Targeted Disruption of01H1Encoding a Particular H1 Histone Variant Causes Changes in Protein Patterns in the DT40 Chicken B Cell Line

J. Mol. Biol. (1995) 254, 869–880

Targeted Disruption of 01H1 Encoding a Particular H1 Histone Variant Causes Changes in Protein Patterns in the DT40 Chicken B Cell Line Kiyomi Seguchi, Yasunari Takami and Tatsuo Nakayama* Department of Biochemistry Miyazaki Medical College Kihara, Kiyotake, Miyazaki 889-16, Japan

Six members of the chicken H1 gene family, all of which are located in two major histone gene clusters, have been shown to encode six different protein variants. The intracellular mRNA level from one of them, 01H1, encoding the 01H1 variant composed of 218 amino acid residues, constitutes 9.9% of the total H1 mRNA in the DT40 chicken B cell line. To study the specific role of this particular H1 variant, besides its well-known functions as a linker in chromatin maintenance and as a general repressor of transcription, we used targeted integration to construct heterozygous and homozygous DT40 mutants with disruption of one and two 01H1 alleles, respectively. Analyses of the stable transfectants showed that the growth rate of DT40 was unchanged in the absence of two 01H1 alleles. Moreover, the remaining H1 genes were shown to be expressed more in these mutants than in the wild-type cell lines. Two-dimensional polyacrylamide gel electrophoresis showed that within an almost constant background in the homozygous mutants several cellular proteins newly appeared or increased, while some other proteins disappeared or decreased quantitatively. These variable proteins all differed from those that varied in DT40 mutants deprived of one of the eight chicken H2B genes, H2B-V, encoding a particular H2B variant. These results suggest that the 01H1 variant is involved in the regulation of expression of genes that encode the proteins that vary in 01H1-deleted mutants of DT40 cells. 7 1995 Academic Press Limited

*Corresponding author

Keywords: H1 histone; targeted integration; RNase protection assay; transcription regulation; 2D-PAGE

Introduction In higher eukaryotes, it is known that the H1 histone is necessary for the condensation of nucleosome chains into higher order structures, whereas core histones (H2A, H2B, H3 and H4) mainly participate in nucleosome assembly (Kornberg, 1977; Felsenfeld, 1978; Isenberg, 1979; Hentschel & Birnstiel, 1981). The presence of multiple copies of the histone genes in most higher eukaryotes is especially well suited as a fundamental explanation of these rapid and precise biological events. In addition, a possible mechanism by which the intracellular levels of all the core histone subtypes remain constant and chromatin structure is precisely maintained during cell proliferation in Abbreviations used: G-418, neomycin (geneticin sulfate); BS, blasticidin S; GAPDH, glyceraldehyde3-phosphate dehydrogenase; IEF, isoelectro-focusing. 0022–2836/95/500869–12 $12.00/0

yeast and the chicken has been reported. Analyses of Saccharomyces cerevisiae mutants with disruptions of the H2A and H2B genes showed that this regulation is mediated through one of two H2A/H2B gene pairs, so that H2A and H2B possibly remain in stoichiometric balance with H3 and H4 (Osley & Hereford, 1981; Hereford et al., 1982; Meeks-Wagner & Hartwell, 1986; Norris & Osley, 1987; Osley & Lycan, 1987; Clark-Adams et al., 1988; Moran et al., 1990). On the other hand, in the chicken DT40 B cell line, which incorporates foreign DNA by targeted integration at high frequency (Buerstedde & Takeda, 1991; Takeda et al., 1992; Bezzubova et al., 1993), both H3-IV and H3-V, which are located in inverted orientation and share a 531 bp 3'-untranslated region (Setoguchi & Nakayama, 1991), produce about 25% of the total intracellular mRNA from all 12 H3 genes. Interestingly, when these two genes were disrupted in DT40 cells, the expression of the residual H3 7 1995 Academic Press Limited

870 genes increased to compensate for this disruption, resulting in the maintenance of a constant total intracellular level of H3 mRNA (Takami et al., 1995). Recently, it was reported that in a yeast mutant deprived of one of the two H2A/H2B gene pairs, the chromatin structure over HIS4 and GAL1 appeared essentially normal, but that over CYH2 and UB14 was dramatically disrupted (Norris et al., 1988). In this mutant, GAL1 repression was unaffected and HIS4 was constitutively expressed. Interestingly, the intracellular level of mRNA from UB14 increased approximately twofold, but the mRNA level from CYH2 was unchanged. Thus, deletion of a particular histone gene pair affected the expression of the several genes analyzed differentially, through chromatin disruption localized in specific regions of the yeast genome. Furthermore, in several eukaryotes, though not in yeast, a new role for the H1 histone has been studied in considerable detail, and results obtained from both in vivo and in vitro experiments have suggested that the H1 histone functions as a general repressor of transcription (Croston et al., 1991; Laybourn & Kadonaga, 1991, 1992; Croston et al., 1992; Kamakaka et al., 1993; Klucher et al., 1993). If the functions of eukaryotic H1 and core histones are restricted to the maintenance of chromatin structure involving nucleosome assembly and the participation in general repression of gene expression, a unique amino acid sequence should be sufficient for each histone subtype. However, there are several different variants of each histone subtype in various eukaryotes, and some of them are differentially synthesized during the cell cycle, development, etc. (Urban & Zweidler, 1983; Lennox & Cohen, 1984; von Holt et al., 1984; Nakayama et al., 1993). These facts led us to

