Nucleic Acid Encoding Mammalian UBR1

Features of proteins that confer metabolic instability are called degradation signals, or degrons (Varshavsky, Cell 64: 13-15 (1991)). The essential component of one degradation signal, termed the N-degron, is a destabilizing N-terminal residue of a protein (Bachmair, et al., Science 234: 179-186 (1986)). A set of N-degrons containing different destabilizing residues is referred to as the N-end rule, which relates the in vivo half-life of a protein to the identity of its N-terminal residue (for review see Varshavsky, Cell 69: 725-735 (1992) and Varshavsky, Cold Spring Harbor Symp. Quant. Biol. 60: 461-478 (1996)). The N-end rule pathway has been found in all species examined, including the eubacterium Escherichia coli, the yeast Saccharomyces cerevisiae, and mammalian cells. The N-end rules of these organisms are similar but distinct.

As discussed in greater detail below, ongoing studies have revealed that the N-end rule pathway participates in a variety of complex functions in eukaryotic systems. Such studies indicate that the ability to intervene at the molecular level to inhibit or modulate the N-end rule pathway offers an important therapeutic avenue. Given the relatively complex enzymology of the pathway, the availability of components in quantity is essential to the development of therapeutic methods.

The subject invention relates, on one aspect, to a nucleic acid sequence encoding a recognition component of the N-end rule pathway. This nucleic acid sequence is characterized by the ability to specifically hybridize to the nucleic acid sequence of SEQ ID NO 1 under stringent hybridization conditions. Such conditions are defined below. In another aspect, the invention relates to a nucleic acid sequence encoding a recognition component of the N-end rule pathway which is characterized by the ability to specifically hybridize to the nucleic acid sequence of SEQ ID NO 2 under stringent hybridization conditions. Other embodiments relate to DNA expression vectors containing nucleic acid sequences of the type described above, as well as cells transformed with such expression vectors. The invention also relates to applications for the compositions described above.

FIG. 1 is a diagrammatic representation of the strategy employed for the cloning of mouse Ubr1 cDNA by intrapeptide PCR, interpeptide PCR, followed by conventional library screening.

The present invention is based on the isolation and cloning of nucleic acids encoding both human and murine forms of the recognition component of the N-end rule pathway, Ubr1. Ubr1, also referred to as E3.alpha., is a ubiquitin-protein ligase which has been linked, in particular, to stress-related deterioration (wasting) of muscle tissue.

More specifically, the rapid loss of muscle proteins associated with a variety of disease states has been shown to be primarily due to enhanced proteolysis via the ubiquitin-proteasome pathway (Mitch and Goldberg, N. Eng. J. Med. 335: 1897 (1996)). In rabbit skeletal muscle extracts, the N-end rule pathway is the dominant pathway for protein degradation--catalyzing the complete degradation of soluble proteins to amino acids (Solomon et al., J. Biol. Chem. 271: 26690 (1996)). In soluble extracts of rabbit muscle tissue, known inhibitors of the Ubr1 gene product were demonstrated to reduce the ATP-dependent degradation of soluble muscle proteins to amino acids by blocking their conjugation to .sup.125 I-ubiquitin (Solomon et al., Abstracts from 1997 FASEB Summer Meeting, Vermont (1997)).

Solomon et al. (Abstracts from 1997 FASEB Summer Meeting, Vermont (1997)) also report the identification of changes in the rate of protein ubiquitination associated with a number of pathological states characterized by muscle proteolysis. More specifically, in atrophying rat muscles, when overall protein degradation increases (e.g., in septic rats with sepsis induced by cecal puncture), hyperthyroid rats (treated with triiodothyronine) or in tumor-bearing rats (carrying Yoshida Ascites Hepatome for 3 to 5 days)), .sup.125 I-ubiquitin conjugation to soluble proteins increased 2-fold above the levels found in control muscles. Introduction of a known inhibitor was found to selectively suppress the increased .sup.125 I-ubiquitin conjugation toward levels in control muscles. In addition, ubiquitination of .sup.125 I-lysozyme (a model N-end rule substrate) was also determined to increase by 2-fold in extracts of atrophying muscles above levels in control extracts. Following hypophysectomy or thyroidectomy, protein degradation was shown to fall in isolated rat muscles. In extracts of such muscles, .sup.125 -I-ubiquitin conjugation to soluble proteins also falls by 50% in parallel with protein degradation. Addition of the lysine-alanine dipeptide suppressed .sup.125 I-ubiquitin conjugation to soluble proteins and eliminated most of the differences in this process between control and hypophysectomized or thyroidectomized rat muscle extracts. Ubiquitination of .sup.125 I-lysozyme was also reduced in these extracts indicating that the activities of components of the N-end rule pathway fall in muscles of thyroidectomized and hypophysectomized rats.

Observations such as those set forth above suggest that inhibitors of the N-end rule pathway will prove useful in connection with the treatment of various diseases characterized by the wasting of muscle tissue. While certain inhibitors of the N-end rule pathway are known (e.g., dipeptides resembling substrates of the N-end rule) the in vivo utility of such inhibitors is limited. Thus, one aspect the present invention relates to methods and compositions for inhibiting the N-end rule pathway-mediated muscle deterioration.

In addition to the treatment of various diseases characterized by the wasting of muscle tissue, the inhibition of the N-end rule pathway is indicated as a means of intervention for treatment of infection by intracellular pathogens such as Lysteria monocytogenes and Yersinia enterocolitica. Intracellular pathogens of this type occupy the cytoplasm of an infected cell, and propagate without killing the cell. An organism attempts to rid itself of infection by such pathogens through degradation of the pathogen's constituent proteins intracellularly, followed by the presentation of bacterial protein fragments to the immune system of the host via the major histocompatibility complex (MHC) class I-associated cytolytic T lymphocyte epitopes.

A recent report by Sijts et al. (J. Biol. Chem. 272: 19261 (1997)) highlights the involvement of the N-end rule pathway in connection with the MHC-associated presentation of Listeria monocytogenes epitopes. More specifically, the investigators focused on the degradation of p60, a Listeria-secreted murein hydrolase. Roughly 3% of degraded p60 gives rise to p60 217-225, a nonamer peptide that is bound by H-2K.sup.d MHC class I molecules. Mutagenesis of the N-terminal residue of p60 to replace the wild-type residue with amino acids known to be either stabilizing or destabilizing according to the N-end rule demonstrated clearly that p60 is a substrate of the N-end rule pathway. Valine substitution dramatically stabilized cytosolic p60 molecules, whereas aspartic acid substitution resulted in rapid degradation.

