|Home | About | Journals | Submit | Contact Us | Français|
Ocular gene therapy is a fast growing area of research. The eye is an ideal organ for gene therapy since it is immune privileged, easily accessible, and direct viral delivery results primarily in local infection. Because the eye is not a vital organ, mutations in eye specific genes tend to be more common. To date, over 40 eye specific genes have been identified which harbor mutations that lead to blindness. Gene therapy with recombinant Adeno Associated Virus (rAAV) holds the promise to treat patients with such mutations. However, proof-of-concept and safety evaluation for gene therapy remains to be established for most of these diseases. This unit describes the in vivo delivery of genes to the mouse eye by rAAV-mediated gene transfer and plasmid DNA electroporation. Advantages and limitations of these methods are discussed, and detailed protocols for gene delivery, required materials, as well as subsequent tissue processing methods are described.
Gene therapy in the eye thus holds promise to treat many genetic and age-related blinding diseases. The eye is considered a prime target for gene therapy since it is a relatively isolated and immune privileged organ. The field of ocular gene therapy received its biggest boost in 2001. Dogs, whose eyes are comparable to humans in size, were successfully treated for Leber’s congenital amaurosis-2 with a rAAV vector (Acland et al., 2001). The disease is caused by a mutation in the retinal-pigmented epithelium (RPE) protein 65. One of the dogs, named Lancelot, became a media star when he visited Congress.
Delivery of rAAVs to the eye is accomplished either by intravitreal or by subretinal injection. Intravitreal injections infect preferentially cells closer to the ganglion cell layer, such as ganglion cells and inner nuclear layer (INL) cells, while subretinal injections tend to infect photoreceptors (PR) and RPE cells. INL cells can also be targeted by subretinal injections depending on the titer and serotype. The protocol describes both, the subretinal and intravitreal injection methods for delivery of rAAV to the mouse eye. Injection methods into both early postnatal (Basic Protocol 1, Alternate Protocol 1) and adult (Basic Protocol 3) mouse eyes are described. In addition, delivery of plasmid DNA by electroporation at postnatal day 0 is also described (Matsuda and Cepko, 2004; Basic Protocol 2). This method targets mainly dividing cells, therefore only cells born from the time of electroporation onwards will be transduced. This includes rod photoreceptor, bipolar cells, Muller Glia cells, amacrine cells and at very low frequency horizontal cells. Electroporation of plasmid DNA circumvents viral production and can thus be used as a fast method to test the promoter activity of the viral construct or to test if overexpression of a protein may have a beneficial effect in a retinal degenerative disease model. However, it is not a viable therapeutic approach to treat humans, nor is it useful to test cell tropism of rAAVs. The protocols presented here discuss advantages and disadvantages of these different methods and describe injection tools that accommodate different budgets (Alternate Protocol 1). Additionally, tissue preparation (Support Protocols 1 and 2) and processing for immunofluorescence (Basic Protocols 5 and 7) and in situ hybridization analyses (Basic Protocols 6 and 8) on either whole mount or cryo- and paraffin sections are described. Basic Protocol 4 describes the use of a fundus scope to monitor the transduced retinal cells.
This protocol describes the delivery of rAAVs to the subretinal or intravitreal space of newborn mice. The advantage of subretinal delivery of virus or plasmid DNA at birth is that PR outer segments are not formed yet. This means that the subretinal space, between the RPE and the ONL, is an actual space in which fluid can be injected and can spread. The biggest advantage of this method is thus that the entire retinal surface area can be infected. However, at the same time this poses the following disadvantages. First, the eye is smaller in size and targeting the subretinal space correctly may be more challenging. Second, since the retina is still developing, injections that result in too much damage may complicate the interpretation of your results if development is interrupted. For example, injections with phosphate buffered saline (PBS) in a retinal degeneration model may result in a protective effect (delay of photoreceptor cell death), from the physical damage to the tissue. Thus if you study photoreceptor degeneration, it is important to perform enough control injections to account for technical variations of the procedure. An extreme case of neuroprotection occurs when too much fluid is injected into the subretinal space. In such a case, the adult retina can take on the shape of a cone instead of a half sphere. This tends to lead to a more profound protective effect. Both procedures, intravitreal and subretinal, can result in cataracts, and in the worst case in an arrest of eye development.
