- Email: [email protected]
Construction and application of a microprojectile system for the transfection of organotypic brain slices
Journal of Neuroscience Methods 101 (2000) 171 – 179 www.elsevier.com/locate/jneumeth
Construction and application of a microprojectile system for the transfection of organotypic brain slices Merdol Ibrahim a,*, Azeddine Si-Ammour b, Marco R. Celio a, Felix Mauch b, Pierre-Alain Menoud a a
Institute of Histology and General Embryology, Uni6ersity of Fribourg, CH-1705 Fribourg, Switzerland b Institute of Plant Biology, Uni6ersity of Fribourg, CH-1705 Fribourg, Switzerland Received 17 March 2000; received in revised form 19 June 2000; accepted 21 June 2000
Abstract In this study we outline a method for constructing an inexpensive chamber used in the transfection of organotypic brain slices. This chamber differs from most commercially available chambers in that DNA-coated gold microcarriers are directly carried by a flow of helium at low pressure (26 psi). Most other chambers employ macrocarriers onto which DNA-coated gold is first loaded, and then released by a shock of helium onto the reverse side of the macrocarriers. This home constructed device has been successfully employed in the transfection of organotypic brain slices cultured using the air – medium interface method. Mammalian expression vectors containing cytomegalovirus (CMV) and simian virus (SV40) enhancers/promoters were used to express enhanced green fluorescence protein (EGFP). DNA was coated onto 0.6-mm gold microcarriers. Transfected cells were visualised under a fluorescence microscope and included identifiable neurones and oligodendrocytes. Also included in this study are step-by-step methods for the preparation of gold microcarriers and organotypic brain slices. © 2000 Elsevier Science B.V. All rights reserved. Keywords: Cytomegalovirus; Simian virus; Green fluorescent protein; Gene gun; Neurone; Oligodendrocyte
1. Introduction Microprojectile methods of gene transfer have been successfully used to transform a variety of plant (Christou et al., 1988; Klein et al., 1988) and animal tissue (Williams et al., 1991). However, these transfection techniques have more recently been used in the field of neuroscience, with interest focusing on the transfection of organotypic brain slice cultures (Arnold et al., 1994; Lo et al., 1994). Many versions of the microprojectile system have been constructed (Christou et al., 1988; Sautter et al., 1991; Takeuchi et al., 1992) and commercially available ones are now being used periodically (Williams et al., 1991; Arnold and Heintz, 1997; Thomas et al., 1998; Wellmann et al., 1999). However, the process of transfection remains either too cumbersome, uses too many consumables and or more importantly the appara-
tus itself is too expensive. An inexpensive particle system used in plant transfections was initially described by Takeuchi et al. (1992), and then further refined by Vain et al. (1993). This ‘Particle Inflow Gun’ (PIG) as termed by Vain et al. (1993) was originally used in the transfection of plant material and was the basis on which the system in this study was based. In addition, the home constructed microprojectile chamber in this study is demonstrated as a successful tool in the transfections of organotypic brain slice cultures, employing the air– medium interface method (Stoppini et al., 1991), using mammalian cytomegalovirus (CMV) and simian virus (SV40) driving the expression of enhanced green fluorescence protein (EGFP). 2. Materials and methods
2.1. Construction of the microprojectile chamber * Corresponding author. Tel.: +41-26-3008490; fax: + 41-263009732. E-mail address: [email protected] (M. Ibrahim).
The full plans and dimensions to construct the microprojectile chamber are shown in Figs. 1 and 2. Note
0165-0270/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 5 - 0 2 7 0 ( 0 0 ) 0 0 2 6 9 - 7
172
M. Ibrahim et al. / Journal of Neuroscience Methods 101 (2000) 171–179
that catalogue numbers have been given to chamber parts that are thought to be necessary. All other parts not indicated can be purchased directly from any local ‘hardware store’. Initially a five-sided chamber was constructed from 12-mm-thick aluminium sheets (Fig. 1). The top panel measured 140×108× 12 mm; two side panels measured 350× 108×12 mm (Fig. 1) and the back panel measured 350 ×140× 12 mm (Fig. 1). A smaller, lower panel measuring 116×108 ×12 mm was placed within the chamber, 14 mm from the bottom (Fig. 1). Before constructing the chamber the two side panels had 2-mm channels, 25 mm apart cut into them (starting 70 mm from the bottom), running the full length of the panels (Figs. 1 and 3A,B). These channels were used for holding perforated aluminium shelves (119× 105 × 1.8 mm) (Fig. 3A,B) in place and for allowing the possibility of adjusting the distance of the shot (Fig. 3A,B). The lower
Fig. 1. Schematic plan and measurements for the construction of the microprojectile chamber. A five-sided panel was initially constructed from 12-mm aluminium and then an overhanging panel (coloured grey) was screwed into place.
ends of the side panels each had a 0.6-mm (14¦) hole drilled into them (Fig. 1) to allow for the attachment of a vacuum pump and air-inlet valve (screw valve) (Figs. 3A and 4). The top panel also had a 0.6-mm hole drilled into its centre and was fitted with a double-threaded (14¦) connector (Bru¨tsch Ru¨egger, CH, cat no. 97542) (Fig. 4). To the top end of the double-threaded connector a compressed air coupler was screwed into place (Bru¨tsch Ru¨egger, CH, cat no. 97531 R14¦) (Fig. 4). This air coupler allowed for the connection of a solenoid valve (Carlo Gavazzi Steinhausen, CH, cat no. SCG262CO14) (Figs. 3A,B and 4) which was connected to a stainless steel quick connect nipple (Bru¨tsch Ru¨egger, CH, cat no. 97534 R14¦) (Figs. 3A,B and 4). To the inner side of the top panel, a stainless steel Luer-lock (Sartorius, cat no. SM17122) was screwed into place for the attachment of a 13-mm Swinney filter holder (Gelman Sciences, cat no. 4317) (Figs. 3A,B and 4). The chamber was initially glued and then held together with 5-mm metal screws (Fig. 1). The five-sided chamber then had an overhanging front panel screwed into place (Figs. 1 and 2). The overhanging front panel was made from an aluminium sheet measuring 370×160× 12 mm, with a middle portion measuring 320× 120 mm cut out (Fig. 2A). From the outer edges (2 mm) of the panel a continuous 3.5-mm channel was made (Fig. 2A), into which a 4-mm rubber ring was glued (Fig. 3A,B), to give an airtight seal to the chamber when a vacuum was applied. Channels, 25 mm apart and 70 mm from the bottom of the panel were made (Fig. 2A), corresponding to those of the inner side panels (Fig. 1). The overhanging front panel was then glued and screwed into place (Figs. 1 and 3A). Finally, a door was made from 15-mm-thick Plexiglass (Figs. 2B and 3A), having the same dimensions (370×160 mm) as the front overhanging panel. To allow the Plexiglass to sit over the overhanging front panel a 2-mm aluminium support was bent at a right angle (Fig. 2B) and screwed to the top of the Plexiglass. Handles were attached to the Plexiglass to allow for easy removal and positioning (Figs. 2B and 3A). To complete the chamber a vacuum cut-off lever was screwed onto one side of the chamber (Fig. 3A). A cross fitting vacuum gauge (Tribold, CH, cat no. 41801213) was attached to the end of the vacuum cut-off lever, leading via flexible reinforced tubing to a vacuum pump capable of applying a vacuum of 20 mmHg (Fig. 3A). To the opposite side panel, an air-inlet valve (screw valve) was attached to enable air back into the chamber once a vacuum was applied (Fig. 3A).From the top of the chamber, helium (He) was carried by steel mesh reinforced flexible pressure hose, to the solenoid, from a He cylinder. The He cylinder had a pressure regulator attached (Fig. 4) (Gloor, CH, cat no. 1500-8) enabling a pre-defined working pressure to be used. The timing of the release of He from the solenoid
M. Ibrahim et al. / Journal of Neuroscience Methods 101 (2000) 171–179
173
Fig. 2. Schematic plans and measurements for the overhanging aluminium front panel (A) and Plexiglass door (B).
was controlled by a time relay control (Fig. 3A) (Bircher, CH Cat no. TRAB230AC) set at 50 ms. A metal dispersion grid (Fig. 3B) (BioRad, cat no. 1652336) was heat-fixed onto a 90-mm Petri dish lid, which had a corresponding hole cut in its centre. The dispersion grid along with the Petri dish was changed when using different DNA constructs. In our studies, transfections were optimal when a dispersion grid was placed at the first level (Fig. 3A,B) and the brain slice culture placed at the third level of the chamber (Fig. 3A,B). The Swinney filter holders screwed directly into the inner Luer-lock holder (Fig. 3A,B). In the final set-up, the distance from the tip of the filter holder to the dispersion net was 10 mm and 60 mm to the brain slice culture.
