WO1993015769A1

WO1993015769A1 - Mutagenesis testing using transgenic non-human animals carrying test dna sequences

Info

Publication number: WO1993015769A1
Application number: PCT/US1993/001293
Authority: WO
Inventors: Jay M. Short; Patricia L. Kretz
Original assignee: Stratagene
Priority date: 1992-02-14
Filing date: 1993-02-12
Publication date: 1993-08-19
Also published as: EP0671955A1; EP0671955A4; CA2130081A1

Abstract

The invention describes an assay for monitoring and assessing the mutagenic potential of agents which involves creating and using transgenic non-human organisms carrying a test DNA sequence or sequences that can be examined for spontaneous mutations or induced mutations following exposure to one or more suspected mutagenic agents. In a preferred assay system, the mutagenized test DNA sequences are selectable, thereby increasing the sensitivity of the assay.

Description

MUTAGENESIS TESTING USING TRANSGENIC NON-HUMAN ANIMALS CARRYING TEST DNA SEQUENCES

Technical Field

This invention relates to transgenic animals and to tests for monitoring mutagenic agents in live animals. More specifically, this invention relates to the creation of transgenic non-human animals carrying test DNA sequences and to methods for monitoring and assessing the mutagenic potential of agents by

exposing the transgenic animal to one or more

suspected mutagens, and optionally recovering the test DNA sequence, and examining the test DNA sequence for mutations. Novel methods for increasing the

efficiency of test DNA sequence recovery and rapid analysis of specific test DNA mutations are also described.

Background

Various agents, such as radiation, ultraviolet light, synthetic chemicals, natural substances, and aberrations in genetic replication and repair can produce mutations in DNA. The results of a

representative study indicate that as many as 60% of the cancers that develop in women and as many as 40% of those that develop in men result from avoidable exposure to mutagens from dietary intake. Vuoto et al, Environ. Mutagen, 7:577-598 (1985). Exposure to environmental mutagens such as nitroaromatic compounds found in automobile exhaust, chlorination by-products used in drinking water, and acrylamide and

formaldehyde used extensively in industrial

laboratories is also of major concern. Quantitative measurement of the effect of suspected mutagens is essential to control exposure to harmful agents. Additionally, whenever a new chemical, drug, or food additive, for example, is to be taken from the

laboratory to the marketplace, it must be tested for its toxicity and cancer-causing potential. As a result, significant effort has gone into the

development of assays that detect the mutagenic potential of various compounds.

Existing tests that assess the mutagenic

potential of substances focus either on alterations of DNA in cultured cells or bacteria or alterations in the health of test animals. However, few tests that monitor alterations in DNA actually expose live animals to the agent to be tested. This is because it is very difficult to rapidly monitor alterations in the genetic code simultaneously in many different organs. Tests to detect these mutations must be very sensitive. They must be able to detect a single mutation amongst millions of normal genetic units.

The difficulty of this task currently makes this approach for live animal studies prohibitively

expensive as well as time intensive. Therefore, most current live animal genotoxicity tests use disease formation or large scale chromosomal alterations as an assay for gene alteration.

DNA alterations that are caused by potential mutagenic agents have generally been approached by performing studies on procaryotic or eukaryotic cells in culture (in vitro tests). The well-known Ames' test uses a special strain of bacteria to detect these mutations. Ames et al, Proc. Nat. Acad. Sci., 70:782-86 (1973). This test and many analogues that use other types of bacterial or animal cells permit the rapid screening of very large numbers of cells for the appearance of an altered phenotype. The appearance of this altered phenotypic trait reflects the occurrence of a mutation within the test gene. These tests are, however, insensitive to or nonspecific for many mutagens that result from metabolic activation of the agent being screened. Although attempts have been made to increase their sensitivity and specificity by activation of such metabolites with liver and other extracts it is noted that, for instance, the

metabolites produced by these extracts are often not present at the same concentrations as in the live tissues of an animal. Metabolites that are only produced in other organs are not detected at all.

Eukaryotic cell lines have also been used to detect mutations. E.g., Glazer et al, Proc. Natl. Acad. Sci. USA, 83:1041-1044 (1986). In this test a target test gene, the amber suppressor tyrosine tRNA gene of E. coli in a bacteriophage shuttle vector, was integrated into a genomic host mammalian cell line by DNA transfection of cultured cells in vitro. After exposing the host cell line to putative mutagenic agents, test genes were re-isolated, propagated in bacteria, and analyzed for mutations. Because the host is only a mammalian cell line and not a live animal, the test is incapable of accurately monitoring mutagenic metabolites of the agent being tested that are only produced at the appropriate concentrations by differentiated cells or the tissue of live animals.

A two year study by the NIH concluded that data obtained from four different prokaryotic and

eukaryotic in vitro assays had only a 60% concordance with whole animal carcinogenicity studies. Tennant et al, Science, 236:933-941 (1987). The study suggests that the high rate of error may result from potential variation in genetic susceptibility between in vitro systems and whole animals. For example, metabolites, frequently involved in activation of promutagens, are not present in in vitro systems, allowing mutagenic potential to go undetected. In addition, differences in DNA repair mechanisms between prokaryotes and eukaryotes may account for some discrepancies in results.

Test genes and large scale screening assays used for in vitro assays are not available for live animal studies. Short of relying on long term animal studies that detect phenotypic changes that require a long time to be identifiable, such as tumors, organ

failure, coat color, etc., current tests do not provide a means for monitoring organ-specific

mutations of DNA. Hence, there exists a need for a system that places a test DNA sequence within an animal and is subsequently assayed on a large scale for mutations. There also exists a need for methods that detect mutations caused by chemical metabolites of the agent being tested. To be most effective the system needs to be capable of monitoring genetic changes in as many tissues of an animal and as easily, rapidly, and inexpensively as possible.

The present invention, providing novel transgenic non-human organism and methods utilizing such

organisms for mutagenesis testing, satisfies these needs. More specifically, the present invention provides a sensitive screen for the mutagenicity of suspected agents and permits the monitoring of the mutagenic effects of such agents and the mutagenic effects of the metabolites of such agents.

Additionally, the invention can permit the

identification of the nature of the mutation, e.g., DNA transition, transversion, deletion, or a point or frameshift mutation. Further, the methods of the invention offer the significant advantage of being rapid to perform, thus permitting the identification of potential mutagens appreciably before other tests can be completed, and is inexpensive relative to other whole animal tests. And, the present invention substantially reduces the number of organisms which must be used for mutagenesis testing.

Summary of the Invention

The present invention contemplates a method for assaying the mutagenic potential of an agent. The method generally comprises administering a

predetermined amount of the agent to an organism containing cells having a genome characterized by the presence of a target gene system containing a test DNA gene sequence that is capable of detection by bioassay in a host cell upon mutation of the test DNA sequence. After a predetermined exposure time period, a

predetermined amount of the target DNA system is then recovered from cells harvested from the exposed organism. The recovered target gene system is then introduced into and expressed in a restriction system deficient host cell, whereupon mutation of the target gene system is determined in bioassay.

In one embodiment, the invention describes a mutagenesis testing method comprising:

(a) exposing a transgenic non-human organism to a test agent, wherein the transgenic organism comprises somatic and germ cells containing a test DNA sequence that is capable of detection in bioassay by expression of a test DNA sequence gene product; and

(b) determining by bioassay in a host cell the frequency of mutation of the exposed test DNA sequence. The host cell contains a reporter gene transcription unit controlled by a promoter that is regulated by the test DNA sequence gene product. In preferred embodiments, the reporter gene product confers a selective growth advantage to the bioassay system, thereby allowing selectable detection of the mutated test DNA sequence over non-mutated test DNA sequences in the target gene system.

In a preferred embodiment, the organism is a transgenic plant or animal, such as a transgenic fish or rodent, and preferably is a transgenic rat or mouse.

Preferred methods have been developed which use a target gene system comprised of a set of two genes, a test gene and a reporter gene. The test gene is incorporated into an organism or its somatic and germ cells to screen for compounds having mutagenic, carcinogenic or teratogenic activity. The reporter gene can be present in a separate assay system such as a prokaryotic or eukaryotic host cell, or it can be incorporated in the organism with the test gene.

Exposure of the organism or its cells to compounds having any of these activities causes mutations resulting in alterations in expression of the test gene. Mutations in the test gene are measured by detecting reporter gene expression, which is affected by test gene expressions. Both the test gene and the reporter gene constructs are each operatively linked to a promoter, preferably a prokaryotic promoter.

Also described are transgenic non-human organisms containing a target gene system of this invention for practicing the mutagenesis testing methods described herein. Further described are host cells that contain a reporter gene transcription unit for detecting a mutated test DNA sequence in a bioassay of this invention.

This method has several advantages over the prior art methods of screening for compounds having mutagenic or teratogenic activity. The most

significant advantage is the ease in detection and decrease in number of false positives. Although the mutation of genes encoding reporter proteins has previously been used to assay for mutagenic activity, the mutational event resulted in the protein not being expressed. Detecting a single cell, or even a few cells, not expressing a protein, while surrounded by cells which express the protein, is difficult, tedious, and subject to a high percentage of error. In contrast, in the present method, the mutational event ultimately results in the expression of a reporter molecule which would otherwise not be

expressed, and which is readily detected. Other permutations of this improved method are described further herein.

Brief Description of the Drawings

In the drawings, forming a portion of this disclosure:

Figure 1 illustrates the sequence of process steps for performing the invention.

Figure 2 illustrates an alternative method for recovering the transgenic test DNA sequence.

Figure 3 illustrates in two panels (3A and 3B) a

DNA fragment containing the mcrB gene and the relative location of various restriction sites therein.

Figure 4 illustrates a schematic depicting the construction of plasmid pBlue MI-.

Figure 5 illustrates a schematic depicting the construction of plasmid pLacI^q.

Figure 6 illustrates a schematic depicting the construction of plasmid plnt.1.

Figure 7 illustrates a schematic depicting the construction of plasmid pPreLacIqZ. Figure 8 illustrates in two panels (Figure 8A and Figure 8B) a schematic depicting the construction of phage Lambda LIZ Alpha.

Figure 9 is a graph that illustrates the mutation frequency for spontaneous (closed symbols) and induced (open symbols) mutations as described in Example 8.

Figure 10 is a schematic in two panels (Figure 10A and Figure 10B) illustrating a gene activating selectable system as described in Example 9. Figure 10A shows the operation of the system under the condition where a wild-type (non-mutated) lacI gene present on the Lambda LIZ Alpha vector (upper vector) represses the reporter gene transcription unit (lower vector), thereby inactivating groE gene expression and inhibiting plaque formation. Figure 10B shows the operation of the system under the condition where a mutated lacI gene present on the Lambda LIZ Alpha vector (upper vector) does not represses the reporter gene transcription unit (lower vector), thereby activating groE gene expression and allowing plaques to form.

Figure 11 is a schematic in two panels (Figure 11A and Figure 11B) illustrating a gene inactivating selectable system as described in Example 10. Figure 11A shows the operation of the system under the condition where a wild-type (non-mutated) lacI gene present on the Lambda LIZ Alpha vector (upper vector) represses the reporter gene transcription unit (lower vector), thereby activating S⁵ gene expression and inhibiting plaque formation. Figure 11B shows the operation of the system under the condition where a mutated lacl gene present on the Lambda LIZ Alpha vector (upper vector) represses the reporter gene transcription unit (lower vector), thereby

inactivating S⁵ gene expression and allowing plaques to form.

Detailed Description of the Invention

The present invention contemplates engineered somatic and germ cells of an organism, such as an animal, animal embryos or differentiated animals, having a genome characterized by the presence of a target gene system useful for the testing of mutagenic potential of a suspected mutagen. Preferably the target gene system is recoverable from the test organism's genome by nucleotide sequences that define an excision means flanking the target gene system, such as an excisably integrated genetic element. The target gene system comprises a test gene

(transcribable, and preferably translatable DNA sequence). Preferably, the test gene is operatively linked to prokaryotic expression signals, such as a promoter, ribosome binding site, stop codon, and the like for expression of a test gene product. In a preferred embodiment, the recoverable target gene system is included within excisably integrated lambda phage DNA. However, those skilled in the art will appreciate that the target gene system need not be included within lambda phage DNA to be recoverable for purposes of the present invention. For example, the target gene system can be present within a plasmid, cosmid, filamentous phage, or present in the genome of the animal and be recoverable.

As used herein, a "test gene" refers to a

sequence of nucleotides which are the direct target for mutation by a suspected mutagen. A "reporter gene" refers to a nucleotide sequence which expresses a detectable phenotype in an assay system, and the expression of which reporter gene is controlled by the test gene. As used herein, unless specifically stated otherwise, "animal cells" will be used to include cells in cell cultures, embryos, and differentiated animals. As also used herein, "mutagen" will be used to include toxins, carcinogens, teratogens, and other agents which alter DNA or RNA sequence or expression, unless stated otherwise.

A recombinant DNA molecule (rDNA) of this

invention (subject rDNA) is used for preparing a non-human transgenic organism of this invention and contains a target gene system as described herein. A target gene system preferably is operatively linked to a lambda phage and can be comprised of any of a variety of test genes (test DNA sequence) whose transcription results ultimately in a detectable phenotype or genotype where a mutation in the

nucleotide sequence of a test gene measurably alters the detectable phenotype or genotype. Typical test genes include genes that confer drug resistance or other selective advantage, or genes whose expression alters the expression of a second reporter gene.

Exemplary drug resistance genes confer resistance to ampicillin, kanamycin, chloramphenicol and the like.

A test gene can be selected from the group of nucleotide sequences which encode regulatory molecules that bind to a sequence controlling reporter gene expression. These can be repressers or other

regulatory molecules, including anti-sense RNA. A preferred test gene is a repressor or activator gene whose expression product directly alters the

detectable expression of a reporter gene. Exemplary is the repressor protein encoded by the lacl gene, and genetic variants of the lacl gene that function to block transcription of the beta-galactosidase gene (lacZ) by binding to the operator region of the lacZ gene's expression signals. In this embodiment, lacZ is a reporter gene. A preferred lacl test gene is the laclq variant that expresses eight- to ten-fold elevated levels of repressor protein and more tightly represses expression of the lacZ reporter gene when under the control of a lac operator.

The inactivation of the lac repressor gene by a mutagenic event causes the transcription and

translation of a defective repressor protein that is no longer able to repress expression of the lacZ

reporter gene encoding beta-galactosidase. Alteration of/the operator region for the reporter gene in a manner that prevents binding of the repressor protein produces the same effect. Derepression of the

reporter gene can then be monitored by assaying for defined functions of the gene product.

In one embodiment, the test gene of the target gene system is operatively linked to a reporter gene, i.e., both test and reporter genes are linked on the same DNA molecule. In other embodiments, the reporter gene is present in a host cell of the assay system and is regulated by the expression of the test gene.

Thus, where the reporter gene transcription unit is present in the host cell for bioassay, such as the E. coli into which the recovered test DNA sequence is to be introduced, the test DNA sequence gene product acts in trans to influence the expression of the reporter gene product.

A reporter gene provides a means for detecting mutations in the test gene. A reporter gene is a gene that encodes a detectable phenotype or genotype and whose expression is under the control of the test gene. Typically, the reporter gene is the final

(endpoint) gene in a biochemical pathway initiated or regulated by the test DNA gene product. The selection of reporter genes is based on the following criteria: (i) the gene product should provide a simple and sensitive detection system for its quantitation, and (ii) non-transformed cells should have a low constitutive background of gene products or activities that will be assayed. In certain embodiments, a reporter is lethal in the assay system, and in other systems the reporter is not lethal. Examples of a non-lethal reporter genes are genes which confer the ability for growth,

amplification or replication of the reporter gene, or cells harboring the reporter gene.

Any of a variety of genes can function as the reporter gene according to the present invention so long as the expressed reporter gene product is

detectable. Genes that encode detectable phenotypes include drug resistance markers, enzymes whose

activity produces a detectable reaction product and the like. Candidate enzymes include beta-galactosidase (Norton et al, Mol. Cell. Biol, 5:281- 290 (1985), peroxidase and luciferase (de Wet et al, Mol. Cell. Biol, 7:725-737 (1987). A preferred reporter gene is the E. coli beta-galactosidase gene (lacZ).

A preferred lacZ gene is one that utilizes alpha complementation, as described herein, whereby

functional lacZ activity requires the association of the alpha portion of the lacZ gene product with the complementary portion of the lacZ referred to as the lacZ▴M15 gene product.

The phenotype produced by the reporter gene can result in detection based on a phenotypic selection such as a colorimetric selection, growth selection, enzymatic activity, and the like. Reporter genes which encode enzymes, antigens or other biologically active proteins which can be monitored easily by biochemical techniques are preferred.

In one embodiment, a reporter gene is expressed when the test gene is not mutated, and is not

expressed when the test gene is expressed as a

functional protein.

In another embodiment, a reporter gene can be expressed only when the test gene is mutated.

Preferably, the reporter gene confers a growth

advantage to the assay system containing the mutated test gene, and the reporter gene thereby provides a mechanism to select for a detection of the mutation event. A growth advantage, in this context, is provided to the reporter gene, whether it replicates (grows) and is amplified in a host cell or in the form of an autonomous genetic element, such as the

exemplary phage. Thus, in this embodiment, the occurrence of a mutation in the test gene is selected for, thereby increasing the efficiency of the system to detect mutagenic activity. Such a system is referred to herein as a "selectable system".

In one example of a selectable system, a reporter gene is under the transcriptional control of an operator that is repressed by a test gene product, i.e. the test gene product is a repressor. Upon mutation, the test gene loses its repressor function, and the reporter transcription unit is expressed providing a growth advantage to the assay system. A preferred system uses a lac operator controlling reporter gene transcription, and uses a lac repressor such as lacl or lacl^q as the test gene.

Preferred reporter genes that can be used to confer a selective growth advantage include groE and lambda S⁵ as described herein, beta-galactosidase- based genes such as lacZ or alpha-lacZ, and E. coli gene that are essential for lambda DNA replication, but dispensable for E. coli, such as grpD, grpE or cro. See, for example, "Lambda II" Hendrix et al, eds., Cold Spring Harbor Press, 1983, p.147.

Additional selectable genes which confer a useful growth advantage as reporter genes are amino acid genes, and tRNA genes

The groE gene is an E. coli gene that is required for lambda phage particle morphogenesis. The groE operon is actually two closely linked genes that encode the GroEL and GroES proteins required for lambda phage head assembly. Mutations in groEL or groES genes block lambda head assembly at an early stage. A mutant E. coli host is utilized for reading the mutagenesis assay that has a deficiency in groEL, groES, or both, where defects in both is designated groESL. The wild type groE, the groEL or groES gene, or both (groESL), is/are supplied by the reporter gene's transcription unit, and upon expression confers the ability (selective advantage) to assemble phage particles.

Preferred GroE systems are described in Examples 9 and 11. In those systems, the expression of a groE reporter gene product is activated by the introduction of a mutated lacl test gene product into a host reporter system, which allows bacteriophage plaques to form, thereby indicating the presence of the mutated test gene in the host. Such a system is referred to as an "activating reporter gene system" because the activation of the reporter gene produces the

detectable event.

Another preferred system utilizes a mutated lambda S gene product designated S⁵. The S⁵ gene product prevents plaque formation by inhibiting the formation of a functional inner membrane pore through which phage particles can extrude during

morphogenesis. The efficient expression of the S⁵ phenotype requires that the reporter gene be expressed in E. coli which is supF. The gene is referred to as a dominant negative inactivating gene because its effect is dominant, not recessive, and because its expression is a negative marker, i.e., it inhibits plaque formation.

The amino acid residue sequence of a lambda S gene is coded for in the wild-type lambda genome at nucleotide base residues 45186 to 45006. The complete nucleotide sequence of wild-type lambda is well known, and is also described in "Lambda II" by Hendrix et al., Cold Spring Harbor Press, 1983. The S⁵

mutations have been identified to be the substitution of an adenine (A) for a thymidine (T) at nucleotide 45214, and the substitution of an adenine (A) for a cytosine (C) at nucleotide 45310. Genes having the S⁵ mutations can thus readily be prepared by synthetic methods such as oligonucleotide synthesis and

hybridization of the synthesized oligonucleotides to form a complete gene, as is well known in the art. A preferred S⁵ gene for use in the invention has the nucleotide sequence shown in SEQ ID NO 18.

A preferred S⁵ system is described in Example 10.

In that system, the expression of a S⁵ reporter gene product is inactivated by the introduction of a mutated lacl test gene product, which allows

bacteriophage plaques to form, thereby indicating the presence of the mutated test gene. Such a system is referred to as an "inactivating reporter gene system" because the inactivation of the reporter gene produces the detectable event. The inactivating reporter gene system utilizes competing transcripts as described further in Example 10 as exemplary. Any other dominant negative inactivating gene can be utilized in a reporter gene transcription unit according to the system described herein for the S⁵ gene.

The test gene is operatively linked to expression signals to facilitate the rapid detection of mutations by the present invention. The type of expression signals depends upon the host cell in which the reporter gene is bioassayed. A preferred host is a prokaryotic cell, and therefore the reporter is preferably under the control of prokaryotic expression signals. Upon recovery of the target gene system, e.g., by one of the various excision means described herein, the test gene system is introduced into a prokaryotic expression system, such as a bacterial cell lawn, so that dilutions of the test genes can be expressed and thereby observed (reported) to quantify the extent of test gene mutation.

Insofar as the expression of a test gene is measured in a prokaryotic expression system such as a bacterial cell, it is understood that the mutations can occur either in the structural portions of a test gene or in the expression control signals the test gene, e.g. procaryotic expression promoter. Thus in preferred embodiments the test gene comprises a lacl, lacl^q, lacl^sq or lacl^cι gene and includes a lacl

promoter region.

The bacterial lac operator-repressor system is preferred because it is one of the most basic and thoroughly studied examples of a protein-nucleic acid interaction that regulates transcription of a gene, as described by Coulondre et al, Mol. Biol., 117:577 (1977), Miller, Ann. Rev. Genet., 17:215 (1983); and in "The Operon", Miller et al, eds., Cold Spring

Harbor, 1980. This bacterial regulatory system has been transfected into mammalian cells and expression detected by addition of an inducer, isopropyl beta-D thiogalactoside (IPTG), as reported by Hu et al, Cell, 48:555 (1987), and Brown et al, Cell, 49:603 (1987). An important difference between previous uses of the lac operator-repressor system and the present method is that mutation rather than induction is used to derepress the reporter genes to express protein whose function is solely to serve as an indicator. Another difference is that, in the preferred embodiment, the target gene system is excisable as an infectious lambda phage.

A lambda phage of this invention comprises a target gene system that is excisably-integrated into the genome of an animal cell or embryo. By excisably-integrated is meant that the lambda phage comprises excision elements operatively linked to the genome that provide a means to conveniently remove the test gene system from the animal, cell or embryo genome subjected to mutagenesis conditions for the purpose of assessing the possible occurrence of mutation.

Exemplary excision elements (excision means) are nucleotide sequences flanking the target gene, and if present the reporter gene, and other elements of the target gene system, that allow site-specific excision out of the genome to which the target gene system is operatively linked (integrated). Excision elements can be site-specific restriction endonuclease

nucleotide sequences, or can be other genetic elements that facilitate site-specific excision. Preferred excision element nucleotide sequences are lambda cos sites, flp recombinase recognition sites, loxP sites recognized by the Cre protein, and the first and second halves of the filamentous bacteriophage (M13, ff or f1) origin of replication (referred to generically as an fl bacteriophage origin of

replication), all of whom are described more fully herein.

