Patent Application
20070212712
Hedgehog peptides (Hh) are signaling ligands with key roles in stem cell specification and growth for tissue patterning in embryos and for stem cell renewal of adult tissues (reviewed by Ingham and McMahon, Genes Dev., 15:3059-3087, 2001; Beachy, et al., Nature, 432:324-31, 2004; U.S. Pat. Nos. 5,789,543 and 5,844,079 to Ingham, et al., incorporated herein by reference). Hh proteins are unique in vertebrates in that they are capable of autoprocessing, i.e., they are capable of self-cleavage thereby converting from a precursor peptide to an active mature peptide. Vertebrates have three Hh proteins, Sonic (Shh), Indian (Ihh) and Desert (Dhh) Hedgehogs. Hh activate the signaling in responsive stem cells by binding to the receptor Patched (Ptc), which relieves Ptc repression of Smoothened, a G-related protein, leading to the nuclear translocation of Gli transcriptional factors to activate and repress specific target genes controlling stem cell growth, apoptosis, differentiation and angiogenesis (Byrd, et al., Development, 129:361-72, 2002).
Hh have been identified as a tumor growth signal in a diversity of human cancers (Berman, etal., Nature, 425:846-851, 2003; Karhadkar, etal., Nature. 431:707-12, 2004). Tumors originating in esophagus, stomach, biliary tract, pancreas and prostate express high levels of Shh and Ihh, which stimulates the growth of stem cell-like tumor cells. Functional neutralization antibodies against Shh and Ihh have been shown to block the growth of these tumors in vitro and in xenografts, establishing that these tumors are completely dependent on Hh ligand for their growth. However, due to the bulky nature of antibodies, the use of antibodies intracellularly in organisms is problematic.
Since hundreds of thousands of people are diagnosed and die from cancer each year, is it critical that novel methods are developed for the detection of agents that can inhibit cancer growth. The identification of modulators (e.g., inhibitors) of Hh autoprocessing is a novel area ripe with possibility. There are no known prior art small molecule inhibitors for Hh autoprocessing. Moreover, there are not any known efficient methods or techniques leading to the development of an in vitro high-throughput (HTS) system for the identification of such inhibitors. Therefore, what is needed are techniques and methods for the efficient high-throughput screening of agent libraries (e.g., small molecule libraries) for the identification of molecules that are effective in the modulation of hedgehog autoprocessing.
The present invention relates to methods for the identification of modulators of Hh autoprocessing. In a preferred embodiment, the invention relates to methods of identifying inhibitors of Hh autoprocessing utilizing in vitro cell-free or cell-based methods. In another aspect, the present invention relates to the screening of small molecule and combinatorial chemical libraries with the high-throughput, cell-free methods of the present invention. In another embodiment, the methods of the present invention are automated.
In other aspects, the present invention relates to the use of rigorous controls in said methods of high-throughput, cell-free screening to lessen false-positive and false-negative identifications of potential small molecule modulators of Hh autoprocessing. In yet another embodiment, the invention relates to using variations of the methods of the present invention in cell-based systems.
Hedgehog (Hh) autoprocessing plays an essential role in embryonic development (U.S. Pat. No.: 6,165,747 to Ingham, et al, incorporated herein by reference) and is involved in stem cell fate and the etiology of certain cancers. Examples are tumors of endodermal origin such as pancreatic, digestive tract and prostate cancers, etc. Hh autoprocessing may also play a role in fat absorption and be responsible for weight gain and hepatic steatosis in adults. The present invention relates to assays for the identification of reagents and agents that are effective in the modulation (i.e., the inhibition and promotion) of Hedgehog protein autoprocessing. One skilled in the art will realize the reagents and molecules identified by the methods of the present invention will be useful in the treatment of various illnesses and disease states related to the aberrant regulation of Hh autoprocessing such as tumors, angiogenesis, weight gain and loss and hepatic steatosis.
In another aspect of the invention, the reagents identified by the methods of the present invention as being effective in the modulation of Hh autoprocessing are also embodied in the presented invention for use in the modulation of Hh autoprocessing for the treatment, for example, of tumors, angiogenesis, weight gain and loss, hepatic steatosis, etc.
There are no known prior art organic small molecule modulators (i.e., inhibitors and promoters) for Hh autoprocessing. Moreover, there are no prior art protocols for an in vitro high-throughput (HTS) system for the identification of such Hh modulators. The following U.S. Patents by Beachy and coworkers have issued, however, and, although they relate to Hh and, in part, the modulation of Hh autoprocessing, none of the cited patents disclose high-throughput cell-free screening systems nor methods comprising rigorous control systems for the identification of modulators of Hh autoprocessing: U.S. Pat. Nos. 6,911,528, 6,867,216, 6,733,971, 6,432,970, 6,288,048, 6,281,332, 6,277,566, 6,214,794, 6,132,728 and 6,057,091 (all of which are incorporated herein by reference).
