Non-FM calls

3.1.2. Females

3.1.2. Females

3.1.1. Males

fects. #: Number.

3.1.1. Males

3.1.2. Females

3.1.1. Males

3.1.2. Females

3.1.2. Females

3.1.2. Females

3.1.2. Females

3.1.2. Females

trols (p = 0.008; Figure 2a).

3.1.2. Females

3.1.1. Males

3.1.2. Females

trols (p = 0.008; Figure 2a).

trols (p = 0.008; Figure 2a).

the F3 generation, regardless of F0 treatment (Table 2).

Few somatic changes were detected in the measured outcomes for EDC-lineage rats. A trend was observed for an F0 treatment effect (EB slightly larger than DMSO) in normalized adrenal weights of F1 (F(2,33) = 2.997, p = 0.064) and F3 (F(2,25) = 2.856, p = 0.076) males (Table 2). Similarly, no changes were detected in serum testosterone levels (F3 hormones not measured) or normalized testes weight. However, in the F2 generation, we observed a significant effect of treatment on male body weight (H(2) = 9.054, p = 0.011) with A1221-lineage males having greater average body weights compared to DMSO controls (p = 0.008; Figure 2a). 3.1.1. Males Few somatic changes were detected in the measured outcomes for EDC-lineage rats. A trend was observed for an F0 treatment effect (EB slightly larger than DMSO) in normalized adrenal weights of F1 (F(2,33) = 2.997, p = 0.064) and F3 (F(2,25) = 2.856, p = 0.076) males (Table 2). Similarly, no changes were detected in serum testosterone levels (F3 hormones not measured) or normalized testes weight. However, in the F2 generation, we observed a significant effect of treatment on male body weight (H(2) = 9.054, p = 0.011) with A1221-lineage males having greater average body weights compared to DMSO conmales (Table 2). Similarly, no changes were detected in serum testosterone levels (F3 hormones not measured) or normalized testes weight. However, in the F2 generation, we observed a significant effect of treatment on male body weight (H(2) = 9.054, p = 0.011) with A1221-lineage males having greater average body weights compared to DMSO controls (p = 0.008; Figure 2a). In females, we found an effect of F0 treatment on normalized uterine weight in the Decreased compared to DMSO (or UNT for Partner Preference). Increased compared to DMSO (or UNT for Partner Preference). FM: Frequency modulated. n.s. and -: No significant effects. #: Number. 3.1. Transgenerational Somatic Changes Few somatic changes were detected in the measured outcomes for EDC-lineage rats. Few somatic changes were detected in the measured outcomes for EDC-lineage rats. A trend was observed for an F0 treatment effect (EB slightly larger than DMSO) in normalized adrenal weights of F1 (F(2,33) = 2.997, p = 0.064) and F3 (F(2,25) = 2.856, p = 0.076) males (Table 2). Similarly, no changes were detected in serum testosterone levels (F3 hormones not measured) or normalized testes weight. However, in the F2 generation, we observed a significant effect of treatment on male body weight (H(2) = 9.054, p = 0.011) with A1221-lineage males having greater average body weights compared to DMSO controls (p = 0.008; Figure 2a). Few somatic changes were detected in the measured outcomes for EDC-lineage rats. A trend was observed for an F0 treatment effect (EB slightly larger than DMSO) in normalized adrenal weights of F1 (F(2,33) = 2.997, p = 0.064) and F3 (F(2,25) = 2.856, p = 0.076) males (Table 2). Similarly, no changes were detected in serum testosterone levels (F3 hormones not measured) or normalized testes weight. However, in the F2 generation, we observed a significant effect of treatment on male body weight (H(2) = 9.054, p = 0.011) with A1221-lineage males having greater average body weights compared to DMSO controls (p = 0.008; Figure 2a). 3.1. Transgenerational Somatic Changes Few somatic changes were detected in the measured outcomes for EDC-lineage rats. A trend was observed for an F0 treatment effect (EB slightly larger than DMSO) in normalized adrenal weights of F1 (F(2,33) = 2.997, p = 0.064) and F3 (F(2,25) = 2.856, p = 0.076) males (Table 2). Similarly, no changes were detected in serum testosterone levels (F3 hormones not measured) or normalized testes weight. However, in the F2 generation, we observed a significant effect of treatment on male body weight (H(2) = 9.054, p = 0.011) 3.1. Transgenerational Somatic Changes 3.1.1. Males Few somatic changes were detected in the measured outcomes for EDC-lineage rats. A trend was observed for an F0 treatment effect (EB slightly larger than DMSO) in normalized adrenal weights of F1 (F(2,33) = 2.997, p = 0.064) and F3 (F(2,25) = 2.856, p = 0.076) males (Table 2). Similarly, no changes were detected in serum testosterone levels (F3 hormones not measured) or normalized testes weight. However, in the F2 generation, we observed a significant effect of treatment on male body weight (H(2) = 9.054, p = 0.011) Ultrasonic Vocalizations (F3) (combined for the sexes) EB A1221 # Total calls # Non-FM calls Decreased compared to DMSO (or UNT for Partner Preference). Increased compared to DMSO (or UNT for Partner Preference). FM: Frequency modulated. n.s. and -: No significant effects. #: Number. 3.1. Transgenerational Somatic Changes Few somatic changes were detected in the measured outcomes for EDC-lineage rats. A trend was observed for an F0 treatment effect (EB slightly larger than DMSO) in normalized adrenal weights of F1 (F(2,33) = 2.997, p = 0.064) and F3 (F(2,25) = 2.856, p = 0.076) males (Table 2). Similarly, no changes were detected in serum testosterone levels (F3 hormones not measured) or normalized testes weight. However, in the F2 generation, we DMSO (or UNT for Partner Preference). FM: Frequency modulated. n.s. and -: No significant ef-3.1. Transgenerational Somatic Changes Few somatic changes were detected in the measured outcomes for EDC-lineage rats. A trend was observed for an F0 treatment effect (EB slightly larger than DMSO) in normalized adrenal weights of F1 (F(2,33) = 2.997, p = 0.064) and F3 (F(2,25) = 2.856, p = 0.076) DMSO (or UNT for Partner Preference). FM: Frequency modulated. n.s. and -: No significant effects. #: Number. 3.1. Transgenerational Somatic Changes 3.1.1. Males Few somatic changes were detected in the measured outcomes for EDC-lineage rats. A trend was observed for an F0 treatment effect (EB slightly larger than DMSO) in normalized adrenal weights of F1 (F(2,33) = 2.997, p = 0.064) and F3 (F(2,25) = 2.856, p = 0.076) # Rearing bouts (trend) - - - Time at Plexiglas (trend) - # Plexiglas bouts - - - Ultrasonic Vocalizations (F3) (combined for the sexes) EB A1221 # Total calls # Non-FM calls Decreased compared to DMSO (or UNT for Partner Preference). Increased compared to DMSO (or UNT for Partner Preference). FM: Frequency modulated. n.s. and -: No significant ef-3.1. Transgenerational Somatic Changes Few somatic changes were detected in the measured outcomes for EDC-lineage rats. A trend was observed for an F0 treatment effect (EB slightly larger than DMSO) in nor-(combined for the sexes) EB A1221 Decreased compared to DMSO (or UNT for Partner Preference). Increased compared to DMSO (or UNT for Partner Preference). FM: Frequency modulated. n.s. and -: No significant ef-3.1. Transgenerational Somatic Changes (combined for the sexes) EB A1221 # Non-FM calls Decreased compared to DMSO (or UNT for Partner Preference). Increased compared to DMSO (or UNT for Partner Preference). FM: Frequency modulated. n.s. and -: No significant effects. #: Number. 3.1. Transgenerational Somatic Changes Ultrasonic Vocalizations (F3) (combined for the sexes) EB A1221 Decreased compared to DMSO (or UNT for Partner Preference). Increased compared to DMSO (or UNT for Partner Preference). FM: Frequency modulated. n.s. and -: No significant ef-# Plexiglas bouts - - - Ultrasonic Vocalizations (F3) (combined for the sexes) EB A1221 Decreased compared to DMSO (or UNT for Partner Preference). Increased compared to DMSO (or UNT for Partner Preference). FM: Frequency modulated. n.s. and -: No significant ef-Mating Behaviors (F1, F2) Mount frequency (F2) - - - Lordosis latency - (F1) - - Partner Preference Behavior (F3) Total time active (trend) - - Time rearing (trend) - # Rearing bouts (trend) - - - Few somatic changes were detected in the measured outcomes for EDC-lineage rats. A trend was observed for an F0 treatment effect (EB slightly larger than DMSO) in normalized adrenal weights of F1 (F(2,33) = 2.997, *p* = 0.064) and F3 (F(2,25) = 2.856, *p* = 0.076) males (Table 2). Similarly, no changes were detected in serum testosterone levels (F3 hormones not measured) or normalized testes weight. However, in the F2 generation, we observed a significant effect of treatment on male body weight (H(2) = 9.054, *p* = 0.011) with A1221 lineage males having greater average body weights compared to DMSO controls (*p* = 0.008; Figure 2a).

