Late H2B Research Articles

Structure of histone H2B and H4 genes of the sea cucumber Holothuria tubulosa.

A genomic library of the sea cucumber Holothuria tubulosa was screened with human and murine histone gene hybridization probes. A recombinant phage carrying an H4 gene was isolated and sequenced. Hybridization analysis of the entire 20 kb phage insert with probes for H1, H2A, H2B and H3 histones was negative except for H2B. This solitary arrangement of the two neighbouring histone H4 and H2B genes is in contrast to the organization of 'cleavage stage' histone genes, which are arranged in tandem quintets of genes encoding the 5 histone classes. Gene organization and sequence features indicate that the two Holothuria genes are the equivalents of a known pair of late H2B and H4 genes which have been described in the genome of sea urchins. This result shows that the simultaneous occurrence of tandem repeats of histone gene quintets and smaller groups of structurally distinct histone gene subtypes is not unique for sea urchins, but also applies to other echinodermata, such as the sea cucumber Holothuria tubulosa.

Direct Induction of DNA Hypermethylation in Sea Urchin Embryos by Microinjection of 5-Methyl dCTP Stimulates Early Histone Gene Expression and Leads to Developmental Arrest

The role of DNA methylation in gene inactivation has been studied extensively in vertebrates but it is not clear whether it serves similar functions in other organisms. We devised a novel approach to induce hypermethylation of both endogenous and injected DNA in the sea urchin Lytechinus pictus in order to study the effect of DNA methylation on gene expression in this invertebrate. By injecting 5-methyl dCTP either alone or together with a cloned DNA construct into fertilized sea urchin eggs, replicating DNA became hypermethylated from the random incorporation of the methylated nucleotide in place of cytosine during DNA synthesis. During subsequent rounds of replication, the injected 5-methyl dCTP became depleted but methylation at CpG sites was still elevated presumably due to the action of a methyltransferase enzyme. Using this approach, we studied the effect of hypermethylation on two members of the sea urchin multigene family, the early H2B and the late H2B genes. De novo methylation was shown to occur at known cis-regulatory regions of the genes. The effect of methylation on gene activity was probed using RNase protection assay. Methylation resulted in increased early H2B histone gene expression but had no effect on late H2B histone gene expression. These results demonstrate that methylation does not necessarily inactivate genes in the sea urchins as previously thought. Interestingly, the development of embryos injected with 5-methyl dCTP typically was arrested at the blastula stage, and analysis of the genomic DNA extracted from injected embryos showed a significant increase in the endogenous methylation content. These data suggest that perturbation of methylation patterns in developing sea urchin embryos may be responsible for the developmental arrest through altering the gene expression pattern.

Histone H2A.F/Z mRNA is stored in the egg cytoplasm and basally regulated in the sea urchin embryo

The sea urchin H2A.F/Z histone is a member of a subclass of highly conserved H2A variants. Sequence analysis confirms that H2A.F/Z mRNA is polyadenylated. In situ hybridization studies demonstrate that maternal H2A.F/Z message is stored in the egg cytoplasm and present at equal levels in all cells of the mesenchyme blastula-stage embryo, suggesting that H2A.F/Z is not coordinately regulated with DNA synthesis. When blastula-stage embryos were exposed to DNA synthesis inhibitors, no effect on the steady-state level of H2A.F/Z mRNA was observed, while the level of late class H2B mRNA decreased substantially. These results provide evidence that the basal mode of regulation of this unusual histone variant is conserved evolutionarily.

An octamer element is required for the expression of the alpha H2B histone gene during the early development of the sea urchin

Early (alpha) histone genes are one of several histone gene families in the sea urchin genome. They are expressed at high levels in blastula-stage embryos and are inactivated by the early gastrula stage. By microinjecting mutant early H2B genes into sea urchin zygotes and monitoring their transcriptional activity in blastula- and gastrula-stage embryos, we sought to identify the cis-regulatory elements responsible for this dramatic change in early H2B gene activity. We found that deletion of DNA 5′ of −71 and 3′ of +591 did not affect the timing or magnitude of early H2B gene expression. Neither was early H2B gene expression affected by the replacement of sequences downstream of −36 with the corresponding region of the L1 late H2B gene, expressed after the peak transcription of the early H2B gene. Further deletion of early H2B promoter sequences from −71 to −56, removing a conserved octamer element, resulted in near-complete inactivation of the early H2B gene in both blastula- and gastrula-stage embryos. Also inactivating early H2B gene expression were an internal deletion of the octamer element and a base substitution mutation that altered its sequence. This base substitution mutation also caused a parallel reduction in the ability of the octamer element to bind a factor present in nuclear extracts of sea urchin blastulae. These data strongly suggest that the proper expression of the early H2B gene in cleavage- and blastula-stage embryos depends on the octamer element and a factor with which it interacts.

