Periplasmic Binding Protein Dimer Has a Second ... - ACS Publications

Sep 6, 2017 - ... and Engineering Division, Oak Ridge National Laboratory, Oak Ridge, ... Caeley Gullett, Dean A. A. Myles, Pratul K. Agarwal, and Mat...
0 downloads 0 Views 19MB Size
Article Biochemistry is published A periplasmic binding by the American Chemical Society. 1155 protein dimer a Sixteenth has Street N.W., Washington, DC 20036 Published by American Chemical Subscriber access provided by Society. UNIVERSITY Copyright © American OF ADELAIDE LIBRARIES Chemical Society. However, no copyright

second allosteric event tied to ligand binding Biochemistry is published by the American

Le Li, Sudipa Chemical Ghimire-Rijal, Society. 1155 Sixteenth Street N.W., Sarah L Lucas, Christopher Washington, DC 20036 PublishedWright, by American B. Stanley, Edward Chemical Subscriber access provided by Society. UNIVERSITY Copyright © American OF ADELAIDE LIBRARIES Chemical Society. However, no copyright

Pratul K. Agarwal, Dean A. A. Myles, and Matthew J. Cuneo Biochemistry is published

Biochemistry, Just by theAccepted American Chemical Society. 1155 Manuscript • DOI: 10.1021/ Sixteenth Street N.W., acs.biochem.7b00657 • Publication Washington, DC 20036 Date (Web): 06 Sep 2017 Published by American Chemical Subscriber access provided by Society. UNIVERSITY Copyright © American OF ADELAIDE LIBRARIES Chemical Society. However, no copyright

Downloaded from http:// pubs.acs.org on September 7, 2017 Biochemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., DC 20036 Just Accepted Washington, Published by American Chemical Subscriber access provided by Society. UNIVERSITY Copyright © American OF ADELAIDE LIBRARIES Chemical Society. However, no copyright

“Just Accepted” manuscripts have been pe online prior to technical editing, formatting f Biochemistry is published the American Society providesby“Just Accepted” as a f Society. 1155 dissemination of Chemical scientific material as soon Sixteenth Street N.W., appear in full in PDF formatDCaccompanied by Washington, 20036 Published by American fully peer reviewed, but should not be consid Chemical Society. Subscriber access provided UNIVERSITY readers and citable byby the Digital Object Ide Copyright © American OF ADELAIDE LIBRARIES Chemical Society. However, no copyright

to authors. Therefore, the “Just Accepted” in the journal. After a manuscript is technic Biochemistry is published as an AS Accepted” Web site and published by the American changes to the manuscript text Chemical Society. 1155and/or gra Sixteenththat Street apply N.W., to the jo and ethical guidelines DC 20036 or consequencesWashington, arisingbyfrom the use of in Published American Chemical Subscriber access provided by Society. UNIVERSITY Copyright © American OF ADELAIDE LIBRARIES Chemical Society. However, no copyright

Page 1 of 37 Biochemistry Homodimer 1 2 3 4 5 6 7

Monomer

Allostery

ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

A periplasmic binding protein dimer has a second allosteric event tied to ligand binding Le Li1,‡ Sudipa Ghimire-Rijal1,‡, Sarah L. Lucas1,2, Christopher B. Stanley1, Edward Wright3, Pratul K. Agarwal3,4, Dean A. Myles1 and Matthew J. Cuneo1,5*

1

Neutron Sciences Directorate, Oak Ridge National Laboratory, Oak Ridge, TN 37831, USA. Department of Biomedical Engineering, North Carolina State University, Raleigh NC 27607. 3 Department of Biochemistry and Cellular and Molecular Biology, University of Tennessee, Knoxville, TN 37996. 4 Computational Biology Institute and Computer Science and Engineering Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831, USA. 2

*

To whom correspondence should be addressed: Matthew J. Cuneo, Biology and Biomedical Sciences Group, Biology and Soft Matter Division, Oak Ridge National Laboratory, Bethel Valley Rd, Oak Ridge TN, USA, Tel: (865) 241-8270; Fax: (865) 574-4403; E-mail: [email protected]

L.L. and S.G. contributed equally to this work.

1 ACS Paragon Plus Environment

Page 2 of 37

Page 3 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

KEYWORDS: ABC transport; allostery; periplasmic binding protein; protein interaction; smallangle neutron scattering ABSTRACT The ligand-induced conformational changes of periplasmic binding proteins (PBP) play a key role in the acquisition of metabolites in ATP binding cassette (ABC) transport systems. This conformational change allows for differential recognition of the ligand occupancy of the PBP by the ABC transporter. This minimizes futile ATP hydrolysis in the transporter, a phenomenon in which ATP hydrolysis is not coupled to metabolite transport.

In many systems, the PBP

conformational change is insufficient at eliminating futile ATP hydrolysis. Here we identify an additional state of the PBP that is also allosterically regulated by the ligand. Ligand binding to the homodimeric apo PBP leads to a tightening of the interface α-helices so that the hydrogen bonding pattern shifts to that of a 310 helix, in-turn altering the contacts and the dynamics of the protein interface so that the monomer exists in the presence of ligand.

2 ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 37

INTRODUCTION Allostery is an important mechanism through which the activities 1, dynamics functional states

3

of proteins are regulated 4.

2

and

In the periplasmic binding protein (PBP)

superfamily the conformation of the PBP is coupled to the ligand occupancy of the interdomain ligand binding site, which also allosterically regulates binding of the PBP to its cognate membrane bound ATP binding cassette (ABC) metabolite transport system. Central to this mechanism is the ligand-induced conformational change, which serves as a means to discriminate between the apo and ligand bound protein by the transporter and other cellular systems where the PBP motif functions in transmembrane signaling 5-7. In the absence of ligand, the two domains of the PBP are in an open conformation, with ligand binding at the interdomain interface stimulating a hinge-bending mediated reorientation of the two domains by as much as 60° 8. The extent of the conformational change of the PBP is thought to provide a mechanism that minimizes a phenomena known as futile ATP hydrolysis, in which ATP is hydrolyzed by ABC transporters in the absence of a bound-transportable ligand 9, 10. The formation of the closed ligand bound state alters the molecular envelope that is presented to the ABC transporters, leading to a preferential interaction over the apo protein. Interaction of the ligand bound PBP with the ABC transporter leads to ligand translocation from the periplasmic space to the cytosol in an ATP-dependent manner. Although there is significant mechanistic diversity among the various ABC transporters, a general theme emerges in which ligand occupancy of the PBP regulates ATP hydrolysis rates

11

. Variations do exist, but in a general sense the ligand bound

PBP has a higher affinity for the inward-facing state of the ABC transporter, thus leading to coupling between ATP hydrolysis and ligand transport 12, 13. In many instances, however, ligand binding only induces small conformational changes, on the order of less than 15°. This leads to

3 ACS Paragon Plus Environment

Page 5 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

little discrimination between ligand bound and apo proteins, which inadequately signals the ligand occupancy state to the ABC transporters, consequently leading to ATP hydrolysis in the absence of ligand

14, 15

. Moreover, because of the large area of the protein:protein interaction

surface formed between the ABC transporter and each PBP domain, even in cases where PBPs undergo hinge bending motions of a greater magnitude, the apo proteins can potentially retain sufficient affinity to stimulate ATP hydrolysis relative to the resting state

16-18

. These scenarios

can lead to little discrimination of the ligand occupancy of the PBP and higher levels of ATP hydrolysis which are uncoupled from the metabolic reward of ligand transport. Allosteric regulation is a commonly emerging theme in ABC transporter biochemistry 19, 20

.

