Team:SUSTech Shenzhen/Design

Load Tools

Back to Top

Design

To spatiotemporally control the gene expression of mammalian cells by using the ' LightON system', we develop three essential ‘modules’ which can be finally assembled into the ‘controlling hoop’. They are ‘multi-level output strategy, automatic illumination and sample collection system, model-based predictive and control system’.


Background

Introduction of LightON system


Figure 1. The principle of LightON system
Figure 1. The principle of LightON system


Light is a more feasible factor to switch the particular gene expression compared to chemicals or other factors because it is easy to be modulated spatiotemporally, and quantitatively.[1] X. Wang et al [2] developed the LightON system, which consists of a single chimeric protein (GAVPO) that can forms homodimer and bind to its promoter upon exposure to blue light, initiate transcription of the target gene. This system is an suitable regulated gene expression system for our project, as it has low background expression, low toxicity (need weak light for induction), low interference with endogenous proteins or genes and the capacity for temporal and spatial control, and can be easy to manipulate. These characteristics provide us with the capability for gene activation with good spatial, temporal and quantitative control in an easy-to-use system.


Construction

To characterize the gene expression in mammalian cells, we first constructed two plasmids, 5xUAS-mRuby-P2A-hGluc and EF1α-GAVPO-Bla. The reason why we choose hGluc as our target protein is that this protein has low molecular weight and its the detection assay is sensitive. Thus it can be used to mimic the production and secretion process of low molecular weight proteins such as cytokines. Also, we construct a plasmid 5xUAS-mRuby-P2A-hGluc-P2A-IL-10 to test the accuracy of regulation in a more realistic condition. Next, we constructed a plasmid 5xUAS-mRuby-P2A-hGluc-P2A-IL-8. Since IL8 can promote the activation and chemotaxis of neutrophil, it can be detected by observing the migration of neutrophil which is a common process in inflammation and immune reaction.


Figure 2. The schematic diagram of plasmids
Figure 2. The schematic diagram of plasmids


And we choose Hela cell as a cell carrier, because it has been widely used in mammalian synthetic biology and laboratory studies in anti-inflammation and immunotherapy.[3]


Module1: Multi-level output strategy


In the previous studies, most engineered cells are regulated by the ‘direct feedback control system,’ which iteratively computes the system input by comparing the real-time system output to the desired tracking objective[4][5]. While this method is effective, the tracking accuracy was rather modest and the controller was incapable of robust, precise regulation in the face of complex environment (eg.tumor microenvironment) and hour-to-hour variability caused by external perturbations and intrinsic change of cells. Thus, it is necessary to understand the cellular behaviors on different scales which contain ‘protein expression at a time-scale (from transcription all the way to protein secretion), cell state(metabolic activity) and migration at a space-scale’.


i.Protein expression in time-scale


Figure 3. Several steps to regulate eukaryotic gene expression
Figure 3. Several steps to regulate eukaryotic gene expression


Since the eukaryotic secretion pathway is complicated, we try to reveal the relationship between light exposure and gene expression at multi-level(transcription level, translation level, and secretion level). And we choose different methods to help characterize these processes, which is essential for Acquisition of experimental parameters.


As mentioned before, we used the 'LightON system' to regulate gene expression in mammalian cells which can be more convenient and sensitive. And In the first place, we need to know the optimal condition of blue light exposure, then we can precisely decide the range of light conditions in the latter experiments. To achieve this goal, we design two gradient experiments. One is to obtain the relationship between target gene expression and blue light intensity, thus we set up a series of light intensities gradient from 0 to 812.4 uW/cm2 and choose 48h as the illuminating duration.


Figure 4. Intensity-gradient experiment. Input: different intensities of light(uW) ; Output: red fluorescence(RFU). Data represent the mean of three independent experiments, and error bars indicate the standard deviations.
Figure 4. Intensity-gradient experiment. Input: different intensities of light(uW) ; Output: red fluorescence(RFU). Data represent the mean of three independent experiments, and error bars indicate the standard deviations.


Another is to obtain the relationship between target gene expression and illumination time. And we set up a series of time gradients from 0 to 60h and choose 102.4 uW as light intensities according to the previous experiment.

