FR2797055A1 - METHOD FOR SCREENING MOLECULES FOR THE TREATMENT OF HUNTINGTON'S DISEASE - Google Patents

Abstract

The invention concerns a method for screening molecules for the treatment of Huntington disease and a method for identifying therapeutic targets for said disease. The methods are based on the indentification of molecules which are capable of modifying the interaction normal htt/ CA 150.

Description

<Desc / Clms Page number 1>

 The subject of the present invention is a method for screening molecules intended for the treatment of Huntington's disease as well as the use of molecules thus demonstrated for the preparation of a medicament intended for the curative or preventive treatment of this disease. It also relates to a method of identifying therapeutic targets for Huntington's disease and the use of nucleotide sequences corresponding to genes involved in the development of this disease in order to treat it.

 Huntington's Disease (Huntington Disease or HD) is an autosomal dominant neurodegenerative polyglutamine expanding disease (polyQ) whose gene codes for huntingtin (htt), an unknown function protein expressed in the brain and several other tissues. The expansion of polyQ (following a dynamic mutation-or expansion-CAG in the htt gene) is located in the N-terminal part of the htt, a region that is conserved and functionally important for the appearance of HD. The normal htt contains a normal number of glutamine repeats (Glns) (typically: 15 to 20 Gins, mainly: less than 39 Glns). The mutated htt contains an abnormally high number of glutamine repeats (typically the pathological threshold is around 39 Glns, the expansion of polyglns can for example reach up to 180 Gins in certain cases) which is associated with the appearance of disease (Koshy and Zoghbi, 1997). Htt is an important protein for the genesis and survival of the neuron (Zeitlin et al., 1995; White et al., 1997). In the normal state, this protein is mainly located in the cytoplasm of neurons where it is for example associated with microtubules (DiFiglia et al., 1997). Normal htt has also been detected in the nucleus, in mouse neuroblastoma cells (de Rooij, 1996), and in the postmortem human brain (Hoogeven et al., 1993).

 On the tissue level, HD is characterized in the first place by a progressive and selective apoptosis of the striatal neurons; more advanced stages, HD affects other regions of the brain like the cortex (Koshy

<Desc / Clms Page number 2>

 & Zoghbi, 1997). The molecular bases of this selectivity are not known.

 As normal and mutated forms of htt are expressed at similar levels, the cause of HD is to be sought in biochemical disturbances in the signaling pathways involved in the genesis, functioning, and survival of the neurons affected in HD.

 Furthermore, there is an inverse correlation between the age of onset and the severity of HD and / or the number of glutamine repeats in the htt (Koshy and Zoghbi, 1997), which indicates that the severity of the disturbance of these biochemical functions is modulated by the length of the polygln domain.

 There is no treatment for this seriously debilitating disease (6000 people in France, apparent prevalence close to 1/10. 000). The identification of the biochemical factors involved in HD is necessary in order to identify therapeutic targets for this disease, and to proceed to the search for therapeutic methods and active compounds making it possible to restore normal functioning of the intracellular signaling pathways important for survival. and the proper functioning of the affected neurons in HD.

 In patients, nuclear aggregates (sometimes perinuclear) composed of N-terminal fragments of htt and ubiquitin have been demonstrated (DiFiglia et al., 1997). The fact that aggregates were detected before the appearance of apoptosis in in vitro cell models and in transgenic mouse models initially led to the hypothesis that nuclear aggregates could be at the origin of the disease (Mangiarini et al., 1996; Davies et al., 1997; Ross, 1997). However, the most recent results strongly support the hypothesis of an effect of the free forms of mutated htt as a fundamental pathophysiological mechanism of HD (Sisodia, 1998). Several studies (on postmortem brains, primary cultures of rat striatal neurons, other in vitro cell models, and mouse models) indicate that there is not a good correlation between the formation of aggregates and apoptosis of target neurons (Bates et al., 1998; and Mandel, 1998; Saudou et al., 1998; 1998; et al., 1999; Schilling et al.,

<Desc / Clms Page number 3>

 1999). These aggregates are moreover likely to fortuitously protect the neuron against an excess of apoptosis caused by the non-aggregated mutated htt (Saudou et al., 1998). On the other hand, the nuclear localization of free Nterminal fragments of the mutated htt is necessary for the induction of neuronal apoptosis (Saudou et al., 1998).

 The nucleus could therefore be the privileged place of the pathophysiological mechanisms of HD, in particular in terms of dysregulation of transcription and inadequate expression of certain genes and proteins.

 Another important aspect of the pathophysiology of HD lies in the fact that neuronal dysfunction before apoptosis is highly likely to be at the origin, in whole or in part, of this disease. In transgenic models of HD in mice, it has indeed been observed that behavioral symptoms in transgenic mice (expression of an N-terminal fragment of the mutated htt corresponding to exon 1 of the gene for this protein) appear before detecting apoptosis of neurons (Bates et al., 1998).

 Furthermore, the expansion of polyglns probably induces a change in conformation of the N-terminal fragments of htt, the pathological consequences of which would go hand in hand with the progressive proteolytic degradation of htt (initially by caspases like caspase-3, then by other unidentified proteases) and the genesis of shorter and shorter N-terminal fragments comprising for example amino acids 1-500,1-300, then 1-150, and 1-100 (Goldberg et al., 1996; Lunkes and Mandel, 1998; Martindale et al., 1998; et al., 1998; Schilling et al., 1999), and their passive diffusion into the nucleus through nuclear pores (Hackam et al., 1999). Indeed, the first 50 to 200 amino acids are sufficient to induce cellular or even behavioral phenotypes similar to those observed in HD in transgenic models of mice (Mangianiri et al., 1996; Bates et al., 1998), of Drosophila (Jackson and al., 1998), and nematode C. elegans (Faber et al., 1999; Néri et al., 1998).

