|
|
||||||||
|
FJ
EXPRESS SUMMARY ARTICLE The Full-length version of this article is also available, published online January 5, 2001 as doi:10.1096/fj.00-0459fje. |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||



* Medizinische Poliklinik, Klinikum Innenstadt, Ludwig-Maximilians-University of Munich, Germany;
Department of Immunology, Berlex Biosciences, Richmond, California;
Institute of Mammalian Genetics, GSF-National Research Center for Environment and Health, Neuherberg, Germany and Genomatix Software GmbH, Munich, Germany; and
National Cancer Institute-Frederick Cancer Research and Development Center, Frederick, Maryland
3Correspondence: Medizinische Poliklinik, Ludwig-Maximilians-Universität München, Schillerstr. 42, 80336 München. E-mail: nelson{at}medpoli.med.uni-muenchen.de
SPECIFIC AIMS
We sought to use the RANTES/CCL5 gene as a molecular probe in order to explore biological mechanisms involved in the control of tissue-specific gene expression. The goal of this study was to characterize transcriptional events responsible for LPS-induced up-regulation of the RANTES/CCL5 gene in monocytic cells. The detailed functional data were then used to identify potential coregulated genes by using unique bioinformatic tools.
PRINCIPAL FINDINGS
1. A Sp1-Rel p50/p50 composite element (promoter module) mediates
LPS responsiveness in monocytic cells
Heterologous promoter reporter gene assays were performed to
assess the functionality of RANTES site AB (R(AB), -73/-34), tandem
B elements, which were identified as an LPS-inducible region in
DNase I footprinting. Transient transfection of MM6 cells with
luciferase constructs containing dimers of R(AB) demonstrated that the
R(AB) region can efficiently confer LPS inducibility. However, the
selective mutation of either R(A) or R(B) in the context of the dimer
resulted in loss of LPS inducibility (Fig. 1
). As both elements are required to transfer LPS inducibility to a
second promoter, region R(AB) meets the criteria for a promoter module.
In subsequent mutation/transient transfection experiments within the
context of the RANTES/CCL5 promoter, R(A) was found to be required for
both constitutive and LPS-induced activity of the RANTES/CCL5 promoter
in monocytic MM6 cells. In contrast, region R(B) did not appear to
affect constitutive promoter reporter gene expression but was required
for LPS inducibility.
|
In EMSA experiments with unstimulated MM6 nuclear extracts, Sp1 and a
slight amount of Rel p50 and p52 proteins were found to bind to R(A).
Slight Rel p50 and p52 binding was seen at R(B). Following LPS
stimulation, pronounced binding of Rel p50/p65 to site R(A) and an
induced binding of Rel p50/p50 to site R(B) were found (Fig. 2
).
|
The functional roles of Sp1 and p50/p65 in transcriptional control
through region R(A), and of Rel protein interaction at R(B), were
investigated by directed modification of sites R(A) and R(B) in the
context of the 194/RANTES promoter. MatInspector
(http://genomatix.gsf.de) was used to design mutants of site R(A) that
resulted in a DNA sequence that would bind either p50/p65 or Sp1, but
not both. A modification of site R(B) was identified that would test
the relative importance of Rel p50/p50 versus p50/p65 binding by
selectively mutating R(B) to bind p50/p65 heterodimers preferentially.
The mutations confirmed by EMSA analysis were then introduced into the
194/RANTES promoter reporter gene construct and were tested in
transient transfection of MM6 cells. The results showed that both Sp1
and NF-
B can activate transcription at site R(A). When region R(B)
was altered to preferentially bind p50/p65, a complete loss of
LPS-induced reporter gene expression was seen again, which suggests
that LPS affects mediated through R(B) depend upon the induced Rel
p50/p50 binding activity.
2. Constitutive C/EBP binding to R(E) is important in LPS control
of RANTES/CCL5 expression
RANTES site E(R(E)) is contiguous, with a large complex identified
by DNase I footprinting (-137/-104). EMSA oligonucleotide competition
and supershift analysis showed that in monocytes, region R(E)
constitutively binds C/EBP-alpha, -beta, -delta, and -epsilon. No
obvious change in the composition of the various complexes was seen
after LPS stimulation. Mutation of the C/EBP consensus caused a
reduction of between 40% and 50% of LPS-stimulated RANTES/CCL5
promoter reporter gene activity and approximately 15% reduction in
constitutive activity.
3. Computer modeling of the organization of functional elements
With FastM, a software tool shaped to design models based on
weight matrix, orientation, and distance parameters, a series of
promoter models for the R(E) and R(AB) elements were developed.
ModelInspector was used to search for promoters showing similar
organization (http://genomatix.gsf.de).
