Permalink
Cannot retrieve contributors at this time
Fetching contributors…
Cannot retrieve contributors at this time
| <?xml version="1.0" encoding="UTF-8"?><!DOCTYPE article PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Archiving and Interchange DTD v1.1d1 20130915//EN" "JATS-archivearticle1.dtd"><article article-type="research-article" dtd-version="1.1d1" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><front><journal-meta><journal-id journal-id-type="nlm-ta">elife</journal-id><journal-id journal-id-type="hwp">eLife</journal-id><journal-id journal-id-type="publisher-id">eLife</journal-id><journal-title-group><journal-title>eLife</journal-title></journal-title-group><issn publication-format="electronic">2050-084X</issn><publisher><publisher-name>eLife Sciences Publications, Ltd</publisher-name></publisher></journal-meta><article-meta><article-id pub-id-type="publisher-id">01833</article-id><article-id pub-id-type="doi">10.7554/eLife.01833</article-id><article-categories><subj-group subj-group-type="display-channel"><subject>Research article</subject></subj-group><subj-group subj-group-type="heading"><subject>Neuroscience</subject></subj-group></article-categories><title-group><article-title>Auditory synapses to song premotor neurons are gated off during vocalization in zebra finches</article-title></title-group><contrib-group><contrib contrib-type="author" equal-contrib="yes" id="author-8673"><name><surname>Hamaguchi</surname><given-names>Kosuke</given-names></name><xref ref-type="aff" rid="aff1"/><xref ref-type="fn" rid="equal-contrib">†</xref><xref ref-type="fn" rid="con1"/><xref ref-type="fn" rid="conf1"/></contrib><contrib contrib-type="author" equal-contrib="yes" id="author-8954"><name><surname>Tschida</surname><given-names>Katherine A</given-names></name><xref ref-type="aff" rid="aff1"/><xref ref-type="fn" rid="equal-contrib">†</xref><xref ref-type="fn" rid="con2"/><xref ref-type="fn" rid="conf1"/></contrib><contrib contrib-type="author" id="author-8955"><name><surname>Yoon</surname><given-names>Inho</given-names></name><xref ref-type="aff" rid="aff2"/><xref ref-type="fn" rid="con3"/><xref ref-type="fn" rid="conf1"/></contrib><contrib contrib-type="author" id="author-8956"><name><surname>Donald</surname><given-names>Bruce R</given-names></name><xref ref-type="aff" rid="aff2"/><xref ref-type="aff" rid="aff3"/><xref ref-type="aff" rid="aff4"/><xref ref-type="other" rid="par-3"/><xref ref-type="fn" rid="con4"/><xref ref-type="fn" rid="conf1"/></contrib><contrib contrib-type="author" corresp="yes" id="author-8890"><name><surname>Mooney</surname><given-names>Richard</given-names></name><xref ref-type="aff" rid="aff1"/><xref ref-type="corresp" rid="cor1">*</xref><xref ref-type="other" rid="par-1"/><xref ref-type="other" rid="par-2"/><xref ref-type="other" rid="par-4"/><xref ref-type="fn" rid="con5"/><xref ref-type="fn" rid="conf1"/></contrib><aff id="aff1"><institution content-type="dept">Department of Neurobiology</institution>, <institution>Duke University Medical Center</institution>, <addr-line><named-content content-type="city">Durham</named-content></addr-line>, <country>United States</country></aff><aff id="aff2"><institution content-type="dept">Department of Electrical and Computer Engineering</institution>, <institution>Duke University</institution>, <addr-line><named-content content-type="city">Durham</named-content></addr-line>, <country>United States</country></aff><aff id="aff3"><institution content-type="dept">Department of Computer Science</institution>, <institution>Duke University</institution>, <addr-line><named-content content-type="city">Durham</named-content></addr-line>, <country>United States</country></aff><aff id="aff4"><institution content-type="dept">Department of Biochemistry</institution>, <institution>Duke University Medical Center</institution>, <addr-line><named-content content-type="city">Durham</named-content></addr-line>, <country>United States</country></aff></contrib-group><contrib-group content-type="section"><contrib contrib-type="editor"><name><surname>Calabrese</surname><given-names>Ronald L</given-names></name><role>Reviewing editor</role><aff><institution>Emory University</institution>, <country>United States</country></aff></contrib></contrib-group><author-notes><corresp id="cor1"><label>*</label>For correspondence: <email>mooney@neuro.duke.edu</email></corresp><fn fn-type="con" id="equal-contrib"><label>†</label><p>These authors contributed equally to this work</p></fn></author-notes><pub-date date-type="pub" publication-format="electronic"><day>18</day><month>02</month><year>2014</year></pub-date><pub-date pub-type="collection"><year>2014</year></pub-date><volume>3</volume><elocation-id>e01833</elocation-id><history><date date-type="received"><day>07</day><month>11</month><year>2013</year></date><date date-type="accepted"><day>10</day><month>01</month><year>2014</year></date></history><permissions><copyright-statement>© 2014, Hamaguchi et al</copyright-statement><copyright-year>2014</copyright-year><copyright-holder>Hamaguchi et al</copyright-holder><license xlink:href="http://creativecommons.org/licenses/by/3.0/"><license-p>This article is distributed under the terms of the <ext-link ext-link-type="uri" xlink:href="http://creativecommons.org/licenses/by/3.0/">Creative Commons Attribution License</ext-link>, which permits unrestricted use and redistribution provided that the original author and source are credited.</license-p></license></permissions><self-uri content-type="pdf" xlink:href="elife01833.pdf"/><related-article ext-link-type="doi" id="ra1" related-article-type="commentary" xlink:href="10.7554/eLife.02289"/><abstract><object-id pub-id-type="doi">10.7554/eLife.01833.001</object-id><p>Songbirds use auditory feedback to learn and maintain their songs, but how feedback interacts with vocal motor circuitry remains unclear. A potential site for this interaction is the song premotor nucleus HVC, which receives auditory input and contains neurons (HVC<sub>X</sub> cells) that innervate an anterior forebrain pathway (AFP) important to feedback-dependent vocal plasticity. Although the singing-related output of HVC<sub>X</sub> cells is unaltered by distorted auditory feedback (DAF), deafening gradually weakens synapses on HVC<sub>X</sub> cells, raising the possibility that they integrate feedback only at subthreshold levels during singing. Using intracellular recordings in singing zebra finches, we found that DAF failed to perturb singing-related synaptic activity of HVC<sub>X</sub> cells, although many of these cells responded to auditory stimuli in non-singing states. Moreover, in vivo multiphoton imaging revealed that deafening-induced changes to HVC<sub>X</sub> synapses require intact AFP output. These findings support a model in which the AFP accesses feedback independent of HVC.</p><p><bold>DOI:</bold><ext-link ext-link-type="doi" xlink:href="10.7554/eLife.01833.001">http://dx.doi.org/10.7554/eLife.01833.001</ext-link></p></abstract><abstract abstract-type="executive-summary"><object-id pub-id-type="doi">10.7554/eLife.01833.002</object-id><title>eLife digest</title><p>Whenever we speak, sing, or play a musical instrument, we use auditory feedback to fine-tune our movements to achieve the sound that we want. This same process is used by songbirds to learn and maintain their songs. As juvenile birds practice singing, they compare their vocalizations with their father’s song, which they will previously have stored in memory, and continually tweak their own song until the two versions match.</p><p>It has been suggested that auditory feedback is integrated with song motor commands—the instructions from the brain to move the muscles required for singing—in a region of the songbird brain called the song premotor nucleus HVC. The structure of certain neurons in this region, known as HVC<sub>X</sub> cells, rapidly changes when a bird is deafened, which suggests that these HVC<sub>X</sub> cells detect auditory feedback.</p><p>Hamaguchi et al. have now tested this idea by using fine electrodes to record the signals in HVC<sub>X</sub> cells in male zebra finches as they sang. The cells changed their activity patterns whenever the birds changed their vocalizations. By contrast, these patterns did not change when the birds heard a distorted version of their own song played back to them as they sang. This suggests that HVC<sub>X</sub> cells are insensitive to auditory feedback, and that they mainly encode song motor commands instead.</p><p>If HVC<sub>X</sub> cells don’t detect feedback, then why does deafening affect them? HVC<sub>X</sub> cells send signals indirectly to a brain region called the LMAN (which is short for the lateral magnocellular nucleus of the anterior nidopallium). Normally, if a bird becomes deaf, the quality of their song begins to deteriorate, but this deterioration can be prevented by destroying the LMAN. Hamaguchi et al. used high resolution imaging to show that destroying the LMAN also prevents deafening from altering the structure of HVC<sub>X</sub> cells. Again, this suggests that auditory feedback is not relayed from the HVC to the LMAN; instead the flow of information is in the opposite direction.</p><p>This surprising finding—namely, that HVC<sub>X</sub> cells do not integrate auditory feedback and song motor commands—raises the question of which brain region is in fact responsible for this process. Further experiments will be required to identify the underlying circuitry in the brains of songbirds.</p><p><bold>DOI:</bold><ext-link ext-link-type="doi" xlink:href="10.7554/eLife.01833.002">http://dx.doi.org/10.7554/eLife.01833.002</ext-link></p></abstract><kwd-group kwd-group-type="author-keywords"><title>Author keywords</title><kwd>zebra finch</kwd><kwd>auditory feedback</kwd><kwd>sensorimotor</kwd><kwd>birdsong</kwd></kwd-group><kwd-group kwd-group-type="research-organism"><title>Research organism</title><kwd>other</kwd></kwd-group><funding-group><award-group id="par-1"><funding-source><institution-wrap><institution>National Institutes of Health</institution></institution-wrap></funding-source><award-id>R01 DC02524</award-id><principal-award-recipient><name><surname>Mooney</surname><given-names>Richard</given-names></name></principal-award-recipient></award-group><award-group id="par-2"><funding-source><institution-wrap><institution>National Institutes of Health</institution></institution-wrap></funding-source><award-id>R21 NS079929</award-id><principal-award-recipient><name><surname>Mooney</surname><given-names>Richard</given-names></name></principal-award-recipient></award-group><award-group id="par-3"><funding-source><institution-wrap><institution>National Institutes of Health</institution></institution-wrap></funding-source><award-id>R01 GM78031</award-id><principal-award-recipient><name><surname>Donald</surname><given-names>Bruce R</given-names></name></principal-award-recipient></award-group><award-group id="par-4"><funding-source><institution-wrap><institution>National Science Foundation Integrative Organismal Systems</institution></institution-wrap></funding-source><award-id>0821914</award-id><principal-award-recipient><name><surname>Mooney</surname><given-names>Richard</given-names></name></principal-award-recipient></award-group><funding-statement>The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.</funding-statement></funding-group><custom-meta-group><custom-meta><meta-name>elife-xml-version</meta-name><meta-value>2</meta-value></custom-meta><custom-meta specific-use="meta-only"><meta-name>Author impact statement</meta-name><meta-value>Intracellular recordings in singing birds and images of synapses in deafened birds provide insights into the neural circuitry that enables songbirds to fine-tune their songs.</meta-value></custom-meta></custom-meta-group></article-meta></front><body><sec id="s1" sec-type="intro"><title>Introduction</title><p>Many elaborate behaviors, ranging from playing a guitar to engaging in a conversation, depend on the brain’s ability to integrate performance-related auditory feedback with the output of motor circuits controlling behavior. The synaptic mechanisms underlying the integration of sensory and motor-related signals remain enigmatic. Songbirds use auditory feedback to learn and maintain their vocalizations (<xref ref-type="bibr" rid="bib22">Konishi, 1965</xref>; <xref ref-type="bibr" rid="bib38">Price, 1979</xref>) and possess well-delineated neural circuits for singing (<xref ref-type="bibr" rid="bib34">Nottebohm et al., 1982</xref>), thus providing an attractive organism in which to identify synaptic mechanisms for auditory–vocal integration. Nevertheless, despite important progress in identifying synaptic mechanisms of auditory-guided song plasticity (<xref ref-type="bibr" rid="bib31">Mooney, 1992</xref>; <xref ref-type="bibr" rid="bib36">Olveczky et al., 2005</xref>; <xref ref-type="bibr" rid="bib2">Andalman and Fee, 2009</xref>; <xref ref-type="bibr" rid="bib40">Roberts et al., 2010</xref>; <xref ref-type="bibr" rid="bib48">Warren et al., 2011</xref>), how auditory feedback is integrated in the brain to affect the neural circuits for singing remains poorly understood.</p><p>One site where feedback could interact with song motor commands is the sensorimotor nucleus HVC, which is necessary for singing (<xref ref-type="bibr" rid="bib35">Nottebohm et al., 1976</xref>), exhibits neural activity precisely time-locked to song (<xref ref-type="bibr" rid="bib15">Hahnloser et al., 2002</xref>; <xref ref-type="bibr" rid="bib23">Kozhevnikov and Fee, 2007</xref>; <xref ref-type="bibr" rid="bib37">Prather et al., 2008</xref>), and receives input from auditory regions that contain auditory feedback-sensitive neurons (<xref ref-type="bibr" rid="bib21">Keller and Hahnloser, 2009</xref>). Within HVC, a distinct type of projection neuron (HVC<sub>X</sub>) provides input to an anterior forebrain pathway (AFP) that is necessary for song learning and that is anatomically similar to cortico-basal ganglia pathways in mammals (<xref ref-type="bibr" rid="bib11">Farries and Perkel, 2002</xref>; <xref ref-type="bibr" rid="bib10">Doupe et al., 2005</xref>). Various studies have shown that the output nucleus of the AFP (the lateral magnocellular nucleus of the anterior nidopallium, or LMAN) is the source of behavioral variations that can adaptively bias song in response to feedback perturbations (<xref ref-type="bibr" rid="bib2">Andalman and Fee, 2009</xref>; <xref ref-type="bibr" rid="bib48">Warren et al., 2011</xref>; <xref ref-type="bibr" rid="bib8">Charlesworth et al., 2012</xref>). Specifically, LMAN lesions prevent deafening-induced song degradation (<xref ref-type="bibr" rid="bib4">Brainard and Doupe, 2000</xref>; <xref ref-type="bibr" rid="bib2">Andalman and Fee, 2009</xref>; <xref ref-type="bibr" rid="bib48">Warren et al., 2011</xref>), and inactivating LMAN can reverse shifts in syllable pitch induced by distorted auditory feedback (DAF) (<xref ref-type="bibr" rid="bib4">Brainard and Doupe, 2000</xref>; <xref ref-type="bibr" rid="bib2">Andalman and Fee, 2009</xref>; <xref ref-type="bibr" rid="bib48">Warren et al., 2011</xref>). These findings suggest that the AFP has access to feedback-related information about song performance. One potential source of this information is the HVC<sub>X</sub> cell population, because auditory responses that can be detected in the AFP of anesthetized birds disappear when HVC is pharmacologically inactivated (<xref ref-type="bibr" rid="bib41">Roy and Mooney, 2009</xref>). Moreover, individual HVC<sub>X</sub> cells can fire in remarkably similar patterns during singing and in response to song playback (<xref ref-type="bibr" rid="bib37">Prather et al., 2008</xref>; <xref ref-type="bibr" rid="bib14">Fujimoto et al., 2011</xref>). However, the singing-related action potential activity of HVC<sub>X</sub> cells does not change when auditory feedback is masked by noise bursts (<xref ref-type="bibr" rid="bib23">Kozhevnikov and Fee, 2007</xref>; <xref ref-type="bibr" rid="bib37">Prather et al., 2008</xref>). These and other observations (<xref ref-type="bibr" rid="bib18">Hessler and Doupe, 1999</xref>) are consistent with the idea that in singing birds, HVC<sub>X</sub> cells provide the AFP with a motor-related efference copy rather than an auditory feedback signal.</p><p>Although extracellular recordings indicate that HVC<sub>X</sub> cell output in singing birds is insensitive to feedback perturbations over tens of minutes, a recent imaging study found that deafening shrinks and destabilizes dendritic spines on HVC<sub>X</sub> neurons within ∼12–48 hr, and these changes precede and predict the severity of song degradation (<xref ref-type="bibr" rid="bib45">Tschida and Mooney, 2012</xref>). Therefore, a remaining possibility is that HVC<sub>X</sub> cells receive subthreshold feedback signals during singing that modify their dendritic spines and more slowly alter HVC<sub>X</sub> action potential output, and thus AFP activity, over a period of hours to days. One prediction of this model is that synaptic inputs onto HVC<sub>X</sub> cells will be acutely sensitive to feedback perturbation. In this study, we tested this idea using sharp intracellular current clamp recordings in freely singing zebra finches exposed to DAF. A second prediction of this model is that the deafening-induced changes in HVC<sub>X</sub> spines will be caused by feedforward changes to HVC’s auditory afferents rather than by AFP-dependent mechanisms that drive deafening-induced vocal plasticity. To test this prediction, we imaged HVC<sub>X</sub> dendritic spines in zebra finches that received LMAN lesions prior to deafening.