Cytosolic antigen degradation is fundamental to the generation of most MHC class I presented peptides. The fact that such degradation is, in the case of intracellular pathogens, mediated by the N-end rule pathway, strongly suggests intervention and manipulation of the pathway can be used to treat such infections. While the definitive demonstration of the importance of the N-end rule pathway in the life cycle of an invading bacterium such as L. monocytogenes remains to be demonstrated directly, the basis for the comments relating to therapy is the fact that intracytosolic parasites such as L. monocytogenes co-evolved with the N-end rule pathway. Therefore, drug-mediated perturbations of this pathway (either its inhibition or activation) are extremely likely to influence the course of infection by these bacteria.

Discussed above are examples of two pathological states, stress-induced deterioration of muscle tissue and infection by an intracellular pathogen, which are amenable to treatment by molecular intervention involving the N-end rule pathway. The N-end rule pathway has been well-studied and is the subject of several comprehensive review articles (see e.g., Varshavsky, Trends Biochem. Sci. 22: 383 (1997) and Varshavsky, Proc. Natl. Acad. Sci. USA 93: 12142 (1996)).

A brief review of the enzymology of the N-end rule pathway is warranted. Eukaryotic cells contain ubiquitin-specific enzymes that catalyze reactions whose product is either a single ubiquitin moiety or a multi-ubiquitin chain covalently linked to an acceptor protein. Ubiquitin is conjugated to other proteins through an amide bond, called the isopeptide bond, between the C-terminal (Gly-76) residue of ubiquitin and the .epsilon.-amino group of a lysine residue in an acceptor protein.

Ubiquitin is activated for conjugation to other proteins by a ubiquitin-activating enzyme (E1), which couples ATP hydrolysis to the formation of a high-energy thioester bond between Gly-76 of ubiquitin and a specific cysteine residue of E1. The E1-linked ubiquitin moiety is moved, in a transesterification reaction, from E1 to a cysteine residue of a ubiquitin-conjugating enzyme (E2), and from there to a lysine residue of an ultimate acceptor protein, yielding a ubiquitin-protein conjugate. This last step requires the participation of another component, called E3 or recognin, which selects a protein for ubiquitylation through an interaction with its degradation signal. Ubr1p, the E3 of the N-end rule pathway, is a bona fide enzyme that acts at the step between an E2 and an ultimate acceptor of ubiquitin. It catalyzes the movement, through transesterification, of the ubiquitin moiety from the cysteine residue of a relevant E2 enzyme to a cysteine residue of E3 itself.

Ultimately, ubiquitin-protein conjugates generated by the cascade of enzyme-catalyzed reactions described above, are specifically degraded by an approximately 2,000 kDa, ATP-dependent protease, termed the 26S proteasome. The 26S proteasome consists of a 20S core proteasome and a complex containing multiple ATPases at both ends of the 20S proteasome.

Ubr1 is one of several E3-type proteins of the ubiquitin system. Ubr1 is specific, in particular, for "destabilizing" residues exposed at the N-terminus of protein substrates. Since the degradation signals recognized by Ubr1 represent only a relatively small subset of the signals recognized by the entire ubiquitin system, inhibition of Ubr1 (and hence inhibition of the N-end rule pathway) would be a relatively mild, non-lethal therapeutic intervention, whereas the inhibition of the entire ubiquitin system would be lethal in mammalian cells, and therefore undesirable for selective therapy.

Inhibition of the Ubr1-encoded function in a cell can be effected in a variety ways. For example, at the nucleic acid level, an inhibitory molecule which specifically hybridizes to the Ubr1 mRNA can be contacted with the Ubr1 mRNA under physiological conditions, thereby inhibiting translation of the Ubr1 mRNA. Alternatively, inhibitors of the translated Ubr1 gene product can be introduced. With respect to inhibition at the translation level, knowledge of the Ubr1 cDNA sequence is essential. With respect to inhibition of the translated Ubr1 gene product, the availability of a cloned nucleic acid sequence encoding Ubr1 is a virtual necessity, for example, for the production of quantities of Ubr1 necessary for screening assays.

Thus, in one aspect, the present invention relates to a nucleic acid sequence encoding the Ubr1 recognition component of the N-end rule pathway. Disclosed herein are cDNA sequences corresponding to murine and human forms of Ubr1.

Given the nucleic acid sequences provided herein, one of skill in the art using no more than routine experimentation could isolate full length cDNAs from virtually any mammalian source. These cDNAs could be inserted into eukaryotic or prokaryotic expression vectors for the production of Ubr1 using recombinant DNA techniques. The scope of Applicants' invention is not limited to the specifically disclosed sequences, but rather encompass variations of such sequences which: 1) hybridize to the disclosed sequences under stringent hybridization conditions; and 2) encode a functional Ubr1 protein. With respect to the first criteria, an example of stringent hybridization conditions includes hybridization in which the disclosed sequences (or a portion thereof) are fixed to a solid support and a second DNA molecule to be tested for the ability to hybridize to the disclosed sequences is detectably labeled and suspended in a hybridization buffer consisting essentially of 50% formamide, 5.times.SSPE (1.times.SSPE is 0.15 mM NaCl, 1 mM Na-EDTA, 10 mM Na-phosphate (pH 7.0), 5.times.Denhardt's solution (0.1% polyvinylpyrrolidone, 0.1% Ficoll)). The hybridization buffer is contacted with the solid support at a temperature of about 45.degree. C. for a period of several hours. The hybridization solution is then removed, and non-specifically bound nucleic acid is removed by repeated washing with 1.times.SSC at increasing temperatures (up to 65.degree. C.). With respect to the second criteria, the functionality of an encoded product can be determined by a variety of assay techniques including, for example, any of the in vitro techniques discussed below.

Inhibition of Ubr1 mRNA translation may be effected by introducing oligonucleotides, or oligonucleotide combinations into a cell of interest. Alternatively, an antisense gene construct encoding a transcription product which is complementary with at least a portion of the Ubr1 mRNA may be employed. Due to the inherent difficulties associated with the introduction of oligonucleotides into the cells of an organisms, the use of an antisense gene construct is preferred in a therapeutic context. An antisense gene construct may be produced, for example, by inserting at least a portion of double-stranded Ubr1 cDNA into an expression vector in reverse orientation, as compared to the wild-type context, relative to a promoter. The expression vector is selected to be suitable for use in the target cell type. For use in connection with therapy in humans, eukaryotic virus-based vectors are preferred. Preferably, the nucleic acid encoded by the reverse gene construct is complementary with the Ubr1 mRNA in a region known to be critical with respect to expression. Typically, initial designs include translation start sites, splice junctions, and other sites critical with respect to expression.

