Transformation of BALB/c-3T3 cells: V. Transformation responses of 168 chemicals compared with mutagenicity in Salmonella and carcinogenicity in rodent bioassays.

This report describes the activities of 168 chemicals tested in a standard transformation assay using A-31-1-13 BALB/c-3T3 cells. The data set includes 84 carcinogens, 77 noncarcinogens, and 7 research chemicals. Carcinogens included 49 mutagens and 35 nonmutagens; noncarcinogens included 24 mutagens and 53 nonmutagens. The transformation assay did not use an exogenous activation system, thus, all chemical responses depended on the inherent target cell metabolic capacity where metabolic activation was required. The upper dose limit was 100 milli-osmolar because the assay could not discriminate active and inactive chemicals tested above this concentration. Certain physicochemical properties resulted in technical problems that affected chemical biological activity. For example, chemicals that reacted with plastic were usually nonmutagenic carcinogens. Similarly, chemicals that were insoluble in medium, or bound metals, were usually nonmutagenic and nontransforming. Multifactorial data analyses revealed that the transformation assay discriminated between nonmutagenic carcinogens and noncarcinogens; it detected 64% of the carcinogens and only 26% of the noncarcinogens. In contrast, the transformation assay detected most mutagenic chemicals, including 94% of the mutagenic carcinogens and 70% of the mutagenic noncarcinogens. Thus, transformation or Salmonella typuimurium mutagenicity assays could not discriminate mutagenic carcinogens from mutagenic noncarcinogens. Data analyses also revealed that mutagenic chemicals were more cytotoxic than nonmutagenic chemicals; 88% of the mutagens had an LD50 < 5 mM, whereas half of the nonmutagens had an LD50 > 5 mM. Binary data analyses of the same data set revealed that the transformation assay and rodent bioassay had a concordance of 71%, a sensitivity for carcinogens of 80.0%, and a specificity for detecting noncarcinogens of 60%. In contrast, Salmonella mutagenicity assays and rodent bioassays had a concordance of 63%, a sensitivity of 58%, and a specificity of 69%. The transformation assay complemented the Salmonella mutagenesis assay in the identification of nonmutagenic carcinogens; thus, the two assays had a combined 83% sensitivity for all carcinogens and a 75% specificity for nonmutagenic noncarcinogens.


Introduction
The National Institute of Environmental Health Sciences (NIEHS) has supported research programs using different cell transformation assay systems because such assays demonstrate phenotypes that respond to carcinogen treatments and mimic certain events in the multistep process of chemical carcinogenesis in vivo (5)(6)(7)(8). The BALB/c-3T3 transformation assay was one of the assays evaluated because chemical-induced morphologically transformed cells are easily recognized and induced at relatively high frequencies in this assay (7,(9)(10)(11)(12). Furthermore, normal BALB/c-3T3 cells have been demonstrated to be biologically different from chemical-induced transformed cells isolated from a type III focus. Whereas normal BALB/c-3T3 cells were nonmalignant and grow at low frequencies in soft agar, transformed cells readily grew in soft agar and were tumorigenic in vivo (7,11,13).
This report summarizes the results we obtained in testing 168 chemicals in a standard BALB/c-3T3 cell transformation assay protocol. The protocol was developed in this laboratory (10,14), and it differed substantially from the method first described by Kakunaga (7) and that currently recommended by government agencies (6,15). Our method modified the standard assay design to enhance the sensitivity for detection of chemical-induced transformation (14,16). The improved sensitivity was achieved without using an exogenous activation system; thus, all chemical responses were dependent on the inherent metabolic capability of the BALB/c-3T3 cells. Each chemical was tested in two or more experiments, and a total of 110 experiments were conducted over a 2.5-year period. The majority of the 168 test chemicals were selected from the NTP database of 301 chemicals tested in rodent bioassay (3); thus, the chemical structures and biological activities of most of these chemicals in several assay systems was readily available for comparative analyses (1-4).
The 168 test chemicals in this data set included comparable numbers of chemicals with three different biological activities (carcinogenicity, cytotoxicity, and mutagenicity). For example, the data set contained roughly equal numbers of carcinogens and noncarcinogens, as well as mutagenic and nonmutagenic chemicals. Furthermore, this data set also included many examples of nonmutagenic carcinogens as well as mutagenic noncarcinogens. Currently both groups of chemicals reduce our ability to predict carcinogenicity in rodents using in vitro tests for genotoxicity. Finally, this set contained many examples of cytotoxic and noncytotoxic chemicals that differed in their carcinogenic and mutagenic activities. The only chemicals tested in the assay that were omitted from this paper were 24 chemicals with unknown carcinogenicity, which were reported separately (14), 10 polycyclic aromatic hydrocarbons (unpublished data), and 21 test chemicals with a unique technical problem. The 21 chemicals rapidly reacted with plastic culture vessels at treatment dose concentrations that were tested for transforming activity and will have to be evaluated separately in a chemicalresistant culture vessel. Taken together, none of the test chemical responses detected during this investigation were either selectively, or arbitrarily, omitted from this report.
This study included a major effort to determine the appropriate upper dose limit for the BALB/c-3T3 cell transformation assay and to investigate the relationship of chemical cytotoxicity to transformation, mutagenicity, and carcinogenicity. Currently most in vitro genotoxicity assays using cultured mammalian cells employ an arbitrary upper dose limit of 5-10 mg/mL. This decision creates two problems. Because test chemicals have widely different molecular weights, the 5-10 mg/mL limit represents a high physiological treatment dose for some chemicals and a relatively lower dose for other chemicals. We avoided this problem by analyzing chemical activities in terms of treatment doses expressed in millimolar (mM) concentrations. In addition, the use of an arbitrary dose limit inhibits one from determining the actual upper dose limit. For the purpose of this investigation, the actual upper dose limit of an assay was defined as the dose at which the assay could not discriminate active and inactive test chemicals. This upper dose limit can only be determined when all chemicals are tested at comparable ranges of cytotoxic responses. This report will provide evidence that the actual upper dose limit for noncytotoxic test chemicals was equivalent to a treatment dose of 100 milliosmolar (mOsM).
The statistical methods used in this report to evaluate the activities of chemicals in one or more experiments, as well as those used to weight and rank-order chemical transformation responses, have been described previously (17)(18). These methods were developed because transformation experiments had different statistical sensitivities (17) and different detection sensitivities for chemicalinduced transformation (18). The statistical weighting procedures used mean and rank t-statistics (18), and these methods solved three data analysis problems. First, statistically weighted chemical responses provided an unbiased method for comparing responses in two independent experiments and could be used to determine whether chemical activities detected in two consecutive experiments were reproducible. Second, the statistical weighting procedure provided an easy and unbiased method for combining the data for a chemical tested in two or more trials. Third, rank-ordered and statistically weighted chemical responses provide a very sensitive means of comparing biological activities of small sets of chemicals.
This report does not present a single table with all ofthe test chemicals and their transformation responses. A binary presentation of positive and negative test chemical responses was too simplistic and masked the multifactorial activities of chemicals in this database. Thus, binary procedures were only used to demonstrate that the data set had a comparable distribution of chemicals to that of other NTP data sets. In contrast, multifactorial procedures were used to compare the activities of chemicals that shared selected biological activities. Multifactorial comparisons of groups of chemicals were examined for many different correlations between biological properties before they were presented in the format of the tables contained herein.

Cell Culture
The investigations in this report used the A31-1-13 clone of BALB/c-3T3 cells (19,20). The materials and methods used to culture the cells have been previously reported in detail (10) and are summarized in part I of these investigations (17).

Standard Clonal Survival Assay
The standard clonal survival assay was used to a) estimate the cytotoxic activity of a test chemical, b) select treatment doses for the preliminary co-culture clonal survival assay, c) assess the reproducibility of the chemicalinduced cytotoxic responses, and d) determine the relative shift in test chemical cytotoxic responses between highand low-density cell cultures. The standard clonal survival assay using low-density cultures of BALB/c-3T3 cells was conducted according to our modification (10,14) of the method described by Kakunaga (7). Briefly, 200 wild type (WT) cells were seeded in either 60-mm culture dishes (or 25-cm2 culture flasks), and chemical treatment doses were applied to triplicate cultures for 48 hr beginning 2 days after seeding. After a total culture period of 8 days, the vessels were washed, fixed in methanol, stained with Giemsa, and colonies of cells were hand tabulated according to the procedure described in part IV of these investigations (14).

Co-culture Clonal Survival Assay
The co-culture clonal survival assay was used to a) select chemical treatment doses for transformation assays, b) assess the reproducibility of chemical-induced cytotoxic responses, and c) verify that the test chemical and positive control treatment doses were cytotoxic in the transformation assay. The procedure used for the co-culture clonal survival assay has been previously reported in detail (11,13) and is summarized in part III of this series (21).

Transformation Assay
Chemical-induced transformation of BALB/c-3T3 cells was evaluated in a standard transformation assay protocol that has been reported in detail (10) and is summarized in part IV of this series (14). Briefly, each transformation assay contained three components: a standard clonal survival assay (10,14), a co-culture clonal survival assays (21), and a transformation assay (10,14). In each experiment, chemial-induced transformation was detected in 18-20 vessels/dose seeded with 3.2 x 104 cells/vessel. Chemical doses were applied to cell cultures for 48 hr, days 2-4, using standard procedures (14). A total of three to six test chemicals were included in each transformation experiment, and each chemical was tested at four treatment doses in two or more independent trials. The procedure for selecting the four doses has been described in part IV of these investigations (14), and the doses covered a range of cytotoxic responses of approximately 10-100% relative cloning efficiency (RCE).

Transformation Assay Acceptance and Evaluation Criteria
A complete explanation of the transformation assay acceptance and evaluation criteria for a test chemical evaluated in a single trial or in multiple trials is provided in part IV of this investigation (14). Briefly, a test chemical evaluated in one experiment had one of four possible transformation responses: sufficient positive (SP), limited activity (LA), sufficient negative (SN), and limited negative (LN). Briefly, an SP transformation response required that a test chemical response was statistically significant at two or more consecutive treatment doses. In contrast, an LA transformation response required that a test chemical response was statistically significant at either one treatment dose alone at the 99% confidence level or at two consecutive doses at the 95% confidence level. An SN transformation response required that a test chemical response did not have a statistically significant increase in transformation responses at any of the four treatment doses. An LN transformation response occurred under two different circumstances. First, the four test chemical treatment doses did not induce a statistically significant transformation response; however, in contrast to an SN transformation response, the test chemical treatments did not have a significant cytotoxic response. Therefore, higher concentrations of the test chemical could have induced a significant cytotoxic response, and this could have resulted in a statistically signiflcant transformation response. Second, the test chemical had the equivalent of an SN transformation response; however, the positive control for the transformation experiment was inactive and did not induce a statistically significant response.

Evaluation of Transformed Foci
The method used to evaluate transformed foci of BALB/ c-3T3 cells has been reported in detail (10) and is summarized in part IV (14) of these investigations. Briefly, the number of type I-III transformed foci of BALB/c-3T3 cells were identified microscopically using published criteria (6)(7)(8)12,17), and type III foci had three phenotypic properties: piling and overlapping cells, disorientation of cells at the periphery of the focus, and invasion of transformed cells into a contact-inhibited monolayer of WT cells. Iype I and II foci also appeared in many different sizes, but they lacked one or more of the three phenotypic properties of the type III transformed focus.

Handling of Test Chemicals
Many chemicals in this investigation had physicochemical properties that could have potentially interfered with them being adequately tested in the BALB/c-3T3 cell transformation assay (Table 1). Therefore, procedures were developed to ensure that all test chemicals would be consistently and adequately evaluated, and the procedures are described in detail in part IV of these investigations (14).

Statistical Analyses and Mathematical
Models Mathematical Transformation of Focus Data. The method used to determine the distribution of spontaneous transformed foci of BALB/c-3T3 cells has been previously reported (10,11) and is described in detail in part I of these investigations (17).
Significance of Transformation Responses. The methods used to determine the statistical significance of a chemical-induced transformation response has been described in detail in part IV of these investigation (14). Briefly, the significance was determined using analysis of variance (F-test) and modified Student's t-tests, and the computations were performed using SAS software (22).
Method for Rank-Ordering Test Chemical Transformation Responses. The method used to rank-order test chemical transformation responses on the basis of the significance of their activity in the transformation assay has been described in detail in part IV of these investigations (14). Briefly, the significance of the test chemical response was observed to vary proportionally to the magnitude ofthe t-statistic, and the t-statistic was independent of the absolute spontaneous transformation response of the solvent control. The average significance of each chemical transformation response, or mean t-statistic, was calculated by averaging the t-statistics of the four test chemical, (or two positive control) treatment doses. TPeatment doses with <5% RCE and incomplete monolayers were deleted, and negative t-statistics were arbitrarily assigned the value of zero. This mean t-statistic was used to rank order chemical transformation responses in individual experiments. The test chemical activity in two or more experimental trials was assessed using a weighted the rank t-statistic. It was calculated using all the t-statistics for test chemical treatments in two or more experimental trials (see Tables A3 and A6 for actual and estimated rank t-statistics of 168 chemical transformation responses). Examples of these calculations are provided in Results.
Effect of Statistical Sensitivity on Detection Sensitivity for BaP. Both the magnitude of the spontaneous and the benzo[a]pyrene (BaP) transformation response varied among the 110 experiments included in this investigation (17,18). Variable spontaneous transformation responses resulted in experiments with different statistical sensitivity to detect test chemical responses (17) and different detection sensitivity for BaP (18). Experiments with significantly low statistical sensitivity were demonstrated to have a low detection sensitivity for BaP (18). Therefore, these experiments had a high probability of underestimating the activity and rank t-statistics oftest chemicals. In contrast, experiments with normal or significantly high statistical sensitivity had normal detection sensitivity for BaP (18). To compensate for the diminished sensitivity to detect chemical-induced transformation, the rank t-statistic was multiplied by a correction factor to obtain an estimated rank t-statistic (14). Example calculations using the actual rank t-statistic and the correction factor to determine the estimated rank t-statistics are provided in Tables A3 and A6.
The seven very noncytotoxic chemicals evaluated in this investigation were all supplied by three companies: Sigma, Fisher Scientific, and U.S. Industrial Products.

Range of Cytotoxic Responses of 168 Chemicals
A co-culture clonal survival assay was used to measure the cytotoxic responses of 168 chemicals (21), and each chemical was tested in two or more experiments. The cytotoxic responses of individual chemicals are presented in detail in Tables Al and A4. The data set had a range of cytotoxic responses of over 7 logs. The most cytotoxic chemical was ziram, and it had an average cytotoxic response, or LD50, of 0.0000373 mM. Based on a molecular weight of 305.81, this concentration was equivalent to approximately 0.0114 ,ug/mL. The least cytotoxic chemical was witch hazel, and it had an LD50 estimated at approximately 540 mM.
The 168 chemicals were arbitrarily divided into three groups according to their relative cytotoxic responses: group 1, cytotoxic chemicals with an LD50 < 5 mM; group 2 noncytotoxic chemicals with an LD50 5 mM-100 mOsM; and group 3, very noncytotoxic chemicals with an LD50 > 100 mOsM (Table 1). There were 114 cytotoxic chemicals, 43 noncytotoxic chemicals and 11 very noncytotoxic chemicals (see Table 1). Chemical cytotoxic responses were divided into groups 1-3 based on three empirical observations. First, using the appropriate solvent vehicles, nearly all cytotoxic chemicals could be tested at treatment doses either at or below their solubility limit in culture medium. In contrast, many noncytotoxic chemicals had to be tested at treatment doses above their solubility limit to obtain cytotoxicity to the BALB/c-3T3 cells. Second, many cytotoxic chemicals (LD50 < 5 mM) were consistently inactive in the transformation assay; however, few noncytotoxic chemicals (LD50> 5 mM) were inactive if they were fully soluble in culture medium. Thus, the solubility of noncytotoxic test chemicals clearly correlated their potential activity in the transformation assay, and nearly all of the noncytotoxic chemicals that were inactive in the transformation assay had solubility problems in culture medium. Third, mutagenic and nonmutagenic test chemicals had very different profiles of cytotoxic responses. Most mutagenic chemicals were cytotoxic chemicals, while only half of the nonmutagenic chemicals were cytotoxic. Data supporting this observation will be presented later in this report.

Distribution of Cytotoxic Responses among Carcinogens and Noncarcinogens
The cytotoxic responses of carcinogenic and noncarcinogenic chemicals were compared in the data set of 168 chemicals (Table 2). This set of chemicals included 84 carcinogens and 77 noncarcinogens, and the remaining 7 test chemicals were model chemicals that had not been evaluated in the NTP rodent bioassay. These analyses of the data revealed that the data set contained a balanced distribution of cytotoxic responses among the carcinogens and noncarcinogens. Furthermore, the data set contained many examples of cytotoxic and noncytotoxic carcinogens and noncarcinogens (Table 2). Thus, these data demonstrated that in vitro cytotoxicity of chemicals to BALB/ c-3T3 cells neither correlated with nor predicted their in vivo carcinogenic activity.

Upper Dose Limit of the Transformation Assay
This investigation did not use an arbitrary upper dose limit of 5-10 mg/mL for the BALB/c-3T3 cell transformation assay. All chemicals were tested over a comparable range of cytotoxicity of 0-100% RCE, and the data from these experiments were retrospectively used to determine an empirical upper dose limit. In addition, the concentration of test chemical treatment doses was expressed in millimoles, and not in micrograms per milliliter because the 168 test chemicals had molecular weights that ranged from 46.07 for ethanol to approximately 1200 for ethoxylated dodecyl alcohol.
The upper dose limit of the BALB/c-3T3 cell transformation assay was set at 100 mOsM based on two empirical observations in this investigation. First, we observed that the test chemicals that were the least cytotoxic to the target cells all had an LD50 over a narrow range of 240-504 mOsM (see Tables 3 and A4). Second, all of the very noncytotoxic chemicals were active in the transformation assay (Appendix H). Furthermore, each of these chemicals began to induce significant transforming activity at an average concentration of 134 mOsM (Table 3). Optimal induction of transforming activity occurred at slightly higher treatment dose concentrations that were close to the chemical's LD50 dose. Taken together, the BALB/ c-3T3 cell transformation assay could not discriminate active and inactive chemicals when they were tested at concentrations above about 134 mOsM; thus, the actual dose limit for the data set of 168 chemicals was set at 100 mOsM.

Physicochemical Properties of 168 Chemicals
We were concerned in this investigation that uncontrolled test chemical technical problems could affect the activity of a chemical in the transformation assay. This concern arose because most of the 168 chemicals in this investigation had physicochemical properties that could potentially have caused technical problems when they were tested in an in vitro assay using cultured mammalian cells (refer to chemical technical problems listed in Tables  Al and A4). Fortunately, the majority of the technical problems were avoided by using specific techniques to handle the test chemicals [see Materials and Methods in part IV of this series (14)].
Nevertheless, six types of technical problems were difficult to control in this investigation, and each of these problems could have influenced the results in these experiments (Table 4). First, 21 chemicals reacted with plastic polystyrene culture vessels; thus, treatment times were reduced from 48 hr to minutes. The chemical reaction with plastic was unusual in that it occurred after the chemical was completely dissolved in the aqueous environment. Because this problem could only be overcome through the use of chemical-resistant culture vessels such as glass bottles, these chemicals were not included in this investigation. A complete list of the 21 chemicals is provided in the Discussion. Second and third, 56 chemicals were oxidized by air and 15 chemicals reacted with water; thus, the BALB/c-3T3 cells were exposed to not only the parent test chemical, but also its oxidized and hydrolyzed byproducts. Fourth, eight chemicals reacted with biochemicals; thus, they could have combined with biochemicals in the culture medium or biochemicals within the target cells. Fifth, seven chemicals bound different metal salts; thus, they could have complexed with critical metals in either the culture medium or the target cell. Finally, over half of the chemicals had solubility problems in an aqueous environment. Fortunately, the use of organic solvents in conjunction with the nonionic surfactant pluronic F68 (14,23) resulted in most of these chemicals being soluble at concentrations that induced cytotoxicity to the BALB/c-3T3 cells. Nevertheless, 14 test chemicals could not be solubilized and were insoluble at a portion or all of the treatment dose concentrations used to test for cytotoxic and transforming activities.
Thus, we predicted that any one of the six technical problems could have affected detection of chemicalinduced transformation of BALB/c-3T3 cells. Furthermore, we anticipated that the same six technical problems might also have affected detection of mutagenicity in Salmonella assays and carcinogenicity in rodent bioassay. Therefore, we examined sets of chemicals with the six technical problems to determine whether any of the problems correlated with the expression of carcinogenicity, mutagenicity, and transformation. If a chemical technical problem had either no effect or a random effect on a biological activity, then there would be equal distributions of active and inactive chemicals with this problem (i.e., ratio of active/inactive chemicals = 1.00). Conversely, if a technical problem had a consistent effect on the biological activity, then the distribution of active and inactive chemicals would be altered (i.e., ratio of active/inactive chemicals <1.00 or >1.00).
The results of these comparisons are summarized in Thble 4. It was found that two of the technical problems, reaction with air and water, had no significant effect on all three biological activities. Three additional technical problems had no effect on carcinogenicity, but they were correlated with suppressed detection of transformation and mutagenic activities. For example, chemicals with severe solubility problems and chemicals that bound metal salts aThe three biological activities included carcinogenicity in rodent bioassay, mutagenicity in Salmonella, and transformation in BALB/c-3T3 cells.
bTest chemicals in this investigation had several difficult problems: 56 chemicals were oxidized upon exposure to air; 21 chemicals reacted with plastic; 15 chemicals reacted with water; 14 chemicals had severe solubility problems in culture medium that was not corrected by the use of pluronic F68; 8 chemicals reacted with biochemicals (i.e., alkylating agents and chemicals that reacted with alcohols and amine groups); and 7 chemicals bound metal salts.
'When a technical problem had no effect on the biological property, it resulted in a relative effect of 1.00 (i.e., equal ratio of inactive and active chemicals). When a technical problem correlated with an enhanced biological activity, it resulted in a relative effect > 2.00. Conversely, when a technical problem correlated with a decreased biological activity, it resulted in a relative effect < 0.500. dChemicals with relative effects either >2.00 or <0.500.
eBecause the 21 chemicals that reacted with plastic could not be tested for transformation, they were all arbitrarily considered inactive to get a relative effect of 0.000. tended to be inactive in both BALB/c-3T3 transformation and Salmonella mutagenicity assays. Conversely, chemicals that reacted with biochemicals tended to be active in both mutagenicity and transformation assays. In contrast, only one of the technical problems had an effect on all three biological properties of carcinogenicity, transformation, and mutagenicity. Nearly all of the 21 chemicals that reacted with plastic culture vessels in BALB/c-3T3 cytotoxicity assays (unpublished observations) were carcinogenic, and they did not induce either transformation or mutagenicity in Salmonella. Thus, the presence of this technical problem significantly correlated with these chemicals being nonmutagenic carcinogens in rodent bioassay.