Involvement of H1 in Transcription Regulation

suppose that they possess specific individual roles in particular biological events including gene expression, as mentioned previously (Huang & Cole, 1984; Coles et al., 1987). The chicken H1 gene family contains six members; 01H1, 02H1, .10H1, 11LH1, 11RH1 and 03H1 (Ruiz-Carrillo et al., 1983; Sugarman et al., 1983; D’Andrea et al., 1985; Coles et al., 1987). All of them differ in both their coding and non-coding regions, but they have considerably similar sequences which encode six different H1 variants (01H1, 02H1, .10H1, 11LH1, 11RH1 and 03H1) composed of 217 to 224 amino acid residues (Coles et al., 1987). To clarify the nature of the 01H1 variant, we have prepared heterozygous and homozygous mutants deprived of 01H1 (which encodes the variant) by targeted integration. The growth rates of all the mutant DT40 cell lines tested were the same; interestingly, however, the electrophoretic patterns of the total cellular proteins of the homozygous mutants were slightly, but definitely, distinct from those of the wild-type cell lines. These results indicate that the 01H1 variant may play a key role in the transcription regulation of putative genes encoding the proteins that varied in the DT40 cells.

Results 01H1 produces 9.9% of the total intracellular level of H1 mRNA in the DT40 cell line In the chicken, six H1 genes (01H1, 02H1, .10H1, 11LH1, 11RH1 and 03H1) are all located in two major histone gene clusters (Grandy & Dodgson, 1987) as shown in Figure 1. There are considerable differences in both their coding and non-coding

Figure 1. Organization of the chicken histone genes. The two major histone gene clusters reported by Grandy & Dodgson (1987) are shown with slight modifications. 1, 2A, 2B, 3 and 4 indicate H1, H2A, H2B, H3 and H4, respectively. The six H1 genes belonging to these clusters are designated as 01, 02, 03, .10, 11L and 11R (Coles et al., 1987). A subcluster, carrying 01H1, H2A-V and H2B-II, is shown enlarged with the coding regions and their orientations indicated. Cleavage sites; upward arrows, EcoRI; downward arrows, BamHI; open arrowheads, HindIII.

Involvement of H1 in Transcription Regulation

871

Figure 2. Expression of H1 genes in the DT40 cell line. A, Schematic illustration of the RNase protection method using probe 01H1. ORF indicates the open reading frame. B, Intracellular mRNA levels from 01H1 and the remaining H1 genes. DT40 cells were grown in Dulbecco’s modified medium to the logarithmic phase, and then total RNA was extracted. A 4 mg sample of total RNA (DT40 lane), together with 4 mg of total RNA from yeast (yeast lane), were analyzed using 32P-labeled antisense RNA probe 01H1. After electrophoresis in denaturating polyacrylamide gels, autoradiography was carried out. The radioactive intensities of the protected fragments of 01H1 and of the residual H1 genes were measured in a Fuji BAS 1000 Image Analyzer.

regions (Coles et al., 1987); the nucleotide (nt) sequence of 01H1, for instance, is almost identical to that of other H1 genes between positions +50 and +207, but considerably different in the region upstream of position +50. To quantify the intracellular levels of mRNA from 01H1 and those from other H1 genes in DT40 cells, we applied the RNase protection method using probe 01H1, which consisted of the antisense RNA fragment derived from the 123 bp 5'-flanking region including the 5'-untranslated sequence plus the 207 bp 5'-coding region of 01H1, in addition to the 33 bp flanking sequence of the plasmid vector (Figure 2A). This antisense RNA probe should therefore protect the 5'-untranslated and 5'-coding sequences of mRNA molecules of about 240 nt from 01H1 completely. On the other hand, it should protect only the internal portions of mRNA molecules of about 160 nt from other H1 genes (Figure 2A). As expected, the 240 nt band of the mRNA from 01H1 was distinguished from the several bands of mRNA of about 160 nt from the remaining H1 genes, probably due to both the similarities and considerable differences in the nucleotide sequences mentioned above (Figure 2B). The 240 nt band disappeared in DT40 subclones in which two 01H1 alleles were deleted by targeted integration, indicating that the band is derived exclusively from 01H1 (see Figure 6 cl 15-4 and cl 15-5 lanes). The intracellular level of 01H1 mRNA constituted 9.9% of the total mRNA from all the H1 genes.

High frequency of targeted integration into the 01H1 locus To clarify the nature of 01H1, we transfected DT40 cells with targeting vectors of the gene, since high frequencies of targeted integration have been achieved at all the different loci analyzed in the DT40 genome (Buerstedde & Takeda, 1991; Takeda et al., 1992; Bezzubova et al., 1993; Takami et al., 1995). An outline of the experimental procedure, together with that for the isolation of drug-resistant control cell lines, is shown in Figure 3. Genomic DNA samples were prepared from the stable transfectants and analyzed by Southern blotting after HindIII or XhoI digestion. The filter was first hybridized with probe 1, originating from sequence upstream of the 7.0 kb EcoRI fragment carrying the 01H1/H2A-V/H2B-II gene set (Figure 4C), followed by hybridization with probes 2 and 3 derived from neo and bsr, respectively (Figure 4D and E). In two drug-resistant control DT40 cell lines (neo-bsr-1 and -2) carrying neo and bsr, respectively, integrated randomly with pb/neo and pb/bsr constructs (Figure 3), probe 1 hybridized to a HindIII fragment of 16.3 kb and to a XhoI fragment of 11.0 kb (Figure 5Aa and b, lanes 1 and 2), as in the DT40 cell line and in several chicken tissues (data not shown). Probe 2 hybridized to HindIII fragments of 2.4 kb, although this probe showed no XhoI fragments, probably because they would be too large to be transferred to the membrane filter

