Visual stimuli guide the behavior of many animals. While the detection of ethologically relevant visual stimuli can elicit innate behavior, often visual stimuli become relevant through learning, leading to goal-directed behavior upon stimulus detection (Morris et al., 2018; Schultz, 2006). For example, many vertebrates can learn to alter their behavior in response to the detection of arbitrary visual stimuli in order to obtain a reward or avoid punishment (Llinás, 1976; Prusky and Douglas, 2004; Skinner, 1963; Valente, 2012). A major challenge to our understanding of the neuronal basis of this elemental form of sensory-based and goal-directed behavior is that any visual stimulus evokes neuronal activity across multiple brain structures (Macé et al., 2018; Seabrook et al., 2017) and, within each structure, across diverse types of neurons (Harris and Mrsic-Flogel, 2013; Harris and Shepherd, 2015; Reinhard et al., 2019).

The striatum and the optic tectum are two evolutionary conserved subcortical structures involved in the learning and performance of simple goal-directed behavior in many vertebrates. The striatum is fundamental for reinforcement learning (Cox and Witten, 2019) and action initiation (Klaus et al., 2019) and the optic tectum (called the superior colliculus in mammals) plays an important role in the detection of salient visual stimuli that trigger innate behaviors (Feinberg and Mallatt, 2019; Grillner and El Manira, 2020). At least in mammals, both structures receive two main sources of visual information: one ascending from the periphery (Klaus et al., 2019; Krauzlis et al., 2013) and the other, descending from visual cortex (VC) via its prominent cortico-fugal pathways (Feinberg and Mallatt, 2019; Smeets et al., 2000; Suryanarayana et al., 2020). Whether the neurons in VC that give rise to these cortico-fugal pathways play a role in simple goal-directed behavior remains poorly understood. In fact, mammals are able to learn and perform simple sensory detection tasks even in the absence of sensory neocortex (Ceballo et al., 2019; Dalmay et al., 2019; Hong et al., 2018; Pöppel et al., 1973; Popper and Fay, 1992; Weiskrantz et al., 1974). Thus, it remains unknown whether cortico-fugal pathways, the main output pathways by which VC can influence the rest of the brain and thus, ultimately, behavior, do actually contribute to simple goal-directed behavior.

Here, we determined the impact of two major populations of cortico-fugal neurons in VC in the speed of learning and in the performance sensitivity of a visual detection task in mice. Using an intersectional viral approach to selectively eliminate specific populations of cortico-fugal neurons in VC, we show that the ablation of neurons projecting to the striatum reduces learning speed during task acquisition, whereas the ablation of neurons projecting to the superior colliculus impairs detection sensitivity during task execution. Furthermore, we show that, with training, both populations of cortico-fugal neurons eventually become dispensable for the task. These data demonstrate the functional dissociation between two distinct populations of cortico-fugal neurons in VC during specific training stages of a visual detection task and highlight that specific cortico-fugal pathways adaptively contribute even to simple goal-directed behavior.

Distinct populations of neurons in VC send visual information to the dorso-medial striatum (dmSt) and the superior colliculus (SC; the mammalian optic tectum) via two prominent cortico-fugal pathways, the cortico-striatal (CSt; Faull et al., 1986; Kemp and Powell, 1970; Khibnik et al., 2014; Saint-Cyr et al., 1990) and the cortico-tectal (CT; Wang and Burkhalter, 2013; Zingg et al., 2017), respectively (Hattox and Nelson, 2007; Jones, 1984; Lur et al., 2016; Norita et al., 1991; Rhoades et al., 1985; Serizawa et al., 1994; Swadlow, 1983; Tang and Higley, 2019). To determine the role of these two populations of cortico-fugal neurons in a simple visual detection task, we trained water restricted, head fixed mice to report the appearance of a visual stimulus with a lick on a waterspout. The visual stimulus consisted of a 30° drifting grating patch presented at full contrast (unless specified otherwise) on a computer monitor placed to the left visual field of the animal (Figure 1A). To be rewarded with water, the animal had to report the presence of the stimulus such that the first lick occurred during the stimulus presentation. To determine the performance of the animal, we used both the first lick latency and the probability of a lick during either stimulus presentation or blank periods. Hits were stimulus trials on which the animal licked; correct rejections were blank trials (no stimulus) in which the animal refrained from licking throughout the response period. Misses and false alarms were omission of licks on stimulus trials, and licks on blank trials, respectively (Figure 1A; for trial structure see Figure 1—figure supplement 1). To determine the visual specificity of the licking behavior, that is the extent to which licking is guided by the visual stimulus, we compared the first lick latency distributions between stimulus and blank trials using Receiver Operating Characteristics (ROC) analysis (Fawcett, 2006; Macmillan and Creelman, 2005). The area under the ROC curve (aROC) reports the probability of an ideal observer to correctly classify trial type (i.e. stimulus versus blank) based on first lick latency. An aROC of 1 corresponds to a complete separability of the two trials types based on the distribution of first lick latencies, and thus a perfect classification. In contrast, an aROC of 0.5 represents a complete overlap of the temporal distributions of first lick latencies and thus, chance performance. With the aROC of first lick latencies, it is thus possible to detect the emergence of visually guided behavior even while hits and false alarms occur at the same rate, as long as the first lick latency in response to a stimulus trial differs from that to a blank trial.

Figure 1 with 6 supplements see all

Download asset Open asset

Figure 1—source data 1

Behavioral performance measurements as a function of training days for intact animals, CT-ablated, and CSt-ablated animals.

https://cdn.elifesciences.org/articles/59247/elife-59247-fig1-data1-v2.xlsx

Download elife-59247-fig1-data1-v2.xlsx

Learning was characterized by a gradual increase in the aROC (linear fit from day 1 to day 14 aROC/day: 0.028 ± 0.003; mean ± SEM; N = 8 mice; population average slope aROC/day: 0.030 ± 0.023; mean ±RMSE, R² = 0.9584, p<0.001; F-test), by a progressive reduction in first lick latency and a reduction in the trial-to-trial variability of the first lick latency on stimulus trials (Figure 1C). Learning was considered complete when it reached an aROC of at least 0.8 for 4 consecutive days. On average learning was completed by the 14th day of training (aROC 0.86 ± 0.04; first lick latency 0.52 ± 0.13 s; variability:0.12 ± 0.05 s; mean ± SEM; N = 8 mice). The temporal distributions of first lick latencies started to differ between stimulus and blank trials early during training (Figure 1B and Figure 1—figure supplement 1) revealing the beginning of the formation of the association between stimulus and reward by the animal (day 4 FL latency: stim. 0.77 ± 0.20 s vs. blank: 1.01 ± 0.16 s; mean ± SEM; N = 8 mice; p<0.01; Wilcoxon signed-rank test; Figure 1C and Figure 1—figure supplement 2). In contrast to ROC analysis, lick probability and thus d-prime strongly depended on the duration of the considered response window (Figure 1B and Figure 1—figure supplement 2A). For example, by day 4 of training, while aROC was well above chance (aROC = 0.60 ± 0.03; mean ± SEM; N = 8 mice; p<0.001; Mann-Whitney U test), there were as many hits as false alarms over the full 4 s response window (Day 4: Hit Rate: 89.9% ± 3.8 False Alarm Rate: 81.6% ± 4.9; mean ± SEM; N = 8 mice; n.s.; Wilcoxon signed-rank test; d-prime: 0.6 ± 0.13, Figure 1—figure supplement 2A–B). That is, the animals licked earlier in the response window following a stimulus as compared to a blank (Day 4: mean FL latency Stim vs Blank), but the lick probability within the whole response window was similar for the two conditions, mainly because of the high spontaneous lick rate of animals early in learning. Accordingly, reducing the time window within which to analyze licks probabilities increased d-prime, up to the ‘optimal’ response window, where d-prime is maximal (d-prime based on optimal response window per animal: 1.49 ± 0.48 s; mean ± SEM; N = 8 mice; Figure 1—figure supplement 2B). The optimal response window, however, varied widely across training days and animals (shortest optimal response window 0.38 ± 0.07 s vs longest optimal response window: 2.90 ± 0.39 s; mean ± SEM, N = 8 mice; p<0.001; Wilcoxon signed-rank test; average optimal response window: 1.19 ± 0.19 s; Figure 1—figure supplement 2B). In contrast, by defining proficiency based on the temporal distribution of first lick latencies using the aROC rather than on the probability to lick, mice were able to reach expertness before maximizing inhibitory control throughout the 4 s response window. Thus, aROC analysis over an extended response window captures the temporal characteristics of the behavioral response and offers a reliable performance metric to determine the specificity of stimulus guided responses.

