Table Of Content4450–4463 Nucleic Acids Research, 2011, Vol. 39, No. 10 Published online 28 January 2011
doi:10.1093/nar/gkr025
Mapping interactions between the RNA chaperone
FinO and its RNA targets
David C. Arthur1, Ross A. Edwards1, Susan Tsutakawa2, John A. Tainer2,3,
Laura S. Frost4 and J. N. Mark Glover1,*
1Department of Biochemistry, University of Alberta, Edmonton, AB T6G 2H7, Canada, 2Life Sciences Division,
Department of Genome Stability, Lawrence Berkeley National Laboratory, Berkeley, CA 94720, 3Department of
Molecular Biology and The Skaggs Institute for Chemical Biology, The Scripps Research Institute, 10550 North
Torrey Pines Road, MB4, La Jolla, CA 92037, USA and 4Department of Biological Sciences, University of
Alberta, Edmonton, AB T6G 2E9, Canada
Received August 11, 2010; Revised January 10, 2011; Accepted January 12, 2011
ABSTRACT INTRODUCTION
Bacterial conjugation is regulated by two- Plasmid conjugation is a major mechanism of horizontal
component repression comprising the antisense gene transfer between bacteria, and is responsible for the
rapid spread of virulence and antibiotic resistance factors
RNA FinP, and its protein co-factor FinO. FinO
throughout bacterial populations. The study of the
mediatesbase-pairingofFinPtothe50-untranslated
conjugative transfer of F and F-like plasmids in
region (UTR) of traJ mRNA, which leads to transla-
Escherichia coli has provided important insights into the
tional inhibition of the transcriptional activator TraJ
mechanisms underlying this process (1,2). Cells newly
and subsequent down regulation of conjugation
infected with many of the plasmids from the F family
genes. Yet, little is known about how FinO binds to are initially competent to express the major plasmid tran-
its RNA targets or how this interaction facilitates scriptional unit, the tra operon, and efficiently transfer
FinP and traJ mRNA pairing. Here, we use solution plasmid torecipientcells. Gradually, however,tra expres-
methods to determine how FinO binds specifically sionisrepressed,andconjugationisattenuated.Theinitial
to its minimal high affinity target, FinP stem–loop II burst of conjugation spreads the plasmid throughout a
(SLII), and its complement SLIIc from traJ mRNA. population. The subsequent repression of the conjugation
machinery likely protects the plasmid-bearing cells from
Ribonuclease footprinting reveals that FinO
infection by pili-specific bacteriophage, and reduces the
contacts the base of the stem and the 30 single-
metabolic burden of plasmid maintenance.
stranded tails of these RNAs. The phosphorylation
Repression of conjugation relies on a plasmid-encoded
or oxidation of the 30-nucleotide blocks FinO
antisenseRNAsystemcalledFinOP.TheantisenseRNA,
binding, suggesting FinO binds the 30-hydroxyl of FinP, is complementary to the 50-UTR of traJ, the tran-
itsRNAtargets.Thecollectiveresultsallowthegen- scriptional activator of the tra operon (3–6) (Figure 1A
eration of an energy-minimized model of the FinO– and B). Binding of FinP to the traJ mRNA occludes the
SLII complex, consistent with small-angle X-ray ribosomalbindingsite,inhibitingtraJtranslationandpre-
scattering data. The repression complex model venting tra operon expression via H-NS de-silencing (7).
was constrained using previously reported cross- FinP alone, however, is unable to repress conjugation, in
linking data and newly developed footprinting part because it is rapidly degraded by RNaseE, a compo-
nentoftheE.coliRNAdegradosome(8,9).FinPrequires
results. Together, these data lead us to propose a
a second plasmid factor, the FinO protein, which binds
model of how FinO mediates FinP/traJ mRNA
FinP and stabilizes it against degradation (8,10). FinO
pairing to down regulate bacterial conjugation.
also binds the traJ 50-UTR, and facilitates FinP–traJ
*To whom correspondence should be addressed. Tel:+1 780 492 2136; Fax:+1 780 492 0886; Email: [email protected]
Correspondence may also be addressed to Ross A. Edwards. Tel:+1 780 492 4575; Fax:+1 780 492 0886; Email: [email protected]
The authors wish it to be known that, in their opinion, the first two authors should be regarded as joint First Authors.
(cid:2)TheAuthor(s)2011.PublishedbyOxfordUniversityPress.
ThisisanOpenAccessarticledistributedunderthetermsoftheCreativeCommonsAttributionNon-CommercialLicense(http://creativecommons.org/licenses/
by-nc/2.5),whichpermitsunrestrictednon-commercialuse,distribution,andreproductioninanymedium,providedtheoriginalworkisproperlycited.
NucleicAcidsResearch,2011,Vol.39,No.10 4451
RNA interactions, to ultimately repress conjugation 100- (Integrated DNA technologies) were designed to have
to 1000-fold. the necessary homologous recombination sites and a
FinO recognizes multiple stem–loop structures in both cleavage site for Prescission protease (GE Healthcare).
