Donate Help Contact The AHA Sign In Home
American Heart Association
Circulation
Search: search_blue_button Advanced Search
« Previous Article | Table of Contents | Next Article »
Circulation. 1996;93:1095-1106

This Article
Right arrow Abstract Freely available
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 Murry, C. E.
Right arrow Articles by Schwartz, S. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Murry, C. E.
Right arrow Articles by Schwartz, S. M.

(Circulation. 1996;93:1095-1106.)
© 1996 American Heart Association, Inc.

Articles

Platelet-Derived Growth Factor–A mRNA Expression in Fetal, Normal Adult, and Atherosclerotic Human Aortas

Analysis by Competitive Polymerase Chain Reaction

Charles E. Murry, MD, PhD; Trudy Bartosek, BS; Cecilia M. Giachelli, PhD; Charles E. Alpers, MD; Stephen M. Schwartz, MD, PhD

From the Department of Pathology, University of Washington, School of Medicine (Seattle).


*    Abstract
up arrowTop
 *Abstract
down arrowIntroduction
 down arrowMethods
down arrowResults
down arrowDiscussion
down arrowReferences
 
Background To understand which growth factors are important for growth of atherosclerotic plaques, it is necessary to know the factor's relative abundance and how its gene is regulated in relation to cell proliferation. We tested whether platelet-derived growth factor–A (PDGF-A) mRNA levels correlated with cell proliferation in developing aorta, normal adult aorta, and atherosclerotic plaques.

Methods and Results We developed a competitive reverse transcription–polymerase chain reaction (RT-PCR) assay to measure human PDGF-A mRNA levels in small tissue samples. A mutated PDGF-A synthetic RNA was used as an internal standard to compete with endogenous PDGF-A mRNA for amplification. The assay is highly sensitive and much more precise than routine RT-PCR. Correction for heteroduplex pairing between the endogenous and mutant PCR products correlates precisely with synthetic RNA standards and quantitative Northern blotting. Immunostaining with the proliferation marker (proliferating cell nuclear antigen) showed the following rank order of proliferation: fetal aorta>>atherosclerotic plaque>normal aortic media. PDGF-A mRNA levels, however, did not correlate with proliferation. Normal adult aorta contained the most PDGF-A mRNA (34.0±7.6 amol/µg total RNA). Fetal aortas were intermediate (10.2±1.6 amol/µg total RNA); advanced atherosclerotic plaques contained the least PDGF-A mRNA (0.3±0.1 amol/µg total RNA). PDGF-A protein was readily detectable in normal media by immunostaining. Advanced plaques generally had less cell-associated PDGF-A protein, although A-chain was also detected in plaque matrix.

Conclusions PDGF-A mRNA and protein do not correlate with proliferation among these three groups. The significance of high levels of PDGF-A mRNA in the "quiescent" aortic media is unknown, but it clearly does not promote cell replication.


Key Words: atherosclerosis • growth substances • polymerase chain reaction • cells


*    Introduction
up arrowTop
 up arrowAbstract
 *Introduction
down arrowMethods
down arrowResults
down arrowDiscussion
down arrowReferences
 
Multiple growth factors and cytokines are now known to be present in human atherosclerotic plaques or arterial lesions in experimental animals.1 2 Despite this, we know very little about how these factors contribute to important events in plaque development, such as cell proliferation, migration, and matrix synthesis. To date, most observations in human tissues have relied on in situ hybridization or immunocytochemistry to determine whether a specific molecule is present and where it is located. Quantitative analyses have usually been limited to determining the frequency of labeled cells. Given the multitude of growth factors present in plaques, it is difficult to infer functional significance based solely on the anatomic distribution of the growth factors. For example, our group is interested in the role of PDGF-A in vascular development and atherogenesis. To understand how PDGF-A gene expression is regulated during these processes, it is necessary to obtain quantitative data.

Quantitative analysis of growth factor protein in plaque has been difficult for small human specimens, such as might be retrieved by coronary atherectomy, due to sensitivity limits. Analysis of RNA expression, however, is amenable to analysis by current methodologies and provides valuable information about gene regulation. In this series of experiments, we developed a competitive RT-PCR assay for human PDGF-A. This method, originally developed by Gilliland et al,3 couples the sensitivity of routine RT-PCR with the precision of solution hybridization. We then tested the hypothesis that PDGF-A mRNA levels would correlate with proliferation rates when comparing fetal aorta, normal adult aorta, and atherosclerotic aortic plaques.


*    Methods
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 *Methods
down arrowResults
down arrowDiscussion
down arrowReferences
 
Tissue Sampling and RNA Extraction
This study was approved by the Human Subjects Committee of the University of Washington. Human fetal aortic tissue (embryonic days, 59 to 120; n=4) was obtained after elective abortions, through the Central Laboratory for Human Embryology (Department of Pediatrics, University of Washington). Aortas were dissected free of surrounding tissue and frozen in liquid nitrogen for RNA extraction. Additional specimens were fixed in 3:1 methanol/acetic acid for morphological analysis. Ischemic time was typically less than 1 hour.

Adult aortic tissue was obtained from cardiac transplant donors and recipients at the University of Washington or from tissue donors via Northwest Tissue Center (Seattle, Wash). Patient characteristics are given in the TableDown. A total of 13 samples were obtained from eight men and four women, ranging in age from 28 to 57 years. The cardiac transplant specimens typically had ischemic times of less than 2 hours. The specimens from tissue donors always had an ischemic time of less than 8 hours—more typically, 4 to 5 hours. Both sources yielded normal tissue for analysis, but atherosclerotic plaques were obtained only from tissue donors. No effect of ischemic time, within the predesignated 8-hour limit, was noted on PDGF-A mRNA levels. Samples from transplant patients were obtained from the ascending aorta, whereas samples from tissue donors were obtained from the descending thoracic and abdominal aortas. No effect of anatomic location was detected on PDGF-A mRNA levels. All samples were dissected under direct visualization after gross examination. The endothelium was removed by gently wiping the luminal surface with gauze, and the adventitia was dissected free. Atherosclerotic plaques were always dissected from the underlying media. Normal segments of aorta were sampled full thickness (intima and media). Tissue samples were frozen in liquid nitrogen and stored at -70°C for RNA extraction. In every case, sections fixed in 4% buffered formalin or methanol Carnoy's solution (60% methanol/30% chloroform/10% glacial acetic acid) were obtained adjacent to the samples for RNA analysis. These sections were used to classify the tissue as normal or atherosclerotic, to verify proper dissection of intimal and medial layers, and to immunostain. Frozen tissue samples were pulverized under liquid nitrogen and then homogenized in 4 mol/L guanidinium isothiocyanate. RNA was prepared by acid phenol/chloroform extraction4 as described,5 and its concentration was determined by absorbance at 260 nm.


View this table:
[in this window]
[in a new window]
  		Table 1. Patient Characteristics

General RT-PCR Procedure
RT was performed with 250 ng of RNA per reaction, unless otherwise specified. A standard PCR buffer (Promega Co) was used for both RT and PCR containing 10 mmol/L Tris, pH 9.0, 50 mmol/L KCl, 2.5 mmol/L MgCl2, and 1% Triton X-100. The RT reaction also contained 5 mmol/L dithiothreitol, 0.5 mmol/L of each deoxynucleotide triphosphate (dNTP), 0.5 µg of random hexamers (Promega), 20 U of RNase inhibitor (Promega), and 200 U of Maloney murine leukemia virus RNase H-negative reverse transcriptase (Superscript, GIBCO-BRL/Life Sciences Technologies), in a total volume of 20 µL. RT was carried out for 1 hour at 37°C, after which the enzyme was denatured by heating to 95°C for 10 minutes. To quantify cDNA synthesis, RT was performed in 5 µL total volume with 13 to 660 fmol of synthetic human PDGF-A cRNA as template. An antisense oligonucleotide spanning nucleotides 1243 to 1266 was used to prime the reaction, which was carried out in the presence of 5 µCi [{alpha}-32P]dCTP. After completion of RT, cDNA synthesis was determined by acid precipitation and scintillation counting.

