Influence of Histochemical and Immunohistochemical Stains on Polymerase Chain Reaction Takayuki Murase, M.D., Hiroshi Inagaki, M.D., Tadaaki Eimoto, M.D. Department of Pathology, Nagoya City University Medical School, Nagoya, Japan longed the CYCLEmin (P < .01), and the PCR ampli- The polymerase chain reaction (PCR) analysis of fication did not reach the “plateau” level with a DNA extracted from tissue sections can be applied maximum of 60 cycles. The PCR efficiency was to a variety of research and diagnostic protocols. To worse in microwave or pressure cooker treatment, analyze selectively the specific areas of tissue, a di- with neither CYCLEmin nor CONCmax being ob- rect microdissection of histochemically or immuno- tained. When target areas from sections for subse- histochemically stained sections, if satisfactory for quent PCR amplification are microdissected, PCR, is helpful. However, the influence of various methyl green is most suitable as a dye for nuclear staining methods on PCR has been poorly investi- staining. The immunohistochemical visualization gated. In this study, paraffin sections of formalin- with DAB or new fuchsin yields no unfavorable ef- fixed lymph node samples were histochemically fects. A successful PCR amplification may not be stained with Mayer’s hematoxylin, eosin Y, methyl expected in sections that are pretreated in a micro- green, or May-Grunwald solution and immuno- wave oven or pressure cooker. stained for CD45 using 3,3؅-diaminobenzidine (DAB), DAB with cobalt ion (DAB-Co), or new fuch- KEY WORDS: Histochemical stain, Immunohisto- sin as the chromogen. In addition, unstained sec- chemical stain, Microdissection, Polymerase chain tions were treated with trypsin, microwave, or pres- reaction, Pretreatment. sure cooker, the techniques frequently used in Mod Pathol 2000;13(2):147–151 immunostains for antigen unmasking. DNA was extracted from each section, and the PCR efficiency Polymerase chain reaction (PCR) used for tissue in amplifying a 110 bp portion of the beta-globin sections as a source of DNA plays an important role gene was evaluated by two parameters: the cycle in pathologic research and diagnosis (1–7). To an- count in which the first visible band was obtained alyze selectively the specific areas of interest, mi- (CYCLEmin) and the maximum amount of PCR crodissection of the sections on glass slides usually products (CONCmax). The hematoxylin stain showed has to be performed. Now the combination of mi- a significantly prolonged CYCLEmin (P < .01) and crodissection and PCR can be applied to even a lower CONCmax (P < .05) in comparison with un- single cell (e.g., Hodgkin or Reed-Sternberg cells in stained and untreated control sections. The May- malignant lymphoma sections) (4–7). Grunwald stain showed a prolonged CYCLEmin (P < In microdissecting such small targets, an accu- .01), although the CONCmax was not significantly rate morphologic identification is necessary, where ؍ different from that of the control (P .051). The the histochemical or immunohistochemical stain is eosin and methyl green stains showed no effects of help. Tissue samples from hematoxylin and eo- against PCR. In immunostains, the DAB-Co method sin–stained sections, however, yield an unsatisfac- showed a lower CONCmax (P < .05), whereas the tory PCR amplification (8, 9). Although various CYCLEmin was not prolonged. The DAB and new dyes, including Wright’s solution instead of hema- fuchsin methods had no untoward effects. Antigen- toxylin, have been applied for this purpose (8), cor- unmasking treatments showed deteriorating effects relation between each dye and efficiency of PCR on PCR. The trypsin treatment significantly pro- remains largely unknown. The conventional phe- nol/chloroform purification of the DNA template may decrease the influence of the various inhibitors Copyright © 2000 by The United States and Canadian Academy of Pathology, Inc. (10–12); however, a considerable amount of DNA VOL. 13, NO. 2, P. 147, 2000 Printed in the U.S.A. Date of acceptance: August 25, 1999. may be lost during the purification (10). Address reprint requests to: Takayuki Murase, M.D., Department of Pa- In the present study, we evaluated in more detail thology, Nagoya City University Medical School, Kawasumi-1, Mizuho- chou, Mizuho-ku, Nagoya, Japan, 467-8601; e-mail: the effects against PCR amplification of histochem- [email protected]; fax: 81-52-851-4166. ical dyes commonly used in microdissection (May- 147 er’s hematoxylin, eosin Y, methyl green, and May- taining 0.2 mg/mL DAB and 0.003% H2O2 (13). In Grunwald solution). We also performed a similar the DAB-Co method, using cobalt ion to intensify analysis on the signal visualizing methods by im- the DAB signals, 0.02% w/v CoCl2 was added to the munohistochemistry (peroxidase-diaminobenzidine above DAB solution; the reaction with peroxidase [DAB], peroxidase-DAB-cobalt [DAB-Co], and alka- resulted in black color (14). In the new fuchsin line phosphatase–new fuchsin). To our knowledge, method, the slides were incubated with streptavidin- no such influence of immunostains on PCR has been alkaline phosphatase, and the red signals were devel- studied. Furthermore, we examined the influence of oped with new fuchsin solution (0.5 mL of a 5% so- signal-enhancing or antigen-unmasking methods lution of new fuchsin in 2N HCl added to 1.25 mL of that frequently are used in modern immunohisto- a freshly prepared 4% solution of sodium nitrite) (15). chemistry (trypsin treatment, microwave heating, and After developing, each slide was rinsed in distilled high-temperature treatment by pressure cooker). water and air dried without counterstaining. To eval- uate the effect of the antigen-unmasking pretreat- ments, dewaxed sections were treated with three dif- MATERIALS AND METHODS ferent methods as follows: incubation in 0.1% w/v trypsin in phosphate buffered saline for 10 min at 37° Tissue Samples C, microwave treatment for 10 min with the sections Six paragastric lymph nodes of similar size (ap- in citrate buffer (0.01 M, pH 6.0), and high- proximately 1 cm in diameter) were collected from temperature heating in a pressure cooker by dipping six patients at surgery for gastric cancer. These slides in pure water for 4 min after the water boiled. nodes were fixed in 10% buffered formalin for 12 to 72 hours and embedded in paraffin. The tissue sec- tions (3 ␮m thick) were deparaffinized and rehy- DNA Extraction and PCR drated in distilled water. The hematoxylin and eo- The above sections as well as unstained and un- sin stain showed a normal nodal architecture with treated (control) sections from six lymph nodes no primary or metastatic tumors in each lymph were totally scraped off the slides and subjected to node. All of the procedures used in this study were DNA extraction. Each section was digested in 0.05 M performed with ample precautions for avoiding Tris buffer (200 ␮l) containing 200 ␮g/mL protein- cross-contaminations. ase K (Boehringer Mannheim, Stuttgart, Germany) at 55° C overnight, then heated at 95° C for 10 min to inactivate proteinase K (16). The influence of Histochemical Stains, Immunohistochemical each staining method on PCR was evaluated by Stains, and Antigen-Unmasking Treatments amplifying a 110 bp portion of beta-globin gene For histochemical stains, the dewaxed sections (17). Subjected to PCR were eight tubes per section, were stained with each of the following dyes of containing 25 ␮l solution with a mixture of 10 mM usual concentration: 1.0% w/v Mayer’s hematoxy- Tris-HCl (pH 8.4), 50 mM KCl, 0.01% gelatin, 1.5 mM ␮ lin, 0.5% w/v eosin Y, 0.5% w/v methyl green, and MgCl2, 0.2 mM dNTPmix, 1.0 M each primer, 0.05 May-Grunwald solution (containing 0.25% w/v unit per microliter of Taq GOLD DNA polymerase nonoxidized methylene blue and eosin Y). When (Perkin-Elmer, New York, NY), and 5 ␮l extracted appropriate contrast was obtained after incubation DNA. The sequences of the primers for the comple- for 10 to 30 min, the sections were rinsed in distilled mentary portions were 5Ј-ACACAACTGTGTTCACT- water and then air dried. For immunohistochemis- AGC-3Ј and 5Ј-CAACTTCATCCACGTTCACC-3Ј, re- try, the dewaxed sections were stained for CD45 spectively (17). The samples were incubated at 95° C (leukocyte common antigen), because most of the for 10 min to activate Taq GOLD DNA polymerase, cells in the lymph node express this antigen. The then PCR was performed for each tube from 25 to 60 signals were visualized using one of the DAB, DAB- cycles at 5-cycle intervals; the cycle condition was for Co, and new fuchsin methods. Briefly, the sections 1 min at 95° C, 1 min at 55° C, and 1 min at 72° C. One were treated with 3% hydrogen peroxide for 15 min ␮l of PCR product in each tube was electrophoresed and reacted with anti-CD45 antibody (2B11 ϩ PD7/ through 2% agarose gel containing ethidium bromide 26; DAKO, Kyoto, Japan) for 1 hour in a humid with a size and concentration marker. The PCR effi- chamber at room temperature, followed by incuba- ciency was evaluated by two parameters: the cycle tion with biotinylated antimouse immunoglobulins count at which the 110 bp PCR product was first (Nichirei, Tokyo, Japan). Then the reaction prod- visible (CYCLEmin) and the maximum concentration ucts were visualized by each of the following meth- of the PCR band (CONCmax). The CONCmax was ob- ods. In the DAB method, the slides were reacted tained when the amount of PCR product reached its with peroxidase-conjugated streptavidin (Nichirei, “plateau” state. The gel images were digitized (Fluor-S Tokyo, Japan), and the brown signals were devel- MultiImager & Multi-Analyst, Bio-Rad Laboratories, oped by immersing slides in the DAB solution con- Melville, NY), and the concentrations of the bands 148 Modern Pathology were quantified using NIH image software (version 1.57; Macintosh, Cupertino, CA).