Handbook of surface plasmon resonance

554 Pages·2017·22.244 MB·English

by  Richard B M Schasfoort, Richard Schasfoort, Arnoud Marquart, Peter Schuck, Erk T. Gedig, Robert Karlsson, Koen Wagner, Noah T. Ditto, David Apiyo, Sylvie Ricard Blum

Handbook of Surface Plasmon Resonance
2nd Edition
1
0
0
P
F
3-
8
2
0
1
0
8
8
7
1
8
7
9
9/
3
0
1
0.
1
oi:
d
g |
or
c.
s
s.r
b
u
p
p://
htt
n
o
7
1
0
2
y
a
M
4
2
n
o
d
e
h
s
bli
u
P
View Online
1
0
0
P
F
3-
8
2
0
1
0
8
8
7
1
8
7
9
9/
3
0
1
0.
1
oi:
d
g |
or
c.
s
s.r
b
u
p
p://
htt
n
o
7
1
0
2
y
a
M
4
2
n
o
d
e
h
s
bli
u
P
View Online
Handbook of Surface Plasmon
Resonance
2nd Edition
1
0
0
P
F
3-
8
2
0
1
0
8
78 Edited by
1
8
7
9
39/ Richard B. M. Schasfoort
0
1
0. University of Twente, Enschede, The Netherlands
1
oi: Email: [email protected]
d
g |
or
c.
s
s.r
b
u
p
p://
htt
n
o
7
1
0
2
y
a
M
4
2
n
o
d
e
h
s
bli
u
P
View Online
1
0
0
P
F
3-
8
2
0
1
0
8
8
7
1
8
7
9
9/
3
0
1
0.
doi:1 PPrDinFteIISSBBNN::997788--11--7788286021--703208--23
g | EPUBeISBN:978-1-78801-139-6
or
c.
s
s.r AcataloguerecordforthisbookisavailablefromtheBritishLibrary
b
u
p
p:// rTheRoyalSocietyofChemistry2017
htt
n Allrightsreserved
o
7
1
20 Apartfromfairdealingforthepurposesofresearchfornon-commercialpurposesorfor
ay privatestudy,criticismorreview,aspermittedundertheCopyright,DesignsandPatents
M
4 Act1988andtheCopyrightandRelatedRightsRegulations2003,thispublicationmaynot
n 2 bereproduced,storedortransmitted,inanyformorbyanymeans,withouttheprior
d o permissioninwritingofTheRoyalSocietyofChemistryorthecopyrightowner,orinthe
he caseofreproductioninaccordancewiththetermsoflicencesissuedbytheCopyright
s
bli LicensingAgencyintheUK,orinaccordancewiththetermsofthelicencesissuedby
Pu theappropriateReproductionRightsOrganizationoutsidetheUK.Enquiriesconcerning
reproductionoutsidethetermsstatedhereshouldbesenttoTheRoyalSocietyof
Chemistryattheaddressprintedonthispage.
Whilstthismaterialhasbeenproducedwithallduecare,TheRoyalSocietyofChemistry
cannotbeheldresponsibleorliableforitsaccuracyandcompleteness,norforany
consequencesarisingfromanyerrorsortheuseoftheinformationcontainedinthis
publication.Thepublicationofadvertisementsdoesnotconstituteanyendorsementby
TheRoyalSocietyofChemistryorAuthorsofanyproductsadvertised.Theviewsand
opinionsadvancedbycontributorsdonotnecessarilyreflectthoseofTheRoyalSocietyof
Chemistrywhichshallnotbeliableforanyresultinglossordamagearisingasaresult
ofrelianceuponthismaterial.
TheRoyalSocietyofChemistryisacharity,registeredinEnglandandWales,
Number207890,andacompanyincorporatedinEnglandbyRoyalCharter
(RegisteredNo.RC000524),registeredoffice:BurlingtonHouse,Piccadilly,
LondonW1J0BA,UK,Telephone:þ44(0)20743786556.
