Utilizing an immunocompetent, autochthonous tumor model driven by common patient mutations for preclinical testing is critical for immunotherapeutic testing. This protocol describes a method to generate brain tumor mouse models using electroporation-based delivery of plasmid DNA that represent common patient mutations, thus providing an accurate, reproducible, and consistent mouse model.
Tumor models are critical for the preclinical testing of brain tumors in terms of exploring new, more efficacious treatments. With significant interest in immunotherapy, it is even more critical to have a consistent, clinically pertinent, immunocompetent mouse model to examine the tumor and immune cell populations in the brain and their response to treatment. While most preclinical models utilize orthotopic transplantation of established tumor cell lines, the modeling system presented here allows for a “personalized” representation of patient-specific tumor mutations in a gradual, yet effective development from DNA constructs inserted into dividing neural precursor cells (NPCs) in vivo. DNA constructs feature the mosaic analysis with the dual-recombinase-mediated cassette exchange (MADR) method, allowing for single-copy, somatic mutagenesis of driver mutations. Using newborn mouse pups between birth and 3 days old, NPCs are targeted by taking advantage of these dividing cells lining the lateral ventricles. Microinjection of DNA plasmids (e.g., MADR-derived, transposons, CRISPR-directed sgRNA) into the ventricles is followed by electroporation using paddles that surround the rostral region of the head. Upon electrical stimulation, the DNA is taken up into the dividing cells, with the potential of integrating into the genome. The use of this method has successfully been demonstrated in developing both pediatric and adult brain tumors, including the most common malignant brain tumor, glioblastoma. This article discusses and demonstrates the different steps of developing a brain tumor model using this technique, including the procedure of anesthetizing young mouse pups, to microinjection of the plasmid mix, followed by electroporation. With this autochthonous, immunocompetent mouse model, researchers will have the ability to expand preclinical modeling approaches, in efforts to improve and examine efficacious cancer treatment.
Murine brain tumor models are crucial for understanding the mechanisms of brain tumor formation and treatment. Current models typically include rapidly produced subcutaneous or orthotopic transplantations of commonly used tumor cell lines, based on a limited number of driver mutations or patient-derived xenograft models, using immunodeficient mice that hinder proper immunotherapy studies1,2,3,4. Additionally, these preclinical results can lead to false positives, in that such models can exhibit dramatic, oftentimes curative effects in response to therapy, but this does not translate to the clinic2,5,6,7. Having the ability to rapidly produce genetically engineered preclinical mouse models that are more reflective of patient mutation signatures is imperative for improving the validity of preclinical results.
Electroporation (EP)-based delivery of DNA plasmids to induce both loss of function (LOF) and gain of function (GOF) mutations allows for the generation of such models. We developed a method for an even more precise representation of GOF driver mutations called mosaic analysis with dual-recombinase-mediated cassette exchange, or MADR8. This method allows for the expression of a gene (or genes) of interest in a controlled, locus-specific manner in somatic cells8. In combination with other molecular tools, such as clustered regularly interspaced short palindromic repeats (CRISPR), different patient mutations can be combined to develop mouse brain tumor models. This method has been used for different pediatric brain tumors, including gliomas and ependymomas8, as well as adult brain tumor models, such as glioblastoma (GBM).
While the EP method of tumor modeling is not as common as a transplant, the following demonstrates heretofore the ease and high reproducibility of this modeling system. mTmG mice are used for the insertion of the MADR-plasmid DNA8,9. This system allows for the recombination of loxP and Flp recombinase target (FRT) sites located at the Rosa26 locus for subsequent insertion of the donor DNA plasmid (i.e., GOF gene of interest)8,9. The following protocol demonstrates the straightforwardness of this method after diligent practice, and the ability to develop mouse brain tumor models in an autochthonous, consistent manner.
All procedures in this protocol were approved by the Cedars Sinai Medical Center Institutional Animal Care and Use Committee (IACUC). Homozygous mTmG mice were bred with C57BL/6J mice to obtain litters of mixed-sex, heterozygous mTmG mice for use in the following protocol. The animals were obtained from a commercial source (see Table of Materials). Mouse pups were electroporated between postnatal days 0 and 3 (P0-P3).
