Postmortem stability of RNA isolated from bovine reproductive tissues

Biochimica et Biophysica Acta 1574 (2002) 10^14

www.bba-direct.com

Postmortem stability of RNA isolated from bovine reproductive tissues Richard Fitzpatrick a; *, Orla M. Casey a;c , Dermot Morris a , Terry Smith b , Richard Powell c , Joseph M. Sreenan a a

Animal Reproduction Department, Teagasc, Agriculture and Food Development Authority, Athenry, Galway, Ireland b National Diagnostics Centre, BioResearch Ireland, National University of Ireland, Galway, Ireland c Department of Microbiology, National University of Ireland, Galway, Ireland Received 6 July 2001; received in revised form 25 September 2001 ; accepted 26 September 2001

Abstract Molecular biology is being increasingly used to address the complex problem of bovine infertility. One common concern shared by many of these studies is the postmortem delay in obtaining reproductive tissues and the effect this may have on RNA dependent studies. To address this concern, bovine ovarian, oviduct and uterine tissue samples, collected over intervals ranging from 0 to 96 h postmortem to freeze storage, were analysed to determine the potential effects on RNA quantity and quality. The analysis showed that total RNA yields were not changed significantly by postmortem interval up to 96 h while 28S ribosomal RNA remained intact up to 24 h postmortem. Specific messenger RNA transcripts encoding L-actin, GAPDH and transforming growth factor-L were detected in all tissues up to 96 h postmortem using reverse transcriptase^polymerase chain reaction and Northern analysis indicated no detectable mRNA degradation up to 24 h postmortem. Finally, using poly(A)þ mRNA isolated from ovarian tissues frozen 2 h postmortem, we constructed corpus luteum and ovarian cortex cDNA libraries containing 7.65U104 and 1.9U106 primary transformants with average cDNA lengths of 2.3 and 1.6 kb respectively. Taken together, these data show that a postmortem delay of up to 24 h does not significantly affect the yield or quality of RNA prepared from bovine reproductive tissues. ß 2002 Elsevier Science B.V. All rights reserved. Keywords : RNA stability ; Bovine; Ovary; Oviduct ; Uterus; Molecular biology

1. Introduction Successful genetic selection programs have resulted in a modern dairy cow, biologically e⁄cient at producing large volumes of milk but with a signi¢cantly increased early embryo loss rate [1]. To address this problem, molecular biology techniques are now being used to analyse bovine embryos and reproductive tissues including the ovary, oviduct and uterus in order to identify the molecular components and mechanisms involved in embryo loss. Most of these approaches depend on the detection and measurement of speci¢c mRNA transcripts and include Northern analysis, reverse transcriptase^polymerase chain reaction (RT^PCR), construction of cDNA libraries and more recently DNA microarray analysis [2^5]. Because access to such tissues is usually postmortem, a common concern potentially a¡ecting the outcome of such investigations is the delay in obtaining tissue from experimental animals

* Corresponding author. Fax: +353-91-845847. E-mail address : r¢[email protected] (R. Fitzpatrick).

postmortem. Given the complexity of the messenger RNA (mRNA) population expressed within cells and the many factors that a¡ect mRNA stability [6,7], it is desirable that the tissues used for analysis be recovered and snap frozen as rapidly as possible. However, postmortem intervals to tissue collection are variable and often exceed 1 h. This delay may be further prolonged when other procedures such as oviduct and uterine £ushing or embryo or follicle recovery are performed. Therefore, it is important to determine the e¡ect(s) of postmortem interval to tissue recovery and freeze storage on RNA integrity. The issue of postmortem RNA stability in human tissues including brain, lung, liver, heart and kidney has been addressed [8^11]. In most cases extensive RNA stability was observed up to 38^48 h postmortem. Similar studies have also been conducted with ligament, tendon and cartilage tissue where neither degradation of ribosomal RNA (rRNA) nor loss of integrity of mRNA were observed up to 96 h postmortem [12]. In tissues suspected of producing large amounts of RNase such as the stomach or pancreas, RNA stability is more problematic. Consequently, it proved necessary to isolate RNA from these

0167-4781 / 02 / $ ^ see front matter ß 2002 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 7 - 4 7 8 1 ( 0 1 ) 0 0 3 2 2 - 0

BBAEXP 93580 8-4-02

R. Fitzpatrick et al. / Biochimica et Biophysica Acta 1574 (2002) 10^14

tissues within 2 h postmortem to prevent RNA degradation [13]. Little is known about postmortem RNA stability in reproductive tissues. We investigated the e¡ects of intervals from 0 to 96 h from tissue recovery to freeze-storage on the quantity and quality of total RNA isolated from bovine ovarian, oviduct and uterine tissues. Subsequently, the presence and integrity of three speci¢c mRNA transcripts encoded by the bovine L-actin, GAPDH and transforming growth factor-L (TGFL) genes were determined by both RT^PCR and Northern hybridisation analyses. Finally, we constructed ovarian cortex and corpus luteum cDNA libraries from tissues frozen 2 h after tissue recovery and assessed their quality in terms of cDNA length.

