Gene, 74 (1988) 135-137 Elsevier
DNA methylation, ~~omatin structure and rotation
of Herpes simplex virus tk gene expression
(Recombinant DNA; in vitro methylation; microinjection)
Adolf Graessmann and Monika Graessmann Instill
Mole~l~rbioIo~e und gi~~ernie, Freie U~iyersit~t Berlin, D-loo0 Berlin 33 (F..R.G.J
Received 25 May 1988 Accepted 18 July 1988 Received by publisher 3 August 1988
(a) Introduction Mammalian cells have acquired a wide spectrum of mechanisms for regulating gene expression, many of which act at the level of transcription. These involve the interplay of DNA with various other macromolecules, changes in DNA or chromatin structure, and last not least DNA modification. Changes in the DNA me~ylation pattern have been studied extensively. In mammalian cells, 5-methylcytosine in the CpG dinucleotide sequence is the only methylated nucleotide found so far and, in somatic cells, the pattern of me~ylation is normally inheritable from generation to generation. For some genes, reduced methylation of the DNA is required but not sufhcient to allow expression. That DNA methylation can indeed block gene expression was clearly demonstrated by DNA transfer experiments. These points have been reviewed (e.g., Dorfler, 1983). However, there is increasing evidence that Correspondence to: Dr. A. Graessmann, Institut Rir Molekularbiologie und Biochemie, Freie Universitfit Berlin, Amimallee 22, D-1000 Berlin 33 (F.R.G.) Tel. (030)838-25~. * Presented at the New England Biolabs Workshop on Biological DNA Modification, Gloucester, MA (U.S.A.) 20-23 May 1988.
Abbreviations: HAT, selective medium containing hypoxanthine, aminopterin and thymidine; HSV, Herpes simplex virus; M ., methyltransferase; pHSVl06-CH,, plasmid pHSV106 DNA that was methylated in vitro with M *HpaII; SDS, sodium dodecyl sulfate; SV40, Simian virus 40; TK, thymidine kinase; Tk, phenotype with respect to TK activity; tk, gene coding for TK.
inhibition of transcription by methylation does not apply to ah e~~otic genes. We have shown that SV40 and polyoma virus DNAs, extensively methylated in vitro, were as active after microinjection as the nonmethylated DNA (Graessmann et al., 1983). Tanaka et al. (1983) described a third class of genes which become hypermethylated before transcription. We now report that methylation of the HSV tk is not sufficient to prevent tk gene expression after microinjection into Tk - culture cells. We have also found that chromatin formation is an essential step in converting the HSV tk gene from a methylation-insensitive to a me~ylation-sensitive state. (b) DNA metbylatioo per se dues not block HSV tk gene expression To test whether DNA methylation per se can block gene expression, pHSV-106 DNA was methylated in vitro with M +hlpaI1 (Busc~ausen et al., 1985). The extent of methylation was determined for each preparation by DNA blot analysis of &a11 and MspI-cieaved DNA. The biological activity of this methylated DNA was tested by microinjection into the nuclei of Tk- rat-2 and mouse L cells. Immediately after microinjection, [ 3H]thymidine (OS [email protected]
) was added to the culture medium for 24 h. After this cultivation period, the cells were tixed and processed for autoradiography to detect TK activity. These experiments revealed that the methylated DNA (pHSV-106-CH3) was as active as the mock-methylated DNA (Table I). To exclude the possibility that TK activity was
0378-l119/88/%03.50 Q 1988Elsevier Saence Publishers B.V. (Biomedical Division)
136 TABLE I Thymidine kinase activity in microinjected rat-2 cells Material injected*
pHSV-106 (200-400 molecules/cell) pHSV-106~CH, (200-400 molecules/cell) pHSV-106 (two to four molecules/cell) pHSV-106~CH, (two to four molecules/cell) pHSV-106 chromatin pHSV-106~CH, chromatin pHSV-106 chromatin/pHSV-106~CH, chromatind pHSV-106-CH, reextracted from pHSV-106-CH, chromatin’
[3H]Thymidine incorporation at different times after injection b O-24 h
120-140 120-140 80-120 SO-120 120-140 0 120-140 120-140
200 160-180 n.t. n.t. 200 0 200 160-180
500 0.5 n.t. nt. 500 0 500 0.5
a Tk - rat-2 cells, grown on glass slides, were used for microinjection. Details of our microinjection techniques are described elsewhere (Graessmann and Graessmann, 1983), and in vitro pHSV-106 chromatin reconstitution in our publication (Buschhausen et al., 1987). b After microinjection, cells were labeled with [‘Hlthymidine (0.5 &i/ml) at 24-h intervals at the times indicated and processed for autoradiography. The number of injected cells was counted as 100%. Higher numbers of positive cells represent cell division. Data are average values from five independent experiments with 50 injected cells each and are expressed as y0 Tk’ cells. nt., not tested. c The number of positive cells (48-72 h after injection) was estimated and not counted. d pHSV-106~CH, chromatin (0.01 mg/ml) was mixed before microinjection with mock-methylated chromatin (0.01 mg/ml) in a 1: 1 ratio (v/v). e pHSV-106-CH, DNA was reextracted from the reconstituted chromatin by SDS-phenol
caused by a small number of undermethylated DNA molecules not detectable by blot analysis, lower DNA copy numbers were injected per cell. These experiments indicated that after the transfer of only two to four methylated DNA molecules per cell, [ 3H]thymidine incorporation still occurred with the same efficiency as in cells injected with unmethylated pHSV-106 DNA (Table I). We can also exclude that demethylation of the DNA occurs after injection. For this purpose, 200 rat-2 cells were microinjected with pHSV-106~CH3 and the DNA was reextracted from recipient cells by the Hirt method 24 h after the transfer. The resulting extract was divided for blot analysis into three aliquots and two of them were incubated with either HpaII or&PI restriction endonuclease. The results exclude intracellular demethylation as the reason for the HSV TK activity (Buschhausen et al., 1985). (c) Inhibition of rk gene expression requires chromatin structure
In long-term experiments, it is known that methylated rk DNA is not expressed (Dbrfler, 1983). To
treatment and microinjected.
