Lab Web Site
Dr. Levens's research is focusing on the transcriptional regulation of
c-myc and the elucidation of the roles of conformation and topology-selective,
sequence-specific DNA binding transcription factors. The c-myc gene employs
multiple cis-elements, upstream and downstream of its promoters, P1 and
P2. The goal is to illuminate the mechanisms through which the input of
multiple factors sets the expression level of this critical proto-oncogene.
Two elements and their associated trans-factors, first described in this
group, display unusual properties in vitro and in vivo which may serve
to highlight important new regulatory modes.
studies identified a cis-element (1500 bp upstream pf P1) which bound
a factor downregulated with the cessation of c-myc transcription during
promonomyelocytic leukemia cell differentiation. This cis-element, FUSE
(far upstream element), was defined by deletional and mutational analysis;
the 70 kDa FUSE binding protein (FBP) was purified and, following peptide
cleavage and microsequencing, the cDNA was cloned. FBP possesses a DNA
binding motif specific for particular sequences only when contained in
single-stranded or negatively supercoiled DNA. Additional dissection of
FBP revealed both a potent transcription activation domain at its carboxyl-terminus
and a strong repression activity within the amino-terminus. Repression
may require a partner protein recently identified in our group. Coupling
these properties with the unusual DNA binding specificity indicates that
FBP may participate in the regulation of c-myc by sensing conformational
perturbation of FUSE in response to processes such as transcription. Thus
the c-myc promoter may directly feedback-regulate itself, requiring no
diffusible mediator. FBP expression is shut off during differentiation
and increased when cells are induced to proliferate.
The conformation of the FUSE was analyzed in vivo. One strand of the FUSE
is occupied in vivo when c-myc is expressed, but is vacant when c-myc
is silent. Thus the opening and closing of the FUSE (perhaps mediated
through transcription-driven supercoiling) may be a key choke-point for
c-myc control. The potential significance of FBP and FUSE is underscored
by the discovery of two homologous proteins, FBP2 and FBP3. The mechanistic
illumination of FUSE and the FBP's should yield new insights into the
integration of multiple pathways onto a complex promoter.
Second, studies of the CT-element, 100 to 150 bp upstream of c-myc P1,
underscore many of the principles operating at the FUSE. The CT-element,
essential for P1, also stimulates expression from P2. A complex array
of factors interacts with this five-times-repeated element. One of these
factors is hnRNP K, which unexpectedly possesses the same general DNA
binding protein architecture as FBP. hnRNP K binds to one strand of the
CT-element either as a single strand or when embedded in negatively supercoiled
DNA. Comparison of the CT-element/hnRNP K complex in vitro and in vivo
provides strong evidence that hnRNP K is at the CT-element in living cells.
Further work indicates that the CT-element is a flexural and torsional
hinge in vivo and in vitro, probably reflecting intrinsic conformational
To prove that hnRNP K interacts with CT-elements in vivo, the powerful
VP16 activation domain was fused to hnRNP K's carboxyl terminus. This
fusion protein activates CT elements in supercoiled but not linearized
plasmids when transfected into Hela cells, showing that DNA conformation
and gene expression are coupled.
hnRNP K activates RNA polymerase II mediated expression both in vivo and
in vitro. Recent experiments show that hnRNP K can bind with linear DNA
and modify transcription in response to ongoing transcriptionally generated
torsion. The potential significance of hnRNP K is underscored by its interactions
with c-src, vav, and TATA binding protein. It also binds tightly with
the retinoblastoma protein. Our recent work suggests that hnRNP K interacts
with major players governing the cell cycle and controlling apoptosis,
and may itself participate in these processes.
The importance of factors interacting with and modifying DNA conformation
in vivo is highlighted by our discovery of protein-dependent conformational
marking of active genes in mitotic chromosomes. This observation suggests
a mechanism for selecting genes for transcriptional restart following
pan-genomic mitotic repression.
Further studies of single-stranded/topology-dependent factors, and of
the rules coupling DNA conformation and topology with transcription, replication,
repair, or recombination of the c-myc gene, promise to help us understand
the expression of c-myc under physiological and pathologicalcircumstances.
- Michelotti, GA,
et al. Mol Cell Biol 1996; 16:2656-69.
- Tomonaga, T; Levens,
D. Proc Natl Acad Sci USA 1996; 93:5830-5.
- Michelotti, EF,
et al. Mol Cell Biol 1996; 16:2350-60.
- Duncan, R, et al.
Genes and Dev 1994; 8:465-80.
- Siegfied Janz,
M.D., D.Sc., NIH
- Dirk Eick, Ph.D.,
Institut fuer Klinische Molekularbiologie