Pavlína Vidláková

MASARYKOVA UNIVERZITA PŘÍRODOVĚDECKÁ FAKULTA

STUDIUM ELEKTROCHEMICKÝCH VLASTNOSTÍ SYNTETICKÝCH OLIGONUKLEOTIDŮ A DNA MODIFIKOVANÝCH REDOX-AKTIVNÍMI SKUPINAMI

Disertační práce

Pavlína Vidláková

Vedoucí práce: Doc. RNDr. Miroslav Fojta CSc. Brno 2015

Bibliografický záznam

Autor: Mgr. Pavlína Vidláková

Studium elektrochemických vlastností syntetických Název práce: oligonukleotidů a DNA modifikovaných redox- aktivními skupinami Studijní program: Fyzika

Studijní obor: Biofyzika

Vedoucí práce: Doc. RNDr. Miroslav Fojta CSc.

Akademický rok: 2014/15

Počet stran: 53+57 Klíčová slova: DNA, elektrochemie, elektrochemické značení, biosensor

Bibliographic Entry

Author Mgr. Pavlína Vidláková

Study of electrochemical properties of of synthetic Title of Thesis: oligonucleotides and DNA modified with redox-active moieties Degree programme: Physics

Field of Study: Biophysics

Supervisor: Doc. RNDr. Miroslav Fojta CSc.

Academic Year: 2014/15

Number of Pages: 53+57 Keywords: DNA, electrochemistry, electrochemical labeling, biosensor

Poděkování

Na tomto místě bych chtěla poděkovat svému školiteli Doc. RNDr. Miroslavu Fojtovi CSc. a také Mgr. Luďkovi Havranovi Dr. za jejich odborné vedení a cenné rady a také za připomínky při zpracování této práce. Déle děkuji kolegům z Biofyzikálního ústavu za příjemné pracovní prostředí a veškerou pomoc, které se mi od nich dostalo.

Prohlášení

Prohlašuji, že jsem svoji rigorózní práci vypracovala samostatně s využitím informačních zdrojů, které jsou v práci citovány.

Brno 6. března 2015 ……………………………… Pavlína Vidláková

Abstrakt

V této disertační práci se věnujeme elektrochemické analýze nukleotidových sekvencí s využitím elektrochemických značek a vlastnostem těchto značek. Teoretická část práce je zaměřena na shrnutí poznatků týkajících se elektrochemického chování nukleových kyselin a oligonukleotidů s využitím různých typů elektrod a elektrochemických metod. Dále jsou zde popsány metody značení nukleových kyselin elektroaktivními značkami a také příprava a použití DNA-biosenzorů.

Experimentální část je přiložena ve formě pěti publikací otištěných v impaktovaných časopisech, viz přílohy 1-5. V kapitole Výsledky a diskuze jsou shrnuty a stručně komentovány výsledky obsažené v těchto článcích. Tyto články jsou zaměřené především na přípravu nových elektrochemickým značek (antrachinonu, nitrofenylu, benzofurazanu a butylakrylátu) využitelných pro značení DNA a přípravu biosenzorů a detailní studium elektrochemického chování těchto látek a to jak samostatně, tak v různých kombinacích. Tyto značky byly pomocí metody prodlužování primeru inkorporovány do molekul oligonukleotidů a takto připravené oligonukleotidy byly opět elektrochemicky studovány.

Poslední článek je zaměřen na elektrochemické studium DNA modifikované kancerostatikem cisplatinou. Cisplatina se kovalentně váže na DNA, nejčastějším vazebným místem jsou dinukleotidové motivy GG a AG. Při elektrochemické redukci komplexních sloučenin platiny na rtuťových elektrodách dochází ke katalytickému vylučování vodíku. Tento proces je možné využít pro citlivé stanovení modifikované DNA.

Abstract

In this thesis we deal with electrochemical analysis of nucleotide sequences using electroactive labels and properties of these labels. The theoretical part is concentrated on summary of findings regarding of electrochemical behavior of nucleic acid and oligonucleotides using of various types of electrodes and electrochemical methods. Further are there methods of DNA labeling with using electroactive tags and methods of preparation of DNA-biosensors.

Experimental part of the thesis encompasses five papers published in international peer- reviewed journals (see appendices 1-5). The results included in these papers are summarized and briefly commented in the chapter Results and Discussion. These papers are concentrated on preparation of new electroactive labels for DNA and for preparation of DNA-biosensors, and on studying electrochemical behavior of these labels separately or in various combinations of several types of labels simultaneously. These tags were incorporated into oligonucleotides with using primer extension and these oligonucleotides were electrochemically studied.

The last paper deals about electrochemical analysis of cisplatin modified DNA. Cisplatin binds covalently on DNA (especially on dinucleotide motifs GG and AG). Electrochemical reduction of cisplatin-DNA adduct at mercury electrodes is accompanied by catalytic hydrogen evolution. This process can be used for sensitive analyses of modified DNA.

Obsah

Seznam zkratek………………………………………………………………………….…..…8 Úvod………………………………………………………………………………………..…10 Literární přehled……………………………………………………………………….….…..12 1. Elektrochemické metody…………………………………………………….…....12 1.1 Pracovní elektrody………………………………………………….….…12 1.1.1 Rtuťové pracovní elektrody…………………………….….….12 1.1.2 Pevné pracovní elektrody……………………………….….....14 1.2 Referentní a pomocné elektrody…………………………………….…....15 1.3 Elektroanalytické metody…………………………………………..….…16 1.4 Rozpouštěcí a přenosové techniky……………………………….……....19 2. Elektrochemické vlastnosti nukleových kyselin………………………….………21 2.1 Elektrochemické chování DNA na rtuťových elektrodách………………21 2.1.1 Redoxní děje na rtuťových elektrodách……...…………....….21 2.1.2 Adsorpčně-desorpční děje na rtuťových elektrodách...…….…22 2.2 Elektrochemické chování DNA na uhlíkových elektrodách……………..24 3. Elektrochemické značení nukleových kyselin……………………………………25 3.1 Nekovalentě se vázající redoxní indikátory……………………………...26 3.2 Kovalentně se vázající elektrochemické značky…………………………26 3.3 Enzymatická inkorporace elektrochemicky značených nukleotidů……...28 3.4 Magnetoseparační techniky………………………………………………30 4. Elektrochemické biosenzory……………………………………………………...31 4.1 Elektrochemické biosenzory pro detekci hybridizace DNA……………..31 4.1.1 Imobilizace hybridizační sondy na povrchu elektrod………....32 4.1.2 Detekční principy využívané v elektrochemických senzorech pro hybridizaci DNA………………………………………….33 4.2 Elektrochemické senzory pro detekci mutací a polymorfismů…………..34 Cíle disertační práce…………………………………………………………………………..36 Seznam publikací……………………………………………………………………… ……37 Výsledky a diskuse……………………………………………………………………………38 Závěr………………………………………………………………………………………….49 Seznam literatury……………………………………………………………………………..50 Přílohy………………………………………………………………………………………...54

Seznam zkratek

A, C, G, T – adenin, cytosin, guanin, thymin

ACV - voltametrie s vkládaným střídavým napětím

AdSV – adsorptivní rozpouštěcí voltametrie

AdTSV – adsorptivní přenosová rozpouštěcí voltametrie

Aox, Gox – oxidační píky adeninu a guaninu

ASV – anodická rozpouštěcí voltametrie

CPSA - chronopotenciometrie s konstantním proudem

CSV – katodická rozpouštěcí voltametrie

CV – cyklická voltametrie

DNA (ds, ss, sc) – deoxyribonukleová kyselina (dvouřetězcová, jednořetězcová, superhelikální) dNTP – deoxyribonukleosid trifosfát

DPP – diferenční pulsní polarografie

DPV – diferenční pulsní voltametrie

HMDE – visící rtuťová kapková elektroda

LSV - voltametrie s lineárně se měnícím potenciálem

MeSAE – pevná amalgámová elektroda (z amalgámu kovu Me)

NK – nukleová kyselina

ON – oligonukleotid

Os,L – komplexy oxidu osmičelého

PEX – prodlužování primeru (primer extension)

PGE – elektroda z pyrolytického grafitu

RNA – ribonukleová kyselina

SAM – samoorganizovaná monovrstva (self-assambled monolayer)

SWV - voltametrie se superponovaným pravoúhlým napětím

TdT – terminální deoxynuxleotidyl transferáza

Úvod

Počátek elektrochemie nukleových kyselin spadá do druhé poloviny padesátých let dvacátého století. V roce 1958 uveřejnil E. Paleček práci, ve které pomocí oscilografické polarografie se střídavým proudem prokázal elektroaktivitu DNA (1). V roce 1961 byla publikována první práce, která se zabývala adsorpcí DNA na povrch rtuťové elektrody (2). Výsledky impedančního měření ukázaly, že je DNA povrchově aktivní látkou schopnou adsorpce na povrchu elektrody a při vkládání negativních potenciálů podléhá charakteristickým adsorpčně/desorpčním dějům. Tyto objevy vedly k dalšímu zkoumání elektrochemické a povrchové aktivity nukleových kyselin. Ukázalo se, že elektrochemické metody mohou přinášet informace jak o přítomnosti a koncentraci nukleových kyselin, tak i o jejich sekundární a terciální struktuře (3-7).

Zpočátku byl vývoj v oblasti elektrochemie nukleových kyselin limitován jak dostupnou instrumentací, tak i omezenými možnostmi přípravy vhodného experimentálního materiálu. Tato situace se výrazně změnila až v posledním desetiletí 20. století, kdy došlo k velkému pokroku jak v biologických i medicínských vědách, tak i k rozvoji analytických zařízení i metod. V současné době se využívá široké spektrum elektrochemických metod, které umožňují stanovení stopových množství nukleových kyselin ve velmi malých objemech vzorků. Ke snížení spotřeby analyzovaného materiálu přispělo mimo jiné zavedení adsorptivních rozpouštěcích technik ve spojení se rtuťovými nebo uhlíkovými elektrodami. Tyto techniky spočívají v akumulaci vzorku na elektrodu před samotným měřením, čím je dosaženo větší citlivosti stanovení. Adsorpce nukleových kyselin na elektrodu je možné provést z velmi malého (jednotky mikrolitrů) množství vzorku, což snižuje spotřebu analytického materiálu (8).

Přestože přirozená DNA je elektrochemicky aktivní, je pro některé analytické aplikace vhodné využít elektrochemického značení. To obvykle dovoluje citlivější detekci, než jaká by byla dosažena měřením vlastních signálů DNA, poskytuje signály v přístupnějších potenciálech mimo oblast vybíjení elektrolytu a elektrodové děje, jimž využívané značky podléhají, jsou často reversibilní. Je také možné detekovat značenou DNA i ve velkém nadbytku neznačené DNA. Při studiu struktury nukleových kyselin je možné využít specifických reakcí některých látek s určitou konformací DNA nebo jejich vazby na konkrétní báze (např. komplexy oxidu osmičelého) (9-11), nebo je možné využít inkorporaci nukleotidů

s kovalentně navázanými elektroaktivními skupinami s využitím metody prodlužování primeru (12-16). Při této metodě je pomocí různých DNA polymeráz prodlužován řetězec DNA od 5´ konce ke 3´ konci. Syntéza probíhá podle templátového řetězce na základě párování komplementárních bazí.

Tato práce je zaměřena především na testování nových elektrochemických značek využitelných pro studium nukleových kyselin, na přípravu oligonukleotidů značených těmito značkami s využitím metody prodlužování primeru a obecně na studium elektrochemického chování chemicky modifikované DNA.

Literární přehled

1. Elektrochemické metody

Základem zařízení pro elektroanalytická měření je pracovní elektroda zapojená do elektrického obvodu umožňujícího přesnou kontrolu vkládaného napětí a citlivé měření elektrického proudu, který obvodem prochází. Pracovní elektroda je obvykle ponořena do roztoku základního elektrolytu. Pokud je v tomto roztoku přítomna látka, která je za určitých podmínek schopná odevzdávat elektrony pracovní elektrodě, nebo je od ní přijímat, případně se na povrchu elektrody adsorbuje, lze v obvodu v závislosti na vloženém potenciálu naměřit proudové signály. Poloha těchto signálů poskytuje informaci o charakteru přítomné látky a jejich intenzita informace o jejím množství.

Při elektrochemické analýze DNA se obvykle používá klasické tříelektrodové zapojení s pracovní, referentní a pomocnou elektrodou.

1.1 Pracovní elektrody

Pracovní elektrody v elektrochemické analýze nukleových kyselin bývají rtuťové nebo z pevných materiálů (ušlechtilé kovy, amalgamy, uhlík). Na materiálu pracovní elektrody a použitém základním elektrolytu závisí rozsah potenciálů, které lze analyticky využít, tzv. potenciálové okno. Ve vodných roztocích je využitelný potenciálový rozsah zpravidla omezen elektrodovými reakcemi souvisejícími s elektrolýzou vody. Rozsah anodického potenciálu je omezen vylučováním kyslíku, katodický rozsah je omezen vylučováním vodíku. Oba tyto procesy jsou závislé na pH. V nevodných roztocích (resp. aprotických rozpouštědlech) je využitelné potenciálová okno zpravidla širší než v roztocích vodných.

1.1.1 Rtuťové pracovní elektrody

Rtuť jako elektrodový materiál má oproti většině tuhých materiálů několik výhod. Jsou to zejména homogenní a snadno obnovitelný povrch a vysoká hodnota přepětí vodíku, která umožňuje měření i při velmi negativních potenciálech. Nevýhodou je naopak snadná oxidace rtuti při kladných potenciálech. Z tohoto důvodu jsou vhodné zejména pro stanovení látek jejich katodickou redukcí.

Typy rtuťových elektrod:

- Kapající rtuťová elektroda DME – elektrodovým povrchem je v čase rostoucí povrch odkapávající rtuťová kapky. - Statická rtuťová kapková elektroda SMDE – u ústí kapiláry je reprodukovatelně vytvářena visící kapka, která je v pravidelných intervalech obnovována. Vzorek je tak na povrchu elektrody absorbován jen několik sekund, během kterých se vložený potenciál mění jen nepatrně, což omezuje např. změny ve struktuře DNA. Jak v případě DME, tak SMDE je nevýhodou je velká spotřeba analyzovaného materiálu, protože tento způsob měření (polarografie) vyžaduje relativně velké objemy vzorků o poměrně vysoké koncentraci (ve strovnání s technikami využívajícími akumulaci analytu na povrchu stacionární elektrody). - Visící rtuťová kapková elektroda HMDE – v tomto případě je celá analýza prováděna na jedné kapce. Silné adsorpce DNA na elektrodě je možné využít při adsorpčních rozpouštěcích technikách včetně přenosových, které umožňují analyzovat i velmi malé množství zředěných vzorku. - Rtuťová filmová elektroda MFE – připravuje se vyloučením malého množství rtuti na povrchu tuhé elektrody. - Amalgamové elektrody - pastové nebo pevné amalgámy, což jsou slitiny rtuti a dalšího kovu (Me). Pevné amalgamové elektrody (MeSAE) jsou obecně netoxické. Po potřeby analýzy DNA se nejčastěji používají stříbrné a měděné amalgámy, často pokryté rtuťovým filmem nebo meniskem. Mají velmi podobné elektrochemické vlastnosti jako elektrody rtuťové (zejména vodíkové přepětí), proto je řadíme do této části.

1.1.2 Pevné pracovní elektrody

Pevné pracovní elektrody (kromě výše zmíněných MeSAE) bývají nejčastěji kovové nebo uhlíkové. Na rozdíl od rtuťových elektrod je povrch tuhých elektrod více či méně nehomogenní a je obtížněji obnovovatelný. Povrch je možné obnovovat mechanickým, chemických nebo elektrochemickým čištěním, případně kombinací těchto postupů. Před samotnou analýzou se často používá elektrochemická aktivace opakovanou cyklickou polarizací v určitém potenciálovém rozsahu. Výhodou pevných pracovních elektrod proti elektrodám rtuťovým je možnost použití při pozitivnějších potenciálech a také jejich použití v detektorech pro separační metody (chromatografie, elektroforéza) či pro měření v průtoku.

Lze je také použít ve formě mikroelektrod ke stanovování látek v biologických objektech. Uhlíkové elektrody se ve srovnání s kovovými elektrodami vyznačují nižším zbytkovým proudem, jsou méně náchylné k oxidaci povrchu a jejich povrch se obvykle snadněji obnovuje.

Typy pevných elektrod:

- Grafitová elektroda – přírodní grafit je porézní materiál, do něhož může vzlínat roztok a geometrický povrch elektrody tak nelze přesně definovat. - Elektroda z pyrolytického grafitu (pyrolytic graphite electrode, PGE) – připravuje se pyrolýzou uhlovodíků. Je méně pórovitá než běžný grafit a má vrstevnatou strukturu, což umožuje její použití ve dvou orientacích – v orientaci základnové roviny (basal plane) a hranové roviny (edge plane), které mají odlišné vlastnosti. - Elektroda ze skelného uhlíku (glassy carbon) – je prakticky nepórovitý a může být vyleštěn do zrcadlového lesku, čímž je jeho geometrický povrch poměrně dobře definovaný. - Uhlíkové pastové elektrody (carbon paste electrode) – připravují se smísením uhlíkové pasty s vhodnou hydrofobní elektrochemicky inertní kapalinou. Jejich výhodou je přizpůsobení jejich složení stanovovanému analytu, včetně biologické či chemické modifikace a také snadné obnovení povrchu otřením. - Diamantové elektrody – jde o film uhlíku se strukturou diamantu, který je pro zavedení elektrické vodivosti dopován atomy boru. Výhodou tohoto materiálu je mimořádně nízký šum a menší problémy s pasivací, než u jiných uhlíkových elektrod. - Zlaté a platinové elektrody – zlato má menší elektrokatalytické účinky než platina a je tedy méně náchylné k otravě katalytickými jedy, proto bývá v elektroanalytické praxi upřednostňováno. Pro imobilizace DNA na povrchu zlaté elektrody se obvykle používají thiolované oligonukleotidy (17).

1.2 Referentní a pomocné elektrody

Referentními elektrodami bývají elektrody II. druhu (jsou tvořeny kovem pokrytým vrstvou jeho málo rozpustné soli v roztoku obsahujícím anion této soli), jako jsou argentochloridová, merkurosulfátová či kalomelová elektroda. Pomocné elektrody bývají z inertního materiálu (Pt, C) a ve srovnání s pracovními elektrodami mívají výrazně větší povrch (drátek, plíšek atd.).

1.3 Elektroanalytické metody

Pro analýzu DNA je možné využít celou řadu elektrochemických metod (např. metody potenciostatické, voltametrické). Nejčastěji používanou metodou je voltametrie, a to zejména voltametrie cyklická a voltametrie se superponovanými pravoúhlými napětovými pulsy (square-wave). Voltametrie (nebo polarografie, v případě, kdy pracovní elektrodou je kapající rtuťová kapka) je metoda, při níž se sleduje závislost proudu procházející pracovní elektrodou ponořenou v analyzovaném roztoku na potenciálu, který se na tuto elektrodu vkládá z vnějšího zdroje. Analytickým signálem je velikost proudu procházející obvodem v přítomnosti analytu při vhodném potenciálu. Závislost proudu na elektrodovém potenciálu se měří buď v ustáleném stavu, nebo za nestacionárních podmínek.

Voltametrie s lineárně se měnícím potenciálem (LSV)

Na pracovní elektrodu se vkládá potenciál, který se lineárně mění s časem. Měří se závislost proudu na vkládaném potenciálu - proudová odezva.

Obr. 1: Potenciálový program při LSV

Cyklická voltametrie (CV)

Na pracovní elektrodu se vkládá potenciál trojúhelníkového průběhu (potenciál nejprve v čase lineárně roste do určité hodnoty – potenciál bodu obratu (Esw) - a později opět lineárně klesá). Důležitým faktorem u LSV a CV je rychlost změny potenciálu. V důsledku toho, že difúze je pomalý proces, při dostatečné rychlosti polarizace nestačí produkty elektrodové reakce oddifundovat od elektrody a za vhodných podmínek je možné je při opačném směru potenciálové změny detekovat. Na základě měření při různých rychlostech polarizace lze usoudit na povahu příslušného elektrochemického děje a jeho reversibilitu.

Obr. 2: Potenciálový program při CV

Voltametrie se superponovaným pravoúhlým napětím (square wave, SWV)

Na elektrodu se vkládá potenciál lineárně se měnící s časem a ten se moduluje střídavým napětím pravoúhlého tvaru o malé amplitudě a vhodné frekvenci. Proud se měří pouze na konci každého (negativního a pozitivního) vloženého pulsu. Tímto způsobem naměřená hodnota proudu odpovídá prakticky faradaickým dějům na elektrodě, protože kapacitní proud je eliminován, což vede ke zvýšení citlivosti měření. Registruje se závislost rozdílu dvou po sobě změřených vzorcích proudu na potenciálu elektrody. Celé měření je tak sérií potenciálových „mikrocyklů“, proto je tato metoda vhodná zejména pro studium reverzibilních dějů.

Obr. 3: Potenciálový program při SWV

Voltametrie s vkládaným střídavým napětím (ACV)

Potenciál vkládaný na elektrodu, který se lineárně mění s časem, se moduluje střídavým napětím sinusového průběhu o malé amplitudě a nízké frekvenci (desítky až stovky Hz). Měří se závislost střídavého proudu procházejícího elektrodou na jejím potenciálu. Metoda je vhodná ke studiu adsorpčních dějů.

Obr. 4: Potenciálový program při ACV

Diferenční pulsní voltametrie (DPV)

Lineárně se měnící potenciál vkládaný na elektrodu je modifikovaný pravidelně vkládanými pulsy s konstantní amplitudou. Stejně jako u SWV se proud měří pouze po určitou, přesně definovanou dobu – před každým pulsem a na konci pulsu. Vynáší se rozdíl těchto proudů proti potenciálu. Zavedení pulsních metod vedlo ke snížení detekčních limitů (18, 19).

Obr. 5: Potenciálový program při DPV

Rozpouštěcí chronopotenciometrie s konstantním prouden CPSA

U CPSA se sleduje časový průběh změny potenciálu pracovní elektrody za vnuceného konstantního proudu. Výhodou této metody oproti voltametrickým metodám je možnost provádět měření v katodické oblasti i v elektrolytech obsahujících kyslík. CPSA se v poslední době osvědčila jako metoda vysoce citlivá k určitým katalytickým dějům, např. katalytickému vyvíjení vodíku v přítomnosti proteinů, kterého lze využít k citlivé analýze jejich struktury (20).

1.4 Rozpouštěcí a přenosové techniky

Rozpouštěcí voltametrie využívá předběžnou akumulaci analytu z roztoku vzorku na povrchu pracovní elektrody. Tím se koncentrace analytu na povrchu elektrody oproti koncentraci analytu v roztoku o několik řádů zvýší, takže při následném rozpouštění z povrchu elektrody zpět do roztoku je naměřený signál (proud) též vyšší. Touto metodou lze dosáhnout mnohem lepšího limitu detekce, než při prostém měření analytu v roztoku bez akumulace. Akumulaci analytu na elektrodu lze provést buď potenciostatickou elektrolýzou nebo adsorpcí. Podle způsobu akumulace látky na elektrodě a jejího následného rozpouštění rozlišujeme následující metody:

- Anodická rozpouštěcí voltametrie (ASV) – analyt se nejprve katodicky vylučuje na elektrodě a následně se anodicky rozpouští zpět do roztoku.

- Katodická rozpouštěcí voltametrie (CSV) – analyt se nejprve anodicky vylučuje na elektrodě a následně se katodicky rozpouští zpět do roztoku.

- Adsorpční rozpouštěcí voltametrie (AdSV) – analyt se na povrchu elektrody akumuluje adsorpcí při určitém potenciálu nebo i při otevřeném obvodu. Pokud jsou adsorbované látky elektrochemicky aktivní, lze je stanovit pomocí jejich oxidačních/redukčních signálů. Pokud jsou elektrochemicky neaktivní, mohou se při dosažení určitého potenciálu desorbovat, což se projeví tzv. tensametrickým píkem, který lze rovněž analyticky využít. Tensametrický pík odpovídá kapacitnímu proudu procházejícímu elektrodou při změně kapacity elektrické dvojvrstvy při desorpci.

Protože se nukleové kyseliny velmi dobře adsorbují na povrch rtuťových i uhlíkových elektrod, je při jejich elektrochemické analýze AdSV hojně využívána. Zavedení této techniky v 70. letech 20. století vedlo ke značnému rozvoji elektrochemické analýzy DNA v důsledku zvýšení citlivosti stanovení a také zmenšením spotřeby analytického materiálu, který nebylo snadné připravit ve větších množstvích. K dalšímu zvýšení citlivosti a selektivity voltametrického stanovení přispívá oddělení akumulačního a detekčního kroku. Naadsorbování DNA na rtuťovou nebo uhlíkovou pracovní elektrodu je možné provést z malého objemu vzorku (obvykle několik l), elektroda s naadsorbovanou DNA je následně opláchnuta vodou, základním elektrolytem nebo vhodným činidlem schopným odstranit rušící látky a přenesena do základního elektrolytu, kde je provedeno elektrochemické stanovení (8). Tato metoda se označuje jako adsorptivní přenosová rozpouštěcí voltametrie (AdTSV). K výhodám AdTSV patří značné snížení spotřeby vzorku, možnost akumulace DNA z roztoků, které neumožňují elektrochemické stanovení, snadné odstranění řady rušících látek v promývacím kroku a v neposlední řadě i možnost využití elektrod s adsorbovanou vrstvou DNA jako jednoduchých elektrochemických biosenzorů (7).

Obr. 6: Schéma AdTSV: 1 – akumulace vzorku, 2 – promytí, 3 – měření

2 Elektrochemické vlastnosti nukleových kyselin

Elektrochemická analýza je využívána ke studiu přirozených i modifikovaných molekul DNA i oligonukleotidů. Elektrochemická stanovení dosahují citlivosti přinejmenším srovnatelné s optickými metodami a často lepší selektivitu při nižších pořizovacích a provozních nákladech. Molekula DNA je elektroaktivní a je možné ji detekovat na rtuťových i uhlíkových elektrodách (21). Při detekci se využívá jak oxidačně-redukčních vlastností NK, tak i charakteristických adsorpčně-desorpčních dějů na povrchu elektrod. Pro detekci DNA je možné využít celé řady elektrochemických metod. Je možné využít jak klasická polarografická a voltametrická měření, LSV, CV, SWV (22, 23), CSV (24), pulsní techniky jako diferenční pulsní polarografie (25), tak i fázově citlivé techniky jako impedanční spektroskopie (26) a ACV (4). Z galvanostatických metod se osvědčila rozpouštěcí chronopotenciometrie s konstantním proudem (27).

2.1 Elektrochemické chování DNA na rtuťových elektrodách

Při potenciálech větších než +0,1 V (proti Ag|AgCl|3M KCl; všechny hodnoty potenciálů v této práci jsou vztaženy na argentchloridovou referentní elektrodu) dochází k elektrolytickému rozpouštění rtuti, proto jsou rtuťové elektrody preferovány pro sledování katodických dějů. Anodické jevy lze sledovat pouze v případě, že probíhají při potenciálech nižších než přibližně 0,0 V.

2.1.1 Redoxní děje na rtuťových elektrodách

Z přirozených složek NK lze na povrchu rtuťových elektrod ve vodném prostředí redukovat cytosin, 5-methylcytosin (28), adenin a guanin. Cytosin, 5-methylcytosin a adenin v DNA se redukují při potenciálu okolo -1,5 V (v závislosti na pH elektrolytu) a poskytují katodický pík CA (7). Guanin se redukuje při potenciálech zápornějších, než -1,6 V. Redukční pík není na voltamogramech pozorovatelný, protože při takto záporných potenciálech dochází k vylučování vodíku na povrchu elektrody a proudové signály příslušející tomuto procesu jsou mnohem výraznější než redukce guaninu. Na voltamogramu můžeme pozorovat anodický pík G při potenciálu -0,25 V příslušející oxidaci 7,8-dihydrogenguaninu generovaného redukcí guaninu (21, 29, 30) (obr. 7). Redukce thyminu (nebo uracilu v RNA) na povrchu rtuti bylo pozorována pouze v nevodném prostředí (dimethylsulfoxid) při velmi negativních potenciálech (31).

Přístupnost zbytků bází je silně závislá na struktuře DNA. Zatímco v ssDNA jsou báze volně přístupné, v dsDNA jsou redukční místa bazí skryty uvnitř dvoušroubovice a nemohou volně komunikovat s prostředím ani s povrchem elektrody (Obr. 8). V důsledku toho se elektrochemické signály ssDNA a dsDNA na rtuťové elektrodě značně liší.

0.1 G 0.0

-0.1

A -0.2



I / I -0.3 CA

-0.4

-0.5 -1.5 -1.0 -0.5 0.0 E [V]

Obr. 7: Cyklický voltamogram DNA z telecího brzlíku na HMDE

Obr. 8: Primární redukční (modrý rámeček) a oxidační (červený kroužek) místa bází nukleových kyselin

2.1.2 Adsorpčně-desorpční děje na rtuťových elektrodách

Nukleové kyseliny jsou na povrchu rtuťových i amalgámových elektrod silně adsorbovány, čehož lze, kromě aplikace adsorptivních přenosových technik, využít i pro

přípravu elektrod modifikovaných nukleovými kyselinami. Adsorpčně-desorpční děje, jimž NK podléhají na rtuťových elektrodách při vkládání negativních potenciálů, vedou ke vzniku tensametrických signálů, které poskytují řadu informací o vlastnostech studovaných NK (především o jejich struktuře) (3, 4, 21). Při vkládání negativních potenciálů na povrch elektrody dochází k elektrostatickému odpuzování záporně nabité cukrfosfátové páteře molekul NK a za určitých podmínek dochází k desorpci těchto molekul nebo jejich částí z povrchu elektrody. Tím se mění diferenciální kapacita elektrodové dvojvrstvy a v důsledku toho se objevují zmiňované tensametrické signály. Charakter těchto signálů závisí na tom, která část molekuly se se adsorpčně-desorpčních dějů účastní, a to dále závisí na jejich struktuře (3).

Vhodným nástrojem pro studium adsorpčně-desorpčních dějů a strukturních změn DNA na povrchu rtuťových kapkových elektrod nebo amalgámových pevných elektrod je AC voltametrie. V katodickém AC voltamogramu lze pozorovat sérii tenzametrických píků, které souvisí a adsorpcí/desorpcí molekul DNA na elektrodě. Při potenciálech okolo -1,2 V dochází k desorpci cukr-fosfátové páteře DNA, která se projevuje vznikem píku označovaného jako pík 1 (Obr. 9). Při potenciálu okolo -1,4 V dochází k desorpci molekul DNA adsorbovaných na povrchu elektrody prostřednictvím bází. Tento tensametrický pík je pozorován pouze tehdy, pokud je molekula DNA jednořetězcová nebo obsahuje jednořetězcové úseky (Obr. 9). V případě dsDNA, která obsahuje volné konce, byl při potenciálu okolo -1,2 V (tzv. U- region) pozorován jev označovaný jako povrchová denaturace (32, 33). Jedná se o odvíjení molekuly DNA řízené potenciálem elektrody. Kromě píků 1 a 3 byl identifikován i pík 2 okolo -1,3 V (často se částečně překrývá s píkem 1), který lze pozorovat při konformačních změnách v dsDNA (34).

Jak již bylo řečeno, tensametrické signály na rtuťových elektrodách mohou být využity jako vysoce citlivý nástroj na sledování i velmi malých změn struktury DNA, jako jsou například odvinutí dvoušroubovice, tvorba jedno- i dvouřetězcových zlomů apod. Tyto změny struktury vedou ke zvýšení přístupnosti zbytků bází, která se projeví na jejich voltametrických signálech. (3, 4, 21, 34, 35).

1.50

1.25 3 scDNA dsDNA 1.00 ssDNA

0.75

 I/ 1 0.50

0.25

0.00 -1.5 -1.4 -1.3 -1.2 -1.1 -1.0 -0.9 -0.8 E/V

Obr. 9: AdTS ACV superhelikální DNA (sc, černá), lineární dvouřetězcová DNA (ds, modrá) a jednořetězcové DNA (ss, červená). Pík 1 odpovídá desorpci cukr-fosfátové páteře, pík dva odpovídá desorpci bází

2.2 Elektrochemické chování DNA na uhlíkových elektrodách

Uhlíkové elektrody na rozdíl od elektrod rtuťových umožňují sledovat i oxidační procesy probíhající při kladných potenciálech. Oxidace purinových bází na povrchu uhlíkové elektrody byla objevena v 70. letech 20. století (36-39). Oxidace guaninu probíhá při potenciálech (v závislosti na pH) okolo +1,0 V o poskytuje pík Gox, oxidace adeninu probíhá při potenciálech okolo +1,2 V o poskytuje pík Aox (Obr. 10). Oxidace pyrimidinových bází probíhá v závislosti na experimentálních podmínkách při ještě pozitivnějších potenciálech (okolo +1,3 V) (40, 41). Měření oxidačních píků pyrimidinových bází v polynukleotidovém řetězci je tak poměrně obtížné. Z tohoto důvodu jsou analyticky využívány obvykle pouze oxidační píky purinů. Ve srovnání s redukčními a tenzametrickými píky DNA na rtuťových a amalgamových elektrodách nejsou oxidační píky tolik citlivé na změnu sekundární struktury dsDNA (41). Změny oxidačních píků purinů je možné využít při studiu interakcí DNA s různými činidly. Tvorba reakčního produktu mezi bází a interagující látkou obvykle vede k poklesu oxidačního píku této báze. Osvědčilo se zejména sledování píku Gox, protože

guanin je vůči většině činidel reaktivnější než ostatní báze. Příkladem těchto reakcí je například reakce s oxidačními činidly, která vede k oxidaci guaninu na 8-oxoguanin, který má nižší oxidační potenciál než guanin a při potenciálu píku Gox už žádný signál neposkytuje (40).

 4

ox 2 G Aox

0.8 1.0 1.2 1.4 1.6 E/[V

Obr. 10: AdTS SWV ssDNA na elektrodě z pyrolytického grafitu (křivky po odečtení základní linie).

3 Elektrochemické značení nukleových kyselin

Přestože nukleové kyseliny jsou elektrochemicky aktivní, a jejich oxidačně-redukční signály i tensametrické píky je možné použít pro celou řadu analytických aplikací, v jiných případech je často výhodné použít elektrochemické značení. Jako značky se využívají látky, které lze velmi dobře elektrochemicky detekovat, jako jsou například komplexy přechodných kovů (6, 42, 43), antrachinon (13), benzofurazan (12), nitro a aminosloučeniny (15) atd. Tyto látky podléhají redoxním reakcím a dávají takto modifikované DNA nové elektrochemické vlastnosti, které je možné využít pro studium struktury a interakcí nukleových kyselin. Díky těmto značkám lze snadno analyzovat přítomnost a množství modifikované DNA ve vzorku a to i v nadbytku neznačené DNA. Další nespornou výhodou použití značení DNA je možnost stanovení DNA i pomocí elektrod, na kterých přirozená DNA žádné analyticky využitelné redoxní signály neposkytuje (např. zlato) (44-46). Využití kombinace více druhů elektrochemických značek (s nepřekrývajícími se potenciály redoxních píků) umožuje specifické redoxní kódování nukleotidových sekvencí pro jejich paralelní stanovení (47), nebo dokonce kódování jednotlivých nukleobází (48-50).

Značená DNA se dá s výhodou použít pro přípravu hybridizačních sond, což jsou úseky DNA o specifické sekvenci, které mohou hybridizovat se sekvencí cílové DNA. Tímto způsoben může být detekována např. přítomnost mutací v různých klinických vzorcích (podrobněji v kapitole 4.1).

3.1 Nekovalentně se vázající redoxní indikátory

Nekovalentně se vázající indikátory se používají především k rozlišení dsDNA a ssDNA imobilizované na povrchu elektrody. Indikátory se váží na DNA buď na základě elektrostatické interakce mezi indikátorem (který je obvykle kationtem) a polyanionickým řetězcem NK nebo se jedná o látky, které se interkalují do dvoušroubovice DNA. Mezi látky používané jako nekovalentně se vázající indikátory patří řada látek známých jako protinádorová léčiva (daunomycin (51), echinomycin (52)), organocyklických sloučenin (methylenová modř (53, 54), antrachinon (55, 56)) a kovových komplexů (57-60). Pro zvýšení citlivosti rozlišení mezi ssDNA a dsDNA bylo zavedeno používání threading (provlékacích) (61, 62) a bis-interkalátorů (63). Např. threading interkalátory mají jednu část, která se interkaluje a k této části jsou připojeny dvě objemné skupiny, které omezují vmezeření interkalující se skupiny do DNA. Jejich vazba na DNA je kineticky blokována, ale jakmile k ní dojde, je velmi stabilní. Příkladem takovéhoto interkalátoru je ferocenylnaftalendiimid, který byl použit např. jako indikátor mutací v genu pro p53 (64).

3.2 Kovalentně se vážící elektrochemické značky

Další skupinou elektrochemických značek jsou látky, které se kovalentně vážou na DNA. Mezi tyto látky patří například ferocen, který se používá pro přípravu hybridizačních sond a k analýzám specifických bodových mutací (46, 65, 66), a také komplexy přechodných kovů, zejména ruthenia a osmia (67, 68).

Mezi nejdéle používané elektrochemické značky pro DNA (už od 80. let minulého století) patří komplexy oxidu osmičelého s dusíkatými ligandy (Os,L, kde L je například 2,2´bipyridin nebo pyridin) (69, 70). Tyto komplexy se v DNA kovalentně vážou na pyrimidinové báze, přednostně na thymin (10, 11, 34, 71). Reakce Os,L s pyrimidinovými bázemi je konformačně specifická a v nativní dsDNA ze sterických důvodů téměř neprobíhá. Proto je možné tyto látky použít jako citlivé sondy pro studium lokálních struktur, ve kterých

se vyskytují nespárované nukleotidy (10, 72, 73). Adukty Os,L s DNA lze detekovat na různých typech elektrod (69, 73-76). Na rtuťových a některých amalgámových elektrodách poskytuje DNA modifikovaná Os,bipy v neutrálním a kyselém pH trojici reversibilních signálů v rozmezí 0 až –1 V a signál katalytického vylučování vodíku při –1,3 V (obr. 11) (77). Na uhlíkových elektrodách rovněž podléhají jak volné Os,L, tak i adukty Os,L-DNA několika faradaickým procesům odpovídajícím postupné redukci/oxidací atomu osmia. Bylo ukázáno, že Os,bipy – DNA adukt poskytuje specifický pík na PGE, který se potenciálové liší od píku volného komplexu (rozdíl 140 mV) (78). Tato skutečnost dovoluje stanovit DNA- Os,bipy i ve zbytkovém množství volného Os,bipy.

Obr. 11: Schéma redoxních potenciálů přirozené DNA (CA – C a A redukce, G guaninový signál na HMDE, Gox, Aox, Cox, Tox- oxidace nukleobazí), 7-deazapurinů (G*ox, A*ox) a příkladů elektroaktivních značek. Převzato z (48).

3.3 Enzymatická inkorporace elektrochemicky značených nukleotidů

Další možností elektrochemického značení DNA a oligonukleotidů je enzymatická inkorporace chemicky modifikovaných nukleotidů. První zprávu o inkorporaci značených dUTP a rUTP podal Langer už v roce 1981 (79). S využitím různých enzymů došlo k různě úspěšné inkorporaci v pozici 5 biotinem značených dUTP a rUTP. Nejúspěšnější byla inkorporace dUbioTP pomocí E. Coli Pol1, nejméně pak byly úspěšné pokusy o zařazení rUbioTP pomocí RNA polymeráz.

Nukleosidtrifosfáty modifikované na zbytku báze elektroaktivní skupinou je možné připravit např. Sonogashira (80-83) nebo Suzuki-Miyaura (84-86) cross-coupling reakcí s halogenovanými nukleosidtrifosfáty, případně nukleosidy, které jsou následně trifosforylovány (87-91). Příklad takovéto reakce je na Obr. 12. Pro značení DNA se používají nukleotidy modifikované elektroaktivními skupinami, které je možné elektrochemicky stanovit i ve velmi nízké koncentraci a které poskytují signály při potenciálech mimo oblast vybíjení elektrolytu a elektrodových dějů odpovídajícím redukci či oxidaci přirozených nukleobází. Pro řadu analytických aplikací je výhodou, pokud jsou redoxní změny elektroaktivní skupiny reverzibilní a/nebo pokud při jejich elektrochemické přeměně dochází k přenosu většího počtu elektronů, zajišťující vyšší citlivost stanovení. Z těchto hledisek se v naší laboratoři osvědčily např. antrachinon (13) ferocen (14) (oba vykazují reverzibilní elektrochemii), nitro skupina (15) (redukuje se ireverzibilně čtyřmi elektrony na hydroxylamin, který lze reverzibilně oxidovat na nitroso skupinu), nebo benzofurazan (12) (poskytuje ireverzibilní šestielektronovou redukci).

Obr. 12: Schéma přípravy dNTP modifikovaných propargylkarbamoylantrachinonem.

i) PAQ, [Pd(PPh3)2Cl2], CuI, (iPr)2EtN, DMF, 1 h, 75°C. ii) PAQ, Pd(OAc)2, TPPTS,

CuI, (iPr)2EtN, CH3CN:H2O (2:1), 1 h, 75°C. Převzato z (13)

Pro přípravu oligonukleotidů s inkorporovanými modifikovanými nukleotidy se používá metoda prodlužování primeru (PEX). Při této metodě je pomocí různých DNA polymeráz prodlužován řetězec DNA od 5´ konce ke 3´ konci. Syntéza probíhá sekvenčně specificky podle templátu (Obr. 13). DNA polymerázy dokáží inkorporovat jak přirozené, tak i vhodně chemicky modifikované nukleotidy (48).

Kromě DNA polymeráz závislých na templátu je známá polymeráza schopná katalyzovat prodloužení řetězce DNA bez použití templátu, tzv. terminální deoxynukleotidyltransferáza (TdT). Pomocí tohoto enzymu je možné připravit DNA nebo oligonukleotidy značené na 3´ konci modifikovanými nukleotidy, často v delších jednořetězcových přesazích (tail-labeling) (92). Bylo zjištěno, že účinnost zařazování nepřirozených dNTP pomocí TdT nezávisí na jejich vodíkových vazbách nebo interakci mezi jejich π elektrony navzájem, ale hlavně na absolutní velikosti. Neschopnost zařazovat objemné nepřirozené dNTP nejspíše vyplývá ze sterického bránění (93).

Obr. 13: Schéma přípravy jedno- nebo dvouřetězcových oligonukleotidů značených elektroaktivní skupinou (X), vázanou v tomto případě na A metodou prodlužování primeru

3.4 Magnetoseparační techniky

Pro separaci a purifikaci produktů PEX nebo jiných metod značení DNA se často využívají magnetoseparační techniky. Tyto techniky využívají magnetický nosič (nejčastěji paramagnetické částice), na který je DNA po inkubaci s interagujícími činidly navázána. Částice s navázanou DNA jsou promyty vhodnými činidly a DNA je následně z nosiče uvolněna (např. zahřátím). Na magnetický nosič lze DNA ukotvit pomocí vhodného adaptoru. V praxi se uplatňují zejména systém streptavidin(avidin) – biotin, kde se DNA s navázaným biotinem váže na magnetické částice nesoucí streptavidin, nebo hybridizace oligo(A) sekvence připojené ke studované DNA s oligo(T) sekvencí kovalentně vázanou na povrchu nosiče (94-96) (Obr. 14). Magnetoseparační techniky umožňují rychlou a účinnou separaci DNA z reakční směsi a ve spojení s AdTS voltametrií představují citlivou metodu pro stanovení velmi malého množství DNA i v nadbytku interferujících látek.

Obr. 14: Schéma separace DNA pomocí magnetických částic s navázaným streptavidinem (DBstr)

4 Elektrochemické biosenzory Senzor je zařízení, které snímá danou veličinu (většinou fyzikální nebo chemickou) a transformuje ji fyzikálním převodem na veličinu výstupní. V případě elektrochemických senzorů je senzor tvořen elektrodou, která je napojena na elektrický zdroj a vyhodnocovací zařízení. Pokud se jedná o biosenzory, je součástí elektrody biologický materiál (tzv. biorekogniční prvek), kterým může být např. DNA, enzym, protein a další (97). První biosenzor byl navržen už v šedesátých letech 20. století L. C. Clarkem. Jedná se o biosenzor na stanovení množství glukózy na základě její oxidace enzymem glukosaoxidázou (98). Tento senzor tvoří glokózooxidáza imobilizovaná na povrch kyslíkové elektrody pomocí dialyzační membrány. Enzym je schopný katalyzovat oxidaci glukózy, což je spojeno s úbytkem kyslíku. Ten byl měřen pomocí kyslíkové elektrody a změna jeho koncentrace odpovídala koncentraci glukózy ve vzorku.

V případě elektrochemického senzoru pro hybridizaci NK je biorekogničním prvkem hybridizační sonda, analytem komplementární cílové vlákno a převodníkem signálu pracovní elektroda, na které je sonda imobilizována.

4.1 Elektrochemické biosenzory pro detekci hybridizace DNA

Hybridizací NK se rozumí tvorba dvoušroubovice NK ze dvou komplementárních vláken. Techniky molekulární hybridizace využívají specifický úsek NK – hybridizační sondu o známé sekvenci bází k detekci komplementárního vlákna NK. Za vhodných podmínek může

dojít k vytvoření dvoušroubovice mezi sondou a cílovým řetězcem NK přítomným v analyzovaném vzorku.

Pro konstrukci hybridizačního biosenzoru jsou důležité výběr pracovní elektrody a detekční strategie a s nimi související způsob imobilizace sondy na povrch elektrody a blokování povrchu elektrody proti nespecifické adsorpci NK a dalších látek (3, 99-101).

Jako senzory se v oblasti detekce hybridizace NK používají pevné pracovní elektrody, zejména elektrody uhlíkové a zlaté.

4.1.1 Imobilizace hybridizační sondy na povrchu elektrod

Hybridizační sondu lze na uhlíkových elektrodách zakotvit buď prostou fyzikální adsorpcí, nebo kovalentně. Fyzikální adsorpce (používaná zejména pro elektrody z pyrolytického grafitu a uhlíkové pastové elektrody) nezajišťuje specifickou orientaci zakotveného vlákna, ale vlákno na povrchu „leží“. Takto připravený senzor je poměrně nestabilní a může dojít k desorpci sondy z povrchu. I přes tyto potenciální nevýhody byly takto připravené senzory úspěšně využity a to zejména v experimentech využívajících krátké časy hybridizace za nízkých teplot a pro experimentální uspořádání, kdy je na povrch elektrody adsorbována přímo cílová DNA a ta je hybridizována se značenou sondou (68).

Ke kovalentnímu zakotvení sondy na povrchu uhlíkových elektrod se využívá úprava povrchu elektrody nebo přímo elektrodového materiálu (např. modifikovaná uhlíková pasta) (102). Lze využít např. derivátů karbodiimidu, který zprostředkuje vazbu mezi aminoskupinami DNA a funkčními skupinami na povrchu elektrod (např. elektrochemicky generované karboxylové skupiny). Takto lze imobilizovat jak přirozené NK, tak NK koncově modifikované primární alifatickou aminoskupinou (tyto se na povrch elektrody vážou orientovaně a mohou se účinněji hybridizovat). Další možností imobilizace NK na povrch elektrody je imobilizace NK koncově značené vhodným adaptorem (např. biotinem) na elektrodu modifikovanou vhodným afinitním partnerem k danému adaptoru (např. streptavidinem).

Pro zakotvení hybridizačních sond na povrchu zlatých elektrod se využívá zejména interakce thiolových skupin kovalentně navázaných na jednom konci syntetických oligonukleotidů s povrchem zlaté elektrody. Za vhodných experimentálních podmínek lze tímto způsobem připravit vysoce samoorganizovanou monovrstvu označovanou jako SAM

(self-assambled monolayer) (103-105). Pro optimální citlivost detekce hybridizace je důležité optimální pokrytí povrchu elektrody sondou. Pro dosažení optimální hustoty pokrytí sondou a také pro zablokování volného povrhu elektrody se využívá derivátů thioalkanů (např. 6- mercapto-1-hexanol).

4.1.2 Detekční principy využívané v elektrochemických senzorech pro hybridizaci DNA

1. Využití vlastní elektrochemické aktivity NK

Vlastní elektrochemické aktivity bází NK lze využít v případě, kdy se obsah některé báze v sondě a cílové DNA výrazně liší. Nejčastěji se využívá oxidačního signálu guaninu, příp. adeninu (např. (106)).

2. Impedanční měření

Elektroda s imobilizovanou ssDNA vykazuje jinou diferenciální kapacitu než elektroda nesoucí dsDNA, takže při vzniku duplexu dojde ke změně vlastností elektrické dvojvrstvy. Tuto změnu je možné detekovat. Pro sledování změn hustoty náboje na povrchu elektrody při tvorbě duplexu lze využít měření redoxních signálů anionických depolarizátorů (např. komplexy železa – FeII/FeIII) (104).

3. Elektroaktivní indikátory

Tato metoda využívá elektrochemicky aktivní látky, která se s vysokou preferencí váží na dsDNA, zejména interkalátory a bis-interkalátory (viz kapitola 3.1). Vznik dsDNA se projeví nárůstem signálu interkalátoru (52, 62).

4. Využití elektroaktivních značek

Na cílovou DNA nebo na signální hybridizační sondu je kovalentně navázána vhodné elektrochemická značka (viz kapitola 3.2) (42, 47, 74).

5. Elektrochemický molekulární maják

Na povrchu elektrody je za jeden konec zakotvena sonda, která zaujímá strukturu vlásenky (107). Na druhém konci je navázána elektrochemicky aktivní značka. Pokud je sonda v podobě vlásenky, značka se nachází v blízkosti povrchu elektrody a je elektrochemicky detekovatelná. Pokud dojde k hybridizaci, vznikne lineární duplex, čímž

dojde k oddálení značky od elektrody a elektrochemický signál značky poklesne, zmizí – vypnutí signálu („signal off“) v důsledku proběhnutí sledovaného děje.

6. Signální sondy značené enzymy

Výhodou značení NK enzymy je amplifikace signálu, kdy jedna molekula enzymu může katalyzovat přeměnu velkého množství substrátu na elektrochemicky aktivní produkt, který je elektrochemicky stanovován. Enzymovou značku lze navázat na NK pomocí biotin-streptavidinové technologie (68, 108, 109).

7. Značení DNA pomocí nanočástic

Jako nanočástice se obvykle označují částice o průměru od 1 do 100 nm. Tyto částice mají charakteristické fyzikální a chemické vlastnosti, které závisí na jejich materiálu, rozměru, tvaru a případně na použité stabilizující látce (110). Vhodná nanočástice (obvykle zlatá, stříbrná nebo např. ze sulfidů zinku nebo kadmia) se naváže na konec cílového vlákna NK (111-114). Po hybridizaci a separaci mohou být nanočástice rozpuštěny a příslušný kov je elektrochemicky stanoven. Aplikace nanočástic zajišťuje vysokou citlivost detekce, protože cílová DNA nese větší počet elektrochemicky aktivních molekul nebo atomů tvořících příslušnou částici.

