Molecular and Cellular Biochemistry 139: 149-157, 1994. 9 1994 Kluwer Academic Publishers. Printed in the Netherlands. The choice of resuspension medium ...
Molecular and Cellular Biochemistry 139: 149-157, 1994. 9 1994 Kluwer Academic Publishers. Printed in the Netherlands.
The choice of resuspension medium for isolated rat liver nuclei: Effects on nuclear morphology and in vitro transcription Roger Strand 1, Roald Boe 2 and Torgeir Flatmark 1 1Department of Biochemistry and Molecular Biology and 2 Department of Anatomy and Cell Biology, University of Bergen, Bergen, Norway Received 8March 1994; accepted 30 June 1994
Abstract Standard protocols for in vitro transcription assay (nuclear run-off) include 10--40% (v/v) glycerol (of various ionic strength) in the medium used for resuspension/storage of the isolated nuclei. In the present work the morphological and functional properties of nuclei isolated from rat liver have been studied as a function of the content of glycerol, sucrose and inorganic ions (K + and Mg z§ in the resuspension medium. In contrast to earlier reports, glycerol was found not to be essential to maintain morphological integrity and RNA polymerase activity in frozen/stored nuclei. Nuclear pellets, resuspended and stored in isoosmotic sucrose media, were found to give morphologically intact and transcriptionally active nuclei. Furthermore, these nuclei displayed a higher specific hybridization signal for the differentially expressed genes encoding peroxisomal fl-oxidation enzymes, relative to the total RNA synthesis, than nuclei resuspended and stored in a hyperosmotic glycerol-containing medium. The concentrations of inorganic ions were also found to affect nuclear morphology. Flow cytometry indicated DNA leakage from nuclei at insufficient concentrations o f K § and Mg 2§ and high ionic strength favoured aggregation and disintegration of nuclei. Our findings indicate that quantitative results from nuclear run-off experiments should be interpreted with caution until the process of transcription in isolated nuclei is better understood. (Mol Cell Biochem 139:149-157, 1994)
Key words: isolated nuclei, nuclear run-off, glycerol, storage media, peroxisomes, rat liver
Introduction Gene expression in eukaryotic ceils is regulated at the level of transcription by several mechanisms, including the rate of transcription [ 1], post-transcriptional processing or premRNA (hnRNA) [2], degradation of cytoplasmic translatable mRNA [3-4] and by antisense RNA [5].An increased steadystate level of specific mRNAs is often observed as a result of a transcriptional upregulation of gene expression, but the mechanism of this increase is experimentally more difficult to define. The method of in vitro transcription in isolated nuclei (nuclear run-off) aims at providing information on transcriptional regulation [6]. The essential parts of an in vitro
transcription assay are the following: Nuclei are isolated from the organ or cell culture of interest. The resuspended, highly purified nuclei are incubated at high concentrations of the ribonucleoside triphosphates ATP, CTP, GTP and UTP, one of which (normally UTP) is labelled, in order to synthesize labelled RNA. Nuclear RNA is then isolated and hybridized against a membrane-bound cDNA or antisense probe of interest and the label is quantified. The assay is termed nuclear run-off as little or no transcriptional initiation is considered to occur after cell disruption, and the RNA synthesis is thought of as a result of already template-bound RNA polymerase complexes 'running off' their DNA templates (i.e. elongation and termination) [7], although the process
Address for offprints: T. Flatmark, Dept of Biochemistry and Molecular Biology, University of Bergen, ~,rstadveien 19, N-5009 Bergen, Norway
