EP0449820A1 - Radiation/microbe bioassay - Google Patents

Abstract

A bioanalytical method and apparatus has been developed for the rapid detection and measurement of the toxic reaction of bacteria to chemicals whose levels are too low to produce acute toxic symptoms, but which could cause a reaction or deterioration. toxic years after chronic exposure. The bacteria used are isogenic strains of Bacillus subtilis subjected to chemical exposure and illuminated by radiation such as a laser beam. Detectors detect the differential light scattered by the bacteria, and the reaction index of cytotoxicity and genotoxicity is calculated.

Description

RADIATION/MICROBE BIOASSAY

BACKGROUND OF THE INVENTION The present invention is directed to the field of bioassays, and in particular to an assay utilizing radiation, or laser, technology in combination with unique genetically engineered bacteria that provide a "fingerprinting" capacity to identify and measure concentrations of toxicants in various environmental media.

It has long been a concern among segments of society that products developed through chemical technology and intended to benefit humanity may be responsible for dramatic increases in the incidence of cancer and may cause long-term decline in genetic health. The ability to efficiently and accurately monitor environmental media such as water, soil, air and food for natural and man-made toxicants capable of mutagenesis and carcinogenesis is therefore one of the most crucial scientific requirements in public health.

Methods of analytical chemistry such as gas chromatography, mass spectroscopy, and high pressure liquid chromatography are available to test air, water or soil samples. However, sample preparation can be complex and different assays must be developed for the identification and quantification of individual chemicals or chemical classes. Moreover,the knowledge that a chemical is present and its concentration does not indicate its cytotoxic or genotoxic activity. These tests do not indicate whether an adverse biological effect is strong, weak, or even present in a sample in order to make reasonable risk evaluations for human beings.

The most widely employed bacterial test system for carcinogens is the Ames test, Ames, Mutat. Res. 31:347- 349 (1974), using Salmonella typhimurium histidine- requiring auxotrophs. Because the Ames assay relies on reverse mutation, a mutant deoxyribonucleic acid (DNA) locus must be mutated back to its wild-type configuration or be suppressed. Thus, chemical mutagens that cannot effect this change in the DNA will go undetected. Researchers have recognized other shortcomings in this test, specifically an inability to devise an in vitro activation system that is reliable and reproducible, and an inability to detect some mutational events that can lead to cancer. Felkner, Microbial Testers; Probing Chemical Carcinogenesis, 177 (1981).

In 1971, Felkner discovered that ordinary transforming strains of Bacillus subtilis recovered from short exposures to the potent mutagenic and carcinogenic agent, 4-nitroquinoline-l-oxide (4NQO), while other strains incapable of recombination and/or dark repair did not. The ability or inability to recover was found to be directly correlated with the blocking of de novo deoxyribonucleic (DNA) synthesis through complexation of 4NQ0 with the DNA of the cell. Laumbach, A.D. and Felkner, I.C., Formation of 4-Nitroquinoline-1-oxide Complex with DNA in Normal and Repair Deficient Strains of Bacillus Subtilis, Mutat. Res. 15: 233-245 (1972). The product of this research was the discovery that a certain chemical structure has a corresponding particular biological function. Based on this principle, the effect of a chemical on a biological system, such as Bacillus subtilis bacteria, may be identified by the change in morphology of the assayed cells.

In 1981, Felkner developed a microbial bioassay from a collection of several hundred Bacillus subtilis mutants based on the principles derived from his earlier work. These selected strains had unique genetic defects causing them to be very sensitive to the effects of specific chemical classes, toxins and radiation. To enable uniformity, the mutant genes from these strains were introduced by recombination technology into identical clones of the wild type (normal) Bacillis subtilis strain. Thus, a battery of strains, differing only by a single character (gene) or limited set of genes was produced. The member strains are therefore isogenic, that is, identical except for single defined genes.

In a liquid suspension and by using conventional bacteriological methods such as microscopy, spectrography and colony or cell counts, it has been documented that chemical toxicants such as 4-nitroquinoline-1-oxide can block cell division, induce cellular swelling and cause cell rupture as a consequence of binding to and damaging the DNA of Bacillus subtilis. Laumbach & Felkner, Formations of a 4-nitroquinoline-l-oxide complex in normal and repair deficient strains of Bacillus subtilis, Mutat. Res. 15:233-245 and 16:444 (1972). These parameters as well as the molecular interactions of many chemical classes have been documented for the Felkner mutant strains, indicating a unique response pattern for any assay chemical that is toxic. Felkner, Development of a Bacillus subtilis system to screen carcinogens/mutagens; DNA-damaging and mutation assays, Microbial Testers: Probing Chemical Carcinogenesis, 89-120 (1981); Song, Photoactivation of furocoumaryl carcinogens/mutagens and their interactions with nucleic acids, Microbial Testers: Probing Chemical Carcinogenesis, 35-66 (1981); Streips, Bacterial Mutation Monitors for active metabolites of chemical carcinogens: Bacillus subtilis assays for mutation and DNA repair, Microbial Testers: Probing Chemical Carcinogenesis, 131-143 (1981). This testing method, however, is dependent on conventional plate and liquid methods. Although deemed acceptable under the United States Environmental Protection Agency Office of Pesticide Program Guidelines as an acceptable way of monitoring genotoxic response, the method lacked the sensitivity, specificity and speed capable of the presently claimed bioassay system.

In the early 1970's, Wyatt and his collaborators discovered a laser light scattering bioassay technique. Exponential phase bacterial cultures were used to inoculate aqueous samples containing compounds that might affect bacterial physiological processes, and indirectly, the morphology of the exposed bacterial populations. A plot was derived of the variation of scattered light intensity as a function of scattering angle. Early instrumentation was based on placing a sample-containing cuvette at the center of a circular arc, about which a collimated photomultiplier detector was rotated. A graph representing the variation of the relative scattered intensity as a function of scattering angle was called the differential light scattering (DLS) pattern.

Bioassays employing bacteria are disclosed in Wyatt U.S. Patent No. 3,730,842, issued May 1, 1973, and Wyatt U.S. Patent No. 4,101,383, issued July 18, 1978. These patents disclose methods of determining bacterial sensitivity or susceptibility to antibiotics by using test differential scattering patterns derived from the test bacteria and then comparing them to patterns of control bacteria. Neither reference provides for generating a pattern to screen for a toxicant having potential cytotoxic or genotoxic activity or to provide a means by which the toxicant may be identified and quantified. Wyatt U.S. patent No. 4,541,719, issued September 17, 1985, discloses a laser apparatus that measures scattered radiation by microparticles such as bacteria at two well-defined scattering angles. The toxic response of bacteria is measured by averaging mean scattered intensities. Other patents naming Wyatt as an inventor include Wyatt U.S. Patent No. 4,693,602; Phillips U.S. Patent No. 4,616,927; Wyatt U.S. patent No. 4,621,063; Wyatt U.S. Patent No. 4,548,500; Wyatt U.S. Patent No. 4,490,042; Wyatt U.S. Patent No. 4,173,415; Wyatt U.S. Patent No. 3,928,140; Wyatt U.S. Patent No. 3,770,351; Wyatt U.S. Patent No. 3,754,830; and Wyatt U.S. Patent No. 3,624,835, which are directed to various methods and apparatus for characterizing microparticles.

Other bioassays have included Bean, U.S. patent No. 3,708,402, issued January 2, 1973, and Bean, U.S. Patent No. 4,061,543, issued December 6, 1977. Neither of these references discloses a method or apparatus by which toxicity or genotoxicity may be assessed, or a test chemical identified. Thilly, U.S. patent No. 4,299,915, issued Nov.10, 1981, discloses an assay for mutagenesis in bacterial cells wherein bacterial cells such as S. typhimurium are exposed to a test chemical and plated in the presence of a purine analog. The procedure is dependent on conventional plating techniques, and suffers the same problems of lack of sensitivity to certain chemicals and the requirement for a laboratory.

SUMMARY OF THE INVENTION Therefore, it is the primary object of this invention to provide a bioassay method and apparatus for rapidly detecting and measuring the toxic response from chemicals in the environment whose levels are too low to produce acute toxic symptoms, but from which a toxic response or damage might develop years following chronic exposure.

Another object of this invention is to provide unique differential light scattering (DLS) patterns for chemical agents using isogenic repair mutant and wild type strains of Bacillus subtilis.

A further object of this invention is to select from the Bacillus subtilis strains a minimum group or "isoset" required to establish response patterns to specific chemical classes.

This invention has as yet another object the provision of a means to score the toxicity responses of the selected bacteria and to compare these responses to the responses of these bacteria to known chemical agents.

Another object of this invention is to establish means for evaluating toxicity responses for compounds whose genotoxic responses have been difficult or even impossible to detect with other state-of-the-art mutagenicity assays.

This invention has the further object to provide, as a reference, DLS patterns generated when these bacteria are treated with identified chemicals, to screen for toxicity and to identify and quantify samples.

A further object of this invention is the provision of a medium for solubilizing water insoluble samples to be tested, this medium being harmless to the assay bacteria and being optically clear when mixed with the bacteria so that radiation may readily penetrate, and scattering may be detected.

It is still another object of this invention to provide for metabolic activation of certain chemicals to be tested so that these chemicals may be taken up by the bacterial cells. The foregoing and other objects of the invention are achieved by the provision of Bacillus subtilis isogenic strains of bacteria lacking certain genetic repair functions and strains that are repair efficient that can be monitored for their unique responses to toxic chemicals using a differential light-scattering laser system. These responses are then scored to determine, for example, acute toxic or genotoxic effects. These scores, calculated as response indices, are then compared to known responses of these bacteria strains to identified chemical agents. It is to be understood that this method is also capable of assessing quantifiably the presence of a chemical to be tested. A better understanding of the disclosed embodiments of the invention will be achieved when the accompanying detailed description is considered in conjunction with the appended drawings, in which reference numerals are used for the same parts as illustrated in the figures.

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a schematic view of the method and apparatus of the claimed invention.

Figures 2a through 2d are differential light scattering patterns generated when the assay bacteria is exposed to certain concentrations of 4-nitroquinoline-N-oxide ("4NQO").

Figures 3a and 3b are differential light scattering patterns generated when the assay bacteria is exposed to certain concentrations of N-methyl-N'-nitro-N-nitroso-quanadine ("MNNG").

Figures 4a through 4d are differential light scattering patterns generated when the assay bacteria is exposed to certain concentrations of benzo(a)pyrene ("B(a)P").

Figures 5a through 5d are differential light scattering patterns generated when the assay bacteria is exposed to certain concentrations of "Lindane".

DESCRIPTION OF THE PREFERRED EMBODIMENTS Figure 1 provides a schematic illustration of the claimed apparatus by which samples are bioassayed in accord with the claimed invention. The bioassay consists of a radiation source 4, such as a laser system, which in the preferred embodiment employs the laser system such as that disclosed in Wyatt U.S. Patent No. 4,541,719, which includes a laser producing monochromatic visible or infrared radiation that is plane polarized with respect to two arrays of detectors. The laser system is capable of making 1200 measurements/second at 15 unique angles. The bioassay further includes an isogenic set of Bacillus subtilis mutants 2. The laser illuminates the cuvette 1 containing the bacteria 2 in suspension, these bacteria causing the light to be differentially scattered. The differential light scattering (DLS) 6 of the laser beam is detected as scattered light intensity as a function of angle. In one embodiment, the detectors 5 detect the DLS and feed the signals to a recorder 7, which records, in the form of a graph, a plot of the scattered intensity. In another embodiment, an analysis apparatus such as a computer 8 receive the data from the detectors 5 directly.

The intensity variation in the scattered light is called the DLS pattern, and is dependent on the average size and structure of the bacteria as well as the population size distribution. The DLS pattern of a suspension shows how many particles are present, their size, shape and the distribution of particles. Thus, if the bacteria undergo cell shrinkage, enlargement, division or lysis, the DLS pattern will shift from the control pattern of untreated bacteria. Bacterial count is expressed as a measure of height, or the highest point, of the DLS pattern. The shift of peaks of a sample DLS pattern indicates the degree of morphological change as compared to the control pattern. It is to be understood that the DLS pattern may be in magnetically recorded digital form for machine readability, or recorded in graphic form such that the scattered intensity pattern is represented as an eye readable curve.

In order to facilitate data entry and calculation where data from the detectors 5 is not transmitted to a recorder 7, a digitizing tablet (not shown) is used to trace each pattern's curve and simultaneously transmit a continuous stream of x and y coordinates to an analysis apparatus such as a computer 8. As discussed, the use of a digitizer can be eliminated by use of an integrated laser-computer system so that input into the system is much more rapid and avoids the potential for input errors. Each embodiment is shown in Figure 1 by phantom line connections. The analysis apparatus, which is in the preferred embodiment a computer 8 carries out the required calculations of response indices representing the shift in the DLS of the control as compared to the sample, and the change in bacterial count. The derived indices provide an immediate preliminary indication of toxic response, the DNA damaging (genotoxic) potential and aid in the identification of "isosets" of the Bacillus subtilis bacteria for compound identification.

The concentration of the chemical sample is determined by the level of response by the bacteria to the given sample. The set of bacteria selected for the bioassay each differ by only one property, which causes them to be more sensitive than other set members because of the mechanism of toxicity. Therefore, a "fingerprint", or profile unique to each chemical is generated from the differential response of the member bacteria. By "scoring" these fingerprints, the bioassay provides data for identifying test chemicals, assessing the concentration of test chemicals, and assessing whether the test chemical is acutely or chronically toxic. Another embodiment of the invention provides a library 9 wherein are stored the DLS patterns and calculated response indices for the bioassay of identified samples. By comparing the sample DLS with the stored pattern, the sample may be screened for toxicity, identified, and/or quantified.

The time required for a complete assay is approximately sixty (60) to sixty-six (66) minutes, the time required for untreated bacteria to divide twice in the absence of any chemical sample. However, individual samples may be analyzed at a rate of one sample per minute to determine the concentrations of a chemical in an aqueous sample. The Bacteria: The bacteria used in the bioassay are of a single species. Bacillus subtilis, which has numerous advantages over other bacteria traditionally used in bioassays. First, Bacillus subtilis is the most widely studied and genetically mapped gram-positive microorganism. Henner, D.J. and Hoch, J.A., The Bacillus subtilis Chromosome, Microbiol. Rev. 44:57-82 (1980).

The morphology, biochemistry, genetics and physiology of these organisms are therefore thoroughly understood. The morphological features of these bacteria are also easily adapted to the requirements of sensitivity and reproducibility of the bioassay. In its dormant form. Bacillus subtilis exists as a spore that is capable of surviving genetically unchanged while resisting heat, drying or other adverse effects for an indefinite period. As a vegetative cell, the bacteria can divide rapidly to double itself in approximately twenty-six (26) minutes during the logarithmic growth phase, and is much more sensitive at this stage to chemical toxicants than are the gramnegative microorganisms. Salmonella typhimurium and Escherichla coli. In its competent state, Bacillus subtilis readily takes up large molecules such as DNA, and can integrate genes derived from members of the same species or other species through traditional recombinant DNA technology.

The bioassay uses Bacillus subtilis mutants that are genetically identical, or isogenic, except for one or more genetic blocks in unique enzymatic repair processes or steps that restore chemically damaged DNA to its functional condition. Among the isogenic set of Bacillus subtilis are specifically selected strains deficient in different recombination (Rec-), excision (exc-), polymerase (Pol-) or spore repair (Spp-) repair steps. These strains may be deficient in one or a multiple of repair steps, and include the Rec- strains, recAl, recA8, recB2, recC5, recD3, recE4, recG13, mc-1, and m45; the Pol- strains T-1, TKJ8201; the Excstrains her-9 and TKJ8206; The Exc- and Rec- strain FH2006-7; the Exc-, Rec- and Pol- strain HJ15; the Exc- and Spp- strain TKJ5211; the Exc-, Pol- and Spp- strain TKJ6321; and repair efficient strains HA101, 168 and 168 wild type. The mutants in this group are therefore more sensitive to the metabolic toxicity or genotoxic activity or a broad range of chemical substances at nannomolar or nannogram concentrations.