Transformation-Responses of 168 Chemicals
Variability among spontaneous transformation responses resulted in experiments with different statistical sensitivities to detect chemical-induced transformation responses (17). Likewise, variability among BaP responses demonstrated that individual experiments had different detection sensitivities for BaP (18). Thus, individual experiments had different sensitivities to measure test chemicalinduced transformation responses. Therefore, the responses of test chemicals in the BALB/c-3T3 cell transformation assay were evaluated in terms of the rankordered sensitivity of individual experiments to detect both spontaneous and BaP-induced transformation responses (14,17,18).
In the current study, the 168 chemicals were tested in two or more transformation assay experiments. The results ofindividual experiments for each test chemical are provided in detail in Appendices B-H. In addition, a summary of transformation responses of all the chemicals are presented in summary Tables A2 and A5. Explanations for the different response calls and evaluation criteria for a single transformation assay experiment have been reported (14) and are summarized in Materials and Methods. The final determination of the rank-ordered activity of each chemical is summarized in Tables A3 and A6. The method used for combining the activities of chemicals tested in two or more experiments has been discussed in detail in part IV of these investigations (14). For the reader who is interested in the cumulative data associated with an individual test chemical, a narrative description of the activities of individual chemicals is provided in Appendix A. To facilitate comparative analyses of chemicals with different biological activities, the same sequence of chemicals has been presented within each of the tables of Appendix A.

Comparison of Carcinogenicity with
Mutagenicity and Transformation Responses The data set of 161 carcinogens and noncarcinogens was compared to the activities of different sets of chemicals tested in other NTP investigations (1-4). In these binary analyses, the concordance of each assay was compared to the rodent bioassay using a chi-square method. In this database the concordance of Salmonella mutagenicity data with rodent bioassay was 63.4% (Table 5). Using the same group of chemicals, Salmonella assays had a sensitivity to detect carcinogens of 58.3% and a specificity for detecting noncarcinogens of 68.8%. Thus, this database was comparable to other NTP data sets (1-4), and it contained a large number of nonmutagenic carcinogens and mutagenic noncarcinogens.
Transformation data were also analyzed using the same method, and the concordance of BALB/c-3T3 transformation responses was compared to carcinogenicity data from rodent bioassay ( Table 6). The transformation assay exhibited a concordance with the rodent bioassay of 70.7%, which was 7.3% higher than Salmonella (i.e., 70.7 versus 63.4%). Likewise, the transformation assay also had a 21.7% higher sensitivity for carcinogens (i.e., 80.0% versus 58.3%) and a 9.2% lower specificity for detecting noncarcinogens (i.e., 59.7% versus 68.9%) compared to Salmonella assays.

Correlation of Test Chemical Cytotoxicity with Mutagenicity
Binary comparisons of the responses of 147 chemicals in BALB/c-3T3 transformation and 161 chemicals in Salmonella mutagenicity assays revealed that the data from both assays had a high concordance with rodent bioassay (Tables 5 and 6). However, this database contained a disproportionate number of cytotoxic, versus noncytotoxic, test chemicals (see Table 2). Thus, the concordance of the transformation and Salmonella mutagenesis assays might have been affected by the relative cytotoxicity of the test chemicals. Because the number of carcinogens and noncarcinogens was roughly equal in both of these groups of chemicals, the correlation of test chemical cytotoxicity with mutagenicity in Salmonella and rodent bioassay carcinogenicity could be directly compared.
The correlation of test chemical cytotoxicity to BALB/ c-3T3 cells with mutagenicity in Salmonella assays was examined first (Table 7). These multifactorial analyses revealed that Salmonella mutagenicity was highly correlated with chemical cytotoxicity. About 88% of the mutagenic chemicals had an LD50 < 5 mM, including both mutagenic carcinogens and noncarcinogens. In contrast, chemical cytotoxicity was not correlated with carcinogenicity; about 57% of both carcinogens and noncarcinogens were cytotoxic. Thus, cytotoxicity of the test chemical to BALB/c-3T3 cells correlated most with its capacity to induce mutations in Salmonella (Table 8). In contrast, cytotoxicity did not correlate with either the induction of transformation in BALB/c-3T3 cells or carcinogenicity in the rodent bioassay (Table 8). Thus, the in vivo capability of a chemical to induce tumors in rodents was not correlated with its in vitro cytotoxicity to a cultured mammalian cell.
Taken together, these data showed that among the four biological variables in this investigation (i.e., carcinogenicity, cytotoxicity, mutagenicity, and transformation), the highest correlation of variables was observed for results from BALB/c-3T3 transformation assays with rodent bioassay (70.7% concordance) and Salmonella mutagenicity assays (69.8% concordance) (Table 8). A less significant correlation was noted for BALB/c-3T3 cytotoxicity and Salmonella mutagenicity (63.4% concordance) and carcinogenicity and Salmonella mutagenicity (63.4% concordance). All other binary comparison of variables were not significantly correlated.

Comparison of Mutagenicity and Transformation
Because BALB/c-3T3 cell transformation and Salmonella mutagenicity assay data both exhibited a high concordance with rodent bioassay data, it was of interest to see whether the two assays detected the same profile of chemicals. If the two assays were to detect the same chemicals, this result would imply, but not prove, that the BALB/c-3T3 transformation assay was detecting primarily mutagenic test chemicals. Thus, a mutation at a gene for the transformed cell phenotype would be the most likely explanation of the activity of chemicals in the assay.
When the BALB/c-3T3 transformation response data was compared to the Salmonella assay data, the transformation assay was observed to detect 92.5% of the mutagenic carcinogens and approximately 70% of the mutagenic noncarcinogens (Table 9). These data demonstrated that the transformation assay detected a high percentage of mutagenic carcinogens and mutagenic noncarcinogens. Most of the mutagenic noncarcinogens in this group were analogues of known carcinogens, and they all had DNA reactive structural alerts (1-4). Thus, neither the BALB/c-3T3 assay nor the Salmonella mutagenesis assays were able to distinguish mutagenic carcinogens from mutagenic noncarcinogens. Fortunately, the frequency of mutagenic noncarcinogens in rodent bioassays has been relatively small. "The concordance of each biological activity with the remaining three biological activities is presented as percentage and as a bar graph. A concordance of 50%°is the control (-) and each X is equivalent to 1% concordance over the control.

Detection of Nonmutagenic Carcinogens
Because the BALB/c-3T3 transformation assay did not detect all of the mutagenic carcinogens (Table 9) but it had a higher sensitivity to detect carcinogens than Salmonella (Table 8), the transformation assay must have detected a substantial number of nonmutagenic carcinogens. This data set included 35 nonmutagenic carcinogens, which was 41.7% of the total of 84 carcinogens. The nonmutagenic carcinogens were approximately equally divided between cytotoxic and noncytotoxic chemicals. A total of 20 of 35 carcinogens were cytotoxic, and 15 of 35 chemicals were noncytotoxic chemicals. 'A total of only 7 chemicals were mutagenic, noncarcinogenic, and nontransforming (i.e., 23 -16 = 7). bA total of 31 chemicals were nonmutagenic, noncarcinogenic and nontransforming (i.e., 42 -11 = 31).
The capability of the BALB/c-3T3 assay to detect nonmutagenic carcinogens is summarized in Table 10. These data revealed that there was a high concordance of 69.3% between nonmutagenic carcinogens detected in rodent bioassay and transformation responses measured in the transformation assay. In addition, the transformation assay had a sensitivity for detecting nonmutagenic carcinogens of 63.6% (21/33), and a high specificity for not detecting noncarcinogens of 73.8% (31/42). The number of nonmutagenic carcinogens used in these analyses was 33 out of a total of 35 because 2 chemicals had an indeterminate activity (Tables A3 and A6).

Activity of Mutagenic and Nonmutagenic Carcinogens
The relative carcinogenic activity of chemicals in rodent bioassay has been evaluated in terms of their level of effect (1-3). The most active carcinogens induced tumors at one or more tissue sites in both species of rodents and were defined as having a level A effect (Table 11). In contrast, carcinogens with lower activities induced tumors in only one species, and they were evaluated as having level B, C, or D effects. Finally, chemicals that did not induce a significant tumor response were evaluated as having an equivocal activity (level E) or as being inactive (level F). Occasionally, a chemical was evaluated as having an indeterminate activity, because it has not been evaluated in a rodent bioassay that fulfilled all of the required prerequisites.
The relative level of activity of mutagenic and nonmutagenic carcinogens in rodent bioassay has also been compared (1,3). Ashby and Tennant (1,3) concluded that mutagenic carcinogens in general induced more multi-site and trans-species effects in the rodent bioassay than nonmutagenic carcinogens. Furthermore, they found evidence that mutagenic carcinogens induced tumors in a different profile of tissues sites than nonmutagenic carcinogens (1). Thus, it was of interest to determine whether the mutagenic and nonmutagenic carcinogens included in the these investigations had a comparable profile of activities as previously reported. It was also of interest to determine whether the BALB/c-3T3 cell transformation assay selectively detected carcinogens of either high or low activity.
The results of these analyses are presented in Table 12. They confirmed the reported observation that the majority of the 49 mutagenic carcinogens in this investigation had a relatively high level of effect in the rodent bioassay (i.e., 37 were A or B versus 12 that were C or D; Tables A3  and A6). In addition, a total of 74% of the carcinogens detected by Salmonella and in the BALB/c-3T3 transformation assay had a level A or B effect. In contrast, the 35 nonmutagenic carcinogens in this investigation contained roughly equal numbers of chemicals with a high or low level of effect (i.e., 19 were A or B and 16 were C or D). In this group the BALB/c-3T3 transformation assay prefer-

Discussion
There were five accomplishments of this investigation. First, we were able to validate the use of the BALB/c-3T3 transformation assay and demonstrate that it selectively detected carcinogenic, versus noncarcinogenic, test chemicals. The data from this study show that the BALB/c-3T3 cell transformation assay exhibits a somewhat higher concordance with the rodent bioassay than Salmonella mutagenicity data, i.e., 70.7 versus 63.4% (Tables 5 and 6). Thus, both of these assays selectively detected carcinogens versus noncarcinogens in this data set. However, the BALB/c-3T3 transformation assay also detected a large number of noncarcinogenic chemicals that were active in Salmonella mutagenesis assays (i.e., mutagenic noncarcinogens). Thus, neither assay could discriminate most matched pairs of carcinogens and noncarcinogens, which have very similar chemical structures. Nearly all of the matched pairs of carcinogens and noncarcinogens, such as 2-and 3-chloromethylpyridine, were active in both assays. The only matched pairs which were discriminated by BALB/c-3T3 transformation assays were the active carcinogens 2-acetylaminofluorene and BaP and the inactive noncarcinogens 4-acetylaminofluorene and benzo[e]pyrene (unpublished observation). One matched pair, HC blue 1 and 2, had either an inactive or an indeterminate activity in the transformation assay.
Second, the data obtained in this investigation demonstrate that the BALB/c-3T3 cell transformation assay can be used to selectively detect some carcinogens that were inactive in the Salmonella mutagenesis assays (i.e., nonmutagenic carcinogens). There were a total of 53 nonmutagenic carcinogens selected for evaluation in the transformation assay; however, only 35 chemicals were tested. The activities of the remaining 18 chemicals will be discussed below. Among the 35 chemicals that were tested in the standard transformation assay, 21 chemicals were active, including 10 of 19 cytotoxic and 11 of 14 noncytotoxic chemicals and 2 chemicals that had an indeterminate activity (Table 10). Of the remaining 12 inactive, nonmutagenic carcinogens, 3 carcinogens (i.e., cinnamyl anthraniliate, methapyrilene, and reserpine) have been demonstrated to be active in a new BALB/c-3T3 cell transformation assay that uses noncytotoxic treatment doses of the test chemical (unpublished data). In this protocol the BALB/c-3T3 cells are exposed continuously to multiple chemical treatment doses, and the assay is only used to evaluate the activities of cytotoxic test chemicals (LD50 < 5 mM). The remaining six cytotoxic, nonmutagenic carcinogens that were inactive (e.g., allyl isovalerate, chlorowax 40, chlorowax 500, D-liminone, tris(2ethylhexyl)phthalate, and 4-vinylcyclohexene) and one equivocal (e.g., 2-mercaptobensothiazole) transformation response await further testing with the multiple treatment (MTA) assay. In contrast, the four noncytotoxic, nonmutagenic, carcinogens could not be evaluated in this assay: decabromodiphenyloxide, di(2-ethylhexyl)adipate, di(2-ethylhexyl)phthalate, and monuron. These carcinogens had severe solubility problems in culture medium, and they were noncytotoxic at treatment dose well above their solubility limit.
Taken together, the Salmonella mutagenesis assays and the standard BALB/c-3T3 transformation assay were complementary and detected 83.3% (70/84) of the carcinogens in this investigation, including 21 nonmutagenic carcinogens. Of the remaining 14 nonmutagenic carcinogens, 10 were cytotoxic and were eligible for evaluation in the BALB/c-3T3 MTA assay. Three of these 10 chemicals have already been shown to be active in a MTA protocol that has a high sensitivity to detect carcinogenic test chemicals. Thus, only four noncytotoxic, nonmutagenic carcinogens with severe solubility problems in culture medium would be predicted to lack activity in the two assays. The sacrifice in specificity in using the two assays could be to detect all 24 mutagenic noncarcinogens and approximately 26% of the nonmutagenic noncarcinogens.
An additional group of 18 nonmutagenic carcinogens were originally selected to be tested in the BALB/c-3T3 cell transformation assay. However, these chemicals were part of a group of 21 chemicals that reacted with polystyrene, plastic culture vessels (Table 4); thus, these chemicals could not be evaluated in the standard transformation assay. This reaction occurred at concentrations that were completely soluble in culture medium and used as treatment doses to detect cytotoxic and transforming activity (Table 4). While most of these chemicals had severe solubility problems in culture medium, they were all completely soluble in culture medium supplemented with pluronic F68. Thus, these chemicals reacted with polystyrene while they were in solution in water. These chemicals are distinguishable from chemicals such as acetone that react with polystyrene as a neat chemical, but not when it is dissolved in culture medium.
Third, we were able to determine the actual upper dose limit for testing chemicals in the BALB/c-3T3 cell transformation assay. To achieve this goal, we tested all of the noncytotoxic chemicals at very high treatment doses to determine the point at which the assay could not dis-tinguish active and inactive chemicals. Furthermore, we tested a number of chemicals with solubility problems in culture medium at concentrations far exceeding their solubility limit. The results of these experiments revealed that the upper dose limit for the standard transformation assay was 100 mOsM because all of the least cytotoxic test chemicals induced significant transforming activity at treatment dose concentrations of about 134 mOsM or higher (Table 3). In the process of conducting these experiments we discovered that many of the chemicals which were tested at doses far above their solubility limit in culture medium were inactive in the transformation assay. In fact, noncytotoxic chemicals with solubility problems in culture medium were far less likely to be active in the transformation assay then chemicals that were freely soluble in culture medium (Table 4).
Fourth, we were able to ascertain that most of the chemicals tested in the standard BALB/c-3T3 transformation assay induced reproducible transformation responses. To accomplish this goal we tested all chemicals in two or more experiments. In addition, five chemicals were tested as both coded and uncoded test chemicals: C. I. basic red 9, dichlorvos, dimethylmorpholinophosphoramidate, HC red 3, and methyl carbamate. The results of these experiments showed that the cytotoxic responses of the paired chemicals were nearly identical (Thbles Al and A4). Likewise, the transformation responses of all 5 pairs of chemicals were not significantly different from one another (Tables A2 and A5). Both sources of dimethylmorpholinophosphoramidate, HC red 3, and methyl carbamate were active in the transformation assay, and C. I. basic red no 9 was inactive. The uncoded source of dichlorvos was inactive in the transformation assay, and the coded source of the chemical was evaluated as having an equivocal response.
Test chemical transformation responses were also observed to be very reproducible for the total group of 168 chemicals tested to at least two consecutive trials. A total of 82.7% (139/168) chemicals were clearly active or inactive in the transformation assay (Tables A3 and A6). Of the remaining 29 test chemicals, 8.9% (15/168) of the chemicals were evaluated as having weakly active or equivocal activities in the transformation assay. Thus, only 8.3% (14/168) of the chemicals had an indeterminate activity which resulted from different transformation responses being detected in two consecutive experiments. Therefore, the majority of the chemicals tested in the transformation assay had reproducible activities detected in two consecutive experiments.
The fifth accomplishment of this investigation has been to use the computer-automated structural evaluation software system (CASE) to investigate quantitative structure-activity relationships (QSAR) for BALB/c-3T3 transformation response data (unpublished observations). Because a combined database of 205 chemicals tested in a standard BALB/c-3T3 transformation assay was available [i.e., 168 chemicals in the current study, 24 chemicals in part IV of this series (14), and 13 polycyclic aromatic hydrocarbons (unpublished observations)], sufficient data were available to investigate a possible correlation of induction of transformation with specific portions of the chemical structure (i.e., biophores). This investigation revealed that the induction of transformation response data was significantly correlated with the presence of only 13 biophores; conversely, just four biophobes were associated with the inhibition of transformation. In addition, the study showed that the four biophobes were present on many ofthe 14 chemicals which had indeterminate activity in the transformation assay. In a companion investigation, CASE utilized the BALB-c-3T3 cell cytotoxicity data from the co-culture clonal survival assay to investigate QSAR for chemical-induced cytotoxic responses. This investigation revealed that a limited number of biophores were highly correlated with certain chemicals being cytotoxic to BALB/c-3T3 and other cultured mammalian cells [unpublished observations (26)]. QSAR investigations have determined that a limited number of biophores are highly correlated with the induction of cytotoxicity and transformation, and this information can be used to predict cytotoxic and transformation responses of chemicals untested in the BALB/c-3T3 transformation assay.
In conclusion, one of the major goals of the NTP Genetic Toxicology program during the 1980s has been to develop and evaluate in vitro assays that selectively detect carcinogenic chemicals that were inactive in Salmonella mutagenesis assays. If such assays could be developed, they could be used to investigate in vitro biological activities in common among the active chemicals and thereby lead to a clearer understanding of the mechanism(s) by which nonmutagenic carcinogens are carcinogenic in rodent bioassays. This report and the companion investigations have demonstrated progress in achieving this goal. The data in this report show that the majority of the 35 nonmutagenic carcinogens (21/35) were selectively detected in a standard BALB/c-3T3 transformation assay. In addition, CASE has identified chemical fragments of each of the nonmutagenic carcinogens that significantly correlated with the effects of the chemical in the transformation assay. Therefore, it now feasible to investigate the several different nonmutagenic carcinogens to determine the mechanism(s) by which they induced a permanent change in the transformed phenotype of BALB/c-3T3 cells. It is hoped that these investigations will help to close the current gap in our understanding of in vitro and in vivo chemical carcinogenesis.
The opinions expressed in this paper are solely those of the authors and do not necessarily reflect the positions of the U.S. Food and Drug Administration.
The experimental work was supported by funds from the National Institute of Environmental Health Sciences, contract no. NO1-ES-65150. The experimental investigations were conducted at Litton Bionetics, Inc. and Hazleton Laboratories America, Inc.  (Table A3) with no serious technical problems reported (Table Al). It was cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.171 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 83 and 89/110, respectively; the detection sensitivities for BaP of trails 1 and 2 were 70 and 87/110, respectively (Table A2). The test chemical had an SP transformation response in two consecutive experiments. 2-Acetylaminofluorene was evaluated as very active in the transformation assay, and its actual and estimated rank t-statistics were 3.12 and 4.67, respectively (Table A3). Acrylonitrile. Acrylonitrile was a level B carcinogen (Table A3). This chemical had one serious technical problem, because it was reported to be oxidized upon exposure to air (Table Al). It was cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.337 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 52 and 41/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 79 and 83/110, respectively (Table A2). In both trials the chemical had an LA transformation responses. Acrylonitrile was evaluated as having had weak activity in the transformation assay. Its actual and estimated rank t-statistics were 3.75 and 4.35, respectively (Table A3).

