In the presence of phospholipid membranes, ESAT-6, but not CFP-10 and the complex, showed an increase in a-helical content and enhanced thermal stability.. The ITC and thermal unfolding
Trang 1complex formation confers thermodynamic and
biochemical stability
Akshaya K Meher1, Naresh Chandra Bal1, Kandala V R Chary2and Ashish Arora1
1 Molecular and Structural Biology, Central Drug Research Institute, Lucknow, India
2 Department of Chemical Science, Tata Institute of Fundamental Research, Mumbai, India
Comparative genomic studies based on whole genome
DNA microarray have led to the identification of 16
regions of deletion (RDs) in Mycobacterium bovis
BCG, which is currently used as a vaccine, with respect
to Mycobacterium tuberculosis and five RDs with
respect to M bovis RD1 is absent from all strains of BCG and Mycobacterium microti, whereas it is present
in all virulent strains of M tuberculosis and M bovis [1] The RD1 region in M tuberculosis is 9455 bp long, and encompasses nine ORFs (Rv3871–Rv3879c)
Keywords
association constant; ESAT-6–CFP-10
complex; limited proteolysis; lipid–protein
interactions; thermal unfolding
Correspondence
A Arora, Molecular and Structural Biology,
Central Drug Research Institute,
Lucknow 226 001, India
Fax: +91 522 223405
Tel: +91 522 261 2411 18 ext 4329
E-mail: ashishcdri@yahoo.com
(Received 4 January 2006, revised 30
January 2006, accepted 6 February 2006)
doi:10.1111/j.1742-4658.2006.05166.x
The 6-kDa early secretory antigenic target (ESAT-6) and culture filtrate protein-10 (CFP-10), expressed from the region of deletion-1 (RD1) of Mycobacterium tuberculosisH37Rv, are known to play a key role in viru-lence In this study, we explored the thermodynamic and biochemical chan-ges associated with the formation of the 1 : 1 heterodimeric complex between ESAT-6 and CFP-10 Using isothermal titration calorimetry (ITC), we precisely determined the association constant and free energy change for formation of the complex to be 2· 107m)1 and )9.95 kcalÆ mol)1, respectively Strikingly, the thermal unfolding of the
ESAT-6–CFP-10 heterodimeric complex was completely reversible, with a Tm of 53.4C and DH of 69 kcalÆmol)1 Mixing of ESAT-6 and CFP-10 at any tempera-ture below the Tmof the complex led to induction of helical conformation, suggesting molecular recognition between specific segments of unfolded ESAT-6 and CFP-10 Enhanced biochemical stability of the complex was indicated by protection of ESAT-6 and an N-terminal fragment of CFP-10 from proteolysis with trypsin However, the flexible C-terminal of CFP-10
in the complex, which has been shown to be responsible for binding to macrophages and monocytes, was cleaved by trypsin In the presence of phospholipid membranes, ESAT-6, but not CFP-10 and the complex, showed an increase in a-helical content and enhanced thermal stability Overall, complex formation resulted in structural changes, enhanced ther-modynamic and biochemical stability, and loss of binding to phospholipid membranes These features of complex formation probably determine the physiological role of ESAT-6, CFP-10 and⁄ or the complex in vivo The ITC and thermal unfolding approach described in this study can readily
be applied to characterization of the 11 other pairs of ESAT-6 family pro-teins and for screening ESAT-6 and CFP-10 mutants
Abbreviations
ANS, 8-anilinonapthalene-1-sulfonate; CFP-10, 10-kDa culture filtrate protein; DPC, dodecylphosphocholine; DSC, differential scanning calorimetry; ESAT-6, 6-kDa early secretory antigenic target; ESAT-6–CFP-10 complex, 1 : 1 complex of ESAT-6 and CFP-10; HSQC,
heteronuclear single quantum correlation; ITC, isothermal titration calorimetry; Myr2PtdCho, dimyristoyl- DL -a-phosphatidylcholine; Ni ⁄ NTA, nickel ⁄ nitrilotriacetic acid; RD1, region of deletion 1; trCFP-10, truncated 10-kDa culture filtrate protein.
