Hepatocellular cancer (HCC) is a disease of poor prognosis. Identifying novel molecular aberrations might present opportunities to identify new therapeutic targets. Due to the similarities between the processes of development and cancer, we used early developing livers to identify genes that might play a primary role in HCC. Platelet-derived growth factor receptor-α (PDGFRα) was identified from microarray using early developing mouse livers. Expression of PDGFRα and its upstream effectors, PDGF-AA and PDGF-CC, were examined in HCC tissues (n = 43) by Western blot, real-time PCR, and immunohistochemistry. Finally, effect of anti-PDGFRα antibody (mAb 3G3, ImClone Systems, Inc.) was examined on human hepatoma cells. A high expression of PDGFRα was observed during early liver development. HCCs (17 of 21) revealed cytoplasmic PDGFRα and activated PDGFRα (phospho-Tyr754) by immunohistochemistry. Additional HCCs (14 of 22) showed elevated PDGFRα levels when compared with the adjacent normal livers by Western blots. Of these 14 patients, 3 showed increased PDGFRα gene expression, 3 showed elevated PDGF-AA, and 4 had higher PDGF-CC levels in the tumors compared with adjacent livers. Multiple hepatoma cell lines, when treated with mAb 3G3, showed significant decreases in cell proliferation and survival (P < 0.05). In conclusion, ∼70% of HCC tissues had elevated PDGFRα levels due to diverse mechanisms. PDGFRα inhibition in hepatoma cells led to diminution of tumor cell survival and proliferation and thus might be of therapeutic significance. [Mol Cancer Ther 2007;6(7):1932–41]

Hepatocellular cancer (HCC) remains a disease of grim prognosis due to poorly understood molecular mechanism. Owing to commonalities between these processes, cancer has also been referred to as recapitulation of development in adults (1). The oncofetal protein α-fetoprotein exists during liver development and reappears only in HCC in adults (2). The proto-oncogene c-kit, a class III receptor tyrosine kinase, was identified during embryonic development and several cancers (3, 4). We investigated early liver development and identified high expression of platelet-derived growth factor receptor-α (PDGFRα), another member of the class III receptor tyrosine kinase family (5).

PDGFRs are related to the CFS-1 receptor/c-fms and the stem cell growth factor/c-kit proto-oncogene family (6). PDGFRs play many critical roles in embryonic and postnatal development (7). In addition, overproduction of PDGFs may be involved in growth stimulation of human tumors (8). In fact, high expression of PDGFRα is detectable in a variety of tumors, such as the basal cell carcinoma, brain tumor, gastrointestinal stromal tumor, prostatic intraepithelial neoplasia, ovarian cancer, osteosarcoma, and leukemia (3, 912). PDGFRβ, an isomer of PDGFR, appears later in liver development and plays a role in hepatic fibrosis (13, 14). PDGFRα undergoes homodimerization of two α-chains following ligand (PDGF-AA, PDGF-AB, or PDGF-CC binding; ref. 15). PDGFs also play key roles during tissue development and maintenance (16). Specifically, PDGF-CC is known to exhibit greater mitogenic potency than PDGF-AA and comparable or greater mitogenic activity than PDGF-AB and PDGF-BB on several mesenchymal cells (17). Our present study used a novel strategy based on the premise that molecular basis of development and cancer cross paths, which led us to identify PDGFRα as a promising new therapeutic target in HCC. In addition, with the availability of new PDGFRα-specific antibody, this finding might have strong clinical implications (18).

Animals

ICR strain mice were obtained from Charles River Laboratories for studies approved by the Institutional Animal Use and Care Committee at the University of Pittsburgh School of Medicine and the NIH.

Collection of Embryos or Tissues

The embryos were obtained from pregnant mice at E10, E11, E12, E14, E15, E16, E17, E18, and E19 stages. For immunohistochemistry, the isolated embryos (E10–E15) or livers (E16–E19) were fixed in 10% buffered formaldehyde. Livers were also isolated from newborn pups, 2-day-old mice, 35-day-old mice, and adult mice. For RNA and protein isolation, fresh livers from same stage were pooled (n ≥ 4).

Cell Lines and Primary Hepatocyte Cultures

The human hepatoma cell lines, Hep3B, SK-Hep1, HepG2, and Snu449, and mouse hepatoma cells, Hepa 1 to Hepa 6 cells, were purchased from the American Type Culture Collection and grown according to their guidelines. The rat hepatoma cell lines (JM1 and JM2) were cultured as described elsewhere (19). Rat hepatocytes were isolated from male Fisher rats (250 g) by collagenase perfusion as described by Berry and Friend and Seglen with some modifications (20, 21).

HCC Tissue

HCC tissues were used from 43 patients (22 frozen tumors with adjacent livers and 21 paraffin blocks) and obtained under an approved Institutional Review Board protocol (exempt) from the Tissue Bank at the University of Pittsburgh, Medical Centre, Pittsburgh and Cooperative Human Tissue Network, Eastern Division, Philadelphia. Comprehensive patient data are given in Supplementary Data.5

5

Supplementary material for this article is available at Molecular Cancer Therapeutics Online (http://mct.aacrjournals.org/).

Immunohistochemistry

Sections from the mouse embryos were subjected to immunohistochemistry for PDGFRα to determine stage-specific expression. Twenty-one paraffin sections from resected HCC were obtained from Department of Pathology, University of Pittsburgh (Supplementary Data).5 Immunohistochemistry was done by an indirect immunoperoxidase method as described previously (22, 23). Primary antibodies are discussed later. All secondary antibodies were horseradish peroxidase conjugated (Chemicon). Slides were viewed on a Zeiss upright research microscope (Axioskop 40) and digital images were obtained using a Nikon Coolpix 4500 camera. Collages were prepared using the Adobe Photoshop 7.01 software.

RNA Isolation and Affymetrix Gene Expression

Fresh pooled livers from E11, E14, E17, and E18 and adult mice were used for isolating and purifying RNA by Qiagen RNeasy kit (Qiagen) and examined by Affymetrix gene expression analysis (Affymetrix) as described elsewhere (22). The Chips were scanned in an HP ChipScanner (Affymetrix) and analysis was done with Affymetrix Microarray Suite 5.0 software. The data were examined for changes in albumin, α-fetoprotein (AFP), c-kit, and Pdfgra expression.

