Patent Application
20240069018
Structural and detailed biochemical investigation of eukaryotic transmembrane receptors is still hampered despite technical advances in expression, purification and crystallization strategies. Important difficulties are the lacking ability to produce the desired receptors in sufficient amounts and their inherent instability.
A method to increase functional expression levels of GPCRs in Escherichia coli by directed evolution is known in the art (Sarkar et al. (2008), PNAS 105:14808-14813; Dodevski & Plückthun (2011), J Mol Biol 408:599-615). Several high-expression variants of GPCRs have been isolated with this method. However, since E. coli is not able to perform post-translational modifications, lacks the secretory quality control and translocation machinery for eukaryotic membrane proteins, and differs in membrane composition, evolution of eukaryotic transmembrane receptors in E. coli is limited to receptors which inherently express in a prokaryotic system above the threshold required for successful evolution, thus not having strict requirements for eukaryotic processing or membrane composition.
Therefore the establishment of a fast, efficient and robust method to increase functional expression of transmembrane receptors in a eukaryotic host is required. In principle the ideal eukaryotic host would be cells that are used for the production of the evolved high-expression variants. However, these mammalian or insect cells are not easily transformed with libraries, which is essential for any type of directed evolution approach. The usage of “lower” eukaryotes such as the yeast Saccharomyces cerevisiae is common for other procedures involving libraries. However, yeast cells comprise a cell wall, which limits functionality assays by restricting access of administered ligands to expressed receptors located in the plasma membrane. For soluble proteins, a common solution to this problem is the use of methods such as yeast display, which expresses the protein of interest as fusion protein that is presented outside of the cell wall. However, this approach is not suitable for selection of functional high-expression variants of insoluble transmembrane receptors, which must be located in the plasma membrane, and thus below the cell wall, and cannot be expressed as fusion proteins attached to the cell wall.
The problem underlying the present invention is to provide novel efficient methods to facilitate and accelerate research and drug design in transmembrane receptors, particularly by identifying transmembrane protein variants resulting in optimized functional expression in eukaryotic cells comprising a cell wall. This problem is solved by the subject-matter of the independent claims.
In the context of the present specification, the term alkaline pH is used in its meaning known in the art of chemistry; it refers to a measure of the basicity of an aqueous solution. A pH greater than 7 indicates alkaline or basic solutions and a pH less than 7 indicates an acidic solution.
In the context of the present specification, the term reducing agent is used in its meaning known in the art of chemistry and biochemistry; it refers to an element or compound that donates an electron to another chemical substance in a redox reaction. In this process the reducing agent is oxidized.
In the context of the present specification, the terms fluorescent cell sorting or fluorescence-activated cell sorting (FACS) are used in their meaning known in the art of cell biology; they refer to a method for cell sorting, wherein cells are suspended in a stream of fluid and pass a detection device. Every passing cell can be directed into different compartments according to the intensity of a specific fluorescent signal of the passing cell.
In the context of the present specification, the term random mutagenesis is used in its meaning known in the art of cell biology and molecular biology; it refers to a method wherein DNA mutations are randomly introduced to produce mutant genes and proteins. A multitude of these mutant genes can then be compiled into a library. Non-limiting examples for random mutagenesis methods are error-prone PCR, UV radiation and chemical mutagens.
In the context of the present specification, the term library is used in its meaning known in the art of cell biology and molecular biology; it refers to a collection of nucleic acid fragments.
One particular type of library is a randomized mutant library, generated by random mutagenesis. Another example would be a designed (or synthetic) library, which comprises specifically engineered DNA fragments.
In the context of the present specification, the term expression level is used in its meaning known in the art of cell biology and molecular biology; it refers to the level of transcription and/or translation of a DNA fragment and the derived mRNA, respectively. In certain embodiments an expression level is deemed to be high if the expression of a mutant gene is higher than a control gene or wild-type gene. In certain embodiments the expression of the mutant gene is at least two-fold higher than that of the control or wild-type gene to be deemed high.
According to a first aspect of the invention a method is provided for selecting a sequence from a library of expressed nucleic acid sequences, wherein the sequence is selected according to its expression level. The method comprises the following steps.
In other words, the method according to the first aspect of the invention allows the expression of a library of transmembrane receptors in eukaryotic cells comprising a cell wall and allows for the selection of functional expression of these receptors. This is made possible by the permeabilization of the cell wall in a way that still maintains viable and structurally stable cells, which is different from other methods of permeabilization of the cell wall such as the preparation of spheroplasts. The permeabilization procedure of this invention allows even the binding of large ligands to the transmembrane receptors. Selection of the cells by the amount of bound ligand enables selection of the cells according to the amount of functional receptor on the plasma membrane and not by total receptor amount, which would also include non-functional receptors (e.g. receptor present in intracellular membranes).
In certain embodiments the method additionally comprises the following steps:
In certain embodiments the selection step comprises a multitude of selection procedures. After a first selection the selection of cells is used again for the selection of a subset of these cells as a function of detectable label present in these cells. Selection of cells is repeated at least 1, 2, 3, 4 or 5 times.
In certain embodiments the library of expressed nucleic acid sequences is obtained by amplification of a nucleic acid sequence encoding the target membrane protein by a process introducing mutations into the amplified sequence.
In certain embodiments the method additionally comprises the following steps:
In certain embodiments the expressed nucleic acid sequences obtained by the method of the invention are characterized by high expression levels and/or high thermodynamic stability of the encoded target membrane protein.
In certain embodiments according to all aspects of the invention in the selection step e), the selection of viable cells comprise the top 0.1% to 5% of the most fluorescent cells.
In certain embodiments the method additionally comprises the following steps:
In certain embodiments the eukaryotic cells comprising a cell wall are yeast cells.
In certain embodiments the target membrane protein is a G-protein coupled receptor (GPCR).
In certain embodiments the permeabilization step comprises exposing the plurality of eukaryotic cells comprising a cell wall to an enzymatic and/or a chemical treatment.
In certain embodiments the enzymatic treatment in the permeabilization step comprises exposing the plurality of eukaryotic cells comprising a cell wall to enzymes or enzyme mixtures that permeabilize the cell wall. Non-limiting examples of such enzymes are glucanases, proteases, mannases and/or sulfatases. Non-limiting examples of such enzyme mixtures are Zymolyase, Lyticase, and/or Glusulase.
In certain embodiments the chemical treatment in the permeabilization step comprises exposing the plurality of eukaryotic cells comprising a cell wall to a buffer of alkaline pH comprising lithium ions, a reducing agent and/or chelating agent.
