Control of the Yeast Mating Pathway by Reconstitution of Functional … Synthetic Biology... ·...

8
Control of the Yeast Mating Pathway by Reconstitution of Functional αFactor Using Split Intein-Catalyzed Reactions Ka-Hei Siu and Wilfred Chen* Department of Chemical and Biomolecular Engineering, University of Delaware, 150 Academy Street, Newark, Delaware 19716, United States * S Supporting Information ABSTRACT: Synthetic control strategies using signaling peptides to regulate and coordinate cellular behaviors in multicellular organisms and synthetic consortia remain largely underdeveloped because of the complexities necessitated by heterologous peptide expression. Using recombinant proteins that exploit split intein-mediated reactions, we presented here a new strategy for reconstituting functional signaling peptides capable of eliciting desired cellular responses in S. cerevisiae. These designs can potentially be tailored to any signaling peptides to be reconstituted, as the split inteins are promiscuous and both the peptides and the reactions are amenable to changes by directed evolution and other protein engineering tools, thereby oering a general strategy to implement synthetic control strategies in a large variety of applications. KEYWORDS: inteins, α-factor, peptide analogues, cell signaling, gene regulation L iving systems of all scales, from unicellular microbes to multispecies communities, must constantly adapt to changing conditions by continually surveying their surround- ings for relevant stimuli. One of the most common mechanisms for this continual process is the use of signaling pathways that are able to sense an external input of interest, transmit the signal inside the cell, and actuate appropriate cellular responses such as transcriptional upregulation. 1 Often, this transduction mechanism involves a signaling molecule and its cognate receptor, as frequently seen in the myriad of G-protein coupled receptors (GPCRs) present in eukaryotes. 2 While a limited number of signaling molecules and their associated pathways (i.e., quorum sensing) have been successfully engineered into dierent species, 3,4 more complex signals such as peptide hormones have not yet been widely incorporated into synthetic control schemes. As opposed to the small molecules used by bacterial quorum sensing systems, peptide signaling molecules are often dicult to secrete, and engineering of their cognate membrane bound GPCRs has proven to be even more challenging because of their very poor expression in recombinant hosts. 5,6 As a result, the use of GPCRs and their associated signaling pathways for orthogonal control of cellular behaviors remain elusive. Despite these diculties, there are tremendous scientic and clinical interests to adopt these peptide-mediated signaling pathways as synthetic extracellular sensing circuits because of their importance in cell-to-cell communication and disease progression. 7,8 We present here a new generalizable framework for designing synthetic extracellular sensing circuits that could be adapted for any signaling peptide of interest. Our design involves the use of split inteins to facilitate the functional reconstitution of inactive signaling peptides through either protein trans-splicing (PTS) or N-terminal cleavage (NTC) reactions (Scheme 1). In PTS, two polypeptides, called the N- and the C-inteins, are able to associate with each other and trigger a series of biochemical reactions that result in the self- excision of the intein sequences from the protein complex and, concomitantly, the formation of a new peptide bond between their anking sequences, called the exteins 9 (Scheme S1A). By mutating a key residue at the catalytic site of the intein complex (N146A), the PTS reaction pathway can be redirected to NTC in which the N-extein is cleaved from the complex instead of being ligated to the C-extein 10,11 (Scheme S1B). To demonstrate the feasibility of our strategy, the well- characterized α-factor (Figure 1A), which activates the yeast mating pheromone response pathway (Figure S1) upon binding onto its native receptor, Ste2p, 12,13 was used as a model signaling peptide. This pathway is ideally suited to establish and validate our approach as it has been used as a model for all eukaryotic GPCR signaling pathway and engineered to perform a variety of synthetic functions with tunable behaviors. 14,15 The yeast α-factor is a 13-amino acid peptide composed of three distinct domains: residues that initiate signaling in the N terminus, residues that mediate binding to Ste2p in the C-terminus, and a exible loop region Received: March 9, 2017 Published: May 15, 2017 Letter pubs.acs.org/synthbio © 2017 American Chemical Society 1453 DOI: 10.1021/acssynbio.7b00078 ACS Synth. Biol. 2017, 6, 14531460

Transcript of Control of the Yeast Mating Pathway by Reconstitution of Functional … Synthetic Biology... ·...

