Context.—The development of biotechnologic platforms capable of high throughput analysis has ushered in a promising new era of translational medicine. However, most studies to date are based on in vitro cell lines or substitute models for human disease. Although these model systems have proven insightful, it is readily becoming apparent that human clinical tissue must be studied in order to fully understand all the nuances of human disease. Studies that are based on human tissue, however, are limited by qualitative and quantitative issues, factors often precluding their use in high throughput studies.

Objective.—To develop a simple and rapid tissue procurement protocol for use in obtaining a homogeneous epithelial cell population from clinical tissue and the recovery of nucleic acids and proteins of high quality and quantity. Also, to determine if the technique preserves tissue, thereby allowing morphologic correlation with molecular findings.

Design.—Performance of manual exfoliation to procure cells from clinical resection specimens and use of immunomagnetic beads embedded with the antibody ber-Ep4 for the positive enrichment of a homogeneous epithelial cell population. Nucleic acids and proteins are then separated using a phenol plus guanidine thiocyante solution. Nucleic acids and proteins are quantitated and qualitatively analyzed using standard laboratory techniques.

Results.—Nucleic acids and proteins of high quality and quantity were recovered following manual exfoliation and immunomagnetic bead separation. Tissue architecture was not destroyed, thus permitting histologic and molecular correlation.

Conclusions.—A simple and reproducible protocol is presented that may enable the molecular profiling of clinically resected tissue. Although the technique is currently limited to certain tissue and tumor types, further research will broaden its overall application.

The era of one gene and one outcome correspondence is over. Today, it is recognized that in certain diseases, such as cancer, the phenotype is determined by a complex interplay at several cellular levels. Alterations in DNA structure and copy number, gene expression levels, and protein expression, as well as posttranslational modifications, all contribute to the complexity of cancer. Fortunately, biotechnology has advanced to the point where molecular analysis on a global scale can be performed in single-event experiments, greatly expanding our understanding of oncogenesis. Applications of these technologies, however, are mostly limited to in vitro cell lines or animal models capable of mimicking human pathologies. Although these alternative models have provided significant insight into disease processes, their application and relevance toward human patient care are not absolute. Cell lines may not reflect the original tumor for a number of reasons. False and contaminated cell lines have recently been discovered in an often-used cell line repository.1 Cell culturing may result in both phenotypic and genotypic drift.2,3 Animal models may not completely mimic their human counterparts.4 These problems, and the eventual need for validation studies on human tissue, have led to a recognition of the importance of studying human tissue.5,6 However, significant technical challenges continue to exist in the study of clinical tissue and are predominantly related to sample collection and processing.

Clinical material used for research can be derived through several means, all of which have inherent drawbacks relating to either quantitative or qualitative issues.7 Bulk tissue that is snap-frozen consists of a heterogeneous mixture of cells that may include inflammatory and stromal cells, with the percentage of tumor cells present often unknown. Some molecular-based studies require that at least 70% to 80% of a specimen be of a homogeneous cell population in order to avoid spurious results.8 Snap-frozen tissue is pulverized prior to the isolation of macromolecules; hence, no morphologic information about the procured cell population can be correlated to subsequent data. A study that is based on heterogeneous tissue makes interpretation problematic, with findings corresponding to the tumor cell of interest as well as the contaminating normal cells. Such problems necessitate additional studies aimed at assigning cell type to experimentally generated signals.9 Alternatively, tissue can be frozen, a histologic slide prepared, and the cells microdissected to ensure a homogeneous cell population for study. However, the recovery of nucleic acids is typically of insufficient quantity for certain high throughput technologic platforms.10 The amount of tissue frozen and dedicated for research and the lower histologic quality compared to formalin-fixed tissue are additional concerns.

