Download Expressing Biologically Active Membrane Proteins in a Cell

bioRxiv preprint first posted online Jan. 30, 2017; doi: http://dx.doi.org/10.1101/104455. The copyright holder for this preprint (which was not
peer-reviewed) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license.
Expressing Biologically Active Membrane Proteins in a Cell-Free
Transcription-TranslationPlatform
ShaobinGuoa,*,AmitVaishb,QingChenb,*andRichardM.Murrayc
a
BiochemistryandMolecularBiophysics,CaliforniaInstituteofTechnology,California,USA
b
DiscoveryAttributeSciences,AmgenInc.,ThousandOaks,California,USA
c
ControlandDynamicalSystems,CaliforniaInstituteofTechnology,California,USA
*
Correspondingauthor:sguo@caltech.edu,qchen@amgen.com
Abstract:
Cell-freetranscription-translationplatformshavebeenwidelyutilizedtoexpresssolubleproteins
inbasicsyntheticbiologicalcircuitprototyping.Fromasyntheticbiologypointofview,itiscritical
to express membrane proteins in cell-free transcription-translation systems, and use them
directly in biocircuits, considering the fact that histidine kinases, G-protein coupled receptors
(GPCRs) and other important biosensors are all membrane proteins. Previous studies have
expressed membrane proteins in cell-free systems with the help of detergents, liposomes or
nanodiscs,buthavenotdemonstratedtheabilitytoprototypecircuitbehaviorforthepurpose
oftestingmorecomplexcircuitfunctionsinvolvingmembrane-boundproteins.Builtonprevious
efforts,inthisworkwedemonstratedthatwecouldco-translationallyexpresssolubilizedand
activemembraneproteinsinourcell-freeTX-TLplatformwithmembrane-likematerials.Wefirst
tested the expression of several constructs with β1 and β2 adrenergic receptors in TX-TL and
observedsignificantinsolublemembraneproteinproduction.Theadditionofnanodiscstothe
cellfreeexpressionsystemenabledsolubilizationofmembraneproteins.Nanodiscislipoproteinbased membrane-like material. The activity of β2 adrenergic receptor was tested with both
fluorescence and Surface Plasmon Resonance (SPR) binding assays by monitoring the specific
bindingresponseofsmall-moleculebinders,carazololandnorepinephrine.Ourresultssuggest
thatitispromisingtousecell-freeexpressionsystemstoprototypesyntheticbiocircuitsinvolving
singlechainmembraneproteinswithoutextraprocedures.Thisdatamadeusonestepcloserto
testingcomplexmembraneproteincircuitsincell-freeenvironment.
Introduction:
Cell-free transcription-translation systems have shown to be extremely useful in synthetic
biologicalcircuitprototyping[1,2].Typicalcell-freetranscription-translationsystemsarebased
onS30E.coliextract[3]andtherehavebeenmanydifferentversions.Thespecificversionused
inthispaperreferredtoas“TX-TL”,hasbeenoptimizedforprototypingsyntheticbiocircuits[4,
5].Unlikeothercell-freeproteinexpressionsystems,includingthePUREsystem,whicharebased
onbacteriophagetranscriptionbysupplementingbacteriophageRNApolymerasestothecrude
cytoplasmicextracts[6],TX-TLhasbeenshowntobesuitableforcomplexbiochemicalsystems
[1,2,4,7].
Membraneproteinsplayimportantrolesintheproperfunctioningofcellsandorganisms.They
laythefoundationformanybiosensorsandsignalpathwaysincells[8].Somemembraneproteins
actasionchannelstotransportionsacrossmembranes[9];othersmakeuptheessentialparts
bioRxiv preprint first posted online Jan. 30, 2017; doi: http://dx.doi.org/10.1101/104455. The copyright holder for this preprint (which was not
peer-reviewed) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license.
ofsensorysystemwhichareresponsibleforcellcommunication[10],whilemembraneenzymes
catalyzeimportantchemicalreactionsnearmembranes[11].Inaddition,membraneproteinsare
themostimportantdrugtargets,forexamplemorethanhalfoftherapeuticsfortreatmentof
various modalities ranging from cancer to cardiovascular diseases target membrane proteins
[12].
