Perinatal Exposure to Aroclor 1254 Impairs Distortion Product Otoacoustic Emissions (DPOAEs) in Rats

Robert E. Lasky*,1, John J. Widholm{dagger}, Kevin M. Crofton{ddagger} and Susan L. Schantz{dagger}

* Center for Clinical Research and Evidence-Based Medicine, University of Texas Health Science Center at Houston Medical School, 6431 Fannin, MSB 2.104, Houston, Texas 77030-1503; {dagger} Department of Veterinary Biosciences, University of Illinois at Urbana-Champaign, Urbana, Illinois 61802; and {ddagger} Neurotoxicology Division, MD-74B, National Health and Environmental Effects Research Laboratory, U.S. Environmental Protection Agency, Research Triangle Park, North Carolina 27711

Received December 24, 2001; accepted April 25, 2002


    ABSTRACT
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Polychlorinated biphenyls (PCBs) are ubiquitous environmental contaminants that are a potential health hazard to human and wildlife populations. Low-frequency auditory impairments have previously been documented in Aroclor 1254 (A 1254)-exposed rats, including elevated behavioral auditory thresholds and decreased amplitude and prolonged latency auditory evoked brain stem responses (ABRs). Furthermore, outer hair-cell loss on the basilar membrane of the cochlea has been documented, demonstrating that the cochlea is a target organ for PCB ototoxicity. The current experiment assessed the effects of A1254 on cochlear function by measuring distortion product otoacoustic emissions (DPOAEs). ABRs were measured to determine the effects of A1254 on the central nervous system auditory pathways. Pregnant Long-Evans rats received either 0 or 6 mg/kg A1254 (po) in corn oil from gestation day 6 to lactational day 21. The auditory function of male and female offspring was assessed at approximately 18 months of age. The rats were anesthetized and a probe-unit, consisting of 2 insert earphones and a microphone, was positioned in the ear canal. DPOAE amplitudes were reduced and thresholds increased in the A1254-exposed rats. The deficits were most pronounced at the lowest frequencies tested (2.1–3.2 kHz), but deficits were also observed at higher frequencies (3.7–8.6 kHz). Males and females were equally affected at the lower frequencies, but females were more impaired at the higher frequencies. In contrast, ABR latencies and amplitudes were not altered by A1254 exposure. These findings provide the first functional evidence supporting a cochlear site of damage in PCB-induced hearing loss.

Key Words: polychlorinated biphenyls; PCBs; Aroclor 1254; distortion product otoacoustic emissions; auditory evoked brainstem responses; Long-Evans rats.


    INTRODUCTION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Polychlorinated biphenyls (PCBs) are members of a class of environmental contaminants known as the halogenated hydrocarbons. They readily enter the food chain, accumulating in adipose tissue (Clarkson, 1995Go). PCBs are transported across the placenta (Jacobson et al., 1984Go) and in breast milk (Yakushiji et al., 1984Go), exposing fetuses and young infants at a time when they may be especially vulnerable to insult.

Exposure to PCBs early in development causes a low-frequency (<=1 kHz) hearing loss in laboratory rats when tested using a modified reflex audiometry task (Crofton and Rice, 1999Go; Crofton et al., 2000aGo,bGo; Goldey et al., 1995aGo; Goldey and Crofton, 1998Go). Rats exposed early in development to a commercial PCB mixture, Aroclor 1254 (A1254), also have significantly smaller amplitude auditory evoked brainstem responses (ABRs) at 1 and 4 kHz but not at 16 or 32 kHz (Herr et al., 1996Go).

Goldey et al. (1995a) have hypothesized that PCB-induced auditory losses are caused by a reduction in thyroid hormones that are necessary for normal cochlear development (Uziel, 1986Go). There is empirical support for this hypothesis. Exposing pregnant and lactating rats to PCBs reduces circulating thyroid hormone concentrations in their offspring (Crofton et al., 2000bGo; Goldey and Crofton, 1998Go; Goldey et al., 1995aGo), while postnatal thyroxine (T4)-replacement therapy partially ameliorates A1254-induced hearing loss (Goldey and Crofton, 1998Go).

Crofton et al. (2000a) documented missing outer hair cells on the basilar membrane of the cochlea following perinatal exposure to A1254. Outer hair cell loss was confined to the apical and upper middle turns of the basilar membrane, consistent with the low-frequency hearing loss associated with A1254 exposure. Outer hair cell loss is the likely proximal cause of PCB-induced hearing loss. The effect of A1254 on outer hair cells was assessed in this study by measuring distortion product otoacoustic emissions (DPOAEs) in rats. ABRs were measured to determine whether A1254 exposure damages nerve cells in the auditory pathway as well as outer hair cells in the cochlea.


