Method for Targeting Degradation of Intracellular Proteins

1. Field of Invention

The current invention concerns a method for targeted degradation of intracellular proteins in situ. In particular, the invention concerns the method for degradation of intracellular proteins by inducing in cells the production of a dual-function protein containing a domain that directs a selective degradation of targeted proteins to which it is attached as well as a domain that acts as a linker between the dual-function protein and the target protein. The protein degradation directing domain is a subregion within 97 amino acids which corresponds to the N-terminus of protein antizyme (NAZ). The invention further concerns a method for inducing in cells the production of NAZ, linker and the dual-function protein in a form directed to destruction of specific cellular target proteins.

2. Background and the Related Art

The body produces a large number of proteins which direct and regulate numerous physiological functions. The rapid turnover of these proteins is important in regulating cell growth and metabolism. These proteins are, under normal circumstances, strictly regulated by physiological feedback, which assures that when the level of certain protein in the body reaches desired level, a regulatory mechanism normally present in the body takes over the control of that particular protein production and either decreases it to the level which is sufficient for a normal physiological function of that protein or temporarily terminates its production.

In many instances, however, due to some pathological or pathophysiological conditions, such endogenous control is disturbed and overproduction or underproduction of certain proteins occurs.

Underproduction of these proteins is typically corrected by providing substitute proteins or drugs which would stimulate the protein production.

In case of overproduction of certain proteins, such as, for example, hormones, the problem is much more serious and typically can be corrected only by inhibiting the production of these proteins by drugs or synthetic enzymatic inhibitors. Very often these drugs are not overly selective for the particular protein but will also affect, that is inhibit, the production of other proteins. Additionally, when administered systemically, these inhibitors have often undesirable secondary systemic symptoms.

It would be, therefore, advantageous to have available a method for selective destruction of targeted proteins in situ.

Proteins that turn over rapidly are of special interest because they can quickly adjust their abundance in response to changes in synthesis or degradation. Some labile proteins have been found to interact with a second protein that promotes their degradation. Examples of these proteins are lysosome-degraded proteins which interact with 70-kD heat shock protein and tumor suppressor p53 which interacts with viral oncoproteins (Science, 346:382-385 (1989) and Cell, 67:547-556 (1991)).

Several methods are presently under consideration as modifiers of protein degradation. The first approach uses inhibitors of the proteasome or other viral or cellular proteases. These methods lack substrate specificity and in general modify the action of the proteolytic machinery rather than its action on specific substrates.

The second approach utilizes ubiquitination. Ubiquitination is a natural process introducing a modification consisting of the attachment of multiple copies of the protein ubiquitin that triggers proteolysis of many short-lived proteins (Nature, 357:375-379 (1992); Microbiol. Rev., 56:592-621 (1992); Ann. Rev. Biochem., 61:761-807 (1992); J. Biol. Chem., 268:6065-6068 (1993).

Attempts were made previously to provide a method for specific proteolytic removal of proteins via the ubiquitin-dependent proteolytic pathway. This method is, however, complicated, laborious and not suitable for clinical use. Degradation via the ubiquitin pathway requires the prior attachment of multiple ubiquitins to the target proteins. This attachment is accomplished by a family of enzymes designated ubiquitin-conjugating enzymes (E2), some of which use domains near their C-termini for target recognition. PNAS, 92:9117-9121 (1995) describes a method for modified ubiquitination by generating chimeric ubiquitin-conjugating enzymes that recognize and ubiquitinate their binding partners with high specificity in vitro. This method comprises forming a fusion containing a linker and one of the enzymes that participates in ubiquitination, thus ubiquitinating specific proteins and targeting them for destruction.

The method described in PNAS, above, however, depends on fusion maintaining enzymatic activity of E2 in general and particularly for the intended target. Multiple ubiquitins must be transferred to a substrate to cause its degradation and additional enzymes collectively called E3 may be needed for substrate ubiquitination. So far, this method has only been tested in vitro and its effectiveness and substrate specificity in vivo have not been demonstrated.

The selective targeted protein degradation method alternative to ubiquitination has not been previously described but would be very useful in controlling excess production and/or inhibition of production of certain proteins or their end-product compounds by way of selective degradation of targeted proteins. Such degradation would be useful in controlling diseases caused by the overproduction of certain proteins or compounds. Examples of such diseases are hormonal disorders, carcinogenic growth, viral, bacterial or parasitic infections, etc.

It would therefore, be advantageous to provide a method for substrate specific and targeted degradation of proteins in situ which method would be effective in vivo, in which the need for ubiquitination would be eliminated and the whole degradation process would be simplified by attaching a protein element that provides a functional alternative to ubiquitination to target proteins.

Physiological feedback control over protein production or termination of such production has been known and studied, for example, on polyamine biosynthetic enzymes. Polyamines presence in the body is essential for cell growth and proliferation. The key enzyme in the biosynthesis of the polyamines, ornithine decarboxylase (ODC), is highly regulated and mammalian ODC is one of the group of the most short lived of proteins. Its turnover can proceed along two different pathways: constitutive pathway and polyamine dependent pathway. The activity of this enzyme is increased by stimulating cell growth in cancer but is decreased by the presence of an excess of polyamines. Polyamine-promoted degradation of ODC seems to need the involvement of another protein.

Feedback regulation of the enzyme by polyamines occurs via induction of a certain protein called antizyme. (Biochem. Biophys. Acta, 428:456-465 (1976); PNAS, 73:1858-1862 (1976); J. Biol. Chem., 259:10036-10040 (1984); J. Biochem., 108:365-371 (1990) and Gene, 113:191-197 (1992)).

In their previous research, inventors discovered that regulated degradation of ODC requires interaction of two ODC regions with a specific polyamine-inducible protein antizyme. These regions are located near C-terminus and near N-terminus of ODC. This specific polyamine-inducible antizyme was found to bind to ODC, inhibit its activity and accelerate its degradation. The antizyme-ODC binding seems to be required for regulated degradation of ODC (Mol. Cell. Biol., 12:3556-3562 (1992)). The interaction of the ODC antizyme with a ODC binding element located near the N-terminus of ODC was found essential but not sufficient for regulation of the enzyme by polyamines. The second ODC element present at the C-terminus of ODC is required for the degradation process. Thus, both the C-terminal degradation region and a binding element near the N-terminus of ODC are required for the degradation process of ODC (Mol. Cell. Biol., 13:2377-2383 (1993)).

In the studies described in (Mol. Cell. Biol., 14:87-92 (1994) selective degradation of ODC was found to be mediated by antizyme which binds to a region near the N-terminus of ODC. This interaction induces a conformational change in ODC that exposes its C-terminus and inactivates the ODC. Additionally, it was found that while the C-terminal half of antizyme alone can bind to and inactivate ODC and alter its conformation, it cannot direct degradation of the enzyme either in vitro or in vivo. A portion of the N-terminal half of antizyme must be present to promote such degradation.

While the above findings concerning the degradation of ODC are scientifically interesting, they are without practical utility. The ODC is closely regulated and as a short life enzyme, it is quickly disposed of. Other than scientific importance and the invention background, the findings described above do not provide a method or means for selective and targeted degradation of proteins intracellularly which would have general utility.

It is, therefore, a primary objective of this invention to provide a method and means for selective and targeted degradation of proteins in situ by administering to a subject in need thereof, or by inducing in the subject in need of such treatment, the endogenous production of an effective amount of dual-function protein comprising a linker domain for specific binding to a target protein as well as a N-terminus domain of antizyme capable, when complexed with the target protein via the linker, of destabilizing it and changing it in such a way that it becomes labile and subject to destruction.

All patents, patent application and publication cited herein are hereby incorporated by reference.

One aspect of the current invention is a method for selective and targeted degradation of proteins.

Another aspect of the current invention is a method for selective and targeted degradation of endogenously present proteins by providing protein compounds comprising two distinct functional domains which compounds are able to form a complex with the target protein and cause its degradation.

Another aspect of the current invention is a method and means for selective and targeted degradation of proteins in situ by administering to a subject in need thereof or by inducing in the subject in need of such treatment the endogenous production of an effective amount of a dual-function protein comprising a linker domain capable of complexing with the target protein, and a N-terminus of antizyme capable of destabilizing the target protein in such a way that it becomes labile and subject to degradation.

Still another aspect of the current invention is a method for degradation of intracellular proteins by inducing in cells the production of a subregion of about 97 amino acids which corresponds to the N-terminus domain of protein antizyme (NAZ) that directs the degradation of proteins to which it attaches, wherein said NAZ domain is able to be attached to the target protein and direct its degradation.

Still another aspect of the current invention is a method for inducing in cells the production of NAZ domain in a form directed to destruction of specific cellular target proteins via its association with a target specific linker.

FIG. 1 illustrates in vitro degradation of mouse ODC (MODC) induced by fusion of ODC to N-terminus of antizyme (NAZ).

FIG. 2 illustrates TbODC degradation induced by the N-terminus of antizyme (NAZ) and C-terminus of mouse ODC.

FIG. 3 illustrates effect of ATP-.gamma.-S on degradation of NAZ-Tb376M and NAZ-Tb422M.

FIG. 4 illustrates NAZ-induced degradation of TbODC-cyclin B 13-90 (with ubiquitination site) and TbODC-cyclin B 13-59 (without ubiquitination site).

FIG. 5 illustrates NAZ-induced degradation of cyclin B 13-90 (with ubiquitination site) and cyclin B 13-59 (without ubiquitination site).

FIG. 6 shows structure of NAZ-p53 and deletions of p53 moiety.

FIG. 7 illustrates NAZ-induced degradation of p53.