#### trols (p = 0.008; Figure 2a). F2 generation (H(2) = 6.434; p = 0.040), in which A1221-lineage females had greater uterine weights compared to EB (Dunn's post hoc, p = 0.044; Table 2). No effect was found for A trend was observed for an F0 treatment effect (EB slightly larger than DMSO) in norwith A1221-lineage males having greater average body weights compared to DMSO controls (p = 0.008; Figure 2a). with A1221-lineage males having greater average body weights compared to DMSO controls (p = 0.008; Figure 2a). 3.1. Transgenerational Somatic Changes observed a significant effect of treatment on male body weight (H(2) = 9.054, p = 0.011) males (Table 2). Similarly, no changes were detected in serum testosterone levels (F3 hormones not measured) or normalized testes weight. However, in the F2 generation, we males (Table 2). Similarly, no changes were detected in serum testosterone levels (F3 hormones not measured) or normalized testes weight. However, in the F2 generation, we DMSO (or UNT for Partner Preference). FM: Frequency modulated. n.s. and -: No significant efmalized adrenal weights of F1 (F(2,33) = 2.997, p = 0.064) and F3 (F(2,25) = 2.856, p = 0.076) 3.1.1. Males 3.1. Transgenerational Somatic Changes Time at Plexiglas (trend) - 3.1.2. Females

the F3 generation, regardless of F0 treatment (Table 2).

the F3 generation, regardless of F0 treatment (Table 2).

the F3 generation, regardless of F0 treatment (Table 2).

the F3 generation, regardless of F0 treatment (Table 2).

the F3 generation, regardless of F0 treatment (Table 2).

the F3 generation, regardless of F0 treatment (Table 2).

the F3 generation, regardless of F0 treatment (Table 2).

3.1.1. Males

3.1.2. Females

3.1.2. Females

F2 generation (H(2) = 6.434; p = 0.040), in which A1221-lineage females had greater uterine weights compared to EB (Dunn's post hoc, p = 0.044; Table 2). No effect was found for female body weight (Figure 2b), normalized adrenal weight, normalized ovarian weight or serum estradiol (F3 hormones not measured). Hormone priming for the sociosexual tests also resulted in an expected increase in normalized ovarian and uterine weights in

3.1.2. Females

mones not measured) or normalized testes weight. However, in the F2 generation, we observed a significant effect of treatment on male body weight (H(2) = 9.054, p = 0.011) with A1221-lineage males having greater average body weights compared to DMSO con-

fects. #: Number.

3.1.2. Females

trols (p = 0.008; Figure 2a).

