In vitro acetylation of histones in rat liver chromatin

Material Information

In vitro acetylation of histones in rat liver chromatin
Added title page title:
Acetylation of histones in rat liver
Racey, Louise Adele, 1941-
Publication Date:
Copyright Date:
Physical Description:
xi, 118 leaves : illus. ; 28 cm.


Subjects / Keywords:
Acetates ( jstor )
Acetylation ( jstor )
DNA ( jstor )
Enzymes ( jstor )
Histones ( jstor )
In vitro fertilization ( jstor )
Incubation ( jstor )
Liver ( jstor )
Rats ( jstor )
RNA ( jstor )
Chromatin ( lcsh )
Dissertations, Academic -- Zoology -- UF
Histones ( lcsh )
Rats ( lcsh )
Zoology thesis Ph. D
bibliography ( marcgt )
non-fiction ( marcgt )


Thesis--University of Florida, 1970.
Bibliography: leaves 108-117.
Additional Physical Form:
Also available on World Wide Web
General Note:
Manuscript copy.
General Note:

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
000565951 ( AlephBibNum )
13608306 ( OCLC )
ACZ2375 ( NOTIS )


This item has the following downloads:

Full Text







I would like to thank the members of my committee for

their help during the progress of this research and the

preparation of this dissertation. I would especially like

to thank Dr. Paul Byvoet for his guidance and critical

evaluation of my research and preparation of this manu-

script. His continued interest in my progress and stimulat-

ing discussions havebeen an invaluable contribution to this

study. I thank Dr. John W. Brookbank, who has been a source

of encouragement and guidance in this work. Dr. E. Marshall

Johnson and Dr. Thomas W. O'Brien contributed greatly in

discussion throughout my graduate career and by critically

reading this manuscript. I would like to thank Dr. Steven

Zam and Dr. Ann Larkin for their suggestions and criticisms

in the preparation of this manuscript. To my friend, Ilse

Ortabasi, I extend many thanks for her help in introducing

me to the techniques of electrophoresis and photography,

and her critical reading of this manuscript. To al] the

faculty of Cellular and Molecular Biology at the University

of Florida 1 extend many thanks.



ACKNOWLEDGMENTS . . . . . . . . .. ii

LIST OF CHARTS . . . . . . . . . . vi

LIST OF TABLES . . . . . . . . . vii

LIST OF FIGURES . . . . . . . . .. ix

ABSTRACT . . . . ... . . . . . x

INTRODUCTION . . . . . . . . . . 1
Composition of the Intcrphase Nucleus . . . 3
Restriction of the Genome . . . . . . 5
Possible Mechanisms for Gene Control. . . . 6
Repetition of DNA sequences. . . . . 6
RNA polymerase and associated factors. . 6
Histones as repressors . . . . . 7
Nonhistone chromosomal proteins as
antagonists to DNA-histone interaction 9
Nonhistone chromosomal proteins as
antagonists to RNA polymerase-
histone interaction . . . . .. . 11
Arguments against histones as repressors . 11
Modification of histone structure. .. .... 13
Methylation of Histones . . . .. 13
Phosphorylation of Histones . . .. 14
Acetylation of Histones . . . .. 16
Rationale of Proposed Study . . . . .. 24

MATERIALS AND MELTHIODS. . . . . . . .. . 28
Isolation of Nuclei . . . . . . . 28
Isolation of Dcoxyribonucleoprotein (DNP) . .29
Conditions of Incubation . . . . . .. 30
Extraction of Histones and Determination of
Specific Radioactivity. . . . . . . 31
Determination of Acidic and Residual Proteins . 33




Methods for Extraction of Acetylating Enzymes
Extraction of acetylating enzyme from
acetone powder of rat liver nuclei .
Extraction of acetylating enzyme from
rat liver nuclei with saline . . .
Extraction of acetylating enzyme from
rat liver DNP . . . . . .
Method for Determining the Extent of O-
Acetylation of Free Histones and that
Occurring in DNP . . . . . .
Determination of Incorporation of Radioactive
Acetate into the Various Histone Fractions.
Fractionation according to Johns . .
Electrophoresis of histone fractions .

. 36

. 36

. 37

. 39

. 41

. 43
. 43
. 47

RESULTS . . . . . . . . . . .
Conditions Influencing the Acetylation of
Histones in DNP . . . . . . . .
Divalent cations . . . . . . .
Medium components . . . . . .
Temperature . . . . . . . .
Concentration of DNP . . . . . .
Concentration of acetyl-CoA and Km
determination . . . . . . .
Effect of pH in different buffer systems
Enzymatic Nature of the DNP Catalyzed
Acetylation of Histones . . . . . .
Inhibition studies . . . . . . .
Comparison of DNP catalyzed acetylation with
spontaneous acetylation of free histones
Buffer Effects . . . . . .
Temperature Effects . . . . .
Determination of the Extent of O-
Acetylation by Hydroxylamine Test
Enzyme isolation . . . . . . .
Acetylating Enzymes from Rat
Liver Nuclei . . . . . .
Acetone powder extracts. . . .
Saline extracts . . . .
Acetylating Enzymes from Rat
Liver DNP . . . . . . .
Relative acetylation of the various
histone fractions . . . . . .


DISCUSSION . . . . . . . . ... . . 84

ADDENDUM . . . . . . . . . . . 92
Comparison of Acetylating Activity in
Novikoff Hepatoma and Liver DNP .. . ... 92
Deacetylating Activity in Nuclear Extracts. . 95
Effects of Treatments Known to Increase RNA
Synthesis on the Acetylation of Histones
in DNP. . . . . . . .. . . 98
Purity of hydrocortisone preparations. . 99
Use of more samples . . . . . ... 100
Use of adrenalectomized rat. .... . .. 100
Preincubation of DNP with hydrocortisone 102
Media changes designed to increase the
solubility of DNP during the incubation. 102
Effect of in vivo administration of
hydrocortisone on the in vitro
acetylation of histones. . . . .. .105
Effect of phenobarbital and liver
regeneration on the in vitro
acetylation of histones in DNP . . .. 105

LIST OF REFERENCES . . . . . . . . .. .108

BIOGRAPHICAL SKETCH. . . . . . . ... . 118









I Effect of Various Medium Components on the
Transfer of Acetate from Acetyl-CoA to
Histones in DNP. .. . . . . . 55

II Effect of Increasing Concentrations of
Acetyl-CoA on the Transfer of Acetate from
Acetyl-CoA to Histones in DNP. . . .. .61

III Effect of Various Conditions on In Vitro
Transfer of Acetate from Acetyl-CoA
to Histones. . . . .... .. . . 67

IV Comparison of the Extent of O-Acetylation
in DNP with that Occurring Spontaneously
as Measured by Lability to Hydroxylamine. .73

V Restoration of Acetylating Activity in
Heated DNP by 0.35 M Extracts from Nuclei. 76

VI Acetylation of Isolated Histones with a
0.35 M NaC1 Enzyme Extract from Nuclei. . 77

VII Extraction of Histone Acetyltransferase
Activity from DNP. . . . .. . . 80

VIII Acetylation of Various Histone Fractions
in DNP . . . . . .. . . 82

IX In Vitro Uptake of Acetate into Histones:
Comparison of Liver and Novikoff Hepatoma. 94

X Deacetylation Activity in Nuclear Sap
Extracts . . . . .... . . .. 97

XI Effect of Increasing Concentrations of
Hydrocortisone on In Vitro Acetylation
of Histones in DNP . . . . . .. 101





XII Effect of Hydrocortisone on In Vitro
Acetylation of Histones in Liver DNP
from an Adrenalectomized Rat . . . .. 101

XIII Effect of Preincubation of DNP with
Hydrocortisone on In Vitro Acetylation of
Histones . . . . . . . ... .103

XIV Effect of Increasing the Solubility of
DNP on In Vitro Acetylation of Histones
in the Presence of Hydrocortisone. . .... 104

XV Effect of In Vivo Administration of Hydro-
cortisone on In Vitro Acetylation of His-
tones in DNP . . . . . . ... 106

XVI Effect of Phenobarbital and Liver Re-
generation on In Vitro Acetylation of
Histones in DNP. . . . . .. . . 106





1 Effect of Temperature on the Transfer
of Acetate from Acetyl-CoA to Histones
in DNP .. . . . . . . . . 57

2 Effect of the Concentration of DNP in the
Incubation Mixture on the In Vitro Transfer
of Acetate from Acetyl-CoA to Histones . 60

3 Effect of Increasing Concentrations of
Acetyl-CoA on the In Vitro Transfer of
Acetate from Acetyl-CoA to Histones
in DNP Plotted According to Lineweaver
and Burk . . . . . . . ... 62

4 Effect of pH in Different Buffer Systems
on the Transfer of Acetate from Acetyl-
CoA to Histones in DNP . . . . .. 63

5 Comparison of the Effect of Increasing pH
in Different Buffer Systems on the Spon-
taneous and Enzymatic Acetylation
Reactions. . . . . . . . .. .71

6 Effect of Temperature on the Transfer of
Acetate from Acetyl-CoA to Histones in
Rat Liver DNP and to Free Histones in
Different Buffer Systems . . . 72

7 Acrylamide Gel Patterns Obtained with
Various Histone Fractions. . . . .. .83

Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment
of the Requirements for the Degree of Doctor of Philosophy



Louise Adele Racey

December, 1970

Chairman: John W. Brookbank
Co-Chairman: Paul Byvoet
Major Department: Zoology

Rat liver chromatin or deoxyribonucleoprotein (DNP) was

found to exhibit acetyltransferase activity in vitro, which

seems to be closely associated with it. Evidence is pro-

vided which indicates that this transfer of acetate from

acetyl-CoA to lysine in "arginine-rich" histones represents

an enzymatic reaction. A number of comparisons have re-

vealed that the in vitro reaction occurring within DNP is

much less random than the spontaneous acetylation of free

histones in solution in the presence of acetyl-CoA. Reports

from other laboratories have indicated that a similar enzyme

could be extracted from an acetone powder of either whole

liver or nuclei. The study reported seems to involve a

different enzyme since no acetyltransferase activity can be

obtained from an acetone powder of chromatin using these

methods. A procedure similar to that used for the extraction

of mammalian RNA polymerase has been successfully applied to

the extraction of acetyltransferase activity from rat liver

chromatin, although even in this case, about half of the

activity remains bound to the DNP complex. The extraction

of acetyltransferase activity from chromatin provides

strong support for the assumption that the described in

vitro reaction is enzymatic.


The mechanism of gene control remains today as one of

the major unexplained areas of biology. It is obvious from

an analysis of living systems that some form of control

must exist. One example is the circadian rhythm of liver
enzyme activity exhibited by rats and other mammals.

Although oscillations in the activity of some enzymes may

be explained by enzyme stimulation or inhibition, many

enzymes are subject to induction and repression at the

level of the genome. In this case the production of a

new enzyme is dependent on deoxyribonucleic acid (DNA)

directed ribonucleic acid (RNA) synthesis as an initial

step. The mechanism by which the control of the expression

of certain regions of the genome occurs in eukaryotic cells

is largely unknown. In prokaryotic cells a model system has

been proposed involving repressor proteins which interact

with DNA, thereby influencing its transcription by RNA

polymerase.2-6 Small effector molecules interact with the

repressor protein, changing its configuration in such a way

as to influence its interaction with DNA and thereby the

transcriptive process. These effector molecules can be

metabolites or other substances capable of allowing the cell

to become responsive and adaptable to changes in its environ-

ment. It is not known whether a similar mechanism is opera-

tive in eukaryotic cells.

The differentiation process undergone by eukaryotic

systems is another area where gene control is operative.

