ABBS 2005,37(09): Study of MMLV RT- Binding with DNA using Surface Plasmon Resonance Biosensor

Jian-Long
ZHAO: Tel, 86-21-62511070-5709; Fax, 86-21-62511070-8714; E-mail,
[email protected]

Meng-Su
YANG: Tel, 852-27887797; Fax, 852-27887406; E-mail, [email protected]

Abstract        Surface
plasmon resonance biosensor technique was used to study the binding of Moloney
murine leukemia virus reverse transcriptase without RNase H domain (MMLV RT–) with DNA in
the absence and in the presence of inhibitors. Different DNA substrates, including
single-stranded DNA (ssDNA), DNA template-primer (T-P) duplex and gapped DNA,
were immobilized on the biosensor chip surface using streptavidin-biotin, and
MMLV RT–-DNA binding kinetics were analyzed by different models.
MMLV RT– could bind
with ssDNA and the binding was involved in conformation change. MMLV RT– binding DNA T-P duplex and gapped DNA
could be analyzed using the simple 1:1 Langmuir model. The lack of RNase H
domain reduced the affinity between MMLV RT– and T-P duplex. The effects of RT
inhibitors, including efavirenz, nevirapine and quercetin, on the interaction
between MMLV RT– and
gapped DNA were analyzed according to recovered kinetics parameters. Efavirenz
slightly interfered with the binding between RT and DNA and the affinity
constant in the presence of the inhibitor (KA=1.21´                                                                                                          

Key words       
surface plasmon resonance biosensor; reverse transcriptase; kinetics;
inhibitor

Reverse transcriptase
(RT) plays an important role in the life of retroviruses. RT possesses ribonuclease
H as well as RNA-directed and DNA-directed DNA polymerase activities. It can
convert a single-stranded RNA of the retrovirus into a double-stranded DNA for
integration into the host genome.

The inhibition of RT
polymerase activity is a major treatment method for human immunodeficiency
virus type 1 (HIV-1). HIV-1 RT inhibitors are subdivided into nucleoside
reverse transcriptase inhibitors (NRTIs) and non-nucleoside reverse
transcriptase inhibitors (NNRTIs). NRTIs are the analogs of nucleotides or nucleosides.
In vivo NRTIs are converted into triphosphate and incorporated into DNA,
which blocks the elongation of DNA. NNRTIs are largely hydrophobic inhibitors
and do not require intracellular metabolism for activity, so they can be
applied directly to study the interaction between RT and its inhibitors.
Structural evidence has shown that the allosteric NNRTIs bind tightly to a
hydrophobic pocket about 10 Å away from the polymerase site [1–3]. Steady-state kinetic studies suggested the inhibitors were non-competive
or uncompetitive with respect to the binding of DNA template-primer (T-P)
duplex [4–6]. Although there are many publications on the
mechanism of the function of NNRTIs, the kinetics of the interaction between RT
and DNA T-P in the presence of NNRTIs has not been well studied. Besides­
synthesized compounds, natural anti-HIV inhibitors have also been studied by
some researchers. In vitro experiments showed that several flavonoids,
including quercetin, myricetin, baicalein and quercetagenin, were inhibitors of
HIV-1 RT and moloney murine leukemia virus RT (MMLV RT) [7–10]. However, it is unclear how these flavonoids
act on RT and function as RT inhibitors.

Surface plasmon
resonance (SPR) biosensor technique has been proven to be a useful tool for obtaining
quantitative kinetic and affinity information on biomolecular interactions. An
SPR biosensor can translate a biospecific interaction between a ligand in
solution and a binding partner­ immobilized on the surface into a detectable
signal that is directly proportional to the extent of the interaction. The SPR
technique offers significant advantages because it is label-free and non-invasive and
results are in real time, which contributes significantly to the understanding
of the interaction between protein and DNA [11–14].

MMLV RT without RNase H domain (MMLV RT–) has been used as a
model to investigate RT binding with DNA in the absence and the presence of
inhibitors using an SPR biosensor. The elimination of the RNase H
domain of MMLV RT does not affect the structural integrity of the polymerase
domain [15]. MMLV RT is a monomer with a molecular weight of 75 kDa and has a
right-hand structure­ similar to HIV-1 RT. The fingers and palm domains of MMLV
RT resemble those of HIV-1 RT except that there are additional 16 residues at
the N-terminal, which relate to the monomer’s resistance to proteolytic
degradation and dimerization [2]. For both RTs, the active site of polymerase
is located at the junction of the fingers and palm domains, which has three
highly conserved aspartate residues­ required for polymerase activity [2].
Because of significant structural homology, the effects of inhibitors on MMLV RT– activity can provide valuable
information to develop agents against HIV-1 RT.