    MATERIALS AND METHODS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Animals and exposure.
Sixty-seven primiparous female Long-Evans rats were shipped from Charles River (Raleigh, NC) to the U.S. Environmental Protection Agency research facility at Research Triangle Park, NC, for dosing. Animals were shipped on gestational day (GD) 2 (appearance of sperm plug = GD 0) in 2 cohorts spaced 8 weeks apart. The pregnant Long-Evans rats received either 0 or 6 mg/kg A1254 po (1 ml/kg) in corn oil from GD 6 to lactational day (LD) 21. A dose of 6 mg/kg A1254 from GD 6 to LD 21 has been reported to induce hearing loss in rats (Goldey et al., 1995aGo). Twenty litters (10 control and 10 exposed) were randomly selected and weaned on LD 21. One male and one female pup from each litter were randomly selected for further testing. All of these rats were tested on a series of cognitive tasks (Roegge et al., 2000Go; Widholm et al., 2001Go). A subsample of 26 rats (7 control males, 8 control females, 5 A1254-exposed males, and 6 A1254-exposed females) from 18 different litters was tested at approximately 18 months of age to assess auditory function. The University of Illinois’s Office of Laboratory Animal Care and Use approved the procedures in this study.

Apparatus and stimuli.
The instrumentation used to record the distortion product otoacoustic emissions (DPOAEs) consisted of a probe unit (2 Etymotic ER-2 insert earphones and an etymotic ER-10B ear canal microphone), an amplifier (etymotic 10B amplifier) providing 40 dB of gain, an Intel microprocessor-based computer, and an Ariel DSP-16+ signal processing board that generated the calibration stimuli, the stimuli evoking the DPOAEs, and the digitization of the analog signal from the ear canal. The sampling rates for both the analog-to-digital and digital-to-analog converters were 50 kHz. The software for generating the DPOAEs was CUBeDISPTM. The software was modified to add capabilities that included artifact rejection and response detection (Lasky et al., 1992Go). The distortion generated by the instrumentation was measured in 1-cc closed syringes to verify that the instrumentation did not distort the results.

The same hardware was used to generate the stimuli and digitize the responses for the auditory brainstem evoked response (ABR) recordings. A Grass P511K A.C. preamplifier amplified the responses 50,000 times and band pass filtered them with the 3 dB cut-offs at 100 Hz and 3 kHz. The roll-offs of the band pass filter were 12 dB/octave. The sampling rates to generate the click stimuli and digitize the responses were 10 kHz. Custom software controlled the recording of the response and generated the ABR stimuli, which were downloaded to the Ariel 16+ DSP board, where they were converted to analog signals.

The etymotic ER-10B ear canal microphone measured the sound pressure levels of the DPOAE stimuli. With the probe in the ear canal, the CUBeDISPTM software initially presented chirps to determine the frequency response of the canal with the probe in place. The frequency response was used to adjust the output of the etymotic ER-2 insert earphones to the specified levels. The 100-µs rarefaction click stimuli for the ABRs was calibrated as peak equivalent, sound-pressure level (peSPL on a linear scale) in a 2-cc coupler with a Bruel and Kjaer sound-level meter (Type 2230) and 4144 microphone (frequency range 2.6–8,000 Hz).

Procedures.
Testing was conducted in an isolated room in a laboratory at the University of Illinois. Background noise level in the testing room was 58 dB, C-weighted scale, 54 dB, A-weighted scale (Radio Shack Realistic Sound Level Meter, catalog # 33–2050) SPL. Rats were sedated with 0.4 ml/kg ketamine/xylazine (87:13 mg/ml) ip. Otoscopy preceded testing. No external or middle ear pathology or obstructions were identified by otoscopy in any of the rats. Subsequent to otoscopy, the rats were placed on their right sides (left ear up) and the probe was positioned in the ear canal. Electrodes were also attached. The ABR testing immediately followed the DPOAE testing.

Total test time was approximately 30–45 min per rat. The room temperature was 20°C. The rats were placed on a heating pad (medium temperature setting) to maintain body temperature, which was not measured during testing. All testing was conducted without prior knowledge of a rat’s experimental condition, to avoid potential bias.

DPOAE testing.
DPOAEs were generated by simultaneously presenting 2 sinusoids differing in frequency (the lower frequency primary is f1 and the higher frequency primary is f2) and recording the sound pressure in the sealed ear canal. Because the cochlear mechanical response to sound is nonlinear, nonlinear distortions can be used to characterize cochlear function. The 2f1–f2 distortion product is the most robust of the distortion products recorded.

