NATURE | VOL 408 | 14 DECEMBER 2000 | www.nature.com 877

letters to nature

.................................................................Dynamic binding of histone H1to chromatin in living cellsTom Misteli*, Akash Gunjan²³, Robert Hock§, Michael Bustink& David T. Brown²

* National Cancer Institute, NIH, Bethesda, Maryland 20892, USA² University of Mississippi Medical Center, Jackson, Mississippi 39216, USA

§ University of Wuerzburg, Biocenter, Germany

kNational Cancer Institute, NIH, Laboratory of Molecular Biology, Bethesda,Maryland, D 20892, USA³ Present address: Imperial Cancer Research Fund, Potters Bar, EN6 3LD, UK...............................................................................................................................................

The linker histone H1 is believed to be involved in chromatinorganization by stabilizing higher-order chromatin structure1±3.Histone H1 is generally viewed as a repressor of transcription as itprevents the access of transcription factors and chromatin remod-elling complexes to DNA4±6. Determining the binding propertiesof histone H1 to chromatin in vivo is central to understandinghow it exerts these functions. We have used photobleachingtechniques to measure the dynamic binding of histone H1±GFPto unperturbed chromatin in living cells. Here we show thatalmost the entire population of H1±GFP is bound to chromatinat any one time; however, H1±GFP is exchanged continuouslybetween chromatin regions. The residence time of H1±GFP onchromatin between exchange events is several minutes in botheuchromatin and heterochromatin. In addition to the mobilefraction, we detected a kinetically distinct, less mobile fraction.After hyperacetylation of core histones, the residence time ofH1±GFP is reduced, suggesting a higher rate of exchange uponchromatin remodelling. These results support a model in which

linker histones bind dynamically to chromatin in a stop-and-gomode.

To visualize the linker histone H1 in living cells, we fused thecoding region of the green ¯uorescent protein (GFP) to that ofmouse histone variants H10 or H1c, placed under the control of azinc-inducible promoter. We established stable cell lines in mouseBALB/c 3T3 ®broblasts as described7. All experiments were done inexponentially growing cells and in the absence of exogenous Zn2+,using only the basal activity of the promoter to prevent alterationsof chromatin structure, as reported in density-arrested cells uponhigh overexpression of H1 variants7,8. On the basis of high perfor-mance liquid chromatography (HPLC) pro®ling, we estimate thatthe cell lines overexpress less than 5% H1±GFP in addition toendogenous H1 (Fig. 1a).

By several criteria, the GFP fusion proteins behaved identically toendogenous H1 variants. Both H10±GFP and H1c±GFP proteinswere released from nuclei upon micrococcal nuclease digestion withkinetics identical to that of the respective unmodi®ed endogenousprotein, suggesting proper positioning of the fusion proteins onDNA (Fig. 1b)7. The salt-extraction properties of the H1±GFPfusion proteins were indistinguishable from those of their endo-genous counterparts (Fig. 1c). Like endogenous H1, the entire poolof both H1±GFP variants was associated with chromosomes inmitotic cells, and expression of the fusion proteins had no effect oncell-cycle behaviour or cell proliferation (data not shown). In cellsstably overexpressing H10±GFP or H1c±GFP, the fusion protein wasfound diffusely distributed throughout the nucleoplasm and con-centrated in 20±40 nuclear foci (Fig. 1d). The foci colocalized inparaformaldehyde-®xed cells with the heterochromatin markerprotein HP1 and Hoechst 33342, an indicator of heterochromatinregions in murine cells (Fig. 1d, top)9. The presence of H1c±GFP inheterochromatin in living cells was con®rmed by colocalization ofthe GFP fusions with Hoechst 33342 (Fig. 1d, bottom). Consistent

d

H1c–GFP HP1 DNA

H1c–GFP DNA

0 2 min 5 min 10 min 15 min

20 min 25 min 30 min 50 min 60 min

b ca

ed

Parent

H10–GFP

H1c–GFP

GFP

H10–GFP

H1c–GFP

H10

H10–GFPH1cH1c–GFP

H10

H10–GFPH1cH1c–GFP

GFP

Elution time (min)

H10H1bH1a

H1d/e

H1c80

60

40

20

0

% r

elea

sed

0.1 0.2 0.3 0.4

Micrococcal nuclease (U m–1) KCl (mM)