REFERENCES
2-Amino-4-Nitrophenol. 2-Amino-4-nitrophenol was a relatively weak level D carcinogen (Table A3). It had one difficult technical problem, because it is oxidized upon exposure to air (Table Al). It was cytotoxic to the BALB/ c-3T3 cells with an average LD50 of 0.933 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 13 and 56/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 66 and 109/110, respectively (Table A2). In trials 1 and 2 the chemical had SN transformation responses. However, there were two problems with the second transformation experiment. The test chemical cytotoxic response in the second experiment did not have a large cytotoxic shift as noted in previous experiments, and the detection sensitivity was very low in the experiment. Therefore, the test chemical should be tested in a third experiment to properly evaluate its activity in the transformation assay. 2-Amino-5-nitrophenol was therefore evaluated as having had an indeterminate activity in the transformation assay. Its actual and estimated rank t-values were both 0.00 (Table A3).
2-Amino-5-Nitrophenol. 2-Amino-5-nitrophenol was a relatively weak level D carcinogen (Table A3). It had one difficult technical problem, because it is oxidized upon exposure to air (Table Al). It was cytotoxic to the BALB/ c-3T3 cells with an average LD50 of 0.409 mM ( Table Al). The statistical sensitivities of transformation assay trials 1-3 were 71, 43 and 21/110, respectively; the detection sensitivites for BaP of trials 1-3 were 23, 9 and 53/110, respectively (Table A2). In a preliminary trial 1 the chemical had a SN transformation response. In trials 2 and 3 the chemical had SP transformation responses. 2-Amino-5nitrophenol was therefore evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were both 5.94 (Table A3). 5-Azacytidine. 5-Azacytidine was a level D carcinogen (Table A3) with no serious technical problems reported (Table Al). It was very cytotoxic to the BALB/c-3T3 cells with an average LD50 of 0.00463 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 45 and 101/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 38 and 104/110, respectively (Table A2). In trials 1 and 2 the chemical had an SP transformation response. 5-Azacytidine was evaluated as one of the most active chemicals in the transformation assay. Its actual and estimated rank t-statistics were 12.8 and 16.8, respectively (Table A3).
Benzidine-2HC1. Benzidine-2HCl was a levelA carcinogen (Table A3). It had one serious technical problem because it was reported to become oxidized upon exposure to air (Table Al). It was cytotoxic to the BALB/c-3T3 cells with an average LD50 of 0.121 mM ( Table Al). The statistical sensitivities oftransformation assay trials 1 and 2 were 30 and 34/110, respectively; the detection sensitivities for BaP oftrials 1 and 2 were 4 and 45/110, respectively (Table  A2). In trial 1 and 2 the chemical had an SP transformation response. Benzidine-2HCl was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were both 3.38 (Table A3).   Abbreviations for Test Chemical Physiochemical Properties: Physicochemical considered in this study included: [1] physical state (S = solid; L = liquid); [2] solvent vehicle (D = dimethyl sulfoxide, C = culture medium, F = pluronic F68, A = acetone, E = ethanol) and [3] technical problems. The technical problems included test chemicals that were a = reactive with acids; ac = reactive with acid chlorides and acid anhydrides; ai = reactive with air; al = reactive with alcohols; alk = alkylating agent and reacts with labile hydrogen; b = reacts with bases; bc = reacts with biochemicals (amino, hydroxyl, and carboxyl groups); hc = reacts with halogenated chemicals; k = reacts with alpha keto acids; ls = light sensitive; m = binds metals; mel = reacts with hexachloro-and trichloromelamine; met = reacts with metals (aluminum, iron, magnesium, potassium, sodium, tin or zinc); mh = metal halides; msc = reacts with miscellaneous organic chemicals (i.e., alpha-aminoethanol, chlorosulfonic acid, ethylene imine, linseed oil, maleic anhydride, oleum, or K-tert-butyloxide); o = reacts with oxidizing agents; p = reacts with plastics; pi = polymerization initiators; r = reacts with reducing agents; ru = reacts with rubber; sp = solubility problem in culture medium; tc = reacts with thiocyanates; ts = temperature sensitive; vts = very temperature sensitive; v = volatile at 37°C; and w = reacts with water [refer to MATERIALS and METHODS].
aTest Chemical: Tables Al and A3 contain 168 chemicals along with their individual CAS registry number and molecular weight. The chemicals were divided into groups of chemicals that correspond to the groups of chemicals that were compared in different text Tables 1-12. Thus, the chemicals were divided into two groups, including 114 cytotoxic test chemicals (LD50 < 5.0 mM) presented in Table Al, and 53 noncytotoxic chemicals (LD50 > 5.0 mM) presented in Table A4. The 114 cytotoxic chemicals in Table Al  bCytotoxic Response: The co-culture clonal survival assay design used to detect the cytotoxic response of the test chemical is described in Materials and Methods. The cytotoxic responses of chemicals in individual experiments are summarized in terms ofthe millimolar (mM) LD50 treatment dose that resulted in 50% survival of the chemically-treated cells relative to the survival of untreated or solvent control treated cell cultures. The LD50 cytotoxic response of each chemical in Tables Al and A4 is an average of two or more experiments with the chemical. The molecular weight of each chemical is provided in order that treatment doses could be converted from mM to ,ug/mL. For example, based upon the molecular weight of 141.61, the LD50 detected for the first chemical in Table Al, 5-chloro-o-toluidine, was 1.69 mM or 239 ,ug/mL.     **Significant transformation response, 0.001 < p --0.01. ***Significant transformation response, p < 0.001. dIBenzo(a)pyrene Transformation Response. The method used to call individual transformation experiments is described in detail in Materials and Methods. The method used to rank order the BaP transformation riesponses from the 110 experiments is based upon statistical comparison of the BaP transformation at the two treatment doses detected in an individual experiment with the median historical activity of the assay. This procedure is described in the Detection Sensitivity versus Benzo(a)pyrene Transformation Response section of the Materials and Methods. The riational for rankordering the experiments is analogous to that described for the spontaneous transformation responses (refer to footnote c above). "Test Chemical Transformation Response: The method used to call individual experiments is described in detail in Materials and Methods, and the abbreviations for the calls are provided above. The significance of the transformation responses of individual chemical treatment doses were calculated using SAS statistical software (22). The mean t-statistic represents the average of the t-statistics of the four test chemical treatment doses in the experiment. The t-statistics for individual chemical treatment doses which were used to calculate the mean t-statistic are provided in the Appendices B-H.       bRodent Bioassay Level of Activity: The relative carcinogenic activity of chemicals in rodent bioassay has been described in terms of the chemical's level of effect (1,3). The highest level A corresponds to chemicals that cause cancer in both mice and rats at one or more sites, and level B refers to chemicals that cause cancer at multiple sites in one species ofrodent. Level C includes chemicals carcinogenic at one site in both sexes of one species, and D includes chemicals carcinogenic at one site in only one sex of a single species. Level E chemicals that only equivocal evidence of carcinogenic activity. Finally, level F includes both noncarcinogens and chemicals that had inadequate carcinogenicity studies. cTransformation Response Rank t-statistic: The method used to calculate the significance of test chemical transformation responses employed SAS statistical software (22) and is described in detail in Materials and Methods. The correct t-statistics of each treatment dose of the test chemical in a single experiment are presented in the Appendices B-H, and these t-statistics were averaged to determine the mean t-statistic of the test chemical for the experiment (refer to Table A2). The mean t-statistics for two or experiments for each chemical was weighted according to the number of treatment doses evaluated and averaged to determine the actual rank t-statistic presented in this dEstimated Rank t-statistic: The estimated rank t-statistic is used to estimate both the historical behavior of the test chemical in the transformation assay, as well as predicting the future behavior or the chemical. It is calculated by correcting the actual rank t-statistic. The data presented in Table A2 showed that individual experiments had very different rank-ordered sensitivities to detect chemical-induced transformation. Therefore, the estimated rank t-statistic modified the actual rank t-statistic to correct for differences in the sensitivities of individual experiments. The method uses the rank ordered sensitivity of individual experiments to detect spontaneous and BaP-induced transformation, and an example calculation is provided below.
The most active test chemical, 5-azacytidine, had statistical sensitivities for spontaneous transformation responses of45 and 101/110 for experiments 6 and 11, respectively, and detection sensitivities for BaP of 38 and 104/110 (Table A2). Therefore, the average rank order of the two experiments was 72.0 (i.e., 45 + 101 + 38 + 104/4 = 72). For a total of 110 experiments, the median experiment has an automatic average rank order of 55.0 (i.e., 110/2 = 55.0). Therefore, the correction factor for the experimental sensitivity to detect chemical transformation was 72.0/55.0 or 1.31. Because the correction factor had been more than one, the actual rank t-statistic would have been multiplied by the correction factor to obtain the estimated rank t-statistic of 16.8. A justification for this correction factor has been reported (18), and it is explained in the Materials and Methods. ePercentage (%) of Active Chemicals: Active chemicals included chemicals with active and weakly active transformation responses. In contrast, inactive chemicals included chemicals with equivocal and inactive transformation responses. Chemicals with an indeterminate activity have to be retested in an additional experiment in order to determine their activity in the standard transformation assay. Therefore, chemicals with indeterminate transformation responses were omitted from the computation of the percentage (%) of the total chemicals that were either active or inactive in the assay.
Mutagenic Carcinogens 2-Biphenylamine. 2-Biphenylamine was a relatively weak level D carcinogen (Table A3) with no serious technical problems reported (Table Al). However, an isomer of the chemical, 4-biphenylamine, has been reported to be oxidized upon exposure to air. It was cytotoxic to the BALB/c-3T3 cells with an average LD50 of 0.421 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 49 and 34/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 24 and 45/110, respectively (Tlable A2). In trial 1 and 2 the chemical had a SP transformation response. 2-Biphenylamine was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were both 3.21 (Table A3).
4-Biphenylamine. 4-Biphenylamine was a level A carcinogen (Table A3). It had one difficult technical problem because it was reported to become oxidized upon exposure to air (Table Al). It was cytotoxic to the BALB/c-3T3 cells with an average LD50 of .479 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 14 and 11/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 72 and 39/110, respectively (Table A2). In a preliminary trial 1 the chemical had an LA transformation response. In trial 2 the chemical had an SP transformation response. 4-Biphenylamine was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were both 2.38 (Table A3).
4-Chloro-o-Phenylenediamine. 4-Chloro-o-phenylenediamine was a potent level A carcinogen (Table A3). It had one difficult technical problem because it was reported to become oxidized upon exposure to air (Table Al). It was very cytotoxic to the BALB/c-3T3 cells with an average LD50 of 0.0318 mM ( Table Al). The statistical sensitivities of transformation assay trials 1-3 were 96, 81 and 62/110, respectively; the detection sensitivities for BaP of trials 1-3 were 86, 22 and 48/110, respectively (Table A2). In a preliminary trial 1, the chemical had a LN transformation response because the test chemical treatment doses did not induce significant cytotoxic activity. In trials 2 and 3 the chemical had an SP transformation response. 4-Chloro-o-phenylenediamine was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were both 5.84 (Table A3).
3-(Chloromethyl)pyridine-HCl. 3-(Chloromethyl)pyridine-HCl was a potent level A carcinogen (Table A3) with no serious technical problems reported (Table Al). It was very cytotoxic to the BALB/c-3T3 cells with an average LD50 of 0.0756 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 99 and 36/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 82 and 28/110, respectively (Table A2). In a preliminary trial 1 the chemical had an LA transformation response. In trial 2 the chemical had an SP transformation response. 3-(Chloromethyl)pyridine-HCl was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were 2.06 and 2.29, respectively (Table A3).
4-Chloro-o-Toluidine-HCl. 4-Chloro-o-toluidine-HCl was a relatively weak level C carcinogen (Table A3) with no serious technical problems reported (Table Al). However, an isomer of the chemical, 5-chloro-o-toluidine, has been reported to be oxidized upon exposure to air (Table Al). It was cytotoxic to the BALB/c-3T3 cells with an average LD50 of 0.650 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 2 and 41/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 2 and 83/110, respectively (Table A2). In trial 1 and 2 the chemical had an SP transformation response. 4-Chloro-o-toluidine-HCl was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were both 5.61 (Table A3). 5-Chloro-o-Toluidine. 5-Chloro-o-toluidine was a potent level B carcinogen (Table A3). It had one serious technical problem because it was reported to become oxidized upon exposure to air (Table Al). It was moderately cytotoxic to the BALB/c-3T3 cells with an average LD50 of 1.69 mM ( Table Al). The stalistical sensitivities of transformation assay trials 1 and 2 were 2 and 41/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 2 and 83/110, respectively (Table A2). In trial 1 and 2 the chemical had an SP transformation response. 5-Chloro-o-toluidine was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were both 3.92 (Table A3).
C. I. Acid Orange 3. C. I. Acid orange 3 was a relatively weak level D carcinogen (Table A3) with no serious technical problems reported (Table Al). It was cytotoxic to the BALB/c-3T3 cells with an average LD50 of 0.102 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 47 and 63/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 43 and 101/110, respectively (Table A2). In a preliminary trial 1 the chemical had an LA transformation response. In trial 2 the chemical had an SP transformation response. C. I. acid orange 3 was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were 2.89 and 3.34, respectively (Table A3).
C. I. Basic Red 9-HCI. C. I. Basic red 9-HCl was one of five chemicals that was tested as both a coded and an uncoded test chemical in this investigation. It was a potent level A carcinogen (Table A3) with no serious technical problems reported (Table Al). Both chemical samples were very cytotoxic to the BALB/c-3T3 cells with an average LD50 of 0.00281 and 0.00216 mM (Table Al). For the uncoded test chemical, the statistical sensitivities of transformation assay trials 1-3 were 64, 108 and 82/110, respectively; the detection sensitivities for BaP of trials 1-3 were 84, 99 and NA/110, respectively (Table A2). For the coded test chemical, the statistical sensitivities of trials 1 and 2 were 80 and 8/110; the detection sensitivities for BaP were 89 and 29/110, respectively. The coded and uncoded test chemical had SN transformation responses in a total of 5 trials. Therefore, C. I. basic red 9-HCl was evaluated as inactive in the transformation assay. The actual and estimated rank t-statistics of the uncoded test chemical were 0.41 and 0.65, respectively (Table A3). The actual and estimated rank t-statistics of the coded test chemical were both 0.05 (Table A3). Taken together, the coded and uncoded test chemicals had nearly identical cytotoxic and transforming activities in the BALB/c-3T3 cell transformation assay.
C. L Disperse Blue 1. C. 1. Disperse blue 1 was a level C carcinogen (Table A3). It had one difficult technical problem. It was insoluble in culture medium at a portion of the treatment doses that were used to evaluate both cytotoxic and transforming activity (Table Al). In addition, this test chemical was observed to bind to the target cells, and it could not be removed using the standard washing procedure. It was cytotoxic to the BALB/c-3T3 cells with an average LD50 of 0.240 mM ( Table Al). The statistical sensitivities of transformation assay trials 1-3 were 80, 54 and 5/110, respectively; the detection sensitivities for BaP of trials 1-3 were 89, 44 and 46/110, respectively (Table  A2). In a preliminary trial 1, the chemical had a SP type III focus transformation response. In trials 2 and 3, the chemical had a SN type III transformation response. In contrast, the test chemical had an SP type I-III transformation response for all three trials. Thus, this test chemical had the unusual and consistent capability of inducing very significant levels of type I and II foci, but not for type III foci. This type of transformation response is shared by two other carcinogens, asbestos and polybrominated biphenyl mixture. Taken together, C. I. acid orange 3 was evaluated as weakly active in the transformation assay. Its actual and estimated rank t-statistics for the type III transformation response were both 1.31; however, the actual and estimated rank t-statistics for the type I-III response were both 7.08 (Table A3).
C. I. Disperse Yellow 3. C. I. Disperse yellow 3 was a level A carcinogen (Table A3) with no serious technical problems reported (Table Al). It was moderately cytotoxic to the BALB/c-3T3 cells with an average LD50 of 1.50 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 24 and 56/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 10 and 109/110, respectively (Table A2). In trials 1 and 2 the chemical had an SN transformation response. C. I. disperse yellow 3 was evaluated as having had an indeterminate activity in the transformation assay. Its actual and estimated rank t-statistics were both 0.07 (Table A3).
C. I. Solvent Yellow 14. C. I. Solvent yellow 14 was a relatively weak level C carcinogen (Table A3). It was insoluble at a portion of the treatment doses that were evaluated for both cytotoxic and transforming activities (Table Al). It was cytotoxic to the BALB/c-3T3 cells with an average LD50 Of 0.199 mM ( Table Al). The statistical sensitivities of transformation assay trials 1-4 were 105, 107, 99 and 105/110, respectively; the detection sensitivities for BaP of trials 1-4 were 69, 106,37, and 39/110, respectively (Table A2). In trials 1, 2 and 4 the chemical had SP transformation responses. In trial 3 the chemical had an LA transformation response. The test chemical was evaluated in more than two experiments, because it was used as a model test chemical in the development of additional assay protocols. C. I. solvent yellow 14 was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were 3.52 and 5.34, respectively (Table A3).
Cytembena. Cytembena was a level B carcinogen (Table  A3) with no serious technical problems reported (Table  Al). It was cytotoxic to the BALB/c-3T3 cells with an average LD50 of 0.153 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 47 and 71/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 43 and 23/110, respectively (Table A2). In a preliminary trial 1 the chemical had an LA transformation response. In trial 2 the chemical had an SP transformation response. Cytembena was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were both 5.93 (Table A3).
1,2-Dibromo-3-Chloropropane. 1,2-Dibromo-3-chloropropane was a potent levelA carcinogen (Table A3) with no serious technical problems reported (Table Al). It was cytotoxic to the BALB/c-3T3 cells with an average LD50 of 0.401 mM ( Table Al). The statistical sensitivities of transformation assay trials 1-3 were 66, 78 and 27/110, respectively; the detection sensitivities for BaP of trials 1-3 were 81,58 and 63/110, respectively (Table A2). In a preliminary trial 1, the test chemical had an LN transformation response, because the chemical treatment doses were noncytotoxic to the target cells. In trials 2 and 3 the chemical had an SP transformation response. 1,2-Dibromo-3-chloropropane was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were 5.59 and 5.74, respectively (Table A3).
2,6-Dichloro-p-Phenylenediamine. 2,6-Dichlorop-phenylenediamine was a relatively weak level C carcinogen (Table A3). It had one difficult technical problem because it was reported to become oxidized upon exposure to air (Tlable Al). It was cytotoxic to the BALB/c-3T3 cells with an average LD50 of 0.921 mM ( Table Al). The statistical sensitivities oftransformation assay trials 1 and 2 were 19 and 91/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 34 and 85/110, respectively (Table A2). In trials 1 and 2 the chemical had an SP transformation response. 2,6-Dichlorop-phenylenediamine was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were 2.26 and 2.35, respectively (Table A3).
1,3-Dichloropropene. 1,3-Dichloropropene was a potent level A carcinogen (Table A3) with no serious technical problems reported (Table Al). It was cytotoxic to the BALB/c-3T3 cells with an average LD50 of 0.280 mM ( Table Al). The statistical sensitivities of transformation assay trials 1-3 were 4, 18 and 26/110, respectively; the detection sensitivities for BaP of trials 1-3 were 1, 31 and 93/110, respectively (Table A2). In a preliminary trial 1 the chemical had a SP transformation response, and in trial 2 the chemical had a SN transformation response. Because these responses were disparate and significantly different from one another, the test chemical was evaluated in a third trial. In the third experiment the chemical had an LA transformation response. 1,3-Dichloropropene was evaluated as weakly active in the transformation assay. Its actual and estimated rank t-statistics were both 1.77, respectively (Table A3).
Dichlorvos. Dichlorvos was one of five chemicals that was tested as both a coded and an uncoded test chemical in this investigation. It was evaluated as a noncarcinogen is its first rodent bioassay trial; however, it was determined in a second trial using a different route of exposure to be a potent level A carcinogen (Table A3). This test chemical had one difficult technical problem because it was rapidly hydrolyzed in water (Table Al). Both chemical samples were cytotoxic to the BALB/c-3T3 cells with an average LD50 of 0.145 and 0.140 mM (Table Al). For the uncoded test chemical, the statistical sensitivities of transformation assay trials 1-3 were 98, 72 and 59/110, respectively; the detection sensitivities for BaP oftrials 1-3 were 60, NA and 17/110, respectively (Table A2). For the coded test chemical, the statistical sensitivities of trials 1 and 2 were 9 and 15/110, the detection sensitivities for BaP were 37 and 21/110, respectively. The uncoded test chemical had SN transformation responses in 2 trials and an LA transformation response in one trial. The coded test chemical had an SN transformation response in a preliminary trial and an SP transformation response in trial 2. The mean t-statistics of the SP and LA transformation responses were not significantly different from the corresponding SN responses, which showed that the test chemical activity in the assay was relatively weak. Taken together, dichlorvos was evaluated as having had equivocal activity in the transfonnation assay. The actual and estimated rank t-statistics of the uncoded test chemical were 1.49 and 1.66, respectively (Table A3). The actual and estimated rank t-statistics of the coded test chemical were both 1.68 (Table A3). Taken together, the coded and uncoded test chemicals had nearly identical cytotoxic and transforming activities in the BALB/c-3T3 cell transformation assay.
Diglycidyl Resorcinol Ether. Diglycidyl resorcinol ether was a potent levelA carcinogen (Table A3). It is an alkylating chemical; thus, it could react with labile hydrogen atoms not only on DNA, but also on a variety of biochemicals in the culture medium (Table Al). It was very cytotoxic to the BALB/c-3T3 cells with an average LD50 of 0.00416 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 45 and 103/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 38 and 73/110, respectively (Table A2). In trials 1 and 2 the chemical had an SP transformation response. Diglycidyl resorcinol ether was evaluated as one ofthe most active chemicals in the transformation assay. Its actual and estimated rank t-statistics were 9.31 and 11.0, respectively (Table A3).
2,4-Dinitrotoluene. 2,4-Dinitrotoluene is a relative potent level B carcinogen (Table A3) with no serious technical problems reported (Table Al). It was cytotoxic to the BALB/c-3T3 cells with an average LD50 of 0.917 mM ( Table Al). The statistical sensitivities of transformation assay trials 1-3 were 88, 104 and 88/110, respectively; the detection sensitivities for BaP of trials 1-3 were 77, 90 and NA/110, respectively (Table A2). In a preliminary trial 1 the chemical had an LA transformation response. In trials 1-3 the chemical had SN transformation responses. 2,4-Dinitrotoluene was evaluated as inactive in the transformation assay. Its actual and estimated rank t-statistics were 0.73 and 1.19, respectively (Table A3).
Epichlorohydrin. Epichlorohydrin was a level A carcinogen (Table A3). It had two serious technical problems, because it was reported to become oxidized upon exposure to air and it reacts with water (Table Al). It was cytotoxic to the BALB/c-3T3 cells with an average LD50 of .364 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 98 and 72/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 60 and NA/110, respectively (Table A2). In a preliminary trial 1 the chemical had an LA transformation response. In trial 2 the chemical had an SP transformation response. Epichlorohydrin was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were 3.03 and 4.22, respectively (Table A3).
1,2-Epoxybutane. 1,2-Epoxybutane was a relatively weak level D carcinogen (Table A3). It was reported to be a highly reactive chemical (Table Al), and it reacts with carboxyl and hydroxyl groups found on constituent biochemicals in culture medium, as well as in the target cells. It was moderately cytotoxic to the BALB/c-3T3 cells with an average LD50 of 1.45 mM ( Table Al). The statistical sensitivities of transformation assay trials 1-3 were 4, 26 and 20/110, respectively; the detection sensitivities for BaP of trials 1-3 were 1, 93 and 19/110, respectively (Table  A2). In a preliminary trial 1 the chemical had an LN transformation response because the chemical treatment doses were too high and completely cytotoxic to the cells. In trials 2 and 3 the chemical had SP transformation responses. 1,2-Epoxybutane was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were both 2.99 (Table A3).
1,2-Epoxypropane. 1,2-Epoxypropane was a potent level A carcinogen (Table A3). It had one difficult technical problem because it was reported to become oxidized upon exposure to air (Table Al). It was moderately cytotoxic to the BALB/c-3T3 cells with an average LD50 of 1.60 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 70 and 57/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 41 and 88/110, respectively (Table A2). In trials 1 and 2 the chemical had an SP transformation response. 1,2-Epoxypropane was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were 4.48 and 5.21, respectively (Table A3).
Ethylene Dibromide. Ethylene dibromide was a level A carcinogen (Table A3). It was reported to be highly reactive chemical, but none of these problems were serious (Table Al). It was moderately cytotoxic to the BALB/c-3T3 cells with an average LD50 of 1.69 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 32 and 41/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 108 and 83/110, respectively (Table A2). In trials 1 and 2 the chemical had an SP transformation response. Ethylene dibromide was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were 5.68 and 6.82, respectively (Table A3).
HC Blue 1. HC Blue 1 was a level A carcinogen (Table  A3). It had one difficult technical problem because it was reported to become oxidized upon exposure to air (Table  Al). It was cytotoxic to the BALB/c-3T3 cells with an average LD50 of 1.96 mM ( Table Al). The statistical sensitivities of transformation assay trials 1-3 were 100, 79 and 88/110, respectively; the detection sensitivities for BaP of trials 1-3 were 42, 75 and NA/110, respectively (Table A2). In trials 1 and 2 the chemical had an SN transformation response. Ordinarily this chemical would not have been tested in a third experiment, but it was selected as a model chemical with an inactive response in the transformation assay. In the third experiment, the test chemical had an SP transformation response. This disparate transformation response could have been caused by the different sample batches of test chemicals that were tested in experiments 1 and 2 versus experiment 3. Therefore, the test chemical has to be tested in a fourth experimental trial. HC blue 1 was evaluated as having had an indeterminate activity in the transformation assay. Its actual and estimated rank t-statistics were 1.72 and 2.40, respectively (Table A3). Iodinated Glycerol. Iodinated glycerol was a level A carcinogen (Table A3) with no serious technical problems reported (Table Al). It was moderately cytotoxic to the BALB/c-3T3 cells with an average LD50 of 3.47 mM ( Table  Al). The statistical sensitivities of transformation assay trials 1 and 2 were 32 and 28/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 108 and 15/110, respectively (Table A2). In a preliminary trial 1 the chemical had an SP transformation response. In trial 2 the chemical had an SN transformation response. The disparate transformation responses were examined further, and the mean t-statistics of the two experiments were not significantly different from one another. In addition, there was a dose-related increase in test chemical activity in the experiment with an SN response at treatment doses that were comparable to that inducing an SP response. Taken together, iodinated glycerol was evaluated as weakly active in the transformation assay. Its actual and estimated rank t-statistics were both 3.68, respectively (Table A3). Melphalan. Melphalan was a level A carcinogen (Table  A3). It had one serious technical problem because it was reported to react with water (Table Al). It was very cytotoxic to the BALB/c-3T3 cells with an average LD50 of 0.00120 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 10 and 31/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 5 and 52/110, respectively (Table A2). In a preliminary trial 1 the chemical had an LA transformation response. In trial 2 the chemical had an SP transformation response. Melphalan was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were both 3.66, respectively (Table A3).
N-Methyl-o-Acrylamide. N-Methyl-o-acrylamide is a level B carcinogen (Table A3) with no insurmountable technical problems (Table Al). It was moderately cytotoxic to the BALB/c-3T3 cells with an average LD50 of 1.75 mM ( Table Al). The statistical sensitivities of trials 1 and 2 were 47 and 67/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 43 and 78/110, respectively (Table A2). In a preliminary trial 1 the chemical had an LA transformation response. In trial 2 the chemical had an SP transformation response. N-Methyl-o-acrylamide was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were 3.10 and 3.31, respectively (Table A3). 4,4-Methylenediamine. 4,4-Methylenediamine was a potent level A carcinogen (Table A3). It had one difficult technical problem because it was reported to become oxidized upon exposure to air (Table Al). It was moderately cytotoxic to the BALB/c-3T3 cells with an average LD50 of 1.56 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 60 and 104/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 27 and 90/110, respectively (Table A2). In trials 1 and 2 the chemical had an SP transformation response. 4,4-Methylenedianiline was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were 1.87 and 2.39, respectively (Table A3). N-Methyl-N'-Nitro-N-Nitrosoguanidine. N-Methyl-N'-Nitro-N-Nitrosoguanidine was a potent level A carcinogen (Table A3). It had one serious technical problem because it reacts with water (Table Al). It was very cytotoxic to the BALB/c-3T3 cells with an average LD50 of 0.0154 mM ( Table Al). The statistical sensitivities oftransformation assay trials 1 and 2 were 43 and 54/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 9 and NA/lO, respectively (Table A2). In trials 1 and 2 the chemical had an SP transformation response. N-Methyl-N'-nitro-N-nitrosoguanidine was one of the most active chemicals in the transformation assay. Its actual and estimated rank t-statistics were both 10.3 (Table A3).
2-Naphthylamine. 2-Naphthylamine was a level B carcinogen (Table A3). It had one serious technical problem because it was reported to become oxidized upon exposure to air (Table Al). It was moderately cytotoxic to the BALB/ c-3T3 cells with an average LD50 of 1.59 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 96 and 37/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 86 and 20/110, respectively (Table A2). In trials 1 and 2 the chemical had an SP transformation response. 2-Naphthylamine was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were 3.66 and 3.98, respectively (Table A3).
Nitrofurantoin. Nitrofurantoin was a potent level A carcinogen (Table A3). It was reported to be a highly reactive chemical, but none of the problems were serious (Table Al). It was cytotoxic to the BALB/c-3T3 cells with an average LD50 of 0.106 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 93 and 43/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 100 and 9/110, respectively (Table A2). In a preliminary trial 1 the chemical had an SN transformation response. In trial 2 the chemical had an SP transformation response. The disparate transformation responses were examined further, and the mean t-statistics of the two experiments were not significantly different from one another. In addition, there was a dose-related increase in test chemical activity in the experiment with an SN response at treatment doses that were comparable to that inducing an SP response. Taken together, nitrofurantoin was evaluated as having had weak activity in the transformation assay. Its actual and estimated rank t-statistics were 2.26 and 2.52, respectively (Table A3).
Nitrofurazone. Nitrofurazone was a potent level A carcinogen (Table A3) with no serious technical problems reported (Table Al). It was very cytotoxic to the BALB/ c-3T3 cells with an average LD50 of 0.0515 mM ( Table Al). The statistical sensitivities of transformation assay trials 1-3 were 24, 17 and 63/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 10, 6 and 101/110, respectively (Table A2). In a preliminary trial 1 the chemical had an LN transformation response, because the chemical treatment doses were too cytotoxic. In trials 2 and 3 the chemical had an SP transformation response. Nitrofurazone was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were both 4.12, respectively (Table A3).
2-Nitro-p-Phenylenediamine. 2-Nitro-p-phenylenediamine was a relatively weak level D carcinogen (Table  A3). It had one difficult technical problem because it was reported to become oxidized upon exposure to air (Table  Al). It was cytotoxic to the BALB/c-3T3 cells with an average LD50 of 0.947 mM ( Table Al). The statistical sensitivities of transformation assay trials 1-3 were 100, 79 and 31/110, respectively; the detection sensitivities for BaP of trials 1-3 were 42, 75 and 52/110, respectively (Table A2). In a preliminary trial 1 the chemical had an SP transformation response. In trial 2 the chemical had an SN transformation response. The disparate transformation responses were examined further, and the mean t-statistics of the two experiments were significantly different from one another. Therefore, the chemical was tested in a third trial, and the activity in this experiment was evaluated as an SP. There was no obvious reason for the absence of activity of the test chemical in the second experiment. Taken together, 2-nitro-p-phenylenediamine was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were 3.60 and 4.13, respectively (Table A3). 4,4-Oxydianiline. 4,4-Oxydianiline was a potent level A carcinogen (Table A3). It had one serious technical problem, because it was reported to become oxidized upon exposure to air (Table Al). It was cytotoxic to the BALB/ c-3T3 cells with an average LD50 of 0.270 mM ( Table Al).
The statistical sensitivities of trials 1 and 2 were 25 and 16/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 36 and 7/110, respectively (Table A2). In trials 1 and 2 the chemical had an SP transformation response. 4,4-Oxydianiline was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were both 2.60 (Table A3). Quinoline. Quinoline was a level C carcinogen (Table  A3). It was reported to be a highly reactive chemical, but none of the problems were serious (Table Al). It was moderately cytotoxic to the BALB/c-3T3 cells with an average LD50 of 4.09 mM ( Table Al). The statistical sensitivities of transformation assay trials 1-4 were 87, 78, 42 and 26/110, respectively; the detection sensitivities for BaP of trials 1-4 were 64, 58, 25 and 93/110, respectively (Table A2). In trials 1 and 2 the chemical had an LN transformation response because the chemical treatment doses did not induce a significant cytotoxic activity. In trial 3 the chemical had an LA transformation response, and trial 4 the response was evaluated as an SP. Quinoline was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were both 2.94 (Table A3).
Selenium Sulfide. Selenium sulfide was a potent level A carcinogen (Table A3) with no serious technical problems reported (Table Al). It was cytotoxic to the BALB/c-3T3 cells with an average LD50 of 0.125 mM ( Table Al). The statistical sensitivities of transformation assay trials 1-3 were 105, 101 and 54/110, respectively; the detection sensitivities for BaP of trials 1-3 were 69, 104 and 44/110, respectively (Table A2). In a preliminary trial 1, the chemical had an SP transformation response. In trial 2, the chemical had an SN transformation response. The disparate transformation responses were examined further, and the mean t-statistics of the two experiments were significantly different from one another. Therefore, the test chemical had to be tested in a third experiment, and the activity in this trial was evaluated as an SR There was no obvious explanation for the disparate transformation responses in second experiment, verses the experiments 1 and 3. Taken together, selenium sulfide was evaluated as weakly active in the transformation assay. Its actual and estimated rank t-statistics were 1.92 and 2.78, respectively (Table A3).
o-Toluidine. o-Toluidine was a potent level A carcinogen (Table A3). It had one difficult technical problem because it was reported to become oxidized upon exposure to air (Table Al). It was moderately cytotoxic to the BALB/c-3T3 cells with an average LD50 of 4.33 mM ( Table Al). The statistical sensitivities of transformation assay trials 1-3 were 87, 106 and 59/110, respectively; the detection sensilivities for BaP of trials 1 and 2 were 64, 51 and 17/110, respectively (Table A2). In a preliminary trial 1 the chemical had an LA transformation response. In trial 2 the chemical had an SP transformation response, but it did not induce a significant increase in type I and II foci (data not presented). Because ofthis unusual activity, the test chemical was tested in a third trial. In the third experiment the chemical response was evaluated as an SP. Taken together, o-toluidine was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were 4.83 and 5.62, respectively (Table A3).
Ziram. Ziram is a relatively weak level D carcinogen (Table A3) with no serious technical problems reported (Table Al). It was the most cytotoxic chemical in this group of test chemicals, and it had an average LD50 of 0.0000373 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 17 and 56/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 6 and 109/110, respectively (Table A2). In a preliminary trial 1 the chemical had an LA transformation response. In trial 2 the chemical had an SP transformation response. Ziram was evaluated as active in the transfonnation assay. Its actual and estimated rank t-statistics were both 2.72 (Table A3).
Cytotoxic, Mutagenic Noncarcinogens 4-Acetylaminonuorene. 2-Acetylaminofluorene is a level F(I) noncarcinogen because it has not been evaluated in a complete rodent bioassay (Table A3). It had one difficult technical problem, because it had a solubility limit in culture medium supplemented with pluronic F68 of 200 pLg/ml ( Table Al). The test chemical was moderately cytotoxic to the BALB/c-3T3 cells and had an average LD50 estimated to be over 900 ,ug/ml or about 0.07 mM (Table Al). Thus, the LD50 was considerably above the solubility limit of the test chemical. The statistical sensitivities of transformation assay trials 1 and 2 were 103 and 83/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 73 and 70/110, respectively (Table A2). The test chemical had an LN transformation response in both experiments, and it was tested at treatment doses that were both above and below its solubility limit. Taken together, 4-acetylaminofluorene was evaluated as an inactive chemical with an indeterminate activity in the transformation assay. Its actual and estimated rank t-statistics were 1.30 and 1.49, respectively (Table A3).
4'-(Chloroacetyl)acetanilide. 4'-(Chloroacetyl)acetanilide is a level F noncarcinogen (Table A3). It had no insurmountable technical problems ( Table Al). The test chemical was very cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.00336 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 58 and 8/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 55 and 29/110, respectively (Table A2). In trials 1 and 2 the test chemical had LA transformation responses. 4'-(Chloroacetyl)acetanilide evaluated as having equivocal activity in the transformation assay, and its actual and estimated rank t-statistics were both 1.65 (Table A3).
2-(Chloromethyl)pyridine-HCI. 2-(Chloromethyl)pyridine is a level F noncarcinogen (Table A3). It had no insurmountable technical problems ( Table Al). The test chemical was cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.118 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 99 and 36/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 82 and 28/110, respectively (Table A2).
In trials 1 and 2 the test chemical had SP transformation responses. 2-(Chloromethyl)pynidine was evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were 2.56 and 2.85, respectively (Table A3).
3-Chloro-p-Toluidine. 3-Chloro-p-toluidine is a level F noncarcinogen (Table A3). It had one difficult technical problem. It was reported to become oxidized upon exposure to air, and it was exposed to air during the treatment period (Table Al). The test chemical was moderately cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 1.17 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 2 and 41/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 2 and 83/110, respectively (Table A2). In trials 1 and 2 the test chemical had SP transformation responses. 3-Chloro-p-toluidine was evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were both 3.26 (Table A3). Coumaphos. Coumaphos is a level F noncarcinogen (Table A3). It had no insurmountable technical problems ( Table Al). The test chemical was cytotoxic to the BALB/ c-3T3 cells and had an average LD50 of 0.218 mM ( Table  Al). The statistical sensitivities of transformation assay trials 1-3 were 40, 8 and 33/110, respectively; the detection sensitivities for BaP of trials 1-3 were 54, 29 and 14/110, respectively (Table A2). The test chemical had an SN transformation response in the first experiment, and an SP response in the second experiment. Because the mean t-statistics ofthe transformation responses in the first two experiments were significantly different from one another, the test chemical was evaluated in a third trial. In the third experiment the test chemical had an SP transformation response. Coumaphos was evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were both 4.12 (Table A3). Dimethoate. Dimethoate is a level F noncarcinogen (Table A3). It had no insurmountable technical problems ( Table Al). The test chemical was cytotoxic to the BALB/ c-3T3 cells and had an average LD50 of 0.602 mM ( Table  Al). The statistical sensitivities of transformation assay trials 1 and 2 were 90 and 18/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 13 and 31/110, respectively (Table A2). The test chemical had an LA transformation response in the first experiment, and an SP response in the second experiment. Dimethoate was evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were both 3.31 (Table A3).
2,4-Dimethoxcyaniline-HCI. 2,4-Dimethoate is a level F noncarcinogen (Table A3). It had no insurmountable technical problems ( Table Al). The test chemical was moderately cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 1.13 mM ( Table Al). The statistical sensitivities of transformation assay trials 1-3 were 7, 63 and 54/110, respectively; the detection sensitivities for BaP of trials 1-3 were 56, 101 and 44/110, respectively (Table A2). The test chemical had a SP transformation response in the first experiment, and a SN response in the second experiment. Because the mean t-statistics of the transformation responses in the first two experiments were significantly different from one another, the test chemical was evaluated in a third trial. The test chemical had a SN transfor-mation response in the third experiment. 2,4-Dimethoxyaniline was evaluated as having had equivocal activity in the transformation assay, and its actual and estimated rank t-statistics were both 3.06 (Table A3).
HC Red 3. HC red 3 was one of five chemicals that was tested as both a coded and an uncoded test chemical in this investigation. It is a level E noncarcinogen (Table A3). It had no insurmountable technical problems ( Table Al). The test chemical was moderately cytotoxic to the BALB/ c-3T3 cells and had an average LD50 of 3.72 mM as a uncoded test chemical and 4.50 mM as a coded chemical (Table Al). For the uncoded test chemical the statistical sensitivities of transformation assay trials 1 and 2 were 60 and 86/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 27 and 74/110, respectively (Table A2). In trials 1 and 2 the uncoded test chemical had SP transformation responses. For the coded test chemical the statistical sensitivities of transformation assay trials 1 and 2 were 93 and 33/110, respectively; the detection sensitivities for BaP were 100 and 14, respectively. The coded test chemical had an LA transformation response in the first experiment, and an SP transformation response in the second experiment. Both the uncoded and the coded HC Red 3 were evaluated as active in the transformation assay. The coded test chemical actual estimated rank t-statistics were 5.60 and 6.11, respectively; the uncoded test chemical actual and estimated rank t-statistics were 2.64 and 2.96, respectively (Table A3).
8-Hydroxyquinoline. 8-Hydroxyquinoline is a level F noncarcinogen (Table A3). It had one difficult technical problem because it was reported to combine with different metal salts (Table Al). Thus, it could have combined with metal salts in FBS and EMEM medium. The test chemical was very cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.00251 mM ( Table Al). The statistical sensitivities of transformation assay trials 1-3 were 94, 102 and 39/110, respectively; the detection sensitivities for BaP of trials 1-3 were 91, 95 and 68/110, respectively (Table A2). The test chemical had an SP transformation response in the first experiment, and an SN response in the second experiment. Because the mean t-statistics of the transformation responses of the first two experiments were significantly different from one another the chemical was evaluated in a third trial. The test chemical had an SN transformation response in the third experiment. 8-Hydroxyquinoline was evaluated as having equivocal activity in the transformation assay, and its actual and estimated rank t-statistics were 0.78 and 1.16, respectively (Table A3). Malaoxon. Malaoxon is a level F noncarcinogen (Table  A3). It had one difficult technical problem. It was reported to become oxidized upon exposure to air, and it was exposed to air during the treatment period ( Table Al). The test chemical was cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.468 mM ( Table Al). The statistical sensitivities oftransformation assay trials 1 and 2 were 87 and 106/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 64 and 51/110, respectively (Table A2). The test chemical had an LA transformation response in the first experiment, and an SP response in the second experiment. Malaoxon was evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were 8.14 and 11.4, respectively (Table A3).
I-Naphthylamine. 1-Naphthylamine is a level F noncarcinogen (Table A3). It had one difficult technical problem. It was reported to become oxidized upon exposure to air, and it was exposed to air during the treatment period ( Table Al). The test chemical was cytotoxic to the BALB/ c-3T3 cells and had an average LD50 of 0.506 mM ( Table  Al). The statistical sensitivities of transformation assay trials 1 and 2 were 96 and 77/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 86 and 67/110, respectively (Table A2). In trials 1 and 2 the test chemical had SP transformation responses. 1-Naphthylamine was evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were 2.82 and 4.18, respectively (Table A3). (Table A3). It had no insurmountable technical problems ( Table Al). The test chemical was cytotoxic to the BALB/ c-3T3 cells and had an average LD50 of 0.125 mM ( Table  Al). The statistical sensitivities of transformation assay trials 1 and 2 were 69 and 63/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 12 and 101/110, respectively (Table A2). In trials 1 and 2 the test chemical had LA transformation responses. N-(1-naphthyl)ethylenediamine-2HCl evaluated as having equivocal activity in the transformation assay, and its actual and estimated rank t-statistics were 1.64 and 1.83, respectively (Table A3).
I-Nitronaphthalene. 1-Nitronaphthalene is a level F noncarcinogen (Table A3). It had no insurmountable technical problems ( Table Al). The test chemical was cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.464 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 49 and 63/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 24 and 101/110, respectively (Tlable A2). The test chemical had an SN transformation response in the first experiment, and an SP response in the second experiment. Because the mean t-statistic responses ofthe two transformation experiments were significantly different from one another, this chemical has to be tested in a third trial. In the absence of these data 1-nitronaphthalene was evaluated as having had an indeterminate activity in the transformation assay, and its actual and estimated rank t-statistics were 1.45 and 1.56, respectively (Table A3). (Table A3). It had one difficult technical problem. It was reported to become oxidized upon exposure to air, and it was exposed to air during the treatment period ( Table Al). The test chemical was cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.292 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 99 and 46/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 82 and 80/110, respectively (Table A2). The test chemical had an SP transformation response in the first experiment, and an LA response in the second experiment. 4-Nitro-o-phenylenediamine was evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were 2.54 and 3.54, respectively (Table A3).
3-Nitropropionic Acid. 3-Nitropropionic acid is a level E noncarcinogen (Table A3). It had one difficult technical problem. It was reported to become oxidized upon exposure to air, and it was exposed to air during the treatment period ( Table Al). The test chemical was moderately cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 1.23 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 82 and 67/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 18 and 78/110, respectively (Table A2). The test chemical had an SP transformation response in the first and second experiments. 3-Nitropropionic acid was evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were 4.69 and 5.22, respectively (Table A3).
p-Phenylenediamine-2HCI. p-Phenylenediamine-2HCl is a level F noncarcinogen (Table A3). It had one difficult technical problem. It was reported to become oxidized upon exposure to air, and it was exposed to air during the treatment period ( Table Al). The test chemical was very cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.0712 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 58 and 51/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 55 and 50/110, respectively (Table A2). In trials 1 and 2 the test chemical had SP transformation responses. p-Phenylenediamine-2HCI was evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were both 3.04 (Table A3).
N-Phenyl-2-Naphthylamide. N-Phenyl-2-naphthylamide is a level F noncarcinogen (Table A3). It had one difficult technical problem. It was reported to become oxidized upon exposure to air, and it was exposed to air during the treatment period (Table Al). The test chemical was cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.195 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 93 and 63/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 100 and 101/110, respectively (Table A2). In trials 1 and 2 the test chemical had SP transformation responses. N-Phenyl-2-naphthylamide was evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were 4.25 and 6.90, respectively (Table A3).
The test chemical was very cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.0437 mM ( Table Al).
The statistical sensitivities of transformation assay trials 1 and 2 were 68 and 43/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 40 and 9/110, respectively (Table A2). In trials 1 and 2 the test chemical had LA transformation responses. 2,3,5,6,-Tetrachloro-4nitroanisole was evaluated as having equivocal activity in the transformation assay, and its actual and estimated rank t-statistics were both 2.08 (Table A3).
Tetraethylthiuram Disulfide. Tetraethylthiuram disulfide is a level F noncarcinogen (Table A3). It had one no insurmountable technical problems (Table Al). The test chemical was very cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.0000583 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 10 and 43/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 5 and 9/110, respectively (Table A2). In trials 1 and 2 the test chemical had SP transformation responses. Tetraethylthiuram disulfide was evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were both 3.88 (Table A3).
2,6-Toluenediamine-2HCl. 2,6-Toluenediamine-2HCl is a level F noncarcinogen (Table A3). It had no insurmountable technical problems ( Table Al). The test chemical was moderately cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 4.11 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 68 and 23/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 40 and 11/110, respectively (Table A2). In trials 1 and 2 the test chemical had SP transformation responses. 2,6-Toluenediamine-2HCI was evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were both 6.95 (Table A3).

