Effect of Fixation Time and Microwave Oven Heating Time on Retrieval of the Ki-67 Antigen from Paraffin-embedded Tissue

Transcription

1 /93/$3.3 The Journal of Histochemistry and Cytochemistry Copyright 1993 by The Histochemical Society, Inc Vol. 41, No. 8. pp , 1993 Printed in US.A. Original Article I Effect of Fixation Time and Microwave Oven Heating Time on Retrieval of the Ki-67 Antigen from Paraffin-embedded Tissue SATORU MUNAKATA' and JAMES B. HENDRICKS' University of Florida Diagnostic Referral Laboratories, Department of Paibology, Gainesville, Florida. Received for publication December 16, 1992 and in revised form March 17, 1993; accepted March 2, 1993 (2A2896). Microwave oven heating has been employed for retrieval of antigens from formalin-fued, padhembedded tissue for immunohistochemical staining. Recently, a Ki-67 equivalent murine monoclonal antibody was generated which can detect tumor proliferative activity in routinely processed tissue with microwave oven heating. We assessed the effect of fixation time (4, 24, or 48 hr) and microwave oven heating time (7, 14, 21,28, 35, or 49 min) on retrieval of the Ki-67 antigen from tonsil tissue. Ki-67 staining was quantitated by image analysis. Owing to the heterogeneity of Ki-67 staining within and between germinal centers, we employed a measurement technique that averages staining across the ger- mina1 centers of each section. Good Ki-67 immunostaining was observed for all microwave oven heating times in tissue fixed for 4 hr. In contrast, poor immunostaining was observed in tissue fued for 48 hr unless a heating time of at least 14 min was used. Tonsil fued for 24 or 48 hr showed a significant increase in percentage positive nuclear area after microwave times of 14 or 21 min. Prolonged heating time (up to 49 min) had no effect on the quality of Ki-67 staining in tissue fued for 4 or 48 hr. (J Histochem Cytochem 41: , 1993) KEY WORDS: Immunohistochemistry; Antigen retrieval; Formalinfixed tissue; Proliferation. Introduction Tumor proliferative activity has prognostic value in many solid tumors (1-4). Immunohistochemical analysis of tumor proliferation has been accomplished with the Ki-67 monoclonal antibody (MAb) which recognizes a nuclear antigen expressed throughout the cell cycle (5). Unfortunately, this MAb cannot be used for retrospective studies that employ formalin-fixed, paraffin-embedded tissue. Recently, however, a Ki-67 equivalent murine MAb was generated which can detect tumor proliferative activity in routinely processed tissue (6,7). Immunohistological detection with this antibody requires antigen retrieval by microwave oven heating according to the method described by Shi et al. (8). The mechanism of action of microwave oven heating for retrieval of antigens is poorly understood. Shi et al. (8) hypothesized that the cross-linking effect of formaldehyde fixation is altered by microwave heating. They reported that tissues fixed in formalin for as long as 2 years could be stained for pancytokeratin and vimentin using antigen retrieval. However, the effect of overfixation could not be quantitatively determined because optimally fixed tissue SM is a Visiting Scientist from the Department of Obstetrics and Gynecology, Hirosaki University School of Medicine, Hirosaki, Aomori, Japan. ' Correspondence to: James B. Hendricks, PhD, U. of Florida, Dept. of Pathology, Gainesville, FL from the same anatomic site was not available. In addition, the relationship between formalin fixation time and microwave oven heating time was not examined. We describe here the effect of fixation time and microwave oven heating time on retrieval of the Ki- 67 antigen. We employed a quantitative approach with image cytometry (9,lO). Materials and Methods Tissue and Fixation. Fresh human tonsil tissue was obtained from routine surgical specimens. Tissue was cut into three equal pieces of approximately 4 mm3 and immediately fixed in 1% neutral buffered formalin for 4, 24, or 48 hr. After fixation the tissue was immersed in 7% ethanol until processing. All tissue were processed simultaneously. Fixed tissues were dehydrated in ethanol, cleared in xylene, and embedded in paraffin blocks, which were cooled before sectioning. Four-pm thick sections were cut and mounted on 3-aminopropyltriethoxysilane (APES)(Sigma; St Louis, M)- coated slides. The use of a tissue adhesive helped to prevent section detachment during microwave processing. Sections were air-dried for at least 15 min and heat-fixed to the slide for an additional 2 min at 6'C. Microwave Oven. A Panasonic model NN-6371WM microwave oven (maximum power 9 W) was employed. Slides were placed in one of two plastic Coplin jars filled with 1 mm sodium citrate buffer, ph 6 (Sigma). The Coplin jars were situated in the center of a plastic pan containing approximately 5 ml water, and were always placed in the same position in the microwave oven. When heated at the indicated settings, the boiling point of the buffer solution was reached in 2.14 *.45 mins. 1241