</p></sec><sec id="s2" sec-type="results"><title>Results</title><sec id="s2-1"><title>Singing-related subthreshold activity of HVC<sub>X</sub> cells is insensitive to DAF</title><p>To determine whether synaptic inputs onto HVC<sub>X</sub> cells convey auditory feedback signals during singing, we made intracellular sharp electrode recordings from HVC neurons in unrestrained, young adult (∼95 days post hatch [dph]) male zebra finches using a modified version of recording techniques developed recently (<xref ref-type="bibr" rid="bib24">Lee et al., 2006</xref>; <xref ref-type="bibr" rid="bib29">Long et al., 2010</xref>) (<xref ref-type="fig" rid="fig1s1">Figure 1—figure supplement 1</xref>). We recorded from a total of 72 HVC<sub>X</sub> neurons in 11 birds as they engaged in spontaneous bouts of singing produced in social isolation (i.e., undirected song) and/or listened to playback of the bird’s own song (BOS). HVC<sub>X</sub> cells were identified either by antidromic stimulation methods (<xref ref-type="fig" rid="fig1">Figure 1A,B</xref>), DC current-evoked firing patterns (<xref ref-type="bibr" rid="bib32">Mooney, 2000</xref>), or characteristic singing-related hyperpolarization (<xref ref-type="bibr" rid="bib29">Long et al., 2010</xref>). As previously reported (<xref ref-type="bibr" rid="bib29">Long et al., 2010</xref>), all HVC<sub>X</sub> cells we recorded without current injection showed spontaneous, regular action potential activity when the bird was not vocalizing (8.2 ± 5.8 Hz, mean ±SD), entered a hyperpolarized state within several hundred milliseconds before song onset (baseline −58.5 ± 7.1 mV, during singing −62.3 ± 7.3 mV), and all except two exhibited one or more action potential bursts during the utterance of the stereotyped sequence of syllables constituting the song motif (<xref ref-type="fig" rid="fig1">Figure 1D</xref>, <xref ref-type="other" rid="video1">Video 1</xref>). Each cell’s pattern of subthreshold membrane potential activity (V<sub>m</sub>) and action potential bursts were highly stereotyped from one motif to the next and from bout to bout (<xref ref-type="fig" rid="fig1">Figure 1D</xref>).<fig-group><fig id="fig1" position="float"><object-id pub-id-type="doi">10.7554/eLife.01833.003</object-id><label>Figure 1.</label><caption><title>Sharp intracellular recordings from sensorimotor neurons in singing birds.</title><p>(<bold>A</bold>) Schematic illustrates the configuration of the in vivo intracellular recording methods used to measure subthreshold activity of HVC<sub>X</sub> neurons. Abbreviations: AFP, anterior forebrain pathway (light blue); SMP, song motor pathway (red); RA, robust nucleus of arcopallium; DLM, dorsolateral division of the medial thalamus; LMAN, lateral magno-cellular nucleus of the anterior nidopallium; VRN, brain stem vocal respiratory network, which includes the tracheosyringeal portion of the hypoglossal motor nucleus (nXIIts) and the respiratory premotor neurons located in the rostral ventrolateral medulla (RVL); R, rostral; D, dorsal. (<bold>B</bold>) Antidromic identification of HVC<sub>X</sub> neurons was achieved by electrically stimulating Area X combined with spike collision tests. (<bold>C</bold>) The experimental design used to generate distorted auditory feedback (DAF). Shortly (∼8 to 10 ms) after detecting that the bird sang the target syllable, a recorded version of the target syllable or a noise burst was played to the bird through a speaker (hit); DAF was suppressed on randomly chosen trials (catch). (<bold>D</bold>) Examples of HVC<sub>X</sub> intracellular membrane potential recordings during hit and catch conditions. Top: sonogram. Middle: simultaneously recorded HVC<sub>X</sub> neuron membrane potential. Bottom: expanded view of membrane potential traces aligned to the onset of the entire motif (iabcd), which was sung twice in this bout. The timing of DAF is shown in boxed and shaded regions.</p><p><bold>DOI:</bold><ext-link ext-link-type="doi" xlink:href="10.7554/eLife.01833.003">http://dx.doi.org/10.7554/eLife.01833.003</ext-link></p></caption><graphic xlink:href="elife01833f001"/></fig><fig id="fig1s1" position="float" specific-use="child-fig"><object-id pub-id-type="doi">10.7554/eLife.01833.004</object-id><label>Figure 1—figure supplement 1.</label><caption><title>The design of the integrated intracellular microdrive used in these experiments.</title><p>(<bold>A</bold>) Schematic diagram of sharp intracellular microdrive, based on a Microdrive design developed by Michale Fee. (<bold>B</bold>) A male zebra finch with a microdrive implanted over HVC.</p><p><bold>DOI:</bold><ext-link ext-link-type="doi" xlink:href="10.7554/eLife.01833.004">http://dx.doi.org/10.7554/eLife.01833.004</ext-link></p></caption><graphic xlink:href="elife01833fs001"/></fig><fig id="fig1s2" position="float" specific-use="child-fig"><object-id pub-id-type="doi">10.7554/eLife.01833.005</object-id><label>Figure 1—figure supplement 2.</label><caption><title>Flow of pre-target and target syllable detection.</title><p>(1) First, a multi-taper estimate of spectrum density (<xref ref-type="bibr" rid="bib30">Mitra and Pesaran, 1999</xref>; <xref ref-type="bibr" rid="bib44">Thomson, 1982</xref>) of the most recent sound segment (5 ms duration) is calculated. We used the first two Slepian sequences (calculated by dpss function in MATLAB) with NW parameter = 4. (2) Features including sound amplitude, mean frequency, Wiener entropy, frequency modulation (FM), and pitch are calculated. A sound event is detected as a continuous sound chunk during which sound amplitude is higher than a threshold (set to ∼10 times from the baseline noise level). (3) At the end of each sound event, mean and variance of spectral features during the sound event are calculated to identify the pre-target syllable, using a support vector machine (SVM) algorithm. Only two dimensions from the test samples are shown. (4) Once the pre-target syllable is detected, the program starts to detect the onset of target syllables defined as the sound segments that satisfy pre-set conditions for at least 3 ms. If all the conditions are not satisfied for a fixed duration (∼100 ms), the program stops detecting the target syllable and starts to look for the pre-target syllable. (5) When the onset of the target syllable is detected, DAF is either executed immediately or suppressed with some probability (usually set to p=0.5).</p><p><bold>DOI:</bold><ext-link ext-link-type="doi" xlink:href="10.7554/eLife.01833.005">http://dx.doi.org/10.7554/eLife.01833.005</ext-link></p></caption><graphic xlink:href="elife01833fs002"/></fig><fig id="fig1s3" position="float" specific-use="child-fig"><object-id pub-id-type="doi">10.7554/eLife.01833.006</object-id><label>Figure 1—figure supplement 3.</label><caption><title>Contingent DAF drives adaptive changes in spectral features of the target syllable.</title><p>When spectral features of a syllable meet certain criteria, white noise is played to the bird. This type of contingent DAF protocol is known to induce adaptive changes in the spectral features of song (<xref ref-type="bibr" rid="bib2">Andalman and Fee, 2009</xref>; <xref ref-type="bibr" rid="bib48">Warren et al., 2011</xref>). The criteria used here to trigger DAF were (1) sound amplitude ∼1–5 dB above baseline and (2) a mean frequency above or below a certain threshold (orange regions in [<bold>C</bold> and <bold>D</bold>]). This frequency threshold was iteratively adjusted to induce vocal changes in the birds. (<bold>A</bold> and <bold>B</bold>) Examples of target syllables (syllable <italic>b</italic>) that did not receive DAF (<bold>A</bold>) and that received DAF (<bold>B</bold>; denoted as <italic>b’</italic>). (<bold>C</bold>) Distribution of target syllable frequency as a function of number of syllable renditions (black dots, 10 ms window measured from target onset; red line is running average of syllable frequency). (<bold>D</bold>) Mean ± SD of the target syllable frequency on each day. The mean frequency of the target syllable shifted significantly after several days of contingent DAF in both upward (day1–4) and downward (day 5–9) directions (<italic>t</italic>-test, day 1 vs day 4, day 4 vs day 7, day 7 vs day 9, p<10<sup>−33</sup>). Here, the mean frequency threshold was set to <5000 Hz (days 1–4, escape in upward shift), >5800 Hz (days 5–7, escape in downward shift), and >5400 Hz (days 8 and 9, escape in downward shift).</p><p><bold>DOI:</bold><ext-link ext-link-type="doi" xlink:href="10.7554/eLife.01833.006">http://dx.doi.org/10.7554/eLife.01833.006</ext-link></p></caption><graphic xlink:href="elife01833fs003"/></fig></fig-group><media content-type="glencoe play-in-place height-250 width-310" id="video1" mime-subtype="mp4" mimetype="video" xlink:href="elife01833v001.mp4"><object-id pub-id-type="doi">10.7554/eLife.01833.017</object-id><label>Video 1.</label><caption><title>Sharp intracellular current clamp recording made in an identified HVC<sub>X</sub> neuron from a freely behaving and singing male zebra finch.</title><p>Left; microphone (blue), membrane potential (green), and speaker output (red) updated at 1 Hz. Right; simultaneously monitored image recorded through a webcam in the recording chamber. In this movie, sound is recorded through the webcam. The whole movie is played in normal speed (x1). Injection current was changed from negative (−0.67 nA), to zero current (0 nA).</p><p><bold>DOI:</bold><ext-link ext-link-type="doi" xlink:href="10.7554/eLife.01833.017">http://dx.doi.org/10.7554/eLife.01833.017</ext-link></p></caption></media></p><p>Although previous studies have shown that singing-related action potentials of HVC<sub>X</sub> neurons are insensitive to DAF (<xref ref-type="bibr" rid="bib23">Kozhevnikov and Fee, 2007</xref>; <xref ref-type="bibr" rid="bib37">Prather et al., 2008</xref>), we tested whether singing-related synaptic inputs to HVC<sub>X</sub> cells encode auditory information by perturbing auditory feedback during singing. To distort the bird’s experience of singing-related auditory feedback, we used a computer-controlled real-time system to detect a specific syllable in the bird’s song motif and trigger sound playback during the production of the ensuing ‘target’ syllable (<xref ref-type="fig" rid="fig1">Figure 1C</xref>; ‘Materials and methods’ and <xref ref-type="fig" rid="fig1s2">Figure 1—figure supplement 2</xref>). Playback sounds included either a 100-ms noise burst or a recorded version of one of the bird’s own syllables; using either of these stimuli to distort singing-related auditory feedback has been shown to induce gradual changes to adult song, indicating the existence of neural circuitry that detects these acute feedback perturbations and induces vocal plasticity (<xref ref-type="bibr" rid="bib27">Leonardo and Konishi, 1999</xref>; <xref ref-type="bibr" rid="bib46">Tumer and Brainard, 2007</xref>; <xref ref-type="bibr" rid="bib2">Andalman and Fee, 2009</xref>). To establish the efficacy of this DAF method, we set the sound amplitude slightly above the level that causes song truncation in some initial trials (∼65 dB at the center of the cage). To ensure that any changes in subthreshold activity we might detect were driven by feedback perturbations rather than by acute changes in motor-related activity, we excluded from analysis any trials in which DAF triggered truncation or changes in song tempo (see below for a description of subthreshold activity in those DAF trials that evoked acute motor effects; also see <xref ref-type="bibr" rid="bib42">Sakata and Brainard, 2008</xref>). In separate behavioral experiments, we also confirmed that applying our perturbation protocol in a pitch-contingent manner over a period of days as previously described (<xref ref-type="bibr" rid="bib46">Tumer and Brainard, 2007</xref>) was sufficient to shift the pitch of target syllables (<xref ref-type="fig" rid="fig1s3">Figure 1—figure supplement 3</xref>; the observed hit rate was 30–70%).</p><p>To measure any rapid effects of DAF on the singing-related synaptic activity of HVC<sub>X</sub> cells, we compared membrane potential records from randomly interleaved trials with (‘hit’) and without (‘catch’) DAF (<xref ref-type="fig" rid="fig1 fig2">Figures 1D and 2A</xref>; target hit rate = 50%; observed rate was 30–70%). In a subset of cells, we used current injection to strongly hyperpolarize the membrane potential to maximize our ability to detect feedback-dependent changes in inhibitory inputs. To quantify the amplitude and timing of singing-related synaptic activity, we computed the average deviation of the baseline subtracted membrane potential between hit and catch trials (ΔV<sub>m</sub>; see ‘Materials and methods’) and estimated the onset times of depolarizing postsynaptic potentials (dPSPs) 20–200 ms after DAF onset, which brackets the auditory-evoked synaptic latencies of HVC neurons (<xref ref-type="bibr" rid="bib25">Lei and Mooney, 2010</xref>) and the offset of the DAF stimulus (<xref ref-type="fig" rid="fig2">Figure 2A–D</xref>). As a population, HVC<sub>X</sub> cells exhibited no systematic changes in ΔV<sub>m</sub> or dPSP onset times in the presence of DAF (<xref ref-type="fig" rid="fig2">Figure 2B–D</xref>; n = 34 cells). When treated on a cell-by-cell basis, none showed changes in dPSP onset times, and only two cells showed slight (<1 mV) but significant changes in ΔV<sub>m</sub> (0.01 ≤ p<0.05; see <xref ref-type="fig" rid="fig2">Figure 2A</xref>, <italic>right</italic>, for one of these positive cases), an outcome that could be accounted for by a false positive rate for this sample size (38 comparisons; 4 cells were sampled at two different ‘resting’ potentials). Taken together, these results indicate that fast synaptic inputs onto HVC<sub>X</sub> neurons are insensitive to auditory feedback perturbation during singing.<fig id="fig2" position="float"><object-id pub-id-type="doi">10.7554/eLife.01833.007</object-id><label>Figure 2.</label><caption><title>Synapses onto HVC<sub>X</sub> neurons do not convey auditory feedback signals during singing.</title><p>(<bold>A</bold>) Each row represents (from top to bottom) the speaker output (Speaker), microphone input (MIC), trial-averaged membrane potential activity in hit (red) and catch (black) trials relative to the baseline (defined as −200 to 0 ms before DAF onset), and the difference in membrane potential between hit and catch conditions (ΔV<sub>m</sub>). Two examples of HVC<sub>X</sub> singing-related subthreshold activity are shown. Two time ranges are set to bracket DAF onset and offset (fast: 20–60 ms, red-shaded region covers the fastest synaptic latency of HVC in response to auditory stimuli [<xref ref-type="bibr" rid="bib25">Lei and Mooney, 2010</xref>]; slow: 60–200 ms, purple shaded region covers the remainder of the DAF period). Red horizontal lines indicate the timing of the DAF stimulus presented during each electrophysiological recording. Data are aligned to the onset of the target syllable, and time zero is set to the mean DAF onset. (<bold>B</bold>) Population analysis of time-averaged ΔV<sub>m</sub> in fast and slow response windows showed no significant changes in response to DAF (n = 38 comparisons made from n = 34 cells; four cells were analyzed at both resting and at hyperpolarized membrane potentials. Individual cell-based analysis revealed that all except two cells showed non-significant changes; triangles [p<0.05], crosses [p≥0.05, paired <italic>t</italic>-test]). (<bold>C</bold>) Examples of dPSP onset timings aligned to target syllable onset obtained from a single cell. (<bold>D</bold>) Onset timings of dPSPs measured in catch vs hit trials were indistinguishable (n = 45 dPSPs [1–4 dPSP onsets per cell] from 18 cells; see ‘Materials and methods’ for more information about dPSP onset detection and clustering methods used for peak detection; mean p=0.52, min p=0.06, <italic>t</italic>-test).</p><p><bold>DOI:</bold><ext-link ext-link-type="doi" xlink:href="10.7554/eLife.01833.007">http://dx.doi.org/10.7554/eLife.01833.007</ext-link></p></caption><graphic xlink:href="elife01833f002"/></fig></p></sec><sec id="s2-2"><title>DAF-insensitive HVC<sub>X</sub> cells can respond to auditory stimuli in non-singing states</title><p>Although previous recordings made in anesthetized zebra finches indicate that most HVC<sub>X</sub> cells can respond to auditory stimulation, and the sample of HVC<sub>X</sub> cells that we tested with DAF is reasonably large (n = 34), it remains possible that we sampled entirely from a subset of cells that do not receive auditory input. However, we found that DAF-insensitive HVC<sub>X</sub> cells could respond to auditory presentation of the bird’s own song during periods when the bird was not singing, including during the daytime or minutes to hours into the night (<xref ref-type="fig" rid="fig3">Figure 3A–B</xref>; 7/8 DAF-insensitive cells showed significant subthreshold responses to BOS playback during the day, and 4/4 DAF-insensitive cells that we tracked across the day–night boundary showed significant sub- and suprathreshold responses during darkness). Moreover, the majority of HVC<sub>X</sub> cells we recorded from, including those in which we did not collect singing-related activity, displayed BOS-evoked auditory activity during the day or night (<xref ref-type="fig" rid="fig3s1">Figure 3—figure supplement 1</xref>, subthreshold responses: n = 15/27 cells in day; 10/11 cells in night; suprathreshold responses: 5/12 cells in day; 7/8 cells in night). Indeed, the increased proportion of auditory-responsive HVC<sub>X</sub> cells we detected during the night is consistent with prior extracellular studies that show auditory responses in the zebra finch HVC are strongly state-dependent (<xref ref-type="bibr" rid="bib5">Cardin and Schmidt, 2003</xref>; <xref ref-type="bibr" rid="bib39">Rauske et al., 2003</xref>; <xref ref-type="bibr" rid="bib33">Nick and Konishi, 2005</xref>). Therefore, we presume that many of the DAF-insensitive cells that we recorded receive auditory input but that these inputs are not activated during singing.<fig-group><fig id="fig3" position="float"><object-id pub-id-type="doi">10.7554/eLife.01833.008</object-id><label>Figure 3.</label><caption><title>DAF-insensitive HVC<sub>X</sub> cells can respond to auditory stimuli in non-singing states.</title><p>(<bold>A</bold>) Examples of trial-averaged singing-related subthreshold activity of DAF-insensitive cells in hit and catch conditions, following the same scheme shown in <xref ref-type="fig" rid="fig2">Figure 2A</xref>. (<bold>B</bold>) Examples of averaged subthreshold responses to BOS playback in the cells shown in (<bold>A</bold>) in day (light on) or night (light off) conditions. The significance of the BOS-evoked auditory response measured in non-singing states was determined either by the difference in the mean (Mann–Whitney <italic>U</italic>-test) or standard deviations (Ansari–Bradley test, a non-parametric test of variance) of the membrane potential fluctuations during the stimulus period and a pre-stimulus baseline period.