How sensory neurons process incoming sensory input thereby leading to perception and behavior remains a central question in systems neuroscience. There is growing evidence showing that neural systems can efficiently process natural sensory input by matching their tuning properties to natural stimulus statistics, thereby removing redundancy and thus maximizing information transmission (Fairhall et al., 2001; Brenner et al., 2000; Maravall et al., 2007). Theory posits that efficient coding is achieved by ensuring that the neural tuning function is inversely proportional to stimulus intensity as a function of frequency, thereby achieving a neural response whose amplitude is independent of frequency (Rieke et al., 1996). While such ‘whitening’ has been observed across species and systems (Dan et al., 1996; Wang et al., 2003; Huang et al., 2016; Pozzorini et al., 2013; Pitkow and Meister, 2012), the nature of the underlying mechanisms remains unclear.

Weakly electric fish offer an attractive model system for studying the mechanisms underlying optimized coding of natural stimuli because of well-characterized anatomy and physiology (Clarke et al., 2015; Berman and Maler, 1999; Chacron et al., 2011). These fish generate a quasi-sinusoidal signal called the electric organ discharge (EOD) around their body, thereby allowing them to explore the environment and communicate with conspecifics. When two conspecifics are located close (<1 m) to one another, each fish experiences an amplitude modulation of its own EOD (i.e., a beat or first-order) whose amplitude (i.e., envelope or second-order) is a function of the distance and relative orientation between two conspecifics (Yu et al., 2012; Fotowat et al., 2013) (see (Stamper et al., 2013) for review). Natural electrosensory envelopes due to movement measured in freely moving animals display scale invariance in that their spectral power decays as a power law as a function of increasing temporal frequency (Fotowat et al., 2013; Metzen and Chacron, 2014). Weakly electric fish furthermore give robust behavioral responses to movement related envelopes in which the EOD frequency tracks the stimulus’ detailed (Metzen and Chacron, 2014).

Envelopes and other electrosensory stimuli are sensed by peripheral receptors scattered over the animal’s skin. These project to pyramidal cells within the electrosensory lateral line lobe (ELL) which in turn project to higher brain structures, thereby giving rise to behavior (Figure 1B). ELL pyramidal cells also receive large amounts of descending inputs (i.e., feedback) from higher brain centers such as the nucleus praeeminentialis (nP) and contact ELL pyramidal cells either directly (i.e., the direct feedback pathway) or indirectly (i.e., the indirect feedback pathway; Figure 1B) (Berman and Maler, 1999; Sas and Maler, 1983; Sas and Maler, 1987). Pyramidal cell responses to beats and other first-order electrosensory stimuli are well characterized in general (see (Clarke et al., 2015; Chacron et al., 2011; Marsat et al., 2012; Krahe and Maler, 2014; Huang and Chacron, 2017) for review) and several studies have shown that both direct and indirect feedback inputs play important roles in shaping (Bastian, 1986a; Bastian et al., 2004; Bastian, 1999; Bol et al., 2011; Chacron, 2006; Chacron et al., 2005) as well as synthesizing (Clarke and Maler, 2017) these. Recent studies have focused on characterizing ELL pyramidal cell responses to envelopes (Huang et al., 2016; Stamper et al., 2013; Huang and Chacron, 2017; Middleton et al., 2006; Zhang and Chacron, 2016; Huang and Chacron, 2016; Martinez et al., 2016). Specifically, it was shown that ELL pyramidal cells can optimally encode natural electrosensory envelopes because their high-pass tuning properties effectively oppose the decaying stimulus power, such that the response power is whitened (Huang et al., 2016; Zhang and Chacron, 2016; Huang and Chacron, 2016; Martinez et al., 2016). However, whether and, if so, how direct and indirect feedback inputs mediate optimized coding of envelopes by ELL pyramidal cells has not been investigated to date.

Figure 1

Download asset Open asset

Here we used a systems level approach to investigate how feedback input mediates optimized processing of natural envelope stimuli by ELL pyramidal cells. Pharmacological inactivation of all sources of descending input strongly reduced pyramidal cell and behavioral responses to envelopes. However, pyramidal cell responses to high envelope frequencies were relatively more attenuated than those to low envelope frequencies, such that the resulting tuning was no longer high-pass but independent of envelope frequency. As a consequence, optimized coding of natural stimuli by temporal whitening was compromised. In contrast, selective inactivation of indirect feedback input strongly increased both pyramidal cell and behavioral sensitivity to envelopes. However, enhancement was primarily seen for low envelope frequencies, such that the resulting pyramidal cell tuning curve was broadband, which also compromised optimized coding of natural stimuli by temporal whitening. Finally, we recorded from two different groups of nP neurons that project either directly or indirectly back to ELL pyramidal cells. nP neurons projecting directly to ELL displayed broadband envelope frequency tuning curves and thus did not perform temporal whitening. In contrast, nP neurons projecting indirectly to ELL displayed high-pass envelope frequency tuning curves that effectively opposed the decaying envelope stimulus spectral power content, thereby enabling temporal whitening. Our results demonstrate clear but distinct functional roles for both direct and indirect feedback pathways in determining how ELL pyramidal cells respond to envelopes. While the direct pathway enhances responses to envelopes independently of frequency, the indirect pathway instead selectively attenuates responses to low frequencies, thereby giving rise to the high-pass tuning that is necessary to optimize coding of natural envelopes via temporal whitening. Interestingly, our results also show that indirect feedback input to ELL is temporally whitened. Our results thus demonstrate an important new function for descending inputs onto sensory neurons in optimizing their responses to natural stimuli and their perception at the organismal level.

We investigated the mechanisms that enable ELL pyramidal cells to optimally encode envelope stimuli. To do so, we performed recordings from these in awake behaving animals (Figure 1A). Data obtained from ON- and OFF-type ELL pyramidal neurons were pooled because, consistent with previous studies (Huang and Chacron, 2016), we found no overall difference between their responses to envelopes. Our stimuli consisted of sinusoidal AMs with constant amplitude as well as noisy EOD AMs whose envelope varied sinusoidally at different frequencies. The left panel of Figure 1A shows example traces of the AM (blue, first-order), envelope (red, second-order), and the full signal (cyan) received by the animal with their respective temporal frequency contents. It is further important to realize that the animal’s EOD is a carrier and that the meaningful stimulus here is the EOD AM. Thus, both first- and second-order features of the stimulus correspond to the second- and third-order features of the full signal received by the animal, respectively. We recorded both the neural activity as well as the animal’s behavioral response that consists of changes in the EOD frequency (Figure 1A). Figure 1B shows a simplified diagram of the relevant anatomy and circuitry of the electrosensory system. Peripheral receptors make synaptic contact onto ELL pyramidal cells either directly with excitation for ON-type cells or indirectly via local inhibitory interneurons for OFF-type cells. Pyramidal cells are the sole output neurons of the ELL and project to the midbrain torus semicircularis (TS) which in turn projects to higher brain areas. As mentioned above, ELL pyramidal cells receive large amounts of feedback input both directly from nP (direct feedback; Figure 1B, brown) and indirectly via the Eminentia Granularis posterior (EGp) (indirect feedback; Figure 1B, orange). As explained above, previous studies have shown that ELL pyramidal cells can optimally encode natural envelope stimuli because their high-pass tuning curves (Figure 2, middle panel) are set to counter the decaying stimulus spectral power (Figure 2, left panel). The resulting response spectrum is thus independent of frequency (Figure 2, right panel) which optimizes information transmission (Huang et al., 2016; Huang and Chacron, 2016) (see (Huang and Chacron, 2017) for review).

