Isolation of full-size mRNA from cells sorted by flow cytometry

J. Biochem. Biophys. Methods 40 (1999) 69–80

Isolation of full-size mRNA from cells sorted by flow cytometry Claudius Diez, Gerald Bertsch, Andreas Simm* ¨ klinische Biochemie und Pathobiochemie, Versbacher Str. 5, D-97078 Wurzburg ¨ , Germany Institut f ur

Abstract Gene expression is one key mechanism to regulate cell growth and differentiation. It is usually determined by Northern blotting or RT-PCR. However, studies with primary cell cultures are frequently hampered due to contaminating cells such as fibroblasts. We have developed a method to isolate intact full-size mRNA from sorted cells. In many cell types, e.g. cardiac myocytes, cell sorting without prior fixation revealed complete RNA breakdown. Based on a murine fibroblast cell line (AKR-2B), ethanol and formaldehyde at various concentrations and pre-treatment with ribonuclease inactivating DEPC were compared with each other. Fixation with 75% ice-cold DEPC–pre-treated ethanol for 5 min yielded mostly intact RNA. In contrast, antibody staining prior to sorting required 15 min fixation. Addition of RNAse-free BSA (0.5%) and 2 mM CaCl 2 optimised the cell recovery ratio and thus a better RNA yield (60% compared to control) after sorting than former studies. Northern blotting and RT-PCR show the intact mRNA species b-actin. Furthermore, dependent on the cellular PCNA content, we have demonstrated the cell cycle dependent cdk2 and cyclin A expression. This fast and reliable method allows to isolate intact full-size mRNA species appropriate for Northern blotting and RT-PCR to monitor gene expression.  1999 Elsevier Science B.V. All rights reserved. Keywords: Flow cytometry; RNA isolation; Fixation; Antibody staining; Northern blotting; RT-PCR

1. Introduction Gene expression studies with primary differentiated post-mitotic cells, such as cardiac Abbreviations: BSA, bovine serum albumin; DEPC, diethyl-pyrocarbonate; PCNA, proliferating cell nuclear antigen; RT-PCR, reverse transcription polymerase chain reaction *Corresponding author. Tel.: 1 49-931-201-3144; fax: 1 49-931-201-3153. E-mail address: [email protected] (A. Simm) 0165-022X / 99 / $ – see front matter  1999 Elsevier Science B.V. All rights reserved. PII: S0165-022X( 99 )00020-2

70

C. Diez et al. / J. Biochem. Biophys. Methods 40 (1999) 69 – 80

myocytes and neuronal cells, are frequently hampered by several limitations. Northern blotting often fails due to the insufficient RNA amount from the cell type under investigation. Polymerase chain reaction, PCR, proved to be a successful alternative in the past and its major advantage lies in the sensitivity because as little as a few transcripts are necessary to study the expression of certain genes. PCR application, however, is heavily dependent upon the purity of the cell culture. Most primary cultured cells, e.g. cardiac myocytes, are contaminated with approximately 5–10% non-myocytes such as fibroblasts and smooth muscle cells. Using PCR or Northern blotting, these contaminations are sufficient to feign false-positive results, especially, if particular genes are abundantly expressed within the contaminating cells. Cell sorting with a fluorescent activated cell sorter (FACS  ) is one possible technical approach to separate cell populations. Our initial experiments unfortunately showed that cardiac myocyte sorting resulted in mechanical damage and RNA degradation. Thus, cells have to be fixed before sorting (Diez and Simm, submitted). Commonly used fixatives are ethanol, formaldehyde, and acetic acid or combinations of these reagents. Using a murine cell line, we decided to investigate different fixation procedures in more detail to establish a method for the isolation of full length mRNA from sorted cells. The approach introduced here comprises ethanol fixation, ribonuclease inactivation as well as RNA isolation according to a standard protocol [2].

