Part:BBa_M50047:Experience

This experience page is provided so that any user may enter their experience using this part.
Please enter how you used this part and how it worked out.

Stanford Location

Glycerol freezer stock information for the Integrase plasmid this composite part interacts with (refer to part BBa_M50042)-

Plasmid name: pD649_Integrase

DNA 2.0 Gene #: 273871

Organism: HEK 293T

Device type: actuator

Barcode #'s: 0133027143, 0133027110, 0133027160

Box label: BIOE44 F16

Applications of BBa_M50047

Methods and Results

I. Device design and creation

For the implementation of our idea, we designed two separate devices to be used in human embryonic kidney (HEK) 293T cells. These cells were first isolated from a healthy aborted fetus in the 1970s [4]. The first device is a simple plasmid which constitutively expresses the necessary Integrase protein in mammalian cells. The second device is a piece of linear DNA which is recognized by the Integrase protein and integrated into the genomic DNA of the cell.

Integrase Plasmid

The Integrase plasmid, pD649_Integrase, uses the premade mammalian vector pD649 from DNA 2.0 [5]. In this plasmid, mammalian expression is driven by the powerful constitutive promoter and enhancer found in cytomegalovirus, or CMV. After the promoter, we added the integrase gene from HIV-1. The amino acid sequence for integrase was sourced from the HIV Databases [6]. We then translated this sequence into DNA and optimized the codons for human expression using IDT’s codon optimizer [7]. Following the integrase gene, we decided to use an internal ribosome entry site, or IRES, in order to express green fluorescent protein, or GFP. This GFP will act as a reporter protein so that we can confirm both the successful transfection of the cells and the expression of the Integrase protein. Finally, there is a poly-adenosine terminator, which will act as a stopping point for transcription. Along with these elements, the plasmid also contains a number of other elements necessary for prokaryotic replication and ampicillin selection. Furthermore, the plasmid also contains puromycin resistance genes and a replication origin for stable expression in mammalian cells; however, we do not expect to use either of these elements in our experiment. The plasmid design is shown in Figure 1.

Linear DNA

When the HIV-1 genome is integrated into a genome, the Integrase protein recognizes specific sequences at the ends of the 5’ and 3’ long terminal repeats, or U3 and U5 region, respectively8. U5 sequence recognized by the Integrase is 5’-CCTCAGACCCTTTTAGTCAGTGTGGAAAATCTCTA GCAGT-3’, and U3 sequence, also recognized by the Integrase is 5’-ACTGGAAGGGCTAATT CACTCCCAAAGAAGACAAGATATC-3’ [8,9,10]. Accordingly, the ends of our linear DNA contain the terminal recognition regions from either the U5 or U3 region. While the interactions between Integrase and the specific bases of these regions are well studied, and it has been shown that only approximately 20 base pairs are necessary for the protein to bind [8,9], we decided to use the terminal 40 base pairs because previous similar experiments have shown that this may increase the efficiency of integration [3,10].

In between the Integrase recognition regions, we have all the necessary elements for the expression of another reporter protein, in this case a red fluorescent protein, or RFP. These elements will be sourced from another plasmid so their exact nature is not yet known, but most certainly these elements will include a strong constitutive promoter, a Kozak sequence, the code for RFP, and a terminator. The design of the linear DNA is shown in Figure 2. The fabrication of this linear piece is detailed in the section explaining linear DNA amplification.

II. Construction of linear DNA fragment and plasmid

In preparation for our experiment, we began by amplifying our RFP-containing plasmid to create the linear DNA described in the previous section. We performed a polymerase chain reaction, or PCR, with the thermocycler on the plasmid using Phusion polymerase. We used specific primers with free-hanging recognition regions from the U3 and U5 regions of HIV-1. The primers were designed to enclose all the necessary components for the constitutive expression of RFP. The forward primer was 5’-ACTGGAAGGGCTAATTCACTCCCAAAGAAGACAAGATATCTCAATATTGGCCATTAGCCAT-3’ and the reverse primer was 5’-ACTGCTAGAGATTTTCCACACTGACTAAAAGGGTCTGAGGTTAAGATA CATTGATGAGTTTGGAC-3’ [6]. Finally, it was purified by gel electrophoresis using the E-Gel CloneWell Gels12.We loaded the samples into the top row of wells and let them electrophorese until the bands migrated into the bottom row, where we pipetted out our purified DNA. To further concentrate this purified DNA, we conducted a PCR purification. Our linear DNA was then ready to be transfected.

