Lighting up yeast cell factories by transcription factor ... · prokaryotes(Fernandez-Lopez´ etal.2015).Inprokaryotes,these one-component protein-based biosensors are able to bind

General rights Copyright and moral rights for the publications made accessible in the public portal are retained by the authors and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the legal requirements associated with these rights.

Users may download and print one copy of any publication from the public portal for the purpose of private study or research.

You may not further distribute the material or use it for any profit-making activity or commercial gain

You may freely distribute the URL identifying the publication in the public portal If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Downloaded from orbit.dtu.dk on: Apr 03, 2020

Lighting up yeast cell factories by transcription factor-based biosensors

D'ambrosio, Vasil; Jensen, Michael Krogh

Published in:FEMS Yeast Research

Link to article, DOI:10.1093/femsyr/fox076

Publication date:2017

Document VersionPublisher's PDF, also known as Version of record

Link back to DTU Orbit

Citation (APA):D'ambrosio, V., & Jensen, M. K. (2017). Lighting up yeast cell factories by transcription factor-based biosensors.FEMS Yeast Research, 17(7), [fox076]. https://doi.org/10.1093/femsyr/fox076

https://doi.org/10.1093/femsyr/fox076

https://orbit.dtu.dk/en/publications/cf3ece6c-7f2c-448d-9d55-0aa3b2f9fb90

https://doi.org/10.1093/femsyr/fox076

FEMS Yeast Research, 17, 2017, fox076

doi: 10.1093/femsyr/fox076Advance Access Publication Date: 13 September 2017Minireview

MINIREVIEW

Lighting up yeast cell factories by transcriptionfactor-based biosensorsVasil D’Ambrosio and Michael K. Jensen∗

Novo Nordisk Foundation Center for Biosustainability, Technical University of Denmark, 2800 Kgs. Lyngby,Denmark∗Corresponding author: Novo Nordisk Foundation Center for Biosustainability, Technical University of Denmark, Kogle Alle 6. 2970 Hørsholm, Denmark.Tel: +4561284850; E-mail: [email protected] sentence summary: Design, engineering and application of biosensors based on transcription factors in yeast.Editor: Zongbao Zhao

ABSTRACT

Our ability to rewire cellular metabolism for the sustainable production of chemicals, fuels and therapeutics based onmicrobial cell factories has advanced rapidly during the last two decades. Especially the speed and precision by whichmicrobial genomes can be engineered now allow for more advanced designs to be implemented and tested. However,compared to the methods developed for engineering cell factories, the methods developed for testing the performanceof newly engineered cell factories in high throughput are lagging far behind, which consequently impacts the overallbiomanufacturing process. For this purpose, there is a need to develop new techniques for screening and selection ofbest-performing cell factory designs in multiplex. Here we review the current status of the sourcing, design and engineeringof biosensors derived from allosterically regulated transcription factors applied to the biotechnology work-horse buddingyeast Saccharomyces cerevisiae. We conclude by providing a perspective on the most important challenges and opportunitieslying ahead in order to harness the full potential of biosensor development for increasing both the throughput of cellfactory development and robustness of overall bioprocesses.

Keywords: synthetic biology; transcription factor; allosteric regulation; transfer function

INTRODUCTION

Biotechnology has flourished in the last two decades as amanufacturing discipline for the future by engineering livingcells to convert renewable carbon sources into products relatedto health, food and transportation (Peralta-Yahya et al. 2012;Galanie et al. 2015; Zhou et al. 2016). Notwithstanding, there isenormous interest to further expand both the production port-folio and the productivity of the organismal hosts used for pro-duction, collectively referred to as cell factories (Van Dien 2013).The most commonly used organisms that have been estab-lished as cell factories include the gram-negative bacteria Es-cherichia coli and the budding yeast Saccharomyces cerevisiae. Forboth of these chassis, the accumulation of the genetic tools and

model-guided efforts derived from genome-scale models haveenabled the engineering of these chassis for the industrialproduction of fuels, platform chemicals and pharmaceuticals(Zhang, Jensen and Keasling 2015). Additionally, therapeuticproducts like opioids and penicillin have been refactored inmicrobial chassis (Galanie et al. 2015; Awan et al. 2017). Theseachievements highlight the versatility and importance of micro-bial biobased production.

Irrespective of host organisms, in order to replace currentpetrochemical pipelines for production of commodity and finechemicals with biobased production using renewable carbonsources, it is critical that cell factories can be constructed ina cost-effective manner and that the cell factories perform at

Received: 29 May 2017; Accepted: 12 September 2017C© FEMS 2017. This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License(http://creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided theoriginal work is properly cited. For commercial re-use, please contact [email protected]

1Downloaded from https://academic.oup.com/femsyr/article-abstract/17/7/fox076/4157790/Lighting-up-yeast-cell-factories-by-transcriptionby DTU Library useron 17 October 2017

http://www.oxfordjournals.org

mailto:[email protected]

http://creativecommons.org/licenses/by-nc/4.0/

mailto:[email protected]

2 FEMS Yeast Research, 2017, Vol. 17, No. 7

scale in relation to titres, rates and yields (Van Dien 2013).To support this, the ongoing sampling and parsing of biochemi-cal and omics data continuously improve the ability ofmetabolicengineers to rationally perturb endogenous metabolism forthe overproduction of metabolites, as well as the introductionof heterologous biosynthetic pathways for production of non-native compounds (Borodina and Nielsen 2014). Likewise, thedecrease in cost of DNA sequencing and synthesis as well asthe emergence of synthetic biology tools aims to support therational engineering of cell factories by focusing on parts char-acterisation and standardized methods which are host-agnosticand can be used broadly to speed the engineering and selectionof cell factory designs (Canton, Labno and Endy 2008). This in-cludes advances in diversity generation of cell factory designsaccommodated by genome engineering tools, such as the recentadvances in CRISPR-derived genome engineering technologies,which allows unprecedented speed and targetedmultiplexing ofstrain building procedures for an expanding number of chassis(Wang et al. 2009; Garst et al. 2016).

Despite these scientific advancements, there are still signifi-cant challenges to overcome in order to develop viable cell fac-tories that enable economically feasible bioprocesses. Given thesize ranges of genomes, numbers of metabolites and enzymaticreactions annotated for most commonly used production hosts,even targeting relatively narrow solution spaces in order to im-prove cell factory productivity, remains a daunting challenge bythe sheer number of individual cell lines and microbial strainsthat must be screened in order to identify the best performingcell factory (Rogers and Church 2016). While in some cases visi-ble phenotypes or growth-coupled production can support high-throughput screening of cell factory designs, testing cell fac-tory performance often relies on analytical methods like massspectrometry and chromatographic techniques for the identifi-cation and quantification of products of interest. Even for well-equipped research laboratories these methods have a maxi-mum throughput in the range of thousands of samples per day,thus creating a costly, laborious and ineffective route to iden-tify best-producing cells. As such there is a need to developmethods, which can enable testing of single cell factory designsin multiplex.

In recent years, the development, characterisation and appli-cation of metabolite biosensors from prokaryotes have spurredsignificant interest. Often used classes of biosensors rely onfluorescence resonance energy transfer, RNA-based aptamersor simple reporter assays (Dahl et al. 2013; Ravasio et al.2014; Strachan et al. 2014). Another class of biosensors ap-plied for metabolic engineering includes allosterically regulatedtranscription factors (aTFs) which are abundantly present inprokaryotes (Fernandez-Lopez et al. 2015). In prokaryotes, theseone-component protein-based biosensors are able to bind atarget metabolite, and upon binding provide a conformationalchange allowing a change in the expression from a target pro-moter. In native hosts, this results in the actuation of a cellularresponse in relation to changing environmental cues (i.e. car-bon source availability). Such a sensing-actuationmechanism isnaturally abundant in prokaryotic signalling pathways as bac-teria continuously exchange, receive and respond to both intra-and extracellular cues in order to coordinate cellular decisionmaking. One particularly well-studied aTF is the tetracyclin-responsive transcriptional repressor TetR (Gossen and Bujard1992). TetR has beenwidely adopted for controlling inducible ex-pression perturbations for almost three decades (Gossen et al.1995; Guet et al. 2002; Stanton et al. 2013). Other more re-cent examples of the successful development and applicationof TF-based biosensors from prokaryotes include the in situ

diagnosis of human gut microbiota, evolution-guided optimisa-tion of biosynthetic pathways in E. coli and synthetic cell–cellcommunication devices (Kotula et al. 2014; Raman et al. 2014;Chen et al. 2015).