Protein Extraction and Western Blots

Whole-cell lysates were prepared from pooled livers (n ≤ 4) from same stages of mice development. Additionally, whole-cell lysates were prepared from 22 frozen human HCC tissue and adjacent or true normal tissue. Protein extraction in radioimmunoprecipitation assay buffer and Western blot analysis has been described elsewhere (22). Blots were visualized by autoradiography. Primary antibodies used were against PDGFRα, phospho-Tyr754-PDGFRα (p-PDGFRα), PDGFRβ, PDGF-AA, and PDGF-CC (Santa Cruz Biotechnology); PDGFRα (Upstate Biotechnology); and β-actin (Chemicon). Specificity of PDGFRα antibody was confirmed by application of the blocking peptide (Santa Cruz Biotechnology).

Extraction of RNA, Reverse Transcription, and Real-time PCR

Total RNA was extracted from HCC (n = 22) and adjacent normal tissues (10–50 mg) using 1 mL Trizol reagent (Invitrogen). Following RNase-free DNase (Promega) treatment, reverse transcription was done using the First-Strand cDNA Synthesis kit (Fermentas). Assays-on-Demand gene expression products Hs00183486_m1 and Hs99999903_m1 (Applied Biosystems) were used to detect PDGFRα and β-actin, respectively. Real-time PCR was done according to Komoroski et al. (24). The relative cDNA content was determined from standard curves constructed from serially diluted cDNA, normalized to β-actin in each sample.

Proliferation Assay

Human hepatoma cell lines, Hep3B, HepG2, Snu449, and SK-Hep1, and mouse hepatoma cell line, Hepa 1 to Hepa 6 tumor cells, were grown to ∼40% confluence. All cultures were done in triplicates for statistical analysis. Cells were treated once with 100 nmol/L monoclonal antibody (mAb) against PDGFRα (3G3, ImClone Systems; ref. 18). One additional treatment was done after 24 h at which time radiolabeled thymidine was added for 24 h and the incorporated thymidine was assayed as described elsewhere (25). The average counts for each cell line (in triplicates) were normalized to the respective untreated controls. The differences were compared for statistical significance by the Student's t test and P < 0.05 was considered significant. The data were presented as a bar graph depicting fold decrease in thymidine incorporation in response to 3G3 treatment compared with untreated controls (KaleidaGraph, Synergy software).

Terminal Deoxynucleotidyl Transferase–Mediated Nick-End Labeling Assay

Human hepatoma cell lines, Hep3B, HepG2, Snu449, and SK-Hep1, and mouse hepatoma cell line, Hepa 1 to Hepa 6 tumor cells, were grown on coverslips in six-well plates to ∼60% confluence. All cultures were done in triplicates for statistical analysis. Cells were treated once with 100 nmol/L 3G3 mAb against PDGFRα. An additional treatment was done after 24 h at which time cells were fixed on coverslips in 4% formaldehyde in PBS. The apoptotic assay was done using the DeadEnd Fluorometric TUNEL system (Promega) according to the manufacturer's instructions. The apoptotic cells were observed by green fluorescence after the cells were counterstained with 4′,6-diamidino-2-phenylindole. The numbers of positive cells were counted in three fields for each cell line (done in triplicates) in the presence and absence of 3G3. The average number of terminal deoxynucleotidyl transferase–mediated nick-end labeling–positive cells for each cell line was normalized to the respective untreated cells. Statistical analysis was done using the Student's t test and the P < 0.05 was considered significant. The data were presented as a bar graph depicting fold change in the numbers of terminal deoxynucleotidyl transferase–mediated nick-end labeling–positive cells normalized to controls by using the KaliedaGraph software.

Statistical Analysis

For immunohistochemistry, staining was arbitrarily scored from + to +++, based on the intensity and number of positive cells, by a single pathologist. Mean scores from controls and tumors were compared for statistical significance by two-tailed Student's t test (InStat) and P > 0.05 was considered significant. The paired t test was used to compare the mean scores of the protein expressions in patients with liver cirrhosis SigmaPlot software (Sigma). All graphs were plotted using the KaleidaGraph or Microsoft Excel software.

For Western blots, the autoradiographs were scanned and subjected to densitometric analysis with NIH Imager software. The mean integrated absorbance was compared between the tumors and controls for statistical significance by the Student's t test or Mann-Whitney test by SigmaPlot software and a two-tailed P < 0.05 was considered significant.

For real-time PCR, due to increased scatter in the observed events, statistical analysis was done using the unpaired Student's t test as well as Mann-Whitney test by SigmaPlot software. P < 0.05 was considered significant.

Peak PDGFRα Expression during Early Mouse Liver Development

Based on the similarities between the ongoing events, such as regulated proliferation, differentiation, and survival, between the processes of development and cancer, we examined early liver development to identify signaling pathways that may be playing a role in HCC (22). Using Affymetrix gene array, we identified high expression of PDGFRα (Pdgfra) at E11 and E14 compared with all later stages including adult liver. This expression of PDGFRα was 37- and 7-fold higher at these stages than adult liver (Fig. 1A). Expression profiles of previously characterized genes in liver development validated our microarray results by serving as internal controls. C-kit, a known hepatic stem cell marker, was highest at E11 (Fig. 1A; ref. 26). The Afp gene is expressed in embryonic and fetal hepatocytes and was significantly higher in E11 to E18 stages than adults (Fig. 1A; ref. 27). The albumin (Alb1) gene expression, a known marker of differentiated hepatocytes, increased with development (Fig. 1A).

Figure 1.