In certain embodiments non-limiting examples of buffering agents contained in the buffer used in the permeabilization step are:
In certain embodiments the chelating agent used in the buffer of the permeabilization step is ethylenediaminetetraacetic acid (EDTA).
In certain embodiments the detectable label is a fluorescent dye and the selection step is accomplished by fluorescent cell sorting.
In certain embodiments non-limiting examples of reducing agent are:
In certain embodiments the permeabilization step comprises the following steps:
In certain embodiments TELi buffer comprises 50 mM Tris-HCl pH 9, 1 mM EDTA and 100 mM lithium acetate.
In certain embodiments the buffer used in the permeabilization step comprises:
In certain embodiments the labeling step comprises exposing the yeast cells to TELi buffer comprising the ligand at 4° C.
In certain embodiments the nucleic acid sequences for the library of expressed nucleic acid sequences are generated by random mutagenesis preferably by error-prone PCR.
In certain embodiments the library of expressed nucleic acid sequences are designed (synthetic) libraries.
In certain embodiments the library of expressed nucleic acid sequences comprises homologous sequences of at least 60% sequence identity with each other.
In certain embodiments the eukaryotic cell comprising a cell wall is a yeast cell, particularly Saccharomyces cerevisiae, Pichia pastoris, Kluyveromyces lactis, Candida boidinii, or Hansenula polymorpha.
In certain embodiments the eukaryotic cells comprising a cell wall are a mutant or genetically engineered yeast strain not comprising a native cell wall.
In certain embodiments the eukaryotic cells comprising a cell wall is Saccharomyces cerevisiae, particularly the S. cerevisiae strain BY4741.
In certain embodiments the ligand specifically capable of binding to the target membrane protein is an agonist, antagonist or allosteric modulator.
In certain embodiments the ligand specifically capable of binding to the target membrane protein is an oligopeptide comprised of at least 3, 4, 5, 6, 8, 10, 14, 18 or 25 amino acids.
In other embodiments, the ligand specifically capable of binding to the receptor is an antibody, antibody fragment or another binding protein, e.g., a scaffold protein from the non-limiting list of examples of DARPins, affibodies, anticalins, nanobodies, affilins, fibronectin-derived scaffolds and other scaffolds.
According to an alternative to this first aspect of the invention a method is provided for selecting a sequence from a library of expressed nucleic acid sequences, wherein the sequence is selected according to its expression level. The method comprises the following steps.
In other words, the method according to this alternative aspect of the first aspect of the invention allows the expression of a library of transmembrane receptors in eukaryotic cells comprising a cell wall and allows for the selection of functional expression of these receptors.
Selection of the cells by the amount of bound ligand enables selection of the cells according to the amount of functional receptor on the plasma membrane and not by total receptor amount, which would also include non-functional receptors (e.g. receptor present in intracellular membranes). This alternative aspect of the first aspect of the invention might be advantageous for the use of small ligands.
According to a second aspect of the invention a method for the selection of an adapted yeast cell with the ability for high expression levels of functional membrane proteins is provided.
The method comprises the following steps:
Wherever alternatives for single separable features such as, for example, a cell strain or permeabilization buffers, sorting method or library type are laid out herein as “embodiments”, it is to be understood that such alternatives may be combined freely to form discrete embodiments of the invention disclosed herein.
The invention is further illustrated by the following examples and figures, from which further embodiments and advantages can be drawn. These examples are meant to illustrate the invention but not to limit its scope.
With about 800 different members in the human genome, G protein-coupled receptors (GPCRs) comprise the largest superfamily of cell surface receptors. GPCRs evolved to highly versatile signaling mediators in eukaryotic life, responding to a wide variety of ligands and transducing signals via heterotrimeric G proteins as well as in a G protein-independent fashion. The pivotal role of GPCRs is reflected by the great number of human diseases linked to aberrant GPCR signaling, for instance obesity, diabetes, cardiovascular diseases, osteoporosis, immunological disorders, neurodegenerative diseases, and cancer. As a consequence, GPCRs represent highly relevant drug targets for the pharmaceutical industry. About 30-50% of marketed drugs act on GPCRs or GPCR-associated mechanisms, among them many top-selling drugs.
The major challenges in structural and detailed biochemical investigation of GPCRs are the difficulties to produce the desired receptors in sufficient amounts as well as their inherent instability and flexibility. Advances in expression and crystallization strategies, like the use of the baculovirus/insect cell expression system, or the establishment of integral membrane protein crystallization in lipidic cubic phases (LCP), lead to breakthroughs in determination of three-dimensional receptor structures by X-ray crystallography. While the available structural datasets provided the scientific community with insights towards the mechanisms of GPCR function at the atomic level and might begin to enable rational structure-based drug design, it is clear that detailed understanding of receptor dynamics and conformational changes has still remained rather incomplete.
Furthermore, the 26 unique GPCR structures deposited in the Protein Data Bank (PDB, http://www.pdb.org) to date still represent only a minor fraction of all receptors (<4%), reflecting that the bottlenecks for structural investigations of GPCRs persist. For instance, while insect cells were successfully used for production of recombinant receptors for about 85% of all GPCR structures determined so far (source: PDB; excluding structures obtained from protein extracted from native tissues), this expression system does not provide a generic solution. Even in insect cells several members of the GPCR superfamily are expressed at low yields, suggesting fundamental issues with the biosynthesis and membrane insertion of these receptors also in eukaryotic cells. This problem cannot be solved by simple optimization of expression conditions, especially since there seems to be no consistency of optimal conditions for individual GPCRs. Thus, to facilitate and accelerate GPCR research, novel approaches are required.
Recently, the inventors developed a method to increase functional expression levels of GPCRs in Escherichia coli by directed evolution (Sarkar et al. (2008), Proc Natl Acad Sci USA 105:14808-14813; Dodevski & Plückthun (2011), J Mol Biol 408:599-615). Random mutagenesis of wild-type GPCRs followed by selection with fluorescence-activated cell sorting (FACS) allowed isolation of high-expression variants from randomized libraries of four different GPCRs, namely neurotensin receptor 1 (NTR1), NK-1 receptor (NK1R, also termed tachykinin receptor 1 or substance-P receptor) and of the alpha-1A and alpha-1B adrenergic receptors. In successive studies for NTR1, the obtained first-generation variants built the basis for creation of second-generation mutants by extensive selection (Schlinkmann et al. (2012), Proc Natl Acad Sci USA 109:9810-9815) and exhaustive recombination (Schlinkmann et al. (2012), J Mol Biol 422:414-428). The synthetic libraries created in these studies were also used to generate NTR1 variants with improved stability in short-chain detergents by a complementary directed evolution approach termed CHESS (Scott & PlIckthun (2013), J Mol Biol 425:662-677; Scott et al. (2014), Biochim Biophys Acta 1838:2817-2824). Ultimately, these efforts resulted in the structures of three different variants of agonist-bound NTR1 from material produced in E. coli (Egloff et al. (2014), Proc Natl Acad Sci USA 111:E655-62), underlining the power of directed evolution in membrane protein engineering.