4.2 Elektrochemické senzory pro detekci mutací a polymorfismů

Analýza odchylek v lidském genomu má význam z hlediska včasné diagnostiky určitých onemocnění a prevence jejich šíření. Elektrochemické metody mohou sloužit jako rychlý a levný nástroj k detekci mutací v molekule DNA. Tyto metody využívají hybridizaci studované DNA se sondou, proto jsou použitelné pouze v případech, kdy je dané onemocnění charakterizováno mutací ve známé sekvenci DNA. V tomto případě je analyzovaná sekvence DNA hybridizována se sondou navrženou speciálně pro odhalení testované mutace. Při hybridizaci může dojít ke dvěma situacím. Buď dvoušroubovici vytvoří dva plně komplementární řetězce a vzniká homoduplex, nebo navržená sonda není plně komplementární s cílovou DNA a vzniká heteroduplex. Heteroduplex je oproti homoduplexu méně stabilní a snáze se denaturuje. Čím je v duplexu více nekomplementárních míst, tím je duplex méně stabilní. Pokud se provede hybridizace za vysoce stringentních podmínek (vyšší

teplota, nižší iontová síla), je výtěžek příslušného heteroduplexu nižší, než výtěžek homoduplexu, což se projeví snížením měřeného signálu. V krajním případě heteroduplex vůbec nevznikne.

V minulosti bylo zjištěno, že je DNA schopná přenášet elektrony prostřednictvím - elektronů aromatických zbytků bazí zapojených do systému stohových (stacking) interakcí uvnitř šroubovice (115-118). Schopnost DNA přenášet náboj je výrazně eliminována až potlačena, jsou-li v duplexu DNA přítomny chybně spárované nebo nespárované báze. Tohoto jevu lze využít pro studium bodových mutací. Např. Kelley a spol. využívá systém, kdy je duplex DNA zakotvený jedním koncem na elektrodě a zprostředkovává přenos elektronů mezi elektrodou a redox aktivním interkalátorem vázaném na druhém konci duplexu (119).

Pro detekci mutací je také možné využít protein MutS, který se váže na chybně spárované nebo nespárovaná baze v molekule DNA. MutS vázaný na heteroduplexy lze elektrochemicky stanovit (120, 121).

Další strategií na detekci mutací v molekule DNA je „minisekvenování“, které využívá metodu prodlužování primeru (viz kapitola 3.3). K cílové sekvenci je připojen primer navržený tak, aby končil těsně před místem předpokládané mutace. K reakční směsi je přidán elektroaktivně značený nukleosid trifosfát komplementární k očekávané mutaci a vhodná DNA polymeráza. Pokud DNA obsahuje mutaci, dojde k zařazení značeného nukleotidu do DNA a ten je elektrochemicky detekován. V principu lze pracovat se čtyřmi různými značenými dNTP a v jednom kroku tak zjistit všechny možné varianty změny v sekvenci (122). Tuto metodu je možné aplikovat přímo na povrchu elektrody (123) nebo ve spojení s dvoupovrchovými technikami založenými na magnetických mikročásticích (124).

Cíle dizertační práce

1. Prozkoumat elektrochemické vlastnosti nových potenciálních elektroaktivních značek pro DNA, zejména takových, které poskytují reverzibilní nebo multielektronové redoxní děje na rtuťových nebo uhlíkových elektrodách

2. Prozkoumat elektrochemické vlastnosti syntetických oligonukleotidů nebo DNA modifikovaných novými typy elektroaktivních značek

3. Nalézt podmínky umožňující současnou detekci dvou nebo více elektroaktivních modifikací DNA pro využití v analýze nukleotidových sekvencí

Seznam publikací

(1) Vidlakova, P.; Pivonkova, H.; Fojta M.; Havran, L.: Electrochemical behavior of anthraquinone- and nitrophenyllabeled deoxynucleoside triphosphates: a contribution to development of multipotential redox labeling of DNA. Monatshefte fur Chemie 2015, 146, 839 - 847

(2) Balintova, J.; Plucnara, M.; Vidlakova, P.; Pohl, R.; Havran, L.; Fojta, M.; Hocek, M.: Benzofurazane as a New Redox Label for Electrochemical Detection of DNA: Towards Multipotential Redox Coding of DNA Bases. Chemistry-a European Journal 2013, 19, 12720-12731.

(3) Balintova, J.; Pohl, R.; Horakova, P.; Vidlakova, P.; Havran, L.; Fojta, M.; Hocek, M.: Anthraquinone as a Redox Label for DNA: Synthesis, Enzymatic Incorporation, and Electrochemistry of Anthraquinone-Modified Nucleosides, Nucleotides, and DNA. Chemistry-a European Journal 2011, 17, 14063-14073.

(4) Dadova, J.; Vidlakova, P.; Pohl, R.; Havran, L.; Fojta, M.; Hocek, M.: Aqueous Heck Cross-Coupling Preparation of Acrylate-Modified Nucleotides and Nucleoside Triphosphates for Polymerase Synthesis of Acrylate-Labeled DNA. Journal of Organic Chemistry 2013, 78, 9627-9637.

(5) Horakova, P.; Tesnohlidkova, L.; Havran, L.; Vidlakova, P.; Pivonkova, H.; Fojta, M.: Determination of the Level of DNA Modification with Cisplatin by Catalytic Hydrogen Evolution at Mercury-Based Electrodes. Analytical Chemistry 2010, 82, 2969-2976.

Výsledky a diskuse

1. Elektrochemické chování antrachinonen a nitrofenylskupinou značených nukleosid trifosfátů Vidlakova, P.; Pivonkova, H.; Fojta M.; Havran, L.: Electrochemical behavior of anthraquinone- and nitrophenyllabeled deoxynucleoside triphosphates: a contribution to development of multipotential redox labeling of DNA. Monatshefte fur Chemie 2015, 146, 839 - 847 Přestože je přirozená DNA sama o sobě elektrochemicky aktivní a je možné ji stanovit na různých typech elektrod (21, 125), je pro řadu analytických aplikací praktické použít DNA značenou elektroaktivními molekulami. Tyto látky podléhají redoxním reakcím a dávají tak modifikované DNA nové elektrochemické vlastnosti. Takto značené molekuly mohou být využity v biologických, medicínských i nanotechnologických aplikacích. Při studiu struktury, poškození i interakcí DNA může být výhodné použít několik elektrochemických značek současně.

V naší práci jsme studovali elektrochemické chování dATP a dCTP značených antrachinonem nebo nitrofenylskupinou za různých podmínek a také možnost stanovit je současně.

Elektrochemické chování nukleosidtrifosfátů značených antrachinonem nebo nitroskupinou bylo studováno pomocí CV na HMDE. Pro elektrochemické chování antrachinonu je charakteristická dvouelektronová redoxní chinon/hydrochinon přeměna. V katodické větvi cyklického voltamogramu antrachinon poskytuje pík AQred při potenciálu okolo -0,4 V, příslušející redukci antrachinonu na antrahydrochinon. V anodické větvi ox cyklického voltamogramu je patrný pík AQH2 příslušející zpětné oxidaci antrahydrochinonu ox (obr.15 A,B). Intenzita píku AQH2 závisí na potenciálu bodu obratu. Intenzita tohoto píku je největší při potenciálech bodu obratu -0,6 - -1,2 V, při potenciálech zápornějších než -1,4 ox V výška píku prudce klesá a při potenciálech zápornějších než -1,6 V pík AQH2 na voltamogramu nepozorujeme (Obr. 16).

Nitroskupina během CV na HMDE poskytuje za daných podmínek při potenciálu okolo - red 0,45 V katodický pík NO2 , příslušející čtyřelektronové redukci nitroskupiny na hydroxylamin. Takto vzniklý hydroxylamin je při potenciálu kolem 0,0 V dvouelektronově

reverzibilně oxidován a poskytuje anodický pík NHOHox (Obr. 15 C,D). Intenzita píku NHOHox je také závislá na potenciálu bodu obratu, ale s posunem bodu obratu ox k negativnějším potenciálům klesá mnohem méně, než intenzita píku AQH2 (Obr. 16).

Zkoumali jsme možnost současného stanovení antrachinonu a nitroskupiny. Vzhledem k blízkým hodnotám potenciálu redukce obou skupin jsou katodické píky antrachinonu a nitrofenylové skupiny často velmi obtížně rozlišitelné. Protože je však redukce nitroskupiny ireverzibilní a neposkytuje žádný oxidační signál v oblasti potenciálů, kde by interferoval s oxidací antrahydrochinonu. Produkt ireverzibilní redukce nitroskupiny navíc poskytuje ox ox oxidační signál NHOH , jehož potenciál se od potenciálu píku AQH2 liší o cca 400 mV a tudíž se oba signály neovlivňují.

Obr. 15: CV dCPAQTP (A), dAPAQTP (B), dNCNO2TP (C) a dNANO2TP (D) na HMDE základní elektrolyt 0,3 M mravenčan amonný, 0,05 M fosforečnan sodný, pH 6,9, počáteční potenciál 0,05 V, potenciál obratu -1,85 V (přerušovaná čára), počáteční potenciál 0,05 V, potenciál bodu obratu -0,6 V (plná čára).

0.7

0.6

0.5

0.4

A  0.3

Ip/

0.2 AQHox (dCPAQTP) 0.1 2 NHOHox (dCPhNO2TP) 0.0 -2.0 -1.8 -1.6 -1.4 -1.2 -1.0 -0.8 -0.6 Esw/V ox ox Obr 16: Závislost intenzity píku AQH2 (černá čára) a píku NHOH (červená čára) na potenciálu bodu obratu.

2. Benzofurazan jako redoxní značka pro analýzu DNA Balintova, J.; Plucnara, M.; Vidlakova, P.; Pohl, R.; Havran, L.; Fojta, M.; Hocek, M.: Benzofurazane as a New Redox Label for Electrochemical Detection of DNA: Towards Multipotential Redox Coding of DNA Bases. Chemistry-a European Journal 2013, 19, 12720- 12731. Tato práce se zabývá použitím benzofurazanu jako redoxní značky pro analýzu sekvence DNA. Benzofurazan je známý pro svoje fluorescenční vlastnosti (126-128) a použití v organických elektronických materiálech (129), ale pro značení DNA zatím nebyl používán.

S využitím cross-coupling reakcí byly připraveny dATP a dCTP s benzfurazanem vázaným buď přímo na bazi (Obr.17A), nebo přes acetylenový linker (Obr. 17B). Tyto dNTP byly následně PEX reakcí s využítim KOD XL polymerázy inkorporovány do ON.

Elektrochemické vlastnosti značených dNTP a oligonukleotidů byly studovány pomocí cyklické a square wave voltametrie na HMDE a PGE. V případě CV na HMDE jsou v cyklickém voltamogramu neznačených ODN měřeném při počátečním potenciálu 0 V a potenciálu bodu obratu -1,85 V patrné dva signály příslušející elektrochemickým reakcím bází. Jedná se o katodický pík CA při potenciálu okolo -1,5 V příslušející ireverzibilní redukci cytosinu a adeninu a anodický pík G při potenciálu -0,25 V příslušející oxidaci 7,8- dihydrogenguaninu generovaného redukcí guaninu při potenciálech nižších než -1,6 V (21, 108). Ve voltamogramech značených oligonukleotidů je kromě píků příslušejících redoxním reakcím bazí i pík benzofurazanu (Obr. 18). Benzofurazan podléhá elektrochemické redukci a poskytuje intenzivní ireverzibilní katodický pík BFred v oblasti mezi -0,7 a -0,85 V (HMDE) nebo mezi -0,9 a -1,0 V (PGE).

Obr. 17: Schéma přípravy dNTP modifikovaných benzfurazanem. A: i), iii) BF-B(OH)2 (1), Pd(OAc)2, 3,3′,3′′-phosphanetriyltris(benzenesulfonic acid) trisodium salt (TPPTS), Cs2CO3, CH3CN/H2O (1:2), 1 h, 75 °C. ii) 1) PO(OMe)3, POCl3, 0 °C; 2) (NHBu3)2H2P2O7, Bu3N, DMF, 0 °C; 3) triethylammonium bicarbonate (TEAB). B: i) BF-C CH (2), [Pd(PPh3)2Cl2], (iPr)2EtN, CuI, DMF, 1 h, 75 °C; ii) 2, Pd(OAc)2, TPPTS, (iPr)2EtN, CuI, CH3CN/H2O (1:2), 1 h, 75 °C.

Obr. 18: CV na HMDE oligonukleotidů značených benzfurazanem vázaným přímo na bázi (BF)nebo přes acetylenový linker (EBF) a kontroly, kdy do reakční směsi nebyla přidána polymeráza.

Kromě PEX produktů značených pouze benzofurazanem byly připraveny i oligonukleotidy značené benzofurazanem (CBF) a nitrofenyl skupinou (ANO2) současně. Tyto produkty byly opět analyzovány pomocí CV a SWV na HMDE a PGE. Nitroskupina podléhá red redukci na hydroxylamin a poskytuje intenzivní pík NO2 v oblasti okolo -0,45 V, což je o red red cca 300 mV pozitivnější, než redukce benzofurazanu (Obr. 19). Intenzita píků BF a NO2

velmi dobře koresponduje s počtem značek v oligonukleotidu i s počtem elektronů účastnících se redukce (6 u BF a 4 u NO2) a může být použita pro sekvenční analýzu DNA, například při detekci mutací, založenou na stanovení poměru dvou značených nukleobazí.

Obr. 19: CV na HMDE (A) a SWV na PGE (B) oligonukleotidů značených benzofurazanem a nitrofenyl skupinou.

3. Antrachinon jako redoxní značka pro analýzu DNA Balintova, J.; Pohl, R.; Horakova, P.; Vidlakova, P.; Havran, L.; Fojta, M.; Hocek, M.: Anthraquinone as a Redox Label for DNA: Synthesis, Enzymatic Incorporation, and Electrochemistry of Anthraquinone-Modified Nucleosides, Nucleotides, and DNA. Chemistry- a European Journal 2011, 17, 14063-14073.

V této práci prezentujeme metodu značení DNA s využitím dATP a dCTP modifikovaných propargylkarbamoylantrachinonem (PAQ) a ethynylantrachinonem (EAQ).

Modifikované nukleosidtrifosfáty byly připraveny Sonogashira cross-coupling reakcí 2- ethynylantrachinonu a N-(-2-propynyl)-antrachinoncarboamidu s halogenovanými nukleosidtrifosfáty. Modifikované dNTP byly pomocí PEX metody s využitím KOD XL DNA polymerasy inkorporovány do oligonukleotidů. Takto připravené oligonukleotidy byly

přečištěny pomocí streptavidinových magnetických kuliček a následně elektrochemicky analyzovány.

Elektrochemické chování PEX produktů značených antrachinonem bylo studováno pomocí cyklické voltametrie a square wave voltametrie na HMDE a PGE. V případě CV na HMDE jsou v cyklickém voltamogramu neznačených ODN měřeném při počátečním potenciálu 0 V a potenciálu bodu obratu -1,85 V patrné dva signály příslušející elektrochemickým reakcím bází. Jedná se o katodický pík CA při potenciálu okolo -1,5 V příslušející ireverzibilní redukci cytosinu a adeninu a anodický pík G při potenciálu -0,25 V příslušející oxidaci 7,8-dihydrogenguaninu generovaného redukcí guaninu při potenciálech nižších než -1,6 V (21, 108). Na cyklickém voltamogramu ODN značených antrachinonem je kromě píků CA a G v katodické větvi dobře vyvinutý pík v potenciálové oblasti okolo -0,4 V příslušející redukci antrachinonu (Obr. 20A ). V „krátkém“ skenu CV (počáteční potenciál 0 V, potenciál bodu obratu -0,8 V, který není dostatečně negativní pro redukci guaninu) měřeném na HMDE nebo PGE neznačené ODN neposkytují žádný signál, zatímco ODN značené antrachinonem poskytují signály příslušející reverzibilní redukci antrachinonu (Obr.20B ).

Obr. 20 : AdTS CV PEX produktů na HMDE - základní elektrolyt 0,3 M mravenčan amonný, 0,05 M fosforečnan sodný, pH 6,9, A - počáteční potenciál 0.0 V, potenciál obratu - 1.85 V, B - počáteční potenciál 0.0 V, potenciál bodu obratu -0,8 V.

Na square wave voltamogramu nemodifikovaných ODN měřeném na PGE jsou patrné dva píky příslušející oxidacím bází guaninu Gox v oblasti okolo 1,2 V a adeninu Aox v oblasti okolo 1,4 V. V případě ODN značených antrachinonem je na voltamogramu kromě signálů příslušejících adeninu a guaninu v DNA patrný i pík antrachinonu při potenciálu okolo -0,4 V (Obr. 21).

Gox 12 neznačené s antrachinonem

9 Aox



I 6 AQ

0 -0.6 -0.4 -0.2 0.8 1.0 1.2 1.4 1.6 E [V]

Obr. 21: AdTS SWV PEX produktů na PGE – základní elektrolyt 0,2 M acetátový pufr, pH 5, počáteční potenciál -1 V, konečný potenciál 1,6 V.

V této práci prezentujeme metodu značení DNA s využitím nukleosidtrifosfátů modifikovaných antrachinonem. Značenou DNA je možné velmi dobře elektrochemicky detekovat na různých typech elektrod. Výhodou antrachinonu jako elektrochemické značky je reverzibilita jeho redoxních změn. Tato vlastnost je výhodná zejména při konstrukci biosenzorů s DNA kovalentně vázanou na povrch pevných elektrod (např. zlatých), protože umožňuje provádět více elektrochemických měření s jednou navázanou DNA.

4. Příprava nukleosid trifosfátů modifikovaných butylakrylátem pro polymerázovou syntézu značené DNA

Dadova, J.; Vidlakova, P.; Pohl, R.; Havran, L.; Fojta, M.; Hocek, M.: Aqueous Heck Cross- Coupling Preparation of Acrylate-Modified Nucleotides and Nucleoside Triphosphates for Polymerase Synthesis of Acrylate-Labeled DNA. Journal of Organic Chemistry 2013, 78, 9627-9637. Tato práce se zabývá přípravou dNTP modifikovaných butylakrylátem a využitím těchto dNTP ke značení DNA inkorporací pomocí DNA polymeráz. Butylakrylátem modifikované nukleosidy, nukleosid monofosfáty a nukleosid trifosfáty byly připraveny cross-coupling reakcemi n-butylakrylátu s jodovanými nukleosidy a nukleosid trifosfáty (Obr. 22). Zatímco syntéza dUBA, dABA a dGBA probíhala velmi dobře s vysokými výtěžky, dCI vykazoval mnohem nižší reaktivitu a výtěžek reakce byl nízký (okolo 14%).

Připravené dNBATP byly použity pro syntézu značených oligonukleotidů metodou prodlužování primeru s využitím KOD XL, Vent(exo-) a Pwo polymeráz. Zatímco dABATP, dCBATP, dUBATP byly všemi polymerázami bez komplikací začleňovány, dGBATP polymerázovou reakci inhiboval a značený oligonukleotid nevznikal.

Obr. 22: Schéma přípravy dNTP modifikovaných butylakrylátem: (i) butyl acrylate, Pd(OAc)2, PPh3, Et3N, DMF; (ii) butyl acrylate, Pd(OAc)2, TPPTS, Et3N, CH3CN/H2O (1:1); (iii)PO(OMe)3, POCl3, 0 °C; (iv) (1) PO(OMe)3, POCl3, 0 °C, (2) NHBu3)2H2P2O7, Bu3N, DMF, 0 °C, (3) 2 M TEAB; (v) butyl acrylate, Pd(OAc)2, TPPTS, Et3N, CH3CN/H2O (1:1).

Jelikož dCBATP bylo obtížné připravit a dGBATP inhibovalo polymerázovou reakci, byly pro studium elektrochemického chování oligonukleotidů značených butylakrylátem připraveny PEX produkty s inkorporovanými ABA a UBA. Značené nukleosidy i oligonukleotidy byly studovány pomocí cyklické voltametrie na HMDE. Na Obr.23A,B jsou cyklické voltamogramy modifikovaných nukleosidů a nukleosid monofosfátů. dABA a dABAMP poskytují dva elektrochemické signály – pík odpovídající redukci A okolo - 1,43 V a pík příslušející redukci butylakrylátu BAred okolo -1,3 V. V cyklickém voltamogramu dUBA a dUBAMP je pouze signál odpovídající redukci butylakrylátu, protože redukce uracilu není na HMDE detekovatelná. V souladu s předchozími studiemi redukce karbonylových sloučenin (130-132) lze předpokládat, že primárním místem redukce butylakrylátu bude dvojná vazba  C=C.

V cyklickém voltamogramu ODN (Obr. 23C) jsou patrné dva signály příslušející elektrochemickým reakcím bází. Jedná se o katodický pík CA při potenciálu okolo -1,5 V příslušející ireverzibilní redukci cytosinu a adeninu a anodický pík G při potenciálu - 0,25 V (21, 108). Ve voltamogramu oligonukleotidů s inkorporovaným ABA nebo UBA (4 BA v jednom ON) je navíc signál příslušející redukci butylakrylátu BAred při potenciálu - 1,4 V (stejný potenciál pro obě modifikované báze).

Obr 23: Cyklické voltamogramy dABA, dABAMP (A), dUBA, dUBAMP (B) a značených i neznačených oligonukleotidů (C)

Z výsledků naší studie vyplývá, že je možné připravit dNTP modifikované butylakrylátem a s výjimkou dGBATP je možné je inkorporovat do DNA a elektrochemicky detekovat. Butylakrylát lze pomocí voltametrie na HMDE velmi dobře detekovat díky vzniku redukčního signálu při potenciálu okolo -1,4 V, což je méně negativní potenciál než je redukce C a A. Zároveň je to výrazně negativnější potenciál, než u ostatních dosud navržených značek (12-15, 133), takže by bylo možné tuto značku použít i pro značení DNA více značkami současně.

5. Stanovení stupně modifikace DNA cisplatinou s využítím katalytického vylučování vodíku na rtuťových elektrodách

Horakova, P.; Tesnohlidkova, L.; Havran, L.; Vidlakova, P.; Pivonkova, H.; Fojta, M.: Determination of the Level of DNA Modification with Cisplatin by Catalytic Hydrogen Evolution at Mercury-Based Electrodes. Analytical Chemistry 2010, 82, 2969-2976.

Cisplatina patří společně s dalšími komplexy platiny oxaliplatinou a carboplatinou mezi cytostatika používaná k léčbě různých druhů nádorů (134). Tyto látky se mohou kovalentně vázat na molekuly nukleových kyselin a vytvářet různé adukty (135). Nejčastějším vazebným místem platinových derivátů je guanin. V naší práci byla použita voltametrická analýza na rtuťových elektrodách pro studium DNA modifikované cisplatinou s využitím katalytického vylučování vodíku doprovázející redoxní jevy u těchto aduktů.

Ke studiu modifikace DNA cisplatinou byla používána square wave voltametrie a cyklická voltametrie. V připadě CV byly měřeny voltamogramy nemodifikované a silně platinované DNA. V případě cyklického voltamogramu nemodifikované DNA byly naměřeny dva signály příslušející elektrochemickým reakcím bází DNA. Jedná se o katodický pík CA v oblasti -1,5 V příslušející ireverzibilní redukci cytosinu a adeninu a pík G příslušející oxidaci 7,8-dihydrogenguaninu generovaného redukcí guaninu při potenciálech menších než -1,6 V. V případě DNA modifikované cisplatinou dojde ve voltamogramu k viditelným změnám. V katodické části dochází k růstu proudu v oblasti kolem -1,2 V. Negativní proud dosahuje maxima v -1,75 V a tvoří široký pík. V anodické části voltamogramu jsou patrné tři vlny (okolo -1.75, -1.45 a -1.3 V), v oblasti mezi -1,53 až -1,18 V má anodická část voltamogramu stejnou polaritu jako katodická část. Toto chování souvisí s katalytickým vyvíjením vodíku při s elektrochemických reakcích platinované složky DNA. Zároveň je ve voltamogramu s nárůstem stupně platinace patrné snižování intenzity píku G a jeho posun do negativnějších potenciálů.

Pro podrobnější studii signálů příslušejících platinované DNA byla použita square wave voltametrie. Byly naměřeny voltamogramy nemodifikované DNA a DNA modifikované cisplatinou do různých rb (poměr cisplatina/nukleotid). Ve voltamogramu DNA modifikované cisplatinou jsou patrné dva signály, pík G v oblasti -0,26 V, který koresponduje se signálem ve voltamogramu nemodifikované DNA a pík P v oblasti -1,25 V, který je charakteristický

pro platinou modifikovanou DNA. Intenzita tohoto píku koresponduje se stupněm modifikace a roste lineárně přibližně do rb 0,12 (Obr 24).

Obr. 24: (A) SWV nemodifikované DNA a DNA modifikované cisplatinou do různého rb, (B) závislost intenzity píku P na koncentraci cisplatiny

V této práci byla prokázána možnost analytického využití katalytických proudů souvisejících s redoxními procesy platinované složky DNA v průběhu anodické polarizace následující po pre-redukci platinou modifikované DNA na HMDE. Z této studie vyplývá, že při modifikaci DNA cisplatinou je možné voltametricky stanovit i modifikace do nízkého stupně (rb 0,01).

Závěr

Tato práce je zaměřena na chemickou modifikaci a značení oligonukleotidů a nukleových kyselin elektrochemicky aktivními molekulami a studium elektrochemického chování značených oligonukleotidů s využitím různých voltametrických metod a různých typů elektrod.

Hlavní část práce je zaměřena na studium vlastností dNTP s kovaletně vázanými elektroaktivními molekulami (antrachinon, nitrofenol, benzofurazan, butylakrylát). U těchto látek bylo pomocí cyklické a square wave voltametrie studováno elektrochemické chování za různých podmínek na rtuťových a uhlíkových elektrodách. Bylo zjištěno, že všechny tyto látky lze velmi dobře elektrochemicky detekovat a to jak samostatně, tak i více značek současně.

Modifikované dNTP byly pomocí polymerázových reakcí inkorporovány do oligonukleotidů a následně byly pomocí elektrochemických metod studovány jejich vlastnosti. DNA polymerázy jsou do oligonukleotidového řetězce schopné začlenit jak přirozené, tak i chemicky modifikované dNTP, což umožňuje připravit oligonukleotid značený elektrochemickými (nebo třeba i fluorescenčními) značkami. Tento způsob elektrochemického značení lze využít při konstrukci různých biosenzorů pro detekci hybridizace, poškození i bodových mutací DNA.

Seznam literatury

1. Palecek E. 1958. Naturwissenschaft 45: 186-7 2. Miller IR. 1961. J Mol Biol 3: 229-+ 3. Palecek E, Fojta M. 2005. Bioelectronics: From Theory to Applications: 127-92 4. Fojta M. 2004. Collect Czech Chem Comm 69: 715-47 5. Palecek E, Jelen F. 2002. Crit Rev Anal Chem 32: 261-70 6. Vacek J, Havran L, Fojta M. 2011. Chemicke Listy 105: 15-26 7. Palecek E, Bartosik M. 2012. Chem Rev 112: 3427-81 8. Palecek E, Postbieglova I. 1986. J Electroanal Chem 214: 359-71 9. Palecek E. 1992. Methods in Enzymology 212: 305-18 10. Palecek E. 1992. Method Enzymol 212: 139-55 11. Jelen F, Karlovsky P, Makaturova E, Pecinka P, Palecek E. 1991. Gen Physiol Biophys 10: 461- 73 12. Balintova J, Plucnara M, Vidlakova P, Pohl R, Havran L, et al. 2013. Chem-Eur J 19: 12720-31 13. Balintova J, Pohl R, Horakova P, Vidlakova P, Havran L, et al. 2011. Chem-Eur J 17: 14063-73 14. Brazdilova P, Vrabel M, Pohl R, Pivonkova H, Havran L, et al. 2007. Chem-Eur J 13: 9527-33 15. Cahova H, Havran L, Brazdilova P, Pivonkova H, Pohl R, et al. 2008. ANGEW CHEM INT EDIT 47: 2059-62 16. Hocek M, Fojta M. 2008. Org Biomol Chem 6: 2233-41 17. Herne TM, Tarlov MJ. 1997. J Am Chem Soc 119: 8916-20 18. Bond AM, Grabaric BS. 1979. Anal Chem 51: 126-8 19. Bond AM, Grabaric BS, Jones RD, Rumble NW. 1979. J Electroanal Chem 100: 625-40 20. Palecek E, Tkac J, Bartosik M, Bertok T, Ostatna V, Palecek J. 2015. Chem Rev 115: 2045-108 21. Palecek E, Jelen F. 2005. In Electrochemistry of Nucleic Acids and Proteins: Towards Electrochemical Sensors for Genomics and Proteomics, ed. E Palecek, F Scheller, J Wang, pp. 73-173 22. Jelen F, Tomschik M, Palecek E. 1997. J Electroanal Chem 423: 141-8 23. Bartosik M, Palecek E. 2011. Electroanalysis 23: 1311-9 24. Palecek E. 1980. Anal Biochem 108: 129-38 25. Palecek E, Frary BD. 1966. Arch Biochem Biophys 115: 431-+ 26. Strasak L, Dvorak J, Hason S, Vetterl V. 2002. Bioelectrochemistry 56: 37-41 27. Palecek E, Tomschik M, Stankova V, Havran L. 1997. Electroanalysis 9: 990-7 28. Bartosik M, Fojta M, Palecek E. 2012. Electrochim Acta 78: 75-81 29. Trnkova L, Studnickova M, Palecek E. 1980. Bioelectrochemistry and Bioenergetics 7: 643-58 30. Palecek E, Fojta M, Jelen F. 2002. Bioelectrochemistry 56: 85-90 31. Cummings TE, Elving PJ. 1979. J Electroanal Chem 102: 237-48 32. Fojta M, Bowater RP, Stankova V, Havran L, Lilley DMJ, Palecek E. 1998. Biochemistry 37: 4853-62 33. Palecek E. 1992. Bioelectrochemistry and Bioenergetics 28: 71-83 34. Fojta M. 2002. Electroanalysis 14: 1449-63 35. Palecek E. 1976. Progress in nucleic acid research and molecular biology 18: 151-213 36. Brabec V, Dryhurst G. 1978. J Electroanal Chem 91: 219-29 37. Brabec V, Dryhurst G. 1978. J Electroanal Chem 89: 161-73 38. Brabec V. 1981. Bioelectrochemistry and Bioenergetics 8: 437-49 39. Dryhurst G, Pace GF. 1970. J Electrochem Soc 117: 1259-&

40. Oliveira-Brett AM, Piedade JAP, Silva LA, Diculescu VC. 2004. Anal Biochem 332: 321-9 41. Boussicault F, Robert M. 2008. Chem Rev 108: 2622-45 42. Fojta M, Havran L, Kizek R, Billova S, Palecek E. 2004. Biosens & Bioelectron 20: 985-94 43. Vrabel M, Horakova P, Pivonkova H, Kalachova L, Cernocka H, et al. 2009. Chemistry-a European Journal 15: 1144-54 44. Surkus A-E, Flechsig G-U. 2009. Electroanalysis 21: 1119-23 45. Peter J, Reske T, Flechsig G-U. 2007. Electroanalysis 19: 1356-61 46. Flechsig GU, Peter J, Hartwich G, Wang J, Grundler P. 2005. Langmuir 21: 7848-53 47. Fojta M, Kostecka P, Trefulka MR, Havran L, Palecek E. 2007. Anal Chem 79: 1022-9 48. Hocek M, Fojta M. 2011. Chem Soc Rev 40: 5802-14 49. Simonova A, Balintova J, Pohl R, Havran L, Fojta M, Hocek M. 2014. Chempluschem 79: 1703- 12 50. Balintova J, Plucnara M, Vidlakova P, Pohl R, Havran L, et al. 2013. Chemistry-a European Journal 19: 12720-31 51. Cheng GF, Zhao J, Tu YH, He PA, Fang YH. 2005. Anal Chim Acta 533: 11-6 52. Jelen F, Erdem A, Palecek E. 2002. Bioelectrochemistry 55: 165-7 53. Kerman K, Ozkan D, Kara P, Meric B, Gooding JJ, Ozsoz M. 2002. Anal Chim Acta 462: 39-47 54. Gu JY, Lu XJ, Ju HX. 2002. Electroanalysis 14: 949-54 55. Wong ELS, Gooding JJ. 2003. Anal Chem 75: 3845-52 56. Yamana K, Kawakami N, Ohtsuka T, Sugie Y, Nakano H, Saito I. 2003. Nuc ac res 89-90 57. Millan KM, Saraullo A, Mikkelsen SR. 1994. Anal Chem 66: 2943-8 58. Erdem A, Meric B, Kerman K, Dalbasti T, Ozsoz M. 1999. Electroanalysis 11: 1372-6 59. Maruyama K, Mishima Y, Minagawa K, Motonaka J. 2002. Anal Chem 74: 3698-703 60. Mugweru A, Wang BQ, Rusling J. 2004. Anal Chem 76: 5557-63 61. Tansil NC, Xie H, Xie F, Gao ZQ. 2005. Anal Chem 77: 126-34 62. Takenaka S, Ohtuka K, Miyahara H, Nojima T, Takagi M. 2002. Nuc ac res 291-2 63. Xie H, Tansil NC, Gao ZQ. 2006. Frontiers in Bioscience-Landmark 11: 1147-57 64. Miyahara H, Yamashita K, Kanai M, Uchida K, Takagi M, et al. 2002. Talanta 56: 829-35 65. Nakayama M, Ihara T, Nakano K, Maeda M. 2002. Talanta 56: 857-66 66. Umek RM, Lin SW, Vielmetter J, Terbrueggen RH, Irvine B, et al. 2001. J Mol Diagnost 3: 74- 84 67. Vrabel M, Hocek M, Havran L, Fojta M, Votruba I, et al. 2007. Eur J Inorg Chem: 1752-69 68. Fojta M, Brazdilova P, Cahova K, Pecinka P. 2006. Electroanalysis 18: 141-51 69. Palecek E, Hung MA. 1983. Anal Biochem 132: 236-42 70. Lukasova E, Jelen F, Palecek E. 1982. Gen Physiol Biophys 1: 53-70 71. Reske T, Surkus A-E, Duwensee H, Flechsig G-U. 2009. Microchim Acta 166: 197-201 72. Palecek E. 1991. Crit Rev Biochem Mol 26: 151-226 73. Jacobsen M, Flechsig G-U. 2013. Electroanalysis 25: 373-9 74. Fojta M, Havran L, Billova S, Kostecka P, Masarik M, Kizek R. 2003. Electroanalysis 15: 431-40 75. Kostecka P, Havran L, Pivonkova H, Fojta M. 2004. Bioelectrochemistry 63: 245-8 76. Trefulka M, Ferreyra N, Ostatna V, Fojta M, Rivas G, Palecek E. 2007. Electroanalysis 19: 1334-8 77. Yosypchuk B, Heyrovsky M, Palecek E, Novotny L. 2002. Electroanalysis 14: 1488-93 78. Fojta M, Havran L, Kizek R, Billova S. 2002. Talanta 56: 867-74 79. Langer PR, Waldrop AA, Ward DC. 1981. Proc Natl Acad Sci USA 78: 6633-7 80. Capek P, Cahova H, Pohl R, Hocek M, Gloeckner C, Marx A. 2007. Chem-Eur J 13: 6196-203 81. Capek P, Pohl R, Hocek M. 2006. Organic & Biomolecular Chemistry 4: 2278-84 82. Cristalli G, Vittori S, Eleuteri A, Volpini R, Camaioni E, et al. 1995. J Med Chem 38: 4019-25 83. Casalnuovo AL, Calabrese JC. 1990. J Am Chem Soc 112: 4324-30 84. Capek P, Pohl R, Hocek M. 2006. Org Biomol Chem 4: 2278-84 85. Western EC, Daft JR, Johnson EM, Gannett PM, Shaughnessy KH. 2003. J Org Chem 68: 6767- 74

86. Western EC, Shaughnessy KH. 2005. J Org Chem 70: 6378-88 87. Kuwahara M, Nagashima J-i, Hasegawa M, Tamura T, Kitagata R, et al. 2006. Nuc Ac Res 34: 5383-94 88. Gramlich PME, Warncke S, Gierlich J, Carell T. 2008. Angew Chem Int Ed 47: 3442-4 89. 2005. Nucleoside Triphosphates and their Analogs: Chemistry, Biotechnology, and Biological Applications 90. Hocek M, Fojta M. 2008. Organic & Biomolecular Chemistry 6: 2233-41 91. Jager S, Rasched G, Kornreich-Leshem H, Engeser M, Thum O, Famulok M. 2005. J Am Chem Soc 127: 15071-82 92. Horakova P, Macickova-Cahova H, Pivonkova H, Spacek J, Havran L, et al. 2011. Org Biomol Chem 9: 1366-71 93. Berdis AJ, McCutcheon D. 2007. Chembiochem 8: 1399-408 94. Wang J, Xu DK, Erdem A, Polsky R, Salazar MA. 2002. Talanta 56: 931-8 95. Palecek E, Fojta M. 2007. Talanta 74: 276-90 96. Palecek E, Kizek R, Havran L, Billova S, Fojta M. 2002. Analytica Chimica Acta 469: 73-83 97. Grieshaber D, MacKenzie R, Voeroes J, Reimhult E. 2008. Sensors 8: 1400-58 98. Clark LC, Lyons C. 1963. ANN NY ACAD SCI 102: 29-& 99. Labuda J, Oliveira Brett AM, Evtugyn G, Fojta M, Mascini M, et al. 2010. Pure Appl Chem 82: 1161-87 100. Wang J. 2005. Electrochemistry of Nucleic Acids and Proteins: Towards Electrochemical Sensors for Genomics and Proteomics 1: 175-94 101. Tarlov MJ, Steel AB. 2003. Biomolecular Films: Design, Function, and Applications 111: 545- 608 102. Drummond TG, Hill MG, Barton JK. 2003. Nat Biotechnol 21: 1192-9 103. Zhao YD, Pang DW, Hu S, Wang ZL, Cheng JK, Dai HP. 1999. Talanta 49: 751-6 104. Keighley SD, Li P, Estrela P, Mighorato P. 2008. Biosens Bioelectron 23: 1291-7 105. Levicky R, Herne TM, Tarlov MJ, Satija SK. 1998. J Am Chem Soc 120: 9787-92 106. Hason S, Pivonkova H, Vetterl V, Fojta M. 2008. Anal Chem 80: 2391-9 107. Palecek E. 2004. Trends in Biotechnology 22: 55-8 108. Fojta M, Havran L, Vojtiskova M, Palecek E. 2004. J Am Chem Soc 126: 6532-3 109. Won BY, Lee DW, Shin SC, Cho D-Y, Lee SS, et al. 2008. Biosensors & Bioelectronics 24: 665-9 110. Niemeyer CM, Simon U. 2005. Eur J Inorg Chem: 3641-55 111. Wang J. 2005. Electrochemistry of Nucleic Acids and Proteins: Towards Electrochemical Sensors for Genomics and Proteomics 1: 369-84 112. Wang J, Xu DK, Kawde AN, Polsky R. 2001. Anal Chem 73: 5576-81 113. Authier L, Grossiord C, Brossier P, Limoges B. 2001. Anal Chem 73: 4450-6 114. Wang J, Liu GD, Polsky R, Merkoci A. 2002. Electrochem Commun 4: 722-6 115. Bixon M, Jortner J. 2002. Chem Phys 281: 393-408 116. Delaney S, Barton JK. 2003. J Org Chem 68: 6475-83 117. Treadway CR, Hill MG, Barton JK. 2002. Chem Phys 281: 409-28 118. Boon EM, Barton JK. 2002. CURR OPIN STRUC BIOL 12: 320-9 119. Kelley SO, Boon EM, Barton JK, Jackson NM, Hill MG. 1999. Nucleid Acid Res 27: 4830-7 120. Palecek E, Masarik M, Kizek R, Kuhlmeier D, Hassmann J, Schulein J. 2004. Anal Chem 76: 5930-6 121. Masarik M, Cahova K, Kizek R, Palecek E, Fojta M. 2007. Anal Bioanal Chem 388: 259-70 122. Di Giusto DA, Wlassoff WA, Giesebrecht S, Gooding JJ, King GC. 2004. J Am Chem Soc 126: 4120-1 123. Patolsky F, Lichtenstein A, Willner I. 2001. Nat Biotechnol 19: 253-7 124. Patolsky F, Weizmann Y, Katz E, Willner I. 2003. Angew Chem In Ed 42: 2372-6 125. Palecek E. 2002. Talanta 56: 807- 126. Goncalves MST. 2010. In Advanced Flourescence Reporters in Chemistry and Biology I: Fundamentals and Molecular Design, ed. AP Demchenko, pp. 27-64

127. Uchiyama S, Santa T, Imai K. 2000. Analyst 125: 1839-45 128. Rohacova J, Luisa Marin M, Martinez-Romero A, Diaz L, Enrique O'Connor J, et al. 2009. Chemmedchem 4: 466-72 129. Zhang B, Hu X, Wang M, Xiao H, Gong X, et al. 2012. New J Chem 36: 2042-7 130. Zimmer JP, Richards JA, Turner JC, Evans DH. 1971. Anal Chem 43: 1000-& 131. Klemm LH, Olson DR. 1979. J Org Chem 44: 4524-7 132. Fahr T, Petr A, Dunsch L. 1997. Ber Bunsenges Phys Chem 101: 1040-4 133. Vrabel M, Horakova P, Pivonkova H, Kalachova L, Cernocka H, et al. 2009. Chem-Eur J 15: 1144-54 134. Hrubisko M. 1987. Bratislavske lekarske listy 88: 223-9 135. Kasparkova J, Vrana O, Farrell N, Brabec V. 2004. J Inorg Biochem 98: 1560-9

Monatsh Chem (2015) 146:839–847 DOI 10.1007/s00706-015-1435-6

ORIGINAL PAPER

Electrochemical behavior of anthraquinone- and nitrophenyllabeled deoxynucleoside triphosphates: a contribution to development of multipotential redox labeling of DNA

Pavlı´na Vidla´kova´ • Hana Pivonˇkova´ • Miroslav Fojta • Ludeˇk Havran

Received: 5 December 2014 / Accepted: 2 February 2015 / Published online: 25 February 2015 Ó Springer-Verlag Wien 2015

Abstract Electrochemical properties of base-modiﬁed Graphical abstract cytosine or 7-deazaadenine nucleoside triphosphates (dNTPs) bearing electrochemically active anthraquinone or 0.7 0.6

3-nitrophenyl moieties were studied using cyclic voltam- 0.5 metry with the hanging mercury drop electrode. The 0.4 /μA p anthraquinone moiety in the dNTPs gives well-pronounced I 0.3 0.2 AQHox (dCPAQ TP) reversible quinone/hydroquinone redox signals around 0.1 2 NHOHox (dCPhNO 2 TP) -0.40 V (against Ag|AgCl|3M KCl reference electrode), 0.0 -2.0 -1.8 -1.6 -1.4 -1.2 -1.0 -0.8 -0.6 while the nitro group in 3-nitrophenyl exhibits irreversible Esw/V reduction to hydroxylamine around -0.45 V that can be reversibly oxidized to corresponding nitroso compound close to 0.0 V. Both anthraquinone and hydroxylamine Keywords Nucleoside triphosphate Á DNA labeling Á redox groups can be selectively switched off by further DNA electrochemistry Á Anthraquinone Á electrochemical transformation, depending on negative Nitro compounds Á Cyclic voltammetry potential applied and composition of the background electrolyte. Results of this study suggest that both nucleobase and the conjugate label moiety influence remarkably Introduction the adsorbability and/or intermolecular interactions taking part at the electrode surface. The potential analytical uti- Labeling of nucleic acids (NA) by various electroactive lization of these phenomena is discussed. tags is of broad interest to scientists in connection with the development of electrochemical methods and biosensors for NA analysis, such as analysis of nucleotide sequences or sensing of DNA damage (reviewed in [1–3]). Redox labels can be introduced into NA via chemical phospho- ramidite-based synthesis of oligonucleotides, by chemical modification of natural NA components (such as thymine residues in DNA by osmium tetroxide reagents [4]or30- terminal ribose in RNA by six-valent osmium complexes & P. Vidla´kova´ Á H. Pivonˇkova´ Á M. Fojta Á L. Havran ( ) [5]), or enzymatically using polymerases and labeled Institute of Biophysics, Academy of Sciences of the Czech Republic, v.v.i., Kralovopolska 135, 612 65 Brno, (deoxy)nucleoside triphosphates [(d)NTPs] as monomer Czech Republic substrates. The latter approach has proved especially effi- e-mail: [email protected] cient and versatile for the preparation of not only the redox- labeled DNA, but also DNA bearing fluorophores [6, 7]or M. Fojta Central European Institute of Technology, Masaryk University, chemically reactive groups for further chemical transfor- Kamenice 753/5, 625 00 Brno, Czech Republic mations on DNA [8, 9] or for bioconjugation with proteins 123 840 P. Vidla´kova´ et al.

[10]. A number of modified dNTP bearing diverse elec- of AQ and PhNO2 signals when anodic responses were trochemically active moieties, such as ferrocene [11], measured [14]. organic nitrocompounds [12], [Ru/Os-(bpy)3] complexes [13], anthraquinone [14], benzofurazane [15], Cyclic voltammetry of AQ-dNTP conjugates methoxyphenol [16], phenylazide [8], and others (reviewed in [2, 17]), have been developed and applied. A combi- Here, we studied in more detail the electrochemical be- PAQ nation of these labels (to encode different nucleotide havior of AQ and PhNO2-labeled dNTPs, dC TP, sequences, or even each nucleobase with different tags dAPAQTP, dCPhNO2 TP, and dAPhNO2 TP. Cyclic voltam- being electrochemically reduced or oxidized at different mograms (CVs) of these four compounds at concentrations potentials) allows parallel analysis of multiple nucleotide of 40 lM, measured in 0.3 M ammonium formate, 0.05 M sequences [18], typing of sequence polymorphisms [12, sodium phosphate, pH 6.9 (a medium optimized for elec- 13], or a simple monitoring of the conversion of one redox trochemical analysis of DNA at mercury electrodes, tag to another, e.g., to probe interactions of the modified suitable for simultaneous detection of natural electroactive DNA with proteins [8]. DNA components [1, 3]) without pre-accumulation of the Anthraquinone (AQ), as a moiety exhibiting well-pro- analytes, are presented in Fig. 1. Red curves correspond to nounced reversible electrochemistry [19–21], has been CVs measured with initial potential Ei = 0.0 V and utilized for redox labeling of biomolecules [22–24]. switching potential Esw =-0.85 V, while black dotted Derivatives of AQ linked via various tethers to nucleosides curves were obtained for Esw =-1.85 V. Both AQ-dNTP were used, for example, to study DNA-mediated charge conjugates (Fig. 1a, b), when measured with Esw = transfer [25–27]. Base-modified cytosine and 7-deaza- -0.85 V, gave a pair of peaks around -0.4 V corresponding adenine dNTPs bearing the AQ labels have recently been to the reversible anthraquinone/anthrahydroquinone redox developed and used for polymerase synthesis of AQ- system: modified oligodeoxynucleotides, and utilization of the AQ À þ AQ + 2 e +2H $ AQH2 ð1Þ tags for dual redox labeling of DNA in combination with PAQ earlier introduced nitrophenyl (PhNO2) labels in simple The behavior of dC TP (Fig. 1a) was nevertheless in model applications have been tested [14]. However, a more some respects different from that of dAPAQTP (Fig. 1b). detailed study of electrochemical properties of First, a considerable difference in the heights of cathodic red ox oligodeoxyribonucleotides (ODNs) bearing AQ or PhNO2 peak AQ and anodic peak AQH2 (the latter being 3 groups, or their combination, is to date missing. In this times higher than the former for 40 lM dCPAQTP; see paper, we present a comparative study of base-modified concentration dependences for further discussion) was PAQ PAQ AQ or PhNO2 dNTP conjugates using cyclic voltammetry observed for dC TP, while for dA TP the intensity with the hanging mercury drop electrode. of the cathodic peak AQred was about 3.5-times higher compared to the analogous signal of dCPAQTP, and the ox PAQ anodic peak AQH2 of dA TP was higher by only 33 % Results and discussion as compared to the peak AQred of the same conjugate (Table 1; compare also solid curves in Fig. 2a, where In our previous study [14], anthraquinone-labeled dCTP details of the voltammograms are shown). Second, peak-to- and 7-deaza-dATP were synthesized and used for DNA peak separation for the AQ/AQH2 redox process was labeling via incorporation of corresponding nucleotides 19 mV in dCPAQTP and 48 mV in dAPAQTP, suggesting into ODNs by DNA polymerases (for general methodolo- more facile electron transfer in the first instance. Third, for gies of this strategy of modified nucleic acids construction, dAPAQTP, clearly developed second pair of peaks at see reviews [2, 17, 28]). Electrochemical measurements potentials more negative by 27 mV were observed. revealed the modified ODNs to give well-developed signals Differences in the relative intensities of the anodic and due to reversible redox electrochemistry of the AQ moiety. cathodic peaks of dCPAQTP and dAPAQTP can be Experiments focused on simultaneous detection of the AQ attributed to different adsorbabilities of the two tags with another type of organic electrochemically active conjugates, with the dAPAQTP adsorbing at the mercury moieties attached to nucleobases in modified ODNs, 3-ni- surface more efficiently. The experiment, the results of trophenyl [12], indicated possibilities of using the two which is shown in Fig. 2, supports such explanation: when labels for convenient dual DNA labeling when optimum dCPAQTP was allowed to accumulate at the electrode conditions for their distinction are applied. Namely, con- surface with open current circuit for 60 s before the CVs red version of the PhNO2 group into the corresponding were measured, the height of the peak AQ increased by ox hydroxylamine derivative (PhNHOH) via irreversible four- about three times. Peak AQH2 became significantly higher electron reduction of the nitro group facilitated resolution after pre-accumulation and its height was practically the 123 Electrochemical behavior of dNTPs 841

Fig. 1 Cyclic voltammograms of dCPAQTP (a), dAPAQTP (b), rate 1 V/s, background electrolyte: 0.3 M ammonium formate, dCPhNO2 TP (c), and dAPhNO2 TP (d). Cyclic voltammetry (CV) at 0.05 M sodium phosphate, pH 6.9, and concentration of all substances HMDE: Ei ?0.0 V (a, b), or ?0.1 (c, d), Esw -1.85 or -0.85 V, scan was 40 lM

Table 1 Heights and potentials red ox red ox Peak AQ AQH2 of peaks AQ and AQH2 PAQ obtained for dA TP and Potential/V Height/lA Potential/V Height/lA PAQ red dC TP and of peaks NO2 and NHOHox for dAPhNO2 TP dAPAQTP -0.428 0.55 -0.380 0.41 PhNO2 and dC TP dCPAQTP -0.394 0.16 -0.375 0.49 red ox Peak NO2 NHOH Potential/V Height/lA Potential/V Height/lA

dAPhNO2 TP -0.457 1.01 -0.037 0.31 dCPhNO2 TP -0.435 2.40 -0.007 0.50

red PAQ ox PAQ PAQ same as the height of the peak AQ .ThedA TP peak AQH2 disappeared in both dC TP and dA TP, exhibited similar effects upon the accumulation, giving suggesting a deeper reduction of the AQ moiety upon ap- wide cathodic and anodic peaks in which the two reversible plying the highly negative potentials. Blocking of the pairs (distinguishable in CV measured without electrode surface by the reduction products then probably accumulation, Fig. 2) were clearly merged. prevents fresh dNPAQTP from the bulk of solution to give ox PAQ When the CVs were measured with the Esw =-1.85 V peak AQH2 in the anodic scan. The dC TP conjugate (i.e., with a setup previously used in DNA analysis to re- (Fig. 1a) exhibited rather complicated behavior in a po- duce guanine residues and obtain an anodic peak G of tential region between -0.9 and -1.7 V, producing guanine at the mercury electrode [1, 3, 14]), the anodic several, under the given conditions, irreversible peaks in

123 842 P. Vidla´kova´ et al.