872 (Figure 5Ba and b; lanes 1 and 2). Probe 3 hybridized to HindIII fragments of about 0.46 kb and to XhoI fragments of 9.5 kb and 4.0 kb, respectively (Figure 5Ca and b, lanes 1 and 2). As an initial step to obtain mutant cells deprived of 01H1, we introduced the 01H1/neo construct into DT40 cells. In this targeting vector, neo under the chicken b-actin promoter was inserted into the 01H1 coding region, and flanked upstream and downstream by sequences surrounding the gene (Figure 4A). As expected, after integration of the 01H1/neo construct into the 01H1 locus (Figure 4D), in two of the 28 stable transfectants selected with neomycin (geneticin; G-418), cl-15 and cl-19, probe 1 now hybridized to a HindIII fragment of 18.7 kb in addition to the endogenous fragment of 16.3 kb (Figure 5Aa, lanes 3 and 4), and to a XhoI fragment of 13.4 kb besides the endogenous fragment of 11.0 kb (Figure 5Ab; lanes 3 and 4). In addition, the intensities of these new bands were equal to those of the bands corresponding to the wild-type fragments. Probe 2 hybridized to a HindIII fragment of 18.7 kb and a XhoI fragment of 13.4 kb (Figure 5Ba and b, lanes 3 and 4), but probe 3 showed no bands (Figure 5Ca and b, lanes 3 and 4). Similar results were obtained with seven other clones. Our findings suggested that in these nine clones, one of two 01H1 alleles had been modified. Only random integration events were observed in the 19 remaining clones (data not shown). One of the nine G-418-resistant clones (cl-15; 01H1, +/−), in which one of two 01H1 alleles had been disrupted by targeted integration of the 01H1/neo construct, was then chosen for transfection of the 01H1/bsr construct. In this targeting construct, bsr transcribed by the chicken b-actin promoter was inserted into the 01H1 coding region (Figure 4B). Genomic DNA samples were isolated from the 16 G-418- and blasticidin S (BS)-resistant clones, and analyzed essentially as described above by Southern blotting. As expected, after integration

Involvement of H1 in Transcription Regulation

of the 01H1/bsr construct into the remaining 01H1 allele (Figure 4E), in two of the analyzed clones (cl-15-14 and cl-15-15), probe 1 now hybridized to a HindIII fragment of 8.7 kb other than the HindIII fragment of 18.7 kb (Figure 5Aa, lanes 5 and 6), and also to a XhoI fragment of 10.0 kb in addition to the XhoI fragment of 13.4 kb (Figure 5Ab, lanes 5 and 6). Probe 3 hybridized to a HindIII fragment of 0.46 kb and a XhoI fragment of 4.0 kb, but probe 2 showed no additional bands. The 14 other clones showed no targeted integration in the remaining 01H1 allele (data not shown). These results therefore, together with those from the first allele, indicated that targeted integration into the 01H1 locus in DT40 had occurred at a high frequency.

Decrease and disappearance of intracellular mRNA from 01H1 in the heterozygous and homozygous mutants We measured the intracellular levels of mRNA from 01H1 and from the remaining H1 genes in DT40 and its five subclones by the RNase protection method using probe 01H1, together with probe GAPDH (see Materials and Methods) to normalize differences in the amounts of total RNA used. The amounts and patterns of H1 mRNA samples from the drug-resistant control cell line (neo-bsr-1; 01H1, +/+) were essentially identical to those of DT40 cells (Figure 6, first two lanes; Figure 6, Table), showing that the two exogenous genes, neo and bsr, were scarcely related to the expression of all the H1 genes. The intracellular levels of mRNA from 01H1 in two heterozygous mutants (cl-15 and cl-19; 01H1, +/−), appearing as a band of 240 nt (Figure 6, cl-15 and cl-19 lanes; Figure 6, Table), were essentially the same, and were about 50% of those in the control cell lines (DT40 and neo-bsr-1; 01H1, +/+). In two homozygous mutants (cl-15-14 and cl-15-15; 01H1, −/−), no band of 240 nt was detected (Figure 6,

Figure 3. Outline of the transfection procedure used to isolate heterozygous and homozygous mutants.

Involvement of H1 in Transcription Regulation

873

Figure 4. Schematic diagram of the homologous recombination resulting in reductions of the first and second alleles of 01H1. a, Targeting vector, 01H1/neo construct. b, Targeting vector, 01H1/bsr construct. The open boxes (neo and bsr) indicate neo and bsr, respectively, under the control of the chicken b-actin promoter represented by the hatched area (b-actin pro). The plasmids were linearized at the ClaI site. c, 01H1 locus in the genome of DT40 cells. d, 01H1 locus in DT40 clones after targeted integration of the 01H1/neo construct. e, 01H1 locus in DT40 clones after targeted integration of the 01H1/bsr construct. 01H1, H2A-V and H2B-II, respectively, indicate their open reading frames. The locations of probes 1, 2 and 3 are indicated by bars. Only relevant restriction sites are indicated. Possible relevant fragments obtained by HindIII or XhoI digestion are shown with lengths in kb.