Unlike the present invention, U.S. Pat. No. 6,057,091 discloses the use of Hh fusion proteins for identifying autoprocessing inhibitors in cell-based systems with non-stringent control systems. Among the disclosures of the other patents are cell lines similar to those used in studying the inhibition of Hh processing in intact cells as well as upstream or downstream inhibitors of hedgehog signaling (affecting either cholesterol synthesis or downstream signaling steps but not Hh autoprocessing itself).
Cell-free Assay Systems
In one aspect of the present invention, a method is contemplated for the screening of compounds that inhibit the autoprocessing of the Hedgehog protein in a cell-free system.
Such cell-free systems are composed of appropriate purified, genetically engineered fusion proteins to be used for various screening assays wherein test reagents are added to mixes of test and/or control peptides suitable for Hh autoprocessing. Such assays are conduced in a physiological environment such as a neutral pH range and physiological temperature. Detection can then be performed via an automated system (such as fluorescence-based plate readers) or by direct visual observation (for example, via the use of fluorescence detection or antibody binding and detection).
In a first compound screening embodiment of the present invention, the cell-free system includes a non-fluorescent or non-luminescent fusion protein, said fusion protein comprising: (1) a C-terminally truncated form of a protein which is fluorescent or luminescent in its full-length form; and (2) a Hedgehog autoprocessing domain fused in-frame to the C-terminal residue of the truncated form of the protein which is fluorescent or luminescent in its full-length form; wherein cleavage of the fusion protein by Hedgehog autoprocessing function results in fluorescence of luminescence of the C-terminally truncated protein.
The luminescent or fluorescent protein of the present invention may be selected from any number of reagents known to those practiced in the art. For example, the luminescent protein may be luciferase whereas the fluorescent protein may be selected, for example, from a group such as GFP, BFP, CFP and YFP and the red fluorescent protein from Discosoma and aequorin and other fluorescent molecules such as DsRed monomer, DsRed2, DsRed-express, AsRd2, HcRed1, AmCyan1, ZsYellow1, and ZsGreen1from Clontech, Mountain View, Calif. or similar products from Evrogen, Moscow, Russia; Invitrogen, Carlsbad, Calif., etc.
In this first compound screening embodiment, the cell-free system also includes a nucleophile initiator. A nucleophile is a reagent that is attracted to centers of positive charge. A nucleophile participates in a chemical reaction by donating electrons to a species known as an electrophile in order to form a chemical bond. In the context of the present invention, a nucleophile initiator is a reagent that serves to initiate the autoprocessing reaction of the Hedgehog protein. In a preferred embodiment, the nucleophile initiator is a sterol (e.g., cholesterol, desmosterol, β-sitosterol, 7β-hydroxycholesterol, ergosterol, or 7-dehydrocholesterol) or a thiol.
Also provided in connection with this embodiment is a compound to be tested for the ability to inhibit (either partially or totally) Hedgehog autoprocessing. The test reagents of the present invention may be selected from small molecule libraries and other libraries including combinatorial chemical libraries. Such libraries are known in the art and are available commercially. Additionally, proprietary libraries are also available for use from collaborators and others. Additionally, the synthesis and screening of small molecule libraries (e.g., combinatorial chemical libraries) are well known in the art (See, for example, U.S. Pat. No. 6,060,596 to Lerner; U.S. Pat. No. 6,185,506 to Cramer, et al.; U.S. Pat. No. 6,377,895 to Horlbeck; U.S. Pat. No. 6,936,477 to Still, et al.; Shipps, et al., Proc. Natl. Acad. Sci. USA, 94:11833-11838, 1997; Stockwell, et al., Chemistry & Biology, 6:71-83, 1999, all of which are incorporated herein by reference). (See, also, for example, www.combichem.net; www.combichemistry.com; www.combinatorial.com and pubs.acs.org/journals/jcchff/).