Since there are many repair processes and many enzymatic steps in these processes, a given chemical will affect only certain of the isogenic mutants, producing a unique DLS pattern. These Bacillus subtilis mutants were developed by treating them with by irradiation or with various chemical mutagens. For example, nitrosoquanidine, a powerful mutagenic chemical, was used to change the genetic makeup of the bacteria and trimethoprim to select for mutant types that could not synthesize DNA unless thymine was provided to them. The mutants from one group isolated by Felkner were designated FH2001 through FH2006 and had been derived from a parent strain JBO1-200 whose lineage is defined in Felkner J. Bacteriol. 104:1030-1032 (1970). As discussed in this reference, conventional genetic analysis determined where the unique thymine gene defect was located in the chromosome of this bacteria. Subsequently, a mutant that was unusually sensitive to ultraviolet light and to the carcinogenic chemical 4- nitroquinoline-1-oxide was isolated by using trimethoprim treatment. Felkner, I.C., Isolation and Characterization of Thymineless Mutants from Ultraviolet-Sensitive Bacillus subtilis Strains, J. Bacteriol. 104: 1030-1032 (1970). This strain, designated FH2006-7 was found to have a gene mutation genetically linked to the thymine locus that cause the mutant to be unable to carry out genetic recombination. The genes from this mutant were transferred to a parental "wild type" strain by DNA isolation, uptake and integration so that the FH2006-7 descendants with the following genotypes were produced:

(1) trp-, thy-, her-, rec-

(2) trp+, thy-, her-, rec-

(3) trp+, thy+, her-, rec- Types 2 and 3 are mutants suitable for assaying her- and rec- mutations to detect DNA damage. These bacteria, constructed by conventional genetic engineering techniques would not occur in nature or result in a surviving bacterial isolate. Among the nineteen strains used with this bioassay, recA8, recA1, recB2, recC5, recD3, recE4, recG13, hcr-9, mc-1 and FH2006-7 are descended from strain 168. This strain is a descendent from the strain trp C of John Spizizen. Spizizen, J. Transformation of biochemically deficient strains of Bacillus subtilis by deoxyribonucleate, Proc. Nat. Acad. Sci. U.S.A. 44:1072-1078. The remainder of the mutants are descended from this strain but are of direct lineage of HA 101 described in Kada, T., Sadaie, Y. and Tutikawa, K., Mutat. Res. 16:165 (1972) and Tanooka, H., Mutat. Res. 42:1 (1977). The mutant strains employed in the preferred embodiment of the bioassay may be obtained through application of conventional genetic engineering and transformation techniques by following the teachings provided in the following, and the references cited therein: Henner, D.J. and Hoch, J.A., The Bacillus subtilis Chromosome, Microbiol. Rev. 44:57-82 (1980); Laumbach, A.D., Felkner, I.C., Formations of a 4-nitroquinoline-1-oxide complex in normal and repair deficient strains of Bacillus subtilis, Mutat. Res. 15:233-245 and 16:444 (1972); Felkner, I.C., Isolation and Characterization of Thymineless Mutants from Ultraviolet-Sensitive Bacillus subtilis Strains, J. Bacteriol. 104:1030 (1970); Felkner, I.C., Kadlubar, F., Parallel Between Ultraviolet Light and 4-Nitroquinoline-1-Oxide Sensitivity in Bacillus subtilis, J. Bacteriol. 96:1448 (1968); and I.C. Felkner, Microbial Testers: Probing Carcinogenesis (1981).

Isosets, defined as a subset of these mutants, constitute the minimum number required to define a response to a sample tested. Data derived form fractional survival tests and data derived from spot assays are used to determine which B. subtilis strains should be selected to comprise an isoset to identify a given compound by the radiation bioassay.

To select isoset members to be used in the bioassay, the isogenic mutants are pre-screened using the spot test with given chemical samples. The DNA-damaging test is performed according to the procedure disclosed in Felkner, I.C., Microbial Testers: Probing Carcinogenesis (1981) and Felkner, I.C. in Medine, A., and Anderson, W. , eds. Proceedings of the National Conference on Environmental Engineering, Am. Soc. Civil Eng., N.Y. pp. 204-209 (1983). Each mutant is inoculated into 5 ml of Brain Heart Infusion (BHI) broth, either from a disc containing approximately 1 × 10⁷ washed spores or a 1 mm loopful from a sporulating slant culture, and incubated by shaking at 37 to 39° C for 16 hours. This culture is designated o/n. Using a 1 mm inoculating loop, inocula from the o/n cultures are streaked radially on a nutrient agar plate to a sensitivity disc containing 10 ul of the assay chemical which may or may not have been pre-incubated with a rat liver microsomal fraction (S-9) in order to achieve metabolic activation. After incubation at 37 to 39° C, the distance (mm) of growth inhibition from the periphery is measured with a vernier caliper (Manostat). If varying concentrations of the assay substance are used, a concentration-dependent growth inhibition curve can be constructed, using the data from responsive mutants. Certain chemicals are not readily uptaken by the assay bacteria without metabolic activation. Liver microsomes, designated an S-9 fraction, were prepared from Sprague-Dawley rats induced with Aroclor 1254 (Litton Bionetics, Charleston, SC 19405). The S-9 fraction was prepared as a KCl homogenate with a protein concentration of 25-28 mg/ml. It is understood that this fraction may be microsomal or post-microsomal and is not limited by the origin of the tissues. The "Cocktail":

Many promutagens/carcinogens are water insoluble. A "cocktail" was therefore derived so that a clear aqueous solution of the test chemical could be mixed with the assay bacteria and the hydrophilic S-9 fraction, thereby minimizing interference with the radiation source and its ability to generate a DLS pattern of the radiation scattered by said bacteria. The "cocktail" has the advantage of solubilizing chemicals in an aqueous system that were not as readily solubilized by other known procedures using dimethylsulfoxide (DMSO), ethanol or acetone before being introduced into the aqueous assay mixture. Moreover, the "cocktail" is nontoxic to the assay bacteria and permits the S-9 fraction enzymes to metabolize the test chemical. The preferred embodiment of the bioassay employs a dispersant sold under the trademark "COREXIT 7664 Oil Dispersant" (Exxon Chemicals, Clark, New Jersey) and a non-ionic surfactant sold under the trademark "EMULPHOR EL-620" (GAF Corp., 140 West 51 Street, New York, N.Y. 10020). Both the "COREXIT" dispersant and the "EMULPHOR" surfactant are readily obtainable at the present time, and are generally known to chemists and commercial users. The chemical compositions of each of these are maintained as trade secrets by the respective owner companies.

The "EMULPHOR" surfactant is a polyoxylated vegetable oil.

At 25° C it is a clear, light brown liquid, having a specific gravity at this temperature of 1.04 to 1.05 and a viscosity of 600 to 1,000 cps. At 25° C, it is readily soluble in water, and its neutralization number is 0.5 maximum. As discussed, this composition was chosen because it is nontoxic to the Bacillus subtilis bacteria and when mixed with the chemical sample to be tested and the bacteria and the hydrophilic S-9 fraction should metabolic activation be required, a clear suspension results that does not interfere with the ability of the laser radiation to penetrate or the intensity of the radiation to scatter. "COREXIT 7664" is a surfactant ester having a light amber color and an alcoholic odor. AT 15.6° C it has a specific gravity of 1.03, a density of 8.59, and a viscosity of 29. It is soluble in fresh water and seawater and is dispersible in hydrocarbons. As discussed, this compound was selected as a component of the "cocktail" because it is nontoxic to bacteria and is capable of mixing with aqueous suspensions of bacteria without clouding and thereby does not interfere with the ability of laser radiation to penetrate the suspension. The Radiation Source:

The radiation source 1 of the bioassay is a laser such as that disclosed in Wyatt, U.S. Patent No. 4,541,719. This instrument records scattering intensity at fifteen angular locations, permitting the relationships between refractive index, size and shape to be measured. Moreover the read head is anodized, sealed and attached directly to the laser head as well as to an integrated circuit board, providing a unitized optical bench and a grounded enclosure for the detector elements. The specifications of the laser employed in the preferred embodiment of the bioassay are as follows: Light source: 5 mW He-Ne laser, λ=632.8 nm. Plane polarized

Angles Monitored Simultaneously: 15 @ 0.2 ≤ sin 0/2 ≤ 0.9 in steps of 0.05

Computer interface: Parallel analog outputs interfaced to 16-channel, 12 bit A/D programmable multiplexer.

Data Collection: 12,500 conversions/second - 25,000 conversions/second

Single Particle Measurements: Concentration ≤ 5 × 10

3/ml Size (depending on refractive index) 0.6 to 50 um

Detectors: Transimpedance photodiodes with built-in amplifiers Detector Output Dynamic Range: 0 - 10 V Dynamic Range: 10⁴ Linearity: ± 0.1%

Dark Current Noise: ± 0.2 mV

It is to be understood that this laser apparatus comprises part of the preferred embodiment of the bioassay, but that any laser instrument capable of the objects of this invention is comprehended. Radiation Bioassay Examples: Examples are provided for the bioassay of 4-nitroquinoline-N-oxide ("4NQO"), N-methyl-N'-nitro-N-Nitroso-guanidine ("MNNG"), benzo(a)pyrene ("B(a)P"), and "Lindane" (1,2,3,4,5,6-hexachlorocyclohexane). The chemicals 4NQO and MNNG are well established as mutagens and DNA damaging agents to bacteria, and are water-soluble and direct-acting genotoxic chemicals that elicit carcinogenic activity. The chemical B(a)P is a well known procarcinogen/mutagen that is insoluble in an aqueous system, and requires metabolic activation by rat liver microsomal fractions (S-9) to become electrophilic and to cause DNA damage or a mutagenic response. "Lindane" is a chemical difficult to assay for its genotoxic potential by state of the art mutagenic assays.

To perform the bioassay according to the preferred embodiment, the following protocol is followed:

1. Lyphilized tablets containing the appropriate Bacillus subtilis strains and dehydrated growth medium are introduced into cuvettes containing water. Suspensions of bacterial cells at the required concentration described in the examples provided infra may also be used.

2. The inoculated cuvettes are placed in an incubation chamber that has been prewarmed to approximately 37 to 39° C, and incubation continued for about one hour to one and a half hours to establish a culture in the "assay-ready condition".

3. Test or control material is introduced into the cuvettes using one pair of cuvettes for each assay organism. A negative control having no test material, and a positive control having an identified toxic chemical at a concentration whose response has been determined is included.

4. The cuvettes are incubated for six minutes at 37-39° C, and each cuvette is read a minimum of fifty

(50) times (the total reading time per sample is 30 seconds).

5. Incubation at 37-39 degrees Centigrade is continued for sixty (60) to seventy (70) minutes and each cuvette is again read a minimum of fifty (50) times.

6. The samples are scored, and the bacterial responses evaluated.

* The bacteria could be lyphilized vegetative cells or spores in a capsule form. However, when spores are used, this form being preferred because of their stability, they must be heat-shocked at 60 degrees

Centigrade for 8 to 10 minutes before incubation. An alternative protocol follows: 1. Solubilize a sample in an aqueous solution.

For water insoluble samples, use a sufficient amount of

"cocktail" to prevent precipitation of the sample.

2. Make serial dilutions from the solubilized sample. 3. Add the capsule containing the bacterial spores and the dehydrated growth medium to a cuvette containing water.

4. Add a measured amount of each solubilized sample dilution to cuvettes containing bacteria and place the cuvette in an incubator at 37-39 degrees

Centigrade. Include a negative control having no test material, and a positive control having a toxic chemical at a concentration whose response has been determined.

5. Read the samples at 0, 6 and 66 minutes to follow the relative growth indicated by the increase in number of particles for each dilution.

6. Score each sample dilution and the controls.

7. Compare the score data to the score data of known compounds to determine a matching response profile.

8. Determine the concentration from a known dose response curve. If the sample is an unknown, search for a matching chemical response to determine its identity and concentration. It is understood that a negative control may also comprise a control sample of bacteria exposed to 4-nitroquinoline-N-oxide or 4-nitroquinoline-1-oxide and a positive control of bacteria exposed to benzo(a)pyrene. Cultivation of bacterial cells appropriate in concentration for the bioassay includes heat shocking strains of Bacillus subtilis in 1 ml of Brain Heart Infusion (BHI) broth for 20 minutes. Additional broth is added until an absorbance of 0.2 A(540) is achieved. The cultures are incubated at 37 to 39° C until the bacterial growth reaches an A(540) of 0.5, a period of approximately 2 hours. Cultures that require a longer period to reach this density are not considered suitable for the bioassay. Cultures found suitable are diluted to an A(540) of 0.3 with BHI broth yielding approximately 2 × 10⁷ bacteria/ml. These cells, when diluted by the amount required for the cuvette assay, yield a cell concentration of 10⁶ bacteria/ml.

1. Bioassay without metabolic activation (Cuvette Assay) - Compounds and bacterial isoset:

4NQO was assayed at final concentrations of 8.5, 4.2 and 2.1 ug/ml using Bacillus subtilis strains recE4 (6), 168 or 168 wild type (11). The DLS patterns for each of these assays comprise Figures 2a through 2d.

MNNG was assayed at a final concentration of 3.6 ug/ml with strains 6 and 11. The DLS patterns for each of these assays comprise Figures 3a and b.

Lindane was assayed at final concentrations of 53.6 ug/ml and 107 ug/ml utilizing Bacillus subtilis strains FH2006-7 (10) and 11. The DLS patterns for each of these assays comprise Figures 5a through 5d. Method:

In a screw cap tube containing 13.6 ml of deionized water at 37 to 39° C is pipetted 0.3 ml of bacteria and 0.1 ml of the test chemical. The tube is inverted gently and its contents poured into a cuvette. The cuvette is illuminated by the laser and the intensities of scattered light measured and scored.

The cuvette is then incubated at 37-39 degrees

Centigrade for approximately one (1) hour and illuminated identically. A negative control of bacteria in deionized water is also concurrently assayed to provide a baseline value for each strain.

The method may also be performed using lyphilized bacteria having with it sufficient dehydrated BHI to support the growth during the assay period. The total volume of fluid required to obtain a reading from the cuvette is 10 ml. Therefore, a volume of about 0.4 ml of fluid will rehydrate the cell-medium. The incubation period remains unchanged, and illuminating is performed at fifteen minute intervals.

2. Bioassay with Metabolic Activation (Preincubation) - Compound and bacterial isoset:

B(a)P was assayed at final concentrations of 15.1 and 7.6 ug/ml using Bacillus subtilis strains 11 and TKJ6321 (19). Method: Into a large capped test tube were pipetted 0.5 ml each of "COREXIT" and "EMULPHOR", B(a)P at the appropriate concentrations, 0.5 ml of S-9 fraction and 0.3 ml bacteria. The volume was adjusted to 14 ml with deionized water pre-warmed to 37° C. A control tube for each strain containing "COREXIT", "EMULPHOR", S-9 fraction and deionized water was prepared according to the same proportions as the sample to be tested. The tubes were incubated in a shaking water bath for approximately one (1) hour. After incubation, the contents were poured into cuvettes and the scattered intensities measured and scored, these readings being taken at zero and sixty minutes, as previously described. When lyphilized bacteria and growth media are used, a smaller volume of said bacteria are used to obtain the desired initial density in one (1) hour before treating the bacteria with the test chemical. The "COREXIT", "EMULPHOR" and S-9 volumes are each approximately 180 ul per 5 ml of sample.

3. Bioassay with Metabolic Activation (Cuvette Incubation) - Compound and bacterial isoset:

B(a)P was assayed at final concentrations of 15.1 and 7.6 ug/ml using Bacillus subtilis strains 11 and TKJ6321 (19). The DLS patterns generated by this assay comprise Figures 4a-4d. Method:

Into a scanning cuvette was pipetted a sufficient volume of Corexit to prevent precipitation of B(a)P when introduced into the aqueous incubation mix of bacteria dn S-9 fraction. The "CORERXIT"- B(a)P mixture was introduced into the cuvette along with 0.5 ml of S-9 fraction and 0.3 ml of bacteria in a total volume of 14 ml of deionized water pre-warmed to 37° C. A concurrent control tube for each strain was also prepared. The cuvettes were read and the scattered intensities scored at zero and sixty (60) minutes as previously described.

In an alternative method, the aqueous cocktail of

"COREXIT", "EMULPHOR" and water is prepared prior to addition of the S-9 fraction and the bacteria. Thus,

180 ul each of "EMULPHOR" and "COREXIT" are added to each 5 ml of deionized water and mixed. At the time of the assay, 180 ul of S-9 fraction is added along with appropriate amount of lyphilized bacteria and medium required to achieve a cell concentration of 10 6 bacteria/ml.

Response Evaluation:

The scoring techniques of the bioassay measure the height of the DLS pattern as a function of bacterial count and the amount of shift in the peaks of the chemical sample pattern and the control sample pattern to determine the degree of morphological change in size and shape of the bacteria. The derived indices provide an immediate preliminary indication of the toxic response, the DNA damaging (genotoxic) potential, and identify isosets for compound identification.

The measurement of each pattern's height provided an indication of bacterial count and is therefore the principal indicator of toxicity. In developing the scoring techniques for the identification of morphological changes, it was noted that there is a function which is continuous on a closed interval from the beginning of the pattern to its end. Each line segment between any two adjacent points formed an arc, and associated with each point on the X-axis is only one point on the pattern. Using a digitizing tablet, it is possible to obtain a sufficiently large number of points so that the sum of the distances between all adjacent points would approximate the length of the pattern. This was defined as the arc length of the curve. Variances of the arc length correspond with the variances in the bacteria morphology, especially as they related to individual cell size. In order to simplify the interpretation of a response to a given treatment by a chemical at a given concentration, a system using three (3) digits was developed, each of which was the indicator for a response. The nature of this response was assigned a value of one (1) through four (4) according to the height and direction of pattern displacement as a function of the type of effects on the bacterial population. To index a set of treatments, the count changes and patterns of a baseline control sample were measured. The subsequent treatment and control samples were compared to the baseline response.

Index calculations:

Four patterns were digitized in the following order: Control at time 0 and 60 minutes; treated sample at time 0 and 60 minutes. For each pattern in order, the response indices were determined by calculating the distance between adjacent points on the curve as the arc length. The mean Y of all coordinate pairs was calculated for the bacterial count index. The control at time "0" was designated as the baseline control. For each of the other curves, the percent change between the pattern's mean Y and the baseline pattern's mean Y were calculated. For the Fourth order Y index, the Y percent change was divided by 5 and, if the change was positive, 1 was added. If the change was negative, 1 was subtracted. The final integer value was retained. The percent change between the test pattern's arc length and the baseline pattern's arc length was calculated. To determine the fourth order arc index, the arc percent change was divided by 5 and, if the change was positive, 1 was added. If the change was negative, 2 was subtracted. The integer value was retained.

Each of the three digits, from left to right in the PRI index, derives from the following table according to the fourth order indices of the patterns, starting with the control at "0" time as the reference and the control at 60 minutes as the first digit.