 The expansion of polyQ (through the change in conformation mentioned above) would also modify the accessibility of huntingtin partner proteins (PPH) to their binding sites in the htt (apart from

<Desc / Clms Page number 4>

 polyQ domain), resulting in a decrease or an increase in the binding of mutated htt with PPH belonging to intracellular signaling pathways essential for the functioning and survival of the neuron. The potential role of abnormal protein-protein interactions in HD has been illustrated by the identification in double-hybrid screening (Fields and Song, 1989) of several PPHs which bind to the N-terminal region of htt (amino acids 1 to 500). The majority of these PPHs seem to bind to htt outside of the polygln domain. In the case where these PPHs have been identified by screening of non-human cDNA libraries (for example, rat or mouse cDNAs), it has been shown that the human counterparts of these PPHs do indeed bind to htt in vitro, and that they co-localize with htt in extracts from the human brain. In general, there is no systematic correlation between the expression profile of PPH identified to date and the sites of neuronal loss observed in HD. These are essentially the following PPHs: - GAPDH (Burke et al., 1996; increased invariant bond with mutated htt): enzyme involved in glycolysis and energy metabolism.

- HIP2 (Kalchman et al., 1996; invariant link with the mutated htt):
Ubiquitin conjugation enzyme.

- Cystathionine ss-synthase or CBS (Boutell et al., 1998; invariant link with mutated htt): synthesis of cysteine from methionine. The absence of CBS is associated with the accumulation of the CBS substrate which is excitotoxic (homocysteine), and with a rare human metabolic disease (homocystinuria). The role of a similar mechanism in HD has not been established.

- HAP1 (Li et al., 1995 and WO 97/17443; increased binding with mutated htt) and HIP1 (Kalchman et al., 1997; et al. 1997; decreased binding with mutated htt): of unknown function associated to the cytoskeleton whose binding sites seem to lie between amino acids 171 and 500. HAP1 interacts with several proteins, in particular the p150Glued subunit of dynactin (Engelender et al., 1997), which confirms the role of HAP1 in formation and activity of the cytoskeleton and trafficking in intracellular vesicles.

<Desc / Clms Page number 5>

- HYPA and HYPB (Faber et al., 1998; increased binding with mutated htt): proteins with WW domain of unknown function, binding at the level of the two polyprolines immediately at the C-terminal of the polyQ domain.

- SH3GL3 (Sittler et al., 1998: increased binding with mutated htt): of unknown function with SH3 domain belonging to the Grb2 family, the members of this family being associated with synaptosomes, binding at the level of the two immediately C-terminal polyprolines at the polyglns.

 Promoter role of SH3GL3 in the formation of aggregates.

It is important to note three points:
1) Several PPHs which show an increase in binding towards the mutated htt (HAP1, SH3GL3, HYP series) are associated with the formation of nuclear aggregates such as SH3GL3 which has a promoter role on their formation (Sittler et al., 1998), or HIP2 which is directly associated with the kinetics of degradation or accumulation of intracellular proteins. As aggregates are not a sine qua non cause of apoptosis (Saudou et al., 1998; Sisodia, 1998), the search for a blockage of an abnormally increased and neurotoxic link between mutated htt and PPH by an active compound (as described by CA Ross et al., 'Huntingtin-associated protein', WO97 / 17443) is in fact likely not to constitute a good therapeutic option. In addition, these PPHs appear to have moderate variations in interaction (generally by a factor of 2 to 5) for mutated htt (40-60 Gins) compared to normal htt (20 Gins). By virtue of the state of the art in the field of small molecules, it is difficult to envisage the specific beneficial blocking of a moderately increased bond between the mutated htt and a PPH, in the absence of an inappropriate side effect of blocking of the link between normal htt and this same PPH.

 2) HIP1 is the only PPH which has so far shown a loss of connection with the mutated htt. HIP1 interacts approximately 10 times less well with mutated htt comprising 125 Gins than with normal htt comprising 25 Gins (Kalchman et al., 1997 and WO 97/18825). This protein is specifically expressed in the brain, more specifically in the cortex. Its function is unknown: the homology of 20-40% with Sla2p

<Desc / Clms Page number 6>

suggests a role in cytoskeleton formation, and the presence of a Leucine zipper-like motif suggests involvement in several protein-protein interactions. The role of HIP1 in the formation of the cytoskeleton is supported by the fact that this protein is associated with cell membranes (cell debris, nucleus, synaptosomes, mitochondria, microsomes and plasma membranes), a common characteristic of all cytoskeleton proteins such as actin. Since HIP1 is not linked to the formation of aggregates, it is an interesting therapeutic target. However, the fact that it is a cytoskeleton protein does not fall within the framework of the latest data in the scientific literature which indicates that the nuclear localization of the mutated htt is necessary for the establishment of the pathophysiological mechanisms of HD (Saudou et al. 1998;
1998).

3) The influence of the length of the N-terminal fragment (in particular those likely to pass into the nucleus) on the binding of htt to PPH has not been described previously.

 In this context, the inventors provide the following new elements: identification of a new WW domain PPH whose human counterpart is a nuclear transcription factor containing a polymorphic domain (repetition of QA or Glu-Ala dipeptides), which does not is therefore not cytosolic and does not concern the formation of aggregates, and identification of biochemical conditions (size of the fragments of the htt) which go hand in hand with notable differences in the binding of this PPH with the mutated htt and the normal htt in the core. The therapeutic targets and the strong differences identified by the inventors are compatible with the concept of therapeutic interventions which aim to act on the pathological effects of certain forms of the mutated htt, without side effects on the normal physiological functions of certain forms of the htt normal.

 Double-hybrid screening for htt partners is a particularly interesting entry point into the biology of affected neurons in HD. The ultimate proof that PPH is actually involved in neuronal impairments comes from additional experiences in

<Desc / Clms Page number 7>

 in vitro and in vivo. For example, it is possible to prove in vivo in an entire organism 1) that the candidate target is actually involved in the disease phenotype (loss of function or apoptosis) reproduced in the model organism used, 2) that it is possible to modulate its biological activity, and 3) that this modulation can make it possible to restore normal functioning of the neurons rendered deficient by the mutated htt.

This so-called cellular validation (establishment of a direct link between a biochemical phenotype and a cellular phenotype) can be carried out in an entire organism and in a relatively short time thanks to the use of the nematode Caenorhabditis elegans. Thus, it is possible to identify the genetic pathways involved in a neuronal deficiency phenotype through the search for suppressor mutants. C. elegans also makes it possible to study PPH or the physical pathways (protein-protein interactions) of interest and of unknown function which were initially identified by double-hybrid screening in yeast, by making it possible to study the profiles of expression and mutant phenotypes for the homologs of these PPHs in C. elegans.