PDGFRα mRNA and protein expression during normal liver development in mice and hepatoma cells. A, results from microarray analysis of pooled multiple livers at E11, E14, E17, and E18, and adult stages. E11 and E14 showed high PDGFRα (Pdgfra) expression (signal), which were 37- and 7-fold higher than adults, respectively. Internal controls, decreasing c-kit expression, high prenatal Afp expression, and loss in adults and increasing Alb1 expression, validated the study. B, PDGFRα localization during mice liver development and in adult liver (Ad). Cytoplasmic or perinuclear PDGFRα is seen in epithelial cells (arrows) at E10, E12, E14, E16, and E18 stages shown at higher magnification (right). Magnification, ×60. Hematopoietic cells show nuclear localization at these times (arrowheads). Adult normal mouse liver (NML) hepatocytes show faint membranous and cytoplasmic localization (arrow). Normal human liver shows faint membranous localization at hepatocytes (arrowhead). C, representative Western blot shows PDGFRα protein expression during normal mouse liver development (top). The protein expression at E12 and E14 was higher than later stages as shown by densitometry (bottom). Equal loading confirmed by β-actin (data not shown). D, normalized integrated absorbance (to normal rat liver) from three representative Western blots was used to plot a graph that showed 3- to 6-fold and significant (P < 0.05) increase in PDGFRα protein expression in hepatoma cells (top). Densitometric analysis from a representative Western blot showed 5.8- to 65-fold increase in activated or p-PDGFRα-Tyr754 in all hepatoma cell lines when compared with NRL (bottom). Equal loading for Western blots was confirmed by β-actin (data not shown). Hepatoma cell lines: M1 and JM2 (rat) and Hep3B, HepG2, and SK-Hep1 (human). NRL, normal rat liver; Hep-culture, hepatocyte culture.

Figure 1.

PDGFRα mRNA and protein expression during normal liver development in mice and hepatoma cells. A, results from microarray analysis of pooled multiple livers at E11, E14, E17, and E18, and adult stages. E11 and E14 showed high PDGFRα (Pdgfra) expression (signal), which were 37- and 7-fold higher than adults, respectively. Internal controls, decreasing c-kit expression, high prenatal Afp expression, and loss in adults and increasing Alb1 expression, validated the study. B, PDGFRα localization during mice liver development and in adult liver (Ad). Cytoplasmic or perinuclear PDGFRα is seen in epithelial cells (arrows) at E10, E12, E14, E16, and E18 stages shown at higher magnification (right). Magnification, ×60. Hematopoietic cells show nuclear localization at these times (arrowheads). Adult normal mouse liver (NML) hepatocytes show faint membranous and cytoplasmic localization (arrow). Normal human liver shows faint membranous localization at hepatocytes (arrowhead). C, representative Western blot shows PDGFRα protein expression during normal mouse liver development (top). The protein expression at E12 and E14 was higher than later stages as shown by densitometry (bottom). Equal loading confirmed by β-actin (data not shown). D, normalized integrated absorbance (to normal rat liver) from three representative Western blots was used to plot a graph that showed 3- to 6-fold and significant (P < 0.05) increase in PDGFRα protein expression in hepatoma cells (top). Densitometric analysis from a representative Western blot showed 5.8- to 65-fold increase in activated or p-PDGFRα-Tyr754 in all hepatoma cell lines when compared with NRL (bottom). Equal loading for Western blots was confirmed by β-actin (data not shown). Hepatoma cell lines: M1 and JM2 (rat) and Hep3B, HepG2, and SK-Hep1 (human). NRL, normal rat liver; Hep-culture, hepatocyte culture.

Close modal

Immunohistochemistry showed cytoplasmic, membranous, and perinuclear localization of PDGFRα in several but not all hepatoblasts and hepatocytes at E10 to E16 stages, with a gradual decrease in staining intensity (Fig. 1B). Some smaller hematopoietic cells were showing nuclear and perinuclear PDGFRα localization at these stages, as well (Fig. 1B). At E18, most hepatocytes showed faint membranous or cytoplasmic localization of PDGFRα, whereas a subset of hematopoietic cells were exhibiting nuclear or perinuclear localization for this protein. Adult mouse as well as donor human liver showed faint membranous localization of PDGFRα (Fig. 1B).

A more accurate and quantitative validation of Pdgfra expression came from Western blot analysis that showed elevated PDGFRα at E12 and E14 followed by a sequential decrease and reappearance in newborn and adult livers (Fig. 1C). Densitometry verified ∼5-fold decrease in PDGFRα protein levels from E12 to adult liver (Fig. 1C).

Increased PDGFRα Activation in Hepatoma Cell Lines

Protein extracts from various hepatoma cell lines were used to determine PDGFRα protein levels (Fig. 1D). Significantly higher levels of PDGFRα were evident in rat and human hepatoma cells compared with normal liver or adult hepatocytes in culture (P < 0.01). Additional analysis revealed increased levels of activated PDGFRα (Tyr754; p-PDGFR), ranging from 5.8- to 65-fold in the hepatoma cells compared with hepatocytes, confirming the importance of PDGFRα in hepatoma cells (Fig. 1D).

Elevated PDGFRα in Predominant HCC Tissues from Patients

Whole-cell lysates from HCC tissues and adjacent livers (n = 22; Supplementary Data)5 were examined by Western blot analysis for PDGFRα protein (Fig. 2A). Overall, >63% of HCC tissues showed up to 7-fold increase in total PDGFRα levels compared with adjacent controls. Whereas 3 of 22 patients showed no change, 5 of 22 patients showed a decrease in PDGFRα (Table 1). Densitometry on Western blots identified ∼3.3-fold increase in total PDGFRα protein expression in tumors compared with their adjacent controls (Fig. 2B), which was statistically significant (P < 0.01).

Figure 2.

PDGFRα and PDGFRβ protein expression in HCC tissues from patients. A, Western blots showing differences in PDGFRα and PDGFRβ protein expression in tumors (T) compared with adjacent normal liver tissue (N) in all patients (n = 22). Equal loading confirmed by actin; for specificity of the antibody, Western blot on the lysates from NIH3T3 cells grown in 10% serum was incubated in primary antibody in presence (−) or absence (+) of respective blocking peptides. B, densitometric analysis from tumors with increased PDGFRα protein (14 of 22) revealed a 3.3-fold and significant (P < 0.01) increase in its expression. C, patients with cirrhosis showed a statistically significant increase in PDGFRβ protein expression when compared with PDGFRα by paired t test (P < 0.01).

