Histone acetylation: lessons from the plant kingdom

Histone acetylation: lessons from the plant kingdom

Review TRENDS in Plant Science Vol.6 No.2 February 2001 59 Histone acetylation: lessons from the plant kingdom Alexandra Lusser, Doris Kölle and Pe...

181KB Sizes 3 Downloads 25 Views

Review

TRENDS in Plant Science Vol.6 No.2 February 2001

59

Histone acetylation: lessons from the plant kingdom Alexandra Lusser, Doris Kölle and Peter Loidl Post-translational acetylation of core histones is an enigmatic process. The identification of histone acetyltransferases and deacetylases as co-regulators of transcription in yeast and vertebrates has advanced our understanding of the biological role of histone acetylation and also improved our general insight into the molecular network of gene regulation. Basic features of histone acetylation in plants resemble those of other eukaryotes but there are differences, which are reflected in novel classes of histone deacetylase. Investigating histone acetylation in higher plants might reveal regulatory pathways distinct from animals, yet of essential importance for gene expression in plants.

Chromatin structure, the mode in which DNA is organized around basic nuclear proteins, the histones, is essential for nuclear processes such as DNA replication, transcription, DNA repair and recombination. Chromatin structure changes in a dynamic way and is continuously remodelled. The basic unit of eukaryotic chromatin is the nucleosome core particle, a repeating element consisting of a histone octamer with 146 bp of DNA wrapped around it. Histones H3, H4, H2A and H2B form the core octamer by protein–protein interactions of their globular domains. Crystallographic analysis has revealed that the N-terminal tails of histones protrude from the octamer1. These highly conserved, flexible N-terminal extensions contain amino acids that are subject to post-translational modifications. Conserved lysine residues therein have attracted considerable interest because they are acetylated. A nucleosome contains a total of 26 potentially acetylated lysine residues. Histone acetyltransferases (HATs) transfer the acetyl moiety of acetyl-CoA to the ε-amino group; this reaction is reversed by the action of histone deacetylases (HDACs). Regulatory functions of histone acetyltransferases and histone deacetylases

Alexandra Lusser* Doris Kölle Peter Loidl Dept Microbiology, University of Innsbruck, Medical School, FritzPregl-Strasse 3, A-6020 Innsbruck, Austria. *e-mail: [email protected] uibk.ac.at

After 30 years of research, a Tetrahymena HAT was shown to be a close homologue of the yeast transcriptional coactivator Gcn5. Soon after, a mammalian homologue of another well known regulatory yeast protein, Rpd3, was found to be an HDAC (Refs 2,3). Since then, many transcriptional regulators and co-regulators have been identified as HATs or HDACs (Tables 1,2). There are several alternative explanations for the effects of lysine acetylation on chromatin structure. First, acetylation might neutralize a positive charge and thus weaken the interaction of the histone octamer with the negatively charged DNA. This would destabilize the

nucleosomes, allowing transcriptional regulators to gain access to the DNA. Second, acetylation might interfere with the higher-order packing of chromatin and thus alter the accessibility of larger chromatin areas for regulatory proteins. Third, acetylation could act as a specific signal that alters histone–protein interactions4. This possibility is supported by the finding that non-histone proteins are also acetylated and deacetylated by HATs and HDACs. Among these proteins are HMG proteins, transcriptional activators, nuclear receptor coactivators, general transcription factors and importin α7 (Ref. 2). The fact that the N-terminal extensions of core histones contain sites for multiple modifications (acetylation, phosphorylation, ubiquitination, ADP-ribosylation and methylation) has raised the question of how these modifications could cross-talk to each other in the sense of a histoneencoded language5. A large body of evidence indicates that acetylation might have a major role in gene expression. Chromatin immunoprecipitation assays have shown that acetylation of H3 and H4 within promoter chromatin is associated with gene expression6,7. By contrast, the use of antibodies that recognize acetylated isoforms of all the core histones has shown that highly acetylated histones are not restricted to promoter regions of active genes8,9. This indicates that the acetylation of distinct histones might have different effects in different regions of chromatin. A variety of HATs have been discovered, many of them related to Gcn5 (Table 1). Gcn5 family members (including PCAF) and CBP/p300 acetylate histones and several transcription factors2,10. CBP/p300 is a central integrator of various signalling pathways; transcriptional regulators such as CREB, Jun/Fos and hormone receptors are associated with CBP/p300 (Ref. 11). Acetylation by CBP can either turn on7 or turn off12 the transcription of a gene, depending on which protein is modified. Acetylation of histones might generally have a positive effect on transcription, whereas acetylation of non-histone factors might have activating or repressing effects. Most HATs (e.g. Gcn5, PCAF, CBP/p300) have a bromodomain, which is a conserved sequence motif found in HATs and a variety of transcription-related proteins whose precise function is unclear. However, it is assumed that the bromodomain is involved in protein–protein recognition and interaction. Bromodomains recognize acetyl-lysines13, therefore

http://plants.trends.com 1360-1385/01/$ – see front matter © 2001 Elsevier Science Ltd. All rights reserved. PII: S1360-1385(00)01839-2

60

Review

TRENDS in Plant Science Vol.6 No.2 February 2001

they might be specific acetylation receptors2. HATs are part of multiprotein complexes (Table 1). Complexes containing Gcn5 are extremely heterogenous2, and the specificity of HATs differs

depending on their partner proteins. Some putative HATs might acetylate transcription factors in vivo, but not histones, and this could modulate the binding of factors to chromatin.