 In terms of prospective therapeutics in neurobiology, C. elegans has the following advantages: well-characterized nervous system, good functional homology with humans in terms of neuronal physiology, many normal neurological phenotypes and mutants available for analysis. Compared to other simple model organisms such as Drosophila (Jackson et al., 1998), C. elegans also has the following advantages: completely sequenced genome, faster mutagenesis, transparent organism allowing visualization of all neurons and a analysis of protein expression in living animals, short reproduction cycle (a few days), knowledge of complete cell phylogeny, ease of use (culture on solid or liquid agar, resistance of larvae to freezing, and possibility high throughput screening (e.g. 96-well microplates).

 Due to the strong conservation of several intracellular signaling pathways between the nematode and other species including humans, especially in terms of neuronal physiology (Bargman, 1998; and Hobert, 1998; The C. elegans Sequencing consortium, 1998) , the study of

<Desc / Clms Page number 8>

 mechanisms of human disease in C. elegans is recognized as a powerful study tool for the identification and validation (at the cellular level, and before anatomo-functional pharmacological study in more complex models such as mice or primates) of targets therapeutic. More particularly, the inventors set out to study the interactions between a human disease protein, htt, and all of the proteins listed in C. elegans, using the double-hybrid screening technique (Fields and Song, 1987) so as to reconstruct, at least in part, the diagram of the signaling pathways which are involved in the normal functioning of htt before polyQ expansion, and in the pathological functioning of htt after polyQ expansion. Following the double-hybrid screening, the analysis of homologies between the gene of interest identified in C. elegans and its human counterparts makes it possible to effectively predict the nature of the pathological mechanisms at play in HD and to define the first bases of 'a therapeutic intervention, to which the inventors have also focused.

 The data obtained by the inventors therefore relate to the identification of phenomena of variation in interaction between a protein of C. elegans which turns out to be a protein with a WW domain of 946 amino acids and certain fragments of normal or mutated htt (loss of 'interaction). Although the function of ZK1127.9 is unknown in the nematode, this protein has characteristics which strongly suggest that the activity of regulating the transcription of the genes of the human homolog of ZK1127.9 (CA150) is modulated by fragments. N-terminals of htt whose molecular weight is compatible with a passage in the nucleus of neuronal cells through nuclear pores.

 Indeed, the N-terminal fragments of the htt used by the inventors, in particular htt 152, have the capacity, due to their molecular weights (about 10-70 kDa), to penetrate into the nucleus of neurons through the nuclear pores, without passively doubt (Hackam et al., 1998).

 In addition, ZK1127.9 ~ 187-758 shows a 92% loss of interaction with the mutated form httl52Q128 compared to httl52Q15. When the htt size is 546 amino acids, the average interaction loss of

<Desc / Clms Page number 9>

 ZK1127.9 ~ 187-758 for htt546Q128 compared to htt546Q 15 is 56%. The phenomenon of loss of interaction between ZK1127.9 ~ 187-758 and htt is all the more marked as the size of the N-terminal fragments of the mutated htt is reduced. This loss of interaction is likely to be complete for any fragment of the htt containing less than 546 amino acids and more than 90 amino acids. Indeed, ZK1127.9 ~ 187-758 is not able to interact with htt90Q15, which indicates the region 90-546 of the htt contains the binding sites for ZK1127.9. The loss of interaction between ZK1127.9 ~ 187-758 and htt is practically complete for htt152Q128. The demonstration of a loss of interaction between mutated htt and WW domain proteins, and the fact that this loss is dependent on the length of the N-terminal fragment of the htt are new facts, which signal a potentially new mechanism importance of neuronal apoptosis or neuronal dysfunction in HD, and underline the great potential value of the httCA150 couple as a therapeutic target.

 Furthermore, ZK1127.9 has significant sequence homology with the human nuclear transcription factor CA150, a factor whose gene appears to be expressed in the human brain. This strongly suggests that certain forms of the mutated htt are unable to bind CA150 properly in the brains of patients. This also suggests that several human genes are not correctly expressed in HD due to a deficiency in the mechanisms of transcription regulation under the control of the htt-CA150 couple.

 The expression profiles of human ESTs corresponding to the htt and to CA150 indicate that these two proteins are expressed in the brain.

Furthermore, the fact that CA150 is a nuclear protein, that the size of the fragments of the htt tested by the inventors is compatible with a passage in the nucleus, and that the htt is moreover detected in the nucleus of neuronal cells, indicates that CA150 and htt most likely co-locate in the nucleus of brain neurons.

 With regard to the interactions 'htt- ZK1127.9' in vitro: After expression and purification of GST fusion proteins :: htt of ZK1127.9 :: epitopeHA in E. Coli (vector pGEX, and vector pGEX- epitopeHA such as modified by the inventors), fixing of the htt on balls

<Desc / Clms Page number 10>

 sepharose (GST affinity resin, Stratagène), and incubation with ZK1127.9 :: HA, the retention of ZK1127.9 :: HA by normal and mutated htt can be tested by western blot using an anti- HA.

 Regarding htt-CA150 interactions: expression and purification of CA150 (source of the complete cDNA: IMAGE network) in E.

Coli, coupling of CA150 on sepharose beads and incubation with protein extracts from normal lymphoblastoid lines or from patients with HD, and from post-mortem brains (healthy or sick individuals), the retention of htt by CA150 can be tested by western blot using anti-htt antibodies whose epitopes are located at the level of the first 600 amino acids of the htt. The co-localization of CA150 with htt in these human tissues can be verified by conventional subcellular fractionation and western blot techniques as previously described (Sittler et al., 1998). In addition, human-CA150 htt interactions can be tested directly in a double-hybrid in yeast.

 Regarding the function of ZK1127.9 in C. elegans: the data described make it possible to study the role of ZK1127.9 in C. elegans using conventional techniques of transgenesis and mutagenesis in C. elegans: profile of expression of ZK1127.9 in C. elegans after expression in transgenesis of ZK1127.9 :: GFP fusion proteins (typically 1-2 kb of sequence upstream from exon 1 plus the first intron are sufficient for expression of d a protein in C. elegans), and analysis of the phenotype of non-lethal mutants for the ZK1127.9 gene (selection of the mutants by PCR screening of DNA pools corresponding to a library of mutants induced by exposure to TMP and to UV; TMP / UV mutagenesis generating a majority of small deletions, it is particularly suitable for PCR screening).

