`
`
`
`Chemistry
`
`1
`
`TFS1031
`
`
`
`Bioconjugate
`Chemtstry
`
`Amencan Chemical Society
`1155 16th St.. N.W.
`Washtngton, DC 20036
`(202) 872-4600
`TOO (202) 452-2120
`FAX (202) 872-4615
`
`EDITOR-IN-CHIEF
`CLAUDE F. MEARES
`Department of Chemistry
`Room 190A
`University of California
`Davis, Califorma 95616
`(916) 752-3360
`FAX {916) 75?-8938
`
`·.
`
`ASSOCIATE EDITORS
`• '
`Paul S. Miller
`· · ... : ..
`The Johns Hopkins University · · '·
`Baltimore, Maryland 21205
`Peter Senter
`Bristol-Myers Squibb
`Seattle, Washington 98121
`
`ADVISORY BOARD
`
`Maciej Adamczyk
`Vernon L. Alvarez
`Lyle Arnold
`Robert W. Baldwin
`Hagan Bayley
`Donald Bergstrom
`PurshOiam Shan
`Gary 81gnam1
`Cynthia J. Burrows
`C. Allen Chang
`Chten-Hstng Ken Chang
`Jik Chtn
`Daniel J. Coughlin
`Masad J. Damha
`Gary S. Davtd
`Peter B. Dervan
`Jean F. Desreux
`David Dolphin
`Robert E. Feeney
`Raymond Firestone
`Alan R. Fritzberg
`Kieran F. Geoghegan
`David A. Goodwin
`Edgar Haber
`Takeshi Hara
`J. Milton Harris
`Richard Haugland
`M. Frederick Hawthorne
`John E. Hearst
`Sidney M. Hecht
`Ned D. Heindel
`Claude Helene
`Leaf Huang
`David K. Johnson
`John A. Katzenellenbogen
`John F. W. Keana
`Ban An Khaw
`
`Eiichi Kimura
`Jindrich Kopecek
`Nenad Kosbc
`Kenneth Krohn
`Robert L. Letsinger
`Joseph Loo
`Damon Meyer
`RICh B. Meyer
`Herbert T. Nagasawa
`Peter E. Nielsen
`Leslie Orgel
`William Pardndge
`Geoffrey A. P1etersz
`Glenn D. Prestwich
`Ralph Reisfeld
`Mark Reynolds
`Stephen Rokita
`Keith Rose
`Peter G. Schultz
`William H. Scouten
`Clay Siegal!
`Caroline Springer
`Su resh Snvastava
`Cy A. Ste1n
`Stanley Stein
`Francis Szoka
`Philip E. Thorpe
`Glen L. Tolman
`Vladimir P. Torchilin
`Janis Upeslacis
`E. Sally Ward
`DavidS. Watt
`Ron Wetzel
`D. Scott Wilbur
`Meir Wilchek
`Samuel Zalipsky
`Michael Zalutsky
`
`Btoconjugate Chemistry (ISSN 1043·1802) is
`published bimonthly by the American Chemical
`Society at 1155 16th Street, N.W., Washing(cid:173)
`ton, DC 20036. Periodicals postage paid at
`Washington , DC, and additional mailing
`offices. POSTMASTER : Send address
`changes to Bioconjugate Chemistry, Member
`& Subscriber Services, P.O. Box 3337, Colum(cid:173)
`bus, OH 43210.
`Canadian GST Reg. No. 127571347.
`Printed in the USA.
`
`Journal Publishing Operations
`Amencan Chemtcal Society
`2540 Olentangy River Road
`P.O. Box 3330
`Columbus, OH 43210
`(614) 447-3600, Ext. 3171
`TELEX 6842086; FAX (614) 447-3745
`
`Member & Subscriber Services
`American Chemical Society
`P.O. Box 3337
`Columbus, 'Oli 4321 o
`(614) 447-37'76; FAX (614) 447·3671
`(800) 333·9511 .
`
`PUBLICATIONS DIVISION
`Robe·rt D. Bovenschulte, Director
`Journal Publishing Operations: Mary E. Scanlan,
`Director
`Editorial Office: Anne C. O'Melia, Manager; Kath·
`leen E. Duffy, Journals Editing Manager; Sean
`Lammers, Assistant Editor
`
`Advertising Office: Centcom, Ltd., 676 East
`Swedesford Road, Suite 202, Wayne, PA 19087-
`1612
`© Copyright 1997 American Chemical Society.
`Copyright permission: Reprographic copying
`beyond that permitted by Section 107 or 108 of
`the US Copyright Act is allowed for mtemal use
`only, provtded that $14.00 per article IS paid to
`the Copyright Clearance Center (CCC), 222
`Rosewood Dr. , Danvers, MA 01923, USA.