Page 1: Control of the Yeast Mating Pathway by Reconstitution of Functional … Synthetic Biology... · Control of the Yeast Mating Pathway by Reconstitution of Functional α‑Factor Using

Control of the Yeast Mating Pathway by Reconstitution of Functionalα‑Factor Using Split Intein-Catalyzed ReactionsKa-Hei Siu and Wilfred Chen*

Department of Chemical and Biomolecular Engineering, University of Delaware, 150 Academy Street, Newark, Delaware 19716,United States

*S Supporting Information

ABSTRACT: Synthetic control strategies using signalingpeptides to regulate and coordinate cellular behaviors inmulticellular organisms and synthetic consortia remain largelyunderdeveloped because of the complexities necessitated byheterologous peptide expression. Using recombinant proteinsthat exploit split intein-mediated reactions, we presented herea new strategy for reconstituting functional signaling peptidescapable of eliciting desired cellular responses in S. cerevisiae.These designs can potentially be tailored to any signalingpeptides to be reconstituted, as the split inteins arepromiscuous and both the peptides and the reactions areamenable to changes by directed evolution and other proteinengineering tools, thereby offering a general strategy toimplement synthetic control strategies in a large variety of applications.

KEYWORDS: inteins, α-factor, peptide analogues, cell signaling, gene regulation

Living systems of all scales, from unicellular microbes tomultispecies communities, must constantly adapt to

changing conditions by continually surveying their surround-ings for relevant stimuli. One of the most common mechanismsfor this continual process is the use of signaling pathways thatare able to sense an external input of interest, transmit thesignal inside the cell, and actuate appropriate cellular responsessuch as transcriptional upregulation.1 Often, this transductionmechanism involves a signaling molecule and its cognatereceptor, as frequently seen in the myriad of G-protein coupledreceptors (GPCRs) present in eukaryotes.2 While a limitednumber of signaling molecules and their associated pathways(i.e., quorum sensing) have been successfully engineered intodifferent species,3,4 more complex signals such as peptidehormones have not yet been widely incorporated into syntheticcontrol schemes. As opposed to the small molecules used bybacterial quorum sensing systems, peptide signaling moleculesare often difficult to secrete, and engineering of their cognatemembrane bound GPCRs has proven to be even morechallenging because of their very poor expression inrecombinant hosts.5,6 As a result, the use of GPCRs and theirassociated signaling pathways for orthogonal control of cellularbehaviors remain elusive.Despite these difficulties, there are tremendous scientific and

clinical interests to adopt these peptide-mediated signalingpathways as synthetic extracellular sensing circuits because oftheir importance in cell-to-cell communication and diseaseprogression.7,8 We present here a new generalizable frameworkfor designing synthetic extracellular sensing circuits that could

be adapted for any signaling peptide of interest. Our designinvolves the use of split inteins to facilitate the functionalreconstitution of inactive signaling peptides through eitherprotein trans-splicing (PTS) or N-terminal cleavage (NTC)reactions (Scheme 1). In PTS, two polypeptides, called the N-and the C-inteins, are able to associate with each other andtrigger a series of biochemical reactions that result in the self-excision of the intein sequences from the protein complex and,concomitantly, the formation of a new peptide bond betweentheir flanking sequences, called the exteins9 (Scheme S1A). Bymutating a key residue at the catalytic site of the intein complex(N146A), the PTS reaction pathway can be redirected to NTCin which the N-extein is cleaved from the complex instead ofbeing ligated to the C-extein10,11 (Scheme S1B).To demonstrate the feasibility of our strategy, the well-

characterized α-factor (Figure 1A), which activates the yeastmating pheromone response pathway (Figure S1) uponbinding onto its native receptor, Ste2p,12,13 was used as amodel signaling peptide. This pathway is ideally suited toestablish and validate our approach as it has been used as amodel for all eukaryotic GPCR signaling pathway andengineered to perform a variety of synthetic functions withtunable behaviors.14,15 The yeast α-factor is a 13-amino acidpeptide composed of three distinct domains: residues thatinitiate signaling in the N terminus, residues that mediatebinding to Ste2p in the C-terminus, and a flexible loop region

Received: March 9, 2017Published: May 15, 2017

Letter

pubs.acs.org/synthbio

© 2017 American Chemical Society 1453 DOI: 10.1021/acssynbio.7b00078ACS Synth. Biol. 2017, 6, 1453−1460

Page 2: Control of the Yeast Mating Pathway by Reconstitution of Functional … Synthetic Biology... · Control of the Yeast Mating Pathway by Reconstitution of Functional α‑Factor Using

in the middle to orient the signaling and binding domains(Figure 1A).13 Since mutations to either the N- or C-terminalregion essentially abolish GPCR activation,14,15 these results

indicate that both GPCR binding and activation are necessaryfor correct pathway signaling. We hypothesized that split α-factor fragments composed of either the N- or C-terminal

Scheme 1. Protein trans-Splicing (PTS) and N-Terminal Cleavage (NTC) Reactions Catalyzed by Split Inteins

Figure 1. Adapting yeast mating pheromone, α-factor, for synthetic control. (A) Wild-type α-factor sequence binds to its native receptor and triggersdownstream cascade that results in many cellular responses, including cell cycle arrest. (B) A small library of α-factor analogues (see Table 1 forcomplete list) with substitutions (marked by X) to insert the Cys residue required for PTS was designed and chemically synthesized, with theprospective split site between residues 8 and 9 (marked by ▲). The synthesized analogues were added to growing yeast on agar in increasingamounts to evaluate their activity. (C) Analogues 1 and 2 displayed the highest activities relative to the wild-type (WT). For variants 1B and 2B, anN-terminal Met residue was added to peptides 1 and 2 to mimic the final spliced product after fusion proteins were expressed from E. coli.