Preferred are cos site excision elements because of the convenience and the efficiency of excision of the genes contained between cos site nucleotide sequences when utilizing lambda bacteriophage in vitro packaging extracts as described herein. In one embodiment using the lambda packaging extract, it is considered useful to boost the packaging reaction by the repeated addition of aliquots of packaging extract to the packaging reaction, as the extract becomes depleted during the packaging reaction, which can be remedied by multiple additions of extract.

The excision elements of a target gene system confer the ability to readily recover the target gene system from the mutagen exposure conditions to the prokaryotic expression medium in which the reporter gene is measured.

An exemplary and preferred test gene system is the Lambda LIZ Alpha vector described herein in which a lacl^q test gene is operatively linked to the alpha-complementation-based lacZ alpha gene, where both test and reporter genes are under the control of

prokaryotic expression signals, namely, lacl promoter and lacZ promoter/operator sequences.

This preferred system further contains nucleotide sequences operatively linked to the test gene that define a prokaryotic origin of replication, a

selectable marker (amp^R) and a filamentous phage origin of replication such that the test gene can readily be transformed into a "f1-type" nucleic acid sequencing vector for rapid determination of the nature of the mutation in the test gene. This latter feature is provided according to the teachings of Short et al., Nucl. Acids Res., 16:7583-7600 (1988), where the terminator and initiator domains of the f1 intergenic region are separated and flank the test gene sequences of this invention to be recovered and sequenced.

A promoter is a sequence of nucleotides that forms an element of a structural gene transcriptional unit which controls the gene's expression by providing a site for RNA polymerase binding resulting in the initiation of the process of transcription whereby a gene is transcribed to form a messenger ribonucleic acid (mRNA) molecule. For the repressor protein to control the expression of a reporter gene, the

operator sequence has to be built into the reporter gene at the location between the transcription

initiation site and the initiation codon ATG.

An operator is a sequence of nucleotides that forms a site for specific repressor binding. Thus, operators are specific for a particular repressor.

A repressor binding site is considered specific if the equilibrium binding constant for repressor binding to the operator is greater than 10^-8 molar (M), preferably greater than 10^-9 M, and more

preferably greater than 10^-10 M. The equilibrium binding constant for a repressor binding to an

operator can readily be measured by well known

equilibrium dialysis methods, or in a nitrocellulose filter binding assay where repressor is immobilized on nitrocellulose and ³²P-labeled operator-containing DNA segment is presented in solution for binding to the immobilized repressor. See, Miller "Experiments in Molecular Genetics", p367-370, Cold Spring Harbor Laboratory, New York, 1972.

The operator for the lac repressor has been well characterized and is used as exemplary herein. See Miller et al, in "The Operon", Cold Spring Harbor Laboratory, New York (1980), for a detailed study. Alternative nucleotide sequences have been described for a lac repressor operator that specifically binds to repressor. See, for example, the description of numerous lac operator variants and the methods for characterizing their repressor-binding activity reported by Sartorius et al, EMBO J., 8:1265-1270 (1989); and Sadler et al, Proc. Natl. Acad. Sci. USA, 80:6785-6789 (1983). Any nucleotide sequence that binds lac repressor specifically can be used in the present invention, although wild type and optimized "ideal" operators are preferred and used as exemplary herein. The original lac operator sequence (5'- GGAATTGTGAGCGGATAACAATCC-3'; SEQ ID NO 1), or a mutant lac operator which binds repressor eight times tighter and has the sequence (5'-ATTGTGAGCGCTCACAAT-3'; SEQ ID NO 2), are preferred for use in vector construction. Two preferred optimized operators derived from the lac operon include the nucleotide sequences as follows:

(SEQ ID NO 3) 5'-TGT GGA ATT GTG AGC GCT CAC AAT TCC ACA-3'

(SEQ ID NO 4) 5'-ATT GTG AGC GCT CAC AAT-3' Operators function to control the promoter for a structural gene by a variety of mechanisms. The operator can be positioned within a promoter such that the binding of the repressor covers the promoter's binding site for RNA polymerase, thereby precluding access of the RNA polymerase to the promoter binding site. Alternatively, the operator can be positioned downstream from the promoter binding site, thereby blocking the movement of RNA polymerase down through the transcriptional unit.

Multiple operators can be positioned on a rDNA molecule to bind more than one repressor. The advantage of multiple operators is several fold.

First, tighter blockage of RNA polymerase binding or translocation down the gene can be effected. Second, when spaced apart by at least about 70 nucleotides and typically no more than about 1000 nucleotides, and preferably spaced by about 200 to 500 nucleotides, a loop can be formed in the nucleic acid by the

interaction between a repressor protein bound to the two operator sites. The loop structure formed

provides strong inhibition of RNA polymerase

interaction with the promoter, if the promoter is present in the loop, and provides inhibition of translocation of RNA polymerase down the

transcriptional unit if the loop is located downstream from the promoter.

In preferred embodiments, a vector contemplated by the present invention includes a procaryotic replicon, i.e., a DNA sequence having the ability to direct autonomous replication and maintenance of the recombinant DNA molecule extrachromosomally in a procaryotic host cell, such as a bacterial host cell, transformed therewith. Such replicons are well known in the art and include OriC as described herein. In addition, those embodiments that include a procaryotic replicon may also include a gene whose expression confers a selective advantage such as amino acid nutrient dependency or drug resistance to a bacterial host transformed therewith as is well known, in order to allow selection of transformed clones. Typical bacterial drug resistance genes are those that confer resistance to ampicillin as used herein, tetracycline, kanamycin, and the like.

Those vectors that include a procaryotic replicon may also include a procaryotic promoter capable of directing the expression (transcription and translation) of the gene transformed therewith. A promoter is an expression control element formed by a DNA sequence that permits binding of RNA polymerase and transcription to occur. Promoter sequences compatible with bacterial hosts are typically provided in plasmid vectors containing convenient restriction sites for insertion of a DNA segment of the present invention. Bacterial expression systems, and choice and use of vectors in those systems is described in detail in "Gene Expression Technology", Meth.

Enzvmol., Vol 185, Goeddel, Ed., Academic Press, NY (1990). Typical of such vector plasmids are pUC8, pUC9, pBR322 and pBR329 available from Bio-Rad

Laboratories, (Richmond, CA) and pPL and pKK233-2, available from Pharmacia, (Piscataway, NJ), or Clone Tech (Palo Alto, Ca).

Insofar as a preferred embodiment involves the use of a lambda bacteriophage system for excision of the test DNA sequence, it is seen that the detectable end point for the bioassay can be either lytic plaque formation or a lysogenic phenotype, depending on the manner in which the genetics of the test DNA sequence and reporter gene are designed. Furthermore, the resulting lysogen or lytic plaque can be designed to exhibit a color screen as the detection means, as shown herein.

In one embodiment, the combined phenotype for the reporter gene of a growth advantage and color

indication is contemplated. A representative combined phenotype reporter gene system is one where the test gene controls two separate reporter genes; one

providing a colorimetric phenotype, such as lacZ, and one providing the growth advantage, such as an

activating reporter gene system. An exemplary

combined phenotype system is the groE system described in Example 9. The system provides a

particular advantage in circumstances where there is a leaky gene. For example, the system can exhibit a low level of false positives in the form of a growth advantage to a reporter gene, but that false positive is determinable as a false positive based on the color phenotype.

Other formats for dual (combined) reporter gene phenotypes can be readily designed by one skilled in the art, and are therefore contemplated by the present invention.

In alternate embodiments, the reporter gene in the host cell can be present in a variety of forms. The reporter gene can be present as a host cell genomic element, or as a transcription unit on a phage genome within the cell or on a plasmid, such as an F' plasmid, within the cell.

Still further embodiments contemplate methods for providing further regulation of the expression of the reporter gene, including methods for preventing expression of the reporter gene transcription unit until the test gene is introduced to methods involving the tight regulation of the reporter gene

transcription unit.

Thus, the invention contemplates the use of

"reversible" reporter gene transcription units in which the reporter structural gene is reversed within the reporter gene transcription unit relative to its promoter such that it cannot be expressed in the reverse order. Upon introduction of the test gene, the reporter structural gene flips around to position in the correct orientation and therefore is under expression control of the reporter gene transcription unit's promoter.

Reversibility of the structural gene can be accomplished in a number of ways. For example, the structural gene can be flanked by nucleotide sequences defining flp sites which are activated by flp

recombinase. Upon introduction of flp recombinase into the host cell, the reporter structural gene flips over and can be transcribed by the transcription unit. Alternatively, the structural gene can be flanked by nucleotides defining the Cre-lox system, and can be flipped upon introduction of means for initiating Cre-lox mediated recombination. Still further, one can flank the structural gene by the att nucleotide sequence of lambda integrase, and flip the structural gene by introduction of lambda integrase.

Another mehcanism for closer regulation of the reporter gene transcription unit is to include a second repressor binding site to the transcription unit which is regulated independently from the test gene product. The second repressor is inducible, and derepression of the second repressor is controlled by the addition of an inducer of the second repressor which is selected to act independently of the test DNA gene product.

An additional mechanism for regulation of

reporter gene transcription is to include

transcription or translation modifiers in the reporter gene transcription unit which repress translation until the test DNA gne is introduced into the host cell. These modifiers include the addition of a poly A transcription terminator after the reporter

structural gene to stabiliize the reporter gene transcripts. Other transcriptional modifiers include the introduction of various ribosome binding sites known to effect the efficiency of transcription, or the alteration of the nucleotide sequence around the ATG start site for translation to alter the effeciency of translation.

The use of host cell strains in which the

endogenous levels of protease are inhibited would be useful for boosting the level of expression of the reporter gene product. An E. coli host cell having a mutation in the Ion gene or hfl gene are preferred examples of mutations that would desirably reduce the amount of endogenous protease in the host cell.

Still further, the invention contemplates multi-level transcription control such as to regulat the reporter gene by a second transcription unit which, in turn, is regulated by the test DNA gene product. For example, the test DNA gene product expresses lacI, and the host cell contains a first transcription unit having the lac operator to which the lac repressor binds. This first transcription unit expresses, for example, the T7 polymerase gene product, which, upon expression, binds to the T7 polymerase operator that is controlling expression of the reporter gene

transcription unit. Other permutations are readily apparent.

In another embodiment, the invention contemplates the DNA molecules, plasmids, nucleotide sequences and the like that are utilized in the test DNA sequences and in the reporter gene transcription units described herein. Also contemplated are host cells containing the DNA molecules of this invention.

Transgenic Organisms and Their Use

In one embodiment, the present invention provides novel transgenic non-human animals and methods for monitoring the mutagenic effects of potential

mutagenic agents. In accordance with this invention, at least one copy of at least one target test DNA sequence is introduced into cells of a non-human organism thereafter bred to produce test systems.

Preferably, substantially all of the cells will contain the test DNA sequence. The test transgenic organism is then exposed to an agent suspected to be mutagenic and the test DNA sequence may be

subsequently recovered from individual tissues of the transgenic organism. The test DNA sequence may be transferred into a host cell containing a reporter gene transcription unit, although such recovery and transfer is not requisite, and assayed for mutations, allowing rapid examination of multiple tissue specific genetic mutations. Other methods to monitor mutations in the test DNA need not rely on rescue and involve either direct examination of the test DNA in situ, PCR amplification of the test DNA, examination of RNA transcription products of the test DNA or protein translation products of said RNA, or effects of said proteins or substrates for said proteins.

Theoretically, any organism suitable for

mutagenic testing may be used as the starting

organism. The organism can be plant or animal, and a preferred and exemplary embodiment is a non-human mammal, preferably a rodent. Although the present methods are readily adaptable to other species, the discussion will refer to mammals as exemplary of the invention.

In order to allow for ubiquitous insertion of the novel test sequence, single cell animal embryos are harvested, although there may be other cells

facilitating the uptake and ultimate ubiquitous presence of the marker DNA in cells of a

differentiated animal.

In accordance with the invention, any number or variety of sequences coding for a phenotype or

genotype that is detectable upon mutation may be used for introduction into the transgenic non-human mammals of the invention. Vectors capable of facilitating the recovery of the test DNA sequence from the host mammal cells, and capable of allowing replication and

expression of the sequence in a bacterial host, are preferably used as carriers for the target test DNA sequence. Accordingly, the construct for such a vector and insert preferably should contain regions for excision from the mammal host genome, and regions that allow replication in a bacterial host cell, as well as regions that permit expression and assay of the test DNA sequence. If integration into the host genome is not required, desired regions that allow for replication of the test DNA sequence in the animal host cells should be present. Elbrecht et al, Mol. Cell. Biol., 7:1276-1279 (1987).

Further, in accordance with the invention, the test DNA sequence is introduced into the host mammal, preferably (but not necessarily) at the single-cell embryo stage, so as to provide the stable presence of the test sequence throughout cells of the

differentiated animal. The use of chimeric animals is also contemplated herein. Typically, this involves the integration of the test DNA sequence into the mammal host genome, although methods that allow the test sequence to be stably and heritably present through the use of autonomously replicating vectors will also be useful. Elbrecht et al, Mol. Cell.

Biol., 7:1276-1279 (1987). At the cellular level, this may be accomplished using the techniques of microinjection, electroporation, dielectrophoresis or various chemically mediated transformation techniques, all of which are well known in the art. At the differentiated tissue level, other techniques may be necessary. The copy number of the test DNA sequence in the genome of a transgenic mammal can be varied to

increase the number of targets for the suspected mutagen. Similarly, the copy number of the reporter gene in the bioassay host cell can be varied to optimize the expression of the reporter gene product. In a related embodiment, the promoter strength can be varied to optimize the expression of the reporter gene product.

Following the introduction of the test DNA sequence and integration into the genome or cell, the transgenic cell or cells must be allowed to

differentiate into a whole organism. This may be accomplished, for example, by embryo implantation into pseudopregnant females, or by other techniques

allowing maturation of transgenic embryos. Once such maturation and differentiation has occurred, the animal is assayed for the presence of the test DNA sequence. Typically this involves removing small portions of tissue from the animal and using standard DNA hybridization assay techniques to detect the presence of the test DNA sequence.

Transgenic animals carrying the test DNA sequence are thereafter bred and offspring carrying the test DNA sequence my be selected for mutagenesis testing. In accordance with the invention, the selected

transgenic mammals are exposed to agents or substances in question under appropriate conditions. Such conditions will depend, for example, on the nature of the agent or substance, the purpose of the mutagenesis study and the type of data desired.

After exposure of test transgenic animals to the agent to be tested under the desired conditions, desired tissue may be removed from the test animal. Because in the preferred embodiment the test DNA sequence is present in essentially all tissues, the tissue type tested is not limited by the process of insertion of the test DNA sequence. Any desired tissue may be removed and assayed at the DNA, RNA, protein or substrate/product level, by various methods including, but not limited to, in situ hybridization to the DNA or RNA, PCR, protein or enzymatic assays (PCR Protocols, A Guide to Methods and Applications, eds. Innis et al, Academic Press, Inc., 1990; Maniatis et al, Molecular Cloning, A Laboratory Manual, Cold Spring Harbor, New York 1982).

Alternatively, genomic DNA may be purified from the tissue. The target test DNA sequence which is integrated may then be rescued (recovered) from the total genomic DNA of the host. This may be

accomplished by excising it from the host genome or by suitable procedures allowing separation by size, weight or charge density. The method of rescue is dependent upon whether target DNA sequence is inserted into the genome or is present on an extra chromosomal element, and whether flanking regions allow for excision, or whether the test DNA sequence is part of a replicating element allowing for separation

techniques.

The rescued test DNA sequences may then be transferred into and expressed by host cells, such as microorganisms, suitable for large scale screening techniques. In a preferred embodiment, this involves excising the test DNA sequence vector from the genomic DNA by packaging the test DNA sequence with

bacteriophage packaging techniques. This may require ligating the test DNA sequence into an appropriate vector or merely involve direct transformation into a microorganism.

The test DNA sequence is thereafter replicated on indicator plates or in selective media. In one embodiment, the test DNA sequence is grown in a host cell such as E. coli. A phenotype indicating mutation of the test DNA sequence will identify a mutated test DNA sequence. The ratio of mutated test DNA sequences to the total number of test DNA sequences is a measure of the mutagenicity of the agent and metabolites thereof.

Bacteriophage packaging techniques involve the use of bacteriophage-infected host cell extracts to supply the mixture of proteins and precursors required for encapsidating the bacteriophage DNA from exogenous sources. We have recently discovered that the rescue efficiency of the test DNA sequence can be

significantly increased by eliminating the restriction systems in the strain of host microorganism used both for preparing the packaging extracts as well as those microorganisms used for plating to detect mutagenesis. Additionally, other recovery systems, e.g., DNA transformation of isolated genomic DNA, would be improved by removal of such restriction systems or activities.

By removing these restriction systems which recognize and deactivate foreign DNA, rescue

efficiencies may be increased up to at least 1,000 to about 10⁶ pfu/μg genomic DNA. These rescue

efficiencies enable several million target genes from each tissue be analyzed, generating a large number of data points and resulting in a significant reduction in the numbers of animals required for mutagenesis testing with greater statistical significance.

Accordingly, the integrated target test DNA sequence is, preferably, recovered from the total genomic DNA of the test organism, e.g., by using a lambda packaging extract deficient in restriction systems which recognize and deactivate foreign DNA. The recovered test DNA sequences may then be

transferred into and expressed by restriction system deficient host cells, having deficiencies as described further herein.

Alternatively, a shuttle vector system can be constructed which provides rapid analysis of test DNA sequence. The test DNA sequence may be contained within a system which allows excision and

recircularization of the test DNA sequence, such as a system that is contained by a bacteriophage genome that is readily rescued. Following rescue of the bacteriophage genome containing test DNA sequence using packaging extracts, the test DNA may be further excised from the bacteriophage genome and

recircularized to provide for rapid mutation analysis.

Further, the present invention contemplates the performance of mutagenesis testing by examining the phenotypes of cells containing the test DNA sequence without recovery of the test DNA sequence from the cell. This may be accomplished by the sectioning of tissues of the transgenic organism of the invention, after exposure to a potential mutagenic agent, and assaying the genotype or phenotype of the test DNA sequence by in situ hybridization or, e.g., by

staining of the tissue sections.

The present invention has application in the genetic transformation of multicellular eukaryotic organisms which undergo syngamy, i.e., sexual

reproduction by union of gamete cells. Preferred organisms include non-human mammals, birds, fish, gymnosperms and angiosperms.

In one embodiment, the present invention

contemplates a transgenic fish for the in vivo

screening for mutagenic compounds. Fish represent a category of animals of great interest for agricultural and ecological reasons in the context of water-borne mutagenic compounds, and provide a convenient system for screening mutagenic compounds in a variety of fish species including, but not limited to, trout, salmon, carp, shark, ray, flounder, sole, tilapia, medaka, goldfish, guppy, molly, platyfish, swordtail,

zebrafish, loach, catfish, and the like.

Transgenic fish of numerous species have been prepared, providing the skilled practitioner with a variety of procedures for developing a transgenic fish having an excisably-integrated target gene according to the present invention. See, for example, the teachings of Ozato et al, Cell Differ., 19:237-244 (1986), Inoue et al, Cell Differ. Dev., 29:123-128

(1990), Rokkones et al, J. Comp. Physiol. B, 158:751-758 (1989), and Guyomard et al, Biochimie, 71:857-863 (1989), that describe preparation of transgenic medaka, medaka, salmon and trout, respectively.

Thus, the present invention contemplates a non-human animal containing a modified lambda

bacteriophage (rDNA) of the present invention

excisably-integrated in the genome of the animal's somatic and germ cells, i.e., a transgenic animal. Particularly preferred are transgenic mammals, and are utilized as exemplary herein.

A particularly preferred transgenic mammal is the transgenic mouse described herein that contains a single copy of the lambda LIZ Alpha vector system. An embryo of the preferred transgenic mouse line

containing the excisably integrated lambda LIZ Alpha phage vector was deposited with the American Type Culture Collection (ATCC) on March 17, 1992, under the ATCC accession number 72011.

Mammals containing a rDNA of the present invention are typically prepared using the standard transgenic technology described in Hogan et al,

Manipulating the Mouse Embryo: A Laboratory Manual, Cold Spring Harbor, NY (1987); and Palmiter et al,

Ann. Rev. Genet., 20:465-499 (1986); which methods are described further herein. Production of transgenic mammals is also possible using the homologous

recombination transgenic systems described by

Capecchi, Science, 244:288-292 (1989). Preparation of transgenic mammals has also been described in U.S.

Patent No. 4,736,866, No. 4,870,009, No. 4,873,191 and No. 4,873,316.

One technique for transgenically altering a mammal is to microinject a rDNA into the male

pronucleus of the fertilized mammalian egg to cause one or more copies of the rDNA to be retained in the cells of the developing mammal. usually up to 40 percent of the mammals developing from the injected eggs contain at least 1 copy of the rDNA in their tissues. These transgenic mammals usually transmit the gene through the germ line to the next generation. The progeny of the transgenically manipulated embryos may be tested for the presence of the construct by Southern blot analysis of a segment of tissue.

Typically, a small part of the tail is used for this purpose. The stable integration of the rDNA into the genome of the transgenic embryos allows permanent transgenic mammal lines carrying the rDNA to be established.

Alternative methods for producing a non-human mammal containing a rDNA of the present invention include infection of fertilized eggs, embryo-derived stem cells, totipotent embryonal carcinoma (Ec) cells, or early cleavage embryos with viral expression vectors containing the rDNA. See for example, Palmiter et al, Ann. Rev. Genet., 20:465-499 (1986) and Capecchi, Science, 244:1288-1292 (1989).

A transgenic mammal can be any species of mammal, including agriculturally significant species, such as sheep, cow, lamb, horse and the like. Preferred are animals significant for scientific purposes, including but not limited to rabbits, primates and rodents, such as mice, rats and the like. A transgenic mammal is not human.

Methods of Genetically Programming a Cell Within an Organism With A Target Gene System

The present invention also contemplates a method of introducing a target gene system into a cell, i.e., genetically programming a cell within an organism by introducing a modified lambda genome containing a target gene system of the present invention into the genome of a zygote to produce a genetically altered zygote, or into the genome of individual somatic cells in the organism. The genetically altered zygote is then maintained under appropriate biological

conditions for a time period equal to a gestation period or a substantial portion of a gestation period that is sufficient for the genetically altered zygote to develop into a transgenic organism containing at least 1 copy of the rDNA.

The term "genetically programming" as used herein means to permanently alter the DNA content of a cell within an organism such as a mammal so that a

prokaryotic target gene system has been introduced into the genome of the cells of the organism.

Any multicellular eukaryotic organism which undergoes sexual reproduction by the union of gamete cells may be genetically programmed using an rDNA of the present invention. Examples of such multicellular eukaryotic organisms include amphibians, reptiles, birds, mammals, bony fishes, cartilaginous fishes, cyclostσmes, arthropods, insects, mollusks,

thallaphytes, embryophytes including gymnosperms and angiosperms. In preferred embodiments, the

multicellular eukaryotic organism is a mammal, bird, fish, gymnosperm or an angiosperm.