150 of transcription in isolated nuclei is poorly characterized. The nuclear run-off method has become a standard technique and was applied in more than 200 publications of 1993 only, according to the Medline| database. Assay conditions vary, but all laboratory handbooks and scientific papers known to us, describing nuclear run-off conditions, recommend the presence of 10-30% (v/v) glycerol during the ribonucleotide incubation [6-12]. Usually, the glycerol is added at the resuspension of the freshly isolated nuclear pellet (in concentrations up to 40% (v/v)), and the addition seems to he motivated by the convenience of storing frozen nuclei until further processing. According to older literature, glycerol is considered essential to maintain morphological integrity and RNA polymerase activity in frozen nuclei [ 13-15]. While performing nuclear run-offs, we were uneased by the recommended high concentrations of glycerol in the nuclear samples. An alternative to glycerol-containing nuclear resuspension media is isoosmotic sucrose media, which has been recommended for enzymological studies on isolated nuclei [ 16-17]. In the present study, morphological evidence is presented to show that sucrose media are indeed suitable as storage media for frozen nuclei. Furthermore, we demonstrate qualitative and quantitative differences between the in vitro transcription when isolated nuclei are resuspended/ stored in glycerol and sucrose media. We have performed nuclear run-offs in a study of the induction of peroxisomal l-oxidation enzymes in the rat liver. The induction of these enzymes upon administration of hypolipidemic compounds ('peroxisome proliferators') represents a well-characterized model system of highly potent transcription activation [ 18--21 ]. Thus, the study of these genes offers a system of easily reproducible >20-fold increases in steady-state mRNA levels. A comparison of an isoosmotic sucrose medium and a hyperosmotic glycerol medium (i.e. a medium with a significantly higher concentration of solutes than 290 mosmol 1-t) for nuclear resuspension/storage revealed that the sucrose-stored nuclei had a higher specific hybridization signal for peroxisoma113-oxidationenzyme genes relative to the total RNA synthesis. This observation indicates that the choice of nuclear resuspension medium affects in vitro transcription in a qualitative as well as a quantitative manner. We have also focused on the concentrations of inorganic ions in the resuspension/storage medium. Previous reports are highly divergent on the question of its recommended ionic strength and composition, ranging from tris buffers with 1 mM MgC1z [13] to storage media with twice the isoosmotic ionic strength (280 mM KC1, 20 mM MgC12and tris buffer) in addition to glycerol [10]. Little morphological evidence has, however, been given to support the recommendations of the alternative media. In the present study it is shown that nuclei are subject to damage at too low or too high concentrations of inorganic ions.
Materials and methods Chemicals and animals
For in vitro transcription experiments nuclease-free sucrose was used (density gradient nuclease-free grade, E. Merck, Darmstadt, Germany). Ultrapure phenol was delivered by Gibco BRL (Gaithersburg, US). 25% (w/v) glutaraldehyde solution (electron microscopy grade) was from E. Merck (Darmstadt, Germany). Osmium tetroxide of 99.99% purity was from Taab (Reading, UK). Isopropanol (HPLC grade) was from Rathburn Chemicals Ltd (Walkerburn, UK). Repelcote was from Hopkin and Williams (Essex, UK). RNase I 'A' was from Pharmacia LKB (Uppsala, Sweden). The restriction enzyme Pst I was delivered by Gibco BRL and Amersham Intl. (Little Chalfont, UK). [ct-32P]-UTP (>400 Ci mol-1) was obtained as a triethylammoniumsalt (Amersham Intl; Prod. No. PB 10163). All other chemicals were of analytical grade unless otherwise stated. Nytal nylon monofilaments were manufactured by SST (Thai, Switzerland). For blotting, a Hoefer PR600 (San Francisco, US) slot blot apparatus together with Trans-Blot 0.45 ~tmnitrocellulose membranes (Bio-Rad Lab., Richmond, US) and GeneScreen 0.2 ~tm membranes (New England Nuclear, Boston, Mass., US) were applied. Wistar rats (Mol:Wist) were from Mollegaard (Ejby, Denmark) and fed a standard pellet (R3 from Ewos A/B (Srdert~ilje, Sweden)) ad libitum. In the in vitro transcription study, male rats weighing 87 + 5 g at the onset ofclofibrate administration were given subcutaneous injections ofclofibrate (ICI Industrial Co., Macclesfield, UK; 65 mg/100 g body weight) every day for 11 days as described elsewhere [22]. Control animals were fed the standard pellet only. For supplemental morphological studies rats of both genders, weighing 200-400 g, were utilized.