FIG. 8 shows effect of ATP-.gamma.-S on p53 degradation.

FIG. 9 illustrates effect of monoclonal antibody PAb246 specific for p53 amino acids 88-109 on E6-mediated degradation of mouse p53.

FIG. 10 shows effect of the p53 degradation domain (amino acids 100-150) on in vivo degradation of trypanosome ODC.

FIG. 11 shows effect of the p53 degradation domain (amino acids 100-150) on polyamine-dependent degradation of chimeric mouse/trypanosome ODC (M314T).

As used herein:

"Dual-function protein" means a protein containing a domain controlling binding and a domain directing protein destabilization and degradation.

"Antizyme" or "AZ" means a reversible, tightly binding polyamine-induced protein inhibitor of ornithine decarboxylase activity.

"Linker" or "link" means a protein or peptide capable of recognizing, identifying and binding to a target protein or peptide. It also means a binding domain present within a dual-function protein typically near its C-terminus, wherein said protein additionally contains the second NAZ domain located near the protein N-terminus. A primary function of the linker is to bind to and in this way to deliver NAZ domain to a target protein.

"ODC" means ornithine decarboxylase.

"NAZ" means N-terminus of antizyme or other dual-function protein which represents a domain, i.e. defined portion of AZ, approximately 97 amino acids in size, which domain carries a critical signal that directs degradation of ODC but which can also act as a functional module which when attached to heterologous proteins other than ODC, causes those proteins to be rapidly degraded.

"CAZ" means C-terminal half of antizyme.

"GST" means glutathione S-transferase.

"Target protein" means any protein containing a degradation region required for cooperation with the NAZ domain. The target protein includes all proteins which are per se already more or less unstable, in that they contain a degradation domain. This includes substantially all proteins important for cellular regulation. These proteins are, under appropriate cellular conditions, subject to rapid intracellular degradation by way of feedback. Consequently, they must contain a degradation domain. Additionally, target proteins are also other stable proteins where the degradation region could be induced or could be incorporated into NAZ-linker protein.

"Covalent association" means association of two protein regions present within one protein chain.

"Non-covalent association" means association of two protein regions present within two separate protein chains.

This invention concerns a method for a substrate specific targeted degradation of intracellular proteins in situ. The method achieves the substrate specific targeted degradation of intracellular proteins by inducing intracellularly the production of a dual-function protein, such as antizyme, containing two domains wherein one domain is a subregion of about 97 amino acids defining the N-terminus of protein antizyme (NAZ) and a second domain is a linker. The NAZ domain directs the degradation of proteins to which it is attached by way of the linker. The second domain, the linker, binds to the target protein.

I. A Method for Degradation of Target Cellular Proteins

This invention is based on finding that a naturally occurring protein called antizyme (AZ) can bind to a second naturally occurring protein. The second protein, when associated with antizyme, is rapidly destroyed by the proteasome, the major cellular protease, but antizyme is not itself destroyed. The invention is further based on additional discovery that a defined portion of antizyme, approximately 97 amino acids in size, carries the critical signal that directs degradation of the second protein. Furthermore, it has been now found that this domain, located near the antizyme N-terminus (NAZ), acts as a functional module for destruction of proteins to which it can be attached. When attached to other proteins by way of the linker, the NAZ domain causes those proteins to be rapidly degraded.

In its broadest form, the invention provides a method and a means for destroying specifically targeted cellular proteins. The invention consists basically of two elements.

One of the elements of the invention is a dual-function protein containing the NAZ domain as well as a second domain, the linker, able to bind to a target protein, thus delivering NAZ domain to that target protein. Through its binding to the target protein NAZ domain provides a signal for the target protein destruction.

The second element of the invention is a means for inducing production of a compound comprising NAZ-linker domains, or production of any protein containing a specific subregion of the NAZ domain coding sequence, by treating cells or organisms with polyamines or any other compound which causes production or synthesis of antizyme or a protein functioning in substantially the same way. NAZ domain and the link proteins act as efficient regulators when delivered to cells.

The method is highly specific and efficient. The linker is selected to have appropriate substrate specificity. Consequently, the protein degradation is highly specific and discriminatory. Antizyme is normally not destroyed during the degradation of the second protein; it is reused. Therefore, one molecule of AZ is able to direct the destruction of many target proteins.

A. Dual-Function Protein

The dual-function protein can be antizyme or any other protein which contains an amino acids region therein identified as NAZ domain or region similar to NAZ domain in its functionality, which protein also contain the linker domain able to identify and recognize the target protein and bind to its degradation region.

Antizyme in quantities necessary to achieve the target protein degradation is intrinsically induced by administration of dietary or injectable polyamines. Polyamines are natural and safe constituents of all living organisms. They are products of the biosynthetic pathway initiated and controlled by ODC. Polyamines have been shown (Cell, 80:51 (1995)) to induce a shift in the reading frame the antizyme mRNA needs for its translation into antizyme protein. Antizyme fusion proteins are therefore easily produced at will by treating the subject with readily available, safe and structurally simple polyamines or their analogs. Tissue specific promoters for expression of the antizyme-linker mRNA with translation conditional on polyamine treatment are used for control of the time and site of protein degradation.

The dual-function protein is selected from the group of proteins or peptides, with exclusion of ubiquitin, that function like antizyme. These proteins, like antizyme, upon association with the target protein cause such target protein to be destroyed by the proteasome. The dual-function protein thus is any protein functionally related to the antizyme whether or not such protein or peptide is structurally related to antizyme. The dual-function protein may act by causing the potential target to be brought to the proteasome or by activating the degradative activity of the proteasome or by any other means and mechanism.

B. NAZ Domain of the Dual-Function Protein

NAZ domain of the dual-function proteins comprises a defined portion comprising about 97 amino acids in the proximity of the amino N-terminus of the dual-function protein. NAZ domain is a region comprising amino acids 1-97 of the Z1 clone (Gene, 113:191-197 (1992)). The NAZ domain function persist if amino acids 1-54 or amino acids 85-98 are removed from antizyme. Therefore, the required region of NAZ domain needed for degradative function is contained within amino acids 55-84. This region contains 30 amino acids.

The N-terminus domain promotes and directs degradation of heterologous proteins when attached to them. This domain is not necessary for the interaction of the dual-function protein with the target protein but it is necessary to induce the target protein degradation. The NAZ domain destabilizes the target protein and confers on the target protein the lability which was not present before its association with the target protein. The NAZ domain thus induces degradation in proteins which are otherwise stable and without the NAZ domain interference would not be subject to degradation. NAZ domain presence drives proteins down a proteasome-mediated degradation pathway with downstream elements common both for non-ubiquitination and ubiquitination and therefore the need for ubiquitination can be bypassed and is eliminated by the method of the invention.

NAZ domain has been shown to be effective in promoting in vitro lability and degradation. This in vitro lability can be conferred on natural substrates of in vivo degradation and used to promote the degradation in vivo.

The propensity of the NAZ domain to act as a functional module on proteins not directly in the ODC metabolic pathway is unexpected and surprising and was until now never described, disclosed or even expected to exist. It is surprising that a region of protein specific for destruction of ODC would assert similar function on other unrelated proteins. Similarly, the dual functionality of ODC specific antizyme has not been previously known and its use for selective and targeted destruction of unrelated proteins is also surprising and unexpected.

C. A Linker Domain of the Dual-Function Protein

The linker domain of the dual-function protein is typically found in the proximity of the C-terminus of the dual-function protein. This domain has a distinct function in the target protein degradation according to the invention in that it recognizes and identifies the target protein, binds to it and in this way delivers NAZ domain to the target protein.

The linker is found or designed in several ways. Naturally occurring interaction partners of the target proteins may be already known or may be found by systemic screening, such as for example by the yeast two hybrid trap or by affinity chromatography. In alternative, monoclonal antibodies are made or other variants are employed such as, for example, phage peptide display libraries screening or dimeric drug ligands.

Candidates for linkers are identified by the methods described in Nucleic Acids Res. 23:1152-6 (1995) for yeast two hybrid trap, in Proc. National Acad. Sci. (USA), 92:11456-60, (1995) for natural partners, peptide library screening etc. Other techniques, such as dimeric drug ligands, according to Science, 262:1019 (1993) may be used to bind target to linker.

Effective linkers should have two properties, high affinity and high specificity. Affinity is formally described by its equilibrium constant, determinable by standard methods, e.g. equilibrium dialysis. Specificity refers to association with the intended target, to the relative exclusion of alternative unintended targets.

Two potential problems exist which are associated with the lack of linker specificity. If the unintended targets themselves contain degradation domains capable of collaborating with NAZ domain to cause degradation, they may themselves be degraded by unintended association with the NAZ-linker. Additionally, if these targets are abundant or of higher affinity for NAZ-linker than the intended target, they may divert the NAZ-linker from the intended target. To determine whether candidate linkers are likely to be of adequate specificity, linkers are tested to see whether there are cellular proteins other than the intended target with which they associate. This is achieved, for example, by immobilizing the candidate linker on an affinity column matrix and by identifying proteins from cell extracts that associate with the column matrix.

D. Target Protein

Target protein is any protein containing a degradation region required for cooperation with the NAZ domain. NAZ domain requires cooperation with a signal region, called the degradation region located on the target protein, to induce degradation of said target protein. Because NAZ domain can function as an independent module when appended to diverse proteins it is a suitable reagent useful for directing degradation of the target protein.

The target protein is any protein which contains degradation domain which makes it, per se, already more or less unstable. The inherent instability requirement would seem to limit the utility of NAZ domain as a general means for targeting protein destruction, but the limitation is not serious. That is because most intracellular proteins important for cellular regulation are, under appropriate cellular conditions, subject to rapid intracellular degradation.