A transgenic organism is an organism that has been transformed by the introduction of a recombinant nucleic acid molecule into its genome. Typically, the recombinant nucleic acid molecule will be present in all of the germ cells and somatic cells of the

transgenic organism. Examples of transgenic organisms include transgenic mammals, transgenic fish,

transgenic mice, transgenic rats and transgenic plants including monocots and dicots. See for example, Gasser et al, Science, 244:1293-1299 (1989); European Patent Application No. 0257472 filed August 13, 1987 by De La Pena et al; PCT Pub. No. WO 88/02405 filed October 1, 1987 by Trulson et al; PCT Pub. No. WO

87/00551 filed July 16, 1986 by Verma, and PCT Pub. No. WO 88/09374 filed May 20, 1988 by Topfer et al.

Methods for producing transgenic organisms containing a rDNA of the present invention include standard transgenic technology; infection of the zygote or organism by viruses including retroviruses; infection of a tissue with viruses and then

reintroducing the tissue into an animal; and

introduction of a rDNA into an embryonic stem cell of a mammal followed by appropriate manipulation of the embryonic stem cell to produce a transgenic animal. See for example, Wagner et al, U.S. Patent No.

4,873,191 (Oct. 10, 1989); Rogers et al, Meth. in Enzymol., 153:253-277 (1987); Verma et al, PCT

Publication No. WO87/00551; Cocking et al, Science, 236:1259-1262 (1987); and Luskin et al, Neuron 1:635-647 (1988).

Transgenic mammals having at least 1 cell

containing the rDNA's of a prokaryotic gene regulation system of the present invention can be produced using methods well known in the art. See for example,

Wagner et al, U.S. Patent No. 4,873,191 (Oct. 10, 1989); Hogan et al, Manipulating the Mouse Embryo: A Laboratory Manual, Cold Springs Harbor, New York

(1987); Capecchi, Science, 244:288-292 (1989); and Luskin et al, Neuron 1:635-647 (1988).

In preferred embodiments the transgenic mammal of the present invention is produced by:

1) microinjecting a subject rDNA into a

fertilized mammalian egg to produce a genetically altered mammalian egg;

2) implanting the genetically altered mammalian egg into a host female mammal;

3) maintaining the host female mammal for a

time period equal to a substantial portion of a gestation period of said mammal;

4) harvesting a transgenic mammal having at

least one cell containing a rDNA that has developed from the genetically altered mammalian egg.

A fertilized mammalian egg may be obtained from a suitable female mammal by inducing superovulation with gonadotropins. Typically, pregnant mare's serum is used to mimic the follicle-stimulating hormone (FSH) in combination with human chorionic gonadotropin (hCG) to mimic luteinizing hormone (LH). The efficient induction of superovulation in mice depends as is well known on several variables including the age and weight of the females, the dose and timing of the gonadotropin administration, and the particular strain of mice used. In addition, the number of

superovulated eggs that become fertilized depends on the reproductive performance of the stud males. See, for example. Manipulating the Embryo: A Laboratory Manual, Hogan et al, eds., Cold Spring Harbor, NY (1986).

The rDNA may be microinjected into the mammalian egg to produce a genetically altered mammalian egg using well known techniques. Typically, the rDNA is microinjected directly into the pronuclei of the fertilized mouse eggs as has been described by Gordon et al, Proc. Natl. Acad. Sci., USA, 77:7380-7384

(1980). This leads to the stable chromosomal

integration of the rDNA in approximately 10 to 40 percent of the surviving embryos. See for example,

Brinster et al, Proc. Natl. Acad. Sci., USA, 82:4438-4442 (1985). In most cases, the integration appears to occur at the 1 cell stage, as a result the rDNA is present in every cell of the transgenic animal, including all of the primordial germ cells. The number of copies of the foreign rDNA that are retained in each cell can range from 1 to several hundred and does not appear to depend on the number of rDNA injected into the egg as is well known.

An alternative method for introducing genes into the mouse germ line is the infection of embryos with virus vectors. The embryos can be infected by either wild-type or recombinant viruses leading to the stable of integration of viral genomes into the host

chromosomes. See, for example, Jaenisch et al, Cell, 24:519-529 (1981). One particularly useful class of viral vectors are virus vector derived from retro-viruses. Retroviral integration occurs through a precise mechanism, leading to the insertion of single copies of the virus on the host chromosome. The frequency of obtaining transgenic animals by

retroviral infection of embryos can be as high as that obtained by microinjection of the rDNA and appears to depend greatly on the titre of virus used. See, for example, van der Putten et al, Proc. Natl. Acad. Sci., USA, 82:6148-6152 (1985).

Another method of transferring new genetic information into the mouse embryo involves the

introduction of the rDNA into embryonic stem cells and then introducing the embryonic stem cells into the embryo. The embryonic stem cells can be derived from normal blastocysts and these cells have been shown to colonize the germ line regularly and the somatic tissues when introduced into the embryo. See, for example, Bradley et al, Nature, 309:255-256 (1984). Typically, the embryo-derived stem cells are

transfected with the rDNA and the embryo-derived stem cells further cultured for a time period sufficient to allow the rDNA to integrate into the genome of the cell. In some situations this integration may occur by homologous recombination with a gene that is present in the genome of the embryo-derived stem cell. See, for example, Capecchi, Science, 244:1288-1292 (1989). The embryo stem cells that have incorporated the rDNA into their genome may be selected and used to produce a purified genetically altered embryo derived stem cell population. See, for example, Mansour et al, Nature, 336:348 (1988). The embryo derived stem cell is then injected into the blastocoel cavity of a preimplantation mouse embryo and the blastocyst is surgically transferred to the uterus of a foster mother where development is allowed to progress to term. The resulting animal is chimeric in that it is composed from cells derived of both the donor embryo derived stem cells and the host blastocyst. Heterozygous siblings are interbred to produce animals that are homozygous for the rDNA. See for example, Capecchi, Science, 244:1288-1292 (1989).

The genetically altered mammalian egg is

implanted into host female mammals. Methods for implanting genetically altered mammalian eggs into host females are well known. See, for example, Hogan et al, Manipulating the Mouse Embryo: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York (1986). Pseudopregnant recipient females may be produced by mating females in natural estrus with vasectomized or genetically sterile males. After mating with a sterile male, the female

reproduction tract becomes receptive for transferred embryos even though her own unfertilized eggs

degenerate. The genetically altered mammalian eggs are then transferred to the ampullae or the uterine horns of the pseudopregnant recipient. If the

genetically altered mammalian egg is transferred into the ampullae it must be enclosed in a zona pellucida membrane. If it is transferred into the uterine horns the genetically altered mammalian egg does not require a zona pellucida membrane.

The host female mammals containing the implanted genetically altered mammalian eggs are maintained for a sufficient time period to give birth to a transgenic mammal having at least 1 cell containing a rDNA of the present invention that has developed from the

genetically altered mammalian egg. Typically this gestation period is between 19 to 20 days depending on the particular mouse strain. The breeding and care of mice is well known. See for example, Manipulating the Mouse Embryo: A Laboratory Manual, Hogan et al, eds., Cold Spring Harbor, New York, (1986).

The infection of cells within an animal using a replication incompetent retroviral vector has been described by Luskin et al, Neuron, 1:635-647 (1988).

In one embodiment, an animal that contains a target gene system in specific tissues or cells is used to test the effect of a material, composition, or compound suspected of being a carcinogen on the specific tissue. The animal is exposed to the

particular material or compound and the mutagenic effect on the animal is determined by the derepression of the operator-regulated reporter gene segment as an indication of the carcinogenicity of the compound or material.

The composition suspected of having carcinogenic activity is introduced into the animal by any suitable method including injection, or ingestion or topical administration.

The animal is then maintained for a predetermined time period that is sufficient to allow the

composition to produce a mutagenic effect on the genes of the target gene system. Typically, this time period ranges from several minutes to several days depending on the time the composition requires to mutagenize the genes.

The physiological process or parameter assayed as an indication of mutagenesis depends upon the

particular physiological alteration produced by the expression of the reporter gene.

A change in a physiologic parameter is determined by measuring that parameter before introduction of the composition into the animal and comparing that

measured value to a measured value determined in identical manner after introduction of the composition into the animal.

The copy number of the test gene in the

transgenic animals alters the sensitivity of transgenic animals to the effects of the suspected carcinogen. Therefore, selection of transgenic animals with varying transgene copy numbers of the test gene will alter the sensitivity of the transgenic mice to the suspected carcinogen.

Selectable Systems

In one embodiment, the invention contemplates the use of mutagenesis testing systems according to the present invention which provide selection for

mutagenic events. Selectable systems are particularly preferred because (1) they provide increased speed and convenience (i.e., efficiency) in detecting the reporter gene product due to the selection of

mutagenized test DNA sequences over a background of non-mutagenized test DNA sequences and may increase sensitivity of the assay, and (2) they allow for the screening of larger numbers of clones per unit of culture medium that is used to bioassay the mutagenic event.

Many possible combinations of test DNA sequences and reporter genes are contemplated by the present invention in which the reporter gene provides a selective growth advantage to the host cell or

bacteriophage which harbors the mutagenized test DNA sequence. In each case, the expression of the

reporter gene in response to the test DNA gene product provides the switch between a selective growth

advantage and a growth disadvantage. Where the expression of a reporter gene provides a selective growth advantage, the non-mutated test DNA gene product functions, directly or indirectly, to repress expression of the reporter gene product. Conversely, where the expression of a reporter gene provides a selective growth disadvantage, the non-mutated test DNA gene product functions, directly or indirectly, to de-repress (or activate/promote) expression of the reporter gene product.

In one preferred embodiment, the reporter gene transcription unit is independent of the test gene, and typically located within the host cell used for bioassaying the mutagenic event. Thus, the regulation of the reporter gene by the test gene occurs between separate genetic elements and is referred to as regulation in trans. This system provides the

advantage of reducing the number of genes in the target gene system of the invention which are present in the test organism which may be subject to a

mutation by the mutagen. By removing genes away from the mutagen, one insures that competing mutations

(mutations in genes other than the test DNA sequence) do not complicate the expression of a detectable phenotype, thereby altering the sensitivity and accuracy of the mutagenic assay.

One selectable system has been described in

Example 8 in which the use of lactose minimal growth medium in the bioassay is combined with the use of the Lac repressor as a test DNA sequence which controls the expression of the LacZ component of the beta-galactosidase gene as the reporter gene. lacZ

expression is required for host cell growth, and the gene is repressed if non-mutated Lacl is present, thereby selecting for mutation events.

An alternative selectable system utilizes a growth advantage provided by the expression of a reporter gene, such that upon mutation of the target gene, the reporter gene is expressed when the target gene is assayed. Thus clones containing mutations on the target gene are "selected" by the growth selection over clones without the mutation. Selectable reporter genes have been described earlier, and include

antibiotic resistance, and the groE and S⁵ systems described herein.

Another selectable system involves the use of a tRNA suppressor gene on the test DNA sequence and the use of a reporter gene which depends of suppression for expression.

Still another selectable system involves the vise of anti-terminator proteins to regulate the reporter gene transcription unit. In this embodiment, the reporter gene transcription unit contains, in addition to the reporter gene itself, terminator and anti-terminator nucleotide sequences in the promoter region of the transcription unit that terminates

transcription unless an anti-terminator protein is present that specifically binds to the anti-terminator sequence. Thus, a controlling element in such a system contains both a terminator sequence and an anti-terminator sequence that overrides termination of transcription and is referred to collectively as a terminator/anti-terminator nucleotide sequence.

Exemplary terminator/anti-terminator nucleotide sequences are described in Examples 9-11.

Furthermore, the target gene system contains, in addition to the test DNA gene sequence, an anti-terminator gene that expresses an anti-terminator protein that specifically binds the anti-terminator sequence in the reporter gene transcription unit and derepresses the terminator sequence, thereby allowing transcription of the reporter gene product.

The use of terminator/anti-terminator sequences in the reporter gene transcription unit provides a particular advantage in the present invention. By providing the anti-terminator protein coding sequence on the test DNA sequence that is introduced into the host cell, the anti-terminator protein acts in trans relative to the terminator/anti-terminator sequence, and is not present until the test DNA sequence is added to the host cell bioassay system. Thus, the reporter gene transcription unit is kept off until the test DNA sequence is added, thereby reducing and preferably preventing any possible leakiness to the reporter gene, and thus any false regulation of reporter gene transcription. No reporter gene product can accumulate prior to the introduction of the test DNA due to low grade transcription of the reporter gene transcription unit. This insures a low

background to the system, and blocks against competing mutations that might independently activate the reporter gene which arise in the host cell rather than in the test DNA sequence.

A preferred anti-terminator protein is the lambda N protein that binds to the lambda nutR anti- terminator sequence or the lambda Q protein that binds to the lambda qut anti-terminator sequence. Numerous terminator and anti-terminator sequences are known in the art that are suitable for use in the present invention. For example, the lambda nutR or nutL anti-terminator sequences are regulated by binding by lambda anti-terminator protein N, and the lambda qut anti-terminator sequence is regulated by binding by lambda anti-terminator protein Q. Terminator

sequences that terminate transcription in the absence of anti-terminator protein can include any combination of one or more terminator sequences. Exemplary terminators include the lambda terminators tI, tL1, tL3, tR1, tR2, t6s (also known as t'Rl1, t'J1, t'J2, t'J3, t'J4 and tRO, the E. coli terminators IS1, IS2 and tryptophan gene terminator (tTRP), phage P82 tR' (t82), phage P22 tANT, tI4 and tfd. Terminator tR1 is particularly preferred and is exemplary of the diversity of numbers and position of terminators possible for use with an anti-terminator sequence. For example, tR1 is known to contain 5 terminator sites, designated as tR1(I-V), and various combinations of the 5 sites can be used, such as tR1(I-II), tR1(I-III) and tR1(I-V), as described herein. In addition, multiple different terminators can be used, such as to combine tR1 with t6s, and the like combinations as described in the Examples.

Exemplary systems that utilize a terminator sequence are described in Examples 9, 10 and 11.

It should be pointed out that in the reporter gene transcription units described herein to provide a selectable system, a number of genetic variables are contemplated which can be modified to optimize the sensitivity of the system that depend, in part, upon the particular phenotype being utilized as the

detection of a reporter gene. These variables include (1) the type and thereby strength of the promoters used to express a reporter gene, the (2) the type and thereby the strength of the operators used to control the promoters, and (3) the copy number of the reporter gene transcription unit within the host cell relative to the copy number of the test DNA sequence. It is understood that for any phenotype dependent upon the regulation of gene expression there are optimum, and therefore preferred, combinations of promoters, operators and gene copy number. The routine

modification of these various genetic components in the reporter gene transcription unit by one skilled in the art is contemplated for the purpose of titering these components relative to each other to achieve optimum levels of control of expression and therefore optimum sensitivity to the regulation of the reporter gene phenotype by the test DNA sequence.

Promoters for use in controlling a reporter gene include Placl^q, lacP, and Ptrp, which each exhibit a different promoter strength, and their selection depends on the promoter strength desired in adjusting the level of transcription for the reporter gene transcription unit.

Many operators are available for use in

controlling a reporter gene construct which exhibit different levels of regulation, and are described by Miller et al., in "The Operon" Cold Spring Harbor Laboratory, 2nd edition, 1980.

Prolysogenic Organisms

The present invention also contemplates a

selectable system for screening (testing) for the presence of mutagenic activity upon the target gene system of this invention that utilizes a prolysogenic organism to provide positive selection for mutagenized target genes. Thus, the invention also contemplates a prolysogenic organism for use in the system and methods of this invention. A preferred prolysogenic organism is a prolysogenic microorganism and will be used as exemplary herein.

A prolysogenic microorganism (prolysogen) is a microorganism containing an isolated bacteriophage cI gene. By "isolated bacteriophage cI gene" is meant a cI gene separated from other bacteriophage genes. The isolated cI gene is present in the microorganism operatively linked to expression control elements for producing a bacteriophage lytic cycle-suppressing amount of cI gene product in the microorganism. The microorganism can be any microorganism, such as a yeast, bacterium and the like, adapted for infection by a bacteriophage, and preferably is a strain of E. coli .

A lytic cycle-suppressing amount of cl gene product is an amount sufficient to prevent a lambda bacteriophage-infected cell from lysing during the lytic phase of the bacteriophage's life cycle. The study of bacteriophage lambda is extensive in the biological arts, and the life cycle, and the lytic and lysogenic phases of the lambda life cycle are

extremely well characterized. Thus assays for

determining whether the cl amount is sufficient for suppression of the lytic cycle, and produce a

lysogenic infection, is well known in the art.

The microorganism expressing a lytic cycle-suppressing amount of cl gene product is referred to as a prolysogen to connote its ability to impose a lysogenic life cycle upon a lambda-infected cell, even if the lambda would otherwise have the ability to be lytic. The control of lytic versus lysogenic life cycles for lambda bacteriophage is well known to reside in the expression of the cl gene product.

In preferred embodiments, the prolysogen is phage-free, i.e., is free of genetic material

recoverable via a bacteriophage packaging extract.

Further preferred is a prolysogen that is

restriction system deficient, e.g., a prolysogenic strain of E. coli deficient in one or more of the mcrA, mcrBC, mrr, mcrF, hsdR restriction systems and the like. It is preferred that the prolysogen not contain any restriction system similar to a

restriction systems found in the minute 98 region of E. coli K-12.

Methods for producing an isolated cl gene are well known in the art. A preferred method utilizes PCR amplification of the gene and its native promoter as a single DNA segment no more than about 1000 nucleotides in length. Typical and preferred are the methods described herein. When genomic integration is desired, the cl gene-containing DNA segment is then typically operatively linked to an genomic insertion element, such as a transposon, or the insertion elements of the transposon. Alternatively, the cl gene-containing DNA segment can be linked to a plasmid capable of low copy number maintenance in the host.

The amount of cl gene, and therefore the amount of cl gene product expressed can vary, so long as the amount is sufficient to suppress lytic cycle, as described previously. To that end, the number of copies of the cl gene can vary, although typically 1 to 4 copies are preferred, particularly 1 copy as demonstrated herein in the preferred embodiment of the SCS-8cI lysogen.

The preparation and use of the SCS-8cI lysogen in the screening/testing to quantitate the amount of mutagenized target gene in a selectable system is described in more detail in Example 8.

Also contemplated is a kit for practicing the methods of the present invention that comprise, in an aliquot, a prolysogen of this invention. A kit can further contain a lambda phage packaging extract of this invention for use with the prolysogen, and having a restriction system deficiency compatible with the prolysogen.

Examples

The following examples are intended to

illustrate, but not limit, the present invention.

Accordingly, variations and equivalents, now known or later developed, that would be within the purview of one skilled in this art are to be considered to fall within the scope of this invention, which is limited only as set forth by the appended claims.

1. Construction of E. coli RecA- Lysogen Strains

Strains BHB2688R- and BHB2690R- are constructed using RecA⁺ transformants of E. coli strains BHB2688 and BHB2690, respectively, as the recipients and any E. coli K-12 strain that carries a Tn10 (tetracycline resistant) in (or near) the mcrA gene (relevant genotype = McrA: Tn10 (tet^R)) as the donor. BHB2688, which is E-, and BHB2690, which is D-, are available from the American Type Culture Collection (ATCC), Rockville, MD under the accession numbers 35131 and 35132, respectively. RecA⁺ transformation is

accomplished by standard methods, typically using a RecA expressing plasmid. Step 1: A P1 lysate is made from the E. coli K-12 strain described above. Step 2: BHB2688 and BHB2690 are transduced with the P1 lysate (Miller, Experiments in Molecular Genetics, Cold

Spring Harbor Lab., Cold Spring Harbor, New York

(1972)). Step 3: Tetracycline (tet^R) resistant colonies are selected and purified. Step 4: Loss of tetracycline resistance is selected for on Bochner plates (Bochner et al, J. Bacteriol., 143:926-933 (1980)), and colonies are purified. Step 5: Lack of McrA restriction activity is tested by comparing transformation efficiency of unmethylated pBR322 versus pBR322 that has been in vitro methylated by Hpall methylase (Raleigh, supra). A McrA⁺ strain will show a greatly reduced efficiency with the methylated plasmid. If McrA activity is absent, this strain is then called BHB2688McrA- and BHB2690McrA-.

To delete the mcrB locus in the above McrA- strains, a donor E. coli K-12 strain with the relevant genotypes McrB: :Tn10 (tet^R), mrr: :Tn5 (kan^R) was used. Make a P1 lysate from an E. coli K-12 strain that carries a Tn10(tet^R) in the mcrB gene. The strain should also have a Tn5(kan^R) in the mrr gene. Step 7: Transduce the BHB2688 McrA- and BHB2690 McrA- recA⁺(tet^s) strain. Step 8: Select for tet^R colonies. Purify one colony that is also kan^R. Step 9: Select for loss of tet^R on Bochner plates (Bochner, supra). Step 10: Purify several colonies and test for

sensitivity to tetracycline and kanamycin. Select colonies that are both tet^s and kan^s. Step 11: Test for lack of McrB restriction activity as done for the McrA test, however in this case, the pBR322 should be in vitro methylated by Alul methylase (Raleigh, supra; Ross, supra). A McrB⁺ strain will show a greatly reduced efficiency with the methylated plasmid. Test for Mrr restriction activity by comparing plating efficiency of lambda versus lambda which has been in vivo methylated by Pst I methylase (Heitman, supra). An Mrr⁺ strain will show reduced efficiency with the methylated lambda. The Mrr- strain is preferably also McrF-. McrF is tested using packaged lambda DNA that has been in vitro methylated with SssI methylasse. An McrF⁺ strain shows a reduced number of plaques when using methylated lambda DNA. Test for HsdR

restriction activity by comparing plating efficiency of lambda versus lambda which has been in vitro

methylated by HsdM methylase (Wood, J. Mol. Biol.. 16:118-133, (1966); Adams, Bacteriophages, New York: Interscience 1959; Bickle, supra. at pp. 95-100). An HsdR+ strain will show reduced efficiency with

unmethylated lambda. If a strain (purified colony) lacks all the restriction activities McrA, McrBC, Mrr, McrF, HsdR and was constructed by this method, it should then contain a deletion throughout the mcrB region (▴mcrB). These strains are designated

BHB2688R- and BHB2690R-. A . E . col i BHB2690R-

(1) Transduction to Obtain BHB2690mcrA- a) Preparation of a P1 Lysate

A bacteriophage P1 lysate, hereinafter referred to as P1, was made from any E. coli K12 strain that carries a tetracyline resistant Transposon 10 (Tn10) in or near the mcrA gene (Tn10: :McrA).

Briefly, one drop from an overnight culture of K12 was admixed into 5 ml of LB broth (Luria-Bertani broth was prepared by admixing and dissolving 10 grams (g) bacto-tryptone, 5 g bacto-yeast extract and 10 g NaCl into 1 liter deionized water) containing 5 X 10^-3 molar (M) CaCl₂. The admixture was aerated by

swirling until the cells were in exponential log phase growth and had reached a density of 2 X 10⁸ cells/ml. P1 was preadsorbed by admixing 10⁷ phage to 1 ml of the above admixture followed by maintenance at 20 minutes in a 30 degrees Celsius (30°C) waterbath to form a phage-cell admixture. LB-top agar, 2.5 ml, maintained at 45°C, was then admixed with the phage-cell admixture. The resultant agar-containing cell suspension was plated onto a freshly made LB plate which was maintained at 30°C for 8 hours. At the end of the maintenance period, the soft agar layers was scraped into a small centrifuge tube. The scraped surface of the plate was then washed with 1 ml broth and the wash was collected for admixture with the scraped soft agar. Five drops of chloroform were added to the centrifuge tube followed by

centrifugation to pellet cell debris. The resultant supernatant, containing the P1 lysate, was collected.

b) Transduction with P1 Lysates

In this invention, E. coli lysogen

BHB2690 (ATCC # 35132) was used as the specific strain for transduction. E. coli BHB2690, which was RecA-, was first transformed with pJC859 to introduce a functional RecA protein into the lysogen. pJC859 was a plasmid in which the nucleotide sequence encoding RecA had been inserted at the Bam HI site of the plasmid E. coli vector, pBR322 (ATCC # 31344).