Trang 2Rv3874 or esxB and Rv3875 or esxA encode the
proteins CFP-10 (10-kDa culture filtrate protein) and
ESAT-6 (6-kDa early secretory antigenic target),
respectively, which play a key role in virulence [2]
Both ESAT-6 and CFP-10 generate a specific Th-1
host immune response and have a strong diagnostic
potential for both the virulent form and latent form
of M tuberculosis [3] Several studies have shown
that RD1 and its flanking regions comprising ORFs
Rv3864–Rv3870 and Rv3880c–Rv3883c code for a
spe-cialized secretion system Esx-1, which is responsible
for secretion of ESAT-6 and CFP-10 [4,5] Recently it
has been shown that the secretion of ESAT-6 and
CFP-10 is also dependent on Esx-1-associated protein
EspA [6]
The genes encoding ESAT-6 and CFP-10 are
organ-ized as an operon and are cotranscribed [7] On the
basis of tryptophan fluorescence, CD and 1D
1H-NMR spectra, Renshaw et al [8] have shown that
ESAT-6 is a molten globule whereas CFP-10 is
unstructured in the native form Together, ESAT-6
and CFP-10 form a tight 1 : 1 complex Recently, the
NMR solution structure of the ESAT-6 and CFP-10
complex has been determined by Renshaw et al [9]
(PDB ID, 1WA8] In the complex, both the proteins
adopt helix–turn–helix hairpin conformation and are
orientated antiparallel to each other The contact
sur-face between ESAT-6 and CFP-10 is primarily
hydro-phobic, and van der Waals interactions between
ESAT-6 and CFP-10 run all along the length of the
helices of both proteins The surface features of the
complex, however, do not indicate its involvement with
any specific function; rather DNA binding, enzyme
activity and pore formation in lipid membranes can be
excluded on the basis of the structure Fluorescence
microscopy studies have shown that the flexible
C-ter-minal of CFP-10 in the complex is responsible for
spe-cific binding to macrophages and monocytes, on the
basis of which a role in receptor-mediated signaling
has been attributed to the complex [9] Whether
CFP-10 alone can bind to macrophages and monocytes in a
specific manner was, however, not explored
The ESAT-6 family contains proteins consisting of
nearly 100 residues M tuberculosis H37Rv has 22
members of this family, all of which are in tandem
pairs arranged in clusters [10] The ESAT-6 family of
protein pairs expressed from Rv0287 and Rv0288 as
well as Rv3019c and Rv3020c are secreted proteins
and form 1 : 1 heterodimeric complexes Moreover
these protein pairs, because of their close sequence
similarity, may also form nongenome Rv0287–
Rv3020c and Rv0288–Rv3019c complexes The
ESAT-6 and CFP-10 interaction is quite specific, and these
proteins do not form nongenome complexes with either Rv0287⁄ Rv0288 or Rv3019c ⁄ Rv3020c pairs Mutational analysis of ESAT-6 has been carried out recently to identify the key residues involved in com-plex formation with CFP-10, secretion, T-cell response and virulence of M tuberculosis H37Rv [11] Several residues essential for complex formation have been identified Mutation of these key residues results in disruption of complex formation and attenuation of virulence The results of mutational analysis have been explained in terms of a coiled-coil model for the ESAT-6–CFP-10 complex, with heptad repeats ‘abc-defg’ harboring positions at sites ‘a’ and ‘d’ for hydro-phobic residues
Hsu et al [12] have demonstrated that either the deletion of RD1 or disruption of the Rv3874-Rv3875 (cfp-10-esat-6) operon of RD1 results in loss of cyto-toxicity towards both pneumocytes and macrophages The behavior of these mutants is similar to that of BCG and in contrast with the well-established cytotox-icity of M tuberculosis H37Rv to macrophages Along similar lines, Guinn et al [13] have reported that H37Rv RD1 mutants with disruption of either of the genes Rv3870, Rv3871, Rv3874 (cfp-10), Rv3875 (esat-6) or Rv3876 grew minimally and produced no cell lysis in human macrophage-like THP-1 cell lines In the studies of both Hsu et al and Guinn et al it was found that the H37Rv RD1 mutants grew inside the host cells but were unable to cause cytolysis It was further demonstrated by Hsu et al that ESAT-6, either alone or in combination with CFP-10, but not CFP-10 alone, could cause disruption and eventual lysis of black lipid membranes prepared from diphytanoyl-phosphatidylcholine On the basis of this, Hsu et al [12] proposed that ESAT-6 may mediate lethal ion fluxes through plasma membranes of the host, leading
to cytolysis In proteomic studies, ESAT-6 has been found in the cell membrane fraction of M tuberculosis H37Rv [14] However, Guinn et al reported that addi-tion of purified ESAT-6, either alone or in combina-tion with CFP-10, did not show any toxic effect on THP-1 cells Therefore, the nature of the interaction of ESAT-6, CFP-10 or the complex with phospholipid membranes is currently not very clear
