Epigenetic mechanisms of cellular memory by sdfsb346f


More Info

        Epigenetic mechanisms of
                                      cellular memory

   The many mitotic cell divisions following the stage of the fertilized egg to the
   adult organism require mechanisms ensuring that defined cellular states,
   determined at specific times during development, are faithfully inherited. Since
   cellular identities are based on a carefully orchestrated gene expression
   pattern of developmental regulators, like the HOX genes, maintaining identity is
   equivalent to maintaining the differential gene expression patterns. This basic
   cellular function is controlled at the level of chromatin. The proteins of the
   Polycomb group (PcG) and trithorax group (trxG) generate stable and heritable
   chromatin structures, maintaining the expression of developmental control
   genes. Our major research focus in the lab deals with the molecular
   mechanisms by which the PcG and trxG proteins control chromatin structure
   and in particular how chromatin-based epigenetic information controlling gene
   expression is faithfully maintained during DNA replication and cell division. This
   mechanism of maintaining determined states has been termed “cellular

Epigenetic mechanisms of cellular memory

   Mechanisms of cellular memory are required to ensure that gene expression
   patterns, defining particular cell lineages, are faithfully inherited during cell
   division. The Polycomb (PcG) and the trithorax group (trxG) genes encode
   proteins that lock a gene expression state at the level of chromatin
   organization, by generating heritable chromatin structures that are transmitted
   during DNA replication and mitosis. PcG and TrxG proteins bind to

                    Lab members
                    Renato Paro
                    University of Heidelberg
                    Im Neuenheimer Feld 282
                    69120 Heidelberg

                    Telefone: +49-62 21 4 68 78
                    Telefax: +49-0 62 21 54 58 91
                    e.mail: paro@zmbh.uni-heidelberg.de


                    Postdoctoral fellows:          Ph.D. students:
                    Dr. Christian Beisel           Nara Lee
                    Dr. Leonie Ringrose            Ana-Laura Monqaut
                    Dr. Muhammad Tariq             Sabine Schmitt
                                                   Gero Strübbe
                    Techn. assistants:
                    Ehret, Heidi
                    Hagar, Sylvia
                    Frank, Andrea

chromosomal elements termed PREs (PcG Response Elements), and interact with the histones and
the basic transcription apparatus to keep their target genes either repressed by the PcG proteins or
active by the TrxG proteins.

Assessing the heritability of the PRE epigenetic state
Changing cellular fates requires changing PRE epigenetic states. What is the process that switches a
silenced PRE into a mitotically heritable activated PRE? Our working hypothesis is that PREs are
silenced by default. Transcription through PREs remodels chromatin by setting positive epigenetic
marks (i.e. acetylated histones) and thus prevents the PcG silencing complexes from binding to
particular PREs, thus keeping the associated target gene active (figure 1). To demonstrate a causal
relationship of this type of non-coding transcription, we established a set of reporter gene constructs to
determine whether transcription through a PRE from a dedicated promoter (Actin 5C flanked by
Cre/lox sites for removal by recombination) would inhibit silencing. Indeed, these constructs always
gave strong miniwhite reporter expression. Upon removal of the promoter silencing of miniwhite is
restored. The attractive hypothesis ensuing from our observation is that the transcriptional memory
mechanism would only require the propagation of a positive epigenetic mark (set by the transcription
process) through DNA replication and mitosis. This positive mark would prevent reestablishment of
PcG-silencing at the particular PRE in the next interphase of the daughter cells and thus ensure an
active expression state of the associated target gene.

                                                                              Figure 1:
                                                                              In situ hybridizations to
                                                                              embryos and larval brain
                                                                              and CNS show for the
                                                                              tailless gene that the PRE
                                                                              is expressed in the same
                                                                              tissue as the mRNA.