The successful results obtained with evolution in E. coli and the proven effectiveness of eukaryotic expression hosts demanded to further develop this approach towards high functional GPCR expression specifically in eukaryotic expression systems. The inventors hypothesized that it might be advantageous to perform evolution of GPCRs directly in a eukaryotic system in order to achieve specific sequence adaptation and thereby improved functional production in eukaryotes. Furthermore, compared to eukaryotic hosts, E. coli is not able to perform post-translational modifications, lacks the secretory quality control and translocation machinery for eukaryotic membrane proteins, and differs in membrane composition. Therefore, directed evolution of GPCRs in E. coli is limited to receptors which inherently express in a prokaryotic system above the threshold required for successful evolution, thus not having strict requirements for eukaryotic processing or membrane composition.
Here the inventors disclose the establishment of a fast. efficient and robust method to increase functional expression of GPCRs in eukaryotic hosts by directed evolution in the yeast Saccharomyces cerevisiae. The method is generally applicable as demonstrated with three different GPCRs, leading, with only two rounds of evolution, to receptor variants which show high functional expression in both yeast and insect cells. In addition, functional expression levels in yeast can be further increased reproducibly by induced host adaptation.
Directed Evolution of Three Different GPCRs in Yeast Results in High Functional Expression in S. cerevisiae
The general approach of this method is depicted in
The inventors aimed to evolve three different GPCRs in parallel: (i) rat NTR1, (ii) human NK1R, and (iii) human kappa-type opioid receptor (KOR1). NTR1 has been shown to be readily evolvable in the E. coli-based system in several studies, and thus was considered a positive control. NK1 R has been successfully subjected to directed evolution towards higher expression in E. coli as well (Dodevski & Plckthun (2011), J Mol Biol 408:599-615). However, despite strong relative improvements, due to the very low expression levels of wild-type NK1 R in the prokaryote, the evolved receptor variants expressed still at only moderate absolute levels compared to the other receptors evolved in E. coli (Sarkar et al. (2008), Proc Natl Acad Sci USA 105:14808-14813; Dodevski & PlUckthun (2011), J Mol Biol 408:599-615). Moreover, expression levels of these evolved NK1R variants were also low in S. cerevisiae (see below), suggesting that further improvement of functional production might be possible, which may benefit future studies on this receptor. Concerning KOR1, no attempts to evolve this receptor have been undertaken so far and it represents a challenging example regarding heterologous expression.
Following the procedure outlined above, for each of the three receptors only two rounds of evolution—each round consisting of one randomization by error-prone PCR and five subsequent selections by FACS—were sufficient to strongly increase functional expression levels.
Strikingly, the expression levels of the selected libraries are not only much higher than the wild-type receptors but also higher than the corresponding variants evolved in E. coli (NTR1-D03, NK1R-E11), illustrated by FC and RLBA analysis. 100,000-150,000 total functional receptors per cell are measured in RLBA for the selected yeast libraries, representing a 25-50-fold and 5-20-fold increase compared to the corresponding wild-type GPCRs and E. coli-evolved variants, respectively (
Such a division into two subpopulations was reproducibly observed in FC experiments upon expression of all GPCRs tested in both yeast single clones and libraries. Hence, this effect appears to be an intrinsic feature of GPCR expression in S. cerevisiae, in which functional expression at the surface is lacking in a substantial fraction of cells. For instance, depicted by the FC data of the E. coli-evolved variants, active surface expression of NTR1-D03 is obtained in about 50% of all cells, and for the low-expressing variant NKR1-E11 only a minority of cells show active surface expression (
Improved Expression Levels of Enriched Receptor Variants in Yeast are Further Increased by Host Adaptation Occurring Concurrently with Selection.
In order to identify enriched clones, plasmid DNA was isolated from the library pools after the first (stage 1.5 libraries) and second round (stage 2.5 libraries) of evolution for DNA sequencing. A summary of the selected clones depicting the different mutations of each clone is given in
Individual clones were analyzed after retransformation and revealed that all enriched receptor variants were indeed significantly better expressed in yeast than the corresponding wild-type receptors (
While only a small fraction of cells not expressing any functional receptor at the surface was detected in the selected GPCR libraries (
The inventors hypothesized that in the course of selection an adaptation of the expression host occurred which led to a further increased level of membrane protein expression in the selected libraries. To test this hypothesis, a yeast clone, expressing a strongly selected variant termed NTR1-Y06, was directly isolated from the NTR1 2.5 library pool. As expected, the isolated strain expresses NTR1-Y06 at high functional levels, with a similar expression profile as the NTR1 2.5 library pool (
In order to test whether the phenotypic adaptation might even have been the main feature of the selection, the inventors also tried to adapt the strain expressing wild-type NTR1. However, after FACS, the sorted cells did not propagate anymore, preventing repetitive selections. This observation clearly shows that the selection of mutations, conferring improved properties to the receptors, is a strict prerequisite of the host adaptation, and that the sequence change is actually the key evolutionary event. The failure to induce adaptation in cells expressing wild-type NTR1 may be explained by the combined stressful effects of expression of the toxic wild-type GPCR followed by selection, together limiting cell survival.
Hence, adaptation induced by repetitive sorting appears to be restricted to evolved GPCR variants, which intrinsically are less toxic for the cells. Indeed, adaptation of strains expressing NK1R-Y09 and KOR1-Y05, evolved NK1R and KOR1 variants, respectively, was possible as well. Both variants were strongly enriched during selection and newly transformed cells expressing NK1R-Y09 or KOR1-Y05 showed lower expression levels than the corresponding library pools. For each of the variants five rounds of phenotypic selection by FACS, identical to the procedure used to adapt the strain expressing NTR1-Y06, were performed. For both variants host adaptation towards higher expression levels was observed, gradually arising in the course of selection (
The Fraction of Functional Receptor is Increased in Adapted Strains, while the Amount of Total Receptor Produced is not Significantly Higher.
Since host adaptation was most pronounced for expression of NTR1-Y06, this variant was used to further study the effect. While ligand-binding FC analysis and RLBA permit the detection of active receptor, neither of these methods allow quantification of total receptor produced regardless of ligand-binding activity. Therefore, to detect total receptor protein produced in different strains by immunoblotting, NTR1 and NTR1-Y06 expression constructs with a hemagglutinin (HA) tag at the C-terminus of the receptors were created.
