Loading menubar.....

Team:UCSF/Results

HOVER TO
ENLARGE

EXPERIMENTS/RESULTS

PROJECT ACHIEVEMENTS


  1. We built a communication circuit in yeast that exhibits stable divergent populations after communication even though their responses to an initial stimulus are similar.
  2. We characterized several communication parameters to further optimize this output.
  3. We submitted and characterized BioBrick parts of almost all components of our circuit.

EXPERIMENTAL DESIGN

Our goal for our circuit is to elicit a divergent community response from two genetically similar cells by varying different communication parameters. To measure distinct community responses in our circuit, we will look at INDIVIDUAL (GFP) response versus COMMUNITY (RFP) response.


Individual GFP Community RFP

We quantify our readouts by inducing our genetic circuit with varying concentrations of doxycycline, our stimulus, and measuring the fluorescent readouts with flow cytometry at 0, 1.5, 3, and 5 hour timepoints.

Doxycycline

BASIC CIRCUIT TESTING

To test the functionality of our basic circuit, we characterized and validated the fluorescent readouts for individual and community response to ensure they would provide accurate and robust measurements. To do so, we ran three different experiments:

  1. Characterization of our Community Response Readout
  2. Validation of LexA Transcription Factor Feedback Sensitivity
  3. Stimulus Induction of Basic Circuit with Altered Communication Parameters

1. Characterization of Community Response Readout:


In our circuit design, our community response is read out by RFP fused with a LexA transcription factor chimera. This readout is regulated by an alpha factor responsive promoter (αFRP), either pFig2c or pAga1. It is vital that these αFRPs have a high dynamic range to distinguish between different alpha factor concentrations and sense a difference between cell populations (i.e. activated vs. unactivated).

aFRP Comparison

Our team characterized pFig2c (BBa_K1829002) and pAga1 (BBa_K1829005) with an alpha factor dose response to measure sensitivity to the community signal. Our testing supports that pAga1 is a better αFRP to use to measure community response in our circuit due to greater dynamic range and a higher expression peak.


pAga1 + pFig2c Dose Curves
Results: Above is our alpha factor dose-response characterizing pFig2c (left) and pAga1 (right). These experiments were done in triplicate with CB008DB yeast cells after two hours induction with varying concentrations of purified alpha factor. We used a CB008DB strain without any fluorescent reporters as the negative control. pFig2c has a 6.8-fold induction of RFP (normalized) at maximum alpha factor concentration (1000 nM) and a dynamic curve for different alpha concentrations. pAga1 has a 14-fold induction of GFP (normalized) at maximum alpha factor concentration (1000 nM) and a dynamic curve for different alpha concentrations.


2. Validation of LexA Transcription Factor Feedback Sensitivity:


Our project this year is aimed toward changing communication parameters to elicit divergent behavior. We’ve engineered these communication parameters into our yeast under the regulation of a LexA transcription factor inducible promoter, LexAOps. This transcription factor is a synthetic fusion of the DNA-binding domain of the repressor LexA with transcription activation domains from VP64 (BBa_J176013). We control production of the LexA-TF with an alpha factor inducible promoter, and when present it binds to the LexA operator sequence to induce transcription of downstream genes. Thus, when our cells sense alpha factor, they will amplify the signal to produce some sort of positive feedback, such as increased secretion or increased reception. (See Circuit design for more detail). Because these feedback loops are essential to our desired phenotype, we constructed a strain to measure the sensitivity of our feedback loops to alpha factor concentrations.

aFRP + LexA GFP

We induced our yeast strain with varying concentrations of purified alpha factor and recorded the cells' level of florescence after two hours. Our experiments verified that our feedback loops were sensitive to alpha factor.


LexA Dose Response
Results: Above is our alpha factor dose response validating our feedback sensitivity to alpha factor. These data are readouts of different interlinked constructs in the same cell. These experiments were done in triplicate with CB008DB yeast cells after two hours induction with varying concentrations of purified alpha factor. We used a CB008DB strain without any fluorescent reporters as the negative control. RFP, fused to our LexA-TF, (left) is measuring the sensitivity of our αFRP (pFig2c) to alpha factor while GFP (right) is measuring the sensitivity of LexAOps to LexA-TF produced by alpha stimulation. Alpha factor concentration is shown in increasing concentration as the color of the peak turns darker. With greater alpha factor concentrations, the RFP peak shifts right, indicating an increasing expression of RFP and LexA transcription factor. This induces the LexAOps promoter driving our positive feedback motifs. On the right we also observe a shift to the right in GFP expression with high concentrations of alpha factor.