Initially, 25 primary stimulus pairs were presented sequentially. These 25 stimuli differed in frequency (f2/f1 = 1.2; f2s = 20.0, 17.4, 15.1, 13.2, 11.4, 9.9, 8.6, 7.5, 6.4, 5.6, 4.9, 4.2, 3.7, 3.2, 2.7, 2.4, 2.1, 1.8, 1.5, 1.3, 1.1, 0.9, 0.8, 0.6, and 0.5 kHz). They were presented from the highest to the lowest frequency pairs. Each stimulus pair was presented for a total of 4096 artifact-free ms. The level of the lower frequency primary (L1) was 59 dB SPL and the level of the higher frequency primary (L2) was 49 dB SPL. L1 was presented 10 dB greater than L2 because larger DPOAEs are associated with those primary stimulus level differences. A fast Fourier transform (FFT) was calculated on the time-averaged response and the amplitudes of the distortion product and noise (estimated by the mean amplitude of the 3 lower- and higher-frequency bins adjacent to the distortion product) were measured. After recording the DPOAEs to the 25 primary stimulus-pairs, a second replicate 25 stimulus-pair sequence was presented. DPOAEs to stimulus pairs differing in frequency are referred to as Dpgrams, which are used to assess cochlear function loss across the range of frequencies presented.

For these data, a DPOAE was defined as a reliable response (and not noise) by a maximum-likelihood ratio test (Lasky et al., 1992Go). That test calculated the probability that the measured DPOAE came from the distribution of background noise near the DPOAE frequency recorded in the ear canal. DPOAEs unlikely (p < 0.05) to be noise were identified as responses. Both replicate DPOAEs to the same stimulus pairs had to be significant (p < 0.05) in order to be considered a response. Suprathreshold responses were determined by the averaged amplitude of the significant replicate DPOAEs to the L1 = 59 and L2 = 49 dB SPL stimuli. All rats had significant DPOAEs for f2s from 11,400 to 3200 Hz. For f2s > 11,400 Hz, the instrumentation could not generate the stimuli at the specified levels for all rats, given the resonance characteristics of their ear canals. Significant responses declined rapidly with decreasing frequency below f2 = 2051 Hz (50% at f2 = 1758 Hz, 0% by f2 = 1074 Hz). The noise levels in the rat ear canal increased with decreasing frequency, contributing to the difficulty recording DPOAEs to low-frequency stimuli. In addition, the input impedances to low frequencies were high because of small outer and middle ears.

The 2f1–f2 DPOAE thresholds were determined by reducing the levels of the primaries in 10 dB steps until no significant responses (p < 0.05 by a maximum-likelihood ratio test) could be recorded at any frequency. Replicates were recorded at each stimulus level. All other stimulus and recording parameters were kept constant. The DPOAE thresholds were defined by the lowest level f2 that elicited a significant DPOAE for both replicates.

ABRs.
The ABR stimuli were presented immediately after the DPOAEs without repositioning the probe. The ABR stimulus was a 65-dB peSPL 100-µs click (broadband stimulus) presented at a rate of 26.8/s. Replicate responses (each representing the averaged responses to 500 clicks) were recorded to all stimuli. The responses were differential voltage recordings from needle electrodes placed in the scalp at the vertex (noninverting electrode) and ipsilateral mastoid (inverting electrode). The ground was placed in the back of the neck.

Figure 1Go presents an example of replicate waveforms recorded to the 65-dB peSPL broadband-stimulus click. Four positive peaks (local amplitude maxima) were recorded in all but 2 of the rats in the first 8 ms following stimulus onset. Peak 3 was absent in the 2 rats (one control and one A1254) with missing peaks. Peak 3 and not peak 4 was judged missing by the latencies of the recorded peaks. The 4 peaks were labeled according to the nomenclature adopted by Herr et al. (1996). The average latency and amplitude (peak to following trough) of the 2 replicates were used in analyzing the data. ABR click thresholds were recorded by attenuating the click stimulus in 10-dB steps until no reproducible response could be recorded. Reproducible and nonreproducible responses were subjectively judged. A response was defined if any of the 4 peaks (invariably peak 2) was scored in both replicates. The stimulus level was raised in 5-dB increments until a response was reestablished, thus determining the ABR threshold within 5 dB. The ABR waveforms and thresholds were recorded with no knowledge of each rat’s experimental condition.



View larger version (17K):
[in this window]
[in a new window]
 
FIG. 1. An example of replicate auditory brainstem response (ABR) waveforms from a control rat in response to the 65-dB peSPL click stimulus. The solid and dashed lines represent the ABR waveform for the first and second stimulus presentations, respectively. The 4 positive peaks are labeled according to the nomenclature adopted by Herr et al. (1996).