100

80

60

40

20

0

% b

ound

250 300 350 400 450 500

Figure 1 Properties of H1±GFP proteins. a, HPLC pro®le of H1±GFP cell lines. Total

histones were extracted from either parental BALB/c 3T3 cells or a cell line stably

expressing H10±GFP or H1c±GFP and separated by HPLC. Recombinant GFP was used

as a control. b, Micrococcal nuclease sensitivity. The amount of H1 released on MNase

digestion was determined. c, Salt-extraction properties. The amount of native H1 released

on extraction with increasing concentrations of KCl was determined. d, Localization of

histone H1±GFP. Top, in ®xed BALB/c 3T3 cells, stably expressed H1c±GFP colocalizes

in nuclear foci with the heterochromatin marker HP1 and Hoechst 33342 (arrows).

Bottom, identical colocalization of H1c±GFP with Hoechst 33342 in heterochromatic

regions of living cells (arrows). e, Time-lapse microscopy of H1c±GFP expressing Balb/c

3T3 cells. Images were taken every 30 s and selected images are shown. Only restricted

local motion of chromatin was observed over extended observation periods (arrows). Scale

bar, 3.4 mm. For experimental details see Supplementary Information.

© 2000 Macmillan Magazines Ltd

with reports on restricted mobility of chromatin regions in vivo10,time-lapse microscopy experiments revealed no signi®cant posi-tional movements of H1c±GFP-positive foci over extended obser-vation periods (Fig. 1e). Identical results were observed for H10±GFP (data not shown).

We have previously used photobleaching techniques to show thatmany proteins are highly mobile in the mammalian cell nucleus11.Here we applied ¯uorescence recovery after photobleaching (FRAP)to probe the dynamic properties of histone H1±GFP in unper-turbed chromatin of living cells. After bleaching an area in thenucleus of a cell expressing either H1c±GFP or H10±GFP, werecorded the recovery of ¯uorescence signal in the bleached areaby time-lapse imaging11 (Fig. 2a). As a standard we used GFP, whichmoves relatively freely in the nucleus and whose ¯uorescence signalrecovers to 98 6 3% of the pre-bleach value in less than 1 s (refs 11,12; Fig. 2b). In contrast, H1c±GFP and H10±GFP recovery was lessthan 1 6 1% during the ®rst second, less than 3 6 1% in the ®rst 5 sand less than 15 6 3% during the ®rst 20 s after bleaching (Fig. 2a,b). These recovery kinetics are signi®cantly slower than thosereported for several nuclear proteins including the chromatin-binding proteins HMG-17 (ref. 11) and HMG-14 (Fig. 2b,P , 0.001). These results were con®rmed by ¯uorescence loss inphotobleaching (FLIP) analysis11 (see Supplementary Information).The reduced mobility of H1±GFP was the consequence of H1±GFPbinding to chromatin, as mutants lacking either the globulardomain (DG) or the carboxy-terminal domain (DC), both ofwhich are involved in DNA binding, showed a signi®cantly morerapid recovery, and their recovery level in 15 s was similar (97 6 2%)to that of GFP alone (Fig. 2c). We conclude that in the nucleus of aliving cell most H1±GFP molecules at any given time are bound tochromatin.

We characterized the dynamics of H1±GFP exchange in unper-turbed chromatin by FRAP experiments using extended recoveryperiods. Upon bleaching a small nuclear area, slow recovery of the¯uorescence signal was observed and recovery reached a plateau in200±250 s (Fig. 3). This observation shows directly that H1±GFP iscontinuously exchanged between chromatin regions in the cell

nucleus (Fig. 3). For these experiments, we de®ned heterochroma-tin on the basis of morphological criteria as areas strongly labelledwith H1±GFP; euchromatin was de®ned morphologically as areasweakly or not labelled with H1±GFP. Recovery kinetics were similarregardless of whether the bleached area was in an euchromatic orheterochromatic region (Fig. 3a±d). In both euchromatin andheterochromatin not all of the ¯uorescence signal was recovered(Fig. 3c, d). The recovery levels were 90.7 6 4.2% and 74.4 6 2.6%for H1c±GFP, and 89.6 6 3.8% and 76.4 6 2.7% for H10±GFP, ineuchromatin and heterochromatin, respectively (Fig. 3c, d). Thesevalues were signi®cantly different (P , 0.001 in heterochromatin;P , 0.005 in euchromatin) from the recovery value for GFP alone(98 6 3%).

