Engineering a GPCR-based biosensor in yeast to measure hormones in saliva

With our project, we attempt to revolutionize hormone-based diagnostics as we know them today. We aim to make a chewing gum containing a GPCR-based yeast biosensor, which will measure the sex hormones Estrogen and Luteinizing Hormones (LH) in saliva. The surge of these hormones indicates the ovulation event of the menstrual cycle. The detection of this event will tell women when they are fertile and has the best chance of conceiving. For more details on the hormones of the menstrual cycle, please refer to our Project Description page.

GPCRs in humans and yeast

G-protein-coupled receptors (GPCRs) are the primary extracellular sensing receptors in higher eukaryotes, with more than 800 receptor types detecting a multitude of structurally diverse agonists in humans ¹. Saccharomyces cerevisiae also has endogenous GPCR signaling pathways ², of which the conserved pheromone response pathway is the most well studied ³. This pathway employs elements of the mitogen-activated protein kinase (MAPK) pathways ⁴. Please see Figure 1B for a schematic representation of elements of the pheromone signaling pathway. Thus, we hypothesize that we can hijack this endogenous pathway by transforming an exogenous GPCR into a refactored version of the pathway. This makes yeast an ideal chassis for GPCR-based biosensor construction.

Refactoring the endogenous yeast pheromone signaling pathway

Constructing a biosensor by hijacking the yeast pheromone signaling pathway has previously been described in research. In 1995, researchers from Cyanamid Agricultural Research Center, Princeton, New Jersey, described how a mammalian GPCR, the rat somatostatin receptor, could be successfully coupled to the yeast pheromone signaling pathway ⁵. In the lab of our Primary investigator, Sotirios Kampranis, researchers are currently working with a laboratory yeast strain, ΔKM111, with a refactored GPCR pheromone signaling pathway. The ΔKM111 strain carries 5 mutations (Fig. 1C), ΔSTE3-0 (knocking out the endogenous receptor that we want to replace with an exogenous GPCR), ΔGPA1-0 (knocking out the adaptor Ga-protein, so we can replace it with a chimera to connect our new receptor to the signaling pathway), ΔSTE12-0 (knocking out the endogenous transcription factor, so we can replace it with a STE12 under the control of optimal promoter) ΔSST2-0 (removing the negative regulator of the pathway), and lastly ΔFAR1-0 (preventing unwanted cell-cycle arrest).

Detecting hormones in saliva

To establish whether the salivary concentrations of Estrogen and LH reflect the serum concentration fluctuations, we investigated the literature. Here, it was evident that salivary concentrations of estrogen and LH can be used for assessment of menstrual cycle phase ^{20, 21} as the fluctuations in the salivary levels of steroid hormones show positive statistical correlations with the changes in the body. ²¹(Table 1).

	Concentration	Reference
Saliva LH concentration (basal level)	17.2 pmol/L	Saibaba et al. (2017)⁹
Saliva LH concentration (peak level)	512.31 pmol/L	Ersyari et al. (2014)¹⁰
Saliva Estrogen concentration (basal level)	20.6 pmol/L	Chiappin et al. (2016)¹¹
Saliva Estrogen concentration (peak level)	40 pmol/L	Chiappin et al. (2016)¹¹

To further investigate the feasibility of using a yeast-based biosensor to detect salivary hormones, we simulated our Ovulaid biosensor system. Through the deterministic model based on numerical integration of 20 ODEs, we can establish that:
1) The salivary concentrations (Table 1) of hormones are sufficient to generate a significant ligand-induced response for the investigated receptors.
2) The functional biosensor will be able to distinguish between the base and peak level concentrations of the menstrual cycle.
3) The final color production is correlated to the decay rate of yeast and the concentration of the substrate.
For more information on our model, please refer to our Modeling page.

Biosensor construction

In order to detect the two hormones, Estrogen and LH, we chose to use G protein-coupled Estrogen Receptor (GPER), Homo sapiens Luteinizing Hormone/Choriogonadotropin receptor (Hu-LHCGR), to detect estrogen and LH respectively. We also chose to express the Xenopus laevis Lutropin-Choriogonadotropic Hormone Receptor (XLHCGR) as it has been shown to react to human LH ⁷. In the end, we were not able to successfully clone the Hu-LHCGR and therefore chose to move on with the XLHCGR.