 
Statistical analysis.
The data were analyzed by analyses of variance tests (ANOVAs) using SPSS for Windows, release 6.0 (SPSS, 1993). For the DPOAE threshold and suprathreshold responses, 2 (group) x 2 (gender) x n (f2 frequency) repeated-measures ANOVAs were calculated. Group and gender were between subjects factors and the f2 frequency was the within the subjects factor. The number of frequencies (n) included in the analyses depended on the number of rats with complete data, resulting in 10 out of 25 frequencies (11.4–3.2 kHz) being included for the suprathreshold amplitudes and 17 out of 25 frequencies (20–2.1 kHz) included for the threshold data. For the suprathreshold amplitude data, 2 (group) x 2 (gender) ANOVAs were calculated at frequencies not included in the three-way ANOVA because of missing data (see Table 1Go). The ABR amplitudes and latencies to the 65-dB peSPL clicks were analyzed separately by 2 (group) x 2 (gender) x 3 (peak) repeated-measures ANOVAs. The 3 peaks included in these analyses were 1A, 2, and 4. Peak 3 was analyzed separately by a 2 (group) x 2 (gender) ANOVA, because Peak 3 could not be reliably scored in all rats. The criterion significance level to reject the null hypothesis was 0.05. All significant results are reported. Nonsignificant results (p > 0.05) are not reported unless specifically identified as such.


View this table:
[in this window]
[in a new window]
 
TABLE 1 Frequencies Assessed, Number of Male and Female Rats Included, and Type of ANOVA Used to Analyze the DPOAE Amplitude Responses for Control (n= 15) and A1254 (n= 11) Groups
 
The responses from each rat were analyzed as independent observations even though 8 rats had siblings in the sample (5 control and 3 A1254-treated sibling pairs). Treating each rat’s response independently is misleading if the responses of siblings are correlated. Consequently, the responses of the male and female siblings from the 8 sibling pairs were correlated (using a Spearman rho because of the small sample size) for each variable. Of the 27 relevant DPOAE variables (the 10 suprathreshold frequencies and 17 threshold frequencies with no missing data), only one correlation coefficient was significant at the 0.05 level. It was a negative correlation. (The ABR correlations were less relevant, because, as will become evident below, none of the ABR results for the total sample analyses proved statistically significant.) Nevertheless, all analyses were repeated including only the 16 rats from the 8 litters contributing 2 rats (one male and one female) per litter. In these analyses of variance, litter rather than individual rat was the unit of analysis, with group a between subject (litter) factor and gender and frequency within subject (litter) factors. The results of these analyses were consistent with the analyses reported below for the total sample, further suggesting that the correlated responses of siblings did not significantly affect the results. Even when the results of these analyses did not achieve the conventional 0.05 level of significance, the trends in the data were in the same direction as for the total sample. These analyses are presented in the Results section as footnotes, referring to the statistically significant results for the total sample.


    RESULTS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
DPOAEs
Figure 2Go presents DPOAE amplitudes as a function of f2 frequency for control and A1254-exposed rats. Control DPOAE amplitudes were larger than those of A1254-exposed rats. The differences diminished with increasing frequency and were larger for females than males.



View larger version (18K):
[in this window]
[in a new window]
 
FIG. 2. Group mean DPOAE amplitudes (in dB SPL) presented as a function of f2 frequency for the control (filled boxes) and A1254-exposed (open boxes) rats. The data are plotted as mean ± SEM.

 
In order to determine whether the trends observed in Figure 1Go were reliable, a 2 (group) x 2 (gender) x 10 (f2 frequency), repeated measures ANOVA was calculated for the frequencies for which we had data from all animals (11.4–3.2 kHz, see Table 1Go). Both the group (F(1,22) = 8.52; p = 0.008) and frequency (F(9,198) = 80.41, p < 0.001) main effects were significant. There was also a significant group x gender interaction (F(1,22) = 6.50, p = 0.018) due to larger group differences in females than males. Finally, there was a significant group x frequency interaction (F(9,198) = 4.85, p < 0.001) resulting from larger group differences at the lower frequencies. The group x gender x frequency three-way interaction was not significant (F(9,198) = 1.39, p = 0.194).2

ANOVAs, 2 (group) x 2 (gender), were calculated for frequencies that had adequate sample sizes (more than half of the subjects in each group with responses) but were not included in the three-way ANOVA reported above because of missing data (see Table 1Go). At 17.4, 15.1, and 13.2 kHz, A1254 rats had smaller DPOAEs than control rats although the differences were significant only at 15.1 kHz (F(1,14) = 5.10; p = 0.040). No other main effects or interactions were significant at these frequencies. There were no data at 20.0 kHz, because our instrumentation could not produce the specified sound levels at that frequency. There were significant group main effects at each of the 2 lowest frequencies (2.7 kHz: F(1,17) = 5.82, p = 0.027; and 2.4 kHz: F(1,19) = 21.75, p < 0.001) with adequate sample sizes. However, neither the gender main effect nor the group x gender interaction was significant at these frequencies, suggesting that both sexes were equally affected by A1254 at the lowest frequencies tested.3

The group differences at the 2 lowest frequencies were a little over 8 dB. Three of the 11 A1254-treated rats at each of these frequencies were excluded because they failed to elicit a reliable DPOAE by a maximum-likelihood ratio test. In contrast, none of the control rats for the 2.4 kHz stimulus and 2 of the 15 control rats were excluded for the 2.7 kHz stimulus. Because more A1254-treated rats failed to elicit reliable DPOAEs, the 8-dB loss may be an underestimate of the actual size of the A1254-induced loss.