Although most of the recently reported characterisationsand applications using these naturally occurring biosensorsbased on aTFs have focused on their transplantation into otherprokaryotic hosts, examples of on-boarding aTFs as biosensorsfor monitoring, selection and therapy in eukaryotic cells are ex-panding. In this review, we provide an overview of the aTFs,which have been successfully used in yeast. We put special at-tention to sourcing of aTFs, their molecular design and their ap-plications in yeast. This review also highlights the various en-gineering efforts, which have been performed to optimize theperformance of biosensors in order to make them applicable assmall-molecule biosensors. Finally, we provide a perspective onfuture directions for further proof of principles and industrialapplications using biosensors.

BIOSENSOR CHARACTERISATION ANDIDENTIFICATION

Characteristics of aTF-based biosensors

Before introducing the design and engineering of aTF-basedbiosensors, it is important first to introduce the main struc-tural components and the functional characteristics associatedwith aTF biosensor performance. The most frequent structuralcharacteristics of aTFs remain the DNA-binding domain (DBD)and the effector-binding domain (EBD), linked by a flexible loop-region, allowing for complex interdomain allostery in relation tobinding of either ligand and/or DNA (Werten et al. 2016). Also, itis important to highlight that monomeric aTFs exert a high de-gree of di- or oligomerisation (Fernandez-Lopez et al. 2015). Thestructure and propensity for oligomerisation together determinethe two main characteristics of aTF functionality, namely (i) theligand specificity and (ii) the transfer function describing the cor-relation between ligand concentration (input) and the biosensoroutput signal (Fig. 1a)(De Paepe et al. 2017). The specificity of abiosensor describes the strength of the binding between ligandand aTF or the difference in biosensor output in relation to a tar-get ligand compared to the output observed when introducingother potential ligands of interest. Inevitably, defining the speci-ficity of a biosensor is of primary importancewhen the biosensoris expected to provide high signal-to-noise ratios within a com-plex cellular environment in which the aTF is exposed to severalpotential ligands (De Paepe et al. 2017). The transfer function ofa biosensor describes its quantitative performance (Fig. 1a). Thisincludes the operational range, in which a significant change inbiosensor output can be measured in relation to ligand concen-tration. Likewise, the dynamic output range refers to the maxi-mum output, which can be obtained compared to the output ob-served in the absence of ligand binding. This is also referred to asthe ON and OFF states, respectively, of the biosensor (Moser et al.2013). From these parameters, the sensitivity of the biosensorcan be inferred as the difference in output of ON and OFF statesdivided by the difference in inducer concentration at the de-tection threshold and the maximum output, also referred to asthe Hill’s coefficient (Dietrich, McKee and Keasling 2010; Moseret al. 2013). Sensitivity is largely dependent on the cooperativ-ity between aTF monomer oligomerisation and ligand binding(Stefan and Le Novere 2013). It should also be mentioned thattransfer functions can have qualitative differences dependingon the mode of action of the biosensor, i.e. negative vs positivecorrelations between biosensor output and ligand concentration

Downloaded from https://academic.oup.com/femsyr/article-abstract/17/7/fox076/4157790/Lighting-up-yeast-cell-factories-by-transcriptionby DTU Library useron 17 October 2017

D’Ambrosio and Jensen 3

Figure 1. The biosensor transfer function and design. (a) The transfer function of a biosensor enables the quantitative characterisation of biosensors showing thecorrelation between input effector concentration and the biosensor output. Transfer functions thereby provide information on product sensitivity, the operationalrange of detection and the dynamic output range as defined by the difference between minimum (OFF) and maximum (ON) biosensor outputs. (b) Expression of

the allosterically regulated transcriptional repressor TrpRRep controls the expression of a gene of interest (GOI) by tryptophan-dependent binding to the promoterharbouring four TrpR binding sites (4xtrpO). To the right a schematic transfer function of TrpR output in relation to tryptophan concentration is illustrated. (c) Expressionof the allosterically regulated transcriptional activator BenMAct controls the expression of a GOI by cis,cis-muconic acid-dependent binding to the promoter harbouringthree BenM binding sites (3xbenO). To the right a schematic transfer function of BenM output in relation to cis,cis-muconic concentration is illustrated. AU; arbitrary

units.

(Fig. 1b and c) (see also section ‘Classes of aTF-based biosensors:repressors vs activators’ and ‘Engineering new specificities andimproving performances in existing aTFs’).

Knowledge about specificity and the transfer function is cru-cial in order to rationally engineer biosensors. In the remainderof the section ‘Biosensor characterisation and identification’, wereview the basic components of aTF-based biosensors sourcedfrom prokaryotes and the engineering strategies adopted for on-boarding these in yeast.

Classes of aTF-based biosensors: repressors vsactivators

Allosterically regulated TF-based biosensors can be dividedin two groups: transcriptional activators and transcrip-tional repressors. The aTF regulates the transcription ofthe reporter gene in a ligand-dependent manner by lo-calising on the operator site and sterically preventing thetranscription, or by supporting the accessibility of the



transcriptional machinery to the transcription start site(Fig. 1b and c).

The most prominent example of the repressing mechanismis the TetR repressor from E. coliwhich blocks the transcription ofthe tetA gene, encoding the tetracycline efflux pump (Hillen andBerens 1994; Møller et al. 2016). By binding directly to TetR, tetra-cycline is able to induce a conformational switch, which pre-vents DNAbinding and therefore allows the transcription of tetA.Subsequently, tetracycline is secreted from the cell and TetR isonce again able to bind to the operator site and block the tran-scription of tetA (Hillen and Berens 1994). This system has beensuccessfully transferred to mammalian and yeast cells in orderto control gene expression by tetracycline and analogs thereof(Gossen andBujard 1992; Garı et al. 1997). Anothermode of actionfor repressors is illustrated by tryptophan-binding TrpR fromE. coli (Arvidson, Bruce and Gunsalus 1986; Jeeves et al. 1999).Here, TrpR binds the co-repressing ligand Trp before bindingits cognate DNA operator site, and thereby regulates aromaticamino acid metabolism (Fig. 1b)(Zhao et al. 2015). However, co-repressors have so far not been adopted as biosensors in yeast.

In addition to repressors, recent studies have also high-lighted successful transfer of aTF-based transcriptional activa-tors into yeast. This was recently illustrated by Skjoedt et al.(2016), who employed BenM from Acinetobacter sp. ADP1, whichis able to induce the transcription synergistically upon the de-tection of benzoate and cis,cis-muconic acid (CCM) (Craven et al.2009). In addition to BenM, Skjoedt et al. were able to showthat several other activators from diverse bacterial species wereable to function as aTF-based biosensors in yeast, allowing invivo sensing of malonate, L-arginine and naringenin, in addi-tion to CCM (Fig. 1c, Table 1). Interestingly, none of the biosensordesigns required the expression of auxiliary transcription ma-chineries (e.g. sigma factors). As such, it is now possible to engi-neer both prokaryotic transcriptional activators and repressorsin eukaryotes, and the choice of aTF type ideally only relies onthe biosensor design and the availability of biosensors for theeffector of interest.

Biosensor prospecting

In the last three decades, a great number of interactions betweenTFs and genes have been discovered, especially from studies ofadaptation of microorganisms to different environmental con-ditions by sensing the presence/absence of metabolites, toxiccompounds or other chemicals (Junker, Kiewitz and Cook 1997;Brzostowicz et al. 2003). Such studies have allowed for the devel-opment of different predictive tools and databases, which can beused for identifying an aTF suitable for user-defined purposes.

In order to mine publicly available literature for new aTFs tobe tested and characterized for use as biosensors in yeast, amin-imum amount of information should be considered as a startingpoint, including if the gene sequence and the operator of the aTFare known. Also, the chemistry of the effector needs to be takeninto account: (i) is the effector toxic; (ii) can it be internalised at apH that is compatible with yeast growth during biosensor char-acterisation; (iii) does it need any complementary genes, such astransporters (Skjoedt et al. 2016).