2.2. Organotypic brain slices 2.2.1. Preparation of organotypic brain slices Under aseptic conditions, rat pups ranging from postnatal (P) day 5 to 18 were decapitated and brains quickly removed into a Petri dish containing cold dissection medium (see below). Using a binocular microscope, the brain was held steady using forceps and initially bifurcated coronally near the superior and inferior colli-
culi. The forebrain was removed from the dissection medium and placed bifurcated side down onto sterile blotting paper. The brain was then glued onto a cutting block and transferred to a vibratome. More fresh dissection medium was added to the vibratome chamber and 300-mm coronal sections were taken throughout the forebrain. Slices were placed into cold dissection medium for a further 30 min. Cerebellar slices were also cut into 300-mm sagittal sections using a McIlwain tissue chopper (McIlwain, Mickle Laboratory, Cambridge). The sectioned cerebellum was transferred to fresh cold dissection medium and the cut folia separated using blunt forceps. The sections were allowed to rest in cold dissection medium for a further 30 min. Up to four cerebellar folia or 1–2 forebrain slices were then transferred to 30-mm Millicell culture inserts (Millipore, cat no. PICM03050), which in turn were placed into individual Petri dishes (Fig. 3D). All dissection medium residues were pipetted from the well and surrounding tissue. Finally, 1 ml of culture medium was added, so that the cut tissue remained open to the air. Brain slices were then placed into an incubator at 37°C containing 5% CO2. For each brain slice, culture medium was replaced after shooting, then the next day, and then every second day, thereafter.
174
M. Ibrahim et al. / Journal of Neuroscience Methods 101 (2000) 171–179
2.2.2. Preparation of dissection and culture medium These detailed protocols are from Stoppini et al. (1991). 2.2.2.1. Dissection medium. For 100 ml of dissection medium, 1.609 g of Modified Eagles Medium powder (100% MEM) (Gibco, Cat no. 11012-010), containing 25 mM Hepes but no NaHCO3 was weighed out and placed in 50 ml double distilled (dd) H2O. To this solution 10 mM of Tris (120 mg, pH 7.2) and 1 ml of PenStrep (Penicillin and streptomycin) (10 000 U; 10 000 mg/ml) was added. The final volume was adjusted to 100 ml with dd H2O. The final solution was then filtered through a Steritop (Millipore) bottle top filter into a sterile bottle. 2.2.2.2. Culture medium. For 100 ml of culture medium 0.8045 g of MEM (50%) powder was weighed out and placed in 25 ml dd H2O. To this solution was added 5 mM Tris (60 mg) along with 4 mM NaHCO3 (35 mg), pH 7.2, 1 ml of PenStrep, 25 ml Horse Serum and 2.5 ml Hanks salts (10× , Gibco, Cat no. 14060-040). The final volume was adjust to 100 ml with dd H2O and filtered as described above. 2.3. Preparation of gold microcarriers and DNA The protocol outlined is a modification originally used by Sanford et al. (1993).
Fig. 3. (A) The completed chamber had a vacuum cut-off lever (1) and vacuum gauge (2) attached to one end of the chamber leading to a vacuum pump (V.P). (3) An air-inlet valve was used to equalise pressure in the chamber after a vacuum had been applied. It was carried by strong tubing and connected directly to a solenoid valve (4). A quick connect nipple (qcn) attached to the solenoid valve allowed for easy attachment/detachment to a compressed air coupler (cac) on top of the chamber. (5) A time relay (set at 50 ms) controlled the release of He from the solenoid. (B) Top half of the microprojectile chamber showing the placement of perforated aluminium shelves (1,2) into the cut channels (arrow heads). The top shelf (1) (level 1) was for the placement of a dispersion grid (DG) heat-fixed onto a Petri dish. A second shelf (2) (shown slightly pulled out at level 3) was for the placement of the brain slices in preparation for transfection. A filter holder (3), containing the DNA-coated gold microcarriers, was screwed into a metal Luer-lock holder (4). (C) An opened filter holder with DNA-coated gold in the centre of the round filter mesh of the filter holder (seen as the darker region in the centre). (D) A 300-mm brain slice on a Millicell porous membrane, placed into a Petri dish holder containing culture medium. This type of organotypic brain slice receives medium from below whilst the surface remains open to the air.
2.3.1. Preparation of gold microcarriers Freshly prepared gold was used for each week’s shooting. For 120 shots, 60 mg of 0.6-mm gold microparticles (BioRad, Cat no.1652262) was weighed out into a microfuge tube. To this, 1 ml of freshly prepared ethanol (95%) was added and vortexed for 5 min. The gold was allowed to settle down and incubated for 15 min at room temperature, centrifuged and ethanol removed. The gold was repeatedly washed by vortexing and centrifugation in dd H2O. After the third wash the gold was allowed to settle for 1 min, centrifuged and H2O discarded. Finally, 1 ml of sterile 50% glycerol was added to bring the gold microparticle concentration to 60 mg/ml. 2.3.2. Coating DNA onto gold microcarriers The gold prepared above was vortexed for 5 min to allow possible clumps to separate. For six shots, using the same DNA construct, 50 ml (3 mg) of microcarriers was pipetted into a microfuge tube and whilst continually vortexing, 5 ml DNA (1 mg/ml), 50 ml CaCl2 (2.5 M) and 20 ml freshly prepared 0.1 M spermidine (Sigma, Cat no. S0266) was added. The mixture was vortexed for a further 5 min, incubated for 15 min at room temperature, centrifuged and liquid discarded. The DNA-coated gold microcarriers were washed in
M. Ibrahim et al. / Journal of Neuroscience Methods 101 (2000) 171–179
175
gently pipetted onto the unscrewed lower portion of the filter unit which housed the filter mesh (Fig. 3C). The DNA-coated gold microparticles were pipetted, drop by drop, centrally onto half the diameter of the filter mesh and then the filter unit was screwed shut (Fig. 3C). The loaded filter holders were carefully placed upright into an airtight box containing Sikon universal desiccant (Fluka, Cat no. 85112). The DNA-coated gold was then allowed to dry for 20–30 min before shooting.
2.4. Plasmids Enhanced green fluorescent protein (EGFP) (Clontech) was cloned into pCI or pSI mammalian expression vectors (Promega), using either cytomegalovirus (CMV) or simian virus (SV40) enhancers/promoters, respectively, to express EGFP. CMV-EGFP and SV40EGFP DNA were used at concentrations of 1 g/ml. A positive transfection result could be directly visualised under a fluorescence microscope.
2.5. Transfection using the completed microprojectile chamber
Fig. 4. Schematic representation showing the operation of the constructed microprojectile chamber. (A) The pressure from the pressure regulator was set to 1.8 bar (26 psi) using the pressure regulator tap. He (shown in grey) was stopped from entering the chamber by the closed solenoid valve. The air inlet valve was closed and vacuum cut-off lever opened until a vacuum of between 20 and 50 mmHg had been applied. (B) The vacuum cut-off lever is closed and the time relay switch pressed, which opens the solenoid valve for 50 ms, allowing He to enter (shown in grey) and disperse the DNA coated gold particles from the filter holder. Air is allowed back into the chamber by opening the air inlet valve.