The biological interpretations that can be made on a digit-by-digit basis are as follows:

"1" indicates a positive increase in bacterial count with a concurrent decrease in size of said bacteria (positive shift), and a non-toxic response.

It is normal for rapidly dividing cells to decrease in size.

"2" indicates a positive increase in bacterial count with a concurrent increase in size (negative shift), and a minimum toxic response.

"3" indicates a decrease in bacterial count with a concurrent increase in size (negative shift), and a toxic and lethal response.

"4" indicates a decrease in bacterial count with a concurrent decrease in size (positive shift), and a toxic and lethal response. The tertiary responses for each of the indexed patterns were obtained by multiplying the fourth order Y index by the fourth order arc index. The secondary response index (SRI) was determined by the calculation of the sum of the tertiary indices for all indexed patterns combined. This method of calculation is illustrated by the following data derived from an assay:

The interpretation of the PRI of 132 is that the control of sixty-six (66) minutes displayed normal growth, hence the number 1 as the first digit. The treated sample at time "0" showed a loss in bacterial count accompanied by size increase, hence the number 3 as the second digit. By sixty-six minutes, the treated sample showed an increase in bacterial count accompanied by a size increase, hence the number 2 in the third digit.

Two extreme examples of PRIs are 133 and 144. These examples are immediately toxic responses from which the cells fail to recover. The 133 PRI shows cell enlargement, and the 144 PRI shows cell shrinkage, both being associated with the failure of the cell to recover but by different mechanisms of action. The presence of a 3 or 4 in the first digit of any PRI should mean that the assay is invalid because the bacterial control culture is not displaying normal growth.

A screening score S provides a precise measure of bacterial growth or inhibition of bacteria in the test sample as compared to the control bacteria. It is defined in terms of DLS placement scores and was implemented by previous studies. Wyatt, P.J., Phillips, E.T., Scher, M.G., Kahn, M.R., and Allen, E.H., J. Argic. Food Chem. 25:1080 (1977); Wyatt, P.J., Scher, M.G., and Phillips, D.T., J. Agric. Food Chem. 25:1086 (1977). Thus,

10(Dbc/300) - 10(Dc/300) S = 300 log

10(Dbt/300) - 10(Dt/300) where Dbc is the DLS displacement of the assay broth plus bacteria relative to the assay broth (control), Dc is the displacement of the control blank relative to itself (0 when there is no background interference), Dbt is displacement of test material in assay broth plus bacteria relative to the control, and Dt is displacement of the test material in assay broth relative to the control blank. The equation converts the data logarithmic DLS patterns back to linear form so that the background is subtracted and hence converted to the usual logarithmic form. The equation yields a factor of 300 for 10-fold increases of control bacteria with respect to test bacteria. applying the index calculations to the samples assayed;

In all the figures, the untreated "0" time control is represented by a solid line the one hour untreated control by a dot broken line ( . . . . ), the

"0" time treated sample by a widely spaced broken line ( - - - - ), and the one hour treated sample by a narrow spaced broken line (_ _ _ _ ).

4NQ0 DLS Toxicity Responses:

Figures 2a through 2d show the DLS response patterns for strains 6 (rec 34) and 11 (168 wild type) respectively. From the mean Y values of the "0" time and 1 hour samples, it is clear that increases in cell counts occurred in all untreated controls. From

Figures 2a and 2b, it is clear that 4.2 ug/ml and 2.1 ug/ml of 4NQO caused increased arc lengths and curves shifting to the left relative to the light scattering angles of the negative time "0" control. The relative intensity increased less from the "0" to 1 hour interval. From Figure 2c, it is clear that 8.5 ug/ml of 4NQO caused a decrease in arc length and a curve shift to the left relative to the light scattering angles of the negative time "0" control. However, at 4.2 ug/ml, the change in arc length was not substantial and the curve shift relative to light scattering angles was slightly toward the larger angles.

In applying the response index calculations, the following interpretations can be made. Strain 6 at 4.2 and 2.1 ug/ml had PRI values of 132 and 131, respectively. At the 4.2 ug/ml level, an immediate, toxic and lethal effect was seen at "0" time. The bacteria showed some recovery because the counts increased by 60 minutes. However, the bacteria remained enlarged. At the 2.1 ug/ml level, an immediate, toxic and lethal effect was seen at "0" time, but the counts increased and cells were of normal size at 60 minutes. The 4.2 ug/ml concentration, however, did not adversely affect strain 11. These data show that strain 11 was more refractile to the effects of 4NP0 than was strain 6, even when strain 11 was treated at higher dose levels. Thus, the genotoxic manifestation is expressed at the primary response level.

MNNG DLS Toxicity Responses: Figures 3a and 3b show the DLS response patterns for strains 6 (rec E4) and 11 (168 wild type) respectively. From the mean Y values of the "0" time and 1 hour samples, it is clear that increases in cell counts occurred in all untreated controls. From Figure 3a, it is clear that 3.6 ug/ml of MNNG caused a decreased arc length and a curve shift to the left, relative to the light scattering angles of the negative time "0" control. The relative intensity increased less than did the negative control from the "0" to 1 hour interval. from Figure 3b, it is clear that 3.6 ug/ml of MNNG did not cause a decrease in arc length. although the arc length increase was slightly less than the negative control. The curve shift relative to light scattering angles was toward the larger angles, almost approximating the negative time "0" control. Applying the calculations of the response indices, the following interpretations can be made. Strain 6 at 3.6 ug/ml had a PRI value of 132, indicating an immediate, toxic and lethal effect at "0" time. The bacteria showed some recovery because the counts increased by 60 minutes. However, the bacteria remained enlarged. Strain 11 at the concentration of 3.6 ug/ml had a PRI value of 111, thus showing no compound-related toxicity expressed either immediately or over the 60-minute period following treatment. This is a genotoxic response expressed at the primary response level. Additionally, the SRI value for strains 6 and 11 were -3 and 19, respectively. The negative secondary response of strain 6, when compared to the positive SRI of strain 11, indicates that there is a genotoxic effect which can be repaired by the wild type (strain 11) but not the Rec- mutant (strain 6). Lindane DLS Toxicity Responses;

Figures 5a through 5d show the DLS response patterns for strains 10 (FH2006-7) and 11 (168 wild type). From the mean Y values of the "0" time and one hour samples, it is clear that increases in cell counts occurred in all untreated controls. From Figures 5a and 5b, it is clear that 107 ug/ml and 53.6 ug/ml of

Lindane initially caused decreased arc lengths and curves shifting to the left relative to the light scattering angles of the negative time "0" control.

However, at the lower level, a positive shift was seen at sixty minutes. The relative intensity did not change substantially from the "0" to one hour intervals at the higher level. However, an increase was seen at the lower concentration. From Figures 5c and 5d, it is clear that both concentrations of Lindane caused an initial decrease in arc length and a curve shift to the left relative to the light scattering angles of the negative time "0" control. However, the arc lengths increased and the scattering angle shifted to the right relative to the negative time "0" control. However, the arc lengths increased and the scattering angle shifted to the right relative to the negative time "0" control after one hour. Applying the response index calculations, the following interpretations can be made. Strain 10 at 107 and 5 3.6 ug/ml had PRI values of 222 and 131, respectively. At the 107 ug/ml level, a toxic and a sustained effect (cell enlargement) was seen. However, at the 53.6 ug/ml level, the bacteria recovered after 60 minutes. Strain 11 had PRI values of 131 at both concentrations. This indicated that at both levels, cell lethality and enlargement occurred but that full recovery resulting in growth and the return to normal size occurred by one hour. Although strain 10 may be slightly more affected than strain 11 at the higher concentration of Lindane, the differences in the PRI values do not appear to be useful in making toxicity assessments. The SRI values, however, were substantially different for strains 10 and 11. The SRI values for strain 10 at 107 ug/ml and 53.6 ug/ml were -19 and 7, respectively. The SRI value was 10 for strain 11 at both 107 ug/ml and 53.6 ug/ml. The dose-related SRI trend seen in strain 10 but not in strain 11 indicates a genotoxic effect of this chemical, as demonstrated by the negative SRI value at the higher dose level. As shown by the positive SRI values, strain 11 at both concentrations, was capable of repairing genetic damage and dividing normally. Strain 10 was capable of repairing genetic damage and dividing normally only at the lower concentration of 53.6 ug/ml. From these results, it is clear that primary cytotoxicity may be expressed by a three digit number. Using the primary response of mutant and wild type bacteria, some compounds such as MNNG and 4NPO which are highly toxic can also be identified as genotoxic. Compounds such as B(a)P when activated with an S-9 fraction cannot be shown to be genotoxic by using the primary response level alone. However, genotoxicity may be determined by calculating the secondary response. Other compounds such as Lindane, whose genotoxic responses have been difficult or even impossible to detect with other state of the art mutagenicity assays are readily detected by the radiation/microbe bioassay by calculating their secondary response indices.

Thus, it will be seen that all embodiments of the present invention provide a unique apparatus and method for radiation bioassay samples. While preferred embodiments of the invention have been disclosed, it should be understood that the spirit and scope of the invention is not to be limited solely by the appended claims, since numerous modifications of the disclosed embodiments will undoubtedly occur to those of skill in the art.

RADIATION/MICROBE BIOASSAY

BACKGROUND OF THE INVENTION

The present invention is directed to the field of bioassays, and in particular to an assay utilizing radiation, or laser, technology in combination with unique genetically engineered bacteria that provide a

"fingerprinting" capacity to identify and measure concentrations of toxicants in various environmental media.

It has long been a concern among segments of society that products developed through chemical technology and intended to benefit humanity may be responsible for dramatic increases in the incidence of cancer and may cause long-term decline in genetic health. The ability to efficiently and accurately monitor environmental media such as water, soil, air and food for natural and man-made toxicants capable of mutagenesis and carcinogenesis is therefore one of the most crucial scientific requirements in public health.

Methods of analytical chemistry such as gas chromatography, mass spectroscopy, and high pressure liquid chromatography are available to test air, water or soil samples. However, sample preparation can be complex and different assays must be developed for the identification and quantification of individual chemicals or chemical classes. Moreover, the knowledge that a chemical is present and its concentration does not indicate its cytotoxic or genotoxic activity. These tests do not indicate whether an adverse biological effect is strong, weak, or even present in a sample in order to make reasonable risk evaluations for human beings.

The most widely employed bacterial test system for carcinogens is the Ames test, Ames, Mutat. Res. 31:347- 349 (1974), using Salmonella typhimurium histidine-requiring auxotrophs. Because the Ames assay relies on reverse mutation, a mutant deoxyribonucleic acid (DNA) locus must be mutated back to its wild-type configuration or be suppressed. Thus, chemical mutagens that cannot effect this change in the DNA will go undetected. Researchers have recognized other shortcomings in this test, specifically an inability to devise an in vitro activation system that is reliable and reproducible, and an inability to detect some mutational events that can lead to cancer. Felkner, Microbial Testers: Probing Chemical Carcinogenesis, 177 (1981).

In 1971, Felkner discovered that ordinary transforming strains of Bacillus subtilis recovered from short exposures to the potent mutagenic and carcinogenic agent, 4-nitroquinoline-1-oxide (4NQO), while other strains incapable of recombination and/or dark repair did not. The ability or inability to recover was found to be directly correlated with the blocking of de novo deoxyribonucleic (DNA) synthesis through complexation of 4NQ0 with the DNA of the cell. Laumbach, A.D. and Felkner, I.C., Formation of 4-Nitroquinoline-1-oxide Complex with DNA in Normal and Repair Deficient Strains of Bacillus Subtilis, Mutat. Res. 15: 233-245 (1972). The product of this research was the discovery that a certain chemical structure has a corresponding particular biological function. Based on this principle, the effect of a chemical on a biological system, such as Bacillus subtilis bacteria, may be identified by the change in morphology of the assayed cells.

In 1981, Felkner developed a microbial bioassay from a collection of several hundred Bacillus subtilis mutants based on the principles derived from his earlier work. These selected strains had uniqu genetic defects causing them to be very sensitive to the effects of specific chemical classes, toxins and radiation. To enable uniformity, the mutant genes from these strains were introduced by recombination technology into identical clones of the wild type (normal) Bacillis subtilis strain. Thus, a battery of strains, differing only by a single character (gene) or limited set of genes was produced. The member strains are therefore isogenic, that is, identical except for single defined genes.

In a liquid suspension and by using conventional bacteriological methods such as microscopy, spectrography and colony or cell counts, it has been documented that chemical toxicants such as 4-nitroquinoline-1-oxide can block cell division, induce cellular swelling and cause cell rupture as a consequence of binding to and damaging the DNA of Bacillus subtilis. Laumbach & Felkner, Formations of a 4-nitroquinoline-1-oxide complex in normal and repair deficient strains of Bacillus subtilis, Mutat. Res. 15:233-245 and 16:444 (1972). These parameters as well as the molecular interactions of many chemical classes have been documented for the Felkner mutant strains, indicating a unique response pattern for any assay chemical that is toxic. Felkner, Development of a Bacillus subtilis system to screen carcinogens/mutagens; DNA-damaging and mutation assays, Microbial Testers: Probing Chemical Carcinogenesis, 89-120 (1981); Song, Photoactivation of furocoumaryl carcinogens/mutagens and their interactions with nucleic acids, Microbial Testers: Probing Chemical Carcinogenesis, 35-66 (1981); Streips, Bacterial Mutation Monitors for active metabolites of chemical carcinogens: Bacillus subtilis assays for mutation and DNA repair, Microbial Testers: Probing Chemical Carcinogenesis, 131-143 (1981). This testing method, however, is dependent on conventional plate and liquid methods. Although deemed acceptable under the United States Environmental Protection Agency Office of Pesticide Program Guidelines as an acceptable way of monitoring genotoxic response, the method lacked the sensitivity, specificity and speed capable of the presently claimed bioassay system.

In the early 1970's, Wyatt and his collaborators discovered a laser light scattering bioassay technique. Exponential phase bacterial cultures were used to inoculate aqueous samples containing compounds that might affect bacterial physiological processes, and indirectly, the morphology of the exposed bacterial populations. A plot was derived of the variation of scattered light intensity as a function of scattering angle. Early instrumentation was based on placing a sample-containing cuvette at the center of a circular arc, about which a collimated photomultiplier detector was rotated. A graph representing the variation of the relative scattered intensity as a function of scattering angle was called the differential light scattering (DLS) pattern.

Bioassays employing bacteria are disclosed in Wyatt U.S. Patent No. 3,730,842, issued May 1, 1973, and Wyatt U.S. Patent No. 4,101,383, issued July 18, 1978. These patents disclose methods of determining bacterial sensitivity or susceptibility to antibiotics by using test differential scattering patterns derived from the test bacteria and then comparing them to patterns of control bacteria. Neither reference provides for generating a pattern to screen for a toxicant having potential cytotoxic or genotoxic activity or to provide a means by which the toxicant may be identified and quantified. Wyatt U.S. patent No. 4,541,719, issued Septembe 17, 1985, discloses a laser apparatus that measure scattered radiation by microparticles such as bacteria at two well-defined scattering angles. The toxic response of bacteria is measured by averaging mean scattered intensities. Other patents naming Wyatt as an inventor include Wyatt U.S. Patent No. 4,693,602; Phillips U.S. Patent No. 4,616,927; Wyatt U.S. patent No. 4,621,063; Wyatt U.S. Patent No. 4,548,500; Wyatt U.S. Patent No. 4,490,042; Wyatt U.S. Patent No. 4,173,415; Wyatt U.S. Patent No. 3,928,140; Wyatt U.S. Patent No. 3,770,351; Wyatt U.S. Patent No. 3,754,830; and Wyatt U.S. Patent No. 3,624,835, which are directed to various methods and apparatus for characterizing microparticles.

Other bioassays have included Bean, U.S. patent No. 3,708,402, issued January 2, 1973, and Bean, U.S. Patent No. 4,061,543, issued December 6, 1977. Neither of these references discloses a method or apparatus by which toxicity or genotoxicity may be assessed, or a test chemical identified. Thilly, U.S. patent No. 4,299,915, issued Nov.10, 1981, discloses an assay for mutagenesis in bacterial cells wherein bacterial cells such as S. typhimurium are exposed to a test chemical and plated in the presence of a purine analog. The procedure is dependent on conventional plating techniques, and suffers the same problems of lack of sensitivity to certain chemicals and the requirement for a laboratory.

SUMMARY OF THE INVENTION Therefore, it is the primary object of this invention to provide a bioassay method and apparatus for rapidly detecting and measuring the toxic response from chemicals in the environment whose levels are too low to produce acute toxic symptoms, but from which a toxic response or damage might develop years following chronic exposure.

Another object of this invention is to provide unique differential light scattering (DLS) patterns for chemical agents using isogenic repair mutant and wild type strains of Bacillus subtilis.

A further object of this invention is to select from the Bacillus subtilis strains a minimum group or

"isoset" required to establish response patterns to specific chemical classes.

This invention has as yet another object the provision of a means to score the toxicity responses of the selected bacteria and to compare these responses to the responses of these bacteria to known chemical agents.

Another object of this invention is to establish means for evaluating toxicity responses for compounds whose genotoxic responses have been difficult or even impossible to detect with other state-of-the-art mutagenicity assays.

This invention has the further object to provide, as a reference, DLS patterns generated when these bacteria are treated with identified chemicals, to screen for toxicity and to identify and quantify samples.