3. Stimulus Induction of Basic Circuit with Altered Communication Parameters:


Now that we have validation that our basic circuit is functional and that the genetic components fulfill their intended purposes, we are ready to begin testing our strains for divergent community phenotypes. We tested our “basic” circuit, containing only the reporters and communication signal in comparison to a Positive Feedback circuit, which produces both mating factor alpha and the receptor Ste2 upon sensing of alpha factor. To test these circuits, we induce our genetic circuit with our stimulus, doxycycline, and measure individual and community readouts with flow cytometry.

Feedback Circuit

Our experiments show genetically identical cells with high individual responses participating in bimodal activation when utilizing a positive feedback loop.We have achieved our goal of creating a population of genetically identical cells that diverge into phenotypically distinct cells.


LexA Dose Response
Results: Above is our doxycycline induction of our genetic circuit at 3 hours. These data are from a high doxycycline responsive cell (high rtTA expressing [pTEF1_m10-rtTA]) with positive feedback for secretion (mFalpha) and receptors (Ste2). We used a CB008DB strain without any fluorescent reporters as the negative control.The light grey peak is our CB008DB negative control, the red is our basic circuit, and the blue is our basic circuit with positive feedback constructs. As you can see in the difference between our “blue” and “red” peaks, positive feedback allows our cellular populations to participate in bimodal activation. The left peak is the “OFF” state, while the right peak is the “ON” state. Note the polarized gap between the states compared to the single convergent red peak from only the basic circuit.


Although we achieved our goal of creating divergent populations from genetically identical cells, we ran into a major problem in our circuit: leaky expression. Because we are working with Bar1 knockout yeast strains, our circuit is hypersensitive to alpha factor signal. So the slightest reception of signal will trigger positive feedback and propagate activation. We see leaky expression in our circuit that activates our circuit at 0 hours, most likely in the pTET system (which drives production of GFP and alpha factor signal). We can also see high basal levels of expression in GFP plots as well.

Leaky Graphs
Results: Above is our data for a strain that has the basic circuit with high expression of rtTA and positive feedback for increased secretion and increased reception. We observe divergent cell populations at 5 hours post induction with doxycycline. However, we see this behavior at all our time points (0-8 hours). The strength of induction increases with time and shifts more cells from the “OFF” state to the “ON” state. From left to right, the dox concentrations are 0 ug/ml, .06 ul.ml, .09 ug/ml, and 6 ug/ml. The left-most peak is our CB008DB negative control and uninduced strain. At each concentration and time point, we observe two distinct peaks: one on the left (“OFF”) and one on the right (“ON”). This displays our goal of bimodal activation in yeast cells, where some cells are activated while the rest remain quiescent. However, our data reveals some leakiness in our circuit. At 0 hours or uninduced time points, we observe a shift in GFP production and some alpha factor activation of cells.

CREATING AN OPTIMAL CIRCUIT

In order to improve the functionality of our circuit, we attempted to control the amount of alpha factor being sensed and secreted through two means:

  1. Signal Degradation
  2. Spatial Clustering

1. Signal Degradation:


We believe that implementing Bar1, a protease that degrades alpha factor, will work in reducing leakiness. We have constructed motifs that express Bar1 under the pTEF1 constitutive promoter and its four mutants characterized by UCSF iGEM 2014.

Feedback Circuit

Thus, we have an array of signal degradation levels which we can implement into our circuit. Previous work has shown Bar1 as an integral mechanism of yeast to desensitize them to alpha factor and sharpen concentration gradients while mating [3]. We hypothesize that Bar1 will create a threshold for activation by degrading background alpha factor secreted into the media. Therefore, only cells with strong enough positive feedback will be able to sense alpha factor and activate a community response. To test this, we measured the expression of an alpha factor responsive promoter under conditions without Bar1 and with Bar1. This experiment verified Bar1’s ability to desensitize yeast to alpha factor and polarize the gaps between “ON” and “OFF.”

Bar1 Decrease
Results: Above is a histogram depicting Bar1 desensitizing yeast cells to the community signal, alpha factor. We tested this by building a construct with an alpha factor responsive promoter (pAga1) driving GFP in a Bar1 knockout strain. We know from literature that using αFRPs as a proxy for community response is reasonable due to the strong correlation between mating and and these mating regulated genes [3]. These experiments were done with CB008DB yeast cells after two hours induction with varying levels of purified alpha factor. We used a CB008DB strain without any fluorescent reporters as the negative control. Without Bar1 (blue), we see maximal GFP expression. By adding Bar1 (red), with high expression from a separate strain, the cells express lower GFP (shifted left) due to alpha factor being degraded faster than it can be sensed.