Using the ΔKM111 strain already available in the lab, we engineered new yeast strains, containing the following new modules: 1) Sensing module: Receptor for our hormone of interest (GPER or XLHCGR) expressed under the strong constitutive promoter pCCW12, which has been shown to work well with this particular chassis. 2) Adaptor module: A modified GPA1-Gα-protein, which is a chimera of yeast endogenous Gα-protein and the human Gα-protein, which will connect the human receptor to the conserved endogenous yeast Gβγ-proteins. Two different types of this chimeric proteins were tested, GPA1-Gαi and GPA1-Gαs. Both were expressed under the pPGK1 promoter, which is a strong constitutive promoter. 3) Signalling module: The STE12 transcription factor, expressed under the constitutive pRET2 promoter. The STE12 expressed is located in a different locus than the endogenous one deleted by mutation ΔSTE12-0, and the combination of this STE12 with pRET2 was previously optimized by the researchers of the Kampranis lab. 4) Reporter module: The reporter gene ZsGreen is a constitutively fluorescent green fluorescent protein. This was chosen as it has previously been described for human GPCR signaling in yeast cells ⁸. ZsGreen was expressed under the inducible promoter pFIG1, which is under the control of STE12 transcription factor. See Figure 1D for a schematic representation of the modules we engineered to create our yeast strains.

Figure 1: Schematic representations of the endogenous yeast pheromone signaling pathway, the ΔKM111 chassis strain, and the refactored yeast strain with engineered modules | 1A: The minimalistic modular design of a biosensor. 1B: Overview of the endogenous yeast pheromone signaling pathway 1C: The ΔKM111 strain, which contains the following mutations: ΔSTE3-0 (knocking out the endogenous receptor that was replaced with an exogenous GPCR), ΔGPA1-0 (knocking out the adaptor Ga-protein, so we can replace it to connect our new receptor to the signaling pathway), ΔSTE12-0 (knocking out the endogenous transcription factor, so we can replace it with a STE12 with an optimized promoter) ΔSST2-0 (removing the negative regulator of the pathway), and lastly ΔFAR1-0 (preventing unwanted cell-cycle arrest). Transparent parts represent knocked out endogenous genes. 1D: The design of a minimal biosensor with modules engineered into our chassis yeast strain. The wrenches annotate the engineered modules, while the padlocks annotate the modules conserved from the ΔKM111 strain.

The multiplex integration cassette

To ensure stable expression of all the engineered modules in our yeast strain, we decided to integrate the modules into the yeast genome. To do this, we employed a multiplex genomic integration system. In the Kampranis lab, cassettes of both 3 and 5 modules were available for us to use. The cassettes consist of modified backbones, which allow integration into the yeast genome 10, site 3. The backbones all carry NotI digestion sites, used for digestion of the plasmids. An overview of the backbones can be seen in Figure 2.

Figure 2: The plasmid backbones used in the 3 and 5-assembler system, respectively | HR1 and HR2: Downstream and upstream homologous recombination site for integration in yeast chromosome 10, site 3. URA: Uracil selection marker. AMP: Ampicillin marker. The X3A and X3C vectors are the same in both systems, while the Ass2 vector employed in the 3-assembler system is split into three different vectors, Ass2A, Ass2B, and Ass2C in the 5-assembler system. All backbones carry NotI digestion sites.

Vector construction

Figure3: Throughout the project we used the same general workflow to create our yeast strains: PCR amplification → USER ligation → E. coli cloning → E. coli colony PCR → Sequencing → Plasmid purification → Yeast transformation → Yeast colony PCR

Design

For construction of all our yeast strains we used the same general workflow described in Figure 3. Initially all fragments were PCR amplified using primers carrying USER overhangs. The fragments were verified through gel electrophoresis. Following PCR amplification, the fragments, were purified and ligated into our integration vectors using USER cloning. See Table 2 for list of constructs.