A detailed characterization of biochemical and ther-modynamic changes associated with complex forma-tion is necessary to fully understand the biological role
of ESAT-6, CFP-10 and the complex In addition, the nature of the interaction of ESAT-6, CFP-10 and the complex with phospholipid membranes needs to be understood clearly The results of our detailed bio-physical studies show that, compared with ESAT-6 or CFP-10, the complex has enhanced thermodynamic
Trang 3and biochemical stability ESAT-6, but not CFP-10 or
the complex, undergoes conformational change on
binding to the phospholipid membranes We also
stud-ied complex formation with CFP-10 and interaction
with phospholipid membranes for four mutants of
ESAT-6 We suggest biophysical characterization of
complex formation as a general approach that can be
used for all 11 pairs of ESAT-6 family proteins in
M tuberculosis H37Rv, and furthermore for screening
the entire set of ESAT-6 and CFP-10 mutants
Results
Thermodynamic parameters governing ESAT-6
and CFP-10 complex formation
Isothermal titration calorimetry (ITC) experiments
were carried out to accurately measure the association
constant for ESAT-6 and CFP-10 complex formation
The raw ITC data, generated by titration of 1.3 mL
0.42 mm ESAT-6 during the 50 injections of 4 lL
0.042 mm CFP-10 are shown in Fig 1A, and the
integ-rated areas under each peak versus molar ratio of
ESAT-6 to CFP-10 are plotted in Fig 1B The binding
isotherm of ESAT-6 with CFP-10 is characterized by
strong heat release, which is indicated by a slope approaching infinity The heat released decreases as ESAT-6 becomes saturated In the last 23 injections of the titration, only heat of dilution is observed The binding isotherm in Fig 1B was fitted to a single-site binding model for determination of thermodynamic parameters The solid line indicates best fit to the plot The parameters used in fitting were the stoichiometry
of association (n), the binding constant (KB) and the change in enthalpy (DHB) The values of these parame-ters obtained from the nonlinear least-squares fit to the binding curve are: n¼ 1.0, DHB¼)40.3 kcalÆ mol)1, and KB¼ 2 · 107m)1 The ITC binding iso-therm can be characterized by a unitless value c [15], which is given by c¼ KB[M]n, where [M] is the con-centration of the macromolecule ESAT-6 For an accu-rate determination of the binding constant, a ‘c’ value between 1 and 1000 is recommended In the case of ESAT-6 and CFP-10, the value of ‘c’ is 840, which is indicative of a tightly bound complex The free energy change (DG) associated with complex formation is given by: DG¼ –RTlnKB, where R is the gas constant and T is the temperature in Kelvin At 25C, DG for complex formation is )9.95 kcalÆmol)1 The entropy change associated with complex formation is deter-mined from the equation: DG¼ DH) TDS At 25 C,
DS is )101 calÆmol)1ÆK)1 Both the entropy change and enthalpy change associated with complex forma-tion are characteristically high However, typical enthalpy–entropy compensation results in a moderate value of DG of )9.95 kcalÆmol)1 The free energy change for complex formation between ESAT-6 and CFP-10 is comparable to the DG associated with simi-larly sized protein–protein interactions, e.g DG of )9.6 ± 0.5 kcalÆmol)1 was observed for interaction between turkey ovomucoid third domain with a-chy-motrypsin and DG of )11.3 ± 0.7 kcalÆmol)1 was observed for interaction between T-cell factor 4 and b-catenin [16,17]
Thermal unfolding of the ESAT-6–CFP-10 complex
is completely reversible Differential scanning calorimetry (DSC) studies were carried out to assess the thermal stability of the ESAT-6–CFP-10 complex and to accurately measure the enthalpy and heat capacity changes involved in the unfolding A DSC thermogram of the thermal unfold-ing of the complex at a concentration of 0.105 mm in phosphate buffer and a scan rate of 60CÆh)1, from 20
to 80C is shown by the solid line curve in Fig 2 After the first heating scan, the sample was cooled from 80 to 20C and then a second heating scan was
A
B
Fig 1 Typical calorimetric isothermal titration measurements of
the interaction of CFP-10 with ESAT-6 in phosphate buffer at
25 C (A) Raw data of heat effect (in lcalÆs)1) of 65 4-lL injections
of 0.42 m M CFP-10 into 1.3 mL 0.042 m M ESAT-6 performed at 4-s
intervals (B) The data points (d) were obtained by integration of
heat signals plotted against the molar ratio of ESAT-6 to CFP-10 in
the reaction cell The solid line represents a calculated curve using
the best-fit parameters obtained by a nonlinear least squares
fit The heat of dilution was subtracted from the raw data of
titra-tion of CFP-10 with ESAT-6.