Although homogeneous cell populations can be microdissected from formalin-fixed, paraffin-embedded tissue, the technique is hindered by the poor recovery and poor quality of macromolecules once they have been exposed to fixatives. The amount of DNA and RNA recovered from formalin-fixed tissue can be as low as 30% for DNA and 1% and 15% for RNA.11,12 To overcome small quantities of nucleic acids, amplification protocols have been developed. For DNA, however, the success of certain nonpolymerase chain reaction–based protocols requires nucleic acids to be of high molecular weight.13 Some polymerase chain reaction–based protocols require intact DNA to be in the range of 500 to 2000 bp.14 In contrast, most DNA from formalin-fixed tissue is fragmented and limited to fewer than 500 bp.15 The highly fragmented DNA recovered from formalin-fixed tissue is therefore a poor template for polymerase chain reaction–based amplification protocols and may lead to biased or limited amplification.16,17 RNA that is subjected to fixatives is even more problematic than DNA. Chemical modification by fixative renders substrates, such as the polyA tail of messenger RNA, a poor template for complementary DNA synthesis, the initial step in some amplification protocols.18 Fragmentation of messenger RNA in fixed tissue limits the number of templates available for amplification.19 The presence of endogenous RNase in tissue, combined with variables such as tissue section thickness and coefficients of diffusion/penetration of fixative, can lead to degradation of RNA before fixation.20 These factors can introduce bias if the amplification of messenger RNA is to be performed.21 

The recovery of proteins poses other unique problems. Proteins cannot be amplified, a procedure that would ensure adequate numbers of a homogeneous cell population for high throughput proteomic studies using techniques such as 2-dimensional polyacrylamide gel electrophoresis, a technique that requires the acquisition of a minimum of thousands of microdissected cells.22 In addition to being laborious, microdissection techniques require tissue staining. Because eosin alters the isoelectric point of proteins, special stains are required to avoid errant results.23 Additionally, although alternatives to formalin fixation have been developed, their use may result in diminished protein recovery.24 Thus, although formalin-fixed, paraffin-embedded tissue represents the largest repository of archived material, it is not an optimal source from which to perform molecular studies.

Two other factors must be taken into consideration when designing a tissue procurement protocol. An ideal tissue procurement protocol would be able to obtain matched normal and neoplastic cells and preserve optimal morphology. The first consideration, obtaining normal and neoplastic cells from the same cellular lineage and the same source/patient, is important to reduce inherent experimental variables.25,26 The second consideration, preservation of tissue morphology, is required to correlate discoveries and new technologies with the established, gold standard of tissue examination, histology.

This pilot project was initiated to assess the feasibility of a novel cellular procurement protocol for obtaining cells from clinically resected tissue for eventual use on high throughput platforms. As a prerequisite, these platforms require large quantities of largely intact macromolecules. The proposed technique, manual exfoliation coupled to immunomagnetic bead separation, was designed to take into consideration all the previously mentioned variables. The successful recovery of a homogeneous cell population with nucleic acids and proteins of both high quality and quantity and preserved tissue morphology would represent forward progress toward translational research.

The Institutional Review Board of the University at Buffalo granted approval for this project. Resected hemicolectomy specimens performed for curative purposes were chosen as the prototypic tissue source on the basis of their availability and histologic growth pattern. Anonymized hemicolectomy specimens were retrieved upon surgical extirpation and immediately placed on ice. The tissue was transported to the Department of Pathology. On arrival, the specimen was opened, and a gross examination was performed by a board-certified pathologist (W.D.M.). Normal-appearing and tumor tissue areas were selected for manual exfoliation on the basis of the gross examination. Separately, manual exfoliation of uninvolved, normal colonic epithelium and tumor tissue was performed. Manual exfoliation consisted of the gentle scraping of selected areas by the edge of a glass slide.27 For nonneoplastic colonic epithelial cells, the surface of the colonic resection specimen several centimeters away from the tumor was scraped. For neoplastic colonic epithelial cells, the tumor was sectioned, and the cut surface was scraped by a glass slide. The exfoliated cells present on the slide were then transferred to a microcentrifuge tube. This was accomplished by rotating the slide perpendicular, with the edge pointing to the ground (Figure 1). By doing this, the exfoliated cells were collected in a tissue droplet that could be deposited in an awaiting microcentrifuge tube containing a 50:50 mixture of RNALater/phosphate-buffered saline. The exfoliated cells were then subjected to positive isolation by introducing magnetic beads bound with immunoglobulin (Figure 1) (Dynal Biotech ASA, Oslo, Norway). Ber-Ep4 was the antibody bound to these beads and is known to recognize the 34- and 39-kd glycoproteins present on the cytoplasmic surface of all epithelial cells, except for squamous, hepatic, and parietal cells. The cells and the immunomagnetic beads were then incubated together for 20 minutes at 4°C on a roller to allow adequate mixing. Afterward, a magnetic particle concentrator was used to collect the beads into a pellet (Dynal MPC, Dynal Biotech). The supernatant was discarded, and the cells were washed twice in a buffered solution. The cells were pelleted by use of the magnetic particle concentrator, and the supernatant was discarded. A small fraction of cells from both the normal and tumor cell pellets was smeared onto separate slides. These slides were stained with hematoxylin-eosin (Figure 2). These slides served as a source to approximate the homogeneity of the exfoliated cells after enrichment with the immunomagnetic beads. The pelleted cells were further processed by the addition of 50 μL of a commercially available phenol plus guanidine thiocyanate solution (TRI Reagent, Molecular Research Center Inc, Cincinnati, Ohio). The cells were then manually disrupted with a sterile disposable micropestal (Argos Technologies Inc, Dundee, Ill), followed by repetitive pipetting to reduce the viscosity of the cell lysates. An additional 700 μL of TRI Reagent was added, and the DNA, RNA, and proteins were separated according to the manufacturer's recommendations. The RNA was further purified by spin columns (RNeasy MinElute Cleanup Kit, Qiagen, Valencia, Calif).