 Regarding cellular validation of therapeutic targets in a living animal: with confirmation of loss of interaction between mutated htt and CA150 and on the basis of information collected on the function of ZK1127.9, the binding sites between CA150 and normal htt can be identified (for example by deletion mapping) in order to develop in vitro and in vivo tests that allow the identification of active compounds (peptides or organic compounds) capable of reconstituting activity

<Desc / Clms Page number 11>

 normal torque htt-CA150. The in vivo efficacy of such compounds can be directly tested in HD animal models which allow pharmacology studies in terms of toxicity, pharmacokinetics, effective doses (for example in primates, cf. Palfi et al., 1998).

 CA150 is a modulator factor of nuclear transcription which interacts with RNA polymerase type II (Shune et al., 1997), and carries a repeating zone (rich in Alanine and Glutamine) which is present in several other modulating factors of the transcription as GAL1I or SSN6 in yeast, or Zest in Drosophila. Using nucleotide primers at the boundaries of the nucleotide region corresponding to the domain (QA) 60 and after conventional analysis of the PCR products on polyacrylamide gel and autoradiography, the inventors have demonstrated that the domain (QA) 60 of CA150 is polymorphic (4 alleles between 57 and 62 QA, 30% heterozygosity) in the normal population after having tested 20 healthy individuals (10 father-mother couples; CEPH reference families of American origin number 102,1332, 1331, 1340, 1347,1362, 1413, 1416,1423, 884). Due to the polymorphism and the position of the polyQA domain (between 2 WW domains), these data suggest that the polymorphism of QA repeats in CA150 may be involved in the onset and progression of HD in patients, and may modulate the physiological activity of CA150, in particular its link with htt.

Analysis of the homologies of CA150 with the human TSEs available in Genbank and in dbEST has made it possible to establish that the CA150 gene is expressed in several human tissues including the brain (uterus, tonsils, breast, muscle, kidney, lung, parathyroid , testicle, heart, whole embryo, colon, ear, blood). 80 human ESTs (representing 71 cDNA clones) correspond to the CA150 gene (chromosomal location: chromosome 5) according to Unigene analysis (http://www.ncbi.nlm.nih.gov/cgi-bin/Unigene ), including 4 ESTs corresponding to 4 brain cDNA clones (clone IMAGE 38349, and clones GEN-158H06, GEN-121H03, and HFBCA63). Four other human brain cDNA clones corresponding to CA150 (Genbank accession numbers M78358, R87450, AI216964, and HUM121H03A) could be detected by the inventors in the databases.

<Desc / Clms Page number 12>

 These ESTs coming from sequences obtained in several different laboratories, it is difficult to predict with precision the level of expression of the CA150 gene in each of these human tissues from the number of ESTs. However, the fact that human brain cDNA clones make up approximately 11% (8/71 clones) of the human cDNA clones corresponding to CA150 suggests that CA150 is not a high level ubiquitous transcription factor. in all tissues (such as TATA-binding protein), and suggests that CA150 may exhibit preferential activity in certain neuronal subpopulations of the brain. The interest of CA150 is supported by the fact that other PPHs such as HYPM or HYPE (Faber et al., 1998) do not seem to be expressed in the human brain due to the absence of ESTs in the human brain corresponding to their genes (Unigene database).

The interest of CA150 is also supported by the fact that the htt presents a high proportion of EST in the brain, which reflects an expression of the htt in all the cells of the brain including those which are not reached in HD, while that CA150 has a low proportion of brain EST which is likely to reflect selective expression of CA150 in certain neurons in the brain, possibly those affected in HD.

 The inventors therefore set out to highlight molecules capable of mimicking the activity of normal htt or CA150 so as to then use them as a palliative for one or the other of these compounds in order to restore functionality. of the normal htt / CA150 pair.

The molecules in question can be any type of compound (natural molecules, synthetic peptides, organic compounds, combinatorial products) capable of mimicking either the activity of normal htt or any N-terminal fragment of normal htt having an activity similar to that of normal htt not only in its interaction with CA150 but also in the consequences of this interaction, in particular with regard to the expressions or repressions of genes induced by the normal functioning of this couple. Preferably, the N-terminal fragment in question will have at most 546 amino acids.

<Desc / Clms Page number 13>

 As regards the molecules capable of mimicking the activity of CA150, in the context of the present invention, these molecules also include molecules capable of mimicking the activity of a functional fragment of CA150, that is to say of a fragment capable of fulfilling a function similar to that of CA150 not only in its interaction with normal htt but also in the consequences of this interaction, in particular with regard to the expressions or repressions of genes induced by the normal functioning of this couple.

 Thus, the subject of the present invention is a method for screening for molecules capable of modifying the normal htt / CA 150 interaction comprising: a) pre-incubation of the normal htt or of one of its Nterminal fragments with a molecule to be tested , b) the elimination of the test molecules which remained free in the pre-incubation medium, c) the addition to the resulting medium of b) CA 150 or a functional fragment of CA150, d) the elimination of CA 150 or functional fragment of CA 150 after incubation of the medium c), e) the measurement of the affinity constant of the normal htt couple or N-terminal fragment thereof / CA 150 or functional fragment of CA 150 derived from d ), f) the comparison of the affinity constant measured in e) with a standard affinity constant of the normal htt / CA 150 pair.

 In order to complete this process or depending on whether the molecule to be tested is suspected of mimicking normal htt or an N-terminal fragment of this or CA150 or a functional fragment of CA150, it may be interesting to reverse the reagents set in # works in this process by carrying out the preincubation in a) with CA150 or a functional fragment of CA150 and by adding in c) the normal htt or an N-terminal fragment of

<Desc / Clms Page number 14>

 this one. Obviously, in this case, it is the normal htt or the N-terminal fragment thereof which remains free in the medium of c) which will then be eliminated in step d).

 In an advantageous embodiment of the process according to the invention, the compound fixed in step a) of the above process is fixed on solid phase which can be chosen from beads, a tube, a plate, etc.

 In fact, the process according to the invention as written above makes it possible to demonstrate a competition occurring between the molecule to be tested and normal htt or an N-terminal fragment of normal htt or else CA150 or a fragment of CA150, depending on the order of the reagents used. Indeed, if the comparison of the affinity constant measured in e) (of the above process) with a finding of standard affinity of the normal htt / CA150 couple reveals an interaction for said couple weaker than that known under standard conditions, then it will be necessary to be interested in the tested molecule which will probably be at the origin of this loss of interaction as a palliative with one or the other of the elements of the couple in question.