In the present study,
the binding characteristics of MMLV RT– to various DNA substrates, including
single-stranded DNA (ssDNA), DNA T-P duplex and gapped DNA, were determined and
compared systematically to establish the binding pattern of RT. Furthermore,
the effects­ of different inhibitors, including two known NNRTIs, efavirenz
(EFV) and nevirapine (NVP), and a natural product­ inhibitor, quercetin, on the
binding affinity of MMLV RT– and
binding modes with DNA were investigated.

Materials and Methods

Materials

Sensor chip CM5 of
research grade, HSB-EP buffer [MMLV RT– (SuperScript II RT) was purchased from
Invitrogen Life Technologies (EFV was generously
provided by Bristol-Myers Squibb Company (One 5‘-biotinylated
oligoribo(deoxy)nucleotide and two nonbiotinylated complementary strands (Table
1) were synthesized and purified by HPLC (Sangon,  

Immobilization of DNA
substrates on the biosensor chip

The CM5 chip was
modified by streptavidin according to the standard protocol (http://www.biacore.com).
After the CM5 sensor chip was fully equilibrated by HSB-EP buffer, 35 ml mixture of EDC ( 

Biosensor measurement of
the MMLV RT– binding with DNA

All SPR measurements
were carried out using BIAcoreX apparatus (Pharmacia Biosensor AB). The basic
principle of the SPR biosensor has been described in detail elsewhere [16].

All binding experiments
were carried out at Ten microliters of HSM
solution comprising MMLV RT– at
different concentrations was injected over the DNA-modified surface for 120 s,
then washed with HSM buffer for 200 s. The DNA-modified surface was regenerated
by washing with 10 ml 1% SDS-HSM solution
for 1 min to remove protein from the DNA substrates. The successive injection
of MMLV RT– solution was
carried out when the baseline reached a level approximate to that before the
previous­ injection.

Biosensor measurement of
MMLV RT– binding with DNA in the presence of different inhibitors

EFV, NVP and quercetin
have poor solubility in aqueous­ buffer, so stock solutions of these potent
inhibitors were prepared in dimethyl sulfoxide (DMSO), all at a concentration
of To study the nature of
MMLV RT– binding with
the gapped DNA in the drug solution, MMLV RT– was fully mixed with a certain inhibitor
at a constant concentration. The inhibitor in appropriate concentration, which
was determined­ to be 50 mM, can induce
a distinct response comparable to the response induced by the free RT binding­
with DNA, and will not increase the non-specific adsorption­ of MMLV RT– with reference surface. The mixtures
were injected over the gapped DNA modified surface for 120 s, then washed with
HSM buffer for 200 s.

Because the
concentration of each inhibitor (50 mM) in the
solution was more than 250-fold in excess of MMLV RT–, and NNRTIs can tightly
bind with RT [3], RT was assumed to be saturated with the compound and the
concentration­ of free enzymes could be omitted before the mixture was
injected.

Kinetics and data
analysis

All experiment data were
analyzed using BIAevaluation software (version 4.1; Pharmacia Biosensor AB).
The numerical­ integration algorithms used by BIAevaluation software are
sensitive to the sets of parameters and may deviate from the true kinetics.
Direct and global curve fitting­ is an optimum approach for data analysis
corresponding to the different possible models. It can avoid deviation caused
by limitation of the mass transport from the bulk solution to the sensor
surface or inhomogeneity of the binding sites [17]. Therefore all kinetic
analyses were performed by global curve fitting. Kinetic para­meters of the binding interactions
were derived from the response curves by non-linear curve fitting with various
possible kinetic models. The degree of randomness of the residual plot and the
reduced c2 value were used to assess the
appropriateness of the various models for analysis of the biosensor data. In all data
fittings, we considered the baseline drift. The value of the drift was less
than 0.05 response units (RU)/s in all the experiments, so the drift could not
cause significant deviation.

Results

The stability of
modified surfaces and specificity of MMLV RT– binding

Schematic
representations of the different DNA substrates­ captured on the
streptavidin-modified surface are shown in Fig. 1. Approximately 1.4´10–14 mol/mm2 of
streptavidin and 3.0´10–14 mol/mm2 of
DNA substrate were immobilized on the sensor chip surface according to the
calculation using the difference in the response levels before­ and after
immobilization, where 1000 RU corresponds to a surface density of approximately
1 ng protein (or 0.8 ng DNA) per square millimeter. Each immobilized
streptavidin molecule can bind with about two DNA molecules. No distinct change
of response level was observed after a typical experiment, so the surface
immobilized with DNA could be used repeatedly.