0. 9 tensammetric processes of the negatively charged dNTPs AQHox a 2 on the negatively charged surface (note the sharp ‘‘spike’’ 0. 6 at -1.45 V, Fig. 1a). On the contrary, dAPAQTP yielded

0. 3 only one well-developed cathodic peak under the same conditions (peak Ared, Fig. 1b), which can be ascribed to

A 0. 0

µ reduction of the 7-deazaadenine nucleobase. Differences / I dAPAQTP t = 0 s between the two dNTPs may be due to different adsorption -0.3 a dAPAQTP t = 60 s a modes (see above), inﬂuencing availability of different dCPA QTP t = 0 s a -0.6 red electroreducible groups for electronic communication with AQ dCPA QTP t = 60 s a the electrode. Taken together, the shapes of CVs of indi- -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 vidual PAQ conjugates, as well as differences observed E/V between dCPAQTP and dAPAQTP both in the region of the AQ reversible electrochemistry and in the more negative AQHox 0.4 b 2 potential region, suggest rather complicated processes un- dCPAQTP 1st sca n dCPAQTP 2nd scan dergone by these complex compounds on the mercury dCPAQTP 3rd scan electrode surface (see also discussion of concentration 0.2 dependences below). A µ / I 0.0 Cyclic voltammetry of PhNO2-dNTP conjugates

PhNO2 -0.2 Both nitrophenyl-labeled dNTPs, dC TP (Fig. 1c) and AQred dAPhNO2 TP (Fig. 1d), gave an irreversible cathodic signal, red -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 peak NO2 , around -0.45 V (Table 1). Nitro group is E/V known to be electrochemically reduced at various types of electrodes [29, 30] to hydroxylamine: Fig. 2 a Effects of adsorptive pre-accumulation (0 or 60 s at open À þ current circuit) on CV responses of dCAQTP and dAPAQTP. ÀNO2 +4e +4H -H2O !ÀNH-OH ð2Þ b Repeated CV scans of dCPAQTP (without pre-accumulation, E i Due to the involvement of four electrons, the latter ?0.1 V, Esw -0.6 V, scan rate 1 V/s, and other conditions as in Fig. 1) electrode reaction gives rise to a strong reduction signal allowing sensitive polarographic or voltammetric

0.7 determination of various nitro compounds [31–39], including environmental pollutants [40–44]. In several 0.6 proof-of-concept applications it has been utilized for 0.5 convenient DNA labeling as well [8, 9, 12, 14]. The hydroxylamime moiety resulting from (2) is reversibly 0.4 oxidizable by two electrons to the nitroso group [30]: A µ 0.3 À þ p/

I ÀNH-OH - 2 e -2H $ NO ð3Þ 0.2 Hydroxylamine reduction is reﬂected in anodic signals AQHox (dCPAQTP) 2 ox 0.1 (peak NHOH ) yielded by both PhNO2 dNTP conjugates NHOHox (dCPhNO2TP) close to 0.0 V (Fig. 1c, d; Table 1). In contrast to the 0.0 ox ox -2.0 -1.8 -1.6 -1.4 -1.2 -1.0 -0.8 -0.6 anodic peak AQH2 , the peak NHOH was observed on

Esw/V the CVs even when the measurements were performed with Esw =-1.85 V, displaying only partial decrease of its ox ox ox Fig. 3 Dependence of the intensity of peaks AQH2 and NHOH on intensity. Dependence of the peak NHOH height on Esw, PAQ PhNO2 switching potential (dC TP and dC TP, CV at HMDE, other measured for dCPhNO2 TP (Fig. 3), shows the peak current conditions as in Fig. 1) to be practically unchanged between Esw =-0.6 V and the cathodic scan that could be ascribed to reduction pro- -1.4, and to decrease gradually with Esw being shifted to cesses of ethynyl and carbamoyl groups in the linker, the more negative potentials; for Esw =-1.85 V the peak ox above proposed reduction of the anthrahydroquinone (see NHOH height corresponded to 55 % of value measured ox the dependence of peak AQH2 height on Esw in Fig. 3, with Esw =-0.6 V. Such behavior suggests that either the showing a steep decrease of the peak current between corresponding redox moiety is not destroyed upon the -1.1 V and -1.6 V), reduction of cytosine as well as electrode polarization to highly negative potentials, or the 123 Electrochemical behavior of dNTPs 843 electrode surface does not get fully blocked by reduction a products and the anodic peak NHOHox is produced by fresh 0.8 PAQ analyte from the bulk of solution. Since the peak NHOHox dC TP 0.6 measured with PhNO2-labeled ODN using ex situ voltammetric procedure (i.e., with adsorbed layer onto AQred A ox µ 0.4 AQH the electrode and no analyte present in the bulk of 2 p/ background electrolyte) completely disappeared for I

Esw =-1.2 V (not shown preliminary data; a complex 0.2 study with labeled ODNs will be published elsewhere), the PhNO explanation based on involvement of fresh dN 2 TP 0.0 from solution appears to be more likely. Notably, 020406080 signiﬁcant decrease of the NHOHox height in dCPhNO2 TP c/µM was observed at Esw values coinciding with potential of 1.4 b reduction of the cytosine nucleobase (Fig. 1c), and partial 1.2 PAQ electrode blocking with products of the latter reaction dA TP could cause the observed decrease of the peak NHOHox 1.0 intensity. Similarly as in the case of corresponding AQ 0.8 A]

PhNO2 µ conjugates (Fig. 1a, b), dA TP differed from 0.6 p/ dCPhNO2 TP by absence of any distinct signals in the I 0.4 AQred potential region between -0.6 and -1.85 V (Fig. 1d). The AQHox dAPhNO2 TP did not yield even the signal of 7-deazaadenine 0.2 2 reduction in the ammonium formate medium (it was 0.0 nevertheless observed in Britton–Robinson buffer at 020406080 pH B 6, see below). On the other hand, the behavior of c/µM 0.15 0.015 PhNO2 dC TP in the same potential region was similar to that c of dCPAQTP, suggesting the nucleobase to be a critical dAPhNO2 TP component of the dNTP conjugate that dictates its behavior 0.010 on the negatively charged mercury surface. 0.10 A A µ µ p/ I Effects of dNTP concentration, pH of background p/ I 0.05 NO red 0.005 electrolyte, and scan rate 2 NHOH ox red ox Dependences of intensities of peaks AQ /AQH2 mea- PAQ sured without pre-accumulation for dC TP and 0.00 0.000 PAQ red ox 020406080 dA TP, and of peaks NO2 and NHOH measured c/µM under the same conditions with dAPhNO2 TP, are shown in red PAQ Fig. 4. The height of the cathodic peak AQ of dC TP red ox red Fig. 4 Dependence of the intensity of peaks AQ , AQH2 ,NO2 , increased more or less linearly within the concentration and NHOHox on dNTP concentration: adCPAQTP, bdAPAQTP, and PhNO2 region between 0 and 80 lM (Fig. 4a). A strikingly dif- cdA TP (Esw -0.85 V and other conditions as in Fig. 1) ferent concentration dependence was observed for the ox anodic peak AQH2 of the same dNTP conjugate. At low between molecules at the surface), which under the given concentrations up to 25 lM dCPAQTP, both peaks AQred conditions is dictated by solution concentration of the ox PAQ PAQ and AQH2 followed an identical trend. However, between dC TP. Alternatively, the reduction of dC TP may PAQ ox 25 and 40 lM dC TP, the height of peak AQH2 in- be accompanied by the formation of an ordered structure of PAQ creased steeply, reached its maximum at 50 lM dC TP the adsorbed layer and orientation of the AQH2 in a way and then gradually decreased. The sigmoidal shape of the facilitating the oxidation process. Since the cooperative ox red concentration dependence suggests intermolecular inter- effect was reﬂected in peak AQH2 , but not AQ heights, actions at the electrode surface: the steep increase of the the presumptive intermolecular interactions were speciﬁc, signal around 30 lM dCPAQTP can be explained by in the case of dCPAQTP, for its reduced form. Moreover, red ox positive cooperative effects of the already adsorbed (and differences in the peak AQ and AQH2 intensities electrochemically reduced) molecules of lM dCPAQTP on measured for 40 lM dCPAQTP were retained in repeated adsorption of more molecules from the solution taking CV scans (Fig. 2b), suggesting the presumptive interaction place from a critical surface coverage (i.e., distances to be reversibly on/off switchable via changing the AQ 123 844 P. Vidla´kova´ et al.

-0.2 0.4 a b -0.4 PAQ PAQ 0.3 dA TP -0.6 dA TP AQred -0.8 AQHox 2 AQred A 0.2 red p/V -1.0

µ A ox E AQH

p/ 2 I -1.2 Ared 0.1 -1.4

-1.6 0.0 345678910 345678910 pH pH

0.2 1.2 c 0.0 d

1.0 -0.2 PhNO2 dA TP -0.4 NOred 0.8 2 NHOHox -0.6 PhNO2 A 0.6 red

p/V dA TP red µ A -0.8 NO E 2 p/ I 0.4 -1.0 NHOHox -1.2 Ared 0.2 -1.4 0.0 -1.6 345678910 345678910 pH pH

Fig. 5 Dependence of the intensity (a) and potential (b) of peaks background electrolyte (Esw -0.85 V, scan rate 1 V/s, background red ox red PAQ AQ , AQH2 , and A for dA TP and intensity (c) and potential electrolyte: Britton–Robinson buffers of the given pH and other red ox red PhNO2 (d) of peaks NO2 , NHOH , and A of dA TP on pH of conditions as in Fig. 1) redox state. In contrast to dCPAQTP, for dAPAQTP the was detectable only in pH [ 5 (Fig. 5c), most probably due S-shaped dependences of signal intensity were obtained for to reduction of the hydroxylamine to amine that took place both cathodic and anodic peaks, indicating that analogous in the acidic media (indeed, an additional pH-dependent intermolecular interactions may have occurred in both re- cathodic peak was detected in pH B 5, but not in pH above duced and oxidized forms of the latter conjugate, 5; not shown). In both conjugates, peak Ared due to re- facilitating accumulation of the oxidized form at the elec- duction of the nucleobase was observed only in pH B 6, in trode surface (see above). Concentration dependence of the agreement with earlier data showing that protonation was a PhNO2 red dA TP peak NO2 (Fig. 4c) involved a linear region prerequisite for the nucleobase reduction at the mercury between 0 and 10 lM, followed by a less steeply increas- electrode (reviewed in [3]). The potentials of all measured ing part at higher concentration. Dependence of the height signals shifted to more negative potentials with increasing of peak NHOHox exhibited certain sign of transition around pH, exhibiting almost parallel trends (Fig. 5b, d). 20 lM; nevertheless, compared to signals of the AQ con- The effects of scan rate were studied in the ammonium jugates this effect was poorly pronounced. formate medium and the results obtained for dCPAQTP and Dependences of heights and potentials of signals yielded dCPhNO2TP are displayed in Fig. 6. As could be expected by dAPAQTP and dAPhNO2TP on pH of the background for rather complex molecules of the dNTP conjugates, in- electrolyte, measured in Britton–Robinson buffer, are volving hydrophobic, hydrophilic, and/or negatively red ox shown in Fig. 5. The heights of peaks AQ and AQH2 charged parts, dependences of the measured signals on scan were almost pH-independent in a wide range between pH 3 rate mostly did not ﬁt into simple models valid for elec- and 9 (Fig. 5a), indicating that the availability of protons trode processes driven by either diffusion or adsorption. red PhNO2 were not limiting for reaction (1) to take place under the Peak NO2 of dC TP followed a slightly concave given conditions. A similar behavior was observed for peak dependence on the scan rate (growing less steeply than a red PhNO2 ox NO2 of dA TP (Fig. 5c). By contrast, peak NHOH linear function, Fig. 6b); when the peak heights were 123 Electrochemical behavior of dNTPs 845

3.5 1.0 PAQ PhNO2 a 0.8 dC TP:dC TP = 1:1 a 3.0 PAQ AQHox dC TP 0.6 2 2.5 0.4

2.0 0.2 NHOHox 2 A A

red µ µ AQ / 0.0 1.5 I 1

p/ ox I AQH 2 -0.2 1.0 -0.4 t 0 s a 0.5 t 60 s -0.6 AQ red a 0.0 -0.8 0.00.51.01.52.02.53.03.54.0 -0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 scan rate/V/s E/V 0.8 PAQ PhNO 2 8 b dC TP:dC TP = 1:2 b 0.6 ox PhNO2 AQH dC TP 2 0.4 ox 6 NHOH 0.2 2 NO red A A 2 0.0 µ µ 4 ox / NHOH I 1 p/ -0.2 I AQred -0.4 t = 0 s 2 a t = 60 s -0.6 a NOre d 2 0 -0.8 0.00.51.01.52.02.53.03.54.0 -0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 scan rate /V/s E/V 0.8 PAQ PhNO2 8 c 0.6 dC TP:dC TP = 1:3 c PhNO 2 AQHox NHOHox dC TP 0.4 2 0.2 6 2 0.0 1

A -0.2 A µ µ

4 /

I -0.4 red

p/ AQ I -0.6 NOred 2 2 -0.8 t = 0 s a -1.0 red t = 60 s NO a -1.2 2 0 0.0 0.5 1.0 1.5 2.0 -0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 1/2 E/V (scan rate) Fig. 7 Voltammetric responses of mixtures of dCPAQTP and red ox red PhNO2 PAQ PhNO2 Fig. 6 Dependence of intensity of peaks AQ , AQH2 NO2 , and dC TP at various ratios: adC TP:dC TP = 1:1, ox PAQ PhNO2 NHOH on scan rate for dC TP (a) and dC TP (b); bdCPAQTP:dCPhNO2TP = 1:2, cdCPAQTP:dCPhNO2TP = 1:3, ac- red dCPhNO2TP dependence of peak NO2 on square root of scan rate for cumulation time 0 s (dashed) or 60 s (solid; Esw -0.85 V, scan rate (c)(Esw -0.85 V, other conditions as in Fig. 1) 1 V/s, and other conditions as in Fig. 1) plotted against square root of the scan rate (Fig. 6c), the charged surface or the above discussed lateral interactions resulting dependence was signiﬁcantly convex (increasing of the AQ conjugates at the electrode surface. more steeply than a line), together indicating the combined effects of diffusion and strong adsorption. The other peaks, Mixtures of AQ- and PhNO2-labeled dNTPs red ox PAQ ox AQ and AQH2 of dC TP and NHOH of dCPhNO2TP, exhibited strongly supralinear dependences of Finally, we have been interested in the possibility of si- their heights on scan rate, suggesting more complex pro- multaneous voltammetric detection of the AQ- and PhNO2- cess possibly involving time-dependent desorption/ labeled dNTPs in mixtures (Fig. 7). For this purpose, reorientation of the dNTP molecules at the negatively dCPAQTP and CPhNO2TP were mixed at ratios 1:1, 1:2,

123 846 P. Vidla´kova´ et al. and 1:3 (keeping the dCPAQTP concentration constant at adsorbed in its oxidized (AQ) form than dCAQTP.The 20 lM), and CVs were measured without pre-accumulation shapes of concentration dependences indicate inter- (dashed lines in Fig. 7) or after a 60-s pre-accumulation molecular interactions of the AQ conjugates at the (solid lines). The results of these experiments demonstrated electrode surface that appear to be redox sensitive AQ the above discussed preferential accumulation of the AQ (specific for the reduced AQH2 form) in dC TP.Inour conjugate at the electrode surface. When the CVs were follow-up study (research in progress), the electro- measured without the pre-accumulation, peak NHOHox chemical properties of oligonucleotides modified with was detectable at the dNTPs ratio 1:1, but not the peak AQ- and PhNO2 conjugates are investigated and the NO2, the potential of which (about -0.45 V) was close to possibilities of the analytical utilization of specific prop- the potential of AQ reduction. Upon increasing the erties of the two types of DNA labels are tested (results PhNO2 PAQ C TP/dC TP ratio to 2:1 and 3:1, the peak NO2 will be published elsewhere). red ox was unmasked and peaks of the AQ /AQH2 depressed. However, after the pre-accumulation, both signals of CPhNO2TP, peak NOred, and peak NHOHox, were strongly 2 Experimental depressed even in threefold excess of the latter dNTP, suggesting that the more strongly adsorbing dCPAQTP Synthesis of anthraquinone- and 3-nitrophenyl-labeled displaced the CPhNO2TP from the electrode surface. In- deoxynucleoside triphosphates terestingly, in the presence of both conjugates, a new cathodic peak appeared around -0.65 V (well developed, Anthraquinone-modified nucleoside triphosphates (dNTPs) at the CPhNO2TP/dCPAQTP ratio of 1:1 when the mea- bearing anthraquinone attached through a propargylcar- surement was performed without pre-accumulation bamoyl linker at the 5-position of cytosine (dCPAQTP)or (Fig. 7a), and in measurements with pre-accumulation its at the 7-position of 7-deazaadenine (dAPAQTP) were pre- intensity exhibited increasing trend with increasing con- pared by Sonogashira cross-coupling of corresponding centration of CPhNO2TP). This signal may indicate a halogenated dNTPs with 2-(2-propynylcarbamoyl)an- chemical reaction between products or intermediate of thraquinone according to [14]. Analogous 3-nitrophenyl- electrochemical reduction of AQ and the nitro group at the modified dNTPs were prepared by the Suzuki–Miyaura electrode surface. reaction of 7-iodo-7-deaza-20-dATP (to obtain dAPh- NO2TP) or 5-iodo-20-deoxycytidine 50-dCTP (to obtain dCPhNO2TP) with 3-nitrophenylboronic acid according to Conclusions [12]. Both modified dNTPs were kindly donated by Prof. Michal Hocek. Cytosine or 7-deazaadenine dNTPs modified at the base residue with electrochemically active anthraquinone or 3-nitrophenyl moieties were studied using cyclic Electrochemical analysis voltammetry with the hanging mercury drop electrode. AQ moiety in the dNTP conjugates is shown to retain its Nucleoside triphosphates were analyzed by conventional well-pronounced reversible electrochemistry around - in situ CV with a hanging mercury drop electrode. CV

0.40 V. The nitro group in PhNO2 exhibits characteristic settings: scan rate 1 V/s, initial potential 0.0 V or ?0.1 V, irreversible reduction around -0.45 V. The product of switching potentials -0.85 or -1.85 V. Background elec- this reduction, phenylhydroxylamine (NHOH), gives a trolyte: 0.3 M ammonium formate, 0.05 mM sodium well-developed signal close to 0.0 V due its reversible phosphate, pH 6.9, if not stated otherwise. All measure- oxidation to the corresponding nitroso compound. Further ments were performed at room temperature by using an electrochemical reduction of the AQH2 and NHOH redox Autolab analyzer (Eco Chemie, The Netherlands) in con- groups can be used for their selective switching off, de- nection with VA-stand 663 (Metrohm, Herisau, pending on the potential applied and composition of the Switzerland). The three-electrode system was used with an background electrolyte (namely, in acidic media the sig- Ag|AgCl|3 M KCl electrode as a reference and platinum nal of NHOH oxidation disappeared, suggesting wire as an auxiliary electrode. Measurements were per- irreversible reduction of the hydroxylamine to amine). formed after deaeration of the solution by argon purging. The modified dNTPs studied differed in their adsorbability at the mercury electrode surface. In general, the Acknowledgments This work was supported by the Czech Science Foundation (grant P206/12/G151 to M.F. and 206/12/2378 to L.H.) tendency to being accumulated at the electrode was higher and by the ASCR (RVO 68081707). The authors thank Jana Balin- in AQ-modified dNTPs than in the PhNO2 derivatives, tova´, Hana Macıćˇkova´-Cahova´, and Michal Hocek (Institute of and among the former dAAQTP was more efficiently Organic Chemistry and Biochemistry, ASCR, Prague, Czech 123 Electrochemical behavior of dNTPs 847

Republic) for providing the modiﬁed nucleoside triphosphates used in 21. Quan M, Sanchez D, Wasylkiw MF, Smith DK (2007) J Am this study. Chem Soc 129:12847 22. Mahajan S, Richardson J, Ben Gaied N, Zhao Z, Brown T, Bartlett PN (2009) Electroanalysis 21:2190 References 23. Wettig SD, Bare GA, Skinner RJS, Lee JS (2003) Nano Lett 3:617 24. Zhang Y-J, He X-P, Hu M, Li Z, Shi X-X, Chen G-R (2011) Dyes 1. Fojta M, Jelen F, Havran L, Palecek E (2008) Curr Anal Chem Pigm 88:391 4:250 25. Abou-Elkhair RAI, Dixon DW, Netzel TL (2009) J Org Chem 2. Hocek M, Fojta M (2011) Chem Soc Rev 40:5802 74:4712 3. Palecek E, Bartosik M (2012) Chem Rev 112:3427 26. Gorodetsky AA, Barton JK (2007) J Am Chem Soc 129:6074 4. Fojta M, Kostecka P, Pivonkova H, Horakova P, Havran L (2011) 27. Jacobsen MF, Ferapontova EE, Gothelf KV (2009) Org Biomol Curr Anal Chem 7:35 Chem 7:905 5. Bartosik M, Trefulka M, Hrstka R, Vojtesek B, Palecek E (2013) 28. Hocek M, Fojta M (2008) Org Biomol Chem 6:2233 Electrochem Commun 33:55 29. Peckova K, Barek J, Navratil T, Yosypchuk B, Zima J (2009) 6. Dziuba D, Pohl R, Hocek M (2014) Bioconjugate Chem 1984 Anal Lett 42:2339 7. Riedl J, Pohl R, Ernsting NP, Orsag P, Fojta M, Hocek M (2012) 30. Zuman P (1993) Collect Czech Chem Commun 58:41 Chem Sci 3:2797 31. Beckett EL, Lawrence NS, Davis J, Compton RG (2002) Anal 8. Balintova J, Spacek J, Pohl R, Brazdova M, Havran L, Fojta M, Lett 35:339 Hocek M (2014) Chem Sci 6:575 32. Boateng A, Brajter-Toth A (2012) Analyst 137:4531 9. Raindlova V, Pohl R, Klepetarova B, Havran L, Simkova E, 33. Cordero-Rando MD, Barea-Zamora M, Barbera-Salvador JM, Horakova P, Pivonkova H, Fojta M, Hocek M (2012) Chem- Naranjo-Rodriguez I, Munoz-Leyva JA, de Cisneros J (1999) PlusChem 77:652 Mikrochim Acta 132:7 10. Dadova J, Orsag P, Pohl R, Brazdova M, Fojta M, Hocek M 34. De Souza D, Mascaro LH, Fatibello-Filho O (2011) Int J Anal (2013) Angew Chem Int Ed 52:10515 Chem 2011:726462 11. Brazdilova P, Vrabel M, Pohl R, Pivonkova H, Havran L, Hocek 35. Gupta S, Agarwal H, Gupta M, Verma PS (2010) J Indian Chem M, Fojta M (2007) Chem Eur J 13:9527 Soc 87:481 12. Cahova H, Havran L, Brazdilova P, Pivonkova H, Pohl R, Fojta 36. Gupta S, Gupta M, Verma PS (2009) Asian J Chem 21:7316 M, Hocek M (2008) Angew Chem Int Ed 47:2059 37. Chu L, Han L, Zhang X (2011) J Appl Electrochem 41:687 13. Vrabel M, Horakova P, Pivonkova H, Kalachova L, Cernocka H, 38. Kawde A-N, Aziz MA (2014) Electroanalysis 26:2484 Cahova H, Pohl R, Sebest P, Havran L, Hocek M, Fojta M (2009) 39. Liu Z, Zhang H, Ma H, Hou S (2011) Electroanalysis 23:2851 Chem Eur J 15:1144 40. Danhel A, Peckova K, Cizek K, Barek J, Zima J, Yosypchuk B, 14. Balintova J, Pohl R, Horakova P, Vidlakova P, Havran L, Fojta Navratil T (2007) Chem List 101:144 M, Hocek M (2011) Chem Eur J 17:14063 41. Dejmkova H, Stoica A-I, Barek J, Zima J (2011) Talanta 85:2594 15. Balintova J, Plucnara M, Vidlakova P, Pohl R, Havran L, Fojta 42. Deylova D, Yosypchuk B, Vyskocil V, Barek J (2011) Electro- M, Hocek M (2013) Chem Eur J 19:12720 analysis 23:1548 16. Simonova A, Balintova J, Pohl R, Havran L, Fojta M, Hocek M 43. Fischer J, Vanourkova L, Danhel A, Vyskocil V, Cizek K, Barek (2014) ChemPlusChem 79:1703 J, Peckova K, Yosypchuk B, Navratil T (2007) Int J Electrochem 17. Hocek M (2014) J Org Chem 79:9914 Sci 2:226 18. Fojta M, Kostecka P, Trefulka MR, Havran L, Palecek E (2007) 44. Niaz A, Fischer J, Barek J, Yosypchuk B, Sirajuddin, Bhanger MI Anal Chem 79:1022 (2009) Electroanalysis 21:1786 19. Ajloo D, Yoonesi B, Soleymanpour A (2010) Int J Electrochem Sci 5:459 20. Batchelor-McAuley C, Li Q, Dapin SM, Compton RG (2010) J Phys Chem B 114:4094

123 DOI: 10.1002/chem.201301868

Benzofurazane as a New Redox Label for Electrochemical Detection of DNA: Towards Multipotential Redox Coding of DNA Bases

Jana Balintov,[a] Medard Plucnara,[b] Pavlna Vidlkov,[b] Radek Pohl,[a] Ludeˇk Havran,[b] Miroslav Fojta,*[b, c] and Michal Hocek*[a, d]

Abstract: Benzofurazane has been at- cleotide probes. In combination with and have explored the relevant electro- tached to nucleosides and dNTPs, nitrophenyl and aminophenyl labels, chemical potentials. The combination either directly or through an acetylene we have successfully developed a of benzofurazane and nitrophenyl re- linker, as a new redox label for electro- three-potential coding of DNA bases ducible labels has proved to be excel- chemical analysis of nucleotide sequen- lent for ratiometric analysis of nucleo- ces. Primer extension incorporation of tide sequences and is suitable for bioa- Keywords: DNA polymerase · elec- the benzofurazane-modified dNTPs by nalytical applications. trochemistry · nucleoside triphos- polymerases has been developed for phates · sequencing · voltammetry the construction of labeled oligonu-

Introduction sulfanylphenyl,[10] hydrazones,[11] and so on, and by combining four different labels for the four nucleobases, established DNA biosensors[1] are broadly applied in the life sciences the first generation of multipotential redox coding[7] of and diagnostics. Electrochemical detection is a comparably DNA and applied it in minisequencing. However, in this sensitive, but less expensive alternative to current techni- first generation of redox labeling, only one or two labels ques of genomics that use optical detection[2] and, therefore, (one reducible and one oxidizable)[6] were incorporated and redox labeling can be a viable economical alternative to flu- detected in one DNA molecule. When trying to combine orescence and microarray techniques in sequencing.[3] Both two different reducible labels (nitro and anthraquinone),[9] the inherent electrochemistry of nucleic acids and the elec- electrochemical analysis of the doubly-labeled DNA gave trochemistry of additional DNA labels have been extensive- one broad signal without distinguishing the ratio of the two ly used in diverse bioanalytical applications.[4] We have pre- labels. Therefore, such first-generation labels were not prac- viously developed several new redox labels, including ferro- tical and there is still a need to develop other redox labels [5] [6] ACHTUNGRE cene, aminophenyl and nitrophenyl, [Os(bpy)3] (bpy= in order to afford a set of four labels that can be readily in- 2,2’-bipyridyl),[7] tetrathiafulvalene,[8] anthraquinone,[9] alkyl- corporated into DNA by polymerase and subsequently be independently readable in the presence of all the other [a] J. Balintov, Dr. R. Pohl, Prof. Dr. M. Hocek labels. Only such a fully orthogonal set of labels could be Institute of Organic Chemistry and Biochemistry used for the simultaneous detection of multiple nucleobase Academy of Sciences of the Czech Republic mutations in short (2–6 nt) sequences in one primer exten- Gilead Sciences and IOCB Research Center sion (as opposed to current sequencing techniques,[3] which Flemingovo nam. 2, 16610 Prague 6 (Czech Republic) use a single label for each DNA molecule for identification Fax : (+420)220-183-559 E-mail: [email protected] of one nucleotide at a time) or for the determination of the [b] M. Plucnara, P. Vidlkov, Dr. L. Havran, Prof. Dr. M. Fojta nucleobase composition of longer sequences. We report Institute of Biophysics, v.v.i. herein the development of a new redox label, benzofurazane Academy of Sciences of the Czech Republic (BF), and its use in combination with nitrophenyl and ami- Kralovopolska 135, 61265 Brno (Czech Republic) nophenyl groups in the first three-potential coding of three Fax : (+420)541-211-293 E-mail: [email protected] nucleobases. Since G is itself electrooxidizable, labeling of [c] Prof. Dr. M. Fojta A, T, and C with external labels can be regarded as provid- Central European Institute of Technology ing a complete set of redox labels for DNA coding. Masaryk University Kamenice 753/5 Oxadiazoles and benzofurazanes are an extensively stud- 62500 Brno (Czech Republic) ied class of compounds with a variety of applications. Oxa- [d] Prof. Dr. M. Hocek diazole derivatives are known antimicrobial agents.[12] Oxa- Department of Organic Chemistry, Faculty of Science diazole carboxamide deoxyribonucleoside analogues can ef- Charles University in Prague, Hlavova 8 [13] 12843 Prague 2 (Czech Republic) fectively mimic natural nucleobases in DNA replication. Supporting information for this article is available on the WWW Benzofurazanes are known for their fluorescent proper- under http://dx.doi.org/10.1002/chem.201301868. ties[14] and are used in polymer solar cells[15] and organic

12720 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Eur. J. 2013, 19, 12720 – 12731 FULL PAPER electronic materials.[16] No use of BFs in nucleic acid labeling had been reported until our recent study on hydrazone modification of DNA,[11,17] in which we first identified the promising electrochemical properties of this heterocyclic system. Therefore, we have focused on the synthesis of two types of BF derivative of nucleosides and dNTPs (connected either directly or through an acetylene linker at the 5-position of cytosine or at the 7-position of 7-deazaadenine) and their polymerase incorporation into DNA.[18]

Results and Discussion

Synthesis of BF-labeled nucleosides and dNXBFTPs: Directly linked BF derivatives of nucleosides dCBF and dABF were prepared in good yields of 69–74% in one step by Suzuki– Miyaura cross-coupling[19] of unprotected halogenated nucleosides 5-iodocytidine (dCI) and 7-deaza-7-iodoadenosine (dAI ; Scheme 1) with benzo[c][1,2,5]oxadiazole-5-boronicACHTUNGRE ACHTUNGRE acid (1) in the presence of Pd(OAc)2, TPPTS, and Cs2CO3 in CH3CN/H2O (1:2) at 758C for 1 h (Table 1, entries 1 and 2). Suzuki–Miyaura cross-coupling of halogenated dNTPs (dCITP and dAITP) under the same aqueous conditions gave the desired BF-modified dNTPs (dCBFTP and dABFTP) in moderate yields (10–22 %, Table 1, entries 3 and 4). To BF prepare larger quantities of these dN TPs, we applied an ACHTUNGRE [20] Scheme 1. Reagents and conditions: i), iii) BF-B(OH)2 (1), Pd(OAc)2, alternative strategy of triphosphorylation of the corre- 3,3’,3’’-phosphanetriyltris(benzenesulfonic acid) trisodium salt (TPPTS), BF ACHTUNGRE sponding nucleosides (dN s; Scheme 1). Thus, treatment of Cs2CO3,CH3CN/H2O (1:2), 1 h, 75 8C. ii) 1) PO(OMe)3, POCl3,08C; BF BF ACHTUNGRE 8 dC and dA with POCl3 in PO(OMe)3 followed by addi- 2) (NHBu3)2H2P2O7,Bu3N, DMF, 0 C; 3) triethylammonium bicarbonate (TEAB). tion of (NHBu3)2H2P2O7 and Bu3N and treatment with TEAB (Scheme 1) gave the desired dNBFTPs (Table 1, en- ACHTUNGRE tries 5 and 6) in yields of 24 and 70%, respectively, after iso- conditions in the presence of Pd(OAc)2, TPPTS, CuI, and lation by RP HPLC. (iPr)2EtN in CH3CN/H2O (1:2) proceeded less efficiently to Sonogashira cross-coupling reactions[19a,21] of 5-ethynyl- give dCEBF and dAEBF in moderate yields (28–45 %; Table 1, benzo[c][1,2,5]oxadiazoleACHTUNGRE (2)[22] were used to attach the BF entries 9 and 10). Analogous aqueous Sonogashira cross- moiety through an acetylene tether. The reactions of 2 with coupling reactions were used to attach the EBF group to 5-iodocytidine (dCI) and 7-deaza-7-iodoadenosine (dAI) dNTPs. Thus, the reactions of dCITP and dAITP with 2 ACHTUNGRE ACHTUNGRE were performed by using [Pd(PPh3)2Cl2], (iPr)2EtN, and CuI (Scheme 2) in the presence of Pd(OAc)2, TPPTS, CuI, and EBF in DMF at 758C for 1 h to give the desired nucleosides (iPr)2EtN in CH3CN/H2O (1:2) gave the desired dC TP dCEBF and dAEBF in good yields (60–70 %; Scheme 2, and dAEBFTP in relatively good yields of 52–54% (Table 1, Table 1, entries 7 and 8). The same reactions under aqueous entries 11 and 12).

Table 1. Preparation of BF-modified nucleosides/nucleotides. Entry Starting Reagent Catalyst Additives Solvent Product Yield compound [%][a] I ACHTUNGRE BF 1 dA 1 Pd(OAc)2, TPPTS Cs2CO3 CH3CN/H2O (1:2) dA 74 I ACHTUNGRE BF 2 dC 1 Pd(OAc)2, TPPTS Cs2CO3 CH3CN/H2O (1:2) dC 69 I ACHTUNGRE BF 3 dA TP 1 Pd(OAc)2, TPPTS Cs2CO3 CH3CN/H2O (1:2) dA TP 22 I ACHTUNGRE BF 4 dC TP 1 Pd(OAc)2, TPPTS Cs2CO3 CH3CN/H2O (1:2) dC TP 10 BF ACHTUNGRE BF 5 dA 1) PO(OMe)3, POCl3,08C; 2) (NHBu3)2H2P2O7,Bu3N, DMF, 08C; 3) TEAB dA TP 70 BF ACHTUNGRE BF 6 dC 1) PO(OMe)3, POCl3,08C; 2) (NHBu3)2H2P2O7,Bu3N, DMF, 08C; 3) TEAB dC TP 24 I ACHTUNGRE EBF 7 dA 2 [Pd(PPh3)2Cl2] CuI, (iPr)2EtN DMF dA 70 I ACHTUNGRE EBF 8 dC 2 [Pd(PPh3)2Cl2] CuI, (iPr)2EtN DMF dC 60 I ACHTUNGRE EBF 9 dA 2 Pd(OAc)2, TPPTS CuI, (iPr)2EtN CH3CN/H2O (1:2) dA 28 I ACHTUNGRE EBF 10 dC 2 Pd(OAc)2, TPPTS CuI, (iPr)2EtN CH3CN/H2O (1:2) dC 45 I ACHTUNGRE EBF 11 dA TP 2 Pd(OAc)2, TPPTS CuI, (iPr)2EtN CH3CN/H2O (1:2) dA TP 54 I ACHTUNGRE EBF 12 dC TP 2 Pd(OAc)2, TPPTS CuI, (iPr)2EtN CH3CN/H2O (1:2) dC TP 52 [a] Yields of the isolated products.

Chem. Eur. J. 2013, 19, 12720 – 12731 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.chemeurj.org 12721 M. Fojta, M. Hocek et al.

Table 2. Primers and templates used for PEX experiments.[a] Sequences Primrnd 5’-CATGGGCGGCATGGG-3’ Temprnd16 5’-CTAGCATGAGCTCAGTCCCATGCCGCCCATG-3’ TempA 5’-CCCTCCCATGCCGCCCATG-3’ TempAterm 5’-TCCCATGCCGCCCATG-3’ TempC 5’-CCCGCCCATGCCGCCCATG-3’ Temp4mod 5’-GTAGCTCACGATCAGTCCCATGCCGCCCATG-3’ TempG 5’-GTAGCATCAGCTCAGTCCCATGCCGCCCATG-3’ Temp3A 5’-TCCTCCTCCCCCATGCCGCCCATG-3’ Temp3T 5’-CACCACACCCCCATGCCGCCCATG-3’ Temp3C 5’-CGCCGCGCCCCCATGCCGCCCATG-3’ Temp3T3C 5’-ACGACGACGCCCATGCCGCCCATG-3’ Temp1T3C 5’-CGCACGCGCCCCATGCCGCCCATG-3’ Temp3T1C 5’-CACACGCACCCCATGCCGCCCATG-3’ Temp3A3C 5’-TGCTGCTCGCCCATGCCGCCCATG-3’ Temp3A1C 5’-TCCTCGCTCCCCATGCCGCCCATG-3’ Temp2A1C 5’-CTCCGCCTCCCCATGCCGCCCATG-3’ Temp1A1C 5’-CCGCCCTCCCCCATGCCGCCCATG-3’ Temp1A2C 5’-CGCCTCCGCCCCATGCCGCCCATG-3’ Temp1A3C 5’-CGCTCGCGCCCCATGCCGCCCATG-3’ Temp3T3A 5’-TCATCATCACCCATGCCGCCCATG-3’ Temp3T1A 5’-CACTCACACCCCATGCCGCCCATG-3’ Temp1T3A 5’-TCTCACTCCCCCATGCCGCCCATG-3’ Temp3A2C1T 5’-CTCACGCTCGCTCCCCATGCCGCCCATG-3’ Temp3A1C3T 5’-GCTCACTCACTCACCCATGCCGCCCATG-3’ Temp1A3C2T 5’-GCTCGCCACGCCACCCATGCCGCCCATG-3’ Temp1A2C3T 5’-CTCAGCCACGCCACCCATGCCGCCCATG-3’ Temp2A1C3T 5’-CTCACGCACCTCACCCATGCCGCCCATG-3’ Temp2A3C1T 5’-GCTCGCCACGCTCCCCATGCCGCCCATG-3’ Temp3A3C3T 5’-GTCAGCTACGTCACCCATGCCGCCCATG-3’ Temp3A0C3T 5’-CTCACTCACCTCACCCATGCCGCCCATG-3’ ACHTUNGRE Scheme 2. Reagents and conditions: i) BF-C CH (2), [Pd(PPh3)2Cl2], Temp3A3C0T 5’-GCTCGCTCCGCTCCCCATGCCGCCCATG-3’ ACHTUNGRE (iPr)2EtN, CuI, DMF, 1 h, 758C; ii) 2,Pd(OAc)2, TPPTS, (iPr)2EtN, CuI, Temp0A3C3T 5’-GCCACGCACGCCACCCATGCCGCCCATG-3’ CH3CN/H2O (1:2), 1 h, 758C. [a] In the template (temp) ONs, segments that form a duplex with primer are printed in italics, and the replicated segments are printed in bold.For magnetic separation of the extended primer strands, the template 5’-end was biotinylated. Polymerase incorporation of dNXBFTPs: In primer extension (PEX) experiments, DNA polymerase incorporates nucleotides (a natural dNTP is replaced by a modified dNXTP) at the 3’-end of a primer in the presence of the complementary template. Each PEX experiment was analyzed by polyacrylamide gel electrophoresis (PAGE). In order to avoid any misincorporation, positive (all natural dNTPs) and negative (absence of one or more dNTPs) control experiments were performed and compared with the PEX experiments by using modified dNXBFTPs. The enzymatic incorporation of the BF-modified dNXBFTPs was studied by using a number of B-family ther- mostable DNA polymerases, namely KOD XL, Pwo, Vent Figure 1. PEX single incorporations of a dNXBFTP into 19-nt DNA by (exo-), Deep Vent, and Deep Vent (exo-), which are known using tempC and tempA templates. P: primer; A+,C+: natural dNTPs; to have a high tolerance for modifications.[23] Incorporations AÀ: dGTP; CÀ: dGTP; ABF : dABFTP, dGTP; AEBF : dAEBFTP, dGTP; of one dNXBFTP were tested (for sequences of templates CBF : dCBFTP, dGTP; CEBF : dCEBFTP, dGTP. and primer, see Table 2) by using 19-mer templates tempA and tempC in combination with KOD XL, Pwo, and Vent (exo-) DNA polymerases. All four dNXBFTPs (Figure 1, modified dNXBF (dNEBFTP, dNBFTP) nucleotides using Pwo lanes 4, 5, 8, and 9) furnished the desired fully extended oli- DNA polymerase with tempC (for C, without natural gonucleotide (ON) in all cases (see Figures S1 and S2 in the dGTP), tempAterm (for A), and primrnd were compared. The Supporting Information). reaction mixtures were incubated for the time intervals indi- To explore the efficiency of the PEX experiment by using cated, and then the reactions were stopped by the addition the dNXBFTPs, we also performed a simple kinetic analysis. of PAGE loading buffer and the mixtures were immediately The rates of the PEX experiments with natural dNTPs or heated. The incorporation of the natural dNTPs was com-

12722 www.chemeurj.org 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Eur. J. 2013, 19, 12720 – 12731 Redox Label for Electrochemical Detection of DNA FULL PAPER plete within 30–60 s, whereas the PEX experiments with PEX product indicated the correct mass expected for the dNXBFTPs took 1–2 min to reach completion (Figures 2 full-length product (Figure S36 in the Supporting Informa- and 3). tion). In almost all previous studies conducted in our own and other laboratories,[18a–f] PEX incorporation of only one functionalized dNXTP (+ three unmodified dNTPs) into DNA has been performed. In principle, when using the natural four-letter genetic alphabet, it is possible to incorporate up to four different labels into one DNA molecule by PEX. Such four-label coding would be highly useful in bioanalytical applications. However, simultaneous PEX incorporation of several functionalized dNXTPs into DNA is much more XBF Figure 2. Kinetics of PEX by using dA TPs in comparison with natural challenging since the polymerase must recognize all of the dATP (+). Time intervals are given in minutes. dNXTPs as substrates and must also be able to extend the primer next to each modified nucleotide. To the best of our knowledge, only one example of the enzymatic synthesis of fully modified (at all four bases) DNA has hitherto been reported.[18a] In order to study multi-potential redox coding of DNA, we needed to develop the PEX incorporation of two or three different labels. We aimed to combine our novel BF labels with previously reported nitrophenyl and amino- Figure 3. Kinetics of PEX by using dCXBFTPs in comparison with natural phenyl groups (Scheme 3).[6] dCTP (+). Time intervals are given in minutes. First, we tested parallel incorporation of dANO2TP and dCBFTP into DNA by using Pwo and KOD XL DNA polymerases and template temprnd16 (Figure 5). In these experi- Multiple incorporations into random sequences were ments, it was necessary to perform more negative control tested by using a temprnd16 template in the presence of several DNA polymerases: KOD XL, Pwo, Vent (exo-), Deep Vent, and Deep Vent (exo-). Three of the modified dNXBFTPs(dCBFTP, dCEBFTP, dABFTP) were successfully incorporated into DNA providing full-length products in PAGE analysis (Figure 4, lanes 5, 7, and 8 and Figures S3– S7 in the Supporting Information), whereas dAEBFTP gave an ON product that appeared shorter on PAGE (Figure 4, lane 6). However, MALDI analysis of this dAEBF-containing

Figure 4. PEX multiple incorporations of dNXBFTPs into 31-nt DNA by using temprnd16 template and DNA polymerase: a) Pwo DNA polymerase; b) KOD XL DNA polymerase. P: primer; + : natural dNTPs; AÀ: dCTP, dGTP, dTTP; CÀ: dATP, dGTP, dTTP; ABF : dABFTP, dCTP, dGTP, dTTP; AEBF : dAEBFTP, dCTP, dGTP, dTTP; CBF :dATP,dCBFTP, Scheme 3. Structures of modified dNTPs (dANO2TP, dCBFTP, dTNH2TP) dGTP, dTTP; CEBF :dATP,dCEBFTP, dGTP, dTTP. and PEX for multipotential coding of DNA.

Chem. Eur. J. 2013, 19, 12720 – 12731 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.chemeurj.org 12723 M. Fojta, M. Hocek et al.

Figure 5. PEX incorporation of dCBFTP and dANO2TP into one ON by using temprnd16 template to form 31-nt DNA products. P: primer; +: nat- BF NO2 NH2 ural dNTPs; AÀ: dCTP, dGTP, dTTP; CÀ: dATP, dGTP, dTTP; A1À: Figure 6. PEX incorporation of dC -, dA -, and dT -labeled dNTPs rnd16 BF NO2 NO2 NO2 dC TP, dGTP, dTTP; C1À: dA TP, dGTP, dTTP; A : dA TP, by using temp template to form 31-nt DNA products. P: primer; +: dCTP, dGTP, dTTP; CBF :dATP,dCBFTP, dGTP, dTTP; ANO2 +CBF : natural dNTPs; AÀ: dCTP, dGTP, dTTP; CÀ: dATP, dGTP, dTTP; TÀ: NO BF BF dA 2TP, dC TP, dGTP, dTTP. dATP, dCTP, dGTP; A1À: dC TP, dGTP, dTTP; C1À: dATP, dGTP, NH2 NO2 NH2 dT TP;T1À: dA TP, dCTP, dGTP; A2À: dCTP, dGTP, dT TP; NO2 BF BF C2À: dA TP, dGTP, dTTP; T2À:dATP,dC TP, dGTP; A3À: dC TP, NH2 NO2 NH2 NO2 BF dGTP, dT TP;C3À: dA TP, dGTP, dT TP;T3À: dA TP, dC TP, tests (absence of each natural dNTP in the presence and ab- NO NO BF BF dGTP; A 2 : dA 2TP, dCTP, dGTP, dTTP; C :dATP,dC TP, dGTP, sence of the other modified dNTPs) in order to fully exclude dTTP; TNH2 : dATP, dCTP, dGTP, dTNH2TP;ANO2 +CBF + TNH2 : dANO2TP, any possible misincorporation case. Figure 5 shows that in dCBFTP, dGTP, dTNH2TP. both cases, the clean, fully extended, doubly-labeled ON products were obtained. After the successful incorporation of two redox labels, the region between À0.70 and À0.85 V, in addition to signals PEX synthesis of ONs with three different redox labels, known to correspond to reduction of cytosine (peak Cred)or dANO2TP, dCBFTP, and dTNH2TP, (Scheme 3) was tested by adenine (peak Ared)[2] at potentials more negative than using the same template. This combination of labels was À1.2 V. Peak BFred was also yielded by building blocks chosen because the oxidative aminophenyl and reductive ni- benzo[c][1,2,5]oxadiazole-5-boronicACHTUNGRE acid (1, BFBA) and EBF trophenyl labels have previously been shown to be orthogo- not containing the nucleoside/nucleotide component (Fig- nal.[6] To our delight, the results indicated that the combina- ure S32 in the Supporting Information). Considering the ab- tion of dANO2TP, dCBFTP, and dTNH2TP also resulted in good sence of any anodic signal in the voltammograms, even incorporation of all modified ONs in the PEX experiments, and the desired full-length ONs were obtained (Figure 6). Therefore, ON probes containing different combinations of these two or three labels in different sequences were prepared by PEX experiments followed by magnetoseparation (25 examples, Table 2, Figures S8–S32, see the Supporting Information) and representative examples were character- ized by MALDI (see the Supporting Information). In a few cases, PAGE analyses and MALDI showed the formation of ONs that were shorter by one nucleotide. Therefore, the templates of these sequences were extended by an additional C at the 5’-end, and this led to the PEX synthesis and isolation of the correct ON probes containing the desired number of different labels (see the Supporting Information). These labeled ON probes were then used for the electrochemical characterization (see below).

Electrochemical analysis: The voltammetric properties of the BF-modified nucleosides and dNTPs were studied by using cyclic voltammetry (CV) at a hanging mercury drop Figure 7. CV responses of BF-labeled nucleosides (A) and dNTPs (B) at HMDE. Concentration of all substances 40 mm, initial potential 0 V, electrode (HMDE) or a basal-plane pyrolytic graphite elec- switching potential À1.6 V, end potential 0 V, scan rate 1 VsÀ1, electro- trode (PGE). At the HMDE (Figure 7), the BF conjugates lyte: 0.2 m acetate buffer (pH 5.0), for other details, see the Experimental produced intense cathodic peaks (denoted BFred) in the Section.

