The Xenopus amygdala mediates socially appropriate vocal communication signals

Abstract

Social interaction requires that relevant sensory information is collected, classified, and distributed to the motor areas that initiate an appropriate behavioral response. Vocal exchanges, in particular, depend on linking auditory processing to an appropriate motor expression. Because of its role in integrating sensory information for the purpose of action selection, the amygdala has been implicated in social behavior in many mammalian species. Here, we show that two nuclei of the extended amygdala play essential roles in vocal communication in the African clawed frog, Xenopus laevis. Transport of fluorescent dextran amines identifies the X. laevis central amygdala (CeA) as a target for ascending auditory information from the central thalamic nucleus and as a major afferent to the vocal pattern generator of the hindbrain. In the isolated (ex vivo) brain, electrical stimulation of the CeA, or the neighboring bed nucleus of the stria terminalis (BNST), initiates bouts of fictive calling. In vivo, lesioning the CeA of males disrupts the production of appropriate vocal responses to females and to broadcasts of female calls. Lesioning the BNST in males produces an overall decrease in calling behavior. Together, these results suggest that the anuran CeA evaluates the valence of acoustic cues and initiates socially appropriate vocal responses to communication signals, whereas the BNST plays a role in the initiation of vocalizations.

Figure 1.

Example oscillograms of the male and female X. laevis calls described in this study. For male advertisement (A) and answer calls (B), the fast and slow trill phase of a single call are underlined. Female rapping (C) and ticking (D) communicate sexual receptivity or nonreceptivity, respectively. Each call segment is 2 s in duration.

Figure 2.

CeA and BNST in X. laevis. A, Schematic view of the caudal portion of the ventral telencephalon in X. laevis, viewed from below (rostral up). The boundaries of BNST and CeA are indicated on each hemisphere, with their respective associated fiber pathways overlaid on the left hemisphere: green represents LFB; orange, ST [stria terminalis]. Diagram single neurons in each nucleus show the typical orientation of cell bodies and dendrites in the nucleus with respect to the forebrain ventricle (not pictured) and the fiber tract. In the right hemisphere, the location of the adjacent nuclei of the basal ganglia is shown in white: Str., Striatum; EA, entopeduncular area, a pallidal homolog. A₁, Schematic view of the entire brain of X. laevis in the sagittal orientation, showing the location of sections (lines) and whole-mount views (gray boxes) from subsequent parts of the figure. B, LFB and axon terminal field within the ventral subpallium labeled by CT injection in a whole-mount brain. z-stack image (500 μm thick) through caudal ventral telencephalon (horizontal plane, rostral up). The CeA is outlined by a white overlay. Arrowhead indicates back-labeled EA cells. Scale bar, 250 μm. C, CT cells back-labeled by CeA injection in a whole-mount brain. z-stack image (500 μm thick) through dorsal thalamus (horizontal plane, rostral up). Scale bar, 250 μm. D, Caudal telencephalon in a whole-mount brain after double-label with CT (green) and DTAM (magenta) injections. z-stack image (375 μm thick) in the horizontal plane (rostral up). Arrowheads indicate back-labeled DTAM-projecting cells (magenta). Although the green CT axons extend into the more rostrally located striatum, the DTAM terminal field and DTAM-projecting cells are restricted to the caudal portion of the subpallium. No back-labeled cells are seen in EA after DTAM injection. Axons connecting CeA and DTAM are also contained within the LFB, but in a more medial location than CT axons (green). Back-labeled cells (both magenta and green) can also be seen in APOA (asterisk). Scale bar, 250 μm. E, High-power (40×) image of anterogradely labeled CT axons (green) within the LFB interacting with locally labeled CeA dendrites (magenta). Horizontal section, rostral up. Scale bar, 50 μm. Dashed boxes indicate the areas shown at expanded magnification in the inset. Arrowheads indicate three example sites of putative synaptic contacts between boutons and spines. F, Transverse section through caudal telencephalon just rostral to APOA after double-label with CT (green) and DTAM (magenta) injections. The CT axon terminal field is clearly restricted to the outer half of the CeA dendritic fields and neatly overlaps the more proximally located terminal field (asterisk) and cell bodies back-labeled by DTAM injection. The combination of DTAM and CT labeling delineates the border of CeA with the BNST medially and the medial amygdala dorsally (see white overlay on contralateral side). Scale bar, 500 μm. G, Synaptophysin (SYP) puncta (magenta) located on CT axons (green). High-power (63×) z-stack image (4.5 μm thick) in the horizontal plane (rostral up). Arrowheads indicate two example puncta located on axons labeled by tracer injection into CT. Scale bar, 10 μm. H, Axon terminal fields within CeA and BNST labeled by contralateral CeA injection. z-stack image (400 μm thick) in the horizontal plane (rostral up). White dashed overlay outlines the BNST area as distinct from CeA. The projection from CeA is bilateral; however, terminal fields within the ipsilateral hemisphere are obscured by the proximity of the dye injection. Back-labeled cells within APOA can be seen at lower right (asterisk). Scale bar, 250 μm.

