765
Modular design of artificial transcription factors
Aseem Z Ansari* and Anna K Mapp†
Eukaryotic transcription factors are composed of
interchangeable modules. This has led to the design of a wide
variety of modular artificial transcription factors (ATFs) that can
stimulate or inhibit the expression of targeted genes. The ability
to regulate the expression of any targeted gene using a
‘programmable’ ATF offers a powerful tool for functional
genomics and bears tremendous promise in developing the
field of transcription-based therapeutics.
Addresses
*Department of Biochemistry and The Genome Center, 433 Babcock
Drive, University of Wisconsin-Madison, Madison, WI 53706, USA
† Department of Chemistry and Department of Medicinal Chemistry,
930 N University Avenue, University of Michigan, Ann Arbor,
MI 48109-1055, USA
Current Opinion in Chemical Biology 2002, 6:765–772
1367-5931/02/$ — see front matter
© 2002 Elsevier Science Ltd. All rights reserved.
Published online 18 October 2002
Abbreviations
ATF
artificial transcription factor
DBD
DNA-binding domain
KRAB Kruppel-associated box
NLS
nuclear localization signal
PNA
peptide nucleic acid
RD
regulatory domain
TFO
triplex-forming oligonucleotide
Introduction: regulation of gene expression
All cells in an organism, with a few exceptions, bear the
same genome, yet they specialize to give rise to tissues
with diverse morphology and function. Clearly, this diversity
is not due to the differences in genomic information
between the tissues in an organism; rather, it is the choice
of genes that are expressed which governs the fate and
function of cells. The recent decoding of the human
genome coupled with genome-wide expression profiling is
clarifying the relationship between specific gene-expression
patterns and cell fate. Patterns of gene expression in cells
are triggered by a variety of intracellular and extracellular
signals that mobilize specific transcriptional regulatory
factors [1••]. Once marshaled, these regulators interact and
recruit various cellular machines to directly govern the
transcription of the specified genes, prescribing both the
time and the scale of positive or negative control of transcription (Figure 1a). In diseases as diverse as diabetes and
cancer, it is often malfunctioning transcriptional regulators
that produce the altered patterns of gene expression at the
heart of the ailment [2,3,4••].
In this context, molecules that can seek out specific genes
and control their expression are attractive targets for study
[5–7,8•]. These artificial transcription factors (ATFs) are
formally defined as functional entities that regulate
transcription either as a single molecule or as an ensemble
of interacting molecules (Figure 1b). To achieve this, an
ATF may (but does not necessarily) bind to DNA and
directly influence the transcription of a specific gene or set
of genes either positively or negatively. ATFs do not
function only on initiation but may also regulate subsequent
steps in transcription. Ideally, ATFs could be designed to
regulate any gene or set of genes without influencing the
expression of any other gene in the genome. Such factors
would be powerful tools for unraveling the key transcriptional events that govern cell fate, and in the long term
would have significant therapeutic potential.
Due to the rich cross-fertilization of ideas at the interface
of chemistry, biology, medicine and genomics, this is an
exciting time in the field of ATF design. Technological
advances coupled with new mechanistic insights into gene
regulation and in vivo delivery make the creation of ATFs
that can reprogram gene expression in living organisms a
tangible goal. In this review, we summarize recent
advances in ATF design within a historical context and
outline a few of the key challenges facing the field.
Modular nature of eukaryotic transcription factors
Natural transcriptional regulators typically bear two essential
yet separable modules: the DNA-binding domain (DBD)
and the regulatory domain (RD) [1••]. The DBD imparts
most of the specificity in targeting the RD to a particular
site in the genome. The regulatory modules typically play
a less critical role in selecting a gene for regulation; instead,
they mediate their effects directly on the gene to which
they are delivered. These two functional modules are
usually exchangeable (Figure 2). In some cases, more often
with repressors, the DBD is not a part of the regulatory
factor [1••]. Rather, proteins containing repressor domains are
targeted to specific genes by interactions with other DNAbinding proteins already bound to the promoter [9,10].