`Republication or reproduction for sale of arttcles
`or abstracts in th1s journalts permitted only under
`license from ACS. Direct these and other per(cid:173)
`mission requests to the ACS Copynght Office at
`the ACS Washtngton address. For quotes and
`information on ordenng bulk reprints, call (202)
`872·4539 or write to the ACS Customer Service
`and Sales Office at the ACS Washington
`address.
`The paper used in th1s publication meets the
`minimum requirements of American National
`Standard for Information Sctences-Permanence
`of Paper for Pnnted Library Matenals. ANSI
`239.48·1984.
`
`Editorial Information
`Instructions for authors and the copyright
`status form are printed in the first issue of each
`volume. Please conform to these instructions
`when submitting manuscnpts. These instructions
`are available via the World Wtde Web by first
`selecting "Electronic Publications", then "Journal
`Home Pages", and finally the journal name at
`URL http://www.ChemCenter.org.
`Manuscripts for publication should be sub·
`milled to the Editor, Claude F. Meares, at his
`Davis, CA, address.
`Correspondence regarding accepted papers
`and proofs should be directed to Journal Pub·
`lishing Operations at the address above.
`The American Chemical Society and its Edi(cid:173)
`tors assume no responsibility for the statements
`and opinions advanced by contributors.
`Registered names and trademarks, etc., used
`in th1s publication, even without specific indication
`
`lh'""'· "' "''"' ,, "'"';'""' ""~"""' by l
`
`law.
`At the end of each document is a umque
`9-character document number that serves as a
`link between the printed and electronic (CJACS
`Plus Images) products and facilitates the retrieval
`of the document in electronic form.
`
`Subscription and Ordering Information
`
`All
`Canada
`Other
`and
`MexiCo Europe· Counlnes·
`
`US
`
`S 35
`S344
`
`S 41
`$350
`
`S 41
`$350
`
`Printed
`S 54
`S 75
`S 67
`Members 1 yr S 35
`S384
`$376
`$363
`$344
`Nonmembers
`Sludenl Members S 26 25 S 45.25 S 58.25 S 66 25
`Microfiche
`Members-! yr S 35
`S344
`Nonmembers
`• A1r serviCe 1ncluded
`New and renewal subscriptions should be
`sent with payment to American Chemical Soc1ety,
`Department L-0011 , Columbus. Ohio 43268·
`0011. Members may share/donate their personal
`subscriptions with/to libranes and the like but only
`after 5 years from publication. Microform, back
`issue, and printed edition single issue orders
`should be sent to Microforms & Back Issues
`Office at the ACS Washington address. Non(cid:173)
`member subscribers in Japan must enter sub·
`scription orders with Maruzen Company Ltd., 3·10
`Nihonbashi 2-chome. Chuo-ku, Tokyo 103.
`Japan. Tel: (03) 272-7211 .
`Claims for issues not received will be honored
`only if submitted within 90 days of the issue date
`(subscribers in North America) or within 180 days
`of the issue date (all other subscribers). Claims
`are handled by Member & Subscriber Services.
`Supporting Information noted on the table of
`contents with a •
`is available as photocopy
`($1 2.00 for up to 3 pages and $1.50 per page for
`each additional page, plus $2.00 for foreign post·
`age). as 24x microfiche ($12.00, plus $1.00 for
`foretgn postage), or in electronic form via the
`Internet (free to subscribers). See paragraph
`below for instructions to access the information
`electrontcally. Canadian residents should add
`7°o GST. For photocopy or microfiche orders,
`see Supporttng lnformatton notice at the end of
`JOurnal article for number of pages. Orders must
`gtve complete title of article, name(s) of author(s),
`JOurnal. 1ssue date, and page numbers. Prepay·
`ment IS reqUJred and prices are subject to
`change. For information on microforms, contact
`Microforms & Back Issues Office at the ACS
`Washington address or phone (202) 872-4554.
`Supporting Information except structure factors
`also appears in the mtcrofilm edition.
`Supporting Information noted on the table of
`
`contents with a * is available tn electrontc form
`
`only. See the followtng paragraph for access
`information.
`Supporting Information 1s also available via the
`World Wtde Web. WWW users should select
`"Electronic Publications", then "Journal Home
`Pages", and finally the journal name at URL
`http://www.ChemCenter.org. Detatled instructions
`for using this service, along with a descriptiOn of
`the file formats, are available at this site. To
`download the Supporting Information, enter the
`journal subscription number from your mailing
`label. For additional information on electronic
`access, send electronic mail to si-help@acs.org
`or phone (202) 872-6333.
