Team:TECHNION-ISRAEL/Results

Results

• Introduction
• Tri-Display
• Timing
• Tri-Display & Timing
• Tolerance assay
• Discussion

Introduction

The goal of our project is to prevent allergies and autoimmune diseases by displaying their characteristic epitopes on the membrane of hematopoietic stem cells. This will, theoretically, lead to deletion of the harmful immune cells, thus precluding immune reaction towards harmless substances or “self.”

Experiments

Our experiments can be divided into three categories:

1. Tri-Display
2. Timing
3. Tolerance assay

Tri-Display

The full protocols for this experiment can be seen here.

One of our goals was to successfully express three different proteins on the cellular membrane using our Tri-Display plasmid. We tested expression in our model cell line HEK293. In order to quantify the expression level of each protein on the cellular membrane we used three different antibodies, specific to three different protein tags, conjugated to fluorophores (APC, FITC and PE) and measured the fluorescence using flow cytometry (figure 1).

Figure 1: Antibodies conjugated to fluorophores bind to the tags on the membrane.

In this experiment we tested four potential Tri-Display plasmids we designed, in order to determine which worked best. For positive control we used the commercial pDisplay vector, which contains one epitope with all three tags . All results were compared to a negative control of non-transfected cells (figure 2).

Figure 2: Median fluorescence of four different optimized Tri-Display constructs

We found that the best Tri-Display vector, capable of sufficiently expressing all three proteins in almost perfectly equimolar ratios, is the T2A-P2A construct. Accordingly, we proceeded to insert the T2A-P2A based display vector into our final ToleGen construct.

Timing

The full protocols for this experiment can be seen here.
In order to test our delay mechanism, we co-transfected the TRE-GFP plasmid (BBa_K2520006) with the CMV-tTA plasmid (BBa_K2520009) into our HSC model (figure 3). After induction, we expected to see that the florescent intensity is inversely proportional to the Doxycycline (analog to Tetracycline inducer) concentration, meaning a higher concentration of Doxycycline would correspond to a lower fluorescent intensity. For negative control we co-transfected TRE-GFP and pUC19 into our HSC model, allowing us to test the basal expression of our system (figure 4).

Figure 3: Co-transfection of TRE-GFP plasmid and CMV-tTA plasmid in our HSC model.

Figure 4: Median fluorescence of cells expressing GFP with increasing concentrations of Doxycycline

The results shown in figure 4 are in line with our expectations. A clear decrease in fluorescence intensity can be seen with increased concentrations of Doxycycline. At a Doxycycline concentration of 10ng/ml, fluorescence intensity stabilizes and reaches plateau, indicating very low expression. The negative control, TRE-GFP plasmid with pUC19, shows no fluorescent intensity for GFP, indicating that the leakiness of our Tet-Off system is very low in the abscence of tTA.

Tri-Display with Timing Mechanism experiment

The full protocols for this experiment can be seen here.
Our final experiment was to test the ToleGen construct we created (figure 5).

Figure 5: Modular ToleGen plasmid TRE-Tri-Display

This experiment combined both induction and immunofluorescent staining. We co-transfected the TRE-Tri-Display (BBa_K2520007)and CMV-tTA (BBa_K2520009) plasmids into our HSC model. TRE-GFP was co-transfected with pUC19 into our model HSCs (HEK293) and served as a negative control. After 24 hours we induced the cells with 6 different concentrations of Doxycycline. After 48 hours, we stained the cells with antibodies and analyzed them using flow cytometry (figure 6). We expected to see a decrease in the fluorescent intensity of all three proteins (as we saw in the delay mechanism experiment).

Figure 6: Median fluorescence of the cells in three different channels (APC, FITC and PE), corresponding to the three different proteins, in different concentrations of Doxycycline

From the results shown in figure 6 we can see a general decrease in the membrane expression of all three proteins as the concentration of doxycycline is increased.
Still, the results of this experiment are not fully satisfactory, as the expression is inconsistent with expectations at two specific concentrations, and the general fluorescence intensity is lower than we expected. We believe this may be due to the very long nature of the experiment, the dependence on co-transfection combined with the complexity of the construct, and complications that delayed our flow cytometry analysis. Unfortunately, there was no time to conduct this experiment again.
In the future, we plan on repeating this experiment, possibly while using a cell line that constitutively expresses the tTA protein. This would allow us to forgo co-transfection and achieve more consistent results.