The strain expressing the HA-tagged NTR1-Y06 was adapted towards higher expression by five rounds of phenotypic selection with FACS (
In non-adapted strains, the amount of total receptor produced increases from wild-type NTR1 to the evolved variants by a factor of two (
However, when comparing the adapted to the non-adapted strain, based on RLBA data, there is a tenfold increase in functional expression of NTR1-Y06 (
Adaptation Increases Surface Expression of Active Receptor, Reduces the Number of Cells Exclusively Expressing Intracellularly Retained GPCRs, and Reduces the Non-Expressing Cell Subpopulation.
To further describe the host adaptation effect by assessing cellular localization and functionality of NTR1 variants in non-adapted and adapted strains, NTR1 and NTR1-Y06 constructs with a C-terminal mCherry (mCh) fusion were used for confocal microscopy studies and FC experiments: while total receptor is quantified by mCh, only functional receptors located at the cell surface in the plasma membrane will exhibit a signal from fluorescent ligand binding.
Identical to the HA-tagged variant, the strain expressing NTR1-Y06 with a mCh fusion was first adapted by phenotypic selection. Next, the non-adapted strains expressing NTR1 and NTR1-Y06, as well as the adapted strain expressing NTR1-Y06 were exposed to fluorescent ligand to perform co-localization studies with confocal fluorescence microscopy (
As seen in ligand-binding FC experiments (
Evolved NTR1-Y06 expressed in the non-adapted strain shows detectable ligand binding at the surface as well as a mCh signal (
The situation is different for NTR1-Y06 expressed in the adapted strain, in which strong signals for ligand binding and mCh are detected in the plasma membrane (
Further investigations by FC experiments (
In summary, host adaptation of S. cerevisiae for GPCR expression increases surface expression of active receptor without increasing the total amount of receptor produced, which is equivalent to a higher percentage of active receptor produced. In contrast to non-adapted cells, for which a large fraction of cells does not show any functional surface expression, adapted cells show only small amounts of intracellularly retained receptor. Since a significant part of intracellular receptor is inactive, host adaptation leads to an overall increase of active receptor. Additionally, the subpopulation of cells not expressing any receptor at all is also decreased in the adapted strains compared to the non-adapted ones, contributing further to higher average functional production levels.
Evolved Variants of all Three GPCRs Show High Functional Expression Levels in Spodoptera frugiperda (Sf9) Insect Cells.
Ultimately, the inventor's goal was to increase functional GPCR expression not only in yeast but in insect cells as well. Thus, the functional expression levels of the evolved GPCRs NTR1-Y06, NK1R-Y09, and KOR1-Y05 were determined in Sf9 insect cells and compared to the corresponding wild-type receptors. For all GPCRs identical standard expression conditions (non-optimized) were used. Indeed, a significant increase of functional expression of the evolved variants was observed. For NTR1-Y06 compared to wild-type NTR1, functional expression is increased 5-fold (
With such high expression levels, as obtained for all of our three GPCR mutants tested, expression cultures in liter scales are sufficient to produce enough material for structural investigations. To test this, purifications of NK1R-Y09 from insect cell expression cultures were performed, reproducibly yielding 4-6 mg/L pure protein.
Given the very high purification yields, NK1R-Y09 represents an interesting and promising candidate for structural investigations. Thus, the functionality of NK1R-Y09 was characterized by signaling assays based on [35S]-GTPyS binding (
Production of recombinant GPCRs in functional form still remains a demanding, laborious, and cost-intensive task. Given that most of the recently determined three-dimensional GPCR structures were obtained from receptors produced in eukaryotic expression hosts and encouraged by the success of protein engineering methods for increased expression of GPCRs in E. coli, the inventors aimed to transfer the directed evolution approach to S. cerevisiae to specifically improve functional GPCR production in eukaryotes.
The yeast S. cerevisiae, characterized by fast growth, cost-effective cultivation, and ease of genetic manipulations, is an ideal eukaryotic host for protein engineering and directed evolution for several additional reasons: First, yeast surface display has become a very powerful technology in a range of different applications. Second, high-diversity libraries are nowadays easily obtained with yeast. Third, the yeast cell is equipped with the cellular machinery required for functional GPCR production, since S. cerevisiae endogenously expresses two different GPCRs with signaling pathways similar to higher eukaryotes. Fourth, yeast has already been extensively used for heterologous expression of several different GPCRs. And fifth, based on a widely used GPCR assay which couples heterologous GPCRs to the yeast endogenous signaling pathway, directed evolution approaches on GPCRs have been successfully used before with the aim to generate designer receptors exclusively activated by designer drugs (DREADDs) activated by new ligands (Armbruster et al. (2007), Proc NatI Acad Sci USA 104:5163-5168).
So far, functional production yields of GPCRs in yeast have generally remained low (Sarramegna et al. (2003), Cell Mol Life Sci 60:1529-1546). Strategies to increase GPCR expression in S. cerevisiae included optimization of expression conditions, co-expression of molecular chaperones, screening for high-expressing host clones, screening of engineered GPCR variants, or host engineering. In this specification the inventors disclose the general applicability of this invention to directly obtain GPCR variants with increased functional GPCR expression levels by evolving three different GPCRs. Remarkably, only two rounds of evolution—meaning two randomizations each followed by selection with FACS—were required to obtain highly increased expression levels, and thus it was possible to evolve all three GPCRs efficiently in a short time. In the E. co/i-based system, usually four rounds of evolution were required. Furthermore, the inventors disclose that the yeast host cell can be adapted towards increased GPCR production. This is an inherent feature of the selection system, which happens concurrently with the selection of improved receptor variants. It is important to reiterate that the directed evolution did lead to sequence changes, whose improved expression phenotype could be transferred to insect cells, and that the host adaptation is only an additional factor occurring in yeast during the selection. Moreover, the inventors found that expression of an evolved receptor is a prerequisite for inducing the host adaptation, as attempts to induce host adaptation in cells expressing a wild-type GPCR failed. The host adaptation of yeast can reproducibly be induced by repetitive phenotypic selection with FACS. This usually requires 4-5 sorts, and the effect gradually increases during the selection.
Adapted strains show a significant increase of surface-expressed active receptor, leading to a much higher fraction of functional receptors in the plasma membrane, with only a small increase of total receptor produced. On the contrary, in non-adapted strains, and especially for expression of wild-type receptors, a large amount of receptors remain in intracellular compartments, of which a significant amount represents inactive misfolded protein. The overall higher average functional production yields in expression cultures of adapted cells can be explained by a decrease of the subpopulation of cells expressing intracellularly retained inactive receptor as well as a decrease of the non-expressing subpopulation.