assess when, after DNA transfer, the methylation block becomes effective, [ 3H]thymidine was added to the culture medium at later time points. These experiments show that thymidine incorporation is not detectable 72 h after injection of the methylated DNA (Table I). To determine more exactly the point of transition from methylation insensitivity to sensitivity, the transcription rate of the injected DNA was analysed. To this end, 200 rat-2 cells were microinjected with two to four DNA molecules each; RNA was extracted from the recipient cells at various times after DNA transfer and subjected to RNA dot-blot analysis. These experiments show that the methylated DNA was as active as the mock-methylated DNA for about 8 h after injection. During this time, the rate of HSV tk RNA accumulation was similar for both methylated and mock-methylated DNA molecules. After this time, a sharp decrease of HSV tk mRNA synthesis occurred in cells injected with the methylated DNA, while transcription of the mock-methylated DNA continued as expected (Buschhausen et al., 1987). Next, we asked whether chromatin formation is a
crucial step for the inhibitory effect of tk gene expression. Since in vitro methylation of chromatin is a very unsatisfactory process, we first methylated the pHSV-106 DNA with the M *HpaII and chromatin was reconstituted in vitro with purified histone octamers. To test the biological activity of reconstituted chromatin, we first microinjected mockmethylated pHSV- 106 minichromosomes into the nuclei of rat-2 and Ltk- cells and analysed [ 3H]thymidine incorporation by autoradiography. These experiments show that the chromatin was as active as the pHSV-106 DNA itself. After microinjection of only two to four chromatin molecules per cell, 120-140% of cells incorporated thymidine during the 24-h cultivation period. We further tested the activity of the HSV-chromatin by RNA dot-blot analysis various times after injection and found no difference in the rate of HSV tk mRNA synthesis between DNA and chromatin-injected cells. Reextraction experiments revealed that the injected chromatin remains stable in the cells without dissociation into DNA and octamers (Buschhausen et al., 1987). However, a significant difference was observed after injection of the reconstituted methylated chromatin. In contrast to the methylated DNA, the methylated chromatin was inactive immediately after microinjection. We could not demonstrate thymidine incorporation, even after injection of 200-400 chromatin molecules per cell. Also HSV tk mRNA synthesis was not demonstrable by dot-blot analysis. Methylated chromatin can be reactivated by 5-azacytidine, and tk’ cell clones were selected in HAT medium (Szybalska and Szybalski, 1962). This reactivation process was linked to demethylation of the HSV tk DNA (Buschhausen et al., 1987). (d) Conclusions
Taken together, our experiments have shown that DNA methylation per se does not prevent expression of the HSV tk gene. Essential for inhibition of tk transcription is chromatin formation, which occurs after injection of DNA with a latency period of about 8 h. DNA and RNA extraction experiments have shown that transition from methylation insensitivity to sensitivity correlates with chromatin formation of the injected DNA. The latency period of 8 h can be
explained by the low free-histone pool of the injected cells during the interphase. The key role of the chromatin structure in the inhibition of tk gene expression by DNA methylation was directly shown by microinjection of DNA methylated in vitro and reconstituted chromatin molecules. So far, however, we do not know what element(s) of the methylated chromatin determines methylation sensitivity. There are several features which can be attributed to many active genes. These include DNase hypersensitive sites and the formation of a nucleosome free region around the promoter region. We now seek to determine whether or not the methylated and mock-methylated chromatin differ in these respects after microinjection into culture cells.
We thank E. Guhl for excellent technical assistance. The work was supported by the Deutsche Forschungsgemeinschaft and the Verband der Chemischen Industrie.
REFERENCES Buschhausen, G., Graessmann, M. and Graessmann, A.: Inhibition of herpes simplex thymidine kinase gene expression by DNA methylation is an indirect effect. Nucleic Acids Res. 13 (1985) 5503-5513. Buschhausen, G., Wittig, B., Graessmann, M. and Graessmann, A.: Chromatin structure is required to block transcription of the methylated herpes simplex virus thymidine kinase gene. Proc. Natl. Acad. Sci. USA 84 (1987) 1177-1181. Dorfler, W.: DNA methylation and gene activity. Annu. Rev. Biochem. 52 (1983) 93-124. Graessmann, M. and Graessmann, A.: Microinjection of tissue culture cells. Methods Enzymol. 101 (1983) 482-491. Graessmann, M., Graessmann, A., Wagner, H., Werner, E. and Simon, D.: Complete DNA methylation does not prevent polyoma and simian virus 40 early gene expression. Proc. Natl. Acad. Sci. USA 80 (1983) 6470-6474. Szybalska, E.H. and Szybalski, W.: Genetics of human cell lines, IV. DNA-mediated heritable transformation of a biochemical trait. Proc. Natl. Acad. Sci. USA 48 (1962) 2026-2034. Tanaka, K., Appella, E. and Jay, G.: Developmental activation of the H-2K gene is correlated with an increase in DNA methylation. Cell 35 (1983) 457-465. Edited by R.M. Blumenthal.