Visitourwebsiteatwww.rsc.org/books
PrintedintheUnitedKingdombyCPIGroup(UK)Ltd,Croydon,CR04YY,UK
5
0
0
P
F Foreword to the 1st Edition
3-
8
2
0
1
0
8
8
7
1
8
7
9
9/
3
0
0.1 Make no bones about it, I love surface plasmon resonance (SPR)-based
1
oi: biosensor technology. After spending three years trying to measure binding
g | d constants using analytical affinity chromatography, I immediately saw the
or benefits of SPR the first time I sat down in front of a Biacore in 1991. Even
c.
s.rs today, no comparable technology exists to characterize molecular inter-
ub actions in real time without labeling in an automated and robust fashion.
p
p:// Butasthetechnologyhasexpandedovertheyears,Ifindthattherearethree
htt general attitudes towards SPR. There are the nay-sayers who hate the tech-
n
o nology. There arelong-time users who thinktheyare experts.And there are
7
01 the users who recognize they do not know everything about SPR but are
2
y eager to improve their skills.
a
M
4 Eversincethefirstcommerciallyviableinstrumentwasunleashedin1990
2
n by the biosensor group at Pharmacia (which was spun out into a separate
o
d company called Biacore in 1996, only to be acquired recently by General
e
h
s Electric, which previously bought Amersham who at one time had merged
bli
u withPharmacia,soinfactnowthebiosensorgrouphascomefullcircle,even
P
though they have always shared the same cafeteria in Uppsala, Sweden),
therehavebeencriticsofSPRtechnology.Somuchsothatin2003Icreateda
character called ‘‘Dr Evil Pessimist’’, who represents a composite of the
various detractors of SPR. Dr Pessimist rants and rages about problems he
has with the technology, including nonspecific binding, instrument drift,
mass transport and avidity effects. He argues that since SPR uses a surface
the rate constants we measure will never reflect solution-based binding
constants.Infact,muchofhisresentmentofthetechnologystemsfromthe
fact that his experiments fail or his data never fit a simple model.
It has been my experience that there are two primary causes of this
SPRaphobia: poor-quality reagents and/or poor experimental design. Per-
haps the molecules Dr Pessimist is studying do not in fact interact or the
HandbookofSurfacePlasmonResonance,2ndEdition
EditedbyRichardB.M.Schasfoort
rTheRoyalSocietyofChemistry2017
PublishedbytheRoyalSocietyofChemistry,www.rsc.org
v
View Online
vi Forewordtothe1stEdition
preparations of samples are not active to begin with. Don’t shoot the
messenger.DrPessimistassertsthathisproteinsareofhighqualitybecause
theyare‘‘asinglebandonanSDSpagegel’’.Hefailstorealizethatthisisnot
evidence of an active preparation or a conformationally homogeneous
sample. I think biosensor experiments are akin to protein crystallography.
05 No structural biologist I know would attempt to crystallize an impure, half-
0
P denatured preparation of protein that has precipitated at the bottom of
F
83- an Eppendorff tube. The sad thing is that garbage into a biosensor will
2
10 often give complex responses that users misinterpret as some interesting
0
88 binding event.
7
81 Ihavefoundthatwhenexperimentsaredesignedappropriatelywithgood-
7
9/9 quality reagents and data are processed and analyzed properly, binding
3
0 responsescanberoutinelyfittoasimpleinteractionmodel.However,unlike
1
10. DrPessimist,IdonotexpecttoobtainperfectbindingresponseswhenIset
oi: up experiments on a new interaction. I realize that obtaining high-quality
d
g | data is an iterative process. In my research group, we usually set up a trial
or
c. experimenttoverifythatthebindingpartnersactuallyinteract.Thenwewill
s
s.r often try different coupling chemistries, surface densities, and/or buffer
b
u
p conditions to optimize surface activity.
