Báo cáo khoa học: Leishmania donovani bisubunit topoisomerase I gene fusion leads to an active enzyme with conserved type IB enzyme function doc

To determine what happens to the enzyme architecture and catalytic property if the two subunits are fused, and to explore the functional relationship between the two subunits, we describ

fusion leads to an active enzyme with conserved type IB enzyme function

Benu B Das1,*, Somdeb Bose Dasgupta1,*, Agneyo Ganguly1, Saumyabrata Mazumder2,

Amit Roy1and Hemanta K Majumder1

1 Department of Molecular Parasitology, Indian Institute of Chemical Biology, Kolkata, India

2 Infectious Diseases Group, Indian Institute of Chemical Biology, Kolkata, India

The type IB DNA topoisomerase family includes

euk-aryotic nuclear topoisomerase I and the

topoisomeras-es encoded by vaccinia, bacteria and other cytoplasmic

poxviruses [1–3] The type IB enzymes relax

super-coiled DNA via a multistep reaction pathway entailing

noncovalent binding of the topoisomerase to duplex

DNA, cleavage of one DNA strand with formation of

a covalent DNA-(3-phosphotyrosyl)–protein

intermedi-ate, strand passage, and strand religation [1,2,4]

Recently, the discovery of the bisubunit

topoiso-merase I enzymes of Trypanosoma [5] and Leishmania

[6] in the kinetoplastid family have brought a new twist

in topoisomerase research related to evolution and func-tional conservation of the type IB family The core DNA-binding domain and the catalytic domain harbor-ing the consensus SKXXY motif are located in separate subunits The two subunits are synthesized by two different genes, and associate with each other through protein–protein interactions to form an active hetero-dimeric topoisomerase I within the parasite This un-usual structure of DNA topoisomerase I may provide

a missing link in the evolution of type IB enzymes

Keywords

camptothecin; gene fusion; Leishmania;

topoisomerase I; SKXXY motif

Correspondence

H K Majumder, Molecular Parasitology

Laboratory, Indian Institute of Chemical

Biology, 4 Raja S.C Mullick Road,

Kolkata-700032, India

Fax: +91 33 2473 5197

Tel: +91 33 2412 3207

E-mail: hkmajumder@iicb.res.in

*These authors contributed equally to this

work

(Received 12 July 2006, revised 1

November 2006, accepted 6 November

2006)

doi:10.1111/j.1742-4658.2006.05572.x

All eukaryotic topoisomerase I enzymes are monomeric enzymes, whereas the kinetoplastid family (Trypanosoma and Leishmania) possess an unusual bisubunit topoisomerase I To determine what happens to the enzyme architecture and catalytic property if the two subunits are fused, and to explore the functional relationship between the two subunits, we describe here in vitro gene fusion of Leishmania bisubunit topoisomerase I into a single ORF encoding a new monomeric topoisomerase I (LdTOPIL-fus-S)

It was found that LdTOPIL-fus-S is active Gene fusion leads to a signifi-cant modulation of in vitro topoisomerase I activity compared to the wild-type heterodimeric enzyme (LdTOPILS) Interestingly, an N-terminal truncation mutant (1–210 amino acids) of the small subunit, when fused

to the intact large subunit [LdTOPIL-fus-D(1–210)S], showed reduced topoisomerase I activity and camptothecin sensitivity in comparison to LdTOPIL-fus-S Investigation of the reduction in enzyme activity indicated that the nonconserved 1–210 residues of LdTOPIS probably act as a

‘pseudolinker’ domain between the core and catalytic domain of the fused Leishmania enzyme, whereas mutational analysis of conserved His453 in the core DNA-binding domain (LdTOPIL) strongly suggested that its role

is to stabilize the enzyme–DNA transition state through hydrogen bonding

to one of the nonbridging oxygens Taken together, our findings provide

an insight into the details of the unusual structure of bisubunit topo-isomerase I of Leishmania donovani

Abbreviation

CPT, camptothecin.

We have previously demonstrated the in vitro

recon-stitution of the two recombinant proteins LdTOPIL

and LdTOPIS, corresponding to the large and small

subunits, and localization of the active enzyme in both

the nucleus and kinetoplast [7] The enzyme is

conven-tional in its Mg2+ independence, site specificity for

eukaryotic type IB-specific recognition sites and

campt-othecin (CPT) sensitivity LdTOPIL and LdTOPIS

form a direct 1 : 1 heterodimer complex through

pro-tein–protein interactions

Davies et al [8] have made a 2.27A˚ crystal

struc-ture of an active truncated Leishmania donovani

TOPIL–TOPIS heterodimer bound to nicked

double-stranded DNA in the presence of vanadate The

vana-date forms covalent linkages between the catalytic

Tyr222 residue of the small subunit (LdTOPIS) and

the nicked ends of the scissile DNA strand, mimicking

the transition state of the topoisomerase I catalytic

cycle The structure predicts that the highly conserved

constellation of the catalytic residues (Arg314,

Lys352, Arg410 and His453 of LdTOPIL and the

consensus catalytic residue Tyr222 in LdTOPIS) share

a common module between Leishmania and human

topoisomerase I

Although the details of catalysis for the unusual

het-erodimeric Leishmania topoisomerase I reaction remain

to be elucidated, based on the crystal structure of

trun-cated LdTOPILS, it appears that His453 forms a

2.6 A˚ hydrogen-bonding contact with a nonbridging

oxygen atom of the vanadate [8] This interaction is

virtually the same as in the noncovalent complex of

human topoisomerase I, where His632 is 2.6 A˚ from a

nonbridging phosphate oxygen atom of the DNA base,

and may be responsible for stabilizing the enzyme–

DNA interaction [9]

