dBigH1, a second histone H1 in Drosophila, and the consequences for histone fold nomenclature

Abstract

Recently, Pérez-Montero and colleagues (Developmental cell, 26: 578–590, 2013) described the occurrence of a new histone H1 variant (dBigH1) in Drosophila. The presence of unusual acidic amino acid patches at the N-terminal end of dBigH1 is in contrast to the arginine patches that exist at the N- and C-terminal domains of other histone H1-related proteins found in the sperm of some organisms. This departure from the strictly lysine-rich composition of the somatic histone H1 raises a question about the true definition of its protein members. Their minimal essential requirements appear to be the presence of a lysine- and alanine–rich, intrinsically disordered C-terminal domain, with a highly helicogenic potential upon binding to the linker DNA regions of chromatin. In metazoans, specific targeting of these regions is further achieved by a linker histone fold domain (LHFD), distinctively different from the characteristic core histone fold domain (CHFD) of the nucleosome core histones.

Introduction

Histone H1 is the historical name given to the histones that bind to linker DNA regions that connect adjacent nucleosomes in the chromatin fiber (Fig. 1). Hence, they are often referred to as linker histones. Structurally, members of the Histone H1 family exhibit a highly characteristic tripartite organization,¹ in which a central winged helix domain (WHD)^1,2 is flanked by two unstructured N- and C-terminal intrinsically disordered regions³ (Fig. 1) that can acquire secondary organization—mainly α-helical conformation—upon binding to DNA. This is in contrast to the core histones (H2A, H2B, H3, and H4), the histones responsible for the organization of the nucleosome, which consist of a dimerizing histone fold domain (HFD)⁴ preceded by a unstructured N-terminal (20–60 amino acids) region. In the case of H2A and H2B, a structurally similar region (20–30 amino acids) is also present at the C-terminal end of the molecule. The HFD allows for the octameric assembly of the nucleosome’s protein core,⁵ which is in turn wrapped by approximately 146 DNA nucleotides to form the nucleosome core particle (Fig. 1). Binding of histone H1 at the entry and exit sites of the linker DNA into the nucleosome⁶ organizes an extra 20 nucleotides of DNA, resulting in a structurally defined particle known as the chromatosome (see Fig. 1).⁷

Figure 1. Structural organization of the chromatin fiber and its fundamental subunit, the nucleosome. Histones are shown in different colors—Histone H1 (red), histone H2A (light yellow), histone H2B (light pink), histone H3 (light blue), histone H4 (light green). The linker histone fold domain (LHFD), the winged helix domain (WHD), and the core histone domain (CHFD) are also shown. The light red helices in the LHFD are meant to indicate that these regions are intrinsically disordered in solution, and only upon interaction with DNA acquire their helical conformation as schematically shown in the image of the nucleosome. N = N-terminus; C = carboxyl terminus. The red question mark indicates that the detailed site of interaction of the N-terminal domain of H1 with the nucleosome is not well determined yet. S1 and S2 indicate the binding sites of WHD to the DNA in the nucleosome.³⁴

The somatic version of histone H1 exhibits a highly typical lysine-enriched composition, mainly at its C-terminal domain.⁸ Histone H1 proteins vary in number between different organisms (e.g., one in the fruit fly Drosophila melanogaster and 7 in humans).^9,10 Interestingly, D. melanogaster has approximately 100 genes of this unique protein. In contrast, the somatic mammalian counterparts have very small primary-structure differences between them and, such microheterogeneity⁸ is the result of their expression by seven single copy genes within the genome. In addition to the somatic versions, a set of developmentally regulated histone H1 variants that participate in gametogenesis and in the early stages of zygotic development also exist, and are expressed in a replication-independent way.¹⁰ The Pérez-Montero et al. publication¹⁰ adds a previously unknown H1 variant to the fold—dBigH1, which is involved in embryonic development in Drosophila.

Members of the histone H1 family play a crucial role in the folding of the chromatin fiber, and play an important role in the modulation of gene expression—mediated by their binding to linker DNA regions of chromatin (Fig. 1). The functional complexity of highly specialized developmental variants being present both in the sperm and in the oocytes of invertebrate and vertebrate organisms (Fig. 2), and the overabundance of posttranslational modifications (PTMs),¹¹ is still very poorly understood.

