Nuclease-Hypersensitive Regions in Chromosomes (Active Genes/Nuclease-Hypersensitive!Cleavage) M

Proc. Nati. Acad. Sci. USA Vol. 82, pp. 854-858, February 1985 Genetics

Nick-translation of metaphase chromosomes: In vitro labeling of nuclease-hypersensitive regions in chromosomes (active genes/nuclease-hypersensitive!cleavage) M. TIEN Kuo* AND WILLIAM PLUNKETTt *Department of Pathology and tChemotherapy Research, The University of Texas, Houston, M. D. Anderson Hospital and Tumor Institute, Houston, TX 77030 Communicated by J. Herbert Taylor, October 4, 1984

ABSTRACT Chinese hamster metaphase chromosomes gions is also preferentially sensitive to single-stranded spe- were labeled by nick-translation, which involved pretreatment cific nuclease S1 (19, 20). However, whether there is single- of metaphase chromosomes with low levels of DNase I followed stranded DNA in these regions has not been settled. by incubation with DNA polymerase I and radioactively la- The present study was initiated to characterize the nature beled nucleotides. The labeled DNA was located on nuclease- of DNase-hypersensitive regions. We employed nick-trans- hypersensitive regions of the chromosomes, as suggested by lation to label metaphase chromosomes. Our data strongly the following observations. (i) The labeled DNA was hypersen- suggest that the nick-translation preferentially labels the sitive to the subsequent DNase I digestion. (ii) The labeled DNase-hypersensitive regions on chromosomes. Biochemi- DNA contained no nucleosomes. DNA reassociation kinetic cal characterizations of the hypersensitive regions are also analysis suggested that the labeled DNA was enriched in repet- presented. itive DNA sequences. Base composition analyses showed that the labeled DNA was highly enriched in guanine and adenine residues, suggesting that the nick-translation reaction was MATERIALS AND METHODS asymmetrical and the strand enriched in purine was preferen- Cell Culture, Cell Synchronization, and Preparation of tially translated. Autoradiographic analysis revealed that the Chromosomes. Chinese hamster ovary and male Chinese label was distributed on every chromosome, but there was a hamster fibroblast cells (Don line) were used for this study. lower grain density on the Y chromosome, which is hetero- Cells were grown in Dulbecco's modified Eagle's medium chromatic and exhibits a relatively low level of gene activity. supplied with 10% fetal calf serum. The locations of silver grains on the Y chromosomes were gen- Synchronization of cells was achieved by thymidine (7 erally consistent with that revealed by the in situ hybridization mM, 16 hr) block, released for 5 hr, followed by mitotic ar- using [3H]cDNA synthesized from the total Chinese hamster rest by Colcemid (0.04 gg/ml, 2 hr). The mitotic cells were messenger RNA. These observations suggest that a specific then gently shaken off. The harvested cells contained 98- subset of genomic DNA on active chromatin is the preferred 99% mitotic cells as monitored by phase-contrast microsco- site of the nick-translation. PY. Preparation of chromosomes were essentially the method It has been well established that active chromatin or poten- described previously (21) using chromosome solution de- tially active chromatin is preferentially sensitive to the diges- scribed by Blumenthal et al. (22). tion by nucleases (see reviews in refs. 1-3). In addition to Nick-Translation of Isolated Chromosomes. The isolated this general nuclease sensitivity in the transcribed regions, chromosomes were washed in DNase I-digesting buffer con- there are chromosome segments that are =10 times more taining 10 mM Tris HCl (pH 7.5), 10 mM NaCl, and 3 mM readily digested by the nucleases than the transcribed chro- MgCl2 and then were resuspended in the same buffer at A260 matin domain (4-6). These DNase-hypersensitive regions - ==0.5. The chromosomes were pretreated with DNase I have been determined almost exclusively by the indirect (0.2-1.0 ,ug/ml) for 10 min at 37°C. The pretreated chromo- end-labeling method (6), which employs a short, radioactive- somes were washed and resuspended in a nick-translation ly labeled probe to hybridize with the restricted DNA frag- buffer that contained 10 mM Tris HCl (pH 7.5), 5 mM ment isolated from low-level nuclease-digested nuclei. The MgCl2, 5 mM 2-mercaptoethanol, 20 ,uCi (1 Ci = 37 GBq) of nuclease-hypersensitive cleavage displays a subband on the [32P]dCTP (50 Ci/mmol, Amersham), 0.3 mg each of unla- autoradiograph with one end defined by the restriction en- beled dATP, dGTP, and dTTP per ml, and 70 units ofEsche- zyme site and the other by the DNase-hypersensitive site. richia coli DNA polymerase I. Three different DNA pQly- By using this method, DNase I-hypersensitive sites have merases (from Boehringer Mannheim) were used: DNA been assigned to certain regulatory sequences such as long polymerase I (Kornberg polymerase, nick-translation terminal repeats of transcribed endogenous provirus (7) and grade), endonUclease-free DNA polyrnerase I, and large the 72-base-pair enhancer sequences of simian virus 40 (8- fragment of DNA polymerase I (Klenow enzyme). The reac- 10). A number of hypersensitive cleavage sites have been tion mixture (0.5 ml) was incubated at 14°C for 2-20 min, and assigned to the 5' side of the genes (11-13), some, but less the reaction was stopped by addition of EDTA (0.2 M) to a frequently (14, 15), at intervening gene sequences. final concentration of 10 mM. The nature of DNase-hypersensitive regions has not been Nick-Translation of Chromosomes in Mitotic Cells. The well characterized. It has been suggested, from restriction harvested mitotic cells were permeabilized with L-a-lyso- enzyme digestion (16, 17), electron microscopic observa- phosphatidylcholine (type 1, Sigma) according to the method tions (10), and "histone image" analysis (18) that these chro- described by Miller et al. (23). Cells were treated with hypo- matin stretches contain no histone. The DNA in these re- tonic solution (1:5 dilution of the regular medium with dis- tilled H20) at 22°C for 15 min, spun down, and resuspended The publication costs of this article were defrayed in part by page charge in a buffer containing 10 mM Tris HCl (pH 7.5), 10 mM payment. This article must therefore be hereby marked "advertisement" NaCl, 3 mM MgCl2, 0.5% Nonidet P-40 (NP-40), and 0.5% in accordance with 18 U.S.C. §1734 solely to indicate this fact. Triton X-100 at 4°C. After 30 min, the cells were washed two 854 Downloaded by guest on September 28, 2021 Genetics: Kuo and Plunkett Proc. Natl. Acad. Sci. USA 82 (1985) 855