Trang 4recorded, which is shown by the dotted line curve in
Fig 2 The peak shaped thermograms indicate
co-op-erativity during unfolding [18] The thermal unfolding
transition is characterized by an enthalpy change (DH)
of 69 kcalÆmol)1, Tm of 53.4C, and T1 ⁄ 2 of 9.01C
However, no change in heat capacity (DCp) was
observed for the thermal unfolding transition DSC
scans recorded at scan rates of 20, 40, 60 and
90CÆh)1showed only a small shift in the Tmfrom 54
to 53.4C and a small decrease in transition enthalpy
from 74 to 69 kcalÆmol)1 As the first and second
heat-ing scans completely overlap at every scan rate, it
strikingly indicates that the thermal unfolding of the
complex is completely reversible
The secondary and tertiary structural changes
asso-ciated with thermal unfolding of the complex were
followed by steady-state CD and 2D15N-1H
heteronu-clear single quantum correlation (HSQC) NMR
experi-ments, respectively Far-UV CD spectra of CFP-10,
ESAT-6 and ESAT-6–CFP-10 complex were similar to
those reported previously by Renshaw et al [8] As
CFP-10 is almost completely unstructured, the thermal
unfolding and refolding experiments were performed
only for ESAT-6 and the complex Steady-state CD
scans were recorded on a sample first at increasing
temperatures in the range 25–75C and then in
decreasing order from 75 to 25C, at 5 C intervals
The thermal unfolding and refolding profiles of
ESAT-6 and the complex are shown in Fig 3A The
midpoints of thermal unfolding transitions (Tm) of
Fig 2 Thermal reversibility of 1 : 1 ESAT-6–CFP-10 complex
monit-ored by DSC DSC thermogram of 0.51 mL 0.105 m M
ESAT-6–CFP-10 from 20 C to 80 C, at a scan rate of 60 C per h The raw data
were baseline-corrected for buffer The plots show excess heat
capacity as a function of temperature in C The complex was
hea-ted to 80 C for the first thermogram shown by the solid line.
The sample was then cooled down to 20 C The second
thermo-gram recorded by reheating the same sample is shown by a
dashed line.
Fig 3 Thermal reversibility of ESAT-6 and the 1 : 1
ESAT-6–CFP-10 complex monitored by CD (A) Normalized transition curves for temperature-induced transition of ESAT-6 and the complex monit-ored in the far-UV CD region at 222 nm Thermal unfolding (h) and thermal refolding (s) profile of ESAT-6 and thermal unfolding (n) and thermal refolding (e) profile of the complex were plotted as fraction of protein folded versus temperature in C (B) Far-UV CD spectrum of ESAT-6 (h) was recorded in phosphate buffer, pH 6.5
at 25 C The sample was heated to 70 C and cooled down to
25 C, and the far-UV CD spectrum was recorded again (s) (C) CD spectrum of the 1 : 1 complex at 25 C was recorded before
therm-al unfolding (h) and after thermtherm-al refolding (s) as described for ESAT-6.