Figure 5. Time-gradient experiment. Input: different illumination time(hours) ; Output: red fluorescence(RFU). Data represent the mean of three independent experiments, and error bars indicate the standard deviations.
Figure 5. Time-gradient experiment. Input: different illumination time(hours) ; Output: red fluorescence(RFU). Data represent the mean of three independent experiments, and error bars indicate the standard deviations.


For both of the experiments, we use Hela-5xUAS-mRuby-P2A-hGluc as experiment group and use Hela-5xUAS-mRuby as control group. After obtaining the optimal illumination condition, we start to quantitatively characterize the whole expression process at different levels. We first try to characterize the transcription process by testing the dynamic change of RNA through quantitative PCR. Since the RNA is less stable, and can be rapidly degraded, its expression level should reach plateau much quicker. We reduce the total measurement time to 20 hours and the interval is 1 hour .


Figure 6. Schematic diagram of 'Multi-level output strategy'
Figure 6. Schematic diagram of 'Multi-level output strategy'


Next, we try to characterize the translation process by testing the dynamic change of mRuby using flow cytometry analysis. And to compare with the change of RNA and secreted protein, we need to simultaneously measure the cell number at each time point (since the output of the red fluorescence is a mean value over single cells). The final step is to characterize the secretion process. Since we have chosen hGluc as our target product, we did it by measuring the chemoluminescence value.


Figure 7. 3D-culture cells
Figure 7. 3D-culture cells


To further increase the complexity of regulation, we perform a functionality test in 3D-culture cells which might reflect the in vivo condition more accurately. A stability test and an intensity gradient test were given to better reveal the light-induced gene expression level in physiological condition.


ii. Cell state(metabolic activity)


For most of our experiments, it is observed that light can have significant effects on the cell state that can lead to large inaccuracies in regulation, especially on the transcription level. Thus, it is essential to consider the cell state during illumination. WST-8 is a kind of water-soluble tetrazolium salts dye, which has been widely used in the detection of cell activity and proliferation.


Figure 8. The principle of CCK8
Figure 8. The principle of CCK8


Therefore, We used this reagent to detect the metabolic activity of cells underlying different illumination time. And it’s also helpful to elucidate the mechanism about how light can alter the cell state, which can further improve the accuracy in regulation.


iii.Migration in space-scale


Although we have characterized the secretion process by measuring the relative amount of secreted protein in previous experiments, it’s still essential to verify the biological activity of secreted protein which is proof of the effectiveness of regulation (Since some protein may lose their biological activity during secretion process). And we choose Chemotaxis as our testing model which is described as the directed migration of cells towards a chemoattractant. We also used IL8 as the chemoattractant. IL8 can promote the activation and chemotaxis of neutrophils. We perform this chemotaxis assay in the ibidi μ-Slide Chemotaxis (a commercial microfluidics instrument), which allows us to investigate the detailed chemotactic behavior of target cells.


Figure 9. IL8 guided trafficking of neutrophils
Figure 9. IL8-guided trafficking of neutrophils


We designed three groups of experiments to verify the ability of IL8 to induce cell migration. The first group is Control-Control, which fills both chambers with cell culture medium. The second group is IL8-IL8, which fills both chambers with secreted IL8 from illuminated cell lines. The third group is IL8-Control, which fills one chamber with secreted IL8, another chamber with a culture medium.


Figure 10. Experiment design of chemotaxis
Figure 10. Experiment design of chemotaxis
Figure 11. Schematic representation of Chemotaxis instrument
Figure 11. Schematic representation of Chemotaxis instrument


Also, in order to give a visual inspection of the living cell trajectories, including the cell velocity and directionality with CONSIDERABLE amount of time (fluorescence imaging might affect the cells because the frequent illumination with intense excitation light), we used brightfield imaging with dark red light, and managed to design an algorithm based on the OpenCV CSRT/CSRDCF tracker which uses the spatial reliability map for adjusting the filter according to a set of learned filter sets.