This behaviour indicates the presence of two distinct kineticpools of H1±GFP in the nucleus: a large mobile pool, whichrepresents the continuously exchanging molecules and is respon-sible for the ¯uorescence signal recovery; and a smaller, less mobilepool, which does not contribute to the recovery. This less mobilefraction was signi®cantly larger in heterochromatin (,25%) than ineuchromatin (,10%), suggesting the presence of more staticallybound H1±GFP molecules in heterochromatin than in euchromatin(Fig. 3e). This less mobile fraction might be distinct biochemically orthe stronger binding might be due to properties of chromatin. Wealso cannot exclude the possibility that the less mobile fraction in`euchromatin' originates from H1±GFP stably associated withinterspersed heterochromatin. Because most H1±GFP moleculesare bound to chromatin at any given time (Fig. 2b), H1±GFPmolecules must ®rst be released from the DNA to which they arebound at the time of bleaching to be able to move into the bleachedarea. Therefore, the residence time of the mobile fraction can beestimated from the recovery graphs as the time point when therecovery plateau is reached. For both histone variants in euchro-matin or heterochromatin, this time was between 220 and 250 s(Fig. 3c, d, f). Our observations show that histone H1±GFP exists inat least two kinetically distinct fractions in the nucleus and that mostH1±GFP molecules reside on chromatin in living cells for ,220 sbefore dissociating and rapidly rebinding to an available binding site.

letters to nature

878 NATURE | VOL 408 | 14 DECEMBER 2000 | www.nature.com

0.2

0.4

0.6

0

0.2

0.4

0.6

0.8

1

0 5 10 15 20 25Recovery time (s)

0 5 10 15 20 25Recovery time (s)

Rel

ativ

e in

tens

ity

H1c

H10

H1cH10

HMG-14

GFPb

0

0.2

0.4

0.6

0.8

1

Rel

ativ

e in

tens

ity

c

a

pre 0 1.6 s 4.8 s 16 s 32 s

∆C

∆G

wt

Figure 2 FRAP analysis of H1±GFP. a, A cell expressing H1c±GFP was imaged before

and during recovery after bleaching of a nucleoplasmic area. Images were taken at the

indicated times after the bleach pulse (yellow circle). The indicated area is shown enlarged

in pseudocolour in the lower panel. Scale bars: 2.6 mm (top); 1.2 mm (bottom). b,

Quantitative analysis of FRAP recovery of wild-type H1±GFP variants. The absence of

rapid recovery demonstrates the absence of a freely mobile pool of H1±GFP variants.

c, FRAP analysis of H1±GFP variants lacking the globular domain (DG) or the C-terminal

domain (DC) and wild-type (wt) H1±GFP. Mutant recovery was signi®cantly faster than

that of wild-type H1±GFP. Values represent means 6 s.d. from at least 10 cells from 3

experiments.

© 2000 Macmillan Magazines Ltd

Core histone acetylation has been implicated in changes inchromatin structure that result in the opening of chromatin andincreased accessibility of DNA to remodelling factors6,13. The effectof core histone hyperacetylation on the binding of histone H1 is notclear14±17. Treatment of Balb/c 3T3 cells with the histone deacetylaseinhibitor trichostatin A (TSA) results in the marked increase in thelevels of acetylation of all four core histones8. In FRAP experimentsof TSA-treated cells, the recovery of H10±GFP and H1c±GFP wascomplete in 100±150 s in both euchromatin and heterochromatin;this was signi®cantly faster than the recovery time of 220±250 s inuntreated cells (Fig. 4a±d; P , 0.001). The increase in recovery rate

translates into a shorter residence time on TSA treatment (Fig. 4e).Although recovery levels in the euchromatin of TSA-treated cellswere similar to those of control cells (,90%), the recovery levelin heterochromatin was increased from ,75% in control cells to87 6 4% for H1c±GFP (P , 0.001) and 92 6 4% for H10±GFP(P , 0.005; Fig. 4c, d, f). This difference indicated that the size of theless mobile fraction of H1±GFP variants in heterochromatin wasreduced from 25% in control cells to 13 6 4% for H10±GFP and to8 6 4% for H1c±GFP on TSA treatment (P , 0.001 for H10±GFP;P , 0.005 for H1c±GFP) (Fig. 4f). The increased exchange rate wasnot a consequence of the transcriptional activity of the cell, as

letters to nature

NATURE | VOL 408 | 14 DECEMBER 2000 | www.nature.com 879

pre 0 16 s 42 s 189 s 320 s

0

0.2

0.4

0.6

0.8

1.0

0 50 100 150 200 250 300 350 400

Recovery time (s)

a

Rel

ativ

e in

tens

ity

0

0.2

0.4

0.6

0.8

1.0

0 50 100 150 200 250 300 350 400

Recovery time (s)