Human topoisomerase I is a monomeric structure

composed of 765 residues with a molecular mass

of 91 kDa Topo70 is a truncated form of human

topo-isomerase I that lacks residues 1–174 of the N-terminal

domain and retains full enzyme activity in vitro The

enzyme contains a central DNA-binding core domain

and a C-terminal catalytic domain harboring an

SKINYL motif The crystal structure of human

topo-isomerase I demonstrates that the core and C-terminal

domains form a clamp-like structure embedding the

DNA helix in a central pore, with two lobes of the

protein binding to either site of the helix [10] The

conserved subdomains I and II contribute the upper

part ‘CAP’, which is connected by a flexible hinge to

the bottom part of the clamp of subdomain III The

linker domain forms a coiled-coil structure that

pro-trudes from the body of the enzyme and connects the

core to the highly conserved C-terminal domain close

to the scissile phosphate in the bound DNA This architecture facilitates the opening and closing of the protein clamp during binding and release of DNA [11,12]

Champoux and his group have previously reported their findings on human topoisomerase I that has been artificially fragmented into two proteins (topo58⁄ 12

or topo58⁄ 6.3) The core and the catalytic domain can reconstitute topoisomerase I activity It was shown that detachment of the linker from the core domain makes the enzyme highly distributive, with 20-fold redu-ced affinity for DNA and less sensitivity to CPT [13] Some of our previous findings on Leishmania topo-isomerase I are in keeping with those of reconstituted human topoisomerase I [7], but a closer look reveals that differences do exist in the sequences, some bio-chemical properties and preferential sensitivities to CPT [14,15] Thus, the key questions arise of what will happen to the enzyme architecture and catalytic pro-perty if the two subunits are fused to a monomeric structure such as human topoisomerase I, and what the role of the conserved His453 in enzyme catalysis is

To address these issues, we describe experiments in which Leishmania bisubunit topoisomerase I large sub-unit (LdTOPIL) and small subsub-unit (LdTOPIS) genes were fused into a single ORF encoding a new topo-isomerase I (LdTOPIL-fus-S) This monomeric enzyme

is active and shows increased activity compared to the wild-type heterodimeric enzyme (LdTOPILS) Interest-ingly, an N-terminal truncation mutant (1–210 amino acids) of the small subunit, when fused to the intact large subunit [LdTOPIL-fus-D(1–210)S], shows reduced topoisomerase I activity compared to LdTOPIL-fus-S The present study also describes the role of the con-served His453 in the core DNA domain (LdTOPIL) in the reaction catalyzed by the fusion enzyme Hence, this study provides substantial information on the mechanistic details and unusual structure of this bisub-unit enzyme

Results Purification of recombinant proteins

A schematic alignment of monomeric (human, vaccinia and bacterial) topoisomerase IB with that of the het-erodimeric topoisomerase IB of Leishmania is shown in Fig 1A, in order to relate the two subunits to the monomeric enzymes All the recombinant constructs used in the present study and the deduced amino acid sequences of the fusion regions are shown in Fig 1B Leishmaniabisubunit topoisomerase I fusion constructs were developed as described in Experimental

procedures The overexpressed proteins from

Escheri-chia coli BL21(DE3)pLysS cells harboring plasmids

pET28cLdTOPIL-fus-S, pET28cLdTOPIS-fus-L and

pET28cLdTOPIL-fus-D(1–210)S (1–210 amino acid

deletion mutant from the N-terminal region of the

small subunit was fused in frame with LdTOPIL) were

purified separately through an Ni2+–nitrilotriacetic

acid agarose column The proteins were further purified

through a phosphocellulose column as described in

Experimental procedures

A recent crystal structure has identified a conserved

His453 of LdTOPIL close to the nonbridging oxygen

atom of the vanadate [8] that potentially mimics the

transient state of the enzyme–DNA covalent complex

To test this possibility directly, we used site-directed

mutagenesis to change His453 of LdTOPIL-fus-S to

glutamine As a control, we also changed His453 of

LdTOPIL-fus-S to alanine, as well to identify its role

The other recombinant proteins, i.e LdTOPIL (large

subunit) and LdTOPIS (small subunit), were purified

as described previously [7] Analysis of the purified proteins by SDS⁄ PAGE (Fig 1C) showed that all the recombinant proteins are essentially homogeneous

LdTOPIL-fus-S fusion protein is a functional topoisomerase I

We assessed the topoisomerase activity of the Leishma-nia bisubunit fused protein encoding a new topoiso-merase I (LdTOPIL-fus-S) by a plasmid relaxation assay Reconstitution of wild-type Leishmania bisub-unit topoisomerase I (LdTOPILS) activity has been described previously [7,14]

