An efficient method for the long-term and specific expression of exogenous cDNAs in cultured Purkinje neurons

Abstract

We present a simple and efficient method for expressing cDNAs in Purkinje neurons (PNs) present in heterogeneous mouse cerebellar cultures. The method combines the transfection of freshly dissociated cerebellar cells via nucleofection with the use of novel expression plasmids containing a fragment of the L7 (Pcp2) gene that, within the cerebellum, drives PN-specific expression. The efficiency of PN transfection (determined 13 days post nucleofection) is approximately 70%. Double and triple transfections are routinely achieved at slightly lower efficiencies. Expression in PNs is obvious after one week in culture and still strong after three weeks, by which time these neurons are well-developed. Moreover, high-level expression is restricted almost exclusively to the PNs present in these mixed cultures, which greatly facilitates the characterization of PN-specific functions. As proof of principle, we used this method to visualize (1) the morphology of living PNs expressing mGFP, (2) the localization and dynamics of the dendritic spine proteins PSD-93 and Homer-3a tagged with mGFP and (3) the interaction of live PNs expressing mGFP with other cerebellar neurons expressing mCherry from a β-Actin promoter plasmid. Finally, we created a series of L7-plasmids containing different fluorescent protein cDNAs that are suited for the expression of cDNAs of interest as N- and C-terminally tagged fluorescent fusion proteins. In summary, this procedure allows for the highly efficient, long-term, and specific expression of multiple cDNAs in differentiated PNs, and provides a favorable alternative to two procedures (viral transduction, ballistic gene delivery) used previously to express genes in cultured PNs.

Figure 1. Plasmids for the expression of exogenous cDNAs in cultured Purkinje neurons

(A) Schematic of plasmid pL7. The L7 DNA fragment (shown in blue) contains a promoter sequence, as well as all of the exons (dark blue) and introns that comprise the small L7 structural gene. All of the AUG codons present in exons 1A, 1B, 2, 3, and 4 were mutated (Serinagaoglu et al., 2007; Smeyne et al., 1995) to ensure that translation starts within the cDNA inserted into the multiple cloning site (MCS; red). The indicated restriction sites are present only once in the vector (except for BamHI). (B) Schematic maps of plasmids pL7-mGFP, pL7-mCerulean and pL7-mCherry, all of which allow fusion of a protein of interest to either the N- or C-terminus of the fluorescent protein. In addition to the L7 DNA fragment (blue) and the L7 exons (dark blue), the cDNA encoding the fluorescent protein is depicted (mGFP, green; mCerulean, blue; mCherry, red). (C) DNA sequences immediately upstream and downstream of the fluorescent protein cDNAs. These sequences are identical in pL7-mGFP, pL7-mCerulean and pL7-mCherry. Also shown are the reading frame and the first two amino acids of the fluorescent proteins and the reading frame at the end of the fluorescent proteins. Restriction sites that occur only once in these plasmids are shown in red. Note that NcoI is not a single-cutter in pL7-mCherry. N-terminal fusions (i.e. the protein of interest precedes the fluorescent protein) are made by inserting cDNAs into the MCS that precedes the fluorescent protein coding sequence. C-terminal fusions (i.e. the protein of interest follows the fluorescent protein) are made by inserting cDNAs using the BglII site or the BglII and SalI-sites that follow the fluorescent protein coding sequence (note that the SalI site cannot be used by itself for cDNA insertion because of an in frame stop codon between the BglII and SalI sites).

Figure 2. Live PNs in dissociated cerebellar culture expressing mGFP following nucleofection with pL7-mGFP

Dissociated cerebellar cells from E18 mouse embryos were nucleofected with 10 μg of pL7-mGFP via nucleofection program O-03, cultured for 13 DIV (A-E) or 22 DIV (F-L), and imaged using confocal microscopy. Image stacks of Z-planes covering the entire thickness of the PNs were projected into a single plane as a maximum projection (A-H) or were processed using the ‘transparent projection’ function of the Zeiss LSM Image software to obtain a 3D reconstruction (I-L). (B) An overlay of the low magnification image shown in (A) with the corresponding transmitted light image is shown. The arrowheads in (G) and (H) indicate examples of neurons other than PNs that are expressing mGFP at low levels. Size bars: 100 μm (A,B); 50 μm (C-H); 20 μm (I); 2 μm (J-L).