ACS Synthetic Biology Letter

DOI: 10.1021/acssynbio.7b00078ACS Synth. Biol. 2017, 6, 1453−1460

1454

Page 3: Control of the Yeast Mating Pathway by Reconstitution of Functional … Synthetic Biology... · Control of the Yeast Mating Pathway by Reconstitution of Functional α‑Factor Using

region that are incapable of activating the mating pathway canbe created for intein-mediated functional reconstitution.For effective intein ligation, an absolute requirement is a Cys

residue at the +1 position after the C-extein.16 Since a largevariety of residues have been substituted into the loop domainof α-factor without affecting biological activity,16 we hypothe-sized that a synthetic α-factor analogue that include therequired residues for PTS while also retaining biologicalactivities could be created. To test this hypothesis, we designeda small library of synthetic α-factor analogues (Table 1) and

screened them using a simple plate-based growth arrest assay(Figure 1B and S2) based on the induction of cell cycle arrestin the G1 phase upon activating the mating pathway.17 Sincethe alpha factor peptides were only 13−14 amino acids inlength, chemical synthesis was the most efficient method ofproduction rather than using recombinant expression. Byexploiting the easily observable growth arrest phenotype, wesuccessfully identified two synthetic analogues (1 and 2) thatretain significant levels of biological activities by correlating thesize of the clearing zone around a lawn of growing yeast cells16

(Figure 1C).On the basis of the results of the semiquantitative plate assay,

we further characterized and quantified the biological activity ofeach synthetic analogue using a liquid-based growth assay(Table 1, Figure S3). While we observed that almost allsynthetic analogues displayed a varying degree of activity,analogues 1 and 2 were again significantly more active thanothers. Additionally, the subset of analogues that mimicked thelength of the wild type peptides (1, 2, and WT) was at least 2times more active than their extended counterparts (1B, 2B,and WT-B) (Table 1). It is likely that the longer peptidesdistorted and/or displaced their termini from effectiveinteractions with the Ste2p receptor. We chose 1 and 2 to begenetically split and fused onto the naturally split Npu inteins18

as constructs N1/C1 and N2/C2 (an N-terminus Met is addedresulting in 1B and 2B as the final reconstituted peptideproducts) (Figure 2A). To ensure these split fragments has nobiological activity, all four fusion proteins were expressed inE. coli and purified using the flanking His6 tag. None of the splitfragments was shown to elicit growth arrest as assessed by the

liquid-based growth assay (Figure 3), an observation consistentwith the requirement of both binding and activation for propersignaling.Once we confirmed that the split fragments are inactive, we

next sought to determine the efficiency of intein ligation. Todemonstrate the splicing of our split synthetic analogues into asingle peptide, we first fused a maltose binding protein (MBP)to the N-terminus of the N-fragments to create constructs N1Band N2B for easy visualization of ligation by SDS-PAGE(Figure 2B) as it has been shown that additional sequences onthe exteins beyond the first few residues do not affect PTSsignificantly.19

The corresponding purified N1B/C1 and N2B/C2 fusionproteins were mixed together and the formation of newproducts were monitored and analyzed by SDS PAGE (Figure2C). Strikingly, the difference of a single residue at the +2position (Gly in C1 vs Ala in C2) resulted in the abolishment ofany observable trans-splicing reaction (Figure 2C, P1 vs P2),confirming previous studies on the effects of the +2 Gly residueon reaction kinetics using Npu split inteins.10,11 Although theestimated t1/2 for reconstitution of N2B/C2 is approximately180 min, well short of the rate reported for the wild-type C+2Phe (t1/2 ∼ 30 s19), this represented a compromise betweenPTS efficiency and biological activity for the split α-factoranalogue; the insertion of a bulky Phe residue into the loopdomain of the signaling peptide would likely prohibit bothtermini from interacting properly with the Ste2p receptor, assuggested by the lack of any bulky residues in the flexibledomain from previous mutational screening studies.12 This was