874 cl-15-14 and cl-15-15 lanes; Figure 6, Table). As mentioned in the previous section, these results confirmed that targeted integration had occurred at one and two 01H1 alleles, respectively, in the heterozygous and homozygous mutants. In the two heterozygous mutants (01H1, +/−), the expressions of other H1 genes increased slightly, and the mRNA levels were about 2% higher than those in the wild-type cell lines (01H1, +/+). In the two homozygous mutants (01H1, −/−), other H1 genes were transcribed at a higher level than in the two wild-type cell lines, the increase being about 9% (Table in Figure 6). In addition, in the heterozygous and homozygous mutants, the average sums of mRNAs from 01H1 and those from the other H1 genes were 96.5% and 98.4% of the normal levels, respectively. Thus, the H1 gene family, like the H3 gene family (Takami et al., 1995), has an ability to maintain the intracellular levels of transcripts.

Involvement of H1 in Transcription Regulation

No change in growth rate even on disruption of two 01H1 alleles The influence of 01H1-disruption on the growth rate was examined (Figure 7). As control cell lines, we also used two G-418- and BS-resistant cell lines (neo-bsr-1 and -2; 01H1, +/+) carrying both neo and bsr inserted randomly. We then studied the growth rates of two heterozygous mutants (cl-15 and cl-19; 01H1, +/−), in which one 01H1 allele was replaced by neo, and of two homozygous mutants (cl-15-14 and cl-15-15; 01H1, −/−), in which the two 01H1 alleles were replaced by neo and bsr. No difference was detected in the growth rates of DT40 and the two drug-resistant control cell lines, the doubling times being about 12 hours, indicating that the expressions of both neo and bsr did not cause any change in cell growth. Interestingly, the doubling times (11.9 to 12.2 hours) of the heterozygous and homozygous mutants analyzed

Figure 5. Southern blot analyses of homologous recombination events. Genomic DNA samples were prepared from two control cell lines carrying both neo and bsr integrated randomly (neo-bsr-1 and neo-bsr-2, lanes 1 and 2), two G-418-resistant mutants after integration of the 01H1/neo construct (cl-15 and cl-19, lanes 3 and 4), and two G-418- and BS-resistant mutants after integration of the 01H1/bsr construct (cl-15-14 and cl-15-15, lanes 5 and 6). The HindIII fragments (a) and XhoI fragments (b) were analyzed with probe 1 (A), 2 (B) or 3 (C).

Involvement of H1 in Transcription Regulation

875

Figure 6. Effects of 01H1 disruption on the intracellular levels of H1 mRNA. Experimental procedures were essentially as for Figure 2. Total RNA samples were prepared from two wild-type (DT40 and neo-bsr-1; 01H1, +/+), two heterozygous (cl-15 and cl-19; 01H1, +/−) and two homozygous (cl-15-14 and cl-15-15; 01H1, −/−) cell lines, and analyzed with a combination of probes 01H1 and GAPDH. In addition, yeast RNA samples were analyzed; the labeled probes were also run. The radioactive intensities of mRNA from other H1 genes in DT40 cells were assigned the values of 100.0, and the relative values obtained are shown in the Table. The ratio of the intensities of the mRNA from 01H1 to those of the mRNA from all the H1 genes are also shown.

were also essentially identical to those of the three wild-type cell lines (01H1, +/+). These results suggested that disruption of 01H1, which produces about 10% of the total intracellular level of H1 mRNA, had no effect on the growth rate of DT40. Changes in protein patterns after disruption of two 01H1 alleles Because 01H1 mRNAs were definitely present in DT40 cells, although at low levels, the 01H1 variant was expected to participate in some particular functions. To clarify the nature of this variant, we compared the total cellular proteins of the homozygous mutants deprived of two 01H1 alleles with those of the wild-type DT40 cell lines. As a control we used the neo-bsr-1 cell line (01H1, +/+) in which neo and bsr were inserted independently by random integration. Total cellular proteins were prepared essentially as described (Sambrook et al., 1989) from exponentially growing DT40 subclones, and analyzed by two-dimensional (2D)-PAGE (Nishine et al., 1991).

Under the conditions used, we separated the proteins, based on differences in both their isoelectric points (pI) in ranges of about 4 to 8, and molecular masses (Mr ) in ranges of about 10 to 200 kDa, respectively. Therefore, all histone subtypes with high pI values of about 12 could not be detected in our 2D-PAGE gels. No difference in the electrophoretic patterns of DT40 and the drug-resistant control cell line was observed (data not shown), and the aminoglycoside-3'-phosphotransferase of 56 kDa and the BS deaminase of 16 kDa, produced from neo and bsr, respectively, could not be detected even in the drug-resistant control cell line, probably because their amounts were very low (Figure 8A). The electrophoretic patterns of the proteins from the drug-resistant wild-type cell line (neo-bsr-1; 01H1, +/+) were almost the same as those from the homozygous mutant (cl-15-14; 01H1, −/−; Figure 8A and B). However, when compared in detail, several distinct variations were observed within this constant background. For instance, the 105 kDa, 28 kDa and 25 kDa proteins, and also some other

876

Involvement of H1 in Transcription Regulation

Figure 7. Growth curves of DT40 and its subclones. DT40 and six subclones were grown, and cell numbers were determined at the indicated times. The numbers are plotted on a log scale. Values are averages for three independent experiments. Symbols for the cell lines are shown in the insert.

