ac.at
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, since inhibition of this factor using si-RNA inhibited up-regulation of secretoneurin. Induction of secretoneurin by hypoxia was dependent on basic fibroblast growth factor in vivo and in vitro, and inhibition of this regulation by heparinase suggests an involvement of low-affinity basic fibroblast growth factor binding sites. In summary, our data show that the angiogenic cytokine secretoneurin is up-regulated by hypoxia in muscle cells by hypoxia-inducible factor-1- and basic fibroblast growth factor-dependent mechanisms.—Egger, M., Schgoer, W., Beer, A. G. E., Jeschke, J., Leierer, J., Theurl, M., Frauscher, S., Tepper, O. M., Niederwanger, A., Ritsch, A., Kearney, M., Wanschitz, J., Gurtner, G. C., Fischer-Colbrie, R., Weiss, G., Piza-Katzer, H., Losordo, D. W., Patsch, J. R., Schratzberger, P., Kirchmair, R. Hypoxia up-regulates the angiogenic cytokine secretoneurin via an HIF-1- and basic FGF-dependent pathway in muscle cells.
Key Words: vascular biology • neuropeptides • ischemia
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2
3
4)
. The prototypical angiogenic cytokine increased by hypoxia is vascular endothelial growth factor (VEGF) and its receptor flt-1; however, other angiogenic factors like angiopoietin-2, placental growth factor, and platelet-derived growth factor B (PDGF-B) are also regulated by hypoxia (5)
. Another angiogenic cytokine, basic fibroblast growth factor (bFGF), exerts enhanced proliferative activity on endothelial cells (ECs) during hypoxia due to an increase of low-affinity heparan sulfate bFGF binding sites (6)
. Hypoxia also induces a hypoxia-inducible factor-1 (HIF-1) -dependent bFGF autocrine loop that drives angiogenesis in human endothelial cells (7)
.
Sensing of oxygen levels in the cell is mediated by HIF-1 (8)
. Under normoxic conditions, HIF-1 is degraded after prolin-hydroxylation and association with the von-Hippel Lindau protein by the proteasome. In hypoxia, HIF-1 is not hydroxylated and degraded; it forms the HIF transcription complex with cofactors, is translocated to the nucleus, and binds to a nucleotide sequence called hypoxia-response element (HRE) within the promoter region of hypoxia-regulated genes.
Secretoneurin (SN) is a neuropeptide derived from prosecretoneurin (pro-SN, also called secretogranin-II), a protein of neuroendocrine storage vesicles, by proteolytic action of prohormone convertases (PCs) (9
10
11)
. We recently showed that SN induces angiogenesis and postnatal vasculogenesis in vivo by inducing vessel growth and incorporation of endothelial progenitor cells (EPCs) in a cornea neovascularization assay. In vitro SN induced capillary tube formation in a Matrigel assay induced proliferation and directed migration of ECs and EPCs, and inhibited apoptosis of these cells by stimulation of mitogen-activated protein kinases (MAPK) and Akt (12
, 13)
.
In this work we tested the hypothesis that SN, like other angiogenic and vasculogenic cytokines, may be regulated by hypoxia.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Animal models
All protocols were approved by the St. Elizabeth's Medical Center Institutional Animal Care and Use Committee and the New York University Medical Center Animal Care and Use Committee. Anesthesia was performed with 2,2,2-tribromoethanol (880 mmol/kg body weight i.p.; Sigma-Aldrich).
Hind limb ischemia, laser Doppler perfusion imaging, and determination of capillary density
Unilateral hind limb ischemia was created in 2-month-old C57BL/6J mice as described previously (14)
. Seven days after induction of ischemia, mice (n=4) were subjected to laser Doppler perfusion imaging (14)
(LDPI, Moor Instruments, Devon, UK) to verify ischemia, then sacrificed, and hind limbs were used for immunohistochemistry.
For application of SN antibody, mice (n=10) were subjected to hind limb ischemia operation and LDPI was performed on days –1, 0, 7, 14, and 21. Perfusion is expressed as the quotient of LDPI values of the ischemic, operated-on side compared with the side that was not operated on. One hundred micrograms of IgG-purified SN antibody (9)
, shown to inhibit angiogenesis in vitro (12)
or rabbit IgG (control group), was injected intraperitoneally (i.p.) on days 0, 2, 4, 6, 11, and 16. After day 21, mice were sacrificed and sections of hind limb muscles were subjected to isolectin B4 staining to determine capillary density as described (15)
. Capillaries of control and ischemic side were counted per mm2 in three different sections of each muscle, and capillary density is expressed as the ratio of capillaries of the operated, ischemic side relative to the control, not operated-on side.