The specificity of MMLV
RT– binding with
DNA was tested by comparing the response level curves on the
streptavidin-modified surface before and after DNA immobilization. The
representative experiment data are shown in Fig. 2. Weak binding between
MMLV RT– and the
streptavidin-modified surface was observed which represented negligible non-specific
adsorption. Distinct non-linear association and dissociation were observed when
MMLV RT– flowed over
the DNA-modified surface. With 120 nM MMLV RT– interacting with the surfaces, about 330
RU was obtained at the end of association after the response data on the
streptavidin-coated surface were subtracted­ from those obtained on the
DNA-coated surface­ (Fig. 2), which verified the specificity of MMLV RT–-DNA binding.

MMLV RT– binding with
different DNA substrates immobilized on the sensor chips

The interactions between MMLV RT– at
different concentrations and the immobilized DNA substrates were measured in
real time (Figs. 3–5). The overall sensor responses­ increased as time went on and as the
concentration of RT increased.

Previous mechanistic
studies suggested a three-step binding­ model of T-P duplex with RT including
an initial binding phase and two subsequent conformation change phases [4,18].
Footprint analysis showed MMLV RT– protected­ the part of the T-P duplex as
far as position –15 and the template as
far as position +6 [15]. The DNA T-P duplex including an overhang single
template part (35-mer) and the gapped DNA including a duplex part (27 bp) in
the vicinity of the biosensor chip surface might provide the second RT binding
site as well as the DNA T-P part away from the surface. Therefore, the 1:1
Langmuir model (Equation 1), the conformation change model (Equation
2) and the parallel reaction model (Equation 3) were used to fit the
response curves of MMLV RT– binding with the different DNA substrates, where E and D
represent MMLV RT– in
solution and DNA immobilized on the biosensor chip, respectively, and (ED)x
represents a complex in another­ conformation different from the ED complex.
The corresponding residual plots are shown in Figs. 3–5.

The conformation change
model was appropriate to represent the MMLV RT– binding with the ssDNA due to the small c2 value and random residual distribution (Fig.
3). When fitting the response curves of MMLV RT– binding­ with the DNA T-P duplex or the
gapped DNA, the c2 value and residual randomness were
acceptable for all three models (Figs. 4 and 5). However, the
conformation change model and the parallel reaction model could not improve the
degree of randomness of the residual plots and reduce the c2 value compared with the 1:1 Langmuir
model. The standard deviations of several kinetics constants calculated­ by
complex models were at the same levels as the values of kinetics constants, indicating
SPR biosensor could not correctly analyze the detailed kinetics under the
present conditions. Therefore the 1:1 Langmuir model was used to calculate the
kinetics constants and the affinity constants­ of MMLV RT– binding with the DNA T-P duplex and the
gapped DNA (Table 2). The affinity of MMLV RT– for the ssDNA (KA=4.31´ 

The effects
of different inhibitors on the binding interaction­ between MMLV RT– and the
gapped DNA 

NNRTIs take effect
during DNA polymerization involving­ RT interaction with T-P. The gapped DNA
only provided one RT binding site. MMLV RT– can fully contact­ with the 15 bp duplex
and the 8-mer template overhang [15]. Therefore the gapped DNA was used to
study MMLV RT– binding with
DNA T-P in the presence of inhibitors. The responses of RT binding with the
gapped DNA in the presence of EFV, NVP or quercetin are shown in Fig. 7;
the residual plots based on the 1:1 Langmuir model showed the binding kinetics
were well described by the model. Analysis of the kinetics data demonstrated
the discrepancy­ in the effects of EFV, NVP and quercetin on RT binding with
the T-P part of the gapped DNA (Table 3). EFV slightly weakened the MMLV
RT– binding
capability with DNA. The affinity decreased approximately three-fold compared
with the affinity measured in the absence of the inhibitor (KA=1.21´106 M–1 vs. KA=4.61´ 

Discussion

This report demonstrated
the use of the SPR biosensor technique in the characterization of RT binding
with different­ DNA substrates, offering an analytical method for studying the
effects of small molecular inhibitors on macrobiomolecule interactions.