the FinP and traJ 50-UTR RNAs, binding with highest The sequences for the forward primers are: FinO WT
affinity to FinP SLII (11). Stem–loop recognition is inde- Fwd (50-GGG GAC AAG TTT GTA CAA AAA AGC
pendent of the precise RNA sequence, but requires 50 AGGCTTCCTGGAAGTTCTGTTCCAGGGGCC
and 30single-stranded tails adjacent to the duplex stem. CAT GAC AGA GCA GAA ACG ACC GGT A-30),
The FinO protein adopts a novel protein fold with an FinO 33W36A Fwd (50-GGG GAC AAG TTT GTA
extended a-helix and an unstructured N-terminal region CAA AAA AGC AGG CTT CCT GGA AGT TCT
(12) (Figure 1D). The C-terminal globular domain, GTT CCA GGG GCC CCC ACC AAA AGC GAA
FinO , which lacks the unstructured region and GGT GAA A-30), FinO 45 Fwd (50-GGG GAC AAG
45–186
much of the extended a-helix, comprises the core RNA TTT GTA CAA AAA AGC AGG CTT CCT GGA
binding domain (12,13). Site-specific protein–RNA AGT TCT GTT CCA GGG GCC CGA GAA GGC
cross-linking studies have indicated that residues on one TGC CCG GGA AGC AGA G-30). The sequence for
faceoftheglobulardomain,aswellasneartheN-terminal the reverse primer is: FinO 186 Rev (50-GGG GAC
tip of the extended a-helix, contact SLII (14). CAC TTT GTA CAA GAA AGC TGG GTC CTA
Perhaps the most intriguing aspect of FinO function is TCA TTG TTC ATC AAG CAC GGC CTG AAG
its ability to facilitate FinP–traJ RNA pairing (10,15,16). TTC-30). The primers were used to PCR off a
In spite of the perfect complementarities of these RNAs, pGEX-KG plasmid containing the FinO 1–186 gene
duplexing is extremely slow in the absence of FinO (13). The amplified inserts were recombined into
because of the significant energy barriers posed by the pDONR201andtransferredtothepDEST-15expression
highly stable internal stem–loops present in both RNAs. plasmid containing the glutathione-S-transferase gene.
FinO not only facilitates duplexing, but also promotes The sequence of the expression clones were verified for
strand exchange between a minimal SLII-like duplex sub- the correct FinO sequence by DNA sequencing
strate and a complementary single strand (15). This (Molecular Biology Services Unit, University of Alberta
suggests that FinO may act by destabilizing the internal Department of Biological Sciences).
duplex structure, thereby lowering the free-energy barrier
to intermolecular base-pairing interactions. Deletion Expression and purification of FinO constructs
analyses reveal that this activity requires the N-terminal
Expression of each FinO-GST construct was described
disordered region as well as the N-terminus of the
previously (13). Gel filtration chromatography was used
extended helix. A solvent-exposed tryptophan residue at
to further purify the FinO constructs. Fractions contain-
thetipoftheexposedN-terminalhelix(Trp36)appearsto
ing FinO were concentrated and loaded onto a Superdex
bethemostcriticalsingleresidueforbothRNAduplexing
7526/60gel-filtrationcolumn(GEHealthcare)whichwas
and strand exchange activities (15).
equilibratedin50mMHEPESpH7,200mMNaCl,1mM
Inthisarticle,weusedsolutiontechniquestodevelopa
EDTA, 5mM Tris(2-carboxyethyl) phosphine hydro-
structural model for the interaction between FinO and
chloride (TCEP) (Sigma-Aldrich). Purified FinO was
minimal, high-affinity target RNAs. Ribonuclease foot-
concentrated to the desired concentration and protein
printing reveals that wild-type FinO and the truncation
concentrations were calculated using extinction coeffi-
mutants, FinO W36A and FinO , all contact
33–186 45–186 cients at 280nm determined experimentally by amino
FinP SLII and traJ SLIIc RNAs in a similar manner
acid analysis (Alberta Peptide Institute): FinO WT
involving the base of the duplex stem and the 30 single- (27872M(cid:2)1cm(cid:2)1), FinO W36A (21263M1–(cid:2)118c6m(cid:2)1)
strandedtail.FinObindingisblockedbyphosphorylation and FinO (20469M3(cid:2)31–c1m86(cid:2)1).
of the 30-end of the RNA, or oxidation of the 30-ribose, 45–186
suggesting a free 30-hydroxyl group is essential for FinO
Preparation and labeling of RNA constructs for
binding. A complex of FinO with SLII was modeled
45–186 electrophoretic mobility shift and RNase
by docking the semi-flexible protein and RNA structures
footprinting experiments
inanautomated,energy-minimizingprocedureemploying
distance restraints defined directly from our biochemical SLII and SLIIc constructs (sequences shown in bold in
data. The model compared favorably with small-angle Figure 1B and A) were chemically synthesized using an
X-ray solution scattering data of the same complex. Applied Biosystems DNA synthesizer which was
These data and the structural model were used to modified for RNA synthesis using Dharmacon 20 ACE
explore the relationship between FinO binding and its chemistry (18). The 50-end-labeled RNAs were labeled
RNA chaperone activity. with g32P-ATP (6000Ci/mmol, Perkin Elmer) and T4
polynucleotide kinase (New England Biolabs). The
30-end-labeled RNAs were labeled with 30,50-cytidine
MATERIALS AND METHODS [50-32P] diphosphate (pCp) (Perkin Elmer) and T4 RNA
Ligase I (New England Biolabs). Both labeling reactions
Cloning of FinO constructs
were purified using denaturing gel electrophoresis. The
FinO constructs were cloned using Gateway Technology 50-end-labeled pellets were resuspended in 1(cid:3)TE pH
(Invitrogen) and the overlap extension PCR technique to 7.5, 10mM NaCl. The 30-end-labeled pellets were resus-
introduce mutations (17). Forward attB PCR primers pended in 50ml of ddH O. An additional T4 kinase
2
4452 NucleicAcidsResearch,2011,Vol.39,No.10
A
U G
G A
U G
20 U G
U - A
RBS
30 C - G
C - G
A A U U
U U A - U
U G A C
G - C G - C
G - C SLIII SLIIc A - U SLIc
U - A U - A
G - C 10 A - U
C - G 1 40 C - G
5'- G U U A A A A U U U G A A A U UG U A U C -3'
45
5' UTR of traJ mRNA
B C
20 20
C A
U C G - C
C A A - U
C A G - C
U - A G - C
C - G 30 G - C 30
C - G A - U
A A C - G
A - U G - C
G - U 10 10 U - A
G - C A - U
SLI A - U SLII G - C
U - A C - G
A - U U G
C - G 1 40 45 1 G - C 40 45
5' - G A U A -3' 5' -G A C A G A U U U U -3'
FinP SLII LV1
D W36A E
N
A
6
C ein WT W3
Prot 86 186 186
No 1-1 33- 45-
FinO-SLII
1 3345 62 186
SLII
Figure 1. Overview of constructs used in the study. (A and B) The secondary structures of traJ mRNA and FinP, respectively. The ribosomal
bindingsiteandstartcodonoftraJmRNAareboxed.ThederivativeRNAsoffocus,SLIIcandSLII,areshowninboldandnumberedaccordingly.
(C) The secondary structure SLII LV1 RNA, used to form a complex with the FinO in the SAXS experiments. SLII LV1 deviates from
45–186
wild-type SLII in its loop region, which is shown in bold. (D) The crystal structure of FinO highlighting the protein constructs used in the
26–186
experiments. The W36A mutation is shown as a ball on the structure. The N-terminal 32 amino acids were not observed due to disorder in the
structureandaredrawnin(dashedsection).BelowisascaledlinearrepresentationoftheprimarystructureofFinOshowingwhereeachconstruct
begins. The shades of gray correspond to the crystal structure coloring. (E) A representative 8% native PAGE demonstrating each of the FinO
constructs in a 1:1 molar complex with 50-32P-SLII.
NucleicAcidsResearch,2011,Vol.39,No.10 4453
dephosphorylation step was needed to remove the the same concentration as the footprinting experiments)
30-phosphatefromtheaddedcytidineofthe30-end-labeled where the concentration of RNase V1 or I was increased
RNAs which prevented binding of the FinO constructs. until non-specific cleavages resulted. The reactions were
The 30-dephosphorylated end-labeled RNAs were resus- incubated at the same temperature and length of time as
pendedin1(cid:3)TEpH7.5,10mMNaCl.RNAscontaining the footprinting experiments. A final RNase V1 concen-
a20,30 dialdehydeterminalriboseweregeneratedbytreat- tration of 0.001U/ml and RNase I concentration of
ment of the 30 dephosphorylated RNA with 10mM 0.01U/ml gave specific cleavages with an adequate signal
NaIO .Foreachperiodatereaction,1mlofdephosphoryl- to noise.
4
ated, annealed, 32P-SLII was added to 99ml of 10mM WeusedtwodifferentmethodstoassigntheRNaseV1
NaIO . The reactions were incubated at 4(cid:4)C for 40min cleavageproductswhich, duetoa30-OH, runslower ona
4
in the dark followed by ethanol precipitation. The pellets denaturing gel. First, for the 50-32P SLII RNase V1
were washed and resuspended in 7ml ddH O. The cleavage assays, we chemically synthesized short RNA
2
periodate-treated 32P-SLII resuspensions were then markers which had the same sequence as SLII and had a
added to 1ml 10(cid:3)structure buffer (500mM Tris–HCl 30-OH. SLII markers were 10 (50-GACAGUCGAU-30)
pH 7.5, 1M NaCl, 100mM MgCl ), 1ml of 2mg/ml and 15 (50-GACAGUCGAUGCAGG-30) nucleotides in
2
tRNA, and 1ml 10mM FinO (wild-type, 33-186 W36A length (Figure 2, left). The oligomers were synthesized,
or 45-186) or 1(cid:3) structure buffer for the no protein purified and labeled in the same manner as SLII and
reaction. Parallel binding reactions were performed for SLIIc(seeabove).TheothermethodusedT4polynucleo-
untreated, annealed, dephosphorylated 32P-SLII RNA tide kinase, in the absence of ATP, to remove the
using 1ml RNA, 1ml 10(cid:3) structure buffer, 1ml of 2mg/ 30-phosphate from the RNase T1 and alkaline hydrolysis
ml tRNA, 1ml 10mM FinO (wild-type, 33–186 W36A, or products (19).
45–186)or1(cid:3)structurebufferforthenoproteinreaction, To quantify the footprinting data, we first normalized
and 7ml ddH O. All 10ml binding reactions were the total counts in each FinO–RNA complex lane to the
2
incubated at 4(cid:4)C for 30min before adding 10ml of 20% total counts of the ‘No Protein’ lane to account for lane
glycerol and loading onto an 8% native polyacrylamide loadingdiscrepancies.Thenforeachbandofinterest,rep-
gel,equilibratedwith1(cid:3)tris–glycine,pH8.0at4(cid:4)C.Prior resentingapositionintheRNA,thefraction(f)ofthetotal
to use in the footprinting experiments, all labeled RNA counts for that band was calculated (f ) where p is a
p, i
stocks were annealed by heating to 95(cid:4)C for 1min FinO–RNA complex (either FinO, FinO W36A or
33–186
followed by slow cooling to room temperature. FinO ), and i is the nucleotide position of the RNA.
45–186
Electrophoretic mobility shift experiments were carried Wealsodeterminedthefractionofthetotalcountsforthe
out as described previously (15). ‘No Protein’ sample (fnp,i) at this position. Finally, we
divide f by f to get the protection factor which is
np,i p,i
RNase footprinting definedasthemagnitudebywhichtheRNAwasprotected
from RNase cleavage by each FinO. For each RNase
Native electrophoretic mobility shift assays were used to cleavage experiment, two independent reactions were
find 1:1 molar ratio FinO–RNA complexes prior to foot- performed and loaded onto the same gel. The values in
printingexperiments.ThebindingreactionsfortheRNase Figures2Band3Bareanaverageofthesetwoindependent
footprinting experiments were performed in 14ml reac- reactions. We decided on a protection value of two or
tions: 1.5ml of 10(cid:3) structure buffer (500mM Tris–HCl greater to represent a significant footprint. This is shown
pH 7.5, 1 M NaCl, 100mM MgCl2), 1.5ml of purified, inthefiguresasahorizontalruleacrossthegraph.