 Thus, in order to deepen such a datum, the method in accordance with the invention may further comprise a step g) in which the affinity constant of the pair tested molecule / normal htt or Nterminal fragment thereof is measured, or well of the couple tested molecule / CA 150 or functional fragment of CA150, and optionally a step h) in which the affinity constant obtained in g) is compared with the standard affinity constant htt normal / CA 150. Such an approach will allow indeed to specify the strength of the interaction of the molecule to be tested with htt or CA150 (or one of their fragment).

 In the case where the above two steps confirm the potential of a tested molecule to interact at the level of the normal htt / CA150 pair, it is necessary to add to the process according to the invention a step i) in which the tested molecule is identified. for which the affinity constant measured in step g) is significant of an interaction with normal htt or an N-terminal fragment thereof or with CA150 or a functional fragment of CA150. This molecule is in fact of particular interest and the subject of the present invention is also its use for the

<Desc / Clms Page number 15>

 preparation of a medicament intended for the curative or preventive treatment of Huntington's disease. Indeed, this molecule is capable of restoring all or part of a normal functionality of the normal htt / CA150 pair and thus of preventing neuronal apoptosis or the neuronal dysfunction induced by mutated htt in the cells concerned. It is not excluded that such a molecule present with normal htt or an N-terminal fragment thereof or with CA150 or a functional fragment of CA150, an affinity constant greater than the known standard affinity constant for the normal htt / CA150 pair. In this case, this type of molecule should be used on purpose.

 The present invention also relates to a method of identifying therapeutic targets for Huntington's disease by the comparative analysis of the expression profiles of mRNAs from healthy cells with expression profiles of mRNAs from cells in which Losses of normal htt / CA150 interactions are suspected. By mRNA is meant not only total mRNA but also certain mRNA (by Northern blot, by sequencing of EST from cDNA libraries, by cDNA display, by subtractive hybridization of mRNA or cDNA, using DNA membranes or microarrays, or any combination of these methods). In the context of the present invention, these mRNAs can be not only derived from cells but also from whole tissues or organisms (human or non-human). When these cells, tissues or organisms are said to be healthy, it should be understood: in the normal state, that is to say showing no expression of the mutated htt. When it comes to cells, tissues or organisms in which losses of normal htt / CA150 interaction are suspected, it should be understood that this is a pathological condition with expression of the mutated htt.

 Such a method makes it possible to demonstrate an under-expression or an over-expression of genes in cells suspected of having a loss of normal htt / CA150 interaction with respect to healthy cells, and the nucleotide sequences corresponding to all or part of these genes can be used in the context of the present invention for the preparation of a medicament intended for the curative or preventive treatment of Huntington's disease.

<Desc / Clms Page number 16>

In the case where the above process reveals that a gene is under-expressed in the diseases suspected of having losses of normal htt / CA150 interaction, the above-mentioned use will consist in introducing the nucleotide sequence corresponding to all or part of this gene in the cells concerned so as to re-express this gene in situ and thus restore the functionality of the normal htt / CA150 pair. Such an approach can for example be carried out by gene therapy
In the case where the process described above makes it possible to demonstrate an overexpression of a gene in cells suspected of having losses of normal htt / CA150 interaction, the above use will consist in blocking the expression in situ of this gene for example by means of an antisense strategy.

 The present invention also relates to the use of a nucleotide sequence corresponding to all or part of the CA150 gene or of a functional fragment of CA150 in order to over-express it in cells suspected of having normal htt interaction losses. / CA150, this as part of a curative or preventive treatment for Huntington's disease. Indeed, in the case where normal CA150 is deficient or in the case where pathological cells contain both mutated htt and normal htt, such an approach consisting in locally over-expressing CA150 would have the effect of shifting the balance of interactions by saturation of the mutated htt and would therefore make available the surplus CA150 for the normal htt present in said cells.

 This could also be the case with hyperactive analogs of CA150 capable of restoring functionality at least equivalent to the standard functionality of the normal htt / CA150 pair.

 The present invention also relates to an anti-mutated htt antibody capable of discriminating between normal htt and mutated htt.

It also relates to the use of such an antibody for the preparation of a medicament intended for the curative or preventive treatment of Huntington's disease. Such a process of neutralizing the mutated huntingtin could moreover be entirely accompanied by a local overexpression of CA150 or a functional fragment of CA150 and / or normal htt or an N-terminal fragment of htt normal so as to reintroduce the

<Desc / Clms Page number 17>

 partners necessary to restore the functionality of the normal htt / CA150 pair. As an example, this could be achieved through gene therapy.

 The present invention is not limited to the above description. It will be better understood in the light of the experiments set out below.

PRESENTATION OF THE FIGURES Figure 1. Quantitative test (ONPG test in liquid phase) for the interactions between N-terminal fragments of the htt and the protein ZK1127.9 ~ 187-758.

Htt90 corresponds to the product of exon 1 of the htt gene. Htt 152 is a fragment that is more likely to enter the neuron nucleus than htt546. The values represented correspond to the means over 45 in vivo measurements (9 independent cultures; 5 measurements / point / culture).

The intervals represent the limits of the values obtained. #: percentage of link difference between allele 15 Gins and allele at 128 Gins. The distributions of ONPG values were compared by ANOVA analysis using Macintosh Statview software.

Figure 2. Comparison of the proteins ZK.1127.9 and CA150. The annotated sizes are in number of amino acids. Black squares: WW domains. Hatched area: polymorphic domain (QA) 60 in the normal population, 4 alleles, 30% heterozygacy.

METHODS AND DATA
The constructs intended for double-hybrid screening and encoding the htt were created in the vector pGBT9 (Genbank accession number U07646, Clontech) in the form of fusion proteins with the binding domain of the GAL4 protein (GAL4BD) for the specific DNA sequences that have been introduced into the genome of host yeasts. The cDNA fragments coding for htt are from the laboratory of Michael Hayden, University of British Columbia, Vancouver, Canada. All constructions have been sequenced in order to verify the phasing

<Desc / Clms Page number 18>

 of reading the htt cDNAs, the number of CAG repeats, and the absence of point mutations or deletions. The C. elegans cDNA library (cDNAs generated using an oligo-dT primer) was subcloned into the vector pACT2 (Genbank accession number U29899, Clontech) in the form of fusion proteins with the GAL4 transcription activation domain (GAL4AD). This cDNA bank is freely available and comes from the laboratory of R. Barstead (Oklahoma Medical Research Foundation, OK, USA). The double-hybrid screening was carried out using a protocol based on the fusion of mat-a and mat-alpha yeasts (Bendixen et al., 1994), the mat-a yeasts (strain CG1945, Clontech) expressing in the occurrence of the fusion proteins [GAL4BD :: htt], and the mat-alpha yeasts (strain Y187, Clontech) expressing the fusions [protein fragments of C. elegans :: GAL4AD].