The DNA-binding modules of transcriptional regulators
consist of a wide variety of structural folds and many have
been well characterized from both a structural and functional standpoint [11••]. By contrast, regulatory modules
are less well-defined and only a few have been characterized
structurally [8•]. Whereas DBDs are often described by
their structural folds (helix-turn-helix, for example) or by
requirements for certain co-factors (zinc fingers), RDs are
typically categorized using simpler criteria such as the
preponderance of certain residues within the regulatory
sequence (acid-rich [1••], glutamine-rich [12,13], prolinerich [14] activating modules and alanine-rich or positivecharge-rich [15] repression modules, for example). RDs are
also cataloged by functional context. For example, some
activator and repressor modules are categorized according
to the distance(s) from the promoter at which they function
766
Model systems
Figure 1
(a)
Chromatin
remodeling
enzyme(s)
Med
(i)
RNA Pol II
Activator
Med
RNA Pol II
Chromatin
condensing
enzyme(s) Co-repressor(s)
Repressor
(ii)
(b)
An overview of the current models of
transcriptional regulation and ATFs. (a) (i)
Genomic DNA is wrapped around histone
octamers to provide nucleosomes that are
further compacted in chromatin fibers. To upregulate a gene, the activator binds to a
specific sequence in the genome and recruits
various cellular machines (such as chromatin
remodeling enzymes) that permit RNA
polymerase and its associated proteins (the
mediator complex, for example) to bind and
transcribe. (ii) One of several mechanisms by
which repressors function. In this example, the
repressor, instructed by an appropriate
cellular signal, binds to a specific DNA site
and recruits protein complexes (corepressors, histone deacetylases, histone
methyltransferases) that modify the histones
and thereby compact the DNA [1••,8•]. (b) An
ATF directly influences the transcription of a
specific gene or set of genes either as (i) a
single entity or (ii,iii) as an ensemble of
interacting molecules. Analogous to natural
transcriptional regulators, ATFs are typically
modular and contain both a DBD (shown in
blue) and an RD (shown in red).
On
Off
(i)
On
Off
(ii)
On
Off
(iii)
Current Opinion in Chemical Biology
or by the specific stages of transcription at which they
exert their influence [16–18]. This includes, for example,
activating domains that work on transcription initiation,
those that work on elongation, and others that stimulate
both steps of the process [16–18].
Modular design of ATFs
The earliest ATFs were those that were generated in the
domain-swapping experiments that defined the modular
nature of natural transcriptional regulators [1••]. Thus,
subsequent efforts to generate ATFs with broader and
‘programmable’ DNA targeting and regulatory features have
replaced the modules present in natural transcription
regulators with designed counterparts [8•]. In Figure 3 we
summarize the domains that have been used to generate
ATFs and discuss their development below.
DNA-binding domains
Because the DNA-binding module, in great measure,
specifies which genes are to be bound and regulated, it has
been the focus of the most attention in the development of
ATFs. Protein DBDs have been attractive targets because
of the high specificity and affinity that they typically
display for their target sequences. Despite a rich base of
structural knowledge of protein DBDs, no clear ‘recognition code’ has been deduced for the rational design of new
recognition motifs to target desired DNA sequences [19••].
To fill this need, genetic selection approaches have been
used to generate protein DNA-binding modules for specific
sequences [20,21•,22]. Using these approaches, zinc-fingerbased DBDs that target unique promoter sequences have
been isolated and shown to function in cultured cell
lines [22].
Modular design of artificial transcription factors Ansari and Mapp
767
Figure 2
Eukaryotic transcription factors are modular
and exchangeable. An activating RD (shown
in red) excised from transcription factor A and
attached to the DBD (shown in blue) from
transcription factor B will stimulate
transcription of the genes that are bound by
the DBD of transcription factor B, and it will
not specifically affect expression of any other
gene. Similarly a repression module will inhibit
the expression of genes dependent on where
it is targeted.
RD
'A'
DBD
'A'
Gene A
RD
'B'
Gene A
DBD
'A'
RD
'B'
DBD
'B'
Gene B
Gene B
Current Opinion in Chemical Biology
The development of non-peptidic DBDs has provided
exciting alternatives for ATF design and function in recent
years [23]. As illustrated in Figure 3, these include the
triplex-forming oligonucleotides (TFOs), which recognize
specific DNA sequences through the formation of
Hoogsteen base pairs and have been utilized as a DNAbinding module [24]. Similarly, peptide nucleic acids
(PNAs), in which the phosphodiester backbone has been
replaced with amide bonds, have been used with some
success to target specific DNA sequences [25]. Also noteworthy are the DNA-binding modules inspired by the
natural product distamycin. Unlike most DBDs, these
‘polyamides’ bind in the minor groove of DNA; Dervan
and co-workers [6] have used rational design to develop
molecules that recognize each of the four possible DNA
base pairs.