Figure 3. Immunofluorescence staining of mGFP-expressing PNs in dissociated cerebellar culture using anti-Calbindin-D-28K antibody

Dissociated cerebellar cells were nucleofected with 10 μg of pL7-mGFP, cultured for 13 DIV, and subjected to immunofluorescence staining using an antibody against the PN-specific protein Calbindin-D-28K. Images show maximum projections of stacks of Z-planes acquired using confocal microscopy and correspond to the anti-Calbindin immunofluorescence signal (A), the mGFP signal (B), and the superimposition of these images (C). Size bar, 50 μm.

Figure 4. Comparison of fluorescent protein cDNA expression from β-Actin- and L7-promoter plasmids in cerebellar cultures

Dissociated cerebellar cells were nucleofected with pL7-mGFP and pβ-Actin-mCherry and cultured for 13 DIV (A, C-E) or 22 DIV (B) before imaging live cultures using confocal microscopy. (A, B) The mGFP (green) and mCherry (red) fluorescence image stacks of Z-planes were processed using the ‘transparent projection’ function of the Zeiss LSM Image software to obtain 3D reconstructions. The superimposed mGFP and mCherry images are shown. (C) The panels show images of a single confocal plane of (from top to bottom) the mGFP fluorescence signal, the mCherry fluorescence signal, their overlay, and an overlay with the corresponding transmitted light image. (D, E) The images of mGFP (green) and mCherry (red) fluorescence signals depict the regions indicated in (C), and were taken from a time series in which images were recorded every two seconds. Time (sec) is indicated; every fourth recorded image is shown. See also Supplemental Movie 1. Size bars: 50 μm (A, B); 2 μm (C).

Figure 5. Comparison of fluorescent protein cDNA expression from CMV- and L7-promoter plasmids in cerebellar cultures

Dissociated cerebellar cells were nucleofected with pL7-mCherry and pCMV-mGFP and cultured for 13 (A, B) or 10 DIV (C) before imaging live cultures. The mCherry and mGFP fluorescence image stacks of Z-planes were projected into a single plane as a maximum projection and are shown in red (mCherry (L7)), green (mGFP (CMV)), and superimposed on each other (Overlay). Also shown is the superimposition of the mCherry, mGFP and the corresponding phase contrast image (Overlay / Phase). The arrowhead indicates a PN expressing mGFP from pCMV-mGFP. Size bar, 100 μm.

Figure 6. Comparison of fluorescent protein cDNA expression from L7-, β-Actin-, and CMV- and promoter plasmids in cerebellar cultures

Dissociated cerebellar cells were nucleofected with pL7-mCerulean, pβ-Actin-mCherry and pCMV-mGFP and cultured for 15 DIV before imaging live cultures. The fluorescence image stacks of Z-planes were processed using the ‘transparent projection’ function of the Zeiss LSM Image software to obtain 3D reconstructions. The mCerulean, mCherry, and mGFP images, as well as their superimposition (Overlay) are shown. Size bar, 50 μm.

Figure 7. Expression of cDNAs encoding mGFP-tagged PSD-93-C3,5 or Homer-3 a_ii in cultured PNs

Dissociated cerebellar cells from E18 mouse embryos were nucleofected with plasmids pL7-mCherry and pL7-PSD-93-C3,5-mGFP (A, B) or pL7-Homer-3a_ii-mGFP (C, D) and cultured for 13 DIV (A, C) or 22 DIV (B, D) before imaging live cultures. Fluorescence image stacks of Z-planes spanning the entire height of the PN were processed to obtain a 3D reconstruction. The resulting PSD-93-C3,5-mGFP (PSD-93-C3,5) and Homer-3a_ii-mGFP (Homer-3a_ii) images are shown, as well as the superimposition of the respective mGFP images with the corresponding mCherry image (Overlay). The inserts show images from a single confocal plane of the fluorescence signal from PSD-93-C3,5-mGFP (PSD-93-C3,5) and Homer-3a_ii-mGFP (Homer-3a_ii), as well as the superimposition of the respective images with the corresponding mCherry image (Overlay). The inserts are taken from time series recordings (see Supplemental Movie 2 and 3). Size bar, 20 μm (A-D) and 2 μm (inserts).