Table 1. Library of Synthetic α-Factor Analogues Screenedfor Growth Arrest Activities

sequenceanumber ofresidues

clearingformationb

EC50(nM)c

1 WHWLRLG^CGQPMY 13 Y 1001-Bd MWHWLRLG^CGQPMY 14 Y 2002 WHWLRL^CAGQPMY 13 Y 1502-Bd MWHWLRL^CAGQPMY 14 Y 4253 WHWLRLG^CFQPMY 13 N N.D.4 WHWLRLGA^CGQPMY 14 Y N.D.5 WHWLRLG^CAGQPMY 14 Y N.D.6 WHWL^CLSPGQPMY 13 Y 4507 WHWL^CLQPGQPMY 13 Y 400

aResidues that differ from WT are underlined. ^ marks the intendedsplit site of the synthetic analogue when fused onto split inteins.bClearing formation observed from agar plate-based growth arrestassays. cEC50 defined as the approximate concentration required toachieve 50% growth inhibition of MATa bar1Δ cells growing in YPD(Figure S3). d“B” denotes an N′-methionine added variant of theoriginal peptide sequence to account for expression of the finalrecombinant proteins in E. coli.

Figure 2. Testing PTS to reconstitute α-factor analogue in vitro. (A)Peptide analogues 1 and 2 were genetically split and fused on thetermini of the naturally split Npu intein for PTS. (B) An additionalmaltose-binding protein (MBP) was added to the N-terminus of the Nfragment of the α-factor to ease analysis of the reaction. (C)Comparison of the two reactions using the two split analoguesequences (∼1 μM of each protein) showed that only the N2B/C2mixture resulted in the accumulation of the expected spliced product(P2) over the course of 16 h.

ACS Synthetic Biology Letter

DOI: 10.1021/acssynbio.7b00078ACS Synth. Biol. 2017, 6, 1453−1460

1455

Page 4: Control of the Yeast Mating Pathway by Reconstitution of Functional … Synthetic Biology... · Control of the Yeast Mating Pathway by Reconstitution of Functional α‑Factor Using

further confirmed by our initial plate-based screen (Figure 1C;Table 1, variant 3).Using the 240 min window, we proceeded to ligate N2/C2

and test the resulting ligation products for the ability to triggercell cycle arrest in liquid cultures (Figure 3A). The addition ofthese reaction mixtures produced a dosage-dependent growthinhibition that strongly suggested the inducible response was adirect result of adding increasing amounts of our reconstitutedα-factor analogue (Figure 3B). This is further supported bydetecting the signature change in cell morphology associatedwith activation of mating response under a microscope (Figure3B, inset). However, the level of growth inhibition was rathermodest even at the highest reactant concentrations of 5 μM,suggesting either a lower ligation efficiency than expected or anartifact of the sensitivity of our growth assay. Since the ultimategoal of our strategy is to induce transcriptional upregulation, wefurther assessed the reaction products to induce GFPexpression from the pFus1 promoter. Again, a dosage-dependent GFP induction was observed through fluorescencemeasurements, flow cytometry, and immunoblotting (Figure3C). Interestingly, the transcriptional response is far moresensitive than the growth arrest assay with ∼7-fold GFPinduction even at 500 nM concentration. This result suggeststhat growth inhibition assay may be biased by the growth

conditions and even a low level of α- factor is sufficient toinduce the required expression phenotype. Collectively, theseresults confirm our ability to reconstitute functional α-factoranalogues for transcriptional activation, albeit at a slightly lowersensitivity than the wild-type version.One key reason why the sensitivity for the protein trans-

splicing approach is lower is the fact that the α-factor analogueitself is more than 10-fold less active compared to wild type(Figure 2; Table 1). One way to bypass this hurdle is to devisean alternative strategy for reconstitution using the wild-type α-factor sequence. Almost all reaction intermediates of PTS aresusceptible to nucleophile-induced cleavage (Figure S1B).10,19

Although these side reactions are typically undesirable as theylower the yields of the final ligated products from PTS, they canalso be exploited as an alternative reaction pathway to producea cleaved product.20 It may be possible to modify our strategyto take advantage of these side reactions in the form of N-terminal cleavage (NTC) in place of PTS to reconstitutefunctional signaling peptides (Scheme 1, Figure 4A). Thiscleavage scheme is promiscuous to the sequence of the excisedN-extein, allowing the use of wild-type α-factor instead of asynthetic analogue with reduced biological activity.However, it was uncertain whether fusion of the N-intein to

the C-terminus of the α-factor is sufficient to block binding to

Figure 3. Protein trans-splicing reaction for reconstitution of functional synthetic α-factor analogue. (A) Schematics of functional assays forreconstituted peptides after PTS or NTC reactions. (B) Addition of reaction mixtures induce cell cycle arrest in a dosage-dependent manner. Inset:Signature change in cell morphology observed in yeast cultures exposed to reaction mixtures containing 5 μM of N2/C2 proteins. Activation ofmating response signaling pathway exploited to drive expression of reporter GFP was confirmed and characterized by (C) microplate assay, (D)immunoblotting, and (E) flow cytometry.