In females, we found an effect of F0 treatment on normalized uterine weight in the F2 generation (H(2) = 6.434; p = 0.040), in which A1221-lineage females had greater uterine weights compared to EB (Dunn's post hoc, p = 0.044; Table 2). No effect was found for female body weight (Figure 2b), normalized adrenal weight, normalized ovarian weight or serum estradiol (F3 hormones not measured). Hormone priming for the sociosexual tests also resulted in an expected increase in normalized ovarian and uterine weights in the F3 generation, regardless of F0 treatment (Table 2). 3.1.2. Females In females, we found an effect of F0 treatment on normalized uterine weight in the F2 generation (H(2) = 6.434; p = 0.040), in which A1221-lineage females had greater uterine weights compared to EB (Dunn's post hoc, p = 0.044; Table 2). No effect was found for female body weight (Figure 2b), normalized adrenal weight, normalized ovarian weight or serum estradiol (F3 hormones not measured). Hormone priming for the sociosexual female body weight (Figure 2b), normalized adrenal weight, normalized ovarian weight or serum estradiol (F3 hormones not measured). Hormone priming for the sociosexual tests also resulted in an expected increase in normalized ovarian and uterine weights in the F3 generation, regardless of F0 treatment (Table 2). malized adrenal weights of F1 (F(2,33) = 2.997, p = 0.064) and F3 (F(2,25) = 2.856, p = 0.076) males (Table 2). Similarly, no changes were detected in serum testosterone levels (F3 hormones not measured) or normalized testes weight. However, in the F2 generation, we observed a significant effect of treatment on male body weight (H(2) = 9.054, p = 0.011) with A1221-lineage males having greater average body weights compared to DMSO controls (p = 0.008; Figure 2a). In females, we found an effect of F0 treatment on normalized uterine weight in the F2 generation (H(2) = 6.434; p = 0.040), in which A1221-lineage females had greater uterine weights compared to EB (Dunn's post hoc, p = 0.044; Table 2). No effect was found for female body weight (Figure 2b), normalized adrenal weight, normalized ovarian weight or serum estradiol (F3 hormones not measured). Hormone priming for the sociosexual tests also resulted in an expected increase in normalized ovarian and uterine weights in 3.1.2. Females In females, we found an effect of F0 treatment on normalized uterine weight in the F2 generation (H(2) = 6.434; p = 0.040), in which A1221-lineage females had greater uterine weights compared to EB (Dunn's post hoc, p = 0.044; Table 2). No effect was found for female body weight (Figure 2b), normalized adrenal weight, normalized ovarian weight or serum estradiol (F3 hormones not measured). Hormone priming for the sociosexual tests also resulted in an expected increase in normalized ovarian and uterine weights in In females, we found an effect of F0 treatment on normalized uterine weight in the F2 generation (H(2) = 6.434; p = 0.040), in which A1221-lineage females had greater uterine weights compared to EB (Dunn's post hoc, p = 0.044; Table 2). No effect was found for female body weight (Figure 2b), normalized adrenal weight, normalized ovarian weight or serum estradiol (F3 hormones not measured). Hormone priming for the sociosexual 3.1.2. Females In females, we found an effect of F0 treatment on normalized uterine weight in the F2 generation (H(2) = 6.434; p = 0.040), in which A1221-lineage females had greater uterine weights compared to EB (Dunn's post hoc, p = 0.044; Table 2). No effect was found for female body weight (Figure 2b), normalized adrenal weight, normalized ovarian weight or serum estradiol (F3 hormones not measured). Hormone priming for the sociosexual 3.1.1. Males Few somatic changes were detected in the measured outcomes for EDC-lineage rats. A trend was observed for an F0 treatment effect (EB slightly larger than DMSO) in normalized adrenal weights of F1 (F(2,33) = 2.997, p = 0.064) and F3 (F(2,25) = 2.856, p = 0.076) males (Table 2). Similarly, no changes were detected in serum testosterone levels (F3 hormones not measured) or normalized testes weight. However, in the F2 generation, we with A1221-lineage males having greater average body weights compared to DMSO controls (p = 0.008; Figure 2a). In females, we found an effect of F0 treatment on normalized uterine weight in the F2 generation (H(2) = 6.434; p = 0.040), in which A1221-lineage females had greater uterine weights compared to EB (Dunn's post hoc, p = 0.044; Table 2). No effect was found for observed a significant effect of treatment on male body weight (H(2) = 9.054, p = 0.011) with A1221-lineage males having greater average body weights compared to DMSO controls (p = 0.008; Figure 2a). In females, we found an effect of F0 treatment on normalized uterine weight in the F2 generation (H(2) = 6.434; p = 0.040), in which A1221-lineage females had greater uterine observed a significant effect of treatment on male body weight (H(2) = 9.054, p = 0.011) with A1221-lineage males having greater average body weights compared to DMSO controls (p = 0.008; Figure 2a). 3.1.2. Females In females, we found an effect of F0 treatment on normalized uterine weight in the F2 generation (H(2) = 6.434; p = 0.040), in which A1221-lineage females had greater uterine fects. #: Number. 3.1. Transgenerational Somatic Changes 3.1.1. Males Few somatic changes were detected in the measured outcomes for EDC-lineage rats. A trend was observed for an F0 treatment effect (EB slightly larger than DMSO) in normalized adrenal weights of F1 (F(2,33) = 2.997, p = 0.064) and F3 (F(2,25) = 2.856, p = 0.076) males (Table 2). Similarly, no changes were detected in serum testosterone levels (F3 hormones not measured) or normalized testes weight. However, in the F2 generation, we observed a significant effect of treatment on male body weight (H(2) = 9.054, p = 0.011) with A1221-lineage males having greater average body weights compared to DMSO controls (p = 0.008; Figure 2a). In females, we found an effect of F0 treatment on normalized uterine weight in the Few somatic changes were detected in the measured outcomes for EDC-lineage rats. A trend was observed for an F0 treatment effect (EB slightly larger than DMSO) in normalized adrenal weights of F1 (F(2,33) = 2.997, p = 0.064) and F3 (F(2,25) = 2.856, p = 0.076) males (Table 2). Similarly, no changes were detected in serum testosterone levels (F3 hormones not measured) or normalized testes weight. However, in the F2 generation, we observed a significant effect of treatment on male body weight (H(2) = 9.054, p = 0.011) with A1221-lineage males having greater average body weights compared to DMSO con-Few somatic changes were detected in the measured outcomes for EDC-lineage rats. A trend was observed for an F0 treatment effect (EB slightly larger than DMSO) in normalized adrenal weights of F1 (F(2,33) = 2.997, p = 0.064) and F3 (F(2,25) = 2.856, p = 0.076) males (Table 2). Similarly, no changes were detected in serum testosterone levels (F3 hormones not measured) or normalized testes weight. However, in the F2 generation, we observed a significant effect of treatment on male body weight (H(2) = 9.054, p = 0.011) with A1221-lineage males having greater average body weights compared to DMSO con-Few somatic changes were detected in the measured outcomes for EDC-lineage rats. A trend was observed for an F0 treatment effect (EB slightly larger than DMSO) in normalized adrenal weights of F1 (F(2,33) = 2.997, p = 0.064) and F3 (F(2,25) = 2.856, p = 0.076) males (Table 2). Similarly, no changes were detected in serum testosterone levels (F3 hormones not measured) or normalized testes weight. However, in the F2 generation, we observed a significant effect of treatment on male body weight (H(2) = 9.054, p = 0.011) 3.1. Transgenerational Somatic Changes Few somatic changes were detected in the measured outcomes for EDC-lineage rats. A trend was observed for an F0 treatment effect (EB slightly larger than DMSO) in normalized adrenal weights of F1 (F(2,33) = 2.997, p = 0.064) and F3 (F(2,25) = 2.856, p = 0.076) males (Table 2). Similarly, no changes were detected in serum testosterone levels (F3 hor-# Plexiglas bouts - - - Ultrasonic Vocalizations (F3) (combined for the sexes) EB A1221 # Total calls # Non-FM calls Decreased compared to DMSO (or UNT for Partner Preference). Increased compared to In females, we found an effect of F0 treatment on normalized uterine weight in the F2 generation (H(2) = 6.434; *p* = 0.040), in which A1221-lineage females had greater uterine weights compared to EB (Dunn's post hoc, *p* = 0.044; Table 2). No effect was found for female body weight (Figure 2b), normalized adrenal weight, normalized ovarian weight or serum estradiol (F3 hormones not measured). Hormone priming for the sociosexual tests also resulted in an expected increase in normalized ovarian and uterine weights in the F3 generation, regardless of F0 treatment (Table 2).

tests also resulted in an expected increase in normalized ovarian and uterine weights in

weights compared to EB (Dunn's post hoc, p = 0.044; Table 2). No effect was found for female body weight (Figure 2b), normalized adrenal weight, normalized ovarian weight or serum estradiol (F3 hormones not measured). Hormone priming for the sociosexual tests also resulted in an expected increase in normalized ovarian and uterine weights in

tests also resulted in an expected increase in normalized ovarian and uterine weights in

DMSO (or UNT for Partner Preference). FM: Frequency modulated. n.s. and -: No significant ef-

observed a significant effect of treatment on male body weight (H(2) = 9.054, p = 0.011) with A1221-lineage males having greater average body weights compared to DMSO con-

In females, we found an effect of F0 treatment on normalized uterine weight in the F2 generation (H(2) = 6.434; p = 0.040), in which A1221-lineage females had greater uterine weights compared to EB (Dunn's post hoc, p = 0.044; Table 2). No effect was found for

Decreased compared to DMSO (or UNT for Partner Preference). Increased compared to

In females, we found an effect of F0 treatment on normalized uterine weight in the

In females, we found an effect of F0 treatment on normalized uterine weight in the F2 generation (H(2) = 6.434; p = 0.040), in which A1221-lineage females had greater uterine weights compared to EB (Dunn's post hoc, p = 0.044; Table 2). No effect was found for female body weight (Figure 2b), normalized adrenal weight, normalized ovarian weight or serum estradiol (F3 hormones not measured). Hormone priming for the sociosexual tests also resulted in an expected increase in normalized ovarian and uterine weights in

males (Table 2). Similarly, no changes were detected in serum testosterone levels (F3 hormones not measured) or normalized testes weight. However, in the F2 generation, we observed a significant effect of treatment on male body weight (H(2) = 9.054, p = 0.011) with A1221-lineage males having greater average body weights compared to DMSO con-

weights compared to EB (Dunn's post hoc, p = 0.044; Table 2). No effect was found for female body weight (Figure 2b), normalized adrenal weight, normalized ovarian weight or serum estradiol (F3 hormones not measured). Hormone priming for the sociosexual tests also resulted in an expected increase in normalized ovarian and uterine weights in