Figure 3.

CeA connects to auditory and vocal circuits. A, Horizontal section through the dorsal thalamus after CeA (green) and IC (magenta) injections (rostral up). Green back-labeled CeA-projecting cells overlap anterogradely labeled terminal field of IC axons (magenta) within CT (see white overlay on contralateral hemisphere). Magenta at lower left is a fiber tract that originates from the injection site (located just below the bottom of the panel). Scale bar, 500 μm. B, Expanded view of region indicated by the dashed box in A, showing the overlap of back-labeled cells and the IC terminal field in greater detail. Proximal dendrites can be seen extending from back-labeled cell bodies (arrowhead). Blue channel is nuclear label from DAPI staining. Scale bar, 50 μm. C, Expanded view of region indicated by the dashed box in B, showing magenta axon boutons making apparent contact with proximal dendrites of green back-labeled CT cells (see arrowheads). Scale bar, 20 μm. D, Horizontal section through the tegmentum and DTAM after CeA (green) and VMN (magenta) injections (rostral up). VMN injection labels both cells and terminal fields within DTAM (dashed box) and overlaps with the green terminal field of labeled CeA axons. White overlay on the contralateral hemisphere outlines the boundaries of DTAM at this level. Green terminal fields from CeA are also visible in the more rostral pedunculopontine tegmental nucleus (asterisk) and the nucleus of the tractus solitarus (arrowhead) located just caudal to DTAM. Scale bar, 500 μm. E, Expanded view of region indicated by the dashed box in D, showing overlap of back-labeled VMN-projecting cells and the CeA terminal field in DTAM. Instances of putative overlap of CeA axonal boutons (green) and VMN-projecting DTAM cells (magenta) are indicated by arrowheads. Blue channel is nuclear label from DAPI staining. Scale bar, 50 μm. F, Expanded view of region indicated by E, showing the overlap of axon boutons and proximal dendrites in greater detail. Scale bar, 20 μm. G, Schematic views of the entire brain of X. laevis in the sagittal orientation, showing the location of sections (lines) and a summary diagram of the connections described in Figures 1 and 2. Green represents auditory nuclei; red represents vocal nuclei.

Figure 4.