`
`2
`
`
`
`552
`
`Bioconjugate Chem. 1997, 8, 552- 559
`
`Biotin- Fluorophore Conjugates with Poly(ethylene glycol) Spacers
`Retain Intense Fluorescence after Binding to Avidin and
`Streptavidin
`
`Hermann J. Gruber,* Markus Marek, Hansgeorg Schindler, and Karl Kaiser
`
`Institute of Biophysics, J. Kepler University, Altenberger Strasse 69, A-4040 Linz, Austria. Received
`February 14, 1997°
`
`Conventional biotin- fluorophore conjugates with ~ 14 atom spacers lose most of their fluorescence
`when binding to avidin or streptavidin, as is demonstrated in the present study. This explains the
`unusual fact that only biotinylated marker enzymes, but not fluorescent biotins, are regularly used
`in bioanalytic assays. Novel biotin- spacer-fluorophore conjugates are presented that retain intense
`fluorescence when binding to avidin or streptavidin. Preservation of fluorescence depends upon the
`use of poly(ethylene glycol) (PEG) spacers, which are shown not to interfere with biotin function.
`The observed absence of nonspecific interactions may also be attributed to the PEG chain. These
`novel fluorescent biotins are expected to be excellent new tools in fluorescence microscopy and related
`techniques.
`
`INTRODUCTION
`
`Specific detection of immobilized biomolecules is a
`standard task in modern bioscience. Generally target
`molecules are recognized by specific probe molecules
`(antibodies, oligonucleotides, etc. ) which are labeled for
`detection (Wilchek and Bayer, 1990a). Direct labeling
`of a probe with a marker function (fluorophore, enzyme,
`etc. ) implies an irreversible restriction to a single detec(cid:173)
`tion method. Biotinylation of a probe, however, allows
`the use of the same probe in combination with almost
`any known detection method because a wide variety of
`biotinylated markers or (strept)avidin-marker conju(cid:173)
`gates is commercially available for the postlabeling of
`biotinylated probes via biotin-(strept)avidin-biotin
`bridges or via biotin-(strept)avidin bridges, respectively
`(Wilchek and Bayer, 1988, 1990b).
`Fluorescent markers play a pre-eminent role in bio(cid:173)
`analytic assayss; therefore, it seems logical to expect
`extensive use of fluorescent biotins. However, conven(cid:173)
`tional biotin- fluorophore conjugates (with ~ 14 atom
`spacers) lose most of their fluorescence when binding to
`(strept)avidin, as is demonstrated in the present study.
`This adverse effect explains why only two fluorescent
`biotins are commercially available-and mostly used for
`purposes other than postlabeling of biotinlyated probes
`(Chu et al., 1994; Schray et al., 1988; Shah et al., 1994).
`In the present study a series of novel biotin-spacer(cid:173)
`fluorophore conjugates is presented that retain intense
`fluorescence after binding to (strept)avidin. Preservation
`1 or PEG800
`of fluorescence depends on the use ofPEG1900
`as spacer elements: The PEG chains minimize dye- dye
`and dye-protein interactions, which cause the quenching
`in complexes of conventional fluorescent biotins with
`(strept)avidin.
`While PEG spacers in biotin- PEG- fluorophore con(cid:173)
`jugates are beneficial to the fluorophore function, the
`question remained to be answered whether the PEG
`spacers would hinder the biotin terminus from binding
`
`* Author to whom correspondence should be addressed [tele(cid:173)
`phone + 43 (732) 2468-9271; fax +43 (732) 2468-822; e-mail
`hermann.gruber@jk.uni-linz.ac.at].
`0 Abstract published in Advance ACS Abstracts, June 15,
`1997.
`
`to (strept)avidin. In the preceding study with nonfluo(cid:173)
`rescent biotin-PEG conjugates (see the first of three
`papers in this issue) it has been demonstrated that 4:1
`complexes with (strept)avidin were indeed formed and
`that at least three biotin- PEG elements per tetrameric
`receptor protein were bound on a time scale of hours. The
`present study shows that nearly four biotin-PEG(cid:173)
`fluorophore ligands/protein remain bound on a time scale
`of minutes, as is desired in bioanalytic fluorescence
`detections.
`
`EXPERIMENTAL PROCEDURES2
`
`Materials. P.a. grade materials were used as far as
`commercially available. Affinity-purified avidin and
`streptavidin, d-biotin, Boc20, and DACA were obtained
`from Sigma. DMF and Et3N were purchased from Fluka.
`NaCl and methanol were obtained from Riedel de Haen.
