A reliable and flexible gene manipulation strategy in posthatch zebra finch brain

Abstract

Songbird models meaningfully contribute to many fields including learned vocal communication, the neurobiology of social interactions, brain development, and ecology. The value of investigating gene-brain-behavior relationships in songbirds is therefore high. Viral infections typically used in other lab animals to deliver gene editing constructs have been less effective in songbirds, likely due to immune system properties. We therefore leveraged the in vivo electroporation strategy used in utero in rodents and in ovo in poultry, and apply it to posthatch zebra finch songbird chicks. We present a series of experiments with a combination of promoters, fluorescent protein genes, and piggyBac transposase vectors to demonstrate that this can be a reliable, efficient, and flexible strategy for genome manipulation. We discuss options for gene delivery experiments to test circuit and behavioral hypotheses using a variety of manipulations, including gene overexpression, CRISPR/Cas9 gene editing, inducible technologies, optogenetic or DREADD cellular control, and cell type-specific expression.

Figure 1. Hyperactive piggyBac transposase (sPBo) paired with 80 V pulses results in transgene expression in a large portion of the targeted area in zebra finch hatchlings.

(a) Plasmid constructs used in this study. Donor pPB plasmids contain inverted terminal repeats (ITRs) surrounding the transgene expression cassettes, each containing a promoter (CAG or human Synapsin 1 [hSyn1] promoters) and fluorescent protein transgenes (eGFP, mRFP, or CFP). (b) Schematic of plasmid infusion targeting and paddle placements. Left panel: Infusions were targeted along a 45° angle visualized from the midline and Y₀ (the anterior-most boundary of the cerebellum). Middle panel: Infusion into the chick’s left ventricle (arrow). Fast green dye in the plasmid mixture allows visual confirmation that the plasmid solution has filled the ventricle (arrow). Right panel: Schematic showing tri-electrode paddle positioning, as seen from the top and back views of the chick’s head. Positively-charged paddles (red) are placed on either side of the head, in line with the posterior-most portion of the filled lateral ventricles. The negatively-charged paddle is placed on top of the head, directly over the posterior-most portions of the lateral ventricles. (c) Quantitative comparison of the effects of various electroporation conditions on the percent of the targeted area containing fluorescent protein-positive (FP+) cells. sPBo is significantly more effective than wildtype (WT) piggyBac at all voltages tested (*p < 0.05; vs. 70, 80, 90 V sPBo). sPBo paired with 100 ms 80 V pulses delivered at an interpulse interval of 900 ms using 3 mm-wide paddles was significantly more efficient than all other conditions (^#p < 0.001). n = 2–3 per group. Bar graphs indicate the group mean ± SEM; ○ = individual birds. (d) Representative sagittal plane images of fluorescent protein-positive (FP+) cells and DAPI-stained cell nuclei along the lateral ventricles 48 h after in vivo electroporation with sPBo, 3 mm-wide paddles, and 100 ms 80 V pulses delivered with an inter-pulse interval of 900 ms. White boxes outline areas magnified in the panels directly above. Scale bars = 100 μm. (e) Summary of the survival, mortality, and efficacy percentages obtained over the course of this study.

Figure 2. Proportion and density of FP+ cells across medial telencephalic brain regions.

(a) Schematics of approximate boundaries and volumes of areas quantified for FP+ cells at P30, P40, P50. Redrawn and modified schematics are patterned on those in the ZEBrA Histological Atlas (www.zebrafinchatlas.org). (b) Representative images of GFP, RFP, and CFP+ cells in the P30 auditory forebrain (AF). Scale bar = 20 μm. (c) Percentage of the total number of FP+ cells located in the AF, the rostral nidopallium (R. Nido), the lateral mesopallium (L. Meso), the NCL, and the lateral nidopallium (L. Nido) at P30 (left graph), P40 (middle graph), and P50 (right graph). *p < 0.05 and **p < 0.001 between indicated groups. (d and e) Density of FP+ in the AF, caudomedial mesopallium (CMM), caudomedial nidopallium (NCM), R. Nido, L. Meso, NCL, and L. Nido at P30, P40, and P50. Data are graphed grouped either within age and across brain areas (d), or within brain area and across age (e). Bar graphs denote the mean number of FP+ cells per 500 μm² in the indicated areas, with individual data points denoting each subject. (d) *p < 0.01, for all indicated groups as compared with CMM; ^#p < 0.01 for all indicated groups as compared with AF. (e) *p < 0.01, as compared with the P30 group. Bar graphs denote the group mean ± SEM; ○ = individual birds. n = 6 (P30 age group), or 3 (P40 and P50 age groups).

Figure 3. Distribution of FP+ cells throughout the AF at P30, P40, P50.

(a) Schematic of AF and its subregions, CMM, NCM, and Field L. Dashed lines indicate the four rostral-caudal planes spanning the dorsal-ventral axis of AF used for Plot profile analysis: Dorsal, Central, Ventral, and Super Ventral. (b) Representative greyscale images of the AF in P30, P40, and P50 males and females. Contrast was adjusted for figure production only. Scale bar = 200 μm. (c) Plot profile analysis, quantifying the mean grey values along the four lines depicted in (b). Sections 165 μm (left column), 495 μm (middle column), and 825 μm (right column) from midline were analyzed. Central, Ventral, and Super Ventral lines included Field L, denoted with grey bars. Graphs are centered around Field L for visual comparison. Note that x-axes vary along the dorsal-ventral axis, reflecting the change in the anatomical length of AF. Lines represent the age group mean. n = 4 (P30), and 3 (P40 and P50).

Figure 4. Colocalization of fluorescent proteins and NeuN.

(a) Schematic of the areas imaged for the dCMM, dNCM, vCMM, and vCMM quantifications. (b) Example image of the visual confirmation of FP+ cells that are also NeuN-positive (NeuN+; arrowhead), and FP+ cells that are NeuN-negative (arrow). Scale bars = 25 μm. (c and e) Proportion of FP+ cells that are also NeuN+. *p < 0.01, as compared with the P30 and P50 age groups within vNCM. (d and f) Proportion of NeuN+ cells that were also FP+. *p < 0.01, as compared with the area-matched P30 age group. Data are presented as the proportion of co-labeled cells within age and across brain area (c and d) or within brain area and across age (e and f). Bar graphs represent the group mean ± SEM; ○ = individual birds. n = 4 (P30), and 3 (P40 and P50).

Figure 5. Synapsin promoter enhances neuronal expression of fluorescent proteins.

The hSyn1 promoter was used to drive eGFP expression and CAG was used to drive universal transcription of RFP. (a) Density of either RFP-positive (RFP+) or eGFP-positive (GFP+) in AF (left graph), CMM (middle graph), or NCM (right graph) at P30 and P40. *p < 0.05 and **p < 0.001, as compared with age-matched RFP+. (b) Proportion of RFP+ or GFP+ cells that are also NeuN+. Data are presented as the mean across the dCMM, vCMM, dNCM, and vNCM subregions of AF. *p < 0.05 between indicated groups. (c) Proportion of NeuN+ cells that are also either RFP+ or GFP+ in each AF subregion at P30 (left graph) and P40 (right graph). *p < 0.05 between indicated groups. All data are presented as the group mean ± SEM; ○=individual birds. n = 3 per age.