Intravitreal injections tend to be easier since the targeting area is larger. Injections can be performed with glass needles or metal needles (Hamilton) and in both cases the injection route can be either at the intersection of the cornea and sclera or through the sclera. Injections with a glass needle through the sclera targets directly the subretinal space. If you prefer that route for vitreal delivery, you need to push the needle through the retina. Injecting at the junction of the cornea and sclera targets the vitreous. However, the same route can also be used for subretinal injections. Basic Protocol 1 will detail injections with glass needles either through the sclera for subretinal injections or through the cornea-scleral margin for subretinal and vitreal injections. This is the most effective route and the one we recommend for delivery. The proper use of fine glass needles results in normal retinal morphology, since it is the least invasive procedure. The alternative protocol will present different tools such as Hamilton syringes and different routes of injection. As you develop your skills, you may prefer one method over the other.
Narrative summary of neonatal injection procedure: Inject neonatal pups subcutaneously with Buprenorphine (0.1 mg/kg) one hour prior to the procedure. After one hour anesthetize pups by hypothermia by placing the pup onto a dry rubber glove over ice. After 2–3 minutes place the pup onto a clean paper towel under the dissecting scope. Clean the skin over the eyelid with Betadine, followed by water and 70% ethanol using cotton swabs. Cut the skin over the eye with a sterile 30-gauge needle in the area where the future eyelid develops. If performed properly the incision will not result in bleeding as this region is undergoing cell death. Push back the skin gently to the side with a pair of sterile forceps to expose the eyeball. Inject by inserting a beveled glass needle directly into the eyeball. Close the eyelid gently with a cotton swab soaked with Betadine and place the pup onto a warm heating mat until fully recovered and then return it to the mother.
Buprenorphine is an analgesic that alleviates pain during and after the procedure.
The titer should be at least 5×1011 genome copies/ml, however we recommend injecting with a titer of 1012 – 5×1013 genome copies/ml, depending on how many cells you intend to transduce.
If you cut to deep you may damage the cornea and if you cut too close to the edge of scar the tissue may bleed.
At this point there are two options depending on the injection route and injection needle. You may pop out the eyeball to better expose the sclera for injections into the subretinal space from the sclera (recommended). You may also choose to use the same procedure to inject into the vitreous by pushing your needle through the retina. To use these procedures continue with step 13 (Movie 1).
For direct injections into the vitreous without damaging the retina you don’t need to pop out the eyeball, you can just push the skin to the side with your thumb and index finger. You can also use this procedure to inject into the subretinal space. To perform the procedure without popping out the eyeball continue with step 19.
IMPORTANT: Follow visually the (blue) solution when injecting. If the center of the eye turns immediately blue, then your needle entered too far and the needle tip crossed the retina, meaning you are injecting into the vitreous (Fig. 1a). If you see the sclera bulging in a blue color then your needle tip did not cross the entire sclera and you are not in the subretinal space yet. If the solution is spreading slowly across the eyeball as seen through the lens then you are in the subretinal space (Fig. 1b). If you inject too much volume, once you pull out the needle, the pressure in the eye may push back some of the injected material. This may be the case if you attempt to transduce the entire retina. TRICK: Before injecting the virus poke a hole into the sclera with the glass needle. Afterwards, reposition the needle and perform the injection. The first hole will allow some of the pressure to escape while you are injecting.Basic Protocol 1. Schematic of injection routes for newborn mouse pups using glass needles. (A–D) Cartoons of mouse eyes showing sclera, choroid, RPE, cornea, lens, retina, subretinal space, and injected solution (blue). (A, B) Example of injection ...