2. Materials and methods

2.1. Solutions All solutions, except those containing formaldehyde and paraformaldehyde, were treated with DEPC at a final concentration of 0.1%, and autoclaved; ethanol solutions received only DEPC without subsequent autoclaving. Two hours before usage, solutions were repeatedly treated with DEPC at a final concentration of 0.1%. Rinsing solutions for the cell sorter were treated with DEPC, left overnight at 48C and autoclaved consecutively. Twelve hours before the planned experiment, the cell sorter (Epics Elite ESP, Coulter, Germany) was extensively rinsed with DEPC-treated PBS. Only sterile glassware was used, plastic tubes were rinsed with freshly prepared DEPC-treated water and autoclaved.

2.2. Cell culture Murine AKR-2B fibroblasts were seeded on 6-well plates (Falcon, Becton Dickinson) or on Petri-dishes (Greiner, Frickenhausen, Germany) at a cell density of 5 3 10 3 cells / cm 2 and grown to confluence over 5 days in McCoy-5A medium containing 5% calf serum (Hyclone) without a medium change. After fixation, cells were directly lysed with 4 M guanidinium isothiocyanate or, in the case of formaldehyde fixation, scraped off with a rubber policeman into the lysis buffer and homogenised with a polytron mixer (Polytron PT 1200, Kinernatica KG, Switzerland).

C. Diez et al. / J. Biochem. Biophys. Methods 40 (1999) 69 – 80

71

2.3. RT-PCR Total RNA from murine AKR-2B fibroblasts was extracted according to the method of Chomczynski and Sacchi [2] with minor modifications as described before [3]. One mg of total RNA was reverse transcribed by 100 units of Molony murine leukaemia virus reverse transcriptase (Gibco, Eggenstein, Germany) in a 15 ml reaction volume, using 10 pmol of the appropriate upstream primer. A 227 bp fragment of the murine b-actin gene was amplified by PCR (see Table 1 for primer specifications). The final 100 ml reaction included 5 ml reverse transcribed template, 20 mmol of each dNTP (ATP, CTP, GTP, and TTP), 0.25 mmol MgCl 2 , 20 pmol of each primer, 10 ml of 10 3 PCR Buffer (AGS, Heidelberg, Germany) and 8 units of Taq polymerase (AGS, Heidelberg, Germany). PCR conditions were as follows: 30 s at 958C, 60 s at 568C and 120 s at 728C. After 30 cycles of amplification, 5 ml of the 100 ml reaction volume were analysed by ethidium bromide stained 1% agarose gel electrophoresis. Subsequently, 95 ml of the reaction volume was purified (Quiagen PCR Purification Kit) and sequenced by using the fluorescent di-deoxy terminator method of cycle sequencing on a Perkin-Elmer, Applied Biosystems Division, 373A automated DNA sequencer.

2.4. Northern blotting Total RNA was isolated according to Ref. [2]. For quantification of RNA loss, the RNA from one well, and for analysis after passage through the FACS, 10 mg of total RNA was denatured at 658C in a solution containing 1.1 M formaldehyde, 35% formamide, and ethidium bromide. RNA was electrophoresed in a 1.25% agarose gel containing 6.1% formaldehyde. After vacuum transfer onto nylon sheets (Hybond N, Amersham) in 6 3 SSC (1 3 SSC 5 150 mM NaCl and 15 mM sodium citrate, pH 7.0), the RNA was cross-linked by UV irradiation (120 mJ / cm 2 for 30 s). The blot was hybridised at 608C in 5 3 SSC, 10 3 Denhard’s solution, 0.2% SDS, and 300 mg / ml Table 1 Characteristics of primer Specificity

Sequences a

Tm (8C)

Product length

Position b

Accession number

b-Actin

TGC TGA TCC ACA TCT GCT GGA GAA GTG TGA CGT TGA CAT CCG TCC ACC ACC CTG TTG CTG TAG C TGG AAA GCT GTG GCG TGA TG CAC TCA CAC ACT TAG TGT CTC TGG TGG G GCC GCG ATG CCG GGC ACC TCG AGG CAT TCG TTG CGA TAA CAG GCT CCG TC CAC AGC CGT GGA TAT CTG GAG ACG CCT GTG GTG GTT ACG CT CCA TCT CTG GCA CCA CTG AC

