Plate Reader Secretion Study

In the first stage of building ICARUS, we set out to identify signal peptides for most efficient secretion of soluble antigens. As one of the measures for secretion efficiency, we investigated the fluorescence intensity of sfGFP tagged with an N-terminal signal peptide in the cell supernatant.

Figure 1: Microscopy images of HEK293T cell line transfected with secretion constructs. Constructs HA-sfGFP and no tag sfGFP were used as negative controls.

50 µL of Lipofectamine 2000-DNA complexes in OptiMEM were formed for transfection experments in 96-well plates. HEK293T cells were reverse transfected at a seeding density of 25,000 per well in 150 µL of DMEM + 10% FBS + 1% Penicillin/Streptomycin and incubated for 2 days post-transfection. First, the cells were analyzed under the fluorescence microscope using a FITC filter cube. The negative control cells shown stable expression of sfGFP (please see our engineering page for construct designs) in the cytoplasm and low background fluorescence. In contrast, experimental samples showns secretion of sfGFP towards the outside of the cells (Figure 1), which can be characterized by lower sfGFP fluorescence in the cytoplasm and higher background fluorescence. In particular, we noticed a difference in the distribution of IFN\(\mu\) and HSA-tagged sfGFP in microscopic analysis, with darker cytoplasmic interiors and what appears to be an outside-facing protein.

We quantified fluorescence intensity in a plate reader assay with secretion constructs. The cell supernatant was collected in 100 \(\mu\)L from the transfected HEK293Ts 2 days post-transfection and suspended in Dulbecco's Modified Eagle Medium (DMEM) for analysis.

Figure 2: Supernatant fluorescence intensity collected from cells transfected with secretion constructs. The fold difference was calculated by comparing the readings to negative control readings. Values are mean \(\pm\) SD. Statistical analysis by 1-way ANOVA with Tukey's post test, **p < 0.01, ***p < 0.001, ****p < 0.0001. The study was conducted with 4 biological repeats.

We measured the fold difference of sfGFP fluorescence for different secretion constructs. We did not see significant difference in sfGFP fold difference between the two negative control samples (no tag vs. HA-tag). The supernatant sfGFP fold difference was significantly higher in the HSA-sfGFP (21 fold difference) and IFN-sfGFP (12 fold difference). The ANOVA analysis revealed significant differences between experimental samples and controls, which suggests efficient secretion as well as a statistically significant difference between the IFN and HSA signal peptides suggesting that HSA might be more efficient at secreting sfGFP. We suspect the negative value of the HA construct can be attributed to high background fluorescence of DMEM, which motivated our calibration curve comparison.

Flow Cytometry

We collaborated with Asimov to determine the effect of promoter strength on secretion efficiency. We used this opportunity to test additional signal peptides as well. As our second way of evaluating secretion efficiency, we looked at fluorescence intensity of single cells compared to fluorescence intensity of a transfection control (please see our engineering page for construct designs).

Figure 3. Histograms of YFP fluorescence intensity comparing secretion tags to HA negative control for constructs with different promoters. All histograms represent BFP+, YFP+ single cell gated HEK293T cells for representative samples.

YFP fluorescence in HEK Freesytle cells, a suspension cell line, was measured using a Beckman Coulter Cytoflex S flow cytometer post-transfection with secretion constructs. Secretion was evaluated for selected peptides driven by different promoters. Shift in YFP+ cell population was observed in constructs with EF1$\alpha$ and CMV promoters. This increase in YFP fluorescence intensity can be attributed to YFP remaining inside the cytoplasm rather than being secreted suggesting that negative secretion controls did not release YFP outside of the cell as expected.

Figure 4: Median YFP fluorescence intensity normalized to median BFP fluorescence intensity in HEK293T cells transfected with secretion constructs. The data was grouped according to promoter that expressed the reporter. HA signal peptide was used as the negative control. Values are mean \(\pm\) SEM.

Numerical quantification of fold difference in YFP fluorescence intensity revealed similar observations. The control HA-sfGFP construct shown heightened fluorescence in the cytoplasm of the cell compared to secretion constructs. SV40 was previously reported as a weaker promoter compared to CMV and EF1\(\alpha\), which might explain the non-significant differences in fluorescence between the negative control and signal peptides in that group.