This protocol describes the transduction of retinal cells by electroporation of plasmid DNA. Since electroporation of plasmid DNA into mitotic cells is more efficient than into postmitotic cells the DNA needs to be delivered in proximity to dividing cells before retinal development is completed. Dividing cells in the retina are located close to the subretinal space thus the same procedure as described in Basic Protocol 1 can be used. At postnatal day 0 mitotic cells give rise to rods, Muller Glia, bipolars, and amacrines. Electroporation at postnatal day 0 will thus result in transduction of only these cell types once retinal development is completed. Because most of the postnatal cells are born within a short window of time after birth, electroporation of plasmid DNA needs to be performed ideally within the first 24h after birth. Electroporations at postnatal day 3 will barely yield any transduced cells. This contrasts transduction of cells by rAAV infection, which can be performed at any time.
Same material as described in Basic Protocol 1 with the following exceptions:
ATTENTION: In contrast to viral injection it is recommended that for electroporation of plasmid DNA only one eye be injected. Injecting and electroporating the second eye generates an electrical filed in the opposite direction to the first eye and may reduce the efficiency of electroporation of the first eye.
This protocol described the delivery of rAAVs into the eye of adult mice. The delivery procedure is similar to the one described in Basic Protocol 1. Since the eye is already exposed, popping out the eyeball is not required. Subretinal injections into adult wild-type mice will always lead to retinal detachment, since the PR outer segment and RPE interactions that exist in adult mice are disrupted by the fluid that is injected. This contrasts injections into newborn mice, as the retina is not attached to the RPE at that age because outer segments are not developed yet. Injecting mice with PR degeneration reduces the amount of retinal detachment caused by the injection, as part of the retina may already be detached due to the disease. However, this increases the efficiency of the viral spread. Targeting the vitreous of adult mice is straightforward. Injecting adult mice has the advantage that development is completed, which reduces some of the undesired procedural effects. Nonetheless, any injection, even PBS, can result in the release of various neuroprotective growth factors. Therefore, when injecting a virus that should result in a neuroprotective effect, enough control injections need to be performed to account for artifacts.
This protocol describes an alternate injection tool and route of delivery for injection into newborn mice. The procedure is very similar to the one described in Basic Protocol 1 and can also be used for electroporation of plasmid DNA or for vitreal injections into adult mice. A less costly way to perform the experiment as described in Basic Protocol 1 is to use a handheld pipetteman that allows mounting a glass needle. Alternatively, a Hamilton syringe, which is also less costly, can be used instead of a glass needle with an injection pump. Here we discuss the use of a Hamilton syringe in combination with a blunt metal needle. Pointy (beveled) metal needles are also available for Hamilton syringes. These needles can be used to perform the injection as described in Basic Protocol 1. The advantage of metal needles is that they do not break easily. However, given the larger size of the needle there is more damage to the tissue. The procedure below explains the use of the Hamilton syringe with a blunt metal needle. This requires a slightly modified procedure to the one explained in Basic Protocol 1.
This protocol describes the use of a fundus scope to visualize retinal cells that have been transduced with a green fluorescent protein (GFP) (Fig. 4a). The technique is non-invasive and allows acquiring retinal photographs of mice that are anesthetized with a Ketamine/Xylazine mixture. Visualizing the area of infection or electroporation is only possible if the expression cassette of your transgene also co-expresses GFP. The advantage of this protocol is that it allows selecting well infected or electroporated animals for further analysis. For example, 30 mice are injected with an rAAV that should delay PR death. Behavioral tests and/or electroretinograms need to be performed to test if the viral transgene leads to improved vision. Preselecting the 10 best-infected mice reduces the overall workload. Additionally, fewer mice need be kept for extend periods of time if long-term effects of the viral transgene are to be tested. However, this procedure is not required to perform histological analyses of GFP transduced retinas. We do not recommend purchasing such equipment to perform the gene delivery protocols described here. Its use is recommended if your institute owns such equipment and if the experimental design benefits from such use.