It has been established that the canonical PBP ligand induced hinge bending motion

allosterically regulates, and is central to, proper functioning of ABC transport and other cellular systems where the PBP motif functions. However, there is evidence that this paradigm may be more complex, with PBPs playing a more active role in the regulation of transport. Indeed, there are an increasing number of PBPs that are not found in a monomeric quaternary state, with the suggestion that higher order oligomers play a unique functional role in the transport process in these systems 21-26. Here we identify an additional protein:protein interaction that is also allosterically regulated by the ligand occupancy of the PBP binding site. Recently, two mannose binding proteins from T. maritima (tmMnBP3 and tmMnBP6) were crystallized in the presence ligand, in addition to characterization of the ligand specificity and affinity profile27.

These two proteins

bind a variety of beta-linked saccharides. In the presence of two of their tight binding ligands (mannobiose and mannohexaose), these proteins were found to be monomeric in the crystal structure27. Although most PBPs are known to be monomeric, we demonstrate that tmMnBP3/6

4 ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

and an additionally distantly-related PBP utilize a conserved interface to form a homodimer, of which the quaternary state is modulated by the presence of ligand at the canonical ligand binding site. The homodimer interaction site is a molecular two-fold axis composed of α-helices. We identify residues in the ligand binding site that transmit the ligand occupancy of the binding site to the protein:protein interface ~15 Å away through tightening of the interface α-helices into 310 helices. The two-fold nature of this interface leads to this simultaneous transformation in both binding half sites and homodimer dissociation. The static snapshots of this allosteric switch were further investigated using detailed computer simulations, which highlight the dynamical communication network that couples active site occupancy to regulate transporter interactions. The identified homodimeric allosteric interface is not a unique case found in a single PBP. Sequence, structural and biochemical analysis of three carbohydrate binding PBPs from Thermotoga maritima reveals the homodimer interface and allosteric switch is conserved. The PBPs analyzed in this study do not undergo large conformational changes and this allosteric dimer to monomer transition serves as an additional means to drastically alter the molecular envelop of the PBP in a ligand dependent manner. We postulate this additional allosteric ligand induced conformational change may be a previously unidentified regulatory mechanism in ABC transport. MATERIALS AND METHODS Protein expression and purification The tmMnBP3 and tmMnBP6 proteins were expressed and purified as previously described 27. Circular Dichroism

5 ACS Paragon Plus Environment

Page 6 of 37

Page 7 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

CD experiments were performed on a Jasco CD spectrophotometer. Thermal denaturation midpoints were determined by measuring the CD signal at 225 nm as a function of temperature using 0.5 µM protein in 10 mM Tris-HCl pH 8.0 and 40 mM NaCl. In the absence of guanidinium chloride, the proteins were too stable to exhibit temperature-induced denaturation and all measurements were carried out in the presence of 2 M guanidine hydrochloride and 0.5 mM ligand. CD measurements were fit to a either a two-state model 28 or a three state model 29. Small angle neutron scattering data collection and analysis Small-angle neutron scattering (SANS) experiments were performed on the extended Qrange small-angle neutron scattering (EQ-SANS, BL-6) beam line at the Spallation Neutron Source (SNS) located at Oak Ridge National Laboratory (ORNL)30. Small-angle X-ray scattering experiments were performed on a Rigaku BIO-SAXS system. Protein was concentrated to 5-7 mg/mL and dialyzed in to 20 mM Tris pD 8.0, 150 mM NaCl in 100 % D2O for SANS measurements that were performed at 20 oC or at 65 oC for SAXS measurements. Ligands were added to the apo protein at a concentration of 0.5 mM for measurements of the ligand bound form.

Data

reduction

followed

standard

procedures

using

MantidPlot

(http://www.mantidproject.org/) 31. Upon verifying a Guinier regime

32

in the SANS profiles, the pair distance distribution

function, P(r), was calculated from the scattering intensity using the indirect Fourier transform method implemented in the GNOM program

33

(Table 1). The real-space radius of gyration, Rg,

and maximum linear dimension, Dmax, were determined from the P(r) solution to the scattering data. The CRYSON program

34

was used to calculate SANS curves from the PDB coordinate

files to compare with the experimental data. Then the OLIGOMER program 35 was used to fit the

6 ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 37

experimental SANS data using a combination of the calculated monomer and dimer SANS curves.

Table 1. SANS data analysis Tm (°C in 2M GdCl)

Dmax (Å)

tmMnBP3 Apo

98.7

108

tmMnBP3 Sat (0.5 mM Mannobiose)

~110

Sample

tmMnBP3 Apo (N231A) tmMnBP3 Sat (N231A) tmMnBP3 Apo (M239R) tmMnBP3 Sat (M239R) tmMnBP6 Apo tmMnBP6 Sat (0.5 mM Mannobiose) tmMnBP6 Apo (N231A) tmMnBP6 Sat (N231A) tmMnBP6 tmMnBP6 (M239R) tmMnBP6 Sat (M239R)

Rg(calc) (Å)

Rg(exp) (Å)

Oligomer Fit Monomer Homodimer (%) (%)

Oligomer χ2

Monomer

Homodimer

36.0± 0.05

24.1

35.4

3

97

5.4

108

27.2± 0.1

23.3

33.6

87

13

9.0

87.7/98.1

100

35.2± 0.1

24.1

35.4

22

78

12.2

88.4/103.1

100

33.5± 0.1

23.3

33.6

48

52

10.3

88.4/98.5

93

25.4± 0.2

24.1

35.4

99

1

11.2

88.5/~110

66

22.7± 0.1

23.3

33.6

100

0

12.4

84.7

104

35.6± 0.03

24.5

36.3

1

99

8.1

89.1

100

28.9± 0.07

23.7

33.2

74

26

6.3

86.6

105

36.0± 0.2

24.5

36.3

9

91

3.4

87.8

102

33.9± 0.2

23.7

33.2

54

46

7.5

86.9

102

27.2± 0.3

24.5

36.3

94

6

5.3

88.7

104

25.8± 0.4

23.7

33.2

100

0

6.4

7 ACS Paragon Plus Environment

Page 9 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Crystallization and X-ray data collection tmMnBP6 was concentrated to 20 mg/mL and dialyzed into 10 mM Tris, 40 mM NaCl 0.5 mM TCEP for crystallization. Crystals of tmMnBP6 were grown by hanging drop vapor diffusion in drops containing 2 µL of the protein solution mixed with 2 µL of 0.2–0.3 M magnesium acetate or calcium acetate, 20–30 % (wt/vol) PEG 3350 equilibrated against 900 µL of the same solution. Crystals were transferred to 35 % (wt/vol) PEG 3350 for cryoprotection, mounted in a nylon loop, and flash frozen in liquid nitrogen. All X-ray diffraction data were collected at 100 K on a Rigaku 007HFmicromax X-ray generator with a Raxis IV++ detector. The diffraction data were scaled and indexed using HKL3000 36. The data collection statistics are listed in Table 2.