12724 www.chemeurj.org 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Eur. J. 2013, 19, 12720 – 12731 Redox Label for Electrochemical Detection of DNA FULL PAPER when the switching potential was set immediately after the cathodic peak BFred (not shown), we concluded that peak BFred corresponded to an irreversible reduction process of the BF moiety, presumably involving six electrons and six protons to reduce two C=N double bonds in the furazane ring and release of a water molecule, thus giving rise to a di- aminobenzene derivative. At the PGE, analogous irreversible signals were observed at potentials between À0.9 and À1.0 V (Figure 8; the nucleobase reduction signals could not be observed at carbon

Figure 9. AdTS CV responses at HMDE of PEX products synthesized with biotinylated Temprnd16 template and dNTP mixes containing one of the dNXBFTP conjugates (as specified in the legend) complemented with three respective unmodified dNTPs. Single-stranded PEX products were purified by using streptavidin-coated magnetic beads. Inset: Detail of the voltammograms showing the BFred and G peaks. Switching potential À1.85 V; for other details, see Figure 7 and the Experimental Section.

ucts synthesized on the temprnd16 template, each containing four BF-modified nucleobases of one type in the extended stretch. All PEX products yielded a cathodic peak CA, corresponding to reduction of adenine and cytosine in DNA at around À1.4 V, and an anodic peak G at around À0.2 V, produced by a guanine reduction product generated in the DNA upon applying potentials more negative than À1.6 V.[2] In addition to these intrinsic DNA signals, the modified PEX products yielded well-developed, symmetrical, and ir- Figure 8. CV responses of BF-labeled nucleosides (A) and dNTPs (B) at reversible cathodic peaks at around À0.8 V, which could be PGE. Switching potential À1.4 V; for other details, see Figure 7 and the assigned to reduction of the BF moieties. Negative control Experimental Section. experiments of PEX reactions (with no polymerase added to a mixture containing a dNXBFTP complemented with electrodes[2]). The potentials of peak BFred depended to three unmodified ONs, as shown for dCBFTP in Figure 9) some extent on the type of nucleobase and on the presence produced the specific signals of the nucleobases but no peak or absence of the ethynyl linker. However, the shifts in the BFred, thus proving that peaks BFred detected for the positive potentials did not exhibit any systematic (or summative) PEX reactions were due to incorporated BF labels and not trend. For example, with nucleosides at the HMDE, the due to unremoved dNXBFTP. peak potentials followed the order dAEBF >dCEBF > dCBF > Since we did not observe any significant qualitative differ- dABF, which might indicate that electronic communication ences in electrochemical properties among the four BF between the nucleobase and the BF moiety is less influ- labels (i.e., differences in peak potentials were too small to enced by the ethynyl bridge in the case of C than in the case allow reliable discrimination between ONs labeled with of A. Nonetheless, the trends observed with the modified AXBF or CXBF or even their independent detection in a mix- dNXBFTPs and even with nucleosides dNXBF at the PGE ture), we selected dCBFTP for ON labeling with BF and fo- were different to those observed at the HMDE. This sug- cused our attention on multi-potential DNA labeling by gests that the phenomena affecting the BF reduction poten- combining BF with other electroactive moieties. For these tial are more complex, and are probably influenced by the experiments, we chose a previously developed nitrophenyl presence (in dNTPs) or absence (in nucleosides) of the neg- (PhNO2) tag as a second label that was irreversibly reduci- atively charged triphosphate group and the mode of adsorp- ble with multiple electrons (four electrons under the condi- tion at the electrode surface. The same phenomena can be tions used in this work), and an irreversibly oxidizable ami- [6] expected to influence the peak intensities, which were ob- nophenyl (PhNH2) label. First, we were interested in ascer- served to vary among the individual substances studied (see taining whether or not incorporation of two reducible tags Figures 7 and 8), including the BFBA and EBF building into one PEX product would allow their independent detec- blocks (Figure S33 in the Supporting Information). tion without significant mutual interference. Figure 10 shows In the next experiments, PEX products containing AXBF the voltammetric responses of PEX products obtained with or CXBF were prepared and subjected to electrochemical the temp3A3C3T template and dNTP mixes containing either analysis. Figure 9 shows cyclic voltammograms of PEX prod- dCBFTP, dANO2TP, or a combination of both simultaneously

Chem. Eur. J. 2013, 19, 12720 – 12731 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.chemeurj.org 12725 M. Fojta, M. Hocek et al.

dANO2TP, dCBFTP, dTNH2TP, and unlabeled dGTP, where the number of individual labels incorporated per ON was dictated by the sequence of the given template (note that the same set of PEX products was analyzed for completeness and accuracy of synthesis by PAGE and MALDI-TOF MS; see above and the Supporting Information). Voltammetric responses of the PEX products were again measured at the HMDE (reductions) and the PGE (both reductions and oxi- dations). Results of the measurements at the HMDE are summarized in Figure 11A, which shows the intensities of

Figure 10. CV responses at HMDE (A) and SWV responses at PGE (B) obtained for PEX product synthesized with temp3A3C3T template and com- X binations of BF and/or PheNO2 dN TP with unmodified dNTPs, as indicated by nucleobase symbols in the legend (valid for both panels). In (A), switching potential À1.3 V; for other details, see the Experimental Section. incorporated in one reaction. The resulting PEX products contained three BF and/or three PhNO2 labels. Signals in voltammograms measured at the HMDE (Figure 10A) reflected the composition of the PEX reaction mixture (i.e., red red peak BF , peak NO2 , or both peaks were detected when dCBFTP, dANO2TP, or both conjugates were present, respectively). Notably, the relative intensities of the signals corresponded to the number of electrons expected to be involved in the given reduction process per label (6 for BF and 4 for

PhNO2), and no significant mutual interference was ob- red red served. When the PEX products were measured at the Figure 11. A) Areas of AdTS CV peaks BF and NO2 obtained at the PGE, analogous results were obtained with the exception of HMDE for PEX products synthesized with tempxAyCzT templates (see red red Table 2; numbers of A, C, and T residues in the synthesized sequences the ratio of the intensities of peaks BF and NO2 . Both are indicated in the graph) and dANO2TP +dCBFTP+ dTNH2TP +dGTP signals exhibited similar peak heights, suggesting an equal red red mix. B) Ratios of areas of peaks BF /NO2 obtained for the same PEX (rather than different) number of electrons being exchanged products. Numbers in blue indicate expected values calculated from the in the respective reduction processes. A detailed study of number of labels per ON and the number of electrons consumed per the electrode reaction mechanisms is beyond the scope of label. this report and will be published elsewhere. Combination of red red either of the reducible labels (or both) with the PhNH2 label peak BF and peak NO2 obtained for the individual PEX red red showed no effect of the PhNH2 on the BF and NO2 products. It is clear that the intensities of the reducible label peaks (not shown). On the other hand, the signal of the signals varied consistently with the variation in the number one-electron primary oxidation of the amino group was of respective conjugates incorporated as dictated by the rather poorly developed (compared to the reduction signals template nucleotide sequence. When a complementary base of BF and PhNO2) and did not exhibit a good correlation was missing from the template, a negligible signal corre- between the number of PhNH2 tags per ON and the signal sponding to the given tag was observed (probably due to a intensity (see below). small amount of misincorporation). The next set of experiments was performed with PEX Absolute signal intensities obtained for the PEX products templates designed for incorporating the three labeled nu- containing identical numbers of a given labeled nucleotide cleotides in different quantities and ratios (Table 2). All (e.g., three CBFs) varied to a small extent within the groups PEX reactions were conducted with equimolar mixtures of of samples incorporating an identical number of the given

12726 www.chemeurj.org 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Eur. J. 2013, 19, 12720 – 12731 Redox Label for Electrochemical Detection of DNA FULL PAPER label, which can be explained in terms of natural variations trodes, occurring at a potential not overlapping with the re- in the yields of individual ONs after the isolation procedure. duction potentials of natural nucleobases (in DNA, more Such variations can, in principle, be eliminated by normali- negative by 500–600 mV) or previously developed reducible [6] [9] zation of the signal for a unit amount of DNA, which can be labels PhNO2 or anthraquinone (less negative by at least accomplished, for example, by referring to a signal inde- 300 mV). These features (at least in applications the princi- pendent of the modification, such as peak G (provided that ple of which has been proven herein) counterbalance an ap- the guanine content does not vary significantly within a parent drawback of BF as a label giving irreversible electro- given set of samples). Another possible means of eliminat- chemistry, that is, the impossibility of coupling it to an elec- ing the effect of DNA concentration variations is calculation trocatalytic system applicable in, for example, amperometric of the ratio between the intensities of signals yielded by two detectors. Nevertheless, in amperometric devices, it is in independently detectable labels. The latter approach was ap- principle rather difficult to determine two or more redox plied to generate Figure 11B, in which ratios of peak areas species independently. On the other hand, simple voltam- red red BF /NO2 are plotted. For measurements at the HMDE, metric analysis of DNA labeled with combinations of BF, we obtained a reasonable correlation between the ratios of PhNO2, and PhNH2 as an oxidizable label revealed no sig- peak areas (directly proportional to the number of electrons nificant interference between BF and PhNO2 reductions and involved) and the ratios of labels incorporated, considering no effect of PhNH2 on the signals of any of the reducible the number of electrons consumed per reduction of the tags. The quantities of BF and PhNO2 labels incorporated given moiety (expected values are indicated by the numbers into a nucleotide sequence could be determined independ- in blue for each sample in Figure 11B). When the same ently, and the relative intensities of their signals exhibited measurements were performed with the PGE, similar results excellent correlation with the number of complementary red red were obtained, but the peak area ratios BF /NO2 were bases in the template, making them applicable for ratiomet- systematically lower (red columns in Figure 11B) than those ric analysis of nucleotide sequences (such as electrochemical resulting from measurements at the HMDE, corresponding detection of mutations in a DNA stretch based on a change to approximately equal numbers of electrons being con- in the ratio of two nucleobases encoded by two different sumed for reduction of both the BF and PhNO2 labels. redox labels). On the other hand, PhNH2 is suitable for Almost identical results were obtained when the PEX reac- qualitative but not (semi)quantitative ratiometric electro- tions were performed with the dANH2TP+dCBFTP + chemical probing of nucleotide sequences (at least when NO2 dT TP+unlabeled dGTP mixture (see Figure S34 in the combined with BF and/or PhNO2). Even though the current- Supporting Information). ly available palette of redox labels is not ready for full DNA In contrast to the combination of two reducible labels, sequencing (i.e., reading of unknown nucleotide sequences), our attempts to use oxidizable PhNH2 in combination with its applicability in single-nucleotide polymorphism (SNP) either BF or PhNO2 tags in analogous measurements did sensing has been demonstrated and extended here towards not result in a good correlation between the relative number analysis of expected base variations in short DNA stretches of tags incorporated into the DNA and the ratio of the re- of known sequence. Our ongoing research is focused on spective signals (for an example, see Figure S35). Although seeking further redox labels or their combinations to devel- the amino group can still be used as a qualitative marker for op a generally applicable and orthogonal “four-color” the presence of a PhNH2-encoded nucleobase in the synthe- coding scheme for electrochemical DNA labeling, which sized ON stretch, its utilization as a ratiometric label was could be applied in the detection of multiple variants of mu- found to be limited. Above and beyond difficulties resulting tations of certain important genes (e.g., the KRAS gene, in from the relatively poorly developed signal of the amino which several mutations in two consecutive codons are asso- group oxidation, another limitation may arise from the fact ciated with colorectal cancer[24]) in one simple experiment, that amino derivatives are frequent products of the reduc- thereby avoiding the expensive instrumentation and exten- tion of a variety of other organic nitrogenous compounds, sive computation needed for full sequencing. Other possible and thus these moieties cannot be used as fully independent applications of orthogonal, independently detectable labels labels in combination with PhNH2. include one-step detection (and typing) of allele-specific mutants (discrimination between homozygous wild types, heterozygotes, and homozygous mutants), determination of Conclusion the ratio between wild types and mutants in multicopy genes, and determination of the number of copies of a repet- We have proposed the use of benzofurazane as a novel re- itive sequence per DNA fragment (e.g., in diagnostics of ducible label for DNA. Adenine and cytosine dNXBFTP con- neurodegenerative disorders associated with triplet repeat jugates have been prepared and successfully tested as sub- expansions).[25] strates for DNA polymerases, and have been reliably and precisely incorporated into different DNA sequences by primer extension. Multi-electron electrochemical reduction of the furazane ring gave rise to an intense cathodic signal that was easily measurable at both mercury and carbon elec-

Chem. Eur. J. 2013, 19, 12720 – 12731 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.chemeurj.org 12727 M. Fojta, M. Hocek et al.

Experimental Section was isolated from the crude reaction mixture by HPLC on a C18 column, eluting with a linear gradient from 0.1m triethylammonium bicarbonate (TEAB) in H O to 0.1 m TEAB in H O/MeOH (1:1). Several co-distilla- General: NMR spectra were measured at 500 MHz for 1H and 2 2 tions with water and conversion to the sodium salt form (Dowex 50WX8 125.7 MHz for 13C, or at 600 MHz for 1H and 150.9 MHz for 13C, from in Na+ cycle) followed by freeze-drying gave a solid product. solutions in D2O (referenced to dioxane as an internal standard, dH = Sonogashira cross-coupling: Method E: A 2:1 mixture of H2O/CH3CN 3.75 ppm, dC =69.3 ppm) or in [D6]DMSO (referenced to the residual solvent signal). Chemical shifts are given in ppm (d scale) and coupling con- (2 mL) was added through a septum to an argon-purged flask containing I stants (J) in Hz. Complete assignment of all NMR signals was achieved halogenated nucleotides dN TP (0.085 mmol, 1 equiv), 2 (19 mg, by using a combination of H,H-COSY, H,C-HSQC, and H,C-HMBC ex- 0.13 mmol, 1.5 equiv), CuI (2 mg, 0.008 mmol, 10 mol%), and (iPr)2EtN ACHTUNGRE periments. Mass spectra were measured on an LCQ Classic (Thermo-Fin- (0.13 mL, 0.85 mmol, 10 equiv). In a separate flask, Pd(OAc)2 (1 mg, nigan) spectrometer using ESI or a Q-TOF Micro (Waters, ESI source, 0.004 mmol, 5 mol %) and TPPTS (6 mg, 0.01 mmol, 2.5 equiv with re- internal calibration with lockspray). Semi-preparative separation of nu- spect to Pd) were combined, the flask was evacuated and purged with cleoside triphosphates was performed by HPLC on a column packed argon, and then a 2:1 mixture of H2O/CH3CN (0.5 mL) was added. This with reversed-phase 10 mm C18 (Phenomenex, Luna C18 (2)). IR spectra catalyst solution was injected into the reaction mixture, which was then were measured by using the attenuated total reflectance (ATR) techni- stirred at 75 8C for 45 min until complete consumption of the starting ma- que or by using KBr pellets. High-resolution mass spectra were measured terial. The product was isolated from the crude reaction mixture by m under ESI conditions. Mass spectra of functionalized DNA were meas- HPLC on a C18 column, eluting with a linear gradient from 0.1 TEAB m ured by MALDI-TOF spectrometry on a Reflex IV spectrometer in H2O to 0.1 TEAB in H2O/MeOH (1:1). Several co-distillations with + (Bruker) with a nitrogen laser. Melting points were determined on a water and conversion to the sodium salt form (Dowex 50WX8 in Na Kofler block. Known starting compounds were prepared according to lit- cycle) followed by freeze-drying gave a solid product. erature procedures (EBF,[22] dANO2TP,[6] dTNH2TP[6]). Triphosphorylation: Method F: Dry trimethyl phosphate (0.32 mL) was BF Synthesis of modified nucleosides: added to an argon-purged flask containing a nucleoside analogue dN (0.17 mmol, 1 equiv) cooled to 08C on ice and then POCl (20 mL, Suzuki–Miyaura cross-coupling: Method A: (dCBF,dABF): A 2:1 mixture 3 0.2 mmol, 1.2 equiv) was added. After 1–3 h, a solution of of H O/CH CN (2 mL) was added through a septum to an argon-purged 2 3 (NHBu ) H P O (480 mg, 0.9 mmol, 5 equiv) and Bu N (0.17 mL, flask containing halogenated nucleosides dNI (0.085 mmol, 1 equiv), bor- 3 2 2 2 7 3 0.7 mmol, 4.2 equiv) in dry DMF (1.3 mL) was added to the reaction mix- onic acid 1 (17 mg, 0.11 mmol, 1.2 equiv), and Cs CO (83 mg, 0.25 mmol, 2 3 ture, which was then stirred for a further 1.5 h and quenched with 2 m 3 equiv). In a separate flask, Pd(OAc)ACHTUNGRE (1 mg, 0.004 mmol, 5 mol %), and 2 TEAB buffer (1 mL). The product was isolated from the crude reaction TPPTS (6 mg, 0.011 mmol, 2.5 equiv with respect to Pd) were combined, mixture by HPLC on a C18 column, eluting with a linear gradient from the flask was evacuated and purged with argon, and then a 2:1 mixture 0.1m TEAB in H O to 0.1m TEAB in H O/MeOH (1:1). Several co-dis- of H O/CH CN (0.5 mL) was added. This catalyst solution was injected 2 2 2 3 tillations with water and conversion to the sodium salt form (Dowex into the reaction mixture, which was then stirred at 75 8C for 1–2 h until 50WX8 in Na+ cycle) followed by freeze-drying gave a solid product. complete consumption of the starting material and then concentrated in BF BF I vacuo. The products were purified by column chromatography on silica dC : Compound dC was prepared from dC according to a general gel eluting with chloroform/methanol (0 to 10%). procedure (Method A). The product was isolated as a white solid (20 mg, 69%). M.p.>300 8C; 1H NMR (600.1 MHz, [D ]DMSO): d =2.12 (ddd, Sonogashira cross-coupling: Method B: (dCEBF,dAEBF): Dry DMF (3 mL) 6 1H, J =13.4, J = 6.6, J = 6.0 Hz; H-2’b), 2.18 (ddd, 1 H, J = was added to an argon-purged flask containing 2 (15 mg, 0.10 mmol, gem 2’b,1’ 2’b,3’ gem 13.4, J =6.2, J =4.0 Hz; H-2’a), 3.52, 3.59 (2 ddd, 2 1H, J = 1.2 equiv), nucleoside analogue dNI (0.08 mmol, 1 equiv), CuI (2 mg, 2’a,1’ 2’a,3’ gem 11.9, J =5.1, J =3.5 Hz; H-5’), 3.78 (q, 1 H, J =J =3.5 Hz; H- 0.008 mmol, 10 mol%), and [Pd(PPhACHTUNGRE ) Cl ] (3 mg, 0.004 mmol, 5 mol%) 5’,OH 5’,4’ 4’,3’ 4’,5’ 3 2 2 4’), 4.24 (m, 1H, J = 6.0, 4.0, J =4.3, J =3.5 Hz; H-3’), 4.98 (t, followed by (iPr) EtN (0.14 mL, 0.84 mmol, 10 equiv). The reaction mix- 3’,2’ 3’,OH 3’,4’ 2 1H, J =5.1 Hz; OH-5’), 5.20 (d, 1 H, J =4.3 Hz; OH-3’), 6.20 (dd, ture was stirred at 75 8C for 1–2 h until complete consumption of the OH,5’ OH,3’ 1H, J =6.6, 6.2 Hz; H-1’), 6.79, 7.49 (2 brs, 21 H; NH ), 7.53 (dd, starting material and then concentrated in vacuo. The products were pu- 1’,2’ 2 1H, J =9.3, J =1.4 Hz; H-6-benzooxadiazole), 7.97 (dd, 1H, J =1.4, rified by column chromatography on silica gel eluting with chloroform/ 6,7 6,4 4,6 J = 1.1 Hz; H-4-benzooxadiazole), 8.07 (dd, 1H, J =9.3, J =1.1 Hz; methanol (0 to 10%). 4,7 7,6 7,4 H-7-benzooxadiazole), 8.11 ppm (s, 1 H; H-6); 13C NMR (150.9 MHz, Sonogashira cross-coupling: Method C: (dCEBF,dAEBF): A 2:1 mixture of [D6]DMSO): d=40.82 (CH2-2’), 61.00 (CH2-5’), 70.04 (CH-3’), 85.39 H2O/CH3CN (2 mL) was added through a septum to an argon-purged (CH-1’), 87.49 (CH-4’), 106.09 (C-5), 115.16 (CH-4-benzooxadiazole), I flask containing halogenated nucleosides dN (0.056 mmol, 1 equiv), 2 116.44 (CH-7-benzooxadiazole), 135.25 (CH-6-benzooxadiazole), 138.47 (10 mg, 0.067 mmol, 1.2 equiv), CuI (1 mg, 0.005 mmol, 10 mol%), and (C-5-benzooxadiazole), 141.49 (CH-6), 148.53 (C-7a-benzooxadiazole), ACHTUNGRE (iPr)2EtN (0.1 mL, 0.56 mmol, 10 equiv). In a separate flask, Pd(OAc)2 149.54 (C-3a-benzooxadiazole), 154.43 (C-2), 163.12 ppm (C-4); IR (1 mg, 0.003 mmol, 5 mol%) and TPPTS (4 mg, 0.007 mmol, 2.5 equiv (KBr): n˜ =3466, 1634, 1603, 1539, 1365, 1265, 1094, 1067, 956, 1482, 1183, with respect to Pd) were combined, the flask was evacuated and purged 782 cmÀ1; MS (ESI +): m/z (%): 368.3 (100) [M++Na]; HRMS (ESI): m/ with argon, and then a 2:1 mixture of H2O/CH3CN (0.5 mL) was added. z calcd for C15H16N5O5: 346.1146; found: 346.11458. This catalyst solution was injected into the reaction mixture, which was dABF : Compound dABF was prepared from dAI according to a general then stirred at 75 8C for 1–2 h until complete consumption of the starting procedure (Method A). The product was isolated as a yellow solid material and then concentrated in vacuo. The products were purified by (23 mg, 74%). M.p. 2188C; 1H NMR (500.0 MHz, [D ]DMSO): d =2.24 column chromatography on silica gel eluting with chloroform/methanol 6 (ddd, 1H, J = 13.1, J =6.0, J =2.8 Hz; H-2’b), 2.58 (ddd, 1H, (0 to 10%). gem 2’b,1’ 2’b,3’ Jgem =13.1, J2’a,1’ =8.1, J2’a,3’ =5.9 Hz; H-2’a), 3.53 (ddd, 1 H, Jgem =11.7, Synthesis of modified nucleoside triphosphates: J5’b,OH =6.0, J5’b,4’ =4.4 Hz; H-5’b), 3.60 (ddd, 1 H, Jgem =11.7, J5’a,OH =5.3,

Suzuki–Miyaura cross-coupling: Method D: A 2:1 mixture of H2O/ J5’a,4’ =4.8 Hz; H-5’a), 3.85 (ddd, 1H, J4’,5’ =4.8, 4.4, J4’,3’ =2.6 Hz; H-4’),

CH3CN (2 mL) was added through a septum to an argon-purged flask 4.38 (m, 1H, J3’,2’ =5.9, 2.8, J3’,OH = 4.1, J3’,4’ =2.6 Hz; H-3’), 5.04 (dd, 1H, I containing halogenated nucleotides dN TP (30 mg, 0.04 mmol, 1 equiv), JOH,5’ =6.0, 5.3 Hz; OH-5’), 5.28 (d, 1H, JOH,3’ =4.1 Hz; OH-3’), 6.57 (brs, boronic acid 1 (10 mg, 0.06 mmol, 1.5 equiv), and Cs2CO3 (43 mg, 2H; NH2), 6.61 (dd, 1 H, J1’,2’ = 8.1, 6.0 Hz; H-1’), 7.77 (dd, 1H, J6,7 =9.3, ACHTUNGRE 0.13 mmol, 3 equiv). In a separate flask, Pd(OAc)2 (0.5 mg, 0.002 mmol, J6,4 = 1.4 Hz; H-6-benzooxadiazole), 7.86 (dd, 1H, J4,6 =1.4, J4,7 =1.0 Hz;

5 mol%) and TPPTS (3 mg, 0.005 mmol, 2.5 equiv with respect to Pd) H-4-benzooxadiazole), 7.87 (s, 1H; H-6), 8.10 (dd, 1H, J7,6 =9.3, J7,4 = were combined, the flask was evacuated and purged with argon, and then 1.0 Hz; H-7-benzooxadiazole), 8.19 ppm (s, 1H; H-2); 13C NMR a 2:1 mixture of H2O/CH3CN (0.5 mL) was added. This catalyst solution (125.7 MHz, [D6]DMSO): d=39.70 (CH2-2’), 62.10 (CH2-5’), 71.14 (CH- was injected into the reaction mixture, which was then stirred at 758C for 3’), 83.21 (CH-1’), 87.66 (CH-4’), 100.08 (C-4a), 112.44 (CH-4-benzooxa- 45 min until complete consumption of the starting material. The product diazole), 115.04 (C-5), 116.22 (CH-7-benzooxadiazole), 123.12 (CH-6),

12728 www.chemeurj.org 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Eur. J. 2013, 19, 12720 – 12731 Redox Label for Electrochemical Detection of DNA FULL PAPER

135.28 (CH-6-benzooxadiazole), 138.05 (C-5-benzooxadiazole), 148.53 HRMS (ESIÀ): m/z calcd for C15H15N5Na2O14P3 : 627.96183; found: (C-7a-benzooxadiazole), 149.78 (C-3a-benzooxadiazole), 151.34 (C-7a), 627.96335. 152.33 (CH-2), 157.66 ppm (C-4); IR (KBr): n˜ =3328, 3218, 3187, 1629, dABFTP: Compound dABFTP was prepared from dAITP in 22% yield ac- À1 1585, 1543, 1469, 1449, 1206, 956, 795 cm ; MS (ESI +): m/z (%): 369.2 cording to Method D or from dABF in 70 % yield according to Method F. + + (10) [M ], 391.2 (100) [M +Na]; HRMS (ESI): m/z calcd for 1 H NMR (499.8 MHz, D2O): d=2.52 (ddd, 1 H, Jgem = 14.0, J2’b,1’ =6.4, C17H16N6NaO4: 391.11252; found: 391.11254. J2’b,3’ =3.4 Hz; H-2’b), 2.78 (ddd, 1 H, Jgem = 14.0, J2’a,1’ =7.6, J2’a,3’ =6.6 Hz; EBF EBF I dC : Compound dC was prepared from dC in 60% yield according H-2’a), 4.13 (ddd, 1 H, Jgem =10.9, JH,P =5.2, J5’b,4’ =4.1 Hz; H-5’b), 4.22 to Method B and in 45% yield according to Method C. The product was (dt, 1H, Jgem = 10.9, JH,P =J5’a,4’ = 5.0 Hz; H-5’a), 4.25 (m, 1H; H-4’), 4.80 1 isolated as a brown solid. M.p.>300 8C; H NMR (600.1 MHz, (overlapped with HDO; H-3’), 6.72 (dd, 1 H, J1’,2’ =7.6, 6.4 Hz; H-1’), 7.82

[D6]DMSO): d= 2.05 (ddd, 1H, Jgem =13.3, J2’b,1’ = 6.6, J2’b,3’ =6.1 Hz; H- (dd, 1 H, J6,7 =9.3, J6,4 =1.2 Hz; H-6-benzooxadiazole), 7.83 (s, 1H; H-6),

2’b), 2.21 (ddd, 1H, Jgem =13.3, J2’a,1’ =6.1, J2’a,3’ =4.1 Hz; H-2’a), 3.61 7.91 (brs, 1 H; H-4-benzooxadiazole), 7.99 (d, 1 H, J7,6 =9.3 Hz; H-7-ben- 13 (ddd, 1H, Jgem =11.8, J5’b,OH =4.7, J5’b,4’ =3.5 Hz; H-5’b), 3.69 (ddd, 1 H, zooxadiazole), 8.23 ppm (s, 1 H; H-2); C NMR (125.7 MHz, D2O): d======Jgem 11.8, J5’a,OH 5.0, J5’a,4’ 3.5 Hz; H-5’a), 3.83 (q, 1 H, J4’,3’ J4’,5’ 41.23 (CH2-2’), 68.14 (d, JC,P =5.5 Hz; CH2-5’), 73.81 (CH-3’), 85.70 (CH- ’ J = J = J = ’ 3.5 Hz; H-4 ), 4.24 (m, 1 H, 3’,2’ 6.1, 4.1, 3’,OH 4.3, 3’,4’ 3.5 Hz; H-3 ), 1’), 88.08 (d, JC,P =8.9 Hz; CH-4’), 103.42 (C-4a), 115.77 (CH-4-benzooxa- 5.20 (dd, 1H, JOH,5’ = 5.0, 4.7 Hz; OH-5’), 5.26 (d, 1H, JOH,3’ =4.3 Hz; OH- diazole), 119.19 (CH-7-benzooxadiazole), 119.21 (C-5), 125.16 (CH-6), 3’), 6.11 (dd, 1H, J1’,2’ = 6.6, 6.1 Hz; H-1’), 7.27 (brs, 1 H; NHaHb), 7.74 137.79 (CH-6-benzooxadiazole), 140.67 (C-5-benzooxadiazole), 151.23 (dd, 1H, J6,7 =9.3, J6,4 =1.3 Hz; H-6-benzooxadiazole), 7.93 (brs, 1 H; (C-7a-benzooxadiazole), 152.22 (C-3a-benzooxadiazole), 153.27 (C-7a), NH H ), 8.11 (dd, 1H, J =9.3, J =1.1 Hz; H-7-benzooxadiazole), 8.32 31 1 a b 7,6 7,4 154.65 (CH-2), 160.10 ppm (C-4); P{ H} NMR (202.3 MHz, D2O): d= (dd, 1 H, J =1.3, J =1.1 Hz; H-4-benzooxadiazole), 8.51 ppm (s, 1H; 4,6 4,7 À21.57 (brt, J=20.0 Hz; Pb), À10.43 (d, J= 20.0 Hz; Pa), À6.19 ppm H-6); 13C NMR (150.9 MHz, [D ]DMSO): d=41.16 (CH -2’), 60.92 (CH - 6 2 2 (brd, J= 20.0 Hz; Pg); MS (ESIÀ): m/z (%): 549.2 (100) 5’), 69.91 (CH-3’), 85.86 (CH-1’), 87.41 (pyrimidine-CC-benzoxadia- À À [MÀHÀH2PO3+Na] , 629.2 (20) [MÀ2H+Na] ; HRMS (ESIÀ): m/z zole), 87.69 (CH-4’), 88.57 (C-5), 93.08 (pyrimidine-CC-benzoxadia- calcd for C17H17N6NaO13P3 : 628.99696; found: 628.99637. zole), 116.59 (CH-7-benzooxadiazole), 117.78 (CH-4-benzooxadiazole), dCEBFTP: Compound dCEBFTP was prepared from dCITP in 52% yield 126.73 (C-5-benzooxadiazole), 135.01 (CH-6-benzooxadiazole), 146.75 according to Method E. 1H NMR (600.1 MHz, D O): d=2.35 (ddd, 1H, (CH-6), 148.11 (C-7a-benzooxadiazole), 148.96 (C-3a-benzooxadiazole), 2 J =14.2, J =6.7, J = 6.4 Hz; H-2’b), 2.50 (ddd, 1H, J =14.2, 153.44 (C-2), 163.84 ppm (C-4); IR (KBr): n˜ =3389, 2201, 1660, 1618, gem 2’b,1’ 2’b,3’ gem J =6.4, J =4.3 Hz; H-2’a), 4.23–4.31 (m, 3H; H-4’,5’), 4.65 (dt, 1 H, 1536, 1493, 1333, 1267, 1090, 1048, 780 cmÀ1; MS (ESI +): m/z (%): 392.3 2’a,1’ 2’a,3’ J =6.4, 4.3, J =4.3 Hz; H-3’), 6.25 (dd, 1H, J = 6.7, 6.4 Hz; H-1’), (100) [M+Na+]; HRMS (ESI +): m/z calcd for C H N O : 370.1146; 3’,2’ 3’,4’ 1’,2’ 17 16 5 5 7.68 (d, 1 H, J =9.3 Hz; H-6-benzooxadiazole), 7.90 (d, 1H, J = found: 370.1146. 6,7 7,6 9.3 Hz; H-7-benzooxadiazole), 8.13 (s, 1H; H-4-benzooxadiazole), EBF EBF I dA : Compound dA was prepared from dA in 70% yield according 13 8.30 ppm (s, 1 H; H-6); C NMR (150.9 MHz, D2O): d=42.23 (CH2-2’), to Method B and in 28% yield according to Method C. The product was 67.81 (d, JC,P = 5.5 Hz; CH2-5’), 72.87 (CH-3’), 87.01 (pyrimidine-CC- isolated as a yellow solid. M.p. 209 8C; 1H NMR (500.0 MHz, benzooxadiazole), 88.39 (d, JC,P =9.0 Hz; CH-4’), 89.22 (CH-1’), 94.46 (C- d= J = J = J = [D6]DMSO): 2.24 (ddd, 1H, gem 13.2, 2’b,1’ 6.0, 2’b,3’ 2.9 Hz; H- 5), 96.83 (pyrimidine-CC-benzooxadiazole), 118.90 (CH-7-benzooxadia- 2’b), 2.51 (ddd, 1H, J =13.2, J =7.9, J =5.7 Hz; H-2’a), 3.55 gem 2’a,1’ 2’a,3’ zole), 121.19 (CH-4-benzooxadiazole), 129.06 (C-5-benzooxadiazole), (ddd, 1H, J =11.8, J =5.9, J =4.3 Hz; H-5’b), 3.61 (ddd, 1 H, gem 5’b,OH 5’b,4’ 137.63 (CH-6-benzooxadiazole), 148.37 (CH-6), 150.94 (C-7a-benzooxa- J =11.8, J =5.3, J =4.5 Hz; H-5’a), 3.86 (ddd, 1 H, J =4.5, 4.3, gem 5’a,OH 5’a,4’ 4’,5’ diazole), 151.63 (C-3a-benzooxadiazole), 158.52 (C-2), 167.16 ppm (C-4); J4’,3’ =2.6 Hz; H-4’), 4.37 (m, 1 H, J3’,2’ =5.7, 2.9, J3’,OH = 4.1, J3’,4’ =2.6 Hz; 31 1 P{ H} NMR (202.3 MHz, D2O): d =À21.92 (t, J =20.0 Hz; Pb), À10.69 H-3’), 5.08 (dd, 1H, JOH,5’ = 5.9, 5.3 Hz; OH-5’), 5.29 (d, 1H, JOH,3’ = (d, J =20.0 Hz; Pa), À7.77 ppm (brd, J =20.0 Hz; Pg); MS (ESIÀ): m/z 4.1 Hz; OH-3’), 6.53 (dd, 1 H, J1’,2’ =7.9, 6.0 Hz; H-1’), 6.85 (brs 2H; À À (%): 448 (100) [MÀH3P2O6] , 528 (90) [MÀH2PO3] , 550.2 (55) NH2), 7.72 (dd, 1 H, J6,7 =9.3, J6,4 =1.3 Hz; H-6-benzooxadiazole), 8.02 (s, À À [MÀHÀH2PO3+Na] , 608.2 (10) [MÀH] ; HRMS (ESIÀ): m/z calcd for 1H; H-6), 8.11 (dd, 1H, J7,6 =9.3, J7,4 =1.1 Hz; H-7-benzooxadiazole), C17H17N5O14P3 : 607.99903; found: 607.99849. 8.17 (s, 1H; H-2), 8.38 ppm (dd, 1H, J4,6 =1.3, J4,7 =1.0 Hz; H-4-benzoox- EBF EBF I 13 dA TP: Compound dA TP was prepared from dA TP in 54% yield adiazole); C NMR (125.7 MHz, [D6]DMSO): d =40.16 (CH2-2’), 62.00 1 according to Method E. H NMR (600.1 MHz, D2O): d=2.52 (ddd, 1H, (CH2-5’), 71.09 (CH-3’), 83.53 (CH-1’), 87.82 (CH-4’), 88.46 (-CC-benzooxadiazole), 90.38 (-CC-benzooxadiazole), 94.04 (C-5), 101.95 (C-4a), Jgem =13.9, J2’b,1’ =6.0, J2’b,3’ = 3.3 Hz; H-2’b), 2.65 (ddd, 1H, Jgem =13.9, 116.56 (CH-7-benzooxadiazole), 117.94 (CH-4-benzooxadiazole), 126.79 J2’a,1’ =8.0, J2’a,3’ =6.0 Hz; H-2’a), 4.15 (dt, 1 H, Jgem = 11.2, JH,P =J5’b,4’ = (C-5-benzooxadiazole), 128.75 (CH-6), 135.18 (CH-6-benzooxadiazole), 4.4 Hz; H-5’b), 4.22 (dt, 1H, Jgem =11.2, JH,P =J5’a,4’ =4.9 Hz; H-5’a), 4.24 148.06 (C-7a-benzooxadiazole), 149.01 (C-3a-benzooxadiazole), 149.88 (m, 1H; H-4’), 4.78 (overlapped with HDO; H-3’), 6.30 (dd, 1 H, J1’,2’ = (C-7a), 153.14 (CH-2), 157.68 (C-4); IR (KBr): n˜ =3458, 3419, 3345, 8.0, 6.0 Hz; H-1’), 7.36 (d, 1H, J6,7 =9.2 Hz; H-6-benzooxadiazole), 7.55 À1 3304, 3112, 2197, 1650, 1623, 1591, 1527, 1467, 1083, 1054, 793, 634 cm ; (d, 1 H, J7,6 =9.2 Hz; H-7-benzooxadiazole), 7.70 (s, 1H; H-4-benzooxa- 13 MS (ESI +): m/z (%): 393.2 (20) [M+], 415.2 (100) [M+Na+]; HRMS diazole), 7.75 (s, 1 H; H-6), 7.83 ppm (s, 1H; H-2); C NMR (150.9 MHz, D O): d= 41.56 (CH -2’), 68.25 (d, J =5.9 Hz; CH -5’), 73.83 (CH-3’), (ESI +): m/z calcd for C19H17N6O4: 393.13058; found: 393.13047. 2 2 C,P 2 85.74 (CH-1’), 88.02 (d, J =8.9 Hz; CH-4’), 89.75 (-CC-benzooxadia- dCBFTP: Compound dCBFTP was prepared from dCITP in 10% yield ac- C,P zole), 94.12 (-CC-benzooxadiazole), 98.70 (C-5), 105.11 (C-4a), 118.17 cording to Method D or from dCBF in 24% yield according to Method F. 1 (CH-7-benzooxadiazole), 119.49 (CH-4-benzooxadiazole), 129.16 (C-5- H NMR (499.8 MHz, D2O): d=2.39 (dt, 1 H, Jgem =14.1, J2’b,1’ =J2’b,3’ = benzooxadiazole), 130.33 (CH-6), 137.12 (CH-6-benzooxadiazole), 150.16 6.8 Hz; H-2’b), 2.475 (ddd, 1 H, Jgem = 14.1, J2’a,1’ =6.4, J2’a,3’ =3.8 Hz; H- 2’a), 4.14–4.24 (m, 3 H; H-4’,5’), 4.65 (dt, 1 H, J =6.8, 3.8, J =3.8 Hz; (C-7a-benzooxadiazole), 150.79 (C-7a), 150.98 (C-3a-benzooxadiazole), 3’,2’ 3’,4’ 31 1 154.50 (CH-2), 159.54 ppm (C-4); P{ H} NMR (202.3 MHz, D2O): d= H-3’), 6.34 (dd, 1H, J1’,2’ = 6.8, 6.4 Hz; H-1’), 7.58 (dd, 1H, J6,7 =9.5, J6,4 = 1.2 Hz; H-6-benzooxadiazole), 7.99–8.04 ppm (m, 3 H; H-6, H-4,7-ben- À21.90 (t, J= 18.3 Hz; Pb), À10.50 (d, J =18.3 Hz; Pa), À8.10 ppm (d, J= À zooxadiazole); 13C NMR (125.7 MHz, D O): d =42.00 (CH -2’), 67.78 (d, 18.3 Hz; Pg); MS (ESIÀ): m/z (%): 551.2 (95) [MÀH2PO3] , 573.2 (100) 2 2 À À À [MÀHÀH2PO3+Na] , 631.2 (18) [MÀH] , 653.2 (30) [MÀ2H+Na] , JC,P =5.4 Hz; CH2-5’), 73.01 (CH-3’), 88.47 (d, JC,P = 8.9 Hz; CH-4’), 88.99 À (CH-1’), 111.36 (C-5), 119.16 (CH-4-benzooxadiazole), 119.82 (CH-7- 675.2 (28) [MÀ3H+2Na] ; HRMS (ESIÀ): m/z calcd for C19H18N6O13P3 : benzooxadiazole), 137.11 (CH-6-benzooxadiazole), 139.54 (C-5-benzoox- 631.01502; found: 631.01445. adiazole), 143.77 (CH-6), 151.35 (C-7a-benzooxadiazole), 152.04 (C-3a- Primer extension experiment: The reaction mixture (20 mL) contained benzooxadiazole), 159.68 (C-2), 166.71 ppm (C-4); 31P{1H} NMR DNA polymerase (KOD XL 0.1–0.2 U, Pwo 0.5 U, Deep Vent 0.2 U, m (202.3 MHz, D2O): d=À21.28 (t, J= 20.0 Hz; Pb), À10.72 (d, J =20.0 Hz; Deep Vent (exo-) 0.2 U, Vent (exo-) 0.1 U), primer (0.15 m ), template m m Pa), À5.33 ppm (d, J= 20.0 Hz; Pg); MS (ESIÀ): m/z (%): 504.0 (95) (0.23 m ), and both natural and modified dNTPs (0.2 m ) in reaction À À À 32 [MÀH2PO3] , 606 (25) [MÀ2H+Na] , 628.0 (20) [MÀ3H+2Na] ; buffer. The primer was labeled with [g P]ATP according to standard

Chem. Eur. J. 2013, 19, 12720 – 12731 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.chemeurj.org 12729 M. Fojta, M. Hocek et al.

techniques. Reaction mixtures were incubated for 15–40 min at 60 8C and Acknowledgements then analyzed by PAGE electrophoresis. Kinetics of PEX: PEX reaction mixtures containing Pwo DNA polymer- This work was supported by the Academy of Sciences of the Czech Re- ase were incubated for time intervals of 0.1–10 min. The reactions were public (RVO 61388963 and institutional research plan AV0Z50040702), then stopped by the addition of PAGE loading buffer and the mixtures the Grant Agency of the Academy of Sciences of the Czech Republic were immediately heated. (IAA400040901) to L.H. and M.F., and by the Czech Science Foundation Polyacrylamide gel electrophoresis: The PEX products were mixed with (P206/12/G151) to J.B., M.P., M.H., and M.F. loading buffer (80% formamide, 10 mm ethylenediaminetetraacetic acid (EDTA), 1 mgmLÀ1 xylene cyanol, 1 mgmLÀ1 bromophenol blue) and subjected to electrophoresis in 12% denaturing polyacrylamide gel con- [1] a) M. J. Heller, Annu. Rev. Biomed. Eng. 2002, 4, 129 –153; b) A. taining 1TBE [TRIS (2-amino-2-hydroxymethylpropane-1,3-diol)/ Sassolas, B. D. Leca-Bouvier, L. J. Blum, Chem. Rev. 2008, 108, 109– borate/EDTA] buffer (pH 8) and 7 m urea at 25 W for 50 min. Gels were 139. dried, autoradiographed, and visualized by using a phosphorimager. [2] a) E. Palecˇek, F. Jelen in Electrochemistry of Nucleic Acids and Pro- Melting temperatures: The oligonucleotides for these measurements teins: Towards Electrochemical Sensors for Genomics and Proteo- were prepared by PEX on a large scale by using Pwo as polymerase, tem- mics (Eds.: E. Palecˇek, F. Scheller, J. Wang), Elsevier, Amsterdam, plate temprnd16, and primrnd as primer. For preparative purposes, a total 2005, pp. 74– 174; b) J. Wang in Electrochemistry of Nucleic Acids volume of 500 mL PEX containing higher concentrations of primer and Proteins: Towards Electrochemical Sensors for Genomics and (10 mm) and template (10 mm) was run and purification was carried out Proteomics (Eds.: E. Palecˇek, F. Scheller, J. Wang), Elsevier, Am- with a QIAquick Nucleotide Removal Kit (Qiagen). Samples were sterdam, 2005, pp. 175 –194. [3] a) D. R. Bentley, S. Balasubramanian, H. P. Swerdlow, Nature 2008, eluted with 100 mLH2O (pH 7.5) and then freeze-dried. DNA duplexes were first dissolved in 160 mL of phosphate buffer (50 mm, pH 6.7) and 456, 53 –59; b) D. A. Wheeler, M. Srinivasan, M. Egholm, Y. Shen, further diluted with the buffer to obtain an optimum concentration with L. Chen, A. McGuire, W. He, Y.-J. Chen, V. Makhijani, G. T. Roth, X. Gomes, K. Tartaro, F. Niazi, C. L. Turcotte, G. P. Irzyk, J. R. aOD260 value between 0.08 and 0.1. Thermal denaturation studies were performed on a Cary 100 Bio UV/Vis spectrophotometer equipped with Lupski, C. Chinault, X.-z. Song, Y. Liu, Y. Yuan, L. Nazareth, X. a temperature controller (Varian). Data were obtained from six individu- Qin, D. M. Muzny, M. Margulies, G. M. Weinstock, R. A. Gibbs, J. M. Rothberg, Nature 2008, 452, 872– 876. al cooling–heating cycles. Melting temperatures (Tm values in 8C) were obtained by plotting temperature versus absorbance and applying a sig- [4] a) M. Hocek, M. Fojta, Chem. Soc. Rev. 2011, 40, 5802–5814; b) E. moidal curve fit. Palecˇek, M. Bartosik, Chem. Rev. 2012, 112, 3427 –3481. [5] a) P. Brzdilov, M. Vrbel, R. Pohl, H. Pivonˇ kov, L. Havran, M. Isolation of single-strand oligonucleotides by magnetoseparation: The re- Hocek, M. Fojta, Chem. Eur. J. 2007, 13, 9527 –9533; b) F. Patolsky, action mixture (100 mL) contained KOD XL DNA polymerase Y. Weizmann, I. Wilner, J. Am. Chem. Soc. 2002, 124, 770 –772. (2.5 UmLÀ1, 1–2 mL), 5’-biotinylated template (100 mm,3mL), primer [6] H. Cahov, L. Havran, P. Brzdilov, H. Pivonˇ kov, R. Pohl, M. (100 mm,3mL), and both natural and modified dNTPs (4 mm,7mL) in re- Fojta, M. Hocek, Angew. Chem. 2008, 120, 2089–2092; Angew. action buffer (40 mL) supplied by the manufacturer. The reaction mixture Chem. Int. Ed. 2008, 47, 2059– 2062. was incubated for 45 min at 608C in a thermal cycler. The reaction was [7] M. Vrbel, P. Horkov, H. Pivonˇ kov, L. Kalachova, H. Cˇ ernock, then stopped by cooling on ice and the mixture was added to streptavidin H. Cahov, R. Pohl, P. Sˇ ebest, L. Havran, M. Hocek, M. Fojta, magnetic particles [Roche; 30 mL of the stock solution was washed with Chem. Eur. J. 2009, 15, 1144– 1154. three 200 mL portions of binding buffer TEN (100mm NaCl, 10mm 100 [8] J. Riedl, P. Horkov, P. Sˇ ebest, R. Pohl, L. Havran, M. Fojta, M. TRIS, 1 mm EDTA, pH 7.5)]. The suspension was shaken at 15 8C for Hocek, Eur. J. Org. Chem. 2009, 3519 –3525. 15 min and 1200 rpm to allow the oligonucleotides to bind to the beads. [9] J. Balintov, R. Pohl, P. Horkov, P. Vidlkov, L. Havran, M. After the incubation, the magnetic beads were collected on a magnet Fojta, M. Hocek, Chem. Eur. J. 2011, 17, 14063– 14073. (Dynal, Invitrogen) and the solution was discarded. The beads were [10] H. Maccˇkov-Cahov, R. Pohl, P. Horkov, L. Havran, J. Sˇ pacˇek, washed with three 200 mL portions of wash buffer TEN (1m NaCl, 1000 M. Fojta, M. Hocek, Chem. Eur. J. 2011, 17, 5833–5841. 10 mm TRIS, 1 mm EDTA, pH 7.5) and then with three 200 mL portions [11] V. Raindlov, R. Pohl, B. Klepetrˇov, L. Havran, E. Sˇimkov, P. of H2O. Water (200 mL) was then added and the sample was denatured Horkov, H. Pivonˇ kov, M. Fojta, M. Hocek, ChemPlusChem 8 for 2 min at 75 C and 900 rpm. The beads were collected on a magnet 2012, 77, 652–662. and the solution was transferred to a clean vial. The product was ana- [12] M. A. Bakht, M. S. Yar, S. G. Abdel-Hamid, S. I. Al Qasoumi, A. lyzed by MALDI-TOF mass spectrometry. Samad, Eur. J. Med. Chem. 2010, 45, 5862 –5869. Electrochemical analysis: Nucleosides, dNTPs, and other building blocks [13] O. Adelfinskaya, W. Wu, V. J. Davisson, D. E. Bergstrom, Nucleo- were analyzed by conventional in situ cyclic voltammetry (CV). PEX sides, Nucleosides Nucleotides Nucleic Acids 2005, 24, 1919– 1945. products were analyzed by ex situ (adsorptive transfer stripping, AdTS) [14] a) M. S. T. GonÅalves, Chem. Rev. 2009, 109, 190–212; b) S. Uchiya- CV or square-wave voltammetry (SWV). The PEX products (purified in ma, T. Santa, K. Imai, Analyst 2000, 125, 1839– 1845; c) J. Rohacova, their single-stranded form by using streptavidin-coated magnetic beads M. L. Marin, A. Martinez-Romero, L. Diaz, J. E. OConnor, M. J. or in their double-stranded forms by using a Qiagen Nucleotide Removal Gomez-Lechon, M. T. Donato, J. V. Castell, M. A. Miranda, Chem- Kit) were accumulated at the surface of a working electrode (hanging MedChem 2009, 4, 466 –472; d) C. Li, E. Henry, N. K. Mani, J. Tang, mercury drop electrode, HMDE, or basal-plane pyrolytic graphite elec- J.-C. Brochon, E. Deprez, J. Xie, Eur. J. Org. Chem. 2010, 2395– trode, PGE) for 60 s from 5 mL aliquots containing 0.2m NaCl. The elec- 2405. trode was then rinsed with deionized water and placed in an electro- [15] B. Zhang, X. Hu, M. Wang, H. Xiao, X. Gong, W. Yang, Y. Cao, chemical cell. CV settings: scan rate 1 V sÀ1, initial potential 0.0 V, for New J. Chem. 2012, 36, 2042– 2047. switching potentials, see figure legends. SWV settings: initial potential [16] P. B. Pati, S. Das, S. S. Zade, J. Polym. Sci. Part A 2012, 50, 3996 – 0 V, for final potentials, see figure legends; frequency 200 Hz, amplitude 4003. 50 mV. Background electrolyte: 0.5m ammonium formate, 0.05m sodium [17] V. Raindlov, R. Pohl, M. Sˇ anda, M. Hocek, Angew. Chem. 2010, phosphate, pH 6.9, or 0.2m sodium acetate, pH 5.0. All measurements 122, 1082 –1084; Angew. Chem. Int. Ed. 2010, 49, 1064– 1066. were performed at room temperature by using an Autolab analyzer (Eco [18] For polymerase incorporations of base-modified dNTPs, see refs. [5– Chemie, The Netherlands) in connection with a VA-stand 663 (Metrohm, 11]. For other examples, see: a) S. Jger, G. Rasched, H. Kornreich- Herisau, Switzerland). A three-electrode system was used, with an Ag/ Leshem, M. Engeser, O. Thum, M. Famulok, J. Am. Chem. Soc. AgCl/3 m KCl electrode as a reference and a platinum wire as an auxili- 2005, 127, 15071 –15082; b) S. Obeid, M. Yulikow, G. Jeschke, A. ary electrode. Reduction signals were measured after deaeration of the Marx, Angew. Chem. 2008, 120, 6886 –6890; Angew. Chem. Int. Ed. solution by argon purging. 2008, 47, 6782–6785; c) C. T. Wirges, J. Timper, M. Fischler, A. S.