In addition to literature mining for known aTFs, publicdatabases are available for browsing the genomes of a vast num-ber of prokaryotes (e.g. www.pseudomonas.com). Here the ap-proach to identify new aTFs can harness the fact that in prokary-otes, aTFs are often encoded divergently and in close genomicproximity of the genes, which they are regulating (Collier et al.1998). Additionally, as DBDs from major aTF families are often

conserved, this enables searching genome sequence databasesfor conserved regions of annotated genes (Maddocks andOyston 2008; Jain 2015). Ideally, in accordance with the divergentexpression of genes encoding the aTF and their target operon,the operator site can be inferred from mining the genomic se-quence upstream of the aTF expression cassette. Such amethodcan enable the search for any sequence that is homologous toknown aTF domains (DBD or EBD), which are encoded in prox-imity of the genes connected to the metabolic reactions of thecandidate ligand. Hence, by way of such in silico approach, newbiosensor candidates can be prospected.

Complementary to the in silico approach, expressiondatabases (e.g. https://www.ebi.ac.uk/arrayexpress/help/GEO data.html) can be used to search for differentially ex-pressed genes in relation to environmental cues or chemicalstimuli of interest, as both genes encoding aTFs and the genesof divergently expressed operons can enable the shortlisting ofboth candidate aTFs and potentially their operators (Strachanet al. 2014).

Alternatively, it is also possible to use different predic-tive tools and databases for transcriptional regulation in or-der to identify aTFs of interest. One example is RegulonDB(http://regulondb.ccg.unam.mx), which contains a list of TFs inE. coli with information on operator sequences, and the genesregulated by the TF (Gama-Castro et al. 2016). RegulonDB also al-lows querying for a known metabolite, gene or TF, and visualisethe connected regulatory systems. Another data mining tool,named SensiPath, allows searching sensing-enabling metabolicpathways (http://sensipath.micalis.fr/)(Delepine et al. 2016). Thistool enables the linking of a specificmetabolite, which otherwisehas no known target aTF, to a chemical that is amaximumof twoenzymatic steps away from the candidate compound which canbe detected by an annotated aTF. Beyond the identification of aproxy for high-throughput screening of yourmetabolite of inter-est, thismethod is also useful as amean to relieve potential toxi-city of yourmetabolite of interest, by identification of enzymaticsteps enabling the conversion of it to a non-toxic chemical.

In general, querying any of these databases can be very use-ful, but in order to identify robust starting points for biosensordesign based on aTF and operator sequences it is advisable toperform a combination of approaches, and search through sev-eral of the mentioned databases.

BIOSENSOR DESIGN ENGINEERING

Once an aTF has been selected, the next step is the design ofa functional transcriptional unit with an optimal transfer func-tion. To accomplish this, several factors must be considered. Inthis section, we review the basic principles of engineering aTFswith focus on designs in eukaryotes, especially yeasts.

aTF selection and aTF expression tuning

Selection of the aTF is a natural first step in the design of abiosensor. This can include the need for discriminating betweenseemingly close homologues and tuning the strength of the pro-moter controlling the expression of the aTF (Teo, Hee and Chang2013; Teo and Chang 2015; Wang, Li and Zhao 2016).

One example highlighting the importance of aTF selectionhas been reported by Teo et al. (2013). In this study, two vari-ants of FadR from E. coli and Vibrio cholera were tested in yeastin order to develop a fatty acid biosensor. While both repres-sors were unable to sufficiently repress the transcription ofthe reporter when cloned under the control of the weak CYC1


http://www.pseudomonas.com

https://www.ebi.ac.uk/arrayexpress/help/GEO_data.html

https://www.ebi.ac.uk/arrayexpress/help/GEO_data.html

http://regulondb.ccg.unam.mx

http://sensipath.micalis.fr/


Table

1.Ex

amplesof

effector

molec

ulesan

dth

eiraT

F-ba

sedbios

enso

rsen

ginee

redin

yeas

t.

Tran

scription

factor

nam

eNLS

Additional

mod

ules

Effector

Dyn

amic

range

(x-fold)

Operational

range

(M)

Referen

ce

FadR

NO

NO

Myristicac

id1.4

0.25

×10

−3Te

o,Hee

andChan

g(201

3)Fa

pR

SV40

NO

Malon

yl-C

oA2

0–13

.5×1

0−6(cerulenin)

Dav

id,N

ielsen

andSiew

ers(201

6)Fa

pR

SV40

NO

Malon

yl-C

oA4.17

0–35

.8×

10−6

(cerulenin)

Liet

al.(20

15)

MetJ

NO

B42

S- Aden

osylmethionine

∼70

Not

spec

ified

Umey

ama,

Oka

daan

dIto(201

3)

XylR

SV40

NO

Xylos

e∼4

26.6–6

66.1

×10

−3Te

oan

dChan

g(201

5)

Repressors

XylR

SV40

NO

Xylos

e14

0–33

.3×

10−3

Wan

g,Li

andZhao

(201

6)XylR

SV40

NO

Xylos

e6.3

0–13

3.3

×10

−3

XylR

SV40

SSN6

Xylos

e>25

1.33

–13.3

×10

−6Hec

toran

dMertens(201

7)Ada(N

-terminal

dom

ain)

NO

Gal4-AD

MeI—methyl

iodide

5.2x

±0.4

2.8

×10

−5–4

×10

−3Mos

eret

al.(20

13)

MMS

6.4x

±1.9

28–3

40×

10−6

DMS

6.9x

±2.0

2–15

0×

10−6

MNNG

3.8x

±0.2

2–64

×10

−6

Ben

MNO

NO

cis,cis-muco

nic

acid

100.2–

1.4

×10

−3Sk

joed

tet

al.(20

16)

Cam

RSV

403x

VP1

6ca

mphor

∼50

Not

spec

ified

Ikush

ima,

Zhao

andBoe

ke(201

5)Cym

RSV

40VP1

6p-C

umate

Not

spec

ified

Not

spec

ified

Ikush

imaan

dBoe

ke(201

7)Erg2

0NO

Gal4-DBD/G

al4-AD

IPP

∼2ND

Chou

andKea

sling(201

3)

Activators

FdeR

NO

NO

Naringe

nin

1.7

5×

10−8

–2

×10

−7Sk

joed

tet

al.(20

16)

Gal4(DBD)-DIG

-VP1

6NO

ND

Digox

in60

10−6

–∼50

0×

10−6

Fenget

al.(20

15)

Gal4(DBD)-PR

O-V

P16

pro

gesteron

e60

∼10−

6–1

0−5

hAR

NO

NO

5α-

Dihyd

rotestos

tero

ne

∼100

–500

∼10−

9–1

0−7

Bov

eeet

al.(20

07)

hAR

NO

NO

17β-T

estosteron

e∼1

00–5

00∼1

0−9–1

0−7

hAR

NO

NO

17β-B

olden

one

∼100

–500

∼10−

8–1

0−6

hAR

NO

NO

17β-Estradiol

∼300

∼10−

7–1

0−5

hER

-αNO

NO

17β-Estradiol

445

×10

−12–2

.8×

10−9

Sanse

verinoet

al.(20

05)

LexA

DBD-ER-

NO

DifferentADs

β-Estradiol

From

15to

60(dep

endingon

the

des

ign)

1–25

0×

10−9

(dep

ending

ondes

ign)

Ottoz

,Rudolfan

dStelling(201

4)

LexA

DBD-V

L/N

LS-V

H-

VP1

6SV

40ND

Bisphen

olA

38.4

±6.41

∼4.4

×10

−8–4

3.8

×10

−8Gionet

al.(20

09)

PhlF

NO

3xVP1

6DAPG

∼50

Not

spec

ified

Ikush

imaan

dBoe

ke(201

7)Te

tRNO

VP1

6Dox

ycyc

line

2000

∼10−

12–1

0−9

Garıetal.(19

97)

Zif26

8-hER

(mutated)-VP1

6NO

Zif26

8(DBD)/VP1

6(AD)

4′-4

′ -Dihyd

roxy

benzy

l∼5

0∼1

0−8–1

0−6

McIsa

acet

al.(20

14)



Figure 2. Engineering the biosensor transfer function of allosterically regulated transcription factors. (a) The transfer function of a reference biosensor. (b) The opera-tional range can be engineered by (i) changing the affinity of the interaction between the aTF and DNA, (ii) changing the number of operators bound by the cognate aTF,(iii) addition of an transcription activation domain and (iv) tuning the expression of the gene encoding the aTF. (c) The dynamic range of a biosensor can be engineeredby (i) mutating the aTF followed by selection, (ii) addition of transcription activation domains, (iii) changing the number of operators bound by the aTF and (iv) tuning

the expression of the gene encoding the aTF. (d) The effector specificity of an aTF towards new ligands can be changed by mutating the aTF followed by selection,(e) A logic gate with a positive correlation between input and output (YES) can be turned into a NOT gate by addition of an activation domain. In (a–e), all black linescorrespond to the transfer function of the reference biosensor, whereas red lines indicate transfer function of engineered biosensors.

