different types of hypersensitive sites in the mouse metallothionein

of 5/5
THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1987 by The American Society of Biological Chemists, Inc. Vol. 262, No. 5, Issue of February 15, pp. 2161-2165,1987 Printed in U. S.A. Different Types of Hypersensitive Sites in theMouse Metallothionein Gene Region* (Received for publication, April 8, 1986) Craig A. MacArthurS and Michael W. Lieberman4 From the Department of Pathology, Fox Chase Cancer Center, Philadelphia, Pennsylvania 19111 We have examined the chromatin structure of the metallothionein (MT)gene region in MT- 549 mouse lymphoma cells and in derivatives which express MT- I alone, MT-I1 alone, or both genes. In all lines, these genes are contained in a 16-kilobase pair region be- tween two DNase Isensitive sites: one site located 5.3 kilobase pairs 5' of MT-I1 (the 5' gene) is present in naked DNA and retained in the chromatin of all lines; the other site located 3.1 kilobase pairs 3' of MT-I is hypersensitive. Hypersensitivity at three other sites is dependent on the expression of MT genes. Two sites 5' of MT-I1 disappear,anda site 3'of MT-I appears regardless of which gene is activated. The fact that these sites respond when either gene is activated sug- gests that the regulation of the two genes is interde- pendent and that the region undergoes a general change in conformation with MT activation. In addi- tion, a single site in the 5' region of MT-I1 becomes hypersensitive with activation of the gene and may be related directly to expression. The mouse metallothionein (MT)' locus is an interesting example of two closely linked genes (MT-I and MT-11) which are expressed and coordinately regulated in many tissues (1, 2). The genes code for two small homologous proteins which appear to function in zinc and copper homeostasis and per- haps in adaptation to stress (3). S49 lymphoma cells do not normally express MTs, although two copies of the locus are present per cell (one/haploid genome); however, the genes may be activated by treatment with carcinogens (4, 5). Little is known about the chromatin structure of this region. One study demonstrateddiffuse DNase I hypersensitivity at a site immediately 5' of MT-I in mouse liver and kidney and in cell lines expressing MT-I (6). In an MT- line, hypersensitivity at this site was not detected. Another study demonstrated that the release of the MT-I gene from mouse liver chromatin by micrococcal nuclease was enhanced by cadmium adminis- tration (7). We have recently treated S49 cells with N-ethyl- nitrosourea and derived a series of variants which express MT-I alone, MT-I1 alone, or both MT-I and MT-II(8). These variants and MT- S49 cells provide the opportunity to study * This work wassupported in part by National Institutes of Health Grant CA39392. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solelyto indicate this fact. $ Trainee of the Medical Scientist Training Program, Washington University, St. Louis, MO63110. Supported by National Institutes of Health Grants 5T32 GM07200 and 5T32 ES07066. § To whom correspondence and requests for reprints should be addressed. ' The abbreviations used are: MT, metallothionein; EGTA, [ethyl- enebis-(oxyethylenenitrilo)]tetraacetic acid; kb, kilobase pair. the chromatin structure of the inactive genes and toexamine changes which occur when individual genes within a cluster are activated. In this paper, we report the identification of three classes of DNase I hypersensitive sites in the mouse metallothionein gene region. MATERIALS AND METHODS Cell Lines and Cell Culture"S49 mouse lymphoma cells are cad- mium-sensitive (Cd") and MT- (4, 5). They were treated with N- ethylnitrosourea to derive MT' lines (5, 8). We have used three of these lines in addition to S49 in this study: line E15 which is MT-I-/ MT-II+;line E16 which is MT-I+/MT-11-;and line E14 which is MT- I+/MT-II+. These lines show constitutive expression of the appropri- ate MT gene(& and MTgene expression is stimulated by cadmium as measured by dot blot analysis and nuclear transcription assays (8). Expressed genes are demethylated in their 5' regions and, as judged from methylation patterns, only one of the two copies of each MT gene is expressed in these lines (8). No deletions, insertions, rear- rangements, or amplifications were detected in the MT gene region in these lines (8); it is not known whether any DNA sequence changes occurred in this region during activation. Cells were grown in Ham's F-12 medium with 10% fetal calf serum (Gibco) as previously de- scribed (4, 5). Three days prior to harvest, cells were diluted 1:10 (final concentration, approximately 1 x 10' cells/ml) and grown for 24 h. For the remaining 48 h prior to harvest, the cellular DNA was labeled with [6-3H]deoxycytidine(Moravek Biochemicals) at a final concentration of 250 nCi/ml. For cadmium-treated cells, the medium was made 20 pM in CdSO, for the final 6 h prior to harvest. Purification of Nuclei-Nuclei were purified according to the method of Siebenlist et al. (9). Briefly, approximately 1-2 X 10' cells pended in 25 ml of ice-cold buffer NA (60 mM KCl, 15 mM NaC1, 5 were washed twice in ice-cold phosphate-buffered saline and resus- mM MgC12, 0.1 mM EGTA, 15 mM Tris-HC1 (pH 7.5), 0.5 M dithio- threitol, 0.1 mM phenylmethylsulfonyl fluoride, and 0.3 M sucrose). Cells (25 ml) were disrupted by mixing with 0.7 ml of ice-cold 10% Nonidet P-40 and Dounce homogenization. This cellular solution was layered over 20 ml of ice-cold buffer NB (NA containing 1.7 M sucrose). The nuclei were pelleted by centrifugation at 28,000 X g (4 "C). DNase Z Digestion of Nuclei and DNA-For DNase I digestion studies, the method of Siebenlist et al. (9) was modified as follows. The nuclear pellet or DNA was dissolved in 6 ml of buffer NC (NA with 5% glycerol) and immediately divided into 12 samples of 500 pl each. Forty pl of H20 was added to one tube, while 0-64 units of DNase I (Sigma, 2540 units/mg) in 40 p1 of 5 mM CaC12 and 1 mM MgC12 was added to 10 others. After incubation for 3-4 min at 25 "C, the reaction was stopped by adding 50 pl of 0.5 M EDTA and placing the sample on ice. A portion (60 pl) of each digestion point was precipitated with ice-cold 7% perchloric acid, and the tritium in the supernatant was measured by liquid scintillation counting. The trit- ium in a fully digested sample was determined by adding 50 pl of proteinase K (10 mg/ml, Sigma) to tube 12 and incubating at 45 "C for 2 h, followed by the addition of 50 pl of 20 units/pl DNase I and incubation at 37 "C. Samples (60 pl) were taken from the tubes at 1, 1.5, and 2 h, and were precipitated with perchloric acid as described above. The extent of digestion for a sample was determined by the following equation: 2161