250 ml of freshly prepared ethanol (70%), vortexed briefly, centrifuged and liquid discarded. Without disturbing the pellet, the gold was given a final wash with 250 ml of absolute ethanol. Finally, 60 ml of absolute ethanol was added and the gold microcarriers re-suspend by vortexing for 2–3 s. Whilst vortexing, 8-ml aliquots (allowing for ethanol evaporation from the original 60 ml), of DNA-coated gold microcarriers were removed and
All transfections were carried out with the chamber under a laminar flow hood. The filter holder containing the gold microcarriers was gently screwed into the metal Luer-lock adapter (Fig. 3A,B). A metal dispersion grid was placed 10 mm beneath the shooting tip (Fig. 3B), corresponding to the first chamber level. Individually, the brain slices were removed from the medium containing Petri dishes and placed on to a 90-mm Petri dish at level 3 of the chamber (60 mm from the tip) (Fig. 3B) and positioned centrally to the filter tip. The shooting pressure from the pressure regulator was set to 1.8 bar (equivalent to 26 psi) using the pressure regulator tap (Fig. 4). The time relay controlled solenoid valve remained closed stopping He from entering the chamber (Fig. 4A). The air inlet valve was closed (Fig. 4A) and while the vacuum pump was on the vacuum cut-off lever was opened (Fig. 4A). As soon as a vacuum of between 20 and 50 mmHg (Fig. 4A) was applied the vacuum cut-off lever was closed (Fig. 4B). The time relay switch was immediately pressed, which opened the solenoid valve (Fig. 4B) for 50 ms, allowing He to enter, which dispersed the goldcoated DNA particles from the filter holder (Fig. 4B). As soon as the gold was dispersed air was allowed back into the chamber by opening the air inlet valve (Fig. 4B). The brain slice was removed and placed back into a culture dish containing fresh medium. The filter holder was removed and the chamber prepared for the next shot by first clearing the chamber with He by repeatedly opening and closing the solenoid valve by pressing the time relay switch. Brain slices were then put back into the incubator at 37°C.
M. Ibrahim et al. / Journal of Neuroscience Methods 101 (2000) 171–179
176
2.6. Visualisation and fixation of transfected slices
3.2. Distribution of transfected cells
Initial controls of transfections were made under a fluorescence microscope using a × 5 objective. Transfected cells were observed as fluorescent green cells after about 16 h. Brain slices were returned to the incubator and incubated for periods ranging from 2 to 7 days in culture. After these periods brain slices were removed from the incubator and immersion fixed in freshly prepared 4% (w/v) paraformaldehyde in phosphate buffered saline (PBS, 0.1 M, pH 7.4), for 2 h at room temperature. The slices were then repeatedly washed in PBS and mounted onto untreated glass slides. After allowing the tissue to air dry, they were mounted in SlowFade™ (Molecular Probes), coverslipped and sealed with clear nail varnish.
Transfected cells tended to be more numerous in the centre of the shot region compared to the outer borders of the slice (not shown). However, even though transfected cells were not evenly dispersed over the brain slice, individually transfected cells tended to be evenly dispersed from one another, whether in the cortex (Fig. 5A), hippocampus (Fig. 5B) or the cerebellum (Fig. 5C).
3. Results In this study a plan of an inexpensive microprojectile system has been given and furthermore is demonstrated to be very useful in the transfection of organotypic brain slices.
3.1. Number of transfected cells
3.3. Morphology of indi6idual cells Both CMV-EGFP and SV40-EGFP transfected cells showed a range of morphologies (Fig. 6). Both CMVEGFP (Fig. 6A) and SV40-EGFP (Fig. 6C) transfected neurones showed EGFP transport throughout the cells, including their dendritic arbourisations. Transfected oligodendrocytes, undergoing various stages of demyelination were also observed throughout the cultured brain slices, irrespective of the type of vector introduced (Fig. 6B,C). Other cell types transfected, included cerebellar Purkinje cells (Fig. 6D) and also cells which were difficult to identify by morphology alone (not shown).
3.4. Cell transfection and distribution of gold microcarriers
The results show the average from six separate microprojectile shots at each age group. The total number of SV40-EGFP and CMV-EGFP microprojectile transfected cells, in each brain slice culture were counted under a fluorescence microscope. The surface area of individual brain slices was estimated and results given as the average number of cells per unit area (Table 1). There was a large range of transfected cell per unit area, irrespective of the animal age groups or brain region. Nevertheless, some experiments resulted in unsuccessful transfections, which have not been taken into account in the cell count estimates. Also shown is the number of days that the slices were kept incubated after the day of shooting.
Cell transfections were first imaged under a fluorescence microscope and then under brightfield (converted into a negative to give better contrast) to visualise the distribution of gold microcarriers. We were able to examine not only the cells transfected (Fig. 6E,G) but also the dispersion of gold microcarriers (Fig. 6F,H). The methods outlined in this study resulted in a more or less uniform spread of individual gold particles (Fig. 6F). However, both clumping (Fig. 6F) and low dispersion (Fig. 6H) of gold was also observed. It is noteworthy that a uniform spread of gold microcarriers did not necessarily result in more cell transfections.
Table 1 Number of transfected cells per slice surface area (cm2)a Postnatal (P) age of animals when culture was prepared and transfected
P5 P9 P10 P11 P18 a
Days in culture Forebrain
4 7 2 4 4
Cerebellum
SV40-EGFP
CMV-EGFP
SV40-EGFP
CMV-EGFP
117 (122–543) – 191 (57–264) 205 (36–1477) 115 (72–142)
209 (200–222) 289 (32–321) 270 (95–447) 287 (50–407) 254 (189–318)
ND – 865 (613–1013) 158 (67–266) ND
ND 386 (100–650) – 125 (0–167) ND
Ranges of cells transfected per unit area are shown in parentheses; –, unsuccessful transfections; ND, not done.
M. Ibrahim et al. / Journal of Neuroscience Methods 101 (2000) 171–179
Fig. 5. Distribution of transfected cells using the CMV-EGFP (A,B) and SV40-EGFP (C) vectors in various regions of the cultured brain slices. (A) CMV-EGFP transfected cells in the cortex of P10 rat brain slice cultured for 48 h after transfection. (B) CMV-EGFP transfected cells in the hippocampus of P10 rat brain slice cultured for 48 h after transfection. SV40-EGFP transfected cells in the cerebellum of P11 rat after 4 days in culture. Scale bar =100 mm for all panels.
4. Discussion In this study, a microprojectile system initially used on plant tissue and based on the helium flow method of transfection (Takeuchi et al., 1992; Vain et al., 1993) has not only been constructed but also successfully employed to transfect organotypic brain slices. This study also gives concise protocols for the preparation of gold particles (Sanford et al., 1993) and organotypic brain slices (Stoppini et al., 1991).
177
Organotypic slice cultures using the air–medium interface method are becoming popular in the research of different cell populations as they retain expression of certain proteins which may otherwise be lost in cell cultures (Gahwiler et al., 1997). Further still, because of the relatively large (up to 400 mm) thickness of organotypic cultures they allow for greater cell–cell interactions and also retain their three-dimensional cytoarchitecture (Gahwiler et al., 1997). However, organotypic cultures have proven difficult to transfect by viral or liposomal methods, with emphasis of transfections being put on a microprojectile method (Thomas et al., 1998). More recently efficient viral transfection of hippocampal slices has been achieved by combining the method of direct microinjection of a virus followed by the roller tube method of slice culturing (Ehrengruber et al., 1999). However, the roller tube method of culturing results in flattening of slices into a monolayer of cells, leaving the air interface (organotypic) culture technique as the method of choice when conservation of cytoarchitecture is required (Gahwiler et al., 1997). The microprojectile chamber in this study differs from other commercially available ones in a few ways. Firstly, this chamber costs a fraction (about 10%) of those commercially available. In this study another big difference from other methods of microprojectile transfection, is that DNA-coated gold microcarriers are carried by the flow of He. Other methods use a shock of He on the back of macrocarriers (Williams et al., 1991; Arnold and Heintz, 1997; Thomas et al., 1998), on which the DNA-coated gold microcarriers are loaded. Because He acts as the carrier in this home constructed chamber there is no need for macrocarriers. The DNA-coated gold microcarriers are directly loaded onto the inner screen of the filter holder, thus allowing the gold to be carried by the flow of He (Takeuchi et al., 1992; Vain et al., 1993). In the original ‘Particle Inflow Gun’ (PIG) (Vain et al., 1993) a ‘prechamber’ was also used between the solenoid and the He inflow. This prechamber as the name suggests acted as a reservoir for the He before being released by the solenoid. In this study a pressure gauge was directly attached to the He cylinder and pressure carrying the gold microcarriers was no higher than 26 psi, compared to other chambers using 1200–1300 psi to disperse gold from the back of macrocarriers (Thomas et al., 1998). The BioRad gene gun (BioRad) has also been successfully used to transfect organotypic cerebellum slices (Wellmann et al., 1999). This compact gene gun not only functions without the need for macrocarriers but also does not require a vacuum to be applied. Nevertheless the device requires high He pressures of 100– 120 psi which may be a factor in causing a ‘deadzone’ of cells in these cultures (Wellmann et al., 1999).