12730 www.chemeurj.org 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Eur. J. 2013, 19, 12720 – 12731 Redox Label for Electrochemical Detection of DNA FULL PAPER

Sologubenko, J. Mayer, U. Simon, T. Carell, Angew. Chem. 2009, des 1998, 17, 15– 27; b) G. Cristalli, E. Camaioni, S. Vittori, R. Vol- 121, 225 –229; Angew. Chem. Int. Ed. 2009, 48, 219 –223; d) N. pini, P. A. Borea, A. Conti, S. Dionisotti, E. Ongini, A. Monopoli, J. Ramsay, A.-S. Jemth, A. Brown, N. Crampton, P. Dear, P. Holliger, Med. Chem. 1995, 38, 1462 –1472; c) C. Amatore, E. Blart, J. P. J. Am. Chem. Soc. 2010, 132, 5096– 5104; e) K. Gutsmiedl, D. Fazio, Genet, A. Jutand, S. Lemaire-Audoire, M. Savignac, J. Org. Chem. T. Carell, Chem. Eur. J. 2010, 16, 6877 –6883; f) A. Baccaro, A. 1995, 60, 6829–6839. Marx, Chem. Eur. J. 2010, 16, 218– 226; g) H. Maccˇkov-Cahov, R. [22] J. A. Key, Ch. W. Cairo, Dyes Pigm. 2011, 88, 95 –102. Pohl, M. Hocek, ChemBioChem 2011, 12, 431– 438; h) P. Kielkow- [23] a) N. Staiger, A. Marx, ChemBioChem 2010, 11, 1963– 1966; b) H. ski, H. Maccˇkov-Cahov, R. Pohl, M. Hocek, Angew. Chem. 2011, Sawai, J. Nagashima, M. Kuwahara, R. Kitagata, T. Tamura, I. 123, 8886–8889; Angew. Chem. Int. Ed. 2011, 50, 8727–8730; i) P. Matsui, Chem. Biodiversity 2007, 4, 1979–1995; c) H. Sawai, A. Mnov, M. Hocek, Chem. Commun. 2012, 48, 6921– 6923; j) P. Ozaki-Nakamura, M. Mine, H. Ozaki, Bioconjugate Chem. 2002, 13, Mnov, H. Cahov, M. Plucnara, L. Havran, M. Fojta, M. Hocek, 309– 316. Chem. Commun. 2013, 49, 4652 –4654. [24] X. Liu, M. Jakubowski, J. L. Hunt, Am. J. Clin. Pathol. 2011, 135, [19] a) P. Cˇ apek, R. Pohl, M. Hocek, Org. Biomol. Chem. 2006, 4, 2278 – 245– 252. 2284; b) E. C. Western, J. R. Daft, E. M. Johnson, P. M. Gannett, [25] M. Fojta, L. Havran, M. Vojtiskova, E. Palecˇek, J. Am. Chem. Soc. K. H. Shaughnessy, J. Org. Chem. 2003, 68, 6767 –6774; c) E. C. 2004, 126, 6532–6533. Western, K. H. Shaughnessy, J. Org. Chem. 2005, 70, 6378– 6388. [20] T. Kovcs, L. tvçs, Tetrahedron Lett. 1988, 29, 4525– 4528. Received: May 15, 2013 [21] a) H. Kumamoto, H. Hayakawa, H. Tanaka, S. Shindoh, K. Kato, T. Revised: June 18, 2013 Miyasaka, K. Endo, H. Michida, A. Matsuda, Nucleosides Nucleoti- Published online: August 9, 2013

Chem. Eur. J. 2013, 19, 12720 – 12731 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.chemeurj.org 12731 FULL PAPER

DOI: 10.1002/chem.201101883

Anthraquinone as a Redox Label for DNA: Synthesis, Enzymatic Incorporation, and Electrochemistry of Anthraquinone-Modified Nucleosides, Nucleotides, and DNA

Jana Balintov,[a] Radek Pohl,[a] Petra Horkov,[b] Pavlna Vidlkov,[b] Ludeˇk Havran,[b] Miroslav Fojta,*[b] and Michal Hocek*[a]

Abstract: Modified 2’-deoxynucleosides none or 2-(2-propynylcarbamoyl)an- XL polymerase have been successfully and deoxynucleoside triphosphates thraquinone. Polymerase incorpora- developed. The electrochemical prop- (dNTPs) bearing anthraquinone (AQ) tions of the AQ-labeled dNTPs into erties of the AQ-labeled nucleosides, attached through an acetylene or prop- DNA by primer extension with KOD nucleotides, and DNA were studied by argylcarbamoyl linker at the 5-position cyclic and square-wave voltammetry, of pyrimidine (C) or at the 7-position which show a distinct reversible couple Keywords: anthraquinone · DNA · of 7-deazaadenine were prepared by of peaks around À0.4 V that make the electrochemistry · nucleosides · oli- Sonogashira cross-coupling of halogen- AQ a suitable redox label for DNA. gonucleotides ated dNTPs with 2-ethynylanthraqui-

Introduction redox labeling of DNA and for DNA minisequencing. Al- though the proof of concept for multicolor redox coding was DNA biosensors[1] are broadly applied in the life sciences. successful, the above-mentioned labels are far from being Electrochemical detection[2] is a less expensive alternative to perfect for real use in diagnostics. The ferrocenes are prone common optical methods, so redox labeling of DNA by di- to oxidation by air,[5] the dNTPs bearing bulky inorganic ACHTUNGRE verse electroactive tags is of increasing importance especial- [Os(bpy)3] are rather poor substrates for DNA polymerases ly in sequence-specific electrochemical DNA sensing.[3] Our and cannot be incorporated into adjacent positions,[8] and teams have recently jointly developed[4] redox labeling by the redox potential of Os2+ is close to that of 7-deazagua- polymerase incorporation of base-modified deoxynucleoside nine base.[8] Therefore, the development of a second genera- triphosphates (dNTPs) bearing a number of redox labels tion of redox labels is highly desirable. The labels should be based on ferrocenes,[5] amino and nitrophenyl groups,[6] tet- air-stable and not too bulky to make the dNTPs good sub- [7] ACHTUNGRE [8] rathiafulvalene, and [Ru/Os(bpy)3] complexes (bpy= strates for DNA polymerases. To increase sensitivity to 2,2’-bipyridine). Their combination (each nucleobase being DNA secondary structures (detection of single strands, mis- labeled by a different marker of different redox potential) matches or deletions), the redox label should preferably be has been used[8] for the first generation of “multicolor” tethered to the nucleobase by a conjugated linker (phenyl- ene or acetylene spacer). Anthraquinone (AQ) is a redox-active[9] molecule fre- quently utilized for electrochemical labeling of biomolecules.[10] Diverse AQ derivatives linked either directly or [a] J. Balintov, Dr. R. Pohl, Prof. Dr. M. Hocek through acetylene or longer flexible tethers were repeatedly Institute of Organic Chemistry and Biochemistry [11,12] [13,14] Academy of Sciences of the Czech Republic used for labeling of purine or pyrimidine nucleo- Gilead Sciences and IOCB Research Center sides and even chemically incorporated into DNA by phos- Flemingovo nam. 2, 16610 Prague 6 (Czech Republic) phoramidite synthesis to study charge transport through Fax : (+420)220-183-559 DNA.[12–14] Oligonucleotides bearing an AQ moiety linked E-mail: [email protected] Homepage: http://uochb.cas.cz/hocekgroup through a long flexible tether were observed to stabilize the [15] [16] [b] Dr. P. Horkov, P. Vidlkov, Dr. L. Havran, Prof. Dr. M. Fojta triplexes. On the other hand, AQ is a photo-oxidant [16] Institute of Biophsics and some AQ derivatives are DNA intercalators, protein v.v.i. Academy of Sciences of the Czech Republic photocleavers,[17] or inhibitors of enzymes.[18] To the best of Kralovopolska 135, 61265 Brno (Czech Republic) our knowledge, no AQ-bearing dNTPs or polymerase syn- Fax : (+420)541-211-293 thesis of AQ-DNA have been reported. Therefore, herein E-mail: [email protected] Homepage: http://www.ibp.cz/labs/LBCMO/ we report the synthesis of AQ derivatives of nucleosides Supporting information for this article is available on the WWW and dNTPs and their enzymatic incorporation and electro- under http://dx.doi.org/10.1002/chem.201101883. chemical properties.

Chem. Eur. J. 2011, 17, 14063 – 14073 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 14063 Results and Discussion

Base-modified dNTPs can be prepared either in the classical way by triphosphorylation[19] of modified nucleosides, or by our straightforward single-step aqueous cross-coupling reactions of halogenated dNTPs. 5-Substituted pyrimidine and 7- substituted 7-deazapurine dNTPs are usually good substrates[4,19,20] for DNA polymerases, whereas 8-substituted purine dNTPs are poor substrates. Also, it is known that an acetylene linker between a bulky aromatic substituent and the nucleobase makes the modified dNTPs better substrates. Therefore, we could not have used the previously reported modified purines[11,12] and directly linked AQ pyrimidines[13] and decided to introduce AQ through conjugated acetylene (analogous to the modification introduced by Barton et al. on deoxyuridine[14]) as well as nonconjugated propargylcarbamoyl linker (analogous to the modification introduced by Grinstaff and Tierney on purine[12]) to position 7 of 7-deaza- dATP (dATP=deoxyadenosine triphosphate) and to position 5 of deoxycytidine triphosphate (dCTP). To develop the Sonogashira cross-coupling reactions, we started with the modifications of halogenated nucleosides dAI and dCI (Schemes 1 and 2, Table 1) with 2-ethynylanthraquinone (EAQ)[12] or 2-(2-propynylcarbamoyl)anthraquinone (PAQ).[11] The reactions of dAI and dCI with EAQ in DMF ACHTUNGRE ACHTUNGRE Scheme 2. i) PAQ, [Pd(PPh3)2Cl2], CuI, (iPr)2EtN, DMF, 1 h, 758C. in the presence of [Pd(PPh3)2Cl2], CuI, and (iPr)2EtN pro- ACHTUNGRE ii) PAQ,Pd(OAc)2, TPPTS, CuI, (iPr)2EtN, CH3CN/H2O (2:1), 1 h, 758C.

ceeded very smoothly (Scheme 1) to give the corresponding AQ–nucleoside conjugates dAEAQ and dCEAQ in good yields of 79% (Table 1, entries 1, 2). On the other hand, under aqueous conditions (suitable for modification of dNTPs) in ACHTUNGRE the presence of Pd(OAc)2, triphenylphosphane 3,3’,3’’-trisul-

fonate (TPPTS), CuI, and (iPr)2EtN in water/acetonitrile (2:1) at 758C for 1 h the reactions of these nucleosides with EAQ did not proceed (Table 1, entries 3, 4). Attempted cross-couplings of EAQ with halogenated dNITPs (dAITP and dCITP) under the same aqueous conditions with the ACHTUNGRE same catalyst or with [Pd(PPh3)2Cl2] did not proceed either (Table 1, entries 5–7). As the reason could be the limited solubility of EAQ in water/acetonitrile, we also attempted the cross-coupling of dAITP and dCITP with EAQ in DMF/ ACHTUNGRE water (4:1) in the presence of [Pd(PPh3)2Cl2]. Under these conditions, the reaction proceeded to give the desired AQ– dNTPs dAEAQTP and dCEAQTP in moderate yields of about 30% (Table 1, entries 8, 9). To prepare larger quantities of these dNTPs, we also applied an alternative strategy of triphosphorylation[21] of the corresponding nucleosides. Thus, EAQ EAQ ACHTUNGRE the treatment of dA and dC with POCl3 in PO(OMe)3

followed by the addition of (NHBu3)2H2P2O7,Bu3N, and treatment with triethylammonium bicarbonate (TEAB; Scheme 1) gave the desired dNEAQTPs in good yields (65– 68%) after isolation by reversed-phase HPLC. The reactions of halogenated nucleosides and dNTPs with ACHTUNGRE Scheme 1. i) EAQ, [Pd(PPh3)2Cl2], CuI, (iPr)2EtN, DMF, 1 h, 75 8C. ACHTUNGRE PAQ were attempted under analogous conditions. Nucleo- ii) 1. PO(OMe)3, POCl3,08C; 2. (NHBu3)2H2P2O7,Bu3N, DMF, 08C; ACHTUNGRE sides dAI and dCI readily reacted with PAQ in DMF in the 3. TEAB. iii) EAQ, [Pd(PPh3)2Cl2], CuI, (iPr)2EtN, DMF/H2O (4:1), 1 h, ACHTUNGRE 758C. presence of [Pd(PPh3)2Cl2], CuI, and (iPr)2EtN to give the

14064 www.chemeurj.org 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Eur. J. 2011, 17, 14063 – 14073 Anthraquinone as a Redox Label for DNA FULL PAPER

Table 1. Synthesis of AQ-modified nucleosides and nucleotides. Entry Starting compound Reagent Catalyst Additives Solvent Product Yield [%] I ACHTUNGRE EAQ 1 dA EAQ [Pd(PPh3)Cl2] CuI, (iPr)2NEt DMF dA 79 I ACHTUNGRE EAQ 2 dC EAQ [Pd(PPh3)2Cl2] CuI, (iPr)2NEt DMF dC 79 I ACHTUNGRE EAQ 3 dA EAQ Pd(OAc)2, TPPTS CuI, (iPr)2NEt CH3CN:H2O (1:2) dA – I ACHTUNGRE EAQ 4 dC EAQ Pd(OAc)2, TPPTS CuI, (iPr)2NEt CH3CN:H2O (1:2) dC – I ACHTUNGRE EAQ 5 dA TP EAQ Pd(OAc)2, TPPTS CuI, (iPr)2NEt CH3CN:H2O (1:2) dA TP – I ACHTUNGRE EAQ 6 dC TP EAQ Pd(OAc)2, TPPTS CuI, (iPr)2NEt CH3CN:H2O (1:2) dC TP – I ACHTUNGRE EAQ 7 dA TP EAQ [Pd(PPh3)2Cl2] CuI, (iPr)2NEt CH3CN:H2O (1:2) dA TP – I ACHTUNGRE EAQ [a] 8 dA TP EAQ [Pd(PPh3)2Cl2] CuI, (iPr)2NEt DMF:H2O (4:1) dA TP 30 (36) I ACHTUNGRE EAQ [a] 9 dC TP EAQ [Pd(PPh3)2Cl2] CuI, (iPr)2NEt DMF:H2O (4:1) dC TP 31 (28) EAQ ACHTUNGRE EAQ 10 dA 1) PO(OMe)3, POCl3,08C; 2) (NHBu3)2H2P2O7,Bu3N, DMF, 0 8C; 3) TEAB dA TP 65 EAQ ACHTUNGRE EAQ 11 dC 1) PO(OMe)3, POCl3,08C; 2) NHBu3)2H2P2O7,Bu3N, DMF, 08C; 3) TEAB dC TP 68 I ACHTUNGRE PAQ 12 dA PAQ [Pd(PPh3)2Cl2] CuI, (iPr)2NEt DMF dA 80 I ACHTUNGRE PAQ 13 dC PAQ [Pd(PPh3)2Cl2] CuI, (iPr)2NEt DMF dC 83 I ACHTUNGRE PAQ 14 dA PAQ Pd(OAc)2, TPPTS CuI, (iPr)2NEt CH3CN:H2O (1:2) dA – I ACHTUNGRE PAQ 15 dC PAQ Pd(OAc)2, TPPTS CuI, (iPr)2NEt CH3CN:H2O (1:2) dC – I ACHTUNGRE PAQ [a] 16 dA TP PAQ Pd(OAc)2, TPPTS CuI, (iPr)2NEt CH3CN:H2O (1:2) dA TP 80 (63) I ACHTUNGRE PAQ [a] 17 dC TP PAQ Pd(OAc)2, TPPTS CuI, (iPr)2NEt CH3CN:H2O (1:2) dC TP 79 (74) [a] In parentheses, the yields of dNXTPs were recalculated from UV spectra and the Lambert–Beer equation. desired dAPAQ and dCPAQ in good yields of 80–83% poration of dNXAQTPs into different sequences and both (Table 1, entries 12, 13), whereas the same reactions under single modification and four modifications (either in sepa- aqueous conditions did not proceed (Table 1, entries 14, 15). rate positions or four in-line). Single nucleotide extension However, the aqueous Sonogashira reactions of dAITP and experiments were tested separately with each of the four dCITP with PAQ in the presence of TPPTS, CuI, and dNXAQTPs by using KOD XL polymerase. All four XAQ (iPr)2EtN in water/acetonitrile (2:1) proceeded very smooth- dN TPs (Figure 1, lanes 4, 5, 8, 9) were successfully incor- ly to give the desired nucleotides dAEAQTP and dCEAQTP in porated into DNA. excellent yields of approximately 80% (Table 1, entries 16, 17). Because the dNXAQTPs were isolated as salts with unde- fined molecular formula, the key yields were also recalculated from their absorbance in UV spectra and the Lambert– Beer equation by using the extinction coefficients of the corresponding nucleosides to confirm similar values (Table 1, entries 8, 9, 16, 17 and the Supporting Information). The enzymatic incorporation of the AQ-modified dNAQTPs in primer extension (PEX) experiments was studied by using KOD XL, Vent (exo-), Deep Vent (exo-), and Pwo polymerases, similarly to our previous works (for sequences of templates and primers, see Table 2). The tem- Figure 1. Denaturing polyacrylamide gel electrophoresis (PAGE) analysis plates and primers were chosen to compare the PEX incor- of incorporations of dNXAQTPs by using KOD XL polymerase, primrnd, tempC, and tempA. Compositions of the dNTP mixes and nucleotide labeling are as follows: “+ ” positive control (dATP or dCTP and deoxy- guanosine triphosphate (dGTP)), negative control experiments NÀ (ab- Table 2. Primers and templates used for PEX experiments. sence of each natural dNTP), NXAQ (dNXAQTP+dGTP). Sequences primrnd 5’-CATGGGCGGCATGGG-3’ primrndA 5’-CATGGGCGGCATGGA-3’ To compare the efficiency in incorporation of the AQ- primrndC 5’-CATGGGCGGCATGGC-3’ modified dNTPs in comparison with the natural ones, we rndT prim 5’-CATGGGCGGCATGGT-3’ performed a simple kinetics study in single-nucleotide PEX temprnd16 5’- CCCATGCCGCCCATG-3’ CTAGCATGAGCTCAGT reaction (Figure 2 and the Supporting Information). In all tempA 5’-CCCTCCCATGCCGCCCATG-3’ tempAterm 5’-TCCCATGCCGCCCATG-3’ cases, the incorporation of the natural dNTP was finished tempC 5’-CCCGCCCATGCCGCCCATG-3’ within 10–30 s whereas the PEX with AQ-modified dNTPs tempAA 5’-CCCTTCCATGCCGCCCATG-3’ took 2–5 min to complete. The dNTPs bearing the rigid AC temp 5’-CCCGTCCATGCCGCCCATG-3’ EAQ label were incorporated more slowly than those bear- tempCA 5’-CCCTGCCATGCCGCCCATG-3’ tempCC 5’-CCCGGCCATGCCGCCCATG-3’ ing the PAQ group. Therefore, the reaction time for multi- tempTA 5’-CCCTACCATGCCGCCCATG-3’ ple incorporations must be prolonged to 30 min to ensure tempTC 5’-CCCGACCATGCCGCCCATG-3’ full-length product formation. Also, we tried the incorpora- C4line temp 5’-TCATCATCATAGGGGCCCATGCCGCCCATG-3’ tion of a single modified dNXAQTP followed by three natural tempA4line 5’- CCCATGCCGCCCATG-3’ CAGCAGCAGCATTTT dGTPs in different sequence contexts (with primers having

Chem. Eur. J. 2011, 17, 14063 – 14073 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.chemeurj.org 14065 M. Hocek, M. Fojta et al.

pected full-length products (see the Supporting Informa- tion), clearly confirming full extension (the differences in PAGE mobilities might arise from formation of different secondary structures of the ONs containing several AQ moieties that could intercalate or stack with each other). To further increase the efficiency of the incorporation of the modified dNXAQTPs by KOD polymerase, we tried to increase Figure 2. Comparison of the rate of the PEX with natural C + (dCTP) or XAQ modified CPAQ (dCPAQTP) nucleotides using KOD XL polymerase with the concentration of the modified dN TPs to ten times tempC and primrnd without natural dGTP. The reaction mixtures were in- that of the natural dNTPs (Figure 3, lanes 9–12). Surprising- cubated for the time intervals indicated, followed by stopping the reac- ly, inhibition of the PEX reaction by dNAQTP resulted. This tion by addition of PAGE loading buffer and immediate heating. inhibition is interesting because it apparently does not proceed by termination of the primer by the modified nucleotide. This may indicate either binding of the modified A, T, or C at the 3’-end) and in all cases the incorporation dNTPs to the polymerase or damage of the enzyme by (pho- was successful but the efficiency was sequence dependent to)oxidation by AQ (analogously to refs. [16–18]). (see Figures S6–S8 in the Supporting Information). Finally, the KOD XL polymerase was challenged by incor- Multiple incorporations into a random sequence contain- poration of four AQ-containing dNTPs at adjacent positions ing all four bases (temprnd16) were also tested. The best re- followed by 11 unmodified nucleotides (Figure 4). Both sults of incorporations were obtained using KOD XL polymerase (Figure 3), whereas other tested polymerases (Vent exo-, Pwo, Deep Vent exo-) gave worse results (see the Sup- porting Information). Both flexible propargylamide-linked dNTPs (dAPAQTP and dCPAQTP; Figure 3, lanes 6, 8) gave the desired fully extended oligonucleotide (ON), whereas the rigid dNEAQTPs gave either a shorter ON (dAEAQTP; Figure 3, lane 5) or full-length but slightly impure product (dCEAQTP; Figure 3, lane 7). The PEX product containing CPAQ showed slightly higher mobility than the positive control and APAQ product, thereby suggesting that the product might be shorter by one nucleotide. However, MALDI analysis of both PEX products gave correct masses of the ex-

Figure 4. Denaturing PAGE analysis of PEX experiments with KOD XL polymerase and the tempA4line, tempC4line templates and primrnd. Composi- tions of the dNTP mixes and nucleotide labeling are as follows: “+” positive control (dATP, dCTP, dTTP, dGTP), negative control experiments NÀ (absence of each natural dNTP), NXAQ (dNXAQTP +three other natural dNTPs).

PAQ-linked nucleotides (dAPAQTP and dCPAQTP; Figure 4, lanes 5, 10) as well as dAEAQTP (Figure 4, lane 4) gave full- length products, whereas the dCEAQTP (Figure 4, lane 9) did not give full extension. Apparently, the more flexible dNPAQTPs are better substrates for KOD polymerase than the rigid dNEAQTPs and thus are more suitable as building blocks for redox labeling of DNA. Thermal denaturation temperatures of our modified double-stranded (ds) DNAs containing NEAQ or dNPAQ nu- Figure 3. Denaturing PAGE analysis of PEX experiments with KOD XL cleotides were also studied. The melting temperatures of the rnd16 rnd polymerase, temp , and prim . Compositions of the dNTP mixes and natural and modified DNAs are summarized in Table 3 (for nucleotide labeling are as follows: “ +” positive control (dATP, dCTP, sequences see Table 4). The presence of rigid EAQ groups deoxythymidine triphosphate (dTTP), dGTP), negative control experiments NÀ (absence of each natural dNTP), NXAQ (dNXAQTP+three resulted in significant destabilization (DTm =À1.4 to À48C) other natural dNTPs). of the duplex. On the other hand, the PAQ substituents

14066 www.chemeurj.org 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Eur. J. 2011, 17, 14063 – 14073 Anthraquinone as a Redox Label for DNA FULL PAPER

Table 3. Melting temperatures of DNA duplexes. DNA1 DNA2 [a] [a] Tm [8C] DTm Tm [8C] DTm A 80.4 – 81.2 – AEAQ 76.4 À4 69.2 À3 APAQ 80.5 +0.1 81.2 0 DNA3 DNA4 C 77.4 0 78 0 CEAQ 76.0 À1.4 69 À2.25 CPAQ 78.4 +1780

[a] DTm =(TmmodÀTm+)/nmod.

Table 4. Oligonucleotides used for measurements of Tm. DNA1 5’-CCCTCCCATGCCGCCCATG-3’ 3’-GGGAGGGTACGGCGGGTAC-5’ DNA2 5’-CTAGCATGAGCTCAGTCCCATGCCGCCCATG-3’ 3’-GATCGTACTCGAGTCAGGGTACGGCGGGTAC-5’ DNA3 5’-CCCGCCCATGCCGCCCATG-3’ 3’-GGGCGGGTACGGCGGGTAC-5’ DNA4 5’-CTAGCATGAGCTCAGTCCCATGCCGCCCATG-3’ 3’-GATCGTACTCGAGTCAGGGTACGGCGGGTAC-5’

slightly stabilized the dsDNA (DTm =0to+ 1). These results also support the conclusion that the PAQ substituents are more suitable as DNA labels. The electrochemical properties of the AQ-labeled nucleosides, dNTPs, and ONs were studied by cyclic voltammetry (CV) and/or square-wave voltammetry (SWV) at hanging mercury drop (HMDE) and pyrolytic graphite electrodes (PGE; Figure 5). Our attention has been oriented mainly towards the characteristic quinone/hydroquinone redox pair[22] close to À0.5 V, which gives rise to the peaks AQred (catho- ox dic, due to reduction of the AQ moiety) and AQH2

(anodic, due to reoxidation of anthrahydroquinone (AQH2), the product of the former electrode reaction). Nevertheless, we investigated the voltammetric responses of the title compounds in a wider potential range, particularly through re- gions where signals of natural nucleobases occur,[2] to assess the possibility of simultaneous monitoring of DNA modification and determination of the DNA amount through measuring its intrinsic responses (see below). Compared to free AQ, the redox potential of the AQ/

AQH2 pair at the HMDE in EAQ and PAQ was shifted to Figure 5. Voltammetric responses of AQ-labeled nucleosides and ONs PAQ less negative potentials by 25–35 mV (Figure 5a, Table 5), (PEX products). a) CV at HMDE: free AQ (black); PAQ (red); dC (blue): concentration of all substances 40 mm, initial potential 0 V, switch- thus suggesting more facile reduction and more difficult oxi- ing potential À1.85 V; inset: switching potential À1.3 V. b) CV at dation in the presence of the substituents on the AQ aro- HMDE: PEX product pexrnd16(ACHTUNGRECPAQ) (black); the same but unmodified matic system. These shifts could in principle be ascribed to (red); negative PEX (no enzyme added: blue). Other conditions (includ- electronic effects of unsaturated conjugate groups (ethynyl ing the inset) as in (a). c) SWV at PGE: PEX product pexrnd16(ACHTUNGRECPAQ) or carbonyl, respectively); on the other hand, measurements (black); the same but unmodified (red). Inset: CV at PGE: samples as in (a). with the same compounds at the PGE revealed more complex phenomena, which resulted in negative potential shifts of the reduction peak and positive shift of the oxidation duction peaks), probably due to less facile electron transfer peak in EAQ (Table 5), whereas PAQ exhibited positive at the PGE compared to the HMDE. In EAQ–nucleoside shifts of both signals. In general, the AQ/AQH2 redox pro- conjugates, further significant (more than 100 mV compared cess at the PGE showed poorer reversibility (indicated by to EAQ) positive shift of the AQ/AQH2 pair measured at remarkable separation of the potentials of oxidation and re- the HMDE was observed, probably due to electron-with-

Chem. Eur. J. 2011, 17, 14063 – 14073 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.chemeurj.org 14067 M. Hocek, M. Fojta et al.

Table 5. Redox potentials of AQ/anthrahydroquinone moiety in building produced by unmodified ONs as well. In agreement with ob- blocks, nucleosides, dNTPs, and ONs. servations made with the nucleosides and dNTPs, no anodic HMDE[a] PGE[b] ox peak AQH2 was detected under these conditions. When, red ox red ox AQ AQH2 AQ AQH2 however, the direction of potential scan was switched at AQ À0.460 À0.460 À0.518 À0.460 À1.3 V, the AQ/AQH2 redox pair exhibited excellent rever- EAQ À0.435 À0.435 À0.540 À0.415 sibility but no peak G was observed (because in this case po- PAQ À0.425 À0.420 À0.480 À0.435 dCEAQ À0.300 À0.300 À0.510 À0.420 tentials applied at the electrode were not sufficiently nega- dCPAQ À0.410 À0.415 À0.490 À0.400 tive for G reduction; see inset in Figure 5b). In general it is dCEAQTP À0.420 À0.415 À0.530 À0.420 useful to measure simultaneously the voltammetric signals PAQ dC TP À0.400 À0.390 À0.520 À0.400 specific for a label introduced and intrinsic DNA responses pexrnd16(ACHTUNGRECPAQ)[c] À0.410 À0.370 À0.520 À0.440 (for example, for normalization of the label signal on the dAEAQ À0.320 À0.320 À0.557 À0.410 [5–8] dAPAQ À0.415 À0.425 À0.490 À0.435 total DNA amount). Here it is apparently impossible to dAEAQTP À0.430 À0.425 À0.520 À0.415 measure the DNA peaks CA and G in one CV scan without dAPAQTP À0.425 À0.390 À0.522 À0.405 loss of the AQ/AQH2 reversibility. Nevertheless, all of the rnd16ACHTUNGRE PAQ [c] pex (A ) À0.410 À0.380 À0.520 À0.420 above signals can easily be obtained in two successive scans [a] CV measured at the HMDE in 0.3 m ammonium formate, 0.05 m by using the less negative switching potential for the first sodium phosphate (pH 6.9) with initial potential 0.0 V and switching po- and the more negative for the second (see the Supporting tential À1.3 V. [b] CV measured in sodium acetate (pH 5) with other conditions as above; for more details see the Experimental Section. [c] Prod- Information for more details). At the PGE, the AQ-specific uct of PEX on temprnd16, bearing four PAQ-labeled C or A residues. peak was obtained in one scan together with purine oxidation peaks Gox and Aox when initial potentials more negative

than the potential of the AQ/AQH2 redox pair and positive drawing effects of the nucleobase electronically coupled scan direction were applied (Figure 5c). through the ethynyl linker.[5,23] In analogous PAQ–nucleo- Furthermore, we tested the possibility of simultaneous de- side conjugates, the nucleobase effects were in minor agree- tection of AQ labels and another electrochemically reduci- [6] ment with the insulating effect of the propargyl bridge. No ble tag attached to nucleobases, 2-nitrophenyl (PhNO2). significant differences in peak potentials between the corre- Previously, for the purposes of multicolor DNA coding, we sponding dAX and dCX conjugates were observed. AQ-la- proposed combinations of various enzymatically introduced beled dNTPs showed in general more negative potentials of electroactive tags differing in their redox potentials and/or the AQ/AQH2 pair compared to the corresponding nucleo- character of the electrochemical process they undergo (re- sides (Table 5), probably due to charge effects of the phos- versible redox, irreversible oxidation or reduction).[6,8] phate groups affecting adsorption of the compounds at the Redox potentials of most of the (reversibly or irreversibly) electrode surfaces and consequently accessibility of the oxidizable tags occur between about + 0.4 and +1.1 V and redox-active groups for the electrode reaction.[24] voltammetric signals of some of them partially interfere

Figure 5a shows that reversibility of the AQ/AQH2 redox with those due to oxidation of natural purines or their 7- process is maintained in CV when the potential scan is deaza analogues. On the other hand, reduction of the nucle- switched from negative to positive at potentials 0À1.3 V obases occurs at potentials around À1.5 V (C, A) or more (inset in Figure 5a). However, when highly negative poten- negative (G), and thus there is no interference between sig- tials are applied (such as À1.85 V to reach G reduction in nals resulting from nucleobase reduction and those due to ox DNA, see below), no anodic AQH2 is observed, which sug- the external reducible tags such as nitrophenyl and AQ, and gests deeper reduction of the AQ moiety resulting in its irre- the only problem is differentiation between these tags. For versible destruction. With nucleosides and nucleotides, re- this purpose we prepared pexrnd16(CX) and pexrnd16(AX) prod- duction signals of the C or A moieties were observed at cor- ucts in which PhNO2 and PAQ were combined at various responding potentials (in Figure 5a shown for dCPAQ, peak ratios (in Figure 6a shown for AX). Reduction peaks of both red red red C ). Depending on the conditions, some of the title com- labels (AQ around À0.41 and NO2 around À0.49 V) pounds produced additional voltammetric signals (Fig- could easily be distinguished when PAQ alone or PhNO2 ure 5a), the detailed analysis of which is beyond the scope alone was introduced. However, when both labels were of this paper and will be presented elsewhere. present in the same ON chain, their reduction signals over- By using the temprnd16 template PEX products bearing lapped due to the relatively small difference in their peak four PAQ-labeled C or A residues were prepared and ana- potentials (ACHTUNGRE80 mV) and it was rather difficult to detect red lyzed by CV at the HMDE (Figure 5b) or PGE (Figure 5c). peak AQ if one PAQ per three PhNO2 was incorporated CV results of the pexrnd16(ACHTUNGRECPAQ) measured at the HMDE for or vice versa. Nevertheless, since the reduction of AQ is re- switching potential À1.85 V revealed peak AQred at À0.41 V versible and that of the nitro group irreversible, only peak ox in addition to peak CA close to À1.5 V (due to irreversible AQH2 was detected in CV on the anodic branch (Fig- reduction of unmodified C and A in the ON) and peak G ure 6a), thus unmasking the minority PAQ label. Moreover,

(due to reoxidation of 7,8-dihydroguanine to which G is the PhNO2 labels could be detected indirectly during the converted at potentials close to the switching potential of anodic voltage scan by using a signal corresponding to rever- À1.85 V).[2] The latter signals—unlike the AQred peak—were sible oxidation of hydroxylamine (the product of four-elec-

14068 www.chemeurj.org 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Eur. J. 2011, 17, 14063 – 14073 Anthraquinone as a Redox Label for DNA FULL PAPER

[6] Scheme 3. Structures of NO2-modified dNTPs used in this study.

proved in principle applicable in single nucleotide polymorphism probing and useful to complement the palette of previously introduced multicolor DNA labels.

Conclusion Figure 6. a) CV responses of pexrnd16(AX) products with incorporated PAQ PhNO2 A and A conjugates at different ratios (given in the panel). Novel AQ-modified dNTPs linked through conjugate acety- b) CV responses obtained for sequence-specific incorporation of a single X lene or nonconjugate propargylcarbamoyl linker have been A bearing either PAQ or PhNO2 label. PEX reactions were performed with tempA template and dN(X)TP mixes given in the panel. CVs were prepared and tested as substrates for DNA polymerases. measured with initial potential 0.0 V, switching potential À0.7 V, and The Sonogashira cross-couplings of halogenated nucleosides final potential +0.05 V; other conditions as in Figure 5. and dNTPs with EAQ proceeded only in DMF and therefore the corresponding dNTPs were prepared by triphosphorylation of nucleosides. On the other hand, the more tron reduction of the nitro group) to the nitroso group (see polar PAQ was sufficiently reactive in aqueous cross-cou- the Supporting Information for more details). Because the plings with halogenated dNTPs. All four AQ-linked corresponding peak NHOHox occurs around À0.01 V dNXAQTPs(dAEAQTP, dCEAQTP, dAPAQTP, and dCPAQTP) (Figure 6), the almost 400 mV peak potential separation be- were tested as substrates of DNA polymerases. KOD XL ox ox tween peaks AQH2 and NHOH allows perfect resolution polymerase was identified as the most suitable enzyme for and independent detection of both labels (Scheme 3). their incorporation. In single incorporation experiments, all Finally, we tested the possibility of electrochemical moni- four dNXAQTPs worked well, whereas for multiple incorpo- toring of the sequence-specific incorporation of a single nu- rations, the more flexible dNPAQTPs were better substrates EAQ cleotide labeled with either PAQ or PhNO2 tag (Figure 6b). than the rigid dN TPs. Under tenfold higher concentra- The PEX reaction was carried out with the tempA template tion of dNXAQTPs, inhibition of the polymerase was ob- and mixtures containing either dAPAQTP+dCPhNO2TP+ served. Electrochemical studies of the AQ-modified nucleo- dGTP, or dAPhNO2TP+ dCPAQTP +dGTP. It is evident from tides and DNA (PEX products) by voltammetry revealed Figure 6 b that CV responses resulting from these experi- well-developed peaks of reversible reduction of the AQ ments revealed perfectly specific incorporation of labeled moiety around À0.4 V. Combination of the AQ modification AX against thymine residue in the template strand: when with previously reported nitrophenyl labeling gave one un- the reaction mixture contained dAPAQTP, peaks AQred and resolved broad reduction peak. However, CV can easily dis- ox red AQH2 were detected, whereas the absence of NO2 and tinguish between these two labels since the reduction of ox NHOH peaks indicated no significant misincorporation of NO2 (unlike AQ) is irreversible, and produces no oxidation the labeled cytosine. For reaction mixtures containing signal interfering with that of AQH2 oxidation. Independent PhNO2 dA TP the opposite was true. Hence, PAQ labels detection of the PhNO2 in the presence of AQ is possible

Chem. Eur. J. 2011, 17, 14063 – 14073 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.chemeurj.org 14069 M. Hocek, M. Fojta et al.

ACHTUNGRE through oxidation of hydroxylamine, product of the NO2 re- 0.05 mmol, 1.2 equiv), [Pd(PPh3)2Cl2] (2 mg, 0.002 mmol, 5 mol%), CuI (1 mg, 0.004 mmol), and (iPr)2EtN (0.74 mL, 0.42 mmol, 10 equiv). The duction. Therefore, the PhNO2 and PAQ labels have a good mixture was stirred and heated to 758C for 1 h. The product was isolated potential for base-specific multiple labeling and selective de- from the crude reaction mixture by HPLC on a C18 column with the use PAQ tection of one in the presence of the other. The dC TP m m of a linear gradient of 0.1 TEAB in H2O to 0.1 TEAB in H2O/MeOH and dAPAQTP linked through the propargylcarbamoyl group (1:1) as eluent. Several co-distillations with water and conversion to are good substrates for polymerase incorporation, so the sodium salt form (Dowex 50WX8 in Na + cycle) followed by freeze- AQ can be used to complete the palette of redox labels for drying from water gave a solid product. EAQ EAQ EAQ multicolor DNA coding in combination with some of the Triphosphosphorylation of the dN : Method C: dA TP, dC TP: Dry trimethyl phosphate (0.16 mL) was added to an argon-purged flask previously reported or future novel redox labels. Studies to- containing nucleoside analogue dNEAQ (0.06 mmol, 1 equiv) cooled to wards practical applications in diagnostics are under way. 08C on ice followed by the addition of POCl3 (7 mL, 0.07 mmol,

1.2 equiv). After 1.5 h, a solution of (NHBu3)2H2P2O7 (181 mg, 0.3 mmol,

5 equiv) and Bu3N (0.04 mL, 0.3 mmol, 4.2 equiv) in dry DMF (1 mL) was added to the reaction mixture which was stirred for another 1 h and Experimental Section quenched by 2 m TEAB buffer (1 mL). The product was isolated from the crude reaction mixture by HPLC on a C18 column with the use of a linear gradient of 0.1 m TEAB in H O to 0.1 m TEAB in H O/MeOH General chemistry: All cross-coupling reactions were performed under 2 2 (1:1) as eluent. Several co-distillations with water and conversion to an argon atmosphere. Compounds 7-I-7-deaza-dATP,[6] 5-I-dCTP,[20] 2-(2- sodium salt form (Dowex 50WX8 in Na + cycle) followed by freeze- propynylcarbamoyl)anthraquinone,[11] and 2-ethynylanthraquinone[12] drying from water gave a solid product. were prepared according to the literature procedures. Other chemicals were purchased from commercial suppliers and were used as received. General procedure for the Sonogashira cross-coupling reaction of 2-(2- NMR spectra were measured on Bruker Avance 500 (500 MHz for 1H, propynylcarbamoyl)anthraquinone with halogenated nucleoside triphos- PAQ PAQ 125.7 MHz for 13C, and 202.3 MHz for 31P) or Bruker 600 (600 MHz phates: Method D: dA TP, dC TP: A 2:1 mixture of H2O/CH3CN (2 mL) followed by (iPr) EtN (0.75 mL, 10 equiv) was added to an argon- for 1H, 150.9 MHz for 13C) instruments in D2O (referenced to dioxane 2 I as internal standard, d(1H)=3.75 ppm, d(ACHTUNGRE13C) =67.19 ppm, standard for purged flask containing halogenated nucleoside triphosphate dA TP or dCITP (0.04 mmol), 2-(2-propynylcarbamoyl)anthraquinone (19 mg, 31P NMR was external H3PO4). Chemical shifts are given in ppm (d scale), coupling constants (J) in Hz. Complete assignment of all NMR 0,06 mmol, 1.2 equiv), and CuI (1 mg, 10 mol%). In a separate flask, Pd- ACHTUNGRE signals was achieved by use of a combination of H,H-COSY, H,C-HSQC, (OAc)2 (1 mg, 0.003 mmol, 5 mol %) and TPPTS (3 mg, 0.005 mmol, and H,C-HMBC experiments. NMR spectra of dNTPs were measured in 2.5 equiv with respect to Pd) were combined, the flask was evacuated phosphate buffer at pH 7.1. Mass spectra were measured on an LCQ and purged with argon, and then a 2:1 mixture of H2O/CH3CN (0.5 mL) classic (Thermo-Finnigan) spectrometer by using ESI or Q-Tof Micro was added. This catalyst solution was injected into the reaction mixture, (Waters, ESI source, internal calibration with lock spray). Preparative which was then stirred at 75 8C for 1 h. The product was isolated from HPLC separations were performed on a column packed with 10 mm C18 the crude reaction mixture by HPLC on a C18 column with the use of a m m reversed-phase material (Phenomenex, Luna C18(2)). IR spectra were linear gradient of 0.1 TEAB in H2O to 0.1 TEAB in H2O/MeOH measured either on a Bruker Alpha FTIR spectrometer with the ATR (1:1) as eluent. Several co-distillations with water and conversion to technique or by using KBr tablets. High-resolution mass spectra were sodium salt form (Dowex 50WX8 in Na + cycle) followed by freeze- measured on an LTQ Orbitrap XL (Hermo Fischer Scientific) spectrome- drying from water gave a solid product. ter by using the ESI technique. Mass spectra of functionalized DNA were dAEAQ : Compound dAEAQ was prepared from dAI according to the gener- measured by MALDI-TOF, Reflex IV (Bruker) with nitrogen laser. UV/ al procedure (Method A). The product was isolated as a dark red solid Vis spectra were measured on a Varian CARY 100 Bio spectrophotome- 1 (63.5 mg, 79%). M.p. 186–192 8C; H NMR (499.8 MHz, [D6]DMSO): ter at room temperature. Melting points were determined on a Kofler 2.24 (ddd, Jgem =13.3, J2’b,1’ 6.0, J2’b,3’ = 2.9 Hz, 1 H; H-2’b), 2.52 (ddd, Jgem = block. 13.3, J2’a,1’ =7.9, J2’a,3’ =5.9 Hz, 1H; H-2’a), 3.55 (ddd, Jgem =11.8, J5’b,OH =

General procedure for the Sonogashira cross-coupling reaction of 2-ethy- 5.8, J5’b,4’ =4.4 Hz, 1H; H-5’b), 3.61 (ddd, Jgem =11.8, J5’a,OH =5.4, J5’a,4’ = nylanthraquinone and 2-(2-propynylcarbamoyl)anthraquinone with halo- 4.4 Hz, 1H; H-5’a), 3.86 (td, J4’,5’ =4.4, J4’,3’ = 2.6 Hz, 1 H; H-4’), 4.38 (m, EAQ EAQ genated nucleosides: Method A: dA , dC : Dry DMF (3 mL) was 1H; H-3’), 5.09 (dd, JOH,5’ =5.8, 5.4 Hz, 1 H; OH-5’), 5.31 (d, JOH,3’ = added to an argon-purged flask containing 2-ethynylanthraquinone 4.1 Hz, 1 H; OH-3’), 6.53 (dd, J1’,2’ =7.9, 6.0 Hz, 1H; H-1’), 6.90 (bs, 2H; I (47 mg, 0.20 mmol, 1.2 equiv), nucleoside analogue dN (0.16 mmol, NH2), 7.95 (m, 2 H; H-6,7-anthr), 8.04 (s, 1 H; H-6), 8.10 (dd, J3,4 =8.0, ACHTUNGRE 1 equiv), CuI (4 mg, 0.02 mmol, 10 mol %), and [Pd(PPh3)2Cl2] (6 mg, J3,1 = 1.7 Hz, 1H; H-3-anthr), 8.17 (s, 1H; H-2), 8.23 (m, 3H; H-4,5,8- 13 0.008 mmol, 5 mol %) followed by (iPr)2EtN (0.3 mL, 1.7 mmol, anthr), 8.35 ppm (dd, J1,3 =1.7, J1,4 =0.6 Hz, 1H; H-1-anthr); C NMR

10 equiv). The reaction mixture was stirred at 758C for 1 h until complete (125.7 MHz, [D6]DMSO): 39.87 (CH2-2’), 62.05 (CH2-5’), 71.12 (CH-3’), consumption of the starting material and then evaporated in vacuo. The 83.49 (CH-1’), 87.83 (CH-4’), 87.33 (C5-CC-anthr), 90.41 (C5-CC- products were purified by silica gel column chromatography with chloro- anthr), 94.31 (C-5), 102.00 (C-4a), 127.00, 127.04, 127.32 (CH-4,5,8- form/methanol (0 to 10 %) as eluent. anthr), 128.72 (CH-6), 129.08 (C-2-anthr), 129.10 (CH-1-anthr), 131.94 dAPAQ, dCPAQ : Dry DMF (3 mL) was added to an argon-purged flask con- (C-4a-anthr), 133.18, 133.34, 133.41 (C-8a,9a,10a-anthr), 134.79, 134.93 taining 2-(2-propynylcarbamoyl)anthraquinone (45 mg, 0.15 mmol, (CH-6,7-anthr), 136.42 (CH-3-anthr), 149.88 (C-7a), 153.76 (CH-2), 1.2 equiv), nucleoside analogue dNI (0.13 mmol, 1 equiv), CuI (3 mg, 157.76 (C-4), 182.04 (C-10-anthr), 182.28 ppm (C-9-anthr); IR (KBr): n˜ = ACHTUNGRE 0.013 mmol, 10 mol%), and [Pd(PPh3)2Cl2] (5 mg, 0.007 mmol, 5 mol%) 3461, 3450, 3290, 2200, 1668, 1628, 1591, 1576, 1568, 1558, 1530, 1479, À1 followed by (iPr)2EtN (0.23 mL, 1.3 mmol, 10 equiv). The reaction mix- 1450, 1329, 1302, 1285, 1263, 1180, 1149, 1055 cm ; MS (ESI +): m/z + ture was stirred at 75 8C for 1 h until complete consumption of the start- (%): 503.1 (100) [M À2H+Na]; HRMS (ESI+): calcd for C27H20N4Na ing material and then evaporated in vacuo. The products were purified O5 : 503.1326; found 503.1324. by silica gel column chromatography with chloroform/methanol (0 to dCEAQ : Compound dCEAQ was prepared from dCI according to the gener- 10%) as eluent. al procedure (Method A). The product was isolated as a yellow-orange 1 General procedure for the Sonogashira cross-coupling reaction of 2-ethy- solid (61 mg, 79%). M.p.>3008C; H NMR (500.0 MHz, [D6]DMSO): nylanthraquinone with halogenated nucleoside triphosphates: Method B: 2.07 (dt, Jgem =13.4, J2’b,1’ =J2’b,3’ =6.1 Hz, 1H; H-2’b), 2.21 (ddd, Jgem = EAQ EAQ dA TP, dC TP: A DMF/H2O mixture (4:1, 2.5 mL) was added 13.4, J2’a,1’ = 6.1, J2’a,3’ =4.0 Hz, 1H; H-2’a), 3.61 and 3.69 (2ddd, Jgem = I through a septum to an argon-purged vial containing dA TP sodium salt 12.0, J5’,OH =5.1, J5’,4’ =3.6 Hz, 21 H; H-5’), 3.83 (q, J4’,5’ =J4’,3’ =3.6 Hz, I or dC TP sodium salt (0.04 mmol), 2-ethynylanthraquinone (12 mg, 1H; H-4’), 4.25 (m, 1H; H-3’), 5.19 (t, JOH,5’ =5.1 Hz, 1 H; OH-5’), 5.25

14070 www.chemeurj.org 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Eur. J. 2011, 17, 14063 – 14073 Anthraquinone as a Redox Label for DNA FULL PAPER

(d, JOH,3’ = 4.3 Hz, 1H; OH-3’), 6.13 (t, J1’,2’ =6.1 Hz, 1H; H-1’), 7.39 and 1H; H-2’a), 3.18 (t, Jvic =7.2 Hz, 12H; CH3CH2N), 4.15–4.27 (m, 3H; H-

7.87 (2bs, 2 1H; NH2), 7.95 (m, 2H; H-6,7-anthr), 8.08 (dd, J3,4 =8.1, 4’,5’), 4.76 (m, 1H; H-3’), 6.39 (dd, J1’,2’ =8.0, 5.9 Hz, 1H; H-1’), 7.65–7.75

J3,1 = 1.7 Hz, 1 H; H-3-anthr), 8.22 (m, 3H; H-4,5,8-anthr), 8.44 (d, J1,3 = (m, 3H; H-6, H-6,7-anthr), 7.82 (d, J3,4 =7.9 Hz, 1H; H-3-anthr), 7.92 (d, 13 1.7 Hz, 1H; H-1-anthr), 8.48 ppm (s, 1 H; H-6); C NMR (125.7 MHz, J8,7 = 7.3 Hz, 1H; H-8-anthr), 7.96 (s, 2 H; H-2, H-1-anthr), 8.04 (d, J5,6 =

[D6]DMSO): 41.10 (CH2-2’), 61.01 (CH2-5’), 70.01 (CH-3’), 85.80 (CH-1’), 7.2 Hz, 1H; H-5-anthr), 8.10 ppm (d, J4,3 =7.9 Hz, 1 H; H-4-anthr); 13 87.33 (C5-CC-anthr), 87.70 (CH-4’), 88.92 (C-5), 93.04 (C5-CC-anthr), C NMR (125.7 MHz, CD3OD+D2O): 9.20 (CH3CH2N), 40.32 (CH2-2’),

126.99, 127.04, 127.19 (CH-4,5,8-anthr), 128.94 (C-2-anthr), 129.24 (CH- 47.64 (CH3CH2N), 66.78 (d, JC,P = 5.8 Hz; CH2-5’), 72.40 (CH-3’), 84.34

1-anthr), 131.98 (C-4a-anthr), 133.20 (C-9a-anthr), 133.33, 133.41 (C- (CH-1’), 86.75 (d, JC,P =9.1 Hz; CH-4’), 88.27 (C5-CC-anthr), 92.24 (C5- 8a,10a-anthr), 134.77, 134.91 (CH-6,7-anthr), 136.34 (CH-3-anthr), 146.53 CC-anthr), 96.76 (C-5), 103.33 (C-4a), 127.75 (CH-8-anthr), 127.97 (CH- (CH-6), 153.48 (C-2), 163.87 (C-4), 182.08 (C-10-anthr), 182.36 ppm (C-9- 5-anthr), 128.46 (CH-4-anthr), 128.83 (CH-6), 129.75 (CH-1-anthr), anthr); IR (KBr): n˜ =3417, 2202, 1675, 1659, 1643, 1590, 1558, 1503, 1360, 130.12 (C-2-anthr), 132.67 (C-4a-anthr), 133.65, 133.80, 134.04 (C- 1328, 1290, 1278, 1259, 1203, 1177, 1093, 1053 cmÀ1; MS (ESIÀ): m/z (%): 8a,9a,10a-anthr), 135.66, 135.72 (CH-6,7-anthr), 137.34 (CH-3-anthr), 456 (15) [M+ÀH], 479 (38) [M +ÀH+Na]; HRMS (ESIÀ): calcd for 149.59 (C-7a), 152.77 (CH-2), 157.87 (C-4), 183.81, 184.04 ppm (C-9,10- 31 1 C25H18N3O6: 456.1201; found 456.1197. anthr); P{ H} NMR (202.3 MHz, CD3OD+ D2O): À21.94 (bdd, J =20.2, PAQ PAQ I dA : Compound dA was prepared from dA according to the gener- 17.4 Hz; Pb), À10.27 (d, J =2.02 Hz; Pa), À9.54 ppm (bd, J =17.4 Hz; Pg); + + al procedure (Method A). The product was isolated as an orange solid MS (ESIÀ): m/z (%): 719 (12) [M ÀH], 741 (45) [M À2H+Na]; HRMS 1 (ESIÀ): calcd for C H N O P : 719.03508; found 719.03426. (58 mg, 80%). M.p. 140–158 8C; H NMR (499.8 MHz, [D6]DMSO): 2.17 27 22 4 14 3 EAQ EAQ I (ddd, Jgem =13.1, J2’b,1’ 6.0, J2’b,3’ =2.9 Hz, 1H; H-2’b), 2.46 (ddd, Jgem = dC TP: Compound dC TP was prepared from dC TP according to EAQ 13.1, J2’a,1’ =8.0, J2’a,3’ =5.8 Hz, 1H; H-2’a), 3.50 (ddd, Jgem =11.8, J5’b,OH = the general procedure (Method B) in 31% yield or from dC according

6.0, J5’b,4’ =4.3 Hz, 1H; H-5’b), 3.56 (ddd, Jgem =11.8, J5’a,OH =5.1, J5’a,4’ = to the general procedure (Method C) in 68% yield. The product was iso- 1 4.3 Hz, 1H; H-5’a), 3.81 (td, J4’,5’ =4.3, J4’,3’ = 2.5 Hz, 1 H; H-4’), 4.33 (m, lated as a yellow solid. H NMR (600.1 MHz, CD3OD+ D2O): 2.29 (ddd,

1H; H-3’), 4.39 (d, J =5.2 Hz, 2 H; CH2N), 5.08 (dd, JOH,5’ = 6.0, 5.1 Hz, Jgem =14.0, J2’b,1’ =7.2, J2’b,3’ = 6.2 Hz, 1 H; H-2’b), 2.45 (ddd, Jgem =14.0,

