The Mechanics And Biology Of Biolistic Transfection

A. Microparticles

The need for ''bullets'' or microparticles in this technology relates to the fact that kinetic energy depends on both mass and velocity (p = m X v). Therefore, applying velocity to a low mass nucleic acid has a low likelihood of imparting sufficient kinetic energy to the molecule to drive it through the membrane and protein structures of a living cell. To circumvent this, nucleic acids and proteins are attached to dense carriers usually in the form of microparticles or ''microbul-lets'' with diameters from 0.25 to 5 ^m(1). Metal microparticles are the carrier of choice for biolistic gene delivery because they have unparalleled density to maximize cell and tissue penetration while minimizing the size of particle needed to achieve this mass. A variety of metals including tungsten and depleted uranium, have been used for biolistics (1,5-7,14). Although any metal could theoretically be used, gold is the favored microparticle material for most mammalian applications because this metal is both dense and inherently biocompatible. For most applications, particles of 1 to 3 ^m are used because these are sufficiently small relative to most mammalian cells to allow the particle to fit in the cytoplasm, but they are also sufficiently large to penetrate through 10 to 20 cell layers when applied in vivo (13,14). For the cornified epithelium of the skin, this size of particle can penetrate into the epidermis and dermis, although efficient gene delivery is usually observed only in the epithelial layers (Fig. 4). Although larger particles can be used, these typically do not bind DNA well and frequently cause more tissue damage than transfec-tion and so are generally not used.

The gold microparticles used for biolistics are usually not made for this specific application, but are made as a by-product of other manufacturing processes. As such, obtaining ''good'' particles of the right size has historically been difficult and required screening of each batch of particles to (1) confirmthey were the correct size, (2) that they were spherical and monodisperse, (3) that they could efficiently bind DNA, and (4) that they mediated transfection. Generally, if the particles satisfied the first 3 requirements, they would function well for transfection. Transfection efficiency generally correlates with the ability of the particles to bind DNA. Particles that bind 50% less DNA per particle generally mediate 50% less transfection. In practice, DNA binding after precipitation in high calcium varies substantially between different gold particles from different vendors. Given this, investigators would usually screen a small batch of a given lot and then buy as much of that lot as they could afford with typical costs on the order of $300 to $400 per gram. BioRad supplies gold particles that they have screened for activity for the gene gun. These cost considerably more per unit gold, but do save the investigator the necessity to screen and buy large amounts of particles.

B. Precipitation of DNA onto Microparticles

Nucleic acids and proteins do not inherently bind to gold particles. A number of methods have been tested to attach nucleic

Figure 4 Cross-section of mouse skin transfected with the gene gun. The panel on the left in black and white shows the tissue and locations of the particles as dark dots. The panel on the right shows epidermal expression of green fluorescent protein (GFP) after delivery of its gene on the particles. Note, expression is restricted to the epidermis, despite the fact that particles have penetrated deeper into the dermis. See the color insert for a color version of this figure.

Figure 4 Cross-section of mouse skin transfected with the gene gun. The panel on the left in black and white shows the tissue and locations of the particles as dark dots. The panel on the right shows epidermal expression of green fluorescent protein (GFP) after delivery of its gene on the particles. Note, expression is restricted to the epidermis, despite the fact that particles have penetrated deeper into the dermis. See the color insert for a color version of this figure.