proteins, were observed in the wild-type cell line, but were absent or present at lower concentrations in the homozygous mutant. On the other hand, the

spots of 42 kDa, 35 kDa, 34 kDa and 21 kDa, possibly with other proteins, either newly appeared or were clearly increased in the mutant. Judging

Figure 8. Comparison of total cellular proteins of the wild-type and homozygous cell lines by two-dimensional polyacrylamide gel electrophoresis. Total cellular proteins were prepared as described in the text from two control cell lines (DT40 and neo-bsr-1; 01H1, +/+; a and A) and two homozygous mutants (cl-15-14 and cl-15-15; 01H1, −/−; B and b). Isoelectric focusing in the first dimension was performed in gels using wide range ampholine (pH 3 to 10). The effective separation range was pH 4 (right) to 8 (left). SDS-PAGE in the second dimension was performed in 12.5% (w/v) acrylamide. Downward arrows indicate proteins that disappeared or decreased in the homozygous mutants. a, 105 kDa; b, 28 kDa; c, 25 kDa. Upward arrows indicate proteins that appeared or increased in the homozygous mutants. d, 42 kDa; e, 35 kDa; f, 34 kDa; g, 21 kDa.

Involvement of H1 in Transcription Regulation

from their Mr and pI, these proteins that varied did not correspond to either the chicken H1 histone (Mr , about 25 kDa; pI, about 12) or the two exogenous drug-related proteins of 56 kDa and 16 kDa. As shown in part in Figure 8a and b, for example, the proteins of 42 kDa and 34 kDa were present at lower concentrations in DT40 but had definitely increased in another homozygous mutant (cl-15-15; 01H1, −/−). Therefore, these variations in the protein patterns were not due simply to clonal deviation. Thus, these results clearly demonstrated that disruption of 01H1 encoding the 01H1 variant caused not only decreases in the amounts of the proteins of 105 kDa, 28 kDa, 25 kDa, etc. but also increases in the amounts of the proteins of 42 kDa, 35 kDa, 34 kDa, 21 kDa, etc.

Discussion It has been suggested that histones are involved in the regulation of gene expression in yeast and several higher eukaryotes. Results of studies with a Saccharomyces cerevisiae mutant with disruption of one of two H2A/H2B gene pairs encoding two different variants H2A and H2B showed that chromatin disruption due to this mutation was localized in specific regions of the yeast genome and influenced the expression of particular genes (Norris & Osley, 1987; Norris et al., 1988). On the other hand, analyses in several higher eukaryotic systems provided biochemical evidence that the H1 histone is a general repressor of transcription (Croston et al., 1991, 1992; Laybourn & Kadonaga, 1991, 1992; Kamakaka et al., 1993; Klucher et al., 1993). There are eight members of the chicken H2B gene family, all of which encode four different variants (classes I to IV; Grandy & Dodgson, 1987; Nakayama & Setoguchi, 1991; Nakayama et al., 1993; our unpublished results). H2B-V, one of these eight H2B genes, encodes the class III H2B variant and has particular transcriptional elements (Ohshige et al., 1993). The mRNA levels from H2B-V in the oviduct and lung of chickens are at most half those in the kidney (Nakayama & Setoguchi, 1992). Recently, we identified a specific role for this class III H2B variant in the DT40 chicken B cell line, by comparison of total cellular proteins by 2DPAGE (Takami et al., unpublished results). In a homozygous mutant bearing a disruption of two H2B-V alleles, several proteins were absent or quantitatively decreased, but some new proteins appeared. Thus the disruption in the class III H2B variant undoubtedly caused changes in the protein pattern, probably through alterations in the expression of putative genes encoding the proteins that varied. All six chicken H1 genes differ considerably in their coding sequences, and in their 5'- and 3'-flanking sequences, and each of them encodes a different H1 variant (01H1, 02H1, .10H1, 11LH1, 11RH1 or 03H1; Coles et al., 1987). One of these genes, 01H1, encoding the 01H1 variant consisting of 218 amino acid residues, normally produces 9.9%

877 of the total intracellular mRNA from all the H1 genes in DT40 cells (Figure 2B). To determine the nature of this 01H1 variant, we carried out targeted integration into the 01H1 locus, and then studied the characteristics of heterozygous (01H1, +/−) and homozygous (01H1, −/−) mutants, respectively, with deletions of one and two 01H1 alleles. The intracellular levels of transcripts from 01H1 in the heterozygous mutants were about half the normal level in wild-type cell lines (01H1, +/+), and no 01H1 transcripts were detected in the homozygous mutants (Figure 6). As with both the H3-IV/H3-V disruption (Takami et al., 1995) and the H2B-V disruption (Takami et al., unpublished results), we observed an alteration in the intracellular mRNA levels from the remaining H1 genes in all the 01H1-disrupted mutants analyzed (Figure 6); nevertheless the H1 mRNA level decreased at most 10% even in the homozygous mutants (Figures 2B and 6). No variation was detected in the growth rates, even after two 01H1 alleles were disrupted (Figure 7). Analyses by 2D-PAGE showed that the proteins in the homozygous mutants were, on the whole, identical to those in the wild-type cell lines. However, interestingly, within an almost constant background, the patterns were slightly, but distinctly, different (Figure 8). In the homozygous mutants, proteins of 105 kDa, 28 kDa and 25 kDa, and some others had disappeared or decreased substantially, whereas those of 42 kDa, 35 kDa, 34 kDa and 21 kDa and some others had increased or newly appeared. These proteins that varied did not correspond to either the chicken 01H1 variant or the two products from neo and bsr, judging from their Mr and pI values. In addition, none of these proteins corresponded to any of the proteins that varied in the mutants deprived of the class III H2B variant (Takami et al., unpublished results). Together, these observations demonstrate that the 01H1 variant, like the class III H2B variant, plays a specific role in transcription regulation in DT40 cells, in addition to a vital role in chromatin organization. To explain the variations in the protein patterns in the 01H1-disrupted mutants, we propose that 01H1 mutation alters the chromatin structure over genes encoding particular proteins, resulting in negative control of the expression of those encoding proteins of 105 kDa, 28 kDa, 25 kDa, etc., and in positive control of the expression of those encoding proteins of 42 kDa, 35 kDa, 34 kDa, 21 kDa, etc. In this model, chromatin over these putative genes would normally include at least one molecule of the 01H1 variant, and in the 01H1-disrupted mutants the variant should be replaced by any of the remaining H1 variants. However, this model is hypothetical as there is no definitive biological evidence that the chromatin structure really varies, not as in the H2B-V-disrupted DT40 mutant (Takami et al., unpublished results). Further studies on the disruption of each of the residual H1 genes are essential to identify whether each member of the H1