Candidate reagents encompass numerous chemical classes, though typically they are organic molecules, preferably small organic compounds having a molecular weight of more than 50 and less than about 3,500 daltons. Candidate reagents comprise functional groups necessary for structural interaction with proteins, particularly hydrogen bonding and typically include at least an amine, carbonyl, hydroxyl or carboxyl group, preferably at least two of the functional chemical groups. The candidate agents often comprise cyclical carbon or heterocyclic structures and/or aromatic or polyaromatic structures substituted with one or more of the above functional groups. Candidate agents are also found among biomolecules including peptides, saccharides, fatty acids, steroids, purines, pyrimidines, derivatives, structural analogs or combinations thereof. Non-limiting examples of suitable small molecule libraries are given in Example 5.
Candidate reagents are obtained from a wide variety of sources including libraries of synthetic or natural compounds. For example, numerous means are available for random and directed synthesis of a wide variety of organic compounds and biomolecules, including expression of randomized oligonucleotides and oligopeptides. Alternatively, libraries of natural compounds in the form of bacterial, fungal, plant and animal extracts are available or readily produced. Additionally, natural or synthetically produced libraries and compounds are readily modified through conventional chemical, physical and biochemical means, and may be used to produce combinatorial libraries. Known pharmacological agents may be subjected to directed or random chemical modifications, such as acylation, alkylation, esterification, amidification, etc. to produce structural analogs.
The contemplated method comprises forming an incubation mixture including the compositions recited above (i.e., the cell-free system, the fusion protein, the nucleophile initiator and the test reagent), incubating the mixture under conditions and for a period of time appropriate for Hedgehog autoprocessing, detecting fluorescence or luminescence in the incubation mixture and comparing the fluorescence or luminescence detected to fluorescence or luminescence detected using an otherwise identical incubation mixture which lacks the test reagent, a decrease in fluorescence or luminescence being indicative of the presence of a compound that inhibits autoprocessing of Hedgehog. Times and conditions appropriate for Hedgehog autoprocessing are typically about 3 hours at about 37° C. and pH 7. However, these conditions are not critical. For example, the temperature can be varied between about 20° C. and 40° C. and the pH from about 6.5 to 8.0.
The detection of luminescence or fluorescence as the result of Hedgehog autoprocessing may be performed by, for example, spectroscopy, automated plate readers, fluorescent (or luminescent) microscopy, etc. Such techniques are well known to those practiced in the art.
Another aspect of the present invention relates to second compound screening assay that is carried out in a cell-free system. In connection with this method, provided are a fusion protein consisting of a Hedgehog protein in which a large portion or most of the N-terminal signaling domains replaced by a small peptide that can be selectively fluorescently labeled. Following the selective labeling, the cleavage of the fusion protein by the Hedgehog autoprocessing function releases the fluorescently labeled small peptide thereby resulting in a detectable change in fluorescence polarization. The small peptide, which can be selectively labeled can be fused to the truncated Hedgehog protein without impairing cholesterol-dependent autoprocessing (Porter, et al., Nature, 374:363-366, 1995).
The method comprises, for example, forming a cell-free incubation mixture comprising the Hedgehog fusion construct as described above, a weakly or non-fluorescent reagent for the fluorescent labeling of the fusion protein, a nucleophile initiator and a compound to be tested for the ability to inhibit (partially or totally) Hedgehog autoprocessing. This mixture is incubated under conditions and for a period of time appropriate for Hedgehog autoprocessing. Fluorescence polarization in the mixture is detected and compared with fluorescence polarization detected in an otherwise identical incubation mixture that lacks the test compound. A decrease in fluorescence polarization in the mixture with the test compound relative to that determined for the otherwise identical incubation mixture which lacks the test compound, is indicative of the ability of the test compound to inhibit autoprocessing of Hedgehog protein.
As mentioned previously, the N-terminal Hedgehog signaling domain is replaced by a small polypeptide without impairing cholesterol-dependent autoprocessing (See, for example, Porter, et al., Nature, 374:363-366, 1995, which is incorporated herein by reference). Since the C-terminal domain of the Hedgehog peptide is the autoprocessing domain, the fusion of a tag on the N-terminal does not affect autoprocessing. A preferred peptide includes the sequence Cys-Cys-X-X-Cys-Cys [SEQ ID NO.: 1], where X is any amino acid. In a preferred embodiment, X-X are the amino acids Gly-Pro giving the sequence Cys-Cys-Gly-Pro-Cys-Cys [SEQ ID NO.: 2]. Such sequences are rare in cells or cell-free systems unless added. These sequences bind to a fluorescent biarsenical ligand with a nanomolar or lower dissociation constant. Biarsenical molecules are weakly fluorescent dyes that bind extremely tightly to very small peptide motifs containing four cysteines, which can be genetically incorporated into proteins, and upon binding become intensely fluorescent. The dyes readily cross biological membranes and are relatively nonfluorescent until they find the target peptide sequences. An example of a suitable fluorescent biarsenical is 4′,5′-bis(1,3,2-thothioarsolan-2-yl)fluorescein, which, upon binding to -Cys-Cys-X-X-Cys-Cys-, yields intense fluorescence emission at 535 nm upon excitation at 480 nm, whereas the free biarsenical reagent has a 50,000-fold lower fluorescence. Numerous other fluorescent biarsenical molecules as well as biarsenical molecules that may be conjugated to fluorescent moieties are known in the art (for example, see, Zhang, et al., Nature Review Molecular Biology, 3:906-918, 2002; and U.S. Pat. Nos. 5,932,474 and 5,998,204 to Tsein, incorporated herein by reference). Upon autoprocessing, either in the presence of cholesterol (or other sterol) or an excess of a thiol (other than a dithiol such as DTT), the approximately 30 kDa fluorescent-tagged fusion protein will be cleaved to an approximately 3 kDa fluorescent peptide, a change that can easily be measured by a decrease in fluorescence polarization.
