Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online before print June 23, 2003, 10.1148/radiol.2282020907
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2282020907v1
228/2/560    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
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 reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Guccione, S.
Right arrow Articles by Bednarski, M. D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Guccione, S.
Right arrow Articles by Bednarski, M. D.
(Radiology 2003;228:560-568.)
© RSNA, 2003

Technical Developments

Functional Genomics Guided with MR Imaging: Mouse Tumor Model Study1

Samira Guccione, PhD, Yi-Shan Yang, PhD, Gongyi Shi, PhD, Daniel Y. Lee, MD, PhD, King C. P. Li, MD, FRCPC2 and Mark D. Bednarski, PhD

1 From the Lucas MR Spectroscopy and Imaging Research Center (S.G., Y.S.Y., G.S., D.Y.L., M.D.B.) and Department of Radiology (S.G., Y.S.Y., G.S., D.Y.L., K.C.P.L., M.D.B.), Stanford University School of Medicine, 1201 Welch Rd, P260, Stanford, CA 94305-5488. Received July 31, 2002; revision requested September 7; revision received October 11; accepted December 10. Supported in part by the Lucas Foundation, the Phil Allen Trust, and National Institutes of Health grant P41 RR09784. Y.S.Y. and S.G. supported by a research fellowship of the National Cancer Institute at the Lucas Center, Stanford University. Address correspondence to M.D.B. (e-mail: mark@s-word.stanford.edu).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 Materials and Methods
Results
Discussion
 REFERENCES
 
To gain a better understanding of gene expression patterns in tumors, the authors used contrast material–enhanced magnetic resonance (MR) imaging to noninvasively characterize regions within the same tumor to provide a correlate for genomic analysis. Gene expression profiles of samples from a mouse tumor model obtained from contrast-enhanced and nonenhanced regions within the same tumor were compared with MR imaging and functional genomics. From these samples, 11,000 genes were analyzed: 10 genes were up-regulated in the contrast-enhanced areas, and one gene was up-regulated in the nonenhanced regions. Several of these genes encode extracellular matrix proteins. Findings in this study demonstrate that MR imaging can serve as a powerful noninvasive tool for characterizing different regions of tumors to guide genomic analysis with high spatial and temporal resolution.

© RSNA, 2003

Index terms: Genes and genetics • Magnetic resonance (MR), experimental studies, **.12143


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
Materials and Methods
Results
Discussion
 REFERENCES
 
To understand the molecular pathophysiology of cancer, genomic expression profiles of tumors are being correlated with clinical presentation patterns, surrogate disease markers, and pathologic evaluations (1–9). Solid tumors, however, are morphologically heterogeneous and display different characteristic phenotypes during disease progression (10–12). Many of these features can be detected with magnetic resonance (MR) imaging, a powerful noninvasive clinical diagnostic tool (13). Syngeneic squamous cell carcinoma VII is a tumor model for human head and neck cancers that exhibits differential imaging characteristics that are similar to those of human solid tumors (14). Enhancement with the clinical MR imaging contrast agent gadopentetate dimeglumine (Magnevist; Berlex Laboratories, Wayne, NJ) can reveal imaging features associated with increased vascular permeability and vessel density that are factors implicated with angiogenesis and prognosis of disease (15–17). We hypothesized that gene expression analysis of the contrast material–enhanced areas of solid tumors relative to areas that do not take up contrast material (nonenhanced areas) within the same tumor can reveal important molecular targets for the development of new therapeutic approaches. Thus, the purpose of our study was to use MR imaging to characterize spatial and temporal tumor heterogeneity to guide tissue sampling for gene expression studies. We propose that MR imaging is a functional tool that can facilitate genomic and proteomic analysis in the identification of targets for molecular diagnosis and drug discovery.


   Materials and Methods
TOP
 ABSTRACT
 INTRODUCTION
 Materials and Methods
Results
Discussion
 REFERENCES
 
Animal Model
Approval for the animal protocol used in this study was obtained from the Stanford University Institutional Animal Care and Use Committee. Mouse squamous cell carcinoma VII cells were transplanted into C3H/Km mice (age range, 10–12 weeks; weight range, 25–30 g). Maintenance and characteristics of this tumor cell line are described elsewhere (14). Eighteen mice were used for each stage of tumor growth in the study, for a total of 54 mice. The stages were homogeneous enhancement, heterogeneous enhancement, and heterogeneous enhancement with changes in the T2-weighted MR images (see the Image Analysis section for details). In each mouse, 2 x 105 cells in 0.05 mL of Hanks solution were implanted intradermally into the left flank. Animals were given standard rodent diet and reverse osmosis water ad libitum before and during the time of tumor growth. The sizes of the tumors were measured twice a week to monitor their growth. Imaging of the tumors was started when they reached a diameter of approximately 5–10 mm at about 1–2 weeks after tumor implantation and was continued twice a week thereafter to monitor the MR contrast-enhancement pattern.

The animals were sacrificed once the tumors reached the stage of progression suitable for further analysis on the basis of the MR contrast-enhancement pattern (see the Image Analysis section for details) or once the tumor began to affect the mobility of the animals. Regions at the periphery and center of the tumor were marked with dissecting pins on the basis of the identifiable anatomic landmarks on the MR images. The tumors were removed, and these regions were sampled at biopsy. Then, the specimens were frozen rapidly in liquid nitrogen. The volume of these samples from the periphery and center of the tumors ranged from approximately 50 to 200 mm3 depending on the size of the tumor at different stages of tumor progression. These samples were processed for gene expression analysis, reverse transcriptase–polymerase chain reaction, immunohistochemistry, in situ hybridization, Western blot analysis, and two-dimensional gel electrophoresis.