In females, we found an effect of F0 treatment on normalized uterine weight in the F2 generation (H(2) = 6.434; p = 0.040), in which A1221-lineage females had greater uterine weights compared to EB (Dunn's post hoc, p = 0.044; Table 2). No effect was found for female body weight (Figure 2b), normalized adrenal weight, normalized ovarian weight or serum estradiol (F3 hormones not measured). Hormone priming for the sociosexual tests also resulted in an expected increase in normalized ovarian and uterine weights in

A trend was observed for an F0 treatment effect (EB slightly larger than DMSO) in normalized adrenal weights of F1 (F(2,33) = 2.997, p = 0.064) and F3 (F(2,25) = 2.856, p = 0.076) males (Table 2). Similarly, no changes were detected in serum testosterone levels (F3 hormones not measured) or normalized testes weight. However, in the F2 generation, we observed a significant effect of treatment on male body weight (H(2) = 9.054, p = 0.011) with A1221-lineage males having greater average body weights compared to DMSO con-

or serum estradiol (F3 hormones not measured). Hormone priming for the sociosexual tests also resulted in an expected increase in normalized ovarian and uterine weights in

the F3 generation, regardless of F0 treatment (Table 2).

In females, we found an effect of F0 treatment on normalized uterine weight in the F2 generation (H(2) = 6.434; p = 0.040), in which A1221-lineage females had greater uterine weights compared to EB (Dunn's post hoc, p = 0.044; Table 2). No effect was found for

In females, we found an effect of F0 treatment on normalized uterine weight in the

the F3 generation, regardless of F0 treatment (Table 2).

the F3 generation, regardless of F0 treatment (Table 2).

tests also resulted in an expected increase in normalized ovarian and uterine weights in

F2 generation (H(2) = 6.434; p = 0.040), in which A1221-lineage females had greater uterine weights compared to EB (Dunn's post hoc, p = 0.044; Table 2). No effect was found for female body weight (Figure 2b), normalized adrenal weight, normalized ovarian weight or serum estradiol (F3 hormones not measured). Hormone priming for the sociosexual tests also resulted in an expected increase in normalized ovarian and uterine weights in

weights compared to EB (Dunn's post hoc, p = 0.044; Table 2). No effect was found for female body weight (Figure 2b), normalized adrenal weight, normalized ovarian weight or serum estradiol (F3 hormones not measured). Hormone priming for the sociosexual tests also resulted in an expected increase in normalized ovarian and uterine weights in

with A1221-lineage males having greater average body weights compared to DMSO con-

3.1. Transgenerational Somatic Changes

trols (p = 0.008; Figure 2a).

trols (p = 0.008; Figure 2a).

female body weight (Figure 2b), normalized adrenal weight, normalized ovarian weight or serum estradiol (F3 hormones not measured). Hormone priming for the sociosexual tests also resulted in an expected increase in normalized ovarian and uterine weights in

trols (p = 0.008; Figure 2a).

or serum estradiol (F3 hormones not measured). Hormone priming for the sociosexual tests also resulted in an expected increase in normalized ovarian and uterine weights in

female body weight (Figure 2b), normalized adrenal weight, normalized ovarian weight or serum estradiol (F3 hormones not measured). Hormone priming for the sociosexual tests also resulted in an expected increase in normalized ovarian and uterine weights in

In females, we found an effect of F0 treatment on normalized uterine weight in the F2 generation (H(2) = 6.434; p = 0.040), in which A1221-lineage females had greater uterine weights compared to EB (Dunn's post hoc, p = 0.044; Table 2). No effect was found for female body weight (Figure 2b), normalized adrenal weight, normalized ovarian weight

the F3 generation, regardless of F0 treatment (Table 2).

the F3 generation, regardless of F0 treatment (Table 2).

the F3 generation, regardless of F0 treatment (Table 2).

the F3 generation, regardless of F0 treatment (Table 2).

the F3 generation, regardless of F0 treatment (Table 2).


**Table 2.** Table of somatic data for each generation.

Body weights are shown for the day of euthanasia, with adrenal, ovarian, uterine, and testicular weights measured postmortem, and the ANOVA *p*-value for a main effect of Treatment. Norm: normalized to body weight. SE: standard error of the mean. n.s.: No significant effects. Bold text indicates significantly different (*p* < 0.05) from DMSO, and italicized text indicates a trend (0.05 < *p* < 0.1) from DMSO in post hoc comparisons. \*, A1221 significantly different from EB (*p* = 0.04).

Toxics 2022, 10, x FOR PEER REVIEW 7 of 20

Figure 2. Boxplots of body weight at euthanasia (~P120) for adult (a) males and (b) females. Data were analyzed by one-way ANOVA or Kruskal–Wallis for effect of F0 treatment, followed by Holm– **Figure 2.** Boxplots of body weight at euthanasia (~P120) for adult (**a**) males and (**b**) females. Data were analyzed by one-way ANOVA or Kruskal–Wallis for effect of F0 treatment, followed by Holm–Sidak or Dunn's pairwise comparisons. \*\* *p* < 0.01. In this and other boxplot graphs, the line represents the median, the lower and upper outline of the boxes the 25th and 75th percentile, respectively, and the lines the 95th percentile.

#### Sidak or Dunn's pairwise comparisons. \*\* p < 0.01. In this and other boxplot graphs, the line represents the median, the lower and upper outline of the boxes the 25th and 75th percentile, respectively, *3.2. Reproductive Behavior in the F1 and F2 Generations* 3.2.1. Males

and the lines the 95th percentile. Table 2. Table of somatic data for each generation. F1 MALES DMSO (n = 14) EB (n = 11) A1221 (n = 12) Treatment Mean ±SE Mean ±SE Mean ±SE p-Values Overall, few effects of perinatal EDC treatment were found in male mating behavior (Figure 3). In the F2 generation, a significant effect of treatment was found for mount frequency (F(2,66) =3.374; *p* = 0.035). Holm–Sidak post hoc analysis showed that EB-lineage males had a lower mount frequency compared to DMSO (*p* = 0.035; Figure 3e) suggesting that EB males required fewer mounts to reach ejaculation. However, no changes were observed in intromission or ejaculation behaviors (Figure 3). All treatments groups showed long average ejaculation latencies with high variability within the groups, presumably due to male subjects being sexually naïve at the time of testing.

Body Weight (g) 475.2 (±10.9) 462.9 (±12.0) 461.7 (±11.6) n.s

Norm Testes Weight (mg) 8.9 (±0.24) 9.0 (±0.07) 9.3 (±0.25) n.s. Serum Testosterone (ng/mL) 1.4 (±0.2) 1.2 (±0.1) 1.1 (±0.2) n.s.