Since anesthesia decreases the body temperature of the mouse, which can cause the lens to become temporarily opaque, dilating the pupil in advance allows starting immediately after the mouse is anesthetized.
This protocol describes the processing of the retina for whole mount analysis. There are two ways the retina can be dissected and prepared depending on the cell type that needs to be visualized. Both methods will be introduced at the beginning. Retinal cells can then be visualized either by whole mount immunofluorescence or immunocytochemistry. These methods will be discussed in this protocol and rely on the use of an antibody that is either cell type specific or directed against the protein that is over-expressed as a result of the transduction of retinal cells. If no antibody is available, whole mount in situ hybridizations can be used to detect either the viral RNA or the mRNA of a cell type specific gene. This procedure will be described in Basic Protocol 6.
Dissect eye in PBS by removing the retina from the rest of the tissue. If target cells are photoreceptors, leaving the retina attached to the lens allows fixing the retina such that it retains its shape as a cup. The lens can be left on during the entire procedure and removed prior to mounting the retina. This avoids curling of the retina and results in better accessibility of the antibody at the periphery of the tissue. To target inner nuclear layer cells and ganglion cells the lens has to be removed. If removed after fixation, most of the curling is prevented. Alternatively, if INL cells are targeted the retina can be left onto the lens and the incubation with the primary and secondary antibody can be performed over a period of 2–3 days each at 4°C. A longer incubation time allows for better penetration of the antibody to the cell in the center of the tissue. However, the length of time needs to be established for each individual antibody. If not otherwise indicated, all steps are performed at room temperature.
This protocol is very similar to Basic Protocol 5 however, gene expression is revealed by in situ hybridization. We recommend removing the lens the latest after the hybridization step even if PRs are targeted. This avoids background from probe that is trapped between the lens and the retina. Similar to the antibody staining protocol, the incubation steps work best if the tissue is in motion during incubation. Make sure that all your solutions up to the hybridization step are free of RNAse. Treat PBS over night with DEPC (Diethylpyrocarbonate, dilution 1:1000) and autoclave next day. Make all your solutions for Day 1 in DEPC treated PBS (except 4% PFA/PBS). If not otherwise noted all steps are performed at room temperature.
Preparation of linearized DNA or PCR product is not described here, nor are the reagents and the equipment needed listed. Purify linearized DNA either by precipitation or by gel electrophoresis. Purify PCR product by gel electrophoresis. For probe synthesis we recommend to use PCR products that have been amplified from a standard cloning vector with a combination of the following primers: T7, T3, SP6.
Follow Basic Protocol 5 to dissect the retina (steps 1–5). Fix over night at 4°C or at RT for 3h.
This protocol describes the processing of the retina for cryo-sectioning. There are two ways the retina can be dissected and prepared depending on the need to retain the RPE attached to the retina. If the RPE is not needed, we recommend the dissection method described in Basic Protocol 5, which leaves initially the lens attached. However after fixation, the lens needs to be removed prior to performing the sucrose gradient and the embedding. Leaving the lens attached ensures a nice cup shaped retina. If you need to retain the RPE attached, follow the dissection method described here. Detection of gene expression by immunofluorescence and in situ hybridization on sections will be presented in Basic Protocol 7 and 8 respectively.
Dehydration of the block causes the OCT to take on a rubber like consistency over time.
ATTENTION: If the gene of interest will be detected by in situ hybridizations use for all steps DEPC treated PBS. If after rehydration of the slides the tissue has holes reduce the incubation time of the sucrose gradient. Sucrose leads to swelling and bursting of cells. Adjusting the window of time of fixation and the sucrose gradient will remedy this problem.
This protocol describes the processing of the retina for paraffin sectioning. Perform dissections as described in Support Protocol 1 according to your needs to retain the RPE attached to the retina. The protocol starts after the initial fixation step of Support Protocol 1 but prior to the over night fixation for cryo-sections.
|a. 25% Ethanol/PBS||10′|
|b. 50% Ethanol/PBS||10′|
|c. 75% Ethanol/H2O||10′|
|d. 100% Ethanol||10′|
|e. 100% Ethanol||10′|
If needed tissue can be stored at −20°C for several months in 100% Ethanol.