ACS Synthetic Biology Letter

DOI: 10.1021/acssynbio.7b00078ACS Synth. Biol. 2017, 6, 1453−1460

1456

Page 5: Control of the Yeast Mating Pathway by Reconstitution of Functional … Synthetic Biology... · Control of the Yeast Mating Pathway by Reconstitution of Functional α‑Factor Using

the Ste2p receptor. To investigate this possibility, a new fusionprotein N3 was purified and was shown to be incapable ofeliciting any observable growth arrest (Figure 4D). This resultindicates that N-intein is effective in blocking Ste2p binding,rendering N3 inactive in mating activation. To further probethe kinetics of NTC, MBP was similarly fused to the N-terminus of N-extein (i.e., the α-factor peptide) to create N3Bfor easy visualization (Figure 4B). The formation of newproducts after mixing N3B/C3 were analyzed by SDS PAGE(Figure 4C), and the expected MBP-α-factor product with yield>50% over the course of an hour was detected. Thesignificantly higher efficiency of NTC relative to PTS waslikely because NTC avoided the rate-limiting step in PTSpathway (Scheme S1), which was exacerbated by thesubstitution of Ala at the C+2 extein residue.Having proven the ability to effectively excise N-exteins using

PTS, we proceeded to test whether the cleaved products arebiologically active and can provide functional signaling in liveyeast cultures. The addition of the N3/C3 reaction mixturesresulted in significant growth inhibition even at 500 nM, aconcentration 10-fold lower than that observed for the ligationreactions (Figure 4D). A similar enhancement in response wasalso detected with the GFP induction assay (Figure 4E). Thishigher sensitivity can be attributed to both the improved

reaction kinetics and the signaling efficiency of the wild-type α-factor sequence.Because of these improved properties, the N3/C3 system

was further evaluated for the ability to elicit signaling responsein an in vivo culture setting. Since yeast cultures are grown atpH 6.6 as compared to the optimum reaction pH of 7.4, theability to carry out the NTC reaction under these lower pHconditions was first investigated. The ligation rate remainedfairly constant for all pH values higher than 5.7, indicating thatthis approach will likely work even for growing yeast cultures(Figure S4). This was verified by observing the ability to induceGFP synthesis by the in situ release of functional α-factors byNTC. Purified N3 and C3 proteins (1 μM each) were addeddirectly into two different yeast cultures grown in either rich orsynthetic medium and up to a 6-fold increase in whole-cellfluorescence was detected within 4 h (Figure 5). A similar levelof response was observed by using an in vitro reaction mixtureat the same N3/C3 concentration (Figure 4E), signifyingrobustness of the ligation mechanism to reactivate functionalsignaling peptides in a culture setting and even in the presenceof live cells.In principle, any GPCRs associated with a peptide signaling

molecule, such as the human somatostatin, can be similarlyactivated by this method. More importantly, the ability toperform in situ activation in live cell cultures opens up the

Figure 4. N-terminal cleavage reactions to reconstitute α-factor. (A) The entire wild-type α-factor was fused to NpuN intein, thereby eliminating thesignal peptide’s ability to bind onto its receptor until NTC reactions cleave it from the fusion protein. (B) An additional maltose-binding protein(MBP) was added to the N-terminus of the α-factor to ease analysis of the reaction. (C) The expected cleavage product (P3) was observed toaccumulate in the reaction mixtures after 1 h of incubation at 37 °C. Cleaved α-factor remained biologically active and was able to induce (D) growtharrest and (E) transcriptional activation of pFUS1-GFP.

ACS Synthetic Biology Letter

DOI: 10.1021/acssynbio.7b00078ACS Synth. Biol. 2017, 6, 1453−1460

1457

Page 6: Control of the Yeast Mating Pathway by Reconstitution of Functional … Synthetic Biology... · Control of the Yeast Mating Pathway by Reconstitution of Functional α‑Factor Using

possibility of creating orthogonal yeast strains secreting eitherN3 or C3 under the control of different external inputs in orderto create a wide range of AND or OR gate logic yeast coculturesystems. More complex logic circuits can be designed byadapting conditional protein splicing (CPS), in which ligation isinitiated by the presence of a small molecule such asrapamycin21 (Figure S6). Other heterodimerization domainsfor proteins or metabolites can be similarly exploited forincreased flexibility. These feasibilities are current underinvestigation.

■ METHODSPeptide Design and Preparation. A small library of

synthetic analogues of α-factor were designed and listed inTable 1.All peptides were purchased from GenScript, Piscataway, NJ.