Time course relaxation experiments were performed

in a standard assay mix where the plasmid DNA and the enzymes (LdTOPILS, LdTOPIL-fus-S and Ld-TOPIS-fus-L) were mixed at a molar ratio of 4 : 1 The velocity for LdTOPIL-fus-S was linear for the first

5 min of the reaction It was observed that LdTOPIL-fus-S relaxed supercoiled DNA at a slower rate than did reconstituted LdTOP1LS (compare lanes 2–9 of Fig 2B with lanes 2–9 of Fig 2A), whereas the reverse fusion LdTOPIS-fus-L failed to show any plasmid DNA relaxation activity (Fig 2C) The smaller num-ber of topoisomer intermediates reacting with Ld-TOPIL-fus-S indicates that LdLd-TOPIL-fus-S completely relaxes the supercoiled DNA substrate in a processive

A

B

C

Fig 1 (A) Schematic alignment Monomeric (human vaccinia and bacterial) topoisomerase IB aligned with bisubunit topoisomerase of Leishmania in order to relate the two subunits with their monomeric counterparts The position of active site pentad residues is also shown (B) Protein constructs Structure of recombinant L donovani topoisomerase I proteins The first line shows the full-length larger subunit (dark) as the core DNA-binding subunit with the conserved catalytic His at position 453 The second line shows the Leishmania bisubunit topoisomerase I fusion construct, LdTOPIL (dark) and LdTOPIS (light shaded) and the deduced amino acid sequences of the fusion regions The third line shows the reverse fusion construct

of Leishmania bisubunit topoisomerase I, LdTOPIS (light shaded) and LdTOPIL (dark) and the deduced amino acid sequences of the fusion regions The fourth line shows the N-terminal truncated small subunit (amino acids 211–262) fused to intact large subunit to gen-erate an ORF, LdTOPIL-fus-D(1–210)S The fifth and sixth lines show the point mutations generated at the His453 position of the LdTOPIL-fus-S gene to H453A and H453Q, respectively The seventh line shows the smaller catalytic subunit (light shaded) with the active site residue The constructs were developed as des-cribed in Experimental procedures (C) Coomassie-stained 10% SDS ⁄ PAGE analysis of the purified recombinant proteins with 5 lg per lane Lanes 1–7, LdTOPIL, LdTOPIL-fus-S, LdTOPIS-fus-L, fus-D(1–210)S, H453A and H453Q mutants of LdTOPIL-fus-S and LdTOPIS proteins purified through an Ni 2+ –nitrilotriacetic acid column, respectively, followed by a phosphocellulose column The positions and molecular masses of protein standards are indica-ted on the left.

fashion before dissociating or reassociating with another DNA molecule However, under these condi-tions, the situation with reconstituted LdTOPILS is different, as partially relaxed topoisomers are visible during the course of the relaxation reaction (compare lanes 2–9 of Fig 2B with lanes 2–9 of Fig 2A)

Relaxation activity of the mutant topoisomerase I The effects of mutations on enzyme activity were ana-lyzed by standard plasmid DNA relaxation assays with

a molar ratio of DNA to enzyme of 3 : 1 (Fig 3A) LdTOPIL-fus-S completely relaxes the DNA within

10 min under these conditions, whereas all three mutant enzymes exhibited slow relaxation kinetics Complete relaxation by the LdTOPIL-fus-D(1–210)S enzyme was not observed until  40 min, and thus the LdTOPIL-fus-D(1–210)S protein appeared to be four-fold less active than the LdTOPILfus-S H453Q had less activity than LdTOPIL-fus-S, failing to completely relax the supercoiled DNA even after 40 min Very little relaxing activity was detectable for the H453A protein We estimate the activity of the H453A protein

to be more than 100-fold reduced compared with that

of LdTOPIL-fus-S

Supercoiled DNA relaxation under conditions of limiting topoisomerase I is stimulated  10-fold in the presence of 10 mm Mg2+, probably because of an increase in the dissociation rate of the enzyme from the DNA [7,16] It follows that the rate-limiting step for DNA relaxation by LdTOPIL-fus-S under normal assay conditions is enzyme dissociation This effect can

A

B

Fig 3 Plasmid relaxation assays for LdTOPILS, LdTOPIL-fus-S and its mutant variants Reaction mixtures containing 90 fmol of supercoiled plasmid DNA in relaxation buffer without (A) and with (B) 10 m M Mg 2+ The reactions were initiated by the addition of 30 fmol of topoiso-merase I variants incubated at 37 C for different time periods as indicated in the figure Reactions were stopped by addition of 0.5% SDS; samples were electrophoresed in 1% agarose gel The zero time point was taken prior to the addition of enzyme.