Here a totipotent cell produces offspring of a particular

type, the characteristics of which are maintained from one

generation to another. Since within an organism each cell

type contains the same amount of DNA and hybridization

studies have shown the DNA of each cell type to be similar
to that of every other type, the question arises as to

how the cells acquire and maintain the selective expression

of specific genetic information from one generation to

another. The eukaryotic system is therefore more compli-

cated than the prokaryotic cell in that it is not only

capable of reversible changes in gene expression, such as

occurs in enzyme induction, but also undergoes more per-

manent changes which are relatively constant from generation

to generation. That these changes are not completely ir-

reversible has been shown by nuclear transplantation studies

wherein nuclei from the intestinal cells of Xenopus appeared

to be capable of supporting the development of a complete

embryo from an enucleated egg. However, Briggs and King.

have demonstrated that nuclei taken from frog embryo cells

at successive stages of development are progressively less
able to support normal development. Although the total

complement of genetic information is present, the expression

of this information is restricted by its passage through

developing cytoplasm.

Composition of the Interphase Nucleus

Attempts to explain the mechanism of genetic control

must take into consideration the structure of the nucleus.

In the eukaryotic cell the nucleus is a defined area sur-

rounded by a semipermeable membrane. This membrane in-

fluences a dynamic interchange of materials between the

nucleus and the cytoplasm. Within the nucleus is contained

DNA and its associated proteins. Nuclear sap proteins can

be removed by isotonic saline washes whereas other proteins

along with the DNA form part of the insoluble chromatin
complex. In the interphase nucleus part of the chromatin

has a diffuse appearance and consists of many fibers. In

some regions the chromatin appears in a more condensed con-

dition. Condensed regions of a single chromosome are called

heterochromatin, as opposed to the less condensed or euchro--

matic regions. The term "heterochromatin" is also used to

refer to a condensed state of chromatin, and is applied to

such regions of the interphase nucleus as well.14


The components of the chromatin include DNA and histone
15, 16
proteins, found in approximately equivalent amounts.15, 16

Nonhistone proteins comprise about 27 percent of the DNA
associated protein in calf thymus, and 43 percent in

liver. These proteins are subdivided into acidic pro-

teins, which are alkali soluble, and residual proteins,

which remain behind after acidic extraction of histones and
basic extraction of acidic proteins. RNA polymerase and

other enzymes are among those which make up the nonhistone
20, 21
chromosomal proteins. Also some species of RNA have

been found associated with the chromatin.22, 23

The arrangement of the chromosome has been a subject

of considerable controversy among cytologists. Interphase

chromatin fibers of 200-300 Ao in diameter have been de-

scribed and Ris has proposed that these consist of two 100

A units, and that each of these further consist of two

40 A units representing single DNA-protein molecules.24, 25

This has been questioned and more recent data utilizing

the techniques of trypsin digestion of chromatin fibers in

conjunction with electron microscopy shows that each 230 Ao

fiber consists of a single DNA molecule packed in a pro-

teinaceous sheath.2 27 According to the model this unit

is further induced to supercoil by the addition of more

protein.28 Different chromosomal proteins may play specific

roles in the maintenance of this structure.

Restriction of the Genome

The need for selective gone expression implies that

certain regions of the chromsome will be active at a given

time whereas others will remain quiescent. Evidence that

this is the case comes from hybridization studies. Georgiev

observed that RNA synthesized on free DNA hybridized to a

much greater extent with DNA than that synthesized on chroma-
tin.9 He explained this by the observation that there are

many repeated DNA sequences in the chromatin which are re-

stricted in transcription. However, RNA synthesized on

chromatin competes with in vivo synthesized messenger RNA

(mRNA) for sites on DNA more efficiently than does RNA

synthesized on free DNA. This implies that RNA synthesized

on a chromatin template more closely approximates the in

vivo situation.

Heterochromatic or tightly condensed regions of chroma-

tin are considered to be in the repressed condition, whereas

the more diffuse euchromatic regions are believed to be

metabolically active. Electron microscope autoradiography

has revealed that the diffuse, presumable euchromatic,

regions actively incorporate 3H-uridine in contrast to the
heterochromatic regions. Frenster has isolated bulk

fractions of condensed and diffuse chromatin from interphase
calf thymus lymphocytes. He has shown that the condensed

regions contain up to 80 percent of the nuclear DNA, but

only 14 percent of the nuclear RNA. This observation would

seem to support the concept that the condensed state of the

chromatin is associated with repression of genetic expression.

Possible Mechanisms for Gene Control

Repetition of DNA sequences

One finding which has recently attracted attention is

that more than one third of the DNA of higher organisms is

made up of sequences which recur anywhere from a 1 x 103
6 32
1 x 10 times per cell. The nucleolus of amphibian

oocytes contains many replicas of the genes which code for
33, 34
ribosomal RNA. This repetition of specific genes has

been viewed as a unique means of controlling the expression

of genetic information by allowing for the production of

certain products in greater quantity at a given time in

the life of the cell.

RNA polymerase and associated factors

In considering the ways in which the regulation of gene

expression may occur, RNA polymerase and its associated

sigma factor must be mentioned. In bacterial systems the

sigma factor has been found to be responsible for the initi-

action of RNA chains by DNA-dependent RNA polymerase. RNA

synthesis in isolated nuclei of avian orthryocytes was

stimulated by a heat stable factor present in extracts of

HeLa cells. More recently, a protein-like factor from

calf thymus with a sedimentation constant of 3 S has been
described. This factor stimulated DNA-dependent RNA

synthesis from the same tissue after the enzyme had already

become bound to DNA. These findings have a bearing on ex-

planations of gene control in that specificity of initiation

site and rate of RNA synthesis are influenced by these


Histones as repressors

In 1951 the Stedmans suggested that histones may play

a role in the regulation of gene activity by acting as
repressors of RNA synthesis. The observation by Huang

and Bonner that histones suppress RNA polymerase activity

when added to DNA in vitro prompted an increased investiga-
tion into this area. The relevancy of these findings has

been questioned and it has been suggested that the inhibi-

tion of RNA polymerase activity is due simply to the pre-
cipitation and removal of DNA from solution. Bonner and

Huang have answered this objection by describing conditions
which prevent DNA precipitation, and Butler and Chipper-

field showed that the amount of inhibition continues to

increase with increasing histone concentrations even after
the DNA has been fully precipitated. Clark and Byvoet

attempted to resolve the question of whether or not a corre-

lation exists between the activity and the solubility of

the DNA template by plotting the logarithm of the percent

inhibition of template activity against the amount of un-

aggregated DNA present in the reaction mixtures at various
histone/DNA ratios. This type of plot revealed a close

correlation between the solubility of the DNA template and

the inhibition of polymerase activity by histone. In any

case, it seems likely that reconstitution of the DNA-

histone complex by simple titration of DNA with histones

preserves very little of the specificity of the original

relationship and there is a question as to whether or not

these in vitro findings have any relation to the in vivo


Further support for the involvement of histones in

repression of RNA synthesis came from studies which

selectively removed different histone fractions from chro-

matin. Successive extractions of pea-bud chromatin with
NaCl solutions resulted in increased template activity.

Extraction of histones from the heterochromatin of the male

mealy bug was shown to derepress RNA synthesis and to in-
crease actinomycin D binding. When all the histone was

removed from the heterochromatin it became equivalent to the

euchromatin in these respects.

Nonhistone chromosomal proteins as
antagonists to DNA-histone interaction

It has been suggested that nonhistone proteins act as

antagonists to the DNA-histone interaction which serves to
repress RNA synthesis. In support of this, it has been

found that basic proteins, such as histones, are closely

associated with DNA, occurring in highest concentrations
where DNA appears to be tightly coiled and condensed.

During puff formation in dipteran giant salivary chromo-

somes, i.e., at regions of intense RNA synthesis, there is

no measurable difference in the amount of stainable basic

protein, implying a continuity of structure during the

transcriptive process. Interestingly, however, there seems

to be an increase in the acidic proteins in these regions
during puff formation. In isolated chromatin fractions the

ratio of total histones to DNA did not significantly differ

within either the repressed or active fractions, and the

relative proportions of each of the distinct types of his-

tones were similar in both forms of chromatin. However,

when nuclear polyanion contents of active and repressed

chromatin fractions were determined relative to the DNA

contents, the active fractions were found to contain more

total nonhistone proteins, RNA and phospholipids, and phos-
phoprotein phosphorous. About 15 percent of the non-
histone proteins consist of such nuclear phosphoproteins.

In vitro experiments have shown that acidic proteins

complex with free histones and prevent histone inhibition
of RNA synthesis by bacterial RPIA polymerase. However,

these proteins cannot dissociate the histone-DNA complex

to reverse histone inhibition of RNA synthesis. Therefore,

Spelsberg and Hnilica have suggested that their action may

reside in the prevention of DNA-histone interaction rather

than the dissociation of DNA--histone complexes.56

It has been shown by Paul and Gilmour that all the

protein components of the chromosome are necessary for the
production of RNA which resembles that found in vivo.

Using hybridization techniques they found that histone re-

pression of DNA directed RNA synthesis in vitro is non-

specific. However, if they added the residual and acidic

proteins back to DNA and histone, the new mRNA synthesized

does compete effectively with mRNA produced in vivo. Similar

experiments by Bekhor, Kung, and Bonner showed that DNA,

histones and chromosomal RNA produced a primer capable of
yielding mRNA equivalent to the natural product. They

therefore suggested that nonhistone protein is not re-

pressive. Marushige, Brutlag, and Bonner, in contrast to

Paul and Gilmour, found that DNA saturated with nonhistone
chromosomal proteins was as good a primer as naked DNA.

These differences may resJlt from contaminated preparations

and have yet to be resolved. It is difficult to imagine

a mechanism for gene control in which any one of the ele-

ments of the chromosome does not play a role, either directly

or indirectly. The observations of these interactions is,

therefore, limited probably by experimental technique.

Nonhistone chromosomal proteins as
antagonists to RNA polymerase-histone interaction

Association of chromosomal proteins with RNA polymerase

is another means by which repression of transcription could

take place. Histones form complexes with bacterial and

mammalian RNA polymerase leading to the inhibition of in
vitro RNA synthesis. Since it has been observed that

salt and polyanions (e.g., acidic nuclear proteins) acti-
63, 64
vate RNA polymerase in chromatin, it has been suggested

that this may be due to dissociation of enzyme-histone com-
plexes, thereby stimulating RNA synthesis. This sugges-

tion is supported by the observation that the ionic strength

required for maximum dissociation of the enzyme-histone

complex is within the range required for the activation of

endogenous RNA polymerase. In contrast, no detectable dis-

sociation of DNA-histone complexes was seen at these ionic


Arguments against histones as repressors

One of the strongest criticisms against the idea that

histones function as genetic repressors is the fact that

histones lack specificity. There are five main histone

fractions: very lysine-rich (fl), moderately lysine-rich

(f 2), and the arginine-rich (f f and f ) histones.*

Although there are major differences between the five his-
tone fractions, there is very little difference in the

composition of similar fractions from different species.

Sequence studies of an arginine-fraction (f2al) from organ-

isms as far removed from each other as calf (thymus) and

pea seeds revealed essentially identical primary struc-
tures. Recent data on very lysine-rich (fl) histones

showed that they are more numerous and have some phenotypic
specificity, but not enough to account for the observed

phenotypic differences. It has been suggested, therefore,

that chromosomal RNA plays a role in gene repression and
derepression. This is an attractive idea since it offers

a means for specific recognition of DNA sites for selected

repression of transcription. In support of this, Huang and

Older terminology referred to just the "arginine-rich"
histones, which included the moderately lysine-rich (f2b)
fraction, and the "lysine-rich" histones, which only
referred to the very lysine-rich (fl) fraction. Hereafter,
"arginine-rich" enclosed in quotation marks refers to the
arginine-rich (f~-al- f2' and f3) and moderately lysine-
rich (f2b) fractions confined.

Bonner have found a chromosomal RNA bound to a nonhistone
22, 23
protein,22, 23 which they have postulated is associated in
complexes with histones.