Additionally, target proteins are also other stable proteins where the degradation region could be induced or could be incorporated into NAZ-linker protein.

E. Modes of Preparation of Selective Targeted Protein Degradation

All modes of preparation of agents for selective targeted protein degradation essentially comprise the NAZ domain or a corresponding degradation directing domain and the linker domain attached covalently or noncovalently.

Any mode of action by which these agents can be prepared using methods known in the art are intended to be within the scope of this invention.

F. Mode of Action for Selective Targeted Protein Degradation

An agent for selective targeted protein degradation prepared intracellularly in any mode described in section D, is complexed with the target protein by binding through the C-terminus linker. The target protein which is susceptible to binding with the C-terminus linker domain of the dual-function protein or NAZ-linker complex typically contains its own degradation region. Basic schematics of the mode of action is as follows:

NAZ-Linker-target protein complex.fwdarw.target protein destroyed+NAZ-Linker protein remains

Modifications of this process, such as for example using instead of NAZ-linker, endogenous AZ delivered as a degradative signal to a novel target via its natural association with a modified form of ODC, i.e., ODC-linker. This allows combinatorial control by introducing ODC-linker.sub.1, ODC-linker.sub.2, etc. within a single class of cell suitable for destruction of different target proteins.

In alternative, by using NAZ-(linker=inhibitor) or endogenous complex AZ+ODC-(linker=inhibitor), with a linker designed to act as an inhibitor of the action of a target protein, inhibition can be made to occur respectively only in the presence or absence of exogenous polyamines.

These and all other modifications of the method are intended to be within the scope of invention.

During the development of this invention, the following investigations were undertaken: antizyme association with ornithine decarboxylase; investigation of distinct regions present on ODC allowing the binding of NAZ domains of antizyme to ODC and its degradation, investigation of NAZ domain function when fused to ornithine decarboxylase, investigation of NAZ domain functions when fused to other proteins and finally NAZ domain directed degradation of multiple target proteins by non-covalent association.

II. Background Studies for Selective Targeted Protein Degradation

Background studies for the current invention include confirmation that antizyme association with ornithine decarboxylase promotes ornithine decarboxylase degradation and that there exist two distinct antizyme domains for association and degradation of proteins.

A. Antizyme Association with Ornithine Decarboxylase Promotes Ornithine Decarboxylase Degradation

The first series of background studies, results of which are described in Mol. Cell. Biol., 12:3556-3562 (1992), addressed differences between intracellular degradation of mammalian (mouse) and parasitic flagellate (Trypanosoma brucei) ornithine decarboxylase (ODC) degradation.

The mouse ODC is known to be regulated by intra-cellular polyamines, and mammalian ODC was found to be selectively degraded by cells when polyamines level increased. The Trypanosoma brucei ODC is not regulated by intracellular polyamines and degradation of this protein is not dependent on polyamines.

Antizyme, a reversible tightly binding protein inhibitor of ODC was found to be involved in the ODC degradation process. While mouse ODC was found to interact with antizyme in vitro, unmodified trypanosome ODC does not interact with antizyme. Thus, the polyamine-mediated decline of ODC activity in mammals seems to be associated with the appearance of a polyamine-induced protein antizyme. Antizyme is capable of binding ODC and reversibly inhibiting enzymatic activity in vertebrates. However, antizyme itself does not and is not capable of degrading ODC.

Antizyme promotes ODC degradation by forming a binding complex with ODC and then acting in some way to promote its degradation. The difference between the polyamine regulation of mouse ODC and trypanosome ODC seems to be associated with ODC sensitivity toward and ability to bind to antizyme. When the region of mouse ODC between amino acids 117 and 140 necessary for antizyme binding was identified and tested, the constructs containing this region were shown to be inhibited and precipitated by antizyme in vitro. Replacing the mouse 117-140 amino acid region with the corresponding 117-140 amino acid region from trypanosome ODC converted mouse ODC into a form resistant to antizyme in vitro and made it unresponsive to polyamines in vivo.

These studies strongly support the conclusion that antizyme binding to certain specific region of ODC promotes degradation of ODC and imply that interaction between antizyme and ODC is a necessary step in polyamine-induced degradation. The amino acid region between amino acid 117 and 140 of the vertebrate ODC is the antizyme-binding region.

Conclusion of this initial study was that ODC possess a very specific amino acid region located between 117-140 amino acids of the 461 amino acid mouse ODC sequence which is responsible for antizyme-ODC binding. When this region is present within ODC, as in vertebrates, the antizyme-ODC binding occurs. When this region is not present or is not substantially identical to that found in vertebrates, such antizyme-ODC binding does not occur. Replacement of one region for another can either promote or prevent antizyme-ODC binding and affect in vivo regulation. Formation of antizyme-ODC binding complex is therefore necessary for regulated degradation of ODC. Such binding can only be achieved when the specific binding region on ODC is present.

B. Antizyme Domains for Association and Degradation of Proteins

In the second series of background studies, described in Mol. Cell. Biol., 13:2377-2383 (1993) and ibid, 14:87-92 (1994) the domains responsible for possible interaction between antizyme and a binding element of ODS were investigated. The studies determined that such interaction is essential but not sufficient for regulation of the ODC by polyamines. These studies additionally determined that while the binding element of ODC is present near the ODC N-terminus, a second element present at the ODC C-terminus is also required for the degradation process and that antizyme somehow causes a conformational change in the ODC. This conformational change made the C-terminus of ODC more accessible.

These findings were confirmed by studies where, when the C-terminus was blocked with antibody, the degradation was prevented. Tethering the C-terminus by a circularly permuted, enzymatically active form of ODC, also prevented antizyme-mediated degradation.

These studies have clearly shown that there are two distinct successive stages of antizyme-directed degradation.

The first stage requires only the association of antizyme and ODC. This results in catalytic inactivation of ODC and exposure of the C-terminus, demonstrated by increased access to antibody specific for the 376-to-461 region. The change in conformation induced by antizyme is not, however, sufficiently global to produce a readily detectable change in the rate or pattern of fragmentation produced by protease. The second stage encompasses proteolytic degradation per se.

Based on the prior findings that ODC contains two different functional regions, one responsible for antizyme-ODC binding near the ODC N-terminus and the second at the ODC C-terminus required for degradation process, additional studies were designed to find whether there are any distinct domains present on antizyme required for binding and proteolysis of ODC.

In these studies, C-terminal domain of antizyme alone was able only to inactivate ODC and alter its conformation, but it was not able to direct degradation of the enzyme, either in vitro or in vivo. For that, the N-terminal domain of antizyme must be present.

This investigation thus determined that antizyme also contains two distinct domains both of which are required for binding and proteolysis of ODC. These domains differ from two regions found to be present on ODC.

Based on obtained results, the following sequence of events was found to lead to the ODC degradation cycle. Elevation of cellular polyamines results in post-transcriptional induction of antizyme which leads to association of antizyme with ODC, resulting in conformational changes in the ODC causing its enzymatic inactivation and carboxy-terminal exposure. The carboxy-terminal exposure allows interaction of the ODC-antizyme complex through the antizyme N-terminus with a protease causing destruction of ODC leading to reduction of polyamine biosynthesis.

In the background studies, the presence within antizyme of a region of unknown biochemical function that is needed to elicit degradation of ODC was discovered.

III. A Selective Targeted Protein Degradation

The studies described in sections A and B show that regulated degradation of ornithine decarboxylase (ODC) is mediated by its association with the inducible protein antizyme and that the N-terminus of antizyme (NAZ), although unnecessary for the interaction with ODC, must be present to induce ODC degradation.

In the follow-up studies leading to this invention, it was surprisingly discovered that covalent fusion of N-terminus domain of antizyme (NAZ) to ODC confers on ODC and other target proteins lability that normally results only from the non-covalent association of native antizyme and ODC.

A. NAZ and Its Function in Degradation of Proteins

To determine whether the function of NAZ domain is specific only for ODC or whether it could act in a similar fashion as a general modular function domain when grafted to other heterologous proteins, recombinant DNA constructs were prepared and used in studies described below.

Results of studies leading to and resulting in the discovery of targeted degradation of proteins in situ are illustrated in FIGS. 1-13.

A. Degradation of Mouse ODC is Induced Solely by NAZ Domain without Presence of CAZ Domain

To test whether NAZ domain alone, without presence of the C-terminus domain CAZ, can promote ODC degradation, recombinant DNA fusion constructs of mouse ODC with NAZ domain (NAZ-MODC) were prepared according to Example 6.

Mouse ODC (MODC) is relatively stable protein in an in vitro degradation system derived from rabbit reticulocytes. Its degradation, however, is promoted and enhanced by the addition of the regulatory protein antizyme. As described above, antizyme is known to contain two functionally distinguishable domains. The C-terminus domain of the protein (CAZ) binds to ODC, whereas the N-terminal part of antizyme (NAZ) has capability to direct degradation of ODC in vitro or in vivo.

To determine whether the NAZ domain is solely responsible for protein degradation or whether the C-terminus (CAZ) domain is also necessary, a portion of antizyme, specifically amino acids 1-97 of antizyme N-terminus was fused to MODC to make fusion protein NAZ-MODC. Results are seen in FIG. 1.