As an alternative to the inhibition of translation of the Ubr1 mRNA, strategies designed to inhibit the activity of the translated Ubr1 gene product are employed. For example, contacting the expressed Ubr1 gene product with a molecule which specifically binds to the Ubr1 gene product and inhibits its activity is a preferred method of inhibition.

A variety of methods may be employed to identify a specific inhibitor of the Ubr1 gene product. For example, using a DNA expression vector containing expressible Ubr1 cDNA, the Ubr1 gene product may be overexpressed in vitro. The in vitro system is supplemented with the relevant ubiquitin conjugating enzyme (E2), the mammalian ubiquitin activating enzyme (E1), free ubiquitin and ATP in amounts sufficient to support ubiquitylation of substrates of the N-end rule pathway. Once established, this in vitro system is employed to screen small molecule libraries for the identification of inhibitors which exert their effect in an ATP and Ubr1-dependent manner. Such small molecule libraries are assembled from sources rich in complex small organic molecules. Bacterial and plant cell extracts are frequently used sources for the isolation of a large number of diverse organic molecules for such screening purposes.

Another method suitable for the identification of a specific inhibitor of the Ubr1 gene product involves the overexpression of the Ubr1 gene product in a mammalian cell line. Again, an expression vector is carefully selected to be compatible with the preferred mammalian cell line. In preferred embodiments, the cell line employed is a human cell line. The overexpression of the Ubr1 gene product in the mammalian cell line would increase the activity of the N-end rule pathway in the cell culture. In such a Ubr1-enhanced assay, inhibitors are detectable at a far greater level of sensitivity than would otherwise be provided by a wild-type cell line.

The availability of the Ubr1 cDNA sequence also provides the opportunity for the identification of specific inhibitors by a rational approach based on a complete understanding of the Ubr1 atomic architecture. The availability of the Ubr1 cDNA enables the production of the Ubr1 gene product by recombinant DNA techniques in milligram quantities. The recombinantly produced Ubri gene product can be crystallized, and its structure determined at atomic resolution by X-ray diffraction techniques. Using such techniques in combination with conventional molecular modelling, rationally designed candidate inhibitor molecules, designed to interact with specific subdomains of the Ubr1 gene product are designed and tested.

Materials and Methods

i) peptide preparation and sequencing

Rabbit Ubr1 protein (previously termed E3.alpha.) was purified from reticulocyte lysate using immobilized-protein column and elution by dipeptides as described (Reiss and Hershko, J. Biol. Chem 265: 3685-3690 (1990)). The sample containing approximately 20 .mu.g (.about.100 pmol) of the purified Ubr1 protein and chicken ovalbumin as a stabilizer for Ubr1 activity, in 50 mM N-ethyl morpholine, pH 8.5, 0.2 mM CaCl.sub.2 and 10% isopropanol, was digested with trypsin at an enzyme-substrate mass ratio of 1:20 at ambient temperature for 24 hrs. The digested sample was dried and resuspended in 6M guanidine-HCl, 0.1% trifluroacetic acid (TFA). The tryptic peptides were fractionated by reverse-phase HPLC on a C18 column and eluted with a gradient of acetonitrile in 0.1% TFA. The isolated peptides were sequenced by automated Edman degradation on a model 470A/900/120A gas phase sequencer/on-line analyzer (Applied Biosystems) using standard chemistry. Fourteen sequences of different peptides were obtained.

ii) Intrapeptide and interpeptide PCR.

Based on the rabbit Ubr1 peptide sequences, the intrapeptide PCR with degenerate oligonucleotides using rabbit genomic DNA (Clontech) as a template was applied to amplify the unique rabbit Ubr1 cDNA sequence corresponding to the internal region of the two PCRs (FIG. 1). Specifically, immediately prior to initiation of the PCR reaction, PCR primers were boiled for 1 min. and immediately cooled on ice. The PCR premixtures (100 .mu.l reaction volume) without AmpliTaq DNA polymerase were preincubated at 72.degree. C. (3 min.) in GeneAmp PCR System 9600 (Perkin-Elmer) and then AmpliTaq polymerase was added to each tube containing the premixture. After 94.degree. C. for 2 min., the first 4 cycles were done at 94.degree. C. for 1 min., 65.degree. C. for 10 min., and 72.degree. C. for 1 min. In the following 20 cycles, the factors in PCR were gradually decreased every four cycles; the denaturation time to 50, 30, 25, 20, and 15 sec; the annealing temperature to 62, 58, 55, 50, and 45.degree. C.; the annealing time to 5, 4, 3, 3, and 2 min.; and the extension time to 50, 30, 25, 25, and 25 sec. Then the final 20 cycles were done at 94.degree. C. for 15 sec, 42.degree. C. for 2 min., and 72.degree. C. for 25 sec. The amplified intrapeptide PCR products were analyzed by electrophoresis in a 4% low melting temperature agarose gel, cloned into PCR2.1 vector (Invitrogen, Calif.) and screened by digestion with restriction enzymes and subsequent sequencing.

Three PCRs gave the intrapeptide PCR products which contained the expected deduced amino acid sequences. The PCR primer pairs used for the positive clones were designated as follows: T122 (forward and reverse); T120 (forward and reverse); and T96 (forward and reverse). These designations correspond to the designations assigned to 3 of the 14 rabbit peptide sequences determined as described above.

Subsequently, the oligonucleotides corresponding the unique sequence of the intrapeptide PCR products using rabbit liver cDNA library (Clontech) as a template were applied to get Ubr1 cDNA fragment between the two peptides (interpeptide PCR) (FIG. 1). Among many combinations of primers, an oligonucleotide containing the unique sequence of T120 and another degenerate oligonucleotide corresponding to T134 (another of the 14 rabbit peptide sequences determined) produced a 392 bp fragment. Subsequently, the 392 bp-mouse Ubr1 cDNA fragment corresponding to the 392 bp-rabbit Ubr1 cDNA fragment was obtained from the mouse cDNA library (Clontech) using an oligonucleotide containing the unique intrapeptide PCR of mouse T120 and a degenerate oligonucleotide corresponding to rabbit T134. The 392 bp-mouse Ubr1 cDNA fragment was used as a probe for the screening of the Ubr1 cDNA clone from the mouse cDNA library as described below.

iii) cDNA Library Screening, DNA Sequencing, and 5'- and 3'-RACEs.