Figure 2

Download asset Open asset

Descending input shapes neural responses and perception of envelopes

We performed several manipulations that either completely or selectively inactivated feedback input onto ELL pyramidal cells that are schematized in Figure 3. The first consists of completely inactivating feedback by injecting the sodium channel antagonist lidocaine bilaterally into nP (Figure 3A). The second consists of selectively inactivating the indirect feedback by injecting lidocaine bilaterally (for behavior) or ipsilaterally (for neurons) into PET (Figure 3B). The third consists of selectively inactivating the indirect feedback by injecting the non-competitive glutamate receptor antagonist 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX) into the ELL molecular layer (Figure 3C). We note that injecting lidocaine into PET will block indirect feedback input to most if not all pyramidal cells within the ipsilateral ELL. In contrast, injecting CNQX into the ELL molecular layer will only block indirect feedback input in the vicinity of the ELL being recorded from (Chacron et al., 2005; Bastian, 1993).

Figure 3

Download asset Open asset

We initially investigated the effects of complete feedback inactivation on neural and behavioral responses to envelopes (Figure 4). We found that complete feedback inactivation strongly attenuated ELL pyramidal neural responses to envelopes (Figure 4A, compare black and purple). Under control conditions, cells typically responded vigorously to the sinusoidal envelope (Figure 4A, top panel, red) through changes in firing rate (Figure 4A, bottom panel, black). However, after complete feedback inactivation, responses to the envelope were strongly reduced (Figure 4A, bottom panel, purple). We next varied the envelope frequency and investigated the effects of complete feedback inactivation on envelope tuning. Under control conditions, the tuning curve of ELL pyramidal cells is high-pass, such that neural gain increases as a power law when envelope frequency is increased (Figure 4B, black). Further, the power law exponent is set such as to oppose the decay of the spectrum of natural envelopes, thereby causing the response power to be independent of frequency (i.e., is ‘whitened’, Figure 4C, black) as quantified by a white index (see methods) value near unity (Figure 4F, left panel, black), which is required for optimal coding. Complete feedback inactivation strongly affected tuning curves as well as temporal whitening. Indeed, we observed a strong attenuation in neural gain for all envelope frequencies tested (Figure 4B, compare black and purple). However, the attenuation was more pronounced for higher envelope frequencies, such as the resulting tuning curve was flat as characterized by a power law exponent near zero (Figure 4B, purple, see inset). Such a change in tuning strongly affected whitening as response power was no longer independent of frequency (Figure 4C, purple), as quantified by a lower white index value (Figure 4F, left panel, purple) together with a decrease in neural sensitivity (Figure 4F, right panel) indicating sub-optimal coding. These results imply that feedback inputs optimize neural coding of envelopes by enhancing neural responses in a frequency dependent manner. Indeed, higher envelope frequencies are more amplified relative to lower envelope frequencies, thereby whitening neural responses to natural envelopes.

Figure 4 with 3 supplements see all

Download asset Open asset

Changes in neural responses are only behaviorally relevant if they are actually decoded by downstream areas. Thus, we next investigated the effects of pharmacological inactivation of feedback pathways on behavioral responses. Under control conditions, the animal’s EOD frequency tracks the envelope (Figure 4D, black). However, changes in EOD frequency are greater for low envelope frequencies, resulting in a behavioral gain that decreases as a power law for increasing envelope frequency, in a power-law manner (Figure 4E, black), consistent with previous results (Metzen and Chacron, 2014). Complete feedback inactivation strongly attenuated the animal’s behavioral responses to envelopes (Figure 4D, purple) in a frequency dependent manner, such that attenuation was strongest for low envelope frequencies (Figure 4E, purple). Decreases in neural sensitivity were accompanied by decreases in behavioral sensitivity (Figure 4F, right panel).

Interestingly, complete feedback inactivation did not affect ELL pyramidal cell responses to AMs (Figure 4—figure supplement 1). We note that performing a ‘Sham complete feedback inactivation’ by injecting saline instead of lidocaine bilaterally into nP did not alter neural or behavioral responses to envelopes (Figure 4—figure supplement 2). Finally, our results showing that complete feedback inactivation strongly affects ELL neural responses to envelopes were robust. This is because injecting lidocaine unilaterally within the contralateral nP while recording from pyramidal cells within the ipsilateral ELL gave rise to similar changes in neural responses to envelopes (Figure 4—figure supplement 3).

Direct feedback enhances while indirect feedback optimizes neural responses to natural envelopes

We next investigated how direct and indirect sources of descending input contribute to determining ELL pyramidal cell and behavioral responses to envelopes (Figure 5). To do so, we inactivated the indirect feedback pathway by injecting lidocaine into PET bilaterally (Figure 3B). Consistent with previous results, lidocaine injection increased ELL pyramidal neuron responses to AMs (Figure 5—figure supplement 1). We found that indirect feedback inactivation enhanced neural responses to envelopes (Figure 5A, compare black and orange). However, the envelope tuning quantified by neural gain was increased for low but not for high envelope frequencies, such that the tuning curve, which was initially high-pass (Figure 5B, black), became independent of frequency after lidocaine injection (Figure 5B, orange). This change in tuning antagonized optimal coding of envelopes, in that the response power spectrum was no longer independent of envelope frequency (Figure 5C, compare black and orange), as quantified by a decrease in the white index (Figure 5F, left, orange). Neural sensitivity was significantly increased (Figure 5F, right, orange). Thus, our results show that indirect sources of descending input actively shape ELL pyramidal cell tuning to envelopes by attenuating responses to low but not high envelope frequencies, thereby optimizing coding through whitening.