1H; OH-5’), 5.27 (d, JOH,3’ =4.1 Hz, 1 H; OH-3’), 6.47 (dd, J1’,2’ =8.0, J2’a,1’ =6.0, J2’a,3’ =3.8 Hz, 1H; H-2’a), 4.20 (q, J4’,3’ =J4’,5’ =3.8 Hz, 1 H; H-

6.0 Hz, 1H; H-1’), 7.73 (s, 1 H; H-6), 7.96 (m, 2 H; H-6,7-anthr), 8.10 (s, 4’), 4.23 (ddd, Jgem =11.1, JH,P =4.9, J5’b,4’ =3.8 Hz, 1H; H-5’b), 4.30 (ddd,

1H; H-2), 8.24, 8.26 (2 ddd, J =5.7, 3.4, 0.6 Hz, 21 H; H-5,8-anthr), Jgem =11.1, JH,P =6.5, J5’a,4’ =3.8 Hz, 1 H; H-5’a), 4.65 (dt, J3’,2’ = 6.2, 3.8,

8.31 (dd, J4,3 =8.1, J4,1 =0.6 Hz, 1 H; H-4-anthr), 8.39 (dd, J3,4 =8.1, J3,1 = J3’,4’ =3.8 Hz, 1H; H-3’), 6.24 (dd, J1’,2’ =7.2, 6.0 Hz, 1 H; H-1’), 7.91 (m,

1.9 Hz, 1 H; H-3-anthr), 8.69 (dd, J1,3 = 1.9, J1,4 = 0.6 Hz, 1H; H-1-anthr), 2H; H-6,7-anthr), 8.10 (d, J3,4 =8.1 Hz, 1H; H-3-anthr), 8.22–8.28 (m, 13 9.61 ppm (t, J =5.2 Hz, 1H; NH); C NMR (125.7 MHz, [D6]DMSO): 3H; H-4,5,8-anthr), 8.31 (s, 1 H; H-6), 8.38 ppm (s, 1H; H-1-anthr); 13 30.38 (CH2N), 39.79 (CH2-2’), 62.06 (CH2-5’), 71.15 (CH-3’), 75.63 (C5- C NMR (150.9 MHz, CD3OD+D2O): 40.99 (CH2-2’), 66.31 (d, JC,P =

CC-anthr), 83.41 (CH-1’), 87.72 (CH-4’), 88.94 (C5-CC-anthr), 94.75 4.9 Hz; CH2-5’), 71.60 (CH-3’), 85.86 (C5-CC-anthr), 87.18 (d, JC,P = (C-5), 102.54 (C-4a), 125.82 (CH-1-anthr), 126.24 (CH-6), 127.07, 127.10 8.7 Hz; CH-4’), 87.65 (CH-1’), 92.43 (C-5), 95.01 (C5-CC-anthr), 128.13, (CH-5,8-anthr), 127.42 (CH-4-anthr), 133.19 (CH-3-anthr), 133.31, 128.25, 128.51 (CH-4,5,8-anthr), 130.10 (C-2-anthr), 130.60 (CH-1-anthr), 133.35, 133.43 (C-2,8a,10a-anthr), 134.93 (CH-6,7-anthr), 135.06 (C-4a- 133.32 (C-4a-anthr), 134.23, 134.34, 134.36 (C-8a,9a,10a-anthr), 135.96, anthr), 138.84 (C-9a), 149.40 (C-7a), 152.92 (CH-2), 157.72 (C-4), 165.16 136.01 (CH-6,7-anthr), 138.18 (CH-3-anthr), 146.72 (CH-6), 156.52 (C-2), (CONH), 182.37, 182.38 ppm (C-9,10-anthr); IR (KBr): n˜ =3432, 3398, 165.55 (C-4), 184.48, 184.62 ppm (C-9,10-anthr); 31P{1H} NMR

3285, 2231, 1676, 1655, 1626, 1591, 1570, 1533, 1473, 1457, 1325, 1298, (202.3 MHz, CD3OD+D2O): À20.75 (t, J=19.3 Hz; Pb), À9.94 (d, J= À1 + 1283, 1255, 1173, 1093, 1058 cm ; MS (ESI+): m/z (%): 538 (25) [M 19.3 Hz; Pa), À7.67 ppm (bd, J =19.3 Hz; Pg); MS (ESIÀ): m/z (%): 696 + + + +H], 560.2 (100) [M +H+Na]; HRMS (ESI +): calcd for C29H24N5O6 : (23) [M ÀH], 718 (55) [M À2H+Na]; HRMS (ESIÀ): calcd for

538.1721; found 538.1725. C25H21N3O15P3 : 696.01910; found 696.01796. dCPAQ : Compound dCPAQ was prepared from dCI according to the general dAPAQTP: Compound dAPAQTP was prepared from dCITP according to procedure (Method A). The product was isolated as a yellow-orange the general procedure (Method D). The product was isolated as a white 1 1 solid (56 mg, 83%). M.p. 165–1718C; H NMR (499.8 MHz, [D6]DMSO): solid (30 mg, 80 %). H NMR (499.8 MHz, CD3OD +D2O): 2.38 (ddd,

1.97 (ddd, Jgem =13.2, J2’b,1’ =7.2, J2’b,3’ =6.0 Hz, 1H; H-2’b), 2.14 (ddd, Jgem =13.7, J2’b,1’ =6.0, J2’b,3’ = 2.8 Hz, 1 H; H-2’b), 2.58 (ddd, Jgem =13.7,

Jgem =13.2, J2’a,1’ =6.0, J2’a,3’ =3.5 Hz, 1 H; H-2’a), 3.54, 3.62 (2ddd, Jgem = J2’a,1’ =7.9, J2’a,3’ = 6.1 Hz, 1 H; H-2’a), 4.11 (bm, 1H; H-5’b), 4.14–4.23

12.1, J5’,OH =4.8, J5’,4’ =3.6 Hz, 21 H; H-5’), 3.79 (q, J4’,3’ =J4’,5’ =3.6 Hz, (bm, 2H; H-4’,5’a), 4.44 (bs, 2 H; CH2N), 4.97 (bm, 1 H; H-3’), 6.54 (dd,

1H; H-4’), 4.20 (m, 1H; H-3’), 4.39 (d, J =5.1 Hz, 2 H; CH2N), 5.06 (bt, J1’,2’ =7.9, 6.0 Hz, 1H; H-1’), 7.71 (s, 1 H; H-6), 7.90 (m, 2H; H-6,7-anthr),

JOH,5’ =4.8 Hz, 1 H; OH-5’), 5.21 (bd, JOH,3’ =3.8 Hz, 1 H; OH-3’), 6.11 8.05 (s, 1 H; H-2), 8.20–8.30 (m, 3 H; H-3,5,8-anthr), 8.32 (d, J4,3 =8.0 Hz, 13 (dd, J1’,2’ = 7.2, 6.0 Hz, 1H; H-1’), 6.90, 7.84 (2 bs, 2 1H; NH2), 7.97 (m, 1H; H-4-anthr), 8.62 ppm (bs, 1 H; H-1-anthr); C NMR (125.7 MHz,

2H; H-6,7-anthr), 8.16 (s, 1H; H-6), 8.24, 8.26 (2 ddd, J=5.7, 3.4, CD3OD+D2O): 31.47 (CH2N), 40.43 (CH2-2’), 66.70 (b, CH2-5’), 72.47

0.6 Hz, 2 1H; H-5,8-anthr), 8.31 (dd, J4,3 = 8.1, J4,1 = 0.6 Hz, 1H; H-4- (CH-3’), 76.23 (C5-CC-anthr), 84.24 (CH-1’), 86.74 (d, JC,P =8.0 Hz; anthr), 8.36 (dd, J3,4 =8.1, J3,1 =1.8 Hz, 1H; H-3-anthr), 8.67 (dd, J1,3 =1.8, CH-4’), 89.17 (C5-CC-anthr), 97.32 (C-5), 103.99 (C-4a), 127.01 (CH-1- 13 J1,4 = 0.6 Hz, 1H; H-1-anthr), 9.44 ppm (t, J=5.2 Hz, 1H; NH); C NMR anthr), 127.32 (CH-6), 128.17, 128.24 (CH-5,8-anthr), 128.80 (CH-4-

(125.7 MHz, [D6]DMSO): 30.37 (CH2N), 41.00 (CH2-2’), 61.25 (CH2-5’), anthr), 133.97 (CH-3-anthr), 134.15, 134.19, 134.52 (C-2,8a,10a-anthr), 70.39 (CH-3’), 74.77 (C5-CC-anthr), 85.58 (CH-1’), 87.69 (CH-4’), 89.56 136.08, 136.11 (CH-6,7-anthr), 136.25 (C-4a-anthr), 139.86 (C-9a-anthr), (C-5), 92.00 (C5-CC-anthr), 125.83 (CH-1-anthr), 127.08, 127.11 (CH- 149.71 (C-7a), 153.17 (CH-2), 158.53 (C-4), 168.67 (CONH), 184.28, 31 1 5,8-anthr), 127.44 (CH-4-anthr), 133.16 (CH-3-anthr), 133.31, 133.35, 184.40 ppm (C-9,10-anthr); P{ H} NMR (202.3 MHz, CD3OD+D2O):

133.41 (C-2,8a,10a-anthr), 134.94 (CH-6,7-anthr), 135.04 (C-4a-anthr), À21.18 (bdd, J= 18.9, 16.5 Hz; Pb), À10.00 (d, J= 18.9 Hz; Pa), + 138.88 (C-9a), 144.23 (CH-6), 153.62 (C-2), 164.61 (C-4), 164.87 À8.73 ppm (bd, J =16.5 Hz; Pg); MS (ESIÀ): m/z (%): 776 (14) [M + (CONH), 182.38 ppm (C-9,10-anthr); IR (KBr): n˜ = 3410, 2927, 2231, ÀH], 798 (12) [M À2H+Na]; HRMS (ESIÀ): calcd for C29H25N5O15P3 : 1676, 1646, 1593, 1533, 1505, 1481, 1416, 1360, 1327, 1283, 1250, 1177, 776.05655; found 776.05532. À1 + 1093, 1053 cm ; MS (ESIÀ): m/z (%): 513 (100) [M ÀH]; HRMS dCPAQTP: Compound dCPAQTP was prepared from dCITP according to (ESIÀ): calcd for C27H21N4O7: 513.14157; found 513.14143. the general procedure (Method D). The product was isolated as a white EAQ EAQ I 1 dA TP: Compound dA TP was prepared from dA TP according to solid (27 mg, 79 %). H NMR (499.8 MHz, CD3OD +D2O): 2.24, 2.40 (2 the general procedure (Method B) in 30% yield or from dAEAQ accord- bm, 2 1H; H-2’), 4.15 (bm, 1H; H-4’), 4.22 (bm, 2H; H-5’), 4.45 (bs, ing to the general procedure (Method C) in 65% yield. The product was 2H; CH2N), 4.60 (bm, 1 H; H-3’), 6.21 (bt, J1’,2’ = 6.0 Hz, 1H; H-1’), 7.85 1 isolated as a dark brown solid. H NMR (499.8 MHz, CD3OD+D2O): (bm, 2H; H-6,7-anthr), 8.10–8.18 (bm, 2 H; H-5,8-anthr), 8.18–8.29 (bm, 13 1.29 (t, Jvic =7.2 Hz, 18H; CH3CH2N), 2.45 (ddd, Jgem =13.8, J2’b,1’ =5.9, 3H; H-6, H-3,4-anthr), 8.49 ppm (bs, 1 H; H-1-anthr); C NMR

J2’b,3’ =3.1 Hz, 1H; H-2’b), 2.60 (ddd, Jgem =13.8, J2’a,1’ = 8.0, J2’a,3’ =5.6 Hz, (125.7 MHz, CD3OD+ D2O): 31.36 (CH2N), 40.84 (CH2-2’), 66.18 (b,

Chem. Eur. J. 2011, 17, 14063 – 14073 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.chemeurj.org 14071 M. Hocek, M. Fojta et al.

CH2-5’), 71.32 (CH-3’), 74.40 (C5-CC-anthr), 86.86 (b, CH-4’), 87.34 urea at 25 W for 50 min. Gels were dried, autoradiographed, and visual- (CH-1’), 92.73 (C-5, C5-CC-anthr), 127.04 (CH-1-anthr), 128.14, 128.19 ized using a phosphorimager. (CH-5,8-anthr), 128.79 (CH-4-anthr), 133.98 (CH-3-anthr), 134.01, Melting temperatures: The oligonucleotides for these measurements 134.03, 134.32 (C-2,8a,10a-anthr), 136.07, 136.10 (CH-6,7-anthr), 136.12 were prepared by PEX on a large scale with KOD XL DNA as poly- (C-4a-anthr), 139.58 (C-9a-anthr), 145.99 (CH-6), 156.69 (C-2), 166.14 merase, templates tempA, tempC, and temprnd16, and primrnd as primer. For (C-4), 168.34 (CONH), 184.25, 184.41 ppm (C-9,10-anthr); 31P{1H} NMR preparative purposes, a total volume of 500 mL PEX with higher concen- (202.3 MHz, CD3OD +D2O): À21.08 (b, Pb), À10.17 (b, Pa), À8.24 ppm trations of primer (10 mm) and template (10 mm) was run and purification + + (b, Pg); MS (ESIÀ): m/z (%): 753 (5) [M ], 775 (22) [M ÀH+Na]; was carried out with QIAquick Nucleotide Removal Kit (Qiagen). Sam-

HRMS (ESIÀ): calcd for C27H24N4O16P3: 753.04056; found 753.04006. ples were eluted with H2O (100 mL, pH 7.5) and then freeze-dried. DNA Materials for biochemistry: Synthetic oligonucleotides were purchased duplexes were first dissolved in phosphate buffer (160 mL, 50 mm, pH 6.7) from VBC Genomics (Austria). Dynabeads M-270 Streptavidin (DBStv) and further diluted with the buffer to optimum concentration OD260 be- were obtained from Dynal A.S. (Norway), Vent (exo-), Pwo, polymerases, tween 0.08 and 0.1. Thermal denaturation studies were performed on a and T4 polynucleotide kinase were from New England Biolabs (Great Cary 100 Bio (UV–visible spectrometer with temperature controller, Britain), KOD XL DNA from Novagen, unmodified nucleoside triphos- Varian). Data were obtained from six individual cooling–heating cycles. phates (dATP, dTTP, dCTP, and dGTP) from Sigma, and g-32P-ATP from Melting temperatures (Tm values in 8C) were obtained by plotting tem- MP Empowered Discovery (USA). Other chemicals were of analytical perature versus absorbance and by applying a sigmoidal curve fit. grade. Isolation of single-strand oligonucleotides by the DBStv magnetosepara- m Primer extension experiment—single incorporation (Figure 1): The reaction procedure: Reaction mixture (50 mL) containing 0.3 NaCl was tion mixture (20 mL) contained KOD XL DNA polymerase (2.5 U mLÀ1, added to DBStv [25 mL of stock solution washed three times by 150 mL m m 0.02 mL), dGTP (natural, 4 mm, 0.05 mL), dNXAQTP, dATP (dCTP) (4 mm, of buffer (0.3 NaCl, 10 m Tris, pH 7.4)]. The suspension was shaken at 1mL), primer (3 mm,1mL, primrnd :3’-GGGTACGGCGGGTAC-5’), and room temperature to allow the oligonucleotides to bind to the DBStv A beads. The DBStv beads were washed three times with phosphate-buf- 19-mer template (3 mm, 1.5 mL, temp :5’-CCCTCCCATGCCGCC- fered saline (PBS) solution (200 mL, 0.14m NaCl, 3 mm KCl, 4 mm CATG-3’ or tempC :5’-CCCGCCCATGCCGCCCATG-3’) in KOD XL sodium phosphate, pH 7.4) with 0.01 % Tween 20 and then three times by reaction buffer (1 mL) supplied by the manufacturer. Primrnd was labeled buffer (200 mL, 0.3m NaCl, 10 mm Tris, pH 7.4) and finally by doubly dis- by use of [g32P]-ATP according to standard techniques. Reaction mixtures tilled H O (200 mL). Single-strand oligonucleotides were released by were incubated for 15 min at 60 8C in a thermal cycler and were stopped 2 shaking and heating the sample at 758C for 2 min. Each medium ex- by addition of stop solution (40 mL, 80% [v/v] formamide, 10 mm ethyle- change was performed by using a magnetoseparator (Dynal, Norway). nediaminetetraacetic acid (EDTA), 0.025 % [w/v] bromophenol blue, 0.025% [w/v] xylene cyanol) and heated for 5 min at 958C. Reaction mix- Electrochemical analysis: Nucleosides, deoxynucleoside monophosphates tures were separated by use of a 12.5% denaturing PAGE. Visualization (dNMPs), and other building blocks were analyzed by conventional in was performed by phosphoimaging. situ CV. PEX products were analyzed by ex situ (adsorptive transfer stripping) CV or SWV. The PEX products were accumulated for 60 s Kinetics of PEX: The PEX reaction mixtures using KOD XL DNA poly- from aliquots (5 mL) containing 0.2 m NaCl at the surface of the working merase were incubated for time intervals (0.1–15 min), followed by stop- electrode (HMDE or basal-plane PGE). The electrode was then rinsed ping the reaction by addition of PAGE loading buffer and immediate with deionized water and placed in the electrochemical cell. CV settings: heating. scan rate 0.5 V sÀ1, initial potential 0.0 V, for switching and final poten- Primer extension experiment—multiple incorporation (Figure 3): The re- tials see figure legends. SWV settings: initial potential À0.6 V, final po- action mixture (20 mL) contained KOD XL DNA polymerase tential À1.6 V, frequency 200 Hz, amplitude 25 mV. Background electro- (2.5 UmLÀ1, 0.3 mL), dNTPs [in samples 2–8: (4 mm,1mL); in samples 9– lyte: 0.3 m ammonium formate, 0.05m sodium phosphate, pH 6.6 (for 12: dNTPs (4 mm,1mL), dNXAQTPs (40 mm,1mL)], primer (3 mm,1mL, measurements at HMDE) or 0.2 m sodium acetate pH 5.0 (for measure- primrnd :3’-GGGTACGGCGGGTAC-5’), and 31-mer template (3 mm, ments at PGE). All measurements were performed at room temperature 1.5 mL, temprnd16 :5’-CTAGCATGAGCTCAGTCCCATGCCGCCCATG- by using an Autolab analyzer (Eco Chemie, The Netherlands) in connec- 3’) in KOD XL reaction buffer (1 mL) supplied by the manufacturer. tion with VA-stand 663 (Metrohm, Herisau, Switzerland). The three-elec- Primrnd was labeled by use of [g32P]-ATP according to standard tech- trode system was used with an Ag/AgCl/3 m KCl electrode as a reference niques. Reaction mixtures were incubated for 30 min at 60 8C in a ther- and platinum wire as an auxiliary electrode. Measurements at the mal cycler and were stopped by addition of stop solution (40 mL, 80% [v/ HMDE were performed after deaeration of the solution by argon purg- v] formamide, 10 mm EDTA, 0.025% [w/v] bromophenol blue, 0.025% ing. [w/v] xylene cyanol) and heated for 5 min at 958C. Reaction mixtures were separated by use of a 12.5 % denaturing PAGE. Visualization was performed by phosphoimaging. Primer extension experiment—multiple incorporation (Figure 4): The reaction mixture (20 mL) contained KOD XL DNA polymerase Acknowledgements (2.5 UmLÀ1, 0.3 mL), dNTPs (4 mm,1mL), primer (3 mm,1mL, primrnd :3’- m GGGTACGGCGGGTAC-5’), and 30-mer template (3 m , 1.5 mL, tem- This work was supported by the Academy of Sciences of the Czech Re- A4line p :5’-CAGCAGCAGCATTTTCCCATGCCGCCCATG-3’ or public (Z4 055 0506, Z5004 0507, and Z5 0040702), the Ministry of Edu- C4line temp :5’-TCATCATCATAGGGGCCCATGCCGCCCATG-3’)in cation (LC06035, LC512), Grant Agency of the Academy of Sciences of rnd KOD XL buffer (1 mL) supplied by the manufacturer. Prim was labeled the Czech Republic (IAA400040901), and by Gilead Sciences, Inc. 32 by use of [g P]-ATP according to standard techniques. Reaction mixtures (Foster City, CA, USA.). were incubated for 30 min at 60 8C in a thermal cycler and were stopped by addition of stop solution (40 mL, 80% [v/v] formamide, 10 mm EDTA, 0.025% [w/v] bromophenol blue, 0.025 % [w/v] xylene cyanol) and [1] a) M. J. Heller, Annu. Rev. Biomed. Eng. 2002, 4, 129 –153; b) A. heated for 5 min at 958C. Reaction mixtures were separated by use of a Sassolas, B. D. Leca-Bouvier, L. J. Blum, Chem. Rev. 2008, 108, 109– 12.5% denaturing PAGE. Visualization was performed by phosphoimag- 139. ing. [2] a) E. Palecˇek, F. Jelen in Electrochemistry of Nucleic Acids and Pro- PAGE : The PEX products were mixed with loading buffer (80% forma- teins: Towards Electrochemical Sensors for Genomics and Proteo- mide, 10 mm EDTA, 1 mg mLÀ1 xylene cyanol, 1 mgmLÀ1 bromophenol mics (Eds.: E. Palecˇek, F. Scheller, J. Wang), Elsevier, Amsterdam, blue) and subjected to electrophoresis in 12.5% denaturing polyacryl- 2005, pp. 74– 174; b) J. Wang in Electrochemistry of Nucleic Acids amide gel containing 1 Tris/borate/EDTA (TBE) buffer (pH 8) and 7 m and Proteins: Towards Electrochemical Sensors for Genomics and

14072 www.chemeurj.org 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Eur. J. 2011, 17, 14063 – 14073 Anthraquinone as a Redox Label for DNA FULL PAPER

Proteomics (Eds.: E. Palecˇek, F. Scheller, J. Wang), Elsevier, Am- [16] a) C. C. Bjorklund, W. B. Davis, Biochemistry 2007, 46, 10745– sterdam, 2005, pp. 175 –194. 10755; b) K. E. Augustyn, E. J. Merino, J. K. Barton, Proc. Natl. [3] T. G. Drummond, M. G. Hill, J. K. Barton, Nat. Biotechnol. 2003, 21, Acad. Sci. USA 2007, 104, 18907– 18912; c) T. T. Williams, C. 1192– 1199. Dohno, E. D. A. Stamp, J. K. Barton, J. Am. Chem. Soc. 2004, 126, [4] Review: M. Hocek, M. Fojta, Org. Biomol. Chem. 2008, 6, 2233– 8148– 8158. 2241. [17] A. Suzuki, M. Hasegawa, M. Ishii, S. Matsumura, K. Toshima, [5] P. Brzdilov, M. Vrbel, R. Pohl, H. Pivonˇ kov, L. Havran, M. Bioorg. Med. Chem. Lett. 2005, 15, 4624 –4627. Hocek, M. Fojta, Chem. Eur. J. 2007, 13, 9527– 9533. [18] a) G. I. Giles, R. P. Sharma, J. Pept. Sci. 2005, 11, 417 –423; b) E. De [6] H. Cahov, L. Havran, P. Brzdilov, H. Pivonˇ kov, R. Pohl, M. Moliner, S. Moro, S. Sarno, G. Zagotto, G. Zanotti, L. A. Pinna, R. Fojta, M. Hocek, Angew. Chem. 2008, 120, 2089–2092; Angew. Battistutta, J. Biol. Chem. 2003, 278, 1831–1836. Chem. Int. Ed. 2008, 47, 2059– 2062. [19] a) T. Gourlain, A. Sidorov, N. Mignet, S. J. Thorpe, S. E. Lee, J. A. [7] J. Riedl, P. Horkov, P. Sˇ ebest, R. Pohl, L. Havran, M. Fojta, M. Grasby, D. M. Williams, Nucleic Acids Res. 2001, 29, 1898 –1905; Hocek, Eur. J. Org. Chem. 2009, 3519 –3525. b) H. A. Held, S. A. Benner, Nucleic Acids Res. 2002, 30, 3857 – [8] M. Vrbel, P. Horkov, H. Pivonˇ kov, L. Kalachova, H. Cˇ ernock, 3869; c) S. Jger, G. Rasched, H. Kornreich-Leshem, M. Engeser, O. Thum, M. Famulok, J. Am. Chem. Soc. , 127, 15071 –15082; H. Cahov, R. Pohl, P. Sˇ ebest, L. Havran, M. Hocek, M. Fojta, 2005 d) M. Kuwahara, J. Nagashima, M. Hasegawa, T. Tamura, R. Kitaga- Chem. Eur. J. 2009, 15, 1144– 1154. ta, K. Hanawa, S. Hososhima, T. Kasamatsu, H. Ozaki, H. Sawai, [9] a) D. Ajloo, B. Yoonesi, A. Soleymanpour, Int. J. Electrochem. Sci. Nucleic Acids Res. 2006, 34, 5383– 5394; e) P. M. E. Gramlich, S. 2010, 5, 459– 477; b) M. Quan, D. Sanchez, M. F. Wasylkiv, D. K. Warncke, J. Gierlich, T. Carell, Angew. Chem. 2008, 120, 3491 –3493; Smith, J. Am. Chem. Soc. 2007, 129, 12847– 12856; c) C. Batchelor- Angew. Chem. Int. Ed. 2008, 47, 3442– 3444; f) V. Borsenberger, M. McAuley, Q. Li, S. M. Dapin, R. G. Compton, J. Phys. Chem. B Kukwikila, S. Howorka, Org. Biomol. Chem. 2009, 7, 3826 –3835; 2010, 114, 4094 –4100; d) C. Batchelor-McAuley, B. R. Kozub, D. g) S. Obeid, M. Yulikow, G. Jeschke, A. Marx, Angew. Chem. 2008, Menshykau, R. G. Compton, J. Phys. Chem. C 2011, 115, 714 –718. 120, 6886 –6890; Angew. Chem. Int. Ed. 2008, 47, 6782– 6785. [10] a) S. D. Wettig, G. A. Bare, R. J. S. Skinner, J. S. Lee, Nano Lett. [20] P. Cˇ apek, H. Cahov, R. Pohl, M. Hocek, C. Gloeckner, A. Marx, 2003, 3, 617–622; b) S. Mahajan, J. Richardson, N. Ben Gaied, Z. Chem. Eur. J. 2007, 13, 6196– 6203. Zhao, T. Brown, P. N. Barlett, Electroanalysis 2009, 21, 2190 –2197; [21] T. Kovacs, L. tvçs, Tetrahedron Lett. 1988, 29, 4525– 4528. c) Y. J. Zhang, X. P. He, M. Hu, Z. Li, X. X. Shi, G. R. Chen, Dyes [22] a) P. Manisankar, A. Gomathi, D. Velayutham, J. Solid State Electro- Pigments 2011, 88, 391– 395. chem. 2005, 9, 601–608; b) J. Vacek, L. Havran, M. Fojta, Electroa- [11] a) R. Abou-Elkhair, T. Netzel, Nucleosides Nucleotides Nucleic nalysis 2009, 21, 2139–2144; c) D. A. Di Giusto, W. A. Wlassoff, S. Acids 2005, 24, 85– 110; b) R. A. I. Abou-Elkhair, D. W. Dixon, Giesebrecht, J. J. Gooding, G. C. King, J. Am. Chem. Soc. 2004, 126, T. L. Netzel, J. Org. Chem. 2009, 74, 4712 –4719. 4120– 4121. [12] M. Tierney, M. Grinstaff, Org. Lett. 2000, 2, 3413 –3416. [23] M. Hocek, P. Stepnicka, J. Ludvik, I. Cisarova, I. Votruba, D. Reha, [13] M. F. Jacobsen, E. E. Ferapontova, K. V. Gothelf, Org. Biomol. P. Hobza, Chem. Eur. J. 2004, 10, 2058 –2066. Chem. 2009, 7, 905– 908. [24] a) P. Zuman, Collect. Czech. Chem. Commun. 1993, 58,41–46;b)K. [14] A. A. Gorodetsky, O. Green, E. Yavin, J. K. Barton, Bioconjugate Peckov, J. Barek, T. Navratil, B. Yosypchuk, J. Zima, Anal. Lett. Chem. 2007, 18, 1434–1441. 2009, 42, 2339–2363. [15] Z. Zhao, G. Peng, J. Michels, K. R. Fox, T. Brown, Nucleosides, Nu- Received: March 28, 2011 cleotides Nucleic Acids 2007, 26, 921–925. Published online: November 17, 2011

Chem. Eur. J. 2011, 17, 14063 – 14073 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.chemeurj.org 14073 Article

pubs.acs.org/joc

Aqueous Heck Cross-Coupling Preparation of Acrylate-Modiﬁed Nucleotides and Nucleoside Triphosphates for Polymerase Synthesis of Acrylate-Labeled DNA † ‡ † ‡ ‡,§ ,†,∥ Jitka Dadova,́Pavlína Vidlaková ,́Radek Pohl, Ludeǩ Havran, Miroslav Fojta, and Michal Hocek* † Institute of Organic Chemistry and Biochemistry, Academy of Sciences of the Czech Republic, Gilead Sciences & IOCB Research Center, Flemingovo nam.́ 2, CZ-16610 Prague 6, Czech Republic ‡ Institute of Biophysics, v.v.i., Academy of Sciences of the Czech Republic, Kralovopolska 135, CZ-61265 Brno, Czech Republic § Central European Institute of Technology, Masaryk University, Kamenice 753/5, CZ-625 00 Brno, Czech Republic ∥ Department of Organic Chemistry, Faculty of Science, Charles University in Prague, Hlavova 8, CZ-12843 Prague 2, Czech Republic

*S Supporting Information

ABSTRACT: Aqueous-phase Heck coupling methodology was developed for direct attachment of butyl acrylate to 5-iodoracil, 5- iodocytosine, 7-iodo-7-deazaadenine, and 7-iodo-7-deazaguanine 2′- deoxyribonucleoside 5′-O-monophosphates (dNMPs) and 5′-O- triphosphates (dNTPs) and compared with the classical approach of phosphorylation of the corresponding modified nucleosides. The 7- substituted 7-deazapurine nucleotides (dABAMP, dABATP, dGBAMP, and dGBATP) were prepared by the direct Heck coupling of nucleotides in good yields (35−55%), whereas the pyrimidine nucleotides reacted poorly and the corresponding BA-modified dNTPs were prepared by triphosphorylation of the modified nucleosides. The acrylate-modified dNBATPs (N = A, C, and U) were good substrates for DNA polymerases and were used for enzymatic synthesis of acrylate-modified DNA by primer extension, whereas dGBATP was an inhibitor of polymerases. The butyl acrylate group was found to be a useful redox label giving a strong reduction peak at −1.3 to −1.4 V in cyclic voltammetry.

■ INTRODUCTION first direct Sonogashira coupling of 5-iodoracil dNTP.5 In our Base-modified 2′-deoxyribonucleoside triphosphates (dNTPs) laboratory, the aqueous Suzuki coupling of halogenated fi nucleotides and dNTPs with diverse boronic acids was bearing chemical modi cations at position 5 of pyrimidines or 16 at position 7 of 7-deazapurines are generally good substrates for developed. More recently, the Suzuki reactions of oligonucleotides have also been reported.17 We extensively use both DNA polymerases. Diverse protocols for enzymatic synthesis of fi base-modified DNA have been developed and have been Suzuki and Sonogashira couplings for modi cation of extensively used in the past decade.1 The methods include pyrimidine and 7-deazapurine dNTPs linked through aryl 2 fi groups (by the Suzuki coupling) or through the alkynyl moiety primer extension (PEX) or PCR, site-speci c single-nucleotide 6,7,10,11 incorporation,3 or nicking-enzyme amplification reaction (by the Sonogashira coupling). Both these reactions are 4 fl 5,6 7 8 performed in water/acetonitrile mixtures using triphenylphos- (NEAR). The applications cover uorescent, redox, spin, ′ ″ barcode,9 and reactive10 labeling, protection,11 and incorpo- phine-3,3 ,3 -trisulfonate (TPPTS) as water-soluble ligand for ration of protein-like groups for catalysis.12 The modified palladium catalyst. Both these reactions are very tolerant to the presence of large variety of functional groups and allow a single- dNTPs are usually synthesized by triphosphorylation of fi modified nucleosides which is laborious and sometimes step preparation of modi ed dNTPs without the need of incompatible with the introduced functionality. protection group manipulation. Development of other aqueous Aqueous-phase cross-coupling reactions are an increasingly coupling reactions applicable for other types of carbon fi 13 substituents is still highly desirable to extend the portfolio of popular and useful tool in modi cations of polar molecules fi and are of particular importance for biomolecules. The very first bioorthogonal modi cations of nucleic acids. ′ The Heck reaction is another very useful type of cross- report was on the Sonogashira coupling of 5-iodo-2 - 18 deoxyuridine and dUMP with propargylamine by Casalnuo- coupling extensively applied in attachment of alkenyl groups. vo.14 Later on, the Shaughnessy group has developed the first aqueous Suzuki−Miyaura coupling of unprotected halogenated Received: May 28, 2013 purine nucleosides,15 whereas the Burgess group reported the Published: August 30, 2013

Scheme 1. Synthesis of Butyl Acrylate Modiﬁed Nucleosides (dNBAs), Nucleoside Mono- (dNBAMPs), and Triphosphates a (dNBATPs)

a Reagents and conditions: (i) butyl acrylate, Pd(OAc)2, PPh3,Et3N, DMF; (ii) butyl acrylate, Pd(OAc)2, TPPTS, Et3N, CH3CN/H2O (1:1); (iii) ° ° ° PO(OMe)3, POCl3,0 C; (iv) (1) PO(OMe)3, POCl3,0 C, (2) (NHBu3)2H2P2O7,Bu3N, DMF, 0 C, (3) 2 M TEAB; (v) butyl acrylate, Pd(OAc)2, TPPTS, Et3N, CH3CN/H2O (1:1).

Table 1. Synthesis of Butyl Acrylate Modiﬁed Nucleosides and Nucleotides

entry starting compd product catalyst additives solvent yield (%) I BA 1 dC dC Pd(OAc)2, PPh3 Et3N DMF 14 I BA 2 dU dU Pd(OAc)2, PPh3 Et3N DMF 93 I BA 3 dA dA Pd(OAc)2, PPh3 Et3N DMF 97 I BA 4 dG dG Pd(OAc)2, PPh3 Et3N DMF 83 I BA 5 dC dC Pd(OAc)2, TPPTS Et3NCH3CN/H2O (1:1) 16 I BA 6 dU dU Pd(OAc)2, TPPTS Et3NCH3CN/H2O (1:1) 98 I BA 7 dA dA Pd(OAc)2, TPPTS Et3NCH3CN/H2O (1:1) 81 I BA 8 dG dG Pd(OAc)2, TPPTS Et3NCH3CN/H2O (1:1) 84 I BA 9 dC MP dC MP Pd(OAc)2, TPPTS Et3NCH3CN/H2O (1:1) - I BA 10 dU MP dU MP Pd(OAc)2, TPPTS Et3NCH3CN/H2O (1:1) 35 I BA 11 dA MP dA MP Pd(OAc)2, TPPTS Et3NCH3CN/H2O (1:1) 55 I BA 12 dG MP dG MP Pd(OAc)2, TPPTS Et3NCH3CN/H2O (1:1) 38 I BA 13 dC TP dC TP Pd(OAc)2, TPPTS Et3NCH3CN/H2O (1:1) - I BA 14 dU TP dU TP Pd(OAc)2, TPPTS Et3NCH3CN/H2O (1:1) 4 I BA 15 dU TP dU TP Pd(OAc)2, PPh3 Et3N DMF 14 I BA 16 dA TP dA TP Pd(OAc)2, TPPTS Et3NCH3CN/H2O (1:1) 43 I BA 17 dG TP dG TP Pd(OAc)2, TPPTS Et3NCH3CN/H2O (1:1) 44 BA BA ° 18 dU dU MP PO(OMe)3, POCl3,0 C42 BA BA ° 19 dA dA MP PO(OMe)3, POCl3,0 C54 BA BA ° ° 20 dC dC TP (1) PO(OMe)3, POCl3,0 C; (2) (NHBu3)2H2P2O7,Bu3N, DMF, 0 C; (3) 19 2 M TEAB BA BA ° ° 21 dU dU TP (1) PO(OMe)3, POCl3,0 C; (2) (NHBu3)2H2P2O7,Bu3N, DMF, 0 C; (3) 40 2 M TEAB BA BA ° ° 22 dA dA TP (1) PO(OMe)3, POCl3,0 C; (2) (NHBu3)2H2P2O7,Bu3N, DMF, 0 C; (3) 28 2 M TEAB BA BA ° ° 23 dG dG TP (1) PO(OMe)3, POCl3,0 C; (2) (NHBu3)2H2P2O7,Bu3N, DMF, 0 C; (3) 25 2 M TEAB

In nucleoside chemistry it has been often used for modifications nucleoside mono- and triphosphates and the use of acrylate- of pyrimidines19 but is difficult for modifications of purines.20 modified dNTPs for enzymatic synthesis of modified DNA. Only very recently, Shaughnessy reported21 the first aqueous- phase Heck reactions for modification of 5-iodo-2′-deoxyur- ■ RESULTS AND DISCUSSION idine. Here we report on the development of the aqueus Heck Synthesis of Acrylate-Modified Nucleosides and coupling for modifications of pyridimine and 7-deazapurine Nucleotides by Aqueous Heck Coupling. The acrylate

9628 dx.doi.org/10.1021/jo4011574 | J. Org. Chem. 2013, 78, 9627−9637 The Journal of Organic Chemistry Article

a Table 2. List of Oligodeoxyribonucleotides Used or Synthesized

oligonucleotide sequence prim 5′-CATGGGCGGCATGGG-3′ temp1C 5′-CCCGCCCATGCCGCCCATG-3′ temp1T 5′-CCCACCCATGCCGCCCATG-3′ temp1A 5′-CCCTCCCATGCCGCCCATG-3′ temp1G 5′-AAACCCCATGCCGCCCATG-3′ temp4 5′-CTAGCATGAGCTCAGTCCCATGCCGCCCATG-3′ ON1C 5′-CATGGGCGGCATGGGCBAGGG-3′ ON1T 5′-CATGGGCGGCATGGGUBAGGG-3′ ON1A 5′-CATGGGCGGCATGGGABAGGG-3′ ON1G 5′-CATGGGCGGCATGGGGBATTT-3′ ON4C 5′-CATGGGCGGCATGGGACBATGAGCBATCBAATGCBATAG-3′ ON4T 5′-CATGGGCGGCATGGGACUBAGAGCUBACAUBAGCUBAAG-3′ ON4A 5′-CATGGGCGGCATGGGABACTGABAGCTCABATGCTABAG-3′ ON4G 5′-CATGGGCGGCATGGGACUGBAAGBACUCAUGBACUAGBA-3′ aIn the template (temp) ONs the segment-forming duplex with the primer are underlined and the replicated segments are in bold. For magnetic separation of the extended primer strands, the templates were 5′-end biotinylated. Acronyms used in the text for primer extension products are analogous to those introduced for the templates (e.g., ssDNA PEX product ON1A was synthesized on temp1A etc.). ester moiety is an attractive label for DNA since it can be in nucleoside triphosphates (dCITP, dUITP, dAITP, and dGITP, principle further derivatized by amide formation, conjugate Scheme 1). To minimize hydrolysis of triphosphate group, the additions, reductions, etc. and also could be reduced on reaction time was shortened to 1 h under the same conditions electrode to serve as redox label for electrochemistry. Jager̈ et as above to reach ca. 50% conversion. The deazapurine dNTPs al.2 prepared methyl acrylate-modified dUTP by triphosphor- reacted well to obtain dABATP and dGBATP in acceptable ylation of nucleoside and showed its successful incorporation yields of 43% and 44%, respectively (Table 1, entries 16 and by Vent(exo-) polymerase but never reported any further use. 17). On the other hand, pyrimidine dNTPs reacted poorly. We wanted to study the aqueous Heck reaction of iodinated dCBATP was not detected in the reaction mixture, whereas pyrimidine and deazapurine nucleotides and dNTPs. However, dUBATP was isolated in 4% yield only due to substantial at first we tested the Heck reactions of 2′-deoxy-5-iodocytidine hydrolysis of both starting dUITP and the product (Table 1, (dCI), 2′-deoxy-5-iodoridine (dUI), 2′-deoxy-7-iodo-7-deazaa- entries 13 and 14). To prevent the hydrolysis, the Heck denosine (dAI) and 2′-deoxy-7-iodo-7-deazaguanosine (dGI) coupling of dUITP was also performed in DMF under the nucleosides with n-butyl acrylate (6 equiv) in the presence of conditions described above for the synthesis of modified Pd(OAc)2 (10 mol %), triphenylphosphine, and triethylamine nucleosides. The conversion increased to ca. 80%, but the in DMF (Scheme 1). While the reactions with dUI, dAI, and isolated yield of dUBATP was 14% (Table 1, entry 15) and dGI proceeded almost quantitatively to give the desired substantive amounts of the corresponding mono- and acrylate-modified nucleosides in excellent yields (93% dUBA, diphosphate were observed. It indicates that dUBATP was 97% dABA, and 83% dGBA; Table 1, entries 2−4), dCI showed largely hydrolyzed during the isolation and purification process. much lower reactivity and the desired dCBA was isolated in For comparison of the efficiency of the aqueous-phase direct poor yield (14%; Table 1, entry 1). Then, for comparison, modification of nucleotides with the classical approach, the aqueous Heck reactions of iodinated nucleosides with n-butyl acrylate modified nucleosides (dCBA, dUBA, dABA, and dGBA) acrylate were carried out in the mixture water/acetonitrile (1:1) were phosphorylated to obtain dNBAMPsanddNBATPs in the presence of water-soluble phosphine ligand tris(3- (Scheme 1, Table 1). The treatment of dUBA or dABA with ° sulfonatophenyl)phosphine (TPPTS, Scheme 1). The results POCl3 in PO(OMe)3 at 0 C followed by quenching the were comparable to the DMF protocol in all cases. The reaction with triethylammonium bicarbonate (TEAB, 2M) gave couplings on dUI, dAI ,and dGI proceeded smoothly with good the desired modified dNBAMPs in acceptable isolated yields isolated yields of modified nucleosides (98% dUBA, 81% dABA, (42% for dUBAMP and 54% for dABAMP). Triphosphorylation and 84% dGBA; Table 1, entries 6−8), whereas dCI gave dCBA of nucleosides dCBA, dUBA, dABA, and dGBA was performed 22 in an isolated yield of 16% (Table 1, entry 5). under standard conditions by treatment with POCl3 in ° Then the aqueous conditions were tested with halogenated PO(OMe)3 at 0 C followed by addition of (NHBu3)2H2P2O7 nucleoside monophosphates (dNMPs) which are relatively in DMF in the presence of tributylamine. The reaction was stable and water-soluble compounds suitable as models for quenched with TEAB and after purification, dNBATPs were nucleic acids (Scheme 1). Thus, the aqueous Heck coupling of obtained in acceptable yields (19% for dCBATP, 40% for dCIMP, dUIMP, dAIMP, and dGIMP with n-butyl acrylate (the dUBATP, 28% for dABATP, and 25% for dGBATP). excess was increased to 10 equiv) were performed in the The yields of both approaches were then compared. The ffi presence of Pd(OAc)2 (10 mol % were necessary) and TPPTS e ciency of direct aqueous Heck reaction of halogenated (25 mol %). The reactions proceeded less efficiently (compared nucleotides was comparable with phosphorylation in the case of to nucleosides) with ca. 60−70% conversion to give modified both tested monophosphates (dUBAMP and dABAMP) giving nucleotides dUBAMP, dABAMP, and dGBAMP in isolated yields yields of 35−55%. It was even more efficient for the synthesis of 35%, 55%, and 38%, respectively (Table 1, entries 10−12). of deazapurine dNTPs: dABATP (43% compared to 28% yield The formation of dCBAMP was not observed (Table 1, entry of triphosphorylation) and dGBATP (44% compared to 25% 9). Finally, the aqueous phase Heck coupling was performed on yield of triphosphorylation). However, in the preparation of

9629 dx.doi.org/10.1021/jo4011574 | J. Org. Chem. 2013, 78, 9627−9637 The Journal of Organic Chemistry Article

Figure 1. Primer extension with (a) temp1C; (b) temp1T; (c) temp1A; (d) temp1G. Key: P, primer; C+, dCTP, dGTP; C−, dGTP; CBA, dCBATP, dGTP; T+, dTTP, dGTP; T−, dGTP; UBA, dUBATP, dGTP; A+, dATP, dGTP; A−, dGTP; ABA, dABATP, dGTP; G+, dGTP, dTTP; G−, dTTP; GBA, dGBATP, dTTP.

Figure 2. Primer extension with temp4 using (a) KOD XL; (b) Vent(exo-); (c) Pwo DNA polymerase; (d) inhibition of KOD XL DNA polymerase by dGBATP. Key: P, primer; +, all natural dNTPs; C−, dATP, dTTP, dGTP; CBA, dCBATP, dATP, dTTP, dGTP; T−, dATP, dCTP, dGTP; UBA, dUBATP, dATP, dCTP, dGTP; A−,dTTP, dCTP, dGTP; ABA, dABATP, dTTP, dCTP, dGTP; G−, dTTP, dCTP, dATP; GBA, dGBATP, dTTP, dCTP, dATP. In (d): lane 4, 140 μM; lane 5, 260 μM; lane 6, 600 μM dGBATP. dUBATP and dCBATP, the Heck reaction of dUITP proceeded in order to introduce one (ON1X) or four modifications only in low yield accompanied by large hydrolysis and the (ON4X) to the extended primer strand. reaction of dCITP did not occur at all. Significantly higher dCBATP, dUBATP, and dABATP were found to be good yields were observed by triphosphorylation of dUBA (40%) or substrates for all of the tested enzymes and were successfully dCBA (19%). incorporated into DNA bearing one modification (ON1X). Incorporation of dNBATPs into DNA by PEX. The Figure 1 shows the denaturing PAGE where only full length enzymatic synthesis of n-butyl acrylate modified oligonucleo- products were observed in all cases (except of weak additional tides (ONBAs) was studied by primer extension experiment bands of n − 1 products with Pwo polymerase). dGBATP was (PEX) using KOD XL, Vent(exo-), or Pwo polymerases. The not incorporated into DNA (ON1G, Figure 1d) by the tested templates and primer (for sequences see Table 2) were chosen enzymes under the same conditions.

9630 dx.doi.org/10.1021/jo4011574 | J. Org. Chem. 2013, 78, 9627−9637 The Journal of Organic Chemistry Article

All dNBATPs were also tested for multiple incorporations into DNA of a mixed sequence bearing four modiﬁcations (Figure 2). All tested DNA polymerases gave full length products with dCBATP and dABATP. KOD XL and Vent(exo-) (but not Pwo) were shown to be suitable enzymes for multiple incorporation of dUBATP (Figure 2a,b; lane 6). On the other hand, dGBATP was not only a poor substrate for these polymerases but also apparently inhibited the enzymatic synthesis of DNA by KOD XL and Vent(exo-) (Figure 2a,b; lane 10) where only unextended primers were detected. Therefore, the PEX using KOD XL was performed with decreasing concentrations of dGBATP. Figure 2d clearly conﬁrms that higher concentrations of dGBATP inhibited polymerase activity, whereas at lower concentration of dGBATP, the desired full-length PEX product has been obtained. All ONBAs prepared by PEX with biotinylated template using KOD XL DNA polymerase were isolated by magneto- separation7 and analyzed by MALDI (data are summarized in Table 3; for copies of spectra see Figures S1−S8 in the Supporting Information). KOD XL DNA polymerase was then also used for preparation of ONBAs for the electrochemical studies.