178
M. Ibrahim et al. / Journal of Neuroscience Methods 101 (2000) 171–179
Further still, different factors such as spreading of gold microcarriers and cell viability play a significant role in successful transfections and appears to be a
universal problem of microprojectile transfections (Sanford et al., 1993; Thomas et al., 1998). The spreading of gold microcarriers can vary considerably using both the
Fig. 6. Individual morphologies of CMV-EGFP (A,B,D,G) and SV40-EGFP (C,E) transfected brain slices from various brain regions at various postnatal ages (P). CMV-EGFP (A) and SV40-EGFP (C) transfected neurones. (B) CMV-EGFP and (C) SV40-EGFP transfected cortical oligodendrocytes exhibiting various stages of demyelination. (D) CMV-EGFP transfected cerebellar Purkinje cell. (E – H) Paired photomicrographs of transfected cells (E,G) and corresponding dispersion of gold microcarriers (F,H) shown in CMV-EGFP (P18) and SV40-EGFP (P11) transfected brain slices. The gold dispersions (seen as white dots) were taken as a brightfield image and then inverted to a negative to give a better contrast. Gold was generally uniformly spread out in the tissue (F), but clumping was still evident (white arrows). Not all gold particles shot in the tissue resulted in a positive transfection. A cortical oligodendrocyte in the cortex (G) transfected even after a low dispersion of gold microcarriers (H). Scale bar in A = 100 mm for A, bar in D=25 mm for B – D and bar in H= 200 mm for E – H.
M. Ibrahim et al. / Journal of Neuroscience Methods 101 (2000) 171–179
commercially available (Thomas et al., 1998; Wellmann et al., 1999) and the home constructed chamber used in this study. It is therefore important to follow the correct procedure for the preparation of the gold and we have found that by allowing the DNA-coated gold microcarriers to dry properly, this decreases the clumping of gold resulting in an increase in the number of transfected cells. However, the particle clumping cannot be completely avoided, but placing the prepared filter holders into a container with desiccant to allow for better drying, appeared to further reduce gold clumping. In this study the dispersion grids from BioRad resulted in gold dispersion comparable to that of the nylon meshes used in the hand-held gene gun (Wellmann et al., 1999). However, gold clumping was not likely to be the only factor in successful transfections but may also depend on the viability of individual cells. Cut slices release toxins that have to be washed away (Gahwiler et al., 1997). Therefore it is important to make sure that the slices are properly bathed in clean dissection medium and left to rest in an incubator for at least 4 h before transfecting. Transfections can also be carried out many days after the slices have been prepared (Thomas et al., 1998; Wellmann et al., 1999), but we have noticed that in longer term cultures the surface of the brain slices dry and form a crust. Our observations indicate that removal of this dry layer over the tissue by gently pipetting some dissection medium on to the surface and then aspirating it off again improves transfection numbers. This home constructed system is built in such a way that the chamber allows for the transfection of larger brain sections or multiple sections placed on single culture inserts, placed at various shooting levels, with adjustments made to the He pressure. Although so far only used in the transfection of organotypic brain slice cultures, this home constructed chamber, may well be efficient in transfecting other animal tissues or even dissociated cell cultures. By using enhancers/promoters, such as CMV and SV40, cell-specific expression could not be controlled, but cell-specific promoters may enable expression in specific cell groups to be transfected (Arnold and Heintz, 1997). Furthermore, as has been shown in this study, transfected cells can be kept alive in an incubator for long periods, therefore further opening the possibility of following changes in individual cell morphologies such as those of oligodendrocytes undergoing demyelination.
Acknowledgements This work was supported by the Swiss National
179
Foundation (NF 3100-047291.96) and Novartis Ltd. We thank the ETH, Zurich for the original design of the microprojectile chamber and R. Graff and J. Gugler from the Institute of Plant Biology, Fribourg, for the construction of the chamber. Thanks also to A. Haunsø for critical reading of the manuscript and Dr Rienk Tuinhof for graphical work on Fig. 4.
References Arnold DB, Heintz N. A calcium responsive element that regulates expression of two calcium binding proteins in Purkinje cells. Proc Natl Acad Sci USA 1997;94:8842 – 7. Arnold D, Feng L, Kim J, Heintz N. A strategy for the analysis of gene expression during neural development. Proc Natl Acad Sci USA 1994;91:9970 – 4. Christou P, McCabe DE, Swain WF. Stable transformation of soyabean callus by DNA-coated gold particles. Plant Physiol 1988;87:671 – 4. Ehrengruber MU, Lundstrom K, Schweitzer C, Heuss C, Schlesinger S, Gahwiler BH. Recombinant Semliki Forest virus and Sindbis virus efficiently infect neurons in hippocampal slice cultures. Proc Natl Acad Sci USA 1999;96:7041 – 6. Gahwiler BH, Capogna M, Debanne D, McKinney RA, Thompson SM. Organotypic slice cultures: a technique has come of age. Trends Neurosci 1997;20:471 – 7. Klein TM, Fromm M, Weissinger A, Tomes D, Schaff S, Sletten M, Sanford JC. Transfer of foreign genes into intact maize cells with high-velocity microprojectiles. Proc Natl Acad Sci USA 1988;85:4305 – 9. Lo DC, McAllister AK, Katz LC. Neuronal transfection in brain slices using particle-mediated gene transfer. Neuron 1994;13:1263 – 8. Sanford JC, Smith FD, Russell JA. Optimizing the biolistic process for different biological applications. Methods Enzymol 1993;217:483 – 509. Sautter C, Waldner H, Neuhaus-Url G, Galli A, Nuehaus G, Potrykus I. Micro-targeting: High efficiency gene transfer using a novel approach for the acceleration of micro-projectiles. Biotechnology 1991;9:1080 – 5. Stoppini L, Buchs PA, Muller D. A simple method for organotypic cultures of nervous tissue. J Neurosci Methods 1991;37:173–82. Takeuchi Y, Dotson M, Keen NT. Plant transformation: A simple particle bombardment device based on flowing helium. Plant Mol Biol 1992;18:835 – 9. Thomas A, Kim DS, Fields RL, Chin H, Gainer H. Quantitative analysis of gene expression in organotypic slice-explant cultures by particle-mediated gene transfer. J Neurosci Methods 1998;84:181 – 91. Vain P, Keen N, Murillo J, Rathus C, Nemes C, Finer JJ. Development of the particle inflow gun. Plant Cell Tissue Organ Cult 1993;33:237 – 46. Wellmann H, Kaltschmidt B, Kaltschmidt C. Optimized protocol for biolistic transfection of brain slices and dissociated cultured neurons with a hand-held gene gun. J Neurosci Methods 1999;92:55– 64. Williams RS, Johnston SA, Riedy M, DeVit MJ, McElligott SG, Sanford JC. Introduction of foreign genes into tissues of living mice by DNA-coated microprojectiles. Proc Natl Acad Sci USA 1991;88:2726 – 30.