Previous studies have
suggested that wild HIV-1 RT bound efficiently with the hybrid duplex but
relatively weakly with single-stranded RNA. However our results indicated that
MMLV RT– bound
efficiently with both ssDNA (Fig. 3) and DNA T-P (Figs. 4 and
5), and RT binding with ssDNA was stronger than that with DNA T-P. The
ssDNA used in our experiments did not have a secondary structure at the temperature The SPR biosensor
technique can resolve complex mechanisms of biomolecular interactions
[11,21,22]. Effective­ use of a complex model to interpret SPR data depends on
many factors, such as quality of fit, structures­ and properties of the
components being studied, and comparison­ with results obtained by other
techniques. In order to minimize the deviation from true kinetics parameters,
all kinetic analyses were performed by global fitting. RT binding with DNA is a
complex kinetic process with initial collision of the enzyme and DNA followed
by conformation change based on the pre-steady kinetics study [4,18]. However,
the conformation change model could not satisfactorily analyze RT binding with
the DNA T-P duplex or the gapped DNA immobilized on the biosensor­ surface in
our experiments because of the large standard deviation of kinetics constants.
Similarly, the parallel reaction­ model was inadequate for the calculation of
kinetics­ parameters although there might be two different RT binding­ sites in
the DNA T-P duplex or the gapped DNA according to the length and components of
DNA [15]. Fig. 6 illustrates that RT binding with the second site was
weak or even negligible. The 75 kDa RT bound with the T-P part away from the
surface, which significantly interfered­ with the contact of the enzyme with
the DNA part in the vicinity of the surface. Because of strong binding­
capability with the ssDNA, a small number of MMLV RT– bound with the ssDNA part of the DNA T-P
duplex and slightly enhanced the total affinity of the binding. The overall­
reaction fitted well with the 1:1 Langmuir model, which could be used to
describe properties of the overall binding due to the predominance of
rate-limiting association­ and dissociation steps.

Both EFV and NVP are
non-nucleotide drugs against HIV-1 RT as approved by the  

References

 1   Arnold Eicalournal, Georgiadis MM, Jessen SM, Ogata CM,
Telesnitsky A, Goff SP, Hendrickson WA. Mechanistic implications from the
structure of a catalytic fragment of moloney murine leukemia virus reverse
transcriptase. Structure 1995, 3: 879–892

 2   Motakis D, Parniak MA. A tight-binding mode of
inhibition is essential for anti-human immunodeficiency virus type 1 virucidal
activity of nonnucleoside reverse transcriptase inhibitors. Antimicrob Agents
Chemother 2002, 46: 1851–1856

 3   Rittinger K, Divita G, Goody RS. Human immunodeficiency
virus reverse transcriptase substrate-induced conformational changes and the mechanism
of inhibition by nonnucleoside inhibitors. Proc Natl Acad Sci  4   Huang H, Chopra R, Verdine GL, Harrison SC. Structure of
a covalently trapped catalytic complex of HIV-1 reverse transcriptase:
Implications for drug resistance. Science 1998, 282: 1669–1675

 5   Maga G, Ubiali D, Salvetti R, Pregnolato M, Spadari S.
Selective interaction of the human immunodeficiency virus type 1 reverse
transcriptase nonnucleoside inhibitor efavirenz and its thio-substituted analog
with different enzyme-substrate complexes. Antimicrob Agents Chemother 2000,
44: 1186–1194

 6   Yang SS, Cragg GM, Newman DJ, Bader JP. Natural
product-based anti-HIV drug discovery and development facilitated by the NCI
developmental therapeutics program. J Nat Prod 2001, 64: 265–277

 7   Ono K, Nakane H, Fukushima M, Chermann JC,
Barre-Sinoussi F. Differential inhibitory effects of various flavonoids on the
activities of reverse transcriptase and cellular DNA and RNA polymerases. Eur J
Biochem 1991, 190: 469–476

 8   Tewtrakul S, Nakamura N, Hattori M, Fujiwara T, Supavita
T. Flavanone and flavonol glycosides from the leaves of Thevetia peruviana
and their HIV-1 reverse transcriptase and HIV-1 integrase inhibitory
activities. Chem Pharm Bull 2002, 50: 630–635

 9   Middleton E Jr, Kandaswami C, Theoharides TC. The
effects of plant flavonoids on mammalian cells: Implications for inflammation,
heart disease and cancer. Pharmacol Rev 2000, 52: 673–751