878 family has a specific function in transcription regulation.

Materials and Methods Cell cultures DT40 cells and all subclones were cultured essentially as described (Buerstedde et al., 1990; Takeda et al., 1992), in Dulbecco’s modified medium containing 10% (v/v) fetal calf serum, 1% (v/v) chicken serum, 2 mM L-glutamine, 0.01 mM b-mercaptoethanol, and penicillin/ streptomycin at 37°C under 5% CO2 in air in a humidified incubator unless otherwise stated. At the indicated times, cell numbers were counted to determine the growth rate. Gene constructs As a source of the chicken genomic DNA of 01H1 we used a 7.0 kb EcoRI fragment carrying a 01H1/H2A-V/ H2B-II gene set (Nakayama & Setoguchi, 1991), and inserted this EcoRI fragment into the EcoRI site of Bluescript II (Stratagene) to prepare the parental chimeric plasmid. To obtain the 01H1/neo targeting construct, we inserted the blunt-ended SalI/BamHI fragment, carrying neo under the chicken b-actin promoter (Fregien & Davidson, 1986; Miyazaki et al., 1989), into the blunt-ended BclI site of the 01H1 coding region of the resultant chimeric plasmid. To obtain the 01H1/bsr targeting construct, the blunt-ended XhoI/BamHI fragment, carrying bsr (Kamakura et al., 1987) under the chicken b-actin promoter, was similarly inserted into the blunt-ended BclI site of the 01H1 coding region of the chimeric plasmid. Before transfection into DT40 cells, we linearized the 01H1/neo or 01H1/bsr construct by ClaI digestion. Probe 1 was the EcoRI fragment of about 600 bp derived from the sequence upstream of the 7.0 kb EcoRI fragment. Probe 2 was the 1.3 kb HindIII/SmaI fragment of neo, and probe 3 was the HindIII fragment of bsr of about 460 bp. Transfection and isolation of transfectants Transfection conditions were essentially as described (Buerstedde et al., 1990; Takeda et al., 1992; Takami et al., 1995). Transfectants with the 01H1/neo construct were selected in medium containing 2 mg/ml neomycin (geneticin; G-418). We transfected the 01H1/bsr construct into clones in which one of two 01H1 alleles had already been disrupted, and selected transfectants in medium containing 2 mg/ml G-418 and 15 mg/ml blasticidin S (BS), respectively. Southern blot analysis Samples of 10 mg of genomic DNA were digested with the indicated enzymes, separated in 0.6% (w/v) agarose gels, transferred to a Hybond N + membrane according to the supplier’s protocol (Amersham), and hybridized by the method of Southern (1975) with probe 1, 2 or 3. The DNA probes were labeled with 32P by random-priming as recommended by the supplier (Amersham).

Involvement of H1 in Transcription Regulation

5'-flanking and 207 bp 5'-coding regions of 01H1, was inserted between the EcoRI and HincII sites of Bluescript II. 32P-labeled antisense RNAs were synthesized with T7 RNA polymerase after cutting with EagI, and used as probe 01H1. This probe was 363 nt long (Figure 6, probe lane), since it consisted of the 330 nt fragment from 01H1, in addition to the 33 nt fragment from the vector on the 5'-end. The HindIII/ApaI fragment of 129 bp of the chicken muscle glyceraldehyde-3-phosphate dehydrogenase (GAPDH) cDNA (Panabieres et al., 1984) was inserted between the HindIII and ApaI sites of Bluescript II. After cutting with BamHI, the labeled antisense RNAs were similarly synthesized, and used as probe GAPDH. This probe was 173 nt long, and consisted of the 129 nt fragment from the HindIII/ApaI fragment of the chicken GAPDH cDNA and the 44 nt fragment from the vector (Figure 6, probe lane). Total RNA was isolated from exponentially growing DT40 subclones as described (Chomczynski & Sacchi, 1987). The intracellular levels of H1 mRNA were quantified by the RNase protection method with [32P]CMP-labeled RNA probe and an Ambious RPAII kit according to the manufacturer’s protocol. Probe 01H1 displayed no bands in yeast RNA fractions (Figure 2B, lane 1; Figure 6, yeast lane), since yeast has no H1 gene. Probe GAPDH was used to normalize differences in the amounts of total RNA used. After electrophoresis in denaturating polyacrylamide gels, autoradiography was carried out. The intensities of the radioactivity of protected fragments for 01H1 and for the remaining H1 genes were then quantified with a Fuji BAS 1000 Image Analyzer. In the former case the value was corrected by the ratio of the number (63) of C nucleotides between positions +50 and +207 to that (101) between positions −30 and +207 of the antisense RNA probe, since the putative initiation site of 01H1 is near position −30. Two-dimensional gel electrophoresis Total cellular proteins were prepared from exponentially growing DT40 subclones as described (Sambrook et al., 1989), and separated in an automated apparatus for two-dimensional electrophoresis, TEP-1 (Shimadzu, Japan) (Nishine et al., 1991). The conditions of the firstdimensional isoelectro-focusing (IEF) gel were 5% (w/v) acrylamide, 6% (v/v) ampholine (pH 3 to 10), 8 M urea and 2% (v/v) NP-40. After electrophoresis for 16 hours at 800 V and 20°C, the IEF-gel was automatically transferred to equilibration buffer (5% (v/v) b-mercaptoethanol, 2.5% (w/v) SDS and 8 M urea) on the second-dimensional SDS-polyacrylamide gel electrophoresis (SDS-PAGE) gel for five minutes. Then the second run was carried out using 0.1% (w/v) SDS and 12.5% (w/v) acrylamide at 15°C, for 30 minutes at 150 V and for 3.5 hours at 300 V. The SDS-PAGE gel was then stained with Coomassie blue.