times with the same buffer lacking NP-40 and Triton X-100 coccal nuclease digestion was somewhat different from that and were resuspended in the nick-translation buffer as de- of the DNase I digestion, it was also evident that the labeled scribed above containing 100 ,Ci of [3H]TTP (117 Ci/mmol, DNA was also highly preferentially sensitive to the micro- New England Nuclear) instead of [32P]dCTP. The reaction coccal nuclease digestion (not shown). Therefore, we con- was carried out at 14'C for 1 hr and stopped by the addition clude that the nick-translated DNA in chromosomes is hy- of EDTA (0.2 M) to a final concentration of 10 mM. Cells persensitive to the nuclease digestion. were washed twice with the nick-translation buffer, fixed We next investigated whether the nick-translated DNA with methanol/acetic acid, 3:1 (vol/vol), and used for air- contained a typical nucleosome structure. The nick-translat- dried preparation of metaphase chromosomes. ed chromosomes were digested with micrococcal nuclease, Metaphase chromosomes on glass slides were rinsed ex- which preferentially cuts the linker region of a nucleosome tensively with 0.3 M sodium chloride/0.03 M sodium citrate, array. DNA was extracted from the digested chromosomes, pH 7, followed by washes with 75% ethanol and 95% etha- separated by agarose gel electrophoresis, stained with ethidi- nol. Autoradiographs were prepared with Kodak thin film um bromide, visualized under UV light, and followed by AR-10 and developed in D-19B. The film was usually ex- autoradiography. A comparison between the autoradiogram posed for 1-2 weeks. The chromosomes were stained with (Fig. 1B) and the ethidium bromide-stained DNA pattern of Giemsa (10% prepared in 0.01 M sodium phosphate at pH the total genomic DNA (Fig. 1A) clearly shows that the nick- 6.8). translated DNA was much more sensitive to nuclease diges- Other Procedures. For DNA base composition determina- tion. Furthermore, no typical nucleosomal DNA repeat was tion, four labeled deoxyribonucleotides were used in the shown on the autoradiogram, whereas the ethidium bromide- nick-translation reaction. DNA (20-50 ,ug) was degraded stained pattern of total genomic DNA showed typical nu- with P1 nuclease and alkaline phosphatase (24). The result- cleosome repeats. These results suggest that the nick-trans- ing nucleosides were separated by reversed phase HPLC lated DNA does not contain a typical nucleosomal structure. (25). The retention times were 9.3 min for dCyd, 23 min for To determine whether the nick-translated DNA is a unique dThd, 25 min for dGuo, and 34.4 min for dAdo. Radioactiv- subset or a random representative of the total genomic ity associated with the eluents of each nucleoside was deter- DNA, we performed a DNA reassociation experiment. DNA mined by liquid scintillation spectroscopy. isolated from nick-translated chromosomes was mixed with Methods for digestion of nuclei with micrococcal nucle- unlabeled genomic DNA, sheared by sonication, denatured, ase, extraction of DNA, agarose gel electrophoresis of nu- and allowed to reassociate. The reassociated DNA was sepa- cleosomal DNA, DNA reassociation (26), and in situ hybrid- rated from the unreassociated DNA by hydroxyapatite col- ization of [3H]cDNA synthesized from total poly(A)+ RNA umn chromatography. Fig. 2 shows that the nick-translated (27) have been described.