Briefly, the concept of molecular movement and rotation is the basis of fluorescence polarization. Fluorescence polarization is defined by the following equation:
P=(V−H)/(V+2H)
where P equals polarization, V equals the vertical component of the emitted light, and H equals the horizontal component of the emitted light of a fluorophore when excited by vertical plane polarized light. As can be seen from the equation, P (“polarization unit”) is a dimensionless entity and is not dependent on the intensity of the emitted light or on the concentration of the fluorophore. This is the fundamental power of FP. Thus, upon autoprocessing of the Hedgehog peptide, the fluorescence polarization of the molecule, as determined by observing the molecule when exposed to polarized light, will be lower because the molecule will rotate faster when released from the larger Hedgehog peptide. The differences in observable fluorescent polarization can, therefore, be used to measure Hedgehog autoprocessing.
Thus, in the present instance, cleavage of the small fluorescent peptide fused to Hedgehog, upon autoprocessing, will result in a small, tagged, molecule, which rapidly tumbles out of the plane of polarization. In contrast, uncleaved Hedgehog, which rotates at a slower rate compared to the cleaved products, will emit a larger amount of energy in the same plane as the excitation energy. Thus, a determination of the activity and rate of the Hedgehog's ability to autoprocess is detectable by measuring the polarization signal with a decrease in the polarization signal (i.e., a decrease in fluorescence polarization) indicative of autoprocessing taking place.
Cell-based Assay Systems
In another aspect of the present invention, a method is contemplated for the screening of compounds that inhibit the autoprocessing of the Hedgehog protein in cell-based systems. In one aspect, the invention utilizes a test cell comprising a first expression vector encoding a functionally active, unprocessed Hedgehog protein and a second expression vector comprising a nucleotide sequence encoding a functional luciferase protein driven by 8 tandem repeats of Gli-binding sequences, or regulatory sequences of other Hedgehog-responsive genes (e.g., Gli, Glil, MIM/BEG4, patched, N-Myc, IGF2). The construction of expression vectors is well known in the art as is exemplified by the teachings in Sambrook, J., Fritsch, E. F., and Maniatis, T., Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, N.Y., Vol. 1, 2, 3 (1989), which is herein incorporated by reference. Suitable expression vectors for the present invention would include, for example, Bluescript™, or any of the Flexi200 Vector systems from Promega (Madison, Wis.).
Luciferase-based assays (both single and dual luciferase assay systems) and variations thereof, all of which are known to those skilled in the art, may be used in the present invention. In the luciferase-based assay, the detection of Hedgehog autoprocessing is via the assays the luciferase expression in response to Hedgehog activity. Luciferase luminescence may be detected, for example, with microscopy, scintillation counters or with the use of automated plate readers. Once Hedgehog is activated by autoprocessing, it is secreted and activates signaling in responsive cells by binding to the receptor Patched (Ptc), which relieves Ptc repression of Smoothened, a G-related protein, leading to the nuclear translocation of Gli transcriptional factors to activate the luciferase gene expression. In the present assay system, the same cells may be transfected with both Hedgehog and with the Hedgehog-responsive luciferase gene thereby allowing for a more efficient assay system.
Such cell-based assays would require the transfection of the expression vectors of the present invention into cells suitable for the assay. Suitable transfection methods include DEAE-dextran, calcium phosphate precipitation, Lipofectamine™ (Invitrogen, Carlsbad, Calif.), Profectin™ (Promega, Madison, Wis.) and other liposome methods, direct microinjection, electroporation and bolistic particle delivery, for example. Any cell line may be used for the present invention although the preferred cell line is the 3T3 mouse cell line.