acids and proteins to particles. Although one could covalently attach DNA or RNA to particles or use very strong noncova-lent methods, the most common method involves preciptation of DNA onto the particles using high concentrations of calcium in the presence of polyamines or polymers to protect the DNA (1,5,6). Simple precipitation is ideal for this approach due to the need for ''reversibility'' when delivering nucleic acids by any method. Producing a vector or a gene gun particle that is essentially like a rock might provide robust protection from physical or nuclease damage, but a nucleic acid ''rock'' is unlikely to mediate transfection because it would remain trapped after entry into the cell. For most nonvi-ral vectors, this need for reversibility is solved by relying heavily on charge interactions to assemble negatively charged nucleic acids with positively charged chemicals such as a ca-tionic lipids or peptides. For the gene gun, a gold particle has little charge or binding affinity for nucleic acids. Therefore, to satisfy the attachment need and the reversibility problem, the simplest and most robust solution is to precipitate the DNA in the presence of 1.5 M CaCl2 in the presence of spermidine or another polymer. By this approach, the nucleic acid becomes insoluble in the high calcium and precipitates onto the gold particles in suspension. The DNA-calcium precipitate is ''sticky,'' so it adsorbs to the particles efficiently. As the particles are loaded with precipitate, they tend to settle out of solution faster than unloaded particles. The gold particles are big enough that they will settle out of suspension normally, but this is accelerated when they are coated with nucleic acids. A good indication that nucleic acids have successfully attached to the particles is that they will have a tendency to stick together somewhat and will need to be triturated (pipetted repeatedly) to be brought back into a state that is easily resus-pended. A good indication that the ratio of nucleic acid to particles is too high is that the particles will form a fairly solid ''clump'' at the bottom of the tube. If observed, too much DNA was added. If not too far gone, the clumps can in some cases be broken apart by trituration, although the resulting particles will likely consist of aggregates of multiple particles that are more likely to cause cell/tissue damage. The particles are then washed in 70% ethanol, then dried in 100% ethanol. The microparticles are then transferred to a macrocarrier, such as a plastic bullet (1) or a plastic disc (7,8), or they can be coated on the inside of a tube for entrained delivery (3) (Fig. 2) in dry 100% ethanol. For macrocarrier-driven guns, transfer of the particles in dry 100% ethanol allows the ethanol to evaporate off the suspension (in a dessicator box) such that the nucleic acid slurry dries and adsorbs to the macrocarrier with sufficient adhesion to remain attached while at rest, but with low enough adhesion that they will release when impacting a stopping plate or screen (Fig. 2A). Working out the minute variations that allowed nucleic acid adhesion to particles and particle adhesion to macrocarriers required substantial effort by early investigators and were pivotal to enabling the gene gun technology.

C. The Effects of Humidity

One critical aspect to the preparation of DNA-coated micro-particles is performing DNA precipitation and storage in a low-humidity environment. For the Sanford-Johnston group, this parameter was discovered only after a few years of frustration where they observed that they could get efficient gene delivery for 6 months and then transfection efficiency would drop for the next 6 months. This cycle continued until they realized that the gun worked well in the winter, but poorly during the humid summers in North Carolina at Duke Univer sity (Stephen Johnston, personal communication). The poor transfection appeared to be due to the fact that DNA-coated microparticles loaded on macrocarriers in a humid environment did not release or ''unstick'' from the macrocarrier efficiently, whereas particles prepared in a dessicator with dry reagents released well and mediated efficient transfection. This humidity variable was unobserved prior to this because Sanford's work was performed in upstate New York, with a considerably lower humidity than North Carolina in the summer. Once only dry reagents were used (e.g., 100% ethanol stored in a dessicator), and macrocarriers and DNA-coated particles were prepared in a dessicator flooded with dry helium gas, efficient transfection was routine at any time of the year.

Particle preparation guidelines for the Helios device are similar to those described with the exception that polyvinyl pyrolidone (PVP) is added to the precipitation reaction to increase the ability of the particles to stick to the inside of the plastic tube for this entrained mode of particle acceleration (Fig. 2C). For this system, the use of too much PVP gives the same problem as humidity in which the particles will not release efficiently from the tube. Although BioRad holds that humidity does not affect their system, in Houston, Texas (which is probably even more humid than Duke), we observed variations in the release of particles from the Helios that may be due to humidity effects. Given this, we prepare and store particles for the Helios using dessicated 100% ethanol, dry the particles with dry N2 gas, and store them in a dessicator.

