Brain Activation Patterns and Dopaminergic Neuron Activity in Response to Conspecific Advertisement Calls in Reproductive versus Non-Reproductive Male Plainfin Midshipman Fish (Porichthys notatus)

Acoustic communication is a key aspect of social behavior across vertebrate groups. Vocalizations are utilized in a variety of social behaviors, including aggressive interactions and reproduction, and an individual’s ability to detect and discriminate between these socially relevant auditory signals above the background noise can be essential for survival and reproduction. The neural pathways responsible for detecting and responding to acoustic cues are posited to be conserved across vertebrate taxa [1‒4] and vocal-acoustic communication is believed to have evolved first in bony fishes (Osteichthyes) [5, 6]. Understanding the neural circuitry involved in detecting socially relevant acoustic cues in bony fishes could provide valuable insights into the conserved and convergent auditory and vocal-acoustic communication networks across vertebrates.

The plainfin midshipman fish (Porichthys notatus) offers an ideal study system for investigating the neural substrates essential for attention and processing of socially relevant acoustic stimuli as their reproductive success depends on vocal-acoustic communication [5, 6]. This species exhibits three adult sexual phenotypes (females and type I and II males) with well-characterized reproductive strategies [7‒10]. Type I or nest guarding males produce long duration vocalizations (“hums”) to attract gravid (reproductive) females to their nest. For a successful reproductive cycle, females localize the type I male’s advertisement call, enter the nest, and deposit eggs, which are promptly fertilized by the male. Type II males “steal” fertilizations from the nest guarding male by entering the nest concurrently with a female and depositing sperm [9]. After fertilization, type I males guard the nest and care for larvae [8]. Nesting type I males will continue to produce advertisement calls to facilitate multiple spawning events throughout the reproductive season [7, 9, 10]. Since multiple type I males establish nesting sites in close proximity, calling males may detect and respond to nearby conspecific competitors [5], especially since spectral aspects of the advertisement call (e.g., amplitude and fundamental frequency) correlate with information about male body size and condition [11].

Male and female midshipman fish undergo seasonal changes in hearing sensitivity related to their reproductive state. It is well documented that during the reproductive season (spring/summer), male (type I and II) and female midshipman fish exhibit decreased hearing thresholds (increased sensitivity) compared to the non-reproductive (NR) season (winter) [12‒18]. However, much of this research has focused on the plasticity of the midshipman peripheral auditory system, with limited studies examining reproductive-state-dependent changes in the central auditory system in plainfin midshipman [14, 19].

The auditory pathways in plainfin midshipman have been well characterized using tract tracing techniques and single-unit recordings. In this species, the saccule is one of the primary peripheral sensory structures for hearing [12, 20‒22]. Saccular afferent neurons project to nuclei in the auditory hindbrain, which in turn project to the nucleus centralis of the torus semicircularis (TS) (TSnc) in the midbrain [23‒25]. The TSnc then projects to the compact division of the central posterior nucleus of the thalamus (CPc), which has reciprocal connections with the hypothalamic anterior tuberal nucleus (AT) [24, 25].

The periventricular posterior tuberculum (TPp) in ray-finned fish is a nucleus in the diencephalon populated with dopaminergic neurons [26 and see 27‒29 for reviews]. Tyrosine hydroxylase (TH) is the rate-limiting enzyme for catecholamine synthesis [30] and thus is a marker for both dopaminergic and noradrenergic neurons in the brain. Consistent with studies in other teleost species, TH immunoreactivity (-ir) but absence of dopamine (DA) beta hydroxylase-ir, an enzyme that is required to catalyze DA to norepinephrine, supports the identity of TH-ir TPp neurons as dopaminergic in midshipman fish [31, 32]. In plainfin midshipman, tract tracing studies have revealed that these dopaminergic neurons in the TPp send direct efferent input to the saccule and the cholinergic octavolateralis efferent nucleus [32, 33]. Additionally, the TH-ir neurons of the TPp send ascending projections to the CPc and areas of the basal forebrain that are thought to be homologous to centers of reward and decision-making across vertebrates [34, 35].

