{Fiat-25;. ‘1"
lllllllllijlfllamllllljjllfflllllfllfllflllll ”gram...
University

THE GENOMIC ORGANIZATION, STRUCTURE
AND EXPRESSION OF THE CHICKEN
H2b HISTONE GENE FAMILY

presented by 0

David Kilgore Grandy

has been accepted towards fulfillment
of the requirements for .

Doctor of Philosophy degree in Microbiology and

 

 

Public Health
6 a r professor O
. dodgson

MS U is an Affirmative Action/Equal Opportunity Institution 0-12771

Date 3-1L-85

 

MSU

LIBRARIES
——

v

 

 

 

RETURNING MATERIALS:
Place in book drop to
remove this checkout from
your record. FINES will
be charged if book is
returned after the date
stamped below.

 

 

 

dfi284fig;
35997 ’87 ass

Ifi"->~‘-.-V‘wo‘ 'l"

fie? 3257

'l

 

 

 

THE GENOMIC ORGANIZATION, STRUCTURE
AND EXPRESSION OF THE CHICKEN

H2b HISTONE GENE FAMILY

By

David Kilgore Grandy

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Microbiology and Public Health

1985

To Laurel and Madeline

ii

ACKNOWLEDGMENTS

I would first like to thank Dr. Jerry Dodgson for providing a
very pleasant and stimulating environment in which to learn and work.
The unselfish sharing of his knowledge, time and financial resources
has made the completion of this research project, in my 4% years in
the Department of Microbiology and Public Health, possible. I wish
him the best in the future.

I should also thank Dr. J.D. Engel and Dr. B.J. Sugarman for
sharing their ideas and unpublished data with me.

I would like to thank my committee members, past and present, who
have offered guidance and assistance during my tenure as a graduate
student: Drs. Conrad, Fluck, Fritsch, Kung, Magee, Miller, Patterson
and Robbins. I am especially grateful to M. Fluck who has stuck it out
with me to the end and P.T. Magee, who almost did.

I would like to acknowledge the support, shared ideas and
friendship of the members of the Dodgson lab: K. Barringer, P. Bates,
P. Boyer, D. Browne, D. Brush, P. Creature, M. Federspiel, J.D. Lee,
W. Lewis, S. Stadt and C. Yoshihara.

A special note of thanks is due my parents, H. Bruce and Betty
Grandy, for literally and figuratively making me what I am today; my
grandfather C.C. Grandy who encouraged my early interest in natural
science; and my other surviving grandparents J. Grandy, M. Kilgore

and L.D. Kilgore for their interest and support.

iii

Finally I thank Laurel for putting up with the trials and tribula-
tions of the most hectic four years of our lives; for her love and
support; for typing a large portion of this manuscript; and for Madeline
(who was a doll in spite of all the commotion). I would also like to
thank Hilary, Maxx and Maynyrd for offering playful diversions when
they were needed.

Chapter 2 is reprinted from Gene Expression (PP.445-455,1983) with

 

the permission of the publisher, A.R. Liss.

iv

TABLE OF CONTENTS

LIST OF TABLES ..... . .............. . ................................ vii
LIST OF FIGURES ................................................... viii
DEFINITION OF HISTONE ............................................... x

INTRODUCTION........ ................................................. 1
The Histone Proteins.................. ......... ............ ..... 2
Biochemistry.. ..... ........... ......................... ....2
Histone protein function....................... ..... .......3
Histone protein variants........................ ........... 7
Secondary modifications of histone proteins.... ..... ......11

The Histone Genes.................................. ........ ....13
Organization....................... ...... ....... ...... ....13

Sea urchin............................... ..... .......13
Drosophila melanogaster.................. ..... .......18
Fungi........................... ..... ................20
Tetrahymena sp. and Artemia sp.. ..... ...... ..... .....22
Amphibia........................ ..... ... ..... ........22

Higher eucaryotes................... ..... .. ...... ....23
Plants.................................... ........ ...28

Fine structure................................... ..... ....29
Histone genes-5' flanking sequences..................30

Histone gene coding regions..........................33

Histone genes-3' flanking sequences..................34
Regulation of histone gene expression........ ....... ......36
Literature Cited...............................................42

 

 

CHAPTER 1. COMPLETE NUCLEOTIDE SEQUENCE OF A CHICKEN HZb HISTONE

GENE........... ...................................................... 49
Abstract.............. ...... .. .......... .... .................. .49
Introduction................ ........ ...... ....... ..............50
Experimental Procedures..................... ..... .. ......... ...52

Material.......................... ..... ......... ....... ...52
Preparation and characterization of pKRla-1.3.............52
DNA sequencing............................................53
Results and Discussion.........................................53
Preparation of a chicken H2b gene plasmid.................53
DNA sequence analysis of pKRla-l.3 ...... ..................55
Sequence of gene-flanking regions............. ..... .......57
Acknowledgments.............................. .............. ....60
References..... ............... .. ........ .. .............. .. ..... 61

V

PAGE

CHAPTER 2. THE CHICKEN H2b GENE FAMILY .............................. 63
Abstract..... .......................................... . ....... 63
Introduction. 0 O O O I O O O O O O O O O O O O O O O O O O O O O O O O O I O O O O O O O O O O O O O O O I O O .64
Results and Discussion... ...... .... ....... .....................65

Analysis of the complete chicken H2b gene family..........65
Construction of a chicken histone H2b gene-containing

plasmid....................... ..... . ....... ..... ........ 67
DNA sequence analysis of pBR2e-3.5........................67
Sequence conservation in the 5' gene-flanking

regions Of H2b-la and H2b-26............... oooooooooo .0069
Acknowledgments............................. ........ .. ......... 73
References............................................ ......... 74

CHAPTER 3. GENOMIC ORGANIZATION, STRUCTURE AND EXPRESSION OF THE

CHICKEN H2b HISTONE GENE FAMILY ......... . ........................... 76
Abstract............ ..... ........... ............. .. ............ 76
Introduction........ ..... ........... ............... .... ..... ...78
Materials and Methods..........................................81

Materials..................................... ...... ......81

Isolation and purification of recombinant DNA molecules...81
Filter hybridization......................................82
Restriction enzyme site mapping...........................83
DNA sequencing............................................83
Sl mapping analysis.......................................84
Results........................................................84
Estimate of the number of chicken H2b histone genes.......84
Isolation of the chicken H2b histone genes................88
DNA sequencing strategy of the H2b histone genes..........92
Chicken H2b coding sequences..............................95
5' sequences flanking the H2b histone genes..............100
81 mapping of the pKRla-l.3 transcript...................104
Chicken H2b histone gene 3' flanking sequences...........105
Identification of nonallelic sibling H2b histone genes...108
Discussion....................................................ll3
References. ........... ..... ............. . ...... ..... .......... 123

CONCLUSIONS ........................................................ 126

vi

TABLE
INTRODUCTION I
CHAPTER 3 I

LIST OF TABLES

PAGE
Classes of histones....... ....... .....4

Comparison of the H2b histone exons and
the proteins they predict with respect

to pKRla-1.3.........................101

vii

LIST OF FIGURES

FIGURE PAGE
INRODUCTION 1 Comparison of the genomic organization
of histone genes from a variety of
species ................................... l6

2 Genomic organization of the chicken
histone genes ............ . ................ 25

3 Genomic organization of a mouse
histone gene cluster on MM221 ............. 27

4 Genomic organization of the human

histone genes.. ........................... 27
5 Histone gene 5' homology blocks ........... 31
CHAPTER 1 1 Restriction maps of ACHla and pKRla-1.3...54
2 The complete nucleotide sequence of the
chicken H2b gene ..... . .................... 58
CHAPTER 2 1 Identification of the chicken H2b histone
gene family ..... . ..... . ...... . ............ 66

2 Restriction maps of ACHZe and pRR2e-3.5...68

3 The nucleotide sequence of two chicken
H2b histone genes ....... . ................. 7O

4 Sequence and spacing homologies found
in the 5' flanking regions of two chicken

H2b genes ............... . ................. 72
5 A conserved nine nucleotide sequence
unique to H2b genes.... ........ .... ....... 72
CHAPTER 3 l Genomic Southern blot of EcoRI/HindIII
digested chicken red blood cell DNA ....... 87

2 Restriction enzyme cleavage maps of the
ten chicken H2b histone gene-containing
phage ..................................... 91

viii

PAGE
Restriction enzyme maps of the
chicken H2b-containing subclones... ....... 94

Chicken H2b exon sequence comparison......97

Distribution of nucleotide base changes
observed in the eight chicken H2b
histone genes with respect to pKRla-1.3...98

DNA sequence analysis of the 5' flanking
regions of eight chicken H2b histone
geDESooooooooooo ...... ......OOOO... ...... 103

S mapping analysis of the pKRla—1.3
1 .
transcrlpt... ...... 0.0.000000000000000107

DNA sequence comparison of the 3' flank-
ing regions of nine chicken H2b histone
geneSoooo00000000000000.0000...one...0000110

Alignment of the phage maps which

contain nonallelic, sibling H2b histone
genes....................................112

ix

HIS'° TONE - n.[<Gr. 1010C , histos (see histo-); the warp,
loom, weave; a web; a ship's mast, lit. 'that which causes
to stand'; + -one]; any of a group of simple proteins that
yield a high proportion of basic amino acids on hydrolysis;
they are often poisonous when injected into an animal and
prevent the clotting of blood; found associated with deoxy-
ribonucleic acid (from Klein, F. 1966. A Comprehensive Ety-
mological Dictionary of the English Language. Elsevier,
Pub., New York).

INTRODUCTION

The histone genes code for a family of polypeptides which
constitute the primary protein component of eucaryotic chromatin. Indeed,
they were the first chromosomal proteins identified (1) and have
received considerable attention since their discovery. Our present
understanding of these small, basic, evolutionarily conserved proteins
places them at the center of the structural framework around which
eucaryotic DNA is organized: the nucleosome. However, interest in
histones is not just limited to this aspect of their function but also
includes several basic questions of molecular genetics concerning, for
example: the evolutionary mechanisms that shape genes and gene families;
the developmental and cell cycle regulation of gene expression; and the
relationship of a gene product to its function.

The first section of this review is intended to briefly summarize
the great body of literature amassed on the histone proteins and their
function as currently understood. This background information will
then serve as a point of departure for the remainder of the review
which will focus on the literature concerning the organization, structure,

and regulation of the genes encoding the histones.

The Histone Proteins

Biochemistry

Early biochemical studies indicated that chromatin contained
roughly equal masses of DNA and basic proteins which were named histones
(2). The initial attempts to fractionate these proteins after their
release from chromatin resulted in a plethora of poorly resolved bands
in polyacrylamide gels. As better preparations of chromatin became
available and band separation improved (after modifying the electrOphoresis
conditions, 3,4) five major classes of histone proteins came to be
recognized (5). These electrophoretic fractions were originally known
by many different nomenclatures but today they are universally referred
to as : H1, H2a, H2b, H3 and H4. Representatives of each of these five
classes of histone have been found in every eucaryotic species

examined with the exception of S, cerevisiae, which lacks H1 (5).

 

Although this situation is unique, it must be recalled that not all
genera have been characterized with respect to their histone content.
It is noteworthy, in this regard, that very little work has been done
on plant histones.

A sixth type of histone, H5, is also recognized. The H5 histones
are unique in that they are limited to species of birds, fish, reptiles
and amphibians (5). It is likely that H5 represents a highly divergent
member of the H1 histone class.

As a result of technical improvements in protein fractionation and
the fact that they are relatively abundant proteins, the isolation and
subsequent amino acid sequencing of histones from many different species
(mostly animal) has been accomplished (5). Upon analyzing the sequence

data from several species it was noted that the histones are

evolutionarily very stable. The arginine rich H3 and H4 histones show
the most sequence conservation. In fact, only two amino acid changes
distinguish a pea H4 histone from a calf thymus H4 (6) and only four
changes separate a pea H3 histone from a calf thymus H3 (7). The lysine
rich H2a and H2b histones exhibit sequence conservation of roughly 80%

and 70%, respectively, when genera as divergent as Saccharomyces (yeast)

 

(8), Parechinus (sea urchin) (9,10), Gallus (chicken) (11) and 30$ (12)

 

are compared. In addition, these four core histones,as they are known,
display a common primary sequence organization. The amino terminal
third of each protein contains most of the molecule's charged amino
acids and the carboxyl two thirds is composed mostly of hydrophobic
residues (5). This shared organization suggests that these histones
themselves derived from a common ancestral protein (13).

Members of the remaining classes of histones, including H1 and H5,
are lysine rich and display the least sequence conservation when genera
are compared. Table 1 presents a list of five of the six histones and

characteristics of each.

Histone Protein Function

The high degree of amino acid sequence conservation maintained
over millions of years of evolution from the simplest protozoan to the
most complex metazoan suggests that the core histones have been under
strong selective pressure. This implies that their sequences,and there-
fore their structures,are extremely important to their function in
chromatin.

Our understanding of histone function was greatly expanded by the

discovery of the nucleosome (14,15), the basic subunit of chromatin.

.omaH .xuo» Bmz .mcom mam zmawz .w .coauawm mam .HH :Ofimmmuaxm mcmu.mw mom .a .m .cHBuA Scum amxmwv

 

 

 

 

 

~m~.HH moa mac NmH Nmm n.~ No.oH m.o No.qH NHH Boo A>H.Hm~chm
~sm.m~ mMH mac ~-H Nan m.~ Nm.mH a.o Nm.m~ NOH zoo fiHHH.mmvm:

umm Na
oom.o~ nH< Nofi Nm.mm m.~ Nm.mH w.¢ No.q Nwa mom I

How NHH
q~m.m~ mNH mH< NOH Nm.~N m.~ Nmfi m.~ Nm.o Nofi zoo ANQHH.n~mVn~:
oom.o~ sac NHH
coo.mH mH< Nsa Nam m.H Nan m.~ Nm.a NHH man .

saw NHH
00¢.MH mNH mH< Nma mom m.H Nm.mH N.H Nm.m “Hg zoo AanHH.~mNmV<Nm
ooo.mm MHN ma< Nmm Nofi q.m Nm.m Hm Nn.H Now mannmu aH .HuvH=
ucwfim3 mcaum mmuzumww mvfium afiunu mvfium wu<\m%A wu< mhq mufiummm mcoumfim

unasumaoz ocHEm umnuo o:HEm uwwfium oafism
umaasz umaoa< uwmmm ofipfiu< mpfium ocfiam dammm

 

mmZOHmHm mo mmmm<qo .H MAQ<H

5

Extensive studies, now at the 7 2 level of resolution (16), have shown
that the nucleosome is composed of an inner core of eight histone
proteins around which roughly 146 base pairs of DNA are wrapped. Recon-
stitution experiments (17), crosslinking data (18) and X-ray crystallo-
graphic data (19) taken together suggest that the core octamer is
composed of an H3:H4 tetramer and two H2a:H2b dimers.

Of special note is the report by McGhee and Felsenfeld (15) that
the amino terminal portion of each core histone is external to the
core particle,based on nuclear magnetic resonance, sensitivity to
trypsin digestion and antibody binding results. Through ip_vi££g
reconstitution experiments the ability of trypsinized core histones to
assemble into complexes resembling nucleosomes has been demonstrated (20).
These results suggest that the basic amino terminal portion of the core
histones are of little functional importance to nucleosome assembly, at
least ipuvitrg. This suprising conclusion was supported and extended
through in viva studies in which a series of yeast H2b gene deletion
mutants were tested for their ability to rescue cells deficient in wild
type H2b (21). Shortened H2b protein lacking amino acid residues 1-31
were found to be completely functional with no detectable phenotype.
However, proteins lacking amino acids 13-70 were nonfunctional. Although
these findings imply that the N-terminal portion of the H2b protein is
dispensible (in yeast, at least) it does not mean it lacks a function
12.2122: It has been suggested that the core histone's extended 'arms'
are involved in the formation of higher order chromatin structures (2)
and that the ability of these structures to form would be destroyed if
more than one class of histone protein were deficient at its amino end.

Furthermore, studies of fusion genes encoding the amino terminus of a

nuclear protein and a carboxy terminal nonnuclear marker protein
demonstrates that the amino end of the marker protein can target the
hybrid polypeptide to the nucleus (22).

With regard to the hydrophobic portion of each histone protein, the
results from the H2b studies underscore the importance of this region
which is involved in protein-protein interactions.

The strong selective pressure on the core histones can now be under-
stood in light of the nucleosome's structure. Any alteration in the pri-
mary structure of these proteins could have an effect on their secondary
and tertiary structures which might, in turn, interfere with their ability
to assemble into a functional core particle.

With regard to the H1 and H5 histones, it has been argued that they
are under less selective pressure than the core histones because they
are located on the outside of the nucleosome, in association with the
DNA found between particles, the linker DNA (2,16). In this position
their function presumably depends on fewer critical protein-protein
interactions than are required of the core histones, thereby permitting
a greater degree of coding freedom. One of the functions proposed for
H1 histones is the 'sealing' of the DNA on the nucleosome where it enters
and leaves the particle. This is suggested by the observation that H1
can be crosslinked to core H23 histones (23,24). Another role for H1
histones is suggested by reconstitution experiments which demonstrate
Hl's involvement in the formation of the 30 nM fiber which consists of
six nucleosomes arranged in a coil (5).

The H5 histones are thought to be involved in the compaction of
chromatin concomitant with the condensation of the nucleus in terminally

differentiating erythrocytes (25). This histone is only found in those

species whose erythrocytes retain their nuclei (5). During this
process H5 histones replace approximately 70% of the H1 molecules in
chromatin (26).

Histone Protein Variants

In spite of the extensive sequence conservation demonstrated by all
classes of histone from many different genera, when individual species
are examined the data reveals that a particular class of histone may be
composed of a number of variants or subtypes. This finding has led to
the development of the concept of histone families. Each family is
composed of individual members that display a significant degree of
similarity. However, microheterogeneity of a few to many amino acid
residues may distinguish one element from another. The corollary to
this concept is that a family of genes must exist which encodes these
histone variants. In some cases histone variants can also arise from
post-translational modifications (5).

The results from many studies make it clear that some variants
only appear at certain stages of an organism's development or during
the maturation of particular tissues. The best characterized processes
during which histone variants are expressed include embryogenesis, sper-
matogenesis and erythropoesis in sea urchin, chicken and mouse species.

The developing sea urchin embryo undergoes many rapid cleavage
divisions following fertilization. The need for an adequate supply of
histones during these very early cell divisions is met by the translation
of stored maternal histone mRNAs (27,28). After the first few cleavage
divisions, however, the main supply of histone protein is translated from

mRNA actively synthesized from these 'early' genes (29,30). These

variant proteins are then replaced by another set of histones known as
the 'late' or larval type histones (31,32). The synthesis of late
histones begins at the conclusion of blastulation (33) and they continue
to be synthesized replacing all other variants entirely in the adult
echinoderm's chromatin.

Recent data suggests that early chicken development is similar in
some respects to sea urchin embryogenesis. In the chicken oocyte a
store of maternal histones and histone messenger RNA is supplied to the
rapidly dividing cells of the embryo. In contrast to sea urchins the
embryonic histones continue to be expressed in the adult (34). However,
after hatching the population of embryonic variants begins to decrease
while a set of 'late' variants becomes more prevalent (34). These 'late'
histones have been characterized by their varying mobilities in gels (35).
The histone component of adult chicken chromatin, and mouse for that
matter, appears to be a function of the mitotic activity of the cell.
If a cell is continuously dividing (e.g.hematopoietic stem or intestinal
mucosa cells) then the early histone variants tend to predominate. On
the other hand the 'late' histones predominate in long-lived, less
mitotically active cells (eg. liver and kidney) (35).

One example of tissue-specific histone variant expression is
found in the gonads of echinoderms and vertebrates during spermatogenesis.
In these species spermatocyte-specific subtypes of H1, H2a, H2b and H3
are expressed during meiotic prophase (36,37) and preferentially replace
the resident replication variants (see below) (34). Their roles
presumably involve the extensive condensation of the sperm's chromatin.

The maturation of nucleated erythrocytes in species of birds, fish

reptiles and amphibians represents an unusual, but well documented
example of tissue-specific histone expression (5,25). In these animals
H5 synthesis is uncoupled from DNA replication and becomes the only
histone synthesized in these cells after cell division stops. Although
H5 histone exists only in non-mammalian vertebrates, it does share
substantial amino acid sequence homology with mammalian H1 variant
H10 (5,38).

Recently Zweidler has proposed that the histone variants define
four sets of histone proteins which differ in their relationship to
DNA replication (34). One set of variants is synthesized in the oocyte,
uncoupled from DNA replication. These histones are made from maternal
mRNA and are used during the first few cleavage divisions, characteristic
of echinoderms, chickens and some amphibians. Another group of variants
is defined as being strictly replication dependent. These proteins,
expressed throughout the life of an organism, are induced during S phase
and synthesized coordinately with DNA replication. These are called
replication variants. A third set of variants is found to be expressed
throughout the cell cycle and have been named replication independent
or replacement variants by Zweidler. These histones are first found in
the early stages of embryogenesis and continue to be expressed through
adulthood. Their synthesis is not affected by inhibitiors of DNA
replication. The ratio of replacement to replication variants is thought
to be determined, to some extent, by the cell's mitotic activity. Recent
experiments performed on regenerating liver (34) and mouse erythro-

leukemia cells (39) have confirmed this prediction: the rapidly dividing

cells contained a larger proportion of replication dependent variants

10

in their chromatin and replacement subtypes predominated in less
mitotically active, mature cells. The fourth set of histone variants
is expressed in a tissue-specific manner. These include the spermato-
cyte-specific Hl.S, H2a.S, H2b.S and H3.S histones and the erythrocyte
H5 histones.

A great deal of work has been devoted to understanding the signifi—
cance of histone variants. One argument suggests that they merely
represent the evolutionary divergence of multiple histone genes. How-
ever, the strength of this argument is weakened to some extent in light
of the tissue-specific variants already discussed. Although the evidence
is less striking for the remaining sets of subtypes, it is tempting to
speculate that each one has been maintained to serve some important
function.

One role for variant histone proteins is suggested by evidence that
indicates newly synthesized DNA becomes associated with newly formed core
particles (40-42) in a conservative (strand-specific) manner. These
core particles could be composed of histone variants which might affect
the functioning of the nucleosome in such a way as to alter the chromatin
structure, for example affecting the accessibility of its DNA to trans-
cription. The conservative nature of the association would lead to a
specific alteration of chromatin structure between the two daughter cells
after cell division.

A promising approach to studying the relationship of chromatin
structure to function has been developed by Weintraub and others (2).
Using nucleases, such as DNase I, they have demonstrated that the

enzymes' ability to nick DNA is a function of the state of the chromatin.

11
Furthermore, genes which are actively transcribed are more susceptible
to enzymatic attack and these domains of sensitivity surround the gene
for roughly several kilobases (2). The nature of this nuclease sensitiv-
ity has been found to be partly attributable to the nonhistone proteins
HMG l4 and 17. These proteins recognize some structural feature in the
nucleosome particle associated with transcriptionally active DNA (2).
One hypothesis suggests that the HMO proteins recognize changes in the

core histones, perhaps particular variants.

Secondary Modifications of Histone Proteins

All histones undergo secondary, posttranslational modifications.
One such modification is acetylation of the amino terminal serine of
H1, H23 and H4 as well as specific N-terminal lysine residues of H3,
H4, H2a and H2b (5). Hyperacetylation has been correllated with an in—
crease in DNase sensitivity, but since this type of acetylation is transi-
tory, appearing during S phase and lost in G2, its significance remains
a mystery (5).

Phosphorylation is another common secondary modification of histone
proteins. All of the histones have been analyzed with respect to phosphor-
ylation and were found to be phosphorylated to some degree (reviewed in
5). The phosphate moiety is covalently attached to specific threonine,
serine and occasionally histidine (H4) and lysine (H1) residues during
S phase, mitosis and in response to hormonal stimulation. In some cases
dephosphorylation of H1 and H3 occurs after anaphase. Gurley et a1.

(43) have suggested that the various organizational levels of chromatin
can be correlated ‘with different forms of phosphorylation. As an

example, if phosphorylation involves H1 then a change in the extent of

12

chromatin compaction (coiling) might be the result. In addition, the
phosphorylation of H23 has been shown to accompany heterochromatin con-
densation (43).

The methylation of histones is reversible (44) and occurs during
late S phase or G2 (45). Only particular lysine residues are methylated
and this probably occurs after the nucleosome has been assembled; how-
ever, this remains to be confirmed (5). A role in altering DNA-histone
interactions has been proposed for histone methylation but, again, this
remains unsubstantiated (46).