Table 2. Data collection and refinement statistics tmMnBP6 Apo Data Collection Resolution Range (Å) Unique reflections a

Redundancy Mean I/σa

Completeness (%)a Rmerged (%)a Unit Cell Dimensions (Å) Space Group Refinement Num. of Reflections (working set/test set)

50.0-2.0 42275 5.7 (5.7) 11.5 (3. 4) 97.1 (96.2) 11.0 (62.7) a=51.3 b=65.8 c=185.7 P21221

42163/2131

17.3/20.9 Rcryst/Rfree Non-hydrogen atoms in refinement All 4834 Protein 4495 r.m.s.d. from ideal 0.003 Bond lengths (Å)

8 ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Bond angles (°) Average B-factors (Å2) Protein Solvent

Page 10 of 37

0.654 28.2 27.9

a

Number in parentheses represent values in the highest resolution shell

Structure determination, model building and refinement The initial structure was solved by molecular replacement using the Phaser program

37

with the ligand bound form of tmMnBP6 (PDB Code 4PFU) as a search model. Manual model building was carried out in COOT

38

and refined using PHENIX39. The models exhibit good

stereochemistry as determined by MolProbity 40; final refinement statistics are listed in Table 2. For analysis of the homodimer interfaces, the apo homodimer of tmMnBP6 used is a crystallographic homodimer generated through application of an defined -x,y,-z crystallographic symmetry operation to produce the complete biological assembly. The asymmetric unit of tmMnBP3 (PDB code 1VR5) represents the same homodimeric state as found in tmMnBP6 yet requires a x,y,z-1 symmetry operation to the PDB coordinates to generate the more biologically relevant asymmetric unit. This transform of the asymmetric units was utilized for computational studies and apo tmMnBP3/6 in solution. Ligand bound homodimer models of tmMnBP3/6 were made through superposition of ligand bound monomeric crystal structures with the structure of the apo tmMnBP homodimer models as described above. Computer Simulations The apo and substrate bound enzymes complexes were modeled using the X-ray coordinate, under explicit solvent conditions. The crystallographic apo homodimers (ApoDimer) 9 ACS Paragon Plus Environment

Page 11 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

used are constructed from the atomic coordinates of the apo proteins as described above. Another state of the proteins used for simulations is a mixed heterodimer model (MixHetDimer), where one monomer was from the apo crystal structure and the other was ligand bound. Both model MixHetDimer states were generated through superposition of the N-terminal domain of a monomer of a mannobiose bound tmMnBP3/6 with the homodimer of the apo protein. The pre-processing steps and simulations were performed using the AMBER simulation package

41

. Simulations were performed with AMBER ff14SB (for protein residues) and

GLYCAM_06EPb (for the substrate ligand) force-fields and SPC/E water model. For system preparation, the protein-ligand complex was solvated and the system was neutralized through the addition of Na+ counterions. After the pre-processing steps, the system was equilibrated using the protocol described previously

42

. All production runs were performed at 300K under NVE

(constant volume and energy) conditions. The apo and substrate bound complex with ligands were simulated for 200 nanoseconds for each system. AMBER’s GPU-enabled pmemd simulation engine was used for equilibration and production runs. Backbone (Cα) and all-atom flexibility of simulation trajectories was determined from the RMSF, computed by aggregating the magnitude of displacement eigenmodes computed using the quasi-harmonic analysis (QHA) in the ptraj analysis module in AMBER. As described previously 43, only the top 10 QHA modes were used in the analysis to focus on the principal dynamics or long time-scale fluctuations in the proteins. For determination of the network, significantly correlated and anti-correlated residue pairs were determined in the following way. AMBER’s ptraj analysis module was used to calculate the correlation matrix (by residues). For comparison between apo dimer and mixed hetero dimer, the matrices were normalized by subtracting the mean value and dividing by the

10 ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 37

standard deviation. The residue pairs over normalized correlation coefficients over +15 and below -5 were considered significant. The residue pairs showing significant correlations were between the ligand binding site and dimer interface used to identify the allosteric networks, as previously described44. Analytical Ultra-centrifugation Sedimentation velocity experiments were performed in a Beckman XL-I analytical ultracentrifuge using an An-50 Ti rotor. Samples were placed in charcoal-filled Epon double sector centerpieces. The reference sector contained 400 µL 20 mM Tris pH 8.0, 150 mM NaCl. The sample sector contained 390 µL of protein solution prepared in the same buffer. All runs were carried out at 25°C after at least one hour of temperature equilibration. Rotor speed for all experiments was 50,000 rpm. Sedimentation was monitored by UV absorbance. Due to the wide range of concentrations tested, detection at different wavelengths was required. For lower concentration samples (3 µM) absorbance was monitored at 230 and 280 nm. For 10 µM, 250 and 280 nm was used, and for concentrations greater than 10 µM, 250 and 295 nm was detected. For data analysis, plots of size distribution as a function of concentration were obtained from the raw data using the continuous c(s) distribution model in Sedfit (v 13.0b)

45

. Protein

partial specific volume and buffer density and viscosity were calculated using the program Sednterp 46. Accession Numbers: Coordinates and structure factors for the apo form of tmMnBP6 have been deposited in the Protein Data Bank 47 under the accession code 5HM4. RESULTS Solution characterization of tmMnBPs

11 ACS Paragon Plus Environment

Page 13 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Although most PBPs are known to be monomeric, there are an increasing number of reports where the quaternary state of PBPs has been suggested to play a functional role in cellular processes

21-26

. Small-angle neutron scattering (SANS) experiments were undertaken to

characterize the oligomerization state and the ligand induced conformational changes of tmMnBP3 and tmMnBP6, two recently structurally characterized PBPs from T. maritima (Table 1 and Figure 1)

27

. Interestingly, the SANS data of either apo or ligand bound tmMnBP3 or

tmMnBP6 could not be rationalized based on known monomeric crystal structures (Figure 1).

Figure 1. SANS analysis of tmMnBP3 and tmMnBP6. Raw SANS data of tmMnBP3 (A) and tmMnBP6 (B). Open black squares are the experimental data for the apo proteins. Solid black squares are the experimental data for the ligand bound protein. Blue line is the computationally derived scattering data based on the monomeric tmMnBP3 apo protein crystal structure34. Green line is the computationally derived scattering data based on the tmMnBP3 mannobiose (A) or tmMnBP6 mannobiose (B) crystal structure. Red line is the computationally derived scattering data based on the tmMnBP3 apo homodimer40.

The experimentally determined radius of gyration (Rg) of apo and ligand bound tmMnBP3 and tmMnBP6 is significantly higher than the Rg computationally derived from the crystal structures, thus suggesting the single-static structures observed in the crystal do not necessarily match those of the proteins in solution (Table 1 and Figure 1)

34

. Moreover, the

addition of mannobiose to these proteins induces larger than expected decrease in the Rg (Table 1 12 ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 37

and Figure 1), which is not consistent with the characteristic hinge-bending motions found in members of this protein superfamily 48, 49. The PISA server

50

was used to analyze the apo tmMnBP3 crystal structure for any

potential protein:protein interfaces. This analysis suggests an alternate arrangement of the two monomers in the crystallographic asymmetric unit of tmMnBP3 (x,y,z-1 transformation matrix to PDB code 1VR5), which results in a protein:protein interface with a buried surface area of 690 Å2,

in

addition

to

multiple

hydrogen

bonds

(Figure

2).

Figure 2. The tmMnBP homodimer. The overall structure of tmMnBP3 (A) and tmMnBP6 (B) homodimer. The protein interface is circled and the helices comprising the protein interface are indicated. Close up stereo view of the tmMnBP3 (C) and tmMnBP6 (D) protein interface. Each monomer is colored as in panels A and B. Inter-protein hydrogen bonds are represented as black dashed lines.

This homodimeric structure of tmMnBP3 was used to interpret the apo protein SANS data

34

.