It is well documented that auditory nuclei in the midshipman’s midbrain, forebrain, and the dopaminergic TPp respond preferentially to the advertisement hum compared to control or ambient sounds in all three sexual phenotypes (females and type I and II males) [30‒33]. The TPp is particularly associated with increased activation in reproductive males (type I and II) and females in response to the male advertisement call [36‒39]. Additionally, in female midshipman, there are differences in TPp neuron area based on reproductive state, with reproductive females exhibiting increased TPp neuron area compared with NR females [14].

We hypothesized that the neurons in three higher level auditory nuclei (TSnc, CPc, AT) and the dopaminergic neurons (DAergic) of the TPp in type I males would be differentially activated when exposed to conspecific advertisement calls based on their reproductive state (reproductive vs. NR) and acoustic stimuli (male advertisement call vs. ambient sound control). To assess neural activation in well-defined auditory nuclei in midshipman [24], we used immunohistochemistry (IHC) methods to label for phospho-S6 ribosomal protein (pS6), a marker for phosphorylated ribosomal proteins and an indicator of increased protein translation in response to a stimulus or behavioral state in combination with an antibody against the Hu protein, which specifically labels mature neuronal somata and nicely delineates the cytoarchitecture of the brain [40‒42]. Phospho-S6 ribosomal protein is becoming a common marker of neural activation in vertebrates, including fish, due to its robust signal and high specificity [43‒50 and see 51‒53 for reviews]. In order to quantify activation of DAergic neurons in the TPp, we double labeled for TH-ir and pS6-ir (see above). We predicted that type I males exposed to a conspecific male advertisement call would show greater pS6 activation in the auditory (TSnc, CPc, AT) and DAergic (TPp) neurons compared to the control but only in the reproductive state.

NR type I male midshipman fish were collected by otter trawl in January (winter) in Monterey Bay near Moss Landing, CA (R/V John H. Martin, Moss Landing Marine Laboratories), at depths from 85 to 100 m. Reproductive type I male midshipman fish were hand-collected during morning low tides in May (spring) from the same nesting sites in Tomales Bay near Marshall, CA, as in previous studies [e.g., 14, 37, 38]. After collection, fish were transported in coolers with aerated sea water to the University of California Bodega Marine Laboratory (BML) in Bodega Bay, CA, where they were housed in flow-through sea water aquaria at natural ambient temperatures (13–14.5°C) until playback trials, which were conducted within 16–48 h after collection. This holding time is consistent with similar studies on type I [38] and type II [37] male midshipman, as well as female midshipman [36]. Furthermore, Sisneros and Bass [12] reported no decrease in auditory sensitivity to frequencies contained within the male mate call up to 25 days post-collection.

Playback trials were performed at BML in a large cylindrical outdoor concrete tank (4 m diameter, 0.5 m water depth) during nighttime sessions. A speaker (Aquasonic AQ339, Clark Synthesis, Littleton, CO, USA) was suspended at the center of the tank, positioned 5 cm above the tank floor, and powered by an audio amplifier (BG-1120, TOA Electronics, Secaucus, NJ, USA). A mesh playback cage (240 cm diameter) was placed around the speaker. At the start of each trial, fish were placed in a plastic mesh cylindrical “release cage” (30 cm diameter) situated in a fixed location along the wall of the playback cage (Fig. 1). The sound pressure level (SPL) at the release cage was 130 dB re: 1 μPa, which is consistent with SPLs recorded from the nest entrance in calling type I males [54] and used in previous studies [36‒39]. As fish were allowed to swim freely, the mesh playback cage ensured that the fish were exposed to at least 130 dB re: 1 μPa SPL during playback of the male advertisement call. SPLs were calibrated nightly using a mini-hydrophone (model 8103, Bruel & Kjaer, Naerum, Denmark), a charge amplifier (Bruel & Kjaer model 2692), and an oscilloscope (TBS 1064, Textronix, Beaverton, OR, USA).