A further object of this invention is the provision of a medium for solubilizing water insoluble samples to be tested, this medium being harmless to the assay bacteria and being optically clear when mixed with the bacteria so that radiation may readily penetrate, and scattering may be detected.

It is still another object of this invention to provide for metabolic activation of certain chemicals to be tested so that these chemicals may be taken up by the bacterial cells. The foregoing and other objects of the invention are achieved by the provision of

Bacillus subtilis isogenic strains of bacteria lacking certain genetic repair functions and strains that are repair efficient that can be monitored for their unique responses to toxic chemicals using a differential light-scattering laser system. These responses are then scored to determine, for example, acute toxic or genotoxic effects. These scores, calculated as response indices, are then compared to known responses of these bacteria strains to identified chemical agents. It is to be understood that this method is also capable of assessing quantifiably the presence of a chemical to be tested. A better understanding of the disclosed embodiments of the invention will be achieved when the accompanying detailed description is considered in conjunction with the appended drawings, in which reference numerals are used for the same parts as illustrated in the figures.

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a schematic view of the method and apparatus of the claimed invention.

Figures 2a through 2d are differential light scattering patterns generated when the assay bacteria is exposed to certain concentrations of 4-nitroquinoline-N-oxide ("4NQO").

Figures 3a and 3b are differential light scattering patterns generated when the assay bacteria is exposed to certain concentrations of N-methyl-N'-nitro-N-nitroso-quanadine ( "MNNG" ). Figures 4a through 4d are differential light scattering patterns generated when the assay bacteria is exposed to certain concentrations of benzo(a)pyrene ("B(a)P").

Figures 5a through 5d are differential light scattering patterns generated when the assay bacteria is exposed to certain concentrations of "Lindane".

DESCRIPTION OF THE PREFERRED EMBODIMENTS Figure 1 provides a schematic illustration of the claimed apparatus by which samples are bioassayed in accord with the claimed invention. The bioassay consists of a radiation source 4, such as a laser system, which in the preferred embodiment employs the laser system such as that disclosed in Wyatt U.S. Patent No. 4,541,719, which includes a laser producing monochromatic visible or infrared radiation that is plane polarized with respect to two arrays of detectors. The laser system is capable of making 1200 measurements/second at 15 unique angles. The bioassay further includes an isogenic set of Bacillus subtilis mutants 2. The laser illuminates the cuvette 1 containing the bacteria 2 in suspension, these bacteria causing the light to be differentially scattered. The differential light scattering (DLS) 6 of the laser beam is detected as scattered light intensity as a function of angle. In one embodiment, the detectors 5 detect the DLS and feed the signals to a recorder 7, which records, in the form of a graph, a plot of the scattered intensity. In another embodiment, an analysis apparatus such as a computer 8 receive the data from the detectors 5 directly.

The intensity variation in the scattered light is called the DLS pattern, and is dependent on the average size and structure of the bacteria as well as the population size distribution. The DLS pattern of a suspension shows how many particles are present, their size, shape and the distribution of particles. Thus, if the bacteria undergo cell shrinkage, enlargement, division or lysis, the DLS pattern will shift from the control pattern of untreated bacteria. Bacterial count is expressed as a measure of height, or the highest point, of the DLS pattern. The shift of peaks of a sample DLS pattern indicates the degree of morphological change as compared to the control pattern. It is to be understood that the DLS pattern may be in magnetically recorded digital form for machine readability, or recorded in graphic form such that the scattered intensity pattern is represented as an eye readable curve.

In order to facilitate data entry and calculation where data from the detectors 5 is not transmitted to a recorder 7, a digitizing tablet (not shown) is used to trace each pattern's curve and simultaneously transmit a continuous stream of x and y coordinates to an analysis apparatus such as a computer 8. As discussed, the use of a digitizer can be eliminated by use of an integrated laser-computer system so that input into the system is much more rapid and avoids the potential for input errors. Each embodiment is shown in Figure 1 by phantom line connections. The analysis apparatus, which is in the preferred embodiment a computer 8 carries out the required calculations of response indices representing the shift in the DLS of the control as compared to the sample, and the change in bacterial count. The derived indices provide an immediate preliminary indication of toxic response, the DNA damaging (genotoxic) potential and aid in the identification of "isosets" of the Bacillus subtilis bacteria for compound identification.

The concentration of the chemical sample is determined by the level of response by the bacteria to the given sample. The set of bacteria selected for the bioassay each differ by only one property, which causes them to be more sensitive than other set members because of the mechanism of toxicity. Therefore, a "fingerprint", or profile unique to each chemical is generated from the differential response of the member bacteria. By "scoring" these fingerprints, the bioassay provides data for identifying test chemicals, assessing the concentration of test chemicals, and assessing whether the test chemical is acutely or chronically toxic. Another embodiment of the invention provides a library 9 wherein are stored the DLS patterns and calculated response indices for the bioassay of identified samples. By comparing the sample DLS with the stored pattern, the sample may be screened for toxicity, identified, and/or quantified.

The time required for a complete assay is approximately sixty (60) to sixty-six (66) minutes, the time required for untreated bacteria to divide twice in the absence of any chemical sample. However, individual samples may be analyzed at a rate of one sample per minute to determine the concentrations of a chemical in an aqueous sample. The Bacteria: The bacteria used in the bioassay are of a single species, Bacillus subtilis, which has numerous advantages over other bacteria traditionally used in bioassays. First, Bacillus subtilis is the most widely studied and genetically mapped gram-positive microorganism. Henner, D.J. and Hoch, J.A., The Bacillus subtilis Chromosome, Microbiol. Rev. 44:57-82 (1980).

The morphology, biochemistry, genetics and physiology of these organisms are therefore thoroughly understood. The morphological features of these bacteria are also easily adapted to the requirements of sensitivity and reproducibility of the bioassay. In its dormant form, Bacillus subtilis exists as a spore that is capable of surviving genetically unchanged while resisting heat, drying or other adverse effects for an indefinite period. As a vegetative cell, the bacteria can divide rapidly to double itself in approximately twenty-six (26) minutes during the logarithmic growth phase, and is much more sensitive at this stage to chemical toxicants than are the gram-negative microorganisms, Salmonella typhimurium and Escherichia coli. In its competent state, Bacillus subtilis readily takes up large molecules such as DNA, and can integrate genes derived from members of the same species or other species through traditional recombinant DNA technology.

The bioassay uses Bacillus subtilis mutants that are genetically identical, or isogenic, except for one or more genetic blocks in unique enzymatic repair processes or steps that restore chemically damaged DNA to its functional condition. Among the isogenic set of Bacillus subtilis are specifically selected strains deficient in different recombination (Rec-), excision (exc-), polymerase (Pol-) or spore repair (Spp-) repair steps. These strains may be deficient in one or a multiple of repair steps, and include the Rec- strains, recA1, recA8, recB2, recC5, recD3, recE4, recG13, mc-1, and m45; the Pol- strains T-1, TKJ8201; the Exc- strains her-9 and TKJ8206; The Exc- and Rec- strain FH2006-7; the Exc-, Rec- and Pol- strain HJ15; the Exc- and Spp- strain TKJ5211; the Exc-, Pol- and Spp- strain TKJ6321; and repair efficient strains HA101, 168 and 168 wild type. The mutants in this group are therefore more sensitive to the metabolic toxicity or genotoxic activity or a broad range of chemical substances at nannomolar or nannogram concentrations.

Since there are many repair processes and many enzymatic steps in these processes, a given chemical will affect only certain of the isogenic mutants, producing a unique DLS pattern. These Bacillus subtilis mutants were developed by treating them with by irradiation or with various chemical mutagens. For example, nitrosoquanidine, a powerful mutagenic chemical, was used to change the genetic makeup of the bacteria and trimethoprim to select for mutant types that could not synthesize DNA unless thymine was provided to them. The mutants from one group isolated by Felkner were designated FH2001 through FH2006 and had been derived from a parent strain JB01-200 whose lineage is defined in Felkner J. Bacteriol. 104:1030-1032 (1970). As discussed in this reference, conventional genetic analysis determined where the unique thymine gene defect was located in the chromosome of this bacteria. Subsequently, a mutant that was unusually sensitive to ultraviolet light and to the carcinogenic chemical 4-nitroquinoline-1-oxide was isolated by using trimethoprim treatment. Felkner, I.C., Isolation and Characterization of Thymineless Mutants from Ultraviolet-Sensitive Bacillus subtilis Strains, J. Bacteriol. 104:1030-1032 (1970). This strain, designated FH2006-7 was found to have a gene mutation genetically linked to the thymine locus that cause the mutant to be unable to carry out genetic recombination. The. genes from this mutant were transferred to a parental "wild type" strain by DNA isolation, uptake and integration so that the FH2006-7 descendants with the following genotypes were produced:

(1) trp-, thy-, her-, rec-

(2) trp+, thy-, her-, rec-

(3) trp+, thy+, her-, rec- Types 2 and 3 are mutants suitable for assaying her- and rec- mutations to detect DNA damage. These bacteria, constructed by conventional genetic engineering techniques would not occur in nature or result in a surviving bacterial isolate. Among the nineteen strains used with this bioassay, recA8, recA1, recB2, recC5, recD3, recE4, recG13, her-9, mc-1 and FH2006-7 are descended from strain 168. This strain is a descendent from the strain trp C of John Spizizen. " Spizizen, J. Transformation of biochemically deficient strains of Bacillus subtilis by deoxyribonucleate, Proc. Nat. Acad. Sci. U.S.A. 44:1072-1078. The remainder of the mutants are descended from this strain but are of direct lineage of HA 101 described in Kada, T., Sadaie, Y. and Tutikawa, K., Mutat. Res. 16:165 (1972) and Tanooka, H., Mutat. Res. 42:1 (1977). The mutant strains employed in the preferred embodiment of the bioassay may be obtained through application of conventional genetic engineering and transformation techniques by following the teachings provided in the following, and the references cited therein: Henner, D.J. and Hoch, J.A., The Bacillus subtilis Chromosome, Microbiol. Rev. 44: 57-82 (1980); Laumbach, A.D., Felkner, I.C., Formations of a 4-nitroquinoline-1-oxide complex in normal and repair deficient strains of Bacillus subtilis, Mutat. Res. 15: 233-245 and 16:444 (1972); Felkner, I.C., Isolation and Characterization of Thymineless Mutants from Ultraviolet-Sensitive Bacillus subtilis Strains, J. Bacteriol. 104:1030 (1970); Felkner, I.C., Kadlubar, F., Parallel Between Ultraviolet Light and 4-Nitroquinoline-1-Oxide Sensitivity in Bacillus subtilis, J. Bacteriol. 96:1448 (1968); and I.C. Felkner, Microbial Testers: Probing Carcinogenesis (1981).

Isosets, defined as a subset of these mutants, constitute the minimum number required to define a response to a sample tested. Data derived form fractional survival tests and data derived from spot assays are used to determine which B. subtilis strains should be selected to comprise an isoset to identify a given compound by the radiation bioassay.

To select isoset members to be used in the bioassay, the isogenic mutants are pre-screened using the spot test with given chemical samples. The DNA-damaging test is performed according to the procedure disclosed in Felkner, I.C., Microbial Testers: Probing Carcinogenesis (1981) and Felkner, I.C. in Medine, A., and Anderson, W., eds. Proceedings of the National Conference on Environmental Engineering, Am. Soc. Civil Eng., N.Y. pp. 204-209 (1983). Each mutant is inoculated into 5 ml of Brain Heart Infusion (BHI) broth, either from a disc containing approximately 1 × 10⁷ washed spores or a 1 mm loopful from a sporulating slant culture, and incubated by shaking at 37 to 39° C for 16 hours. This culture is designated o/n. Using a 1 mm inoculating loop, inocula from the o/n cultures are streaked radially on a nutrient agar plate to a sensitivity disc containing 10 ul of the assay chemical which may or may not have been pre-incubated with a rat liver microsomal fraction (S-9) in order to achieve metabolic activation. After incubation at 37 to 39° C, the distance (mm) of growth inhibition from the periphery is measured with a vernier caliper (Manostat). If varying concentrations of the assay substance are used, a concentration-dependent growth inhibition curve can be constructed, using the data from responsive mutants. Certain chemicals are not readily uptaken by the assay bacteria without metabolic activation. Liver microsomes, designated an S-9 fraction, were prepared from Sprague-Dawley rats induced with Aroclor 1254 (Litton Bionetics, Charleston, SC 19405). The S-9 fraction was prepared as a KCl homogenate with a protein concentration of 25-28 mg/ml. It is understood that this fraction may be microsomal or post-microsomal and is not limited by the origin of the tissues. The "Cocktail":

Many promutagens/carcinogens are water insoluble. A "cocktail" was therefore derived so that a clear aqueous solution of the test chemical could be mixed with the assay bacteria and the hydrophilic S-9 fraction, thereby minimizing interference with the radiation source and its ability to generate a DLS pattern of the radiation scattered by said bacteria. The "cocktail" has the advantage of solubilizing chemicals in an aqueous system that were not as readily solubilized by other known procedures using dimethylsulfoxide (DMSO), ethanol or acetone before being introduced into the aqueous assay mixture. Moreover, the "cocktail" is nontoxic to the assay bacteria and permits the S-9 fraction enzymes to metabolize the test chemical. The preferred embodiment of the bioassay employs a dispersant sold under the trademark ".COREXIT 7664 Oil Dispersant" (Exxon Chemicals, Clark, New Jersey) and a non-ionic surfactant sold under the trademark "EMULPHOR EL-620" (GAF Corp., 140 West 51 Street, New York, N.Y. 10020). Both the "COREXIT" dispersant and the "EMULPHOR" surfactant are readily obtainable at the present time, and are generally known to chemists and commercial users. The chemical compositions of each of these are maintained as trade secrets by the respective owner companies.

The "EMULPHOR" surfactant is a polyoxylated vegetable oil.

At 25° C it is a clear, light brown liquid, having a specific gravity at this temperature of 1.04 to 1.05 and a viscosity of 600 to 1,000 cps. At 25° C, it is readily soluble in water, and its neutralization number is 0.5 maximum. As discussed, this composition was chosen because it is nontoxic to the Bacillus subtilis bacteria and when mixed with the chemical sample to be tested and the bacteria and the hydrophilic S-9 fraction should metabolic activation be required, a clear suspension results that does not interfere with the ability of the laser radiation to penetrate or the intensity of the radiation to scatter. "COREXIT 7664" is a surfactant ester having a light amber color and an alcoholic odor. AT 15.6° C it has a specific gravity of 1.03, a density of 8.59, and a viscosity of 29. It is soluble in fresh water and seawater and is dispersible in hydrocarbons. As discussed, this compound was selected as a component of the "cocktail" because it is nontoxic to bacteria and is capable of mixing with aqueous suspensions of bacteria without clouding and thereby does not interfere with the ability of laser radiation to penetrate the suspension. The Radiation Source:

The radiation source 1 of the bioassay is a laser such as that disclosed in Wyatt, U.S. Patent No. 4,541,719. This instrument records scattering intensity at fifteen angular locations, permitting the relationships between refractive index, size and shape to be measured. Moreover the read head is anodized, sealed and attached directly to the laser head as well as to an integrated circuit board, providing a unitized optical bench and a grounded enclosure for the detector elements. The specifications of the laser employed in the preferred embodiment of the bioassay are as follows: Light source: 5 mW He-Ne laser, λ=632.8 nm. Plane polarized

Angles Monitored Simultaneously: 15 @ 0.2 ≤ sin 0/2 ≤ 0.9 in steps of 0.05

Computer interface: Parallel analog outputs interfaced to 16-channel, 12 bit A/D programmable multiplexer.

Data Collection: 12,500 conversions/second - 25,000 conversions/second

Single Particle Measurements: Concentration ≤ 5 × 10

3/ml Size (depending on refractive index) 0.6 to 50 um

Detectors: Transimpedance photodiodes with built-in amplifiers Detector Output Dynamic Range: 0 - 10 V Dynamic Range: 10⁴ Linearity: ± 0.1%

Dark Current Noise: ± 0.2 mV

It is to be understood that this laser apparatus comprises part of the preferred embodiment of the bioassay, but that any laser instrument capable of the objects of this invention is comprehended. Radiation Bioassay Examples: Examples are provided for the bioassay of 4-nitroquinoline-N-oxide ("4NQO"), N-methyl-N'-nitro-N-Nitroso-guanidine ("MNNG"), benzo(a)pyrene ("B(a)P"), and "Lindane" (1,2,3,4,5,6-hexachlorocyclohexane). The chemicals 4NQO and MNNG are well established as mutagens and DNA damaging agents to bacteria, and are water-soluble and direct-acting genotoxic chemicals that elicit carcinogenic activity. The chemical B(a)P is a well known procarcinogen/mutagen that is insoluble in an aqueous system, and requires metabolic activation by rat liver microsomal fractions (S-9) to become electrophilic and to cause DNA damage or a mutagenic response. "Lindane" is a chemical difficult to assay for its genotoxic potential by state of the art mutagenic assays.

To perform the bioassay according to the preferred embodiment, the following protocol is followed:

1. Lyphilized tablets *, containing the appropriate Bacillus subtilis strains and dehydrated growth medium are introduced into cuvettes containing water. Suspensions of bacterial cells at the required concentration described in the examples provided infra may also be used.

2. The inoculated cuvettes are placed in an incubation chamber that has been prewarmed to approximately 37 to 39° C, and incubation continued for about one hour to one and a half hours to establish a culture in the "assay-ready condition".

3. Test or control material is introduced into the cuvettes using one pair of cuvettes for each assay organism. A negative control having no test material, and a positive control having an identified toxic chemical at a concentration whose response has been determined is included.