As the inventors did not wish to be limited to S. cerevisiae as a production host, it was explored whether the obtained GPCR variants would permit higher expression in other eukaryotic hosts as well. Indeed, functional expression of GPCR variants evolved in yeast was also increased in insect cells. Notably, variants evolved in the E. co/i-based system show less improved functional expression when moved to eukaryotic hosts like insect cells, compared to the variants generated in yeast (Schlinkmann et al. (2012), J Mol Biol 422:414-428). The high expression levels in Sf9 cells obtained with the novel yeast-evolved variants are sufficient for crystallographic investigations with expression cultures in liter scales under standard conditions. More specifically, purification of the NK1R variant NK1R-Y09 yielded very large quantities of pure protein. Furthermore, NK1R-Y09 showed signaling activities similar to wild-type NK1 R, illustrating that the disclosed method can give rise to biologically active variants able to perform their naturally intended function, which consequently can be studied conveniently with these variants.
In summary, the method for directed evolution of GPCRs in S. cerevisiae disclosed herein allows efficient generation of GPCR variants with high functional expression in eukaryotic expression hosts in a short time. Using S. cerevisiae as evolution host resulted in GPCR variants with high functional expression levels in both yeast and insect cells, in the latter system by up to a factor 27. Thus, the mutations are transferable and selected variants showed a greater increase in expression in yeast as well as in insect cells than when evolution had been performed in E. coli.
The transferability of the beneficial effects of the receptor mutations to insect cells further promotes Sf9 cells as expression host for large-scale production of the generated mutants. Nonetheless, the fact that high expression levels in S. cerevisiae can now also be obtained as a result of GPCR evolution, combined with induced host adaptation, implies a potential of yeast as a large-scale production host. From a practical point of view, the fact that in adapted strains the ratio of functional receptor to total receptor produced is improved is a great advantage for purification strategies like IMAC, which cannot discriminate between non-functional and functional protein. This is of special interest for purifications of receptors for which no ligand affinity column, for instance like the cleavable ligand column as established for purification of NTR1 (Egloff et al. (2014), Protein Expr Purif, in press), is available.
The versatility of S. cerevisiae broadens the portfolio of GPCRs amenable to our approach, allowing evolution of GPCRs for which the previously established system in E. coli is inadequate. For instance, for GPCRs like KOR1, which show very low functional expression levels even in eukaryotes, expression levels in E. coli may be below the threshold needed for successful evolution. Furthermore, while neither prokaryotes nor native yeast do produce cholesterol—on which activity of some GPCRs relies—cholesterol-producing S. cerevisiae strains have been engineered. In principle, the disclosed method should be easily transferable to such or any other engineered yeast strains.
Yeast Strain, Vectors, Cultivation, and Expression.
For all experiments S. cerevisiae strain BY4741 (MATa his3Δ1 Ieu2Δ0 metl5Δ0 ura3Δ0), obtained from EUROSCARF, was used. Standard expression was performed with pMSO3het, derived from p415 GAL1. For more details on different yeast expression vectors see below. BY4741 transformed with pMSO3het vectors were cultivated at 30° C. in SDD-Leu medium (6.9 g/L yeast nitrogen base without amino acids (Formedium), 690 mg/L complete supplement mixture without leucine (Formedium), 20 g/L glucose, 35 mM sodium citrate tribasic, 35 mM citric acid). For expression, yeast cells in the logarithmic growth phase grown in SDD-Leu− medium at 30° C. were centrifuged and subsequently resuspended in SDG-Leu-medium (identical to SDD-Leu but with 20 g/L galactose instead of glucose). Initial OD600 was always chosen to be 1.0 after resuspension in SDG-Leu˜ and expression was performed at 20° C. for 24 h.
Yeast Expression Vectors
To obtain pMSO3het, the α-mating factor prepro sequence was cloned from pPICZa A (Life Technologies) into the multiple cloning site (Xhol/Spel) of p415 GAL1. pMS03het contains NheIlBamHl restriction sites which allow efficient vector linearization for high-efficiency transformation or in-frame cloning of genes preceded by the α-mating factor prepro sequence. For expression of GPCRs with a C-terminal HA tag or fusion to mCherry, vectors pMSO3het_HA or pMSO3het_mCh were used, respectively. To obtain pMSO3het_HA and pMSO3het_mCh, sequences coding for HA tag or mCherry were cloned via BamHI into pMSO3het.
Yeast Library Construction and Transformation.
The wild-type gene of rat NTR1 (N-terminally truncated from amino acids 1-42) was a kind gift from Reinhard Grisshammer (National Institutes of Health). Wild-type cDNA of human NK1R and human KOR1 was obtained from the Missouri S&T cDNA Resource Center. All wild-type GPCR genes were cloned into pMS03het (Nhel/BamHI). DNA library construction was performed by amplification of wild-type genes or isolated DNA after the first round of evolution with error-prone PCR using the GeneMorph 11 Random Mutagenesis Kit (Agilent Technologies) according to the manufacturer's protocol. Transformation of BY4741 with DNA libraries was performed by square wave electroporation on a GenePulser Xcell electroporator (Bio-Rad). On average, libraries with a diversity of 5×107-1×108 were obtained. For more details on library construction and high-efficiency transformation see below.
DNA Library Construction and High-Efficiency Transformation
DNA libraries of wild-type GPCRs or isolated versions after the first round of evolution were generated by performing two error-prone PCRs (epPCRs), one epPCR with 20 and one with 25 cycles, of which the obtained products were subsequently pooled. Primers used for epPCR introduced sites homologous to linearized pMSO3het (digested with Nhel/BamHl) at each end of the gene.
Forward Primer:
5′-CTAAAGAAGAAGGGGTATCTCTCGAGAAACGTGAGGCGGAAGCGGCTAGC-3′;
Reverse Primer:
5′-ATTACATGACTCGACTCGATGCCGACGAGAGCGGCCGCCTATTAGGATCC-3′.
The obtained epPCR products were further amplified by standard PCR with the identical primers in order to obtain enough DNA material for transformation.