PCR amplification was performed using the entire cDNA product from an RT reaction. PCR was performed in a 50 µL reaction mix containing 0.2 mmol/L dNTPs, 0.5 µmol/L PCR primers, and 0.25 U Taq polymerase (Promega). All PCR primers were obtained from Operon. For each set of PCR templates and primers, reaction conditions were optimized for MgCl2 concentration and annealing temperature.

Construction of a Competitive Template for Human PDGF-A
Site-specific mutagenesis was performed according to the method of Higuchi,6 as modified by Perrin and Gilliland.7 A PCR primer was designed spanning nucleotides 850 to 873 of human PDGF-A with the sequence 5'-GTGGAATTCGTCAGGAAGAAGCCA-3'.8 The primer contained an A->T substitution at position 857, creating a new EcoRI restriction site. With a wild-type human PDGF-A cDNA (pADIR) as template, an initial round of PCR amplification was performed with this mutant primer in conjunction with an antisense primer corresponding to nucleotides 1243 to 1266. The antisense primer's sequence was 5'-TTTCTGCAGGGAAGCTTCTTACTGCTTCACCGA-3' and contained a Pst I restriction site for subsequent cloning (underlined). The resulting PCR product spanned nucleotides 850 to 1266 and incorporated the new EcoRI restriction site. After purification, this product was used as a "giant primer" for a second round of PCR, in conjunction with a sense primer spanning nucleotides 300 to 323. The sense primer's sequence was 5'-TTTGGATCCGTACTGATTTTCGCCGCCACAGGA-3' and contained a BamHI restriction site (underlined). The product of the second round of PCR spanned nucleotides 300 to 1266 and contained the new EcoRI site. This product was cloned into the BamHI/Pst I sites of pBluescript SK- and designated pMT1. DNA sequencing confirmed the single base-pair substitution at position 857. A wild-type human PDGF-A clone spanning nucleotides 300 to 1266 was cloned into pBluescript SK- and designated pWT1.

To construct sense cRNA molecules for use as competitive PCR standards, both plasmids were linearized with Xho I and then subjected to in vitro transcription with a T7 polymerase kit according to the manufacturer's instructions (Ambion). To obtain precise quantification of the cRNA product, trace amounts of 3H-CTP were added to the reaction mix. The cRNA product was separated from unincorporated nucleotides by centrifugation through a Sephadex G50 column (Pharmacia). The concentration of the final product was determined by scintillation counting of acid precipitates. Analysis of the product by denaturing agarose gel showed a single band of {approx}1000 bases. Northern blot analysis was performed to verify that the final product consisted of full-length transcript. Samples were aliquoted in DEPC water and stored at -70°C.

Competitive RT-PCR
All RT reactions were prepared from a common master mix containing 250 ng of total RNA per reaction. The master mix was aliquoted into individual reaction tubes, and a precisely known amount of mutant human PDGF-A synthetic RNA was added to each tube, typically ranging from 100 to 0.25 amol per reaction. After RT, a PCR master mix was prepared as described above, containing {approx}1 µCi of [{alpha}-32P]dCTP per reaction. For competitive PCR, a sense primer spanning nucleotides 640 to 663 with the sequence 5'-AAGAGAAGCATGGAGGAAGCTGTC-3' and an antisense primer spanning nucleotides 946 to 969 with the sequence 5'-TCCCGTGTCCTCTTCCCGATAATC-3' were used. These primers flanked two large introns (3 and 4),9 preventing amplification of potentially contaminating genomic DNA. Thirty microliters of this mix was aliquoted to each RT reaction, and the samples were overlaid with mineral oil. Thermal cycling was performed with either a Coy TempCycler model 50/60 or an MJ Research model PTC-60 thermal cycler. The cycling parameters began with a 5-minute denaturation at 94°C, followed by 45 cycles for 1 minute at 94°C for denaturation and 2 minutes at 68°C for annealing and extension. Samples were then held at 4°C. The resulting 330–base-pair product included sequence from exons 2 to 5 but did not include any portion of the variably spliced exon 6.10 11

After amplification, samples were digested with 40 U of EcoRI for 3 hours, followed by heat inactivation. Forty microliters of each PCR reaction was electrophoresed through a 2% NuSieve (FMC Bioproducts)/1% agarose gel. The wild-type product gave a 330–base-pair band, and the digested mutant product gave bands of 214 and 116 base-pairs. The ethidium bromide–stained bands were excised, and the 32P activity was determined by scintillation counting; identical plugs of gel trailing the bands were excised for subtraction of background counts. Each competitive PCR run included the following controls: (1) omission of reverse transcriptase; (2) omission of template; (3) a DNA PCR reaction with mutant PDGF-A template, which was used to ensure complete restriction digestion of the mutant product; and (4) a control sample of cellular RNA, which contained a known amount of PDGF-A.

Each sample was analyzed by five competitive RT-PCRs, with the mutant template initially in excess of the wild-type, tapering progressively until the wild-type template was in excess of the mutant. After measurement of 32P activity of the PCR products, regression analysis was performed to determine the concentration of PDGF-A mRNA in the sample. The regression analyses were typically linear with correlation coefficients of >.95. Some of the titrations had one point that was a clear outlier, and it was excluded if this point was not near the equivalency point at which the mutant and wild-type products were at equal concentrations. If two outliers were present or if an outlier occurred near the equivalency point, the titration was repeated.

Cell Culture and Northern Analysis
Human fetal aortic smooth muscle cells12 13 were grown to confluence in Dulbecco's modified Eagle's medium containing 10% calf serum. Cells were growth-arrested by serum withdrawal for 48 hours, after which {alpha}-thrombin (Sigma Chemical Co) was added to a final concentration of 10 nmol/L. Pilot studies demonstrated that this dose of {alpha}-thrombin gave maximal stimulation of DNA synthesis (data not shown). Cells were harvested for RNA isolation at 0, 2, 4, 8, and 24 hours after the addition of thrombin. Twenty micrograms of total RNA per lane was electrophoresed through a 1.2% agarose gel containing formaldehyde, blotted to a Hybond-N membrane (Amersham), and immobilized by cross-linking with UV light. Prehybridization and hybridization were performed as described,5 using a radiolabeled human PDGF-A probe prepared from a PCR product containing base-pairs 640 to 969; the probe therefore corresponded to the same region amplified by competitive PCR. The PDGF-A signal was quantified with a Molecular Diagnostics phosphor imager. To correct for potential variation in loading, the blot was then stripped and reprobed with an end-labeled oligonucleotide corresponding to the 28S ribosomal subunit, as described previously.5 The 28S signal was also quantified with a phosphor imager. The PDGF-A signal was then normalized by dividing by the 28S signal.