4. The cuvettes are incubated for six minutes at 37-39° C, and each cuvette is read a minimum of fifty

(50) times (the total reading time per sample is 30 seconds).

5. Incubation at 37-39 degrees Centigrade is continued for sixty (60) to seventy (70) minutes and each cuvette is again read a minimum of fifty (50) times.

6. The samples are scored, and the bacterial responses evaluated.

* The bacteria could be lyphilized vegetative cells or spores in a capsule form. However, when spores are used, this form being preferred because of their stability, they must be heat-shocked at 60 degrees

Centigrade for 8 to 10 minutes before incubation. An alternative protocol follows: 1. Solubilize a sample in an aqueous solution.

For water insoluble samples, use a sufficient amount of

"cocktail" to prevent precipitation of the sample.

2. Make serial dilutions from the solubilized sample. 3. Add the capsule containing the bacterial spores and the dehydrated growth medium to a cuvette containing water.

4. Add a measured amount of each solubilized sample dilution to cuvettes containing bacteria and place the cuvette in an incubator at 37-39 degrees

Centigrade. Include a negative control having no test material, and a positive control having a toxic chemical at a concentration whose response has been determined.

5. Read the samples at 0, 6 and 66 minutes to follow the relative growth indicated by the increase in number of particles for each dilution.

6. Score each sample dilution and the controls.

7. Compare the score data to the score data of known compounds to determine a matching response profile.

8. Determine the concentration from a known dose response curve. If the sample is an unknown, search for a matching chemical response to determine its identity and concentration. It is understood that a negative control may also comprise a control sample of bacteria exposed to 4-nitroquinoline-N-oxide or 4-nitroquinoline-1-oxide and a positive control of bacteria exposed to benzo(a)pyrene. Cultivation of bacterial cells appropriate in concentration for the bioassay includes heat shocking strains of Bacillus subtilis in 1 ml of Brain Heart Infusion (BHI) broth for 20 minutes. Additional broth is added until an absorbance of 0.2 A(540) is achieved. The cultures are incubated at 37 to 39° C until the bacterial growth reaches an A(540) of 0.5, a period of approximately 2 hours. Cultures that require a longer period to reach this density are not considered suitable for the bioassay. Cultures found suitable are diluted to an A(540) of 0.3 with BHI broth yielding approximately 2 × 10⁷ bacteria/ml. These cells, when diluted by the amount required for the cuvette assay, yield a cell concentration of 10⁶ bacteria/ml.

1. Bioassay without metabolic activation (Cuvette Assay) - Compounds and bacterial isoset:

4NQO was assayed at final concentrations of 8.5, 4.2 and 2.1 ug/ml using Bacillus subtilis strains recE4 (6), 168 or 168 wild type (11). The DLS patterns for each of these assays comprise Figures 2a through 2d.

MNNG was assayed at a final concentration of 3.6 ug/ml with strains 6 and 11. The DLS patterns for each of these assays comprise Figures 3a and b.

Lindane was assayed at final concentrations of 53.6 ug/ml and 107 ug/ml utilizing Bacillus subtilis strains FH2006-7 (10) and 11. The DLS patterns for each of these assays comprise Figures 5a through 5d. Method:

In a screw cap tube containing 13.6 ml of deionized water at 37 to 39° C is pipetted 0.3 ml of bacteria and 0.1 ml of the test chemical. The tube is inverted gently and its contents poured into a cuvette. The cuvette is illuminated by the laser and the intensities of scattered light measured and scored.

The cuvette is then incubated at 37-39 degrees

Centigrade for approximately one (1) hour and illuminated identically. A negative control of bacteria in deionized water is also concurrently assayed to provide a baseline value for each strain.

The method may also be performed using lyphilized bacteria having with it sufficient dehydrated BHI to support the growth during the assay period. The total volume of fluid required to obtain a reading from the cuvette is 10 ml. Therefore, a volume of about 0.4 ml of fluid will rehydrate the cell-medium. The incubation period remains unchanged, and illuminating is performed at fifteen minute intervals.

2. Bioassay with Metabolic Activation (Preincubation) - Compound and bacterial isoset:

B(a)P was assayed at final concentrations of 15.1 and 7.6 ug/ml using Bacillus subtilis strains 11 and TKJ6321 (19). Method: Into a large capped test tube were pipetted 0.5 ml each of "COREXIT" and "EMULPHOR", B(a)P at the appropriate concentrations, 0.5 ml of S-9 fraction and 0.3 ml bacteria. The volume was adjusted to 14 ml with deionized water pre-warmed to 37° C. A control tube for each strain containing "COREXIT", "EMULPHOR", S-9 fraction and deionized water was prepared according to the same proportions as the sample to be tested. The tubes were incubated in a shaking water bath for approximately one (1) hour. After incubation, the contents were poured into cuvettes and the scattered intensities measured and scored, these readings being taken at zero and sixty minutes, as previously described. When lyphilized bacteria and growth media are used, a smaller volume of said bacteria are used to obtain the desired initial density in one (1) hour before treating the bacteria with the test chemical. The "COREXIT", "EMULPHOR" and S-9 volumes are each approximately 180 ul per 5 ml of sample.

3. -Bioassay with Metabolic Activation (Cuvette Incubation) - Compound and bacterial isoset:

B(a)P was assayed at final concentrations of 15.1 and 7.6 ug/ml using Bacillus subtilis strains 11 and TKJ6321 (19). The DLS patterns generated by this assay comprise Figures 4a-4d. Method:

Into a scanning cuvette was pipetted a sufficient volume of Corexit to prevent precipitation of B(a)P when introduced into the aqueous incubation mix of bacteria dn S-9 fraction. The "CORERXIT"- B(a)P mixture was introduced into the cuvette along with 0.5 ml of S-9 fraction and 0.3 ml of bacteria in a total volume of 14 ml of deionized water pre-warmed to 37° C. A concurrent control tube for each strain was also prepared. The cuvettes were read and the scattered intensities scored at zero and sixty (60) minutes as previously described.

In an alternative method, the aqueous cocktail of "COREXIT", "EMULPHOR" and water is prepared prior to addition of the S-9 fraction and the bacteria. Thus, 180 ul each of "EMULPHOR" and "COREXIT" are added to each 5 ml of deionized water and mixed. At the time of the assay, 180 ul of S-9 fraction is added along with appropriate amount of lyphilized bacteria and medium required to achieve a cell concentration of 10 6 bacteria/ml.

Response Evaluation:

The scoring techniques of the bioassay measure the height of the DLS pattern as a function of bacterial count and the amount of shift in the peaks of the chemical sample pattern and the control sample pattern to determine the degree of morphological change in size and shape of the bacteria. The derived indices provide an immediate preliminary indication of the toxic response, the DNA damaging (genotoxic) potential, and identify isosets for compound identification.

The measurement of each pattern's height provided an indication of bacterial count and is therefore the principal indicator of toxicity. In developing the scoring techniques for the identification of morphological changes, it was noted that there is a function which is continuous on a closed interval from the beginning of the pattern to its end. Each line segment between any two adjacent points formed an arc, and associated with each point on the X-axis is only one point on the pattern. Using a digitizing tablet, it is possible to obtain a sufficiently large number of points so that the sum of the distances between all adjacent points would approximate the length of the pattern. This was defined as the arc length of the curve. Variances of the arc length correspond with the variances in the bacteria morphology, especially as they related to individual cell size. In order to simplify the interpretation of a response to a given treatment by a chemical at a given concentration, a system using three (3) digits was developed, each of which was the indicator for a response. The nature of this response was assigned a value of one (1) through four (4) according to the height and direction of pattern displacement as a function of the type of effects on the bacterial population. To index a set of treatments, the count changes and patterns of a baseline control sample were measured. The subsequent treatment and control samples were compared to the baseline response.

Index calculations:

Four patterns were digitized in the following order: Control at time 0 and 60 minutes; treated sample at time 0 and 60 minutes. For each pattern in order, the response indices were determined by calculating the distance between adjacent points on the curve as the arc length. The mean Y of all coordinate pairs was calculated for the bacterial count index. The control at time "0" was designated as the baseline control. For each of the other curves, the percent change between the pattern's mean Y and the baseline pattern's mean Y were calculated. For the Fourth order Y index, the Y percent change was divided by 5 and, if the change was positive, 1 was added. If the change was negative, 1 was subtracted. The final integer value was retained. The percent change between the test pattern's arc length and the baseline pattern's arc length was calculated. To determine the fourth order arc index, the arc percent change was divided by 5 and, if the change was positive, 1 was added. If the change was negative, 2 was subtracted. The integer value was retained.

Each of the three digits, from left to right in the PRI index, derives from the following table according to the fourth order indices of the patterns, starting with the control at "0" time as the reference and the control at 60 minutes as the first digit.

The biological interpretations that can be made on a digit-by-digit basis are as follows:

"1" indicates a positive increase in bacterial count with a concurrent decrease in size of said bacteria (positive shift), and a non-toxic response.

It is normal for rapidly dividing cells to decrease in size.

"2" indicates a positive increase in bacterial count with a concurrent increase in size (negative shift), and a minimum toxic response.

"3" indicates a decrease in bacterial count with a concurrent increase in size (negative shift), and a toxic and lethal response.

"4" indicates a decrease in bacterial count with a concurrent decrease in size (positive shift), and a toxic and lethal response. The tertiary responses for each of the indexed patterns were obtained by multiplying the fourth order Y index by the fourth order arc index. The secondary response index (SRI) was determined by the calculation of the sum of the tertiary indices for all indexed patterns combined. This method of calculation is illustrated by the following data derived from an assay:

The interpretation of the PRI of 132 is that the control of sixty-six (66) minutes displayed normal growth, hence the number 1 as the first digit. The treated sample at time "0" showed a loss in bacterial count accompanied by size increase, hence the number 3 as the second digit. By sixty-six minutes, the treated sample showed an increase in bacterial count accompanied by a size increase, hence the number 2 in the third digit.

Two extreme examples of PRIs are 133 and 144. These examples are immediately toxic responses from which the cells fail, to recover. The 133 PRI shows cell enlargement, and the 144 PRI shows cell shrinkage, both being associated with the failure of the cell to recover but by different mechanisms of action. The presence of a 3 or 4 in the first digit of any PRI should mean that the assay is invalid because the bacterial control culture is not displaying normal growth.

A screening score S provides a precise measure of bacterial growth or inhibition by bacteria in the test sample as compared to the control bacteria. It is defined in terms of DLS placement scores and was implemented by previous studies. Wyatt, P.J., Phillips, E.T., Scher, M.G., Kahn, M.R., and Allen, E.H., J. Argic. Food Chem. 25:1080 (1977); Wyatt, P.J., Scher, M.G., and Phillips, D.T., J. Agric. Food Chem. 25:1086 (1977). Thus,

10(Dbc/300) - 10(Dc/300) S = 300 log

10(Dbt/300) - 10(Dt/300) where Dbc is the DLS displacement of the assay broth plus bacteria relative to the assay broth (control), Dc is the displacement of the control blank relative to itself (0 when there is no background interference), Dbt is displacement of test material in assay broth plus bacteria relative to the control, and Dt is displacement of the test material in assay broth relative to the control blank. The equation converts the data logarithmic DLS patterns back to linear form so that the background is subtracted and hence converted to the usual logarithmic form. The equation yields a factor of 300 for 10-fold increases of control bacteria with respect to test bacteria. applying the index calculations to the samples assayed:

In all the figures, the untreated "0" time control is represented by a solid line the one hour untreated control by a dot broken line ( . . . . ), the "0" time treated sample by a widely spaced broken line ( - - - - ), and the one hour treated sample by a narrow spaced broken line (_ _ _ _ ).

4NQ0 DLS Toxicity Responses:

Figures 2a through 2d show the DLS response patterns for strains 6 (rec 34) and 11 (168 wild type) respectively. From the mean Y values of the "0" time and 1 hour samples, it is clear that increases in cell counts occurred in all untreated controls. From Figures 2a and 2b, it is clear that 4.2 ug/ml and 2.1 ug/ml of 4NQ0 caused increased arc lengths and curves shifting to the left relative to the light scattering angles of the negative time "0" control. The relative intensity increased less from the "0" to 1 hour interval. From Figure 2c, it is clear that 8.5 ug/ml of 4NQO caused a decrease in arc length and a curve shift to the left relative to the light scattering angles of the negative time "0" control. However, at 4.2 ug/ml, the change in arc length was not substantial and the curve shift relative to light scattering angles was slightly toward the larger angles.

In applying the response index calculations, the following interpretations can be made. Strain 6 at 4.2 and 2.1 ug/ml had PRI values of 132 and 131, respectively. At the 4.2 ug/ml level, an immediate, toxic and lethal effect was seen at "0" time. The bacteria showed some recovery because the counts increased by 60 minutes. However, the bacteria remained enlarged. At the 2.1 ug/ml level, an immediate, toxic and lethal effect was seen at "0" time, but the counts increased and cells were of normal size at 60 minutes. The 4.2 ug/ml concentration, however, did not adversely affect strain 11. These data show that strain 11 was more refractile to the effects of 4NPO than was strain 6, even when strain 11 was treated at higher dose levels. Thus, the genotoxic manifestation is expressed at the primary response level.

MNNG DLS Toxicity Responses: Figures 3a and 3b show the DLS response patterns for strains 6 (rec E4 ) and 11 (168 wild type) respectively. From the mean Y values of the "0" time and 1 hour samples, it is clear that increases in cell counts occurred in all untreated controls. From Figure 3a, it is clear that 3.6 ug/ml of MNNG caused a decreased arc length and a curve shift to the left, relative to the light scattering angles of the negative time "0" control. The relative intensity increased less than did the negative control from the "0" to 1 hour interval. from Figure 3b, it is clear that 3.6 ug/ml of MNNG did not cause a decrease in arc length, although the arc length increase was slightly less than the negative control. The curve shift relative to light scattering angles was toward the larger angles, almost approximating the negative time "0" control. Applying the calculations of the response indices, the following interpretations can be made. Strain 6 at 3.6 ug/ml had a PRI value of 132, indicating an immediate, toxic and lethal effect at "0" time. The bacteria showed some recovery because the counts increased by 60 minutes. However, the bacteria remained enlarged. Strain 11 at the concentration of 3.6 ug/ml had a PRI value of 111, thus showing no compound-related toxicity expressed either immediately or over the 60-minute period following treatment. This is a genotoxic response expressed at the primary response level. Additionally, the SRI value for strains 6 and 11 were -3 and 19, respectively. The negative secondary response of strain 6, when compared to the positive SRI of strain 11, indicates that there is a genotoxic effect which can be repaired by the wild type (strain 11) but not the Rec- mutant (strain 6). Lindane DLS Toxicity Responses:

Figures 5a through 5d show the DLS response patterns for strains 10 (FH2006-7) and 11 (168 wild type). From the mean Y values of the "0" time and one hour samples, it is clear that increases in cell counts occurred in all untreated controls. From Figures 5a and 5b, it is clear that 107 ug/ml and 53.6 ug/ml of Lindane initially caused decreased arc lengths and curves shifting to the left relative to the light scattering angles of the negative time "0" control. However, at the lower level, a positive shift was seen at sixty minutes. The relative intensity did not change substantially from the "0" to one hour intervals at the higher level. However, an increase was seen at the lower concentration. From Figures 5c and 5d, it is clear that both concentrations of Lindane caused an initial decrease in arc length and a curve shift to the left relative to the light scattering angles of the negative time "0" control. However, the arc lengths increased and the scattering angle shifted to the right relative to the negative time "0" control. However, the arc lengths increased and the scattering angle shifted to the right relative to the negative time "0" control after one hour. Applying the response index calculations, the following interpretations can be made. Strain 10 at 107 and 5 3.6 ug/ml had PRI values of 222 and 131, respectively. At the 107 ug/ml level, a toxic and a sustained effect (cell enlargement) was seen. However, at the 53.6 ug/ml level, the bacteria recovered after 60 minutes. Strain 11 had PRI values of 131 at both concentrations. This indicated that at both levels, cell lethality and enlargement occurred but that full recovery resulting in growth and the return to normal size occurred by one hour. Although strain 10 may be slightly more affected than strain 11 at the higher concentration of Lindane, the differences in the PRI values do not appear to be useful in making toxicity assessments. The SRI values, however, were substantially different for strains 10 and 11. The SRI values for strain 10 at 107 ug/ml and 53.6 ug/ml were -19 and 7, respectively. The SRI value was 10 for strain 11 at both 107 ug/ml and 53.6 ug/ml. The does-relates SRI trend seen in strain 10 but not in strain 11 indicates a genotoxic effect of this chemical, as demonstrated by the negative SRI value at the higher dose level. As shown by the positive SRI values, strain 11 at both concentrations, was capable of repairing genetic damage and dividing normally. Strain 10 was capable of repairing genetic damage and dividing normally only at the lower concentration of 53.6 ug/ml. From these results, it is clear that primary cytotoxicity may be expressed by a three digit number. Using the primary response of mutant and wild type bacteria, some compounds such as MNNG and 4NPO which are highly toxic can also be identified as genotoxic. Compounds such as B(a)P when activated with an S-9 fraction cannot be shown to be genotoxic by using the primary response level alone. However, genotoxicity may be determined by calculating the secondary response. Other compounds such as Lindane, whose genotoxic responses have been difficult or even impossible to detect with other state of the art mutagenicity assays are readily detected by the radiation/microbe bioassay by calculating their secondary response indices.

