Note: Descriptions are shown in the official language in which they were submitted.
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
1
Modulation of Chromosome Function
by Chromatin Remodeling Aces nts
The present invention relates to a process for the
modulation of chromosome function using sequence-specific
chromatin remodeling agents. The invention also relates to
chromatin remodeling agents which specificially target
chromatin responsive elements in the genome, and to their
use in modulating endogenous and heterologous gene function
in a cell. The invention further relates to fluorescent
chromatin remodeling agents, and their use in probing
epigenetic status and location of DNA in nuclei, and their
use for cytological / structural determinations, including
quantitative estimations of specific DNA sequences in cells
and chromosomal material.
In eukaryotic cells, DNA is folded first in the chromatin
fiber and then into chromosomes by several hierarchical
levels of organisation. This organisation changes
dynamically around the cell cycle to facilitate different
chromosomal functions. These structural levels of
chromosomes are brought about by the evolutionarily
conserved histones, various non-histone proteins and
oligonucleotide protein complexes. The complex formed
between DNA and these components is referred to as
chromatin.
The first order of chromatin structure can be considered to
be the « beads on a string » structure created when the DNA
is wrapped around individual histones. Here the chromatin
has the appearance of spherical particles connected by thin
fibers. Each of these bead structures is known as a
SUBSTITUTE SHEET (RULE 26)
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
2
nucleosome, associated with approximately 200 base pairs of
DNA. The nucleosome « beads » or « core particles » comprise
highly protected DNA segments of 146 base pairs tightly
wrapped around two each of histones H2A, H2B, H3 and H4 (the
histone octamer). The stretch of 146 base pairs makes almost
two full turns around the disc-shaped histone octamer. The
remaining DNA provides the linker between adjacent
nucleosomes. A single molecule of histone H1 is associated
with each nucleosome, and serves to « seal » the two turns
of DNA around the histone octamer.
Higher order chromatin structure involves the further
assembly of the nucleosomes into filaments of 300 Angstroms,
where the chromatin is proposed to be compacted by winding
into a « solenoid »-type structure containing six
nucleosomes per turn. Most chromatin is present in this form
in interphase nuclei. Differences in the degree of folding,
i.e. in the chromatin structure, in different regions of the
chromosome play an important role in determining whether a
particular gene is active in a particular cell. Indeed,
chromatin status is one of the main upstream steps in gene
regulation since it determines whether or not DNA-binding
factors can gain access to the DNA. Chromatin state is thus
an important element in epigenetic control of gene
expression.
The chromatin fiber is thought to be partitioned into
transcriptionally active (competent) and inactive domains.
Experimentally these domains are revealed by their different
sensitivity to digestion by enzymes such as DNase I,
restriction enzymes or cleavage by topoisomerase II. Active
chromatin (also called open) domains are generally more
sensitive (more accessible) to digestion by these enzymes
than inactive (less accessible) ones. At the global
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
3
morphological level, it is possible to observe by light
microscopy in nuclei two extreme structural (epigenetic)
states of chromatin organisation. One is called
heterochromatin and the other euchromatin. Heterochromatin
generally reflects transcriptionally inactive (structurally
compact) chromatin and euchromatin is generally enriched in
transcriptionally active (competent), more open chromatin.
A growing number of activities have been described that
mediate changes in chromatin structure (chromatin
remodeling) that facilitate chromosome function by rendering
the chromatin fiber more accessible to nuclear factors. The
high molecular weight ATPase dependent activities include
complexes such as SWI/SNF, a highly conserved 2MDa multisub-
unit assembly. The modifications brought about by such
complexes include changes in the DNA conformation on the
histone, alteration of histone conformation, and changes in
histone / DNA interactions (Peterson C., and Workman J.,
Curr. Opinion in Genetics & Development, 2000, 10 . 1~7-
192 ) .
Chromatin remodeling not only gives rise to gene activation
but can also lead to gene silencing. This affects both
endogenous and heterologous DNA elements. The precise
mechanisms involved in chromatin-mediated gene-silencing are
as yet unclear, but may involve the facilitated binding of
silencing factors (including enzymes), or the spreading of
the heterochromatin-like states and/or chemical
modifications. If the inserted gene is juxtaposed within or
near such chromatin states, then gene silencing can occur.
Furthermore, integration of multiple copies of heterologous
genes can give rise to interactions between repeated
sequences which in turn trigger the formation of inactive
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
4
genetic states. Indeed, methylation induced by repeats can
lead to chromatin modification.
Cis-acting DNA elements involved in chromatin remodeling
have to date not been clearly identified, nor has their
mechanism of action been elucidated.
Candidate elements are « Scaffold Associated Regions
(SARs) which are very AT-rich fragments several hundred base
pairs in length composed of numerous clustered irregularly
spaced runs of As and Ts (called A .tracts). These DNA
sequences specifically associate with the nuclear scaffold
and possibly define the bases of chromatin loops. The
possible involvement of SARs in higher order chromatin
structure and gene expression has been suggested by a number
of authors, although direct proof has not been obtained and
the precise relationship between SAR function and nuclear
organisation remains to be elucidated.
Data obtained in Drosophila using a high-affinity high-
molecular weight SAR-binding protein called MATH20,
expressed specifically in the larval eye imaginal discs
(Girard et al., 1998). MATH20 was found to suppress the
position effect variegation (PEV) phenotype manifested by
the silencing of the white gene which, as a result of an
inversion, is juxtaposed close to the heterochromatin of the
X chromosome. On the basis of one current model, the authors
suggested that MATH20 may be binding to a giant
approximately 11 Mb reiterated SAR in the form of satellite
III repeats, thereby disrupting the cooperative interaction
of compacting proteins responsible for heterochromatin
formation and transmission into the juxtaposed euchromatic
region. According to this hypothesis, the binding of MATH20
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
would energetically disfavour the spreading of the
polymerizing proteins into the surrounding euchromatic
region, thus restoring the activity of the white gene.
An alternative, mutually non exclusive model has also been
proposed according to which SAR function is mediated by
sequences that readily unwind under torsional stress (Bode
et al., Science 1992 255, 195-197).
The nature of the cis-DNA elements involved in chromatin
remodeling, and their mechanism of action, has therefore not
been unambiguously identified to date. The use of chromatin
remodeling as a means of epigenetic control of gene
function, and more generally of chromosome function, has
therefore not been possible.
It is an object of the present invention to establish the
existence and identity of DNA sequence motifs involved in
chromatin remodeling.
It is also an object of the present invention to identify a
means by which the chromatin state of these elements can be
specifically remodelled and consequently by which specific
regulation of chromosome function in cis or in trans can be
effected.
It is a further object of the invention to provide chemical
compounds which specifically interact with these elements
and bring about chromatin remodeling to specifically
regulate chromosome function.
It is also an object of this invention to establish that
small DNA sequence-specific compounds binding to chromatin
responsive elements (CRE) can mediate modeling.
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
6
The objectives of the present invention have been fulfilled
by the identification, by the inventors, of chromatin
responsive elements (CRE) in the genome. The inventors have
also demonstrated that molecules capable of binding in a
sequence-specific manner to the CREs trigger chromatin
remodeling and thereby directly or indirectly regulate gene
function in a predetermined genomic segment or segments.
Specifically, the invention concerns a process for
modulating the function of a DNA element in a eukaryotic
cell,
comprising the step of contacting a genomic DNA
element, so-called e< chromatin responsive element » (CRE),
with a compound having the capacity to bind in a
sequence-specific manner to said CRE, and preferably having
a molecular weight of less than approximately 5 KDa,
said step of contacting being carried out in
conditions permitting chromatin remodeling of the CRE by
said compound,
wherein said chromatin remodeling of the CRE alters
the activity of one or more other DNA sequences in the
genome.
The present inventors have thus shown that chromosome
function is regulated in cis and in traps by DNA sequences,
designated CREs, whose chromatin status affects the activity
of other genomic sequences. They have also shown that the
binding of sequence-specific compounds to the CREs causes
chromatin remodeling of the CRE, and thereby elicit the
regulatory action. Consquently chromosome function can be
regulated by contacting the CREs with binding compounds.
In the context of the present invention, the term Chromatin
Responsive Element (CRE) signifies a DNA sequence whose
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
7
chromatin status allows the modulation of chromosome
function in cis or in trans. The remodeling of the chromatin
of the CRE, brought about by the DNA-binding molecule,
causes a change in the function of (a) chromosome segments)
different from, or including, the CRE.
In the context of the present invention, « chromatin
remodeling » signifies any change in the chromatin,
including changes to DNA conformation with respect to the
histones such as . modification of the rotational phasing of
the DNA on the histone octamer or modified accessibility of
the DNA to DNA-binding proteins. The histone conformation
may also be modified for example by rearranging or evicting
components of the histone octamer, histone Hl or non-histone
proteins. Changes in DNA/histone interactions may be made,
for example by modifying the total length of DNA per
nucleosome, by reducing or increasing nucleosome stability,
or by modifiying nucleosome mobility in cis or in trans.
The epigenetic state of the CRE is modified as a result of
the chromatin remodeling. The epigenetic state of a DNA
element can be considered to be the information content of
the element which arises from characteristics other than its
sequence.
The DNA element (s) whose function is modified in cis or in
trans by the chromatin remodeling of the CRE will be
referred to herein as « the modulated DNA element ». This
element is DNA and associated proteins. It may also undergo
an epigenetic alteration, including chemical modification
such as methylation, as a result of the change in the CRE,
for example the change in chromatin state of the CRE may
give rise to a change in the chromatin state of the
modulated DNA element, thus modifying its function. However,
the modulation of function of this DNA element may
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
8
ultimately arise from other types of changes such as
redistribution, displacment, inhibition, enhancement of
binding factors, initially caused by, the chromatin
remodeling of the CRE.
The CRE or the modulated DNA elements) may comprise
heterochromatin, heterochromatin-like DNA, euchromatin or
naked DNA. According to a preferred embodiment the CRE is
heterochromatin or heterochromatin-like and its remodeling
converts it to a euchromatin-like accessible state.
The CRE may comprise single copy DNA or multicopy DNA, and
it may contain identical or non-identical sequence motifs,
or functionally interacting multipartite DNA segments.
Particularly preferred CREs comprise repeat sequences such
as satellite DNA, for example a series of GAGAA repeats. The
CRE may comprise a DNA element involved in chromosome
structure and function such as Scaffold Associated Regions
(SARs), which are AT-rich fragments composed of numerous
clustered, irregularly spaced runs of As and Ts.
Alternativley, the CRE may comprise unwinding motifs, non-B
type DNA structure (containing kinks or bends), or DNA
elements with a propensity to position nucleosomes.
The CREs are usually, but not always, in non-coding,
transcriptionally inactive sequences. The CRE may have a
length ranging from about 6 to several thousand base-pairs.
In the latter case, only part of the CRE is targeted by the
sequence-specific DNA binding molecule. If the CRE
encompasses repeat sequences, multiple binding molecules
will bind within the CRE.
The modulated DNA element may be on the same DNA molecule as
the CRE, in which case the CRE is said to be cis-acting. The
modulatory effect can be exerted by the CRE in a local
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
9
manner (i.e. over several tens of base pairs), or in a long
distance manner (i.e. from about 100 upto several thousand
base pairs), and can indeed extend over the whole of the
chromosome. Thus the modulated DNA element may be positioned
immediately flanking the CRE, or may be separated from the
CRE by tens, hundreds or thousands of base pairs. An example
of such a situation is where a heterologous gene has
integrated into the genome in a position juxtaposing a
heterochromatic satellite region. This embodiment of the
invention is illustrated in the examples below by the white
mottled PEV phenotype experiments.
The CRE and the modulated DNA element may coincide. In this
case, the chromatin remodeling of the CRE gives rise to a
direct effect on the function of the CRE-containing DNA
element.
According to a further embodiment of the invehtion, the CRE
may also be trans-acting in that the modulated DNA is or are
not on the same DNA molecule as the CRE, or are not directly
linked to the CRE. The modulatory effect exerted by
chromatin remodeling of trans-acting CREs can arise as a
result of displacement, redistribution, inhibition, or
enhancement of DNA-binding factors which affect gene
function. This embodiment of the invention is
illustrated in the examples below by the brown-dominant
PEV phenotype experiments.
According to the invention, the DNA element whose function
is modulated by the CRE (i.e. the « modulated DNA element »)
can be any potentially active or inactive DNA element.
Particularly preferred DNA elements comprise regions
involved in the binding of DNA-binding proteins, for example
transcription regulatory regions, locus-control regions,
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
origins of replication, boundary / insulation elements,
chromosome structural elements.
The chromatin state of the modulated DNA element, prior to
modulation, may be heterochromatic, heterochromatin-like, or
euchromatic. It may also be naked DNA. After modulation
mediated by the chromatin remodeling of the CRE, the
chromatin state of the modulated DNA element may be changed
or unchanged with respect to its state before modulation.
Often, the modulation involves the conversion of the
modulated DNA element from a heterochromatin-like state to a
euchromatin-like state.
The modulatory effect of a given CRE is specific in so far
as it is exerted on a particular DNA or series of DNA
segments. Depending on the CRE in question, one unique DNA
may be modulated, or on the contrary a~ multiplicity of DNA
segments may be affected. Moreover, the effect exerted on
the modulated DNA may in itself give rise to a cascade of
further cis or trans modulatory effects. Thus the spectrum
of effects which can be achieved using the present invention
is broad and can be controlled by choice of the CRE, and of
the CRE-binding molecule. Preferred CREs are unique in the
genome.
The modulation induced by the chromatin remodeling of the
CRE can involve one or more of the following effects .
restoration of chromosome function, loss of chromosome
function, enhancement of chromosome function, reduction of
chromosome function, prevention of chromosome function,
modification of the temporal or spatial specificity of gene
function, and maintenance of chromosome function.
Particularly preferred effects include restoration of gene
function, for example by suppression of cis or trans
epigentic gene silencing. This variant of the invention is
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
11
particularly applicable for ensuring the function of
heterologous genes, but can also be employed in therapy for
activating endogenous genes which have become epigenetically
silenced.
Another preferred embodiment is the loss of gene function by
redistribution, displacement or inhibition of euchromatic
binding factors involved in chromosome function, or by
allowing the binding of such factors.
According to the invention, the CREs and the modulated
DNA(s) may both be endogenous to the cell. In such a
situation, the process of modulating chromosome function in
accordance with the invention comprises simply the
introduction into the cell of the sequence-specific CRE
binding compound. This variant of the invention is
particularly applicable for activating endogenous genes
which are epigenetically silenced, either as a result of
chromosome rearrangements or for example as a result of
tissue and developmental specificity. It can also be used to
induce loss of function of endogenous genes which are
otherwise active.
In another variant, a CRE which is endogenous to the cell is
used in combination with a modulated DNA which is
heterologous to the cell. According to this variant, a
heterologous DNA to be modulated is introduced into the cell
in conditions allowing its integration in a chromosomal
location where the endogenous CRE can exert its modulatory
effect. Since the CREs of the invention can be chosen such
as to exert their effect in cis over a short or long
distance, or even in trans, it is possible to achieve the
functional interaction between the endogenous CRE and the
heterologous modulated DNA without undue effort. According
to this variant, the sequence-specific CRE-binding compound
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
12
is selected to bind to a CRE which has the capacity to
functionally interact with the particular heterologous DNA
of interest.
A further variant involves the use of a heterologous CRE in
association with an endogenous modulated DNA. This variant
of the invention involves introducing the heterologous CRE
into the cell, in conditions allowing its integration in a
chromosomal location where it can exert its modulatory
effect on the endogenous gene in question. The CRE-binding
compound must also be introduced. Again, the fact that the
CREs of the invention can be chosen to exert their effect in
cis over a short or long distance, or even in traps, allows
the functional interaction between the heterologous CRE and
the endogenous modulated DNA to be achieved without undue
effort. This variant of the invention can be used for
example to modulate the activity of chromosomal sequences
which cannot be sufficiently regulated by endogenous CREs
either as a result of the nature of the sequences involved,
or as a result of positioning on the chromosome.
A further variant involves the use of a heterologous CRE in
association with a heterologous modulated DNA. This variant
of the invention is particularly useful for modifiying the
epigenetic state of heterologous genes, for example for
preventing or reducing epigenetic gene silencing. The
process according to this variant of the invention
comprises .
- transforming a cell, preferably in a stable manner,
with a nucleic acid sequence comprising the heterologous
gene, and with a nucleic acid sequence comprising the
heterologous CRE,
- introducing into the cell a compound which has the
capacity to bind in a sequence-specific manner to said
heterologous CRE,
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
13
said step of contacting being carried out in conditions
permitting chromatin remodeling of the heterologous CRE by
said compound,
wherein said chromatin modelling of the CRE modulates the
epigenetic state of the heterologous gene.