To address the contribution of VC neurons projecting to the SC or to the dmSt to the learning of this simple detection task, we selectively ablated CT or CSt neurons, respectively. For this, we used an intersectional viral strategy that takes advantage of a designer AAV-Cre virus (Madisen et al., 2015; Tervo et al., 2016) to conditionally express taCaspase3 (taCasp3; Yang et al., 2013) in select populations of cortico-fugal neurons. Using the retrograde tracer Cholera toxin B (CTB; Luppi et al., 1990; Wan et al., 1982), we verified that CT and CSt neurons indeed represent distinct populations in VC (overlap <3.4%; 107 of 3187 CTB-labeled cells; 4.62 ± 1.13% of the CSt and 8.82 ± 4.88% of the CT population; Figure 1—figure supplement 3), consistent with previous reports (Brown and Hestrin, 2009; Lur et al., 2016; Norita et al., 1991; Serizawa et al., 1994). The vast majority of CT and CSt neurons were located in layer 5 and, while the two populations of neurons largely intermingled, CT neurons tended to be more tightly distributed toward the lower part of the layer (Figure 1—figure supplement 3).

To determine whether CT neurons are necessary to learn the detection task, we selectively ablated them by injecting, 3 weeks before training onset, Ai14 reporter mice with retroAAV-Cre (Tervo et al., 2016) in the right SC to conditionally express taCasp3 (Yang et al., 2013) in the right VC, contralateral to the visual stimulus. The efficiency of retroAAV-Cre in retrogradely labeling cortico-fugal neurons was similar to CTB (retroAAV-Cre: 30 ± 14 cells/100 µm³; retrograde tracer CTB: 24 ± 9 cells/100 µm³ in layer 5; mean ± SD; N = 3 mice). We also directly compared the efficiency of the retroAAV-Cre by co-injecting it with CTB in the Rosa26 LSL H2B reporter mouse. The retroAAV-Cre was highly efficient as the majority of CTB-labeled cells co-expressed nuclear mCherry (85.5%, 864 cells of 1010, N = 2 mice; Figure 1—figure supplement 4). The viral ablation of Cre recombinase expressing CT neurons in VC was highly efficient with over 90% of ablated neurons 3 weeks after the injection, as verified histologically (retroAAV-Cre only: 26 ± 8 cells/100 µm³ in layer 5 vs. retroAAV-Cre and AAV-taCasp3: 1 ± 1 cells/100 µm³ in layer 5; mean ± SD; N = 3 mice; p<0.001 Mann-Whitney U test; Figure 1—figure supplement 5A). The expression of taCasp3 reduced the population of NeuN-labeled neurons in layer 5 by 20.06 ± 3.95% (NeuN-labeled neurons control: 27.55 ± 1.25 cells/100 µm² vs. casp3: 21.88 ± 0.32 cells/100 µm²; mean ± SEM; N = 3 mice, p<0.01 Mann-Whitney U test; cells per animal: control: 1203.2 ± 13.52; casp3: 853.7 ± 3.34; mean ± SEM). This is in line with the number of CT neurons in this layer (calculated using retrogradely labeled CTB neurons 4.04 ± 0.55 cells/100 µm²; N = 3 mice) and consistent with taCasp3 toxicity being limited to Cre-expressing neurons (Gray et al., 2010).

Ablation of CT neurons had little impact on task learning as compared to controls in which only the retroAAV-Cre had been injected (aROC/day CT-ablation: 0.03 ± 0.006 vs. retroAAV-Cre only: 0.03 ± 0.004; mean ± SEM; N = 5 mice, n.s. Mann-Whitney U test; population average CT-ablation slope aROC/day: 0.029 ± 0.0014 vs. retroAAV-Cre only slope aROC/day: 0.026 ± 0.001; mean ±RMSE; n.s.; Fisher z-test; Figure 1D and Figure 2D). Accordingly, by the 14th day of training, CT-ablated mice reached an aROC similar to controls animals (CT-ablation aROC:0.85 ± 0.08 vs. retroAAV-Cre only aROC 0.84 ± 0.04; mean ± SEM; n.s. Mann-Whitney U test; Figure 1D and Figure 1—figure supplement 6). Consistent with the lack of impact on visual specificity of the licking behavior, also the first lick latency and variability decreased with training in CT-ablated animals as in controls (Figure 1D and Figure 1—figure supplement 6). Thus, the ablation cortico-fugal neurons that project to the SC has no effect on the acquisition of the task.

Figure 2 with 1 supplement see all

Download asset Open asset

Figure 2—source data 1

Behavioral performance measurements as a function of training days during task acquisition for VC-lesioned animals and comparison of behavioral metrics across experimental groups.

https://cdn.elifesciences.org/articles/59247/elife-59247-fig2-data1-v2.xlsx

Download elife-59247-fig2-data1-v2.xlsx

To determine whether neurons that project to the dmSt affect the ability to learn the detection task, we selectively ablated them using the same intersectional approach used for ablating CT neurons, this time however, by injecting the retroAAV-Cre in the dmSt. The viral ablation of Cre recombinase expressing CSt neurons in VC was also highly efficient with over 90% of ablated cells 3 weeks after the injection, as verified histologically (retroAAV-Cre: 20 ± 6 cells/100 µm³ CSt-ablation: 1 ± 1 cells/100 µm³ in layer 5; mean ± SD; N = 3 mice; p<0.001 Mann-Whitney U test; Figure 1—figure supplement 5). In striking contrast to control and CT-ablated mice, in CSt-ablated mice the visual specificity of the licking behavior increased much slower (aROC/day CSt-ablation: 0.014 ± 0.003 vs. retroAAV-Cre only: 0.03 ± 0.004; mean ± SEM; p<0.05; Mann-Whitney U test; population average CSt-ablation slope aROC/day: 0.014 ± 0.0018 vs. retroAAV-Cre only slope aROC/day: 0.026 ± 0.001; mean ±RMSE; Fisher z-test; p<0.001; CSt-ablation slope aROC/day: 0.014 ± 0.0018 vs. CT-ablation slope aROC/day: 0.029 ± 0.0014; mean ±RMSE; Fisher z-test; p<0.001; Figure 1E and Figure 2D). As a consequence, by day 14 of training, CSt-ablated mice had an aROC of only 0.68 ± 0.04 compared with 0.84 ± 0.04 in control animals in which only the retroAAV-Cre had been injected (p<0.001 Mann-Whitney U test; Figure 1E and Figure 1—figure supplement 6). Furthermore, both the latency and trial-to-trial variability of the first licks on stimulus trials decreased slower in CSt-lesioned as compared to retroAAV-Cre control animals (Figure 1E and Figure 1—figure supplement 6B). The spontaneous lick frequency was similar between control and CSt-ablated mice across the 2 weeks of training (inter-lick-interval: intact: 1.29 ± 0.03 s; retroAAV-Cre control: 1.27 ± 0.03 s vs. 1.40 ± 0.04 s mean ± SEM; n.s.; Mann-Whitney U test, Figure 2D). In addition, the impairments in task learning was not due to a decrease in the number of trials the animals performed during the training period neither to a reduced spontaneous lick frequency because these behavioral variables were similar in CSt-ablated and control mice (number of trials per training session: intact: 242.5 ± 3.2; retroAAV-Cre only: 232.8 ± 2.6 vs. CSt-ablation: 237.6 ± 5.3; mean ± SEM, n.s. Mann-Whitney U test; Figure 2D). Thus, ablation of cortico-fugal neurons that project to the dmSt led to slower learning such that by 2 weeks of training, learning was still incomplete.