Immunocytochemistry
Immunoperoxidase staining was performed on methanol Carnoy's–fixed, paraffin-embedded aortic sections as described previously,14 with diaminobenzidine as chromogen. Rabbit polyclonal antiserum to human PDGF-A was obtained from Santa Cruz Biotechnology and used at a titer of 1:50. This antiserum was raised against a synthetic peptide comprising the 30 residues of the amino terminus of the "long" form of PDGF-A, including the last 12 amino acids of exon 5 and all 18 amino acids of the variably present exon 6.10 11 The antiserum recognizes a single band of 10 to 14 kD on reduced gels of human fetal kidney by Western blotting, and immunostaining can be inhibited by preincubation of the antiserum with recombinant human PDGF-A containing exon 6.15 Cell proliferation was studied with the use of an antibody to PCNA (clone PC10; DAKO). The PCNA antibody marks cells at all points in the cell cycle, giving higher replication rates than measures of S-phase such as 3H-thymidine incorporation (reviewed in Reference 16). Both methods are internally consistent, permitting comparison of relative replication rates. Smooth muscle cells were identified with a monoclonal antibody specific to smooth muscle {alpha}-actin (IA4; DAKO).17 Macrophages were identified with an antibody to CD68 (KP1; DAKO),18 and endothelial cells were identified with the use of lectin staining with Ulex europaeus (Vector Laboratories).19 20

Statistical Analysis
PDGF-A mRNA levels were normalized to total RNA and expressed as amol/µg total RNA. Linear regressions were performed by the least-squares method. All data are presented as mean±SEM. For comparisons involving three groups, a Kruskal-Wallis test for nonparametric data sets was used, followed by a Mann-Whitney U test with a Bonferroni correction for multiple comparisons to determine which groups differed significantly. A value of P<.05 was considered statistically significant.


*    Results
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 up arrowMethods
*Results
down arrowDiscussion
down arrowReferences
 
Evaluation of Routine PCR as a Quantitative Assay
Before undertaking the development of a competitive RT-PCR assay, we determined the precision of conventional PCR. We used a PDGF-A plasmid to examine the relation between the amount of initial template and the amount of PCR product generated. Master mixes were used to maximize the chemical similarity of each reaction, and four identical aliquots were used at each concentration of template. The data from one such experiment are plotted in Fig 1Down. Although there was a positive relation between the amount of template and the amount of PCR product, there was a great degree of variability among individual samples. For example, at a given concentration of PDGF template, there often was twofold to fourfold variation in the amount of PCR product generated, even though the reactions were aliquoted from the same initial mix. If one were to use this as a typical standard curve for analytical chemistry, extrapolating a sample product against this curve would yield template levels that varied by as much as sixfold. We concluded that this amount of analytical variability could well exceed the biological differences we were trying to measure and that routine PCR was not suited for precise quantitative work. Similar problems with variability have been reported by Gilliland et al.3



View larger version (20K):
[in this window]
[in a new window]
  		Figure 1. Quantitative analysis by routine PCR. A human PDGF-A plasmid was used as template for PCR as described in "Methods." PCR products were quantified after gel electrophoresis by 32P incorporation and are expressed in arbitrary units (cpm of product band minus background). Although there was a general positive relation between the amount of starting template and final PCR product, there was considerable variability, such that chemically identical reactions varied twofold to fourfold in product yield. The data are representative of three separate experiments.

Validation of RT and Competitive PCR
We chose to use competitive PCR to circumvent the problem of variability encountered with routine PCR. This approach, initially developed by Gilliland et al,3 uses an internal standard within each reaction tube. Amplification of this standard serves as a reference for the gene of interest. For our studies, a mutant human PDGF-A competitive template was constructed by mutating a single base-pair to yield a new EcoRI restriction site. The mutant gene is added as a synthetic cRNA to the RT reaction. Because the mutant is virtually identical to the wild-type mRNA, the mutant is reverse-transcribed and PCR-amplified with identical efficiency. Thus, the ratio of mutant to wild-type PCR products reflects the starting ratio of mutant to wild-type templates. After amplification, the products are digested with EcoRI, which cuts only the mutant gene, and the products are quantified after separation by gel electrophoresis.

Because this assay should depend on both the RT reaction and PCR being quantitative, we validated the steps independently. To determine whether the cDNA product stoichiometrically reflected the initial RNA template level, we synthesized human PDGF-A cRNA and used it as template for RT. The results of this assay, shown in Fig 2Down, show that the PDGF-A cDNA product is linearly related to the amount of PDGF-A cRNA template over a wide range of template concentrations.



View larger version (30K):
[in this window]
[in a new window]
  		Figure 2. RT stoichiometrically reflects starting RNA levels. Synthetic human PDGF-A RNA was reverse-transcribed as described in "Methods," and the cDNA product was determined by 32P incorporation of acid precipitates. There was a linear relation between starting cRNA template concentration and the yield of the cDNA product. The reaction had an overall efficiency of 3.4%, ie, 3.4 molecules of cDNA/100 cRNA molecules.

To validate competitive PCR, a series of titrations was performed with precisely known amounts of mutant and wild-type PDGF-A plasmids as templates. PCR products were measured to determine whether the expected ratios were obtained. We demonstrated that the assay was highly linear as long as ratios of mutant to wild-type templates were between 1:40 and 40:1 (Fig 3Down). When ratios exceeded 40:1, the product ratios plateaued despite increasing template ratios. The slope of the reaction, however, was significantly different from the predicted slope of 1.0. In fact, the ratio of mutant to wild-type product was typically 40% to 60% of that predicted from the ratio of the templates, suggesting underestimation of the mutant product or overestimation of the wild-type product (Fig 3Down, lower line).



View larger version (26K):
[in this window]
[in a new window]
  		Figure 3. Heteroduplex formation after competitive PCR. Competitive PCR was performed with precisely known amounts of mutant and wild-type PDGF-A plasmids as described in "Methods." The titration curve should pass through the origin with a slope of 1.0. The uncorrected titration curve has a slope significantly lower than the predicted value, indicating underestimation of the mutant product and/or overestimation of the wild-type product. After correction by driving heteroduplex formation to equilibrium and solving mathematically for the true concentrations of mutant and wild-type products, the slope of the regression line did not differ significantly from 1.0. The data are representative of at least five independent experiments.

We eventually determined that this problem was due to the formation of heteroduplex pairs between the mutant and wild-type products. During the separation and reannealing of DNA strands that occurred in PCR, a significant proportion of the product consisted of one strand of mutant and one strand of wild-type DNA. Because only one of the two strands contained the EcoRI restriction site, the heteroduplex product could not be digested and therefore comigrated with the wild-type product. This resulted in overestimation of the wild-type product with corresponding underestimation of the mutant product. To correct for this problem, we first purified the PCR products and then drove heteroduplex formation to equilibrium by heat-denaturing the PCR products and allowing them to reanneal through slow cooling. The products were then digested and electrophoresed as before. Assuming identical hybridization affinities, the distribution of homoduplexes and heteroduplexes at equilibrium is given by the following binomial equation: a2+2ab+b2=1, where a2 represents the proportion of wild-type homoduplex, b2 represents the proportion of mutant homoduplex, and 2ab represents the proportion of heteroduplex. The true proportions of mutant and wild-type products are given by a and b, respectively. This correction consistently gave slopes to the titration curve that were not significantly different from 1.0 (Fig 3Up, upper line). We then used synthetic wild-type and mutant PDGF cRNA templates in precisely known quantities to test whether competitive RT-PCR would give the expected results in a defined system. After correction for heteroduplex formation, the titration curves were highly linear with slopes of 1.0 (Fig 4Down).



View larger version (24K):
[in this window]
[in a new window]
  		Figure 4. Competitive RT-PCR gives the predicted yield with synthetic mutant and wild-type cRNA standards. The titration curve should follow the line of unity (x=y). The experimental curve does not differ significantly from this relation.

