Patent Application
20090233317
The present invention relates to a novel corepressor whose function has been substantially elucidated. Various possible uses of this corepressor are disclosed.
The designation GRIM1 which has been chosen for the corepressor of the invention represents the abbreviation of the designation “Global Repressor Involved in Myogenic Differentiation”. This name was chosen because the cloned factor plays an important part in particular during skeletal muscle differentiation.
In recent years a large number of molecules which prove to be responsible for direct transcriptional regulation has been found. Besides the DNA-binding factors (in the narrower sense referred to as transcription factors), there are regulatory molecules which, as coactivators, facilitate gene expression or which, as corepressors, actively bring about transcriptional repression. Cofactors are promiscuous and can be recruited in various combinations by various DNA-binding partners. The alternation between associated coactivator and corepressor complexes is an important regulatory step within cellular differentiation processes.
For the example of skeletal muscle differentiation, over the course of years a cascade of transcription factors which are necessary for myogenesis has been found. Besides other transcription factors, the bHLH proteins MyoD, myf5, MRF4 and myogenin bring about execution, in a proliferating myoblast precursor cell, of the genetic program which results in a terminally differentiated functional skeletal muscle cell. During the course of this phylogenesis, the cell passes through a cell cycle arrest, and it fuses with other committed muscle cells to give multinuclear myoblasts and expresses skeletal muscle-specific structural and metabolic enzymes.
The associated cofactors acting within this process have been investigated only in recent years. Positively regulated expression of myogenically specific genes takes place for example through a functional association of MyoD with the coactivator protein p300. Acetylated MyoD has far greater transcriptional activity than unmodified protein.
On the other hand, there are certain corepressors which actively prevent targeted repression of the expression of muscle-specific genes in proliferating cells. Thus, for example, a functional association between MyoD and the corepressor N—CoR has been described, and in this way revealed a new field of activity for the cofactors which had previously been assigned only to nuclear hormone receptors. In addition, members of the histone deacetylase (HDAC) family have been described as being involved in skeletal muscle differentiation (C2C12 cell culture model) and showing a subcellular relocalization during the first steps of differentiation.
Association of HDACs with active repression is regarded as one of the basic requirements for negative transcriptional regulation, because the reaction catalyzed by the HDACs is transmitted directly to the chromatin and provides a mechanistic explanation for the observed effects.
The data shown in this application describe GRIM1 as a novel “bona fide” transcriptional repressor with a novel mechanism of repression. GRIM1 is not associated with an HDAC activity, and the GRIM1-mediated repression cannot be influenced by specific HDAC inhibitors. GRIM1 is able to inhibit directly, via acidic domains in the N and C termini, the acetylation of histone N termini (one of the preconditions for directed transcriptional activation). Repression domains which have not to date been characterized in detail within GRIM1 likewise have high potential for repression of transcription activity. GRIM1 has the potential in transient transfections for dose-dependent repression of both complex and synthetic minimal promoters. Based on these data, the newly cloned factor has been designated GRIM1 (Global Repressor Involved in Myogenic differentiation).
GRIM1 shows relocalization during skeletal muscle differentiation, although at a distinctly later time than previously described for the HDACs 4, 5 and 7 in connection with MEF2-specific transcription. The available data suggest that the repression potential of GRIM1 must not be switched off until later times during skeletal muscle differentiation so that terminal differentiation of the myotubes is possible. In addition, GRIM1 likewise shows a subcellular change of localization during preadipocyte differentiation.
One function of GRIM1 is involvement as repressor in regulating the differentiation of skeletal muscle cells. The data show that GRIM1 shows a distinctly different role than the involved factors previously described. Elucidation of the in vivo role of GRIM1 takes place via GRIM1 knockout animals and transgenic mice. In the transgenic animals, GRIM1 mutants will express under skeletal muscle-specific promoters which can no longer reach the nucleus or which can no longer be transported out of the nucleus. The phenotype of the mice which results from the generated atypical localization of GRIM1 will contribute further to understanding the function of GRIM1 within differentiation processes.
Application of the present invention is particularly in the area of degenerative muscle disorders or in controlling muscle atrophy in the aging man. The therapeutic modification of the function of GRIM1 may in such cases counteract premature muscle degeneration or actively intervene in processes which build up muscle.
The present invention therefore relates firstly to polypeptides having a sequence of at least 20 consecutive amino acids from the sequence of hsGRIM1 (hs=homo sapiens) having the sequence ID No. 3.
The polypeptides of the invention preferably have a sequence of at least 40 consecutive amino acids.
The polypeptides more preferably have a sequence of at least 80 consecutive amino acids, even more preferably a sequence of at least 120 consecutive amino acids and most preferably the polypeptides of the invention have a sequence of at least 200 consecutive amino acids from Seq. ID No. 3.
The invention further relates to polypeptides which have a sequence of at least 20 consecutive amino acids from the mmGRIM sequence (mm=mus musculus) having the Seq. ID No. 4.
These polypeptides more preferably have a sequence of at least 40 consecutive amino acids, even more preferably a sequence of at least 120 consecutive amino acids and very particularly preferably a sequence of at least 200 consecutive amino acids of Seq. ID No. 4.
The present invention further relates to antibodies which bind specifically to an epitope of a polypeptide of the invention. The antibodies of the invention can be prepared by customary standard methods. Either the polypeptides coding for GRIM1 can be used to immunize suitable laboratory animals such as, for example, rabbits or goats, and it is possible in this way to prepare suitable polyclonal antibodies. The antibodies of the invention may be directed against epitopes such as conformational epitopes, but also against peptides of the invention.
As an alternative to this it is possible to prepare suitable monoclonal antibodies by methods which have been well known since the publication by Köhler and Milstein in 1975 and belong to the standard repertoire of an average molecular biologist.
Antibodies of this type can be employed in therapy. However, to do this, it is usually necessary to prepare humanized antibodies, because antibodies having constituents derived from animals may cause unwanted side effects. If the binding regions of a suitable antibody have been sequenced, these sequences can be incorporated into a human basic antibody structure, and an antibody of this type can be used in therapy.
An alternative area of use of antibodies of this type is in diagnosis. It is perfectly possible for simple polyclonal antibodies, with which qualitative and/or quantitative detection of the presence of GRIM1 polypeptides in the cells or cell compartments of interest is possible, to suffice for this purpose. The average skilled worker can carry out suitable diagnostic methods with antibodies of this type. The present invention further relates to medicaments which comprise a polypeptide of the invention.
These medicaments are preferably employed for the treatment of disturbances of skeletal muscle differentiation and for the treatment of disturbances of fat cell differentiation.
The medicaments of the invention may comprise a polypeptide having the complete GRIM1 sequence or a suitable part of the sequence. The medicaments are administered in a suitable form to the patients to be treated. Oral or preferably parenteral pharmaceutical formulations are suitable in this connection.
The mechanism of action of the medicaments of the invention relies on modulation of the function of GRIM1. During skeletal muscle differentiation, GRIM1 is transported from the nucleus into the cytoplasm. The medicaments of the invention can intervene in this differentiation process by either specifically blocking importation or else blocking exportation. This can be achieved by incorporating the suitable segments of the polypeptides into appropriate formulations so that the target site of interest can be reached.
