The brain is required for normal muscle and nerve patterning during early Xenopus development

Our data suggest a revision of the view of the brain as quiescent prior to the animal’s independent activity, showing that its signaling role spans the control of pattern formation and behavior. This is consistent with recent proposals that the important mechanistic and conceptual commonalities exist between the algorithms of neuroscience and those that guide pattern regulation64. This model system and dataset serves as a base for future studies of local and long-range influences over large-scale patterning. The relationship of instructive morphogenetic signals mediated by bioelectric events to the computational capabilities of the brain is an exciting direction for future work. Moreover, our data point to widely-available and already human-approved drugs as potential ‘morphoceuticals’ – agents that can be capitalized upon to prevent or perhaps even reverse specific kinds of anatomical defects. More broadly, these results suggest a direction for regenerative medicine towards the development of implanted organoids and hybrid electrochemical constructs to provide bioelectric and neurotransmitter stimulation.

Xenopus laevis embryos were fertilized in vitro according to standard protocols25 in 0.1X Marc’s Modified Ringer’s solution (MMR; 10 mM Na+, 0.2 mM K+, 10.5 mM Cl–, 0.2 mM Ca2+, pH 7.8). Xenopus embryos were housed at 14 °C and staged according to Nieuwkoop and Faber65. All experimental procedures involving Xenopus embryos were approved by the Institutional Animal Care and Use Committees and Tufts University Department of Laboratory Animal Medicine under protocol M2014-79.

Stage-25 embryos were randomly grouped in one of the two experimental groups: Control (Ctrl) and Brainless (BR−; see Fig. 1a for a schematic representation of the experimental design). Randomized controlled trial conditions were maintained throughout the experiment. Embryos were anesthetized in a 0.02% tricaine solution (pH 7.5) in 0.1X MMR. Brain removal in BR− group was performed under dissecting microscope and using a dissecting knife (FST #10055-12). Once movement ceased, a single cut removed the anterodorsal region corresponding to the brain (from the cement gland to the most anterior somite; Fig. 1b). After brain removal, animals were allowed to heal in 0.75X MMR for 1 h. After washing, untreated and surgically treated animals were raised at 14 °C and scored and analyzed for phenotype distribution and the different morphological parameters, respectively, at stages 30–41 (early stage) and 42–48 (late stage). Following the same experimental and care conditions, additional spinal cord (SC) resection (SC−; Fig. 6a) experiments were performed on stage-25 embryos. To prevent regeneration66, one segment of the spinal cord, sizing 50–100 μm in length, was completely removed by using two forceps with super-fine tips (Dumont #5ST, FST 11252-00). SC segments were removed at cervical level, immediately posterior to the hindbrain, and taking special care of minimizing the damage on the most rostral myotomes.

Two different Sham surgeries were performed by removing pieces of tissue of comparable size at different locations of the stage-25 animal body (Supplementary Fig. 1): part of the endodermal yolk mass (Yolk− or yolk resection) or the most posterior part of the embryo body (Tail− or tailbud resection).

In order to feed the BR− tadpoles, a nutritive medium was modified from ref. 67. The nutritive medium consisted of 9.5% Ham’s nutrient mixture F12 (with 1.0 mM L-glutamine, Sigma 51651 C) and 0.5% calf serum (from formula-fed bovine calves, iron supplemented, Sigma C8056) dissolved in 0.1X MMR. From stage 45, Ctrl and BR− tadpoles were fed with this nutritive medium for 2 h every day.

Capped synthetic mRNAs generated using mMessage mMachine kit (Ambion) were dissolved in nuclease-free water and injected into embryos immersed in 3% Ficoll solution using standard methods25. The mRNA injections were made using borosilicate glass needles calibrated to bubble pressures of 55 to 60 kPa in water, delivering 100- to 130-ms pulses. Each injection delivered between 0.5–1 nl or 0.5–1 ng of mRNA per blastomere into the embryos. Two different sets of injections were performed. Firstly, at the two-cell stage (Fig. 3a, b), either one (1/2 HCN2 Inj) or both (HCN2 Inj) blastomeres were injected. Secondly, at the four-cell stage (Fig. 6b, c), either the two ventral (HCN2 ventral Inj) or the two dorsal (HCN2 dorsal Inj) blastomeres were injected. Injections were done into the center of cells at the animal pole.