Figure 1.

Technique of manual exfoliation and immunomagnetic bead separation. A glass slide is used to gently exfoliate cells from fresh tissue. The flexibility of this technique allows the operator to select the desired cells for exfoliation and study. A scraping of the colon away from the exophytic mass will allow the collection of uninvolved, normal colonic epithelial cells. A scraping of the tumor surface will allow the collection of dysplastic cells. The scraping of the cut section of the tumor would collect invasive tumor cells. The slide is then angled, and the exfoliated cells are collected at the bottom edge. The exfoliated cells are then incubated with magnetic beads bound with the ber-Ep4 antibody. Enrichment of epithelial cells is achieved after the contaminating cells are washed off.

Figure 1.

Technique of manual exfoliation and immunomagnetic bead separation. A glass slide is used to gently exfoliate cells from fresh tissue. The flexibility of this technique allows the operator to select the desired cells for exfoliation and study. A scraping of the colon away from the exophytic mass will allow the collection of uninvolved, normal colonic epithelial cells. A scraping of the tumor surface will allow the collection of dysplastic cells. The scraping of the cut section of the tumor would collect invasive tumor cells. The slide is then angled, and the exfoliated cells are collected at the bottom edge. The exfoliated cells are then incubated with magnetic beads bound with the ber-Ep4 antibody. Enrichment of epithelial cells is achieved after the contaminating cells are washed off.

Close modal
Figure 2.

Visual verification of the cellular enrichment of colonic epithelial cells recovered by manual exfoliation and immunomagnetic bead separation. a, Nonneoplastic colonic epithelial cells. b, Neoplastic colonic epithelial cells (hematoxylin-eosin, original magnification ×40)

Figure 2.

Visual verification of the cellular enrichment of colonic epithelial cells recovered by manual exfoliation and immunomagnetic bead separation. a, Nonneoplastic colonic epithelial cells. b, Neoplastic colonic epithelial cells (hematoxylin-eosin, original magnification ×40)

Close modal

Recovered DNA was resuspended in 8 mM NaOH and quantitated on a Nanodrop ND-1000 spectrophotometer (Ambion, Austin, Tex). One microgram of DNA was aliquoted and further digested overnight with the restriction enzyme EcoRI. One microgram of undigested and 1 μg of digested DNA from both the normal and neoplastic cells were electrophoresed on a 0.7% ethidium bromide–stained nondenaturing agarose gel with a 1000-kilobase pair DNA ladder (catalog no. 10381-010, Invitrogen, Carlsbad, Calif). The DNA was visualized using an ultraviolet light source. Recovered RNA was eluted from the spin columns in nuclease-free water and analyzed on an Agilent 2100 BioAnalyzer using the RNA 6000 Lab Chip from Caliper Technologies (Mountain View, Calif). Recovered proteins were resuspended in 1% sodium dodecyl sulfate and quantitated using a commercially available colorimetric assay based on the Bradford dye-binding procedure (Bio-Rad Protein Assay, catalog no. 500-0006, Bio-Rad, Hercules, Calif). Proteins were denatured for 5 minutes at 95°C prior to electrophoresis. Equimolar concentrations of proteins derived from normal colonic epithelial cells and neoplastic colonic epithelial cells were electrophoresed in parallel on a 4% to 20% Pre-Cast Tris glycine gradient gel with a prestained sodium dodecyl sulfate–polyacrylamide gel electrophoresis protein molecular-weight standard (catalog no. 161-0318, Bio-Rad). After electrophoresis, the gel was stained with Coomassie blue for 1 hour, followed by overnight de-staining. A digital camera was used for imaging.