According to this variant, the heterologous CRE and the
heterologous gene to be modulated may be introduced into the
cell on the same or separate molecules of DNA, and they may
be introduced simultaneously or subsequently one to the
other. Furthermore, the introduction of the sequence-
specific binding compound may be carried out prior to,
simultaneously with, or subsequent to the introduction of
the nucleic acids carrying the heterologous sequences of
interest. In a preferred embodiment, the CRE and
heterologous gene are introduced into the cell on the same
molecule of DNA and the CRE-binding molecule is introduced
subsequently when stable transformation has been
established.
Again according to this embodiment, the heterologous CRE may
act in cis or in trans on the heterologous gene.
Preferred examples of this variant of the invention include
the use of satellite sequences as a CRE, in association with
heterologous genes encoding growth factors, hormones,
receptor proteins, viral proteins, regulatory RNAs, tumour
suppressor genes, haemoglobin gens, genes involved in the
immune response, therapeutic protein factors.
The process of the invention may be carried out in vivo, in
vitro or ex vivo. In vivo use is particularly preferred. For
ex vivo use, cells are taken from an organism and
genetically modified to contain either a heterologous CRE or
a hetereologous gene to be modulated, or both, and are
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
14
reimplanted in the body. The sequence-specific GRE-binding
molecule is introduced into the modified cells by any
approriate method such as ingestion, injection, topical
application etc.
For in vitro use, the three essential components of the
invention, that is the CRE, the DNA whose function is to be
modulated and the CRE-binding compound are combined in vitro
in conditions allowing binding of the compound to the CRE.
According to the invention, the cell in which the chromatin-
remodeling mediated modulation is effected can be eukaryotic
or prokaryotic. Eukaryotic is particularly preferred.
Suitable examples are vertebrate cells, invertebrate cells,
plant cells, particularly mammalian cells, insect cells, or
yeast cells. Human cells may be used. Cells from animal
species useful in the production of heterologous proteins or
in animal models, for example, bovine, ovine, avian, fish,
equine, simian cells etc are all suitable. Plant cells are
also particularly preferred Indeed, heterologous genes
inserted into plants are particularly susceptible to gene
silencing and thus the technique of the invention is
advantageous.
The preferred CREs of the invention are satellite sequences.
However, the invention is not limited to such sequences.
Further CREs can be identified using the teaching of the
invention. Specifically, a series of compounds which
specifically bind in the vicinity of the DNA element to be
modulated are used, and any compounds) affecting the
epigenetic state e.g. facilitating the interaction of
factors, is selected. Knowledge of the chromatin structure,
for example the position of the nucleosomes and the DNA-
binding factors (e.g. transcription factors), is useful to
select candidate CRE motifs. Initially the epigenetic state
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
of said DNA element to be modulated and the alteration
thereof by compounds is monitored in vitro with the help of
chromatin probes such as nucleases. This rapidly identifies
compounds with the desired property. Subsequently, in vivo
experiments are carried out to evaluate the phenotypical
changes
The sequence-specific CRE binding compound may bind to the
CRE through major groove interactions, minor-groove
interactions, phosphate back-bone interactions, or a
combination of these types of binding. According to a
particularly preferred embodiment of the invention, the
sequence-specific compound binds to the DNA minor groove.
The compound preferably has a molecular weight of less than
5 KDa, for example less than 4.5 kDA, or less than 4kDa.
The sequence-specific CRE binding compound ~is preferably
cell-permeable, greatly facilitating its introduction into
the cell. For administration to animals, including
administration to humans for therapeutic purposes, the
molecules can thus be administered orally, topically, by
injection etc.
In the context of the present invention, a molecule is said
to bind in a sequence specific manner to the CRE target if
the cell or organism in which the binding occurs presents no
intolerable side-effects or toxicity as a result of the
binding. By intolerable side effects or toxicity is meant
life-threatening, or of sufficient gravity to cause
undesired disruption of metabolism and biological function.
Preferably, a molecule which binds in a sequence specific
manner is capable of specifically recognising a DNA target
sequence of at least 6, preferably at least 8, more
preferably at least 10, even more preferably at least 12 and
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
16
most preferably at least 18 nucleotides, in a chromatin
context
The CRE-binding compounds of the invention preferably have
an apparent binding affinity with respect to the CRE, of at
least 5 x 10' M-1, as measured by in vitro techniques, such
as footprinting techniques. More preferably the compound has
an apparent binding affinity of at least 1 x 109 M-1' and
even more preferably of at least 5 x 101° M-1.
Particularly preferred examples of CRE-binding compounds of
the invention are DNA-binding organic o~ligomers comprising
heterocycles, for example wherein the heterocycles have at
least one annular nitrogen, oxygen or sulphur.
Examples of such oligomers include heterocycles chosen from
pyrrole, imidazole, triazole, pyrazole, furan, thiazole,
thiophene, oxazole, pyridine, or derivatives of any of these
compounds wherein the ring NH group is substituted. Such
compounds are described in Bailly C., and Chaires J.,
Bioconjugate Chem, Vol 9 N°5, 1998, 513-538.
Particularly preferred compounds contain N-methylpyrrole
(Py) and / or N-methylimidazole (Im), and may further
contain aliphatic amino acids such as (3-alanine and y-
aminobutyric acid. The synthesis of DNA-specific compounds
of this type containing N-methylpyrrole (Py) and / or N-
methylimidazole (Im), has been described (Geierstanger et al
1994). These pseudo-peptides, based on the structure of
naturally occurring distamycin, bind DNA in the minor groove
as antiparallel dimers (Pelton and Wemmer 1989). Their
sequence-specificity depends on the side-by side pairing of
this dimer where an Im opposite a Py (Im/Py) targets a GC
base pair, a Py/Im recognizes a CG base pair and a Py/Py
pair is degenerate for both AT or TA base pairs (White et
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
I7
al., 1997). Py-Im compounds have been shown to be cell
permeable (Gottesfeld et al., 1997). These compounds will be
referred to hereinafter as polyamides.
Using the pairing rules mentioned above, an appropriate CRE-
polyamide is synthesised to recognise a given CRE. The
sequence of the CRE determines the stucture and composition
of the polyamide. Many CRE-binding compounds can be made
applying these rules.
A non-limiting example of a general formula which can be
adapted to fit a particular CRE sequence is the following
Formula I .
N C
L tA~ a - a~ -z
(I)
wherein
- A is a monomer unit selected from the group
consisting of an aromatic amino acid residue,
particularly a heterocycle having at least one
annular nitrogen, or the aliphatic amino acid (3-
alanine ((3), or fluorescent derivatives of said
aromatic amino acid residues ;
- a represents an integer from 6 to 9,
- ~3 represents (3-alanine
- Z represents dimethylaminopropylamide (Dp) or another
end group, or a fluorescent derivative thereof,
- each solid line represents a covalent bond,
- N and C indicate the N- and C-terminal extremities of
the molecule, respectively,
with the proviso that .
- the multiple A monomer units may be the same or
different.
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
18
According to a preferred embodiment, in the above Formula I,
a is 7 or 8.
According to a further embodiment, [A]a comprises at least
four aromatic amino acids, and [A]a does not comprise a
stretch of more than three contiguous aromatic amino acids.
Preferably, the multiple A units of Formula I comprise N-
methylpyrrole (Py) and / or N-methylimidazole (Im). An
example of this type of molecule has the formula (II) .
[Ai]-[Az]-[A3]-[Aa]-[AS]-[A6]-[A~]-a-Z (II)
wherein (3 and Z are as previously defined,
[A4] is (3,
[Al] to [A3~, and [AS] to [A.,] are chosen from N-
methylpyrrole (Py) and / or N-methylimidazole (Im).
Preferred embodiments of Formula II are those wherein [A1]
to [A3~, and [AS] to [A~] are each N-methylpyrrole (Py) .
Another preferred molecule has the formula (III):
[Ai]-[Az]-[As]-[A4]-[AS]-[A6]-[A~]-[AB]-(3-Z (III)
wherein (3 and Z are as previously defined,
[A1] to [As] are chosen from N-methylpyrrole (Py) , N-
methylimidazole (Im) and a (3 alanine residue,
with the proviso that the [A] immediately adjacent to
each Im on the N-terminal side is a (3 alanine residue.
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
19
A specific examples of the CRE-binding molecules of the
invention are .
[ Im-[3-Im-Py-(3-Im-(3-Im-(3] -Dp
The above molecules are well suited for binding to CREs
since they show high affinity and specificity and are cell
permeable. Further details are provided in the Examples.
The invention also relates to a gene expression kit suitable
for modulating the epigenetic state of a heterologous gene
in a cell, said kit comprising .
- a nucleic acid molecule comprising said heterologous
gene ;
- a nucleic acid molecule comprising a so-called
heterologous ee CRE », said heterologous CRE being a
sequence whose chromatin status allows the modulation
of chromosome function in cis or trans ;
- a compound having a molecular weight of less than
approximately 5 KDa, and having the capacity to bind in
a sequence-specific manner to said CRE.
The heterolog'ous CRE in such a kit may comprise a satellite
sequence, for example a SAR-like AT tract or a GAGAA repeat
sequence. Such kits may be used in gene therapy. The
heterologous gene may be any gene of interest for example
those cited earlier.
The invention also relates to a cell containing a compound,
preferably having a molecular weight of less than 5KDa, and
having the capacity to bind in a sequence-specific manner to
a genomic CRE, said CRE being a sequence whose chromatin
status allows the modulation of chromosome function in cis
or trans. Preferably the compound within the cell binds the
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
DNA-minor groove, although major groove binders and
phosphate back-bone binders may also be contemplated.
The cell according to the invention may additionally
contain
- a nucleic acid molecule comprising a heterologous gene ;
- a nucleic acid molecule comprising a so-called
heterologous « CRE », said heterologous CRE being a sequence
whose chromatin status allows the modulation of chromosome
function in cis or trans.
The cell of the invention may be a eukaryotic cell or a
prokaryotic cell, as cited earlier.
Non-human organisms comprising such cells also form part of
the invention. The organism may be a non-human animal. It
may or may not be transgenic, depending upon whether any of
the three components of the invention, i.e. the CRE, the
modulated DNA element and the CRE-binding molecule are
stably introduced into the cell by stable transformation.
Particularly preferred organisms are plants, which may be
non-transgenic or transgenic. Gymnosperms and angiosperms
are particularly suitable for use in the present invention,
the latter group including monocotyledons and dicotyledons.
The invention also relates to a compound having the capacity
to bind, in a sequence-specific manner, to a predetermined
CRE, said CRE being a sequence whose chromatin status allows
modulation of chromosome function in cis or in trans, with
the proviso that said compound is not distamycin, HMG-I/Y,
MATH2 0 . - .
In a preferred embodiment, the compound has a molecular
weight of less than 5KDa and has the capacity to.
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
2I
specifically recognise a sequence of at least 6 nucleotides.
Even more preferably, the compound has the capacity to
specifically recognise a sequence of at least 8, or at least
nucleotides, for example at least 15 or 16 nucleotides .
The compound is preferably cell permeable.
Examples of such compounds are discussed earlier.
The compounds of the invention may be combined with a
suitable physiologically acceptable excipient to prepare
pharmaceutical compositions for use in humans or animals.
Particularly preferred excipients are those for oral,
topical, sub-cutaneous, intramuscular administration.
The pharmaceutical compositions of the invention may be used
together with other pharmaceutical compositions comprising
the necessary nucleic acid molecules for production of
~heterologous CREs and modulated elements. Such associations
comprise a first pharmaceutical composition containing
a nucleic acid molecule comprising a heterologous
gene ; a nucleic acid molecule comprising a so-called
heterologous « CRE », said heterologous CRE being a sequence
whose chromatin status allows the modulation of chromosome
function in cis or trans, said nucleic acid molecules being
in association with a physiologically acceptable excipient,
and
a second pharmaceutical composition comprising a compound
having the capacity to bind, in a sequence-specific manner,
to said CRE, in association with a physiologically
acceptable excipient.
The compounds, compositions associations o'f compositions and
kits according to the invention can be used in therapy, for
example in therapy of genetic disorders resulting from
epigenetic status. Examples of such disorders are fragile X
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
22
syndrome and imprinting disorders such as Wilm's tumour, and
Prader-Willi syndrome.
The compounds and kits of the invention can also be used in
a non-therapeutic manner for modulation of the expression of
heterologus genes in cells, particularly eukaryotic cells.
The cells may be in culture or in vivo. When the method is
carried out in vivo, the organism may be transgenic or non-
transgenic.
According to a particularly preferred embodiment, the
compounds of the invention are fluorescent or fluorescently
labelled. More particularly, this aspect relates to a DNA-
binding compound capable of sequence specific binding to
genomic DNA, said compound being an oligomer comprising
cyclic heterocycles having at least one annular nitrogen,
and optionally at least one aliphatic amino acid residue,
wherein said compound is fluorescent or fluorescently
labelled. The CRE-binding compounds described above are
particularly preferred variants of this aspect of the
invention.
It has surprisingly been shown by the present inventors that
the addition of a fluorescent tag to the DNA-binding
molecules does not alter the specificity of the binding.
This permits the use of the fluorescent derivatives for
cytological / structural determinations, including
quantitative estimations of specific DNA sequences in cells
and chromosomal material.
The fluorescent tags are usually added at the N or C
terminal of the molecule, and can be a fluorescent dye such
as fluorescein, dansyl, Texas red, isosulfan blue, ethyl
red, malachite green, rhodamine and cyanine dyes.
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
23
The fluorescent CER-binding molecules can be used for
probing the epigenetic state and location of DNA in
chromosomes and nuclei, and for diagnosis of pathological
conditions arising from epigenetic status, including pre-
symptomatic diagnosis.
Further uses of the fluorescent derivatives of the invention
are chromosome marking, diagnosis, forensic studies, and
affiliation studies.
Figure legends
Various aspects of the invention are illustrated in the
figures .
Figure 1: DNase I footprint assays with P9 and P7.
DNase I cleavage pattern in the presence of P9 and P7.
higand concentrations are indicated at the top of each lane.
The position of each of, the AT-tracts is indicated by square
brackets. Panel A shows the footprints of P9 and P7 on probe
W9. This probe is composed of head-to-tail tandem repeats of_
an oligonucleotide with a 9 by AT-tract.
Figure 2: Staining of Drosophila nuclei with fluorescently
tagged oligopyrroles.
Isolated Itc nuclei were stained with ethidium bromide and
fluorescein-tagged oligopyrroles as indicated. Note that P9
(panel A) highlights as intense green foci satellites I and
III and that the general nucleoplasmic background staining of
P9 is low.
Figure 3 . Binding specificity of P31 and GAGA factor
Panel A: DNAse I footprinting experiment with P31 and
affinity cleavage with P31E are shown on GAF31 and the Brown
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
24
I probes. The GAF31 and Brown I probes contains a (AAGAG)2
motif and GAGA factor (GAF) binding site from the Ubx
promoter (Biggin et al., 1988). Note that P31 does not bind
the typical GAF binding (Ubx) . The Brown I oligo (a. tandem
repeat) includes an (AAGAG)5 binding site and a degenerate
P31 binding site (AACAC)2 as indicated. P31 concentrations
used (nM) are indicated. Lanes labeled P31E (top) are
affinity cleavage reactions with 1 nM of P31E on either
probe. Binding orientations of P31E on these probes are
indicated by arrowheads on the brackets pointing towards the
N-terminus of the molecule. The letter G refers the G
nucleotide cleavage reaction. Panel B: DNAse I footprinting
experiment with purified GAGA factor (GAF) on the GAF31
probe . Note that GAF binds both the (AAGAG) 2 motif and the
binding site from the Ubx promoter.
Figure 4 . The fluorescent polyamide P31T specifically
highlights the GAGAA satellite V
Isolated ECc nuclei and polytene chromosomes were
stained with DAPI (blue), P31T (Texas red-labeled P31), P9F
(Fluorescein tagged P9). Panel A: The green P9F foci are
proposed to highlight satellites I and III. P31T marks the
separate positions of the GAGAA satellites. Panels B & C:
The black and white panels display the red and green
channels of panel A, respectively. Panel D: Staining of
brown-dominant polytene chromosome with DAPI, P31T and P9F.
The polytene banding pattern is shown in blue (DAPI). P31T
highlights in red the heterochromatic GAGAA repeats of the
allele bwD at 59E.
Figure 5 . Oligopyrroles induced chromatin opening of
satellite III.
Kc nuclei were incubated with mitotic Xenopus egg
extracts in the presence of the various polyamides and then
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
further treated with VM26 to accumulate the so-called
cleavable complexes of topoisomerase II. Cleavage in
Drosophila satellite III was revealed by southern blotting.
Satellite III contains a major topoisomerase II cleavage site
once per 359-by repeat. The extent of the cleavage activity is
reflected by the development of the ladder of multimers of the
basic repeat. All panels included controls with (+) and lanes
without Vm26 (-). The massive activation of cleavage
(chromatin opening) mediated by P9 and the reduced activity
P31 in this assay is shown.
