Fluorescence Sensing of a Ribonucleoside 5′-Triphosphate

@

Tetrahedrontitters, Vol. 38, No. 36, pp.6323.6326, 1997 01997 ElsevierScienceLtd

Pergamon

All rights reserved. %irrted in Great Britain 0040-4039/97 $17.00+ 0.00

PII: S0040-4039(97)01454-8

Fluorescence Sensing of a Ribonucleoside 5’-Triphosphate Steven Patterson, Bradley D. Smith*, and Richard E. Taylor* Departmentof Chemistryand Biochemistry,Universityof Notre Dame,NotreDame,Indiana46556,USA

Abstract: Conjugation of a fluorescent srrccharidesensor to poly(allylamine) produces a sensing compound with selectivity for riborrucleoside5’-triphosphate. An associationconstant (log Ka) of 5.6 was determined for uridine 5’-triphosphatein pH 7.3 phosphatesolution (15 mM, 298 K), whereas log Ka waaless than 2 for uridine,uridine5’-monophosphate, and undine 5’-diphosphate. 01997 Elsevier Science Ltd.

A primaryaimin abioticsupramolectdarchemistryis to inventmoleculardeviceswithspeeiticfunctions suchas recognitionand catalysis. The majorityof casesreportedin the literaturedescribea compoundthathas promisingattributes but there is a need to convertthe lead candidateinto a second-generationversion with improvedbindingaffinityand/orbindingselectivity. At presentthere are few waysto addressthis problem.] Wehaveinitiateda programto developgeneralmethodsof improvingreceptorbindingthroughthe incorporation of a tunableseeondaryrecognitiondomain. As a startingpoint,we havechosento exploremethodsthat expand upon the binding characteristicsof boronic acids which are knownto form covalent complexeswith diolcontairtingmoleculessuchas carbohydrates.2~3 We have modifiedthe carbohydratereceptor,1, inventedby Shinkaiand coworkers.z.gThis receptor design was chosenfor a numberof reasons. It binds neutralmonosaccharidesin aqueousmethanolsolution witha selectivityfor fructose(Ka= 103M-1).2Compound1 is fluorescentand is thoughtto signalbindingvia modulationof a photoinducedelectrontransferquenchingmechanism.zAlthoughfluorescenceprovideslimited informationabout the molecular details of binding, it is nonethelessa straightforwardway of measuring association.Tbisis an importanttechnicalpointwithsugarreceptorsbecausemostcarbohydrateslackan easily observablechromophore.

\\\,, Tm c @ 1,

1;

;

;

N

B(OH)2

T #

ti

1>

1 R=H 2 R = 4-C#14C02H

o

t.la+0-

1;

3

In this report we describethe versatilederivative2, whichcontainsa benzoic acid linker groupthat allowsconjugationto a rangeof secondmybindingfunctionality.As a demonstrationof its versatility,we have attached 2 to a cationic polyamineand produceda fluorescentsensor with greatly enhanced affinity for a ribonucleoside5’-triphosphate.Nucleotidesareimportantbiologicalcompoundsandthereis a need to develop fluorescentnucleotidesensors.sTo date,onlya fewattemptshavebeenreported.6.7 6323

6324

Compound2 was preparedby the sequenceshownin Scheme 1. The secondaryammonium5 was treatedwiththe knownorgartoboron62to provide7,8whichwaseasilyhydrolysizedto 2 withaqueouslithium hydroxide.A controlversion,3, whichdoesnot containa boronicacidwaspreparedby a similarsequence. Scheme

+ tfH3 Br

P

o

:1

CHSO

4

x

w

-

00

‘B’

<+ NH2 Cl-

a

—

0 & CH30

1;

(a) 9-anthraldehyde,CH30H,

1

,

1; U*

~ b

\\\ T “d 4 / / / :.0

Br

o

d1;

N 1;

aieves; NaBH4; HCI; (71 %); (b) 6, K2C03;

(4B %).