Membrane proteins are difficult targets as compared to soluble proteins because of the
challengesassociatedwiththeirexpression,solubilization,andstabilization.Typicalstudiesof
membraneproteinsrelyonproteinsproducedfromcellsandsolubilizedcellmembraneusing
detergents, liposomes or other membrane-like materials. These approaches may have
advantagesintermsofproteinyield,theyarenotwell-suitedforhigh-throughputselectionof
protein constructs, since transformation, cell growth, lysis, membrane solubilization and
purificationareinvolvedforeachconstruct[13].Itisalsodifficulttousethesetechniquesfor
biocircuitsprototyping,whichrequirepromptfunctionofthenewlyexpressedproteins.
Recently cell free expression of membrane protein have been reported using detergents,
liposomesornanodiscstosolubilizeandstabilizetheproteins[13-19].Ourdataindicatedthe
possibilitytointegratemembraneproteinsdirectlyintofuturesyntheticbiocircuitswithcomplex
functions.Amongalltheimportantmembraneproteins,wepickedGproteincoupledreceptors
(GPCRs), beta-2(and 1) adrenergic receptors (β2AR/β1AR) as model proteins. With more than
900members,GPCRsareoneofthemostimportantandthelargestintegralmembraneproteins
familyinhumancellsandthemostimportantclinicaldrugtargetsastheyplayimportantrolesin
manyphysiologicalfunctionsandimplicatedinmanydiseases[12,20,21].GPCRsallsharethe
sametopology–seventransmembranea-helices,andtheyarethoughttofunctioninmonomeric
form[22],althoughtherehavebeenstudiesindicatingtheirdimerization[23,24].WeusedTXTL platform in combination with nanodisc for expressing β2AR/β1AR and subsequent in situ
stabilization.Furthermore,biologicalactivityoftheseproteinswasconfirmedbybindingtoits
ligands.
bioRxiv preprint first posted online Jan. 30, 2017; doi: http://dx.doi.org/10.1101/104455. The copyright holder for this preprint (which was not
peer-reviewed) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license.
A
B
Figure1.IllustrationoftheplasmidmapandlinearDNAoftheconstructs.pSG73isusedasanexample.pSG74and
pSG75sharethesamefeaturesaspSG73.A:CircularplasmidmapofpSG73.B:LinearDNAversionofpSG73.
ResultsandDiscussion:
We ordered gene synthesis services and then used golden gate assembly [27] to make three
differentGPCRconstructswiththreedifferentvariants,whichsharesimilarbackbone,promoter,
ribosomebindingsitesandterminator(Figure1);Figure2liststhedifferenceincodingsequences
correspondingtovariousadrenergicreceptorconstructs.Allthreeconstructssharedthesame
superfolder GFP (sfGFP) fusion protein topology, which was tagged with 6xHis tag at the Cterminaloftheprotein.sfGFPisusedformonitoringandestimationoftargetproteinexpression
levelduringandattheendofTX-TL.Whereas,6xHistagwasusedtodetectproteinsinWestern,
tocapturetheproteinforbindingassays,andforaffinitypurification,ifnecessary.
OneoftheadvantagesofTX-TLexpressionplatformisthatwecoulduseeitherlinearDNAor
plasmid DNA for expression in TX-TL [7]. Because of that, we could implement fast construct
prototypinginTX-TLbyligatingpartstogetherandamplifyingthelinearDNAswithPCR,avoiding
cloningthelinearfragmentsintoplasmid.ToexpresstheseconstructsinTX-TL,weonlyadded
GPCR%Protein
Construct
Protein%Size
β1#AR#ts:)3ZPR)Thermostabilized)turkey)β1)adrenergic)receptor
pSG73:)protein#sfGFP#His6
65kD)(35kD)minus)sfGFP)
β2#AR)wild)type:)ADRB2_HUMAN)β2)adrenergic)receptor
pSG74:)protein#sfGFP#His6
75kD)(45kD)minus)sfGFP)
β2#AR#T4L:)2RH1)β2#adrenergic)receptor/T4#lysozyme)chimera
pSG75:)protein#sfGFP#His6
85kD)(55kD)minus)sfGFP)
Figure2.Informationof GPCR constructs usedinexperiments.pSG73andpSG75havethecorrespondingPDBnumber
noted.Proteinsizeofeachfusionproteinisalsolisted.
bioRxiv preprint first posted online Jan. 30, 2017; doi: http://dx.doi.org/10.1101/104455. The copyright holder for this preprint (which was not
peer-reviewed) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license.