High-efficiency transformation by square wave electroporation was performed analogously to a previously published method (Van Deventer & Wittrup (2014), Methods Mol Biol 1131:151-181) with some minor adaptations. Yeast cells were grown in 60 mL YPD at 30° C. to an OD600=1.8-2.0. As soon as this cell density was reached, 50 mL of culture were centrifuged, the medium aspirated, and cells were treated in 25 mL conditioning solution (100 mM lithium acetate, 10 mM DTT) at 30° C. for 15 min. Subsequently, cells were pelleted, washed in 25 mL cold ddH2O, pelleted again, and resuspended in cold ddH2O to a total volume of 500 μL. Henceforward, cells were always kept at 4° C. For one transformation, 250 μL of yeast cells were mixed with 4 μg of linearized pMSO3het and 12 μg PCR product and the transformation mixture was transferred to a 2 mm electroporation cuvette. Square wave electroporation was performed with one pulse with a voltage of 500 V and a pulse length of 15 ms. After electroporation, cells were allowed to recover in 5 mL YPD without shaking at 30° C. for 1 h. Finally, recovered cells were pelleted, transferred to 500 mL SDD-Leu− for selective growth at 30° C. for 20-24 h, and stored in glycerol stocks at −80° C. To obtain high-diversity libraries, always two transformations per library were performed.
Permeabilization of Yeast Cells and Binding of Fluorescent Ligand.
After expression, cultures were centrifuged, medium was aspirated, and cells were resuspended in TELi buffer (50 mM Tris-HCl pH 9.0 (at 4° C.), 1 mM EDTA, 100 mM lithium acetate) at RT. Next, cells were incubated in TELi Buffer supplemented with 50 mM DTT at 20° C. for 30 min and subsequently washed twice in cold TELi Buffer. Henceforward, cells were always kept at 4° C. For fluorescent ligand binding, permeabilized cells were incubated with ligand labeled with HiLyte Fluor 488 (AnaSpec) (NTR1 variants: 25 nM fluorescent neurotensin (8-13); NK1R variants: 20 nM fluorescent substance P; KOR1 variants: 10 nM fluorescent dynorphin A (1-11)) in TELi buffer at 4° C. without exposure to light for 2 h. After incubation, cells were washed once in TELi buffer prior to measurements. Non-specific binding was determined in the presence of a 1000-fold excess of unlabeled ligand (NTR1 variants: 25 μM neurotensin (8-13) (AnaSpec); NK1R variants: 20 μM substance P (AnaSpec); KOR1 variants: 10 μM dynorphin A (1-11) (GenScript)). For more details on fluorescent ligands see below.
Fluorescent Ligands
All fluorescent ligands were labeled with HiLyte Fluor 488 (Anaspec). Neurotensin (8-13) (KKPYIL) was covalently labeled at the N-terminal amino group. Substance P (RPKPQQFFGLM) was covalently labeled at the amino group of lysine-3. Dynorphin A (1-11) (YGGFLRRIRPK) was covalently labeled at the amino group of lysine-11.
Flow Cytometry and FACS.
Cells fluorescently labeled by ligand binding were kept in TELi buffer for measurements. Flow cytometry was performed on a BD FACSCanto II cytometer (BD Biosciences) or on a BD LSRFortessa cell analyzer (BD Biosciences) and FACS was performed on a BD FACSAria Ill sorter (BD Biosciences). For analytical measurements always 50,000 events were recorded. In FACS, 3×105-5×105 of the 0.5-1.0% most fluorescent cells were sorted into SDD-Leu− medium for subsequent cultivation at 30° C. for 24 h. For all samples identical acquisition settings were used in order to allow comparative analysis. Data were analyzed with FlowJo vX.0.7.
Radioligand Binding Assays.
RLBAs on whole yeast cells were performed as follows: First, 1×108 cells (assuming OD600=1.0 corresponds to 107 cells/mL) were harvested after expression and treated by consecutive washing first in ddH2O, then SPH1 buffer (1 M sorbitol, 25 mM EDTA, 50 mM DTT, pH 8.0), and finally in 1 M sorbitol. Next, cells were resuspended in SPH2 buffer (1 M sorbitol, 1 mM EDTA, 10 mM potassium citrate tribasic, pH 5.8) and cell wall digestion was performed by addition of 6 U/mL Zymolyase 20T (AMS Biotechnology) followed by incubation at 30° C. for 30 min. Henceforward, cells were kept at 4° C. Subsequently, cells were incubated at 4° C. for 2 h in 50 mM Tris-HCl pH 7.4 (at 4° C.) containing [3H]-labeled ligand (NTR1 variants: 20 nM [3,11-Tyrosyl-3,5-3H(N)]-neurotensin (Perkin Elmer); NK1R variants: 15 nM [Leucyl-3,4,5-3H(N)]-substance P (Perkin Elmer); KOR1 variants: 15 nM [15,16−3H]-diprenorphine (Perkin Elmer)). Nonspecific binding was determined in the presence of a 1000-fold excess of unlabeled ligand (NTR1 variants: 20 μM neurotensin (8-13) (AnaSpec); NK1R variants: 15 μM substance P (AnaSpec); KOR1 variants: 15 μM diprenorphine (Tocris Bioscience)). After incubation, cells were filtered on MultiScreen filter plates (Merck Millipore) with a vacuum manifold, filters were washed four times with cold 50 mM Tris-HCl pH 7.4 (at 4° C.), transferred to Isoplate-96 scintillation plates (Perkin Elmer), dried at 65° C. for 2 h, and Optiphase Supermix scintillation cocktail (Perkin Elmer) was added. RLBA measurements were performed on a 1450 MicroBeta Plus liquid scintillation counter (Wallac).
RLBAs on whole Sf9 cells were performed as described previously (Schlinkmann et al. (2012), J Mol Biol 422:414-428; Egloff et al. (2014), Proc Natl Acad Sci USA 111:E655-62.). Whole cells were incubated in binding buffer (50 mM Tris-HCl pH 7.4 (at 4° C.), 1 mM EDTA, 0.1% (w/v) BSA and 40 μg/mL bacitracin) containing 15 nM [3H]-labeled ligand. Nonspecific binding was determined in the presence of 15 μM unlabeled ligand. The same ligands as for RLBAs with yeast cells were used.
Quantitative Western Blot.
After expression, 4×107 cells (assuming OD600=1.0 corresponds to 107 cells/mL) for each sample were centrifuged and whole cell protein extraction was performed according to a previously published protocol (Zhang et al. (2011), Yeast 28:795-798.). Protein detection was performed with the primary antibodies rabbit anti-HA (Sigma-Aldrich, H6908) and mouse anti-actin (Abcam, ab8224) and the secondary antibodies goat anti-rabbit conjugated to Alexa Fluor 680 (Life Technologies, A-21076) and donkey anti-mouse conjugated to IRDye800 (Rockland Immunochemicals, 610-732-124). Image acquisition was performed on an Odyssey system (LI-COR Biosciences) and quantification was performed with Image Studio Lite version 3.1.4 (LI-COR Biosciences). For more details see below.