Figure 2.

PDGFRα and PDGFRβ protein expression in HCC tissues from patients. A, Western blots showing differences in PDGFRα and PDGFRβ protein expression in tumors (T) compared with adjacent normal liver tissue (N) in all patients (n = 22). Equal loading confirmed by actin; for specificity of the antibody, Western blot on the lysates from NIH3T3 cells grown in 10% serum was incubated in primary antibody in presence (−) or absence (+) of respective blocking peptides. B, densitometric analysis from tumors with increased PDGFRα protein (14 of 22) revealed a 3.3-fold and significant (P < 0.01) increase in its expression. C, patients with cirrhosis showed a statistically significant increase in PDGFRβ protein expression when compared with PDGFRα by paired t test (P < 0.01).

Close modal
Table 1.

Summary of PDGFR protein expression in patient samples with corresponding results for PDGFR mRNA expression and the ligands PDGF-AA and PDGF-CC

Patient no.PDGFRα proteinPDGFRα RNA
PDGF-AA protein
PDGF-CC protein
RegulationFold changeRegulationFold changeRegulationFold change
0.03 0.6 0.66 
1.6 0.1 1.53 
0.04 0.75 0.81 
0.28 0.04 0.23 
1.2 1.83 0.55 
0.12 2.39 0.26 
NC 0.03 7.31 0.62 
1.22 NC 0.61 
0.01 0.005 0.58 
10 0.68 0.6 1.57 
11 NC 0.3 3.64 NC 0.91 
12 0.32 0.41 0.8 
13 NC 0.62 0.56 NC 1.1 
14 0.83 NC 1.1 NC 0.98 
15 0.53 0.46 0.28 
16 0.15 0.66 1.39 
17 0.0009 0.05 NC 0.97 
18 NC 0.98 1.68 NC 0.91 
19 0.77 1.33 NC 1.1 
20 1.5 3.63 1.3 
21 0.22 NC 0.97 1,3 
22 0.47 1.44 0.55 
Patient no.PDGFRα proteinPDGFRα RNA
PDGF-AA protein
PDGF-CC protein
RegulationFold changeRegulationFold changeRegulationFold change
0.03 0.6 0.66 
1.6 0.1 1.53 
0.04 0.75 0.81 
0.28 0.04 0.23 
1.2 1.83 0.55 
0.12 2.39 0.26 
NC 0.03 7.31 0.62 
1.22 NC 0.61 
0.01 0.005 0.58 
10 0.68 0.6 1.57 
11 NC 0.3 3.64 NC 0.91 
12 0.32 0.41 0.8 
13 NC 0.62 0.56 NC 1.1 
14 0.83 NC 1.1 NC 0.98 
15 0.53 0.46 0.28 
16 0.15 0.66 1.39 
17 0.0009 0.05 NC 0.97 
18 NC 0.98 1.68 NC 0.91 
19 0.77 1.33 NC 1.1 
20 1.5 3.63 1.3 
21 0.22 NC 0.97 1,3 
22 0.47 1.44 0.55 

Abbreviations: I, increase; D, decrease; NC, no change.

In addition, expression of PDGFRβ was examined in these samples (Fig. 2A). Increase in PDGFRβ was identified in only 6 of 22 tumors compared with their adjacent controls. In addition, five of six tumors had concomitant increase in PDGFRα levels. Interestingly, when overall expression of PDGFRα and PDGFRβ was compared in tumors with cirrhosis, PDGFRβ was significantly higher in these samples confirming its association with cirrhosis (P < 0.01; Fig. 2C).

Next, 21 additional HCC paraffin sections were examined for total and p-PDGFRα expression by immunohistochemistry (Fig. 3A). In normal donor livers, PDGFRα was observed at the membrane only (Fig. 3A, a and e). In addition, no p-PDGFRα was detected in normal livers (Fig. 3A, b and f). Tumors (17 of 21) showed aberrant PDGFRα protein accumulation in hepatocytes as shown in representative tumors (Fig. 3A, c and g). Simultaneously, these samples displayed cytoplasmic p-PDGFRα staining (Fig. 3A, d and h). The staining was graded based on intensity to obtain an arbitrary immunohistochemistry score as discussed in Materials and Methods (Table 2). Statistical analysis using these scores showed a significant increase in total and activated PDGFRα in tumors compared with normal tissue (P < 0.01; Fig. 3B).

Figure 3.

Elevated PDGFRα expression and activation in HCC. A, representative photomicrographs of PDGFRα staining in normal donor liver (a and e) and HCC (c and g). Representative photomicrographs of activated or p-PDGFRα (Tyr754) in normal donor livers (b and f) and HCC (d and h). B, statistical analysis from mean immunohistochemical (IHC) scores of PDGFRα and p-PDGFRα staining in normal samples (n = 5; C1–C5; Table 2) and tumor tissues with elevated protein and activation (17 of 21), showing a significant difference (P < 0.01).

Figure 3.

Elevated PDGFRα expression and activation in HCC. A, representative photomicrographs of PDGFRα staining in normal donor liver (a and e) and HCC (c and g). Representative photomicrographs of activated or p-PDGFRα (Tyr754) in normal donor livers (b and f) and HCC (d and h). B, statistical analysis from mean immunohistochemical (IHC) scores of PDGFRα and p-PDGFRα staining in normal samples (n = 5; C1–C5; Table 2) and tumor tissues with elevated protein and activation (17 of 21), showing a significant difference (P < 0.01).

Close modal
Table 2.