F2 MALES DMSO (n = 26) EB (n = 22) A1221 (n = 23)

Norm Adrenal Weight (mg) 0.108 (±1.8 × 10-3) 0.111 (±2.3 × 10-3) 0.109 (±2.7 × 10-3) n.s. Norm Testes Weight (mg) 9.3 (±0.009) 9.2 (±0.16) 9.0 (±0.18) n.s. Serum Testosterone (ng/mL) 1.1 (±0.1) 1.2 (±0.2) 0.8 (±0.1) n.s.

F3 MALES DMSO (n = 10) EB (n = 9) A1221 (n = 9)

Body Weight (g) 472.4 (±15.2) 464.0 (±12.9) 491.3 (±7.0) n.s.

Norm Adrenal Weight (mg) 0.108 (±4.7 × 10-3) 0.114 (±3.5 × 10-3) 0.100 (±2.7 × 10-3) p = 0.076

Body Weight (g) 449.3 (±5.5) 461.5 (±8.2) 482.2 (±11.4) p = 0.011

Mean ±SE Mean ±SE Mean ±SE

Mean ±SE Mean ±SE Mean ±SE

#### 3.2.2. Females

Sexually naïve F1 and F2 females were examined for copulatory, proceptive and receptive behaviors (Figure 4). All females were in behavioral estrus during mating trials and were successfully able to lordose in response to male mounting and intromitting behavior. In F1 females, the latency to display the first lordotic response was affected by treatment (H(33) = 3.83; *p* = 0.032). Post hoc analysis revealed A1221-exposed females had significantly longer latencies compared to DMSO-exposed females (*p* = 0.015) despite mounting attempts by a sexually experienced male (Figure 4c). Overall, females displayed high levels of aversive behavior and few proceptive behaviors, likely due to the non-paced setting of the mating trials. Toxics 2022, 10, x FOR PEER REVIEW 9 of 20

Figure 3. Boxplots of male mating behavior at P60 for F1 and F2 generations. Sexually naïve males were scored for (a) latency to mount, (b) latency to intromit, (c) latency to ejaculate, (d) postejaculatory interval, (e) mount frequency, (f) intromission frequency, (g) intromission ratio (calculated as number of intromissions divided by number of mounts), and (h) copulatory rate (calculated as the number of mounts and intromissions from the start time until ejaculation). Data were analyzed by one-way ANOVA or Kruskal–Wallis for effect of F0 treatment, followed by Holm–Sidak or Dunn's pairwise comparisons. F1: n = 14 DMSO, 11 EB, 11 A1221; F2: n = 26 DMSO, 22 EB, 22 A1221. \* p < **Figure 3.** Boxplots of male mating behavior at P60 for F1 and F2 generations. Sexually naïve males were scored for (**a**) latency to mount, (**b**) latency to intromit, (**c**) latency to ejaculate, (**d**) postejaculatory interval, (**e**) mount frequency, (**f**) intromission frequency, (**g**) intromission ratio (calculated as number of intromissions divided by number of mounts), and (**h**) copulatory rate (calculated as the number of mounts and intromissions from the start time until ejaculation). Data were analyzed by one-way ANOVA or Kruskal–Wallis for effect of F0 treatment, followed by Holm–Sidak or Dunn's pairwise comparisons. F1: *n* = 14 DMSO, 11 EB, 11 A1221; F2: *n* = 26 DMSO, 22 EB, 22 A1221. \* *p* < 0.05.

#### 0.05. *3.3. Sociosexual Behaviors in the F3 Generation*

#### 3.2.2. Females 3.3.1. Ultrasonic Vocalizations (USVs)

setting of the mating trials.

Sexually naïve F1 and F2 females were examined for copulatory, proceptive and receptive behaviors (Figure 4). All females were in behavioral estrus during mating trials and were successfully able to lordose in response to male mounting and intromitting behavior. In F1 females, the latency to display the first lordotic response was affected by treatment (H(33) = 3.83; p = 0.032). Post hoc analysis revealed A1221-exposed females had significantly longer latencies compared to DMSO-exposed females (p = 0.015) despite We examined the number and duration of appetitive 50 kHz USVs in F3 adults within a mating context (Figure 5). Experimental females were ovarian-intact but hormone primed to ensure receptivity during testing. To reduce potential variability caused by mixed maternal vs. paternal lineages, only maternal, maternal F3 females and paternal, paternal F3 males were used (see Figure 1). Two-way ANOVA tests revealed significant sex and treatment effects in USVs within the first 5 min of separation from the stimulus rat. In

mounting attempts by a sexually experienced male (Figure 4c). Overall, females displayed high levels of aversive behavior and few proceptive behaviors, likely due to the non-paced

particular, the total call number was significantly affected by sex (F(1,46) = 34.96; *p* < 0.001), F0 treatment (F(3,46) = 6.80; *p* < 0.001) and their interaction (F(3,46) = 5.72; *p* = 0.002; Figure 5c). Post hoc treatment contrasts revealed a trend for EB-lineage males to call less frequently than DMSO controls (*p* = 0.065). When sexes were combined to further examine the treatment main effect, we found that DMSO-lineage controls emitted more USVs than both EB (*p* = 0.003) and A1221 (*p* = 0.007) groups (Figure 5e). There was also a trend for reduced call number between EB-lineage rats and our in-house bred untreated controls (UNT; *p* = 0.063). Toxics 2022, 10, x FOR PEER REVIEW 10 of 20

Figure 4. Boxplots of female mating behavior at P60 for F1 and F2 generations. Sexually naïve females were scored for (a) lordosis quotient, (b) lordosis intensity score, (c) latency to first lordosis response, (d) proceptive behavior frequency, (e) proceptive rate, (f) rejection behavior frequency and (g) rejection rate. Data were analyzed by one-way ANOVA or Kruskal–Wallis for effect of F0 treatment, followed by Holm–Sidak or Dunn's pairwise comparisons. F1: n = 14 DMSO, 11 EB, 12 A1221; F2: n = 23 DMSO, 22 EB, 23 A1221. \* p < 0.05. **Figure 4.** Boxplots of female mating behavior at P60 for F1 and F2 generations. Sexually naïve females were scored for (**a**) lordosis quotient, (**b**) lordosis intensity score, (**c**) latency to first lordosis response, (**d**) proceptive behavior frequency, (**e**) proceptive rate, (**f**) rejection behavior frequency and (**g**) rejection rate. Data were analyzed by one-way ANOVA or Kruskal–Wallis for effect of F0 treatment, followed by Holm–Sidak or Dunn's pairwise comparisons. F1: *n* = 14 DMSO, 11 EB, 12 A1221; F2: *n* = 23 DMSO, 22 EB, 23 A1221. \* *p* < 0.05.