These washes are important to remove all residual Xylene from the tissue. If Xylene is not removed properly, when stretching the tissue in a warm water bath the lower melting point of the Xylene will leave holes in your section.
|a. 100% Ethanol||2 × 5′|
|b. 75% Ethanol/H2O||5′|
|c. 50% Ethanol/PBS||5′|
|d. 25% Ethanol/PBS||5′|
|e. PBS||2 × 5′|
For antibody stainings follow Basic Protocol 7, for in situ hybridizations follow Basic Protocol 8. ATTENTION: If the gene of interest will be detected by in situ hybridizations use for all steps DEPC treated PBS.
This protocol describes immunofluorescence analysis on retinal cross-section. The protocol works equally well for cryo- and paraffin sections.
Alternatively, DAPI can also be added with the secondary antibody.
This protocol describes in situ hybridizations on sections. It differs from Basic Protocol 6 since it is optimized for section in situ hybridizations. The same RNA probe synthesis procedure as described in Basic Protocol 6 can be used for section in situ hybridizations. The recipes for solutions that are the same between both protocols are not described here. ATTENTION: The hybridization buffer (HB) for section in situ hybridizations differs from the whole mount in situ hybridization buffer.
Use solution immediately after mixing. Do not prepare in advance!
Plastic cover slides are easier to remove and do not sheer the sections.
Reagents and solutions are described in each protocol. The following stock solutions are recommended:
Gene therapy has long been viewed as one of the tools of modern molecular medicine to treat many human disease conditions. However, less than a decade after the first clinical trials which began in 1990, the field suffered a major setback. In 1999, an 18-year-old boy died only 4 days after receiving an injection of a therapeutic Adenovirus. His death was likely caused by a severe innate immune response to the virus. The rAAVs used today elicit a minimal immune response and are thus much safer. Compared to the first generation of rAAVs, which were based on rAAV2, the ones used today achieve sustained and efficient gene expression. This new generation of rAAVs and their successful use to treat blind dogs (Acland et al., 2001) has paved the road for the future ocular gene therapy in humans.
The retina is a thin neuronal tissue at the back of the eye, which initiates the process of vision. Three distinct nuclear layers characterize it. Each nuclear layer is composed of a subset of specialized neurons (Fig. 5) (Masland, 2001, 2011). The outer nuclear layer (ONL) harbors rod and cone photoreceptors (PR), the cells that absorb photons. Rods are 1000 times more sensitive to light than cones and function primarily in dim light, while cones are used for daylight, color, and high-acuity vision. Although humans are diurnal and mice are nocturnal, in both species rods outnumber cones 20:1 (Masland, 2001), with the exception of a small region in the human eye referred to as the fovea. The fovea is composed only of cones and is the center for high acuity vision in humans. Outside the fovea the mouse and human retina are alike. Upon absorption of a photon, PRs hyperpolarize and signal to bipolar cells, which reside in the inner nuclear layer (INL). Bipolar cells connect to ganglion cells in the third nuclear layer, which send their axons through the optic nerve to the visual centers of the brain (Masland, 2001). In addition to bipolar cells, the INL is also populated by amacrine and horizontal cells (Masland, 2001), which modulate the signal, and Muller Glia cells, which are the only non-neuronal cell type in the retina and form the retinal blood barrier. Nutrition for retinal cells is provided by the retinal vasculature and the retinal-pigmented epithelium (RPE). The RPE is in intimate contact with PR outer segments. It provides nutrients and oxygen for PRs and is involved in the visual cycle (Parker and Crouch, 2010; Wang and Kefalov, 2011).