R(AB) represents a composite element or promoter module. Together with
the TATA box, R(A) and R(B) were used as the basic elements for the
models illustrated in Fig. 3B
. The RANTES/CCL5 promoter contains consensus sequences with
defined core and matrix parameters for p50/p65 and Sp1 in region R(A),
and p50/p65 and p50/p50 in region R(B) (MatInspector). Three possible
combinations of factors binding at R(A) and R(B) were chosen to
generate models 1, 2, and 3. Models 2 and 3 represent combinations that
were functional in LPS induced RANTES/CCL5 promoter reporter activity
in MM6 cells. Model 1 was tested with both R(A) and R(B) binding Rel
p50/p65 heterodimers. Models 4, 5, and 6 were derived from models 1, 2,
and 3 by adding a weight matrix for C/EBP. ModelInspector was applied
to search human, rodent, other mammalian, other vertebrate, and viral
sections of EMBL database (release 63) and Philipp Buchers Eukaryotic
Promoter Database (EPD) for the six models. The output was filtered for
sequences with promoter annotation.
|
CONCLUSIONS
The RANTES/CCL5 gene is an excellent example of the flexibility and selectivity that underlie the tissue- /signal-specific regulation of gene expression. Diverse cell types express the RANTES/CCL5 gene. It is expressed constitutively by some cells (e.g., megakaryocytes) and is up-regulated by other cells in response to specific stimuli. The molecular mechanisms involved in the up-regulation of RANTES/CCL5 expression can vary significantly between the various tissues that express the gene. This condition is seen in the kinetics, duration, and stimulus-specific control of its expression.
Human monocytic cells up-regulate RANTES/CCL5 mRNA rapidly and transiently in response to lipopolysaccharide. Here we have characterized transcriptional elements within the RANTES/CCL5 promoter responsible for this up-regulation and used the experimental data to model higher-order organization of promoters. A FastM model (model 3), which contained the Sp1-p50/p50 promoter module functionally described here identified a series of promoters within the human and vertebrate (rodent/other mammalian/other vertebrate) database sections, including genes described as transcriptionally regulated by LPS, for example, the high mobility group protein-1 (HMG-1) and immunoglobulin heavy chains. A second group of promoters identified by the model represents promoters for gene products known to be involved in the biology of LPS, which are therefore excellent candidates for regulation by LPS; for example, the human phospholipid transferase protein (PLTP), which will complex and neutralize LPS.
Model 6, an extension of model 3, which included a matrix for a C/EBP element, identified a subset of the promoters found for model 3. Among these are the promoters for human and porcine beta-myosin heavy chains, rat GTP cyclohydrolase I, human and sea urchin metallothioneins 1B, and snake phospholipase A2. All of these genes are described as up-regulated by LPS, which suggest that model 6 can describe the higher-level organization of a class of LPS-responsive genes.
The kinetics and molecular control mechanisms involved in
RANTES/CCL5 transcription/expression vary among cell types. The
modeling approach described here may be used to predict genes that are
coregulated with RANTES/CCL5 in other tissue types in response to other
signals. Regions R(E) and R(AB) are important for RANTES/CCL5
expression in T cells. T cells up-regulate RANTES/CCL5 late (35
days) after mitogenic activation of resting peripheral blood T cells.
In T cells, Rel p50/p50 constitutively binds R(B). A novel
transcription factor called RFLAT-1 (RANTES factor of late activated T
cells 1) is induced late after T cell activation and mediates
transcription through R(A). In addition, a yet unidentified factor
up-regulated late in T cell activation (R(E)FLAT) binds to R(E) and
regulates transcription late after activation. In astrocytes,
RANTES/CCL5 induction by IL-1ß involves Rel p50/p65 binding to both
R(A) and R(B) but no C/EBP interaction at site R(E). However, other
elements proximal to the R(E) site, and further upstream in the
promoter sequence, were found to play important roles (Fig. 3A
).
The organization of factor interaction on the RANTES/CCL5 promoter may help characterize the pathways used for tissue-specific control of expression. The integration of functional biology with computer modeling of promoter organization will provide an invaluable tool to the study of transcriptional regulation in the approaching post-human genome project era. The approaches used here demonstrate the efficacy of using computer models to predict potential co-regulated genes. These techniques can be applied to diverse questions in biology, including the engineering of synthetic promoters, the higher-order analysis of clustered genes from micro array analysis, and the identification of additional target genes.
FOOTNOTES
1 To read the full text of this article, go to
http://www.fasebj.org/cgi/doi/10.1096/fj.00-0459fje ; to cite this
article, use FASEB J. (January 5, 2001)
10.1096/fj.00-0459fje ![]()
2 These authors contributed equally to this
work. ![]()
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |