9.3Photocontrol of RNA Processing

Steven G. Chaulk, Oliver A. Kent, and Andrew M. MacMillan

9.3.1Introduction

In addition to being a carrier of genetic information, RNA is involved in theregulation of both transcription and translation, plays a key role in the splicingof pre-mRNA in eukaryotes, and is responsible for catalysis of amide bond for-mation at the heart of the ribosome [1, 2]. The study of many RNA systems iscomplicated by the dynamic nature of RNA secondary and tertiary structures,the transient nature of RNA–RNA or RNA–protein complexes, and in somecases the chemical reactivity of the RNA itself. One useful approach to the studyof discrete RNA (and DNA) structures has been to limit available conformationsthrough site-specific intra-strand crosslinking [3–7]. A complementary approachwhich allows the isolation of specific RNA structures or complexes involves thetransient blocking or “caging” of RNA functional groups involved in the transi-tion between two different states. The caging of cofactors or reactive substrateswith photolabile groups has proven useful in a wide variety of investigationsranging from mechanistic enzymology to cell biology [8–10]. Caging of specificfunctionalities within proteins has been used to control reactivity in these sys-tems, e.g. Noren et al. caged a specific serine residue in Thermococcus litoralisVent polymerase for studies on the mechanism of protein splicing [11].

In the context of an RNA molecule, a caging approach might be used to blockeither chemical reactivity of the RNA or formation of secondary or tertiary struc-ture. The caged RNA system can be studied both before and after photolysis,thus permitting characterization of the two states and the transition betweenthem. The first application of the caging approach to studies of RNA structureand function has involved blocking the chemical reactivity associated with anRNA functionality, the 2�-hydroxyl group, since specific RNA 2�-hydroxyls act asnucleophiles in a number of biologically important transesterifications [12–17].Caging of a single 2�-hydroxyl in a short synthetic oligonucleotide blocks thecleavage reaction catalyzed by the hammerhead ribozyme; cleavage is initiatedby photolysis of the ribozyme–substrate complex [18]. This approach has alsobeen extended to the caging of the branch adenosine in a full-length pre-mRNAfor studies of RNA processing by the mammalian spliceosome [19]. Caging ef-fectively isolates spliceosome assembly from catalysis of the splicing transesteri-fications permitting a closer examination of the mechanisms of each.

9.3 Photocontrol of RNA Processing 495

9.3.2Caging the Hammerhead Ribozyme

The hammerhead system was chosen for preliminary studies because it hasbeen the subject of intense investigation culminating in high-resolution X-raystructures of both model and native hammerhead–substrate complexes [20–22].The hammerhead ribozyme is a site-specific RNA endonuclease derived fromself-cleaving plant viroid and satellite RNAs [12]. The ribozyme–substrate com-plex contains three base-paired stems and a central core region of 13 conservednucleotides which form the catalytic site (Fig. 9.3.1A). Cleavage of the substrateRNA (RNA 1 in Fig. 9.3.1A) is magnesium dependent and occurs through intra-molecular attack of a 2�-hydroxyl on the adjacent phosphodiester (Fig. 9.3.1B).As part of our efforts to understand the hammerhead system, many studieshave involved the use of altered reaction conditions or mutant ribozymes or

9 New Challenges496

Fig. 9.3.1 (A) Secondary structure of a cano-nical hammerhead–ribozyme·substrate com-plex showing the scissile phosphodiesterlinkage. (B) Intramolecular transesterifica-

tion catalyzed by the hammerhead ribozyme.(C) Photolysis of a 2�-caged hammerheadsubstrate to initiate the hammerhead cleav-age reaction.

substrates. For example, since the ribozyme-mediated cleavage is dependent onthe presence of a 2�-hydroxyl 5� to the site of cleavage, replacement of this func-tionality with hydrogen or OMe substituents has allowed the study of catalyti-cally inactive ribozyme–substrate complexes [20, 21]. Modification of this single2�-hydroxyl of the hammerhead with a photodissociable group would representa caging of this reactive functionality, thus permitting photocontrol of the ham-merhead mediated reaction (Fig. 9.3.1C).

For initial studies on the hammerhead system, the 2-nitrobenzyl group waschosen as the caging functionality because its photochemistry has been wellcharacterized and because it can be removed, by irradiation at 308 nm, underconditions which will not damage nucleic acids or proteins. The residue 5� tothe cleavage site in the hammerhead substrate can be a C, U or A residue.Caged adenosine was chosen for incorporation into a hammerhead substrate be-cause of applications of this monomer to studies of Group II self-splicing in-trons and pre-mRNA splicing. The nucleoside 2-nitro-benzyladenosine wassynthesized [23–25] and its reactivity studied under a variety of conditions.Upon photolysis at 308 nm with an excimer laser, the 2-nitrobenzyladenosinewas quickly, cleanly and quantitatively converted to adenosine as monitored byC18 reverse-phase high-performance liquid chromatography (HPLC) with theyield of uncaged adenosine independent of pH in the range of 6–8. Next, thecompatibility of the 2-nitrobenzyl modification with standard RNA synthesisconditions was assessed. While the 2-nitrobenzyl ether was stable to most condi-tions of RNA synthesis, it was partially removed by fluoride under the condi-tions used in deprotection of 2�-silyl protected RNAs (�20% cleavage observed).Thus the 2�-tert-butyldimethylsilyl (TBDMS) group commonly employed in RNAsynthesis was incompatible with the introduction of the caged monomer intoRNA. As an alternative, the acid labile [1-(2-fluorophenyl)-4-methoxypiperidin-1-yl (Fpmp)] group was used as a 2�-protecting functionality [26]. The 2�-Fpmpgroup is quantitatively removed upon treatment with mild aqueous acid over30–40 h, conditions under which the nitrobenzyl ether is stable, with RNA syn-thesis yields comparable with those obtained using the 2�-TBDMS group. In or-der to incorporate the caged monomer into RNA, the 2�-modified adenosine waselaborated into a 5�-dimethoxytrityl-3�-phosphoramidite by standard procedures.This monomer was used, along with 2�-Fpmp phosphoramidites, in the solidphase synthesis of the RNA 2: 5�-GGGUGUA*UGGUU-3� (Fig. 1A; a modifiedversion of a well-characterized hammerhead substrate in which A* represents acaged nucleotide 5� to the cleavage site [27]). Deprotection of the crude productwith methanolic ammonia followed by aqueous acid afforded an RNA in whicha single residue was modified at the 2� position with the desired caging func-tionality. The unmodified control oligonucleotide substrate 1: 5�-GGGUGUAUG-GUU-3� was synthesized by standard solid phase synthesis procedures and theribozyme 3 was synthesized by T7 transcription from a synthetic DNA template[28]. HPLC analysis of the purified caged RNA, 2, following enzymatic digestionwith snake venom phosphodiesterase and calf intestinal alkaline phosphatase,showed the presence of a single nitrobenzyl modified adenosine residue. Upon