To determine the role of bFGF in the up-regulation of SN in the hind limb ischemia model, a neutralizing bFGF antibody (Sigma) was injected i.p. at a dose of 100 µg on days 0, 2, 4, and 6 after induction of ischemia. Controls received rabbit IgG at the same dose. Mice were sacrificed on day 7 after induction of ischemia and muscles were stained for SN by immunohistochemistry.
Soft tissue ischemia model
Soft tissue ischemia model was performed as described (16)
. Briefly, lateral skin incisions were performed on the dorsal back of mice (n=4). The overlying skin was undermined and a silicon sheet was inserted to separate the skin from the underlying tissue bed. After sacrifice, tissue was fixed in methanol and sections from normoxic and hypoxic areas were used for immunohistochemistry.
Secretoneurin antibody, radioimmunoassay, immunohistochemistry, immunofluorescence, and NADH staining
SN antibody was generated and radioimmunoassay (RIA) was performed as described in detail before (9)
. SN immunohistochemistry was performed by standard procedures. Briefly, sections were deparaffinized, washed with PBS, treated with H2O2 (3%) for 30 min, washed again in PBS, blocked with 10% normal goat serum (NGS) for 30 min, and incubated with the primary SN antibody overnight (dilution 1:1000). Preabsorption of the antibody with the peptide was used as negative control. A biotinylated goat immunoglobulin antibody against rabbit (Signet, Dedham, MA, USA) was used as secondary antibody. Staining was performed using 3-amino-9-ethylcarbazol (AEC: L6 myoblast cell culture) or 3,3'diaminobenzidine: hind limb ischemia and soft tissue ischemia model) as substrate.
For immunofluorescence, sections were washed for 1 h at room temperature in Tris-buffered saline (TBS; 50 mM Tris pH 7.5+0.9% NaCl) containing 0.3% Triton X-100. For blocking, the sections were preincubated for 1 h at room temperature with 20% NGS in TBS containing 0.3% Triton X-100 in disposable immunostaining chambers (Shandon Coverplate, Cat. No. 7211013, Thermo Electron Corp., Woburn, MA, USA) and subsequently coincubated overnight at 4°C with rabbit secretoneurin antiserum (9)
at a dilution of 1:1000 and sheep antineuropeptide Y antibody at a dilution of 1:5000 in TBS containing 2% NGS and 0.3% Triton X-100. After three washes with TBS, the sections were incubated with secondary antibodies (Cy3-conjugated AffiniPure goat anti-rabbit IgG, 111–165-006, and Cy2 AffiniPure donkey anti-sheep IgG, 713–225-003; Jackson ImmunoResearch Labs, West Grove, PA, USA), each diluted 1:400 in TBS containing 2% NGS and 0.3% Triton X-100 for 24 h at 4°C. Stained sections were washed three times with TBS, and before being coverslipped were mounted with 0.5% gelatin containing 0.05% chrom(III)sulfate. Sections were visualized with a Zeiss Axioplan optical microscope (Jena, Germany) and pictures were taken with a Zeiss AxioCam HR.
NADH staining of muscles was performed by standard procedures. The histochemical reaction NADH-Tr (nicotinamide adenin dinucleotide tetrazolium reductase) visualizes the mitochondrial oxidative enzyme complex I. Dark stained fibers represent type I fibers, and fibers with lower oxidative activity are type II fibers.
Cell culture
L6 rat myoblasts from American Type Culture Collection (ATCC, Manassas, VA, USA) were cultured in minimum essential medium (Sigma), human umbilical vein endothelial cells (HUVEC, from ATCC) in EBM-2 (Clonetics, San Diego, CA, USA), AtT20 cells (from ATCC) in DMEM (1x) high glucose medium (PAA Laboratories, Pasching, Austria), and human vascular smooth muscle cells (SMC, from PromoCell, Heidelberg, Germany) in SMC growth medium 2 (PromoCell). 10% heat-inactivated fetal bovine serum (FBS), 100 U/ml penicillin, 100 µg/ml streptomycin, and 4 mM L-glutamine were added to each medium. In some experiments cells were treated with bFGF (10 ng/ml), VEGF (50 ng/ml), PDGF-BB (50 ng/ml), bFGF antibody (used at a dilution of 1:1000), or heparinase I (50 mU/ml).