III. Cells and transfection procedure After acquiring the linear DNA and the pD649_Integrase plasmid purified using midiprep [13], we transfected our HEK 293T cells by following the transfection protocol from Practical 3 [14]. There were four groups of three samples, making twelve samples in total. We had two baseline control groups, one transfected with only linear DNA (Linear) and one transfected with only pD649_Integrase (Plasmid). The experimental group was transfected with both the plasmid and the linear DNA at the same time (Both). We decided to transfect with both DNA constructs simultaneously due to the success of a similar experiment with Rous Sarcoma Virus [3]. Moreover, when transfecting the experimental group, we used 1 µg of DNA total. This 1 µg was comprised of a 1:1 ratio of plasmid to linear DNA, which was shown to have an optimal success rate [3]. We calculated the weights by comparing the number of base pairs: Linear DNA : p649_Integrase plasmid = 2151 bps : 7068 bps, thus we used 2125/(2151+7968) = 230 ng of linear DNA and 7068/(2151+7068) = 770 ng of plasmid.

Each baseline group was transfected with the same amount of DNA for that particular construct, plasmid or linear DNA, as the experimental group. In addition, our control group was not transfected but only treated with Lipofectamine as a mock transfection, allowing us to compare our results with unadapted, functioning HEK 293T cells (Control). Table 1 shows the amounts of substances used for each group.

We also kept another plate of HEK 293T cells that we passaged alongside with our experimental plate as a backup and also as a another layer of negative control group. To determine whether the Integrase successfully integrated the linear DNA into the cell genome, we passaged the cells every 4-6 days for 4 passages and observed the GFP and RFP levels when we changed media or passaged. To acquire more accurate results, we normalized the cell density when possible, aiming for 0.1 x 106 cells each time we passaged, which is the optimal seeding density for 12-well plates [15].

IV. Fluorescence Assays

Fluorescent Microscopes

The fluorescent microscope allowed for cell-level interpretation of our results even when the GFP and RFP levels were too low to be distinguishably measured by the plate reader. For example, shortly after we transfected our cells and after several passages, we noticed only a small proportion of cells expressing fluorescence. This fluorescence was only noticeable under the microscope allowing us to qualitatively compare our experimental groups even though the levels were too low for the plate reader to read reliably in the wake of other variables. Additionally, we could identify regions of particularly high concentrations of fluorescence, one such region being the strange conglomerations of cells or potential contamination.

We observed both GFP and RFP levels for each experimental group. We conducted measurements immediately after the initial transfections, every time we changed the media, and immediately before we passaged. Over the span of the experiment, we achieved four passages. Immediately after our transfections, we observed no fluorescence. This is to be expected since it takes several hours for the HEK 293T cells to be transfected with the floating DNA and express the necessary fluorescents. Within the next few days, we found that the first passage of cells provided the expected results. Despite having seen very little fluorescence when changing the media the day after transfecting, at confluence, the cells exhibited the most GFP and RFP seen throughout the entire experiment. Fortunately, our control had no GFP or RFP expression, while the Linear and Plasmid groups exhibited RFP and GFP respectively. Our Both group exhibited approximately the same amounts of RFP and GFP, proving that our transfections were successful with relatively high efficiencies (Figure 3).

We continued taking images into the second and third passages, and saw that the fluorescence naturally subdued over time. Seemingly, the RFP in the Linear group deteriorated at the same rate as the RFP in the Both group, implying that the RFP in the Both group was not successfully integrated and replicating with the cells. Naturally, the GFP in groups Plasmid and Both deteriorated at the same rate, due to the transient nature of the plasmid. An unexpected observed phenomenon became increasingly evident between passages two and three. Large conglomerations of cells had built up in every well – the control included. These conglomerations were originally predicted to be contaminants, but they ominously contained much of the fluorescence in all groups (Figure 4). When aspirating the media, these clumps of cells readily detached and were mostly removed with the media, possibly contributing to our diminution of fluorescence over time. See the discussion section for more about this phenomenon.

Passage four allowed us to make the most relevant observations with respect to the desired result. In the Linear and Both groups, we continued to see some RFP expression, and when increasing the magnification, we noticed that this may have been due to natural integration of the linear DNA without any aid of Integrase. We made this conclusion based on the close proximity of cells expressing RFP with approximately the same intensities in a sea of little other fluorescence – this possibly resembled a successfully integrated cell multiplying. This observed integration seemed to occur more frequently in the Linear group than the Both group. Still, it was difficult to discern these results due to the many confounding variables (Figure 5). On the other hand, as expected, during this passage, a significant amount of GFP had deteriorated in groups Plasmid and Both. The entirety of the microscope images can be found in the supplement.