DNAoftheseconstructstoTX-TLreactionmix.Theiterationofeachprototypingtestsignificantly
faster(overnight)comparedtoweeksusingcloningforcell(E.coli)expression.
A
B
300
1200
1000
GFP-expression- in-nM
GFP+expression+ in+nM
250
200
150
100
50
800
600
400
200
0
0
pSG73:+
b1ARts+
linear+
10nM
pSG73:+
b1ARts+
linear+
20nM
pSG73:+
b1ARts+
linear+
40nM
pSG74:+
b2AR+
linear+
10nM
pSG74:+
b2AR+
linear+
20nM
pSG74:+ pSG75:+ pSG75:+ pSG75:+
b2AR+ b2ART4L+ b2ART4L+ b2ART4L+
linear+
linear+
linear+
linear+
40nM
10nM
20nM
40nM
pSG73:b1ARtslinear20nM
pSG73:- pSG73:- pSG74:- pSG74:- pSG74:- pSG75:- pSG75:- pSG75:b1ARts- b1ARtsb2ARb2ARb2AR- b2ART4L- b2ART4L- b2ART4Llinear- w/- linear- w/- linear- linear- -w/- linear- w/- linear- linear- w/- linear- w/ND1
ND3
20nM
ND1
ND3
20nM
ND1
ND3
Figure3.ExpressionlevelofpSG73-75constructsinTXTLmeasuredbysfGFPA:EndpointmeasurementofGFPexpression
from10nM,20nM and40nMlinearDNAofpSG73,pSG74 andpSG75 in10µLTX-TL.B: Endpoint measurementof GFP
expressionfrom20nMlinearDNAofpSG73-75withorwithout24 µMnanodiscsinTX-TL.ND1isMSP1D1-DMPCandND3is
MSP1E3D1-DMPG.Measurements weredoneat29°C in BIOTEKSynergyH1HybridMulti-ModeMicroplateReaderusing
ex485nm/em525nm.
WefirstestimatedtheexpressionlevelofpSG73-75inTX-TLbymeasuringthefluorescenceof
sfGFP which was fused at the C-terminal of β2AR or 1AR, using a plate reader at 485 nm
(absorbance)/525nm(emission).AllthelinearDNAconstructsshowedaGFPfluorescencesignal,
indicatingsuccessfulexpressionofthefusionproteinsinTX-TL(Figure3A).Differentconstructs,
despitehavingexactlythesamepromoter,ribosomebindingsite,fusionproteinframeworkand
DNA concentration, showed different expression levels of sfGFP, especially pSG74. This
suggestedthatthedifferenceincodingsequencescouldaffecttranscriptionandtranslationlevel.
AnotherobservationwasthatincreasinglinearDNAconcentrationcouldhelpincreasethefusion
proteinexpression,butthisincreasewaslimitedbyTX-TLresourcesand/ortoxicsaccumulated
inbatchmode,asshownelsewhere[28].SettingupTX-TLreactionsindialysissystemsislikelyto
improvetheproteinexpressionlevel.
After testing the fusion protein expression level, we explored the possibility of stabilizing
hydrophobicmembraneproteinbyprovidingamembranemimicintothereaction.Detergent
micelle,liposomeandnanodiscarecommonlyusedtoprovideartificialbilayerenvironmentto
membraneprotein(Ref).WeobservedthatmajorityofdetergentsweredetrimentaltoTX-TL
reactioncomprisingmembraneprotein(datanotshown).Conversely,reconstitutingproteininto
liposome resulted into a poor yield of the folded protein, which could be attributed to the
liposome’s closed topography. Therefore, we employed nanodisc − a lipid-protein complex,
composed of lipids constrained by a membrane scaffold protein (MSP). Nanodisc provides a
robustplatformwithtwo-dimensionaltopographyforreconstitutingTX-TLexpressedmembrane
protein.Wechosetwoofthemostcommonlyusednanodiscs:MSP1D1-DMPC(ND1)~10nm
diameterandalongerversion,MSP1E3D1-DMPG(ND3)~13nmindiameter[13].
bioRxiv preprint first posted online Jan. 30, 2017; doi: http://dx.doi.org/10.1101/104455. The copyright holder for this preprint (which was not
peer-reviewed) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license.