Post on 14-Feb-2017

217 views

Category:

Documents

2 download

Embed Size (px)

TRANSCRIPT

  • THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1987 by The American Society of Biological Chemists, Inc.

    Vol. 262, No. 5, Issue of February 15, pp. 2161-2165,1987 Printed in U. S.A.

    Different Types of Hypersensitive Sites in the Mouse Metallothionein Gene Region*

    (Received for publication, April 8, 1986)

    Craig A. MacArthurS and Michael W. Lieberman4 From the Department of Pathology, Fox Chase Cancer Center, Philadelphia, Pennsylvania 19111

    We have examined the chromatin structure of the metallothionein (MT) gene region in MT- 549 mouse lymphoma cells and in derivatives which express MT- I alone, MT-I1 alone, or both genes. In all lines, these genes are contained in a 16-kilobase pair region be- tween two DNase I sensitive sites: one site located 5.3 kilobase pairs 5' of MT-I1 (the 5' gene) is present in naked DNA and retained in the chromatin of all lines; the other site located 3.1 kilobase pairs 3' of MT-I is hypersensitive. Hypersensitivity at three other sites is dependent on the expression of MT genes. Two sites 5' of MT-I1 disappear, and a site 3' of MT-I appears regardless of which gene is activated. The fact that these sites respond when either gene is activated sug- gests that the regulation of the two genes is interde- pendent and that the region undergoes a general change in conformation with MT activation. In addi- tion, a single site in the 5' region of MT-I1 becomes hypersensitive with activation of the gene and may be related directly to expression.