The covalent linkage of poly (ADP-ribose) to glutamate or aspartate
residues of H1, H23, H2b and H3 has been recognized for some time (5).
Although poly(ADP) ribosylation has been proposed to affect histone-
histone interactions, no conclusive evidence is yet available.

The most unusual secondary modification of a histone is the link-
age of ubiquitin, through an isopeptide bond, to the terminal amino
group of lysine 119 of certain H23 proteins (47). This complex is
known as A24 (48). Ubiquitination of 3 small number of H2b histones
has also been demonstrated (49). Ubiquitin, an acidic protein of 76
amino acid residues, is one of the most evolutionarily conserved proteins
known and is found in both eucaryotes and procaryotes (50). Ubiquitin-
ation can affect as much as 10% of a cell's H23 component and about 1%
of its H2b molecules (51). The degree to which a cell's H23 histones
are found as A24 appears to be correlated with the cell's mitotic
activity such that nondividing cells contain more A24 than dividing
ones. Furthermore, the results of studies performed on chicken

chromatin suggest that ubiquitinated H23 histones tend to be

13

associated with nucleosomes involved in DNA transcription and are
essentially excluded from heterochromatin (52). In the cytoplasm
ubiquitin has been reported to tag proteins which will be targets of
proteolytic enzyme degradation (53). The exact nature of either A24's

or UH2b's role in the nucleus remains to be clarified.

The Histone Genes
Organization

Sea Urchin

The study of the histone genes began when Borun, Scharff and Robins
(54) and Gallwitz and Mueller (55) identified 98 mRNA, in mammalian
tissue culture cells, that was associated with polysomes engaged in
histone synthesis in S phase. The criteria of size, time of synthesis
and association with ribosomes were then used to identify similar 98
RNA species in sea urchin embryos (56, 57). Sea urchins were well suited
for such studies because a great deal of developmental biology was
known about them, large numbers of synchronized embryos could be ob-
tained and these embryos could be easily manipulated physically and
biochemically.

_During the early stages of sea urchin embryogenesis a series of
rapidly occuring cell divisions take place and the embryo's requirement
for histone proteins is at a maximum. It was known that about 60% of
the embryo's translational activity at this time of development is de-
voted to the synthesis of nuclear proteins (56, 58). Therefore, the
sea urchin embryo represents an enriched source of histone mRNA.

Using unfractionated, and then fractionated embryonic 9S mRNA,

radiolabled as probe, several investigators were able to demonstrate

14

by solution hybridization kinetics that the genes coding for these RNAs
are repeated several hundred times in the sea urchin genome (59). Proof
that these genes were in actuality histone genes was supplied by comparing
in_zitrg_translation products (60) and RNA sequence results (61) to
histone amino acid sequence data.

The first analysis of the genomic organization of these highly
reiterated genes was performed by subjecting high molecular weight sea
urchin DNA to equilibrium density gradient centrifugation (59,62). At
equilibrium all of these histone genes were found to band as a satellite
fraction, suggesting that they were extensively linked to one another.
This interpretation was confirmed when the detailed maps of these sea
urchin genes were determined using recombinant DNA technology.

When individual histone mRNAs were used to probe cloned and high
molecular weight genomic DNA the remarkable finding was made that all
five histone genes are organized in clusters 5-7 kb in size (60) that
are tandemly repeated 300 - 600 times in the genome (63). Electron
microscopic studies of mRNA/DNA hybrids (R-loop mapping) was then used
to map the individual histone genes on cloned fragments (64). These
assignments were later confirmed by DNA sequencing experiments (29,

30).

The picture which has emerged from these studies is diagrammed in
Figure 1A. The salient organizational features are that: the early
sea urchin histone genes are clustered in the order from 5' to 3': H4,
H2b, H3, H23 and H1; each gene is separated from its neighbors by
intergenic spacer DNA that is AT—rich, of varying length and which does
not encode protein; the clusters of genes are 537 kb in length and they

are tandemly repeated several hundred times in the genome; the genes in

15

Figure 1. Comparison of the genomic organization of histone genes from
a variety of species.

A.

Genomic organization of the major early histone gene repeat and larval
stage H3 and H4 histone genes of the sea urchin L, pictus (from
Roberts, S.B. et al. 1984. J. Mol. Biol. 174:647-662.).

. Genomic organization of the major D, melanogaster histone gene repeats.

 

The site of the 270 bp insertion element found in the more prevalent
Skb repeat is indicated. The open horizontal arrows point in the di-
rection of transcription (from ref. 71).

. Genomic organization of the entire S. cerevisiae histone gene com—

plement. The histone genes are shaded and the points indicate the
direction of transcription. p1,2,3,4 and SMTl do not code for his-
tone proteins (from ref. 81).

 

. The genomic organization of two of the histone gene repeats from the

brine shrimp Artemia sp.(from ref. 85).

. The genomic organization of the major 5.8 kb X. laevis histone gene

repeat. The arrows indicate the direction of transcription (from ref.86).

. The major histone gene repeat of the newt N, viridescens. The horizon-

 

tal arrows above and below the histone loci point in the direction of
transcription (from ref. 87).

16

 

 

—o -9 —b —9 —9
HI H_d "—28 H3 HZA
early r m H H- ' i I?
IOOO boss-pairs
'—-——t o— —-o
A H4 H3
‘ larval F {E i H H _on8 S: 1: 5H to a

IOOO base-pours
t——-l

Band-11(8), EwRHE). Hindlll(H). HinclHHi) SslHS). xmuxm XhoHX)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mseanou
v-fi
\‘ '1
\l
8' H1 H3 H4 H2a H2b H1
W
A—T RICH v
REPEATED ....1kb_,.,
' SEQUENCE
P R a p1 HZA H233 a P
P l J l I
TRT1
l.
1.. 1 l Cl“ 80’- D Cl ,
TRTZ TRT3
C. 3 P P Harman a 3
L4 ' W L .4
5:201 'Sczoz B = BamHI,
5:82 I; = EcoRI.
= Pstl,
”,srmnam. a u
5:191
xasnz t I v u
l "“~--..._. nag;
D’ 9P5" ' “ i i N’ l-Hoelll
was CECE] cm
xasm- * t i o i
W m
_§
'3' RH: H4 H2A H28
um
_.""’ H1 H3 H2A H4
"-.
“——-{ j-...
E. ‘—
H .3

Figure 1.

17

a cluster are all transcribed from the same strand of DNA but not as
a polycistronic message; and none of the histones contain intervening
sequences.

These early histone genes become transcriptionally active after the
first few cell divisions following fertilization and it has been estimated
that given a moderate rate of transcription they can supply all of the
message required through blastulation. However, after blastulation,
when the rate of cell division slows, a switch in histone gene expression
takes place. At this time the early embryonic genes are turned off and
the late or larval types become transcriptionally active (33).

Based on the length of their mRNAs, the melting characteristics of
late RNA/early gene hybrids and oligonucleotide fingerprinting (33,65-
67) the late mRNAs were predicted to be transcribed from a different set
of genes than the early histones. Using the early histone genes as
probes under mild hybridization and washing conditions the late histone
genes have been identified and some of them have been cloned (68).
Genomic blots indicate that these genes are present in 5-12 copies per
haploid genome and are organized in a very different way than their
early gene counterparts. Furthermore, DNA sequence analysis has
confirmed that these genes have diverged considerably with respect to
the early histone gene sequences.

Recent evidence has revealed that not all of the sea urchin histone
genes fall into either the early or late gene clusters. These 'orphons',
as they have come to be called, appear to be located in moderatedly re-
petitive sequences throughout the genome (69). Their location in the
genome of one animal may differ from their lbcation in another animal.

Whether these genes are functional or are pseudogenes remains to be

18

demonstrated.

Drosophila melanogaster

 

The extensive histone protein amino acid sequence conservation
across species barriers suggested that the mRNA from one species should
cross—hybridize, to some extent, with sequences of another species.
Based on this premise Karp and Hogness used sea urchin early histone

mRNA sequences to identify D, melanogaster histone genes. They

 

screened a library of genomic DNA fragments cloned in ColEl plasmids
under low stringency conditions and obtained a clone which hybridized
to the sea urchin probe (70). This clone, designated cDMSOO, is about

9 kb in length. It was mapped with restriction enzymes and the organ-
ization of the histone genes it contains was determined by hybridizing
fly poly A- mRNA from embryos and tissues to the cDM500 clone (71).

The selected RNA was eluted and five different RNA species were purified
by gel electrophoresis. The region of cDM500 which encodes each RNA
was identified in hybridization experiments. 'DNA sequence analysis of
the various hybridizing regions was performed permitting the assign-
ment of 3 particular histone gene to its appropriate map coordinatesjfi
The cDMSOO clone was found to code for each of the five major histones.
They are organized into a cluster whose length is about 4.8 kb and
which was predicted to be repeated approximately 150 times in the
fly's genome. The results of R-loop mapping indicated that the genes
contained on cDM500 are not all transcribed from the same strand of

DNA (71). When the DNA sequence analysis was complete, the R-loop

data was confirmed. The five histone genes of D, melanoggster are

 

arranged differently than in sea urchin, being H1, H3, H4, H23 and H2b.

19

The order and direction of histone gene transcription is given in Figure
13.

When cDMSOO was used to probe 3 D, mealanogaster genomic blot, two \

 

types of repeating elements were identified, one corresponding to the
4.8 kb cDM500 repeat and the other to a repeat approximately 270 bp
larger (71). On fine structure analysis the second type of repeat
displays the same organization as cDM500 but is represented about three
times more frequently. The copy number of this larger more frequent
repeat was estimated to be in the range of 110 copies per haploid

genome based on reassociation kinetics. The chromosomal location of

 

most, if not all, of D, melanogaster histone sequences is on the left
arm of chromosome 2 at position 39DE (72). How the two types of repeats
are ordered with respect to each other is only partially known. The
available evidence suggests that the organization is not random; the
two repeats are interspersed and occur in small sets or clusters which
are interrupted by spacer sequences of unknown constitution or
function (71).

In order to address the significance of maintaining two nearly
identical sets of histone genes investigators have relied on molecular
biological and classical genetic techniques. At the molecular level no

stage-specific D, melanogaster histone variants have been identified

 

among these repeats by restriction enzyme polymorphism, DNA sequencing

or thermal elution studies (73). Drawing on genetic analysis, viable
flies have been bred which lack one or the other type of

repeat or carry a duplicated set of histone genes (74-76). Furthermore,
the chromatin structure is the same for both types of repeat in all cells

and stages of development examined (73). These findings indicate that

20

either type of histone repeat can supply the animal with the histones
it requires but clearly more work must be done if we are to under-
stand why two sets are maintained and how a fly compensates for

alterations in its histone gene dosage.

Fungi
The discovery of multiple forms of histone gene organization
in echinoderms and diptera prompted the investigation of other

genera. One obvious candidate for study was Saccharomyces cerevisiae

 

because extensive genetic analysis of yeast has been performed, it
undergoes developmental processes which can be controlled, its genome
is less complex than other eucaryotes and it is an organism amenable
to molecular biological manipulation.

The characterization of the yeast histone genes began when Hereford
et al. used a sea urchin early histone gene probe to screen a yeast
genomic library (77). They identified several clones which later turned
out to be false-positives. Therefore, a second approach was used
which took advantage of the fact that, unlike most histone messages,
yeast histone mRNAs are polyadenylated (78). Using low molecular
weight poly A+ mRNA to screen a yeast genomic library 100 clones were
identified. These clones were used to hybrid-select mRNA that was then
translated iguvitrg and the protein products were electrophoresed in
2D gels along with yeast histone standards. This approach led to the
identification of two, nonallelic H23—H2b loci (77). The H3 and H4
yeast histone genes were isolated from a lambda library by employing
a double screening procedure in which two sets of identical filters

were prepared (79). One set was probed with sea urchin histone DNA and

21

the other was probed with cDM500. One clone was identified in this
manner which contains both an H3 and an H4 gene. A second, nonallelic
H3-H4 pair was cloned by rescreening the lambda library with the
homologous yeast H3 and H4 genes as probes. The organization of these
genes is illustrated in Figure 1C. As was mentioned previously no Hl
histones have been found in yeast chromatin and to date no Hl genes
have been identified.

The entire complement of yeast histone genes has been cloned and
reveals an organizational pattern which is different from both echin-
oderms and diptera. The yeast histone genes are arranged in dispersed
clusters containing either an H23-H2b pair or an H3-H4 pair. In all
cases the pair of genes are divergently transcribed. Experiments in
which one or both of the H2b genes were inactivated by site-specific
frameshift mutagenesis have revealed that each H2b gene is equipotent.
Spores however, which lack both H2b genes can only germinate and bud;
no further development is possible (80).

The yeast histones are located on four nonallelic clusters per
haploid genome. These four histone loci are unlinked based on results
from two different approaches. The first relied on developing a
detailed restriction map for each lambda clone and comparing these maps
for regions of overlap. No overlapping areas were identified suggesting
that at this level of resolution the clusters were separated by 15-35 kb
of sequence (81). The second approach took advantage of the extensive
linkage data available for nutritional markers, tetrad analysis and
restriction site polymorphisms. Hereford et al. (77) using this

approach showed that the H23-H2b gene pairs are separated by at least

22

235 kb and may actually reside on separate chromosomes. Similar
results were obtained for the H3-H4 gene pairs (82).

Recently another fungus, Neurospora crassa has been studied with

 

respect to its histone genes. Using sea urchin or X, laevis H3 and H4
histone genes as probes, Woudt et al. (83) have identified a 6.9 kb
5311 genomic fragment that contains both H3 and H4 sequences. Fine
structure restriction enzyme mapping and DNA sequencing have revealed
that these two genes are divergently transcribed and that they contain
intervening sequences. Genomic blots indicate that there are one or

two copies of each of these genes per haploid genome.

Tetrahymena sp. and Artemia sp.

 

In addition to sea urchin, D, melanogaster and the fungi two other

 

lower eucaryotes have recently been examined with respect to their

histone gene complement. These organisms are the protozoan Tetrahymena

 

and the brine shrimp Artemia. The number and organization of the

Tetrahymena histone genes is not completely known, however several H3

 

and H4 genes have been identified and they appear to be linked in the
genome (84). The genomic organization of Artemia's histones displays
yet another organizational motif as illustrated in Figure 1D (85).

Whether or not these genes are reiterated and if so how many times and

in what arrangement are questions that remain to be answered.

Amphibia
The intermediate eucaryotes that have been investigated from the
standpoint of histone gene organization include the vertebrate genera

Xenopus and Notophthalmus. Xenopus laevis genomic DNA was probed with

 

 

the major sea urchin early histone repeat and a 5.8 kb EcoRI restriction

23

fragment was identified that hybridized to all four core histone gene
probes (Figure 1E) (86).This cluster of genes appear to be repeated

45—50 times in the frog's genome. Not all of these repeats are identical,
however, as evidenced by some restriction site polymorphisms. To date

no H1 gene has been identified in Xenopus.

The genomic organization of the major newt histone cluster was de-
termined by mapping blots probed with sea urchin histone sequences (87).
This repeat, shown in Figure 1F is present in 600-800 copies with each
cluster apparently identical at the level of restriction site poly-
morphisms. These clusters of genes are separated by reiterated satellite

DNA.

Higher Eucaryotes

The histone gene families of three phylogenetically more advanced
and genomically more complex genera, Gallus (88), Mu§_(89) and H929 (90),
have also been investigated. The early results based on solution
hybridization kinetics indicated that these organisms differed from the
other species that had been studied in that despite their genomic
complexity,they only contain 10-20 copies of each histone gene per
haploid genome. Furthermore, as the amount of fine structure informa-
tion has increased.the complexity of the genomic organization of each
species' histone complement has come to be appreciated.

The chicken genomic library constructed in l Charon 4A by
Dodgson et 31. (91) was screened with a sea urchin probe which contained
early H3 and H23 genes (92,93). After several cycles of screening
about 50 positive phage clones were identified and isolated (92).

Roughly half of these have been analyzed by restriction enzyme mapping

24

and in some cases histone-containing restriction fragments have been
subcloned and sequenced (94,95). This effort represents the most
detailed analysis of any eucaryote's histone complement, with the excep-
tion of yeast. One result of this extensive survey of the chicken
histone genes was the finding that many of them are clustered and each
cluster is different with respect to the genes it contains and how these
genes are arranged (92,94). Furthermore, it is clear that the various
clusters are not tandemly repeated in the genome although some clusters
are linked by a few kb. An example of some of the genomic chicken
histone clones is presented in Figure 2. In spite of the fact that most
clusters differ from each other, a few common elements of organization
tend to be repeated. One such arrangement is the pairing of H23 with
H2b genes such that they are divergently transcribed. Another common
pair is the H3-H4 gene couple. Furthermore, most histone genes in
chicken are found in clusters. The few genes that are genomically
isolated appear to be very different in the way they are regulated and
in their fine structure (see below).

When genomic chicken DNA is cut with a number of different restric-
tion enzymes, blotted and probed with individual chicken histone genes
it appears that the four core histone genes are represented roughly 10-
16 times each; H1 is present in about 6 copies while the H5 gene is
represented only once per haploid genome (96,97). This solitary H5 gene

is unlinked to any other histone sequence (97,98).

25

 

 

 

 

 

 

 

 

 

 

 

64%;» v *2» P' ACHIO
.. a #4
h
L I. J, 1‘ I" 1‘ ' Wi‘w j 9 \lr .9 I) ACHZO
9—3
{LL ’I‘ ’F v +_m_‘1 9 v w “7 ’l‘ ‘1’: 0
Q If ‘1" v win“; 9 v v 7 1‘ 159 0
{—1 u . .' o ACHZC
(E 9 4r v 10
a a up
as .... ..., . . p . .. .I—“m'h XCHZe

 

ACHSb

 

MEI“, \Ir ! sly t 9 Jr .:
g 331 I. . u.» ACHSc

L ,5
(.3 — 0
W ' . M

 

 

 

 

 

 

 

 

mfg—Ame 4“— 0 ACHIOO
E. , 3* .1, g g
L . 4' :fi a u XCHSe

I}
T F 5 ‘F ”r

G -EcoRI Luann-Jr - EcoRI, "anon, haul-1|; a meal, a «nu; 8 033a

 

 

£55 Momma firm Scouructo Ruoon
HI — I —9

H20 = zc—fi

H20 m Zb—fi

H3 m 3 —')

"9 animus 4-—+

Figure 2. Genomic organization of the chicken histone genes.
Restriction maps for 15 histone encoding recombinants
with the regions corresponding to various histone genes

indicated (from 94).

XCH5d
XCHSq

XCHZd

ACHSd

ACHSC

ACH7b

26

Mouse

First recognized in chicken, the clustered non—tandem organ-
izational pattern of the histone genes has also been demonstrated to be
a hallmark of mouse and human histone gene organization. A cDNA probe
prepared from the polyA- mRNA fraction of mouse cells was used to screen
a genomic library (99). The phage identified by hybridization with the
cDNA probe were then rescreened with a sea urchin H3 sequence. Two phage
hybridized to both probes and so were analyzed further by DNA sequence
analysis. One of the two clones codes for three histones: H3, H23 and
H2b. The H3 and H23 genes are transcribed from the same strand of DNA
whereas the H2b gene is divergently transcribed. The arrangement of
these genes is apparently unique to mouse, given the limited information
available. The other histone—containing clone may contain an H3 pseudo-
gene or a variant sequence.

When the individual mouse histone genes were used to probe genomic
Southern blots it was clear that each of the ten or so genes are not
tandemly repeated. Figure 3 illustrates the best characterized arrange-

ment of mouse histone genes in clone MM221.

Man

The human histone genes were originally identified in a A Charon 4A
library using sea urchin and chicken histone sequences (100,101). Ten
different clones have now been isolated. The arrangement of the genes
in each clone is unique but some common features do emerge (Figure 4).
H23 and H2b genes are coupled as are H3 and H4. One clone contains all
four of the core histones and the others contain incomplete clusters.

No unlinked genes have been discovered and only one H1 gene has been

27

 

MM221 l__s_£_. 1 I L;

H 3.2 Haul HBJ H20

J, = Eco R] I“
Q: 8311

Figure 3. Genomic organization of a mouse histone gene clus-
ter present on MM221. The horizontal arrows point in the di-
rection of transcription (from 99).

 

 

 

" m 9 n; g m to.
l l
41 Liil l l l [I I]
i .. I" ‘" " t
55 L l l l l I
(an) I .. I» r 'T m ‘57.;
m)

......LL 1 1 1
' 3.

EcoRI (l). Hind "1(1‘ ), and Ham HI (‘)

Figure 4. Genomic organization of the human histone genes.
-Ten lambda recombinants with the histone loci they con-
tain are shown (from 100).

28

isolated.

Plants

Oryza sativa (rice) is the only plant species for which any infor-

 

mation regarding its histone genes is available. A library of genomic

DNA was prepared and screened and a 6.6 kb fragment was eventually sub-
cloned that encoded a cluster of H23, H2b and H4 genes (102). The prelim-
inary results indicate that the H23 and H2b genes are closely linked.

This pair is unusual in that the two genes are transcribed from the same

strand of DNA.

Evolution of the Histone Genes and Gene Families

As a result of the extensive, albeit incomplete, survey of the
histone genes and gene families it is apparent that the simple concept
of a typical organization is no longer valid. Each species is unique
and so must be considered on an individual basis. How this variety
evolved is an interesting question to ponder but beyond our current
ability to answer. We can speculate though as to what types of mechan-
isms may have acted in the past and continue to act, on the histone
genes and gene families.

It seems possible that the gene coding for an ancestral histone
protein became duplicated as a result of a chromosome breakage event
followed by replication of the fragments and then their cross-union
resulting in a tandem duplication. Once this rare duplication occurred,
the more frequent recombination process of unequal crossing over could
have been responsible for the multiple reduplications which characterize
the larger histone families. Each gene would then be subjected to

mutation events which could, over time, and under selective pressure,

29

result in considerable sequence divergence eventually giving rise to
the various histone sequences recognized today.

However, another recombinational mechanism, in addition to unequal
crossing over may have played an important role in the evolution of the
histone gene families. This process is known as gene conversion and as
its name implies it results from the conversion of one nucleotide sequence
to another (reviewed in 103). Gene conversion can occur under a number
of different circumstances including some which exclude unequal crossing
over. One result of gene conversion is the homogenization of sequences.
This may be how the extensive sequence similarity seen among the
large histone families has been maintained over millions of years.
Furthermore, since the locations of the interacting sequences do not
change nor does the number of genes, conversion may also explain how, as
homogenization occurs, the copy number and organization of the family

is maintained.

Fine Structure
The analysis of histone gene nucleotide sequences has proven to be
of considerable interest. Not only has such information been used to
predict amino acid sequences and locate specific protein coding regions,
but the comparison of many genes has allowed the identification of a
number of conserved structural elements. These consensus sequences can
be analyzed in functional assays in order to determine what role, if any,

they play in regulating histone gene expression.

30

Histone Genes-5' Flanking Sequences

The region which lies 5' of the protein encoding sequence has been
found to contain most of a gene's transcriptional regulatory sequences
(104). Recently a downstream regulator of transcription was reported
for a human B-globin gene (105) and for yeast H23 and H2b (106).

When the 5' nucleotide sequences from a number of histone genes
of different organisms are compared several sequence homologies can be
identified. Figure 5 presents one such comparison. The fact that such
homologies have been maintained over millions of years of evolution
suggests that they serve an important function. One of these 5' consen-
sus sequences is identified by the nucleotides GTATAAATAG and is known
as the TATA or Goldberg-Hogness box (107). This sequence has been
identified in almost every histone gene analyzed, located 15-35 bp
upstream of where mRNA transcription begins. The similarity in nucleo-
tide sequence of the histone TATA box to the promoter element found
in procaryotes by Pribnow et al. (108) suggested that the two shared
a common function (109). The functional studies which have been performed
on this sequence in vivg indicate that in eucaryotes TATA helps to fix
the point where the mRNA transcript is initiated by RNA polymerase II
(110). When TATA is removed or mutated, transcription is greatly
reduced but can still occur, initiating at a number of sites. Therefore,
the eucaryotic TATA is not the exact equivalent of the related procary-
otic sequence. TATA is currently believed to interact with RNA polymer-
ase II resulting in transcription initiation at a particular site.

Upstream of the TATA homology is a region in histone genes

characterized by the canonical sequence CCAAT (111). This 'Cat' box

31

 

cI'-d"'o

Box"

 

7") IIIII
n25 [\UAC.--
‘r‘.