Acknowledgement This work was supported in part by a Grant-in-Aid from the Ministry of Education, Science, Sports and Culture of Japan.

RNase protection method An antisense RNA probe to measure the amounts of H1 mRNA was prepared as follows. The EcoRI/NaeI fragment of about 870 bp, containing the 660 bp

References Bezzubova, O., Shinohara, A., Mueller, R. G., Ogawa, H. & Buerstedde, J.-M. (1993). A chicken RAD51

Involvement of H1 in Transcription Regulation

homologue is expressed at high levels in lymphoid and reproductive organs. Nucl. Acids Res. 21, 1577–1580. Buerstedde, J.-M. & Takeda, S. (1991). Increased ratio of targeted to random integration after transfection of chicken B cell lines. Cell, 67, 179–188. Buerstedde, J.-M., Reynaud, C.-A., Humphries, E. H., Olson, W., Ewert, D. L. & Weill, J. C. (1990). Light chain gene conversion continues at a high rate in an ALV-induced cell line. EMBO J. 9, 921–927. Chomczynski, P. & Sacchi, N. (1987). Single-step method of RNA isolation by acid guanidinium thiocyanatephenol-chloroform extraction. Anal. Biochem. 162, 156–159. Clark-Adams, C. D., Norris, D., Osley, M. A., Fassler, J. S. & Winston, F. (1988). Changes in histone gene dosage alter transcription in yeast. Genes Dev. 2, 150–159. Coles, L. S., Robins, A. J., Madley, L. K. & Wells, J. R. E. (1987). Characterization of the chicken histone H1 gene complement. Generation of a complete set of vertebrate H1 protein sequences. J. Biol. Chem. 262, 9656–9663. Croston, G. E., Kerrigan, L. A., Lira, L. M., Marshak, D. R. & Kadonaga, J. T. (1991). Sequence-specific antirepression of histone H1-mediated inhibition of basal RNA polymerase II. Science, 251, 643–649. Croston, G. E., Laybourn, P. J., Paranjape, S. M. & Kadonaga, J. T. (1992). Mechanism of transcriptional antirepression by GAL4-VP16. Genes Dev. 6, 2270– 2281. D’Andrea, R. J., Coles, L. S., Lesnikowski, C., Tabe, L. & Wells, J. R. E. (1985). Chromosomal organization of chicken histone genes: preferred associations and inverted duplications. Mol. Cell. Biol. 5, 3108–3115. Felsenfeld, G. (1978). Chromatin. Nature, 271, 115–121. Fregien, N. & Davidson, N. (1986). Activating elements in the promoter region of the chicken b-actin gene. Gene, 48, 1–11. Grandy, D. K. & Dodgson, J. B. (1987). Structure and organization of the chicken H2B histone gene family. Nucl. Acids Res. 15, 1063–1080. Hentschel, C. C. & Birnstiel, M. L. (1981). The organization and expression of histone gene families. Cell, 25, 301–313. Hereford, L., Bromley, S. & Osley, M. A. (1982). Periodic transcription of yeast histone genes. Cell, 30, 305–310. Huang, H.-C. & Cole, R. D. (1984). The distribution of H1 histone is nonuniform in chromatin and correlates with different degrees of condensation. J. Biol. Chem. 259, 14237–14242. Isenberg, I. (1979). Histones. Annu. Rev. Biochem. 48, 159–191. Kamakaka, R. T., Bulger, M. & Kadonaga, J. T. (1993). Potentiation of RNA polymerase II transcription by Gal4-VP16 during but not after DNA replication and chromatin assembly. Genes Dev. 7, 1779–1795. Kamakura, T., Kobayashi, K., Tanaka, T., Yamaguchi, I. & Endo, T. (1987). Cloning and expression of a new structural gene for blasticidin S deaminase, a nucleotide aminohydrolase. Agric. Biol. Chem. 51, 3165–3168. Klucher, K. M., Sommer, M, Kadonaga, J. T. & Spector, D. H. (1993). In vivo and in vitro analysis of transcriptional activation mediated by the human cytomegalovirus major immediate-early proteins. Mol. Cell. Biol. 13, 1238–1250. Kornberg, R. D. (1977). Structure of chromatin. Annu. Rev. Biochem. 46, 931–954.