Male midshipman fish were exposed to either 15 min of continuous male advertisement calls (“hum”; NR: N = 5; reproductive: N = 6; Fig. 2) or a no-sound ambient control (105–120 dB re: 1 μPa; NR: N = 6; reproductive: N = 7). The hums were recorded from reproductive type I midshipman collected from nests in Brinnon, WA, during low tides in June/July and housed at the Friday Harbor Marine Laboratory (FHL) in outdoor tanks outfitted with artificial nests. Brantley and Bass [9] showed that the fundamental frequency of male advertisement calls increases with temperature, and female preference reflects this correlation [20]. Therefore, the water temperature at BML was taken nightly (January: 13°C; May 14–14.5°C), and we used FHL recordings taken at the corresponding temperatures to create 15-min audio files from a single calling male with fundamental frequencies of 85 Hz (January trials) and 87–90 Hz (May trials) (Fig. 2).

After the 15-min playback trial, fish were placed in a 5-gallon bucket for 15 min and then sacrificed by deep anesthetization in 0.025% benzocaine in seawater. Immediately after sacrifice, the standard length (SL) and weight were recorded for each animal, followed by transcardial perfusion to preserve the brain tissue. Fish were first perfused with ice cold teleost ringers followed by 4% paraformaldehyde in 0.1 m phosphate buffer (PB; pH 7.2). The brains were then harvested, post-fixed for 1 h, rinsed in 0.1 m PB (3X), and stored at 4°C in 0.1 m PB for <1 week. Next, brains were cryoprotected in 30% sucrose-PB solution, embedded in cryogenic gel, and stored in a −80°C freezer until sectioning. Brains were sectioned on a cryostat in the transverse plane at 25 μm in two series, collected onto subbed slides, and stored at −80°C until labeling. For this experiment, one of two series was used for IHC. Finally, the testes of each fish were dissected and weighed and the gonadosomatic index (GSI) calculated ([gonad mass/body mass – gonad mass] × 100).

Slides were first brought to room temperature, then perimeter of each slide was traced with a hydrophobic pen and they were soaked 3X for 10 min in phosphate-buffered saline (PBS; pH 7.2), followed by a 1-h soak in a blocking solution made of PBS with 0.3% Triton X-100 and 5% normal donkey serum (PBS-DS; donkey serum from Jackson Immunolab, West Grove, PA, USA). Following the blocking procedure, slides were incubated for 16–17 h at room temperature in PBS-DS containing rabbit anti-pS6 (1:500; lot No. 16; cat No. 4858; Cell Signaling Technology, Danvers, MA, USA), sheep anti-TH (rate-limiting enzyme for DA synthesis;1:500; lot No. 3261295; cat No. AB1542; Millipore Sigma, Burlington, MA, USA), and mouse anti-Hu (1:1,000, lot No. 2105721; cat No. A-21271; Thermo Fisher Scientific, Waltham, MA, USA), which labels neurons [40‒42]. After incubation, slides were briefly dipped and then rinsed in PBS +0.5% normal donkey serum 5X for 10 min. The slides were then incubated for 2 h at room temperature with PBS-DS containing donkey anti-rabbit conjugated to Alexa Fluor 568 (1:200; lot No. 2044343; cat No. A10042; Thermo Fisher Scientific) for pS6, anti-sheep Alexa Fluor 488 (1:200; lot No. 2079356; cat No. A-11015; Thermo Fisher Scientific) for TH, and anti-mouse Alexa Fluor 350 (1:200; lot No. 2045306; cat No. A10035; Thermo Fisher Scientific) for Hu. Finally, slides were dipped and rinsed 4X for 10 min in PBS before being coverslipped using ProLong Gold Antifade (without DAPI; cat No. P36930; Thermo Fisher Scientific) and allowed to cure for 48–72 h in a dark room. Once dried, the slides were sealed with nail polish and stored at 4°C. The IHC methods for TH and Hu were based on previous studies with midshipman [37, 38]. Phospho-S6 ribosomal protein (pS6) is a well-established marker used in studies across multiple teleost species [43‒47, 49], including plainfin midshipman [50]. The specificity of the anti-pS6 primary antibody has been demonstrated in a prior study with Astatotilapia burtoni, a teleost fish [43], and confirmed for midshipman using Western blot analysis [50].

Images were obtained using a Nikon Eclipse 80i epifluorescence compound microscope (Nikon Instruments Inc., Melville, NY, USA) with a Prime 95B sCMOS camera (Teledyne Photometrics, Tuscon, AZ, USA) and NIS-Elements software (Nikon Instruments Inc.). For the three “auditory nuclei” (TSnc, CPc, AT), the regions were first identified using the DAPI filter to visualize neurons (Hu immunoreactive or Hu-ir cells) before being imaged using a 10X objective. Each photomicrograph was taken consecutively using DAPI and Texas Red filter sets, respectively. For the dopaminergic nucleus (TPp), the region was first identified using the GFP filter to visualize the dopaminergic TH immunoreactive (TH-ir) neurons before being imaged using a 10X objective. TH is the rate-limiting enzyme for DA synthesis and is a proxy for DA neuron activation [30]. Each photomicrograph was taken consecutively using GFP, Texas Red, and DAPI filter sets, respectively. For all regions examined, the exposure times and light levels were held constant for each channel across all conditions. Photomicrographs were then merged (TPp: “Batch RGB Merge” macro; auditory nuclei: “Batch RB Merge” modified macro) and analyzed using ImageJ (National Institutes of Health, Bethesda, MD, USA) [55]. For all auditory nuclei, images were first analyzed in the blue (Hu-ir) channel and the region of interest was outlined and isolated. The region of interest was then processed using a custom ImageJ macro “Object Count with Find Maxima” in the red channel to identify pS6 immunoreactive (pS6-ir) neurons [56]. The pS6-ir identified cells were then manually verified, ensuring only colocalized Hu-ir and pS6-ir cells were counted. The average number of pS6-ir cells per section in each auditory nucleus was recorded in each animal. For the TPp, the total number of TH-ir neurons were first counted using the custom “Manual Count” ImageJ macro in the green channel. pS6-ir neurons were then identified using the “Object Count with Find Maxima” macro and manually verified to confirm that only pS6-ir and TH-ir neurons were counted. For each animal, the percentage of total TH-ir neurons that also contained pS6 (hereafter referred to as colocalization) was then determined for each section.

The auditory pathway in the midshipman midbrain and forebrain has been well described [23, 24] and similar to previous work [38, 39], we imaged the following regions: the nucleus centralis of the TS (TSnc) of the midbrain and two forebrain regions: the compact division of central posterior nucleus of the thalamus (CPc) and the anterior tuberal nucleus of the ventral hypothalamus (AT) (Fig. 3). For all auditory regions, sections were sampled unilaterally on the right side throughout their entire extent. No left/right differences in hemispheric brain activation were predicted because fish were allowed to move freely above the underwater playback speaker as in Mohr et al. [39]. The regions were identified using landmarks from previously published neuroanatomical and physiological studies [e.g., 23, 24] and the sampling techniques were replicated from Petersen et al. [38] and Mohr et al. [39]. For the midbrain TSnc, similar to Peterson et al. [38], we sampled every fourth section (moving caudally to rostral), and we used five sections throughout the TSnc for analysis in all animals (except one individual in the reproductive control group for which four sections were analyzed). In the forebrain, both the CPc (identified as a wing-shaped band of cells lateral to the midline) and the AT receive projections from the TSnc and for each nucleus, we sampled three sections [38].

Analysis of pS6-ir colocalization with the TH-ir neurons of the TPp began caudally with appearance of large pear-shaped dopaminergic (TH-ir) cells clustered medially to the medial forebrain bundle and extending ventrolaterally along the lateral border of the paraventricular organ (Fig. 3) [32, 36‒38, 57]. TPp was analyzed serially for an average of 6.48 ± 0.75 (standard deviation [SD]) sections in the caudal to rostral direction until the disappearance of the large, pear-shaped TH-ir cells adjacent to the midline. There was no difference in the number of sections analyzed between groups (ANOVA F(3, 20) = 1.47; p = 0.26).