Table 1. Summary of putative histone acetyltransferases of various organismsa HAT

HAT complex

Gcn5-related N-acetyltransferases (GNAT) Gcn5 ADA, SAGA, STAGA Hat1 Complex with Rbap48

Function

Organism

Ubiquitous Ubiquitous Mammals Yeast Yeast Mammals

PCAF Hpa2 Elp3 ATF-2

PCAF complex – RNA polymerase II complex Complex with c-jun

Coactivator of transcription Cytoplasmic acetylation of H4 (deposition-related) Coactivator of transcription Unknown Transcription (elongation) Coactivator of transcription

CBP/p300

Associates with different regulatory proteins

Coactivator of transcription

Ubiquitous

Nuclear receptor coactivators ACTR – SRC-1 – TIF2 –

Coactivator of transcription Coactivator of transcription Coactivator of transcription

Mammals Mammals Mammals

TAFII250

TFIID

Factor associated with TBP

Ubiquitous

TFIIIC (90, 110, 220)

TFIIIC-complex

RNA polymerase III transcription

Human

MYST-family Sas3 Esa1 MOF MOZ Tip60 HBO1

NuA NuA MSL complex – Tip60 complex HBO-complex

Silencing Cell cycle regulation Gene dosage compensation Malignant diseases HIV-Tat interaction Interacts with replication origin recognition complex

Yeast Yeast Insects Human Human Human

aFor

details of the individual HAT-types see Ref. 2.

Abbreviations: HATs, histone acetyltransferases; –, not known.

Table 2. Summary of histone deacetylases of various organismsa HDAC-family (examples)

Enzymes

Organisms

Proteins associated directly or indirectly with HDAC-complexes

RPD3, HOS1-3 RPDA, HOSA dHDAC1-3 HDA1-3 HDm HDAC1-3 HDAC7, 8 RPD3/HD1-B

Yeast Aspergillus Drosophila Caenorhabditis elegans Xenopus laevis Chicken, mammals Chicken, mammals Maize

Sin3, Rbap, SAP, MAD, MAX, NCoR, SMRT, Mi2, MTA2, MBD3, MeCP1, MeCP2, Ikaros, UME6, Ski, p53, HPV E7, PcG, YY1, LIM, Hunchback, Groucho, LAZ3, PLZF, BRCA1, HDAC4, HDAC5

HDA1 dHDA2 mHDA1, 2 HDAC4-6

Yeast Drosophila Mouse Human

HDAC3, MEF2A, NCoR, SMRT

HD2

Plants

Homopolymer of HD2-p39 and phosphorylated forms

SIR2 SIR2-homolog

Yeast Mouse

Sir3, Sir4, Net 1

RPD3-like

HDA1-like

HD2-like

SIR2-like NAD-dependent

aFor

details of the individual HDAC-types see Ref. 3.

Abbreviation: HDACs, histone deacetylases.

http://plants.trends.com

Review

Fig. 1. Similarity of Gcn5-related N-acetyltransferase (GNAT) family members from plants and yeast. The most conserved domains of the proteins harboring the acetyl-CoA substrate recognition and binding sites were aligned. EST-encoded proteins from various plants can be grouped into three subfamilies related to Esa, Gcn5 or Hat1 based on their similarity to the respective yeast proteins. Amino acids identical in all sequences are boxed in red. Residues that show 65% identity among all sequences of a subfamily are boxed in yellow (Esa family), green (Gcn5 family) or blue (Hat1 family). A GenBank Accession number is indicated for each protein. Abbreviations: At, Arabidopsis thaliana; Ee, Euphorbia esula; Gm, Glycine max; Lj, Lotus japonicus; Os, Oryza sativa; y, Saccharomyces cerevisiae; Zm, Zea mays.

TRENDS in Plant Science Vol.6 No.2 February 2001

61

GmEsa: AW395340 OsEsa: AU058160 ZmEsa: T14677 LjEsa: AV415198 yEsa1p: NP014887

H H H -

AtHat1: CAB77581 EeGcn5: AW874991 yGcn5: NP011768

H S QK FG E I A FC A I T A D E Q V K G YG TR LM N H LK Q H A R D V D G L T - - - -R FG E I A FC A I T A D E Q V K G YG TR LM N H LK Q H A R D V D G L T H - - -E FA E I V FC A I S S T E Q V R G YG A H LM N H LK D YV R N TS N IK Y F

Gcn5

ZmHATB: AF171927 yHat1p: NP015324

P E SI R LR IS Q I LV LP P YQ G E G H G LG L LE A - -I N YI A Q S E N I - - - - F R A K IS Q F LIF P P YQ N K G H G S C L YE E A II Q S W L E D K S I TE