The animals were chosen for each technique on the basis of the following criteria: Tumors with similar T1- and T2-weighted MR image patterns were grouped together (see the Image Analysis section for details). Tissue samples from different tumors in the same group were used for these techniques. The sample sizes for each of the protocols were determined on the basis of previous literature precedents (18–22).

MR Imaging
A clinical 3.0-T MR imager (Signa Horizon LX8.2; GE Medical Systems, Milwaukee, Wis) was used with a custom-designed quadrature high-pass birdcage coil tuned to 127 MHz for signal reception. Fifty-four animals were imaged. The following MR imaging protocols were performed in the transverse plane with a field of view of 8 cm, 256 x 192-pixel matrix, section thickness of 1.5 mm, and one signal acquired: (a) precontrast T1-weighted spin echo (repetition time msec/echo time msec of 300/13), (b) precontrast T2-weighted fast spin echo (4,000/85, echo train length of eight), and (c) postcontrast T1-weighted spin echo (300/13) at 2 minutes after injection of gadopentetate dimeglumine (200 µL of 0.5 mol/L solution administered via a tail vein at a rate of 40 µL/sec with no flushing solution). Postprocessing was performed with the software package in the MR imager.

Image Analysis
Images were reviewed by two clinicians (K.C.P.L., D.Y.L.). The reviewers were blinded to information about the stage of tumor progression. The animals were then categorized into three groups on the basis of the MR images: group 1 (18 mice) had homogeneously enhanced postcontrast T1-weighted MR images; group 2 (18 mice) had heterogeneously enhanced postcontrast T1-weighted MR images; and group 3 (18 mice) had heterogeneously enhanced postcontrast T1-weighted MR images, with elevated signal intensity in the T2-weighted MR images that corresponded to the contrast-enhanced areas in the T1-weighted images. The relative tumor pixel intensities of the center and peripheral areas on the T2-weighted images were analyzed (MRVISION, version 1.5.4b; MRVision, Menlo Park, Calif). For groups 1 and 2, the mean relative pixel intensity was 1.0 ± 0.1 (SD). However, the center areas with elevated signal intensity in the more advanced tumors on heterogeneous T2-weighted MR images had higher pixel intensities relative to those in the peripheral region of the tumor (1.6 ± 0.1) (group 3). The former tumors were excluded from the current study.

Hematoxylin-Eosin Staining
Two mice from each group were used for hematoxylin-eosin staining. Tissue samples were preserved in 10% formalin solution for 24–48 hours. Then they were embedded in paraffin, sectioned, stained with hematoxylin-eosin, and mounted on glass slides. Morphologic features used to identify viable nuclei included the absence of overall shrinkage and homogenous dark basophilia. The slides were reviewed by a veterinary pathologist at Stanford University Medical School.

Oligonucleotide Microarray Analysis
Four mice were used from each group for microarray analysis. Total RNA was isolated (GStract Total RNA Isolation Kit; Maxim Biotech, South San Francisco, Calif). The procedure involves (a) homogenizing tissue or cells, (b) adding salt solution to precipitate the DNA and RNA, (c) extracting the RNA from the precipitate by using phenol chloroform, and (d) washing the RNA pellet to eliminate all contaminants. The polyadenylated messenger RNA (mRNA) was then purified from the total RNA (Oligotex mRNA Kit; Qiagen, Valencia, Calif). The purified mRNA was handled for analysis by using the GeneChip Expression Analysis protocol (Affymetrix, Santa Clara, Calif): (a) The preparation was targeted by synthesizing double-stranded complementary DNA (cDNA) from purified mRNA by using reverse transcriptase. (b) The purified cDNA was used to produce biotin-labeled complementary RNA (cRNA) by means of in vitro transcription reaction. (c) The target biotin-labeled cRNA was fragmented and hybridized with the control cRNA to the oligonucleotide probes on the probe array for 16 hours at 45°C. (d) The probe array was washed and stained with streptavidin phycoerythrin conjugate immediately after hybridization. (e) The probe array was scanned with a confocal microscope scanner (Hewlett-Packard, Palo Alto, Calif) at the excitation wavelength of 488 nm. The amount of light emitted at 570 nm was proportional to that of the bound target at each location on the probe array. Algorithms (GeneChip Expression Analysis; Affymetrix) were used to analyze data generated from the oligonucleotide microarrays (Mu11subA and Mu11subB; Affymetrix) (total of 11,000 genes) to determine the intensity of expressed genes. Data were analyzed (Microarray Suite, version 4.0, Affymetrix; Gene Spring, version 4.0, Silicon Genetics, Redwood City, Calif).

Quantitative Reverse Transcriptase–Polymerase Chain Reaction Analysis
With three mice from each group, quantitative reverse transcriptase–polymerase chain reaction was performed according to standard procedures. RNA was extracted from tumor tissues with the following steps: Tissue was homogenized in 5 mL of lysis buffer (6 mol/L urea, 3 mol/L lithium chloride, 50 mmol/L sodium acetate, 200 µg/mL heparin, and 0.1% sodium dodecyl sulfate). The homogenized tissue was centrifuged at a rate of 13,100 rpm (16,000 g) for 20 minutes, extracted twice with an equal volume of phenol and chloroform, and precipitated with ethanol. The RNA pellet was air dried and dissolved in water treated with diethylpyrocarbonate. Reverse transcriptase–polymerase chain reaction was performed (DNA Thermal Cycler 480; Perkin-Elmer; Oak Brook, Ill). The chain reaction involved incubation of the samples at (a) 42°C for 15 minutes, 99°C for 5 minutes, and 5°C for 5 minutes for one cycle; (b) 95°C for 45 seconds and 60°C for 45 seconds for 35 cycles; and (c) 72°C for 7 minutes for one cycle. The products were checked in 2% agarose gel, along with the 100–base pair ladder (Promega, Madison, Wis).