Maniatis et al, Molecular Cloning: A Laboratory

Manual, Cold Spring Harbor Laboratory, 2nd ed.,

Sections 1.12 and 5.88, 1989; and US Patent 4,843,006 (in which the RecA promoter/operator and 150 amino acid residues of RecA coupled to a heterologous polypeptide sequence is described). For the

transformation, E. coli BHB2690 competent cells were prepared following standard procedures familiar to one skilled in the art. Maniatis et al, supra, Section 1.76. Alternatively, competent cells can be obtained commercially.

Five ml of a fresh overnight culture of BHB2690 RecA⁺ to be transduced was resuspended in 0.5 ml of buffer consisting of 0.1 M MgSO₄ and .5 mM CaCl₂ according to the procedure by Miller. Miller,

Experiments in Molecular Genetics.. Cold Spring Harbor Laboratory, 1972. The BHB2690 RecA⁺ cell suspension was then aerated by swirling at 30°C for 15 minutes. To each of 5 small test tubes, 0.1 ml of the aerated suspended cells was added. One hundred microliters (ul) of P1 lysate, prepared above, was added to the first tube. The P1 lysate was serially diluted 10-fold for addition to the remaining tubes except for the last tube which did not receive P1 and, thus, served as a control. A tube without cells containing P1 was also used as an additional control. The P1 lysate was preabsorbed by maintaining the tubes at 30ºC in a water bath for 20 minutes. Two hundred ul of 1 M sodium citrate was then added to each of the prepared tubes and the contents of each tube was then plated on tetracycline-containing plates to select for tetracycline resistant (tet^R) colonies.

After maintaining the plates at 30°C to allow for growth of colonies, tet^R colonies were selected and purified following procedures well known to one skilled in the art. The tet^R colonies were then replated on Bochner plates to select for the loss of tetR as described by Maloy. Maloy et al, J.

Bacteriol.. 145:1110-1112 (1981). Briefly, tet-sensitive, tet^s, colonies were selected on a medium consisting of the following: 15 grams/liter (g/1) agar; 5 g/1 tryptone broth; 5 g/1 yeast extract; 4 milliliters/1 (ml/1) chlortetracycline hydrochloride (12.5 milligram (mg)/ml); 10 g/1 NaCl; 10 g/1 NaH₂PO₄-H₂O; 6 ml/1 fusaric acid (2 mg/ml); and 5 g/1 ZnCl₂ (20 millimolar (mM)). Chemicals were obtained from Sigma. (Sigma, St. Louis, MO).

Selected tet^s colonies were then purified and tested for the lack of McrA restriction activity. The determination of McrA- strains was accomplished by comparing transformation efficiency of unmethylated pBR322 versus pBR322 that had been in vitro methylated by Hpa II methylase. A McrA- strain showed a greatly reduced efficiency with the methylated plasmid.

BHB2690RecA⁺McrA-; hereinafter designated BHB2690McrA-, strains were, thus, determined and used to make

BHB2690McrB- transductions as described below.

(2) Transduction to Obtain BHB2690mcrB- The BHB2690McrA- strains prepared above were used in similar transductions to select for

BHB2690McrB- strains. For this procedure, a P1 lysate was prepared as described above from any E. coli K12 strain that carried a Tn10 (tet^R) in the mcrB gene (McrB: :Tn10 (tet^R). The strain selected also carried the Tn5 with kanamycin (antibiotic) resistant gene (kan^R) in the mrr gene (Mrr : : Tn5 (kan^R) .

The E. coli BHB2690McrA- (tet^s) strains were then transduced with P1 lysate [(McrB::Tn10 (tet^R) and (Mrr::Tn5 (kan^R) as described in Example la above.

Tet^R colonies that were also kan^R were selected and purified. The loss of tet^R on Bochner plates was measured as described above. Colonies that were both tet^s and kan^s after selection on Bochner plates were purified.

The lack of McrB restriction activity was

performed as described for determining the lack of McrA activity with the exception that pBR322 was in vitro methylated by Alu I methylase. A McrB⁺ strain showed a greatly reduced efficiency with the

methylated plasmid. The test for Mrr restriction activity was accomplished by comparing plating

efficiency of lambda versus in vivo methylated lambda (by Pst I methylase). A Mrr- strain showed reduced efficiency with the methylated lambda. The Mrr-strain is preferably also McrF-, as described before.

A separate test for HsdR restriction activity was also performed as the lack of activity confirmed the deletion of the entire McrB region. The HsdR

restriction activity test was performed by comparing plating efficiencies of lambda versus lambda which had been in vivo methylated by HsdM methylase. A HsdR+ strain showed reduced efficiency with the unmethylated lambda. With these tests, a selected colony which lacks all restriction activity, McrA, McrBC, McrF, Mrr, and HsdR, and constructed using this transduction approach was shown to contain a deletion throughout the McrB region. The resulting strain was grown in liquid culture and then plated to isolate colonies on NZY media. The isolated colonies were screened for sensitivity to ampicillin, indicating a loss of the RecA plasmid pJC859. An Amp^s isolate was thereby purified. This strain was designated BHB2690R- and used in the preparation of extract for prehead as described in Example 2a.

B. E. coli BHB2688R-

E. coli BHB2688 strains containing RecA⁺ but lacking McrA, McrBC, McrF, Mrr and HsdR were prepared using the approach described above for preparing

E. coli BHB2690R-. For the transductions, E. coli lysogen BHB2688 (ATCC # 35131) was used. The

resultant strain, designated BHB2688R-, was used in the preparation of extract for protein donor a

described in Example 2 below.

2. Preparation of Packaging Extracts from Two

Lysogens

A. Preparation of Sonicated Extract from

Induced E. coli Strain BHB2690R- Cells

For preparing a sonicated extract of the

E. coli lysogen, strain BHB2690R- (prehead donor) prepared in Example la, the genotype of the strain is first verified before large-scale culturing. The presence of the mutation that renders the

bacteriophage cl gene product temperature-sensitive is determined by streaking from the master stocks of E. coli BHB2690R- onto two LB agar plates. One of the plates is maintained at 32°C while the other is maintained at 45°C. Bacteria with intact cl only grow on the plates maintained at 32°C. A single small colony of E. coli BHB2690R- is picked and maintained overnight at 32°C and 45°C. The bacteria with the mutation only grow at 32°C and grow slowly due to the RecA- mutation present in the BHB strains. A 100 ml subculture of the verified master stock of E. coli strain BHB2690R- is then prepared and maintained overnight at 32°C.

After maintaining the E. coli BHB2690R- culture overnight, the optical density (OD) is measured at a wavelength of 600 nm. An aliquot of the overnight culture is admixed into 500 ml of NZM broth (NZM broth is prepared by admixing 10 g NZ amine, 5 g NaCl, and 2 g MgSO₄-7H₂O to 950 ml of deionized water; the pH of the solution containing dissolved solutes is adjusted to pH 7.0 with 5 N NaOH), prewarmed to 32°C, in a 2- liter flask, to result in a starting OD₆₀₀ of

approximately 0.1. The bacterial admixture is then maintained at 32°C with vigorous agitation (300 cycles/minute in a rotary shaker) until an OD₆₀₀ of approximately 0.3 is reached. The OD₆₀₀ of 0.3 is generally attained within 2 to three hours of

maintaining the culture. The cultures must be in the mid-log phase of growth prior to induction as

described below.

The lysogen is induced by placing the flask in a water bath preheated to 45°C. The flask is swirled continuously for 15 minutes. An alternative approach for inducing lysogen is to immerse the flask in a shaking water bath set at 65°C. The temperature of the fluid contents of the flask is monitored. When the fluid reaches 45°C , the flask is then transferred to a water bath set at 45°C and maintained for 15 minutes. The induced cells are then maintained for 2 to 3 hours at 38 to 39°C with vigorous agitation as described above. A successful induction is verified by the visual clearance of an added drop of chloroform to the culture.

Following the 2 to 3 hour maintenance period, the cells are recovered from the admixture by

centrifugation at 4000g for 10 minutes at 4°C. The resultant supernatant is decanted and any remaining liquid is removed with a pasteur pipette and a cotton swab. The walls of the centrifuge bottle are wiped dry with towels. To the pelleted induced bacterial cells, 3.6 ml of freshly prepared sonication buffer is admixed. Sonication buffer consists of 20 mM Tris-HCl, pH 8.0, (Tris [hydroxymethyl]-aminomethane

hydrochloride), 1 mM EDTA, pH 8.0,

(ethylenediaminetetraacetic acid) and 5 mM beta-mercaptoethanol. The bacterial cell pellet is

resuspended in the sonication buffer by mixing to form a homogenized cell suspension.

The resultant suspension is transferred to a small, clear plastic tube (Falcon 2054 or 2057,

Falcon, Oxnard, California) for subsequent sonication. The cells are disrupted by sonication with 10 second bursts at maximum power using a microtip probe. For sonication, the tube containing the suspension is immersed in ice water and the temperature of the sonication buffer should not be allowed to exceed 4°C. The sample is cooled for 30 seconds in between each sonication burst. The sonication procedure is

continued until the solution in the tube clears and its viscosity decreases. The sonicated bacterial sample is transferred to a centrifuge tube and debris is pelleted by centrifugation at 12,000g for 10 minutes at 4°C forming a clear supernatant.

The resultant supernatant containing preheads is removed and admixed with an equal volume of cold sonication buffer and one-sixth volume of freshly prepared packaging buffer to form a diluted prehead admixture. Packaging buffer consists of the

following: 6 mM Tris-HCl, pH 8.0; 50 mM spermidine; 50 mM putrescine; 20 mM MgCl₂; 30 mM ATP, pH 7.0; and 30 mM beta-mercaptoethanol. The admixture is then dispensed into pre-cooled to 4°C 1.5-ml microfuge tube in 15 ul aliquots. The caps of the microfuge tubes are then closed and the tubes are immersed briefly in liquid nitrogen for freezing. The frozen preheads in packaging buffer are then stored at -70°C for long-term storage.

B. Preparation of Frozen-Thawed Lysate of

Induced E. coli BHB2688R- Cells

A subculture of E. coli BHB2688R- , prepared in Example 1b above, for obtaining an extract of packaging protein donor is verified for the genotype and is prepared as described above for preparing E. coli BHB2690R-. Overnight cultures are maintained and lysogen is induced also as described above.

The induced E. coli BHB2688R- cells are pelleted by centrifugation at 4000g for 10 minutes at 4°C. The resultant supernatant is removed and any excess liquid is removed. The pelleted cells are resuspended in a total of 3 ml of ice-cold sucrose solution (10% sucrose in 50 mM Tris-HCl, pH 8.0) to form a cell suspension. The resultant cell suspension is

dispensed in 0.5 ml aliquots into each of six

precooled to 4°C microfuge tubes. Twenty-five ul of fresh, ice-cold lysozyme solution (2 mg/ml lysozyme in 10 mM Tris-HCl, pH 8.0) is admixed to each tube containing the cell suspension. The cell-lysozyme admixture is gently mixed to form an E. coli extract and then immersed in liquid nitrogen for freezing.

The frozen tubes are removed from the liquid nitrogen and the extracts are thawed on ice. Twenty-five ul of packaging buffer, as prepared above, is admixed to each tube containing thawed extract to form a packaging buffer-extract admixture. The separately prepared admixtures are then combined in a centrifuge tube and centrifuged at 45,000g for 1 hour at 4°C to form an supernatant containing packaging protein donor.

The resultant supernatant is removed and

dispensed in 10 ul aliquots into precooled at 4°C microfuge tubes. The caps of the tubes are closed and the tubes are then immersed in liquid nitrogen. The tubes are then removed from the liquid nitrogen and stored at -70°C for long term storage.

3. In Vitro Packaging Using Two Extracts

The frozen tubes containing prehead and packaging donor extracts prepared in Example 2a and 2b,

respectively, are removed from storage at -70°C and allowed to thaw on ice. The frozen-thawed lysate containing the protein donor thaws first and is admixed to the still-frozen sonicated prehead extract to form a prehead-protein donor admixture. The resultant admixture is mixed gently until almost totally thawed. The DNA to be packaged (up to 1 ug dissolved in 5 ul of 10 mM Tris-HCl, pH 8.0, 10 mM MgCl₂) is admixed with the thawed combined extracts and mixed with a fine glass stirring rod to form a DNA-extract admixture. The admixture is then

maintained for 1 hour at room temperature. To the admixture, 0.5 to 1 ml of SM (SM buffer is prepared by admixing 5.8 g NaCl, 2g MgSO₄-7H₂O, 50 ml Tris-HCl, pH 7.5, and 5 ml 2% gelatin (w/v) to 1 liter of deionized water and adjusting the pH to 7.5) and a drop of chloroform is added and gently mixed. Debris is removed by centrifugation at 12,000g for 30 seconds at room temperature in a microfuge. The resultant supernatant is removed and contains packaged

bacteriophage DNA particles.

The titer of the viable bacteriophage particles is measured by plating on the appropriate indicator strains. Recombinant DNAs that are 90% or 80% of wild-type bacteriophage lambda in length are packaged with efficiencies that are 20-fold to 50-fold lower, respectively, than those obtained with unit-length bacteriophage lambda. The same packaging extracts may be used for the packaging of both bacteriophage lambda and cosmids.

As an alternate embodiment for a packaging reaction, it was determined that the packaging

extracts can be admixed with the packaging reaction either simultaneously or by multiple additions of the packaging extract. For example, a comparison between the single addition and multiple addition of extracts to packaging reaction admixture was made as follows. A conventional procedure using 4 ul of DNA to be packaged was admixed with 10 ul of freeze-thaw extract and 15 ul of sonicate extract, and the admixture maintained over 3 hours at 30°C. Thereafter, the titer of phage packaged was determined. In contrast, 4 ul of DNA was admixed with 5 ul of freeze-thaw extract and 7 ul of sonicate extract and maintained for 1.5 hours at 30°C. Thereafter, an additional 5 ul of freeze-thaw extract and 7 ul of sonicate extract was added to the maintained admixture, and further maintained for 1.5 hours at 30°C. Thereafter the titre of phage packaged was determined. About 1.5 to at least 4 times as many infectious phage particles were present in the final packaging reaction when the packaging extracts were added repeatedly. In testing other permutations, it was determined that from 2 to about 5 sequential additions, preferably 2, of

packaging extracts could be added to the packaging reaction during the packaging process to produce higher yields when compared to the result when a single admixture and incubation was conducted.

4. Preparation of Transgenic Mice and Their Use

The following studies provide details of the manner in which the present invention may be made and used in order to achieve the rapid recovery and examination of test DNA sequences from transgenic animals.

A. DNA Test Seguence

The test sequence DNA can, theoretically, contain any number or variety of genes or other identifiable test DNA sequences. In one prototype described herein, an E. coli bacteriophage lambda genome has been engineered to carry lacZ, a beta-galactosidase test DNA sequence. Lambda shuttle vectors L2B (46.5kb) or C2B (48.0kb) may be used. The genotype of the modified lambda genome L2B is lac5 delta (shind III lambda 2-3°) srl lambda 3-5° cI857 sXhL lambda 1° sScII lambda 4°. Before injecting it into mouse embryos as described below, this lambda DNA was diluted to a concentration of 10 micrograms per milliliter and the cos ends were annealed and ligated under conditions predominantly forming circular lambda phage monomers. Maniatis et al, Molecular Cloning, A Laboratory Manual, pp. 109-110, 383-389 (Cold Spring Harbor, New York 1982).

In another embodiment, a variation of L2B was constructed that contains a plasmid sequence that can be readily excised from the lambda phage and contains the lacl repressor gene. This variation has several advantages. First, as discussed below, physical identification of phage carrying mutations is

facilitated since they grow as blue plaques on a white background in the presence of X-gal (5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside) without IPTG (isopropylβ-D-thiogalacto-pyranoside). This advantage also simplifies and reduces the cost of the assay since it permits an increase in the density of phage per plate. Additionally, the lacl genetic systems of E. coli are the first systems that conveniently permitted the study of large numbers of mutations within procaryotes at the DNA level (Miller et al, J. Mol. Biol., 109:275-302 (1977), Coulondre et al, J. Mol. Biol., 117:275-302 (1977), Schaaper, J. Mol.

Biol., 189:273-284, (1986)), and the use of lacl

provides a test gene with significant historical mutational data for comparison between mutagenesis assays.

Still further permutations of a test DNA sequence are described hereinbelow.

B. Creation of a Transgenic Animal

Mice were used as the test animal. (Hogan et al, Manipulating the Mouse Embryo: A Laboratory Manual, Cold Spring Harbor Laboratory, 1986). Single cell mouse embryos were harvested from female mice that were impregnated the evening before. The embryos were treated with hyaluronidase and briefly cultured in M16 medium. The embryos were transferred to M2 medium on a microscope glass depression slide. The embryos were observed with a 40X objective and a 10X eyepiece using a Nikon Diaphot microscope equipped with Hoffman optics. The embryos were held in place with a holding pipet that had been rounded with a microforge. The positions of both the holding pipets and the injection pipets were controlled with

micromanipulators. DNA as described above was loaded in the injection pipet at a concentration of 1 to 10 micrograms per milliliter. Approximately one

picoliter, as judged by a refractile change (Hogan et al, supra) of the pronucleus, of DNA solution was injected into the male pronucleus.

After DNA injection, four hundred embryos were transferred to M16 medium and incubated at 37°C in a 5% CO₂ atmosphere for one to two hours. One hundred fifty embryos survived microinjection. Lysed embryos were discarded and 30 embryos that appeared normal were transferred to one of the fallopian tubes of each of 5 pseudopregnant foster mothers. The transfers were performed under a dissecting microscope using general anesthesia (avertin).

Seven pups were born. After birth, newborn mice were kept with their foster mothers for 2 weeks, at which point they were then weaned and screened for DNA integration. A 2 cm portion of the tail was removed and homogenized in 2 ml of a solution of 0.1 M NaCl, 50 mM Tris-HCl, pH 7.5, lmM EDTA for short duration, but long enough to disrupt cell and nuclear membranes. The homogenized tissue was treated with 50 U/ml RNaseA and 0.1% SDS for 15 minutes at 37°C. The mixture was exposed to Proteinase K digestion for 3 hours at 55°C followed by three extractions with phenol/chloroform. DNA was then precipitated by the addition of ethanol. After resuspending the precipitated DNA in 10 mM Tris pH 8.0, 0.5 mM EDTA, some of it was digested with

BamHl endonuclease and electrophoresed through an 0.8% agarose gel. The DNA was denatured by soaking the gel in 1.5 M NaCl, 0.5 M NaOH for one hour and then neutralizing the DNA by soaking it in 1.5 M NaCl, 0.5 M Tris, pH 7.4 for 30 minutes. The gel was then soaked in 10X SSC for one hour. The DNA was then transferred from the gel into a nitrocellulose filter by the method of Southern, as described in Maniatis, supra.

The filter with transferred DNA was hybridized overnight with ³²P labeled lambda DNA prepared

according to standard procedures by the method of nick translation. Maniatis, supra. Following this

overnight hybridization, the filter was washed in 0.1 x SSC, 0.1% SDS at 50°C and Kodak XAR film was exposed to it in order to identify lambda DNA present within the mouse genome. Lambda DNA, used as standards, that had been electrophoresed alongside the mouse genomic DNA were compared in intensity to the transgenic mouse DNA hybridized to the ³²P labeled lambda DNA to

estimate the number of copies of test DNA per mouse cell. Three transgenic animals have been produced and identified by this technique.

Newborn mice tested for the presence of the test DNA sequence by the tail-blotting procedure (Hogan et al, Manipulating the Mouse Embryo: A Laboratory

Manual, pp. 174-183, Cold Spring Harbor Laboratory, 1986) were found to carry the test DNA sequence in DNA isolated from their tails. Eight weeks after birth these transgenic mice were mated and their progeny were examined for the test DNA sequence.

Approximately 50% of the resulting offspring carried the test DNA sequence, demonstrating that the original transgenic mice carried the test DNA sequence in their germ line and that this sequence was inherited

normally. While transgenic lines having approximately one copy of the test DNA sequence per cell can be obtained, it will be understood by one skilled in the art that multiple copy numbers per cell are obtainable and may be useful for many different applications.

The above procedure has been utilized with the L2B vector as the source of test DNA sequence, and has also been utilized with other test DNA sequences, including lambda LIZ Alpha, described herein.

The resulting transgenic mouse containing the test DNA sequence lambda LIZ alpha was deposited with the ATCC in the form of a frozen mouse embryo as described further herein. C. Mutagen Treatments

Six to eight week old transgenic male mice were treated on day 1 and day 4 by intraperitoneal injection of either 125 or 250 mg N-ethyl-N-nitrosourea (EtNu), per kg body weight. Control animals were injected with 100 mM phosphate buffer at 10 ml/kg body weight. Tissues were collected two hours after final injection.

D. Recovery of the Test DNA Seguence and

Mutagenesis Testing

(1) Recovery

In the embodiment described here, rescue of the marker (test) DNA sequence was

accomplished by containing it within a lambda

bacteriophage genome. The entire lambda bacteriophage genome is excised from the mouse chromosome by the in vitro packaging extract. The packaging extract recognizes the cos sites of the integrated lambda DNA and packages the sequences between the cos sites into lambda phage particles, as shown in Figure 1.

The test DNA sequence may be found within the genomic DNA purified from any tissue of the transgenic mouse. Since the test DNA sequence is contained within a lambda phage genome, it can be excised away from the remainder of genomic DNA by using a lambda phage packaging extract. Packaged lambda phage such as L2B or C2B, may then be plated on E. coli cells for further evaluation.

Bacteriophage lambda DNA can be packaged in vitro using protein extracts prepared from bacteria infected with lambda phage lacking one or more genes for producing the proteins required for assembly of infectious phage particles. Typical in vitro

packaging reactions are routinely capable of achieving efficiencies of 10⁸ plaque forming units (pfu) per μg of intact bacteriophage lambda DNA. About 0.05 - 0.5 percent of the DNA molecules present in the reaction can be packaged into infectious virions.

Various genetic mutations affect different stages of bacteriophage lambda DNA packaging. For instance, the E protein is the major component of the

bacteriophage head and is required for the assembly of the earliest identifiable precursor. Bacteriophages mutant in the E gene (E-) accumulate all of the components of the viral capsid. The D protein is localized on the outside of the bacteriophage head and is involved in the coupled process of insertion of bacteriophage lambda DNA into the "prehead" precursor and the subsequent maturation of the head.

Bacteriophages mutant in the D gene (D-) accumulate the immature prehead but do not allow insertion of bacteriophage lambda DNA into the head. The A protein is involved in the insertion of bacteriophage lambda DNA into the bacteriophage prehead and cleavage of the concatenated precursor DNA at the cos sites.