Figure 2. Compositional heterogeneity of developmentally regulated H1 histones. Amino acid sequence of dBig H1 from D. melanogaster (fruit fly),¹⁴ PL-I from Spisula solidissima (surf clam),⁶⁸ B4 from Xenopus laevis (African clawed frog),⁶⁹ and PL-I from Mullus surmuletus (striped red mullet, fish).⁵⁴ The occurrence of the histones in the sperm or oocyte is indicated by the cartoon representation on the side. The boxes indicate the regions corresponding to the WHD. Arginine is shown in red, lysine in orange, and glutamic/aspartic acid is shown in green.

The discovery of a zygotic genome activation (ZGA) histone (dBigH1) in Drosophila, which is also abundant in the oocyte, is important—not only from a developmental perspective, but also because this is the second histone H1 that has been reported in this organism. The presence of abundant acidic amino acid (glutamic/aspartic) patches at the N-terminal end of dBigH1 is highly unusual for a histone—especially for a histone H1. It raises important new questions about true histone identity and our conventional perception of histone H1 family members.

Histone H1 and development

The functional relevance of somatic histone H1 microheterogeneity and its potential unique functional specificity has long been a reason for debate.¹² While each of the microheterogeneity variants appears to exhibit functional redundancy to some extent, it is now clear that their overall presence is essential to proper organism development.¹³

Developmentally regulated histone H1s (such as dBigH1,¹⁴ which is involved in gametogenesis and early embryogenesis) exhibit a significant compositional deviation from somatic histone H1 members, and they possess highly specialized functions.^12,15 Previously, only one histone H1 type was believed to be present in D. melanogaster, but Pérez-Montero et al.¹⁴ have uncovered the presence of dBigH1 (Figs. 2 and 3), which is much larger than the somatic counterpart—hence its name. dBigH1 is an essential embryonic histone H1 variant that plays an important role in the early development stages of the zygote, before cellularization. Following cellular division and the progressive activation of the zygotic genome, dBigH1 is replaced in Drosophila by dH1—arrest of dBigH1 prevents premature zygotic genome activation.¹⁴

Figure 3. Developmental structural diversity of invertebrate and vertebrate histone H1. (A) Schematic representation of the structural organization of the histone H1s shown in Figure 2. B/A indicates the ratio of basic (B) to acidic (A) amino acids present in the whole protein; the percentage of lysine (%K), alanine (%A), and proline (%P) within the C-terminal tail are also indicated for each of the histones; vL = very large. (B) Tertiary structure organization of the WHD of dBigH1 and somatic dH1 from D. melanogaster. The predicted structures were determined by Phyre 2 server (http://www.sbg.bio.ic.ac.uk/~phyre/) displaying 96% of amino acid coverage with 99.5% confidence (dBigH1) and 99% of amino acid coverage with 99.9% confidence (dH1).

Another extreme example of the developmental specificity of H1 family members is found in the large and small protamine-like proteins (PL-I) present in the sperm of invertebrate and vertebrate organisms (Fig. 2).¹⁵ In this instance, the lysine-rich composition of protamine-like proteins is accompanied by an abundance of arginine-rich clusters (such as those found in the protamines of mature sperm from a large variety of deuterostome and protostome organisms¹⁵) which, like PL, displace and replace most of the somatic histones during spermiogenesis.

The unusual amino acid composition, atypical structural organization, highly specialized function, and DNA binding observed in the above mentioned variants calls for a reconsideration of the minimal structural features that define a protein member of the histone H1 family.

Defining histone H1

The importance of the C-terminal domain

One of the first hints of the structural and functional relevance of the C-terminal tail of histone H1 comes from evolution. In contrast to core histones, in which tailless ancestral forms (present in archaebacteria) consisted exclusively of the histone fold, the histone H1 C-terminal tail (present in some protists) appears to have developed earlier than the WHD—it became fully established during the course of metazoan evolution.¹ Previous studies showed that some trypanosomes contain a microheterogeneous mixture (70–100 amino acids) of lysine-alanine-proline (KAP)-rich peptides^16,17 with a composition very similar to the C-terminal tail of somatic histones H1 found in vertebrates (see Fig. 3)—but with the absence of WHD.

