Engineering Success

Design

Our main goal was to build a plasmid which would incorporate EHMT2, a CMV promoter, and a puromycin resistance gene (which could be used in future studies to ensure maximum transfection rate into mammalian cells) that could be produced in E-coli DH5-α.

We first started by assessing which plasmid parts we would take, and sourced them out. We used a pre-existing primer provided to us by a professor (named GFP 3.3) which was essentially the backbone of our plasmid and contained a CMV promoter. We also sourced another plasmid similarly with a puromycin resistance gene for mammalian cells, and an EHMT2 plasmid (sourced from addgene). We planned to take the GFP 3.3 sequence which contained the neomycin resistance gene, replace it with the puromycin resistance gene, and take out the “GFP” protein and insert in “EHMT2” instead. We also designed primers using Primer Blast for each portion of the PCR.

The dark blue portion of the sequence is labeled as the CMV promoter, and the yellow portion is the EHMT2 gene, with the dark green being EHMT2 vector primer (reverse).

Similarly, the PuroR (puromycin resistance gene) is in the deep yellow color, and its forward and reverse vector primers are in neon green.

The ideal plasmid in construction with its sizes were:

The full sequence of primers were:
​​PuroR vector reverse: GCGAAACGATCCTCATCCTGT
PuroR vector forward: GCGGGACTCTGGGGTTCGCGA
PuroR insert forward: ACAGGATGAGGATCGTTTCGCATGACCGAGTACAAGCCCACG
PuroR insert reverse: TCGCGAACCCCAGAGTCCCGCTCAGGCACCGGGCTTGCGGGT
EHMT2 vector forward: GCTGCCTTCTGCGGGGCTTGC
EHMT2 vector reverse: GGTGGCTCTTATATTTCTTCT
EHMT2 insert forward: AGAAGAAATATAAGAGCCACCATGGCGGCGGCGGCGGGAGCT
EHMT2 insert reverse: GCAAGCCCCGCAGAAGGCAGCCTAACCGGTACGCGTAGAATC

Build

We then conducted a Plasmid Extraction, PCR, Gel Electrophoresis, and Gibson Assembly in order to construct our plasmid. We first tested our PCR products, and everything matched in the gel electrophoresis stage, except for the EHMT2 band.

Therefore, we tried to run another PCR with the EHMT2, and got clearer results, but still had 2 bright bands. We therefore resolved to cut out the one band that looked closer to the size we needed

Then, for Gibson Assembly, we combined the parts of the plasmid into one.

Test

For the first iteration of our gibson assembly, we ran a gel and did not see any favorable results, so we regrouped and tried again with new-grown EHMT2 plasmids cultured in LB overnight at 37°C. This yielded much better results which was confirmed for us after running the gel, so we grew them in culture and decided to send it in for sequencing.

Our sequencing results for these three plasmids yielded almost 100% alignment except for a few letters in the sequence (none in parts which mattered most: EHMT2, puroR, CMV promoter), and therefore we considered the plasmid production to be a success.

Iteration 1: (Matches: 9124; Mismatches: 8; Gaps: 6)

Iteration 2: (Matches: 9125; Mismatches: 7; Gaps: 6)

Iteration 3: (Matches: 9127; Mismatches: 5; Gaps: 6)

Design & Build

We initially aimed to design a plasmid which was constructed to contain a CMV enhancer and the EHMT2 protein to increase production in cell lines. We sourced the plasmid’s sequence through gene synthesis, obtained the target gene, and transformed it into E-coli DH5-α. After, we transfected the said plasmid into HEK293 cell lines and ran the transfection through RT-PCR to detect the RNA levels in the cells after 4 days.

First, because we were working with cell lines and not just E-coli alone, we went through a validation process in order to confirm that our plasmid would be expressed in our cell line through The Human Protein Atlas (image shown below). We ended up settling on HEK-293 tissue cells due to its second highest expression level of our protein and was safe for us to work with.

We also kept in mind certain details of the plasmid which we believed would be optimal for cell production, including a CMV enhancer which is a strong driver of a gene, especially in mammalian cells.

*Image sourced from Addgene
Further, we looked into different isoforms of EHMT2 to make sure that the comparisons between our plasmid and primers/probes designed for qPCR aligned by using the NIH catalog and PrimerBlast which compared our sequenced primers and probes constructed in similarity to the plasmid made in cycle 1. (images below of NIH and PrimerBlast)

The primers and probes we selected are as follows for our qPCR run:
EHMT2 qPCR forward primer: TGACTGCGTGCTGTTATTCC
EHMT2 qPCR reverse primer: GAGCTTGCGGTTGAGTTGAA
EHMT2 qPCR probe: CCGCAGCTCAGGGTTGGCCCCACGT
GAPDH qPCR forward primer: TTGTCAAGCTCATTTCCTGGTAT
GAPDH qPCR reverse primer: AGGGTCTCTCTCTTCCTCTTGT
GAPDH qPCR probe: AGACCCCTGGACCACCAGCCCCAGCA

After extensive rounds of validation, we then performed a plasmid extraction protocol and sent out our plasmid for sequencing. The results came back to match our expected sequence and thus we moved on to the testing stage.

Test

For our test, we transfected the said plasmid into HEK293T cells using Lipofectamine 3000 from Thermofisher (see experiments or notebook page for more detail about our protocol), and we compared these results after a few days with cells that were not given the treatment with the plasmids transfected through qPCR. We expected to see a higher Cq value for the cells that were not given the plasmid, as it would have taken longer for them to amplify compared to those given a plasmid meant to overexpress this protein.
Note: The black line on the right is the line we are comparing for EHMT2. When referring to cells, we mean “[cells] per well in a 24 well plate”

The results of our test can be shown below:

0.5 x 10^5 cells, Treatment; transfected 3 days: Cq value of 21.41

0.5 x 10^5 cells, Control; left for 3 days: Cq value of 24.88

Normalization expression graph compared to GAPDH:

2.0 x 10^5 cells, Treatment; transfected 4 days: Cq value of 19.88

Due to the 3 day transfected well falling through, we were forced to make a comparison between transfected 4 days wells with a control of 3 days which could prove insufficient in data.

2.0 x10^5 cells, Control; left for 3 days: Cq value of 29.68

Note: Our 1.0 x 10^5 cells control group had not amplified, and therefore had no sufficient data for us to extrapolate from.

Normalization expression graph compared to GAPDH:

Learn

Although we do see a clear difference in the Cq value between cells transfected with the plasmid and cells without in the qPCR runs with genes that did amplify, we believe that the data is not comprehensive enough for a conclusive set of results. We are able to create a vague conclusion that our plasmid was effective in the sense of EHMT2 production, but our biggest block in fully trusting the validity was the test we ran with 1x10^5 cells which fell through as the control variable did not amplify. Another large problem we ran into was that the third transfection day’s well of 2x10^5 cells in the treatment group did not amplify when run through a qPCR test, therefore forcing us to compare results across two days: transfection day 4 and control of day 3. Therefore, our only set of data that we can chalk up as completely accurate is the 0.5x10^5 cells group.

For the next iteration, we could focus on not only trying to validate our 0.5x10^5 cells samples, but also test 1.0x10^5 cells and 2.0x10^5 cells once more with more than one well of control and treatment. Because we were doubling up this experiment with one testing our siRNA technology simultaneously, we were limited in the trials we attempted. Further, we could have tested our predesigned primers prior to running the qPCR in order to ensure maximum amplification, and kept transfection time constant throughout.