Regulatory domains: activation
Activating modules that bear multiple acidic and
hydrophobic residues (‘acidic activators’) function quite
robustly in all eukaryotes tested and, as a result, have been
frequently used in ATFs. Designed zinc fingers fused to
one such acidic activating domain, the potent viral co-activator
VP16, have been used to up-regulate the endogenous
genes erbB-2 by 6- to 8-fold [26] and VEGF-A by 15- to
30-fold [27] in cultured cell lines. The minimal segment of
VP16 that retains activator function [28] has also been used
in several ATFs (Figure 3). Recently, Yaghmai and Cutting
[29••] demonstrated that the potency of zinc finger ATFs
is dependent upon cell type, the distance of the DBD sites
from the transcription start site, and the number of VP16
activating domain repeats present. Attachment of one or
two repeats of VP16 activating domain to TFOs also
provides ATFs with activation levels in cell culture ranging
from 5- to 40-fold [30•,31•]. When an analogous VP16
activating region was fused to hairpin polyamides, activation
levels of 18- to 34-fold were observed in an in vitro system
[32••,33]. Verdine and co-workers [34] have also reported
that the minimal VP16 activating region composed of
D-amino acids functions almost as well as the counterpart
bearing L-amino acids.
The generation of novel peptidic activators has been
accomplished by several approaches. One of the earliest
studies used genetic screens with random sequences of
Escherichia coli DNA attached to natural DBDs. The
activating modules identified contained an excess of acidic
residues, analogous to natural activating regions [1,35]. A
recent study utilizing randomized octapeptides provided
several hydrophobic sequences that activated transcription
in yeast as robustly as the strongest natural activators when
fused to protein DBDs [36•]. More recently, attempts at
broadening the repertoire by creating regulatory modules
that specifically interact with one or another component of
the transcriptional machinery have yielded peptides that
interact with a human histone acetyl-transferase [37],
the yeast repressor Gal80 [38], or the yeast mediator
component Gal11 (AK Mapp, AR Minter, G Belanger and
JK Lum, unpublished data). The Gal80-binding peptides,
despite targeting a yeast repressor, function as activators
both in yeast and in mammalian cells (T Kodadek and
Y Han, personal communication).
Genetic screens have also yielded RNA molecules that
activate up to 100-fold in budding yeast (Saccharomyces
cerevisiae) when tethered to DNA. This is the first reported
example of a non-peptidic activator [39] (S Saha,
AZ Ansari, K Jarrell and M Ptashne, unpublished data).
Regulatory domains: repression
RDs that repress transcription have not been as clearly
defined. A few short peptides that function as repressor
768
Model systems
Figure 3
DBD
Linker
RD
Protein
Activation domain
PEG
Gal4
O
N
H
O
O
VP16
[...DFDLDMLG...]N
O
O
N
H
O
Peptide
HTH
O
H2N
O
H
N
N
H
N
H
OH
O
O
Zinc finger
O
N
N
O
N
(GGSGGS)
CO2H
N
H
O
AH
PEFPGIELQELQE
LQALLQQQ
OH
O
H
N
∗ RNA
O
(PPPP)N
N
N
FK506-CsA
Repression domain
MeO
Non-protein
O
MeO
H
Me
N
Me
N
N
Me
O
∗ PNA
N
H
O
H
N
O
HO
O
O
NMe
H
N
N
H
∗ KRAB
N
O
H
O
OH
O
O
H
N
N
Me
O
Me
N
N
H
O
HO
NMe
O
O
O
H
O
OMe
OH
O
O
∗ WRPW
OMe
NH2
Polyamide
N
H2N
NH
N
N
N
N
HO2C
H
N
O
OMe
H
N
8
WRPW
Mtx-SLF
O
ch o
O
O
ou
∗ TFO
Gr
O
O
O
N
O
O
O
∗ CCVC
Current Opinion in Chemical Biology
Modules used in the design of ATFs. Representative examples of the
DBDs, RDs and linkers used in the design of most ATFs are
summarized. Asterisks denote modules for which little structural data
are available. Common protein-based DBDs include Gal4 [59], helixturn-helix proteins (HTH; a generic binding scheme is shown) such
as LexA, LacI and TetR, and both natural and designed zinc-finger
motifs (structure from [60]) [11••]. The most frequently used nonprotein DBDs are TFOs, PNAs and polyamides (for structure see
[61]). While the linker is most commonly a flexible (GGSGGS) or
rigid (polyproline [50]) peptide, a simple flexible linker derived from
polyethylene glycol (PEG) can also be used without loss of function.
Small molecules such as the FK506-CsA dimer [49] or the dimer of
methotrexate (Mtx) and a synthetic analog of FK506 (SLF) [52] can
also bridge the DBDs and RDs by non-covalent interactions [48].