1. Surgical setup
2. Pre-surgical preparation
3. Microinjection of DNA plasmid mix into the brain ventricle
4. Electroporation
5. Post-surgical steps
The protocol described above has been used to successfully develop both pediatric and adult brain tumor mouse models, with the former published in extensive detail in Kim et al.8. With proper technique and careful planning of plasmid design, the success for EP development of tumors is typically 100%. Histology is the quickest and easiest way to check for successful DNA plasmid insertion when a reporter protein is used. This protocol involves steps on how to develop a GBM brain tumor model with 100% penetrance, as confirmed by histological analysis. An MADR donor plasmid expressed both a reporter protein (smTagBFP2-V5) and SpCas9 (Supplementary File 1). Two strong driver LOF tumor suppressor gene single-guide RNAs (sgRNAs) were also included, directed at Nf1 and Trp53 (see reference12 for sgRNA cloning strategy; see Supplementary File 1 for oligonucleotide sequences used in this protocol). Finally, the Cre and Flp recombinase plasmids were added to the plasmid mix for recombination at the Rosa26 locus (Figure 1A; see Supplementary File 1). Mice were moribund by 5 months post-EP, with full tumor growth detected through the reporter protein and histological analysis (Figure 1B).
Figure 1: Immunofluorescent staining of successful tumor growth at 5 months post-EP. (A) MADR donor plasmid for spCas9 and the reporter protein TagBFP2-V5. Also included in the plasmid mix was a Cre-Flp recombinase plasmid, along with sgRNAs directed at Nf1 and Trp53. (B) A coronal section taken at 5 months post-EP of a GBM tumor driven by Nf1 and Trp53 LOF. Tumor cells are labeled with the TagBFP-V5 reporter protein linked to spCas9 (B3). Sparse EGFP labeling is also detected (B2) along with all cells that did not express the plasmids labeled with tdTomato (B1). Scale bar = 1,000 µm. Abbreviations: Ctx = cortex; CC = corpus callosum; LV = lateral ventricle; Str = striatum. Please click here to view a larger version of this figure.
Supplementary File 1: Plasmid sequences. Complete sequences of all plasmids and oligonucleotides used in this protocol, including pDonor-SM-TagBFP2-V5-P2A-spCas9-WPRE, pCag-FlpO-2A-Cre, and the oligonucleotides for sgRNA targeting Trp53 and Nf1. Please click here to download this File.
Electroporation-based delivery of plasmid DNA allows for the in vivo use of molecular biology, similar to that used in genetically engineered mouse models, but with the speed, localization, and efficiency of viral transduction8,13,14. With the latter, however, comes safety concerns as well as immune responses. We have shown in our modeling system using EP-delivery of plasmid DNA that minimal immune response occurs due to the initial insertion of the glass capillary pipette into the brain ventricle8. Therefore, to improve upon the current tumor modeling systems for immunotherapy, the protocol presented above allows for a more expedient and robust preclinical modeling system for brain tumors.
The first crucial step for successful brain tumor modeling using this method is plasmid design. While not described in this protocol, it has been extensively discussed in Kim et al.8 and Rincon Fernandez Pacheco et al.10 for pediatric brain tumor models, as well as other sources for additional uses15,16. While our lab has successfully electroporated up to six plasmids, there will be an upper limit to how much DNA can be delivered. This upper limit is constrained by plasmid DNA concentration, plasmid sizes, voltage, and the volume of DNA delivered, which is itself limited by the size of the brain ventricle at the age of electroporation. In addition, with multiple DNA plasmids in the mix, it is imperative to have a well-suspended plasmid mix prior to injection. If the mix is not resuspended well, it will be difficult to take up in the narrow glass capillary pipette. Also, with the proper plasmid mix, the addition of fast green dye (10% v/v) allows for clear viewing of the plasmid mix taken up in the brain ventricle during administration (step 3.4). One important caveat to this, however, is that the dye can interfere with antibody staining in the far-red wavelengths (e.g., Cy5) of tissue processed within several weeks after EP8. Although the fast green dye is very helpful when first learning EP, it can be eliminated to avoid false-positive staining. It is worth keeping in mind that, for step 3.4, the delivery of the plasmid mix in the ventricle will not be visible, so it will be crucial to watch the glass pipette for the volume to decrease as it is injected.
For consistency and proper handling of injections, there are several factors to keep in mind: 1) the injection volume. Though it may take several attempts to cut the glass capillary pipette accurately (step 2.1.2), it is crucial that only 1 µL of plasmid mix is injected for each animal. Increases or decreases in the volume between animals will certainly affect the latency of tumor formation. 2) once the injection volume is set, the parameters of the microinjector must not be changed, as this will affect the volume. 3) once the plasmid mix is brought up in the glass capillary pipette, the pipette should be kept out of the way so as to not accidentally poke oneself or others. Using a micropipette stand (soldering aid; see Table of Materials) that clips onto the microinjector holder is useful for this. However, if the plasmid mix is not injected soon after, it will likely clog the tip. Therefore, it is important to do one test microinjection on a piece of parafilm immediately prior to injecting into the brain ventricle to make sure there is no clog blocking the orifice. Finally, similar to every other surgical technique, the execution of the protocol is different for each individual. It is therefore, imperative to have the same technician performing the microinjection and/or EP for all groups of mice in an experiment.