Statistical Analysis
Statistical analysis was performed by means of the Student t test. Differences with a P value of less than .05 were considered statistically significant. This analysis was performed by comparing the relative gene expression levels (fold change [Table 1]) of genes in the homogeneously enhanced tumors with those same genes in the heterogeneously enhanced tumors.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Gene Expression Profile for Squamous Cell Carcinoma VII Mouse Tumors

 
In Situ Hybridization
Two mice were used from each group for in situ hybridization analysis. In situ hybridization was performed according to standard protocols (InnoGenex, San Ramon, Calif). Formalin-fixed paraffin-embedded tissue sections were dewaxed and rehydrated in ribonuclease-free condition. Washed slides were placed in ribonuclease–free water for 3–5 minutes. Proteinase K was administered in drops to cover the sections, and the slides were incubated for 5–10 minutes. Next, the slides were washed in ribonuclease-free water; then 1% formaldehyde in phosphate-buffered saline was added to cover the sections. The slides were incubated for 10 minutes and then washed. The probe for the matrix Gla protein (MGP) gene was constructed by using standard protocol, and it was hybridized onto the slides at 37°C for 1 hour. The primary and secondary antibodies were linked after the streptavidin-peroxide reagent and 3-amino,9-ethyl-carbazole chromogen were added. Excess liquid was washed out by using buffer solution. The tissue sections were then stained with hematoxylin and eosin and mounted. The slides were reviewed by the veterinary pathologist. An increased level of staining was scored as indicative of the presence of MGP transcript.

Two-dimensional Electrophoresis Argorase Gel and Sequencing
Three mice from each group were used for two-dimensional gel electrophoresis and sequencing. Gel electrophoresis was performed with standard protocol (23): Cytosolic fractions of the periphery (contrast-enhanced) and center (nonenhanced) regions of the tumor lysate (250 µg) were subjected to isoelectric focusing on impedance plethysmographic strips (isoelectric point at pH of 4–7); they were separated by 4%–20% at polyacrylamide gel electrophoresis with sodium dodecyl sulfate. The gels were then stained with silver. The band of interest (marked by arrow on the right side in Fig 1, B) was digested with trypsin. The resulting fragments were analyzed by means of matrix-assisted laser desorption ionization time-of-flight mass spectrometry to determine their molecular weights. These fragments were identified as MGP (Profound Peptide Mapping software, version 4.10.5; Rockefeller University, New York, NY).



View larger version (64K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1. Validation of protein expression of MGP in heterogeneously contrast-enhanced tumor. A, Photomicrographs depict in situ hybridization, or ISH, of a homogeneously contrast-enhanced tumor. The expression level of MGP (red stain) was distinctly higher in the peripheral region, in which chromogen can be observed, compared with that in the center of the tumor. (Red chromogen stain; original magnification, x40.) B, Photographs of two-dimensional agarose gel electrophoresis of tumor samples obtained from the contrast-enhanced and nonenhanced regions. The band indicated by the arrow was removed and fragmented by means of trypsin digestion. The resulting fragments were analyzed; their mean masses were 1,344.49, 1,555.7, 1,792.01, 1,835.08, and 2,462.78. These fragments were identified as mouse MGP.

 
Immunohistochemistry Testing
Two mice were used from each group for immunohistochemistry testing, which was performed with the standard method. Approximately 5-µm-thick tissue sections were deparaffinized, rehydrated, incubated with fresh 3% hydrogen peroxide in methanol for 10 minutes, and then washed with phosphate-buffered saline solution. Normal rabbit serum (5%) was applied for 30 minutes and removed by means of blotting. Sections were then incubated with biotinylated monoclonal platelet-derived growth factor (PDGF)-receptor antibody (BD Bioscience, San Diego, Calif) for 60 minutes at room temperature, washed three times in buffer, and incubated with biotinylated secondary antibody for 20 minutes. After they were washed, the sections were incubated with avidin-biotin complex reagent (Vector Laboratories, Burlingame, Calif) for 20 minutes, followed by repeated washing. Sections were then incubated with diaminobenzodine in 0.03% hydrogen peroxide for 10 minutes, washed, counterstained with hematoxylin-eosin, rinsed, and mounted. The slides were reviewed by the veterinary pathologist. An increased level of staining was scored as indicative of the presence of PDGF-receptor protein.

Western Blot Analysis
Two mice from each group were used for Western blot analysis, which was performed according to standard protocols (24). Membrane fractions of the peripheral and center regions of the tumor lysate (50 µg) were mixed with sample-loading buffer, heated for 5 minutes at 95°C in 50 mmol/L dithiothreitol, separated at polyacrylamide gel electrophoresis with sodium dodecyl sulfate, and transferred to nitrocellulose. The nitrocellulose filters were blocked with 5% reconstituted dry milk and incubated with an anti–PDGF-receptor {alpha} chain antibody. Bound antibodies were visualized by using a chemiluminescence detection kit (Amersham Biosciences, England). The quantity of the expression level was determined by means of densitometric analysis.