It has also been found within the scope of the present invention that GRIM1 is involved in further differentiation processes. GRIM1 shows a subnuclear relocalization during adipocyte differentiation. Involvement of GRIM1 in various other regulatory processes emerges from the findings of the present invention, in particular from the experimental data which show that GRIM1 shows a repression of a wide variety of complexes and synthetic promoters, and expression of GRIM1 is ubiquitous.
The present invention further relates to a method for identifying substances which influence the biological function of a polypeptide of the invention. This entails the substance to be identified being brought into contact with the polypeptide in a test system. It is possible with this test system to identify substances which interact with the GRIM1 polypeptide and inhibit or enhance the biological function of GRIM1.
The systems used according to the invention are preferably cellular differentiation systems which make use in particular of myogenic C2C12 cells or rapidly and inducibly differentiating 10T1/2 MyoD-ER cells. Detection of GRIM1 or the GRIM1 localization is possible for example immunohistologically. Suitable methods are, inter alia, the use of fluorescence-coupled or enzyme-coupled antibodies in a direct coloring step.
It is also possible to provide suitable cell lines which comprise an appropriate genetic construct with a GRIM1 gene. The substances to be investigated can then be brought into contact with the cells, and the effect of the substance to be investigated can be determined by appropriate comparative tests.
It is possible in a further embodiment to introduce suitable genetic constructs into laboratory animals and to use a transgenic animal for the test methods.
It is possible in a further embodiment of the invention to incubate recombinantly produced GRIM1 fragments in cell-free test systems with substances to be tested, and to obtain the desired result of the test by use of a suitable reporter molecule.
A further aspect of the invention is represented by the cDNA which codes for hsGRIM1 (homo sapiens) and which has a sequence of at least 1200 consecutive nucleotides from Seq. ID No. 1.
Another aspect of the invention is represented by the cDNA which codes for mmGRIM1 (mus musculus) and which has a sequence of at least 1200 consecutive nucleotides from Seq. ID No. 2.
The cDNAs of the invention can be employed in transfection vectors which have a cDNA of the invention. A preferred possibility in this connection is an adenoviral vector.
These transfection vectors are used to produce cells which have been transfected with such a vector.
In the examples which are described in detail below, besides customary standard methods, the following methods were preferably used, and the underlined sections of the methods are explained in more detail hereinafter:
The cells were cultured in the media recommended by the ATCC on 15 cm cell culture dishes under sterile conditions. Cells are subcultured in a ratio of 1:5 to 1:20 every two to three days and kept at a maximum of 60% confluence. Old medium is removed, and the cells are washed with PBS buffer and treated with trypsin/EDTA solution (0.25% trypsin, 0.04% EDTA in PBS buffer) until they are detached from the culture dish. The cells are isolated in fresh medium and subcultured on new subculture dishes.
To determine the transcriptional properties of hsGRIM1 proteins in eukaryotic cells, pCMX-hsGRIM1 expression plasmids were cotransfected with luciferase (LUC) reporter plasmids. Expression of the particular hsGRIM1 protein in a pCMX expression plasmid is under the control of the cytomegalovirus promoter. Luciferase expression in the LUC reporter plasmids used is controlled either by the thymidine kinase (TK) promoter, an E1b minimal promoter (TATA box), or by the respectively indicated complex promoters.
Transfection takes place by the calcium phosphate method or by the DOTAP liposomes method (Roche Diagnostics, product information). Unless otherwise indicated, 105 cells/well were seeded in 1 ml of culture medium in 12-well dishes. In the CaPO4 method, the plasmid DNA to be transfected is coprecipitated with calcium phosphate, and the precipitate is distributed to the cells capable of division. The calcium phosphate precipitate is prepared by preparing the mixture indicated below for duplicates in 12-well dishes:
While agitating (vortexing), 80 μl of 2×BES buffer are slowly and uniformly added dropwise to this mixture. The precipitate is incubated at RT for 10 min and distributed uniformly to the cells (80 μl per well) and, after 8 hours, removed by washing with PBS. The cells are incubated in fresh culture medium for a further 18 to 24 h.
Stable transfection of cell culture cells is carried out with the pcDNA6-His system supplied by Invitrogen. Cells are transfected as described above, and stable integration of the plasmids used is detected by the selection antibiotic blasticidin S. Individual clones are picked, expanded and tested by Western blotting and immunodetection for stable expression.
Yeast two-hybrid interaction assays, mammalian THS, pull-down analyses, immunoprecipitation of cellular proteins followed by immunodetection or in combination with in vitro translated 35S-methionine-labeled interactants, Phast gel system.
A protein is expressed as GST fusion protein in E. coli, immobilized on a GST-binding carrier material, and incubated with the potential interactant which is usually radiolabeled. In the subsequent washing steps, proteins which interact only weakly or not at all are washed off the carrier material or the carrier material-bound GST fusion protein. The amount of associated interactant can be determined by fractionation of the proteins by SDS-PAGE and subsequent autoradiography. To check the specificity of interaction between potential interactant and GST fusion protein, the interaction with GST alone is examined.
Aliquots of GST-GRIM1 fragments or GST-containing E. coli crude lysate are mixed in each case with 30 μl of glutathione (GSH)-Sepharose and incubated on a bohemian wheel at 4° C. for 1 h. The GSH-Sepharose is pelleted by centrifugation (1′ at 2000 rpm) and washed twice in pulldown buffer. The pellet is then resuspended in 500 μl of pulldown buffer and incubated with the in vitro translated, 35S-methionine-labeled potential interactant at 4° C. on the bohemian wheel for 1 h. After washing three times with pulldown buffer, the GSH-Sepharose pellet is boiled in 30 μl of SDS sample buffer, the proteins are fractionated by SDS-PAGE, and the gel is dried and then subjected to autoradiography.
If two proteins interact in solution, the stable complex of the two proteins can be precipitated with an antibody which is directed against only one of the two interactants. In this case, the protein-antibody complex is isolated from the solution with the aid of an antibody-binding carrier material (γ-bind G Sepharose, Pharmacia). To investigate interactions between GRIM1 and putative interactants, either native cell extracts or GRIM1-containing cell extracts are mixed with in vitro translated, 35S-methionine-labeled target proteins in 500 μl of IP buffer. 5 μg portions of anti-GRIM1 or nonspecific antibodies (rabbit IgG1) are mixed with 30 μl of preequilibrated γ-bind G Sepharose, and the mixture is incubated on a bohemian wheel at 4° C. for 60 min. The G-Sepharose is then pelleted at 2000 rpm for 1′ and washed three times with 300 μl each time of ice-cold IP buffer and finally the pellet is boiled in 30 μl of SDS sample buffer. The samples were finally separated by SDS-PAGE, and interaction was detected either by Western blotting or, in the case of the proteins translated in vitro in the presence of 35S-methionine, by autoradiography.