Another major objection to the simple repression of

transcription by histone binding to DNA comes from the ob-

servations that histones turn over at the same rate as
DNA, are synthesized at about the same time of the
cell cycle as DNA, ,and are present at the same levels

within the chromosome during RNA synthesis.1 These obser-

vations mediate against a simple explanation of histones

acting alone as gene repressors simply by combination with

and dissociation from DNA. It is for these reasons that

attention has focused on structural modifications of his-

tones, taking place after their synthesis, which may influence

their binding to DNA, and thereby perhaps the transcriptive

Modification of histone structure

Three methods of histone modification are known. These

are methylation of lysine and histidine residues,

71, 74, 93, 94
acetylation of lysine residues, and phos-
phorylation of serine residues.

Methylation of Hlistones

Isolated nuclei from calf thymus and chromatin from
96-98 90
peas, rat liver, and Ehrlich ascites cells can

mcthylate histones in vitro. Methionine is the methyl
donor via S-adenosyl-methionine. Gershey at al have found

that the methylation of histones occurs much more rapidly

in immature erythroid cells, but as the nucleus matures and

loses its biosynthetic capacities, it also loses its capacity
to methylate histidine in its histones. Tidwell et al.

have compared the time courses of histone synthesis and
methylation during regeneration of the liver in the rat.

They found that maximal histone methylation occurs at a time

when the rates of histone and DNA synthesis have already

begun to decline and does not correlate with an increase in

RNA or nonhistone protein synthesis. They suggest, there-

fore, that methylation is not involved in gene activation

for RNA synthesis.

Phosphorylation of Histones

The phosphorylation of histones occurs both in vivo

and in vitro. Ord and Stocken have found a 2-fold increase

in the rate of phosphorylation and phosphate content of

very lysine-rich (fl) histone of regenerating rat liver near
the first peak of DNA synthesis after partial hepatectomy.

Langan and Smith have isolated an enzyme from liver which

specifically transfers phosphate from adenosine triphosphate
(ATP) to shrine residues of protamines and histones.

The most rapidly phosphorylated histones are the lysine-rich

fractions, fl and f It appears that there is more than

one histone kinase and that they differ in their rate of
phosphorylation of f and f histones. A phosphatase
1 2b

has also been partially purified which is specific for
phosphorylated histones and protamine.

Since phosphorylation by histone kinase is stimulated

by adenosine-3', 5'-monophosphate (cyclic AMP) Langan has

suggested a mechanism for the induction of RNA synthesis

by those hormones that cause increases in the concentration
of cyclic AMP. The administration of glucagon or insulin

to rats causes an in vivo increase in phosphorylation of

lysine-rich histones in the liver involving the same specific
serine residue that becomes phosphorylated in vitro.

The phosphorylation of chromosomal proteins is not

restricted to histones. Kleinsmith et al.have observed an
increase in P-phosphate incorporation into nuclear phos-

phoproteins of lymphocytes stimulated to grow and divide by
phytohemagglutinin. This increase was particularly evi-

dent in nonhistone protein and occurs prior to an increase

in the rate of RNA synthesis occurring under these con-

ditions. Gershey and Kleinsmith have followed changes in

nuclear phosphoproteins during the course of development of
the avian erythrocyte. As maturation proceeds, the

nuclear levels of both phosphoprotein kinase and protein-

bound phosphorous correlate with a decrease in the bio-

synthetic activities of the nucleus. The bulk of the

incorporation of phosphate occurs as in the case of the

lymphocyte in the nonhistone protein fraction. Since

Langan has reported that phosphoproteins can interact with

histones in vitro, and that complex formation with phos-

phoprotein decreased the inhibitory effects of added histone

on DNA dependent RNA synthesis, it has been suggested that

this interaction may play a role in the regulation of gene
expression. Further support for this idea comes from

studies on isolated chromatin fractions, which revealed that

phosphoprotein concentrations are highest in fractions which
are active in RNA synthesis. Dipteran salivary gland

chromosomes also show active phosphorylation of nonhistone
proteins, particularly in the puff regions.1

Acetylation of Histones

The acetylation of histones has been studied in intact
109, 110 111-113
animals, in cells in tissue culture, in iso-
94, 96, 114-116
lated nuclei, and in isolated enzyme prep-
arations. Early studies measured the incorporation
of C-acetate, but it was later found that acetyl Coenzyme
A (acetyl-CoA) is the direct acetyl donor. In every case

examined the incorporation of labeled acetate occurred pri-

marily in the arginine-rich (f3' f2al and f2a2) fractions,


while very lysine-rich (fl) and moderately lysine-rich (f2b)

fractions incorporated minor amounts. This is in contrast

to the predominant phosphorylation of the lysine-rich frac-
tions. The low level of acetate incorporation into thymus

fl histone also contrasts with the finding that this fraction

is almost completely acetylated at its amino terminal

120, 121
end.120 121 These alpha-N-acetyl groups appear to be rela-

tively stable, whereas the internal epsilon-N-acetyl groups in

the f2al and f3 fractions are involved in a dynamic process

of acetylation and deacetylation. Evidence for the fact that

acetyl groups are attached to histones after synthesis of the

polypeptide chain is provided by the finding that puromycin

in concentrations sufficient to block protein synthesis in

calf thymus nuclei did not inhibit the acetylation of his-
tones. In agreement with this, Byvoet has observed that

the turnover of acetyl groups in histones is much more rapid

than the turnover of histones themselves and the ratio of the
two turnover rates vary with the tissues.81 Gershey et al.

have prepared histone fractions from calf thymus nuclei which
14 93
had been incubated in the presence of sodium- C-acetate.

These were digested with trypsin and pronase and the result-

ing peptides and amino acids were separated by chromatog-

raphy. Only the arginine-rich fractions f2al and f3 were

appreciably labeled and the radioactivity was associated with


epsilon-N-acetyl-lysine. Studies on the sequence of f2al his-

tones have shown that 50 percent of the lysine residues at
position 16 are acetylated in calf thymus, whereas in pea
bud the epsilon-N-acetyl content is only 6 percent.

Enzyme fractions which acetylate histones have been iso-
lated front an acetone powder of whole pigeon liver. Two

fractions were found, of which one activates acetate by con-

verting it in the presence of ATP, Mg and coenzyme A to

acetyl-CoA, while the other transfers the acetate from

acetyl-CoA to histones. Using similar procedures Gallwitz

reported the isolation of acetokinase from an acetone powder
of rat liver nuclei. This enzyme was capable of trans-

ferring acetate from acetyl-CoA to isolated histones and was

especially active for arginine-rich (f and f ) fractions.
3 2al
More recently Gallwitz reported the isolation of two

histone-specific transacetylases from rat liver nuclei
using a different approach. Rat liver nuclei were

sonicated in the presence of 1 M ammonium sulfate and 20

percent glycerol, followed by precipitation with 3.5 M

ammonium sulfate. A 50-fold purification was accomplished

by chromatography. No preference of the two enzymes has

been found for one specific histone fraction.

The isolation of a deaetylating enzyme was reported by
Inoue and Fujimoto from a salt extract of ca1 f thymus.

This enzyme showed a preference for histones which were
acetylated by incubating calf thymus nuclei with C-acetic

anhydride. Libby has also reported deacetylating activity

in nuclei of rat liver and Novikoff hepatoma.1

Studies by Allfrey, Mirsky, and their colleagues have

attempted to correlate histone acetylation and chromosomal

activity. Arginine-rich histones from calf thymus inhibited

RNA polymerase in an "aggregate enzyme" from calf thymus as

124, 125
well as purified RNA polymerase from bacterial sources.

In the case of the mammalian RNA polymerase studies, the

problem was complicated by difficulties in the purification

of the enzyme because it remained bound to the chromatin
complex. When external histones are added to this com-

plex, inhibition of RNA synthesis is observed, but the

significance of this observation is questionable and it may

represent a nonspecific effect. When RNA polymerase from

Escherichia coli was used, this problem was circumvented as

it can be obtained in purified form. In vitro RNA synthesis
by E. Coli RNA polymerase was followed using 1C-ATP (ade-
nosine triphosphate) or C-UTP (uridine triphosphate) as

precursors. It was found that the incorporation into RNA

was decreased by the addition of arginine-rich histone.

This degree of inhibition of RNA synthesis was progressively

decreased experimentally by the chemical addition of acetyl

groups to the isolated histones. The technique used to

acetylate the histones for the inhibition of RNA snythesis

was selected to give maximum acetylation of alpha amino

groups and minimum acetylation of epsilon amino groups of

lysine. It has subsequently been found that in vivo ace-

tylation involves primarily the epsilon amino group of ly-

sine. Therefore, it would be interesting to study this

effect using histones which have been acetylated in a more

physiological manner.

Histone acetylation has been compared in chromatin

fractions which are active or inactive in RNA synthesis.

Fractionation of thymus chromatin after labeling nuclei in

vitro with isotopic RNA precursors has shown that RNA syn-

thesis proceeds faster in the diffuse chromatin than in the
30, 31
condensed chromatin. Comparisons of the rates of

acetate incorporation into the histones of each of these

fractions show that histone acetylation, like that of RNA

synthesis, proceeded 2-3 times more rapidly in the diffuse

chromatin fraction. Autoradiography of calf thymus nuclei

after 3H-acetate incorporation indicates that acetylation

occurs in both diffuse and condensed regions but is par-

ticularly evident at the boundaries between the two re-
gions. This would be expected if acetylation is associ-

ated with the transition of chromatin from a condensed to a

more diffuse, active state.


Since puffing in dipteran polytene chromosomes has been

associated with RNA synthesis. attempts to correlate this
with acetylation of histones have been made. Autoradio-

graphic studies indicated uptake of radioactivity in chromo-

somes when salivary glands of Chironomus thummi were incu-

bated in the presence of 3H-acetate. The grain densities

over the puff regions however were not particularly intense

and Allfrey has suggested that the methods used in the fix-

ation process extract most of the acetylated arginine-rich

histones from the chromatin. However, using new methods of

fixation, which retain incorporated acetate, Clever and

Ellga1rd also failed to observe any accumulation of label

over puff regions and it is their opinion that puff forma-
tion does not include acetylation of histones.

The mature nucleated red cells of birds synthesize

little or no RNA, whereas lymphocytes do. Comparisons have

been made of the rates of histone acetylation and RNA syn-

thesis in nucleated erythrocytes and in isolated lymphocyte

nuclei. These reveal that the lymphocyte nuclei were far

more active in acetylating their histones than were the
erythrocytes. Although this may be coincidental, it is

further evidence for the hypothesis that acetylation of

histones may be associated with gene activation.

Comparison of the rates of histone acetylation and RNA

synthesis in lymphocytes at different times after the addi-

tion of phytohemagglutinin (PILA) to Lhe culture medium shows

that RNA synthesis is stimulated and that there is an increase

in acetylation of arginine-rich histones preceding the in-

crease in RNA synthetic activity. In contrast to the be-

havior of lymphocytes upon exposure to PIIA, polymorphonu-

clear leucocytes curtail the synthesis of RNi- and under

these conditions histone acetylation is also depressed.111 112

Monjardino and MacGillivray have suggested, however, that

these effects may be nonspecific since some preparations of

PHA were found to increase RNA synthesis in lymphocytes
while decreasing histone acetylation. Ki.llander and

Rigler have shown that the amount of acridinc orange dye

bound to the chromatin of PHA stimulated lymphocytes in-

creased rapidly over a time course which is similar to that
for histone acetylation. This reflects a change in

chromatin molecular structure as a result of PHA treatment,

which as suggested by Allfrey may be initiated by the acety-
lation reaction.

It has been shown that in liver cells gone activation

occurs with consequent appearance of new species of RNA as
a result of partial hepatectomy.1 Increa:;s in DNA tem-
plate activity of isolated liver nuclei, -and increases

in Lemplate activity and RNA polymerase activity in chro-

martin and in "aggregate" enzyme preparations have been re-
44, 133 135
ported. With this in mind Allfrey and others

have studied the acetylation of histones during the course
110, r1_26
of liver regeneration. The specific activity of

different hisLone fractions from control and regenerating

rat liver were measured at various time intervals after an
in vivo injection of H-acetate. The results in sham-

operated controls indicated a high rate of acetate uptake,

the maximum specific activity being reached in 15 minutes.