2 1242 MUNAKATA. HENDRICKS Immunohistochemistry. Slides were deparaffinized in two changes of xylene for 1 min each. They were hydrated in decreasing concentrations of ethanol and rinsed in PBS. For each experiment, slides prepared from tissue fixed for 4, 24, or 48 hr were processed simultaneously. Six slides were placed in each of the two Coplin jars for microwaving. Slides were microwaved at 7-min intervals for a total of 7, 14, 21, 28, 35, or 49 min at maximal power level. The maximal attainable temperature in each Coplin jar was not affected by slide removal (data not shown). At each 7-min interval the fluid level in the Coplin jar was topped off with water. No special precautions were employed for prolonged microwaving. After heating, the Coplin jars were removed from the microwave and allowed to cool. For each experiment, two slides prepared from tissue fixed for 4, 24, or 48 hr were not exposed to microwaves. Endogenous peroxidase was blocked with 3% H22 for 1 min at room temperature. Slides were incubated overnight at 4 C with a 1:2 dilution of primary antibody (MIB1) (a gift from AMAC, Westbrook, ME) or control. A biotin-streptavidin detection system was employed with diaminobenzidine (DAB) as the chromogen (BioGenex; San Ramon, CA). Slides were washed twice with PBS and incubated with the linking reagent (biotinylated anti-immunoglobulins) for 2 min at room temperature (RT). After rinsing in PBS, the slides were incubated with the peroxidase-conjugated streptavidin label for 2 min at RT. The sections were again rinsed with PBS and incubated with DAB for 1 min in the dark. After chromogen development, slides were washed in two changes of water for 8 min each and counterstained with.2% ethyl green (Cell Analysis Systems; Elmhurst, IL) in sodium acetate buffer, ph 4 The sections were then dehydrated, cleared in xylene, and mounted with permount. Image Analysis. Quantitation of Ki-67 h nostaining was accomplished using the CAS 2 Image Analysis System (Cell Analysis Systems), which employs two solid-state image-sensing channels especially matched to identify nuclei stained with ethyl green in one channel and nuclei stained with DAB in the other channel. Staining was expressed as positive nuclear area defined as the percentage of nuclei with chromogen intensity above the control antibody threshold. The variation (CV) in percent of positive area from image field to field was reported as heterogeneity. The quantitative immunocytochemical (QIC) score was generated by the following equation: positive area x positive staid1 (1). The antibody threshold was determined on the control slide. To allow uniformity of analysis between slides, 25 % of each of four germinal centers (GC) was measured according to the scheme outlined in Figure 1. Scene segmentation was employed as necessary to confine measurement to the germinal center. The total nuclear area measured from each slide was always greater than 1, pmz. Results Without microwave oven heating, formalin-fixed, paraffin-embedded tonsil tissue exhibited little MIBl immunostaining. In contrast, after microwave heating tonsil tissue exhibited a strong staining pattern with the MIBl antibody in the GC and the basal layer of the epithelium. Occasional positivity in the diffuse lymphoid tissue was observed. Staining was restricted to the nucleus, with no evidence of cytoplasmic staining except in cells containing mitotic figures. The pattern of nuclear staining was homogeneous or blotchy, with irregular densities scattered throughout the nucleoplasm. Frozen sections prepared from the same tonsils exhibited a similar staining pattern (data not shown). Evaluation of inter-gc staining heterogeneity was performed by measuring 1 complete GCs from the same optimally fixed (4 hr) tissue section. The heterogeneity between measurements in a single GC varied from 5% to 22% for the 1 GCs studied (Figure 2). Furthermore, the percentage positive area between individual GCs varied from 39% to 73% (Figure 2). On the basis of these experiments, we developed a measurement technique that averages staining across the germinal centers of each section (Figure 1). The effect of fixation time and microwave oven heating time on Ki-67 immunostaining is shown in Figure 3 for one representative experiment of three. Good Ki-67 immunostaining was observed for all heating times in tissue fixed for 4 hr. However, staining intensity appeared to improve with heating times of 14 min or more. The most dramatic change in stain intensity as a function of microwave oven heating time was observed for overfixed tissue. In tissue fixed for 24 or 48 hr, poor immunostaining was observed unless -\\\ 25 1 Figure 1. Each germinal center was divided into quarters; staining was assessed in the lower right quadrant of the first germinal center, the upper left quadrant of the second germinal center, the lower left quadrant of the third germinal center, and the upper right quadrant of the fourth germinal center. 3 I Figure 2. Percentage positive area and heterogeneity assessed by measuring 1 complete germinal centers from the same optimally fixed (4-hr) tonsil tissue section. Mean is indicated by the horizontal line.