56

227 bp

J00691 c

58

401 bp

59

1277 bp

56

233 bp

56

279 bp

3079– 2729 1001– 600 1393– 116 962– 729 910– 631

GAPDH Cyclin A cdk 2 cdk 4 a

The upper sequence indicates the antisense primer. The binding sites listed below point to the according mRNA. c The numbers indicate the primer binding positions on the corresponding cDNA. b

Reference

M17701 Z26580

[10]

D28753

[11]

LI 1007

[11]

72

C. Diez et al. / J. Biochem. Biophys. Methods 40 (1999) 69 – 80

denatured herring sperm DNA. The DNA probes were labelled by random priming using an oligolabelling kit (Pharmacia, Freiburg, Germany) with [a 32 P]dCTP (Hartmann Analytic, Braunschweig, Germany). The filters were washed in solutions containing decreasing contents of SSC in 0.1% SDS and 2 mM EDTA. The final wash stringency was 0.4% SSC at 658C. Radioactivity was detected and quantified using a PhosphorImager  . As the b-actin probe served the 227 bp fragment obtained in the PCR as described before, or a 770 bp fragment (Oncor, Hamburg, Germany).

2.5. Cell sorting and antibody staining AKR-2B cells from two 10 cm Petri-dishes (Greiner, Frickenhausen, Germany) were recovered with trypsin and fixed for 15 min as described under Section 4 with 75% ice-cold ethanol pre-treated with DEPC. After collecting cells by centrifugation (400 g, 5 min), cells were re-suspended in DEPC-treated PBS. A total of 10 6 cells / ml were incubated with 5 mg FITC-conjugated anti-PCNA (Pharmingen, San Diego, USA) antibodies for 30 min in DEPC-treated PBS / 1% RNase-free BSA. After centrifugation at 400 g for 5 min cells were re-suspended in PBS and sorted using an EPICS Elite ESP flow cytometer (Coulter, Krefeld, Germany). The foreward scatter signal of the 488 nm Argon laser was used as a trigger and the FITC fluorescence was quantified after excitation at 488 nm in the green light channel (525 / 10 nm bandpass filter). In a control experiment, anti-PCNA marked cells were additionally stained for DNA using 1.2 mg / ml Hoechst 33258 and cell cycle analysis was done using a gated amp and the 450 nm emission after excitation with the 325 nm line of a Helium–Cadmium laser. All cells, independently whether stained or not, were sorted into autoclaved vials containing 0.5% RNAse-free BSA and 2 ml CaCl 2 in DEPC-treated PBS. After sorting, the cell suspension was spun down at 14 000 g for 3 min and the cell pellet was lysed in 4 M guanidinium isothiocyanate. We used 10 mg total RNA for Northern blotting and 1 mg for RT-PCR.

2.6. Statistical analysis Radioactivity was measured by a PhosphorImager  and data were stored in a graphic file (.gel). ImageQuant  software was used to determine the RNA content of the probes. Data were subsequently transferred to Origin  4.1 to calculate means and standard deviations. All data are presented as means6standard deviation unless otherwise indicated.

3. Results

3.1. Formaldehyde versus ethanol fixation Formaldehyde and ethanol both are commonly used fixatives in clinical research. First experiments included paraformaldehyde as a third species. As we have never observed any differences between formaldehyde and paraformaldehyde with respect to RNA

C. Diez et al. / J. Biochem. Biophys. Methods 40 (1999) 69 – 80

73

integrity and subsequent Northern blotting, we exclusively used formaldehyde in our further experiments (data not shown). As shown in Fig. 1A, only ethanol fixation (25 or 75%, respectively) led to obviously intact RNA independent from fixation time (5 or 15 min, respectively). In contrast, formaldehyde fixation (0.5, 0.2, and 0.05%, respectively) prevented ‘normal cell lysis’ with guanidinium isothiocyanate buffer and thus RNA could not be isolated. Light microscopy revealed the morphological integrity of