1 2 3 4 1 2 3 4 RESULTS t _ When the isolated chromosomes were incubated under the nick-translation reaction conditions as described above, incorporation of radioactively labeled deoxyribonucleotides into acid-insoluble fractions (chromosomes) was found. Without DNA polymerase, very little incorporation was seen, suggesting that the nick-translation reaction was polymerase-dependent. The rate of incorporation was very rapid in the initial 2 min and then gradually leveled off. It appears that no further incorporation of the radioactive material was evident after 10 min of incubation (not shown). The level of incorporation of radioactive nucleotide into chromosomes was a function ofconcentrations of DNase I in the pretreatment reaction, suggesting that the nick-translation reaction was also DNase I-dependent. In the absence of DNase I pretreatment, no incorporation of radionucleotides was observed when large fragments of DNA polymerase I (Klenow enzyme) or endonuclease-free DNA polymerase I was used. A significant amount of incorporation was observed, however, when the nick-translation grade of DNA polymerase I was used. These results suggest that "endogenous" nicks were not significant in our nick-translation reaction. We chose 0.2 ,g of DNase I per ml for pretreatment and an incubation time of 10 min in our subsequent nick- translation reaction, using nick-translation-grade DNA polymerase I unless otherwise indicated. We next investigated the nuclease sensitivity of the nick- translated DNA in the chromosomes. The labeled chromosomes were mildly digested with either DNase I or micrococcal nuclease and the extent of digestion was measured. It B was shown that the labeled DNA was very sensitive to the digestion by both enzymes. Under conditions (1.0 ,ug of DN- ase I per FIG. 1. Agarose gel electrophoresis of DNA isolated from nick- ml, digestion for 10 min) in which about 80% of the translated chromosomes digested by micrococcal nuclease. (A) labeled DNA was rendered acid-soluble, only about 5% of Ethidium bromide staining pattern. (B) Autoradiogram. Lanes: 1, the total genomic DNA was in the acid-soluble fraction (not undigested control; 2-4, micrococcal nuclease-digested samples shown). Although the response of labeled DNA to the micro- with increasing concentrations of enzyme. Downloaded by guest on September 28, 2021 856 Genetics: Kuo and Plunkett Proc. NatL Acad Sci. USA 82 (1985) ou portion of the hamster Y chromosome is heterochromatic and supports very low levels of transcription activity, this D result suggests that the nick-translation-labeled DNA is pref- c 60- erentially in the active chromatin 0 40 regions. ._ To address this issue more precisely, we compared the lo- Mu0) cations of silver grains in the nick-translated chromosomes 40- Cu with those in the chromosomes that were hybridized in situ x with [3H]cDNA synthesized from total mRNA (Fig. 4A). We 0 specifically focused our analysis on the Y chromosome, V- 20 i_ since the Y chromosome is readily identifiable and contains 0-0 a lower grain density than other chromosomes. Most (65%) of the silver grains were located at the telomere of the short 1 10 10 100 1 ,0 3000 arm, and about 14% were located at the proximal region to- 10 100 1000 10,000 30,000 ward the centromere in the long arm (Fig. 4B). The distribu- Equivalent Cot tion of the nick-translated DNA was similar to that of the mRNA-specifying sequences identified by the in situ hybrid- FIG. 2. Reassociation of nick-translated DNA (0) and total geno- ization (Fig. 4A). This result is consistent with the idea that mic DNA (e). the nick-translated DNA in chromosomes is preferentially DNA reassociated at a faster rate than the bulk DNA. At Cot located on the active chromatin. 