Assessment of chemicals using a battery of neurobehavioral tests: a comparative study

Single-dose LD10S and LD50S were determined in male, Fischer-344 rats for acrylamide monomer, arsenic trioxide, chlordecone, lead acetate, methylmercury hydroxide, monosodium salicylate, tetraethyl tin, and triethyl lead chloride. Proportions of the single-dose LD10S were used in a subacute study to estimate the 28-day LD20S for each chemical. Proportions of the 28-day LD20S were used in a subchronic (105 days of dosing) study to determine the effectiveness of a battery of neurobehavioral tests for detecting and characterizing the neurotoxic effects of each chemical. The battery consisted of undifferentiated motor activity, forelimb and hindlimb grip strengths, rotation orientation, thermal sensitivity, startle responsiveness to acoustic and air-puff stimuli, and performance of a multisensory conditioned pole-climb avoidance response task; body weight and rectal temperature were also monitored. The battery of tests was administered on eight occasions, that is, before, at three-week intervals during dosing (PO or IP, five days each week for 15 weeks), and at three and six weeks after dosing. Normative data (controls from each experiment) indicated fair overall stability of the measures over the eight test sessions, but experiment-to-experiment variability in this regard was clearly evident. The inherent statistical sensitivity of the tests varied greatly as estimated by their coefficients of variation, which ranged from 1% (rectal temperature) to over 100% (rotation orientation). Intercorrelations among the various measures were low to moderate indicating relatively little redundancy. The various measures were differentially affected by the eight chemicals: body weight by all eight; rectal temperature by one; undifferentiated motor activity by three; forelimb grip strength by two; hindlimb grip strength by four; rotation orientation by one; thermal sensitivity by one; startle responsiveness by three; and CAR performance by five. A profile analysis using the slopes of the dose-response functions after 15 weeks of dosing indicated clearly different patterns of effect among the eight chemicals. A composite score derived from a multidimensional analysis of the data suggested that the eight chemicals could be ranked from most to least neurotoxic as defined by these tests as acrylamide greater than methylmercury greater than chlordecone greater than tetraethyl tin greater than triethyl lead greater than lead acetate greater than arsenic greater than monosodium salicylate. This ranking and the results in general are in good agreement with what is known about these chemicals from other experiments with animals and from human experience. This battery of tests, or subsets thereof, may therefore have utility in the assessment of the potential neurobehavioral toxicity of various chemicals.