Cell Homogeneity

The cells that were aliquoted from the separate, pelleted normal and neoplastic cells were smeared onto glass slides, and a peripheral blood type smear was created. The slides were visually examined using a microscope to assess cell enrichment. The separate cell populations, normal colonic epithelial cells, and neoplastic colonic epithelial cells showed different cytologic features (Figure 2). The normal cell population displayed polarity and low nuclear-cytoplasmic ratios. The neoplastic cell population showed loss of polarity, nuclear enlargement, and hyperchromasia. Both samples displayed cell enrichment and consisted of the targeted colonic epithelial cell population and magnetic beads. Visual examination of the cell population smeared onto a slide showed that greater than 95% were epithelial cells.

Recovery of DNA

High-molecular-weight DNA was recovered, as evidenced by a slow migrating band during electrophoresis (Figure 3, a). A second, fainter band corresponding to approximately 16 000 bp was present that was compatible with mitochondrial DNA. The DNA digested with EcoRI showed up as a smear, confirming the undigested high-molecular-weight band as DNA. On average, 242 μg of DNA was recovered per sample.

Figure 3.

Qualitative analysis of recovered nucleic acids and proteins. a, Simple electrophoresis of DNA. Lane M, Molecular-weight marker with top band = 12 000 bp. Lane A, DNA from normal colonic cells. Lane B, DNA from normal colonic cells digested with the restriction enzyme EcoRI. Lane C, DNA from neoplastic colonic cells. Lane D, DNA from neoplastic colonic cells digested with the restriction enzyme EcoRI. b, Histogram of RNA evaluated on an Agilent 2100 BioAnalyzer. RNA shows the presence of sharp peaks corresponding to 18S and 28S (ribosomal RNA). c, Electrophoresis of proteins extracted from normal colonic cells (N) and tumor (T). M indicates protein molecular-weight ladder

Figure 3.

Qualitative analysis of recovered nucleic acids and proteins. a, Simple electrophoresis of DNA. Lane M, Molecular-weight marker with top band = 12 000 bp. Lane A, DNA from normal colonic cells. Lane B, DNA from normal colonic cells digested with the restriction enzyme EcoRI. Lane C, DNA from neoplastic colonic cells. Lane D, DNA from neoplastic colonic cells digested with the restriction enzyme EcoRI. b, Histogram of RNA evaluated on an Agilent 2100 BioAnalyzer. RNA shows the presence of sharp peaks corresponding to 18S and 28S (ribosomal RNA). c, Electrophoresis of proteins extracted from normal colonic cells (N) and tumor (T). M indicates protein molecular-weight ladder

Close modal

Recovery of RNA

Analysis of RNA on the Agilent 2100 BioAnalyzer showed the presence of 2 sharp peaks corresponding to intact 18S and 28S ribosomal RNA. From 6 separate samples, the ratio between these 2 peaks was between 1.47 and 2.36 (Figure 3, b). In most cases, the intervening distance between these 2 peaks did not contain any well-defined smaller peaks, and the baseline between 29 seconds and the 18S peak was relatively flat. These features are compatible with intact RNA.28 An average of 16.48 μg of total RNA was recovered per sample.

Recovery of Protein

An average of 146 μg of proteins was recovered from the cell samples. The electrophoresed samples showed protein bands of similar molecular weight present between the normal and tumor samples. However, some bands differed between the 2 samples by the amount of protein present (Figure 3, c). This represented quantitative differences between the normal and tumor cell populations for similarly migrating proteins.

Preservation of Tissue Architecture

The areas from which the cells were scraped off were fixed in formalin after manual exfoliation was performed. The tissue was subsequently fixed in formalin and embedded in paraffin, and a routine hematoxylin-eosin– stained slide was created. Microscopic examination showed preservation of the overall tissue architecture (Figure 4, a).