Turnover at 60 minutes wcs such that only one third of the

acetyl-groups originally incorporated into the arginine-rich

fraction v re left. This pattern was altered for regenerating

Jiver. In this case histone acetylation was increased by

300 percent at 3 hours after partial hepatectomy and in the

period between 1-2 hours after the operation the histones

lost 13.7 percent of their original acetyl content while

the controls had lost 70 percent. Allfrey has suggested

that this is due to an increase in the rate of acetylation

and lower rates of deacetylation for regenerating liver.

The increase in the percent retention of the acetyl groups

in histones from regenerating liver just precedes a rise in

RNA polymerase activity in regenerating rat liver nuclei

and reaches its peak 2 hours before the nuclei reach the

first plateau in PRNA polymerase activity. These findings

are consistent with the view that acetylation of histones

modifies DNA-histone interactions, and the subsequent changes

influence the template activity of the chromatin for RNA


A correlation has also been observed between the pat-

terns of RNA synthesis and histone acetylation in liver re-
109, 126
spending to stimulation by steroid hormones. Ad-

ministration of cortisol to adrenalectomized rats leads to

increases in the amounts and changes in the types of RNA

synthesized. As in the case of liver regeneration, cortisol

treatment of adrenalectomized rats leads to early increases

in the rate of acetylation of arginine-rich histones and a

suppression of turnover of previously incorporated acetyl

groups. Cortisol stimulation of the liver is only one

example of hormone-induced gene activation. Takaku et al.

observed increases in the acetylation of histones in spleen

cells of polycythemic mice just preceding an increase in

RNA synthesis at 4 and 8 hours after erythropoietin injec-
tion. Another example is provided by estradiol which in-
creases RNA synthesis in the uterus. Estradiol has

also been reported to stimulate the acetylation of histones
by ce]l free extracts of the rat uterus.

Rationale of Proposed Study

Although none of these observations show a direct cause

and effect relationship, they suggest that the acetylation

of histones may be related to genetic expression. The fact

that only two of the five known histone fractions are acety-

lated upon incubation of nuclei in the presence of radio-

active acetate suggests that this is an extremely specific

reaction. The complete elucidation of the amino acid se--

quence of the fl histone has even established that in the
entire molecule only one specific lysine residue is acety-
71, 74
lated at the epsilon amino position. It does not seem

likely that the acetylation of only one lysine residue would

change the structure of the histone sufficiently to influ-

ence gene activity. However, the location oE this lysine

residue within an unusual cluster of five basic residues

would seem rather suggestive of some role which this acety-

lation may play in histone-DNA interaction.

It appeared that the biological implications of the

postulated changes in histone-DNA interactions, resulting

from histone acetylation, are sufficiently important to

warrant a thorough study of this process at Lhe molecular

level. In order to carry out such a study, the development

of an in vitro system is essential in view of the many ad-

vantages of such a system over the myriad of possible in-

direct effects which may influence the outcome of in vivo

studies. Some of these may be hormonal influences, pool

sizes, and degradation of labeled precursor. Even in the

case of isolated nuclei there is a possibility of cytoplas-

mic contamination. In addition, nuclei contain an enzyme

which deacetylates histones (see Addendum). The development

of the in vitro system described in this study which utilizes

deoxyribonucleoprotein (DNP) has eliminated many of these

disadvantages and possesses all the advantages of in vitro

systems in general. The most important of these is the

isolation of the enzyme to be studied from the many vari-

ables which cannot be controlled in vivo. Other advantages

are greater reproducibility, control of medium composition,

temperature, and pH. In the first stages of the development

of this system, it appeared that DNP could be employed in-

stead of isolated nuclei, especially if the most immediate

precursor, acetyl-CoA was used. It was found that isolated

rat liver DNP was capable of transferring acetate from

acetyl-CoA to histones. A problem which presented itself

at this point was the possibility that this transfer might

not be enzymatically catalyzed since it was found that free

histones are spontaneously acetylated upon incubation with

acetyl-CoA. It, therefore, became essential to establish

that this transfer is an enzymatically catalyzed reaction

and attention has been directed toward this problem. This

dissertation describes the conditions influencing the trans-


fer of acetate from acetyl-CoA to histones in rat liver

chromatin and provides evidence that this is an enzymatic



Isolation of Nuclei

Although in the beginning these studies were done with

calf thymus, this was replaced by rat liver since fresh

preparations were found to exhibit higher acetylating

activities in contrast to frozen material. Other factors

in favor of rat liver include availability of previously

described methods for the isolation and purification of

nuclei and deoxyribonucleoprotein (DNP). It was also

desirable to compare experiments on Novikoff hepatoma and

the tissue from which the tumor presumably originated.

Male Holtzman rats weighing approximately 300-500 gm

were decapitated and the liver excised. All preparations

were carried out at 40C. The liver was homogenized in 10

volumes (weight/volume) 0.25 M sucrose containing 1.5 mM

CaC12 in a Potter type homogenizer, filtered through cheese-

cloth to remove connective tissue, and centrifuged at 600 x
g for 10 minutes. The pellet was resuspended in 5

volumes of 0.25 M sucrose, 1.5 mM CaCl2 and recentifuged

as above. The nuclear fraction thus obtained was purified

by resuspension and centrifugation at 40,000 x g for 30
minutes in 2.1 M sucrose, 0.5 mM CaCl2.

It has been found that this procedure removes surface

cytoplasm from rat liver nuclei even when loose homogenizers
were employed. Examination of the 2.1 M sucrose, 0.5 mM

CaCl2 pellet by phase microscopy showed clean nuclei with

little visible debris in the preparation. Nuclei have also

been prepared according to more recent procedures using

detergent, Triton N-101, which has been shown by electron

microscopy to remove the outer nuclear membrane and perinu-
clear ribosomes.

Where calf thymus was used, the 2.1 M sucrose purifi-

cation of nuclei was omitted because there is less of a

problem of cytoplasmic contamination with this tissue. If

Novikoff hepatoma was used, fibrous connective tissue was

removed by filtration through a wire screen rather than


Isolation of Deoxyribonucleoprotein (DNP)

DNP was prepared from nuclei by homogenization in 10-

15 volumes of 0.14 M NaC1, 0.01 M sodium citrate in a micro

Waring Blendor (Virtis), for 1 minute followed by centri-

fugation at 2000 x g for 10 minutes. The sediment was re-

suspended by light homogenization in a loose fitting Potter

type homogenizer and recentrifuged. This procedure, which

removes soluble nuclear proteins, was repeated once.

Chart 1 illustrates the outline of the general procedure


Conditions of Incubation

Unless otherwise stated, samples of isolated DNP or

purified nuclei containing about 0.3-0.5 mg of histones were

incubated at 37 C in a medium originally designed for studies
of protein synthesis by isolated calf thymus nuclei.

This medium contained 0.19 M sucrose, 20 mM glucose, 25 nmM4

phosphate buffer, pll 6.75, 12 mM NaC1, 0.75 mM Ca+, 5 mM

++ 14
Mg and 0.01 pic acctyl- C-CoA (spec. act. 56 mc/mM, New Eng-

land Nuclear), in a final volume of 2 ml (step 3 in Chart

1). In preliminary experiments H-acetate was used as a

precursor, but replaced by acetyl-14C-CoA since it was
found that this improved incorporation of C-acetate. Low

incorporation of acetate into histones when 3H-acetate was

used as a precursor may be due to lack of activating enzymes

converting acetate to acetyl-CoA (see Introduction), es-

pecially since there was no stimulation of H-acetate in-

corporation into histones upon the addition of ATP. Studies

reported below revealed that divalent cations, glucose, and

sucrose were not necessary for the transfer of acetate from

acety].-CoA to histones. They were therefore eliminated from

the medium. Lat--.r, experiments were also run at a pH of 8,

rather than 6.75.


After a 15 minute incubation, the reaction was stopped

by cooling the flasks and adding trichloroacetic acid (TCA)

to 5%. The resulting mixture was centrifuged at 2000 x g

for 10 minutes (step 4 in Chart 1).

If nuclei were incubated in the above described manner,

the reaction was stopped after 10 minutes by rapid cooling

and centrifugaLion of the samples at 2000 x g for 10 min-

utes. The nuclei were then washed in saline-citrate to re-

move soluble nucleoproteins as described for the preparation

of DNP.

In experiments where proteinaceous extracts were added

to DNP preparations, the reaction was not stopped by TCA,

but only by centrifugation after cooling, to prevent co-

precipitation of the added proteins with the DNP upon ad-

dition of TCA. These proteins could contaminate subse-

quently extracted histones.

Extraction of Histones and Determination
of Specific Radioactivity

After incubation, the total histones were extracted

from the DNP sediment with 0.25 N HC1 (step 5 of Chart 1).

Chromatographic analysis and subsequent amino acid analysis

of histones extracted in this manner have shown that less

than 0.5% of the total nitrogen is due to contamination

145, 146
from other proteins. After centrifugation the

supernatant was made 5% with TCA which precipitates the

"arginine-rich" (includes f2b fraction) histones leaving

the very lysine-rich h.istones behind in solution. After

standing overnight the TCA precipitate was washed with ace-

tone-l% HCl and ether and allowed to dry (step 6 in Chart 1).

This procedure, therefore, selectively isolates the "ar-

ginine-rich" histones, which have been found to be by far

the most actively acetylated fraction (see Introduction).

Determinations of total counts contained in histones,

DNA, acidic and residual proteins (described in next sec-

tion) show that 76% of the radioactive label is contained in

the "arginine-rich" histone fraction and therefore cannot be

ascribed to contamination with a hypothetical nonhistone

fraction which is very highly labeled. Almost all of the

remaining label was contained in the acidic protein frac-

tion. Purity of histone fractions was tested by amino acid

analysis and acrylamide gel electrophoresis of "arginine-

rich" histones isolated according to this procedure.

In experiments where the spontaneous acetylation of

free histones was studied, "arginine-rich" calf thymus

histories were used. Essentially, these histones are com-

parable to those obtainable from rat liver,1 but were

used in place of rat liver histones because larger quan-

tities could be obtained more easily from calf thymus.

These histones were extracted from 0.14 M NaCI washed calf

thymus with 0.25 N HCl and the "arginine-rich" fractions

were precipitated by 5% TCA and dried in acetone-l% HCl

and ether.

Protein concentrations were determined by the method of
Lowry et al. using bovine serum albumin as a standard.4

Radioactivity was determined by liquid scintillation count-

ing after dissolving the histones in a small aliquot of

water (step 7 in Chart 1 lists this procedure). One liter

of liquid scintillation fluid contained 72 ml of spectra-

fluor Butyl-PBD (Nuclear Chicago), 300 ml of ethanol, and

628 ml of toluene.

Determination of Acidic and Residual Proteins

Amounts of DNA in DNP subjected to various treatments

were determined prior to incubation to ensure that histone/

DNA ratios were unchanged by these treatments. Therefore,

any differences in acetylating activity observed between

pretreated DNP and control preparations cannot have been

due to changes in the essential composition of the complex,

such as might occur by the selective removal of histones.

The precipitate remaining after the 0.25 N HCl extraction

of histones was washed with cold 5% TCA, to remove any small

organic molecules, such as sucrose, which might interfere

with the DNA determinations. After centrifugation, the

precipitate was washed with 95% ethanol, absolute ethanol,

chloroform:menthanol 2:1, and twice with ether to remove

lipids. When DNA was determined by the H2SO4 reaction,

the mixture was dissolved in 1 N NaOH and incubated at 37C
148 o
for 16-20 hours. Upon acidification with HC1 at 0 C the

DNA precipitates leaving hydrolyzed RNA in solution. After

hydrolysis in perchloric acid at 900C for 20 minutes it
was reacted with H2SO4 to give a colorimetric reaction.

The Burton test which is a modified diphenylamine test for
DNA was also used. In this case it was not necessary to

remove RNA as this reaction is specific for deoxyribose.