`1,12-diamino-4,9-dioxadodecane was obtained from Ald(cid:173)
`rich. Acetic acid and chloroform were purchased from
`Baker. fluorescein-biotin (see footnote 1 for full struc(cid:173)
`ture), ANS, 5-(and 6)-carboxyfluorescein succinimidyl
`ester, and 5-(and 6)-carboxytetramethylrhodamine suc(cid:173)
`cinimidyl ester were obtained from Molecular Probes. Cy3
`
`1 Abbreviations: ANS, 2-anilinonaphthalene-6-suJfonic acid;
`biotin, biotinoyl group; biotin- NHS, succinimidyl ester ofbiotin;
`biotin - dode-TMR, 5-(and 6)-[[N-(12-biotinamido-4,9-dioxa(cid:173)
`dodecyl)]aminocarbonyl]tetra methylrhodarnine; biotin- NH(cid:173)
`PEGsoo or 1900-NH2· HCl, N-biotinoyl-0,0' -bis(2-aminopropyl)poly(cid:173)
`(ethylene glycoDsoo or 1900 hydrochloride; biotin- PEGsooor 1900-
`dye, see Scheme 1; biotin- PEGsoo or 1900-Flu, see Scheme 1;
`biotin- PEGsoo or 1900-TMR, see Scheme 1; biotin- PEGsoo or 1900-
`Cy3, see Scheme 1; biotin- PEGsooorl900-Cy5, see Scheme 1;
`Boc20, di-tert-butyl pyrocarbonate; Boc, tert-butyloxycarbonyl
`group; DACA, p-(dimethylamino)cinnamaldehyde; DMF, N,N(cid:173)
`dimethylforma mide; EDTA, ethylenediamine-N,N,N',N'-tet(cid:173)
`raacetic acid; Et3N, N,N,N-triethylamine; Flu, 5-(and 6)(cid:173)
`carboxyfluorescein res idue; fluorescein- biotin, 5-[[N-[5-[N-[6-
`(biotinoyl)amino]hexanoyl]amino]pentyl]thioureidyl]fluore(cid:173)
`scein; (strept)avidin, streptavidin and/or avidin; FRET, fluores(cid:173)
`cence resonance energy transfer; NHS, N-hydroxysuccinimide;
`PEG, poly(ethylene glycol); RT, room temperature; TMR, 5-(and
`6)-carboxytetramethylrhodamine residue.
`2 Detailed procedures and NMR data can be found in the
`Supporting Information.
`
`51043·1802(97)00087·6 CCC: $14.00 © 1997 American Chemical Society
`
`(
`
`3
`
`
`
`Biotin-Fiuorophore Conjugates
`
`Bioconjugate Chem., Vol. 8, No. 4, 1997 553
`
`and Cy5 monofunctional dyes (succinimidyl esters) were
`purchased from Amersham. Sephadex-based gels were
`obtained from Pharmacia. All other materials were
`purchased from Merck. Biotin-NHS was prepared as
`described (Wilchek and Bayer, 1990c). Absolute DMF
`was prepared according to a standard procecdure. Bi(cid:173)
`otin- NH-PEGaoo-NH2·HCl and biotin-NH-PEG1900-NH2·(cid:173)
`HCl were synthesized as described before (see the first
`of three papers in this issue).
`Buffers. Buffer A contained 100 mM NaCl, 50 mM
`NaH2P04 , and 1 mM EDTA, adjusted to pH 7.5 with
`NaOH. Buffer B contained 1 mM NaH2P04 (adjusted to
`pH 7.5 with NaOH) and variable NaCl concentrations
`(as specified).
`Methods. Synthesis ofN-Boc-4,9-dioxa-1,12-diamino(cid:173)
`dodecane·CHJCOOH . 4,9-Dioxa-1,12-diaminododecane
`(29.4 mmol) in methanol was reacted with 28.4 mmol of
`Boc20 under Ar. After addition of toluene and 5 mL of
`acetic acid, the mixture was taken to dryness (10.5 g of
`crude product). Silica chromatography of 3 g of crude
`product in chloroform/methanol/acetic acid mixtures (90:
`10:0.1 and 70:30:5) gave 3.20 mmol of product, which was
`pure by TLC.
`Synthesis of N-Boc-N'-biotin-4,9-dioxa-1,12-diamin(cid:173)
`ododecane. N-Boc-4,9-dioxa-1, 12-diaminododecane· HCl
`(1.8 mmol) was reacted with 2.7 mmol ofbiotin-NHS and
`240 _uL ofEt3N in 10 mL of absolute DMF. Excess biotin(cid:173)
`NHS was hydrolyzed with water. After solvent removal,
`the crude product was purified on silica 60 (eluent
`chloroform/methanol/acetic acid 120:30:0.5), resulting in
`1.11 mmol of product (pure by TLC).
`Synthesis of N-Biotin - 4,9-dioxa-1,12-diaminododecane·(cid:173)
`HCl. N-Boc-N' -biotin-4,9-dioxa-1,12-diaminododecane
`(1.11 mmol) was deprotected with 98% formic acid, and
`the crude product was purified by ion exch ange chroma(cid:173)
`tography on SP Sephadex C-25. Salt was removed by
`extracting deprotonated product into chloroform. Drying
`and lyophilization from dilute HCl gave 0.64 mmol of
`product (pure by TLC).