The chemically synthesized peptides were purified, lyophilized,and analyzed by MS and HPLC to ensure ≥70% purity by thecompany. Prior to use, lyophilized peptides were reconstitutedin 100% DMSO to stock concentrations of 10 mg/mL. Sterilewater was used to dilute stock peptide solutions where needed.Screening of Synthetic α-Factor Analogues. Individual

yeast BY4741 (MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0bar1::G418R) colonies were picked from YPD agar plates andgrown overnight in 3 mL YPD. Overnight cultures were used toinoculate 3 mL of fresh YPD at an initial OD600 of ∼0.05 andgrow to exponential phase (OD600 of ∼0.3−0.5). The growingyeast cultures were thoroughly mixed into sterile, melted 0.75%agar at 50 °C at an OD600 of ∼0.005 and quickly poured overYPD agar plates. Once the soft overlay containing yeastsolidified, 5 μL of α-factor analogue solutions of desiredconcentrations were added (Figure 1) and air-dried. Driedplated were incubated at 30 °C for 36 h prior to visualinspection and imaging. Images were taken using the Gel DocEz system (Biorad Laboratories).All synthetic analogues that resulted in clearings on plates

were further subjected to a liquid growth assay to quantify theirbiological activities. Similar to the plate-based assay, individualcolonies were picked and grown overnight. Overnight cultureswere used to inoculate 3 mL of fresh YPD at an initial OD600 of∼0.05 and grow to early exponential phase (OD600 of ∼0.2−

0.25). Varying concentrations of α-factor analogues were addedto these cultures, which were then incubated at 30 °C for 16 hwith shaking before OD600 were measured. The final OD600were normalized to the OD600 of control samples with nopeptides added (see Supporting Information). By plotting thenormalized OD600 against the concentrations of peptides addedon a semilog plot, the half maximal effective concentration(EC50) of each analogue could be identified (Figure S3).

Construction of Expression Plasmids. The plasmidsencoding all constructs were created by standard subcloningtechniques and are listed in Table S1. Briefly, genes for the Npusplit inteins were amplified from pI plasmids from Ramirez etal.22 using primers with additional flanking sequences encodingsplit α-factors, His-tag, and restriction sites as needed (TableS2). Amplified products were purified by gel electrophoresis.Purified products and pET-24a(+) vector (Novagen) weredigested using restriction enzymes (New England Biolabs) asneeded. Recombinant constructs were made by ligatingdigested vectors and inserts using T4 ligase (New EnglandBiolabs), followed by heat-shock transformation into NEB5αcells (New England Biolabs). Transformed cells were screenedby restriction digests and the sequences of each expressionconstruct were further confirmed by sequencing.

Protein Expression and Purification. E. coli BL21-Gold(DE3) cells (Agilent Technologies, Cedar Creek, TX) weretransformed with individual expression plasmids. Transformedcells were grown in 3 mL TB (Terrific Broth, 12g/L tryptone,24g/L yeast extract, 4 mL/L glycerol, 17 mM KH2PO4, 72 mMK2HPO4) at 37 °C overnight. Overnight cultures were used toinoculate 25 mL of fresh TB to an initial OD600 of ∼0.05 andgrow to midexponential phase (OD600 ∼ 0.75) at 37 °C.Protein expression was induced by addition of 200 μMisopropyl-β-thiogalactopyranoside (IPTG) at 20 °C for ∼16h. Induced cells were then harvested by centrifugation at 3000gfor 10 min, resuspended in lysis buffer (composition here) toan OD600 of ∼15, and lysed by sonication. Insoluble materialswere removed by centrifugation at 10 000g for 30 min. Desiredfusion proteins were purified from soluble supernatants byaffinity chromatography using His-Bind resin (Novagen)columns. Purified proteins were dialyzed against ReactionBuffer (50 mM Tris-HCl, 150 mM NaCl, 1 mM EDTA, 10% v/v glycerol, 5 mM TCEP, pH 7.4). Dialyzed protein solutionswere analyzed by SDS-PAGE and Bradford assays to estimatetheir concentrations and purities. These protein solutions wereused directly in in vitro reactions as described below.

In Vitro Reconstitution of Functional α-Factor fromSplit Intein Fusions. All in vitro reactions were carried out inReaction Buffer with TCEP added immediately prior to thestart of each reaction to a final concentration of 5 mM. Fusionproteins were mixed into the reaction mixtures at concen-trations as indicated. The assembled reaction mixtures wereincubated at room temperature on a rotator. Samples werecollected at times as noted.For reactions involving constructs N1B/C1, N2B/C2, and

N3B/C3, where further analyses by SDS-PAGE were needed,the reactions were quenched by addition of Gel Loading Buffer(5×, 300 mM Tris-HCl, 10% w/v SDS, 25% v/v glycerol, 0.5%w/v bromophenol blue, 200 mM β-mercaptoethanol, pH 6.8)to collected samples and heated to 95 °C for 5 min. Afterquenching, the reaction products were loaded onto 10% or 12%polyacrylamide gels followed by electrophoresis. StandardCoomaisse staining or Western Blotting techniques were used

Figure 5. Reconstitution of α-factor peptide by NTC in various mediawith growing yeast cultures. Addition of 1 μM of N3/C3 induced GFPexpression in either synthetic dropout medium (SD-2xSCAA) orenriched complex medium (YPD), albeit at slightly lower efficiencyrelative to in vitro reactions (Figure 4E). Each bar represents the highand low values obtained from two independent experimentsperformed on separate days.