Specific activity of acidic proteins was determined by

dissolving the proteins remaining in the precipitate after

hot acid extraction in 1 N NaOH. Any residual protein un-

dissolved by this procedure was dissolved in NCS solubilizer

(Nuclear Chicago) and counted.


1. Isolated nuclei washed 3x in saline-citrate--9

2. Deoxyribonucleoprotein (DNP).

3. Nuclei or DNP incubated in 2 ml final volume, containing:

0.19 M sucrose, 20 mM glucose, 25 mM phosphate buffer,
pH 6.75, 12 mM NaCI, 0.75 mM Ca+-!-, 5 mM Mg1+, 0.01 Pc
acetyl-14C-CoA (0.6 mUM).

4. After 15 minutes brought to 00C and trichloroacetic
acid (TCA) added to 5% and centrifuged.

5. Sediment extracted with 0.25 N HCl "Arginine-rich

6. Precipitated with 5% TCA, washed with acetcne-l% HCl,
ether, dry.

7. Dissolved in 0.6 ml 1120, protein concentration (Lowry),
radioactivity (liquid scintillation), DNA (H2SO4).

Methods for Extraction of Acetylating Enzymes

Extraction of acetylating enzyme from
acetone powder of rat liver nuclei

Nohara et al. have reported the isolation of acetylating
enzymes from an acetone powder of whole pigeon liver. In

addition Gallwitz has extracted an acetokinase from an ace-
tone powder of rat liver nuclei. An attempt was there-

fore made to use a similar procedure to extract the acetyl-

transferase from nuclei and DNP. A modification of the

original procedure was used, which was claimed by Bondy and

Roberts to be successful in the isolation of a histone
acetokinase from rat brain and liver nuclei.

Nuclei or DNP obtained from 25 gm of rat liver were

suspended in 7 ml of water. This preparation was then added

dropwise with continuous stirring to 70 ml of acetone kept

at -30 C. The resulting suspension was passed through

Whatman No. 1 filter paper on a Buchner funnel. The pre-

cipitate was washed 3 times with 50 ml acetone at -30 C and

dried in a dessicator at 0 C. The dry powder obtained was

ften homogenized in 13 ml of 0.1 M tris-HCl buffer at pH 8.2

and centrifuged at 2000 x g for 10 minutes. The resulting

supernatant was treated with neutral saturated ammonium

sulfate. The fraction which precipitated between 35-60%

saturation with ammonium sulfate was dissolved in 6.5 ml of

tris buffer and dialyzed against two successive 500 ml

portions of a solution which contained 0.068 M KCl, 0.001

M 2--mercaptoethanol and 0.02 M NaHCO3 at pHI 8.0 for 1 hour

each. This dialysed preparation contained the soluble ace-

tylating enzyme and 1 ml containing 0.1-0.5 my of protein

was incubated at 370C with 0.5 mg of isolated "arginine-

rich" calf thymus histones in 25 nmM phosphate buffer, pH 8,

12 mM NaCl in the presence of acetyl-- 14C-CoA. The reaction

was stopped by precipitating the protein onto filter paper

with cold 15% TCA.

In a preliminary experiment Bio Gel filters were used,

which were washed twice with 2 ml of incubation medium

containing 10 times the original concentration of unlabeled

acetyl-CoA, twice with 2 ml of acetone-1% HCI, and twice

with 1 ml of ether. Since this filter paper did not dis-

solve in scintillation fluid, corrections were made for loss

of counts by counting a sample before and after absorption

onto filter paper. In subsequent experiments millipore

(AAWP02500) filters were used which could be dissolved in
Bray's scintillation fluid. These were dried at 90 C for

15 minutes after the TCA precipitation step and subsequently


Extraction of acetylating enzyme from
rat liver nuclei with saline

Johlrxand Forrester have found that acidic proteins


which remain bound to the chromatin complex during the 0.14

M NaCI procedures used to prepare DNP from nuclei can be

removed by extraction with 0.35 M NaC1. It seemed

plausible that the acetylating enzyme was among this group

of proteins and could be extracted from the DNP complex

with 0.35 M NaCl. Therefore, attempts were made to extract

the acetylLransferase with 0.35 M NaCI from nuclei and DNP.

Purified rat liver nuclei or DNP from 6 gn of liver

were therefore blended for 1 minute in the Virtis homogeni-

zer in 4 ml of 0.35 M NaCl. After a preliminary centrifuga-

tion at 2000 x g for 10 minutes to remove the major portion

of the DNP, the supernatant was diluted to 0.14 M NaCl

and centrifuged at 40,000 x g for 30 minutes. To check

whether any DNP remained in the supernatant, the DNA con-

centration of the supernatants was determined which showed

less than 0.15 mg of total DNA present. This supernatant

was then used as an enzyme source.

DNP was inactivated by heating at 65C for 10 minutes

in 0.14 M NaC1-0.01 M citrate. The salt extract from

nuclei or DNP derived from 1 gm of liver was incubated with

inactivated DNP from 1 gm of liver at 370C for .5 minutes

in 0.1 M NaCl, 25 mM phosphate buffer, pH 8. To prevent

precipitation of the extract with the DNP, the reaction was

not stopped with TCA, but tih DNP C. -un down and hi stones

extracted and dried as usual. Where isolated "arginine-

rich" calf thymus histones were used as a substrate, the in-

cubation reaction was stopped by precipitation with 15% TCA

(since isolated histones are soluble in the medium) and the

resulting precipitate was dissolved in 1 M hyamine and

counted. Extracts alone and inactivated extracts plus sub-

strate were used as controls.

Extraction of acetylating enzyme
from rat liver DNP

Since previous extraction procedures were effective for

the isolation of acetylating enzymes from rat liver nuclei,

but not from DNP, the possibility was considered that these

enzymes represented cytoplasmic contaminants and were not

the acetyltransferase in the chromatin complex. Therefore,

an attempt was made to extract the acetylating activity from

the DNP complex using a method which had been proven to be
successful in the isolation of RNA polymerase. Earlier

studies on the isolation of RNA polymerase had revealed

that this enzyme is tightly bound to the chromatin complex,

similar to the chromatin acetyltransferase. In studies on

endogenous RNA polymerase activity the crude chromatin was
consequently often referred to as the "aggregate enzyme."

Nuclei and DNP from 12 gm of rat liver were gently

homogenized in 12 ml of 15 mN phosphate buffer, pH 8, 0.5

mra ethelonediamine teLraacetic acid (EDTA), and 1.0 mM

2-mcrcaptoethanol. This mixture was incubated at 30 C for

50 minutes with gentle shaking. In the original polymerase

extraction procedure a tris-phosphate buffer, pH 8.8, was
used, 1 but since phosphate buffer had been shown to be

much more favorable to the DNP catalyzed transfer of acetate

from acetyl-CoA to histones than tris buffer, the phosphate

buffer system was substituted. After the extraction, the

mixture was centrifuged at 115,000 x g for 40 minutes and

the supernatant served as the enzyme source. One ml of

this supernatant contained the enzyme extracted from DNP

isolated from 1 gm of rat liver. One ml of enzyme solution

was incubated in the presence of acetyl- C-CoA either with

inactivated (by heating at 65 C, 10 minutes) DNP from 1 gm

of rat liver or 0.5 mg of isolated "arginine-rich" calf

thymus histones. The enzyme was also added back to DNP

from 1 gm of liver which had been extracted by the above

procedure, i.e., to the 115,000 x g precipitate, in an

attempt to restore the preparation to full activity. The

final volume was 2 ml and contained 0.25 mM EDTA, 0.5 mM

2-mercaptoethanol, 33 rmM phosphate buffer, pH 8, 12 mM

NaCl, and 0.01 pc of acetyl-14C-CoA (spec. act. 56 mc/mM).

After 15 minutes at 370C, the fractions containing isolated

histones were cooled and made 15% with respect to TCA, and

were filtered on millipore filters. The filters were

dried in a hot air oven, dissolved in Bray's scintillation

fluid, and counted. The samples containing DNP were cooled

rapidly and centrifuged to prevent co-precipitation of the

enzyme in 15% TCA. The histones were extracted from the

sediment as described above.

Method for Determining the Extent of O-Acetylation of
Free Histones and that Occurring in DNP

It is possible to estimate the degree of O-acetylation

occurring in proteins by a method devised by Narita.1 In

this procedure one takes advantage of the liability of the

O-acetyl bond in the presence of hydroxylamine. Histones

or DNP prelabeled with radioactive acetyl-CoA can therefore

be incubated in the presence of hydroxylamine. The degree

of O-acetylation can subsequently be estimated by measuring

the amount of radioactivity released as a result of this


Pogo et al. have shown that the acetate incorporated

in the f histone fraction of regenerating rat liver was
stable to treatment with 2 M hydroxylamine,0 indicating

that this fraction does not contain O-acetyl groups. Other

studies on the site of acetylation of histones in calf

thymus nuclei f2al fraction showed that acetylation only
2at the epsilon-amino group in lysine.
takes place at the epsilon-amino group in lysine. The

f3 fraction of regenerating rat liver, however, showed a

release of 55% of the acetyl groups as a result of hydro-
xylamine treatment, which agrees with the work of Nohara

et al. who reported considerable O-acetylation of the f3

fraction by pigeon liver enzymes in vitro. In the case

of calf thymus, all of the acetate incorporated in the f3

fraction upon incubation of nuclei in the presence of 14C

acetate was recovered as epsilon-N-acetyllysine after en-

zymatic digestion and ion exchange chromatography. Thus,

there is,at least in calf thymus, no evidence for the forma-

tion of O-acetyl linkages under these conditions.

Since it may be expected that an enzymatically cata-

lyzed reaction will show greater specificity of acetylation

and therefore perhaps a lower level of O-acetylation than

that occurring spontaneously, DNP catalyzed acetylation in

the presence of acetyl-CoA was compared with that occurring

spontaneously by free histones with respect to O-acetyla-

tion. Conditions were selected to yield a similar amount

of radioactivity by either process. Therefore, for DNP

catalyzed acetylation, DNP equivalent to 1 gm of liver was

incubated in the presence of 0.01 uc of acetyl- C-CoA,

12 mM NaCl and 25 mM phosphate buffer, pH 8, for 15 minutes

at 37C. The reaction was stopped with 5% TCA, and the

histones were dried with acctone-l% HCl and ether.

For the hydroxylamine test, 0.45 ml of hydroxylamine

solution (pH 6.4) containing 3 volumes of 40% hydroxylamine

hydrochloride and 2 volumes of 3.5 N NaOH were added to 0.3

ml of 0.1 N acetate buffer, pH 5.4, and 0.3 ml of H20 con-

training 0.3-0.5 mg of 1C-acetylated histones obtained as

described above. After standing at room temperature for 5

hours the histones were precipitated with 15% TCA, centri-

fuged after 4 hours, and dried in acetone-l% HCl and ether.

The percent 0 acetylation was determined by counting the

radioactivity released in an aliquot of the 15% TCA super-

natant, whereas the extent of N-acetylation was deduced from

the radioactivity remaining in the histones after the

hydroxylamine treatment.

Determination of Incorporation of Radioactive
Acetate into the Various Histone Fractions

Fractionation according to Johns157

Since it has been observed by others that acetylation

of histones occurs primarily in the arginine-rich (fal and
f3) fractions (see Introduction), it was deemed important to

compare this labeling pattern with that of histones ace-

tylated in vitro in DNP. Such a comparison would determine

how closely the in vitro reaction approximated the in vivo

situation. Therefore, histones which had been acetylated

were separated into the five major fractions according to

Johns and their specific activity determined.1 These

fractions were examined by acrylamidc gel electrophoresis to

determine the purity of the separation.