Trang 5ESAT-6 and the complex are at 33C and 53 C,
respectively For the complex, the Tmdetermined from
CD (53C) matches well with that determined by
DSC (53.4C) CD spectra recorded before and after
unfolding, at 25C, for ESAT-6 and the complex are
shown in Fig 3B,C, respectively Similar to the
unfold-ing and refoldunfold-ing profiles mentioned above, entire CD
spectra before and after unfolding overlapped at every
temperature, suggesting that the molecular steps
lead-ing to thermal unfoldlead-ing are retraced on refoldlead-ing for
both ESAT-6 and the complex
The 2D 15N-1H-HSQC spectrum serves as a
finger-print of the overall structure of a protein The HSQC
spectrum recorded with 15N-labeled CFP-10 at 30 C
is shown in Fig 4A The spectrum is characterized by
sharp but narrowly dispersed peaks along the 1HN
dimension (within 7–8.5 p.p.m), which is consistent
with CFP-10 being unstructured in its native form The 2D15N-1H-HSQC spectrum of15N-labeled
ESAT-6 is shown in the Fig 4B The broad peaks and peak dispersion pattern in the HSQC spectrum are consis-tent with the previously reported molten globular state
of free ESAT-6 The HSQC spectrum of the complex formed between 15N-labeled CFP-10 and unlabeled ESAT-6 is shown in Fig 4D, and that of the complex formed between 15N-labeled ESAT-6 with unlabeled CFP-10 is shown in the Fig 4E Figure 4C shows the 2D 15N-1H-HSQC spectrum of the complex in which both the proteins are 13C,15N-labeled The sum of the HSQC spectra of individually labeled proteins in com-plex, i.e the sum of spectra in Fig 4D,E, is shown in the Fig 4F The spectrum in Fig 4F overlaps very well with the spectrum of the complex shown in Fig 4C To find any change in tertiary structure of the
Fig 4 Conformational change observed individually in ESAT-6 and CFP-10 on complex formation (A) and (D) show 15 N- 1 H-HSQC spectra of 15
N-labeled CFP-10 in the free state and in complex with unlabeled ESAT-6, respectively (B) and (E) show15N-1H-HSQC spectra of15
N-label-ed ESAT-6 in the free state and in complex with unlabelN-label-ed ESAT-6, respectively (C) 15 N- 1 H-HSQC spectrum of 1 : 1 [ 13 C, 15 N]ESAT-6– [ 13 C, 15 N]CFP-10 complex (F) Spectrum produced by addition of the spectra in (D) and (E) All spectra were recorded in NMR buffer (see Experimental procedures) containing 5% (v ⁄ v) D 2 O at 30 C on a 600-MHz NMR spectrometer.
Trang 6complex during the unfolding and refolding process,
15N-1H-HSQC spectra on 1 mm complex in phosphate
buffer were first recorded at 30, 40, 50, 55, 60 and
65C, in increasing order (Fig 5A,C,E,G,I,K,
respect-ively), after which HSQC spectra on the same sample
were recorded at 60, 55, 50, 40 and 30C
(Fig 5J,H,F,D,B, respectively), in decreasing order
The tertiary structure is retained up until 60C
Strik-ingly, the peaks in the HSQC spectrum at any
partic-ular temperature before and after unfolding almost
completely overlap, and are representative of the HSQC spectrum of the complex, but not the HSQC spectra of the individual proteins ESAT-6 and
CFP-10 This indicates that the tertiary structure of the complex is also completely regained after thermal unfolding
Molecular recognition between ESAT-6 and CFP-10 exists even when the two proteins are in unstructured form
As the secondary structure of ESAT-6 is highly dependent on the temperature, we investigated whether any residual secondary structure of ESAT-6 is neces-sary for complex formation with CFP-10 CD scans were recorded for samples in which ESAT-6 and
CFP-10 were mixed at 25, 30, 35, 40, 45, 50 and 55 C, and compared with CD scans of the complex formed between the two proteins at 25 C and heated to equivalent temperatures Fig 6 shows thermograms generated by plotting mean residue ellipticity at
222 nm as a function of temperature for ESAT-6, CFP-10, the 1 : 1 complex of ESAT-6–CFP-10, and equimolar CFP-10 and ESAT-6 mixed at different temperatures As can be seen, there was an increase in helical content equivalent to that of the complex when ESAT-6 and CFP-10 were mixed together at tempera-tures up to 55 C, indicating formation of helices locally by interactions between specific segments of CFP-10 and ESAT-6 These results indicate that the secondary structure of ESAT-6 is not necessary for the
I
K
J
Fig 5 Thermal reversibility of 1 : 1 ESAT-6–CFP-10 complex
monit-ored by NMR spectroscopy 1 m M [ 15 N]ESAT-6–[ 15 N]CFP-10
com-plex in NMR buffer, pH 6.5, with 5% (v ⁄ v) D 2 O was used to
monitor thermal reversibility of the complex 15 N- 1 H-HSQC spectra
were recorded on a 500-MHz NMR spectrometer at 30 C (A),
40 C (C), 50 C (E), 55 C (G), 60 C (I) and 65 C (K), in increasing
order, after which 15 N- 1 H-HSQC spectra on the same sample were
recorded at 60 (J), 55 (H), 50 (F), 40 (D) and 30 C (B), in
decreas-ing order.