10, the point at which repetitive DNA sequences of about 300 copies per genome should have reassociated, only about DISCUSSION 20% of the total genomic DNA was in the double-stranded The nick-translation procedure has been commonly used for structure, whereas about 30% of the labeled DNA was in the in vitro labeling of DNA for various purposes (28). Levitt et double-stranded form. These results suggest that the nick- al. (29) first applied this method to investigate the nick-trans- translated DNA on chromosomes was enriched in repetitive lated DNA in chicken oviduct chromatin and observed that DNA sequences. the active chromatin was preferentially labeled by nick- We further investigated the base composition of DNA iso- translation. Kerem et al. (30) applied this method to label lated from nick-translated chromosomes by HPLC. As active regions in metaphase chromosomes on fixed slides. shown in Table 1, the nick-translated DNA was enriched in The procedure they used to fix chromosomes has been purine (guanine and adenine) and depleted in pyrimidine (cy- known to cause drastic changes in chromatin structure (31). tosine and thymine) residues. The depletion of pyrimidines In the present study, we utilized this method to preferential- in the nick-translated DNA was consistently observed in the ly label DNA on metaphase chromosomes that were isolated samples with different incubation times. In contrast, nick- under conditions that have been best known for preserving translation of purified total cellular DNA did not show such their authentic structure (31). Our results, which showed drastic enrichment in purine (or depletion in pyrimidine) (Ta- that the nick-translated DNA in chromosomes was hyper- ble 1). These observations suggest that the nick-translation sensitive to the nuclease digestion and contained no typical reaction in metaphase chromosomes was asymmetrical and nucleosomal structure, are consistent with the notion that the strand that was enriched in purine residues was preferen- the nuclease-hypersensitive region is preferentially labeled. tially translated. It is unlikely that the nick-translation reaction itself would We also used autoradiographic analysis to determine the have ruined the nucleosome structure, since Levitt et al. (27) locations of nick-translated DNA on chromosomes by using have observed nucleosome structure in the nick-translation permeabilized male Chinese hamster (Don) cells. The label reaction in which higher concentrations of DNase I were was distributed in virtually every chromosome, with specifi- used. cally a lower silver grain density (number of silver grains per The observation that the nick-translated DNA was en- unit length of chromosome) on the Y chromosome (Fig. 3B). riched in repetitive DNA sequences is consistent with the Analysis of 35 metaphase spreads indicated that the grain notion that conserved DNA sequences are present in the density in the Y chromosome was lower by a factor of -3 DNase-hypersensitive region. This suggestion is consistent than that of chromosome 1 (data not shown). Since a major with the recent results published in the literature: DNase I- hypersensitive regions at the 5' side upstream of genes may Table 1. DNA base composition of nick-translated regions of be responsible for transcriptional initiation machinery and chromatin and total cellular DNA some consensus DNA sequences have been reported to be in Time of this region (32). Enhancer sequences that are also hypersen- reaction, % total sitive to DNase I digestion in chromatin (8, 9, 14, 15) also min dCyd dThd dGuo dAdo contain a common core sequence (33). That the nick-translated DNA in metaphase chromosomes Chromatin was enriched in guanine and adenine residues is consistent Exp. 1 10 10 10 36 44 with the notion that the nick-translation reaction is asymmet- 20 12 11 31 46 rical 30 12 12 29 47 and the purine-rich strand in a DNA duplex is preferen- Exp. 2 11 tially labeled. Although DNase I is not a single-strand-specif- 10 9 44 36 ic nuclease, it produces single-stranded nicks. By contrast, 20 12 12 29 44 micrococcal nuclease usually cuts DNA across the double 30 10 11 29 50 strand (3). It has been reported that the DNase I-hypersensi- Total cellular DNA tive sites in the chromatin are also recognizable by S1 nucle- Exp. 10 21 23 ase (34). When the DNA sequences containing the hypersen- 1 25 28 sitive sites were cloned into a bacterial plasmid, it was re- 20 22 22 26 27 ported that these sites were also recognized by single- 30 22 22 27 26 stranded Exp. 2 10 26 21 26 25 nuclease, as long as the superhelical structure was maintained in the cloned DNA (34, 35). It is important to 30 23 21 28 26 note that nicking supercoiled DNA by the single-strand-spe- Downloaded by guest on September 28, 2021 Genetics: Kuo and Plunkett Proc. Natl. Acad. Sci. USA 82 (1985) 857