Altered cell fate in LiCl-treated sea urchin embryos

Endo16 is a lineage-specific protein expressed during embryogenesis in the sea urchin, Strongylocentrotus purpuratus. We have examined the effects of various concentrations of lithium chloride, a well-known vegetalizing agent, on Endo16 expression in whole sea urchin embryos. Our results show that treatment with LiCl causes increased steady-state levels of Endo16 transcripts. An increase in the number of endodermal cells in treated embryos demonstrates a change in cell lineage. In addition, we observed a delay in development of lithium-treated embryos that is accompanied by a delay in the expression of a late class histone H2B gene.

Activation of the L1 late H2B histone gene in blastula-stage sea urchin embryos by an Antennapedia-class homeoprotein

The L1 late H2B histone gene of the sea urchin Strongylocentrotus purpuratus is transcriptionally activated in late blastula stage embryos by a mechanism that depends on an enhancer element located 3' of the gene (Zhao et al., 1990). A protein factor, designated H2B abp 1, binds this element at a site that resembles the consensus recognition sequence of Antennapedia-class homeodomain proteins. We demonstrate here that Antennapedia (Antp) and Hbox4 proteins, members of the Antennapedia class of homeoproteins from Drosophila and sea urchin respectively, bind the L1 H2B abp 1 site, and that the Drosophila Antp protein acts through this site to trans-activate the L1 H2B gene, in vivo. In addition, RNA gel blot analysis demonstrated that Hbox4 transcripts accumulate in developing embryos with a time course that closely resembles that of H2B adp 1 DNA binding activity and the activity and the transcription rate of the L1 late H2B gene. Finally, we show that antibody prepared against the sea urchin Hbox4 protein, a member of the Abd-B subclass of the Antennapedia class, specifically inhibits binding of the H2B abp 1 factor to the L1 H2B enhancer, suggesting that H2B abp 1 is encoded by Hbox4 or a closely related gene.

Activation of a Late H2B Histone Gene in Blastula-Stage Sea Urchin Embryos by an Unusual Enhancer Element Located 3′ of the Gene

In the sea urchin embryo, late histone genes are transcribed at low levels during cleavage and blastula formation and at substantially higher levels in later stages of embryogenesis. To investigate the molecular basis of the stage-specific expression of a late H2B histone gene, we injected mutant genes lacking portions of 5'- and 3'-flanking regions into Lytechinus pictus embryos and monitored their expression by RNase protection. A 200-bp region located 489 bp downstream of the mRNA 3' terminus was necessary for the increase in transcription of the late H2B gene at the mid-blastula stage of development. DNase I and methylation interference footprint analyses located only one factor-binding site in this region, and gel mobility shift experiments showed that the DNA-binding activity of this factor (designated H2B abp 1) paralleled the transcriptional activity of the L1 H2B gene. Additional mutagenesis and microinjection experiments located the activator element to a 32-bp DNA segment that includes the H2B abp 1-binding site. These experiments also showed that the 32-bp fragment functions independently of position and orientation and therefore has the hallmarks of an enhancer. That this fragment contains most or all of the L1 H2B gene transcription-stimulatory activity makes it unusual among enhancerlike elements, which generally consist of several clustered factor-binding sites that act additively or cooperatively to affect transcription. The nucleotide sequence of the L1 H2B enhancer element suggests that the trans-acting factor that interacts with it is a member of the antennapedia or engrailed class of homeodomain proteins.

Activation of a late H2B histone gene in blastula-stage sea urchin embryos by an unusual enhancer element located 3' of the gene.