The DPOAE threshold data for males and females are presented in Figure 3Go. DPOAE thresholds in the A1254-exposed rats were elevated relative to controls, particularly at the lowest frequencies tested. The gender differences were not as large as they were for DPOAE suprathreshold amplitudes.



View larger version (19K):
[in this window]
[in a new window]
 
FIG. 3. Group mean DPOAE thresholds (in dB SPL) presented as a function of f2 frequency for the control (filled boxes) and A1254-exposed (open boxes) rats. The data are plotted as mean ± SEM.

 
The DPOAE-threshold data were assessed statistically by a 2 (group) x 2 (gender) x 17 (f2 frequency: 20–2.1 kHz) repeated measures ANOVA. The group (F(1,22) = 7.69, p = 0.011) and frequency (F(16,352) = 65.53, p < 0.001) main effects were significant, as was the group x frequency interaction (F(16,352) = 1.80; p = 0.03). Neither the group x gender nor the group x gender x frequency interactions were significant for DPOAE thresholds.4 The significant group x frequency interaction was explored by 2(group) one-way ANOVAs at each f2 frequency. Significant group differences were recorded at 3.2–2.1 kHz (F(1,24) = 4.73, p = 0.040; F(1,24) = 12.51, p = 0.002; F(1,24) = 6.77, p = 0.015; and F(1,24) = 6.90, p = 0.015, respectively) and also at 5.6 kHz (F(1,24) = 6.97, p = 0.014) but not at the other frequencies.5

Again, at the lowest frequencies, the group differences may have been underestimates, because reliable DPOAEs could not be recorded at the highest stimulus levels presented. For rats that failed to produce a reliable DPOAE at the highest stimulus levels, we assigned a threshold of 54-dB SPL, 5 dB higher than the most intense L2 presented to assess thresholds. Thus, these rats were assigned a conservative estimate of their threshold, which happened more often for the A1254-exposed rats than controls for the 2.1 (1 of 15 controls and 6 of 11 A1254-exposed rats) and the 2.4 and 2.7 kHz stimuli, as described above. Therefore, the thresholds for the A1254-exposed rats at these frequencies may have been underestimated.

Auditory Evoked Brainstem Responses
Figure 4Go presents the latencies and amplitudes for the 4 peaks to the 65-dB peSPL stimulus. Data were not presented separately by gender, because there were no gender-related differences associated with A1254 exposure on these measures. The differences in peak latencies and amplitudes between the A1254-exposed rats and the controls were small and did not consistently favor either group. Separate 2 (group) x 2 (gender) x 3 (peak: 1A, 2, and 4) repeated-measures ANOVAs were calculated for the ABR latency and amplitude data. The only significant effect was the peak main effect (F(2,40) = 2358.41, p < 0.001 for latency and F(2,40) = 157.35, p < 0.001 for amplitude). Peak 3 could not be reliably scored in all rats and was not included in these analyses. Instead, separate 2 (group) x 2 (gender) ANOVAs were calculated for peak 3 latency and amplitude. Neither the main effects nor the interaction were significant for those analyses.



View larger version (16K):
[in this window]
[in a new window]
 
FIG. 4. Mean ABR peak latencies in milliseconds (ms) and amplitudes (in microvolts) to the 65-dB peSPL click stimulus for the control (dark bars) and A1254-exposed (light bars) rats. The data are plotted as mean ± SEM.

 
The ABR thresholds are presented in Figure 5Go, by exposure group. Again, gender was not a significant factor, and therefore, the data were not plotted separately by gender. A 2 (group) x 2 (gender) ANOVA was calculated. ABR thresholds for A1254 rats were not found to be significantly elevated relative to controls, (F(1,15) = 2.56, p = 0.131). The main effect of gender (F(1,15) = 1.50, p = 0.240) and the group x gender interaction (F(1,15) = 1.50, p = 0.240) also failed to reach the 0.05 level of significance.



View larger version (19K):
[in this window]
[in a new window]
 
FIG. 5. Mean ABR thresholds (in dB peSPL) for the control (dark bar) and A1254-exposed (light bar) rats. The data are plotted as mean ± SEM.