This analysis results in improved agreement among the experimentally and 13 ACS Paragon Plus Environment

Page 15 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

computationally derived data (Table 1 and Figure 1), suggesting the apo tmMnBPs are in a homodimeric quaternary state in solution. However, the modelling of the SANS data, collected in the presence of saturating amounts of ligand, with the structure of the ligand bound proteins does not have high agreement to the solution scattering. Through superposition with the crystal structure of the apo tmMnBP3 homodimer, ligand bound homodimer models of tmMnBP3 and tmMnBP6 were made; however, these models have a significantly larger radius of gyration than the experimentally determined Rg values (Table 1) 34. We then tested to see if this discrepancy could be resolved by assuming a mixed population of monomer and dimer using the small-angle scattering analysis program OLIGOMER 35. For the apo proteins, this analysis showed that both tmMnBP3 and tmMnBP6 are essentially all homodimeric. By contrast, the data for the ligand bound proteins is better modeled as a mixture of monomer and homodimer, with monomer being the predominant species, suggesting a previously unidentified ligand induced change in the quaternary state of these PBPs (Table 1). The same trend in ligand induced dissociation is also observed at 65 ˚C in small-angle X-ray scattering experiments of tmMnBP3, which is within the permissible growth range of T. maritima and 15 ˚C away from the optimal growth temperature (Figure S1). Structural characterization of the apo-tmMnBPs Analysis of the SANS data of the apo tmMnBPs suggests that the homodimer present in the crystallographic asymmetric unit of the unpublished crystal structure of tmMnBP3 (Protein Data Bank Code 1VR5) is also present in solution. To see if this homodimer is also present in tmMnBP6, X-ray crystallography was used to solve the structure of the apo tmMnBP6. The structure of the apo form of tmMnBP6 was determined to a resolution of 2.0 Å and refined to an Rcryst/Rfree of 17.3/20.9% (Table 2 and Figure 2). A single molecule is found in the

14 ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 37

crystallographic asymmetric unit; however, a crystallographic symmetry mate is also found forming an essentially identical homodimer to the one found in the tmMnBP3 crystal structure (Figure 2). The tmMnBP3 and tmMnBP6 homodimers superimpose with a Cα r.m.s.d. of 2.5 Å. tmMnBPs protein:protein interface The tmMnBP protein interface resembles a canonical protein interaction site with a nonpolar core surrounded by polar hydrogen bonding amino acids. Four α-helices, two from each monomer (helixA/helixB or helixA’/helixB’), comprise the two-fold symmetric protein interaction site (Figure 2). Residues 230-240 and residues 252-260 comprise helices A and B, respectively in tmMnBP3 and tmMnBP6. The overall features of the tmMnBP3 and tmMnBP6 protein interfaces are very similar. A total of 690 Å2 of buried surface area is found at the tmMnBP3 protein interface, in addition to seven hydrogen bonds, four of which are salt bridges. 740 Å2 of buried surface area is found at the tmMnBP6 interaction site, in addition to eight hydrogen bonds, four of which are salt bridges. All of the amino acids involved in hydrogen bonds are identical among tmMnBP3 and tmMnBP6 (Figure 2). The differential hydrogen bonding observed among the two complexes is due to small differences in the packing of interface α-helices resulting in the loss of the hydrogen bond between the main chain carbonyl of Met239 and the side chain of Asn232 in tmMnBP3 in only one of the helices. Dissociation of the tmMnBP interface Analysis of the SANS data and the crystal structures of the tmMnBPs suggests a ligand dependent allosteric switch is involved in the dissociation of the protein interface. Comparison of the superimposed apo and ligand bound crystal structures allows for identification of the molecular mechanisms that potentially underlie this allosteric switch (Figure 3). In tmMnBP6, Met249 and Asn247 are located on the loop adjacent to the N-terminus of helix B/B’ (Figure 3).

15 ACS Paragon Plus Environment

Page 17 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

In addition to forming Van der Waals interactions with the saccharide, Met249 and Asn247 hydrogen bond to the first sugar ring. To accommodate these molecular interactions, the loop encompassing these residues shifts by 1.8 Å. This ligand induced conformational change has little effect on the C-terminal residues of this loop, however helixB/B’ is also shifted ~2.4 Å. The hydrogen

bond

between

Pro251

and

Tyr260

is

lost

as

a

result

(Figure

3).

Figure 3. The tmMnBP6 allosteric switch. (A) Close-up view of the ligand induced changes in the B/B’ helix. Ligand binding leads to a shifting of the B/B’ helix and dissociation of the hydrogen bonds between Tyr260 and Pro251. Mannobiose bounds structure is shown in green ribbon model with mannobiose colored yellow. The two monomers of the homodimeric tmMnBP6 are colored in magenta and cyan with the mannobiose bound structure superimposed on only the magenta monomer for clarity. (B) Stereo view of the ligand induced changes in the A/A’ helix, colored as in panel A. Hydrogen bonds in the ligand bound form are shown as black dashed lines whereas hydrogen bonds are represented as red dashed lines in the apo protein.

16 ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 37

Additionally, without movement of the B/B’ helix, the carbonyl of Asn231 in the A/A’ helix would clash with the side chain of Leu250 in the B/B’ helix upon binding of ligand. The amino acid side chains found in the B/B’ helix are in essentially identical conformations in both the monomer and homodimer; however, this positional shift in the helix alters the molecular interactions at the protein interface. In tmMnBP6, the A/A’ helix undergoes a more dramatic ligand induced conformational change (Figure 3). Asn231 forms two hydrogen bonds to the first sugar ring. To accommodate these hydrogen bonds, significant structural rearrangements occur in the helixA/A’ between the apo and ligand bound forms of the protein. The hydrogen bonding pattern of this helix in the apo form is that of a canonical alpha helix, with hydrogen bonds between the backbone carbonyl of the i residue to the amide hydrogen of the i+4 residue (Figure 3). To form the hydrogen bonding interactions with the ligand, Asn231 must alter its conformation away from the surface towards the binding pocket. This is accomplished through tightening of part the A/A’ helix so that the hydrogen bonding pattern shifts to that of the higher energy 310 helix (carbonyl of i, hydrogen bonding to the amide hydrogen of the i+3) (Figure 3). A total of 4 residues shift to a 310 helix, encompassing Ser230 to Ala234. The tightening of the four residues of the A/A’ helix significantly alters the protein interface. Asn232 is no longer in a conformation that permits hydrogen bonding to Tyr260 in B/B’ helix, in addition to the hydrogen bond formed with Met239. The loss of these hydrogen bonds may also facilitate the breaking of others between Pro251 and Tyr260 in the B/B’ helix and also the positional shift of this helix. Interestingly, residues of Lys240 and Glu242 in the A/A’ helix are in essentially identical conformation, suggesting little disruption of the salt bridge hydrogen bonds. In light of all the changes at the

17 ACS Paragon Plus Environment

Page 19 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

rest of the protein interface, this may partially account for the residual homodimer observed in the presence of ligand. All but one of the residues postulated to be involved in the transduction of the ligand occupancy of the tmMnBP6 binding pocket are conserved in tmMnBP3. In tmMnBP3, Met249 is replaced with a phenylalanine (Phe249) which subsequently no longer occludes the binding site in the apo form (Figure S2). Accordingly, accompanying motions of the loop encompassing Phe249 are of a smaller magnitude (1.5Å), yet, the net result on the B/B’ helix and hydrogen bonding pattern of Tyr260 is the same.