879 Laybourn, P. J. & Kadonaga, J. T. (1991). Role of nucleosomal cores and histone H1 in regulation of transcription by RNA polymerase II. Science, 254, 238–245. Laybourn, P. J. & Kadonaga, J. T. (1992). Threshold phenomena and long-distance activation of transcription by RNA polymease II. Science, 257, 1682–1685. Lennox, R. W. & Cohen, L. H. (1984). The H1 subtypes of mammals: metabolic characteristics and tissue distribution. In Histone Genes; Structure, Organization and Regulation (Stein, G. S., Stein, J. L. & Marzluff, W. F., eds), pp. 373–395, John Wiley & Sons, New York. Meeks-Wagner, D. & Hartwell, L. H. (1986). Normal stoichiometry of histone dimer sets is necessary for high fidelity of mitotic chromosome transmission. Cell, 44, 43–52. Miyazaki, J., Takaki, S., Araki, K., Tashiro, F., Tominaga, A., Takatsu, K. & Yamamura, K. (1989). Expression vector system based on the chicken b-actin promoter directs efficient production of interleukin-5. Gene, 79, 269–277. Moran, L., Norris, D. & Osley, M. A. (1990). A yeast H2A-H2B promoter can be regulated by changes in histone gene copy number. Genes Dev. 4, 752–763. Nakayama, T. & Setoguchi, Y. (1991). Nucleotide sequence of a member of the chicken H2B histone-encoding gene family. Gene, 98, 299–300. Nakayama, T. & Setoguchi, Y. (1992). Structural change of a particular H2B histone gene possibly results in differences in its transcriptional regulation in different chicken tissues. Nucl. Acids Res. 20, 1309–1314. Nakayama, T., Takechi, S. & Takami, Y. (1993). The chicken histone gene family. Comp. Biochem. Physiol. 104B, 635–639. Nishine, T., Nakamura, S., Hazama, M., Koga, M. & Nokihara, K. (1991). Development and applications of a novel apparatus for two-dimensional electrophoresis involved a transfer robot. Shimadzu Rev. 48, 247–251. Norris, D. & Osley, M. A. (1987). The two gene pairs encoding H2A and H2B play different roles in the Saccharomyces cerevisiae life cycle. Mol. Cell. Biol. 7, 3473–3481. Norris, D., Dunn, B. & Osley, M. A. (1988). The effect of histone gene deletions on chromatin structure in Saccharomyces cerevisiae. Science, 242, 759–761. Ohshige, T., Takechi, S. & Nakayama, T. (1993). Presence of particular transcription regulatory elements in the 5'-intergenic region shared by the chicken H2A-III and H2B-V pair. Gene, 131, 193–199. Osley, M. A. & Hereford, L. M. (1981). Yeast histone genes show dosage compensation. Cell, 24, 377–384. Osley, M. A. & Lycan, D. (1987). Trans-acting mutations that alter transcription of Saccharomyces cerevisiae histone genes. Mol. Cell. Biol. 7, 4204–4210. Panabieres, F., Piechaczyk, M., Rainer, B., Dani, C., Fort, P., Riaad, S., Marty, L., Imbach, J. L., Jeanteur, P. & Blanchard, J. M. (1984). Complete nucleotide sequence of the messenger RNA coding for chicken muscle glyceraldehyde-3-phosphate dehydrogenase. Biochem. Biophys. Res. Commun. 118, 767–773. Ruiz-Carrillo, A., Affolter, M. & Renaud, J. (1983). Genomic organization of the genes coding for the six main histones of the chicken: complete sequence of the H5 gene. J. Mol. Biol. 170, 843–859. Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning. A Laboratory Manual, 2nd edit., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York.

880 Setoguchi, Y. & Nakayama, T. (1991). Nucleotide sequences of new members (H3-IV and H3-V) of the chicken H3 histone-encoding gene family. Nucl. Acids Res. 19, 6327. Southern, E. M. (1975). Detection of specific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol. 98, 503–517. Sugarman, B. J., Dodgson, J. B. & Engel, J. D. (1983). Genomic organization, DNA sequence, and expression of chicken embryonic histone genes. J. Biol. Chem. 258, 9005–9016. Takami, Y., Takeda, S. & Nakayama, T. (1995). Targeted disruption of an H3-IV/H3-V gene pair causes increased expression of the remaining H3 genes

Involvement of H1 in Transcription Regulation

in the chicken DT40 cell line. J. Mol. Biol. 250, 420–433. Takeda, S., Masteller, E. L., Thompson, C. B. & Buerstedde, J.-M. (1992). RAG-2 expression is not essential for chicken immunoglobulin gene conversion. Proc. Natl Acad. Sci. USA, 89, 4023–4027. Urban, M. K. & Zweidler, A. (1983). Changes in nucleosomal core histone variants during chicken development and maturation. Dev. Biol. 95, 421–428. von Holt, C., De Groot, P., Schwager, S. & Brandt, W. F. (1984). Organization and expression of mouse histone genes. In Histone Genes; Structure, Organization and Regulation (Stein, G. S., Stein, J. L. & Marzluff, W. F., eds), pp. 65–105, John Wiley & Sons, New York.

Edited by J. Karn (Received 26 May 1995; accepted 19 September 1995)