For this experiment, DNP was obtained as usual from

15 gm of rat liver and incubated at 370C for 15 minutes in

5 gm samples containing 10 ml total volume per sample of 12

mM NaC1, 25 mM phosphate buffer, pH 8, and 0.1 uc acetyl-
C-CoA. After the incubation, the reaction was stopped by

centrifugation and the unbound label was rinsed out by

washing once with a large volume of saline-citrate. Then,

the very lysine-rich (fl) fraction was extracted from the

DNP with 8 ml of 5% perchloric acid (PCA), and centrifuged

at 1.100 x g for 20 minutes. This procedure was repeated

once with 4 ml of 5% PCA. This fraction was precipitated

from the combined supernatants with 25% TCA, dried with

acetone--l% HCl, and ether. The residual DNP was extracted

for 18 hours with 20 ml of ethanol-1.25 N HCI (4:1), with

stirring, and contrifuged at 1100 x g for 15 minutes. This

extraction was repeated once with 10 ml of ethanol-HCl

(4:1), for 2 hours. After centrifugation the combined

supernatants contained the arginine-rich (f2a and f3) frac--

tions, whereas the moderately lysineo-rich (f2b) fraction

remained still bound to the residual DNP. The supernatants

containing the f.2 and f3 fractions were dialyzed against
4P a 3

ethanol for 1.8 hours, which causes the f fraction to pre-
cipitate, leaving the f2a fraction in solution. After cen-

trifugation, Lhe f2a histones were further separated by the

addition of an equal volume of acetone. This caused the

precipitation of f2a2. After centrifugation f2al was pre-

cipitated from the supernatant with 3 volumes of acetone.

The precipitates were dried in ether. The moderately lysine-

rich (f2b) fraction was then extracted from the residue with

0.25 N HCl, and precipitated with 5 volumes of acetone, and

dried in ether.


DNP from
15 grams of rat liver

extract with 8 ml 5% PCA
centrifuge 1100 x g, 20 min.

repeat with 4 ml 5%PCA


extract 18 hours with 20 ml
ethanol-1.25 N HCl (4:1)

centrifuge 1100 x g, 20 mir

repeat with 10 ml for 2 hrs


1 add TCA to
.- 1100 x g 20 min.

. Precipitate=f1
wash with acetone-l% HC1, ether

ecipitate superhatant
extract with 0.25 N HCI
dialyze against ethanol
centrifuge 1100 x g, 20 min. 18 hrs.
SI ~ supernatant
- precipitate=f3

ecipitate supernatant wash with ether add 1 volume
add 5 vol. acetone
I supernatant


1100 x g
20 min.

wash with ether

add 3 vol. acetone


wash with ether



Electrophoresis of histone fractions

The histone fractions obtained according to Johns were

checked for purity by polyacrylamide gel electrophoresis

in a vertical electrophoresis apparatus from the E. C.

Apparatus Co., Philadelphia, Pennsylvania. Acid gel con-
editions, designed for the resolution of basic proteins,

were used as follows:

Gel and Sample Buffer: Tris 0.12 M adjusted to pH 2.9
with citric acid.

Electrode Buffer: Glycine 0.37 M, adjusted to pH 4.0
with citric acid.

Sample Solvent: Tris-citric acid buffer pH 2.9 con-
taining 6 M urea. The sample solvent
was saturated with sucrose to facili-
tate the settling of the sample in
the slots.

Polymer Solution: 12% Cyanogum-41 in tris-citric
acid buffer, pH 2.9, containing
3 M urea. Total volume was 150
ml'for each gel.

Catalyst: 0.1% ascorbic acid, 0.0025% ferrous sulfate,
and 0.02% H202.

Besides the addition of sucrose to the sample solvent,

the only modification of the procedure as originally de-

signed was that the H202 concentration was dropped from

0.03% to 0.02% to increase the polymerization time long

enough to pour the gel.

After mixing the polymer solLition, the ascorbic acid

and ferrous sulfate were added with stirring. Immediately

after mixing, the T1202 was added and the gel was poured.

The electrophoresis apparatus was precooled before pouring

the gel by running water at 80C through the circulation

system. This was done to prevent contraction of the gel

upon polymerization. Since the gel hardened within a few

minutes, it was necessary to remove any bubbles that formed

immediately. After about 10 minutes the gel had sufficiently

hardened to fill the apparatus with 2 liters of the electrode

buffer. It was only after this step that the teflon slot

former could be easily removed.

Samples of 15 ul containing 30 ,ug of protein were

placed in each slot and ran at 250 v., 80C, for 4 hours.

It was found in preliminary experiments that no pre-run

was necessary using these gel conditions. Staining of the

gel was done in 0.2% Amido Schwartz, 7% acetic acid, and 40%

ethanol, for 20 minutes, and destaining accomplished elec-

trophoretically by a destainer from E. C. Apparatus Company.


In early experiments the in vitro acetylation of his-

tones by isolated nuclei was studied. These nuclei were

purified by centrifugation through 2.1 M sucrose as described

in Materials and Methods. In the course of these investi-

gations it was discovered that if these purified nuclei

were washed with isotonic saline, the resulting DNP prepara-

tion possessed an acetylating activity approximately 3 times

that of the nuclei. This was an indication that perhaps an

acetylating enzyme responsible for the transfer of acetate

from acetyl-CoA to histones was present in or at least

closely associated with the chromatin complex. Since other

laboratories had only reported the isolation of acetylating

enzymes from whole tissue or nuclei, this finding

stimulated interest to characterize this reaction more

carefully. An enzyme obtained from the DNP complex had less

chance of being a cytoplasmic contaminant than those iso-

lated from nuclei, and, in view of its localization in the

chromatin itself, took on more meaning in view of the pos-

sibility that histone acetylation plays a role in chromo-

somal function, e.g., the control of gene expression.



To verify that the acetylation reaction was not due to

cytoplasmic contamination, DNP was also prepared from

nuclei isolated in detergent. This procedure removes the

outer nuclear membrane. DNP prepared from these nuclei was

found to possess an acetylating activity approximately

equivalent to that of DNP obtained from nuclei isolated by

the usual procedure. This would support the contention

that the acetylating activity observed in DNP prepared from

nuclei by the usual procedure isnot of cytoplasmic origin.

Further support for this was obtained from experiments

in which DNP isolated according to the usual procedure was

further purified by centrifugation at 22,000 rpm for 3. hours

in the Spinco S. W, 25 head in 1.7 M sucrose. Marushige

and Bonner have shown that rat liver chromatin purified by
this procedure is characterized by a low RNA content, and

others have reported that the nonhistone protein contained

in this preparation is not a cytoplasmic contaminant, but a
real constituent of chromatin. DNP treated in this

manner showed acetylating activity two or three times

greater than that of the usual preparation. These results

confirmed that the acetylation reaction was not due to a

cytoplasmic contaminant and was closely associated with the

chromatin. Having established the localization of the

transferase activity in DNP, the following experiments were


designed to characterize the conditions influencing this

reaction, using rat liver DNP as the source for enzyme

activity as well as the acetate acceptor.

Conditions Influencing the
Acetylation oE Histones in DNP

Among the conditions influencing the in vitro reaction

that were studied were divalent cations, medium components,

temperature, concentration of substrates, and pH. These

are discussed below in that order.

Divalent cations
Nohara et al. has found a Mg requirement for two

pigeon liver fractions which act respectively as acetate

activating and transferring enzymes in the acetylation of
11"7 ++
isolated histones. Gallwitz has suggested that Mg is

only necessary for the activating enzyme since a trans-

ferring enzyme which he has isolated from rat liver nuclei
does not require it. In view of these findings it was

decided to study the effect of several divalent cations

on the DNP catalyzed transfer of acetate from acetyl-CoA

to histones.

Rat liver DNP washed in 0.075 M NaCI and 0.024 M EDTA
chelatess divalent cations), p1H 8, was incubated in the

presence of 5 mM EDTA or different concentrations of divalent

cations. The results indicate a depression in activity at

10 mM Mn and 8 mM Ca The presence of Mg from 0-10

mM did not seem to affect the activity significantly.

Medium components (Table I)

Since the original incubation medium was designed for

the study of protein synthesis by isolated calf thymus
nuclei, it seemed desirable to analyze the effects of the

various components of this medium on the acetylation re-

action. The results of 1 experiment using 4 samples for

each condition is reported in Table I. The term "complete

medium" refers to the one described in Materials and Methods,

except that phosphate buffer, pH 8, rather than pH 6.75, was

used and the divalent cations were omitted. These changes

were made on the basis of other experiments to be reported

here. Components of the complete medium were omitted and

the molarity of the buffer was gradually lowered. The re-

sults indicate that the presence of sucrose and glucose is

not essential for the DNP catalyzed acetylation of histones,

and the optimum concentration of the phosphate buffer was

25 mM. The effect of ionic strength on the reaction mix-

ture was studied by changing the concentration of NaCl in

the incubation medium. Between 0.01 M to 0.2 M NaCl there

was a slight rise in activity up to 0.1 M NaCl, followed

by a decrease at 0.2 M NaC1. This is in agreement with a

report by Gallwitz and Sekeris on the acetylation of histones
by rat liver nuclei. The optimum conditions for this

reaction were therefore assumed to be 12 mM NaCI, and 25 mM

phosphate buffer, pH 8, according to the parameters studied.

Temperature (Figure 1)

Figure 1 shows the course of the reaction with time

at two different temperatures. The medium for this experi-

ment was as described for the original procedure (Materials
and Methods) except that the concentration of acetyl- C-CoA

was 4 times higher. From these data a Q10 of about 1.8

could be calculated. Allfrey has reported a Q0 of 2.09
for this reaction in nuclei.126 In general the velocity

of enzymatic reactions is doubled for a 100 rise in


Table I

Effect of Various Medium Components on the Transfer of
Acetate from Acetyl-CoA to Histones in DNP

Incubation Media Specific Activity of Phosphate
(cpm/mg) Buffer

Complete medium 8243 + 611 25 mM

-glucose 9075 + 910 25 mM

-NaCl 7037 + 444 25 mM

-NaC1 1729 + 224 2.5 mM

-NaCl 998 + 85 0.25 mM

-NaC1 769 + 76 0.025 mM

The results are expressed as specific activity (counts
per minute per mg) of histones plus or minus the standard

Figure 1: Effect of Temperature on the Transfer of
Acetate From Acetyl-CoA to His ones in DNP.

The results are expressed as specific acLivity (counts
per minute per mg x J0 ) of histones as a function
of time.



0 5 10 15

Concentration of DNP (Figure 2)

Since in this particular experimental set-up the ace-

Late-accepting substrate and enzyme are both present in

the DNP preparation, the histone concentration varies with

that of the enzyme when the DNP concentration in the re-

action mixture changes. Therefore, it seemed interesting

to determine the optimum concentration of DNP for the trans-

fer reaction. Rat liver DNP was incubated at different

concentrations in the presence of acetyl- C-CoA in the

usual manner. The results are represented graphically in

Figure 2, either as total counts per minute or as specific

activity (counts per minute per mg histone). As can be

seen, there is a rise in both curves with an optimum con-

centration of DNP equivalent to 0.3-0.5 mg of histones per

2 ml incubation mixture. Past this point increasing con-

centrations of DNP affected the total incorporation to a

minor degree, but caused a rapid drop in the specific

activity of histones.

Concentration of acetyl-CoA (Table II)
and Km determination (Figure 3)

The Km of the transfer reaction for acetyl-CoA was

determined by incubating the optimal amount of DNP with in-

creasing concentrations of acctyl-CoA. Results were ob-

tained from 2 separate experiments using isotope solutions

containing 3 different specific activities of acetyl-14C-

CoA. The data are given in Table II below. The results
were plotted (Figure 3) according to Lineweaver and Burk,

giving a KIn value of 2.5 x 10 M.

Effect of pl1 in different buffer systems (Figure 4)

In these experiments 25 mM phosphate buffer was sub-

stituted by 25 mM tris-HCl buffer or 25 mM glycine-NaOH

buffer. The optimum pH for the DNP catalyzed acetylation of

histones in glycine-NaOH buffer was found to be 8.5-9.6.