Retinal degeneration is a major cause for blindness in the industrialized world. The degeneration affects either ganglion cells or PRs. Loss of ganglion cells results in glaucoma while loss of PRs is associated with a variety of retinal degenerative diseases such as dry and wet age-related macular degeneration, diabetic retinopathy, Retinitis Pigmentosa (RP), Leber’s congenital amaurosis (LCA), etc. Retinitis Pigmentosa and LCA are inherited retinal degenerative diseases. In such cases, rAAV mediated gene therapy entails either the replacement of a nonfunctional gene or the knockdown of a dominant allele. In contrast, age-related macular degeneration, diabetic retinopathy, and glaucoma are caused by a combination of environmental and genetic factors. rAAV mediated gene therapy is still possible for these diseases however, it requires an understanding of the molecular mechanisms that lead to the disease pathology. For example, in wet age-related macular degeneration and diabetic retinopathy, neovascularization of either the choroidal or the retinal vasculature, respectively, causes leakage of fluid into the retinal proper, which then results in PR death. This neovascularization is stimulated by the vascular endothelium growth factor (VEGF) and overexpressing the soluble form of the VEGF receptor-1 (sFLT-1) reduces the incidence of new blood vessel formation (Lai et al., 2005). Retinal gene therapy can thus be applied to a wide range of genetic and non-genetic eye diseases. Recently a new treatment strategy for PR degenerative diseases has emerged, referred to as optogenetics. This strategy uses the endogenous remaining retinal circuit after PRs have died to reactivate vision. rAAVs are engineered to overexpress light sensitive ion channels such as channel rhodopsin2 in bipolar cells and/or ganglion cells. Such an approach is independent of the initial insult that resulted in PR death and can thus in principle be applied to almost all PR degenerative diseases. Finally, the most successful rAAV mediated gene therapy in the eye of humans thus far targeted the RPE cells. LCA-2 is an early onset disease caused by a mutation in the RPE protein RPE65 (Gu et al., 1997; Marlhens et al., 1997). Delivery of the RPE65 gene by rAAV to the RPE of individuals suffering from LCA-2 has restored vision in blind people (Bainbridge et al., 2008; Maguire et al., 2008), giving hope to many others suffering from vision loss. Since every retinal cell type, including the RPE, is a potential target for gene therapy, a large arsenal of rAAVs capable to infect all different cell types is needed. Cell tropism of rAAV serotypes in the eye is only known for the most commonly used serotypes (Stieger et al., 2011). Knowing the tropism is particularly important since serotypes that preferentially infect only a subset of cells can reduce unwanted side effects. In combination with cell type specific promoters or microRNA regulation (Xie et al., 2011), such rAAVs can potentially restrict transgene expression to a specific cell type.
The procedure for subretinal delivery of viruses in rodents was initially described by the Cepko laboratory (Price et al., 1987) using a replication incompetent retrovirus for lineage tracing. The same laboratory also published the DNA transduction technique of retinal cells in newborn rodents by electroporation (Matsuda and Cepko, 2004) and the embryonic transduction of retinal cells (Punzo and Cepko, 2008; Turner et al., 1990). Since then, many laboratories have used these procedures successfully and modified them according to their needs.
Viral injections yield in general, successful infections even for beginners. The simple act of inserting a needle into the eye and injecting virus will result in infection. The amount of infected target cells will increase with experience. The most important variable regarding viral injections is the viral titer and the infectivity of the viral preparation. rAAV titers are generally determined by genome copies. While this number reflects the actual amount of virus particles, the number of infectious particle can be quite different and as much as 100–1000 times lower. If little infection is seen with a high titer virus, then most likely the infectivity of the viral preparation is low. A rough assessment of infectivity can be performed in cell culture by adding 5μl of the viral preparation to one well of a 6-well culture plate with HEK239 cells. This is only possible if a) the viral cassette uses a broad expressing promoter such as CMV, b) if the cassette overexpresses an easy detectable marker such as GFP, and c) if the serotype used is capable of infecting HEK293 cells. In such a case, if 2 days post infection many cells are GFP positive, the infectivity of the viral preparation should be sufficient to successfully transduce many retinal cells. However, not all rAAV serotypes will lead to such a fast expression. This method is not meant to replace the standard quantification method. It relies on the fact that super-infection of cells with more than 100 virus particles per cell will lead to a fast expression. While a high enough viral titer is usually the most important concern regarding expression in a tissue, a too high titer can lead to unwanted effects in the retina. For example, we have observed that injections with a high titer virus 5 x1013 preparation, which shows also good infectivity, can lead to PR death. However, it remains unclear if this is solely due to a high number of virus particles per PR cell. If PR death occurs due to a high titer, diluting the virus or re-purifying it over a CsCl gradient or specialized exchange columns may help mitigate the problem.