Hat1

S S S S -

E E E -

E E E E D

S S S S G

Y Y Y Y Y

N N N N N

LA LA LA LA VA

C C C C C

I I I I I

L L L L L

T T T T T

LP LP LP LP LP

P P P P Q

YQ YQ YQ YQ YQ

R R R R R

K K K K M

G G G G G

YG YG YG YG YG

K K K K K

F F F F L

LIA LIA LIA LIA LIE

FS FS FS FS FS

YE YE YE YE YE

LS LS LS LS LS

K K K K K

K K K K K

E E E E E

G G G G N

K K K K K

V V V V V

G G G G G

TP TP TP TP SP

Esa

TRENDS in Plant Science

Structural rearrangements of chromatin are necessary to allow functional changes. The identification of HATs and HDACs as modulators of transcription has largely focused the attention of molecular biologists on roles in the regulation of gene expression. However, there are many other processes that depend on changes in histone acetylation. It has recently been shown that V(D)J recombination in T and B cells is regulated by histone acetylation14. In mammalian cells, a HAT has been identified that associates with a replication-origin recognition complex2. This HAT, HBO1, is a member of the MYST family of acetyltransferases, which includes other putative HATs, some of which are implicated in human disease15. In general, HATs have attracted considerably more attention than HDACs. However, the identification of yeast Rpd3 and related proteins as transcriptional co-repressors and the fact that HDACs exert their function as multisubunit complexes (Table 2) have led to the conclusion that gene regulation is a complex interrelation of histone-acetylating and -deacetylating activities. Repression of transcription is often linked to recruitment of multisubunit complexes containing one or more HDACs. Moreover, some of the known HDACs are part of chromatin remodelling machines and evidence has accumulated that implicates HDACs in human cancer development3. In the past, a great deal of research on functional aspects of histone acetylation was based on HDAC inhibitors, whereas no such inhibitors are available for HATs. The inhibitors used are butyrate and a panel of highly specific and potent natural compounds of diverse chemical composition, such as trichostatin A and cyclic tetrapeptides (HC toxin, chlamydocin, trapoxin); cyclic tetrapeptides play a role in plant–microorganism interactions16.

of H4 lysines 5, 8 and 12, but not 16, in heterochromatic areas of the genome18. Intriguingly, no difference in H4 acetylation pattern was detected between the two X chromosomes of Silene latifolia homogametic female cells, even though one X chromosome is supposed to be inactivated as a consequence of dosage compensation19. In a recent report20, it was shown that the H3 and H4 acetylation patterns of field bean (Vicia faba) nuclei change significantly during the cell cycle and correlate better with replication than with transcriptional activity. H4 from Medicago, Arabidopsis, tobacco and carrot was detected in five acetylated isoforms (mono- to penta-acetylated at lysines 5, 8, 12, 16 and 20). In animals and yeast, Lys20 of H4 is not acetylated but is methylated instead21,22. Core histones and histone H1 undergo phosphorylation on specific serine/threonine residues. Phosphorylation of Ser10 of H3 has been observed during activation of c-Jun and c-Fos in response to stimulation by growth factors and plays an essential role in chromosome condensation during mitosis5. Recently, mitosis-associated phosphorylation of H3 has been shown in plants by indirect immunofluorescence studies of mitotically dividing cells, and the staining pattern in meiotic cells has been shown to differ significantly between plant and animal chromosomes23. Enzymes of histone acetylation in plants

The monocot maize (Zea mays) is a model organism for the detailed biochemical, enzymatic and molecular characterization of different HAT and HDAC types. In germinating seedlings of maize, at least three HAT and four distinct HDAC activities can be detected by chromatographic fractionation of cellular extracts4,24.

Post-translational acetylation of plant histones

Whereas H4 is the predominant target of acetylation in animals and fungi, H3 was found to be the most extensively acetylated histone in plants17. Immunofluorescence studies of mitotic chromosomes from several plants using antisera against specifically acetylated H4 isoforms confirmed the general staining pattern observed in animal cells: highly acetylated H4 subspecies were associated with transcriptionally active regions and underacetylation http://plants.trends.com

Histone acetyltransferase

Apart from a predominantly cytoplasmic HAT (HATB), which is homologous to yeast Hat1 (Ref. 25), two nuclear HATs (HATA1, HATA2) have been biochemically distinguished in maize, although they have not been further characterized4. An antibody against a maize Gcn5-related protein revealed the presence of ZmGcn5 in both HATA fractions (R. Thompson, pers. commun.), suggesting the

62

Review

ZmRpd3: P56521 AtRpd3: O22446 ZmHD1BII: AAD10139 McRpd3: BE036122 ZmRpd3-3: W099/51731 MtRpd3: AW980746 yRpd3p: P32561 yHda1p: P53973 GmHda1: AI940895 GmHda1-2: AW508008 GmHda1-3: AI988530 HvHda1: BE412922