The above result suggests that CSt neurons specifically contribute to the learning speed of the animal. Alternatively, CSt neurons may be required for learning itself and slow learning in CSt-ablated animals could simply result from an incomplete ablation of neurons projecting to the dmSt. To completely eliminate the CSt neurons, we surgically removed the entire VC, that is, primary VC and the surrounding higher visual areas contralateral to the stimulus 10 days before starting behavioral training (in three out of eight animals both ipsi and contralateral VC were ablated; Figure 2 and Figure 2—figure supplement 1). Both the spontaneous lick frequency and the number of performed trials per session were similar between control and VC-lesioned mice (inter lick time interval day 1: control: 0.95 ± 0.14 s; vs. VC lesion: 0.99 ± 1.47 s; mean ± SEM; number of trials per training session: control 244.7 ± 4.0 vs. VC lesion 239.3 ± 9.9 mean ± SEM; n.s. Mann-Whitney U test; Figure 2D). Like in CSt-ablated animals, VC lesions led to slower learning (population average VC lesion slope aROC/day: 0.014 ± 0.0014 vs. CSt ablation slope aROC/day: 0.014 ± 0.0018; mean ±RMSE; n.s.; Fisher z-test) as compared to intact animals (VC lesion slope aROC/day: 0.014 ± 0.0014 vs. control slope aROC/day: 0.025 ± 0.0014; mean ±RMSE; p<0.001 Fisher z-test) such that by 2 weeks of training none of the VC-lesioned mice reached criterion (VC lesion: aROC of 0.60 ± 0.03 vs. intact: 0.86 ± 0.04; mean ± SEM; p=0.0040; Mann-Whitney U test; Figure 2). Thus, VC lesions recapitulate the effect of CSt neuron ablation. These results indicate that ablation of CSt neurons impairs learning speed.

If ablation of CSt neurons selectively impairs learning speed without limiting the animal’s ability to learn, it should be possible, with additional training, for CSt-ablated animals to eventually reach the same visual specificity as control animals. We thus determined the impact of additional training on animals in which, before the onset of training, we either ablated CSt neurons or lesioned VC and compared their performance with control animals. As described above, by the 14th day of training, intact animals had reached the learning criterion while CSt-ablated and VC-ablated animals were still below criterion. Importantly, by the 14th day of training, intact animals had reached plateau because further training did not improve their performance (slope aROC/day 2.46 × 10⁻⁴ ± 0.021; mean ±RMSE; R² = 9.51 × 10⁻⁴, n.s.; F-test; population average linear fit from day 14 to day 21; Figure 3A). In striking contrast to the plateau performance of intact animals, both CSt-ablated and VC-lesioned animals continued to improve in the visual specificity of their licking behavior with additional training (CSt-ablation lesion slope aROC/day 0.033 ± 0.054, mean ±RMSE; R² = 0.83, F-test; p<0.001; VC lesion slope aROC/day 0.022 ± 0.063, mean ±RMSE; R² = 0.78, F-test; p<0.001; Figure 3B). By the end of the third week, CSt-ablated and VC-ablated animals reached an aROC of 0.87 ± 0.02 and 0.76 ± 0.05, hence similar to the plateau levels of control animals (intact aROC: 0.84 ± 0.03 vs. VC-lesion: 0.76 ± 0.05; n.s.; retroAAV-Cre only: 0.81 ± 0.05 vs. CSt-ablation: 0.87 ± 0.02 n.s., Mann-Whitney U test; Figure 3B).

Figure 3

Download asset Open asset

Figure 3—source data 1

Behavioral performance measurements throughout the third week training across experimental groups.

https://cdn.elifesciences.org/articles/59247/elife-59247-fig3-data1-v2.xlsx

Download elife-59247-fig3-data1-v2.xlsx

If ablation of CSt neurons before task acquisition specifically impairs learning speed, ablation of CSt neurons in animals that master the task, should not affect performance. To address this question, we completely eliminated CSt neurons in VC, as above, by surgically removing VC in mice who had learned the detection task and tested them after a 10 days training gap following the surgery. Strikingly, mice that had learned the task before VC lesion maintained a high visual specificity of the licking behavior even after VC lesions (aROC pre-lesion: 0.92 ± 0.02 vs. aROC post-lesion: 0.91 ± 0.08, mean ± SEM; N = 5 mice, n.s. Wilcoxon signed-rank test; Figure 4A). VC-lesioned animals also showed no change in first lick latency (pre-lesion: 0.58 ± 0.10 s vs. post-Lesion: 0.45 ± 0.04; mean ± SEM; n.s. Wilcoxon signed-rank test) and in its variability (pre-lesion:0.54 ± 0.22 s vs. post-lesion:0.49 ± 0.24 s; mean ± SEM; n.s. Wilcoxon signed-rank test; Figure 4—figure supplement 1). Thus, once learning has occurred, CSt neurons are no longer required for the animal’s ability to execute the task.

Figure 4 with 1 supplement see all

Download asset Open asset

Figure 4—source data 1

Behavioral performance measurements as a function of VC lesion.

https://cdn.elifesciences.org/articles/59247/elife-59247-fig4-data1-v2.xlsx

Download elife-59247-fig4-data1-v2.xlsx

What is the function of the CT neurons in the detection task? While CT neurons are not involved in learning, they may play a role in stimulus detection, given the role of the SC, the main target of CT neurons, in detecting visual stimuli that elicit innate behavior (Evans et al., 2018; Liang et al., 2015; Shang et al., 2015; Shang et al., 2018). To assess the animal’s sensitivity in detecting visual stimuli, we tested their performance by presenting full field visual stimuli at various contrasts. To be able to detect potentially subtle changes in performance across a large range of stimulus contrasts, we opted to acutely and reversibly silence VC using an optogenetic approach, thereby obtaining trials with and without VC silencing within the same session. Using this approach, we compared the impact of VC silencing on behavior in the same animals before and after ablation of CT neurons (Figure 5). Trained animals were tested on stimuli presented at seven different contrasts (0%, 4%, 8%, 16%, 32%, 64%, 100%) to obtain a psychometric function of visual sensitivity. On a third of the trials, VC was silenced by optogenetically activating GABAergic neurons expressing Channelrhodopsin 2, as described previously (Lien and Scanziani, 2013; Olsen et al., 2012). Consistent with the above results in which we surgically removed VC, optogenetic silencing of VC did not impair performance in response to stimuli presented at full contrast (aROC control: 0.88 ± 0.02 vs. aROC VC Silencing: 0.90 ± 0.04; mean ± SEM; N = 6 mice, n.s. Wilcoxon signed rank test; Figure 5A). However, at lower contrasts, silencing VC significantly reduced performance leading to a rightward shift of the psychometric function and a corresponding increase in the contrast threshold for detection (threshold contrast: control 7.8 ± 1% vs. VC silencing: 28 ± 3%, Weibull fit; see Materials and methods, p<0.01 Wilcoxon signed rank test; Figure 5A). Thus, VC modulates sensitivity by lowering the contrast threshold for stimulus detection. To determine whether VC modulates detection sensitivity via cortico-fugal neurons that target the SC, we ablated CT neurons using the same intersectional strategy described above. We noted that the animals did not exhibit the same level of performance at maximal stimulus contrast under control conditions before and after CT-ablation (contrast 100%: control pre CT-ablation aROC 0.88 ± 0.01 vs. control post CT-ablation aROC 0.83 ± 0.02; N = 3 mice; p<0.05; Mann-Whitney U test; Figure 5B), possibly because of the 3-week long training gap between the two conditions. Therefore, we assessed the role of CT-ablation by comparing the relative impact of VC silencing before and after the CT-ablation on the contrast threshold within the same test session. Strikingly, upon ablation of CT neurons the acute silencing of VC no longer lead to an increase in the detection threshold (pre CT-ablation: threshold contrast: control 13.9 ± 3.3% vs. VC silencing: 35.4 ± 5.4%, p<0.05; post CT-ablation: threshold contrast: control 25.8 ± 9.5% vs. VC silencing: 31.2 ± 15.8%, N = 3 mice; n.s. Wilcoxon signed rank test; Figure 5B). These data indicate that cortico-fugal neurons projecting to the SC, while not involved in learning, increase detection sensitivity and thus enhances an animal’s ability to detect less salient stimuli.

