Improved preservation of chromatin structure in ethanol-fixed cells.

The chromatin structure of differently expressed genes is generally probed by nuclease digestion of isolated nuclei (1). It is then vital that the chromatin adequately represents the transcriptional and structural status of the intact living cell. We have encountered difficulties in reproducibly detecting nuclease hypersensitivity in the chromatin of the promoter region of the actively transcribing Balbiani ring (BR) 1 gene in the dipteran Chironomus tentans (2), using conventional procedures for obtaining cell nuclei (1). Isolation of an adequate number of salivary glands required some time, why chromatin rearrangements during collection of glands and/or nuclei isolation could cause the variability observed. We therefore decided to collect the salivary glands in ice cold 70% ethanol immediately after dissection, to rapidly fix the cells and the cromatin structure and subsequently open the cells by homogenisation or detergent treatment. In a typical experiment, 200-400 salivary glands were collected in 1.5 ml of ice-cold 70% ethanol. Ethanol was discarded and the salivary glands were rinsed with nuclei isolation buffer (15 mM Tris-HCl (pH7.4), 60 mM KC1, 15 mM NaCl, 0.15 mM spermine, 0.5 mM spermidine, 0.1 mM phenylmethylsulfonyl fluoride). Nonidet P-40 was added to 1 % and the salivary glands were homogenized in a total volume of 2—3 ml in a Dounce homogenizer (10 strokes), pelleted by centrifugation and resuspended in 100 n\ nuclei buffer containing 3 mM MgCl2 (for DNase I digestions) or 1 mM CaCl2 (for micrococcal nuclease). Alternatively rinsed salivary glands were resuspended in 100 /il nuclei buffer containing 1 % NP-40 and 3 mM MgCl2 or 1 mM CaCl2. DNase I or micrococcal nuclease was added and digestions were performed at 20°C for 3 minutes and at 37°C for 5 minutes respectively. The reactions were stopped with EDTA and SDS brought to a concentration of 10 mM and 0.5%, respectively. DNA was then isolated as described previously (1). In Figure 1 A, it is demonstrated that micrococcal nuclease digestion of the chromatin in general is efficient in gland cells fixed in 70% ethanol prior to permeabilisation with 0.5 1 % NP-40 as compared to chromatin in non-fixed gland cells permeabilised with NP-40 (3,4) and in nuclei isolated by the conventional method (1). Higher concentrations (500—1000 U/ml, 37°C) of enzyme were needed in permeabilised cells. The same result was obtained for DNase I digestion (data not shown). Enzyme digestions were also efficient in cultured C. tentans cells fixed in 70% (data not shown). For a more detailed analysis, salivary glands were fixed in 70% ethanol and opened by homogenisation in a Dounce homogeniser. For comparison, salivary gland nuclei as well as nuclei from a tissue culture cell line in which the BR1 gene is inactive were isolated by the conventional method (1) and analysed in parallel. Perfect micrococcal nuclease (5) and DNase I (6) digestion patterns were observed in all three instances in the -246 to 5 4 region of the BR1 gene upstream region (Fig. IB), which has a nucleosomal organisation also in the active gene (Belikov et al. manuscript in preparation). The same filter was hybridised with a transcription start site probe, position 5 3 to +102. With ethanol fixed cells, a quite irregular picture was then observed in the case of micrococcal nuclease and an 'almost nothing' picture in the case of DNase I (Fig. 1C), consistent with nucleosomal disruption in this region. In contrast, chromatin isolated by the conventional method exhibited a disorganised but still obvious digestion pattern for both enzymes. Our results show that both methods produce nucleosomal digestion patterns both in bulk chromatin and in a specific promoter region. In a juxtaposed region, covering the TATAbox and the transcription start region, local nucleosomal disruption specific to the actively transcribing gene was suggested only after ethanol fixation. The latter is expected from results on other genes (7) and from earlier electron microscope data, showing a nucleosome-free region in the promoter region of BR genes (8). As we in addition, can obtain reproducible and strong DNase I hypersensitive sites correlated with transcriptional activity only after ethanol fixation (Belikov et al., in preparation), we believe this method to better preserve the native chromatin structure. We interpret these different results in terms of chromatin transitions occurring during salivary gland dissection and/or nuclei isolation employing the conventional method. These transitions lead to the partial restoration of the nucleosomal structure within the transcription start site region. A possible explanation for such transitions could be that the salivary gland cells continue to be biologically active after dissection. Removal from the normal environment could then result in cessation of transcription, possibly as a result of the antistress system switch (hsp genes and probably others). Fixation in ethanol appears to be a general method to 'freeze' the chromatin structure that can be used when transcriptional shut down during collection of cells, nuclei isolation or cell