A
No*DNA
pSG73:*β14AR4ts
10nM 20nM 40nM
pSG74:*β24AR
10nM 20nM 40nM
pSG75:*β24AR4T4L
10nM 20nM 40nM
Dimer
Monomer
B
pSG73:*β14AR4ts
Nanodisc
!
ND1%%%%%ND3
pSG74:*β24AR
!
ND1%%%%%ND3
pSG75:*β24AR4T4L
!
ND1%%%%%ND3
Dimer
Monomer
Figure4.WesternblotsofTX-TLreactions.A:ResultsofwholereactionTX-TLsamples.Aftermeasuringthefluorescence
oftheTX-TLreaction,theywereusedforwesternblotusingthemethodinMaterialsandMethods.TX-TLreactionwith
noDNAwasusedasnegativecontrolandthreedifferentconcentrationsofthreedifferentconstructswererununderthe
same condition. B: Results of only supernatant from TX-TL samples. Only soluble samples were run on this blot.
SupernatantswithoutanynanodiscswererunatthesameconditionasoneswitheitherND1orND3.
WerepeatedtheTX-TLexpressionexperimentinthepresence/absenceofND1orND3.Asshown
inFigure3B,thepresenceofnanodiscimprovestheTX-TLefficiency,asindicatedbyenhanced
GFP fluorescence. Nanodisc doesn’t have any intrinsic fluorescence, therefore the increased
fluorescenceshouldbearisingfromthefusionprotein.GPCR-sfGFPfusionproteinprecipitated
in the TXTL reaction mix without nanodisc. However, in the presence of nanodisc, the
supernatantofthereactionmixtureexhibitedincreasedGFPfluorescencesignal,indicatingthat
nanodiscslikelytohelpstabilizingfusionmembraneproteinsinTX-TL,andkeepthefoldedfusion
proteininsupernatant.
TofurtherconfirmthatthetargetfusionproteinswereexpressedinTX-TLandnanodiscscould
helpsolubilizeGPCR-sfGFPfusionproteins,weranSDS-PAGEofTX-TLsamplesandwesternblots
withananti-His6antibody.First,weranthewholereactionsamplesofthreedifferentconstructs
atthreedifferentconcentrations(Figure4A).
As illustrated on the blot, samples that had one of the three constructs showed significant
detection of His-tagged proteins, protein size was confirmed on the blot using a ladder. In
contrast,controlsamplewithwithoutDNAhadnoHis-taggedsignal.Additionally,therewerenot
significantdifferencesbetweendifferentconcentrationsinpSG73andpSG75,suggestingmasstransfernutrientslimitationandtoxicaccumulationintheTX-TLreactionasdiscussedearlier.
AnotherinterestingobservationwasthatpSG73-β1-AR-tsdidnotshowdimerizationbutpSG74
bioRxiv preprint first posted online Jan. 30, 2017; doi: http://dx.doi.org/10.1101/104455. The copyright holder for this preprint (which was not
peer-reviewed) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license.
andpSG75,whichwerebothβ2-ARbasedfusionproteins,showedstrongdimerbandsonthe
blot.ThepresenceofGPCRdimerizationisconsistentwithpreviousreports[24].
We further tested the presence of ND by spinning down the TX-TL reaction mix and took
supernatantonlytorunthewesternblot.Alltheinsolubleproteinswouldprecipitateoutand
endedupinthepelletaftercentrifugation.Onlysolubleproteinswouldbeinthesupernatant
andcanbedetectedinthewesternblot.InFigure4B,wehadthreedifferentconstructsandeach
ofthemhadthreedifferentexperimentalconditions:nonanodisc;withND1:MSP1D1-DMPC;
and with ND3: MSP1E3D1-DMPG. All target proteins generated in TX-TL precipitated without
nanodiscsandleftinthepellet(confirmedbyrunningpelletonwesternblot,datanotshown).
Onthecontrary,wheneitherND1orND3wasaddedintothereaction,wesawsignificantbands
of target membrane proteins by the Western, suggesting that these hydrophobic membrane
proteinsbecamesolublewithhelpfromnanodiscs.
SofarwehavedemonstratedthattargetmembraneproteinsproducedinTXTLreactionscanbe
solubilizedandstabilizedwithin-situpresenceofnanodiscs.However,proteinassociationwith
nanodiscdoesn’tinsurethattheseproteinsarebiologicallyactive.Tofurthertestwhetherthese
soluble proteins were active, we designed two binding assays: fluorescence-based carazolol
bindingassayandSPRbasednorepinephrinebindingassay.