TRENDS in Plant Science Vol.6 No.2 February 2001

SVGGAVKFNHG-HDIAINWSGGLHHAKKCEASGFCYVNDIVLAILELLKHHE----RVLYV DIDI H H G D G VEEAF YTTDRVMTVS SVGGSVKLNHGLCDIAINWAGGLHHAKKCEASGFCYVNDIVLAILELLKQHE----RVLYV DIDI H H G D G VEEAF YATDRVMTVS SIGAAVKLNRGDADITVNWAGGLHHAKKSEASGFCYVNDIVLAILELLKFHR----RVLYV DIDV H H G D G VEEAF FTTNRVMTVS SIGAAVKLNRRDADIAINWAGGLHHAKKSEASGFCYVNDIVLGILELLKVHR----RVLYI DIDV H H G D G VEEAF YTTDRVMTVS TLDAARRLNHKICDIAINWAGGLHHAKKCEASGFCYINDLVLGILELLKYHA----RVLYI DIDV H H G D G VEEAF YFTDRVMTVS TIDAARRLNNKLCDIAINWAGGLHHAKKCEASGFCYINDLVLGILELLKHHP----RVLYI DIDV H H G D G VEEAF YFTDKVMTVS SMEGAARLNRGKCDVAVNYAGGLHHAKKSEASGFCYLNDIVLGIIELLRYHP----RVLYIDIDV H H G D G VEEAF YTTDRVMTCS EACKAVVEGRVKNSLAVVRPPG-HHAEPQAAGGFCLFSNVAVAAKNILKNYPESVRRIMILDWDI H H G NGTQKSFYQDDQVLYVS DLASAIVSERAKNGFALVRPPG-HHAGVRHAMGFCLHNNAAVAA---LAAQAAGARKVLILDWDV H H G NGTQEIFEQNKSVLYIS SAMKHLLNGDGKVSYALVRPPG-HHAQPSLADGYCFLNNAGLAVQ---LALDSGCKKVAVIDIDV H Y G NGTAEGFYRSNKVLTIS SAMKHLLNXDGKVSYALVRPPG-HHAQPSLADGYCFLNNAGLAVQ---LALEFRCKKVAVIDIDV H Y G NG--------------SAMRHILDGHGKIAYALXRPPG-HHXQPDHADGYCFLNNAGLAVH---LALHSGRAXVS------ H W-----------------TRENDS in Plant Science

Fig. 2. Alignment of the catalytic domains of various plant and yeast histone deacetylases (HDACs). Three different types of Rpd3-like HDACs can be detected in Zea mays, each of which has at least one close homologue in other plant species. Multiple members within one plant species can also be detected for the Hda1 family. Amino acid residues that are crucial for catalytic activity and conserved in all sequences are boxed in red, whereas amino acids conserved in 65% of the sequences of either the Rpd3 or the Hda family are shown in blue and green, respectively. A GenBank Accession number is indicated for each protein. Abbreviations: At, Arabidopsis thaliana; Gm, Glycine max; Hv, Hordeum vulgare; Mc, Mesembryanthemum crystallinum; Mt, Medicago truncatula; Zm, Zea mays.

existence of complexes with distinct compositions reminiscent of the yeast SAGA and ADA complexes2. Searching available plant DNA databases revealed various homologues to the Gcn5-related Nacetyltransferase (GNAT) superfamily, the MYST family of HATs and CBP/p300-related HATs (Table 1, Fig. 1). Zea mays HATB is responsible for the acetylation of newly synthesized H4 at lysines 5 and 12 (Ref. 26) before chromatin assembly. H4 acetylation was suggested to be important for the transport of newly synthesized H4 into the nucleus and/or correct assembly into the nucleosome27. Consequently, HATB activity would be required to provide free H4 with a tag for its subsequent fate. However, deletion of the Hat1 gene from yeast revealed no apparent mutant phenotype and, recently, it was shown that histones need not to be acetylated to interact with chromatin assembly factor CAF-1 or to be deposited onto chromatin28. Because the H3 and H4 N-termini in yeast are functionally redundant, it was assumed that acetylation of H3 N-termini by another, yet unidentified, enzyme could complement the lack in H4 acetylation28. Histone deacetylases from maize and other plant species

HDACs have been categorized into three classes11. Classes 1 and 2 contain enzymes that are homologous to the yeast proteins Rpd3 and Hda1, respectively. Proteins related to maize HD2 belong to class 3. Recently, the yeast-silencing information protein Sir2 has been shown to be an NAD-dependent HDAC (Ref. 29), thus defining a fourth class. Three biochemically distinct HDAC activities have been identified in pea and four in maize (HD1A, HD1BI, HD1BII and HD2)4,24. Maize HD1BI and HD1BII are class-1 HDACs, and there is at least one additional member of this family in the databases (Fig. 2). Several EST clones from maize, Arabidopsis and other plant species are also available that are http://plants.trends.com