    The mouse metallothionein (MT)' locus is an interesting example of two closely linked genes (MT-I and MT-11) which are expressed and coordinately regulated in many tissues (1, 2). The genes code for two small homologous proteins which appear to function in zinc and copper homeostasis and per- haps in adaptation to stress (3). S49 lymphoma cells do not normally express MTs, although two copies of the locus are present per cell (one/haploid genome); however, the genes may be activated by treatment with carcinogens (4, 5 ) . Little is known about the chromatin structure of this region. One study demonstrated diffuse DNase I hypersensitivity at a site immediately 5' of MT-I in mouse liver and kidney and in cell lines expressing MT-I ( 6 ) . In an MT- line, hypersensitivity at this site was not detected. Another study demonstrated that the release of the MT-I gene from mouse liver chromatin by micrococcal nuclease was enhanced by cadmium adminis- tration (7). We have recently treated S49 cells with N-ethyl- nitrosourea and derived a series of variants which express MT-I alone, MT-I1 alone, or both MT-I and MT-II(8). These variants and MT- S49 cells provide the opportunity to study

    * This work was supported in part by National Institutes of Health Grant CA39392. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

    $ Trainee of the Medical Scientist Training Program, Washington University, St. Louis, MO 63110. Supported by National Institutes of Health Grants 5T32 GM07200 and 5T32 ES07066.

    To whom correspondence and requests for reprints should be addressed.

    ' The abbreviations used are: MT, metallothionein; EGTA, [ethyl- enebis-(oxyethylenenitrilo)]tetraacetic acid; kb, kilobase pair.

    the chromatin structure of the inactive genes and to examine changes which occur when individual genes within a cluster are activated. In this paper, we report the identification of three classes of DNase I hypersensitive sites in the mouse metallothionein gene region.

    MATERIALS AND METHODS

    Cell Lines and Cell Culture"S49 mouse lymphoma cells are cad- mium-sensitive (Cd") and MT- (4, 5). They were treated with N- ethylnitrosourea to derive MT' lines (5, 8). We have used three of these lines in addition to S49 in this study: line E15 which is MT-I-/ MT-II+; line E16 which is MT-I+/MT-11-; and line E14 which is MT- I+/MT-II+. These lines show constitutive expression of the appropri- ate MT gene(& and MT gene expression is stimulated by cadmium as measured by dot blot analysis and nuclear transcription assays (8). Expressed genes are demethylated in their 5' regions and, as judged from methylation patterns, only one of the two copies of each MT gene is expressed in these lines (8). No deletions, insertions, rear- rangements, or amplifications were detected in the MT gene region in these lines (8); it is not known whether any DNA sequence changes occurred in this region during activation. Cells were grown in Ham's F-12 medium with 10% fetal calf serum (Gibco) as previously de- scribed (4, 5). Three days prior to harvest, cells were diluted 1:10 (final concentration, approximately 1 x 10' cells/ml) and grown for 24 h. For the remaining 48 h prior to harvest, the cellular DNA was labeled with [6-3H]deoxycytidine (Moravek Biochemicals) at a final concentration of 250 nCi/ml. For cadmium-treated cells, the medium was made 20 p M in CdSO, for the final 6 h prior to harvest.

    Purification of Nuclei-Nuclei were purified according to the method of Siebenlist et al. (9). Briefly, approximately 1-2 X 10' cells

    pended in 25 ml of ice-cold buffer NA (60 mM KCl, 15 mM NaC1, 5 were washed twice in ice-cold phosphate-buffered saline and resus-

    mM MgC12, 0.1 mM EGTA, 15 mM Tris-HC1 (pH 7.5), 0.5 M dithio- threitol, 0.1 mM phenylmethylsulfonyl fluoride, and 0.3 M sucrose). Cells (25 ml) were disrupted by mixing with 0.7 ml of ice-cold 10% Nonidet P-40 and Dounce homogenization. This cellular solution was layered over 20 ml of ice-cold buffer NB (NA containing 1.7 M sucrose). The nuclei were pelleted by centrifugation at 28,000 X g (4 "C).