Althoughthereisnoneedtopurifytheproteinforcircuitprototyping,purificationisrequiredto
confirmthebindingactivityofTX-TLexpressedproteintoavoidinterferencecausedbyE.coli
endogenousproteinsintheTX-TLreactionmix.
We used Ni-NTA affinity chromatography to purify His-tagged target protein. The protocol is
describedinMaterialsandMethods.
Fluorescence-based carazolol binding assay uses (S)-carazolol, a derivative of the potent β
blockercarazololwithfluorescenceproperties(ex633nm/em650nm).Thisligandcanbeusedas
afluorescencetrackerforβ2ARbindingactivity.Thedetailedexperimentalsetupisdescribedin
Materials and Methods. Briefly, purified membrane proteins were incubated with or without
carazolol for one hour. These samples were further dialyzed against 100X volume buffer
overnight.Subsequently,allthesampleswereconcentrateddowntotheirstartingvolumeand
fluorescence signal was measured in plate reader. Green fluorescence (GFP) indicates the
amountoffusionproteinsinthesamplesandredfluorescencewouldrepresenttheamountof
carazololboundtotargetmembraneproteinsasanindicationofproteinactivity.
WeusedtwodifferentND1(MSP1D1-DMPC)inthisassay:1)HisNDuses6xHistaggedMSP1D1
andwaspurchasedfromCubeBiotech;2)BiotinNDusesbiotinylatedMSP1D1andwasmadein
house.AsseeninFigure5B,therewasnotmuchdifferenceingreenfluorescencesignalbetween
the samples. However, there is a significant difference between ND1 samples incubated with
bioRxiv preprint first posted online Jan. 30, 2017; doi: http://dx.doi.org/10.1101/104455. The copyright holder for this preprint (which was not
peer-reviewed) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license.
carazolol. We attribute high carazolol signal on HisND reconstituted β2AR to the presence of
emptyNDafterNi-NTAaffinitychromatographypurification.
One limitation of fluorescence-based assay is that non-specific binding of carazolol to lipids
surroundingmembraneproteincouldcausehighersignal-to-noiseratio.Toovercomethis,we
developedaSPR-basedbindingassayusingnorepinephrine,whichisapartialagonistforβ2AR.
Since norepinephrine shares the same binding pocket with carazolol [31], we tethered
norepinephrinetothesurfaceviaamidebondbetweenitsprimaryamineandthecarboxylgroup
on the surface. We hypothesized that carazolol could be used as a competitor for β2AR to
norepinephrinebindingonSPR.AsshowninFigure5C.,SPRbindingresponseofβ2ARdecreased
~60%whenitwasincubatedwith1µMcarazolol,indicatingspecificinteractionbetweenβ2AR
anditsbindingpartners.
A
B
20
1.2
1
16
GFP+fluorescence+fold+change
Carazolol+fluorescence+fold+change
18
14
12
10
8
6
4
0.8
0.6
0.4
0.2
2
0
C
0
D
140
B in d in g 0R e s p o n s e 0o f 0N D 8 R e c o n s t it u t e d
120
b2AR
Response+Unit
100
80
b2AR_car
60
40
S P R 0R e s p o n s e 0( R U )
B 2 A R 0( W T ) 0o n 0N E 8 F u n c t io n a liz e d 00S u r f a c e s
100
50
0
A
2
B
A
R
_
C
a
R
r
20
B
2
0
0
200
400
600
Time+(s)
800
1000
1200
Figure5.BindingassaysofTX-TLmadeβ1andβ2adrenergicreceptor.A:Barchartofcarazololfluorescencefoldchange.
Thenumberwastheratioofredfluorescencesignalfromsamplesincubatedwithcarazololtotheoneswithoutcarazolol.
DetailedexperimentalmethodisinMaterialsandMethods.Allsamplesweredialyzedagainstbufferwithoutcarazololto
removenon-bindingcarazolol.B:BarchartofGFPfluorescencefoldchangewiththesamesamplesfromA.Thenumberwas
theratioof greenfluorescence signalfrom samplesincubated with carazolol to theones withoutcarazolol. C: Response
curvesofβ2adrenergicreceptorproteinsbindingtoSPRsurfacecoatedwithnorepinephrine.Redcurvewasresponsesignal
fromβ2-AR sample withoutcarazolol. Green curve wasresponse signalfromsample incubated with1 µM carazolol asa
control.D:BarchartoftheendpointoftheresponsecurvesinC.
bioRxiv preprint first posted online Jan. 30, 2017; doi: http://dx.doi.org/10.1101/104455. The copyright holder for this preprint (which was not
peer-reviewed) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license.