For each auditory nucleus (TSnc, CPc, AT), the number of pS6-ir cells per section was averaged for each animal and compared between the four treatment groups (nonreproductive control, NR hum, reproductive control, reproductive hum) using a two-way ANOVA (α set at 0.05) and Tukey HSD post hoc tests. For the dopaminergic TPp, average percent of TH-ir cells colocalized with pS6-ir per section was determined for each animal, and the percent data were arcsine transformed. A two-way ANOVA (α set at 0.05) and Tukey HSD post hoc tests were performed on the residuals. Shapiro-Wilk tests confirmed normal distribution of the data (p > 0.05). The Benjamini-Hochberg procedure was used to correct for multiple comparisons, with a false discovery rate of 0.25 [14, 36‒39, 57, 58]. Statistical analyses were performed in RStudio (version 2023.06.0+421, PBC, Boston, MA, USA) [59]. All averages are reported as mean ± SD. In cases where the power is <0.8, we still report the significance; however, these results are associated with a higher likelihood of a type II error (Table 1).

Of the 24 type I male midshipman used in the study, 11 were collected during the winter in January (N = 6 control; N = 5 hum) and 13 were collected during the spring in May (N = 7 control; N = 6 hum). In the NR (winter) condition, fish in the control group (ambient noise) had an SL of 24.38 ± 3.42 cm (mean ± SD), body mass of 203.63 ± 83.57 g, and a GSI of 1.76 ± 0.48. For the NR fish exposed to the male hum, the SL was 23.38 ± 5.77 cm, body mass was 181.67 ± 110.30 g, and GSI was 1.77 ± 0.87. In the reproductive (spring) condition, fish in the control group had an SL of 19.59 ± 4.73 cm, body mass of 110.26 ± 75.49 g, and GSI of 1.60 ± 0.65. The reproductive males exposed to the male hum had a SL of 18.12 ± 4.1 cm, body mass of 92.56 ± 61.0 g, and GSI of 2.09 ± 0.21. There were no differences between groups for any of the morphometric data analyzed (SL: ANOVA F(3, 20) = 2.37; p = 0.10; weight: ANOVA F(3, 20) = 2.35; p = 0.10; GSI: ANOVA F(3, 20) = 0.72; p = 0.55). Although GSI is a measure of reproductive capacity, Sisneros et al. [60] reported that while mean GSI in type I males did not significantly differ between the reproductive and NR seasons, histological analysis revealed that testes from NR fish (collected via trawl in the winter) did not contain sperm, while the testes from reproductive fish (collected from nests) contained mature sperm. Furthermore, reproductive state can also be verified by the presence of robust sonic muscles, which develop in reproductive-state type I males [61, 62]. We are confident that the fish collected for this study in May (spring) represent the reproductive state as these fish were collected from nesting sites at low tide and had well-developed sonic muscles around the swim bladder [61, 62], while we observed diminished sonic muscles in NR males collected in January (winter).

In the midbrain, TSnc neurons were differentially activated based on both reproductive state and playback stimulus (ambient control vs. hum). A two-way ANOVA revealed a significant effect of both reproductive state (F(1, 20) = 6.61, p = 0.028) and playback stimulus (F(1, 20) = 6.24, p = 0.021), with a significant interaction between reproductive state and stimulus (F(1, 20) = 12.41, p = 0.0021) (Table 1). Reproductive (spring) males exposed to the advertisement call (hum) (mean ± SD pS6-ir cells/section; 34.52 ± 10.17) had significantly more pS6-ir TSnc neurons compared with reproductive males in the ambient control group (19.06 ± 8.88, p = 0.0075; Fig. 4) and the NR (winter) male groups (control: 18.57 ± 3.56, p = 0.027; hum: 14.12 ± 11.4, p = 0.0075; Fig. 4) (Table 1).