Initially,the bindingabilitiesof the free receptors,1 and 2, were determinedin aqueoussolution(pH 7.3, 15rnM phosphatebuffer)by fluorescencetitrationmethods. Compound2 (log Ka = 4.6) binds fructose more tightly than parent 1 (log Ka = 3.7),9 but the fluorescenceenhancementat sensor saturationis smaller (two-foldfluorescenceincreasefor 2 comparedto a three-foldenhancementfor 1). Compounds1 and2 were rdsoassayed for associationwith nucleosidesand nucelotides. In all cases there was little or no change in receptorfluorescencesuggestingveryweakbinding. To enhancenucleotideaffinity,receptor2 was attachedto poly(allylamine)PA (MW 50,000- 65,000) via an amidelinkage.lo A loadinglevelof107. was employed (i.e., the ratio of sensor 2 to allylamine monomer was 0.1). A control polymer was also prepared by functionalizing10% of PA withfluorophore3. The PA-2 conjugatewas titratedwith fructose,uridineand variousuridine5’-phosphatederivativesin aqueous solution (pH 7.3, 15 mM phosphatebuffer). The resulting changes in fluorescenceare shown in Figure 1. Titrationof PA-2 with fructoseshowedweakenedaffinity(log Ka = 2.7) comparedto boronicacid receptor 2 alone. It is possiblethis is due to aggregationof the hydrophobicsensinggroups. However,the fluorescencespectrumfor PA-2 in aqueousphosphatesolutionshowsno excimeremission,indicatingthat the sensinggroupsare dispersedalongthe PA chain. Therefore,it appearsthat the cationicPA and its solvation shellof phosphateionsstericallyencumberfructosebindingto the sensingboronicacidgroups. 1.30-

1.20-

I/lo

1.101.004

0.9000.20 solute cone (mM)

Figure 1. Change in fluorescence emission (kex 389 nm, l.em 423 nm) for PA-2 upon titration with: O 5’-UT#-; ■ thctose; A 2’-deoxy-5’-UT#-,+ 5’-UMP2-,and x 5’-UDP3-, in phosphatebuffer solution(PH 7.3, 15 mM), T = 298 K.

6325

Titrationof PA-2 with uridine,5’-UMP2-and 5’-UDP3-producednegligiblechangesin fluorescence. However,5’-UTP4-was foundto bind very strongly(log Ka = 5.6).9 Althoughthe changein fluorescenceis small,the highassociationconstantfor 5’-UT~-is particularlyimpressivesincebindingoccursin thepresence of competingphosphatebuffer. The associationconstantis comparableto the best aftlnities observedwith Lehn’sbis(intercrdand)and tris-acridinecryptand.bIn addition,PA-2exhibitsa bindingselectivitythat is quite differentto Lehn’sreceptors. The lack of responseto 2’-deoxy-5’-UTP, a compoundincapableof chelating with the boronicacid, is evidencein favorof the bindingmodeshownin Scheme2. Furtherevidencethat the boronic acid is crucial for sensingis the observationthat in phosphatebuffer the control polymer PA-3 is unresponsiveto 5’-UTP’f_ or anyof the compoundsdescribedin Figure 1. Scheme 2

P

w+

PA-2 (10 % loading)

HO? ..-= ‘ Anth-

N.D

\

UTP&

:1 a

a

@“

&

weak fluorescence

a

-2 H20

enhancedfluorescence

Competitive titration experiments provided a further demonstrationof the sensing selectivity for nucleoside triphosphate. Titration of PA-2 with 5’-UTP4-in the presence of excess 5’-UMP2-(1 mM) produceda titration curve that was essentiallyidenticalto that seen in Figure 1. The high selectivityfor 5’UTP4-over5’-UDP3-and 5’-UMP2is rationalizedin termsof a competitionbetweenthe analyteandphosphate bufferfor the cationicpolymerPA-2. It appearsthat onlywhenthe analytechargereaches-4 doesbindingto the cationic PA-2 become highly favorable. When the titrations are repeated in HEPES (N-[2hydroxyethyl]piperazine-N’-[2-ethanesulfonic acid)bufferall of the nucleotidesinduce a modest fluorescent responsein the order 5’-UTP4-> 5’-UDP3-> 5’-UMP2-> fructose. In this case, the anionicanalytescan more readilydisplacethe zwitterionicHEPESfromthecationicPA-2chainandgain accessto the sensinggroups. A similar result was very recently reported by Anslyn.l1 In HEPESbuffer, the control polymer PA-3 also showed weak fluorescentresponsesto the differentnucleotides,suggestingthat in HEPES there are more signalingpathwaysavailableto PA-2thantheboron-mediatedmechanismshownin Scheme2. The fluorescenceenhancementfor PA-2 saturatedwith 5’-UT@-is lower than that observedwith fructose. This suggests that the anthracenefluorescenceis partially quenchedby the pyrimidinebase, in agreement with the known fluorescence quencing properties of nucleobases.b,lz Thus the observed fluorescenceincrease in phosphatebuffer is a compromisebetween the enhancementdue to boronic acid chelation,andquenchingdueto interactionbetweenthesnthraceneandnucleobase. In summary,conjugationof a fluorescentdiol-sensorto a cationicpolymerchain producesa sensing compoundwith selectivityfor ribonucleoside5’-triphosphate.Thus,we havedemonstratedan applicationof a generalconceptwheretheintroductionof secondaryinteractionenhancesthe aftlnityandselectivityinherentto a molecularreceptor. Studiesin progressare attemptingto furtherimprovethese propertiesthroughthe use of morehighlyfictionalized secondruyrecognitiondomains.