Rel

ativ

e in

tens

ity

c d

0

5

10

15

20

25

30

Imm

obile

frac

tion

(%)

100

50

0

150

200

250

300

Res

iden

ce t

ime

(s)

e f

eu

het eu

het

H10 H1c

eu het eu het

H10 H1c

b

H1c euchromatin H1c heterochromatin Fixed cells H1o euchromatin H1o heterochromatin Fixed cells

Figure 3 FRAP on H1±GFP in heterochromatin or euchromatin. a, b, Cells expressing

H1c±GFP were imaged before and during recovery after bleaching of either a

heterochromatic (a) or an euchromatic (b) area (yellow circle). Images were taken at the

indicated times after end of the bleach pulse. The indicated area is enlarged in

pseudocolour in the bottom panels. Scale bars: 3.2 mm (top); 1 mm (bottom).

c, d, Quantitative analysis of FRAP experiments after bleaching of H1c±GFP (c) or

H10±GFP (d) in either an euchromatic (open circles) or heterochromatic (®lled circles)

region of the nucleus. e, f, Immobile fraction (e) and residence time (f) of H10±GFP or

H1c±GFP in heterochromatin or euchromatin. The residence time is similar in

heterochromatin and euchromatin. The immobile fraction is larger in heterochromatin

than in euchromatin. Values represent means 6 s.d. from at least 10 cells from 3

experiments.

© 2000 Macmillan Magazines Ltd

inhibition of transcription alone by a-amanitin did not have anyeffect on the recovery rate (data not shown). These observationssuggest that hyperacetylation increases the exchange rate of histoneH1±GFP from chromatin of living cells.

We have used photobleaching techniques to characterize thebinding dynamics of the linker histone H1±GFP to unperturbedchromatin in vivo. To our knowledge this is the ®rst measurement ofthe residence time of a DNA-binding protein in chromatin of livingcells. We ®nd that at any given time, most histone H1±GFPmolecules are bound to chromatin. However, most H1±GFPmolecules are continuously exchanged between chromatin regions.These observations lead to the conclusion that most H1±GFPmolecules bind chromatin by a `stop-and-go' mechanism: a histoneH1±GFP molecule resides on chromatin for ,220 s. The moleculethen dissociates and rapidly binds to another binding site. On thebasis of the mobility properties of other nuclear proteins, wespeculate that the intermediate in this exchange is freely mobileand moves by diffusion in the nucleoplasm11.

Transfer of H1±GFP between chromatin segments has beensuggested by in vitro salt-extraction studies and cell fusion experi-ments of cells micro-injected with labelled H1 (refs 18±21).

However, those earlier studies could not determine the residencetime of the mobile fraction of histone H1, nor did they reveal thepresence of kinetically distinct pools of H1. At present, we do notknow the functional or physiological signi®cance of the twofractions, and it is not clear how these two fractions differ atthe molecular level. An interesting possibility is that proteinphosphorylation, which alters the DNA interaction properties ofH1 (refs 22, 23), can interconvert the two pools. We also ®ndevidence for increased exchange of H1±GFP between chromatinregions on hyperacetylation. Increased core histone acetylationis a hallmark of transcriptionally active chromatin, and results,in many cases, from the action of chromatin remodellingactivities6,24,25. On treatment of cells with TSA, we observed anincreased exchange rate of the mobile fraction of H1±GFP, suggest-ing a `looser' binding of histone H1±GFP to chromatin. Theseresults extend into living-cells biochemical studies showing thatfractionated `active' chromatin is relatively depleted in H1 andenriched in acetylated core histones26,27. We propose that thedynamic nature of binding is an essential feature of linker histonesin their functions as regulators of chromatin remodelling andchromatin structure in vivo. M

letters to nature

880 NATURE | VOL 408 | 14 DECEMBER 2000 | www.nature.com

0

5

10

15

20

25

30

Imm

obile

frac

tion

(%)

0

0.2

0.4

0.6

0.8

1.0

0 50 100 150 200 250 300 350 400

Recovery time (s)