promoter, expressing them under the control of the strong con-stitutive TEF1 promoter enabled repression of the reporter genecontrolled by each of the two FadR variant, albeit with significantdifference in the level of repression. While strong expression ofthe E. coli FadR enabled repression of the reporter fluorescenceby almost 90%, FadR from V. cholerae was only able to reduce ex-pression by 50% compared to a reference strain without FapR.As such, the promoter driving the expression of the aTF canimpact the OFF state of the biosensor. Additionally, the opera-tional range can be influenced by the choice of promoter drivingthe expression of the gene encoding the aTF. In fact, a weakerpromoter controlling the aTF was shown to enable a change inthe operational range, with lower expression driving the oper-ational range towards lower concentrations of the effector andstronger promoter driving the operational range towards highereffector concentrations (Teo, Hee and Chang 2013) (Fig. 2a andb). Still, strong expression is generally the first choice for con-trolling the expression of genes encoding repressor-based aTFsin order to support lower OFF states (Teo, Hee and Chang 2013; Liet al. 2015; David, Nielsen and Siewers 2016; Hector and Mertens2017) Complementary to this, when transcriptional activatorsare employed, a weaker promoter driving the expression of theaTF may be preferred since a strong promoter can lead to a highOFF state expression of the reporter gene or prolonged lag phaseduring growth (Teo and Chang 2015; Skjoedt et al. 2016). As such,the promoter controlling the aTF should be chosen carefully inorder to ensure control of both OFF state, operational range andpotential growth effects.

Engineering the reporter promoter

The architecture of the promoter controlling the output of thebiosensor is another essential engineering target. While nativepromoters are often well characterised and therefore can be ad-vantageous to use for control over biosensor output over non-native ones, it is important to take into consideration the na-

tive regulation of the promoter. Several groups have engineerednative yeast promoters for the design of biosensors. A com-mon strategy has been to insert aTF operator sites (see sec-tion ‘Operator sequence’) into a native yeast promoter chassisin which upstream activation sequences or upstream repressionsequences have been deletedwhenever present, in order tomin-imise the complexity of native regulation (Ikushima, Zhao andBoeke 2015; Li et al. 2015; Skjoedt et al. 2016; Wang, Li and Zhao2016; Ikushima and Boeke 2017). In the example by Skjoedt et al.(2016) engineering of the promoter driving the reporter gene in-cluded testing of four different ‘crippled’ variants of the CYC1promoter and identified the shortest variant (209 bp) to be theone allowing for the lowest OFF state, while still maintainingapproximately 4-fold dynamic output range in the presence ofCCM. In general, the reporter promoter should be strong enoughto allow for a detectable output but, overexpressing the reportergene can cause ‘leaky’ expression, which is of particular im-portance when the biosensor is to be coupled to selection. Inaddition to controlling ON and OFF states, when carefully en-gineered, the reporter promoter design can also tune both thedynamic output range and the operational range (Fig. 2b andc)(McIsaac et al. 2014; Skjoedt et al. 2016).

Operator sequence

The operator is a DNA sequence that is recognised by the DBD ofthe aTF. Its presence is essential for aTF binding to DNA, and itimpacts both the dynamic and operational range. Adding opera-tor sequences to a native or truncated promoter intended to con-trol the biosensor output has generally proved to have a negativeimpact on the basal expression (Li et al. 2015; David, Nielsen andSiewers 2016; Skjoedt et al. 2016; Hector and Mertens 2017). Like-wise, the position of the operator is important, and due to thedifficulty predicting the effect operator positioning will have onthe reporter gene, several groups have mined different designsin order to find the best positions. In one example, Hector and



Mertens (2017) tested three different operator designs: (i) one op-erator sequence immediately downstream of the TATA element,(ii) two operators after the TATA element and (iii) two opera-tors flanking the TATA sequence. The addition of the first oper-ator immediately downstream of the TATA sequence decreasedthe promoter basal activity by almost 50%, and allowed for a 4-fold de-repression when the effector was added. The additionof a second operator downstream of the TATA sequence causedan even further decrease in the promoter activity by strongerrepression. Finally, the best performing design, two operatorsflanking the TATA element, had a basal expression comparableto the promoter with only one operator, but allowed for approx-imately 8-fold de-repression. This is a clear example of how thesame number of operators can have significantly different out-puts when positioned differently. However, the position of op-erator(s) is not the only design parameter impacting reporterpromoter output. Wang et al. (2016) showed that different opera-tor sequences can also perturb the operational range of biosen-sors. This was particularly exemplified by exchanging the op-erator site of XylR from Staphylococcus xylosus with the operatorsequence of the homologous from Bacillus subtilis. Swapping be-tween these two operators changed the operational range from0 to 20 g L−1 of xylose, to be saturated at <5 g L−1 (Table 1)(Wang,Li and Zhao 2016). By designing a degenerate operator sequence,the authors identified several operational ranges, and were alsoable to achieve an increase in dynamic range. This behaviourbrings new importance to the interaction between the aTF DBDand the operator of choice, and provides additional means todeveloping biosensors with sought-for transfer function charac-teristics.

Nuclear localisation signal

The transport of molecules between the nucleus and cytoplasmis made possible by the nuclear pore complex (NPC). NPCs func-tion as canals and allow free diffusion of molecules and smallproteins in and out of the nucleus. Even though the NPCs arelarge enough to allow free diffusion of macromolecules smallerthan 50 kDa, they function as a barrier by preventing the diffu-sion of larger proteins to the nucleus, unless a nuclear locali-sation signal (NLS) is encoded in their primary structure (Hahnet al. 2008). A common choice of NLS to support nuclear local-isation of aTFs is the simian virus 40 (SV40) NLS (Feng et al.2015; Ikushima, Zhao and Boeke 2015; Li et al. 2015; Garst et al.2016; Hector and Mertens 2017; Ikushima and Boeke 2017), andit has been shown that the absence of the NLS can lead to anon-functional biosensor (Li et al. 2015). Specifically, Li and co-workers compared the ability of FapR from B. subtilis to repressthe transcription of the reporter gene when expressed in pres-ence or absence of the NLS and showed that the strain harbour-ing FapR with NLS showed 30% of the fluorescence intensity ofthe control strain not expressing FapR, while the strain harbour-ing the repressor without the NLS showed fluorescence compa-rable to the strain without the repressor. Yet, the presence ofthe NLS does not seem strictly necessary when employing aTFs,which have a molecular mass lower or comparable to the nu-clear pore diffusion limit (Teo, Hee and Chang 2013; Ikushima,Zhao and Boeke 2015; Skjoedt et al. 2016), but NLS should indeedbe considered a design parameter when engineering aTF-basedbiosensors in yeast.

Module transferability

The simple road-block of RNA polymerase progression pro-vided by DNA-bound aTF repressors in the absence of effectors

allows engineering of transcriptional activation by addition ofone or several activator domains. Here, the strategy is to addsingle or multiple VP16 activation domain(s) from herpes sim-plex virus to the repressor (Gion et al. 2009; Ottoz, Rudolf andStelling 2014; Feng et al. 2015; Ikushima, Zhao and Boeke 2015;Ikushima andBoeke 2017) (Table 1). In the seminal study onTetR,a TetR-dependent expression system was developed in mam-malian cells in which the aTF was converted from a repressor toan activator by fusing a VP16 activator domain on the C-terminalof TetR (Gossen and Bujard 1992) (Fig. 2e). This system, referredto as Tet-Off, relies on the ability of the TF to bind a syntheticpromoter, where seven operator sequenceswere linked to amin-imal promoter fragment derived from the CMV promoter, andthereby induce the transcription of the reporter gene in the ab-sence of effectors. The activity of the aTF can then be relievedby tetracyclin or doxycycline administration (Loew et al. 2010).Another variant, referred to as Tet-On, where transcription ofthe reporter by TetR fused to VP16 is only achieved in the pres-ence of tetracycline or doxycycline, was later developed throughrandommutagenesis (Gossen et al. 1995) (Fig. 2e). Both these sys-temswere employed in yeast in order to control gene expressionwithout perturbing the cellular metabolism, as in the case ofgalactose or methionine addition (Bellı et al. 1998; Urlinger et al.2000; Giuraniuc, MacPherson and Saka 2013; Roney et al. 2016).