Western blot
Phospho-p44/42 MAP kinase (Thr202/Tyr204) antibody and p44/42 MAP kinase antibody were purchased from Cell Signaling (Beverly, MA, USA). Cells were lysed and lysates were processed for Western blot as suggested by the manufacturer.
Induction of hypoxia
Cultured cells were kept under hypoxic condition as described (17)
. Briefly, normal humidified tissue culture incubators with 5% CO2 were used for the normoxic cultures. For decreased oxygen cultures, plates were inserted into gas-tight modular incubator chambers (Billups-Rothenberg, Del Mar, CA, USA), which were flushed with a custom gas mixture containing 5% CO2 and 95% N2 for 15 min twice a day. pO2 was determined in the culture medium by a gas analyzer (Bayer/Ciba-Corning 278, Diamond Diagnostics, Holliston, MA, USA). Six hours after CO2/N2 gas incubation, pO2 was 35 mm Hg, which persisted for 24 h.
Polymerase chain reaction (PCR)
PCR was performed to detect mRNA expression of respective proteins and RNA of cells was extracted (Qiagen, RNeasy Mini Kit, Vienna, Austria). To eliminate genomic DNA, samples were digested with DNase I (desoxyribonuclease I, Amplification Grade, Invitrogen, Vienna, Austria), then subjected to reverse transcription (Superscript TM First-Strand Synthesis System for RT-PCR, Invitrogen). cDNA was finally used as template for PCR reactions (Advantage-GC cDNA Polymerase Mix, Clontech, Heidelberg, Germany) under the following protocol: 94°C for 2 min, 35 cycles of 94°C 30 s, and 64°C 3 min, followed by 8 min of 64°C. PCR for actin was used as loading control.
Primers used
Human primers for HUVEC and SMC
Pro-SN (550 bp) forward: 5'-AGAAGCCGAATGGATCAGTGGAACC
backward: 5'-TCAGTCTCCTGAGAGCTGCCTTGAT
PC1 (830 bp) forward: 5'-ATGACTTTGAGCCCAGAGCCCTGAA
backward: 5'-GAATGTCCACCAGAGCTTGAAGCAG
Actin (734 bp) forward: 5'-CGCGAGAAGATGACCCAGATCATGT
backward: 5'-TGCTTGCTGATCCACATCTGCTGGA
VEGF (230 bp): forward: 5'-TGTCTATCAGCGCAGCTACTGCCAT
backward: 5'-GGAAGCTCATCTCTCCTATGTGCTG
Rat primers for L6 myoblast
Pro-SN (673bp) forward: 5'-AGCATGCTTGGAGCCTTCCACACAA
backward: 5'-CAAGCTCTTGGAACACAGACTCCAG
PACE4 (434bp) forward: 5'-CGAGCTGTCAGAAGTGTGTGGATGA
backward: 5'-TACCATCTCGCAGAAGGTCTCATCG
PC7 (442 bp) forward: 5'-ACCTGCATGATGACTTCACTCTGCC
backward: 5'-CAGATCTGTCCATCCTTCAGGTCTG
Furin (526 bp) forward: 5'-TCACTCTCGTCCTATATGGCACAGC
backward: 5'-AAGCCTGAACACCGATGCAGGAAGA
bFGF receptor-1 (775bp) forward: 5'-TACGTGCTTGGCGGGTAACTCTATC
backward: 5'-GAAGACAGCTGTTCCTCAGGGTTGT
HIF-1 (585 bp) forward: 5'-GGATTCCAGCAGACCCAGTTACAGA
backward: 5'-CAGCTGTGGTAATCCACTCTCATCC
Heparan sulfate 2-O-sulfotransferase (HS2ST; 368 bp)
forward: 5'-TGATTACAGACCGGGACTAAGGAGG
backward: 5'-AGCTTCGCGATGGTCTGTTTAGTGG
1,4-N-acetylglucosamine transferase (GlcNAcT-I; 615 bp)
forward: 5'-CTGATGGAGGTGACAATGCAGACAC
backward: 5'-CTAGACAAGTGAGCTGAGAAGCTGC
Actin (825 bp) forward: 5'-TTTCTACAATGAGCTGCGTGTGGCC
backward: 5'-TCATCGTACTCCTGCTTGCTGATCC
VEGF (603 bp) forward: 5'-ATGAACTTTCTGCTCTCTTGGGTGC
backward: 5'-ACACAGGACGGCTTGAAGATATACT
Mouse primers for AtT20 cells
Pro-SN (781 bp) forward: 5'-ACCAACTCTCAGAGGATGCCTCCAA
backward: 5'-TCTGCCTGCTCTTGGTTGAGGTACT
PC1 (667 bp) forward: 5'-TGCTGTTGGAACCAGCACTGTACTG
backward: 5'-TAGGATGTCCATGAGAGCTTGCAGC