D. Particle Acceleration

Once particles are attached to their macrocarrier or lined on the inside of their tube for entrainment, they are ready to be accelerated. Kinetic energy can be applied to the macrocarrier or the microparticles directly from a variety of sources, including gun powder (1), and high-pressure gases, including nitrogen (12), air (16), and helium (7), and by electrical discharge to vaporize a droplet of water explosively (8). Efficient acceleration of microparticles for most guns does not occur simply by ''blowing'' the particles into the target tissue, but appears to require the production of a supersonic shockwave by the explosive release of gases, whether the energy source is gunpowder, gas, or electrical [(7) and Figs. 2A andB]. Shockwave efficiency appears to increase with decreasing molecular weight of the gas. For this reason, helium is the current propel-lant of choice for most gene guns. Although hydrogen might be more efficient than helium, the risk of combustion and explosion by this gas precludes its use.

Creating a supersonic shockwave requires the explosive release of gas or vapor. Generating this much force with helium generally employs gas pressures of 1000 to 2000 psi to create sufficient explosive energy to produce shockwaves. Release of this much gas is comparable to the force generated by a high-powered rifle and, if unrestrained, this gas blast will deafen anyone the room. This amount of gas will also destroy any cells in the path of the blast. Given this, much of the early development of gene guns involved a balancing act of generating sufficient force to produce a shockwave, while en gineering a robust mechanism to protect the target from the gas blast used to produce the shockwave. One part of the solution to this problem was to generate the gas blast in a vacuum (Figs. 2A and B). Shooting in a vacuum not only improved shockwave production, but also absorbed some of the gas blast after gene gun firing. In addition, removal of ambient air prior to firing also increased the final velocity of the particles because the gas molecules in air are sufficiently large to create drag on the flying particles. Accelerating the particles in a vacuum also necessitates that the target cells be under vacuum. Fortunately, mammalian cells can tolerate transient vacuum of 25 in. of mercury for long enough to allow transfection by gene guns like the PDS1000. Unfortunately, the same cannot be said for intact animals, who would likely explode if exposed to high vacuum. To protect animals from the required vacuum, the initial solution was to design an adaptor that could be placed in the vacuum chamber of the PDS1000 that would allow exposure of only a small area of skin or organ to the vacuum of the gun (14). The subsequent solution for the vacuum problem was the design of ''handheld'' gene guns (Fig. 3), in which the vacuum was internal to the gun and only a small area of tissue was exposed to vacuum via a small exit port (Fig. 2B). Intact tissues like the skin and liver are resilient to the application of vacuum levels of 30 to 20 in. of Hg, respectively, through the exit port of the gene gun (7). Even relatively fragile structures like the eye will tolerate vacuum levels of 5 in. Hg (17,18).

The second key feature that enabled the shockwave-generating gene guns was the use of kapton or mylar plastic disks as macrocarriers (7,8). These incredibly tough plastics are sufficiently durable to be struck by a supersonic shockwave and also hold up to the gas blast that creates the shockwave. The tough nature of these macrocarriers allowed the disks to perform several critical roles in the gene gun. Their primary function was to carry the microparticles as a ''flying disk'' macrocarrier that could be accelerated by the shockwave, fly to and impact with the stopping screen (Fig. 2A). When the macrocarrier impacts the stopping screen, the disk is stopped but the loosely held microparticles on the front of the disk do not stop and fly through the screen to impact the target cells below (Fig. 2A). The shockwave that is the driving force for particle acceleration travels at a speed greater than the speed of sound. The helium blast that produced the shockwave is chasing after the wave at the speed of sound. If this blast reaches the target cells or tissue, the target will be destroyed by the gas impact. To protect the target cells from this blast, the macrocarrier plays its final role upon impact with the stopping screen, where it seals the upper chamber with the helium away from the lower chamber containing the target cells. The helium blast chasing after the shockwave therefore bounces off the macrocarrier on the stopping screen and the excess pressure is absorbed by the vacuum pump (Fig. 2A). Thus, the flying disc macrocarrier not only carries the particles, but also shields the target from the gas blast that is the driving force of the gun.