The compositional abundance of lysine in histone H1, dating from its ancestral origins, is not surprising. The preference of lysine and arginine polypeptides for binding, respectively, to AT- and GC-rich DNA sequences has long been known.^18,19 Since the linker DNA regions of chromatin are AT-rich,^20,21 it is not then surprising that a lysine-rich protein was selected to bind to these chromatin domains.

Chromatin reconstitution experiments have shown that a histone H1 fragment, consisting of folded domain (WHD) and the C-terminal tail, was able to fold the chromatin fiber in the same way as the intact histone. As this ability was not observed when only globular WHD was used,²² this experiment emphasizes the importance of the C-terminal tail of linker histones in the involvement of chromatin fiber folding through charge neutralization.^23,24 Interestingly, this region appears to bestow the binding of somatic microheterogeneous forms with their differential binding affinity and their specificity^25,26 for different genomic loci.^27,28

Further evidence for the importance of the C-terminal tail of H1 histones comes from a series of experiments in fluorescence recovery after photobleaching (FRAP), a method that allows researchers to carry out in vitro experiments within the cell setting. These experiments showed that the C-terminal tail is the primary determinant of histone H1 binding to chromatin,²⁹ and that the majority of initial contact with a nucleosome occurs through this region, followed by the subsequent binding of the WHD.³⁰

The winged helix domain

It appears from previous evidence that, despite its well defined tertiary structure, the WHD plays a secondary role in chromatin binding. The tertiary structure of this central globular domain of linker histones was first determined in chicken histones H1 by NMR studies, and then in chicken histones H1³¹ and H5^2,32 using X-ray crystallography. These studies showed a remarkable similarity of globular domain to the winged helix domain of the transcription factor HNF3 (hepatocyte nuclear factor-3).³³

The WHD has two DNA binding sites³⁴ (see Fig. 1) that allow it to bind the cruciform-like DNA structure arising from the linker regions entering and exiting the nucleosome. Unlike the C-terminal tail, the winged histone globular domain of histone H1 was unable to fold chromatin.²²

From all of the above, it is clear that the functional and structural requirements of the linker histones in metazoans for the proper folding of chromatin are the WHD and the C-terminal tail. The presence of both of them is essential for the chromatosome arrangement^7,35,36 (see Fig. 1), with its characteristic 3–5 nm stem^37-39 (see Fig. 1).

The true histone nature of linker histones

The extensive compositional and structural variability of the members of the histone H1 (Figs. 2 and 3) family in addition to the presence of a WHD (instead of a histone fold) in their structural organization calls for an updated definition of histone H1. It furthermore raises the question of whether or not the name “histone,” the definition of which also includes the nucleosomal core histones, is an appropriate encompassing name for all these proteins.

The name “histone” was coined in 1884 by Albrecht Kossel to reflect the tissue (from the Greek “histos,” tissue) and peptone (pepton) origins of proteins extracted from the nuclei of different cells; histones were, in fact, the most abundant proteins extracted from purified nuclei by using dilute acids (i.e., 0.5 N hydrochloric/0.4 N sulfuric acid).⁴⁰ The name includes histone H1 members, which exhibit a unique solubility in 5% perchloric acid,⁴¹ likely as a result of their high lysine composition (≥ 20%, see Fig. 3A).

Mammals contain a single copy of somatic histone H1 genes (7 in humans) in comparison to the multiple copies of core histone genes present (16 H2A, 22 H2B, 14 H3, and 14 H4 in humans).⁴² Core histone genes exist in clusters, and are dispersed throughout several chromosomes; somatic histone H1 genes are interspersed between the core histone gene clusters, suggesting a genomic linkage. In Drosophila, each repeating histone gene cluster contains one copy each of H2A, H2B, H3, H4, and H1. This arrangement suggests also an integrated relationship between the core histones and H1. The existence of a single copy of the linker histones is not surprising, assuming that there is approximately one histone H1 per nucleosome, and two copies for each one of the core histones. Therefore, for every one H1 gene (approx. 7 in mammals, assuming equal or compensatory expression), two of each core histone gene (approx. 14) would be required to retain a similar expression during DNA replication. Interestingly, histone H1 is the last to be deposited into chromatin during S phase,⁴³ and the ratio of H1 to nucleosome is, in many tissues, <1.⁴⁴ Hence, it would be plausible that these protein genes, required to chromatinize the newly synthesized DNA, are present in near stoichiometric amounts.