Most activating regions have been derived from natural activator
proteins such as VP16 [8•,12] or designed to mimic natural
activating regions (AH) [32••,35]. One exception is the use of RNA
hairpins [39] (S Saha, AZ Ansari, K Jarrell and M Ptashne
unpublished results). The repression domain most often used is
derived from the KRAB protein [26,29••] but small peptides such as
WRPW [40] and CCVC [41] can also function as repressors when
bound to DNA. In many cases, the modules shown can be readily
swapped to generate ATFs that target different DNA sequences or
have a different regulatory function.
modules have been identified, and these commonly contain
a preponderance of alanine and/or basic residues; these, however, function only modestly [15]. One interesting example
is the tetrapeptide WRPW that, when fused to a DBD,
interacts with the Drosophila repressor Groucho (TLE in
humans), which then recruits proteins that modify chromatin
structure and actively repress the expression of neighboring
genes [40]. In yeast, cysteine-rich peptides have been reported
to localize the gene to which they are tethered via a DBD to
the nuclear envelope and, by an unknown mechanism,
‘silence’ its expression [41]. Repressive ATFs have also been
generated by fusing large segments of repressor proteins to a
DBD. For example, when the Kruppel-associated box
(KRAB) repressor module is fused to designed zinc fingers, a
modest repression of the adjacent gene is observed [26,29••].
It is not yet clear, however, how the various repressor
modules described above will function when fused to
synthetic DBDs. This remains a relatively unexplored area
of ATF design and function.
More typically, inhibition of gene expression with an ATF
is achieved by competition for the DNA binding site of an
Modular design of artificial transcription factors Ansari and Mapp
endogenous transcription factor. A breakthrough in this
approach was achieved when polyamides designed to
target transcription factor binding sites were successfully
used to inhibit the expression of HIV genes in cell culture
[42]. Recently, Laemmli and co-workers [43•] designed
polyamides to target specific sequences in Drosophila
melanogaster. When fed to developing flies, a distinct
phenotypic change was caused by the specific polyamides
but not by those designed to target a different sequence.
Alternatively, PNA-based inhibitors have been used to
inhibit transcriptional elongation of the oncogenic c-myc
gene in Burkitt’s lymphoma and in prostate cancer cells
[44•,45]. In general, the inhibition strategy demands precise
knowledge of which DNA binding sites to target in vivo.
Moreover, the requirement that the ATF out-compete the
endogenous factor for binding to its cognate DNA site may
be hard to achieve for every factor [46••], especially in
cases where a large number of transcription factors bind
cooperatively to adjacent sites [1••]. Thus, it may be more
desirable in the long term to generate ATFs bearing
repression modules that can actively silence transcription
of a gene even when bound at a distal site on the promoter
that does not overlap with a binding site of an endogenous
transcription factor.
Linking the DNA-binding and regulatory domains
To achieve activation or active repression, the DBD and
RD must be linked together to function as an ATF [32••].
Most commonly, the two domains are tethered via peptide
linkers. However, a covalent linker is not essential; noncovalent interactions that bridge a DBD to an RD suffice
to generate an ATF (Figure 3). These bridging non-covalent
interactions can be mediated by protein–protein [1••],
protein–RNA [47], or protein–small molecule binding
events [48,49]. Systematic studies with flexible chemical
linkers as well as rigid peptide linkers have shown that the
linker length is an important variable in optimizing the
function of the regulatory module [32••,33,50].
Some of the small molecule linkers mentioned above have
the additional benefit of providing external control over
the function of intracellular ATFs. Two pioneering examples
of ligand-dependent ATFs were provided by the groups of
Bujard and Schreiber. Bujard and co-workers [51] developed
ATFs that bind to DNA only in the presence of doxycyclin,
whereas Schreiber and co-workers [49] used chemical
inducers of dimerization (CIDs) to mediate the interaction
of a DBD and a RD in eukaryotes. In a recent development, Cornish and co-workers [52] reported a novel CID
composed of methotrexate and a synthetic analog of the
natural product FK506 to manipulate the interaction of a
DBD with an activating region that functions robustly in
bacteria. Inspired by earlier work on steroid-responsive
nuclear hormone receptors, Barbas and co-workers [53•]
described an allosteric ligand-responsive ATF constructed
by linking a zinc finger DBD to a VP16 AD via the ligandbinding domain of a steroid receptor. These strategies
continue to provide powerful tools for dissecting regulatory
769
Figure 4
(b)
(a)
(c)
On
(d)
Off
Current Opinion in Chemical Biology
ATFs that respond to physiological cues. An intriguing goal for the
design of future ATFs involves the harnessing of signal transduction
pathways to activate dormant ATFs. (a) Physiological signals are often
monitored by cell surface receptors. (b) The interaction between a
signal and the receptor typically triggers an intracellular signal
transduction pathway. (c) A dormant ATF containing a signalresponsive functional moiety would be converted to its functional state
upon sensing the signal. (d) The functional ATF would then regulate
specifically targeted genes in either a positive or negative manner.
ATFs that could integrate instructions from different signal transduction
pathways or respond only to signals in certain tissues would provide
an even greater level of temporal and spatial control over the regulation
of specific genes or gene clusters.
pathways in transcription and may also prove important for
the development of ATF-based therapies.