The activator domain is not the only module that can beadded to aTFs. As shown by several groups, both DBDs andwhole proteins can be added (Chou and Keasling 2013; Moseret al. 2013). Moser et al. (2013), while developing a sensor forstrong methylating compounds, exploited the characteristic ofthe Ada protein from E. coli to detect methyl groups on DNA andto transfer them to a cysteine residue on its amino acidic se-quence, which then leads to its activation as an aTF in yeast. Toaccomplish this, they fused the N-terminal domain of the Adaprotein, which is the region that allows for both DNA bindingand methyltransferase activity, to the Gal4 transactivation do-main (Gal4-AD) in order to induce transcriptional activation inyeast upon detection of methyl groups (Moser et al. 2013).

In another study, Feng et al. (2015) fused the DBD of Gal4to a synthetic EBD and to a VP16 module to create sensors fordigoxin,which they later rationally engineered to induce a speci-ficity switch towards progesterone (Fig. 2d). Both sensors couldinduce the transcription of the reporter gene by 60-fold upon thedetection of the ligand (Table 1).

Finally, another hybrid biosensor design has been reported byChou and Keasling (2013) illustrating the flexibility supported bythe modular transferability of TF protein domains. Here, the au-thors exploited the ability of an isopentenyl diphosphate (IPP)isomerase (idi) to dimerise in order to create an IPP sensor inyeast (Chou and Keasling 2013). To do so, they first created twochimeric proteins: the first one consisted of the Gal4 DBD fusedto the IPP isomerase, while the second chimeric protein was IPPisomerase fused to the Gal4 AD. The sensor then relied on theability of the IPP isomerase dimer to bring the two Gal4 domainsinto close proximity to activate transcription of a reporter genecontrolled by a GAL promoter. After validating the sensor, theyreplaced the IPP isomerasemodulewith IPP-binding Erg20 to fur-ther improve the dynamic range of the IPP biosensor in yeast(Chou and Keasling 2013).

In summary, the addition of new modules can allow for thecreation of new biosensors when naturally occurring ones fora specific target are unknown. By adding a DBD to a syntheti-cally designed EBD, or by adding an AD, it is possible to inducea stronger response upon the detection of the ligand, since thede-repression level allowed by repressor-based aTFs is generally



lower than the induction of aTFs fused with activation domains(Table 1). Also, by testing different modules, such as differentADs or EBDs, it is possible to improve the sensor’s response,leading to an improved dynamic range (Chou and Keasling 2013;Ottoz, Rudolf and Stelling 2014) (Fig. 2c).

Engineering new specificities and improvingperformances in existing aTFs

Many aTFs have been identified based on traditional forward en-gineering (Shamanna and Sanderson 1979; Neidle, Hartnett andOrnston 1989). However, it still remains a challenge to infer lig-and specificity of an aTF from such studies. Even after exhaus-tive database mining, it may still not be possible to correlatean aTF to a specific compound of interest (see section ‘Biosen-sor prospecting’). In such cases, it may be necessary to engineera preexisting aTF with new ligand specificity of interest. How-ever, one of the challenges researchers are facing when aimingto engineer new specificities into existing aTF-based biosensorsis the need to simultaneously maintain critical properties likeDNA binding, allosteric signal transduction, dimerisation andgeneral protein structure, as all of these functionalities are cru-cial for ligand-induced transcriptional regulation (Raman et al.2014). However, learning from nature’s rich diversity of aTFs,several studies have shown that both rational and directed evo-lution approaches can be used to successfully engineer new lig-and specificities into existing aTF-based biosensors (Fig. 2d).

As for directed evolution, several examples have been re-ported in bacteria using random mutagenesis based on error-prone PCR (epPCR) coupled with ligand-specific selection in or-der to affinity-mature aTFs for new ligand specificities (Cebolla,Sousa and De Lorenzo 1997; Collins, Arnold and Leadbetter 2005;Taylor et al. 2015; Xiong et al. 2017). While most of these exam-ples refer to studies in bacteria, a few studies highlight the powerof directed evolution for improving the biosensor performanceof aTFs when engineer in yeast. In relation to ligand specificity,Chockalingam et al. (2005) adopted directed evolution to alterthe specificity of the human estrogen receptor alpha (hER-alpha)from 17-beta estradiol to 4,4′-dihydrobenzyl (DHB) by site sat-uration mutagenesis on ligand contacting residues, and epPCRon the whole receptor. Library variants with improved responseto DHB relative to parental hER-alpha were selected based ongrowth of the host yeast cells in medium containing an appro-priate concentration of DHB. Taken together, these studies havedemonstrated that mutating residues involved in the bindingsite can change ligand specificity without affecting allostericfunction. However, it should be noted that the specific residuesthat give rise to allostery are generally unknown and do not tendto be localised in certain parts of the protein (Suel et al. 2003).Furthermore, mutations in EBD tend to alter allosteric commu-nication with the DBD and mutations can have far-reaching ef-fects (Raman et al. 2014; Taylor et al. 2015).

In addition to studies related to changes in specificity, di-rected evolution can also be used to improve the transfer func-tion of an existing aTF. By increasing aTFs specificity and sen-sitivity towards the ligand and altering the binding affinity tothe operator sequence, it is possible to modify most of thecharacteristics of the transfer function exemplified in Fig. 2.For instance, Skjoedt et al. (2016) showed that directed evolu-tion on BenM’s EBD via epPCR, yeast plasmid gap repair andfluorescence-activated cell sorting (FACS) enabled the identifi-cation of aTF variants with >10-fold increase in reporter geneoutput upon the presence of CCM, while wild-type BenM was

only able to induce 4-fold (Fig. 2c). In another example, yet per-formed in bacteria, Richards et al. (2017) recently showed thatby random mutagenesis it is possible to induce a YES/NOT logicgate inversion in a LacI variant which was previously made in-sensitive to the ligand by removing the allosteric communica-tion between the EBD and the DBD (Fig. 2e). These variants, gen-erated via epPCR and transformed in E. coli, also show differentresponses to increasing concentrations of the effector and dif-ferent operational ranges.

In addition to random mutagenesis, computational ap-proaches are also being developed and have been employed tocreate sensors with new specificities. An example that clearlystates the potential of this method comes from De Los Santoset al. (2016), where the authors were able to induce a vanillinresponse into QacR from S. aureus by computational design, anaTF that is induced by a broad range of structurally dissimilarcompounds. Similarly, Taylor et al. (2015) adopted both randomepPCR-based and Rosetta-guided mutagenesis to change ligandspecificity of LacI.

To summarise, by using a combination of computational,random and site saturationmutagenesis methods, coupled withadequate selection, it is possible to create sensorswith improvedor novel characteristics, and therefore allow a broadening of therange of applications where aTFs can be employed.

APPLICATIONS OF TF-BASED BIOSENSORS INYEAST

TF-based biosensors described in section ‘Biosensor design en-gineering’ have been implemented in yeast for different appli-cations that range from the detection of pollutants to actuatorson native yeast metabolism to improve overall cell factory per-formances. In this section, we review applications of aTF biosen-sors in yeast, while a list of aTF biosensors currently known tohave been employed in yeast is found in Table 1.

aTF biosensor-reporter systems in yeast

Industrially relevant organisms like Sa. cerevisiae are able tometabolise glucose, which is the most abundant sugar in thelignocellulosic biomass. However, in order to bring down costof biobased production from yeast, it is important to improve itsfeedstock utilisation. In addition to glucose, xylose represents animportant fraction of available sugars in lignocellulosic biomass(Hector and Mertens 2017). In order to create a sensor that couldallow for protein expression or gene circuits development in-duced by xylose, various groups have developed xylose biosen-sors based on XylR, a xylose-responsive transcriptional regula-tor found in bacteria (Table 1) (Teo and Chang 2015;Wang, Li andZhao 2016; Hector and Mertens 2017).