Summary of PDGFR protein expression in patient samples (paraffin sections)

Patient no.PDGFRα ProteinPDGFRα (Tyr745)Patient no.PDGFRα ProteinPDGFRα (Tyr745)
23 ++ ++ 36 +,++ 
24 37 ++ ++ 
25 +,++ 38 ++ ++ 
26 +,++ 39 ++ ++ 
27 ++ ++ 40 
28 ++ 41 
29 +,++ 42 
30 ++ 43 
31 C1 (adenoma) 
32 +,++ ++ C2 (adenoma) 
33 +,++ C3 (metastatic colorectal) 
34 +,++ C4 (inguinal hernia) 
35 ++ ++ C5 (adenoma) 
Patient no.PDGFRα ProteinPDGFRα (Tyr745)Patient no.PDGFRα ProteinPDGFRα (Tyr745)
23 ++ ++ 36 +,++ 
24 37 ++ ++ 
25 +,++ 38 ++ ++ 
26 +,++ 39 ++ ++ 
27 ++ ++ 40 
28 ++ 41 
29 +,++ 42 
30 ++ 43 
31 C1 (adenoma) 
32 +,++ ++ C2 (adenoma) 
33 +,++ C3 (metastatic colorectal) 
34 +,++ C4 (inguinal hernia) 
35 ++ ++ C5 (adenoma) 

NOTE: ++, moderate; +, mild; 0, no staining.

Mechanism for PDGFRα Protein Up-regulation and Activation in HCC

To address the mechanism of PDGFRα up-regulation, we examined frozen HCC and control samples for Pdgfra mRNA expression by real-time PCR. Interestingly, an overall significant decrease in Pdgfra mRNA expression was observed in HCC (n = 22) compared with controls (P < 0.05; Fig. 4A). Only 3 of the 22 tumors, which displayed elevated PDGFRα levels, showed a concomitant increase in Pdgfra expression (Table 1). The increase ranged from 1.22- to 1.6-fold and was significant as judged by Mann-Whitney test (P < 0.05).

Figure 4.

Mechanism for PDGFRα protein up-regulation and activation in HCC tissues. A, real-time PCR analysis of PDGFRα mRNA expression in tumor and adjacent normal tissue (n = 22) showed an overall decrease in tumors, which was significant (P < 0.05). B, mean integrated absorbance, from densitometric analysis of PDGF-AA and PDGF-CC Western blots using lysates from all patients, showed no significant differences in their expression levels between tumors and adjacent liver. Columns, mean integrated absorbance; bars, SE.

Figure 4.

Mechanism for PDGFRα protein up-regulation and activation in HCC tissues. A, real-time PCR analysis of PDGFRα mRNA expression in tumor and adjacent normal tissue (n = 22) showed an overall decrease in tumors, which was significant (P < 0.05). B, mean integrated absorbance, from densitometric analysis of PDGF-AA and PDGF-CC Western blots using lysates from all patients, showed no significant differences in their expression levels between tumors and adjacent liver. Columns, mean integrated absorbance; bars, SE.

Close modal

To investigate additional mechanisms of PDGFRα activation, frozen tumors and controls were examined for its known ligands, PDGF-AA and PDGF-CC (Table 1). HCC samples (8 of 22) showed an up-regulation of PDGF-AA protein, which was coexistent with elevated PDGFRα protein in three patients. In addition, out of these three tumors, one just had simultaneous increases in Pdgfra expression and PDGF-CC protein and the remaining two had a unique increase in PDGF-AA only. Cumulative analysis in 22 tumors and normal samples showed insignificant differences in PDGF-AA levels except in three samples that showed a significant increase (2.39- to 3.63-fold; P < 0.05; Fig. 4B). PDGF-CC, a selective PDGFRα ligand, was statistically increased in four HCC samples with elevated PDGFRα protein (P < 0.05; Table 1). Out of these four tumors, one showed concomitant increase in Pdgfra expression and PDGF-AA levels and one showed a concurrent increase in Pdgfra expression only. However, overall insignificant difference in PDGF-CC was observed in tumors versus normal adjacent controls (Fig. 4B).

Anti-PDGFRα mAb Inhibits Hepatoma Cell Proliferation and Induces Apoptosis

Next, we wanted to examine the effect of mAb anti-PDGFRα at 100 nmol/L PDGFRα on the growth and survival on various human hepatoma cell lines and in mouse hepatoma cells. Efficacy of mAb (3G3) to inhibit PDGFRα activity has been reported recently (18). Tumor cell lines were grown to ∼40% confluence and treated with 3G3 as described in Materials and Methods. This treatment affected proliferation of all cell lines significantly (P < 0.05) tested with a decrease in thymidine incorporation ranging from 10% to 40% compared with the untreated cells (Fig. 5A). The greatest difference was observed in the SK-Hep1 cells and the least in mouse Hepa 1 to Hepa 6 cells.

Figure 5.

Effects of mAb 3G3 on human hepatoma cell proliferation and survival. A, modest decrease in thymidine incorporation (range, 10–40%), in response to mAb 3G3 in multiple hepatoma cell lines, which was statistically significant (P < 0.05). The values are normalized to the respective untreated controls and presented as fold change. B, terminal deoxynucleotidyl transferase–mediated nick-end labeling (TUNEL) assay identified a significant increase in the number of apoptotic nuclei in response to mAb 3G3 treatment in multiple hepatoma cell lines, which ranged from 4- to 18-fold and was statistically significant (P < 0.05). The values are normalized to respective untreated controls and presented as fold change. C, representative photomicrographs show an increase in apoptotic nuclei (green) in Hep3B and SK-Hep1 cells in the presence of mAb 3G3.

Figure 5.

Effects of mAb 3G3 on human hepatoma cell proliferation and survival. A, modest decrease in thymidine incorporation (range, 10–40%), in response to mAb 3G3 in multiple hepatoma cell lines, which was statistically significant (P < 0.05). The values are normalized to the respective untreated controls and presented as fold change. B, terminal deoxynucleotidyl transferase–mediated nick-end labeling (TUNEL) assay identified a significant increase in the number of apoptotic nuclei in response to mAb 3G3 treatment in multiple hepatoma cell lines, which ranged from 4- to 18-fold and was statistically significant (P < 0.05). The values are normalized to respective untreated controls and presented as fold change. C, representative photomicrographs show an increase in apoptotic nuclei (green) in Hep3B and SK-Hep1 cells in the presence of mAb 3G3.