Preparation o f isolated rat liver nuclei
Rat liver nuclei were isolated as described by Blobel and Potter [23], with the following minor modifications: The liver was minced thoroughly with scissors in 2 volumes of TKM/ 0.25 M sucrose medium (the TKM medium is 25 mM KCI, 5 mM MgC12 and 50 mM tris-HC1, pH 7.50 at 20~ and homogenized in a Potter-Elvehjem homogenizer with a motor-driven pestle (3-4 stokes at 700 rpm), immersing the homogenizer glass in an ice-water mixture at all times. The homogenate was filtered through one layer of Nytal monofilaments (pore size 100 Ixm) with a recovery of more than 80%. Sedimentation was performed as described by Blobel and Potter [23]. In order to obtain single-nuclei suspensions, the
151 nuclear pellet was carefully resuspended by dropwise addition of the resuspension medium and gentle mixing with a glass rod at 0--4~ Higher temperatures, addition of large portions of resuspension medium, as well as pipetting or vortexing, resulted in massive aggregation of the nuclei.
Storage and thawing of isolated nuclei Samples of isolated rat liver nuclei were immersed into liquid nitrogen within minutes after resuspension and stored in liquid nitrogen. The following thawing procedure was found useful to prevent aggregation or disintegration of the nuclei: The samples were transferred from liquid nitrogen to--80~ for 15 min, then to-34~ for 15 min, and finally 0-4~ until complete thawing. It was found essential to avoid mechanical disturbances during the actual thawing at 0--4~ Full transcriptional activity in the nuclei was maintained after at least 6 months of storage. Morphological intactness, as judged by flow cytometry and light microscopy, was maintained after 24 months of storage.
Morphological studies Transmission electron microscopy (TEM) was performed on a Philips EM 300 microscope. Samples of isolated nuclei were prepared for TEM as described by Reynolds [24]. Flow cytometry was performed on a Skatron Argus 100 Flow Cytometer (Tranby, Norway). Samples of isolated nuclei were stained with ethidium bromide (final concentration 100 lag ml-~) and left on ice in the dark for at least 15 min before analysis. The green band (530-550 nm) of a mercury arc lamp served as excitation light, and integrated (2 las) red fluorescence (600-720 nm) was recorded for at least 5000 nuclei per sample.
In vitro transcription assay (nuclear run-of39 Glassware, plastic tubes and aqueous solutions were treated with the RNase inhibitor diethylpyrocarbonate as described elsewhere [25], except for the tris-solutions, nucleotides and 'solution D' (see below). Prior to diethylpyrocarbonate-treatment, the glassware was siliconized with Repelcote. Incorporation of [a-32P]-UTP into RNA of isolated nuclei was performed essentially as described by Nevins [10]. Aliquots of 100 lal thawed nuclei suspension were incubated with 100 tal reaction mixture at 30~ for 25 min. 100 lal reaction mixture (made just before the experiment began) contained I00 laCi [~-32p]-UTP, 0.2 lamol ATP, CTP and GTP, 4 units of ereatine kinase, 2 lamoles phosphocreatine and 0.1 ~tmolDTT, pH 7.5. Reaction was quenched by the addition of 0.9 ml of
'solution D' of Chomczynski and Sacchi [26] (4 M guanidinium thiocyanate, 0.5% (v/v) sodium N-lauroylsarcosine, 0.1 M 2-thioethanol and 25 mM sodium citrate, pH 7.0); this led to the immediate disintegration of the nuclei. RNA was isolated and purified as described by Chomczynski and Saachi [26] with minor modifications: The quenched samples were transferred to 15 ml disposable plastic tubes (with tight screw caps) containing 0.9 ml solution D, 100 ~12M NaAc, pH 4.0 and 32 lag E. Coli tRNA (which improved the yield). 