 The construct used by the inventors for screening (fusion protein with GAL4BD) was htt546Q15 (Figure 1). After screening 3.8 108 clones (fusion efficiency greater than 33%), 56 positive clones were detected. After elimination of false positives (based on a double-hybrid interaction with the fusion protein GAL4BD :: lamine) and verification of the redundancy of true positives by analysis of restriction profiles (enzymatic digestion of cDNAs selected at using the restriction enzymes EcoRI and BamHI), the inventors sequenced the 5 ′ and 3 ′ ends of the resulting unique cDNA clones (ie 2 clones). Search for sequence homology between the 3 'and 5' ESTs (Expressed Sequenced Tag) obtained for these clones and the proteins of C. elegans on the BLAST C. elegans server at the Sanger Center, Cambridge, UK (http: // www.sanger.ac.uk/Projects/C~elegans/blast~server.shtml) using BLAST programs (Altschul et al., 1997) enabled the inventors to identify the protein ZK1127.9 as a partner for htt546Q15. As indicated in the Acedb public database, the gene for ZK1127.9 is found on chromosome II (cosmid ZK1127) of C. elegans. The sequence of the protein ZK1127.9 was predicted from the analysis of the genomic sequence of C. elegans (The C. elegans Sequencing consortium, 1998) and contains 946 amino acids. Comparison of the 3 'end sequences and

<Desc / Clms Page number 19>

 5 'of the cDNA clone corresponding to ZK 1127.9 with the C. elegans library of the Sanger Center using the BLASTN and BLASTX programs made it possible to establish that the cDNA clone corresponding to the gene for the protein ZK1127.9 codes for amino acids 187 to 758 of this protein. The binding sites of htt546Q15 on ZK1127.9 are therefore located between amino acids 187 to 758 of ZK1127.9 (ZK1127.9 ~ 187-758). The interaction profile of ZK1127.9 187-758 with different forms of htt (htt546Q15, htt546Q128, httl52Q15, httl52Q128) was analyzed by the inventors using double-hybrid tests (Bendixen et al., 1996 ), initially qualitatively after selection of positive clones on agar and transfer to nylon filters, and then quantitatively after selection of positive clones in liquid phase, protein extraction, and ONPG test in liquid phase (Kalchman et al., 1996). The ONPG tests were carried out 45 times, ie from 9 independent yeast cultures, and from 5 measurements / case / culture (Figure 1). Statistical analysis of the data was performed using Statview software (Macintosh) and analysis of variances on the distribution of values obtained for each case (Figure 1).

Sequence homologies of ZK1127.9 and ZK1127.9 ~ 187-758 with other genes have been sought by the inventors in several public databases including: - Genbank at NIH, MD (http: //www.ncbi nlm.nih.gov) - the WWW server BLAST C. elegans of the Sanger Center, Cambridge,
UK (http://www.sanger.ac.uk/Projects/C~elegans/blast~server.shtml) - the EBI WWW server, Cambridge, UK (http://www2.ebi.ac.uk / servicestmp)
This research was carried out using BLAST programs (Altschul et al., 1997) or using programs based on the Smith and Waterman algorithm (Smith and Waterman, 1981) such as the bic2 program (EBI) . In addition, conserved domains were searched in ZK1127.9, ZK1127.9 ~ 187-758, and CA150 using the Pfam program (http://www.sanger.ac.uk/Softwares/Pfam/search. shtml) These analyzes made it possible to highlight the following facts (Figure 2):

<Desc / Clms Page number 20>

- ZK1127.9 ~ 187-758 contains 3 WW domains (Pirozzi et al., 1997) at positions 200-225,262-292, and 378-407 amino acids (0.00019 <Evaluated <0.011). The WW domains are 35-40 amino acid motifs characterized by the presence of 4 conserved aromatic residues, two of which are tryptophans, are mediators of protein-protein interactions, and classically interact with polyproline domains (Pirozzi et al., 1997). It is therefore very likely that ZK1127.9 ~ 187-758 interacts with the proline-rich region of the htt (htt: Genbank accession number L12392) which is located immediately in
C-terminal of the polyQ domain (between amino acids 40 and 100) and which contains 2 polyprolines of 11 and 10 consecutive prolines. It is also very likely that the binding of htt to CA150 involves the polyprolines and the immediately C-terminal region (on at least 30 amino acids) to the polyproline domains in the htt and the WW domains at positions 133-
162,431-460 (corresponding to the WW 262-292 domain of ZK1127.9), and
532-559 aminocides (10 <Evaluated <0.0016) from CA150. The mediating role of the htt polyproline-rich region for interaction with WW domain proteins has been illustrated elsewhere (Faber et al., 1998).

- the region of ZK1127.9 going from amino acids 267 to 944 exhibits significant homology (identical amino acids: approximately 38%; similar amino acids: approximately 69%; p = 1213 in Smith and Waterman analysis, and p (N) = 10-137 in BLAST analysis) with the region 435 to 1095 amino acids of the human protein CA150 (Genbank accession number AF017789). There are also other weaker and non-significant homologies for short segments of ZK1127.9, such as for example homologies with WW domains present in other proteins like HYPA or
HYPB (Faber et al., 1998).

<Desc / Clms Page number 21>

 REFERENCES Altschul SF, Thomas LM, Schaffer AA, Zhang J, Zhang Z, Miller W, and Lipman DJ. Gapped BLAST and PSI-BLAST. a new generation of protein database search programs. Nucleic Acids Res 25, 3389-3402, 1997.

Bates GP, Mangiarini L, Davies SW. Transgenic mice in the study of polyglutamine repeat expansion diseases. Brain Pathol 8, 699-714,1998.

Bargman CI. Neurobiology of the C. elegans genome Science 282,2028-2033, 1998.