Figure 6: Binding Mode of Polyamide P9 and P31
Synthesis and characterization of these DNA satellite-
specific polyamides is described in the examples. Both
compounds, P9 and P31 were found to bind their targets with
subnanomolar affinity in a 1:l drug to DNA complex by hydrogen
bonding schemes as proposed in this figure.
Panel A: Proposed 1:1 binding model for the complex of P9 (Py-
Py-Py-[3-Py-Py-Py-(3-Dp) where Py - N-methylpyrrole, (3=~i-
alanine, Dp=dimethylaminopropylamide with AATTAATAT. White
balls and diamonds represent pyrrole and (3-alanine
respectively. Circles with two dots represent lone pairs of
electrons on N3 of purines and 02 of pyrimidines at the edges
of the bases. Putative bifurcated hydrogen bonds to the amide
NH's are illustrated by dashed lines.
Panel B: Proposed 1:1 binding model for the complex of P31
(Im-(3-Im-Py-~3-Im-~3-Im-(3-Dp where Im - N-methylimidazole)
with AAGAGAAGAG. Black balls represent imidazole. Circles
containing an H represent the N2 hydrogen of guanine. Dashed
lines illustrate putative hydrogen bonds. Consensus binding
sequences are indicated.
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
26
Figure 7: Specific suppression of the white-mottled eye
phenotype with P9
Different polyamides drugs (indicated) were fed to
developing r~t"4 flies and representative eye-phenotypes of 5-
day old flies are shown. Polyamides (final concentration 100
~,M) were present in the fly food from egg laying to
hatching. Only oligopyrrole P9, which opens satellite III,
was found to suppress the ~"4 eye phenotype.
Figure 8: Eye-pigment level determination
Eye-pigments were extracted . and determined
spectrometrically from 30 carefully dissected fly heads. The
pigment levels of 5-day old wm4 male flies of the experiment
presented in Figure (2) are shown. Included are also the
eye-pigment levels of heterozygous brown-dominant (bwD/+),
white-eye (w67) or wild-type (Canton S.) flies. Genotypes
are indicated. Note, that only compound P9 is a suppressor
of PEV in wm4 flies, leading to an increased activity of the
white gene. Neither P9 nor P31 modify PEV of brown-dominant
flies.
Figure 9: P31 induces homeotic abdomen transformations in
bw° flies
Representative eye phenotypes and homeotic
transformations are shown. Panels A and B show eye
phenotypes of heterozygous brown-dominant (bwD/+) flies
raised in the presence of P9, P31 or no drug (indicated). A
slight increase in the eye pigmentation (red ommatidia) is
observed in P31 treated flies (panel B). This increase is
thought to reflect the more advanced age (65-75 hours) of
these flies (delayed development) rather than a genuine
suppression of PEV. Panel C shows the abdomen of dissected,
heterozygous bwD pupae raised in the presence of P31 or P9
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
27
(indicated) . The A6 to A5 transformation induced by P31 is
manifested by the formation of sternite bristles on the A6
segment of flies (arrows). Wild type males are devoid of
sternite bristles in A6 (left). Panel D shows the abdomen of
dissected pupae raised in the presence of P31 (indicated).
The A6 to A5 transformation is more penetrating (Table 1) in
the homozygous bwD/bwD flies (arrows). This transformation
requires the bwD allele and is not observed in Ubxl/+ flies
or other genotypes.
Figure 10 . P31 induces sex comb reduced phenotype in bud
flies
Photographs of the first thoracic male leg showing
examples of sex comb phenotypes of heterozygous and
homozygous bwD flies obtained after P31 treatment (+P31)
compared to untreated flies. The sex comb reduced phenotype
is induced by P31 (not P9) only in a bwD genetic background.
Mean values for number of teeth are indicated.
Figure 11 . P31 enhances the halters-to-wing homeotic
transformation of Ubx 1 in bc~° flies
Photographs of the halteres of animals with different
genotypes raised in the presence of P31 or absence (-).
Note that P31 enhances the halters-to-wing transformation of
Ubx 1 only in a bwD genetic background. Hence, P31 mimics
the genetic interaction of Tr113C and Ubx.
Figure 12 . P31 induces recruitment of GAF to the bwD
insert in interphase
Photographs of homozygous bwD .polytene nuclei
immunostained for GAF (green). DNA is highlighted in blue by
DAPI and the bwD insert is highlighted in red by P31T. Black
and white inserted panels show the red (right) and green
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
28
(left) channels separately for the bwD foci. Drug used are
indicated. Note that the GAF and P31T signals only overlap
following exposure of the permeabilized glands to P31.
EXAMPLES
SECTION I Synthesis and Characterization of DNA Satellite-
specific Drugs
Genome projects not only discover a daunting number
of new genes, they also yield an enormous amount of non-
coding sequence data which must inevitably include
'architectural' DNA elements. Architectural DNA is proposed
to harbor sequences that mediate nuclear order, chromosome
stability and dynamics, sister chromatid cohesion,
centromere and telomere formation. While it is conceivable
that the tools of proteomics combined with new technologies
will eventually allow the assignment of tentative functions
to many of the discovered genes (Frederickson, 1999), we are
poorly equipped to discover, predict and assign functions to
non-genic and architectural DNA. Yet, to understand
chromosome biology, we must not only understand gene
function but also how these DNA elements impose and then
transmit the inheritance of chromosome structural features
through cell division.
Biological assays to study architectural DNA are
extremely limited. For example, although the phenomena of
position effect variation (PEV) is attributed to the
positioning of genes near centric heterochromatin (Henikoff,
2000; Karpen, 1994), genetic tools to dissect the functions
(if any) of centric satellite DNA are lacking. Is PEV
mediated by the sequence satellite repeats, by base
composition, by their epigenetic state or simply by the
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
29
repetitive nature of its chromatin? In view of the
difficulties we encounter of assigning functions to large
fractions of the genome, we consider the development of new
approaches and tools of major importance.
The approach we successfully developed here is based on the
synthesis of DNA sequence-specific pseudopeptides
(Geierstanger et al., 1994) to study the biological role of
architectural DNA involved in chromosome condensation and
PEV.
Our interest in these compounds stems from their potential
as a tool for molecular and cell biology. Sequence specific
minor groove binding drugs may permit a dissection of. the
role of repeated DNA and of difficult cis-acting elements.
Py-Im compounds may be used, if fluorescently labeled, for
in situ localization of specific sequences, possibly
allowing the study of repeated DNA in living cells.
Here we -describe the synthesis and characterization of
polyamides that target different DNA satellites and SARs.
Interestingly, we observed that compounds targeted to
satellite III massively unfold this heterochromatic repeat and
that SAR-specific polyamides inhibit chromosome condensation
in mitotic Xenopus egg extracts. We also show that pyrrole-
imidazole compounds targeted to two different DNA satellites
of Drosophila melanogaster have a dramatic effect on PEV and
gene expression (Janssen et al., 2000). These observations
illustrate the powerful utility of sequence-specific minor
groove binding drugs for chromosome research.
Example 1
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
Synthesis of oligopyrroles for taraetinct AT-tracts
To explore the biological potential of polyamides, we
aimed at synthesizing compounds that target DNA satellite I,
III, V and the interspersed SAR elements. Satellite I (1.672
density) consists of AATAT units encompassing about 6
megabases (Mb) Satellite V (1.705 density) is composed of
AAGAG repeats amounting to about 7 Mb (Lohe et al., 1993).
Satellite III (1.688 density) has a much longer repeating unit
(359 bp) and covers about 10 Mb (Hsieh and Brutlag, 1979).
Satellite III repeats behave operationally like SARs (Kas and
Laemmli, 1992), the sequence hallmarks of which are numerous
clustered AT-tracts. For example, the SAR associated with the
Drosophila histone gene cluster is defined by a 656 by
EcoRl/Hinf1 fragment containing 26 AT-tracts of 8 or more Ws
(A or T bases) with an average length of 10 base pairs (Gasser
and Laemmli, 1986; Mirkovitch et al., 1984). Twenty of these
AT-tracts are clustered and separated by a spacer of only a
few nucleotides (average 4.5) of mixed base pair sequence.
To enlarge binding site size and improve affinity,
the number of N-methylpyrrole units can be increased, since
each pyrrole carboxamide contacts one AT base pair. However,
for compounds Containing more than six pyrroles this
prediction is no longer valid since the molecule gets out of
phase with the base pairs along the minor groove floor.
Indeed, the pyrrole-pyrrole distance is about 20% longer then
required for perfect match (Goodsell and Dickerson, 1986). In
addition, compounds with five or more pyrrole rings are found
to be over-bent relative to the pitch of the DNA helix
resulting in decreased binding affinities for longer
oligopyrroles (de Clairac et al., 1999). To circumvent this
mismatch problem, a flexible amino acid ((3-alanine) can be
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
31
introduced in the center of the pyrrole ring system to restore
register of the recognition elements and relax the curvature
of these crescent-shaped molecules (Youngquist and Dervan,
1987). A pyrrole hexamer termed P9 was synthesized containing
a central (3-alanine (PyPyPy-[3-PyPyPy-(3-Dp) and was observed
that to bind W9 with 100-fold better affinity (Kapp about 0.75
nM) than P7 (Figure 1A) . This latter value was obtained from
footprints that extended to lower ligand concentrations than
those shown in Figure (1A).
Example 2
Selective staining of DNA satellites and SARs in nuclei and
polytene chromosomes.
Drosophila Kc nuclei:
To address the question of the specificity of
oligopyrroles when probed on DNA packaged by histones into
chromatin the possibility of fluorescently tagging pyrrole
ligands in order to stain isolated Kc nuclei and polytene
chromosomes for examination by epifluorescence microscopy was
explored. If sequence preference is maintained upon tagging
and also extends to chromatin, it should be possible to
highlight in stained nuclei the positions of the main targets
of these fluorescent oligopyrroles (satellites I and III).
Fluorescent groups were coupled to oligopyrroles
using commercially available succinimidyl active esters of
fluorescein. DNase I footprinting of the fluorescent ligands
revealed that these derivatives are differently affected upon
tagging. In general, tagging resulted in reduced binding
affinity but never affected AT-specificity. Interestingly, for
some compounds an improved SAR specificity factor was observed
(see Table 1). The SAR specificity of P9F was increased about
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
32
4 fold. The fluorescent moiety of this molecule may serve to
improve discrimination.
Drosophila Kc nuclei were double stained with
ethidium bromide and fluorescein-tagged pyrrole compounds.
Ethidium bromide (red) stains nuclear chromatin generally but
it also markedly outlines the nucleolus due to the high RNA
concentration of this subnuclear domain.
The staining patterns observed with P9F (green) show
striking features; the ligand accumulated at one or two
subnuclear locations (Figure 2A and B) resulting in strong
green foci. These foci are generally abutting the nucleolus
and are proposed to arise from the expected localization at
the abundant AT-rich Drosophila satellites I and III (see
below). A low green signal throughout the nucleoplasm is
observed with the P9F.
This intense nucleoplasmic localization obtained with
the P9F is interpreted to arise from binding to isolated/short
AT-tracts that abundantly occur throughout the genome.
Example 3
Targeting the GAGA.A repeats of Satellite V with P31.
A polyamide that targets the abundant satellite V composed
of GAGAA repeats (Lohe et al., 1993) was synthesized.
Designing molecules that would bind to this repeat motif
represented a challenge since with current knowledge,
targeting of sequences containing 5'-GNG-3' or 5'-GA-3' with
drugs composed of pyrrole and imidazole is difficult.
However, successful targeting to sequences containing 5'-
GTG-3' was previously achieved using an Im-(3-Im motif where
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
33
[3-alanine replaces the function of pyrrole (Turner et al.,
1998) . Since j3-alanine, like pyrroles, is degenerate for A~T
and T~A base pairs, we designed a compound based on these
observations, to recognize a sequence composed of two tandem
GAGAA repeats by systematic placement of -alanine at the N-
terminal neighbor of imidazole. The binding affinity and
specificity of this compound, termed P31 (=Im-(3-Im-Py-~3-Im-
(3-Im-(3-Dp), were evaluated by DNAse I footprinting. For this
purpose, two different probes were examined, both containing
GAGAA repeats. Figure (3A) shows that P31 binds with
subnanomolar affinity to its target binding site, in this
case two GAGAA repeats (lanes 2-8). The apparent binding
constant of P31 for this sequence was estimated at 0.25
nM. At higher concentrations, protection of two mismatch
binding sites was observed. One of these sites contains an
AAGTG motif (Figure 3A).
To determine binding orientation and stoichiometry
for P31, we prepared a Fe (II) -EDTA analogue of P31, termed
P31E (Im-~3-Im-Py-(3-Im-(3-Im-(3-Dp-EDTA) . Affinity cleavage was
carried out on the footprint probe containing two GAGAA
repeats (lane 9) and revealed one major cleavage site
flanking the two GAGAA repeats, thereby confirming the
assumption that one P31 molecule binds two GAGAA repeats in
a 1:1 drug to DNA complex.
A drawback of this binding model, as opposed to
conventional 2:1 drug to DNA complexes, is that P31 is
expected to bind degenerate GC and CG base pairs, albeit
with different affinity. The consensus sequence can thus be
defined as SWSWWSWSWW, where S stands for a G or C and W for
A or T. To evaluate binding of P31 to CACAA repeats, we used
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
34
a second probe that contains two of these repeats as well as
five tandem GAGAA repeats. Figure (3A) shows that P31
protects CACAA repeats with approximately five fold lower
affinity than GAGAA repeats (lanes 11-15). Furthermore,
affinity cleavage reactions using P31E revealed two major
cleavage sites in the GAGAA region (lane 16), showing that
in this case, two P31 molecules are bound in tandem to the
pentameric GAGAA repeat. Again, it is observed than this
molecule binds as a 1:l drug to DNA complex in an
orientation as indicated by arrowheads (Figure 3A). We
propose that special structural features of AT-tracts and
GAGAA repeats might favor 1:1 DNA to drug complexes.
In the Examples below it is demonstrated that P31
fed to developing Drosophila melanogaster of the brown-
dominant genotype interferes with the function of the GAGA
factor (GAF). A footprint experiment was therefore carred
out with this protein. The DNA probe (GAF31) used for this
purpose contains besides the (AAGAG)2 motif (the target of
P31) a typical promoter proximal GAF binding site derived
from the Ubx gene (Biggin et al., 1988). This Ubx site
contains the pentameric consensus sequence GAGAG of GAF
(Omichinski et al., 1997). The DNase I footprint studies
show that, while GAF binds both the (AAGAG)z and Ubx motifs,
P31 interacts only with the former satellite repeats
(compare panels A and B of Figure 3).
Selective Staining of GAGAA Satellite V in Nuclei and
Polytene Chromosomes:
Fluorescent derivatives of P31 were synthesized to
visually assess their binding targets by staining of nuclei
anal chromosomes. DNase I footprinting of the fluorescent
ligands revealed that P31T bound the GAGAA sequence with
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
unaltered specificity but with 100 fold reduced binding
affinity. Drosophila FCc nuclei were triple stained with
DAPI, P9F and P31T and recorded by epifluorescent
microscopy. The micrographs obtained again are striking
since one notes against the blue DAPI background of nuclear
DNA, separate green and red foci stemming from P9F and P31T
staining, respectively (Figure 4A). Closer inspection
reveals that these foci are largely non-overlapping (compare
panels A and B).
In situ hybridization analysis.showed that it is
possible to detect satellite I but not satellite V
((GAGAA)n) in polytene chromosomes obtained from wild type
flies, supposedly due to a more severe under-replication of
satellite V (Platero et al., 1998). Hence, due-to this
apparent absence of GAGAA repeats, the specificity of P31T
for its target binding site cannot be evaluated using
'normal' polytene chromosomes. Therefore, to circumvent this
limitation, we prepared polytene chromosomes from bOwndominane
(bud') flies which harbor an large block of heterochromatin
(about 1.7 megabases) composed of GAGAA repeats inserted
into the coding region of the brown (bw+) gene. This
heterochromatic insert appears to be normally polytenized
(Csink and Henikoff, 1996; Dernburg et al., 1996; Platero et
al., 1998) probably due to its euchromatic localization.
Polytene chromosomes were prepared from these flies and
stained with P9F, P31T and DAPI. The results obtained were
striking (Figure 4). P31T (red) highlighted conspicuously
the bc~ GAGAA insert at locus 59E on the right arm of
chromosome 2 (2R). No other P31T foci were observed, neither
at the chromocenter nor along the euchromatic arms. The
familiar band/interband pattern of polytene chromosomes is
revealed in blue by DAPI staining.
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
36
In summary, we synthesized different satellite-
specific polyamides as established by footprinting and
epifluorescence microscopy. The Im-Py compound P31 was shown
to specifically bind satellite V. All these compounds bind
their DNA targets as 1:1 drug to DNA complexes.
