Brian Earley and Dr. Laura Bridgewater, Department of Microbiology and Molecular Biology

Bmp2 is a secretory mammalian protein that plays important roles in bone development and adult pathology. Recently, a nuclear variant (nBmp2) was discovered with an unknown function for this typically secretory protein in the nucleus (Felin et al., 2010). The purpose of this project is to find an in vitro DNA binding sequence for nBmp2. This discovery would lead to further research for how nBmp2 could affect transcription in vivo. However, since the conception of this CASTing project 16 months ago, I have faced challenges in plasmid construction, cell culture, and concessions with a fused nBmp2:GFP plasmid. While the plasmid problems require more research, the project is only a couple months from completion.

In order to find a DNA binding sequence for nBmp2, the nBmp2 protein must be synthesized in parallel with several control proteins. We elected to tag the nBmp2 protein with an epitope. Epitope-tagged proteins make experimentation more practical by allowing the use of inexpensive, high-affinity antibodies for popular epitopes.

To increase our chances of finding an epitope-protein combination that would synthesize enough protein of interest, we initially used three different epitope tags at either the carboxyl terminus or amino terminus of our protein. We needed just one of these six unique combinations to produce the necessary protein, however, the only epitope-tagged protein plasmid to express any protein had very low expression. This setback necessitated finding another way to synthesize adequate amounts of protein. First, we removed the nBmp2 gene from our common expression plasmid and ligated it into a special epitope-tagged expression plasmid to assess if this would improve expression of our protein. When this failed, we became suspicious of the sequence of the gene itself. I was able to sequence the entire nBmp2 gene in our plasmid and, with considerable confidence, annotate three nonsynonymous mutations in the gene that may affect function. However, all three mutations represented changes from the rat amino acid to the mouse amino acid sequence. Since these mutations were so closely related evolutionarily we suspected that the mutations were not the cause of our poor protein expression.

In the midst of these plasmid challenges, I ran into a flurry of cell culture difficulties as well. We chose the 10T1/2 cell line for several reasons. First, it was one of the cell lines used to initially identify nBmp2 (Felin et al., 2010) so producing nBmp2 proteins from this cell line is possible. Second, past research assistants had used the 10T1/2 cell line for CASTing experiments on different proteins, so I could reference their work in my attempts to culture this cell line and transfect it with the tagged nBmp2 plasmid and control plasmids. However, 10T1/2 cells were a challenge to culture, and unfortunately, I learned quickly that the previous research assistants who worked with 10T1/2 cells kept records that were too succinct to reproduce.

My main indicator of extraction-ready cell culture was packed cell volume (PCV). During each passage I measured the volume of the pellet of packed cells before transferring some portion of the cells to a fresh flask. Initially my PCVs for passaging cells were in the range of 10 microliters for a seemingly 100% confluent, 75 square centimeter flask. My best results came when I passaged my cells into a fresh flask after 3 and sometimes 4 days changing the nutrient media once in between passages. I also learned to condition my cells to grow more densely by transferring about a quarter of the cells from the previous flask. Together these procedures improved my PCV to as high as 40 microliters from a 75 square centimeter flask and as much as 100 microliters from a 175 square centimeter flask. I needed at least 100 microliters of PCV in order to get sufficient protein yields from extractions.

Luckily, the information from past research assistants was more prevalent on transfection efficiency, so I gathered from previous work the proper reagent amounts for different flask sizes. I was able to get transfections of as many as 20% of cells which would give adequate amounts of protein if an effective plasmid could be created.

After months of failing to find an expression plasmid that would express our tagged protein effectively, I made some concessions on the plasmid. The first discoveries of nBmp2 were made by use of a fused Bmp2:GFP protein (Felin et al., 2010). Green Fluorescent Protein (GFP) fluoresces green when exposed to UV light. GFP is a well-studied protein with validated antibodies for all the experiments performed for the CASTing project. The obvious drawback, however, is the size of GFP. Epitope tags are typically less than ten amino acids long while GFP is composed of 238 amino acids. It is possible that fusing the two large proteins together could disallow the uncharacterized DNA binding function of nBmp2.

Preliminary results of the protein extractions of the GFP fused proteins have been ambiguous. Western blot analysis showed that the GFP antibody bound to GFP presenting a band formed next to the 27 kilodalton ladder protein. However, no discrete band could be found for the fused nBmp2:GFP protein sample. Perhaps the GFP antibody binds at or near the point of fusion. But this wouldn’t coincide with the immunoprecipitation validation of the GFP antibody. Perhaps there was not enough protein sample extracted to allow a discrete band above the background. If this is the case, then we have the same problem with a fused protein as we had with our previous plasmids. As before, I sequenced the expression plasmid to check for any mutations. This sequencing missed about 100 bases which could include a deletion, insertion or frame-shift mutation that may disable the expression of the fused proteins.

At this point, the project is two months from completion if the problem was just a poor protein extraction. Fixing the expression plasmid would add perhaps another month to the project. I expect that the next western blot analysis will show sufficient protein to start CASTing experiments, and that those experiments will confirm the DNA binding sequence that is enhanced by nBmp2 before I graduate in April. A biotinylated oligonucleotide of the same sequence will be used in a radiation free EMSA to confirm DNA-protein binding completely.

References

Felin, J., Mayo, J. L., Loos, T. J., Jensen, J. D., Sperry, D. K., Gaufin, S. L., Meinhart, C. A., et al. (2010). Nuclear variants of bone morphogenetic proteins. BMC Cell Biology, 11.