The phosphate buffer system appeared the most favorable for

the DNP catalyzed reaction, and since pH 8 was the highest

pH that can be obtained with this buffer, it represents the

optimum buffer conditions for DNP catalyzed acetylation of




2000 ,


E 1000


0 .5 1.0 mg. Hist.

Figure 2: Effect of the Concentration of DNP in
the Incubation Mixture on the In Vitro Transfer of
Acetate from Acetyl-CoA to Histones

DNP was incubated at increasing concentrations in the
presence of acetyl-14C-CoA as described in Materials
and Methods. Results are expressed as total counts
per minute recovered in isolated histones []--
or as specific activity (counts per minute per mg)
6--- of histones. DNP concentrations
are expressed in terms of histone content of DNP
sample in 2 ml medium.

Table II

Effect of Increasing Concentrations of Acetyl-CoA
on the Transfer of Acetate
from Acetyl-CoA to Histones in DNP

of Acetyl-CoA

EXP. 1 1.6 x

1.6 x

1.6 x

1.6 x

1 x


EXP. 2














10 M

10 M

10 -5M
10 M






10 M

Spec. Act.













Spec. Act.
(cpm/jj mole)

28.8 x 106

2.3 x 106

1.0 x 10

Moles Acetate/
mg Histones

27.7 x 10

213.2 x 10
829.5 x 10

1176.8 x 10-6

1631.0 x 10-
1377.0 x 10

212, 4



160. 7







----I---- ---------)

-- --- ------------



"o 0.6

x -6

-> 0.4 Km= 2.5 x 10


0 0.2 0.4 0.6

x 107

Figure 3: Effect of Increasing Concentrations of Acetyl-CoA
on the In Vitro Transfer of Acetate from AceLyl-
CoA to Histones in DNP Plotted According
to Lineweaver and Burk

Rat liver DNP equivalent to 0.3-0.5 mg histones was incubated in
25 mM phosphate buffer, p11 8, 12 mM NaCl, 0.19 M sucrose
and 20 mM glucose in the presence of increasing concentra-
tions of acetyl-CoA as described in Materials and Methods.
Calcium and magnesium were omitted. Ordinate: V=pmoles
of acetate incorporated per mg histone. Abscissa: S-molar
concentration of acetyl-CoA.


" /(phosphate)




Z (tris-HCI)




/ \



-2000 I
(glycne-N OH)

8.5 9.0


Figure 4: Effect of pH in Different Buffer Systems on
the Transfer of Acetate from Acetyl-CoA to
Histones in DNP

The results are expressed as specific activity (counts per
minute per mg) of histones as a function of pH.

8.0 8.0


Enzymatic Nature of the DNP Cata.lyzed
Acetylation of Histones

Since it had been reported, and subsequently recon-

firmed in this laboratory, that isolated histones will be-

come acetylated spontaneously when incubated in the presence

of acetyl-CoA, considerable attention has been given to the

possibility that the acetylation of histones occurring with-

in the DNP complex may not be enzymatic.

Inhibition studies (Table III)

The first approach taken to rule out nonenzymatic

transfer of acetate from acetyl-CoA to histones was an in-

direct one. In these studies the DNP complex was treated

prior to incubation in various ways to either wash out or

extract the acetylating enzyme, or to inhibit its action by

methods which are in general considered to be deleterious to

enzymes. The effects of these various treatments on the

transfer of labeled acetate during subsequent incubation

are shown in Table III.

Treatment of DNP in 1 M NaCl causes it to dissociate
17, 162, 163
into histones and DNA. Subsequent dilution of this

solution to 0.14 M will result in recombination of histone

and DNA. NonhisLone proteins, however, remain in solution.

It was found that if liver DNP was subjected to such a

treatment the activity decreased to 66% of untreated con-

trols. Although this treatment is expected to remove non-

histone proteins bound within the DNP complex, it appears

as though most of the activity remains tightly bound to

the DNP.

Recently Johns and Forrester reported that extraction

of calf thymus DNP with 0.35 M NaCl removes acidic proteins

which have become bound to it during the isolation proce-
dure. An attempt was therefore made to extract the

factor responsible for acetylating activity from rat liver

DNP with 0.35 M NaCl (see Materials and Methods). The re-

sults of these experiments indicate that DNP subjected to

such treatment retains 80% of its acetylating activity.

Heating briefly at 650C, or washing with ethanol,

acetone, and ether, practically abolished activity. The

histone acetylation in these preparations appeared to be

negligible, both under conditions (phosphate buffer) which

favor an enzymatic reaction and those (tris buffer) favoring

a chemical process (see below). Prolonged incubation of DNP

in the regular medium at 00C and 37 C for different time

periods resulted in a decreased activity. Neither the salt

extractions, treatment with organic solvents,or heat,

changed the histone/DNA ratio of the DNP. It could be

argued that the decrease in acetylating activity observed

upon incubation of DNP for prolonged periods at different

temperatures could be due to an alteration in the histones


rather than a change in an enzyme within the DNP. To check

this possibility, "arginine-rich" histones from DNP which

had been incubated for 20 hours at 0C and for 4 hours at

37 C were isolated and their ability to become acetylated

spontaneously or enzymatically in the presence of the ace-

tyltransferase (see below) was compared with that of his-

tones isolated from control preparations of DNP. This ex-

periment showed that histones from these pretreated prepara-

tions were still capable of becoming acetylated to the same

degree as the controls or even higher. These findings argue

against the possibility that the histones were degraded or

extracted from the DNP complex during these treatments.

Therefore the observed decrease in acetylation cannot be

explained by a defective substrate, but presumably results

from enzymatic denaturation.

Free histones subjected to treatment with organic

solvents or heated at 65 C were still capable of becoming

spontaneously acetylated. In contrast the acetylation re-

action within the DNP is completely abolished by these

treatments. These findings suggest that acetylation is

somehow prevented in histones which are bound within the

DNP comr.plex.

Table III

Effect of Various Conditions on In Vitro Transfer of
Acetate from Acetyl-CoA to Histones

Pretreatment of Deoxyribonucleoprotein

Extraction with 1 M NaCi

Extraction with 0.35 M NaCl

Organic solvents

650C for 5 minutes

Organic solvents (Incub. at pH 9, 12 mM tris)

65C for 5 minutes (Incub. at pH 9, 12 mM tris)

Incubated at 0C for 20 hours in regular medium

Incubated at 370C for 4 hours in regular medium










The results are expressed as percent of untreated con-
trols on the basis of specific activity (counts per minute
per mg of histones).

Comparison of DNP catalyzed acetylation with
spontaneous acetylation of free histones

Buffer Effects (Figure 5)

Comparisons between the pH and buffer conditions in-

fluencing the spontaneous acetylation of free histones and

that occurring within the DNP were made in an attempt to

show that they are, indeed, two different reactions. Figure

5 represents a composite graph which summarizes data from a

number of separate experiments in which the spontaneous

acetylation of free histones was compared with that occurring

within DNP in different buffer systems, at increasing pH.

As can be seen the DNP catalyzed reaction is greatly favored

by the phosphate buffer system, whereas the spontaneous

reaction is more active in tris-HCl buffer. The spon-

taneous reaction shows an increase with pH in tris buffer,

and at pH 9 actually exceeded that occurring in the DNP.

Due to the range of the phosphate buffer system, it was im-

possible to observe the DNP catalyzed reaction under optimum

phosphate buffer conditions at a p1l higher than 8. Both

the spontaneous and the DNP mediated reactions show a pH

optimum at pH 8.6-9.6 in glycine-NaOH buffer.

Temperatuce Effects (Figure 6)

Since most mammalian enzyme reactions show temperature

optima close to 370C, whereas nonenzymatic reactions in-

crease with temperature, it was decided to compare the

spontaneous acetylatiun of hisLones with that occurring

in DNP at increasing temperatures.

Previous experiments indicated that a phosphate buffer

system, pH 8, represented the most favorable conditions ob-

tainable for the DNP catalyzed acetylation of histones and

that a tris-HCl, pH 9, buffer represented the optimum con-

dition for the spontaneous acetylation of free histones.

Therefore, the transfer of acetate from acetyl-CoA to his--

tones in DNP as well as free calf thymus histones were com-

pared under these conditions at increasing temperature.

The data shown in Figure 6A were obtained at pH 8 in

25 mM phosphate buffer (most favorable conditions for DNP

acetylation); those in 6B at pH 9 in 25 umM tris-HCl buffer

nost favorable for spontaneous acetylation of free his-

tones). As can be seen, there is a pronounced differential

effect of the ionic environment on the two reactions. In

both media, however, the qualitative effect of increasing

temperature on each of the two processes was the same: the

enzymatic reaction in the DNP complex indicating an optimum

at 370C, followed by a sharp decline; while the spontaneous

acetylation of isolated histones showed a progressive in-

crease with increasing temperature up to 770C. This ex-

periment offers further evidence that the acetate transfer

occurring in the DNP complex is catalyzed by an enzyme.

Determination of the Extent of O-Acetylation
by _Iycroxyl]amine Test (Table IV)

Since most evidence available at present indicates that

the acetylation of histones occurs only at the epsilon-

amino group of internal lysine residues and at alpha-

amino groups of terminal amino acids, it may be expected

that O-acetylation often found in in vitro systems is an

artifact occurring only as a result of spontaneous acetyla-

tion. The DNP catalyzed acetylation of histones was there-

fore compared with that occurring spontaneously by comparing

the extent of O-acetylation occurring in either case. The

results in Table IV show that 3% of the acetyl-groups of

"arginine-rich" histones acetylated by DNP can be released

by hydroxylamine treatment, whereas about 28% can be re-

leased from "arginine-rich" histones acetylated spon-

taneously. This suggests a greater degree of specificity

occurring in the DNP catalyzed reaction and implies an

enzymatic rather than a random spontaneous process.

II Spontaneous reaction with free histones
8000 Enzyme reaction within DNP



' 2000

7.0 8.0 9.0 PH 7.0 8.0
Tris Buffer Phosphate Buffer

Figure 5: Comparison.of the Effect of
Increasing pH in Different Buffer Systems
on the Spontaneous and Enzymatic Acetylation Reactions

Results are expressed as specific activity (counts per
minute per mg) of histones. Range of data is given in

20,000 r






j 4000


Figure 6: Effect of Temperature on the Transfer of
Acetate from Acetyl-CoA to Histones in Rat Liver
DNP -- and to Free
Histones - - -i in Different Buffer Systems

In Figure 6A, the incubations were carried out in 25 mMI
phosphate buffer, pH 8; in Figure 6B, in 25 mM tris buffer,
pH 9. Calcium and magnesium were omitted in both. Results
are expressed as specific activity (counts per minute per
mg) of histones.

30 50 70 *C

Table IV

Comparison of the Extent of O-Acetylation in DNP with that
Occurring Spontaneously as
Measured by Lability to Hydroxylamine

Type of Labile Acetyl Stable Acetyl Percent
Reaction Group (cpm) Group (cpm) O-Acetylation

Transferase Exp. 1 67 2267 2.8
DNP Exp. 2 47 1149' 3.9

Acetylation Exp. 1 1340 3994 25.1
of Free
Histones Exp. 2 579 1287 31.0

Results are expressed as total counts per minute. Per-
cent O-acetylation was derived from the amount of label re-
leased by the hydroxylamine and the total counts recovered.

Enzyme isolation

Although the evidence reported above suggests the

presence of an enzyme within the DNP complex which is re-

sponsible for the transfer of acetate from acetyl-CoA to

histones, conclusive proof of this hypothesis would be the

isolation of the enzyme (acetyltransferase). The results

of attempts to this effect are reported below.

Acetylating Enzymes from Rat Liver Nuclei

Acetone powder extracts

The first attempts at isolation followed procedures re-

ported to have been successful for the isolation of histone

acetokinases from rat brain and liver nuclei.118' 151

Following the procedure given in Materials and Methods, it

was found that a low activity acetylating enzyme could be

isolated from nuclei using this procedure, but not from DNP.