1119 50“»-
1-”) Ir-w-
SF‘-‘.§ hu‘-¥-

-A-\
DRDUC

Anmfl-R

mus-nz-l CC.CC-
__a-\ ‘PR II!- _“o ~~~~~ '
DC- \anC N‘A ‘- -R'C\.CC'L
LIQ ~~~~~~~~~ -':.- nnnnn ~
.... Jr.C_—_ CA _-c Jn CC 3
”u- -— . was “a -----
fr" ‘- Unyy"... --"--"-—‘."-o\---
R._‘ ~~~~~~~~~ . ' ~~~~~ n
:21 0 UACC.V..Cn'-N'Jn.CC'-

:ac "..-—.... - .....
DP-U CC». W..C’H- CCUCC
' .mAI-e‘naa
A.-hl-l HAK.» .nuvx.

HER
h22 GGACAATTT-Z9- GATCC-
hl9 GGACAATAG- 26- GATCC-

-45-
ACGCC-

AGTCAATTC

 

"rca- 50 ea c:
AATGA- -45-"T”‘-

Uh CCl
mTGA-44-GATCC-

C531 CACCAATGA ’39-
g;

hl9

Spl i/ GAGCC‘

" lfiownons‘mum HA MAo‘mA-r- 1 Humps-«H
6’...un-n.n-er‘-."‘CiunCC‘42‘Cr-u-u- .-
A-F“I§Ah--nfl AA 5. FFFFFF : "sang-n t-
O Ca.m\C.~v-. .er’--’,C.-.1JCC’JC‘CnlCn.C
f‘ app-mafiP—I-'-- A.
9'uarxnC.-‘U:\T rsC‘: ".C_r_§..Cu DI Cn.Cn.C
~ Q--~Iflm*fl-44- ‘hTF‘
agtnrm.n::' -. AANUQAC J—“uv-aq
I If“ l“!!- f‘ 'r. q. u fl-,- ”Iv-om-
9'flTh.n..mu.’-0"xC_\CCMU‘24’nC:C.n.CI
Ciro..- Q app-Quinlan. so Pea-f-
C"3;n..C-v1.z:x:"2x'.Cr-s..Cn" "I-mCC.-..,;
.‘---.-o -fi- -~\~\~A‘-l- \- of-
C n..-<..r.rv«. —.S- .. .C.'\..‘Cn 5 '.~.-..:-...
’ ~Iv~ - -' Input-“alum. -_ u‘ .uu.
C'.-.-;\.:-.r.n..:x.;'-C'.C.-...Cr'.“-".~..-.C.n.C
" vuhI-Foifin-ufl- «f a” cowl-'4‘. ( uuuuuuu
C'n..-\.rJ-\.~.x.-..:’-C'.Cr..CA'NE’AA ..... ..2
" cw-o—n-ofl-wn- -- 17:... ”~- “" 9‘ - A
t"n.nA-v-\CJ.:\C’-."A.nLTv.’D."CaCr-.gr.v
Cy. ~n—n --\~ IF‘~.‘§F :- nnnnnnn
.n..~.n.v._.C --‘nCr.ur\C1‘C"C."'CnC.r .:
4mgm‘h-mou 5. nmohmm-r ’_ ~~~~~~~
6’C.n.~v1.rC-.‘.C‘CAAC.it’lc UCunr C:

on,-

9-GTATAAATAG-24-CC5TTCA-6l- CnxCATG
9-GTATAAATGG-22-CCATTCA-61- CATCATG
GTATAAATAG-22-CC5TTCA-60-CATCATG
GTATAAATAT-Zl-GCATTCG-4l-CAAAATG
9-TTATAACTGC-l9-CGATTC-137-TGCGATC

9- -GTATAAACHG- 20- -GCACTCA~ 67-CA CCATG
0- CTATAAAAAG- 16- CCATTCA- 70-CATCATC
8- MGTAT "' G- 16- CAnTTCA- -68- -CAT CATS

STAT nAACGT-IS-CCATTCG- 37-AAC"TG

CTR TAAnTnA- 21-CCATTCA-48-CACT TC

AT‘TAGCTGA-IE- -TC§CTTG-29-C:‘CRTG
GTATHTGGAT-l9- -TTATTCG-29-C.AAGATC
S-TTATATTGAC-lT- -TTATTCC- 29- 'CM 'CnT G

 

GATCC

GTATAAATAG ECATTCE CA--ATG

 

Figure 5. Histone gene

The terms "CCAAT Box",

5' homology blocks.
"TATA Box" and "Cap

Box" are explained in the text (from 111).

32

is usually positioned 30—60 nucleotides 5' of TATA and may be repeated

a second time further upstream. The CCAAT sequence, or something very
similar to it, has been implicated as an entry site for the RNA polymer-
ase II (110), thereby, acting as a modulatory element of a gene's
promoter. This sequence is lacking in a few H1, H4 and H2a genes as
well as a number of other RNA polymerase II genes. Whether another
sequence acts as a surrogate in these genes is not known. In the case
of H2a genes that are associated with a divergently transcribed H2b
gene, the two may actually share one or more CCAAT boxes. This
intriguing possibility is currently being investigated in chicken (112).

The site where RNA polymerase II initiates transcription is
chosen based on the location of TATA, as has been mentioned, but there
is also some indication that the preferred site is marked by a sequence
element known as the 'CAP' box (104). In histone genes the sequence
requirement of the CAP box can be met in several forms with the basic
motif being: €CATTC§ (111). In some cases this element is not readily
apparent and SI nuclease protection experiments must be performed in
order to identify the site where the message begins.

In addition to sequence homologies shared by the histones in
general, particular classes of histone genes have been analyzed in hopes
of identifying regulatory elements responsible for tissue-specific or
developmental restricted gene expression. Much remains to be done in
this regard since only a few gene families are currently known in
sufficient detail. The H2b gene family of chicken represents the most

completely characterized of all the higher eucaryotic histone gene

families (95,113,114). An analysis of the sequence data from chicken

33

and other genera has revealed a nine base sequence that is shared by
all H2b genes so far examined. This nonomer is located typically 5-10
bases upstream of TATA. The role of this sequence, if any, is currently

under investigation (112).

Histone Gene Coding Regions

Several interesting facets of histone genes have been revealed by
the characterization of their exons. Knowing the nucleotide sequence
has provided independent confirmation of the amino acid sequences
determined for a number of histones and has predicted the existence of
others not yet isolated. It has become clear that these genes, which
encode the conserved histone proteins, are themselves much less well
conserved. Although the members of a gene family which code for one class
of histone protein resemble each other, when that family of genes is
compared to other families (eg. early versus late) from the same or a
different species, significant sequence divergence is found (68,115).
This point further emphasizes that most current collections of histone
genes are almost certainly incomplete.

Another interesting observation, which has been made from the
analysis of coding sequence data, is that codon usage is usually
nonrandom and often exhibits species specificity. The preference for
a particular codon or set of codons over others is not understood.

Finally the DNA sequencing of histone coding regions has revealed
that most genes are composed of a single exon. However, at least four
intron containing examples have been reported. Two of these occur in

chicken (116) and two in Neurospora crassa (83). The N, crassa genes

 

appear to be the only ones used by this species. The H3.3 genes are

clearly replacement variants.

33

and other genera has revealed a nine base sequence that is shared by
all H2b genes so far examined. This nonomer is located typically 5-10
bases upstream of TATA. The role of this sequence, if any, is currently

under investigation (112).

Histone Gene Coding Regions

Several interesting facets of histone genes have been revealed by
the characterization of their exons. Knowing the nucleotide sequence
has provided independent confirmation of the amino acid sequences
determined for a number of histones and has predicted the existence of
others not yet isolated. It has become clear that these genes, which
encode the conserved histone proteins, are themselves much less well
conserved. Although the members of a gene family which code for one class
of histone protein resemble each other, when that family of genes is
compared to other families (eg. early versus late) from the same or a
different species, significant sequence divergence is found (68,115).
This point further emphasizes that most current collections of histone
genes are almost certainly incomplete.

Another interesting observation, which has been made from the
analysis of coding sequence data, is that codon usage is usually
nonrandom and often exhibits species specificity. The preference for
a particular codon or set of codons over others is not understood.

Finally the DNA sequencing of histone coding regions has revealed
that most genes are composed of a single exon. However, at least four
intron containing examples have been reported. Two of these occur in

chicken (116) and two in Neurospora crassa (83). The N, crassa genes

 

appear to be the only ones used by this species. The H3.3 genes are

clearly replacement variants.

34

Histone Genes-3' Flanking Sequences

The DNA sequences that follow genes transcribed by RNA polymerase II
contain' elements which signal the termination of transcription and the
processing of the primary transcript (117). The exact nature of the
transcription termination signal is not known. The histone genes have
been extensively investigated with respect to the processing of their
mRNA's 3' terminus. Birchmeier, et al. (118) have used site-specific
mutagenesis and sea urchin extract supplemented Xenopus oocytes in
transcription assays to demonstrate that there are two target sites
required for authentic 3' processing. One site, located 20-50 bases
from the termination codon, is a highly conserved palindrome known as
the hyphenated symmetrical dyad. This palindrome can potentially form
a hairpin stem-loop structure in the mRNA. The second site required
for the correct processing of histone mRNA is referred to as the spacer
sequence and extends from the dyad about 100 bases downstream. Deletion
mutagenesis has defined a nine base purine rich sequence in this spacer
RNA which lies 10 bases 3' of the dyad and is the basic spacer sequence
required for any correct 3' processing. The site of the mRNA's processed
3' end falls between the dyad and this purine rich nanomer. When the
remaining portion of the spacer sequence is present, the wild-type levels
of accurately processed mRNA can be obtained. These studies were carried
out on sea urchin H2a genes, but the results can probably be extra-
polated to other histone genes as well. Similar sequences have almost
always been found in the 3' flanking sequences of histone genes (111,
115). This region tends to be remarkably well conserved in comparison
to the 5' flanking regions. Birnstiel's group has been able to further

dissect the mechanism by which the 3' end of the message is processed

35

by identifying a small nuclear RNA-protein complex called a 'SNURP'
which is involved in the reaction (118).

Another processing function often associated with RNA polymerase
II transcripts is the addition of a poly A tract of nucleotides to the
end of the mRNA. The mechanism by which this reaction occurs is still
incompletely understood, but site-specific mutagenesis has clearly
identified the conserved sequence AAUAAA as an important component of
the signal (120).

Histone transcripts are generally not polyadenylated and DNA se—
quence analysis has confirmed that most histone genes lack the canonical
3' poly A addition signal sequence. It has been known for some time

that §, cerevisiae histone mRNAs are polyadenylated and DNA sequencing

 

has revealed that these genes carry the signal sequence (81). Some
higher eucaryotic histone messages also carry a poly A tail. In the
chicken, one H2b gene has been identified whose 3' flanking sequence
contains the poly A addition signal and its transcript appears to be
polyadenylated (95,115). Chicken transcripts from H3.3 histone genes,
genes that contain introns, were found to partition with polyA+ mRNA(116).
The genomic sequences of these variants also appear to contain the
appropriate signal sequence (121). The oocyte-specific amphibian
histone transcripts all possess poly A tracts, but whether or not their
genes carry the poly A signal is not known (122). Further, there have
been reports of polyadenylated histone messages in HeLa cells (123),
sea urchin (124) and clams (125). Why some histone genes code for
polyadenylated mRNAs while others specifying the same protein do

not is a question which remains to be answered. What role the poly A

tract plays in the life cycle of any mRNA molecule is currently being

36
actively investigated (see below)

Regulation of Histone Gene Expression

There are several mechanisms by which a gene's expression may be
regulated. These include transcriptional controls; mRNA processing,
transport and stability; translational mechanisms; and DNA modifications
and rearrangements. Although the major regulatory mechanism is
thought to be transcriptional in nature other levels of control are most
likely superimposed on this one. Two regulatory mechanisms of general
interest are those which exert cell cycle-specific control and those
which alter the state of the chromatin where the gene of interest is
located.

The histones provide an interesting system in which to study
each of these levels of gene control. From a transcriptional stand-
point, since each histone family is composed of a set of genes, how
are these sets controlled to ensure that the appropriate amounts of
each protein (and the appropriate variants thereof), are available for
nucleosomal assembly? Furthermore, what controls individual histone
genes whose expression is developmentally regulated? With respect to
cell cycle regulation, what element(s) distinguishes a gene that codes
lfor a replacement histone variant whose synthesis occurs throughout
the cell cycle from a DNA replication, cell cycle-sensitive gene?
Finally, since the histone genes are scattered fairly extensively
throughout an organism's genome, they provide targets for studies on the
various states of chromatin (eg., methylation, heterochromatinization,
DNase sensitivity and B or 2 forms of DNA) and their effect on gene

expression.

37

The investigation of 'classical' promoters has so far resolved three
important regions in the 5' flanking sequence of RNA polymerase II
genes; CCAAT, TATA and CAP. To these promoter regions should be
added an enhancer element; which, by definition, need not be directly
5' to the gene of interest. An enhancer is a short nucleotide sequence
located 5' or 3' of a transcriptional unit which can augment the trans-
criptional activity of that gene (126).

One of the first chromosomal genes found to be associated with
a putative enhancer element was a sea urchin early histone H2a gene
(127). Using deletion mutants in in 3239 and in 31532 transcription
assays Grosschedl and Birnstiel defined a region located between -185
to -111 (5' coordinates) that is important for enhanced transcription
from that gene (125). However, the criteria of direction independence
and location independence have not been demonstrated. The association
of this putative enhancing element with this gene may or may not be a
characteristic shared by other histone genes.

Of interest in this regard is the recent report by Osley and
Hereford (106) of a putative enhancer sequence associated with a yeast
H2b gene (see below).' A great deal more work is required before the
significance of these observations can be properly evaluated and put
into perspective (see below).

Gene—specific regulatory elements have been proposed to play a
role in the selective activation of genes. With respect to the histones
these studies are only to the point of identifying stretches of sequence
homology shared by genes which encode similar products. The most ad-
vanced studies in this regard have focused on the chicken H2b, H3, H1

and H5 genes (129). The DNA sequences of nine H2b genes have been

38

analyzed and all have been found to possess the nine base sequence

6 %TGCATA 5-10 bp 5' of TATA (115). The functional significance of
this sequence is being addressed by site-specific mutagenesis, hybrid
gene constructions and in_zizg transcription assays.

Once the primary transcript is synthesized it must undergo a
number of processing events in the nucleus, described previously,
before the message can be transported to the cytOplasm. Each of these
steps is a potential point for controlling the amount of message that
is ultimately available for translation. It is now clear that all
histone genes are transcribed individually and not as polycistronic
messages (60).

The nascent transcripts are processed rapidly at their 5' termini
where they are 'capped' with m7G (130). This structure may assist in
the transcript's interaction with the ribosome (131), as well as
influence its stability (132). Once the primary transcript is formed,
its 3' end is processed (132). The efficiency of this step could influence
the rate at which histone transcripts appear in the cytoplasm. The
addition of a poly A tract to some histone messages is another point at
which some degree of regulatory control could be exerted. The poly A
sequence may contribute to the cytoplasmic stability of the message (132).

The fact that some histone genes are interrupted by introns
suggests another potential means of regulating a gene's expression. The
splicing reaction does not occur as quickly as the capping or polyadenyl-
ation reactions (133). Although progress has been made in clarifying

part of the mechanism by which splicing is achieved, how it is controlled

remains unknown (134). There is no clear evidence for alternative

39

splicing of histone transcripts.

The influence of mRNA transport from the nucleus to the cytoplasm
on the control of gene expression remains an interesting possibility,
but is only speculative with respect to histones.

The cytoplasmic stability of histone mRNAs has been investigated
and isthought to be the second most important level of control next to
transcription (132). When Sittman et al. (135) investigated the effect
of inhibitors of DNA synthesis on histone mRNA transcripts and their
stability, it was determined that both transcription and stability were
affected. Transcription was depressed by a factor of five and histone
mRNA stability decreased by the same amount. These effects were only
seen when the replication-dependent variants were analyzed; the replace-
ment histones were unaffected. These investigators concluded that a
signal molecule was involved in both the suppression of transcription
and the destabilization of the message. They proposed that this molecule
could be an unstable protein and ruled out the likelihood that the
histones themselves were involved in their own regulation. However,
Hereford's data in yeast suggests that they are (106).

When DNA synthesis is blocked by inhibitorssuch as hydroxyurea
a concomitant drop in histone levels is observed (60). These types of
experiments were interpreted to mean that all histone synthesis is
tightly coupled to DNA replication during the S phase of the cell cycle.
Now, however, the bulk of the evidence indicates that only some histone
synthesis is tightly coupled to DNA replication (34). Furthermore,
the transcription of even those genes that are periodically activated

may begin before replication of the DNA has commenced (136). Therefore,

40

the concept of histone gene expression being limited to S phase must be
replaced with a more complex scheme.

Experiments designed to address the role of the cell cycle in
regulating histone gene expression have been most successful in S.

cerevisiae. There are several reasons why yeast has been the organ-

 

ism of choice in which to conduct these studies. First, yeast cultures
can be easily synchronized to near homogeneity. A second important
development has been the identification of a number of temperature
sensitive cell cycle mutants which can be blocked at several points in
61 and S phases. Third, the complete yeast histone gene complement is
cloned providing specific probes for hybridization experiments. Fourth,
specific hybrid gene constructions can be introduced into yeast in a
stably integrated form or in an unstable episomal form permitting the
functional analysis of a region of DNA sequence.

The results from a number of studies indicate that there are
sequences associated with some yeast histone genes, in particular 3: of
an H2b gene, which regulate its transcriptional activity in a periodic
fashion (106). When a plasmid lacking a functional origin of replica-
tion is fused to this H2b-associated sequence the plasmid becomes capable
of autonomously replicating as an episome. In this type of assay the
H2b-associated sequence behaves as a number of other yeast sequences do
(137), namely to act as origins of replication or as autonomously repli-
cating sequences (ARS). Whether they are in fact origins of DNA
replication remains to be proven.

In order to evaluate this H2b ARS in a chromosomal environment,

a fusion gene was constructed which could be mutagenized in vitro, used

41

to transform competent yeast cells and be stably integrated into the
chromosome. These constructs were then analyzed for their transcrip-
tional activity throughout the cell cycle (106). The results from this
type of analysis indicate that the ARS may be acting as an enhancer
element but that it constitutes only part of the 'sensor' that

responds in a periodic way to some form of cell cycle control.

The association of an element of sequence that confers transcrip-
tional periodicity upon histone genes, whatever its exact nature, is
exciting. These results, for the first time suggest a way in which the
expression of a histone gene could be made sensitive to cell cycle signals
and in addition imply that local alterations in DNA structure may play
an important role in activating at least some histone genes.

That changes in the chromatin structure of a given histone gene
accompany its activation or inactivation was recently demonstrated by
Bryan et al. (138) in sea urchin. Using the nuclease assay described
earlier, they were able to show that when a particular histone gene is
activated, the 5' and 3' regions surrounding it are hypersensitive to
nuclease attack. When the gene is no longer transcriptionally active,

this nuclease sensitivity is lost. y

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

42

LITERATURE CITED

Kossel, A. 1884. Cited by H. Bush. 1965. Histones and Other Proteins.
Academic Press, New York.

Igo-Kemins, T., W. Horz and H.G. Zachau. 1982. Ann. Rev. Biochem.
51:89-121.

Zweidler, A. 1978. Met. Cell Biol. 17:223-233.
Hardison, R. and R. Chalkley. 1978. Met. Cell Biol. 17:235-251.
Isenberg, I. 1979. 48:159-191.

DeLange, R.J., D.M. Fambrough, E.L. Smith and J. Bonner. 1969.
J. Biol. Chem. 244:5669-5679.

Patthy,L., E.L. Smith and J. Johnson. 1973. J. Biol. Chem. 248:6834-
6840.

Wallis, J.W., L. Hereford and M. Grunstein. 1980. Cell. 22:799-805.

Strickland, M., W.N. Strickland, W.F. Brandt and C. von Holt. 1977.
Eur. J. Biochem. 77:263-275.

Strickland, W.N., M. Strickland, W.F. Brandt and C. von Holt. 1977.
Eur. J. Biol. Chem. 77:277-286.

van Helden, P.D., W.N. Strickland, M. Strickland and C. von Holt.
1982. Biochim. et Biophys. Acta. 703:17-20.

Iwai, K., H. Hayashi and K. Ishikawa. 1972. J. Biochem. 72:357-367.
Elgin, S.C.R. and H. Weintraub. 1975. Ann. Rev. Biochem. 44:725-774.
Kornberg, R.D. 1977. Ann. Rev. Biochem. 46:931-954.

McGhee, J.D. and G. Felsenfeld. 1980. Ann. Rev. Biochem. 49:1115-1156.

Richmond, T.J., J.T. Finch, B. Rushton, D. Rhodes and A. Klug. 1984.
Nature. 311:532-537.

Laskey, R.A. and W.C. Earnshaw. 1980. Nature. 286:763-767.

Mirzabekov, A.D., A.V. Belyavsky, S.G. Bavykin and V.V. Shick. 1980.
BioSystems. 12:265-271.

Finch, J.T., R.S. Brown, D. Rhodes, T. Richmond, B. Rushton, L. Lut-
ter and A. Klug. 1981. J. Mol. Biol. 145:757-769.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

39.

39.

40.

41.

43
Whitlock, J.P. and A. Stein. 1978. J. Biol. Chem. 253:3857-3861.
Wallis, J.W., M. Rykowski and M. Grunstein. 1983. Cell. 35:711-719.
Hall, M.N., L. Hereford and I. Herskowitz. 1984. Cell. 36:1057-1065.

Boulikas, T., J.M. Wiseman and W.T. Garrad. 1980. Proc. Natl. Acad.
Sci. U.S.A. 77:127-131.

Thomas, J.0. and A.J. Khabaza. 1980. Eur. J. Biochem. 112:501-511.
Thoma, F., T. Koller and A. Klug. 1979. J. Cell Biol. 83:403-427.
Bates, D.L. and J.O. Thomas. 1981. Nucl. Acids Res. 9:5883-5894.

Skoultchi, A. and P.R. Gross. 1973. Proc. Natl. Acad. Sci. U.S.A.
70:2840-2844.

Brandhorst, B.P. 1980. Dev. Biol. 79:139-148.

Schaffer, W., G. Kunz, H. Daetwyler, J. Telford, H.O. Smith and M.
L. Birnstiel. 19780 C811. 14:655-671.

Sures, I., J. Lowry and L.H. Kedes. 1978. Cell. 15:1033—1044.

Cohen, L.H., K.M. Newrock and A. Zweidler. 1975. Science. 190:994-
997.

Newrock, K.M., L.H. Cohen, M.B. Hendricks, R.J. Donnelly and E.S.
Weinberg. 1978. Cell. 14:327-336.

Childs, G., R. Maxson and L. Kedes. 1979. Dev. Biol. 73:153-173.

Zweidler, A. 1984. pp.339-371. In_Histone Genes. G.S. Stein, J.L.
Stein and W.F. Marzluff, eds. J. Wiley and Sons, New York.

Urban,M.K. and A.Zweidler.l983. Dev. Biol. 95:421-428.

Louie, A.J., B.P.M. Candido and G.H. Dixon. 1974. Cold Spring Har-
bor Symp. Quant. Biol. 38:803-807.

Trostle-Weige, P.R., M.L. Meistrich, W.A. Brock, K. Nishioka and
J.W. Bremer. 1982. J. Biol. Chem. 257:5560-5570.

Smith, B.J., J.M.Walker and E.W. Johns. 1980. FEBS Letters. 112:42-
44. '

Grove, G.W. and A. Zweidler. 1981. J. Cell Biol. 91:673.
Seidman, M.N., A.J. Levine and H. Weintraub. 1979. Cell. 18:439-449.

Weintraub, H., S.J. Flint, I.M. Leffak, M. Groudine and R.M. Graing-
er. 1978. Cold Spring Harbor Symp. Quant. Biol. 42:401.

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

44

Jackson, V., S. Marshall and R. Chalkey. 1981. Nucl. Acids Res.
9:4563-4581.

Gurley, L.H., R.A. Walters, S.S. Barham and L.L. Deaven. 1978. Exp.
Cell Res. 111:373-383.

Honda, B.M., B.P.M. Candido and G.H. Dixon. 1975. J. Biol. Chem.
250:8686-8689.

Borun, T., D. Pearson and W.K. Paik. 1973. J. Biol. Chem. 247:4288-
4298.

Byvoet, P. and C.S. Baxter. 1975. pp. 125-151. lg Chromosomal Pro-
teins and Their Role in the Regulation of Gene Expression. G.S.
Stein and L.J. Kleinsmith, eds. Acaddemic Press, New York.

Hunt, L.T. and M.D. Dayhoff. 1977. Biochem. Biophys. Res. Comm.
74:650-655.

Goldknopf, I.L. and H. Busch. 1977. Proc. Natl. Acad. Sci. U.S.A.
74:864-868.

West, M.H.P. and W.M. Bonner. 1980. Nucl. Acids Res. 8:4671-4680.