Cytotoxic, Nonmutagenic Carcinogens
Allyl Isothicyanate. Allyl isothiocyanate is a level D carcinogen (Table A3). It had one difficult technical problem because it was reported to react with water ( Table Al). The test chemical was a very cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.00712 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 90 and 98/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 13 and 60/110, respectively (Table A2). The test chemical had an LA transformation response in the first experiment, and an SP response in the second experiment. Allyl isothiocyanate evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were 2.86 and 3.39, respectively (Table A3).
Allyl Isovalerate. Allyl isovalerate is a level A carcinogen (Table A3). It had no insurmountable technical problems (Table Al). The test chemical was moderately cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 4.51 mM ( Table Al). The statistical sensitivities of transformation assay trials 1-4 were 66, 78, 42, and 27/110, respectively; the detection sensitivities for BaP of trials 1-4 were 81, 58, 35, and 63/110, respectively (Table  A2). The test chemical had an LN transformation response in the first two experiments, an SN response in the third experiment, and an LA response in the fourth experiment. Allyl isovalerate was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were both 1.65 (Table A3).
Chlorendic Acid. Chlorendic acid was a level A carcinogen (Table A3). This chemical was reported to bind metal salts (Table Al); thus, it could have affected the concentration of metal salts in FBS and culture medium.
The test chemical was moderately cytotoxic chemical with an average LD50 of 4.07 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 9 and 57/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 37 and 88/110, respectively (Table A2). In a preliminary trial 1 the chemical had an SP transformation response. In trial 2 the chemical had an LA transformation response. Chlorendic acid was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were both 1.98 (Table A3).
3-Chloro-2-Methylpropene. 3-Chloro-2-methylpropene was a level A carcinogen (Table A3). It had many technical problems including its reported reaction ( Table Al). The test chemical was a cytotoxic chemical with an average LD50 of 0.662 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 9 and 57/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 37 and 88/110, respectively (Table A2). In a preliminary trial 1 the chemical had an LA transformation responses. In trial 2 the chemical had an SP transformation response. 3-Chloro-2-methylpropene was evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were both 2.14 ( Table A3).
Chlorowax 40. Chlorowax 40 is a level D carcinogen (Table A3). It had no insurmountable technical problems ( Table Al). The test chemical was moderately cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 1.43 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 12 and 26/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 98 and 93/110, respectively (Table A2). The test chemical had an LA transformation response in the first experiments and an SN response in the second experiment. Chlorowax 40 was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were 0.46 and 0.48, respectively (Table A3) Chlorowax 500. Chlorowax 500 is a level A carcinogen (Table A3). It had no insurmountable technical problems ( Table Al). The test chemical was moderately cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 1.58 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 32 and 15/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 108 and 21/110, respectively (Table A2). The test chemical had an SN transformation response in both experiments. Chlorowax 500 was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were both 0.06 (Table A3).
Cinnamyl Anthranilate. Cinnamyl anthranilate is a level A carcinogen (Table A3). It had one difficult technical problem. It was reported to become oxidized by air; thus, it could have reacted with air during the treatment period ( Table Al). The test chemical was very cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.0947 mM ( Table Al). The statistical sensitivities of transformation assay trials 1-3 were 53, 102 and 92/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 30, 95 and ND/110, respectively (Table A2). The test chemical had an SN transformation response in all three experiments. Cinnamyl anthranilate was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were 0.004 and 0.005, respectively (Table A3). Diethylstilbestrol. Diethylstilbestrol is a level A carcinogen (Table A3) with no technical problems (Table Al). It was very cytotoxic to the BALB/c-3T3 cells with an average LD50 of 0.0858 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 55 and 31/110, respectively; the detection sensitivities for BaP oftrials 1 and 2 were 8 and 52/110, respectively (Table  A2). In a preliminary trial 1 the chemical had an LA transformation response. In trial 2 the chemical had an SP transformation response. Diethylstilbestrol was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were both 2.91, respectively (Table A3).
Dimethylvinyl Chloride. Dimethylvinyl chloride is a level A carcinogen (Table A3). It had one serious technical problem because it was noted to be oxidized upon exposure to air (Table Al). It was moderately cytotoxic to the BALB/ c-3T3 cells with an average LD50 of 4.74 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 12 and 27/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 98 and 63/110, respectively (Table A2). In a preliminary trial 1 the chemical had an LA transformation response. In trial 2 the chemical had an SP transformation response. Dimethylvinyl chloride was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were both 2.59 (Table A3).
Ethyl Acrylate. Ethyl acrylate is a level A carcinogen (Table A3). It has one difficult technical problem because it was reported to react with water (Table Al). It was very cytotoxic to the BALB/c-3T3 cells with an average LD50 of 0.0746 mM ( Table Al). The statistical sensitivities oftransformation assay trials 1 and 2 were 66 and 74/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 81 and 35/110, respectively (Table A2). In a preliminary trial 1 the chemical had an LA transformation response. In trial 2 the chemical had an SP transformation response. Ethyl acrylate was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were 4.51 and 5.25, respectively (Table A3). Isophorone. Isophorone is a level D carcinogen (Table  A3) with no insurmountable technical problems (Table Al). It was moderately cytotoxic to the BALB/c-3T3 cells with an average LD50 of 5.18 mM ( Table Al). The statistical sensitivities of transformation assay trials 1-3 were 106, 74 and 26/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 51, 35 and 93/110, respectively (Table A2). In a preliminary trial 1 the chemical had an LN transformation response because the test chemical treatment doses were noncytotoxic to the cells. In trials 2 and 3 the chemical had an SP transformation response. Isophorone was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were 2.76 and 2.86, respectively (Table A3).
D-Limonene. D-Limonene is a level D carcinogen (Table  A3). It had one difficult technical problem. It was reported to become oxidized by air; thus, it could have reacted with air during the treatment period ( Table Al). The test chemical was cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.988 mM ( Table Al). The statistical sensitivities of transformation assay trials 1-3 were 70, 12 and 57/110, respectively; the detection sensitivities for BaP of trials 1-3 were 41, 98 and 88/110, respectively (Table A2). The test chemical had an LN transformation response in the first experiments and an SN response in the second and third experiments. D-Limonene was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were 0.24 and 0.27, respectively (Table A3).
Malonaldehyde, Sodium Salt. Malonaldehyde, sodium salt, is a level C carcinogen (Table A3). It was had one serious technical problem because it is very temperature sensitive ( Table 1). The test chemical was moderately cytotoxic to the BALB/c-3T3 cells with an average LD50 of 3.74 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 21 and 54/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 26 and 44/110, respectively (Table A2). In a preliminary trial 1 the chemical had an SP transformation response. In trial 2 the chemical had an LA transformation response. Malonaldehyde, sodium salt, was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were 1.87 and 1.87, respectively (Table A3).
Methapyrilene-HCI. Methapyrilene-HCl is a level C carcinogen (Table A3). It had no insurmountable technical problems ( Table Al). The test chemical was cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.812 mM ( Table Al). The statistical sensitivities of transformation assay trials 1-3 were 60, 91 and 82/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 27, 85 and ND/110, respectively (Table A2). The test chemical had an SN transformation response in all three experiments. Methapyrilene-HCL was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were 0.14 and 0.17, respectively (Table A3).
2-Mercaptobenzothiazole. 2-Mercaptobenzothiazole is a level C carcinogen (Table A3). It had one serious technical problem. It was reported to react with water; thus, its activity in the transformation assay could have been unavoidably affected by its exposure to an aqueous environment during the 48-hr treatment period ( Table Al).
The test chemical was cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.130 mM ( Table Al). The statistical sensitivities of transformation assay trials 1-3 were 3,17 and 51/110, respectively, the detection sensitivities for BaP of trials 1-3 were 110, 6 and 50/110, respectively (Table A2). The test chemical had a LN transformation response in the first experiment, because the positive control did not induce significant transformation in this experiment. In the second and third experiments the test chemical had LA transformation responses. 2-Mercaptobenzothiazole was evaluated as having had equivocal activity in the transformation assay, and its actual and estimated rank t-statistics were both 1.23 (Table A3).
Because this chemical was reported to binds metal salts, it could have affected the concentration ofmetal salts in FBS and culture medium (Table Al). It was moderately cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 5.98 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 80 and 33/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 89 and 14/110, respectively (Table A2). In trials 1 and 2 the chemical had an SP transformation response. Nitrilotriacetic acid, trisodium salt, was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were both 5.86, respectively (Table  A3).
Polybrominated Biphenyl Mixture. Polybrominated biphenyl mixture is a level A carcinogen (Table A3). This test chemical was insoluble in culture medium at a portion of the treatment doses that were used to induce both cytotoxic and transforming activity (Table Al). It was cytotoxic to the BALB/c-3T3 cells with an average LD50 of 0.291 mM ( Table Al). The statlstical sensitivities of transformation assay trials 1 and 2 were 81 and 39/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 22 and 68/110, respectively (Table A2). In trials 1 and 2 the chemical had an SP transformation responses for both type III and type I-III focus transformation responses. However, this test chemical was very unusual in that it induced proportionally a much higher response for the type I and II foci, than for the type III foci (refer to the Discussion). Polybrominated biphenyl mixture was evaluated as active in the transformation assay. Its actual and estimated rank t-statistics were both 1.93 for the type III focus response, and 5.81 for the type I and II focus response (Table A3).
Reserpine. Reserpine is a level A carcinogen (Table A3). It had one difficult technical problem. It was reported to become oxidized by air; thus, it could have reacted with air during the treatment period ( Table Al). The test chemical was very cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.0133 mM ( Table Al). The statistical sensitivities of transformation assay trials 1-3 were 25, 16 and 92/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 36 and 7/110, respectively (Table A2). The test chemical had an LA transformation response in the first experiment and an SN response in the second and third experiments. Reserpine was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were both 0.24, respectively (Table A3). 71is(2-ethylhexyl)phosphate. Tris(2-ethylhexyl)phosphate is a level D carcinogen (Table A3). It had no insurmountable technical problems ( Table Al). The test chemical was cytotoxic to the BALB/c3T3 cells and had an average LD50 of 0.338 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 57 and 59/110, respectively, the detection sensitivities for BaP of trials 1 and 2 were 88 and 17/110, respectively (Table A2). The test chemical had an SN transformation response in both experiments. Tris(2-ethylhexyl)phosphate was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were both 0.00 (Table A31 4-Vinylcyclohexene. 4-Vinylcyclohexene is a level D carcinogen (Table A3). It had one difficult technical problem. It was reported to become oxidized upon exposure to air; thus, it could have reacted with air during the treatment period ( Table Al). The test chemical was moderately cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 3.88 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 32 and 38/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 108 and 16/110, respectively (Table A2). The test chemical had an SN transformation response in both experiments. Vinylcyclohexane was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were both 0.00 (Table A3).