`Synthesis of Biotin-dode - TMR. N -Biotin-4,9-dioxa-
`1,12-diaminododecane·HCl (19 _umol) was reacted with
`30 _umol of 5-(and 6)-carboxytetramethylrhodamine suc(cid:173)
`cinimidyl ester and 10 _uL of Et3N in chloroform under
`Ar. TLC showed quantitative labeling of the primary
`amine. The mixture was taken to dryness and purified
`by chromatography on silica 60 (chloroform/methanol/
`water 70:26:4). In spite of using 100 g of silica 60 it was
`impossible to remove those two TMR derivatives, which
`were already present in the commercial TMR reagent and
`whose Rr values were just below (0.45) or above (0.56)
`that of the product (0.51). For characterization, TLC
`spots were harvested quantitatively, extracted with
`chloroform/ethanol/water (10:15:2), clarified by centrifu(cid:173)
`gation, and checked for TMR contents by their UV - vis
`spectra. The two byproducts together gave rise to 17%
`of the absorption at 550 nm, while the correct main
`product contributed with 83% to ~. From determina(cid:173)
`tion of biotin end group contents (by the ANS method,
`see below) a similar estimate of purity was obtained (87%
`as compared to the TMR group contents estimated from
`UV - vis absorption).
`Synthesis ofBiotin- PEGfJ()I;-TMR. Biotin- NH-PEGaoo(cid:173)
`NH2·HCl (8.4 _umol) was quantitatively labeled with 23
`_umol of 5-(and 6)-carboxytetramethylrhodamine succin(cid:173)
`imidyl ester in chloroform!Et3N under Ar. After evapora(cid:173)
`tion, the crude product was gel filtered on Sephadex G-25
`in distilled water, yielding 4.8 _umol of biotin- PEGsoo(cid:173)
`TMR according to biotin end group assay (see below).
`Quantitative TLC (as performed with biotin- dode- TMR,
`
`see above) showed that free TMR gave rise to 2% of the
`absorption at 550 nm, while the product contributed 98%
`to A55o·
`Synthesis of Biotin-PEG1900- TMR. The procedure
`was the same as for the PEGsoo derivative. Forty-three
`milligrams of biotin-NH-PEG1900-NH2·HCl (nominally
`18.7 _umol, - 20% water content) was reacted with 36
`_umol of 5-(and 6)-carboxytetramethylrhodamine succin(cid:173)
`imidyl ester. Yield= 12.8_umol ofbiotin-PEG1900- TMR
`according to biotin end group assay (see below). Quan(cid:173)
`titative TLC (as performed with biotin- dode-TMR, see
`above) showed that free TMR gave rise to 3% of the
`absorption at 550 nm, while the product contributed 97%
`to Asso.
`Synthesis of Biotin- PEG8oo-Flu. Twenty-two micro(cid:173)
`moles of biotin- NH-PEG800-NH2·HC1 was labeled with
`43 _umol of 5-(and 6)-carboxyfluorescein succinimidyl ester
`in DMF/Et 3N under Ar. After solvent removal, the
`residue was dissolved in chloroform and successively
`washed with 200 mM Na2C03 (saturated with NaCl) and
`with dilute acetic acid (saturated with NaCI). The
`organic layer was dried, evaporated, redissolved in 2 mL
`of3 mM Na2C03 , and chromatographed on QAE Sepha(cid:173)
`dex A-25 using buffer B with increasing NaCl concentra(cid:173)
`tions. Salt was removed by extraction into chloroform.
`Yield = 18 _umol, pure by TLC.
`Synthesis ofBiotin - PEGt9oo-Flu. Fifty milligrams of
`biotin- NH-PEG1900-NH2·HCl (22_umol) was reacted with
`15.5 mg of 5-(and 6)-carboxyfluorescein succinimidyl ester
`(33 .umol). The procedure was the same as for the PEG800
`homologue, except that 50 mM instead of 150 mM NaCl
`in buffer B was used to elute the product from QAE(cid:173)
`Sephadex A-25. Yield = 12 _umol product, pure by TLC.
`Synthesis of Biotin -PEG800- Cy3. Biotin- NH-PEG800-
`NH2· HCl (1. 7 _umol) was reacted with one vial of Cy3
`monofunctional dye ("reactive dye to label 1 mg of
`antibody" according to Amersham) in absolute DMF/
`Et3N. After solvent removal, the residue was dissolved
`in buffer B and chromatographed on QAE Sephadex
`A-25, using 125 mM NaCl in buffer B for product elution.
`Yield = 63 nmol determined from A 550 ( E550 = 150 000
`M- 1 cm- 1 for Cy3 according to Amersham) or 54 nmol
`according to biotin end group assay with ANS (see below).
`Th e uncoupled dye was eluted from QAE-Sephadex A-25
`with 1 M NaCl and amounted to 74 nmol according to
`Asso.
`Synthesis ofBiotin-PEG1900- Cy3. The procedure was
`the same as for the PEG800 derivative, except that 0.44
`_umol of biotin- NH-PEG1900-NH2·HCl was reacted, a nd
`75 mM NaCl in buffer B was used for elution from the
`ion exchange column. Yield = 55 nmol determined from
`A 550 or 52 nmol according to biotin end group assay with
`ANS (see below).