    DNase Z Digestion of Nuclei and DNA-For DNase I digestion studies, the method of Siebenlist et al. (9) was modified as follows. The nuclear pellet or DNA was dissolved in 6 ml of buffer NC (NA with 5% glycerol) and immediately divided into 12 samples of 500 pl each. Forty pl of H20 was added to one tube, while 0-64 units of DNase I (Sigma, 2540 units/mg) in 40 p1 of 5 mM CaC12 and 1 mM MgC12 was added to 10 others. After incubation for 3-4 min at 25 "C, the reaction was stopped by adding 50 pl of 0.5 M EDTA and placing the sample on ice. A portion (60 pl) of each digestion point was precipitated with ice-cold 7% perchloric acid, and the tritium in the supernatant was measured by liquid scintillation counting. The trit- ium in a fully digested sample was determined by adding 50 pl of proteinase K (10 mg/ml, Sigma) to tube 12 and incubating at 45 "C for 2 h, followed by the addition of 50 pl of 20 units/pl DNase I and incubation at 37 "C. Samples (60 pl) were taken from the tubes at 1, 1.5, and 2 h, and were precipitated with perchloric acid as described above. The extent of digestion for a sample was determined by the following equation:

    2161

  • 2162 Hypersensitive Sites in Metallothionein Chromatin where (CPM),,I, is the perchloric acid-soluble 3H cpm for the sample, (CPM)o is the perchloric acid-soluble 3H cpm for the control, and (CPM)lm is the perchloric acid-soluble 'H cpm for the fully digested sample.

    Purification of DNA and Indirect End-labeling Analysis-To the remaining 540-p1 samples from each digestion point, we added 11 pl of 20% sodium dodecyl sulfate and 60 pl of 10 mg/ml proteinase K. The samples were incubated at 45 "C for 4 h, and the DNA was purified by repeated organic extractions, ethanol precipitation, diges- tion with ribonuclease A (0.1 mg/ml, 2 h, 37 "C; Sigma), additional organic extractions, and two more ethanol precipitations. The DNA was dissolved in 10 mM Tris-HCI (pH 7.5), 1 mM EDTA and quan- titated spectrophotometrically. For indirect end-labeling analyses (lo), we used probes for the MT region (Fig. 1). DNA (10 pgllane) was digested with 5-10 units/pg DNA of XbaI (Amersham), HindIII (Amersham), or EcoRI (New England Biolabs) and electrophoresed on 1% or 1.5% agarose gels. Gels were blotted in the usual fashion,

    - M T - I I M T - I - U

    8 U W U

    g f f!

    . . . . . . . . . . . . . . . 0 1 8 9 IO 11 I2 13 14 I5 16 I7 18 19 26

    NE450 EHBPO H m

    FIG. 1. Map of mouse metallothionein gene region. The data for this map are from Searle et al. (1) and MacArthur et al. (8). Only the relevant restriction sites are included. The scale for the map is indicated by dots below the map and is numbered in kilobase pairs. The 5' EcoRI site is arbitrarily labeled as 0. Note that the scale is interrupted in two places. The probes used in Southern blots (NE450, EH820, XH1000) are indicated below the map. The exons of the MT- I and MT-I1 genes are indicated by solid boxes, and the direction of transcription of each gene is indicated by an arrow above the map.

    FIG. 2. Analysis of DNase I hy- persensitive sites in MT- S49 cells. Nuclei from cadmium-treated S49 cells were isolated and digested with DNase I, and DNA was purified (see "Materials and Methods"). DNA (10 pgllane) was digested with EcoRI (A) XbaI ( B ) or HindIII ( C and D), electrophoresed on 1% (A and B ) or 1.5% (C and D) agarose gels, and blotted to nitrocellulose. The resulting blots were probed with NE450 (A), XHlOOO ( B and C ) , or EH820 (D). The number to the left of the figures are sizes of marker fragments ( M ) in kb. The numbers to the right show the bands that result from DNase I cutting and corre- spond to the DNase I sites shown in Fig. 5. The numbers at the top of the figures are the tube numbers of the DNase I digestions (see "Materials and Meth- ods"). The percent of ['Hjdeoxycytidine rendered acid-soluble by DNase I is as follows. A and B: Tube I, 0%; Tube 3,

  • Hypersensitive Sites in Metallothionein Chromatin 2163

    A. M 1 2 3 4 5 6 7 8 91011

    19 - 9.0- 6.6-

    3.8 - 4.3 -

    2.3-

    1.3-

    0.56 - 0.77-

    0.39 -

    c. M I 2 3 4 5 6 7 891011 19. 9.0 6.6. 4.3 - 3.8 .