10  Tsoi
PY, Yang M. Surface plasmon resonance study of the molecular recognition
between polymerase and DNA containing various mismatches and conformational
changes of DNA-protein complexes. Biosens Bioelectron 2004, 19: 1209–1218

11  Tsoi
PY, Yang J, Sun YT, Sui SF, Yang MS. Surface plasmon resonance study of DNA
polymerases binding to template/primer DNA duplexes immobilized on supported
lipid monolayers. Langmuir 2000, 16: 6590–6596

12  Gorshkova
II, Rausch JW, le Grice SF, Crouch RJ. HIV-1 reverse transcriptase interaction
with model RNA-DNA duplexes. Anal Biochem 2001, 291: 198–206

13  Linnell
J, Mott R, Field S, Kwiatkowski DP, Ragoussis J, Udalova IA. Quantitative
high-throughput analysis of transcription factor binding specificities. Nucleic
Acids Res 2004, 32: e44

14  Wohrl
BM, Georgiadis MM, Telesnitsky A, Hendrickson WA, le Grice SF. Footprint
analysis of replicating murine leukemia virus reverse transcriptase. Science
1995, 267: 96–99

15  Sjolander
S, Urbaniczky C. Integrated fluid handling system for biomolecular interaction
analysis. Anal Chem 1991, 63: 2338–2345

16  Luo
J, Zhou J, Zou W, Shen P. Antibody-antigen interactions measured by surface
plasmon resonance: Global fitting of numerical integration algorithms. J
Biochem 2001, 130: 553–559

17  Wohrl
BM, Krebs R, Goody RS, Restle T. Refined model for primer/template binding by
HIV-1 reverse transcriptase: Pre-steady-state kinetic analyses of
primer/template binding and nucleotide incorporation events distinguish between
different binding modes depending on the nature of the nucleic acid substrate.
J Mol Biol 1999, 292: 333–344

18  Farmerie
WG, Loeb DD, Casavant NC, Hutchison CA 3rd, Edgell MH, Swanstrom R. Expression
and processing of the AIDS virus reverse transcriptase in E. coli.
Science 1987, 236: 305–308

19  Telesnitsky
A, Goff SP. RNase H domain mutations affect the interaction between Moloney
murine leukemia virus reverse transcriptase and its primer-template. Proc Natl
Acad Sci 20  Tsoi
PY, Yang MS. Kinetic study of various binding modes between human DNA
polymerase b and different DNA substrates by
surface-plasmon-resonance biosensor. Biochem J 2002, 361: 317–325

21  Lipschultz
CA, Li Y, Smith-Gill S. Experimental design for analysis of complex kinetics
using surface plasmon resonance. Methods 2000, 20: 310–318

22  Spence
RA, Kati WM, 23  Sluis-Cremer
N, Tachedjian G. Modulation of the oligomeric structures of HIV-1 retroviral
enzymes by synthetic peptides and small molecules. Eur J Biochem 2002, 269:
5103–5111

24  Tachedjian
G, Goff SP. The effect of NNRTIs on HIV reverse transcriptase dimerization.
Curr Opin Investig Drugs 2003, 4: 966–973

25  Tachedjian
G, Aronson HE, de los Santos M, Seehra J, McCoy JM, Goff SP. Role of residues
in the tryptophan repeat motif for HIV-1 reverse transcriptase dimerization. J
Mol Biol 2003, 326: 381–396

26  Tachedjian
G, Orlova M, Sarafianos SG, Arnold E, Goff SP. From the Cover: Nonnucleoside
reverse transcriptase inhibitors are chemical enhancers of dimerization of the
HIV type 1 reverse transcriptase. Proc Natl Acad Sci USA 2001, 98: 7188–7193

27  Kohlstaedt
LA, Wang J, Friedman JM, Rice PA, Steitz TA. 28  Pauwels
R, Peletskaya EN, Kogon AA, Tuske S, Arnold E, Hughes SH. Nonnucleoside
inhibitor binding affects the interactions of the fingers subdomain of human
immunodeficiency virus type 1 reverse transcriptase with DNA. J Virol 2004, 78:
3387–3397

29  Mizushina
Y, Iida A, Ohta K, Sugawara F, Sakaguchi K. Novel triterpenoids inhibit both
DNA polymerase and DNA topoisomerase. Biochem J 2000, 350: 757–763

30     Kim HJ Woo ER, Shin CG, Park H. A new
flavonol glycoside gallate ester from Acer okamotoanum and its
inhibitory activity against human immunodeficiency virus-1 (HIV-1) integrase. J
Nat Prod 1998, 61: 145–148