3.3. Sociosexual Behaviors in the F3 Generation 3.3.1. Ultrasonic Vocalizations (USVs) We examined the number and duration of appetitive 50 kHz USVs in F3 adults within a mating context (Figure 5). Experimental females were ovarian-intact but hormone primed to ensure receptivity during testing. To reduce potential variability caused by mixed maternal vs. paternal lineages, only maternal, maternal F3 females and paternal, Similar effects were seen when analyzing two subtypes of USVs: frequency-modulated (FM) and non-FM calls (Table 3). Males emitted both types of calls more frequently (non-FM: F(1,42) = 31.87, *p* < 0.001; FM: F(1,41) = 19.89, *p* < 0.001) and called for longer average durations (F(1,44) = 4.932, *p* = 0.032) than females. The number of non-FM calls was also affected by F0 treatment (F(3,42) = 2.79, *p* = 0.024), with EB males having fewer calls of this subtype than DMSO (*p* = 0.031).

paternal F3 males were used (see Figure 1). Two-way ANOVA tests revealed significant sex and treatment effects in USVs within the first 5 min of separation from the stimulus rat. In particular, the total call number was significantly affected by sex (F(1,46) = 34.96; p < 0.001), F0 treatment (F(3,46) = 6.80; p < 0.001) and their interaction (F(3,46) = 5.72; p = 0.002; Figure 5c). Post hoc treatment contrasts revealed a trend for EB-lineage males to call less frequently than DMSO controls (p = 0.065). When sexes were combined to further

trend for reduced call number between EB-lineage rats and our in-house bred untreated

Similar effects were seen when analyzing two subtypes of USVs: frequency-modulated (FM) and non-FM calls (Table 3). Males emitted both types of calls more frequently (non-FM: F(1,42) = 31.87, p < 0.001; FM: F(1,41) = 19.89, p < 0.001) and called for longer

controls (UNT; p = 0.063).

#### 3.3.2. Partner Preference

The partner preference paradigm was used to determine the extent to which F3 EDCor control-lineage rats would be preferred (or avoided) to untreated animals in a mating context. An F3 experimental rat (A1221-, EB-, DMSO-lineage or UNT) was placed opposite a naïve rat purchased from Harlan as the Stimulus animals. A separate set of naïve, Harlan-purchased Chooser rats (of the opposite sex) interacted with the stimulus rats through wire mesh dividers held in place by Plexiglas. Similar to the USV experiments, females were gonadally intact and hormone primed to be receptive. A blind experimenter scored the Chooser rats' behaviors including grooming, rearing, facial investigation of the stimulus rats through the wire mesh and physical contact with the adjacent Plexiglas. As most behaviors occurred in proximity to the stimulus rats, we focused our analysis to the region adjacent to the wire mesh divider (called the wire zone; Figure 6b). The full set of parameters scored within AnyMaze are listed in Supplemental Table S1.

**Table 3.** Ultrasonic vocalization parameters for F3 males and females.


Two-way ANOVA *p*-values for a main effect of Treatment and Sex are provided for the sexes combined (shown next to the male data, but applicable to both sexes). Bold text indicates significantly different at *p* < 0.05 from DMSO in post hoc pairwise comparisons within each sex, and italicized text indicates a trend (0.05 < *p* < 0.1). n.s.: No significant effects, FM: frequency modulated.

In trials where naïve female Harlan Choosers were exposed to F3 experimental males, linear mixed modeling (LMM) analysis showed that the females' time spent rearing (*p* = 0.044) and time spent contacting the Plexiglas (*p* = 0.020) were significantly affected by F0 treatment (Figure 6c). Post hoc analysis revealed that Chooser females preferred the EB-lineage males more often than they preferred UNT controls (time rearing, trend *p* = 0.082; time Plexiglas, *p* = 0.030).

When naïve male Harlan Choosers were tested, their total time active (*p* = 0.017), time spent (*p* = 0.009) and number (*p* = 0.022) of rearing bouts, time spent (*p* = 0.014) and number (*p* = 0.026) of bouts contacting the Plexiglas, and number of facial investigation bouts (trend, *p* = 0.075) were affected by F0 treatment (Figure 6d). Post hoc analysis demonstrated that naïve Chooser males avoided EDC-lineage females more frequently than UNT controls (time active UNT vs. A1221 (trend, *p* = 0.063) and UNT vs. EB (*p* = 0.049); rearing number UNT vs. EB (trend, *p* = 0.060); rearing time UNT vs. A1221 (*p* = 0.032) and UNT vs. EB (*p* = 0.025); Plexiglas number UNT vs. EB (*p* = 0.045); Plexiglas time UNT vs. A1221 (trend, *p* = 0.056) and UNT vs. EB ((*p* = 0.032); Figure 6d and Supplemental Table S1).

average durations (F(1,44) = 4.932, p = 0.032) than females. The number of non-FM calls

Toxics 2022, 10, x FOR PEER REVIEW 11 of 20

calls of this subtype than DMSO (p = 0.031).

DMSO, 6 EB, 8 A1221. + p < 0.07; \*\* p < 0.01.

FEMALES UNT (n = 5) DMSO (n = 9) EB (n = 6) A1221 (n = 8)

Figure 5. Ultrasonic vocalizations (USVs) emitted by F3−generation males and females (P60–120) in response to an opposite−sex rat. An untreated control group (UNT) was raised across generations in−house alongside the F3 litters. (a) Timeline of USV experiment; (b) Diagram of experiment on day 3; (c) example spectrogram of recorded USVs; (d) Boxplots of the total number of USV calls (frequency modulated [FM] and non−FM) during the first 5 min of recording by sex; (e) Boxplots of the total number of USV calls with sex combined. Data were analyzed by two−way ANOVA or Aligned Rank Transformation (ART) for effect of F0 treatment and sex, followed by Holm−Sidak or ART−C pairwise comparisons. Males: n = 6 UNT, 10 DMSO, 6 EB, 6 A1221; females: n = 5 UNT, 9 **Figure 5.** Ultrasonic vocalizations (USVs) emitted by F3−generation males and females (P60–120) in response to an opposite−sex rat. An untreated control group (UNT) was raised across generations in−house alongside the F3 litters. (**a**) Timeline of USV experiment; (**b**) Diagram of experiment on day 3; (**c**) example spectrogram of recorded USVs; (**d**) Boxplots of the total number of USV calls (frequency modulated [FM] and non−FM) during the first 5 min of recording by sex; (**e**) Boxplots of the total number of USV calls with sex combined. Data were analyzed by two−way ANOVA or Aligned Rank Transformation (ART) for effect of F0 treatment and sex, followed by Holm−Sidak or ART−C pairwise comparisons. Males: *n* = 6 UNT, 10 DMSO, 6 EB, 6 A1221; females: *n* = 5 UNT, 9 DMSO, 6 EB, 8 A1221. + *p* < 0.07; \*\* *p* < 0.01.

bined)

Mean ±SE Mean ±SE Mean ±SE Mean ±SE Treatment Sex

Table 3. Ultrasonic vocalization parameters for F3 males and females.

Number of total calls 101.2 (±23.6) 124.5 (±28.3) 33.5 (±12.3) 68.8 (±7.8) p < 0.001 p < 0.001 Number of non-FM calls 48.2 (±10.5) 62.8 (±13.9) 22.8 (±7.5) 36.5 (±10.1) p = 0.024 p < 0.001 Number of FM calls 53.0 (±14.5) 47.2 (±11.5) 17.4 (±6.2) 49.5 (±9.4) n.s. p < 0.001

Percentage of FM calls 50.3 (±4.8) 47.3 (±3.8) 46.2 (±7.1) 59.7 (±4.8) n.s. n.s. Average call duration (msec) 1.44 (±0.22) 1.41 (±0.13) 1.13 (±0.13) 1.61 (±0.19) n.s. p = 0.032