Table 2: The constructs made by USER ligation

	Backbone	Promoter	Gene fragment
1	X3A	pCCW12	GPER
2	X3A	pCCW12	XLHCGR
3	X3A	pCCW12	GPER-linker + sfGFP
4	X3A	pCCW12	XLHCGR-linker + sfGFP
5	Ass2A	pPGK1	GPA1-Gαi
6	Ass2A	pPGK1	GPA1-Gαs
7	Ass2B	pRET2	STE12
8	X3C	pFIG1	ZsGreen

Engineered yeast strains

Using the multiplex integration system described above, we created 2 yeast strains (OV4 and OV5) expressing our full biosensor system with GPER and XLHCGR, respectively (Fig. 4A). We also expressed both strains with both GPA1-Gαi and GPA1-Gαs, however, we were not able to express a positive strain with Gαi, and we therefore moved on with Gαs.

In order to confirm the correct expression and localization of our receptors, we created 2 additional yeast strains (OV2 and OV3) in which we conjugated the GPER and XLHCGR to superfolded GFP (sfGFP) in the C-terminal end (Fig. 4B). We later performed western blot assays and confocal microscopy with these strains to verify the expression and localization of the receptors of strain OV4 and OV5.

Figure 4: The multiplex genomic integration system | 4A: The 5-assembler genomic integration cassette. The first plasmid (X3A) contains the receptor (GPER/XLHCGR). Second plasmid (Ass2A) contains the GPA1- Gαi/Gαs. Third plasmid (Ass2B) contains the transcription factor STE12. Fourth plasmid (Ass2C) contains an empty vector backbone and the fifth plasmid (X3A) contains the reporter gene ZsGreen. 4B: The 3-assembler genomic integration cassette. The first plasmid (X3A) contains the receptor tagged with sfGFP. The second and the third plasmids (Ass2 and X3C) contain empty backbone vectors.
Furthermore, we also transformed a strain with empty vectors to use this as negative control. See Figure 5 for a full overview of the yeast strains created for the biosensor assays.

Figure 5: Overview of the 5 different cassettes transformed into yeast | OV1 is a negative control strain transformed with empty vector backbones. OV2 and OV3 were transformed with our 2 receptors conjugated to sfGFP in module 1, and two empty vectors in module 2 and 3. OV4 and OV5 were transformed with our receptors in module 1, the adaptor GPA1 in module 2, and the transcription factor STE12 in module 3. Module 4 was kept empty in these strains, while module 5 contains the reporter ZsGreen.

For verification of successful integration, we employed PCR. Here, we used 3 specific yeast genotyping primers, one forward and one reverse primer for the homologous recombination sites in the yeast chromosome 10, and one for one of the backbones in the integration system (Fig. 6). In the presence of our construct, we expect to see a band at 1000 bp as, which is the size of the fragment between the forward primer of the backbone, and the reverse primer for the yeast chromosome. In the absence of the constructs, we expect to see the bands at 1500 bp, as this is the size of site 3 of chromosome 10; the fragment amplified by the two yeast chromosome primers.

Figure 6: Expected yeast colony PCR band sizes | Overview of the expected band sizes from yeast transformed with the integration cassette vs. the size of site 3 of chromosome 10.

Biosensor design considerations

We foresaw certain shortcomings with the incorrect post-translational modifications of the HuLHCGR in S. cerevisiae, leading to inactive protein formation. In vitro studies on the LHCGR showed that the sulfated tyrosine 331 (sTyr331) is essential for HuLHCGR ligand binding, but less important for hCG function ¹⁴. We therefore cloned the coding sequence of human tyrosylprotein sulfotransferase-1 enzyme isoforms TPST1 and TPST2 that catalyze post-translational tyrosine sulfation of proteins to transform yeast strain expressing LHCGR receptor ¹⁵. Unfortunately, we were not able to successfully express the HuLHCGR or carry out the experiments with TPST within the timeframe of the iGEM competition.