</p><p><bold>DOI:</bold><ext-link ext-link-type="doi" xlink:href="10.7554/eLife.01833.008">http://dx.doi.org/10.7554/eLife.01833.008</ext-link></p></caption><graphic xlink:href="elife01833f003"/></fig><fig id="fig3s1" position="float" specific-use="child-fig"><object-id pub-id-type="doi">10.7554/eLife.01833.009</object-id><label>Figure 3—figure supplement 1.</label><caption><title>Summary of auditory responses of HVC<sub>x</sub> cells in non-singing states.</title><p>(<bold>A</bold>) Plots of subthreshold (V<sub>m</sub>, top row) and suprathreshold (bottom row) responses to BOS playback measured by the mean and standard deviations (SDs). N = 30 HVC<sub>X</sub> cells in total. Significance level of auditory responses are tested using either the Mann–Whitney <italic>U</italic>-test (p<0.05 using ranksum function in MATLAB) or the Ansari–Bradley test (non-parametric test of variance, p<0.05, applied only when there is no significant difference in mean. Using ansaribradley function in MATLAB.) Circles, significant cells; Crosses, non-significant cells. (<bold>B</bold>) Summary of auditory responsive cells in day and night time. Subthreshold responses: 56% in day (15/27), 91% in night (10/11). Suprathreshold responses: 42% (5/12) in day, 88% (7/8) in night. (<bold>C</bold> and <bold>D</bold>) Bar plots showing the increased strength of subthreshold (<bold>C</bold>) and suprathreshold (<bold>D</bold>) responses to BOS playback during day and night-time recordings.</p><p><bold>DOI:</bold><ext-link ext-link-type="doi" xlink:href="10.7554/eLife.01833.009">http://dx.doi.org/10.7554/eLife.01833.009</ext-link></p></caption><graphic xlink:href="elife01833fs004"/></fig></fig-group></p><p>Prior studies in other songbirds indicate that a subset of HVC<sub>X</sub> cells display a strong correspondence in their auditory and motor-related activity (<xref ref-type="bibr" rid="bib37">Prather et al., 2008</xref>; <xref ref-type="bibr" rid="bib14">Fujimoto et al., 2011</xref>), and this sensorimotor ‘mirroring’ is speculated to arise from the integration of feedback- and motor-related signals by these cells (<xref ref-type="bibr" rid="bib17">Hanuschkin et al., 2013</xref>). Although prior studies have not described auditory–vocal mirroring in HVC of zebra finches, longer intracellular recordings (>3 hr) that we made in two DAF-insensitive HVC<sub>X</sub> cells revealed clear evidence of auditory–vocal mirroring (<xref ref-type="fig" rid="fig4">Figure 4A,B</xref>). Specifically, at later times during the night, we detected BOS-evoked sub- and supra-threshold activity that closely resembled the singing-related activity recorded earlier during the daytime. Moreover, in a larger subset of HVC<sub>X</sub> cells in which we recorded singing-related activity with or without DAF during the day and tracked across the day–night boundary, we found that the correlation in membrane potential activity recorded during singing and in response to BOS playback increased at later times in the night (<xref ref-type="fig" rid="fig4">Figure 4C</xref>). These findings show that auditory–vocal mirror neurons in the zebra finch HVC are insensitive to auditory feedback perturbation during singing and also suggest that sensorimotor mirroring may be a latent feature of a large cohort of HVC<sub>X</sub> cells in this species.<fig id="fig4" position="float"><object-id pub-id-type="doi">10.7554/eLife.01833.010</object-id><label>Figure 4.</label><caption><title>Auditory-vocal mirror neurons in zebra finches.</title><p>(<bold>A</bold>) Examples of singing and auditory-related membrane potential activity of two HVC<sub>X</sub> cells aligned to each bird’s motif. Initially weak but significant auditory responses during the day (black) become robust during the night (blue) and precisely mirror the cell’s singing-related activity (red). The cell at the top is the same cell shown in <xref ref-type="fig" rid="fig3">Figure 3</xref> (<bold>A</bold>) right, which lacked DAF sensitivity. (<bold>B</bold>) Emergence of robust BOS-evoked subthreshold activity as the night progresses. Top: BOS responses near the day–night boundary (same cell shown in <xref ref-type="fig" rid="fig4">Figure 4A</xref>, top). Baseline-subtracted membrane potential responses aligned to the BOS playback (top panel) and shown in the order of the recording time. Bottom: normalized voltage trace of the cell’s BOS evoked activity at day (black), night (blue) and singing-related activity (red). (<bold>C</bold>) The same cell’s action potential response to BOS playback, plotted in the same scheme used in (<bold>B</bold>). 1–2 hr after nightfall, strong BOS-evoked action potential activity was detected that closely mirrored the singing-related activity recorded from this same cell 4–5 hr earlier, during the daytime. (<bold>D</bold>) Similarity of a cell’s subthreshold activity during singing and hearing the same motif in the day or night measured by the correlation coefficient (C.C.) of averaged membrane potential records reveals enhanced similarity of singing-related and BOS-evoked activity later in the night (≥30 min after light off). Data are only from auditory-responsive cells in which we also collected singing data (singing vs hearing data during the day: n = 7, C.C. = 0.076 ± 0.041 (SEM); Early night, < 30 min after light off: n = 7, C.C. = 0.077 ± 0.061; Late night, ≥ 30 min after light off: n = 4, C.C. = 0.302 ± 0.105. Mann–Whitney <italic>U</italic>-test, p=0.04 compared to day).</p><p><bold>DOI:</bold><ext-link ext-link-type="doi" xlink:href="10.7554/eLife.01833.010">http://dx.doi.org/10.7554/eLife.01833.010</ext-link></p></caption><graphic xlink:href="elife01833f004"/></fig></p></sec><sec id="s2-3"><title>Intracellular methods can detect DAF-related synaptic activity in cells ventral to HVC</title><p>A remaining issue is whether the intracellular recording methods we used can detect small subthreshold membrane potential fluctuations in cells that are likely to receive input from DAF-sensing cells. In a small subset of experiments, we made intracellular recordings (n = 14 neurons in 5 birds) in the region immediately ventral to HVC (i.e., HVC shelf, <xref ref-type="fig" rid="fig5s1">Figure 5—figure supplement 1A</xref>), which is known to receive input from auditory regions (<xref ref-type="bibr" rid="bib12">Fortune and Margoliash, 1995</xref>; <xref ref-type="bibr" rid="bib47">Vates et al., 1996</xref>) that contain DAF-sensitive neurons (<xref ref-type="bibr" rid="bib21">Keller and Hahnloser, 2009</xref>). During singing, we found that most (5/9) HVC shelf cells showed significantly greater subthreshold depolarizing responses in the presence of DAF, consistent with the notion that they receive synaptic input from neurons that exhibit DAF-sensitive action potential output (<xref ref-type="fig" rid="fig5">Figure 5A,C, D</xref>; one of these cells also showed a significant suprathreshold response to DAF, <xref ref-type="fig" rid="fig5">Figure 5B</xref>). These recordings also revealed that synaptic activity in HVC shelf cells can be modulated during singing, although the pattern of this activity was less stereotyped from one motif to the next than observed in HVC<sub>X</sub> cells (<xref ref-type="fig" rid="fig5s1">Figure 5—figure supplement 1B</xref>. HVC shelf: n = 14 cells, C.C. = 0.12 ± 0.03 (SEM). HVC<sub>X</sub> cells: n = 53 cells, C.C. = 0.70 ± 0.02, without time warping). These data confirm that the intracellular recording technique used here can detect subthreshold membrane potential responses to DAF in the singing zebra finch, further strengthening the idea that auditory-related synaptic inputs to HVC<sub>X</sub> cells are gated in an ‘off’ position during singing.<fig-group><fig id="fig5" position="float"><object-id pub-id-type="doi">10.7554/eLife.01833.011</object-id><label>Figure 5.</label><caption><title>Synapses onto HVC shelf neurons convey DAF signals during singing.</title><p>(<bold>A</bold>) Examples of trial-averaged subthreshold activity patterns of DAF-sensitive HVC shelf cells in hit and catch conditions, following the same scheme as <xref ref-type="fig" rid="fig2">Figure 2A</xref>. Exposure to DAF significantly increased subthreshold depolarizations during singing. (<bold>B</bold>) An HVC shelf cell that showed a significant suprathreshold response to DAF, following the same scheme shown in <xref ref-type="fig" rid="fig2">Figure 2A</xref> except for the addition of a spike-rastergram in the second row. (<bold>C</bold>) Time-averaged ΔV<sub>m</sub> in fast and slow response windows (n = 9 cells, triangles [p<0.05], crosses [p≥0.05, paired t-test]). (<bold>D</bold>) DAF sensitivity plot for HVC shelf neurons and HVC<sub>X</sub> cells. Nearly half of the HVC shelf neurons showed DAF sensitivity. The minimum p-values of time-averaged ΔV<sub>m</sub> either in fast or slow time window are plotted, Mann–Whitney <italic>U</italic>-test.</p><p><bold>DOI:</bold><ext-link ext-link-type="doi" xlink:href="10.7554/eLife.01833.011">http://dx.doi.org/10.7554/eLife.01833.011</ext-link></p></caption><graphic xlink:href="elife01833f005"/></fig><fig id="fig5s1" position="float" specific-use="child-fig"><object-id pub-id-type="doi">10.7554/eLife.01833.012</object-id><label>Figure 5—figure supplement 1.</label><caption><title>HVC shelf cells show subthreshold DAF sensitivity.</title><p>(<bold>A</bold>) An example of intracellular membrane potential recordings from an HVC shelf neuron during DAF hit and catch conditions, following the same scheme as <xref ref-type="fig" rid="fig1">Figure 1D</xref>. Scale bars, 500 ms × 10 mV. (<bold>B</bold>). Correlation Coefficient (C.C.) of membrane potential traces of individual HVC<sub>X</sub> and HVC shelf cells recorded while birds sang many renditions of the same motif. No time warping is applied to the data. Mann–Whitney <italic>U</italic>-test, p<10<sup>−7</sup>.</p><p><bold>DOI:</bold><ext-link ext-link-type="doi" xlink:href="10.7554/eLife.01833.012">http://dx.doi.org/10.7554/eLife.01833.012</ext-link></p></caption><graphic xlink:href="elife01833fs005"/></fig></fig-group></p></sec><sec id="s2-4"><title>Evidence that HVC<sub>X</sub> cells encode motor-related information</title><p>Our findings indicate that HVC<sub>X</sub> cells are insensitive to auditory feedback perturbation during singing, raising the possibility that they primarily encode motor-related information. One expectation of a motor-related signal is that it should precede sound onset during singing. To test this idea, we examined the relationship between the onsets of dPSPs and syllables during bouts of singing (<xref ref-type="fig" rid="fig6">Figure 6A–C</xref>). Population-averaged cross-correlation analysis revealed that dPSP onsets in HVC<sub>X</sub> cells occur on average ∼25 ms before syllable onsets (n = 53 cells, <xref ref-type="fig" rid="fig6">Figure 6D</xref>), similar to the findings from prior extracellular studies (<xref ref-type="bibr" rid="bib23">Kozhevnikov and Fee, 2007</xref>; <xref ref-type="bibr" rid="bib14">Fujimoto et al., 2011</xref>). In contrast, dPSP onsets were not correlated with syllable offsets (<xref ref-type="fig" rid="fig6">Figure 6D</xref>), indicating that singing-related synaptic inputs to HVC<sub>X</sub> cells are correlated to future, rather than ongoing or past, vocal output. Another expectation of a motor-related signal is that, should change in parallel to changes in vocal output. As previously mentioned, DAF can trigger acute effects, such as truncation or slowing of song tempo (<xref ref-type="fig" rid="fig6">Figure 6E,G</xref>) (<xref ref-type="bibr" rid="bib42">Sakata and Brainard, 2008</xref>), which are distinct from the slower changes in song that are associated with feedback-dependent error correction mechanisms. In fact, we observed significant changes in membrane potential fluctuations of HVC<sub>X</sub> cells when DAF-triggered song truncation or delays in syllable onsets (<xref ref-type="fig" rid="fig6">Figure 6F,H</xref>). Therefore, the intracellular method we used is capable of detecting DAF-related changes in the subthreshold activity of HVC<sub>X</sub> cells, but only when DAF triggers acute vocal motor effects. Taken together, these findings suggest that synaptic inputs to HVC<sub>X</sub> cells do not convey auditory feedback signals during singing and instead encode information about future vocal output, consistent with a motor-related signal.<fig id="fig6" position="float"><object-id pub-id-type="doi">10.7554/eLife.01833.013</object-id><label>Figure 6.</label><caption><title>Synaptic inputs onto HVC<sub>X</sub> cells encode future syllable onsets.</title><p>(<bold>A</bold>) An example of a stereotyped syllable sequence, with syllables underlined by orange bars. (<bold>B</bold>) Pseudo-color representation of the membrane potential patterns relative to baseline while the bird sang many repetitions of the syllable sequence shown in (<bold>A</bold>). Black dots; syllable onsets. Data are aligned to a specific syllable and sorted from long to short motifs, which reflect natural variations in zebra finch song tempo. (<bold>C</bold>) Same data shown in (<bold>B</bold>), but dPSP onsets are shown by black dots, and the timing of individual syllables is represented by the orange regions. (<bold>D</bold>) Syllable onset- and offset-triggered average of dPSP onset rate (n = 53 HVC<sub>X</sub> cells). (<bold>E</bold>) Example spectrograms from catch and ‘hit-and-truncated’ songs. (<bold>F</bold>) Trial-averaged subthreshold activity during hit-and-truncated songs showed a clear deviation from activity during catch trials. (<bold>G</bold>) Cumulative histogram of syllable onset timings relative to DAF onset in hit trials including delayed syllables (red dashed line), hit trials without delayed syllables (red solid line), and catch trials (black line). (<bold>H</bold>) Examples of trial-averaged subthreshold activity reveal that delayed motor output drives changes in the subthreshold activity (compare black and red dashed lines). After removing hit trials with delayed syllables, the subthreshold activity in hit and catch conditions is nearly identical (black vs red solid lines).</p><p><bold>DOI:</bold><ext-link ext-link-type="doi" xlink:href="10.7554/eLife.01833.013">http://dx.doi.org/10.7554/eLife.01833.013</ext-link></p></caption><graphic xlink:href="elife01833f006"/></fig></p></sec><sec id="s2-5"><title>Deafening-induced changes to HVC<sub>X</sub> spines depends on an intact AFP</title><p>Our finding that HVC<sub>X</sub> cells lack feedback-sensitive synaptic activity during singing raises the question of how deafening drives structural and functional changes to synapses on HVC<sub>X</sub> cells that precede song degradation (<xref ref-type="bibr" rid="bib45">Tschida and Mooney, 2012</xref>). One possibility is that feedback perturbations are detected by auditory neurons indirectly presynaptic to HVC<sub>X</sub> cells, and information from these feedback-sensing cells alters synapses on HVC<sub>X</sub> cells more gradually as a prelude to the AFP-mediated error correction processes that result in vocal plasticity (<xref ref-type="fig" rid="fig7">Figure 7A</xref>). In this ‘feedforward model’, disrupting auditory feedback should be sufficient to induce synaptic changes in HVC<sub>X</sub> cells regardless of any downstream processes in the AFP. Another possibility is that deafening-induced changes to synapses on HVC<sub>X</sub> cells depend on downstream processes in the AFP that are known to be critical for feedback-dependent changes to song (<xref ref-type="fig" rid="fig7">Figure 7B</xref>). Indeed, contrary to the long-held assumption that activity only propagates from HVC to the AFP, a recent study established that activity could propagate from LMAN through recurrent circuitry to HVC (<xref ref-type="bibr" rid="bib16">Hamaguchi and Mooney, 2012</xref>). Therefore, a remaining possibility is that deafening-induced changes to synapses on HVC<sub>X</sub> cells depend on LMAN activity and not simply on the removal of auditory feedback to HVC, as predicted by the feedforward model.<fig id="fig7" position="float"><object-id pub-id-type="doi">10.7554/eLife.01833.014</object-id><label>Figure 7.</label><caption><title>Two models of auditory feedback-dependent vocal and HVC dendritic plasticity.</title><p>(<bold>A</bold>) A model where deafening triggers slow changes to HVC<sub>X</sub> spines, subsequently altering AFP activity, which in turn drives vocal plasticity. In this model, LMAN lesions prevent deafening-induced song degradation but will not prevent deafening-induced changes to HVC<sub>X</sub> spines. (<bold>B</bold>) A model where deafening acts through LMAN to trigger song plasticity and also to drive changes to HVC<sub>X</sub> spines. In this model, LMAN lesions will prevent both deafening-induced song degradation and deafening-induced changes to HVC<sub>X</sub> spines.