Figure 5

Download asset Open asset

Figure 5—source data 1

Behavioral performance measurements as a function of optogenetic silencing of VC and CT ablation.

https://cdn.elifesciences.org/articles/59247/elife-59247-fig5-data1-v2.xlsx

Download elife-59247-fig5-data1-v2.xlsx

Do CT neurons maintain their role on detection sensitivity or, like the CSt neurons, eventually become dispensable for performance? To address this question, we compared the impact of VC silencing on detection sensitivity as a function of training duration. Interestingly, while animals trained 2–7 weeks (31 ± 12 days median ± SD; N = 6 mice) showed a strong rightward shift in the psychometric function and an increase in detection threshold of approximately 3.4-fold upon VC silencing (threshold contrast: control 7.75 ± 1.8%, VC silencing: 26.6 ± 5.0%, p<0.01; Wilcoxon signed rank test; N = 6 mice; Figure 6A), detection sensitivity of animals trained 10–17 weeks (93 ± 9 days, median ± SD; N = 4 mice) was much less affected by VC silencing as the detection threshold upon VC silencing increased only 1.5 times (threshold contrast: control 10.02 ± 2.1%, VC silencing.: 15.5 ± 3.7%, n.s.; Wilcoxon signed rank test; Figure 6B). Thus, the impact of VC on detection sensitivity diminishes with training (Pearson correlation coefficient R = −0.612; p=0.013; R² = 0.37; N = 16 mice; Figure 6C).

Figure 6

Download asset Open asset

Figure 6—source data 1

Measurements of the cortical impact on psychometric data as a function of prolonged training and on neuronal activity in the Superior Colliculus.

https://cdn.elifesciences.org/articles/59247/elife-59247-fig6-data1-v2.xlsx

Download elife-59247-fig6-data1-v2.xlsx

Given that VC modulates detection sensitivity via CT neurons (see above), we tested the impact of VC on visual responses in SC as a function of training duration. To this end, we recorded neuronal activity across the depth of SC using linear extracellular probes as the animal’s performed the task. Early in training, VC silencing robustly reduced visual responses in the cortical recipient layers of SC, namely the stratum opticum and stratum griseum intermediale, and less so in the retino-recipient layer (Figure 6D). Later in training, visual responses across SC layers were less dependent on VC input (Figure 6E,F). Thus, the impact of the CT neurons both on detection sensitivity and on visual-evoked activity in the SC diminishes progressively with prolonged training.

The mammalian sensory cortex is an important stage in the representation of sensory stimuli, yet its contribution to the learning and performance in goal-directed behaviors is still controversial and may depend on whether the task involves detection or discrimination, and on the sensory modality required for the task (Guo et al., 2014; Hutson and Masterton, 1986; Lashley, 1931; Miyashita and Feldman, 2013; Talwar et al., 2001). Our results show that VC contributes to two select behavioral components of a simple visual detection task through two distinct populations of cortico-fugal neurons. On the one hand, VC, via the CSt neurons, plays a crucial role in the speed at which an animal learns to report the presence of visual stimuli. On the other hand, once the task has been learned VC modulates detection sensitivity via the CT neurons. Thus, this study reveals the functional dissociation of two major populations of cortico-fugal neurons in VC during specific stages in the learning and performance of a simple goal-directed behavior. Furthermore, eventually, both populations of cortico-fugal neurons become dispensable for the task.

Neurons projecting to the dmSt play their main role during the acquisition phase of the detection task as the ablation of CSt neurons impairs the learning speed but not the animal’s ability to detect salient visual stimuli once the task has been learned. This is consistent with the role of the dmSt in action selection/initiation and reinforcement learning of goal-directed behavior (Cox and Witten, 2019; Klaus et al., 2019). The finding that VC via CSt neurons plays a preferential role during learning is reminiscent of the necessity of motor cortex in the learning but not for the performance of some motor tasks (Kawai et al., 2015). Given that the elimination of CSt neurons once the task has been learned does not affect performance, CSt neurons may function as a tutor to mediate learning-related plasticity in subcortical circuits that underlie the performance of the task. It will be interesting to address which properties of the subcortical circuitry that are not dependent on the CSt pathway once learning is achieved, are affected by the CSt pathway during learning. Our data suggest that the plastic events that occur during learning, while relying on the CSt input, do not reside at the CSt synapse. Otherwise removing VC after learning would impair performance. Given the existence of multiple subcortical loops between the SC and the basal ganglia (Chevalier and Deniau, 1984; Harting et al., 2001; Krauzlis et al., 2013; Krout et al., 2001; Lin et al., 1984; McHaffie et al., 2005; Takada et al., 1985), one could speculate that CSt neurons modify a SC-basal ganglia loop during learning. We cannot exclude that the ablation of CSt neurons also affects detection sensitivity, similar to the ablation of CT neurons. However, this possibility seems unlikely because silencing the entire VC (thus also including CSt neurons) following the ablation of CT neurons has little effect on the psychometric function.

The ability of neurons in sensory cortex to encode specific spatial and temporal features of the stimulus may likely be the neural basis for the role of VC (Glickfeld et al., 2013; Jurjut et al., 2017; Marques et al., 2018; Petruno et al., 2013; Poort et al., 2015; Resulaj et al., 2018) and in particular of CSt neurons (Xiong et al., 2015; Znamenskiy and Zador, 2013) in feature discrimination tasks. Also for the learning of the detection task, like the one used in this study, CSt neurons may convey an instructive signal about the spatio-temporal properties of the visual stimulus. Alternatively, for a detection task, CSt neurons may simply provide a permissive, and possibly unspecific, drive to subcortical structures necessary for animal’s ability to form an association between a stimulus, an action and the outcome.

Contrary to the ablation of CSt neurons, the ablation of neurons that project to the SC did not not impact learning, at least when assessing learning with stimuli at 100% contrast. We cannot exclude that if learning had been assessed with lower contrast stimuli, that ablation of CT neurons could have revealed impairments in other aspects of learning, like perceptual learning (Hua et al., 2010; Lu et al., 2011; Yan et al., 2014). Ablation of CT neurons, however, impaired detection sensitivity. This result is consistent with the impact of cortical silencing on the sensory threshold for visually guided innate behavior mediated by the SC (Liang et al., 2015) and the involvement of the SC in simple orienting behaviors including voluntary licking (Rossi et al., 2016). Thus, CT neurons may represent the neural substrate underlying reported impairments in the perception of low-contrast stimuli following VC lesions (Cowey and Stoerig, 1995; Glickfeld et al., 2013; Pöppel et al., 1973; Weiskrantz et al., 1974). In humans and primates, this phenomenon is called blindsight and, interestingly, relies on the SC (Cowey and Stoerig, 1995; Kinoshita et al., 2019; Mohler and Wurtz, 1977), the main target of CT neurons. We show that VC enhances neuronal activity in SC in response to visual stimuli, which may account, at least in part, for the deficits observed following VC lesions. With prolonged training, VC gradually loses its impact on behavior and on visual responses in SC, suggesting the existence of plasticity at the CT synapse or local plasticity within the SC.

The ablation of a specific population of cortico-fugal neurons may have effects on the connectivity or function of the remaining network in the VC. However, given that cortical neurons with distinct projection targets have been shown to form segregated subnetworks (Brown and Hestrin, 2009; Harris and Mrsic-Flogel, 2013; Kim et al., 2018; Lur et al., 2016; Zhang et al., 2016), we believe that the connectivity and function of the spared neurons is largely unaffected. In the future, optogenetic or pharmacogenetic approaches to transiently perturb the activity of cortico-fugal neurons may function as a complementary approach to mitigate any putative long-term compensatory changes in cortex. Ironically, however, even the impact of acute perturbations on behavior can be difficult to interpret specifically because of the lack of compensation in the downstream targets (Otchy et al., 2015).

Which structures may mediate learning, albeit slow, in the absence of VC? Given that, following VC lesions, the SC is the main visual processing stage in the brain and given the role of the cortico-fugal pathway targeting the dmSt in learning, we hypothesize that a subcortical SC-basal ganglia loop mediates the slower, VC-independent learning. Along these lines, recent findings indicate that a somatosensory detection task can be learned to proficiency in the absence of somatosensory cortex (Hong et al., 2018), possibly involving subcortical basal ganglia loops (Bosman et al., 2011; Redgrave et al., 2010). The fact that, with prolonged training, the behavior becomes independent of VC implies that the underlying subcortical structures can entirely rely on ascending sensory input from the periphery rather than on descending cortico-fugal pathways.

