Serotonergic stimulation induces nerve growth and promotes visual learning via posterior eye grafts in a vertebrate model of induced sensory plasticity

For animals exposed to the 5-HT1B/D activator Zolmitriptan (Sigma SML0248), 50 mM stocks were prepared in DMSO and frozen at −20 °C until used. Once operated animals were given 30 min to heal, they were moved to new dishes containing 50 µM Zolmitriptan in 0.1× MMR. Pharmaceuticals were refreshed every other day and removed at stage 46, after which all animals were reared in standard 0.1× MMR.

Host innervation was visualized through immunohistochemistry with the monoclonal anti-acetylated alpha tubulin antibody (Sigma T7451) using a previously described protocol.39 Briefly, animals were anesthetized in 0.02% tricaine for 30 min, then fixed for 2 h at room temperature in MEMFA (100 mM MOPS (pH 7.4), 2 mM EGTA, 1 mM MgSO4, 3.7% (v/v) formaldehyde). Following fixation, animals were washed three times, 10 min per wash, in phosphate buffered saline + Tween 20 (PBST), and then blocked for 1 h at room temperature with 10% goat serum in PBST. Samples were then rocked overnight at 4 °C in monoclonal anti-acetylated alpha tubulin antibody, diluted 1:500 in PBST + 10% goat serum. Following primary exposure, samples were washed three times for 15 min in PBST before a 60 min secondary incubation with AlexaFlour-555 conjugated secondary at 1:1000 diluted in PBST. Following secondary incubation, samples were washed three times for 15 min in PBST and imaged on an Olympus BX-61 microscope.

All wavelength-learning trials were performed with a custom built automated training system.26, 40 The device consists of an array of 12 individual chambers, each containing a disposable 60 × 15 mm petri dish filled with 15 ml of 0.1× MMR. Below each dish is a machine vision camera (Insight-Micro 1400, Cognex Corporation, Natick, MA, USA), which uses a firmware-embedded background subtraction algorithm to plot the location of any animals in the chamber in a Cartesian manner. Illumination can be provided to each chamber independently from above, by an illumination control module which can specify color and intensity by quadrant. Red or blue light is delivered by light-emitting diodes (Osram Semiconductors, blue LED; 470 nm part no. LBW5SM, red LED, 635 nm part no. LRG6SP) and Xenopus are expected to see both given the spectral profiles of their three known cone classes.41 Finally, within each chamber is also a set of six iridium oxide-coated titanium electrodes allowing the delivery of mild to strong electric shocks. All shocks delivered during wavelength-mediated training experiments were 1.2 mA AC currents, pulsed for 100 ms followed by 300 ms of no shock. This value was previously determined to be the lowest that elicits a behavioral response and no animals displayed physiological or behavioral abnormalities upon completion of testing.

To determine if tadpoles could learn in a color-based associative assay, the following trial was performed26: individuals were first presented with an arena which was illuminated half with red light, half with blue light, in the absence of any punishment to probe innate color preferences. Innate testing lasted 30 min and the position of the colors was inverted after 15 min to avoid having stationary tadpoles be scored as having a 100% preference for either color (inverting the lights would result in a 50/50 preference). After innate testing, tadpoles enter the learning acquisition phase, where individuals are again presented with red and blue halves of the dish, but in this case the tadpole receives a 1.2 mA shock if it occupies a red quadrant. Acquisition duration is 20 min, with the colors in the chamber being inverted every 5 min. Following acquisition, tadpoles are given a 90-min rest in which the entire chamber is illuminated with blue light, and no punishment is delivered. Finally, individuals are probed for learning by giving them a choice between red and blue for 5 min, with neither sample receiving punishment. The entire block of acquisition-rest-probe is repeated six times. All tadpoles used in learning trials were stage 48, and an individual was determined to have learned if their preference for red was below 30%, averaged across the final three probes of the experiment. Further, tadpoles were fed directly before associative learning experiments, and food was added to each arena during training, as hungry animals fail to learn in the assay. Animal location was varied across trials to ensure treatments were evenly distributed between chambers, avoiding any confounding effect of position.

In addition to red/blue wavelength discrimination, tadpoles were tested to determine whether they could respond to the movement of patterns on a white background. For this assay, 3–5 tadpoles were tested simultaneously in a 15-cm diameter petri dish filled with 0.1× MMR. Dishes were placed on top of a LCD monitor screen (model Lenovo D221), and animated patterns were displayed under the bottom of the dish. The animated pattern consisted of groups of triangles around the outer portion of the dish, which would rotate in a clockwise or counterclockwise direction at a speed of 1 revolution per 20 sec (or 18°/s)—the approximate speed of a stage 48 tadpole swimming around the edge of the dish. Tadpoles were placed in the dish for 15 min with the pattern rotating clockwise, after which their direction of swimming was scored [with the possible outcomes being clockwise, counterclockwise, or other—which included no movement, traversing the center of the dish (defined as any open area not covered by the pattern), or colliding with another animal]. After the clockwise test, the rotation of the pattern was set to counterclockwise and the tadpole swimming direction was scored again after 15 min had passed. For all pattern movement trials, food was withheld for 24 h pre-experiment, as satiated animals often did not swim often enough to score directionality.

To image innervation by implanted eye grafts, stage 46 animals were anesthetized for 10 min in 0.02% tricaine solution, pH 7.5, and transferred to depression slides. Donor fluorescence was imaged with an Olympus BX-61 microscope using a TRITC filter set, which was then overlaid with bright field images. Individuals were imaged at 40× and 100× magnifications, and widefield stacks were used to resolve the region of interests. Host innervation was imaged following immunohistochemistry using the same Olympus BX-61 and TRITC filter set.

Statistical analyses were performed using Prism v.5 (GraphPad Software, La Jolla, CA, USA). Animal numbers were chosen based on previous studies as well as sample size estimates given a type I error rate of 5% and power of 0.8. Host innervation data were collected as binomial outputs (presence or absence of innervation phenotype) and analyzed using non-parametric statistics. For color-learning assays, animals were considered to have learned if their average preference over the last three testing periods was equal to or below 30%. Results were then compared using a two proportion z-test, with Bonferroni corrections to adjust experiment-wide α level, where multiple comparisons were made within an experiment. Comparison of punishment between tadpole learners and non-leaners was performed using a two-way ANOVA.

We thank Joshua Finkelstein for comments on drafts of this manuscript, Erin Switzer for Xenopus husbandry assistance, as well as Gerhart Ryffel and Roger Tsien for the TdTomato clones. This research was supported by the Allen Discovery Center program through The Paul G. Allen Frontiers Group, and by The G. Harold and Leila Y. Mathers Charitable Foundation.

D.B. and M.L. conceived and designed the experiments. K.V. created, and collected data for, the behavioral rotation assay and D.B. performed all other experiments. D.B., K.V., and M.L. all contributed to writing the manuscript.