</p><p><bold>DOI:</bold><ext-link ext-link-type="doi" xlink:href="10.7554/eLife.01833.014">http://dx.doi.org/10.7554/eLife.01833.014</ext-link></p></caption><graphic xlink:href="elife01833f007"/></fig></p><p>To distinguish between these possibilities, we used multiphoton in vivo imaging methods to monitor the size and stability of dendritic spines on HVC<sub>X</sub> cells in adult male zebra finches that received bilateral LMAN lesions prior to surgical deafening (n = 3 male zebra finches, 120–140 dph). As previously described (<xref ref-type="bibr" rid="bib40">Roberts et al., 2010</xref>; <xref ref-type="bibr" rid="bib45">Tschida and Mooney, 2012</xref>), a GFP-lentivirus was injected into HVC to label HVC cells, and a retrograde tracer was injected bilaterally into Area X to facilitate identification of HVC<sub>X</sub> cells. Additionally, we placed bilateral lesions in LMAN with ibotenic acid 4–5 days before the first imaging session, which was conducted during the bird’s subjective nighttime (<xref ref-type="fig" rid="fig8">Figure 8A</xref>). The morning following the first imaging session, the birds were deafened by bilateral surgical removal of the cochlea, which triggers a form of AFP-dependent song degradation that we confirmed here can be dramatically reduced by prior LMAN lesions (<xref ref-type="fig" rid="fig8">Figure 8A</xref>; only syllables that underwent significant degradation within the first 3d post-deafening are shown; p values reported in <xref ref-type="fig" rid="fig8">Figure 8A</xref> for difference in magnitude of spectral change between LMAN lesion and LMAN intact birds; also see below). Surprisingly, in addition to greatly reducing song degradation, LMAN lesions completely prevented the decrease in HVC<sub>X</sub> cell spine size index that normally follows deafening (<xref ref-type="fig" rid="fig8">Figure 8B–C</xref>; 471 spines from 7 HVC<sub>X</sub> neurons in 3 LMAN lesion-deafened birds; 495 spines from 7 HVC<sub>X</sub> cells in 6 LMAN-intact deafened birds; spine data from LMAN intact-deafened birds were previously reported in <xref ref-type="bibr" rid="bib45">Tschida and Mooney, 2012</xref>; p=0.001 for difference between groups across all time bins). In fact, the post-deafening spine size index values from HVC<sub>X</sub> cells in deafened birds with LMAN lesions measured here did not differ from values measured from longitudinally imaged birds with normal hearing (data not shown; spine data from birds with normal hearing were previously reported in <xref ref-type="bibr" rid="bib45">Tschida and Mooney, 2012</xref>). Lesions to LMAN also prevented deafening-induced decreases in HVC<sub>X</sub> cell spine stability, and spine stability measurements from HVC<sub>X</sub> cells in LMAN lesion-deafened birds were significantly higher than those from LMAN intact-deafened HVC<sub>X</sub> cells in all post-song degradation time bins (<xref ref-type="fig" rid="fig8">Figure 8C</xref>; 2401 spines from 7 HVC<sub>X</sub> cells in 3 LMAN lesion-deafened birds; 3562 spines from 14 HVC<sub>X</sub> cells in 9 LMAN intact-deafened birds; spine data from LMAN intact-deafened birds were previously reported in <xref ref-type="bibr" rid="bib45">Tschida and Mooney, 2012</xref>; p≤0.02 for difference between groups in last 3 time bins). Therefore, in contrast to a feedforward model of auditory feedback processing by HVC, deafening-induced changes to dendritic spines on HVC<sub>X</sub> cells depend on an intact LMAN.<fig id="fig8" position="float"><object-id pub-id-type="doi">10.7554/eLife.01833.015</object-id><label>Figure 8.</label><caption><title>LMAN lesions prevent decreases in spine size and spine stability in HVC<sub>X</sub> neurons following deafening.</title><p>(<bold>A</bold>) Upper left: schematic illustrates surgical manipulations for in vivo imaging experiments. Lower left: quantification of deafening-induced changes to syllable entropy and EV at 3 days post-deafening in LMAN intact (black) and LMAN lesion (red) birds (only syllables that underwent significant degradation by 3 days post-deafening are shown; song degradation analyzed in a total of 6 LMAN lesion birds (3 used for imaging) and 19 LMAN intact birds (13 used for imaging in <xref ref-type="bibr" rid="bib45">Tschida and Mooney, 2012</xref>); entropy: 14 syllables from 10 LMAN intact birds, 5 syllables from 5 LMAN lesion birds; EV: 12 syllables from 9 LMAN intact birds, 8 syllables from 4 LMAN intact birds). Right: representative songs before and after deafening from LMAN lesion and LMAN intact birds. Deafened birds that received LMAN lesions still undergo subtle but significant song degradation. (<bold>B</bold>) Representative in vivo, two-photon images of HVC<sub>X</sub> neurons showing changes in spine size between 2 and 3 days post-deafening in deafened birds with or without LMAN lesions. LMAN lesions prevent the decrease in the size of HVC<sub>X</sub> spines that normally follows deafening (size index <1 in LMAN intact birds). Scale bars, 5 µm. (<bold>C</bold>) Spine size index (left) and spine stability (right) is significantly higher in deafened birds with LMAN lesions. Quantification of spine size (left) and stability (right) in HVC<sub>X</sub> neurons from deafened birds with (red) and without (black) LMAN lesions (LMAN intact data were previously reported in <xref ref-type="bibr" rid="bib45">Tschida and Mooney (2012)</xref>, except the values reported here have not been normalized to baseline, pre-deafening measurements). Time bins with significant differences between intact and LMAN lesion groups are indicated with asterisks (p<0.05).</p><p><bold>DOI:</bold><ext-link ext-link-type="doi" xlink:href="10.7554/eLife.01833.015">http://dx.doi.org/10.7554/eLife.01833.015</ext-link></p></caption><graphic xlink:href="elife01833f008"/><permissions><copyright-statement>© 2012, Elsevier Inc., All Rights Reserved</copyright-statement><copyright-year>2012</copyright-year><copyright-holder>Elsevier Inc.</copyright-holder><license><license-p>Figure 8, part (<italic>B</italic>) (images of HVC<sub>X</sub> dendrites from LMAN intact bird) is reprinted with permission from Figure 1B in <xref ref-type="bibr" rid="bib45">Tschida and Mooney, 2012</xref>.</license-p></license></permissions></fig></p><p>We also conducted several control measurements to exclude the possibility that LMAN lesions by themselves trigger reactive changes in HVC circuitry or to the bird’s song that might occlude or otherwise alter spine dynamics in HVC. First, spine dynamics in HVC<sub>X</sub> cells following LMAN lesions but prior to deafening did not differ from values obtained from similarly aged birds with normal hearing and an intact LMAN (<xref ref-type="fig" rid="fig9">Figure 9A</xref>; HVC<sub>X</sub> cell spine size index: 1.10 ± 0.06 in LMAN lesion group; 1.07 ± 0.03 in LMAN intact group, mean ±SEM, p=0.62; HVC<sub>X</sub> cell spine stability: 94.0 ± 1.1% in LMAN lesion group, 92.0 ± 1.6% in LMAN intact group, p=0.61; LMAN intact data previously reported in <xref ref-type="bibr" rid="bib45">Tschida and Mooney, 2012</xref>). Second, prior to deafening, LMAN lesions did not affect the mean or variability of syllable entropy or entropy variance (EV), two features that are sensitive to small changes in song’s acoustic structure (<xref ref-type="fig" rid="fig9">Figure 9B</xref>; Wilcoxon signed-ranks test used to compare pre-lesion to post-lesion values for 25 syllables from 6 birds [includes 3 birds used for imaging]; data shown only for mean entropy and EV; p=0.09 for mean entropy, p=0.92 for coefficient of variation [CV] of entropy, p=0.25 for mean EV, p=0.31 for CV of EV). Third, a post mortem analysis of brain tissue indicated that the lesions we made in LMAN were comparable in volume to those reported in previous studies that have established LMAN’s role in preventing deafening-induced song degradation (40–100% lesion extent, <xref ref-type="bibr" rid="bib4">Brainard and Doupe, 2000</xref>). Finally, although the LMAN lesions we made largely prevented deafening-induced song degradation, birds that received LMAN lesions still showed extremely subtle but significant song degradation following deafening, as previously reported (<xref ref-type="fig" rid="fig8">Figure 8A</xref>) (<xref ref-type="bibr" rid="bib4">Brainard and Doupe, 2000</xref>; <xref ref-type="bibr" rid="bib19">Horita et al., 2008</xref>). Therefore, neither the complete abolition of auditory feedback by deafening nor the expression of small amounts of deafening-induced vocal plasticity in LMAN-lesioned birds are sufficient to drive structural changes to HVC dendritic spines.<fig id="fig9" position="float"><object-id pub-id-type="doi">10.7554/eLife.01833.016</object-id><label>Figure 9.</label><caption><title>In birds with normal hearing, LMAN lesions do not significantly affect HVC<sub>X</sub> spine dynamics or spectral features of song.</title><p>(<bold>A</bold>) LMAN lesions affect neither spine size index nor spine stability in HVC<sub>X</sub> neurons (p=0.62 for size, p=0.61 for stability; LMAN intact data were previously reported in <xref ref-type="bibr" rid="bib45">Tschida and Mooney, 2012</xref>). (<bold>B</bold>) LMAN lesions do not affect mean syllable entropy or EV (pre-lesion songs compared to songs from 4 to 5 days post-lesion; 25 syllables from 6 birds, p=0.09 for difference in mean entropy, p=0.25 for mean EV.</p><p><bold>DOI:</bold><ext-link ext-link-type="doi" xlink:href="10.7554/eLife.01833.016">http://dx.doi.org/10.7554/eLife.01833.016</ext-link></p></caption><graphic xlink:href="elife01833f009"/></fig></p></sec></sec><sec id="s3" sec-type="discussion"><title>Discussion</title><p>Resolving how feedback-related information enters the AFP is a critical step to understand neural mechanisms of vocal learning. HVC<sub>X</sub> cells are attractive candidates for providing this information because they provide auditory input to the AFP (<xref ref-type="bibr" rid="bib41">Roy and Mooney, 2009</xref>), receive inputs that are weakened by deafening (<xref ref-type="bibr" rid="bib45">Tschida and Mooney, 2012</xref>), and may receive inputs from DAF-sensitive neurons, including HVC interneurons (<xref ref-type="bibr" rid="bib42">Sakata and Brainard, 2008</xref>; <xref ref-type="bibr" rid="bib21">Keller and Hahnloser, 2009</xref>). Furthermore, some forms of mammalian cortical plasticity initially involve changes to inhibitory rather than excitatory synapses (<xref ref-type="bibr" rid="bib13">Froemke et al., 2007</xref>; <xref ref-type="bibr" rid="bib20">Keck et al., 2011</xref>), raising the possibility that DAF might alter inhibitory input onto HVC<sub>X</sub> cells prior to any changes in excitatory input. Therefore, although the singing-related action potential output of HVC<sub>X</sub> neurons is insensitive to DAF (<xref ref-type="bibr" rid="bib23">Kozhevnikov and Fee, 2007</xref>; <xref ref-type="bibr" rid="bib37">Prather et al., 2008</xref>), the possibility remained that these cells integrate input from feedback-sensing cells during singing at subthreshold levels, which could then trigger slower changes to synapses on HVC<sub>X</sub> neurons as a prelude to vocal plasticity (<xref ref-type="bibr" rid="bib45">Tschida and Mooney, 2012</xref>). By directly measuring synaptic activity in singing birds, we found that singing-related synaptic activity in HVC<sub>X</sub> neurons lacks real-time sensitivity to feedback perturbation. Moreover, many HVC<sub>X</sub> cells that we recorded displayed auditory responses in non-singing states, including a subset of cells that exhibited auditory-motor mirroring, a property which has been theorized to arise through integration of real-time feedback signals (<xref ref-type="bibr" rid="bib17">Hanuschkin et al., 2013</xref>). Therefore, although HVC<sub>X</sub> cells receive fast synaptic inputs from several sources that may have the potential to convey auditory feedback during singing, these inputs appear to be gated to an ‘off’ position during vocalization in zebra finches (<xref ref-type="bibr" rid="bib43">Shea and Margoliash, 2003</xref>; <xref ref-type="bibr" rid="bib6">Cardin and Schmidt, 2004</xref>; <xref ref-type="bibr" rid="bib9">Coleman et al., 2007</xref>).</p><p>One potential concern is that a synaptic feedback ‘gate’ onto HVC<sub>X</sub> cells may only open after the bird has experienced frequent and prolonged exposure to DAF. Several observations allay this concern. First, the hit rate for applying DAF we used during intracellular recordings (i.e., ∼50%) has been shown here (<xref ref-type="fig" rid="fig1s3">Figure 1—figure supplement 3</xref>) and by others to be sufficiently frequent to drive adaptive changes in pitch (<xref ref-type="bibr" rid="bib2">Andalman and Fee, 2009</xref>; <xref ref-type="bibr" rid="bib8">Charlesworth et al., 2012</xref>; <xref ref-type="bibr" rid="bib1">Ali et al., 2013</xref>). Second, all of the analyses of subthreshold DAF sensitivity we performed were conducted after the bird had been exposed to DAF for at least an entire preceding day, the time required for the bird to habituate to the exposure, as determined by the decreased frequency of song truncations we observed in response to DAF. Finally, because we recorded from individual birds for 1 to 2 weeks and exposed them to DAF throughout much of that time window, many of the cells we recorded from were monitored after the animal had experienced many days or even more than a week of DAF. That is, our intracellular recordings and analyses of subthreshold DAF sensitivity were made in conditions where the bird’s exposure to DAF was both frequent and prolonged. And although we did not track any single cell’s activity while DAF exposure induced song plasticity, a recent study by <xref ref-type="bibr" rid="bib1">Ali et al. (2013)</xref> has shown that singing-related HVC multiunit activity remains unchanged even after contingent DAF is used to drive changes in song pitch. Therefore, we support a model in which auditory synapses onto HVC<sub>X</sub> cells are always gated off when the bird is singing, while also recognizing that experiments involving longer-term intracellular recordings in birds that are undergoing DAF-induced song plasticity are required to fully resolve this issue.</p><p>The absence of any detectable synaptic trace of auditory feedback in HVC<sub>X</sub> cells during singing raises the question of how deafening drives changes to synapses on these cells. One possibility is that auditory feedback information is first transmitted to HVC<sub>X</sub> cells and then to downstream error correction circuits in the AFP, but that this information transfer occurs slowly and only when the bird is not singing. Regardless of the plausibility of such an offline feedforward model, it predicts that removal of auditory feedback independent of downstream error correction processes in the AFP should be sufficient to trigger changes to dendritic spines in HVC. In contrast, we found that LMAN lesions prevent the shrinkage and destabilization of HVC<sub>X</sub> dendritic spines that normally follow deafening. Therefore, deafening-induced changes to HVC<sub>X</sub> spines depend on an intact AFP, rather than resulting simply from the removal of auditory drive to HVC, thus raising the possibility that HVC receives error-correction signals from the AFP (<xref ref-type="bibr" rid="bib1">Ali et al., 2013</xref>). Although the current experiments cannot resolve whether deafening-induced changes in HVC are driven solely by changes in LMAN activity or instead by LMAN acting in concert with the removal of auditory inputs to HVC<sub>X</sub> cells, the total insensitivity of HVC<sub>X</sub> cells to acute feedback perturbations suggests to us that deafening-induced synaptic changes in HVC<sub>X</sub> cells are a consequence rather than a cause of error correction processes mediated by the AFP. The indirect, recurrent pathways that can convey activity from LMAN to HVC (<xref ref-type="bibr" rid="bib16">Hamaguchi and Mooney, 2012</xref>) provide a substrate over which the consequences of error correction processes in the AFP could be transmitted back to the early stages of the vocal motor hierarchy. Furthermore, because LMAN can access multiple stages of the motor hierarchy in tens of milliseconds (<xref ref-type="bibr" rid="bib16">Hamaguchi and Mooney, 2012</xref>) and song may only change in response to altered feedback over hours or days, the consequences of performance evaluation may be widely distributed in the song motor network even before behavioral changes are first evident.</p><p>A remaining important question is how and in what form auditory feedback information enters the AFP to facilitate error correction. Previous studies have shown that input and output cells of the AFP (HVC<sub>X</sub> and LMAN neurons) are not real-time sensors of auditory feedback (<xref ref-type="bibr" rid="bib26">Leonardo, 2004</xref>; <xref ref-type="bibr" rid="bib23">Kozhevnikov and Fee, 2007</xref>; <xref ref-type="bibr" rid="bib37">Prather et al., 2008</xref>), and the current study shows that even those feedback-dependent changes that can be detected in HVC depend on the AFP rather than fast synaptic signaling onto HVC<sub>X</sub> cells from feedback-sensing auditory neurons. These findings suggest that HVC is downstream of the performance evaluation provided by real-time feedback sensors, and also raises the possibility that these sensors communicate error-related information to the AFP independent of HVC. Notably, the ventral tegmental area provides input to Area X (<xref ref-type="bibr" rid="bib28">Lewis et al., 1981</xref>) and receives indirect input from several auditory regions, including the HVC shelf, that contain DAF-sensitive cells (<xref ref-type="bibr" rid="bib49">Wild et al., 1993</xref>; <xref ref-type="bibr" rid="bib12">Fortune and Margoliash, 1995</xref>; <xref ref-type="bibr" rid="bib47">Vates et al., 1996</xref>; <xref ref-type="bibr" rid="bib21">Keller and Hahnloser, 2009</xref>). In this scenario, the consequences of error detection would reach the AFP in the form of neuromodulatory signals from the VTA, an arrangement that would buffer the entire song premotor circuitry from the role of acute feedback processing. In fact, this buffered feedback architecture, along with the gating of auditory inputs to HVC, may be necessary when the frequency of motor control is so fast that a relatively slow feedback signal becomes uninformative for the ongoing motor command.</p></sec><sec id="s4" sec-type="materials|methods"><title>Materials and Methods</title><sec id="s4-1"><title>Subjects</title><p>All experiments were carried out in accordance with a protocol approved by Duke University Institutional Animal Care and Use Committee. Data were collected from 14 adult (age range, 85-150 post hatch days) male zebra finches (<italic>Taeniopygia guttata</italic>): n = 11 for intracellular recordings in freely singing birds, n = 3 for HVC longitudinal imaging combined with LMAN lesion and deafening.</p></sec><sec id="s4-2"><title>Intracellular recordings in singing birds</title><p>Intracellular recordings were made in zebra finches using a modified version of a microdrive (<xref ref-type="fig" rid="fig1s1">Figure 1—figure supplement 1</xref>, original design by M Fee, MIT, <xref ref-type="bibr" rid="bib29">Long et al., 2010</xref>) constructed by stereolithography (Agile Manufacturing, Inc, Ontario, Canada). A miniature motor (Part # 0206A001B+02/1 47:1-Y2825; Micromo, FL) was used to move the micropipette attached to the shuttle, and two additional stainless steel screws were used to stabilize the movement of the micropipette. A miniaturized headstage (by IY, equivalent to HS-2A headstage with gain x0.1, Axon Instruments [now Molecular Devices, CA]) was mounted on the back of the base to send signals and receive command current from the intracellular recording amplifier (AxoClamp-2B; Axon Instruments) through a flexible tether cable (Omnetics, MN). The microdrive was surgically implanted over HVC using stereotaxic coordinates. All recordings were made from the right hemisphere. HVC<sub>X</sub> neurons were identified by antidromic stimulation of Area X with a bipolar stimulating electrode (a pair of 75 μm diameter silver wires, ∼500 μm apart, A-M systems, WA) implanted into Area X, or by their spontaneous, DC-evoked, and singing-related activity. Recordings were attempted for approximately 7–21 days per bird and recording pipettes were changed daily.