Bacteriophages mutant in the A gene (A-) also

accumulate empty preheads. Complementing extracts have been prepared from cells infected with A- and E- or D- and E- strains; alternatively, extracts prepared from cells infected with A- mutants can be

complemented by the addition of purified wild-type A protein.

A bacteriophage lambda DNA packaging extract is a proteinaceous composition that is capable of packaging bacteriophage lambda DNA into infectious virus particles. Preferably, the lambda DNA packaging extracts useful in this invention have a packaging efficiency of at least 10⁸, and more preferably at least 10⁹, pfu/μg of intact lambda DNA.

The packaging extracts of this invention are usually prepared from cells containing bacteriophage lambda lysogens of the appropriate genotype, e.g., amber mutations in genes A, D, E and the like. In addition to lacking a functional lacZ gene, useful lysogens preferably have one or more of the following mutations:

clts857 - specifies a temperature-sensitive

bacteriophage lambda repressor molecule. This mutation causes lambda DNA to be maintained in the lysogenic state when the host bacteria are grown at 32°C;

bacteriophage growth is induced by transiently raising the temperature to 42- 45°C to inactivate the repressor specified by the cl gene.

Sam7 - an amber mutation in the bacteriophage S gene that is required for cell lysis. This mutation causes capsid components to

accumulate within SuIII- bacterial cells for 2-3 hours following induction of the clts857 lysogen.

b-region deletion (b2 or b1007) - a deletion in the bacteriophage genome that effectively removes the lambda DNA attachment site

(att). This mutation reduces, but does not entirely eliminate, the packaging of

endogenous lambda DNA molecules in extracts made from the induced cells.

Red3 (in lambda) and RecA (in E. coli) - mutations that inactivate the generalized recombination systems of bacteriophage lambda and the host, thereby minimizing recombination between the endogenous lambda DNA in the extract and the exogenously added recombinant genomes.

Thus, in preferred embodiments, a lambda lysogen useful for producing a packaging extract is one deficient in one or more of the McrA, McrBC, McrF, HsdR and Mrr restriction systems. The genes

comprising these systems can be removed or inactivated by well known methods, such as by transduction, transposon (Tn) mutagenesis, and the like.

Packaging extracts are usually prepared from a lysogenic bacteria having one or more of the following mutations: McrA-, McrBC-, McrF-, Mrr-, HsdR-, and preferably prepared from K-12 ▴mcrB region, BHB2690R-, or BHB2668R-, by growing the appropriate lysogenic bacteria to mid-log phase at 32°C, inducing lytic functions by inactivating the cl repressor protein by raising the temperature to 45°C for 15 minutes, and then growing the cultures for an additional 2-3 hours at 38-39°C to allow packaging components to

accumulate. Cell extracts are then prepared as described further herein.

(2) Testing for Mutagenesis

For plating bacteria, β-galactosidase deficient E. coli, are grown in IX TB (5g/L NaCL, 10g/L tryptone) supplemented with 0.2% maltose and 10mM MgSO₄ overnight at 30° C. Cells are harvested by centrifugation and resuspended in 10mM MgSO₄ in preparation for plating (Maniatis, supra).

In a typical experiment, 1-5 μg of genomic DNA are exposed to in vitro lambda phage packaging extract and incubated for 2 hours at room temperature. The packaging reaction is then diluted in 500 μl SM buffer (100mM NaCL, 8mM MgSO₄, 50mM Tris, pH 7.5, and 0.01% gelatin) and incubated with the above described bacteria (2.0 mL of OD₆₀₀ = 0.5), and then plated onto NZY/agar Nunc Bioassay Dishes (245mm × 245mm × 20mm) with molten top agar containing 1.25 mg/mL X-gal and 2.5 mM IPTG at a density of less than 20,000 pfu per plate. The plates are incubated overnight at 37°C.

For the test DNA sequence-containing lambda genomes that have the β-gal (not the lacI) gene, in the presence of X-gal and IPTG, the phage plaques turn blue if the beta-galactosidase sequence within the lambda genome has not mutated. However, a white plaque or faint blue plaque on the petri dish is evidence that a mutation in the beta-galactosidase sequence has, for example, altered the reading freme, altered essential codons, or has created stop codons in the sequence. These white or faint blue plaques will be scored as positive for mutations and they can be plaque purified and saved for further analysis. The ratio of white or faint blue to blue plaques minus background provides a numerical measure of the

mutagenic potential of the mutagen being tested.

Background is the measured mutation rate of the agent being tested when compared with DNA extracted from mice that have not been treated with potentially mutagenic agents.

Other testing for mutagenesis protocols are described elsewhere herein, such as those protocols using the lac repressor (lac I or lac I^q) as the target gene, and those using selectable systems that include a reporter gene that selects for the growth of host cells containing the mutagenized test DNA

sequence. E. Methods for Increasing Efficiency of Test DNA Sequence Rescue

(1) Demethylation

It is anticipated that test DNA sequence rescue efficiency can be influenced by the state of CpG methylation in the mouse chromosome. Highly methylated DNA may not be efficiently excised by lambda packaging extract, presumably because of inhibition of cleavage at the cos sites, inhibition of expression of lambda genes encoded on lambda phage, or restriction by E. coli restriction systems. This may be alleviated by placing transcriptional enhancers, promoters and/or other regions of the DNA which inhibit methylation near critical sites such as the cos site to reduce CpG methylation. The drug 5'-azacytidine can also be used to reduce the level of DNA methylation in the target cells prior to DNA purification and rescue. Jaenisch et al, Proc. Natl. Acad. Sci. USA, 82:1451-1455 (1985). In such a procedure, fibroblast cell lines are obtained from organisms containing the test DNA sequence of

interest. Adams, Cell Culture for Biochemists, pages 68-83 (1980) Elselvier/North Hollan Biomedical Press). The cells are exposed in vitro at 37°C, within 50 μM 5'azacytidine supplementing the culture medium. Upon DNA replication, the daughter DNA loses its CpG methylation, which eliminates the methylation of sites in the target vector, where the target vector is a lambda phage. The DNA from these fibroblasts is then exposed to in vitro packaging extract, as previously described.

Alternatively, organisms containing the test DNA sequence can be directly injected with a 1 mg/ml solution of 5'-azacytidine in 0.15 M NaCl. This is done over a period of at least about 4 days, with a total of 400 μg administered. Jaenisch, supra. After this treatment, DNA can be extracted from various tissues and packaged as before.

(2) Removal of Packaging Extract and Plating

Strain Restriction Systems

We have determined that the efficiency of test DNA sequence recovery is dependent on the

genotype of both the bacterial strain used to generate the packaging extract as well as the plating strains used for mutagenesis testing. This is due to, at least in part, host-controlled restriction

modification systems (or restriction systems) that enable a bacterial cell to identify and inactivate foreign DNA by endonuclease cleavage. DNA is

susceptible to restriction by the endonucleic activity of the host unless it is protected by modifications, such as methylation of specific nucleotides. While methylation of specific nucleotides usually serves to protect DNA from restriction by the endonucleolytic activity of the host, methylation at some DNA

sequences actually confers sensitivity to restriction. One example, the McrB restriction system of E. coli K-12, is responsible for the biological inactivation of foreign DNA that contains 5-methylcytosine residues. Ross et al, Journal of Bacteriology, 171:1974-1981 (1989).

There are a number of restriction/methylation systems endogenous to E. coli which are capable of inactivating foreign DNA by endonuclease cleavage.

The most widely known systems are hsd (Bickle,

Nucleases, p. 85, Cold Spring Harbor Lab., Cold Spring Harbor, N.Y. 1982), Mrr (Heitman et al, J. Bacteriol., 169:3243-3250 (1987)), McrF (Kretz et al, X.

Bacteriol., 173:4707-4716 (1991); Kelleher et al, J. Bacteriiol., 173:5220-5223, (1991)). McrA (Raleigh et al, PNAS. 83:9070-9074 (1986)) and McrBC (Raleigh, supra). The Hsd system works by selectively

restricting DNA that is not protected by adenine methylation at the N-6 position in the sequence, A^6meACNNNNNNGTGC or GC^6me-ACNNNNNNGTT. The Mrr system also involves adenine methylation, however, in this case the methylation does not serve to protect the DNA, but serves to make the DNA vulnerable to the restriction system. The mrr gene system has recently been shown to also recognize and restrict cytosine methylated sequences. This activity of the mrr gene has been named McrF. The systems McrA and McrBC are similar to Mrr in that they recognize and restrict methylated DNA. However, these two systems differ from Mrr in that they have been shown to recognize only methylated cytosine. Further, the McrB function is provided by the products of at least two genes, mcrB and mcrC (Ross et al, J. Bacteriol.. 171:1974-1981 (1989)). The recognition sequences for mcr and mrr are contemplated in the literature, but precise sequences are as yet unknown.

We found that efficiency of recovery of the lacZ construct from the transgenic animal genome was increased without the use of 5-azacytidine, by using lambda packaging extracts and E. coli plating strains lacking restriction systems that cleave eukaryotic DNA. By removing these restriction systems, rescue efficiencies have been increased up to at least 70,000 pfu/μg genomic DNA. Of course, one skilled in the art will recognize that "removal" of these restriction systems may be effected by deleting or inhibiting the activity of these restriction systems, and the term "restriction system deficient" includes, but is not limited to, removal of the restriction systems by either method. In addition, naturally occurring strains of E. coli that are deficient in these

restriction systems may be isolated and used.

Identification of the genes responsible for the

E. coli restriction systems was achieved by

examination of the inhibitory effect of certain E.

coli strains on the ability to recover lambda phage.

Isolation of the responsible genes was achieved

through the use of interrupted matings and P1

transduction. An approximately 200 kb region of DNA

in E. coli K-12 was found to produce an inhibitory

effect on the plating efficiency of the rescued

vector. Further, the region responsible for

decreasing rescue efficiency was found to be near 98

minutes in the E. coli K-12 genetic map (Bachmann, E.

coli and S. Typhimurium: Cellular and Molecular

Biology, eds. Neidhart et al, ASM, WA, DC, 1987) in

the mcrB region containing mcrBC-RMS-hsd-mrr.

A comparison study of the rescue efficiency using

E. coli strains with different restriction genotypes

was performed using the strains shown in Table 1. The

bacterial strains listed in Table 1 are available from

the following source or reference: ED8767 (Ishiura et

al. Anal. Biochem.. 176:117-127 (1988); ER1451 (New

England BioLabs, Beverly, MA); LCK8 (Bachman, Yale E.

coli Center); NM621 (Murray, Univ. of Edinburgh);

K802, LE392, NM554 , PLK-A, PLK-17, Y1088, E. coli C,

Sure (Stratagene, La Jolla, CA)).

Table 1

Strain Restriction Genotype Plating Efficiency hsdR mcrA mcrB mrr

ED8767 - - - + - ER1451 - - - + - K802 - - - + - LCK8 - - - + -

LE392 - - + + -

NM554 - - - + -

NM621 - - - + -

PLK-A - - - + -

PLK-17 - - - + -

Y1088 - - + (+ ) -

E. coli C - - - - +

RR1-A - - - - +

K-12 -uncrB ▴ - ▴ ▴ +

Sure^Tm ▴ - ▴ ▴ +

Strain RR1-A and K-12▴mcrB are constructed as described below.

Strain RR1-A is constructed with strain RR1

(Maniatis, supra) (relevant genotype = McrA⁺, (tet^s)) as the recipient and any E. coli K-12 strain that carries a Tn10 (tetracycline resistant) in (or near) the mcrA gene (relevant genotype = McrA: Tn10 (tet^R)) as the donor. Step 1: A P1 lysate is made from the E. coli K-12 strain described above. Step 2: RR1 is transduced (Miller, J., Experiments in Molecular

Genetics, Cold Spring Harbor Lab., Cold Spring Harbor, New York (1972)). Step 3: Tetracycline resistant colonies are selected and purified. Step 4: Loss of tetracycline resistance is selected for on Bochner plates (Bochner, B.R., et al., J. Bacteriol., 143:926-933 (1980)), and colonies are purified. Step 5: lack of McrA restriction activity is tested by comparing transformation efficiency of unmethylated pBR322 versus pBR322 that has been in vitro methylated by Hpall methylase (Raleigh, supra). A McrA⁺ strain will show a greatly reduced efficiency with the methylated plasmid. If McrA activity is absent, this strain is then called RR1-A. Strain K-12AmcrB is constructed using two donor E. coli K-12 strains with the relevant genotypes McrB: :Tn10 (tet^R), Mrr: :Tn5 (Kan^R) and McrA: :Tn10 (tet^R) and a recipient E. coli K-12 with the relevant

genotype RecA⁺, tet^s. Steps 1-5: Perform steps 1-5 as described for construction of RR1-A. In step 2 , transduce any E. coli K-12 RecA⁺ strain. Step 6:

Make a P1 lysate from an E. coli K-12 strain that carries a Tn10(tet^R) in the mcrB gene. Transduce the K-12 RecA⁺(tet^s) strain. Step 8: Select the tet^R colonies. Purify one colony that is also kan^R. Step 9 : Select for loss of tet^R on Bochner plates

(Bochner, supra). Step 10: Purify several colonies and test for sensitivity to tetracycline and

kanamycin. Select colonies that are both tet^s and kan^s. Step 11: Test for lack of McrB restriction activity as done for the McrA tet, however in this case, the pBR322 should be in vitro methylated by Alul methylase (Raleigh, supra; Ross, supra). A McrB⁺ strain will show a greatly reduced efficiency with the methylated plasmid. Test for Mrr restriction activity by comparing plating efficiency of lambda versus lambda which has been in vivo methylated by Pst I methylase (Heitman, supra). An Mrr⁺ strain will show reduced efficiency with the methylated lambda. Test for HsdR restriction activity by comparing plating efficiency of lambda versus lambda which has been in vitro methylated by Hsd methylase (Wood, J. Mol.

Biol., 16:118-133 (1966); Adams, Bacteriophages, New York: Interscience 1959; Bickle, supra, at pp. 95-100). An HsdR+ strain will show reduced efficiency with unmethylated lambda. If a strain (purified colony) lacks all restriction activities, namely, McrA, McrBC, McrF, Mrr, HsdR and was constructed by this method, it should then contain a deletion throughout the McrB region (▴mcrB). It will then also very efficiently plate lambda that has been rescued from the mouse. This strain is called K-12▴mcrB.

The "A" symbol in Table 1 indicates that the strain contains a large deletion in the mcrB region. All other McrB- strains listed in Table 1 are K-12 derivatives believed to contain a small mutation in the mcrB region, with the exception of E. coli C which does not contain the K-12 mcrB region, and RR1-A which carries the wild type mcrB locus of E. coli B. It is known that all of these strains plate control L2B phage (amplified in HsdM+ E. coli K-12 rather than rescued from the mouse) with equal efficiency (within 1-4 fold). Rescued L2B phase were recovered from the mouse genome using Mcr- E. coli K-12 lambda packaging extracts (Gigapack II - Stratagene, La Jolla, CA) and plated onto the indicated bacterial strains. A "+" plating efficiency of phage indicates that

approximately 500 pfu/0.05 μg of transgenic mouse genome DNA was observed, while a "-" plating

efficiency indicates that less than 5 pfu/0.05 μg of transgenic mouse genome DNA was observed. Note also that (+) indicates that the Mrr activity has not been confirmed in Y1088.

In order to determine more precisely the region of DNA responsible for the inhibition of dC-methylated lambda phase genome, the 98 minute region of E. coli K-12 LCK8 was cloned. A partial LCK8 genomic library was made in pOU61cos. (Knott, Nucleic Acid Res., 16:2601-2612 (1988)), packaged with Gigapack™ II XL

(Stratagene, La Jolla, Calif.), and plated on E. coli C. Clones containing the 98 minute region were identified by colony hybridization using an

oligonucleotide (ATGAGTGCGGGGAAATTG) probe specific to the Hsd region (Gough et al, J. Mol. Biol., 166:1-19 (1983)). All clones were propagated in the host RR1-A when tested for plating efficiency of phage. As shown in Figure 3, in panels 3A and 3B, the activity was isolated to a 2.6 kb fragment containing the mcrB gene. The mcrB region including open reading frames (Ross et al, supra) is shown in Figure 3A. The subclones corresponding to these groups are shown directly below. The table on the far left gives information pertaining to the DNA fragment shown on the right. (The restriction map depicted in Figure 3A and 3B, showing the location of the hsdS gene and adjacent mcrB region of the E. coli K12 chromosome, is from Ross et al, J. Bact.. 171:1974-1981 (1989).)

The results in the above study support the observation that the restriction activities of the minute 98 region have a negative effect on rescue efficiency. To obtain high plating efficiencies, a complete deletion of the minute 98 mcrB region (mcrB through mrr) is preferred, as opposed to a small mutation of mcrB present in most commonly used McrB- lab strains. This is because despite the McrB- phenotype exhibited by these McrB- strains (using Alul methylase modified pBR322 transformation as the assay (Ross, supra.)) some inhibitory activity of the mcrB region remains. Complete deletion resulted in optimal efficiency, accounting for a greater than 1000-fold improvement in rescue efficiency using eukaryotic modified DNA.

Preferred E. coli strains for rescue of the lacZ construct from the transgenic animal genome are SCS-8 (Catalog Number 200,288) and VCS257 (Catalog Number 200,256) which are commercially available from

Stratagene Cloning Systems and are also contained in a kit (Big Blue™ Mouse Mutagenesis System, Catalog Number 720,000). SCS-8 has the following genotype: RecA1, endA1, mcrA, ▴(mcrBC-msdRMS-mrr), ▴(argF-lac)U169, phi80▴lacZ▴M15, Tn10(tet^r). SCS-8 provides the lacZ▴M15 gene which allows for alpha-complementation when SCS-8 is infected by the packaged bacteriophage. Additional commercially available E. coli strains which contain the lacZ▴M15 genotype for use in this invention include the following: XL1-Blue

(Stratagene, Catalog Number 200,236); SureTM

(Stratagene, Catalog Number 200,238); PLK-F'

(Stratagene, Catalog Number 200,237); JM101

(Stratagene, Catalog Number 200,234) ; JM109

(Stratagene, Catalog Number 200,235); and NM522

(Stratagene, Catalog Number 200,233).

While the use of mcrB deletion strains is

described herein for use in mutagenesis testing and recovery of lambda phage DNA from mammalian cells, it is apparent that restriction system deficient strains may be used for other eukaryotic DNA cloning projects.

Of course, any number or variety of test DNA sequences or genes can be inserted between lambda cos sites. The in vitro packaging extract would still excise the DNA between the cos sites and package it into a lambda phage particle. Thus, a variety of recombinant lambda genomes or cosmids may be used for this excision event.

F. Construction of Shuttle Vector Systems for Rapid DNA Sequence Identification of

Mutations in Test DNA

Mutations evidenced by the production of white plaques resulting from disruption of the β-galactosidase (β-gal) gene are useful for determining the mutation rate of a mutagen, but give little information regarding the specific mutation within the DNA. In addition, analysis of the specific mutation is hampered somewhat by the size of the test β-gal gene (i.e., about 3200 b.p.).

To help increase the effectiveness of the

procedure, the target lambda phage can be made to provide a target gene with reduced size (e.g., the lacl gene having about 1000 b.p.), and a rapid means with which the target gene is transferred from the lambda phage into plasmid vectors for sequence

analysis.

Both the lacl and β-gal genes are inserted within a lambda vector, such that if the mutation occurs within the lacl gene, the repressor activity is lost allowing the β-gal gene to be expressed giving rise to blue plaques in the absence of IPTG. In the described embodiment, the lacI gene is positioned upstream of the alpha portion of the lacZ gene in the vector

(Miller et al. The Operon, 2nd Ed. Cold Spring Harbor Laboratory, 1980, pp. 104-105). When the host E. coli (which is infected by the bacteriophage vector) provides the complementary portion of the lacZ gene (referred to as lacZ▴M15) (Miller et al, supra), the gene products synthesized by these two partial genes combine to form a functional β-galactosidase protein (referred to as alpha-complementation) giving rise to blue plaques in the presence of Xgal when a mutation has occurred in the lacI gene or in the presence Xgal and IPTG when the lacI gene is not mutated. The AM15 portion of the lacZ gene provided by the host is provided either episomally (via a low copy number plasmid or F-factor) or stably integrated into the bacterial chromosome. The alpha portion of lacZ is used because 1) the β-gal protein formed by alpha-complementation is known to be weaker in activity than the contiguous protein, minimizing the possibility of background blue plaques due to inefficient repression by lacI, and 2) to provide a smaller and thus more easily characterized lacZ target should this gene be used in mutagenesis studies. The requirements of the host E. coli in this system are the following:

lacl(-), lacZ▴M15, restriction (-). All cloning steps are outlined in the figures 4 through 8 and are done using standard procedures (Sambrook et al, Molecular Cloning, A Laboratory Manual, 2nd. Ed. Cold Spring Harbor Laboratory 1989).

The embodiment described utilizes the alpha portion of lacZ with lacI. The complete lacZ can also be used by providing a means to maintain complete repression by lacI until induction is desired. This can be done in a variety of ways including control of▴M15 laZ expression by a lambda specific promoter

(PR') which prevents lacZ expression in the host E. coli until several minutes following infection by the bacteriophage, allowing lacI levels to build up to suitable levels to enable complete repression.

Additionally, low levels of lac repressor can be maintained in the host to assist in repression by lacI until induction occurs, either by a mutation in lacI or by addition of IPTG to the system. A third

alternative is to use an altered lacI gene which gives rise to a repressor protein with higher specific activity, thereby allowing stronger repression of β-galactosidase production.

The source of starting materials for the cloning procedures are as follows: the pBluescript II SK+ and SK-, pBS(+), lambda gtll, and lambda L2B are available from Stratagene Cloning Systems, La Jolla, CA. Lambda L47.1 and pPreB: Short et al, Nucleic Acids Res., 16:7583-7600 (1988). pMJR1560 is available from

Amersham Corp., Arlington Heights, Illinois.

Rapid sequencing of the mutagenized lacI gene within the lambda vector is facilitated by

incorporating "lambda ZAP" excision sequence within the lambda vector. (Short et al, Nucleic Acids Res., 16:7583-7600 (1988). Lambda ZAP is a lambda phage vector which permits in vivo excision of inserts from the lambda vector to a plasmid. This is possible because the lambda phage contains the two halves of an fl bacteriophage origin of replication. In the presence of proteins supplied by f1 helper phage, all DNA present between the two partial f1 origins is automatically excised from the lambda phage. The two halves come together to form an intact f1 origin. The resulting phagemid contains a Col E1 origin of

replication and an ampicillin resistance gene, thus the insert is effectively subcloned into a plasmid vector. All sequences between the two partial f1 origins are excised as a plasmid within hours.

In the mutation analysis vector, these fl origins are positioned so that the lacI gene can be

automatically excised from the lambda phage from the mouse genomic DNA. Following this conversion from phage to plasmid, the insert may be rapidly sequenced or characterized by other known methods.

Characterization of a large number of mutations within the lacI gene can be completed within 3 days following isolation of mouse genomic DNA, as opposed to several months using standard techniques.