Competitive RT-PCR Correlates With Quantitative Northern Blot Analysis
To validate the method with complex RNA samples from living cells, we compared competitive RT-PCR with Northern blotting, analyzed as quantitatively as possible. PDGF-A mRNA was induced in cultured human fetal smooth muscle cells by the addition of 10 nmol/L {alpha}-thrombin.21 RNA was isolated at timed intervals thereafter. Competitive PCR was performed on samples of 250 ng total RNA. Northern blots (20 µg/lane) were probed with a 32P-labeled PDGF-A probe corresponding to the region amplified for PCR analysis, and band intensity was measured with a phosphor imager. The blot was subsequently stripped, reprobed with a 32P end-labeled oligonucleotide probe to the 28S ribosomal RNA subunit, and phosphor-imaged to correct for variation in loading. Both methods demonstrated a time-dependent increase in PDGF-A mRNA, peaking at 8 hours and returning toward baseline by 24 hours (Fig 5ADown). When the competitive RT-PCR values were plotted against the Northern blot analysis values, there was an excellent linear correlation, which did not differ from the predicted line of unity (Fig 5BDown).



View larger version (23K):
[in this window]
[in a new window]
  		Figure 5. A, PDGF-A mRNA was induced in cultured fetal human aortic smooth muscle cells by incubation with {alpha}-thrombin for various lengths of time. Total RNA was analyzed by quantitative Northern hybridization with a phosphor imager and by competitive RT-PCR. Both methods gave a comparable time course for PDGF-A induction, peaking at 8 hours and returning toward baseline by 24 hours. B, Regression analysis of Northern versus competitive RT-PCR measurements of PDGF-A mRNA in cultured smooth muscle cells regardless of time after {alpha}-thrombin stimulation. There is a close linear relation that does not differ significantly from the line of unity.

Cell Proliferation in Human Aortas
Cell proliferation was studied by immunostaining with an antibody directed against PCNA.16 22 In sections of human fetal aorta ranging from embryonic days 74 to 120, there was exuberant PCNA staining (Fig 6ADown). In the media, proliferative rates ranged from 13% to 56% and averaged 24%. To the best of our knowledge, this represents the first measurement of replication rates in human fetal aorta. Proliferation was also detected in the endothelium and in the small amount of adventitia present, albeit at lower rates. In agreement with previous data from our group,16 22 the normal adult aortic media displayed virtually no PCNA staining (Fig 6CDown). Sections of atherosclerotic aortic plaque (Fig 6BDown) showed an intermediate level of proliferation, typically none to one cell per high-power field. The PCNA staining in plaque was clearly 10- to 100-fold lower than in fetal aorta and clearly greater than in the normal media. We did not attempt to quantify the low levels of proliferation present in normal and atherosclerotic aortas, however. This quantification has been done in detail previously,16 22 and the present data appear to be quite consistent with those results.





View larger version (446K):
[in this window]
[in a new window]
  		Figure 6. Cell proliferation measured by PCNA staining in fetal, atherosclerotic, and normal adult aorta. A, In fetal aorta (day 117), there was intense proliferative activity in endothelial cells, medial smooth muscle cells, and, to a lesser extent, adventitial cells. Black nuclei indicate proliferation; L, lumen; M, media; A, adventitia; and bar, 7 µm. B, In atherosclerotic aortic plaques, there was substantially less replication, typically none to one cell per high-power field (arrow). Bar, 3.5 µm. C, In normal aortic media, there was virtually no replication detected. Bar, 3.5 µm.

PDGF-A mRNA Levels and Immunostaining in Human Aortas
We obtained microscopic sections of adult aorta adjacent to the samples taken for RNA extraction to classify the samples. Normal aortic samples generally had mild-to-moderate intimal thickening, without significant lipid accumulation or fibrosis. The atherosclerotic plaques in this series were advanced lesions containing dense fibrous caps, extensive necrotic cores with extracellular lipid accumulation, a mild-to-moderate inflammatory cell infiltrate, and mild-to-moderate neovascularization. All samples of plaque were free of underlying media. Based on the above proliferation data, we hypothesized the following rank order of PDGF-A expression: fetal>atherosclerotic plaque>normal aorta. In contrast to our hypothesis, PDGF-A did not correlate with proliferative activity among these tissues (Fig 7Down). Normal adult aorta, the least proliferative tissue examined, expressed almost four times more PDGF-A than did fetal aorta (adult, 34.0±7.6 amol/µg RNA; fetal, 10.2±1.6 amol/µg RNA). Atherosclerotic plaques had far less PDGF-A mRNA than either adult or fetal aorta (0.31±0.11 amol/µg). Statistical analysis was complicated by the limited availability of samples in the atherosclerotic and fetal aortas (n=4 in each group). The Kruskal-Wallis test for multiple comparisons indicated a highly significant difference among the three groups, with P=.003. Pairwise comparison with a Mann-Whitney U test showed a significant difference for normal adult aorta versus atherosclerotic aorta (P=.004) and for fetal aorta versus atherosclerotic aorta (P=.02). Comparison of normal adult aorta versus fetal aorta achieved borderline significance (P=.055). When corrections for multiple comparisons were performed, the atherosclerotic group was still significantly different from the normal adult aorta (P=.01), but the fetal aorta versus atherosclerotic aorta was of borderline significance (P=.06). The normal adult aortas were not significantly different from fetal aortas after correction for multiple comparisons (P=.18).



View larger version (23K):
[in this window]
[in a new window]
  		Figure 7. PDGF-A mRNA in fetal, adult, and atherosclerotic aortas. Fetal aortas averaged 10.2±1.6 amol PDGF-A/µg total RNA. Adult aortas averaged 34.0±7.6 amol/µg total RNA, although there was considerable variability in PDGF-A content among specimens. Atherosclerotic plaques contained substantially less PDGF-A mRNA than either fetal or normal adult vessels, averaging 0.3±0.1 amol/µg RNA. Each data point represents one sample.

Immunostaining was performed to determine the anatomic location of PDGF-A protein. In fetal aortas (Fig 8ADown), there was relatively weak staining for PDGF-A within the media. Close inspection suggested that most of the protein was intracellular, although weak extracellular staining could not be ruled out. The media stained vigorously with antibodies to smooth muscle {alpha}-actin (Fig 8BDown). In the normal adult aorta, PDGF-A was readily detected within smooth muscle cells of the media and the intima (Fig 8CDown). The distribution of PDGF-A protein correlated anatomically with the distribution of smooth muscle {alpha}-actin (Fig 8DDown), indicating that PDGF-A protein resided in smooth muscle cells. No extracellular staining was present in normal adult aorta. In atherosclerotic aortas (Fig 8EDown), the staining pattern was more complex. All atherosclerotic vessels demonstrated PDGF-A in medial smooth muscle cells (Fig 8FDown). Interestingly, medial PDGF-A staining appeared to be less prominent in atherosclerotic than in normal arteries, although no attempt was made to quantify this possible difference. Within the plaque, some areas rich in smooth muscle cells showed virtually no PDGF-A staining, suggesting downregulation of protein production. In other areas, PDGF-A was present within plaque smooth muscle cells, similar to that shown for the intima in Fig 8CDown. We also observed regions of plaque exhibiting diffuse, extracellular staining, which presumably represents PDGF-A bound to the extracellular matrix (data not shown). Extracellular staining was most prominent in the less cellular, fibrotic regions of the plaque but was also present in areas containing numerous smooth muscle cells. No association of PDGF-A was seen with macrophages or foam cells. Occasional cells of the plaque vasa vasorum contained PDGF-A; this was more prominent in vasa that were invested with smooth muscle cells.