</p></sec><sec id="s4-3"><title>Data acquisition for intracellular recordings</title><p>Custom MATLAB software with xPCtarget toolbox (by KH) (Mathworks, MA) was used for recording the microphone signal and membrane potential data, real-time detection of the target syllable, and generation of distorted auditory feedback [DAF] (<xref ref-type="fig" rid="fig1s2">Figure 1—figure supplement 2</xref>). In most intracellular recording sessions, we randomly selected a subset of songs as ‘catch’ trials for which DAF playbacks were suppressed (target hit rate 50%, actual hit rate, 30–70%). The microphone and membrane potential data were simultaneously recorded at 44 kHz for feature detection, down-sampled to 22 kHz, and saved for post-hoc analysis.</p></sec><sec id="s4-4"><title>Detection of target syllable and execution of DAF</title><p>For the online detection of the pre-target syllable, spectral features were calculated in short (5 ms) segments of sound and updated at 1000 Hz (<xref ref-type="fig" rid="fig1s2">Figure 1—figure supplement 2</xref>, step 1, 2). Before the experiments, a support-vector machine (SVM) algorithm was used to define the pre-target syllable (SVM algorithm with a Gaussian kernel using soft margins was written by KH in MATLAB). During the experiment, the same SVM algorithm and the same parameter sets (written in Simulink and xPC Target, MATLAB by KH) were used for the online detection of the pre-target syllable (<xref ref-type="fig" rid="fig1s2">Figure 1—figure supplement 2</xref>, step 3). Once the pre-target syllable was detected, spectral features including sound amplitude, mean frequency, Wiener entropy, FM, and pitch values were measured for a duration ∼100 ms (<xref ref-type="fig" rid="fig1s2">Figure 1—figure supplement 2</xref>, step 4). Because the occurrence of the target syllable was almost completely predicted by the detection of the pre-target syllable in zebra finch songs, we defined the occurrence of the target syllable simply as the time when sound amplitude crossed a threshold. The occurrence of the target syllable was confirmed in post-hoc analysis. To avoid false positives (triggering DAF by detecting a natural fluctuation of syllable features), the criteria need to be met for more than 3 ms before executing DAF playback; therefore, the minimum latency between the detection of target syllable and the delivery of auditory feedback is 3–4 ms. We set the sound amplitude slightly above the level where birds tend to truncate their song (∼65 dB at the center of the cage) to establish the saliency of DAF. Birds gradually begin to sing motifs of normal duration and tempo even with DAF, and our analysis is limited to the songs without truncation and delayed syllables except for <xref ref-type="fig" rid="fig6">Figure 6</xref>. Song truncation was detected by the absence of expected syllables following DAF onset. Delayed syllables were detected by referencing the expected distribution of the syllable onset timings in catch conditions with the probability p(τ | catch), where τ is the syllable onset timings relative to DAF. By randomly selecting the hit trials with the probability of p(τ | catch)/p(τ | hit), we made the syllable onset timing distributions in hit and catch conditions equivalent.</p><p>In separate experiments, we used sound amplitude and mean frequency as the contingency criteria (<xref ref-type="fig" rid="fig1s2">Figure 1—figure supplement 2</xref>, step 4). In this case, the bird can escape from DAF by slowly changing its syllable structure toward a certain direction of spectral features, such as higher or lower frequency. These experiments confirmed that our perturbation protocol can shift spectral features of target syllables when DAF was applied in a contingent manner as previously reported (<xref ref-type="bibr" rid="bib46">Tumer and Brainard, 2007</xref>; <xref ref-type="bibr" rid="bib2">Andalman and Fee, 2009</xref>) (<xref ref-type="fig" rid="fig1s3">Figure 1—figure supplement 3</xref>).</p></sec><sec id="s4-5"><title>Analysis of bird song for intracellular recordings</title><p>To precisely align intracellular recording data to the singing data, we first analyzed the microphone data to identify distinct syllables. The mean and variance of spectral features (sound amplitude, entropy, FM, mean frequency, pitch, goodness of pitch) during each sound event were saved in a MySQL database (<ext-link ext-link-type="uri" xlink:href="http://www.mysql.com">http://www.mysql.com</ext-link>). We classified them into mutually exclusive labels, such as introductory notes, syllables, and cage noise, using an SVM-based classification algorithm (<xref ref-type="bibr" rid="bib7">Chang and Lin, 2001</xref>), using a Gaussian kernel with soft margins implemented in a Matlab-based GUI interface written by KH. Manually classified data were used to train the SVM algorithm until classification accuracy exceeds 98% correctness, estimated by the cross-validation method.</p></sec><sec id="s4-6"><title>Analysis of intracellular data</title><p>To measure subthreshold DAF sensitivity, raw voltage measurements were first median-filtered to eliminate spikes (5 ms window). The filtered membrane potential data were then grouped into similar voltage ranges to ensure that average membrane potential did not differ between hit and catch conditions. For each voltage group, neurons from which we obtained at least 14 motifs of associated membrane potential data in total were used to measure DAF sensitivity. The differences of the baseline-subtracted V<sub>m</sub> generated from catch and hit trials averaged over fast and slow time windows (20–60 ms and 60–200 ms) were compared using <italic>t</italic>-tests. Here the baseline-subtracted V<sub>m</sub> is defined as <inline-formula><mml:math id="inf1"><mml:mrow><mml:msubsup><mml:mi>V</mml:mi><mml:mi>m</mml:mi><mml:mi>i</mml:mi></mml:msubsup><mml:mrow><mml:mrow><mml:mo>(</mml:mo><mml:mrow><mml:mi>t</mml:mi><mml:mtext>|S</mml:mtext></mml:mrow><mml:mo>)</mml:mo></mml:mrow></mml:mrow><mml:mo>−</mml:mo><mml:mi>b</mml:mi><mml:mi>a</mml:mi><mml:mi>s</mml:mi><mml:mi>e</mml:mi><mml:mi>l</mml:mi><mml:mi>i</mml:mi><mml:mi>n</mml:mi><mml:msup><mml:mi>e</mml:mi><mml:mi>i</mml:mi></mml:msup></mml:mrow></mml:math></inline-formula> where <inline-formula><mml:math id="inf2"><mml:mrow><mml:msubsup><mml:mi>V</mml:mi><mml:mi>m</mml:mi><mml:mi>i</mml:mi></mml:msubsup><mml:mrow><mml:mrow><mml:mo>(</mml:mo><mml:mrow><mml:mi>t</mml:mi><mml:mtext>|S</mml:mtext></mml:mrow><mml:mo>)</mml:mo></mml:mrow></mml:mrow></mml:mrow></mml:math></inline-formula> is the median filtered voltage trace at time <italic>t</italic> in <italic>i</italic>–th trial in condition S = {‘hit’, ’catch’}. The baseline of <italic>i</italic>–th trial was calculated as the mean of <inline-formula><mml:math id="inf3"><mml:mrow><mml:msubsup><mml:mi>V</mml:mi><mml:mi>m</mml:mi><mml:mi>i</mml:mi></mml:msubsup><mml:mrow><mml:mrow><mml:mo>(</mml:mo><mml:mrow><mml:mi>t</mml:mi><mml:mtext>|S</mml:mtext></mml:mrow><mml:mo>)</mml:mo></mml:mrow></mml:mrow></mml:mrow></mml:math></inline-formula> during −200 to 0 ms relative to DAF onset. The difference of the mean of baseline-subtracted V<sub>m</sub> (ΔV<sub>m</sub>) are defined as follows:<disp-formula id="equ1"><mml:math id="m1"><mml:mrow><mml:mi>Δ</mml:mi><mml:msub><mml:mi>V</mml:mi><mml:mi>m</mml:mi></mml:msub><mml:mrow><mml:mrow><mml:mo>(</mml:mo><mml:mi>t</mml:mi><mml:mo>)</mml:mo></mml:mrow></mml:mrow><mml:mo>=</mml:mo><mml:mo> </mml:mo><mml:mfrac><mml:mn>1</mml:mn><mml:mi>M</mml:mi></mml:mfrac><mml:mstyle displaystyle="true"><mml:msubsup><mml:mo>∑</mml:mo><mml:mi>i</mml:mi><mml:mi>M</mml:mi></mml:msubsup><mml:mrow><mml:mrow><mml:mrow><mml:mo>(</mml:mo><mml:mrow><mml:msubsup><mml:mi>V</mml:mi><mml:mi>m</mml:mi><mml:mi>i</mml:mi></mml:msubsup><mml:mrow><mml:mrow><mml:mo>(</mml:mo><mml:mrow><mml:mi>t</mml:mi><mml:mtext>|</mml:mtext><mml:mi>h</mml:mi><mml:mi>i</mml:mi><mml:mi>t</mml:mi></mml:mrow><mml:mo>)</mml:mo></mml:mrow></mml:mrow><mml:mo>−</mml:mo><mml:mi>b</mml:mi><mml:mi>a</mml:mi><mml:mi>s</mml:mi><mml:mi>e</mml:mi><mml:mi>l</mml:mi><mml:mi>i</mml:mi><mml:mi>n</mml:mi><mml:msup><mml:mi>e</mml:mi><mml:mi>i</mml:mi></mml:msup></mml:mrow><mml:mo>)</mml:mo></mml:mrow></mml:mrow></mml:mrow></mml:mstyle><mml:mo>−</mml:mo><mml:mfrac><mml:mn>1</mml:mn><mml:mi>N</mml:mi></mml:mfrac><mml:mstyle displaystyle="true"><mml:msubsup><mml:mo>∑</mml:mo><mml:mi>j</mml:mi><mml:mi>N</mml:mi></mml:msubsup><mml:mrow><mml:mrow><mml:mrow><mml:mo>(</mml:mo><mml:mrow><mml:msubsup><mml:mi>V</mml:mi><mml:mi>m</mml:mi><mml:mi>j</mml:mi></mml:msubsup><mml:mrow><mml:mrow><mml:mo>(</mml:mo><mml:mrow><mml:mi>t</mml:mi><mml:mtext>|</mml:mtext><mml:mi>c</mml:mi><mml:mi>a</mml:mi><mml:mi>t</mml:mi><mml:mi>c</mml:mi><mml:mi>h</mml:mi></mml:mrow><mml:mo>)</mml:mo></mml:mrow></mml:mrow><mml:mo>−</mml:mo><mml:mi>b</mml:mi><mml:mi>a</mml:mi><mml:mi>s</mml:mi><mml:mi>e</mml:mi><mml:mi>l</mml:mi><mml:mi>i</mml:mi><mml:mi>n</mml:mi><mml:msup><mml:mi>e</mml:mi><mml:mi>j</mml:mi></mml:msup></mml:mrow><mml:mo>)</mml:mo></mml:mrow></mml:mrow></mml:mrow></mml:mstyle></mml:mrow></mml:math></disp-formula>.</p><p>To calculate the mean firing rate and mean membrane potential before and during singing, mean values of the median-filtered voltage were calculated from a 0.5-s period before song (−1 to −0.5 s before bout onset) and a 0.5 s period during singing (0–0.5 s after bout onset). Here, the onset of a song is defined as the first introductory note or syllable proceeded by at least 1 s of silence. For each cell, up to the first 50 bouts that occurred during the recording were used for calculating mean firing rate and resting potential. We excluded the period from −0.5 to 0 s before song onset from these calculations because all HVC neurons recorded here displayed sharp transitions in firing rate and/or membrane potential within this period. Spikes were detected as events where the membrane potential deviates >+30 mV from the high-pass filtered membrane potential (>400Hz).</p></sec><sec id="s4-7"><title>Synaptic onset analysis</title><p>Depolarizing post-synaptic potentials (dPSPs) were detected using a modified version of the algorithm previously reported (<xref ref-type="bibr" rid="bib3">Ankri et al., 1994</xref>). Briefly, the onsets of dPSPs were defined as the time points where dV/dt exceeds a threshold defined as 0.8–1.5 times higher than its standard deviation. The threshold was set relatively high; therefore, our analysis is limited to sharp-rising dPSPs. Other parameter values were set to dt = 4 ms and refractoriness = 5 ms. Because zebra finch songs are highly stereotyped, the onset timing of each dPSP is highly stereotyped across bouts, and dPSP onsets naturally form clusters in dPSP raster plot (<xref ref-type="fig" rid="fig2">Figure 2C</xref>). To identify each dPSP onset cluster, we used a <italic>k</italic>-means clustering algorithm to identify dPSP onsets. The number of clusters is given by inspecting the dPSP histogram and raster plots. The time range for this analysis is 20–200 ms after DAF onset. We did not include cells with overlapping dPSPs because the identification of the corresponding dPSPs across hit and catch conditions is difficult.</p></sec><sec id="s4-8"><title>Measurement of auditory responses</title><p>To quantify the auditory responses of the intracellularly recorded neurons, we first measured the subthreshold (median-filtered voltage trace relative to baseline) and suprathreshold (firing rate) activity during and before playback of the bird’s own song (BOS), a stimulus known to evoke robust and selective responses in HVC neurons (stimulus duration was 1–2.4 s, played to birds at random intervals of 15-60 s). We analyzed data from neurons with 7 or more trials and the significance of each neuron’s response was evaluated by either a statistical test for differences in median value (Mann–Whitney <italic>U</italic>-test) or variance (Ansari–Bradley test). The Ansari–Bradley test is applicable only when there is no significant difference in median value; therefore, we first tested for a significant difference in the median response using the Mann–Whitney <italic>U</italic>-test and then applied the Ansari–Bradley test if there was no significant difference in median response. Cells that exhibited a significant difference in either the median or variance were defined as ‘responsive’. The threshold for statistical significance was set to p=0.025 for both tests. Response index of variable <italic>r</italic> (V<sub>m</sub> or firing rate) is defined as follows;<disp-formula id="equ2"><mml:math id="m2"><mml:mrow><mml:mtext>Response</mml:mtext><mml:mo> </mml:mo><mml:mtext>index</mml:mtext><mml:mo>=</mml:mo><mml:mo>|</mml:mo><mml:msub><mml:mi>r</mml:mi><mml:mrow><mml:mi>r</mml:mi><mml:mi>e</mml:mi><mml:mi>s</mml:mi><mml:mi>p</mml:mi><mml:mi>o</mml:mi><mml:mi>n</mml:mi><mml:mi>s</mml:mi><mml:mi>e</mml:mi></mml:mrow></mml:msub><mml:mo>−</mml:mo><mml:msub><mml:mi>r</mml:mi><mml:mrow><mml:mi>r</mml:mi><mml:mi>e</mml:mi><mml:mi>s</mml:mi><mml:mi>p</mml:mi><mml:mi>o</mml:mi><mml:mi>n</mml:mi><mml:mi>s</mml:mi><mml:mi>e</mml:mi></mml:mrow></mml:msub><mml:mo>|</mml:mo><mml:mo>/</mml:mo><mml:mo>|</mml:mo><mml:msub><mml:mi>r</mml:mi><mml:mrow><mml:mi>r</mml:mi><mml:mi>e</mml:mi><mml:mi>s</mml:mi><mml:mi>p</mml:mi><mml:mi>o</mml:mi><mml:mi>n</mml:mi><mml:mi>s</mml:mi><mml:mi>e</mml:mi></mml:mrow></mml:msub><mml:mo>+</mml:mo><mml:msub><mml:mi>r</mml:mi><mml:mrow><mml:mi>b</mml:mi><mml:mi>a</mml:mi><mml:mi>s</mml:mi><mml:mi>e</mml:mi></mml:mrow></mml:msub><mml:mo>|</mml:mo><mml:mo>,</mml:mo></mml:mrow></mml:math></disp-formula>where, <inline-formula><mml:math id="inf4"><mml:mrow><mml:msub><mml:mi>r</mml:mi><mml:mrow><mml:mi>r</mml:mi><mml:mi>e</mml:mi><mml:mi>s</mml:mi><mml:mi>p</mml:mi><mml:mi>o</mml:mi><mml:mi>n</mml:mi><mml:mi>s</mml:mi><mml:mi>e</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula> and <inline-formula><mml:math id="inf5"><mml:mrow><mml:msub><mml:mi>r</mml:mi><mml:mrow><mml:mi>b</mml:mi><mml:mi>a</mml:mi><mml:mi>s</mml:mi><mml:mi>e</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:math></inline-formula> are the mean or standard deviation of variable <italic>r</italic> measured during response and baseline periods, respectively.</p><p>The expected playback amplitude was subtracted from the sound amplitude recorded through the microphone to detect and eliminate trials when the bird vocalized during BOS playback. The sound amplitude was set to around 60 dB at the center of the cage during day-time recordings (light on) and ∼45 dB during night-time recordings (light off).</p><p>We classified our dataset based on both recording time and the bird’s behavioral state. For day-time recordings, the bird’s wakefulness was monitored by video, and we excluded data from periods when the bird closed its eyes and perched quietly for >1 min. For night-time recordings, the light was turned off and we excluded the trials when the bird was making noise. The light was turned on at a fixed time each day (9AM) and was turned off around 9 PM (±2 hr; day/night ratio between 14:10 and 10:14 hr).</p></sec><sec id="s4-9"><title>Deafening and quantification of song degradation</title><p>Male zebra finches (120–150 dph) were anesthetized with isoflurane and deafened by bilateral cochlear removal. Undirected song was recorded continuously starting at least 2 days before until at least 1 week after deafening. Wiener entropy and entropy variance (EV) of each syllable in a bird’s song were quantified using Sound Analysis Pro (<xref ref-type="bibr" rid="bib43a">Tchernichovski et al., 2000</xref>). 30 examples of each syllable were measured on each day, and values from two pre-deafening days were pooled to obtain a baseline distribution for each syllable. The onset of song degradation for each bird was defined as the day on which the distribution of values for either the entropy or EV of any syllable differed significantly from the baseline distribution and remained significantly different on all subsequent days (one-way ANOVA).</p></sec><sec id="s4-10"><title>Fluorescent labeling of HVC neurons and in vivo two-photon imaging</title><p>As previously described (<xref ref-type="bibr" rid="bib45">Tschida and Mooney, 2012</xref>), GFP-lentivirus (∼1 µL; eGFP expressed under the control of the Rous Sarcoma Virus LTR [FRGW]) was injected into HVC and Fast Blue or Alexa-Fluor 594 conjugated dextran amine were injected into Area X and RA (64-160 nL) in isoflurane-anesthetized male finches 4 days to 2 weeks prior to imaging. The birds were placed on a reverse day–night cycle 1 week before the first imaging session to minimize effects of imaging on their daytime behavior and were imaged longitudinally starting 1–2 nights prior to deafening. On the first night of imaging, birds were anesthetized with isoflurane and placed in a stereotaxic apparatus. A headpost was affixed to the skull using dental acrylic, and bilateral craniotomies (1–2 mm<sup>2</sup>) were made over HVC. The dura was excised, and a custom-cut coverslip (No. 1 thickness) was placed over the pial surface and sealed in with dental acrylic. The birds were placed on a custom stage under a Zeiss Laser Scanning Two-Photon Microscope 510. Only GFP-labeled neurons within a field of retrogradely labeled neurons were classified as HVC neurons and imaged. Dendritic segments of identified HVC neurons were imaged twice nightly at 2 hr intervals (Zeiss Laser Scanning Two-Photon Microscope 510, 1024 × 1024 pixels, 76 × 76 µm<sup>2</sup> image size, 3.2 µs/pixel, averaging 2 samples per pixel with 1 µm z-steps, using a 40x/0.8NA Zeiss IR-Archoplan immersion objective).