After expression of the protein had been ascertained (e.g., by Western blotting, cell sorting, immunoflourescence or historical reference), a test compound is added to the transfected cells and the mixture is incubated under conditions and for a period of time appropriate for Hedgehog autoprocessing. Luminescence is then detected in the incubation mixture and compared to the luminescence of an otherwise identical incubation mixture that lacks the test compound, a decrease in luminescence of the incubation mixture being indicative of the presence of a test compound that inhibits autoprocessing of Hedgehog.
Another aspect of the present invention is a cell-based system for the screening of Hedgehog inhibiting compounds wherein a test cell has been transfected with an expression vector encoding a full-length, unprocessed Hedgehog protein and the test cell is contacted with a test compound under conditions and for a period of time appropriate for Hedgehog autoprocessing. The efficiency of Hedgehog processing is then detected by Western blotting. In brief, Western blotting is performed, for example, as follows: Proteins are separated using SDS-polyacrylamide gel electrophoresis (SDS-PAGE). They are then transferred to a (for example) nitrocellulose membrane. The nitrocellulose is blocked with a blocking agent (e.g., milk). The nitrocellulose membrane is incubated with, first, a primary antibody and, then, with a secondary antibody. The secondary antibody should be an antibody-enzyme conjugate. The labeled secondary antibody is then detected with, for example, x-ray film, phosphor imager or other detection device. The efficiency of Hedgehog autoprocessing is than compared to a otherwise identical mixture that lacks the test compound, wherein a lower efficiency of Hedgehog protein autoprocessing in the incubation mixture, relative to that determined using the otherwise identical incubation mixture, is indicative of the ability of the test compound of step to inhibit Hedgehog protein autoprocessing.
The assay described above may also be performed by using mass spectrometry to detect Hedgehog autoprocessing instead of, or in addition to, Western blotting. In brief, mass spectrometry is performed, for example, as follows: In MALDI-TOF mass spectrometry, a substance is deposited on a matrix and bombarded with a laser beam having sufficient energy to volatilize and impart a charge on the molecule. The charged molecules are accelerated in a vacuum through a magnetic field and are sorted on the basis of mass-to-charge ratio. Since the bulk of the ions produced in the mass spectrometer carry a unit charge, the value m/e is equivalent to the molecular weight of the fragment.
Another aspect of the present invention is a cell-based system for testing the specificity of Hedgehog inhibiting compounds comprising both a test cell and a control cell. The test cell comprises two expression vectors, the first expression vector comprising a nucleotide sequence encoding a functionally active, unprocessed Hedgehog protein; and the second expression vector comprising a nucleotide sequence encoding a functional luciferase protein driven by 8 tandem repeats of Gli-binding sequences. The control cell comprises two expression vectors, the first expression vector comprising a nucleotide sequence encoding a pre-processed Hedgehog protein the second expression vector comprising a nucleotide sequence encoding a functional luciferase protein driven by 8 tandem repeats of Gli-binding sequences. After expression of the protein had been ascertained (e.g., by Western blotting, cell sorting, immunofluorescence or historical reference), a test compound is added to the transfected test cells and control cell. A test incubation mixture is made by contacting the test cells with a test compound and a control incubation mixture is made by contacting the control cells with the same test compound, both incubation mixtures under conditions and for a period of time appropriate for Hedgehog autoprocessing. Luminescence is then detected in the test incubation mixture which has the expression vector encoding the unprocessed Hedgehog protein and compared to the luminescence of the control incubation mixture that has the expression vector encoding the preprocessed Hedgehog protein, a decrease in luminescence of the incubation mixture being indicative of the presence of a test compound that inhibits autoprocessing of Hedgehog.
Another cell-based assay for Hedgehog autoprocessing inhibitors is to analyze the gene expression of Hedgehog-responsive gene (also known as “gene transactivity” since the activation of one gene induces the activation of another gene) (e.g., Gli, Glil, MIM/BEG4, patched, N-Myc, IGF2). The test cell comprises the expression vector comprising a nucleotide sequence encoding a full-length Hedgehog protein. After expression of the protein had been ascertained (e.g., by Western blotting, cell sorting, immunofluorescence or historical reference), a test compound is added to the transfected cells and the mixture is incubated under conditions and for a period of time appropriate for Hedgehog autoprocessing. The expression of Hedgehog-responsive genes will be assayed by RT-PCR or Western blot and compared between incubation mixture and an otherwise identical incubation mixture that lacks the test compound, a decrease in Hedgehog-responsive gene expression being indicative of the presence of a test compound that inhibits autoprocessing of Hedgehog.