Cytotoxic, Nonmutagenic Noncarcinogens
Anilazine. Anilazine is a level F noncarcinogen (Table  A3). It had no insurmountable technical problems ( Table  1). The test chemical was a very cytotoxic to the BALB/ c-3T3 cells and had an average LD50 of 0.0475 (Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 68 and 67/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 40 and 78/110, respectively (Table A2). In trials 1 and 2 the test chemical had LA transformation responses. Aniline was evaluated as having equivocal activity in the transformation assay, and its actual and estimated rank t-statistics were 1.82 and 2.09, respectively (Table A3).
L-Ascorbic Acid. L-Ascorbic acid is a level F noncarcinogen (Table A3). It had two difficult technical problems. It was reported to become oxidized upon exposure to air; thus, it could have reacted with air during the treatment period. In addition, it was noted to bind metal salts; thus, it could have combined with metal salts in FBS and EMEM medium ( Table Al). The test chemical was cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.363 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 50 and 101/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 59 and 104/110, respectively (Table A2). In trials 1 and 2 the test chemical had SN transformation responses. L-Ascorbic acid was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were 0.32 and 0.46, respectively (Table A3).
Bisphenol A. Bisphenol A is a level E noncarcinogen (Table A3). It had no insurmountable technical problems ( Table Al). The test chemical was cytotoxic to the BALB/ c-3T3 cells and had an average LD50 of 0.147 mM ( Table  Al). The statistical sensitivities of transformation assay trials 1-4 were 53, 16, 99 and 105/110, respectively; the detection sensitivities for BaP of trials 1-4 were 30, 7, 37 and 39/110, respectively (Table A2). The test chemical had an SN transformation response in all four experiments. Bisphenol A was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were both 0.79 (Table A3).
Carbromal. Carbromal is a level F noncarcinogen (Table  A3). It had one diffiecult technical problem. It was reported to become oxidized by air; thus, it could have reacted with air during the treatment period (Table Al). The test chemical was moderately cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 3.60 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 14 and 23/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 72 and 11/110, respectively (Table A2). The test chemical had an SN transformation response in the first experiment, and an SP response in the second experiment. Although the test chemical had disparate transformation responses in two experiments, the mean t-statistics of the two responses were not significantly different from one another. Carbromal was evaluated as weakly active in the transformation assay, and its actual and estimated rank t-statistics were both 2.26 (Table A3).
Chlorpheniramine-Maleate. Chlorpheniramine maleate is a level F noncarcinogen (Table A3). It had no insurmountable technical problems ( Table Al). The test chemical was cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.287 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 47 and 67/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 43 and 78/110, respectively (Table A2). The test chemical had an SN transformation response in the first experiment, and an SP response in the second experiment. Although the test chemical had disparate transformation responses, the mean t-statistics of the two responses were not significantly different from one another. Chlorpheniramine-maleate was evaluated as weakly active in the transformation assay, and its actual and estimated rank t-statistics were 1.18 and 1.26, respectively (Table A3).
C. L Acid Red 14. C. I. Acid red 14 is a level F noncarcinogen (Table A3). It had no insurmountable technical problems ( Table Al). The test chemical was moderately cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 3.38 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 40 and 35/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 54 and 57/110, respectively (Table A2). In trials 1 and 2 the test chemical had SP transformation responses. C. I. Acid red 14 was evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were 60th 2.15 (Table A3).
C. L Acid Yellow 73. C. I. Acid yellow 73 is a level F(I) noncarcinogen which has been reclassified as an incomplete bioassay study (Table A3). It had no insurmountable technical problems ( Table Al). The test chemical was moderately cytotoxic to the BALB/c-3T3 cells and had an average LD50 of4.65 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 17 and 71/110, respectively; the detection sensitivities for BaP oftrials 1 and 2 were 6 and 23/110, respectively (Table  A2). In trials 1 and 2 the test chemical had SN transformation responses. C. I. Acid yellow 73 was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were both 0.03 (Table A3).
Ephedrine Sulfate. Ephedrine sulfate is a level F noncarcinogen (Table A3). It had no insurmountable technical problems ( Table Al). The test chemical was moderately cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 1.53 mM ( Table Al). The statistical sensitivities of transformation assay trials 1-3 were 24, 17 and 51/110, respectively; the detection sensitivities for BaP of trials 1-3 were 10,6 and 50/110, respectively (Table A2). The test chemical had an LN transformation response in the firsl experiment because the test chemical did not have significant cytotoxic activity. The test chemical had an SN response in the second and third experiments. Ephedrine sulfate evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were both 0.16 (Table A3).
Erythromycin Stearate. Erythromycin stearate is a level F noncarcinogen (Table A3). It had no insurmountable technical problems ( Table Al). The test chemical was very cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.0746 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 24 and 51/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 10 and 50/110, respectively (Table A2). The test chemical had an SN transformation response in both the first and second experiments. Erythromycin stearate was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were both 0.08 (Table A3).
Ethoxylated Dodecyl Alcohol. Ethoxylated dodecyl alcohol is a level F(I) noncarcinogen which has been reclassified as an incomplete study (Table A3). It had no insurmountable technical problems (Tlable Al). The test chemical was very cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.0172 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 1 and 15/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 3 and 21/110, respectively (Table A2). The test chemical had an SN transformation response in both the first and second experiments. Ethoxylated dodecyl alcohol was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were both 0.29 (Table A3).
Ethylenediamine Tetraacetic Acid, 11risodium Salt. Ethylenediamine tetraacetic acid, trisodium salt, is a level F noncarcinogen (Table A3). It had one difficult technical problem. It was reported to bind with certain metal salts; thus, it could have reacted with metal salts in FBS and EMEM medium ( Table Al). The test chemical was moderately cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 1.89 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 80 and 67/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 89 and 78/110, respectively (Table A2). The test chemical had an LA transformation response in the first experiment, and an SN response in the second experiment. Ethylenediamine tetraacetic acid, trisodium salt, was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were 1.41 and 2.01, respectively (Table A3). Eugenol. Eugenol is a level E noncarcinogen (Table A3). It had one difficult technical problem. It was reported to become oxidized upon exposure to air; thus, it could have reacted with air during the treatment period ( Table Al).
The test chemical was cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.875 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 32 and 18/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 108 and 31/110, respectively (Table A2). The test chemical had an LN transformation response in the first experiment, because the test chemical did not induce significant cytotoxic activity. In the second experiment the test chemical had an SN transformation response. Since the test chemical had an LN transformation response in the first experiment, it had to be tested in two additional trials. Therefore, eugenol was evaluated as having had an indeterminate activity in the transformation assay, and its actual and estimated rank t-statistics were both 1.07 (Table A3).
Geranyl Acetate. Geranyl acetate is a level F noncarcinogen (Table A3). It had no insurmountable technical problems ( Table Al). The test chemical was cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.302 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 44 and 41/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 65 and 83/110, respectively (Table A2). The test chemical had an SN transformation response in the first and second experiments. Geranyl acetate was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were 0.35 and 0.37, respectively (Table  A3).
4-Hexylresorcinol. 4-Hexylresorcinol is a level E noncarcinogen (Table A3). It had no insurmountable technical problems ( Table Al). The test chemical was cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.103 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 13 and 67/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 66 and 78/110, respectively (Table A2). The test chemical had an SN transformation response in the first experiment, and an LA response in the second experiment. 4-Hexylresorcinol was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were 0.46 and 0.47, respectively (Table A3). D,L-Menthol. D,L-Menthol is a level F noncarcinogen (Table A3). It had no insurmountable technical problems ( Table Al). The test chemical was moderately cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 4.63 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 46 and 89/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 80 and 87/110, respectively (Table A2). The test chemical had an SN transformation response in the first experiment and second experiments. D,L-Menthol was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were 0.16 and 0.22, respectively (Table A3).
Methoxychlor. Methoxychlor is a level F noncarcinogen (Table A3). It had no insurmountable technical problems ( Table Al). The test chemical was very cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.0978 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 58 and 51/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 55 and 50/110, respectively (Table A2). The test chemical had an SN transformation response in the first experiment, and an LA response in the second experiment. Methoxychlor was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were both 1.08 (Table A3).
Methyldopa Sesquihydrate. Methyldopa sesquihydrate is a level E noncarcinogen (Table A3). It had no insurmountable technical problems ( Table Al). The test chemical was very cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.0810 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 21 and 56/110, respectively, the detection sensitivities for BaP of trials 1 and 2 were 26 and 109/110, respectively (Table A2). The test chemical had an LA transformation response in the first experiment, and an SN response in the second experiment. Methyldopa sesquihydrate was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were both 1.21, respectively (Table A3).
Methylphenidate-HCI. Methylphenidate is a level I noncarcinogen because it has not been evaluated in a complete rodent bioassay study (Table A3). It had no insurmountable technical problems ( Table Al). The test chemical was moderately cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 5.63 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 64 and 86/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 84 and 74/110, respectively (Table A2). In trials 1 and 2 the test chemical had SN transformation responses. C. I. Acid red 14 was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were 1.05 and 1.47, respectively (Table A3).
Oxytetracycline-HCI. Oxytetracycline is a level E noncarcinogen (Table A3). It had one difficult technical problem. It was reported to hydrolyze in water; thus, it could have reacted with water during the treatment period ( Table Al). The test chemical was cytotoxic to the BALB/ c-3T3 cells and had an average LD50 of 0.523 mM ( Table  Al). The statistical sensitivities of transformation assay trials 1-3 were 80, 22 and 5/110, respectively; the detection sensitivities for BaP of trials 1-3 were 89, 53 and 46/110, respectively (Table A2). The test chemical had an SN transformation response in all three experiments. The test chemical was tested in the third experiment, because the its cytotoxic activity in the first experiment was excessive. Oxytetracycline-HCI was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were both 0.00, respectively (Table A3). Phenol. Phenol is a level F noncarcinogen (Table A3). It had two difficult technical problems. It was reported to become oxidized by air; thus, it could have reacted with air during the treatment period. In addition, it was reported to react with sulfate groups on chemicals; thus, it could have reacted with biochemicals in culture medium, as well as biochemicals in the target cells (Table Al). The test chemical was moderately cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 3.29 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 12 and 15/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 26 and 21/110, respectively (Table A2). The test chemical had an SP transformation response in the first and second experiments. Phenol was evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were both 7.60 (Table A3).
Phenylephrine-HCI. Phenylephrine-HCl is a level F noncaranogen (Table A3). It had one difficult technical problem. It was reported to become oxidized by air; thus, it could have reacted with air during the treatment period ( Table Al). The test chemical was moderately cytotoxic to the BALB/c3T3 cells and had an average LD50 of 3.52 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 80 and 29/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 89 and 97/110, respectively (Table A2). The test chemical had an SN transformation response in the first and second experiments. Phenylephrine-HCl was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were 0.31 and 0.42, respectively (Table A3).
Propyl Gallate. Propyl gallate is a level E noncarcinogen (Table A3). It had one difficult technical problem. It was reported to react with iron; thus, it could have reacted with the iron in FBS (Table Al). The test chemical was very cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.0631 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 94 and 102/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 91 and 95/110, respectively (Table A2). The test chemical had an SP transformation response in the first experiment, and an LA response in the second experiment. Propyl gallate was evaluated as active in the transformation assay. Because the statistical sensitivity and detection sensitivity for BaP in both of the experiments were significantly low, and the test chemicals actual and estimated rank t-statistics were very different (i.e., 1.70 and 2.95, respectively (Table A3). Rotenone. Rotenone is a level E noncarcinogen (Table  A3). It had one difficult technical problem. It was reported to become oxidized by air; thus, it could have reacted with air during the treatment period ( Table Al). The test chemical was very cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.000464 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 21 and 31/110, respectively, the detection sensitivities for BaP of trials 1 and 2 were 26 and 52/110, respectively (Table A2). The test chemical had an SN transformation response in the first and second experiments. Rotenone was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were both 0.26 (Table A3). Sodium Diethyldithiocarbamate. Sodium diethyldithiocarbamate is a level F noncarcinogen (Table A3). It had no insurmountable technical problems ( Table Al). The test chemical was very cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.000142 mM ( Table Al). The statistical sensilivities of transformation assay trials 1 and 2 were 69 and 31/110, respectively; the detection sen-sitivities for BaP of trials 1 and 2 were 12 and 52/110, respectively (Table A2). The test chemical had an SP transformation response in the first and second experiments. Sodium diethyldithiocarbamate was evaluated as active in the transforrnation assay, and its actual and estimated rank t-statistics were both 2.94 (Table A3).
Stannous Chloride. Stannous chloride is a level E noncarcinogen (Table A3). It had two difficult technical problems. It was reported to react with both alcohols and amines; thus, it could have reacted with biochemicals with this groups in both FBS and EMEM medium ( Table Al). The test chemical was very cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.0285 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 77 and 37/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 67 and 20/110, respectively (Table A2). The test chemical had an SN transformation response in the first and second experiments. Stannous chloride was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were both 0.78 (Table A3).
Tetracycline-HCI. Tetracycline-HCl is a level F noncarcinogen (Table A3). It had one difficult technical problem, because it was reported to become oxidized upon exposure to air; thus, it could have reacted with air during the treatment period (Table Al). The test chemical was moderately cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 3.24 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 24 and 51/110, respectively, the detection sensitivities for BaP of trials 1 and 2 were 10 and 50/110, respectively (Table A2). The test chemical had an SN transformation response in the first experiment and second experiments.
Tetracycline-HCl was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were both 0.021 (Table A3).
Tetrakis (hydroxymethyl)phosphonium Chloride. Tetrakis(hydrosymethyl)-phosphonium chloride is a level F noncarcinogen (Table A3). It had no insurmountable technical problems (Table Al). The test chemical was very cytotoxic to the BALB/c-3T3 cells and had an average LD50 of0.00825 mM ( Table Al). The slatistical sensitivities of transformation assay trials 1-3 were 70, 15 and 59/110, respectively; the detection sensitivities for BaP of trials 1-3 were 41, 21 and 17/110, respectively (Table A2). The test chemical had an SN transformation response in the first experiment, and an SP response in the second experiment. Because the mean t-statistics of the test chemical transformation responses in the first two experiments were significantly different from one another, the test chemical had to be tested in a third trial. The test chemical had a SN transformation response in the third experiment. Tetrakis(hydroxymethyl)phosphonium chloride was evaluated as having had equivocal activity in the transformation assay, and its actual and estimated rank t-statistics were both 1.27 (Table A3).
Tetrakis(hydroxymethyl)phosphonium Sulfate. Tetrakis(hydroxymethyl)phosphonium sulfate is a level F noncarcinogen (Table A3). It had no insurmountable technical problems ( Table Al). The test chemical was very cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.00438 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 70 and 44/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 41 and 65/110, respectively (Table A2). The test chemical had an SN transformation response in the first and second experiments. Tetrakis-(hydroxymethyl)phosphonium sulfate evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were both 0.083 (Table A3).
Triphenyltin Hydroxide. Triphenyltin hydroxide is a level F noncarcinogen (Table A3). It had no insurmountable technical problems ( Table Al). The test chemical was very cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 0.000134 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 82 and 43/110, respectively; the detection sensitivities for BaP oftrials 1 and 2 were 18 and 9/110, respectively (Table  A2). The test chemical had an SN transformation response in the first experiment, and an SP response in the second experiment. Since the mean t-statistics ofthe test chemical transformation responses in the first two experiments were significantly different from one another, it had to be tested in a third experiment. Therefore, triphenyltin hydroxide was evaluated as having had an indeterminate activity in the transformation assay, and its actual and estimated rank t-statistics were both 1.64 (Table A3).
Xylenes, (mixed). Xylenes (mixed) is a level F noncarcinogen (Table A3). It had no insurmountable technical problems ( Table Al). The test chemical was moderately cytotoxic to the BALB/c-3T3 cells and had an average LD50 of 3.20 mM ( Table Al). The statistical sensitivities of transformation assay trials 1 and 2 were 70 and 73/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 41 and 49/110, respectively (Table A2). The test chemical had an SN transformation response in the first and second experiments. Xylenes (mixed) was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were 0.47 and 0.50, respectively (Table A3).
Noncytotoxic, Mutagenic Carcinogens DC Red No. 9. DC Red no. 9 is a level B carcinogen (Table A6). It had no insurmountable technical problems; however it had a solubility limit in culture medium of about 500 ,ug/ml. This improved to 2250 pLg/ml when the medium was supplemented with the solvent vehicle pluronic F68 (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 3.82 mM and 6.52 mM, either with or without using pluronic F68 (Table A4). The statistical sensitivities of transformation assay trials 1-3 were 30, 91 and 107/110, respectively; the detection sensitivities for BaP of trials 1-3 were 4, 85 and 106/110, respectively (Table A5). The test chemical had an SP transformation response in two experiments, and an LA response in one experiment. Significant test chemical transforming activity was detected at doses both above and below its solubility limit in culture medium. DC Red No. 9 was evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were 4.33 and 5.50, respectively (Table A6). Diethanolnitrosamine. Diethanolnitrosamine is a level I carcinogen (Table A6). It had no insurmountable technical problems (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 61.1 mM (Table A4). The test chemical was only evaluated in one trial due to the limited availability during this investigation. The statistical sensitivity of transformation assay trial 1 was 52/110; the detection sensitivity for BaP of trial 1 was 79/110, respectively (Table A5). The test chemical had an SP transformation response in the only experiment conducted for this test chemical. Diethanolnitrosamine was evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were 4.01 and 4.87, respectively (Table A6). Diethylnitrosamine. Diethylnitrosamine is a level A carcinogen (Table A6). It had no insurmountable technical problems (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 46.0 mM (Table A4). The statistical sensitivities of transformation assay trials 1 and 2 were 4 and 27/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 1 and 63/110, respectively (Table A5). The test chemical had an SP transformation response in the two consecutive experiments. Diethylnitrosamine was evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were both 4.69 (Table A6).  [2] solvent vehicle (D = dimethyl sulfoxide, C = culture medium, F = pluronic F68, A = acetone, E = ethanol) and [3] technical problems. The technical problems included test chemicals that were a = reacts with acids; ac = reacts with acid chlorides and acid anhydrides; ai = reacts with air, al = reacts with alcohols; alk = alkylating agent and reacts with labile hydrogen; b = reacts with bases; bc = reacts with biochemicals (amino, hydroxyl, and carboxyl groups); hc = reacts with halogenated chemicals; k = reacts with alpha keto acids; ls = light sensitive; m = binds metals; mel = reacts with hexachloroand trichloromelamine; met = reacts with metals (aluminum, iron, magnesium, potassium, sodium, tin or zinc); mh = metal halides; msc = reacts with miscellaneous organic chemicals (i.e. alpha-aminoethanol, chlorosulfonic acid, ethylene imine, linseed oil, maleic anhydride, oleum, or K-tert-butyloxide); o = reacts with oxidizing agents; p = reacts with plastics; pi = polymerization initiators; r = reacts with reducing agents; ru = reacts with rubber; sp = solubility problem in culture medium; te = reacts with thiocyanates; ts = temperature sensitive; vts = very temperature sensitive; v = volatile at 37,C; and w = reacts with water [refer to MATERIALS and METHODS].
aTest Chemical: Tables Al and A3 contain 168 chemicals along with their individual CAS registry number and molecular weight. The chemicals were divided into groups of chemicals that correspond to the groups of chemicals that were compared in different text Tables 1-12. Thus, the chemicals were divided into two groups, including 114 cytotoxic test chemicals (LD50 < 5.0 mM) presented in Table Al, and 53 noncytotoxic chemicals (LD50 >5.0 mM presented in Table A4. The 114 cytotoxic chemicals in Table Al  bCytotoxic Response: The co-culture clonal sulvival assay design used to detect the cytotoxic response of the test chemical is described in Materials and Methods. The cytotoxic responses of chemicals in individual experiments are summarized in terms ofthe millimolar (mM) LD50 treatment dose that resulted in 50% survival of the chemically-treated cells relative to the survival of untreated or solvent control treated cell cultures. The LD50 cytotoxic response of each chemical in Tables Al and A4 is an average of two or more experiments with the chemical. The molecular weight of each chemical is provided in order that treatment doses could be converted from mM to ,ug/ml. For example, based upon the molecular weight of 84.08, the LD50 detected for the first chemical in Table A4, 3-amino-1,2,4-triazole, was 109 mM or 9165 ,ug/ml.  cSpontaneous Transformation Response: The method used to caIculate the spontaneous transformation response, as well as the positive control and test chemical responses, is explained in the Materials and Methods. The transformation responses are expressed as type III foci/vessel and were calculated using a loglo mathematical transformation procedure. The arithmetic value for foci/vessel in this table is the antilog of the log1o mean transformation response minus one.
The procedure for rank ordering the spontaneous responses from 110 experiments is based upon the different statistical sensitivities of transformation experirnents with different spontaneous responses is explained in the Statistical Sensitivity versus Spontaneous Transformation Response section of the Materials and Methods. Experiments with high spontaneous responses had a high statistical sensitivity and have relatively low rank-order numbers. For example, diethylnitrosamine had a high spontaneous response of 5.12 foci/vessel in exp. 79, which had a high statistical sensitivity and rank order number 4/110. Conversely, experiments with a low statistical sensitivity and have high rank-order numbers. For example, 11aminoundecanoic acid had a low spontaneous response of .085 foci/vessel in exp. 67, which had a low statistical sensitivity with a high rank order number 107/110. dBenzo(a)pyrene Transformation Response. The method used to call individual transformation experiments is described in detail in Materials and Methods. The method used to rank order the BaP transformation responses from the 110 experiments is based upon statistical comparison of the BaP transformation at the two treatment doses detected in an individual experiment with the median historical activity of the assay. This procedure is described in the Detection Sensitivity versus Benzo(a)pyrene Transformation Response section of the Materials and Methods. The rational for rankordering the experiments is analogous to that described for the spontaneous transformation responses (refer to footnote c above).
eTest Chemical Transformation Response: The method used to call individual experiments is described in detail in Materials and Methods, and the abbreviations for the calls are provided in footnote d above. The significance of the transformation responses of individual chemical treatment doses were calculated using SAS statistical software (22). The mean t-statistic represents the average of the t-statistics of the four test chemical treatment doses in the experiment. The t-statistics for individual chemical treatment doses which were used to calculate the mean t-statistic are provided in Appendices B-H. *Significant spontaneous or BaP transformation response, 0.01 < p < 0.05. **Significant spontaneous or BaP transformation response, 0.001 < p < 0.01. 'Significant spontaneous or BaP transformation response, p < 0.001.    Table A6 are idenical to those in Table A4, and they are subdivided into groups of 21 carcinogens, 26 noncarcinogens and 7 model very noncytotoxic chemicals.
i'Rodent Bioassay Level of Activity: The relative carcinogenic activity of chemicals in rodent bioassay has been described in terms of the chemical's level of effect (1,.). The highest level A corresponds to chemicals that cause cancer in both mice and rats at one or more sites, and level B refers to chemicals that cause cancer at multiple sites in one species of rodent. Level C includes chemicals carcinogenic at one site in both sexes of one species, and D includes chemicals carcinogenic at one site in only one sex of a single species. Level E includes chemicals that only equivocal evidence of carcinogenic activity. Finally, level F includes both noncarcinogens and chemicals that had inadequate carcinogenicity studies.
'Transformation Response Rank t-statistic: The method used to calculate the significance of test chemical transformation responses employed SAS statistical software (22) and is described in detail in Materials and Methods. The correct t-statistics of each treatment dose of the test chemical in a single experiment are presented in the Appendices B-H, and these t-statistics were averaged to determine the mean t-statistic of the test chemical for the experiment (refer to Table A2). The mean t-statistics for two or experiments for each chemical was weighted according to the number of treatment doses evaluated and averaged to determine the actual rank t-statistic presented in this table. For example, the actual rank t-statistic of D&C ried no. 9 in experiments 43, 54 and 67 is equal to 4.33 [i.e., 10.9 + 10.5 + 7.16 + 5.26 (exp. 43) + 2.96 + 1.39 + .04 + .00 (Exp. 54) + 5.47 + 3.37 + 3.07 + 1.78 (exp. 67); Appendix F]. d'Estimated Rank t-statistic: The estimated riank t-statistic is used to estimate both the historical behavior of the test chemical in the transformation assay, as well as predicting the future behavior or the chemical. It is calculated by correcting the actual rank t-statistic. The data presented in Table A5 showed that individual experiments had very different rank-ordered sensitivities to detect chemical-induced transformation. Therefore, the estimated rank t-statistic modified the actual rank t-statistic to correct for differences in the sensitivities of individual experiments. The method uses the rank ordered sensitivity of individual experiments to detect spontaneous and BaP-induced transformation, and an example calculation is provided below.
Thus, these two experiments had a combined statistical sensitivity and detection sensitivity that was above the median of 55.0. The actual rank t-statistic was multiplied by the corirection factor to obtain the estimated rank t-statistic (i.e., 5.50). A justification for this correction factoi has been reported (18), and it is explained in the Materials and Methods. ePercentage (%) of Active Chemicals: Active chemicals included chemicals with active and weakly active transformation responses. In contrast, inactive chemicals included chemicals with equivocal and inactive transformation responses. Chemicals with an indeterminate activity have to be retested in an addlitional experiment in order to determine their activity in the standard transformation assay. Therefore, chemicals with indeterminate transformation responses were omitted from the computation of the percentage (%) of the total chemicals that were either active or inactive in the assay. Dimethylnitrosamine. Dimethylnitrosamine is a level A carcinogen (Table A6). It had one difficult technical problem. It was reported to be oxidized upon exposure to air, and it was exposed to air during the standard treatment period (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 256 mM (Table A4). The statistical sensitivities of transformation assay trials 1 and 2 were 42 and 73/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 25 and 49/110, respectively (Table  A5). The test chemical had an SP transformation response in the two consecutive experiments. Dimethylnitrosamine was evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were both 4.45 (Table A6).
Phenobarbital, Sodium Salt. Phenobarbital, sodium salt, is a level A carcinogen (Table A6). It had one difficult technical problem. It was reported to be oxidized by air, and it was exposed to air during the standard treatment period (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 6.11 mM (Table A4). The statistical sensitivities of transformation assay trials 1 and 2 were 75 and 6/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 76 and 92/110, respectively (Table A5). The test chemical had an SP transformation response in the two consecutive experiments. Phenobarbital was evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were 2.77 and 3.14, respectively (Table A6).
2,4-and 2,6-Tbluene Diisothiocyanate. 2,4-and 2,6-Toluene diisothiocyanate is a level A carcinogen (Table A6). It had three difficult technical problems. It was reported to react with strong bases such as NaOH, and stock solutions were acidic and had to be neutralized before testing. It was also reported to react with water, and treatments were performed in an aqueous environment. Finally, it was noted to react with amines; thus, it could have reacted with the amine portion of biochemicals in culture medium, as well as in the target BALB/c-3T3 cells (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 7.93 mM (Table A4). The statistical sensitivities of transformation assay trials 1 and 2 were 12 and 28/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 98 and 15/110, respectively (Table A5). The test chemical had an SP transformation response in the two consecutive experiments. 2,4 and 2,6-Toluene diisothiocyanate was evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were both 3.01 (Table A6).
Noncytotoxic, Nonmutagenic Carcinogens 3-Amino-1,2,4-Triazole. 3-Amino-1,2,4-triazole is a level A carcinogen (Table A6). It had one difficult technical problem. It was reported to bind metals; thus, it could have bound metals contained in FBS and EMEM medium (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 109 mM (Table A4). The statistical sensitivities of transformation assay trials 1 and 2 were 85 and 33/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 102 and 14/110, respectively (Table A5). The test chemical had an SP transformation response in the two consecutive experiments. 3-Amino-1,2,4-triazole was evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were 6.14 and 6.53, respectively (Table A6).
Il-Aminoundecanoic Acid. 11-Aminoundecanoic acid is a level B carcinogen (Table A6). It had no insurmountable technical problems; however, it had a limited solubility of 1500 ,ug/ml in culture medium supplemented with pluronic F68 (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 19.4 mM ( Table A4). The statistical sensitivities of transformation assay trials 1-4 were 83, 89,19 and 107/110, respectively; the detection sensitivities for BaP of trials 1-4 were 70, 87, 34 and 106/110, respectively (Table  A5). The test chemical had a LN transformation response in the first two experiments because it was not tested at cytotoxic treatment doses. In contrast, it had an SP transformation response in the last two experiments. Significant transforming activity was only detected at treatment doses that were slightly above the solubility limit of the test chemical in culture medium. 11-Aminoundecanoic acid was evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were 2.11 and 2.55, respectively (Table A6).
Cyclamate, Sodium Salt. Cyclamate, sodium salt, is a level A carcinogen (Table A6). It had one difficult technical problem. It was reported to bind potassium salts; thus, it could have bound the potassium in culture medium (Table  A4). The test chemical was noncytotoxic to the BALB/ c-3T3 cells and had an average LD50 of 132 mM (Table A4). The statistical sensitivities of transformation assay trials 1 and 2 were 24 and 5/110, respectively, the detection sensitivities for BaP of trials 1 and 2 were 10 and 46/110, respectively (Table A5). The test chemical had an SP transformation response in the two consecutive experiments. Cyclamate was evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were both 4.09 (Table A6). Decabromodiphenyloaxide. Decabromodiphenyloxide is a level C carcinogen (Table A6). It had one difficult technical problem. It had a limited solubility in culture medium of 250 ,ug/ml (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 6.26 mM (Table A4). Thus, the test chemical LD50 was about 24-fold higher than its solubility limit in culture medium. The statistical sensitivities of transformation assay trials 1 and 2 were 21 and 62/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 26 and 48/110, respectively (Table A5). The test chemical had an LN transformation response in the two consecutive experiments. It was tested at treatment doses that far exceeded its solubility limit in culture medium, but these doses were not cytotoxic to the target cells. Taken together, decabromodiphenyloxide was evaluated as both inactive and indeterminate in the transformation assay, and its actual and estimated rank t-statistics were both 0.29 (Table A6).
Di(2-ethylhexyl)adipate. Di(2-ethylhexyl)adipate is a level C carcinogen (Table A6). It had one difficult technical problem. Its solubility limit in culture medium supplemented with pluronic F68 was only 1000 nl/ml (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 98.4 mM (Table A4). Thus, this LD50 far exceeded the solubility limit of the test chemical. The statistical sensitivities of transformation assay trials 1 and 2 were 57 and 20/110 respectively; the detection sensitivities for BaP of trials 1 and 2 were 88 and 19/110, respectively (Table A5). The test chemical had an SN transformation response in the two consecutive experiments. The test chemical was tested using treatment doses that were both above and below its solubility limit. Di(2-ethylhexyl)adipate was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were both 0.00 (Table A6).
Di(2-ethylhexyl)phthalate. Di(2-elhylhexyl)phthalate is a level A carcinogen (Table A6). It had no insurmountable technical problems, and its solubility limit in culture medium supplemented with pluronic F68 was 12000 nl/ml (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 21.4 mM ( Table A4). The statistical sensitivities of transformation assay trials 1 and 2 were 74 and 73/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 35 and 49/110, respectively (Table A5). The test chemical had an SN transformation response in the two consecutive experiments. Di(2-ethylhexyl)phthalate was evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were both 0.00 (Table A6). Dimethyl Hydrogen Phosphite. Dimethyl hydrogen phosphite is a level B carcinogen (Table A6). It had two difficult technical problems. It was a very acidic test chemical, and stock solutions had to be neutralized with NaOH. Unfortunately, the test chemical was reported to react with strong bases and with water; thus, it could have been altered during the treatment period (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 130 mM (Table A4). Since it required an equal molar concentration of NaOH to neutralize the test chemical, this LD50 was actually equal to 260 mOsM. The statistical sensitivities of transformation assay trials 1 and 2 were 44 and 26/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 65 and 93/110, respectively (Table A5). The test chemical had an SP transformation response in the two consecutive experiments; however,' significant transforming activity was detected at treatment doses that exceeded the upper dose limit of the assay of 100 mOsM. Taken together, dimethyl hydrogen phosphite was evaluated as active and a false positive in the transformation assay. Its actual and estimated rank t-statistics were 5.97 and 6.19, respectively (Table A6).
Dimethyl Methyl Phosphonate. Dimethyl methyl phosphonate is a level D carcinogen (Table A6). It had two difficult technical problems. It was reported to react with air, and it was exposed to air during the treatment period. In addition, it was noted to be an alkylating agent and reacted with basic nitrogen compounds. Thus, this test chemical could have reacted not only with biochemicals in culture medium, but also with biochemicals in the target cells (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 172 mM (Table A4). The statistical sensitivities of transformation assay trials 1 and 2 were 44 and 27/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 65 and 63/110, respectively (Table A5). The test chemical had an LA transformation response in the first experiment, and an SP transformation response in the second experiment. Significant transforming activity for this test chemical was detected using treatment doses that exceeded the upper dose limit of the assay of 100 mOsM. Taken together, dimethyl methyl phosphonate was evaluated as active and a false positive in the transformation assay. Its actual and estimated rank t-statistics were both 2.90 (Table A6).
Dimethylmorpholinophosphoramidate. Dimethylmorpholinophosphoramidate is one of five chemicals that was tested as a coded and as a uncoded test chemical in this investigation. It is a level C carcinogen (Table A6), and it had no insurmountable technical problems (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells, and the uncoded and coded test chemicals had an average LD50 of 17.1 and 24.4 mM, respectively (Table A4). For the uncoded test chemical the statistical sensitivities of transformation assay trials 1 and 2 were 52 and 28/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 79 and 15/110, respectively (Table A5). For the coded test chemical the statistical sensitivities of transformation assay trials 1 and 2 were 52 and 20/110; the detection sensitivities for BaP of trials 1 and 2 were 79 and 19/110, respectively. Both the uncoded and coded test chemical had an SP transformation response in the two consecutive experiments. Dimethylmorpholinophosphoramidate was evaluated as active in the transformation assay. The actual and estimated rank t-statistics for the uncoded test chemical were both 3.19; the actual and estimated rank t-statistics for the coded test chemical were both 4.23 (Table A6). Taken together, the coded and uncoded test chemicals had virtually identical cytotoxic and transforming activities in the BALB/c-3T3 cell transformation assay.
Ethylene Thiourea. Ethylene thiourea is a level A carcinogen (Table A6). It had no insurmountable technical problems (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 91.4 mM ( Table A4). The statistical sensitivities of transformation assay trials 1 and 2 were 75 and 95/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 76 and 61/110, respectively (Table A5). The test chemical had an LN transformation response in the first experiment and an SP transformation response in the second experiment. Ethylene thiourea was evaluated as weakly active in the transformation assay, and its actual and estimated rank t-statistics were 2.84 and 3.96, respectively (Table A6).
Hexamethylphosphoramide. Hexamethylphosphoramide is a level C carcinogen (Table A6). It had no insurmountable technical problems (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 64.4 mM CTable A4). The statisti-cal sensitivities oftransformation assay trials 1 and 2 were 9 and 59/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 37 and 17/110, respectively (Table A5). The test chemical had an SP transformation response in the two consecutive experiments. Hexamethylphosphoramide was evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were both 2.08 (Table A6).
Melamine. Melamine is a level D carcinogen (Table A6). It had one difficult technical problem, because its solubility limit in culture medium supplemented with pluronic F68 was about 1000 ,ug/ml (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 39.6 mM (Table A4). Thus, this LD50 far exceeded the solubility limit of the test chemical in culture medium. The statistical sensitivities of transformation assay trials 1 and 2 were 30 and 97/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 4 and 94/110, respectively (Table A5). The test chemical had an SP transformation response in the first experiment, and an LA transformation response in the second experiment. Melamine was evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were 2.10 and 2.15, respectively (Table A6).
Methyl Carbamate. Methyl carbamate is one of five chemicals that was tested as a coded and as a uncoded test chemical in this investigation. It is a level C carcinogen (Table A6). It had no insurmountable technical problems (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells, and the uncoded and coded test chemicals had an average LD50 of 225 and 195 mM, respectively (Table A4). For the uncoded test chemical the statistical sensitivities of transformation assay trials 1-3 were 55, 108 and 10/110, respectively; the detection sensitivities for BaP of trials 1-3 were 8, 99 and 5/110, respectively (Table  A5). For the coded test chemical the statistical sensitivities of transformation assay trials 1 and 2 were 71 and 33/110; the detection sensitivities for BaP of trials 1 and 2 were 23 and 14/110, respectively. The uncoded test chemical had two SP and a LA transformation responses in three experiments. The coded test chemical had an SP transformation response in the two consecutive experiments. Methyl carbamate was evaluated as active in the transformation assay. The actual and estimated rank t-statistics for the uncoded test chemical were both 4.48; the actual and estimated rank t-statistics for the coded test chemical were 3.76 (Table A6). Taken together, the coded and uncoded test chemicals had virtually identical cytotoxic and transformation responses in the BALB/c-3T3 cell transformation assay. Monuron. Monuron is a level B carcinogen (Table A6). It had one difficult technical problem. It had a limited solubility in culture medium supplemented with pluronic F68 of about 25 Fg/ml (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 5.54 mM (Table A4). Thus, this LD50 far exceeded the solubility limit of the test chemical in culture medium.
The statistical sensitivities of transformation assay trials 1 and 2 were 81 and 39/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 22 and 68/110, respectively ( Table 2). The test chemical had an LA transformation response in the first experiment and an SN transformation response in the second experiment. Monuron was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were both 1.01 (Table A6).
Saccharin, Sodium Salt. Saccharin, sodium salt, is a level A caranogen (Table A6). It had no insurmountable technical problems (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 76.5 mM and 153 mOsM (Table A4). The statistical sensitivities of transformation assay lrials 1 and 2 were 21 and 62/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 26 and 48/110, respectively ( Table A5). The test chemical had an SP transformation response in the two consecutive experiments. Significant transforming activity was detected at treatment doses that were below the upper dose limit of the assay of 100 mOsM. Saccharin was evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were both 7.99 (Table A6).
Noncytotoxic, Mutagenic Noncarcinogens 2-Chloroethanol. 2-Chloroethanol is a level F noncarcinogen (Table A6). It had one difficult technical problem. It was reported to react with water; thus, it may have been altered by the aqueous environment during the treatment period (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 81.0 mM (Table A4). The statistical sensitivities of transformation assay trials 1 and 2 were 9 and 73/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 37 and 49/110, respectively (Table A5). The test chemical had an SP transformation response in the both the first and second experiments. 2-Chloroethanol was evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were both 3.22 (Table A6).
Methyl Methacrylate. Methyl methacrylate is a level F noncarcinogen (Table A6). It had two difficult technical problems. It was reported to react with air; thus, it may have been altered by exposure to air during the treatment period. In addition, it was noted to react with amines; therefore, it may have reacted with amines on biochemicals in culture medium and in the target BALB/c-3T3 cells (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 10.8 mM (Table A4). The statistical sensitivities of transformation assay trials 1 and 2 were 4 and 28/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 1 and 15/110, respectively (Table A5). The test chemical had an SP transformation response in the first experiment and an SN transformation response in the second experiment. Since the mean t-statistics of the two transformation responses were not significantly different from one another, the test chemical did not have to be evaluated in a third trial. Methyl methacrylate was evaluated as having been weakly active in the transformation assay, and its actual and estimated rank t-statistics were both 2.54 (Table A6). 4-Nitroanthranilic Acid. 4-Nitroanthranilic acid is a level F noncarcinogen (Table A6). It had no insurmountable technical problems (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 8.58 mM (Table A4). The statistical sensitivities of transformation assay trials 1 and 2 were 7 and 22/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 56 and 53/110, respectively (Table A5). The test chemical had an SN transformation response in the first experiment and an SP transformation response in the second experiment. Because the mean t-statistics of the two transformation responses were not significantly different from one another, the test chemical was not evaluated in a third trial. 4-Nitroanthranilic acid was evaluated as having been weakly active in the transformation assay, and its actual and estimated rank t-statistics were both 1.09 (Table A6).