Goldstein, G., H. Scheid, U. Hammerling, E.A. Boyse, D.H. Schlesinger
and H.D. Niall. 1975. Proc. Natl. Acad. Sci. 72:11-15.

Wu, R.S., K.W. Kohn and W.M. Bonner. 1981. J. Biol. Chem. 256:5916-
5920.

Goldknopf, I. L., C. Wilson, R. N. Ballal and H. Busch. 1980.
J. Biol. Chem. 255:10555-10558.

Hershko, A. and A. Ciechanover. 1982. Ann. Rev. Biochem. 51:335-364.

Borun, T.W., M.D. Scharff and E. Robbins. 1967. Proc. Natl. Acad.
Sci. U.S.A. 58:1977-1983.

Gallwitz, D. and C.C. Mueller. 1969. J. Biol. Chem. 244:5947-5952.
Kedes, L. and P. Gross. 1969. J. Mol. Biol. 42:559-575.

Nemer, M. and D.T. Lindsay. 1969. Biochem. Biophys. Res. Comm. 35:
156-160.

Moav, B. and M. Nemer. 1971. Biochemistry. 10:881-888.
Kedes, L.H. and M.L. Birnstiel. 1971. Nature New Biol. 230:165-169.
Kedes, L.H. 1979. Ann. Rev. Biochem. 48:837-870.

Grunstein, M., S. Levy, P. Schedl and L.H.Kedes. 1973. Cold Spring
Harbor Symp. Quant. Biol. 38:717-724.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71.

72.

73.

74.

75.

76.

77.

78.

79.

80.

45

Birnstiel, M., J. Telford, E.S. Weinberg and D. Stafford. 1974.
Proc. Natl. Acad. Sci. U.S.A. 71:2900-2904.

Gross, K.W., W. Schaffner, J. Telford and M. Birnstiel. 1976. Cell.
8:471-478.

Holmes, D.S., R.H. Cohn, L.H.Kedes and N.Davidson. 1977. Biochemistry.
16:1504-1512.

Kunkel, N.S. and E.S. Weinberg. 1978. Cell. 14:313-326.

Spinelli, G. F. Gianguzza, C. Casano, P. Acierno and J. Burckhardt.
1979. Nucl. Acids Res. 6:454-460.

Grunstein, M. 1978. Proc. Natl. Acad. Sci. U.S.A. 75:4135-4139.

Maxson, R., T. Mohun, G. Gormezano, G. Childs and L. Kedes. 1983.
Nature. 301:120-125.

Childs, G., R. Maxson, R. Cohn and L. Kedes. 1981. Cell. 23:651-663.
Karp, R. and D.S. Hogness. 1976. Fed. Proc. 35:1623-

Lifton, R.P., M.L. Goldberg, R.W. Karp and D.S. Hogness. 1977. Cold
Spring Harbor Symp. Quant. Biol. 42:1047-1051.

Pardue, M.L., L.H. Kedes, E.S. Weinberg and M.L. Birnstiel. 1977.
Chromosoma. 63:135-151.

Anderson, K.V. and J.A. Lengyel. 1984. pp. 135-161. lg_Histone Genes.
G.S. Stein, J.L.Stein and W.F. Marzluff, eds. J. Wiley and Sons,
New York.

Khesin, R.B. and B.A. Leibovitch. 1978. Molec. Gen. Genet. 162:323-
329.

Moore, G.D., J.P. Procunier, D.P. Cross and T.A. Grigliatti. 1979.

Wright, T.R.F., R.B. Hodgetts and A.F. Sherald. 1976. Genetics.
84:267-285.

Hereford, L. K. Fahrner, J. Woolford, M. Rosbash and D.B. Kabak.
1979. Cell. 18:1261-1271.

Fahrner, K., J. Yarger and L. Hereford. 1980. Nucl. Acids Res. 8:5725-
5737.

Smith, M.M. and K. Murray. 1983. J. Mol. Biol. 169:641-661.

Rykowski, M.C., J.W. Wallis, J. Choe and M. Grunstein. 1981. Cell.
25:477-487.

81.

82.

83.

84.

85.

86.

87.

88.

89.

90.

91.

92.

93.

94.

95.

96.

97.

98.

46

Smith, M.M. 1984. pp. 3-33. In Histone Genes. G.S. Stein, J.L. Stein
and W.F. Marzluff, eds. J. Wiley and Sons, New York.

Olson, M.V., K Laughney and B.D. Hall. 1979. J. Mol. Biol. 132:387-
410.

Woudt, L.P., A.Pastink, A.F. Kempers-Veenstra, A.F.M. Jansen, W.H.
Mager and R.J. Planta. 1983. Nucl. Acids Res. 11:5347-5360.

Bannon, G.A. and M. Gorovsky. 1984. pp.163-179. In_Histone Genes.
G.S. Stein, J.L. Stein and W.F. Marzluff, eds. J. Wiley and Sons,
New York.

Bagshaw, J.C., M.T. Andrews and B.A. Perry. 1984. pp.181-196. In
Hisotne Genes. G.S. Stein, J.L.Stein and W.F. Marzluff, eds. J.
Wiley and Sons, New York.

 

Moorman, A.M., R.T.M. de Laaf, D.H.J. Destrée, J. Telford and M.L.
Birnstiel. 1980. Gene. 10:185-193.

Stephenson, E.C. 1984. pp.225-242. In Histone Genes. G.S. Stein, J.L.
Stein and W.F. Marzluff, eds. J. Wiley and Sons, New York.

Crawford, R.J., P. Krieg, R.P.Harvey, D.A. Hewish and J.R.E. Wells.
1979. Nature. 279:132-136.

Jacob, E. 1976. Eur. J. Biochem. 65:275-284.

Wilson, M.C., M. Melli and M. Birnstiel. 1974. Biochem. Biophys. Res.
Comm. 61:404-414.

Dodgson, J.B., J. Strommer and J.D. Engel. 1979. Cell. 17:879-887.

Engel, J.D. and J.B. Dodgson. 1981. Proc. Natl. Acad. Sci. U.S.A.
78:2856-2860.

Bruschi, S. and J.R.E. Wells. 1981. Nucl. Acids Res. 9:1591-

Sugarman, B.J., J.B. Dodgson and J.D. Engel. 1983. J. Biol. Chem.
258:9005-9016.

Grandy, D.K., J.D. Engel and J.B. Dodgson. 1982. J. Biol. Chem.
257:8577-8580.

Ruiz-Carrillo, A., M. Affolter and J. Renaud. 1983. J. Mol. Biol.
170:843-859.

Harvey, R.P. and J.R.E. Wells. 1984. pp.245-262. IE_Histone Genes.
G.S. Stein, J.L. Stein and W.F. Marzluff, eds. J. Wiley and Sons,
New York.

Krieg, P.A., A.J. Robins, A. Colman and J.R.E. Wells. 1982. Nucl.
Acids Res. 10:6777-6785.

99.

100.

101.

102.

103.

104.

105.

106.

107.

108.

109.

110.

111.

112.

113.

114.

115.

116.

117.

118.

47

Sittman, D.B., I.M. Chiu, C.J. Pan, R.H. Cohn, L.H. Kedes and W.F.
Marzluff. 1981. Proc. Natl. Acad. Sci. U.S.A. 78:4078-4082.

Stein, G.S., F. Sierra, M. Plumb, F. Marashi, L. Baumbach, J.L. Stein,
N. Carolli and K. Prokopp. 1984. pp.397—455. In Histone Genes. G.S.
Stein, J.L. Stein and W.F. Marzluff, eds. J. Wiley and Sons, New
York.

Carozzi, N., F. Marashi, M. Plumb, S. Zimmerman, A. Zimmerman, L.S.
Coles, J.R.E. Wells, G. Stein and J. Stein. 1984. Science. 224:1115-
1117.

Thomas, G. and J.D. Padayatty. 1983. Nature. 306:82-84.

Baltimore, D. 1981. Cell. 24:592-594.

Breathnach, R. and P. Chambon. 1981. Ann. Rev. Biochem. 50:349-383.

Wright, S., A. Rosenthal, R. Flavell and F. Grosveld. 1984. Cell.
38:265-273.

Osley, M.A. and L. Hereford. 1982. Proc. Natl. Acad. Sci. U.S.A.
79:7689-7693.

Goldberg, M. 1979. Thesis. Stanford University.
Pribnow, D. 1975. Proc. Natl. Acad. Sci. U.S.A. 72:784-788.

Hu, S.L. and J.L. Manley. 1981. Proc. Natl. Acad. Sci. U.S.A. 78:
820-824.

McKnight, S.L. and R. Kingsbury. 1982. Science. 217:316-324.

Hentschel, C.C. and M.L. Birnstiel. 1981. Cell. 25:301-313.
Grandy, D.K. and J.B. Dodgson. unpublished observations.

Grandy, D.K., J.D. Engel and J.B. Dodgson. 1983. pp.445-455. In
Gene Expression. D. Hamer and M. Rosenberg, eds. A.R. Liss, New

York.

Harvey, R.P., A.J. Robins and J.R.E. Wells. 1982. Nucl. Acids Res.
10:7851-7863.

Grandy, D.K. unpublished observations.

Engel, J.D., B.J. Sugarman and J.B. Dodgson. 1982. Nature. 297:434-
436.

Proudfoot, N. 1984. Nature. 307:412-413.

Birchmeier, C.)W. Folk and M.L. Birnstiel. 1983. Cell. 35:433-440.

48
119. DeRoberts, E.M. 1983. Cell. 32:1021-1025.
120. Proudfoot, N. and G.G. Brownlee. 1976. Nature. 263:211-214.
121. Brush, D.B. and J.B. Dodgson. unpublished observations.
122. van Dongen, W.M.A.M., A.F.M. Moorman and G.H.J. Destree. 1984.
pp.199-224. In_Histone Genes. G.S. Stein, J.L. Stein and W.F. Marz-
luff, eds. J. Wiley and Sons, New York.

123. Borun, T.W., R. Ajiro, A. Zweidler, T.W. Dolby and R.B. Stephens.
1977. J. Biol. Chem. 252:173-180.

124. Ruderman, J.V. and M.L. Pardue. 1978. J. Biol. Chem. 253:2018-2025.
125. Gabrielli, F. and C. Gablioni. 1975. Dev. Biol. 43:254-263.
126. Khoury, G. and P. Gross. 1983. Cell. 33:313-314.

127. Grosschedl, R. and M.L. Birnstiel. 1980. Proc. Natl. Acad. Sci. U.S.A.
77:1432-1436.

128. Grosschedl, R. and M.L. Birnstiel. 1982. Proc. Natl. Acad. Sci. U.S.A.
79:297-301.

129. Grandy, D.K., D. Browne, P. Boyer, D. Brush, and J.B. Dodgson. un-
published observations.

130. Shatkin, A.J. 1976. Cell. 9:645—653.

131. Banerjee, A.K. 1980. Microbiol. Rev. 44:175-205.
132. Nevins, J.R. 1983. Ann. Rev. Biochem. 52:441-466.
133. Nevins, J.R. 1979. J. Mol. Biol. 130:493-506.

134. Ruskin, B., A.R. Krainer, T. Maniatis and M.R. Green. 1984. Cell.
38:317-331.

135. Sittman, D.B., R.A. Graves and W.F. Marzluff. 1983. Proc. Natl.
Acad. Sci. U.S.A. 80:1849-1853.

136. Coffino, P., E. Stimac, V.E. Groppi and D. Bieber. 1984. pp.317-338.
In Histone Genes. G.S. Stein, J.L.Stein and W.F. Marzluff, eds.

J. Wiley and Sons, New York.

137. Struhl, K., D.T. Stinchcomb, S. Scherer and R.D. Davis. 1979. Proc.
Natl. Acad. Sci. U.S.A. 76:1035-1039.

138. Bryan, P.N., J. Olah and M.L. Birnstiel. 1983. Cell. 33:843-848.

CHAPTER 1
COMPLETE NUCLEOTIDE SEQUENCE OF A CHICKEN H2b HISTONE GENE*
David K. Grandy, James Douglas Engel and Jerry B. Dodgson

Abstract

The complete nucleotide sequence of a chicken H2b histone gene has
been determined along with extensive flanking sequence both 5' and 3'
from the gene. This H2b gene was isolated on a 1 Charon 4A-chicken DNA
recombinant in which it is closely linked to two H3 histone genes. The
H2b histone gene predicts a chicken H2b histone protein sequence differ-
ing in 5 of 125 amino acids from the sequence of a calf thymus H2b
histone. The gene is uninterrupted and is flanked by several consensus
sequences seen in many other eucaryotic genes. In particular, unlike
many other histone genes it contains a 3'-AATAAA sequence usually assoc—

iated with genes coding for polyadenylated mRNA.

*
Reprinted from The Journal of Biological Chemistry 257: 8577-8580,
1982.

49

50

INTRODUCTION

The chromosomal DNA of all eucaryotes is found in the cell tightly
complexed to a small set of basic proteins called histones, whose struc-
ture has been highly conserved over evolutionary time (1). The genes
which code for these proteins are generally found in multiple copies
per haploid genome, and they form a large gene family which is quite
diverse in its arrangement in different species (2,3). The most inten-
sively studied histone genes are those of various sea urchin species

and, to a lesser extent, those of Drosophila melanogaster (4). Most

 

of the histone genes in these organisms are arranged in tandemly repeated
blocks of DNA containing one of each of the five gene types per repeat
(2-4). The histone genes of Xenopus (5,6) and newt (7) maintain a
clustered arrangement but the organization of these clusters appears to
be considerably more diverse than that of invertebrates. Our own

studies (8) and those of others (9-11) on chicken histone genes show
that, while these genes are often (but by no means always) clustered,
essentially no evidence for conservation of linkage arrangement can be
found when one cluster is compared to another. This has also been shown
to be the case for mammalian histone genes (12,13).

The diversity of organization seen in the chicken histone gene
family and the possible importance of histone-DNA interactions in regu—
lating cell development encouraged us to begin detailed structural
studies on several representative chicken histone genes. These
experiments serve as a necessary prelude to detailed studies of the

regulation of expression of the histone gene family. We isolated a

51

large set of chicken histone gene-containing recombinant DNAs from a
chicken DNA clone library prepared in the vector A Charon 4A (14).
These clones were identified by cross-hybridization (at low stringency)
to the previously isolated sea urchin histone genes (8). One of three
such clones whose characterization was described previously (8) is
designated ACHla and contains regions of homology to both invertebrate
H3 and H2b histone genes. Recently, a similar, possibly allelic, clone
has been isolated (from the same library) by Harvey et a1. (10). These
workers identified this clone by hybridization to complementary DNA made
to total histone mRNA. Two divergently transcribed H3 histone genes and
a linked H2b gene were located on the 1 clone (their ACHOZ).

This paper reports the complete nucleotide sequence analysis of the
H2b gene contained within ACHla. The gene codes for a chicken H2b histone
which is closely related to the known amino acid sequence of a calf thymus
H2b histone. The gene is continuous and contains in its flanking regions
several consensus sequences related toIhose seen in other eucaryotic genes,
especially other members of histone gene families. Such comparisons may
help delineate regulatory elements that function in the control of histone

gene expression.

52

EXPERIMENTAL PROCEDURES

Material-Restriction enzymes and nucleic acid enzymes were obtained
from New England Biolabs Bethesda Research Laboratories, and Biotec Inc.
(Madison, WI), and used according to the specifications of the supplier.
Polynucleotide kinase was from Bethesda Research Laboratores and AMV1
reverse transcriptase was supplied by J.W. Beard of Life Sciences, Inc.
through the Office of Program Resources and Logistics, Viral Cancer
Program, National Institutes of Health. Other material and strains were
as described previously (16,22). All manipulations of recombinant DNA
molecules in phage or bacteria were done in accordance with current
National Institutes of Health guidelines.

Preparation and Characterization of pKRla-1.3-The 1.3-kb DNA

 

fragment generated by digestion of lCHla with Kpn I and Eco RI enzymes
was isolated by gel electrophoresis and electroelution as described (15).
The chicken a-globin gene-containing plasmid pBRa7-1.7 (16) was also
digested with Kpn I and Eco RI, and the 4.5-kb fragment containing 4.0-kb
of pBR322 and 0.5-kb (noncoding) of the chicken aA globin gene was
isolated. These two fragments were ligated at equimolar ratio and

transformed into Escherichia coli HBlOl. The appropriate plasmid con-

 

taining the 1.3-kb fragment was identified by restriction digestion of

small scale plasmid DNA isolates of selected AmpR colonies. APlasmid

1The abbreviations used are: AMV, avian myeloblastosis virus; kb,

kilobase pairs; bp, base pairs; AmpR, ampicillin-resistant.

53

DNA was prepared and isolated as described (17). A detailed restriction
map of the plasmid was prepared by single and multiple digests of the
plasmid DNA with a variety of restriction enzymes followed by gel electro-
phoresis on 1% agarose or 5% acrylamide gels (18). In some cases, the map
was further characterized by partial restriction digests of singly labeled
DNA fragments prepared for sequencing reactions (19).

DNA Sequencing-DNA sequencing was performed by the chemical degrada-

 

tion technique of Maxam and Gilbert (20), in some cases as modified by
Smith and Calvo (21). DNA fragments were end-labeled by polynucleotide
kinase as described (20) or by AMV reverse transcriptase-catalyzed DNA
synthesis following limited exonuclease III digestion of the 3' termini
of blunt-ended restriction fragments as described (21). In rare cases,
DNA to be labeled by polynucleotide kinase were similarly digested with
exonuclease III to improve labeling efficiency. Sequencing gels were run

as described previously (20,22).

RESULTS AND DISCUSSION

Preparation of a Chicken H2b Gene Plasmid-We previously identified

 

a region of the chicken histone gene-containing recombinant ACHla which

specifically hybridized to sea urchin and Drosophila H2b sequences (8).

 

This region spans 1.3 kb and is flanked by an Eco RI restriction site and
a Kpn I restriction site (Fig. 1A). The flanking region of DNA to the
left of the Eco RI site in Fig. 1A has been shown to contain H3 histone
gene sequences (8)2. Harvey et al. (10) have shown that in a clone

closely related to ACHla the analogous flanking region contains two

2J.D. Engel, unpublished observation.

54

 

 

 
   
 

 

/ 1:? gr \
B . pK Rla- 1 . sfiml u t. t, ‘30 A i? 6 ‘:q 5 b ......”
3 A

W 7‘03.
115"“ LINKER f lent-II YfllnII 6 agar
VHindIII .v Ale! 1 SCIIA a] Hindu
¢EeeRI Y loan 0 In!!!

a Real 1' Neon!

FIG. 1. Restriction maps of ACHla and pKRla-1.3. A. a por-
tion of the clone ACHla containing the chicken DNA insert is shown.
The thin line represents chicken DNA and the heavy line corresponds
to A Charon 4 DNA. The location of chicken DNA sequences which
hybridized to invertebrate H3 (m) and H2b (I) histone gene probes
is indicated. B. a partial restriction enzyme map of the recombinant
plasmid pKRla-l.3 showing the location of cleavage sites employed
in the DNA sequence analysis is diagrammed. The 1.3-kb Eco RU
Kpn 1 chicken DNA fragment subcloned from ACHla is indicated by
the thin line. pBR322 sequences are represented by is) and the
noncoding aoglobin fragment is identified by (in). The arrows indicate
the direction and extent to which the DNA sequence was read from
one experiment The arrows above the map refer to the DNA strand
corresponding to the mRNA sequence and the arrows below the map
refer to the complementary strand. The H2b coding region is delim-
ited on this map by the initiation codon (ATG) and the termination
oodon (TAG).

55

divergently transcribed histone H3 genes, and they have located a single
H2b gene in the region corresponding to the H2b hybridizing region
indicated in Fig. 1A. The 1.3-kb Eco RI/Kpn I H2b DNA fragment was
inserted into a chicken a-globin gene plasmid pBRa7-l.7 which contains

a 1.7-kb DNA fragment cloned into the BAM HI and Eco RI sites of pBR322.
The vector was cleaved at its single Kpn I site in the globin gene insert
and at the regenerated Eco RI site and the 1.3-kb H2b DNA fragment was
inserted into those sites (see "Experimental Procedures"). The resultant
recombinant pKRla-1.3 contains 4.0 kb of pBR322, 0.5 kb (noncoding) of the
a-globin gene region and 1.3 kb of ACHla DNA containing the H2b gene
(Fig. 1B).

The restriction enzyme cleavage map of pKRla-l.3 was determined by
single and multiple enzyme digestions and, in some cases, confirmed by
the technique of Smith and Birnstiel (l9). Harvey et al. (10) have also
subcloned this region of DNA (or an essentially identical region) and,
where the maps overlap, our results are in agreement with theirs. A
partial restriction map which emphasizes the sites used.in the DNA
sequence analysis is shown in Fig. 1B.

DNA Sequence Analysis of pKRla-1.3-The complete nucleotide sequence

 

of the 1.3-kb H2b-containing DNA fragment has been determined by chemical
degradation methods (20,21). That portion which contains the H2b gene

and its flanking regions is shown in Fig. 2. Although the sequence of

both strands of DNA in most of this region has been determined (Fig. 13),
only the strand corresponding to the mRNA sequence is shown. The predicted
amino acid sequence of the corresponding H2b histone is given above the
nucleotide sequence. This sequence can be clearly identified since it

is identical in all but five residues (of 125) to the amino acid sequence

56

of a calf thymus H2b histone (23). Furthermore, the predicted amino
acid sequence is identical to a partial sequence (covering amino acids
1 to 29 and 63 to 89) of a chicken H2b histone purified from adult
erythrocytes (34). Harvey et al. (10) have also identified this gene
from the nucleotide sequence of the first 10 codons, and our sequence
is identical to theirs in this region. The complete nucleotide sequence
confirms the striking level of identity between the chicken H2b histone
and that of the calf. Interestingly, all five differences between the
chicken sequence and the calf are clustered between residues 21 to 40,
and each amino acid change can be accounted for by a single transition
mutation event.

As is typically the case for histone genes(2,3) the chicken H2b
coding sequence is quite GC-rich and exhibits nonrandom codon usage.
Only four of the 123 degenerate third bases in the coding sequence are
A or T rather than G or C. A similar situation was observed in the
chicken H2a gene by D'Andrea et a1. (11). That this GC-rich property
is selectively maintained in the histone H2b gene is further suggested
by the fact that the 375-bp coding sequence contains 30 CpG dinucleo-
tides which are silent with respect to the H2b histone amino acid
sequence. CpG dinucleotides are normally rare in eucaryotic DNA possibly
because these sites are hot spots for mutations resulting from methyl-
ation at the 5 position of the C residue (24). Deamination of the 5
methyl cytosine would lead to a thymine base which could not be recog-
nized by the uracil-specific repair system, thus leading to a comparatively
high rate of transition mutations of CpG sequences to TpG. Therefore,
the high frequency of CpG dinucleotides in the H2b gene which are silent

with respect to the protein sequence suggests that the GC-rich character

57

of the gene must be selectively maintained during evolution. Alterna-
tively, it could be argued that histone genes are rarely methylated

in germ line DNA because they may be active in germ cells (active genes
being generally undermethylated (25)), and, therefore, that they would
not mutate CpG sequences at a higher than normal rate.

Sequence of Gene-flanking Regions-Putative 5' and 3' flanking

 

signal sequences of eucaryotic histone genes have been identified by
comparison of the sequences of different histone DNAs and have recently
been described in detail (2,3,11). While no other complete H2b gene
sequences of higher vertebrates are presently available, the chicken H2b
gene described above has several interesting features in common with
invertebrate H2b genes and histone genes in general. We have not yet
proven that this particular H2b gene is actively transcribed, but the
fact that it can code for an apparently normal H2b protein and in several
other respects resembles an active eucaryotic gene (see below) suggests
that it probably is expressed.

In common with genes transcribed by RNA polymerase II, this chicken
H2b histone gene contains a "TATA" or "Hogness" box (24) (here TATAAATA),
67 bp upstream for the AUG inititator codon (Fig. 2). The H2b gene also
contains the commonly observed CCAAT sequence (26) at postition -104
(with the A of ATG as +1) along with another such CCAAT box at -l62, The
possible significance, if any, of such a double CCAAT sequence is unknown,
although this duplication has also been observed in the chicken a-globin
genes3 among others (27). In the case of the H2b histone gene, the
CCAAT duplicated region extends at least ten base pairs beyond CCAAT in the
direction of the gene as shown in Fig. 2 (7 of 10 base pairs identical).

3Dodgson, J.B. and Engel, J.D., manuscript in preparation.

