HDAC3 Inhibitor RGFP966 Modulates Neuronal Memory for Vocal Communication Signals in a Songbird Model

Abstract

Epigenetic mechanisms that modify chromatin conformation have recently been under investigation for their contributions to learning and the formation of memory. For example, the role of enzymes involved in histone acetylation are studied in the formation of long-lasting memories because memory consolidation requires gene expression events that are facilitated by an open state of chromatin. We recently proposed that epigenetic events may control the entry of specific sensory features into long-term memory by enabling transcription-mediated neuronal plasticity in sensory brain areas. Histone deacetylases, like HDAC3, may thereby regulate the specific sensory information that is captured for entry into long-term memory stores (Phan and Bieszczad, 2016). To test this hypothesis, we used an HDAC3-selective inhibitor (RGFP966) to determine whether its application after an experience with a sound stimulus with unique acoustic features could contribute to the formation of a memory that would assist in mediating its later recognition. We gave adult male zebra finches limited exposure to unique conspecific songs (20 repetitions each, well below the normal threshold to form long-term memory), followed by treatment with RGFP966 or vehicle. In different groups, we either made multi-electrode recordings in the higher auditory area NCM (caudal medial nidopallidum), or determined expression of an immediate early gene, zenk (also identified as zif268, egr-1, ngfi-a and krox24), known to participate in neuronal memory in this system. We found that birds treated with RGFP966 showed neuronal memory after only limited exposure, while birds treated with vehicle did not. Strikingly, evidence of neuronal memory in NCM induced by HDAC3-inhibition was lateralized to the left-hemisphere, consistent with our finding that RGFP966-treatment also elevated zenk expression only in the left hemisphere. The present findings show feasibility for epigenetic mechanisms to control neural plasticity underlying the formation of specific memories for conspecific communication sounds. This is the first evidence in zebra finches that epigenetic mechanisms may contribute to gene expression events for memory of acoustically-rich sensory cues.

Keywords: NCM; electrophysiology; epigenetics; gene expression; lateralization; memory; stimulus specific adaptation (SSA).

Figure 1

Stimulus Specific Adaptation (SSA) to song stimuli is evident in caudal medial nidopallidum (NCM) electrophysiological responses. (A) Multi-unit recordings from two unique sites in NCM show evoked responses to a zebra finch song. The sound spectrogram is representative of the stimuli presented during to evoke electrophysiological responses. Note the complex features in the sound stimulus, such as the differences in fundamental frequencies and harmonic acoustic structures that vary among the individual syllables. These features distinguish each song motif, resulting in unique identifiers for an individual bird vocalization. (B) NCM neuronal responses change with repeated exposures to the same song stimulus. The response magnitude of multiunit activity recorded at one site in NCM are plotted over time during the presentation of songs that were either previously heard during the limited song exposure (20X; F1, F2) or that were novel (N1, N2). Both stimulus classes were presented 20 h after the limited song exposure. Each unique song was played 25 times in shuffled order, but responses have been reordered in the plot for clarity. SSA rates were always calculated for trials 6–25 as indicated by the box outline in N2 (see “Materials and Methods” Section “Electrophysiological Analysis for Detecting Neuronal Memory”) and are shown at the top of each panel.

Figure 2

Histone deacetylase 3 (HDAC3) is evident in nuclei of NCM cells. (A–D) Representative sagittal section of NCM taken 0.75 mm from the midline. Top row: 40× magnification, scale bar = 2500 μm; Middle row: 40X magnification, scale bar = 50 μm; Bottom row: 100X magnification, scale bar = 25 μm. Columns show immunolabeling with antibodies against DAPI, NeuN, or HDAC3 as indicated. Last column shows a merged image. Images are from the target area of NCM at higher resolution (E–H) and magnification (I–L). NCM, caudal medial nidopallidum.

Figure 3

HDAC3-i enabled long-term neural memory in NCM, measured after only limited song exposure (20X), is lateralized. The Familiarity Index (FI) is a measurement of specific song memory derived from the neuronal responses in the NCM. An FI ratio at 1 (gray line) indicates that the previously exposed songs are not distinguished from novel, never before heard songs; an FI ratio > 1 indicates that the previously exposed song is more familiar (remembered) than the novel songs. FI measurements are significantly lateralized with a bias to higher FI’s in the left hemisphere in HDAC3-i treated birds (Kolmogorov-Smirnov Test, d_stat = 0.223; p = 0.03; median_left = 1.28; IQR_left = 0.69; median_right = 1.15; IQR_right = 0.61). In contrast, vehicle treated birds did not show a significant lateralization (Kolmogorov-Smirnov Test, d_stat = 0.162, p = 0.341). In the box and whisker plot, the box extends from the 25th to 75th percentiles, the line in the middle indicates median, and upper whiskers report maximum data values. *Indicates significant difference between groups, alpha = 0.05.

Figure 4

Effects of HDAC3-i on transcription demonstrated gene and region specificity. HDAC3-i induced significant increases in zenk (aka egr-1) expression selectively in left NCM with respect to vehicle controls. In contrast, this pattern of gene expression was not evident in c-fos expression (left panel; zenk: left NCM, t₍₈₎ = 2.46, p = 0.039; right NCM, t₍₈₎ = 0.232, p = 0.822; c-fos: left NCM, t₍₈₎ = 0.083, p = 0.936; right NCM, t₍₈₎ = 0.440, p = 0.672; one-sample t-test, Bonferroni corrected for the two hemisphere comparisons). This lateralized increased zenk expression was not observed in the Anterior Pole (AP, right panel; left AP, t₍₄₎ = 0.479, p = 0.657; right AP, t₍₄₎ = 0.919, p = 0.410; Bonferroni corrected for the two hemisphere comparisons). Schematized sagittal section of the zebra finch brain, sampled at 0.75 mm from the midline. The dashed curve line indicates the area dissected from the AP used in the gene expression study. NCM, caudal medial nidopallidum. Scale bar = 2500 μm. *Indicates significant difference between groups, alpha = 0.05.

Figure 5

HDAC3-i coupled with the limited exposure paradigm (20X) reveals hemispheric differences in FI which is absent in the “typical” extended exposure condition (200X). The extended exposure paradigm (200X, no HDAC3-i condition) resulted in assessment of auditory memories that were more specific for previously exposed stimuli (Familiar) than for never before heard stimuli (Novel). However, in contrast, to the HDAC3-i treated birds, 200X birds do not show significant lateralized memory responses in FI (K-S d_stat = 0.159; p = 0.346). Figure conventions as in Figure 3.