In the example described herein, a lambda ZAP is used to convert the test DNA inserts from integration in the lambda vector to a plasmid. Other systems may also be used which allow excision and

recircularization of a linear sequence of DNA thereby providing a rapid means with which the test DNA sequence may be transferred from the phage to a form suitable for analysis. Such other systems include, but are not limited to, the use of FLP-mediated

(Senecoff et al, Proc. Natl. Acad. Sci. USA, 82:7270-7274 (1985); Jayaram, Proc. Natl. Acad. Sci. USA, 82:5875-5879 (1985); McLeod, Mol. Cell. Biol., 6:3357-3367 (1986); Lebreton, B. et al., Genetics, 118:393-400 (1988)) or Cre-lox site specific recombination techniques (Hoess et al, J. Mol. Biol., 181:351-362 (1985); Hoess et al, Proc. Natl. Acad. Sci. USA, 81:1026-1029 (1984)).

The embodiments described above utilize the E. coli beta-galactosidase gene as a test DNA sequence, which allows phenotypes that are positive and negative for mutation to be observed. Other potential test DNA sequences include (but are not limited to): the lac I repressor, the cl repressor, any antibiotic resistance gene sequence (ampicillin, kanamycin, tetracycline, neomycin, chloramphenicol, etc.), the lambda Red and Gam gene sequences, a thymidine kinase gene sequence, a xanthine-guanine phosphoribosyl transferase gene sequence, sequences that code for restriction enzymes or methylation enzymes, a gene sequence that codes for luciferase, and/or a tRNA stop codon or frameshift suppressor gene sequence.

Even more general models can be made that

eliminate the cos sites, although the excision

mechanism now becomes different. By bracketing the test (marker) DNA sequence (s) with convenient

restriction sites, as shown in Figure 2, the test sequence (s) can be separated away from the mouse DNA with restriction enzymes (enzyme rX) and subsequently ligated with restriction sites of lambda or cosmid vectors which contain cos sites, or if the test sequence is linked to a replication origin it can be transformed directly. Background can be reduced in such a system by including with the test DNA sequences a sequence that is necessary for lambda phage

replication, which is then cloned with the test DNA sequence into a lambda genome deficient or defective in that sequence.

5. Preparation of a Modified Lambda Genome

Containing a Target Gene System

An exemplary modified Lambda genome, designated Lambda LIZ alpha, is prepared through a series of molecular gene manipulations as diagrammed in Figures 4 through 8.

Figure 4 depicts the construction of pBlue MI-. pBluescript SK- (Stratagene, La Jolla, California) is modified using site directed mutagenesis to introduce an AvaIII restriction site at a position 5' to the open reading frame for the lacI gene, but downstream from the ampicillin resistance gene and the ColE1 origin of replication present on pBluescript to form pBlue MI-.

Figure 5 depicts the construction of pLadq.

pBluescript II SK+ (Stratagene) is digested with the restriction enzymes Pstl and EcoRI, both which cleave in the polylinker region to form linearized

pBluescript SK+ lacking the small fragment derived from the polylinker. pMJR1560 (Amersham Corporation, Arlington Heights, Illinois) is digested with the restriction enzymes Pstl and EcoRI to release a lacl^q-containing fragment that is separated by agarose gel electrophoresis and eluted from the gel. The lacl^q-containing fragment is then ligated into the

linearized pBluescript SK+ to form pLadq.

Figure 6 depicts the construction of plnt.l. A double stranded DNA segment defining multiple cloning sites (a polylinker) is produced by synthetic

oligonucleotide synthesis and annealing. The polylinker contains multiple restriction endonuclease recognition sequences including two AvaIII sites flanking Xbal, Kpnl and Pvul sites. The polylinker is digested with AvaIII to form AvaIII cohesive termini on the polylinker. pBlueMI- is digested with AvaIII and the polylinker is ligated into pBlueMI-to

introduce the Pvul site into the AvaIII site and form pint.1

Figure 7 depicts the construction of pPreLadqZ (pPRIAZ). To that end, pPre B is first prepared as described by Short et al, Nucl. Acids Res., 16:7583-7600 (1988).

To prepare pPre B, plasmid pUC 19 (ATCC #37254) described by Yanisch-Perron et al, Gene, 33:103-119 (1985), was digested with EcoRI, dephosphorylated and ligated to complementary oligonucleotides, each having compatible EcoRI ends and defining a T7 RNA polymerase promoter as described by McAllister et al, Nucl. Acids Res., 8:4821-4837 (1980) to form pJF3 having the T7 promoter oriented to direct RNA synthesis towards the multiple cloning site of pUC 19. pJF3 was digested with Hindlll, dephosphorylated and ligated to

complementary oligonucleotides having HindIII-compatible ends and defining a T3 RNA polymerase promoter as described by Morris et al, Gene, 41:193-2000 (1986). A resulting plasmid, designated

pBluescribe (pBS), was isolated that contained the T3 promoter oriented to direct RNA synthesis towards the multiple cloning site of pUC 19. pBS was then

digested with Aatll and Narl, treated with mung bean nuclease and alkaline phosphatase, and ligated to a 456 base pair (bp) Rsal/DraII blunt-end fragment isolated from the pEMBL8 plasmid described by Dente et al, Nucl.Acids Res., 11:1645-1655 (1983). The 456 bp fragment contains the intergenic region of fl phage, but does not contain the f1 gene II promoter sequence. Phagemid clones were isolated from the resulting ligation mixture and clones were isolated containing both orientations (+ or -) of the intergenic region and are designated pBS(+) or pBS(-), where "+"

indicates that the intergenic region is in the same orientation as the lacZ gene. pBluescript SK(-) and SK(+) were produced from pBS(-) and pBS(+),

respectively, by digestion of each with EcoRI and Hindlll, followed by blunt ending with Klenow fragment of DNA polymerase I. The blunt-ended molecules were ligated to a blunt-ended synthetic polylinker

containing 21 unique restriction sites as described by Short et al, Nucl.Acids Res., 16:7583-7600 (1988), to form pBluescript SK(-) or SK(+), respectively. A majority of the terminator portion of the f1

intergenic region is contained on the Rsal (position 5587) to Hinfl (position 5767) restriction fragment isolated from pEMBL8. The remaining terminator sequences were provided by preparing synthetic

oligonucleotides as described by Short et al, supra, to provide a complete terminator, a gene II cleavage signal and unique restriction sites for EcoRV and Ndel. The synthetic oligonucleotide and the Rsal/Hinfl fragment were ligated with a 3009 bp Dral/Ndel

fragment obtained from pBS to form the plasmid pBST-B. The initiator domain of the fl intergenic region was separately cloned by digesting pEMBL8 with Sau961 and Dral to form a 217 bp fragment that was then blunt-ended with Klenow and then subcloned into the Narl site of pBST-B to form pBSITO#12. pBluescript SK(-) was digested with Nael and partially digested to cut only at the Pvul site located adjacent to the f1 origin, and the resulting fragment lacking the f1 origin was isolated. The isolated fragment was ligated to the Nael/Pvul fragment of pBSITO#12 that contains the terminator and initiator regions of the fl intergenic region to form plasmid pPre B.

Lambda gtll (ATCC #37194) was digested with Kpnl and Xbal to produce a 6.3 kilobase (kb) fragment containing the lacZ gene, which was agarose gel purified. pBS(+) prepared above and available from Stratagene was digested with Kpnl and XbaI, and the resulting lacZ fragment was ligated into pBS(+) to form pBS (lacZ). pLadq from above was digested with Narl and Sail and the resulting small fragment

containing the lacl^q gene was isolated. pBS(lacZ) was digested with Narl and Sail and the resulting large fragment containing the lacZ gene was isolated and ligated to the small lacI^q-containing fragment to form pladqZ. plnt.l prepared above was digested with Kpnl, and the resulting linear molecule was ligated to the lacI-lacZ fragment, produced by digesting pladqZ prepared above with Kpnl, to form plntladqZ.

plntladqZ was then digested with Xbal, blunt-ended with Klenow, digested with Seal, and the resulting large fragment containing lacZ-lacI^q and most of the ampicillin resistance gene was isolated. pPre B prepared above was digested with Pvul, blunt ended with mung bean nuclease, digested with Seal and the resulting fragment containing the terminator and initiator fl intergenic region components was isolated and ligated to the plfntladqZ-derived large fragment to form plasmid pPreladqZ (pPRIAZ)

Figure 8, shown in two panels 8A and 8B, depicts the construction of Lambda LIZ Alpha. To that end, Lambda L47.1, described by Loenen et al, Gene, 10:249-259 (1980), by Maniatis et al, in "Molecular Cloning: a Laboratory Manual" , at page 41, Cold Spring Harbor, New York (1982), and by Short et al, supra, and having the genetic markers (srIlambdal-2) delta, imm434 cl-, NIN5, and chi A131, was digested with EcoRI, Hindlll and Smal to form a Lambda L47.1 digestion mixture. Lambda L2B, available from Stratagene, was first digested with Xbal, then treated with Klenow to fill-in the 5¹ Xbal overhang, then digested with MluI to form a L2B digestion mixture. The L47.1 and L2b digestion mixtures were ethanol-precipitated to prepare the DNA for ligation, and were then ligated to pPRIAZ prepared above that had been linearized with Ndel to form Lambda LIZ alpha.

The final construction, Lambda LIZ alpha, is a preferred modified Lambda bacteriophage for use in the present invention because it combines the elements of 1) a reporter gene in the form of the alpha component of lacZ, 2) the lambda bacteriophage excision

capability provided by the presence of the cos sites at the termini, 3) an indicator gene system in the form of the lacl^q target gene, including a lacI

promoter and the repressor structural gene sequences, and 4) an fl origin of replication arranged according to the present invention and as described for the in vivo excision system of Short et al, supra, that allows quick isolation of the mutated test gene from positive colonies containing the mutated test gene for sequencing.

6. Production of Chimeric SCID/hu Mice

A. Preparation of Human Lymphocytes

Peripheral blood lymphocytes (PBLs) are isolated from venous blood drawn from volunteer donors. First, one hundred milliliters (ml) of blood are anticoagulated with a mixture of 0.14 M citric acid, 0.2 M trisodium-citrate, and 0.22 M dextrose. The treated blood is layered on Histopaque-1077 (Sigma, St. Louis, Missouri) to form isolated PBLS which are recovered by centrifugation at 400 X g for 30 minutes at 25°C. Isolated PBLs are then washed twice with phosphate buffered saline (PBS) (150 mM sodium chloride and 150 mM sodium phosphate, pH 7.2 at 25°C). The resulting PBL cell pellet is resuspended in PBS to a concentration of 1 X 10⁸ cells/ml.

Lymphocytes are also isolated from tonsils obtained from therapeutic tonsillectomies from

consenting patients. The tonsils are first

homogenized and then lymphocytes are isolated over Histopaque as described above.

Subject rDNA are then inserted into the isolated lymphocytes using techniques known to one skilled in the art. Preferred techniques are electroporation of lymphocytes and calcium chloride permeabilization of the lymphocytes.

B. Preparation of SCID Mice

SCID mice having the autosomal recessive mouse mutations scid are obtained from Imdyme (San Diego, California). Alternatively, SCID mice are derived from an inbred strain of mice, C.B-17 (Balb/c-C57BL/Ka-Igh-1b/ICR (N17F34) as described by Bosma et al, Nature, 301: 527-530 (1983). Analysis of the pedigree of mice lacking IgM, IgGl or IgG2a determined that the defect was inheritable and under the control of the recessive scid gene. Bosma et al, supra. A colony of mice can be established which are homozygous for the defective gene. The SCID mice are maintained in microisolator cages (Lab Products, Maywood, New Jersey) containing sterilized food and water.

C. Preparation of SCID/hu Chimeras

SCID mice obtained in Example 6b are reconstituted by intraperitoneal injection with at least 5 X 10⁷ human PBLs or tonsil lymphocytes prepared in Example 6a. The recipient SCID mice are designated SCID/hu chimeras which contain the subject rDNA. The human PBL reconstituted SCID mouse model is then used for assaying the effects of mutagens on human cells as described in this invention.

7. Preparation of a Prolysogenic Microorganism

A prolysogenic microorganism (a prolysogen) was prepared as described below, and was used as exemplary of the methods of this invention. The prolysogen was constructed in E. coli strain SCS-8 available from Stratagene (La Jolla, California) using the procedures of Herrero et al, J. Bacterial., 172:6557-6567 (1990), to introduce a stably integrated copy of the lambda cl gene into the genome, and form the E. coli designated SCS-8cI. The genotype of E. coli strain SCS-8 has been described in some detail by Kohler et al,

Proc.Natl .Acad. Sci.USA, 88:7958-7962 (1991). SCS-8 is resistant to the antibiotic tetracycline.

The prolysogen E. coli strain SCS-8cI has been deposited with the American Tissue Culture Collection (ATCC; Rockville, MD) on February 14, 1992, and has been assigned an ATCC accession number 55297.

For the construction of SCS-8cI, the plasmid pUC18Sfi I was first prepared as described by Herrero et al, supra. pUC18Sfi I was derived from pUC18 by adding two SfiI restriction sites flanking the

polylinker. Two oligonucleotide primers were

synthesized that correspond to the termini of the wild type cl gene nucleotide sequence described in "Lambda II" by Hendrix et al, at page 631. These two primers (pi and p2) were designed for use as PCR primers to amplify a cl gene-containing fragment when used on wild type lambda in he form of a E. coli lysogen extract of strain N99. The primers have the nucleotide sequences as follows:

p1 5'-ATCAGCGAATTCCAACCTCCTTAGTACATGCAA-3', and p2 5'-CATACGGTCGACGATCAGCCAAACGTCTCTTCAGG-3'.

p1 (SEQ ID NO 5) includes to nucleotides 38019 to 38039 of lambda, and p2 (SEQ ID NO 6) includes

nucleotides 37225 to 37255 of lambda. The resulting PCR product was inserted into the polylinker region of the plasmid pUC18Sfi I, and then removed from the plasmid by digestion with SfiI, to form a PCR fragment having SfiI cohesive termini. The plasmid pUTKm described and available from Herrero et al,

J. Bacteriol., 172:6557-6567 (1990), was digested with SfiI, and the PCR fragment was ligated into pUTKm to form plasmid pUTKm-cI.

pUTKm contains the 19 base pair ends of the transposon Tn5 required for insertion of DNA fragments into genomic DNA. Between the flanking ends, termed insertion elements, is a selectable marker, the gene for kanamycin resistance (kan), and a restriction endonuclease site for SfiI into which the PCR fragment containing the cl gene was ligated. Outside the insertion elements is the tnp gene encoding the transposase protein required for the transposition function. Thus, upon transposition, the tnp gene is left behind, so that the genomically integrated insertion fragment having the cl and kan genes cannot be excised by transposition in the absence of a transposase protein.

Plasmid pUTKm also contains an R6K origin of replication (ori) that requires the pir gene product for replication. pUTKm was propagated in the S17-1 E. coli strain that contains the pir gene (pir⁺) also described and provided by Herrero et al, supra. S17-1 is sensitive to both of the antibiotics kanamycin and tetracycline. Plasmid pUTKm-cI was transformed into E. coli

S17-1 according to standard bacterial transformation methods, and cultured in kanamycin to select for transformants containing the plasmid, and were

designated sl7-l/pUTKm-d.

For production of a stable integration of the cl gene in pUTKm-cI into the genome of the SCS-8 strain of E. coli, a standard mating protocol was conducted where cultures of SCS-8 and s17-1/pUTKm-cI were coincubated together. Plasmid pUTKm-cI was

transferred by conjugation to SCS-8 cells, and the resulting conjugants were plated onto plates

containing kanamycin and tetracycline. The

tetracycline kills the S17-1 cells, leaving only the SCS-8 cells, which are tet^R. The absence of the pir gene product in the SCS-8 cells prevents the plasmid pUTKm-cI from replicating, thereby providing the selective pressure for transposition to occur in order to maintain the kanamycin gene in the E. coli cells. The resulting viable cells are SCS-8 E. coli

containing integrated cl by transposition, and having antibiotic resistance to both kanamycin and

tetracycline at 50 and 15 ug/ml, respectively. These viable cells are designated SCS-8cI cells, and are exemplary herein as a prolysogenic microorganism because the expression of the cl gene product prevents the cell from entering the lytic phase upon infection by a lysis-competent lambda bacteriophage when

cultured at 30°C as described in Example 8.

8. In Vivo Mutagenesis Testing Assay Using a

Prolysogen

The previous examples described in vivo

mutagenesis assays using either a lacZ or lacI target gene within a lambda shuttle vector. In that assay. genomic DNA from a transgenic mouse was exposed to in vitro lambda packaging extract which allows lambda phage genomes containing the lacZ or lacI target gene to be recovered from the mouse genome. The resulting phage particles were then adsorbed to an E. coli host and plated with top agar on rich NZY media.

Incubation of the plates allowed the formation of plaques which were then scored according to color. When a "phenotypic" mutation was present in the lacI gene, for example, the lac repressor is no longer able to block expression of the alpha-lacZ gene that is present in the recovered phage genome. This alpha-lacZ protein is then able to complement with the omega-lacZ protein that is produced constitutively in the host E. coli strain to form a functional beta-galactosidase protein, referred to as alpha-complementing beta-galactosidase. This protein then breaks down the chromogenic substrate, X-gal, that is present in the top agar media of the plate. As the phage genome replicates in the E. coli cells, a blue plaque will form indicating the presence of a mutant lacI target gene. If a non-mutant lacI target is present, a colorless plaque will form.

Because both mutant and non-mutant targets are scored in this assay, it is considered a non-selectable system for screening for mutagenesis of the target gene. Although this system generated easily identifiable mutants, plating densities cannot exceed 50,000 plaques per 25x25 cm plate and are optimal at densities of 15,000 plaques per plate. Based on these numbers, 10-20 plates are required for each mouse tissue analyzed. The number of plates contributes significantly to the cost of the assay in terms of plates, media, X-gal substrate and labor. The system described in this example is a selectable, and preferred, version of this assay in which only phage that harbor mutant lacI target genes can survive and be identified on the plate. This selection allows a higher plating density to be used: up to 500,000 phage and thus 500,000 target genes, can be screened on one 25x25 cm plate, significantly decreasing the cost and time to perform the assay.

The identification of mutants using the present selectable system depends on the expression of the alpha-lacZ gene to create a functional beta-galactosidase protein. This protein allows the cell to utilize lactose for growth. Thus, cells carrying phenotypically mutant lacI genes can grow on minimal media containing lactose as the sole carbon source. Cells carrying non-mutant lacI genes code for a functional lac repressor and therefore cannot express alpha-lacZ, and die from carbon starvation. Thus the system is selectable for only those target genes which have been mutagenized, thereby inactivating the lacl gene.

The plating media required for this selection is referred to as minimal media containing lactose as the sole carbon source and consists of the following components expressed per liter and are obtained from Sigma Chemical Co. (St. Louis, MO) unless noted differently: 6.0 g Na₂HPO₄, 1.0 g NH₄Cl, 0.5 g NaCl, 20.0 g Bacto Agar, 0.5 g lactose, 0.34 g thiamine HCl, 1.0 ml of 1M MgsO₄, 0.1 ml of 1M CaCl₂. The top agar consists of the following components expressed per liter: 6.0 g Na₂HPO₄, 3.0 g KH₂PO₄, 1.0 g NH₄Cl, 0.5 g NaCl, 3.5 g Seakem Agarose (FMC, Rockland, ME), 0.4 g Difco casamino acids, 2.0 g X-gal (Stratagene Cloning Systems, La Jolla, CA). Note that X-gal was used in this media not to allow distinction between mutants and non-mutants but to allow easier identification of the mutants on the light colored minimal media plates.

Variations in the media formulation can be utilized, however, it was determined that lactose should not be included in the top agar, as this contributed to elevated levels of false positive bacterial cell growth (higher backgrounds). In addition, the rate of growth for the lactose-dependent cells was improved when casamino acids were included in the top agar.

In addition to the change in media that is required for the selectable system to function, a new E. coli strain was also required. Because the

selection depends on survival of the host cells as opposed to plaque formation, it is necessary to inhibit lytic growth of the rescued phage particles once they have adsorbed to the host E. coli cells.

This was done by using a strain of E. coli that carries the lambda cl gene, which strain being

referred to herein as a prolysogenic strain because it maintains the lambda infection in a non-lytic life cycle. The cl gene specifies the lambda repressor protein which allows lambda genomes to be maintained in the lysogenic state as opposed to replicating lytically. In the strain used in this system, SCS-8cl, the cl gene is stably integrated in the E. coli chromosome as described above.

E. coli strain SCS-8cI and the minimal media described above are the two major changes that are incorporated into the protocol described in the earlier Examples 1-6 for the non-selectable version of mutagenesis testing assay of this invention. The general method of the assay was as follows: First, genomic DNA was isolated from transgenic mouse tissue. Approximately 20.0 ul of the DNA was then packaged using Transpack in vitro lambda packaging extract (produced as described earlier) for three hours at 30°C. SM buffer was then added to the reaction tube to give a final volume of one ml. Fifty ul of the reaction was then plated according to the protocol described for the previous non-selectable system. The remaining 900 ul of the packaging reaction was then adsorbed to SCS-8cI cells (prepared as previously described with the exception that the cells are resuspended to an OD of 2.0 after spinning), for 20 minutes at 30°C followed by mixing with 2.5 ml of minimal top agar and pouring onto a 25 × 25 cm plate containing lactose minimal media. This plate was then incubated for approximately 60 hours at 30°C before scoring the blue mutant colonies. The non-selectable plate was scored for total number of plaques to determine the rescue efficiency as previously

described. Note that in this system one packaging reaction was used to set-up two plates: one non-selectable plate that allows determination of rescue efficiency and one selectable plate that allows

~500,000 plaques to be scored for the presence of mutant lacI target genes. In the non-selectable system, ~10~12 plates would be required to screen the same number of targets.

In initial experiments to test the frequency of mutagenic events in a target (test) DNA gene (also referred to as a bioassay) using this selectable system, several different stocks of mutant lacl phage particles (phage having known defects in lacI) were used to determine the rate of lysogenization

obtainable with each particular mutant. First, the various phage were adsorbed to SCS-8cI and plated in top agar on the selective minimal media described above. The plates were incubated at 30°C for 60-65 hours after which time the total number of colony forming units (cfu) was determined.

Thereafter, a second plate was set up using the same kinds and numbers of phage particles used for the first plate, however in this case SCS-8 was the

E. coli host and the non-selectable rich NZY media was used. Following incubation at 37°C for 12 hours the total number of plaque forming units (pfu) was

determined. The total number of cfu was divided by the total number of pfu and multiplied by 100 to give the rate of lysogenization (Rate = cfu/pfu × 100). A similar test was done using a wild type lacI phage stock.

As shown in Table 2 , this selectable system allows very high rates of lysogenization. The data also shows that cells carrying a lysogenized phage with a mutant lacI gene are able to form colonies on this media while those cells carrying phage with non-mutant lacI genes are not able to form colonies. Thus the lysogenization/selection system functions

extremely efficiently.

Table 2

Lysogenization Rates

lacI Phage Stock cfu pfu Rate

Mutant A1 71 68 104%

Mutant D9 425 421 101%

Mutant D10 60 83 75%

Wildtype 0 80 0%

The selectable system was then evaluated for its ability to detect both spontaneous and induced mutant lacI targets rescued from the mouse genome. Following a standard dosing regimen to treat mice with

methylnitrosylurea (MNU), genomic DNA was prepared from the liver of animals treated as described in previous assays, as well as from untreated control animals. Briefly, 6-8 week old B6C3F1 mice were injected intraperitoneally with 100 mg MNU/kg/day for five days. The animals were sacrificed 12 days after the injections, and the genomic DNA prepared and assayed as described. The selectable mouse

mutagenesis assay was then performed on three separate days (experiments 1, 2, 3) as described above.