 

    DISCUSSION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
A Cochlear Loss
Distortion product otoacoustic emission (DPOAE) amplitudes were reduced and DPOAE thresholds elevated in adult rats exposed early in development to A1254. Deficits were most pronounced at the lowest frequencies tested. These results are consistent with previous studies. They implicate the cochlea and complement histological findings of A1254-induced outer hair cell loss (Crofton et al., 2000aGo). Nevertheless, some of our results do differ from results previously reported.

Frequency Effects
Our results confirm that A1254-induced hearing loss is greater at low frequencies. However, we also reported losses at frequencies for which Goldey, Crofton, and colleagues (Goldey and Crofton, 1995a; Crofton et al., 2000bGo) have reported no significant loss. The threshold elevations recorded at frequencies greater than 1 kHz in this study were smaller than those reported, but yet, were statistically significant because of the smaller variability in DPOAE measurements. The standard errors of the mean for DPOAE amplitudes and DPOAE thresholds in this study were about half those for behavioral thresholds in the reflex audiometry paradigm used by Goldey, Crofton, and colleagues (Goldey and Crofton, 1995a; Crofton et al., 2000bGo).

Gender Effects
While gender-specific effects on neurobehavioral function have been reported following perinatal PCB exposure (Davenport, et al., 1976Go; Geller et al., 2001Go; Widholm et al., 2001Go), previous studies have shown that males and females are equally affected with respect to thyroid hormone reductions and hearing loss (Crofton and Rice, 1999Go; Crofton et al., 2000bGo; Goldey et al., 1995aGo; Rice and Hayward, 1999Go). Similarly, in this study there was no gender effect on DPOAE amplitudes at the lowest frequencies reliably assessed (2.7 and 2.4 kHz), consistent with the results of prior studies. Gender differences were only recorded at higher frequencies. Why there would be a gender effect at higher but not lower frequencies is not obvious but could be explained if females were adversely affected by A1254 earlier in development. Alternatively, this result could be explained by chance (a type I error). Additional research is needed to interpret the recorded gender effects.

Effects on the ABR
The discrepancy between the ABR results in this study (no effect) and those of Herr et al. (1996, reduced amplitudes and prolonged latencies) may be explained, in part, by the frequencies and levels of the stimuli presented. Herr et al. used equal energy pure tone pips. We used a 100-µs click, a broadband stimulus with a null at 10,000 Hz. However, the acoustic waveform is shaped by the auditory transducer, and therefore, the spectrum of the click presented in this study rolls off at both the low- and high-frequency ends. The effective stimulus is also determined by the anatomy and physiology of the rat. ABRs to 100-µs clicks correlate best with hearing thresholds from 2000 to 4000 Hz (Stapells et al., 1994Go) in humans. Rat outer and middle ears have higher resonance frequencies than humans (Hemila et al., 1995Go). Rats also have more sensitive higher-frequency hearing. Thus, it is likely that rat ABRs to clicks are generated by higher frequencies than in the human. At the stimulus level (65-dB peSPL) and effective frequencies (probably higher than 2–4 kHz) presented in this study, Herr et al. (1996) reported no effect of A1254 exposure on the ABR, consistent with the results of this study.

The rational for recording the ABR was to determine whether there were A1254-induced neural deficits indicated by a greater effect on later peaks than on the earlier peaks. Neural insult is unlikely to be frequency-specific, because there is little reason to expect that A1254 exposure would only affect neurons relaying information from restricted regions of the cochlea. Thus, frequency-specific stimuli were not presented. There is little evidence from this study or the Herr et al. (1996) study to indicate that A1254 exposure affects neural transmission in the auditory nerve and brainstem.

Implications for Humans
Species differences in the timing of cochlear development are significant enough to warrant caution in generalizing results from rodent models to humans. Cochlear functional development occurs postnatally in rats. In contrast, the functional maturation of the human cochlea begins about the 20th week postconception and is completed by the 35th week (Pujol and Uziel, 1988Go; Sulik, 1995Go). Many of the anatomical changes that occur during that time interval concern the outer hair cells (the targets of PCB ototoxicity in rats and the likely generators of DPOAEs) and their support cells. Assuming that the critical period for PCB-induced auditory insult corresponds to the period of functional maturation of the cochlea, the human would be most vulnerable to PCB ototoxicity during the last half of pregnancy when the only route of PCB exposure is via placental transfer. A recent epidemiological study suggests that prenatal PCB exposure may be associated with subtle auditory deficits in children (Grandjean et al., 2001Go). Based on these findings, further epidemiological studies of auditory function in children exposed to PCBs prenatally are warranted. Recording otoacoustic emissions could contribute significantly to that research. Otoacoustic emissions reflect the hypothesized site of PCB damage, and they can be recorded reliably and efficiently in humans of all ages (Robinette and Glattke, 1997Go).