annealed 50 or 30-32P-SLII or SLIIc (in 10mM Tris–HCl
pH 7.5, 10mM NaCl, 1mM EDTA), 1.5ml of protein Preparation of samples for SAXS experiments
sample (at 10(cid:3) final concentration in 50mM HEPES
Forthesmall-angleX-rayscattering(SAXS)experiments,
pH 7, 200mM NaCl, 1mM EDTA, 5mM TCEP), 1.5ml
the FinO construct was the same as for the RNase
of 2mg/mL tRNA (Ambion), and 8ml ddH O. Reactions 45–186
2 footprinting studies. We used a full-length SLII construct
wereincubatedonicefor30minandaliquotsof5mlofthe
with a different loop sequence denoted SLII LV1 (the
reaction was removed and added to 5ml of 20% glycerol
sequence is shown in Figure 1C). The RNA was chem-
and loaded onto an 8% native gel to assay binding. One
ically synthesized and gel purified using the same proced-
microliter of RNase V1 or I (at the appropriate concen-
ures as for SLII and SLIIc (above). Precipitated RNA
tration; see below) was added to the remaining 9ml of the
samples were resuspended in 1(cid:3) TE pH 7.5, 10mM
binding reaction. RNase cleavage experiments were
incubated at 4(cid:4)C for 1h and stopped immediately with NaCl at the appropriate concentration and quantified
using extinction coefficients determined by an online
120ml of 0.3M NaOAc pH 5.3 and 130m l of phenol/
calculator (Ambion).
chloroform/IAA.Sampleswerethenchloroformextracted
andethanolprecipitated.Pelletswereresuspendedin4mL
SAXS data collection and analysis
of formamide gel load buffer and loaded on a 15%
urea-denaturing sequencing gel. Gels were exposed over- SAXS data were collected at beamline 12.3.1 of the
night and quantitated using ImageQuant software (GE Advance Light Source, Lawrence Berkeley National
Healthcare). To determine the optimal amount of Laboratory on a MAR165 detector (20,21), at a 1.6m
RNaseV1 orItoaddforthefootprinting assays,we per- sample to detector distance and a wavelength of 1.12A˚ .
formedcleavageassayswithend-labeledSLIIorSLIIc(at The hutch temperature was maintained at 17(cid:4)C during
4454 NucleicAcidsResearch,2011,Vol.39,No.10
A A A B
6 6
einWTW3 einWTW3
ut 05Prot86 186 186 ut Prot86 186 186 6 5' 32P-SLII
InpOHT1M1M1No 1-133-45- Inp No 1-133-45- n
ei
ot
39 Pr 4
o
32 N
30 o
15 e t 2
v
ati
el
20 V1 r U6 C7 G8 A9 U31G32C33A34U35C36G37G38C39
e
as 40
RN 3' 32P-SLII
15 m
30 o 30
n fr
o
10 34 ecti 20
ot
8 Pr
7 10
39
G32 C33 A34 U35 C36 G37 G38 C39 G40
5' 32P-SLII 3' 32P-SLII Nucleotide Position
Figure 2. RNase V1 cleavage of 50- and 30-end-labeled SLII in the absence and presence of various FinO constructs. (A) 15% urea-denaturing
polyacrylamidegelsshowingtheproductsofRNaseV1(0.001U/mlfinalconcentration)cleavagereactions.M10andM15aresynthesizedSLIIRNA
markersof10and15nucleotidesinlength(see‘MaterialsandMethods’section).ThelanesOHandT1representthealkalinehydrolysisandRNase
T1 cleavage of denatured SLII, respectively. SLII nucleotide positions are indicated at the right of the gels. Large and small arrowheads indicate
majororminorcleavagesbyRNaseV1inthepresenceofFinO.VerticalbracketsrepresentsignificantfootprintsonSLIIRNAresultingfromFinO
protectionofSLIIfromRNaseV1attack.(B)Bargraphsshowingthequantificationofthefootprintareasofthegelsin(A).Theleftaxisshowsthe
degreeofFinOprotectionfromRNaseV1relativetothe‘Noprotein’reaction.InblackisFinO WT,whiteisFinO W36A,andingrayis
1–186 33–186
FinO .Dataabovethehorizontaldashedruleineachgraphrepresentsignificantprotection((cid:5)2-fold)byFinO.Theblackbarsbelowthex-axis
45–186
highlightthefootprint.Theshiftinthefootprintwhen50-and30-end-labelingarecomparedislikelyduetotheeffectofaddingpCptothe30-endof
the RNA when 30-end-labeling. In Figure 6, we demonstrate that this subtly alters protein binding.
data collection. Solvent-only exposures of 7 and 70s for restraint-driven program HADDOCK2.0 (26,27). On the
background subtraction were taken from the dialysis protein, active ambiguous air restraints (AIRs) were
supernatant of each construct before and after each defined as residues found to form chemical cross-links
sample data collection. A sequence of consecutive expos- with the RNA as determined in Ghetu et al. (14) (Lys46,
ureslasting7,7,70and7swererecordedforeachsample Lys81, Arg121, Lys125, Arg165 and Lys176). Passive
and the appropriate background exposure subtracted and AIRs on SLII were those nucleotides determined by ribo-
circularly averaged using the program OGRE2 provided nuclease footprinting to be protected in the complex,
atthebeamline.Nosignofradiation-inducedaggregation namely nucleotides 7, 8 and 36–45 (numbering as SLII
was observed in the consecutive exposures. The program in Figure 7A). No passive AIRS were defined for FinO,
AUTORG(22)wasusedtocheckthelinearityoflog(I(s)) andnoactiveAIRsweredefinedforSLII.Defaultsettings
versus s2 at low angle (s<0.05A˚ (cid:2)1) and confirm the were used for most HADDOCK parameters. No AIRs
absence of aggregation. were randomly removed during separate docking trials.
The stem of SLII was further geometry-restrained to be
Protein–RNA docking an A-form helix with Watson–Crick base pairing. The 50-
AmodelofSLIIincludingthe50-and30-tailswasinitially and 30-tails of SLII were defined as both semi- and fully
constructed based on the NMR structure of the initiator flexible regions. Semi-flexible protein residues were auto-
tRNA anti-codon loop from yeast, a 21-nt RNA stem– maticallychosenbyHADDOCKiftheywerewithin5.0A˚
loop (RCSB accession code 1SZY) (23). The nucleotides of the RNA. A control experiment was carried out, in
weremutatedtocorrespondtothesequenceofSLIIusing which 2500 docking trials used randomly defined AIRs
COOT (24) and the stem–loop geometry-regularized in with random removal of 50% of AIRs per trial.
CNS (25). A model for FinO was derived by Filtering of all docked models against the SAXS data
45–186
removing residues 33–44 from the crystal structure used the program CRYSOL (28) to calculate theoretical
(RCSB accession code 1DVO). The docking of SLII to scatteringcurvesandobtainvaluesofR andD forthe
G max
FinO was performed using the chemical models. The theoretical scattering curves were fit to the
45–186
NucleicAcidsResearch,2011,Vol.39,No.10 4455
A A A B
6 6
ut Protein86 WT186 W3186 Protein86 WT186 W3186ut 5 5' 32P-SLIIc
InpOHT1No 1-133-45- No 1-133-45-Inp ein 4
ot
Pr
37 o 3
N
32 14 e to 2
30 v
ati
el
20 1 r C30 C31 C32 U33 G34 C35 A36 U37 C38 G39
V
e 14
s
20 Na 12 3' 32P-SLIIc
R
30 m 10
o
15 n fr 8
o
cti 6
e
36 ot
Pr 4
*
10
2
40
G34 C35 A36 U37 C38 G39 A40
5' 32P-SLIIc 3' 32P-SLIIc Nucleotide Position
Figure 3. RNase V1 cleavage of 50- and 30-end-labeled SLIIc in the absence and presence of various FinO constructs. (A) 15% urea-denaturing
polyacrylamide gels showing the products of RNase V1 (0.01U) cleavage reactions. The lanes OH and T1 represent the alkaline hydrolysis and
RNase T1 cleavage of denatured SLIIc, respectively. SLIIc nucleotide positions are indicated at the right of the gels. Large and small arrowheads
indicatemajororminorcleavagesbyRNaseV1inthepresenceofFinO.VerticalbracketsrepresentsignificantfootprintsonSLIIcRNAresulting
fromFinOprotectionofSLIIcfromRNaseV1attack.TheasteriskatpositionC10marksveryweakprotectionofthelowerstematthe50-endof
SLIIc. (B) Bar graphs showing the quantification of the footprint areas of the gels in (A). The left axis shows the degree of FinO protection from
RNase V1 relative to the ‘No protein’ reaction. In black is FinO WT, white is FinO W36A, and in gray is FinO . Data above the
1–186 33–186 45–186
horizontal dashed rule in each graph represent significant protection ((cid:5)2-fold) by FinO. The black bars below the x-axis highlight the footprint.
experimental SAXS data using CRYSOL in the range footprintingexperimentstodeterminewhichareasofthese
0<s<0.3A˚ (cid:2)1 to provide a goodness-of-fit value, (cid:2)2. stem–loop target RNAs are contacted by FinO. We
The top 250 models, ordered on fit to the data, were sub- compared interactions made by full length FinO, with
jected to pair-wise cluster analysis based on structural those of two mutants in which RNA strand exchange
similarity. The RMSD cutoff of 13A˚ was chosen as and duplexing have been compromised (Figure 1D).
small as possible while still providing enough models in FinO lacks part of the N-terminal region
33–186 W36A
a cluster to be meaningful. The five clusters obtained had required for RNA duplexing/strand exchange, and also
their members aligned using the maximum-likelihood bears a substitution at the critical Trp36 residue. This
superpositioning program THESEUS (29). variant binds RNA with a significantly enhanced affinity,
butitsstrandexchangeandduplexingactivitiesarereduced
Calculation of electrostatic potentials compared to wild-type FinO. FinO corresponds to
45–186
the core RNA binding domain and binds SLII with a
ElectrostaticpotentialswerecalculatedusingtheAdaptive
20-fold higher affinity than wild-type FinO, but has no
Poisson–Boltzmann Solver (30) and the PARSE force
detectablestrandexchangeorRNAduplexingactivity.
field. PDB2PQR was used to prepare the RCSB file
RNase V1 (32) was used to map interactions between
1DVO for this calculation (31).
thevariousFinOproteinsanddouble-strandedorstacked
single-stranded regions of RNA targets. SLII and SLIIc
RESULTS RNAswereradiolabeledatthe50-or30-endandsubjected
to limited RNase V1 digestion alone or in complex with
FinO binds to the lower duplex region of SLII and SLIIc
FinO at a 1:1 molar ratio. Native gel electrophoresis was
In order to facilitate FinP–traJ mRNA pairing, FinO used to ensure complex formation under these conditions
must first bind to its target RNAs. Previous studies have (Figure1E).TheresultsoftheRNaseV1cleavageexperi-
shown that FinO binds specifically and with high affinity ments with 50 and 30-end-labeled SLII are shown in
totheSLIIandSLIIcdomainsofFinPandtraJmRNAs, Figure 2A. In the absence of FinO, strong cleavage was
respectively(Figure1AandB)(11,13).Weusedenzymatic observed on both strands of the duplex stem, consistent
4456 NucleicAcidsResearch,2011,Vol.39,No.10
with previous ribonuclease mapping carried out on intact at position C10 (marked by an asterisk in Figure 3A, left
FinP (33). FinO binding protected residues G32-C39 on gel). Both 50- and 30-end-labeled SLIIc gels have strong
50-32P-SLII (left gel) and residues A34-C39 on 30-32P-SLII RNase V1 cleavages at the upper portion of the stem on
(right gel) corresponding to the lower 30-portion of the the 50 side. These cleavages at positions C14 to G18, with
SLII double-stranded stem region. A minor footprint weaker cleavages nearby, were observed in the presence
was observed on the 50-32P-SLII gel at positions C7 and of all FinO constructs. The upper 30 portion of the
G8, which maps to the lower 50-portion of the SLII stem. SLIIc stem had weaker cleavages from C28 to G34. The
Essentiallyidenticalprotectionpatternswereobservedfor results indicate that SLIIc binds to FinO and its deriva-
all three FinO constructs. Quantification of the footprint tives in a similar fashion as SLII, with the interaction
regions is shown in Figure 2B. For this study, we chose occurring at the lower part of the SLIIc stem region,
protectionvaluesgreaterthanorequalto2-foldrelativeto leaving the upper half of the stem exposed to cleavage
the ‘No Protein’ sample as significant (see ‘Materials and by RNase V1.
Methods’ section).
AnareaofintenseRNaseV1cleavagewasobservedon
the50-sideoftheSLIIstemproximaltotheloop(residues FinO does not bind to the loop region of its target RNAs
G11–G14,withminor cleavages atU10andG15–G16) in Itwaspreviouslyshownthatdeletionofthesevennucleo-
the presence of all FinO constructs. Minor cleavages also tide loop from SLII does not affect the affinity of FinO
wereobserved onthe30-sideoftheSLIIstemproximalto binding, suggesting the loop plays little if any role in
theloop(C26toG32).Takentogether,theresultssuggest specific, high affinity interactions with FinO (15). We
thatFinOcontactstheendoftheSLIIduplexproximalto usedRNaseI(34)footprintingtodirectlyprobeforinter-
thesingle-strandedtailsbutdoesnotcontacttheregionof actionsbetweenFinOandsingle-strandedareasofSLIIor
the stem proximal to the loop. SLIIc. Figure 4 shows the results for the limited RNase I
The RNase V1 cleavage results for SLIIc were similar digestion of SLII and SLIIc both free, and in a 1:1
for SLII (Figure 3). FinO protected the lower portion of complex with the three FinO constructs. For both
the30-sideofthestemofSLIIcfromRNaseV1cleavageat RNAs, intense cleavage was observed, as expected, at
positions C32-U37 on 50-32P-SLIIc and A36-A40 on the single-stranded loop regions, and FinO binding con-
30-32P-SLIIc. Like SLII, there appeared to be a very sistently resulted in a significant enhancement in this
weak protection of the lower stem at the 50-end of SLIIc cleavage.
A A A A
6 6 6 6
einWTW3 einWTW3 einWTW3 einWTW3
ut Prot86 186 186 ut Prot86 186 186 ut Prot86 186 186 ut Prot86 186 186
InpOHT1No 1-133-45- InpOHT1No 1-133-45- InpOHT1No 1-133-45- InpOHT1No 1-133-45-
25
25
20 20
20
20
25
25
15
15
30
30
10 10
35
35
5' 32P-SLII 3' 32P-SLII 5' 32P-SLIIc 3' 32P-SLIIc
Figure 4. LimitedRNaseIdigestionof50-and30-end-labeledSLIIandSLIIcintheabsenceandpresenceofvariousFinOconstructs.Productsof
the RNase I (0.01U/ml final concentration) cleavage reactions were resolved on 15% urea-denaturing polyacrylamide gels. The radiolabeled RNA
constructisnotedbeloweachofthegels.ThelanesOHandT1representthealkalinehydrolysisandRNaseT1cleavageofdenaturedSLIIorSLIIc,
respectively. The RNA nucleotide positions are indicated at the left of the gels. Large and small arrowheads indicate major or minor cleavages by
RNase I in the presence of FinO while the vertical bracket indicates protection from RNase I.
NucleicAcidsResearch,2011,Vol.39,No.10 4457
FinOprotectsthe30-tailsofSLIIandSLIIcfromRNaseI with increasing FinO concentrations, suggesting a strong
degradationandenhancescleavageofthe50-tailofSLIIc interaction between FinO and the 30-tail. Less dramatic
but significant protection was observed for the 30-tail of
Ithasbeenshownbyelectrophoretic mobilityshiftassays
SLIIc (Figure 5B). Similar results were also observed for
that a 30 single-stranded tail of at least six nucleotides on
FinO W36A and FinO (data not shown).
SLII (and FinP) is required to bind FinO (11). Removing 33–186 45–186
In contrast, the 50-tails of SLII and SLIIc did not
the tail in its entirety led to a 5.5-fold decrease in affinity
appear to be protected by FinO. This was most clearly
forFinO-GST.Toamuchlesserextent,removingthefour
demonstrated for the SLIIc construct (Figure 5C). Here,
nucleotide50-tailofSLII(spacerbetweenSLIandSLIIin
modestcleavageofthe50-tailwasobservedintheabsence
FinP) led to a modest 1.3-fold decrease in affinity, and
of FinO, however, binding of any of the FinO variants
removal of both tails decreased affinity for FinO by
resulted in a more significant cleavage of the 50-tail. This
nearly 14-fold. We used RNase I footprinting to test for
suggests that FinO binding frees the 50-tail of SLIIc for
interactions between FinO and the 50- and 30-tails. The
attack by RNase I. We were unable to detect RNase I
50-32P-labeled SLII cleavage gel in Figure 4 suggests that
cleavage at the shorter, four nucleotide 50-tail of SLII
protection of the 30-tail does occur.
(Figure 5D). A summary of the RNase I and RNase VI
To more clearly assess this protection, we carried out
cleavage data for SLII and SLIIc is shown in Figure 7A.
RNaseIfootprintingexperimentsatamuchhigherRNase
I concentration, to enhance cleavage of the 30-tail and
reducetheamountoffull-lengthRNAwhichwouldother- FinO selectively binds to the 30-hydroxyl group of the
wise obscure the footprint. The clearest results were 30-tail of SLII
obtained for 50-labeled SLII RNA (Figure 5A). Under TofurthertestthenatureofrecognitionoftheSLII30-tail
these conditions, the 30-tail was completely removed in by FinO, we created a version of SLII in which
the absence of FinO. Dramatic protection was observed 30,50-cytidine diphosphate (pCp) was appended at the
A Final [FinO 1-186 WT] (µM) B A C
0 .25 .5 1 2.5 5 SFiL(niI)OI -SLII Input OHT1 No Protein1-186 WT33-186 W36 45-186 No Protein 1-186 WT 33-186 W36A 45-186 T1 OH Input
ut
Inp OHT1 G39 (i)
(ii) G7
5' 32P-SLIIc
G40
G23
G25
D
(ii)
A
6
ein WTW3
(iii) No Prot 1-186 33-186 45-186 T1 OH Input 3' 32P-SLIIc
G18
G16
G5
G15
3' 32P-SLII
5' 32P-SLII
Figure 5. RNaseIoverdigestionof50-and30-end-labeledSLIIandSLIIcinabsenceandpresenceofvariousFinOconstructs.Inallexperiments,the
RNAs were digested with RNase I at a final concentration of 0.1U/ml. (Ai) A 8% native EMSA showing binding reactions of 50-32P-SLII with
increasing amounts of wild-type FinO before the addition of RNase I. The final concentration of FinO WT in micromolar in each reaction is
indicatedontopofthegel.(Aii)Slicefroma15%urea-denaturinggelshowingtheproductsoftheRNaseIdigestof50-32P-SLIIatthe30-endinthe
presence of increasing amounts of FinO. The lanes correspond to the binding reactions in (Ai). The lanes OH and T1 represent the alkaline
hydrolysis and RNase T1 cleavage of denatured 50-32P-SLII, respectively. SLII nucleotide positions are indicated at the right of the gel. (B) Slice
from a 15% urea-denaturing gel showing the products of the RNase I digest of 50-32P-SLIIc at the 30-end in the absence and presence of various
FinOconstructs.NucleotidepositionG39isindicatednexttothegel.(C)Slicefroma15%urea-denaturinggelshowingtheproductsoftheRNaseI
digestof30-32P-SLIIcatthe50-endintheabsenceandpresenceofvariousFinOconstructs.SLIIcnucleotidepositionsareindicatednexttothegel.
(D) Slice from a 15% urea-denaturing gel showing the products of the RNase I digest of 30-32P-SLII at the 50-end in the absence and presence of
variousFinOconstructs.NucleotidepositionG5isindicatedattherightofthegel.Forexperimentsin(B–D),a1:1molarratioFinO-RNAcomplex
was formed prior to exposure to RNase I.
4458 NucleicAcidsResearch,2011,Vol.39,No.10
30-end using T4 RNA ligase I. Intriguingly, none of the different models resulted, all consistent with the input
FinO constructs bound this RNA in electrophoretic restraints. These models were then further filtered based
mobility shift assays; however, binding could be restored ontheiragreementtothesolutionscatteringprofileofthe
by treatment of this RNA with T4 kinase/phosphatase to FinO –SLII complex. A single orientation of the
45–186
remove the terminal 30-phosphate (Figure 6). To further RNAstem–loopdockedonFinOagreedwithboththebio-
probe the specificity for the RNA 30-hydroxyl group, we chemicalandSAXSdata.Therestrainedenergyminimiza-
next treated the RNA with NaIO , which selectively tion was carried out using the program HADDOCK
4
oxidizes the 30 ribose sugar to a 20,30 dialdehyde (35). (26,27).
This treatment abrogated FinO binding, further demons- Docking in HADDOCK is a multi-step procedure con-
trating the selective recognition of the 30-hydroxyl group sisting of rigid-body docking by energy minimization,
by FinO. torsion-angle refinement of residues or nucleotides
involved in the docking interface, and a final Cartesian
An energy-minimized, biochemical knowledge-driven model dynamics refinement in water. The energy minimization
of the FinO45–186–SLII complex is constrained by chemical distance restraints called
High resolution structural information for FinO–RNA AIRs. AIRs are defined by the user from knowledge
complexes has been difficult to obtain using either X-ray derived from biochemical or biophysical experiments
crystallographic or NMR approaches (D.C.A., R.A.E. about the docking components and their interactions.
and J.N.M.G., unpublished data). In the absence of such We have used a combination of site-directed protein–
information, we turned to small-angle X-ray scattering RNA cross-linking (14) and RNA footprinting (this
(SAXS) to provide structural information to help to work) to define AIRs between the protein and the RNA.
define FinO–RNA interactions, as it is suitable to define The RNA footprinting results define regions of the RNA
flexibleand dynamiccomplexes (36).Wedecided tofocus that are protected from ribonuclease digestion by FinO
on complexes between FinO and the minimal SLII and are therefore likely involved in, or adjacent to, the
45-186
RNA target for three reasons. First, the footprinting protein–RNA interface. The protein–RNA cross-linking
results described above indicate that FinO contacts data defines which regions of the protein are in close
45–186
the same regions within stem-loop RNA targets as enough proximity to the RNA to form a chemical
full-length FinO. Second, FinO is more soluble and cross-link. The crosslinker used, azidophenacyl bromide,
lesspronetoaggregationthanth45e–1la86rgerFinOconstructs, reacts non-specifically with RNA within a 10A˚ radius. In
and therefore is more appropriate for SAXS measure- HADDOCK, two types of AIRs are formally defined,
ments. Third, the fact that FinO binds SLII with active and passive. Active restraints are between residues
45–186
(cid:6)20-fold higher affinity than full-length FinO helps shownexperimentallytobeinvolvedintheinteractionand
ensure that the experimental sample is fully bound under have a solvent-exposed surface area of >50%. Passive
the SAXS solution conditions and does not contain a sig- restraints are between similarly solvent-exposed residues
nificant fraction of unbound components. SAXS data that are direct neighbors of active residues. Distance
were collected at the SIBYLS beamline at the Advanced restraints during the energy-minimization are applied
Light Source on FinO with the minimal SLII RNA between pairs of residues being either active–active or
45–186
target. Analysis was consistent with complex formation. active–passive, but not passive–passive combinations.
In order to present a rigorous structural model of the Sixproteinresiduesshowedappreciablecross-linkingto
FinO –SLII interaction, we docked existing X-ray theRNAandwereassignedasactiveAIRsontheprotein.
45–186
crystal structures of the individual components in a re- Notably, Arg121 and Lys125, both part of a positively
strained energy-minimization procedure. The restraints charged patch on one side of the base of FinO, exhibited
were derived from biochemical information presented significantly stronger crosslinks than Lys46, Lys81,
here and in previous work (14). A number of spatially Arg165 or Lys176. Nucleotides of SLII protected from
A B A C A
NP FinO 33-186 W36A 45-186 No ProteinFinO 33-186 W3465-186 No ProteinFinO 33-186 W3465-186
T4 kinase/
FinO-SLII phosphatase NaIO
4
32P-SLII
3’ phosphate 2’,3’ cis-diol 2’,3’ dialdehyde
Figure 6. FinObindingtoSLIIrequiresaterminal30-OHonthe30-tailofSLII.Nativegels(8%)ofbindingreactionsbetweenFinOconstructsand
SLIIRNA derivatives.(A)FinOdoes notbind SLIIRNAcontaininga30,50-cytidinediphosphate 30-terminus. T4 RNA ligaseIwasusedtoligate
30,50-cytidine [50-32P] disphosphate (pCp) to the 30-tail resulting in a 30-phosphate. NP, no protein. Triangles represent increasing concentrations of
FinO or the indicated mutants: 0.25, 0.5, 1, 2.5, 5 and 10mM. The positions of free 32P-SLII and the FinO-32P-SLII are noted by arrows. (B)
TreatmentofSLIIRNAcontaininga30,50-cytidinediphosphate30-terminustogivea20,30 cis-diol(30-hydroxyl)restoresFinObinding.(C)Oxidation
of SLII with sodium periodate to give a 20,30 dialdehyde reduces binding affinity. The protein concentrations were 1mM in each of the binding
reactions in B and C.
NucleicAcidsResearch,2011,Vol.39,No.10 4459
RNaseV1 or RNase I cleavage upon FinO binding largely constrained by the distance restraints. The models
(7,8,36–45, Figure 7A), were assigned as passive AIRs. sample a broad selection of size and shape, with R
G
HADDOCK was used to generate 2500 models ranging from 23.0 to 33.9A˚ . The mean fit to the data
sampling a subset of the complex’s conformation space for this model set had a (cid:2)2 of 11.5 with a best fit of 3.4.
defined by the biochemical information encoded in the A further improvement in fit could be obtained by
form of ambiguous interaction restraints (see ‘Materials allowing those residues determined to be flexible by
and Methods’ section). RNA duplexes were primarily re- normal mode analysis of FinO , namely residues
45–186
stricted to the positively charged face of the protein and 45–64 of the N-terminal a-helix, to be fully flexible
oriented with the 50-and 30-tails contacting the base of during the docking. This flexible model set had a mean
FinO with the loop exposed into the solvent, orientations and best fit (cid:2)2 of 9.7 and 2.3, respectively. The full model
set is represented in Supplementary Figure 1 along with a
subset of the best 50 models ordered on (cid:2)2.
A For comparison, HADDOCK was run using randomly
A selectedAIRsbetweenresiduesandnucleotidescontaining
C C >20% relative accessibility to generate 2500 models. This
20 U G U
U U G G random-AIRmodelsethadamean(cid:2)2of17.5,significant-
G - C C A ly worse than that of the restrained model set, with the
A - U A A
G - C G - C single best-fit to the experimental scattering of 2.1.
G - C 20 A - U Although this best-fit randomly restrained model slightly
G - C 30 G - C 30 improves the fit to the SAXS data relative to the best-fit
A - U G - C
C - G G - C restrained model, inspection shows a poor fit to the bio-
G - C A - U chemical data. The best 250 RNA–protein complex
10 U - A C - G models, ordered on (cid:2)2 were subjected to RMSD-based
A - U G - C
G - C U - A pair-wise cluster analysis. Five clusters (Figure 8) met
C - G A - U the criteria of having a pair-wise RMSD cutoff less than
1 UG - GC 45 VVVVV 10GC -- GC 45 13A˚ and containing at least six members. In every case,
5'-G A C A G A U U U U - 3' 5'-A A A A U C G C A C U G U C -3' the RNA docked to the side of FinO containing the posi-
40 1 40 tively charged patch and residues Arg121, Lys125 and
SLII SLIIc Arg165. Alignment of the stem–loop was also consistent
amongstallclusterswiththebaseoftheRNAclosertothe
B
globular portion of FinO and the stem-loop generally co-
incidentwiththeN-terminalhelicalextension.Arepresen-
tative model from one of these clusters was chosen to
Lys46
illustrate the satisfaction of a number of the programmed
distancerestraintsandthegeneralorientationoftheRNA
stem (Figure 9A) while still providing a reasonable fit to
°
180 the SAXS data (Figure 9B). Nucleotides C7 and G8 are
partially buried against the N-terminal a-helix of FinO
and the 30-tail extends along the positively charged face
of FinO within contact distance of residues Arg121,
Lys176
Lys125 and Arg165. The fit of this particular model to
the experimental data, (cid:2)2 of 4.2, is comparable to the
mean (cid:2)2 of 4.6 for all 39 models in the five clusters.
Arg165 Lys125 Lys81
Arg121
DISCUSSION
FinO binds to SLII and SLIIc at the lower half of the
duplex and the 30-tail
OurfootprintingresultsindicatethatFinObindsitstarget
Figure 7. SummaryofRNaseV1andIcleavagereactionsofSLIIand
RNAs through recognition of single strand–duplex RNA
SLIIc. (A) Secondary structuresofSLII andSLIIc showingthe results
fromthe RNasecleavagereactions.Large andsmallblack arrowheads junctions. This interaction is asymmetric; that is, the 30
denote strong and weak RNase V1 cleavages, respectively in the single strand and 30 side of the duplex is contacted more
presence of the FinO constructs. Large and small open arrowheads strongly by the protein than the 50 single-stranded tail or
denote strong and weak RNase I cleavages, respectively in the the 50 side of the duplex. While these interactions were
presence of FinO. Boxes indicate footprints where FinO protected the
RNAs from RNase V1 cleavage. Dashed boxes indicate areas of pro- clearly evident in ribonuclease footprinting experiments,
tection from RNase I cleavage by FinO. An area where a ‘V’ resides we were unable to detect any interactions using the
indicates enhanced cleavage by RNase I in the presence of the FinO chemical probes diethylpyrocarbonate (DEPC), dimethyl-
constructs. (B) Electrostatic potentials at the solvent accessible surface
sulphate (DMS), or hydroxyl radical (data not shown).
of FinO , contoured at ±10 kT/e. Approximate surface locations
33–184 This suggests that FinO does not interact tightly with
ofthesixFinOside-chainsknowntocross-linktoSLIIarelabeledand
shown with semi-transparent circles. the bases or sugars of the RNA, and largely relies on