3 CONDITIONS AFFECTING E-67 ANTIGEN RETRIEVAL 1243 a heating time of at least 14 min was employed. The results of quantitative image analysis are shown in Tgble 1 for one representative experiment of four. Without microwave oven heating the K;-67- positive area was less than 1% regardless of formalin fixation time. No statistically significant increase in percentage positive area was observed for optimally fmed tissue (4 hr fixation) as a function of microwave oven heating time. However, there was a tendency towards increased positive area in tonsil microwaved for 14 or 21 min compared with 7 min. Tonsil fixed for 24 hr showed a significant increase in percentage positive area after microwave times of 14 min (p <.5) or 21 min (p <.1) relative to a 7-min microwave time. Likewise, for tonsil fixed for 48 hr, a significant increase in percentage positive area was observed after microwave times of 14 min (p <.5) or 21 min (p <.5) relative to 7 min. Regardless of fixation time, no significant dlfference in percentage positive area was observed between tissue microwaved for 14 or 21 min. Tonsil fixed for 48 hr showed a significant increase in heterogeneity (p <.2) relative to tissue fixed for 4 hr with a microwave time of 7 min. In each experiment the greatest variation in percentage positive area from field to field (heterogeneity) was observed for tonsil tissue that was overfixed (48 hr) and underheated (7 min). To determine the effect of prolonged microwave heating time on Ki-67 immunostaining, tonsil tissue fixed for 4 or 48 hr was microwaved for 7-min intervals up to 49 min. Ki-67 immunostaining as measured by percentage positive area or QIC score increased significantly between microwave times of 7 and 14 min but thereafter plateaued (Figure 4). Discussion The microwave oven has been used extensively for tissue fixation (11). Recently a method was described for retrieving antigens for immunohistochemical staining from formalin-fixed, paraffinembedded tissue by microwave Oven hearing (8). It was hypothesized that the cross-linking effect of formaldehyde fixation is altered by microwaving. It is not known ifoverfition requires increased heating time for optimal retrieval of tissue antigens. The Ki-67 equivalent MAb recently generated by Gerdes and colleagues for use in routinely processed tissue requires microwave heating (6). These authors microwaved tissue sections for a total of 1 min either in a household microwave oven (7 W) or a Bio- Rad Polaron H25 (BioRad; Richmond, CA) microwave processor (45 W). Although they suggested that the MIBl antibody gave identical results with lymphoid tissues fixed in formalin for as long as 72 hr, they did not examine the effect of formalin fixation time on contiguous pieces from the same tissue sample. Furthermore, these authors did not quantitatively examine Ki-67 immunostaining. In this study we examined the effect of fition time and microwave oven heating time on retrieval of the Ki-67 antigen from contiguous pieces of tonsil tissue. Ki-67 staining was quantitated by image analysis. Owing to the heterogeneity in Ki-67 staining within and between germinal centers (5,12), we employed a measurement technique that attempted to average staining across the germinal centers of each section. Tonsil fixed for 4 hr exhibited excellent staining with Ki-67; there was a tendency for increased positive area in tonsil microwaved for 14 or 21 min relative to 7 min. Overfixed tissue (24 or 48 hr) required a minimum of 14 min for good staining. In every experiment the greatest variation in percentage positive area from field to field (heterogeneity) was observed for tonsil tissue that was overfixed and underheated. Because there is no standardization of tissue fixation among laboratories, protocols requiring antigen retrieval for immunohistochemical staining must employ a microwave heating time that ensures maximal and complete staining. This is particularly true for antigens like Ki-67 that require quantitative measurements for clinical significance. The microwave time that produces independence from length of fixation may vary from antigen to antigen. Based on these studies, however, it appears that prolonged microwaving (up to 49 min) does not have a negative effect on staining. Therefore, in laboratories that lack control over tissue fixation, it may be advisable to use longer heating times. Our laboratory employs a heating time of 21 min for retrieval of the Ki-67 antigen. Ki-67 immunohistochemical staining is a simple and reliable method of studying tumor proliferative activity in formalin-fixed, paraffin-embedded tissue. The information obtained on paraffinembedded material by Ki-67 evaluation may prove superior to S-phase fraction by flow cytometry, which has been shown to suffer from various technical limitations (13). Acknowledgments We thank Robin Foss, PA, for technical'assistance andamac, Inc. (a subsidiary of Immunotech SA) for kind4 providing the MIBl antibody for these studies. Literature Cited 1. Bauer KD, Lincoln ST, Vera-Roman JM, Wallemark CB, Chime1 JS, Madurski ML, Murad T, Scarpelli DG. Prognostic implications of proliferative activity and DNA aneuploidy in colonic adenocarcinomas. Lab Invest 1987;57: Hedley DW, Rugg CA, Gelber RD. Association of DNA index and Table 1. Effect offixation time ana' microwave oven heating time on Ki-67 immunostaining, measured by image analysis for one representative experiment of foura Fix Mw POS. area Hetero N. area Fix, foation time in formalin (hr); MW, microwave oven heating time (min); Pos. area, positive area (%); Hetero, heterogeneity; N. area, cumulative nuclear area analyzed ( Id).