Plate Reader

While the fluorescent microscopes let us gauge qualitative comparisons between wells and over time, the plate reader provided a much more structured and quantitative method of collecting the GFP and RFP fluorescence of the wells. When structuring our plate reader experiment, we allowed the instrument to conduct twenty-five measurements per well and subsequently average them together for a finalized value of RFP or GFP. These twenty-five measurements were to account for the uneven distributions of fluorescence in each well. At the conclusion of passage one, the results showed that our transfections were successful and correlated with our microscope images. During passage two and three, however, cells began to grow at increasingly variable rates causing drastic differences in cell densities. Consequently, the red Dulbecco's Modified Eagle Medium, or DMEM complete growth media, had widely variable discoloration among wells. We observed that the redness of the media affected the reading of RFP in the plate reader scans: the less cells per well, the redder the media, and the more RFP expression read by the plate reader. This factor most definitely had the potential to completely wash out any actual RFP expression. Due to decreased growth rates in the Plasmid and Both groups, some data falsely suggests that these wells had high levels of RFP expression. During passage four, our plate reader results were too skewed to draw reasonable conclusions due to potential foreign microorganism contamination and drastically differing growth rates.

Figure 6 displays the plate reader results for the first three passages. Unfortunately, we could not conclude that RFP expression remained constant in the Both group. Rather, by passage three we had a number of confounding variables that altered our results. Bleed-over from the high levels of RFP expression in the Linear group potentially caused high reading levels of GFP. Therefore, the group with linear DNA appears to have expressed more GFP than the Both group by passage three. Additionally, the Plasmid group seemingly displayed more RFP than the Linear group in passage 2. This could have been due to bleed-over or differing discoloration of DMEM media. Scaling according to cell density may have skewed the results as well because the cell density measurements are not always exact and had to be estimated in some cases due to lack of slides for the cell counter. We believe that relatively low cell density during passage one and two in the Both group caused the overrepresentation of the fluorescence in the this group despite trying to control for this variable. Hence, our expression is excessively high for the Both group during our first two passages. All in all the graph is not perfectly representative of our microscope images due to a variety of variables which were unforeseen and/or difficult to control. The raw data for this graph and code in R and can be found in the supplement.

Discussion

Passage one and two provided clear evidence that our transfections were successful based on the fluorescent microscope images and the plate readings; we had comparatively large amounts of GFP and RFP both quantitatively and qualitatively in the appropriate wells. Yet, our transfection efficiency could have been improved had we waited a day after performing our first passage to transfect. Passaging the cells the second time, we had a limit on the number of cell counter slides. Because of this material constraint, we could not measure all of the cell densities and hence could not dilute accordingly to the optimal seeding density of a 12-well plate. Therefore, we proceeded in diluting all the wells by a factor of one to seven and subsequently estimated cell densities from microscope images for fluorescence scaling. Since each well had a different cell density upon passaging – potentially because the transfections killed many cells or because the replication cycle of a cell can be reduced or increased based on the presence of other DNA – we began passage two with inconsistent cell densities, further propagating this inconsistency. Without this material constraint and by optimizing our transfection procedure we could potentially see better and more conclusive results.

During passage three, the large cells clumps became another issue. Under the 40X objective, these large clumps appeared to be made of cells, potentially rounded HEK 293T cells or another type of mammalian cell. We also made the observation that most of our fluorescence derived from these clumped cells, suggesting that they were present during the initial transfection. At first we thought that this may have occurred due to insertional mutagenesis. During integration vital genes for the cells may have been spliced, affecting their life cycles and potentially their morphology. Therefore, these cells with successfully integrated genes conglomerated into the observed clumps. The caveat to this reasoning is that the Control group with no DNA had these clumps as well, although to a slightly lesser degree. Our next hypothesis was that the clumps were caused by some sort of contamination during the experiment either from the media or the environment. While we cannot rule this out completely, we grew the original cell line we received alongside the experiment throughout its entirely. It was subject to the same media and passage procedure, but it did not undergo a transfection procedure. There were no strange conglomerations observed in these cells suggesting that the source of the problem was not media contamination or contamination from the environment. This left us with one final variable which could be causing the phenomenon: the transfection procedure, specifically the Lipofectamine 3000 or the Opti-MEM. Our most probable explanation for these conglomerations of cells based on all the evidence is that they are caused by a morphological shift in the HEK 293T cells induced by contamination in the Lipofectamine 3000 or the opti-MEM. In any case, the conglomerations of cells greatly hindered our ability to interpret our results because the rounded morphology prevented these cells from adhering to the floor of the well. Therefore, these cells were aspirated away during standard media changes and while passing, and if these cells really tended to be those which received the Lipofectamine, then the majority of our transfected cells would have been lost.