2. Materials and methods 2.1. Tissue sampling In order to precisely time the interval from tissue recovery to freeze storage, tissues were obtained from cattle during mid-ventral laparotomy. Cross bred heifers were anaesthetised with thiopentone sodium (5 g i.v. ; Rhone Merieux, Harlow, UK) and maintained by inhalation of halothane (May and Baker, Dagenham, UK). Samples of ovarian, oviduct and uterine tissues were surgically removed in accordance with the European Community Directive, 86-609-EC. To simulate postmortem intervals, tissues were trimmed free of excess adipose tissue, rinsed in sterile RNase-free phosphate bu¡er, blotted on sterile sponges and stored at room temperature for 0, 0.5, 1, 24, 48, 72 and 96 h. Tissues were then snap-frozen in liquid nitrogen and stored at 380‡C until RNA extraction. 2.2. RNA preparation and RT^PCR analysis Total RNA was prepared from 200^300 mg of fragmented frozen tissue and homogenised in TRIzol, 1 ml per 100 mg of tissue (Gibco BRL Life Technologies, Renfrewshire, UK) using a Heidolph Diax 900 homogeniser. The same amount of tissue was used for all tissue types and at all time points. 0.2 ml of chloroform per ml of TRIzol was added, samples were mixed well and centrifuged at 12 000Ug for 15 min at 4‡C. The upper aqueous

11

phase containing the RNA was removed and precipitated with 0.6 volumes of isopropanol. The RNA pellet was then washed in 1 ml of 75% ethanol, air dried for 5^10 min and redissolved in RNase-free water. The quantity of RNA was determined by absorbance at 260 nm and the degree of protein contamination assessed by the ratio of absorbance at 260 and 280 nm. Samples were stored at 380‡C until further analysis. RNA integrity was assessed by examining the 28S and 18S rRNA transcripts on ethidium bromide-stained agarose formaldehyde gels. For RT^ PCR analysis, samples were treated with RNase-free DNase (Gibco BRL) for 15 min at room temperature. The reaction was stopped by the addition of 25 mM EDTA and heating to 65‡C for 10 min. One Wg of total RNA was reverse transcribed using M-MLV reverse transcriptase (Promega, WI, USA) and 500 ng of oligo(dT)15 in a 50-Wl reaction. Four-Wl aliquots from the ¢rst strand cDNA synthesis reaction were then ampli¢ed in a 100 Wl volume containing 1U PCR bu¡er, 0.2 WM dNTP mixture, 1.5 mM MgCl2 , 100 ng of each primer and 2.5 U of Taq DNA polymerase (Promega). Thirty-¢ve cycles of PCR consisting of 94‡C for 30 s, 60‡C for 30 s and 72‡C for 1 min were performed. Non-reverse transcribed total RNA was included as a control for the presence of genomic DNA contaminants. The identity of the PCR products was veri¢ed by restriction analysis and agarose gel electrophoresis. Gels were stained with ethidium bromide and visualised using a Molecular Imager FX (BioRad). The primer sequences speci¢c for bovine GAPDH, TGFL and L-actin genes were obtained from GenBank and are shown in Table 1. 2.3. Northern blot and hybridisation analysis Total RNA (15 Wg), pooled from three animals, was electrophoresed through a denaturing 2.2 M formaldehyde 1% agarose gel in 1U MOPS bu¡er [14] and transferred to nylon membranes (Schleicher and Schuell, Dassel, Germany) with 5U SSC [15]. Membranes were baked at 80‡C for 2 h and stored dry until hybridisation. Blots were prehybridised for 4 h at 65‡C in a solution containing 6U SSC, 5U Denhardt’s, 0.5% SDS and 50 Wg/ml denatured salmon sperm DNA. Membranes were hybridised overnight at 62‡C with 32 P-radiolabelled DNA fragments of GAPDH (858 bp) and L-actin (226 bp). DNA probes

Table 1 Bovine speci¢c primer sequences and expected size of ampli¢ed products Primers

Sequences

Size (bp)

GAPDHf GAPDHr L-actinf L-actinr TGF-L1f TGF-L1r

5P-CCTTCATTGACCTTCACTACATGGTCTA-3P 5P-TGCTGTAGCCAAATTCATTGTCGTACCA-3P 5P-CGTGGGCCGCCCTAGGCACCA-3P 5P-GGGGGCCTCGGTCAGCAGCAC-3P 5P-GCTCCACAGAAAAGAACTGCT-3P 5P-AGGAGCGCACGATCATGTTGG-3P

858

Primer sequences were derived from bovine gene sequences found in GenBank.