Figure 5 with 3 supplements see all

Download asset Open asset

Indirect feedback inactivation significantly enhanced the animal’s behavioral responses to envelopes (Figure 5D, compare black and orange) only for low frequencies (Figure 5E, compare black and orange). The behavioral response curve thus decreased more steeply with increasing envelope frequency as reflected by a significant decrease in the behavioral exponent (Figure 5E, inset), leading to an increased behavioral sensitivity to envelopes only at low frequencies (Figure 5F, right, orange). Taken together, our results show that both direct and indirect feedback inputs onto ELL pyramidal cells have differential effects on envelope tuning and optimized coding. Specifically, they suggest that the function of the direct input is to enhance envelope responses independently of frequency while the indirect input selectively attenuates low frequencies, thereby optimizing coding.

We note that inactivating indirect feedback by injecting CNQX in the ELL molecular layer (Figure 3C) gave rise to similar effects on ELL pyramidal cell responses to envelopes (Figure 5—figure supplement 2). Also, consistent with previous results (Bastian et al., 2004), this manipulation significantly increased ELL pyramidal cell responses to AMs (Figure 5—figure supplement 3).

Responses of nP neurons that give rise to descending input onto ELL pyramidal cells

Finally, we investigated the nature of the descending signals that are received by ELL pyramidal cells. To do so, we recorded from nP neurons that project both directly and indirectly to ELL (Figure 6A). Specifically, nP stellate cells project directly to ELL pyramidal cells while nP multipolar cells instead project indirectly through the EGp (Sas and Maler, 1983; Sas and Maler, 1987). Both cell types can easily be distinguished from one another based on their electrophysiological properties (Bastian and Bratton, 1990; Bratton and Bastian, 1990) (Figure 6—figure supplement 1). Overall, our results show that both stellate and multipolar cells responded strongly to envelopes (Figure 6B) but showed differential envelope frequency tuning (Figure 6C). Specifically, stellate cell sensitivity was largely independent of envelope frequency (Figure 6C, brown) as quantified by a power law exponent near zero (Figure 6E, left panel, brown). As such, these cells did not perform temporal whitening of natural envelopes as their response power spectra decayed with increasing frequency (Figure 6D and E right, brown). These results confirm our hypothesis that the function of the direct feedback input is to enhance ELL pyramidal cell responses to envelopes independently of temporal frequency.

Figure 6 with 1 supplement see all

Download asset Open asset

In contrast, multipolar cells instead displayed high-pass tuning to envelopes (Figure 6C, orange) as quantified by a power law exponent near 0.4 (Figure 6E left, orange) that is similar to that observed for ELL pyramidal cells (compare with Figure 4B). As a result, we found that multipolar cells perform temporal whitening of envelopes as their response spectra was independent of frequency (Figure 6D, orange) as quantified by a white index near unity (Figure 6E right, orange). Thus, our results reveal that the feedback input that is sent indirectly to ELL pyramidal cells via the EGp is already temporally whitened. This result has important implications for understanding how temporal whitening of ELL pyramidal cell responses is achieved as discussed below.

We investigated the roles of both direct and indirect sources of descending input onto ELL pyramidal cells in determining their responses to envelopes. Pharmacological inactivation of both direct and indirect sources strongly attenuated pyramidal cell and behavioral responses to envelopes. Because responses to higher envelope frequencies were more attenuated, the resulting tuning curve became independent of frequency, thereby compromising optimized coding through temporal whitening. Pharmacological inactivation of indirect input instead increased pyramidal cell and behavioral responses to envelopes. However, enhancement was observed primarily for low envelope frequencies, such that the resulting tuning curve was independent of frequency, which also compromised optimized coding through temporal whitening. Finally, we investigated the nature of the feedback signals being received both directly and indirectly by ELL pyramidal cells. Specifically, nP stellate cells that project directly to ELL displayed tuning curves that were independent of envelope frequency and did not perform temporal whitening. In contrast, nP multipolar cells that project indirectly to ELL displayed high-pass tuning and optimally encoded envelopes through temporal whitening. Thus, our results provide the first experimental evidence showing how descending pathways mediate optimized coding of stimuli by sensory neurons. While direct feedback input enhances neural responses independently of frequency, our results show that indirect feedback input selectively attenuates responses to low envelope frequencies, thereby giving rise to a high-pass tuning that opposes natural envelope statistics and optimizes coding through temporal whitening.

Our results provide a new function for this feedback pathway by showing that nP stellate cells enhance the responses of ELL pyramidal cells to envelopes. Indeed, while previous studies have suggested that the function of this feedback pathway was to enhance responses to salient stimuli (Berman and Maler, 1999; Bratton and Bastian, 1990; Berman and Maler, 1998), experimental evidence supporting this hypothesis was lacking until recently when a clear role in synthesizing responses to motion stimuli consisting exclusively of first-order stimulus features was established (Clarke and Maler, 2017). A recent study has furthermore shown that the direct feedback pathway enables neural responses to weak envelope stimuli (Metzen et al., 2018). Our results show an important novel functional role for the direct feedback pathway in enhancing both neural responses to and perception of behaviorally relevant second-order (i.e., envelope) stimuli that is independent of frequency. Interestingly, previous studies have suggested that the direct feedback pathway could function as a sensory searchlight (Berman and Maler, 1999), as originally proposed by Crick (Crick, 1984), thereby enhancing salient stimulus features and attenuating others. Our results are consistent with descending input onto ELL pyramidal cells acting in this manner but instead suggest that concerted action from both direct and indirect sources is necessary. Indeed, while the direct pathway enhances responses to envelopes independently of frequency, the indirect pathway instead selectively attenuates responses to low frequencies, thereby favoring responses to higher frequencies.

Overall, our results are consistent with previous ones showing that the tuning function of ELL pyramidal cells must be matched to natural statistics in order to optimize coding, which in turn ensures that behavioral sensitivity is greatest for frequencies at which natural stimuli contain the most power (Huang et al., 2016). However, we found that feedback inactivation altered ELL pyramidal cell tuning, thereby rendering coding sub-optimal, which caused a mismatch between behavioral sensitivity and natural statistics. However, our results suggest that the relationship between ELL pyramidal cell and behavioral responses to envelopes is more complicated than previously expected. Specifically, it was assumed that changes in the neural exponent of ELL pyramidal cells alone would determine changes in the behavioral exponent. Our current results show that this is not the case as pharmacological inactivation of both direct and indirect feedback input led to similar changes in the neural tuning exponent but led to opposite changes in the behavioral tuning exponent. Thus, it is not only the tuning exponent of ELL pyramidal cells that determines the behavioral exponent, but also the overall sensitivity. Further studies are needed to better understand how information transmitted by ELL pyramidal cells is decoded by higher brain areas in order to lead to behavioral responses.