Traffic and delivery
Delivery of ATFs into cells and organisms continues to be
a considerable obstacle. For example, a recent report by
Dervan and co-workers demonstrates that the trafficking
of polyamide ATFs into cell nuclei is strongly dependent
upon cell type, with cultured and primary T-cells showing
the most nuclear localization [54•]. TFO-based regulators
are typically delivered into cells either by electroporation
or by cationic liposomes [30•,31•]. This is an unsustainable
way to modulate transcriptional regulation and it severely
limits the tissues that can be targeted. By contrast, the
covalent attachment of a short peptide such as the nuclear
localization signal (NLS) or of the steroid dihydroxytestosterone to PNA-containing ATFs results in effective
delivery into cell nuclei [44•,45]. In addition, an argininerich peptide derived from the HIV transcriptional regulator
Tat, peptidomimetic derivatives of Tat, and certain
hydrophobic peptides can impart potent cell membrane
transducing properties on the molecules to which they are
coupled [55–58]. Peptides that serve as delivery modules
could also be expressed contiguous with the protein-based
770
Model systems
ATFs; this may enable delivery of these molecules into
cells without gene therapy. This strategy could make
ATFs of all varieties more viable in vivo.
regulatory molecules that may provide a powerful tool to
dissect transcriptional networks (functional genomics) and
perhaps create a new class of transcription-based therapeutics.
Future design considerations
Acknowledgements
As summarized in Figure 3, most ATFs to date have been
generated by mixing and matching the DBD, the RD, and,
in some cases, the linker between the two. Although the
ability to substitute the DBD and RD has been an essential
design feature of the current generation of ATFs, additional features must be included in the coming generations
to achieve greater control in regulating targeted genes. To
achieve a further degree of subtlety in transcriptional
control, the next generation of ATFs may also have to
include properties found in natural transcriptional regulators
such as the ability to interact only with cell type-specific
components of the transcriptional machinery and the
ability to respond to physiological cues.
AZA thanks the University of Wisconsin for a Steenbock Career
Development award. AKM acknowledges financial support for this work
from the NIH (GM65330), the Burroughs Wellcome Fund, and the
March of Dimes.
The ability to sustain the desired expression level of
targeted genes over a prolonged period also remains a
significant hurdle. Small molecule RDs are attractive targets
in this context because they would probably be resistant to
proteolytic degradation in the cell. Small molecule regulators
might also overcome some of the limitations faced by
peptidic RDs, such as poor cell permeability, the potential
for triggering an immune response, or possible effects on
function of other endogenous regulators by competing for
common cellular targets. Furthermore, small molecules
that interact with cell type-specific or species-specific
components of the transcriptional machinery could provide
an additional degree of functional specificity.
References and recommended reading
Papers of particular interest, published within the annual period of review,
have been highlighted as:
• of special interest
•• of outstanding interest
1. Ptashne M, Gann A: Genes & Signals. New York: Cold Spring Harbor
•• Laboratory; 2001.
This book summarizes much that is known about transcriptional mechanisms,
activators, repressors and the transcriptional machinery in an accessible
format with many primary literature citations. The authors propose that
‘regulated recruitment’ is a guiding principle in most cellular processes.
2.
Duncan SA, Navas MA, Dufort D, Rossant J, Stoffel M: Regulation of
a transcription factor network required for differentiation and
metabolism. Science 1998, 281:692-695.
3.
Perou CM, Sorlie T, Eisen MB, van de Rijn M, Jeffrey SS, Rees CA,
Pollack JR, Ross DT, Johnsen H, Aksien LA et al.: Molecular portraits
of human breast tumours. Nature 2000, 406:747-752.
4. Pandolfi PP: Transcription therapy for cancer. Oncogene 2001,
•• 20:3116-3127.
This recent review discusses the relationship between disruption of a
transcription factor and the onset of acute promyelocytic leukemia. It also
proposes the use of ATFs as therapeutic agents for treatment of this disease.
5.
Gottesfeld JM, Turner JM, Dervan PB: Chemical approaches to
control gene expression. Gene Expr 2000, 9:77-91.
6.
Dervan PB: Molecular recognition of DNA by small molecules.
Bioorg Med Chem 2001, 9:2215-2235.
7.
Denison C, Kodadek T: Small-molecule-based strategies for
controlling gene expression. Chem Biol 1998, 5:R129-R145.
As mentioned above, a desirable RD feature is the capacity
to provide a graded potency in response to physiological or
therapeutic cues. Initial steps towards this control were
achieved with the development of ligand-responsive ATFs
as well as RDs with tunable potency. A more ambitious
goal would be to design ATFs that are functionally
dormant in cells until activated by physiological cues
(Figure 4). Thus, ATFs that respond to natural cellular
signal transduction pathways, and perhaps even integrate
information from a combination of pathways, to regulate
the expression of a set of genes, present an exciting horizon
for the future.