The .lambda.gt11 liver cDNA library (Clontech) was plated on Escherichia coli Y1090 (about 3.times.10.sup.4 plaque-forming units/150-mm plate, total 30 plates), and the plaques were lifted onto nylon membrane (Hybond-N, Amersham) and screened by hybridization with the 392 bp-mouse cDNA fragment (obtained from interpeptide PCR) that was labeled with ›.sup.32 P!dCTP. The putative positive clones were rescreened until they were plaque-purified. This initial screening using a .lambda.gt11 liver cDNA library gave two positive plaques. The purified DNA was digested with EcoRI and then analyzed on 1% agarose gels. Both of the selected positive clones turned out to contain identical 2.45 kb insert by the partial sequencing of the eluted PCR products produced with .lambda.DNA and 392 bp-probe specific primers. The cDNA inserts of the two clones were then subcloned into the pBluescript II SK.sup.+ plasmid vector, one of them (MR3) was sequenced on both strands. The complete sequencing of MR3 revealed nine regions in deduced amino acid sequence, including the regions of the 392 bp-mouse probe corresponding to T120, T100 and T134, which showed strong identity (62%-100%) to that of the rabbit peptide sequences. The overall identity and similarity of the sum of the nine regions (196 aa) to those of rabbit peptide sequences were 89% and 90%, respectively. This fact indicates that the cloned 2.45 kb insert encodes the mouse homolog of rabbit Ubr1. Furthermore, although the overall homology is relatively low (24% identity and 50% similarity), the deduced amino acid sequence (812 aa from N-terminus) of the cloned 2.45 kb insert showed considerable homology to that (aa 3-960) of S. cerevisiae UBR1, further supporting that the cloned 2.45 kb insert encodes the mouse homolog of S. cerevisiae UBR1.

The region of similarity began from the ATG codon located between nt 12 and 14 in 2.45 kb sequence, suggesting that this ATG codon may be the start codon in the mouse Ubr1 ORF. There was no stop codon, the poly(A) addition signal or poly(A) tail downstream of the ATG codon, suggesting that the cloned 2.45 kb insert encodes partial N-terminal portion of the mouse Ubr1 ORF considering the observed molecular weight of the purified rabbit Ubr1 (.about.180 kDa) on SDS-polyacrylamide gel electrophoresis (Reiss and Hershko, J. Biol. Chem 265: 3685-3690 (1990)), which is slightly smaller than S. cerevisiae UBR1 (225 kDa). Since there was no inframe stop codon upstream of the ATG codon, 5'-RACE PCR (Frohman et al., Proc. Natl. Acad. Sci. USA 85: 8998-9002 (1988)) was employed to amplify the upstream region, which contains the inframe stop codon, using an oligonucleotide primer specific for the 2.45 kb clone and a primer complementary to an in vitro-produced 5'oligo(dA) tract.

Specifically, 500 ng of poly(A).sup.30 RNA, isolated from mouse L-cells using Oligotex Direct mRNA kit (QIAGEN), was mixed with diethylpyrocarbonate-treated water to make the final volume of 20 .mu.l, incubated at 70.degree. C. for 5 min., and cooled on ice. To this sample were added 20 pmols of a primer corresponding to the antisense strand of the mouse Ubr1 ORF (nt 313-338 in SEQ ID NO:1), 2 .mu.l of 10.times. PCR buffer (Perkin-Elmer), 2 .mu.l of 0.1M dithiothreitol (DTT), 1 .mu.l of 10 mM dNTPs, 1 .mu.l of 25 mM MgCl.sub.2, and 1 .mu.l of bovine serum albumin (BSA; 2 mg/ml). The sample was incubated at 42.degree. C. for 2 min., followed by the addition of 1 .mu.l of Superscript II reverse transcriptase (Gibco-BRL) and an incubation at 42.degree. C. for another 40 min. The temperature was increased to 55.degree. C., followed by the addition of 1 .mu.l of RNAase H (Gibco-BRL; 2 units/.mu.l) and incubation for 20 min. The resulting cDNA products were purified with QIAquick PCR purification kit (QIAGEN), and were eluted with 50 .mu.l of water. To produce a cDNA-linked 5'oligo(dA) extension, 5 .mu.l of purified cDNA was diluted with 11.5 .mu.l of water, incubated at 70.degree. C. for 5 min., cooled on ice, then mixed with 1 .mu.l of 10.times. PCR buffer (to the final concentration of 0.5.times.), 1 .mu.l of 25 mM MgCl.sub.2, 0.5 .mu.l of BSA (2 mg/ml), 0.5 .mu.l of 10 mM dCTP, and 0.25 .mu.l of terminal transferase (Boehringer Mannheim). After an incubation at 37.degree. C. for 5 min., the enzyme was inactivated by heating the sample at 65.degree. C. for 10 min. The first round of RACE-PCR amplification was carried out in a 100 .mu.l sample containing 10 .mu.l of 10.times. PCR buffer, 2.5 mM MgCl.sub.2, 5 .mu.l of cDNA linked to oligo(dC), 20 pmols of a primer corresponding to the antisense strand of Ubr1 ORF (nt 306-332 in SEQ ID NO: 1), and 20 pmoles of oligo(dA) anchor primer. The sample was incubated at 94.degree. C. for 5 min., then at 57.degree. C. for 8 min., followed by the addition of AmpliTaq DNA polymerase (Perkin-Elmer) and incubation at 72.degree. C. for 8 min. to produce the complementary cDNA strand. Thereafter, 35 cycles of a 3-step PCR amplification were carried out. Each step involved consecutive incubations for 30 sec at 94.degree. C., 1 min. at 57.degree. C., and 2 min. at 72.degree. C. 2 .mu.l of the first-round PCR product was used for the second-round PCR that utilized, in the same total volume, 0.4 nmols of T-adapter primer (same as the T-anchor primer but lacking T.sub.17), and 0.4 nmols of a primer corresponding to the antisense strand of Ubr1 ORF (nt 271-293 in SEQ ID NO:1). The PCR-produced DNA fragments were inserted into pCR2.1 vector (Invitrogen). Two of the resulting clones gave the 114 bp-upstream sequence of the ATG codon, which contains two successive in-frame stop codons 48 bp and 93 bp upstream of the above ATG codon, suggesting that the putative Met start codon is the likely in vivo start codon of Ubr1 ORF.