e 4A

0 *Bo 0 ..

A ; : * 7 0n

;Ml::,*s 0 ' I .. ik 0'

al0 0~~ _ .

._ V PM at0 0 At ft I a* A

CAo c. '

aro ., '% .w

a Ye.*.W^, . n

B #ta.. FIG. 3. Autoradiographs of metaphase plates of male Chinese hamster cells. (A) In situ hybridization using [3H]cDNA synthesized from total poly(A)+ RNA. (B) Nick-translation of metaphase chromosomes.

0000000000 00000 0

000000

0 A~~~~~~~~~~~~0

0000 t~r1 0 00 B 00

FIG. 4. Distribution of silver grains on Chinese hamster Y chromosomes. (A) In situ hybridization. (B) Nick-translation. Downloaded by guest on September 28, 2021 858 Genetics: Kuo and Plunkett Proc. NatL Acad Sci. USA 82 (1985) cific nuclease is asymmetrical and only the strand enriched 11. Keene, M. A. & Elgin, S. C. R. (1981) Cell 27, 57-64. in purine is preferentially nicked (20, 35, 36). The observa- 12. Stalder, J., Larsen, A., Engel, J. D., Dolan, M., Groudine, M. tion presented in the present report is consistent with the & Weintraub, H. (1980) Cell 20, 451-460. results published in these reports. 13. Kaye, J. S., Bellard, M., Dretzen, G., Bellard, F. & Chambon, Our cytogenetic analyses described in this P. (1984) EMBO J. 3, 1137-1144. report indicate 14. Chung, S. Y., Folsom, Y. & Wooley, J. (1983) Proc. Nat!. that the location of the nick-translated DNA on the Y chro- Acad. Sci. USA 80, 2427-2431. mosome is consistent with that revealed by in situ hybridiza- 15. Mills, F. C., Fisher, M. L., Kuroda, R., Ford, A. M. & Gould, tion using labeled DNA complementary to total mRNA. This H. J. (1983) Nature (London) 306, 809-812. result is consistent with the idea that the nick-translated 16. McGhee, J. D., Wood, W. I., Dolan, M., Engel, J. D. & Fel- DNA (or the DNase I-hypersensitive region) in the genome senfeld, G. (1981) Cell 27, 45-55. is located in the active chromatin regions (27, 28, 37). The 17. Sweet, R. W., Chao, M. V. & Axel, R. (1982) Cell 31, 347-353. present report also provides information concerning the pos- 18. Karpov, V. L., Preobrazhenskaya, 0. V. & Mirzabekov, sible locations of active genes on the Y chromosomes. Al- A. D. (1984) Cell 36, 423-431. 19. Kohwi-Shigematsu, T., Gelinas, R. & Weintraub, H. (1983) though there has been some speculation concerning the exis- Proc. Natl. Acad. Sci. USA 80, 4389-4393. tence of male-specific genes, no functional genes have been 20. Schou, E., Evans, T., Welsh, J. & Efstratiadis, A. (1983) Cell conclusively identified and assigned on Y chromosomes in 35, 837-848. mammals, including man (38). Our in situ and nick-transla- 21. Kuo, M. T. (1982) J. Cell Biol. 93, 278-284. tion data, which show that the telomere of the short arm and 22. Blumenthal, A. B., Dieden, J. D., Kapp, L. N. & Sedat, J. W. the regions proximal to the