4 Figure 3. Representativetonsil tissue sections from a single experiment showing nuclear immunoreactivity with Ki-67 (MIBI) after (A, E) 4-hr fixation or (C, D) 48-hr fixation and microwave oven heating times of (A, C) 7 min or (E. D) 14 min. Original magnification x 5. Bars = 1 pm. 1244

5 CONDITIONS AFFECTING Ki-67 ANTIGEN RETRIEVAL t d, _I :.. - I.. c.

6 ~ MUNAKATA, HENDRICKS : 4 W In a v - 4 $ I I 1 I I I v?! < 55 - (U 5 2 d._ g 45 a Microwave Time (minutes) S-phase fraction with prognosis of nodes positive early breast cancer. Cancer Res 1987;47: Kallioniemi -P, Punnonen R, Mattila J. Lehtinen M. Koivula T Prognostic significance of DNA index, multiploidy, and S-phase fraction in ovarian carcinoma. Cancer 1988;61: Merkel DE, McGuire WL. Ploidy, proliferative activity and prognosis. DNA flow cytometry of solid tumors. Cancer 199;65: Gerdes J, Schwab U, Lemke H, Stein H. Production of a mouse monoclonal antibody reactive with a human nuclear antigen associated with cell proliferation. Int J Cancer 1983;31:13 6. Cattoretti G, Becker MHG, Key G, Duchrow M, Schluter C, Galle J, Gerdes J. Monoclonal antibodies against recombinant parts of the Ki- 67 antigen (MIB 1 and MIB 3) detect proliferating cells in microwaveprocessed formalin-fixed paraffin sections. J Pathol 1992;168: Gerdes J, Becker MHG, Key G, Cattoretti G. Immunohistological detection of tumour growth fraction (Ki-67 antigen) in formalin-fixed and routinely processed tissue. J Pathol, 1992;168:85 8. Shi SR. Key ME, Kalra KL. Antigen retrieval in formalin-fixed, paraffinembedded tissues: an enhancement method for immunohistochemical staining based on microwave oven heating of tissue sections. J Histochem Cytochem 1991;39: Bacus SS, Goldschmidt R, Chin D, Moran G, Weinberg D, Bacus JW. Biological grading of breast cancer using antibodies to proliferating cells and other markers. Am J Pathol 1989;135: Bacus S, Flowers JL, Press MF, BacusJW, McCarty KS. The evaluation of estrogen receptor in primary breast carcinoma by computer-assisted image analysis. Am J Clin Pathol 1988;9: Leong ASY, Milios J, Duncis CG. Antigen preservation in microwaveirradiated tissues: a comparison with formaldehyde fixation. J Pathol 1988;156: Kame1 OW, LeBrun DP, Davis RE, Berry GJ, Warnke RA. Growth fraction estimation of malignant lymphomas in formalin-fixed paraffinembedded tissue using anti-pcnakyclin 19A2. Correlation with Ki- 67 labeling. Am J Pathol 1991;138: Frierson HE The need for improvement in flow cytometric analysis of ploidy and S-phase fraction. Am J Clin Pathol 1991;95: I I I I I I I Microwave Time (minutes) 4 15 Figure 4. Effect of prolonged microwave heating time on Ki-67 immunastaining. (A) Tonsil fixed for 4 hr. (e) Tonsil fixed for 48 hr.