Bendixen C, Gangloff S, and Rothstein R. A yeast mating-selection scheme for detection of protein-protein interactions. Nucleic Acids Res 22, 1778-1779, 1994.

Bertaux F, Sharp AH, Ross CA, Lehrach H, Bates GP, and Wanker E.

HAP1-huntingtin interactions do not contribute to the molecular pathology in the Huntington's disease transgenic mice. FEBS letter 17, 229-232, 1998.

Boutell JM, Wood J, Harper PS, and Jones L. Huntingtin interacts with cystathionine B-synthase. Hum Mol Genet 7, 371-378, 1998.

Burke JR, Enghild JJ, Martin ME, Jou YS, Myers RM, Roses AD, Vance JM, and Strittmater WJ. Huntingtin and DRPLA proteins selectively interact with the enzyme GAPDH. Nature Medicine 2, 347-350, 1996.

Davies SW, Turmaine M, Cozens B, DiFiglia M, Sharp AH, Ross CR, Scherzinger E, Wanker EE, Mangiarini L, and Bates G. Formation of neuronal intranuclear inclusions underlies the neurological dysfunction in mice transgenic for the HD mutation. Cell 90,537-548, 1997.

De Rooij KE, Dorsman JC, Smoor MA, Den Dunnen JT, and Van Ommen GJB. Subcellular localization of the Huntington's disease gene product in

<Desc / Clms Page number 22>

 cell lines by immunofluorescence and biochemical subcellular fractionation. Human Molecular Genet 5,1093-1099, 1996.

DiFiglia M, Sapp E, Chase KO, Davies SW, Bates GP, Vonsattel JP, & Aronin N. Aggregation of huntingtin in neuronal intranuclear inclusions and dystrophic neurites in brain. Science 277,1990-1993, 1997.

Engelender et al. 1997 Faber PW, Barnes GT, Srinidhi J, Chen J, Gusella JF, and MacDonald ME. Huntingtin interacts with a family of WW domain proteins. Hum Mol Genet 7, 1463-1474, 1998.

Faber PW, Alter JR, MacDonald ME, and Hart A. Polyglutamine-mediated dysfunction and apoptotic death of a C. elegans sensory neuron. Proc Natl Acad Sci USA 96, 170-184, 1999.

Fields S and Song OK. A novel genetic system to detect protein-protein interactions. Nature 340,245-246, 1989.

Goldberg YP, Nicholson DW, Rasper DM, Kalchman MA, Koide HB, Graham RK, Bromm M, Kazemi-Esfarjani P, Thornberry NA, Vaillancourt JP, and Hayden MR. Cleavage of huntingtin by apopain, a proapoptotic cysteine protease, is modulated by the polyglutamine tract. Nature Genet 13, 442-449,1996.

Hackam AS, Singaraja R, Zhang T, Gan L, and Hayden MR. In vitro evidence for both the nucleus and cytoplasmas subcellular sites of pathogenesis in Huntington's disease. Hum Mol Genet 8, 25-33, 1999.

Hoogeveen AT, Willemsen R, Meyer N, De Rooij KE, Roos RAC, and Van Ommen GJB, and Galjaard H. Characterization and localization of the Huntington disease gene product. Hum Mol Genet 2,2069-2073, 1993.

<Desc / Clms Page number 23>

 Jackson GR, Salecker I, Dong X, Yao X, Arhheim N, Faber PW, MacDonald ME, and Zipursky SL. Polyglutamine-expanded human huntingtin transgenes induce degeneration of Drosophila photoreceptors neurons. Neuron 21, 633-642, 1998.

Kalchman MA, Graham RK, Xia G, Koide HB, Hodgson JG, Graham KC, Goldberg YP, Gietz RD, Pickart CM, and Hayden MR. Huntingtin is ubiquitinated and interacts with a specifies ubiquitin-conjugating enzyme J Biol Chem 271,19385-19394, 1996.

Kalchman MA, Koide HB, McCutcheon K, Graham RK, Nichol K, Nishiyama K, Kazemi-Esfarjani P, Lynn FC, Wellington C, Metzler M, Goldberg YP, Kanazawa I, Gietz RD, and Hayden MR. HIP1, a human homologue of S. cerevisae Sla2p, interacts with membrane-associated huntingtin in the brain. Nat Genet 16, 44-52, 1997.

Koshy BT and Zoghbi HY. The CAG / polyglutamine tract diseases: gene products and molecular pathogenesis. Brain Pathol 7, 927-942, 1997.

Li X-J, Li S-H, Sharp AH, Nucifora Jr FC, Schilling G, Lanahan A, Worley P, Snyder SH, Ross CA. A Hungtintin-associated protein enriched in brain with implications for pathology. Nature 378, 398-402, 1995.

Lunkes A and Mandel JL. A cellular model that recapitulates major pathogenic steps of Huntington's disease Hum Mol Genet 7,1355-1361, 1998.

Mangiarini L, Sathasivam K, Seller M, Cozens B, Harper A, Hetherington C, Lawton M, Trottier Y, Lehrach H, Davies SW, and Bates G. Exon 1 of the HD gene with an expanded CAG repeat is sufficient to cause a progressive neurological phenotype in transgenic mice Cell 87,493-506, 1996.

<Desc / Clms Page number 24>

Martindale D, Hackam A, Wieczoreck A, Ellerby L, Wellington C, McCutcheon K, Singaraja R, Kazemi-Esfarjani P, Devon R, Kim SU, Bredesen DE, Tufaro F, and Hayden MR. Length of huntingtin and its polyglutamine tract influences localization and frequency of intracellular aggregates. Nature Genet 18,150-154, 1998.

Néri C, Du Hongping, Morinière S, and Chalfie C. A C. elegans model of neuron dysfunction in polyglutamine expansion diseases: Suggestiveevidence for polyglutamine-dependent intracellular aggregation and touch receptor neuron dysfunction. European Worm Meeting, August 1998, Hinxton, UK, p. 43.

Palfi S, Conde F, Riche D, Brouillet E, Dautry C, Mittoux V, Chibois A, Peschanski M, and Hantraye P. Fetal striatal allografts reverse cognitive deficits in a primate model of Huntington disease. Nat Med 4,963-966,
1998.