View larger version (1031K):
[in this window]
[in a new window]
  		Figure 8. Immunostaining for PDGF-A protein and smooth muscle {alpha}-actin in fetal, normal adult, and atherosclerotic aortas. A, Fetal aorta, anti–PDGF-A. There is weak PDGF-A staining within medial smooth muscle cells. L indicates lumen; M, media; and A, adventitia. Bar, 7 µm. B, Fetal aorta, anti–smooth muscle {alpha}-actin. The medial smooth muscle cells stain strongly for {alpha}-actin. Bar, 7 µm. C, Normal aorta, anti–PDGF-A. PDGF-A protein is present in the media and in smooth muscle cells within the intima (arrow). Bar, 35 µm. D, Adjacent section of normal aorta stained with an antibody to smooth muscle actin. PDGF-A positive region within the intima corresponds to focus of smooth muscle cells (arrow). Bar, 35 µm. E, Atherosclerotic aorta, anti–PDGF-A. PDGF-A staining is present within smooth muscle cells of the media. No PDGF-A staining is present within the atherosclerotic intima. Arrows denote the position of the internal elastic lamina. I indicates intima. Bar, 14 µm. F, Adjacent section of atherosclerotic aorta stained with anti–smooth muscle actin. There are numerous actin-positive smooth muscle cells within the intimal and medial layers. Arrows denote the position of the internal elastic lamina. Bar, 14 µm. G, Adjacent section of atherosclerotic aorta stained with hematoxylin and eosin. The media contain smooth muscle cells with orderly lamellae. The intima contains smooth muscle cells, scattered inflammatory cells, and lipid-laden foam cells. Arrows denote the position of the internal elastic lamina. Bar, 14 µm.


*    Discussion
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 up arrowMethods
up arrowResults
*Discussion
down arrowReferences
 
Critique of Method
In this study, we sought to develop a method for quantitative analysis of gene expression that would be suitable for working with small tissue samples, such as human vascular specimens or vessels from small experimental animals. The limited amount of RNA available from such specimens requires a highly sensitive assay, and we therefore chose to use RT-PCR. Our initial experiments clearly showed that routine PCR is an imprecise quantitative tool. Using chemically identical reaction mixes aliquoted into multiple tubes, we obtained twofold to fourfold variation in the amount of PCR product (Fig 1Up). Furthermore, extrapolation from the standard curve to determine the concentration of an unknown sample would yield up to sixfold variation in the amount of template. The reason for this marked tube-to-tube variability is not known, but it may relate to very slight temperature variations within the heating block of the thermal cycler, which translate to large differences in yield after exponential amplification. This problem with variability has been reported by others.3 Despite this, the literature contains many reports in which investigators have used the yield of a PCR reaction to predict template concentrations. Although this may be acceptable for all-or-none expression patterns or when differences exceed 5- to 10-fold, many important biological variations do not exceed the noise threshold for measurement by routine PCR. Thus, quantitative data obtained with routine PCR must be interpreted with great caution.

In contrast, competitive PCR offers a highly sensitive assay with excellent linearity and good tube-to-tube and run-to-run reproducibilities. Using 250 ng of total RNA, we were able to amplify PDGF-A from virtually every sample. Although we did not determine the minimum amount of RNA required for consistent amplification, it is likely that we could use substantially less than 250 ng for each assay.

One pitfall we encountered in the use of competitive PCR was the formation of heteroduplex pairs between mutant and wild-type products. Heteroduplex pairs contain only one strand with the EcoRI restriction site. Therefore, they are not digested and comigrate with the wild-type product in agarose gels. This leads to overestimation of the wild-type PDGF product and, consequently, overestimation of the amount of PDGF mRNA in the sample. This problem can be overcome by driving heteroduplex formation to equilibrium and then solving mathematically for the true concentration of mutant and wild-type products. Heteroduplex pairs can form in our assay because the mutant and wild-type strands differ by just 1 base-pair in 330. There are other types of mutant templates for competitive PCR that have substantial insertions or deletions between the primer sequences, which permit detection without restriction digestion.3 A further variant is the competitive template that shares only primer sequences with the wild-type gene, with the intervening nucleotide sequence completely different.23 It seems likely that competitors that have less sequence homology would be less likely to form heteroduplexes due to the necessity of creating thermodynamically unfavorable stem-loop structures. As sequence homology diverges, however, it also becomes increasingly likely that the efficiency of RT and PCR amplification will differ between mutant and wild-type templates. Thus, one may not obtain an accurate copy number with such an approach, although relative differences between two samples could still be measured.

In analyzing gene expression by competitive RT-PCR, it is critical to normalize the data to a proper denominator. Given that most PDGF-A appears to be produced by smooth muscle cells, we considered normalizing our data to a smooth muscle–specific gene such as smooth muscle {alpha}-actin. This would have the advantage of correcting for dilution of smooth muscle cells by endothelial cells and inflammatory cells in plaques. However, smooth muscle cells from plaque also downregulate their {alpha}-actin levels relative to medial cells. The same is probably true for many other smooth muscle genes. Thus, it would be difficult to find a constant, smooth muscle–specific denominator to permit comparison of medial with plaque smooth muscle cells. Because of this difficulty, we chose to use total RNA as a denominator. Based on the histology of the lesions, it seems highly unlikely that the 100-fold difference we observed between normal aorta and plaque could be explained entirely by other cell types.

Finally, it should be emphasized that for technical reasons we have not quantified distribution of PDGF-A protein. Our studies showed a marked difference in the steady state levels of PDGF-A mRNA in fetal, normal adult, and atherosclerotic aortas. This finding must reflect either differing transcriptional activity or rates of message degradation, but it remains to be determined whether there are similar differences in protein content. We addressed this issue qualitatively with the use of immunostaining and found abundant PDGF-A protein in normal aortic media, with an apparent reduction in plaque smooth muscle cells (Fig 8Up). A precise comparison of PDGF-A protein content is not possible from these data, however.

PDGF in Atherosclerosis and Vascular Injury
PDGF was one of the first mitogens defined for cultured smooth muscle cells, and it has been postulated to induce smooth muscle proliferation in atherosclerosis.24 It is well known that PDGF exists in two isoforms—an A chain and a B chain—which are products of different genes. PDGF functions as a dimer, composed of either AA, AB, or BB chains, depending on the source. Similarly, there are two separate PDGF receptor genes, designated PDGFR-{alpha} and PDGFR-ß. The current model for PDGF receptor/ligand interactions suggests that the A-chain can only bind to the {alpha}-receptor, whereas the B-chain can bind to either the {alpha}- or ß-receptor. Therefore, as dimers, PDGF-AA ligand can bind only to {alpha}{alpha}-receptor dimers, AB ligand can bind to {alpha}{alpha}- or {alpha}ß-receptor dimers, and BB ligand can bind to all receptor combinations.25

Despite the interest in the role of PDGF in atherosclerosis, there have been relatively few studies of PDGF expression in human vascular specimens. Initial studies by Barrett and Benditt26 demonstrated that both PDGF-A and -B chain transcripts could be detected in human carotid plaques with the use of Northern blot analysis. By reprobing blots with markers for macrophages (c-fms), endothelial cells (von Willebrand factor), or smooth muscle cells (smooth muscle {alpha}-actin) and by performing multiple regression analysis, the authors concluded that PDGF-A was produced predominantly by smooth muscle cells, whereas PDGF-B was produced predominantly by macrophages. Interestingly, Barrett and Benditt also found relatively abundant PDGF-A transcript in the normal aortic media. Libby et al27 demonstrated that cultured smooth muscle cells from atherosclerotic plaques constitutively synthesized PDGF-A mRNA and secreted the protein into the culture medium. Subsequent in situ hybridization studies by Wilcox et al28 showed that PDGF-A mRNA was associated with plaque mesenchymal cells (presumably of smooth muscle origin), particularly those in perivascular locations, whereas PDGF-B mRNA was associated with endothelial cells of the plaque vasa vasorum.