homologous to the Hda1 family, although none of them has been studied in detail (Fig. 2). Immunological data indicated that maize HD1BI, HD1BII and HATB co-fractionate with a protein that is related to tomato LeMSI and the human Rbap46/48-like proteins24,25 (Fig. 3). These WDrepeat-containing proteins are thought to be responsible for targeting the enzymes to histones because they have been found associated with HDACs, B-type HATs, CAF-1 and the chromatin remodelling complex NURF (Ref. 28). In maize, only the HD1B activity can deacetylate the characteristic diacetylation pattern introduced by HATB on H4, suggesting a possible function for HD1B in the histone deposition process and substantiating the idea of Rbap-like proteins as histone chaperones30. ZmRpd3/HD1BI can functionally complement a yeast rpd3 null mutant31 and antisense downregulation of Arabidopsis AtRPD3A resulted in a delayedflowering phenotype, indicating that histone acetylation might play an important role in plant development32. Another HDAC activity identified in germinating maize embryos is the loosely chromatin-associated HD1A (Fig. 3). The catalytic activity of this enzyme is regulated by phosphorylation. Dephosphorylation of HD1A results in an increase in enzyme activity and a change in substrate specificity30. Whether HD1A belongs to one of the known HDAC families or represents a novel class remains to be determined. HD2-like HDACs (Ref. 33) form multigene families of highly similar members within the plant kingdom34 but no closely related proteins have been identified so far in animals or fungi. HD2 was purified from maize chromatin as a high molecular weight complex composed of three almost identical acidic polypeptides (p39, p42, p45; Fig. 3). This enzyme, too, is subject to phosphorylation but, in contrast with HD1A, whose activity is elevated upon dephosphorylation, dephosphorylation of the HD2 complex almost abolishes its activity. The nucleolar location of HD2 in maize cells suggests a possible role in the regulation of rRNA genes33. Functional characterization of the Arabidopsis homologue AtHD2A (Ref. 35) revealed that it can repress transcription when targeted to a reporter gene in vivo, a feature shared with other HDACs (Ref. 3).

Review

TRENDS in Plant Science Vol.6 No.2 February 2001

Fig. 3. (a) Subcellular location of histone acetyltransferases (HATs) and histone deacetylases (HDACs) in germinating maize embryos. Two nuclear HAT activities (A1, A2) can be distinguished by chromatography. A cytoplasmic B-type HAT (p50) is associated with an Rbap-related protein that is homologous to yeast MSI proteins. HAT-B is highly specific for diacetylation of histone H4 at lysines 12 and 5 in a sequential manner26, and it is also present in the nucleus, as shown by immunofluorescence microscopy of isolated nuclei25. At least three biochemically distinct HDACs can be distinguished; HD1B, the maize homologue of yeast RPD3 exists in at least two forms (HD1BI/ZmRpd3 and ZmHD1BII), both of which are associated with Rbap. HD1A represents a novel type of HDAC whose substrate specificity is regulated by phosphorylation30. This enzyme is only loosely bound to chromatin and is therefore found in the cytoplasmic fraction after cellular fractionation. HD1A and HD1B might shuttle between nuclei and cytoplasm in a differentiation-dependent manner. The plant-specific HD2 family is a high molecular weight oligomeric complex that is exclusively located in the nucleolus and thus likely to be involved in ribosomal chromatin structure and function33. (b) Properties of maize HDACs.

(a) Nucleolus HD1BI/ ZmRpd3

Rbap/ MSI

P p39

p42

Transcriptional repression HD1BII

p45

Rbap/ MSI

HD2 P

P HD1A

63

HD1A HAT-A2 Transcriptional co-activation HAT-A1

HATB- Rbap/ MSI p50

Nucleus Cytoplasm

HAT-B

(b) Nucleolar HD2 complex 400 kDa complex (putative composition) p42 p45

P p39 p39 P p39 p42 p39 p45 p42 P

HDAC active as monomers Phosphorylation HD1A

Phosphorylation or dephosphorylation

p39

HD2p39 non-phosphorylated

p42

HD2p42 phosphorylated

P

Change in substrate specificity

P HD1A

Dephosphorylation = activation

RPD3-type HDAC complexes HD1BI/ ZmRpd3

Rbap/ MSI

HD1BII

Rbap/ MSI

HD2p45 phosphorylated

p45 P

TRENDS in Plant Science

However, downregulation of AtHD2A expression by an antisense approach did not affect expression of the rRNA genes. Instead, reproductive development was disturbed severely, resulting in aborted seed development. Proteins more distantly related to HD2 have been identified in insects, yeast and two parasitic apicomplexans36. The insect proteins belong to a family of FK506-binding proteins (FKBPs) that exhibit peptidyl–prolyl cis–trans isomerase (PPIase) activity. FKBPs have a functional PPIase domain in addition to the HD2-related putative HDAC domain and the charged middle part, suggesting that two http://plants.trends.com

enzymatic activities might be included in one protein. Although HDAC activity could not be detected for any of the homologous proteins (G. Brosch and P. Loidl, unpublished), the potential combination of HDAC and PPIase activity in one protein is intriguing, considering that both activities might be involved in chromatin rearrangements. In spite of the reasonable biochemical characterization of class-1 and -3 HDACs in maize and other plants, information about their physiological roles or downstream targets is largely missing. Even less is known about Sir2-like HDACs, because Sir2 HDAC activity depends on NAD and therefore might have escaped detection using standard assays. Nevertheless, there are certainly also members of this HDAC family in plants29 and it will be interesting to elucidate their functions. Chromatin modifications and silencing