Intein Autoprocessing Inhibitors
Furthermore, another aspect of the present invention relates to the compounds identified as inhibitors of intein autoprocessing by the method described in Example 5 and illustrated in
Hedgehog Molecular Biology
Although the present invention is not limited to any particular theory, it is believed that the following discussion of Hedgehog peptides (Hh) will aid in the understanding of the invention. Hedgehog peptides are signaling ligands with key roles in stem cell specification and growth for tissue patterning in embryos and for stem cell renewal of adult tissues (reviewed by Ingham and McMahon, Genes Dev., 15:3059-3087, 2001; Beachy, et al., Nature, 432:324-331, 2004). Vertebrates have three Hh proteins, Sonic (Shh), Indian (Ihh) and Desert (Dhh) Hedgehogs. Hh activate signaling in responsive stem cells by binding to the receptor Patched (Ptc), which relieves Ptc repression of Smoothened, a G-related protein, leading to the nuclear translocation of Gli transcriptional factors to activate and repress specific target genes controlling stem cell growth, apoptosis and differentiation.
Hh have been identified as a tumor growth signal in a diversity of human cancers (Berman, et al., Nature, 425:846-851, 2003; Karhadkar, et al., Nature, 431:707-712, 2004). Tumors originating in esophagus, stomach, biliary tract, pancreas and prostate express high levels of Shh and Ihh, which stimulates the growth of stem cell-like tumor cells. Functional neutralization antibodies against Shh and lhh have been shown to block the growth of these tumors in vitro and in xenografts, establishing that these tumors are completely dependent on Hh ligand for their growth. As such, the present invention, as detailed above and in the Experimental section, relates to methods for the identification of small molecule drugs that inhibit the autoprocessing of Hh proteins, a process that is essential for the production and functional activation of Hh proteins and is unique to Hh among all vertebrate (e.g., human) known proteins. Inhibitors of Hh autoprocessing, therefore, should provide highly specific and effective drugs to block the growth of ligand-dependent (for example) Hh tumors.
The regulation of Hh autoprocessing has also been shown to be necessary for both weight regulation (Buhman, et al., J. Nutr., 134:2979-84, 2004) and stem cell differentiation (Ahn and Joyner, Nature 437:894-897, 2005). Thus, it is one aspect of the invention that the compounds identified as regulators (modulators) of Hh autoprocessing by the methods of the present invention are useful for both the regulation of weight and stem cell differentiation in animals.
Hedgehog are the only known self-spliced proteins in vertebrates. All three Hh proteins are expressed as a ˜45 kDa precursor, which undergoes autoprocessing to generate a secreted amino-terminal fragment Hh-N and a carboxyl-terminal fragment Hh-C. Hh-N, which is produced with a cholesterol moiety covalently attached to its carboxyl terminus, is the functional signaling ligand, and Hh-C is the auto-processing domain that is responsible for peptide bond cleavage and cholesterol transfer. Previous crystallographic study and in vitro splicing assays have demonstrated that Drosophila Hh-C not only shares highly similar sequence homology and crystal structure with the autoprocessed regions of inteins, Hh-C also utilizes highly similar reaction mechanisms as inteins in autoprocessing, which include an initial conversion of a peptide bond to a thioester linkage and subsequent attack by nucleophilic groups (Hall, et al., Cell, 91:85-97, 1997). However, in contrast to other autoprocessed proteins, Hh autoprocessing involves attack on the thioester group by cholesterol rather than by a Cys/Ser/Thr residue at the intein/C-extein junction (Hall, et al., Cell, 91:85-97, 1997; Paulus, Annu. Rev. Biochem., 69:447-496, 2000). All three human Hh contain conserved amino acid residues required for autoprocessing. Recently, as exemplified below, it has been shown human Shh-C also displays autoprocessing activities in vitro. Therefore, human Hh precursors undergo similar, but not identical, autoprocessing process to generate functional Hh-N signaling ligands.
For this project, a directed biochemical screen to identify small molecule inhibitors of Hh autoprocessing as drugs for treatment of, for example, ligand-dependent Hh cancers, was constructed.