9.3 Photocontrol of RNA Processing 497

photolysis of 2 at 308 nm, the caged adenosine was cleanly and quantitativelyconverted to adenosine with no changes in the composition of the oligonucleo-tide as monitored by denaturing polyacrylamide gel electrophoresis (PAGE) andHPLC of the digested RNA.

As expected, incubation of the control hammerhead substrate 1 with catalyticamounts of ribozyme 3 led to site-specific cleavage of 1 (Fig. 9.3.2A, lane 2)with kinetics similar to those reported for related substrates [27]. Under identi-cal conditions, the caged RNA substrate 2 was not cleaved (Fig. 9.3.2A, lane 5).Following photolysis, however, the uncaged RNA was site-specifically cleaved ina magnesium-dependent reaction to the same extent as the unmodified sub-strate 1 (Fig. 9.3.2A, lane 6; typically, 70–80% cleavage of either unmodified oruncaged substrate was observed). In order to further characterize the cagedRNA substrate 2, it was compared with the unmodified RNA substrate 1 in thepresence of saturating ribozyme. Under these conditions, following photolysis,the uncaged substrate was cleaved as quickly as the unmodified RNA substrate(kobs�0.2 min–1 for RNAs 1 and 2; Fig. 9.3.2B). In order to measure the effect,if any, of the caging functionality on the stability of the ribozyme–substratecomplex, equilibrium binding studies measuring the dissociation constants forinteraction of the ribozyme 3 with the caged RNA 2 and a modified version ofRNA 1 containing 2�-deoxy-adenosine 5� to the cleavage site were performed[29]. The Kds for both of the ribozyme complexes were measured to be�200 nM suggesting that the caging functionality does not appreciably disruptthe ribozyme–substrate complex. Together, these results demonstrate that the re-active 2�-hydroxyl functionality in a hammerhead substrate can be caged andthat an efficient and accurate hammerhead reaction is initiated upon photolysisof the caged substrate. Furthermore, neither the photolysis conditions nor thenitroso-aldehyde product of photolysis adversely affect the course of the reactioneven under saturating conditions. Finally, the presence of the caging functional-ity does not disrupt the ribozyme–substrate complex as evidenced by equilib-rium binding studies.

9.3.3Splicing

9.3.3.1 Photocontrol of Splicing using Caged pre-mRNAsPre-mRNAs in eukaryotes are characterized by a split-gene structure in whichcoding exon sequences are separated by non-coding intron sequences [30]. Theprocess by which the introns are excised from the pre-mRNA and the exons arejoined together is known as pre-mRNA splicing, and is catalyzed by the spliceo-some – a biochemical machine containing both protein and RNA components[31]. The spliceosome includes the U1, U2 and U4/U5/U6 small nuclear ribo-nucleoprotein particles (snRNPs), each containing a unique RNA and associatedproteins. The functional spliceosome is not formed independent of its substrate;instead the active splicing catalyst assembles in an ATP dependent, stepwise, or-dered fashion on pre-mRNA substrates through the A, B and C complexes

9 New Challenges498

9.3 Photocontrol of RNA Processing 499

Fig. 9.3.2 (A) Denaturing PAGE analysis of32P-end-labeled unmodified (RNA 1; S) andcaged (RNA 2; S) substrates and 7-nt prod-uct (P) of ribozyme-mediated cleavage underconditions of catalytic ribozyme. Lane 1, un-modified substrate; lane 2, unmodified sub-strate and ribozyme, t=120 min; lane 3, un-modified substrate and ribozyme, photo-lysed, t= 120 min; lane 4, caged substrate;lane 5, caged substrate and ribozyme,t= 120 min; lane 6, caged substrate and ribo-zyme, photolysed, t=120 min; lane 7, unmo-dified substrate and ribozyme in the absence

of magnesium, t= 120 min; lane 8, cagedsubstrate and ribozyme, photolysed, in theabsence of magnesium, t= 120 min. (B) Ki-netics of hammerhead catalyzed cleavage ofRNA 1 and RNA 2, following photolysis withan excimer laser, under conditions of satur-ating ribozyme: normalized percent cleavageas a function of time for RNA 1 (�) andRNA 2 (O); plots of ln fraction substrate pre-sent versus time give kobs values of0.20± 0.02 min–1 for RNA 1 and0.22± 0.02 min–1 for RNA 2.

(Fig. 9.3.3A) which may be visualized by native gel electrophoresis [32]. It wouldclearly be useful to separate spliceosome assembly from the subsequent chemi-cal steps of pre-mRNA splicing such that the requirements for each could bestudied independently.

The chemistry of pre-mRNA splicing involves two sequential transesterifica-tion reactions. During the course of spliceosome assembly, a specific adenosineresidue in the branch region of the intron becomes bulged from an RNA duplexformed between the pre-mRNA and the U2 snRNA. The 2�-hydroxyl of this resi-due carries out a nucleophilic attack at the 5� splice site to generate a free 5�exon and a cyclic intermediate containing a 2�–5� phosphodiester branch [33].Attack of the free 5� exon at the 3� splice site then yields ligated exons and a cy-clic intron product (Fig. 9.3.3 B). Removal of the nucleophilic hydroxyl of thebranch adenosine by placement of a 2�-deoxy residue at this position preventsbranch formation at this position [34].

The demonstration in the hammerhead system that the reactivity of a uniqueRNA 2�-hydroxyl can be transiently blocked by the presence of a photolabile ni-tro-benzyl ether at this position suggested that a similar strategy could be ap-plied to study of the splicing reaction. In order to prepare a caged pre-mRNA, a10-base oligomer was chemically synthesized in which the branch adenosinewas modified at the 2� position as an ortho-nitro-benzyl ether and all other 2� po-sitions were protected the Fpmp group. Following standard work-up, the puri-fied caged oligomer was ligated to products of T7 RNA polymerase transcriptionrepresenting the 5� and 3� portions of the PIP85.B pre-mRNA substrate in thepresence of a DNA-bridging oligonucleotide to yield a full-length pre-mRNAwith caged branch nucleotide (Fig. 9.3.4A) [34–36]. The resulting RNA con-tained a single 32P radiolabel that was 2 nt 3� of the caged nucleotide(Fig. 9.3.4 A) facilitating analysis by denaturing PAGE.

The reactivity of the caged RNA was examined by carrying out photolysis ex-periments using a xenon arc lamp as a light source followed by RNase A andRNase T1 digestion, and thin-layer chromatography (TLC) analysis of the result-ing RNA fragments. Although control RNAs were unaffected by photolysis, thecaging group was 90% removed upon irradiation to yield the free 2�-hydroxyl(Fig. 9.3.4B).

The behavior of the caged pre-mRNA under splicing conditions was tested byincubating it in HeLa nuclear extract in the presence of ATP and Mg2+ – bothrequired cofactors for pre-mRNA splicing. Products associated with the first andsecond transesterification reactions of splicing are typically detected after an ap-proximately 20-min lag time required for the assembly of the spliceosome.However, in the case of the caged pre-mRNA, a negligible amount of splicingproduct was observed (�3% due to some uncaging during handling;Fig. 9.3.5A). The formation of spliceosomes on the caged pre-mRNA was con-firmed by non-denaturing PAGE (Fig. 9.3.5B) [32]. Thus, while the caginggroup blocks the first chemical step of splicing, the caged pre-mRNA is stillable to direct formation of the spliceosome.

9 New Challenges500

9.3 Photocontrol of RNA Processing 501

Fig.

9.3.

3(A

)A

ssem

bly

ofin

divi

dual

snR

NP

part

icle

son

apr

e-m

RN

Ato

form

the

activ

esp

liceo

som

esh

owin

gco

nser

ved

5�sp

lice

site

(GU

),br

anch

site

(A)

and

3�sp

lice

site

(AG

)se

quen

ces

onth

epr

e-m

RN

A.

(B)

Sequ

entia

ltr

anse

ster

ifica

tion

reac

tions

cata

lyze

dby

the

splic

eoso

me.

Inth

efir

stst

ep,

the

2�-h

ydro

xyl

ofth

ebr

anch

aden

osin

ew

ithin

the

intr

onca

rrie

sou

ta

nu-

cleo

phili

cdi

spla

cem

ent

atth

e5�

splic

esi

tege

nera

ting

afr

ee5�

exon

and

acy

clic

(lar

iat)

inte

rmed

iate

.A

ttac

kof

the

free

5�ex

onat

the

3�sp

lice

site

then

yiel

dslig

ated

exon

san

da

lari

atin

tron

prod

uct.

In order to determine whether the spliceosomes formed on the caged substrateswere catalytically competent, caged pre-mRNA was incubated in HeLa extractunder splicing conditions, the resulting complexes were photolysed, aliquotswere removed over time and the RNA analyzed by denaturing PAGE. These ex-periments showed that uncaging of the pre-mRNA occurred cleanly within thespliceosome and furthermore that the two chemical steps of splicing proceededwithout a delay, consistent with the fact that the caged RNA is sequestered with-in a fully formed spliceosome (Fig. 9.3.5 C).

9 New Challenges502

Fig. 9.3.4 (A) Synthesis of pre-mRNAs con-taining a caged 2�-hydroxyl at the branch po-sition. A synthetic oligomer containing a sin-gle modified adenosine residue was ligatedto T7 RNA polymerase transcripts to yieldfull length caged pre-mRNA containing a sin-gle 32P radio-label (p*), right: denaturing

PAGE purification of ligation; (B) TLC analy-sis of photolysis of caged pre-mRNA. Lane1: unphotolysed caged pre-mRNA digestedwith RNase A and RNase T1; lane 2: photo-lysed uncaged pre-mRNA digested withRNase A and RNase T1.

9.3.3.2ATP Requirements for Splicing

The ATP requirements of splicing were examined by incubating caged pre-mRNA in HeLa extract in the presence of ATP to allow the ATP-dependent for-mation of the spliceosome. Following a 60-min incubation, the reaction mixturewas depleted of ATP by the addition of glucose and the enzyme hexokinase [37].The reactions were irradiated to uncage the adenosine nucleophile and productswere analyzed by denaturing PAGE. Although the first step of splicing was ob-

9.3 Photocontrol of RNA Processing 503

Fig. 9.3.5 (A) Splicing analysis of unmodi-fied and caged pre-mRNAs. Unmodified andcaged 32P-labeled pre-mRNAs were incu-bated under splicing conditions in HeLa cellnuclear extract for 60 min. Left: unmodifiedpre-mRNA incubated for 60 min under spli-cing conditions showing production of lariatintermediate (first chemical step) and lariatintron (second chemical step) products.Right: caged pre-mRNA incubated for 60 minunder splicing conditions showing no spli-cing products. (B) Spliceosome formationon uncaged and caged pre-mRNAs. Uncagedand caged 32P-labeled pre-mRNAs were incu-

bated under splicing conditions in HeLa cellnuclear extract for the indicated times to al-low formation of the A, B and C (spliceo-some) complexes, and analyzed by nativegel electrophoresis (4%; 80 : 1; 50 mM Tris-glycine). (C) Uncaging of pre-mRNA in HeLacell nuclear extract; denaturing PAGE analy-sis of 32P-labeled RNA. Left: caged pre-mRNA incubated for 60 min under splicingconditions; right: time course of splicing fol-lowing photolysis of pre-mRNA–spliceosomecomplexes showing production of lariat in-termediate (first chemical step) and lariat in-tron (second chemical step) products.

served under these conditions, depletion of ATP had the effect of blocking thesecond step (Fig. 9.3.6). Addition of excess ATP resulted in complete recovery ofthe second step of splicing in these reactions.