Indirect immunofluorescence, cell fractionation followed by immunodetection.
Cells are seeded in the desired cell density on autoclaved slides and fixed on the support with 5% paraformaldehyde for 10′. Cell membranes are made permeable for antibodies by incubation in 0.2% TritonX100 for 10′. Before the incubation with the primary antibodies, the cells are blocked in 0.2% gelatin (in PBS) at RT for 1 h. Anti-hsGRIM1 antibodies are diluted 1:500 in 0.2% strength gelatin solution and incubated at RT for 1 h. After washing three times with PBS for 5′ each time, the respective fluorochrome-coupled secondary antibody is added 1:2000 in 0.2% gelatin solution and incubated in the dark at RT for 30′. After the washing steps, the nucleus is counterstained with DAPI (1 μg/ml PBS, Roche Diagnostics) and the cells are finally washed twice in 0.1% TritonX100 in order to minimize autofluorescences. The slides are preserved in Fluoromount M (Southern Biotechnology Associates) and analyzed using a fluorescence microscope.
Intact nuclei can be isolated and lyzed separately by hypotonic lysis of the cell membrane. Cells are harvested in ice-cold PBS, pelleted and swollen in hypotonic lysis buffer (LB). The cell membrane is permeated by adding NP40 ad 0.1%, and the cytoplasmic constituents are released by gentle shaking and incubation on ice for 15′. Intact nuclei are pelleted by centrifugation at 14000 rpm for 15′, the CP fraction is removed quantitatively, and the protein concentration is determined. Nuclei are lyzed by incubation in nuclear lysis buffer (NLB) while vortexing for 15′. Debris is pelleted and the soluble nuclear preparation is removed. The quality of the preparation is examined by Western blotting and immunodetection with nucleus-specific (e.g. TIF2) and cytoplasmic (e.g. RasGAP) markers.
HDAC assay, INHAT assay, luciferase assays.
The histone deacetylation reactions which were carried out were carried out within the framework of a kit obtained from Upstate Biotechnology (3H-labeled histone H3 N-terminus as substrate). Alternative substrates used were histones 14C-acetyl-labeled with p300 or in vivo labeled histones (kindly provided by Dr Martin Göttlicher, Kernforschungszentrum Karlsruhe).
INHAT assays were carried out as described by Eckner et al. and by Seo et al. (Eckner et al., 1996, Genes and Development 10 (19), pp 2478-2490, Seo et al., 2001, Cell 104 (1), pp 119-130). Histones or histone mixtures are preincubated with the appropriate peptides in a total of 50 μl of HAT buffer, then mixed with bacterially expressed p300 HAT domains and 0.5 μCi of 14C-acetyl-coenzyme A, and incubated at 30° C. and 1000 rpm for 2 h. Reactions are stopped by adding SDS Laemmli buffer. Proteins are fractionated in denaturing SDS-PAGE, and the gels are incubated in Amplify (NEN) for 1 h. Acetylations are detected in autoradiographies.
Transiently transfected cells are washed once with PBS and lyzed by adding 50 μl of reporter lysis buffer (Promega) and further disrupted by freezing at −80° C. Cell lysates are transferred into Eppendorf reaction vessels, and cell detritus is pelleted at 14000 rpm for 2′. 10 μl portions of the lysates are pipetted into 96-well microtiter plates to determine the luciferase activity. The luciferase assay took place in an EG&G Berthold Microluminomat. 40 μl of luciferase assay buffer are injected into the 10 μl of cell lysate, and the resulting chemiluminescence is determined in a time integral of 10″. The measured values for the luciferase activity are related to the amount of protein present in the cell lysate. For this purpose, 3 μl portions of the lysates are mixed in 96-well microtiter plates with 100 μl of Bradford assay reagent (BioRad) diluted 1:5 with H2O, and the absorption of the solution is determined at 595 nm. Relative luciferase activities are found from the ratio of the luciferase activity to the protein absorption.
The invention is explained in more detail by the following examples:
The cloned human GRIM1 gene includes 2250 base pairs and codes for 749 amino acids. The cloned human GRIM1 sequence obtained by sequencing was deposited for the first time as unknown cDNA in 1999 under the accession number AL050019 (protein DKFZp564C186) as part of a “Random EST sequencing” project of the DKFZ in Heidelberg. A sequence which had been revised further was deposited in March 2001 (Wiemann et al., 2001). The function of the protein DKFZp564c186 has not previously been described. The cDNA of the invention is depicted in
It was possible to calculate from mouse EST databases by comparative analysis with the hsGRIM1 aa sequence from independent ESTs a full-length protein which has a total of 750 amino acids and a total of 60% identical and more than 85% homologous residues to the known hsGRIM1 sequence. Analysis of mouse cDNA libraries and mouse genomic libraries (Celera) have confirmed the calculated and expected mmGRIM1 aa sequence. It was possible to confirm the putative cDNA by RT-PCR analyses. The full-length cDNA of mmGRIM1 has not yet been described or deposited. The sequence having Seq. ID No. 2 is depicted in
Seq. ID No. 2 was obtained by screening mmEST dbs with the hsGRIM1 aa sequence. The resulting data were compared with mm cDNA dbs and mouse genomic dbs and the virtual mmGRIM1 cDNA was constructed. The sequence was confirmed by RT-PCR and genomic sequencing, and by a direct sequence comparison with the mouse genomic database of Celera Inc.
Demonstration of homology between hs and mmGRIM1 at the cDNA and at the protein level is shown in
The amino acid sequence of mmGRIM1 (mouse) is Seq. ID No. 4 and the amino acid sequence of hsGRIM1 (human) is Seq. ID No. 3.
All the functional and potentially interesting sequence motifs described in detail below are sequence-identically conserved in the mouse and the human protein (nuclear localization sequence NLS, nuclear export signal NES, so-called CoRNR box with which corepressors with their associated transcription factors are able to interact). Further motifs are HMG box-like domains within the C terminus and N terminus of GRIM1 of both species, which may possibly bring about contact with DNA and/or histones. In addition, in total 5 so-called LXXLL or related motifs are present within the primary hsGRIM1 aa sequence, an alpha-helical structure which has been reported to be responsible solely for contact of cofactors with their associated transcription factors. Within the mmGRIM1 primary sequence, only 2 of these motifs are conserved vis-á-vis the human sequence. No functional data are available on these motifs as yet. According to calculations, hsGRIM1 mainly has an alpha-helical structure (Columbia University PHD predict program) and is non-globular (GLOBE of the Columbia University PredictProtein server). GRIM1 is localized in the nuclei of proliferating cells, while GRIM1 is relocated from the nucleus to the cytoplasm within cellular differentiation systems (e.g. in C2C12, L6, 10T½ and 3T3-L1 cells).
For the purposes of the present invention, only hsGRIM1 was used on the basis of cDNA. However, it is assumed that corresponding results will be obtained on use of mouse cDNA. Various constructs were produced by deletion- and sequence-specific point mutagenesis from the human GRIM1 for detailed functional characterization and are compiled in the following tables.