Fictive calling can be elicited by stimulation of CeA and BNST. A, Transverse sections through the X. laevis forebrain. Locations of stimulation sites within the telencephalon are coded by type of stimulation and success in eliciting calling on the left hemisphere, whereas colored regions on the right hemisphere indicate the extent of the different telencephalic nuclei (or groups of nuclei, in the case of the septum) that were stimulated. Nu. Acc., Nucleus accumbens; MP, medial pallium; Sep., septum; Stri., striatum. The approximate locations of each section (i–v) are indicated by the vertical lines on the frog brain schematic at the bottom left (sagittal view, rostral to the left). B, Fictive calling in response to 5-HT. The trace shows a 20 s excerpt of laryngeal nerve activity taken from a longer calling bout (∼2 m) during a 3 min bath application of 5-HT (indicated by the blue box above the trace). In this case, 5-HT was applied ∼25 s before the beginning of this trace. Inset to the right, Dashed area with an enlarged time-scale, demonstrating the alternating biphasic pattern of the fictive calling episode. Nerve activity consists of highly synchronized CAPs, organized into biphasic alternating trills. Calibration: B, 2 s; inset, 200 ms. C, Biphasic fictive calling elicited by short pulse stimulation of CeA. Trace represents laryngeal nerve activity during and after the stimulus train. Red box represents the timing and duration of the stimulus train (0.2 ms pulse duration, 30 Hz). A short monophasic trill occurs during the second half of the stimulus train (dashed overline, CAPs obscured by stimulus artifacts), followed ∼4 s later by a repeating biphasic trill. Inset, Dashed area with an enlarged time-scale, demonstrating the alternating biphasic pattern of the fictive calling episode. Calibration: C, 2 s; inset, 200 ms. D, Monophasic calling episode after stimulation in CeA. Nerve activity during and after a long pulse stimulus train (indicated by the red box, long pulse: 2 ms duration, 60 Hz) shows two monophasic trills after stimulation: one immediately after the cessation of the stimulus train (expanded time-scale of the dashed region shown in inset) and the second ∼5 s later. Calibration: D, 2 s; inset, 200 ms. E, Individual CAPs elicited by 5-HT and electrical stimulation (stim) from the same brain. Fast trill CAPs have similar shapes and heights regardless of mode of stimulation, whereas slow trill CAPs are less stereotyped both within and between modes of stimulation. Individual CAPs were extracted from 1- to 2-s-long stretches of fictive calling activity for periods of both 5-HT and electrically stimulated fictive calling in a single brain (stimulation CAPs in this example come from the trace shown in B). Plots represent 5-ms-long stretches of N.IX-X nerve activity centered on the peak of the CAP with individual CAPs overlaid. All CAPs originated from periods of alternating biphasic calling, but only those that could be reliably assigned to fast or slow trills were used for this plot. Upper row, Raw CAP traces separated into fast and slow trill groups and color-coded by the mode of stimulation that elicited them. Fast and slow CAPs are both plotted with the same vertical scale (vertical scale bar represents 1000 mV). Thicker lines represent the mean across all individual CAPs of that type represented on this plot. Lower row, Same CAPs normalized by peak height for both fast and slow trills, again color-coded by the mode of stimulation, with bold lines representing the mean of all CAPs in each group. Horizontal scale bar for each plot represents 1 ms. F, CAP width, time to peak, and CAP half-width displayed as box-and-whisker plots for fast trill and slow trill CAPs elicited by 5-HT and electrical stimulation. Measurements from 25 instances of each type of CAP (fast and slow for 5-HT and stim-elicited episodes) were averaged to produce a single value for each type of CAP for each brain that produced fictive calling episodes to both modes of stimulation. Upper plots, Values for fast trill CAPs; lower plots, those for slow; both plots, color coded for mode of stimulation. Thick line indicates the median across all brains; box represents the upper and lower quartiles, with tails showing the upper and lower extremes. Black dots indicate outliers (>3 SD above median). G, ICI histograms showing the bimodal distribution of in vivo and fictive calling. All measured clicks (in vivo) and CAPs (fictive) were pooled across brains and calling episodes, and in aggregate show bimodal distributions with a “fast” peak centered at ∼16 ms and a broader “slow” peak whose mean ICI is more variable between conditions. The n value indicates the number of frogs or brains that data were pooled from for each condition.

Figure 5.