Particle acceleration by entrainment differs substantially from shockwave-mediated gene guns. For entrainment, the microparticles are literally ''blown'' out of a tube by high-pressure helium without any supersonic Shockwave production in a manner analogous to blowing powder out of a straw (Fig. 2C). This entrainment approach originally described by Sanford did not become practical until gas valves could be used in the gene guns that could open and close rapidly enough to create a sharp gas accelerant burst. The entrainment approach to particle acceleration was first demonstrated by the group from Agracetus and later applied commercially in the Helios gene gun. Because a supersonic shockwave is not required for entrainment, much lower helium pressures (i.e., 200-400 psi) can be used for these types of gene guns. This approach also allows particle acceleration to occur in the absence of a vacuum. Although the gas pressure is lower, it is still sufficient to shred target tissues if not deflected. To deflect the helium, these guns typically use deflectors or a ''horn'' that dissipates the gas to the sides of the shot site using Ber-nouli's law (Fig. 2C). These guns are therefore easier to use than shockwave gene guns because entrained guns do not require vacuum or very high pressures of gas. They are, however, more prone to aerosolizing DNA or target materials into the environment because the helium surge is blown to the sides of the gun. One should be aware of the potential for aerosols with this type of gun, particularly when shooting potentially biohazardous materials on the particles or when addressing a biohazardous target (e.g., an infected cell culture, animal, or person). It is advisable to perform the shooting with this type of gun (or any gun) in a biosafety cabinet designed for personnel protection to avoid distribution of random agents into the air. The shockwave gene guns largely avoid aerosol-ization problems by applying shots under vacuum, but are more cumbersome to use than the Helios gene gun. Regardless of which gene gun is used, one should always consider the production of aerosols, how potential contaminants are released from the gun in the laboratory, and effective methods to decontaminate the gun after use.

Figure 5 Gene gun-mediated transfection of terminally differentiated skeletal myotubes. Cells were cultured and differentiated and transfected with 1 to 3 ^m gold particles coated with nuclear-localized p-galactosidase gene. One day later, the cells were fixed and stained with X-gal. The large dark circles are the p-galactosi-dase-positive nuclei of the multinucleated myotubes.

Figure 5 Gene gun-mediated transfection of terminally differentiated skeletal myotubes. Cells were cultured and differentiated and transfected with 1 to 3 ^m gold particles coated with nuclear-localized p-galactosidase gene. One day later, the cells were fixed and stained with X-gal. The large dark circles are the p-galactosi-dase-positive nuclei of the multinucleated myotubes.

receptor, enter an endosome, and escape into the cytoplasm (Fig. 1).

When tissues or cells in culture are examined to determine the location of the particles in the cells, they are in most cases found in the cytoplasm of the transfected cell, not in the nucleus (Fig. 5 and data not shown). Experiments using fluores-cently labeled DNA have demonstrated that immediately after impact in the cytoplasm, DNA on the particles comes back

E. Particle-Cell Interactions

Gene guns use kinetic energy to deliver nucleic acids, proteins, and chemicals directly into cells. As such, gene guns have the capacity to modify cells that are generally refractory to vectored gene delivery. For example, differentiated skeletal myotubes have historically been difficult to transfect with nonviral or viral vectors (19) due in part to lower receptor-mediated uptake by myotubes, as well as mismatch between the receptors of the cell and the ligands of the vectors (20). Although we could not transfect differentiated myotubes with liposomes or other vectors, these cells could be transfected by the gene gun (Fig. 5). Similarly, more complex targets such as whole mouse embryos can be transfected by the gene gun ex vivo with little damage to the organism (Fig. 6). Therefore, refractory cells can be addressed by the gun using kinetic energy to deliver the DNA directly into the cytoplasm of the cell, thereby avoiding the need to cooperatively bind a cell

1000 psi

2800 psi

0 0

Post a comment