Previous results have demonstrated important functions for the indirect feedback pathway such as gain control (Bastian, 1986a; Bastian, 1986b) as well as cancelation of both self and externally-generated low (<15 Hz) frequency spatially diffuse AM (i.e., first-order) stimuli (Bastian et al., 2004; Bastian, 1999; Chacron, 2006; Chacron et al., 2005; Clarke and Maler, 2017; Bastian, 1986b; Bastian, 1996a; Bastian, 1996b; Bastian, 1998), thereby allowing better detection of spatially localized stimuli (e.g., those caused by prey) (Bastian et al., 2004; Litwin-Kumar et al., 2012). Specifically, previous studies have shown that this pathway generates a negative image of the stimulus, thereby strongly reducing ELL pyramidal cell responses. More recent studies have shed light on how short-term burst-timing dependent depression helps in sculpting this negative image in an adaptive fashion (Bol et al., 2011; Bol et al., 2013; Mejias et al., 2013). In contrast, our results have demonstrated an important new function for indirect feedback in mediating optimized coding of envelope (i.e., second-order) stimuli by selectively attenuating responses to low frequencies.

How then can the same feedback pathway mediate attenuation of responses to both low-frequency AMs and envelopes? One possibility is that in both cases it is the same mechanisms that facilitate response attenuation. This is however unlikely to be the case because of the important difference between the frequency ranges of AMs and envelopes for which responses are attenuated, as mentioned above. Specifically, while the response to a 1 Hz AM will be strongly attenuated by feedback (Bol et al., 2011), our results show that this is not the case for a 1 Hz envelope. Previous results have shown that SK1 channels strongly determine envelope but not AM tuning properties. Specifically, pharmacological inactivation of SK1 channels gave rise to minimal effects on responses to AMs (Toporikova and Chacron, 2009) but compromised optimized coding of envelopes by causing the tuning curve to become broadband (Huang et al., 2016). In fact, the effects of indirect feedback inactivation and of SK1 channel antagonists on ELL pyramidal cell tuning to envelopes were strikingly similar (compare our Figure 4B to Figure 6B of (Huang et al., 2016)). We propose that indirect feedback excitation provides the necessary intracellular calcium entry that activates SK1 channels. These in turn give rise to spike frequency adaptation, thereby selectively attenuating responses to low envelope frequencies and causing the envelope tuning curve to become high-pass (Huang et al., 2016; Benda and Herz, 2003; Deemyad et al., 2012) (see (Huang and Chacron, 2017) for review). Thus, blocking indirect feedback input would strongly attenuate calcium entry via NMDA receptors, thereby effectively inhibiting SK1 channels and explaining why the effects on ELL pyramidal cell tuning to envelopes are similar to those previously observed after application of SK channel antagonists (Huang et al., 2016). The fact that calcium-permeable NMDA receptors are highly expressed within the apical dendrites of ELL pyramidal cells (Harvey-Girard and Dunn, 2003; Bottai et al., 1997), where indirect feedback input terminates (Berman and Maler, 1999) and SK1 channels are located (Ellis et al., 2008), supports our hypothesis. However, the exact mechanisms by which SK1 channels determine responses to envelopes but not AMs, though critical in order to enable the indirect pathway to perform multiple functions, are not well-understood and should be the focus of future studies. We also note that anatomical studies have shown that multiple cell types within nP project to EGp (Sas and Maler, 1983; Sas and Maler, 1987). However, with the exception of multipolar cells, the electrophysiological properties of these neurons are unknown to this day. It is very likely that projections from these multiple types underlie the different functions of the indirect feedback pathway but further studies are needed to test this hypothesis.

It is important to note that the envelope stimuli considered here are behaviorally relevant and contain important information as to the relative position between both conspecifics. Since our results show that the indirect feedback did not attenuate responses to envelopes with higher (>0.2 Hz) frequencies, it is conceivable that these could interfere with the detection of other behaviorally-relevant low frequency AM stimuli (e.g., those caused by prey) since the responses to both stimuli should be enhanced by the direct feedback pathway. This is unlikely to be the case however because: 1) the AM stimuli caused by prey are spatially localized whereas envelope stimuli are spatially diffuse and; 2) the temporal frequency content of envelopes tends to be much lower (0 – 1 Hz) than that of prey stimuli (0 – 10 Hz) (Stamper et al., 2013; Nelson and Maciver, 1999). We hypothesize that these differences in spatial extent and temporal frequency content are used by the animal to distinguish between both stimuli, but further studies are needed to verify that this is the case.

We have shown for the first time how neurons within nP that project either directly or indirectly back to ELL respond to envelopes. Stellate cells projecting directly to ELL pyramidal cells displayed broadband tuning curves. The fact that ELL pyramidal cells displayed broadband tuning curves after pharmacological inactivation of indirect descending input suggests that they perform little filtering of input from stellate cells. Interestingly, multipolar cells projecting indirectly have tuning properties that are similar to those of ELL pyramidal cells and optimize coding via temporal whitening. This result has important implications for understanding how temporal whitening of envelopes occurs in ELL pyramidal cells as it implies that the output of the nP received by EGp granule cells that project back to ELL is temporally whitened. It is very likely that the output of nP multipolar cells, which is critical in determining ELL pyramidal cell tuning and responses to envelopes, undergoes significant filtering both by EGp granule cells to enable filtering by SK1 channels in ELL pyramidal cells. Further studies are needed to gain further understanding as to the underlying mechanisms.

Our results provide novel functional roles for the indirect feedback pathway in optimizing ELL pyramidal cell responses to envelopes through temporal whitening by selective attenuation of low frequencies. It should be noted that the architecture of the indirect feedback pathway by which parallel fibers emanating from the EGp terminate onto distal ELL pyramidal cell apical dendrites shares many similarities with the cerebellum and other cerebellum-like structures (e.g., the dorsal cochlear nucleus). Studies performed in cerebellum as well as cerebellum-like structures have shown clear common functions and mechanisms mediating how such descending input in attenuating or even cancelling responses to self-generated sensory input (Cullen, 2011; Bell et al., 2008; Requarth and Sawtell, 2011; Warren and Sawtell, 2016; Sawtell, 2017; Singla et al., 2017). It is likely that our results showing that feedback input from parallel fibers in the cerebellar-like ELL mediate optimized coding of natural stimuli through whitening will be applicable to the cerebellum as well as other cerebellum-like structures.

Finally, recent results have emphasized the critical role of descending pathways, which are found ubiquitously in sensory systems (Cajal, 1909; Holländer, 1970; Ostapoff et al., 1990; Sherman and Guillery, 2002), in determining accurate sensory neural and behavioral responses to sensory input (Manita et al., 2015; Kwon et al., 2016; Takahashi et al., 2016). In particular, it was shown that feedback terminating on the apical dendrites of cortical pyramidal neurons was essential in determining both responses and perception to somatosensory input (Takahashi et al., 2016). Critically, the electrosensory system of weakly electric fish displays many documented similarities with the mammalian visual, auditory, and vestibular systems (Clarke et al., 2015; Chacron et al., 2011; Metzen et al., 2015). Thus, given that temporal whitening has been observed across sensory systems (Dan et al., 1996; Wang et al., 2003; Pozzorini et al., 2013; Lundstrom et al., 2010), it is likely that our results showing how feedback pathways mediate temporal whitening of sensory input by ELL pyramidal cells in the electrosensory system of weakly electric fish will be generally applicable to other systems.