Table 3. MALDI Data of ONs Bearing Butyl Acrylate

oligonucleotide M (calcd) (Da) M (found) [M + H]+ (Da) ON1C 6077.1 6078.6 ON1T 6078.0 6079.1 ON1A 6100.1 6101.2 ON1G 6041.1 6042.2 ON4C 10122.1 10123.2 ON4T 10065.6 10066.1 ON4A 10118.1 10122.3 ON4G 10118.1 10119.2 Figure 3. (A, B) Cyclic voltammograms of dABA, dABAMP (A), dUBA Electrochemical Study. Electrochemistry has proved to be and dUBAMP (B). Concentrations of all substances 40 μM, potent, widely applicable tool for analysis of nucleic acids background electrolyte sodium acetate buffer pH 5.0. (C) Ex situ modified with diverse oxidizable or reducible extrinsic CVs of unmodified ON and ON4BAs with ABA or UBA incorporated; moieties.7 In our previous work, we demonstrated that such measured in ammonium formate/sodium phosphate background species can be applied for redox coding of nucleotide sequences electrolyte. See the Experimental Section for details. and/or individual nucleobases with potential applications in DNA diagnostics. To complete the palette of redox tags for reduction of uracil cannot be measured at mercury electrodes in independent coding of all four nucleobases, new potential labels aqueous media. In analogy with previously studied α,β- are sought among all newly synthesized dNTP conjugates. unsaturated carbonyl compounds,24 including ketones,24a Moreover, not only species applied purposely as redox DNA esters, amides,24b and acrylate anion,24c the primary electro- labels but also functional groups serving for further reduction of BA moiety can be expected at the α,β CC derivatization (DNA postsynthetic modification) introduced double bond probably via two successive one-electron steps, in into DNA can be determined and their chemical conversion aqueous medium resulting in the double bond hydrogenation. monitored by means of simple voltammetric techniques.23 Different shapes and potential shifts of the BAred peak among Thus, we were interested in electrochemical properties of the the BA-modified nucleos(t)ides can be ascribed to different BA nucleos(t)ide conjugates. Since the dCBA nucleotides were adsorbability of these species, effect of negative charge of the difficult to synthesize (due to low reactivity) and dGBATP was phosphate group and influence of these phenomena on the an inhibitor of DNA polymerases, we have chosen dUBA- and reduction mechanism of the given compound. dABA-modified ONs for the electrochemical studies. As evident Figure 3C shows ex situ CVs obtained for unmodified ON from cyclic voltammograms (CVs) measured on a hanging (product of PEX with all four unmodified dNTPs on temp4 mercury electrode (HMDE) shown in Figure 3A, dABA as well template) and ON4BAs with incorporated either ABA or UBA dABAMP produced two separated irreversible cathodic peaks conjugates. All ONs yielded a cathodic peak CA at −1.490 V which can be assigned to electrochemical reduction of adenine due to irreversible reduction of adenines and cytosines and an (peak Ared at −1.430 V) and of the BA moiety at a less negative anodic peak G at −0.255 V due to reoxidation of guanine potential (peak BAred at −1.295 V; compare with the behavior reduction product generated at potentials more negative than of modified DNA, vide infra). For dUBA and dUBAMP only one −1.6 V.25 Both ON4BAs produced and additional reduction irreversible signal was observed in agreement with the fact that signal, peak BAred,at−1.400 V, i.e., at a potential less negative

9631 dx.doi.org/10.1021/jo4011574 | J. Org. Chem. 2013, 78, 9627−9637 The Journal of Organic Chemistry Article

BA 13 31 than potential of peak CA (compare behavior of dA (MP) for C NMR). P chemical shifts were referenced to H3PO4 as above). In contrast to nucleo(s)tides, the modified ON produce external reference or to phosphate buffer signal 2.35 ppm in the case of practically identical shape of the voltammograms regardless of measurement in phosphate buffer. Coupling constants (J) are given in which modified nucleotide was incorporated. Such behavior hertz. NMR spectra of dNTPs were measured in phosphate buffer at could be expected considering the major contribution of the pH 7.1. Complete assignment of all NMR signals was achieved by polyanionic ON molecule containing excess of unmodified using a combination of H,H-COSY, H,C-HSQC, and H,C-HMBC nucleotides, dictating the overall interaction of the ON with the experiments. Mass spectra and high-resolution mass spectra were negatively charged electrode surface. Some differences were measured using ESI ionization technique. Mass spectra of functionalized ONs were measured by MALDI-TOF with 1 kHz smartbeam II observed in the anodic branch of CVs where broad peaks fi − laser. Water used in synthetic part was of HPLC quality. Ultrapure appeared with the modi ed ONs in the region between 1.160 Ω − water (18 M .cm) was used for all biochemical experiments. All and 0.750 V and intensities and potential of peak G were also chemicals, oligonucleotides, enzymes, streptavidine magnetic particles, influenced by the modifications (Figure 3C) which may be ff and isolation kits were purchased from commercial suppliers. Synthesis ascribed to e ects of the BA reduction products on electrode and characterization data for 2′-deoxy-5-iodocytidine 5′-O-triphospha- processes at negatively charged HMDE surface. Explanation of te,7b 2′-deoxy-5-iodouridine 5′-O-triphosphate,26a 2′-deoxy-7-iodo-7- these phenomena will require more detailed study which is deazaadenosine 5′-O-triphosphate,16b 2′-deoxy-7-iodo-7-deazaguano- beyond the scope of this report. Nevertheless, data presented sine 5′-O-triphosphate26b were reported previously. here demonstrate applicability of the BA electrochemical General Procedure I: Preparation of Butylacrylate-Modified reduction for monitoring of DNA modification with this Nucleosides (dNBAs). Method Ia. Nucleoside (dNI), butyl acrylate, functional group. Pd(OAc)2, and TPPTS were dissolved in a mixture of water/ acetonitrile (1:1, 3 mL) under argon atmosphere followed by addition ° ■ CONCLUSIONS of trielthylamine. The reaction mixture was stirred at 80 C for 1.5 h and then evaporated in vacuo. The products were purified by column The first aqueous Heck cross-coupling of halogenated chromatography. I nucleotides and dNTPs has been developed and tested on Method Ib. Nucleoside (dN ), butyl acrylate (6 equiv), Pd(OAc)2 fi BA BA the synthesis of acrylate-modi ed dN MPs and dN TPs. For (10 mol %), and PPh3 (20 mol %) were dissolved in DMF (3 mL) modification of 7-deazapurine nucleotides (synthesis of under argon atmosphere followed by addition of trielthylamine (2 dABAMP, dABATP, dGBAMP, and dGBATP), the direct aqueous equiv). The reaction mixture was stirred at 100 °C and then coupling procedure is comparable or more efficient than evaporated in vacuo. The products were purified by column phosphorylation of modified nucleosides. However, the Heck chromatography. BA coupling of pyrimidine dNTPs gave only traces of dUBATP or (E)-5-[2-(n-Butyloxycarbonyl)vinyl]-2′-deoxyuridine (dU ). no reaction at all (for dCIMP and dCITP), and therefore, the phosphorylation approach is necessary for the synthesis of these compounds. It can be concluded that the aqueous Heck cross-coupling is a possible reaction for modification of deazapurine dNTPs; however, apparently it is far less general and efficient than the Suzuki and Sonogashira reactions − developed previously.5 7,10,11 dCBATP, dUBATP, and dABATP were found to be good According to general method Ia, 2′-deoxy-5-iodouridine (71 mg, 0.201 ffi μ substrates for DNA polymerases and were e ciently mmol), butyl acrylate (172 L, 1.202 mmol), Pd(OAc)2 (3.4 mg, BA μ incorporated to ssONs and dsDNAs by PEX, whereas dG TP 0.010 mmol), TPPTS (17.1 mg, 0.020 mmol), and Et3N (172 L, was found to be inhibitor at higher concentrations. Electro- 0.402 mmol) were heated. The crude product was purified by column chemical properties of the butyl acrylate group were also chromatography using chloroform/methanol (10:1) as a mobile phase. studied on dUBA- and dABA-modified ONs to show that it gives dUBA was isolated as white powder (70 mg, 98%). ′ an analytically useful signal of reduction at −1.4 V in cyclic According to general method Ib, 2 -deoxy-5-iodouridine (100 mg, μ fi 0.283 mmol), butyl acrylate (243 L, 1.698 mmol), Pd(OAc)2 (6.3 voltammetry suitable for monitoring DNA modi cation with μ mg, 0.028 mmol), PPh3 (15.0 mg, 0.057 mmol), and Et3N (79 L, BA. Considering the position of the BA reduction signal at a fi potential less negative than potential of nucleobase reduction 0.566 mmol) were heated for 45 min. The crude product was puri ed by column chromatography using chloroform/methanol (7:1) as a but more negative than potential of reduction of previously BA 1 7 mobile phase. dU was isolated as white powder (93 mg, 93%). H introduced redox tags, it is promising also for possible redox ′″ 7 NMR (600.1 MHz, CD3OD): 0.97 (t, 3H, J4′″,3′″ = 7.4, H-4 ); 1.43 coding of nucleobases in combination with other redox labels ′″ ′″ (m, 2H, H-3 ); 1.66 (m, 2H, H-2 ); 2.27 (dt, 1H, Jgem = 13.6, J2′b,1′ = as an extension of the available palette of reducible tags toward ′ J2′b,3′ = 6.5, H-2 b); 2.34 (ddd, 1H, Jgem = 13.6, J2′a,1′ = 6.2, J2′a,3′ = 4.1, more negative potential region. Another field of potential ′ ′ H-2 a); 3.76 (dd, 1H, Jgem = 12.1, J5′b,4′ = 3.4, H-5 b); 3.86 (dd, 1H, applications are further postsynthetic chemical transformations ′ Jgem = 12.1, J5′a,4′ = 3.0, H-5 a); 3.95 (ddd, 1H, J4′,3′ = 3.6, J4′,5′ = 3.4, ′ ′″ of the butyl acrylate group in DNA. Further studies along these 3.0, H-4 ); 4.16 (t, 2H, J1′″,2′″ = 6.7, H-1 ); 4.42 (dddd, 1H, J3′,2′ = 6.5, ′ lines are under way. 4.1, J3′,4′ = 3.6, J3′,1′ = 0.5, H-3 ); 6.26 (ddd, 1H, J1′,2′ = 6.5, 6.2, J1′,3′ = ′ ″ 0.5, H-1 ); 6.89 (dd, 1H, J2″,3″ = 15.8, J2″,6 = 0.3, H-2 ); 7.39 (dd, 1H, ″ ■ EXPERIMENTAL SECTION J3″,2″ = 15.8, J3″,6 = 0.6, H-3 ); 8.49 (dd, 1H, J6,3″ = 0.6, J6,2″ = 0.3, H-6). 13 ′″ ′″ NMR spectra were recorded on a 600 MHz (600.1 MHz for 1H, 150.9 C NMR (150.9 MHz, CD3OD): 14.1 (CH3-4 ); 20.2 (CH2-3 ); ′″ ′ ′ ′″ MHz for 13C) or a 500 MHz (499.8 or 500.0 MHz for 1H, 202.3 or 31.9 (CH2-2 ); 41.9 (CH2-2 ); 62.4 (CH2-5 ); 65.3 (CH2-1 ); 71.7 ′ ′ ′ ″ 202.4 MHz for 31P, 125.7 MHz for 13C) spectrometer from sample (CH-3 ); 87.1 (CH-1 ); 89.2 (CH-4 ); 110.4 (C-5); 118.8 (CH-2 ); δ ″ solutions in D2OorCD3OD. Chemical shifts (in ppm, scale) were 138.6 (CH-3 ); 144.8 (CH-6); 151.1 (C-2); 163.7 (C-4); 169.4 (C- ″ + + + referenced as follows: D2O (referenced to dioxane as internal standard: 1 ). MS (ESI ): m/z 377.2 (100) [M + Na] ; 731.5 (50) [2M + Na] . 1 13 + + 3.75 ppm for H NMR and 69.30 ppm C NMR); CD3OD HR/MS (ESI ) for C16H22O7N2Na: [M + Na] calcd 377.1319, found (referenced to solvent signal: 3.31 ppm for 1H NMR and 49.00 ppm 377.1318.

9632 dx.doi.org/10.1021/jo4011574 | J. Org. Chem. 2013, 78, 9627−9637 The Journal of Organic Chemistry Article E n ′ ″ ( )-7-[2-( -Butyloxycarbonyl)vinyl]-2 -deoxy-7-deazaadeno- (d, 1H, J2″,3″ = 15.7, H-2 ); 7.59 (dd, 1H, J3″,2″ = 15.7, J3″,6 = 0.7, H- BA ″ 13 sine (dA ). 3 ); 8.72 (d, 1H, J6,3″ = 0.6, H-6). C NMR (125.7 MHz, CD3OD): ′″ ′″ ′″ ′ 14.07 (CH3-4 ); 20.19 (CH2-3 ); 31.93 (CH2-2 ); 42.63 (CH2-2 ); ′ ′″ ′ ′ 62.00 (CH2-5 ); 65.53 (CH2-1 ); 71.08 (CH-3 ); 88.04 (CH-1 ); 89.06 (CH-4′); 103.98 (C-5); 118.25 (CH-2″); 136.88 (CH-3″); 142.85 (CH-6); 157.16 (C-2); 165.26 (C-4); 168.48 (C-1″). MS (ESI+): m/z 354.3 (50) [M + H]+; 376.2 (100) [M + Na]+. HR/MS + + (ESI ) for C16H24O6N3:[M+H] calcd 354.16596, found 354.16598. (E)-7-[2-(n-Butyloxycarbonyl)vinyl]-2′-deoxy-7-deazaguanosine (dGBA).

According to general method Ia, 2′-deoxy-7-iodo-7-deazaadenosine (100 mg, 0.266 mmol), butyl acrylate (381 μL, 2.660 mmol), Pd(OAc)2 (6.1 mg, 0.027 mmol), TPPTS (37.8 mg, 0.067 mmol), and μ Et3N (111 L, 0.798 mmol) were reacted. The crude product was purified by column chromatography using chloroform/methanol (10:1). dABA was isolated as white powder (81 mg, 81%). According to general method Ib, 2′-deoxy-7-iodo-7-deazaadenosine (100 mg, 0.266 mmol), butyl acrylate (229 μL, 1.596 mmol), Pd(OAc)2 (6.1 mg, 0.027 mmol), PPh3 (13.9 mg, 0.053 mmol), and According to general method Ia, 2′-deoxy-7-iodo-7-deazaguanosine μ μ Et3N (74 L, 0.532 mmol) were heated for 1.5 h. The crude product (50 mg, 0.128 mmol), butyl acrylate (182 L, 1.28 mmol), Pd(OAc)2 fi was puri ed by column chromatography using chloroform/methanol (2.9 mg, 0.013 mmol), TPPTS (18.2 mg, 0.032 mmol), and Et3N (54 (10:1) as a mobile phase. dABA was isolated as white powder (97 mg, μL, 0.384 mmol) were heated. The crude product was purified by 1 97%). H NMR (600.1 MHz, CD3OD): 0.98 (t, 3H, J4′″,3′″ = 7.4, H- column chromatography using chloroform/methanol (10:1) as a ′″ ′″ ′″ BA 4 ); 1.46 (m, 2H, H-3 ); 1.70 (m, 2H, H-2 ); 2.35 (ddd, 1H, Jgem = mobile phase. dG was isolated as white powder (42 mg, 84%) after ′ 13.4, J2′b,1′ = 6.0, J2′b,3′ = 2.8, H-2 b); 2.64 (ddd, 1H, Jgem = 13.4, J2′a,1′ = final purification using reversed-phase HPLC (water/methanol, 5− ′ ′ 8.0, J2′a,3′ = 6.0, H-2 a); 3.75 (dd, 1H, Jgem = 12.2, J5′b,4′ = 3.6, H-5 b); 100%). ′ 3.82 (dd, 1H, Jgem = 12.2, J5′a,4′ = 3.2, H-5 a); 4.02 (ddd, 1H, J4′,5′ = 3.6, According to general method Ib, 2′-deoxy-7-iodo-7-deazaguanosine ′ ′″ 3.2, J4′,3′ = 2.8, H-4 ); 4.21 (t, 2H, J1′″,2′″ = 6.6, H-1 ); 4.53 (dtd, 1H, (100 mg, 0.256 mmol), butyl acrylate (366 μL, 2.560 mmol), ′ J3′,2′ = 6.0, 2.8, J3′,4′ = 2.8, J3′,1′ = 0.6, H-3 ); 6.41 (d, 1H, J2″,3″ = 15.8, Pd(OAc)2 (5.7 mg, 0.026 mmol), PPh3 (13.4 mg, 0.051 mmol), and ″ ′ μ H-2 ); 6.54 (dd, 1H, J1′,2′ = 8.0, 6.0, H-1 ); 7.95 (dd, 1H, J3″,2″ = 15.8, Et3N (72 L, 0.512 mmol) were heated for 1 h. The crude product ″ 13 J3″,6 = 0.8, H-3 ); 7.98 (bd, 1H, J6,3″ = 0.8, H-6); 8.12 (s, 1H, H-2). C was purified by column chromatography using chloroform/methanol ′″ ′″ BA NMR (150.9 MHz, CD3OD): 14.1 (CH3-4 ); 20.2 (CH2-3 ); 32.0 (7:1) as a mobile phase. dG was isolated as white powder (83 mg, ′″ ′ ′ ′″ 1 (CH2-2 ); 41.7 (CH2-2 ); 63.5 (CH2-5 ); 65.5 (CH2-1 ); 72.9 (CH- 83%). H NMR (500.0 MHz, CD3OD): 0.97 (t, 3H, J4′″,3′″ = 7.4, H- ′ ′ ′ ′″ ′″ ′″ 3 ); 86.6 (CH-1 ); 89.2 (CH-4 ); 103.0 (C-4a); 113.2 (C-5); 117.6 4 ); 1.44 (m, 2H, H-3 ); 1.67 (m, 2H, H-2 ); 2.28 (ddd, 1H, Jgem = ″ ″ ′ (CH-2 ); 125.2 (CH-6); 137.9 (CH-3 ); 152.0 (C-7a); 152.8 (CH- 13.4, J2′b,1′ = 6.0, J2′b,3′ = 3.0, H-2 b); 2.50 (ddd, 1H, Jgem = 13.4, J2′a,1′ = ″ + ′ ′ 2); 159.4 (C-4); 169.4 (C-1 ). MS (ESI ): m/z 377.1 (100) [M + 7.9, J2′a,3′ = 6.1, H-2 a); 3.71 (dd, 1H, Jgem = 12.0, J5′b,4′ = 4.2, H-5 b); + + + ′ H] ; 399.1 (20) [M + Na] . HR/MS (ESI ) for C18H25O5N4:[M+ 3.77 (dd, 1H, Jgem = 12.0, J5′a,4′ = 3.8, H-5 a); 3.94 (td, 1H, J4′,5′a = J4′,5′b + ′ ′″ H] calcd 377.1820, found 377.1819. = 4.0, J4′,3′ = 2.9, H-4 ); 4.15 (t, 2H, J1′″,2′″ = 6.6, H-1 ); 4.47 (dtd, 1H, E n ′ BA ′ ( )-5-[2-( -Butyloxycarbonyl)vinyl]-2 -deoxycytidine (dC ). J3′,2′a = 6.0, J3′,2′b = J3′,4′ = 3.0, J3′,1′ = 0.6, H-3 ); 6.40 (bdd, 1H, J1′,2′a = ′ ″ 7.9, J1′,2′b = 6.0, H-1 ); 7.19 (dd, 1H, J2″,3″ = 15.7, J2″,6 = 0.6, H-2 ); 7.43 (q, 1H, J6,2″ = J6,3″ = J6,1′ = 0.5, H-6); 7.66 (dd, 1H, J3″,2″ = 15.7, ″ 13 ′″ J3″,6 = 0.6, H-3 ). C NMR (125.7 MHz, CD3OD): 14.08 (CH3-4 ); ′″ ′″ ′ ′ 20.27 (CH2-3 ); 32.02 (CH2-2 ); 41.40 (CH2-2 ); 63.51 (CH2-5 ); ′″ ′ ′ ′ 65.04 (CH2-1 ); 72.79 (CH-3 ); 85.29 (CH-1 ); 88.76 (CH-4 ); 99.44 (C-4a); 118.05 (C-5); 118.14 (CH-2″); 124.42 (CH-6); 138.60 (CH-3″); 154.51 (C-7a); 154.67 (C-2); 161.52 (C-4); 170.38 (C-1″). MS (ESI+): m/z 393.2 (100) [M + H]+; 415.2 (35) [M + Na]+; 785.5 2′-Deoxy-5-iodocytidine (100 mg, 0.284 mmol), butyl acrylate (404 (42) [2M + H]+; 807.5 (26) [2M + Na]+. HR/MS (ESI+) for μ + L, 2.840 mmol), Pd(OAc)2 (6.4 mg, 0.028 mmol), and TPPTS (32.0 C18H25O6N4:[M+H] calcd 393.1769, found 393.1768. mg, 0.056 mmol) were dissolved in mixture water/acetonitrile (1:1, 6 General Procedure II: Preparation of Butyl Acrylate mL) under argon atmosphere followed by addition of triethylamine Modified Nucleoside Monophosphates (dNBAMPs). Method I (80 μL, 0.568 mmol). The reaction mixture was stirred at 80 °C for 2 IIa: Heck Coupling of Butyl Acrylate to dN MPs. Nucleoside I h and then evaporated in vacuo. The product was purified by column monophosphate (dN MP), butyl acrylate (10 equiv), Pd(OAc)2 (10 chromatography using chloroform/methanol (7:1) as a mobile phase. mol %), and TPPTS (25 mol %) were dissolved in a mixture water/ dCBA was isolated as pale yellow powder (16 mg, 16%) after final acetonitrile (1:1, 2 mL) under argon atmosphere followed by addition purification using reversed-phase HPLC (water/methanol, 5−100%). of triethylamine (3 equiv). The reaction mixture was stirred at 80 °C According to general method Ib, 2′-deoxy-5-iodocytidine (100 mg, for 2 h and then evaporated in vacuo. The products were purified by μ 0.284 mmol), butyl acrylate (243 L, 1.704 mmol), Pd(OAc)2 (6.4 C18 reversed-phase HPLC using water/methanol (5 to 100%) μ ff mg, 0.028 mmol), PPh3 (14.7 mg, 0.056 mmol), and Et3N (79 L, containing 0.1 M TEAB bu er as eluent. Several codistillations with 0.566 mmol) were heated for 2 h. The crude product was purified by water and conversion to sodium salt (Dowex 50WX8 in Na+ cycle) column chromatography using chloroform/methanol (7:1) as a mobile followed by freeze-drying from water gave the desired dNBAMPsas phase. dCBA was isolated as pale yellow powder (14 mg, 14%). 1H white solids. ′″ dNBA fi NMR (499.8 MHz, CD3OD): 0.97 (t, 3H, J4′″,3′″ = 7.4, H-4 ); 1.44 Method IIb: Phosphorylation of s. Acrylate-modi ed nucleo- ′″ ′″ BA ° (m, 2H, H-3 ); 1.68 (m, 2H, H-2 ); 2.22 (ddd, 1H, Jgem = 13.6, J2′b,3′ side (dN ) was dried at 80 C for 2 h in vacuo. After cooling, ′ = 6.3, J2′b,1′ = 5.8, H-2 b); 2.42 (ddd, 1H, Jgem = 13.6, J2′a,1′ = 6.4, J2′a,3′ PO(OMe)3 and POCl3 were added on ice under argon atmosphere. ′ ′ ° = 4.9, H-2 a); 3.77 (dd, 1H, Jgem = 12.1, J5′b,4′ = 3.2, H-5 b); 3.89 (dd, The resulting mixture was stirred at 0 C. The phosphorylation was ′ 1H, Jgem = 12.1, J5′a,4′ = 2.9, H-5 a); 3.96 (ddd, 1H, J4′,3′ = 4.3, J4′,5′ = stopped by addition of TEAB (2 M, 2 mL) and water (2 mL). The ′ ′″ fi 3.2, 2.9, H-4 ); 4.18 (t, 2H, J1′″,2′″ = 6.7, H-1 ); 4.40 (ddd, 1H, J3′,2′ = products were puri ed by C18 reversed-phase HPLC using water/ ′ ′ ff 6.53, 4.9, J3′,4′ = 4.3, H-3 ); 6.21 (dd, 1H, J1′,2′ = 6.4, 5.8, H-1 ); 6.33 methanol (5 to 100%) containing 0.1 M TEAB bu er as eluent.

9633 dx.doi.org/10.1021/jo4011574 | J. Org. Chem. 2013, 78, 9627−9637 The Journal of Organic Chemistry Article

Several codistillations with water and conversion to sodium salt (E)-7-[2-(n-Butyloxycarbonyl)vinyl]-2′-deoxy-7-deazaguano- (Dowex 50WX8 in Na+ cycle) followed by freeze-drying from water sine 5′-O-Phosphate (dGBAMP). gave the desired dNBAMPs as white solids. (E)-5-[2-(n-Butyloxycarbonyl)vinyl]-2′-deoxyuridine 5′-O- Phosphate (dUBAMP).

According to general method IIa, dGIMP (30.0 mg, 63.8 μmol), butyl μ μ acrylate (91 L, 0.638 mmol), Pd(OAc)2 (1.4 mg, 6.4 mol), TPPTS μ μ I (9.1 mg, 16.0 mol), and Et3N (27 L, 0.191 mmol) were heated to According to general method IIa, dU MP (50.0 mg, 0.110 mmol), yield dGBAMP (11.5 mg, 38%). 1H NMR (500.0 MHz, D O, butyl acrylate (150 μL, 1.100 mmol), Pd(OAc) (2.3 mg, 11 μmol), 2 2 ref(dioxane) = 3.75 ppm): 0.94 (t, 3H, J ′″ ′″ = 7.4, H-4′″); 1.42 (m, μ μ 4 ,3 TPPTS (14.9 mg, 27 mol), and Et3N (44 L, 0.330 mmol) were ′″ ′″ 2H, H-3 ); 1.69 (m, 2H, H-2 ); 2.41 (ddd, 1H, Jgem = 13.9, J2′b,1′ = reacted to yield dUBAMP (17.5 mg, 35%). ′ 6.3, J2′b,3′ = 3.3, H-2 b); 2.65 (ddd, 1H, Jgem = 13.9, J2′a,1′ = 7.9, J2′a,3′ = According to general method IIb, dUBA (60 mg, 0.169 mmol), ′ ′ 6.3, H-2 a); 3.87 (t, 2H, JH,P = J5′,4′ = 5.3, H-5 ); 4.14 (td, 1H, J4′,5′ = PO(OMe) (0.6 mL), and POCl (60 μL) were stirred at 0 °C for 4 h ′ ′″ 3 3 5.3, J4′,3′ = 3.3, H-4 ); 4.20 (t, 2H, J1′″,2′″ = 6.7, H-1 ); 4.64 (dt, 1H, and then kept in the refrigerator overnight. dUBAMP was isolated as a ′ ′ J3′,2′ = 6.3, 3.3, J3′,4′ = 3.3, H-3 ); 6.34 (dd, 1H, J1′,2′ = 7.9, 6.3, H-1 ); 1 ″ white powder (25 mg, 42%). H NMR (600.1 MHz, D2O): 0.92 (t, 6.90 (d, 1H, J2″,3″ = 15.7, H-2 ); 7.47 (s, 1H, H-6); 7.63 (d, 1H, J3″,2″ = ′″ ′″ ′″ ′″ ′″ ″ 13 3H, J4 ,3 = 7.4, H-4 ); 1.39 (m, 2H, H-3 ); 1.68 (m, 2H, H-2 ); 15.7 H-3 ). C NMR (125.7 MHz, D2O, ref(dioxane) = 69.3 ppm): ′ ′ ′ ′″ ′″ ′″ ′ 2.42 (m, 2H, H-2 ); 4.03 (m, 2H, H-5 ); 4.20 (m, 1H, H-4 ); 4.21 (t, 15.76 (CH3-4 ); 21.35 (CH2-3 ); 32.82 (CH2-2 ); 40.82 (CH2-2 ); ′″ ′ ′ ′″ ′ 2H, J1′″,2′″ = 6.7, H-1 ); 4.57 (dt, 1H, J3′,2′ = 5.6, 3.7, J3′,4′ = 3.7, H-3 ); 66.70 (d, JC,P = 4.5, CH2-5 ); 67.96 (CH2-1 ); 74.33 (CH-3 ); 85.56 ′ ″ ′ ′ 6.30 (t, 1H, J1′,2′ = 6.9, H-1 ); 6.90 (d, 1H, J2″,3″ = 15.9, H-2 ); 7.46 (d, (CH-1 ); 88.38 (d, JC,P = 8.3, CH-4 ); 100.79 (C-4a); 119.06 (C-5); ″ 13 ″ ″ 1H, J3″,2″ = 15.9, H-3 ); 8.19 (s, 1H, H-6). C NMR (150.9 MHz, 119.14 (CH-2 ); 126.10 (CH-6); 140.64 (CH-3 ); 155.69 (C-7a); ′″ ′″ ′″ ″ 31 1 D2O): 15.7 (CH3-4 ); 21.3 (CH2-3 ); 32.7 (CH2-2 ); 41.8 (CH2- 155.99 (C-2); 163.52 (C-4); 173.31 (C-1 ). P{ H} NMR (202.4 ′ ′ ′″ ′ − − 2 ); 67.0 (d, JC,P = 4.7, CH2-5 ); 68.2 (CH2-1 ); 73.9 (CH-3 ); 88.8 MHz, D2O): 3.83. MS (ESI ): m/z 471.1 (100) [M + H] . HR/MS ′ ′ ″ − − (CH-1 ); 88.9 (d, JC,P = 8.4, CH-4 ); 112.6 (C-5); 120.9 (CH-2 ); (ESI ) for C18H24O9N4P: [M + H] calcd 471.1286, found 471.1287. 140.7 (CH-3″); 146.9 (CH-6); 153.3 (C-2); 166.4 (C-4); 172.7 (C- General Procedure III: Preparation of Butyl Acrylate ″ 31 1 − Modified Nucleoside Triphosphates (dNBATPs). Method IIIa: 1 ). P{ H} NMR (202.4 MHz, D2O): 2.35. MS (ESI ): m/z 433.1 I − − − Heck Coupling of Butyl Acrylate to dN TPs. Nucleoside mono- (100) [M + H] . HR/MS (ESI ) for C16H22O10N2P: [M + H] calcd I 433.1018, found 433.1019. phosphate (dN TP), butyl acrylate (10 equiv), Pd(OAc)2 (10 mol %), (E)-7-[2-(n-Butyloxycarbonyl)vinyl]-2′-deoxy-7-deazaadeno- and TPPTS (25 mol %) were dissolved in a mixture water/acetonitrile sine 5′-O-Phosphate (dABAMP). (1:1, 2 mL) under argon atmosphere followed by addition of triethylamine (3 equiv). The reaction mixture was stirred at 80 °C for 1 h and then evaporated in vacuo. The products were purified by C18 reversed-phase HPLC using water/methanol (5 to 100%) containing 0.1 M TEAB buffer as eluent. Several codistillations with water and conversion to sodium salt (Dowex 50WX8 in Na+ cycle) followed by freeze-drying from water gave the desired dNBATPs as white solids. Method IIIb: Triphosphorylation of dNBAs. Acrylate-modified nucleoside (dNBA) was dried at 80 °C for 2 h in vacuo. After cooling, PO(OMe)3 and POCl3 were added on ice under argon atmosphere. According to general method IIa, dAIMP (20.0 mg, 41.9 μmol), butyl The resulting mixture was stirred at 0 °C. In a separate flask, the μ μ acrylate (57 L, 0.419 mmol), Pd(OAc)2 (0.9 mg, 4.2 mol), TPPTS mixture of (NHBu3)2H2P2O7 and tributylamine in dry DMF was μ μ ° (5.7 mg, 10.5 mol), and Et3N (17 L, 0.125 mmol) were heated to prepared under argon atmosphere, cooled to 0 C, and then added by yield dABAMP (10.5 mg, 55%). syringe to the reaction mixture. The mixture was stirred at 0 °C. The According to general method IIb, dABA (50.0 mg, 0.133 mmol), phosphorylation was stopped by addition of TEAB (2 M, 2 mL) and μ ° fi PO(OMe)3 (0.5 mL), and POCl3 (25 L) were stirred at 0 C for 45 water (2 mL). The products were puri ed by C18 reversed-phase min. dABAMP was isolated as white powder (27.1 mg, 54%). 1H NMR HPLC using water/methanol (5 to 100%) containing 0.1 M TEAB ′″ buffer as eluent. Several codistillations with water and conversion to (600.1 MHz, D2O): 0.95 (t, 3H, J4′″,3′″ = 7.4, H-4 ); 1.42 (m, 2H, H- ′″ ′″ sodium salt (Dowex 50WX8 in Na+ cycle) followed by freeze-drying 3 ); 1.71 (m, 2H, H-2 ); 2.49 (ddd, 1H, Jgem = 14.0, J2′b,1′ = 6.2, J2′b,3′ ′ from water gave the desired dNBATPs as white solids. = 3.1, H-2 b); 2.73 (ddd, 1H, Jgem = 14.0, J2′a,1′ = 8.0, J2′a,3′ = 6.2, H- ′ ′ (E)-5-[2-(n-Butyloxycarbonyl)vinyl]-2′-deoxyuridine 5′-O- 2 a); 3.85 (dt, 1H, Jgem = 11.1, JH,P = J5′b,4′ = 5.0, H-5 b); 3.89 (ddd, BA ′ Triphosphate (dU TP). 1H, Jgem = 11.1, JH,P = 5.4, J5′a,4′ = 5.0, H-5 a); 4.18 (td, 1H, J4′,5′ = 5.0, ′ ′″ J4′,3′ = 3.1, H-4 ); 4.23 (t, 2H, J1′″,2′″ = 6.8, H-1 ); 4.68 (dt, 1H, J3′,2′ = ′ ″ 6.2, 3.1, J3′,4′ = 3.1, H-3 ); 6.33 (d, 1H, J2″,3″ = 15.8, H-2 ); 6.56 (dd, ′ ″ 1H, J1′,2′ = 8.0, 6.2, H-1 ); 7.75 (dd, 1H, J3″,2″ = 15.8, J3″,6 = 0.6, H-3 ); 13 7.91 (s, 1H, H-6); 8.10 (s, 1H, H-2). C NMR (150.9 MHz, D2O): ′″ ′″ ′″ ′ 15.8 (CH3-4 ); 21.3 (CH2-3 ); 32.8 (CH2-2 ); 40.9 (CH2-2 ); 66.6 ′ ′″ ′ ′ (d, JC,P = 4.5, CH2-5 ); 68.2 (CH2-1 ); 74.4 (CH-3 ); 85.6 (CH-1 ); ′ ″ 88.6 (d, JC,P = 8.4, CH-4 ); 104.0 (C-4a); 115.2 (C-5); 119.4 (CH-2 ); 126.0 (CH-6); 139.8 (CH-3″); 153.3 (C-7a); 154.5 (CH-2); 160.2 According to general method IIIa, dUITP (40 mg, 58.7 μmol), butyl ″ 31 1 μ μ (C-4); 172.3 (C-1 ). P{ H} NMR (202.4 MHz, D2O): 4.45. MS acrylate (84 L, 0.587 mmol), Pd(OAc)2 (1.3 mg, 5.9 mol), TPPTS − − − μ μ (ESI ): m/z 455.1 (100) [M + H] .HR/MS(ESI)for (8.3 mg, 14.7 mol), and Et3N (25 L, 0.176 mmol) were reacted to − BA C18H24O8N4P: [M + H] calcd 455.1337, found 455.1341. yield dU TP (1.5 mg, 4%).

9634 dx.doi.org/10.1021/jo4011574 | J. Org. Chem. 2013, 78, 9627−9637 The Journal of Organic Chemistry Article

BA μ − − According to general method IIIb, dU (30.0 mg, 84.7 mol), (ESI ) for C17H20O13N5P3Na: [M + 2H + Na] calcd 618.0174, found μ ° PO(OMe)3 (0.3 mL), and POCl3 (30 L) were stirred at 0 C for 4 h 618.0169. and then kept in the refrigerator overnight. A cool solution of (E)-5-[2-(n-Butyloxycarbonyl)vinyl]-2′-deoxycytidine 5′-O- μ BA (NHBu3)2H2P2O7 (0.3 g) and tributylamine (120 L) in dry DMF Triphosphate (dC TP). (1.2 mL) was added, and the resulting mixture was stirred at 0 °C for 2 h. dUBATP was isolated as a white powder (22.0 mg, 40%). Method IIIc. Triethylammonium salt of dUITP (50 mg, 55.8 μmol), μ μ butyl acrylate (48 L, 0.335 mmol), Pd(OAc)2 (1.3 mg, 5.6 mol), μ and PPh3 (2.9 mg, 11.2 mol) were dissolved in DMF (3 mL) under argon atmosphere followed by addition of trielthylamine (16 μL, 0.112 mmol). The reaction mixture was stirred at 100 °C for 1 h and then evaporated in vacuo. The product was purified by C18 reversed-phase HPLC using water/methanol (5 to 100%) containing 0.1 M TEAB According to general method IIIb, dCBA (30.0 mg, 84.9 μmol), μ ° buffer as eluent. Several codistillations with water and conversion to PO(OMe)3 (0.3 mL), and POCl3 (16 L) were stirred at 0 C for 1 h. + sodium salt (Dowex 50WX8 in Na cycle) followed by freeze-drying Cool solution of (NHBu3)2H2P2O7 (0.3 g) and tributylamine (120 BA 1 μ from water gave dU TP (5.0 mg, 14%). H NMR (600.1 MHz, D2O, L) in dry DMF (1.3 mL) was added and the resulting mixture was ff ′″ ° BA pD = 7.1, phosphate bu er): 0.92 (t, 3H, J4′″,3′″ = 7.4, H-4 ); 1.39 (m, stirred at 0 C for 1.5 h. dC TP was isolated as a white powder (9.2 1 2H, H-3′″); 1.68 (m, 2H, H-2′″); 2.43 (m, 2H, H-2′); 4.20−4.27 (m, mg, 19%). H NMR (600.1 MHz, D2O, ref(dioxane) = 3.75 ppm, pD ′″ ′ ′ ′ ′ ff ′″ 5H, H-1 ,4 ,5 ); 4.67 (m, 1H, H-3 ); 6.29 (t, 1H, J1′,2′ = 6.6, H-1 ); = 7.1, phosphate bu er): 0.92 (t, 3H, J4′″,3′″ = 7.4, H-4 ); 1.40 (m, 2H, ″ ″ ′″ ′″ 6.91 (d, 1H, J2″,3″ = 15.9, H-2 ); 7.48 (d, 1H, J3″,2″ = 15.9, H-3 ); 8.17 H-3 ); 1.69 (m, 2H, H-2 ); 2.37 (ddd, 1H, Jgem = 14.2, J2′b,1′ = 7.1, 13 ′ (s, 1H, H-6). C NMR (150.9 MHz, D2O, pD = 7.1, phosphate J2′b,3′ = 6.4, H-2 b); 2.46 (ddd, 1H, Jgem = 14.2, J2′a,1′ = 6.2, J2′a,3′ = 3.7, ff ′″ ′″ ′″ ′ − ′″ ′ ′ bu er): 15.7 (CH3-4 ); 21.3 (CH2-3 ); 32.7 (CH2-2 ); 41.5 (CH2- H-2 a); 4.19 4.27 (m, 5H, H-1 ,4 ,5 ); 4.63 (dt, 1H, J3′,2′ = 6.2, 3.7, ′ ′ ′″ ′ ′ ′ 2 ); 67.9 (d, JC,P = 5.7, CH2-5 ); 68.2 (CH2-1 ); 73.1 (CH-3 ); 88.5 J3′,4′ = 3.7, H-3 ); 6.25 (dd, 1H, J1′,2′ = 7.1, 6.2, H-1 ); 6.39 (d, 1H, J2″,3″ ′ ′ ″ ″ ″ 13 (d, JC,P = 8.8, CH-4 ); 88.7 (CH-1 ); 112.6 (C-5); 120.9 (CH-2 ); = 15.9, H-2 ); 7.55 (d, 1H, J3″,2″ = 15.9, H-3 ); 8.17 (s, 1H, H-6). C 140.7 (CH-3″); 146.9 (CH-6); 153.3 (C-2); 166.4 (C-4); 172.7 (C- NMR (150.9 MHz, D2O, ref(dioxane) = 69.3 ppm, pD = 7.1, 31 1 ff ′″ ′″ 1″). P{ H} NMR (202.3 MHz, D O, pD = 7.1, phosphate buffer): phosphate bu er): 15.71 (CH3-4 ); 21.24 (CH2-3 ); 32.71 (CH2- 2 ′″ ′ ′ ′″ −21.31 (t, J = 19.8, Pβ); −10.34 (d, J = 19.8, Pα); −6.77 (bd, J = 19.8, 2 ); 42.05 (CH2-2 ); 67.95 (d, JC,P = 5.7, CH2-5 ); 68.33 (CH2-1 ); − − − − ′ ′ ′ Pγ). MS (ESI ): m/z 297.0 (60) [M 3PO3 C4H9] ; 433.1 (50) 73.23 (CH-3 ); 88.50 (d, JC,P = 8.6, CH-4 ); 89.43 (CH-1 ); 106.90 − − − − ″ ″ [M 2PO3 +H] ; 513.1 (100) [M + H PO3] ; 535.1 (90) [M + (C-5); 121.53 (CH-2 ); 138.81 (CH-3 ); 144.33 (CH-6); 159.24 (C- − − − − ″ 31 1 Na PO3] ; 593.1 (10) [M + 3H] ; 615.1 (30) [M + 2H + Na] ; 2); 166.62 (C-4); 171.66 (C-1 ). P{ H} NMR (202.4 MHz, D2O, − − ff ff 637.1 (20) [M + H + 2Na] . HR/MS (ESI ) for C16H24O16N2P3:[M ref(phosphate bu er) = 2.35 ppm, pD = 7.1, phosphate bu er): − + 3H] calcd 593.0344, found 593.0341. −21.38 (t, J = 19.8, Pβ); −10.44 (d, J = 19.8, Pα); −6.84 (bd, J = 19.8, − − − (E)-7-[2-(n-Butyloxycarbonyl)vinyl]-2′-deoxy-7-deazaadeno- Pγ). MS (ESI ): m/z 512.2 (100) [M + 3H PO3H] ; 534.2 (98) [M ′ O BA − − − − sine 5 - -Triphosphate (dA TP). +2H+Na PO3H] ; 432.2 (45) [M + H 2PO3H] ; 614.2 (30) [M + 2H + Na]−; 636.2 (26) [M + H + 2Na]−. HR/MS (ESI−) for − C16H24O15N3P3Na: [M + 2H + Na] calcd 614.0323, found 614.0328. (E)-7-[2-(n-Butyloxycarbonyl)vinyl]-2′-deoxy-7-deazaguanosine 5′-O-Triphosphate (dGBATP).

According to general method IIIa, dAITP (40 mg, 56.8 μmol), butyl μ μ acrylate (81 L, 0.568 mmol), Pd(OAc)2 (1.3 mg, 5.7 mol), TPPTS μ μ (8.1 mg, 14.2 mol), and Et3N (24 L, 0.170 mmol) were reacted to yield dABATP (15.5 mg, 43%). According to general method IIIb, dABA (30 mg, 79.7 μmol), According to general method IIIa, dGITP (30 mg, 47.8 μmol), butyl μ ° μ μ PO(OMe)3 (0.3 mL), and POCl3 (15 L) were stirred at 0 C for 45 acrylate (68 L, 0.478 mmol), Pd(OAc)2 (1.1 mg, 4.8 mol), TPPTS μ μ min. A cool solution of (NHBu3)2H2P2O7 (0.3 g) and tributylamine (6.8 mg, 12.0 mol), and Et3N (20 L, 0.143 mmol) were reacted to (120 μL) in dry DMF (1.2 mL) was added, and the resulting mixture yield dGBATP (13.0 mg, 44%). was stirred at 0 °C for 2 h. dABATP was isolated as a white powder According to general Method IIIb, dGBA (38.0 mg, 96.9 μmol), 1 μ ° (15.2 mg, 28%). H NMR (600.1 MHz, D2O, pD = 7.1, phosphate PO(OMe)3 (0.3 mL), and POCl3 (18 L) were stirred at 0 C for 1 h. ff ′″ ′″ bu er): 0.95 (t, 3H, J4′″,3′″ = 7.4, H-4 ); 1.42 (m, 2H, H-3 ); 1.71 (m, A cool solution of (NHBu3)2H2P2O7 (360 mg) and tributylamine (146 ′″ ′ μ 2H, H-2 ); 2.51 (ddd, 1H, Jgem = 13.8, J2′b,1′ = 6.2, J2′b,3′ = 3.2, H-2 b); L) in dry DMF (1.5 mL) was added, and the resulting mixture was ′ ° BA 2.71 (ddd, 1H, Jgem = 13.8, J2′a,1′ = 8.0, J2′a,3′ = 6.6, H-2 a); 4.12 (dt, 1H, stirred at 0 C for 1.5 h. dG TP was isolated as a white powder (15.4 ′ 1 Jgem = 11.2, JH,P = J5′b,4′ = 5.3, H-5 b); 4.18 (ddd, 1H, Jgem = 11.2, JH,P = mg, 25%). H NMR (500.0 MHz, D2O, ref(dioxane) = 3.75 ppm): ′ ′″ ′″ ′″ 6.2, J5′a,4′ = 4.4, H-5 a); 4.23 (t, 2H, J1′″,2′″ = 6.7, H-1 ); 4.24 (m, 1H, 0.94 (t, 3H, J4′″,3′″ = 7.4, H-4 ); 1.42 (m, 2H, H-3 ); 1.69 (m, 2H, H- ′ ′ − ′″ ′ H-4 ); 4.76 (m, 1H, H-3 overlapped with HDO signal); 6.31 (d, 2 ); 2.42 (ddd, 1H, Jgem = 14.0, J2′b,1′ = 6.2, J2′b,3′ = 3.3, H-2 b); 2.65 ″ ′ ′ − 1H, J2″,3″ = 15.7, H-2 ); 6.55 (dd, 1H, J1′,2′ = 8.0, 6.2, H-1 ); 7.70 (d, (ddd, 1H, Jgem = 14.0, J2′a,1′ = 7.9, J2′a,3′ = 6.3, H-2 a); 4.10 4.23 (m, ″ 13 ′ ′ ′″ 1H, J3″,2″ = 15.7, H-3 ); 7.87 (s, 1H, H-6); 8.08 (s, 1H, H-2). C 3H, H-4 ,5 ); 4.20 (t, 2H, J1′″,2′″ = 6.7, H-1 ); 4.72 (bdt, 1H, J3′,2′a = ff ′ NMR (150.9 MHz, D2O, pD = 7.1, phosphate bu er): 15.8 (CH3- 6.2, J3′,2′b = J3′,4′ = 3.2, H-3 ); 6.35 (dd, 1H, J1′,2′a = 7.9, J1′,2′b = 6.2, H- ′″ ′″ ′″ ′ ′ ″ 4 ); 21.3 (CH2-3 ); 32.8 (CH2-2 ); 41.1 (CH2-2 ); 68.2 (d, JC,P = 1 ); 7.91 (d, 1H, J2″,3″ = 15.8, H-2 ); 7.46 (s, 1H, H-6); 7.63 (d, 1H, ′ ′″ ′ ′ ″ 13 5.7, CH2-5 ); 68.2 (CH2-1 ); 73.8 (CH-3 ); 85.7 (CH-1 ); 88.0 (d, J3″,2″ = 15.8, H-3 ). C NMR (125.7 MHz, D2O, ref(dioxane) = 69.3 ′ ″ ′″ ′″ ′″ JC,P = 8.8, CH-4 ); 103.9 (C-4a); 115.2 (C-5); 119.4 (CH-2 ); 125.9 ppm): 15.77 (CH3-4 ); 21.35 (CH2-3 ); 32.82 (CH2-2 ); 40.95 ″ ′ ′″ ′ (CH-6); 139.6 (CH-3 ); 153.2 (C-7a); 154.3 (CH-2); 160.0 (C-4); (CH2-2 ); 67.96 (CH2-1 ); 68.32 (d, JC,P = 5.9, CH2-5 ); 73.87 (CH- ″ 31 1 ′ ′ ′ 172.3 (C-1 ). P{ H} NMR (202.3 MHz, D2O, pD = 7.1, phosphate 3 ); 85.80 (CH-1 ); 87.80 (d, JC,P = 8.8, CH-4 ); 100.82 (C-4a); buffer): −21.22 (t, J = 19.4, Pβ); −10.27 (d, J = 19.4, Pα); −6.53 (bd, J 119.09 (C-5); 119.17 (CH-2″); 126.10 (CH-6); 140.62 (CH-3″); − − − ″ 31 1 = 19.4, Pγ). MS (ESI ): m/z 516.1 (100) [M + 3H PO3H] ; 538.1 155.68 (C-7a); 155.94 (C-2); 163.49 (C-4); 173.31 (C-1 ). P{ H} − − − − − (90) [M + 2H + Na PO3H] ; 618.0 (15) [M + 2H + Na] . HR/MS NMR (202.4 MHz, D2O): 21.84 (t, J = 19.7, Pβ); 10.47 (d, J =

9635 dx.doi.org/10.1021/jo4011574 | J. Org. Chem. 2013, 78, 9627−9637 The Journal of Organic Chemistry Article

− 19.7, Pα); −7.95 (bd, J = 19.6, Pγ). MS (ESI ): m/z 551.2 (100) [M + Notes − − − 3H-PO3H] ; 573.2 (90) [M + 2H + Na PO3H] ; 471.3 (43) [M + The authors declare no competing financial interest. 3H + 2PO H]−; 653.2 (30) [M + 2H + Na]−; 631.2 (15) [M + 3H]−. 3 − − HR/MS (ESI ) for C18H26O15N4P3: [M + 3H] calcd 631.0613, found 631.0614. ■ ACKNOWLEDGMENTS Incorporation of Butyl Acrylate Modified Triphosphates into DNA by PEX. The reaction mixture (20 μL) contained primer (4 This work was supported by the Academy of Sciences of the μM), template (4 μM), DNA polymerase (0.075 U KOD XL, 0.1 U Czech Republic (RVO 61388963 and institutional research Vent(exo-) or 0.5 U Pwo), and dNTPs (either all natural or 3 natural plan AV0Z50040702) and the Grant Agency of the Academy of and 1 modified, 260 μM; for the inhibition studies shown in Figure 2d, Sciences of the Czech Republic (IAA400040901). 140; 260 and 600 μM dGBATP was used) in enzyme reaction buffer supplied by the manufacturer. Primer was labeled on its 5′-end by use 32 ■ REFERENCES of [γ P]-ATP according to standard techniques. The reaction mixture was incubated for 40 min at 60 °C in a thermal cycler. Primer (1) Reviews: (a) Kuwahara, M.; Sugimoto, N. Molecules 2010, 15, extension was stopped by addition of stop solution (40 μL, 80% (v/v) 5423−5444. (b) Hollenstein, M. Molecules 2012, 17, 13569−13591. formamide, 20 mM EDTA, 0.025% (w/v) bromophenol blue, 0.025% (2) Jager,̈ S.; Rasched, G.; Kornreich-Leshem, H.; Engeser, M.; (w/v) xylene cyanol) and heated for 5 min at 95 °C. Samples were Thum, O.; Famulok, M. J. Am. Chem. Soc. 2005, 127, 15071−15082. separated by 12.5% PAGE under denaturing conditions (42 mA, 1 h). (3) Menová ,́ P.; Cahova,́ H.; Plucnara, M.; Havran, L.; Fojta, M.; Visualization was performed by phosphoimaging (Figures 1 and 2). Hocek, M. Chem. Commun. 2013, 49, 4652−4654. MALDI-TOF Experiments. The MALDI-TOF spectra were (4) Menová ,́ P.; Raindlova,́ V.; Hocek, M. Bioconjugate Chem. 2013, measured on a MALDI-TOF/TOF mass spectrometer with 1 kHz 24, 1081−1093. smartbeam II laser. The measurements were done in reflectron mode (5) Thoresen, L. H.; Jiao, G.-S.; Haaland, W. C.; Metzker, M. L.; by droplet technique, with the mass range up to 30 kDa. The matrix Burgess, K. Chem.Eur. J. 2003, 9, 4603−4610. consisted of 3-hydroxypicolinic acid (HPA)/picolinic acid (PA)/ (6) (a) Riedl, J.; Pohl, R.; Ernsting, N. P.; Orsag, P.; Fojta, M.; ammonium tartrate in ratio 9/1/1. The matrix (1 μL) was applied on Hocek, M. Chem. Sci. 2012, 3, 2797−2806. (b) Riedl, J.; Menová ,́ P.; the target (ground steel) and dried down at room temperature. The Pohl, R.; Orsag,́ P.; Fojta, M.; Hocek, M. J. Org. Chem. 2012, 77, μ μ sample (1 L) and matrix (1 L) were mixed and added on the top of 8287−8293. the dried matrix preparation spot and dried at room temperature. ́ ́ ́ ̌ ́ BA (7) (a) Brazdilova, P.; Vrabel, M.; Pohl, R.; Pivonkova, H.; Havran, Preparation of ON s for MALDI-TOF Analysis. Streptavidin L.; Hocek, M.; Fojta, M. Chem.Eur. J. 2007, 13, 9527−9533. magnetic particles stock solution (Roche, 50 μL) was washed with ́ ́ ́ ̌ ́ ff × μ (b) Cahova, H.; Havran, L.; Brazdilova, P.; Pivonkova, H.; Pohl, R.; binding bu er (3 200 L, 10 mM Tris, 1 mM EDTA, 100 mM Fojta, M.; Hocek, M. Angew. Chem., Int. Ed. 2008, 47, 2059−2062. NaCl, pH 7.5). The PEX solution (prepared as described above) and (c) Vrabel,́ M.; Horaková ,́ P.; Pivoňkova,́ H.; Kalachova,́ L.; Černocka,́ binding buffer (50 μL) were added. Suspension was shaken (1200 ́ ̌ ° H.; Cahova, H.; Pohl, R.; Sebest, P.; Havran, L.; Hocek, M.; Fojta, M. rpm) for 30 min at 15 C. The magnetic beads were collected on a Chem.Eur. J. 2009, 15, 1144−1154. (d) Balintova,́ J.; Pohl, R.; magnet (DynaMag-2, Invitrogen) and washed with wash buffer (3 × Horaková ,́ P.; Vidlaková ,́ P.; Havran, L.; Fojta, M.; Hocek, M. Chem. 200 μL, 10 mM Tris, 1 mM EDTA, 500 mM NaCl, pH 7.5) and water Eur. J. 2011, 17, 14063−14073. (4 × 200 μL). Then water (50 μL) was added and the sample was (8) Obeid, S.; Yulikow, M.; Jeschke, G.; Marx, A. Angew. Chem., Int. denatured for 2 min at 55 °C and 900 rpm. The beads were collected 2008 − on a magnet and the solution was transferred into a clean vial. The Ed. , 47, 6782 6785. (9) Baccaro, A.; Steck, A.-L.; Marx, A. Angew. Chem., Int. Ed. 2012, product was analyzed by MALDI-TOF mass spectrometry (the results − are summarized in Table 3, for copies of mass spectra, see Figures S1− 51, 254 257. (10) (a) Borsenberger, V.; Kukwikila, M.; Howorka, S. Org. Biomol. S8, Supporting Information). − Electrochemistry. Nucleosides and dNMPs were analyzed by Chem. 2009, 7, 3826 3835. (b) Wirges, C. T.; Timper, J.; Fischler, M.; conventional in situ cyclic voltammetry (CV) while ONs (PEX Sologubenko, A. S.; Mayer, J.; Simon, U.; Carell, T. Angew. Chem., Int. − products) by ex situ (adsorptive transfer stripping, AdTS) CV. The Ed. 2009, 48, 219 223. (c) Borsenberger, V.; Howorka, S. Nucleic − PEX products (purified in their single-stranded form using Acids Res. 2009, 37, 1477 1485. (d) Gutschmiedl, K.; Fazio, D.; Carell, T. Chem.Eur. J. 2010, 16, 6877−6883. (e) Raindlova,́ V.; streptavidin-coated magnetic beads as above) were accumulated at ̌ the surface of a hanging mercury drop electrode (HMDE) for 60 s Pohl, R.; Sanda, M.; Hocek, M. Angew. Chem., Int. Ed. 2010, 49, 1064− from 5-μL aliquots containing 0.2 M NaCl. The electrode was then 1066. (f) Raindlova,́ V.; Pohl, R.; Hocek, M. Chem.Eur. J. 2012, 18, rinsed with deionized water and placed into an electrochemical cell. 4080−4087. CV settings: scan rate 1 V s−1, initial potential 0.0 V, switching (11) (a) Kielkowski, P.; Macíckovǎ -Cahová ,́ H.; Pohl, R.; Hocek, M. potential −1.85 V. Background electrolyte: 0.3 M ammonium formate, Angew. Chem., Int. Ed. 2011, 50, 8727−8730. (b) Kielkowski, P.; 0.05 M sodium phosphate, pH 6.9 (for ON measurements) or 0.2 M Brock, N. L.; Dickschat, J. S.; Hocek, M. ChemBioChem 2013, 14, sodium acetate pH 5.0 (for measurements of nucleos(t)ides). All 801−804. measurements were performed at room temperature using an Autolab (12) (a) Sidorov, A. V.; Grasby, J. A.; Williams, D. M. Nucleic Acids analyzer (Eco Chemie, The Netherlands) in connection with VA-stand Res. 2004, 32,1591−1601. (b) Kuwahara, M.; Nagashima, J.; 663 (Metrohm, Herisau, Switzerland) using a three-electrode system Hasegawa, M.; Tamura, T.; Kitagata, R.; Hanawa, K.; Hososhima, S.; with a Ag/AgCl/3 M KCl electrode as a reference and platinum wire Katsmatsu, T.; Ozaki, H.; Sawai, H. Nucleic Acids Res. 2006, 34, 5383− as an auxiliary electrode in solution deareated by argon purging. 5394. (c) Hollenstein, M.; Hipolito, C. J.; Lam, C. H.; Perrin, D. M. Nucleic Acids Res. 2009, 37, 1638−1649. (d) Hollenstein, M. Chem. ■ ASSOCIATED CONTENT Eur. J. 2012, 18, 13320−13330. * (13) (a) Shaughnessy, K. H. Eur. J. Org. Chem. 2006, 1827−1835. S Supporting Information (b) Shaughnessy, K. H. Chem. Rev. 2009, 109, 643−710. Copies of NMR and MALDI spectra. This material is available (14) Casalnuovo, A. L.; Calabrese, J. C. J. Am. Chem. Soc. 1990, 112, free of charge via the Internet at http://pubs.acs.org. 4324−4330. (15) Western, E. C.; Daft, J. R.; Johnson, E. M., II; Gannett, P. M.; ■ AUTHOR INFORMATION Shaughnessy, K. H. J. Org. Chem. 2003, 68, 6767−6774. (16) (a) Čapek, P.; Pohl, R.; Hocek, M. Org. Biomol. Chem. 2006, 4, Corresponding Author 2278−2284. (b) Čapek, P.; Cahova,́ H.; Pohl, R.; Hocek, M.; *E-mail: [email protected]. Gloeckner, Ch.; Marx, A. Chem.Eur. J. 2007, 13, 6196−6203.