A recent article by Rekhter and Gordon29 examined the relation between smooth muscle cell proliferation and PDGF-A protein in human carotid plaques. They found PDGF-A in smooth muscle–like cells within plaques and also found diffuse regions of extracellular staining. These findings correlate well with our current immunostaining results in aortic plaques. They performed a detailed morphometric evaluation and demonstrated that the most potent predictors of smooth muscle cell proliferation were either the presence of microvessels or the presence of PDGF-A staining. These data are consistent with PDGF-A acting as a mitogen or as a permissive comitogen, enabling cells to replicate in response to other growth factors. The significance of PDGF-A bound to plaque matrix is unclear. Ostman et al11 and Raines and Ross10 have shown that binding of PDGF-A to the cell surface or matrix is facilitated by the presence of the variably spliced exon 6. It is possible that matrix-bound PDGF-A is functionally sequestered but could be released by as-yet-undefined stimuli to bind its receptor.

These reports of PDGF-A expression in human atherosclerosis may seem discrepant with the present finding that normal media contain many times more PDGF-A mRNA than is found in plaques. Indeed, this result was initially surprising to us, but after further analysis, we do not believe there is a real discrepancy. The previous in vivo studies used either immunostaining or in situ hybridization to detect PDGF-A. These morphological approaches are by nature qualitative, whereas analysis by competitive RT-PCR is quantitative. We found PDGF-A mRNA and protein in every atherosclerotic sample studied, which is consistent with the previous reports. What the previous studies failed to note, however, was the abundance of PDGF-A in the normal media. In situ hybridization for PDGF-A gives a diffuse signal over the media, which can be overlooked due to lack of contrast. In plaque, PDGF-A expression is focal and eye-catching. This may have contributed to the (erroneous) notion that PDGF-A was upregulated in plaque. Quantitative analysis of gene expression, such as afforded by competitive RT-PCR, provides important information that complements more conventional morphological approaches.

PDGF expression has been studied more extensively in cultured cells and in animal models of vascular injury and repair. In cultured smooth muscle cells, exogenously added PDGF-AA is a relatively weak mitogen. On the other hand, induction of PDGF-A synthesis is apparently required for the full mitogenic effects of other growth factors such as transforming growth factor–ß30 and {alpha}-thrombin,31 since antibodies to PDGF-A inhibit cell proliferation induced by these factors. These data suggest that PDGF-A may serve as a permissive agent or mitogenic cofactor. In the rat model of carotid balloon injury, Majesky et al32 showed that medial PDGF-A mRNA levels increased 10- to 12-fold within the first 6 hours after injury and then declined to moderately elevated levels (fourfold) by 24 hours. At 2 weeks after injury, when a neointima had formed, PDGF-A mRNA was localized by in situ hybridization to smooth muscle cells closest to the lumen. These luminal smooth muscle cells show the greatest proliferative activity remaining after 2 weeks, but it has not been determined whether PDGF-A actually causes cell replication in this setting.

Studies aimed at determining the function of PDGF after balloon injury suggest that its greatest effect may be as a chemotactic factor rather than as a mitogen. Infusion of PDGF-BB (which binds to all receptor combinations) results in intimal thickening with only a modest increase in cell replication, suggesting a stimulation of cell migration.33 Similarly, neutralization of PDGF with a polyclonal antibody recognizing all isoforms inhibits neointima formation without influencing cell replication.34 In healing synthetic vascular grafts in baboons, it has been shown35 36 37 that proliferation of intimal smooth muscle cells was accompanied by synthesis of PDGF-A mRNA by endothelial cells and smooth muscle cells. Geary et al35 38 hypothesized that shear-dependent remodeling in vascular grafts is mediated in part by induction of PDGF-A.

PDGF-A in the Normal Vessel Wall
One of the most striking findings of the present study was that PDGF-A transcripts were approximately 100 times more abundant in the normal aortic media than in advanced atherosclerotic plaques. In the rat, we recently found with the use of competitive PCR that normal carotid arteries have substantially more PDGF-A mRNA than either injured media or neointima (J. Taguchi, C. Murry, F. Tsui, B. Herrin, M. Pech, S. Schwartz, unpublished observations). Because proliferation is an extremely rare event in the normal media (Fig 6CUp), PDGF-A clearly is not an active mitogen in this tissue. Several functions for PDGF-A seem possible based on cell culture experiments or extrapolation from other in vivo systems. One possibility is that PDGF-A is a trophic factor, promoting vessel wall mass either by cell hypertrophy39 40 or by increasing extracellular matrix synthesis.41 Negoro et al42 recently demonstrated that spontaneously hypertensive rats had twofold higher PDGF-A mRNA content in their aortas than did normotensive Wistar-Kyoto rats. Furthermore, treatment with a diverse group of antihypertensive drugs lowered A-chain mRNA levels and blood pressures in parallel. The authors suggested that PDGF-A functioned as an autocrine hypertrophic factor in hypertension. A second possibility lies in the ability of PDGF to stimulate vascular contraction,43 maintaining contractile tone in the normal aorta. A third possibility is that PDGF-A might be important in inhibiting smooth muscle migration from the media. Although PDGF-BB is a potent chemotactic factor, there is evidence that PDGF-AA inhibits cell migration in response to several chemotactic agents, including PDGF-BB.44 45 This is a controversial area, however, in that other investigators46 have reported that both PDGF-AA and PDGF-BB can stimulate smooth muscle cell migration. Finally, it is possible that PDGF-A could serve as a "survival factor," preventing programmed cell death. PDGF-AA has been shown to be a survival factor for oligodendroglia in vivo.47 Furthermore, Bennett et al recently demonstrated that PDGF-AA and -BB are survival factors for smooth muscle cells that have been induced to undergo apoptosis by constitutive expression of c-myc48 49 and also for plaque smooth muscle cells, which show high rates of spontaneous apoptosis in vitro.48 Thus, there are many possible functions for the surprisingly high level of PDGF-A mRNA within the normal aortic media. It will be important to study the PDGF-{alpha} receptor to determine which cells are capable of responding to PDGF-A produced by the media.

We wondered whether different anatomic locations within the aorta might differ in PDGF-A mRNA content, given known differences in developmental origin or susceptibility to atherosclerosis. No consistent variation was observed, however, in either proliferation or PDGF-A levels as a function of location in normal or atherosclerotic aorta. For example, no differences were found when comparing normal thoracic aorta with abdominal aorta (data not shown). This finding must be tempered by the fact that we did not set out to test for regional variation within the aorta. This would best be done by taking multiple samples from the same patient; given the considerable variation of PDGF-A expression levels among different patients, we might have missed relatively subtle regional variations with our approach.

Proliferation in the Fetal Aorta
We observed strikingly high rates of medial smooth muscle cell proliferation in fetal aortas, with a mean proliferation rate of 24%. This is in marked contrast to the normal adult media, in which replication is an exceedingly rare event. To the best of our knowledge, this represents the first report of proliferation rates in fetal human aorta. The basis for the high replication rates in fetal aorta is not known, but similar patterns are seen in cultured cells. Studies by Cook et al50 have shown that smooth muscle cells from embryonic rat aortas proliferate rapidly in culture without serum or exogenous mitogens, whereas adult smooth muscle cells require exogenous growth factors. It is not known whether the embryonic smooth muscle cells secrete a mitogen into the culture medium or whether intracellular factors are responsible. Heterokaryons between fetal and adult smooth muscle cells show that the adult phenotype is dominant.51 Presumably, the autonomous proliferation in fetal cells is extinguished in heterokaryons by some inhibitory factor produced in the adult cells. The identity of the inhibitory factor has not yet been determined.