All data have been deposited on Figshare under the URL: https://figshare.com/s/b9e094a67a38e29212e8

1. 1. Huang CG
   2. Metzen MG
   3. Chacron MJ

   (2018) Figure source data for: Feedback optimizes neural coding and perception of natural stimuli

   Available at figshare under a CC0 Public Domain licence (https://figshare.com/).

   https://figshare.com/s/b9e094a67a38e29212e8

1. 1. Bastian J

   (1986a) Gain control in the electrosensory system: a role for the descending projections to the electrosensory lateral line lobe

   Journal of Comparative Physiology A 158:505–515.

   https://doi.org/10.1007/BF00603796

   • Google Scholar
2. 1. Bastian J

   (1986b) Gain control in the electrosensory system mediated by descending inputs to the electrosensory lateral line lobe

   The Journal of Neuroscience 6:553–562.

   https://doi.org/10.1523/JNEUROSCI.06-02-00553.1986

   • PubMed
   • Google Scholar
3. 1. Bastian J
   2. Bratton B

   (1990) Descending control of electroreception. I. Properties of nucleus praeeminentialis neurons projecting indirectly to the electrosensory lateral line lobe

   The Journal of Neuroscience 10:1226–1240.

   https://doi.org/10.1523/JNEUROSCI.10-04-01226.1990

   • PubMed
   • Google Scholar
4. 1. Bastian J

   (1993) The role of amino acid neurotransmitters in the descending control of electroreception

   Journal of Comparative Physiology A 172:409–423.

   https://doi.org/10.1007/BF00213523

   • PubMed
   • Google Scholar
5. 1. Bastian J

   (1996a) Plasticity in an electrosensory system. II. Postsynaptic events associated with a dynamic sensory filter

   Journal of Neurophysiology 76:2497–2507.

   https://doi.org/10.1152/jn.1996.76.4.2497

   • PubMed
   • Google Scholar
6. 1. Bastian J

   (1996b) Plasticity in an electrosensory system. I. General features of a dynamic sensory filter

   Journal of Neurophysiology 76:2483–2496.

   https://doi.org/10.1152/jn.1996.76.4.2483

   • PubMed
   • Google Scholar
7. 1. Bastian J

   (1998) Plasticity in an electrosensory system. III. Contrasting properties of spatially segregated dendritic inputs

   Journal of Neurophysiology 79:1839–1857.

   https://doi.org/10.1152/jn.1998.79.4.1839

   • PubMed
   • Google Scholar
8. 1. Bastian J

   (1999)

   Plasticity of feedback inputs in the apteronotid electrosensory system

   The Journal of Experimental Biology 202:1327–1337.

   • PubMed
   • Google Scholar
9. 1. Bastian J
   2. Chacron MJ
   3. Maler L

   (2002) Receptive field organization determines pyramidal cell stimulus-encoding capability and spatial stimulus selectivity

   The Journal of Neuroscience 22:4577–4590.

   https://doi.org/10.1523/JNEUROSCI.22-11-04577.2002

   • PubMed
   • Google Scholar
10. 1. Bastian J
    2. Chacron MJ
    3. Maler L

    (2004) Plastic and nonplastic pyramidal cells perform unique roles in a network capable of adaptive redundancy reduction

    Neuron 41:767–779.

    https://doi.org/10.1016/S0896-6273(04)00071-6

    • PubMed
    • Google Scholar
11. 1. Bell CC
    2. Han V
    3. Sawtell NB

    (2008) Cerebellum-like structures and their implications for cerebellar function

    Annual Review of Neuroscience 31:1–24.

    https://doi.org/10.1146/annurev.neuro.30.051606.094225

    • PubMed
    • Google Scholar
12. 1. Benda J
    2. Herz AV

    (2003) A universal model for spike-frequency adaptation

    Neural Computation 15:2523–2564.

    https://doi.org/10.1162/089976603322385063

    • PubMed
    • Google Scholar
13. 1. Berman NJ
    2. Maler L

    (1998) Interaction of GABAB-mediated inhibition with voltage-gated currents of pyramidal cells: computational mechanism of a sensory searchlight

    Journal of Neurophysiology 80:3197–3213.

    https://doi.org/10.1152/jn.1998.80.6.3197

    • PubMed
    • Google Scholar
14. 1. Berman NJ
    2. Maler L

    (1999)

    Neural architecture of the electrosensory lateral line lobe: adaptations for coincidence detection, a sensory searchlight and frequency-dependent adaptive filtering

    The Journal of Experimental Biology 202:1243–1253.

    • PubMed
    • Google Scholar
15. 1. Bol K
    2. Marsat G
    3. Harvey-Girard E
    4. Longtin A
    5. Maler L

    (2011) Frequency-tuned cerebellar channels and burst-induced LTD lead to the cancellation of redundant sensory inputs

    Journal of Neuroscience 31:11028–11038.

    https://doi.org/10.1523/JNEUROSCI.0193-11.2011

    • PubMed
    • Google Scholar
16. 1. Bol K
    2. Marsat G
    3. Mejias JF
    4. Maler L
    5. Longtin A

    (2013) Modeling cancelation of periodic inputs with burst-STDP and feedback

    Neural Networks 47:120–133.

    https://doi.org/10.1016/j.neunet.2012.12.011

    • PubMed
    • Google Scholar
17. 1. Bottai D
    2. Dunn RJ
    3. Ellis W
    4. Maler L

    (1997) N-methyl-D-aspartate receptor 1 mRNA distribution in the central nervous system of the weakly electric fish Apteronotus leptorhynchus

    The Journal of Comparative Neurology 389:65–80.

    https://doi.org/10.1002/(SICI)1096-9861(19971208)389:1<65::AID-CNE5>3.0.CO;2-C

    • PubMed
    • Google Scholar
18. 1. Bratton B
    2. Bastian J

    (1990) Descending control of electroreception. II. Properties of nucleus praeeminentialis neurons projecting directly to the electrosensory lateral line lobe

    The Journal of Neuroscience 10:1241–1253.

    https://doi.org/10.1523/JNEUROSCI.10-04-01241.1990

    • PubMed
    • Google Scholar
19. 1. Brenner N
    2. Bialek W
    3. de Ruyter van Steveninck R

    (2000) Adaptive rescaling maximizes information transmission

    Neuron 26:695–702.

    https://doi.org/10.1016/S0896-6273(00)81205-2

    • PubMed
    • Google Scholar
20. Book

    1. Cajal RS

    (1909)

    Histologie Du Système Nerveux De l'Homme Et Des Vertébrés 

    Paris: Maloine.