PC2 (579 bp) forward: 5'-TGAGGTTCATCAGTGGCGACGGAAT
backward: 5'-TGCTTCTTGGACATGGCCAGCTTAG
PACE4 (744 bp) forward: 5'-GTGCTTGAACTGTGTCCACTTCAGC
backward: 5'-TGGATGAGAGCTTCCGTTCGCAGA
Actin (852 bp) forward: 5'-TGGCATTGTTACCAACTGGGACGAC
backward: 5'-ATCCACATCTGCTGGAAGGTGGACA.
Omitting reverse transcriptase enzyme or cDNA template was used as negative control. Experiments showing regulation of pro-SN mRNA were confirmed by RT-PCR and quantifications were performed as described (18)
. Quantification of other mRNAs was performed by densitometry of semiquantitative PCR.
Transfection with si-RNA
Myoblasts were transfected with si-RNA of HIF-1 as described (19)
. Briefly, myoblasts were cultured in 6-well plates and transfected with HIF-1 si-RNA (5'-AGAGGUGGAUAUGUCUGGG-3') or scrambled si-RNA (5'-AGGAUGUGACGGAUUGUGGTT-3'), used as negative control (MWG Biotech AG, Ebersberg, Germany), in the presence of Oligofectamine 228 (Invitrogen) for 4 h as suggested by the manufacturer. Afterward, 10% serum was added and cells were cultured for 36 h under normoxic or hypoxic conditions.
To verify inhibition of HIF-1, PCR and Western blot (using an antibody against HIF-1 from Novus Biologicals, Littleton, CO, USA) were performed as suggested by the manufacturer.
A fluorescently labeled nontarget control for si-RNA was purchased from Cell Signaling and used as a negative control in addition to scrambled si-RNA. Fluorescence microscopy showed high transfection efficiency for nontarget control si-RNA (data not shown). Cells transfected with nontarget control si-RNA were exposed to hypoxia, and analysis of pro-SN mRNA by real-time RT-PCR showed a comparable increase of pro-SN by hypoxia in cells transfected with nontarget control si-RNA, with scrambled si-RNA, or cells without si-RNA transfection (data not shown).
Statistics
ANOVA was used for statistical analysis and a P value of <0.05 was considered to denote statistical significance.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
A). SN was not uniformly distributed along all muscle fibers but showed distinct immunoreactivity of mostly smaller muscle fibers between other muscle cells negative for SN expression (Fig. 1B
). Negative control obtained by preabsorption of the antibody by SN peptide (Fig. 1B
, bottom panel) did not show specific staining.
|
In the ischemic skin, SN was up-regulated in distinct muscle fibers of the subcutaneous layer compared with nonischemic areas (Fig. 1C
). Ischemic fibroblasts and adipocytes of subcutaneous tissue did not show increased SN immunoreactivity compared with normoxic areas.
NPY represents another classical neuropeptide, also demonstrated to induce angiogenesis and to be up-regulated by hypoxia. We therefore performed double immunofluorescence of ischemic hind limb muscles for NPY and SN. We found that some ischemic muscle cells express both neuropeptides (Fig. 1D, F
, arrows) whereas NPY immunostaining partially localized to structures outside of muscle fibers, negative for SN and possibly representing capillaries (Fig. 1D, F
, arrowheads). In contrast to NPY, other members of the chromogranin/secretogranin family besides pro-SN, chromogranin A and B were not detectable in ischemic hind limb muscles.
