Reliable and inexpensive expression of large, tagged, exogenous proteins in murine bone marrow-derived macrophages using a second generation lentiviral system

Over the past two decades, researchers have struggled to efficiently express foreign DNA in primary macrophages, impeding research progress. The applications of lipofection, electroporation, microinjection, and viral-mediated transfer typically result in disruptions in macrophage differentiation and function, low expression levels of exogenous proteins, limited efficiency and high cell mortality. In this report, after extensive optimization, we present a method of expressing large tagged proteins at high efficiency, consistency, and low cost using lentiviral infection. This method utilizes laboratory-propagated second generation plasmids to produce efficient virus that can be stored for later use. The expression of proteins up to 150 kDa in size is achieved in 30-70% of cells while maintaining normal macrophage differentiation and morphology as determined by fluorescence microscopy and Western blot analysis. This manuscript delineates the reagents and methods used to produce lentivirus to express exogenous DNA in murine bone marrow-derived macrophages sufficient for single cell microscopy as well as functional assays requiring large numbers of murine bone marrow-derived macrophages.


INTRODUCTION
Macrophages are white blood cells (leukocytes), which are essential components of the immune system. Macrophages play key roles in innate and adaptive immune responses as immediate responders to eliminate microbial infections, promote wound healing, and present antigen to lymphocytes. Macrophages are derived from circulating monocytes which, upon activation, extravasate through the vascular wall and differentiate into macrophages. After terminal differentiation macrophages can migrate through complex extracellular matrices to sites of injury or infection to elicit an immune response.
As essential components of both innate and adaptive immunity, efforts to fully understand macrophage functions involve in vitro studies of primary macrophage populations. Murine macrophages are generally purified for in vitro study utilizing two different methods. One approach entails harvesting bone marrow and differentiating multipotent stem cells into murine macrophages (BMDMs) in culture using cytokines such as murine monocyte colony-stimulating factor (MCSF) or granulocyte-macrophage colony-stimulating factor (GMCSF) [1]. Murine macrophages can also be isolated by inducing peritonitis and harvesting murine macrophages that infiltrate the peritoneal cavity, which can be directly assayed or cultured [1].
To study the structures and proteins that mediate macrophage function it is advantageous for researchers to have the ability to express constructs in these cells, including shRNA and exogenous DNA. Macrophages and macrophage cell lines are notoriously difficult to manipulate genetically. With many studies utilizing knock-out (KO) mouse generated macrophages, the ability to trace molecular mechanisms relies on reconstitution with wild type (WT), mutant, and truncated versions of the deleted protein. Most techniques utilized cause detrimental effects to macrophage differentiation and viability. One likely reason for the difficulty of expressing exogenous DNA within macrophages is that these cells undergo terminal differentiation [2]. Terminal differentiation of macrophages from precursor cells causes macrophages to adopt a quiescent state, no longer undergoing proliferation [2]. Reliable and stable expression of exogenous DNA requires gene integration into the genome; however this process is very difficult if not impossible in cells that no longer proliferate. Therefore, the most common available methods of introducing exogenous DNA in primary differentiated macrophages permit only non-integrated transient expression. Early reported studies and methods used to attempt DNA delivery and transient expression in primary murine macrophages include liposomal transfection, ionic precipitation, electroporation, and chemical transfection. However, the dawn of gene delivery viruses has yielded numerous advances in research and has begun to be utilized for macrophage research.