6326

Acknowledgments. This work was supportedby a BayerPostdoctoralFellowshipand the NationalScience

Foundation(USA). References

and Notes

1. Gennari,C.; Nestler,H. P.; Piarulli,U.; Salem,B. LiebigsAnn.lRecueil 1997, 637-647. 2. (a) James, T. D.; Linnane, P.; Shinkai, S. .1. Chem. Soc., Chetn. Commun. 1996, 281-288. (b) James,

T. D.; SamankumaraSandanayakeK. R. A.; Shinkai,S. Angew. Chem. Int. Ed. Engl. 1996,35, 19101922. 3. Smith,B. D. Supramol. Chem. 1996,7, 55-60. 4. To datetherehavebeenfew attemptsto alterthe sensingpropertiesof 1 by syntheticmodification.The R groupin 1 has beenchangedfromhydrogento a secondboronicacid,2a crownether,zand a photochromic diarylethene;Takeshita,M.; Uchida,K.; Irie, M. J. Chem. Soc., Chem. Commun. 1996, 1807-1808. 5. Czarnik,A. W. Chemistry and Biology 1995,2,423-428. 6. (a) Cudic,P.; Zinic, M.; Tornisic,V.; Simeon,V.; Vigneron,J. P.; Lehn,J.-M. J. Chem. Soc., Chem. Commun. 1995, 1073-1074. (b) Teulade-Fichou,M, -P.; Vigneron,J. -P.; Lehn, J. M. J. Chem. Soc., Perkin Trans. 21996, 2169-2174. 7. Takeuchi,M; Taguchi,M.; Shinmori,H.; Shinkai,S. BuZZ.Chem. Soc. Jpn. 1996,69, 2613-2618. 8. Characterizationdata for 7: mp 157-159“C; IH NMR(500MHz,CDC13)50.92 (s, 6H), 3.37(s, 4H),

3.59 (s, 2H), 3.89 (s, 3H), 3.96 (s, 2H), 4.66 (s, 2H), 7.22 (app. t, J = 7.5 Hz, IH), 7.30-7.45(m, 8H), 7.64 (d, J = 7.5 Hz, IH), 7.90 (d, J= 8 Hz, 2H), 7.94-7.96(m, 2H), 8.25-8.27(m, 2H), 8.35 (s, IH); 13c NMR (75 MHz, CDC13)621.87, 31.42,50.77,51.98, 58.63, 58.70,71.74, 124.72,125.28, 125.35, 126.36, 127.39, 128.60,128.89, 129.20,129.43,129.47, 130.30,131.39, 131.59, 134.04, 143.95,145.70,167.14;MS (FAB+)mlz:557 (M+, 12),366(13), 191(100);HRMS557.2788(calcdfor C36H36BN04:557.2744). 9. AssociationconstantsweredeterminedbytheBenesi-Hildebrand methodornordinearcurvefittingof the equationfor 1:1binding. Tsukube,H.; Furata,H.; Odani,A.; Takeda,Y.; Kudo,Y.; Inoue,Y.; Liu, Y.; Sakarnoto,H.; Kimura,K. in Comprehensive Supramolecular Chemist~, Atwood,J. L., Davies,J. E. D., MacNicol,D. D., Vogtle,F., Eds.; Pergamon,New York, 1996,vol. 8, Ch. 10. 10. Menger,F. M.; Eliseev,A. V.; Migulin,V. A. J. Org. Chem. 1995,60,6666-6667. Poly(allylarnine) hydrochloride(MW50,00- 65,000)wasconvertedto its freeamineformusinga ten-foldexcessof Dowex l-X8 resin. An aqueoussolutionof poly(allylamine)(0.5M of monomerunits)wastreatedwithsolutions of 2 (0.1molarequivalentsof monomerunits)in dimethylforrnamide and l-(3-dimethylaminopropyl)-3ethylcarbodiimidehydrochloride(0.5molarequivalentsof monomerunits)in water. The reactionwas stirredat roomtemperaturefor 14h beforepurificationby gelfiltration(SephadexG15)andremowdof the solvent. The residue(IR 1620cm-l)wasredissolvedin methanol/waterto a makestocksolutionthatwas 20 @f in sensorunit. A 200~L solutionof this stocksolutionwasdilutedto 3.0MLwithphosphate buffer(15MM,pH 7.3)to givea finalsensorunitconcentrationof 1.3@f. 11. Metzger,A.; Lynch, V. M.; Anslyn,E. V. Angew. Chem. Int Ed. Engl. 1997,36, 862-865. 12. Lewis, F. D., Zhang, Y.; Letsinger,R. L. J. Am. Chem.Soc.,1997, 119, 5451-5452.

(Received in USA 10 June 1997;accepted 10 July 1997)