   Results
TOP
 ABSTRACT
 INTRODUCTION
 Materials and Methods
Results
Discussion
 REFERENCES
 
MR Imaging
We observed three dramatically different imaging patterns as the tumor progressed. Figure 2, A–C, shows the imaging characteristics at the early stages of tumor growth, at approximately 1–2 weeks after tumor implantation. Precontrast (Fig 2, A, B) and postcontrast (Fig 2, C) MR images show tumors with a homogeneous pattern. Figure 2, D–F, shows representative MR images, obtained approximately 2–4 weeks after tumor implantation, for the next stage of tumor progression. A heterogeneous pattern is seen after administration of contrast material (Fig 2, F); the contrast-enhanced and nonenhanced areas are easily visible and can be spatially demarcated for biopsy or tissue sampling after tumor resection. Similar to the homogeneous tumors, no changes were observed on the T2-weighted images (compare Fig 2, B and E; this finding indicates that no necrosis is present at this stage of tumor growth).



View larger version (98K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2. Three-tesla MR images obtained in C3H/K mice (age range, 8-10 weeks; weight range, 25-30 g) with an implanted subcutaneous squamous cell carcinoma VII tumor. The following images were obtained in a homogeneously enhanced tumor: A, Precontrast T1-weighted image. B, Precontrast T2-weighted image. C, Contrast-enhanced T1-weighted image. The following images were obtained in a heterogeneously enhanced tumor: D, Precontrast T1-weighted image. E, Precontrast T2-weighted image. F, Contrast-enhanced T1-weighted image. The following images were obtained in a heterogeneously enhanced tumor with necrosis: G, Precontrast T1-weighted image. H, Precontrast T2-weighted image. I, Contrast-enhanced T1-weighted image. All contrast-enhanced images were obtained 2 minutes after injection of the contrast agent. Contrast enhancement was observed in the periphery of the heterogeneously enhanced tumors (F), but no differential contrast enhancement was observed in the homogeneously enhanced tumors (C). In C and F, thick arrows indicate central region of tumor, and thin arrows indicate periphery of tumor. Tissue was removed from these areas for genomic analysis. In H and I, dotted arrows indicate necrosis. T1-weighted images were obtained with a spin-echo pulse sequence (300/13, one signal acquired, 256 x 192-pixel matrix, field of view of 8 cm, section thickness of 1.5 mm). T2-weighted images were obtained with a fast spin-echo pulse sequence (4,000/85, one signal acquired, echo train length of eight, 256 x 192 matrix, field of view of 8 cm, section thickness of 1.5 mm).

 
As the tumors progressed further, at approximately 4–5 weeks after tumor implantation, they exhibited focal changes on T2-weighted MR images (compare Fig 2, H [dotted arrow] and E). These areas also correspond to regions in the tumor where low signal intensity was observed after administration of contrast material (compare Fig 2, H and I). These patterns (high signal intensity on T2-weighted MR images and corresponding low signal intensity on T1-weighted MR images) are characteristic of focal necrotic areas within individual tumors. This stage of tumor growth, which includes development of necrotic tissues, is difficult to analyze with microarray analysis because the techniques that rely on mRNA production, such as the functional genomic analysis used in this study, are valid only when the tissue is viable. In addition, necrotic areas are not representative of the associated matrix with viable cells, which is useful in target identification (Yang YS, written communication, 2002).

Hematoxylin-Eosin Staining
With hematoxylin-eosin staining, we observed no distinguishable difference between the periphery and center regions of the homogeneously enhanced tumors (Fig 3, A, B) or between the contrast-enhanced and nonenhanced regions of the heterogeneously enhanced tumors (Fig 3, C, D). At this stage of tumor growth, neither region showed any necrosis or substantial inflammatory infiltrates (Fig 2, C), as was reflected in the T2-weighted MR image (Fig 2, E). In contrast, hematoxylin-eosin staining for the nonenhanced region (Fig 3, F) in heterogeneously enhanced tumor showed a different pattern compared with those in the enhanced area (Fig 3, E). Significant generalized necrosis is evident with the presence of inflammatory infiltrates and karyorrhexis.



View larger version (96K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3. Photomicrographs depict protein expression of MGP. In homogeneously enhanced tumor: A, Contrast-enhanced area. B, Nonenhanced area. In heterogeneously enhanced tumor: C, Contrast-enhanced area. D, Nonenhanced area. In heterogeneously enhanced tumor with changes on T2-weighted MR images: E, Contrast-enhanced area. F, Nonenhanced area. In A-E, these regions of tumor are composed of sheets of moderately pleomorphic large cells with eccentrically placed nuclei and abundant eosinophilic cytoplasm with distinct cell borders. In F, this region of tumor shows overall shrinkage of cells and absence of nucleus. (Hematoxylin-eosin stain; original magnification, x20.)