Multiple reports have claimed to express exogenous DNA in primary murine macrophages and murine macrophage cell lines. An early report detailed an approach to express exogenous DNA in primary murine macrophages utilizing and optimizing various methods [3]. Rupprecht et al. performed calcium-phosphate, diethylaminoethyl(DEAE)-dextran, and lipofection, using an enzyme assay to determines expression profiles, and concluded that DEAE-dextran was optimal [3]. However, another study performed a few years later by Thompson et al. incorporated additional reagents and techniques to optimize expression in a murine macrophage cell line to measure expression using a luciferase assay. The results of this study correlated with reports of Ruprecht et al. concluding that electroporation yielded optimal DNA expression, while DEAE-dextran was the least efficient technique [4]. These studies were in agreement that lipofection was very ineffective, however neither study measured actual expression efficiency levels, and Thompson et al. reported over 75% cell death with their optimal DNA expression technique [3,4].
Dokka et al. attempted to optimize lipofection techniques utilizing protamine sulfate to enhance lipid-mediated transfection of murine macrophage cell lines and additionally compared these with other transfection techniques that have been reported to be successful by other investigators [5]. Protamine sulfate is suggested to condense DNA for easier nuclear permeability as well as containing a nuclear localization sequence, which promotes trafficking of transfected DNA into the nucleus [6]. Dokka et al. also utilized the luciferase assay to assess transfection efficiencies and reported that lipofection combined with protamine sulfate yielded high expression levels in comparison to other transfection techniques, with little effect on cell viability [5]. Thus, there is extensive controversy found regarding these reported techniques utilized for effective transfection and expression of exogenous DNA in primary murine macrophages and murine macrophage cell lines.
Additional techniques have been described to aid in the expression of exogenous DNA in a murine macrophage cell line by blocking the silencing of foreign DNA caused by methylation after DNA has been delivered inside macrophages [7]. 5-azacytidine is a chemical that blocks the methylation of DNA by inhibiting DNA methyltransferases [7]. In one study, DEAE-dextran combined with 5-azacytidine was found to be more efficient than several liposomal reagents, concluding that DEAE-dextran was optimal for gene delivery, yet DEAE-dextran alone was reported to have poor efficiency [7]. To our knowledge, this is the only documented report utilizing 5-azacytidine to express exogenous DNA in macrophages. In our own studies 5-azacytidine treatment on murine macrophage cell lines or primary macrophages did not increase foreign DNA expression efficiencies (data not shown).
Electroporation is another method that has been reported to effectively deliver DNA into murine macrophages. This method uses electrical current to deliver charged particles, such as DNA through the plasma membrane of cells to reach the cytoplasm for expression. A variation to electroporation has been more recently developed, termed "nucleofection", that is purported to be better suited for DNA transfer into macrophages, offering more system adjustments such as voltage and cell-specific reagents for electroporation permitting additional optimization. Using nucleofection, Van De Parre et al. compared transfection efficiencies of DNA, dsRNA, and mRNA of green fluorescent protein (GFP) using a murine macrophage cell line. Interestingly, they showed that DNA and dsRNA caused extensive cell death through causing apoptosis and no expression of GFP was observed [8]. However, mRNA was observed to have GFP expression efficiencies up to 75% with little effect on cell viability, as evaluated by flow cytometry and Western blot analysis [8]. Another group showed success in using the nucleofection technique to transfect shRNA and DNA constructs into a murine macrophage cell line in two different studies [9,10]. They transfected a macrophage cell line with GFP-tagged shRNA and using fluorescence microscopy to detect GFP expression generated a stably transfected macrophage cell line [9,10]. Knockdown of protein levels were demonstrated by Western blot analysis, yet efficiency levels of exogenous DNA expression were not specified [9,10]. These studies have suggested that the more advanced electroporation system has a substantial impact on increasing exogenous DNA expression in murine macrophage cell lines. The use of nucleofection on primary cells seems unfeasible as primary murine BMDMs have a short lifespan, and are unlikely to survive a drug selection process to create a stable cell line.