Oligopyrroles mediate chromatin remodelling in a sequence-
specific fashion
Previously, we reported that exposure of nuclei to
distamycin (Py-Py-Py) causes opening of .the chromatin fiber,
thereby facilitating cleavage by restriction enzymes and
topoisomerase II at satellite III (Kas and Laemmli, 1992). Do
synthetic polyamides have similar effects on chromatin? As
mentioned above, satellite III consists of 359-by repeats and
each repeat unit is packaged in two nucleosomes.
Biochemically, satellite III repeats behave as SARs; they
preferentially bind nuclear scaffolds, topoisomerase II, HMG-
I/Y and MATH20 (Girard et al., 1998; Kas and Laemmli, 1992).
Topoisomerase II is also enriched at satellite III in vivo as
demonstrated by microinjection of fluorescent topoisomerase~II
into Drosophila embryos (Marshall et al., 1997). Satellite III
contains one prominent topoisomerase II cleavage site per
repeat located in every second nucleosomal linker (Kas and
Laemmli, 1992). Topoisomerase II cleavage products accumulate
in the presence of the cytostatic drug VM26 when Kc nuclei are
exposed to Xenopus egg extracts, rich in topoisomerase II.
This treatment generates a DNA ladder with a repeat length of
359 by as revealed by hybridization. The ladder is observed
only upon addition of VM26 (Figure 7A, left). Interestingly,
cleavage is massively stimulated by addition P9 (also P7, not
shown). Cleavage stimulation is evidenced by an increased
intensity of the main repeat band (marked M, one cut per 359-
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
37
by repeat) and a shift of the ladder to shorter fragments.
Stimulation is maximal at 500 nM and starts to diminish at
higher concentrations (Figure 7A). P9 exposure also results in
the appearance of additional, minor bands (marked m) that most
likely arise from cleavage within nucleosomes (see
discussion). These minor bands are not observed without the
drug, even after extended exposure (data not shown).
Next, we tested the potency of P31 in this assay. The
results, shown in Figure (7A), demonstrate that P31 stimulates
cleavage considerable less well than P9. That is, while,
massive cleavage stimulation is observed with the lowest
concentration of P9 (62 nM, Figure 7A, lane 3), no significant
reinforcement of the pattern is observed with P31 up to a
concentration of 200 nM (Figure 7A, lanes 8 to 11). Only at
500 nM is cleavage stimulation by P31 comparable to that
obtained with 62 nM of P9 (compare lane 3A to lane 12).
Stimulation with P9 is maximal at 500-1000 nM and starts to
diminish at higher concentrations. The cleavage ladder induced
by P31 at these concentrations is also less pronounced than
that of P9 in keeping with the dose response observed. These
dosage experiments demonstrate that P9 opens the
heterochromatic satellite TII at a roughly 10 fold lower
concentration than P31.
The data presented above demonstrate that the
synthetic oligopyrrole compounds P9 and P7 (not shown)
strongly facilitate cleavage by topoisomerase II. The
stimulation response ,to drug treatment is thought to reflect
the initial opening of chromatin, that facilitating cleavage.
An additional observation that supports the notion of
chromatin opening is that P9 also facilitated cleavage within
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
38
satellite III by restriction enzymes. Satellite III repeats
contain near the topoisomerase II cleavage site a HaeIII
restriction sequence. We previously demonstrated that cutting
by HaeIII in chromatin (not DNA) is facilitated by distamycin
(Kas and Laemmli, 1992) . We made a similar observation using
P9 (data not shown).
Discussion of Section I
We explored the potential of sequence-specific minor groove
binding polyamides as novel tools to address issues of
chromosomal structure, dynamics and the biological functions
of non-genie DNA. To this end, we synthesized compounds that
interact with satellite I (AATAT), V (GAGAA) and SARs,
including the SAR-like satellite III. Although targeting
satellite I and SARs can be achieved with 'conventional'
minor groove binding drugs such as Distamycin, Hoechst and
DAPI, their relatively short binding site give rise to high
background signals.
Synthesizing compounds that bind GAGAA repeats with
high affinity is chemically more challenging since this
sequence includes a 'difficult' motif. However, impressive
targeting to satellite V repeats was obtained with the
monomer P31 which is composed of both imidazole and pyrrole
units. Structurally, P31 extends recent observations that
the 'difficult' triplet GWG sequence can be targeted by a
Im-(3-Im motif where [3-alanine is positioned N-terminal of
imidazoles (Turner et al., 1998). In P31, this design
principal was systematically extended to achieve
subnanomolar affinity for two consecutive GAGAA repeats.
This design expands the number of sequences that can be
targeted, by including GA and GAG motifs.
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
39
Pyrrole-Imidazole drugs generally bind the DNA
minor groove as antiparallel 2:1 drug to DNA complexes
(White et al., 1997). However, the affinity cleavage
experiments presented here suggest a 1:1 drug to DNA complex
for oligopyrrole P31F. Since binding of two antiparallel
oriented molecules requires the expansion of the minor
groove (Kielkopf et al., 1998), widening the AT-tract might
energetically be too costly. Likewise, crystal structures of
B-DNA oligomers demonstrated that GpA steps tend to narrow
the minor groove more than GpT steps (Yanagi et al., 1991)
which in turn may disfavor 2:1 complexes between P31 and
GAGAA repeats.
Epifuorescent microscopy Fluorescent DNA dyes with sequence
preference, such as DAPI or Hoechst, are useful, everyday
tools of cell biology, medicine and cytogenetics. Sequence
specific compounds, if successfully rendered fluorescent,
could extend the scientific potential enormously, since
innumerable basic questions about chromosome structure,
function and dynamics could be addressed using sequence
specific dyes. Also, such molecules could facilitate and
improve more routine work such as chromosome typing.
Although conjugation of a fluorescent label either at the N-
or C-terminal end of oligopyrroles is straightforward, tagging
at these positions altered affinity (Table 1).
The main nuclear targets of P3lwere also demonstrated by
staining isolated Kc nuclei and polytene chromosomes with the
Texas red derivative, P31T. P31T foci must, represent the GAGAA
repeats of the centric satellite V (Figure 4A-C). Positive
identification of the main DNA target of P31T was obtained by
staining of bw° polytene chromosome whose GAGAA repeat was
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
sharply highlighted by this compound (Figure 4D). We observed
no other P31 signals along the euchromatic arms or at the
chromocenter of polytene chromosomes derived from bwp or
Canton S. flies. The repetitiveness of these satellite
sequences and the polyteny of these chromosomes facilitate the
detection of the staining signals. Labeling chromosomes with
sequence-specific polyamides is experimentally
straightforward, allowing the application of such dyes in
innumerable scientific and diagnostic applications. Needless
to say, polytene chromosomes might be the ideal object to
asses the specificity of sequence-specific hairpin polyamides.
Chromatin opening The chromatin studies revealed that.
titration of AT-tracts with oligopyrrole P9 massively
unfolds the heterochromatic satellite III. Chromatin opening
of satellite III is evidenced by the massive stimulation of
cleavage by endogenous topoisomerase II when Kc nuclei were
exposed to Xenopus egg extracts. We previously made similar,
although less pronounced observations, using distamycin and
speculated, that unfolding might arise from a displacement
of histone H1 or another protein from the nucleosomal linker
region (Kas and Laemmli, 1992; Kas et al., 1993).
Alternatively, minor groove contacts of the core histones
could be of importance for maintaining the heterochromatic
state of the chromatin fiber. In contrast to P9, chromatin
opening of satellite III required high concentrations of
compound P31. In contrast to this, in the accompanying
paper, we present data suggesting that, P31 but not P9 can
open the heterochromatic GAGAA insert which constitutes the
brown-dominant allele (b"'D). These observations suggest the
DNA minor groove binding polyamides may serve as sequence-
specific chromatin openers for silenced genes.
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
41
Materials and Methods.
Boc-(3-PAM-resin, HBTU, Fmoc-Glu(otBu)-OH, Boc-(3-alanine and
Boo-y-aminobutyric acid were purchased from Novabiochem AG,
Switzerland. HOBt was from Bachem. The methylester of 4-amino-
1-methylpyrrole-2-carboxylic acid hydrochloride was
synthesized by Bachem on special request. DMF, acetonitrile
(HPLC grade) and 3,3'-diamino-N-methyldipropylamine were
purchased from Aldrich. N,N-diisopropylethylamine (DIEA) was
from Sigma. Dichloromethane (DCM), thiophenol (PhSH),
ethanedithiol (EDT), trifluoroacetic acid (TFA), thiodiglycol,
piperidine, N,N'-diisipropylcardodiimide (DIC),
dicyclohexyloarbodiimide (DCC) and 3-dimethylamino-1-
propylamine were from Fluka. FLUOS (5(6)-carboxy-fluorescein-
N-hydroxysuccinimide ester) was purchased from Boehringer-
Mannheim. All reagents were used without further purification.
Glass peptide synthesis reaction vessels (5 ml) with a # 2
sintered glass filter frit were obtained from Verrerie Carouge
(Geneva, Switzerland). Analytical and semi-preparatory HPLC
was performed as previously described (Baird and Dervan,
1996). Electrospray Ionization mass spectra were obtained in
the positive ion mode on a Trio 2000 instrument at the
University Medical Center (Geneva, Switzerland).
Syntheses of pyrroles for solid phase synthesis.
1,2,3-Benzotriazole-1-y1 4-[tert-Butoxycarbonyl)amino]-1-
methylpyrrole-2-carboxylate or Boc-Py-Obt was synthesized from
4-amino-1-methylpyrrole-2-carboxylic acid methylester
hydrochloride (Baird and Dervan, 1996).
Manual Solid phase synthesis of pyrrole compounds .
Couplings of Boc-Pyrrole were performed as previously
described (Baird and Dervan, 1996). Boc deprotections were
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
42
carried out with 90% TFA, 5% EDT and 5% PhSH (2x 30 s, 1 x 20
min). Cleavage from the resin with 3-dimethylamino-1-
propylamine or 3,3'-diamino-N-methyldipropylamine was
performed as described (Baird and Dervan, 1996). After
cleavage, most of the excess organic base was removed prior to
HPLC purification by precipitation of ~pyrrolic peptides. For
this purpose, the reaction mixture was mixed with 3-4 volumes
of DCM, followed by the addition of 10 volumes of cold (-20
C) petroleum ether. The precipitated product was collected by
centrifugation and dissolved in 1% TFA to obtain acidic pH.
Fluorescein-labeling of compounds.
Oligopyrroles with a unique primary amine were obtained by
either cleavage of oligopeptides from solid phase with a
diamine (3,3'-diamino-N-methyldipropylamine) or deprotection
of an N-terninal y-aminobutyric acid spacer. The N-hydroxy
succinimide active ester of fluorescein was added in 3 fold
excess together with 6 or more equivalents of DIEA. Reactions
were allowed to proceed at room temperature for 15 minutes and
the fluorescein labeled oligopeptide was purified by HPLC.
Synthesis of P31 and P31T
P31 (Im-(3-Im-Py-(3-Im-~i-Im-(3-Dp) was synthesized in a
stepwise fashion by manual solid-phase synthesis from Boc-(3-
PAM resin as previously described for Tmidazole and Pyrrole
containing hairpin polyamides (Baird and Dervan, 1996).
Since acylation of the imidazole amine on solid phase gives
unsatisfactory results, Boc-[3-alanine couplings were
performed by preparing a Boc-(3-Im-OH dimer in solution. The
synthesis and activation was as described for dimers of Boc-
y-aminobutyric acid and Imidazole (Baird and Dervan, 1996).
For fluorescent labeling of P31, cleavage from the solid
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
43
support was performed with 3,3'-diamino-N-
methyldipropylamine. After HPLC purification, the C-terminal
amine was acylated using an commercially available
(Molecular Probes) N-hydroxy succinimide active ester of
Texas red. The resulting compound was then again purified by
HPLC.
Preparation of probes for DNase.I footprinting.
Synthetic oligonucleotides
GATCTAGACGCATATTAATTGCGCTGTCGACGCATTAGTG and
GATCCACTAATGCGTCGACAGCGCAATTAATATGCGTCTA , were hybridized to
obtain the W9 probe, oligomerized by ligation and digested
with BamHl and BgIII to obtain different tandem repeats . The
following oligonucleotides were prepared identically: GAF31 is
composed of the oligonucleotides
GATCCTCAGAGAGAGCGCAAGAGCGTCCCGGGAGAAGAGAAGAGAGTA and
GATCTACTCTCTTCTCTTCTCCCGGGACGCTCTTGCGCTCTCTCTGAG and BrownI of
oligonucleotides
GATCCAAGAGAAGAGAAGAGAAGAGAAGAGTACTTATTAACACAACACA and
GATCTTGTGTTGTGTTAATAAGTACTCTTCTCTTCTCTTCTCTTCTCTTG. Fragments
were purified on low-melt agarose gels and then cloned into a
modified pSP64 vector, cut by BamHI and BglII. End-labeling
was carried out following digestion with HindIII and a fill-in
reaction with Klenow DNA polymerase. The labeled plasmid was
cut with PvuII and the target fragments purified from low-
melting agarose gels. The 657 by EcoR1/Hinf1 fragment of the
Drosophila histone SAR was cloned into the SmaI site of the
modified pSP64 plasmid. This SAR probe was end-labeled
following digestion with EcoRl, then cut with ClaI and the
resulting 347 by fragment purified from low-melting agarose
gels.
DNase I footprinting.
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
44
All reactions were performed in a total volume of 40 ~,1. A
polyamide stock solution or buffer (for reference lanes) was
added to an assay buffer containing 20 kcpm radiolabeled DNA,
affording final concentrations of 10 mM Tris-HC1 (pH 7.4) , ZO
mM KC1, 10 mM MgCl, 5 mM CaCl2, 0.5 mM EDTA, 0.5 mM EGTA, 1 mM
DTT and 0.1% digitonine. The solutions were allowed to
equilibrate for at least 2 h at room temperature. Footprinting
reactions were initiated by the addition of 2 ~,1 of a DNase
stock solution (containing 100 pg DNase T in buffer) and
allowed to proceed for 2 min at room temperature. The
reactions were stopped by addition of .10 ~,l of a solution
containing 1.25 M NaCl, 100 mM EDTA. Next, 5 ~.1 of a 1% SDS
solution was added, followed by 2 ~.l of a solution containing
1 ~.g poly(dA-dT), 1 ~,g salmon sperm DNA and 10 ~g glycogen and
the DNA was ethanol precipitated (20 min at -20 C). The
reactions were resuspended in 4 ~1 of 80% formamide loading
buffer, denatured 10 min at 85 C, cooled on ice and
electrophoresed on 8o polyacrylamide denaturing gels (5%
cross-link, 8 M urea) at 30 W for 1h. The gels were dried and
exposed o/n at -70 C.
Staining of Drosophila nuclei.
Kc Drosophila nuclei were isolated (Mirkovitch et
al., 1984), diluted into XBE (10 mM Hepes, pH 7.7, 2 mM
MgCl~, 0.1 mM CaCl2, 100 mM KC1, 5 mM EGTA and 50 mM
sucrose), fixed with 0.8% fresh paraformaldehyde for 15
minutes and spun onto a round coverslip (10 mm) as described
previously (Boy de la Tour and Laemmli, 1988). For washing
and staining, coverslips were floated on 60 ~,1 drops of XBE
deposited on parafilms. After centrifugation coverslips were
washed twice (1 minute), stained for 60 minutes, washed four
times (1 minute) and then mounted in PPDI (5 mM Hepes pH
7 . 8 , 10 0 mM NaCl , 2 0 mM KCl , 1 mM EGTA, 10 mM Mg S04 , 2 mM
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
CaCl2, 78% glycerol, 1 mgr/ml paraphenylene diamine). Figure
(4) panel A was stained with 0.5 ~.M P9F and 15 ~M ethidium
bromide (EB).
Other methods
Topoisomerase II inhibition and chromosome assembly were as
described previously (Girard et al., 1998; Strick and Laemmli,
1995). Affinity cleavage experiments was performed as
described elsewhere (Turner et al., 1997).
Section II . Specific Gain and Loss of Function phenotypes
induced by Satellite-specific DNA-binding Drugs fed to
Drosophila
Position effect variegation (PEV) is an epigenetic
.gene inactivation phenomenon discovered by Muller (Muller,
1930) arising from chromosomal arrangements that juxtapose
euchromatic genes to heterochromatin. Heterochromatin-
mediated gene silencing is heritable, epigenetic event that
involves no alterations in DNA sequence but instead is due
to heritable changes in chromatin structure.