FIG. 1 shows in vitro degradation of mouse ODC induced by fusion of amino acids 1-97 of NAZ domain to mouse ODC (MODC). As a control, the C-terminus (amino acids 106-212) of antizyme (CAZ) was also fused to MODC to make CAZ-MODC. The time course of in vitro degradation of ›.sup.35 S!-methionine labelled proteins was examined and the labeled proteins remaining undestroyed after the indicated period of incubation were examined by SDS-PAGE and autoradiography. FIG. 1A shows structures of antizyme and fusion proteins NAZ-MODC and CAZ-MODC. Antizyme N-terminal half (NAZ) is depicted as a solid block. Antizyme C-terminal half (CAZ) is depicted as a fine cross-hatch block. MODC is depicted as a coarse cross-hatch block.

FIG. 1B shows MODC fused to the C-terminal half of antizyme to form CAZ-MODC fusion protein or to the N-terminal half of antizyme to form NAZ-MODC fusion protein. The fusion proteins seen in FIG. 1 were subjected to degradation for 0, 1, and 2 hours. The position of migration of each protein is indicated by arrow.

As seen from FIG. 1B, NAZ-MODC fusion protein was gradually destroyed with time and was largely degraded in 2 hours. In contrast, the CAZ-MODC fusion protein was relatively stable and no degradation was observed during the same time when the NAZ-MODC was degraded. The fusion construct NAZ-MODC contains the two domains needed for degradation, the NAZ domain and the mouse ODC C-terminus. CAZ-MODC does not contain NAZ domain. NAZ domain alone was, therefore, responsible for such degradation of ODC. When provided with a means to associate with its target, whether that association is non-covalent as with intact antizyme, or covalent as in the AZ-MODC fusion protein, NAZ domain was capable to direct the ODC protein degradation.

Results of this study show that NAZ, when coupled directly to ODC by way of fusion constructs, is solely responsible for antizyme NAZ domain degradative capability.

B. Two Functional Domains Jointly Direct Degradation of Trypanosome ODC

To test whether NAZ domain can destabilize otherwise stable protein and promote degradation in a protein which is insensitive to polyamines and not subject to antizyme induced degradation, recombinant DNA fusion constructs of NAZ domain with Tb367M (NAZ-Tb367M) were prepared. Tb367M is a chimeric protein in which a junction between mouse and Trypanosome brucei ODC (Tb376M) was created at 367 amino acid. NAZ-Tb367M was prepared according to Example 6. Results are shown in FIG. 2.

Ornithine decarboxylase from Trypanosome brucei (TbODC), although almost 70% identical to MODC in its core structure, is a stable protein. TbODC is insensitive to regulation by polyamines in the native context of the parasite or when expressed in mammalian cells. It lacks both the C-terminal degradation domain and the antizyme-binding domain present in the mouse ODC. The protein can be converted into one that is unstable in vivo, but still unresponsive to polyamines, by replacing its C-terminus with the degradation domain of mouse ODC. Such conversion has been done by making chimeric proteins, in which a junction between mouse and trypanosome ODC, without insertion or deletion, is created at amino acid 376 to make Tb376M or at 422 to make Tb422M, as shown in FIG. 2A.

FIG. 2 shows TbODC degradation induced by the N-terminus of antizyme (NAZ) and C-terminus of mouse ODC.

To test whether NAZ domain was capable of inducing Tb376M degradation in vitro, NAZ domain was coupled in front of the Tb367M protein, to make NAZ-Tb376M, and NAZ-MODC. Structures of these proteins are seen in FIG. 2A.

FIG. 2A shows structures of trypanosome ODC (TbODC), Tb327M and Tb422M chimeras, fusion of NAZ domain with trypanosome ODC (NAZ-Tb ODC) and fusion proteins comprising trypanosome ODC having attached NAZ domain at its N-terminal region and mouse C-terminal domain containing amino acids 376-461 (NAZ-T376M) or 422-461 (NAZ-Tb422M) at the C-terminal region. TbODC is depicted as an open block. NAZ domain is depicted as a solid block. MODC C-terminus is depicted as a cross-hatched block.

Results of degradation analysis of chimeric proteins TbODC, T376M alone, NAZ-TbODC and NAZ-Tb376M is seen in FIG. 2B.

FIG. 2B shows in vitro translation of TbODC, Tb376M, NAZ-TbODC and NAZ-T376M proteins. These proteins were labelled with ›.sup.35 S!-methionine, subjected to degradation from 0 to 2 hours and analyzed as in FIG. 1. The Tb376M and Tb422M chimeras, like most other short-lived proteins, were relatively stable in the in vitro degradation assay.

In vitro degradation analysis illustrated in FIG. 2 showed that NAZ-Tb376M was gradually degraded during the two hours and that both NAZ domain and the C-terminus of mouse ODC were needed to induce efficient degradation of TbODC. Neither the C-terminus domain of mouse ODC attached to TbODC (T376M) nor NAZ domain attached to TbODC (NAZ-TbODC) alone were capable of changing the stability of TbODC. Densitometric analysis of the data of FIG. 2B is displayed as FIG. 2C. FIG. 2C shows quantitation of degradation of TbODC. As seen in FIG. 2C, while there was some degradation observed for TbODC, T367M and NAZ-TbODC, the degradation of NAZ-Tb367M protein was more than double.

Degradation of protein with NAZ domain was not limited to the Tb376M but was also observed for Tb422M protein. Results are seen in FIG. 2D. As seen in FIG. 2D, NAZ domain also stimulated the degradation of Tb422M. NAZ-Tb422M contains the entire open reading frame of trypanosome ODC except for the last 2 amino acids, sandwiched between NAZ domain and the C-terminus (amino acids 422-461) of mouse ODC.

The studies described above show that both functional domains, that is NAZ domain and the C- terminus domain together, are necessary and sufficient to cause in vitro destabilization and degradation of a stable trypanosomal ODC protein.

For studies illustrated in FIG. 3, control fusion proteins were made according to Example 6. These proteins consisted of NAZ-TbODC extended in its C-terminus by either full length human papilomavirus 16 E6 (151 amino acids) or the N-terminal 90 amino acids of that protein. These proteins were prepared in order to determine whether any extension, regardless of sequence specificity, can induce degradation of NAZ-TbODC. Both fusion proteins were subjected to the degradation assay as described in Example 7. During the 2 hour incubation, these proteins were not degraded, confirming that the C-terminus of mouse ODC must contain a specific functional sequence motif that is able to cooperate with NAZ.

To determine whether antizyme-mediated degradation of ODC by the 26S proteasome is independent of ubiquitination, the studies were designed to examine whether degradation induced by appending NAZ domain involves ubiquitin modification. In these studies, NAZ-Tb376M and NAZ-Tb422M were incubated as for in vitro degradation, but ATP was substituted with ATP-.gamma.-S. ATP-.gamma.-S has been previously shown to block degradation but not ubiquitination (Cell, 63:1129-1136 (1990)), thereby leading to the accumulation of high molecular weight forms of a target protein decorated by multiple ubiquitin chains.

FIG. 3 illustrates effect of ATP-.gamma.-S on degradation of NAZ-Tb376M and NAZ-Tb422M proteins.

In FIG. 3A, proteolysis was induced by co-incubation of oncoprotein p53 with human papilomavirus 16 E6 (left three lanes) or by fusion of NAZ domain to Tb376M and Tb422M (right six lanes). Samples were analyzed at time 0 or after 2 hours of incubation. As indicated, reactions were performed with ATP or replacing ATP with ATP-.gamma.-S.

As seen in FIG. 3A, both NAZ-Tb376M and NAZ-Tb422M were degraded in the presence of ATP. Substitution of ATP-.gamma.-S for ATP blocked degradation, but high molecular weight conjugates of NAZ-Tb376M and NAZ-Tb422M did not appear. As a positive control, HPV16 E6-mediated degradation of p53 which is ubiquitin depended, was used. In the presence of E6 and ATP, p53 was degraded. When ATP-.gamma.-S was used instead of ATP, degradation was blocked and high molecular weight forms of p53 accumulated.

In FIG. 3B, the composition of conjugates produced as in FIG. 3A was tested by immunoprecipitation with anti-p53 antibody PAb421, followed by separation on SDS-PAGE, and autoradiography (lanes 1 and 2), or by transfer to nitrocellulose paper and immunodetection with anti-ubiquitin antibody (lanes 3 and 4). Lanes 1 and 3 show p53 without E6. Lanes 2 and 4 show p53 incubated with E6 and ATP-.gamma.-S. The position of high molecular weight ubiquitin conjugates and of IgG heavy chains reactive with the secondary antibody is indicated.

As seen in FIG. 3B, immunoprecipitation with a monoclonal antibody to p53 (PAb42) followed by Western blot analysis using anti-ubiquitin antiserum confirmed that the high molecular weight proteins enhanced in the presence of ATP-.gamma.-S consisted of poly-ubiquitinated p53.

These results show that NAZ-induced degradation of NAZ-Tb376M and NAZ-Tb422M, like antizyme-mediated degradation of ODC, is ubiquitin independent.

C. The Destruction Box of Cyclin B Cooperates with NAZ Domain in Protein Degradation

To test whether other short-lived proteins can confer lability on the otherwise stable TbODC protein, the cyclin B degradation domain within amino acids 13-90 was fused to the C-terminus of TbODC. Results are seen in FIG. 4.

Sea urchin cyclin B is well-characterized short-lived protein that accumulates at interphase and is degraded at metaphase. Degradation of the protein allows cells to exit metaphase and enter interphase. The protein is stable in frog oocyte extracts prepared from interphase cells, but it is rapidly degraded in metaphase extracts. The specific capacity of mitotic cell extracts to degrade cyclin B thus reproduces faithfully the property of the cell of origin.