Fig 6 Temperature dependence of the interaction of ESAT-6 and CFP-10 Isothermal CD spectra were recorded at 5 C temperature interval from 25 to 55 C A plot is shown of mean residue
elliptici-ty values at 222 nm as a function of temperature, recorded for ESAT-6 (h), CFP-10 (e), and 1 : 1 ESAT-6–CFP-10 complex formed
by mixing equimolar proteins at 25 C (n), and equimolar ESAT-6 and CFP-10 mixed together at 25, 30, 35, 40, 45, 50 and 55 C (d).
Trang 7complex formation, and specific molecular recognition
between the interacting segments of ESAT-6 and
CFP-10 exists even when the two proteins are in
unstructured form
CFP-10 reduces its susceptibility to trypsin
digestion on forming a complex with ESAT-6
To investigate the biochemical stability of the proteins,
limited proteolysis with trypsin was performed at 4C,
for ESAT-6, CFP-10 and the 1 : 1 ESAT-6–CFP-10
complex, and the digested products thus obtained were
analyzed by SDS⁄ PAGE (15% gel) The
Coomassie-stained SDS⁄ polyacrylamide gels are shown in
Fig 7A On trypsinolysis, CFP-10 showed multiple
bands on SDS⁄ PAGE after 1 min of digestion at 4 C,
and was completely digested to oligopeptides in
20 min ESAT-6 was stable for 60 min at 4C
Fur-ther degradation of ESAT-6 yielded two bands
corres-ponding to molecular masses of 14 kDa and 3 kDa
The 14-kDa band may be an aggregate of
trypsin-degraded products of 6 In contrast with
ESAT-6 and CFP-10, the complex displayed a characteristic
pattern on trypsinolysis On treatment of the complex
with trypsin at 4C, one additional band appeared
after 1 min incubation The largest and smallest of
these bands corresponded to CFP-10 and ESAT-6,
respectively A third band labeled trCFP-10 (for
trun-cated CFP-10), in between CFP-10 and ESAT-6, with
molecular mass 2 kDa lower than CFP-10 was observed, which apparently results from truncation of CFP-10 by cleavage at a particular site by trypsin On continued incubation, the intensity of the band corres-ponding to CFP-10 decreased, whereas that of
trCFP-10 increased with time, and no change in the intensity
of the band corresponding to ESAT-6 was observed After 2 h of trypsin treatment, the band corresponding
to intact CFP-10 had disappeared completely, whereas the bands corresponding to trCFP-10 and ESAT-6 were still present An essentially similar pattern of bands was observed for the complex after 3 h of tryp-sinolysis except that a weak band with an apparent mass of 6 kDa was observed, which resulted from fur-ther degradation of tr10 Both ESAT-6 and
CFP-10 have C-terminal hexa-histidine tags Western blots with antibody to histidine are shown in Fig 7B trCFP-10 was not detected, indicating that it results from cleavage of the C-terminus of CFP-10 Overall, these results indicate that complex formation leads to interdependent protection of an N-terminal fragment
of CFP-10 and ESAT-6 from trypsinolysis
ESAT-6 possesses solvent-exposed hydrophobic clusters
To assess the solvent-exposed hydrophobic surface of the proteins, we studied the change in fluorescence intensity of 8-anilino-1-naphthalenesulfonate (ANS) on
A
B
Fig 7 Limited proteolysis with trypsin of ESAT-6, CFP-10 and 1 : 1 ESAT-6–CFP-10 complex (A) SDS⁄ PAGE of aliquots removed at differ-ent time points for reaction of 40 l M ESAT-6, or CFP-10, or 1 : 1 ESAT-6–CFP-10 complex with 1 lg trypsin at 4 C Lanes 1, 4, 7, 10, 13,
16, and 19, CFP-10; lanes 2, 5, 8, 11, 14, 17, and 20, ESAT-6; lanes 3, 6, 9, 12, 15, 18, and 21 ESAT-6–CFP-10 correspond to aliquots withdrawn after 0, 1, 5, 20, 60, 120 and 180 min of trypsinolysis LMW is low-molecular-mass protein marker (B) Western blot developed with antibody to histidine The lanes of the blot correspond to the lanes of SDS ⁄ PAGE, except for LMW.