    2.3 .

    1.3.

    0.77.

    0.39 - 0.56-

    1

    19 - 9.0 - 6.6-

    4.3- 3.8 -

    2.3-

    1.3-

    0.77- 0.56- 0.39-

    --c

    -0 P

    19 - 9.0 - 6.6- 4.3- 3.8 -

    -8

    -7 -6

    -4

    1.3-

    0.77 - 0.56- 0.39-

    -4

    FIG. 3. Analysis of DNase I hypersensitive sites in MT-I-/MT-II+ E15 cells. Nuclei from cadmium- stimulated E15 cells (A-C) and unstimulated E15 cells (D) were digested with DNase I (see Materials and Methods). DNA (10 &lane) was digested with EcoRI ( A ) , XbaI ( B ) , or HindIII (C and D) as in Fig. 2. The resulting blots were probed with NE450 ( A ) , XHlOOO (R) , or EH820 (C and D). Marker fragments (M) and DNase I hypersensitive sites are indicated as in Fig. 2. The percent of [HH]deoxycytidine rendered acid-soluble by DNase I is as follows. A-C: Tube 1, 0%; Tubes 2 and 3,

  • 2164 Hypersensitive Sites in Metallothionein Chromatin

    19

    9.0 6.6

    3.8' 4.3

    2.3

    1.3

    0.77 0.56 0.39

    C. M i 2 3 4 5 6 7 8 9 1 0 1 1

    19 - 9.0 - 6.6-

    4.3- 3.8-

    2.3 -

    1.3-

    0.77- 0.56- 0.39-

    M I 2 3 4 5 6 7 8 9 1 0 1 1

    19 - 9.0 - 6.6 -

    "1 4.3- 3.8-

    2.3-

    1.3-

    0.77 - 0.56 - 0.39 -

    -8

    -7

    D. M i 2 3 4 5 6 7 8 9 1 0 1 1

    , . ie; ., .

  • Hypersensitive Sites in Metallothionein Chromatin 2165

    FIG. 5. DNase I hypersensitive site map of MT gene% in S49 cells and Cd' variants E14, E16, and E16. The restriction sites are the same as shown in Fig. 1, with the following

    HindIII (H), XbaI (X). The exons for the designations: EcoRI (E) , NcoI (N),

    MT genes are indicated by solid blocks. The direction of transcription for each MT gene is indicated by an arrow above the exons. The identity of the cell line and its MT status are given above each map. DNase I hypersensitive sites are indicated below the maps and are num- bered as in Figs. 2-4. DNase I sensitive sites that do not change upon activation of MT genes by carcinogens are indi- cated by a plain arrow ct). Sites that disappear when the MT genes are acti- vated by carcinogens are indicated by a broken arrow ( I ). Sites that appear when the MT genes are activated are indicated by an arrow with two cross marks ( ). The arrow in region 5 is thicker, reflecting its diffuse nature.

    M T - I I M T - I

    549 MT-1-1 P- - E N H E E H H X H X L 1 .d I J .A 1

    # t

    I h b 7 E

    E 15 MT-1-1 P+ X H X I .,I 1

    t t 1 4 6 7 8

    E E 16 MT-I+/ IT-

    N H E E H H X H X I .-I 1

    t 1 f t 5 7 8

    E 14 MT-I+/II+ X H X

    I .-I

    I 4 4 " 5 4 1 6 7 8

    H 1 kb

    DISCUSSION

    We found three types of DNase I hypersensitive sites in MT- S49 cells and MT' variants. The first type (site 7, Fig. 5) is present in all S49 cells regardless of whether the MT genes are expressed. Related to this finding is the presence of a nuclease-sensitive site found in naked DNA and in the chromatin of S49 cells and the MT' variants (site 1, see below). The second type of hypersensitive site appears or disappears in association with the activation of either of the MT genes: Two DNase I hypersensitive sites 5' of the MT-I1 gene in MT- S49 cells (sites 2 and 3, Fig. 5) disappear and a new DNase I hypersensitive site 3' of the MT-I gene appears (site 8, Fig. 5). An additional site 3' of the MT-I gene (site 6, Fig. 5) also appears when the MT-I1 gene is activated. These findings suggest that the MT locus in chromatin functions as a single unit. A third type of site occurs in the 5' region of an activated gene. In cell lines in which the MT-I1 gene is activated, a DNase I hypersensitive site (site 4 , Fig. 5) appears in the first intron of activated MT-I1 genes.