Tolerance assay

Apoptosis test with soluble anti-IgM

In this experiment we confirmed that our B cell model (WEHI-231 cells) undergoes apoptosis with the addition of soluble anti-IgM. We checked the percentage of apoptotic cells using the commercial “Dead Cells Apoptosis Kit” (Invitrogen). Apoptosis was measured at 24 hours, and compared to a control (B cell model without anti-IgM).

Figure 7: Detection of apoptosis in WEHI-231 cells using Annexin V FITC and PI assay

As can be seen from the above graphs, the percentage of model B cells that were double positive (upper right quadrant, signifying dead cells) was higher in the experiment plates (with anti-IgM) as compared to the control (without anti-IgM). In addition, there were no positive cells for Annexin V alone, meaning that there were no early apoptotic cells.

Testing Our Assay Plasmid

After we proved that our B cell model undergoes apoptosis when exposed to soluble anti-IgM, we moved on to our final assay- proof that our HSC model (HEK293), transfected with our assay plasmid can induce apoptosis in our model B cells, and thus, can induce tolerance.
First, we transfected our HSC model with our assay plasmid in order to verify that the anti-IgM construct we created was functional. The anti-IgM was expressed under two promoters- a mutant EF1a ( BBa_K2520023) promoter and a CMV promoter. We tested the functionality of our anti-IgM scFv-FC-Fusion using soluble murine IgM conjugated to a fluorophore, and compared the median fluorescence to non-transfected cells using flow cytometry (figure 8).

Figure 8: Fold change of the median fluorescence after extracellular staining with IgM conjugated to FITC, as compared with a negative control (non-transfected cells).

As can be seen, our construct was able to bind IgM more specifically than the control, though further optimization is clearly necessary. It is important to mention that although the experiment plates bound IgM more specifically than the control, the percentage of cells expressing anti-IgM was very low (0.0175% in the control, 1.535% under the CMV promoter and 1.46% under the EF1a promoter). We believe this is due to the complex nature of the antibody being displayed, and the need for dimerization on the membrane. For more on how we designed this construct, please see our Tolerance Assay page.

Tolerance assay

Finally, we attempted to induce apoptosis in our immature B cell model with an anti-IgM presenting HSC model.

We transfected the HSC model with our assay plasmid. After 24 hours, we washed the cells, and cocultured them with our B cell model. After an additional 24 hours (48 hours after transfection) we quantified the apoptosis of the B cell model using a “Dead Cells Apoptosis Kit” (Invitrogen) and compared the results to a control plate of our B cell model in mono-culture (figure 9).

Figure 9: Detection of apoptosis in WEHI-231 cells using Annexin V FITC and PI assay

As can be seen from the above graphs (figure 9), and in accordance with our model's prediction, the percentage of dead B cell model was nearly twice that of the control (cocultured with non-transfected HSC model). Therefore, we can conclude that the B cell model underwent apoptosis as a result of our model HSCs presenting anti-IgM. This supports the claim that our display system is capable of presenting epitopes effectively and can induce tolerance.

Discussion

When we began working in the lab, we had two goals:

1. Create a functional and inducible Tri-Display plasmid that incorporates the different elements of our design.
2. Demonstrate that our system induces immune tolerance

We believe that we have achieved both these goals. We successfully tested our display mechanism and inducible promoter separately, demonstrated that cells transfected with our modular display plasmid can induce immune tolerance, and lastly, we showed that the inducible Tri-Display plasmid showed promising results, but still requires further testing and optimization.
In the future, we would like to further optimize both the display and delay mechanisms in order to achieve more robust and consistent expression. Furthermore, because of an unfortunate shipping delay, we did not have time to test our “Kill Switch” mechanism. In the future we plan on combining all three systems, the inducible promoter, Tri-Display system, and “Kill Switch,” thus creating the final ToleGen plasmid.
Finally, a tremendous amount of work went in to working with our second model cell line, HPC-7 cells. Not every endeavor can be successful, and within the time frame of iGEM we were not able to conduct our project experiments in these cells. Towards the end of our project we managed to successfully transfect these cells, but time ran out before we could conduct further experiments. In an effort to aid future teams and scientists we documented our efforts in detail HERE.

1. Gottschalk, Alexander R., and José Quintáns. "Apoptosis in B lymphocytes: the WEHI-231 perspective." Immunology & Cell Biology 73.1 (1995).