Many of the mutants used carry specific tags such as the Gal4 DNA-binding domain of the yeast GAL4 transcription factor [Gal4-DBD], green-fluorescent protein [GFP], herpes simplex virus transactivator protein 16 [VP16], the antibody-specific signal sequences FLAG tag [FLAG], Xpress tag [XPRESS], c-myc tag [MYC], nickel-binding epitope His tag [His], maltose-binding epitopes [MBP], glutathione-binding portions of glutathione S-transferase [GST]. The use of proteins modified in this way is identified through use of the epitope name in the GRIM1 construct employed. The function of the individual regions of the polypeptides of the invention was investigated by these experiments.
Database analyses have shown that hsGRIM1-homologous proteins are to be found in a large number of species. Homologous sequences are to be found in Shizosaccharomyces pombe, Saccharomyces cerevisiae, Rattus spec., Bos bovis, Mus musculus, Drosophila melanogaster, Caenorhabditis elegans, Danio rerio and also in Arabidopsis thaliana. A summarizing overview is obtained from the databases UNIGENE (National Center for Biotechnological Information, NCBI) and of the PredictProtein server of Columbia University and is shown in Table 3 below:
H. sapiens:
M. musculus:
R. norvegicus:
A. thaliana:
C. elegans:
D. melanogaster:
S. cerevisiae:
The Swissprot accession numbers, UNIGENE accession numbers, investigated region and maximum homology are indicated (mouse Q9WV70 is not the full-length clone; compare with the described mouse clone described under Seq. ID No. 4).
The characteristics of the genomic sequences of mmGRIM1 and hsGRIM1 are compared in Table 4 below. The genomic sequence of both species, including the detailed exon-intron boundaries, is indicated in
Many different strategies and expression systems were tested for expression of recombinant hsGRIM1 protein, but only expression of short protein fragments either in the His6x or in the GST context was successful. The following lists briefly summarize the results and the methods used.
Bacterial systems: all the systems used employ Escherichia coli safety strains which were purchased from Stratagene.
BL21 (DE3) lysGOLD His-fused: Expression in the pRSET system or in a modified pRSET vector with unique hexahistidine tag, purification on a nickel affinity matrix (TALON, Clontech). Coexpression with GroESL and hspHJK chaperones.
BL21 (DE3) lysGOLD GST-fused: Expression in the pGEX4 or 6 systems from Pharmacia, purification on GSH affinity columns, expression inducible by IPTG. Coexpression with GroESL and hspHJK chaperones.
BL21 (DE3) lysGOLD NusA-fused: Increase in solubility of the target protein via the NusA content, purification by GPC (Stratagene, Novagen)
BL21 (DE3) lysGOLD MBP-fused: periplasmic expression, and thus reduced cytotoxicity (Stratagene, New England Biolabs pMAL system)
BL21 (DE3) lysRIL: Codon-optimized BL21 strain (Stratagene). Both hexahistidine- and GST-fused protein expression.
ABLE C: Safety strain which downregulates endogenous plasmid copy numbers by a factor of 4× (Stratagene). Both hexahistidine- and GST-fused protein expression.
ABLE K: Safety strain which downregulates endogenous plasmid copy numbers by a factor of 10× (Stratagene). Both hexahistidine- and GST-fused protein expression.
Physcomitrella patens moss: stable integration of a CaMV-driven expression plasmid for GST-hsGRIM1 (pRT 99/35S-GST-hsGRIM1) via non-homologous recombination, triple selection via two different selection antibiotics and multiple seedings (in cooperation with Prof. Ralf Reski, Plant Biotechnology Institute of Freiburg University).
Expression of GST fusion protein in eukaryotic cell culture cells under the control of a constitutively active CMV promoter.
Short hsGRIM1 protein fragments were obtained both as hexahistidine- and as GST-fusion proteins only in BL21 (DE3) lysGOLD without coexpression of chaperones. The following table summarizes the expression results obtained; unlisted fragments (compare table of the hsGRIM1 constructs used) could not be expressed:
To produce hsGRIM1-specific antibodies, two hexahistidine-fused hsGRIM1 protein fragments were expressed in E. coli, purified and used to immunize rabbits. Amino acids 3-147 and 609-749 of hsGRIM1 served as epitopes for the antibody.
The antibodies 2719 and 2720 resulted therefrom.
The antibodies 2910 and 2911 resulted therefrom.
These regions correspond to the putative INHAT domains of hsGRIM1 and are distinguished in particular by acidic (polyE/polyD clusters). The corresponding regions of the hsGRIM1 cDNA sequence were generated by restriction digestion with endogenous EcoRI (aa147) and XmaI (aa609) cleavage sites, and the products were cloned into the vector PRSET-B (Invitrogen) (aa 609-749) or pRSET-N+/H based on the PRSET-B with the difference that further tags have been deleted and the protein is expressed only with hexahistidine tag) (aa 3-147). The hexahistidine-fused hsGRIM1 protein fragments (aa 3-147 “E”, aa 609-749 “I”) were overexpressed in E. coli BL21 (DE3) lysGOLD and, after disruption of the bacteria, purified by TALON affinity chromatography. After isolation and purification from E. coli, both polypeptides had an estimated purity of about 80-90%. Rabbits were immunized with the polypeptides, and the raised antiserum was purified. The purified monospecific antibodies against aa 3-147 (2719, 2720) and against aa 609-749 (2910 and 2911) were functionally characterized.
The antibodies obtained in this way were employed in various immunological test methods that made it possible to detect expression of GRIM1. The two antibodies directed against the N terminus show exclusively nuclear localization of GRIM1 in subnuclear compartments which are generally referred to as “speckles”. The number and size of the nuclear speckles varies from cell line to cell line, only the exclusively nuclear localization of GRIM1 being common to proliferating cells. The two antibodies directed against the C terminus recognize exclusively nuclear GRIM1 in proliferating cells, but not in speckles, rather in a pan-nuclear localization in the entire nucleoplasm.
hsGRIM1 cDNA expression is ubiquitous according to the report of the Unigene database (UniGene Cluster Hs. 134200, Homo sapiens DKFZP564C186/DKFZP564C186 protein with the cDNA library 2334BT0407 used). Expression was detected:
Cervix, Pancreas, amnion_normal, bone, bone marrow, brain breast, cervix, colon, colon_ins, denis_drash, ear, epid_tumor, esophagus, eye, genitourinary tract, germ cell, head_neck, heart, kidney, kidney_tumor, liver, lung, lung_normal, lung_tumor, lymph, muscle, nervous_normal, ovary, pancreas, placenta, pool, pooled, pooled brain, lung, testis, pooled colon, kidney, stomach, pooled lung and spleen, prostate, prostate_tumor, salivary gland, skin, small intestine, stomach, testis, testis-cell line, thymus pooled, tonsil, uterus, uterus_tumor, whole embryo
The expression pattern of hsGRIM1 was visualized by Northern blot analysis (CLONTECH hsMTN I and hsMTN II, probe: both full-length hGRIM1 cDNA and N- and C-terminal fragments). The hsGRIM1 transcript size is about 3.3 kb (consistent with the deposited full-length mRNA under Acc. # AL 050019):
The expression of mmGRIM1 in various tissues of the adult mouse (4-6 weeks) was determined by RT-PCR. The following tissues and organs were classified as mmGRIM1-positive:
Expression of mmGRIM1 mRNA in embryonic development was determined via RT-PCR with complete embryo mRNA. A distinct mmGRIM1 signal is detected at all times, and no large change in mmGRIM1 mRNA expression during development is observable.