</p></sec><sec id="s4-11"><title>Image analysis</title><p>Three-dimensional image stacks were smoothed using a Gaussian filter (ImageJ); brightness and contrast adjustments were not made for data analysis, although images were contrast-enhanced for figure presentation. Dendritic segments to be analyzed were selected and identified in image stacks collected either 2 hr or 24 hr apart. Spine size (measured across nights, 24 hr interval) was measured as the integrated optical density of each spine head, background-subtracted (using the optical density of neuropil next to the dendrite) and normalized to the mean brightness of the adjacent dendritic shaft. Spine size index was calculated as the ratio of spine size values measured during imaging sessions separated by 24 hr (time 24 size/time 0 size). Spine stability was calculated as the percentage of spines maintained each night (2 hr interval). Data from LMAN intact, deafened birds and LMAN intact, hearing birds (previously described in <xref ref-type="bibr" rid="bib45">Tschida and Mooney, 2012</xref>) are reported as raw values (i.e., post-deafening measurements of spine size index and spine stability were not normalized to baseline, pre-deafening measurements for data presentation).</p></sec><sec id="s4-12"><title>LMAN lesions</title><p>The boundaries of LMAN were mapped out using extracellular recordings, and LMAN was lesioned bilaterally by 7% ibotenic acid injection at least 4 days prior to deafening or imaging. Nissl-stained tissue from 2 LMAN intact control birds was used to estimate the average volume of LMAN, and the volume of LMAN remaining in LMAN lesion birds was divided by this value to estimate lesion extent. Additionally, injections of retrograde tracer were made into RA 4–5 days prior to sacrificing the animals to retrogradely label any remaining RA-projecting neurons in LMAN and to aid in the identification of remaining, non-lesioned portions of LMAN in Nissl-stained tissue sections.</p></sec><sec id="s4-13"><title>Source code</title><p>MATLAB and xPCtarget based data acquisition program for realtime detection of target syllable and playback of sound. This program also contains cell search mode and antidromic cell type identification mode used to conduct intracellular recording in freely behaving birds. All the codes are uploaded in GitHub (<ext-link ext-link-type="uri" xlink:href="https://github.com/hamaguchikosuke/RealSong_control">https://github.com/hamaguchikosuke/RealSong_control</ext-link>).</p></sec></sec></body><back><ack id="ack"><title>Acknowledgements</title><p>The authors thank David Schneider for reading earlier versions of the manuscript and Jennifer Baltzegar and Michael Booze for technical assistance.</p></ack><sec sec-type="additional-information"><title>Additional information</title><fn-group content-type="competing-interest"><title>Competing interests</title><fn fn-type="conflict" id="conf1"><p>The authors declare that no competing interests exist.</p></fn></fn-group><fn-group content-type="author-contribution"><title>Author contributions</title><fn fn-type="con" id="con1"><p>KH, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article</p></fn><fn fn-type="con" id="con2"><p>KAT, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article</p></fn><fn fn-type="con" id="con3"><p>IY, Contributed unpublished essential data or reagents</p></fn><fn fn-type="con" id="con4"><p>BRD, Contributed unpublished essential data or reagents</p></fn><fn fn-type="con" id="con5"><p>RM, Conception and design, Drafting or revising the article</p></fn></fn-group><fn-group content-type="ethics-information"><title>Ethics</title><fn fn-type="other"><p>Animal experimentation: All experiments were carried out in accordance with a protocol (A292-11-11) approved by Duke University Institutional Animal Care and Use Committee.</p></fn></fn-group></sec><ref-list><title>References</title><ref id="bib1"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Ali</surname><given-names>F</given-names></name><name><surname>Otchy</surname><given-names>TM</given-names></name><name><surname>Pehlevan</surname><given-names>C</given-names></name><name><surname>Fantana</surname><given-names>AL</given-names></name><name><surname>Burak</surname><given-names>Y</given-names></name><name><surname>Olveczky</surname><given-names>BP</given-names></name></person-group><year>2013</year><article-title>The Basal Ganglia is necessary for learning spectral, but not temporal, features of birdsong</article-title><source>Neuron</source><volume>80</volume><fpage>494</fpage><lpage>506</lpage><pub-id pub-id-type="doi">10.1016/j.neuron.2013.07.049</pub-id></element-citation></ref><ref id="bib2"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Andalman</surname><given-names>AS</given-names></name><name><surname>Fee</surname><given-names>MS</given-names></name></person-group><year>2009</year><article-title>A basal ganglia-forebrain circuit in the songbird biases motor output to avoid vocal errors</article-title><source>Proceedings of the National Academy of Sciences of the United States of America</source><volume>106</volume><fpage>12518</fpage><lpage>12523</lpage><pub-id pub-id-type="doi">10.1073/pnas.0903214106</pub-id></element-citation></ref><ref id="bib3"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Ankri</surname><given-names>N</given-names></name><name><surname>Legendre</surname><given-names>P</given-names></name><name><surname>Faber</surname><given-names>DS</given-names></name><name><surname>Korn</surname><given-names>H</given-names></name></person-group><year>1994</year><article-title>Automatic detection of spontaneous synaptic responses in central neurons</article-title><source>Journal of Neuroscience Methods</source><volume>52</volume><fpage>87</fpage><lpage>100</lpage><pub-id pub-id-type="doi">10.1016/0165-0270(94)90060-4</pub-id></element-citation></ref><ref id="bib4"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Brainard</surname><given-names>MS</given-names></name><name><surname>Doupe</surname><given-names>AJ</given-names></name></person-group><year>2000</year><article-title>Interruption of a basal ganglia-forebrain circuit prevents plasticity of learned vocalizations</article-title><source>Nature</source><volume>404</volume><fpage>762</fpage><lpage>766</lpage><pub-id pub-id-type="doi">10.1038/35008083</pub-id></element-citation></ref><ref id="bib5"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Cardin</surname><given-names>JA</given-names></name><name><surname>Schmidt</surname><given-names>MF</given-names></name></person-group><year>2003</year><article-title>Song system auditory responses are stable and highly tuned during sedation, rapidly modulated and unselective during wakefulness, and suppressed by arousal</article-title><source>Journal of Neurophysiology</source><volume>90</volume><fpage>2884</fpage><lpage>2899</lpage><pub-id pub-id-type="doi">10.1152/jn.00391.2003</pub-id></element-citation></ref><ref id="bib6"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Cardin</surname><given-names>JA</given-names></name><name><surname>Schmidt</surname><given-names>MF</given-names></name></person-group><year>2004</year><article-title>Noradrenergic inputs mediate state dependence of auditory responses in the avian song system</article-title><source>Journal of Neuroscience</source><volume>24</volume><fpage>7745</fpage><lpage>7753</lpage><pub-id pub-id-type="doi">10.1523/JNEUROSCI.1951-04.2004</pub-id></element-citation></ref><ref id="bib7"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Chang</surname><given-names>C-C</given-names></name><name><surname>Lin</surname><given-names>C-J</given-names></name></person-group><year>2001</year><article-title>LIBSVM: a library for support vector machines</article-title><source>ACM Transactions on Intelligent Systems and Technology</source><volume>2</volume><fpage>27:1</fpage><lpage>27:27</lpage><pub-id pub-id-type="doi">10.1109/72.963765</pub-id></element-citation></ref><ref id="bib8"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Charlesworth</surname><given-names>JD</given-names></name><name><surname>Warren</surname><given-names>TL</given-names></name><name><surname>Brainard</surname><given-names>MS</given-names></name></person-group><year>2012</year><article-title>Covert skill learning in a cortical-basal ganglia circuit</article-title><source>Nature</source><volume>486</volume><fpage>251</fpage><lpage>255</lpage><pub-id pub-id-type="doi">10.1038/nature11078</pub-id></element-citation></ref><ref id="bib9"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Coleman</surname><given-names>MJ</given-names></name><name><surname>Roy</surname><given-names>A</given-names></name><name><surname>Wild</surname><given-names>JM</given-names></name><name><surname>Mooney</surname><given-names>R</given-names></name></person-group><year>2007</year><article-title>Thalamic gating of auditory responses in telencephalic song control Nuclei</article-title><source>Journal of Neuroscience</source><volume>27</volume><fpage>10024</fpage><lpage>10036</lpage><pub-id pub-id-type="doi">10.1523/JNEUROSCI.2215-07.2007</pub-id></element-citation></ref><ref id="bib10"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Doupe</surname><given-names>AJ</given-names></name><name><surname>Perkel</surname><given-names>DJ</given-names></name><name><surname>Reiner</surname><given-names>A</given-names></name><name><surname>Stern</surname><given-names>EA</given-names></name></person-group><year>2005</year><article-title>Birdbrains could teach basal ganglia research a new song</article-title><source>Trends in Neurosciences</source><volume>28</volume><fpage>353</fpage><lpage>363</lpage><pub-id pub-id-type="doi">10.1016/j.tins.2005.05.005</pub-id></element-citation></ref><ref id="bib11"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Farries</surname><given-names>MA</given-names></name><name><surname>Perkel</surname><given-names>DJ</given-names></name></person-group><year>2002</year><article-title>A telencephalic nucleus essential for song learning contains neurons with physiological characteristics of both striatum and globus pallidus</article-title><source>The Journal of Neuroscience: the Official Journal of the Society for Neuroscience</source><volume>22</volume><fpage>3776</fpage><lpage>3787</lpage></element-citation></ref><ref id="bib12"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Fortune</surname><given-names>ES</given-names></name><name><surname>Margoliash</surname><given-names>D</given-names></name></person-group><year>1995</year><article-title>Parallel pathways and convergence onto HVc and adjacent neostriatum of adult male zebra finches (<italic>Taeniopygia guttata</italic>)</article-title><source>The Journal of Comparative Neurology</source><volume>360</volume><fpage>413</fpage><lpage>441</lpage><pub-id pub-id-type="doi">10.1002/cne.903600305</pub-id></element-citation></ref><ref id="bib13"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Froemke</surname><given-names>RC</given-names></name><name><surname>Merzenich</surname><given-names>MM</given-names></name><name><surname>Schreiner</surname><given-names>CE</given-names></name></person-group><year>2007</year><article-title>A synaptic memory trace for cortical receptive field plasticity</article-title><source>Nature</source><volume>450</volume><fpage>425</fpage><lpage>429</lpage><pub-id pub-id-type="doi">10.1038/nature06289</pub-id></element-citation></ref><ref id="bib14"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Fujimoto</surname><given-names>H</given-names></name><name><surname>Hasegawa</surname><given-names>T</given-names></name><name><surname>Watanabe</surname><given-names>D</given-names></name></person-group><year>2011</year><article-title>Neural coding of syntactic structure in learned vocalizations in the songbird</article-title><source>The Journal of Neuroscience: the Official Journal of the Society for Neuroscience</source><volume>31</volume><fpage>10023</fpage><lpage>10033</lpage><pub-id pub-id-type="doi">10.1523/JNEUROSCI.1606-11.2011</pub-id></element-citation></ref><ref id="bib15"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Hahnloser</surname><given-names>RHR</given-names></name><name><surname>Kozhevnikov</surname><given-names>AA</given-names></name><name><surname>Fee</surname><given-names>MS</given-names></name></person-group><year>2002</year><article-title>An ultra-sparse code underliesthe generation of neural sequences in a songbird</article-title><source>Nature</source><volume>419</volume><fpage>65</fpage><lpage>70</lpage><pub-id pub-id-type="doi">10.1038/nature00974</pub-id></element-citation></ref><ref id="bib16"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Hamaguchi</surname><given-names>K</given-names></name><name><surname>Mooney</surname><given-names>R</given-names></name></person-group><year>2012</year><article-title>Recurrent interactions between the input and output of a songbird cortico-basal ganglia pathway are implicated in vocal sequence variability</article-title><source>The Journal of Neuroscience: the Official Journal of the Society for Neuroscience</source><volume>32</volume><fpage>11671</fpage><lpage>11687</lpage><pub-id pub-id-type="doi">10.1523/JNEUROSCI.1666-12.2012</pub-id></element-citation></ref><ref id="bib17"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Hanuschkin</surname><given-names>A</given-names></name><name><surname>Ganguli</surname><given-names>S</given-names></name><name><surname>Hahnloser</surname><given-names>RH</given-names></name></person-group><year>2013</year><article-title>A Hebbian learning rule gives rise to mirror neurons and links them to control theoretic inverse models</article-title><source>Frontiers in Neural Circuits</source><volume>7</volume><fpage>106</fpage><pub-id pub-id-type="doi">10.3389/fncir.2013.00106</pub-id></element-citation></ref><ref id="bib18"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Hessler</surname><given-names>NA</given-names></name><name><surname>Doupe</surname><given-names>AJ</given-names></name></person-group><year>1999</year><article-title>Singing-related neural activity in a dorsal forebrain-basal ganglia circuit of adult zebra finches</article-title><source>The Journal of Neuroscience: the Official Journal of the Society for Neuroscience</source><volume>19</volume><fpage>10461</fpage><lpage>10481</lpage></element-citation></ref><ref id="bib19"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Horita</surname><given-names>H</given-names></name><name><surname>Wada</surname><given-names>K</given-names></name><name><surname>Jarvis</surname><given-names>ED</given-names></name></person-group><year>2008</year><article-title>Early onset of deafening-induced song deterioration and differential requirements of the pallial-basal ganglia vocal pathway</article-title><source>European Journal of Neuroscience</source><volume>28</volume><fpage>2519</fpage><lpage>2532</lpage><pub-id pub-id-type="doi">10.1111/j.1460-9568.2008.06535.x</pub-id></element-citation></ref><ref id="bib20"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Keck</surname><given-names>T</given-names></name><name><surname>Scheuss</surname><given-names>V</given-names></name><name><surname>Jacobsen</surname><given-names>RI</given-names></name><name><surname>Wierenga</surname><given-names>CJ</given-names></name><name><surname>Eysel</surname><given-names>UT</given-names></name><name><surname>Bonhoeffer</surname><given-names>T</given-names></name><name><surname>Hubener</surname><given-names>M</given-names></name></person-group><year>2011</year><article-title>Loss of sensory input causes rapid structural changes of inhibitory neurons in adult mouse visual cortex</article-title><source>Neuron</source><volume>71</volume><fpage>869</fpage><lpage>882</lpage><pub-id pub-id-type="doi">10.1016/j.neuron.2011.06.034</pub-id></element-citation></ref><ref id="bib21"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Keller</surname><given-names>GB</given-names></name><name><surname>Hahnloser</surname><given-names>RHR</given-names></name></person-group><year>2009</year><article-title>Neural processing of auditory feedback during vocal practice in a songbird</article-title><source>Nature</source><volume>456</volume><fpage>187</fpage><lpage>190</lpage><pub-id pub-id-type="doi">10.1038/nature07467</pub-id></element-citation></ref><ref id="bib22"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Konishi</surname><given-names>M</given-names></name></person-group><year>1965</year><article-title>The role of auditory feedback in the control of vocalization in the white-crowned sparrow</article-title><source>Zeitschrift fuer Tierpsychologie</source><volume>22</volume><fpage>770</fpage><lpage>783</lpage></element-citation></ref><ref id="bib23"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Kozhevnikov</surname><given-names>AA</given-names></name><name><surname>Fee</surname><given-names>MS</given-names></name></person-group><year>2007</year><article-title>Singing-related activity of identified HVC neurons in the zebra finch</article-title><source>Journal of Neurophysiology</source><volume>97</volume><fpage>4271</fpage><lpage>4283</lpage><pub-id pub-id-type="doi">10.1152/jn.00952.2006</pub-id></element-citation></ref><ref id="bib24"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Lee</surname><given-names>AK</given-names></name><name><surname>Manns</surname><given-names>ID</given-names></name><name><surname>Sakmann</surname><given-names>B</given-names></name><name><surname>Brecht</surname><given-names>M</given-names></name></person-group><year>2006</year><article-title>Whole-cell recordings in freely moving rats</article-title><source>Neuron</source><volume>51</volume><fpage>399</fpage><lpage>407</lpage><pub-id pub-id-type="doi">10.1016/j.neuron.2006.07.004</pub-id></element-citation></ref><ref id="bib25"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Lei</surname><given-names>H</given-names></name><name><surname>Mooney</surname><given-names>R</given-names></name></person-group><year>2010</year><article-title>Manipulation of a central auditory representation shapes learned vocal output</article-title><source>Neuron</source><volume>65</volume><fpage>122</fpage><lpage>134</lpage><pub-id pub-id-type="doi">10.1016/j.neuron.2009.12.008</pub-id></element-citation></ref><ref id="bib26"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Leonardo</surname><given-names>A</given-names></name></person-group><year>2004</year><article-title>Experimental test of the birdsong error-correction model</article-title><source>Proceedings of the Montana Academy of Sciences</source><volume>101</volume><fpage>16935</fpage><lpage>16940</lpage><pub-id