Close modal

Next, we examined apoptosis in these cells in response to 3G3 treatment for 2 days at 100 nmol/L as described in Materials and Methods. A significant increase in the number of apoptotic nuclei was observed in response to 3G3 treatment (P < 0.05), which ranged from 4- to 18-fold in various cell lines when normalized to the respective controls (Fig. 5B and C). The most pronounced effect was in Hep3B cells and the least in Snu449.

These results show a modest antiproliferative and a prominent antisurvival role of PDGFRα inhibition on the various hepatoma cells of human and mouse origin. These combinatorial effects show the usefulness of PDGFRα inhibition in HCC.

Despite several advances in understanding the pathogenesis of HCC, its prognosis remains ominous. Although some reports have examined the expression patterns of multiple genes in these tumors, these studies are limited by reproducibility and ability to address ‘cause or effect’ relationship of observed changes (28, 29). Although inclusion of controls and tumors at various stages of progression might assist in addressing the latter issue to an extent, the interpretation might be complicated by heterogeneity in the tumors or patient population. Such variations become apparent as differences in gene expression patterns continue to be observed among various studies (3032). Thus, these studies have only limited practicability to identify novel molecular targets for therapy.

Instead, we used a novel approach to initiate analysis in early liver development, based on similarities in the processes of development and cancer (1). Cancer has been appropriately called as recapitulation of development in adults, or development and organogenesis might represent “physiologic tumorigenesis.” Examining the molecular basis of these stages thus might reveal primary pathways of physiologic or, in cancer, more pathologic relevance and thus be more appropriately a primary aberration. This strategy is also of essence in identifying molecular basis of potential cancer stem cells, which might be similar to developmental stem cells as shown in gastrointestinal tract (33, 34). Thus, based on ongoing events at E11 and E14 in developing livers, such as regulated cell proliferation, differentiation, and survival, we examined global gene expression to identify PDGFRα as one of the highly expressed genes and proteins for its possible role in HCC (22).

Involvement of PDGF proteins in liver fibrosis and cirrhosis, especially PDGFRβ, is known (13, 35). Lately, studies are focused on ameliorating in vitro and in vivo hepatic fibrogenesis (36). Hepatic fibrosis, a sequel to infectious, inflammatory, or metabolic disorders of the liver, also predisposes to HCC (37, 38). Although a number of reports have examined the PDGF pathway in fibrosis, its role and expression in HCC remains elusive. More recently, mice overexpressing (liver specific) PDGF-CC, a ligand for PDGFRα, go on to develop cirrhosis and HCC (39). Getting cues from our developmental analysis, we expected a promising role of PDGFRα in HCC.

PDGFRα plays a role in development in many tissues, assisting in proliferation, morphogenesis, angiogenesis, or epithelial-mesenchymal interactions (40, 41). Activation of PDGFRα secondary to mutations or other causes has been reported in many cancers as well, where it is associated to tumor cell proliferation, progression, or angiogenesis (3, 9). In our study, aberrant increase and activation of PDGFRα was observed in hepatoma cell lines of human and nonhuman origin. Similarly, >70% of all HCC samples displayed a significant increase in total and p-PDGFRα, indicating aberrant activation of this pathway (n = 43). Whereas increase in PDGFRα was reported in endothelial cells in highly metastatic HCCs, we observed an increase in parenchymal liver cells by immunohistochemistry, suggesting that PDGFRα might be playing a role in events critical for tumor cells, such as proliferation and survival (42).

Analysis from frozen tissues revealed 3 out of 22 tumors with elevated PDGFRα levels, showing a significant increase in Pdgfra expression indicating a contributing mechanism for its elevated protein. None of these tumors were examined for mutations in Pdgfra gene as has been reported in gastrointestinal stromal tumor and brain tumors (3, 9, 15). Samples (11 of 14) showed increased PDGFRα protein despite decreased mRNA, reflecting possible posttranscriptional modification of mRNA to increase half-life or a posttranslational event affecting its degradation. Both forms of regulation are known for PDGFRα (43, 44). PDGFRα up-regulation or constitutive activation has been reported as a consequence of fusion, chromosomal translocations, and up-regulation of ligands leading to autocrine or paracrine stimulation (45). PDGF-AA and PDGF-CC are known ligands for this receptor. Tumors (three of eight) showed concomitant elevation in PDGF-AA and PDGFRα proteins. Tumors (four of five) showed simultaneous increase in PDGF-CC and PDGFRα proteins. These observations reveal heterogeneity in the mechanisms of activation of PDGFRα in these tumors. Additionally, this analysis supports the presence of autocrine or paracrine loops, which might be contributing to the disease process. Also of relevance is the recently reported study where PDGF-CC overexpression in hepatocytes leads to the development of HCC (39).

Because PDGF signaling has been associated to hepatic fibrosis and cirrhosis, we investigated any correlation of the observed PDGFRα up-regulation in HCC to cirrhosis. Six of 14 tumors that showed elevated PDGFRα had coexisting cirrhosis and 8 of 14 tumors had no cirrhosis. Out of the eight tumors without cirrhosis but increased PDGFRα levels, three were fibrolamellar carcinomas. In fact, three of four fibrolamellar HCCs presented up-regulated PDGFRα protein without any evidence of cirrhosis, whereas one showed no change. Additionally, out of the 8 of 22 frozen tumors that showed either decrease or no change in PDGFRα protein, 4 had cirrhosis. Thus, any correlation between PDGFRα up-regulation and cirrhosis was coincidental. On the other hand, PDGFRβ protein showed significant expression in cirrhotic livers (P < 0.05), an observation well described previously (13, 46, 47).