Table 8 depicts a UNIGENE database analysis on EST expression selected in homology to mm/hsGRIM1 protein. The accession numbers of the IMAGE clone or of the sequences which are unknown in some cases, and the tissue type from which the cDNAs were isolated, are indicated. GRIM1 appears to be expressed ubiquitously also at the protein level. In total, 308 ESTs were found, and those which show an unambiguous tissue assignment are listed below:
vary
15932
I1476
007523
765941
817292
346323
449674
60700
ourinary
U636573
79097
indicates data missing or illegible when filed
Since various experiments cannot be carried out directly on humans, cell culture and animal experiments which, however, also have validity for the human situation were carried out.
a) The expression of mmGRIM1 protein in primary cells and tissues is indicated below (all the listed cells are classified as GRIM1-positive in a Western blot (86 kD), and for cells the expression was also checked in IIF). Signals are obtained with both antibodies (N- and C-terminal epitope):
Expression of an hsGRIM1-like protein in Drosophila melanogaster was investigated by Western blotting and whole-fly in situ hybridizations.
No specific signal was detected with the anti-hsGRIM1-specific antibodies (resolution from 36 to 210 kD) either in whole-larva lysate or in whole-cell extracts from fly heads. Likewise, no specific dmGRIM1 (dm=Drosophila melanogaster) signal was detectable with the available GRIM1 antibodies in in situ hybridizations with various larval stages.
In the protein-protein interaction studies carried out to date, the following potential mm/hsGRIM1 interactants were investigated for direct or indirect interaction with the following methods:
The underlined factors showed an interaction with hs/mm/maGRIM1 and are described separately at the end of the list (ma=mesocricetus aureus, BHK cells).
hsAR(PD, utfIP, tfIP, yTHS, tT), hsERα (utfIP, tfIP), hsERβ, rnGR (tT, tfIP), hsMR (tT), hsPR (tT), PPARγ (tT), hsRAR α(tfIP), rnTRα (tT), hsTRβ-2 (tT), mmERR1 (tfIP, tT), mmGCNF (PD, ivIP, tfIP, tT), hsRevErbA (tT), mmRVR (tT, tfIP), mmSF1 (tT).
Myogenin (ivIP, tT), MEF2A (ivIP), MEF2B (ivIP), MEF2C (ivIP, tT), eHand (ivIP), dHand (ivIP), E47 (ivIP), E12 (ivIP), GATA4 (ivIP), MyoD (ivIP, tT), SEF (ivIP, tT), MRF4 (ivIP, tT), Myf5 (ivIP), Sp1 (tT), CBF (tT).
N—CoR (ivIP, tfIP, utfIP, PD), Sin3A (ivIP, tfIP, utfIP), Sin3B (tfIP, utfIP), SuN—CoR (ivIP, tT), SMRT (ivIP, tfIP, utfIP), HDAC1 (tfIP, utfIP), HDAC2 (tfIP, utfIP), RIP140 (ivIP).
P300 (tT), FHL2 (ivIP, PD), FHL3 (PD), TIF2 (ivIP), RIP140 (ivIP).
Further factors: Histones H3, H4, H2A, H2B (PD)
It has been possible to date to show a strong direct interaction between hsGRIM1/mmGRIM1 only with the androgen receptor (hsAR) (detected by tfIP in both directions, utfIP in both directions, yTHS, PD and tT). The interaction with the androgen receptor is ligand-independent and cannot be influenced by the presence of antagonists (Casodex, hydroxyflutamide, cyproterone acetate). GRIM1 interacts both with the “normal” AR having 17 glutamines and with AR isoforms which are associated with Kennedy's disease and comprise polyglutamine-rich insertions (17, 44 and 77 Q), likewise ligand-independently. The PD analyses were carried out with the GRIM1 fragments GST-I/E/H/M, and all the tested fragments apart from GRIM1 “E” (aa 3-147) interacted with the AR, in the sequence M≧H>>I. The physiological relevance of the AR-GRIM1 interaction is described in detail in a separate section.
Weak interactions were detected between hsGRIM1 and the nuclear orphan receptors Germ Cell Nuclear Factor (mmGCNF) and Estrogen receptor Related Receptor 1 (mmERR1) (only ivIP, tfIP and PD), and a weak interaction with the estrogen receptor α (hsERα; interaction was detected only in tfIP and proved to be ligand-independent both in the presence of agonists (estradiol), of antagonists (raloxifene, tamoxifen, ICI) and in the absence of a ligand). The physiological relevance, in vivo colocalization or in vivo interaction is very substantially unclear. It was not possible to show any direct association between GRIM1 and mmGCNF, mmERR1 or hsERα in transient transfections.
GRIM1 shows a strong interaction with the core histones H4, H2B, H2A and H3 in GST pulldown analyses. Recombinantly expressed GST-GRIM1 fusion proteins (GST-aa 3-147 and GST-aa 609-749) show a marked interaction with all four core histones, providing further support for the observed INHAT activity of the N- and C-terminal GRIM1 domains.
Potential mm/hsGRIM1 interactants can be identified by conventional methods. These methods include yeast two-hybrid screen with various cDNA libraries and direct protein biochemical methods such as Co-IP of GRIM1 and associated proteins, fractionation by 2-dimensional gel electrophoresis, mass spectrometric analysis and microspray sequencing of the potentially interacting proteins from various cell culture cell lines (C2C12 differentiated, C2C12 undifferentiated, 3T3L1 differentiated, 3T3L1 undifferentiated, BHK and many more).
All the described data were obtained with the hsGRIM1 clone. GRIM1 is a potent transcriptional repressor when used as Gal4-DBD fusion protein in transient transfections. Gal-GRIM1 represses expression of reporter genes both of Gal4-RE-controlled complex and synthetic minimal reporter constructs. The repression potential of GRIM1 is distinctly greater for example compared with the corepressor Gal-SuN—CoR. In these experiments, the appropriate corepressors are transiently expressed as heterologous Gal4-DBD fusion proteins in cells. The repression capacity on various synthetic promoter-reporter constructs compared with equal quantities of transfected wild-type Gal4-DBD is determined by luciferase reporter gene expression.
293 cells were cotransfected with G5E1bTATA-LUC reporter and the stated amounts of corepressors. The same effect can be observed on testing complex promoters (Gal43x-tk-LUC).