Since several lines of evidences (see above and Results) indicated that the cloned 2.45 kb insert encodes partial N-terminal portion of the mouse Ubr1 ORF and since another hybridization screening of the same filters with the cloned 2.45 kb insert as a probe gave no additional positive clones, we employed 3'-RACE to amplify the downstream region of the 2.45 kb insert to use it as a probe for the next cDNA library screening. The 3'-RACE, which was done similarly to 5'-RACE (above) except omitting homopolymer tailing in 3'-RACE, gave 1.3 kb product (corresponding nt 1985-3313 in SEQ ID NO:1) which overlapped with 2.45 kb sequence by 465 bp. Based on the sequence of the 1.3 kb 3'-RACE product, one more 3'-RACE was done and gave 1.2 kb (corresponding nt 3039-3835 in mouse Ubr1 cDNA sequence), in which 797 bp Ubr1 cDNA sequence was fused with the 3'-UTR region of mouse glutathione S-transferase (GST) mRNA, which seems to be an artifact.

To get full length cDNA, .lambda.gt10 cDNA library (Clontech) from MEL-C19 cells was plated on Escherichia coli C600Hfl, hybridized with a labeled 998 bp-probe (nt 2470-3467 SEQ ID NO:1) synthesized by PCR on the basis of the sequence from 3'-RACE. By PCR analysis of the positive plaque lysates (or phage .lambda.DNA) followed by partial sequencing, the insert size and relative location of each of fourteen independent positive clones ranging in size from 0.6 to 4.6 kb were determined. Among them, five clones which overlap each other and cover the full length cDNA were subcloned into Bluescript II SK.sup.+ (MR16 with size of 3.0 kb, MR17 with size of 2.8 kb, MR19 with size of 2.2 kb, MR20 with size of 1.4 kb, and MR23 with size of 4.6 kb) and sequenced on both strands. Especially in the ORF region, at least two independent clones (from cDNA library screening, 5'- or 3'-RACE) were sequenced. Among them, MR16 contained the putative ATG start codon of the initial clone (2.45 kb, see above) preceded by 57 bp-mouse Ubr1 5'-UTR containing an inframe stop codon 48 bp upstream of the ATG codon. The 57 bp-5'-UTR region of MR16 was preceded by 360 bp mouse 18S ribosomal RNA sequence (EMBL accession number X00686), which is thought to be an artifact during library construction. MR19 contained an ORF showing considerable homology to yeast UBR1 followed by a stop codon preceding a poly(A) addition signal 41 bp downstream which was followed by poly(A.sub.21) tail 9 bp downstream, suggesting that MR19 contains Ubr1 C-terminal region and 3'-UTR. MR20 overlapped with 3'-region of MR16 and 5'-region of MR19, suggesting that MR16, MR20 and MR19 covers the full length ORF of mouse Ubr1 cDNA and also 3'- and 5'-UTR regions. These three clones (MR16, MR20 and MR19) were joined into a single contiguous fragment to make MR26 which contains 57 bp-5'UTR, 5271 bp Ubr1 ORF (1757 residues) and 58 bp-3'-UTR. Specifically, the 1.2 kb-EcoRI-XbaI fragment of MR20 was subcloned into pBluescript II SK.sup.+ to yield MR24. Subsequently, the 2.2 kb-XbaI fragment of MR19 was subcloned into MR24 to make MR25. Then the 3 kb-MscI-NotI fragment of MR25 was inserted into MR16 to make MR26 which contains the full length mouse Ubr1 ORF shown in SEQ ID NO:1.

iv) Cloning of Partial Human UBR1 cDNA fragment (1 kb) using RT-PCR

Poly(A).sup.+ RNA was isolated from human 293 cells using Oligotex Direct mRNA kit (QIAGEN). The first-strand cDNA was synthesized from 500 ng of Poly(A).sup.+ RNA using oligo(dT) priming and Superscript II reverse transcriptase (Gibco-BRL), followed by treatment of 2 units of RNAase H (Gibco-BRL) and purified with QIAquick PCR purification kit (QIAGEN). 30 ng of the synthesized cDNA was used for PCR using AmpliTaq DNA polymerase (Perkin-Elmer) and several different primers sets corresponding to mouse Ubr1 cDNA sequence. One of the reactions gave the 1 kb product which was subcloned into pCR2.1 vector (Invitrogen) and sequenced. The sequence of the partial human UBR1 cDNA fragment is shown in SEQ ID NO:2.

v) Cloning of Partial Human UBR1 Genomic DNA Fragments using Genomic PCR

The human genomic DNA was isolated from human 293 cells by conventional method and was used for PCR using Expand High Fidelity PCR System (Boehringer Mannheim) and exon specific primers. The PCR products were subcloned into pCR2.1 vector (Invitrogen) to give HR8 (insert size 6.3 kb), HR6-4 (insert size 5.8 kb), HR2-25 (insert size 3.6 kb), HR7-2 (insert size 5.4 kb). The four inserts described above were analyzed by partial DNA sequencing and were shown to cover .about.21 kb of the human UBR1 gene with overlapping of 100-150 bp. The exon/intron junctions were determined by partial sequencing.

vi) Northern and Southern Hybridizations

Mouse and human multiple tissue Northern blots with 2 .mu.g of poly(A).sup.+ RNA per lane (Clontech), isolated from various adult mouse or human tissues, were hybridized with the P.sup.32 -labeled probes (1 kb-human UBR1 cDNA fragment for human blot and 2 kb and 648 bp mouse Ubr1 CDNA fragments corresponding to nt 116-2124 and nt 4738-5385, respectively, in mouse Ubr1 cDNA sequence (SEQ ID NO: 1) eluted from the gel after PCR. Hybridization was carried out as suggested in the manufacturer's protocol. The intactness of the RNA samples on the blots were checked with the .beta.-actin probe provided with them. For Southern blot analysis genomic DNAs, isolated by conventional method from mouse L-cell and human 293 cell and digested with various restriction enzymes, were hybridized with 1228 bp or 1169 bp mouse Ubr1 cDNA probes for mouse blot (corresponding to nt 105-1332 and nt 610-1778, of SEQ ID NO: 1, respectively) or 1 kb human UBR1 CDNA probes for human blot under either high stringency (final washing; 0.1.times. SSC/0.1% SDS at 55.degree. C.) or low stringency conditions (final washing; 0.2.times. SSC/0.1% SDS at 42.degree. C.).