The primary and 3-D structures of inteins and the Hh autoprocessed domain are highly similar and the first step of protein splicing and Hh autoprocessing are identical (Paulus, Annu Rev Biochem 69:447-496, 2000). These similarities imply that there should be a class of protein splicing inhibitors that also inhibit Hh autoprocessing. Since our high-throughput screen for protein splicing inhibitors has led to a large collection of reagents, it was worthwhile to test these for anti-Hh activity. Indeed, of the 41 protein splicing inhibitors tested in a cell-based assay for Hh autoprocessing, two were found also to inhibit Hh autoprocessing (see FIGS. 1 and 2). This constitutes proof-of-concept that small molecules can inhibit Hh autoprocessing and suggests a method for obtaining such inhibitors.
This cell-based assay (
This high-throughput assay, summarized in
GFPuv, the same form used for our intein constructs (Gangopadhyay, et al., Anal Chem 75:2456-2462, 2003), will be fused to the Hh autoprocessing domain, at residues between 224 through 232. The initial choice is insertion downstream of Gly228 or Gly232, which allows optimal autoprocessing, but other sites (i.e., downstream of Val224, Thr225, Ala226, Ala227, Gly228, Ile229, Thr230, or His231, for example) are also suitable as insertion sites for producing the desired results. These fusions are easily constructed by “gene SOEing (splicing by overlap extension)” (Horton, et al., Biotechniques 8:528-535, 1990). At 37° C. the fusion proteins are expressed as inclusion bodies and are denatured and then renatured at lower temperatures allowing a certain fraction of soluble fusion proteins will be produced. Persons skilled in the art will see that multiple variations of the denature-renature process will produce suitable fusion proteins. The fusion proteins of choice are non-fluorescent. These are monitored for gain of fluorescence upon Hh autoprocessing either in the presence of 20 mM DTT (dithiothreitol; which leads to the removal of the Hh domain) or cholesterol (which leads to the replacement of the Hh domain by cholesterol). The non-fluorescent constructs of choice are those that achieve fluorescence upon cholesterol-induced autoprocessing.
It is also worthwhile to focus on constructs that fluoresce even in the absence of autoprocessing as models for optimizing the autoprocessing conditions of GFP-Hh fusion proteins. Indeed, a fluorescent GFP-Hh fusion protein, in which the Hh autoprocessing domain is fused to GFP downstream of residue 252, can be the basis for an alternate, FRET-(fluorescence resonance energy transfer) based high throughput screening system, which is an embodiment of the present invention.
This type of assay is based on the observation that the NTA-Rhodamine derivative (NTA-I) in metal ion complex with an N-terminal His-tag almost completely quenches GFP fluorescence (Guignet, et al., Nature Biotechnology 22:440-444, 2004). Since the N- and C-termini of the Hh autoprocessing domain are in close proximity, a C-terminal His-tag of a GFP-Hh fusion protein will show a similar relationship to the GFP chromophore as an N-terminal GFP His-tag and NTA-I complexed to that C-terminal His-tag of the GFP-Hh fusion protein causes almost complete quenching of GFP fluorescence.
However, upon autocleavage, the covalent link between GFP and Hh is broken and fluorescence quenching is relieved. Thus, Hh autoprocessing leads to a substantial increase in GFP fluorescence (about 20-fold), thereby providing for a sensitive assay.
The assay (
This assay is considerably simpler in its structural requirements than the assay in Example 2 and does not involve GFP chromophore formation as a secondary reaction that can be inhibited by in a separate reaction unrelated to the autoprocessing step, but suffers from the disadvantage that the signal-to-noise ratio is less.
Similar assays as the one described above are based on other fluorescent proteins such as the GFP variants BFP, CFP and YFP, the red fluorescent protein from Discosoma and aequorin or based on luciferase proteins such as the firefly and Renilla luciferase. The principles of all these assays is basically the same: The fusion of the hedgehog autoprocessing domain to an amino acid residue near the C-termninus of the fluorescent protein or the luciferase in such a manner that the presence of the hedgehog protein blocks the fluorescence or luciferase function but that the replacement of the bulky hedgehog protein by cholesterol as a result of autoprocessing restores the ability of the protein to fluoresce or omit light. Since the atomic structures of all of these proteins are known, the selection of an appropriate fusion site for the hedgehog autoprocessing domain is a relatively straightforward matter based on the teachings contained herein.