Although ATP is required for spliceosome assembly, depletion of ATP fromreactions containing fully formed spliceosomes did not prevent first step chem-istry. This suggests that the caged pre-mRNA was poised to perform the firsttransesterification step and further indicates that ATP hydrolysis is in no waycoupled to the nucleophilic attack at the 5� splice site. ATP depletion did preventthe second step of splicing which is in agreement with both genetic studies inyeast [38] and also the ATP requirement for second step chemistry in a trans-splicing system [39]. This requirement most likely reflects the ATP requirementof RNA helicases involved in the rearrangement of RNA structure for the sec-ond step [31].

9.3.3.3 Phosphatase Activity and Spliceosome AssemblyLamond et al. have performed a series of studies suggesting that protein phos-phorylation and dephosphorylation are critical for both spliceosome assemblyand the splicing process itself. Addition of the protein phosphatases PP1 toHeLa extract blocked the formation of E complex on a pre-mRNA during thecourse of spliceosome assembly [40]. However, if phosphatase activity wasblocked by the addition of tautomycin, okadaic acid or microcystin LR spliceo-some assembly proceeded through to C complex, as assayed by native gel elec-trophoresis, but the transesterifications themselves were blocked [41]. Specificinhibition of PP2A using low concentrations of okadaic acid resulted in normalfirst-step chemistry, but blocked the second step of splicing. Together, these re-sults suggest the requirement of a PP1-specific dephosphorylation immediately

9 New Challenges504

Fig. 9.3.6 ATP requirement following assem-bly of the spliceosome; denaturing PAGEanalysis of 32P-labeled RNA. Caged pre-mRNA was incubated in HeLa extract undersplicing conditions to form pre-mRNA–spli-ceosome complexes after which ATP was de-pleted (�ATP) from the reaction by the addi-tion of glucose and hexokinase. Left: caged

pre-mRNA incubated for 60 min under spli-cing conditions; right: time course of spli-cing following ATP depletion and photolysisof pre-mRNA–spliceosome complexes show-ing production of lariat intermediate (firstchemical step), but no lariat intron (secondchemical step) products.

prior to the first step of splicing and a PP2-catalyzed dephosphorylation betweenthe two steps of splicing.

In a preliminary series of experiments, the phosphatase requirements for pre-mRNA splicing have been examined using caged pre-mRNA. First, it was con-firmed that pre-incubation of HeLa nuclear extract with the phosphatase inhibi-tors microcystin or okadaic acid inhibited the splicing of an unmodified pre-mRNA (Fig. 9.3.7 A). In a second set of experiments, caged pre-mRNA was incu-bated in HeLa extract to allow spliceosome assembly to occur, high concentra-tions of microcystin or okadaic acid were added and the reactions were photo-lysed to remove the caging group. No inhibition of either step of splicing wasobserved (Fig. 9.3.7 B). Taken together with the observations of Lamond, theseresults indicate that a phosphatase activity for pre-mRNA splicing is active with-in the fully assembled spliceosome before the first step of splicing, but is not di-rectly coupled to that step. Second, if a phosphatase activity is required for thesecond step of splicing, the relevant dephosphorylation occurs before the firststep of splicing.

9.3.4Future Prospects

Caging of the 2�-hydroxyl in an RNA can be used to control RNA processing bythe hammerhead ribozyme and the catalytic center of the spliceosome. Thismethodology could be applied to studies of other systems in which a 2�-hydroxylis the reactive functionality including the Group II intron, the hepatitis �, hair-pin and VS ribozymes. Finally, the hydrogen bond donor/acceptors of the nu-

9.3 Photocontrol of RNA Processing 505

Fig. 9.3.7 Phosphatase requirements of thechemical steps of splicing using 32P-labeledpre-mRNA. (A) Generation of spliced prod-ucts (left) from unmodified pre-mRNA isblocked by the addition of either microcystin

LR or okadaic acid (right). (B) Neither mi-crocystin LR nor okadaic acid block the gen-eration of spliced products from a photo-lysed caged spliceosome.

cleic acid bases are the principal determinants governing most nucleic acid–nu-cleic acid and protein–nucleic acid interactions through both Watson-Crick andnon-Watson-Crick hydrogen bonding patterns [42]. Caging of these base func-tionalities will allow the disruption of specific recognition events allowing char-acterization of the formation of and transition between different nucleic acidcomplexes.

9.3.5Experimental

9.3.5.1 Caged Hammerhead Ribozyme

9.3.5.1.1 Synthesis of 2�-O-(2-nitrobenzyl)adenosine PhosphoramiditeChemical reagents were purchased from Aldrich Chemical Company. Silica gel(0.03–0.07 mm) for flash chromatography was purchased from Acros. Solventswere distilled as follows: tetrahydrofuran from Na/benzophenone; pyridine, di-chloromethane and triethylamine from CaH2. 1H-NMR spectra were collectedon a Varian Gemini-200 instrument (200 MHz) and 31P-NMR on a Varian Ge-mini-300 instrument (300 MHz). All FAB-HRMS analyses were performed on aVG ZAB-SE machine (Medical Science Mass Spectrometry Facility, University ofToronto). HPLC was performed using a Waters 501 HPLC pump and an analyti-cal C18 column (3.9–300 mm; Bondapak) with detection at 254 nm on a Waters486 detector.

9.3.5.1.2 2�-O-(2-nitrobenzyl)adenosineAdenosine (1.0 g, 3.75 mmol; evaporated 3 times from dry pyridine) was dis-solved in 34 mL of hot DMF. To this solution was added NaH (225 mg, 60% inoil, washed 3 times with hexanes) as a suspension in 4 mL of DMF. The result-ing solution was stirred at 0 �C for 45 min after which 2-nitrobenzylbromide(1.21 g, 5.6 mmol) in 2 mL of DMF was added. The reaction mixture was thenstirred at room temperature under nitrogen for 5 h after which it was pouredinto 375 mL of ice-cold H2O and stirred overnight. The resulting yellow precipi-tate was collected by vacuum filtration and concentrated in vacuo. The 1H-NMR(200 MHz) of the crude product was consistent with the literature [23]. Theyield was assumed to be 100% for the next step in the preparation.