RIA of tissue extracts of ischemic and normal hind limb muscles showed an increase of SN immunoreactivity in ischemic muscles (control leg: 7.5±1.9 fmol/100 mg wet weight; ischemic leg: 15.1±2.2 fmol/100 mg wet weight; P<0.05; n=4).
NADH staining revealed that small SN-positive cells were NADH-positive type I muscle fibers (Fig. 1E
, arrows), whereas large muscle cells that showed weak SN staining were NADH-negative type II fibers (Fig. 1E
, asterisks).
Inhibition of secretoneurin impairs angiogenesis in the hind limb ischemia model
To evaluate the role of SN in physiological angiogenesis, we injected an SN neutralizing antibody (9
, 12)
into mice after hind limb ischemia operation. We observed that the SN antibody impairs the angiogenic response determined by reduced blood perfusion in LDPI (Fig. 2
, top panel), as indicated by reduced recovery of LDPI ratio of the operated vs. the control leg in the SN antibody group on days 7, 14, and 21 (LDPI ratio controls 0.99±0.06 on day 21, indicating complete recovery of perfusion in the operated leg; LDPI ratio of SN antibody group 0.62±0.06 on day 21, P<0.05 vs. control group). Capillary density (Fig. 2
, bottom panel) was decreased in SN antibody-injected mice compared with control, IgG-injected animals (ratio of capillaries of operated vs. not operated-on side was 0.94±0.08 in control, and 0.51±0.13 in SN antibody-injected mice; P<0.05 control vs. SN antibody).
|
In vitro experiments
L6 myoblasts express pro-secretoneurin, Furin, PACE4, and PC7
To evaluate whether a non-neuroendocrine cell expresses pro-SN and its processing enzymes, we used PCR to detect mRNA of respective proteins and were able to show that L6 myoblasts express mRNAs of pro-SN, Furin, PACE4, and PC7 (Fig. 3
). Adrenal medulla was used as positive control. Other processing enzymes like PC1, PC2, and PC5 were not expressed. To quantify pro-SN expression between adrenal medulla and L6 myoblast cells, we performed real-time RT-PCR and found an 
100-fold higher expression of pro-SN in adrenal medulla (relative pro-SN mRNA: adrenal medulla: 1, L6 normoxic: 0.011±0.001; n=3). We also investigated the expression of other neuroendocrine proteins, chromogranin A and B, in these cells but could not detect these peptides in L6 myoblasts in normoxic or hypoxic conditions (data not shown).
|
Secretoneurin and PACE4 are up-regulated by hypoxia in L6 myoblasts
Exposure of myoblasts to hypoxia for 36 h led to a pronounced up-regulation of pro-SN mRNA and, to a lesser extent, to an up-regulation of the processing enzyme PACE-4 (3.2±0.9-fold, P<0.05, n=3) (Fig. 4
A, semiquantitative PCR, left upper graph; quantification of PACE4: lower panel, middle graph). We also performed real-time RT-PCR for pro-SN mRNA after different time points of hypoxia (Fig. 4A
, lower panel, left graph). Compared with normoxic cells, pro-SN mRNA was –1 ± 0.2% after 3 h of hypoxia, –5 ± 2% after 6 h of hypoxia, +30 ± 10% after 12 h of hypoxia, +330 ± 90% after 24 h of hypoxia and + 820 ± 190% after 36 h of hypoxia (n=3; P<0.05 after 24 h, P<0.01 after 36 h, p=n.s. after 3, 6, and 12 h of hypoxia). Furin and PC7 showed no regulation by hypoxia.
|
To evaluate whether secretoneurin is also up-regulated at the protein level, L6 myoblast cells were subjected to hypoxia for 36 h and stained for SN by immunohistochemistry. Cells exposed to hypoxia showed increased SN immunoreactivity compared with normoxic cells (Fig. 4B
). RIA revealed that hypoxia induced a significant increase of SN in the medium of hypoxic myoblasts (48.2±3.5 fmol SN/106 cells) compared with normoxic cells (28.2±2.4 fmol SN/106 cells; P<0.05, n=3) (Fig. 4B
, lower panel, right graph).