Figure 4.

Limitations of technique based on the histology of the tumor type. a and b, Colonic adenocarcinoma and renal clear cell carcinoma, respectively, show total effacement of tissue by tumor cells. No residual normal epithelial cells from that organ site are present. c and d, Adenocarcinoma of the prostate and ductal carcinoma of the breast, respectively, show neoplastic glands infiltrating between residual normal prostatic glands and mammary ducts (hematoxylin-eosin, original magnification ×4)

Figure 4.

Limitations of technique based on the histology of the tumor type. a and b, Colonic adenocarcinoma and renal clear cell carcinoma, respectively, show total effacement of tissue by tumor cells. No residual normal epithelial cells from that organ site are present. c and d, Adenocarcinoma of the prostate and ductal carcinoma of the breast, respectively, show neoplastic glands infiltrating between residual normal prostatic glands and mammary ducts (hematoxylin-eosin, original magnification ×4)

Close modal

Although the definition of translational medicine is broad, one aspect that characterizes it are “investigations in humans which define the biology of disease.”29 The recent development of several biotechnologic platforms capable of analyzing thousands of individual molecules in single experimental events heralded the advent of high throughput studies and, with them, the hope of better defining human disease. Most studies to date, however, have used these technologies only on in vitro cell lines or animal model tissue. Because these sources have inherent drawbacks, there is a growing consensus for the need to study human tissue. Despite this understanding, several factors continue to hinder the widespread application of these high throughput techniques on human tissue. Tissue heterogeneity, macromolecular degradation, insufficient material for analysis, and histologic correlation remain problematic issues. Recent advances in tissue processing have been developed but have also proven to possess drawbacks.30,31 The inability to completely circumvent these concerns may influence experimental designs and trials, thereby effectively preventing the application of these high throughput technologies to human tissue.

A tissue procurement protocol capable of retrieving macromolecules of good quality and quantity was recently reported that used manual exfoliation and laser capture microdissection.32,33 We have modified this protocol by using immunomagnetic beads in lieu of laser capture microdissection. Unlike 2 other reported protocols that have also incorporated the use of immunomagnetic beads, our protocol maintains the use of manual exfoliation as the initial step.34,35 In our experience, the use of manual exfoliation of cells possesses several advantages over the 2 latter reported techniques. Foremost is its simplicity. Secondly, it partially purifies the targeted epithelial cells from the host tissue in the fresh state in one easy, rapid movement. This greatly diminishes the amount of time required to work with the tissue and the use of other reagents and equipment. In contrast, other reported techniques require shock freezing the tissue source, sectioning, staining, and then manual dissection before incubation with immunomagnetic beads. Another technique also freezes the tissue and then requires filtering the tissue on a nylon mesh, followed by repeated washes and centrifugation steps prior to incubation with immunomagnetic beads. In this regard, the manual exfoliation technique is simpler and faster and uses less equipment and reagents than these other reported methods.

Coupling manually exfoliated cells with magnetic beads bound with the antibody ber-Ep4 can greatly enrich an epithelial cell population. Because the technique of manual exfoliation occurs at the macroscopic level, it nonselectively scrapes cells off from the host tissue. This may include inflammatory cells, vessels, and stromal cells. The use of magnetic beads bound with antibodies to ber-Ep4 ensures that only epithelial cells are retained after the wash and concentration steps. However, the lack of an antibody specific to the plasma membrane of neoplastic cells currently limits this technique to certain types of tumors. The epithelial-specific nature of this antibody linked to the magnetic beads means that tumors characterized by the infiltration of neoplastic cells among and between normal epithelial cells and glands are not candidates for this technique. For example, manual exfoliation of an adenocarcinoma of the prostate or ductal carcinoma of the breast could result in the acquisition of inflammatory cells, vessels, stromal cells, and a mixed population of normal and neoplastic epithelial cells (Figure 4, c and d). The use of magnetic beads bound with ber-Ep4 will only isolate an epithelial cell population from other cell types but will not separate a normal from a neoplastic epithelial population. Therefore, under its current design, this technique is only capable of isolating a homogeneous normal or neoplastic epithelial cell population from certain tissues and tumors. Tumors that are characterized by total effacement of the tissue by neoplastic cells are suitable candidates for this technique. Since manual exfoliation occurs at the macroscopic level, the center of the tumor is the best place for manual exfoliation. In this regard, adenocarcinoma of the colon and clear cell renal carcinoma are suitable tumors for examination (Figure 4, a and b). Histologically, it can be observed that in the center of these neoplasms, the tumor consists only of neoplastic glands and vessels, stroma, and inflammatory cells. Although the vessels, stroma, and inflammatory cells may also be exfoliated when scraped by a slide, they can be washed off after incubation with immunomagnetic beads. Depending on the area manually exfoliated (ie, whether normal or neoplastic), the end product after immunomagnetic bead separation is a homogeneous population of (normal or neoplastic) epithelial cells.