Wang et al. (2016) developed a xylose-dependent biosensorbased on S. xylosus XylR, which is characterised by a broad op-erational range (Table 1). To demonstrate the utility of the sen-sor, the authors used directed evolution to improve the xylosetransportation capacity of hexose transporter HXT14 from Sa.cerevisiae. Here, a mutant library of the gene encoding HXT14was created via epPCR and the library was introduced in thestrain already containing the xylose biosensor. After the addi-tion of xylose, cells with higher fluorescence were selected byFACS, thereby allowing the identification of a xylose transporterwith a 6.5-fold higher transportation capacity (Fig. 3a).

Another example of biosensor used as a screening systemcomes from Li et al. (2015) where authors used an aTF-based



Figure 3. Applications of aTF-based biosensors in yeast. (a) Biosensors based on aTFs have allowed the screening of cDNA libraries, protein variants and combinatorial

pathway designs in high throughput, by coupling accumulation of carbon sources and metabolites with expression of reporters, selectors or actuators. Such effortshave allowed the screening of cell factories with improved carbon source uptake and productivity of various chemicals. (b) Environmental biosensing using aTF-basedbiosensors in yeast has been applied for monitoring of hazardous methylating and estrogenic compounds.

biosensor to screen a genome-wide overexpression cDNA libraryto identify enzymes that could improve 3-hydroxypropionic acid(3-HP)(Fig. 3a). The sensor is based on FapR (Table 1) from B. sub-tilis, a TF that is able to detect malonyl-CoA, a key intermediatefor the biosynthesis of several industrially relevant compoundssuch as 3-HP, an attractive value-added chemical. This approachallowed them to identify two enzymes, PMP1 and TPI1, whichlead to higher GFP de-repression. The first gene encodes for aplasma membrane that regulates a plasma membrane protonATPase, while the latter encodes a triose phosphate isomerase.The overexpression of these two genes singularly lead to a 116%and 120% increase in 3-HP production respectively, compared tothe control strain. Interestingly, the overexpression of both en-zymes simultaneously led to lower titres, comparable to the con-trol strain (Li et al. 2015).

More recently, BenM from Acinetobacter sp. ADP1 and FdeRfrom Herbaspirillum seropedicae (Table 1) were used to screen cellfactories in order to identify best-performing pathway designsfor naringenin and CCM, the latter being an important precur-sor for bioplastic production (Curran et al. 2013; Skjoedt et al.2016). Subsequently, the production strains were tested by HPLCand flow cytometry. Here, the authors observed a strong correla-tion (r = 0.98) between the fluorescence intensity and biobasedproduction of CCM and narigenin, demonstrating the abilityof these biosensor as orthogonal screening systems for high-producing strains (Fig. 3a)(Skjoedt et al. 2016).

Biosensors can be employed outside the field of cell factorydesign, and can be used as tools to screen for environmentalcontaminants found in water and soil. One of the earliest ex-amples of applying aTFs as environmental screening tools isfounded on the human estrogen receptor alpha (Sanseverinoet al. 2005). This biosensor allowed the detection of chemicalswith estrogenic activity, which are able to damage the endocrinesystem in vertebrates. To detect those chemicals Sanseverinoet al. (2005) developed a biosensor based on hER-alpha and thebacterial lux cassette (luxCDABE)(Table 1), which allowed for fastand easy detection of estrogenic compounds within hours ofsampling (Fig. 3b). Apart from the almost real-time monitoringpotential, this system allows for, there are still different limita-tions, the most important of which is that yeast cells responddifferently when subjected to different estrogenic moleculesand may not be able to detect specific estrogenic compounds(Sanseverino et al. 2005).

Another class of toxic molecules that are common both inindustry and in agriculture is represented bymethylating chem-icals, and in order to detect the presence of these compounds inthe environment, Moser et al. (2013) developed a system that re-lies on the E. coli Ada protein to be directly methylated and/or todetect DNA methylation, and to transduce a quantifiable signalvia a fluorescent output. This sensor is particularly interestingbecause it is able to detect methyl iodide (MeI) which is used inagriculture as a controversial fumigant. To validate the sensor,the authors added an aqueous solution of MeI to a soil sampleand monitored fluorescence as a proxy for MeI levels over timeusing the biosensor (Fig. 3b)(Moser et al. 2013)

aTF biosensor selection in yeast

Biosensors as a tool to screen libraries based on fluorescencehave proved very useful. However, screening large librariesbased on their fluorescence levels requires expensive instru-mentation and, in order to screen for even larger libraries, thecoupling of the biosensor to a selection assay has proved veryuseful (Dietrich, McKee and Keasling 2010). Two such examplescome fromUmeyama, Okada and Ito (2013) and Feng et al. (2015).The first example relies on the MetJ repressor fused to a B42 ac-tivation domain to induce the transcription of HIS3 LEU2 andVenus genes in the presence of S-adenosylmethionine (SAM).Additionally, the system is repressed by the Dox-responsive re-pressor TetR to allow for further tuning of the sensor-selectorsystem (Fig 3a). By this dual-input system, GAL11 was identi-fied as an enhancer of SAM accumulation. In the second exam-ple, the authors developed a systemwhere a hybrid synthetic TFbased on the DBD of Gal4, an evolved progesterone-responsiveEBD and a VP16 module was used to control progesterone-dependent expression of HIS3 (Fig. 3a). This is particularly in-teresting because yeast-based platforms for the biosynthesisprogesterone have already been developed (Duport et al. 1998).Progesterone is synthesised by the conversion of pregnenoloneto progesterone by the enzyme 3β-hydroxysteroid dehydroge-nase (3β-HSD). To develop a selection assay, they added a His3-inhibitor to plates supplemented with pregnenolone, so that thecells harbouring the wild-type 3β-HSD were not able to pro-duce enough progesterone to complement the histidine aux-otrophy. Next, they mutagenised the 3β-HSD by epPCR and se-lected clones that could grow on the selective medium. Two of



the surviving clones, harbouring a single amino acid mutation,were analysed for progesterone production by gas chromatog-raphy and mass spectroscopy. Both the clones were able to pro-duce twice asmuch progesterone compared to the cells with thewild-type enzyme (Feng et al. 2015).

aTF biosensor actuation in yeast

Apart from providing a detectable output, such as fluorescenceor auxotrophy complementation, biosensors can be directlyplugged into the metabolism by cloning an enzyme under thecontrol of the aTF. A clear example of this method comes fromDavid et al. (2016). Here FapR from B. subtiliswas used to dynam-ically control the flux at the malonyl-CoA node and direct theprecursors towards 3-HP production instead of producing fattyacids. Since 3-HP can be produced fromMalonyl-CoA by one en-zymatic step via a Malonyl-CoA reductase (MCRCa) from Chlo-roflexus aurantiacus, by putting the expression of this enzymeunder the control of FapR, the authors were able to create a self-regulatory system that increased the expression of MCRCa at alevel related to Malonyl-CoA availability (Fig. 3a). The combina-tion of this approach with the development of a two-stage dy-namic system where 3-HP is produced under glucose limitingconditions led ultimately to a 10-fold increase in 3-HP produc-tivity, reaching a final titre of 1 g L−1(David et al. 2016). The abil-ity to control the flux at different nodes as shown in this studycan be extremely useful to prevent metabolic imbalances and toincrease productivity, as also evidenced from synthetic sensor–actuator systems in bacteria (Kobayashi et al. 2004; Xu et al. 2014).To date, however, the use of FapR is the only example of dynamicregulation of metabolism reported in yeast.

CONCLUSIONS AND PERSPECTIVES

In order to improve the applicability of biosensors, it is criticalto expand the chemical space that biosensors are able to de-tect, or can be engineered to detect (Delepine et al. 2016). Fromthe studies performed on biosensor development, it is clear thatachieving an optimal tuning between the elements composingthe biosensor can indeed be performed by both rational andrandomised approaches, as demonstrated using targeted en-gineering of modular domains and directed evolution, respec-tively (Fig. 2). However, forward engineering of aTFs with novelspecificities and enhanced transfer functions is generally notfeasible to date. In terms of transfer function, the goal is to gainenough understanding to design a functional biosensor by cor-relating the level of expression of the aTF to the synthetic pro-moter, containing a specific operator design, and operator vari-ants that have different strengths, controlling the reporter gene,allowing for both relevant dynamic and operational ranges. Sim-ilarly, this also applies to engineering of new specificities at theaTF level. Here a deeper understanding of the structure to func-tion knowledge is needed as not only operational and dynamicranges depend on the aTF protein structure but also the speci-ficity. Gathering more data focusing on the structure to func-tion relationships of allostery is a prominent path towards in-creasing our understanding of aTF functionality and the under-lying design principles (Taylor et al. 2015; Richards, Meyer andWilson 2017). Emerging and future studies based on labora-tory evolution coupled with next-generation sequencing are ex-pected to enable the build-up of data-sets supporting the con-struction of machine-learning algorithms with predictive poweron aTF biosensor designs in relation to both transfer functionand ligand specificity.