Degradation of cyclin B was previously (Nature, 349:132-138 (1991)) shown to require that it contain both ubiquitination sites and a destruction box. Its N-terminus contains all the structural information needed for degradation. Its amino acid 13-90 region fused to staphylococcal protein A is able to induce degradation in metaphase cell extracts. A smaller region, amino acids 13-59, does not contain the ubiquitination sites and does not undergo proteolysis, presumably because the lysines within amino acids 60-66 are not present to serve as ubiquitin modification sites. Likewise, a single mutation at the conserved arginine in the destruction box (amino acids 42-50), a conserved region in the N-terminus, prevents cycle-specific degradation.

In these studies, NAZ domain was fused to two amino acid regions of cyclin B protein which has been shown to contain two region in its N-terminus. The region of cyclin B.sub.13-90 (amino acids 13-90) is capable of undergoing degradation when fused to Protein A. The other region of cyclin B.sub.13-59 (amino acids 13-59), is not subject to such degradation.

To test whether the cyclin B degradation domain within amino acids 13-90 can confer lability on the otherwise stable TbODC, this region was fused with C-terminus of TbODC in the presence or in the absence of NAZ domain.

FIG. 4 shows NAZ-induced degradation of TbODC-cyclin B.sub.13-90, with ubiquitination site, and TbODC-cyclin B.sub.13-59, without ubiquitination site.

FIG. 4A shows a structure of Tb-cyclin B.sub.13-90, NAZ-Tb-cyclin B.sub.13-90 and NAZ-Tb-cyclin B.sub.13-59 fusion proteins. NAZ domain is depicted as a solid block. TbODC is depicted as an open block. Cyclin B is depicted as a hatched block. The region of cyclin B.sub.13-90 containing lysines that are the ubiquitination sites of cyclin B is marked "K".

FIG. 4B shows Tb-cyclin B.sub.13-90, NAZ-Tb-cyclin B.sub.13-90 and NAZ-Tb-cyclin B.sub.3-59 which were subjected to degradation under the conditions described above. The results were analyzed as described in FIG. 1.

As shown in FIG. 4B, the TbODC-cyclin B.sub.13-90 fusion protein was stable in the reticulocyte lysate. This result agreed with the conclusion that this portion of cyclin alone cannot promote degradation of protein A. However, the TbODC-cyclin B.sub.13-90 fusion protein became unstable after NAZ domain was appended to form NAZ-TbODC-cyclin B.sub.13-90. Furthermore, deleting the lysine residues that are putative sites of ubiquitination from NAZ-TbODC-cyclin B.sub.13-90, to form NAZ-TbODC-cyclin B.sub.13-59, did not alter its degradation.

The association of these regions of cyclin B with NAZ domain made both NAZ-cyclin B.sub.13-90 and NAZ-cyclin B.sub.13-59 unstable and subject to degradation. Furthermore, NAZ domain cyclin B.sub.13-59 complex was able to induce in vitro degradation of Trypanosoma brucei ODC, a stable protein.

These results confirm that cyclin B amino acids 13-59 contains a degradation domain that can cooperate with NAZ domain to promote trypanosome ODC degradation in vitro. Such degradation is independent of the presence of lysine residues contained within amino acids 60-90, which are the normal targets for ubiquitination, further confirming that NAZ domain can provide an alternate signal for degradation that bypasses the requirement for ubiquitin modification.

To test whether the NAZ domain has a capability to directly induce degradation of the cyclin B degradation region 13-90 in a rabbit reticulocyte lysate, NAZ domain was coupled in front of either amino acids 13-90 or amino acids 13-59. Results are seen in FIG. 5.

FIG. 5 illustrates NAZ-induced degradation of cyclin B 13-90 (with ubiquitination site) and cyclin B 13-59 (without ubiquitination site). Unfused cyclin B 13-59 was used as control.

FIG. 5A shows a structure of NAZ-cyclin fusion proteins. NAZ domain is depicted as a solid block. Cyclin B is depicted as a hatched block.

In studies shown in FIG. 5B, the proteins were subjected to degradation and the results were analyzed as previously described in FIG. 1. As seen in FIG. 5B, both fusion proteins were degraded while the control cyclin B 13-59 remained intact.

FIG. 5C shows quantitation of degradation of above fusion proteins. Triangles represent cyclin B 13-90 used as a control. Open circles represent NAZ-cyclin B 13-90. Solid circles represent NAZ-cyclin B 13-59. Fusion, NAZ domain containing proteins were quantitatively destroyed about twice as fast as the control proteins.

As expected for a metaphase-specific substrate subjected to degradation in the reticulocyte-based degradation assay system, the cyclin B degradation region 13-90 alone used as a control was relatively stable. NAZ, on the other hand was able to direct the degradation of both fusion proteins, regardless of presence or absence of ubiquitination sites. NAZ domain therefore can alter the degradation properties of cyclin B amino acids 13-90 in two ways, by promoting its lability in the absence of cycle specific signals and by diverting it to a pathway that does not require ubiquitination.

D. Identification of Degradation Region of the Target Protein that Confers Lability

In order to determine which region of the NAZ domains, antizyme or the target protein confers lability, the series of additional studies were performed.

As was seen in prior sections, for polyamine-dependent degradation to occur, two elements are necessary to be present within ODC, namely the antizyme binding region where the C-terminal domain of the antizyme attaches antizyme to the ODC and the C-terminal degradation region where the N-terminal domain of antizyme promotes the ODC degradation. The N-terminus half of antizyme is thus not involved in the antizyme/ODC interaction, but it is required for promoting ODC degradation. Because NAZ domain is a functional module that only functions in collaboration with a degradation domain of the target protein, identification of such domains within target proteins was investigated.

For these studies, a labile p53 protein was selected as a candidate.

Tumor suppressor protein p53 imposes negative regulation on the cell cycle, and abnormal inactivation of the protein has been shown to be related to cell transformation and tumorigenesis. Elimination of p53 function takes place by multiple means, including degradation, interaction with other proteins, interaction with mutant forms of p53 itself, or somatic or germline mutation. Degradation of p53 is influenced by protein-protein interactions. A variety of viral oncoproteins have been shown to act on p53. The association of p53 with SV40 large T antigen or with adenovirus 2 E1B-55K protein extends its half-life, but the E6 protein of human papillomavirus (HPV) 16 shortens it. Mutation of p53 can change its function and stability. The multiple influences that can play on p53 stability make its degradation a suitable subject for study of effect of NAZ attachment to p53 degradation domain.

Normal degradation of p53 is mediated by the ATP-dependent and nonlysosomal large 26 S protease complex. Ubiquitination is required for p53 degradation. E6 and the associated host cell protein E6-AP act as a ubiquitin ligase E3, which leads to ubiquitination of p53 and its degradation. The structural elements of p53 needed for recognition and destruction by the proteolytic machinery remain unclear, as is true for most short-lived proteins.

For the 26 S protease to act and destroy the target protein, the target protein must contain a structural motif, so called a degradation region. In order to undergo protein degradation, whether via ubiquitin-dependent or ubiquitin-independent pathways, this region is required.

In the studies described below, a degradation region of p53 was identified by fusing it to NAZ domain and making deletions within the p53 moiety. The identified degradation region of p53 was able to function like the degradation region of mouse ODC. Like the degradation region of mouse ODC, p53 was able to induce degradation of trypanosome ODC, otherwise a stable protein.

E. Identification of the p53 Degradation Domain

To determine if p53 has a region that can collaborate with NAZ domain, NAZ domain was fused directly to p53 and the lability of NAZ-p53, as well as a series of deletions within the 393-amino acid p53 moiety of the NAZ-p53 parent molecule were analyzed. Deletions were made from each end or from both ends. Results are seen in FIGS. 8 and 9.

FIG. 6 shows structure of entities depicting p53, NAZ-p53, NAZ-p53 fragments and numerous deletions of p53 moiety. Deletions were made from each end or from both ends. NAZ is depicted as a solid block. p53 is depicted as a hatched block. Entities depicted in FIG. 6 were submitted to NAZ domain induced degradation.

FIG. 7 shows results of NAZ domain induced degradation of entities depicted in FIG. 6.

FIG. 7A shows degradation pattern of p53 and NAZ-p53. As seen in FIG. 7A, there was slight degree of degradation of p53 observed using the degradation conditions described in Example 7. Under the same conditions, NAZ-p53 was degraded about twice as fast.

FIG. 7B shows degradation pattern of fusion protein entities containing the C-terminal deletions of p53, namely NAZ-p53 1-304, NAZ-p53 1-202, and NAZ-p53 1-103. As seen in FIG. 7B, both NAZ-p53 1-304 and NAZ-p53 1-202 and NAZ-p53 1-103 were degraded by about 50% at two hours. NAZ-p53 1-202, on the other hand was almost completely degraded during the same time.

FIG. 7C is shows degradation pattern of fusion protein entities containing the N-terminal deletions of p53, namely NAZ-53 101-393, NAZ-p53 201-393, and NAZ-p53 301-393. As seen in FIG. 7C, N-terminus deletions of p53 resulted only in small degree of degradation in all NAZ-p53 101-393, NAZ-p53 201-393 and NAZ-p53 301-393. Such degradation was qual to the degradation observed for p53.

FIG. 7D shows a degradation pattern of fusion protein entity containing C terminal NAZ-p53 1-150 having removed amino acids 151-200 and entity containing internal deletion of p53, namely deletion of NAZ-p53 100-139 from NAZ-p53 1-200. As seen in FIG. 7B, the internal deletion of amino acids 100-139 increased a degree of p53 degradation to only about 30% compared to about 74% of degradation observed for NAZ-p53 1-150.