Forebrain nuclei were also differentially activated based on both reproductive state and playback stimulus, particularly in the CPc. For the CPc, a two-way ANOVA showed that there was a significant effect of both reproductive state (F(1, 20) = 63.9, p < 0.0001) and playback stimulus (F(1, 20) = 19.12, p = 0.00029), with a significant interaction between reproductive state and playback stimulus (F(1, 20) = 11.94, p = 0.0025) (Table 1). Reproductive males exposed to the hum had significantly more pS6-ir CPc neurons (67.03 ± 12.11, Fig. 5a) compared to reproductive males in the ambient control group (38.71 ± 10.03, p = 0.00075, Fig. 5a) and the NR male groups (ambient control: 22.55 ± 9.77, p < 0.0001; hum: 22.60 ± 7.36, p < 0.0001, Fig. 5a) (Table 1). Additionally, reproductive males in the ambient control group had more pS6-ir CPc neurons compared to both the NR male groups; however, after the Benjamini-Hochberg correction this difference was only significant when compared to the NR ambient control (p = 0.026; reproductive ambient control vs. NR hum: p = 0.068 [uncorrected: p = 0.034]) (Table 1).

In the AT, a two-way ANOVA showed a significant effect of reproductive state (F(1, 20) = 4.79, p = 0.041) and playback stimulus (F(1, 20) = 19.50, p = 0.00027), with an interaction effect between reproductive state and playback stimulus (F(1, 20) = 5.73, p = 0.027) (Table 1). Reproductive males exposed to the hum (108.78 ± 15.81, Fig. 5b) had significantly more pS6-ir AT neurons compared with reproductive males in the ambient control group (75.99 ± 11.23, p = 0.0027, Fig. 5b) and NR male groups (ambient control: 76.05 ± 11.35, p = 0.0033; hum: 85.00 ± 8.51, p = 0.042, Fig. 5b) (Table 1).

The dopaminergic neurons of the TPp were differentially activated based on playback condition. Although the TPp data are reported as percent colocalization of pS6-ir and TH-ir neurons, the percentage data were arcsine transformed for statistical tests. A two-way ANOVA on the residuals showed a significant effect of playback stimulus (F(1, 20) = 4.95, p = 0.038) and an interaction effect between reproductive state and playback stimulus (F(1, 20) = 8.78, p = 0.0077) (Table 1). Reproductive males exposed to the hum had a significantly higher percent colocalization (38.52% ± 22.02% [mean ± SD]) of pS6-ir and TH-ir neurons in the TPp compared to reproductive males in the ambient control group (7.98% ± 7.54%, p = 0.021, Fig. 6) (Table 1). Although the reproductive males exposed to the hum had higher activation in dopaminergic neurons compared to the NR groups (ambient control: 21.91% ± 11.67%, hum: 15.70% ± 16.61%), these differences were not significant (p = 0.44 and p = 0.17, respectively) (Table 1).

The goal of this study was to compare neural activation in auditory midbrain, forebrain, and dopaminergic (DAergic) regions in type I male midshipman across two reproductive states (reproductive vs. NR) and under two auditory stimulus conditions (conspecific male advertisement call, i.e., the “hum” vs. no playback ambient control). Using phospho-S6 ribosomal protein (pS6) as a proxy for neural activity, we observed increased activation in the midbrain and forebrain auditory nuclei in reproductive male midshipman exposed to the conspecific male hum. Additionally, reproductive male midshipman exposed to the advertisement call showed increased pS6-ir colocalization in the TH-ir cells of the TPp, as a proxy for the activation of the DAergic neurons. Our results are consistent with previous studies in male type I [38] and female midshipman [39], which showed increased activation, as reported by a cFos antibody, in midbrain (TSnc), forebrain (AT, CPc), and DAergic (TPp) nuclei in response to male advertisement calls within the reproductive season. We also report differential response of the DAergic TPp in reproductive male midshipman based on playback condition, which is consistent with previous work in midshipman [36‒38].