Summary
Competitive RT-PCR provides extremely sensitive and precise measurement of growth factor mRNA levels within human and experimental vascular specimens. When coupled with qualitative methods such as in situ hybridization and immunocytochemistry, it is possible to map both the amount of a gene product and the cell type responsible for its production. This approach has demonstrated that PDGF-A, traditionally considered a mitogen and possible chemotactic factor, is expressed at much higher levels in quiescent adult aorta than either fetal aorta or atherosclerotic plaques. This provides evidence for an important nonmitogenic role for PDGF-A in the maintenance of the normal media.


*    Selected Abbreviations and Acronyms
 
PCNA = proliferating cell nuclear antigen
PCR = polymerase chain reaction
PDGF = platelet-derived growth factor
RT = reverse transcription


*    Acknowledgments
 
This work was supported by NIH grants HL-47151, HL-07312, HL-42270, and DK-47659. The authors are indebted to Chris Covin and Francine Stanton for assistance with RNA extractions and to Kelly Hudkins for performing the immunostaining. We thank Drs Joan Lemire, Luis Borges, and Gary Gilliland for fruitful discussions and advice. We also wish to express gratitude to the staff of the Central Laboratory for Human Embryology at the University of Washington (supported by NIH grant HD-00836) and to the staff of Northwest Tissue Center for providing us with many of our vascular specimens.


*    Footnotes
 
Reprint requests to Charles E. Murry, MD, PhD, Department of Pathology, Box 357335, University of Washington School of Medicine, Seattle, WA 98195. E-mail murry@u.washington.edu.

Received July 31, 1995; revision received October 23, 1995; accepted October 26, 1995.


*    References
up arrowTop
 up arrowAbstract
 up arrowIntroduction
 up arrowMethods
up arrowResults
up arrowDiscussion
*References
 