A

B

C

-Time

(min)

RL/NM

Lane 1 2 3 4 5 6 7 8 9

Lane 1 2 3 4 5 6 7 8 9

Lane 1 2 3 4 5 6 7 8 9

SM

Time

(min)

RL/NM

SM

RL/NM

SM

0.5

LdTOP1LS

LdTOP1L-fus-S

LdTOP1S-fus-L

1 5 10 15 20 30 40

- 0.5 1 5 10 15 20 30 40

Time

(min) - 0.5 1 5 10 15 20 30 40

Fig 2 The LdTOPIL-fus-S fusion protein is a functional

topoiso-merase I Relaxation of supercoiled pBS (SK + ) DNA with

reconstitu-ted enzyme LdTOPILS (A), LdTOPIL-fus-S (B), and reverse fusion

LdTOPIS-fus-S (C), at a molar ratio of 4 : 1 Lane 1, 80 fmol of pBS

(SK +) DNA Lanes 2–9, same as lane 1, but incubated with

20 fmol of topoisomerase I variants at 37 C for different time

peri-ods as indicated in the figure All reactions were stopped by

addi-tion of 0.5% SDS; samples were electrophoresed in 1% agarose

gel The positions of supercoiled monomer (SM) and relaxed and

nicked monomer (RL ⁄ NM) are indicated.

be seen in Fig 3B, where the addition of 10 mm Mg2+

to the LdTOPIL-fus-S reaction increased the rate

approximately 10-fold (reaction complete in 1 min)

Although addition of 10 mm Mg2+enhances the

relax-ation rate of LdTOPIL-fus-D(1–210)S, it appears to be

20-fold reduced in comparison to LdTOPIL-fus-S

(Fig 3B) However, the presence of Mg2+ in the

reac-tions for the His453 mutant proteins (H453Q and

H453A) had no effect on the relaxation rates (Fig 3B),

suggesting that enzyme chemistry rather than enzyme

dissociation was the rate-limiting step for all of the

His453 mutant enzymes Moreover, in the presence of

10 mm Mg2+, the differences between the estimated

activity for LdTOPIL-fus-S and the activities of the

mutant proteins were magnified H453Q was at least

40-fold less active than LdTOPIL-fus-S, whereas

H453A was more than 100-fold less active than

LdTOPIL-fus-S As a wild-type control, reconstituted

LdTOPIL and LdTOPIS were used both in the

absence and the presence of Mg2+-containing buffer

Effect of CPT on the relaxation activity and

equilibrium cleavage activity of fused

topoisomerase I variants

We examined the effect of CPT on the relaxation

activ-ity of wild-type control reconstituted LdTOPILS and

the fused enzyme [S and

LdTOPIL-fus-D(1–210)S] The reverse fusion LdTOPIS-fus-L was not

included in this experiment, as it was enzymatically

inac-tive in the relaxation assay Time course relaxation

experiments were performed in a standard assay mix

where the plasmid DNA and the enzyme [LdTOPILS,

LdTOPIL-fus-S and LdTOPIL-fus-D(1–210)S] were

mixed at a molar ratio of 1 : 2, to circumvent possible

effects due to a slow dissociation rate and enzyme

turn-over number in the presence of CPT The wild-type was

more distributive in nature and less sensitive to CPT

compared to LdTOPIL-fus-S (Fig 4A) In the absence

of CPT, the rate of relaxation of LdTOPIL-fus-S was

greater than that of LdTOPIL-fus-D(1–210)S (compare

lanes 2 and 3 of Fig 4B with lanes 2 and 3 of Fig 4C)

In the presence of CPT, it can be seen that the time

required to complete relaxation for LdTOPIL-fus-S was

increased approximately 25-fold (from 1 min to 25 min;

Fig 4B, compare lane 3 with lane 18), whereas the

drug had a reduced effect on the rate of relaxation by

LdTOPIL-fus-D(1–210)S (compare lane 4 with lane 15

of Fig 4C)

CPT, the most established topoisomerase I inhibitor,

has been shown to stabilize the cleavable complex

Here, we investigated the characteristics of

LdTOPIL-fus-S and LdTOPIL-fus-D(1–210)S in a cleavage assay

and compared it with LdTOPIL.S Transesterification was examined under equilibrium conditions by reacting LdTOPIL.S and LdTOPIL-fus-S with 5¢-32 P-end-labe-led 25-mer duplex oligonucleotides containing the high-affinity topoisomerase IB-binding site [7,14] The fact that cleavage activity with LdTOPIL-fus-S is enhanced in the presence of the drug suggests that CPT binds to the covalent complex between LdTOPIL-fus-S and DNA (Fig 4D, lanes 4 and 5) CPT enhanced the formation of the cleavable complex by 40% with respect to the extent of cleavage observed in the absence of the drug This result is similar to that obtained for the wild-type enzyme LdTOPILS (Fig 4D, lanes 2 and 3) Interestingly, LdTOPIL-fus-D(1–210)S showed reduced efficiency in cleaving 25-mer duplex oligonucleotides both in the absence and the presence of CPT (Fig 4C, lanes 6 and 7) The lower cleavage activity obtained for LdTOPIL-fus-D(1–210)S was consistent with the modest reduction in relaxation activity and CPT sensitivity

These results indicate that the single ORF resulting from fusion of the large and small subunit genes of Leishmania bisubunit topoisomerase I (LdTOPIL-fus-S) encodes for a new functional topoisomerase I The enzyme is conventional in its CPT sensitivity and shows cleavage specificity similar to that of LdTOPILS [7,14], whereas 1–210 amino acids residues from the N-terminal end of the small subunit (LdTOPIS) have a probable role in CPT sensitivity in the fused enzyme (LdTOPIL-fus-S)