#### *Toxics* **2022**, *10*, 47 Toxics 2022, 10, x FOR PEER REVIEW 13 of 20

Figure 6. Partner preference (PP) in a mating context by F3-generation males and females (P60–120). The Chooser was a naïve rat purchased from Harlan, given a choice between two opposite-sex rats: an in-lab-generated F3 rat (UNT, DMSO, EB, A1221) and a purchased rat. A preference score was calculated by subtracting time spent with the F3 rat minus time spent with the Harlan rat, in which positive numbers indicate more time spent near the F3-lineage rat and a negative score indicating time towards the Harlan rat. (a) Timeline of PP experiment; (b) diagram of experiment on day 3, with the wire zone shaded in gray; preference scores from the wire zone for (c) naïve female Choosers with F3-lineage males and (d) naïve male Choosers with F3-lineage females. Data were analyzed by linear mixed model (LMM) for effect of F0 treatment within each sex followed by Holm–Sidak pairwise comparisons. LMM estimated marginal means and standard errors are graphed. Males: n = 5 UNT, 9 DMSO, 7 EB, 8 A1221; females: n = 5 UNT, 10 DMSO, 7 EB, 7 A1221. \* = p < 0.05 vs. UNT; + = p ≤ 0.083 vs. UNT. **Figure 6.** Partner preference (PP) in a mating context by F3-generation males and females (P60–120). The Chooser was a naïve rat purchased from Harlan, given a choice between two opposite-sex rats: an in-lab-generated F3 rat (UNT, DMSO, EB, A1221) and a purchased rat. A preference score was calculated by subtracting time spent with the F3 rat minus time spent with the Harlan rat, in which positive numbers indicate more time spent near the F3-lineage rat and a negative score indicating time towards the Harlan rat. (**a**) Timeline of PP experiment; (**b**) diagram of experiment on day 3, with the wire zone shaded in gray; preference scores from the wire zone for (**c**) naïve female Choosers with F3-lineage males and (**d**) naïve male Choosers with F3-lineage females. Data were analyzed by linear mixed model (LMM) for effect of F0 treatment within each sex followed by Holm–Sidak pairwise comparisons. LMM estimated marginal means and standard errors are graphed. Males: *n* = 5 UNT, 9 DMSO, 7 EB, 8 A1221; females: *n* = 5 UNT, 10 DMSO, 7 EB, 7 A1221. \* = *p* < 0.05 vs. UNT; + = *p* ≤ 0.083 vs. UNT.

#### **4. Discussion**

The current study demonstrates that a transient gestational exposure to estrogenic EDCs can significantly alter behaviorally relevant endpoints for at least three generations. Interestingly, this occurred in a sex- and generation-specific manner. We found modest but significant effects on copulatory behavior in F1 females and F2 males. In the F3 generation, EDC treatment decreased appetitive 50 kHz ultrasonic vocalizations in response to a rat of the opposite sex and affected the preference of EDC-lineage males and females for a naïve conspecific in a mating context. Finally, we found few somatic changes in adulthood. It is interesting that multigenerational effects of EDCs are preferentially manifested, at least in this paradigm, in neurobehavioral rather than somatic outcomes, a result that may relate to the exquisite sensitivity of the brain to developmental hormones and its potential for epigenetic programming [51]. However, it is possible that other somatic or biochemical outcomes not examined here are altered by prenatal EDC exposure, as we only measured a few specific endpoints and only at one timepoint in adulthood. For example, previous studies on prenatal PCB exposure found delays in the timing of puberty in males [52] and transgenerational effects on anogenital index and female sex steroid hormone levels at P60 [36].

A1221 has weakly estrogenic activity but also has other mechanistic actions including through thyroid and aromatase-mediated pathways [53,54]. Thus, while A1221 can produce similar effects to EB, it is not a pure estrogen and will often deviate from the EB group due to its non-estrogenic-mediated actions as shown in this and previous studies [6]. Here, EDC exposure was given on days E16 and E18 of gestation during the period of germline epigenetic changes and the beginning of brain sexual differentiation in the rat. Both processes are vulnerable to environmental perturbations and A1221 exposure at this time can cause epimutations that become embedded in the germline, leading to changes in somatic gene expression in later generations [24] and lifelong alterations in sex-typical reproductive physiology and behavior. Studies have found differences in maternal versus paternal lineage transmission of disease phenotypes as there are many sex differences in germline de- and re-methylation dynamics [55]. In this study, both maternal and paternal lineages were investigated in the F2 generation; however, we did not find any significant lineage effects on our endpoints. Finally, due to experimental constraints we were unable to perform experiments on every lineage combination in the F3 generation and instead selected F3 females of maternal, maternal lineage and F3 males of paternal, paternal lineage. This is an important area of future study.

#### *4.1. Transgenerational Somatic Endpoints*

Of the somatic changes monitored in the current study, only a few changes were observed in adulthood of EDC-lineage rats, mirroring previous results using this treatment model [36]. Here, we reproduced an increase in body weight at euthanasia in F2-A1221 males. A1221 males had a modest ~7% increase in body weight at euthanasia compared to controls, all given the same ad libitum diet of low-phytoestrogen rat chow. Whether this weight increase is due to increased consumption or a difference in metabolism and energy expenditure between groups should be addressed in future studies. For instance, additional markers of altered metabolism, such as serum insulin or adipokines, could be examined. This finding suggests that PCBs may act according to the "obesogen hypothesis," in which EDC activity can predispose organisms to obesity and metabolic dysfunction. Future research should investigate the extent to which the transgenerational effects of A1221 can synergistically increase weight gain with a high fat diet in adulthood. In the Mennigen et al. (2018) study [36], both F2 and F3 males with A1221 lineage had increased body weight; however, this was primarily driven through the maternal line and our study used only paternal F3 males. Therefore, this discrepancy is likely due to mechanisms of maternal vs. paternal inheritance.

Increased body weight was also previously found in the female F2- and F3-A1221 littermates [36]; however, there are major differences between these subjects and those

in the current study. Here, the F2 females carried a litter to term and were euthanized after weaning, and F3 females were euthanized after the completion of all sociosexual experiments at P120. This resulted in females whose age and postpartum status were vastly different from those in the previous study. Similarly, our finding of an increase in the normalized uterine weight of the F2-A1221 females compared to EB, which was not seen previously, could be due to an interaction between EDC-lineage and postpartum status or might be due to differences in cycle status between the groups. Unfortunately, we did not track the cycle status of the females in the present study, as we presumed that females would be roughly distributed throughout the estrous cycle, and this precludes our ability to rule out cycle effects. Finally, our findings agree with previous work showing that treatment of dams with EDCs on gestational days 16 and 18 do not significantly influence serum testosterone or estradiol concentrations in the F1 and F2 generations [36].

#### *4.2. F1 and F2 Generation Adult Mating Behavior*

Perturbation of the reproductive axis by estrogenic compounds may affect the expression of sexual behavior in adulthood [56]; thus, we studied the copulatory behavior of the F1 and F2 generations as they were mated to propagate litters for the transgenerational experiment. The timing and setup of the mating trials were designed to replicate the conditions from previous experiments on perinatal EDC treatment in our lab. Therefore, sexually inexperienced EDC-lineage rats were mated to untreated, Harlan-raised rats under non-paced mating conditions.