    • Google Scholar
21. 1. Chacron MJ
    2. Maler L
    3. Bastian J

    (2005) Feedback and feedforward control of frequency tuning to naturalistic stimuli

    Journal of Neuroscience 25:5521–5532.

    https://doi.org/10.1523/JNEUROSCI.0445-05.2005

    • PubMed
    • Google Scholar
22. 1. Chacron MJ

    (2006) Nonlinear information processing in a model sensory system

    Journal of Neurophysiology 95:2933–2946.

    https://doi.org/10.1152/jn.01296.2005

    • PubMed
    • Google Scholar
23. 1. Chacron MJ
    2. Longtin A
    3. Maler L

    (2011) Efficient computation via sparse coding in electrosensory neural networks

    Current Opinion in Neurobiology 21:752–760.

    https://doi.org/10.1016/j.conb.2011.05.016

    • PubMed
    • Google Scholar
24. 1. Clarke SE
    2. Longtin A
    3. Maler L

    (2015) Contrast coding in the electrosensory system: parallels with visual computation

    Nature Reviews Neuroscience 16:733–744.

    https://doi.org/10.1038/nrn4037

    • PubMed
    • Google Scholar
25. 1. Clarke SE
    2. Maler L

    (2017) Feedback Synthesizes Neural Codes for Motion

    Current Biology 27:1356–1361.

    https://doi.org/10.1016/j.cub.2017.03.068

    • PubMed
    • Google Scholar
26. 1. Crick F

    (1984) Function of the thalamic reticular complex: the searchlight hypothesis

    PNAS 81:4586–4590.

    https://doi.org/10.1073/pnas.81.14.4586

    • PubMed
    • Google Scholar
27. 1. Cullen KE

    (2011) The neural encoding of self-motion

    Current Opinion in Neurobiology 21:587–595.

    https://doi.org/10.1016/j.conb.2011.05.022

    • PubMed
    • Google Scholar
28. 1. Dan Y
    2. Atick JJ
    3. Reid RC

    (1996) Efficient coding of natural scenes in the lateral geniculate nucleus: experimental test of a computational theory

    The Journal of Neuroscience 16:3351–3362.

    https://doi.org/10.1523/JNEUROSCI.16-10-03351.1996

    • PubMed
    • Google Scholar
29. 1. Deemyad T
    2. Kroeger J
    3. Chacron MJ

    (2012) Sub- and suprathreshold adaptation currents have opposite effects on frequency tuning

    The Journal of Physiology 590:4839–4858.

    https://doi.org/10.1113/jphysiol.2012.234401

    • PubMed
    • Google Scholar
30. 1. Deemyad T
    2. Metzen MG
    3. Pan Y
    4. Chacron MJ

    (2013) Serotonin selectively enhances perception and sensory neural responses to stimuli generated by same-sex conspecifics

    PNAS 110:19609–19614.

    https://doi.org/10.1073/pnas.1314008110

    • PubMed
    • Google Scholar
31. 1. Ellis LD
    2. Maler L
    3. Dunn RJ

    (2008) Differential distribution of SK channel subtypes in the brain of the weakly electric fish Apteronotus leptorhynchus

    The Journal of Comparative Neurology 507:1964–1978.

    https://doi.org/10.1002/cne.21597

    • PubMed
    • Google Scholar
32. 1. Fairhall AL
    2. Lewen GD
    3. Bialek W
    4. de Ruyter Van Steveninck RR

    (2001) Efficiency and ambiguity in an adaptive neural code

    Nature 412:787–792.

    https://doi.org/10.1038/35090500

    • PubMed
    • Google Scholar
33. 1. Fotowat H
    2. Harrison RR
    3. Krahe R

    (2013) Statistics of the electrosensory input in the freely swimming weakly electric fish Apteronotus leptorhynchus

    Journal of Neuroscience 33:13758–13772.

    https://doi.org/10.1523/JNEUROSCI.0998-13.2013

    • PubMed
    • Google Scholar
34. 1. Frank K
    2. Becker MC

    (1964)

    Physical Techniques in Biological Research

    23–84, Microelectrodes for recording and stimulation :, Physical Techniques in Biological Research, 5, Nastuk WL, New York, Academic, 10.1016/B978-0-12-514105-5.50009-5.

    • Google Scholar
35. 1. Harvey-Girard E
    2. Dunn RJ

    (2003) Excitatory amino acid receptors of the electrosensory system: the NR1/NR2B N-methyl-D-aspartate receptor

    Journal of Neurophysiology 89:822–832.

    https://doi.org/10.1152/jn.00629.2002

    • PubMed
    • Google Scholar
36. 1. Hitschfeld EM
    2. Stamper SA
    3. Vonderschen K
    4. Fortune ES
    5. Chacron MJ

    (2009) Effects of restraint and immobilization on electrosensory behaviors of weakly electric fish

    ILAR Journal 50:361–372.

    https://doi.org/10.1093/ilar.50.4.361

    • PubMed
    • Google Scholar
37. 1. Holländer H

    (1970) The projection from the visual cortex to the lateral geniculate body (LGB). An experimental study with silver impregnation methods in the cat

    Experimental Brain Research 10:219–235.

    https://doi.org/10.1007/BF00235047

    • PubMed
    • Google Scholar
38. 1. Huang CG
    2. Chacron MJ

    (2016) Optimized Parallel Coding of Second-Order Stimulus Features by Heterogeneous Neural Populations

    Journal of Neuroscience 36:9859–9872.

    https://doi.org/10.1523/JNEUROSCI.1433-16.2016

    • PubMed
    • Google Scholar
39. 1. Huang CG
    2. Zhang ZD
    3. Chacron MJ

    (2016) Temporal decorrelation by SK channels enables efficient neural coding and perception of natural stimuli

    Nature Communications 7:11353.

    https://doi.org/10.1038/ncomms11353

    • PubMed
    • Google Scholar
40. 1. Huang CG
    2. Chacron MJ

    (2017) SK channel subtypes enable parallel optimized coding of behaviorally relevant stimulus attributes: a review

    Channels 11:281–304.

    https://doi.org/10.1080/19336950.2017.1299835

    • PubMed
    • Google Scholar
41. Website

    1. Huang CG
    2. Metzen MG
    3. Chacron MJ

    (2018) Figure source data from feedback optimizes neural coding and perception of natural stimuli

    Accessed October 2018.

    https://figshare.com/s/b9e094a67a38e29212e8
42. 1. Krahe R
    2. Bastian J
    3. Chacron MJ