Construction and application of a microprojectile system for the transfection of organotypic brain slices Merdol Ibrahim a,*, Azeddine Si-Ammour b, Marco R. Celio a, Felix Mauch b, Pierre-Alain Menoud a a
Institute of Histology and General Embryology, Uni6ersity of Fribourg, CH-1705 Fribourg, Switzerland b Institute of Plant Biology, Uni6ersity of Fribourg, CH-1705 Fribourg, Switzerland Received 17 March 2000; received in revised form 19 June 2000; accepted 21 June 2000
Abstract In this study we outline a method for constructing an inexpensive chamber used in the transfection of organotypic brain slices. This chamber differs from most commercially available chambers in that DNA-coated gold microcarriers are directly carried by a flow of helium at low pressure (26 psi). Most other chambers employ macrocarriers onto which DNA-coated gold is first loaded, and then released by a shock of helium onto the reverse side of the macrocarriers. This home constructed device has been successfully employed in the transfection of organotypic brain slices cultured using the air – medium interface method. Mammalian expression vectors containing cytomegalovirus (CMV) and simian virus (SV40) enhancers/promoters were used to express enhanced green fluorescence protein (EGFP). DNA was coated onto 0.6-mm gold microcarriers. Transfected cells were visualised under a fluorescence microscope and included identifiable neurones and oligodendrocytes. Also included in this study are step-by-step methods for the preparation of gold microcarriers and organotypic brain slices. © 2000 Elsevier Science B.V. All rights reserved. Keywords: Cytomegalovirus; Simian virus; Green fluorescent protein; Gene gun; Neurone; Oligodendrocyte
1. Introduction Microprojectile methods of gene transfer have been successfully used to transform a variety of plant (Christou et al., 1988; Klein et al., 1988) and animal tissue (Williams et al., 1991). However, these transfection techniques have more recently been used in the field of neuroscience, with interest focusing on the transfection of organotypic brain slice cultures (Arnold et al., 1994; Lo et al., 1994). Many versions of the microprojectile system have been constructed (Christou et al., 1988; Sautter et al., 1991; Takeuchi et al., 1992) and commercially available ones are now being used periodically (Williams et al., 1991; Arnold and Heintz, 1997; Thomas et al., 1998; Wellmann et al., 1999). However, the process of transfection remains either too cumbersome, uses too many consumables and or more importantly the appara-
tus itself is too expensive. An inexpensive particle system used in plant transfections was initially described by Takeuchi et al. (1992), and then further refined by Vain et al. (1993). This ‘Particle Inflow Gun’ (PIG) as termed by Vain et al. (1993) was originally used in the transfection of plant material and was the basis on which the system in this study was based. In addition, the home constructed microprojectile chamber in this study is demonstrated as a successful tool in the transfections of organotypic brain slice cultures, employing the air– medium interface method (Stoppini et al., 1991), using mammalian cytomegalovirus (CMV) and simian virus (SV40) driving the expression of enhanced green fluorescence protein (EGFP). 2. Materials and methods
2.1. Construction of the microprojectile chamber * Corresponding author. Tel.: +41-26-3008490; fax: + 41-263009732. E-mail address: [email protected] (M. Ibrahim).
The full plans and dimensions to construct the microprojectile chamber are shown in Figs. 1 and 2. Note
0165-0270/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 5 - 0 2 7 0 ( 0 0 ) 0 0 2 6 9 - 7
172
M. Ibrahim et al. / Journal of Neuroscience Methods 101 (2000) 171–179
that catalogue numbers have been given to chamber parts that are thought to be necessary. All other parts not indicated can be purchased directly from any local ‘hardware store’. Initially a five-sided chamber was constructed from 12-mm-thick aluminium sheets (Fig. 1). The top panel measured 140×108× 12 mm; two side panels measured 350× 108×12 mm (Fig. 1) and the back panel measured 350 ×140× 12 mm (Fig. 1). A smaller, lower panel measuring 116×108 ×12 mm was placed within the chamber, 14 mm from the bottom (Fig. 1). Before constructing the chamber the two side panels had 2-mm channels, 25 mm apart cut into them (starting 70 mm from the bottom), running the full length of the panels (Figs. 1 and 3A,B). These channels were used for holding perforated aluminium shelves (119× 105 × 1.8 mm) (Fig. 3A,B) in place and for allowing the possibility of adjusting the distance of the shot (Fig. 3A,B). The lower
Fig. 1. Schematic plan and measurements for the construction of the microprojectile chamber. A five-sided panel was initially constructed from 12-mm aluminium and then an overhanging panel (coloured grey) was screwed into place.
ends of the side panels each had a 0.6-mm (14¦) hole drilled into them (Fig. 1) to allow for the attachment of a vacuum pump and air-inlet valve (screw valve) (Figs. 3A and 4). The top panel also had a 0.6-mm hole drilled into its centre and was fitted with a double-threaded (14¦) connector (Bru¨tsch Ru¨egger, CH, cat no. 97542) (Fig. 4). To the top end of the double-threaded connector a compressed air coupler was screwed into place (Bru¨tsch Ru¨egger, CH, cat no. 97531 R14¦) (Fig. 4). This air coupler allowed for the connection of a solenoid valve (Carlo Gavazzi Steinhausen, CH, cat no. SCG262CO14) (Figs. 3A,B and 4) which was connected to a stainless steel quick connect nipple (Bru¨tsch Ru¨egger, CH, cat no. 97534 R14¦) (Figs. 3A,B and 4). To the inner side of the top panel, a stainless steel Luer-lock (Sartorius, cat no. SM17122) was screwed into place for the attachment of a 13-mm Swinney filter holder (Gelman Sciences, cat no. 4317) (Figs. 3A,B and 4). The chamber was initially glued and then held together with 5-mm metal screws (Fig. 1). The five-sided chamber then had an overhanging front panel screwed into place (Figs. 1 and 2). The overhanging front panel was made from an aluminium sheet measuring 370×160× 12 mm, with a middle portion measuring 320× 120 mm cut out (Fig. 2A). From the outer edges (2 mm) of the panel a continuous 3.5-mm channel was made (Fig. 2A), into which a 4-mm rubber ring was glued (Fig. 3A,B), to give an airtight seal to the chamber when a vacuum was applied. Channels, 25 mm apart and 70 mm from the bottom of the panel were made (Fig. 2A), corresponding to those of the inner side panels (Fig. 1). The overhanging front panel was then glued and screwed into place (Figs. 1 and 3A). Finally, a door was made from 15-mm-thick Plexiglass (Figs. 2B and 3A), having the same dimensions (370×160 mm) as the front overhanging panel. To allow the Plexiglass to sit over the overhanging front panel a 2-mm aluminium support was bent at a right angle (Fig. 2B) and screwed to the top of the Plexiglass. Handles were attached to the Plexiglass to allow for easy removal and positioning (Figs. 2B and 3A). To complete the chamber a vacuum cut-off lever was screwed onto one side of the chamber (Fig. 3A). A cross fitting vacuum gauge (Tribold, CH, cat no. 41801213) was attached to the end of the vacuum cut-off lever, leading via flexible reinforced tubing to a vacuum pump capable of applying a vacuum of 20 mmHg (Fig. 3A). To the opposite side panel, an air-inlet valve (screw valve) was attached to enable air back into the chamber once a vacuum was applied (Fig. 3A).From the top of the chamber, helium (He) was carried by steel mesh reinforced flexible pressure hose, to the solenoid, from a He cylinder. The He cylinder had a pressure regulator attached (Fig. 4) (Gloor, CH, cat no. 1500-8) enabling a pre-defined working pressure to be used. The timing of the release of He from the solenoid
M. Ibrahim et al. / Journal of Neuroscience Methods 101 (2000) 171–179
173
Fig. 2. Schematic plans and measurements for the overhanging aluminium front panel (A) and Plexiglass door (B).
was controlled by a time relay control (Fig. 3A) (Bircher, CH Cat no. TRAB230AC) set at 50 ms. A metal dispersion grid (Fig. 3B) (BioRad, cat no. 1652336) was heat-fixed onto a 90-mm Petri dish lid, which had a corresponding hole cut in its centre. The dispersion grid along with the Petri dish was changed when using different DNA constructs. In our studies, transfections were optimal when a dispersion grid was placed at the first level (Fig. 3A,B) and the brain slice culture placed at the third level of the chamber (Fig. 3A,B). The Swinney filter holders screwed directly into the inner Luer-lock holder (Fig. 3A,B). In the final set-up, the distance from the tip of the filter holder to the dispersion net was 10 mm and 60 mm to the brain slice culture.