Electroporation of plasmid DNA tends to be more complicated than viral injections. Simply inserting a needle into the eye and injecting DNA will not lead to GFP positive cells. Although the procedure is technically the same, the biggest hurdle is targeting the subretinal space properly, delivering enough DNA into that space and placing the electrode pads adequately. It may take a couple of mouse litters to master the technique. When dissecting the retina, if GFP is seen only on top of the PR outer segments but not in PRs itself, it suggests that the DNA was targeted correctly to the subretinal space but not electroporated efficiently. Macrophages that enter the retina and clean up the remaining plasmid DNA will be GFP positive due to the plasmid they took up. In such a case, either the electroporator is not delivering enough current or the electrode pads were not placed properly. The plus pole of the tweezer electrodes tends to oxidize over time, which will result in a reduction of the electric field. The electrode pads need to be kept clean and the tweezer electrode needs to be replaced every so often. The most critical parameters for this procedure are the quality and concentration of the DNA, the proper targeting of the subretinal space, the electroporator, and the tweezer electrodes. We recommend practicing subretinal injections into newborn mice with the CD1 mouse strain. CD1 mice have large litters and are albino, which helps initially to visualize the spread of the injection solution.
In situ hybridizations on retinal whole mount or sections can be quite challenging. In general, the quality of the probe and the target sequence that was chosen for hybridization are the most critical parameters. If the whole mount in situ hybridization is not working, we recommend testing the probe on sections. After resuspending the probe with H2O, we recommend adding 80μl of the section in situ HB. Diluting the probe afterwards in the whole mount HB for the actual hybridization does not affect the whole mount in situ hybridization, while the probe can still be used for section in situ hybridizations. This allows testing the probe on sections if whole mounts are not working. If the probe does not work on sections we recommend to either resynthesize the probe, or to choose a different target region of the gene. As a positive control, we recommend to start with a gene that is expressed at high levels such as one of the PR specific opsins. If there is too much background, increasing the number of washes and the wash temperature may mitigate the problem. Incubation with the antibody alone will determine if the background stems primarily from the antibody being trapped in the tissue or from unspecific trapping of the probe. A sense probe can be used as a control for the hybridization. However, many genes have coding sequence of another gene running in the antisense direction. Thus it is important to determine on the UCSC genome browser if there is a gene running in the opposite direction prior to designing and synthesizing a sense control probe. We also recommend using the UCSC genome browser to determine if there are alternative splice isoforms of your gene of interest. In such a case designing a probe against the most common region may be a good starting point.
Antibody staining tends to be straightforward when it works and difficult to troubleshoot when it does not work. The antibody and the epitope that is recognized by the antibody are the most important factors, which are unfortunately difficult to control. Here are some tips if the staining does not work. In general, more antibodies tend to work better on cryo-sections than on paraffin section. However, there are exceptions and some antibodies work better on paraffin sections. If your antibody does not work, adding SDS at a final concentration of 0.01%–0.03% in PBTB may help. Alternatively an antigen retrieval protocol can help. There are many different protocols for antigen retrieval and it is difficult to predict which ones works best for a specific antibody. If the signal is weak by immunofluorescence, then immunocytochemistry may yield a better result. In such a case we would recommend to use a secondary antibody coupled to horseradish peroxidase. If such an approach is used the tissue needs to be pretreated with H2O2 to inactivate the endogenous peroxidase in order to reduce background.