An altogether different type of Hedgehog processing assay takes advantage of the fact that the N-terminal Hedgehog signaling domain can be replaced by a small polypeptide without impairing cholesterol-dependent autoprocessing (Porter, et al., Nature, 374:363-366, 1995). Replacing the N-terminal signaling domain with a peptide that selectively binds a fluorescent moiety allows a change in fluorescent polarization following autoprocessing by Hedgehog. There are many ways in which such N-terminal labeling can be effected. Perhaps the simplest approach involves the introduction of an affinity tag in the N-terminal domain to which a fluorescent probe can selectively bind with high affinity. One such labeling approach involves the use of Flash-tags (Griffin, et al., Science, 281:269-272 1998 and U.S. Pat. Nos. 5,932,474; 6,008,378; 6,054,271; 6,451,569 and 6,686,458, all to Tsien and Griffin, all of which are incorporated herein by reference). This approach can easily be implemented by expressing a fusion protein in which the N-terminal signaling domain of a functional Hedgehog protein is replaced by a small polypeptide that includes the sequence Cys-Cys-X-X-Cys-Cys [SEQ ID NO.: 1], where X is any amino acid. In a preferred embodiment, X-X are the amino acids Gly-Pro giving the sequence Cys-Cys-Gly-Pro-Cys-Cys [SEQ ID NO.: 2]. This peptide selectively binds a fluorescent biarsenical such as 4′,5′-bis(1,3,2-dithioarsolan-2-yl)fluorescein to yield intense fluorescence at 535 nm emission upon excitation at 480 nm, whereas the free biarsenical reagent has a 50,000-fold lower fluorescence. Upon autoprocessing, either in the presence of cholesterol or an excess of a thiol other than a dithiol such as DTT, the approximately 30 kDa fluorescent protein is cleaved to an approximately 3 kDa fluorescent peptide, a change that can easily be measured by fluorescence polarization. The assay for the inhibition of Hedgehog autoprocessing inhibitors therefore involves treating the Flash-tagged Hedgehog autoprocessing domain with the fluorescent biarsenical and comparing the fluorescence polarization signal in the absence and the presence of inhibitor candidates. Inhibition is indicated by a reduction in the fluorescence polarization signal. A similar assay for Hedgehog autoprocessing is implemented with fluorescent fusion proteins in which the N-terminal moiety is fluorescently labeled by other methods, a technique that could be practiced by one skilled in the art with the teachings contained herein.
The fact that both the primary and 3-D structures of inteins and the Hh autoprocessing domain are highly similar and that the first step of protein splicing and Hh autoprocessing are identical (Paulus, Annu. Rev. Biochem., 69:447-496, 2000) implies that there should be a class of protein splicing inhibitors that also inhibits Hh autoprocessing. This screening system makes use of GFP as a fluorescent indicator and is based on the observation that GFP, interrupted by a Hh autoprocessing site, can undergo splicing to produce a functional protein (Ozawa, et al., Anal Chem 72:5151-5157, 2000). The coding sequence for a RecA mini-intein derived from Mycobacterium tuberculosis was inserted adjacent to the codon of residue 129 in the gene of GFPuv to yield a hybrid plasmid. In E. coli JM109(DE3), the GFP-RecA mini-intein fusion protein is expressed as inclusion bodies and is solubilized in, for example, 8 M urea and purified as a non-fluorescent protein, which is stabilized to allow prolonged storage by blocking the thiol groups with 4,4′-dithiopyridine. Upon renaturation by dilution and dialysis and reduction of the disulfide bonds with TCEP, autoprocessing is observed, accompanied by the appearance of fluorescence, as shown in
The 384-well assay was used to screen about 80,000 reagents from diverse libraries, utilizing the facilities of the Institute of Chemistry and Cell Biology (ICCB) of Harvard University. The reagent library screening will involve, for example, the following libraries: Commercial Diversity Set (5,056 reagents), Diversity Oriented Synthesis (16,030 reagents), ICCB Bioactives (489 reagents), ChemDiv Combilab/Intemational (28,864 reagents), Bionet (6,168 reagents), Maybridge (16,807 reagents), Peakdale (3,168 reagents) and Mixed Commercial Plates (1,254 reagents).
Secondary screening of the hits to eliminate inhibitors of GFP chromophore formation and reagents with IC50 values greater than 40 μM are reagents were purchased and studied in more detail. Sixty compounds that inhibited the RecA intein in the low micromolar range were tested for inhibition of hedgehog autoprocessing in the cell-based assay systems described in Example 1 and two compounds were found (Reagents #3 and #4) that significantly inhibited hedgehog autoprocessing. This demonstrates that selecting intein inhibitors from compound libraries can serve as a method of enrichment of compounds that inhibit hedgehog autoprocessing.
This work was funded in part through grants from the National Institutes of Health (RO1 CA94089 and R43 AI53583). The government may have certain rights in this invention.
| Number | Date | Country | |
|---|---|---|---|
| 60742472 | Dec 2005 | US |