Thus, it will be seen that all embodiments of the present invention provide a unique apparatus and method for radiation bioassay samples. While preferred embodiments of the invention have been disclosed, it should be understood that the spirit and scope of the invention is not to be limited solely by the appended claims, since numerous modifications of the disclosed embodiments will undoubtedly occur to those of skill in the art.

RADIATION/MICROBE BIOASSAY

BACKGROUND OF THE INVENTION The present invention is directed to the field of bioassays, and in particular to an assay utilizing radiation, or laser, technology in combination with unique genetically engineered bacteria that provide a "fingerprinting" capacity to identify and measure concentrations of toxicants in various environmental media.

It has long been a concern among segments of society that products developed through chemical technology and intended to benefit humanity may be responsible for dramatic increases in the incidence of cancer and may cause long-term decline in genetic health. The ability to efficiently and accurately monitor environmental media such as water, soil, air and food for natural and man-made toxicants capable of mutagenesis and carcinogenesis is therefore one of the most crucial scientific requirements in public health.

Methods of analytical chemistry such as gas chromatography, mass spectroscopy, and high pressure liquid chromatography are available to test air, water or soil samples. However, sample preparation can be complex and different assays must be developed for the identification and quantification of individual chemicals or chemical classes. Moreover,the knowledge that a chemical is present and its concentration does not indicate its cytotoxic or genotoxic activity. These tests do not indicate whether an adverse biological effect is strong, weak, or even present in a sample in order to make reasonable risk evaluations for human beings.

The most widely employed bacterial test system for carcinogens is the Ames test, Ames, Mutat. Res. 31:347- 349 (1974), using Salmonella typhimurium histidine- requiring auxotrophs. Because the Ames assay relies on reverse mutation, a mutant deoxyribonucleic acid (DNA) locus must be mutated back to its wild-type configuration or be suppressed. Thus, chemical mutagens that cannot effect this change in the DNA will go undetected. Researchers have recognized other shortcomings in this test, specifically an inability to devise an in vitro activation system that is reliable and reproducible, and an inability to detect some mutational events that can lead to cancer. Felkner, Microbial Testers: Probing Chemical Carcinogenesis, 177 (1981).

In 1971, Felkner discovered that ordinary transforming strains of Bacillus subtilis recovered from short exposures to the potent mutagenic and carcinogenic agent, 4-nitroquinoline-1-oxide (4NQO), while other strains incapable of recombination and/or dark repair did not. The ability or inability to recover was found to be directly correlated with the blocking of de novo deoxyribonucleic (DNA) synthesis through complexation of 4NQO with the DNA of the cell. Laumbach, A.D. and Felkner, I.C., Formation of 4-Nitroquinoline-1-oxide Complex with DNA in Normal and Repair Deficient Strains of Bacillus Subtilis, Mutat. Res. 15: 233-245 (1972). The product of this research was the discovery that a certain chemical structure has a corresponding particular biological function. Based on this principle, the effect of a chemical on a biological system, such as Bacillus subtilis bacteria, may be identified by the change in morphology of the assayed cells.

In 1981, Felkner developed a microbial bioassay from a collection of several hundred Bacillus subtilis mutants based on the principles derived from his earlier work. These selected strains had unique genetic defects causing them to be very sensitive to the effects of specific chemical classes, toxins and radiation. To enable uniformity, the mutant genes from these strains were introduced by recombination technology into identical clones of the wild type (normal) Bacillis subtilis strain. Thus, a battery of strains, differing only by a single character (gene) or limited set of genes was produced. The member strains are therefore isogenic, that is, identical except for single defined genes.

In a liquid suspension and by using conventional bacteriological methods such as microscopy, spectrography and colony or cell counts, it has been documented that chemical toxicants such as 4-nitroquinoline-1-oxide can block cell division, induce cellular swelling and cause cell rupture as a consequence of binding to and damaging the DNA of Bacillus subtilis. Laumbach & Felkner, Formations of a 4-nitroquinoline-1-oxide complex in normal and repair deficient strains of Bacillus subtilis, Mutat. Res. 15: 233-245 and 16:444 (1972). These parameters as well as the molecular interactions of many chemical classes have been documented for the Felkner mutant strains, indicating a unique response pattern for any assay chemical that is toxic. Felkner, Development of a Bacillus subtilis system to screen carcinogens/mutagens; DNA-damaging and mutation assays, Microbial Testers: Probing Chemical Carcinogenesis, 89-120 (1981); Song, Photoactivation of furocoumaryl carcinogens/mutagens and their interactions with nucleic acids, Microbial Testers: Probing Chemical Carcinogenesis, 35-66 (1981); Streips, Bacterial Mutation Monitors for active metabolites of chemical carcinogens: Bacillus subtilis assays for mutation and DNA repair, Microbial Testers: Probing Chemical Carcinogenesis, 131-143 (1981). This testing method, however, is dependent on conventional plate and liquid methods. Although deemed acceptable under the United States Environmental Protection Agency Office of Pesticide Program Guidelines as an acceptable way of monitoring genotoxic response, the method lacked the sensitivity, specificity and speed capable of the presently claimed bioassay system.

In the early 1970's, Wyatt and his collaborators discovered a laser light scattering bioassay technique. Exponential phase bacterial cultures were used to inoculate aqueous samples containing compounds that might affect bacterial physiological processes, and indirectly, the morphology of the exposed bacterial populations. A plot was derived of the variation of scattered light intensity as a function of scattering angle. Early instrumentation was based on placing a sample-containing cuvette at the center of a circular arc, about which a collimated photomultiplier detector was rotated. A graph representing the variation of the relative scattered intensity as a function of scattering angle was called the differential light scattering (DLS) pattern.

Bioassays employing bacteria are disclosed in Wyatt U.S. Patent No. 3,730,842, issued May 1, 1973, and Wyatt U.S. Patent No. 4,101,383, issued July 18, 1978. These patents disclose methods of determining bacterial sensitivity or susceptibility to antibiotics by using test differential scattering patterns derived from the test bacteria and then comparing them to patterns of control bacteria. Neither reference provides for generating a pattern to screen for a toxicant having potential cytotoxic or genotoxic activity or to provide a means by which the toxicant may be identified and quantified. Wyatt U.S. patent No. 4,541,719, issued September 17, 1985, discloses a laser apparatus that measures scattered radiation by microparticles such as bacteria at two well-defined scattering angles. The toxic response of bacteria is measured by averaging mean scattered intensities. Other patents naming Wyatt as an inventor include Wyatt U.S. Patent No. 4,693,602; Phillips U.S. Patent No. 4,616,927; Wyatt U.S. patent No. 4,621,063; Wyatt U.S. Patent No. 4,548,500; Wyatt U.S. Patent No. 4,490,042; Wyatt U.S. Patent No. 4,173,415; Wyatt U.S. Patent No. 3,928,140; Wyatt U.S. Patent No. 3,770,351; Wyatt U.S. Patent No. 3,754,830; and Wyatt U.S. Patent No. 3,624,835, which are directed to various methods and apparatus for characterizing microparticles.

Other bioassays have included Bean, U.S. patent No. 3,708,402, issued January 2, 1973, and Bean, U.S. Patent No. 4,061,543, issued December 6, 1977. Neither of these references discloses a method or apparatus by which toxicity or genotoxicity may be assessed, or a test chemical identified. Thilly, U.S. patent No. 4,299,915, issued Nov.10, 1981, discloses an assay for mutagenesis in bacterial cells wherein bacterial cells such as S. typhimurium are exposed to a test chemical and plated in the presence of a purine analog. The procedure is dependent on conventional plating techniques, and suffers the same problems of lack of sensitivity to certain chemicals and the requirement for a laboratory.

SUMMARY OF THE INVENTION Therefore, it is the primary object of this invention to provide a bioassay method and apparatus for rapidly detecting and measuring the toxic response from chemicals in the environment whose levels are too low to produce acute toxic symptoms, but from which a toxic response or damage might develop years following chronic exposure.

Another object of this invention is to provide unique differential light scattering (DLS) patterns for chemical agents using isogenic repair mutant and wild type strains of Bacillus subtilis.

A further object of this invention is to select from the Bacillus subtilis strains a minimum group or "isoset" required to establish response patterns to specific chemical classes.

This invention has as yet another object the provision of a means to score the toxicity responses of the selected bacteria and to compare these responses to the responses of these bacteria to known chemical agents.

Another object of this invention is to establish means for evaluating toxicity responses for compounds whose genotoxic responses have been difficult or even impossible to detect with other state-of-the-art mutagenicity assays.

This invention has the further object to provide, as a reference, DLS patterns generated when these bacteria are treated with identified chemicals, to screen for toxicity and to identify and quantify samples.

A further object of this invention is the provision of a medium for solubilizing water insoluble samples to be tested, this medium being harmless to the assay bacteria and being optically clear when mixed with the bacteria so that radiation may readily penetrate, and scattering may be detected.

It is still another object of this invention to provide for metabolic activation of certain chemicals to be tested so that these chemicals may be taken up by the bacterial cells. The foregoing and other objects of the invention are achieved by the provision of Bacillus subtilis isogenic strains of bacteria lacking certain genetic repair functions and strains that are repair efficient that can be monitored for their unique responses to toxic chemicals using a differential light-scattering laser system. These responses are then scored to determine, for example, acute toxic or genotoxic effects. These scores, calculated as response indices, are then compared to known responses of these bacteria strains to identified chemical agents. It is to be understood that this method is also capable of assessing quantifiably the presence of a chemical to be tested. A better understanding of the disclosed embodiments of the invention will be achieved when the accompanying detailed description is considered in conjunction with the appended drawings, in which reference numerals are used for the same parts as illustrated in the figures.

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a schematic view of the method and apparatus of the claimed invention.

Figures 2a through 2d are differential light scattering patterns generated when the assay bacteria is exposed to certain concentrations of 4-nitroquinoline-N-oxide ("4NQO").

Figures 3a and 3b are differential light scattering patterns generated when the assay bacteria is exposed to certain concentrations of N-methyl-N'-nitro-N-nitroso-quanadine ( "MNNG").

Figures 4a through 4d are differential light scattering patterns generated when the assay bacteria is exposed to certain concentrations of benzo(a)pyrene ("B(a)P").

Figures 5a through 5d are differential light scattering patterns generated when the assay bacteria is exposed to certain concentrations of "Lindane".

DESCRIPTION OF THE PREFERRED EMBODIMENTS Figure 1 provides a schematic illustration of the claimed apparatus by which samples are bioassayed in accord with the claimed invention. The bioassay consists of a radiation source 4, such as a laser system, which in the preferred embodiment employs the laser system such as that disclosed in Wyatt U.S. Patent No. 4,541,719, which includes a laser producing monochromatic visible or infrared radiation that is plane polarized with respect to two arrays of detectors. The laser system is capable of making 1200 measurements/second at 15 unique angles. The bioassay further includes an isogenic set of Bacillus subtilis mutants 2. The laser illuminates the cuvette 1 containing the bacteria 2 in suspension, these bacteria causing the light to be differentially scattered. The differential light scattering (DLS) 6 of the laser beam is detected as scattered light intensity as a function of angle. In one embodiment, the detectors 5 detect the DLS and feed the signals to a recorder 7, which records, in the form of a graph, a plot of the scattered intensity. In another embodiment, an analysis apparatus such as a computer 8 receive the data from the detectors 5 directly.

The intensity variation in the scattered light is called the DLS pattern, and is dependent on the average size and structure of the bacteria as well as the population size distribution. The DLS pattern of a suspension shows how many particles are present, their size, shape and the distribution of particles. Thus, if the bacteria undergo cell shrinkage, enlargement, division or lysis, the DLS pattern will shift from the control pattern of untreated bacteria. Bacterial count is expressed as a measure of height, or the highest point, of the DLS pattern. The shift of peaks of a sample DLS pattern indicates the degree of morphological change as compared to the control pattern. It is to be understood that the DLS pattern may be in magnetically recorded digital form for machine readability, or recorded in graphic form such that the scattered intensity pattern is represented as an eye readable curve.

In order to facilitate data entry and calculation where data from the detectors 5 is not transmitted to a recorder 7, a digitizing tablet (not shown) is used to trace each pattern's curve and simultaneously transmit a continuous stream of x and y coordinates to an analysis apparatus such as a computer 8. As discussed, the use of a digitizer can be eliminated by use of an integrated laser-computer system so that input into the system is much more rapid and avoids the potential for input errors. Each embodiment is shown in Figure 1 by phantom line connections. The analysis apparatus, which is in the preferred embodiment a computer 8 carries out the required calculations of response indices representing the shift in the DLS of the control as compared to the sample, and the change in bacterial count. The derived indices provide an immediate preliminary indication of toxic response, the DNA damaging (genotoxic) potential and aid in the identification of "isosets" of the Bacillus subtilis bacteria for compound identification.

The concentration of the chemical sample is determined by the level of response by the bacteria to the given sample. The set of bacteria selected for the bioassay each differ by only one property, which causes them to be more sensitive than other set members because of the mechanism of toxicity. Therefore, a "fingerprint", or profile unique to each chemical is generated from the differential response of the member bacteria. By "scoring" these fingerprints, the bioassay provides data for identifying test chemicals, assessing the concentration of test chemicals, and assessing whether the test chemical is acutely or chronically toxic. Another embodiment of the invention provides a library 9 wherein are stored the DLS patterns and calculated response indices for the bioassay of identified samples. By comparing the sample DLS with the stored pattern, the sample may be screened for toxicity, identified, and/or quantified.

The time required for a complete assay is approximately sixty (60) to sixty-six (66) minutes, the time required for untreated bacteria to divide twice in the absence of any chemical sample. However, individual samples may be analyzed at a rate of one sample per minute to determine the concentrations of a chemical in an aqueous sample. The Bacteria: The bacteria used in the bioassay are of a single species, Bacillus subtilis, which has numerous advantages over other bacteria traditionally used in bioassays. First, Bacillus subtilis is the most widely studied and genetically mapped gram-positive microorganism. Henner, D.J. and Hoch, J.A., The Bacillus subtilis Chromosome, Microbiol. Rev. 44:57-82 (1980).

The morphology, biochemistry, genetics and physiology of these organisms are therefore thoroughly understood. The morphological features of these bacteria are also easily adapted to the requirements of sensitivity and reproducibility of the bioassay. In its dormant form, Bacillus subtilis exists as a spore that is capable of surviving genetically unchanged while resisting heat, drying or other adverse effects for an indefinite period. As a vegetative cell, the bacteria can divide rapidly to double itself in approximately twenty-six (26) minutes during the logarithmic growth phase, and is much more sensitive at this stage to chemical toxicants than are the gram-negative microorganisms, Salmonella typhimurium and Escherichia coli. In its competent state, Bacillus subtilis readily takes up large molecules such as DNA, and can integrate genes derived from members of the same species or other species through traditional recombinant DNA technology.

The bioassay uses Bacillus subtilis mutants that are genetically identical, or isogenic, except for one or more genetic blocks in unique enzymatic repair processes or steps that restore chemically damaged DNA to its functional condition. Among the isogenic set of Bacillus subtilis are specifically selected strains deficient in different recombination (Rec-), excision (exc-), polymerase (Pol-) or spore repair (Spp-) repair steps. These strains may be deficient in one or a multiple of repair steps, and include the Rec- strains, recA1, recA8, recB2, recC5, recD3, recE4, recG13, mc-1, and m45; the Pol- strains T-1, TKJ8201; the Exc- strains her-9 and TKJ8206; The Exc- and Rec- strain FH2006-7; the Exc-, Rec- and Pol- strain HJ15; the Exc- and Spp- strain TKJ5211; the Exc-, Pol- and Spp- strain TKJ6321; and repair efficient strains HA101, 168 and 168 wild type. The mutants in this group are therefore more sensitive to the metabolic toxicity or genotoxic activity or a broad range of chemical substances at nannomolar or nannogram concentrations.

Since there are many repair processes and many enzymatic steps in these processes, a given chemical will affect only certain of the isogenic mutants, producing a unique DLS pattern. These Bacillus subtilis mutants were developed by treating them with by irradiation or with various chemical mutagens. For example, nitrosoquanidine, a powerful mutagenic chemical, was used to change the genetic makeup of the bacteria and trimethoprim to select for mutant types that could not synthesize DNA unless thymine was provided to them. The mutants from one group isolated by Felkner were designated FH2001 through FH2006 and had been derived from a parent strain JB01-200 whose lineage is defined in Felkner J. Bacteriol. 104:1030-1032 (1970). As discussed in this reference, conventional genetic analysis determined where the unique thymine gene defect was located in the chromosome of this bacteria. Subsequently, a mutant that was unusually sensitive to ultraviolet light and to the carcinogenic chemical 4- nitroquinoline-1-oxide was isolated by using trimethoprim treatment. Felkner, I.C., Isolation and Characterization of Thymineless Mutants from Ultraviolet-Sensitive Bacillus subtilis Strains, J. Bacteriol. 104:1030-1032 (1970). This strain, designated FH2006-7 was found to have a gene mutation genetically linked to the thymine locus that cause the mutant to be unable to carry out genetic recombination. The genes from this mutant were transferred to a parental "wild type" strain by DNA isolation, uptake and integration so that the FH2006-7 descendants with the following genotypes were produced:

(1) trp-, thy-, her-, rec-

(2) trp+, thy-, her-, rec-

(3) trp+, thy+, her-, rec- Types 2 and 3 are mutants suitable for assaying her- and rec- mutations to detect DNA damage. These bacteria, constructed by conventional genetic engineering techniques would not occur in nature or result in a surviving bacterial isolate. Among the nineteen strains used with this bioassay, recA8, recA1, recB2, recC5, recD3, recE4, recG13, hcr-9, mc-1 and FH2006-7 are descended from strain 168. This strain is a descendent from the strain trp C of John Spizizen. Spizizen, J. Transformation of biochemically deficient strains of Bacillus subtilis by deoxyribonucleate, Proc. Nat. Acad. Sci. U.S.A. 44:1072-1078. The remainder of the mutants are descended from this strain but are of direct lineage of HA 101 described in Kada, T., Sadaie, Y. and Tutikawa, K., Mutat. Res. 16:165 (1972) and Tanooka, H., Mutat. Res. 42:1 (1977). The mutant strains employed in the preferred embodiment of the bioassay may be obtained through application of conventional genetic engineering and transformation techniques by following the teachings provided in the following, and the references cited therein: Henner, D.J. and Hoch, J.A., The Bacillus subtilis Chromosome, Microbiol. Rev. 44:57-82 (1980); Laumbach, A.D., Felkner, I.C., Formations of a 4-nitroquinoline-1-oxide complex in normal and repair deficient strains of Bacillus subtilis, Mutat. Res. 1J5: 233-245 and 16:444 (1972); Felkner, I.C., Isolation and Characterization of Thymineless Mutants from Ultraviolet-Sensitive Bacillus subtilis Strains, J. Bacteriol. 104:1030 (1970); Felkner, I.C., Kadlubar, F., Parallel Between Ultraviolet Light and 4-Nitroquinoline-1-Oxide Sensitivity in Bacillus subtilis, J. Bacteriol. 96:1448 (1968); and I.C. Felkner, Microbial Testers: Probing Carcinogenesis (1981).

Isosets, defined as a subset of these mutants, constitute the minimum number required to define a response to a sample tested. Data derived form fractional survival tests and data derived from spot assays are used to determine which B. subtilis strains should be selected to comprise an isoset to identify a given compound by the radiation bioassay.

To select isoset members to be used in the bioassay, the isogenic mutants are pre-screened using the spot test with given chemical samples. The DNA-damaging test is performed according to the procedure disclosed in Felkner, I.C., Microbial Testers: Probing Carcinogenesis (1981) and Felkner, I.C. in Medine, A., and Anderson, W. , eds. Proceedings of the National Conference on Environmental Engineering, Am. Soc. Civil Eng., N.Y. pp. 204-209 (1983). Each mutant is inoculated into 5 ml of Brain Heart Infusion (BHI) broth, either from a disc containing approximately 1 × 10⁷ washed spores or a 1 mm loopful from a sporulating slant culture, and incubated by shaking at 37 to 39° C for 16 hours. This culture is designated o/n. Using a 1 mm inoculating loop, inocula from the o/n cultures are streaked radially on a nutrient agar plate to a sensitivity disc containing 10 ul of the assay chemical which may or may not have been pre-incubated with a rat liver microsomal fraction (S-9) in order to achieve metabolic activation. After incubation at 37 to 39° C, the distance (mm) of growth inhibition from the periphery is measured with a vernier caliper (Manostat). If varying concentrations of the assay substance are used, a concentration-dependent growth inhibition curve can be constructed, using the data from responsive mutants. Certain chemicals are not readily uptaken by the assay bacteria without metabolic activation. Liver microsomes, designated an S-9 fraction, were prepared from Sprague-Dawley rats induced with Aroclor 1254 (Litton Bionetics, Charleston, SC 19405). The S-9 fraction was prepared as a KCl homogenate with a protein concentration of 25-28 mg/ml. It is understood that this fraction may be microsomal or post-microsomal and is not limited by the origin of the tissues. The "Cocktail":

Many promutagens/carcinogens are water insoluble. A "cocktail" was therefore derived so that a clear aqueous solution of the test chemical could be mixed with the assay bacteria and the hydrophilic S-9 fraction, thereby minimizing interference with the radiation source and its ability to generate a DLS pattern of the radiation scattered by said bacteria. The "cocktail" has the advantage of solubilizing chemicals in an aqueous system that were not as readily solubilized by other known procedures using dimethylsulfoxide (DMSO), ethanol or acetone before being introduced into the aqueous assay mixture. Moreover, the "cocktail" is nontoxic to the assay bacteria and permits the S-9 fraction enzymes to metabolize the test chemical. The preferred embodiment of the bioassay employs a dispersant sold under the trademark "COREXIT 7664 Oil Dispersant" (Exxon Chemicals, Clark, New Jersey) and a non-ionic surfactant sold under the trademark "EMULPHOR EL-620" (GAF Corp., 140 West 51 Street, New York, N.Y. 10020). Both the "COREXIT" dispersant and the "EMULPHOR" surfactant are readily obtainable at the present time, and are generally known to chemists and commercial users. The chemical compositions of each of these are maintained as trade secrets by the respective owner companies.

The "EMULPHOR" surfactant is a polyoxylated vegetable oil.

At 25° C it is a clear, light brown liquid, having a specific gravity at this temperature of 1.04 to 1.05 and a viscosity of 600 to 1,000 cps. At 25° C, it is readily soluble in water, and its neutralization number is 0.5 maximum. As discussed, this composition was chosen because it is nontoxic to the Bacillus subtilis bacteria and when mixed with the chemical sample to be tested and the bacteria and the hydrophilic S-9 fraction should metabolic activation be required, a clear suspension results that does not interfere with the ability of the laser radiation to penetrate or the intensity of the radiation to scatter. "COREXIT 7664" is a surfactant ester having a light amber color and an alcoholic odor. AT 15.6° C it has a specific gravity of 1.03, a density of 8.59, and a viscosity of 29. It is soluble in fresh water and seawater and is dispersible in hydrocarbons. As discussed, this compound was selected as a component of the "cocktail" because it is nontoxic to bacteria and is capable of mixing with aqueous suspensions of bacteria without clouding and thereby does not interfere with the ability of laser radiation to penetrate the suspension. The Radiation Source:

The radiation source 1 of the bioassay is a laser such as that disclosed in Wyatt, U.S. Patent No. 4,541,719. This instrument records scattering intensity at fifteen angular locations, permitting the relationships between refractive index, size and shape to be measured. Moreover the read head is anodized, sealed and attached directly to the laser head as well as to an integrated circuit board, providing a unitized optical bench and a grounded enclosure for the detector elements. The specifications of the laser employed in the preferred embodiment of the bioassay are as follows: Light source: 5 mW He-Ne laser, λ=632.8 nm. Plane polarized

Angles Monitored Simultaneously: 15 @ 0.2 ≤ sin 0/2 ≤ 0.9 in steps of 0.05

Computer interface: Parallel analog outputs interfaced to 16-channel, 12 bit A/D programmable multiplexer.

Data Collection: 12,500 conversions/second - 25,000 conversions/second

Single Particle Measurements: Concentration ≤ 5 × 10

3/ml Size (depending on refractive index) 0.6 to 50 um

Detectors: Transimpedance photodiodes with built-in amplifiers Detector Output Dynamic Range: 0 - 10 V Dynamic Range: 10⁴ Linearity: ± 0.1%

Dark Current Noise: ± 0.2 mV

It is to be understood that this laser apparatus comprises part of the preferred embodiment of the bioassay, but that any laser instrument capable of the objects of this invention is comprehended. Radiation Bioassay Examples: Examples are provided for the bioassay of 4-nitroquinoline-N-oxide ("4NQO"), N-methyl-N'-nitro-N-Nitroso-guanidine ("MNNG"), benzo(a)pyrene ("B(a)P"), and "Lindane" (1,2,3,4,5, 6-hexachlorocyclohexane). The chemicals 4NQO and MNNG are well established as mutagens and DNA damaging agents to bacteria, and are water-soluble and direct-acting genotoxic chemicals that elicit carcinogenic activity. The chemical B(a)P is a well known procarcinogen/mutagen that is insoluble in an aqueous system, and requires metabolic activation by rat liver microsomal fractions (S-9) to become electrophilic and to cause DNA damage or a mutagenic response. "Lindane" is a chemical difficult to assay for its genotoxic potential by state of the art mutagenic assays.

To perform the bioassay according to the preferred embodiment, the following protocol is followed:

1. Lyphilized tablets containing the appropriate Bacillus subtilis strains and dehydrated growth medium are introduced into cuvettes containing water. Suspensions of bacterial cells at the required concentration described in the examples provided infra may also be used.

2. The inoculated cuvettes are placed in an incubation chamber that has been prewarmed to approximately 37 to 39° C, and incubation continued for about one hour to one and a half hours to establish a culture in the "assay-ready condition" .

3. Test or control material is introduced into the cuvettes using one pair of cuvettes for each assay organism. A negative control having no test material, and a positive control having an identified toxic chemical at a concentration whose response has been determined is included.

4. The cuvettes are incubated for six minutes at 37-39° C, and each cuvette is read a minimum of fifty

(50) times (the total reading time per sample is 30 seconds).

5. Incubation at 37-39 degrees Centigrade is continued for sixty (60) to seventy (70) minutes and each cuvette is again read a minimum of fifty (50) times.

6. The samples are scored, and the bacterial responses evaluated.

* The bacteria could be lyphilized vegetative cells or spores in a capsule form. However, when spores are used, this form being preferred because of their stability, they must be heat-shocked at 60 degrees Centigrade for 8 to 10 minutes before incubation. An alternative protocol follows: 1. Solubilize a sample in an aqueous solution. For water insoluble samples, use a sufficient amount of "cocktail" to prevent precipitation of the sample.

2. Make serial dilutions from the solubilized sample. 3. Add the capsule containing the bacterial spores and the dehydrated growth medium to a cuvette containing water.

4. Add a measured amount of each solubilized sample dilution to cuvettes containing bacteria and place the cuvette in an incubator at 37-39 degrees

Centigrade. Include a negative control having no test material, and a positive control having a toxic chemical at a concentration whose response has been determined.

5. Read the samples at 0, 6 and 66 minutes to follow the relative growth indicated by the increase in number of particles for each dilution.

6. Score each sample dilution and the controls.

7. Compare the score data to the score data of known compounds to determine a matching response profile.

8. Determine the concentration from a known dose response curve. If the sample is an unknown, search for a matching chemical response to determine its identity and concentration. It is understood that a negative control may also comprise a control sample of bacteria exposed to 4-nitroquinoline-N-oxide or 4-nitroquinoline-1-oxide and a positive control of bacteria exposed to benzo(a)pyrene. Cultivation of bacterial cells appropriate in concentration for the bioassay includes heat shocking strains of Bacillus subtilis in 1 ml of Brain Heart Infusion (BHI) broth for 20 minutes. Additional broth is added until an absorbance of 0.2 A(540) is achieved. The cultures are incubated at 37 to 39° C until the bacterial growth reaches an A( 540 ) of 0.5 , a period of approximately 2 hours. Cultures that require a longer period to reach this density are not considered suitable for the bioassay. Cultures found suitable are diluted to an A(540) of 0.3 with BHI broth yielding approximately 2 × 10⁷ bacteria/ml. These cells, when diluted by the amount required for the cuvette assay, yield a cell concentration of 10⁶ bacteria/ml.

1. Bioassay without metabolic activation (Cuvette Assay) - Compounds and bacterial isoset:

4NQO was assayed at final concentrations of 8.5, 4.2 and 2.1 ug/ml using Bacillus subtilis strains recE4 (6), 168 or 168 wild type (11). The DLS patterns for each of these assays comprise Figures 2a through 2d.

MNNG was assayed at a final concentration of 3.6 ug/ml with strains 6 and 11. The DLS patterns for each of these assays comprise Figures 3a and b.

Lindane was assayed at final concentrations of 53.6 ug/ml and 107 ug/ml utilizing Bacillus subtilis strains FH2006-7 (10) and 11. The DLS patterns for each of these assays comprise Figures 5a through 5d. Method:

In a screw cap tube containing 13.6 ml of deionized water at 37 to 39° C is pipetted 0.3 ml of bacteria and 0.1 ml of the test chemical. The tube is inverted gently and its contents poured into a cuvette. The cuvette is illuminated by the laser and the intensities of scattered light measured and scored.

The cuvette is then incubated at 37-39 degrees

Centigrade for approximately one (1) hour and illuminated identically. A negative control of bacteria in deionized water is also concurrently assayed to provide a baseline value for each strain.

The method may also be performed using lyphilized bacteria having with it sufficient dehydrated BHI to support the growth during the assay period. The total volume of fluid required to obtain a reading from the cuvette is 10 ml. Therefore, a volume of about 0.4 ml of fluid will rehydrate the cell-medium. The incubation period remains unchanged, and illuminating is performed at fifteen minute intervals.

2. Bioassay with Metabolic Activation (Preincubation) - Compound and bacterial isoset:

B(a)P was assayed at final concentrations of 15.1 and 7.6 ug/ml using Bacillus subtilis strains 11 and TKJ6321 (19). Method: Into a large capped test tube were pipetted 0.5 ml each of "COREXIT" and "EMULPHOR", B(a)P at the appropriate concentrations, 0.5 ml of S-9 fraction and 0.3 ml bacteria. The volume was adjusted to 14 ml with deionized water pre-warmed to 37° C. A control tube for each strain containing "COREXIT", "EMULPHOR", S-9 fraction and deionized water was prepared according to the same proportions as the sample to be tested. The tubes were incubated in a shaking water bath for approximately one (1) hour. After incubation, the contents were poured into cuvettes and the scattered intensities measured and scored, these readings being taken at zero and sixty minutes, as previously described. When lyphilized bacteria and growth media are used, a smaller volume of said bacteria are used to obtain the desired initial density in one (1) hour before treating the bacteria with the test chemical. The "COREXIT", "EMULPHOR" and S-9 volumes are each approximately 180 ul per 5 ml of sample.

3. Bioassay with Metabolic Activation (Cuvette Incubation) - Compound and bacterial isoset:

B(a)P was assayed at final concentrations of 15.1 and 7.6 ug/ml using Bacillus subtilis strains 11 and TKJ6321 (19). The DLS patterns generated by this assay comprise Figures 4a-4d. Method:

Into a scanning cuvette was pipetted a sufficient volume of Corexit to prevent precipitation of B(a)P when introduced into the aqueous incubation mix of bacteria dn S-9 fraction. The "CORERXIT"- B(a)P mixture was introduced into the cuvette along with 0.5 ml of S-9 fraction and 0.3 ml of bacteria in a total volume of 14 ml of deionized water pre-warmed to 37° C. A concurrent control tube for each strain was also prepared. The cuvettes were read and the scattered intensities scored at zero and sixty (60) minutes as previously described.

In an alternative method, the aqueous cocktail of

"COREXIT", "EMULPHOR" and water is prepared prior to addition of the S-9 fraction and the bacteria. Thus,

180 ul each of "EMULPHOR" and "COREXIT" are added to each 5 ml of deionized water and mixed. At the time of the assay, 180 ul of S-9 fraction is added along with appropriate amount of lyphilized bacteria and medium required to achieve a cell concentration of 10 6 bacteria/ml.

Response Evaluation:

The scoring techniques of the bioassay measure the height of the DLS pattern as a function of bacterial count and the amount of shift in the peaks of the chemical sample pattern and the control sample pattern to determine the degree of morphological change in size and shape of the bacteria. The derived indices provide an immediate preliminary indication of the toxic response, the DNA damaging (genotoxic) potential, and identify isosets for compound identification.

The measurement of each pattern's height provided an indication of bacterial count and is therefore the principal indicator of toxicity. In developing the scoring techniques for the identification of morphological changes, it was noted that there is a function which is continuous on a closed interval from the beginning of the pattern to its end. Each line segment between any two adjacent points formed an arc, and associated with each point on the X-axis is only one point on the pattern. Using a digitizing tablet, it is possible to obtain a sufficiently large number of points so that the sum of the distances between all adjacent points would approximate the length of the pattern. This was defined as the arc length of the curve. Variances of the arc length correspond with the variances in the bacteria morphology, especially as they related to individual cell size. In order to simplify the interpretation of a response to a given treatment by a chemical at a given concentration, a system using three (3) digits was developed, each of which was the indicator for a response. The nature of this response was assigned a value of one (1) through four (4) according to the height and direction of pattern displacement as a function of the type of effects on the bacterial population. To index a set of treatments, the count changes and patterns of a baseline control sample were measured. The subsequent treatment and control samples were compared to the baseline response.

Index calculations:

Four patterns were digitized in the following order: Control at time 0 and 60 minutes; treated sample at time 0 and 60 minutes. For each pattern in order, the response indices were determined by calculating the distance between adjacent points on the curve as the arc length. The mean Y of all coordinate pairs was calculated for the bacterial count index. The control at time "0" was designated as the baseline control. For each of the other curves, the percent change between the pattern's mean Y and the baseline pattern's mean Y were calculated. For the Fourth order Y index, the Y percent change was divided by 5 and, if the change was positive, 1 was added. If the change was negative, 1 was subtracted. The final integer value was retained. The percent change between the test pattern's arc length and the baseline pattern's arc length was calculated. To determine the fourth order arc index, the arc percent change was divided by 5 and, if the change was positive, 1 was added. If the change was negative, 2 was subtracted. The integer value was retained.

Each of the three digits, from left to right in the PRI index, derives from the following table according to the fourth order indices of the patterns, starting with the control at "0" time as the reference and the control at 60 minutes as the first digit.

The biological interpretations that can be made on a digit-by-digit basis are as follows:

"1" indicates a positive increase in bacterial count with a concurrent decrease in size of said bacteria (positive shift), and a non-toxic response.

It is normal for rapidly dividing cells to decrease in size.

"2" indicates a positive increase in bacterial count with a concurrent increase in size (negative shift), and a minimum toxic response.

"3" indicates a decrease in bacterial count with a concurrent increase in size (negative shift), and a toxic and lethal response.

"4" indicates a decrease in bacterial count with a concurrent decrease in size (positive shift), and a toxic and lethal response. The tertiary responses for each of the indexed patterns were obtained by multiplying the fourth order Y index by the fourth order arc index. The secondary response index (SRI) was determined by the calculation of the sum of the tertiary indices for all indexed patterns combined. This method of calculation is illustrated by the following data derived from an assay:

The interpretation of the PRI of 132 is that the control of sixty-six (66) minutes displayed normal growth, hence the number 1 as the first digit. The treated sample at time "0" showed a loss in bacterial count accompanied by size increase, hence the number 3 as the second digit. By sixty-six minutes, the treated sample showed an increase in bacterial count accompanied by a size increase, hence the number 2 in the third digit.

Two extreme examples of PRIs are 133 and 144. These examples are immediately toxic responses from which the cells fail to recover. The 133 PRI shows cell enlargement, and the 144 PRI shows cell shrinkage, both being associated with the failure of the cell to recover but by different mechanisms of action. The presence of a 3 or 4 in the first digit of any PRI should mean that the assay is invalid because the bacterial control culture is not displaying normal growth.

A screening score S provides a precise measure of bacterial growth or inhibition of bacteria in the test sample as compared to the control bacteria. It is defined in terms of DLS placement scores and was implemented by previous studies. Wyatt, P.J., Phillips, E.T., Scher, M.G., Kahn, M.R., and Allen, E.H., J. Argic. Food Chem. 25:1080 (1977); Wyatt, P.J., Scher, M.G., and Phillips, D.T., J. Agric. Food Chem. 25:1086 (1977). Thus,

10(Dbc/300) - 10(Dc/300) S = 300 log

10(Dbt/300) - 10(Dt/300) where Dbc is the DLS displacement of the assay broth plus bacteria relative to the assay broth (control), Dc is the displacement of the control blank relative to itself (0 when there is no background interference), Dbt is displacement of test material in assay broth plus bacteria relative to the control, and Dt is displacement of the test material in assay broth relative to the control blank. The equation converts the data logarithmic DLS patterns back to linear form so that the background is subtracted and hence converted to the usual logarithmic form. The equation yields a factor of 300 for 10-fold increases of control bacteria with respect to test bacteria.

Applying the index calculations to the samples assayed:

In all the figures, the untreated "0" time control is represented by a solid line the one hour untreated control by a dot broken line ( . . . . ) , the

"0" time treated sample by a widely spaced broken line ( - - - - ), and the one hour treated sample by a narrow spaced broken line (_ _ _ _ ).

4NQO DLS Toxicity Responses:

Figures 2a through 2d show the DLS response patterns for strains 6 (rec 34) and 11 (168 wild type) respectively. From the mean Y values of the "0" time and 1 hour samples, it is clear that increases in cell counts occurred in all untreated controls. From

Figures 2a and 2b, it is clear that 4.2 ug/ml and 2.1 ug/ml of 4NQO caused increased arc lengths and curves shifting to the left relative to the light scattering angles of the negative time "0" control. The relative intensity increased less from the "0" to 1 hour interval. From Figure 2c, it is clear that 8.5 ug/ml of 4NQO caused a decrease in arc length and a curve shift to the left relative to the light scattering angles of the negative time "0" control. However, at 4.2 ug/ml, the change in arc length was not substantial and the curve shift relative to light scattering angles was slightly toward the larger angles.

In applying the response index calculations, the following interpretations can be made. Strain 6 at 4.2 and 2.1 ug/ml had PRI values of 132 and 131, respectively. At the 4.2 ug/ml level, an immediate, toxic and lethal effect was seen at "0" time. The bacteria showed some recovery because the counts increased by 60 minutes. However, the bacteria remained enlarged. At the 2.1 ug/ml level, an immediate, toxic and lethal effect was seen at "0" time, but the counts increased and cells were of normal size at 60 minutes. The 4.2 ug/ml concentration, however, did not adversely affect strain 11. These data show that strain 11 was more refractile to the effects of 4NPO than was strain 6, even when strain 11 was treated at higher dose levels. Thus, the genotoxic manifestation is expressed at the primary response level.

MNNG DLS Toxicity Responses: Figures 3a and 3b show the DLS response patterns for strains 6 (rec E4) and 11 (168 wild type) respectively. From the mean Y values of the "0" time and 1 hour samples, it is clear that increases in cell counts occurred in all untreated controls. From Figure 3a, it is clear that 3.6 ug/ml of MNNG caused a decreased arc length and a curve shift to the left, relative to the light scattering angles of the negative time "0" control. The relative intensity increased less than did the negative control from the "0" to 1 hour interval. from Figure 3b, it is clear that 3.6 ug/ml of MNNG did not cause a decrease in arc length, although the arc length increase was slightly less than the negative control. The curve shift relative to light scattering angles was toward the larger angles, almost approximating the negative time "0" control. Applying the calculations of the response indices, the following interpretations can be made. Strain 6 at 3.6 ug/ml had a PRI value of 132, indicating an immediate, toxic and lethal effect at "0" time. The bacteria showed some recovery because the counts increased by 60 minutes. However, the bacteria remained enlarged. Strain 11 at the concentration of 3.6 ug/ml had a PRI value of 111, thus showing no compound-related toxicity expressed either immediately or over the 60-minute period following treatment. This is a genotoxic response expressed at the primary response level. Additionally, the SRI value for strains 6 and 11 were -3 and 19, respectively. The negative secondary response of strain 6, when compared to the positive SRI of strain 11, indicates that there is a genotoxic effect which can be repaired by the wild type (strain 11) but not the Rec- mutant (strain 6). Lindane DLS Toxicity Responses:

Figures 5a through 5d show the DLS response patterns for strains 10 (FH2006-7) and 11 (168 wild type). From the mean Y values of the "0" time and one hour samples, it is clear that increases in cell counts occurred in all untreated controls. From Figures 5a and 5b, it is clear that 107 ug/ml and 53.6 ug/ml of

Lindane initially caused decreased arc lengths and curves shifting to the left relative to the light scattering angles of the negative time "0" control.

However, at the lower level, a positive shift was seen at sixty minutes. The relative intensity did not change substantially from the "0" to one hour intervals at the higher level. However, an increase was seen at the lower concentration. From Figures 5c and 5d, it is clear that both concentrations of Lindane caused an initial decrease in arc length and a curve shift to the left relative to the light scattering angles of the negative time "0" control. However, the arc lengths increased and the scattering angle shifted to the right relative to the negative time "0" control. However, the arc lengths increased and the scattering angle shifted to the right relative to the negative time "0" control after one hour. Applying the response index calculations, the following interpretations can be made. Strain 10 at 107 and 5 3.6 ug/ml had PRI values of 222 and 131, respectively. At the 107 ug/ml level, a toxic and a sustained effect (cell enlargement) was seen. However, at the 53.6 ug/ml level, the bacteria recovered after 60 minutes. Strain 11 had PRI values of 131 at both concentrations. This indicated that at both levels, cell lethality and enlargement occurred but that full recovery resulting in growth and the return to normal size occurred by one hour. Although strain 10 may be slightly more affected than strain 11 at the higher concentration of Lindane, the differences in the PRI values do not appear to be useful in making toxicity assessments. The SRI values, however, were substantially different for strains 10 and 11. The SRI values for strain 10 at 107 ug/ml and 53.6 ug/ml were -19 and 7, respectively. The SRI value was 10 for strain 11 at both 107 ug/ml and 53.6 ug/ml. The dose-related SRI trend seen in strain 10 but not in strai 11 indicates a genotoxic effect of this chemical, as demonstrated by the negative SRI value at the higher dose level. As shown by the positive SRI values, strain 11 at both concentrations, was capable of repairing genetic damage and dividing normally. Strain 10 was capable of repairing genetic damage and dividing normally only at the lower concentration of 53.6 ug/ml. From these results, it is clear that primary cytotoxicity may be expressed by a three digit number. Using the primary response of mutant and wild type bacteria, some compounds such as MNNG and 4NPO which are highly toxic can also be identified as genotoxic. Compounds such as B(a)P when activated with an S-9 fraction cannot be shown to be genotoxic by using the primary response level alone. However, genotoxicity may be determined by calculating the secondary response. Other compounds such as Lindane, whose genotoxic responses have been difficult or even impossible to detect with other state of the art mutagenicity assays are readily detected by the radiation/microbe bioassay by calculating their secondary response indices.

Thus, it will be seen that all embodiments of the present invention provide a unique apparatus and method for radiation bioassay samples. While preferred embodiments of the invention have been disclosed, it should be understood that the spirit and scope of the invention is not to be limited solely by the appended claims, since numerous modifications of the disclosed embodiments will undoubtedly occur to those of skill in the art.

Claims

WHAT IS CLAIMED IS:

1. An apparatus for bioassaying samples by determining a morphological change in bacteria comprising: (a) means for generating scattered radiation from a morphologically changed assay bacteria subjected to treatment by a sample;

(b) means for detecting said scattered radiation and means for generating a pattern representative of the scattered radiation detected.

2. An apparatus according to claim 1, further including:

(c) reference means having at least one reference pattern generated when bacteria are exposed to an identified sample; and

(d) means for comparing said reference pattern to said generated pattern to identify and quantify said sample.

3. An apparatus according to claim 2, wherein said bacteria is selected form the group comprised of:

Bacillus Subtilis, an isogenic mutant strain of Bacillus Subtilis, Bacillus Subtilis strains selected from a group consisting of recA1, recA8, recB2, recC5, recD3, recE4, recG13, mc-1, m45, T-1, TKJ8201, her-9, TKJ8206, FH2006-7, HJ15, TKJ5211, TKJ6321, HS 101, 168, and 168 (wild type).

4. An apparatus according to claim 2, wherein said assay bacteria includes a subset of mutant strains, said subset being a minimum number required to define a response to said test chemical.

5. An apparatus according to claim 1, wherein said means for detecting comprises means for detecting the intensity of scattered radiation for a control sample and for a test sample at zero time and at least one selected time interval.

6. An apparatus according to claim 5, including means for plotting said pattern on an x and a y axis, the x axis being scattering angle and the y axis being relative scattered intensity; means for scoring said pattern by measuring the height of said pattern and by calculating shift of at least one pattern peak; said means for scoring includes means for generating x and y coordinates for said pattern wherein said x axis corresponds to a count of said bacteria and the y axis corresponds to morphological change in size and shape of said bacteria; means for determining an arc length of the curve by summing the distances between adjacent x coordinates; means for calculating a mean Y value for all coordinate pairs of said pattern; means for calculating a percent change between the mean Y value for a control pattern and the mean Y value for a sample pattern; means for determining a fourth order Y index by dividing said Y percent change by five, and for adding 1 when said change is positive, and for subtracting 1 when said change is negative; means for calculating a percent change between the arc length of said control pattern and the arc length of said sample pattern; means for determining a fourth order arc index by dividing said arc length percent change by 5, and for adding 1 when said change is positive, and for subtracting 1 when said change is negative; and means for generating a primary response index digit according to the following table:

wherein,

"1" indicates a positive increase in bacterial count with a concurrent decrease in size of said bacteria, and a non-toxic response;

"2" indicates a positive increase in bacterial count with a concurrent increase in size, and a minimum toxic response; "3" indicates a decrease in bacterial count with a concurrent increase in size, and a lethal toxic response;

"4" indicates a decrease in bacterial count with a concurrent decrease in size, and a lethal toxic response.

7. An apparatus according to claim 5, further comprising means for calculating a tertiary response index for said pattern by multiplying the fourth order Y index by the fourth order arc index.

8. An apparatus according to claim 5, further comprising means for calculating a secondary response index for said pattern by summing the tertiary indices for all the indexed patterns.

9. An apparatus according to claim 1, wherein said means for generating comprises a means for generating a negative control by plotting a pattern for a control sample of bacteria that is exposed to 4-nitroquinoline-N-oxide and means for generating a positive control by plotting a pattern for a control sample of bacteria that is exposed to benzo(a)pyrene.

10. An apparatus according to claim 1, further comprising means for metabolically activating said assay bacteria by treating said bacteria with a post-microsomal or a microsomal fraction.

11. An apparatus according to claim 1, wherein said sample is a water insoluble sample; and further comprises a surfactant means, and a dispersant means whereby said water insoluble sample in suspension is solubilized and substantial optical clarity of said suspension is maintained.

12. An apparatus according to claim 1, wherein said sample is solubilized in a polyoxyethylated vegetable oil such as "EMULPHOR EL-620" and a dispersant such as "COREXIT 7664" Oil Dispersant.

13. An apparatus according to claim 1, wherein said radiation generating means is a monochromatic laser.

14. A process for bioassaying a sample by determining a morphological change in bacteria comprising the steps of: (a) generating scattered radiation from a morphologically changed assay bacteria subjected to treatment by a sample; and

(b) detecting said scattered radiation and generating a pattern representative of the scattered radiation detected.

15. A process according to claim 14, further comprising the steps of providing reference means having at least one reference pattern generated from exposing bacteria to an identified sample; comparing said reference pattern to said generated pattern to assist in identifying and quantifying said sample.

16. A process according to claim 14, wherein said assay bacteria and said sample are prepared by performing the steps of: (a) solubilizing the sample;

(b) making serial dilutions of said solubilized sample;

(c) adding said assay bacteria to at least one clear optical cuvette; and

(d) providing a measured amount of each solubilized sample dilution to said cuvette containing said bacteria.

17. A process according to claim 14, wherein said sample and said bacteria are incubated at a temperature in the range of approximately 37 to 39° C.

18. A process according to claim 14, wherein said assay bacteria is selected from the group comprised of:

Bacillus Subtilis, an isogenic mutant strain of

Bacillus Subtilis, or Bacillus Subtilis strains selected from a group consisting of recA1, recA8, recB2, recC5, recD3, recE4, recG13, mc-1, m45, T-1, TKJ8201, hcr-9, TKJ8206, FH2006-7, HJ15, TKJ5211, TKJ6321, HA 101, 168, and 168 (wild type).

19. An apparatus according to claim 14, wherein said assay bacteria is a subset of mutant strains, said subset being a minimum number required to define a response to said test chemical.

20. A process according to claim 14, wherein said detecting step includes detecting the intensity of scattered radiation of a control sample and of a test sample at zero time and at least one selected time interval.

21. A process according to claim 14, wherein said generating a pattern step includes plotting said pattern on an x and a y axis, the x axis being scattering angle and the y axis being relative scattered intensity.

22. A process according to claim 15, further comprising the steps of: generating x and y coordinates for said pattern wherein said x axis corresponds to a count of said bacteria and the y axis corresponds to morphological change in size and shape of said bacteria; determining an arc length of said pattern by summing the distances between adjacent x coordinates; calculating a mean Y value for all coordinate pairs for said pattern; calculating a percentage change between the mean Y value for a reference pattern and said mean Y value for said pattern; determining a fourth order Y index by dividing said Y percent change by five, and adding 1 when said change is positive, and subtracting 1 when said change is negative; calculating a percent change between the arc length of said reference pattern and the arc length of said pattern; determining a fourth order arc index by dividing said arc length percent change by 5, and adding 1 when said change is positive, and subtracting 1 when said change is negative; and generating a primary response index digit according to the following table:

wherein, "1" indicates a positive increase in bacterial count with a concurrent decrease in size of said bacteria, and a non-toxic response;

"2" indicates a positive increase in bacterial count with a concurrent increase in size-, and a minimum toxic response;

"3" indicates a decrease in bacterial count with a concurrent increase in size, and a lethal toxic response;

"4" indicates a decrease in bacterial count with a concurrent decrease in size, and a lethal toxic response.

23. A process according to claim 22, further including the step of: calculating a tertiary response index for said pattern by multiplying the fourth order Y index by the fourth order arc index.

24. A process according to claim 22, further comprising the step of: calculating a secondary response index for said patten by summing the tertiary indices for all the indexed patterns.

25. A process according to claim 14, wherein said sample is water insoluble, and further comprising the steps of: solubilizing said water insoluble sample in suspension and maintaining substantial optical clarity of said suspension by using a surfactant means and a dispersant means.

26. A process according to claim 25, wherein said solubilizing step includes the use of a polyoxyethylated vegetable oil such as "EMULPHOR EL-620" and a dispersant means such as "COREXIT 7664" Oil Dispersant.

27. A process according to claim 14, wherein said radiation for said radiation generating step is provided by a monochromatic laser.

28. A process according to claim 14, further comprising the steps of: generating a negative control by plotting a curve for a control sample of bacteria that is exposed to 4-nitroquinoline-N-oxide; and generating a positive control by plotting said curve for a control sample of bacteria that is exposed to benzo(a)pyrene.

29. A process according to claim 14, including metabolically activating said assay bacteria by treating said bacteria with a microsomal or post microsomal fraction.

30. A plurality of Bacillus Subtilis bacteria, each being uniquely responsive by changes in morphology to chemical exposure, and means for recording and representing said changes.

31. A composition for solubilizing a chemical sample which comprises a surfactant and a dispersant, said composition being optically clear when mixed with assay bacteria and being harmless to said assay bacteria.

32. A process for providing an optically clear medium for use in a radiation bioassay comprising mixing a water insoluble sample with "COREXIT 7664" Oil Dispersant and "EMULPHOR EL-620" non-ionic surfactant.