Rel

ativ

e in

tens

ity

H10 euchromatin H10 euchromatin TSA

0

50

100

150

200

250

300

Res

iden

ce t

ime

(s)

Control+ TSA

Control+ TSA

a

0

0.2

0.4

0.6

0.8

1.0

0 50 100 150 200 250 300 350 400

Recovery time (s)

Rel

ativ

e in

tens

ity

H1c euchromatin H1c euchromatin TSA

b

0

0.2

0.4

0.6

0.8

1.0

0 50 100 150 200 250 300 350 400

Recovery time (s)

Rel

ativ

e in

tens

ity

H10 heterochromatin H10 heterochromatin TSA

c

0

0.2

0.4

0.6

0.8

1.0

0 50 100 150 200 250 300 350 400

Recovery time (s)

Rel

ativ

e in

tens

ity

H1c heterochromatin H1c heterochromatin TSA

d

e f

eu het eu het

H10 H1c eu het eu het

H10 H1c

Figure 4 FRAP on H1±GFP in heterochromatin or euchromatin after TSA treatment. Cells

were treated with 50 ng ml-1 TSA for 2 h before FRAP analysis. a±d, Quantitative analysis

of FRAP experiments after bleaching of either variant in heterochromatin (open circles) or

euchromatin (®lled circles). e, f, Residence time (e) and immobile fractions (f) of H10±GFP

or H1c±GFP in heterochromatin or euchromatin after TSA treatment as compared with

control cells. The residence time of H1±GFP in euchromatin and heterochromatin was

reduced upon TSA treatment. The immobile fraction of H1±GFP in heterochromatin, but

not euchromatin, was reduced. Values represent means 6 s.d. from at least 10 cells from

3 experiments.

© 2000 Macmillan Magazines Ltd

MethodsConstructs and cell lines

The coding sequence for enhanced GFP was excised from pEGFP-C1 (Clontech) and wasinserted after the last lysine residue of the H1 variants in pMTH1cneo or pMTH10neo(ref. 7). An additional alanine residue was inserted between GFP and the H1 variant duringthe cloning process. Deletion mutants lacking the C-terminal domain (DC) or thecentral globular domain (DG) were generated by standard subcloning methods (seeSupplementary Information). H10DC contained residues 1±91; H1cDC, 1±104; H10DG,1±20 and 77±193; H1cDG, 1±35 and 142±211. Stable cell lines expressing H1±GFP fusionproteins were generated from BALB/c 3T3 ®broblasts and grown as described7. More than15 cell lines for each variant were generated. For details see Supplementary Information.

HPLC pro®ling, nuclease digestion and salt extraction

Expression levels of H1±GFP fusion proteins were determined by extraction of totalchromatin-bound histones and quantitative separation by HPLC as described28. Release ofH1±GFP proteins on micrococcal nuclease (MNase) digestion of nuclei was determinedby ¯uorimetry of MNase-soluble material as described7. Release of H1 proteins on MNasedigestion of the same sample was determined by quantitative HPLC of MNase-insolublematerial as described7. To measure the salt extraction of H1 variants, crude nuclei wereextracted with high salt buffer containing increasing concentrations of KCl. Total histonesin the pelleted material were analysed by quantitative HPLC as described28. For details seeSupplementary Information.

Indirect ¯uorescence and in vivo microscopy, and FRAP

Indirect immuno¯uorescence microscopy was carried out as described29, using ®xation in2% buffered paraformaldehyde in PBS for 15 min at room temperature. A rabbitpolyclonal antibody against HP1 (kindly provided by E. Minc, Institute Pasteur, Paris) wasused at 1:500 in PBS9. Hoechst stain 33342 was used at 500 ng ml-1 for 10 min at roomtemperature. Cells were mounted using Vectashield and observed on a Nikon E800microscope using a ´60 Planapo Nikon objective (N.A. 1.4) and a PhotometricsMicroMax cooled CCD camera (1,300 ´ 1,030 array, 6.7 mm pixel size, 5 MHz, image pixelsize 80 nm). Images were generated and analysed in Metamorph (Universal Imaging). Forlive-cell microscopy, we grew cells to 50% con¯uence in Nalgene LabTek II chambers andcarried out microscopy with a Leica TCS-SP confocal microscope using the 488-nm laserline of an argon laser (20 mW nominal output, beam width at specimen 0.2 mm, detection500±575 nm using a photon multiplier tube with pinhole setting at 1, image pixel size47 nm). All experiments were done at 37 8C.