  1. Schwartz SM, Heimark RL, Majesky MW. Developmental mechanisms underlying pathology of arteries. Physiol Rev. 1990;70:1177-1209. [Abstract/Free Full Text]
  2. Ross R. The pathogenesis of atherosclerosis: a perspective for the 1990s. Nature. 1993;362:801-813. [Medline] [Order article via Infotrieve]
  3. Gilliland G, Perrin S, Blanchard K, Bunn HF. Analysis of cytokine mRNA and DNA: detection and quantitation by competitive polymerase chain reaction. Proc Natl Acad Sci U S A. 1990;87:2725-2729. [Abstract/Free Full Text]
  4. Chomczynski P, Sacchi N. Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem. 1987;162:156-159. [Medline] [Order article via Infotrieve]
  5. Murry CE, Giachelli CM, Schwartz SM, Vracko R. Macrophages express osteopontin during repair of myocardial necrosis. Am J Pathol. 1994;145:1450-1462. [Abstract]
  6. Higuchi R. Recombinant PCR. In: Innis MA, Gelfand DH, Sninsky JJ, White TJ, eds. PCR Protocols: A Guide to Methods and Applications. San Diego, Calif: Academic Press; 1990:177-183.
  7. Perrin S, Gilliland G. Site-specific mutagenesis using asymmetric polymerase chain reaction and a single mutant primer. Nucleic Acids Res. 1990;18:7433-7438. [Abstract/Free Full Text]
  8. Betsholtz C, Johnsson A, Heldin C, Westermark B, Lind P, Urdea MS, Eddy R, Shows TB, Philpott K, Mellor AL, Knott TJ, Scott J. cDNA sequence and chromosomal localization of human platelet-derived growth factor A-chain and its expression in tumour cell lines. Nature. 1986;320:695-699. [Medline] [Order article via Infotrieve]
  9. Bonthron DT, Morton CC, Orkin SH, Collins T. Platelet-derived growth factor A chain: gene structure, chromosomal location, and basis for alternative mRNA splicing. Proc Natl Acad Sci U S A. 1988;85:1492-1496. [Abstract/Free Full Text]
  10. Raines EW, Ross R. Compartmentalization of PDGF on extracellular binding sites dependent on exon-6-encoded sequences. J Cell Biol. 1992;116:533-543. [Abstract/Free Full Text]
  11. Ostman A, Andersson M, Betsholtz C, Westermark B, Heldin CH. Identification of a cell retention signal in the B-chain of platelet-derived growth factor and in the long splice version of the A-chain. Cell Regul. 1991;2:503-512. [Medline] [Order article via Infotrieve]
  12. Perez-Reyes N, Halbert CL, Smith PP, Benditt EP, McDougall JK. Immortalization of primary human smooth muscle cells. Proc Natl Acad Sci U S A. 1992;89:1224-1228. [Abstract/Free Full Text]
  13. Conroy SC, Hart CE, Giachelli CM, McDougall JK. Characterization of human aortic smooth muscle cells transformed with HPV16 E6/E7 open reading frames. Am J Pathol. 1995;147:753-762. [Abstract]
  14. Giachelli CM, Bae N, Almeida M, Denhardt DT, Alpers CE, Schwartz SM. Osteopontin is elevated during neointima formation in rat arteries and is a novel component of human atherosclerotic plaques. J Clin Invest. 1993;92:1686-1696.
  15. Alpers CE, Hudkins KL, Ferguson M, Johnson RJ, Rutledge JC. Platelet-derived growth factor A-chain expression in developing and mature human kidneys and in Wilms' tumor. Kidney Int. 1995;48:146-154. [Medline] [Order article via Infotrieve]
  16. Gordon D, Reidy MA, Benditt EP, Schwartz SM. Cell proliferation in human coronary arteries. Proc Natl Acad Sci U S A. 1990;87:4600-4604. [Abstract/Free Full Text]
  17. Skalli O, Roprax P, Trzeciak A, Benxonana G, Gillessen D, Gabbiani G. A monoclonal antibody against alpha-smooth muscle actin: a new probe for smooth muscle differentiation. J Cell Biol. 1986;103:2787-2796. [Abstract/Free Full Text]
  18. Warnke RA, Pulford KAF, Pallesen G, Ralfkiaer E, Brown DC, Gatter KC, Mason DY. Diagnosis of myelomonocytic and macrophage neoplasms in routinely processed tissue biopsies with monoclonal antibody KP1. Am J Pathol. 1989;135:1089-1095. [Abstract]
  19. Holthofer H, Virtanen I, Kariniemi A, Hormia M, Linder E, Miettinen A. Ulex europaeus I lectin as a marker for vascular endothelium in human tissues. Lab Invest. 1982;47:60-66. [Medline] [Order article via Infotrieve]
  20. Alpers CE, Beckstead JH. Monocyte/macrophage-derived cells in normal and transplanted human kidneys. Clin Immunol Immunopathol. 1985;36:129-140. [Medline] [Order article via Infotrieve]
  21. Okazaki H, Majesky MW, Harker LA, Schwartz SM. Regulation of platelet-derived growth factor ligand and receptor gene expression by {alpha}-thrombin in vascular smooth muscle cells. Circ Res. 1992;71:1285-1293. [Abstract/Free Full Text]
  22. O'Brien ER, Alpers CE, Stewart DK, Ferguson M, Tran N, Gordon D, Benditt EP, Hinohara T, Simpson JB, Schwartz SM. Proliferation in primary and restenotic coronary atherectomy tissue: implications for antiproliferative therapy. Circ Res. 1993;73:223-231. [Abstract/Free Full Text]
  23. Wang AM, Doyle MV, Mark DF. Quantitation of mRNA by the polymerase chain reaction. Proc Natl Acad Sci U S A. 1989;86:9717-9721. [Abstract/Free Full Text]
  24. Ross R. The pathogenesis of atherosclerosis: an update. N Engl J Med. 1986;314:488-500. [Medline] [Order article via Infotrieve]
  25. Hart CE, Bowen-Pope DF. Platelet-derived growth factor receptor: current views of the two-subunit model. J Invest Dermatol. 1990;94(suppl 6):53S-57S.
  26. Barrett TB, Benditt EP. Platelet-derived growth factor gene expression in human atherosclerotic plaques and normal artery wall. Proc Natl Acad Sci U S A. 1988;85:2810-2814. [Abstract/Free Full Text]
  27. Libby P, Warner SJ, Salomon RN, Birinyi LK. Production of platelet-derived growth factor-like mitogen by smooth-muscle cells from human atheroma. N Engl J Med. 1988;318:1493-1498. [Abstract]
  28. Wilcox JN, Smith KM, Williams LT, Schwartz SM, Gordon D. Platelet-derived growth factor mRNA detection in human atherosclerotic plaques by in situ hybridization. J Clin Invest. 1988;82:1134-1143.
  29. Rekhter MD, Gordon D. Does platelet-derived growth factor-A chain stimulate proliferation of arterial mesenchymal cells in human atherosclerotic plaques? Circ Res. 1994;75:410-417. [Abstract/Free Full Text]
  30. Battegay E, Raines E, Seifert RA, Bowen-Pope DF, Ross R. TGF-beta induces bimodal proliferation of connective tissue cells via complex control of an autocrine PDGF loop. Cell. 1990;63:515-524. [Medline] [Order article via Infotrieve]
  31. Stouffer GA, Sarembock IJ, McNamara C, Gimple LW, Owens GK. Thrombin-induced mitogenesis of vascular SMC is partially mediated by autocrine production of PDGF-AA. Am J Physiol. 1993;265:C806-C811. [Abstract/Free Full Text]
  32. Majesky MW, Reidy MA, Bowen-Pope DF, Wilcox JN, Schwartz SM. Platelet-derived growth factor (PDGF) ligand and receptor gene expression during repair of arterial injury. J Cell Biol. 1990;111:2149-2158. [Abstract/Free Full Text]
  33. Jawien A, Bowen-Pope DF, Lindner V, Schwartz SM, Clowes AW. Platelet-derived growth factor promotes smooth muscle migration and intimal thickening in a rat model of balloon angioplasty. J Clin Invest. 1992;89:507-511.
  34. Ferns GAA, Raines EW, Sprugel KH, Motani AS, Reidy MA, Ross R. Inhibition of neointimal smooth muscle accumulation after angioplasty by an antibody to PDGF. Science. 1991;253:1129-1132. [Abstract/Free Full Text]
  35. Geary RL, Kohler TR, Vergel S, Kirkman TR, Clowes AW. Time course of flow-induced smooth muscle cell proliferation and intimal thickening in endothelialized baboon vascular grafts. Circ Res. 1994;74:14-23. [Abstract/Free Full Text]
  36. Kraiss LW, Raines EW, Wilcox JN, Seifert RA, Barrett TB, Kirkman TR, Hart CE, Bowen-Pope DF, Ross R, Clowes AW. Regional expression of the platelet-derived growth factor and its receptors in a primate graft model of vessel wall assembly. J Clin Invest. 1993;92:338-348.
  37. Golden MA, Au YP, Kirkman TR, Wilcox JN, Raines EW, Ross R, Clowes AW. Platelet-derived growth factor activity and mRNA expression in healing vascular grafts in baboons: association in vivo of platelet-derived growth factor mRNA and protein with cellular proliferation. J Clin Invest. 1991;87:406-414.
  38. Geary RL, Kraiss LW, Au YP, Vergel S, Mattsson E, Ferguson MS, Alpers CE, Clowes AW. Reduced shear induces platelet-derived growth factor-A (PDGF-A) expression in a baboon model of graft intimal thickening. FASEB J. 1994;8:A392. Abstract.
  39. Stouffer GA, Shimizu RT, Turla MB, Owens GK. ANG II- or AVP-induced increases in protein synthesis are not dependent on autocrine secretion of PDGF-AA. Am J Physiol. 1993;264:C390-C395. [Abstract/Free Full Text]
  40. Berk BC, Rao GN. Angiotensin II-induced vascular smooth muscle cell hypertrophy: PDGF A-chain mediates the increase in cell size. J Cell Physiol. 1993;154:368-380. [Medline] [Order article via Infotrieve]
  41. Amento EP, Ehsani N, Palmer H, Libby P. Cytokines and growth factors positively and negatively regulate interstitial collagen gene expression in human vascular smooth muscle cells. Arterioscler Thromb. 1991;11:1223-1230. [Abstract/Free Full Text]
  42. Negoro N, Kanayama Y, Haraguchi M, Umetani N, Nishimura M, Konishi Y, Iwai J, Okamura M, Inoue T, Takeda T. Blood pressure regulates platelet-derived growth factor A-chain gene expression in vascular smooth muscle cells in vivo: an autocrine mechanism promoting hypertensive vascular hypertrophy. J Clin Invest. 1995;95:1140-1150.
  43. Berk BC, Alexander RW, Brock TA, Gimbrone MAJ, Webb RC. Vasoconstriction: a new activity for platelet-derived growth factor. Science (Wash DC). 1986;232:87-90. [Abstract/Free Full Text]
  44. Koyama N, Morisaki N, Saito Y, Yoshida S. Regulatory effects of platelet-derived growth factor-AA homodimer on migration of vascular smooth muscle cells. J Biol Chem. 1992;267:22806-22812. [Abstract/Free Full Text]
  45. Koyama N, Hart CE, Clowes AW. Different functions of the platelet-derived growth factor-{alpha} and -ß receptors for the migration and proliferation of cultured baboon smooth muscle cells. Circ Res. 1994;75:682-691. [Abstract/Free Full Text]
  46. Ferns GA, Sprugel KH, Seifert RA, Bowen-Pope DF, Murray M, Raines EW, Ross R. Relative platelet-derived growth factor receptor subunit expression determines cell migration to different dimeric forms of PDGF. Growth Factors. 1990;3:315-324. [Medline] [Order article via Infotrieve]
  47. Barres BA, Hart IK, Coles HSR, Burne JF, Voyvodic JT, Richardson WD, Raff MC. Cell death in the oligodendrocyte lineage. J Neurobiol. 1992;23:1221-1230. [Medline] [Order article via Infotrieve]
  48. Bennett MR, Evan GI, Schwartz SM. Apoptosis of human vascular smooth muscle cells derived from normal vessels and coronary atherosclerotic plaques. J Clin Invest. 1995;95:2266-2274.
  49. Harrington EA, Bennett MR, Fanidi A, Evan GI. c-Myc-induced apoptosis in fibroblasts is inhibited by specific cytokines. EMBO J. 1994;13:3286-3295. [Medline] [Order article via Infotrieve]
  50. Cook CL, Weiser MC, Schwartz PE, Jones CL, Majack RA. Developmentally timed expression of an embryonic growth phenotype in vascular smooth muscle cells. Circ Res. 1994;74:189-196. [Abstract/Free Full Text]