The TS in anamniotes is thought to be homologous to the amniote inferior colliculus [63‒65], a region associated with detection of social acoustic/communication signals [66]. In fishes and amphibians, the TS also plays a key role in processing auditory signals [38, 39, 67‒70]. In anurans, several TS subdivisions are associated with auditory processing [71‒74]. For instance, in a playback experiment using male túngara frogs (Engystomops pustulosus), Hoke et al. [72] found a relationship between egr-1 (an immediate early gene) activation in the TS and the behavioral salience of mating calls. In midshipman and other bony fishes, the nucleus centralis of the TS (TSnc) is the primary auditory center in the midbrain, receiving direct input from auditory regions of the hindbrain [3, 24, 75]. In the current study, we report increased activation of pS6 in TSnc neurons only in reproductive male midshipman exposed to the male hum. Our findings are consistent with the increased cFos (an immediate early gene) induction reported in the TSnc in reproductive male [38] and female [39] midshipman in response to the male hum but expand those initial findings to indicate this differential response to the advertisement call only occurs, at least in males, while in a reproductive state.

The compact division of the central posterior nucleus in the auditory thalamus (CPc) receives projections from the TSnc and is theorized to be involved in higher-order processing of auditory information in midshipman and other bony fishes [3, 24, 25]. We observed the highest pS6 activation in the CPc in reproductive male plainfin midshipman exposed to the male hum. Interestingly, the CPc was the only region where reproductive male midshipman in the ambient sound (reproductive control) group showed greater pS6 activation compared to NR control (i.e., winter state) individuals. Previous studies on male [38] and female [39] midshipman reported the highest cFos activation in groups exposed to the male hum, but all these fish were in reproductive condition. These results suggest that basal level of activity in the auditory thalamus is higher in the reproductive versus NR state. Based on single-unit recordings in the goldfish (Carrasus auratus), Lu and Fay [76] concluded that the CPc, in concert with the TSnc, is involved in discriminating acoustic signals. Northcutt [77] further theorized that the goldfish thalamic subdivisions likely integrate both auditory and visual sensory inputs to modulate sensorimotor response behavior. Together with these previous studies, our findings suggest that the increased activation in CPc and TSnc in reproductive male midshipman could reflect a reproductive-state-dependent enhancement in attention to/processing of conspecific signals.

Both male and female midshipman fish exhibit seasonal plasticity in the auditory periphery, resulting in decreased hearing thresholds (higher sensitivity) during the reproductive season [12‒18]. Forlano et al. [14] reported that reproductive female midshipman had increased density of TH immunoreactive (TH-ir) fibers in the CPc compared to NR females, a seasonal difference not observed in the TSnc or AT. These fibers originate, at least in part, from the periventricular posterior tuberculum (TPp) [14, 32], and since they also project to the saccule of the inner ear, they are thought to modulate seasonal changes in hearing sensitivity at the periphery and auditory-driven reproductive behaviors. Mohr et al. [39] reported greater cFos activation in the CPc for female midshipman exposed to the male advertisement call compared with a heterospecific vocalization. Together with Mohr et al. [39], the results of the present study, which show higher pS6 activation in reproductive male midshipman, particularly in the male hum playback group (vs. ambient sound control), provide strong evidence that the CPc plays a crucial role in the processing of social acoustic signals.

The anterior tuberal nucleus (AT) also receives direct input from the TSnc and shares reciprocal connections with the CPc [3, 24, 25, 45]. We report that reproductive male midshipman exposed to the conspecific male advertisement call exhibited the highest pS6 activation. This finding is consistent with previous studies that reported increased cFos activation in the AT of male [38] and female [39] midshipman exposed to the male hum. Similar to TSnc and CPc, this effect was specific to the reproductive season.

Reproductive male midshipman exposed to the conspecific male advertisement call showed an increased percentage of dopaminergic (DAergic) neurons (expressing TH-ir) colocalized with pS6-ir. These results are consistent with the increased cFos-ir/TH-ir colocalized cells reported by Peterson et al. [38] in reproductive male midshipman exposed to the conspecific hum compared to controls. Similarly, Forlano et al. [36] showed that activation of TPp TH-ir neurons strongly correlated with the time gravid female midshipman attended to a speaker playing a synthetic hum. Interestingly, our results indicate that differential responsiveness of TPp TH-ir neurons to the male advertisement call is only found in the reproductive state. These findings provide supporting evidence that DAergic neurons in the TPp (1) aid in coordinating motivated acoustic-driven reproductive behaviors [36] and (2) may play a role in modulating the seasonal plasticity of the central auditory system, especially in CPc [14].