The gradual reduction in the role of VC on behavior may, furthermore, reflect a transition from goal-directed to habitual behavior (Balleine, 2019; Dickinson, 1994). On one hand, VC, through its cortico-fugal projection to the dmSt, may facilitate behavioral flexibility in an ever-changing environment during goal-directed behavior. On the other hand, the subcortical sources of visual input to the basal ganglia may be sufficient to drive habitual responses upon prolonged training in stable contingencies.

In summary, two distinct populations of cortico-fugal neurons in the VC that give rise to major descending projections targeting phylogenetically older subcortical structures, substantially contribute to visually guided behavior of a simple detection task, by increasing learning speed and improving detection sensitivity. These results establish a causal relationship between anatomically defined classes of cortico-fugal neurons and the emergence of a visually guided behavior and highlight the selective role of distinct classes of cortico-fugal neurons in specific aspects of simple goal-directed behavior. The improvement in learning and performance by defined classes of cortico-fugal neurons may represent an early adaptive benefit to goal-directed behavior associated with the expansion of cortex in mammals.

All data generated or analyzed during this study are included in the manuscript and supporting files. Numerical data for graphs represented in figures 1-6, figure 1-figure supplement 2,3,4,5, figure 2-figure supplement 1, figure 4-figure supplement 1 are provided as source data files. The software used to generate visual stimuli and record neuronal activity is available at: https://github.com/mscaudill/neuroGit and https://github.com/aresulaj/ResRueOlsSca18.

1. 1. Balleine BW

   (2019) The meaning of behavior: discriminating reflex and volition in the brain

   Neuron 104:47–62.

   https://doi.org/10.1016/j.neuron.2019.09.024

   • PubMed
   • Google Scholar
2. 1. Bosman LW
   2. Houweling AR
   3. Owens CB
   4. Tanke N
   5. Shevchouk OT
   6. Rahmati N
   7. Teunissen WH
   8. Ju C
   9. Gong W
   10. Koekkoek SK
   11. De Zeeuw CI

   (2011) Anatomical pathways involved in generating and sensing rhythmic whisker movements

   Frontiers in Integrative Neuroscience 5:53.

   https://doi.org/10.3389/fnint.2011.00053

   • PubMed
   • Google Scholar
3. 1. Brown SP
   2. Hestrin S

   (2009) Intracortical circuits of pyramidal neurons reflect their long-range axonal targets

   Nature 457:1133–1136.

   https://doi.org/10.1038/nature07658

   • PubMed
   • Google Scholar
4. 1. Ceballo S
   2. Piwkowska Z
   3. Bourg J
   4. Daret A
   5. Bathellier B

   (2019) Targeted Cortical Manipulation of Auditory Perception

   Neuron 104:1168–1179.

   https://doi.org/10.1016/j.neuron.2019.09.043

   • Google Scholar
5. 1. Chevalier G
   2. Deniau JM

   (1984) Spatio-temporal organization of a branched tecto-spinal/ tecto-diencephalic neuronal system

   Neuroscience 12:427–439.

   https://doi.org/10.1016/0306-4522(84)90063-0

   • Google Scholar
6. 1. Cowey A
   2. Stoerig P

   (1995) Blindsight in monkeys

   Nature 373:247–249.

   https://doi.org/10.1038/373247a0

   • Google Scholar
7. 1. Cox J
   2. Witten IB

   (2019) Striatal circuits for reward learning and decision-making

   Nature Reviews Neuroscience 20:482–494.

   https://doi.org/10.1038/s41583-019-0189-2

   • PubMed
   • Google Scholar
8. 1. Dalmay T
   2. Abs E
   3. Poorthuis RB
   4. Hartung J
   5. Pu DL
   6. Onasch S
   7. Lozano YR
   8. Signoret-Genest J
   9. Tovote P
   10. Gjorgjieva J
   11. Letzkus JJ

   (2019) A critical role for neocortical processing of threat memory

   Neuron 104:1180–1194.

   https://doi.org/10.1016/j.neuron.2019.09.025

   • PubMed
   • Google Scholar
9. Book

   1. Dickinson A

   (1994) Instrumental Conditioning

   In: Mackintosh N, editors. Animal Learning and Cognition. Academic Press. pp. 4–79.

   https://doi.org/10.1016/C2009-0-02316-2

   • Google Scholar
10. 1. Evans DA
    2. Stempel AV
    3. Vale R
    4. Ruehle S
    5. Lefler Y
    6. Branco T

    (2018) A synaptic threshold mechanism for computing escape decisions

    Nature 558:590–594.

    https://doi.org/10.1038/s41586-018-0244-6

    • PubMed
    • Google Scholar
11. 1. Faull RLM
    2. Nauta WJH
    3. Domesick VB

    (1986) The visual cortico-striato-nigral pathway in the rat

    Neuroscience 19:1119–1132.

    https://doi.org/10.1016/0306-4522(86)90128-4

    • Google Scholar
12. 1. Fawcett T

    (2006) An introduction to ROC analysis

    Pattern Recognition Letters 27:861–874.

    https://doi.org/10.1016/j.patrec.2005.10.010

    • Google Scholar
13. Book

    1. Feinberg TE
    2. Mallatt JM

    (2019)

    The Ancient Origins of Consciousness

    The MIT Press.

    • Google Scholar
14. 1. Glickfeld LL
    2. Histed MH
    3. Maunsell JH

    (2013) Mouse primary visual cortex is used to detect both orientation and contrast changes

    Journal of Neuroscience 33:19416–19422.

    https://doi.org/10.1523/JNEUROSCI.3560-13.2013

    • PubMed
    • Google Scholar
15. 1. Gray DC
    2. Mahrus S
    3. Wells JA

    (2010) Activation of specific apoptotic caspases with an engineered small-molecule-activated protease

    Cell 142:637–646.

    https://doi.org/10.1016/j.cell.2010.07.014

    • PubMed
    • Google Scholar
16. 1. Grillner S
    2. El Manira A

    (2020) Current principles of motor control, with special reference to vertebrate locomotion

    Physiological Reviews 100:271–320.

    https://doi.org/10.1152/physrev.00015.2019

    • PubMed
    • Google Scholar
17. 1. Guo ZV
    2. Li N
    3. Huber D
    4. Ophir E
    5. Gutnisky D
    6. Ting JT
    7. Feng G
    8. Svoboda K

    (2014) Flow of cortical activity underlying a tactile decision in mice

    Neuron 81:179–194.

    https://doi.org/10.1016/j.neuron.2013.10.020

    • PubMed
    • Google Scholar
18. 1. Harris KD
    2. Mrsic-Flogel TD

    (2013) Cortical connectivity and sensory coding

    Nature 503:51–58.

    https://doi.org/10.1038/nature12654

    • PubMed
    • Google Scholar
19. 1. Harris KD
    2. Shepherd GM

    (2015) The neocortical circuit: themes and variations

    Nature Neuroscience 18:170–181.

    https://doi.org/10.1038/nn.3917

    • PubMed
    • Google Scholar
20. 1. Harting JK
    2. Updyke BV
    3. Van Lieshout DP

    (2001) The visual-oculomotor striatum of the cat: functional relationship to the superior colliculus

    Experimental Brain Research 136:138–142.

    https://doi.org/10.1007/s002210000606

    • PubMed
    • Google Scholar
21. 1. Hattox AM
    2. Nelson SB

    (2007) Layer V neurons in mouse cortex projecting to different targets have distinct physiological properties

    Journal of Neurophysiology 98:3330–3340.

    https://doi.org/10.1152/jn.00397.2007

    • PubMed
    • Google Scholar
22. 1. Hong YK
    2. Lacefield CO
    3. Rodgers CC
    4. Bruno RM

    (2018) Sensation, movement and learning in the absence of barrel cortex

    Nature 561:542–546.