However, when testing this motif in our circuit, we discovered that Bar1 was too strong at all levels (low to high) for all yeast cells in the population to overcome. This meant that our lowest expressing Bar1 construct was still too effective at degrading alpha factor to allow any cell in the population to activate. For a more feasible activation threshold, we would need to regulate Bar1 with a weaker expressing promoter.

Bar1 Pos Feed
Results: Above is data showing a basic circuit with positive feedback increasing secretion and reception after 5 hours post induction with 6 ug/ml of doxycycline.The one in blue is divergent, without any expression of Bar1. The one in yellow is convergent, with high Bar1 expression. We believe that the single peak is due to Bar1 being too efficient and degrading all the alpha factor, keeping all cells in the population from activating. However, how does this explain why the “OFF” peak for the basic circuit is more left-shifted than the “OFF” peak for the Bar1 strain? We hypothesize that leaky expression may play a role in this, as RFP is stable and will not degrade after it has been made. Also, since the mating pathway is an intricate signaling cascade comprised of multiple intermediate genes, continuous transcription may be a result of sensing alpha factor. Although we do not have a good explanation for this, we believe that this artifact may be due to biochemical mechanisms in the cascade necessary for yeast mating.

2. Spatial Clustering:


We believe that spatial retention of cells will stabilize the divergence we see in our population. We have constructed motifs that express a fusion yeast surface display protein, Mgfp5.

Cluster Circuit

This protein is endogenously found in mussel feet to secrete a bioadhesive that allows them to stick to rocks. We have engineered this protein to be displayed on the cell surface of yeast, and only expressed by activated cells, through a similar positive feedback motif as Ste2 and mFα. Activated cells will cluster and form a smaller, exclusive community and increase their own activation while not engaging with unactivated cells (who remain silent). Similar models are utilized by T-Cells through lymph nodes [9], and V. fischeri through symbiotic organs in squids [12] in order to communicate with their community more efficiently. Once in a cluster, cells will amplify a local concentration of alpha factor and participate in asocial, exclusive behavior [3]. This motif should lend to a more stable, replicable, and efficient method of eliciting bimodal activation.

We have constructed and implemented this clustering motif (LexAOps-Mgfp5) into our basic circuit with positive feedback. We have done some preliminary qualitiative testing, but hope to proceed with quantitative testing of cluster size with software such as CellProfiler or ImageJ in the near future. With visual analysis using light microscopy, we observed large clusters after induction with alpha factor.


Clustering Image
Results: Above is a picture of our clustering circuit (shown above) visualized with 40x magnification on a light microscope. After induction with 1000 nM alpha factor for two hours, we observed large cell clusters not found in wild type strains (on the left). Highlighted in red are the gene-activated cell clusters from Mgfp5 expressing cells (right).

FUTURE DIRECTION

Future Direction

Although we have achieved our goal of eliciting bimodal activation in our yeast cell populations, we hope to optimize this system to create a more stable and robust divergent response. The obvious next step is to work to tune the Bar1 level to both reduce background/set a threshold and allow for activation. We characterized the activity of the endogenous promoter for Bar1 (pBar1, BBa_K1829001), and found that it exhibits little to no activation when induced with alpha factor. Based on the overwhelming activity of Bar1 in our circuit, we suspect that pBar1 has low activity in order to prevent production of too much Bar1. Perhaps with less Bar1 we can create more robust divergent populations in our community. Alternatively, increasing the strength of positive feedback -- either by spatially promoting it through clustering or altering the components of the feedback portion to express at higher levels -- could help overcome Bar1 degradation to induce activation of the circuit.

In addition to divergent behaviors, we still hope to model more complex community phenotypes. Other communication parameters we could look at could be the temporal dynamics of activation and paracrine vs. autocrine signaling. Temporal dynamics are crucial to look at to understand the efficiency and timescale of activation in cellular populations. For example, time is an important aspect in T cell proliferation as extended activation can result in autoimmune disorders. So populations must somehow be able to tune temporal scales and scale back activation after the response is no longer necessary. As a necessary step in looking in dynamics, we may want to alter the system to optimize our reporters. Because RFP is a stable protein, expression of RFP will degrade slowly and stay in the system, offering remnant readouts from previous time points. By creating a less stable community response readout, we can eliminate this artifact in our data. This time-dependent community response will then provide new insights into how activation is started, maintained, and degraded over time.

Ultimately, we hope that by building a model system for studying how cell-to-cell communication can help process responses of a community of cells to a stimulus, we might learn more about how to perturb or harness natural systems to produce new synthetic biology systems, tools, or therapeutics.