During our fourth and final passage, we made some striking observations, including the potentially higher-yielding integration by natural means over Integrase-mediated integration. The lower integration in the Both group may have even been attributed to the presence of plasmid. Still, a number of confounding variables, including inconsistency of cell densities, contamination, and transfection efficiencies, present these results as largely inconclusive.

In the future, to better these results, we should revisit our constructs. Particularly we should revisit our integrase gene and Integrase protein. We have no direct evidence that the protein was expressed besides the GFP reporter, nor do we know if the protein is functional. Furthermore, we should optimize our transfection procedure and try it again with fresh Lipofectamine and Opti-MEM. This would both increase our transfection efficiency and eliminate the conglomerations of cells. To improve our measurements using the plate reader, we should replace the media with a clear alternative such as Dulbecco's phosphate-buffered saline before reading, and we should keep our dilution techniques consistent across all passages to better control for cell density. With these changes, we believe we could demonstrate the the effectiveness of HIV-1 Integrase-mediated transfection as it has been demonstrated with other viruses such as Rous Sarcoma Virus [3].

References:

1. Kim, T.K., Eberwine, J.H. 2010. Mammalian cell transfection: the present and the future. Anal Bioanal Chem. 397(8):3173–3178.

2. Nayerossadat, N., Maedeh, T., and Ali, P.A. 2012. Viral and nonviral delivery systems for gene delivery. Adv Biomed Res. 1:27.

3. Mizuarai S, et al. 1999. Integrase-mediated nonviral gene transfection with enhanced integration efficiency. J Biosci Bioeng. 88(5):461-7.

4. HEK 293 Cell Line [Internet]. [cited 2016 Oct 30]. Available from: http://www.hek293.com/

5. Mammalian Expression Vectors [Internet]. DNA 2.0 [cited 2016 Oct 30]. Available from: https://www.dna20.com/eCommerce/catalog/datasheet/448

6. HIV Databases [Internet]. Los Alamos: Los Alamos National Security; c2005-2006 [cited 2016 Oct 30]. Available from: http://www.hiv.lanl.gov/

7. Codon Optimization Tool [Internet]. Redwood City: Integrated DNA Technologies [cited 2016 Oct 30]. Available from: https://www.idtdna.com/CodonOpt

8. Kessl, J.J., et al. 2009. HIV-1 Integrase-DNA Recognition Mechanisms. Viruses. 1(3):713–736.

9. Esposito D., Craigie R. 1998. Sequence specificity of viral end DNA binding by HIV-1 Integrase reveals critical regions for protein-DNA interaction. EMBO J. 17(19):5832-43.

10. 2006. HIV-1, complete genome. GenBank:AF033819.3.

11. HEK 293T Cell Line: Cat #HCL4517 [Internet]. Buckinghamshire, UK: GE Healthcare; c2015 [cited 2016 Oct 30]. Available from: http://dharmacon.gelifesciences.com/uploadedFiles/Resources/tla-hek293t-cell-line-manual.pdf

12. Thermo Fisher Scientific Inc. 2016. Gel-purify your DNA in 3 simple steps [Internet]. [cited 2016 Nov 15]. Available from: https://www.thermofisher.com/us/en/home/life-science/dna-rna-purification-analysis/nucleic-acid-gel-electrophoresis/e-gel-electrophoresis-system/e-gel-pre-cast-agarose-gels/e-gel-clonewell.html

13. QIAGEN. 2012. QIAGEN Plasmid Purification Handbook. “Protocol: Plasmid or Cosmid DNA Purification using QIAGEN Plasmid Midi and Maxi Kits.” 17-21.

14. Qi, S., Rogers, K. 2016. Lab: Practical #3 - Getting DNA into cells.

15. Thermo Fisher Scientific Inc. 2016. Useful information for various sizes of cell culture dishes and flasks [Internet]. [cited 2016 Nov 17]. Available from: https://www.thermofisher.com/us/en/home/references/gibco-cell-culture-basics/cell-culture-protocols/cell-culture-useful-numbers.html

User Reviews

UNIQae5c14b851fa9c53-partinfo-00000000-QINU UNIQae5c14b851fa9c53-partinfo-00000001-QINU