Noncytotoxic, Nonmutagenic Noncarcinogens
Aldicarb. Aldicarb is a level F noncarcinogen (Table A6). It had no insurmountable technical problems (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 10.7 mM (Table A4). The statistical sensitivities of transformation assay trials 1 and 2 were 19 and 33/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 34 and 14/110, respectively (Table A5). The test chemical had an SN transformation response in the first experiment, and an SP response in the second experiment. Since the mean t-statistics of the transformation responses in the two experiments were significantly different from one another, the test chemical has to be evaluated in a third trial. In the absence of the data from a third trial, aldicarb was evaluated as having an indeterminate activity in the transformation assay. Its actual and estimated rank t-statistics were both 1.93 (Table A6). Ampicillin 7hihydrate. Ampicillin trihydrate is a level E noncarcinogen (Table A6). It had no insurmountable technical problems (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 23.8 mM (Table A4). The statistical sensitivities of transformation assay trials 1 and 2 were 71 and 29/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 23 and 97/110, respectively (Table A5). The test chemical had an LA transformation response in the both the first and second experiments. Ampicillin trihydrate was evaluated as having an equivocal activity in the transformation assay, and its actual and estimated rank t-statistics were both 1.32 (Table A6). o-Anthranilic Acid. o-Anthranilic acid is a level F noncarcinogen (Table A6). It had one difficult technical problem. It was reported to become oxidized by air; thus, it may have been altered by exposure to air during the treatment (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 72.9 mM (Table A4). The statistical sensitivities of transformation assay trials 1 and 2 were 100 and 36/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 42 and 28/110, respectively (Table A5). The test chemical had an SN transformation response in the first experiment and an SP transformation response in the second experiment. Since the mean t-statistics of the two transformation responses were significantly different from one another, the test chemical has to be evaluated in a third trial. In the absence of data from a third trial, o-anthranilic acid was evaluated as having an indeterminate activity in the transformation assay. Its actual and estimated rank t-statistics were both 2.19 (Table A6). Benzoin. Benzoin is a level E noncaranogen (Table A6). It had one difficult technical problem. The solubility limit of this test chemical in culture medium supplemented with pluronic F68 was about 500 ,ug/ml (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 14.8 mM (Table A4). Thus, this test chemical had a LD50 that far exceeded its solubility in culture medium. The statistical sensitivities of transformation assay trials 1 and 2 were 50 and 109/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 59 and 103/110, respectively (Table A5). The test chemical had an SN transformation response in the both the first and second experiments. The test chemical was tested at treatment doses that far exceeded its solubility limit in culture medium. Benzoin was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were 0.74 and 1.08, respectively (Table  A6).
Benzyl Alcohol. Benzyl alcohol is a level F noncarcinogen (Table A6). It had one difficult technical problem. It was reported to become oxidized by air; thus, it may have been altered by exposure to air during the treatment period (Table A4. The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 17.9 mM (Table A4). The statistical sensitivities of transformation assay trials 1 and 2 were 2 and 38/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 2 and 16/110, respectively (Table A5). The test chemical had an LA transformation response in the first experiment and an SP transformation response in the second experiment. Benyl alcohol was evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were both 1.95 (Table A6). Caprolactam. Caprolactam is a level F noncarcinogen (Table A6). It had no insurmountable technical problems (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 71.8 mM ( Table A4). The statistical sensitivities of transformation assay trials 1 and 2 were 110 and 109/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 107 and 103/110, respectively (Table A5). Thus, both of these experiments had an unusually low statistical sensitivity and detection sensitivity for BaR The test chemical had an SN transformation response in the both the first and second experiments. Caprolactam was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were 1.20 and 2.34, respectively (Table A6).
(2-Chloroethyl)trimethylammonium Chloride. (2-Chloroethyl)trimethylammonium chloride is a level F non-carcinogen (Table A6). It had no insurmountable technical problems (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 62.0 mM (Table A4). The statistical sensitivities of transformation assay trials 1 and 2 were 40 and 35/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 54 and 57/110, respectively (Table A5). The test chemical had an SP transformation response in the first experiment and an LA transformation response in the second experiment.
(2-Chloroethyl)trimethylammonium chloride was evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were both 1.74 (Table A6).
C. I. Acid Orange 10. C. I. Acid orange 10 is a level F noncarcinogen (Table A6). It had no insurmountable technical problems (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 26.5 mM (Table A4). The statistical sensitivities of transformation assay trials 1 and 2 were 92 and 22/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 71 and 53/110, respectively (Table A5). The test chemical had an SN transformation response in the both the first and second experiments. C. I. Acid orange 10 was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were 0.63 and 0.68, respectively (Table A6) Dimethyl Terephthalate. Dimethyl terephthalate is a level E noncarcinogen (Table A6). It two difficult technical problems. It was reported to be very temperature sensitive and react with water. Because the test chemical was insoluble in water, it had to be sonicated and warmed at 37°C for 30 minutes or more to become a fine particulate suspension in culture medium supplemented with pluronic F68. Its solubility in culture medium was about 125 pLg/ml (Table A4). Thus, it is possible that the test chemical was altered during the procedure to solubilize the test chemical as well as the treatment period. The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 15.4 mM ( Table A4). The statistical sensitivities of transformation assay trials 1 and 2 were 22 and 5/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 53 and 46/110, respectively (Table A5). The test chemical had an SP transformation response in the first experiment and an LA transformation response in the second experiment. The test chemical was tested at doses that far exceeded its solubility in culture medium. Taken together, dimethyl terephthalate was evaluated as having an indeterminate activity in the transformation assay, because it needs to be retested at treatment doses closer to its solubility limit in culture medium. Furthermore, it must be evaluated using a procedure less likely to alter it while it is being solubilized for testing. Its actual and estimated rank t-statistics were both 1.79 (Table A6). Diphenylhydantoin. Diphenylhydantoin is a level I chemical, because its testing in rodent bioassay is incomplete (Table A6). It had no insurmountable technical problems (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 5.02 mM (Table A4). The solubility limit of this test chemical was about 500 ,ug/ml in culture medium supplemented with pluronic F68. The statistical sensitivities of transforma-tion assay trials 1 and 2 were 84 and 95/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 105 and 61/110, respectively (Table A5). The test chemical had an SN transformation response in two consecutive experiments, and it was tested at treatment doses above its solubility limit. Diphenylhydantoin was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were both 0.00 (Table A6).
FD and C Yellow No. 6. FD and C Yellow no. 6 is a level F noncarcinogen (Table A6). It had no insurmountable technical problems (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 67.7 mM (Table A4). The statistical sensitivities of transformation assay trials 1-3 were 7, 95 and 22/110, respectively; the detection sensitivities for BaP of trials 1-3 were 56, 61 and 53/110, respectively (Table A5). The test chemical had an SP transformation response in the all three experiments. FD and C Yellow No. 6 was evaluated as active in the transformation assay, and its actual and estimated rank t-statistics were both 7.65 (Table A6).
D-Mannitol. D-Mannitol is a level F noncarcinogen (Table A6). It had no insurmountable technical problems (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 estimated to be over 324 mM (Table A4). The statistical sensitivities of transformation assay trials 1-3 were 46, 35 and 38/110, respectively; the detection sensitivities for BaP of trials 1-3 were 80, 57 and 16/110, respectively (Table A5). The test chemical had an LN transformation response in the first experiment, an LA response in the second experiment, and an SP response in the third experiments. In all three experiments significant transforming activity was only detected at treatment doses that exceeded the upper dose limit of the assay of 100 mOsM. Taken together, D-mannitol was evaluated as active and a false positive in the transformation assay, and its actual and estimated rank t-statistics were 3.00 and 1.99, respectively (Table A6).
Molybdenum Riaoxide. Molybdenum trioxide is a level I noncarcinogen (Table A6). It had one difficult technical problem. It was reported to form polymeric compounds when it was exposed to acids and bases. Since stocks of the test chemical were acidic and had to be neutralized with NaOH in order to tested, it is possible that the test chemical was altered during preparation of the dosing solutions (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 9.38 mM ( Table A4). The statistical sensitivities of transformation assay trials 1 and 2 were 61 and 84/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 47 and 105/110, respectively (Table A5). The test chemical had an LA transforrnation response in the first experiment and an SN transformation response in the second experiment. Molybdenum trioxide was evaluated as inactive in the transformation assay, and ils actual and estimated rank t-statistics were 0.64 and 0.86, respectively (Table A6).
Penicillin VK+. Penicillin VK + is a level F noncarcinogen (Table A6). It had no insurmountable technical problems (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 17.8 mM (Table A4). The statistical sensitivities of transformation assay trials 1 and 2 were 10 and 62/110, respectively, the detection sensitivities for BaP of trials 1 and 2 were 5 and 48/110, respectively (Table A5). The test chemical had an SP transformation response in two consecutive experiments. Penicillin VK+ was evaluated as active in the transforrnation assay, and its actual and estimated rank t-statistics were both 5.96 (Table A6). Phthalamide. Phthalamide is a level F noncarcinogen (Table A6). It had no insurmountable technical problems (Table A4). The lest chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 73.1 mM (Table A4). This LD50 was far above the solubility limit of the test chemical of 1000 ,ug/ml in culture medium supplemented with pluronic F68. The statistical sensitivities of transformation assay trials 1 and 2 were 14 and 38/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 72 and 16/110, respectively (Table A5). The test chemical had an LN transformation response in the first experiment and an LA transformation response in the second experiment. Phthalamide was evaluated as having an indeterminate activity in the transformation assay, and its actual and estimated rank t-statistics were 1.01 and 0.64, respectively (Table A6).
Phthalic Anhydride. Phthalic anhydride is a level F noncarcinogen (Table A6). It had three difficult technical problems. It was reported to react with water and strong bases. Because the test chemical stock solutions were very acidic, they had to be neutralized with NaOH; thus, the test chemical may have been altered during the preparation of dosing solutions. In addition, the test chemical was noted to react with amine groups, thus it may have reacted with amine groups on biochemicals in culture medium, as well as biochemicals in the target cells (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 13.2 mM (Table A4). The statlstical sensitivities of transformation assay trials 1 and 2 were 82 and 5/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 18 and 46/110, respectively (Table  A5). The test chemical had an LA transformation response in the first experiment and an SN transformation response in the second experiment. Phthalic anhydride was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were both 0.80 (Table A6). (Table  A4). The test chemical was noncytotoxic to the BALB/ c-3I3 cells and had an average LD50 of43.8 mM (Table A4).