    https://doi.org/10.1038/s41586-018-0527-y

    • PubMed
    • Google Scholar
23. 1. Hua T
    2. Bao P
    3. Huang C-B
    4. Wang Z
    5. Xu J
    6. Zhou Y
    7. Lu Z-L

    (2010) Perceptual Learning Improves Contrast Sensitivity of V1 Neurons in Cats

    Current Biology 20:887–894.

    https://doi.org/10.1016/j.cub.2010.03.066

    • Google Scholar
24. 1. Hutson KA
    2. Masterton RB

    (1986) The sensory contribution of a single vibrissa's cortical barrel

    Journal of Neurophysiology 56:1196–1223.

    https://doi.org/10.1152/jn.1986.56.4.1196

    • PubMed
    • Google Scholar
25. 1. Jones EG

    (1984) Cerebral cortex. vol. 1. cellular components of the cerebral cortex

    Cerebral Cortex 140:1–32.

    https://doi.org/10.1113/expphysiol.1985.sp002910

    • Google Scholar
26. 1. Jurjut O
    2. Georgieva P
    3. Busse L
    4. Katzner S

    (2017) Learning enhances sensory processing in mouse V1 before improving behavior

    The Journal of Neuroscience 37:6460–6474.

    https://doi.org/10.1523/JNEUROSCI.3485-16.2017

    • PubMed
    • Google Scholar
27. 1. Kawai R
    2. Markman T
    3. Poddar R
    4. Ko R
    5. Fantana AL
    6. Dhawale AK
    7. Kampff AR
    8. Ölveczky BP

    (2015) Motor cortex is required for learning but not for executing a motor skill

    Neuron 86:800–812.

    https://doi.org/10.1016/j.neuron.2015.03.024

    • PubMed
    • Google Scholar
28. 1. Kemp JM
    2. Powell TP

    (1970) The cortico-striate projection in the monkey

    Brain 93:525–546.

    https://doi.org/10.1093/brain/93.3.525

    • PubMed
    • Google Scholar
29. 1. Khibnik LA
    2. Tritsch NX
    3. Sabatini BL

    (2014) A direct projection from mouse primary visual cortex to dorsomedial striatum

    PLOS ONE 9:e104501.

    https://doi.org/10.1371/journal.pone.0104501

    • PubMed
    • Google Scholar
30. 1. Kim MH
    2. Znamenskiy P
    3. Iacaruso MF
    4. Mrsic-Flogel TD

    (2018) Segregated subnetworks of intracortical projection neurons in primary visual cortex

    Neuron 100:1313–1321.

    https://doi.org/10.1016/j.neuron.2018.10.023

    • PubMed
    • Google Scholar
31. Book

    1. Kingdom FAA
    2. Prins N

    (2010)

    Psychophysics: A Practical Introduction

    Academic Press.

    • Google Scholar
32. 1. Kinoshita M
    2. Kato R
    3. Isa K
    4. Kobayashi K
    5. Kobayashi K
    6. Onoe H
    7. Isa T

    (2019) Dissecting the circuit for blindsight to reveal the critical role of pulvinar and superior colliculus

    Nature Communications 10:8058.

    https://doi.org/10.1038/s41467-018-08058-0

    • Google Scholar
33. 1. Klaus A
    2. Alves da Silva J
    3. Costa RM

    (2019) What, if, and when to move: basal ganglia circuits and Self-Paced action initiation

    Annual Review of Neuroscience 42:459–483.

    https://doi.org/10.1146/annurev-neuro-072116-031033

    • PubMed
    • Google Scholar
34. 1. Krauzlis RJ
    2. Lovejoy LP
    3. Zénon A

    (2013) Superior colliculus and visual spatial attention

    Annual Review of Neuroscience 36:165–182.

    https://doi.org/10.1146/annurev-neuro-062012-170249

    • PubMed
    • Google Scholar
35. 1. Krout KE
    2. Loewy AD
    3. Westby GW
    4. Redgrave P

    (2001) Superior colliculus projections to midline and intralaminar thalamic nuclei of the rat

    The Journal of Comparative Neurology 431:198–216.

    https://doi.org/10.1002/1096-9861(20010305)431:2<198::AID-CNE1065>3.0.CO;2-8

    • PubMed
    • Google Scholar
36. 1. Lashley KS

    (1931) The mechanism of vision IV. the cerebral Areas necessary for pattern vision in the rat

    The Journal of Comparative Neurology 53:419–478.

    https://doi.org/10.1002/cne.900530304

    • Google Scholar
37. 1. Liang F
    2. Xiong XR
    3. Zingg B
    4. Ji XY
    5. Zhang LI
    6. Tao HW

    (2015) Sensory cortical control of a visually induced arrest behavior via corticotectal projections

    Neuron 86:755–767.

    https://doi.org/10.1016/j.neuron.2015.03.048

    • PubMed
    • Google Scholar
38. 1. Lien AD
    2. Scanziani M

    (2013) Tuned thalamic excitation is amplified by visual cortical circuits

    Nature Neuroscience 16:1315–1323.

    https://doi.org/10.1038/nn.3488

    • PubMed
    • Google Scholar
39. 1. Lin CS
    2. May PJ
    3. Hall WC

    (1984) Nonintralaminar thalamostriatal projections in the gray squirrel (Sciurus carolinensis) and tree shrew (Tupaia glis)

    The Journal of Comparative Neurology 230:33–46.

    https://doi.org/10.1002/cne.902300104

    • PubMed
    • Google Scholar
40. Book

    1. Llinás R

    (1976) Frog Neurobiology

    Springer Science & Business Media.

    https://doi.org/10.1007/978-3-642-66316-1

    • Google Scholar
41. 1. Lu ZL
    2. Hua T
    3. Huang CB
    4. Zhou Y
    5. Dosher BA

    (2011) Visual perceptual learning

    Neurobiology of Learning and Memory 95:145–151.

    https://doi.org/10.1016/j.nlm.2010.09.010

    • PubMed
    • Google Scholar
42. 1. Luppi PH
    2. Fort P
    3. Jouvet M

    (1990) Iontophoretic application of unconjugated cholera toxin B subunit (CTb) combined with immunohistochemistry of neurochemical substances: a method for transmitter identification of retrogradely labeled neurons

    Brain Research 534:209–224.

    https://doi.org/10.1016/0006-8993(90)90131-t

    • PubMed
    • Google Scholar
43. 1. Lur G
    2. Vinck MA
    3. Tang L
    4. Cardin JA
    5. Higley MJ

    (2016) Projection-Specific visual feature encoding by layer 5 cortical subnetworks

    Cell Reports 14:2538–2545.

    https://doi.org/10.1016/j.celrep.2016.02.050

    • PubMed
    • Google Scholar
44. 1. Macé É
    2. Montaldo G
    3. Trenholm S
    4. Cowan C
    5. Brignall A
    6. Urban A
    7. Roska B

    (2018) Whole-Brain functional ultrasound imaging reveals brain modules for visuomotor integration

    Neuron 100:1241–1251.

    https://doi.org/10.1016/j.neuron.2018.11.031

    • PubMed
    • Google Scholar
45. Book

    1. Macmillan N
    2. Creelman CD

    (2005)

    Detection Theory: A User’s Guide

    Lawrence Erlbaum Associates.