In the F1 generation, prenatal EDC treatment did not alter male copulatory behavior during their first exposure to sexual experience. While a study using a PCB mixture (PCB 126, 138, 153 and 180) found that prenatal exposure delayed latencies in first and subsequent testing of F1 males [57] our model used a differing PCB mixture that may have differing mechanisms of action. On the other hand, our F1-A1221 females significantly delayed their first lordotic event in response to mounting attempts by a sexually experienced male compared to DMSO. A delay in receptive behavior may indicate a deficiency in copulatory motivation. Similarly, using the same A1221 dose, F1 females in a paced mating paradigm also delayed the pacing of mating encounters and event-return latencies [37] although, in both cases, female lordosis remained intact. Some studies have found EDC effects on lordosis and proceptivity using prenatal endocrine active UV filters [58] or exogenous estradiol [59]. However, other specific PCB mixtures had no influence on female lordosis [57,60] as we found here.

The F2 generation showed a different pattern of results. F2 females had no effects of EDCs in their mating behavior; however, F2 males of EB lineage had a decrease in the mount frequency compared to DMSO. The decrease in the number of mounts did not affect the average intromission ratio, or copulatory efficiency, in which a higher percentage of intromissions to mounts may indicate greater ease to achieve an erection [61]. The decrease is also unlikely to reflect a decreased motivation for sexual activity because the latencies to mount and intromit, better indices of motivation, were not affected. In any case, F2 male mating behavior was not severely impacted by either EDC treatments, at least when comparing the initial sexual event. Future studies should address whether reproductive behavior after repeated sexual experience trials reveals other significant effects.

#### *4.3. F3 Generation Adult Sociosexual Behaviors*

Ultrasonic vocalizations are emitted by rodents throughout development and are thought to represent affective states and possibly facilitate communication. In adulthood, rat USVs can be characterized by two main types: 22 kHz calls, associated with aversive stimuli, and those in the 50 kHz or above range, associated with arousal states and positive affect [62]. Rats produce a high rate of 50 kHz calls during positive social interactions such as reproductive behavior, juvenile play and tickling by an experimenter. While the 22 kHz calls are emitted by males after ejaculation, the 50 kHz calls are associated with solicitation and copulatory acts [63]. In this study, we used a well-documented paradigm for

inducing 50 kHz calls through a brief exposure to a hormonally receptive rat of the opposite sex [40]. Upon removal of the stimulus animal, rats will reliably produce 50 kHz calls. We also added an additional negative control group of untreated rats (UNT) bred in-house alongside our F3 generation. Both negative control groups, UNT and F3-DMSO, behaved similarly. We found a decrease in 50 kHz USV production with EDC lineage, particularly in males. Unfortunately, due to a low *n* per group, our study was underpowered. However, when sexes were combined, we were able to see statistically significant decreases in both EB and A1221 groups compared to DMSO control. As 50 kHz calls appear to facilitate mating interactions by signaling a readiness to mate and orienting the activity of the estrous female [63,64], a decrease in USV calls may indicate a deficit in reproductive fitness.

The 50 kHz calls often display variation in subtype and can be roughly separated into frequency-modulated (FM) or non-FM calls. Although the functional implications are not fully understood for these subtypes, FM calls may signal a dopamine-dependent reward state and are preferentially increased in anticipation of cocaine and amphetamine [41,65]. In contrast, flat calls appear to help coordinate social behavior as they are evoked after separation from cage-mates or potential mates and can induce approach behavior in both mating and non-mating environments [66]. Our findings show a decrease in non-FM calls, which would include the flat subtype, with EB lineage. This may suggest a deficit in the coordination of reproductive behavior instead of a decreased motivation to mate. Interestingly, when F3-A1221 pups were separated from their mother, the rate of neonatal USVs were also decreased in paternal-lineage pups [29], so this effect appears to be consistent throughout development.

In this study, we observed notable sex differences in USV calls, with males calling more frequently and for longer call duration than females. While males are known to emit more 50 kHz calls during rough-and-tumble play behavior than females [67], the two sexes generally produce similar call rates during mating encounters [29,68]. Acquisition of sexual experience and hormonal status of both the experimental and stimulus rats can affect the number of vocalizations [68–70]. In naturally cycling females, calls are maximized during proestrus compared to the other cycle states as well as after hormone administration in ovariectomized females [68,71]. The sexually inexperienced females of this study remained ovarian-intact but were supplemented with both estradiol and progesterone to induce the appropriate physiological state. Further, receptivity was confirmed (a lordosis response to an experienced male's mount) prior to the experiment. Unfortunately, this setup failed to induce vocalizations in the females, while males produced calls at a similar rate to that seen in sexually naïve males in the same paradigm [40]. Future studies should assess female USV production during the appropriate stage of their estrous cycle to determine if calls are increased during their natural behavioral estrus. Thus, while the EDC effects appear to be driven solely by paternal-lineage males, our interpretation of F3 female behavior must take into account that this floor effect may mask further decreases in females USV production.

Finally, we investigated whether F3 rats inherited indicators of reproductive deficits from their EDC ancestry. To test this hypothesis, we allowed naïve Chooser rats to select from an F3 experimental rat or a naïve rat (raised at Harlan), using a partner preference paradigm that previously showed a female preference for F3-vehicle males over F3 vinclozolin males (in that study, males showed no preference for either type of female; [32]). Here, we made the surprising observation of a distinction between our in-house negative controls (UNT and F3-DMSO) and the naïve Harlan stimulus rats, especially when males were choosing between females. While males on average tended to prefer UNT and DMSO females compared to naïve Harlan females, they tended to avoid F3-EB and F3-A1221 females. These results emphasize the importance of negative controls, as environmental factors such as rearing environment (in-house vs. Harlan) can affect behavior. Conversely, when naïve females were Choosers, they showed higher preference scores for F3-EB males than for the UNT controls. This study did not attempt to determine the basis for the differences in choice, although this result is particularly interesting in the context of decreased USV production seen in F3-EB males. Other physical stimuli, such as pheromones, and

behavioral cues, also play a role in mate choice, and may outweigh any deficits in social USV calls.

#### **5. Conclusions**

These results show that prenatal EDC treatment has distinct effects within each generation, in a sexually dimorphic manner, showing the complexity of studying inheritance of EDC exposure. These results extend and complement other data showing transgenerational studies on EDCs as well as other environmental stressors [72] that influence health and disease.

**Supplementary Materials:** The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/toxics10020047/s1, Table S1: Table of estimated marginal means and standard errors for F3 partner preference scores.

**Author Contributions:** Conceptualization, B.A.K. and A.C.G.; investigation, B.A.K., L.M.T., J.R.J. and M.H.B.W.; methodology, B.A.K., A.C.G. and L.M.T.; data curation, B.A.K. and L.M.T.; formal analysis, B.A.K.; writing—original draft preparation, B.A.K. and A.C.G.; writing—review and editing, A.C.G.; visualization, B.A.K. and A.C.G.; project administration, L.M.T. and B.A.K.; funding acquisition, A.C.G. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by NIH (NIEHS) R01 ES023254 and R01 ES029464 grants.

**Institutional Review Board Statement:** This study was conducted according to the guidelines of the NIH and approved by the IACUC at the University of Texas at Austin (AUP-2016-00029, approved 3 July 2016).

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Data will be made available upon request.

**Acknowledgments:** We thank Fay Guarraci and Juan Dominguez for their guidance in scoring female and male reproductive behavior, respectively. We also thank Spurthi Tarugu for scoring the USV data and Mandee Bell for help with animal husbandry. Vector icons were obtained through Vecteezy.com (syringe) and Shutterstock (rats).

**Conflicts of Interest:** The authors declare no conflict of interest. The funder had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

#### **References**