In plants, silencing involves transgenes and transposons, but also endogenous genes37. Silenced regions in the plant genome often contain repeated DNA sequences and exhibit extensive methylation. Methylation of DNA occurs in various species and a large body of evidence indicates that methylation represses transcription, although the underlying molecular mechanism has remained unclear for a long time. Silenced genomic regions also exhibit a repressive chromatin structure characterized by reduced nuclease sensitivity and hypoacetylation of histones38. DNA methyltransferases and methylCpG-binding proteins have been shown to mediate silencing of methylated promoters in animals by recruiting HDACs (Ref. 11). This important finding suggests a mechanism by which histone deacetylation could affect silencing in a methylation-dependent manner. In addition, a link between methylation and histone deacetylation has been established in plants: silencing of underdominant rRNA genes from the allotetraploid hybrid Brassica napus can be relieved by treatment with the methyltransferase inhibitor 5-aza2′deoxycytosine (aza-dC) as well as by the HDAC inhibitor trichostatin A. Both DNA methylation and

64

Review

TRENDS in Plant Science Vol.6 No.2 February 2001

histone deacetylation are likely to function in the same pathway38 because treatment with both compounds did not have an additive effect on the reactivation of the underdominant rRNA genes. Although the protein players in this case have not yet been identified, the scenario is reminiscent of that described for animals. One possibility is that a putative methyl-cytosine-binding protein recruits HDACs and/or other co-repressors to generate a repressive chromatin environment at the underdominant rRNA genes39. Additional methylation-independent but chromatin-dependent mechanisms might be involved in silencing plant transgenes and/or endogenous genes. Evidence for this comes from the recent discovery that the Arabidopsis MOM protein shares sequence homology with SWI2/SNF2-like proteins, which are components of ATP-dependent chromatin remodelling complexes in other organisms. MOM is required to silence a methylated transgene conferring resistance to hygromycin (HPT)40. Mutation of MOM results in reactivation of the HPT locus but does not affect its methylation state. Considering that SWI2/SNF2 ATPases and HDACs have been found associated in transcription repression complexes41, similar mechanisms could also account for silencing in plants, at least for a subset of genes or genomic regions. Histone acetylation and gene regulation

Acknowledgements We gratefully acknowledge discussions, sharing of unpublished data and critical reading of the manuscript by Gerald Brosch, Stefan Graessle, Florian Jesacher, Florentine Marx, Antonius and Marjori Matzke, Mario Motto, Alexandra Pipal, Richard Thompson and Vincenzo Rossi. We apologize for not being able to cite all the primary literature owing to space limitations. Experimental work in our laboratory was supported by grants from the Austrian Science Foundation (FWF-P-11741, FWF-P-15428) and the Austrian National Bank (P-7415) to P.L.

We have only limited knowledge about the regulation of specific endogenous genes by chromatin-dependent mechanisms in plants. As in animals and fungi, nucleosome-modifying enzymes might be involved in regulation. One example is the PICKLE (PKL) gene42, which encodes a CHD3-class protein that contains SNF2-like domains. In pkl mutants, the expression of embryo-specific genes after germination is disturbed, indicating that PKL might be a transcriptional repressor in the regulation of embryo-specific transcription factors. Interestingly, CHD3 ATPases have been described in animals as components of transcription repression complexes that contain HDACs (Ref. 41). Hence, transcriptional regulation by Arabidopsis PKL might involve modification of the chromatin structure at the target genes by a combination of active remodelling and histone deacetylation. The regulation of genes that respond to hormones and growth factors has been extensively studied in mammalian systems. There is a detailed picture of the mechanistic events at certain hormone-regulated promoters. This involves the recruitment of HATs and remodelling complexes to create an active chromatin structure when the ligand is present. In the absence of the hormone, HDACs are recruited, resulting in transcriptional repression43. Similar mechanisms have been proposed to account for the regulation of the phas gene of bean by abscisic acid44. Activation of the phas gene, encoding the storage protein βhttp://plants.trends.com

phaseolin, by the seed-specific transcriptional activator PvALF is accompanied by the formation of a more open chromatin structure. Whether PvALF achieves this by binding to the promoter before chromatin assembly during replication or by recruiting chromatin-modifying complexes, presumably HATs and/or remodelling activities, is not yet clear44. Histone acetylation and cell cycle regulation

Cell cycle progression in yeast and animals is regulated by cyclin-dependent kinases (cdks), cyclins and many other components. In animals, a key player in cell cycle regulation is the retinoblastoma protein (Rb). Active Rb is hypophosphorylated and, in this state, can bind to a variety of proteins, including the S-phase-specific transcription factor E2F. Rb binding to E2F blocks the transcriptional activation of E2Fregulated genes by inactivating E2F and creating a repressive chromatin environment through recruitment of HDAC (Ref. 45). Cell cycle regulation in plants appears to be similar to that in animals because many of the crucial regulators, such as cdks, cyclins and cdk inhibitors, found in animals have also been identified in plants46. Rb-related proteins (RRB1, RRB2) and E2F-like proteins have been identified in maize47. Maize RRBs appear to be able to interact with E2F and to repress E2F-dependent transcription in human cells, and can be phosphorylated by human cdk–cyclin complexes47. Furthermore, maize RRB proteins interact with the HDAC ZmRpd3 (Ref. 31) and cooperate with it in repressing gene transcription (V. Rossi, pers. commun.). This is again reminiscent of regulation in animal cells and provides additional evidence for conservation of the principal mechanisms. Conclusion and perspectives