The ligand-induced tightening of the A/A’ helix also

occurs in a similar manner, although 310 hydrogen bonding occurs only for Ser230 and Asn231, and the water molecules satisfying the carbonyl hydrogen bonds of Asn232 and V233 (Figure S2). The remainder of the tmMnBP3 helix in is a canonical α-helix form as observed in tmMnBP6. Dissection of the tmMnBP allosteric mechanism Two mutant forms of the tmMnBPs were made to dissect the mechanisms underlying the allosteric switch. A M239R mutation was made to have a protein that was an obligate monomer, and a N231A mutant was made to have a protein that was deficient in the allosteric switch. Ligand induced shifts in thermal melting points were used to confirm these proteins still bound ligand, albeit inducing a significantly reduced ∆Tm in tmMnBP6 and abnormal non-two state unfolding in tmMnBP3 (Table 1 and Figure S3).

SANS was used to characterize the

stoichiometric behavior of the mutant proteins in solution (Table 1). For both tmMnBPs, the M239R mutation, which is found at the protein interface, essentially leads to complete dissociation into a monomer regardless of the presence or absence of ligand (Table 1). Mutation of Asn231 to alanine, which is located in the binding site and also a proposed component of the

18 ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 37

allosteric switch, has a significant effect on the interaction behavior of the protein in both the presence and absence of ligand. Both N231A tmMnBP mutants show an increase in the fraction of protein that is found to be monomeric in the absence of ligand (20% and 10% increase in monomer for apo tmMnBP3 and tmMnBP6 respectively). The addition of ligand to the N231A mutants suggests that these mutants, with essentially an equal mixture of monomer and dimer in the presence of ligand, are deficient in the allosteric switch (Table 1). This suggests that while Asn231 plays a significant role in the allosteric switch, it is not the only residue involved. Computer Simulations Molecular dynamics simulations were used to probe ligand mediated changes in dynamics to the protein:protein interface. Two states of the proteins were used for simulations, one consisted of the crystallographic apo homodimer (ApoDimer), the other was a mixed heterodimer model generated through superposition at the protein:protein interface, where one monomer was apo and the other was ligand bound (MixHetDimer). The MixHetDimer model is utilized as a means to uncover the communication of ligand occupancy of one ligand bound subunit to the other apo subunit, which is potentially a relevant biological state under nutrient limiting environments that could only be studies in-silico. In the tmMnBP3 ApoDimer, the A/A’ helices are rigid, whereas the B/B’ helices are more dynamic (Figure 4).

There is apparent

symmetry in the dynamics of the interface. The same trends are also captured in the B-factors of the crystal structure of the apo tmMnBP3 (Figure 4 and Figure S5). The thickness of the tubes in this figure corresponds to the aggregate flexibility computed using top 10 quasi-harmonic modes (see methods section for more details). Interestingly, in the tmMnBP3 MixHetDimer model simulations the dynamics of the B/B’ helices are reduced compared to the ApoDimer, and the apparent symmetry in dynamics of this interface is still conserved (Figure 4). This is unexpected

19 ACS Paragon Plus Environment

Page 21 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

as you would expect asymmetry, which can be observed in the B-factors of the MixHetDimer model (Figure 4 and Figure S5). These results suggest that ligand binding in one half site is communicated across the protein interface, and in-turn alters the dynamics the entire interface to mimic the ligand bound state, even though ligand is present in only one monomer.

20 ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 37

Figure 4. Dynamics of the tmMnBP3 protein interface. (A) Overall structure of the tmMnBP3 ApoDimer, where r.ms.f. from molecular dynamics simulations is represented by the width of the backbone. Interface helices are indicated. (B) Close-up view of panel A. (C) Overall structure of the tmMnBP3 ApoDimer, where atomic B-factors are represented by the width of the backbone. (D) Close-up view of panel C. (E) Overall structure of the tmMnBP3 MixHetDimer, where r.ms.f. from molecular dynamics simulations is represented by the width of the backbone. (F) Close-up view of panel E. (G) Overall structure of the tmMnBP3 MixHetDimer, where atomic B-factors are represented by the width of the backbone. (H) Closeup view of panel G. 21 ACS Paragon Plus Environment

Page 23 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

A similar trend in B-factors and dynamics is observed when in the tmMnBP6 ApoDimer structure and MixHetDimer model; however, the dynamical behavior of the B/B’ helices in the tmMNBP6 ApoDimer is of a lower magnitude than in the tmMnBP3 ApoDimer (Figure S4 and Figure S5). The symmetry of the tmMnBP6 MixHetDimer model is also conserved as in the tmMnBP3 model. Detailed analysis of the computer simulations allows identification of correlated residues and, therefore, a network of dynamically coupled residues between the ligand and the dimer interface. Such networks have previously shown to influence enzyme function through allosteric effects

44

. The network shown is shown in Figure S6, is based on pairs and groups of residues

that show large correlated motions between the ligand binding site and the dimer interface. Residues pairs that show large positive correlations show highly coordinated movements over the course of MD simulations, while the residues that show large negative correlations are anticorrelated (correlated but opposite phase) corresponding to movements towards and away from each other (Figure S6). Several pairs of residues that are correlated with each other form a group of correlated residues. It can be hypothesized that through correlated and anti-correlated motions functional changes are transmitted through this network.

It is observed that in both apo

tmMnBPs, each half site of the protein interface is also correlated to the other (Figure S6). In both MixHetDimer models, the half sites are also correlated to one another even though there is asymmetry in the ligand occupancy (Figure S6). There also is an increase in the anti-correlation motions of the MixHetDimer model, compared to the apo proteins. Large scale anti-correlated motions, such as the one indicated by the mixed apo/mannobiose system are indicative of large domains moving towards each other. Therefore, it is possible that the presence of ligand enhances the movement of monomers towards each other, mediated by hinge bending motions,

22 ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 37

in what may be a potentially relevant state of the protein in the transport process. This further suggests ligand binding to one monomer is communicated not only through structure, but also dynamics. Analytical ultra-centrifugation Sedimentation velocity analytical ultracentrifugation (AUC) experiments with tmMnBP6 were performed to gain further insight into the ligand induced allosteric dissociation. Size distribution analysis of the AUC data indicates three species of tmMnBP6 are present in solution (Figure S7). The protein exists primarily as a homodimer in apo form. Integration of the peaks in the distribution plot shows that at 10 µM, approximately 80% of the apo protein is dimeric with a weight average sedimentation coefficient of 7.1S for the major peak. The large fraction of dimeric protein at 10 µM suggests that the homodimer dissociation constant is less than 10 µM and lower than the typical periplasmic concentration of PBPs

51

. Upon addition of mannobiose,

tmMnBP6 becomes primarily monomeric. For the 10 µM sample with ligand bound, the major peak shifts to 4.9S. This peak corresponds to 63% of the loaded protein with 19% of the remaining signal corresponding to dimer. In both cases a significant amount (15-20%) of higher molecular weight species is present. This larger species corresponds to a molecular weight of 240,000 at a frictional coefficient (f/f0) of 1.2, indicating the presence of tetrameric tmMnBP6. The possibility that the larger species is a trimer with a compact structure (f/f0 approaching 1.0) cannot be ruled out using this data, but is less likely as it would require an apo monomeric state to form a trimer. Curiously, the fraction of larger molecular weight species changes little as a function of either protein concentration or ligand presence (Figure S7). In agreement with the SANS data, the tmMnBP6 M239R mutation results in a monomeric protein even in the absence of mannobiose. The c(s) plots show a single, major peak