BBAEXP 93580 8-4-02

226 300

12

R. Fitzpatrick et al. / Biochimica et Biophysica Acta 1574 (2002) 10^14

Table 2 Total RNA yield (Wg/mg) for postmortem tissues

Ovary Oviduct Uterus

0h

0.5 h

1h

24 h

48 h

72 h

96 h

1.88 þ 0.601* 1.04 þ 0.455 1.20 þ 0.516*

2.26 þ 0.533* 1.04 þ 0.220 1.49 þ 0.355

2.49 þ 0.320 1.25 þ 0.526 1.25 þ 0.412

1.76 þ 0.407 1.30 þ 0.233 1.38 þ 0.528

2.26 þ 0.503 1.23 þ 0.579 0.97 þ 0.230

2.00 þ 0.250 1.11 þ 0.350 1.29 þ 0.220

1.93 þ 0.420 1.20 þ 0.430 1.12 þ 0.539

n = 3; *n = 2.

were generated using the High Prime labelling kit (Boehringer Mannheim, Germany) according to manufacturer instructions. Membranes were washed three times for 30 min in 0.2U SSC, 0.1% SDS at 65‡C. Membranes were exposed to imaging plates (Kodac) for 24 h and scanned using a Molecular Imager FX (Bio-Rad). 2.4. Construction and evaluation of cDNA libraries Corpus luteum tissue from heifers at day 6, 8 and day 14 of the oestrous cycle (n = 3) and ovarian cortex tissue from the same animals (and also from a heifer at day 2 of the oestrus cycle) were frozen in liquid nitrogen 2 h postmortem. Total RNA was extracted as described above, pooled and used to construct cDNA libraries using the VZap Express cDNA synthesis/Gigapack III Gold cloning kit (Stratagene Cloning Systems, CA, USA). cDNA synthesis was performed using an oligo(dT)18 primer for the reverse transcription of approximately 5 Wg of mRNA. Average cDNA clone length was established after in vivo excision of 12 random clones into pBK-CMV phagemid form according to manufacturer instructions (Stratagene). Excised clones were grown overnight in LB containing kanamycin (50 Wg/ml) and phagemid DNA was prepared from 3 ml of overnight bacterial culture using a High Pure Plasmid Isolation kit (Roche Diagnostics, Mannheim, Germany). The cloned cDNA inserts were released by digestion with the restriction enzymes PstI and XbaI and analysed by electrophoresis through 1% agarose gels [15]. The lengths of the cloned cDNA fragments were determined by visual comparison to VHindIII DNA length standards (New England Biolabs, MA, USA) and 100 bp ladder (Amersham Pharmacia Biotech, NJ, USA).

three reproductive tissues analysed. The average RNA yields from ovarian, oviduct and uterine tissue frozen 96 h postmortem were 1.93, 1.2 and 1.12 Wg/mg, respectively, and these were similar to the respective yields of 1.88, 1.04 and 1.2 Wg/mg obtained from immediately frozen tissues. With respect to RNA purity, typically 260:280 nm ratios of 1.7 were consistently obtained for all tissue types and at all time points. With respect to RNA quality, all the RNA preparations were examined by agarose gel electrophoresis for the presence of intact 18S and 28S rRNA transcripts. The results showed that intact 28S and 18S rRNA transcripts were obtained for all tissues up to 24 h postmortem, while the 48^96 h samples showed a decreased intensity for the 28S rRNA signal accompanied by lower molecular bands that are indicative of RNA degradation (Fig. 1). 3.2. RT^PCR analysis The presence of speci¢c mRNA transcripts encoding bovine GAPDH, and L-actin (transcribed from two high

3. Results 3.1. RNA analysis To simulate postmortem, bovine reproductive tissues were surgically removed, stored at room temperature for intervals of 0^96 h and subsequently snap frozen in liquid nitrogen as described in Section 2. The yields of total RNA obtained from bovine ovarian, oviduct and uterine tissues, analysed over the postmortem interval range of 0^ 96 h are shown in Table 2. There were no signi¢cant changes in RNA yield, as quanti¢ed by spectrophotometry, throughout the postmortem interval in any of the

Fig. 1. Ethidium bromide-stained gel of total RNA isolated from postmortem tissue stored at room temperature for increasing periods. (A) ovary; (B) oviduct; (C) uterus. Lanes: 1, 0 h; 2, 0.5 h; 3, 1 h; 4, 24 h; 5, 48 h; 6, 72 h; 7, 96 h.