Recent therapies against cancers are based on identification of unique molecular aberrations and have shown varying degree of success (48). The presence of PDGFRα, as well as activated PDGFRα in HCC in >70% of patients and its relationship to proliferation in liver development and other tumors, justifies its consideration as a valid therapeutic target in HCC. We used a mAb against PDGFRα to test its effect on multiple human hepatoma cells, which showed aberrant PDGFRα activation. This antibody has been characterized recently as a selective PDGFRα inhibitor (18). This resulted in a modest and significant decrease in the tumor cell proliferation. Additionally, there was a marked increase in tumor cell death in response to PDGFRα antibody. Because high PDGFRα expression was evident in the tumor core compared with the adjacent liver, it suggested a role of this protein in events associated with advanced stages of tumor development, such as tumor cell survival rather than proliferation, which might be critical for early tumors or cells at the tumor margin. Indeed, PDGFRα inhibition dramatically affected hepatoma cell survival, only modestly reducing their proliferation. Nonetheless, this combined effect on tumor cell proliferation and their viability indicates PDGFRα as a promising therapeutic target in HCC. Additional studies examining more cell lines as well as primary cultures from resected human tumors to further evaluate the efficacy of targeting PDGFRα are under way. Interestingly, there are several existing case reports that have shown a favorable therapeutic response to tyrosine kinase inhibitors that also possess anti-PDGFRα in both HCCs and cholangiocarcinomas (49, 50). In addition, targeting PDGF in HCC has also been suggested in transforming growth factor β–mediated tumor progression in hepatocytes (51). As mentioned previously, a recent study reporting generation and characterization of PDGF-CC transgenic mice also supports the role of aberrant PDGFRα activation in HCC and validates its therapeutic targeting in this tumor (39).

Grant support: American Cancer Society grant RSG-03-141-01-CNE and NIH grant 1RO1DK62277 (S.P.S. Monga). Cleveland Foundation and Rango's Fund for Enhancement of Pathology Research.