Gene fusion and its analysis for DNA-binding efficiency

To test whether the observed changes in relaxing activ-ity of the fused proteins resulted from increased or decreased affinity of the enzymes for DNA, we carried out native gel mobility shift assays with reconstituted LdTOPILS, monomeric S, LdTOPIL-fus-D(1–210)S, H453Q and H453A mutant LdTOPIL-fus-S complexed with the 5¢-32P-labeled duplex oligomer containing the high-affinity topoisomerase IB binding site [9], as previously described [7,14]

Like LdTOPILS, S, D(1–210)S, H453Q and H453A mutant

LdTOPIL-fus-S are positively charged, and because the bound oligonucleotide only partially neutralizes the positive charge, the protein–DNA complexes failed to enter the native gel Figure 5A shows the extent of unbound oligonucleotide compared to the oligonucle-otide control when binding was carried out with increasing concentrations of the enzymes Under these conditions, Kd is equal to the protein concentration

at which the amount of unbound oligonucleotides

observed in the gel has been reduced by a factor of

two [17,18] The binding assays yielded a Kd value of

3.2· 10)7m for the interaction of LdTOPILS with

the DNA substrate, which is about 5.5-fold higher

than the value measured for the interaction of

LdTOPIL-fus-S (0.6· 10)7m) with DNA (Fig 5B),

whereas LdTOPIL-fus-D(1–210)S interacts with DNA

substrate with a Kd value of 1.9· 10)7m, indicating

 3-fold lower affinity than LdTOPIL-fus-S Thus, gene fusion increases the DNA-binding efficiency of LdTOPIL-fus-S approximately 5.5-fold compared to the reconstituted enzyme On the other hand, a

 3-fold decrease in the DNA-binding efficiency of LdTOPIL-fus-D(1–210)S compared to LdTOPIL-fus-S correlates well with the decrease in topoisomerase activity of LdTOPIL-fus-D(1–210)S in the plasmid relaxation assay (Fig 5B)

A

B

C

D

Fig 4 Effect of CPT on the relaxation activity equilibrium cleavage with LdTOPILS, LdTOPIL-fus-S and LdTOPIL-fus-D(1–210)S Relaxation of supercoiled pBS (SK + ) DNA with LdTOPILS (A) LdTOPIL-fus-S (B) or LdTOPIL-fus-D(1–210)S (C) at a molar ratio of 1 : 2 assayed in the presence or absence of CPT Lanes 1 and 11, 50 fmol of pBS (SK + ) DNA Lanes 2–10, same as lane 1 but incubated with 100 fmol of LdTO-PIL-fus-S or LdTOPIL-fus-D(1–210)S in the absence of CPT Lanes 11–20, same as lanes 2–10, but in the presence of 60 l M CPT incubated

at 37 C for the time periods indicated in the figure All reactions were stopped by addition of SDS to a final concentration of 0.5% (w ⁄ v); samples were electrophoresed in 1% agarose gel (D) Equilibrium cleavage reactions and electrophoresis in a denaturating polyacrylamide gel were performed as described in Experimental procedures Lane 1, 10 n M 5¢- 32

P-end-labeled 25-mer duplex oligonucleotides as indicated above Lanes 2–3, same as lane 1, but incubated with equal amounts (0.15 l M ) of LdTOPILS Lanes 4–5, incubated with LdTOPIL-fus-S Lanes 6–7, incubated with LdTOPIL-fus-D(1–210)S, in the absence or presence of CPT (60 l M ) as indicated Positions of uncleaved oligo-nucleotide (25-mer) and the cleavage product (12-mer oligooligo-nucleotide complexed with residual topoisomerase I) and the scheme of the reac-tion are indicated.

The binding profiles revealed that the affinity of the

H453Q and H453A mutant LdTOPIL-fus-S for DNA

substrate was about the same as that of

LdTOPIL-fus-S, with Kd values of 0.7 · 10)7m (data not shown)