BBAEXP 93580 8-4-02

R. Fitzpatrick et al. / Biochimica et Biophysica Acta 1574 (2002) 10^14

13

copy number house-keeping genes) and TGFL (transcribed from a low copy number gene) was assessed by RT^PCR. Products of the expected lengths, i.e., 858 bp for GAPDH, 226 bp for L-actin and 300 bp for TGFL, were successfully ampli¢ed by RT^PCR from RNA isolated from all three tissues sampled at 0, 48 and 96 h postmortem. Fig. 2. shows the RT^PCR results obtained from the uterine RNA preparations. The ampli¢ed products were shown to be derived from reverse transcribed mRNA and not from genomic DNA by completing parallel reactions with total RNA that had not being reverse transcribed. While quantitative PCR analysis was not performed, there was an apparent decline in ampli¢ed product correlating with increased postmortem time. 3.3. Northern blot and hybridisation analysis To further assess mRNA integrity, 15 Wg of total RNA isolated from ovarian, oviduct and uterine tissues, were transferred from agarose gels onto nylon membranes and screened with the 858 bp DNA fragment of the bovine GAPDH gene. A single mRNA transcript of the expected length of 1.35 kb was detected in all samples from all three bovine tissues (Fig. 3). Evidence of mRNA degradation was apparent in the 48^96 h postmortem samples and was accompanied by a reduction in signal intensity. Densitometric analysis of the RNA signals indicated a relative reduction in undegraded mRNA of approximately 0.79fold in the 48 h postmortem samples, and of approximately 2.0-fold in the 72^96 h postmortem samples. Similar results were also obtained when these RNA samples were screened with a L-actin-derived DNA probe (data not shown).

Fig. 3. Northern blot analysis of GAPDH for postmortem tissues stored at room temperature for increasing periods. (A) ovary; (B) oviduct; (C) uterus. Lanes: 1, 0 h; 2, 0.5 h; 3, 1 h; 4, 24 h; 5, 48 h; 6, 72 h; 7, 96 h.

parental background of 6 1%. Analysis of 12 randomly chosen clones from each cDNA library indicated that the average corpus luteum cDNA length was 2.3 kb and the longest was 4.5 kb, while the average ovarian cortex cDNA length was 1.6 kb and the longest was 3.2 kb (Fig. 4). These data indicate that the mRNA populations were not unduly composed of short-length degraded molecules and also that the transcripts contained su⁄ciently polyadenylated 3P-termini.

4. Discussion 3.4. cDNA library analysis As a more general assessment of mRNA stability, corpus luteum and ovarian cortex cDNA libraries were constructed from mRNA prepared from tissue dissected and frozen 2 h postmortem. The corpus luteum cDNA library contained 7.65U104 primary transformants with a parental background of 6 2%. The ovarian cortex cDNA library contained 1.9U106 primary transformants with a

Fig. 2. RT^PCR analysis of 1 Wg of total RNA isolated from uterine tissue stored at room temperature for 0, 48 and 96 h. Lanes: 1, 0 h, TGFL; 2, 48 h, TGFL; 3, 96 h, TGFL; 4, 0 h, GAPDH ; 5, 48 h, GAPDH ; 6, 96 h, GAPDH; 7, 0 h, L-actin; 8, 48 h, L-actin, 9, 96 h, Lactin.

Postmortem stability of RNA isolated from bovine reproductive tissues was assessed in terms of recovery yields, rRNA stability, integrity of speci¢c mRNA transcripts and the ability to construct quality cDNA libraries. The results show that yields of RNA isolated from bovine ovarian, oviduct and uterine tissues were not signi¢cantly a¡ected by postmortem delay. Similarly, postmortem intervals up to 24 h had little a¡ect on RNA quality. Ribosomal RNA degradation was only apparent in the 48^96 h postmortem samples, while RT^PCR and Northern analyses indicated stability of speci¢c mRNA transcripts up to 24 h postmortem in all of the tissues analysed. A decline in L-actin and GAPDH mRNA stability was apparent in the 48^96 h postmortem samples as judged by reduced Northern hybridisation signals. However, speci¢c mRNA transcripts encoding L-actin, GAPDH and TGFL were still detected by RT^PCR. Accordingly, RNA isolated from the 48^96 h postmortem tissues may still be of su⁄cient quality to enable qualitative gene expression studies which should be of general interest, particularly to human research where for practi-