Experimental details of FRAP have been described11. Five single imaging scans wereacquired, followed by a single bleach pulse of 200±500 ms using a spot of ,1 mm indiameter without scanning. Single-section images were then collected at 1-s intervals (15images), followed by 10-s intervals (10 images) and then by 30-s intervals (8 images). Forimaging, the laser power was attenuated to 1% of the bleach intensity. FRAP recoverycurves were generated from background-subtracted images. The total cellular ¯uorescencewas determined for each image and compared with the initial total ¯uorescence to accountfor the amount of signal lost during the bleach pulse and during imaging as described11.Relative recoveries were determined as described11. In FRAP experiments, typically ,10%of total ¯uorescence was lost during the bleach pulse and less than 5% of ¯uorescence waslost during the entire imaging phase. The level of auto¯uorescence in cells not expressingthe fusion proteins was 1% or less of that of expressing cells. The standard Student's t-testwas used to determine the statistical signi®cance of results. All quantitative valuesrepresent averages from at least 10 cells from 3 independent experiments.

Received 29 August; accepted 25 September 2000.

1. Thoma, F. & Koller, T. In¯uence of histone H1 on chromatin structure. Cell 12, 101±107 (1977).

2. Ramakrishnan, V. Histone H1 and chromatin higher order structure. Crit. Rev. Eukaryot. Gene Expr. 7,

215±230 (1997).

3. Thomas, J. O. Histone H1: location and role. Curr. Opin. Cell Biol. 11, 312±317 (1999).

4. Croston, G. E., Kerrigan, L. A., Lira, L. M., Marshak, D. R. & Kadonaga, J. T. Sequence-speci®c

antirepression of histone H1-mediated inhibition of basal RNA polymerase II transcription. Science

251, 643±649 (1991).

5. Zlatanova, J. & van Holde, K. Linker histones versus HMG1/2: a struggle for dominance? BioEssays 20,

588±588 (1998).

6. Strahl, B. D. & Allis, C. D. The language of covalent histone modi®cations. Nature 403, 41±45 (2000).

7. Gunjan, A., Alexander, B. T., Sittman, D. B. & Brown, D. T. Effects of H1 histone variant

overexpression on chromatin structure. J. Biol. Chem. 274, 37950±37956 (1999).

8. Gunjan, A. & Brown, D. T. Overproduction of histone H1 variants in vivo increases basal and induced

activity of the mouse mammary tumor virus promoter. Nucleic Acids Res. 27, 3355±3363 (1999).

9. Minc, E., Allory, Y., Worman, H. J., Courvalin, J.-C. & Buenida, B. Localization and phosphorylation

of HP1 proteins during the cell cycle in mammalian cells. Chromosoma 108, 220±234 (1999).

10. Marshall, W. F. et al. Interphase chromosomes undergo constrained diffusional motion in living cells.

Curr. Biol. 7, 930±939 (1997).

11. Phair, R. D. & Misteli, T. High mobility of proteins in the mammalian cell nucleus. Nature 404, 604±

609 (2000).

12. Houtsmuller, A. B. et al. Action of DNA repair endonuclease ERCC1/XPF in living cells. Science 284,

958±961 (1999).

13. Struhl, K. Histoneacetylation and transcriptional regulatory mechanisms. Genes Dev. 12, 599±606

(1998).

14. Ura, K., Wolffe, A. P. & Hayers, J. J. Core histone acetylation does not block linker histone binding to a

nucleosome including a Xenopus borealis 5S rRNA gene. J. Biol. Chem. 269, 27171±27174 (1994).

15. Krajewski, W. A. & Becker, P. B. Reconstitution of hyperacetylated, DNase I-sensitive chromatin

characterized by high conformational ¯exibility of nucleosomal DNA. Proc. Natl Acad. Sci. 95, 1540±

1545 (1998).

16. Ridsdale, J. A., Hendzel, M. J., Delcuve, G. P. & Davie, J. R. Histone acetylation alters the capacity of the

H1 histones to condense transcriptionally active/competent chromatin. J. Biol. Chem. 265, 5150±5156

(1990).

17. Juan, L.-J., Utley, R. T., Adams, C. C., Vettese-Dadey, M. & Workman, J. L. Differential repression of

transcription factor binding by histone H1 is regulated by the core histone amino termini. EMBO J.

13, 6031±6040 (1994).