2.0 ml ultrapure phenol and then 1.0 ml chloroform-isoamyl alcohol (49:1) was added. Upon addition, the capped tube was inverted once and left for 2-5 min before release of pressure (to avoid spills). The capped tube was then vigorously shaken for 10 sec and the content was poured into a 15 ml Corex tube kept on ice. RNA precipitation was done with isopropanol and for 8-15 h at -34~ Supernatants were not poured off but discarded by pipetting, since RNA pellets do not adhere to the walls of siliconized glass or plastic. The step of precipitation in Eppendorftubes was repeated once. The cpm of the final supernatant was less than 1% of the cpm of the RNA pellet. The final pellet of RNA was dissolved in 240-300 ~tl ETS (10 mM EDTA, 10 mM tris-HCl, 0.2% (w/v) SDS, pH 7.4) at 56~ Aliquots were "taken for scintillation counting, and the RNA samples were stored at -34~ Hybridization of the [3~P]-labelled nuclear RNA to membrane-bound cDNA probes was accomplished according to Greenberg [11] for 28--40 h with shaking at 65~ with 1.0 x 106 cpm [32p]-RNA in a total hybridization volume of 500 lal per plasmid-bound strip and scintillation vial. After washing twice with 2x SSC (60 rain, 65~ RNase I 'A' treatment (10 lag m1-1 RNase in 2x SSC for 35 min, 37~ and a final wash (2x SSC, 60 min, 37~ strips were carefully blotted, placed onto double-sided Scotch tape (with the DNA facing upwards), wrapped in Saran wrap and autoradiographed for several days at---80~ Occasional small spots along the pictured slots on the autoradiograms were considered to be small precipitates, possibly [azp]-RNA/SDS. Thus, densitometry of autoradiograms combined with inspection (in order to exclude 'false' signals) was preferred to liquid scintillation counting of the filters. Densitometry was performed at )~ = 633 nm on an LKB UltroScan Densitometer (Pharmacia LKB, Uppsala, Sweden). Each filter was scanned at four sites along the slot, and the mean optical density was calculated. The recombinant plasmids pMJ 115 [27], pMJ 26 [28] and plivS 6 [29] were used as cDNA probes. These plasmids are pBR 322-derived plasmids with inserts at the Pst I site (inside the ampicilin resistance gene). The 2.8 kb insert in pMJ 115 contains a sequence encoding 330 amino acid residues at the C-terminal end of the peroxisomal fatty acyl-CoA oxidase, pMJ 26 has a 2.3 kb insert that includes a sequence encoding all but the six N-terminal amino acid residues (i.e. aa 7-722) of the peroxisomai trifunctional protein (A3,Az-
152 I
I
I
I
I
A
I
I
I
I
I
I
B
o~
J I
I
[
I
II
C
D i
I
Fluorescence intensity Fig. 1. Ethidium bromide-DNA fluorescence flow cytometry histograms of isolated rat liver nuclei. Nuclei were stained and flow cytometry was performed as described in the text. In each histogram, axes are linear and data of at least 5000 nuclei are presented. A. Isolated nuclei resuspended in TKM/ 40% (v/v) glycerol. (TKM is 50 mM tris-HC1,25 mM KC1 and 5 mM MgC1v pH 7.50.) Equivalent samples in TKM/0.25 M sucrose resulted in histograms identical to this one. B. Isolated nuclei resuspended in TKM/0.25 M sucrose, stored in liquid nitrogen for 24 months and thawed as described in the text. C. Isolated nuclei resuspended in a medium of very low ionic strength (1 mM MgCI2,25% (v/v) glycerol and 50 mM tris-HCl, pH 7.90). The left-sided tail of the peaks indicates partial fragmentation. D. Isolated nuclei resuspended in a strongly hyperosmotic medium (280 mM KC1, 20 mM MgCIv 0.25 M sucrose or 37% (v/v) glycerol, and 50 mM tris-HC1,pH 7.50). The nuclei are clearly not intact.
enoyl-CoA isomerase/2-enoyl-CoA hydratase/3-hydroxya c y l - C o A dehydrogenase). The p B R 322-derived p l a s m i d plivS 6 was used as a reference probe, plivS 6 carries an insert o f 550 b p at the Pst I site [29] which is complementary to the m R N A o f the protein transferrin expressed in rat liver [30]. A b o u t 250 ~tg plasmid D N A was digested with Pst I as p r e s c r i b e d b y the manufacturer. A g a r o s e gel e l e c t r o phoresis [25] o f double-stranded D N A showed that digestion was complete at the restriction site. The digested plasmid was
bound to membranes as described by Greenberg [11], using a slot blot apparatus at a negative pressure o f 0.066 arm. 10 ~tg D N A per slot was applied. This amount o f D N A was seen to ensure saturation o f the binding capacity o f the m e m b r a n e slots. Edges o f each slot were m a r k e d with water-proof pencils and membranes were air-dried, fixed for 2 - 3 h at 7 0 - 8 0 ~ in an evacuated desiccator and stored in the desiccator until further use. Prior to hybridization, each slot was cut out and
153 trimmed close to the edges with ethanol-rinsed and air-dried scissors and forceps.