BBAEXP 93580 8-4-02

14

R. Fitzpatrick et al. / Biochimica et Biophysica Acta 1574 (2002) 10^14

that other mRNAs could be adversely a¡ected by postmortem interval. However, in agreement with Marchuk et al. [12] our results indicate that there is not a general activation of cellular nucleases postmortem. This is supported by the ability to construct quality corpus luteum and ovarian cortex cDNA libraries containing 7.65U104 and 1.9U106 primary transformants with an average cDNA length of 2.3 and 1.6 kb, respectively, from RNA prepared from tissues frozen 2 h after animal death. Taken together, these results indicate that bovine reproductive tissues collected within 24 h postmortem su¡er only modestly in terms of RNA quantity and quality and provides further evidence implicating the appropriate use of postmortem tissues for molecular biology research.

Acknowledgements O.M.C. is currently supported by the Teagasc Walsh postgraduate fellowship scheme.

References

Fig. 4. Analysis of (A) corpus luteum and (B) ovarian cortex cDNA insert length. Lane 1, VHindIII; lanes 2^13, respective cDNA clones digested with PstI and XbaI; lane 14, 100 bp ladder. Upper bands represent vector, lower bands represent cDNA inserts.

cal and ethical reasons postmortem intervals of 3^5 days are common. Extensive RNA stability for extended postmortem intervals has also been demonstrated in several human tissues including the brain, liver, lung, heart and kidney [8^11]. However, Humphreys-Beher et al. [13] have shown that time postmortem is a critical factor in the recovery of intact RNA from stomach and pancreatic tissue. By comparison, no di¡erence in postmortem RNA stability was observed between ovarian, oviduct and uterine tissues, despite their diverse morphological and physiological characteristics. While only a limited number of mRNA transcripts were investigated in this study, and given the complexity of mRNA population expressed within cells, it is possible

[1] W.R. Butler, J. Dairy Sci. 81 (1998) 2533^2539. [2] G.A. Johnson, K.J. Austin, A.M. Collins, W.J. Murdoch, T.R. Hansen, Endocrine 10 (1999) 243^252. [3] H. Niemann, C. Wrenzychi, Theriogenology 53 (2000) 21^34. [4] T.P. Smith, W.M. Grosse, B.A. Freking, A.J. Roberts, R.T. Stone, E. Casas, J.E. Wray, J. White, J. Cho, S.C. Fahrenkrug, G.L. Bennett, M.P. Heaton, W.W. Laegreid, G.A. Rohrer, C.G. Chitko-McKown, G. Pertea, I. Holt, S. Karamycheva, F. Liang, J. Quackenbush, J.W. Keele, Genome Res. 11 (2001) 626^630. [5] R.S. Hines, Semin. Reprod. Med. 18 (2000) 91^96. [6] P. Mitchell, D. Tollervey, Curr. Opin. Genet. Dev. 10 (2000) 193^ 198. [7] J. Ross, Microbiol. Rev. 59 (1995) 423^450. [8] S. Larsen, K. Rygaard, S. Asnaes, M. Spang-Thomsen, APMIS 100 (1992) 498^502. [9] G.N. Davies, I.S. Bevan, J.B. Lundemose, H. Smith, C. Sweet, J. Clin. Pathol. Mol. Pathol. 49 (1996) 364^367. [10] A.J.L. Barton, R.C.A. Pearson, A. Najlerahim, P.J. Harrison, J. Neurochem. 61 (1993) 1^11. [11] M. Schramm, P. Falkai, R. Tepest, T. Schneider-Axmann, R. Przkora, A. Waha, T. Pietsch, W. Bonte, T.A. Bayer, J. Neural. Transm. 106 (1999) 329^335. [12] L. Marchuk, P. Sciore, C. Reno, C.B. Frank, D.A. Hart, Biochim. Biophys. Acta 1379 (1998) 171^177. [13] M.G. Humphreys-Beher, F.K. King, B. Bunnel, B. Brody, Biotechnol. Appl. Biochem. 8 (1986) 392^403. [14] H. Lehrach, D. Diamond, J.M. Wozney, H. Boedtkey, Biochemistry 16 (1977) 4743^4751. [15] J. Sambrook, E.F. Fritsch, T. Maniatis, Molecular Cloning : A Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, 1989.

BBAEXP 93580 8-4-02