58

S.
-194 CTTACCACIAGCAGCGAGCCA{IIQIGAIGGACCAA
" III-

‘158 ITACACAGCCCTCCIAICGAGCGACACCTCACGGACTCTCTTATCCAATCAGAGACCACATACACAAGGCACTCEAII:
'_—'—" _—_ ‘
-79 GCAIACIGCCCCIATAAATACGCGACCACTGCTCGCAGCCGGCACTCCGCTCCGCCGAAGGCAICCTGCACAGTTCCAC

Pro Glu Pro Ala Lys Ser Ala Pro Ala Pro Lys Lys Gly Ser Lys Lys Ala Val Ihr
+1 AIC CCI GAG CCG GCC AAC ICC CCA CCC GCC CCC AAG AAG GGC ICC AAC AAC GCC GIC ACC

Lys Ihr Gln Lys Lys Gly Asp Lys Lys Arg Lys Lya Ser Arg Lys Clu Ser Iyr Ser Ile
+61 AAC ACC CAG AAG AAG GCC GAC AAG AAG CGC AAC AAG AGC CGC AAG GAG AGC IAC ICG AIC
Ala Asp Gly Arg Val

Iyr Val Iyr Lys Val Len Lys Clo Val His Pro Asp Ihr Gly Ila Ser Ser Lys Ala Hat
+121 IAC GIG IAC AAC GIG CIG AAG CAG GIG CAC CCC GAC ACG GCC AIC ICC ICC AAG GCC AIC

Cly lle Met Asp Ser Phe Val Asn Asp Ila Phe Clu Arg lle Ala Cly Glu Ala Ser Arg
+181 GGC AIC AIG AAC ICC ITC GIG AAC GAC AIC IIC GAG CGC AIC GCC GCC GAG GCG ICC CGC

Lao Ala His Iyr Asa Lys Arg Ser Ihr Ila Ihr Ser Arg Glu lle Clo Ihr Ala Val Arg
+241 CIG GCC CAC IAC AAC AAG CGC ICG Acc AIC ACG ICG CGG GAG AIC CAG ACA GCC GIG CGG

Leu Lau Leu Pro Gly Glu Lau Ala Lya His Ala Val Ser Glu Cly Ihr Lys Ala Val Ihr
+301 CIG CIG CIG CCC GCC GAG CIG GCC AAC CAC GCC GIC ICC GAG GGI ACC AAG GCC GIC ACC

Lys Iyr Thr Ser Sar Lys .
+361 AAG IAC ACC AGC ICC AAG IAG AGCGGIGCGGAIIACTCAJIIIAACCCAAAGCCICIIIICGGAGCCACCAI

«.33 maurwmminxqmmggcmnmmmmmm

+512 AGCGAGIGCCIGAGGTATCTAIAIAATIGCITAACITCGCAGTTCGCAGGTCTCCGTTCCGAGITAATTCACCAGIAGC

I
+591 AACTCCCGGAGTIAACTGGGTTGGTCGCTAGCCGTCCTATTACTCCAGCACGGIACCA +668 3

FIG. 2. The complete nucleotide
sequence of the chicken 32b gene.
Only the DNA sequence corresponding
to the mRN A strand is given. The amino
acid sequence specified by the coding
region is shown above the nucleotide
sequence. Amino acid residues of a calf
thymus H2b protein (23) which difl'er
from the chicken sequence are given be-
low the DNA sequence. The symbols
located below the DNA sequence flank-
ing the H2b gene identify regions of in-
terest discussed in the text (GA'I'I'T,
”‘; CCAAT box and associated se-
quence. -’ ; TATA box. . possible
cap site. +++; symmetrical dyad. -«—;
putative polyadenylation signal. as;
polypyrimidine sequence, . . .1.

 

59

Busslinger et al. (28) have identified a less well conserved common
sequence about 10 bp upstream of the TATA box by its canonical sequence,
GATCC. This sequence seems to be common to histone gene sequences, but
not necessarily to other eucaryotic genes. In the chicken H2b gene, this
sequence is present as GATTT at -84 (fig. 2). Finally, a putative cap
site for the H2b gene mRNA can be identified by the sequence CACTC 30 bp
downstream form the TATA box (Fig. 2). Unlike the chicken histone H2a
gene (11), only one region in reasonable agreement with consensus cap
sites (3,29) is present in this area of the H2b gene. Further studies
to accurately identify the transcriptional initiation site of this gene
are in progress.

In the 3' flanking region of the chicken H2b gene, additional
consensus sequences can be identified. A wide variety of histone genes
contain a region of dyad symmetry split by a sequence containing four
contiguous T residues (3). This element is seen around nucleotide
417 (Fig. 2). The dyad symmetry is composed of two S-base inverted repeat
sequences. Downstream from this region, most histone genes contain the
sequence -CAAGAAGA- (3). The chicken H2b gene lacks such a sequence but
rather contains in this region at about +440 a sequence more commonly
associated with polyadenylated messages, AATAAA (30). Since most histone
messages appear not to be polyadenylated (2,3) with, however, signifi-
cant exceptions (31,32), it is unusual to find this sequence at the 3'
end of the H2b gene. Studies are underway to determine if any fraction
of the mRNA corresponding to this gene is polyadenylated. D'Andrea et al.
(11) did not find such a poly(A) signal sequence at the 3' end of a

chicken H2a gene; however, they do find an AATAAAA sequence at a similar

60

position in a different chicken H2b gene (unpublished data reported
in Ref. 3). Thus, it remains to be seen if such a sequence has
functional significance for chicken histone genes. The chicken H2b gene
3' end is also unusual in possessing a long polypyrimidine-polypurine
stretch just downstream from the AATAAA sequence (Fig. 2). In the strand
shown, this sequence is comprised of 19 contiguous pyrimidines (15 T,
4 C). The functional significance of such a sequence, if any, is as yet
unknown.

In summary, the complete sequence of this chicken H2b histone gene
confirms and extends previous observations of histone gene relatedness
at both the amino acid and nucleotide sequence levels. However, the
diversity of histone gene organization in the chicken and other higher
vertebrates (5-13) and our studies of other members of the chicken histone
gene family (35) suggest that the conserved elements of gene organization
described above may not, in fact, be ubiquitous among the complete
histone gene family. Further studies which can examine the expression
of the complete family of chicken histone genes (about 10 of each type
per haploid genome (33) throughout development will be required to address

the functional role of histone gene diversity.

Acknowledgments-We thank Drs. R.P. Harvey and J.R.E. Wells for

 

communication of results cited in Ref. 10 prior to publication and Sara
Stadt for technical assistance. We also thank Drs. Michele Fluck and

Ed Fritsch for comments on the manuscript.

61

REFERENCES

l.Isenberg, I. (1979) Annu. Rev. Biochem. 48, 159-191

2. Kedes, L.H. (1979) Annu. Rev. Biochem. 48, 837-870

3. Hentschel, C.C., and Birnstiel, M.L. (1981) Cell 25, 301-313

4. LIfton, R.P., Goldberg, M.L., Karp, R.W., and Hogness, D.S. (1977)
Cold Spring Harbor Symp. Quant. Biol. 42, 1047-1051

5. Zernik, M., Heintz, N., Boime, L., and Roeder, R.G. (1980) Cell
22, 807-815

6. Moorman, A.F.M., de Laaf, R.T.M., Destree, O.H.J., Telford, J., and
Birnstiel, M.L. (1980) Gene 29, 185-193

7. Stephenson, E.C., Erba, H.P., and Gail, J.G. (1981) Cell 24, 639-647

8, Engel, J.D., and Dodgson, J.H, (1981) Proc. Natl. Acad. Sci. USA
28, 2856-2860

9. Harvey, R.P. and Wells, J.R.E. (1979) Nucleic Acids Res. 2, 1787-1797.

10.Harvey, R.P., Krieg, P.A., Robins, A.J., Coles, L.S., and Wells, J.R.E.
(1981) Nature (Lond.) 224, 49-52

11.D'Andrea, R., Harvey, R., and Wells, J.R.E. (1981) Nucleic Acids Res.
2,3119-3128

12.Heintz, N., Zernik, M., and Roeder, R.G. (1981) Cell 24, 661-668

13.Sittman, K.B., Chin, L., Pan, C., Cohn, R.G., Kedes, L.H., and
Marzluff, W.F. (1981) Proc. Natl Acad. Sci. USA ‘28, 4078-4082

14.Dodgson, J.B., Strommer, J. and Engel, J.D. (1979) Cell 22, 879-887

15.Girvitz, S.C., Bacchetti, 8., Rainbow, A.J., and Graham, E.L. (1980)
Anal. Biochem. 206, 492-496

16.Dodgson, J.B., McCune, K.C., Rusling, D.J., Krust, A., and Engel, J.D.
(1981) Proc. Natl. Acad. Sci. USA 28, 5998-6002

l7.Colman, A., Byers, M.J., Primrose, S.B., and Lyons, A. (1978)Eur. J.
Biochem 22, 303-310

18.Maniatis, T., and Efstratiadis, A. (1980) Methods Enzymol. 45, 299-305

l9.Smith, H.D., and Birnstiel, M.L. (1976) Nucleic Acids Res. 2, 2387-
2398

20.Maxam, A.M., and Gilbert, W. (1980) Methods Enzymol. 65,499-560

21.Smith, D.R., and Calvo, J.M. (1980) Nucleic Acids Res. 8, 2255-2274

22.Dolan, M., Sugarman, B.J., Dodgson, J.B., and Engel, J.D. (1981)
Cell 24, 669-677

23.Iwai, K. Ishikawa, K., and Hayashi, H. (1970) Nature (Lond.)22g, 1056-
1058

24.Salser, W. (1978) Cold Spring Harbor Symp. Quant. Biol. 42, 985-1002

25.McGhee, J., and Ginder, G.D. (1979) Nature (Lond.) 280, 419-420

26.Gannon, F., O'Hare, K., Perrin, F., LePennec, J.P., Benoist, C.Cochet,
M., Breathnach, R., Royal, A., Garapin, A., Cami, B., and Chambon, P.
£1979) Nature (Lond.) 278. 428:434. , .

27.Efstratiadis, A., Posakony, J.W., Maniatis, T., Lawn, R.M., O'Connell,
C., Spritz, R.A., DeRiel, J.K., Forget, B.G., Weissmann, S.M., Slightom,
J.L. Blichl, A.E., Smithies, 0., Baraeel, F.E., Shoulders, 8.8., and
Proudfoot, N.J. (1980) Cell 22, 653-668

28.Busslinger, M., Portmann, R., Irminger, J.C., and Birnstiel, M.L.

(1980) Nucleic Acids Res. 8, 957-977

62

29.Hentschel,C.C., Irminger, J.C., Bucher,P., and Birnstiel, M.L.
Nature (Lond.) 285, 147-151

30.Proudfoot, N.J., and Brownlee, G.G. (1976) Nature (Lond.) 242,211-214

31.Hereford, L., Fahrner, K., Woolford, J., Jr., Rosbach, M., and Kaback,
D.B. (1979) Cell 28, 1261-1271

32.Molgaard, H.V., Perucho, M., and Ruiz-Carrillo, A. (1980) Nature (Lond.)
222, 502-505

33.Scott, A.C., and Wells, J.R.E., (1976) Nature (Lond.) 259, 635-638

34.Van Helden, P., Strickland, W.N., Brandt, W.F., and Von Holt, C.
(1978) Biochim. Biophys. Acta 222, 278-281

35.Engel, J.D., Sugarmann, B.J., and Dodgson, J.B. (1982) Nature
(Lond.) 292, 434-436

CHAPTER 2

THE CHICKEN H2b GENE FAMILY *

David K. Grandy, James D. Engel, and Jerry B. Dodgson

Abstract

This paper describes recent progress in our analysis of the
chicken H2b histone gene family. Using a previously identified and
sequenced H2b gene we have shown that chickens contain 7-8 H2b histone
genes per haploid genome. The nucleotide sequence of a second H2b
histone gene is also described. A comparison between this DNA sequence
and that of the previously sequenced chicken H2b gene, H2b-la, revealed
that both genes are continuous and code for identical proteins 125 amino
acids long. In contrast the noncoding flanking sequence 5' to the two
genes share little sequence homology. A small set of sequences has
been conserved, however, whose elements are found associated with
certain other eucaryotic genes. One of these conserved sequences,
ATTTGCATA, is positioned 26 bp 3' to the CCAAT box and 6 bp 5' to the
TATA box in both genes. An analysis of DNA sequence data from other
species suggests that this nine nucleotide sequence is unique to
members of the H2b gene family and may serve a gene-specific regula-

tory function.

*
Reprinted from Gene Expression, pages 445-455 © 1983 A.R.Liss, Inc.,
New York.

63

64

INTRODUCTION

The histones form the primary protein constituent of chromatin
in eucaryotic organisms. Therefore, the regulation of the structure
and expression of the histone proteins represent possible controlling
elements at which major alterations in total cellular gene expression
could occur. The synthesis of histone proteins is generally under cell
cycle control (1,2) with the exception of rapidly dividing sea urchin
embryos (3). The expression of certain histones is also developmentally
regulated. The two best documented examples of this are metazoan
spermatogenesis (4,5) and the maturation of nucleated erythrocytes in
birds, fish, reptiles and amphibians (6-9).

The possible importance of histone-DNA interactions in achieving
cellular differentiation together with the existence of developmental
programs regulating the expression of certain histone genes has led us
to examine the fine structure of these genes in the chicken. The
examination of a wide variety of histone gene-containing chicken DNA
lCharon 4A recombinants showed that members of the five primary his-
tone gene families (H1, H2a, H2b, H3 and H4) were often but not always
linked in chicken chromosomal DNA (10-13), but not in any ordered or
repeating manner such as had previously been observed in invertebrate
histone gene families (1,14,15). Moreover, restriction mapping of
these recombinants demonstrated that each histone gene was located in
a unique region of chicken chromosomal DNA that bore no obvious (large
scale) resemblance to the DNA surrounding other members of the same

family.

65

We have therefore chosen to make a more thorough investigation
of all members of a specific histone gene family, the H2b histone genes,
to see if these genes are expressed in an equivalent or differential
manner and to examine their nucleotide sequence for elements which
might regulate this expression. We initially reported the first complete
sequence of a vertebrate H2b gene (16). This paper describes the
sequence of a separate unlinked H2b histone gene and the comparison
of the two H2b gene structures. Similar (probably identical or allelic)
chicken H2b gene sequences have been recently described by Harvey et a1.

(17).

RESULTS AND DISCUSSION

Analysis of the Complete Chicken H2b Gene Family.

Our initial molecular cloning results (10) were in approximate
agreement with the initial report (18) that chickens contain about
10 members of each primary histone gene family per haploid genome.
We have analyzed the repetition frequency of the H2b gene family in
greater detail by Southern blotting analysis of total chicken DNA using
our previously isolated and sequenced chicken H2b gene (H2b-1a) (16)
as hybridization probe. Total chicken DNA was cut with EcoRI,
electrophoresed on 0.7% agarose, blotted, and hybridized to a labeled
H2b histone gene BstEII fragment internal to the coding regions of
H2b-la (Figure 1). Six primary bands are observed, one or two of
which seem to contain two gene equivalents, linked on a single DNA
fragment or on fragments of identical size. One faint band of consid-
erably less than one gene equivalent is also observed. Thus it appears

that there are about 7-8 H2b histone genes per haploid chicken genome.

66

 

Figure 1. Identification of the chicken H2b histone gene family.
Chicken sperm DNA was cut with EcoRI, electrophoresed, transferred to
nitrocellulose and hybridized to a nick-translated 0.3 kb BstEII DNA
fragment internal to the coding region of H2b-la.

67

(The chicken DNA used was sperm DNA from two or three roosters of an
inbred line provided by the USDA Regional Poultry Research Labs, East
Lansing, Michigan, so it is unlikely that any of the bands represent

heterozygous polymorphisms.)

Construction of a Chicken Histone H2b Gene-Containing Plasmid.

The isolation of the A Charon 4A histone gene recombinants
ACHla and ACHZe have been described (10), as has the subcloning and
sequence analysis of the H2b gene on ACHla (H2b-la) (16). The
restriction map of 1CH2e is shown in Figure 2A. The 3.5 kb* EcoRI
fragment containing H2b sequences (as identified previously, 10) was
subcloned by standard procedures (10). A partial restriction map of

this plasmid, designated pRRZe-3.5, is shown in Figure 2B.

DNA sequence analysis of pRR2e-3.5.

The nucleotide sequence of the H2b gene and its 5' flanking
sequence on pRR2e-3.5 was determined by the method of Maxam and
Gilbert (19) as modified by Smith and Calvo (20) and is presented in
Figure 3. This H2b gene was identified by the 125 amino acid sequence
it predicts, which is identical to that of the chicken H2b gene pre-
viously isolated (on plasmid pKRla-l.3) (16). A comparison of the two
H2b genes reveal that their coding sequences differ in eight nucleo-
tides (Figure 3). These point mutations are clustered at the 5' end
of the H2b-2e gene in codons l, 4, 7, 10 and 14 and at the 3' end in

codons 96, 112 and 114. In each case the nucleotide polymorphism occurs

*The abbreviations used are: kb, kilobase pairs; bp, base pairs.

68

 

 

3.0 no ,I \
x \
x \\
I
II” \\
1” \
I \
,’ \
I’ \\
I
I” \\
I
fine-”W 1 I w
10 660] 0 a

fast one‘

p—fi
0.4KB

f a.- m ' sum usu- Vuuu an Ass.
Ann luau Alan! '1“
0 an a a mu II ? P-H ...asaaa

Figure 2. Restriction maps of ACHZe and pRRZe-3.5.
A, a portion of the clone ACHZe containing the chicken DNA
insert is shown. B, a partial restriction map of the re-
combinant plasmid pRR2e-3.5. The arrows indicate the di-
rection and'extent to which the DNA sequence was read from
individual experiments. The H2b coding region is positioned
on this map within the initiation codon (ATG) and the term-
ination codon (TAA) as designated.

69

in the degenerate third position of the codon and is silent with
respect to the amino acid sequence. The effect of these changes on
the primary structure of the H2b-2e gene is to increase the number of
rare CpG dinucleotides to 40 compared to 37 in H2b-la and to reduce
the number of A or T residues in the third codon position from 4 in la
to 2 in 2e. As described previously (13,16), most histone genes
contain a very high percentage of G-C base pairs in degenerate sites
leading to the high ZGC of the coding sequence and the comparatively
high frequency of CpG dinucleotides. In addition to the silent coding
sequence polymorphisms the stop codon of H2b-2e differs from that of

H2b-la; the former being UAA (ochre) and the latter being UAG (amber).

Sequence conservation in the 5' gene-flanking regions of H2b-1a and
H2b-2e.

In an attempt to identify conserved sequences which may play a
role in regulating the expression of chicken H2b histone genes we have
examined the noncoding 5' regions flanking the two genes. It is
apparent from the sequence data (Figure 3) that H2b-1a and H2b-2e
present very divergent 5' gene-flanking sequences. However, contained
within these divergent sequences are blocks of homology which have been
conserved. The general lack of homology outside of the coding sequences
and consensus promoter sequences suggests that individual H2b histone
gene-specific probes for transcribed sequences can be constructed to
analyze the expression of specific chicken H2b genes. Further work
along these lines is in progress. By analogy with other histone genes
(14) a putative mRNA initiation locus or cap site, CAgTC, has been

identified in the 5' region of both genes. This "cap box" is located

70

I
la 5 AAT1C11AlCCACTAUCACCCACCGATTTCTCATCCACEA
29-ZOOCCAATCCAATTTGAACCAAICAAAACCACTTATAACCCAC
QIIACACACCCCTTCCTATCCACCCACACCTCACCCACTCTCTTAlCCAATCACACACCACATACAGAACCCACTCCAII
GAAACGCCTCTTCTCCACTTCCCACCAATCAAACACTCCCAAACCAATCCTTCTCATTTCCATAGACCCCCTATAAATAA

 

 

-80
rCCATACTCCCCCIATAAAIAGGCCACCACICCICCCAGCCCCCACICCCCICCCCLCAAGGCAICGIGCAGACIICGAC
AIGCCIACCACCCCITCGITTCCATICACCCtttttxeertbtt11ICIILCLLILCLItteteACtCLCttetetcACI

 

T C A C C
l .. .
ATC LLC GAG CCC GCT AAC TCC CCC CCC CCC CCC AAC AAC CCC TCT AAC AAC CCC CTC ACC
Pro Clu Pro Ala Lye Scr Ala Pro Ala Pro Lys Lys Cly Ser Lys Lys Ala Val Ihr

+01
AAC ACC CAG AAC AAC CCC CAC AAC AAC CCC AAC AAC ACC CCC AAC CAC ACC TAC ICC ATC

Lye Ihr Cln Lye Lys Cly Asp Lye Lye Arg Lye Lye Ser Arg Lye Clu Ser Iyr Ssr lle

+ 2
' IIAC GIG IAC AAC GIG CIG AAC CAC GIG CAC CCC GAC ACC GCC AIC ICC ICC AAC GCC AIG

Iyr Val Iyr Lys Val Lcu Lye Gln Val His Pro Asp Ihr Cly Ila Set Set Lye Ala Hat

+ .
IslCCC ATC ATC AAC TCC TTC CTC AAC CAC ATC TTC CAC CCC ATC CCC CCC CAC CCC TCC CCC

Cly llc Met Asn Scr Phc Val Aen Asp 11c Phc Glu Arg 11c Ala Cly Clu Ala Scr Arg

72"CTC GCC CAC IAC AAC AAC CCC ICC ACC AIC ACC ICC CCC GAG AIC CAG ACC GCC GIG CCC
Lau Als His Iyr Aen Lye Arg Ser Ihr lla Ihr Scr Arg Clu lle Gln Ihr Ala Val Arg

*JO'CIC CIC CIC CCC GCC GAG CIC CCC AAC CAC GCC GIC ICC GAG CCC ACC AAC GCC GIC ACC
Lsu Leu_Lau Pro Cly Glu Lao Ala Lye His Ala Val Sar Glu Cly Ihr Lye Ala Val Ihr

C 9
”“Mc mc acc «cc ch AAC IM ,2“

Lys Iyr Thr Scr Sar Lye

Figure 3. The nucleotide sequence of two chicken H2b
histone genes. The DNA sequence corresponding to the mRNA
strand is presented for both genes. The predicted amino
acid sequence is given below the sequence of H2b-2e. The
entire 5' region flanking H2b-1a is shown above the 2e se-
quence. The underlined sequences are those which are shared
by both 1a and 2e and are discussed in the text. Only those
bases which are unique to H2b-1a are shown above the protein-
coding sequences of H2b-2e.

71

32 bp and 53 bp upstream of the initiation codon and 22 bp and 23 bp
downstream of the TATA box (see below) in H2b-1a and H2b-2e respectively.
The precise locations of the sites at which the transcripts are initiated
have not yet been determined. The AT-rich block which bears a very
strong resemblance to the Hogness (21) or TATA box consensus sequence
associated with histones and other genes transcribed by RNA polymerase

II (14,21) is identified in H2b-la and H2b-2e by the sequence, CTATAAATA.

Another distinct nucleotide sequence often found associated with
RNA polymerase II-transcribed genes is given by the canonical sequence
CCAAT (14,22). Both H2b genes in fact contain two such elements in
their promoter regions. The first such element is found in both H2b
genes 42 bp 5' to the TATA box. The second CCAAT box is located 43 bp
and 44 bp 5' to the first CCAAT box in H2b-la and H2b-2e respectively.
The significance of a repeated CCAAT box is unknown, however similar
duplications have been observed in an embryonic a-globin gene (23) and
in certain y-globulin genes (22). A summary of these consensus
sequences and their locations in the flanking regions of H2b-la and
H2b-2e is presented in Figure 4.

In addition to the putative sequence elements recognized by the
polymerase II transcription apparatus a sequence of nine nucleotides,
ATTTGCATA, has been identified 26 bp 3' from the 3'-most CCAAT boxes
and 6 bp 5' to the TATA box in both H2b gene flanking regions. The
function, if any, of this AT-rich sequence is unknown, however it may
serve some H2b gene-specific regulatory role. This supposition is
supported by the fact that the sequence has been well conserved through
evolution. An analysis of the DNA sequence data available for a

variety of H2b genes reveals that this sequence occurs in the sea

'72

Ins-16239
lax-185sv *1

3 la 0
ccm- 32a» -ccm- 2639 armam- 639 -CTATAAATA- fir» -CA§Tc- fill; 416
2a 2a 2a

TANDEM CCAAT BOXES TATA 30! CAP BOX START

H28
SPECIFIC

Figure 4. Sequence and spacing homologies found in the
5' flanking regions of two chicken H2b genes. This is a
summary of the conserved sequences found in both genes and
discussed in the text.