Constructs used were HCN2-WT-2A-GFP or HCN2-WT68 and β-galactosidase in PCS2 and both RNAs (HCN2-WT and β-galactosidase) were mixed at 3:5 and 1:5 dilutions respectively, for microinjections. HCN2 is a potassium/sodium hyperpolarization-activated cyclic nucleotide-gated ion channel, type 269. Embryos were injected in 3% Ficoll solution and after 30 min, were washed and then reared in 0.1X MMR until desired stages.

Ctrl and BR− Xenopus embryos were exposed to specific pharmacological agents, dissolved in 0.1X MMR, from stage 25 (immediately after brain removal) to stage 48. The drugs were refreshed every three days. Two drugs targeting cholinergic receptors and one directed to NMDA-glutamatergic receptor were used: 10 μM scopolamine (a muscarinic-receptor antagonist; Tocris 1414), 10 μM carbamoylcholine chloride (carbachol, a dual muscarinic- and nicotinic-receptor agonist; Tocris 2810) and 10 μM (RS)-(Tetrazol-5-yl)glycine (RS; an NMDA receptor agonist; Tocris 0312), respectively. All drug treatments were performed using embryos from mixed batches of fertilizations, using at least three biological replicates.

Stock solutions of all pharmaceuticals were created by dissolving the compound in Millipore water (to a final drug concentration of 100 mM for both scopolamine and carbachol, and to 50 mM for RS) and then stored in aliquots at –20 °C. Further dilution of all compounds was made in normal media (0.1X MMR). Control experiments were performed using embryos in 0.1X MMR, both for Ctrl and BR− groups. Drug concentrations were determined through toxicity screens and were applied at levels that did not result in lethality or observable developmental defects.

Immediately after treatment, at the relevant time points (immediately after brain removal and first 2 weeks and third and fourth weeks after brain removal for early- and late-stage studies, respectively), anesthetized tadpoles were fixed in MEMFA overnight at 4 °C for whole-mount immunofluorescence25. Briefly, embryos were washed twice in 1× phosphate buffered saline (PBS), and permeabilized in PBS 0.1% Triton X-100 (PBST) for 30 min. Animals were then blocked with 10% normal goat serum in PBST for 1 h at room temperature (RT). Samples were rocked overnight at 4 °C using znp1 (Developmental Studies Hybridoma Bank, used at 1:250 dilution) and anti-acetylated alpha-tubulin antibody (Tub; Sigma T7451 used at 1:500 dilution). Following primary exposure, embryos were washed three times in PBST before a 60-min RT incubation with AlexaFluor-555 conjugated secondary antibody (Invitrogen) used at 1:500 diluted in PBST. Following secondary incubation, animals were washed three times for 15 min in PBST and imaged on an Olympus BX-61 microscope equipped with a Hamamatsu ORCA AG CCD camera, and controlled by Metamorph software. Particular care was taken to ensure that embryos from all the different groups were processed in the same batch at the same immunofluorescence session.

Embryos injected with β-galactosidase (β-gal) mRNA were fixed (30 min in MEMFA at RT) at the relevant stages, washed twice in PBS with 2 mM MgCl2, and stained with X-gal (Roche Applied Sciences, Indianapolis, IN) staining solution at 37 °C for 3 h. Embryos were then rinsed three times in PBS and analyzed.

Xenopus embryos at stage 35 were collected and fixed in MEMFA25 and in situ hybridization was performed25, 70. The embryos were washed with phosphate buffered saline 0.1% Tween-20 (PBST) and transferred through a series of methanol washes 25%-50%-75%-100%. In situ antisense probe was generated in vitro from linearized template using DIG labeling mix (Roche). Chromogenic reaction times were optimized for signal to background ratio. Antisense RNA probes for Xenopus HCN2 were generated from X. laevis hcn2.L IMAGE clone 5514485 (purchased from Dharmacon): a HindIII fragment was deleted, leaving exons 2-4 and part of exon 5 as probe. In situ hybridized embryos were then agarose embedded and sectioned. Briefly, 4% low melting point agarose was melted at poured into plastic scaffolds. Gently dried, in situ hybridized embryos were oriented within the agarose for transverse sectioning. The agarose blocks were then allowed to solidify, trimmed and sectioned using Leica VT1000S vibratome to obtain transverse sections.