In the sea urchin embryo, late histone genes are transcribed at low levels during cleavage and blastula formation and at substantially higher levels in later stages of embryogenesis. To investigate the molecular basis of the stage-specific expression of a late H2B histone gene, we injected mutant genes lacking portions of 5'- and 3'-flanking regions into Lytechinus pictus embryos and monitored their expression by RNase protection. A 200-bp region located 489 bp downstream of the mRNA 3' terminus was necessary for the increase in transcription of the late H2B gene at the mid-blastula stage of development. DNase I and methylation interference footprint analyses located only one factor-binding site in this region, and gel mobility shift experiments showed that the DNA-binding activity of this factor (designated H2B abp 1) paralleled the transcriptional activity of the L1 H2B gene. Additional mutagenesis and microinjection experiments located the activator element to a 32-bp DNA segment that includes the H2B abp 1-binding site. These experiments also showed that the 32-bp fragment functions independently of position and orientation and therefore has the hallmarks of an enhancer. That this fragment contains most or all of the L1 H2B gene transcription-stimulatory activity makes it unusual among enhancerlike elements, which generally consist of several clustered factor-binding sites that act additively or cooperatively to affect transcription. The nucleotide sequence of the L1 H2B enhancer element suggests that the trans-acting factor that interacts with it is a member of the antennapedia or engrailed class of homeodomain proteins.

Synthesis and turnover of late H2b histone mRNA in developing embryos of the sea urchin, Strongylocentrotus purpuratus

In sea urchins, “early” histone proteins are synthesized during cleavage and blastula formation, “late” histone proteins in subsequent stages of development. To understand the molecular mechanisms responsible for this ontogenic switch in histone subtype synthesis, we determined the absolute amounts, rates of synthesis, and rates of turnover of late H2b histone mRNAs during development. We showed previously that late H2b mRNA comprises several mRNA isotypes. In this study, we used both a class-specific DNA probe to measure the amounts of the late H2b mRNA isotypes collectively, and a gene-specific probe to measure amounts of a particular late H2b mRNA encoded by a gene known as L1. We found that the amount of late H2b mRNA increased dramatically from 85,000 molecules per embryo in the 16-hr blastula to a peak of 670,000 molecules per embryo in the 24-hr mesenchyme blastula, and fell to 380,000 molecules per embryo in the 72-hr pluteus larva. The L1 late H2b mRNA achieved its maximum abundance earlier than the late H2b mRNA class as a whole, reaching a peak of 34% of total late H2b in the 14-hr blastula and declining to 7% in the pluteus larva. Measurements of the rate of incorporation of [ 3H]uridine into late class H2b mRNA, performed by a novel in vivo isotope incorporation method, enabled us to calculate both synthesis rates and half-lives of late H2b mRNA during development. These calculations showed (1) that the increase in late H2b mRNA level between 16 and 24 hr postfertilization is regulated primarily if not entirely at the level of mRNA synthesis; and (2) that the half-life of late H2b mRNA is comparatively short, around 20 min, at all stages examined.

Developmental regulation of micro-injected histone genes in sea urchin embryos

The developmental behavior of cloned histone genes of Psammechinus miliaris was studied by injection into eggs of two related sea urchin species followed by fertilization. All five early histone genes were faithfully expressed in early blastula embryos as shown by SP6 RNA mapping. A 5–10 times lower expression rate was estimated for the injected early H2A gene from its competition strength with the endogenous gene. Transcripts of this early H2A gene accumulated during the cleavage stages and decayed in late embryos in parallel with the endogenous early H2A mRNA. However, an introduced late H2B gene was incorrectly regulated, since its mRNA level did not increase from the blastula to the gastrula stage. The sperm H2B-1 gene, normally inactive in development, was 80 times less well expressed than the early H2A gene in transformed blastulae. A fusion gene with the early H2A promoter linked to the structural sperm H2B gene was, however, efficiently transcribed suggesting that all essential information for an early expression pattern is contained within the 5′ region of the early H2A gene.

Differential stimulation of sea urchin early and late H2B histone gene expression by a gastrula nuclear extract after injection into Xenopus laevis oocytes.