The technique of manual exfoliation serves 2 purposes important in translational research. It acquires fresh cells in a viable state for further isolation and, more importantly, does not destroy the underlying host tissue. Cells are exfoliated from the underlying stromal support by this technique and retained for downstream studies. The result is the acquisition of cells of interest that are of high quality and quantity. We chose a commercially available phenol plus guanidine thiocyanate reagent for the isolation of molecules because of the ability of this reagent to separate DNA, total RNA, and proteins through separate steps. We were able to recover sufficient quantities of DNA and RNA for high throughput platforms, as required by a regional microarray facility. This facility recommends a minimum of 1 μg of DNA to perform the requested analysis on a 19K bacterial artificial chromosome array comparative genomic hybridization platform. Either a total of 500 ng of total RNA is recommended for an oligonucleotide expression array or a total of 2.5 μg of total RNA for a complementary DNA expression array. Two-dimensional differential in-gel electrophoresis requires a minimum of 50 μg of protein. In all instances, we recovered sufficient quantities for these high throughput platforms. If more starting material is required, the flexibility of the technique permits compensation. By manually exfoliating the tissue several more times, additional pellets of cells can be generated. The macromolecular components can be isolated in parallel and eventually pooled. These additional scrapes ensure that adequate amounts of nucleic acids or proteins are procured for analysis. By doing this, amplification steps and the bias they can incur can be avoided.21 Additionally, the area from which these cells were exfoliated can subsequently be submitted for formalin fixation and paraffin embedding. In the future, results from analytic studies performed on the exfoliated cells may be directly correlated with the tissue's histologic features.

Although this technique obviates several of the negative variables associated with standard tissue procurement processing, one factor remains removed from the pathologist's domain. That factor concerns in vivo tissue hypoxia or warm ischemia, the time between ligation of the vascular supply and surgical extirpation.36 Under current surgical techniques, this factor cannot be circumvented, and so for now, it should be duly recorded by any pathologist using this technique for research.

In conclusion, a straightforward protocol has been presented that can serve to procure cells from certain tissue types and tumors. The simplicity of manual exfoliation renders this technique readily adaptable by other pathologists interested in obtaining high-quality tissue for molecular studies. The technique is rapid and has been shown to be capable of isolating a homogeneous epithelial cell population with recovery of nucleic acids and proteins of high quantity and quality. The preservation of tissue architecture allows a histologic correlation of the tissue source with downstream molecular studies. Although limitations currently exist in regard to the type of tissue and tumor that can be examined, further experimentation with the protocol may lead to the discovery of unique targets, and the development of antibodies toward them may broaden the application of the protocol to other tissues and tumor types.

This project was funded by a grant from the Kaleida Health Foundation.