 
Oligonucleotide Microarray Analysis
With oligonucleotide microarray analysis, we analyzed four syngeneic squamous cell carcinoma VII tumors that exhibited heterogeneous contrast-enhancement patterns on T1-weighted MR images (Table 1). Significant up-regulation (mean of twofold or more) was observed in 10 genes compared with samples obtained from the contrast-enhanced and nonenhanced areas. Six of these 10 genes encoded for extracellular matrix proteins (Table 1, entries 1–6). These proteins are MGP (25,26), fisp-12 (27), CTLA-2-{alpha} cysteine protease (28), extracellular matrix-associated protein Sc1 (29), insulin-like growth factor binding protein-3 (30), and apolipoprotein E (31). Three additional genes ({alpha}-globin, ß-globin complex, and ß-1-globin) are associated with globin synthesis and hypoxia (Table 1, entries 7–9) (32,33), and one gene is associated with vasculogenesis (PDGF receptor: Table 1, entry 10) (34). We observed one elevated mRNA level in the nonenhanced region that corresponded to mast cell protease-12 (Table 1, entry 11) (35). Interestingly, many genes such as fibronectin were highly up-regulated in the tumor but were also spatially homogeneous throughout the mass (Table 1, entry 12, compare mean intensity, contrast enhanced [3,822] and nonenhanced [3,639] with fold change [x1.08]) (36).

To ensure that differences in gene expression profiles are related to the changes we observed at MR imaging, we also analyzed four tumors that were homogeneously enhanced in the postcontrast MR images. In this case, tissue samples were taken from the periphery and center of the tumor, similar to the regions sampled at biopsy in the heterogeneously enhanced tumors (Fig 2, C, thin and thick arrows); the samples were analyzed in the same way as were the heterogeneous tumors. The only gene that was significantly up-regulated (more than twofold) in the homogeneously enhanced tumors was {alpha}-actin (x3.2 ± 0.4).

Statistical Analysis
A Student t test was performed to compare the mean fold change in gene expression profile between the homogeneously and heterogeneously enhanced tumor sets (Table 2). The results showed a statistically significant difference (P < .05) for MGP, PDGF receptor, and apolipoprotein E. No statistical significance was observed between the mean fold change of fibronectin in the homogeneously and heterogeneously enhanced tumors (P = .483).


View this table:
[in this window]
[in a new window]

  		TABLE 2. Validation of Selected Genes for Squamous Cell Carcinoma Mouse Tumors

 
Quantitative Reverse Transcriptase–Polymerase Chain Reaction
We performed quantitative reverse transcriptase–polymerase chain reaction with the same samples that were analyzed statistically from the heterogeneously enhanced tumors (Table 2). Results showed that the three genes that were demonstrated to have statistically significant differences in gene expression between the two distinct imaging patterns were also up-regulated more than twofold. As observed in the microarray analysis, no difference in gene expression was observed for fibronectin between the contrast-enhanced and nonenhanced regions of the tumors.

Target Validation
To further confirm the heterogeneous spatial distribution of these genes, we performed in situ hybridization studies, immunohistochemical analysis, two-dimensional gel electrophoresis, protein sequencing, and Western blot analysis on MGP and PDGF receptor.

MGP: In Situ Hybridization, Two-dimensional Gel Electrophoresis, and Protein Sequence
We performed in situ hybridization for MGP, which showed the highest difference in mRNA expression in both microarray and quantitative reverse transcriptase–polymerase chain reaction analysis. As demonstrated in Figure 1, A, the periphery of the tumor showed intense chromogen staining compared with that in the center of the tumor, which reflects a higher degree of probe hybridization in the periphery. This pattern of mRNA up-regulation in the tumor periphery was consistent with findings on the corresponding contrast–enhanced MR image (Fig 2, F). Next, protein purification and sequence identification was performed for MGP by means of mass spectroscopy. This approach was used as opposed to a Western blot analysis because MGP may be distinctly altered by {gamma}-carboxylation, which makes antibody-based detection less reliable. These forms of MGP were resolved at two-dimensional gel electrophoresis (Fig 1, B). A series of bands at 11,000 Da were observed from three distinct tissue samples obtained from the contrast-enhanced region of the tumors. These bands represented similar molecular weights but different isoelectric points consistent with the expected carboxylation state of MGP (Fig 1, B, right). The bands and patterns were distinctly different for the tissue obtained from the nonenhanced region in the same tumor (Fig 1, B, left). The most intense band present in the enhanced region of the tumor (arrow on Fig 1, B, left) was isolated and digested with trypsin. The resulting fragments were analyzed by means of matrix-assisted laser desorption ionization time-of-flight mass spectrometry and were identified as mouse MGP.

PDGF Receptor: Immunohisto-chemical and Western Blot Analysis
For PDGF receptor, we performed immunohistochemical analysis by means of a biotinylated monoclonal antibody (BD Bioscience) on tissue sections. In the heterogeneously enhanced tumors, the results showed a distinctly darker stain in the periphery of the tumor compared with that in the center, which represented chromogen deposition related to the presence of PDGF-receptor {alpha} (Fig 4, A). The staining pattern corresponded precisely with areas of contrast enhancement seen at MR imaging (Fig 2, F). To further confirm that the differences in mRNA expression in the contrast-enhanced regions of the tumor are related to protein expression, Western blot analysis was performed to determine the level of protein expression. Membrane fractions of the peripheral and central regions of the tumor were analyzed by means of immunoblot analysis with the anti–PDGF-receptor {alpha} chain antibody. Results of densitometric analysis showed that PDGF receptor was up-regulated in the peripheral region by x2.7 ± 0.2 (n = 3) (Fig 4, B).