Fig. 1. RNA fractionation and b-actin blots after different fixations at room temperature. (A) Total RNA of cells from one well was extracted. RNA was size fractionated on a denaturing formaldehyde gel. Unfixed cells were used as a control (C). Duration and concentration of the fixative point to the fixation conditions. DEPC was not added to the fixatives in this experiment. A (upper panel) shows the ethidium bromide stained total ] RNA in the gel after size fractionation and B (lower panel) the according Northern blot with a b-actin probe (arrow indicates the size of the 18S rRNA). After formaldehyde fixation, no significant amount of 18S and 28S rRNA can be seen. There is no significant difference in RNA yield between controls and ethanol fixation at various concentrations (A), whereas the Northern blot (B) shows nearly complete degradation of the b-actin messenger RNA after fixation with ethanol at room temperature and without DEPC pre-treatment.

74

C. Diez et al. / J. Biochem. Biophys. Methods 40 (1999) 69 – 80

formaldehyde-treated cells (G. Bertsch, unpublished data). Only a short exposure (5 min) to 0.05% formaldehyde results in a low amount of 28S and 18S RNA as seen in Fig. 1A. It is noteworthy that we almost always find tRNA after formaldehyde fixation and mechanical homogenisation. Using Northern blotting, we evaluated the fixation quality, i.e. the RNA loss and the extent of degradation during fixation. As shown in Fig. 1B, Northern blotting with b-actin revealed degraded RNA after ethanol fixation.

3.2. Effects of temperature and DEPC pre-treatment on ethanol fixation As mentioned above, cell lysis after ethanol fixation resulted in degraded RNA. To reduce the degree of degradation, we tested ethanol fixation at room temperature and on ice as well as the pre-treatment with DEPC. We observe the best results after 5 min fixation with DEPC pre-treated 75% ethanol at room temperature and on ice, respectively (Fig. 2). Additionally, ice-cold fixatives resulted less frequently in degradation (data not shown). In contrast, lower ethanol concentration (25%) led to an enhanced degradation. Table 2 presents the quantitative RNA loss of approximately 40% considering the indicated fixation conditions.

3.3. Cell sorting yields intact RNA convenient for Northern blotting and RT-PCR Further experiments were performed with ice-cold and DEPC pre-treated 75% ethanol. As we wished to apply our method to antibody staining, we considered 15 min fixation, because the fluorescence intensity proved to be more stable than after 5 min fixation (data not shown). We isolated intact RNA after sorting and successfully tested this RNA for RT-PCR and Northern blotting as shown in Fig. 3. Moreover, we mixed proliferating and density arrested AKR-2B fibroblasts and stained for PCNA as a proliferation marker within this mixed population. Fig. 4A shows the PCNA fluorescence versus the DNA fluorescence and Fig. 4B the PCNA distribution within the population. After sorting cells with a high PCNA content, i.e. proliferating cells, we screened for cell cycle dependent cdk2 and cyclin A. Fig. 4C demonstrates the presence of both cdk2 and cyclin A only within cells expressing abundantly PCNA, whereas GAPDH and cdk4 remain unchanged. This finding supports our method to check for gene expression within FACS sorted cells.

3.4. Discussion The objective of this study was to find an experimental approach to isolate intact RNA from fixed cells after FACS based cell sorting. The findings of this study then should be transferred to primary cell culture, where a reliable and simple protocol supports to study gene expression within these cells. Instead of primary cells, we used a well established murine cell line, AKR-2B, to search for optimal fixation and sorting conditions. AKR-2B cell themselves cannot be sorted without prior fixation (A. Simm, unpublished data) and thus are an appropriate cell model for our experiments. Isolation of intact RNAs from sorted cells without