Yeast Whole Cell Protein Extraction, SDS-PAGE, and Western Blot
For whole cell protein extraction, samples after expression were centrifuged, medium was aspirated, and pelleted cells were resuspended in 500 μL 2 M lithium acetate for incubation on ice for 5 min. Next, cells were pelleted, the supernatant was removed, 100 μL of 0.4 M NaOH were added, and samples were incubated on ice for 5 min. After incubation, the samples were centrifuged, the supernatant was removed and the cell pellets were resuspended in 200 μL reducing NuPAGE LDS sample buffer (Life Technologies). Samples were incubated at 20° C. for 15 min, centrifuged, and 5 μL of each sample were run on a NuPAGE Novex 4-12% Bis-Tris protein gel (Life Technologies) in NuPAGE MES SDS running buffer (Life Technologies). Wet blotting was performed onto Immobilon-FL membranes (Merck Millipore). Blocking of membranes was performed in 1× Casein blocking buffer (Sigma-Aldrich) in PBS at RT for 20 min. Antibody binding was performed in 1× Casein blocking buffer in PBST (PBS, 0.05% (v/v) Tween-20) at RT for 1 h, and PBST was used for all membrane washing steps. Primary rabbit anti-HA antibody was used at a dilution of 1:5,000, primary mouse anti-actin antibody at a dilution of 1:1,000, and the secondary antibodies (goat anti-rabbit conjugated to Alexa Fluor 680 and donkey anti-mouse conjugated to IRDye800) both at a dilution of 1:10,000.
Confocal Fluorescence Microscopy.
Yeast cells were permeabilized and binding of fluorescent neurotensin was performed as described above. After washing, cells were transferred into Nunc Lab-Tek II chambered coverglasses (Thermo Scientific) and confocal microscopy was performed on a Leica TCS SP5 microscope (Leica Microsystems). For all samples magnification was 630-fold and identical acquisition settings were used in order to allow comparative analysis.
Spodoptera frugiperda (Sf9) Vectors and Expression.
Wild-type and evolved receptor constructs were amplified by PCR from yeast expression vector pMS03het and cloned via SLIC into a modified MultiBac pFL vector. The vector designated as pFL_mFLAG_His10_TEV_SLIC contains an expression cassette with an N-terminal melittin signal sequence followed by a FLAG tag, a deca-histidine tag, a TEV protease cleavage site, and a SLIC cloning site. E. coli DH10 EMBacY cells were transformed with pFL vectors containing the different receptor genes and the resulting baculovirus genome was isolated. For details on generation of recombinant baculovirus and baculovirus-infected insect cell stocks (BIICs) see below. Expression was performed in Sf-900 II SFM medium (Life Technologies) by infection of Sf9 cells at a density of 3×106 cells/mL with 100-fold diluted BIIC stocks and cultivation at 27° C. for 4 d. After expression, cells were harvested by centrifugation, washed in cold PBS and stored at −80° C. until use.
Generation of Recombinant Baculovirus and Baculovirus-Infected Insect Cell Stocks (BIlCs).
Recombinant baculovirus was generated by transfecting 8×105 Sf9 cells in 2 mL of Sf-900 II SFM medium (Life Technologies) using 8 μL Cellfectin II reagent (Life Technologies). After 4 h of incubation in a humidified incubator at 27° C., the transfection medium was removed and replaced by 2 mL of fresh Sf-900 II SFM. V0 viral stock was harvested after 5 d at 27° C. and used to generate V1 high-titer virus stock (108-109 viral particles per mL). V1 virus stock was then used to generate BIICs. Briefly, Sf9 cells at a density of 106 cells/mL were infected with a multiplicity of infection (MOI) of 5, incubated for 24 h in suspension, harvested and frozen at −80° C. in aliquots in Sf-900 II SFM containing Penicillin-Streptomycin (Life Technologies) and 10% (v/v) DMSO.
[35S]-GTPyS Binding Assay.
Membranes used for [35S]-GTPyS binding assays were isolated as previously described (Egloff et al. (2014), Proc Natl Acad Sci USA 111:E655-62). Briefly, cells were disrupted by shear force after osmotic shock. In order to reduce unspecific [35S]-GTPyS binding, membranes were washed in a urea containing buffer. The isolated membranes were frozen in aliquots and stored at −80° C. Receptor levels on urea-washed membranes were determined by radioligand binding assays as described above.
The [35S]-GTPyS binding assay was performed as previously described (Egloff et al. (2014), Proc Natl Acad Sci USA 111:E655-62). Briefly, 1 nM GPCR in urea-washed membranes and 100 nM G protein (Gai1lp1y1, purified according to (Rasmussen et al. (2011), Nature 477:549-555.)) in assay buffer (50 mM Tris-HCl pH 7.4 (at 4° C.), 1 mM EDTA, 100 mM NaCl, 1 mM DTT, 3 mM MgSO4, 0.3% (w/v) BSA, 2 μM GDP, 4 nM [35S]-GTPyS (Perkin Elmer)) were incubated at 25° C. for 20 min in the presence or absence of 200 μM substance P (AnaSpec). Background counts arising from buffer, GPCR and G protein alone have been taken into account and subtracted. Therefore, given counts represent the GPCR-induced [35S]-GTPyS binding to G protein in the presence and absence of agonist.