Sea urchin early histone genes are active in preblastula embryos; late histone genes are maximally expressed during subsequent stages of embryogenesis. We used the Xenopus laevis oocyte to assay for trans-acting factors involved in this differential regulation. Sea urchin nuclear proteins were prepared by extracting gastrula-stage chromatin successively with 0.45, 1, and 2 M NaCl. We injected three fractions into oocytes along with plasmids bearing sea urchin early and late H2b histone genes. While neither the 0 to 0.45 M nor the 1 to 2 M salt fraction affected H2b gene expression, the 0.45 to 1 M salt fraction stimulated early and late H2b mRNA levels significantly. Late H2b gene expression was stimulated preferentially when the early and late genes were coinjected into the same oocytes. This extract did not stimulate the accumulation of transcripts of injected herpesvirus thymidine kinase genes or of the sea urchin Spec 1 gene, suggesting that the stimulatory activity is not a general transcription factor. We localized the DNA sequence required for the stimulatory effect to a region of the late H2b gene located between -43 and +62 relative to the transcription start site. A component of the 0.45 to 1 M salt wash fraction specifically bound to the 105-base-pair late gene DNA sequence and to the corresponding early gene fragment. The abundance of this binding activity decreased on a per genome basis during early development of the sea urchin.

Differential stimulation of sea urchin early and late H2B histone gene expression by a gastrula nuclear extract after injection into Xenopus laevis oocytes.

Sea urchin early histone genes are active in preblastula embryos; late histone genes are maximally expressed during subsequent stages of embryogenesis. We used the Xenopus laevis oocyte to assay for trans-acting factors involved in this differential regulation. Sea urchin nuclear proteins were prepared by extracting gastrula-stage chromatin successively with 0.45, 1, and 2 M NaCl. We injected three fractions into oocytes along with plasmids bearing sea urchin early and late H2b histone genes. While neither the 0 to 0.45 M nor the 1 to 2 M salt fraction affected H2b gene expression, the 0.45 to 1 M salt fraction stimulated early and late H2b mRNA levels significantly. Late H2b gene expression was stimulated preferentially when the early and late genes were coinjected into the same oocytes. This extract did not stimulate the accumulation of transcripts of injected herpesvirus thymidine kinase genes or of the sea urchin Spec 1 gene, suggesting that the stimulatory activity is not a general transcription factor. We localized the DNA sequence required for the stimulatory effect to a region of the late H2b gene located between -43 and +62 relative to the transcription start site. A component of the 0.45 to 1 M salt wash fraction specifically bound to the 105-base-pair late gene DNA sequence and to the corresponding early gene fragment. The abundance of this binding activity decreased on a per genome basis during early development of the sea urchin.

Closely linked early and late histone H2B genes are differentially expressed after microinjection into sea urchin zygotes.

An early and a late histone H2B gene from the sea urchin Stronglyocentrotus purpuratus were linked in a single plasmid and injected into the eggs of the sea urchin Lytechinus pictus. The levels of transcripts of injected early and late genes and of endogenous early genes were monitored during development by a ribonuclease protection assay. Transcripts of both the injected and endogenous early genes peaked during the blastula stage and decreased severalfold by the mesenchyme blastula stage. Transcripts of the injected late gene became detectable at the blastula stage and increased in amount subsequently, until at least the early gastrula stage, 28 hr after fertilization. Thus, the pattern of expression of the injected early and late H2B genes is similar to that of their endogenous counterparts. These results show that DNA sequences regulating the temporal pattern of early and late H2B gene expression must lie within the cloned DNA segments; i.e., within 600 base pairs of the early H2B gene and 3 kilobases of the late H2B gene.

Differential expression of early and late embryonic histone genes in adult tissues of the sea urchin Strongylocentrotus purpuratus

The sea urchin synthesizes distinct classes of histone mRNAs at different stages of development. “Early” embryonic histone mRNAs are synthesized in large amount in cleavage and blastula stage embryos. “Late” embryonic histone mRNAs are the predominant forms in postblastula embryos. To learn more about how early and late histone genes are regulated during the life cycle of the sea urchin and to search for additional classes of developmentally regulated histone mRNAs, we examined histone mRNAs in sea urchin adult tissues. Using methods of primer extension and S1 nuclease protection, we found that tube foot, intestine, testis, and ovary contain a subset of the several H2b mRNA species synthesized by the embryo. We detected early H2b mRNA in ovary, but not in other tissues. Three late H2b mRNA species were present in all tissues tested, while a fourth late H2b was not detected. Using a probe that hybridized specifically with transcripts of a single-copy late H2b gene, we found that this gene was transcribed in both embryos and adults. Interestingly, its level of expression relative to other late H2b genes varied among tissues. Finally, we identified two H2b mRNA species that were distinct from early and late embryonic forms and were synthesized only in adult tissues.