The brain tumor models developed in our lab have focused on gliomas. From a neurobiological perspective, performing this method between postnatal days 0-3 targets the gliogenesis period and the period post-embryonic neurogenesis. Interest in alternative brain tumors may require adjustment in the time point and/or position of electrodes. Several protocols are available for in utero EP that would be pertinent for neural-based tumors15,17,18.
There are a few limitations to this method. The ability to clone and mix and match both LOF and GOF DNA plasmids using CRISPR, transposons, and, more recently, the MADR system allows for endless combinations of patient mutations to be modeled. The techniques of microinjection and EP are relatively simple, though they do take a few practice attempts to improve and become consistent. One limitation of the model presented is the length of time it takes for the GBM tumor to fully develop (5 months). Despite this, we are currently working on additional models using common patient tumor mutations that lead to aggressive tumor growth lasting less than 2 months. In addition to limitations, the equipment itself is a substantial initial investment (around $20,000-$30,000); however, the ability to electroporate tens if not hundreds of mice in one sitting allows for incredible power for a preclinical experiment. With that being said, if there are issues with mouse breeding, this can become the biggest limitation, as it is imperative to have healthy breeders and healthy litters.
It has been several decades since advancements in brain tumor treatments have occurred, with the last greatest improvement of the chemotherapeutic Temozolomide extending survival by only a few months19. Immunotherapy has revolutionized many different cancers, but is yet to make a significant impact on the standard of care in neuro-oncology. Interestingly, currently used mouse brain tumor models have shown great success with immunotherapy, but continuously fail to translate to the clinic and show success with patients. It is therefore imperative to take a step back and re-evaluate the mouse tumor models used. Current models depend on using older cell lines with limited mutations not commonly found in patients while injecting thousands of cells into the brain. With the technique discussed in this protocol, different patient tumor mutation profiles in a mouse model are recapitulated, allowing for an autochthonous, gradual development of tumors within a time frame that would allow for preclinical testing in immunocompetent mice.
The authors have nothing to disclose.
We thank Gi Bum Kim for the immunofluorescent staining and images. We also thank Emily Hatanaka, Naomi Kobritz, and Paul Linesch for helpful advice on the protocol.
0.1-2.5 µL 1-channel pipette | Eppendorf | 3123000012 | |
2 µL pipette tips | Fisher Scientific | 02-707-442 | |
20 µL pipette tips | Fisher Scientific | 02-707-432 | |
2-20 µL 1-channel pipette | Eppendorf | 3123000098 | |
DNAZap PCR DNA Degradation Solutions | Fisher Scientific | AM9890 | |
ECM 830 Square Porator Electroporator | BTX | 45-0662 | |
Electrode Gel | Parker Labs | PLI152CSZ | |
Fast Green Dye | Sigma-Aldrich | F7258-25G | |
Helping Hands Soldering Aid | Pro'sKit | 900-015 | |
Micro Dissecting Scissors, 4.5" Straight Sharp | Roboz | RS-5916 | |
Mouse Strain: C57BL/6J | The Jackson Laboratory | JAX: 000664 | |
Mouse Strain: Gt(ROSA)26Sortm4(ACTB-tdTomato,-EGFP)Luo/J | The Jackson Laboratory | JAX: 007676 | |
Parafilm | Grainger | 16Y894 | |
Plasmid: pCag-FlpO-2A-Cre EV | Addgene | 129419 | |
Platinum Tweezertrode, 7 mm Diameter | BTX | 45-0488 | |
Sharps container, 1-quart | Uline | S-15307 | |
Standard Glass Capillaries, 4 in, 1 mm OD, 0.58 mm ID | World Precision Instruments | 1B100F-4 | Capillary pipettes need to be pulled – see reference 10 for details. |
Vertical Micropipette Puller | Sutter Instruments | P-30 | Heat settings: Heat #1 at 880, Heat #2 at 680; pull at 800. See reference 10 for more details on pulling. |
Vimoba Tablet Solution | Quip Laboratories | VIMTAB | |
XenoWorks Digital Microinjector | Sutter Instruments | BRE | |
XenoWorks Micropipette Holder | Sutter Instruments | BR-MH |