AS seen in FIG. 7B, NAZ-p53 1-202 was found to be very extensively degraded and deleting 99 amino acids from its C-terminus (NAZ-p53 1-103) greatly stabilized it. Degradation of NAZ-p53 1-202 is more extensive than that of NAZ-p53, the parent molecule, or of NAZ-p53 1-304. This seem to occur because either amino acids 203-304 protect amino acid 103-202 region by altering its structure or denying access to a protease; or the degradation domain in p53 amino acid 103-202 region needs, like the degradation domain of mouse ODC, to be topologically at or close to the C-terminus of the fusion protein.

The obtained results indicate that amino acids 101-202 contain a degradation domain. A more refined deletion from the C-terminus of p53 further demonstrated, as seen in FIG. 7D that removal of amino acids 151-202 from NAZ-p53 1-202 to form NAZ-p53 1-150 did not interfere with degradation, but the internal deletion of amino acids 101-139 from .DELTA. NAZ-p53 1-200 to form NAZ-p53 1-200 A 101-139 prevented degradation.

These results show that p53 amino acids 100-150 region contains or is a degradation region of p53 that is active in a fusion with NAZ domain.

F. NAZ-mediated Degradation of p53.sub.1-150 is Ubiquitin-independent

To examine whether the NAZ-p53 fusion protein could in principle undergo degradation via a ubiquitinated intermediate, as is normal for p53, or by a ubiquitin-independent pathway, as is the case for ODC when it is associated with antizyme, in vitro degradation was carried out in the presence of ATP-.gamma.-S. ATP-.gamma.-S has been shown to block degradation but not ubiquitination. By grafting NAZ to p53, a fusion protein that is labile in vitro was created and deletion analysis was used to identify a degradation-determining region within p53. As a positive control, HPV16 E6-mediated degradation of p53, which is ubiquitin-dependent was used. Results are seen in FIG. 8.

FIG. 8 shows effect of ATP-.gamma.-S on p53 degradation. Proteolysis was induced by non-covalent association of p53 with E6 (left three lanes) or by fusion of NAZ to p53 amino acids 1-150 (right three lanes), and the products of the reaction were analyzed by SDS-polyacrylamide gel electrophoresis and autoradiography. In both cases, time-dependent proteolysis occurred in the presence of ATP. When ATP-.gamma.-S was used instead of ATP, proteolysis was inhibited in both cases. In the presence of ATP-.gamma.-S, high molecular weight forms of p53 accumulated in the presence of E6 but not in its absence. Under identical conditions of incubation, no high molecular weight forms of NAZ-p53 1-150 were detected. Substitution of ATP-.gamma.-S for ATP blocked degradation, but no high molecular mass form of the NAZ-p53 1-150 fusion protein appeared. It is clear that NAZ-induced degradation of p53 protein is ubiquitin-independent. The N-terminus of antizyme seems to bypass the modification normally required for p53 degradation and imposes instead its own properties otherwise associated with degradation of ODC.

G. A Monoclonal Antibody Blocks E6-mediated p53 Degradation

In this study, the degradation region, identified within the fusion proteins was investigated to show if it itself is involved in degradation of p53 itself by using a monoclonal antibody (PAb246) that specifically recognizes mouse p53 amino acids 88-109. Antibody PAb421 was used as a control. The PAb246 was used to determine whether it can block E6-mediated p53 degradation. Because there is no cross-reaction of the antibody with human p53 and because similarly directed antibodies to the human protein were not available, mouse p53 was used in this study. PAb246 recognizes the native conformation of p53 epitope. Results are seen in FIG. 9.

FIG. 9 shows the effect of monoclonal antibody PAb246 which is specific for p53 amino acids 88-109 on E6-mediated degradation of mouse p53. ›.sup.35 S!-Methionine-labeled mouse p53 was subjected to degradation according to example 7 in the absence of HPV16 E6 (Lane 1) and in the presence of E6 (Lanes 2-4). p53 was preincubated with control monoclonal antibody PAb421 (Lane 3) or with monoclonal antibody PAb246 (Lane 4).

As seen in FIG. 9, mouse p53 degradation was induced by human E6, but the presence of the antibody prevented this degradation. A control monoclonal antibody (PAb421) against the C-terminus of p53, however, had no effect on degradation. These results show that the region of p53 identified as necessary for NAZ domain degradation must be accessible for that process to occur and that the epitope specific antibody can obstruct such access.

H. The Degradation Domain of p53 Induces Trypanosome ODC Degradation

It was previously shown that the mouse ODC C-terminus is required for its in vivo degradation and that it confers lability on the homologous but stable ODC (TbODC) found in Trypanosoma brucei.

To find whether amino acids 100-150 of p53 contain the degradation domain which can act in place of the mouse ODC C-terminus, this region was fused to the C-terminus of TbODC to make TbODC-p53 100-150. C55.7 ODC-deficient Chinese hamster ovary cells was stably transfected with TbODC-p53 100-150. The stability of the fusion protein was determined by measuring the rate at which ODC activity, provided by the fusion protein, declined in cells treated with cycloheximide to inhibit protein synthesis. Measurement of ODC activity provides an accurate surrogate determination of its protein level. Results are seen in FIG. 10.

FIG. 10 illustrates effect of the p53 degradation domain (amino acids 100-150) on in vivo degradation of trypanosome ODC (TbODC). Components of TbODC chimeric proteins are depicted as open block. Components of p53 chimeric proteins are depicted as hatched blocks. Triangle represent TbODC activity, circles represent TbODC-p53 100-150 activity, solid symbols represent treated cells and open symbols represent untreated control cells. TbODC and TbODC-p53 100-150 initial activities were 51.0 and 15 pmol/mg/min, respectively. ODC stability in cells expressing TbODC or Tb-p53 100-150 was assessed by inhibiting protein synthesis with cycloheximide and determining residual enzymatic activity immediately and after 1, 2, or 3 hours. Control cells were similarly incubated, harvested, and analyzed but were not treated with cycloheximide. ODC activities are represented as a percentage of initial activity.

As seen in FIG. 10, ODC activity fell with a half-life of about 30 min. In contrast, the ODC activity provided by a construct encoding TbODC without the p53 degradation domain remained unaffected by cycloheximide treatment. These results shows that amino acids 100-150 of p53 can confer in vivo lability on the stable protein TbODC.

I. The Degradation Domain of p53 Plus the Antizyme Binding Domain Confers Polyamine Regulation on Trypanosome ODC

To determine whether the degradation domain of p53 can cooperate with the antizyme binding domain to subject TbODC to polyamine-mediated degradation, a binding site for antizyme was provided. The N-terminus of the construct TbODC-p53 100-150 was replaced with an equivalent region of mouse ODC containing the antizyme binding domain to form M314T-p53 100-150. For this study, cells were transformed as above and treated with the polyamine precursor putrescine for the indicated times to induce antizyme. ODC activity in the form of M314T-p53 100-150 was regulated by polyamines. Results are seen in FIG. 11.

FIG. 11 shows effect of the p53 degradation region amino acids 100-150, on polyamine-dependent degradation of chimeric mouse/trypanosome ODC (MS14T). Components of TbODC chimeric proteins are depicted as open blocks. Components of mouse ODC chimeric proteins are depicted as cross-hatched blocks. Components of p53 chimeric proteins are depicted as hatched blocks.

To elicit polyamine-mediated regulation and induce endogenous antizyme, putrescine was added to the culture medium of cells expressing M314Tb or M314Tb-p53 100-150 to a final concentration of 100 .mu.M. Cell lysates were prepared after the indicated time of treatment and assayed for ODC activity as described in Mol.Cell Biol., 13:2377-2388 (1993). Control cells were similarly incubated, harvested, and analyzed but were not treated with putrescine. ODC activities are represented as the percentage of initial activity. M314Tb and M314Tb-p53 100-150 initial activities were 72.9 and 6.6 pmol/mg/min, respectively. M314T is depicted as circles. M314T-p53 100-150 is depicted as triangles. Treated cells are depicted as solid symbols. Untreated control cells are depicted as open symbols.

As seen in FIG. 11, within 4 hours of treatment ODC declined by about 65% from control values and subsequently remained at this level. However, M314Tb, identical to M314Tb-p53 100-150 but for the absence of the degradation domain of p53, was insensitive to regulation by polyamines. Therefore, p53 in this context can provide the same functional properties as the C-terminal degradation region of ODC.

In the above studies, NAZ-p53 protein constructs were utilized to find, using deletion analysis of the p53 moiety, whether these constructs would be labile. Results confirmed that these constructs would be labile but only if they contain amino acids 100-150 of p53. The domain containing amino acids 100-150 was identified as putative degradation domain of p53. The domain was able to destabilize trypanosome ODC, as does the degradation domain of mouse ODS. Additionally, it was found that in mouse p53, E6-mediated degradation was blocked by an antibody specific for this region.

Previously, the mammal binding region of p53 for large T antigen targeting has been identified as amino acids 94-293. The degradation domain of p53, which was identified here, is located within this binding region. Therefore, the association of p53 with large T antigen could cover the degradation domain and prevent p53 degradation. Second, the findings are consistent with the distribution of mutations that increase the half-life of p53. Mutations in p53 are commonly found in tumors, and the mutated protein often has a much longer half-life than wild type p53. The mutations are widely distributed within the protein but are mainly located in four regions. One of the regions lies in the amino acid 100-150 degradation domain. Mutations in the region may either directly change the degradation domain itself to prevent protease targeting or change protein conformation to prevent ubiquitination.