vii) Interspecific Mouse Backcross Mapping

The chromosomal position of mouse Ubr1 was determined using the interspecific backcross analysis. Interspecific backcross progeny were generated by mating (C57BL/6J.times.M. spretus) F.sub.1 females and C57BL/6J males as described (Copeland and Jenkins, Trends Genet. 7: 113-118 (1991)). A total of 205 N.sub.2 mice were used to map the Ubr1 locus (see text for details). DNA isolation, restriction enzyme digestion, agarose gel electrophoresis, Southern blot transfer and hybridization were performed essentially as described (Jenkins et al., J. Virol. 43: 26-36 (1982)). All blots were prepared with Hybond-N.sup.+ nylon membrane (Amersham). A 1169 bp-fragment of mouse cDNA corresponding to nt 610-1778, was labeled with ›.alpha..sup.32 P! dCTP using a random primed labeling kit (Stratagene); washing was done to a final stringency of 1.0.times.SSC, 0.1% SDS, 65.degree. C. Fragments of 5.6, 5.4, and 4.3 kb were detected in ScaI digested C57BL/6J DNA and a fragment of 15.0 kb was detected in ScaI digested M. spretus DNA. The presence or absence of the 15.0 kb ScaI M. spretus-specific fragment was followed in backcross mice. A description of the probes and RFLPs for the loci linked to Ubr1 including Thbs1 and B2m has been reported previously (Lawler et al., Genomics 11: 587-600 (1991)). One locus has not been reported previously for this interspecific backcross. The probe for erythrocyte protein band 4.2 (Epb4.2) was an .about.800 bp EcoRI fragment of human cDNA that detected a 7.8 kb SphI fragment in C57BL/6J DNA and a 14.0 kb SphI fragment in M. spretus DNA. Recombination distances were calculated using Map Manager, version 2.6.5. Gene order was determined by minimizing the number of recombination events required to explain the allele distribution patterns.

vii) Human chromosome mapping. Fluorescence in situ hybridization (FISH) was performed on human chromosomes prepared from synchronized cultures of lymphocytes isolated from cord blood (Heng et al, Proc. Natl. Acad. Sci. USA 89: 9509-9513 (1992)). Human chromosomes were probed with mixture of plasmids (HR8, HR6-4, HR2-25, and HR7-2) containing human UBR1 genomic DNA fragments (.about.21 kb) corresponding to partial human UBR1 cDNA fragment (1 kb, nt 2540-3532 in SEQ ID NO: 1). Probes were labeled with biotinylated DATP using the BRL BioNick labeling kit (15.degree. C., 1 hr), hybridized to the chromosome spreads, and detected with FITC-avidin. Signals were amplified by incubation with biotinylated goat anti-avidin followed by a second round of incubation with FITC-avidin. Chromosome banding patterns were obtained with the chromatin-binding fluorescent dye 4'-6-diamino-2-phenylindole (DAPI). Chromosomal localization of human UBR1 was made by superimposing photographs of the hybridization signals with photographs of the DAPI banding patterns.

Results

i) Peptide sequencing and PCR cloning

The fourteen peptide sequences of the purified rabbit Ubr1 protein showed no significant homology to any of the proteins deposited in the database, even to S. cerevisiae UBR1 protein, a counterpart of rabbit Ubr1 protein. A second independent purification of rabbit Ubr1 protein from reticulocyte lysates using similar method with that of the first approach (Reiss and Hershko, J. Biol. Chem 265: 3685-3690 (1990)) followed by determination of tryptic peptide sequences was done and gave three peptide sequences (PEP1, PEP2 and PEP3) which are identical to those of the first approach, supporting that the purified protein and the determined sequences are authentic.

On the basis of the peptide sequences, an initial attempt was made to obtain the unique sequence (not degenerate) of rabbit Ubr1 cDNA using degenerate oligonucleotide primers and intrapeptide PCR, which amplify the internal region of a peptide sequence (FIG. 1 and see Materials and Methods). Unique sequences were determined from intrapeptide PCR and corresponded to regions of sequence encoding several of the 14 rabbit peptide sequences (e.g., T122, T120 and PEP2). The intrapeptide PCR of T120 from mouse .lambda.gt11 liver cDNA gave a unique sequence encoding the same amino acids (with third codon redundancies) with that from rabbit genomic DNA of T120. The unique sequence of PEP1 was variable (four different types in 8 clones). The reason for this variable unique sequences is unclear. Subsequently, the oligonucleotide PCR primers, bearing the unique sequence from intrapeptide PCR, together with the original degenerate PCR primers were used for the amplification of rabbit Ubr1 cDNA fragment between the peptide sequences from rabbit liver cDNA library (interpeptide PCR) (FIG. 1). One of the interpeptide PCRs yielded a 392 bp-PCR product bearing T120 and T134 on both ends, and also internally bearing T100, indicating it to be authentic rabbit Ubr1 cDNA fragment corresponding to the purified protein. A 392 bp-mouse Ubr1 cDNA fragment corresponding to the 392 bp-rabbit UBR1 cDNA fragment was also obtained from the mouse cDNA library (see Materials and Methods). Sequencing of the 392 bp-mouse Ubr1 cDNA fragment revealed three regions corresponding to T120, T100 and T134 peptide sequences of 392 bp-rabbit Ubr1 cDNA fragment. The 392 bp-rabbit and mouse Ubr1 cDNA fragments shared 88% and 89% identity in nucleotide and protein sequence, respectively. They showed no significant homology to any sequence in data base including S. cerevisiae UBR1.

ii) Isolation of mouse Ubr1 cDNA

In the initial cDNA screening using the 392 bp-mouse cDNA fragment and subsequent screening using a probe based on the 3-RACEs gave several positive clones ranging in size from 0.6 to 4.6 kb. Among them, MR16 with size of 3.0 kb containing the ATG start codon preceded by 57 bp-5'-UTR with an in-frame stop codon 48 bp upstream of the ATG start codon, MR20 with size of 1.4 kb containing the middle region of the Ubr1 ORF, and MR19 with size of 2.2 kb containing the C-terminal region of the Ubr1 ORF and 58 bp-3'-UTR covered the full length cDNA ORF. The comparison of the partial sequence of MR23 (4.6 kb insert) with other clones (MR3, MR16, MR20, and MR19) and its sequence search revealed that it contains the C-terminal half of Ubr1 ORF (from aa 2703) flanked in 5'-region by the polyprotein sequence of Friend murine leukemia virus with the orientation reversed, which is believed to be a result from an artifact. Furthermore, the Ubr1 ORF of MR23 was followed by a long 3'-UTR (1010 bp), in which the poly(A) addition site in MR19 was bypassed, the significance of which is unclear.