18. Bates, D. L., Butler, P. J., Pearson, E. C. & Thomas, J. O. Stability of the higher order structure of

chicken erythrocyte chromatin in solution. Eur. J. Biochem. 119, 469±476 (1981).

19. Caron, F. & Thomas, J. O. Exchange of histone H1 between chromatin segments. J. Mol. Biol. 146,

513±537 (1981).

20. Louters, L. & Chalkley, R. Exchange of histones H1, H2A, and H2B in vivo. Biochemistry 24, 3080±

3085 (1985).

21. Wu, L. H., Kuehl, L. & Rechsteiner, M. Dynamic behavior of histone H1 microinjected into HeLa cells.

J. Cell Biol. 103, 565±574 (1986).

22. Lu, M. J., Mpoke, S. S., Dadd, C. A. & Allis, C. D. Phosphorylated and dephosphorylated linker histone

H1 reside in distinct chromatin domains in Tetrahymena macronuclei. Mol. Biol. Cell 6, 1077±1087

(1995).

23. Dou, Y., Mizzen, C. A., Abrams, M., Allis, C. D. & Gorovsky, M. A. Phosphorylation of linker histone

H1 regulates gene expression in vivo by mimicking H1 removal. Mol. Cell 4, 641±647 (1999).

24. Grunstein, M. Histone acetylation in chromatin structure and transcription. Nature 389, 349±352

(1997).

25. Berger, S. L. Gene activation by histone and factor acetyltransferases. Curr. Opin. Cell Biol. 11, 336±

341 (1999).

26. Hebbes, T. R., Thorne, A. W. & Crane-Robinson, C. A direct link between core histone acetylation and

transcriptionally active chromatin. EMBO J. 7, 1395±1402 (1988).

27. Bresnick, E. H., Bustin, M., Marsaud, V., Richard-Foy, H. & Hager, G. L. The transcriptionally-active

MMTV promoter is depleted of histone H1. Nucleic Acids Res. 20, 273±278 (1992).

28. Brown, D. T., Alexander, B. T. & Sittman, D. B. Differential effect of H1 variant overexpression on cell

cycle progression and gene expression. Nucleic Acids Res. 24, 486±493 (1996).

29. Misteli, T. & Spector, D. L. Serine/threonine phosphatase 1 modulates the subnuclear distribution of

pre-mRNA splicing factors. Mol. Biol. Cell 7, 1559±1572 (1996).

Supplementary information is available on Nature's World-Wide Web site(http://www.nature.com) or as paper copy from the London editorial of®ce of Nature.

Acknowledgements

We thank D. Sittman for valuable discussion, S. Leuba for critical comments on themanuscript and E. Minc for providing HP1 antibodies.

Correspondence and requests for materials should be addressed to T.M.(e-mail: mistelit@mail.nih.gov).

letters to nature

NATURE | VOL 408 | 14 DECEMBER 2000 | www.nature.com 881

.................................................................Metal-ion coordination by U6 smallnuclear RNA contributes tocatalysis in the spliceosomeShyue-Lee Yean, Gerald Wuenschell, John Termini & Ren-Jang Lin

Department of Molecular Biology, Beckman Research Institute of the City of Hope,Duarte, California 91010-3011, USA

..............................................................................................................................................

Introns are removed from nuclear messenger RNA precursorsthrough two sequential phospho-transesteri®cation reactions in adynamic RNA±protein complex called the spliceosome1,2. Butwhether splicing is catalysed by small nuclear RNAs3,4 in thespliceosome is unresolved. As the spliceosome is a metalloen-zyme5±7, it is important to determine whether snRNAs coordinatecatalytic metals. Here we show that yeast U6 snRNA coordinates ametal ion that is required for the catalytic activity of the spliceo-some. With Mg2+, U6 snRNA with a sulphur substitution for thepro-RP or pro-SP non-bridging phosphoryl oxygen of nucleotideU80 reconstitutes a fully assembled yet catalytically inactivespliceosome. Adding a thiophilic ion such as Mn2+ allows the®rst transesteri®cation reaction to occur in the U6/sU80(SP)- butnot the U6/sU80(RP)-reconstituted spliceosome. Mg2+ competi-tively inhibits the Mn2+-rescued reaction, indicating that themetal-binding site at U6/U80 exists in the wild-type spliceosomeand that the site changes its metal requirement for activity in theSP spliceosome. Thus, U6 snRNA contributes to pre-messenger

© 2000 Macmillan Magazines Ltd