Cu H23-1A CCAAT"' 2639 "'ATTTGCATA"' Bap '°'CTATAAATA"° 593? "'ATG
Cu H23-2: CCAAT"' 263p "'ATTTGCATA"' be? "'CTATAAATA°" 8139 "'ATG
X. LAEVlS CCAGT"' 2639 "'ATTTGCATG"' 639 "'CTATAAATA"' 5939 '°'ATG
S. P.-2 CCAAT"' 2239 "'ATTTGCATA"° 3039 "TATAAAAAG°" 963p °"ATG
P. n. «19 CCAATH' 223p '“ATTTGCATA’” 323v “TATAAAAAG'” 9639 “'ATG
P. n. H22 CCAAT°" 3139 "°ATTTGCATA"° 263? "TATAAAGAG"' 97s? "'ATG

H23 SPECIFIC
consensus sequence: ATTTGCATé

Figure 5. A conserved nine nucleotide sequence unique
to H2b genes. A comparison of the available DNA sequence
data from several species reveals the presence of a well con-
served sequence element located between the CCAAT and TATA
boxes in the noncoding 5' region flanking H2b genes. Se-
quences are from chicken (see figure 3), S, purpuratus (1,
24), g, miliaris (1, 25, 26) and 2, laevis (27).

73

urchin species S, purpuratus (1,24) and P, miliaris (1,25,26) and a
nearly identical sequence has been found 6 bp 5' to the TATA box in
an 2, laevis H2b gene (27). A summary of these homologies is presented
in Figure 5. A similar observation has been made by Harvey et al. (17)
in their analysis of chicken H2b gene sequences.

The consensus sequence, GATCC. which has been found associated
with several histone genes (14) was not found associated with H2b-la
or H2b-2e. This sequence is also missing from the 5' region flanking
Drosophila (l4) and Xenopus (27) H2b genes and therefore its signif-

icance is unclear.

ACKNOWLEDGMENTS
This work was supported by Grant GM 29687 from the National
Institutes of Health and by Grant I-759 from the March of Dimes Birth
Defect Foundation. Journal Article No. 10823 from the Michigan

Agricultural Experiment Station.

9.

74

REFERENCES

Kedes, LR (1979). Histone genes and histone messengers. Ann Rev
Biochem 48:837.

Hereford LM, Osley, MA, Ludwig, JRIII, McLaughlin CS (1981) Cell
cycle regulation of yeast histone mRNA. Cell 24:367.

Woodland, HR (1980). Histone synthesis during the development of
Xenopus. FEBS Letters 121:1.

Geraci, G, Lanciere, M, Marchi, P, Noviello, L (1979). The sea
urchin (Sphaerechinus granularis) codes different H2b histones to
assemble sperm and embryo chromatin. Cell Diff 8:187.

 

. Seyedin, SM, Kistler, WS (1979). H1 histone subfractions of mammal-

ian testes. 2. Organ specificity in mice and rabbits. Biochemistry
18:1376.

. Molgaard, HV, Perrucho, M, Ruiz-Corrillo, A (1980). Histone H5

messenger RNA is polyadenylated in chicken erythrocytes. Nature
(London) 283:502.

Levin, B (1980). "Gene Expression 2." New York: Wiley-Interscience,
pp 311-312.

. Destree, OHS, Hoenders, HJ, Moorman, AFM, Charles, R (1979).

Histone of Xenopus laevis erythrocytes. Purification and character-
ization of the lysine-rich fractions. Biochim Biophys ACTA 577:

61.

Risley, MS, Eckhardt, RA (1981). H1 histone variants in Xenopus
laevis. Dev Biol 84:79.

 

10.Engel, JD, Dodgson, JB (1981). Histone genes are clustered but not

tandemly repeated in the chicken genome. Proc Nat Acad Sci USA
78:2856.

11.Harvey, RP, Wells, JRE (1979). Isolation of a genomal clone contain-

ing chicken histone genes. Nucl Acids Res 7:1787.

12.Harvey, RP, Krieg, PA, Robins, AJ, Coles, LS, Wells, JRE (1981).

Non-tandem arrangement and divergent transcription of chicken
histone genes. Nature (London) 294:49.

13. D'Andes, R, Harvey, RP, Wells, JRE (1981). Vertebrate histone

genes: nucleotide sequence of a chicken H2a gene and regulatory
flanking sequences. Nucl Acids Res 9:3119.

l4.Hentschel, CC, Birnstiel, ML (1981). The organization and expression

of histone gene families. Cell 25:301.

15.Lifton, RP, Goldberg, ML, Darp, RW, Hogness DS (1977). The organ-

ization of the histone genes in Drosophila melanogaster: functional
and evolutionary implications. Cold Spring Harbor Symp Quant Biol
42:1147.

 

l6.Grandy, DK, Engel, JD, Dodgson, JB (1982). Complete nucleotide sequence

of a chicken H2b histone gene. J Biol Chem 257:8577.

17.Harvey, RP, Robins, AJ, Wells, JRE (1982). Independently evolving

chicken histon H2b genes: identification of a ubiquitous H2b-specific
5' element. Nucl Acids Res 10:7851.

l8.Crawford, RJ, Krieg, P, Harvey, RP, Hewish, DA, Wells, JRE (1979).

Histone genes are clustered with a lS-kilobase repeat in the chicken
genome. Nature (London) 279:132.

75

19.Maxam, AM, Gilbert, W (1980). Sequencing end-labeled DNA with base-
specific chemical cleavages. Methods Enzymol 65:499.

20.Smith, DR, Calvo, JM (1980). Nucleotide sequence of the S, coli
gene coding for dihydrofolate reductase. Nucl Acids Res 8:2255.

21.Salser, W (1978). Globin mRNA sequences: analysis of base pairing
and evolutionary implications. Cold Spring Harbor Symp Quant Biol
42:985.

22.Efstradiatis, A, Posakony, JW, Maniatis, T, Lawn, RM, O'Connell, C,
Slightom, JL, Blechl, AE, Smithies, 0, Baralle, FE, SHoulders, CC,
Proudfoot, NJ (1980). The structure and evolution of the human
B-globin gene family. Cell 21:653.

23.Enge1, JD, Rusling, DJ, McCune, KC, Dodgson, JB (1983). Unusual
structure of the chicken embryonic a-globin gene,fl'. Proc Natl
Acad Sci USA 80:1392.

24.Sures, I, Lowry, J, Kedes, LB (1978). The DNA sequence of sea
urchin (S, purpuratus) H2a, H2b and H3 histone coding and spacer
regions. Cell 15:1033.

25.Busslinger, M, Portmann, R, Irminger, JC, Birnstiel, ML (1980).
Ubiquitous and gene-specific regulatory 5' sequences in a sea
urchin histone DNA clone coding fonrhistone protein variants.
Nucl Acids Res 8:957.

26.Schaffner, W, Kung, G, Daetwyler, H, Telford, J, Smith, HO, Birnstiel,
ML (1978). Genes and spacers of cloned sea urchin histone DNA
analyzed by sequencing. Cell 14:655.

27.Moorman, AFM, DeBoer, PAJ, DeLaaf, RTM, Destree, OHJ (1982). Primary
structure of the histones H2a and H2b genes and their flanking
sequences in a minor histone gene cluster of Xenopus laevis.
FEBS Letters 144:235.

 

 

 

CHAPTER 3

GENOMIC ORGANIZATION, STRUCTURE AND EXPRESSION

OF THE CHICKEN H2b HISTONE GENE FAMILY

David K. Grandy and Jerry B. Dodgson

The organization, structure and expression of the chicken H2b his-
tone gene family has been examined. A 0.3 kb chicken H2b gene-coding
sequence was used to probe Southern blots of genomic DNA. The results
of these experiments suggest that the chicken H2b histone gene family
contains an estimated 14-16 loci per haploid genome and that the sequence
of one of these genes has diverged considerably from the others. The
chicken H2b gene probe was also used to screen a A Charon 4A chicken
genomic library. Eight unique phage clones were isolated and their re-
spective restriction maps were determined. These clones are in addition
to the two H2b histone gene-containing phage clones previously isolated
as the result of their linkage to H3 and H2a histone gene sequences. The
chicken H2b histone genes are found in clusters of other histone genes
and these clusters are not tandemly repeated. The portions of the phage
which contain the H2b gene sequences were subcloned into pBR322. Eleven
subclones were obtained, mapped with restriction enzymes and their DNA
was sequenced by a modification of the chemical degradation method. Sever-

al nearly identical H2b genes were discovered which, however, represent

76

77

different H2b loci. Nine unique H2b histone coding sequences were ident—
ified. None of these H2b genes contain introns. Three different H2b his-
tone protein variants are predicted by the DNA sequence data. An analysis
of the 5' flanking sequence data reveals that these genes possess CCAAT
and TATA boxes, elements commonly associated with genes transcribed by
RNA polymerase II. In addition, these genes share an H2b-specific sequence
element in their 5' untranscribed region of the form: % T E TGCATA g ,
located between the CCAAT and TATA boxes. The sequences which are 3' of
the chicken H2b histone exons contain the hyphenated symmetrical dyad
homology: GGCTCTTTTC % GAGCC, which is shared by histone genes in general.
In addition to the conserved symmetrical dyad,eight of the nine chicken
H2b histone genes possess a sequence element downstream of the coding
region that resembles: CAAGAAAGA, commonly found in the 3' region follow-
ing other histone exons. One chicken H2b histone gene lacks this sequence
and possesses the canonical poly A addition signal: AATAAA, instead. The
transcripts from this particular gene are polyadenylated. These tran-

scripts are found at low levels in adult chicken tissues and at higher

levels in chick embryos.

78

INTRODUCTION

The DNA of all organisms is organized through its association
with protein. In eucaryotic chromatin the basic subunit of this
organization is the nucleosome (Kornberg, 1977; McGhee & Felsenfeld,
1980), a structure consisting of a core particle of histone proteins
around which the DNA is wrapped and thereby compacted. The histones
are, therefore, responsible for presenting the DNA to its environment
in a particular conformation, which may or may not lead to the express-
ion of the information contained within the nucleotide sequence.

The role, if any, of histone proteins in the regulation of gene
expression remains unclear. In this regard the diversity of histone
families in several species has not been fully appreciated. Early work

with sea urchins and 2, melanogaster histone gene families (Kedes, 1979)

 

showed that they were organized into tandemly repeated clusters in
which each repeat unit contained a complete set of H1, H2a, H2b, H3
and H4 sequences (a single copy of each gene per repeat). The histone
genes in these families are coded in single continuous units (no
introns), and their transcripts lack the 3' polyadenylic acid polymer
common to most eucaryotic mRNA (Kedes, 1979; Nevins, 1983).

In recent years this early portrait of the histones has been
considerably broadened. It is now evident that not all histone
expression is coupled to DNA synthesis (Zweidler, 1984). A number of
histone protein variants can be synthesized in the absence of DNA
replication (Sittman, et al., 1983; Groppi & Coffino, 1980), at

different stages of development (Urban & Zweidler, 1983; Childs,

79

et al., 1979; Roberts, et al., 1984) and/or in specific tissues
(Edwards & Hnilica, 1968; Risley & Eckhardt, 1981; Zweidler, 1984).
Some of these variants are coded for by genes which contain introns
(Engel et al., 1982; Woudt et al., 1983) and several histone trans-
cripts are known to be polyadenylated (Kreig et al., 1983; Engel et al.,
1982; Ruderman & Pardue, 1978; Borun et al., 1977). Furthermore, the
relatively simple organization of histone gene families characteristic
of the lower eucaryotes is not shared by vertebrates such as chicken
(Engel & Dodgson, 1981; Harvey et al., 1981), mice (Sittman et al.,
1981), humans (Sierra et al., 1982), nor even the more recently
examined late histone genes of the sea urchin (Maxson et al., 1983;
Roberts et al., 1984).

In order to better understand the processes involved in the regu-
lation of eucaryotic gene expression we have continued our studies into
the genomic organization, structure and expression of the chicken H2b
histone gene family. The relatively limited number of chicken histone
genes, predicted by solution hybridization (Crawford et al., 1979) and
genomic blotting experiments (Engel & Dodgson,1981; Grandy et al., 1982),
made this family an attractive candidate in which to study histone
expression in a higher eucaryote. Through the characterization of an
entire histone gene family we hope to correlate specific histones with
the genes which encode them. When these proteins are expressed in a
regulated manner, an analysis of the DNA sequences of their corresponding
genes should reveal regulatory elements associated with, if not
responsible for, the control of their expression. Furthermore, when

all the members of a particular gene family are compared at the

80

nucleotide sequence level homologies can be identified. In this way
one can take advantage of millions of years of evolutionary diver-
gence and focus site-specific mutagenesis experiments on the conserved
sequences in hopes of identifying their function.

In this paper we report the characterization of 11 chicken H2b
genes subcloned along with extensive flanking sequences from 9 genomic
phage clones. Much of the DNA sequence of these genes has been deter-
mined. A comparative analysis of these sequences has revealed that each
gene is composed of a unique nucleotide sequence, and that three H2b
protein variants are predicted from these sequences. The 5' flanking
regions display sequences common to other genes transcribed by RNA
polymerase II. One or more CCAAT elements, a TATA box and a less well
conserved capping site are associated with each gene. In addition,
an H2b-specific element is associated with each H2b gene's TATA box.
The 3' flanking regions are rich in conserved sequences, presumably
involved in the proper processing of the primary transcript (Birchmeier
et al., 1983). In their 3' flanking regions all of these genes share
a palindromic sequence which is followed, in all but one case, by a
well-conserved spacer element. One gene lacks this spacer and in its
place exhibits the canonical poly A addition signal AATAAA (Proudfoot &

Brownlee, 1976).

81

MATERIALS AND METHODS
(a) Materials

Restriction enzymes and other nucleic acid modifying enzymes were
purchased from International Biotechnologies, Inc., New England Biolabs,
Inc. and Bethesda Research Labs, Inc. and used according to the manufac-
turer's specifications. Polynucleotide kinase was obtained from
PL Biochemicals. Ultrapure urea, acrylamide and bis-acrylamide were
purchased from International Biotechnologies, Inc. Other materials and

bacterial strains used were as described previously (Grandy et al.,

1982).

(b) Isolation and Purification of Recombinant DNA molecules

The 1 Charon 4A chicken DNA library used in these studies was
constructed by Dodgson et al., 1979 and contains genomic fragments
generated by AluI/HaeIII partial digestion of chicken DNA ligated to
synthetic EcoRI linkers. Phage were plated at a density of 50,000
plaques per plate and approximately ten genomic equivalents were
screened. The library was screened according to the filter hybrid-
ization procedure of Benton and Davis (1977) using a previously
described H2b gene-specific 0.3 kb BstEII fragment from pKRla-l.3 as
probe (Grandy et al., 1982). The probe was made radioactive by nick-
translation according to the procedure of Maniatis et al. (1982).
Positive plaques were picked and purified by at least one additional

screening step. Phage were prepared by banding over a glycerol step

82

gradient (Maniatis et al., 1982) and the DNA was prepared according to
Maniatis et al. (1982). Phage were restriction mapped and hybrid-
ization mapped for H2b sequences as described previously (Grandy et al.,
1983). H2b sequence-containing DNA fragments were purified from phage
DNA digests and subcloned into pBR322 as described (Girvitz et al.,
1980; Grandy et al., 1982). Bacteria containing recombinant subclones
were screened for the insert of interest by the filter hybridization
procedure of Grunstein and Hogness (1975). DNA was prepared by the
alkaline lysis procedure (Birnboim & Doly, 1979) and purified over a

two step gradient of cesium chloride and ethidium bromide (Garger

et al., 1983). The banded DNA was collected, extracted with isobutanol,
extensively dialyzed against 10 mM Tris HCl (pH7.5), 1 mM EDTA (TE).

All manipulations of recombinant DNA molecules were done in accordance

with current National Institutes of Health guidelines.

(c) Filter Hybridization

High molecular weight genomic chicken red blood cell DNA was a
generous gift of W. Lewis. This DNA was digested with restriction
enzymes, precipitated, loaded onto 0.7% agarose gels at a concentration
of 15 micrograms per lane, electrophoresed in a vertical gel support,
stained with ethidium bromide and photographed. The DNA was denatured
in alkali, the gel was then neutralized and DNA was transferred to
nitrocellulose filters by capillary action as described by Southern
(1975). The filters were prehybridized and probed in a hybridization
solution previously described (Engel & Dodgson, 1981) at 370 or 420
overnight. The Southern blots were washed in 0.1xSSC (leSC = 0.15 M

NaCl 0.015 M Na citrate, pH 7.0) at either 370 or 65°, respectively,

83

and exposed to Kodak XAR film with DuPont Cronex intensifying screens

0
at -70 for one to two weeks.

(d) Restriction Enzyme Site Mapping

Restriction enzyme sites were mapped to coordinates on the phage
clones as well as on the subclones. This was achieved by digesting
the DNA of interest with several enzymes singly and in pairwise combin-
ations, electrophoresing the fragments in vertical agarose or poly-
acrylamide gels, staining the gels with ethidium bromide, photographing
and blotting the DNA to nitrocellulose as described above. These
mapping blots were then probed with the H2b-specific fragment and the
hybridizing fragment(s) identified and mapped to a particular locus.
The generation of accurate maps was essential to the development of an

effective DNA sequencing strategy..

(e) DNA Sequencing

DNA sequencing was performed by a modified Maxam and Gilbert
chemical degradation technique as described by Smith and Calvo (1980).
DNA fragments with 5' protruding ends were phosphatase treated and end-
labeled with [y-32P] and polynucleotide kinase as described (Maxam &
Gilbert, 1980). Blunt-ended fragments were labeled either by exonucle-
ase III digestion followed by filling in with the appropriate labeled
nucleotide as described (Smith and Calvo, 1980) or by directly kinasing
the phosphatased ends as described by Maxam and Gilbert (1980).
Sequencing gels of 6% and 8% were loaded three times on a schedule of
14 hours at 30 Watts for the first load; 6 hours at 40 Watts for the
second and approximately two hours at 50 Watts for the third load.

The gels were then mounted on Whatman 3MM filter paper, cut in half,

84

covered with plastic wrap and dried in a Biolab gel dryer for 20
minutes at 80°. The gel was then exposed to Kodak XAR film for one to

ten days employing intensifying screens as required.

(f) Sl Mapping Analysis

51 mapping experiments were performed essentially as described
by Sollner—Webb & Reeder (1979). The DNA probe was precipitated with
total adult liver, total breast muscle, total chick embryo fibroblast
(a gift from M. B. Raines), total 4.5 day old embryo (gift from J. D.
Engel), red blood cell cytoplasmic and total HD3 or poly A+ HD3 mRNAs
( gift from M. Federspiel). The DNA and RNA were heat-denatured in
S1 hybridization buffer and allowed to slowly reanneal at 550 for
4 hours. The resulting hybrids were then digested with 103 units of
S1 nuclease at room temperature for 15 minutes. The S1 nuclease was
purchased from Sigma Chemical Co. The reaction was stopped and
precipitated. The precipitates were dissolved in formamide loading

mix, heat-denatured and electrophoresed in 6% polyacrylamide sequencing

gels along with Maxam and Gilbert sequencing reactions as markers, the

gels were dried and exposed to film.

RESULTS
(a) Estimate of the number of chicken H2b histone genes
In order to assess the number of H2b histone genes present in the
haploid chicken genome, Southern blotting experiments were performed.
Chicken red blood cell DNA was prepared and digested with the restri-
tion enzymes EcoRI and HindIII. These enzymes were chosen because

they do not cut within either of the two H2b genes previously

85

characterized (Grandy et al., 1983). The DNA fragments were electro-
phoresed, blotted to nitrocellulose filters and then probed with a

nick translated 0.3 kb BstEII restriction fragment prepared from
pKRla-l.3 (Grandy et al., 1982). As previously described, this clone
contains a single H2b gene, and the 0.3 kb BstEII fragment consists
solely of H2b coding sequence. Hybridization was carried out in 50% for-
mamide at 370 or 420 for 16 hours. The filters were then washed several
times at 370 or 650 in 0.1xSSC.

The results of such an experiment are shown in Figure 1. The
H2b-specific probe hybridizes to eight distinct bands under the more
stringent conditions (Figure 1, Lane A). These bands correspond to sizes:
10 kb, 8.0 kb, 7.8 kb, 7.2 kb, 5.0 kb, 4.0 kb, 3.6 kb and 2.0 kb. The
pattern observed is in agreement with the results of Ruiz-Carrillo et
al. (1983). One striking feature of this autoradiogram is that the
bands are not of equal intensity. This differential signal strength
is most likely a reflection of the number of H2b genes in each band,
since the probe sequence is completely within the region of conserved
H2b coding sequence. This suggests that a particular restriction
fragment may contain more than one H2b gene; some fragments may be
represented more than once per haploid genome; an H2b gene may
contain an internal EcoRI or HindIII site resulting in fragments
containing fractions of an H2b gene; and/or there are divergent
H2b genes which are only partially homologous to the coding sequence
probe. Except for the 1.6 kb band in Figure 1, Lane B discussed
below, the last explanation appears unlikely. 0f the 9 completely
sequenced H2b genes (see below), all show a very high level of nucleotide

sequence homology to the gene used as probe. Also, none of the

86

Figure 1. Genomic Southern blot of EcoRI/ HindIII digested chicken red
blood cell DNA. The blots were probed with a BstEII 0.3 kb H2b-specific
sequence nick translated to a specific activity of 1x108 cpm. 100ng of
the 0.3 kb probe was used in 7 ml of hybridization solution. The washed
blots were exposed to film with an intensifying screen for 10 days at
-70°. (A) The hybridizations were carried out at 42° and then the filter

was wahed at 650 in 0.1xSSC. (B) The hybridization was performed at 37°
and the filter was washed at 37°.

87

 

 

Figure 1.

88
cloned genes contain an internal EcoRI or HindlII site, and it seems un-

likely that any of the H2b genes yet to be cloned contain such sites be-
cause of their small size and high degree of sequence homology. Thus we
favor one or both of the first two possibilities mentioned above. It is
therefore likely that the faintest bands at 3.5 and 4.0 kb correspond to
restriction fragments which contain one gene copy. The stronger bands at
10, 8.0, 7.2, 5.0 and 2.0 kb represent twice the copy number represented
by the fainter bands and the very strong signal at 7.8 kb may correspond
to three H2b gene equivalents.

When the experiment is performed at 370 under less stringent condi-
tions a ninth band at 1.6 kb appears (Figure 1, lane B). The fact that
the probe can be washed off at an elevated temperature is indicative of
reduced sequence homology between the probe and the sequence contained in
the 1.6 kb band. This result defines a different class of H2b gene se-
quence from those which share more sequence homology with the probe.
Based on these results, the chicken H2b histone gene family consists of

an estimated 14-16 different members per haploid genome.

(b) Isolation of the chicken H2b histone genes

Upon screening a chicken genomic library in A Ch 4A with sea urchin
H3 and H2a histone gene probes, about 50 histone gene-containing recom-
binants were isolated (Engel & Dodgson, 1981; Sugarman at al.,1983).
About 20 of these have been mapped and characterized in detail. Several
of these clones were shown to contain H2b-hybridizing regions. We have
previously described the complete sequence of two of these H2b genes
(Grandy et a1, 1982; Grandy et al., 1983). These two genes were found
to differ in both their coding and flanking sequences suggesting that

it would be necessary to examine a number of different H2b genes in

89
in order to better understand the diversity of this gene family and the
sequences regulating its expression.

Since an H2b probe was not used in the initial screening, the chick-
en genomic library was rescreened with an internal 0.3 kb BstEII H2b-
specific fragment in order to expand our collection of chicken H2b genes.
Twentynine positive plaques were picked and their phage DNAs were pre-
pared. A comparison of BstEII digestion patterns and their hybridization
to the coding sequence probe revealed that eight of these phage clones
were different enough to warrant further analysis.

In order to map these phage DNAs the restriction enzymes EcoRI,
BamHI and HindIII were used singly and in pairwise combinations. These
digests were electrophoresed, photographed, blotted and probed with the
H2b-specific fragment. From these data restriction maps were constructed
for each of the phage clones. Figure 2 summarizes the mapping and hybri-
dization data for all of the unique chicken H2b histone gene-containing
recombinants analyzed to date.

The results of these mapping experiments reveal that the organization
of the chicken H2b gene family is complex. All of the H2b genes contained
on these phage clones occur within clusters of other histone genes and most
commonly are found closely linked to an H2a gene. Three of these phage
clones contain two different H2b genes whereas the remaining seven con-
tain one H2b gene each. One of these clones (A Ch5e) contains the same
H2b gene that is on A Ch2e. Thus our collection contains 12 different
H2b loci.