The resulting mouse Ubr1 cDNA ORF was composed of 5271 bp encoding a 1757-residues (200 kDa) protein, which is largely similar to that (225 kDa) of S. cerevisiae UBR1 and the observed molecular weight (180 kDa) of rabbit Ubr1, purified from reticulocyte lysate, on SDS-PAGE (Reiss and Hershko, J. Biol. Chem 265: 3685-3690 (1990)). The upstream sequence of the putative (first) ATG start codon, preceded by two in-frame stop codons 48 bp and 93 bp upstream, was largely in an agreement with Kozak's rules (Kozak, M., J. Biol. Chem. 266: 19867-19870 (1991)) in that A in position -3 and G in position +4. Immediately downstream of the first ATG codon there are two more ATG codons. Both the second and third ATG codons (the 6th and 12th amino acids in the ORF) have a purine (G and A, respectively) in -3 position and G in +4 position, indication them to be potential alternative start codons. One prominent feature of the N-terminus of the ORF is that among the first 13 residues in ORF 7 are charged (6 negative and 1 positive) amino acids. The meaning of the second and third ATG codons and the highly charged (negative) N-terminal 13 residues are unclear.

iii) Deduced amino acid sequence of mouse Ubr1

Although mouse Ubr1 protein sequence showed relatively low overall homology to S. cerevisiae UBR1 (22% identity and 48% similarity) (for S. cerevisiae UBR1 sequence see GenBank Accession No. P19812), certain subdomains showed significant homology. More specifically, six specific regions of homology were identified indicating functional relationship of the two proteins. These regions were arbitrarily designated Regions I-VI with designations assigned in order based on location from N-terminus to C-terminus (i.e., Region I is the most C-terminal of the VI regions of homology). Furthermore, comparison of mouse Ubr1 sequence with those available in sequence databases using BLAST programs (Altschul et al., J. Mol. Biol. 215: 403-410 (1990)) through the National Center for Biotechnology Information revealed several proteins showing significant homology; a 1927 aa-Caenorhabditis elegans ORF (GenBank Accession number U88308) (32% identity and 53% similarity; termed C. elegans Ubr1.sup.2), a 1872 aa-S. cerevisiae ORF (GenBank Accession number Z73196) (21% identity and 47% similarity; termed S. cerevisiae UBR2), a 2168aa-C. elegans ORF (GenBank Accession number U40029) (21% identity and 45% similarity; termed C. elegans Ubr2) and a 794 aa-Arabidopsis thaliana CER3 (eceriferum 3; 26% identity and 49% similarity). Besides these proteins, a 147 aa-partial ORF of Candida albicans corresponding to the region near N-terminus of mouse Ubr1 (http://alces.med.umn.edu/bin/genelist?LUBR1; termed C. albicans UBR1), and a 272 aa-partial ORF of Schizosaccharomyces pombe corresponding to the region near C-terminus of mouse Ubr1 (GenBank Accession number M26699; termed S. pombe UBR2) also showed significant homologies to the members of UBR1 family. Although the overall homology in these proteins are relatively low (17-32% identity), alignment of their protein sequences revealed several distinctive regions, suggesting that mouse Ubr1 and yeast UBR1, together with their related proteins of previously unknown function, belong to a distinct UBR1 family. The recognition of this fact was made possible exclusively by the isolation and identification of the mouse and human Ubr1 sequences as disclosed herein.

One of the prominent regions of UBR1 proteins is a 66 aa-region (Region I) near N-terminus (in mouse Ubr1), which shows the highest homology among all the regions (61% identity and 75% similarity between mouse Ubr1 and C. elegans Ubr1). This region contains a distinctive Cys/His rich domain which does not fit to any other known Cys/His rich motifs. This Cys/His domain is conserved in all the UBR1 family members including mouse Ubr1 , C. elegans Ubr1 , S. cerevisiae UBR1, S. cerevisiae UBR2, C. elegans Ubr2 and a partial N-terminal ORF of Candida albicans ORF (http://alces.med.umn.edu/bin/genelist?LUBR1) (termed C. albicans UBR1), except Arabidopsis thaliana CER3 which contains only Region V and VI. Although this Cys/His structure is likely to be a zinc finger, the number and spacing of Cys and His residues in this structure did not fit to any other known zinc finger. Region V also contains a distinctive Cys/His domain which is conserved in all the UBR1 family members. By comparison of this Cys/His domain with the already known Cys/His structures, the Cys/His domain in Region V was turned out to belong to a RING-H2 finger, a subfamily RING fingers (Borden and Freemont, Curr Opin Struct Biol 6: 395-401 (1996)). Several known examples of RING-H2 finger-containing proteins are PSMP, CELG, FAR1, PEP3 and PEP5 (for references of each sequence, see Freemont, P. S., Ann. NY Acad. Sci. 684: 174-192 (1993)). One distinctive feature of the RING-H2 of UBR1 family is that the length of loop1 (53 aa-85 aa) is longer than that (12aa-35aa) of those known RING-H2 finger proteins. Other extensive homologies between UBR1 proteins are observed in the 115-aa Region VI (in mouse Ubr1) near C-terminus (24%-50% identity and 46-70% similarity to Region VI of mouse Ubr1). Region VI of C. elegans Ubr1 showed highest homology (50% identity and 70% similarity) to that of mouse Ubr1. However, the region VI of S. cerevisiae UBR1, the homolog of mouse Ubr1, showed the lowest homology (24% identity and 46% similarity) among UBR1 family members including S. cerevisiae UBR2, C. elegans Ubr2, A. thaliana CER3 and S. pombe UBR2. Furthermore, while region VI of all the other related proteins was located 4-14aa from C-terminus of each protein, S. cerevisiae UBR1 had an additional 132 and 159 residue-tail which is highly rich in (mainly negative) charged residues (36% and 33%). The significance of the tails is unclear. Region IV also shows high homology in all the UBR1 family members except C. elegans UBR2. No protein showed considerable homology to this region when searched using BLAST.

iv) Cloning of Partial Human CDNA and Genomic DNA

To obtain probes for chromosome mapping of human UBR1, a partial human UBR1 CDNA (1 kb), corresponding nt 2218-3227 of mouse Ubr1 cDNA sequence, was cloned by RT-PCR using Poly(A).sup.+ RNA isolated from human 293 cells. The nucleotide and deduced amino acid sequences shared 90% and 93% identities, respectively, with mouse Ubr1 cDNA sequence. Partial human UBR1 genomic DNA fragments (HR8, HR6-4, HR2-25 and HR7-2 with insert sizes of 6.3 kb, 5.8 kb, 3.6 kb and 5.4 kb, respectively, with overlapping of 100-150 bp), corresponding to 1 kb cDNA and .about.21 kb genomic DNA, were cloned by genomic PCR using genomic DNA from human 293 cells as a template and the primers based on the human UBR1 cDNA sequence. Partial DNA sequencing of the cloned genomic DNA fragments showed that the .about.21 kb genomic DNA region was composed of 11 exons ranging in length from 49 bp to 155 bp. All of the exon/intron junctions contained the consensus GT and AG dinucleotides characteristic of mammalian nuclear pre-mRNA splice sites (Shapiro and Senapathy, Nucleic Acids Res. 15: 7155-7174 (1987)).