Gal-GRIM1 is equally strong as a repressor as a large number of described corepressors such as, for example, Gal-HDAC1-10, Gal-CBF1, Gal-Alien and represses only insignificantly weaker than Gal-N—CoR-RD or Gal-SMRT-RD. The repression potential of Gal-GRIM1 was investigated in conjunction with the complex Gal43x-tk-LUC and Gal-MMTV-LUC promoter/reporter constructs and in conjunction with the minimal promoter/reporter constructs Gal5x-E1b-TATA-box and Gal3x-βglobin-TATA-LUC. In addition, the repression capacity of Gal-GRIM1 was investigated on an autonomous VP16 transactivation domain bound to DNA by means of LexA-DBD, via a LexA8x-Gal5x-LUC reporter (provided by Dr S. Khochbhin). The data obtained on Gal-GRIM1 has been tested experimentally in the following cell lines: A7r5, C2C12, CV1, BHK, 293, N2a, COS-7. The following figure shows representative results for BHK and C2C12 cells.
The Gal-GRIM1 repression is not impaired by addition of the specific histone deacetylase (HDAC) inhibitors trichostatin A (TSA, 300 nM) and of sodium butyrate (Na-But, 5 mM), suggesting that the mechanism of repression is fundamentally different from that described for other corepressors. Comparable cases have been described for other repressors, although no explanation of the method is described. In addition, the described factors show no involvement in skeletal muscle differentiation or adipogenesis. Data on the independence of HDACs in transient transfections are depicted in the following figure for the example of BHK cells. The transcriptional activity of Gal-GRIM1 is uninfluenced by TSA in any cell line already tested.
Transient transfections were carried out in BHK cells with a Gal4-tk-LUC reporter plasmid.
To determine the GRIM1 domains involved in repression, a total of 11 deletion mutants of GRIM1 were constructed and all were tested in the Gal4-DBD context for their repression potential in all the reporter systems already described. All the GRIM1 domains used show transcriptional repression in the range from 2× (Gal-E and Gal-I) to 15-20× (full-length Gal-GRIM1). All the tested deletion mutants proved to be insensitive to HDAC inhibitors.
In summary, it was possible to show that GRIM1 has at least 5 independent repression domains, with the 3 main repression domains being located in the hydrophobic core region of the protein (mutQ, mutR, mutP). However, every analyzed deletion mutant shows a repression capacity which is significantly above the background in the Gal4-DBD context.
The GRIM1 subdomains “E” and “I” described in the section “INHAT activity” show the smallest repressive effect as Gal4-DBD fusions in the transcriptional assays, but are involved in preventing histone acetylation by p300. No further information on the Q, P and R domains is available, and it was not possible to prove a direct involvement in the recruitment of additional corepressors by the interaction studies cited above. At the moment there is only speculation about the mechanism of repression used by GRIM1, although involvement of HDACs can be ruled out, and it appears that at least two different mechanisms are used (E and I in contrast to Q, R and P).
Gal-GRIM1 deletion mutants were investigated for their repression potential in transient transfections. The central GRIM1 repressor domain is located in the region between aa 246-609 (depicted in white). The central RD can be divided further by further deletion studies into a total of three independent domains (depicted with dark shading).
Results with the LexA8xGal5x system confirm these results but show a different relative repression of the individual fragments investigated (data not shown).
Full-length GRIM1 is able to repress both complex naturally occurring and synthetic minimal promoters. The following systems have been tested in transient transfections: GRIM1 and complex promoters (p21 (−4542)LUC, SKA-LUC, calponin-LUC, myogenin-LUC, myogenin proximal promoter-LUC, thymidine kinase-LUC, probasin-LUC, MMTV-LUC and many more).
GRIM1 dose-dependently represses the ApoA1 promoter, the calponin (−3000) promoter, the SKA promoter, and the myogenin promoter and its proximal element (myogenin 133-LUC).
GRIM1 dose-dependently represses the calponin (−3000) promoter, the SKA promoter, the p21 (−4542) and the myogenin promoter.
Cotransfections of GRIM1 and E1b-TATA-box-containing minimal promoters with synthetic TF binding sites (AR, Gal4 and SEF binding sites) or with the complex MMTV promoter leads to repression of basal transcriptional activity (all transfections shown were carried out in BHK cells).
500 ng of reporter construct were used in transient transfections of BHK cells. G5E1bTATA-LUC is an adenovirus E1b TATA box-containing minimal construct with 5 Gal4 binding sites, ARE2× (tk)TATA-LUC is a minimal construct with 2 copies of an AR binding site in front of a thymidine kinase TATA box, E-Box(tk)TATA-LUC includes a binding site for the bHLH transcription factor SEF in the same context and MMTV describes the complex Moloney mouse tumor virus promoter.
Repression of GRIM1 with minimal promoters without synthetic TF binding sites cannot be demonstrated more accurately owing to the low basal value of the constructs used.
The repression potential of GRIM1 on promoter-luciferase-reporter constructs which have been stably integrated into the genome of the particular cell line (HRL+N with RARE3x-LUC, NIH3T3 with cycA-LUC, 293 with NFkB-RE4x-LUC) is summarized in the next figure. Transfection of the cells with GRIM1 shows no effect on reporter gene transcription in the induced and in the basal state.
A HRL+N cells with stably integrated retinoic acid-inducible promoter. B NIH3T3 cells with stably integrated CycA promoter-LUC. C 293 cells with TPA-inducible NFkB promoter.
In addition, direct effects of GRIM1 on various transcription factors were investigated in transient transfections (compare also Example 11). Besides the AR, in particular mmGCNF, mmERR1 and mmRVR were analyzed experimentally as full-length proteins with the corresponding reporter constructs for a functional interaction with GRIM1. No significant effect was observable in any of the transfections.
Interactions of GRIM1 with further transcription factors or cofactors was investigated in the Gal4-DBD-based context, with the respective Gal4-DBD-fused factor being cotransfected in constant amount with an increasing amount of wild-type GRIM1. Factors investigated in this system are: Gal4-p300, Gal4-DBD, Gal-Sp1-4, Gal4-CBF). Once again, no significant effect was observed in any of the transfections.
a) HDAC activity
The mechanism of repression by recruitment of histone deacetylases (HDACs), which is the most widespread to date, appears not to be involved with GRIM1. The repression potential of Gal4-DBD-GRIM1 in transient transfections is not impaired by application of specific HDAC inhibitors in a concentration of up to 5-fold above the KD (final concentration of trichostatin A 300 nM, of sodium butyrate 5 mM). Immunoaffinity-purified and high-stringency-washed GRIM1 shows no endogenous HDAC activity. Nor do associated complexes which have been coimmunoprecipitated with GRIM1 by mild washing conditions show any HDAC activity. In addition, direct involvement in HDAC/N—CoR/SMRT complexes have been ruled out by Co-IP studies.
Released radioactivity above the background of 850 cpm is assessed as HDAC activity. Samples were washed either 3 times physiologically (SC) or with increasing amounts of NaCl (300 and 400 mM). At 400 mM salt, it is assumed that no associated proteins were coprecipitated (confirmed by Western blotting and Ponceau stains). The figure summarizes 5 HDAC assays which were carried out independently. The name Arco which is used refers to the GRIM1 protein.
b) INHAT activity
A further possibility for transcriptional repression consists of direct masking of the histone N termini to prevent acetylation and thus loss of cohesion of the chromatin in the affected region. It was found by database analysis that the N terminus (aa 25-135) and the absolute C terminus (aa 638-749) of hsGRIM1 has acidic domains which display homology with a described INHAT domain of the Set protein. INHAT is an acronym for protein domains/proteins able to bind lysines and thus prevent histone acetylation.