In terms of application, it is not hard to imagine that in thenear future, biosensors will allow for complex gene circuit regu-lations by aTFs acting as core unit actuators of yeast cell facto-ries. Here, dynamic shunting of competing pathway fluxes, de-coupling of growth and production phases, and improved overallrobustness of cell factories performing under industrial biopro-cesses are but a few examples for which carefully engineeredaTFs can be applied. Indeed, even though dynamic regulation ofmetabolic pathways for improved biobased production in E. colihas been reported by several groups (Dahl et al. 2013; Xiao et al.2016), the recent study byDavid, Nielsen and Siewers (2016) on 3-HP production in yeast is expected to lead further developmentof synthetic control over cell metabolism by the use of product-activated biosensors.

Overall, inspired by nature, it is expected that further devel-opment in the field of synthetic biology will allow for more com-plex aTF-controlled regulations in order to improve biobasedproduction of both monoculture and mixed cultures, ultimatelyenabling higher throughput of the iterative design-build-test cy-cle and overall cost-reduced biomanufacturing processes.

FUNDING

This work was supported by the Novo Nordisk Foundation andby the European Union’s Horizon 2020 research and innovationprogramme under the Marie Sklodowska-Curie action PAcMEN(grant agreement No. 722287).

Conflict of interest. None to declared.

REFERENCES

Arvidson DN, Bruce C, Gunsalus RP. Interaction of the Escherichiacoli trp aporepressor with its ligand, L-tryptophan. J Biol Chem1986;261:238–43.

Awan AR, Blount BA, Bell DJ et al. Biosynthesis of the antibioticnonribosomal peptide penicillin in baker’s yeast. Nat Com-mun 2017;8:15202.

Bellı G, Garı E, Piedrafita L et al. An activator/repressor dual sys-tem allows tight tetracycline-regulated gene expression inbudding yeast. Nucleic Acids Res 1998;26:942–947.

Borodina I, Nielsen J. Advances inmetabolic engineering of yeastSaccharomyces cerevisiae for production of chemicals. Biotech-nol J 2014;9:609–20.

Bovee TFH, Helsdingen RJR, Hamers ARM et al. A new highlyspecific and robust yeast androgen bioassay for the detec-tion of agonists and antagonists. Anal Bioanal Chem 2007;389:1549–58.

Brzostowicz PC, Reams AB, Clark TJ et al. Transcriptionalcross-regulation of the catechol and protocatechuatebranches of the beta-ketoadipate pathway contributes tocarbon source-dependent expression of the Acinetobac-ter sp. strain ADP1 pobA gene. Appl Environ Microb 2003;69:1598–606.

Canton B, Labno A, Endy D. Refinement and standardizationof synthetic biological parts and devices. Nat Biotechnol2008;26:787–93.

Cebolla A, Sousa C, De Lorenzo V. Effector specificity mutants ofthe transcriptional activator NahR of naphthalene degradingPseudomonas define protein sites involved in binding of aro-matic inducers. J Biol Chem 1997;272:3986–92.

Chen Y, Kim JK, Hirning AJ et al. Emergent genetic oscillationsin a synthetic microbial consortium. Science (80-) 2015;349:986–9.



Chockalingam K, Chen Z, Katzenellenbogen JA et al. Directedevolution of specific receptor–ligand pairs for use in thecreation of gene switches. P Natl Acad Sci USA 2005;102:5691–6.

Chou HH, Keasling JD. Programming adaptive control to evolveincreased metabolite production. Nat Commun 2013;4:2595.

Collier LS, Gaines GL, Neidle EL et al. Regulation of benzoatedegradation in Acinetobacter sp. strain ADP1 by BenM, a LysR-type transcriptional activator. J Bacteriol 1998;180:2493–501.

Collins CH, Arnold FH, Leadbetter JR. Directed evolution of Vibriofischeri LuxR for increased sensitivity to a broad spectrum ofacyl-homoserine lactones. Mol Microbiol 2005;55:712–23.

Craven SH, Ezezika OC, Haddad S et al. Inducer responses ofBenM, a LysR-type transcriptional regulator from Acinetobac-ter baylyi ADP1. Mol Microbiol 2009;72:881–94.

Curran KA, Leavitt JM, Karim AS et al. Metabolic engineering ofmuconic acid production in Saccharomyces cerevisiae. MetabEng 2013;15:55–66.

Dahl RH, Zhang F, Alonso-Gutierrez J et al. Engineering dynamicpathway regulation using stress-response promoters. NatBiotechnol 2013;31:1039–46.

David F, Nielsen J, Siewers V. Flux control at the malonyl-CoAnode through hierarchical dynamic pathway regulation inSaccharomyces cerevisiae. ACS Synth Biol 2016;5:224–33.

Delepine B, Libis V, Carbonell P et al. SensiPath: computer-aideddesign of sensing-enablingmetabolic pathways.Nucleic AcidsRes 2016;44:W226–31.

Dietrich JA, McKee AE, Keasling JD. High-throughput metabolicengineering: advances in small-molecule screening and se-lection. Annu Rev Biochem 2010;79:563–90.

De Los Santos ELC, Meyerowitz JT, Mayo SL et al. Engineeringtranscriptional regulator effector specificity using computa-tional design and in vitro rapid prototyping: developing avanillin sensor. ACS Synth Biol 2016;5:287–95.

De Paepe B, Peters G, Coussement P et al. Tailor-made transcrip-tional biosensors for optimizing microbial cell factories. J IndMicrobiol Biot 2017;44:623–45.

Duport C, Spagnoli R, Degryse E et al. Self-sufficient biosynthesisof pregnenolone and progesterone in engineered yeast. NatBiotechnol 1998;16:186–9.

Feng J, Jester BW, Tinberg CE et al.A general strategy to constructsmall molecule biosensors in eukaryotes. Elife 2015;4:e10606.

Fernandez-Lopez R, Ruiz R, de la Cruz F et al. Transcriptionfactor-based biosensors enlightened by the analyte. Front Mi-crobiol 2015;6:648.

Galanie S, Thodey K, Trenchard IJ et al. Complete biosynthesis ofopioids in yeast. Science (80-) 2015;349:1095–100.

Gama-Castro S, Salgado H, Santos-Zavaleta A et al. RegulonDBversion 9.0: high-level integration of gene regulation, co-expression, motif clustering and beyond. Nucleic Acids Res2016;44:D133–43.

Garı E, Piedrafita L, Aldea M et al. A set of vectors witha tetracycline-regulatable promoter system for modu-lated gene expression in Saccharomyces cerevisiae. Yeast1997;13:837–48.

Garst AD, Bassalo MC, Pines G et al. Genome-wide map-ping of mutations at single-nucleotide resolution for pro-tein, metabolic and genome engineering. Nat Biotechnol2016;35:48–55.

Gion K, Sakurai Y, Watari A et al. Designed recombinant tran-scription factor with antibody-variable regions. Anal Chem2009;81:10162–6.

Giuraniuc CV, MacPherson M, Saka Y. Gateway vectors for effi-cient artificial gene assembly in vitro and expression in yeast

Saccharomyces cerevisiae. PLoS One 2013;8, DOI: 10.1371/jour-nal.pone.0064419.

Gossen M, Bujard H. Tight control of gene expression in mam-malian cells by tetracycline-responsive promoters. P NatlAcad Sci USA 1992;89:5547–51.

Gossen M, Freundlieb S, Bender G et al. Transcriptional ac-tivation by tetracyclines in mammalian cells. Science (80-)1995;268:1766–9.

Guet CC, Elowitz MB, Hsing W et al. Combinatorial synthesis ofgenetic networks. Science (80-) 2002;296:1466–70.

Hahn S, Maurer P, Caesar S et al. Classical NLS proteins from Sac-charomyces cerevisiae. J Mol Biol 2008;379:678–94.

Hector RE, Mertens JA. A synthetic hybrid promoter for xylose-regulated control of gene expression in Saccharomyces yeasts.Mol Biotechnol 2017;59:24–33.