9636 dx.doi.org/10.1021/jo4011574 | J. Org. Chem. 2013, 78, 9627−9637 The Journal of Organic Chemistry Article

(17) (a) Omumi, A.; Beach, D. G.; Baker, M.; Gabryelski, W.; Manderville, R. A. J. Am. Chem. Soc. 2010, 133,42−50. (b) Cahova,́ H.; Jaschke,̈ A. Angew. Chem., Int. Ed. 2013, 52, 3186−3190. (18) Reviews: (a) Heck, R. F. Acc. Chem. Res. 1979, 12, 146−151. (b) Beletskaya, I. P.; Cheprakov, A. V. Chem. Rev. 2000, 100, 3009− 3066. (19) General review: (a) Agrofoglio, L. A.; Gillaizeau, I.; Saito, Y. Chem. Rev. 2003, 103, 1875−1916. Recent examples: (b) Lee, S. E.; Sidorov, A.; Gourlain, T.; Mignet, N.; Thorpe, S. J.; Brazier, J. A.; Dickman, M. J.; Hornby, D. P.; Grasby, J. A.; Williams, D. M. Nucleic Acids Res. 2001, 29, 1565−1573. (c) Garg, N. K.; Woodroofe, C. C.; Lacenere, C. J.; Quake, S. R.; Stoltz, B. M. Chem. Commun. 2005, 4551−4553. (d) Ding, H.; Greenberg, M. M. J. Am. Chem. Soc. 2007, 129, 772−773. (e) Ogino, M.; Taya, Y.; Fujimoto, K. Org. Biomol. Chem. 2009, 7, 3163−3167. (20) Tobrman, T.; Dvorǎk,́ D. Eur. J. Org. Chem. 2008, 2923−2928. (21) Cho, J. H.; Shaughnessy, K. H. Synlett 2011, 2963−2966. (22) Ludwig, J. Acta Biochim. Biophys. Acad. Sci. Hung. 1981, 16, 131−133. (23) Raindlova,́ V.; Pohl, R.; Klepetaŕovǎ ,́ B.; Havran, L.; Šimkova,́ E.; Horaková ,́ P.; Pivoňkova,́ H.; Fojta, M.; Hocek, M. ChemPlusChem 2012, 77, 652−662. (24) (a) Zimmer, J. P.; Richards, J. A.; Turner, J. C.; Evans, D. H. Anal. Chem. 1971, 43, 1000−1006. (b) Klemm, L. H.; Olson, D. R. J. Org. Chem. 1979, 44, 4524−4527. (c) Fahr, T.; Petr, A.; Dunsch, L. Ber. Bunsenges. Phys. Chem. 1997, 101, 1040−1044. (25) Palecek,̌ E.; Bartosík,̌ M. Chem. Rev. 2012, 112, 3427−3481. (26) (a) Kovacs, T.; Otvös, L. Tetrahedron Lett. 1988, 29, 4525− 4528. (b) McDougall, M. G.; Hosta, L. P.; Kumar, S.; Fuller, C. W. Nucleosides Nucleotides 1999, 18, 1009−1011.

9637 dx.doi.org/10.1021/jo4011574 | J. Org. Chem. 2013, 78, 9627−9637 Anal. Chem. 2010, 82, 2969–2976

Determination of the Level of DNA Modiﬁcation with Cisplatin by Catalytic Hydrogen Evolution at Mercury-Based Electrodes

Petra Hora´ kova´,†,‡ Lucie Teˇ snohlı´dkova´,† Ludeˇ k Havran,† Pavlı´na Vidla´ kova´,† Hana Pivonˇ kova´,*,† and Miroslav Fojta†

Institute of Biophysics, Academy of Sciences of the Czech Republic, v.v.i., Kra´lovopolska´ 135, 612 65 Brno, Czech Republic, and Department of Analytical Chemistry, Faculty of Chemical Technology, University of Pardubice, Studentska´ 573, 532 10 Pardubice, Czech Republic

Electrochemical methods proved useful as simple and electrochemical activity related to redox processes of the metal inexpensive tools for the analysis of natural as well as moieties. Metal chelates and organometallics have been used for chemically modified nucleic acids. In particular, co- labeling of DNA in electrochemical DNA hybridization sensors valently attached metal-containing groups usually render and other techniques designed for the sequence-specific DNA the DNA well-pronounced electrochemical activity related sensing or detecting DNA damage.6-10 Reversible redox re- to redox processes of the metal moieties, which can in sponses of ferrocene derivatives,11-13 ruthenium, osmium,14-17 some cases be coupled to catalytic hydrogen evolution at or other metal complexes attached to DNA via chemical oligo- mercury or some types of amalgam electrodes. In this nucleotide synthesis,12 using modified nucleoside triphosphates paper we used voltammetry at the mercury-based elec- and DNA polymerases,11,13,17,18 or via chemical modification of trodes for the monitoring of DNA modification with cis- natural DNAs or standard (unmodified) synthetic oligonucleo- diamminedichloroplatinum (cisplatin), a representative of tides,7,14–16,19 have been measured at different working electrodes. metallodrugs used in the treatment of various types of cancer or being developed for such purpose. In cyclic (2) Fojta, M. Collect. Czech. Chem. Commun. 2004, 69, 715–747. voltammetry at the mercury electrode, the cisplatin- (3) Fojta, M. In Electrochemistry of Nucleic Acids and Proteins. Towards Electrochemical Sensors for Genomics and Proteomics; Palecek, E., Scheller, modified DNA yielded catalytic currents the intensity of F., Wang, J., Eds.; Elsevier: Amsterdam, The Netherlands, 2005; pp 386- which reflected DNA modification extent. In square-wave 431. voltammetry, during anodic polarization after prereduc- (4) Fojta, M.; Jelen, F.; Havran, L.; Palecek, E. Curr. Anal. Chem. 2008, 4, 250–262. tion of the cisplatinated DNA, a well-developed, sym- (5) Palecek, E.; Jelen, F. In Electrochemistry of Nucleic Acids and Proteins. metrical signal (peak P) was obtained. Intensity of the Towards Electrochemical Sensors for Genomics and Proteomics; Palecek, E., peak P linearly responded to the extent of DNA modifica- Scheller, F., Wang, J., Eds.; Elsevier: Amsterdam, The Netherlands, 2005; ) pp 74-174. tion at levels relevant for biochemical studies (rb (6) Fojta, M.; Horakova, P.; Cahova, K.; Fojtova, M.; Hason, S.; Havran, L.; 0.01-0.10, where rb is the number of platinum atoms Kostecka, P.; Nemcova, K.; Pivonkova, H.; Brazdova, M. In Bioelectrochem- bound per DNA nucleotide). We demonstrate a correla- istry Research Developments; Bernstein, E. M., Ed.; Nova Publishers: Hauppauge, NY, 2008. tion between the peak P intensity and a loss of sequence- (7) Havran, L.; Vacek, J.; Cahova, K.; Fojta, M. Anal. Bioanal. Chem. 2008, specific DNA binding by tumor suppressor protein p53, 391, 1751–1758. as well as blockage of DNA digestion by a restriction (8) Labuda, J.; Fojta, M.; Jelen, F.; Palecek, E. In Encyclopedia of Sensors; Grimes, C. A., Dickey, E. C., Pishko, M. V., Eds.; American Scientific endonuclease Msp I (both caused by the DNA cisplati- Publishers: Stevenson Ranch, CA, 2006; Vol. 3E-F, pp 201-228. nation). Application of the electrochemical technique in (9) Palecek, E.; Fojta, M. In Bioelectronics; Wilner, I., Katz, E., Eds.; Wiley studies of DNA reactivity with various anticancer platinum VCH: Weinheim, Germany, 2005; pp 127-192. (10) Palecek, E.; Fojta, M. Talanta 2007, 74, 276–290. compounds, as well as for an easy determination of the (11) Brazdilova, P.; Vrabel, M.; Pohl, R.; Pivonkova, H.; Havran, L.; Hocek, M.; extent of DNA platination in studies of its biochemical Fojta, M. Chem.sEur. J. 2007, 13, 9527–9533. effects, is discussed. (12) Fan, C.; Plaxco, K. W.; Heeger, A. J. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 9134–9137. (13) Patolsky, F.; Weizmann, Y.; Willner, I. J. Am. Chem. Soc. 2002, 124, 770– Electrochemical methods proved useful as simple and inex- 772. (14) Flechsig, G. U.; Reske, T. Anal. Chem. 2007, 79, 2125–2130. pensive tools for the analysis of natural as well as chemically (15) Fojta, M.; Havran, L.; Kizek, R.; Billova, S.; Palecek, E. Biosens. Bioelectron. 1-5 modified nucleic acids. In particular, covalently attached metal- 2004, 20, 985–994. containing groups usually render the DNA well-pronounced (16) Fojta, M.; Kostecka, P.; Trefulka, M.; Havran, L.; Palecek, E. Anal. Chem. 2007, 79, 1022–1029. * To whom correspondence should be addressed. E-mail: [email protected]. (17) Vrabel, M.; Horakova, P.; Pivonkova, H.; Kalachova, L.; Cernocka, H.; † Academy of Sciences of the Czech Republic. Cahova, H.; Pohl, R.; Sebest, P.; Havran, L.; Hocek, M.; Fojta, M. ‡ University of Pardubice. Chem.sEur. J. 2009, 15, 1144–1154. (1) Brabec, V.; Vetterl, V.; Vrana, O. In Experimental Techniques in Bioelec- (18) Hocek, M.; Fojta, M. Org. Biomol. Chem. 2008, 6, 2233–2241. trochemistry; Brabec, V., Walz, D., Milazzo, G., Eds.; Birkhauser Verlag: (19) Trefulka, M.; Ferreyra, N.; Ostatna, V.; Fojta, M.; Rivas, G.; Palecek, E. Basel, Switzerland, 1996; Vol. 3, pp 287-359. Electroanalysis 2007, 19, 1334–1338.

10.1021/ac902987x  2010 American Chemical Society Analytical Chemistry, Vol. 82, No. 7, April 1, 2010 2969 Published on Web 02/26/2010 Electrochemical reduction of some of these species is coupled to been utilized to monitor DNA cisplatination.36 Other approaches catalytic hydrogen evolution at mercury or some types of amalgam to electrochemical analysis of DNA modified with platinum or electrodes, and the high electron-yield catalytic processes can be other metal complexes rely in measurements of the changes of utilized to attain a high sensitivity of the determination of the structure-selective DNA signals. Modification of DNA with the modified DNA (in the sense of either detection of low amounts metallodrugs is connected with characteristic conformational of the densely modified DNA or detection of low levels of DNA changes in the DNA double helix, which may be denaturational modification in the presence of a considerable excess of unmodi- (resulting in opening of base pairs) or nondenaturational (changes fied nucleotides). Typical examples of transition metal-based of DNA conformation without opening the base pairs).28,37 These nucleic acid adducts yielding well-pronounced catalytic response events can be monitored and distinguished by polarographic at the mercury-based electrodes are the products of DNA16,20-22 techniques,1,2,5 which represent a potent tool for studies of specific or PNA23 modification with osmium tetroxide complexes24,25 or DNA modifications. Moreover, bending of DNA due to cisplati- adducts of ribonucleosides with analogous osmate complexes.26 nation has recently been reported to hamper long-range electron Cisplatin [cis-diamminedichloroplatinum(II)] is a representative transfer through DNA, which has been monitored electrochemi- of cytotoxic and antineoplastic metallodrugs used for the treatment cally.38 Differential pulse polarography has been used for indirect of various malignancies or being developed for such purpose.23–28 determination of the extent of global DNA platination via deter- The drug binds covalently to DNA, forming several kinds of mination of the residual unbound platinum complex.39 Surpris- adducts. In double-stranded chromosomal DNA, the most frequent ingly, reports on direct determination of DNA platination via of them are intrastrand cross-links (IAC) between neighboring measuring electrochemical responses related to electrochemical purine residues: 65% of 1,2-GG IAC and 25% of 1,2-AG IAC. Another activity of the DNA-bound platinum moieties are missing in the 6-10% belong to 1,3-GNG IAC and interstrand cross-links and literature (with an exception of stripping voltammetric technique 2-3% of various monofunctional adducts. Similar products are proposed for the determination of DNA-bound platinum in formed by clinically used cisplatin analogues carboplatin and mineralized samples from patients treated with oxaliplatin40). oxaliplatin, whereas other types of platinum complexes may form In this paper we applied voltammetry at the mercury-based other types of DNA adducts, depending on the mutual steric electrodes for the monitoring of DNA modification with cisplatin positioning of DNA and the given complex upon their physical using catalytic hydrogen evolution that accompanies redox interaction.27–29 processes of the cisplatin-DNA adducts. We demonstrate that Besides utilization of other biochemical and physicochemical intensities of the catalytic currents, measured by cyclic or square- techniques, modification of DNA with cisplatin or its analogues wave voltammetry, linearly respond to the extent of DNA has been studied by electrochemical methods. Guanine is the modification at levels relevant in the studies of biochemical effects primary target for the platinum complexes, and chemical modifica- of DNA cisplatination (0.01-0.10 platinum atoms per nucleotide). tion of this nucleobase often affects its electrochemical responses A correlation between the intensity of the electrochemical at both carbon- and mercury-based electrodes.3,4,30 Several authors response (providing information about the level of DNA cisplati- thus focused their attention to changes in the guanine (and/or nation) and a loss of sequence-specific DNA binding by tumor adenine) oxidation response at various types of carbon electrodes suppressor protein p53, as well as blockage of DNA digestion by to develop techniques suitable for monitoring DNA modification a restriction endonuclease Msp I due to DNA cisplatination, is with the platinum complexes and/or biosensors for the platinum demonstrated. drugs using DNA as the recognition layer.31-35 Similarly, changes in the guanine response at the mercury electrode have recently MATERIALS AND METHODS (20) Havran, L.; Fojta, M.; Palecek, E. Bioelectrochemistry 2004, 63, 239–243. Material and Reagents. Supercoiled (sc) pBSK(-) DNA was (21) Reske, T.; Surkus, A. E.; Duwensee, H.; Flechsig, G. U. Microchim. Acta prepared as described,41 synthetic oligonucleotides (ODNs; 2009, 166, 197–201. Table 1) were purchased from VBC Biotech, calf thymus DNA (22) Yosypchuk, B.; Fojta, M.; Havran, L.; Heyrovsky, M.; Palecek, E. Elec- troanalysis 2006, 18, 186–194. and cisplatin were from Sigma, restriction endonuclease Msp I (23) Palecek, E.; Trefulka, M.; Fojta, M. Electrochem. Commun. 2009, 11, 359– and T4 polynucleotide kinase were from NEB, and γ-32P-ATP was 362. from ICN. Other chemicals were of analytical grade. Dynabeads (24) Jelen, F.; Karlovsky, P.; Pecinka, P.; Makaturova, E.; Palecek, E. Gen. Physiol. Biophys. 1991, 10, 461–473. Oligo(dT)25 (DBT) and magnetic concentrator were supplied (25) Palecek, E. In Methods in Enzymology; Abelson, J. N., Simon, M. I., Eds.; by Invitrogen. - Academic Press: New York, 1992; Vol. 212, pp 139 155. DNA Modification with Cisplatin. DNA was incubated with (26) Trefulka, M.; Ostatna, V.; Havran, L.; Fojta, M.; Palecek, E. Electroanalysis 2007, 19, 1281–1287. cisplatin in 0.1 M NaClO4 at 37 °C overnight (usually single- (27) Kasparkova, J.; Fojta, M.; Farrell, N.; Brabec, V. Nucleic Acids Res. 2004, 32, 5546–5552. (35) Ravera, M.; Bagni, G.; Mascini, M.; Dabrowiak, J. C.; Osella, D. J. Inorg. (28) Marini, V.; Kasparkova, J.; Novakova, O.; Scolaro, L. M.; Romeo, R.; Brabec, Biochem. 2007, 101, 1023–1027. V. J. Biol. Inorg. Chem. 2002, 7, 725–734. (36) Krizkova, S.; Adam, V.; Petrlova, J.; Zitka, O.; Stejskal, K.; Zehnalek, J.; (29) Kloster, M.; Kostrhunova, H.; Zaludova, R.; Malina, J.; Kasparkova, J.; Sures, B.; Trnkova, L.; Beklova, M.; Kizek, R. Electroanalysis 2006, 19, Brabec, V.; Farrell, N. Biochemistry 2004, 43, 7776–7786. 331–338. (30) Fojta, M. Electroanalysis 2002, 14, 1449–1463. (37) Novakova, O.; Chen, H. M.; Vrana, O.; Rodger, A.; Sadler, P. J.; Brabec, V. (31) Bagni, G.; Osella, D.; Sturchio, E.; Mascini, M. Anal. Chim. Acta 2005, Biochemistry 2003, 42, 11544–11554. 573-574, 81–89. (38) Wong, E. L. S.; Gooding, J. J. J. Am. Chem. Soc. 2007, 129, 8950–8951. (32) Brabec, V. Electrochim. Acta 2000, 45, 2929–2932. (39) Kim, S. D.; Vrańa, O.; Kleinwa¨chter, V.; Niki, K.; Brabec, V. Anal. Lett. (33) Brett, A. M. O.; Serrano, S. H. P.; Macedo, T. A.; Raimundo, D.; Marques, 1990, 23, 1505–1518. M. H.; LaScalea, M. A. Electroanalysis 1996, 8, 992–995. (40) Weber, G.; Messerschmidt, J.; Pieck, A. C.; Junker, A. M.; Wehmeier, A.; (34) Mascini, M.; Bagni, G.; Di Pietro, M. L.; Ravera, M.; Baracco, S.; Osella, Jaehde, U. Anal. Bioanal. Chem. 2004, 380, 54–58. D. Biometals 2006, 19, 409–418. (41) Fojta, M.; Palecek, E. Anal. Chim. Acta 1997, 342, 1–12.

2970 Analytical Chemistry, Vol. 82, No. 7, April 1, 2010 Switzerland) in three-electrode setup (Ag/AgCl/3 M KCl electrode Table 1. Nucleotide Sequences of Synthetic Oligonucleotides Used in This Worka as a reference and platinum wire as an auxiliary electrode), under argon. Cyclic voltammetry (CV) at HMDE was performed as acronym nucleotide sequence follows: background electrolyte, 0.3 M ammonium formate, 0.05 noG CATCATCATCATCATCATCATAAAAAAAAAAAAA M sodium phosphate, pH 6.9 (AFP) or Britton-Robinson buffer AAAAAAA G25 GGGGGGGGGGGGGGGGGGGGGGGGGAAAAAAA of different pHs; initial potential, 0.0 V; switching potential, -1.85 AAAAAAAAAAAAAAAAAA V; scan rate,1Vs-1 (if not stated otherwise). Square-wave GTT GTTGTTGTTGTTGTTGTTGTTAAAAAAAAAAAAA voltammetry (SWV) at a HMDE was performed as follows: AAAAAAA GGT GGTGGTGGTGGTGGTGGTGGTAAAAAAAAAAAA background electrolyte, AFP or Britton-Robinson buffer of AAAAAAAA different pHs; initial potential, -1.85 V; end potential, 0.0 V; ds50 GACGATATGCTAGAGGCATGTTTAAACATGTTT quiescent time, 2 s; frequency, 200 Hz; amplitude, 50 mV; ACCGGTTGATATCGAA potential step, 5 mV, if not stated otherwise. SWV at a pyrolytic a The noG, G25, GTT and GGT ODNs were designed for the graphite electrode (PGE) was performed as follows: back- experiments involving magnetoseparation; they contain An adaptors for capture at the DBT and the “probe” stretches differing in reactivity ground electrolyte, 0.2 M sodium acetate, pH 5.0; initial towards cisplatin. The ds50 ODN (shown top strand only, for the duplex potential, 0.0 V; end potential, +1.5 V; quiescent time, 2 s; see Figure 5 legend) involves a p53 target site (underlined) and a Msp I restriction site (bold italics) for testing biochemical effects of DNA frequency, 200 Hz; amplitude, 25 mV; potential step, 5 mV. cisplatination. Renewal of electrode surfaces: for m-AgSAE, after each measurement, the electrode was cleaned by applying of -2.0 V for 120 s in 0.2 M KCl; for PGE, the surface was renewed by stranded calf thymus DNA and the ODN probes) or for 48 h peeling-off the graphite top layer using sticky tape after applying (double-stranded ODN) in the dark. DNA concentration in the +1.7 V for 60 s in the background electrolyte. -1 reaction mixture was 20 µgmL ; in the case of the modifica- Restriction Cleavage and p53-DNA Binding. Restriction tion of ODN probes in the presence of the competitor DNA, cleavage of the pBSK(-) plasmid competitor DNA and of the -1 the reaction mixture contained 20 µgmL of the probe and ds50 ODN (radiolabeled with 32P using γ-32P-ATP and T4 -1 20 µgmL of the pBSK(-) plasmid. The plasmid DNA was polynucleotide kinase) was conducted under standard condi- chosen for this purpose because it is suitable for the parallel tions recommended by the enzyme supplier. Products of the Msp I cleavage tests (see below and the Supporting Informa- enzymatic cleavage were analyzed using electrophoresis in 1% tion), as it contains defined number of the restriction sites. agarose gel stained with ethidium bromide (plasmid DNA) or Cisplatin concentrations and the corresponding rb values (number in 15% polyacrylamide gel followed by autoradiography. Binding of cisplatin moieties per DNA nucleotide, always related to total of the p53 protein to the ds50 ODN protein was tested in 2 DNA content in the sample) are given in the text and figure mM DTT, 50 mM KCl, 5 mM Tris pH 7.6, 0.01% Triton-X 100 legends. Note: cisplatin is a genotoxic compound which has to be (total volume 20 µL), after 30 min of incubation on ice, using handled with care. a protein/DNA molar ratio of 5/1. The reaction mixture Magnetoseparation Procedure. The purpose of the magnetic contained 50 ng of the 32P-labeled ds50 ODN and 2 µgof separation was to isolate the cisplatin-treated ODNs from the nonspecific competitor calf thymus DNA. The protein-ds50 mixture with the competitor plasmid DNA, and thus to determine complexes were detected by electrophoretic mobility shift assay the portion of cisplatin bound the given ODN probe. Aliquots of in 5% native polyacrylamide gel followed by autoradiography. 20 µL of the DBT were washed thrice in 50 µL of binding buffer Band intensities on the gels were quantified by ImageJ + (0.3 M NaCl 10 mM Tris, pH 7.4), followed by incubation with software. 40 µL of the cisplatin-treated samples in the binding buffer on a shaker at 20 °C for 30 min to allow binding of the ODN to the RESULTS AND DISCUSSION DBT via hybridization of the immobilized T25 stretches with An adaptors of the ODN. Then the beads were washed four times For the monitoring of DNA modification with platinum by 50 µL of the binding buffer (using repeated magnetosepa- complexes, several authors used voltammetry at carbon electrodes ration and resuspending) and transferred into 10 µL of deion- to follow changes in the intensity of oxidation signal due to ized water. ODNs were released from DBT by heating at 85 oxidation of purine residues, particularly the peak Gox due to °C for 2 min. After addition of NaCl to a concentration of 0.2 guanine.31–34 Such an approach is rational because N7 of guanine M, the recovered ODNs were analyzed electrochemically. is the primary target of cisplatin and other platinum drugs, and Voltammetric Measurements. Voltammetric responses of chemical damage within the guanine imidazole ring results in the modified DNA or ODNs were measured using the adsorptive diminution of the peak Gox. Prolonged exposure of DNA transfer stripping, AdTS,2,5 procedure with DNA-modified elec- recognition layer immobilized on a carbon electrode to a trodes. A hanging mercury drop electrode (HMDE) or mercury solution containing a platinum complex forming covalent meniscus-modified silver solid amalgam electrode (m-AgSAE) was adducts with the DNA results in accumulation of damage to immersed in a 5 µL aliquot of the sample. After a 60 s accumulation guanine residues with concomitant decrease of the peak Gox. at open current circuit, the electrode was subsequently washed On the other hand, such an approach is less well suited for by deionized water and by background electrolyte and placed in the determination of low levels of DNA modification because a voltammetric cell. All measurements were performed at room of its “signal off” nature. DNA contains many guanine residues, temperature with an Autolab analyzer (Eco Chemie, Utrecht, The and relative decrease of the peak Gox depends on the fraction Netherlands) connected to VA-Stand 663 (Metrohm, Herisau, of guanines modified. Reasonable levels of DNA platination

Analytical Chemistry, Vol. 82, No. 7, April 1, 2010 2971 used in biochemical studies27,28,37,42-44 are usually below 0.1 platinum atoms per nucleotide (further referred to as rb). Our experiments revealed rb ) 0.1 as the lowest degree of DNA modification reliably detectable via decrease of the peak Gox; changes of the signal reached at lower platination levels did not exceed relative experimental error (see below). We therefore tested the possibility of determination of the DNA platination level using an electrochemical signal specific for the platinum adducts instead of following diminution of the response of the unmodified DNA. Cyclic Voltammetry. We first measured AdTS cyclic voltammograms of unmodified and strongly cisplatinated denatured calf thymus DNA in AFP at the HMDE (Figure 1A). The unmodified DNA yielded two signals due to electrode processes undergone by the DNA bases: a cathodic peak CA at -1.51 V (corresponding to irreversible reduction of cytosine and adenine residues) and peak G (corresponding to chemically reversible oxidation of 7,8- dihydrogen guanine generated at the electrode upon guanine reduction at potentials <-1.6 V).2,5,45 For the cisplatinated DNA, the shape of the voltammograms was changed significantly. In the cathodic part, the negative current started to increase sharply around -1.2 V, resulting in complete obscuring of the peak CA. The negative current reached its maximum at -1.75 V, forming a wide peak. The anodic part of the CV displayed three waves (around -1.75, -1.45, and -1.3 V), and between -1.53 and -1.18 V it was going through higher negative current values than the cathodic part in the same region. Such behavior suggested a kinetic process coupled to reversible electron-transfer reactions,46,47 most likely catalytic hydrogen evolution accompanying redox processes of the platinum moieties which has been observed earlier as a common feature of platinum group metals and their compounds.48 The catalytic nature of the processes undergone by the cisplatin-modified DNA at the HMDE was further assessed in studies of the effects of parameters such as scan rate or pH, the results of which are shown and discussed briefly in the Supporting Information. As shown in Figure 1B, intensity of the catalytic currents responded to the degree of DNA modification. For rb ) 0.1, the cathodic waves on the anodic part of the CV were well-developed, Figure 1. (A) Adsorptive transfer stripping cyclic voltammograms ) but in contrast to rb 1.0, the broad peak on the cathodic part at HMDE of unmodified denatured calf thymus DNA (blue) and the around -1.75 V was not observed. A clear effect of DNA same DNA strongly modified with cisplatin (black): DNA concentration, cisplatination was observable at rb as low as 0.01 (Figure 1B). 20 µgmL-1; cisplatin/nucleotide ratio, rb ) 1.0; accumulation time, - Peak G produced by the cisplatinated DNA in the first cycle was 60 s; initial potential, 0.0 V; switching potential, 1.85 V; scan rate, 1.0Vs-1; background electrolyte (dashed curve), 0.3 M ammonium by about 30% lower than the same signal produced by the formate, 50 mM sodium phosphate, pH 6.9. Inset: details of peak G unmodified DNA and was shifted by 30 mV to more negative for the same samples. (B) Sections of AdTS CVs obtained for potentials. Notably, in contrast to the peak Gox due to guanine unmodified or cisplatinated calf thymus DNA: unmodified (blue); rb electrooxidation at carbon electrodes (Figure S1 in the Sup- ) 0.01 (green); rb ) 0.1 (red); rb ) 1.0 (black). Other conditions are porting Information), which was strongly depressed at the same as in panel A. (C) Effects of repeated potential cycling on the peak G intensity for unmodified (red) and strongly (rb ) 1) cisplatinated (black) DNA. Inset: details of peak G for unmodified DNA, first scan (blue), (42) Pivonkova, H.; Brazdova, M.; Kasparkova, J.; Brabec, V.; Fojta, M. Biochem. Biophys. Res. Commun. 2006, 339, 477–484. and cisplatinated DNA, first (black), third (green), and fifth (red) scan. (43) Brabec, V. In Progess in Nucleic Acid Research and Molecular Biology; Conditions as in panel A. Moldave, K., Ed.; Academic Press Inc.: San Diego, CA, 2002; pp 1-68. (44) Kasparkova, J.; Pospisilova, S.; Brabec, V. J. Biol. Chem. 2001, 276, 16064– degree of DNA cisplatination, the peak G at the HMDE was only 16069. partially decreased. Moreover, the intensity of peak G produced (45) Trnkova, L.; Studnickova, M.; Palecek, E. Bioelectrochem. Bioenerg. 1980, 7, 644–658. by the cisplatinated DNA at HMDE increased with the number (46) Bond, A. M. Modern Polarographic Methods in Analytical Chemistry; Marcel of successive potential cycles between 0 and -1.85 V, and starting Dekker: New York, 1980. from the third cycle it was even higher than observed for the same (47) Nicholson, R. S.; Shain, I. Anal. Chem. 1965, 36, 706–723. (48) Heyrovsky, J.; Kuta, J. Principles of Polarogarphy; Czechoslovak Academy but unmodified DNA under the same conditions (Figure 1C). Such of Sciences: Prague, Czechoslovakia, 1965. behavior suggests that during their electrochemical reduction, the

2972 Analytical Chemistry, Vol. 82, No. 7, April 1, 2010 Figure 2. Monitoring of selective cisplatination of probe oligonucleotides in the presence of competitor plasmid DNA using magnetoseparation and AdTS CV. (A) Separation of the cisplatinated ODN probes using magnetic beads. A mixture of either noG or G25 ODN (20 µgmL-1), -1 comprising a stretch tested for formation of the platinum adducts and an A20 adaptor, with plasmid competitor DNA (20 µgmL ), was treated with different concentrations of cisplatin in 40 µL of 0.1 M NaClO4. After overnight incubation at 37 °C in the dark, 20 µL aliquots of the reaction mixtures were withdrawn, mixed with the binding buffer to a total volume of 40 µL, and incubated with the DBT in a shaker at 20 °C for 30 min. After binding, the beads were washed and transferred into 20 µL of deionized water. ODNs were released from the DBT by heating at 85 °C for 2 min. The recovered ODNs were analyzed by AdTS CV. (B) Sections of AdTS CVs obtained for the ODNs noG (black) or G25 (red) treated with cisplatin (rb ) 0.1) in the mixture with plasmid DNA. Inset: sections of anodic parts of AdTS CVs obtained for G25 treated with various concentrations of cisplatin corresponding to rb values of 0, 0.01, 0.02, 0.04, 0.06, 0.08, and 0.1 (increasing rb is denoted by the arrow). (C) Dependence of the current value measured at the anodic part of the AdTS CV at -1.3 V (as indicated by dashed line in panel B, inset) on the concentration of cisplatin used for modification of the ODN probes in the presence of plasmid DNA: noG (black); G25 (red). Other conditions are as in Figure 1. cisplatin adducts were decomposed with concomitant regeneration all ODNs showed considerable catalytic currents, including of guanine residues in DNA adsorbed at the electrode (in contrast, the noG ODN lacking any guanine residue (see Figure S3 in the during DNA oxidation at the PGE the adducts were not decom- Supporting Information). This observations suggests that the posed and the peak Gox intensity was close to zero for the same single-stranded ODNs were forming various types of platinum level of calf thymus DNA modification). Our preliminary results adducts regardless of the presence or absence of guanine- (not shown) further suggest that reduction of guanine at containing sequence motifs, and that in the absence of guanine HMDE (to 7,8-dihydroguanine whose electrooxidation back to residues other nucleobases accommodated the platinum moieties. guanine is reflected in the peak G2,5,45) is facilitated by the Therefore, we further performed modification of the ODNs in presence of products of cisplatin reduction at the HMDE, probably the presence of plasmid DNA serving as an indifferent competitor through a mechanism involving the catalytic hydrogen evolution DNA (featured by pBSK(-) plasmid offering multiple guanine (more details will be published elsewhere). motifs for the formation of the stable bifunctional adducts). In The main goal of this work was to develop a simple electro- such situation, distribution of the cisplatin adducts between the chemical technique suitable for the determination of biochemically given probe ODN and the competitor plasmid DNA was relevant levels of DNA modification with platinum complexes, expected to reflect occurrence of motifs forming preferentially useful also in studies of selective platination of various DNA stable cisplatin adducts. Mixtures of either noG or G25 ODN substrates differing in the content of guanine and/or nucleotide with the competitor plasmid DNA were treated with various sequence. We therefore compared CV responses of ODNs concentrations of cisplatin, followed by separation of the ODN differing in the content and sequence context of guanine residues probes from mixtures with the modified plasmid DNA using (see Table 1) modified with various concentrations of cisplatin. DBT (Figure 2A) and AdTS CV measurements (Figure 2, parts

Each ODN possessed an A20 stretch designed for the separation B and C). In agreement with the above assumption, the noG ODN of the modiﬁed ODN probe from the reaction mixture using displayed no signiﬁcant changes in its CV responses after 6,10 magnetic beads bearing T25 capture probes (DBT). However, treatment with 0-12 µM cisplatin (corresponding to rb values after cisplatination of individual ODNs and magnetic separation, between 0 and 0.1, as related to total DNA in the mixture) (Figure

Analytical Chemistry, Vol. 82, No. 7, April 1, 2010 2973 2, parts B and C). On the other hand, the G25 ODN, comprising numerous sites for formation of stable bifunctional cisplatin adducts, exhibited a clear dependence of the catalytic currents on the concentration of cisplatin used for the DNA modification. Negative current measured within the wave at -1.3 V on the anodic part of the CV (inset in Figure 2B) was increasing linearly with the DNA cisplatination level in the rb range between 0 and 0.1 (Figure 2, parts B and C). Modification of the competitor plasmid DNA was tested in parallel by cleavage with a restriction endonuclease Msp I specific for CCGG tetranucleotide. Preferential DNA cisplatination within the GG doublets resulted in inhibition of the DNA cleavage, which was monitored by agarose gel electrophoresis. These experiments confirmed modification of the indifferent competitor in the presence of any of the ODN probes. Moreover, the extent of restriction cleavage inhibition was reflecting inversely the distribution of cisplatin between the probe and competitor DNA in dependence on the probe reactivity, as assessed by the voltammetric measurements (Figure 2, parts B and C; results of the Msp I digestion experiments are shown in the Supporting Information, Figure S4). Square-Wave Voltammetry. The CV studies revealed catalytic currents at the HMDE during the anodic polarization to respond proportionally to the DNA cisplatination level. To obtain better developed signal specific for the DNA modification, we used anodic AdTS SWV instead of CV in the following experiments (Figure 3). Denatured calf thymus DNA modified with cisplatin to rb 0.01, 0.05, and 0.1 was prepared as above and measured by the AdTS SWV under conditions previously optimized49 for the measurements of peak G. The latter signal, occurring at -0.26 V, was the only peak observed on the voltammogram corresponding Figure 3. AdTS square-wave voltammetry of single-stranded calf -1 to the unmodified DNA (black curve in Figure 3A). DNA modified thymus DNA (20 µgmL ) treated with cisplatin. (A) AdTS SWV at HMDE: unmodified DNA (black); rb ) 0.01 (green); rb ) 0.05 (blue); with cisplatin produced, in addition to peak G, another distinct rb ) 0.1 (red); 6 µM cisplatin in the absence of DNA (corresponding signal at -1.25 V (in Figure 3A denoted as peak P), the intensity to rb ) 0.1 in samples with DNA). SWV: amplitude 50 mV, frequency of which was about proportional to the DNA modification level. 200 Hz, initial potential -1.85 V, quiescent time 2 s, final potential Cisplatin alone, at the same concentration as was corresponding 0.0 V; background electrolyte, AFP. Other conditions are as in Figure ) to rb ) 0.1 when used for DNA modification in this experiment, 1. Inset: AdTS SWV at m-AgSAE: unmodified DNA (black); rb 0.1 (red); parameters of the measurements were the same as used for yielded only a faint signal at a potential close to that of peak P, HMDE. (B) Comparison of the effects of various degrees of cisplati- suggesting a weak adsorption of the DNA-unbound platinum nation on AdTS peak Gox due to guanine oxidation measured at a complex at the HMDE surface and its efficient removal and pyrolytic graphite electrode (empty), peak G measured at HMDE separation of the firmly adsorbed DNA50 during the washing steps (blue), and peak P due to platinum adducts measured at HMDE (red). (even at a concentration an order of magnitude higher, the response of cisplatin alone was negligible, not shown). Similarly We tested also the possibility of using a mercury meniscus- as observed in the CV measurements (Figure 1), the peak G modified solid silver amalgam electrode instead of the HMDE in tended to shift toward more negative potentials (Figure 3A). To measurements of the cisplatin-modified DNA. The m-AgSAE optimize parameters for the SWV measurements, we followed proved a potent substitute for the HMDE in many analytical SWV responses of the cisplatinated DNA as functions of pulse applications, including nucleic acids studies.22,51,52 Besides sensi- amplitude, frequency, and pH of background electrolyte. We tive determination of purine nucleobases52 and label-free electro- selected an amplitude of 50 mV and a frequency of 200 Hz as chemical analysis of natural DNAs,51 it was successfully applied values giving rise to the best developed, symmetrical, and from also for the measurements of catalytic response of DNA modified background separated peak P. Although the peak P intensity was with an osmium tetroxide complex.22 The inset in Figure 3A shows increasing with pH shifting to acid values (in the Britton-Robinson a well-developed peak P of the cisplatin-modified DNA (rb ) 0.1) buffer) in agreement with the involvement of catalytic hydrogen when measured by AdTS SWV at the m-AgSAE, suggesting the evolution in the electrode process, we have chosen AFP as a latter to be applicable for this purpose as well. medium suitable for simultaneous measurements of both peak P In Figure 3B we compare effects of DNA cisplatination at levels and peak G (see below). Results of these experiments are shown corresponding to rb e 0.1 on the peak Gox measured at a PGE, in detail in the Supporting Information (Figure S5). (51) Fadrna, R.; Yosypchuk, B.; Fojta, M.; Navra´til, T.; Novotny, L. Anal. Lett. (49) Jelen, F.; Tomschik, M.; Palecek, E. J. Electroanal. Chem. 1997, 423, 141– 2004, 37, 399–413. 148. (52) Yosypchuk, B.; Heyrovsky, M.; Palecek, E.; Novotny, L. Electroanalysis (50) Palecek, E.; Postbieglova, I. J. Electroanal. Chem. 1986, 214, 359–371. 2002, 14, 1488–1493.

2974 Analytical Chemistry, Vol. 82, No. 7, April 1, 2010 concentration (Figure 4). This observation was in accord with our expectations since the latter ODN comprised adjacent guanine residues for formation of the 1,2-GG IAC featuring the most frequent of DNA modification with cisplatin. Finally, we designed a 50-mer double-stranded ODN ds50 (Table 1) comprising a target binding site for tumor suppressor protein p53 (p53CON) and an Msp I restriction site (see scheme in Figure 5A). The sensitivity of Msp I cleavage within the CCGG site to DNA cisplatination has been mentioned above. The p53CON involved a single GG doublet (Figure 5A). Previously it was shown that DNA cisplatination inhibits p53 binding to target p53CONs containing cisplatin-reactive motifs.27,44,54 Here we modified the ds50 ODN with various concentrations of cisplatin (corresponding to rb values 0.02, 0.04, and 0.08) and followed effects of this modification on the Msp I cleavage (Figure 5B), p53-DNA binding (Figure 5C), and voltammetric response of the cisplatin adducts represented by the SWV peak P. Figure 5D Figure 4. Monitoring of selective cisplatination of probe ODNs in shows relative correlation between these three features. Intensity the presence of competitor plasmid DNA using magnetic separation of the peak P was increasing with the cisplatin concentration, albeit and AdTS SWV. Dependences of the peak P on the rb obtained for following a significantly sublinear trend (whereas the cisplatin noG (black), GTT (blue), and GGT (red). The ODNs were treated concentration increased in the ratio 1:2:4, the peak P intensity with cisplatin in the presence of plasmid competitor DNA and was only in the ratio 1:1.7:2.2). This nonlinearity was nevertheless separated using DBT as in Figure 2. For other details see Figure 3A. natural in this case considering the facts that the ds50 ODN was modified in the absence of any other DNA and that the rb was peak G and peak P (both measured at the HMDE). Whereas calculated relative to all nucleotides in the ds50 ODN, whereas peak Gox at PGE exhibited certain (albeit within the experi- only three preferentially modified guanine doublets were present mental error) decrease for rb ) 0.1, but no measurable changes in the ODN. Thus, three platinum moieties per ODN molecule for lower modification levels, peak G at the HMDE showed a were in principle sufficient to create the 1,2-GG IAC at all of these small increase in its intensity (not exceeding 10% within the doublets, which formally corresponds to rb ) 0.03 (the 50-mer given rb range). On the contrary, the cisplatination-specific duplex contains 100 nucleotides). Further increasing the cisplatin peak P, which was naturally absent on the voltammogram of concentration resulted in exceeding the equivalence point and in unmodified DNA, clearly indicated DNA cisplatination and the observed sublinearity. Binding of the p53 protein to the ds50 changes in its level within the given rb range. The peak P can 50-mer, giving rise the band R due to the protein-DNA complex thus be utilized for the determination of the extent of DNA in the electrophoretic mobility shift assay (Figure 5C), was modification at rb levels relevant for biochemical studies. gradually decreasing as the rb was increasing, inversely reflecting Owing to its weak sensitivity to DNA cisplatination at rb e 0.1, the trend exhibited by the peak P intensity. The Msp I cleavage, the peak G (which is measured simultaneously with the peak resulting in formation of the ds37 fragment (Figure 5A) and a P during a single voltammetric scan, see Figure 3A) can be corresponding band in polyacrylamide electrophoresis (Figure employed as an independent signal for the determination of total 5B), was strongly (practically by an order of magnitude) inhibited DNA concentration and normalization of the platination-specific at rb as low as 0.02, and no cleavage product was detected at signal per unit of DNA (with the limitation given by the require- higher rb values. The stronger effect of DNA cisplatination on ment for constant G content). the restriction cleavage, compared to the p53-DNA binding, Further, we used the AdTS SWV to monitor modification of accorded well with the presence of only one GG doublet in the the noG, GTT, and GGT ODN probes in the presence of the p53CON but two doublets in the Msp I site (one in the top and plasmid DNA competitor, separated after the cisplatin treatment the other in the bottom strand, see Figure 5A). Probability of the using the DBT (as in Figure 2). For the noG ODN, only a small formation of a 1,2-GG IAC within the Msp I site at lower cisplatin peak P was observed at rb ) 0.1, whereas practically zero signal concentrations was thus 2-fold, compared to probability of modi- was detected for lower cisplatination levels (Figure 4). GTT, fication of the GG doublet in the p53CON; at the same time, although not containing optimum sequence motifs identified as cisplatination of only one doublet in the restriction site was major targets for cisplatin in double-stranded DNA,43,53 exhibited sufficient for preventing the cleavage. When the trends in relative almost linear increase of the peak P intensity with increasing intensities of the peak P, the amount of ds50 ODN bound by p53 concentration of cisplatin in the reaction mixture. The flexible protein, and the amount of the Msp I cleavage product ds37 single-stranded (GTT)7 stretch was thus able to accommodate, (Figure 5D) are compared, one can see that about 2-fold amount in the presence of the plasmid DNA competitor, a remarkable of cisplatin bound to the ODN substrate was required to cause number of platinum moieties yielding in SWV the peak P. The the same relative effect on the p53-DNA binding, compared to best target for cisplatination was the GGT ODN probe, as the restriction cleavage inhibition. Taken together, we demon- revealed by the most intense peak P obtained for any cisplatin (54) Pivonkova, H.; Pecinka, P.; Ceskova, P.; Fojta, M. FEBS J. 2006, 273, 4693– (53) Jamieson, E. R.; Lippard, S. J. Chem. Rev. 1999, 99, 2467–2498. 4706.

Analytical Chemistry, Vol. 82, No. 7, April 1, 2010 2975 Figure 5. Correlation between biochemical effects of DNA cisplatination and the intensity of peak P due to the cisplatin adducts. (A) Scheme of a 50-mer ODN, bearing a target sequence of p53 protein (p53CON) and an Msp I restriction site. The ODN was labeled with 32P and modified to various levels (see panels B-D) with cisplatin. (B) Effect of cisplatination on the ODN cleavage with the Msp I enzyme. The ds50 band corresponds to full-length 50-mer, ds37 band to the cleaved ODN. (C) Effect of cisplatination on binding of the p53 protein to the 50-mer ODN (R band corresponds to the p53-DNA complex). (D) Bar graph comparing the effect of the cisplatin/nucleotide ratio (rb) on the relative intensity of AdTS SWV peak P (gray; value obtained for rb ) 0.08 taken as 100%), cleavability of the Msp I site (blue; relative intensity of the ds37 band, taken as 100% for rb ) 0), and the p53-DNA sequence-specific interaction (red; relative intensity of the R band, taken as 100% for rb ) 0). For AdTS SWV measurements, the 50-mer ODN was modified under the same conditions as used for the cleavage and binding experiments but was not radiolabeled. strate an excellent correlation between electrochemical response reasonable levels of DNA cisplatination (such as rb e 0.1), the of the cisplatinated DNA and biochemical consequences of the latter signal can be utilized as an internal control for the deter- DNA modification. mination of DNA concentration and normalization of the number of cisplatin adducts per DNA unit. The proposed approaches can CONCLUSIONS find application in studies of DNA modification with platinum We propose an electrochemical technique suitable for the complexes and development of novel metallodrugs as a simple determination of DNA modification with cisplatin at levels relevant and inexpensive alternative to other methods of determination of for studies of biochemical effects of the DNA platination. For the transition metals such as AAS or ICPMS. first time we demonstrate analytically useful catalytic current responses accompanying redox processes of the cisplatin moieties ACKNOWLEDGMENT in DNA, without any sample pretreatment, during anodic polariza- This work was supported by the Czech Science Foundation tion following prereduction of the cisplatinated DNA at HMDE (Grant 204/07/P476 to H.P.) and partially by the Grant Agency or m-AgSAE. Using a competition approach, relying in cisplatin of the ASCR (Grant IAA500040701 to M.F. and IAA400040903 to treatment of ODN probes in the presence of indifferent DNA and L.H.), by the ASCR (Grant 1QS500040581, institutional research magnetic separation of the probes, followed by AdTS voltammetric plans AV0Z50040507 and AV0Z50040702), and by the MEYS CR measurements, selective modification of various sequence motifs (LC06035). can simply be monitored. Results of our preliminary experiments with cisplatin analogues carboplatin and oxaliplatin (H. Pivonˇkova´, SUPPORTING INFORMATION AVAILABLE L. Teˇsnohlı´dkova´, and M. Fojta, unpublished) suggest that Results of experiments focused on the effect of parameters application of the technique is not restricted to cisplatin. Hence, such as scan rate (in CV), pulse amplitude, frequency (in SWV), the proposed approach can easily be adapted for different single- and pH of background electrolyte on the cisplatinated DNA or double-stranded ODN probes designed to form adducts with voltammetric responses, as well as results of cisplatin treatment various platinum complexes, as well as for testing reactivity of of ODN probes in the absence of competitor DNA and of various platinum complexes toward different DNA sequence independent enzymatic probing of the extent of competitor plasmid motifs. Variations in the intensity of the SWV peak P, providing DNA platination. This material is available free of charge via the information about changes in the extent of the modification of Internet at http://pubs.acs.org. double-stranded DNA substrate, correlated well with effects of the DNA cisplatination on the binding of the p53 as well as on the susceptibility of the DNA to cleavage with a restrictase Msp Received for review December 31, 2009. Accepted I. Since the intensity of an intrinsic DNA signal, peak G due to February 12, 2010. guanine measured by SWV, is not significantly affected by AC902987X

2976 Analytical Chemistry, Vol. 82, No. 7, April 1, 2010