Figure 12. The workflow of designing cell tracking algorithm
Figure 12. The workflow of designing cell tracking algorithm


In the next step, we want to further explore the secretion mode of real-time lighting engineered cells by observing other cells’ migration, which can help us better understand this process. Thus, we designed a new microfluidics chip that allows for in real-time illuminating engineered cells and observing cells’ migration.


Figure 13. Schematic diagram of the new microfluidics chip
Figure 13. Schematic diagram of the new microfluidics chip


Module2: Automatic illumination and sample collection system


Unfortunately, well-to-well variation and disturbance in a biochemical system are unavoidable for most of the control systems.[6] However, the performance features of our control system, such as well-to-well variability and external disturbance, could be improved with further engineering, thereby extending the applicability of our method to more biological problems.


Figure 14. Automatic illumination and sample collection system
Figure 14. Automatic illumination and sample collection system


To eliminate the variability and increase data collection throughput, we design an automatic illumination and sample collection system, which is capable of projection a constant light signal, as well as continuously refreshing and collecting the cell culture medium during illumination. We verified the Automatic-Culturing System with cell illumination experiment in two scales, time interval, and light intensity by comparing the result with the same experiment we have done artificially.


For More Details


Module3: Model-based predictive and control system


To quantitatively understand and predict the light-induced gene expression, a fine-tuned mathematical model obtained through a long characterization process of cells is required.[7] Thus, we try to establish and parameterize ordinary differential equations based on previous results in ‘ Protein expression in time-scale ’ and ‘ cell state ’. Furthermore, we design another three experiments to help us better characterize the dynamics of the model.


Experiment1: Switch off kinetics of secreted protein


For parameterizing the light-dependent shut-on and shut-off, cells were first illuminated for 15h, then incubated for 30h in the dark. The measured output is the concentration of hGluc and the concentration of hGluc-mRNA in a single cell normalized to the basal mRNA expression at different time points.


Experiment2: Multi-intensity time gradient experiment


Since some parameters in our model can be intensity-dependent, We need to fit them under different intensities at different time points. Thus we design a time gradient experiment with five different intensities according to the functional relationship between relevant parameters and light intensity.


Experiment3: Regulation test


In this experiment, relevant control sequences that are returned from our simulation algorithm are input into the computer to make the concentration of hGluc fluctuate within a certain range, which is a proof-of-concept that we can effectively control its concentration changes under in vitro conditions.


For More Details

References

  1. ↑ Mansouri, M. , Strittmatter, T. , & Fussenegger, M. . (2018). Light-controlled mammalian cells and their therapeutic applications in synthetic biology. Advanced Science.
  2. ↑ Wang, X. , Chen, X. , & Yang, Y. . (2012). Spatiotemporal control of gene expression by a light-switchable transgene system. Nature Methods, 9(3), 266-269.
  3. ↑ Mansouri, M. , Strittmatter, T. , & Fussenegger, M. . (2018). Light-controlled mammalian cells and their therapeutic applications in synthetic biology. Advanced Science
  4. ↑ Milias-Argeitis, A. , Rullan, M. , Aoki, S. K. , Buchmann, P. , & Khammash, M. . (2016). Automated optogenetic feedback control for precise and robust regulation of gene expression and cell growth. Nature Communications, 7, 12546.
  5. ↑ Shao, J. , Xue, S. , Yu, G. , Yu, Y. , Yang, X. , & Bai, Y. , et al. (2017). Smartphone-controlled optogenetically engineered cells enable semiautomatic glucose homeostasis in diabetic mice. Science Translational Medicine, 9(387), eaal2298
  6. ↑ Olson, E. J. , Hartsough, L. A. , Landry, B. P. , Shroff, R. , & Tabor, J. J. . (2014). Characterizing bacterial gene circuit dynamics with optically programmed gene expression signals. Nature Methods,11(4), 449-455.
  7. ↑ Müller, K., Engesser, R., Metzger, S., Schulz, S., Kämpf, M. M., Busacker, M., … Weber, W. (2013). A red/far-red light-responsive bi-stable toggle switch to control gene expression in mammalian cells. Nucleic acids research, 41(7), e77. doi:10.1093/nar/gkt002