    ACKNOWLEDGMENTS
 
The research presented in this document has been funded wholly (or in part) by the U.S. Environmental Protection Agency. This research was supported by a grant from NIEHS (F32-ES05885) and a grant from the Center for Clinical Research and Evidence-Based Medicine.


    NOTES
 
This paper has been subjected to review by the National Health and Environmental Effects Research Laboratory, U.S. Environmental Protection Agency, and approved for publication. Approval does not signify that the contents reflect the views of the Agency, nor does mention of trade names or commercial products constitute endorsement or recommendation for use.

1 To whom correspondence should be addressed. Fax: (713) 500-0519. E-mail: robert.e.lasky{at}uth.tmc.edu. Back

2 A 2 (group) x 2 (gender) x 10 (f2 frequency) repeated-measures ANOVA was calculated for the 8 litters, each contributing a male and a female rat. Group was a between-subjects (litters) factor and gender and frequency were within-subjects (litters) factors. The group (F(1,6) = 6.69, p = 0.041) and the frequency (F(9,54) = 55.40, p < 0.001) main effects were significant. The group x gender interaction approached significance (F(1,6) = 3.70, p = 0.103). The controls had larger amplitude DPOAEs than A1254-exposed rats. That difference was more apparent for females and at lower frequencies. The trends in the data were the same as those observed in the total sample. Back

3 ANOVAs, 2 (group) x 2 (gender), with gender as a within-subjects (litters) factor, were calculated at 17.4, 15.1, 13.2, 2.7, 2.4, and 2.1 kHz for the 8 litters, each litter contributing a male and a female rat. Only 2 effects were significant: the group main effect for the 2.7 (F(1,3) = 18.09, p = 0.024) and 2.4 (F(1,5) = 7.32, p = 0.042) kHz stimuli. Group x gender interactions were not significant at the lowest frequencies analyzed (F(1,3) = 0.14, p = 0.737 for f2 = 2.7 kHz; F(1,5) = 0.13, p = 0.734 for f2 = 2.4 kHz; F(1,3) = 0.47, p = 0.544 for f2 = 2.1 kHz). Back

4 A 2 (group) x 2 (gender) x 17 (f2 frequency) repeated-measures ANOVA was calculated for the 8 litters, each contributing a male and a female rat. Group was a between-subjects (litters) factor and gender and frequency were within-subjects (litters) factors. The frequency (F(16,96) = 44.23, p < 0.001) main effect and the group x frequency interaction (F(16,96) = 1.81, p = 0.040) were significant. Elevated thresholds for the A1254-treated rats were more apparent at low than high frequencies, as they were for the total sample. Back

5 The complementary analyses on the sub-sample of rats from the 8 litters, contributing both a male and a female each, were 2 (group) x 2 (gender) ANOVAs. Gender was a within-subjects (litters) factor. Only the main effect for group (F(1,6) = 11.94, p = 0.014) at 3.2 kHz was significant, although the group main effect at 2.7 kHz approached significance (F(1,6) = 4.77, p = 0.072). Back


    REFERENCES
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Clarkson, T. W. (1995). Environmental contaminants in the food chain. Am. J. Clin. Nutr. 61(Suppl. 3), 682–686S.

Crofton, K. M., Ding, D.-L., Padich, R., Taylor, M., and Henderson, D. (2000a). Hearing loss, following exposure during development to polychlorinated biphenyls: A cochlear site of action. Hear. Res. 144, 196–204.[ISI][Medline]

Crofton, K.M., Kodavanti, P.R.S., Derr-Yellin, E.C., Casey, A.C., and Kehn, L.S. (2000b). PCBs, thyroid hormones, and ototoxicity in rats: Cross-fostering experiments demonstrate the impact of postnatal lactation exposure. Toxicol. Sci. 57, 131–140.[Abstract/Free Full Text]

Crofton, K. M., and Rice, D. C. (1999). Low-frequency hearing loss following perinatal exposure to 3,3',4,4',5-pentachlorobiphenyl (PCB 126) in rats. Neurotoxicol. Teratol. 21, 299–301.[ISI][Medline]

Davenport, J. W., Gonzalez, L. M., Hennies, R. S., and Hagquist, W. M. (1976). Severity and timing of early thyroid deficiency as factors in the induction of learning disorders in rats. Horm. Behav. 7, 139–157.[ISI][Medline]

Geller, A. M., Oshiro, W. M., Haykal-Coates, N., Kodavanti, P. R. S., and Bushnell, P. J. (2001). Gender-dependent behavioral and sensory effects of a commercial mixture of polychlorinated biphenyls (Aroclor 1254) in rats. Toxicol. Sci. 59, 268–277.[Abstract/Free Full Text]