    • Google Scholar
46. 1. Madisen L
    2. Garner AR
    3. Shimaoka D
    4. Chuong AS
    5. Klapoetke NC
    6. Li L
    7. van der Bourg A
    8. Niino Y
    9. Egolf L
    10. Monetti C
    11. Gu H
    12. Mills M
    13. Cheng A
    14. Tasic B
    15. Nguyen TN
    16. Sunkin SM
    17. Benucci A
    18. Nagy A
    19. Miyawaki A
    20. Helmchen F
    21. Empson RM
    22. Knöpfel T
    23. Boyden ES
    24. Reid RC
    25. Carandini M
    26. Zeng H

    (2015) Transgenic mice for intersectional targeting of neural sensors and effectors with high specificity and performance

    Neuron 85:942–958.

    https://doi.org/10.1016/j.neuron.2015.02.022

    • PubMed
    • Google Scholar
47. 1. Marques T
    2. Summers MT
    3. Fioreze G
    4. Fridman M
    5. Dias RF
    6. Feller MB
    7. Petreanu L

    (2018) A role for mouse primary visual cortex in motion perception

    Current Biology 28:1703–1713.

    https://doi.org/10.1016/j.cub.2018.04.012

    • PubMed
    • Google Scholar
48. 1. McHaffie JG
    2. Stanford TR
    3. Stein BE
    4. Coizet V
    5. Redgrave P

    (2005) Subcortical loops through the basal ganglia

    Trends in Neurosciences 28:401–407.

    https://doi.org/10.1016/j.tins.2005.06.006

    • PubMed
    • Google Scholar
49. 1. Miyashita T
    2. Feldman DE

    (2013) Behavioral detection of passive whisker stimuli requires somatosensory cortex

    Cerebral Cortex 23:1655–1662.

    https://doi.org/10.1093/cercor/bhs155

    • PubMed
    • Google Scholar
50. 1. Mohler CW
    2. Wurtz RH

    (1977) Role of striate cortex and superior colliculus in visual guidance of saccadic eye movements in monkeys

    Journal of Neurophysiology 40:74–94.

    https://doi.org/10.1152/jn.1977.40.1.74

    • PubMed
    • Google Scholar
51. Book

    1. Morris R
    2. Bornstein A
    3. Shenhav A

    (2018)

    Goal-Directed Decision Making: Computations and Neural Circuits

    Elsevier.

    • Google Scholar
52. 1. Norita M
    2. McHaffie JG
    3. Shimizu H
    4. Stein BE

    (1991) The corticostriatal and corticotectal projections of the feline lateral suprasylvian cortex demonstrated with anterograde biocytin and retrograde fluorescent techniques

    Neuroscience Research 10:149–155.

    https://doi.org/10.1016/0168-0102(91)90037-Y

    • PubMed
    • Google Scholar
53. Software

    1. Olen S

    (2020) ResRueOlsSca18, version 59a6772

    GitHub.

    https://github.com/aresulaj/ResRueOlsSca18
54. 1. Olsen SR
    2. Bortone DS
    3. Adesnik H
    4. Scanziani M

    (2012) Gain control by layer six in cortical circuits of vision

    Nature 483:47–52.

    https://doi.org/10.1038/nature10835

    • PubMed
    • Google Scholar
55. 1. Otchy TM
    2. Wolff SB
    3. Rhee JY
    4. Pehlevan C
    5. Kawai R
    6. Kempf A
    7. Gobes SM
    8. Ölveczky BP

    (2015) Acute off-target effects of neural circuit manipulations

    Nature 528:358–363.

    https://doi.org/10.1038/nature16442

    • PubMed
    • Google Scholar
56. Book

    1. Paxinos G
    2. Franklin K

    (2008)

    The Mouse Brain in Stereotaxic Coordinates, Compact

    Academic Press.

    • Google Scholar
57. 1. Petruno SK
    2. Clark RE
    3. Reinagel P

    (2013) Evidence that primary visual cortex is required for image, orientation, and motion discrimination by rats

    PLOS ONE 8:e56543.

    https://doi.org/10.1371/journal.pone.0056543

    • PubMed
    • Google Scholar
58. 1. Poort J
    2. Khan AG
    3. Pachitariu M
    4. Nemri A
    5. Orsolic I
    6. Krupic J
    7. Bauza M
    8. Sahani M
    9. Keller GB
    10. Mrsic-Flogel TD
    11. Hofer SB

    (2015) Learning enhances sensory and multiple Non-sensory representations in primary visual cortex

    Neuron 86:1478–1490.

    https://doi.org/10.1016/j.neuron.2015.05.037

    • PubMed
    • Google Scholar
59. 1. Pöppel E
    2. Held R
    3. Frost D

    (1973) Residual Visual Function after Brain Wounds involving the Central Visual Pathways in Man

    Nature 243:295–296.

    https://doi.org/10.1038/243295a0

    • Google Scholar
60. Book

    1. Popper A
    2. Fay R

    (1992) The Mammalian Auditory Pathway: Neurophysiology

    Springer.

    https://doi.org/10.1007/978-1-4612-2838-7

    • Google Scholar
61. 1. Prusky GT
    2. Douglas RM

    (2004) Characterization of mouse cortical spatial vision

    Vision Research 44:3411–3418.

    https://doi.org/10.1016/j.visres.2004.09.001

    • PubMed
    • Google Scholar
62. 1. Redgrave P
    2. Coizet V
    3. Comoli E
    4. McHaffie JG
    5. Leriche M
    6. Vautrelle N
    7. Hayes LM
    8. Overton P

    (2010) Interactions between the midbrain superior colliculus and the basal ganglia

    Frontiers in Neuroanatomy 4:132.

    https://doi.org/10.3389/fnana.2010.00132

    • PubMed
    • Google Scholar
63. 1. Reinhard K
    2. Li C
    3. Do Q
    4. Burke EG
    5. Heynderickx S
    6. Farrow K

    (2019) A projection specific logic to sampling visual inputs in mouse superior colliculus

    eLife 8:e50697.

    https://doi.org/10.7554/eLife.50697

    • PubMed
    • Google Scholar
64. 1. Resulaj A
    2. Ruediger S
    3. Olsen SR
    4. Scanziani M

    (2018) First spikes in visual cortex enable perceptual discrimination

    eLife 7:e34044.

    https://doi.org/10.7554/eLife.34044

    • PubMed
    • Google Scholar
65. 1. Rhoades RW
    2. Mooney RD
    3. Fish SE

    (1985) Subcortical projections of area 17 in the anophthalmic mouse

    Developmental Brain Research 17:171–181.

    https://doi.org/10.1016/0165-3806(85)90141-5

    • Google Scholar
66. 1. Rossi MA
    2. Li HE
    3. Lu D
    4. Kim IH
    5. Bartholomew RA
    6. Gaidis E
    7. Barter JW
    8. Kim N
    9. Cai MT
    10. Soderling SH
    11. Yin HH

    (2016) A GABAergic nigrotectal pathway for coordination of drinking behavior

    Nature Neuroscience 19:742–748.

    https://doi.org/10.1038/nn.4285

    • PubMed
    • Google Scholar
67. Software

    1. Ruediger S

    (2020) neuroGit - A suite of Matlab packages for visual neuroscience, version adfd43a

    GitHub.

    https://github.com/mscaudill/neuroGit
68. 1. Saint-Cyr JA
    2. Ungerleider LG
    3. Desimone R

    (1990) Organization of visual cortical inputs to the striatum and subsequent outputs to the pallido-nigral complex in the monkey

    The Journal of Comparative Neurology 298:129–156.

    https://doi.org/10.1002/cne.902980202

    • PubMed
    • Google Scholar
69. 1. Schultz W

    (2006) Behavioral theories and the neurophysiology of reward

    Annual Review of Psychology 57:87–115.

    https://doi.org/10.1146/annurev.psych.56.091103.070229

    • PubMed
    • Google Scholar
70. 1. Seabrook TA
    2. Burbridge TJ
    3. Crair MC
    4. Huberman AD

    (2017) Architecture, function, and assembly of the mouse visual system

    Annual Review of Neuroscience 40:499–538.

    https://doi.org/10.1146/annurev-neuro-071714-033842

    • PubMed
    • Google Scholar
71. 1. Serizawa M
    2. McHaffie JG
    3. Hoshino K
    4. Norita M

    (1994) Corticostriatal and corticotectal projections from visual cortical Areas 17, 18 and 18a in the pigmented rat

    Archives of Histology and Cytology 57:493–507.

    https://doi.org/10.1679/aohc.57.493

    • PubMed
    • Google Scholar
72. 1. Shang C
    2. Liu Z
    3. Chen Z
    4. Shi Y
    5. Wang Q
    6. Liu S
    7. Li D
    8. Cao P

    (2015) BRAIN CIRCUITS. A parvalbumin-positive excitatory visual pathway to trigger fear responses in mice

    Science 348:1472–1477.

    https://doi.org/10.1126/science.aaa8694

    • PubMed
    • Google Scholar
73. 1. Shang C
    2. Chen Z
    3. Liu A
    4. Li Y
    5. Zhang J
    6. Qu B
    7. Yan F
    8. Zhang Y
    9. Liu W
    10. Liu Z
    11. Guo X
    12. Li D
    13. Wang Y
    14. Cao P