The results are shown in Figure 9 as mutant frequency versus time. The mutation frequency is the number of mutant colonies obtained divided by the total number of phage particles screened as calculated from the non-selectable plate. The time represents the hours at which the mutant colonies were counted after being placed in the 30°C incubator. The data show that both spontaneous and induced mutant lacI targets can be detected with this selectable assay allowing the calculation of mutant frequencies. The data also demonstrates the reproducibility of the assay as determined by three independent experiments. Additionally, induction rates can be detected with this system as seen by the increased number of mutants observed with the treated (induced) DNA relative to the untreated (control) DNA.

The data in Figure 9 is expressed over time because mutants continue to arise over time. It is possible that a correlation exists between the time that a mutant colony appears and the specific type of mutation that is present in the lacI gene. This may aid in classifying the mutants. For example, a colony that appears early may contain a "strong" mutation while a colony that appears late may contain a "weak" mutation that confers more of a "leaky" lacZ

phenotype.

The selectable system described above is flexible in that the components and concentrations of the components of the minimal media may be varied as can the incubation temperature of the plates. These changes can effect the growth rate of the cells and thus alter the time at which the plates can be scored.

Thus, the selectable version of the previously described mouse mutagenesis assay can be used to determine mutation frequencies and induction rates. In order for this selectable system to function, two major modifications to the original assay were

required: the new E. coli strain (designated as a prolysogen) was constructed, and a different plating media was developed. This system allows ~10~20 fewer plates to be used, thus significantly reducing the cost of the assay in terms of plates, media, X-gal chromogenic substrate, and labor. In addition, the selectable system allows the classification of lacI mutants and permits a wider and/or different spectrum of mutants to be detected.

In a related embodiment, the above selectable system utilizing the SCS-8cI host cell and minimal lactose medium was conducted as above but at 37°C, rather than at 30°C, in order to reduce the growth time. Under these conditions, although lysogenization was induced, the lac repressor was unable to sustain lysogeny at 37°C for prolonged periods, and plaques form in the minimal media top agar after about 40 hours. The casamino acids in the top agar permitted the growth of a low density lawn of E. coli necessary for the plaque formation by the phage harboring the mutant lacI gene. Without the supplement, plaques were not visualized. Thus, the present embodiment allows mutants to be scored as lytic plaques after shorter incubation times of about 40 hours, rather than the longer 60-65 hours required when growth is at 30°C with the lysogenic assay. 9. Construction of a Gene Activating Selectable

System Utilizing GroE and Anti-Terminator Protein Lambda O

A plasmid vector containing a reporter gene under the transcriptional control of a lac promoter and lac operator is constructed having the E. coli groE as the reporter gene and the lambda terminator nucleotide sequences qut-t6s that is regulated by lambda

terminator protein Q. A strain of E. coli that is GroE- is required as the host cell for use of the reporter gene construct.

A. Preparation of a GroE- E. coli Strain

The E. coli strain SCS-8 tet^r described earlier was selected for sensitivity to tetracycline by culturing on Bochner media as described by Maloy et al, J. Bacteriol. 145:1110-1112 (1981) to form SCS-8 tet^s. The strain SCS-8 tet^s was transformed with the RecA⁺ plasmid pJC859 described earlier herein, to form SCS-8 tet^s/pJC859.

A P1 lysate was prepared according to standard procedures from E. coli strain CAG9269 described by Tilly et al, Proc. Natl. Acad. Sci USA, 78:1629-1633 (1981), that is groEL140 Tn10 tet^r, the lysate was used to perform a standard P1 transduction of SCS-8 tet^s/pJC859, and the resulting transductants were selected for tet^r colonies to form SCS-8 tet^r/pJC859. The resulting colonies were screened for the GroE mutant phenotype by infection by lambda phage and identifying colonies that inhibit plaque formation, indicating the presence of the groEL140 mutation.

GroE- colonies were selected and designated SCS-8 tet^r groEL140/pJC859.

Thereafter, E. coli strain SCS-8 tet^r

groEL140/pJC859 was cultured under conditions of rich liquid media (NZY) culture to select for the loss of pJC859 to form SCS-8 groEL140 (tet^r), which has the relevant genotype of groEL140, recA, deltaM151acZ, mcrA, delta (mcrCB-hsdR-mrr) Supº.

The resulting GroE mutant E. coli strain is extremely efficient at inhibiting lambda plaque formation, and no plaques are observed when 4.5 × 10⁸ phage particles are plated on strain SCS-8 groEL140 (tet^r). This level of inhibition is well below the amount of spontaneous mutations of about 10^-5 required for use in a mutagenic testing animal of this

invention.

A second E. coli strain designated SCS-8 groES30 (tet^r) was similarly prepared using the E. coli strain CAG759 described by Fayet et al, J. Bacteriol.,

171:1379-1385, (1989), that is groES30 Tn10 tet^r as the source of the P1 lysate.

In tests of the ability of these GroE mutant strains to inhibit plaque formation, 250, 2.5 × 10⁴, or 2.5 x 10⁶ particles of Lambda LIZ Alpha were plated on several strains to evaluate mutant activity. At all levels tested, no plaques were observed for any of the GroE mutant strains SCS-8 groES30, SCS-8 groEL140 or the parental strains CAG759 groES30 or CAG9269 groEL140, indicating that GroE mutants effectively inhibit plaque formation. In contrast, using the control strain SCS-8 tet^r, 250 input particles yielded 250 plaques, 2.5 x 10⁴ particles yielded a plate confluent with plaques, and 2.5 × 10⁶ particles yielded a plate that was totally cleared, indicating efficient plaque formation at expected levels for wild-type GroE.

B- Preparation of groE Reporter Gene Plasmids To that end, the groE gene was PCR-amplified from a colony of E. coli strain XL1-Blue (Stratagene) using Taq DNA polymerase under standard PCR procedures with the following primers p3 (SEQ ID NO 7) and p4 (SEQ ID NO 8):

p3 5'-GCCGGTCGACCTAGTAAGGACTTTCTCAAAGGAGAGT-3' p4 5'-GTCAGGGCCCCGTGCATGTTATTCCCCATA-3'

The resulting PCR-amplified fragment contains a ribosome binding site, an ATG translational start sequence, and both the groES and groEL genes. A groE gene can readily be isolated from other E. coli

strains. Thereafter, the fragment was restriction endonuclease digested with Apal and Sail, treated with calf intestine alkaline phosphatase (CIAP), gel purified and cloned into pBluescript SK (pBS;

Stratagene) that had been pre-cut with Apal and Sail to form pBSG3.

The resulting plasmid pBSG3 contained the groES and groEL genes under the control of the lac promoter and lac operator in pBS. Plasmid pBSG3 is designated a high copy number plasmid because it contains a ColE1 prokaryotic origin of replication that maintains the plasmid in the host cell at high copy number. Thus host cells used in the assay system of the present invention will contain high copy numbers of plasmid where the reporter gene transcription unit is based on pBSG3.

A second plasmid, having the ability to maintain a lower copy number within a host cell due to the presence of the pl5A prokaryotic origin of replication (replicon) described by Selzer et al, Cell, 32:119-129 (1983), was constructed as follows. Plasmid p15 SK was provided by Ronald Fisher (Federal Republic of Germany). Plasmid p15 SK contains the p15A replicon and is based on the pBluescript SK plasmid except for the replicon, and the presence of the chloramphenicol acetyltransferase gene (cat) selectable marker. Plasmid pBSG3 was digested with Apal and Sail to release the groE gene, and the resulting digested plasmid treated with CIAP. The resulting 2.2 kilobase (kb) fragment containing the groE gene was isolated, purified, and cloned into pl5 SK that was predigested with Apal and Sail to form plasmid p15G10.

Plasmids p15G10 and pBSG3 were tested for their ability to complement GroE mutant E. coli strain

CAG9269 groEL140. First, each plasmid was transformed into the mutant strain to form CAG9269 groEL140/p15G10 and CAG9269 groEL140/pBSG3. About 280 particles of lambda L2B were plated on XL1-Blue, on CAG9269

groEL140, on CAG9269 groEL140/p15G10 and on CAG9269 groEL140/pBSG3, to yield 288, 0, 260 and 161 plaques, respectively. Similarly, about 200 particles of

Lambda LIZ Alpha were plated on XL1-Blue, on CAG9269 groEL140, on CAG9269 groEL140/p15G10 and on CAG9269 groEL140/pBSG3, to yield 203, 0, 212 and 0 plaques, respectively. These results indicate that the

wildtype groE gene present on the plasmid complemented the mutant groE present in the mutant E. coli, thereby providing a suitable host and reporter gene system for practicing the present invention.

Terminator sequences are cloned and inserted into the reporter gene transcription unit of the above described groE plasmids pBSG3 and p15G10. To that end, a DNA fragment containing the lambda terminator sequence qut-t6s was isolated from wildtype lambda by PCR amplification using the primers p5 (SEQ ID NO 9) and p6 (SEQ ID NO 10):

p5 5'-ATCAGCCTGCAGGTCGACATGGGTTAATTCGCTCGTTGT-3' p6 5'-AGCATCGTCGACGGATCCTCTTACCTGTTGTGCAGATAT-3' The resulting PCR-amplified fragment contains the lambda terminator sequence qut-t6s. Thereafter, the fragment was restriction endonuclease digested with BamHI and Sail, treated with CIAP, gel purified and cloned into pDR720 (BRL; Bethesda, MD) that had been pre-cut with the BamHl and Sail to form pDR720Qut. The construction placed the qut-t6s termination sequence upstream of a galactosidase (galK) reporter gene in the vector, which can be utilized to determine if termination of transcription is being regulated by the lambda anti-terminator protein Q.

A galK mutant strain of E. coli HB101 was grown on MacConkey galactose color indicator agar plates, and forms white colonies indicating that no galK gene product is formed. Cells that utilize galactose on MacConkey plates form pink colonies while cells that do not utilize galactose form white or clear colonies. When HB101 was transformed with plasmid pDR720, that expressed the galK gene product, pink colonies were formed, indicating that the pDR720 transcription unit was expressing the galK gene. When HB101 was

transformed with plasmid pDR720Qut, that contained the termination sequence qut-ts6 in the galK transcription unit, white colonies were observed, indicating that the termination sequence prevented transcription of the reporter galK gene.

An E. coli strain designated N99 lambda⁺ is a lambda lysogen that expresses the lambda anti-terminator Q protein at a low level, and that lacks the galK gene. On MacConkey plates, N99 lambda⁺ forms white colonies, indicating that no galK gene product is expressed. When N99 lambda⁺ is transformed with pDR720Qut, pink colonies are formed, indicating that the galK reporter gene transcription unit normally terminated by the qut-t6s sequence is expressed because the Q protein in the host cell is able to anti-terminate (i.e. activate) the terminated galK reporter gene transcription unit. Thus, it is seen that the lambda qut-t6s

termination sequence is effective at terminating the expression of a reporter gene transcription unit, and that the expression of the lambda Q anti-terminator protein is effective in trans to control transcription of the terminated reporter gene.

The cloned lambda terminator sequence qut-t6s was then inserted into the groE reporter gene

transcription unit of both the high and the low copy number plasmids pBSG3 and pl5G10 described above.

To that end, pBSG3 was digested with EcoRI and Spel, phosphatased with CIAP, and the resulting linear vector containing the groE reporter gene transcription unit was gel purified. The terminator sequence was isolated from pDR720Qut by PCR amplification using the following oligonucleotide primers pl3 (SEQ ID NO 19) and pl4 (SEQ ID NO 20):

p13 5'-CGGATCAGCTCTAGAACTAGTATGGGTTAATTCGCTCGTTGT-3' p14 5'-GCGAGCATCAAGCTTGAATTCTCTTACCTGTTGTGCAGATAT-3'

The resulting PCR amplified product was digested with EcoRI and Spel, and gel purified. Thereafter, the fragment containing the terminator sequence was cloned into the EcoRI/ Spel site of pBSG3 to form pBSG3Qut.

A similar construction was prepared using pl5G10, adding the terminator sequence to form pl5G10Qut.

Plasmid pl5G10 was digested with EcoRI and Spel as for pBSG3, and the qut-t6s terminator sequence was

inserted as before.

Thus, for both pBSG3Qut and pl5G10Qut, the qut-t6s terminator sequence was placed inside the groE reporter gene transcription unit upstream of the structural gene, and downstream of the lac promoter, thereby rendering transcription susceptible to

regulation by the anti-termination protein Q. The in trans regulation of the groE reporter gene product by the Q protein is illustrated in Figure 10, showing the use of the test DNA sequence Lambda LIZ Alpha, which contains the lacI test DNA sequence and the lambda Q anti-terminator protein. In addition, the reporter gene transcription unit illustrated in Figure 10 shows the promotion of the groE reporter gene by a lac promoter/operator in which the lacI repressor also controls the transcription of the reporter gene. As shown, when lacI is mutated, the lac promoter/operator is activated, groE is expressed and phage particles are produced in proportion to the amount of mutagenesis on the lacI test DNA sequence. When lacI is not mutated, the lac operator is

repressed, groE is not expressed and plaques do not form.

The presence of the terminator sequence in the reporter gene transcription insures that reporter gene transcription is inhibited in the absence of a test DNA sequence, thereby increasing the sensitivity of the assay.

10. Construction of a Gene Inactivating Selectable System Utilizing Lambda S⁵ and Anti-Terminator Protein Lambda O

In another embodiment, the invention describes a system related to the anti-terminator system described in Example 9, but which includes a dominant negative inactivator gene as the reporter gene (lambda S⁵ mutant) and which includes the use of competing transcription as the means for controlling expression of the reporter gene. The system is illustrated in Figure 11, and is referred to as a reporter gene inactivating system in which plaque formation is inhibited until a mutated lacI test DNA sequence is introduced that inactivates expression of the reporter gene.

Lambda S⁵ is a mutant protein of the multi-subunit membrane pore complex that lambda forms in the inner membrane of an E. coli host cell that is SupF in order for lysis of the host cell during a lytic bacteriophage cycle. The complete membrane pore complex is required for a functional pore to be formed. During the assay, the test DNA sequence

(e.g., Lambda LIZ Alpha) brings in the wild-type S protein. If a single S⁵ mutant protein is present in the pore, the phage particle cannot pass through the inner membrane, and the host cell does not lyse.

Thus, the S⁵ mutation is referred to as a dominant negative inactivator gene, because its phenotype is dominant, not recessive, (i.e., defective even in the presence of wild type S protein) and inactivator because the effect is to inactivate plaque formation.

Other examples of lambda genes suitable as dominant negative inactivator genes for use in a gene inactivating system include any lambda gene in which multiple protein subunits are required for a

functional protein complex such that the presence of a single mutant protein in the complex disrupts the protein complex's function. The lac repressor lacl^d- gene is an example of another dominant negative inactivator gene suitable for use in the present gene inactivating system. The lacl^d- gene is described by Miller et al., in "The Operon", 2nd edition, Cold Spring Harbor Laboratory, 1980.

Competing transcription is also required in the gene inactivating system shown in Figure 11. The system utilizes two transcriptional promoters of unequal transcriptional promotion strength that compete for transcription in converging orientations across the same reporter gene, the weaker promoter (PR¹) producing a transcript that yields reporter gene product, and the stronger promoter (TacP) producing a transcript in the wrong orientation for the reporter gene, thereby blocking expression of reporter gene product. Thus, this system provides an example where the test gene controls expression of the reporter gene by competing transcription.

The gene inactivating system produces a

detectable signal in the form of lytic plaque

formation when mutated test DNA sequences are

introduced because the mutated lacI gene cannot repress the competing TacP transcriptional promoter which is controlled by a lac operator, thus the competing transcript overpowers the weaker PR¹

promoter, the mutant S⁵ reporter gene is inactivated, no S⁵ protein is made, and lytic plaque formation is no longer inhibited. In the presence of functional lac repressor (lacI), the weaker promoter transcribes the S⁵ reporter gene which inhibits plaque formation.

The gene inactivating system shown in Figure 11 preferably contains, but does not require, the in trans terminator/anti-terminator system (Q protein on the test DNA sequence and qut-t6s in the weaker reporter gene transcriptional unit). The

terminator/anti-terminator system ensures that no S⁵ protein can be produced in the host cell until the test DNA sequence is introduced into the host cell.

The construction of a host plasmid vector

containing the reporter gene inactivating system is as follows. Lambda ZA5 (Stratagene) contains the

genotype PR'-qut-t6s-S⁵ on a section of the lambda genome in the order indicated and as required in the reporter gene transcription unit. Therefore a single PCR reaction using Lambda ZA5 produces a PCR-amplified fragment that contains PR'-qut-t6s-S⁵ when

oligonucleotide primers p5 and p6 are used in the PCR reaction. p7 (SEQ ID NO 11) and p8 (SEQ ID NO 12) have the following sequences:

p7 5'-ATCAGCGCTAGCACATATTAACGGCATGAT-3'

p8 5'-AGCATCAGATCTCCACGCCAGCATATCGAGGA-3'

The resulting PCR-amplified fragment was digested with Nhel and BglII, gel purified and ligated into

pBlueMI-, described in Example 5, which had been predigested with Nhel and BglII. The resulting plasmid containing PR--qut-t6s-S⁵ is designated pMI/S⁵.

The nucleotide sequence of the lambda S gene is mutated at two positions to form the S⁵ mutation as described herein. The nucleotide sequence of the S⁵ gene is shown in SEQ ID NO 18. The other lambda sequences defining PR'-qut-t6s in plasmid pMI/S⁵ have wild-type sequences and therefore can be readily prepared from a variety of sources of lambda DNA. If needed, the S⁵ mutation can be introduced into wild-type lambda using well known mutagenesis procedures.

The S⁵ mutation in the lambda S gene requires that the E. coli host cells in which inhibition of plaque formation occurs must harbor the SupF mutation, and when used in this system is referred to as the lambda SY5 system. Thus, the genotype of a host cell for the S⁵ system is restriction minus, RecA-, supF, and preferably lacZdeltaMIS, although the lacZ

mutation is not required unless a color selection is desired in addition to the growth selection.

E. coli strain LE392.23 (Stratagene) which is SupE, SupF was made SupF only by using a P1

transduction lysate as before prepared from E. coli strain CAG1E7 having a wild type SupE/Tn10. The transposition corrected the SupE mutation. The resulting E. coli was grown on tetracycline to insure stabilization of the transduced supE gene.

Thereafter, the strain was screened for its inability to support plaque formation when infected by the lambda strain lambda: : 1105. This lambda strain requires the host cell to provide SupE. Thus those strains which do not support plaque formation have lost the SupE trait. The resulting strain was

designated SCS-15 SupF. E. coli strain SCS-15 SupF is rendered restriction minus using the procedures described in Example 4.E.2. Thereafter, the strain is rendered Rec A- by subjecting it to a standard P1 transduction using a lysate from an E. coli which contained a Tn10 transposon near a mutant recA gene.

The plasmid pMI/S⁵ was then transformed into E. coli strain SCS-15 (supF), and tested for inhibition of lytic plaque formation using Lambda LIZ Alpha. The predicted amount of plaques were observed when the host cell was wild-type SCS-15 (supF) lacking plasmid pMI/S⁵.

In two individual clones of transformed host SCS- 15 (supF) cells, the pMI/S⁵ plasmid produced

sufficient S⁵ protein to block plaque formation. At dosages of 250 plated phage particles, no plaques were observed, and at dosages of 2.5 × 10⁶ plated phage particles, 7-10 plaques were observed, producing a leakiness frequency of about 1 plaque per 10⁵ phage particles. This amount of lysis is acceptable for the assay if color screening is also used because the plaques that do form will be clear rather than blue because only plaques formed in the presence of lacI mutations will appear blue; in the presence of wild type lacI, lacZ alpha on Lambda LIZ Alpha will be repressed and no beta-galactosidase will be formed to yield the blue color.

The S⁵ system provides an example of a selectable system in which mutation in the test gene can be measured by two modes of reporter gene phenotype, simultaneously, namely by a selective growth

advantage, and by the presence of a color selection.

Thus, the test gene product (laql) controls both a first reporter gene transcription unit, namely the S⁵ reporter gene, and controls a second gene (lacZ alpha) that can also function as a second reporter gene. Furthermore, the second reporter gene allows for the capability to produce a detectable color indicator in the host cell.

The above result indicates that the S⁵ mutant is a suitable reporter gene for use in the reporter gene inactivating system described in Figure 11. The completed host vector is prepared by introducing the TacP promoter under the control of a lac operator (lacO). To that end, a PCR-generated fragment

containing TacP-lacO (the "ideal" lac operator

sequence described herein) is produced using the plasmid vector pTL21 (Stratagene) and oligonucleotide primers p9 (SEQ ID NO 13) and plO (SEQ ID NO 14) having the sequences as follows:

p9 5'-ATCAGCACGCGTTTAATGTGAGTTAGCTCACTCA-3' PlO 5'-AGCATCAGATCTCAGCTTTTGTTCCCTTTAG-3'

The resulting PCR-amplified fragment was digested with MluI and BglII, treated as before and ligated into the pMI/S⁵ that was pre-digested with MluI and BglII to form a plasmid designated pMI/S⁵ tac-lac that contains all the genetic elements shown for the host plasmid vector in Figure 11.

The resulting plasmid pMI/S⁵ tac-lac, upon transformation into E. coli SCS-8, is used in the host cell for assaying a test DNA sequence such as Lambda LIZ Alpha recovered from a mutagenized mouse.

Detection of the mutagenized lacI gene is indicated by the presence of blue plaques when the host cell is cultured as described earlier in the presence of X-gal.

The system described in Figure 11 is a prototype for the use of first and second convergent promoters, where the first promoter (lacO) is regulated by the test DNA gene product (lacI) and the second promoter (PR¹) is regulated by an anti-terminator protein.

Other terminator/anti-terminator protein systems can be readily utilized in the groE or S⁵ based assay systems described in Examples 9 and 10. For example, other suitable terminator nucleotide sequences are well characterized, including the Box-A like consensus sequence C/TGCTCTT(T)A (SEQ ID NO 15), and the related transcription terminator sequences for crp, trp, his. phe, thr, ampC, ilv, and rrnB.

Particularly preferred is the terminator/anti-terminator system provided by lambda in the nutR-tR1 nucleotide sequence and the N protein. The nutR-tR1 nucleotide sequence can readily be substituted into the above described terminator/anti-terminator

constructions in place of qut-t6s. A PCR-amplified fragment containing the nutR-tR1 sequence [that contains tR1(I-III)] is isolated by conducting PCR on wild-type lambda in the presence of oligonucleotide primers pll (SEQ ID NO 16) and pl2 (SEQ ID NO 17) having the sequences:

pll 5'-ATCAGCCTGCAGGTCGACTAAATAAACCCGCTCTTACAC-3' P12 5'-AGCATCGTCGACGGATCCCACGAACCATATGTAAGTATT-3' The resulting PCR fragment is cloned into the above host vector plasmids as described above. In addition, the anti-terminator protein N is provided in trans by the lambda test DNA sequence Lambda LIZ Alpha.