`Synthesis ofBiotin- PEGt900-Cy5. The procedure was
`the same as for the corresponding Cy3 analogue, except
`that 5 mg of biotin-NH-PEG1900-NH2• HCl (2.2 _umol) was
`reacted with one vial of Cy5 monofunctional dye. Yield
`=55 nmol determined fromA647 (E647 = 250 000 M- 1 cm- 1
`for Cy5 according to Amersham) or 39 nmol according to
`biotin end group assay with ANS (see below).
`Quantitative Assay for Biotin End Groups and fo r
`Biotin Binding Sites. A published fluorescence assay for
`avidin- biotin interaction (Mock and Horowitz, 1990) was
`modified as described before (see the first of three papers
`in this issue). Typically, 2 mL of buffer A containing 1
`nmol of avidin was mixed with 20 _uL of 5 mM ANS, and
`"-2 nmol (estimated from UV- vis absorption) of a biotin(cid:173)
`fluorophore conjugate was added to saturate about half
`of the receptor sites on avidin. The unoccupied sites were
`then titrated with 5 _uL increments of exactly 100 _uM
`
`4
`
`
`
`554 Bioconjugate Chern. , Vol. 8, No. 4, 1997
`
`d-biotin in buffer A while the fluorescence of the pseudo(cid:173)
`ligand ANS was monitored at 328 nm excitation (10 nm
`slit) and 408 nm emission wavelength (10 nm slit). The
`inflection point in the titration curve indicated saturation
`of all biotin binding sites (determined by parallel stan(cid:173)
`dardization of the avidin stock solution with d-biotin
`alone). Fortunately, none of the biotin- fluorophore
`conjugates caused significant background fluorescence
`under these assay conditions.
`The ANS assay was also used to determine functional
`biotin binding sites in avidin stock solutions as described
`before (see the first of three papers in this issue), whereas
`streptavidin was functionally characterized by titration
`with biotin-PEG800- pyrene (see the third of three papers
`in this issue) because streptavidin is known not to bind
`the pseudoligand ANS (Mock and Horowitz, 1990).
`Gel Filtration Assay for Binding of Biotin- PEG-Dye
`to Avidin and Streptavidin. Specificity and metastability
`of biotin- PEG- dye binding to avidin or streptavidin was
`tested by gel filtration as previously described (see the
`first of three papers in this issue). Typically, 0.5 mL
`samples of buffer A containing 50 ,uM "functional" avidin
`(see above) or 2 ,uM "functional" streptavidin (see above)
`and various amounts ofbiotin- PEG- dye were incubated
`for 1 h at 25 oc and subjected to gel filtration on a 1 x
`48 em column of Sephadex G-100 at RT. Elution was at
`0.25 mUmin with buffer A while fractions were collected
`at 5 min intervals. All fractions were assayed for avidin
`or streptavidin by A282 (corrected for E2a2 = 0.08Esso or
`0.20Esso or 0.18E496 in the case ofCy3-, TMR-, and Flu(cid:173)
`PEG conjugates, respectively) and for dye contents by
`UV - vis absorbance at Amax· Molar extinction coefficients
`for avidin and streptavidin were taken from the literature
`(Green, 1990).
`Measurement of Fluorescence in Complexes of Biotin(cid:173)
`PEG- Dye with Avidin and Streptavidin. In a "forward
`titration" 2 mL of receptor protein (::5 80 nM) in buffer A
`was titrated with a stock solution of biotin-PEG-dye
`(7- 15 ,uM). In the "reverse titration" mode 2 mL of
`fluorescent ligand (::5320 nM) in buffer A was titrated
`with stock solutions of avidin or streptavidin (2-4 ,uM).
`All fluorescence signals were corrected for the small
`dilution factors. Time intervals were 3- 5 min as re(cid:173)
`quired for equilibration at RT (except for titration of
`streptavidin with fluorescein-biotin, for which 10 min
`intervals were required). The concentrations of receptor
`proteins and ligands refer to biotin binding sites and
`biotin termini, respectively, as determined by functional
`titrations (see above).
`
`RESULTS
`Syntheses of Biotin-Spacer- Dye Conjugates.
`The goal of the present study was the identification of
`biotin-spacer-fluorophore structures that retain high
`fluorescence yield when binding to avidin or streptavidin.
`PEG1900, PEG800, and a 14-atom homologue were chosen
`as spacers, and the most popular/promising fluorochrome
`labels were tested (see Scheme 1).
`Biotin-PEG-dyes with anionic fluorophores could be
`subjected to ion exchange chromatography, resulting in
`homogeneous products as evidenced by TLC. In contrast,
`special precautions were necessary to arrive at 97- 98%
`purity with the zwitterionic TMR derivatives (see Ex(cid:173)
`perimental Procedures). Purity and 1:1 ratios of biotin/
`fluorophore end groups were also confirmed by 1H NMR.