2.2. Organotypic brain slices 2.2.1. Preparation of organotypic brain slices Under aseptic conditions, rat pups ranging from postnatal (P) day 5 to 18 were decapitated and brains quickly removed into a Petri dish containing cold dissection medium (see below). Using a binocular microscope, the brain was held steady using forceps and initially bifurcated coronally near the superior and inferior colli-
culi. The forebrain was removed from the dissection medium and placed bifurcated side down onto sterile blotting paper. The brain was then glued onto a cutting block and transferred to a vibratome. More fresh dissection medium was added to the vibratome chamber and 300-mm coronal sections were taken throughout the forebrain. Slices were placed into cold dissection medium for a further 30 min. Cerebellar slices were also cut into 300-mm sagittal sections using a McIlwain tissue chopper (McIlwain, Mickle Laboratory, Cambridge). The sectioned cerebellum was transferred to fresh cold dissection medium and the cut folia separated using blunt forceps. The sections were allowed to rest in cold dissection medium for a further 30 min. Up to four cerebellar folia or 1–2 forebrain slices were then transferred to 30-mm Millicell culture inserts (Millipore, cat no. PICM03050), which in turn were placed into individual Petri dishes (Fig. 3D). All dissection medium residues were pipetted from the well and surrounding tissue. Finally, 1 ml of culture medium was added, so that the cut tissue remained open to the air. Brain slices were then placed into an incubator at 37°C containing 5% CO2. For each brain slice, culture medium was replaced after shooting, then the next day, and then every second day, thereafter.
174
M. Ibrahim et al. / Journal of Neuroscience Methods 101 (2000) 171–179
2.2.2. Preparation of dissection and culture medium These detailed protocols are from Stoppini et al. (1991). 2.2.2.1. Dissection medium. For 100 ml of dissection medium, 1.609 g of Modified Eagles Medium powder (100% MEM) (Gibco, Cat no. 11012-010), containing 25 mM Hepes but no NaHCO3 was weighed out and placed in 50 ml double distilled (dd) H2O. To this solution 10 mM of Tris (120 mg, pH 7.2) and 1 ml of PenStrep (Penicillin and streptomycin) (10 000 U; 10 000 mg/ml) was added. The final volume was adjusted to 100 ml with dd H2O. The final solution was then filtered through a Steritop (Millipore) bottle top filter into a sterile bottle. 2.2.2.2. Culture medium. For 100 ml of culture medium 0.8045 g of MEM (50%) powder was weighed out and placed in 25 ml dd H2O. To this solution was added 5 mM Tris (60 mg) along with 4 mM NaHCO3 (35 mg), pH 7.2, 1 ml of PenStrep, 25 ml Horse Serum and 2.5 ml Hanks salts (10× , Gibco, Cat no. 14060-040). The final volume was adjust to 100 ml with dd H2O and filtered as described above. 2.3. Preparation of gold microcarriers and DNA The protocol outlined is a modification originally used by Sanford et al. (1993).
Fig. 3. (A) The completed chamber had a vacuum cut-off lever (1) and vacuum gauge (2) attached to one end of the chamber leading to a vacuum pump (V.P). (3) An air-inlet valve was used to equalise pressure in the chamber after a vacuum had been applied. It was carried by strong tubing and connected directly to a solenoid valve (4). A quick connect nipple (qcn) attached to the solenoid valve allowed for easy attachment/detachment to a compressed air coupler (cac) on top of the chamber. (5) A time relay (set at 50 ms) controlled the release of He from the solenoid. (B) Top half of the microprojectile chamber showing the placement of perforated aluminium shelves (1,2) into the cut channels (arrow heads). The top shelf (1) (level 1) was for the placement of a dispersion grid (DG) heat-fixed onto a Petri dish. A second shelf (2) (shown slightly pulled out at level 3) was for the placement of the brain slices in preparation for transfection. A filter holder (3), containing the DNA-coated gold microcarriers, was screwed into a metal Luer-lock holder (4). (C) An opened filter holder with DNA-coated gold in the centre of the round filter mesh of the filter holder (seen as the darker region in the centre). (D) A 300-mm brain slice on a Millicell porous membrane, placed into a Petri dish holder containing culture medium. This type of organotypic brain slice receives medium from below whilst the surface remains open to the air.
2.3.1. Preparation of gold microcarriers Freshly prepared gold was used for each week’s shooting. For 120 shots, 60 mg of 0.6-mm gold microparticles (BioRad, Cat no.1652262) was weighed out into a microfuge tube. To this, 1 ml of freshly prepared ethanol (95%) was added and vortexed for 5 min. The gold was allowed to settle down and incubated for 15 min at room temperature, centrifuged and ethanol removed. The gold was repeatedly washed by vortexing and centrifugation in dd H2O. After the third wash the gold was allowed to settle for 1 min, centrifuged and H2O discarded. Finally, 1 ml of sterile 50% glycerol was added to bring the gold microparticle concentration to 60 mg/ml. 2.3.2. Coating DNA onto gold microcarriers The gold prepared above was vortexed for 5 min to allow possible clumps to separate. For six shots, using the same DNA construct, 50 ml (3 mg) of microcarriers was pipetted into a microfuge tube and whilst continually vortexing, 5 ml DNA (1 mg/ml), 50 ml CaCl2 (2.5 M) and 20 ml freshly prepared 0.1 M spermidine (Sigma, Cat no. S0266) was added. The mixture was vortexed for a further 5 min, incubated for 15 min at room temperature, centrifuged and liquid discarded. The DNA-coated gold microcarriers were washed in
M. Ibrahim et al. / Journal of Neuroscience Methods 101 (2000) 171–179
175
gently pipetted onto the unscrewed lower portion of the filter unit which housed the filter mesh (Fig. 3C). The DNA-coated gold microparticles were pipetted, drop by drop, centrally onto half the diameter of the filter mesh and then the filter unit was screwed shut (Fig. 3C). The loaded filter holders were carefully placed upright into an airtight box containing Sikon universal desiccant (Fluka, Cat no. 85112). The DNA-coated gold was then allowed to dry for 20–30 min before shooting.
2.4. Plasmids Enhanced green fluorescent protein (EGFP) (Clontech) was cloned into pCI or pSI mammalian expression vectors (Promega), using either cytomegalovirus (CMV) or simian virus (SV40) enhancers/promoters, respectively, to express EGFP. CMV-EGFP and SV40EGFP DNA were used at concentrations of 1 g/ml. A positive transfection result could be directly visualised under a fluorescence microscope.
2.5. Transfection using the completed microprojectile chamber
Fig. 4. Schematic representation showing the operation of the constructed microprojectile chamber. (A) The pressure from the pressure regulator was set to 1.8 bar (26 psi) using the pressure regulator tap. He (shown in grey) was stopped from entering the chamber by the closed solenoid valve. The air inlet valve was closed and vacuum cut-off lever opened until a vacuum of between 20 and 50 mmHg had been applied. (B) The vacuum cut-off lever is closed and the time relay switch pressed, which opens the solenoid valve for 50 ms, allowing He to enter (shown in grey) and disperse the DNA coated gold particles from the filter holder. Air is allowed back into the chamber by opening the air inlet valve.