C. Diez et al. / J. Biochem. Biophys. Methods 40 (1999) 69 – 80

75

Fig. 2. Effect of temperature and DEPC on RNA yield and degradation. Cells on a 6-well plate were fixed as indicated for 5 min. Total RNA from one well was isolated and loaded onto a denaturing formaldehyde gel and size fractionated. RNA gel (A) and the corresponding b-actin Northern blot (B) are shown. The 18S RNA was indicated as a size control. No RNA degradation can be seen for 75% ice-cold ethanol with and without addition of DEPC and with 75% ethanol at room temperature after addition of DEPC.

preceding fixation has only been described for peripheral blood cells and thymocytes [12–14]. Compared to primary cells, AKR-2B cells are usually less affected by preparation techniques (e.g. different cell number, contaminations). Furthermore, experiments with a cell line yield more reliable and consistent results than experiments with primary cells.

76

C. Diez et al. / J. Biochem. Biophys. Methods 40 (1999) 69 – 80

Table 2 Effect of time and ethanol concentration on RNA yield abc

Control 25% ethanol, 5 min 75% ethanol, 5 min Control 25% ethanol, 15 min 75% ethanol, 15 min

RNA [% of control]

Number of experiments

100.065.9 66.864.4 62.5610.6 100.063.9 64.6613.5 56.368.5

6 5 5 8 6 6

a To quantify RNA loss after fixation, several Northern blots with an actin probe were scanned and then density of the actin bands was measured using a Phosphorlmager  . b Data are shown as mean6SEM and are expressed as percent of control. c Fixation times and ethanol concentrations are indicated.

To our knowledge, there is only one published study describing the isolation of mRNA from ethanol / acetic acid fixed hybridoma cells before cell sorting [1]. Routinely used fixatives such as ethanol and formaldehyde do not completely inactivate RNAses [1]. There are also a few reports about RT-PCR with RNA isolated from formalin fixed and paraffin-embedded tissue [4–6]. The product size ranged from 75 to 802 bp [7–9] and thus, not all met the criteria for full-size mRNA. It remains difficult to correctly judge these RNA preparations, because PCR usually needs just a few intact transcripts for amplification. Another paper shows PCR analysis from a pathologic specimen after different fixations in Omnifix (an ethanol based fixative), ethanol, Zenkers’, Bouin and B5 [4]. Formaldehyde fixation (0.05–4%) resulted in morphologically intact cells, which had to be lysed and homogenised mechanically. Our data indicate an immense RNA loss, especially larger RNA molecules were not detectable after formaldehyde fixation. In contrast, small tRNAs were present after fixation at various concentrations. One

Fig. 3. Northern blot and RT-PCR of b-actin from sorted cells. AKR-2B fibroblasts were fixed for 15 min with 75% DEPC-treated ice-cold ethanol and passaged through an Epics Elite ESP cell sorter. Thereafter, the RNA was isolated according to Ref. [2]. (A) b-actin Northern Blot of 10 mg RNA from FACS sorted cells. The size of the 28 and 18S rRNA was indicated as a size control. (B) b-actin RT-PCR of 1 mg RNA from FACS sorted cells from two independent experiments. The appropriated 227 bp b-actin band was indicated with an arrow.

C. Diez et al. / J. Biochem. Biophys. Methods 40 (1999) 69 – 80

77

Fig. 4. RT-PCR from sorted cells with high or low PCNA content. (A) Shows a density plot of AKR-2B cells in different phases of the cell cycle. The plot shows DNA-fluorescence on the x-axis against the PCNAfluorescence on the y-axis. (B) Anti-PCNA stained AKR-2B cells were analysed by FACS. Cells with low PCNA content (M1 ) and cells with high PCNA content (M2 ) were sorted and collected. (C) PCR analysis of cells with low and high PCNA content as described under B. Shown are the cell cycle markers cyclin A, cdk2, cdk4 and as an internal standard GAPDH and two size-standards. The cell cycle dependent cdk2 and cyclin A are expressed in cells with high PCNA content, whereas cdk4 and GAPDH remained unchanged independently of the PCNA content.