It was possible to show in GST pulldown analyses that the N- and C-terminal fragments of hsGRIM1 can associate directly with histones in vitro. In addition, it can be shown in in vitro acetylation assays that recombinantly expressed GRIM1 aa 3-147 and aa 609-749 efficiently blocks acetylation of all four core histones (H3, H2A, H2B and H4) by p300. It was possible thereby to show that a possible mechanism of the GRIM1-mediated repression might be ensured by histone masking.
There are differences in the efficiency of inhibition of acetylation of individual histones by GRIM1. As yet, only blockage of acetylation of H2A and H2B have been unambiguously demonstrable.
c) Further mechanisms of repression
Lysine Methylation of Histones
Methylation of specific lysines in the free histone N termini (H3K9, H3K4) ensure compaction of the chromatin structure by preventing acetylation of the lysines which are important for regulation and thus preventing a loss of cohesion of the chromatin. In addition, methylated lysines represent a high-affinity binding site for the HP1 protein (heterochromatic protein 1). HP1 may subsequently recruit DNA methylases, MeCPs and also HDACs and lead to a targeted formation of transient or constitutive heterochromatin and thus repress transcriptional processes. Examples of HMTases described to date are Suv39h1, Suv39h2 and PRMT1.
Homology analyses show that no conserved SET domain responsible for histone methyltransferase (HMTase) activity is present in the GRIM1 sequence. It is therefore assumed that GRIM1 has no endogenous HMTase activity.
Recruitment of chromatin-Reorganizing Complexes
A further possibility for specific repression would be targeted recruitment of chromatin-reorganizing complexes (e.g. BRG, SWI/SNF, CHRAC, ARC, RSC etc.). Promoter regions of the genes to be repressed can be wrapped up in compact chromatin by active chromatin reorganization in such a way that transcriptional initiation is distinctly impeded.
It is not possible to rule out a functional interaction of GRIM1 with one of the components of one of the remodeling complexes.
Blocking PIC Formation/Impeding the Initiation Reaction/Impeding Elongation
A further form of repression is impeding or preventing the formation of the preinitiation complex (PIC) at the start of transcription. Successful initiation is possible only if the TATA box-binding portion of TFIID can recruit the polymerase holocomplex at the starting point of transcription and if all the basal transcription factors involved are available in sufficient quantity. A further prerequisite for successful transcription is the need for the C terminus of RNAPII to be phosphorylated so that leaving of the initiation site is possible.
It is possible in an experiment familiar to the skilled worker to find whether GRIM1 directly interacts with one of the basal transcription factors or elongation factors and thus blocks efficient transcription (yeast two-hybrid system, pulldown analyses). No phosphatase domain can be found within the GRIM1 structure by homology analyses, and it is yet to be proven experimentally that GRIM1 is able to dephosphorylate RNAPII.
Titration Out of coactivators by Binding to GRIM1
The potential of GRIM1 to inactivate basal or general coactivators by binding and thus to repress the initiation or elongation process is currently under investigation by searching for potential hsGRIM1 interactants.
A total of at least 5 independent repressor domains (RDs) have been found within the GRIM1 protein in transient transfections. The fragments E and I of GRIM1 showed the smallest repression potential in transient transfections, but their function in repression by GRIM1 was demonstrated in the INHAT analyses.
However, the deletion mutants P, Q, R and also S show a far stronger repression function in Gal-based transfections, suggesting further mechanisms beyond the known facts. It is possible to find the contribution to repression made by the other regions of GRIM1 by conventional experiments. These were intended to identify interacting GRIM1 partner proteins in order to find whether there is an association with non-HDAC components of transcriptional corepressor complexes.
Functional motifs within the hsGRIM1 structure were found by database homology analyses. A summary of the motifs found, their localization within the complete protein and the experimental strategy used is shown in the following table.
I RD LI
L QN LL
L AF LV
L TE LL
II GC I
L PF IL
IL RP L
L RV L AF L V L
KRRK NAD RK DED RK QF K
Detection of the NES within hsGRIM1
For direct analysis of the motif, all four relevant lysines were inactivated by directed point mutagenesis so that the α-helical structure of the NES was destroyed. The inactivating mutations are depicted in the following figure. A directed inactivation of the two 3′-located lysines on its own proved to be insufficient to destroy the functionality of the NES.
To demonstrate the functionality of the export signal, hsGRIM1 protein fragments were fused to GFP (green fluorescent protein) (depicted in the following figure) and expressed in cells by transient transfection. The localization of the respective fragments was detected by fluorescence microscopy.
ΔNES characterizes the point mutations described above.
Only GFP showed a pan-cellular localization. Full-length GFP-hsGRIM1 was located exclusively in the nucleus through the dominance of the NLS, as were the NES deletion mutant and the fragment from which a region flanking the NES had been deleted (hsGRIM1 Δaa246-334). On fusion of the hsGRIM1 fragment which contains the NES to GFP (hsGRIM1 aa246-334), the resulting protein was located exclusively in the cytoplasm. On coupling of the point-mutated fragment of aa 246-334 to GFP, the pan-cellular localization was restored so that it was possible to demonstrate the functionality of the motif.
Detection of the NLS within hsGRIM1
To analyze the motif, besides the deletion mutants detailed hereinafter, all seven relevant basic residues were inactivated by directed point mutagenesis. Inactivation of the 3′-located 3 basic residues alone proved to be insufficient to inactivate the NLS.
To demonstrate the functionality of the import signal, hsGRIM1 protein fragments were fused to GFP (depicted in the following figure) and expressed in cells by transient transfection. The localization of the respective fragments was detected by fluorescence microscopy.
Only GFP showed a pan-cellular localization. Full-length GFP-hsGRIM1 was located exclusively in the nucleus. On deletion of the last 140 amino acids (hsGRIM1 aa1-609), the GFP fusion protein showed both a cytoplasmic and a weak nuclear localization, while the corresponding counterpart fused to GFP (GFP-hsGRIM1 aa609-749) shows an exclusively nuclear localization. The full-length hsGRIM1 with the seven-fold point mutation within the NLS showed the same distribution pattern as the hsGRIM1 aa1-609 deletion mutant and confirmed the functionality of the NLS. On coupling of the point-mutated fragment of aa609-749 to GFP, the fusion protein was likewise located in the cytoplasm.
Detection of the CoRNR Box within hsGRIM1
To analyze the CoRNR box, all three highly conserved and necessary isoleucines in the hydrophobic core of the helix were replaced by alanines or by a serine: I436A, I437S, I444A. Further functional characterization of the point mutations made have revealed to date only that the repression potential of hsGRIM1 is not impaired, and that the interaction with the AR is not impaired by destroying the CoRNR box.