Trang 8binding to ESAT-6, CFP-10 and ESAT-6–CFP-10.
Figure 8 shows extrinsic fluorescence spectra of ANS
in the presence of ESAT-6, CFP-10 and the complex,
at 25 C The fluorescence intensities have been nor-malized with respect to the maximum fluorescence intensity of ANS bound to ESAT-6 As expected from its molten globule state, ESAT-6 showed high ANS binding No change in fluorescence intensity of ANS was observed in the presence of CFP-10, indicating that ANS did not bind to CFP-10, as expected from the unstructured form of CFP-10 A decrease of
65 ± 5% in ANS fluorescence intensity was obtained
on ESAT-6–CFP-10 complex formation
Myr2PtdCho vesicles stabilize the secondary structure of ESAT-6 above its melting temperature
To investigate the binding of ESAT-6, CFP-10 and the complex to lipid membranes, 6 lm protein samples were incubated with dimyristoyl-dl-a-phosphatidylcho-line (Myr2PtdCho) vesicles in phosphate buffer, and the change in conformation was monitored by CD spectroscopy CD spectra of CFP-10, ESAT-6–CFP-10 and ESAT-6 in the absence and presence of Myr2 Ptd-Cho vesicles are shown in Fig 9A,B,C At 25 C, the
Fig 9 Far-UV CD spectra of ESAT-6, CFP-10 and the 1 : 1 ESAT-6–CFP-10 complex in the presence of Myr 2 PtdCho vesicles CD spectra of
6 l M CFP-10, ESAT-6–CFP-10 and ESAT-6 without Myr2PtdCho vesicles in phosphate buffer, pH 6.5, at 25 C (h) and 37 C (s) and with Myr2PtdCho vesicles in phosphate buffer, pH 6.5, at 25 C (n) and 37 C (,) are shown The spectra obtained at 25 C after cooling the pro-tein samples containing Myr 2 PtdCho vesicles from 37 C, are shown with symbols (e).
Fig 8 Binding of ANS to ESAT-6, CFP-10 and the 1 : 1 ESAT-6–
CFP-10 complex The fluorescence emission spectra of 100 l M
ANS in the presence of 10 l M 6 (s), CFP-10 (h) and
ESAT-6–CFP-10 complex (m) in phosphate buffer, pH 6.5, at 25 C.