23 ACS Paragon Plus Environment

Page 25 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

at 4.9S at all concentrations tested (Figure S7). In addition, this mutation also prevented tetramer formation, indicating that dimer formation is a prerequisite to the association of the larger species observed at greater than 10S in wild type enzyme. AUC data also confirms the importance of Asn231 in the allosteric switch. A comparison of ligand-bound wild type at 10 µM shows that mannobiose induced dimer dissociation is less efficient in the mutant enzyme (Figure S7). Data for the N231A mutant at 10 µM shows a broadened peak centered at 5.4S. This shift and broadening indicates that monomer-dimer interconversion is occurring during the course of the experiment. This contrasts with the location and shape of the monomer peak observed in both wild type and M239R tmMnBP6 (Figure S7). Due to the multi-state nature of the mutant tmMnBP3 proteins, AUC was not carried out. Conservation of the protein interface tmMnBP3 and tmMnBP6 are postulated to be functionally diverged gene duplicates. Overall these proteins share 61 % amino acid identity and 74 % similarity, however the carbohydrate binding site diverged to an amino acid identity of 48 %

27

. Interestingly, the

residues comprising the protein interaction site retained an amino acid identity of 80 % and a similarity of 95 % suggesting this interaction site is under a selective pressure that is different from that of the binding site and the remainder of the protein (Figure 5). The T. maritima genome contains a total of ten other proteins that share the same oligopeptide binding protein PBP fold as the two tmMnBPs

52

. The closest sequence (30 % identity

and 50 % similarity) and structural homolog to the tmMnBPs is the T. maritima cellobiose binding protein (tmCBP)

26

. The sequence identity of the analogous residues in tmCBP to the

24 ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 37

tmMnBPs protein interaction site is only 15 %, whereas the similarity is 40% (Figure 5).

Figure 5. Conservation of the allosteric switch. (A) Sequence alignment of the tmMnBP3, tmMnBP6 and tmCBP residues that comprise the protein interface (blue text). Residues involved in interprotein hydrogen bonds are underlined and the binding site Asn231 is in red text. (B) Re-analysis of the apo tmCBP SANS data as a mixture of monomer and dimer. The raw scattering data is shown as solid black squares. Blue line is the calculated SANS curve based on a model of an apo monomeric tmCBP, red line is the calculated SANS curve based on a model of an apo dimeric tmCBP. The green line is a modelled as a mixture of apo monomer and homodimer.

Interestingly, although the apparent conservation of the interaction site is not high, the residues involved in the hydrogen bonding interactions across the protein interface are, as is the binding site asparagine (Asn217) that is involved in tightening of the A/A’ interface helix in the tmMnBP proteins (Figure 5). Seeing as though the A/A’ helix in the ligand bound tmCBP crystal structure has an alpha helical hydrogen bonding pattern, ligand binding to it may stimulate relocation of the entire helix rather than tightening of it as observed in the tmMnBPs.

25 ACS Paragon Plus Environment

Page 27 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Many of the other components of the switch are present in tmCBP suggesting it also undergoes a dimer to monomer transition. Recently, tmCBP was analyzed with SANS in both the absence and presence of its cognate ligand cellobiose

26

. Whereas the SANS data in the presence of ligand could be

interpreted in terms of the known ligand bound crystal structure, the apo data had a much larger Rg than would be expected for a monomeric protein, and it was suggested that higher order tmCBP oligomers existed. Indeed, a 7.1 Å decrease in the Rg was observed upon addition of ligand, which is similar to the Rg change observed in the tmMnBPs and larger than expected for monomeric PBPs 48, 49. To see if, similar to the tmMnBPs, apo tmCBP is also in a homodimeric quaternary state, a tmCBP model was generated to analyze the SANS data by superposing the domains from the laminaribiose bound crystal structure of tmCBP with those of the tmMnBP6 homodimer 26. The apo tmCBP SANS data is not well modelled using this homodimer, whereas modelling as a mixture of 45% monomer and 55% dimer significantly improves the fit to the experimental data (Figure 5) 34. The reanalysis of the tmCBP SANS data suggests this allosteric switch is found in other ABC transport systems as well. DISCUSSION Understanding of what occurs in ABC transport, subsequent to the interaction of the PBP with the transmembrane transporters, is thought to be a solved problem. In this process, PBPs serve strictly as a specificity-determining subunit of the transport process by delivering cognate ligands to their respective transmembrane transporters. Central to this is the ligand-induced conformational change that serves as a means to discriminate between the apo and ligand bound protein by the transport systems and other cellular systems where the PBP motif functions. Through the PBP hinge bending motion, ligand binding allosterically regulates the protein

26 ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 37

interaction with the ABC transporters in addition to minimizing ATP hydrolysis that is uncoupled from ligand transport. There is, however, an increasing mass of evidence that this paradigm may be more complex, with PBPs playing a more active role in the regulation of transport. Indeed, there is a growing number of PBPs that are not found in a monomeric quaternary state, with the higher order oligomers proposed to play a unique functional role in the transport process 21-26. The dual coupling of ligand binding to both the hinge bending motion and homodimer dissociation represents an efficient strategy to further differentiate the molecular envelope of the apo and ligand bound states of the PBP. Although the current works lacks the contextualization of the effect of the PBP homodimer on ABC transport, the data presented suggests there being an important-conserved biological role for this additional ligand induced allostery. Moreover, an alternate model for ABC transport has proposed a functional role PBP dimers in the transport process53.

Regardless of the exact biological role, interpretation of the importance of this

dimerization is aided by the 80% identity of this switch and interface with the tmMnBP gene duplicates, while the remainder of the protein shares only 61% amino acids similarity. This is in stark contrast to the binding sites of the two tmMnBPs, which have neofunctionalized from one another, and share only 48% identity. This supports there being a strong selective pressure to conserve this protein interaction and the underlying allosteric switch. Moreover, this allostery is conserved in the tmMnBP homolog, tmCBP. Although the amino acid sequence of tmCBP has significantly diverged from the tmMnBPs, key residues underlying the formation of the protein interaction and allosteric switch are conserved. Interestingly, the tmMnBPs and tmCBP have significant overlap in their substrate specificities. Although the tmMnBPs are annotated as beta(1,4) mannose binding proteins, they both bind tightest to beta(1,4) glucose ligands, which

27 ACS Paragon Plus Environment

Page 29 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

are also the substrates of tmCBP. As glucose is preferentially metabolized, it is possible that under conditions of saturating beta(1,4) glucose that the tmMnBPs and tmCBP can use identical transporters and therefore have conserved this allosteric switch. As most bacteria genomes encode an extensive network of PBPs it remains to be determined the prevalence of this allostery in other ABC transport operons and the underlying biological role. Furthermore, measuring transporter ATP hydrolysis rates with wild-type and obligate monomers/homodimer PBPs will be required to fully understand the functional biological role of this conserved allostery in ABC transport. ACKNOWLEDGMENTS The Oak Ridge National Laboratory Center for Structural Molecular Biology (FWP ERKP291) is supported by the Office of Biological and Environmental Research of the US Department of Energy. Research at the Spallation Neutron Source of Oak Ridge National Laboratory was sponsored by the Scientific User Facilities Division, Office of Basic Energy Sciences, US Department of Energy. The authors would like to thank X. Lu for help with crystallographic data. PKA was supported by a grant from NIH (GM105978). Author Contributions: The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Protein crystallization and small-angle neutron scattering were carried out by S.G. and C.B.S. Structure analysis was carried out by L.L. Analytical ultracentrifugation experiments were carried out by E.W. Molecular simulations were carried out by P.K.A. The project was coordinated by D.A.M and M.J.C. All authors reviewed the results and approved the final version of the manuscript. ASSOCIATED CONTENT