It can therefore be concluded that, in spite of the highly diverse composition of H1 histones when compared with core histones, and of the different tertiary organization of their globular domains, these two families of chromosomal proteins are genomically, biochemically, and structurally related to each other.

The CHFD and LHFD histone folds

A careful observation of the primary structure of all the histone H1 variants shown in Figures 2 and 3 reveals that (1) the C-terminal tail is more size-constrained than its N-terminal counterpart (Fig. 3A), (2) the WHD of dBigH1 is structurally very similar to the somatic histone H1 (Fig. 3B), and (3) there is a consistent presence of conserved lysine, alanine and proline (KAP) composition (Figs. 2 and 3A) in the C-terminal tail.

The limited variation in the length of the C-terminal tail may simply be a reflection of the constraints imposed by the relatively small variation in the length of linker DNA between different tissues⁴⁴ and in the way this region binds to the nucleosome (Fig. 1).

The presence of the WHD in histone H1 members is in contrast to the rather different histone fold (HF)⁴ of core histones (Fig. 1); the nomenclature used here appears to be at odds with the true histone identity of linker histones—as discussed above. The rich, acidic amino acid composition of some egg-specific H1 histones, including dBigH1 (Figs. 2 and 3), add controversy to this situation, as does the relation to highly arginine-rich protamines (i.e., the protamine-like nature of some sperm-specific H1s [PL-Is]).¹⁵

It has long been known that the C-terminal tail of somatic histone H1 includes a rich composition of highly helicogenic amino acids, lysine and alanine. Based on this composition, and on the sequence distribution of alanine and lysine, the existence of a putative helical conformation had been proposed.^23,45 While in aqueous solution, this region has a disordered conformation⁴⁵; recent studies have documented that, upon interaction with DNA, this region acquires a helical conformation^46-48 (see ref. 49 for a review) (see also Fig. 1). The KAP composition of the C-terminal tail of all proteins shown in Figure 3A, including dBigH1, is very similar to the dH1somatic version (34% K, 23% A, and 5% P) and to the bovine somatic histone H1 mixture (average 38 ± 2% K, 24 ± 6% A, and 12 ± 2% P taken as a mammalian representative). Indeed, many of the regions within this domain (in the sequences shown in Fig. 2) exhibit alternating occurrences of lysine and alanine, and show the charge distribution that would be expected for the generation of helical regions.²³

When all of this is put together within an evolutionary frame, in the context of DNA binding within the nucleosome, a clearer picture starts to emerge: In core histones, the HF⁵⁰ appeared first in archaebacteria (kingdom euryarchae)^51-53 and, in the course of evolution, preceded the acquisition of histone N- and C-terminal. Conversely, in the H1 family, a KAP rich C-terminal domain appeared first, while the WHD was acquired later on, for further binding refinement. This addition of WHD likely enhanced linker histones’ ability to perform specific targeting of the intertwining organization found in linker DNA at the entry and exit of the nucleosome (Fig. 1), or of the linker-like DNA intertwining that is present in the sperm of some species (like in the fish Mullus surmuletus⁵⁴) after the displacement of histones during spermiogenesis. Such DNA crossover arrangements are structurally similar to the four-way junctions preferably bound by histone H1 in cruciform DNA.⁵⁵ Thus, throughout the highly variable forms of H1 family histones, both the KAP C-terminal tail and the WHD define an ancestral DNA binding domain in metazoans, which could be referred to as the linker histone fold domain (LHFD) (Fig. 1).