Regulation of pro-secretoneurin and prohormone convertases in other cell types
We also investigated the regulation of pro-SN and PCs by hypoxia in vascular cells (HUVEC, vascular SMC) and in an endocrine tumor cell line (AtT20 cells) known to produce pro-SN and processing enzymes. We observed tissue-specific regulation in these cells and could demonstrate up-regulation of PC1 (3.0±0.3-fold, P<0.05, n=3) but not of pro-SN in HUVEC (Fig. 5
A). In vascular SMC, hypoxia induced a strong up-regulation of VEGF, used as positive control, but neither pro-SN nor PC1 were increased (Fig. 5B
). In AtT20 cells, hypoxia increased mRNA of PACE4 (7.6±1.8-fold, P<0.05, n=4) but not of pro-SN, PC1, or PC2 (Fig. 5C
). SN protein levels as determined by RIA were not increased in HUVEC, SMC, or AtT20 cells by hypoxia (Figs. 5A-C
), further confirming tissue-specific regulation of SN and indicating that up-regulation of processing enzymes alone does not increase SN levels. Levels of SN as determined by RIA were 50-fold higher in AtT20 cells compared with non-neuroendocrine cells.
|
Mechanism of up-regulation of pro-secretoneurin in myoblasts by hypoxia
We did not observe up-regulation of pro-SN in myoblasts when cells were cultured without serum and exposed to hypoxia (Fig. 6
A, left panel, pro-SN, two left lanes). In contrast to pro-SN, regulation of VEGF by hypoxia was not dependent on serum as shown by up-regulation of VEGF in L6 myoblasts cultured without serum (Fig. 6A
, left panel, VEGF). The lack of up-regulation of pro-SN by hypoxia without serum indicates that a factor in serum mediates an increase of pro-SN by hypoxia.
|
We demonstrate that bFGF mediates pro-SN increase by hypoxia as shown by the increase of pro-SN mRNA in hypoxic myoblasts cultured without serum and the addition of bFGF (Fig. 6A
, left panel, pro-SN, two right lanes) (8.1±1.3-fold increase by real-time RT-PCR compared with normoxic cells, P<0.01, n=3). Incubation of serum with a bFGF-neutralizing antibody inhibited hypoxia-induced up-regulation of pro-SN: as determined by real-time PCR, pro-SN was increased 7.1 ± 1.3-fold by hypoxia when cells were cultured with 10% serum and 1.9 ± 1.1-fold by hypoxia when cells were cultured with 10% serum incubated with a bFGF antibody (P<0.01 hypoxia vs. control and hypoxia+bFGF antibody, n=3; Fig. 6A
, middle graph). Treatment of myoblasts with other angiogenic factors like VEGF or PDGF-BB did not increase pro-SN mRNA in hypoxic cells cultured without serum (data not shown). To extend these studies to an in vivo model of ischemia, a neutralizing bFGF antibody was injected systemically into mice after hind limb ischemia operation. We found that SN immunoreactivity in ischemic muscles was largely reduced by the bFGF antibody compared with a control antibody (Fig. 6A
, right panels).
PCR analysis revealed that L6 myoblasts express FGF receptor 1 (Fig. 6B
, upper panel), and stimulation of MAPK phosphorylation by bFGF in L6 myoblasts (especially p42) indicates that this receptor is functionally active in these cells (Fig. 6B
, lower panel).
To evaluate whether an increase of pro-SN in myoblasts is mediated by HIF-1, myoblasts were transfected with si-RNA against HIF-1; blockade of this transcription factor inhibits up-regulation of pro-SN by hypoxia to a large extent, whereas scrambled si-RNA had no effect (8.5±2.2-fold increase by hypoxia in cells transfected with scrambled si-RNA, 1.5±0.09-fold increase by hypoxia in cells transfected with HIF-1 si-RNA, P<0.05, n=4) (Fig. 6C
). Western blot revealed that HIF-1 si-RNA inhibited an increase of HIF-1 protein by hypoxia by 92% (as determined by densitometry normalized to actin, used as loading control; Fig. 6D
).