8. Ansari AZ: Regulating gene expression: the design of synthetic
•
transcriptional regulators. Curr Org Chem 2001, 5:903-921.
This review outlines in detail the structure and function of natural and
artificial transcriptional regulators. It contains an overview of the advantages
and limitations of both protein and synthetic DBDs and a discussion of the
current understanding of the structure–function relationship of RDs.
Conclusions
12. Gerber HP, Seipel K, Georgiev O, Hofferer M, Hug M, Rusconi S,
Schaffner W: Transcriptional activation modulated by
homopolymeric glutamine and proline stretches. Science 1994,
263:808-811.
The modular nature of natural transcriptional regulators
has inspired a modular design of ATFs. The success of this
approach has allayed early concerns that stereospecific
presentation of regulatory modules would be necessary to
generate a functional ATF. This modularity allows tremendous
flexibility in design where desired features can be appended
to the essential functional modules. Current developments
in the design of ATFs allow us to probe the mechanistic
role of various architectural and temporal features of transcriptional regulators. As our design of ATFs becomes
more sophisticated, there is the real potential of developing
9.
Mannervik M, Nibu Y, Zhang H, Levine M: Transcriptional
coregulators in development. Science 1999, 284:606-609.
10. Smith RL, Johnson AD: Turning genes off by Ssn6-Tup1:
a conserved system of transcriptional repression in eukaryotes.
Trends Biochem Sci 2000, 25:325-330.
11. Garvie CW, Wolberger C: Recognition of specific DNA sequences.
•• Mol Cell 2001, 8:937-946.
This recent review outlines the architectural principles governing
protein–DNA interaction with relevant citations of the primary literature as
well as other reviews on this topic.
13. Courey AJ, Holtzman DA, Jackson SP, Tjian R: Synergistic activation
by the glutamine-rich domains of human transcription factor Sp1.
Cell 1989, 59:827-836.
14. Mermod N, Oneill EA, Kelly TJ, Tjian R: The proline-rich
transcriptional activator of Ctf/Nf-I is distinct from the replication
and DNA-binding domain. Cell 1989, 58:741-753.
15. Saha S, Brickman JM, Lehming N, Ptashne M: New eukaryotic
transcriptional repressors. Nature 1993, 363:648-652.
16. Bentley D: Coupling RNA polymerase II transcription with
pre-mRNA processing. Curr Opin Cell Biol 1999, 11:347-351.
Modular design of artificial transcription factors Ansari and Mapp
17.
Brown SA, Weirich CS, Newton EM, Kingston RE: Transcriptional
activation domains stimulate initiation and elongation at different
times and via different residues. EMBO J 1998, 17:3146-3154.
18. Blau J, Xiao H, McCracken S, Ohare P, Greenblatt J, Bentley D: Three
functional classes of transcriptional activation domains. Mol Cell
Biol 1996, 16:2044-2055.
19. Pabo CO, Nekludova L: Geometric analysis and comparison of
•• protein–DNA interfaces: why is there no simple code for
recognition? J Mol Biol 2000, 301:597-624.
The authors outline the problems that have been encountered in the numerous
attempts to deduce a straightforward code for the recognition of DNA
by proteins.
20. Urlinger S, Baron U, Thellmann M, Hasan M, Bujard H, Hillen W:
Exploring the sequence space for tetracycline-dependent
transcriptional activators: novel mutations yield expanded range
and sensitivity. Proc Natl Acad Sci USA 2000, 97:7963-7968.
21. Chin JW, Schepartz A: Concerted evolution of structure and
•
function in a miniature protein. J Am Chem Soc 2001,
123:2929-2930.
The authors describe a genetic selection with a novel minimal protein
scaffold to obtain compact DBDs.
22. Beerli RR, Barbas CF: Engineering polydactyl zinc-finger
transcription factors. Nat Biotechnol 2002, 20:135-141.
23. Neidle S: DNA minor-groove recognition by small molecules. Nat
Prod Rep 2001, 18:291-309.
24. Faria M, Giovannangeli C: Triplex-forming molecules: from
concepts to applications. J Gene Med 2001, 3:299-310.
25. Nielsen PE: Peptide nucleic acid targeting of double-stranded
DNA. Methods Enzymol 2001, 340:329-340.
26. Beerli RR, Dreier B, Barbas CF: Positive and negative regulation of
endogenous genes by designed transcription factors. Proc Natl
Acad Sci USA 2000, 97:1495-1500.
27.