McLeod
,
R. A. F.
,
W. G.
Dirks
, and
Y.
Matsui
.
et al
.
Widespread intraspecies cross-contamination of human tumor cell lines arising at source.
Int J Cancer
1999
.
83
:
555
563
.
Osborne
,
C. K.
,
K.
Hobbs
, and
J. M.
Trent
.
Biological differences among MCF-7 human breast cancer cell lines from different laboratories.
Breast Cancer Res
1987
.
9
:
111
121
.
Celis
,
A.
,
H. H.
Rasmussen
, and
P.
Celis
.
et al
.
Short term culturing of low grade superficial bladder transitional cell carcinoma leads to changes in the expression levels of several proteins involved in key cellular activities.
Electrophoresis
1999
.
20
:
355
361
.
European Organization for Research and Treatment of Cancer (EORTC).
Laboratory research division workshop on the role of in vivo pre-clinical models in the development of contemporary cancer therapeutics, Verona, Italy, February 3, 2001.
Eur J Cancer
2001
.
37
:
1976
1980
.
Emmert-Buck
,
M. R.
,
R. L.
Strausberg
, and
D. B.
Krizman
.
et al
.
Molecular profiling of clinical tissue specimens.
J Mol Diagn
2000
.
2
:
60
66
.
Duggan
,
B. J.
,
J. J.
McKnight
, and
K. E.
Williamson
.
et al
.
The need to embrace molecular profiling of tumor cells in prostate and bladder cancer.
Clin Cancer Res
2003
.
9
:
1240
1247
.
Emmert-Buck
,
M. R.
Molecular profiling initiative.
Available at: http://cgap-mf.nih.gov/. Accessed August 25, 2004
.
Liku
,
J.
,
V. I.
Zabarovska
, and
E.
Braga
.
et al
.
Loss of heterozygosity in tumor cells requires re-evaluation: the data are biased by the size dependent differential sensitivity of allele detection.
FEBS Lett
1999
.
462
:
121
128
.
Tureci
,
O.
,
J.
Ding
, and
H.
Hilton
.
et al
.
Computational dissection of tissue contamination for identification of colon cancer specific expression profiles.
FASEB J
2003
.
17
:
376
385
.
McClintick
,
J. N.
,
M. A.
Kuczyk
, and
U.
Paeslack
.
et al
.
Reproducibility of oligonucleotide arrays using small samples.
BMC Genomics
2003
.
4
:
e4
.
Available at: www.biomedcentral.com/bmcgenomics/. Accessed September 10, 2004
.
Serth
,
J.
,
M. A.
Kuczyk
, and
U.
Paeslack
.
et al
.
Quantitation of DNA extracted after micropreparation of cells from frozen and formalin fixed tissue sections.
Am J Pathol
2000
.
156
:
1189
1196
.
Abrahamsen
,
H. N.
,
T.
Steiniche
, and
E.
Nexo
.
et al
.
Towards quantitative mRNA analysis in paraffin embedded tissues using real-time reverse transcriptase-polymerase chain reaction.
J Mol Diagn
2003
.
5
:
34
41
.
Lage
,
J. M.
,
J. H.
Leamon
, and
T.
Pejovic
.
et al
.
Whole genome analysis of genetic alterations in small DNA samples using hyperbranched strand displacement amplification and array-CGH.
Genome Res
2003
.
13
:
294
397
.
James
,
L. A.
Comparative genomic hybridization as a tool in tumor cytogenetics.
J Pathol
1999
.
187
:
385
395
.
Sikowski
,
A.
,
A.
Ishkanian
, and
C.
Garnis
.
et al
.
An efficient method for the assessment of DNA quality of archival microdissected specimens.
Mod Pathol
2002
.
15
:
889
892
.
Roemen
,
G.
,
M.
Lentjes
, and
M.
Luchtenborg
.
et al
.
Reliable mutation analysis of archival tissue: critical issues and procedures.
Proc Am Assoc Cancer Res
2003
.
44
:
1213
1214
.
Wand
,
G.
,
C.
Brennan
, and
M.
Rook
.
et al
.
Balanced PCR amplification allows unbiased identification of genomic copy changes in minute cell and tissue samples.
Nucleic Acids Res
2004
.
32
:
e76
.
Available at: www.nar.oupjournals.org/. Accessed August 18, 2004
.
Masuda
,
N.
,
T.
Ohnishi
, and
S.
Kawamoto
.
et al
.
Analysis of chemical modification of RNA from formalin fixed samples and optimization of molecular biology applications for such samples.
Nucleic Acids Res
1999
.
27
:
4436
4443
.
Specht
,
K.
,
T.
Richter
, and
U.
Muller
.
et al
.