These results demonstrate that the reduction in

relax-ing activity with the various changes at position 453 of

mutant proteins of LdTOPILfus-S did not result from

a defect in DNA binding

Suicidal cleavage activity of LdTOPIL-fus-S and

its mutant variants

We examined the transesterification reaction under

suicidal conditions by reacting LdTOPILS,

LdTOPIL-fus-S, LdTOPIL-fus-D(1–210)S, H453Q and H453A

mutant LdTOPIL-fus-S with synthetic suicide DNA

substrate The substrate consisted of a 5¢-32P-labeled

14 bp duplex with an 11 bp 5¢-tail [11,14] Upon

clea-vage and formation of a covalent protein–DNA

com-plex, the AG dinucleotide at the 3¢-end of the scissile

strand is released Cleavage was performed at 230C

for the time periods given in Experimental procedures

The cleavage activities of the enzymes, as determined

by the percentage of substrate converted to products,

were plotted as a function of time [19] In the suicidal

cleavage assay for LdTOPILS, about 75–80% of the

input DNA became covalently bound to protein and

reached its cleavage plateau after 30 min of incubation,

whereas LdTOPIL-fus-S completed the reaction after

6 min of incubation; however, interestingly, the cleavage

pattern with LdTOPIL-fus-D(1–210)S was

approxi-mately 10-fold reduced compared to that with the fused

enzyme LdTOPIL-fus-S LdTOPIL-fus-D(1–210)S

rea-ched its cleavage plateau after 60 min These

observa-tions indicate that gene fusion leads to a five-fold

enhancement of the apparent suicidal cleavage rate of

LdTOPIL-fus-S over LdTOPILS, whereas the deletion

mutant LdTOPIL-fus-D(1–210)S was defective in the

cleavage reaction compared to LdTOPIL-fus-S This

difference probably accounts for the relatively slow

plasmid relaxation rate caused by LdTOPUIL-fus-D(1–210)S

The cleavage reaction with H453Q mutant LdTOPIL-fus-S reached a plateau after 240 min of incubation, whereas the cleavage rates for the H453A protein were just detectable above the background compared to LdTOPIL-fus-S (Fig 6A) Thus, the effects of the vari-ous changes at position 453 on the cleavage rates quanti-tatively parallel the reductions in the rates of relaxation described above, indicating its role either in transesterifi-cation chemistry per se or in a step in the reaction path-way that occurs after initial binding prior to strand rotation

To gain further insight into the fate of the covalent complexes produced by LdTOPILS, LdTOPIL-fus-S, and LdTOPIS-fus-L with labeled oligonucleotide substrate, the reaction mixtures were analyzed by SDS⁄ PAGE Coomassie blue-stained SDS⁄ PAGE shows the mobility of free enzymes (Fig 6B, lanes 1–3) An autoradiograph of the same dried gel shows that the label appears to be associated with LdTOPIS and LdTOPIL-fus-S (Fig 6C, lanes 4 and 5), and this association causes slightly slower migration of enzyme– DNA complex compared to free proteins No LdTOPIS DNA or LdTOPIL-fus-S DNA bands are visible with Coomassie blue staining (Fig 6B), as only a small amount of protein became covalently attached to the DNA, and this became visible after autoradiography (Fig 6C, lanes 4 and 5) Suicide cleavage by LdTOPIS-fus-L was not achieved under the same conditions (Fig 6C, lane 6) These results demonstrate that the reverse fusion product (LdTOPIL-fus-S) was unable to show topoisomerase I cleavage activity

Religation activity of LdTOPIL-fus-S and its mutant variants

Religation was studied under single-turnover conditions

by assaying the ability of the covalent intermediate to attach a 5¢-hydroxyl-terminated 11-mer to the covalently

Fig 5 DNA-binding assays The native gel shift assay was carried out as described in Experimental procedures (A) Autoradiograph

of the unbound oligonucleotide for each con-centration of protein used corresponding to the DNA control for the enzymes LdTOPILS, LdTOPIL-fus-S and LdTOPIl-fus-D(1–210)S (B) The percentage of unbound duplex oligo-nucleotide present in the gel was quantified

by Phosphorimager and plotted against the protein concentrations The binding profiles for LdTOPILS, fus-S and LdTOPIL-fus-D(1–210)S are indicated.

cleaved 12-mer to form a 23-mer product [11,14] The

ligation reactions of LdTOPILS, and LdTOPIL-fus-S

and LdTOPIL-fus-D(1–210)S, were performed as

described in Experimental procedures

The results indicated that the religation kinetics for

LdTOPILS was approximately two-fold faster than

that of LdTOPIL-fus-S However, the religation

kinet-ics for LdTOPIL-fus-D(1–210)S were more or less

sim-ilar to those of LdTOPIL-fus-S (Fig 7A,B) Therefore,

the five-fold faster cleavage rate and the two-fold

reduced religation rate of Leishmania fused

topoiso-merase I (LdTOPIL-fus-S) accounts for a small shift in

the cleavage–religation equilibrium towards cleavage compared to the reconstituted enzyme, and correlates with the increase of activity in the plasmid DNA relax-ation assay

Role of His453 of LdTOPIL in the fused enzyme construct

The crystal structure of Leishmania heterodimeric topo-isomerase I shows that the Ne2 atom of His453 of

A

B

Fig 7 Religation activity (A) Religation activity of LdTOPILS and LdTOPIL-fus-S Active cleavage complexes containing LdTOPILS or LdTOPIL-fus-S covalently attached to the covalently cleaved 12-mer

of the suicide substrate were reacted with 5¢-hydroxyl-terminated 11-mer to form a 23-mer product for 15, 30 and 60 s at 37 C, and the products were analyzed as above Religated product, active covalent complex and uncleaved product are indicated (B) The relative amount of cleavage product converted to ligation product in each sample for LdTOPILS and LdTOPIL-fus-S was plotted as func-tion of time The religafunc-tion reacfunc-tions were stopped after 15, 30, 60,

120, 150 s at 37 C, and the products were analyzed as above The results depicted were from experiments performed three times, and representative data from one set of these experiments are expressed as means ± SD Variations among different set of experiments were < 5%.