ACS Synthetic Biology Letter

DOI: 10.1021/acssynbio.7b00078ACS Synth. Biol. 2017, 6, 1453−1460

1458

Page 7: Control of the Yeast Mating Pathway by Reconstitution of Functional … Synthetic Biology... · Control of the Yeast Mating Pathway by Reconstitution of Functional α‑Factor Using

to visualize the results. Densitometry analysis was performedusing ImageJ.Yeast Growth Arrest Assays Using Products of Intein-

Catalyzed Reactions. Yeast BY4741 bar1Δ cells (MATahis3Δ1 leu2Δ0 met15Δ0 ura3Δ0 bar1::G418R) were grownfrom single colonies in 3 mL YPD at 30 °C overnight.Overnight cultures were used to inoculate 25 mL of fresh YPDto an initial OD600 of ∼0.05 and grow to early exponentialphase (OD600 ∼ 0.2−0.25) at 30 °C. 900 μL of this culture wasinduced with 100 μL of reaction products, resulting in 10-folddilutions of final concentration of α-factor where products werepresent. The resultant mixtures were incubated at 30 °C for 16h to induce growth arrest. OD600 of these cultures were thenmeasured and normalized to the average of cultures with onlyreaction buffer added.Transcriptional Induction of GFP in Yeast. Yeast W303-

derived cells with GFP under the control of pFUS1 matingresponsive promoter (MATa far1Δmfa2::pFUS1-GFP bar1::-NatR his3 trp1 leu2 ura3, courtesy of Prof. Wendell Lim ofUCSF15) were grown and induced using the same proceduresas in the growth arrest assays described above. Instead of 16 hof induction, OD600 and fluorescence (Ex: 475 nm, Em: 515nm) of the induced cultures were measured using a microplatereader after 3 h of induction. The specific fluorescence weredetermined by dividing the fluorescence reading by OD600

measurements. Samples were also examined using flowcytometry. Collected samples were further pelleted, lysed bydisruption with glass beads, and analyzed with SDS PAGE andblotted against anti-GFP. For testing NTC reactions in livingcultures, 1 μM of purified proteins in Reaction Buffer wereadded as denoted in Figure 5 directly into the culture mediumwith growing yeast at OD600 ∼ 0.5. Specific GFP fluorescencewere determined using the same protocol as noted above 3 hafter the addition of the proteins.

■ ASSOCIATED CONTENT

*S Supporting InformationThe Supporting Information is available free of charge on theACS Publications website at DOI: 10.1021/acssynbio.7b00078.

Detailed proposed reaction schemes of PTS and NTC,Outline of yeast mating pathway, growth arrest assaysdata, SDS PAGE analyses of reaction for assessing extentof reaction and under different pH, proposed futureschemes, strains and constructs used in the study, andoligonucleotides used for construction of plasmids(PDF)

■ AUTHOR INFORMATION

Corresponding Author*E-mail: [email protected].

ORCID

Wilfred Chen: 0000-0002-6386-6958Author ContributionsK.-H. S. and W. C. conceived the biochemical framework,designed experimental assays, and wrote the manuscript. K.-H.S. performed the experiments.

NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTS

This work was supported by NSF (CBET1263719). We thankProf. Zhilei Chen at TAMU and Prof. Wendell Lim at UCSFfor their gracious gifts of a few noted plasmids and strains usedin this study.