pub-id-type="doi">10.1073/pnas.0407870101</pub-id></element-citation></ref><ref id="bib27"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Leonardo</surname><given-names>A</given-names></name><name><surname>Konishi</surname><given-names>M</given-names></name></person-group><year>1999</year><article-title>Decrystallization of adult birdsong by perturbation of auditory feedback</article-title><source>Nature</source><volume>399</volume><fpage>466</fpage><lpage>470</lpage><pub-id pub-id-type="doi">10.1038/20933</pub-id></element-citation></ref><ref id="bib28"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Lewis</surname><given-names>JW</given-names></name><name><surname>Ryan</surname><given-names>SM</given-names></name><name><surname>Arnold</surname><given-names>AP</given-names></name><name><surname>Butcher</surname><given-names>LL</given-names></name></person-group><year>1981</year><article-title>Evidence for a catecholaminergic projection to area X in the zebra finch</article-title><source>The Journal of Comparative Neurology</source><volume>196</volume><fpage>347</fpage><lpage>354</lpage><pub-id pub-id-type="doi">10.1002/cne.901960212</pub-id></element-citation></ref><ref id="bib29"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Long</surname><given-names>MA</given-names></name><name><surname>Jin</surname><given-names>DZ</given-names></name><name><surname>Fee</surname><given-names>MS</given-names></name></person-group><year>2010</year><article-title>Support for a synaptic chain model of neuronal sequence generation</article-title><source>Nature</source><volume>468</volume><fpage>394</fpage><lpage>399</lpage><pub-id pub-id-type="doi">10.1038/nature09514</pub-id></element-citation></ref><ref id="bib30"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Mitra</surname><given-names>PP</given-names></name><name><surname>Pesaran</surname><given-names>B</given-names></name></person-group><year>1999</year><article-title>Analysis of dynamic brain imaging data</article-title><source>Biophysical Journal</source><volume>76</volume><fpage>691</fpage><lpage>708</lpage><pub-id pub-id-type="doi">10.1016/S0006-3495(99)77236-X</pub-id></element-citation></ref><ref id="bib31"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Mooney</surname><given-names>R</given-names></name></person-group><year>1992</year><article-title>Synaptic basis for developmental plasticity in a birdsong nucleus</article-title><source>The Journal of Neuroscience: the Official Journal of the Society for Neuroscience</source><volume>12</volume><fpage>2464</fpage><lpage>2477</lpage></element-citation></ref><ref id="bib32"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Mooney</surname><given-names>R</given-names></name></person-group><year>2000</year><article-title>Different subthreshold mechanisms underlie song selectivity in identified HVc neurons of the zebra finch</article-title><source>Journal of Neuroscience</source><volume>20</volume><fpage>5420</fpage><lpage>5436</lpage></element-citation></ref><ref id="bib33"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Nick</surname><given-names>TA</given-names></name><name><surname>Konishi</surname><given-names>M</given-names></name></person-group><year>2005</year><article-title>Neural song preference during vocal learning in the zebra finch depends on age and state</article-title><source>Journal of Neurobiology</source><volume>62</volume><fpage>231</fpage><lpage>242</lpage><pub-id pub-id-type="doi">10.1002/Neu.20087</pub-id></element-citation></ref><ref id="bib34"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Nottebohm</surname><given-names>F</given-names></name><name><surname>Kelley</surname><given-names>DB</given-names></name><name><surname>Paton</surname><given-names>JA</given-names></name></person-group><year>1982</year><article-title>Connections of vocal control nuclei in the Canary telencephalon</article-title><source>The Journal of Comparative Neurology</source><volume>207</volume><fpage>344</fpage><lpage>357</lpage><pub-id pub-id-type="doi">10.1002/cne.902070406</pub-id></element-citation></ref><ref id="bib35"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Nottebohm</surname><given-names>F</given-names></name><name><surname>Stokes</surname><given-names>TM</given-names></name><name><surname>Leonardo</surname><given-names>CM</given-names></name></person-group><year>1976</year><article-title>Central control of song in the Canary, Serinus canarius</article-title><source>The Journal of Comparative Neurology</source><volume>165</volume><fpage>457</fpage><lpage>486</lpage><pub-id pub-id-type="doi">10.1002/cne.901650405</pub-id></element-citation></ref><ref id="bib36"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Olveczky</surname><given-names>BP</given-names></name><name><surname>Andalman</surname><given-names>AS</given-names></name><name><surname>Fee</surname><given-names>MS</given-names></name></person-group><year>2005</year><article-title>Vocal experimentation in the juvenile songbird requires a basal ganglia circuit</article-title><source>PLOS Biology</source><volume>3</volume><fpage>e153</fpage><pub-id pub-id-type="doi">10.1371/journal.pbio.0030153</pub-id></element-citation></ref><ref id="bib37"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Prather</surname><given-names>JF</given-names></name><name><surname>Peters</surname><given-names>S</given-names></name><name><surname>Nowicki</surname><given-names>S</given-names></name><name><surname>Mooney</surname><given-names>R</given-names></name></person-group><year>2008</year><article-title>Precise auditory-vocal mirroring in neurons for learned vocal communication</article-title><source>Nature</source><volume>451</volume><fpage>305</fpage><lpage>310</lpage><pub-id pub-id-type="doi">10.1038/nature06492</pub-id></element-citation></ref><ref id="bib38"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Price</surname><given-names>P</given-names></name></person-group><year>1979</year><article-title>Developmental determinants of structure in zebra finch song</article-title><source>Journal of Comparative and Physiological Psychology</source><volume>93</volume><fpage>260</fpage><lpage>277</lpage><pub-id pub-id-type="doi">10.1037/h0077553</pub-id></element-citation></ref><ref id="bib39"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Rauske</surname><given-names>PL</given-names></name><name><surname>Shea</surname><given-names>SD</given-names></name><name><surname>Margoliash</surname><given-names>D</given-names></name></person-group><year>2003</year><article-title>State and neuronal class-dependent reconfiguration in the avian song system</article-title><source>Journal of Neurophysiology</source><volume>89</volume><fpage>1688</fpage><lpage>1701</lpage><pub-id pub-id-type="doi">10.1152/jn.00655.2002</pub-id></element-citation></ref><ref id="bib40"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Roberts</surname><given-names>TF</given-names></name><name><surname>Tschida</surname><given-names>KA</given-names></name><name><surname>Klein</surname><given-names>ME</given-names></name><name><surname>Mooney</surname><given-names>R</given-names></name></person-group><year>2010</year><article-title>Rapid spine stabilization and synaptic enhancement at the onset of behavioural learning</article-title><source>Nature</source><volume>463</volume><fpage>948</fpage><lpage>952</lpage><pub-id pub-id-type="doi">10.1038/nature08759</pub-id></element-citation></ref><ref id="bib41"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Roy</surname><given-names>A</given-names></name><name><surname>Mooney</surname><given-names>R</given-names></name></person-group><year>2009</year><article-title>Song Decrystallization in adult zebra finches does not require the song nucleus NIf</article-title><source>Journal of Neurophysiology</source><volume>102</volume><fpage>979</fpage><lpage>991</lpage><pub-id pub-id-type="doi">10.1152/jn.00293.2009</pub-id></element-citation></ref><ref id="bib42"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Sakata</surname><given-names>JT</given-names></name><name><surname>Brainard</surname><given-names>MS</given-names></name></person-group><year>2008</year><article-title>Online contributions of auditory feedback to neural activity in avian song control circuitry</article-title><source>Journal of Neuroscience</source><volume>28</volume><fpage>11378</fpage><lpage>11390</lpage><pub-id pub-id-type="doi">10.1523/JNEUROSCI.3254-08.2008</pub-id></element-citation></ref><ref id="bib43"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Shea</surname><given-names>SD</given-names></name><name><surname>Margoliash</surname><given-names>D</given-names></name></person-group><year>2003</year><article-title>Basal forebrain cholinergic modulation of auditory activity in the zebra finch song system</article-title><source>Neuron</source><volume>40</volume><fpage>1213</fpage><lpage>1226</lpage><pub-id pub-id-type="doi">10.1016/S0896-6273(03)00723-2</pub-id></element-citation></ref><ref id="bib43a"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Tchernichovski</surname><given-names>O</given-names></name><name><surname>Nottebohm</surname><given-names>F</given-names></name><name><surname>Ho</surname><given-names>CE</given-names></name><name><surname>Pesaran</surname><given-names>B</given-names></name><name><surname>Mitra</surname><given-names>PP</given-names></name></person-group><year>2000</year><article-title>A procedure for an automated measurement of song similarity</article-title><source>Animal Behaviour</source><volume>59</volume><fpage>1167</fpage><lpage>1176</lpage><pub-id pub-id-type="doi">10.1006/anbe.1999.1416</pub-id></element-citation></ref><ref id="bib44"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Thomson</surname><given-names>DJ</given-names></name></person-group><year>1982</year><article-title>Spectrum estimation and harmonic-analysis</article-title><source>Proceedings of the IEEE</source><volume>70</volume><fpage>1055</fpage><lpage>1096</lpage><pub-id pub-id-type="doi">10.1109/PROC.1982.12433</pub-id></element-citation></ref><ref id="bib45"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Tschida</surname><given-names>KA</given-names></name><name><surname>Mooney</surname><given-names>R</given-names></name></person-group><year>2012</year><article-title>Deafening drives cell-type-specific changes to dendritic spines in a sensorimotor nucleus important to learned vocalizations</article-title><source>Neuron</source><volume>73</volume><fpage>1028</fpage><lpage>1039</lpage><pub-id pub-id-type="doi">10.1016/j.neuron.2011.12.038</pub-id></element-citation></ref><ref id="bib46"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Tumer</surname><given-names>EC</given-names></name><name><surname>Brainard</surname><given-names>MS</given-names></name></person-group><year>2007</year><article-title>Performance variability enables adaptive plasticity of ‘crystallized’ adult birdsong</article-title><source>Nature</source><volume>450</volume><fpage>1240</fpage><lpage>1244</lpage><pub-id pub-id-type="doi">10.1038/Nature06390</pub-id></element-citation></ref><ref id="bib47"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Vates</surname><given-names>GE</given-names></name><name><surname>Broome</surname><given-names>BM</given-names></name><name><surname>Mello</surname><given-names>CV</given-names></name><name><surname>Nottebohm</surname><given-names>F</given-names></name></person-group><year>1996</year><article-title>Auditory pathways of caudal telencephalon and their relation to the song system of adult male zebra finches</article-title><source>The Journal of Comparative Neurology</source><volume>366</volume><fpage>613</fpage><lpage>642</lpage><pub-id pub-id-type="doi">10.1002/(SICI)1096-9861(19960318)366:4<613::AID-CNE5>3.0.CO;2-7</pub-id></element-citation></ref><ref id="bib48"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Warren</surname><given-names>TL</given-names></name><name><surname>Tumer</surname><given-names>EC</given-names></name><name><surname>Charlesworth</surname><given-names>JD</given-names></name><name><surname>Brainard</surname><given-names>MS</given-names></name></person-group><year>2011</year><article-title>Mechanisms and time course of vocal learning and consolidation in the adult songbird</article-title><source>Journal of Neurophysiology</source><volume>106</volume><fpage>1806</fpage><lpage>1821</lpage><pub-id pub-id-type="doi">10.1152/jn.00311.2011</pub-id></element-citation></ref><ref id="bib49"><element-citation publication-type="journal"><person-group person-group-type="author"><name><surname>Wild</surname><given-names>JM</given-names></name><name><surname>Karten</surname><given-names>HJ</given-names></name><name><surname>Frost</surname><given-names>BJ</given-names></name></person-group><year>1993</year><article-title>Connections of the auditory forebrain in the pigeon (Columba livia)</article-title><source>The Journal of Comparative Neurology</source><volume>337</volume><fpage>32</fpage><lpage>62</lpage><pub-id pub-id-type="doi">10.1002/cne.903370103</pub-id></element-citation></ref></ref-list></back><sub-article article-type="article-commentary" id="SA1"><front-stub><article-id pub-id-type="doi">10.7554/eLife.01833.018</article-id><title-group><article-title>Decision letter</article-title></title-group><contrib-group content-type="section"><contrib contrib-type="editor"><name><surname>Calabrese</surname><given-names>Ronald L</given-names></name><role>Reviewing editor</role><aff><institution>Emory University</institution>, <country>United States</country></aff></contrib></contrib-group></front-stub><body><boxed-text><p>eLife posts the editorial decision letter and author response on a selection of the published articles (subject to the approval of the authors). An edited version of the letter sent to the authors after peer review is shown, indicating the substantive concerns or comments; minor concerns are not usually shown. Reviewers have the opportunity to discuss the decision before the letter is sent (see <ext-link ext-link-type="uri" xlink:href="http://elife.elifesciences.org/review-process">review process</ext-link>). Similarly, the author response typically shows only responses to the major concerns raised by the reviewers.</p></boxed-text><p>[Editors’ note: although it is not typical of the review process at <italic>eLife</italic>, in this case the editors decided to include the reviews in their entirety for the authors’ consideration as they prepared their revised submission.]</p><p>Thank you for sending your work entitled “Mechanisms underlying a synaptic trace of auditory feedback in a sensorimotor nucleus important to learned birdsong” for consideration at <italic>eLife</italic>. Your article has been favorably evaluated by a Senior editor and 3 reviewers, one of whom is a member of our Board of Reviewing Editors.</p><p>The following individuals responsible for the peer review of your submission have agreed to reveal their identity: Ronald L Calabrese, Reviewing editor; Marc Schmidt, peer reviewer.</p><p>The Reviewing editor and the other reviewers discussed their comments before we reached this decision, and the Reviewing editor has assembled the following comments to help you prepare a revised submission.</p><p>In this study, Hamaguchi et al. perform a careful electrophysiological analysis of subthreshold activity in HCV<sub>X</sub> neurons using sharp microelectrode recordings of the awake unrestrained songbird to assess the role of auditory feedback during singing in these neurons. Positioned as they are in the auditory/singing pathway these HCV<sub>X</sub> neuron seem a likely target for feedback during singing that leads to behavioral plasticity, but extracellular recordings have not shown an effect of acoustic feedback on spiking during singing: thus the motivation in this study to look for subthreshold effects of feedback. The results show convincingly that synaptic input to HVC<sub>X</sub> cells can be driven by sound input but not during singing corroborating the previous extracellular results. These experiments are supplemented by an ablation (AFP) experiment that shows that deafening induced plasticity of spines on HVC<sub>X</sub> neurons requires an anterior forebrain pathway. This structural study provides novel conclusions. This elegant analysis in freely singing birds will generate wide interest in the songbird and sensory motor integration communities.</p><p>All the reviewers thought the work was elegant and important but there were some minor concerns that should be addressed. The reviews are highly congruent and complementary, and they are provided in their entirety so that the authors can benefit from them directly. The authors should address reviewer concerns and implement all the minor edits. They should particularly address this minor concern:</p><p>“A remote possibility with DAF is that auditory responses in HVC<sub>X</sub> might only occur after a window of time. One could envisage, for example, that HVC only processes auditory feedback errors if these occur for a critical amount of time. In other words, auditory flow might be gated for small infrequent errors (this might prevent the motor system from changing to any random perturbation of auditory feedback during singing) but might become “un-gated” once the perturbations are sensed as occurring in a consistent manner (this “consistency” could potentially be monitored in the auditory forebrain and need not rely on the AFP or VTA). A prediction of this idea is that HVC<sub>X</sub> neurons would not show any changes in activity if (1) DAF is too infrequent or (2) when DAF is first introduced. It would be helpful for the authors to address this issue. This argument could easily be resolved if the authors are able to show that DAF-induced changes in song also occur in the birds they record from. Ideally, the most convincing argument would be to show a lack of change in HVC<sub>X</sub> neurons during DAF even while the birds undergo changes in song.”</p><p>Because such a demonstration would constitute a new study they should squarely address this issue in Discussion.</p><p>Another point that the authors should address is making the title more descriptive of what the manuscript actually presents.