28 ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 30 of 37

Supporting Information: SAXS at 65 degrees, the tmMnBP3 allosteric switch, ligand binding to wild-type and mutant tmMnBP proteins, dynamics of the tmMnBP6 protein interface, dynamics and B-factors of the tmMnBP proteins, dynamically coupled residues in tmMnBPs and analytical ultracentrifugation of tmMnBP6. REFERENCES (1) Kadaba, N. S., Kaiser, J. T., Johnson, E., Lee, A., and Rees, D. C. (2008) The high-affinity E. coli methionine ABC transporter: structure and allosteric regulation, Science 321, 250-253. (2) Yonetani, T., and Laberge, M. (2008) Protein dynamics explain the allosteric behaviors of hemoglobin, BBA-Prot. and Proteo. 1784, 1146-1158. (3) van Dijk, W. J., Klaassen, R. V., Schuurmans, M., van der Oost, J., Smit, A. B., and Sixma, T. K. (2001) Crystal structure of an ACh-binding protein reveals the ligand-binding domain of nicotinic receptors, Nature 411, 269-276. (4) Fenton, A. W. (2008) Allostery: an illustrated definition for the ‘second secret of life’, Trends Biochem. Sci. 33, 420-425. (5) Gardina, P., Conway, C., Kossman, M., and Manson, M. (1992) Aspartate and maltose-binding protein interact with adjacent sites in the Tar chemotactic signal transducer of Escherichia coli, J. Bacteriol. 174, 1528-1536. (6) Björkman, A. J., and Mowbray, S. L. (1998) Multiple open forms of ribose-binding protein trace the path of its conformational change, J. Mol. Biol. 279, 651-664. (7) Sharff, A. J., Rodseth, L. E., Spurlino, J. C., and Quiocho, F. A. (1992) Crystallographic evidence of a large ligand-induced hinge-twist motion between the two domains of the maltodextrin binding protein involved in active transport and chemotaxis, Biochemistry 31, 10657-10663. (8) Cuneo, M. J., Beese, L. S., and Hellinga, H. W. (2008) Ligand-induced conformational changes in a thermophilic ribose-binding protein, BMC Struct. Biol. 8, 50. (9) Petronilli, V., and Ames, G. (1991) Binding protein-independent histidine permease mutants. Uncoupling of ATP hydrolysis from transmembrane signaling, J. Biol. Chem. 266, 16293-16296. (10) Locher, K. P. (2009) Structure and mechanism of ATP-binding cassette transporters, Phil. Trans. of the Royal Society of London B: Biological Sciences 364, 239-245. (11) Oldham, M. L., Khare, D., Quiocho, F. A., Davidson, A. L., and Chen, J. (2007) Crystal structure of a catalytic intermediate of the maltose transporter, Nature 450, 515-521. (12) Oldham, M. L., and Chen, J. (2011) Crystal structure of the maltose transporter in a pretranslocation intermediate state, Science 332, 1202-1205. (13) Alvarez, F. J. D., Orelle, C., Huang, Y., Bajaj, R., Everly, R. M., Klug, C. S., and Davidson, A. L. (2015) Full engagement of liganded maltose‐binding protein stabilizes a semi‐open ATP‐ binding cassette dimer in the maltose transporter, Mol. Microbiol. 98, 878-894. (14) Hollenstein, K., Frei, D. C., and Locher, K. P. (2007) Structure of an ABC transporter in complex with its binding protein, Nature 446, 213-216. (15) Borths, E. L., Locher, K. P., Lee, A. T., and Rees, D. C. (2002) The structure of Escherichia coli BtuF and binding to its cognate ATP binding cassette transporter, Proc. Natl. Acad. Sci. 99, 16642-16647. (16) Davidson, A. L., Shuman, H. A., and Nikaido, H. (1992) Mechanism of maltose transport in Escherichia coli: transmembrane signaling by periplasmic binding proteins, Proc. Natl. Acad. Sci. 89, 2360-2364. 29 ACS Paragon Plus Environment

Page 31 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

(17) Ames, G. F.-L., Liu, C. E., Joshi, A. K., and Nikaido, K. (1996) Liganded and unliganded receptors interact with equal affinity with the membrane complex of periplasmic permeases, a subfamily of traffic ATPases, J. Biol. Chem. 271, 14264-14270. (18) Gould, A. D., Telmer, P. G., and Shilton, B. H. (2009) Stimulation of the maltose transporter ATPase by unliganded maltose binding protein, Biochemistry 48, 8051-8061. (19) Yang, J. G., and Rees, D. C. (2015) The allosteric regulatory mechanism of the Escherichia coli MetNI methionine ATP Binding Cassette (ABC) transporter, J. Biol. Chem. 290, 9135-9140. (20) Gorbulev, S., Abele, R., and Tampé, R. (2001) Allosteric crosstalk between peptide-binding, transport, and ATP hydrolysis of the ABC transporter TAP, Proc. Natl. Acad. Sci. 98, 3732-3737. (21) Cuneo, M. J., Changela, A., Miklos, A. E., Beese, L. S., Krueger, J. K., and Hellinga, H. W. (2008) Structural analysis of a periplasmic binding protein in the tripartite ATP-independent transporter family reveals a tetrameric assembly that may have a role in ligand transport, J. Biol. Chem. 283, 32812-32820. (22) Gonin, S., Arnoux, P., Pierru, B., Lavergne, J., Alonso, B., Sabaty, M., and Pignol, D. (2007) Crystal structures of an Extracytoplasmic Solute Receptor from a TRAP transporter in its open and closed forms reveal a helix-swapped dimer requiring a cation for α-keto acid binding, BMC Struct. Biol. 7, 1. (23) Luchansky, M. S., Der, B. S., D’Auria, S., Pocsfalvi, G., Iozzino, L., Marasco, D., and Dattelbaum, J. D. (2009) Amino acid transport in thermophiles: characterization of an arginine-binding protein in Thermotoga maritima, Mol. Biosys. 6, 142-151. (24) van der Heide, T., and Poolman, B. (2002) ABC transporters: one, two or four extracytoplasmic substrate‐binding sites?, EMBO reports 3, 938-943. (25) Shi, R., Proteau, A., Wagner, J., Cui, Q., Purisima, E. O., Matte, A., and Cygler, M. (2009) Trapping open and closed forms of FitE—A group III periplasmic binding protein, Proteins. 75, 598-609. (26) Munshi, P., Stanley, C. B., Ghimire-Rijal, S., Lu, X., Myles, D. A., and Cuneo, M. J. (2013) Molecular details of ligand selectivity determinants in a promiscuous β-glucan periplasmic binding protein, BMC Struct. Biol. 13, 18. (27) Ghimire-Rijal, S., Lu, X., Myles, D. A., and Cuneo, M. J. (2014) Duplication of genes in an ATPbinding cassette transport system increases dynamic range while maintaining ligand specificity, J. Biol. Chem. 289, 30090-30100. (28) Cohen, D. S., and Pielak, G. J. (1994) Stability of yeast iso-1-ferricytochrome c as a function of pH and temperature, Protein Sci. 3, 1253-1260. (29) Niklasson, M., Andresen, C., Helander, S., Roth, M. G., Zimdahl Kahlin, A., Lindqvist Appell, M., Mårtensson, L. G., and Lundström, P. (2015) Robust and convenient analysis of protein thermal and chemical stability, Protein Sci. 24, 2055-2062. (30) Zhao, J. K., Gao, C. Y., and Liu, D. (2010) The extended Q-range small-angle neutron scattering diffractometer at the SNS, J. Appl. Cryst. 43, 1068-1077. (31) Wignall, G. D., and Bates, F. S. (1987) Absolute calibration of small-angle neutron scattering data, J. Appl. Cryst. 20, 28-40. (32) Guinier, A., and Fournet, G. (1955) Small-angle scattering of X-rays, Wiley, New York. (33) Svergun, D. I. (1992) Determination of the regularization parameter in indirect-transform methods using perceptual criteria, J. Appl. Cryst. 25, 495-503. (34) Svergun, D., Barberato, C., and Koch, M. (1995) CRYSOL–a program to evaluate X-ray solution scattering of biological macromolecules from atomic coordinates, J. Appl. Crystallogr. 28, 768773. (35) Konarev, P. V., Volkov, V. V., Sokolova, A. V., Koch, M. H., and Svergun, D. I. (2003) PRIMUS: a Windows PC-based system for small-angle scattering data analysis, J. Appl. Crystallogr. 36, 1277-1282. (36) Otwinowski, Z. a. M., W. (1997) Processing of X-ray diffaction data collected in oscillation mode., Methods Enzymol. 276A, 307-326.