A classical example of PEV is the Drosophila
melanogaster allele white-mottled (w"'4),which arises from a
large inversion that juxtaposes the white gene close to the
heterochromatin of the X chromosome. The variegated
phenotype of the eyes of c~"' flies is noted as red, clonally-
derived patches of transcriptionally active cells in an
otherwise white colored background where the white gene is
silenced (Elgin, 1996; Karpen, 1994; Wakimoto, 1998). A well
studied, different case of PEV concerns the brown (bw+) gene
of Drosophila melanogaster (reviewed by (Henikoff and Comai,
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
46
1998). Insertion of a large unit (about 1.7 megabases) of
heterochromatic GAGAA satellite repeats into the coding
region of this gene causes the brOwnD~minant (bye) phenotype
(Henikoff et al., 1995). This dominant allele becomes
manifested in heterozygous flies (bcn~/+), where the
heterochromatic insertion inactivates in trans (trans-
inactivation) the paired wild type copy of brown (bw+) . The
phenotype of bw+/bw° flies is observed as pale brown,
variegated eyes due to lack of the pteridine pigment
(Henikoff and Dreesen, 1989).
Various molecular models have been discussed to explain
PEV. It is often proposed, in the context of white-mottled,
that epigenetic inactivation results from spreading of the
heterochromatic state along the chromosomes (Locke et al.,
1988; Tartof et al., 1989). Other models, largely derived
from the brown-dominant studies, suggest that gene silencing
is due to local or long-range pairing of heterochromatin
and/or altered positioning in the nucleus (Csink and
Henikoff, 1996; Csink and Henikoff, 1998; Platero et al.,
1998 ) .
Genetic studies in Drosophila melanogaster led to
the identification of numerous trans-acting factors
implicated in PEV (Elgin and Jackson, 1997) . Although these
studies allow a better understanding of the biochemistry of
heterochromatin, very little is known in higher organism
about cis-acting DNA motifs implicated in PEV. It has been
suggested that the heterochromatic chromatin state may
simply be a consequence of tandem sequence repetition
(reviewed by Henikoff, 1998 #97). This suggestion is linked
to a phenomenon termed repeat-induced silencing (RIGS) that
describes gene silencing caused by tandem gene repetition.
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
47
RIGS was first described in Arabidopsis thaliana (Assaad et
al., 1993) but is also observed in transgene arrays of mice
and flies (borer and Henikoff, 1994; borer and Henikoff,
1997; Garrick et al., 1998).
Eukaryotic genomes contain a vast amount of 'non-
genic' DNA (e.g. satellite) and PEV provides a rare but
very limited experimental opportunity to address the role of
this DNA fraction. In view of the difficulties in dissecting
and assigning functions to large fractions of the genome, we
consider the development of new approaches and tools of
major importance. One approach that we successfully applied
is the synthesis of artificial DNA sequence specific
inhibitors (Girard et al., 1998; Strick and Laemmli, 1995).
We were led along this path in attempts to dissect the
biological role of scaffold (or matrix) associated regions,
called SARs or MARS. These elements are intriguing, since
they appear to mediate their biological effect by unusual
('nonconformist') mechanisms which may well be
representative of the ways non-genic DNA elements implement
their functions (Laemmli et al., 1992). The sequence
hallmark of SARs are numerous AT-tracts (short sequences of
A and T bases) that are generally separated by short, mixed
sequence spacers, resulting in clustered AT-tracts. In
contrast to standard cis-acting DNA elements, the specific
interactions of SARs are not mediated by precise base
sequence, but by structural DNA features, such as the narrow
minor groove of the AT-tracts and possibly bends (Adachi et
al., 1989; Bode et al., 1992; Kas and Laemmli, 1992).
Based largely on techniques developed by Dervan and
collaborators (Dervan and Burli, 1999), compounds that
target different DNA satellites of Drosophila melanogaster
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
48
were synthesized and characterized (Janssen et al., 2000).
When fed to developing Drosophila melanogaster, it was
observed that these satellite-specific drugs can lead to
defined gain or loss of function phenotypes. Details are set
out below
Example 4
Suppression of PEV of white-mottled flies by oligopyrrole P9
Two monomeric satellite-specific compounds, termed
P9 and P31, were synthesized and characterized as describes
in the previous examples . P9 (sequence = PyPyPy-(3-PyPyPy-(3-Dp
where Py - N-methylpyrrole, . [3 - (3-alanine and Dp -
dimethylaminopropylamide) was found to bind AT-tracts of 9
Ws (A or T bases) with subnanomolar affinity. We
demonstrated further by staining of Kc nuclei and polytene
chromosomes using fluorescently tagged P9, that this
compound predominantly binds satellite I and III. Satellite
I is composed of short AATAT repeats and is therefore a high
affinity target for P9. Similarly, due to the numerous AT-
tracts in the 359-by unit of the SAR-like satellite III,
specific binding of P9 to this repeat was expected. Compound
P31 (sequence - Im-(3-Im-Py-(3-Im-~i-Im-(3-Dp where Im - N-
methylimidazole) binds two consecutive GAGAA repeats of
satellite V. Again, targeting of P31 was confirmed by
epifluorescence using a fluorescently labeled compound for
staining of nuclei and polytene chromosomes. Both compounds,
P9 and P31 were found to bind their targets with
subnanomolar affinity in a 1:1 drug to DNA complex by
hydrogen bonding schemes as proposed in Figure (6A, B).
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
49
We established further that titration of the AT-tracts of
satellite III with 'P9 resulted in the opening of this
heterochromatin block, as revealed by a facilitated cleavage
with restriction enzymes or topoisomerase II at
internucleosomal linkers (Janssen et al., 2000). Similar
observations were made previously with distamycin (Kas and
Laemmli, 1992). Furthermore, we observed that P31 was
considerably less potent in opening satellite IIT as
compared to P9, presumably due to its low affinity for AT-
tracts. In summary, we synthesized two satellite-specific
polyamides of similar molecular weight,, binding affinity,
target site size and interaction mode.
Satellite III is the major component of centric
heterochromatin of chromosome X and PEV of white-mottled
Drosophila flies is due to an inversion that juxtaposes the
white gene (allele w"'4) to this centric heterochromatin.
Since P9 opens satellite III we asked whether this drug
could affect PEV of white-mottled Drosophila flies.
The eye phenotype of c~"' flies is quite heterogeneous. About
650 of the eyes are strongly variegated, ranging from quasi-
white with little pigment to those containing a generally
white background and a number of red patches (defined as the
white-mottled class). The remaining flies have eyes with a
darker appearance with often larger red patches in an orange
background (red-mottled class). Given the phenotypic
heterogeneity of w'"4 flies, we carried out our experiments on
a relatively large scale by mixing these compounds directly
with semi-synthetic fly medium. Vials were prepared
containing a final concentration of either 100 ~.M P9, P31 or
no compound. Equal numbers of 5 to 10 days old w'"' flies were
allowed to lay eggs for 36 hours. The parents were then
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
removed and progeny development was allowed to proceed at a
constant temperature of 18 °C. We observed no significant
toxicity of P9 and P31 when fed to developing t~t"4 or wild-
type flies (Canton S). The timing of the developmental
stages was also normal and generally, around 90 to 160 flies
hatched per vial in this experiment (Table 2).
For eye phenotype analysis, young flies born on the same day
were transferred to a drug-free vial and scored .5 days
thereafter. Eye phenotypes were categorized into the two
classes defined as white-mottled or .red-mottled, after
examination under a dissection microscope. This analysis
identified P9 as a strong suppressor of PEV. Where in the
absence of the drug, 620 of the eyes were scored as white-
mottled, only 11o retained this phenotype for the flies
raised in the presence of P9. About 90% of the flies had a
red-mottled to red eye phenotype upon P9 treatment (Figure
7, compare A and B, Table 2). In contrast, P31 had no effect
on the rn~"4 eye phenotype, since 64% of the flies remained
white-mottled (Figure 7C, Table 2).
To quantify these results further, the red-eye pigments were
extracted from 30 heads of males and their relative
concentration determined by spectrometry. Figure (8) shows
that the red pigment level corresponding to flies treated
with P9 is about 3 times higher than that of the control
flies (no drug) or flies fed with P31. Hence, P9 very
markedly and specifically suppresses PEV of c~"' restoring the
red pigment level to about 50% of wild type flies (Figure
3).
Furthermore, we found that P9 did not suppress PEV of a
variegating white reporter transgene inserted at the basis
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
51
of chromosome 2T, (data not shown), indicating that
suppression of PEV by P9 is very unlikely the result of a
direct interaction of P9 with the promoter of the white gene
(see Discussion).
Taken together, these results strongly suggest that
suppression of PEV by P9 (not P31) is mediated by specific
chromatin opening resulting from titration of AT-tracts
which in turn reduces silencing of the rearranged white
gene.
Example 5
P31 (not P9) causes a developmental delay in brown-dominant
flies
The P9-induced suppression of the white-mottled phenotype is
a remarkable result. It encouraged us to examine a different
PEV phenomenon concerning the .brown (bw+) gene . In contrast
to white-mottled, where the proximity of heterochromatin
brings about cis-inactivation, PEV mediated by the bcnf
allele occurs by traps-inactivation (Henikoff and Comai,
1998) . In bt~', the brown gene contains an insertion of a
large unit of heterochromatic GAGAA satellite repeats in its
coding region. This dominant allele becomes manifested in
heterozygous flies (bw+/bw°) where the heterochromatic
insertion is traps-inactivating the paired wild type copy
(bw+) of .brown.
It was demonstrated by footprinting techniques and
immunofluorescence the impressive specificity of P31 for
GAGAA repeats of bc~ (also referred to as br,~ repeats) . It
was therefore of great interest to test whether P31 would
affect the eye phenotype of heterozygous bw° flies. Here, P9
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
52
serves as the control compound since it does not bind GAGAA
repeats. The experiment is similar to that described above
for w"'4. Vials with semi-synthetic fly food medium were
prepared containing a final concentration of 100 ~.M P31, P9
or no compound. Egg laying was allowed to proceed for 36
hours by homozygous (brvn/ .hr~) females that were crossed with
scarlet (st/st) males. The progeny from this cross is
heterozygous for the brown locus and homozygous for scarlet
(bcnf/+; st/st). In a scarlet background, a modification of
the bc~' eye color is much easier to observe (Talbert et al . ,
1994) .
In the control experiment, the eye color phenotype of the
heterozygous progeny exposed to P9 was pale (light-yellow)
and indistinguishable from that of the no-drug control as
judged visually or by determination of eye pigment
concentration (Figures 9A). As observed in the w"'4 experiment,
we noted no toxicity or alteration in the timing of
developmental stages upon treatment of b~ flies with P9
(Table 3). This contrasts markedly with the results obtained
with P31. This compound not only severely affected the
timing of the fly developmental stages, but it also
dramatically reduced the viability of the resulting progeny
(Table 3). Although a roughly normal timing was noted for
the appearance of the second instar larvae, we observed a
serious delay of 65-75 hours in both pupation and hatching.
Despite this delay, about a normal number of progeny
hatched, but these new born flies appeared feeble and often
drowned in the fly flood.
Examination of the eye phenotypes ofthe P31 treated progeny
revealed that they contain a larger number of spots with red
ommatidia as compared to P9 or untreated flies (Figure 9A
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
53
and B) . This is also reflected by a slight increase in the
eye pigment concentration (Figure 8). The small, but
reproducible, increase in eye pigmentation may reflect their
more advanced age (hatching was delayed by 65-75 hours)
rather than a genuine increase of bw+ function. It is well
known that eye pigmentation augments following birth, and
e.g. untreated flies of the (bw°/+; st/st) genotype are
white at birth but pale yellow (slightly pigmented) at day
two. In contrast, P31 treated flies have at birth a level of
eye pigmentation that roughly corresponds to that of the
controls at day two to three.
In summary, treating heterozygous bra' flies with P31 does
not overcome t:rans-inactivation to a significant extent. The
eye phenotype remains pale and strongly variegated. But
interestingly, P31 (not P9) induces a serious developmental
delay of over 2 days and yields a progeny of very feeble
flies, that was not observed upon treatments of Canton S or
rt"4 flies.
The striking dependence of the P31-induced developmental
delay on the bcvn allele, suggested that a direct interaction
between P31 and the GAGAA insert somehow causes a
developmental defect(s). Consistent with this idea, we found
that the effect of the bra insert is quantitative, since we
found that the effect of P31 on the developmental delay was
more severe in homozygous than in heterozygous brn~ flies. We
observed that about two thirds of the progeny died at the
pupal stage and that only a few feeble flies hatched with a
delay of about 65 to 75 hours. These experiments established
a remarkable molecular interplay between the GAGAA insert of
the bc~ allele and its target compound P31.
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
54
Example 6
In brown-dominant flies, P31 mimics the Trithorax-like
allele Trllsc
Homeotic transformation of A6 to A5: Surprisingly,
closer inspection of the P31 progeny revealed a pronounced
transformation of the abdominal segment 6 (A6) into A5
(Figure 9C). This homeotic transformation is manifested by
bristles on the ventral A6 which, in contrast to A5, is
otherwise bristle-free in males. Bristle.induction was found
in over 90% of the P31 bw° /+ progeny (Table 3), each
displaying between 1 and 10 bristles on A6 (mean = 3.75). In
contrast, no bristles were observed on A6 in the P9 treated,
control males. Importantly, the transformation of A6 to A5
requires both P31 and the bw° genotype, since bristles were
never found on segment A6 in flies that are wild type at the
brown locus (e.g.uf"4, see Table 3).
As mentioned above, bw° homozygosity increased the
developmental delay mediated by P31. This observation also
extends to the A6 to A5 transformation. While about 10% of
heterozygous progeny exposed to P31 lacked bristles on A6,
all 40 homozygous bw° males scored had between 3 to 10
bristles (mean value - 6, Table 3). Hence, this homeotic
transformation and the developmental delay require at least
one b~' allele and both have a greater penetration in the
bc~ homozygous progeny. This observation. further reinforces
a molecular link between the bc,~ allele and sensitivity to
P31 .
A transformation of A6 into A5 is characteristic of certain
loss-of-function alleles of the Abd-B gene but is also
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
observed for mutations in genes belonging to the Trithorax-
group (Kennison et al., 1998). In particular, the A6 to A5
transformation was previously also observed for the Trllsc
allele of the Trithorax-like gene which encodes the GAGA
factor (GAF, (Farkas et al., 1994)). Tr113c has a P-element
insertion in the first intron of the Trl gene which appears
to result in a moderately reduced GAF protein level (Bhat et
al., 1996; Farkas et al., 1994). Two presumptive null
mutations (TrlR6', TrlRes ) of the GAF gene led to lethality at
late larval stages (Farkas et al., 1994). It has been
suggested that lethality may be late since a large maternal
deposition of GAF in the egg allows developmental
progression up to larval stages (Bhat et al., 1996).
Example 7
Sex-comb reduced phenotype: The phenotypic parallels (A6 to
A5 transformation) observed between Tr113c flies and bw°
flies raised in the presence of P31 suggested that this drug
somehow interferes with GAF function. Hence we examined the
chemical mimicry by P31 of additional Tr113c phenotypes.
Several of the trithorax-Group (trx-G) genes are known to be
implicated in the expression of the sex comb reduced (Scr)
gene (Kennison et al., 1998). We tested whether the Tr1 gene
might also be involved. For this, we used the Tr113callele,
which occasionally gives rise to viable homozygous flies
with no other described additional phenotype (Farkas et al.,
1994) . We scored homozygous TrIl3c males for the number of
teeth per sex comb and observed that they were significantly
reduced compared to wild-type flies (Table 3). While on
average, about 11 teeth per sex comb are found in Canton S.
flies and Oregon R., we measured a mean number of 8.6 in
homozygous Tr113c flies and also found that about one third
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
56
had only 7 teeth per sex comb . Furthermore, to test if P31
would also mimic the sex comb reduced phenotype in
homo/heterozygous bw° flies, we dissected male legs after
treatment of these flies with P31 or P9 (Table 3) . Our data
show that heterozygous bwn progeny raised in the presence of
P31 had mostly 7 to 8 (mean =8.1) sex comb teeth and that
this distribution was often shifted to lower numbers, mostly
6 (mean = 7.2) in the homozygous case (Figure 10 and Table
3). Thus, this phenotype depends on the bcnf allele and is
enhanced by homozygosity. No teeth reduction was observed
with w"'4 flies (Table3) or other fly stocks (see below)
raised in the presence of drugs.
In summary, P31 treatments of br,,~ flies mimic the A6 to A5
homeotic transformation and the Scr phenotypes of Trhac
mutants. Moreover, homozygous Trhsc/Tr113c flies are known to
display a rough-eye phenotype (Farkas et al., 1994), we also
observed a slight roughening of the eye after treatment with
P31 in bw° progeny (data not shown).
P31 enhances haltere-to-wing transformation in heterozygous
bcap/bw; Ubxl/+ flies: The GAGA factor is know to bind to the
promoter of Ubx and to stimulate its transcription in vitro
(Biggin and Tjian, 1988). The Trl locus also genetically
interacts with Ubx, since Tr113c dominantly enhances the
segmental transformation observed in Ubx heterozygotes. That
is, flies doubly heterozygous for Trl and Ubx, possess
halteres that are further transformed to wing-like
structures than Ubx/+ flies (Farkas et al., 1994). We asked,
does P31 also chemically mimic the genetic interaction of
Ubx and Tr113c~ To address this issue, we fed P31 to two
genotypes, +/+; Ubx/+ and b~/bu~;Ubxl/+. The latter genotype
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
57
was obtained by crossing b~/bwp~ st/st females and bc~/br~
Ubx'~/Tm3, sb males.