Tosummarize,wehavedemonstratedthat1)membraneproteincanbeexpressedinTX-TLat
analytical scale; 2) presence of nanodisc during TX-TL reaction facilitates folding and
solubilizationofsinglechainmembraneprotein;3)fluorescenceandSPRbindingassayswere
developedtodemonstratespecificinteractionbetweensmall-moleculeandnanodisc-stabilized
membraneprotein.
Conclusions:
Inthiswork,wetestedproteinsfromtheGPCRfamilyinourcell-freetranscription-translation
(TX-TL) system, which would be ideal for synthetic biocircuit prototyping. We expressed
β1AR/β2AR in TX-TL with nanodiscs and were able to show that not only these membrane
proteins are soluble in TX-TL with nanodiscs, but they were also active. Nanodiscs are cotranslationally associated with membrane proteins without extra processing, which enables
directprototypingofacompletebiocircuitwithTX-TLproducedmembraneprotein(s).
Weintendtooptimizeourbindingassaysandperformcompetitivebindingexperimentstotest
the stringency of this assay. We envision that GPCR co-expressed with G protein can be
expressedbyTX-TLtotestthebiologicalcircuitincludingsignaltransduction.
Additionally, histidine kinases, which phosphorylate corresponding response regulators, can
activatedownstreamtranscriptionandtranslation.Wehavestartedtestedsomehybridhistidine
kinases and have seen promising results. Our goal is to prototype a logic biocircuit with
membraneenzymeinit,expandingTX-TLplatformtobroadertopics.
bioRxiv preprint first posted online Jan. 30, 2017; doi: http://dx.doi.org/10.1101/104455. The copyright holder for this preprint (which was not
peer-reviewed) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license.
MaterialsandMethods:
PlasmidsandlinearDNAs:
DNA and oligonucleotides primers were ordered from Integrated DNA Technologies (IDT,
Coralville,Iowa).PlasmidsinthisstudyweredesignedinGeneious8(Biomatters,Ltd.)andwere
madeusingstandardgoldengateassembly(GGA)protocols.BsaI-HF(R3535S)enzymeusedin
GGAwaspurchasefromNewEnglandBiolabs(NEB).LinearDNAsweremadebyPCRingprotein
expressionrelatedsequencesoutofGGAconstructsusingPhusionHotStartFlex2XMasterMix
(M0536L)fromNEB.
TX-TLreactions:
TX-TLreactionmixwassetupaccordingtopreviousJOVEpaper[5].Briefly,TX-TLextractand
bufferweremixedtogetherwithcalculatedlinearDNAsorplasmidswithorwithoutnanodiscs.
Reactionvolumesvariedfrom10µL(initialscreening)to1mL(proteinpurificationandanalysis).
Gelandwesternblot:
GelsusedinthisworkwereBolt4–12%Bis-TrisPlusGelsfromThermoFisherScientific.Running
bufferwasBoltMESSDSRunningBuffer.Gelswererunwithoutreducingagents.Proteinsamples
weremixedwithLDSsamplebufferbeforeloadingintogels.iBlot2GelTransferDeviceand
iBlot Nitrocellulose Regular Stacks were used for transfer proteins from gel to membrane.
MembranewasthentransferredtoiBinddeviceandincubatedwithPenta-HisHRPConjugatein
1:500dilutions.BlotsweredetectedusingSuperSignalChemiluminescentHRPSubstratesfrom
ThermoFisherScientific.
Proteinpurification:
TX-TLreactionmixwasfirstspun@14,000gfor10minat4°C.Supernatantwasthentransferred
to buffer-equilibrated HisPur Ni-NTA Spin Purification column (ThermoFisher Scientific) and
incubatedwithshakingfor1h.Thenspundowntheflowthrough@2000gfor2minandwashed
withnanodiscbuffer(20mMTrispH7.4,0.1MNaCl)addedwith20mMimidazolethreetimes.