View larger version (62K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4. Validation of protein expression of PDGF receptor precursor in a heterogeneously contrast-enhanced tumor. A, Immunohistochemical analysis. The stain intensity was substantially higher in the periphery compared with that in the center region of the tumor. Immunohistochemical analysis performed on a homogeneously contrast-enhanced tumor showed no difference in stain intensity between the periphery and center regions of the tumor. (Brown chromogen stain; original magnification, x40.) B, Bar graph depicts results of quantification of Western blot analysis with individual densitometric readings from the contrast-enhanced and nonenhanced regions. Membrane fractions of the periphery and center regions of the tumor lysate (50 µg) were also analyzed by means of immunoblot analysis by using the monoclonal antibody against PDGF-receptor {alpha} chain. Densitometric analysis showed that PDGF receptor was up-regulated in the peripheral region (x2.7 ± 0.2 [n = 3]).

 

   Discussion
TOP
 ABSTRACT
 INTRODUCTION
 Materials and Methods
Results
Discussion
REFERENCES
 
We found that contrast-enhanced MR imaging can serve as a tool for characterizing different regions of tumors for genomic analysis. We observed differences in spatial and temporal resolution over the tumor by correlating MR imaging findings with changes in gene expression profiles. Oligonucleotide microarray analysis allowed examination of a large number of genes at the same time, which helped us understand the expression behavior of the genes in a particular system regardless of their function or role in cancer development or other biologic processes. Thus, the analysis is an efficient and nonbiased way to discover potential targets for imaging diagnosis and therapeutic treatment. In the present study, we demonstrated that regions of increased contrast enhancement seem to correlate best with changes in the mRNA of proteins that are strongly associated with the extracellular matrix. The expression level of selected genes—including MGP, apolipoprotein E, PDGF receptor, and fibronectin–were further validated with independent methods because these genes encode for extracellular matrix proteins and have a potential role in tumor development and angiogenesis.

MGP is an extracellular protein that is dependent on vitamin-K {gamma}-carboxylation. Findings in previous studies identified MGP in metastatic breast cancer cell lines (26); therefore, it may be involved in the process of metastasis in cancer (22). The overproduction of PDGF is associated with autocrine and paracrine growth stimulation of human tumors, and activation of their corresponding receptors may be critically important in tumor progression (34). Both PDGF and its receptors are implicated in the process of angiogenesis and neovascular maturation (34). Fibronectin, an extracellular matrix protein, is a ligand for the integrin family of receptors and functions in the mediation of cell adhesion (36). Evidence from both genomic analysis and immunohistologic examination show correlations between fibronectin, tumorigenesis, and metastasis (22). Apolipoprotein E is believed to be synthesized and secreted by astrocytes and has been of great interest as a participant in the pathogenesis of Alzheimer disease. This gene is expressed in a variety of tumors, but its role in tumor biology has not been defined (31).

Quantitative reverse transcriptase– polymerase chain reaction was performed for these four genes, and the results were consistent with the results at oligonucleotide microarray analysis. MGP, apolipoprotein E, and PDGF receptor all had a higher level of expression in the contrast-enhanced area compared with the nonenhanced area, while fibronectin showed a similar expression level in both regions. Thus, we conclude that the findings at quantitative reverse transcriptase–polymerase chain reaction supported the microarray data: The levels of MGP, PDGF receptor, and apolipoprotein E are associated with the differential MR contrast-enhancement pattern.

In addition, the up-regulation of MGP in the contrast-enhanced regions of the tumor was confirmed with in situ hybridization, two-dimensional electrophoresis, and protein sequencing. On the basis of these results, we conclude that MGP was expressed differentially and was present mostly in the contrast-enhanced region of the tumor. Immunohistochemical examination and Western blot analysis were performed to demonstrate the up-regulation of PDGF receptor in the contrast-enhanced area of the tumor. We conclude that the level of PDGF receptor {alpha} was more than twofold the protein level in the contrast-enhanced region relative to the nonenhanced region in the heterogeneously enhanced tumors.

Thus, we hypothesize that strongly contrast-enhanced regions in tumors are associated with changes in the extracellular matrix (the "milieu") within the environment of the tumor. This matrix seems to contain factors that increase permeability to small molecules, such as gadopentetate dimeglumine, or has properties to retain these molecules long enough for them to be observed with standard clinical imaging sequences. This milieu is probably a combination of proteins from both the tumor cells themselves and the cells associated with the extracellular matrix. In the present study, we showed that the mRNA of proteins associated with metastatic cell lines and tumors, such as MGP and PDGF receptor, are up-regulated in the contrast-enhanced areas. Thus, the environment to which tumor cells are exposed may be as important as the cell taxonomy when tumors are staged.

New imaging techniques can also provide discrimination of tissue samples for genomic analysis. With dynamic contrast-enhanced MR imaging, the kinetics of contrast enhancement can be studied to provide more parameters for characterizing tissues in vivo (13). Similarly, MR imaging can be performed with tissue-specific contrast agents and molecular imaging techniques that provide even more information about the tissue of interest (37,38). In addition, clinical imaging modalities—such as computed tomography (CT), positron emission tomography, and single photon emission CT—can provide new ways of correlating imaging findings with genomic analysis findings. In addition, optical imaging methods combined with laser-capture microdissection can provide discrimination to the cellular level of samples obtained at image-guided biopsy (39).

In conclusion, we demonstrated that contrast-enhanced MR imaging can serve as a noninvasive tool for characterizing different regions of tumors and can provide correlates for genomic analysis with high spatial and temporal resolution. The differential expression of genes and proteins related to the heterogeneity within solid tumors as depicted at MR imaging has important consequences. First, an in vivo evaluation with noninvasive imaging of the tumor in its unperturbed state can guide tissue sampling to areas of the tumor with distinct pathophysiology. Second, by comparing the differences in gene and protein expression between spatially distinct regions, important targets can be identified that may otherwise be overlooked. Finally, since the clinical imaging protocols used in this study are implemented routinely in cancer patients, they can be easily added as a parameter to annotate gene expression analysis. MR imaging can be used as a noninvasive tool to characterize the distinct pathophysiology within cancers for genomic and proteomic analysis.