Note: P. Stock and D. Monga contributed equally to this work.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1
Kho AT, Zhao Q, Cai Z, et al. Conserved mechanisms across development and tumorigenesis revealed by a mouse development perspective of human cancers.
Genes Dev
2004
;
18
:
629
–40.
2
Wu AH, Sell S. Markers for hepatocellular carcinoma.
Immunol Ser
1990
;
53
:
403
–22.
3
Heinrich MC, Corless CL, Duensing A, et al. PDGFRA activating mutations in gastrointestinal stromal tumors.
Science
2003
;
299
:
708
–10.
4
Thorgeirsson SS, Evarts RP, Bisgaard HC, Fujio K, Hu Z. Hepatic stem cell compartment: activation and lineage commitment.
Proc Soc Exp Biol Med
1993
;
204
:
253
–60.
5
Claesson-Welsh L. PDGF receptors: structure and mechanism of action. Basel (Switzerland): Karger; 1993.
6
Westermark B, Heldin CH. Platelet-derived growth factor. Structure, function, and implications in normal and malignant cell growth.
Acta Oncol
1993
;
32
:
101
–5.
7
Betsholtz C. Insight into the physiological functions of PDGF through genetic studies in mice.
Cytokine Growth Factor Rev
2004
;
15
:
215
–28.
8
Heldin CH, Westermark B. Mechanism of action and in vivo role of platelet-derived growth factor.
Physiol Rev
1999
;
79
:
1283
–316.
9
Clarke ID, Dirks PB. A human brain tumor-derived PDGFR-α deletion mutant is transforming.
Oncogene
2003
;
22
:
722
–33.
10
Dabrow MB, Francesco MR, McBrearty FX, Caradonna S. The effects of platelet-derived growth factor and receptor on normal and neoplastic human ovarian surface epithelium.
Gynecol Oncol
1998
;
71
:
29
–37.
11
Fudge K, Bostwick DG, Stearns ME. Platelet-derived growth factor A and B chains and the α and β receptors in prostatic intraepithelial neoplasia.
Prostate
1996
;
29
:
282
–6.
12
Sulzbacher I, Birner P, Trieb K, Traxler M, Lang S, Chott A. Expression of platelet-derived growth factor-AA is associated with tumor progression in osteosarcoma.
Mod Pathol
2003
;
16
:
66
–71.
13
Friedman SL. Closing in on the signals of hepatic fibrosis.
Gastroenterology
1997
;
112
:
1406
–9.
14
Ikura Y, Morimoto H, Ogami M, Jomura H, Ikeoka N, Sakurai M. Expression of platelet-derived growth factor and its receptor in livers of patients with chronic liver disease.
J Gastroenterol
1997
;
32
:
496
–501.
15
Fletcher JA. Role of KIT and platelet-derived growth factor receptors as oncoproteins.
Semin Oncol
2004
;
31
:
4
–11.
16
Betsholtz C. Role of platelet-derived growth factors in mouse development.
Int J Dev Biol
1995
;
39
:
817
–25.
17
Gilbertson DG, Duff ME, West JW, et al. Platelet-derived growth factor C (PDGF-C), a novel growth factor that binds to PDGF α and β receptor.
J Biol Chem
2001
;
276
:
27406
–14.
18
Loizos N, Xu Y, Huber J, et al. Targeting the platelet-derived growth factor receptor α with a neutralizing human monoclonal antibody inhibits the growth of tumor xenografts: implications as a potential therapeutic target.
Mol Cancer Ther
2005
;
4
:
369
–79.
19
Novicki DL, Jirtle RL, Michalopoulos G. Establishment of two rat hepatoma cell strains produced by a carcinogen initiation, phenobarbital promotion protocol.
In Vitro
1983
;
19
:
191
–202.
20
Berry MN, Friend DS. High-yield preparation of isolated rat liver parenchymal cells: a biochemical and fine structural study.
J Cell Biol
1969
;
43
:
506
–20.
21
Seglen PO. Preparation of isolated rat liver cells.
Methods Cell Biol
1976
;
13
:
29
–83.
22
Micsenyi A, Tan X, Sneddon T, Luo JH, Michalopoulos GK, Monga SP. β-Catenin is temporally regulated during normal liver development.
Gastroenterology
2004
;
126
:
1134
–46.
23
Monga SP, Monga HK, Tan X, Mule K, Pediaditakis P, Michalopoulos GK. β-catenin antisense studies in embryonic liver cultures: role in proliferation, apoptosis, and lineage specification.
Gastroenterology
2003
;
124
:
202
–16.
24
Komoroski BJ, Zhang S, Cai H, et al. Induction and inhibition of cytochromes P450 by the St. John's wort constituent hyperforin in human hepatocyte cultures.
Drug Metab Dispos
2004
;
32
:
512
–8.
25
Gebhardt R, Cruise J, Houck KA, et al. Differential effect of growth factors on growth stimulation and phenotypic stability of glutamine-synthetase-positive and -negative hepatocytes in primary culture.
Differentiation
1986
;
33
:
45
–55.
26
Omori M, Omori N, Evarts RP, Teramoto T, Thorgeirsson SS. Coexpression of flt-3 ligand/flt-3 and SCF/c-kit signal transduction system in bile-duct-ligated SI and W mice.
Am J Pathol
1997
;
150
:
1179
–87.
27
Shiojiri N, Lemire JM, Fausto N. Cell lineages and oval cell progenitors in rat liver development.
Cancer Res
1991
;
51
:
2611
–20.
28
Kurokawa Y, Matoba R, Takemasa I, et al. Molecular features of non-B, non-C hepatocellular carcinoma: a PCR-array gene expression profiling study.
J Hepatol
2003
;
39
:
1004
–12.
29
Tackels-Horne D, Goodman MD, Williams AJ, et al. Identification of differentially expressed genes in hepatocellular carcinoma and metastatic liver tumors by oligonucleotide expression profiling.
Cancer
2001
;
92
:
395
–405.
30
Garber ME, Troyanskaya OG, Schluens K, et al. Diversity of gene expression in adenocarcinoma of the lung.
Proc Natl Acad Sci U S A
2001
;
98
:
13784
–9.
31
Luo JH, Ren B, Keryanov S, et al. Transcriptomic and genomic analysis of human hepatocellular carcinomas and hepatoblastomas.
Hepatology
2006
;
44
:
1012
–24.
32
Nimgaonkar A, Sanoudou D, Butte AJ, et al. Reproducibility of gene expression across generations of Affymetrix microarrays.
BMC Bioinformatics
2003
;
4
:
27
.
33
Brittan M, Wright NA. Stem cell in gastrointestinal structure and neoplastic development.
Gut
2004
;
53
:
899
–910.
34
Sancho E, Batlle E, Clevers H. Signaling pathways in intestinal development and cancer.
Annu Rev Cell Dev Biol
2004
;
20
:
695
–723.
35
Pinzani M, Milani S, Grappone C, Weber FL, Jr., Gentilini P, Abboud HE. Expression of platelet-derived growth factor in a model of acute liver injury.
Hepatology
1994
;
19
:
701
–7.
36
Yoshiji H, Noguchi R, Kuriyama S, et al. Imatinib mesylate (STI571: Gleevec) attenuates liver fibrosis development in rats.
Am J Physiol Gastrointest Liver Physiol
2005
;
288
:
G907
–13.
37
Alter HJ, Seeff LB. Recovery, persistence, and sequelae in hepatitis C virus infection: a perspective on long-term outcome.
Semin Liver Dis
2000
;
20
:
17
–35.
38
Caldwell SH, Crespo DM, Kang HS, Al-Osaimi AM. Obesity and hepatocellular carcinoma.
Gastroenterology
2004
;
127
:
S97
–103.
39
Campbell JS, Hughes SD, Gilbertson DG, et al. Platelet-derived growth factor C induces liver fibrosis, steatosis, and hepatocellular carcinoma.
Proc Natl Acad Sci U S A
2005
;
102
:
3389
–94.
40
Liu L, Chong SW, Balasubramaniyan NV, Korzh V, Ge R. Platelet-derived growth factor receptor α (pdgfr-α) gene in zebrafish embryonic development.
Mech Dev
2002
;
116
:
227
–30.
41
Utoh R, Shigenaga S, Watanabe Y, Yoshizato K. Platelet-derived growth factor signaling as a cue of the epithelial-mesenchymal interaction required for anuran skin metamorphosis.
Dev Dyn
2003
;
227
:
157
–69.
42
Zhang T, Sun HC, Xu Y, et al. Overexpression of platelet-derived growth factor receptor α in endothelial cells of hepatocellular carcinoma associated with high metastatic potential.
Clin Cancer Res
2005
;
11
:
8557
–63.
43
Langerak AW, van der Linden-van Beurden CA, Versnel MA. Regulation of differential expression of platelet-derived growth factor α- and β-receptor mRNA in normal and malignant human mesothelial cell lines.
Biochim Biophys Acta
1996
;
1305
:
63
–70.
44
Rosenkranz S, Ikuno Y, Leong FL, et al. Src family kinases negatively regulate platelet-derived growth factor α receptor-dependent signaling and disease progression.
J Biol Chem
2000
;
275
:
9620
–7.
45
Jones AV, Cross NC. Oncogenic derivatives of platelet-derived growth factor receptors.
Cell Mol Life Sci
2004
;
61
:
2912
–23.
46
Davis BH, Rapp UR, Davidson NO. Retinoic acid and transforming growth factor β differentially inhibit platelet-derived-growth-factor-induced Ito-cell activation.
Biochem J
1991
;
278
:
43
–7.
47
Pinzani M, Milani S, Herbst H, et al. Expression of platelet-derived growth factor and its receptors in normal human liver and during active hepatic fibrogenesis.
Am J Pathol
1996
;
148
:
785
–800.
48
Griffin J. The biology of signal transduction inhibition: basic science to novel therapies.
Semin Oncol
2001
;
28
:
3
–8.
49
Holcombe RF, Gu M, Imagawa D, Milovanovic T. Expression of Kit and platelet-derived growth factor receptors α and β in cholangiocarcinoma, and case report of therapy with imatinib mesylate (STI571).
Anticancer Drugs
2003
;
14
:
651
–7.
50
Ramadori G, Fuzesi L, Grabbe E, Pieler T, Armbrust T. Successful treatment of hepatocellular carcinoma with the tyrosine kinase inhibitor imatinib in a patient with liver cirrhosis.
Anticancer Drugs
2004
;
15
:
405
–9.
51
Gotzmann J, Fischer AN, Zojer M, et al. A crucial function of PDGF in TGF-β-mediated cancer progression of hepatocytes.
Oncogene
2006
;
25
:
3170
–85.