IV. A Method for Selective Targeted Protein Degradation

Based on obtained results, the following sequence of events illustrates the target protein degradation cycle. Elevation of cellular polyamines results in post-transcriptional induction of antizyme including NAZ domain and C-terminus domain. Induction of antizyme leads to the association of antizyme with a target protein through its C-terminus target specific domain. This allows the interaction of the target protein antizyme complex through the antizyme N-terminus with a protease causing destruction of the target protein.

The above outlined general method is used for specific destruction of target proteins intracellularly.

A. Treatment of Cancer

Human cancer is commonly caused or exacerbated by the presence of proteins in abnormal form or in excess amount within cells. Examples of these proteins are products of the genes ras, myc, fos or p53, that disrupt signals limiting normal growth. For destruction of these proteins, a fusion protein is made consisting of antizyme linked to a targeting peptide that binds to these cancer-producing proteins and causes the fusion protein to be expressed in cancer cells. In this way, normal control of cell growth is restored.

Cancer cells themselves commonly produce polyamines in excess amounts compared to normal cells. Precise delivery of a gene encoding the fusion protein to cancer cells exclusively typically present difficulty. These difficulties are overcome by the method of the invention. By using a gene requiring polyamine-dependent frameshifting for expression of the encoded fusion protein according to the invention, cancer cells would be subject to a more significant degree of inhibition of growth than normal cells, even if both kinds of cells contain the gene encoding the fusion protein, because endogenous polyamines in the cancer cells cause polyamine-dependent frameshifting.

B. Specific Protection of Normal Growing Cells from Toxicity Caused by Cancer Therapy

Commonly used forms of cancer therapy attack growing cells, regardless of whether these are cancer cells or innocent-bystander normal cells. The resultant damage or destruction of normal cells can limit the extent of existing therapeutic treatment that is tolerable.

Cellular proteins that control cell growth include the cyclins, cyclin-dependent kinases with which these cyclins interact and additional proteins that effect the activity or stability of the cyclin-kinase pairs. A fusion protein prepared according to the invention consisting of antizyme linked to a targeting peptide that binds to one of these proteins and causes the fusion protein to be expressed in normal cells, impedes cell growth, thereby protecting the normal cells from the toxicity associated with therapy. Because growth of normal cells must be allowed to remain unimpeded between periodic applications of cancer therapeutic agents, expression of the fusion protein consisting of antizyme linked to a targeting peptide must be limited to the time of treatment. By using a gene requiring polyamine-dependent frameshifting for expression of the encoded fusion protein, it is possible to impede cell growth in normal cells at the desired time by administration of polyamines.

C. Transgenic Plants and Animals

Using the method of the invention, transgenic plants or animals are created in which a gene expressed in one or more tissues encodes a fusion protein, as above, that is expressed as a protein only when polyamines are administered to induce polyamine-dependent frameshifting. By using antizyme fusion proteins that cause the destruction of hormones or the destruction of proteins that either induce or limit the production of hormones in this manner, commercially useful physiologic alterations of the transgenic plants or animals are made to happen at desired times. These include, for example, the ability to alter nutritional content, cause or retard fruit ripening, cause or retard seed germination, and, to facilitate harvesting, cause abscission.

Specific proteins within cells regulate many essential processes, such as growth or differentiation. Proteins with such regulatory roles are often short-lived. NAZ-linker protein makes target proteins much more labile, irreversibly destroying their ability to influence cellular behavior. Antizyme is intrinsically inducible by dietary or injectable polyamines, which are the natural and safe constituents of all living organisms and products of the biosynthetic pathway initiated by ODC. Polyamines induces a frameshift in the reading of the antizyme mRNA needed for its translation into antizyme protein. Antizyme fusion proteins can thus be made present or absent at will by treating with readily available, safe and structurally simple polyamines or analogs. By combining use of tissue-specific promoters for expression of the antizyme-linker mRNA with translation conditional on polyamine treatment, both the time and place of application can be controlled.

This example describes methods used for production of antizyme (AZ) cDNA and the expression vectors used.

The rat antizyme .lambda. gt11 partial cDNA clone Z1 was prepared according to J. Biochem., 108:365-371 (1990) and Gene, 113:191-197 (1992). The dexamethasone-inducible vector pMaMneoZ1 was obtained according to J. Biol. Chem., 167:13138-13141 (1992). Z1 and pMAMneoZ1 have a 212-amino-acid open reading frame. A numbering convention that specifies the start of this reading frame to be amino acid 1 of antizyme was used.

This example describes methods used for preparation of antizyme fusion proteins and for construction of plasmids encoding antizyme glutathione S-transferase (GST) fusions.

Rat AZcDNA Z1 of Example 1 was expressed as a fusion protein C-terminal to glutathione S-transferase (GST) by cloning into the BamHI-EcoRI sites of the pGEX-3 vector according to Gene, 67:31-40 (1988). Deletions within the AZ Z1 212-amino-acid open reading frame were made by PCR, using oligonucleotides that included BamHI or EcoRI sites.

The resulting amplified PCR fragments were restricted and ligated into the pGEX-3 vector to create fusion proteins with various deletions in the antizyme portion. To express the fusion proteins, Escherichia coli carrying these recombinant plasmids was induced with isopropylthiogalactopyranoside, (IPTG), and the proteins were then purified with glutathione-Sepharose 4B beads (Pharmacia). For use in the degradation assay, proteins were eluted from beads with 0.1M glycine and 0.1M NaCl (pH 2.5), and the eluates were neutralized with 1.5M Tris-HCl (pH 8.8).

This example describes methods used to achieve in vitro transcription of antizyme cDNA into cRNA and subsequent translation into an appropriate protein.

Plasmid DNA was used as a template to amplify chimeric ODC cDNA or AZ cDNA by PCR, using as the 5' oligonucleotide a primer containing a T7 RNA polymerase recognition site, and the amplified DNA was then transcribed into cRNA with T7 RNA polymerase according to Mol. Cell. Biol., 12:3556-3562 (1992). cRNA was translated in vitro by using a rabbit reticulocyte lysate (Promega). The mouse ODC, radiolabeled with ›.sup.35 S!methionine, was used for a binding assay or, in unlabeled form, for an inhibition assay.

ANTIZYME BINDING AND INHIBITORY ACTIVITY

This example describes assays used for determination of antizyme binding and inhibitory activity.

Ten microliters of translation lysate containing ›.sup.35 S!methionine-labeled mouse ODC was added to 50 .mu.l of phosphate-buffered saline containing 0.05% Triton X-100 and 15 .mu.l of glutathione-Sepharose 4B beads (Pharmacia) bearing the fusion proteins. After shaking at 4.degree. C. for 1 hour, the beads were washed three times with phosphate-buffered saline. The proteins associated with the beads were made soluble in sodium dodecyl sulfate (SDS) sample buffer, and the radiolabeled ODC was analyzed by SDS-polyacrylamide gel electrophoresis (PAGE) and autoradiography.

For the inhibition assay, 10 .mu.l of beads bearing the fusion proteins was mixed with 10 .mu.l of in vitro-translated mouse ODC and incubated on ice for 5 min. ODC activity of the mixture was then determined as described in Mol. Cell. Biol., 12:3556:3562 (1992).

This example illustrates assays used to determine in vivo regulation using antizyme 106-212.

DNA encoding antizyme amino acids 106 to 212 (AZ 106-212) was amplified by PCR, using the rat AZ .lambda. gt11 cDNA clone Z1 as the template. An Nhel restriction site and start codon ATG were incorporated into the 5' oligonucleotide primer, and a SalI site was incorporated into the 3' primer. The PCR fragment so generated was then cloned into the NheI and SalI sites of the dexamethasone-inducible expression vector pMAM-neoZ1 to generate pMAMneoAZ106-212.

Plasmid pMAMneoAZ106-212 is identical to pMAMncoZ1 except that it encodes a truncated form of AZ composed of amino acids 106 to 212 in place of amino acids 1 to 212. Transformants were selected with G418. Pools of stably transformed clones were plated (approximately 10.sup.6 cells per 100-mm-diameter Falcon dish) and incubated at 37.degree. C. for 16 hours. Dexamethasone was added to the cell culture to a final concentration of 1 .mu.M to induce AZ or AZ 106-212 protein. Cell lysates were prepared after the indicated time of treatment and assayed for ODC activity as described in Mol. Cell. Biol., 12:3556:3562 (1992).

This example illustrates method used for immunoprecipitation of .sup.35 S-labeled ornithine.

For immunoprecipitation of .sup.35 S-labeled ornithine, AZ or AZ 106-212 stably transfected HTC cells were plated at a density of about 10.sup.6 cells per 10-mm-diameter dish (Falcon) and grown overnight. The cells were treated with 1 mM dexamethasone or were left untreated for 4 h and then labeled with ›.sup.35 S!methionine for 30 min. Cells were collected in Nonidet P-40 lysis buffer (150 mM NaCl, 1% Nonidet P-40, 50 mM Tris ›pH 8.0!). Cell lysates containing 5.times.10.sup.6 acid-precipitable cpm were immuno-precipitated with anti-mouse ODC-specific polyclonal rabbit antiserum. Immunoreactive protein was collected by using Pansorbin and subjected to SDS-PAGE and autoradiography.

This example describes procedure used for preparation of recombinant DNA constructs used for the expression of fusion proteins by in vitro transcription and translation.

Amplification

The DNA sequence encoding each constituent protein fragment to be expressed was copied by PCR using oligonucleotides that incorporated a common restriction endonuclease recognition site at the point of fusion. The PCR fragments encoding the protein regions to be fused were digested with the common restriction enzyme, ligated, and reamplified using the distal 5' and 3' oligonucleotides. The 5' oligonucleotide used to copy the N-terminal element of each fusion contained a T7 RNA polymerase recognition site placed upstream of the translation initiation AUG codon. The resultant PCR product after the second round of amplification thus contained at the 5' end a T7 RNA polymerase recognition site, followed by an open reading frame encoding the fusion protein.