250 ml of freshly prepared ethanol (70%), vortexed briefly, centrifuged and liquid discarded. Without disturbing the pellet, the gold was given a final wash with 250 ml of absolute ethanol. Finally, 60 ml of absolute ethanol was added and the gold microcarriers re-suspend by vortexing for 2–3 s. Whilst vortexing, 8-ml aliquots (allowing for ethanol evaporation from the original 60 ml), of DNA-coated gold microcarriers were removed and
All transfections were carried out with the chamber under a laminar flow hood. The filter holder containing the gold microcarriers was gently screwed into the metal Luer-lock adapter (Fig. 3A,B). A metal dispersion grid was placed 10 mm beneath the shooting tip (Fig. 3B), corresponding to the first chamber level. Individually, the brain slices were removed from the medium containing Petri dishes and placed on to a 90-mm Petri dish at level 3 of the chamber (60 mm from the tip) (Fig. 3B) and positioned centrally to the filter tip. The shooting pressure from the pressure regulator was set to 1.8 bar (equivalent to 26 psi) using the pressure regulator tap (Fig. 4). The time relay controlled solenoid valve remained closed stopping He from entering the chamber (Fig. 4A). The air inlet valve was closed (Fig. 4A) and while the vacuum pump was on the vacuum cut-off lever was opened (Fig. 4A). As soon as a vacuum of between 20 and 50 mmHg (Fig. 4A) was applied the vacuum cut-off lever was closed (Fig. 4B). The time relay switch was immediately pressed, which opened the solenoid valve (Fig. 4B) for 50 ms, allowing He to enter, which dispersed the goldcoated DNA particles from the filter holder (Fig. 4B). As soon as the gold was dispersed air was allowed back into the chamber by opening the air inlet valve (Fig. 4B). The brain slice was removed and placed back into a culture dish containing fresh medium. The filter holder was removed and the chamber prepared for the next shot by first clearing the chamber with He by repeatedly opening and closing the solenoid valve by pressing the time relay switch. Brain slices were then put back into the incubator at 37°C.
M. Ibrahim et al. / Journal of Neuroscience Methods 101 (2000) 171–179
176
2.6. Visualisation and fixation of transfected slices
3.2. Distribution of transfected cells
Initial controls of transfections were made under a fluorescence microscope using a × 5 objective. Transfected cells were observed as fluorescent green cells after about 16 h. Brain slices were returned to the incubator and incubated for periods ranging from 2 to 7 days in culture. After these periods brain slices were removed from the incubator and immersion fixed in freshly prepared 4% (w/v) paraformaldehyde in phosphate buffered saline (PBS, 0.1 M, pH 7.4), for 2 h at room temperature. The slices were then repeatedly washed in PBS and mounted onto untreated glass slides. After allowing the tissue to air dry, they were mounted in SlowFade™ (Molecular Probes), coverslipped and sealed with clear nail varnish.
Transfected cells tended to be more numerous in the centre of the shot region compared to the outer borders of the slice (not shown). However, even though transfected cells were not evenly dispersed over the brain slice, individually transfected cells tended to be evenly dispersed from one another, whether in the cortex (Fig. 5A), hippocampus (Fig. 5B) or the cerebellum (Fig. 5C).
3. Results In this study a plan of an inexpensive microprojectile system has been given and furthermore is demonstrated to be very useful in the transfection of organotypic brain slices.
3.1. Number of transfected cells
3.3. Morphology of indi6idual cells Both CMV-EGFP and SV40-EGFP transfected cells showed a range of morphologies (Fig. 6). Both CMVEGFP (Fig. 6A) and SV40-EGFP (Fig. 6C) transfected neurones showed EGFP transport throughout the cells, including their dendritic arbourisations. Transfected oligodendrocytes, undergoing various stages of demyelination were also observed throughout the cultured brain slices, irrespective of the type of vector introduced (Fig. 6B,C). Other cell types transfected, included cerebellar Purkinje cells (Fig. 6D) and also cells which were difficult to identify by morphology alone (not shown).
3.4. Cell transfection and distribution of gold microcarriers
The results show the average from six separate microprojectile shots at each age group. The total number of SV40-EGFP and CMV-EGFP microprojectile transfected cells, in each brain slice culture were counted under a fluorescence microscope. The surface area of individual brain slices was estimated and results given as the average number of cells per unit area (Table 1). There was a large range of transfected cell per unit area, irrespective of the animal age groups or brain region. Nevertheless, some experiments resulted in unsuccessful transfections, which have not been taken into account in the cell count estimates. Also shown is the number of days that the slices were kept incubated after the day of shooting.
Cell transfections were first imaged under a fluorescence microscope and then under brightfield (converted into a negative to give better contrast) to visualise the distribution of gold microcarriers. We were able to examine not only the cells transfected (Fig. 6E,G) but also the dispersion of gold microcarriers (Fig. 6F,H). The methods outlined in this study resulted in a more or less uniform spread of individual gold particles (Fig. 6F). However, both clumping (Fig. 6F) and low dispersion (Fig. 6H) of gold was also observed. It is noteworthy that a uniform spread of gold microcarriers did not necessarily result in more cell transfections.
Table 1 Number of transfected cells per slice surface area (cm2)a Postnatal (P) age of animals when culture was prepared and transfected
P5 P9 P10 P11 P18 a
Days in culture Forebrain
4 7 2 4 4
Cerebellum
SV40-EGFP
CMV-EGFP
SV40-EGFP
CMV-EGFP
117 (122–543) – 191 (57–264) 205 (36–1477) 115 (72–142)
209 (200–222) 289 (32–321) 270 (95–447) 287 (50–407) 254 (189–318)
ND – 865 (613–1013) 158 (67–266) ND
ND 386 (100–650) – 125 (0–167) ND
Ranges of cells transfected per unit area are shown in parentheses; –, unsuccessful transfections; ND, not done.
M. Ibrahim et al. / Journal of Neuroscience Methods 101 (2000) 171–179
Fig. 5. Distribution of transfected cells using the CMV-EGFP (A,B) and SV40-EGFP (C) vectors in various regions of the cultured brain slices. (A) CMV-EGFP transfected cells in the cortex of P10 rat brain slice cultured for 48 h after transfection. (B) CMV-EGFP transfected cells in the hippocampus of P10 rat brain slice cultured for 48 h after transfection. SV40-EGFP transfected cells in the cerebellum of P11 rat after 4 days in culture. Scale bar =100 mm for all panels.
4. Discussion In this study, a microprojectile system initially used on plant tissue and based on the helium flow method of transfection (Takeuchi et al., 1992; Vain et al., 1993) has not only been constructed but also successfully employed to transfect organotypic brain slices. This study also gives concise protocols for the preparation of gold particles (Sanford et al., 1993) and organotypic brain slices (Stoppini et al., 1991).
177
Organotypic slice cultures using the air–medium interface method are becoming popular in the research of different cell populations as they retain expression of certain proteins which may otherwise be lost in cell cultures (Gahwiler et al., 1997). Further still, because of the relatively large (up to 400 mm) thickness of organotypic cultures they allow for greater cell–cell interactions and also retain their three-dimensional cytoarchitecture (Gahwiler et al., 1997). However, organotypic cultures have proven difficult to transfect by viral or liposomal methods, with emphasis of transfections being put on a microprojectile method (Thomas et al., 1998). More recently efficient viral transfection of hippocampal slices has been achieved by combining the method of direct microinjection of a virus followed by the roller tube method of slice culturing (Ehrengruber et al., 1999). However, the roller tube method of culturing results in flattening of slices into a monolayer of cells, leaving the air interface (organotypic) culture technique as the method of choice when conservation of cytoarchitecture is required (Gahwiler et al., 1997). The microprojectile chamber in this study differs from other commercially available ones in a few ways. Firstly, this chamber costs a fraction (about 10%) of those commercially available. In this study another big difference from other methods of microprojectile transfection, is that DNA-coated gold microcarriers are carried by the flow of He. Other methods use a shock of He on the back of macrocarriers (Williams et al., 1991; Arnold and Heintz, 1997; Thomas et al., 1998), on which the DNA-coated gold microcarriers are loaded. Because He acts as the carrier in this home constructed chamber there is no need for macrocarriers. The DNA-coated gold microcarriers are directly loaded onto the inner screen of the filter holder, thus allowing the gold to be carried by the flow of He (Takeuchi et al., 1992; Vain et al., 1993). In the original ‘Particle Inflow Gun’ (PIG) (Vain et al., 1993) a ‘prechamber’ was also used between the solenoid and the He inflow. This prechamber as the name suggests acted as a reservoir for the He before being released by the solenoid. In this study a pressure gauge was directly attached to the He cylinder and pressure carrying the gold microcarriers was no higher than 26 psi, compared to other chambers using 1200–1300 psi to disperse gold from the back of macrocarriers (Thomas et al., 1998). The BioRad gene gun (BioRad) has also been successfully used to transfect organotypic cerebellum slices (Wellmann et al., 1999). This compact gene gun not only functions without the need for macrocarriers but also does not require a vacuum to be applied. Nevertheless the device requires high He pressures of 100– 120 psi which may be a factor in causing a ‘deadzone’ of cells in these cultures (Wellmann et al., 1999).