The use of virus-based gene delivery into cells has gained much momentum over the last decade. Reports of increased efficacy in DNA delivery and expression in both murine macrophage cell lines and primary murine macrophages, utilizing viral systems, have begun to phase out other common techniques used for DNA transfection. The two most frequently used viral systems used for expressing exogenous DNA in murine macrophages are adenovirus and lentivirus; however, adenovirus can only be used for dividing cells, and thus is not suitable for primary bone marrow-derived macrophages. Moreover, variations in techniques and/or reagents have been utilized in efforts to increase lentiviral transduction efficiency, including polybrene and/or spinfection, also termed spinoculation [11,12]. Spinfection is thought to help sediment viral particles to increase virus contact with cell membranes, while polybrene reduces charge repulsion between viral particles and the cell membrane [11,12]. A study by Cunnick et al. demonstrated lentivirus transduction efficiency between 30% and 50% in murine bone marrow-derived macrophages determined using fluorescence microscopy [13]. Interestingly, this group reported that polybrene did not increase lentiviral transduction efficiencies [13]. Pan et al. demonstrated the use of a lentiviral system with a dual promoter, allowing for targeted and inducible expression in specific cells of macrophage lineage [14]. This study reported high expression efficiency in murine macrophage cell lines and bone marrow-derived macrophages, however multiple rounds of cell sorting was required to achieve this, which can ultimately have a detrimental impact on cell viability [14].
Prior to developing and optimizing the method presented here, we utilized a complete virus producing kit from Clontech (Cat # 631247). The expression vector ultimately utilized in our method is derived from the Clontech system. The complete Clontech kit exploited a forth generation packaging system, which allotted for a higher degree of safety, due to 4 recombination events being required for release of replicating virus. In our hands, this system did not produce infectious virus that was capable of expressing quantifiable exogenous DNA in primary macrophages. Additionally the use of four packaging vectors was probably less efficient in transferring viral DNA in to the 293T cells for viral production. These packaging vectors are purchased premixed, requiring repeated purchasing instead of having the opportunity to propagate the vectors in lab, increasing lab expenses. Fortunately the expression vector containing GFP was compatible with a different packaging system that was commercially available and can be propagated in the lab. We also initially attempted to concentrate virus utilizing a concentration solution for lentivirus from Clontech (Cat # 631231), however this proved to be cytoxic to primary macrophages. We report here an efficient, consistent, inexpensive protocol of exogenous DNA expression in primary murine bone marrow-derived macrophages, using lentivirus transduction. This second generation system would require two recombination events during viral production for release of replicating virus and is considered by the CDC to be a BSL-2 level activity, suitable for most tissue culture containment hoods. While DMEM with 10% FBS, 15% L929 cell supernatant, 2 mM Glutamax, 10 units/ml Penicillin, 10 μg/ml Streptomycin 9 70% Ethanol Solution: 70 ml of 100% EtOH with 30 ml sterile water 9 Fixing Solution: PBS with 3.7% formaldehyde final concentration 9 Permeabilization Solution: PBS with 0.002% NP-40 9 Blocking Solution: PBS with 3% heat-inactivated goat serum and 1% TritonX-100 9 Mounting Solution: (Gelvatol) 24 ml of distilled H 2 O with 9.6 g PVA, 19.05 ml glycerol, 48 ml Tris pH 8.5, 2.5% DABCO