Liu PQ, Rebar EJ, Zhang L, Liu Q, Jamieson AC, Liang YX, Qi H,
Li PX, Chen BL, Mendel MC et al.: Regulation of an endogenous
locus using a panel of designed zinc finger proteins targeted to
accessible chromatin regions — activation of vascular endothelial
growth factor A. J Biol Chem 2001, 276:11323-11334.
28. Seipel K, Georgiev O, Schaffner W: A minimal transcription
activation domain consisting of a specific array of aspartic-acid
and leucine residues. Biol Chem Hoppe Seyler 1994, 375:463-470.
29. Yaghmai R, Cutting GR: Optimized regulation of gene expression
•• using artificial transcription factors. Mol Ther 2002, 5:685-694.
This recent report outlines a systematic study of the regulatory potential of
zinc-finger- based ATFs in a variety of cultured cell lines.
30. Kuznetsova S, Ait-Si-Ali S, Nagibneva I, Troalen F, Le Villain JP, Harel
•
Bellan A, Svinarchuk F: Gene activation by triplex-forming
oligonucleotide coupled to the activating domain of protein VP16.
Nucleic Acids Res 1999, 27:3995-4000.
This is the first report demonstrating that a TFO can be covalently conjugated
to an activating region to generate an ATF that stimulates expression of a
targeted reporter in cultured cell ines.
31. Stanojevic D, Young RA: A highly potent artificial transcription
•
factor. Biochemistry 2002, 41:7209-7216.
A recent extension of [30•] in which the authors observe that adding a
flexible linker between the TFO and the activating region increases the
potency of the ATF.
32. Mapp AK, Ansari AZ, Ptashne M, Dervan PB: Activation of gene
•• expression by small molecule transcription factors. Proc Natl Acad
Sci USA 2000, 97:3930-3935.
This is the first demonstration of the idea that each component of a natural
transcriptional regulator can be replaced with a synthetic counterpart to yield
an ATF that functions well in an in vitro system. Perhaps most relevant for
ATF design, it was observed that the length and composition of the linker has
a profound effect on the potency of the ATF.
33. Ansari AZ, Mapp AK, Nguyen DH, Dervan PB, Ptashne M: Towards a
minimal motif for artificial transcriptional activators. Chem Biol
2001, 8:583-592.
34. Nyanguile O, Uesugi M, Austin DJ, Verdine GL: A nonnatural
transcriptional coactivator. Proc Natl Acad Sci USA 1997,
94:13402-13406.
35. Ma J, Ptashne M: A new class of yeast transcriptional activators.
Cell 1987, 51:113-119.
771
36. Lu XY, Ansari AZ, Ptashne M: An artificial transcriptional activating
•
region with unusual properties. Proc Natl Acad Sci USA 2000,
97:1988-1992.
The authors report the isolation of ‘non-acidic’ octapeptides that function as
well as the strongest natural yeast activators. As such, they are the smallest
eukaryotic artificial activators.
37.
Frangioni JV, LaRiccia LM, Cantley LC, Montminy MR: Minimal
activators that bind to the KIX domain of p300/CBP identified by
phage display screening. Nat Biotechnol 2000, 18:1080-1085.
38. Han Y, Kodadek T: Peptides selected to bind the Ga180 repressor
are potent transcriptional activation domains in yeast. J Biol Chem
2000, 275:14979-14984.
39. Sengupta DJ, Wickens M, Fields S: Identification of RNAs that bind
to a specific protein using the yeast three-hybrid system. RNA
1999, 5:596-601.
40. Fisher AL, Caudy M: Groucho proteins: transcriptional
corepressors for specific subsets of DNA-binding transcription
factors in vertebrates and invertebrates. Genes Dev 1998,
12:1931-1940.
41. Andrulis ED, Neiman AM, Appulla DC, Sternglanz R: Perinuclear
localization of chromatin facilitates transcriptional silencing.
Nature 1998, 394:592-595.
42. Dickinson LA, Gulizia RJ, Trauger JW, Baird EE, Mosier DE, Gottesfeld
JM, Dervan PB: Inhibition of RNA polymerase II transcription in
human cells by synthetic DNA-binding ligands. Proc Natl Acad Sci
USA 1998, 95:12890-12895.
43. Janssen S, Cuvier O, Muller M, Laemmli UK: Specific gain- and loss
•
of-function phenotypes induced by satellite-specific DNA-binding
drugs fed to Drosophila melanogaster. Mol Cell 2000, 6:1013-1024.
In this elegant study, the authors use polyamide-based ATFs to induce
phenotypic changes in D. melanogaster. Here, the ATF functioned in an
organism rather than in cultured cell lines.
44. Cutrona G, Carpaneto EM, Ulivi M, Roncella S, Landt O, Ferrarini M,
•
Boffa LC: Effects in live cells of a c-myc anti-gene PNA linked to a
nuclear localization signal. Nat Biotechnol 2000, 18:300-303.