   ACKNOWLEDGMENTS
 
We thank Shoucheng Ning, MD, PhD, in the laboratory of Susan Knox, MD, PhD, Department of Radiation Oncology, Stanford University Medical Center, for expert help in tumor cell line preparation.


   FOOTNOTES
 
2 Current address: Warren Grant Magnuson Clinical Center, National Institutes of Health, Bethesda, Md. Back

3 **. Multiple body systems Back

Abbreviations: MGP = matrix Gla protein, mRNA = messenger RNA, PDGF = platelet-derived growth factor

Author contributions: Guarantors of integrity of entire study, K.C.P.L., M.D.B.; study concepts, K.C.P.L., M.D.B.; study design, K.C.P.L., M.D.B., S.G., Y.S.Y.; literature research, S.G., Y.S.Y.; experimental studies, S.G., Y.S.Y., G.S.; data acquisition and analysis/interpretation, S.G., Y.S.Y., G.S.; statistical analysis, S.G., Y.S.Y.; manuscript preparation, S.G., Y.S.Y., M.D.B.; manuscript definition of intellectual content, K.C.P.L., M.D.B., S.G., Y.S.Y., D.Y.L.; manuscript editing, all authors; manuscript revision/review, K.C.P.L., M.D.B., S.G., Y.S.Y., D.Y.L., G.S.; manuscript final version approval, S.G., Y.S.Y., M.D.B.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 Materials and Methods
Results
Discussion
 REFERENCES
 