Elutionwasdonebyaddingelutionbuffer(20mMTrispH7.4,0.1MNaCl,250mMimidazole)for
3times2columnvolume.ProteinswerethenconcentratedusingAmiconUltraCentrifugalFilter
Units(Millipore)withUltracel-30membrane.
Fluorescence-basedcarazololbindingassay:
Carazolol was purchased from Abcam (S)-Carazolol Fluorescent ligand (Red) ab118171. Each
purified protein was first divided into two equal volume samples. One was added 100nM
carazolol(dissolvedinwater)andtheotheronewasaddedthesamevolumeofwater.Theywere
incubated at 4°C for 1h before transferred to mini 10k MW D-Tube Dialyzers (Millipore) and
dialyzed against 100x volume of nanodisc buffer for overnight. Then dialyzed samples were
concentratedusingAmiconUltraCentrifugalFilterUnits(Millipore)withUltracel-30membrane
to starting volume. Samples were then put into BIOTEK Synergy H1 Hybrid Multi-Mode
MicroplateReaderandmeasuredforGFPfluorescence(ex485nm/em525nm)withgain61orgain
bioRxiv preprint first posted online Jan. 30, 2017; doi: http://dx.doi.org/10.1101/104455. The copyright holder for this preprint (which was not
peer-reviewed) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license.
100orRedfluorescence(ex633nm/650nm)withoptimalgain.GFPfluorescencewasconverted
tonMusingcalibrationdatafrompurifiedGFPprotein.
SurfacePlasmonResonance(SPR)basednorepinephrinebindingassay:
GEBiacoreT-200SPRsystemwasusedforSPRexperiment.Goldplatedchipwasfirstimmobilized
withnorepinephrineandthenwashedawayextrachemical.Therearefourchannelsononechip.
Twowereusedasexperimentalchannelsandtheremainingtwowereusedasnegativecontrols
toprovidebackgroundresponsefrombuffer.Onesamplewasincubatedwith1µMcarazololfor
1hattotestbindingspecificityandtheothersamplewasincubatedwithsamevolumeofwater.
Samples were then loaded and flowed through corresponding experimental channels and
responsecurveswererecorded.
Acknowledgments
WewouldliketothankHaoChenforreceptorplasmiddesign.MarieWrightandRobertKurzeja
fornanodiscexpression,andAmgenInc.forfinancialsupport.
Reference:
1.
Takahashi,M.K.,etal.,Characterizingandprototypinggeneticnetworkswithcell-free
transcription-translationreactions.Methods,2015.86:p.60-72.
2.
Niederholtmeyer,H.,etal.,Rapidcell-freeforwardengineeringofnovelgeneticring
oscillators.Elife,2015.4:p.e09771.
3.
Nirenberg,M.W.andJ.H.Matthaei,Thedependenceofcell-freeproteinsynthesisinE.
coliuponnaturallyoccurringorsyntheticpolyribonucleotides.ProcNatlAcadSciUSA,
1961.47:p.1588-602.
4.
Shin,J.andV.Noireaux,AnE.colicell-freeexpressiontoolbox:applicationtosynthetic
genecircuitsandartificialcells.ACSSynthBiol,2012.1(1):p.29-41.
5.
Sun,Z.Z.,etal.,ProtocolsforimplementinganEscherichiacolibasedTX-TLcell-free
expressionsystemforsyntheticbiology.JVisExp,2013(79):p.e50762.
6.
Shimizu,Y.,etal.,Cell-freetranslationreconstitutedwithpurifiedcomponents.Nat
Biotechnol,2001.19(8):p.751-5.
7.
Sun,Z.Z.,etal.,LinearDNAforrapidprototypingofsyntheticbiologicalcircuitsinan
EscherichiacolibasedTX-TLcell-freesystem.ACSSynthBiol,2014.3(6):p.387-97.
8.
Cournia,Z.,etal.,MembraneProteinStructure,Function,andDynamics:aPerspective
fromExperimentsandTheory.JMembrBiol,2015.248(4):p.611-40.
9.
Choe,S.,Potassiumchannelstructures.NatRevNeurosci,2002.3(2):p.115-21.
10.
Pierce,K.L.,R.T.Premont,andR.J.Lefkowitz,Seven-transmembranereceptors.NatRev
MolCellBiol,2002.3(9):p.639-50.
11.
Bhate,M.P.,etal.,Signaltransductioninhistidinekinases:insightsfromnewstructures.