Šídák's multiple comparisons test	Mean Diff.	95.00% CI of diff.	Below threshold?	Summary	Adjusted P Value
HA vs. Tcadherin	4.512	2.712 to 6.312	Yes	****	\[\lt0.0001\]
HA vs. IgKappa	4.771	2.971 to 6.571	Yes	****	\[\lt0.0001\]
HA vs. BM40	4.649	2.848 to 6.449	Yes	****	\[\lt0.0001\]
HA vs. Trypsinogen	4.887	2.825 to 6.950	Yes	****	\[\lt0.0001\]
HA vs. IL2	5.803	3.741 to 7.865	Yes	****	\[\lt0.0001\]

Table 1: Results of the Sidak post-test after 2-way ANOVA statistical analysis of the flow cytometry data. Only secretion peptide effect was considered in the analysis.

The 2-way ANOVA analysis with post-hoc testing revealed significant effect of signal peptide type on secretion of the cargo, which indicates successful secretion of YFP out of the cell. Flow cytometry analysis can be viewed as the “inverse” of our secreted study, with an improved perspective on the distribution of internal fluorescence within cells that may correlate to secretion efficiency. Further analysis of the coefficients of variation and single-cell analyses of high secretor cells vs. low secretor cells may reveal cell properties that promote better secretion, such as plasmid copy number.

Calibration Curve

Our team wanted to match the fluorescence readings to the concentration of sfGFP in the cell supernatant using the iGEM Measurement Committee Fluorescein calibration protocol. Given that different plate readers may exhibit different optics and data collection schemes, we wanted to provide a controlled measurement for comparison of signal peptides between labs. During the process of establishing a fluorescein calibration curve with PBS, as is recommended, and comparing to DMEM, our secretion diluent, we noticed high background fluorescence of DMEM, which motivated us to compare the calibration curves in phosphate buffered saline (PBS), DMEM and OptiMEM, a serum-free and more optically neutral media compared to phenol red found in DMEM.

Figure 5: Fluorescein calibration curve in common cell culture solutions. The lines are the result of simple linear regression and the values are mean \(\pm\) SD of four replicates.

Linear regression was fitted to fluorescence readings of fluorescein in the different media and the slopes of regression lines were statistically compared. The slopes were found to be significantly different between the PBS and DMEM dilution series, as well as the PBS and OptiMEM dilution series. This calibration experiment initially demonstrated that DMEM with FBS and phenol red may not be the best optically neutral media to perform our secretion studies, whether this quenches fluorescence or interferes with readings.

We were surprised that OptiMEM as the diluent still exhibited significant variation between the PBS standard curve and thus, may require further study on appropriate buffers to perform the fluorescein calibration curve to better compare measurements across experiments, labs, and plate readers. Further experiments with DMEM + FBS but no phenol red or an appropriate dilution of supernatant with PBS would elucidate the effectiveness of this calibration protocol for future iGEM teams working on mammalian synthetic biology projects.

References

1. Qin, J. Y., Zhang, L., Clift, K. L., Hulur, I., Xiang, A. P., Ren, B.-Z., & Lahn, B. T. (2010). Systematic Comparison of Constitutive Promoters and the Doxycycline-Inducible Promoter. PLOS ONE, 5(5), e10611. https://doi.org/10.1371/journal.pone.0010611
2. Richard Tennant, Paul Rutten 2019. Calibration Protocol - Fluorescence Standard Curve with Fluorescein. protocols.io https://dx.doi.org/10.17504/protocols.io.zgkf3uw
3. Chen, H.-Y., & Chen, C. (2022). Evaluation of Calibration Equations by Using Regression Analysis: An Example of Chemical Analysis. Sensors (Basel, Switzerland), 22(2), 447. https://doi.org/10.3390/s22020447
4. Grasso, S., Dabene, V., Hendriks, M. M. W. B., Zwartjens, P., Pellaux, R., Held, M., Panke, S., van Dijl, J. M., Meyer, A., & van Rij, T. (2023). Signal Peptide Efficiency: From High-Throughput Data to Prediction and Explanation. ACS Synthetic Biology, 12(2), 390–404. https://doi.org/10.1021/acssynbio.2c00328