PROCEDURE
All reagents and materials used during viral production, harvesting, concentrating, and infection should be used in a sterile tissue culture hood equipped with ultra-violet light for decontamination. Treatment with 15% bleach was used as an additional decontamination measure.
1. Transfection of 293T-X cells for virus production.
1.1. The 293T-X cells are maintained in DMEM + 10% FBS + 1% penicillin-streptomycin + 10 mM glutamax. Cell passage should be performed before cells reach full confluency using trypsin. During the first wash of cell passage, an equal volume of complete media must be added to cells in trypsin to allow for proper cell pelleting.

1.2.
To transfect plasmids for viral production, 293T-X cells are cultured in 225 cm 2 flasks. Cells should be at ~90% confluency for transfection. Two 225 cm 2 flasks of 293T-X cells should be cultured and transfected for each virus.

1.3.
On the day of 293T-X cell transfection, media should be removed from cells, and cells should be supplemented with 20 ml of fresh complete media.

1.6.
Place the flasks back in the incubator at 37°C with 5% CO 2 .

1.7.
After 6 h, replace the transfection media with 40 ml of fresh complete media and place flasks back into the incubator.

1.8.
Harvest viral media after 72 h. Pool together viral media from identical flasks and add to a 250 ml 0.45 μM acetate filter flask to filter the viral media to remove any debris or cells in suspension.

1.9.
Ultracentrifuge tubes are then filled with 13 ml of the filtered viral media and balanced using the single beam balance after being placed into the buckets of a SW40Ti rotor.
1.10. Centrifuge the viral media at 100,000 × g for 2 h at 4°C, with a SLOW brake setting.
1.11. After centrifugation, carefully remove buckets and transfer to hood. Remove the ultracentrifuge tubes from the rotor buckets and gently pour off supernatant. Add 500 μl of bone marrow-derived macrophage media to each tube. Wrap the top of each tube with parafilm and store at 4°C overnight.
1.12. After overnight incubation, place tubes on orbital shaker at 100 rpm at room temperature for 1 h.
1.13. Transfer tubes to a sterile tissue culture hood and remove parafilm. Pass media through a 5 ml serological pipet to break up remaining pellet, and pool together resuspended viral pellets from identical transfections.
2. Harvesting, culturing, and differentiation of murine bone marrow-derived macrophages for viral infection.

2.1.
In our experiments, mice were sacrificed between the ages of 5 and 9 weeks using CO 2 .
2.2. Spray mice down with 70% ethanol and pin mice down to dissection mat with the abdomen facing down.
2.3. Using forceps and surgical scissors, remove the femur and tibia and place them in a 10 cm petri dish containing wash media.

2.4.
After removal of leg bones, in a sterile tissue culture hood further remove any remaining muscle tissue from the leg bones.

POL Scientific
2.5. Using surgical hemostats to hold and cut the ends off of each femur and tibia using surgical scissors. Using wash media in 3 ml syringes equipped with a 25 G needle flush the bone marrow from each bone into a 50 ml conical tube.
2.6. Centrifuge the bone marrow for 5 min at 1000 rpm at room temperature.

2.7.
Wash the pellet with 5 ml of 10 mM EDTA in PBS.
2.8. Centrifuge the bone marrow for 5 min at 1000 rpm at room temperature.

2.9.
Resuspend the pellet with 3 ml of ACK buffer and incubate for 2 min to lyse RBCs.
2.10. After the incubation time, immediately add 10 ml of wash buffer to prevent lysis of bone marrow cells.
2.11. Centrifuge the bone marrow for 5 min at 1000 rpm at room temperature.
2.12. Resuspend the pellet with 20 ml of adhesion exclusion buffer, and transfer the cells to a 10 cm petri dish, and place dish into an incubator at 37°C with 5% CO 2 for 2 h.
2.13. Gently swirl media in the dish, and then carefully remove media with unattached cells while leaving any observable pieces of tissue.
2.14. Centrifuge the unattached bone marrow cells for 5 min at 1000 rpm at room temperature.

ANTICIPATED RESULTS
The methods detailed here offer an optimized lentiviral system for expressing exogenous DNA in extremely hard-to-transfect cells (e.g., BMDMs) with efficiencies of ~70% with GFP alone and ~30% with large GFP-tagged proteins, as determined by visual counting of fluorescence microscopy ( Fig. 1 and Fig. 2) and Western blotting (Fig.  3). For expression efficiency shown in Figure 2, cells were visualized and counted at a higher magnification (60 ×) than are shown in Figure  1. This was necessary due to the discreet localization of FMNL1γ to macrophage podosomes. Fluorescent cells due to GFP-alone were easily visible at all magnifications ( Fig. 1 and 2). The inset images in Figure 2 show the difference in fluorescence distribution between GFP and GFP-FMNL1γ. GFP is broadly distributed throughout the cytoplasm while GFP-FMNL1γ is found in punctuate structures of the ventral surface, an adhesion structure termed a podosome. Fluorescence microscopy may not be sufficient to detect extremely low levels of expression in cells, thus these reported expression efficiencies may largely underestimate the actual number of expressing cells. Therefore, flow cytometry may be a more accurate method to determine actual expression levels, but was not available for this study. Furthermore, viral production may be less efficient with large, GFP-tagged proteins, so increasing the number of flasks or media harvests from the 293T-X packaging cells producing virus for concentration can increase the amount of virus. We have also periodically performed virus titers using both an ELISA-based assay to quantify levels of virus specific protein, and traditional plaque assays. Both methods appear to underestimate the total virus transduction occurring, as measured by microscopy. Two important components that contribute to the success of this protocol are concentrating the lentivirus and using it quickly after production without any freeze/thaw cycles. were fixed, permeabilized, and stained with DAPI for nuclear visualization. Field was imaged using a 20x objective.