Purification of NK1R Variant NK1R-YO9
All steps were performed at 4° C. Frozen Sf9 cells were thawed and swelled in a hypotonic buffer (10 mM HEPES pH 7.4, 20 mM KCl, 10 mM MgCl2, cOmplete ULTRA EDTA-free tablets (Roche), 1 μM substance P) for 1 h. Cell membranes were then disrupted by homogenization (Dounce homogenizer) and collected by centrifugation at 130,000 rcf. The isolated membranes were extensively washed twice by repeated homogenization (Dounce homogenizer) and centrifugation in hypotonic buffer, followed by three repetitions of this procedure using a hypertonic buffer (10 mM HEPES pH 7.4, 1 M NaCl, 20 mM KCl, 10 mM MgCl2, cOmplete ULTRA EDTA-free tablets, 1 μM substance P). Purified membranes were resuspended in solubilization buffer (30 mM HEPES pH 7.4, 150 mM NaCl, 10 mM MgCl2, cOmplete ULTRA EDTA-free tablets, 1 μM substance P, 2 mg/mL iodoacetamide (Sigma)) and stirred for 45 min. Membranes were then solubilized in 1.5% (w/v) n-dodecyl-p-D-maltopyranoside (DDM, Anatrace) and 0.3% (w/v) cholesteryl hemisuccinate (CHS, Sigma). After 3 h of stirring, non-solubilized material was removed by centrifugation (130,000 rcf, 40 min, 4° C.). The supernatant was adjusted to contain an end concentration of 800 mM NaCl and 25 mM imidazole before being incubated overnight with 1.5 mL TALON Superflow resin (Clontech). The protein-bound resin was transferred into gravity flow columns and then washed with 10 column volumes (CV) each of Wash 1 buffer (25 mM HEPES pH 7.4, 800 mM NaCl, 10 mM MgCl2, 25 mM imidazole, 10% (v/v) glycerol, 0.5 μM substance P, 0.3%/0.06% DDM/CHS), Wash 2 buffer (25 mM HEPES pH 7.4, 400 mM NaCl, 10 mM MgCl2, 40 mM imidazole. 10% (v/v) glycerol, 0.5 μM substance P, 0.2%/0.04% DDM/CHS, 10 mM ATP), and Wash 3 buffer (25 mM HEPES pH 7.4, 200 mM NaCl, 40 mM imidazole, 10% (v/v) glycerol, 0.5 μM substance P, 0.2%/0.04% DDM/CHS). NK1R-Y09 was eluted in 4 CV of elution buffer (25 mM HEPES pH 7.4, 200 mM NaCl, 300 mM imidazole, 10% (v/v) glycerol, 0.5 μM substance P, 0.2%/0.04% DDM/CHS). Purified receptor was concentrated to 0.5 mL with 100 kDa molecular weight cut-off Vivaspin centrifuge concentrators (Sartorius Stedim Biotech). Size-exclusion chromatography (SEC) was performed on an Aekta Pure FPLC system (GE Healthcare) with a Superdex S200 Increase 10/300 GL column (GE Healthcare) equilibrated with SEC buffer (20 mM HEPES pH 7.4, 150 mM NaCl, 0.05%0.01% DDM/CHS, 0.1 μM substance P).
Directed evolution of GPCRs towards higher expression levels in Escherichia coli has been established previously (Sarkar et al. (2008), Proc Natl Acad Sci USA 105:14808-14813; Dodevski & Plückthun (2011), J Mol Biol 408:599-615). This invention discloses the analogous method for directed evolution in the yeast Saccharomyces cerevisiae. Several GPCRs have been evolved with this method in S. cerevisiae, leading to high functional expression in yeast as well as in insect cells and outperforming expression levels obtained from variants evolved in the E. coli-based system. Usually, only two rounds of evolution were required to obtain highly expressed variants.
Interestingly, the yeast strain used can also be adapted towards increased GPCR production. This is automatically induced during evolution by repetitive sorting with fluorescence-activated cell sorting (FACS) and gradually arises in the course of selection. Thus, obtained expression levels of selected libraries are a result of GPCR evolution combined with host adaption. If individual GPCR variants are expressed in newly transformed yeast cells, the average expression level will be lower than for selected libraries, since these cells were not adapted. However, simply by repetitive FACS selections the adaption can be induced for any evolved variant.
GPCR Expression Constructs
Several different GPCR expression constructs are available. All vectors are derived from p415 GAL1. The vectors are maintained as low-copy plasmids in S. cerevisiae providing a functional LEU2 gene for selective growth of strains auxotrophic for leucine and can be easily re-isolated from yeast after transformation. Being shuttle-vectors, they can be propagated as high-copy vectors in E. coli, in which correspondingly all cloning steps are performed.
Every vector encodes the α-mating factor prepro sequence preceding the cloning site for insertion of the GPCR gene. Cloning is done via Nhel and BamHI restriction sites, allowing in-frame insertion of the gene of interest. Expression is under control of the inducible GAL1 promoter. The different vectors vary in the sequences after the cloning site. The standard vector used for evolution, pMSO3het, does not encode any tag or protein after the cloning site, thus GPCRs without any fusions are produced with pMSO3het. The other vectors encode constructs containing a C-terminal cleavable deca-histidine-tag (pMSO3het_3CHis10), a HA-tag (pMSO3het HA), a fusion to mCherry (pMSO3het mCh), or an AviTag™ (pMS03het_Avi).
Permeabilization of Yeast Cells for Ligand Binding
This section describes the permeabilization of yeast cells for ligand binding experiments. Conditioning with TELi buffer and DTT permeabilizes the cell wall and allows diffusion of ligands to the GPCRs in the cell membrane. The advantage of this permeabilization method is that the cells remain viable and are not as fragile as spheroplasts. Therefore, this procedure can be used for analytical flow cytometry and FACS, as well as for any other application in which ligand binding is detected (e.g. confocal fluorescence microscopy, RLBA).
Protocol:
Measurement of GPCR Expression with Flow Cytometry
Detection of functional GPCR expression at the cell surface can in principle be performed by fluorescent ligand binding measured in flow cytometry. Upon incubation with fluorescent ligand, only functionally expressed GPCRs located in the plasma membrane will bind the ligand. After removal of unbound ligand by washing, ligand binding is detected by measuring fluorescence intensity. However, the yeast cell wall acts as a protective barrier, not allowing larger molecules like peptide ligands to pass through. Therefore, yeast cells need to be permeabilized prior to incubation with fluorescent ligand. Next to analytical quantification of functional expression levels, fluorescent ligand binding is also used for selections with FACS. Cells expressing a variant with a high-expression phenotype will correspondingly exhibit a higher fluorescence intensity. By sorting the most fluorescent cells, highly expressing GPCR variants can be isolated.
While incubation with fluorescent ligand allows measurement of the total signal, it is important to determine the non-specific signal as well. This can be done by incubation of fluorescent ligand with an excess of non-labeled ligand in a competition binding experiment. The obtained non-specific signal in such measurements depicts the background. Usually, GPCR expression in yeast leads to two subpopulation of cells. Whereas one fraction of cells shows surface expression of active GPCRs, there is also a subpopulation for which no active surface expression is observed. Typically, this fraction of non-expressing cells decreases in adapted strains.
Workflow of Selections with FACS
An overview of the workflow of selections is depicted in
For adaptation of strains expressing an evolved GPCR variant, the workflow is almost identical. The only difference is that no library needs to be prepared. Strain adaptation starts with a single clone expressing the GPCR variant of choice which is subjected to 5 rounds of selection with FACS. After the selections, glycerol stocks of the adapted strains can be prepared.
| Number | Date | Country | Kind |
|---|---|---|---|
| 14199729.6 | Dec 2014 | EP | regional |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 15538479 | Jun 2017 | US |
| Child | 18349312 | US |