Quantitative gene expression analysis in microdissected archival formalin fixed and paraffin embedded tumor tissue.
Am J Pathol
2001
.
158
:
419
429
.
Srinivasan
,
M.
,
D.
Sedmak
, and
S.
Jewell
.
Effect of fixatives and tissue processing on the content and integrity of nucleic acids.
Am J Pathol
2002
.
161
:
1961
1971
.
Wilson
,
C. L.
,
S. D.
Pepper
,
Y.
Hey
, and
C. J.
Miller
.
Amplification protocols introduce systematic but reproducible errors into gene expression studies.
Biotechniques
2004
.
36
:
498
506
.
Emmert-Buck
,
M. R.
,
J. W.
Gillespie
, and
C. P.
Paweletz
.
et al
.
An approach to proteomic analysis of human tumors.
Mol Carcinog
2000
.
27
:
158
165
.
Craven
,
R. A.
,
N.
Totty
, and
P.
Harden
.
et al
.
Laser capture microdissection and two dimensional polyacrylamide gel electrophoresis: evaluation of tissue preparation and sample limitations.
Am J Pathol
2002
.
160
:
815
822
.
Ahram
,
M.
,
M. J.
Flaig
, and
J. W.
Gillespie
.
et al
.
Evaluation of ethanol fixed, paraffin embedded tissues for proteomic applications.
Proteomics
2003
.
3
:
413
421
.
Bakay
,
M.
,
Y. W.
Chen
, and
R.
Borup
.
et al
.
Sources of variability and effect of experimental approach on expression profiling data interpretation.
BMC Bioinformatics
2002
.
3
:
e4
.
Available at: www.biomedcentral.com/bmcbioinformatics/. Accessed September 12, 2004
.
Ornstein
,
D. K.
,
J. W.
Gillespie
, and
C. P.
Paweletz
.
et al
.
Proteomic analysis of laser capture microdissected human prostate cancer and in vitro prostate cell lines.
Electrophoresis
2000
.
21
:
2235
2242
.
Yang
,
G. C.
and
S. A.
Hoda
.
Combined use of the “scratch and smear” sampling technique and Ultrafast Papinicolaou stain for intraoperative cytology.
Acta Cytol
1997
.
41
:
1513
1518
.
Accessed September 30, 2004
.
Vitale
,
D.
The total RNA story.
Available at: http://www.chem.agilent.com/. Accessed September 30, 2004
.
Mankoff
,
S. P.
,
C.
Brander
,
S.
Ferrone
, and
F. M.
Marincola
.
Lost in translation: obstacles to translational medicine [commentary].
J Transl Med
2004
.
2
:
14
15
.
Florell
,
S. R.
,
C. M.
Coffin
, and
J. A.
Holden
.
et al
.
Preservation of RNA for functional genomic studies: multidisciplinary tumor bank protocol.
Mod Pathol
2001
.
14
:
116
128
.
Jungbluth
,
A. A.
,
D.
Kolb
, and
S. K.
Eastlake-Wade
.
et al
.
Tissues fixed in stabilization solution ‘RNAlater’: effect on immunohistochemistry [abstract].
Mod Pathol
2005
.
18
:(
suppl 1
).
330A
.
Mojica
,
W. D.
,
R.
Saxena
, and
R. T.
Cheney
.
A cytology based tissue procurement protocol for homogeneous cell populations for molecular profiling [abstract].
Am J Clin Pathol
2003
.
120
:
445
.
Mojica
,
W. D.
,
L. A.
Liotta
, and
V.
Espina
.
Manual exfoliation of fresh tissue obviates the need for frozen sections for molecular profiling [abstract].
Mod Pathol
2004
.
17
:(
suppl 1
).
1627
1629
.
Sarto
,
C.
,
C.
Valsecchi
, and
P.
Mocarelli
.
Renal cell carcinoma: handling and treatment.
Proteomics
2002
.
2
:
1627
1629
.
Croner
,
R. S.
,
K.
Guenther
, and
T.
Foertsch
.
et al
.
Tissue preparation for gene expression profiling of colorectal carcinoma: three alternatives to laser capture microdissection with preamplification.
J Clin Lab Med
2004
.
143
:
344
351
.
Huang
,
J.
,
R.
Qi
, and
J.
Quakenbush
.
et al
.
Effects of ischaemia on gene expression.
J Surg Res
2001
.
99
:
222
227
.

The authors have no relevant financial interest in the products or companies described in this article.

Presented in poster form at the 94th Annual Meeting of the United States and Canadian Academy of Pathologists, San Antonio, Tex, February 2005.

Author notes

Reprints: Wilfrido D. Mojica, MD, University at Buffalo, Department of Pathology, 100 High St, Buffalo, NY 14203 (mojica@buffalo.edu)