Ctrl-, Sham-Yolk−-, and BR−-Xenopus embryos were scored for abnormalities in the macroscopic phenotype. Average number of animals with clearly identifiable left-right axis bending and/or tail defects were used to evaluate the percent abnormal within each population. Embryos were photographed (lateral view) using a Nikon AZ100 with an attached QImaging CD camera controlled by QCapture software.

Mean collagen density, angle and length of myotome fibers, and number of somites. All muscle structure studies were performed on 4× or 10× images taken under polarized light. Birefringence microscopy was performed on an Olympus BX-61 compound microscope with a universal condenser (U-UCD8). The transmitted light DIC slider (U-DICTS) was pulled out and the polarizing filter was rotated such the background appeared darkest. Embryos were positioned at a 45° angle for imaging. Analysis of birefringence images was performed in ImageJ (National Institutes of Health, Maryland, US).

Collagen density in early-staged embryos (Fig. 1c, d; short arrows) was evaluated by means of optical density (OD) using a gray scale gradient of 0-255 (white to black), on 4× polarized-light photomicrographs31. Each embryo was sampled using a systematic procedure in order to measure the OD at different anatomical levels, along the anteroposterior axis. Six OD-myotome measurements were used to determine mean value for each animal. At least 30 embryos within each experimental group were analyzed. Possible aberrations originating from the optical system of the camera were corrected by background image subtraction.

Mean angle and length of myotome fibers (Fig. 1c–g; overlaid dashed arrowhead-like lines and double-headed arrows, respectively) for both early- and late-staged embryos was performed on 10× polarized-light images, using ImageJ software. For each animal, three somites at anterior, central and posterior levels, respectively, were systematically random sampled. To obtain a representative mean value of length, nine myotome fibers, from dorsal to ventral axis, were measured for each somite at each anatomical level. The number of somites, between end of the gut and tip of the tail, for both early- and late-staged embryos was evaluated on 4x birefringence images.

Analysis of the neural network branching (see Fig. 5, unfilled triangles) in early- and late-staged embryos was performed by evaluating the intensity of znp1 and acetylated alpha-tubulin (Tub) immunostaining, respectively, through gray-level measures (OD). In order to obtain one OD-mean value per animal, multiple measurements were taken along the anteroposterior axis. Each measurement consisted of the mean value of the pixels of a fixed-size window, placed between two consecutive somites (intermyotomal or internal neuropil). The size of the window remained constant across subjects. All individuals among whom comparisons are being made were produced in the same batch, treated identically for processing and imaging–conditions were not changed. In addition, possible aberrations originating from the optical system of the camera were corrected by background image subtraction.

All statistical analysis was performed using GraphPad Prism (GraphPad Software, Inc., CA, US) and Microsoft Excel (Microsoft Corporation, WA, US) software. Each dish of tadpoles was considered a replicate.

When comparing numerical variables (OD, number of somites, angle and length of the myotome fibers), data from, at least, two replicates (minimal replicate size: n = 25 embryos) were analyzed per each experimental condition. Firstly, data were tested for homogeneity of variances by Bartlett’s test. If variances are similar, we applied unpaired and two-tailed Student’s t-test (two independent groups), one-way ANOVA test (multiple independent groups), or two-way ANOVA test (two independent variables) followed by post-hoc Bonferroni’s test (when P < 0.05). In case of unequal variances (or non-normal distributions for a sample size of less than 50), two-tailed Mann–Whitney and Kruskal–Wallis test followed by pot-hoc Dunn’s test (when P < 0.05), respectively, were used.

For differences in sample proportion (percent of aberrant embryos) or distribution analysis (frequency of phenotypes), data from, at least, two replicates and sample size equal or greater than 50 embryos were pooled and analyzed by multiple-sample z-test or Χ2 test, respectively.

The significance level (α) was set to 0.05 in all cases. The statistical values are reported as mean ± s.d.

The authors declare that all data supporting the findings of this study are available within the article and its Supplementary Information Files or from the corresponding author upon reasonable request.