centromere in the long arm of the (1979) J. Cell Biol. 81, 255-258. Y chromosome contain coding sequences, may provide use- 23. Miller, M. R., Jr., Castellot, J. J. & Pardee, A. B. (1978) Bio- ful information for further investigation of Y chromosome- chemistry 17, 351-355. specific functional genes in Chinese hamster cells. With the 24. Gehrke, C. W., Kuo, K. C., McCune, R. A. & Gerhardt, K. 0. (1982) J. Chromatogr. 230, 297-308. newly developed techniques in chromosome microdissection 25. Avramis, V. T. & Plunkett, W. (1983) Biochem. Biophys. Res. and microcloning, the experimental approach has become Commun. 113, 35-43. feasible (39). 26. Kuo, M. T., Iyer, B. & Schwartz, R. J. (1982) Nucleic Acids Res. 10, 4565-4578. This research was supported in part by The Robert A. Welch 27. Kuo, M. T. & Saunders, G. F. (1977) Chromosoma 63, 241- Foundation (G831) and the National Institute of General Medical 252. Sciences (GM28573). 28. Rigby, P. J. W., Dieckmann, M. J., Rhodes, C. & Berg, P. (1977) J. Mol. Biol. 113, 237-252. 29. Levitt, A., Axel, R. & Cedar, H. (1979) Dev. Biol. 69,496-505. 1. Mathis, D., Oudet, P. & Chambon, P. (1980) Prog. Nucleic 30. Kerem, B. S., Goitein, R., Richler, C., Marcus, M. & Cedar, Acids Res. Mol. Biol. 24, 1-55. H. (1983) Nature (London) 304, 88-90. 2. Weisbrod, S. (1982) Nature (London) 297, 289-295. 31. Kuo, M. T. (1982) Exp. Cell Res. 138, 221-229. 3. McGhee, J. D. & Felsenfeld, G. (1980) Annu. Rev. Biochem. 32. Breathnach, R. & Chambon, P. (1981) Annu. Rev. Biochem. 493, 1155-1156. 50, 349-383. 4. Wu, C., Bingham, P. M., Liwak, K. J., Holmgren, R. & Elgin, 33. Hen, R., Borrelli, E., Sassone-Corsi, P. & Chambon, P. (1983) S. C. R. (1979) Cell 16, 797-806. Nucleic Acids Res. 11, 8747-8760. 5. Elgin, S. C. R. (1981) Cell 47, 413-415. 34. Larsen, A. & Weintraub, H. (1982) Cell 29, 609-622. 6. Wu, C. (1980) Nature (London) 286, 854-860. 35. Nickol, J. M. & Felsenfeld, G. (1983) Cell 35, 467-477. 7. Groudine, M., Eisenman, R. & Weintraub, H. (1981) Nature 36. Finer, M. H., Fodor, E. J. B., Boedtker, H. & Doty, P. (1984) (London) 292, 311-317. Proc. Natl. Acad. Sci. USA 81, 1659-1663. 8. Scott, W. A. & Wigmore, D. J. (1978) Cell 15, 1511-1518. 37. Javaherian, K. & Fasman, G. (1984) J. Biol. Chem. 259, 3343- 9. Saragosti, S., Cereghini, S. & Yaniv, M. (1982) J. Mol. Biol. 3349. 160, 133-146. 38. Kuo, M. T. (1984) Am. J. Human Genet., in press. 10. Jongstra, J., Reudelhuber, T. L., Oudet, P., Benoist, C., 39. Rohme, D., Fox, H., Herrmann, B., Frischauf, A. M., Ed- Chae, C.-B., Jeltsch, J. M., Mathis, D. J. & Chambon, P. strom, J. E., Marius, P., Silver, L. & Lehrach, H. (1984) Cell (1984) Nature (London) 307, 708-714. 36, 783-788. Downloaded by guest on September 28, 2021