Structure,2015.23(6):p.981-94.
12.
Lappano,R.andM.Maggiolini,Gprotein-coupledreceptors:noveltargetsfordrug
discoveryincancer.NatRevDrugDiscov,2011.10(1):p.47-60.
13.
Bayburt,T.H.andS.G.Sligar,MembraneproteinassemblyintoNanodiscs.FEBSLett,
2010.584(9):p.1721-7.
bioRxiv preprint first posted online Jan. 30, 2017; doi: http://dx.doi.org/10.1101/104455. The copyright holder for this preprint (which was not
peer-reviewed) is the author/funder. It is made available under a CC-BY-NC-ND 4.0 International license.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
Ishihara,G.,etal.,ExpressionofGproteincoupledreceptorsinacell-freetranslational
systemusingdetergentsandthioredoxin-fusionvectors.ProteinExprPurif,2005.41(1):
p.27-37.
Roos,C.,etal.,Co-translationalassociationofcell-freeexpressedmembraneproteins
withsuppliedlipidbilayers.MolMembrBiol,2013.30(1):p.75-89.
Roos,C.,etal.,Characterizationofco-translationallyformednanodisccomplexeswith
smallmultidrugtransporters,proteorhodopsinandwiththeE.coliMraYtranslocase.
BiochimBiophysActa,2012.1818(12):p.3098-106.
Rues,R.B.,V.Dotsch,andF.Bernhard,Co-translationalformationandpharmacological
characterizationofbeta1-adrenergicreceptor/nanodisccomplexeswithdifferentlipid
environments.BiochimBiophysActa,2016.1858(6):p.1306-16.
Yang,J.P.,etal.,Cell-freesynthesisofafunctionalGprotein-coupledreceptorcomplexed
withnanometerscalebilayerdiscs.BMCBiotechnol,2011.11:p.57.
Schneider,B.,etal.,Membraneproteinexpressionincell-freesystems.MethodsMol
Biol,2010.601:p.165-86.
Fredriksson,R.,etal.,TheG-protein-coupledreceptorsinthehumangenomeformfive
mainfamilies.Phylogeneticanalysis,paralogongroups,andfingerprints.Mol
Pharmacol,2003.63(6):p.1256-72.
Lefkowitz,R.J.,Seventransmembranereceptors:somethingold,somethingnew.Acta
Physiol(Oxf),2007.190(1):p.9-19.
Rosenbaum,D.M.,S.G.Rasmussen,andB.K.Kobilka,ThestructureandfunctionofGprotein-coupledreceptors.Nature,2009.459(7245):p.356-63.
Terrillon,S.andM.Bouvier,RolesofG-protein-coupledreceptordimerization.EMBO
Rep,2004.5(1):p.30-4.
Kasai,R.S.andA.Kusumi,Single-moleculeimagingrevealeddynamicGPCRdimerization.
CurrOpinCellBiol,2014.27:p.78-86.
Christopher,J.A.,etal.,Biophysicalfragmentscreeningofthebeta1-adrenergic
receptor:identificationofhighaffinityarylpiperazineleadsusingstructure-baseddrug
design.JMedChem,2013.56(9):p.3446-55.
Cherezov,V.,etal.,High-resolutioncrystalstructureofanengineeredhumanbeta2adrenergicGprotein-coupledreceptor.Science,2007.318(5854):p.1258-65.
Engler,C.,etal.,Goldengateshuffling:aone-potDNAshufflingmethodbasedontype
IIsrestrictionenzymes.PLoSOne,2009.4(5):p.e5553.
Siegal-Gaskins,D.,etal.,Genecircuitperformancecharacterizationandresourceusage
inacell-free"breadboard".ACSSynthBiol,2014.3(6):p.416-25.
Nath,A.,W.M.Atkins,andS.G.Sligar,Applicationsofphospholipidbilayernanodiscsin
thestudyofmembranesandmembraneproteins.Biochemistry,2007.46(8):p.2059-69.
Innis,R.B.,F.M.Correa,andS.H.Synder,Carazolol,anextremelypotentbeta-adrenergic
blocker:bindingtobeta-receptorsinbrainmembranes.LifeSci,1979.24(24):p.225564.
Liu,J.J.,etal.,Biasedsignalingpathwaysinbeta2-adrenergicreceptorcharacterizedby
19F-NMR.Science,2012.335(6072):p.1106-10.