Subretinal injections at postnatal day 0 take some time to master. Viral injections tend to yield sooner positive results than electroporations, as injecting a virus into the eye will result in infection even if only few of the desired target cells are hit. Most people tend to have positive transduction with both methods after injecting 3 litters. With some practice, 50–70% of electroporated retinas will be successfully transduced. In such cases, the retinal surface area that is positively transduced ranges anywhere between 5%–50%, with 25% being the norm for most people (Fig. 6). Viral injections yield similar results however, it is easier to transduce a larger area of the retina. With some experience, it is possible to transduce even the entire retinal surface area with a single injection (Fig. 6). This may take quite some practice and even with a lot of experience only 10–20% of virally infected retinas will show infection across the entire retinal surface area.
The time point of electroporation determines the cell types that can be transduced efficiently. To broaden the range of cells types by electroporation, the procedure needs to be performed at embryonic time points. Such a procedure has been described previously (Punzo and Cepko, 2008). In contrast, viral injections with rAAVs can target all retinal cell types at any age. While the cell type that is efficiently infected is dependent on the serotype, the viral titer and the infectivity of the viral preparation can influence the perceived tropism. Additionally, the route of delivery may also influence the cell types that are preferentially infected. Figure 6 shows examples of subretinal injections with a rAAV(2/5) that carries nuclear GFP. At a low titer and low infectivity, preferentially cone PRs and some rods are infected. With increasing titer and infectivity, INL cells appear infected as well. It is thus important to determine if the tropism of a virus has been tested to its maximum potential before concluding which cells can be infected with a specific serotype. Performing both subretinal and intravitreal injections with a high titer virus that shows good infectivity should yield a comprehensive picture of the cell tropism of a specific rAAV serotype. Of note, the tropism observed in the mouse retina with a specific serotype may change when performing injections with the same viral preparation in a different organism.
Immunofluorescence or in situ hybridization analyses depend on the antibody and probe used. A compilation of immunofluorescence and in situ hybridization data is shown in Figure 7.
Viral injections or electroporations at postnatal day 0 take roughly 1½ to 2 hours per litter depending on your experience. This includes the 1 hour lag time, in which pups are injected with Buprenorphine. The individual injection of a pup takes only 2–3 minutes. Viral injections into adults take a similar amount of time. Transduction of retinal cells either by rAAV or by electroporation differs mainly in the time it takes for gene expression to occur. Gene expression upon electroporation can be detected as early as 2 days post electroporation and stays stable for several months. In contrast, gene expression from rAAV vectors may take 1–3 weeks, however expression tends to persist for years. Two variables determine the speed of expression for rAAVs. One is the viral serotype and the other is the multiplicity of infection (MOI), which depends on the amount of virus that is injected in a particular area and the actual infectious titer of the virus. In general, when more than one virus particle enters a cell, expression tends to start earlier. In our hands, the rAAV(2/5) shows robust GFP expression within 10 days post infection. Preparation of the virus is not discussed here but takes generally 1–2 weeks including, cell line expansion, purification and virus titering.
Retinal analysis by immunofluorescence or in situ hybridization may take 2 days to 2 weeks depending on the protocol and if sections or whole mounts are processed. The specific time windows for the different procedures are indicated in the individual protocols. When performing an antibody staining, we recommend to incubate with the primary antibody over night until you have determined that the antibody works well, at which point that step can be shortened to 2 hours.
Movie 1: Subretinal space injections into newborn mouse pups (Basic Protocol 1 & 2).
Movie 2: Dissections and tissue processing for whole mount and section analysis (Basic Protocols 5–6, Support Protocols 1–2).
This work was supported in part by grants from the National Institutes of Health to G.G (UL1RR031982, 2P01 HL059407, 1P01AI100263-01, and 2R01NS076991-01).