Pirozzi G, McConnell SJ, Uveges AJ, Carter JM, Sparks AB, Kay BK, Fowlkes DM Identification of novel human WW domain-containing proteins by cloning of ligand targets. J Biol Chem 272, 14611-14616, 1997.

Ross CA. Intranuclear neuronal inclusions: common pathogenic mechanism for glutamine-repeat neurodegenerative diseases? 19,
1147-1150, 1997.

Ruvkun G and Hobert O. The taxonomy of developmental control in C. elegans. Science 282, 2033-2041, 1998.

Saudou F, Finkbeiner S, Devys D, and Greenberg M. Huntingtin acts in the nucleus to induce apoptosis but death does not correlate with the formation of intranuclear inclusions. Cell 95, 55-66, 1998
Schilling G, Becher MW, Sharp AH, Jinnah HA, Duan K, Kotzuk JA, Slunt HH, Ratovitski T, Cooper JK, Jenkins NA, Copeland NG, Price DL, Ross

<Desc / Clms Page number 25>

 CA, and Borchelt DR. Intranuclear inclusions and neuritic aggregates in transgenic mice expressing a mutant N-terminal fragment of huntingtin.

Hum Mol Genet 8, 397-407, 1999.

Sisodia SS. Nuclear inclusions in glutamine repeat disorders: they pernicious, coincidental, or beneficial? Cell 95,1-4, 1998.

Sittler A, Walter S, Wedemeyer N, Hasenbank R, Scherzinger E, Eickhoff H, Bates GP, Lehrach H, and Wanker E. SH3GL3 associates with the huntingtin Exon 1 protein and promotes the formation of polygln-containing aggregates. Molecular Cell 2,427-436, 1998.

Sharp AH, Zhang C, Cannon M, Becher MW, Cooper JK, Schilling G, Borchel DR, Ross CR. Possible cleavage products of huntingtin in HD nuclear inclusions and huntingtin transgenic mice. Abstract in: Huntington's Disease: progress. Proceedings of the Hereditary Disease Foundation meeting, New Seabury, MA; 9-10.1998.

Shune C, Hayashi T, Liu Y, Lane WS, Young RA, Garcia-Blanco MA.

CA150, a nuclear protein associated with the RNA polymerase II holoenzyme, is involved in Tat-activated immunodeficiency virus type 1 transcription. Mol Cell Biol 17,6029-6039, 1997.

Smith and Waterman. Identification of common molecular subsequences. J Mol Biol 147, 195-197, 1981.

The C. elegans sequencing consortium. Genome sequence of the nematode C. elegans: A platform for investigating biology. Science 282,2012-2018, 1998.

Wanker EE, Rovira C, Scherzinger E, Hasenbank R, Walter S, Trait D, Colicelli J, and Lehrach H. HIP-I: a huntingtin interacting protein isolated by the yeast two-hybrid system. Hum Mol Genet 6,487-495, 1997.

<Desc / Clms Page number 26>

 White JK, Auerbach W, Duyao MP, Vonsattel JP, Gusella JF, Joyner AL, and MacDonald M. Huntingtin is required for neurogenesis and is not impaired by the Huntington's disease CAG expansion. Nature Genet 17, 404-410,1997.

Zeitlin S, Liu J-P, Chapman DL, Papaioannou VE, Efstratiadis A. Increased apoptosis and early embryonic lethality in mice nullizygous for the Huntington's disease gene homologe. Nature Genet 11,155-163, 1995.

Claims (12)

1. Method for screening molecules capable of modifying the normal htt / CA150 interaction comprising: a) pre-incubation of the normal htt or one of its Nterminal fragments with a molecule to be tested, b) elimination of the molecules to be tested remained free in the pre-incubation medium, c) the addition to the resulting medium of b) of CA150 or of a functional fragment of CA150, d) elimination of CA150 or functional fragment of CA150 after incubation of the medium c), e) the measurement of the affinity constant of the normal htt couple or N-terminal fragment thereof / CA150 or functional fragment of CA150 from d), f) the comparison of the affinity constant measured in e) with a standard affinity constant of the normal htt / CA150 couple.  2. Method according to claim 1 in which: - the pre-incubation in a) is carried out with CA150 or a functional fragment of CA150, - in c) the normal htt or N-terminal fragment thereof is added, and - in d) the normal htt or N-terminal fragment thereof which remains free in the medium of c) is eliminated.  3. Method according to claim 1 or 2 in which step a) is carried out by fixing the normal htt, an N-terminal fragment thereof, CA150 or a functional fragment of CA150, on solid phase. <Desc / Clms Page number 28>  4. Method according to one of claims 1 to 3, further comprising a step g) in which the affinity constant of the pair tested molecule / normal htt or N-terminal fragment thereof or CA150 or functional fragment is measured of CA150, and optionally a step h) in which the affinity constant obtained in g) is compared with the standard affinity constant htt normal / CA150.  5. Method according to claim 4 further comprising a step i) in which the tested molecule is identified for which the affinity constant measured in step g) is significant of an interaction with normal htt or N-terminal fragment of this or CA150 or functional fragment of CA150.  6. Use of the molecule identified according to claim 5 for the preparation of a medicament intended for the curative or preventive treatment of Huntington's disease.  7. Method for identifying therapeutic targets for Huntington's disease by comparative analysis of expression profiles of mRNAs from healthy cells with expression profiles of mRNAs from cells in which normal htt interaction losses CA150 are suspected.  8. Use of a nucleotide sequence corresponding in whole or in part to a gene under-expressed in cells suspected of having losses of normal htt / CA150 interaction with respect to healthy cells in accordance with the method according to claim 7, for the preparation of a medicament intended for the curative or preventive treatment of Huntington's disease.  9. Use of a nucleotide sequence corresponding in whole or in part to a gene overexpressed in cells suspected of having losses of normal htt / CA150 interaction with respect to healthy cells according to the method according to claim 7, for the preparation of a <Desc / Clms Page number 29>  medicine intended for the curative or preventive treatment of Huntington's disease.  10. Use of a nucleotide sequence corresponding to all or part of the CA150 gene or of a functional fragment of CA150 for the preparation of a medicament intended for the curative or preventive treatment of Huntington's disease by over-expression of said sequence in cells suspected of having normal htt / CA 150 interaction losses.  11. Mutated anti-htt antibodies.  12. Use of an antibody according to claim 11 for the preparation of a medicament intended for the curative or preventive treatment of Huntington's disease.