■ REFERENCES(1) Lim, W. A. (2010) Designing customized cell signalling circuits.Nat. Rev. Mol. Cell Biol. 11, 393−403.(2) Venkatakrishnan, A. J., Deupi, X., Lebon, G., Tate, C. G.,Schertler, G. F., and Babu, M. M. (2013) Molecular signatures of G-protein-coupled receptors. Nature 494, 185−94.(3) Chen, M.-T., and Weiss, R. (2005) Artificial cell-cellcommunication in yeast Saccharomyces cerevisiae using signalingelements from Arabidopsis thaliana. Nat. Biotechnol. 23, 1551−5.(4) Collins, C. H., Leadbetter, J. R., and Arnold, F. H. (2006) Dualselection enhances the signaling specificity of a variant of the quorum-sensing transcriptional activator LuxR. Nat. Biotechnol. 24, 708−712.(5) Sarramegna, V., Muller, I., Milon, A., and Talmont, F. (2006)Recombinant G protein-coupled receptors from expression torenaturation: A challenge towards structure. Cell. Mol. Life Sci. 63,1149.(6) Lundstrom, K., Wagner, R., Reinhart, C., Desmyter, A.,Cherouati, N., Magnin, T., Zeder-Lutz, G., Courtot, M., Prual, C.,Andre, N., Hassaine, G., Michel, H., Cambillau, C., and Pattus, F.(2007) Structural genomics on membrane proteins: Comparison ofmore than 100 GPCRs in 3 expression systems. J. Struct. Funct.Genomics 7, 77−91.(7) Rozengurt, E., Sinnett-Smith, J., and Kisfalvi, K. (2010) Crosstalkbetween insulin/insulin-like growth factor-1 receptors and G protein-coupled receptor signaling systems: a novel target for the antidiabeticdrug metformin in pancreatic cancer. Clin. Cancer Res. 16, 2505−11.(8) Lappano, R., and Maggiolini, M. (2011) G protein-coupledreceptors: novel targets for drug discovery in cancer. Nat. Rev. DrugDiscovery 10, 47−60.(9) Muir, T. W. (2003) Semisynthesis of proteins by expressedprotein ligation. Annu. Rev. Biochem. 72, 249−89.(10) Amitai, G., Callahan, B. P., Stanger, M. J., Belfort, G., andBelfort, M. (2009) Modulation of intein activity by its neighboringextein substrates. Proc. Natl. Acad. Sci. U. S. A. 106, 11005−10.(11) Shah, N. H., Dann, G. P., Vila-Perello, M., Liu, Z., and Muir, T.W. (2012) Ultrafast protein splicing is common among cyanobacterialsplit inteins: implications for protein engineering. J. Am. Chem. Soc.134, 11338−41.(12) Naider, F., and Becker, J. (1986) Structure-Activity Relation-ships of the Yeast α-Factor. Crit. Rev. Biochem. 21, 225−248.(13) Naider, F., and Becker, J. M. (2004) The alpha-factor matingpheromone of Saccharomyces cerevisiae: a model for studying theinteraction of peptide hormones and G protein-coupled receptors.Peptides 25, 1441−63.(14) Park, S.-H., Zarrinpar, A., and Lim, W. A. (2003) Rewiring MAPkinase pathways using alternative scaffold assembly mechanisms.Science 299, 1061−4.(15) Youk, H., and Lim, W. A. (2014) Secreting and sensing thesame molecule allows cells to achieve versatile social behaviors. Science343, 1242782.(16) Manfredi, J. P., Klein, C., Herrero, J. J., Byrd, D. R., Trueheart, J.,Wiesler, W. T., Fowlkes, D. M., and Broach, J. R. (1996) Yeast alphamating factor structure-activity relationship derived from geneticallyselected peptide agonists and antagonists of Ste2p. Mol. Cell. Biol. 16,4700−9.(17) Ishii, J., Matsumura, S., Kimura, S., Tatematsu, K., Kuroda, S.,Fukuda, H., and Kondo, A. (2006) Quantitative and dynamic analysesof G protein-coupled receptor signaling in yeast using Fus1, enhancedgreen fluorescence protein (EGFP), and His3 fusion protein.Biotechnol. Prog. 22, 954−60.

ACS Synthetic Biology Letter

DOI: 10.1021/acssynbio.7b00078ACS Synth. Biol. 2017, 6, 1453−1460

1459

Page 8: Control of the Yeast Mating Pathway by Reconstitution of Functional … Synthetic Biology... · Control of the Yeast Mating Pathway by Reconstitution of Functional α‑Factor Using

(18) Zettler, J., Schutz, V., and Mootz, H. D. (2009) The naturallysplit Npu DnaE intein exhibits an extraordinarily high rate in theprotein trans-splicing reaction. FEBS Lett. 583, 909−14.(19) Shah, N. H., Eryilmaz, E., Cowburn, D., and Muir, T. W. (2013)Extein residues play an intimate role in the rate-limiting step of proteintrans-splicing. J. Am. Chem. Soc. 135, 5839−47.(20) Wood, D. W., Wu, W., Belfort, G., Derbyshire, V., and Belfort,M. (1999) A genetic system yields self-cleaving inteins forbioseparations. Nat. Biotechnol. 17, 889−92.(21) Mootz, H. D., Blum, E. S., Tyszkiewicz, A. B., and Muir, T. W.(2003) Conditional protein splicing: a new tool to control proteinstructure and function in vitro and in vivo. J. Am. Chem. Soc. 125,10561−9.(22) Ramirez, M., Guan, D., Ugaz, V., and Chen, Z. (2013) Intein-triggered artificial protein hydrogels that support the immobilization ofbioactive proteins. J. Am. Chem. Soc. 135, 5290−5293.

ACS Synthetic Biology Letter

DOI: 10.1021/acssynbio.7b00078ACS Synth. Biol. 2017, 6, 1453−1460

1460