A

Fig 6 Suicide cleavage assays (A) DNA cleavage rate for LdTOPILS,

LdTOPIL-fus-S, LdTOPIL-fus-D(1–210)S, H453Q and H453A mutant

LdTOPIL-fus-S with the 5¢- 32

P-end-labeled suicide DNA substrate (14-mer⁄ 25-mer) shown in the figure The reaction mixtures were

incubated with the topoisomerase I variants for 1, 5, 10, 15, 30, 60,

120, 180, 240, 300 min at 23 C as described in Experimental

proce-dures Cleavage products were analyzed by denaturating PAGE, and

the percentage of cleaved DNA substrate was plotted as a function

of time The results depicted were from experiments performed

three times, and representative data from one set of these

experi-ments are expressed as means ± SD Variations among different set

of experiments were < 5% (B) Coomassie blue-stained

SDS-poly-acrylamide gel (C) Autoradiograph of the same gel Lanes 1–3,

5¢- 32 P-end-labeled suicide DNA substrate (14-mer ⁄ 25-mer) was

incu-bated with 3 lg of reconstituted LdTOPILS, LdTOPIL-fus-S and

LdTOPIS-fus-L, respectively, in the reaction buffer for 3 h at 23 C,

and reactions were stopped with SDS ⁄ PAGE sample buffer; samples

were boiled and loaded onto 10% SDS ⁄ PAGE gel.

LdTOPIL forms a hydrogen bond with the nonbridging

oxygen atom of the vanadate Hence, we assumed that

the His side chain might possibly serve as a general acid

that donates a proton to the leaving 5¢-hydroxyl as

cleavage occurs [10] If His453 were to act as a general

acid, deprotonation of the imidazole ring with

increased pH should reduce the rate of the cleavage

reaction for LdTOPIL-fus-S, but a similar increase in

pH should have no effect on cleavage by the H453Q

mutant enzyme To test this prediction, we measured

the cleavage rates of both LdTOPIL-fus-S and the

H453Q mutant LdTOPIL-fus-S proteins at the

follow-ing pH values: 6, 6.5, 7, 7.5, 8, 8.5, and 9.5 As shown

in Fig 8, the activity of LdTOPIL-fus-S decreases

slightly over the pH range from 7.5 to 9.5, but the

response of the H453Q mutant enzyme was very

sim-ilar Thus, it appears unlikely that His453 of LdTOPIL

for the fused enzyme acts as a general acid that donates

a proton to the leaving 5¢-oxygen In eukaryotic

type IB enzymes, the conserved His residue is involved

solely in phosphate binding and transition state

stabil-ization Some bacterial type IB enzymes have an Asn

residue [3] in place of this His residue This further

sug-gests a generalized role of His453 rather than a specific

role as a proton donor

Discussion

The crystal structure of monomeric human

topoiso-merase I seems compatible with a rotational model for

the relief of supercoils during DNA relaxation

Mode-ling studies have indicated that the DNA would

prob-ably contact both the CAP and linker regions of the

protein during strand rotation [10,12] The linker

domain, which is poorly conserved and variable in length, links the core and catalytic domains of the monomeric enzyme and is responsible for the activity

of the enzyme and CPT sensitivity [10]

However, interestingly, L donovani topoisomerase I

is an unusual bisubunit enzyme in which the functional linker is absent between the core DNA-binding domain and the catalytic domain which is harbored in a separate subunit [6,7,23] Our recent findings reveal that 1–39 amino acid residues of the large subunit that resemble the CAP region of the monomeric enzymes have a modulating role in noncovalent interactions with DNA and sensitivity towards CPT [14] Thus, it is interesting

to observe the change in the catalytic properties of the heterodimeric enzyme when the two subunits are fused

We also investigated the role of the conserved His453 residue in the large subunit (LdTOPIL) during enzyme catalysis Our studies provide insights into the mechanis-tic conservation of topoisomerase IB function in the Leishmaniaheterodimeric enzyme

Change in the catalytic efficiency due to gene fusion

We describe here the significant modulation of in vitro DNA relaxation due to gene fusion We have previ-ously shown that the reconstituted LdTOPILS has reduced activity in plasmid relaxation assays Ld-TOPILS appears to leave intermediate substrates after removing only a few supercoils at a time This accounts for the higher dissociation rate, yielding a higher turnover number [7], whereas under the same conditions, the relaxation mode of the fused enzyme LdTOPIL-fus-S was found to be more processive, and

is going through multiple rounds of relaxation before dissociating from its substrate DNA (Fig 2B), which

is well manifested by decreases in the enzyme dissoci-ation rate and turnover number (data not shown) These observations were further supported by a

 5.5-fold increase in the DNA-binding affinity of LdTOPIL-fus-S (Kd of 0.6· 10)7m) compared to the reconstituted enzyme LdTOPILS (Kdof 3.2 · 10)7m) This observation is consistent with that of reconstitu-ted human topoisomerase I that has been artificially divided into two proteins (topo58⁄ 12 or topo58 ⁄ 6.3), which are highly distributive, and bind DNA at a lower affinity than that of the intact enzyme [13] The enhanced activity in the plasmid DNA relaxa-tion assay shown by LdTOPIL-fus-S correlated well with the increase in cleavage rates seen under single-turnover condition; that is, LdTOPIL-fus-S shows a five-fold increase in cleavage rate over reconstituted LdTOPILS Interestingly, the fused enzyme shows an

Fig 8 Effect of pH on suicide cleavage rate for LdTOPIL-fus-S and

H632Q mutants of LdTOPIL-fus-S proteins The rate of suicide

clea-vage was measured as described in Experimental procedures, and

the logarithm (base 10) of the rates was plotted as a function

of pH.