78

C. Diez et al. / J. Biochem. Biophys. Methods 40 (1999) 69 – 80

explanation for this finding might be due to cross-linked RNA molecules, especially 18S and 28S RNA, to proteins. This might lead to RNA loss during purification. A possible degradation can be excluded as we cannot detect the typical pattern of degradation after denaturing formaldehyde gel electrophoresis. Ethanol fixation proved to be the method of choice for fixation before subsequent sorting procedures. Ethanol neither affects secondary nor tertiary protein structures and thus allows to work with histochemical methods in parallel. As pointed out by Ref. [1], ethanol fixation leads to a time-dependent increase in loss of macromolecules through pores in the cell membrane as well as to rapid RNA degradation by active or insufficiently inactivated ribonucleases. In contrast to Esser et al. [1], we do not observe significant differences of RNA integrity after 5 and 15 min fixation with 75% ethanol, respectively. One explanation for this finding might be the use of DEPC instead of ribonucleoside vanadyl complexes to inactivate ribonucleases. While the latter is a transition state analogue and inactivates RNAse A, DEPC inactivates besides RNAse A, RNAse B and RNAse C, and RNAse H as well. Whereas higher ethanol concentrations and longer fixation times might cause more or larger membrane holes leading to an increased RNA loss, lower ethanol concentrations yielded much more degraded RNA inconvenient for Northern blotting. There seems to be a difference between mRNA and complete RNA loss. Ribosomal RNA content is comparable with the control after fixation, probing with b-actin revealed a 40% loss of mRNA. Perhaps, smaller mRNA molecules cross the leaky cell membrane earlier than larger rRNA species. As fixed cells are ‘sticky’, i.e. they easily adhere to plastic or glass surfaces, the addition of appropriate additives is necessary to prevent adhesion, which would result in immense cell loss. Cell sorting augments adhesion due to electrostatic loading during sorter passage. Whereas Esser et al. [1] found tryptophane, a mixture of amino acids and salts, adequate, we used 0.5% RNAse-free BSA and 2 mM CaCl 2 . All additives did not alter RNA integrity and were necessary to pellet the sorted cells. The RNA yield of about 60% in this study is 50% better than in the comparable study from Ref. [1]. The RNA yield and the cell recovery ratio are important parameters before planning the experiments. Our results after PCNA staining indeed indicate that the method described here provides reliable results.

4. Conclusion By combining several methodological approaches, we found adequate conditions to sort and isolate intact RNA from fixed cells. RNA yield and cell recovery ratio heavily based upon preceding ethanol fixation, which also allows histochemical staining for cytoplasmic and nuclear antigens. In summary, our protocol comprised the following steps: 1. Rinsing the cell layers with 1 3 PBS. 2. Cell detachment with 2 ml 0.5% trypsin / EDTA in PBS for 2–5 min. 3. Rinsing the cells with 10 ml DEPC-treated 1 3 PBS.

C. Diez et al. / J. Biochem. Biophys. Methods 40 (1999) 69 – 80

79

4. Cell transfer to Falcon vials, centrifugation at 200 g for 5 min. 5. Re-suspension in 1 ml DEPC pre-treated 1 3 PBS. 6. Cell transfer into a syringe (with a 20G cannula) and injection into 3 ml DEPCtreated 100% ethanol (enables rapid fixation without cell clumping). 7. Fixation for 15 min on ice. 8. Centrifugation at 400 g for 5 min. 9. Re-suspension in DEPC-treated 1 3 PBS, cell sorting. 10. Collection of sorted cells into 2 mM CaCl 2 and 0.5% RNAse-free BSA containing PBS, transfer into 2 ml Eppendorf caps. 11. Spinning down at 13 000 g at 48C, cell lysis according to Ref. [2]. Further applications of this method will focus on gene expression within fixed and sorted single cells to investigate heterogeneity in tissues or cell culture. We have already applied this protocol to sort single cardiac myocytes from contaminated cell cultures to study gene expression within these cells (Diez and Simm, submitted). As most organs and tissues consist of at least of two different cell populations, gene expression studies in co-cultured cell models might be a further source to apply this method. Clinically important specimens from pleural or pericardial effusions might also be screened for malignant transformation, leading to better grading of cancer cells or the detection and exact determination of viral or bacterial infection.