Intracellular Relocalization of mmGRIM1 in Cell Culture-Based Differentiation Models
GRIM1 shows a change in its subcellular localization during specific differentiation processes. In cell culture systems, GRIM1 is transported from the nucleus into the surrounding cytoplasm during skeletal muscle differentiation of mouse C2C12 cells. The timeframe for the observed translocation of GRIM1 is in the range for the expression of the myogenic bHLH transcription factor myogenin (My-14) and of the structural protein skeletal muscle myosin heavy chain (MHC), i.e. at a more advanced time during muscle differentiation than previously described factors such as HDAC4, HDAC5 and HDAC7. C2C12 cells differentiate under low-serum conditions within 6-8 days from proliferating myoblasts to fused, multinuclear myotubes. During this, the cells undergo a G0 arrest, the CDK inhibitor p21 is upregulated, and an ordered cascade of myogenic transcription factors such as Myf5, MyoD and myogenin provide the transcriptional precondition for myogenesis. In differentiating C2C12 skeletal muscle cells, GRIM1 leaves the nuclei of fused myotubes after about 6 to 8 days, whereas GRIM1 is always located in the nucleus in mononuclear myoblasts. Expression of the structural protein of muscle MHC takes place at the time when the first GRIM1-free nuclei are to be observed, and was used in IIFs as temporal marker for GRIM relocalization.
The data determined for C2C12 also apply to other cellular skeletal muscle differentiation systems. The same effect was also detectable in rat L6 skeletal muscle cells. The cells have different differentiation kinetics and do not fuse until day 10-12 in differentiation medium (DM). GRIM1 also relocalizes in L6 cells, and GRIM1-free nuclei are detectable only from day 16 onwards in DM. In a further differentiation system, mouse 10T½ fibroblasts were stably transfected with an estrogen-inducible MyoD expression cassette (cells kindly made available by Dr. D. Bergstrom). Addition of 1 μM estradiol enables skeletal muscle differentiation via MyoD expression. After only about 3-4 days in DM, many cells can be detected as multinuclear myotubes.
During skeletal muscle differentiation, the GRIM1 protein is not degraded and/or broken down but is present to the full extent also in the cytoplasm. GRIM1 immunoreactivity can be detected unchanged in Western blot analyses with whole cell extracts of C2C12 cell extracts which have been differentiated to myotubes in DM for 8 days (on day 8 after addition of DM, ≧60% of all the cells present were myotubes). The results can also be further confirmed by cell fractionations. In addition, RT-PCR analyses have confirmed that the amount of mmGRIM1-mRNA is not changed during C2C12 differentiation.
As a further check, it was investigated whether GRIM1 also experiences subcellular relocalization during smooth muscle differentiation. For this purpose, mouse smooth muscle cells were obtained ex vivo and dedifferentiated by growth factor addition. Cells were additionally differentiated again further to smooth muscle cells via differentiation medium in vitro. The results show that GRIM1 does not leave the nucleus during smooth muscle differentiation. In addition, a cell culture smooth muscle differentiation system (Monc-1 cells) was used in order to obtain in vitro data about possible GRIM1 relocalization. As already shown for the ex vivo cells, no GRIM1 relocalization in smooth muscle cells was observable during the investigated differentiation times.
A further in vitro differentiation model used for GRIM1 relocalization analyses is differentiation of 3T3-L1 preadipocytes to mature fat cells. In the total of three differentiation protocols carried out, GRIM1 twice showed a detectable relocalization from the core into the cytoplasm, while in one case no change in the localization of the GRIM1 immunoreactivity was detectable. Western blot analyses have shown that no decrease in GRIM1 immunoreactivity in whole-cell extracts was detectable in any of the three protocols.
Complete experimental results are not yet available on the regulation of GRIM1 function and on signal pathways which might be involved in the subcellular relocalization of GRIM1. No signal pathway modulating the transcriptional activity of GRIM1 has yet been unambiguously identifiable. Nor is there any experimental evidence of involvement of a signal cascade intervening in the GRIM1 relocalization process. The data available on this topic can be briefly summarized below.
It was possible to demonstrate in a yeast two-hybrid interaction assay that GRIM1 interacts with the nuclear hormone receptor androgen receptor (AR). The GRIM1 interaction domain in the yeast-based assay is located in the C terminus between aa 606 to 749. Further functional characterization of the interaction domains within GRIM1 was carried out by pulldown analyses with the expressible fragments of GRIM1. The interaction of GRIM1 with the AR is independent of the type of ligand used (agonists, antagonists or no ligands). The available data are summarized in the following picture:
Regions of the protein depicted in white have not yet been expressed recombinantly.
The C terminus of GRIM1 shows, in contrast to the data obtained in yeast, no in vitro interaction with the AR, but the fragment from aa 460 to aa 609 shows a strong interaction independent of the ligand used.
The interaction between the AR and GRIM1 was likewise detectable in in vivo interaction assays. In co-immunoprecipitations, the interaction between GRIM1 and the AR was likewise ligand-independent and could not be significantly blocked even in washing buffer with 400 mM NaCl. It was possible to show the interaction with IP studies using antibodies both against the AR and against GRIM1.
The AR is a member of the superfamily of ligand-activatable nuclear hormone receptors. After binding of the natural AR ligand 5α-dihydrotestosterone, the AR is transported from the cytoplasm into the nucleus and brings about transcriptional activation of its target genes. Besides the activating ligands, each receptor has specific antagonists which likewise bring about translocation into the nucleus, but bring about a repressive conformation of the receptor ligand binding domain. Antagonist-bound receptor is able to recruit corepressor complexes and actively repress transcription. However, in some cell lines or in certain pathological situations, antagonists may become partial agonists and unnaturally bring about a transactivation by the receptor. It is known in the case of the AR that substances employed therapeutically for prostate carcinomas, such as cyproterone acetate, Casodex or hydroxyflutamide, may have partial agonistic effects in hormone-refractory tumors.
It was possible to show in the cell culture system that GRIM1 is able specifically to repress an AR transactivation brought about by CPA, Casodex or OH-flutamide. Non-liganded or DHT-loaded AR is not transcriptionally affected by GRIM1. The observed effect varies in the range between 2-3-fold repression of the CPA-induced AR transactivation.
Transient transfections were carried out in BHK cells in which the antagonists used act as partial agonists. A minimal promoter with synthetic AR binding sites was used as reporter (ARE2x-(tk)TATA-LUC). The same effect was obtained on use of 1 μM Casodex or OH-flutamide instead of 1 μM CPA.
A further indication of the functionality of the effect is that the superactivation of the CPA-loaded AR by the coactivator TIF2 in BHK cells can be repressed or at least markedly restricted by cotransfection of GRIM1.
Transient transfections were carried out in BHK cells in which CPA (1 μM) acts as partial agonist. ARE2x-(tk)TATA-LUC was used as reporter.
| Number | Date | Country | Kind |
|---|---|---|---|
| 102 12 397.7 | Mar 2002 | DE | national |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 10507780 | Nov 2004 | US |
| Child | 12341655 | US |