In addition to the high and low copy number plasmid constructs described herein, the host reporter gene constructs can also be present as lower copy number, or as single copy genomic elements within the E. coli chromosome. The preparation of stably-integrated genetic elements was described earlier using the suicide vector system for the construction of the prolysogen strain of E. coli. That system can readily be utilized to produce stably integrated groE or S⁵ reporter gene constructions as described herein. 11. Construction of a Gene Activating Selectable

System Utilizing GroE and Anti-Terminator Protein Lambda N

A plasmid vector containing a reporter gene under the transcriptional control of a lac promoter and lac operator is constructed having the E. coli groE as the reporter gene and the lambda anti-terminator

nucleotide sequence nutR, that is regulated by lambda terminator protein N, and one or more of either of the lambda terminator sequences tR1 or t6s, or both.

A strain of E. coli that is GroE- is required as the host cell for use of this reporter gene construct, and was prepared earlier in Example 9A, designated SCS-8 groES30, and is tet^r.

The reporter gene plasmid construct used in a GroE system with the N anti-terminator protein can take a number of various forms, which can vary in the precise reporter gene, including groES, groEL or

groESL, and can vary in the structure of the

terminator (s).

A. Plasmid pALS74

The reporter construct on plasmid pALS74 is very similar to that shown in Figure 10, and can be represented by the following schematic:

...-lacP-lacO-nutR-tR1(I-V)-t6s-groES-.... Plasmid pALS74 carries a nutR site for anti-termination, and has two primary terminators. The first terminator is the lambda terminator tR1 that has been shown to consist of a series of five terminator sequences, designated tR1(I-V), to connote the

presence of all five sequences. The second terminator is the lambda t6s terminator. In addition, pALS74 utilizes the groES reporter gene.

Plasmid pALS74 was produced by cloning the lambda nutR-tR1 (I-V) region by PCR into plasmid p15G10Qut described earlier in Example 9 to form plasmid

p15G10Nut, and then moving the fragment that contains the elements lacP-lacO-nutR-tR1 (I-V) -t6s-groES into the BglII site of the low copy number vector pALlow to form pALS74.

Plasmid pALS74 was then tranformed into the E . coli host SCS-8 groES30 described in Example 9 to form SCS-8 groES30(pALS74).

The plasmid pALS74 has been deposited with the ATCC in the form of E. coli strain SCS-8 groES30

described herein and containing the plasmid pALS74.

Plasmid pALS74 is a low copy number plasmid because it contains the RSF1010 bacterial origin of replication.

B. Plasmid pBO

The reporter construct on plasmid pBQ is very similar to pALS74 and can be represented by the following schematic:

... -lacP-lacO-nutR-tR1(I-V)-t6s-t6s-groESL....

Plasmid pBQ carries a nutR site for anti-termination, and has three primary terminators. The first

terminator is the lambda terminator tR1 (I-V), and the second and third terminators are each the lambda t6s terminator. In addition, pBQ utilizes the groESL reporter gene.

To prepare pBQ, plasmid pALS74 is restriction endonuclease digested with Apal and Sail to release a fragment containing the groES gene. The resulting fragments are treated with CIAP and the larger

(vector) fragment is gel purified to form a vector fragment. Plasmid pBSG3 was prepared as described in Example 9, and is restriction endonuclease digested with Apal and Sail to release a 2.2 kilobase (Kb) fragment containing the groESL gene. The groESL

fragment is similarly gel purified and ligated with the above vector fragment to form pALS74 (SL).

pALS74(SL) is then restriction endonuclease digested with. EcoRI and Sail, treated with CIAP, and the largest (vector) fragment is gel purified to form phosphatased pALS74 (SL) linear vector.

Wild-type lambda is then PCR-amplified using Pfu DNA polymerase (Stratagene) under standard PCR

conditions with the following primers pl3 (SEQ ID NO

) and p14 (SEQ ID NO

):

pl3 5'-CGGATCGAATTCATGGGTTAATTCGCTCGTTGT-3' pl4 5'-GCGAGTGTCGACTCTTACCTGTTGTGCAGATAT-3' The resulting PCR-amplified fragment of about 215 bp contains the lambda t6s terminator sequence.

Thereafter, the fragment is restriction endonuclease digested with EcoRI and Sail, gel purified and ligated with the phosphatased pALS74(SL) linear vector to form pBQ. C. Plasmid pBQ2

The reporter construct on plasmid pBQ2 is very similar to pBQ, and can be represented by the following schematic:

... -lacP-lacO-nutR-tR1(I-II)-t6s-t6s-groESL....

Plasmid pBQ2 carries a nutR site for anti-termination, and has three primary terminators. The first

terminator is the lambda terminator tR1(I-II), and the second and third terminators are each the lambda t6s terminator. In addition, pBQ2 utilizes the groESL reporter gene.

To prepare pBQ2, pBQ is restriction endonuclease digested with NotI and Spel to release a fragment of about 270 bp containing the nutR-tR1 (I-V) genes. The resulting fragments are treated with CIAP and the larger (vector) fragment is gel purified to form phosphatased linear pBQ vector fragment.

Wild-type lambda is then PCR-amplified using Pfu DNA polymerase under standard PCR conditions with the following primers pl5 (SEQ ID NO

):

5'-ATCGACAGCGAATTCGCGGCCGCATAAATAACCCCGCTCTTACAC-3' and p16 (SEQ ID NO

):

5'CGTACTGGCACTAGTCATGTCCAGTGCGCGCTAGATAACAATTGATTGAATG-3.'

The resulting PCR-amplified fragment of about 150 bp contains the nutR-tR1 (I-II) region of lambda.

Thereafter, the fragment is restriction endonuclease digested with NotI and Spel, gel purified and ligated with the phosphatased linear pBQ vector fragment to form pBQ2.

D. Plasmid pBQ2C

The reporter construct on plasmid pBQ2C is very similar to pBQ, and can be represented by the following schematic:

... -lacP-lacO-nutR-tR1(I-II)-t6s-t6s-T7rbs-groES.... Plasmid pBQ2C carries a nutR site for anti-termination, and has three primary terminators, and a T7 gene 10 ribosome binding site (T7rbs). The first terminator is the lambda terminator tR1 (I-II), and the second and third terminators are each the lambda t6s terminator. In addition, pBQ2C utilizes the groES reporter gene.

To prepare pBQ2C, pBQ2 is restriction

endonuclease digested with Sail and Apal to release a fragment containing the groESL gene. The resulting fragments are treated with CIAP and the larger

(vector) fragment is gel purified to form phosphatased linear pBQ2 vector fragment.

Plasmid pBSG3 produced in Example 9 is then PCR-amplified using Pfu DNA polymerase under standard PCR conditions with the following primers pl7 (SEQ ID NO

):

5'GCGGGTCGACTAGTAACTGCGGCCGCCTTTAAGAAGGAGATATATATATGAA

TATTCGTCCATTGCA-3', and p18 (SEQ ID NO

):

5'-CCTCTCGAGGGGCCCCTATTACGCTTCAACAATTGCCA-3'

The resulting PCR-amplified fragment of about 300 bp contains the groES gene, and has a T7 gene10 ribosome binding site (rbs) in place of the groES rbs based on the design of the PCR primer. Thereafter, the fragment is restriction endonuclease digested with Apal and Sail, gel purified and ligated with the phosphatased linear pBQ2 vector fragment to form

PBQ2C. E. Selection and Mutagenesis Testing With

GroE/N System

Using any one of the above plasmids pALS74, pBQ, pBQ2 or pBQ2C as a reporter gene

construct, selection is carried out essentially as described in Example 9 by transforming an E. coli host cell that is GroE-, such as SCS-8 groES30. and

screening a target DNA sequence such as Lambda LIZ Alpha which expresses the anti-terminator protein N. Phage plaque formation is inhibited in a manner analogous to the mechanism shown in Figure 10 because groE gene products are required for phage assembly and the transcription of the groE reporter gene

transcription unit is inhibited by the terminator sequences until the lambda infects and produces the lambda N anti-terminator protein. Expression of wild type lacl from Lambda LIZ Alpha then inhibits reporter gene transcription. Transcription will occur if mutant lad (non-repressing) is present.

Low copy vectors using the groE selection system are preferred because they exhibit greater sensitivity than the medium copy vectors. All of the repoerter vector constructs described in Example 11 are low copy vectors having the RSF1010 origin of replication.

The selection systems based on GroE, Sy5 or lactose dependent growth in prolysogenic host cells described in Examples 11, 10 and 8, respectively, have been compared for efficiency and sensitivity. The selection systems allow for significantly more rapid analysis because larger numbers of packaged phage particles can be screened per assay plate. For example, whereas about 15,000 phage particles are plated per plate when using color detection in a non-selectable system based on lacI as the target gene as described in Example 4D(2), about 100,000 phage particles are plated per plate when using GroE or approximately 45,000 particles per plate for SY5 selection, and about 200,000 phage particles are plated per plate when using lactose selection with a prolysogenic host.

Mutagenesis testing was also compared using the various systems described herein. For example a comparison was made for induction by mutagen (5×100 mg/kg MNU) compared to spontaneous induction using (1) non-selectable color detection, (2) lactose selection using prolysogen strain SCS-8cI, (3) SY5 selection, and (4) GroE selection using pALS74. The results, including fold induction ( induced/spontaneous), are shown in Table 3.

Table 3

Plaques/ Coloniesi Formed^a

System^b Spontaneous Induced

color 2.0 138.0

lactose 4.0 48.9

SY5 5.4 54.3

GroE 4.9 76.5 a Plaques or colonies measured are expressed as x10^-5.

b The systems used are (1) the non-selectable color based assay, (2) lactose selection using prolysogen SCS-8cI, (3) S5 selection described in Example 10, and (4) GroE selection using pALS74.

The results in Table 3 illustrate that the various selection systems described herein allow the detection of mutagenic activity in transgenic animals.

12. Deposit of Materials

The following cell lines and plasmids have been deposited on or before February 12, 1993, with the American Type Culture Collection, 1301 Parklawn Drive, Rockville, MD, USA (ATCC):

Material ATCC Accession No, LIZ Alpha transgenic embryo ATCC 72011 Prolysogenic E. coli SCS-8cI ATCC 55297 E. coli SCS-8 groES30(pALS74) ATCC - These deposits were made under the provisions of the Budapest Treaty on the International Recognition of the Deposit of Microorganisms for the Purpose of Patent Procedure and the Regulations thereunder

(Budapest Treaty). This assures maintenance of a viable culture for 30 years from the date of deposit. The deposits will be made available by ATCC under the terms of the Budapest Treaty which assures permanent and unrestricted availability of the progeny of the culture to the public upon issuance of the pertinent U.S. patent or upon laying open to the public of any U.S. or foreign patent application, whichever comes first, and assures availability of the progeny to one determined by the U.S. Commissioner of Patents and Trademarks to be entitled thereto according to 35 U.S.C. §122 and the Commissioner's rules pursuant thereto (including 37 CFR §1.14 with particular reference to 886 OG 638). The assignee of the present application has agreed that if the culture deposit should die or be lost or destroyed when cultivated under suitable conditions , it will be promptly

replaced on notification with a viable specimen of the same culture. Availability of the deposited strain is not to be construed as a license to practice the invention in contravention of the rights granted under the authority of any government in accordance with its patent laws.

The foregoing is intended as illustrative of the present invention but not limiting. Numerous

variations and modifications can be effected without departing from the true spirit and scope of the invention.

Claims

What is claimed is:

1. A mutagenesis testing method comprising:

(a) exposing a transgenic non-human organism to a test agent, said transgenic organism comprising somatic and germ cells containing a test DNA sequence being capable of direct or indirect detection in bioassay by expression of a test DNA sequence gene product; and

(b) determining by bioassay in a host cell the frequency of mutation of said exposed test DNA sequence, said host cell containing a reporter gene transcription unit controlled by a promoter that is regulated by said test DNA sequence gene product.

2. The method of claim 1 wherein said promoter is a lac operon and said test DNA sequence gene product is a lac repressor.

3. The method of claim 1 wherein said lac repressor is lacI, lacl^q, laql^sq, or lacl^cι.

4. The method of claim 1 wherein said reporter gene transcription unit is present in said host cell in the form of a plasmid, an F¹ plasmid, a phage genome or a host cell genome.

5. The method of claim 1 wherein said reporter gene transcription unit expresses a reporter gene product that confers a selective growth advantage to said test DNA sequence having a mutation.

6. The method of claim 5 wherein said reporter gene product is selected from the group consisting of beta-galactosidase, alpha-complementing beta-galactosidase, GroE, GroES, GroEL, GroESL and lambda SY5.

7. The method of claim 5 wherein said reporter gene transcription unit further contains a

terminator/antiterminator sequence and said test DNA sequence includes a gene that expresses an anti- terminator protein that controls transcription of said transcription unit.

8. The method of claim 7 wherein said

terminator sequence is selected from the group

consisting of lambda tR1, tR1(I-II), tR1(I-III), tR1(I-V) and t6s.

9. The method of claim 7 wherein said anti-terminator sequence is selected from the group

consisting of lambda nutR, nutL and qut.

10. The method of claim 7 wherein said

terminator/anti-terminator sequence is selected from the group consisting of lambda nutR-tR1(I-II), nutR-tR1(I-III), nutR-tR1(I-V), nutR-tR1(I-V)-t6s, nutR- tR1(I-II)-t6s-t6s and nutR-tR1(I-V)-t6s-t6s, and said anti-terminator protein is the lambda N protein.

11. The method of claim 7 wherein said

terminator/anti-terminator sequence is lambda qut-t6s and said anti-terminator protein is the lambda Q protein.

12. The method of claim 7 wherein said reporter gene transcription unit is controlled by first and second convergent promoters, wherein said first promoter is regulated by said test DNA sequence gene product and said second promoter is regulated by said anti-terminator protein.

13. The method of claim 5 wherein said test DNA sequence gene product also controls the expression of a second reporter gene transcription unit that

expresses a second reporter gene product such that said bioassy contains two different detectable

reporter gene products.

14. The method of claim 13 wherein said second reporter gene product is capable of producing a color indicator in said host cell.

15. The method of claim 1 wherein said host cell is restriction system deficient.

16. The method of claim 15 wherein said

restriction system deficiency is selected from the group consisting of McrA-, McrBC-, McrF-, Mrr- and HsdR-.

17. The method of claim 1 wherein said test DNA sequence further comprises nucleotide sequences that define excision means that permit excision of said test DNA sequence whereby said test DNA sequence is recoverable from said cells.

18. The method of claim 17 wherein said excision means are selected from the group consisting of restriction endonuclease cleavage sites, f1

bacteriophage origins of replication at the termini of said test DNA sequence, FLP-mediated recombination sites, Cre-lox site-specific recombination sites and lambda cos sites.

19. The method of claim 18 wherein said excision means are lambda cos sites and said determining comprises (i) recovering a sample of said test DNA sequence from said exposed cells via said cos sites by admixture of said sample with a lambda packaging extract to form a lambda phage packaged test DNA sequence, and (ii) conducting said bioassay on said lambda phage packaged test DNA sequence.

20. The method of claim 19 wherein said lambda phage packaging extract has a restriction system deficiency.

21. The method of claim 20 wherein said

22. A transgenic non-human organism for mutagen testing, comprising somatic and germ cells containing a test DNA sequence recoverable from said cells via excision sites, said sequence being capable of

detection by bioassay in a host cell by expression of a test DNA sequence gene product and said sequence defining genes that code for said test DNA sequence gene product and that code for an anti-terminator protein capable of derepressing a reporter gene transcription unit in said bacteria.

23. The organism of claim 22 wherein said anti-terminator protein is selected from the group

consisting of lambda N protein that binds to lambda anti-terminator nucleotide sequence nutR or nutL, and lambda Q protein that binds to lambda anti-terminator nucleotide sequence qut.

24. The organism of claim 22 wherein said test DNA sequence gene product is lacI, lacl^q, laql^sq, or lacI^CI.

25. The organism of claim 22 wherein said mammal is a mouse.

26. The organism of claim 22 wherein said excision sites are nucleotide sequences that define a system that permits enzymatic excision of the test DNA sequence.

27. A host cell for detecting mutations in a test DNA sequence by bioassay, said host cell

containing a reporter gene transcription unit

controlled by a promoter that is regulated by binding to a test DNA sequence gene product that is expressed by said test DNA sequence, said reporter gene

transcription unit containing nucleotide sequences that define (1) a reporter gene capable of expressing a reporter gene product that is detectable in said bioassay and that define (2) a terminator/anti-terminator sequence, said terminator/anti-terminator sequence allowing transcription of said transcription unit upon binding by an anti-terminator protein.

28. The host cell of claim 27 wherein said terminator/anti-terminator sequence is selected from the group consisting of lambda nutR-tR1(I-II), nutR-tR1(I-III), nutR-tR1(I-V), nutR-tR1(I-V)-t6s, nutR-tR1(I-II)-t6s-t6s and nutR-tR1(I-V)-t6s-t6s, and said anti-terminator protein is the lambda N protein.

29. The host cell of claim 27 wherein said terminator sequence is selected from the group

consisting of lambda qut-t6s, qut-tR1(I-II), qut-tR1(I-III), qut-tR1(I-V), qut-tR1(I-V)-t6s, qut-tR1(I-II)-t6s-t6s and qut-tR1(I-V)-t6s-t6s, and said anti-terminator protein is the lambda Q protein.

30. The host cell of claim 27 wherein said promoter is from the lac operon and said test DNA sequence gene product is Lac repressor.

31. The host cell of claim 27 wherein said reporter gene transcription unit expresses a reporter gene product that confers a selective growth advantage to said host cell or to said test DNA sequence.

32. The host cell of claim 31 wherein said reporter gene product is selected from the group consisting of beta galactosidase, alpha-complementing beta-galactosidase, groE, groES, groEL, groESL and lambda SY5.

33. The host cell of claim 27 wherein said reporter gene transcription unit is controlled by first and second convergent promoters, wherein said first promoter is regulated by said test DNA sequence gene product and said second promoter is regulated by said anti-terminator protein.

34. The host cell of claim 27 wherein said host cell is restriction system deficient.

35. The host cell of claim 34 wherein said restriction system deficiency is selected from the group consisting of McrA-, McrBC-, McrF-, Mrr- and HsdR- .

36. The host cell of claim 27 selected from the group consisting of SCS-8groES30 and SCS-8groEL140.

37. The host cell of claim 27 wherein said cell contains the reporter gene transcription unit selected from the group consisting of the reporter gene

transcription units present in plasmids pBSG3Qut, p15G10Qut, pALS74, pBQ, pBQ2, pBQ2C and pMI/S⁵ tac-lac.

38. A kit for determining by bioassay

mutagenesis in a test DNA sequence recoverable from the genome of cells of a non-human transgenic

organism, said test DNA sequence being capable of expression of a test DNA gene product, said kit comprising:

(1) a restriction deficient host cell comprising a reporter gene transcription unit controlled by a promoter repressed upon binding by said test DNA sequence gene product, said reporter gene

transcription unit comprising a reporter gene and a terminator/anti-terminator sequence, wherein said reporter gene is capable of expressing a reporter gene product that is directly or indirectly detectable in said bioassay, and wherein said terrainator/anti- terminator sequence allows transcription of said transcription unit upon binding to an anti-terminator protein; and

(2) a reagent for recovery of said test DNA sequence from said genome of said cells.

39. The kit of claim 38 wherein said test DNA sequence is contained within an excisably integrated lambda phage, and wherein said reagent is an aliquot of a bacteriophage lambda DNA packaging extract isolated from a bacteriophage lysogen that is

restriction deficient, the amount of said extract present in said aliquot being sufficient to excise and package a transforming amount of said excisably- integrated lambda phage.

40. The kit of claim 38 wherein said host cell has a restriction deficiency selected from the group consisting of McrA-, McrBC-, McrF-, Mrr- and HsdR-.

41. A prolysogenic microorganism that is

restriction system deficient.

42. The prolysogenic microorganism of claim 41 wherein said restriction system is a restriction system found within the minute 98 region of the E. coli K-12 genome.

43. The prolysogenic microorganism of claim 41 wherein said microorganism is a strain of E. coli that has a restriction system deficiency selected from the group consisting of McrA-, McrBC-, McrF-, Mrr- and HsdR-.

44. A phage-free, prolysogenic strain of E.

coli.

45. The E. coli strain of claim 44 wherein said strain is SCS-8cI.

46. A kit for determining by bioassay the extent of mutagenesis in a test DNA sequence recoverable from the genome of cells via enzymatic excision sites and containing an excisably integrated lambda phage comprising a prolysogenic strain of E. coli that is restriction system deficient and an aliquot of a bacteriophage lambda DNA packaging extract isolated from a bacteriophage lysogen that is McrA-, McrBC, McrF-, Mrr- and HsdR-, the amount of said extract present in said aliquot being sufficient to excise and package a transforming amount of said excisably- integrated lambda phage.

47. A mutagenesis testing method comprising:

(a) exposing a transgenic non-human mammal to a test agent, said transgenic mammal comprising somatic and germ cells containing a test DNA sequence recoverable from said cells via enzymatic excision sites, said sequence being capable of detection in bacteria by bioassay;

(b) recovering a sample of said test DNA sequence from said exposed mammal; and

(c) determining by bioassay in a prolysogenic microorganism that is restriction system deficient the frequency of mutation of said test DNA sequence in said recovered sample.

48. The method in accordance with claim 47 wherein said determining of step (c) comprises:

(d) treating said recovered sample of step (b) with a restriction endonuclease to form ligatable termini flanking said test DNA sequence;

(e) circularizing said test DNA sequence via said ligatable termini;

(f) transforming said prolysogenic restriction system deficient microorganism of step (c) with said circularized target DNA sequence of step (e).

49. The method in accordance with claim 47 wherein said test DNA sequence includes Lad or Lacl^q.

50. The method in accordance with claim 49 wherein said test DNA sequence further includes LacZα.

51. The method in accordance with claim 47 wherein said recovering of step (b) comprises rescue of the test DNA sequence by admixture of a sample of test DNA sequence from said exposed mammal with a lambda phage packaging extract to form lambda phage packaged test DNA sequence.

52. The method in accordance with claim 51 wherein said lambda phage packaging extract has a restriction system deficiency selected from the group consisting of McrA-, McrBC-, McrF-, Mrr- and HsdR-.

53. The method in accordance with claim 47 wherein said determining of step (c) comprises:

(c) treating said recovered sample of step (b) with a restriction system deficient bacteriophage packaging extract to package said recovered test DNA sequence of step (b);

(d) infecting a phage-free, prolysogenic strain of E. coli with said packaged test DNA

sequence; and

(e) determining by bioassay of said

infected E. coli of step (d) the frequency of mutation of said test DNA sequence.

54. The method in accordance with claim 47 wherein said enzymatic excision sites define a system for excision and circularization selected from the group consisting of the system of lambda ZAP, FLP-mediated recombination and Cre-lox site-specific recombination.

55. A mutagenesis testing method comprising:

(a) exposing a transgenic non-human mammal to a test agent, said transgenic mammal comprising somatic and germ cells containing a test DNA sequence flanked on both sides by one of two halves of an f1 bacteriophage origin of replication contained by a bacteriophage origin of replication contained by a bacteriophage derivative genome;

(b) rescuing with fl helper phage proteins a sample of said test DNA sequence from said exposed mammal; and

(c) determining the frequency of mutation of said test DNA sequence in said rescued sample.