The progeny of these two genotypes raised in the presence of
P31 were scored both for the A6 to A5 and the haltere-to-
wing transformations as well as for the sex comb reduced
phenotypes (Table 3). We observed no altered phenotype in
the progeny of Ub.~/+ flies upon treatment with P31 (Figure
11C, Table 3). In contrast; severe alterations of the
phenotypes were noted in brn~"/bw°; Ubx'/+, flies . P31 blocked
their development completely, often before the pupation
stage, yielding no newborn flies (Table 3). However, eight
of them (out of approximately 100) reached a late pupae
stage and thus could be dissected to inspect their
morphology. All of these pupae had a strong A6 to A5
homeotic transformation with 2 to 6 bristles (Table 2) and a
significant reduction of sex comb teeth (Figure 5).
Furthermore, five of eight pupae that were heterozygous for
Ub~ and homozygous for br,,~ (bwD/bwD; Ub~/+) displayed clear
signs of haltere-to-wing transformations (Figure 11). This
transformation was manifested by an enlarged size and gray
color of the halteres, and also by the appearance of
numerous bristles at the base of the halteres that are
characteristic of the anterior wing margin. Hence, the b~
allele, in combination with the drug P31, enhance the Ubx
phenotype. We conclude that the chemical mimicry by P31 in a
bcvn genetic backgrounds extends to the genetic interaction
of Tr113c with Ubx.
Example 8
Massive opening of the GAGAA satellite insert with P31
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
58
GAF binds GA-rich sequences (Lu et al., 1993). Although
there is considerable variability in the binding sequences
of GAF, NMR studies identified the pentameric sequence GAGAG
as its optimal consensus, where single base pair mutations
except the central G have only moderate effects on GAF
binding (Omichinski et al., 1997). Recent studies showed
that a single trinucleotide repeat GAG is often sufficient
to define a specific GAF interaction (Wilkins and Lis, 1998)
and that this protein can bind multiple binding sites
cooperatively (Katsani et al., 1999)
To test whether P31 binds the GAF consensus sequence
(GAGAG), we performed footprinting analysis using a DNA
probe that includes, besides 2 consecutive bw° repeats
(GAGAAGAGAA), a high affinity binding site for GAF
corresponding to the Ubx promoter sequence (Biggin and
Tjian, 1988). Inspection of the footprint data showed that,
although P31 protects the brn~ repeats at a 0.1 nM
concentration, no binding is noted at the Ubx GAF site with
a ligand concentration up to 25 nM (see Figure 7A, (Janssen
et al. , 2000) . We also studied the interaction of GAF with
this same DNA probe and observed that this protein, in
contrast to P31, protects both the bw° repeats and the Ubx
site (Figure 2B, (Janssen et al., 2000))
Database searches indicated that promoter proximal GAF
binding sites are normally not composed of bwn repeats, but
are either defined by the GAGAG consensus or by multiple GAG
motifs . Since these sites are poor targets for P31, it can
be concluded that this compound is not likely. to interfere
directly with gene regulation. The observation that P31 is
active only in a bcv" genetic background strongly supports
this suggestion.
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
59
What is then the molecular link between the chemically
induced P31 phenotype and the bra satellite insert?
Immunofluorescence studies demonstrated that GAF shuttles
during the cell cycle between euchromatin and
heterochromatic binding sites. In mitosis, GAF is bound to
the heterochromatic AG-rich satellites of metaphase
chromosomes (Platero et al., 1998; Raff et al., 1994). In
contrast, GAF appears to be actively excluded from the
heterochromatic chromocenter of interphase polytene
chromosomes but instead is bound to hundreds of sites along
the euchromatic arms (Raff et al., 1994; Tsukiyama et al.,
1994). The lack of GAF staining at the chromocenter could
possibly be due to a detection problem and to the selective
under-replication of this heterochromatic satellite (we were
unable to detect a br,~ satellite signal in the chromocenter
with fluorescent P31). But detailed immunofluorescence
studies by (Platero et al., 1998) demonstrated that GAF is
indeed bound to euchromatin but not to heterochromatin in
interphase. In contrast to interphase nuclei, intense GAF
staining was observed on heterochromatin GAGAA satellites of
mitotic chromosomes. Hence, GAF is clearly heterochromatin
bound in mitosis and is then redistributed to numerous
euchromatin loci during interphase.
It was demonstrated that P9 opens the heterochromatic
satellite III as measured by a massive stimulation of
cleavage by topoisomerase II (Janssen et al., 2000). We
argued that perhaps P31 may open the bw° insert as to render
this heterochromatin more accessible for GAF binding. To
address this question, we incubated permeabilized bcvn
salivary glands either with P31, P9 or no drug and then
triply stained the polytene . nuclei for GAF (by
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
immunofluorescence), total DNA (with DAPI) and for the br,~
insert (with P31T). Stained polytene glands were gently
mounted for microscopy without squashing in order examine a
large number of nuclei.
Figure (12) shows polytene nuclei, incubated without
compound or with P9 (top row). The thick polytene arms
(blue) of intact nuclei are not spread but the euchromatic
positions of GAF are observed as sharp, green bands. In red,
the position of the bc,~ insert is highlighted with P31T as
single spot. Careful examination revealed that no GAF
staining occurs at the position of the P31T signal which
marks bcvn. This is particularly evident in the black and
white inserts since no green GAF signal is overlapping with
that of the red P31T (Figure 12, top row). The staining
pattern observed following exposure of salivary glands to P9
is identical to that obtained without drug. These
micrographs confirm observations reported by (Platero et
al., 1998), who noted no GAF signal over the b~' insert.
Evidently, P9 does not alter this pattern.
In contrast, incubation of permeabilized polytene glands
with P31 leads to a dramatic GAF redistribution, manifested
by a massive co-localization of GAF (green) and P31T (red)
(Figure 7, bottom row). Again, this is best observed in the
black and white inserts, where the co-localization of the
green GAF and red P31F is evident. Examination of these
nuclei further show that GAF staining at euchromatic sites
appears considerably weaker; fewer sites are observed and
the intensity of the signal is reduced. This observation
strongly suggest that P31 (not P9) opens up the
heterochromatic bwn satellite to allow binding of GAF
resulting in a massive redistribution of GAF from
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
61
euchromatic sites to heterochromatic sites. It is therefore
reasonable to suggest that a similar redistribution occurs
in P31 treated bra flies, where a reduced availability of
GAF for euchromatic function leads to a 'chemical' gene
dosage effect of GAF, as manifested by the observed homeotic
transformations.
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
62
Discussion of Section II
The experimental potential of sequence-specific
polyamides as tools to better define DNA sequence motifs
implicated in PEV has been explored. This epigenetic
phenomenon arises from a stochastic gene inactivation either
mediated in cis or trans by large blocks of satellite
heterochromatin. Two satellite-specific DNA binding drugs
were synthesized and fed to developing Drosophila
melanogaster flies that display PEV phenotypes. Remarkably,
this led to a gain or loss of function, depending on the
drug used and the genetic fly background. Most satisfactory
is the reciprocity of the experimental observations made.
While polyamide drug P9 (not P31) suppressed PEV of white-
mottled flies (gain of function), P31 (not P9) mediated
homeotic transformations (loss of function) in brown-
dominant flies. Both phenomena are in molecular terms
explained by chromatin opening of drug-targeted DNA
satellites.
Compounds The satellite-specific, DNA minor groove binding
drugs used in this study were characterized in detail in
Section I. Briefly, compound, P9, is a pyrrole hexamer that
targets AT-tracts of 9 (or more) Ws and, P31, is composed of
both imidazole and pyrrole units and binds two consecutive
GAGAA repeats. Both compounds possess subnanomolar affinity
for their DNA target sequence and both were found to bind as
1:1 (drug to DNA) complexes. The main nuclear targets of
these compounds where directly revealed using epifluorescent
microcopy by staining isolated Kc nuclei and polytene
chromosomes. with the fluorescent derivatives (P9F and P31T).
Both dyes conspicuously marked separate foci in Kc nuclei.
P9F targets primarily satellite I and III and P31T targets
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
63
satellite V which is composed of GAGAA repeats (see Figures
4 & 5, (Janssen et al., 2000)).
White-mottled We demonstrated here that feeding oligopyrrole
P9 to developing ~"4 flies significantly suppressed PEV in
the resulting progeny. In contrast, P31 and 2 other
compounds had no activity on this fly line. This conclusion
is based on a statistical and visual analysis of the eye
phenotypes and quantitative measurements of the pigment
level obtained by extraction of isolated eyes (Figure 2).
This pharmacological experiment was carried out on a
relatively large scale, with 90 or more flies per treated
progeny, and was found to be highly reproducible.
PEV of white-mottled (w"'4) flies arises from a large
inversion that juxtaposes the white gene close to the
heterochromatin of the X chromosome. The major component of
this centric chromatin is satellite III, which operationally
behaves like a SAR (scaffold associated region (Kas and
Laemmli, 1992)), whose sequence hallmarks are clustered,
variably sized, generally long AT-tracts. The 359-by repeat
unit of satellite III contains 10 AT-tracts of 7 or more W
bases (average 9.4), accommodated in two phased nucleosomes.
We demonstrate that titration of these AT-tracts in vitro
with the oligopyrrole P9 unfolds satellite III (Janssen et
al., 2000). This is manifested by a strongly facilitated
cleavage by topoisomerase II. Cleavage is known to occur in
one of the two nucleosomal linker DNA regions (Kas and
Laemmli, 1992) .
Based on these results, it is reasonable to suggest that P9,
when fed to developing flies acts similarly by opening of
satellite III and reducing the extend of spreading of the
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
64
heterochromatic state towards ~"'. Our experiments revealed a
perfect correlation between chromatin opening in vitro and
PEV suppression in flies. P31, which opened satellite III
considerably less well, did not suppress PEV of w'"° flies.
Cis-acting elements involved in PEV have not been identified
genetically. The pharmacological studies presented here and
our previous MATH20 expression experiment (Girard et al.,
1998) offer a novel approach to study this important
epigenetic phenomenon. Both series of experiments are
strongly congruent, implicating the numerous AT-tracts of
satellite III in the establishment of. a heterochromatic
state and gene silencing at w"'4.
Although we favor the possibility that P9 and MA.TH20 reduce
the spreading of heterochromatin toward w'"' through long-
range effects, we cannot completely rule out the possibility
that a local binding site near the white gene (a local SAR)
is involved. We attempted to address this issue by studying
the effect of P9 on two different variegating minrwhite
reporter genes called C3/2L and BL2/Y (Lu et al., 1996;
Wallrath and Elgin, 1995). We observed that P9 did not
suppress PEV in these stocks (data not shown). This
observation demonstrates that suppression of PEV by P9 is
not a general phenomenon, but does not rule out the
possibility that P9 may act more locally on w"'4, since the
minrwhite reporter gene, in C3/2L and BL2/Y flies , may lack
a nearby P9-responsive element, Nevertheless, we consider a
local-acting mode less likely; the enormous target size of
satellite III, 11 Mb versus 0.5 kb for a typical SAR, favors
a long-range spreading model.
The notion that PEV occurs by spreading of heterochromatin
towards w'"4 stems from morphological studies demonstrating
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
that cis-silencing correlates with a cytological change from
a euchromatin to a heterochromatin appearance (Belyaeva et
al . , 1993 ; Umbetova et al . , 1991 ) . It would therefore be of
interest to study whether this cytological change is
reversed in flies raised in the presence of P9.
Brown-dominant The brown-dominant phenotype is due to the
pairing of homologous chromosomes, one carrying the bra
allele and the other being wild type for brown. Important
genetic and cytological experiments by Henifkoff's group led
to an attractive model suggesting that the bra insert
mediates an aberrant nuclear interaction with the centric
heterochromatin, located on the same chromosome. Hence,
according to this model, the bwn insert ,tethers the paired
bw' gene into a heterochromatic environment. This unusual
localization, a heterochromatic nuclear compartment, is then
proposed to mediate silencing of the bw+ gene (Csink and
Henikoff, 1996; Talbert et al., 1994)
Our pharmacological experiments showed that feeding P9 or
P31 to developing heterozygous bx~' flies did not affect
their eye phenotype significantly. Our observations
discussed below established that feeding P31 to developing
b~/+ or bc~/bw° flies mimics the phenotypic effects of the
Tr113c allele of the Tr1 (GAF) gene. The Tr113c allele is
known to enhance (not suppress) the variegated phenotype of
w"'', but it does not affect the eye phenotype of br~/+ flies
(Sass and Henikoff, 1998). Hence, since P31 is mimicking the
Tr113c mutation (partial loss of GAF function) in developing
flies, it is not surprising that P31 does not affect the eye
phenotype of bwn/+ flies.
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
66
Surprisingly, P31 mediated a developmental. delay
and several defined homeotic transformations. All these
phenotypes depended on the presence of the bud allele and
are dependent on the dosage of bra. Homozygous Tr113c male
flies are known to homeotically transform the abdominal
segment A6 to A5. This is observed by the appearance of
bristles on A6 (Farkas et al., 1994). This transformation is
'phenocopied' by feeding P31 (not P9) to bwn heterozygous
flies and enhanced by homozygosity for bw° (Figure 4,
Table3). The same parallel exists for the sex comb reduced
phenotype. In homozygous Tr113c/Trl~3c male flies, we counted
a reduced number of teeth per sex comb. This establishes a
genetic interaction between the Scr and Tr1 genes. A similar
reduction is observed in b~/+ males raised in the presence
of P31 and the number of teeth per sex comb is further
reduced in bw°/b~ males {Figure 10, Table 3). Hence, P31
pharmacologically mimics the Tr113c allele surprisingly well,
but this is restricted to bwa flies that carry the
heterochromatic GAGAA insert. P31 feeding to other genetic
backgrounds never led to these homeotic functions.
The Tr113c carries a P-element insertion in intron 1 of the
GAF encoding gene. Detailed studies suggest that this allele
is hypomorphic. Therefore, the observed phenotypes arise
from a partial Trl loss of function, supposedly by a reduced
dose of GAF (Bhat et al., 1996; Farkas et al., 1994). The
observed pharmacological phenotypes correspond to loss of
function alleles of Abd-B and Scr. The Abd-B gene of the
bithorax complex is required for the normal development of
abdominal segments A5 through A8. The observation that
allele Trl'-sc and drug P31 affect only segment A6 implicates
the cis-acting element iab-6 as a target for GAF (Celniker
et al., 1990). Similarly, chromatin immuno-precipitation
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
67
experiments revealed that GAF is bound to the iab-6 element
(Strutt et al., 1997). No molecular information is available
concerning the Scr gene, but it is well known that mutations
in genes of the Trx-G can Lead to a Scr loss of function
phenotype.
The strong phenotypic parallels between the effect
of P31 and the hypomorphic allele Trllac suggest, that the
pharmacological mimicry by P31 arises from a reduced GAF
dose. However, we can practically rule out that P31
primarily acts directly on gene regulatory binding sites in
Abd-B and Scr. First, P31 does not interact with typical GAF
binding sites (see Figure 3, (Janssen et al., 2000)) nor is
it expected to do so. Secondly, the bx~ allele requirement
establishes that the P31 effect is molecularly mediated by
this heterochromatic insert.
GAF has a very interesting cell-cycle behavior. It
binds centric heterochromatin in metaphase and is displaced
in interphase to numerous euchromatic sites (Granok et al . ,
1995; Platero et al., 1999; Raff et al., 1994). Our
experiments with permeabilized polytene glands showed that
P31 (not P9) mediated a massive redistribution of GAF from
the euchromatic binding sites to the heterochromatic hw~
insert . The b~' insert of polytene chromosomes is devoid of
GAF without drug treatment or exposure to P9 but is highly
enriched of GAF in the presence of P31 (Figure 7). We
propose that the br,~ insert (and supposedly centric~GAGAA
repeats) in P31 treated bc~' flies, serves as a molecular
sink for GAF. This is achieved through chromatin opening
mediated by P31. We propose that the inopportune
redistribution of GAF to heterochromatin during interphase
leads to a depletion of GAF at euchromatic sites. In turn,
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
68
the reduced availability of GAF results in the observed
mimicry of the Tr113° allele, that is a partial loss of GAF
function. Although our redistribution experiments were not
carried out with living fly embryos, but with permeabilized
glands, it demonstrates that P31 can interfere with the
chromosomal distribution of GAF.
Chromatin opening by P31 of the GAGAA insert could
occur long-range, but it was of interest to determine
whether P31 and GAF can co-bind the same target sequence.