Catecholamines have been associated with a variety of conspecific interactions, across vertebrate taxa including reproductive behaviors, male-male competition, and territory defense. For instance, DAergic neurons in the posterior tuberculum are associated with positive phonotaxic response behavior in female gray tree frogs (Hyla versicolor) responding to the male advertisement call [78]. Numerous studies in mammals and songbirds have implicated DA in male-motivated sexual behavior [79, 80] and vocal behavior in reproductive male zebra finches [81‒83] and European starlings (Sturnus vulgaris) [84]. Furthermore, catecholamines have been linked with antagonistic behaviors in reproductive males. For instance, both DA and noradrenaline were increased in green anoles (Anolis carolinensis) responding to the presentation of actual [85] or simulated [86] male competitors. Sewall et al. [87] found an increase in noradrenaline metabolites in the auditory forebrain in male Lincoln’s sparrows (Melospiza lincolnii) exposed to rival male songs.

Here, we show differential responsiveness of TH-ir neurons in the TPp in reproductive male midshipman based on playback condition. Most vertebrates (likely all Sarcopterygii, which includes tetrapods) possess an ascending DA system that originates in the midbrain; however, this system is not present in teleost fishes (largest group of bony fishes) [26, 28, 88]. In zebrafish (a species of teleost fishes), the large DAergic cells in the TPp project anteriorly to the ventral telencephalon [89, 90] and caudally to the hindbrain and spinal cord [89, 91‒94]. In addition to directly innervating the forebrain CPc, the DAergic cells of the midshipman TPp also send projections to the saccule, one of the primary hearing end organs of the inner ear [32]. Forlano et al. [14] examined seasonal plasticity of these TPp innervations and reported increased TH-ir density in the CPc but decreased TH-ir density in the saccule in reproductive female midshipman. The results of the present study show that the reproductive males exposed to the hum had the highest percent co-localization of pS6-ir and TH-ir cells in the TPp, although these differences were not significant when compared with the NR condition. Differential activation of DAergic neurons in the TPp during the reproductive season may serve to increase auditory attention and coordinate directed behavioral responses to hearing male competitors [38].

In conclusion, research on midshipman suggests that the DAergic TPp (1) is preferentially activated by social acoustic signals [37, 38], (2) is the major contributor of DAergic input to both the central and peripheral auditory system [32, 33], and (3) has seasonally regulated DAergic innervation in females, in part by steroid hormones [14, 95]. The present study further supports a role of catecholamines, including DA, in modulating reproductive behaviors driven by the auditory system.

We would like to thank Kitty Brown and the UC Davis Bodega Marine Laboratory for logistical support. We also thank Jennifer Liang and the University of Minnesota Duluth Biology Department for use of their epifluorescence compound microscope and supplemental imaging was done at the University of St. Thomas thanks to Carolyn Zeiner. We would also like to thank Kelsie Murchy for sharing spectrogram code and Rob Mohr for providing feedback on an early draft of this manuscript.

All experimental procedures were approved by the University of Washington Institutional Animal Care and Use Committee (protocol: 4079-06).

The authors declare no conflict of interests.

This work was supported by the National Institutes of Health (Auditory Neuroscience Training Grant, T32: 4T32DC005361-15 to B.J.V.), the National Science Foundation (IOS-1456700 to J.A.S. and IOS-1456743 to P.M.F.), and Tow Professorship Award to P.M.F.

Conceptualization, methodology, and writing – review and editing: B.J.V, J.P.T., N.R.L, J.A.S., and P.M.F.; formal analysis and funding acquisition: B.J.V, J.A.S., and P.M.F.; resources and supervision: J.A.S. and P.M.F; writing – original draft: B.J.V.

All data generated or analyzed during this study are included in this article. Further inquiries can be directed to the corresponding author.