For the experimental part of our project, we aimed to develop in yeast a high-throughput method to engineer proteases with distinct specificity and later to use the method to obtain a protease that would specifically cleave the SALSA protein.
We constructed a synthetic transcription factor where the DNA binding and activation domains are separated by an unstructured linker that contains a potential protease cleavage site. The synthetic transcription factor activates the expression of sic1ΔN to inhibit yeast growth. However, when a protease cleaves the linker sequence, the two domains of the transcription factor are separated, the inhibitor is not expressed, and the cells grow normally.
Overexpression of Sic1ΔN arrests yeast growth
We first set out to test the inhibition of yeast growth by inducing the expression of sic1ΔN.. Sic1 is an inhibitor of the master cell cycle regulator, Cdk1, that is targeted for degradation via phosphorylation of its N terminus, enabling cells to enter S phase (Kõivomägi et al., 2011). Previous studies have found that mutation of the phosphorylation sites in Sic1 leads to stabilization of Sic1 and a delay in cell cycle progression (Cross et al., 2007) (Figure 1).
We first tested whether Sic1, where the degradation motifs in the N terminus have been removed (Sic1ΔN), is capable of triggering efficient cell cycle arrest when overexpressed. For this, we cloned the coding sequence for Sic1 residues 90-284 to an expression vector with the LexA-dependent promoter. This expression cassette was transformed to a yeast strain expressing an estradiol-regulated synthetic transcription factor that regulates the LexA promoter (Ottoz et al., 2014).
To characterize the effect of sic1ΔN expression on growth rate, we need to activate the LexA promoter at different levels using a range of estradiol concentrations (the work is underway). Using sic1ΔN as a reporter gene allows to translate the activity of the transcription factor to growth rate. Importantly, if the growth rate is dependent on estradiol concentration in this assay, this indicates that the setup can also be used to monitor the level of the transcription factor, enabling us to use it as a quantitative readout for the protease activity towards the transcription factor.
Figure 1. Dosing of transcription factor activity to growth rate via expression of Sic1ΔN. (A) The scheme of Sic1, showing the phosphorylation sites in the N terminus, necessary for Sic1 degradation, and the C-terminal part used in Sic1ΔN that inhibits Cdk1. (B) Induced overexpression of sic1ΔN. leads to dosage-dependent suppression of yeast growth.
Our protease engineering method potentially allows simultaneous screening of the proteases’ activity towards the target sequence in the LexA-linker-VP16 transcription factor and its activity towards essential cellular proteins. Overexpression of a protease with high activity and broad specificity is expected to have a negative effect on cell growth due to the cleavage of cellular proteins. This would result in these cells gradually dropping out in the competitive growth assay, with the fastest dividing cells containing a protease that efficiently cleaves the LexA-linker-VP16 transcription factor to suppress sic1ΔN. expression but does not cleave additional targets that would decrease the cell’s fitness.
We first used TEV protease as the model enzyme to test our experimental setup, as it is highly studied and has a distinct cleavage site. To control the expression of the protease, we cloned TEV protease in an expression vector with a GAL1 promoter that is inactive in the glucose-containing medium but highly expressed in the galactose-containing medium.
The TEV expression plasmid was transformed to yeast, and the effect of protease expression on cell viability was studied by comparing the yeast growth in the presence or absence of galactose. The assay showed that overexpression of TEV protease does not cause a growth defect (Figure 2). The slightly faster growth rate on the glucose-containing plate is due yeasts’ preference for glucose over galactose as a carbon source. Due to the high specificity of TEV protease, this result could be expected, but it is a critical step in the high-throughput screening process involving proteases with very different substrate specificities.
Figure 2. Viability assay of the yeast cells growing on the glucose-containing plate (left panel, TEV protease expression is not induced) and raffinose/galactose-containing plate (right panel, TEV protease expression is induced). TU3-TEV strain was pre-grown in three biological replicates (col1, col2, col3) in 1 ml of liquid CSM medium (with 2% glucose) to OD600 0.5. Before plating, cultures were washed three times with CSM to remove glucose. After the last wash and resuspension in 1 ml of CSM medium, cultures were diluted according to the factor on top of the figure. 5 µl of each dilution was plated on the CSM/glucose and CSM/raffinose/galactose plates.
The analysis of the functionality of our transcription factor
The final component of our protease activity screening assay is a synthetic transcription factor that contains the desired sequence for proteolytic targeting between the LexA DNA binding domain and the VP16 activation domain. We first have to verify whether the transcription factor is able to efficiently activate sic1ΔN expression from the LexA-dependent promoter. For this, the transcription factor was cloned to an expression cassette with GAL1 promoter and transformed to yeast that contains the Sic1ΔN reporter cassette.
For comparison, we will use the estradiol-controlled transcription factor as above (Ottoz et al., 2014). This experiment will allow us to verify that the created transcription factor is functional and would indicate that the setup is flexible in terms of the sequence inserted between the transcription factor domains as the protease target sequence (work is underway).
The analysis of the functionality of our transcription factor
As a proof-of-concept for this method, we first want to test it with TEV protease, as it has a very specific cleavage sequence. For this, we use a LexA-linker-VP16 transcription factor, where the TEV cleavage site (ENLYFQS) was introduced to the linker. All three components of the assay (Figure 3) are transformed into yeast, followed by monitoring of yeast growth. The experiment to analyze cell growth after TEV protease induction is ongoing. We expect that the protease will cleave the LexA-linker-VP16 transcription factor, leading to suppression of sic1ΔN expression and allowing yeast growth. The next steps will include validation of the cleavage of the transcription factor by TEV using Western Blot.
Transcriptional units transformed into yeast.
TU1 - Transcriptional Unit 1 consists of sic1ΔN under the control of LexA binding promoter. TU2 - Transcriptional Unit 2 includes LexA binding domain (LexA BD) and activation VP16 domain separated by SRCR-SID linker from SALSA protein (under control of GAL1 promoter). TU3 - transcriptional unit 3 consists of protease fused to ligand facilitating protease binding to SALSA (under control of GAL1 promoter).
The method can next be used with a library of proteases or alternatively a library of linker sequences as a pooled input, allowing for the selection of efficient cleavage pairs in a competitive growth assay in a high-throughput manner. Next, we aim to implement the method using trypsin and prolyl peptidase, as suggested by our modelling results. These proteases will be tested both as unmodified enzymes and also as fusions with the SRCR-binding peptide from S. mutans. The peptide is expected to dock the protease to the transcription factor, increasing the local concentration and driving the specific proteolytic cleavage. Further, we will create a library of protease mutants using error-prone PCR to find proteases with altered active site-specificity to obtain an enzyme that is optimal for targeting the SALSA protein.