    (2008) Temporal processing across multiple topographic maps in the electrosensory system

    Journal of Neurophysiology 100:852–867.

    https://doi.org/10.1152/jn.90300.2008

    • PubMed
    • Google Scholar
43. 1. Krahe R
    2. Maler L

    (2014) Neural maps in the electrosensory system of weakly electric fish

    Current Opinion in Neurobiology 24:13–21.

    https://doi.org/10.1016/j.conb.2013.08.013

    • PubMed
    • Google Scholar
44. 1. Kwon SE
    2. Yang H
    3. Minamisawa G
    4. O'Connor DH

    (2016) Sensory and decision-related activity propagate in a cortical feedback loop during touch perception

    Nature Neuroscience 19:1243–1249.

    https://doi.org/10.1038/nn.4356

    • PubMed
    • Google Scholar
45. 1. Litwin-Kumar A
    2. Chacron MJ
    3. Doiron B

    (2012) The spatial structure of stimuli shapes the timescale of correlations in population spiking activity

    PLoS Computational Biology 8:e1002667.

    https://doi.org/10.1371/journal.pcbi.1002667

    • PubMed
    • Google Scholar
46. 1. Lundstrom BN
    2. Fairhall AL
    3. Maravall M

    (2010) Multiple timescale encoding of slowly varying whisker stimulus envelope in cortical and thalamic neurons in vivo

    Journal of Neuroscience 30:5071–5077.

    https://doi.org/10.1523/JNEUROSCI.2193-09.2010

    • PubMed
    • Google Scholar
47. 1. Manita S
    2. Suzuki T
    3. Homma C
    4. Matsumoto T
    5. Odagawa M
    6. Yamada K
    7. Ota K
    8. Matsubara C
    9. Inutsuka A
    10. Sato M
    11. Ohkura M
    12. Yamanaka A
    13. Yanagawa Y
    14. Nakai J
    15. Hayashi Y
    16. Larkum ME
    17. Murayama M

    (2015) A Top-Down Cortical Circuit for Accurate Sensory Perception

    Neuron 86:1304–1316.

    https://doi.org/10.1016/j.neuron.2015.05.006

    • PubMed
    • Google Scholar
48. 1. Maravall M
    2. Petersen RS
    3. Fairhall AL
    4. Arabzadeh E
    5. Diamond ME

    (2007) Shifts in coding properties and maintenance of information transmission during adaptation in barrel cortex

    PLoS Biology 5:e19.

    https://doi.org/10.1371/journal.pbio.0050019

    • PubMed
    • Google Scholar
49. 1. Marsat G
    2. Longtin A
    3. Maler L

    (2012) Cellular and circuit properties supporting different sensory coding strategies in electric fish and other systems

    Current Opinion in Neurobiology 22:686–692.

    https://doi.org/10.1016/j.conb.2012.01.009

    • PubMed
    • Google Scholar
50. 1. Martinez D
    2. Metzen MG
    3. Chacron MJ

    (2016) Electrosensory processing in Apteronotus albifrons: implications for general and specific neural coding strategies across wave-type weakly electric fish species

    Journal of Neurophysiology 116:2909–2921.

    https://doi.org/10.1152/jn.00594.2016

    • PubMed
    • Google Scholar
51. 1. McGillivray P
    2. Vonderschen K
    3. Fortune ES
    4. Chacron MJ

    (2012) Parallel coding of first- and second-order stimulus attributes by midbrain electrosensory neurons

    Journal of Neuroscience 32:5510–5524.

    https://doi.org/10.1523/JNEUROSCI.0478-12.2012

    • PubMed
    • Google Scholar
52. 1. Mejias JF
    2. Marsat G
    3. Bol K
    4. Maler L
    5. Longtin A

    (2013) Learning contrast-invariant cancellation of redundant signals in neural systems

    PLoS Computational Biology 9:e1003180.

    https://doi.org/10.1371/journal.pcbi.1003180

    • PubMed
    • Google Scholar
53. 1. Metzen MG
    2. Chacron MJ

    (2014) Weakly electric fish display behavioral responses to envelopes naturally occurring during movement: implications for neural processing

    Journal of Experimental Biology 217:1381–1391.

    https://doi.org/10.1242/jeb.098574

    • PubMed
    • Google Scholar
54. 1. Metzen MG
    2. Jamali M
    3. Carriot J
    4. Ávila-Ǻkerberg O
    5. Cullen KE
    6. Chacron MJ

    (2015) Coding of envelopes by correlated but not single-neuron activity requires neural variability

    PNAS 112:4791–4796.

    https://doi.org/10.1073/pnas.1418224112

    • PubMed
    • Google Scholar
55. 1. Metzen MG
    2. Hofmann V
    3. Chacron MJ

    (2016) Neural correlations enable invariant coding and perception of natural stimuli in weakly electric fish

    eLife 5:e12993.

    https://doi.org/10.7554/eLife.12993

    • PubMed
    • Google Scholar
56. 1. Metzen MG
    2. Chacron MJ

    (2017) Stimulus background influences phase invariant coding by correlated neural activity

    eLife 6:e24482.

    https://doi.org/10.7554/eLife.24482

    • PubMed
    • Google Scholar
57. 1. Metzen MG
    2. Huang CG
    3. Chacron MJ

    (2018) Descending pathways generate perception of and neural responses to weak sensory input

    PLoS Biology 16:e2005239.

    https://doi.org/10.1371/journal.pbio.2005239

    • PubMed
    • Google Scholar
58. 1. Middleton JW
    2. Longtin A
    3. Benda J
    4. Maler L

    (2006) The cellular basis for parallel neural transmission of a high-frequency stimulus and its low-frequency envelope

    PNAS 103:14596–14601.

    https://doi.org/10.1073/pnas.0604103103

    • PubMed
    • Google Scholar
59. 1. Nelson ME
    2. Maciver MA

    (1999)

    Prey capture in the weakly electric fish Apteronotus albifrons: sensory acquisition strategies and electrosensory consequences

    The Journal of Experimental Biology 202:1195–1203.

    • PubMed
    • Google Scholar
60. 1. Ostapoff EM
    2. Morest DK
    3. Potashner SJ

    (1990)

    Uptake and retrograde transport of [3H]GABA from the cochlear nucleus to the superior olive in the guinea pig

    Journal of Chemical Neuroanatomy 3:285–295.