Plasmids encoding sea urchin cyclin B were provided by Andrey W. Murray (Univ. of Calif., San Francisco).

Antizyme .lambda. gt11

The rat antizyme .lambda. gt11 partial cDNA clone Z1 was obtained from S. Hayashi (Jekei Univ., Tokyo). Z1 has a 212 amino acid open reading frame. A numbering convention that specifies the start of this reading frame to be amino acid 1 of antizyme was used.

The sources and sequences of mouse ODC (M-ODC) and Trypanosoma brucei ODC (Tb-ODC) were as described in Mol. Cell. Biol., 14:897-92 (1994).

Fusion NAZ to Proteins To fuse the 1-97 N-terminal amino acids of antizyme (NAZ) to various proteins, Z1 DNA was amplified by PCR using a 3' oligonucleotide with a HindIII site. Similarly, using PCR, a HindIII site was introduced at the 5' ends of M- ODC, TbODC, Tb376M, Tb422M, and cyclin B amino acids 13-90. PCR amplified fragments were digested or partially digested with HindIII and ligated to make NAZ-M-ODC, NAZ-Tb-ODC, NAZ-Tb376M, NAZ-Tb422M, and NAZ-cyclin B 13-90. To make the C-terminal deletion NAZ-cyclin B 13-90, sequence changes were incorporated into the 3' oligonucleotides which created translation stop codons immediately after cyclin B amino acid 59. CAZ-MODC was made using a similar strategy by ligating antizyme cDNA encoding amino acids 106-212 and mouse ODC cDNA with a Hind III linker. To make fusion proteins of TbODC coupled to the degradation domain of p53, DNA was also amplified by PCR using oligonucleotides that contained a BamHI sites at the TbODC 3' end and at the 5' end of p53 degradation domain (amino acids 100-150). Tb376M, Tb422M and M314T were as described in Mol. Cell Biol., 12:3556-3562 (1992).

This example describes assay used for protein degradation studies.

Recombinant DNA constructs produced by PCR in Example 6 were used as templates for in vitro transcription by T7 RNA polymerase. RNA was translated in vitro using a rabbit reticulocyte lysate (Promega) in the presence of ›.sup.35 S!-methionine. The translated proteins were subjected to degradation in a rabbit reticulocyte lysate extract as described in Mol. Cell. Biol., 13:2377-2383 (1993).

Degradation was carried out at 30.degree. C. and the labeled protein remaining undestroyed was examined by SDS-PAGE and autoradiography. For densitometric analysis of autoradiograms, images were captured and digitized with a Lighttools Research gel documentation system and quantitated with PPLab Gel version 1.5 c software.

To inhibit degradation, incubation was carried out for 2 hours as above, except that 2 mM ATP-S was used in place of ATP and the ATP regenerating system. To test whether the high molecular conjugate of p53 accumulated by ATP-S is the poly-ubiquitinated form of substrate protein, p53 and its conjugate were immuno-precipitated with PAb42I (Ongogene Science) and immuno-blotted with ubiquitin antibodies (Sigma).

This example describes procedures used for study of ability of antizyme, ornithine decarboxylase or their associated complex to form a complex with the 26S proteasome.

Mouse ODC, GST, GST-AZ or GST-ODC was transcribed and translated in vitro in the presence of ›.sup.35 S!-methionine. The in vitro translated GST or GST fusion proteins were directly purified with Glutathione-Sepharose 4B. To isolate a complex of ODC/AZ, ODC was in vitro translated in the presence of .sup.35 S-methionine and GST-AZ fusion proteins were expressed in E. coli. These fusion proteins include antizyme amino acids 55-212 (G-AZ-55-212) or antizyme amino acids 106-212 (G-AZ 106-212). These proteins were coupled to glutathione-Sepharose 4B and mixed with the .sup.35 S-labeled in vitro translated ODC. The ODC/AZ complexes were eluted from the beads with 20 mM glutathione and mixed with the 26S protease, the 20S proteasome or the 11S activator, prepared from rabbit reticulocyte lysate as described in Example 7. The mixtures were then loaded on non-denaturing gel for electrophoretic separation, the 26S protease and the 20S proteasome were identified by the fluorogenic peptide overlay assay according to J. Biol. Chem., 267:22362-22368 (1992) and Coommassie blue staining. The association of ODC/AZ complex with the 26S protease was determined by autoradiography.

This example describes in vitro transcription and translation procedures used for study of fusion proteins.

Recombinant DNAs used for expression of fusion proteins by in vitro transcription and translation were made as follows. The DNA encoding each constituent protein fragment to be expressed was copied by PCR, using oligonucleotides that incorporated a common restriction endonuclease recognition site at the point of fusion. The PCR fragments encoding the protein regions to be fused were digested with the common restriction enzyme, ligated, and reamplified using the distal 5'- and 3'-oligonucleotides. The 5'-oligonucleotide used to copy the N-terminal element of each fusion contained a T7 RNA polymerase recognition site placed upstream of the translation initiation AUG codon. The resultant PCR product after the second round of amplification thus contained at the 5'-end a T7 RNA polymerase recognition site, followed by an open reading frame encoding the fusion protein.

This example describes transfection and in vivo procedures used for study of in vivo NAZ-induced degradation of fusion proteins.

Constructs TbODC-p53 100-150 and M314T-p53 100-150 were cloned into a mammalian expression vector under the control of the SV40 early promoter according to Genes &Dev.,4:764-778 (1990). The plasmids, along with pMC1 conferring neomycin/G418 resistance were prepared according to Cell, 51:503-512 (1987), were used to transfect mutant Chinese hamster ovary C55.7 cells devoid of endogenous ODC activity according to J. Biol. Chem., 257:4603-4509 (1982). Transformants were obtained by subjecting the cells simultaneously to two forms of selection: 1) with the drug G418, selective for cells expressing the neo gene encoded by pMC1, and 2) by incubation in polyamine-deficient medium, selective in the mutant ODC-deficient cells for expression of the transfected ODC gene, which is needed for polyamine biosynthesis. Pools of stably transformed clones were plated (approximately 10.sup.6 cells/100-mm Falcon dish) and incubated at 37.degree. C. for 16 hours. To elicit polyamine-mediated regulation and induce endogenous antizyme, putrescine was added to the culture medium to a final concentration of 100 um. ODC stability was assessed in some experiments by inhibiting protein synthesis with cycloheximide, 100 mg/ml of cell culture medium, and determining the rate of decay of enzymatic activity. Cell lysates were prepared after the indicated time of treatment and assayed for ODC activity as described in Mol.Cell Biol, 13:2377-2388 (1993).

This example illustrates a method for specific protection of normal growing cells from toxicity caused by cancer therapy.

Commonly used forms of cancer therapy attack growing cells, regardless of whether these are cancer cells or innocent-bystander normal cells. The resultant damage or destruction of normal cells can limit the extent of existing therapeutic treatment that is tolerable.

Cellular proteins that control cell growth include the cyclins, cyclin-dependent kinases with which these cyclins interact and additional proteins that effect the activity or stability of the cyclin-kinase pairs. A fusion protein prepared according to the invention consisting of antizyme linked to a targeting peptide that binds to one of these proteins and causes the fusion protein to be expressed in normal cells, impedes cell growth, thereby protecting the normal cells from the toxicity associated with therapy. Because growth of normal cells must be allowed to remain unimpeded between periodic applications of cancer therapeutic agents, expression of the fusion protein consisting of antizyme linked to a targeting peptide must be limited to the time of treatment. By using a gene requiring polyamine-dependent frameshifting for expression of the encoded fusion protein, it is possible to impede cell growth in normal cells at the desired time by administration of polyamines. Prior to cancer chemotherapy, patient bone marrow cells is removed by aspiration. The mixture of bone marrow cells is treated to deplete the cell population of cancer cells and to enrich the cell population with stem cells capable of regenerating normal bone marrow using methods known in the art for selective cultivation and fractionation based on physical cell characteristics or cell surface markers.

The resultant population enriched in stem cells is transfected with DNA encoding an NAZ-linker fusion protein designed to block cell growth when synthesis of the NAZ-linker is induced within cells by polyamine-dependent frameshifting.

The DNA is produced within bacteria as a plasmid containing a gene composed of NAZ with polyamine-dependent frameshifting region fused in frame to a linker with high affinity and specificity for a cyclin, for example cyclin B. A linker is designed as described in Section I, C. For cyclin B, a natural interaction partner exists in the form of a cyclin-dependent kinase and it or a portion of it capable of maintaining such interaction constitutes a natural candidate linker.

DNA is transfected into the cell population by chemical or physical means using standard methods. The cell population so transfected is then reinfused into the patient and allowed to populate the bone marrow. A portion of the normal bone marrow stem cell population thus consists of cells the growth of which is prevented or permitted depending on whether polyamines are fed or infused (cell population may not grow) or withheld (cell population may grow).

Prior to and during treatment with agents that selectively kill growing cells, i.e. cancer cells by intention and normal cells as an undesired toxic effect, patient is treated with polyamines to deplete cells expressing the NAZ-linker of cyclin B by inducing destruction of cyclin B. Cells so depleted of cyclin B are incapable of growth and replication and are thus protected from damage or death.

Polyamine treatment is repeated to coincide with each round of cancer therapy. Because multiple rounds of therapy over a protracted period are commonly employed in cancer treatment, the method described is more effective than infusing the stem cell population after therapy is completed, at which time the patient may be suffering from lethal bone marrow failure.