The solution structure of GAF, complexed to its GAGAG
consensus sequences revealed a modular binding mode where
its single zinc finger and a basic domain (BR2) make
contracts in the major groove, while an other basic domain
(BR1), wraps around into the minor groove (Omichinski et
al . , 1997) . Although these basic domains are required for a
high affinity interaction (Pedone et al., 1996), we observed
in competition binding experiments that P31 and GAF can co-
bind GAGAA repeats (data not shown).
Is the chromatin opening model quantitatively
reasonable? The additional fraction of GAGAA repeats in bc~
animals amounts to roughly 17 or 34% for the heterozygous or
homozygous genotypes, respectively (Platero et al., 1999).
If one assumes that P31 opens the bra' and centric repeats
similarly, then the GAF concentration available for
euchromatic function would proportionally be reduced by 17
(bwD/bw+) or 34% (bc~'/bw°) as compared to bw+ flies. We
consider it reasonable to propose that a fractional
reduction to this extent (at euchromatic sites) could affect
GAF gene function. Most genes that require GAF for
expression, such as Ubx, engrailed and hsp70, contain
multiple binding sites and GAF is known to bind such
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
69
elements highly cooperatively by oligomerization via its POZ
domain (Katsani et al., 1999). It is thus conceivable, that
a relative small change in GAF concentration could
significantly affect the activity of certain genes. If the
above mentioned model for chromatin opening and GAF
recruitment model is correct, overexpression of GAF should
then reduce the effects of P31 in bw° flies. Other molecular
scenarios could be envisaged, e.g. centric and bra
heterochromatin could be different biochemically such that
P31 only opens the brn~ insert. If that were the case, then a
much greater relative reduction of the. GAF concentration
could be achieved in .bra versus bw+ flies.
Polyamides were shown to be cell-permeant and to
inhibit the expression of targeted genes when added to the
media of tissue culture cells (Dickinson et al., 1998;
Dickinson et al., 1999; Dickinson et al., 1999; McBryant et
al., 1999). Here, we demonstrate that polyamides can affect
gene expression of an entire developing organism.. In these
experiments, flies were raised in the presence of food
containing these compounds from egg laying to hatching. Our
fly MATH20 expression data showed the suppression of PEV of
r,~"4 required its expression around 48 to 72 hours of
development when the differentiation of the eye imaginal
disc occurs (Girard et al. , 1998) . It can thus be concluded
that these compounds are chemically stable for several days
under the experimental conditions. Previous studies also
demonstrated that tetracyline added to fly food can also
regulate a transactivation system dependent on this
antibiotic (Bello et al., 1998). The remarkably unambiguous
quality of the fly phenotypes observed, combined with
progress in synthesizing sequence-specific polyamides,
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
emphasizes the utility of these chemicals as novel gene
tools.
The structure of the heterochromatic fiber is
unknown and the experiments presented do not bear on the
mechanism whereby heterochromatic satellites are opened by
DNA minor groove binding compounds. Such molecules are known
to interfere with the binding of proteins that make DNA
minor groove contacts (Dorn et al., 1992) . Hence chromatin
opening might be due to a displacement of heterochromatin-
associated proteins, such as HP1, D1 or even the linker
histone H1. The N-terminal tails of the core histones are
know to make minor groove contracts and it has been
speculated that the tail of histone H4 may be involved in
mediating the higher-order folding of the chromatin fiber
(Luger et al., 1997). DNA minor groove binding drugs could
compete for such interactions made by the histone tails and
then promote sliding of nucleosomes or unfolding the higher-
order chromatin e.g. by disrupting nucleosome-nucleosome
interactions. We do not know how long-range spreading of an
open chromatin structure is mediated. It has often been
discussed that heterochromatinization arises by
cooperatively interacting components which 'polymerize'
along the chromatin fiber where the extent of spreading is
governed my mass action. It is easy to see that disruption
or displacement of units from this 'polymer' will
energetically disfavor spreading.
The findings reported here also emphasize that
chromatin accessibility may not only be regulated by
sophisticated large machines but may be constitutive, that
is, determined by the intrinsic property of given chromatin
sections to breath (opening/closing) and the general
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
71
availability of factors that compete by mass action for
chromatin opening or closing. It is interesting to speculate
that evolution may have positioned of chromatin sections
that 'breath easily' (e. g. SARs) adjacent to gene regulatory
sequences so as to facilitate constitutive accessibility.
Numerous ATP-driven chromatin remodeling complexes
have recently been described which facilitate the binding of
factors involved in gene expression or, conversely, promote
the assembly of repressed heterochromatin (Tyler and
ICadonaga, 1999). Such activities are thought to catalyze
nucleosome mobility (sliding) or the disruption of
nucleosome structure so as to enhance access of DNA-binding
factors (experimentally, nucleases) to DNA packaged into
chromatin. Chromatin remodeling complexes are very large and
composed of several protein subunits. Here we demonstrate
that small molecules can serve in flies as heterochromatin
remodeling activities. DNA satellites are composed of very
large blocks and are therefore relatively easily targeted
with sequence-specific compounds. It would be of great
interest to explore further whether polyamides can be used
as activators of epigenetically silenced genes. Gene
silencing is a major problem of most genetic manipulations
such as gene therapy, constitutive or regulated expression
of genes introduced into plants, animals and microorganism.
It might be possible to revert or prevent epigenetic
silencing by targeting high affinity polyamides to natural
or 'synthetic' cis-acting elements of gene expression
vectors.
Experimental procedures for Section II
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
72
Polyamide treatment of flies and determination of the aye
phenotypes
Vials for egg laying containing polyamides were prepared as
follows: 250 nmoles of P9, P31 was disolved in 150,1 ddH20
and then mixed with 2.35 g of semi-synthetic fly media pre-
heated at 60° C. These vials contained a concentration 100
~tM of compound in final. After cooling down to room
temperature, - 300 w'"4 flies (age 5-10 days) were added to
the vial for egg laying. Eggs were collected for a period of
36 hours at 18° C, mother flies were then removed and fly
development was allowed to proceed at a constant temperature
(18° C ~ 1). New-born flies of the same day were transferred
to a fresh vial and scored after 5 days for eye phenotypes
(Girard et al., 1998). For visual inspection of eye-
phenotypes (Tablel), both female and male flies were scored.
For optical density measurements of the eye-pigments, 30
heads of males were selected, pigments extracted and
measured as described (Hazelrigg et al., 1984).
Drug treatments of the bc~ flies were performed as for w"'4
flies except that 30 parental females were allowed to lay
eggs. To obtain the heterozygous bc,~/+; st/st progeny, 30
virgin bw°/b~; st/st homozygous females were mated with
scarlet homozygous males for about 5 days before
transferring them to experimental vials for egg laying.
Development delay, lethality and homeotic transformations
The lethality of the progenies was calculated by
counting the dead bodies and the total empty pupal shells .
The A6 to A5 transformation was detected by bristles that
form on abdominal segment A6 on males. All male adult flies
obtained (~ 50) were scored. Sex combs and halteres were
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
73
individually inspected and photographed under the microscope
(Axiophot, ~eiss) after dissecting at least 30 individuals
and incubating them into 30 ~,1 of 0.6g/ml gum arabic; 4g/ml
chlorohydrate; 35% glycerol under a 24x24mm coverslip as
described (Ashburner, 1989). Since raising double
heterozygous flies (Ubx'/+ ; bwn/+) led to lethality, we
morphologically dissected eight late stage pupae after
carefully removing the pupal envelope.
We also found that two males of these progenies had
a 90°shift of the genital plate, as previously found in
hemizygous bithorax mutants (Karch et al., 1985). Additional
mutant phenotypes were occasionally observed for the bc,~/+
in P31 treated flies (data not shown). These phenotypes are
a missing pastvertical bristle on the dorsal head and a
triplication of antererior scutellars on segment T2. A11 of
the P31 brn~/bw+ progenies had a rough eye phenotype (data not
shown).
Staining of nuclei and polytene chromosomes with polyamides
and immunofluorescence
Kc nuclei and polytene chromosomes were stained
with fluorescent polyamides as described elsewhere (Janssen
et al., 2000). For immunofluorescence, polytene glands were
carefully dissected from b~ or Canton S. flies in lx PBS as
described (Ashburner, 1989). Whole glands in sol P (lxPBS
supplemented with 1 mM MgCl2 and O.lo digitonin) were then
incubated for 60 minutes with various concentrations of
either P31, P9 or no compound at room temperature. P31 (not
P9) induced redistribution of GAF during this incubation
step. Glands were washed then twice for 5 min with sol P and
fixed with fresh paraformaldehyde at a final concentration
of 0.8%. Glands were washed 5 times in sol P for 5 min and
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
74
then blocked for 60 min in the same buffer, supplemented
with 5o non-fat milk. Glands were then incubated overnight
at 4°C in a humid chamber with sol P supplemented with 0.5%
non-fat milk and the rabbit a-GAGA antibodies kindly
provided by Dr. Jordan Raff (Raff et al., 1994). The primary
antibodies were removed by washing (5 x 5 min) and then
incubated with the goat anti-rabbit secondary antibodies
tagged with FITC (Nordic). Glands were washed as above and
stained with 200 nM of P31T to highlight the b~ insert as
described (Janssen et al., 2000). Images were recorded with
a wide field, deconvolution-type imaging system from
DeltaVision.
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
Table 1 Apparent Bindina Affinities of OligopYrroles
KdaPP KdaPP
Compound Sequence ' Ratio
W9 SAR
(~) (nI'4)
P10 (py) 5-(3-Dp 80 35 2 .3
P9 ( PY) a-~- (PY) 0 . 75 0 . 55 1. 4
a'~'~P
P13 ( (Py) 3-(3) 3-Dp 1. 0 1 . 25 0 . 8
P9F (Py) 3-~3- (Py) 4 . 0 1 .0 4
3-~3-Dp-F*
P10F F*-y-(Py)5-(3-Dp 3000 1200 2.5
y = 4-amino-butyric acid
(3 = beta-alanine
F* = Fluorescein
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
76
Table 2
Compound "White mottled""red Mottled" Total Scored
(Strongly (Weakly'
Variegated) Variegated)
(~)
- 62 38 219
P31 64 36 91
P9 8.9 91 161
Table 2 . Suppression of PEV eye phenotype of v~°4 flies by P9
New born flies were transferred to fresh vials and scored
after five days for eye phenotype. Note that only P9
strongly reduced the fraction of flies that remained
strongly variegated ("white mottled"). Suppression of PEV
results in the increased activity of the white gene, that
is, an augmented level of red pigmentation ("red mottled" as
displayed in Figure 7
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
77
0
~ I I I I I I I I I + + + +
x
l0 N 00 Lf7 ~ M .~ ~ O M O
N ~H
O O O O O O O O
~,''
U r1 v-I~-i r1 '~ ~ ~ L~ ~ rt v-1 r1 L~
~ N N M N ~ ~ ~ ~ N N N 41
'
~
, r1 r-1~-i r-I I I I y -1 ~ r1
Q
v
U] 1 1 I I ~ ~ lfl ~ I I I U7
I~ O ~ r, ~D
1 I I
0 0\ o\ o\ o\ o\ O o\ M O o\o o\ o\ N
1J O O O O O O ' O O O
o\ o\
\
0
O 00
r1 o\ o\ o\ o\ o\ o\ o\ o\
rl O O O O O O O o\ o\ o o\
\ o\ O
(~ r1 r1 r1 r1 r-Iri r! l0 N ~ O
,L; I I 1 1 I I I l0 01 r-1
1~ Ln LO LO LO Lf1111 Ln
M U
a
w
a
m
8
H ,-~ ~t
~ ~ <n
~
d
a
In
w
w , + I
w +
~ ~
I M I M I M I I M I M
~'
W W Ly W G4
a
U N ~ R~
\ \ \ \
\ \ \ \ \
H
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
78
nt~s~nnwtr~n a
Adachi, Y., Luke, M., and Laemmli, U. K. (1991). Chromosome
assembly in vitro: topoisomerase II is required for
condensation. Cell 64, 137-48.
around naked DNA in vitro. J Cell Biol 115, 1479-89.
Ashburner, M. (1989). "Drosophila": A Laboratory Manual."
(Cold Spring Harbor, N.Y.: Cold Spring Harbor Laboratory
Press ) .
Assaad, F. F., Tucker, K. L., and Signer, E. R. (1993).
Epigenetic repeat-induced gene silencing (RIGS) in
Arabidopsis. Plant Mol Biol 22, 1067-85.
Baird, E. E., and Dervan, P. B. (1996). Solid phase
synthesis of polyamides containing imidazole and pyrrole
amino acids. J Am Chem Soc 118, 6141-6146.
Beerman, T. A., Woynarowski, J. M., Sigmund, R. D., Gawron,
L. S., Rao, K. E., and Lown, J. W. (1991). Netropsin and
bis-netropsin analogs as inhibitors of the catalytic
activity of mammalian DNA topoisomerase II and topoisomerase
cleavable complexes. Biochim Biophys Acta 1090, 52-60.
Bello, B., Resendez-Perez, D., and Gehring, W. J. (1998).
Spatial and temporal targeting of gene expression in
Drosophila by means of a tetracycline-dependent
tr~ansactivator system. Development 125, 2193-202.
Belyaeva, E. S., Demakova, O. V., Umbetova, G. H., and
Zhimulev, I. F. (1993). Cytogenetic and molecular aspects of
position-effect variegation in Drosophila melanogaster. V.
Heterochromatin-associated protein HP1 appears in
euchromatic chromosomal regions that are inactivated as a
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
79
result of position-effect variegation. Chromosome 102, 583-
90.
Bhat, K. M., Farkas, G., Karch, F., Gyurkovics, H., Gausz,
J., and Schedl, P. (1996). The GAGA factor is required in
the early Drosophila embryo not only for transcriptional
regulation but also for nuclear division. Development I22,
1113-24.
Biggin, M. D., and Tjian, R. (1988). Transcription factors
that activate the Ultrabithorax promoter in developmentally
staged extracts. Cell 53, 699-711.
Biggin, M. D., Bickel, S., Benson, M., Pirrotta, V., and
Tjian, R. (1988). Zeste encodes a sequence-specific
transcription factor that activates the Ultrabithorax
promoter in vitro. Cell 53, 713-22.
Bode, J. , Kohwi, Y. , Dickinson, L. , Joh, T. , Klehr, I?. ,
Mielke, C., and Kohwi-Shigematsu, T. (1992). Biological
significance of unwinding capability of nuclear matrix-
associating DNAs. Science 255, 195-7.
Bode, J., Stengert-Iber, M., Kay, V., Schlake, T., and
Dietz-Pfeilstetter, A. (1996). Scaffold/matrix-attached
regions: topological switches with multiple regulatory
functions. Crit Rev Eukaryot Gene Expr 6, 115-38.
Boy de la Tour, E., and Laemmli, U. K. (1988). The metaphase
scaffold is helically folded: sister chromatids have
predominantly opposite helical handedness. Cell 55, 937-44.
Celniker, S. E., Sharma, S., Keelan, D. J., and Lewis, E. B.
(1990). The molecular genetics of the bithorax complex of
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
Drosophila: cis- regulation in the Abdominal-B domain. Embo
J 9, 4277-86.
Coll, M., Frederick, C. A., Wang, A. H., and Rich, A.
(1987). A bifurcated hydrogen-bonded conformation in the
d(A.T) base pairs of the DNA dodecamer d(CGCAAATTTGCG) and
its complex with distamycin. Proc Natl Acad Sci U S A 84,
8385-9.
Csink, A. K., and Henikoff, S. (1996). Genetic modification
of heterochromatic association and nuclear organization in
Drosophila. Nature 381, 529-31.
Csink, A. K., and Henikoff, S. (1998), Large-scale
chromosomal movements during interphase progression in
Drosophila. J Cell Biol 143, 13-22.
de Clairac, R. P. L., Seel, C. J., Geierstanger, B. H.,
Mrksich, M., Baird, E. E., Dervan, P. B., and Wemmer, D. E.
(1999). NMR characterization of the aliphatic b/b pairing
for recognition of AT/TA base pairs in the minor groove of
DNA. J Am Chem Soc 121, 2956-2964.
Dernburg, A. F., Broman, K. W., Fung, J. C., Marshall, W.
F., Philips, J., Agard, D. A., and Sedat, J. W. (1996).
Perturbation of nuclear architecture by long-distance
chromosome interactions. Cell 85, 745-59.
Dervan, P. B., and Burli, R. W. (1999). Sequence-specific
DNA recognition by polyamides. Curr Opin Chem Biol 3, 688-
93.
Dickinson, L. A., Gulizia, R. J., Trauger,.J. W., Baird, E.
E., Mosier, D. E., Gottesfeld, J. M., and Dervan, P. B.
(1998). Inhibition of RNA polymerase II transcription in
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
8I
human cells by synthetic DNA-binding ligands [see comments].
Proc Natl Acad Sci U S A 95, 12890-5.