The response to playbacks was altered by CeA lesion. A, The time to suppression of calling (left y-axis) and the time to the resumption of calling (right y-axis) during playback are plotted for advertisement (Ad), rapping, and ticking. Sham-lesioned males suppressed earlier and resumed calling later during advertisement playbacks than rap playbacks (stars; Friedman, Dunn's multiple comparison, Q = 6.5, p = 0.03 for suppression; Q = 9.2, p = 0.006 for resumption). Advertisement playbacks were 15 min long; rap and tick playbacks were 5 min long, so that the x-axis is normalized for the duration of each playback. B, Using the same conventions as in A, CeA-lesioned animals show prolonged vocal suppression during advertisement, rap, and ticking playbacks. After IBO or electrolytic lesion, there was no significant difference in the time to suppression or resumption of calling during any playback (suppression: Friedman, IBO, Q = 6.0, p = 0.057; electrolytic, Q = 3.211, p = 0.1873; resumption: Friedman, IBO, Q = 0.9630, p = 0.5690; electrolytic, Q = 3.0, p = 0.2223). C, D, The prelesion to postlesion change in time to suppression (C) and time to resumption (D) was correlated with the number of cells remaining in the posterior CeA after lesion. The x-axes are reversed, with higher cell counts on the left and lower on the right. The y-axis was calculated as ([post lesion behavior] − [prelesion behavior]), so positive numbers indicate greater postlesion values. The fewer cells remaining in the CeA, the more rapid the time to calling suppression (C, Pearson's r = 0.54, p = 0.020) and the longer the time until call resumption (D, Pearson's r = −0.62, p = 0.01). E, F, The prelesion to postlesion change in the duration of calling during rapping (E) and ticking (F) playbacks was correlated with the number of cells remaining in the posterior CeA after lesion. The fewer cells remaining in the CeA, the less animals called during rapping (E, Pearson's r = 0.69, p = 0.001) or ticking (F, Pearson's r = 0.68, p = 0.001).

Figure 6.

Transverse sections through the X. laevis forebrain at the level of the CeA and BNST. Electrolytic lesion sites for (A) the CeA (n = 12) and (B) the BNST (n = 9) were traced, colored gray, and overlain. Inset, Schematic of a sagittal view of the frog brain (rostral left) with vertical lines indicating the approximate level of the sections.

Figure 7.

Answer call production decreases after CeA lesion. A, The proportion of answer calls to all calls produced by a lesioned frog when paired with another male frog for 10 min was significantly reduced after electrolytic lesion of the CeA. Star and brackets indicate significantly different populations (Kruskal–Wallis K = 14.8, p = 0.001). B, The prelesion to postlesion change in the number of answer calls produced during an interaction with a male frog was correlated with the number of cells remaining in the CeA (male: Pearson's r = 0.39, p = 0.046). C, The proportion of answer calls directed at a female frog during a 10 min interaction was significantly reduced after electrolytic lesion of the CeA (Kruskal–Wallis K = 13.05, p = 0.002). D, The prelesion to postlesion change in the number of answer calls produced during an interaction with a female frog was correlated with the number of cells remaining in the CeA (Pearson's r = 0.44, p = 0.026).

Figure 8.

Vocalization rates decrease after BNST lesion. A, The prelesion to postlesion change in the time of spontaneous vocalization differed for sham- and BNST-lesioned males. Star and brackets indicate significantly different populations (Kruskal–Wallis K = 5.756, p = 0.053). B, The prelesion to postlesion change in the time of spontaneous vocalization was correlated with the number of cells remaining in the BNST (Pearson's r = 0.78, p = 0.001). C, The prelesion to postlesion change in the time spent vocalizing when paired with a male differed for sham- and BNST-lesioned males. Star and brackets indicate significantly different populations (Kruskal–Wallis K = 7.683, p = 0.0215). D, There was no prelesion to postlesion difference in the time spent vocalizing with a female for sham-, CeA-, or BNST-lesioned males.