While the N-terminal tail modifications in both linker and core histones are important for chromatin metabolic functions, they are not in most instances required for proper chromatin binding and organization. Nucleosome particles—which are almost structurally identical to native nucleosome core particles—can be reconstituted from tailless core histones,⁵⁶ and the folding of the chromatin fiber is not restricted to the N-terminal tail of H1.²² Thus, from a nomenclature perspective, the histone fold might be better referred to as the core histone fold domain (CHFD), in contrast to the LFHD (Fig. 1).

Despite the different evolutionary origin of the linker and the core families,^1,15 their genetic linkage is strong, and the difference in their structural folds are merely a reflection of their complementary, yet well-defined, distinctive chromatin functions. Hence, the nomenclature issue could be easily resolved if HF were referred to as core histone fold domain CHFD, and the WHD together with the C-terminal tail were referred to as the linker histone fold domain LHFD; that system of nomenclature would not detract from their shared (histone) structural and functional attributes.

Conclusions and Outlook

The existence of extremely compositionally divergent histones H1, such as the recently discovered dBigH1 (despite its unusually large size¹⁴), provide a unique opportunity to create a better structural definition of the histone H1 chromosomal proteins. Here, we identify the compositionally lysine and alanine-rich helicogenic C-terminal tail and the WHD as integral parts of a unique LHFD (Fig. 1) that confers members of the histone H1 family with their ability to specifically bind to the linker regions of chromatin. This domain functionally distinguishes H1 histones from core histones. Hence, we propose the use of the term CHFD (core histone fold domain) rather than the generic “histone fold” to refer to the dimerizing domain that allows histones H2A-H2B-H3-H4 to come together to form the nucleosome protein core (Fig. 1). Use of the generic term “histone fold” for core histones detracts from the true histone nature of H1 histones that, albeit with a different function, are equally important structural and functional chromosomal protein partners of the chromatin fiber.

In addition to all of this, it is important to recognize that linker histones of the H1 family, like their core histone counterparts, contain an additional disordered N-terminal domain, the roles of which require further attention. It is likely that the acidic patches present within this region of dBigH1 are important for its developmental functions, and are critical to its survival.¹⁴ Yet, the molecular mechanism involved remains completely unknown. Also, like the core histone counterparts, linker histones are subject to a plethora of PTMs^57-59 (see ref. 11 for a review) that may potentially regulate their function and interaction with other functionally relevant protein complexes of chromatin metabolism. Nevertheless, in contrast to the core histone PTMs,⁶⁰ the functional and structural role of the linker histone PTMs (with the exception of phosphorylation^61,62) and their potential epigenetic implications remain, in many instances, largely unknown or unexplored. It is quite possible that access to the epigenetic marks introduced by some core histone PTMs, such as those of the N-terminal domain of histone H3, are highly modulated by the presence of histone H1. Histone H3 binds to the nucleosome in close proximity to H1, and both are involved in the maintenance of the chromatin organization.⁶³ For instance, steric hindrance by the tails of histone H1 has been shown to inhibit acetylation of histone H3 via the histone acetyl transferasep300/CBP-associated PCAF factor.⁶⁴ However, to explore this area of research, more experimental work is needed. This leads to an investigation into the unknown detailed molecular mechanisms by which histone H1 regulates gene expression (besides its obvious involvement in chromatin folding). An example of this is provided by the paper by Pérez-Montero et al.,¹⁴ which serves as a focus of this review. Questions to be considered are: What is the role of the N-terminal acidic patches in such regulation? Similarly, what is the role of the glutamic-rich C-terminal domains of other oocyte-specific H1s, such as the Xenopus B4 (see Figs. 2 and 3)?

The answers to the above questions are challenging. H1 is a highly mobile^65-67 and dynamic histone, and genome-wide analysis of its distribution requires particular crosslinking conditions. Collecting information for such questions, however, is important because it is very likely that deregulation of histone H1 PTMs and/or histone H1 itself plays an important role in illnesses that involve chromosome rearrangements and loss of genomic integrity, as occurs in many cancers.

14S Pérez-Montero, A Carbonell, T Morán, A Vaquero, F Azorín.The embryonic linker histone H1 variant of Drosophila, dBigH1, regulates zygotic genome activationDev Cell2013265789010.1016/j.devcel.2013.08.011

10.4161/epi.28427