The restriction enzyme and mapping blot data obtained for these
phage served as the basis upon which a subcloning strategy for each H2b

gene was developed. The phage DNA was digested with the appropriate

90

Figure 2. Restriction enzyme cleavage maps of the ten chicken H2b histone
gene-containing phage. These phage clones have not been completely mapped
with respect to the H2a, H3 and H4. The solid horizontal arrows mark the
H2b hybridizing regions and point in the direction of transcription, where
this information is known. The solid horizontal bars identify H2b-hybidiz-
ing regions but are incompletely characterized. The 1 phage Chla, 2e, 5e,
3c and 2d were independently isolated by Sugarman et'al. (1983). The phage
clones Chla and 2e were previously described by Grandy et al. (1982;1983).

91

i-——i 2K3
WW X01118
H3 H3 H1

1L: [loll Acmo

H2b

 

 

11—1 4'1 l l l .11 A0111

H2b

ML '0 \‘2 ml (11 )\Ch4a
H2b

“'LC—D—CTQV a $5M A0129

H3 H28 H4 ’1‘ H1 H28H2b

 

 

11 A 1% H \U chse

5 H1 H28H2b

 

lid/Q £7 \1/ fiflLJ/C. _ _______‘U >\Ch3c
7F H28 H28 H28 H2b

L ‘ ‘V \1 \1/ v 11711 )\Ch2d

H28 H2b H1 [1‘

.IL . ..91 v1 J) Ac“
H2b H2aH2b 7F

lit—4.? <7 0 _S? v TT 1) Xensa

H28 H2b

 

 

 

 

== H2a
_ = H4

\U=EcofiI‘ ,‘llecoRlv- =HINDIIIA= KPNI' ;/1\=BAMHI‘ -= Direction oi
linker Transcription

Figure 2.

92

restriction enzyme(s) and the H2b-containing fragment was then gel puri-
fied and subcloned into the appropriate restriction sites of the plasmid
pBR322. The plasmid DNA was prepared and then digested with a variety

of restriction enzymes, The electrophoretic and blotting data obtained
for each clone was organized into a fine structure restriction map. These
maps are diagrammed in Figure 3 for the eleven subclones obtained. The

H2b-containing sequences present on A Chl have not been subcloned.

(c) DNA sequencing strategy for the H2b histone genes

The fine structure restriction maps generated for each subclone
were used to develop a sequencing strategy for the individual H2b genes.
The method of Maxam and Gilbert (1980) as modified by Smith and Calvo
(1980) was used to determine the DNA sequence for each of the H2b genes
subcloned. The horizontal arrows in Figure 3 indicate the extent of the
region sequenced and the direction in which the sequence was read. The
arrows above the clone map indicate that the DNA strand with the same
sense as the mRNA was read and the arrows under the clone identify se-
quences read which correspond to the template strand of the DNA molecule.

From this sequence data the exact location of each H2b gene was
identified as was the direction in which they are transcribed. This di-
rection is indicated in Figure 3 by the open horizontal arrows. Some
sequence data was also obtained for a few H2a and H1 genes (Browne, 1984)
which are closely linked to particular H2b genes. The location and
transcriptional orientation of these genes are indicated on the maps.
The DNA sequence results indicate that in the four instances where an
H2b gene and an H2a gene are closely linked to each other, they are in-
variably transcribed in a divergent fashion. This interesting arrangement

fixes the promoter region for both genes in the 340 bp region which

93

Figure 3. Restriction enzyme maps of the chicken H2b-containing subclones.

The open horizontal arrows mark the various genes and point in the direction
of transcription. The horizontal arrows above the map indicate the extent
of the DNA sequence read which corresponds to the mRNA. The horizontal arrows
below the map indicate the extent of the DNA sequence determined for the tem-
plate strand. .

94

 

 

 

 

 

 

 

 

 

—-s ‘= ——>
O
.. ' KR1 -1.3
02 h—Z: p a
<—<——
—4
.. l a
E ‘1 pRR4a-2.2
, C- .
b
H2b - 1 mam-4.4
T l
e—
!
“—5 H1 ,P H2a H2b pFiR2e-3.5
E pRRac-5.4
——L__,
«a H2b 1; "’
. o 1
6; (———
l T TTTVIA i? pRRac- -.3 s
'_'L__>
1w 1' TPPP2°'4-°
6 6
(—
——->
.l 11 fl__’pPP2d-2 o
E: HINFI E
i: HINCH ——9
5: 1(er ’1 L126 , pBBA-3.0
i: ssrn 3
= BAMHI
= HlNOlII /r s e T ’
"= BGLI , (l {gar—‘1‘” ”BRA-5'4
’ = Xhol ' s__°
:: BSTEH
I: AVAI
= PSTI 39923 L T <1 an 1135' HRSa-3.3
1= EcoRI m °
£= EcoRI A
LINKER

2: 3m; Figure 3.

95

separates the two genes. Eight H2b genes have been completely sequenced.
In addition all of the coding sequences of the pPP2d-4.0 gene which is
incomplete due to the insertion of an EcoRI linker in the original phage
construction have been obtained. Partial DNA sequencing (mainly flanking
sequences, see below) has been performed on the pRR4a-2.2 and pRRlOa-4.4

H2b genes.

(d) Chicken H2b coding sequences

The H2b coding regions were identified based on sequence comparisons
with the two previously characterized genes in pKRla-1.3 and pRR2e-3.5
(Grandy et al., 1982; Grandy et al., 1983). Figure 4 presents a compil-
ation of nine (eight complete plus that available for the pPP2d-4.0 gene)
H2b gene coding sequences. The H2b exon in pKRla-l.3 was chosen as the
prototype against which the others were compared.

An analysis of these sequences reveals that all of the chicken H2b
genes that have been completely sequenced are uninterrupted and code for
an H2b protein composed of 125 amino acids. The coding region of each
H2b gene is 375 nucleotides long. These exons tend to be rich in G and
C residues which can account for 58%(pBBA-3.0) to 64%(pRR3c-3.5) of
each coding region. This GC bias is reflected in the number of CpG di-
nucleotides which can number as many as 40 in certain H2b genes, whereas
CpG is generally relatively rare in eucaryotic DNA, although less so
within coding sequences (Salser, 1978).

Each of the nine H2b coding regions is unique, but they are all found
to share extensive sequence homology to the pKRla-1.3 H2b exon ranging
from 92% to 99.7%. Most of the differences in sequence are silent, falling
in the third or, occasionally, the first or second codon position. The

distribution of these nucleotide substitutions with respect to the pKRla-1.3

96

Figure 4. Chicken H2b exon sequence comparison. The gene on pKRla-1.3
was used as the prototype against which the other sequences were compared.
The amino acid substitutions are indicated where they occur. Nine unique

H2b genes are represented. The gene on pPP2d-4.0 contains the sequence 3'
of residue 106.

pKRla-l.
pRRZe-3.
pRRBc-S.
pRREc-3.
pPPZd-Z.
pBBA- 3.

OWUIJ-hlflw

pKRla-l.
pHRSa-3.
pBBA- 3.

a n -
“—13 - N
Hasn‘t Jo

JSOWKJ

pKRla-l.

pBBA- 3.

ow

pKRla-l.
pPPZd-Z.
pBBA- 3.

OWN

pKRla-l.3
pBBA- 3.0

pKRla-l.3

pKRla-l.
pRRZe-B.
pRR3c-5.
pRR3c-3.
pPPZd-Z.
pPPZd-4.

cum-hunt.)

pKRla-l.
pRR2e-3.
pRR3c-S.
pRR3c-3.
pPPZd-Z.
pPPZd-4.
pHRSa-3.
pBBA- 3.

owouma-tnw

pKRla-l.
pRRZe-3.
pRR3c-5.
pPPZd-d.

(3:th

O

pBBA- 3.

Figure 4.

UI

ATG

15
Lys
AAG

30
-y-

AAG

4S
Leu

CTG

60
as
91']

GGC

Ser
75

Gly
GGC

90
Thr
AC6

105
Glu

GAG

120
Lys
AAG

(‘1 '1
Donn—i O

16
Lys

AAG

J -
' I
4.; 3'5

AAG

Arg
46

Lys
AAG

61
Ile
ATC

76
Glu
GAG

91
Ser

TCG

106
Leu

CTG

121
Tyr
TAC

a)
(-

Glu
GAG

Se:
AGC
GCG
Ala
47

Gln
CA6

62
Me:
ATG

77
Ala
GCG

92
Arg
C66

107
Ala

GCC

122
Thr

ACC

Ile

3
ro

CCG

108
Lys
AAG

123
Ser

4
Ala

GCC

I: —4—+4

94
Ile
ATC

109
His

CAC

124
Ser

97

5
Lys
AAG

20
Lys
AAG

I) r4 (n

>05 (1

CD (.) L.)

50
Pro

CCC

65

- Phe

TTC

80

_ Leu

CTG

95
Gln
CAG

110
Ala

GCG

125
Ly s

AGC TCC AAG

6
Ser

TCC

21
Thr

ACC

(T) (1) O\

r111

13(1) (.1

31
Asp

GAC

66
Val
GTC

81
Ala
GCG

96
Thr
ACA

mmmmm

111
Val

GTC

3T

Tyr
TAC

52
Thr

ACG

67
Asn

AAC

82
His
CAC

97
Ala
GCC

112
Set

TCC
G

Pro

CCC

TCG

53
Gly
GGC

68
Asp

GAC

83
Tyr
TAC

98
Val

GTG

113
Glu

GAG

Ala
GCC

114
Gly
GGT

non (WHO

85

. Lys

AAG

100
Leu

CTG

115
Thr

ACC

ll
Lys
AAG

116
Lys
AAG

12
Lys
AAG

117
Ala

GCG

13
Gly
GGC

118
Val

GTC

14
Ser

TCC

cw —4—+4

29

Arg

~ CGC
‘ A A

44

GTE

S9
Met

ATG

74
Ala

GCC

89
Ile

ATC

104
Gly
GGC

119
Thr

ACC

98

.uauoo wmwcmno wafiwoo vfium ocfiam mamas mumu«wcfi mxmflumumm any .m.~lm~mxa cu uomammu nu“? mmcww
accumas nmm cmxofiso uswwm msu a“ vm>ummno mmwcmno ammo mwauomausa mo coauanfiuumfin .m muswwm

coxm nmm ca mmswfimmu wwwuomaosc mo acmscwfimmm HmowumEDZ

mum com ocm omm oom owm com oqm CNN com owH Goa oq~ ONH cod cm 00 De om
U<H CH<

__=._._ L . _ .L . . Er . .55: _= -

 

 

 

 

 

 

o
\Dlnx'rMNv—a

saSueqo
aseq go Jaqmnu

99
sequence is depicted in Figure 5. This histogram illustrates that these
mutations are clustered and define three regions of the H2b exon. The
greatest variability is located between the initiation codon and nucleo-
tide residue 110. In spite of the large number of alterations in this
region, only two amino acid substitutions result from these changes.
The exon in pBBA-3.0 codes for AGA at amino acid residue 31, which is an
arginine codon. pKRla-1.3 codes for lysine at this position with the
codon AAC. pBBA-3.0 also differs from the other H2b genes at amino acid
residue 32 where,instead of the codon ACC, for serine, it specifies GCG,
which corresponds to alanine. This amino acid substitution replaces a
polar residue with a nonpolar one which would result in a more hydropho-
bic protein molecule.

The second cluster of sequence differences is located between nucleo-
tides 145-186. The codon GGC at residue 60 specifies glycine but in
pBBa-3.0 the codon is AGC, serine, representing a conservative amino acid
change. Therefore, the protein encoded by pBBA-3.0 exhibits three amino
acid coding differences with respect to pKRla-l.3.

The third cluster of mutations is localized at the 3' end of the
H2b exon extending from nucleotide residue 288 through the termination
codon. One of these changes, in pPP2d-A.O, results in a coding change
at amino acid residue 122. The triplet ACC for polar threonine becomes
ATC in pPPZd-4.0. ATC encodes isoleucine, a nonpolar amino acid rendering
the predicted 2d H2b protein more hydrophobic. Recall that the H2b gene
present on the subclone designated pPPZd-4.0 is only partially represented.
The nucleotides 5' of residue 106 are missing and the missing 5' portion

of this gene's exon has not yet been identified.

100

Table I presents a summary of the differences at both the nucleotide
and amino acid sequence levels which distinguish each of the H2b genes
from pKRla-l.3. From this comparison, it is seen that the major class
of H2b protein is encoded by seven different genes as typified by the
pKRla-1.3 sequence. Each of the two remaining H2b variants isolated to
date is encoded by a single gene.

When the nine H2b exons are compared, a certain degree of biased
codon usage is observed. For example, of the six possible leucine codons
only the triplet CTG is used by chicken H2b genes. Other notable codon
biases are observed for arginine, asparagine, glutamine, glycine, histi-
dine, isoleucine and tyrosine. A termination codon bias is also evident
with a two to one to zero preferance found for TAG (amber) to TAA (ochre)

to TGA (opal), respectively.

(e) 5' sequences flanking the H2b histone genes

When the 5' flanking sequences of the eight completely sequenced
chicken H2b genes are compared, each one is seen to be unique. However,
within this variable background a number of blocks of sequence homology
can be identified. These are listed in Figure 6. Many of these conserved
sequences are shared by other genes transcribed by RNA polymerase II.
One such canonical sequence is the CCAAT pentamer or "cat box." At least
one copy of this sequence is associated with every chicken H2b gene ana-
lyzed and in five cases two or three copies are found. These sequences
reside at a distance of 124-198 nucleotides prior to the ATG start codon.
The distance separating multiple CCAAT pentamers does not appear to be
rigidly conserved but varies from between 22 to 44 bp. In every chicken
H2b gene examined, downstream of the 3'-most cat box is another component

of the RNA polymerase 11 promoter. This AT-rich region, which conforms to

101

Table I. Comparison of the H2b histone exons and the proteins

they predict with respect to pKRla-l.3.

 

Number of NT 2 Protein coding differences
differences sequence
compared to la divergence # residue substitution _
ClassI
pKRla-l.3 - - - - -
pRRZe-3.5 8 2.1 - - -
pRR3c-3.S 8 2.1 - - -
pRR3c-5.4 6 1.6 — - —
pPPZd-2.3 5 1.3 - - -
pHRSa-3.3 2 0.5 - - -
pMM—SA. 1 m3 - - -
Class II
pBBA- 3.0 29 7.7 3 31 Lys * Arg
32 Ser + Ala
60 Gly + Ser
Class III

pPPZd-A.O 3 0.8 1 122 Thr + Ile

 

102

Figure 6. DNA sequence analysis of the 5' flanking regions of eight
chicken H2b histone genes. The consensus sequences are discussed in the
text. Abbreviations are: Y=pyrimidine; R=purine; N=any nucleotide.

When two nucleotides occur at the the same position in the sequence

the most frequent one is listed above the line. When three nucleotides
occur at the same position the most frequently used base is written in
line with the rest of the sequence with the next most frequently used
residue above the line.

 

103

.m

.o ouswwm

hzmzmow
xom m<u xom <2H u_m_umam-mm: mmxom b<<uu
an A H an puma u u an an
op< mm Om zmzpzz :vm 5N az<<<p<pmmmm new z<~<uopmp¢ pm cw H<<uu as am p<<uu-a;m~ _<<uu
"cvcosvom wsmzcmcou
op<-mnsv-phopoou-ansN-<H<<<H<p»uuuuu<<a<uoppp<-anm~-h<<uu o.m .ame
«Na-
o»<-dn~m-<upho<u-anem-<<<<<b<puuuo<u<<p<uwhhp<-can~-h<<uu-mnmmuh<<uu-an~m-»<<uu v.m u<mma
ems-
up<-an~m-<UHpo<u-mneN-<<<<<P<huuuo<u<<~<uo-»<-ana~-H<<uu-mnmmup<<uu-anm~-H<<uu M.M-mmzza
ems-
op<-a:oe-HUHh»ou-aa~m-<o<<<<p<huouo<u<p<uo~up<-anvm-p<<uu-mnmc-p<<uu-an--»<<uu ..m-cmaai
. was-
op<-in_m-ouoeu<p-m2_N-oo<<<~<puouuouo<p<uobp_u-ansm-»<<uu c.m-cmmaa
«Ma-
op<-mcmm-<ohe<uu-anmN-<m<<<b<_uoooc<o<~<uobbp<-gscm-Ȣ<uu v.6-tmzea
on"-
ob<-a;mm-<u~p<uu-;;mm-<H<<<H<~uoooo<c<+<uobsh<-626m-H<<uu-nn¢V-H<<uu r.»-cmaea
was-
oh<-;;:m-ouueu<u-;;--<_<<<_<PuuuuoFu<+<uosha<-anem-~<<uu-nnvv-p<<uu m._-a_zxa

mc_- ,
.m

104
the consensus TATA or Goldberg box (1979) is of the form TATAAA in
chicken H2b genes. It is consistently located 39-43 hp downstream from
a cat box and between 61-83 bp upstream of ATC. Positioned between the
CCAAT and TATA homologies is a seven bp sequence which is shared by all
chicken H2b genes but not, generally,by other histone genes (Grandy et a1,
1983; Harvey et al., 1982). This sequence: % T § TGCATA, is invariably
located 6-7 bp 5' of TATA and 24-27 hp 3' of a CCAAT sequence. The site
where transcription initiates, referred to as the "cap" site, is usually
located 25-30 bp downstream of TATA in the histone genes which have been
examined (Hentschel and Birnstiel, 1981). Attempts to identify a specific
block of sequence, or cap box, associated with this site have not been
entirely successful because no real homology is evident. The exact place-
ment of each cap site will require mapping of the transcript (see below)
and then comparison of this information with the sequence data for each

gene. Putative cap boxes have been indicated based on their distance

from TATA in Figure 6.

(f) 81 mapping of the pKRla-l.3 transcript

We have begun S mapping experiments of individual transcripts in order

1
to address the questions of whether these genes are expressed, when and
where they are expressed and where their transcripts are initiated. The
H2b gene on pKRla-1.3 was chosen to be the first gene in our collection
to be analyzed for several reasons. One of the most important of these
was that it possesses a poly(A) signal sequence (see below). We were
therefore interested to know if it was expressed, and, if so, whether or
not it was polyadenylated.

A DNA probe was prepared by digesting the pKRla-1.3 plasmid sequen—

tially with EcoRI and BstEII. The 0.75 kb fragment of interest was then

105
gel purified, its ends were phosphatased and kinase-labeled with 32p.
The BstEII site lies 59 bp inside the coding sequence, and its protru-
ding 5' end allows labeling of the DNA stand which is complimentary to
the mRNA. The probe was precipitated with either total adult liver, total
breast muscle,tota1 chick embryo fibroblast, 4.5 day chick embryo, red blood
cell cytoplasmic, total HD3 or poly A+ HD3 mRNAs. The HD3 cell line is
a chicken erythroblast line transformed by a temperature sensitive mutant
of avian erythroblastosis virus which can be induced to differentiate
at the nonpermisssive temperature (42°) (Beug et al., 1982). The results
of this experiment are illustrated in Figure 7. A strong signal is seen
at a position in the gel which corresponds to 59 b. This is the distance
from the labeled BstEII site to the start codon and represents exon se-
quences protected by 82b transcripts which are identical, or nearly
identical, in their coding sequence to the pKRla-1.3 probe but differ
extensively from the probe in their 5' flanking sequence. The signal seen
at the position in the gel which corresponds to a 95 b fragment indicates
the size of the probe specifically protected by the mRNA transcribed
from the pKRla-1.3 gene (or any with identical 5' flanking regions, see
below). This therefore defines where the cap is located, i.e. 95 bp
upstream from the BstEII site. This site falls at the A residue in the
predicted cap box for this gene: CACTCCG. The H2b histone encoded by the
pKRla-l.3 exon appears to be expressed in all of the RNA preparations
examined. However, the strongest signal is seen in the total 4.5 day

embryo as expected (Sugarman et al., 1983).

(g) Chicken 82b histone gene 3' flanking sequences
A comparison of the flanking sequences 3' of the H2b genes reveals

that each is unique, yet they all have in common some extensive stretches

106

Figure 7. S mapping analysis of the pKRla—1.3 transcript. The 0.75
kb probe was fiybridized with various mRNAs. The RNAs used were: red blood
cell cytoplasmic(507); total chick embryo fibrblast(50y); HD3 poly A+
(1.0y); total HD3(20y); total adult liver(50y); total 4.5-day chick embryo
(20y) and total adult breast muscle(50y). The protected band seen at

59 bp corresponds to the size of a fragment extending from the labeled
BstEII site to the ATG start codon. The fragment seen at 95 bp corresponds
to a fragment protected by a transcript identical to the pKRla-l.3 probe

in its 5' leader sequence. A faint band is visible in the original at
95 bp in the red cell lane.

107

 

 

 

Figure 7.

108
of sequence. A comparison of each of the nine H2b genes whose 3' sequences
have been determined is found in Figure 8. The 3' data for pRR2e-3.5
was obtained from Harvey et al. (1982). The most striking sequence homol-
ogy that is shared by all nine genes is a 27 bp element which is located
17-33 nucleotides from the termination codon. Contained within this
large sequence is an inverted repeat which is interrupted by a stretch
of four thymine residues followed by a C and then a purine. This sequence,
GGCTCTTTTCQGAGCC, is highly conserved in the 3' regions of almost all

histone genes examined (Hentschel and Birnstiel, 1981) and is referred to

as the hyphenated symmetrical dyad. Another sequence element shared by

‘3‘ EGAAEGA
eight of the nine H2b genes is a decamer of the form: CA T T - - - A — -
5 A A G G C T G

located invariably 11—12 hp 3' of the hyphenated symmetrical dyad. The
H2b gene in pKRla-1.3 lacks this sequence. In its place is found the
hexanucleotide AATAAA, which is the signal sequence recognized as being

involved in mRNA polyadenylation (Proudfoot and Brownlee, 1976).

(h) Identification of nonallelic, sibling H2b histone genes
A very intriguing discovery made as a result of analyzing all of

the H2b nucleotide sequence data is that several of these sequences are
more closely related to each other than to the rest of the collection.
The H2b gene contained on subclone pRR2e-3.5 is nearly identical to the
gene found on subclone pRR3c-5.4, not only in the coding region, but also
in flanking regions. However, when the restriction maps of the phage from
which these clones were obtained are examined, it is apparent that they
each represent nonallelic loci. An alignment of these phage is shown in
Figure 9A which shows that the homology ends approximately 1 kb 5' of the
linked H2a genes. A second set of sibling H2b genes is defined by the

pHRSa-3.3 and pBRA-5.4 subclones. These two subclones share extensive

109

Figure 8. DNA sequence comparison of the 3' flanking regions of nine
chicken H2b histone genes. The hyphenated symmetrical dyad is marked

by the horizontal arrows. The chicken H2b consensus sequence is shown
along with the consensus sequence derived by Hentschel and Birnstiel
(1981) for histone genes in general shown for comparison. When two
nucleotides occur at the same position in the sequence the most frequently
used base is above the line. When three nucleotides can occur at a given
position in the sequence the most frequently used base is placed in line
with the rest of the sequence and the next most frequently used base
written above this one.

pKRla-l.

pRRZe-3.

pRR3c-3.

pPPZd-Z.

Chicken

110

3 SlAG-Z3bP-ACCCAAAGECTCTTTTCGGAGCCACC-9:3-AATAAA3'
3 TAA-3Eb;-ACCCAAAGGCTCTTTTCAGAGCCACC-8:;-CA£AEAAAGA
4 TAA-33b;-ACCCAAAGGCTCTTTTCAGAGCCACC-abs-CAGAGAAACA
S TAG-23bp-ACCCAAAGGCTCTTTTCAGAGCCACC-8bp-CAATGAAAGA
3 TAG-23b?-ACCCAAAGGCTCTTTTCGGAGCCACC-Bbp-CAGTAAAAGA
0 TAG-23bp-ACCCAAAGGCTCTTTTCAGAGCCATT-9bp-CATGGAAGGG
TAG-23b;-ACCCAAAGGCTCTTTTCAGAGCCACC-pr-CAATGAAAGA
TAG-23bp-ACCCAAAGGCTCTTTTCAGAGCCACC-Bbp-CAATGAAAGA
TAA-l7bp-ACCCAAACGGCTCTTTTCGGAGCCCAA-Bbp-CAAAGGGCTG
H2b Consensus Sequence: A A G A G
TA7-17-33—ACCCAAAGGCTCTTTTC76AG C7 é é-B—9-CAA 4 7 7 7 A 9 7
A bp --9 (5€--C 7 4 bp 4 6 A 6.6 é T G

Consensus Sequence for Histone Genes in General:

Figure 8.