http:// And when it comes to the number one complaint about SPR technology
n (that the surface will automatically change the thermodynamics of the sys-
o
7 tem), what Dr Pessimist fails to realize is that most biosensor experiments
1
0
2 donot useaflatsurface.Instead,thesurfaceiscoated withadextranlayer,
y
Ma which suspends the molecule in solution. We and others have shown with
24 numerous systems that when experiments are performed properly, binding
on constants (including thermodynamic parameters) measuredwith SPR do in
d
he fact match those obtained from solution-based measurements.
s
bli However,Iagree withDrPessimistinoneregard.Since1991, Ihaveread
u
P everypaperthatreportedusingacommercialSPRbiosensorandDrRebecca
RichandIhavecomposedayearlyreviewoftheliteraturesince1998.Thisis
becoming a fairly daunting task since more than 1000 research papers are
published annually. More, unfortunately, is not always better. We find that
thedatainmostbiosensorarticlesarenotworththepapertheyareprinted
on. For example, about half the time authors even fail to present figures
showing the binding responses and yet they expect us to believe the rate
constants they report for their interactions. Without a visual inspection of
the data, we have no idea if the experiments were run properly. And often-
times,evenwhendataarepresented,itisclearthattheinvestigatorsdonot
know how to utilize the technology properly. Also, while a fundamental
dogma of science is to replicate and randomize samples, less than 3% of
published biosensor data include replicate injections even within a single
experiment. An overlay of replicate injections demonstrates the stability of
the reagents and multiple independent experiments yield an average and
standard deviation for the reported binding constants, yet this attention to
detailinabiosensorexperimentismorerarethanfindingafour-leafclover
intheoutfieldatFenwayPark.Inaddition,lessthan5%oftheauthorswho
View Online
Forewordtothe1stEdition vii
reportkineticconstantsincludeanoverlayofthebindingresponsewiththe
fittedmodel.Andfinally,evenfromabriefglancethroughtheliterature,itis
apparentthatthemajorityofinvestigatorsdonotunderstandthattheshape
oftheresponseprofileshouldbeanexponentialinboththeassociationand
dissociation phases (maybe many users do not even understand what an
05 exponential is). It is no wonder that scientists outside the biosensor use
0
P community think SPR technology does not work. I would think the same
F
83- thing if all I had to rely on was the published data.
2
10 Youmightaskyourself,‘‘howdiditgettothispoint?’’Ioftenwonderifall
0
88 scientists are so poorly educated in basic scientific technique (which could
7
81 actually explain why we haven’t found a cure for the common cold). I place
7
9/9 theblameonthe‘‘kitmentality’’thatwasintroducedwithmolecularbiology
3
0 backintheearly1990s,backwhenwewerelisteningtoourWalkmanswhile
1
10. typing on our IBM 286 personal computers. Nowadays you can buy a kit to
oi: clone,mutate,expressandpurifyaprotein.Well,thekitmentalitycontinued
d
g | when these same investigators got access to commercially available bio-
or
c. sensor technology. Since these instruments are so easy to use, anyone can
s
s.r walk up to the machine, chuck in their proteins, collect some response, fit
b
u
p the data and publish the results, believing that the results must be correct
http:// because they came out of this very expensive machine. Unfortunately, it
n actually takes some skill and know-how to set up, execute and analyze a
o
7 biosensor experiment properly.
1
0
2 Thisleadsmetothenextgroupofbiosensorusersthatgivethetechnology
y
Ma ablackeye.Thesepeoplearetheoneswhohavebeenusinginstrumentsfor
24 alongtimeandthinktheyareexperts.Icallthem‘‘SPiRts’’.SPiRtsareeven
on more threatening than Pessimists because their complacency often leads
d
he them to perpetuate poor experimental technique. A common SPiRt mistake
s
bli publishedintheliteratureistheuseofmultivalentanalytesinsolution(e.g.
u
P monoclonal antibodies or GST fusion proteins), which can produce avidity
effects.Alltoooften,SPiRtspresentelaboratebiologicaljustificationsforthe
shape of their unusual binding profiles when in fact the responses are
simply indicative of poor reagent quality and/or inadequate experimental
optimization or data processing. Even worse,SPiRts usecomplex modelsto
describe their poor-quality data. It seems that the latest fad of these model
surfers is to apply a conformational change mechanism. ‘‘My data fit a
conformationalchange model, which must mean there is a conformational
change, right?’’ Wrong!