approximately two-fold slower religation rate

com-pared to the reconstituted Leishmania enzyme

There-fore, the greater cleavage rates for LdTOPIL-fus-S and

slower religation compared to LdTOPILS account for

a shift in the cleavage–religation equilibrium towards

cleavage, as observed in monomeric human

topoisom-erase I [10,12] However, both the Leishmania enzymes

(LdTOPILS and LdTOPIL-fus-S) show functional

con-servation, i.e substrate specificity and CPT sensitivity

similar to those of eukaryotic topoisomerase IB

(Fig 4A,C) Therefore, gene fusion may account for

the control of noncovalent DNA binding or

coordina-tion of DNA contacts by other parts of the enzyme

Comparing the crystal structure of human and

vac-cinia topoisomerase I enzymes, it is evident that a

pre-cleavage conformational change in the core and

catalytic domains is necessary to establish the correct

position of the active site Tyr for nucleophilic attack

on DNA [19–21] This implies that the reverse fusion

(LdTOPIS-fus-L) may lead to a conformational change

in the topoisomerase architecture, leading to loss of

activity

Effect of deletion of 1–210 amino acids from the

N-terminus of LdTOPIS

The small subunit (LdTOPIS) shares 43.5% sequence

identity with the C-terminal domain of human

topoiso-merase I, including alignment of conserved sequences

surrounding the catalytic Tyr residue LdTOPIS

con-tains a large nonconserved N-terminal extension

(start-Met-Asn210), enriched in serine residues that might be

potential sites of phosphorylation [8] Reconstitution

of LdTOPIL with truncated LdTOPI-D(1–210)S shows

topoisomerase I activity (data not shown)

Interest-ingly when LdTOPI-D(1–210)S was fused to intact

LdTOPIL to create a single ORF LdTOPIL-fus-D(1–

210)S, it showed decreased topoisomerase I activity

and sensitivity towards CPT in plasmid DNA

relaxa-tion experiments compared to LdTOPIL-fus-S (Fig 4)

The reduced relaxation activities of LdTOPIL-fus-D(1–

210)S (Fig 3) correlated well with decreased cleavage

rates under suicidal conditions; that is,

LdTOPIL-fus-D(1–210)S showed a 10-fold reduction in cleavage rate

compared to LdTOPIL-fus-S This finding is consistent

with the results of the 25-mer duplex oligonucleotide

equilibrium cleavage assay A low level of cleavage

was observed for LdTOPIL-fus-D(1–210)S in the

pres-ence or abspres-ence of CPT compared to cleavage by

LdTOPIL-fus-S or LdTOPILS (Fig 4D)

Hence, we surmise that 1–210 amino acid residues

from the N-terminal end of the small subunit probably

act as a ‘pseudolinker’ in the fused LdTOPIL-fus-S

construct Owing to gene fusion, some additional contacts (1–210 amino acid of LdTOPIS) perhaps account for the prominent role in the cleavage step or

in the steps preceding cleavage, i.e DNA binding The later possibility is supported by  3-fold decreased binding affinity of the mutant LdTOPIL-fus-D(1–210)S (Kd of 1.9· 10)7m) compared to that of LdTOPIL-fus-S (Kdof 0.6· 10)7m) These findings are in keep-ing with those for human topoisomerase I, where it was demonstrated that the linker domain participates in a network of correlated movements with key regions of the enzyme involved in the human topoisomerase I cat-alytic cycle, providing a structural–dynamic explanation for the better DNA relaxation activity and CPT sensi-tivity of topo70 when compared to topo58⁄ 6.3 [21]

Role of His453 (LdTOPIL) in enzyme catalysis The catalytic activity of type IB topoisomerases is derived chiefly from five strictly conserved amino acid residues In human topoisomerase I, the residues con-stituting this active site pentad are Arg488, Lys532, Arg590, His632, and Tyr723 The analogous residues Arg314, Lys352, Arg410 and His453 are also con-served in the large subunit of the Leishmania enzyme, and the smaller catalytic subunit harbors the consensus SKXXY motif [8,23] In the present study we also investigated the role of His453 of LdTOPIL in the transesterification reaction Point mutations leading to changes in His265 of the structurally similar vaccinia topoisomerase I and His632 of human topoisomerase I have adverse effects on the transesterification reaction catalyzed by the two enzymes, and changes at this position appear to perturb the corresponding active sites somewhat differently [8,9,21] Unlike His, the Glu and Ala side chain mutants of LdTOPIL-fus-S show appreciable variation in their effects on concerted topo-isomerase I action We found that replacing His453 in fused Leishmania topoisomerase I with Glu caused a 40-fold reduction in the rate of relaxation and suicide cleavage With Ala, both relaxation and suicide clea-vage were reduced to nearly undetectable levels From the mutational analysis, it seems most likely that the active site His453 of LdTOPIL plays a major role in stabilizing the pentavalent transition state of the enzyme through an interaction with the nonbridging oxygen of the scissile phosphate of the DNA [10]

In conclusion, our gene fusion studies improve our knowledge of the unusual structure of L donovani het-erodimeric topoisomerase I Our study also shows that unconserved N-terminal extended regions of the small subunit (amino acids 1–210) have a role in controlling noncovalent DNA binding and CPT sensitivity Thus,