  1. Bittner M, Meltzer P, Chen Y, et al. Molecular classification of cutaneous malignant melanoma by gene expression profiling. Nature 2000; 406:536-540.[CrossRef][Medline]
  2. Perou CM, Sorlie T, Eisen MB, et al. Molecular portraits of human breast tumours. Nature 2000; 406:747-752.[CrossRef][Medline]
  3. Alizadeh AA, Eisen MB, Davis RE, et al. Distinct types of diffuse large B-cell lymphoma identified by gene expression profiling. Nature 2000; 403:503-511.[CrossRef][Medline]
  4. Golub TR, Slonim DK, Tamayo P, et al. Molecular classification of cancer: class discovery and class prediction by gene expression monitoring. Science 1999; 286:531-537.[Abstract/Free Full Text]
  5. Scherf U, Ross DT, Waltham M, et al. A gene expression database for the molecular pharmacology of cancer. Nat Genet 2000; 24:236-244.[CrossRef][Medline]
  6. Khan J, Simon R, Bittner M, et al. Gene expression profiling of alveolar rhabdomyosarcoma with cDNA microarrays. Cancer Res 1998; 58:5009-5013.[Abstract/Free Full Text]
  7. Coller HA, Grandori C, Tamayo P, et al. Expression analysis with oligonucleotide microarrays reveals that MYC regulates genes involved in growth, cell cycle, signaling, and adhesion. Proc Natl Acad Sci U S A 2000; 97:3260-3265.[Abstract/Free Full Text]
  8. Pinkel D. Cancer cells, chemotherapy and gene clusters. Nat Genet 2000; 24:208-209.[CrossRef][Medline]
  9. Duggan DJ, Bittner M, Chen Y, Meltzer P, Trent JM. Expression profiling using cDNA microarrays. Nat Genet 1999; 21:10-14.[CrossRef][Medline]
  10. Wilkinson N, Rollason TP. Recent advances in the pathology of smooth muscle tumours of the uterus. Histopathology 2001; 39:331-341.[CrossRef][Medline]
  11. Ross JS, Sheehan CE, Dolen EM, Kallakury BVS. Morphologic and molecular prognostic markers in prostate cancer. Adv Anat Pathol 2002; 9:115-128.[CrossRef][Medline]
  12. Kuhl CK, Schild HH. Dynamic image interpretation of MRI of the breast. J Magn Reson Imaging 2000; 12:965-974.[CrossRef][Medline]
  13. Taylor JS, Tofts PS, Port R, et al. MR Imaging of tumor microcirculation: promise for the new millennium. Magn Reson Imaging 1999; 10:903-907.
  14. Ning S, Yu N, Brown DM, Kanekal S, Knox SJ. Radiosensitization by intratumoral administration of cisplatin in a sustained-release drug delivery system. Radiother Oncol 1999; 50:215-223.[CrossRef][Medline]
  15. Kashanian FK, Goldstein HA, Blumetti RF, Holyoak WL, Hugo FP, Dolker M. Rapid bolus injection of gadopentetate dimeglumine: absence of side-effects in normal volunteers. AJNR Am J Neuroradiol 1990; 11:853-856.[Abstract]
  16. Hudgins PA, Elster AD, Runge VM, Kashanian FK, Goldstein HA. Efficacy and safety of gadopentetate dimeglumine in the evaluation of patients with a suspected tumor of the extracranial head and neck. J Magn Reson Imaging 1993; 3:345-349.[Medline]
  17. Hawighorst H, Knapstein PG, Knopp MV, et al. Uterine cervical carcinoma: comparison of standard and pharmacokinetic analysis of time-intensity curves for assessment of tumor angiogenesis and patient survival. Cancer Res 1998; 58:3598-3602.[Abstract/Free Full Text]
  18. Markert JM, Fuller CM, Gillespie GY, et al. Differential gene expression profiling in human brain tumors. Physiol Genomics 2001; 5:21-33.[Abstract/Free Full Text]
  19. Ljubimova JY, Lakhter AJ, Loksh A, et al. Overexpression of a4 chain-containing laminins in human glial tumors identified by gene microarray analysis. Cancer Res 2001; 61:5601-5610.[Abstract/Free Full Text]
  20. Mariani L, Beaudry C, McDonough WS, et al. Glioma cell motility is associated with reduced transcription of proapoptotic and proliferation genes: a cDNA microarray analysis. J Neurooncol 2001; 53:161-176.[CrossRef][Medline]
  21. Huang H, Colella S, Kurrer M, Yonekawa Y, Kleihues P, Ohgaki H. Gene expression profiling of low-grade diffuse astrocytomas by cDNA arrays. Cancer Res 2000; 60:6868-6874.[Abstract/Free Full Text]
  22. Clark EA, Golub TR, Lander ES, Hynes RO. Genomic analysis of metastasis reveals an essential role for RhoC. Nature 2000; 406:532-535.[CrossRef][Medline]
  23. Wilkins MR, Williams KL, Appel RD, Hochstrasser DF, eds. Proteome research: new frontiers in functional genomics Berlin, Germany: Springer-Verlag, 1997.
  24. Fruttiger M, Calver AR, Kruger WH, et al. PDGF mediates a neuron-astrocyte interaction in the developing retina. Neuron 1996; 17:1117-1131.[CrossRef][Medline]
  25. Price PA, Fraser JD, Metz-Virca G. Molecular cloning of matrix Gla protein: implications for substrate recognition by vitamin K–dependent gamma-carboxylase. Proc Natl Acad Sci U S A 1987; 84:8335-8339.[Abstract/Free Full Text]
  26. Chen L, O’Bryan JP, Smith HS, Liu E. Overexpression of matrix Gla protein mRNA in malignant human breast cells: isolation by differential cDNA hybridization. Oncogene 1990; 5:1391-1395.[Medline]
  27. Bork P. The modular architecture of a new family of growth regulators related to connective tissue growth factor. FEBS Lett 1993; 327:125-130.[CrossRef][Medline]
  28. Denizot F, Brunet JF, Roustan P, et al. Novel structures CTLA-2 alpha and CTLA-2 beta expressed in mouse activated T cells and mast cells and homologous to cysteine proteinase proregions. Eur J Immunol 1989; 19:631-635.[Medline]
  29. McKinnon PJ, McLaughlin SK, Kapsetaki M, Margolskee RF. Extracellular matrix-associated protein Sc1 is not essential for mouse development. Mol Cell Biol 2000; 20:656-660.[Abstract/Free Full Text]
  30. Grimberg A, Cohen P. Role of insulin-like growth factors and their binding proteins in growth control and carcinogenesis. J Cell Physiol 2000; 183:1-9.[CrossRef][Medline]
  31. Vogel T, Guo N, Guy R, et al. Apolipoprotein E: a potent inhibitor of endothelial and tumor cell proliferation. J Cell Biochem 1994; 54:299-308.[CrossRef][Medline]
  32. Trimborn T, Gribnau J, Grosveld F, Fraser P. Mechanisms of developmental control of transcription in the murine alpha- and beta-globin loci. Genes Dev 1999; 13:112-124.[Abstract/Free Full Text]
  33. Gourdon G, Sharpe JA, Higgs DR, Wood WG. The mouse alpha-globin locus regulatory element. Blood 1995; 86:766-775.[Abstract/Free Full Text]
  34. Heldin CH, Westermark B. Mechanism of action and in vivo role of platelet-derived growth factor. Physiol Rev 1999; 79:1283-1316.[Abstract/Free Full Text]
  35. Dimitriadou V, Koutsilieris M. Mast cell-tumor cell interactions: for or against tumour growth and metastasis? Anticancer Res 1997; 17:1541-1550.[Medline]
  36. Ruoslahti E. Fibronectin and its integrin receptors in cancer. Adv Cancer Res 1999; 76:1-20.[Medline]
  37. Sipkins DA, Cheresh DA, Kazemi MR, Nevin LM, Bednarski MD, Li KC. Detection of tumor angiogenesis in vivo by alphaVbeta3-targeted magnetic resonance imaging. Nat Med 1998; 4:623-626.[CrossRef][Medline]
  38. Weissleder R. Molecular imaging: exploring the next frontier. Radiology 1999; 212:609-614.[Free Full Text]
  39. Simone NL, Bonner RF, Gillespie JW, Emmert-Buck MR, Liotta LA. Laser-capture microdissection: opening the microscopic frontier to molecular analysis. Trends Genet 1998; 14:272-276.[CrossRef][Medline]



This article has been cited by other articles:


Home page
 PhysiologyHome page
R. G. Pautler
Mouse MRI: Concepts and Applications in Physiology
Physiology, August 1, 2004; 19(4): 168 - 175.
[Abstract] [Full Text] [PDF]

Home page
 Br. J. Radiol.Home page
K C P Li
Angiogenesis imaging in the post-genomic era
Br. J. Radiol., December 1, 2003; 76(suppl_1): S1 - S2.
[Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2282020907v1
228/2/560    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
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 reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Guccione, S.
Right arrow Articles by Bednarski, M. D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Guccione, S.
Right arrow Articles by Bednarski, M. D.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE