Bioassay of genotoxic effects of environmental particles in a feeding ciliate.

The ciliated protozoan, Paramecium, can be used to quantitate cytotoxic and genotoxic effects of ingested complex environmental particles. Cytotoxicity is quantitated by the increased proportion of treated versus control cells which do not retain their capacity for normal cell replication. Genotoxic effects are assessed by the increased fraction of nonviable offspring from treated versus control parent cells after the self-fertilization process of autogamy. Since these cells ingest nonnutrient respirable-sized particles, biological activity of intracellular extraction of dusts and fly ash can be compared before and after extraction with polar and nonpolar solvents. Previous studies indicated that coal fly ash was mutagenic in these eukaryotic cells. Mutagenicity of coal fly ash was not detectable after extraction with a concentration of HCl known to remove nonmatrix trace elements. These results suggested that this ciliate bioassay might be a detector of mineral mutagens. Fine particles of the carcinogenic nickel compounds, alpha-nickel subsulfide, and beta-nickel sulfide were compared for their biological activity in this bioassay. Both nickel compounds were ingested by the ciliates and induced heritable damage in the progeny of the treated parent cells.

In P. tetraurelia, there are two kinds of nuclei, the "germ line" micronucleus and the "somatic" macronucleus. The micronucleus is normally silent (22), and any lethal dominant or recessive damage present in the micronucleus will be expressed only when the zygote micronucleus differentiates a new macronucleus for the progeny cell after fertilization. Dominant lethal mutations can be expressed as early as the first cell division after autogamy, while recessive lethals can show phenotypic lag. The pres-*Zoology and Physiology Department, University of Wyoming, Laramie, WY 82071.
tBattelle Columbus Laboratories, Columbus, OH 43201. ence of dominant wild-type genes from the old macronucleus and their products must be degenerated and/or diluted by successive cell generations prior to expression of recessive traits (23). Thus, induction of dominant versus recessive mutations can be monitored by examining the timing of the expression of the damage after autogamy. The fraction of dead or slow growing progeny from treated parents reflects the presence of lethal and detrimental mutations in the parental "germ line" micronuclei (24)(25)(26)(27)(28)(29)(30)(31)(32).
Cytotoxicity versus genotoxic effects can be easily distinguished and monitored. In general, the bioassay involves treatment of parent cells with the test agent, and if the agent is cytotoxic, the cells will be unable to successfully complete division. Genotoxic effects are assayed by determination of the damage expressed in the self-fertilization offspring of parent cells which have undergone a minimum of seven to ten cell divisions in an agent-free medium to deplete the food supply which induces autogamy. Thus, for detection, induced nuclear damage must be heritable and sufficiently nontoxic to allow successive cell division and successful completion of self-fertilization to result in expression of any induced "germ line" micronuclear damage.
We have taken advantage of the established mutagenesis assay in P. tetraurelia as a prescreen for environmental hazards (33,34) which can be inhaled by man (35,36). Upon inhalation, particles show a size-dependent deposition in the lung where phagocytosis and particle surface interactions occur (37)(38)(39). The ciliated protozoan, Paramecium tetraureliz, engulfs respirable-sized particles (40) and undergoes the self-fertilization process of autogamy resulting in homozygosity. Thus induced damage to the parental genome by exposure to test agents is expressed in the next generation of progeny cells (22,41).
In the present study, we have capitalized on the fact that protozoans ingest nonnutritive particles (40,42) of a size respired by man. Intracellular extraction of these airborne particles more closely mimics the phagocytosis seen in macrophages of lungs of mice chronically exposed to fly ash particles (37) than does the Ames assay or viral bioassays which require extraction prior to mutagenicity testing. The two finest and most respirable coal fly ash fractions collected from the smoke-stack of a power plant were more mutagenic in the Ames Salmonella assay than two coarser fractions (43). The acute toxicity of a single intratracheal instillation of Ni1S, particles was 12 times as toxic for fine (VMD = 1.8 ,um) versus coarse (13.3 ,um) particles (44). Thus particle size was found to be a significant factor for biological activity for both coal fly ash and Ni3S2.
Our previous studies with paramecia indicated that coal fly ash was mutagenic before and after heat treatment. Extraction of particles with aqueous acid, but not with the organic solvent dimethyl sulfoxide, removed detectable mutagenic activity. The positive mutagenicity of the heat-treated fly ash as well as the decreased mutagenic activity with acid extraction found in the protozoan bioassay (13., 36), but not in the Ames assay (45,46), suggested the possible increased sensitivity of the protozoan bioassay to mineral mutagens. Therefore, the sensitivity of paramecia to the known metal carcinogen nickel subsulfide (a-Ni3S2) and the crystalline nickel monosulfide (P-NiS) (18,19) was investigated in the present study.

Stock and Culture Conditions
Strains of P. tetraurelia stock 51 were used. Strain A was derived by crosses of stock 51 with stock 169; strain B was derived by crosses of stock 51 with stock 242. Both derived stocks were kindly supplied by Dr. D. Nyberg and have been described previously by Nyberg (47).
The culture medium was inoculated 24 hr before use with Klebsiella aerogenes and adjusted to pH 6.8 prior to addition of the ciliates. Cells of known age (35,36) were used as a source of parent cells for the bioassay.

Preparation and Use of Sized Nickel Compounds
The compounds were kindly prepared by D. Mc-Neill, Battelle Columbus Laboratories. The dusts were prepared by grinding repeatedly in a carbon steel grinding chamber of a Spex Model 8000 mixer mill. Dusts were then suspended in the enthanol settling media and settled in an Andraessen pipet for the required time to remove particles greater than the desired size. The desired fine particles had a volume median diameter (VMD) of 1.82 ,m with associated geometric standard deviation of approximately 1.55. The fine particles were gently drawn off, centrifuged to remove excess solvent and then vacuum dried. Particle size and morphology were evaluated by light microscopy. X-ray diffraction studies showed no contaminants or species other than nickel subsulfide before and after size separation by sedimentation (McNeill, personal communication). Nickel monosulfide was purchased from Alpha Products and was ground and sized using the same procedures as those employed for nickel subsulfide. X-ray diffraction analysis demonstrated a mixed crystalline structure of nickel monosulfide with the major phase present as P-NiS.
Although it was the intent of the authors to also utilize the noncarcinogenic amorphous nickel monosulfide in this investigation, some confusion arose regarding the availability of this compound. Subsequent discussions with Drs. Warner and Timberg (personal communication) of INCO Limited indicated that their attempts to synthesize this compound using previously cited techniques resulted in formation of a product with the sulfur to nickel ratio greater than unity and with a total Ni + S content of less than 100%. They have concluded that X-ray diffraction alone is not sufficient to characterize the product precipitated from a solution of nickel chloride with ammonium sulfide. Thus it appears that results of previous investigations using amorphous nickel sulfide should be interpreted with care and that future studies should provide analysis of both the elemental composition and the crystallinity.

Cytotoxicity Assay
For biological testing, the particles were suspended in sterile distilled water and sonicated for 15 min to deagglomerate all particles. Suspensions were made 1-2 hr before use to minimize solubility of both nickel monosulfide and nickel subsulfide.
The particles were then diluted in bacterized culture medium before use in the cytotoxicity and genotoxicity bioassays.
Twenty cells were exposed to various dilutions of particles for 2 hr at 340C and/or 1 day at 270C. The proportion of viable cells was estimated by the fraction of 20 cells still motile after incubation with the dust suspensions.

Genotoxicity Assay
The main steps are: (1) treatment of synchronized cells; (2) isolation of single treated cells and growth of each into a clone; (3) induction of autogamy in treated clone; (4) isolation of 16 cells in autogamy, i.e., 16 randomly chosen haploid chromosomal sets from each clone; and (5) determination of how many of these isolates die or grow poorly.
The proportion of autogamous isolates which are nonvigorous or nonviable are presumed to represent an index of the deterimental or lethal mutations in parent cells which are heritable and therefore expressed in their progeny. Self-fertilization results in homozygosity of all genetic loci and therefore expression of induced and/or spontaneous genetic damage.
Synchronized cells were obtained by removing cells with the typical morphology of dividers using a micropipet under a dissecting microscope from a mass culture of exponentially growing cells of known clonal age. The dividers were then incubated at 270C for 2 hr to provide cells which were at a stage refractory to DNA repair (29). These parent cells were then exposed to bacterized culture medium for 2 hr at 340C, including (1) medium only; (2) sterile washed glass beads, 1-3 pm diameter as a negative control for nonnutritive particles; (3) certain dilutions of the nickel compounds; or (4) the positive control mutagen, 4-nitroquinoline-N-oxide. After treatment, the cells were washed by three repeated transfers of cells to agent-free medium with a micropipet under a dissecting microscope. The washed cells were placed as single isolates into fresh media, allowed to multiply and starved to induce the self-fertilization process of autogamy. Samples of cells were stained with acridine orange (48), and when 90-100% of the cells showed the nuclear changes typical of autogamy, 16 cells were taken as a source of progeny cells and placed as single isolates into fresh food. Normally 20 parent cells were used for each experimental group to provide 320 progeny (20 x 16). The cells were scored as follows: viable, cells which have cleared the medium of bacteria; detrimentals or slow growers, cells which have not cleared the medium in 3 days; and lethals, no surviving or a few moribund cells. The statistical analysis for determination of significance involved placing confidence intervals on the difference between any two proportions in pairwise comparisons using the normal approximation (49).
If the confidence interval did not capture zero, then the pair compared were deemed significantly different. The variance of that difference was computed on the basis of stratified random sample (50). The detailed procedure has been described previously (35,36).
Experimental results are presented only for those studies in which the negative control did not differ significantly from mass culture media, since contaminated clones can sometimes yield false positive results.

Results
Light microscopic observation revealed that both nickel compounds were ingested by paramecia into aThe percentage of 20 isolated cells still alive after 2 hr and 1 day is given; clones A and B were 20-26 fissions old. bCells were seen to aggregate around only these nickel particles at this concentration.
eGrowth inhibition was observed at this concentration. tSignificantly different from glass beads, but not from media only.
bDied indicates cytotoxicity to parent.
food vacuoles, and these vacuoles appeared swollen relative to those vacuoles which appeared to contain only bacteria.
Cytotoxicity was generally not observed at 54 l.gl mL using NiS or Ni3S2 unless exposed overnight (Table 1). However, older clones (61 fissions old) of strain B died at this concentration and at a 10-fold dilution of nickel subsulfide after a 2 hr exposure. The age or differences in strain sensitivity to nickel compounds may contribute to the observed cytotoxicity since strain B was more sensitive than strain A after 2 hr exposures to the nickel compounds (Tables 1 and 2). In four of six experiments, Ni3S2 showed significantly (p < 0.01) higher lethality and detrimentals compared to the media only or negative glass beads controls ( Table 2). In the one negative experiment, the baseline mutagenesis was high (12%), and the most genotoxic concentration (0.5 pg/mL) was not tested. Nickel monosulfide yielded progeny with increased lethality and detrimentals relative to controls in three of five experiments. Both nickel compounds show a decrease in biological effect with increase in concentration (Table 3).

Discussion
Both the nickel subsulfide and the crystalline nickel monosulfide dusts significantly increased lethality among the self-fertilization offspring from treated versus control parents. The monosulfide exhibited more variability and less severity than the subsulfide with respect to loss of progeny viability and vigor. No difference in the ability of the paramecia to ingest the two nickel dusts was observed, although quantitative studies were not carried out.  Table 2 with the exception of experiment 3. The exclusion was made because this clone contained high baseline mutagenesis not seen in the other series.
bThe positive control used was the direct-acting mutagen, 4nitroquinoline-N-oxide, derived from cumulative experiments. In some experiments, the carcinogen was toxic to the parent cells used and no value could be obtained.
*These values are significantly different (p < 0.01) from the negative coiitro1 glass-bead& and riedia only.
The decrease in genotoxicity with increased dose of nickel dusts was not due solely to elimination from the population of damaged cells by death since lethality and detrimentals are derived from offspring of viable parents. Some damaged parent cells could be eliminated from the experimental pool by cytotoxicity.
The inverse dose-response relationship may reflect nickel-induced changes in the physiology of the cell which promotes protection of the cell from heri-Expt. 5 Expt. 6 208 table damage. Any reduction in the rate of uptake and/or reduction in growth rate could decrease damage and promote repair before DNA replication. The Paramecium bioassay did not show a striking difference in the biological activity of the a-Ni3S2 and f3-NiS, in agreement with Sunderman and Hopfer (14), who indicated that both are metal carcinogens in rats. Similarly Costa et al. (51) reported that crystalline nickel subsulfide and nickel monosulfide described herein also produce neoplastic transformations in Syrian hamster embryo cultures.
There is also a correlation between induction of heritable damage in paramecia and nickel-compound induced damage in plant cells. Abnormalities have been observed in root cells of the broad bean induced by nickel compounds (52)(53)(54).
In summary, with respect to the carcinogenicity of these two nickel compounds in experimental animals, this bioassay demonstrated a correlation between the eukaryotic cell mutagenicity and previously published reports on the animal carcinogenicity or the two agents.