178
M. Ibrahim et al. / Journal of Neuroscience Methods 101 (2000) 171–179
Further still, different factors such as spreading of gold microcarriers and cell viability play a significant role in successful transfections and appears to be a
universal problem of microprojectile transfections (Sanford et al., 1993; Thomas et al., 1998). The spreading of gold microcarriers can vary considerably using both the
Fig. 6. Individual morphologies of CMV-EGFP (A,B,D,G) and SV40-EGFP (C,E) transfected brain slices from various brain regions at various postnatal ages (P). CMV-EGFP (A) and SV40-EGFP (C) transfected neurones. (B) CMV-EGFP and (C) SV40-EGFP transfected cortical oligodendrocytes exhibiting various stages of demyelination. (D) CMV-EGFP transfected cerebellar Purkinje cell. (E – H) Paired photomicrographs of transfected cells (E,G) and corresponding dispersion of gold microcarriers (F,H) shown in CMV-EGFP (P18) and SV40-EGFP (P11) transfected brain slices. The gold dispersions (seen as white dots) were taken as a brightfield image and then inverted to a negative to give a better contrast. Gold was generally uniformly spread out in the tissue (F), but clumping was still evident (white arrows). Not all gold particles shot in the tissue resulted in a positive transfection. A cortical oligodendrocyte in the cortex (G) transfected even after a low dispersion of gold microcarriers (H). Scale bar in A = 100 mm for A, bar in D=25 mm for B – D and bar in H= 200 mm for E – H.
M. Ibrahim et al. / Journal of Neuroscience Methods 101 (2000) 171–179
commercially available (Thomas et al., 1998; Wellmann et al., 1999) and the home constructed chamber used in this study. It is therefore important to follow the correct procedure for the preparation of the gold and we have found that by allowing the DNA-coated gold microcarriers to dry properly, this decreases the clumping of gold resulting in an increase in the number of transfected cells. However, the particle clumping cannot be completely avoided, but placing the prepared filter holders into a container with desiccant to allow for better drying, appeared to further reduce gold clumping. In this study the dispersion grids from BioRad resulted in gold dispersion comparable to that of the nylon meshes used in the hand-held gene gun (Wellmann et al., 1999). However, gold clumping was not likely to be the only factor in successful transfections but may also depend on the viability of individual cells. Cut slices release toxins that have to be washed away (Gahwiler et al., 1997). Therefore it is important to make sure that the slices are properly bathed in clean dissection medium and left to rest in an incubator for at least 4 h before transfecting. Transfections can also be carried out many days after the slices have been prepared (Thomas et al., 1998; Wellmann et al., 1999), but we have noticed that in longer term cultures the surface of the brain slices dry and form a crust. Our observations indicate that removal of this dry layer over the tissue by gently pipetting some dissection medium on to the surface and then aspirating it off again improves transfection numbers. This home constructed system is built in such a way that the chamber allows for the transfection of larger brain sections or multiple sections placed on single culture inserts, placed at various shooting levels, with adjustments made to the He pressure. Although so far only used in the transfection of organotypic brain slice cultures, this home constructed chamber, may well be efficient in transfecting other animal tissues or even dissociated cell cultures. By using enhancers/promoters, such as CMV and SV40, cell-specific expression could not be controlled, but cell-specific promoters may enable expression in specific cell groups to be transfected (Arnold and Heintz, 1997). Furthermore, as has been shown in this study, transfected cells can be kept alive in an incubator for long periods, therefore further opening the possibility of following changes in individual cell morphologies such as those of oligodendrocytes undergoing demyelination.
Acknowledgements This work was supported by the Swiss National
179
Foundation (NF 3100-047291.96) and Novartis Ltd. We thank the ETH, Zurich for the original design of the microprojectile chamber and R. Graff and J. Gugler from the Institute of Plant Biology, Fribourg, for the construction of the chamber. Thanks also to A. Haunsø for critical reading of the manuscript and Dr Rienk Tuinhof for graphical work on Fig. 4.
References Arnold DB, Heintz N. A calcium responsive element that regulates expression of two calcium binding proteins in Purkinje cells. Proc Natl Acad Sci USA 1997;94:8842 – 7. Arnold D, Feng L, Kim J, Heintz N. A strategy for the analysis of gene expression during neural development. Proc Natl Acad Sci USA 1994;91:9970 – 4. Christou P, McCabe DE, Swain WF. Stable transformation of soyabean callus by DNA-coated gold particles. Plant Physiol 1988;87:671 – 4. Ehrengruber MU, Lundstrom K, Schweitzer C, Heuss C, Schlesinger S, Gahwiler BH. Recombinant Semliki Forest virus and Sindbis virus efficiently infect neurons in hippocampal slice cultures. Proc Natl Acad Sci USA 1999;96:7041 – 6. Gahwiler BH, Capogna M, Debanne D, McKinney RA, Thompson SM. Organotypic slice cultures: a technique has come of age. Trends Neurosci 1997;20:471 – 7. Klein TM, Fromm M, Weissinger A, Tomes D, Schaff S, Sletten M, Sanford JC. Transfer of foreign genes into intact maize cells with high-velocity microprojectiles. Proc Natl Acad Sci USA 1988;85:4305 – 9. Lo DC, McAllister AK, Katz LC. Neuronal transfection in brain slices using particle-mediated gene transfer. Neuron 1994;13:1263 – 8. Sanford JC, Smith FD, Russell JA. Optimizing the biolistic process for different biological applications. Methods Enzymol 1993;217:483 – 509. Sautter C, Waldner H, Neuhaus-Url G, Galli A, Nuehaus G, Potrykus I. Micro-targeting: High efficiency gene transfer using a novel approach for the acceleration of micro-projectiles. Biotechnology 1991;9:1080 – 5. Stoppini L, Buchs PA, Muller D. A simple method for organotypic cultures of nervous tissue. J Neurosci Methods 1991;37:173–82. Takeuchi Y, Dotson M, Keen NT. Plant transformation: A simple particle bombardment device based on flowing helium. Plant Mol Biol 1992;18:835 – 9. Thomas A, Kim DS, Fields RL, Chin H, Gainer H. Quantitative analysis of gene expression in organotypic slice-explant cultures by particle-mediated gene transfer. J Neurosci Methods 1998;84:181 – 91. Vain P, Keen N, Murillo J, Rathus C, Nemes C, Finer JJ. Development of the particle inflow gun. Plant Cell Tissue Organ Cult 1993;33:237 – 46. Wellmann H, Kaltschmidt B, Kaltschmidt C. Optimized protocol for biolistic transfection of brain slices and dissociated cultured neurons with a hand-held gene gun. J Neurosci Methods 1999;92:55– 64. Williams RS, Johnston SA, Riedy M, DeVit MJ, McElligott SG, Sanford JC. Introduction of foreign genes into tissues of living mice by DNA-coated microprojectiles. Proc Natl Acad Sci USA 1991;88:2726 – 30.