Laser Capture Microdissection of Single Neurons with Morphological Visualization Using Fluorescent Proteins Fused to Transmembrane Proteins

Abstract

Gene expression analysis in individual neuronal types helps in understanding brain function. Genetic methods expressing fluorescent proteins are widely used to label specific neuronal populations. However, because cell type specificity of genetic labeling is often limited, it is advantageous to combine genetic labeling with additional methods to select specific cell/neuronal types. Laser capture microdissection is one of such techniques with which one can select a specific cell/neuronal population based on morphological observation. However, a major issue is the disappearance of fluorescence signals during the tissue processing that is required for high-quality sample preparation. Here, we developed a simple, novel method in which fluorescence signals are preserved. We use genetic labeling with fluorescence proteins fused to transmembrane proteins, which shows highly stable fluorescence retention and allows for the selection of fluorescent neurons/cells based on morphology. Using this method in mice, we laser-captured neuronal somata and successfully isolated RNA. We determined that ∼100 cells are sufficient to obtain a sample required for downstream applications such as quantitative PCR. Capability to specifically microdissect targeted neurons was demonstrated by an ∼10-fold increase in mRNA for fluorescent proteins in visually identified neurons expressing the fluorescent proteins compared with neighboring cells not expressing it. We applied this method to validate virus-mediated single-cell knockout, which showed up to 92% reduction in knocked-out gene RNA compared with wild-type neurons. This method using fluorescent proteins fused to transmembrane proteins provides a new, simple solution to perform gene expression analysis in sparsely labeled neuronal/cellular populations, which is especially advantageous when genetic labeling has limited specificity.

Keywords: channelrhodopsin; gene expression; halorhodopsin; optogenetics; viral vector.

Figure 1.

Flowchart of methodology.

Figure 2.

The signals of fluorescent proteins fused to transmembrane proteins are preserved after the standard tissue-processing procedure for laser capture microdissection. A, Schematic structures of the viral vector genomes. B–F, Images of fluorescently labeled granule cells in the dentate gyrus of mice injected with a mixture of two retroviral vectors expressing YFP-fused enhanced halorhodopsin, eNpHR3.0-YFP, or mCherry (B); Pomc-ChR2-YFP transgenic mice (C) and mice injected with a retroviral vector expressing ChR2(SSFO)-YFP (D); an adeno-associated viral vector expressing mCherry-TRPV2 (E) and a retroviral vector expressing a membrane-targeted form of GFP, GFP-F (F). Brain sections were prepared from paraformaldehyde fixed or freshly frozen brains and went through the standard tissue-processing procedure. Note that fluorescence signals of eNpHR3.0-YFP (B), ChR2-YFP (C), ChR2(SSFO)-YFP (D), and mCherry-TRPV2 (E), but not mCherry (B) and GFP-F (F), were preserved in sections prepared from freshly frozen brains. Arrows and arrowheads in B indicate somata labeled by eNpHR3.0-YFP and mCherry. RV, Retroviral vector; AAV, adeno-associated viral vector; LTR, long-terminal repeat; CAG, CAG promoter; WPRE, Woodchuck hepatitis virus post-transcriptional regulatory element; CaMKIIα, the CaMKIIα promoter; ITR, inverted terminal repeat. Scale bars, 100 μm.

Figure 3.

Cryoprotectant prevents the loss of fluorescence signals from regular fluorescent proteins in HEK293 cell culture after a freeze–thaw cycle. A, Representative images showing fluorescence signals from regular, cytosolic fluorescent proteins (GFP and mCherry) after a freeze–thaw cycle with or without cryoprotectant. Fluorescence signals disappeared without cryoprotectant while cryoprotectant preserved fluorescence signals. B, Representative images showing fluorescence signals from fluorescent proteins fused to transmembrane proteins [eNpHR3.0-YFP, ChR2-mCherry, ChR2(SSFO)-YFP]. Fluorescence signals were preserved without cryoprotectant after a freeze–thaw cycle. Scale bar, 50 μm.

Figure 4.

Laser capture microdissection of fluorescence labeled neurons. A, A schematic structure of the viral vector genome. B, Images of eNpHR3.0-YFP-expressing neurons and non-neuronal cells in the dentate gyrus of mice that were injected with a retroviral vector expressing eNpHR3.0-YFP. C, Fluorescence images of sections before and after laser capture microdissection. Red dotted ovals show the location of somata of fluorescently labeled neurons selected for laser capture microdissection. Blue dots with yellow labels indicate the center of a selected area used for laser capture. Scale bar, 20 μm. D, An electrophoresis image showing PCR amplification end products for the DCX, CaMKIIα, and β-actin genes. Note the high expression of neuronal markers in fluorescently labeled neurons compared with a total liver RNA sample where DCX and CaMKIIα were not detected. M, Molecular marker; NTC, nontemplate control.

Figure 5.

RNA quality assay. A, Electrophoresis images of RNA samples. Clear bands of 18S and 28S ribosomal RNAs are visible. Ladder: 0.2, 0.5, 1.0, 2.0, 4.0, and 6.0 kb. B, Electropherogram showing the distinct peaks of 18S and 28S ribosomal RNAs. FU, Fluorescence unit; Pink and thick green solid lines, baseline for 18S and 20S peaks, respectively; thin green solid line: overall baseline.

Figure 6.

Evaluation of specific RNA collection from neurons targeted by laser capture microdissection. A, Images of brain sections sparsely containing YFP⁺ neurons in the dentate gyrus. The brain sections were prepared from Pomc-ChR2-YFP mice. YFP⁺ neurons (top) and their neighboring YFP⁻ cells (bottom) were visually selected for microdissection (red dotted ovals). Scale bar, 20 μm. B, Relative amount of YFP RNA in YFP⁺ neurons and their neighboring YFP⁻ cells, quantified by qPCR analysis.

Figure 7.

Validation of retrovirus-mediated, single-cell gene knockout in adult-born neurons in the dentate gyrus. A, Schematic structures of the viral vector genomes. B, Experimental time line. C, Images of the dentate gyrus of floxed NR1 mice, which was injected with a bicistronic retroviral vector expressing eNpHR3.0-YFP and Cre. Scale bar, 50 μm. D, qPCR amplification end products in gel electrophoresis. M, Molecular marker; KO, NR1 knock-out neurons with Cre expression; Control, control neurons without Cre expression; NTC, nontemplate control. E, qPCR quantification of NR1 RNA in Cre-expressing (KO) and control adult-born neurons. Note the ∼92% reduction in Cre-expressing neurons.