9.3.5.1.3 2�-O-(2-nitrobenzyl)-N6-benzoyladenosineTrimethylsilylchloride (3.8 mL, 30 mmol) was added to a suspension of 2�-O-(2-nitrobenzyl)adenosine (1.5 g, 3.73 mmol) in 20 mL pyridine under nitrogen.The mixture was then stirred for 30 min at room temperature after which timebenzoyl chloride (2.4 mL, 20.6 mmol) was added and the reaction stirred for an-other 2.5 h. The resulting mixture was then cooled to 0 �C, H2O (4 mL) wasadded, the reaction was stirred for 5 min, NH4OH (33%, 8 mL) was added and

9 New Challenges506

stirring continued for an additional 30 min. The mixture was concentrated invacuo and subjected to silica gel flash chromatography in 5% MeOH/CH2Cl2(Rf = 0.3) to yield 647 mg of 24-O-(2-nitrobenzoyl)-N6-benzoyladenosine (43%from adenosine).

1H-NMR (CDCl3, 200 MHz): � (p.p.m.) 8.72 (s, 1H, H8), 8.12 (s, 1H, H2),8.04 (d, 2H, NO2-ArH), 7.85–7.82 (m, 1H, NO2-ArH), 7.61–7.28 (m, 6H, ArH),6.00 (d, 1H, 1�H), 5.09–4.91 (m, 2H, methylene), 4.71 (s, 1H, 3�H), 4.65 (d, 1H,2�H), 4.21 (s, 1H, 4�H), 3.87 (dd, 2H, 5�H). The 2� position of the nitrobenzylgroup was confirmed by an NOED experiment. FAB-HRMS: calculated forC24H22N6O7 (MH+), 507.1628; observed, 507.1612.

9.3.5.1.4 5�-O-(4,4�-dimethoxytrityl)-2�-O-(2-nitrobenzyl)-N6-benzoyladenosine-3�-O-(2-cyanoethyl-N,N-diisopropylamino) Phosphoramidite

5�-O-(4,4�-dimethoxytrityl)-2�-O-(2-nitrobenzyl)-N6-benzoyl-adenosine (520 mg,0.6 mmol) was dissolved in dry distilled THF (3 mL) under nitrogen in flame-driedglassware. Then Et(iPr)2N (0.71 mL, 4 mmol) and 2-cyanoethyl-N,N-diisopropyla-minochloro-phosphite (0.29 mL, 1.3 mmol) were added and the reaction was al-lowed to proceed for 6 h. The resulting mixture was poured into EtOAc(200 mL), washed 3 times with 5% NaHCO3 (100 mL), dried over MgSO4, concen-trated in vacuo and subjected to silica gel flash chromatography in 80% EtOAc/hex-anes (Rf = 0.5–0.6, both diastereomers) to yield 622 mg (92%) of 5�-O-(4,4�-di-methoxy-trityl)-2�-O-(2-nitrobenzyl)-N6-benzoyladen osine-3�-O-(2-cyano-ethyl-N,N-diisopropylamino) phosphoramidite.

1H-NMR (CDCl3, 200 MHz): � (p.p.m.) 9.55 (bs, 2H, NHi, NHii ), 8.61 (d, 2H,H8i, H8ii), 8.23 (d, 2H, NO2-ArHi, NO2-ArHii), 7.94 (s, 2H, 2Hi, 2Hii), 7.94 (d,4H, NO2-ArHi NO2-ArHii), 7.64 (d, 2H, NO2-ArHi, NO2-ArHii), 7.51–7.12 (m,30H, ArHi, ArHii), 6.80–6.74 (m, 8H, DMT-ArHi, DMT-ArHii), 6.26 (m, 2H,1�Hi, 1�Hii), 5.26–4.90 [m, 4H, methylene(i), methylene(ii)], 4.69 (m, 2H, 3�Hi,3�Hii), 4.46 (m, 2H, 2�Hi, 2�Hii), 4.39 (m, 2H, 4�Hi, 4�Hii), 4.09–3.99 (m, 4H,5�Hi, 5�Hii), 3.78–3.31 [m, 8H, ethylene, cyanoethyl(i), cyanoethyl(ii)], 3.70 [d,12H, methoxy(i), methoxy(ii)],2.49–2.33 [m, 4H, methine, iPr(i), iPr(ii)], 1.29–0.91[m, 24H, methyl, iPr(i), iPr(ii)].

31P-NMR (CDCl3, 300 MHz): �, 151.27 p.p.m., �, 150.94 p.p.m. (85% H3PO4

in H2O as external standard). FAB-HRMS: calculated for C54H57N8O10P (MH+),1009.4013; observed, 1009.4019.

9.3.5.1.5 Nucleoside Model StudiesA 200-ml solution of 2�-O-(2-nitrobenzyl)-adenosine (32 mM) in 20 mM Tris, pH8.0 in a Pyrex reaction vessel was irradiated with 100 pulses (308 nm;300 mJ pulse–1; full beam: 1 �3 cm) from a Lambda Physik EMG 201 MSC exci-mer laser. The reaction mixture was directly injected onto a C18 reverse-phaseHPLC column (Waters) and was analyzed using an elution gradient from

9.3 Photocontrol of RNA Processing 507

0.05 M TEAA (triethylammonium acetate) to 1 : 1 0.1 M triethylammonium ace-tate/acetonitrile (16 min; 1 mL/min). The chromatogram was monitored at254 nm, peaks were identified by comparison with authentic standards and con-version yields determined by integration of the peaks. Irradiation at300 mJ pulse–1 resulted in 100% conversion, while irradiation at 45 mJ pulse–1

resulted in 80–85% conversion.

9.3.5.1.6 T7 TranscriptionThe hammerhead ribozyme, 3: 5�-GGGACCACUGAUGAGGCCGUUAGGCC-GAAACACC-3�, was synthesized by in vitro transcription with T7 RNA polymer-ase from synthetic DNA templates. Transcription reactions (1 mL) contained40 mM Tris, pH 7.5, 5 mM DTT, 1 mM spermidine, 3 mM nucleoside triphos-phates (Pharmacia), 0.01% Triton X-100, 0.2 mM DNA template (from a tem-plate stock solution which contained 10 mM Tris, pH 7.5, 10 mM MgCl2 and2 mM DNA template containing the T7 promoter region), 2 U T7 RNA polymer-ase (Promega) and 20 mM MgCl2. The transcriptions were performed at 37 �Cfor 3 h after which time the reaction mixture was extracted with phenol/chloro-form/isoamyl alcohol and chloroform and then ethanol precipitated. The crudeproduct was then purified by denaturing 20% (19 : 1) PAGE.