In fact, protein determinations of the tris extract of the

acetone powder preparation from DNP showed that no protein

whatsoever could be extracted by this procedure. When the

acetone powder of DNP or nuclei remaining after tris ex-

traction was tested for acetylating activity, it showed

about 50-70% of the activity usually observed in normal DNP

preparations. This suggests that the enzyme responsible for

the reaction in the DNP complex is still tightly bound to the

DNP and must therefore be decidedly different from that

which is extractable from nuclei.

Saline extracts (Tables V, VI)

Since it has been reported that washing DNP in 0.35 M
NaC1 removes acidic proteins, saline extracts were made

of nuclei and DNP which were examined for acetylating ac-

tivity. Table V shows the results of 3 different ex-

periments in which these extracts, in quantities comparable

to the native preparation, were added back to inactivated

DNP. As can be seen, no activity was restored to inacti-

vated preparations by the addition of extracts from DNP,

although the addition of extracts from nuclei did restore

the acetylating activity to 10% of that of the controls.

Table VI shows the activity of salt extracts of nuclei

using isolated "arginine-rich" calf thymus histones and

polylysine as acetate acceptors. It can be seen that under

these conditions this enzyme lacks specificity, as poly-

lysine was acetylated to the same degree as isolated his-


Table V

Restoration of Acetylating Activity
in Heated DNP by 0.35 M Extracts from Nuclei

Experiment 1 2 3

Control DNP 8505 + 1340 10997 + 758 6718 + 474

65C inactivated DNP 306 + 64 292 + 156 54 + 46

650C inactivated DNP and
0.35 M extract of DNP 284 + 101 171 + 6

650C inactivated DNP and
650C inactivated 0.35 M
extract of DNP 196 + 36 0

650C inactivated DNP and
0.35 M extract of nuclei 1306 + 27 792 + 50

Results are expressed as specific activity (counts per
minute per mg) of histones plus or minus the standard

Table VI

0.35 M NaCl

Incubated Material

Isolated Histones

Histones and extract

Histones and extract

Histones and 650C
inactivated extract

Extract alone


Polylysine and extract

of Isolated Histones with a
Enzyme Extract from Nuclei

Total Counts per Minute

131 + 23

1677 + 275 (enzyme extracted from
1 gm liver)

1199 + 1075 (enzyme extracted from
0.5 gm liver)

113 + 26 (enzyme extracted from
0.5 gm liver)

268 + 4 (enzyme extracted from 0.5
gm liver)

267 + 48

1288 + 368 (enzyme extracted from
0.6 gm liver)

Results are expressed as total counts per minute plus
or minus the standard error.

Acetylating Enzymes from Rat Liver DNP (Table VTI)

A method similar to one used for the isolation of RNA

polymerase from the "aggregate enzyme complex" proved to be

successful for the extraction of the acetyltransferase ac--
tivity from DNP. Table VII shows the results of 2

experiments in which the extract containing acetyltrans-

ferase activity from DNP from 1 gm of liver was added back

l:o inactivated DNP, free histones, or to DNP which had been

extracted by this procedure. In each instance, the sub-

strate contained histones in quantities equivalent to that

found in 1 gm of liver. In the case of samples in which

isolated histones served as the substrate, the specific

activity was calculated by dividing the total counts ob-

tained by the amount of histones added to the reaction mix-

ture. Since there were no histones present in the case of

the incubation of the extract alone, this hypothetical

figure represents the specific activity calculated by as-

suming that the standard amount of histones was present.

This figure was derived by dividing the total counts per

minute in these samples by 0.5 mg of histones.

As can be seen, the extract can only restore activity

to heated DNP Lo 10% of the control level, although about

50% of the acetylating activity appears to be extracted by

this procedure. When the extract is added back to DNP


which has been extracted by this procedure, again, only

approximately a 10% increase is observed. However, when

the extract is added to free histones, there is a high rate

of acetylation.


Table VII

Extraction of Histone Acetyltransferase Activity from DNP

Exp. 1 Exp 2
Specific Specific
Incubated Material Activity Activity
S___ (cpm/q)_ (cpm/ij)

Normal DNP

11204 + 1608

5857 + 1046

DNP heated at 65C, 10 min.

Heated DNP and extract

"Arginine-rich" histones (0.5 mg)


"Argininc-rich" histones (0.5 mg)
and extract

Extracted DNP

172 + 131 134 + 82

1140 +

1142 +

75 537 +


528 +

474 + 18" 440 +

9668 + 335 4445 + 287

6107 + 228 3523 + 79

Extracted DNP and extract
added back

7013 + 198 4898 + 708

The amount of DNP or extract used for each


was obtained from 1 gm of rat liver. The results are ex-
pressed as specific activity (counts per minute per mg) of
histones plus or minus the standard error.

Specific activity was calculated on the assumption that
0.5 mg of histoncs was present in the reaction mixture.

Relative acetylation of the various
histone fractions (Table VIII) (Figure 7)

After incubation of DNP in the presence of acetyl-CoA

as described, the histones were extracted and fractionated
according to Johns into the five major histone groups.

Table VIII shows the specific activity of each of these

fractions. As can be seen, the arginine-rich histones,

f 2 and f are the most actively acetylated fractions,
2al 3

with some activity also in the f fraction. The radio-
activity of the lysine-rich fractions, f and f2b' was low

which correlated with in vivo findings.

The purity of these fractions was checked by acrylamide

gel electrophoresis and Figure 7 illustrates the patterns

observed. As can be seen the f sample shows a band
corresponding to the f band and therefore this fraction
is probably slightly contaminated with f2al histones. This

could account for some of the activity observed in the f
sample, as the f fraction is very highly labeled. Frac-
2al 1
tions f3 f2a2 and f2b are not completely separated by

this procedure, when all the fractions are combined, but

when run separately f3 and f2b each show only one distinct

major band and therefore were considered to be relatively


Table VIII

Acctylation of Various Histone Fractions in DNP

__Fraction (cpm/mg)

Lysine-rich histones fl 557

f2b 2,909

fal 16,041

Arginine-rich histones f2a2 8,713

f3 12,904

The results are expressed as specific activity (counts
per minute per mg) of histones.

PJ. I: PI o
nJ H

p- 0 (D
& *



f H
H- 0

0 0 1
O0 0

S 0
h- H

-n Oj n
rf r

H ,-

p-C r1 W| "I NJ
i rn NJ

h -h (n

Ch (D _

cn ( t

Ni H
C Ph h-
o H UJ H-


Since previous work has revealed that free histones are
spontaneously acetylated upon incubation with acetyl-CoA,

the possibility arose that the acetylation of histones oc-

curring in DNP might not be enzymatically catalyzed. There-

fore, attention has been given to this problem and evidence

is provided for the presence of an acetyltransferase bound

to DNP which is responsible for this reaction.

Indirect evidence suggesting that the in vitro acetyla--

tion is catalyzed by an enzyme was provided by experiments

in which DNP was pretreal-ed in a way which would be expected

to denature a completed enzyme. Heat or organic solvents,

which are generally considered to be inhibitory to enzyme

reactions, practically destroyed the acetylating activity

of DNP. The fact that after these procedures the histone/

DNA ratios were unchanged implies that the physical compo-

sition of the DNP complex was the same even after treatment.

Therefore, the inhibitory action observed can be attributed

to denaturation of an enzyme responsible for the acetylation

reaction, especially since similar pretreatment of histones

did not affE-ct their spontaneous acetylation. DNP subjected


Lo heat or organic solvents does not catalyze the acetyla-

tion reaction, even under conditions most favorable for a

spontaneous transfer of acetate to hisLones. Apparently,

spontaneous acetylation is somehow prevented in histones

bound to DNA in the DNP complex. This may have biological

implications for the mechanism of gene control in that only

an enzymatically catalyzed reaction, capable of a great deal

of specificity, rather than a random process, is permitted

within the DNP complex.

In the case where DNP showed a loss in acetylating

activity as a result of preincubation at 370C for 4 hours,

it could be argued that the histones themselves were de-

graded by this treatment and thereby were incapable of be-.

coming acetylated. However, if histones are isolated from

thusly treated DNP, they are still capable of becoming spon-

taneously acetylated or acetylated in the presence of a

subsequently prepared extract containing acetyltransferase.

These histones as well as the acetyltransferase are there-

fore relatively unchanged with respect to their ability to

become acetylated.

Although the above mentioned pretreatments of DNP,

especially heating at 65 C, did not change the composition

of the DNP with respect to histone/DNA ratios or the ability

of the subsequently isolated histones to become spontane-

ously acetylated, it could be argued that the structure of

the complex was altered in such a way as to prevent a non-

enzymatic transfer of acetate causing the observed inhibi-

tion. This seems unlikely, however, as it has been demon-

strated that the DNA of liver chromatin js stabilized against

heat denaturation as compared with deproteinized liver
DNA. The temperature of half--melting (T ) is 680C in the

case of naked DNA, but is increased to 81 C for chromatin.

Preparations of DNP heated at 650C according to the methods

described herein, therefore, probably represent a native


Comparisons between acetylation of completed histones

in DNP and that of free histone occurring spontaneously re-

vealed that, although the pH optima of the two reactions

was the same, the DNP reaction was much more efficient in a

phosphate buffer system, whereas the spontaneous reaction

was greatly favored by a tris-HCl buffer. Furthermore, the

DNP catalyzed reaction showed a marked temperature optimum

of 370C, which is common for mammalian enzyme reactions,

while the spontaneous reaction rate increased with increas-

ing temperature. These data would seem to indicate that

one is dealing with two completely different processes.

Another difference between the acetylation of histones

in DNP and that of free histones was the degree of O-acetyla-


tion. Acetyl groups of "arginine-rich" histones acebylated

by DNP showed a much greater degree of stability to hydroxyla-

mine treatment than those acetylated spontaneously. This

indicates that most of the acetate is N-linked, which corre-

lates with findings by others on the acetylation of histones
93, 94
by calf thymus nuclei. Although findings by others

on regenerating rat liver showed 55% O-acetylation for the

f3 fraction, acetate incorporation in the f2al fraction was
stable to hydroxylamine treatment. Using the same pro-

cedure Gallwitz and Sekeris also found 35% O-acetylation

among the acetate groups of the f3 fraction acetylated in
vitro in rat liver nuclei, but none in the other fractions.

These findings are probably an exaggerated estimate of the

degree of O-acetylation since Perlmann has reported that the

epsilon-N-acetyl groups of the lysine residues of pepsinogen
are also split by hydroxylamine treatment. Furthermore,

direct chromatographic analysis of tryptic and pronase

digests of labeled histones according to Gershey at al.

indicates that 80% of the radioactivity was present as

Using similar methods, Vidali et al. have also reported

that the radioactive acetate of isolated histones acetylated

in calf thymus nuclei in vitro can be recovered as a single

chromatographic peak which was identified as epsilon-N-

acetyl lysine.9 Since the latter represent more careful

and reliable studies, it is probably safe to assume that N-

acetylation is the physiological mode of acetylation which

occurs in vivo, whereas O-acetylation is an artifact of the

spontaneous reaction. The results reported in this study

show only a 3% release of acetate from "arginine-rich"

histones acetylated by DNP upon exposure to hydroxylamine.

The conclusion that the DNP catalyzed acetylation of

histones approximates the natural process and does not

represent an artifact is supported by the finding that the

in vitro pattern of labeling, in which the most actively

acetylated fractions are f2al and f3, is similar to that

obtained in vivo after administration of 1C-acetate.0

Therefore, in studies using whole animals, nuclei or DNP,

only the arginine-rich fractions are capable of becoming

acetylated whereas the lysine-rich (fl and f2b) are not.

A further similarity between the acetyltransferase

acting in DNP with acetyltransferring enzymes described by
others is the lack of a Mg requirement. Previous investi-
gations have shown that Mg is required for the transfer

of acetate to acetyl-CoA by a fraction from an acetone powder
extract from pigeon liver, but not for the transfer of
acetate from acetyl-CoA to p-aminobonzoic acid.

Similarly Gallwitz showed that there is no Mg required for