Goldey, E. S., and Crofton, K. M. (1998). Thyroxine replacement attenuates hypothyroxinemia, hearing loss, and motor deficits following developmental exposure to Aroclor 1254 in rats. Toxicol. Sci. 45, 94–105.[Abstract]

Goldey, E. S., Kehn, L. S., Lau, C., Rehnberg, G. L., and Crofton, K. M. (1995a). Developmental exposure to polychlorinated biphenyls (Aroclor 1254) reduces circulating thyroid hormone concentrations and causes hearing deficits in rats. Toxicol. Appl. Pharmacol. 135, 77–88.[ISI][Medline]

Goldey, E. S., Kehn, L. S., Rehnberg, G. L., and Crofton, K. M. (1995b). Effects of developmental hypothyroidism on auditory and motor function in the rat. Toxicol. Appl. Pharmacol. 135, 67–76.[ISI][Medline]

Grandjean, P., Weihe, P., Burse, V. W., Needham, L. L., Storr-Hansen, E., Heinzow, B., Debes, F., Murata, K., Simonsen, H., Ellefsen, P., Budtz-Jorgensen, E., Keiding, N., and White, R. F. (2001). Neurobehavioral deficits associated with PCB in 7-year-old children prenatally exposed to seafood neurotoxicants. Neurotoxicol. Teratol. 23, 305–317.[ISI][Medline]

Hemila, S., Nummela, S., and Reuter, T. (1995). What middle ear parameters tell about impedance matching and high frequency hearing. Hear. Res. 85, 31–44.[ISI][Medline]

Herr, D. W., Goldey, E. S., and Crofton, K. M. (1996). Developmental exposure to Aroclor 1254 produces low-frequency alterations in adult rat brainstem auditory evoked responses. Fundam. Appl. Toxicol. 33, 120–128.[ISI][Medline]

Jacobson, J. L., Fein, G. G., Jacobson, S. W., Schwartz, P. M., and Dowler, J. K. (1984). The transfer of polychlorinated biphenyls (PCBs) and polybrominated biphenyls (PBBs) across the human placenta and into maternal milk. Am. J Public. Health, 74, 378–379.[Abstract]

Lasky, R. E., Perlman, J., and Hecox, K. E. (1992). Distortion-product otoacoustic emissions in human newborns and adults. Ear Hear. 13, 430–441.[ISI][Medline]

Pujol, R., and Uziel, A. (1988). Auditory development: Peripheral aspects. In Handbook of Human Growth and Developmental Biology, Vol IB (E. Meisami and P.S. Timiras, Eds.), pp. 109–130. CRC Press, Boca Raton, Florida.

Rice, D. C., and Hayward, S. (1999). Effects of exposure to 3,3',4,4',5-pentachlorobiphenyl (PCB 126) throughout gestation and lactation on behavior (concurrent random interval-random interval and progressive ratio performance) in rats. Neurotoxicol. Teratol. 21, 679–687.[ISI][Medline]

Robinette, M. S., and Glattke, T. J. (1997). Otoacoustic Emissions: Clinical Applications. Thieme, New York.

Roegge, C. S., Seo, B. W., Crofton, K. M., and Schantz, S. L. (2000). Gestational-lactational exposure to Aroclor 1254 impairs radial-arm maze performance in male rats. Toxicol. Sci. 57, 121–130.[Abstract/Free Full Text]

Stapells, D. R., Picton, T. W., and Durieux-Smith, A. (1994). Electrophysiologic measures of frequency-specific auditory function. In Principles and Applications in Auditory Evoked Potentials (J.T. Jacobson, Ed.), pp. 251–283. Allyn and Bacon, Boston.

Sulik, K. K. (1995). Embryology of the Ear. In Hereditary Hearing Loss and Its Syndromes (R. J. Gorlin, H. V. Toriello, and M. M. Cohen, Eds.), pp. 23–42. Oxford University Press, New York.

Uziel, A. (1986). Periods of sensitivity to thyroid hormone during the development of the organ of Corti. Acta Otolaryngol. Suppl. 429, 23–27.[Medline]

Widholm, J. J., Clarkson, G. B., Strupp, B. J., Crofton, K. M., Seegal, R. F., and Schantz, S. L. (2001). Spatial reversal learning in Aroclor 1254-exposed rats: Sex-specific deficits in associative ability and inhibitory control. Toxicol. Appl. Pharmacol. 174, 188–198.[ISI][Medline]

Yakushiji, T., Watanabe, I., Kuwabara, K., Tanaka, R., Kashimoto, T., Kunita, N., and Hara, I. (1984). Postnatal transfer of PCBs from exposed mothers to their babies: Influence on breast-feeding. Arch. Environ. Health 39, 368–375.[ISI][Medline]