POL Scientific
The GFP-tagged proteins described in these experiments are splice variants of the formin, FMNL1. Formins are actin-associated proteins that perform multiple actin cytoskeleton regulating events including actin filament nucleation, polymerization, bundling, and/or severing [15,16,17]. FMNL1 is a large protein with 1069 amino acids, making it approximately 110 kDa, and with a GFP tag (26 kDa) this protein is approximately 136 kDa. 3 different isoforms of FMNL1 have been identified, and we have shown expression of two of these using our lentiviral system (Fig. 3) [18,19]. PCR was used to copy Superscript-derived cDNA encoding the FMNL1 isoforms from purified macrophage RNA. Oligonucleotides specific to the 5' end of FMNL1γ (accession FJ534522) and FMNL1β (accession BC001710) incorporated a Xba1 restriction site in frame with initiating methionine. Oligonucleotides specific for the 3' end of each FMNL1 isoform were engineered to include a BamH1 restriction site. Following PCR, the product was subjected to Xba1 and BamH1 restriction and ligated into Xba1/BamH1 restricted pLVX-AcGFP1-C1 for viral production. Therefore, this expression system is well suited for expressing large proteins in primary murine bone marrow-derived macrophages with high efficiency (Fig.  3). Moreover, this system permits expression of full-length proteins that are fluorescently labeled, making this ideal for studies in primary murine macrophages derived from murine KO models.
Moreover, there are a few additional benefits to utilizing our developed lentiviral system for expressing exogenous DNA. For example, this system can be applied towards the expression of fluorescently labeled proteins in non-macrophage cell lines, as we have used this system to reach nearly 100% efficiency in HeLa, U2OS, and MDA cell lines (data not shown). In addition, our method incorporates transforming competent E. coli with viral vectors, allowing for unlimited propagation. Thus, our method is very cost-effective as it does not require the constant purchasing of viral vector packaging mixes. Also, the lentivirus produced from this procedure can be frozen and used at a later date; however, freeze/thaw cycles will likely affect some of the viral infectivity. Additionally, it may be possible to increase BMDM expression efficiency by performing a secondary transduction 48 h after the initial transduction, however this would require producing an exceptionally large volume of virus for each independent experiment.
Another very important aspect of our expression system is that it can yield high enough expression efficiencies to perform functional assays that require large numbers of cells. This would prove to be difficult if not impossible using other systems because of their lower expression efficiencies, where functional assays often require a large portion of the cell population to express the exogenous DNA for measurement and quantification. On another note, attempts to use this protocol to express foreign DNA in primary human peripheral blood monocyte-derived macrophages yielded poor expression efficiencies. However, more extensive optimizing could potentially permit higher expression levels in primary human macrophages.
Therefore, in this report, we fully detail how to produce infectious lentivirus with the capacity to express exogenous DNA in primary murine bone marrow-derived macrophages at high efficiencies, consistently, and at low cost. This easily adaptable protocol will have a profound effect on enlightening and enhancing the field of macrophage biology with more efficient and comprehensive studies working on macrophage structures and proteins.