Nuclear receptor design

To further investigate the potential of a functional yeast estrogen biosensor, we ordered the Plasmid #69100 (pRS416-yZ3EV-Z3pr-yEGFP (RB3579)) from the Addgene catalog. The plasmid was a gift from Dr. David Botstein. It has never been published in a scientific journal, but parts of it were used by Pothoulakis and Ellis (2018)¹⁶.

The construct we used us a nuclear estrogen receptor. The vector carries the ACT1 promoter controlling the expression of Z3EV. This gene encodes an artificial transcription factor that consists of three parts, namely the ligand binding domain of the human estrogen receptor, viral protein 16, and the Zif268 DNA binding domain ¹⁷. The plasmid further contains the Z3 promoter driving the expression of yEGFP. Human estrogen receptor can recognize a specific DNA sequence in the regulatory regions of genes, the estrogen regulatory element. When bound to estrogen, this element leads to transcription activation ¹⁹.

Biosafety device

We are aware that the introduction of a diagnostic tool containing a genetically modified organisms (GMO) would likely be met with certain skepticism, especially in a potential European market. Wanting to confront the challenges that arise from creating an organism that meets safety standards, we expanded the existing research on the Bax/BI-1 system, i.e. a human toxin-antitoxin-based kill switch compatible with S. cerevisiae. This biosafety device was previously described in the iGEM competition by Team NAU China 2017.

The basic principle is to express the toxin Bax under the regulation of a constitutive promoter and the antitoxin BI-1 under the regulation of an inducible promoter. Hence, the antitoxin is only able to rescue the yeast in the presence of an inducer molecule. We exploit this system to make sure that the yeast would survive as long as it is in the environment of our diagnostic tool and will cease to survive due to apoptosis in case of escape.

Figure 7: Overview of the BAX/BI-1 Kill switch device | The figure depicts the genes encoding for the toxin BAX under the expression of a constitutive promoter while the anti-toxin BI-I under the expression of an inducible promoter pGAL1. In the absence of galactose, BI-1 will not be expressed, and the toxin will be allowed to kill the yeast cell.

We imagined the biosensor and the kill switch modules to be incorporated into the same yeast strain and therefore chose a different site for the genomic integration for the kill switch sequences, namely chromosome XI site 2, using a 3-plasmid assembler system.
Furthermore, we also planned to test the system when expressed in high copy plasmids, as opposed to when integrated into the yeast genome. In order to do so, we employed two different plasmids, pUUS and pWUS, with selection markers for Uracil and Tryptophan, respectively (Fig. 8).

Figure 8: Simplified overview of the pUUS and pWUS vectors used for dual plasmid transformation | pUUS carries a selectiokn marker for Uracil, pWUS carries a selection marker for Tryptophan. Both plasmids carry Ampicillin selection marker.