9.3.5.1.7 Automated SynthesisAll oligonucleotide synthesis was carried out on an Applied Biosystems 392DNA/RNA Synthesizer. DNA phosphoramidites and 2�-TBDMS-protected RNAphosphoramidites, and 500-Å controlled-pore glass resin for the synthesis ofDNA and unmodified RNA substrate 1 were from CPG. The 2�-Fpmp RNAphosphoramidites and controlled-pore glass resin used in the synthesis of cagedRNA 2 were from Cruachem. Standard DNA and RNA synthesis cycles wereused in all cases except that the standard 1-mmol RNA synthesis cycle wasmodified to give a coupling time of 15 min. All phosphoramidites were 0.1 Min acetonitrile. The unmodified hammerhead substrate RNA 1: 5�-GGGU-GUAUGGUU-3� was synthesized using phosphoramidites with 2�-O-tert-butyldi-methylsilyl protecting groups on a 1-�mol scale. The RNA was cleaved from theresin and deprotected by treatment with saturated NH3/MeOH (1 mL) at 55 �Cfor 20 h. The supernatant was recovered and lyophilized to dryness. Deprotec-tion of the 2�-hydroxyls was carried out by treatment with tetrabutylammoniumfluoride (600 mL, 0.1 M in THF) at room temperature for 24 h. The oligonucleo-tide solution was added to 100 mL of 1 M triethylammonium bicarbonate,loaded onto a C18 cartridge (Waters/Millipore), washed with 20 mL of 20 mMtriethylammonium bicarbonate, eluted with 30% CH3CN/0.1 M triethylammo-nium bicarbonate and lyophilized to dryness. The resulting crude product wasthen purified by denaturing 20% (19 : 1) PAGE. The modified hammerhead sub-strate RNA 2: 5�-GGGUG-UA*UGGUU-3� (A* represents 2� caged adenosine)was synthesized using the 2�-O-(2-nitrobenzyl)adenosine phosphoramidite and

9 New Challenges508

2�-Fpmp phosphoramidites on a 1-mmol scale. The RNA was cleaved from theresin and deprotected by treatment with saturated NH3/MeOH (1 mL) at 55 �Cfor 20 h. The supernatant was recovered and lyophilized to dryness. The 2�-Fpmp groups were removed by treatment with 500 �L NaOAc (pH 3.25) for40 h at room temperature after which the mixture was neutralized with 500 �LTris buffer (3.15 M, pH 9). The mixture was ethanol precipitated, the residuelyophilized to dryness and the crude product purified by denaturing 20% (19 : 1)PAGE. The recovered RNA consisted of an 85 : 15 mixture of RNAs 2 and 1 andwas subjected to C18 reverse-phase HPLC to yield pure RNA 2 (gradient elutionfrom 9 : 1 0.05 M triethylammonium acetate/acetonitrile to 7 : 3 0.05 M triethyl-ammonium acetate/acetonitrile over 25 min; 1 mL/min). Yields of unmodifiedRNA 1 and modified RNA 2 were 20% as determined by UV absorption at260 nm.

9.3.5.1.8 Nucleoside Composition AnalysisEnzymatic RNA digests (37 �C, 8 h) of modified and unmodified RNAs wereperformed in 60 �L reactions containing RNA substrate (5 nmol), 0.2 mMZnCl2, 16 mM MgCl2, 250 mM Tris, pH 6.0, 0.2 U snake venom phosphodies-terase (Pharmacia) and 4 U calf-intestinal alkaline phosphatase (Boehringer-Mannheim). Following digestion, samples were injected onto a reverse-phaseC18 HPLC column (Waters) with a gradient elution from 0.05 M triethylammo-nium acetate to 1 : 1 0.1 M triethylammonium acetate/acetonitrile (16 min;1 mL/min). Peaks corresponding to U, G, A and the modified nucleoside 2�-O-(2-nitrobenzyl)adenosine were identified by co-injection of nucleoside standardswith the chromatogram monitored at 254 nm.

9.3.5.1.9 RNA PhotolysisRNAs (5 nmol) were photolysed in a Pyrex reaction vessel essentially as de-scribed above (308 nm; 140 pulses; 250 mJ pulse–1), enzymatically digested andanalyzed by reverse-phase HPLC as described above. Hammerhead·substrate so-lutions were cooled to 0 �C before flashing and were immediately transferred toa 30 �C bath following photolysis.

9.3.5.1.10 Catalytic Ribozyme Cleavage ReactionRNA was labeled in reactions (20 �L) containing: 10 pmol RNA, 70 mM Tris pH7.6, 10 mM MgCl2, 5 mM DTT, 1 U T4 polynucleotide kinase (New EnglandBiolabs) and 10 �L [�-32P]ATP (3000 Ci/mmol, 5 mCi/mL, NEN). Reactions werecarried out at 37 �C for 10 min, extracted with phenol/chloroform/isoamyl alco-hol and chloroform, and then ethanol precipitated. Excess substrate cleavage re-actions contained 40 nM ribozyme 3, 100 nM substrate (RNA 1 or 2), 50 mMTris pH 6.0 and 10 mM MgCl2 (or 10 mM EDTA for control reactions) in a reac-tion volume of 20 �L. Ribozyme and substrate solutions in reaction buffer were

9.3 Photocontrol of RNA Processing 509

heated to 90–95 �C for 1 min and then allowed to cool to 22 �C over 15 min afterwhich the substrate solution was brought to 10 mM in MgCl2 (or 10 mM inEDTA). The cleavage reaction was initiated by adding ribozyme to the substrateand then flashed, if required, as described above (308 nm; 140 pulses;250 mJ pulse–1). After 2 h, the reaction was stopped by the addition of 20 �L ofstop solution (0.05% bromophenol blue, 0.05% xylene cyanol, 8 M urea and100 mM EDTA) and heated to 90–95 �C for 1 min before loading onto 20%(19 : 1) polyacrylamide denaturing sequencing gels for quantification. Gels werequantified using a Molecular Dynamics PhosphorImager and ImageQuant soft-ware version 3.22.