As an increase in heparan sulfate containing low-affinity bFGF binding sites mediates increased responsiveness of ECs to bFGF in hypoxia by HIF-1-dependent up-regulation of heparan sulfate synthesizing enzymes (6)
, we investigated the effect of hypoxia on regulation of the two key enzymes of heparan sulfate synthesis [i.e., heparan sulfate 2-O-sulfotransferase (HS2ST) and 1,4-N-acetylglucosamine transferase (GlcNAcT-I)] in L6 myoblasts as well as the effect of heparinase treatment on SN regulation by hypoxia. We observed up-regulation of mRNAs of both enzymes by hypoxia in L6 cells (Fig. 6E
, left panel; 3.4±0.7-fold increase for GlcNAcT-I and 4.0±0.9-fold increase for HS2ST; P<0.05, n=3). Furthermore, we found that heparinase treatment completely inhibited pro-SN increase by hypoxia in myoblasts (real-time PCR: 11.4±0.8-fold increase by hypoxia, 1.2±0.07-fold increase by hypoxia and heparinase treatment, P<0.01, n=3; Fig. 6E
, right graph) indicating the necessity of heparan sulfate side chains for regulation of SN by hypoxia in L6 cells.
To investigate whether bFGF is necessary for HIF-1 up-regulation by hypoxia as described for endothelial cells (7)
, L6 cells were subjected to hypoxia in the absence or presence of a bFGF antibody or heparinase (Fig. 6F
). Both inhibited HIF-1 up-regulation by hypoxia to a large extent, indicating the necessity of bFGF and heparan sulfate containing low-affinity bFGF binding sites for HIF-1 up-regulation by hypoxia.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
21
22
23)
. Furthermore, factors from the nervous system like nerve growth factor or neuropeptides like substance P, neuropeptide Y (NPY), or SN have been shown to induce angiogenesis and postnatal vasculogenesis (12
, 13
, 24
25
26)
. The sensitivity of nerve cells to hypoxia might be the reason for this close interaction because angiogenesis counteracts hypoxia by producing new blood vessels.
Hypoxic tissues express angiogenic factors in a cell-specific manner mediated by the action of the transcription factor HIF-1. Whereas VEGF is up-regulated in most tissues by hypoxia, angiopoietin-2 and PDGF-B are expressed especially in endothelial cells (5)
. We show in this study in two in vivo models of ischemia (hind limb ischemia model and a novel model of soft tissue ischemia (16)
) that the angiogenic cytokine SN is up-regulated in ischemic skeletal muscle cells. Inhibition of SN by a neutralizing antibody impaired the angiogenic response in the hind limb ischemia model, indicating a role for SN in physiological angiogenesis in this in vivo model. SN and processing enzymes are expressed in vitro in the rat muscle cell line L6 and SN is up-regulated in L6 by prolonged hypoxia. Although SN is usually confined to storage vesicles of neuroendocrine cells, we were able to detect SN in non-neuroendocrine L6 cells at amounts 50- to 100-fold lower under basal conditions than in neuroendocrine cells like adrenal medulla or AtT20 cells, as determined by RIA and real-time RT-PCR. It has already been reported that non-neuroendocrine cells like tumor cells express SN (27)
, and so it might be conceivable that non-neuroendocrine cells express peptides like SN, usually confined to neuroendocrine cells under special circumstances like tumor growth or hypoxia, where we were able to demonstrate substantial up-regulation of this peptide in myoblasts.
In contrast to chromogranin A and B, which were not expressed in muscle cells, we could also detect immunostaining for another neuropeptide, NPY, in muscle cells of ischemic hind limbs positive for SN and in structures probably representing capillaries. Previous studies have demonstrated regulation of neuropeptides by hypoxia: increased expression of SN was observed in neuronal cells of the brain after transient forebrain ischemia and in a neuroblastoma cell line exposed to hypoxia (28
, 29)
. In the ischemic hind limb, the Y2 receptor of the angiogenic neuropeptide NPY as well as the processing enzyme dipeptidyl-peptidase IV, important mediators of the angiogenic action of NPY, were up-regulated (30)
, underlining the regulation of NPY-induced angiogenesis by ischemia. In hypoxic neuroblastoma cells, NPY was not changed, suggesting tissue-specific regulation by hypoxia for this neuropeptide as well (29)
. In neuroblastoma cells, however, two other peptides—galanin and adrenomedullin—were increased by hypoxia (29)
. Adrenomedullin also induces angiogenesis (31)
and therefore serves as another example of an angiogenic neuropeptide regulated by hypoxia.