The basic patterns of growth and development differ between higher plants and animals. When animals have developed into adult organisms, growth and morphogenesis stop and cell division mainly replaces dead cells or specialized cells that undergo continuous turnover. In plants, morphogenesis and growth continue throughout life. Owing to the lack of significant migration of cells, plant morphogenesis is directed mainly by cell division and expansion. Because cell division primarily occurs in distinct meristematic regions, the identity of a cell that leaves the meristematic region is determined predominantly by the position of the cell relative to its neighbour cells. By contrast, in animals, cell identity depends essentially on cell lineage. These basic differences indicate that plant cells have developed unique molecular mechanisms of growth and developmental regulation. Therefore, we might expect that the regulation of chromatin structure and function, and of histone acetylation, would be characterized by novel features distinct from the mechanisms that are conserved among all eukaryotes. An example is the recent

Review

TRENDS in Plant Science Vol.6 No.2 February 2001

finding that H3 and H4 acetylation patterns in field bean nuclei change during the cell cycle and correlate with replication rather than with transcriptional activity20. The fact that higher plants have at least two classes of HDACs, which have not yet been found in animals, indicates that there might be additional regulatory pathways in plant gene regulation and structural chromatin rearrangements. It will also be particularly interesting to undertake structural studies with maize HD1A because this enzyme acts as a monomer, in contrast with animal HDACs, which exert their function in multiprotein complexes of varying composition. The existence of a distinct nucleolar deacetylase (HD2) indicates that the replication and/or References 1 Luger, K. et al. (1997) Crystal structure of the nucleosome core particle at 2.8 Å resolution. Nature 389, 251–260 2 Sterner, D.E. and Berger, S.L. (2000) Acetylation of histones and transcription-related factors. Microbiol. Mol. Biol. Rev. 64, 435–459 3 Cress, W.D. and Seto, E. (2000) Histone deacetylases, transcriptional control, and cancer. J. Cell. Physiol. 184, 1–16 4 Loidl, P. (1994) Histone acetylation: facts and questions. Chromosoma 103, 441–449 5 Strahl, B.D. and Allis, C.D. (2000) The language of covalent histone modifications. Nature 403, 41–45 6 Krebs, J.E. et al. (1999) Cell cycle-regulated histone acetylation required for expression of the yeast HO gene. Genes Dev. 13, 1412–1421 7 Parekh, B.S. and Maniatis, T. (1999) Virus infection leads to localized hyperacetylation of histones H3 and H4 at the IFN-beta promoter. Mol. Cell 3, 125–129 8 Madisen, L. et al. (1998) The immunoglobulin heavy chain locus control region increases histone acetylation along linked c-myc genes. Mol. Cell. Biol. 18, 6281–6292 9 Crane-Robinson, C. et al. (1999) Chromatin immunoprecipitation assays in acetylation mapping of higher eukaryotes. Methods Enzymol. 304, 533–547 10 Imhof, A. and Wolffe, A.P. (1998) Transcription: gene control by targeted histone acetylation. Curr. Biol. 8, R422–R424 11 Davie, J.R. and Chadee, D.N. (1998) Regulation and regulatory parameters of histone modifications. J. Cell. Biochem. 31 (Suppl.), 203–213 12 Munshi, N. et al. (1998) Acetylation of HMG I(Y) by CBP turns off IFN beta expression by disrupting the enhanceosome. Mol. Cell 2, 457–467 13 Dhalluin, C. et al. (1999) Structure and ligand of a histone acetyltransferase bromodomain. Nature 399, 491–496 14 McMurry, M.T. and Krangel, M.S. (2000) A role for histone acetylation in the developmental regulation of VDJ recombination. Science 287, 495–498 15 Jacobson, S. and Pillus, L. (1999) Modifying chromatin and concepts of cancer. Curr. Opin. Genet. Dev. 9, 175–184 16 Brosch, G. et al. (1995) Inhibition of maize histone deacetylases by HC toxin, the host-selective toxin of Cochliobolus carbonum. Plant Cell 7, 1941–1950 http://plants.trends.com

65

transcription of ribosomal gene chromatin might be controlled by other histone-modifying enzymes than those involved in animals. Exploiting these differences will be an interesting challenge not only for basic molecular research but also for applied plant technologies. Histone acetylation is involved in plant–microorganism interactions16. The maize pathogen Cochliobolus carbonum produces the hostspecific HC toxin, and its inhibitory effect on HDACs seems to be the main pathogenicity mechanism. It will be interesting to find out whether the interaction between pathogenic microorganisms and mammals also involves the specific interference with host histone acetylation patterns. This could establish a new line of treatment for infectious disease in humans.