Results Morphological characteristics of isolated and resuspended nuclei Nuclear pellets resuspended in the 'standard' TKM media resulted in intact nuclei, as judged by light microscopy, flow cytometry and transmission electron microscopy (TEM). Light microscopy showed that drops of nuclear suspensions were composed mainly of single, intact nuclei with visible nucleoli, but also small aggregates, normally of less than 10 nuclei (data not shown). Flow cytometrical histograms of DNA fluorescence (Fig. 1a) displayed narrow peaks of diploid and tetraploid nuclei, the coefficient of variance of the peaks being < 8%. Less than 5% of the counts had a fluorescence higher than 4 times the diploid nucleus (>8N), confirming a limited degree of aggregation. Freezing and thawing of nuclei in TKM media (with either 0.25 M sucrose or 40% (v/ v) glycerol) did not affect their morphology, DNA fluorescence histograms showed sharp peaks (Fig. lb), and both light microscopy and TEM showed intact nuclei (Fig. 2). In particular, there seemed to be no difference between frozen/ thawed sucrose- and glycerol-stored nuclei, as judged by TEM (Fig. 2a-b). Smearing, staining with haematoxyline and subsequent light microscopy was used as a marker of mechanical robustness. In general, this technique induces some damage to nuclei, observed as DNA leakage ('threads'). The results confirmed that frozen/thawed nuclei were more fragile than those freshly prepared, and subsequent incubations at 30~ for 15 min. (as in the nuclear run-off experiments) increased the fragility.
Effect of ionic strength on nuclear morphology Pellets of isolated nuclei were resuspended in media of various ionic strength (ranging from 1 mM MgCI 2 to 280 mM KCI + 20 mM MgC12) and with 0.25 M sucrose or 25-40% (v/v) glycerol. The same morphological characteristics as of nuclei in the TKM media were observed (by light microscopy and flow cytometry) with nuclei resuspended in sucrose or glycerol media of intermediate ionic strength (60-140 mM
Fig. 2. Nuclear morphologyin frozenand thawed samples of isolated rat livernuclei,as seenby transmissionelectronmicroscopy.A. Isolatednuclei resuspended in TKM/0.25M sucrose,frozenin liquidnitrogenand thawed as describedin the text. B. Isolatednuclei resuspendedin TKM/40%(v/ v) glycerol,frozenin liquid nitrogenand thawed as describedin the text. C. Hepatocytein primaryculture, for reference.
154 Table 1. Effect of storage medium on the total incorporation of lc~-32p]-UTPinto total RNA of isolated nuclei and the ratios of hybridization signals of [32P]-UTPsynthesized in vitro by isolated liver nuclei from clofibrate-treatedand control rats. Type of storage medium
Total incorporation/numberof nucleia (epm)
pMJ 115 (Clf.:Ctr.)
pMJ 26 (Clf.:Ctr.)
plivS 6 (Clf.:Ctr.)
Isoosmotic sucrose Hyperosmotie glycerol
0.58 • 0.11 1.09 • 0.28~
3.8 • 0.3b 6.1d
3.8 • 0.9b 4.3d
0.9 • 0.1 0.54a
Nuclear run-offs were performed as described in the text? Figures on total incorporation are aritmethie means of 6 triplicate samples (3 control and 3 clofibrate samples), in which isoosmotie sucrose samples and hyperosmotic glycerol samples pairwise contained the same amount of nuclei from the same animals. There was no significant differencebetween the level of total incorporation in control and clofibrate samples stored in identical media. bSignificantly higher than 1 (p