We also found regulation of prohormone convertases by hypoxia. These enzymes cleave large proteins to the biological active hormones or neuropeptides and mediate the generation of SN from pro-SN (32
, 33)
. We demonstrate that one of these enzymes, PACE4, is regulated in L6 myoblasts and in the pituitary tumor cell line AtT20; another one, PC1, is increased by hypoxia in ECs. In contrast to L6 myoblasts, SN itself was not up-regulated in AtT20 cells or ECs. In vascular SMCs, neither pro-SN nor prohormone convertases were increased by hypoxia in contrast to VEGF. These data underline the tissue-specific regulation of SN by hypoxia in skeletal muscle cells. While tissue-specific regulation of angiogenic cytokines like angiopoietin-2 and PDGF-B (5)
is poorly understood, it has been reported that kidney-specific expression of erythropoietin is mediated by specific sequences of regulatory DNA elements and tissue-specific transcription factors (34)
. The factors that mediate SN expression specifically in ischemic skeletal myocytes remain to be determined.
Regulation of the angiogenic cytokine SN in the ischemic skeletal muscle might indicate a possible role for SN in clinical situations of muscle hypoxia such as in peripheral artery disease. Preliminary data of patients with critical limb ischemia (unpublished data and ref. 13
) showed increased levels of SN in the plasma of affected patients compared with controls. Furthermore, our data of impaired angiogenesis in the hind limb ischemia model after injection of a SN-neutralizing antibody indeed indicate a role for SN in physiological/pathophysiological angiogenesis. Under normoxic conditions, however, myocytes express low levels of SN (compared with neuroendocrine cells or hypoxic myoblasts); it therefore is unlikely that this peptide plays a role in vascular homeostasis in normoxic areas.
Increased production of angiogenic cytokines by hypoxia usually is regulated by HIF-1 (3)
. Using si-RNA against HIF-1 (19)
, we were able to demonstrate that an increase of SN by hypoxia is HIF-1 dependent. However, the fact that exposure of L6 myoblasts to hypoxia without serum failed to increase pro-SN mRNA (in contrast to VEGF) indicates that additional factors are required for pro-SN up-regulation. We hypothesized that bFGF might mediate an increase of SN by hypoxia, because bFGF is known to regulate pro-SN (35)
and to exert enhanced action in hypoxia by a HIF-1-dependent increase of low-affinity heparan sulfate bFGF binding sites (6)
. Several observations indicate that SN regulation by hypoxia in myoblasts is indeed mediated by bFGF. First, we could demonstrate that addition of bFGF to muscle cells without serum restores up-regulation of pro-SN by hypoxia. Treatment with a neutralizing bFGF antibody abolished an increase in SN by hypoxia in vitro and in vivo. Furthermore, key enzymes for heparan sulfate side chain synthesis, GlcNAcT-I and HS2ST (6)
, were also increased by hypoxia in myoblasts; heparinase treatment abolished the increase in pro-SN by hypoxia, indicating that low-affinity heparan sulfate bFGF binding sites might mediate this effect. In contrast, we did not observe an increase in mRNA of the high-affinity bFGF receptor 1 in these cells by hypoxia (data not shown). Finally, other angiogenic factors like VEGF, (the proto-typical angiogenic cytokine increased by hypoxia) as well as PDGF-BB, did not mediate pro-SN regulation by hypoxia in muscle cells. Thus, our data indicate that during hypoxia not only ECs (6)
but also muscle cells show increased sensitivity to bFGF, leading to up-regulation of the angiogenic factor SN. We also observed that bFGF and intact heparan sulfate side chains are necessary for up-regulation of HIF-1 by hypoxia in L6 cells as shown by impaired up-regulation of HIF when hypoxic cells were incubated with a bFGF antibody or heparinase. In HUVEC, inhibition of bFGF was also shown to inhibit HIF-1 up-regulation by hypoxia by blocking an autocrine amplification loop between HIF-1 and bFGF (7)
.
In summary, our results demonstrate that the angiogenic cytokine SN is regulated by hypoxia in skeletal muscle cells via HIF-1 and bFGF-dependent mechanisms.
|
ACKNOWLEDGMENTS |
|---|
Received for publication September 28, 2006. Accepted for publication March 26, 2007.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