</p><p><italic>Reviewer</italic><italic>#1</italic>:</p><p>In this manuscript the authors perform a careful electrophysiological analysis of subthreshold activity in HCV<sub>X</sub> neurons using sharp electrode recordings of the awake unrestrained songbird during singing to assess the role of auditory feedback during singing in these neurons. Positioned as they are in the auditory/singing pathway these HCV<sub>X</sub> neuron seem a likely target for feedback during singing that leads to behavioral plasticity, but extracellular recordings have not shown acute feedback during singing: thus the motivation in this study to look for subthreshold feedback. The results show convincingly that synaptic input to HVC<sub>X</sub> cells can be driven by sound input but not during singing corroborating the previous extracellular results. These experiments are supplemented by an ablation experiment that shows that deafening induced plasticity of spines on HVC<sub>X</sub> neurons requires an anterior forebrain pathway. This structural study provides novel conclusions. This elegant analysis in freely singing birds will generate wide interest in the songbird and sensory motor integration communities.</p><p>Concerns:</p><p>The writing is succinct but the paper is not always easy to follow and could benefit from more explanation for the general reader. The work is carefully done with detailed analysis and the figures appropriately illustrate the data. Nevertheless the figures are very complicated and could benefit from more detailed legends and perhaps better labeling. A better general figure of the system than <xref ref-type="fig" rid="fig1">Figure 1A</xref> would help, including some of the detail of <xref ref-type="fig" rid="fig7">Figure 7</xref> up front. Essential information from the supplemental figures should be incorporated into the text according to <italic>eLife</italic> policy – I defer to the expert reviewers on what is essential.</p><p><italic>Reviewer</italic><italic>#2</italic>:</p><p>The authors convincingly demonstrate their two main points: lack of subthreshold responses to auditory feedback perturbations in HVC<sub>X</sub> neurons and a dependence of deafening-induced changes in spine morphology on an intact AFP. The work represents a very high level of technical proficiency and, except in rare cases, the presentation is very clear.</p><p>My primary concern centers on the impact of the paper’s findings. As the authors correctly point out, a number of prior studies have used extracellular recordings to show that the pattern of action potentials fired by HVC<sub>X</sub> neurons is insensitive to perturbations of auditory feedback. The first main finding of the present manuscript <italic>–</italic> that subthreshold activity is also not affected by feedback perturbations <italic>–</italic> therefore confirms current thinking in the field and appears to have limited interest except to specialists in the system.</p><p>The paper’s second main finding – that post-deafening changes in spine morphology in X-projecting HVC neurons depends on an intact AFP – has somewhat higher impact. Earlier work from the same lab (Tschida and Mooney, Neuron 2012) convincingly demonstrated that deafening-induced changes in spine morphology precede the degradation of song and that the changes in spine size predicted the magnitude of behavioral changes. These earlier results raise the question of whether the morphological changes reflect the sensory perturbation (deafening) or the beginning of AFP-dependent motor changes that are first detectable days later. The present results demonstrating that spine changes depend on an intact AFP show that such changes are not solely a function of altered auditory feedback, ruling out a simple model (<xref ref-type="fig" rid="fig7">Figure 7A</xref>) in which changes in spine morphology depend only on the removal of ascending auditory inputs. Although ruling out this model leaves several important questions unanswered (especially the mechanisms by which patterns of AFP activity could regulate spine changes in HVC and how LMAN’s indirect inputs to HVC might interact with ascending auditory signals), it nevertheless represents an advance in knowledge over prior studies.</p><p>In summary, I have no substantial concerns about how the research was performed or the validity of the results. My primary concern is that the first finding described above does not substantially change current thinking about the system. The second finding advances our understanding of sensorimotor plasticity by ruling out a simple model of auditory dependence and therefore will be of greater interest to readers.</p><p><italic>Reviewer</italic><italic>#3</italic>:</p><p>This is a very carefully executed experimental study investigating, in songbirds, how and where auditory feedback modifies vocal motor circuits. This is a question that lies at the core of how auditory feedback shapes vocal learning. Over the past several decades, a number of increasingly refined experiments have shown that auditory feedback is necessary for vocal learning and that perturbations of auditory feedback will cause, even the normally stereotyped song of adult zebra finches, to degrade. The time course over which this degradation occurs is not immediate. Depending on the experimental paradigm and the precision with which song is analyzed, changes in song acoustic structure can be observed within hours to days. Because lesions of a specialized basal ganglia circuit (known as the anterior forebrain pathway (AFP)) prevents song degradation following feedback perturbation (e.g., deafening by cochlear removal or distorted auditory feedback (DAF)), it has been hypothesized and shown that song degradation is an active process that can be prevented by interrupting this basal ganglia circuit, for example with lesions of its output nucleus, LMAN.</p><p>One of the primary assumptions has always been that auditory feedback information is processed by nucleus HVC. This nucleus receives direct input from the auditory forebrain, is necessary for song production and was thought to provide the only known auditory input to the basal ganglia circuit by way of neurons that project directly to the basal ganglia, known as HVC<sub>X</sub>-projection neurons. It had been assumed that these neurons would relay auditory feedback information to the basal ganglia. It was therefore surprising that these neurons did not shown any changes in firing pattern during DAF even though feedback distortion was sufficient to perturb song.</p><p>In the present study, the authors used intracellular electrodes (technically heroic experiments) to record from HVC<sub>X</sub>-projecting neurons during singing under both normal and DAF conditions. They provide compelling evidence that DAF does not alter (at all!) the highly stereotyped membrane potential fluctuations that are recorded during song production. This despite recent findings, by the same lab, that DAF modifies synaptic inputs onto these cells. While these results do not provide a direct answer as to where auditory feedback is first monitored in the vocal motor system, these experiments are extremely important in ruling out HVC as the site for such monitoring.</p><p>In this study the authors also show that changes in synaptic strength (as measured by spine dynamics) of HVC<sub>X</sub>-projection neurons following auditory feedback perturbation (here they use deafening) are dependent on LMAN. They use this finding to argue that auditory feedback dependent changes in HVC<sub>X</sub>-projecting neurons are dependent on LMAN and that HVC is therefore downstream of where the real-time sensors of auditory feedback. Because the authors have previously show that LMAN receives direct auditory input from the VTA, they suggest that this circuit might be a possible candidate for auditory feedback processing, but they do not show any evidence for this.</p><p>While the authors do not, in the end, show where auditory feedback processing occurs, this study is extremely important because it makes a strong case that it does not occur in HVC. The paper is well written and the experiments are performed with great care and the conclusions sounds.</p><p>Concerns:</p><p>A remote possibility with DAF is that auditory responses in HVC<sub>X</sub> might only occur after a window of time. One could envisage, for example, that HVC only processes auditory feedback errors if these occur for a critical amount of time. In other words, auditory flow might be gated for small infrequent errors (this might prevent the motor system from changing to any random perturbation of auditory feedback during singing) but might become “un-gated” once the perturbations are sensed as occurring in a consistent manner (this “consistency” could potentially be monitored in the auditory forebrain and need not rely on the AFP or VTA). A prediction of this idea is that HVC<sub>X</sub> neurons would not show any changes in activity if (1) DAF is too infrequent or (2) when DAF is first introduced. It would be helpful for the authors to address this issue. This argument could easily be resolved if the authors are able to show that DAF-induced changes in song also occur in the birds they record from. Ideally, the most convincing argument would be to show a lack of change in HVC<sub>X</sub> neurons during DAF even while the birds undergo changes in song.</p></body></sub-article><sub-article article-type="reply" id="SA2"><front-stub><article-id pub-id-type="doi">10.7554/eLife.01833.019</article-id><title-group><article-title>Author response</article-title></title-group></front-stub><body><p>We thank the three reviewers for their helpful comments, and we are grateful for their constructive criticism. We have done our best to follow their various suggestions, as detailed below. As instructed by the Reviewing editor, we have addressed Reviewer 3’s concern about “ungating” in the Discussion. We have also changed the title to render it more descriptive of what the manuscript actually presents.</p><p>Reviewer #1:</p><p><italic>The writing is succinct but the paper is not always easy to follow and could benefit from more explanation for the general reader</italic>.</p><p>We have done our best to edit the manuscript for clarity and accessibility, in part by providing more explanation for the general reader in the Introduction, Methods, Discussion, and figure legends.</p><p><italic>The work is carefully done with detailed analysis and the figures appropriately illustrate the data. Nevertheless the figures are very complicated and could benefit from more detailed legends and perhaps better labeling. A better general figure of the system than</italic><xref ref-type="fig" rid="fig1"><italic>Figure 1A</italic></xref><italic>would help, including some of the detail of</italic><xref ref-type="fig" rid="fig7"><italic>Figure 7</italic></xref><italic>up front</italic>.</p><p>We modified <xref ref-type="fig" rid="fig1">Figure 1A</xref> for clarity by including some of the details from <xref ref-type="fig" rid="fig7">Figure 7</xref>. We also included more complete labeling of song system structures. We have tried where possible to provide more detailed legends, and to further simplify or expand the figures.</p><p><italic>Essential information from the supplemental figures should be incorporated into the text according to eLife policy – I defer to the expert reviewers on what is essential</italic>.</p><p>We have incorporated information from the supplemental figures into the text to conform to <italic>eLife</italic> policy.</p><p>Reviewer #2:</p><p><italic>My primary concern centers on the impact of the paper’s findings. As the authors correctly point out, a number of prior studies have used extracellular recordings to show that the pattern of action potentials fired by HVC</italic><sub><italic>X</italic></sub><italic> neurons is insensitive to perturbations of auditory feedback. The first main finding of the present manuscript – that subthreshold activity is also not affected by feedback perturbations – therefore confirms current thinking in the field and appears to have limited interest except to specialists in the system</italic>.</p><p><italic>The paper’s second main finding – that post-deafening changes in spine morphology in X-projecting HVC neurons depends on an intact AFP – has somewhat higher impact. Earlier work from the same lab (Tschida and Mooney, Neuron 2012) convincingly demonstrated that deafening-induced changes in spine morphology precede the degradation of song and that the changes in spine size predicted the magnitude of behavioral changes. These earlier results raise the question of whether the morphological changes reflect the sensory perturbation (deafening) or the beginning of AFP-dependent motor changes that are first detectable days later. The present results demonstrating that spine changes depend on an intact AFP show that such changes are not solely a function of altered auditory feedback, ruling out a simple model (</italic><xref ref-type="fig" rid="fig7"><italic>Figure 7A</italic></xref><italic>) in which changes in spine morphology depend only on the removal of ascending auditory inputs. Although ruling out this model leaves several important questions unanswered (especially the mechanisms by which patterns of AFP activity could regulate spine changes in HVC and how LMAN’s indirect inputs to HVC might interact with ascending auditory signals), it nevertheless represents an advance in knowledge over prior studies</italic>.</p><p><italic>In summary, I have no substantial concerns about how the research was performed or the validity of the results. My primary concern is that the first finding described above does not substantially change current thinking about the system. The second finding advances our understanding of sensorimotor plasticity by ruling out a simple model of auditory dependence and therefore will be of greater interest to readers</italic>.</p><p>We respect the reviewer’s perspective on the novelty and/or potential impact of the intracellular data we collected in singing birds that fails to detect evidence of auditory feedback perturbation. However, we believe that the clear negative results we obtained with these intracellular recordings are critical to distinguishing between two different models of auditory feedback processing in songbirds. Together, the intracellular recordings and imaging datasets we obtained allow us to advance a model where the song premotor elements in HVC are buffered from feedback during singing. This model stands in contrast to that developed by Sakata and Brainard (albeit in a different species) and also clarifies pathways by which deafening affects synapses on song premotor neurons (<xref ref-type="bibr" rid="bib45">Tschida and Mooney, 2012</xref>). As the third reviewer noted, “…these experiments are extremely important in ruling out HVC as the site for such [feedback] monitoring.”</p><p>Reviewer #3:</p><p><italic>A remote possibility with DAF is that auditory responses in HVC</italic><sub><italic>X</italic></sub><italic> might only occur after a window of time. One could envisage, for example, that HVC only processes auditory feedback errors if these occur for a critical amount of time. In other words, auditory flow might be gated for small infrequent errors (this might prevent the motor system from changing to any random perturbation of auditory feedback during singing) but might become “un-gated” once the perturbations are sensed as occurring in a consistent manner (this “consistency” could potentially be monitored in the auditory forebrain and need not rely on the AFP or VTA). A prediction of this idea is that HVC</italic><sub><italic>X</italic></sub><italic> neurons would not show any changes in activity if (1) DAF is too infrequent or (2) when DAF is first introduced. It would be helpful for the authors to address this issue. This argument could easily be resolved if the authors are able to show that DAF-induced changes in song also occur in the birds they record from. Ideally, the most convincing argument would be to show a lack of change in HVC</italic><sub><italic>X</italic></sub><italic> neurons during DAF even while the birds undergo changes in song</italic>.</p><p>The reviewer brings up an interesting idea, namely that a synaptic feedback “gate” onto HVC<sub>X</sub> cells may only open after the bird has experienced frequent and prolonged exposure to DAF. Several observations allay this concern. First, the hit rate for applying DAF we used during intracellular recordings (i.e., ∼50%) has been shown here (<xref ref-type="fig" rid="fig1s3">Figure 1–figure supplement 3</xref>) and by others to be sufficiently frequent to drive adaptive changes in pitch (<xref ref-type="bibr" rid="bib2">Andalman and Fee, 2009</xref>; <xref ref-type="bibr" rid="bib8">Charlesworth et al., 2012</xref>; <xref ref-type="bibr" rid="bib1">Ali et al., 2013</xref>). Second, all of the analyses of subthreshold DAF sensitivity we performed were conducted after the bird had been exposed to DAF for at least an entire preceding day, the time required for the bird to habituate to the exposure, as determined by the decreased frequency of song truncations we observed in response to DAF. Finally, because we recorded from individual birds for one to two weeks and exposed them to DAF throughout much of that time window, many of the cells we recorded from were monitored after the animal had experienced many days or even more than a week of DAF. That is, our intracellular recordings and analyses of subthreshold DAF sensitivity were made in conditions where the bird’s exposure to DAF was both frequent and prolonged. And although we did not track any single cell’s activity while DAF exposure induced song plasticity, a recent study by <xref ref-type="bibr" rid="bib1">Ali et al. (2013)</xref> has shown that singing-related HVC multiunit activity remains unchanged even after contingent DAF is used to drive changes in song pitch. Therefore, we support a model in which auditory synapses onto HVC<sub>X</sub> cells are always gated off when the bird is singing, while also recognizing that experiments involving longer-term intracellular recordings in birds that are undergoing DAF-induced song plasticity are required to fully resolve this issue. These potential concerns and the points made above that address them are now considered in the second paragraph of the Discussion, as directed by the Reviewing editor.</p></body></sub-article></article> |