Evolution of late H2A, H2B, and H4 histone genes of the sea urchin, Strongylocentrotus purpuratus.

Sea urchins possess several distinct sets of histone genes, including "early" genes, maximally active in cleavage and blastula stages, and "late" genes, active from the late blastula stage onwards. We determined the nucleotide sequences of six sea urchin (Strongylocentrotus purpuratus) late histone genes located on four genomic segments. Comparative analysis of these sequences identified several conserved elements in 5' flanking regions, including the sequences ATGPyATANTATA shared by all late genes and GGCGGGAAATTGAAAA shared by two late H4s. Comparisons of protein-coding sequences of late H4 and H2B genes with their early counterparts showed that silent sites have diverged to the theoretical maximum, indicating that early and late histone gene classes diverged at least 200 million years ago. Since extant echinoderms evolved from a common ancestor at about that time, it is likely that early and late histone gene sets are characteristic of all echinoderm groups. Amino acid sequences derived from nucleotide sequences of late H2A and H2B gistone genes differ substantially from amino acid sequences of their late counterparts. Most such differences are in highly mutable positions. A few, however, occur in positions that do not mutate frequently and thus may reflect functional differences between the early and late forms of the H2A and H2B proteins.

Isolation, characterization, and expression of the gene encoding the late histone subtype H1-gamma of the sea urchin Strongylocentrotus purpuratus.

We cloned and characterized the gene encoding H1-gamma, a late histone subtype of the sea urchin species Strongylocentrotus purpuratus. The predicted primary sequence of H1-gamma is 216 amino acids in length and has a net charge of +70, which is high for a somatic H1 histone. The H1-gamma gene appears to be a unique sequence gene that is not tightly linked to the core histone genes. The 770-base-pair transcribed region of the H1-gamma gene is bordered on the 5' side by two previously described H1-specific sequence elements and on the 3' side by a hairpin loop structure and CAGA box sequences. We detected 3,900 stored maternal H1-gamma mRNA transcripts per egg. The number of H1-gamma transcripts per embryo rises by 9.5 h postfertilization, but the maximum rate of accumulation (4,300 molecules per min per embryo) occurs in the late-blastula-stage embryo between 14 and 21 h after fertilization. The number of H1-gamma mRNA molecules peaks 21 h after fertilization when there are 2.0 X 10(6) molecules per embryo (a 500-fold increase) and then decreases over the next 3.25 h to 1.3 million molecules per embryo. Between 24 and 82 h after fertilization the number of H1-gamma transcripts declines steadily (210 molecules per min per embryo) to reach approximately 5.4 X 10(5) H1-gamma mRNAs by 82 h postfertilization. Surprisingly, the number of late H1 mRNA molecules per embryo is greater than the number of late H2B mRNA molecules beginning at the early gastrula stage of development.

Isolation, characterization, and expression of the gene encoding the late histone subtype H1-gamma of the sea urchin Strongylocentrotus purpuratus.

We cloned and characterized the gene encoding H1-gamma, a late histone subtype of the sea urchin species Strongylocentrotus purpuratus. The predicted primary sequence of H1-gamma is 216 amino acids in length and has a net charge of +70, which is high for a somatic H1 histone. The H1-gamma gene appears to be a unique sequence gene that is not tightly linked to the core histone genes. The 770-base-pair transcribed region of the H1-gamma gene is bordered on the 5' side by two previously described H1-specific sequence elements and on the 3' side by a hairpin loop structure and CAGA box sequences. We detected 3,900 stored maternal H1-gamma mRNA transcripts per egg. The number of H1-gamma transcripts per embryo rises by 9.5 h postfertilization, but the maximum rate of accumulation (4,300 molecules per min per embryo) occurs in the late-blastula-stage embryo between 14 and 21 h after fertilization. The number of H1-gamma mRNA molecules peaks 21 h after fertilization when there are 2.0 X 10(6) molecules per embryo (a 500-fold increase) and then decreases over the next 3.25 h to 1.3 million molecules per embryo. Between 24 and 82 h after fertilization the number of H1-gamma transcripts declines steadily (210 molecules per min per embryo) to reach approximately 5.4 X 10(5) H1-gamma mRNAs by 82 h postfertilization. Surprisingly, the number of late H1 mRNA molecules per embryo is greater than the number of late H2B mRNA molecules beginning at the early gastrula stage of development.