9.3.5.1.11 Saturating Ribozyme KineticsHammerhead cleavage reactions were performed under saturating ribozymeconditions at pH 7.5. The cleavage reactions contained 7500 nM ribozyme 3,10 nM substrate (RNA 1 or 2), 50 mM Tris, pH 7.5 and 10 mM MgCl2 in a totalvolume of 100 �L. Separate solutions of ribozyme and substrate were prepared,heated to 90–95 �C for 1 min and allowed to cool over 5 min to 30 �C. Then bothsolutions were brought to 10 mM in MgCl2 and 50 mM in Tris, pH 7.5, andcombined to initiate the reaction. Reactions containing the modified hammer-head substrate were photolysed as described above (308 nm; 140 pulses;250 mJ pulse–1). Aliquots, from 0 to 60 min, taken from the reactions were com-bined with an equal volume of stop solution (0.05% bromophenol blue, 0.05%xylene cyanol, 8 M urea and 100 mM EDTA) and heated to 90–95 �C for 1 minbefore loading onto 20% (19 : 1) polyacrylamide denaturing sequencing gels forquantification. Gels were quantified using a Molecular Dynamics Phosphor-Imager and ImageQuant software version 3.22.

9.3.5.1.12 Measurement of Equilibrium Dissociation Constantsfor Ribozyme–RNA Complexes

Equilibrium dissociation constants (Kd values) for ribozyme–RNA complexeswere determined as reported elsewhere [29]. Binding experiments were carriedout under native conditions (50 mM Tris, pH 7.5, 10 mM MgCl2, 5% sucrose,0.02% bromophenol blue, 0.02% xylene cyanol) comparing the caged RNA 2with an RNA containing a single deoxy-adenosine residue 5� to the cleavage site.Binding reactions were allowed to equilibrate for 15 h at room temperature andresolved on 15% native polyacrylamide gels (19 : 1; 13–22–0.15 cm) in 50 mMTris acetate, pH 7.5, 10 mM magnesium acetate buffer (pre-run for 2 h at 6 Wfollowed by buffer exchange and then run at 6 W for 6 h at 4 �C). Gels werequantified using a Molecular Dynamics PhosphorImager and ImageQuant soft-ware version 3.22. Kd values were determined from Scatchard analysis of thebinding data.

9 New Challenges510

9.3.5.2 Caged Pre-mRNA

9.3.5.2.1 Preparation of Caged Pre-mRNAA 10-base oligoribonucleotide containing a single caged adenosine was synthe-sized and purified as described above. Full length caged pre-mRNA was synthe-sized by incubating 5�-phosphorylated synthetic branch oligomer (300 pmol)with upstream (300 pmol) and 5� 32P-labeled downstream (100 pmol) T7 RNAtranscription products representing the 5� and 3� portions of the PIP85.B pre-mRNA in the presence of a bridging DNA oligonucleotide and 40 U T4 DNA li-gase (60 mM Tris, pH 7.8, 20 mM MgCl2, 36 U ribonuclease inhibitor, 1.2 mMATP, 2.4% PVP-40, 5 mM DTT) at 30 �C for 3 h. Ligations were purified directlyby 15% PAGE. The products were visualized by autoradiography, extracted, dis-solved in double-distilled water and stored at –20 �C.

9.3.5.2.2 TLC Analysis of pre-mRNACaged pre-mRNAs containing a single 32P label two nucleotides 3� to the modi-fication were digested with 2 U each of RNase T1 and RNase A (20 �L reactionvolume, 10 mM Tris, pH 7) both before and after photolysis (irradiation at a dis-tance of 1 cm with a 1000 W Oriel Xenon arc lamp for 4 s in 5-mm Pyrex reac-tion vessels). Reactions were then concentrated and loaded onto a cellulose PEITLC plate (4 �10 cm, J.T. Baker) and eluted for 1 h in 79 : 19 : 1 saturated(NH4)2SO4/1 M NH4OAc/isopropanol. The air-dried TLC plate was exposed to aMolecular Dynamics phosphor screen and then scanned using a Molecular Dy-namics Storm 860 PhosphorImager.

9.3.5.2.3 Splicing and Photolysis of Caged pre-mRNARNAs (50–100 �103 c.p.m.) were incubated in 10 �L reactions containing 40%HeLa nuclear extract, 2 mM MgCl2, 20 mM KCl, 1 mM ATP, 5 mM creatinephosphate, 4 U ribonuclease inhibitor and 4 �g tRNA. Following a 60-min incu-bation period at 30 �C, splicing complexes were photolysed as described above.ATP depletions were effected by adding glucose to a final concentration of7 mM, adding 0.3 U of hexokinase and incubating at 37 �C for 10 min. Phospha-tase inhibition was effected by the addition to splicing extracts of either Micro-cystin LR (Sigma) or okadaic acid (Sigma) to a final concentration of 11 �M.During the course of splicing, aliquots were removed at various times, the reac-tions were quenched by extraction with phenol/chloroform/isoamyl alcohol,ethanol precipitated and then subjected to denaturing PAGE (15%, 29 : 1). Driedgels were exposed to a Molecular Dynamics phosphor screen which was thenscanned using a Molecular Dynamics Storm 860 PhosphorImager.

9.3 Photocontrol of RNA Processing 511

Acknowledgments

This work was supported by the Natural Sciences and Engineering ResearchCouncil of Canada (NSERC), by a grant from the Connaught Fund of the Uni-versity of Toronto, and by a Research Opportunity Award (Research Corpora-tion). A. M.M. acknowledges support from the Natural Sciences and Engineer-ing Research Council of Canada (NSERC) and the Alberta Heritage Foundationfor Medical Research (AHFMR). O.A. K. is supported by an AHFMR graduatefellowship.

9 New Challenges512

References

1 R. F. Gesteland, J. F. Atkins (eds), TheRNA World. Cold Spring Harbor Labora-tory Press, Cold Spring Harbor, NY,1992.

2 M.V. Rodnina, W. Wintermeyer, Curr.Opin. Struct. Biol. 2003, 13, 334–340.

3 A. E. Ferentz, G. L. Verdine, J. Am.Chem. Soc. 1991, 113, 4000–4002.

4 S.A. Wolfe, G. L. Verdine, J. Am.Chem. Soc. 1993, 115, 12585–12586.

5 C.R. Allerson, G. L. Verdine, Chem.Biol. 1995, 2, 667–675.

6 J. T. Goodwin, G.D. Glick, TetrahedronLett. 1994, 35, 1647–1649.

7 S. T Sigurdsson, T. Tuschl, F. Eck-

stein, RNA 1995, 1, 575–583.8 J. A. McCray, D. R. Trentham, Annu.