Hillen W, Berens C. Mechanisms underlying expression ofTN10 encoded tetracycline resistance. Annu Rev Microbiol1994;48:345–69.

Ikushima S, Boeke JD. New orthogonal transcriptional switchesderived from Tet repressor homologues for Saccharomycescerevisiae regulated by 2,4-diacetylphloroglucinol and otherligands. ACS Synth Biol 2017;6:497–506.

Ikushima S, Zhao Y, Boeke JD. Development of a tightly con-trolled off switch for Saccharomyces cerevisiae regulatedby camphor, a low-cost natural product. G3 2015;5:1983–90.

Jain D. Allosteric control of transcription in GntR family oftranscription regulators: a structural overview. IUBMB Life2015;67:556–63.

Jeeves M, Evans PD, Parslow RA et al. Studies of the Escherichiacoli Trp repressor binding to its five operators and to variantoperator sequences. Eur J Biochem 1999;265:919–28.

Junker F, Kiewitz R, Cook AM. Characterization of the p-toluenesulfonate operon tsaMBCD and tsaR in Comamonastestosteroni T-2. J Bacteriol 1997;179:919–27.

Kobayashi H, Kaern M, Araki M et al. Programmable cells: inter-facing natural and engineered gene networks. P Natl Acad SciUSA 2004;101:8414–9.

Kotula JW, Kerns SJ, Shaket LA et al. Programmable bacteria de-tect and record an environmental signal in the mammaliangut. P Natl Acad Sci USA 2014;111:4838–43.

Li S, Si T, Wang M et al. Development of a synthetic malonyl-CoA sensor in Saccharomyces cerevisiae for intracellularmetabolite monitoring and genetic screening. ACS Synth Biol2015;4:1308–15.

Loew R, Heinz N, Hampf M et al. Improved Tet-responsive pro-moters withminimized background expression. BMC Biotech-nol 2010;10:81.

McIsaac RS, Gibney PA, Chandran SS et al. Synthetic biologytools for programming gene expression without nutritionalperturbations in Saccharomyces cerevisiae. Nucleic Acids Res2014;42:e48.

Maddocks SE, Oyston PCF. Structure and function of the LysR-type transcriptional regulator (LTTR) family proteins. Micro-biology 2008;154:3609–23.

Møller TSB, Overgaard M, Nielsen SS et al. Relation between tetRand tetA expression in tetracycline resistant Escherichia coli.BMC Microbiol 2016;16, DOI: 10.1186/s12866-016-0649-z.

Moser F, Horwitz A, Chen J et al.Genetic sensor for strongmethy-lating compounds. ACS Synth Biol 2013;2:614–24.

Neidle EL, Hartnett C, Ornston LN. Characterization of Acine-tobacter calcoaceticus catM, a repressor gene homologousin sequence to transcriptional activator genes. J Bacteriol1989;171:5410–21.



Ottoz DSM, Rudolf F, Stelling J. Inducible, tightly regulated andgrowth condition-independent transcription factor in Sac-charomyces cerevisiae. Nucleic Acids Res 2014;42:e130.

Peralta-Yahya PP, Zhang F, del Cardayre SB et al. Microbial en-gineering for the production of advanced biofuels. Nature2012;488:320–8.

Raman S, Rogers JK, Taylor ND et al. Evolution-guided opti-mization of biosynthetic pathways. P Natl Acad Sci USA2014;111:17803–8.

Ravasio D, Walther A, Trost K et al. An indirect assay for volatilecompound production in yeast strains. Sci Rep 2014;4:3707.

Richards DH, Meyer S, Wilson CJ. Fourteen ways to reroute coop-erative communication in the lactose repressor: engineeringregulatory proteins with alternate repressive functions. ACSSynth Biol 2017;6:6–12.

Rogers JK, Church GM.Multiplexed engineering in biology. TrendsBiotechnol 2016;34:198–206.

Roney IJ, Rudner AD, Couture J-F et al. Improvement of the re-verse tetracycline transactivator by single amino acid sub-stitutions that reduce leaky target gene expression to unde-tectable levels. Sci Rep 2016;6:27697.

Sanseverino J, Gupta RK, Layton AC et al. Use of Saccharomycescerevisiae BLYES expressing bacterial bioluminescence forrapid, sensitive detection of estrogenic compounds. Appl En-viron Microb 2005;71:4455–60.

Shamanna DK, Sanderson KE. Genetics and regulation of D-xylose utilization in Salmonella typhimurium LT2. J Bacteriol1979;139:71–9.

SkjoedtML, Snoek T, Kildegaard KR et al. Engineering prokaryotictranscriptional activators as metabolite biosensors in yeast.Nat Chem Biol 2016;12:951–8.

Stanton BC, Nielsen AAK, Tamsir A et al. Genomic mining ofprokaryotic repressors for orthogonal logic gates. Nat ChemBiol 2013;10:99–105.

Stefan MI, Le Novere N. Cooperative Binding. PLoS Comput Biol2013;9:e1003106.

Strachan CR, Singh R, VanInsberghe D et al. Metagenomic scaf-folds enable combinatorial lignin transformation. P Natl AcadSci USA 2014;111:10143–8.

Suel GM, Lockless SW, Wall MA et al. Evolutionarily conservednetworks of residues mediate allosteric communication inproteins. Nat Struct Biol 2003;10:59–69.

Taylor ND, Garruss AS, Moretti R et al. Engineering an allosterictranscription factor to respond to new ligands. Nat Methods2015;13:177–83.

Teo WS, Chang MW. Bacterial XylRs and synthetic promotersfunction as genetically encoded xylose biosensors in Saccha-romyces cerevisiae. Biotechnol J 2015;10:315–22.

Teo WS, Hee KS, Chang MW. Bacterial FadR and syntheticpromoters function as modular fatty acid sensor- reg-ulators in Saccharomyces cerevisiae. Eng Life Sci 2013;13:456–63.

UmeyamaT, Okada S, Ito T. Synthetic gene circuit-mediatedmo-nitoring of endogenous metabolites: identification of GAL11as a novel multicopy enhancer of S-adenosylmethioninelevel in yeast. ACS Synth Biol 2013;2:425–30.

Urlinger S, Baron U, Thellmann M et al. Exploring the sequencespace for tetracycline- dependent transcriptional activators:novel mutations yield expanded range and sensitivity. P NatlAcad Sci USA 2000;97:7963–8.

Van Dien S. From the first drop to the first truckload: commer-cialization of microbial processes for renewable chemicals.Curr Opin Biotechnol 2013;24:1061–8.

Wang HH, Isaacs FJ, Carr PA et al. Programming cells by multi-plex genome engineering and accelerated evolution. Nature2009;460:894–8.

Wang M, Li S, Zhao H. Design and engineering of intracellular-metabolite-sensing/regulation gene circuits in Saccharomycescerevisiae. Biotechnol Bioeng 2016;113:206–15.

Werten S, Schneider J, Palm GJ et al. Modular organisation of in-ducer recognition and allostery in the tetracycline repressor.FEBS J 2016;283:2102–14.

Xiao Y, Bowen CH, Liu D et al. Exploiting nongenetic cell-to-cell variation for enhanced biosynthesis. Nat Chem Biol2016;12:339–44.

Xiong D, Lu S, Wu J et al. Improving key enzyme activity inphenylpropanoid pathway with a designed biosensor. MetabEng 2017;40:115–23.

Xu P, Li L, Zhang F et al. Improving fatty acids production by engi-neering dynamic pathway regulation and metabolic control.P Natl Acad Sci USA 2014;111:11299–304.

Zhang J, Jensen MK, Keasling JD. Development of biosensors andtheir application in metabolic engineering. Curr Opin ChemBiol 2015;28:1–8.

Zhao G, Hu T, Li J et al. A novel strategy to analyze l-tryptophanthrough allosteric Trp repressor based on rolling circle am-plification. Biosens Bioelectron 2015;71:103–7.

Zhou YJ, Buijs NA, Zhu Z et al. Production of fatty acid-derivedoleochemicals and biofuels by synthetic yeast cell factories.Nat Commun 2016;7, DOI: 10.1038/ncomms11709.


Documents

Lighting up yeast cell factories by transcription factor ... · prokaryotes(Fernandez-Lopez´ etal.2015).Inprokaryotes,these one-component protein-based biosensors are able to bind