Acknowledgements The technical expertise of Ms Sabine Ebeling and Ms Anja Scheidler is gratefully acknowledged. This work was supported by a grant from the Deutsche Forschungsgemeinschaft (SFB 355TP C2) and the Graduiertenkolleg ‘Regulation des Zellwachstums’.

References [1] Esser C, Gottlinger C, Kremer J, Hundeiker C, Radbruch A. Isolation of full-size mRNA from ethanol fixed cells after cellular immunofluorescence staining and fluorescence-activated cell sorting (FACS). Cytometry 1995;21:382–6. [2] Chomczynski P, Sacchi N. Single-step method of RNA isolation by acid guanidinium thiocyanate– phenol–chloroform extraction. Anal Biochem 1987;162:156–9. [3] Leicht M, Simm A, Bertsch G, Hoppe J. Okadaic acid induces cellular hypertrophy in AKR-2B fibroblasts: Involvement of the p70 S6 kinase in the onset of protein and rRNA synthesis. Cell Growth & Differentiation 1996;7:1199–209. [4] Ben-Ezra J, Johnson DA, Rossi J, Cook N, Wu A. Effect of fixation on the amplification of nucleic acids from paraffin-embedded material by the polymerase chain reaction. J Histochem Cytochem 1991;3:351– 4. [5] Finke J, Fritzen R, Ternes P, Lange W, Dolken G. An improved strategy and a useful housekeeping gene for RNA analysis from formalin-fixed, paraffin-embedded tissues by PCR. BioTechniques 1993;14:448– 53.

80

C. Diez et al. / J. Biochem. Biophys. Methods 40 (1999) 69 – 80

[6] Jiang YH, Davidson LA, Lupton JR, Chapkin RS. A rapid RT-PCR method for detection of intact RNA in formalin-fixed paraffin-embedded tissues. Nucleic Acid Res 1995;23:3071–2. [7] Foss RD, Guha-Thakurta N, Conran RM, Gutman P. Effects of fixative and fixation time on the extraction and polymerase chain reaction amplification of RNA from paraffin-embedded tissue. Comparison of two housekeeping gene mRNA controls. Diag Mol Pathol 1994;3:148–55. [8] Gruber AD, Greiser-Wilke IM, Haas L, Hewicker-Trautwein M, Moennig V. Detection of bovine viral diarrhea virus RNA in formalin-fixed, paraffin-embedded brain tissue by nested polymerase chain reaction. J Virol Methods 1993;43:309–20. [9] Rupp GM, Locker J. Purification and analysis of RNA from paraffin-embedded tissues. Biotechnology 1988;6:56–60. [10] Kang MJ, Koh GY. Differential and dramatic changes of cyclin-dependent kinase activities in cardiomyocytes during the neonatal period. J Mol Cell Cardiol 1997;29:1767–77. [11] Moore GD, Ayabe T, Kopf GS, Schultz RM. Temporal patterns of gene expression of GI-S cyclins and cdks during the first and second mitotic cell cycles in mouse embryos. Mol Reprod Dev 1996;45:264–75. [12] Schmid I, Uittenbogaart CH, Keld B, Giorgi JV. A rapid method for measuring apoptosis and dual-color immunofluorescence by single laser flow cytometry. J Immunol Methods 1994;170:145–57. [13] Belloc F, Dumain P, Boisseau MR, Jalloustre C, Reiffers J, Bernard P, Lacombe F. A flow cytometric method using Hoechst 33342 and propidium iodide for simultaneous cell cycle analysis and apoptosis determination in unfixed cells. Cytometry 1994;17:59–65. [14] Ferlini C, Di Cesare S, Rainaldi G, Malorni W, Samoggia P, Biselli R, Fattorossi A. Flow cytometric analysis of the early phases of apoptosis by cellular and nuclear techniques. Cytometry 1996;24:106–15.