    (2018) Divergent midbrain circuits orchestrate escape and freezing responses to looming stimuli in mice

    Nature Communications 9:7.

    https://doi.org/10.1038/s41467-018-03580-7

    • Google Scholar
74. 1. Skinner BF

    (1963) Operant behavior

    American Psychologist 18:503–515.

    https://doi.org/10.1037/h0045185

    • Google Scholar
75. 1. Smeets WJ
    2. Marín O
    3. González A

    (2000) Evolution of the basal ganglia: new perspectives through a comparative approach

    Journal of Anatomy 196 ( Pt 4:501–517.

    https://doi.org/10.1046/j.1469-7580.2000.19640501.x

    • PubMed
    • Google Scholar
76. 1. Stitt I
    2. Galindo-Leon E
    3. Pieper F
    4. Engler G
    5. Engel AK

    (2013) Laminar profile of visual response properties in ferret superior colliculus

    Journal of Neurophysiology 110:1333–1345.

    https://doi.org/10.1152/jn.00957.2012

    • PubMed
    • Google Scholar
77. 1. Suryanarayana SM
    2. Pérez-Fernández J
    3. Robertson B
    4. Grillner S

    (2020) The evolutionary origin of visual and somatosensory representation in the vertebrate pallium

    Nature Ecology & Evolution 4:639–651.

    https://doi.org/10.1038/s41559-020-1137-2

    • PubMed
    • Google Scholar
78. 1. Swadlow HA

    (1983) Efferent systems of primary visual cortex: a review of structure and function

    Brain Research Reviews 6:1–24.

    https://doi.org/10.1016/0165-0173(83)90002-4

    • Google Scholar
79. 1. Takada M
    2. Itoh K
    3. Yasui Y
    4. Sugimoto T
    5. Mizuno N

    (1985) Topographical projections from the posterior thalamic regions to the striatum in the cat, with reference to possible tecto-thalamo-striatal connections

    Experimental Brain Research 60:385–396.

    https://doi.org/10.1007/BF00235934

    • Google Scholar
80. 1. Talwar SK
    2. Musial PG
    3. Gerstein GL

    (2001) Role of Mammalian Auditory Cortex in the Perception of Elementary Sound Properties

    Journal of Neurophysiology 85:2350–2358.

    https://doi.org/10.1152/jn.2001.85.6.2350

    • Google Scholar
81. 1. Tang L
    2. Higley MJ

    (2019) Layer 5 circuits in V1 differentially control visuomotor behavior

    Neuron 105:14.

    https://doi.org/10.1016/j.neuron.2019.10.014

    • Google Scholar
82. 1. Tervo DGR
    2. Hwang B-Y
    3. Viswanathan S
    4. Gaj T
    5. Lavzin M
    6. Ritola KD
    7. Lindo S
    8. Michael S
    9. Kuleshova E
    10. Ojala D
    11. Huang C-C
    12. Gerfen CR
    13. Schiller J
    14. Dudman JT
    15. Hantman AW
    16. Looger LL
    17. Schaffer DV
    18. Karpova AY

    (2016) A Designer AAV Variant Permits Efficient Retrograde Access to Projection Neurons

    Neuron 92:372–382.

    https://doi.org/10.1016/j.neuron.2016.09.021

    • Google Scholar
83. 1. Valente A

    (2012)

    Ontogeny of classical and operant learning behaviors in zebrafish

    Learning & Memory 112:.

    • Google Scholar
84. 1. Wan XC
    2. Trojanowski JQ
    3. Gonatas JO

    (1982) Cholera toxin and wheat germ agglutinin conjugates as neuroanatomical probes: their uptake and clearance, transganglionic and retrograde transport and sensitivity

    Brain Research 243:215–224.

    https://doi.org/10.1016/0006-8993(82)90244-x

    • PubMed
    • Google Scholar
85. 1. Wang Q
    2. Burkhalter A

    (2013) Stream-related preferences of inputs to the superior colliculus from Areas of dorsal and ventral streams of mouse visual cortex

    Journal of Neuroscience 33:1696–1705.

    https://doi.org/10.1523/JNEUROSCI.3067-12.2013

    • PubMed
    • Google Scholar
86. 1. Weiskrantz L
    2. Warrington EK
    3. Sanders MD
    4. Marshall J

    (1974) Visual capacity in the hemianopic field following a restricted occipital ablation

    Brain 97:709–728.

    https://doi.org/10.1093/brain/97.1.709

    • PubMed
    • Google Scholar
87. 1. Xiong Q
    2. Znamenskiy P
    3. Zador AM

    (2015) Selective corticostriatal plasticity during acquisition of an auditory discrimination task

    Nature 521:348–351.

    https://doi.org/10.1038/nature14225

    • PubMed
    • Google Scholar
88. 1. Yan Y
    2. Rasch MJ
    3. Chen M
    4. Xiang X
    5. Huang M
    6. Wu S
    7. Li W

    (2014) Perceptual training continuously refines neuronal population codes in primary visual cortex

    Nature Neuroscience 17:1380–1387.

    https://doi.org/10.1038/nn.3805

    • Google Scholar
89. 1. Yang CF
    2. Chiang MC
    3. Gray DC
    4. Prabhakaran M
    5. Alvarado M
    6. Juntti SA
    7. Unger EK
    8. Wells JA
    9. Shah NM

    (2013) Sexually dimorphic neurons in the ventromedial hypothalamus govern mating in both sexes and aggression in males

    Cell 153:896–909.

    https://doi.org/10.1016/j.cell.2013.04.017

    • PubMed
    • Google Scholar
90. 1. Zhang S
    2. Xu M
    3. Chang WC
    4. Ma C
    5. Hoang Do JP
    6. Jeong D
    7. Lei T
    8. Fan JL
    9. Dan Y

    (2016) Organization of long-range inputs and outputs of frontal cortex for top-down control

    Nature Neuroscience 19:1733–1742.

    https://doi.org/10.1038/nn.4417

    • PubMed
    • Google Scholar
91. 1. Zhao X
    2. Liu M
    3. Cang J

    (2014) Visual cortex modulates the magnitude but not the selectivity of looming-evoked responses in the superior colliculus of awake mice

    Neuron 84:202–213.

    https://doi.org/10.1016/j.neuron.2014.08.037

    • PubMed
    • Google Scholar
92. 1. Zingg B
    2. Chou XL
    3. Zhang ZG
    4. Mesik L
    5. Liang F
    6. Tao HW
    7. Zhang LI

    (2017) AAV-Mediated anterograde transsynaptic tagging: mapping corticocollicular Input-Defined neural pathways for defense behaviors

    Neuron 93:33–47.

    https://doi.org/10.1016/j.neuron.2016.11.045

    • PubMed
    • Google Scholar
93. 1. Znamenskiy P
    2. Zador AM

    (2013) Corticostriatal neurons in auditory cortex drive decisions during auditory discrimination

    Nature 497:482–485.

    https://doi.org/10.1038/nature12077

    • PubMed
    • Google Scholar

Author details

1. Sarah Ruediger

   1. Center for Neural Circuits and Behavior, Neurobiology Section and Department of Neuroscience, University of California, San Diego, La Jolla, United States
   2. Department of Physiology, University of California, San Francisco, San Francisco, United States
   3. Howard Hughes Medical Institute, University of California, San Francisco, San Francisco, United States

   Contribution

   Conceptualization, Data curation, Software, Formal analysis, Funding acquisition, Validation, Investigation, Visualization, Methodology, Writing - original draft

   For correspondence

   sarruedi@gmail.com

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-6684-4693

2. Massimo Scanziani

   1. Center for Neural Circuits and Behavior, Neurobiology Section and Department of Neuroscience, University of California, San Diego, La Jolla, United States
   2. Department of Physiology, University of California, San Francisco, San Francisco, United States
   3. Howard Hughes Medical Institute, University of California, San Francisco, San Francisco, United States

   Contribution

   Conceptualization, Supervision, Funding acquisition, Visualization, Writing - original draft, Project administration

   For correspondence

   massimo@ucsf.edu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-5331-9686

• 3,857

  views

• 521

  downloads

• 18

  citations

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