    • PubMed
    • Google Scholar
61. 1. Pitkow X
    2. Meister M

    (2012) Decorrelation and efficient coding by retinal ganglion cells

    Nature Neuroscience 15:628–635.

    https://doi.org/10.1038/nn.3064

    • PubMed
    • Google Scholar
62. 1. Pozzorini C
    2. Naud R
    3. Mensi S
    4. Gerstner W

    (2013) Temporal whitening by power-law adaptation in neocortical neurons

    Nature Neuroscience 16:942–948.

    https://doi.org/10.1038/nn.3431

    • PubMed
    • Google Scholar
63. 1. Requarth T
    2. Sawtell NB

    (2011) Neural mechanisms for filtering self-generated sensory signals in cerebellum-like circuits

    Current Opinion in Neurobiology 21:602–608.

    https://doi.org/10.1016/j.conb.2011.05.031

    • PubMed
    • Google Scholar
64. Book

    1. Rieke F
    2. Warland D
    3. de Ruyter van Steveninck RR
    4. Bialek W

    (1996)

    Spikes: Exploring the Neural Code

    Cambridge, MA: MIT Press.

    • Google Scholar
65. 1. Sas E
    2. Maler L

    (1983) The nucleus praeeminentialis: a golgi study of a feedback center in the electrosensory system of gymnotid fish

    The Journal of Comparative Neurology 221:127–144.

    https://doi.org/10.1002/cne.902210202

    • PubMed
    • Google Scholar
66. 1. Sas E
    2. Maler L

    (1987) The organization of afferent input to the caudal lobe of the cerebellum of the gymnotid fish Apteronotus leptorhynchus

    Anatomy and Embryology 177:55–79.

    https://doi.org/10.1007/BF00325290

    • PubMed
    • Google Scholar
67. 1. Savard M
    2. Krahe R
    3. Chacron MJ

    (2011) Neural heterogeneities influence envelope and temporal coding at the sensory periphery

    Neuroscience 172:270–284.

    https://doi.org/10.1016/j.neuroscience.2010.10.061

    • PubMed
    • Google Scholar
68. 1. Sawtell NB

    (2017) Neural Mechanisms for Predicting the Sensory Consequences of Behavior: Insights from Electrosensory Systems

    Annual Review of Physiology 79:381–399.

    https://doi.org/10.1146/annurev-physiol-021115-105003

    • PubMed
    • Google Scholar
69. 1. Sherman SM
    2. Guillery RW

    (2002) The role of the thalamus in the flow of information to the cortex

    Philosophical Transactions of the Royal Society B: Biological Sciences 357:1695–1708.

    https://doi.org/10.1098/rstb.2002.1161

    • PubMed
    • Google Scholar
70. 1. Singla S
    2. Dempsey C
    3. Warren R
    4. Enikolopov AG
    5. Sawtell NB

    (2017) A cerebellum-like circuit in the auditory system cancels responses to self-generated sounds

    Nature Neuroscience 20:943–950.

    https://doi.org/10.1038/nn.4567

    • PubMed
    • Google Scholar
71. 1. Stamper SA
    2. Carrera-G E
    3. Tan EW
    4. Fugère V
    5. Krahe R
    6. Fortune ES

    (2010) Species differences in group size and electrosensory interference in weakly electric fishes: implications for electrosensory processing

    Behavioural Brain Research 207:368–376.

    https://doi.org/10.1016/j.bbr.2009.10.023

    • PubMed
    • Google Scholar
72. 1. Stamper SA
    2. Madhav MS
    3. Cowan NJ
    4. Fortune ES

    (2012) Beyond the Jamming Avoidance Response: weakly electric fish respond to the envelope of social electrosensory signals

    Journal of Experimental Biology 215:4196–4207.

    https://doi.org/10.1242/jeb.076513

    • PubMed
    • Google Scholar
73. 1. Stamper SA
    2. Fortune ES
    3. Chacron MJ

    (2013) Perception and coding of envelopes in weakly electric fishes

    Journal of Experimental Biology 216:2393–2402.

    https://doi.org/10.1242/jeb.082321

    • PubMed
    • Google Scholar
74. 1. Takahashi N
    2. Oertner TG
    3. Hegemann P
    4. Larkum ME

    (2016) Active cortical dendrites modulate perception

    Science 354:1587–1590.

    https://doi.org/10.1126/science.aah6066

    • PubMed
    • Google Scholar
75. 1. Thomas RA
    2. Metzen MG
    3. Chacron MJ

    (2018) Weakly electric fish distinguish between envelope stimuli arising from different behavioral contexts

    The Journal of Experimental Biology 221:jeb178244.

    https://doi.org/10.1242/jeb.178244

    • PubMed
    • Google Scholar
76. 1. Toporikova N
    2. Chacron MJ

    (2009) SK channels gate information processing in vivo by regulating an intrinsic bursting mechanism seen in vitro

    Journal of Neurophysiology 102:2273–2287.

    https://doi.org/10.1152/jn.00282.2009

    • PubMed
    • Google Scholar
77. 1. Wang XJ
    2. Liu Y
    3. Sanchez-Vives MV
    4. McCormick DA

    (2003) Adaptation and temporal decorrelation by single neurons in the primary visual cortex

    Journal of Neurophysiology 89:3279–3293.

    https://doi.org/10.1152/jn.00242.2003

    • PubMed
    • Google Scholar
78. 1. Warren R
    2. Sawtell NB

    (2016) A comparative approach to cerebellar function: insights from electrosensory systems

    Current Opinion in Neurobiology 41:31–37.

    https://doi.org/10.1016/j.conb.2016.07.012

    • PubMed
    • Google Scholar
79. 1. Yu N
    2. Hupé G
    3. Garfinkle C
    4. Lewis JE
    5. Longtin A

    (2012) Coding conspecific identity and motion in the electric sense

    PLoS Computational Biology 8:e1002564.

    https://doi.org/10.1371/journal.pcbi.1002564

    • PubMed
    • Google Scholar
80. 1. Zhang ZD
    2. Chacron MJ

    (2016) Adaptation to second order stimulus features by electrosensory neurons causes ambiguity

    Scientific Reports 6:28716.

    https://doi.org/10.1038/srep28716

    • PubMed
    • Google Scholar

Author details

1. Chengjie G Huang

   Department of Physiology, McGill University, Montreal, Canada

   Contribution

   Conceptualization, Data curation, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing—original draft, Writing—review and editing

   Contributed equally with

   Michael G Metzen

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-7491-6060

2. Michael G Metzen

   Department of Physiology, McGill University, Montreal, Canada

   Contribution

   Conceptualization, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing—original draft, Writing—review and editing

   Contributed equally with

   Chengjie G Huang

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-2365-4192

3. Maurice J Chacron

   Department of Physiology, McGill University, Montreal, Canada

   Contribution

   Conceptualization, Resources, Software, Supervision, Funding acquisition, Validation, Writing—original draft, Project administration, Writing—review and editing

   For correspondence

   maurice.chacron@mcgill.ca

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-3032-452X

• 1,429

  views

• 203

  downloads

• 20

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.