Trang 9CD spectra of CFP-10 and the complex did not show
any significant change, whereas ESAT-6 showed a
minor increase in helicity (from 49% to 52%) in the
presence of Myr2PtdCho vesicles When the
tempera-ture of the sample was increased to 37C, CFP-10 and
the complex still showed no change However, ESAT-6
retained an a-helical content of 32% in contrast with
19% in the absence of Myr2PtdCho vesicles at 37C
On cooling the same ESAT-6⁄ Myr2PtdCho vesicle
sample from 37C to 25 C, the a-helical content
increased further to 63%, which is significantly higher
than the helicity obtained on mixing ESAT-6 and
Myr2PtdCho vesicles at 25C
Interaction of ESAT-6 mutants with CFP-10 and
phospholipid membranes
We have used a novel approach to select residues for
mutations from the 26 residues of ESAT-6 that are at
the interface between ESAT-6 and CFP-10 in the com-plex, as reported by Renshaw et al [9] Our approach was based on detection of NOEs from the backbone amide protons of ESAT-6 to the side chain protons of CFP-10 Residues of ESAT-6, the amide protons of which showed strongest NOEs with the side chain protons of CFP-10 in the labeled complex, were selected for mutation For detecting NOEs, we prepared the complex from13C,15N-labeled CFP-10 and2H,13C,15 N-labeled ESAT-6 A set of 3D triple-resonance experi-ments HNCO, HNCA, and HN(CA)CB were recorded
to validate our sample Strips from HNCA and HN(CA)CB spectra demonstrating the sequential assignments of residues Leu39 to Trp43 are shown in Fig 10A,B, respectively These assignments are similar
to those reported by Renshaw et al [9] An15N-edited NOESY-HSQC spectrum was recorded for the complex for detecting the NOEs NOEs from backbone amide protons of ESAT-6 and side chain protons of CFP-10
A
B
C
Fig 10 Sequential assignments and
inter-protein NOEs for a segment of ESAT-6
interacting with CFP-10 (A) and (B) Strips
showing the sequential assignments from
3D HNCA and HN(CA)CB spectra,
respect-ively, recorded from 1 m M 1 : 1 complex of
2
H,13C,15N-labeled ESAT-6 and13C,15
N-labe-led CFP-10 in NMR buffer with 5% (v ⁄ v)
D2O at 30 C on a 600-MHz NMR
spectro-meter The strips are taken at the indicated
15 N chemical shifts that were assigned to
residues 39–43 of ESAT-6 They are
cen-tered about the corresponding amide proton
chemical shifts The top of the
sequence-specific assignments is indicated by
one-letter amino-acid code and by sequence
number The one directional arrows in these
figures indicate a sequential walk through
2D 13 Ca- 1 H N and 13 Cb- 1 H N planes taken
in the position of the corresponding
1 HN, 15 N, 13 Caand 1 HN, 15 N, 13 Cbresonances
in 3D HNCA and HN(CA)CB spectra,
respectively (C) Strips from1H,15
N-NOESY-HSQC spectrum recorded with smixof
150 ms In these strips, NOEs are shown
between downfield amide protons and
upfield aliphatic protons The amide protons
correspond to the sequentially assigned
seg-ment 39–43 of ESAT-6 The backbone
amide protons of this segment show NOEs
with protons at 0.808 p.p.m from a side
chain of CFP-10.
Trang 10were observed for the segments Ala14-Ala15-Ser16
(1.187 p.p.m.), Ala17-Ile18 (1.200 p.p.m.), Ser24-Ile25
(0.934 p.p.m.), Leu28-Leu29-Asp30 (0.897 p.p.m.),
Glu31-Gly32-Lys33-Gln34-Ser35-Leu36 (0.745 p.p.m.),
Leu39-Ala40-Ala41-Ala42-Trp43 (0.808 p.p.m.), and
Glu64-Leu65-Asn66 (1.415 p.p.m.) Values in
paren-theses are the chemical shift of the side chain protons of
CFP-10 with which backbone amide proton of ESAT-6
show the NOE Figure 10C shows the NOE between the
amide protons for the segment Leu39 to Trp43 from
ESAT-6 to the side chain proton of CFP-10 Strongest
NOEs were observed for the residues Leu29, Gly32, Ala41 and Leu65 On the basis of this, four point mutants L29D, G32D, A41D and L65D of ESAT-6 were generated We studied complex formation between ESAT-6 mutants and CFP-10 by CFP-10 pull-down assays and CD spectroscopy In parallel, we also studied the interaction of ESAT-6 mutants with Myr2PtdCho membranes by CD spectroscopy
SDS⁄ PAGE of the CFP-10 pull-down assay is shown in Fig 11A Two prominent low-molecular-mass bands corresponding to untagged CFP-10 and
A
B
Fig 11 Study of complex formation between ESAT-6 mutants and CFP-10 (A) A SDS ⁄ 15% polyacrylamide gel showing results of CFP-10 pull-down assay LMW, low-molecular-mass protein marker The rest of the lanes show purified ESAT-6 or ESAT-6 mutants and Ni ⁄ NTA eluate (see Experimental procedures) (B) Far-UV CD spectra of CFP-10 (h), ESAT-6 mutants (n) and 1 : 1 mixture of ESAT-6 mutant and CFP-10 (s) recorded in phosphate buffer, pH 6.5, at 25 C.