17 Waterborg, J.H. et al. (1990) Dynamic histone acetylation in alfalfa cells. Butyrate interference with acetate labeling. Biochim. Biophys. Acta 1049, 324–330 18 Belyaev, N.D. et al. (1998) The acetylation patterns of histones H3 and H4 along Vicia faba chromosomes are different. Chromosome Res. 6, 59–63 19 Vyskot, B. et al. (1999) Euchromatic domains in plant chromosomes as revealed by H4 histone acetylation and early DNA replication. Genome 42, 343–350 20 Jasencakova, Z. et al. (2000) Histone H4 acetylation of euchromatin and heterochromatin is cell cycle dependent and correlated with replication rather than with transcription. Plant Cell 12, 2087–2100 21 Waterborg, J.H. (1992) Identification of five sites of acetylation in alfalfa histone H4. Biochemistry 31, 6211–6219 22 Strahl, B.D. et al. (1999) Methylation of histone H3 at lysine 4 is highly conserved and correlates with transcriptionally active nuclei in Tetrahymena. Proc. Natl. Acad. Sci. U. S. A. 96, 14967–14972 23 Manzanero, S. et al. (2000) The chromosomal distribution of phosphorylated histone H3 differs between plants and animals at meiosis. Chromosoma 109, 308–317 24 Lechner, T. et al. (2000) RPD3-type histone deacetylases in maize embryos. Biochemistry 39, 1683–1692 25 Lusser, A. et al. (1999) Analysis of the histone acetyltransferase B complex of maize embryos. Nucleic Acids Res. 27, 4427–4435 26 Kolle, D. et al. (1998) Substrate and sequential site specificity of cytoplasmic histone acetyltransferases of maize and rat liver. FEBS Lett. 421, 109–114 27 Grunstein, M. (1997) Histone acetylation in chromatin structure and transcription. Nature 389, 349–352 28 Verreault, A. (2000) De novo nucleosome assembly: new pieces in an old puzzle. Genes Dev. 14, 1430–1438 29 Guarente, L. (2000) Sir2 links chromatin silencing, metabolism, and aging. Genes Dev. 14, 1021–1026 30 Kolle, D. et al. (1999) Different types of maize histone deacetylases are distinguished by a highly complex substrate and site specificity. Biochemistry 38, 6769–6773 31 Rossi, V. et al. (1998) Identification and characterisation of an RPD3 homologue from maize (Zea mays L.) that is able to complement an

32

33

34

35

36 37

38

39

40

41

42

43 44

45

46

47

rpd3 null mutant of Saccharomyces cerevisiae. Mol. Gen. Genet. 258, 288–296 Wu, K. et al. (2000) Functional analysis of a RPD3 histone deacetylase homologue in Arabidopsis thaliana. Plant Mol. Biol. 44, 167–176 Lusser, A. et al. (1997) Identification of maize histone deacetylase HD2 as an acidic nucleolar phosphoprotein. Science 277, 88–91 Dangl, M. et al. Comparative analysis of HD2 type histone deacetylases in higher plants. Planta (in press) Wu, K. et al. (2000) Functional analysis of HD2 histone deacetylase homologues in Arabidopsis thaliana. Plant J. 22, 19–27 Aravind, L. and Koonin, E.V. (1998) Second family of histone deacetylases. Science 280, 1167a Matzke, A.J. and Matzke, M.A. (1998) Position effects and epigenetic silencing of plant transgenes. Curr. Opin. Plant Biol. 1, 142–148 Meyer, P. (1999) The role of chromatin remodeling in transgene silencing and plant development. In vitro Cell. Dev. Biol. Plant 35, 29–36 Pikaard, C.S. (1999) Nucleolar dominance and silencing of transcription. Trends Plant Sci. 4, 478–483 Amedeo, P. et al. (2000) Disruption of the plant gene MOM releases transcriptional silencing of methylated genes. Nature 405, 203–206 Wade, P.A. et al. (1998) A multiple subunit Mi-2 histone deacetylase from Xenopus laevis cofractionates with an associated Snf2 superfamily ATPase. Curr. Biol. 8, 843–846 Ogas, J. et al. (1999) PICKLE is a CHD3 chromatin-remodeling factor that regulates the transition from embryonic to vegetative development in Arabidopsis. Proc. Natl. Acad. Sci. U. S. A. 96, 13839–13844 Leo, C. and Chen, J.D. (2000) The SRC family of nuclear receptor coactivators. Gene 245, 1–11 Li, G. et al. (1999) Beta-phaseolin gene activation is a two-step process: PvALF-facilitated chromatin modification followed by abscisic acidmediated gene activation. Proc. Natl. Acad. Sci. U. S. A. 96, 7104–7109 Brehm, A. et al. (1998) Retinoblastoma protein recruits histone deacetylase to repress transcription. Nature 391, 597–601 den Boer, B.G. and Murray, J.A. (2000) Triggering the cell cycle in plants. Trends Cell Biol. 10, 245–250 Gutierrez, C. (1998) The retinoblastoma pathway in plant cell cycle and development. Curr. Opin. Plant Biol. 1, 492–497