Rev. Biophys. Biophys. Chem. 1989, 18,270–270.

9 S.R. Adams, R. H. Symons, Annu. Rev.Biochem. 1992, 61, 641–671.

10 R. Y. Tsien, Annu. Rev. Physiol. 1993, 55,755–784.

11 S.N. Cook, W. E. Jack, X. Xiong, L.E.

Danley, J. A. Ellman, P.G. Schultz,

C. J. Noren, Angew. Chem. Int. Ed.1995, 34, 1629–1630.

12 O.C. Uhlenbeck, Nature 1987, 328,596–600.

13 A. D. Branch, H.D. Robertson, Proc.Natl. Acad. Sci. USA 1991, 88, 10163–10167.

14 J. M. Buzayan, W.L. Gerlach, G. Brue-

ning, Nature 1986, 323, 349–353.15 H.C. Guo, R. A. Collins, EMBO J.

1995, 14, 368–376.16 C.L. Peebles, P.S. Perlman, K.L.

Mecklenburg, M.L. Petrillo, J. H. Ta-

bor, K.A. Jarrell, H.L. Cheng, Cell1986, 44, 213–223.

17 M.M. Konarska, P. J. Grabowski, R. A.

Padgett, P.A. Sharp, Nature 1985, 313,552–557.

18 S.G. Chaulk, A. M. MacMillan, Nu-cleic Acids Res. 1998, 26, 3173–3178.

19 S.G. Chaulk, A. M. MacMillan, An-gew. Chem. Int. Ed. 2001, 40, 2149–2152.

20 H.W. Pley, K. M. Flaherty, D.B.

McKay, Nature 1994, 372, 68–74.21 W. G. Scott, J. T. Finch, A. Klug, Cell

1995, 81, 991–1002.22 W. G. Scott, J. B. Murray, J. R.P. Ar-

nold, B.L. Stoddard, A. Klug, Science1996, 274, 2065–2069.

23 E. Ohtsuka, S. Tanaka, M. Ikehara,

Chem. Pharm. Bull. 1977, 25, 949–959.24 D.G. Bartholomew, A. D. Broom, J.

Chem. Soc. Chem. Commun. 1975, 38.25 E. Ohtsuka, T. Wakabayashi, S. Tana-

ka, T. Tanaka, K. Oshie, A. Hasegawa,

M. Ikehara, Chem. Pharm. Bull. 1981,29, 318–324.

26 D.C. Capaldi, C.B. Reese, Nucleic AcidsRes. 1994, 22, 2209–2216.

27 D.M. Williams, W. A. Pieken, F. Eck-

stein, Proc. Natl Acad. Sci. USA 1992,89, 918–921.

28 J. F. Milligan, O.C. Uhlenbeck, Meth-ods Enzymol. 1989, 180, 51–62.

29 M.J. Fedor, O.C. Uhlenbeck, Biochem-istry 1992, 31, 12042–12054.

30 P.A. Sharp, Angew. Chem. Int. Ed.1994, 33, 1229–1240.

31 J. P. Staley, C. Guthrie, Cell 1998, 92,315–326.

32 M.M. Konarska, P. A. Sharp, Cell1987, 49, 763–774.

9.4 Light Reversible Suppression of DNA Bioactivity with Cage Compounds 513

33 M.M. Konarska, P. J. Grabowski, P. A.

Padgett, P.A. Sharp, Nature 1985, 313,552–557.

34 C.C. Query, M. J. Moore, P. A. Sharp,

Genes Dev. 1994, 8, 587–597.35 M.J. Moore, P.A. Sharp, Science 1992,

256, 992–997.36 A. M. MacMillan, C.C. Query, C.R.

Allerson, S. Chen, G.L. Verdine,

P.A. Sharp, Genes Dev. 1994, 8, 3008–3020.

37 L.A. Lindsey, A. J. Crow, M.A. Garcia-

Blanco, J. Biol. Chem. 1995, 270,13415–13421.

38 B. Schwer, C. Guthrie, EMBO J.1992, 11, 5033–5039.

39 K. Anderson, M.J. Moore, Science1997, 276, 1712–1716.

40 J. E. Mermoud, P.T. Cohen, A. I. La-

mond, EMBO J. 1994, 13 5679–5688.41 J. E. Mermoud, P. Cohen, A. I. La-

mond, Nucleic Acids Res. 1992, 20,5263–5269.

42 W. Saenger, Principles of Nucleic AcidStructure, Springer, New York, 1984.

9.4Light Reversible Suppression of DNA Bioactivity with Cage Compounds

W. Todd Monroe and Frederick R. Haselton

9.4.1Introduction

In this chapter we examine progress in photochemical control of DNA-depen-dent processes such as transcription, translation, DNA–DNA hybridization andDNA–protein interactions. Both in vitro and in vivo examples are described. Em-phasis is placed on the functional regulation of these processes and their utilityin biology, rather than on the chemical synthesis and photolysis of the cagedcomplexes, which is covered in greater detail in other chapters of this book. Asin other chapters, the basis of these efforts is the stable covalent modification ofDNA structure by photocleavable cage structures. The presence of the cagestructures suppresses or blocks the native DNA bioactivity of interest, but uponexposure to light native bioactivity is restored.

The native bioactivities of DNA are manifold and it would be impossible to at-tempt to describe all of the potential uses for caged DNA. Some examples aresketched in Fig. 9.4.1. In this chapter we focus on demonstrating several applica-tions, generally presented in order of increasing complexity. Reports documentingreversible changes in DNA structure are followed by examples of reversible block-ade of restriction enzyme digestion, in vitro translation, in vitro hybridization, cel-lular blockade of expression by exogenous plasmids, blockade of an antisenseknock down in cells and, finally, examples of blockade of expression in vivo.

To date, there have been few publications in this area. Our efforts to cageDNA with 1-(4,5-dimethoxy-2-nitrophenyl)diazoethane (DMNPE) were the firstdemonstration of photoactivated gene expression with cage compounds [1, 2]. A