    The importance of site 1 is difficult to assess since it is present in both naked DNA and chromatin. Whether it rep- resents a chance occurrence in the DNA sequence 5' to MT- I1 or contains information important to the formation of chromatin cannot readily be determined. It is possible that sites 1 and 7 which bracket the two MT genes are involved in forming the boundaries of an MT domain or are even related to matrix association regions and/or topoisomerase I1 binding sites (12-15), but additional investigation would be needed to test this hypothesis directly.

    The activation of either MT gene results in the disappear- ance of DNase I hypersensitive sites 2 and 3 upstream of MT- I1 (Fig. 5) and the appearance of site 8 downstream of MT-I (Fig. 5). These data suggest that the activation of the MT genes is associated with alterations of chromatin structure throughout the MT gene region and that functionally this small gene cluster behaves as a single unit. Precisely how this functional unit might relate to the persistent hypersensitive sites at either end of the cluster (sites 1 and 7) remains to be resolved. However, we suggest that a functional unit can extend beyond a persistent hypersensitive site since site 8 lies

    -3.5 kb 3' of site 7 and appears when either MT gene is activated.

    Senear and Palmiter (6) were able to detect a 5' MT-I hypersensitive site in mouse kidney, but found only weak bands in cultured mouse cells in which the MT locus had been amplified by selection in cadmium. Without amplifica- tion, this band was undetectable in their experiments. Thus, our inability to demonstrate a definitive hypersensitive site 5' of MT-I is probably inherent in the structure of this region and not the result of blotting procedures.

    The significance of our findings is that we are able to distinguish among different classes of hypersensitive sites; thus, these and similar variants should allow examination of the relationship of different classes of hypersensitive sites to chromatin structure and transcription.

    ard H. Cohen for their thoughtful comments. Acknowledgments-We thank Dr. John B. E. Burch and Dr. Leon-

    I.

    2.

    3. 4.

    5.

    6.

    7.

    8.

    9.

    10. 11.

    12. 13.

    14.

    15.

    REFERENCES Searle, P. F., Davison, B. L., Stewart, G. W., Wilkie, T. M.,

    Norstedt, G., and Palmiter, R. D. (1984) Mol. Cell. Biol. 4, 1221-1230

    Yagle, M. K.., and Palmiter, R. D. (1985) Mol. Cell. Biol. 6, 291- 294

    Karin. M. 11985) Cell 41. 9-10 Lieberman; M. W., Beach, L. R., and Palmiter, R. D. (1983) Cell

    36.207-214 MacArthur, C. A., Ramabhadran, R., Godwin, A. K., Lebovitz,

    R. M., and Lieberman, M. W. (1985) Carcinogenesis (Lond.).

    Senear, A. W., and Palmiter, R. D. (1982) Cold Spring Harbor 6,887-892

    SymD. Qwlnt. Biol. 47,539-547 Koiopatnick, J., Andrew; G., and Duerksen, J. D. (1983) Nucleic

    Acids Res. 11,3255-3267 MacArthur, C. A., Hawley, T., and Lieberman, M. W. (1986)

    Siebenlist, U., Hennighausen, L., Battey, J., and Leder, P. (1984)

    Wu, C. (1980) Nature 286,854-860 Mayo, K. E., Warren, R., and Palmiter, R. D. (1982) Cell 29,99-

    Elgin, S. C. R. (1982) Nature 300,402-403 Mirkovitch, J., Mirault, M.-E., and Laemmli, U. K. (1984) Cell

    Earnshaw, W. C., and Heck, M. M. S. (1985) J. Cell Biol. 100,

    Cockerill, P. N., and Garrard, W. T. (1986) Cell 44,273-282

    Carcinogenesis ( L o n d . ) 7,1487-1495

    Cell 37,381-391

    108

    39,223-232

    1716-1725