To set the record straight, in 1994 my colleague and software engineer
extraordinaire, Tom Morton (who I refer to as SoftEE), developed the nu-
merical integration approach to data analysis that allows one to apply any
interaction model. Before then, we were in the caveman days of linear
transformation and, believe me, you don’t want to go back there. We were
the first to show that a change in conformation that stabilized a bound
complexwouldinfactproduceachangeinresponseeventhoughtherewas
no additional change in mass. However, in the intervening 13 years I have
never needed to apply this model to describe the responses obtained from
View Online
viii Forewordtothe1stEdition
more than 1000 systems I have examined. The reason I am reluctant to use
this model is that typically a data set that fits a conformational change
modelcanbeequallywelldescribedbyothermodelssuchasthoseforligand
and/or analyte heterogeneity. Even more alarmingly, the rates for the sup-
posed conformational changes measured on the biosensor are extremely
05 slow,oftenwithhalf-livesof20–60minutesifyoutakethetimetocalculate
0
P them.Theseratesdonotmakebiologicalsensetome.Aquicksearchofthe
F
83- classical conformational change literature shows that re-organizational
2
10 eventswhichoccurduringbindinghappenonananosecondtomillisecond
0
88 time-scale.Thehot‘‘new’’trendwiththeSPiRtsistofittheirbiosensordata
7
81 withaconformationalchangemodelandthenpresentcrystalstructuredata
7
9/9 of unbound and bound complexes and say ‘‘See, this change in conform-
3
0 ationprovesit’’.Butanobjectiveviewerwoulddisagree.Thefactthatyousee
1
10. a change in conformation in the structure actually may not relate to the
oi: complex binding response you are measuring on the biosensor. Don’t be
d
g | fooled by these sleight-of-hand arguments. (What would help confirm the
or
c. conformational change suggested by SPR would actually be to use a time-
s
s.r resolved structural method such as circular dichroism or fluorescence res-
b
u
p onance energy transfer and demonstrate that the time-dependent changes
http:// arethesame.)Thecauseofthecomplexbindingresponseonthebiosensor
n is actually more likely due to surface aggregation, nonspecific binding,
o
7 molecular crowding, avidity effects or sample heterogeneity.
1
0
2 ThisbringsmetomyfavoriteSPRusers,whoIrefertoasSPiRits.SPiRits
y
Ma arenewusersorthosehavingsomefamiliarityofbiosensortechnologywho
24 have a deep desire to learn more about its features, applications and po-
on tential. They are the ones who are participating in our yearly benchmark
d
he studies,whicharegearedtowardcalibratingusers’experimentaltechnique.
s
bli Theyarewillingtoputintheefforttotroubleshoottheirsystemsandwantto
u
P improvethequalityofthedataandnotjustsettleforwhateverthemachine
spits out. SPiRits will be the users who develop novel applications and im-
plement new technologies in the future.