Dickinson, h. A., Trauger, J. W., Baird, E. E., Dervan, P.
B., Graves, B. J., and Gottesfeld, J. M. (1999). Inhibition
of Ets-1 DNA binding and ternary complex formation between
Ets-1, NF-kappaB, and DNA by a designed DNA-binding ligand.
J Biol Chem 274, 12765-73.
Dimitrov S, D. M., Wolffe AP (1994). Remodeling sperm
chromatin in Xenopus laevis egg
borer, D. R., and Henikoff, S. (1994). Expansions of
transgene repeats cause heterochromatin formation and gene
silencing in Drosophila. Cell '77, 993-1002.
borer, D. R., and Henikoff, S. (1997). Transgene repeat
'arrays interact with distant heterochromatin and cause
silencing in cis and trans. Genetics 147, 1181-90.
born, A., Affolter, M., Muller, M., Gehring, W. J., and
heupin, W. (1992). Distamycin-induced inhibition of
homeodomain-DNA complexes. Embo J 11, 279-86.
Elgin, S. C. (1996). Heterochromatin and gene regulation in
Drosophila. Curr Opin Genet Dev 6, 193-202.
Elgin, S. C., and Jackson, S. P. (1997). Chromosomes and
expression mechanisms [editorial]. Curr Opin Genet Dev 7,
149-51.
extracts: the role of core histone phosphorylation and
Farkas, G., Gausz, J., Galloni, M., Reuter, G., Gyurkovics,
H., and Karch, F. (1994). The Trithorax-like gene encodes
the Drosophila GAGA factor. Nature 371, 806-8.
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
82
Fernandez, L. A., Winkler, M., Forrester, W., Jenuwein, T.,
and Grosschedl, R. (1998). Nuclear matrix attachment regions
confer long-range function upon the immunoglobulin mu
enhancer. Cold Spring Harb Symp Quant Biol 63, 515-24.
for the Notch gene are in the interband. Cell 54, 46I-72.
Forrester, W. C., Fernandez, L. A., and Grosschedl, R.
(1999). Nuclear matrix attachment regions antagonize
methylation-dependent repression of long-range enhancer-
promoter interactions. Genes Dev 13, 3003-3014.
Frederickson, R. (1999). "Functional" proteomics? Nat
Biotechnol 17, 1050.
Garrick, D., Fiering, S., Martin, D. I., and Whitelaw, E.
(1998). Repeat-induced gene silencing in mammals [see
comments]. Nat Genet 18, 56-9.
Gasser, S. M., and Laemmli, U. K. (1986). Cohabitation of
scaffold binding regions with upstream/enhancer elements of
three developmentally regulated genes of D, melanogaster.
Cell 46, 521-30.
Geierstanger, B. H., Mrksich, M., Dervan, P. B., and Wemmer,
D. E. (1994). Design of a G.C-specific DNA minor groove-
binding peptide. Science 266, 646-50.
Guard, F., Bello, B., Laemmli, U. K., and Gehring, W. J.
(1998). In vivo
analysis of scaffold-associated regions in Drosophila: a
synthetic high-affinity SAR binding protein suppresses
position effect variegation. Embo J 17, 2079-85.
Goodsell, D., and Dickerson, R. E. (1986). Isohelical
analysis of DNA groove-binding drugs. J Med Chem 29, 727-33,
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
83
Gottesfeld, J. M., Neely, L., Trauger, J. W., Baird, E. E.,
and Dervan, P. B. (1997). Regulation of gene expression by
small molecules. Nature 387, 202-5.
Granok, H., Leibovitch, B. A., Shaffer, C. D., and Elgin, S.
C. (1995). Chromatin. Ga-ga over GAGA factor. Curr Biol 5,
238-41.
Hart, C. M., and Laemmli, U. K. (1998). Facilitation of
chromatin dynamics by SARs. Curr Opin Genet Dev 8, 519-25.
Hazelrigg, T., Levis, R., and Rubin, G. M. (1984).
Transformation of white locus DNA in drosophila: dosage
compensation, zeste interaction, and position effects. Cell
36, 469-81.
Henikoff, S. (2000). Heterochromatin function in complex
genomes. Biochim Biophys Acta 1470, 01-08.
Henikoff, S., and Comai, L. (1998). Traps-sensing effects:
the ups and downs of being together. Cell 93, 329-32.
Henikoff, S., and Dreesen, T. D. (1989). Traps-inactivation
of the Drosophila brown gene: evidence for transcriptional
repression and somatic pairing dependence. Proc Natl Acad
Sci U S A 86, 6704-8.
Henikoff, S., Jackson, J. M., and Talbert, P. B. (1995).
Distance and pairing effects on the brownDominant
heterochromatic element in Drosophila. Genetics 140, 1007-
17.
heterochromatin of Drosophila melanogaster. Genetics 134,
1149-74.
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
84
Hirano T, K. R., Hirano M (1997). Condensins, chromosome
condensation protein complexes containing XCAP-C, XCAP-E and
a Xenopus homolog of
Hirano, T., and Mitchison, T. J. (1991). Cell cycle control
of higher-order chromatin assembly the Drosophila Barren
protein. Cell 89, 511-21.
Hsieh, T., and Brutlag, D. (1979). Sequence and sequence
variation within the 1.688 g/cm3 satellite DNA of Drosophila
melanogaster. J Mol Biol 135, 465-81.
immunodeficiency virus type 1 integrase.~Mol Pharmacol 54,
280-90.
Janssen, S., Cuvier, 0., Muller, M., and Laemmli, U. K.
(2000). Specific Gain and Loss of Function Phenotypes
induced by Satellite-specific DNA-binding Drugs fed to
Drosophila melanogaster.
Janssen, S., Durussel, T., and Laemmli, U. K. (2000).
Synthesis and Characterization of DNA Satellite-specific
Drugs: Chromatin Opening and Chromosome Condensation. .
Karch, F., Weiffenbach, B., Peifer, M., Bender, W., Duncan,
I., Celniker, S., Crosby, M., and Lewis, E. B. (1985). The
abdominal region of the bithorax complex. Cell 43, 81-96.
Karpen, G. H. (1994). Position-effect variegation and the
new biology of heterochromatin. Curr Opin Genet Dev 4, 281-
91.
Kas, E., and Laemmli, U. K. (1992). In vivo topoisomerase II
cleavage of the Drosophila histone and satellite III
repeats: DNA sequence and structural characteristics. Embo J
11, 705-16.
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
Kas, E., Poljak, L., Adachi, Y., and Laemmli, U. K. (1993).
A model for chromatin opening: stimulation of topoisomerase
II and restriction enzyme cleavage of chromatin by
distamycin. Embo J 12, 215-26.
Katsani, K. R., Hajibagheri, M. A., and Verrijzer, C. P.
(1999). Co-operative DNA binding by GAGA transcription
factor requires the conserved BTB/POZ domain and reorganizes
promoter topology. Embo J 18, 698-708.
Kennison, J. A., Vazquez, M., and Brizuela, B. J. (1998).
Regulation of the Sex combs reduced gene in Drosophila. Ann
N Y Acad Sci 842, 28-35.
Kielkopf, C. L., Baird, E. E., Dervan, P. B., and Rees, D.
C. (1998). Structural basis for G.C recognition in the DNA
minor groove. Nat Struct Biol 5, 104-9.
Kirillov, A., Kistler, B., Mostoslavsky, R., Cedar, H.,
Wirth, T., and Bergman, Y. (1996). A role for nuclear NF-
kappaB in B-cell-specific demethylation of the Igkappa
locus. Nat Genet 13, 435-41.
Laemmli, U. K., Kas, E., Poljak, L., and Adachi, Y. (1992) .
Scaffold-associated regions: cis-acting determinants of
chromatin structural loops and functional domains. Curr Opin
Genet Dev 2, 275-85.
Locke, J., Kotarski, M. A., and Tartof, K. D. (1988).
Dosage-dependent modifiers of position effect variegation in
Drosophila and a mass action model that explains. their
effect. Genetics 120, 181-98.
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
86
Lohe, A. R., Hilliker, A. J., and Roberts, P. A. (1993).
Mapping simple repeated DNA sequences in Heterochromatin of
Drosophila melanogaster. Genetics, 134,1149-1174
Lu, B. Y., Bishop, C. P., and Eissenberg, J. C. (1996).
Developmental timing and tissue specificity of
heterochromatin-mediated silencing. Embo J 15, 1323-32.
Lu, Q., Wallrath, L. L., Granok, H., and Elgin, S. C.
(1993). (CT)n (GA)n repeats and heat shock elements have
distinct roles in chromatin structure and transcriptional
activation of the Drosophila hsp26 gene.~Mol Cell Biol 13,
2802-14.
Luger, K., Mader, A. W., Richmond, R. K., Sargent, D. F.,
and Richmond, T. J. (1997). Crystal structure of the
nucleosome core particle at 2.8 A resolution [see comments].
Nature 389, 251-60.
Marshall, W. F., Straight, A., Marko, J. F., Swedlow, J.,
Dernburg, A., Belmont, A., Murray, A. W., Agard, D. A., and
Sedat, J. W. (1997). Interphase chromosomes undergo
constrained diffusional motion in living cells. Curr Biol 7,
930-9.
Martello, P. A., Bruzik, J. P., deHaseth, P., Youngquist, R.
S., and Dervan, P. B. (1989). Specific activation of open
complex formation at an Escherichia coli promoter by
McBryant, S. J., Baird, E. E., Trauger, J. W., Dervan, P.
B., and Gottesfeld, J. M. {1999). Minor groove DNA-protein
contacts upstream of a tRNA gene detected with a synthetic
DNA binding ligand. J Mol Biol 286, 973-81._
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
87
Miklos, G. Z., and Cotsell, J. N. (1990). Chromosome
structure at interfaces between major chromatin types:alpha-
and beta-heterochromatin. Bioessays 12, 1-6.
Mirkovitch, J., Mirault, M. E., and Zaemmli, U. K. (1984).
Organization of the higher-order chromatin loop: specific
DNA attachment sites on nuclear scaffold. Cell 39, 223-32.
Muller, H. J. (1930). Types of Visible Variations Induced by
X-Rays in Drosophila. J. Genet 22, 299-334.
Neamati, N., Mazumder, A., Sunder, S., M., 0. J., M., T.,
hown, J. W., and Pommier, Y. (1998). Highly potent synthetic
polyamides, bisdistamycins, and lexitropsins as inhibitors
of human
oligo(N-methylpyrrolecarboxamide)s: effects of peptide
length and identification of DNA target
Omichinski, J. G., Pedone, P. V., Felsenfeld, G.,
Gronenborn, A. M., and Clore, G. M. (1997). The solution
structure of a specific GAGA factor-DNA complex reveals a
modular binding mode [see comments]. Nat Struct Biol 4, 122-
32.
Pedone, P. V., Ghirlando, R., Clore, G. M., Gronenborn, A.
M., Felsenfeld, G., and 0michinski, J. G. (1996). The single
Cys2-His2 zinc finger domain of the GAGA protein flanked by
basic residues is sufficient for high-affinity specific DNA
binding [published erratum appears in Proc Natl Acad Sci U S
A 1996 Jun 25;93(13):6846]. Proc Natl Acad Sci U S A 93,
2822-6.
Pelton, J. G., and Wemmer, D. E. (1989). Structural
characterization of a 2:1 distamycin A.d(CGCAAATTGGC)
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
88
complex by two-dimensional NMR. Proc Natl Acad Sci U S A 86,
5723-7.
Platero, J. S., Ahmad, K., and Henikoff, S. (1999). A distal
heterochromatic block displays centromeric activity when
detached from a natural centromere [In Process Citation].
Mol Cell 4, 995-1004.
Platero, J. S., Csink, A. K., Quintanilla, A., and Henikoff,
S. (1998). Changes in chromosomal localization of
heterochromatin-binding proteins during the cell cycle in
Drosophila. J Cell Biol 140, 1297-306.
Poljak, Z., Seum, C., Mattioni, T., and Laemmli, U. K.
(1994). SARs stimulate but do not confer position
independent gene expression. Nucleic Acids Res 22, 4386-94.
Raff, J. W., Kellum, R., and Alberts, B. (1994). The
Drosophila GAGA transcription factor is associated with
specific regions of heterochromatin throughout the cell
cycle. Embo J 13, 5977-83.
Reeves, R., and Nissen, M. S. (1990). The A.T-DNA-binding
domain of mammalian high mobility group I chromosomal
proteins. A novel peptide motif for recognizing DNA
structure. J Biol Chem 265, 8573-82.
Rykowski, M. C., Parmelee, S. J., Agard, D. A., and Sedat,
J. W. (1988). Precise determination of the molecular limits
of a polytene chromosome band: regulatory sequences
Saitoh, Y., and Zaemmli, U. K. (1994). Metaphase chromosome
structure: bands arise from a differential folding path of
the highly AT-rich scaffold. Cell 76, 609-22.
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
89
Sass, G. L., and Henikoff, S. (1998). Comparative analysis
of position-effect variegation mutations in Drosophila
melanogaster delineates the targets of modifiers. Genetics
148, 733-41.
sites. Biochemistry 28, 4455-61.
Spierer, A., and Spierer, P. (1984). Similar level of
polyteny in bands and interbands of Drosophila giant
chromosomes. Nature 307, 176-8.
Strick, R., and Laemmli, U. K. (1995). SARs are cis DNA
elements of chromosome dynamics: synthesis of a SAR
repressor protein. Cell 83, 1137-48.
Strick, R. , and Laemmli, U. K. (1995) . SARs are cis DNA
elements of chromosome dynamics: synthesis of a SAR
repressor protein. Cell 83, 1137-48.
Strutt, H., Cavalli, G., and Paro, R. (1997). Co-
localization of Polycomb protein and GAGA factor on
regulatory elements responsible for the maintenance of
homeotic gene expression. Embo J 16, 3621-32.
Talbert, P. B., LeCiel, C. D., and Henikoff, S. (1994).
Modification of the Drosophila heterochromatic mutation
brownDominant by linkage alterations. Genetics 136, 559-71.
Tartof, K. D., Bishop, C., Jones, M., Hobbs, C. A., and
Locke, J. (1989). Towards an understanding of position
effect variegation. Dev Genet 10, 162-76.
Taylor, J. S., Schultz, P. G., and Dervan, P. B. (1984).
Sequence specific cleavage of DNA by distamycin-EDTA .
Fe(II) and EDTA-distamycin . Fe(II). Tetrahedron 40, 457-
465.
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
Tsukiyama, T., Becker, P. B., and Wu, C. (1994). ATP-
dependent nucleosome disruption at a heat-shock promoter
mediated by binding of GAGA transcription factor [see
comments]. Nature 367, 525-32.
Turner, J. M., Swalley, S. E., Baird, E. E., and Dervan, P.
B. (1998). Aliphatic/aromatic amino acid pairings for
polyamide recognition in the minor groove of DNA. J. Am.
Chem. Soc. 120, 6219-6226.
Tyler, J. K., and Kadonaga, J. T. (1999). The "dark side" of
chromatin remodeling: repressive effects~on transcription.
Cell 99, 443-6.
Umbetova, G. H., Belyaeva, E. S., Baricheva, E. M., and
Zhimulev, I. F. (1991). Cytogenetic and molecular aspects of
position effect variegation in Drosophila melanogaster. IV.
Underreplication of chromosomal material as a result of gene
inactivation. Chromosoma 101, 55-61.
Wakimoto, B. T. (1998). Beyond the nucleosome: epigenetic
aspects of position-effect variegation in Drosophila. Cell
93, 321-4.
Wallrath, L. L., and Elgin, S. C. (1995). Position effect
variegation in Drosophila is associated with an altered
chromatin structure. Genes Dev 9, 1263-77.
Wang, W., and Lown, J. W. (1992). Anti-HIV-I activity of
linked lexitropsins. J Med Chem 35, 2890-7.
White, S., Baird, E. E., and Dervan, P. B. (1997). On the
pairing rules for recognition in the minor groove of DNA by
pyrrole-imidazole polyamides. Chem Biol 4, 569-78.
CA 02409408 2002-11-19
WO 02/00262 PCT/EPO1/08097
91
Wilkins, R. C., and Zis, J. T. (1998). GAGA factor binding
to DNA via a single trinucleotide sequence element. Nucleic
Acids Res 26, 2672-8.
Yanagi, K., Prive, G. G., and Dickerson, R. E. (1991).
Analysis of local helix geometry in three B-DNA decamers and
eight dodecamers. J. Mol. Biol. 217, 201-214.
Youngquist, R. S., and Dervan, P. B. (1985). Sequence-
specific recognition of B-DNA by oligo(N-
methylpyrrolecarboxamide)s. Proc Natl Acad Sci U S A 82,
2565-9.
Youngquist, R. S., and Dervan, P. B. (1987). A synthetic
Peptide binds 16 base pairs of A,T double helical DNA. J Am
Chem Soc 109, 564-7566.