Roaxarsone. Roxarsone is a level E noncarcinogen (Tlable A6). It had no insurmountable technical problems
The statistical sensitivities of transformation assay trials 1 and 2 were 10 and 27/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 5 and 92/110, respectively (Table A5). The test chemical had an SP transformation response in the first experiment and an SN transformation response in the second experiment. Since the mean t-statistics of the two test chemical transformation responses were significantly different from one another, it has to be evaluated in a third trial. In the absence of data from a third trial, roxarsone was evaluated as having an indeterminate activity in the transformation assay, and its actual and estimated rank t-statistics were both 1.67 (Table A6). (Table A6). It had no insurmountable technical problems (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 12.5 mM ( Table A4). The statistical sensitivities of transformation assay trials 1 and 2 were 1 and 6/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 3 and 19/110, respectively (Table A5). The test chemical had an SN transformation response in the first experiment and an SP transformation response in the second experiment. Since the mean t-statistics of the two test chemical transformation responses were significantly different from one another, the chemical has to be evaluated in a third trial. In the absence of data from a third experiment, sodium(2ethylhexyl alcohol sulfate was evaluated as having an indeterminute activity in the transformation assay, and its actual and estimated rank t-statistics were both 2.04 (Table A6). Sulfisoaazole. Sulfisoxazole is a level F noncarcinogen (Table A6). It had one difficult technical problem. It was reported to become oxidized upon exposure to air; thus, it may have been altered by exposure to air during the treatment period (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 estimated to be 18.7 mM (Table A4). The statistical sensitivities of transformation assay trials 1 and 2 were 1 and 6/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 3 and 19/1 10, respectively (Table A5). The test chemical had an LN transformation response in the first experiment and an SP transformation response in the second experiment. Since the LN transformation response did not qualify as one of the two required trials, the test chemical has to be tested in a third experiment. Taken together, sulfisoxazole was evaluated as having an indeterminate activity in the transformation assay, and its actual and estimated rank t-statistics were both 1.26 (Table A6). (Table A6). It had no insurmountable technical problems (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 117 mM (Table A4). The statistical sensitivities of transformation assay trials 1 and 2 were 49 and 23/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 24 and 11/110, respectively (Table A5). The test chemical had an SP transformation response in two consecutive experiments. 3-Sulfolene was evaluated as having been weakly active in the transformation assay, and its actual and estimated rank t-statistics were both 3.24 (Table A6). Tetrahydrofuran. Tetrahydrofuran is a level I chemical because its testing in rodent bioassay is incomplete (Table  A6). It had one difficult technical problem because it was reported to react with water (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 90.3 mM (Table A4). The statistical sensitivities of transformation assay trials 1 and 2 were 1 and 28/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 3 and 15/110, respectively (Table A5). The test chemical had an SN transformation response in two consecutive experiments. Tetrahydrofuran was evaluated as inactive in the transformation assay, and its actual and estimated rank t-statistics were both 0.72 (Table A6).