We need SPiRits because the number and types of SPR instruments are
exploding. An Internet search reveals more than 20 companies developing
SPR-based biosensor systems. Lately, biosensor advances have occurred on
two fronts. First, many of the recently released instruments (and others
currentlyunderdevelopment)arededicatedtospecificapplicationsranging
from small-molecule drug discovery to the characterization of complex
mixturesintheclinicalandfoodsciences.Corning’sEpicplate-basedsystem
isanexampleoftargetingthetechnologyforscreeningapplications.Second,
we have seen a push to increase the throughput of biosensor analyses. In
the past few years, the launches of BioRad’s ProteOn XPR36 and Biacore’s
A100 have dramatically impacted the biosensor field since they allow for
parallel processing of multiple analytes over multiple targets simul-
taneously. Array-based platforms represent the next wave in biosensor
development. Biacore’s Flexchip and instruments being developed by GWC
Technologies,Lumera,IBISTechnologies,GenopticsandMavenopenupthe
View Online
Forewordtothe1stEdition ix
possibility of characterizing hundreds to thousands of interactions at one
time. But not surprisingly, these array formats come with their own sets of
challenges. The methods used for spotting DNA may not be optimal for
producing protein arrays. Clearly, a lot of work remains to be done before
protein array systems meet their full potential.
05 As biosensor applications expand and new instruments are released, the
0
P technology’s user base also increases. I worry that higher-throughput sys-
F
83- tems may allow more userssimply to generate more bad data faster. So, we
2
10 clearly need to improve the skill level of both novice and seasoned users.
0
88 This book is a great resource to obtain the fundamental knowledge of
7
81 biosensor technology, and also discover recent developments in both in-
7
9/9 strumentation and applications. But in order to turn professional, remem-
3
0 ber that the biosensor is just a tool. Use it wisely. Be skeptical, but keep an
1
10. open mind. Know when to say when (not all systems are amenable to bio-
oi: sensor analysis). Go forth and become a good ShePaRd of my favorite
d
g | technology.
or
c.
s
s.r David G. Myszka
b
u
p University of Utah
p://
htt
n
o
7
1
0
2
y
a
M
4
2
n
o
d
e
h
s
bli
u
P
0
1
0
P
F Foreword to the 2nd Edition
3-
8
2
0
1
0
8
8
7
1
8
7
9
9/
3
0
0.1 ItisagreathonorformetobewritingtheForewordtothe2ndeditionofthe
1
oi: HandbookofSurfacePlasmonResonance,notonlybecauseofthehighcaliber
g | d of the contributing authors, forming an ‘‘all-star cast’’ in the label-free
or biosensorfield,butalsobecausetheForewordtothe1steditionwaswritten
c.
s.rs by my post-doctoral mentor, Dr David Myszka, to whom I am grateful for
ub giving me the incredible opportunity to work in his laboratory, which
p
p:// launchedmycareerinSPR.InreadinghisForeword,Iespeciallyrelatetohis
htt comments regarding the love of SPR due to the truly awesome insight it
n
o offers into how molecules interact with one another and how this can be
7
01 observed in real time and without the need for labeling. Indeed, ‘‘seeing is
2
y believing’’! Biosensors are devices that use biological molecules such as
a
M
4 proteins to detect the presence of other molecules due to their binding
2
n interactions. This rather general definition sets ground zero for the myriad
o
d of applications that are now made possible on commercial biosensors,
e
h
s which employ SPR or other label-free detection methods, such as biolayer
bli
u interferometry (BLI), quartz crystal microbalance (QCM), electrical im-
P
pedance, or microcalorimetry.
I first learned about SPR in 1997 during my PhD studies when I was
seekingaquantitativemethodforassessingtheDNA-bindingpropertiesofa
panelofmutantNFkBconstructs,withtheproject’sgoalbeinggenetherapy.
I had turned to SPR because I was frustrated at the tedium and lack of
quantitation associated with other methods, such as gel shift assays. I was
instantly hooked by SPR’s real-time label-free method and to this day I am
passionate about SPR technologies and excited by the versatility of experi-
ments that can be performed quickly, with relative ease, and with minimal
sample consumption to answer basic to more exploratory questions about
molecular-levelinteractionanalysis.Intheearly1990s,theBiacore2000was
HandbookofSurfacePlasmonResonance,2ndEdition
EditedbyRichardB.M.Schasfoort
rTheRoyalSocietyofChemistry2017
PublishedbytheRoyalSocietyofChemistry,www.rsc.org
x