The authors demonstrate that covalent attachment of a nuclear localization
sequence to a PNA results in transport into cells and nuclei. The NLS does,
however, appear to interfere with the ability of the PNA to target its specific
DNA site.
45. Boffa LC, Scarfi S, Mariani MR, Damonte G, Allfrey VG, Benatti U,
Morris PL: Dihydrotestosterone as a selective cellular/nuclear
localization vector for anti-gene peptide nucleic acid in prostatic
carcinoma cells. Cancer Res 2000, 60:2258-2262.
46. Supekova L, Pezacki JP, Su AI, Loweth CJ, Riedl R, Geierstanger B,
•• Schultz PG, Wemmer DE: Genomic effects of polyamide/DNA
Interactions on mRNA expression. Chem Biol 2002, 9:821-827.
The authors used genome-wide transcript profiling in colon cancer cells to
investigate the targeting specificity of a polyamide ATF designed to inhibit
an endogenous gene. They conclude that transcript profiling will be an
important tool in future design efforts.
47.
Zhang B, Kraemer B, Sengupta DJ, Fields S, Wickens M: Yeast
three-hybrid system to detect and analyze interactions between
RNA and protein. Methods Enzymol 1999, 306:93-113.
48. Lin H, Cornish VW: In vivo protein–protein interaction assays:
beyond proteins. Angew Chem 2001, 40:871-875.
49. Belshaw PJ, Ho SN, Crabtree GR, Schreiber SL: Controlling protein
association and subcellular localization with a synthetic ligand
that induces heterodimerization of proteins. Proc Natl Acad Sci
USA 1996, 93:4604-4607.
50. Arora PS, Ansari AZ, Best T, Ptashne M, Dervan PB: Design of
artificial transcriptional activators with rigid poly-L-proline linkers.
J Am Chem Soc 2002, in press.
51. Baron U, Bujard H: Tet repressor-based system for regulated gene
expression in eukaryotic cells: principles and advances. Methods
Enzymol 2000, 327:401-421.
52. Altoff EA, Cornish VW: A bacterial small-molecule three-hybrid
system. Angew Chem Int Ed Engl 2002, 41:2327-2330.
53. Beerli RR, Schopfer U, Dreier B, Barbas CF: Chemically regulated
•
zinc finger transcription factors. J Biol Chem 2000,
275:32617-32627.
This study represents a first step in designing ATFs that can be used to
regulate gene expression in a temporal manner. This approach offers much
772
Model systems
promise for the design of ATFs that respond to one or more physiological
cues (Figure 4).
54. Belitsky JM, Leslie SJ, Arora PS, Beerman TA, Dervan PB: Cellular
•
uptake of N-methylpyrrole/N-methylimidazole polyamide-dye
conjugates. Bioorg Med Chem 2002, 10:3313-3318.
In this recent report, the authors examined the cellular uptake of fluorescently labeled polyamides in seven different cell lines and observed that five
cell lines showed no discernable nuclear localization of the conjugates. Also
noteworthy is the observation that cell fixing caused rapid uptake into the
nuclei of all cell types examined, indicating the importance of live cell controls.
55. Schwarze SR, Dowdy SF: In vivo protein transduction: intracellular
delivery of biologically active proteins, compounds and DNA.
Trends Pharmacol Sci 2000, 21:45-48.
56. Hawiger J: Noninvasive intracellular delivery of functional peptides
and proteins. Curr Opin Chem Biol 1999, 3:89-94.
57.
Rothbard JB, Kreider E, VanDeusen CL, Wright L, Wylie BL,
Wender PA: Arginine-rich molecular transporters for drug delivery:
role of backbone spacing in cellular uptake. J Med Chem 2002,
45:3612-3618.
58. Torgerson TR, Colosia AD, Donahue JP, Lin YZ, Hawiger J:
Regulation of NF-kappa B, AP-1, NFAT, and STAT1 nuclear import
in T lymphocytes by noninvasive delivery of peptide carrying the
nuclear localization sequence of NF-kappa B p50. J Immunol
1998, 161:6084-6092.
59. Marmorstein R, Carey M, Ptashne M, Harrison SC: DNA recognition
by GAL4 – structure of a protein••DNA complex. Nature 1992,
356:408-415.
60. Wang BS, Grant RA, Pabo CO: Selected peptide extension
contacts hydrophobic patch on neighboring zinc finger and
mediates dimerization on DNA. Nat Struct Biol 2001,
8:589-593.
61. Kielkopf CL, White S, Szewczyk JW, Turner JM, Baird EE, Dervan PB,
Rees DC: A structural basis for recognition of A•T and T•A base
pairs in the minor groove of B-DNA. Science 1998, 282:111-115.