Characterization of two nonallelic pairs of late histone H2A and H2B genes of the sea urchin: differential regulation in the embryo and tissue-specific expression in the adult.

Two nonallelic pairs of late H2A and H2B genes of the sea urchin Psammechinus miliaris were isolated on two different cosmid clones. The genes of cosmid PmL1 are separated by 11 kilobases of DNA and code for the late H2A-2 and H2B-2 variants. The genes of clone PmL2 are divergently transcribed with 1,060 base pairs of intergenic spacer DNA and code for novel variants of the H2A-2 and H2B-2 type. A comparison of the promoter sequences revealed little homology upstream of the TATA box with the exception of a 24-base-pair-long conserved sequence which is present at the same position in both late H2B promoters and part of which is identical with the "H2B-specific" 5' element. The mRNAs of the H2A and H2B genes of cosmid PmL1 reach their maximal levels early in the mesenchyme blastula embryo, whereas the transcripts of both genes of clone PmL2 accumulate maximally only later in the pluteus larva. In the adult sea urchin all four mRNAs are present in the tube foot but not in the intestine and lantern muscle. This pattern of differential expression in the embryo and tissue-specific expression in the adult suggests cell lineage-specific regulation of the late H2A-2 and H2B-2 genes. Another class of late histone genes represented by the H2A-3 and H2B-1 genes was shown to be expressed in all three adult tissues tested, whereas transcripts of the late H2A-1 genes could not be detected, suggesting that these genes are active exclusively during sea urchin development.

Characterization of two nonallelic pairs of late histone H2A and H2B genes of the sea urchin: differential regulation in the embryo and tissue-specific expression in the adult.

Two nonallelic pairs of late H2A and H2B genes of the sea urchin Psammechinus miliaris were isolated on two different cosmid clones. The genes of cosmid PmL1 are separated by 11 kilobases of DNA and code for the late H2A-2 and H2B-2 variants. The genes of clone PmL2 are divergently transcribed with 1,060 base pairs of intergenic spacer DNA and code for novel variants of the H2A-2 and H2B-2 type. A comparison of the promoter sequences revealed little homology upstream of the TATA box with the exception of a 24-base-pair-long conserved sequence which is present at the same position in both late H2B promoters and part of which is identical with the "H2B-specific" 5' element. The mRNAs of the H2A and H2B genes of cosmid PmL1 reach their maximal levels early in the mesenchyme blastula embryo, whereas the transcripts of both genes of clone PmL2 accumulate maximally only later in the pluteus larva. In the adult sea urchin all four mRNAs are present in the tube foot but not in the intestine and lantern muscle. This pattern of differential expression in the embryo and tissue-specific expression in the adult suggests cell lineage-specific regulation of the late H2A-2 and H2B-2 genes. Another class of late histone genes represented by the H2A-3 and H2B-1 genes was shown to be expressed in all three adult tissues tested, whereas transcripts of the late H2A-1 genes could not be detected, suggesting that these genes are active exclusively during sea urchin development.

Cell-type-specific transcription of an immunoglobulin kappa light chain gene in vitro.

We have established a cell-free system, derived from a human B-cell lymphoma, in which immunoglobulin kappa light chain gene promoters are both accurately transcribed and regulated in a cell-type-specific manner. Thus, accurate transcription from the T1 kappa light chain gene promoter was much more efficient in B-cell extracts than in HeLa cell extracts, whereas control promoters (adenovirus major late and histone H2B) were transcribed equally well in either extract. More important, the increased kappa light chain gene transcription in B-cell extracts was dependent upon upstream sequences (containing the conserved decanucleotide element) previously shown to be necessary for B-cell-specific transcription in vivo; in contrast, removal of these sequences had no effect on the low level of kappa transcription in HeLa extracts. The maximal level of upstream sequence-mediated transcription was dependent upon template topology. These studies show that there is at least one B-cell-specific factor that stimulates transcription from purified DNA templates, and they further suggest that the in vivo action of the factor(s) on other components of the transcription machinery is direct rather than indirect (e.g., via the maintenance of an open chromatin structure). The cell-free system described here should facilitate both purification and functional studies of the B-cell-specific factor(s).