v) Northern and Southern blot hybridizations of mouse Ubr1 and human UBR1

The expression of mouse Ubr1 and human UBR1 was tested by Northern blot analysis using 2 kb (N-terminal region) or 640 bp (C-terminal region) mouse Ubr1 cDNA fragments for mouse blot and 1 kb-human UBR1 cDNA fragment for human blot. A poly A.sup.+ transcript of .about.8.0 kb was ubiquitously detected in different mouse tissues using either of the probes (N- or C-terminal region), with relatively high level in skeletal muscle, heart and brain and with lowest level in kidney. The testis-derived Ubr1 mRNA existed as two species: the minor one comigrated with the .about.8.0 kb Ubr1 mRNA of the other tissues, while the major one had the apparent size of .about.6 kb, which is similar to Northern blot of testis-derived Ntan1 mRNA in which the minor one comigrated with the .about.1.4 kb Ntan1 mRNA of the other tissues, while the major one had the apparent size of .about.1.1 kb (Grigoryev et al., J. Biol. Chem. 271: 28521-28532 (1996)). It is unclear whether the testis specific Ubr1 and Ntan1 Northern patterns were the result of RNA degradation during isolation, specific cleavage of RNA, or two distinct primary transcripts (from different poly(A) addition site or alternative splicing), like E2.sub.14K mRNA. The mouse E2.sub.14K, the mouse homologs of S. cerevisiae UBC2 which is a component of the yeast N-end rule, also shows the highest mRNA expression level in skeletal (Grigoryev et al., J. Biol. Chem. 271: 28521-28532 (1996)). The upstream region of the rabbit E2.sub.14K ORF contains several putative binding sites for MyoD, a muscle-specific transcription factor (Weintraub et al., Genes Devel. 5: 1377-1386 (1991)). The human UBR1 mRNA showed similar Northern blot pattern with that of mouse Ubr1.

Southern blotting analysis of mouse or human genomic DNA has revealed rather simple band patterns under either high (final washing; 0.1.times. SSC/0.1% SDS at 55.degree. C.) or low stringency conditions (final washing; 0.2.times. SSC/0.1% SDS at 42.degree. C.), suggesting the presence of a single copy of Ubr1 gene in genome and the absence of genes whose structures are closely related to Ubr1 at nucleotide level. However, we cannot exclude the presence of mammalian E3(s) closely related with Ubr1 only at amino acid level. Indeed, several lines of evidences indicate the presence of E3.beta. in rabbit reticulocyte lysates which is believed to be another mammalian ubiquitin-protein ligase recognizing small uncharged N-termini (Ala, Ser, Thr: type III N-terminal destabilizing residues) of the N-end rule (Gonda et al., J. Biol. Chem. 264: 16700-16712 (1989)). Although they have different substrate specificities, rabbit Ubr1 and E3.beta. share several properties (Hershko and Ciechanover, Annu. Rev. Biochem. 61: 761-807 (1992)). Therefore, it is likely that the sequences of these two proteins are similar.

vi) Interspecific Mouse Backcross Mapping

The mouse chromosomal location of Ubr1 was determined by interspecific backcross analysis using progeny derived from matings of ›(C57BL/6J.times.Mus spretus) F.sub.1 X C57BL/6J! mice. This interspecific backcross mapping panel has been typed for over 2400 loci that are well distributed among all the autosomes as well as the X chromosome (Copeland and Jenkins, Trends Genet. 7: 113-118 (1991)). C57BL/6J and M. spretus DNAs were digested with several enzymes and analyzed by Southern blot hybridization for informative restriction fragment length polymorphisms (RFLPs) using a mouse cDNA Ubr1 probe. The 15.0 kb ScaI M. spretus RFLP (see Materials and Methods) was used to follow the segregation of the Ubr1 locus in backcross mice. The mapping results indicated that Ubr1 is located in the central region of mouse chromosome 2 linked to Thbs1, Epb4.2, and B2m. Although 66 mice were analyzed for every marker and are shown in the segregation analysis, up to 133 mice were typed for some pairs of markers. Each locus was analyzed in pairwise combinations for recombination frequencies using the additional data. The ratios of the total number of mice exhibiting recombinant chromosomes to the total number of mice analyzed for each pair of loci and the most likely gene order are: centromere-Thbs1-4/133-Ubr1-0/113-Epb4.2-1/122-B2m. The recombination frequencies (expressed as genetic distances in centiMorgans (cM).+-.the standard error) are -Thbs1-3.0.+-.1.5-›Ubr1, Epb4.2!-0.8.+-.0.8-B2m. No recombinants were detected between Ubri and Epb4.2 in 113 animals typed in common suggesting that the two loci are within 2.7 cM of each other (upper 95% confidence limit).

The interspecific map of chromosome 2 was compared with a composite mouse linkage map that reports the map location of many uncloned mouse mutations (provided from Mouse Genome Database, a computerized database maintained at The Jackson Laboratory, Bar Harbor, Me.). Ubr1 mapped in a region of the composite map that lacks mouse mutations with a phenotype that might be expected for an alteration in this locus. The central region of mouse chromosome 2 shares a region of homology with human chromosome 15q. The placement of Ubr1 in this interval in mouse suggests that human homolog will map to 15q, as well.

vii) Chromosomal localization of the human UBR1 locus

The chromosome localization of mouse Ubr1 was independently confirmed and refined by chromosome mapping of human UBR1 by FISH using human UBR1 genomic clones (HR8, HR6-4, HR2-25 and HR7-2) as the hybridization probes. Under the conditions described in Materials and Methods, the hybridization efficiency was approximately 91% for these probes (among 100 checked mitotic figures, 91 of them showed signals on one pair of the chromosomes). Since the DAPI banding was used to identify the specific chromosome, the assignment between signal from probes and the long arm of chromosome 15 was obtained. The detailed position was further determined based on the summary from 10 photos. There was no additional locus picked by FISH detection under the condition used, therefore, UBR1 is located at human chromosome 15q15-15q21.1, which is in a good agreement with the result of mouse chromosome mapping of Ubr1.