References

1. Fredriksson, R., Lagerström, M., Lundin, L. and Schiöth, H. (2003). The G-Protein-Coupled Receptors in the Human Genome Form Five Main Families. Phylogenetic Analysis, Paralogon Groups, and Fingerprints. Molecular Pharmacology, 63(6), pp.1256-1272.
2. Versele, M., Lemaire, K., & Thevelein, J. M. (2001). Sex and sugar in yeast: two distinct GPCR systems. EMBO reports, 2(7), 574-579.
3. Bardwell, L. (2004). A walk-through of the yeast mating pheromone response pathway. Peptides, 25(9), pp.1465-1476.
4. Drogen, F.V., Stucke, V.M., Jorritsma, G., Peter, M., 2001. MAP kinase dynamics in response to pheromones in budding yeast. Nature Cell Biology 3, 1051–1059. doi:10.1038/ncb1201-1051
5. Price, L.A., Kajkowski, E.M., Hadcock, J.R., Ozenberger, B.A., Pausch, M.H., 1995. Functional coupling of a mammalian somatostatin receptor to the yeast pheromone response pathway. Molecular and Cellular Biology 15, 6188–6195. doi:10.1128/mcb.15.11.6188
6. Lu, Y. (1999). Salivary estradiol and progesterone levels in conception and nonconception cycles in women: evaluation of a new assay for salivary estradiol. Fertility and Sterility, 71(5), pp.863-868
7. Wlizla, M., Falco, R., Peshkin, L., Parlow, A. F., & Horb, M. E. (2017). Luteinizing Hormone is an effective replacement for hCG to induce ovulation in Xenopus. Developmental biology, 426(2), 442-448.
8. Nakamura, Y., Ishii, J., Kondo, A., 2013. Bright Fluorescence Monitoring System Utilizing Zoanthus sp. Green Fluorescent Protein (ZsGreen) for Human G-Protein-Coupled Receptor Signaling in Microbial Yeast Cells. PLoS ONE 8. doi:10.1371/journal.pone.0082237
9. Saibaba, G., Srinivasan, M., Priya Aarthy, A., Silambarasan, V. and Archunan, G. (2017). Ultrastructural and physico-chemical characterization of saliva during menstrual cycle in perspective of ovulation in human. Drug Discoveries & Therapeutics, 11(2), pp.91-97.
10. Ersyari, R. M., Wihardja, R., & Dardjan, M. (2014). Determination of ovulation in women using saliva ferning test. Padjadjaran Journal of Dentistry, 26(3)
11. Chiappin, S., Antonelli, G., Gatti, R. and De Palo, E. (2007). Saliva specimen: A new laboratory tool for diagnostic and basic investigation. Clinica Chimica Acta, 383(1-2), pp.30-40.
12. Keutmann, H.T. et al., Structure of Human Luteinizing Hormone Alpha Subunit. Endocr. Res. Commun., 5(1), 57-70 (1978).
13. Keutmann, H.T. et al., Structure of Human Luteinizing Hormone Beta Subunit: Evidence For a Related Carboxyl-terminal Sequence Among Certain Peptide Hormones. Biochem. Biophys. Res. Commun., 90(3), 842-848 (1979).
14. Grzesik, P., Kreuchwig, A., Rutz, C., Furkert, J., Wiesner, B., Schuelein, R., Kleinau, G., Gromoll, J. and Krause, G. (2015). Differences in Signal Activation by LH and hCG are Mediated by the LH/CG Receptor’s Extracellular Hinge Region. Frontiers in Endocrinology, 6.
15. Tanaka, S., Nishiyori, T., Kojo, H., Otsubo, R., Tsuruta, M., Kurogi, K., Liu, M., Suiko, M., Sakakibara, Y. and Kakuta, Y. (2017). Structural basis for the broad substrate specificity of the human tyrosylprotein sulfotransferase-1. Scientific Reports, 7(1).
16. Pothoulakis, G. and Ellis, T. (2018). Synthetic gene regulation for independent external induction of the Saccharomyces cerevisiae pseudohyphal growth phenotype. Communications biology 1, pp. 1-11.
17. McIsaac, R. S., Gibney, P. A., Chandran, S. S., Benjamin, K. R.; Botstein, D. (2014). Synthetic biology tools for programming gene expression without nutritional perturbations in Saccharomyces cerevisiae. Nucleic acids research 42 (6), e48.
18. pRS416-yZ3EV-Z3pr-yEGFP (RB3579) was a gift from David Botstein (Addgene plasmid # 69100 ; http://n2t.net/addgene:69100 ; RRID:Addgene_69100)
19. Waterman, M. L., Adler, S.; Nelson, C., Greene, G. L., Evans, R. M., Rosenfeld, M. G. (1988). A single domain of the estrogen receptor confers deoxyribonucleic acid binding and transcriptional activation of the rat prolactin gene. Molecular endocrinology (Baltimore, Md.) 2 (1), pp. 14–21.
20. Lu, Y. (1999). Salivary estradiol and progesterone levels in conception and nonconception cycles in women: evaluation of a new assay for salivary estradiol. Fertility and Sterility, 71(5), pp.863-868.
21. Stanescu, I. (2016). Saliva as a Monitoring Fluid for Hormonal Activity in Systemic Lupus Erythematosus. Rheumatology, pp.654-659.