Quantitative studies to determine interactions of epigenetic components
Based on the fact that PRE targeting of PcG complexes is dependent on a set of defined DNA binding
                                  proteins, we have developed an algorithm capable of predicting
                                  PRE sequences in the Drosophila genome. This allowed us to
                                  identify over 150 potential candidate genes subjected to PcG/TrxG
                                  regulation. The gene categories represented cover early functions
                                  involved in segmentation and organogenesis as well as basic
                                  cellular regulatory features like cell cycle and cancer control. We
                                  are currently developing tools to isolate native PREs in order to
                                  identify the bound constituents and assess the tissue specificity of
                                  the regulatory complexes.
                                  Target gene specificity of PcG-silencing is primarily achieved
                                  through the PRE sequence. However, epigenetic marks like histone
                                  H3 methylated at K9 (H3K9me) and K27 (H3K27me) contribute to
                                  PcG repression, as the Polycomb (PC) protein can interact with
                                  these moieties through its chromo domain. Using polytene
                                  chromosome immunostainings we show that the two methyl marks
                                  colocalize with PC in distinct but overlapping patterns. We find that
                                  high levels of methylation and PC binding at the promoter do not
                                  prevent strong transcription, suggesting rather that multiple,

                                   Figure 2:
                                   Schematic model of PC binding to PREs used for the mathematical
                                   modeling of peptide competition (top) and simulation of binding
                                   shown in the bottom part.

regulated interactions between methylated PREs and promoters are required to create a silenced
locus. In a cellular assay where we compete PC binding with various modified histone tail peptides, we
show that the stability of PC binding is different at different loci and correlates with local differences in
histone methylation and transcriptional status. We applied mathematical modeling (figure 2) to assess
the kinetic nature of this observed behavior, showing theoretically that weak interactions are sufficient
to maintain stable local concentrations of PcG proteins, which are in constant dynamic exchange with
their target sites, offering opportunities for locus-specific regulation.

Developing a systematic approach to study the epigenetic network on a whole-genome scale
We have developed microarray and high-throughput approaches to study PcG/TrxG function at a
genome-wide level. We have generated a new gene annotation of the Drosophila genome yielding the
number of approximately 17.000 – 17.500 total Drosophila genes. A microarray containing cDNA
sequences (PCR fragments of approx. 500 bp in length) has been developed and successfully
employed to assess gene expression patterns during various stages of development (for more
information see http://hdflyarray.zmbh.uni-heidelberg.de). Material from this study (Oligos, PCR
fragments and microarrays) is being provided to the scientific community by the company Eurogentec.
In addition, the PCR-set was utilized for the production of dsRNAs applicable in whole genome RNAi
screens of Drosophila cells. This collaborative effort with the group of Norbert Perrimon (Harvard
Genetics) culminated in the establishment of an RNAi Screening Center (www.flyrnai.org ) funded by
NIH and servicing the scientific community utilizing these new tools for high-throughput whole-genome
We have used the knowledge gained in this genomics project, to establish the ChIP-on-Chip
methodology in order to map protein distributions using microarrays. We have characterized the
binding profiles of several members of the PcG/TrxG as well as compared them to the profiles of
histones with particular modifications on a tiling array of the BX-C and ANT-C as well as other PRE-
containing genes (figure 3). Besides establishing a genomic microarray, we have elaborated methods
to produce homogeneous material for ChIP. To this purpose, we have established the BirA ligase
technology in Drosophila, allowing particular proteins in defined cells to be marked with a biotin
moiety, in order to affinity purify the protein complexes.

                                                                 Result of a ChIP-on-Chip experiment
                                                                 comparing the binding profile of the PC
                                                                 protein with acetylated histones in the

The availability of the microarray and RNAi technology will give us a very powerful tool to
systematically assess PcG/TrxG function on a genome wide scale. The questions that can be tackled
at this level of complexity are i) Do the candidate genes identified by the PRE algorithm become
globally deregulated in PcG or trxG mutants, ii) How is tissue-specific control of these broad classes of
different PcG/TrxG target genes achieved, iii) How do chromatin protein distributions in the BX-C/ANT-
C relate to other candidate PRE genes, is there a common profile motif. Indeed, it is becoming quite
evident that PcG/TrxG regulation is widespread and involved in many diverse processes from plant
seed development to mammalian stem cell renewal. Thus, it is important to continue on the large basis
of information that has been generated in the model Drosophila to better understand the basic
mechanism and, thus, to be able to extrapolate knowledge to more complex related phenomena like
tissue remodeling or cancer epigenetics in mammalian organisms.

To top