3-Sulfolene. 3-Sulfolene is a level F noncarcinogen
Titanium Dioxide. Titanium dioxide is a level F noncarcinogen (Table A6). It had one difficult technical problem. It was reported to be reduced by metals such as calcium, magnesium, potassium and sodium; thus, it could have been altered by these metals in culture medium. In addition, this test chemical was very insoluble in culture medium supplemented with pluronic F68, and had a solubility limit of about 125 ,ug/ml (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 12.1 mM (Table A4). Thus, this LD50, far exceeded the solubility limit of the test chemical in culture medium. The statistical sensitivities of transformation assay trials 1 and 2 were 82 and 6/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 12 and 92/110, respectively (Table A5). The test chemical had an LN transformation response in two consecutive experiments, and it was tested at treatment doses that far exceeded its solubility limit. Titanium dioxide was evaluated as an inactive with an indeterminate activity in the transfonnation assay. Its actual and estimated rank t-statistics were both 0.00 (Table A6).
Witch Hazel. Witch hazel is a mixture of chemicals that include 15% v/v ethanol, 85% v/v water and some inert ingredients. It has been shown to be a level F noncarcinogen (Table A6). It had no insurmountable technical problems (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 estimated to be 540 mOsM (Table A4). The statistical sensitivities of transformation assay trials 1 and 2 were 2 and 38/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 2 and 16/110, respectively (Table A5). The test chemical had an LA transformation response in two consecutive experiments. This test chemical only induced significant transforming activity at treatment doses that exceeded the upper dose limit of the assay of 100 mOsM. Taken together, witch hazel was evaluated as having an equivocal activity and a false positive in the transformation assay. Its actual and estimated rank t-statistics were both 1.95 (Table A6).

Very Noncytotoxic Chemicals
Acetone. Acetone has not been evaluated in rodent bioassay, therefore, its level of carcinogenicity is classified as a level I chemical (Table A6). It had no insurmountable technical problems (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of about 257 mOsM (Table A4). The statistical sensitivities of transformation assay trials 1 and 2 were 1 and 27/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 3 and 63/110, respectively (Table A5). The test chemical had an LA transformation response in the first experiment, and an SP response in the second experiment. This test chemical induced significant transforming activity only at treatment doses above the upper dose limit of the assay of 100 mOsM. Taken together, acetone was evaluated as active and a false positive in the transfonnation assay. Its actual and estimated rank t-statistics were both 4.49 (Table A6). Dimethyl Sulfoxide. Dimethyl sulfoxide has not been evaluated in rodent bioassay; therefore, its level of carcinogenicity is classified as a level I chemical (Table A6). It had no insurmountable technical problems (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 507 mM (Table A4). The statistical sensitivities of transformation assay trials 1 and 2 were 90 and 73/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 13 and 49/110, respectively ( Table A5). The test chemical had an LA transformation response in the first experiment, and an SP response in the second experiment. This test chemical induced significant transforming activity only at treatment doses above the upper dose limit of the assay of 100 mOsM. Taken together, dimethyl sulfoxide was evaluated as active and a false positive in the transformation assay. Its actual and estimated rank t-statistics were 3.38 and 3.45, respectively (Table A6).
Ethanol. Ethanol has not been evaluated in rodent bioassay; therefore, its level of carcinogenicity is classified as a level I chemical (Table A6). It had no insurmountable technical problems (Table A4). The test chemical was noncytotoxic to the BALBtc-3T3 cells and had an average LD50 of 429 mM (Table A4). The statistical sensitivities of transformation assay trials 1 and 2 were 2 and 20/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 2 anid 19/110, respectively (Table A5). The test chemical had an LA transformation response in the first experiment, and an SP response in the second experiment. This test chemical induced significant transforming activity only at treatment doses above the upper dose limit of the assay of 100 mOsM. Taken together, ethanol was evaluated as active and a false positive in the transformation assay. Its actual and estimated rank t-statistics were both 2.50 (Table A6).
Glycerol. Glycerol has not been evaluated in rodent bioassay; therefore, its level of carcinogenicity is classified as a level I chemical (Table A6). It had no insurmountable technical problems (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 401 mM (Table A4). The statistical sensitivities of transformation assay trials 1 and 2 were 1 and 20/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 3 and 19/110, respectively (Table A5). The test chemical had an SP transformation response in both the first and second experiments. This test chemical induced significant transforming activity only at treatment doses above the upper dose limit of the assay of 100 mOsM. Taken together, glycerol was evaluated as active and a false positive in the transformation assay. Its actual and estimated rank t-statistics were both 3.09 (Table A6).
Sodium Chloride. Sodium chloride has not been evaluated in rodent bioassay; therefore, its level of carcinogenicity is classified as a level I chemical (Table A6). It had no insurmountable technical problems (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of about 144 mM (288 mOsM) ( Table A4). The statistical sensitivities of transformation assay trials 1-3 were 10, 6 and 106/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 5, 92 and ND/110, respectively (Table A5). The test chemical had an SP transformation response in the first and third experiments, and an SN response in the second experiment. There was no apparent explanation for the absence of test chemical transforming activity in the second experiment. This test chemical induced significant transforming activity only at treatment doses above the upper dose limit of the assay of 100 mOsM. Taken together, sodium chloride was evaluated as active and a false positive in the transformation assay. Its actual and estimated rank t-statistics were both 6.53 (Table A6).
Sucrose. Sucrose has not been evaluated in rodent bioassay, therefore, its level of carcinogenicity is classified as a level I chemical (Table A6). It had no insurmountable technical problems (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of 240 mM (Table A4). The statistical sensitivities of transformation assay trials 1 and 2 were 62 and 5/110, respectively; the detection sensitivities for BaP of trials 1 and 2 were 48 and 46/110, respectively (Table A5). The test chemical had an SP transformation response in both the first and second experiments. This test chemical induced significant transforming activity only at treatment doses above the upper dose limit of the assay of 100 mOsM. Taken together, sucrose was evaluated as active and a false positive in the transformation assay. Its actual and estimated rank t-statistics were both 5.73 (Table A6).
Urea. Urea has not been evaluated in rodent bioassay; therefore, its level of carcinogenicity is classified as a level I chemical (Table A6). It had no insurmountable technical problems (Table A4). The test chemical was noncytotoxic to the BALB/c-3T3 cells and had an average LD50 of about 254 mM (Table A4). The statistical sensitivity of transfonnation assay trial 1 was 6/110; the detection sensitivity for BaP oftrial 1 was 92/110 (Table A5). The test chemical had an SP transformation response in the first experiment. This test chemical induced significant transforming activity only at treatment doses above the upper dose limit of the assay of 100 mOsM. Taken together, urea was evaluated as active and a false positive in the transformation assay. Its actual and estimated rank t-statistics were both 1.81 (Table A6) (20) 316 (20) 24 (74) 128 (20) 37 (20) 182 (20) 248 (20) 52 (20) 11 (20)  Abbreviations: BaP, benzo(a)pyrene; CCA, co-culture clonal survival assay; Conc, concentration; mM, millimole; M.W., molecular weight; N, number of culture vessels, NC, negative control; ND; not determined; %RCE, percent relative cloning efficiency; S.A, standard clonal survival assay. aTreatment condition: The experimental design for the transformation assay is described in detail in the Materials and Methods. The concentration of the positive control and test chemical treatment are presented in mM, but they can be converted to ,ug/ml using the molecular weight that is provided with each chemical. The solvent vehicles used for the individual test chemicals were listed in Appendix Tables Al and A3, and the concentrations of the solvent vehicles are presented in the Materials and Methods. bCytotoxic activity: The experimental design for the standard survival assay (SA) and the co-culture clonal survival assay (CCA) were described in the Materials and Methods. The test chemical cytotoxic response was expressed as % RCE and was calculated as described in the Materials and Methods.
cThe criteria used to evaluate the transformed foci of BALB/c-3T3 cells is described in the Materials and Methods. The number of type III foci > 2-mm in diameter per culture vessel scored are recorded in this table.
dTransformation response: The transformation responses are expressed as type III foci/vessel and were calculated using a log1o, mathematical transformation procedure (refer to Materials and Methods). The arithmetic value or foci/vessel represents the antilog of the log1o mean transformation response minus one. eSignificance: The significance of test chemical transformation responses was calculated by a computer using the SAS statistical. software (22), and the method is described in detail in Materials and Methods. The correct t-statistic according to the F-test is presented in this table. The t-statistics of each treatment dose of the test chemical in a single experiment were averaged to determine the mean t-statistic of the test chemical for the experiment (refer to Appendix Tables A2 and A5). The mean t-statistics for two or experiments for each chemical was weighted to the number of treatment doses evaluated and averaged to determine the rank t-statistic which was used to rank-order the test chemical transformation responses in Appendix Tables A3 and A6. Arbitrarily, transformation responses with negative (-) t-statistics were given a value of zero (0). *Significant BaP or test chemical transformation response, 0.01 < p < 0.05. **Significant BaP or test chemical transformation response, 0.001 < p < 0.01. ***Significant or BaP or test chemical transformation response, p < 0.001.
Treatment condition: The experimental design for the transformation assay is described in detail in the Materials and Methods. The concentration of the positive control and test chemical treatment are presented in mM, but they can be converted to ,ug/ml using the molecular weight that is provided with each chemical. The solvent vehicles used for the individual test chemicals were listed in Appendix Tables Al and A3, and the concentrations of the solvent vehicles are presented in the Materials and Methods. bCytotoxic activity: The experimental design for the standard survival assay (SA) and the co-culture clonal survival assay (CCA) were described in the Materials and Methods. The test chemical cytotoxic response was expressed as % RCE and was calculated as described in the Materials and Methods.
cThe criteria used to evaluate the transformed foci of BALB/c-3T3 cells is described in the Materials and Methods. The number of type III foci > 2-mm in diameter per culture vessel scored are recorded in this table.
dTransformation response: The transformation responses are expressed as type III foci/vessel and were calculated using a log1o mathematical transformation procedure (refer to Materials and Methods). The arithmetic value or foci/vessel represents the antilog of the log1o mean transformation response minus one. eSignificance: The significance of test chemical transformation responses was calculated by a computer using the SAS statistical software (22), and the method is described in detail in Materials and Methods. The correct t-statistic according to the F-test is presented in this table. The t-statistics of each treatment dose of the test chemical in a single experiment were averaged to determine the mean t-statistic of the test chemical for the experiment (refer to Appendix Tables A2 and A5). The mean t-statistics for two or experiments for each chemical was weighted to the number of treatment doses evaluated and averaged to determine the rank t-statistic which was used to rank-order the test chemical transformation responses in Appendix Tables A3 and A6. Arbitrarily, transformation responses with negative (-) t-statistics were given a value of zero (0). *Significant BaP or test chemical transformation response, 0.01 < p < 0.05. **Significant BaP or test chemical transformation response, 0.001 < p < 0.01. ***Significant or BaP or test chemical transformation response, p <. 0.001.
cThe criteria used to evaluate the transformed foci of BALB/c-3T3 cells is described in the Materials and Methods. The number of type III foci > 2-mm in diameter per culture vessel scored are recorded in this table.
aTreatment Condition: The experimental design for the transformation assay is described in detail in the Materials and Methods. The concentration of the positive control and test chemical treatment are presented in mM, but they can be converted to ,ug/ml using the molecular weight that is provided with each chemical. The solvent vehicles used for the individual test chemicals were listed in Appendix Tables Al and A3, and the concentrations of the solvent vehicles are presented in the Materials and Methods. bCytotoxic activity: The experimental design for the standard survival assay (SA) and the co-culture clonal survival assay (CCA) were described in the Materials and Methods. The test chemical cytotoxic response was expressed as & RCE and was calculated as described in the Materials and Methods.
cThe criteria used to evaluate the transformed foci of BALB/c-3T3 cells is described in the Materials and Methods. The number of type III foci > 2-mm in diameter per culture vessel scored are recorded in this table.
dTransformation response: The transformation responses are expressed as type III foci/vessel and were calculated using a log1o mathematical transformation procedure (refer to Materials and Methods). The arithmetic value or foci/vessel represents the antilog of the log1o mean transformation response minus one. eSignificance: The significance of test chemical transformation responses was calculated by a computer using the SAS statistical software (22), and the method is described in detail in Materials and Methods. The correct t-statistic according to the F-test is presented in this table. The t-statistics of each treatment dose of the test chemical in a single experiment were averaged to determine the mean t-statistic of the test chemical for the experiment (refer to Appendix Tables A2 and A5). The mean t-statistics for two or experiments for each chemical was weighted to the number of treatment doses evaluated and averaged to determine the rank t-statistic which was used to rank-order the test chemical transformation responses in Appendix Tables A3 and A6. Arbitrarily, transformation responses with negative (-) t-statistics were given a value of zero (0). *Significant BaP or test chemical transformation response, 0.01 < p < 0.05. **Significant BaP or test chemical transformation response, 0.001 < p < 0.01. ***Significant or BaP or test chemical transformation response, p < 0.001. Summary of the transformation responses of 12 noncytotoxic carcinogens.  (40) 197 (19) 115 (20) 100 (20)  87 (20) 42 (20)  53 (20)  91 (38) 48 (20) 39 (20) 17 (20) 10 (20)  .00 (-5.88) .00 (-7.33) .00 (-3.07 .00(-1. 14) .00 (-5.11) .00 (-1 .66) .00 ( Abbreviations: B(a)P, benzo(a)pyrene; CC.A, co-culture clonal survival assay; Cone, concentration; mM, millimole; M.W., molecular weight; N, number of culture vessels, NC, negative control; ND, not determined; %RCE, percent relative cloning efficiency; S.A, standard clonal survival assay.
'Treatment Condition: The experimental design for the transformation assay is described in detail in the Materials and Methods. The concentration of the positive control and test chemical treatment are presented in mM, but they can be converted to j±g/ml using the molecular weight that is provided with each chemical. The solvent vehicles used for the individual test chemicals were listed in Appendix Tables Al and A3, and the concentrations of the solvent vehicles are presented in the Materials and Methods. bCytotoxic activity: The experimental design for the standard survival assay (SA) and the co-culture clonal survival assay (CCA) were described in the Materials and Methods. The test chemical cytotoxic response was expressed as & RCE and was calculated as described in the Materials and Methods.
cThe criteria used to evaluate the transformed foci of BALB/c-3T3 cells is described in the Materials and Methods. The number of type III foci > 2-mm in diameter per culture vessel scored are recorded in this table.
dThansformation response: The transformation responses are expressed as type III foci/vessel and were calculated using a log1o mathematical transformation procedure (refer to Materials and Methods). The arithmetic value or foci/vessel represents the antilog of the log1o mean transformation response minus one. eSignificance: The significance of test chemical transformation responses was calculated by a computer using the SAS statistical software (22), and the method is described in detail in Materials and Methods. The correct t-statistic according to the F-test is presented in this table. The t-statistics of each treatment dose of the test chemical in a single experiment were averaged to determine the mean t-statistic of the test chemical for the experiment (refer to Appendix Tables A2 and A5). The mean t-statistics for two or experiments for each chemical was weighted to the number of treatment doses evaluated and averaged to determine the rank t-statistic which was used to rank-order the test chemical transformation responses in Appendix Tables A3 and A6. Arbitrarily, transformation responses with negative (-) t-statistics were given a value of zero (0). *Significant BaP or test chemical transformation response, 0.01 < p < 0.05. 'Significant BaP or test chemical transformation response, 0.001 < p < 0.01. ***Significant or BaP or test chemical transformation response, p < 0.001. Appendix G.  (20) 196 (20) 119 (20) 108 (40) 85 (18) 43 (20) 40 (20)  Abbreviations: B(a)P, benzo(a)pyrene; CC.A., co-culture clonal survival assay; Cone., concentration; mM, millimole; M.W., molecular weight; N, number of culture vessels, NC, negative control; ND, not determined; %RCE, percent relative cloning efficiency; S.A., standard clonal survival assay.
aTreatment Condition: The experimental design for the transformation assay is described in detail in the Materials and Methods. The concentration of the positive control and test chemical treatment are presented in mM, but they can be converted to ,ug/ml using the molecular weight that is provided with each chemical. The solvent vehicles used for the individual test chemicals were listed in Appendix Tables Al and A3, and the concentrations of the solvent vehicles are presented in the Materials and Methods. bCytotoxic activity: The experimental design for the standard survival assay (SA) and the co-culture clonal survival assay (CCA) were described in the Materials and Methods. The test chemical cytotoxic response was expressed as & RCE and was calculated as described in the Materials and Methods.
cThe criteria used to evaluate the transformed foci of BALB/c-3T3 cells is described in the Materials and Methods. The number of type III foci > 2-mm in diameter per culture vessel scored are recorded in this table.
dTransformation response: The transformation responses are expressed as type III foci/vessel and were calculated using a log1o mathematical transformation procedure (refer to Materials and Methods). The arithmetic value or foci/vessel represents the antilog of the log1o mean transformation response minus one. eSignificance: The significance of test chemical transformation responses was calculated by a computer using the SAS statistical software (22), and the method is described in detail in Materials and Methods. The correct t-statistic according to the F-test is presented in this table. The t-statistics of each treatment dose of the test chemical in a single experiment were averaged to determine the mean t-statistic of the test chemical for the experiment (refer to Appendix Tables A2 and A5). The mean t-statistics for two or experiments for each chemical was weighted to the number of treatment doses evaluated and averaged to determine the rank t-statistic which was used to rank-order the test chemical transformation responses in Appendix Tables A3 and A6. Arbitrarily, transformation responses with negative (-) t-statistics were given a value of zero (0). *Significant BaP or test chemical transformation response, 0.01 < p < 0.05. **Significant BaP or test chemical transformation response, 0.001 < p < 0.01. ***Significant or BaP or test chemical transformation response, p < 0.001.
"Treatment Condition: The experimental design for the transformation assay is described in detail in the Materials and Methods. The concentration of the positive control and test chemical treatment are presented in mM, but they can be converted to ,ug/ml using the molecular weight that is provided with each chemical. The solvent vehicles used for the individual test chemicals were listed in Appendix Tables Al and A3, and the concentrations of the solvent vehicles are presented in the Materials and Methods. bCytotoxic activity: The experimental design for the standard survival assay (SA) and the co-culture clonal survival assay (CCA) were described in the Materials and Methods. The test chemical cytotoxic response was expressed as & RCE and was calculated as described in the Materials and Methods.
cThe criteria used to evaluate the transformed foci of BALB/c-3T3 cells is described in the Materials and Methods. The number of type III foci > 2-mm in diameter per culture vessel scored are recorded in this table.
dTransformation response: The transformation responses are expressed as type III foci/vessel and were calculated using a log1o mathematical transformation procedure (refer to Materials and Methods). The arithmetic value or foci/vessel represents the antilog of the log1o mean transformation response minus one. eSignificance: The significance of test chemical transformation responses was calculated by a computer using the SAS statistical software (22), and the method is described in detail in Materials and Methods. The correct t-statistic according to the F-test is presented in this table. The t-statistics of each treatment dose of the test chemical in a single experiment were averaged to determine the mean t-statistic of the test chemical for the experiment (refer to Appendix Tables A2 and A5). The mean t-statistics for two or experiments for each chemical was weighted to the number of treatment doses evaluated and averaged to determine the rank t-statistic which was used to rank-order the test chemical transformation responses in Appendix Tables A3 and A6. Arbitrarily, transformation responses with negative (-) t-statistics were given a value of zero (0). *Significant BaP or test chemical transformation response, 0.01 < p < 0.05. **Significant BaP or test chemical transformation response, 0.001 < p < 0.01. ***Significant or BaP or test chemical transformation response, p < 0.001.