`Virtually noise-free single-pulse spectra were recorded
`at 500 MHz to obtain correct integrals from unsaturated
`signals. Moreover, the specific bindability of all biotin(cid:173)
`PEG- dyes to (strept)avidin was close to 100%, as deter(cid:173)
`,mined by gel filtration assays (compare Figure 5) which
`
`Gruber et al.
`
`Scheme 1. Synthesis and Structure of Biotin- Fluoro(cid:173)
`phore Conjugates
`A: biotin-NH~O~O~NH-TMR
`
`Biotin-dode-TMR
`
`B:
`
`biotin-NH~ ~0~
`I L 0
`J
`NH2
`n= 18 or43
`
`biotirrNftPEG·N~
`
`! Dye-NHS
`biotin-NHYi ~0~
`-j
`NH-Dye
`n=18or43
`
`0
`
`Biotin-PEG-Dye
`
`Biotin-PEG·HJ (Dye= cart>oxyfll.oresceine)
`Biotin-PEG-TMR (Dye= cart>oxytetramethytrtlodamine)
`Biotin-PEG-Cy3 (Dye= Cy3 monofulctional dye)
`Biotin-PEG-CyS (Dye= CyS monofl.nctional dye)
`
`further excluded the presence of fluorescent molecules
`without a biotin terminus.
`One short fluorescent biotin (fluorescein- biotin) with
`a 14-atom spacer was commercially available. For a more
`systematic study a second example was synthesized in
`which biotin is linked to TMR via a 14-atom spacer also
`(see Scheme 1). In spite of moderate purity the product
`fully served its intended role as a poor fluorescent ligand
`for avidin and streptavidin (see Figures 2 and 9).
`Fluorescence Properties ofBiotin- Spacer- Fluo(cid:173)
`rophore Conjugates before/after Binding to Avidin
`or Streptavidin. The superiority of novel biotin-PEG(cid:173)
`dyes over conventional fluorescent biotins is demon(cid:173)
`strated in Figures 1 and 2. Short fluorescein- biotin is
`highly quenched when binding to avidin (Figure 1B, open
`squares) or streptavidin (Figure 1C, open squares), and
`the same is true for short biotin- dode- TMR (Figure 2C,
`circles). Avidin and streptavidin are tetravalent receptor
`proteins for biotin; therefore, the abrupt rise in fluores(cid:173)
`cence at ligand/receptor ratios >4:1 indicates stoichio(cid:173)
`metric binding, and the absence of nonspecific binding
`is evidenced by the strictly parallel nature of this linear
`rise with the linear dose response in the absence of
`receptor protein. Saturation at 4:1 stoichiometry and
`absence of further binding were also observed with
`biotin- PEG- dyes (see Figures 1, 2A,B, and 3A) except
`for the Cy3 derivatives (see Figures 3B,C and 4).
`In contrast to short fluorescein- biotin, the long biotin(cid:173)
`PEG1900- Flu showed little quenching when bound to
`avidin (Figure 1A). More quenching was observed with
`the intermediate conjugate biotin-PEG8oo-Flu when
`bound to avidin (Figure lB, circles) or streptavidin
`(Figure 1C, circles), but the fluorescence signals in 4:1
`complexes with receptor protein were still very intense
`in comparison to those of short fluorescein-biotin (Figure
`1B,C, open squares).
`Biotin- PEG-TMR conjugates (Figure 2A,B) differ
`from corresponding fluorescein analogues in two as(cid:173)
`pects: The effect of PEG chain length is much less
`pronounced, and the fluorescence quenching in the bound
`state is increased, reaching ~50% in the 4:1 complexes
`with avidin and streptavidin. Taking into account the
`zwitterionic nature ofTMR and the smaller Stokes shift
`as compared to fluorescein, it is not surprising that self-
`
`5
`
`
`
`Biotin-Fiuorophore Conjugates
`
`Bioconjugate Chern., Vol. 8, No. 4, 1997 555
`
`[Biotin-spacer-Flu] I [avidin]
`0
`6
`2
`4
`
`8
`
`[Biotin-spacer-TMR) I [strept-/avidin)
`0
`2
`4
`6
`8
`
`100
`
`B c:
`8 tJl
`~
`0
`::J
`c;::
`
`80
`
`60
`
`40
`
`20
`
`0
`
`60
`
`40
`
`20
`
`0
`
`80
`
`60
`
`40
`
`20
`
`0
`
`8
`
`c
`
`•• 0
`••
`"
`
`.•
`
`0 '
`
`..
`•
`• 0
`I
`• 0
`•• 0
`
`.. o
`
`. . .. :AJ
`
`• ·' 0
`.. 0
`••
`..
`• p
`
`0
`
`0
`
`0
`
`I •
`•• ,
`
`0
`
`•
`•
`0
`• . eeo
`~'\~~!!