T A A
AACGGCéCTTTTCéG/GCCACC-6-9-CAAGAAAGA
—-) G<—— bp

111

Figure 9. Alignment of the phage maps which contain nonallelic,
sibling H2b histone genes. The solid black box in A Chl and 4a indi-
cate H2b hybridizing regions in which the genes' transcriptional orien—
tation is not known. (A) 1Ch5e overlaps with 2e. The H2b genes which
are similar in ACh2e and 3c are aligned to indicate the regions of
identity and nonidentity. (B) The two H2b:H2a gene pairs which share ex-
tensive sequence homology are aligned to indicate the regions of identi-
ty and nonidentity. (C) The H2b gene in XChla has been completely sequenced.
The genes contained on ACth and 4a have only been partially characterized
and the H2b gene in lChl has not been subcloned but is included with this
set of sibling genes based on blotting and restriction mapping data (not
shown).

112

 

 

‘11 311A“ \l/ \l/ ll XChse

 

A H1 H2. szT
JL :7 SC 0 \L73J/Hvl'L A Chze
T H1H28 H2b

 

T H2b H20 H2. H2!)

 

{L . ..o \L oxl J1 kan
H2b H2: H2!) T

11 #9 9 L S \‘7 all x cns.

H2. H2b [1‘ If

= t——-« 2K8
LEE. 9 a Xenia

 

 

 

H3 H3 H2b H1
1. 1 s i i is...
H2b T
H \l/ \L Jr J: 11' ACM
H26 ’1

 

H1 9 g ml ll Acnn

==H23
—: H4

U=Eco RI;\L=Eco 8157=H|N0m36= KPNI;T= BAMHI; - = Direction of
linker Transcription

Figure 9.

113

sequence homology, but a comparison of their respective parental phage
maps, in Figure 9B, reveals that they are indeed distinct loci. In this
particular pair the 3' extent of the homology can be located approximately
2.0 kb downstream of the H2b genes but the 5' end can not. A third set

of related genes are defined by the sequences in subclones pKRla-1.3,
pRR10-4.4, pRRAa-2.2 and probably the 2.0 kb H2b-hybridizing region of

A Chl. Here again it is clear that the phage maps, aligned in Figure 9C,
are not coincident. Although the sequence data is incomplete for this
last set of genes, the information that is available strongly suggests
that these clones all contain essentially the same H2b gene (identical

in both coding and flanking regions) in four genetically different regions.

DISCUSSION

The H2b genes isolated and characterized in this study were identi-
fied by their hybridization with a 0.3 kb BstEII restriction fragment
prepared from pKRla-1.3. This gene was originally identified by its as—
sociation with an H3 sequence which hybridized to sea urchin embryonic
histone gene probes.

Based on the genomic Southern blotting data the chicken H2b gene
family is estimated to contain between 14 and 16 members. This estimate
assumes that each band seen in the autoradiogram in Figure 1 corresponds
to one, two or three H2b gene equivalents and not to fragments containing
partial gene sequences. This seems a safe assumption based on the fact
that none of the eleven of the H2b genes we have isolated has been found
to contain a natural EcoRI or HindIII site within its coding region.
Therefore, the least intense bands at 1.6 kb, 3.5 kb and 4.0 kb probably

represent a single 82b gene equivalent each. The stronger signals seen

114

at 2.0 kb, 5.0 kb, 7.2 kb, 8.0 kb and 10 kb may contain two gene equiva-
lents either on a single fragment or on two fragments which comigrated

in this experiment. The most intense band, seen at 7.8 kb in Figure 1,
may represent three H2b gene equivalents contained on one fragment or on
three comigrating fragments or on two comigrating fragments one of which
contains two genes. Unfortunately, most of the isolated H2b genes are
not flanked by both natural EcoRI and HindIII sites in the cloned genomic
DNA, so it is difficult to substantiate this assessment at this time.

One of the bands, seen at 2.0 kb in the EcoRI/HindIII genomic Southerns,
corresponds to the fragment size predicted by the l Ch3c map. Several
of the bands observed in other Southern blotting experiments, where the
DNA was digested with both EcoRI and BamHI (data not shown) also corre-
spond to fragments of the appropriate size as predicted by the phage maps.
These gene accounting experiments leave six genes in our chicken H2b gene
collection unassigned to a particular band in either type of genomic
Southern blot. In order to identify the genomic fragments which corre-
spond to each of these six sequences it will be necessary to prepare
gene-specific probes from the unique flanking regions of the individual
genes or to identify overlapping phage or cosmid clones whose restriction
maps supply the information currently lacking. Both approaches are being
pursued at this time.

The gene corresponding to the sequence present in the 1.6 kb band in
Figure 18 is probably not represented in any of the clones isolated be-
cause our library screens were carried out at high stringency. The fact
that this fragment only cross hybridizes with the 0.3 kb BstEII H2b probe
under low stringency conditions is indicative of a lack of sequence homol-

ogy between the two. The sequence divergence that is reflected in this

115

hybridization behavior suggests that the 1.6 kb fragment may contain
sequences coding for a variant H2b protein, perhaps the spermatocyte-
specific H2b. It is therefore of interest to obtain a genomic clone of
this sequence since it may exhibit unique regulatory elements. We are
currently analyzing the phage in our collection under less stringent hy-
bridization and washing conditions in an attempt to determine whether or
not this sequence may fortuitously be linked to other cloned H2b sequences.
If necessary the phage library described above or a newly constructed
cosmid library will be screened in an attempt to obtain this sequence.

We are able to account for thirteen of the estimated 14-16 H2b genes
in the chicken genome. It is possible that the remaining 1-3 genes which
our collection lacks were missed in the original screen as a result of
their being under-represented in the library. We are screening a chicken
genomic cosmid library in order to identify the remaining H2b sequences.

The genomic organization of the chicken H2b gene family is complex
and no common theme has emerged. The clustering of several different his-
tone genes without tandem duplication is an organizational theme first
described in chicken histone families (Engel & Dodgson, 1981; Harvey
et al., 1981) and shared by those of higher vertebrates that have been
examined, in particular, mice (Sittman et al., 1981) and humans (Sierra
et al., 1982). This arrangement is very different from what had been
previously described for lower vertebrates (Moorman et al., 1980; van
Dongen et al., 1981) and a number of invertebrate species (Kedes, 1979),
with the exception of the more recently described late sea urchin histone
genes (Moxson et al., 1983; Roberts et al., 1984). The variety of histone
organizational themes exhibited throughout the eucaryotic phyla implies

that each s ecies' enome ma have evolved in a unique manner so as to
P 8 Y

116

supply the developing organism through adulthood, with an adequate and
appropriate complement of histone protein. This study presents the most
detailed description of a histone gene family of any vertebrate species
examined to date. From this analysis it can be concluded that at least
six H2b genes occur as doublets as seen on A ChA, 2d and 3c with three

of these genes closely linked to divergently transcribed H23 genes(Figure
2)- The remaining six phage in the collection contain only one H2b gene
each and one of these, A Ch2e, is closely linked to a divergently trans-
cribed H2a gene. Therefore, if any common organizational scheme can be
said to exist with respect to the H2b gene family, it appears that these
genes occur with other histone genes in clusters, each of which is unique,
and that the H2b genes are most often associated with H2a genes. The tight
linkage between these two genes may be evolutionarily maintained so that
they can be coordinately regulated. A similar organization is found in

S, cerevisiae ( Hereford et al., 1979). Why, then, only one third of the

 

H2b genes in chicken are arranged in this fashion is unclear.

A sequence analysis of the nine most completely characterized H2b
genes reveals that, although each is uniquely distinguishable at the nuc-
leotide level, their overall coding sequences are highly conserved. The
chicken H2b genes are unusual among eucaryotic genes in that their
exons are composed of 58-64% G:C base pairs, with many of these
existing as CpG dinucleotides. This pair of bases is rare in eucaryotic
DNA because it tends to be methylated at C-5 and subsequent deamination
will yield a thymine residue which is often not repaired before replica-
tion (Salser, 1978). Such transition mutation events should, over time,
result in the shift from CpG's to TpG's. The maintenance of CpG dinucleo—

tides, especially in noncoding positions, suggests that they are

117

selectively maintained, perhaps as a result of some unknown selective
pressure at silent base pairs or due to undermethylation in the germ cell
DNA. Such undermethylation could reflect a heightened transcriptional
activity of H2b genes in these cells. Often a hallmark of histone
genes is their biased use of amino acid codons, and the chicken H2b genes
are no exception. Furthermore, a preferance for certain termination co-
dons is demonstrated by the use of TAG twice as frequently as TAA, where-
as the triplet TGA is not used in any of the genes sequenced. The advan-
tage of maintaining codon biases is unclear but may reflect an individual
species' preferences due to its tRNA complement.

The three protein variants predicted by the nine H2b genes have
not been previously identified in chickens. In spite of this, we feel
that each is likely to be expressed since they all resemble typical eu-
caryotic transcription units recognized by RNA polymerase II and there
are no internal nonsense codons that often exist in unexpressed pseudo-
genes (see below). Urban and Zweidler (1983) have identified two classes
of chicken H2b proteins using their gel electrophoretic system and they
have labeled these variants H2b.1 and H2b.2. The amino acid sequences
of these two proteins are not known. The H2b.1 protein is the predominant
species and is expressed in several tissues throughout the life of the
organism. The H2b.2 protein is a minor variant expressed only in adult
liver and spleen. A third variant may also be produced in adult testes.
van Helden et a1. (1982) have determined the complete amino acid sequence
of the major histone H2b protein present in the erythrocytes of adult
chickens( which may be equivalent to Urban and Zweidler's H2b.l dis-
cussed above). The sequence which van Helden et al. (1982) obtained cor-

responds to the protein predicted by the gene contained on pKRla—1.3 and

118
six other H2b genes in 124 of 125 residues. The pKRla—1.3 gene codes
for isoleucine at residue 61 whereas serine is found in the erythrocyte
derived protein. This difference could be the result of genetic varia-
tion between the two chickens used, or it may be that the gene coding
for the serine variant is one we have not yet isolated. Alternatively,
it is possible that the protein sequence is in error.

The two variants predicted by clones pBBA-3.0 and pPP2d—4.0 in our
collection differ from the pKRla-1.3 class (class I) of protein as a re-
sult of several amino acid changes. The variant encoded by pBBA-3.0
(class II), which displays three amino acid differences with respect to
class I H2b proteins, becomes less polar as a result of substituting alanine
for serine at amino acid residue 32. The H2b variant, partially encoded
by pPP2d-4.0 (class III), would contain an isoleucine residue at position
122 instead of a polar threonine, so this variant would also be more hydro-
phobic. It should be recalled that this clone encodes only the carboxy
two thirds of an H2b protein. Each of these variants should be identifi-
able in Zweidler's electrophoretic system. It is unclear which band in
Urban and Zweidler's experiments(1983) corresponds to the protein sequenced
by van Helden et a1. (1982). Therefore, it is not possible at this time
to assign any of our genes to a particular gene product. The fact that
only two H2b proteins, and perhaps a third, are seen in Urban and Zweidler's
experiment may indicate that the others are rare or that two or more species
comigrate in their gel system. Until further evidence is obtained, we
would guess that the class I H2b sequence we have determined corresponds
to Urban and Zweidler's more common H2b.l variant and one or both of the

other two different classes corresponds to H2b.2.

119

The analysis of the nucleotide sequences which flank the H2b coding
regions has been very informative. Although each of the eight fully se-
quenced genes is unique in both its 5' and 3' flanking sequences, when
properly aligned, several blocks of homology shared by all of the chicken
H2b genes can be identified. The 5' flanking DNA sequence is the region
where the promoter elements recognized by RNA polymerase II are located.
These signal sequences include the CCAAT pentamer, the TATA or Goldberg
box and a cap site. The H2b genes of chicken possess one, two or three
CCAAT boxes which are located upstream of the exon between postions -198 to
-124 (where the A of ATC is numbered as one). The significance of multiple
CCAAT boxes associated with five of the chicken H2b genes is unclear but
the relative density of these elements may be an important factor in
the control of H2b gene expression. The CCAAT box multiplicity observed
in chicken H2b genes does not appear to be correlated with its linkage
to a divergently transcribed H2a gene.The spacing of the multiple cat boxes
relative to one another appears to be flexible. However, with regard to
the TATA box a distance of 39-42 base pairs invariably separate these two
elements. The TATA box is a consensus sequence associated with other
genes transcribed by RNA polymerase II and is thought to play a role in
the selection of the transcription initiation site, approximately 30 base
pairs downstream (McKnight & Kingsbury, 1982). In chicken H2b genes, the
sequence TATAAA is found 56 to 92 base pairs upstream of the ATG initia-
tion codon. Each of the TATA sequences is preceded by a GC-rich tetramer
followed by a pyrimidine. The importance of this sequence is unknown.

In addition to the promoter elements commonly shared with other genes
transcribed by RNA polymerase II, the chicken H2b genes display a conserved

seven base pair sequence invariably positioned six or seven bp 5' of TATA

120

and 24 to 27 bp 3' of a CCAAT sequence. A survey of the sequence data
available for other H2b genes indicates that this heptamer is shared by
yeast, sea urchin and Xenopus species (Grandy et al., 1983; Harvey et al.,
1982). The function of this H2b gene-specific element is unknown, but

its location relative to recognized promoter elements is intriguing and
suggests a possible regulatory role.

The DNA sequence which follows the nine different types of H2b genes
is also unique. However, consensus sequences are revealed when the genes
are properly aligned. The most striking block of homology to be uncovered
is the 22 bp sequence, located 17-37 bp downstream of the termination codon,
which contains the hyphenated symmetrical dyad. The symmetrical dyad
is not only conserved in chicken H2b genes but is also found to be a com-
ponent of most 3' histone gene-flanking regions (Hentschel and Birnstiel,
1981). In addition, eight to ten base pairs downstream of the symmetrical
dyad is another consensus sequence shared by other histone genes.In eight of
the nine H2b genes a purine-rich sequence is present, located 11-12bp
3' of the hyphenated dyad. Both the hyphenated palindrome and the purine-
rich sequence which follows it have been analyzed by deletion and point
mutagenesis ( Birchmeier et al., 1983). Such experiments, performed on
sea urchin H2a genes, have revealed that the symmetrical dyad forms a
secondary structure in the mRNA molecule, and if this stem-loop is absent,
no 3' processing will occur. Furthermore, it has been demonstrated that
the purine-rich sequence approximately 10 bp downstream from the inverted
repeat is very important for accurate and efficient processing of the
primary transcript. Without this sequence, no mRNAs with authentic 3'
ends are produced. When this sequence and the hyphenated dyad are present

alone, low levels of accurately processed mRNA transcripts are observed,

121
and as more of the 3' flanking sequence (up to 80 bp) is left intact, the
wild type levels of processing can be obtained. The extensive 3' blocks
of homology shared by sea urchin H2a genes and chicken H2b histone genes
suggests that the same sort of primary transcript processing mechanism may
be employed in both species.

One interesting exception to the conserved H2b 3' structural organiza-
tion is found in pKRla-1.3. Instead of the conserved purine-rich sequence
3' of the palindrome,the consensus signal sequence AATAAA for poly A addi-
tion, is found. The S1 mapping analysis of this gene indicates that its
transcript is indeed polyadenylated and that it is expressed in adult tis-
sues but not at the same high levels as seen in embryos. This data also
suggests that at least one other H2b mRNA species is polyadenylated based
on the strong signal seen at 59 bp when the HD3 poly A+ mRNA population is
analyzed (Figure 7). This is further evidence that our current collection
is lacking some of the H2b histone gene family members.

The discovery that nearly identical copies of some of the chicken H2b
genes, and their flanking sequences, are found repeated two or more times
throughout the genome is interesting. At the present level of resolution
it is not possible to speculate as to the linkage of these sibling genes
except to say that they do not appear to be closely linked. It seems like-
ly that the H2b genes with identical flanking sequences are more closely
related evolutionarily than those whose flanking sequences have extensively
diverged. Note that if the region of identity were to include an EcoRI or
Hind III site on both sides of a gene, this would account for the bands
seen in Figure 1 that appear to contain two or more genes.

The finding of several nearly identical H2b sequences which, however,

clearly represent different loci may reflect the mechanisms and / or

122

the timing by which the H2b gene family has evolved. One correction
mechanism which may be responsible for the homogeneity observed among
different H2b loci is gene conversion. If gene conversion is the
mechanism responsible for maintaining sequence homology among different
loci,it is unclear why in,some cases,extensive flanking sequences have
been converted whereas many very similar H2b coding regions have little
similarity in their flanking sequences. It is clear, however, that based
on the extensive sequence homology shared by all nine of the H2b gene
types characterized, they must have evolved from a common ancestral se-
quence. After the initial duplication events, recombination, gene con-
version and mutation have presumably generated the extensive diversity
within and surrounding the chicken H2b histone genes observed in this

study.

123

REFERENCES
Benton, W.D. & Davis, R.W. (1977). Science, 196, 180-182.

Beug, H., Doederlein, G., Freudenstein, C. & Graf, T. (1982). J. Cell.
Physiol. Supp. 1, 195—207.

Birchmeier, C., Folk, W. & Birnstiel, M.L. (1983). Cell, 35, 433-440.
Birnboim, H.C. & Doly, J. (1979). Nucl. Acids Res. 1, 1513-1523.

Borun, T.W., Ajiro, R., Zweidler, A., Dolby, T.W. & Stephens, R.B. (1977).
J. Biol. Chem. 252, 173-180.

Browne, D. (1984). unpublished results.
Childs, G., Maxson, R. & Kedes, L.H. (1979). Dev. Biol. 13, 153—173.

Crawford, R.J., Kreig, P., Harvey, R.P., Hewish, D.A. & Wells, J.R.E.
(1979). Nature(London), 279, 132—136.

Dodgson, J.B., Strommer, J. & Engel, J.D. (1979). Cell, 11, 879-887.
Edwards,, L.J. & Hnilica, L.S. (1968). Experimentia (Basel) 24, 228-229.

Engel, J.D. & Dodgson, J.B. (1981). Proc. Natl. Acad. Sci., U.S.A., 28,
2856-2860.

Engel, J.D., Sugarman, B.J. & Dodgson, J.B. (1982). Nature (London), 297,

Garger, S.J., Griffith, O.M. & Grill, L.K. (1983). Biochem. Biophys. Res.
Comm. 117, 835-842.

Girvitz, S.C., Bacchetti, 8., Rainbow, A.J. & Graham, F.L. (1980). Anal.
Biochem. 106, 492-496.

Goldberg, M. (1979). Ph.D. thesis, Stanford University.

Grandy, D.K., Engel, J.D. & Dodgson, J.B. (1982). J. Biol. Chem. 257,
8577-8580.

Grandy, D.K., Engel, J.D. & Dodgson, J.B. (1983) In Gene Expression
Hamer, D. & Rosenberg, M., eds.) pp.445-455, A.R. Liss, Inc., New
York.

 

Groppi, V.E. & Coffino, P. (1980). Cell, 21, 195-204.

Grunstein, M. & Hogness, D. (1975). Proc. Natl. Acad. Sci., U.S.A. 3961-
3965.

Harvey, R.P., Krieg, P.A., Robins, A.J., Coles, L.S. & Wells, J.R.E. (1981).
Nature (London), 294, 49-52.

124
Harvey, R.P., Robins, A.J. 6 Wells, J.R.E. (1982). Nucl. Acids Res., 19,
7851-7863.
Hentschel, C.C. 6 Birnstiel, M.L. (1981). Cell, 25, 310-313.

Hereford, L., Fahrner, K., Woolford, J.Jr., Rosbash, M. 6 Kaback, D.
(1979). Cell, ifin 1261—1271.

Kedes, L.H. (1979). Ann. Rev. Biochem. 48, 837-870.
Kornberg, R.D. (1977). Ann. Rev. Biochem. 46, 931-954.

Krieg, P.A., Robins, A.J., D'Andrea, R. 6 Wells, J.R.E. (1983). Nucl.

Maniatis, T., Fritsch, B.P. 6 Stambrook, J. (1982). In Molecular Cloning,
a Laboratory Manual. Cold Spring Harbor New York.

 

 

Maxam, A.M. 6 Gilbert, W. (1980). Methods Enzymol. 65, 499-560.

Maxson, R., Mohun, T., Gormezano, G., Childs, G. 6 Kedes, L.H. (1983).
Nature (London), 301, 120—125.

McGhee, J.D. 6 Felsenfeld, G. (1980). Ann. Rev. Biochem. 49, 1115-1156.

McKnight, S.L. 6 Kingsbury, R. (1982). Science, 17, 316-324.

Moorman, A.F.M., delaaf, R.T.M., Destree, C., Telford, J. 6 Birnstiel, M.
L. (1983). Gene 19. 185—193.

Nevins, J.R. (1983). Ann. Rev. Biochem., 52) 441-466.

Proudfoot, N. 6 Brownlee, G.G. (1976). Nature (London), 263, 211-214.

Risley, M.S. 6 Eckhardt, R.A. (1981). Dev. Biol. 4, 79-87.

Roberts, S.B., Weisser, R.B. 6 Childs, G. (1984). J. Mol. Biol. 174, 647-
662.

Ruderman, J.V. 6 Pardue, M.L. (1978). J. Biol. Chem. 253, 2018-2025.

Ruiz-Carrillo, A., Affolter, M. 6 Renaud, J. (1983). J. Mol. Biol. 170,
843-859.

Salser, W. (1978). Cold Spring Harbor Symp. Quant. Biol. 42, 985-1002.

Sierra, F.,Lichtler, A., Marashi, F., Rickles, R., Van Dyke, T., Clark,
8., Wells, J., Stein, G., 6 Stein, L. (1982). Proc. Natl. Acad. Sci.,
U.S.A., 12) 1795-1799.

Sittman, D.B., Chiu, I.M., Pan, C.J., Cohn, R.H., Kedes, L.H. 6 Marzluff,
W.F. (1981). Proc. Natl. Acad. Sci., U.S.A., Z8, 4078-4082.

125

Sittman, D.B., Graves, R.A. 6 Marzluff, W.F. (1983). Proc. Natl. Acad.
Sci., U.S.A., 88, 184 -1853.

Smith, D.R. 6 Calvo, J.M. (1980). Nucl. Acids Res. 8, 2255-2274.
Sollner—Webb, B. 6 Reeder, RyH. (1979). Cell, 18, 485-499.
Southern, E. (1975). J. Mol. Biol. 88, 503-518.

Sugarman, B.J., Dodgson, J.B. 6 Engel, J.D. (1983). J. Biol. Chem. 258,
9005-9016.

Urban, M.R. 6 Zweidler,A. (1983). Dev. Biol. 88, 421-428.

Van Dongen, W., DeLaaf, L., Zaal, R., Moorman, A. 6 Destree, O. (1981).
Nucl. Acids Res., 8, 2297-2311.

Van Helden, P.D., Strickland, W.N., Strickland, M. 6 von Holt, C. (1982).
Biochim. et Biophys. Acta., 703, 17-20.

Woudt, L.P., Pastink, A., Kempers-Veenstra, A.E., Jansen, A.F.M., Mager,
W.H. 6 Planta, R.J. (1983). Nucl. Acids Res., 11, 5347-5360.

Zweidler, A. (1984). In Histone Genes. (Stein, G.S., Stein, J.L. 6 Marzluff,
W.F., eds.) pp.339-371, J. Wiley and Sons, New York.

 

126

CONCLUSIONS
The objective of this research project was to characterize the
chicken H2b histone gene family. As a result of my efforts the following
conclusions have been reached:

1. The chicken H2b histone gene family contains an estimated 14-16
loci per haploid genome.

2. The nucleotide sequence of one of these genes has diverged consid-
erably from the others.

3. Ten unique chicken H2b-containing lambda genomic clones have been
isolated and each cloned DNA has been mapped with restriction
enzymes.

4. Eleven unique chicken H2b histone genes have been subcloned and
the DNA of nine of these genes has been completely sequenced.

5. Only a partial gene is present on pPP2d-4.0.

6. Three different chicken H2b histone proteins are predicted by
the nucleotide sequence data.

7. The greatest degree of nucleotide sequence flexibility occurs at
the 5' end of chicken H2b histone exons.

8. The noncoding sequence homologies 5' of H2b histone exons include
a CCAAT box (or boxes), TATA box, an H2b-specific element (% T
% TGCATA g), and a putative cap box.

9. The noncoding sequence homologies 3' of chicken H2b histone exons
include a 22 base pair element which contains the hyphenated
symmetrical dyad: GGCTCTTTTC g GAGCC.

10. Eight of nine sequences 3' of chicken H2b exons contain a

purine-rich element downstream of the hyphenated symmetrical dyad.

127

11. One of the nine sequences 3' of chicken H2b exons contains
the consensus signal sequence for polyadenylation: AATAAA, and
a polyadenylated transcript from this gene (pKRla-1.3) has been
identified.

12. Three sets of sibling H2b histone gene sequences have been

identified.