30 ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 32 of 37

(37) Collaborative Computational Project, N. (1994) The CCP4 suite: programs for protein crystallography, Acta Crystallogr. D Biol. Crystallogr. 50, 760-763. (38) Emsley, P., and Cowtan, K. (2004) Coot: model-building tools for molecular graphics, Acta Crystallogr D Biol Crystallogr 60, 2126-2132. (39) Adams, P. D., Afonine, P. V., Bunkoczi, G., Chen, V. B., Davis, I. W., Echols, N., Headd, J. J., Hung, L. W., Kapral, G. J., Grosse-Kunstleve, R. W., McCoy, A. J., Moriarty, N. W., Oeffner, R., Read, R. J., Richardson, D. C., Richardson, J. S., Terwilliger, T. C., and Zwart, P. H. (2010) PHENIX: a comprehensive Python-based system for macromolecular structure solution, Acta Crystallogr. D Biol. Crystallogr. 66, 213-221. (40) Davis, I. W., Murray, L. W., Richardson, J. S., and Richardson, D. C. (2004) MOLPROBITY: structure validation and all-atom contact analysis for nucleic acids and their complexes, Nucleic Acids Res. 32, W615-619. (41) Case, D. A., Cheatham, T. E., Darden, T., Gohlke, H., Luo, R., Merz, K. M., Onufriev, A., Simmerling, C., Wang, B., and Woods, R. J. (2005) The Amber biomolecular simulation programs. J. Comp. Chem. 26, 1668-1688. (42) Agarwal, P. K. (2004) Computational Studies of the mechanism of cis/trans isomerization in HIV-1 catalyzed by cyclophilin A, Proteins. 56, 449-463. (43) Ramanathan, A., and Agarwal, P. K. (2011) Evolutionarily Conserved Linkage between Enzyme Fold, Flexibility, and Catalysis, PLoS Biology 9, e1001193. (44) Ramanathan, A., and Agarwal, P. K. (2011) Evolutionarily conserved linkage between enzyme fold, flexibility, and catalysis, PLoS Biol. 9, e1001193. (45) Schuck, P. (2000) Size-distribution analysis of macromolecules by sedimentation velocity ultracentrifugation and lamm equation modeling, Biophys. J. 78, 1606-1619. (46) Laue, T. M., Shah, B. D., Ridgeway, T. M., and Pelletier, S. L. (1992) Computer-aided interpretation of analytical sedimentation data for proteins, In Analytical Ultracentrifugation in Biochemistry and Polymer Science (Harding, S. E., Rowe, A. J., and Horton, J. C., Eds.), pp 90-125, Royal Society of Chemistry, Cambridge, U.K. (47) Berman, H. M., Westbrook, J., Feng, Z., Gilliland, G., Bhat, T. N., Weissig, H., Shindyalov, I. N., and Bourne, P. E. (2000) The Protein Data Bank, Nucleic Acids Res. 28, 235-242. (48) Shilton, B. H., Flocco, M. M., Nilsson, M., and Mowbray, S. L. (1996) Conformational changes of three periplasmic receptors for bacterial chemotaxis and transport: the maltose-, glucose/galactose-and ribose-binding proteins, J. Mol. Biol. 264, 350-363. (49) Olah, G., Trakhanov, S., Trewhella, J., and Quiocho, F. (1993) Leucine/isoleucine/valine-binding protein contracts upon binding of ligand, J. Biol. Chem. 268, 16241-16247. (50) Krissinel, E., and Henrick, K. (2007) Inference of macromolecular assemblies from crystalline state, J. Mol. Biol. 372, 774-797. (51) Manson, M. D., Boos, W., Bassford, P., and Rasmussen, B. (1985) Dependence of maltose transport and chemotaxis on the amount of maltose-binding protein, J. Biol. Chem. 260, 9727-9733. (52) Nelson, K. E., Clayton, R. A., Gill, S. R., Gwinn, M. L., Dodson, R. J., Haft, D. H., Hickey, E. K., Peterson, J. D., Nelson, W. C., Ketchum, K. A., McDonald, L., Utterback, T. R., Malek, J. A., Linher, K. D., Garrett, M. M., Stewart, A. M., Cotton, M. D., Pratt, M. S., Phillips, C. A., Richardson, D., Heidelberg, J., Sutton, G. G., Fleischmann, R. D., Eisen, J. A., White, O., Salzberg, S. L., Smith, H. O., Venter, J. C., and Fraser, C. M. (1999) Evidence for lateral gene transfer between Archaea and bacteria from genome sequence of Thermotoga maritima, Nature 399, 323-329. (53) Jones, P. M., and George, A. M. (2014) A reciprocating twin-channel model for ABC transporters, Q. Rev. Biophys. 47, 189-220.

31 ACS Paragon Plus Environment

Page 33 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

A

Biochemistry

B

ACS Paragon Plus Environment

A 1 2 3 4 5 6 7 8 9 10 11 12 C 13 14 M237 15 16 A 17 18 L250 19 20 M239 21 22 23 24 B 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

B Biochemistry

Page 34 of 37

A’ A

A’ A

B’ B

K240

E242

B’

E242

K240

M237 M237 V233 A’

L250 L250 M239

N232 Y260

N231 P251 B’

V255

M237 A

K240

D

L250

N232 Y260

V233 A’

L250 M239

N231

M239 N231

B’

P251 B

V255

B’ I255

ACS Paragon Plus Environment

V233 A’

L250

N232 Y260

M237 A

P251

B

E242

M237 M237 A

A’

K240

E242

M237

V233

B

N232 Y260

N231

P251 B

B’ I255

Page 35 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

A

Y260 N247 P251 Y260

P251

M249

B

236

236 235

235

233

233

232

232

231

231

230

230 T388

G387

ACS Paragon Plus Environment

T388

G387

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

A

Page 36 of 37

B

A’

A A

A’

B

B’

B

C

B’

D

A’

A A

A’

B

B’

B

E

B’

F

A’

A A

A’

B

B’

B

G

B’

H

A’

B’

A A

A’

B

B ACS Paragon Plus Environment

B’

Page 37 of 37 1 A 2 tmMnBP3 3 tmMnBP6 4 tmCBP 5 6 B 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

LSNNVAVGMLMKGELDWSNFFLPGVP-VLKKAYG 261 LSNNVALGMLMKGELDFSNFMLPGVP-ILKKVYN 261 KDNPSASLAFERGDIDWNGLFIPSVWELWEKKGL 247

ACS Paragon Plus Environment