`'
`o a
`60~----~----~~--+-----1
`
`8
`
`80
`
`60
`
`~ c
`~ 40
`~
`0
`~ 20
`
`c
`
`50
`
`40
`
`30
`
`20
`
`+
`
`I
`
`e
`00
`
`. I
`•'
`0
`
`•
`
`Ill
`
`Ill
`
`m m
`
`Ill •
`
`Ill
`
`!! e
`88
`
`Ill
`
`10
`!! !!
`Ill
`•
`0 o oooooooo
`
`8
`6
`4
`2
`0
`[Biotin-spacer-Flu) I [streptavidin]
`Figure 1. Fluorescence of biotin- spacer - fluorescein conju(cid:173)
`gates in the presence (open symbols) and absence (controls, solid
`symbols) of avidin or streptavidin. Excitation was at 485 nm (5
`nm slit), and emission was at 525 nm (5 nm slit). (A) 50 nM
`avidin was titrated with 9.2,uM biotin- PEG1ooo-Fiu. (B ) 50 nM
`avidin was titrated with 14.7 .uM biotin- PEGaoo- Flu (circles)
`or with 12.2 .uM fluorescein-biotin (squares). (C) 55 nM strepta(cid:173)
`vidin was titrated with 14.7 p M biotin- PEGaoo-Flu (circles) or
`with 12.2pM fluorescein- biotin (squares). In the presence of 4
`Jt M d-biotin neither avidin nor streptavidin had an effect on any
`of the biotin-spacer- fluorescein cof\iugates (not shown to avoid
`overloading of the figures).
`
`association and/or self-quenching by FRET will be fa(cid:173)
`vored in TMR. Nevertheless, biotin- PEG- TMR conju(cid:173)
`gates (Figure 2A,B) compare well with short fluorescent
`biotins (open squares in Figure 1B,C, circles in Figure
`2C).
`The dark red-emitting conjugate biotin- PEG1900-Cy5
`showed much similarity with biotin- PEG1900- Flu in that
`fluorescence was mostly preserved after binding to avidin
`(Figure 3A). The small reduction just allowed visualiza(cid:173)
`tion of the 4:1 complex formation.
`Best results were obtained with biotin-PEG-Cy3
`conjugates (Figures 3B,C and 4). The optimal case is
`represented by biotin- PEG1900- Cy3, where fluorophore
`performance was virtually independent from binding to
`avidin (Figure 3B) while biotin- PEG800-Cy3 seemed
`"more than optimal" because binding to avidin enhanced
`fluorescence yields above the control values observed in
`the absence of avidin (Figure 3C). This enhancement was
`further verified by inverse titration of a constant fluo(cid:173)
`rescent ligand concentration with a concentrated avidin
`stock solution (Figure 3D). In contrast to avidin, strepta(cid:173)
`vidin caused little increase in the fluorescence of biotin(cid:173)
`PEGaoo-Cy3 (Figure 4).
`Specificity of Biotin-PEG- Dye Bind ing to Avi(cid:173)
`din and Streptavidin. While emphasizing the out-
`
`8
`
`6
`4
`2
`0
`[biotin-spacer-TMR) I [strept-/avidin)
`Figure 2. Fluorescence ofbiotin-spacer-TMR cof\iugates in
`the presence and absence (controls, • > of avidin or streptavidin.
`Excitation was at 540 nm (5 nm slit) and emission at 580 nm
`(slit 5 nm except for biotin- dode- TMR, for which a 10 nm slit
`was used). 80 nM avidin (e ) or streptavidin (0) was titrated
`with 10.0 ,uM biotin- PEGlooo-TMR (A), 8.2pM biotin- PEGaoo(cid:173)
`TMR (B ), or 22,uM biotin - dode- TMR (C; 17% of the TMR labels
`had no biotin terminus-see Experimental Procedures-therefore,
`17% of the corresponding control signal in the absence of protein
`was subtracted from the signal in the presence of protein to
`estimate the fluorescence that originates from biotin- dode(cid:173)
`TMR). In the presence of 8 ,uM d-biotin neither avidin (t:.) nor
`streptavidin (+) had an effect on any of the biotin- spacer (cid:173)
`fluorescein cof\iugates.
`
`standing fluorescence properties of biotin-PEG800- Cy3,
`we have so far ignored a puzzling feature of the titration
`curves in Figures 3C and 4: They are obviously triphasic.
`At ligand/receptor ratios <4:1 the slopes are steeper than
`normal, between 4 and 8 ligands/receptor protein the
`slopes are flater than normal, and above 8 ligands/
`receptor the linear increases appear parallel to control
`series in the absence of protein (see lines in Figure 4).
`We are thus confronted with an apparent 8:1 stoichiom(cid:173)
`etry between biotin- PEG800- Cy3 and avidin (Figure 3C)
`or streptavidin (Figure 4), which has never been reported
`for any other biotin derivative. Yet binding of all 8
`ligands fully depends on specific interaction with biotin
`end groups, as is evidenced by complete block with
`d-biotin (see triangles in Figures 3C and 4).
`Fortunately, ligand