3% (INHIHIHIHIHIIIHIHII‘IIITIIPIIHIINIHHWWWI

 

1007

This is to certify that the
thesis entitled

Biophysical and Structural Studies of HrpA, a Type Ill
Secretion Pilin from Gamma-proteobacterial Plant Pathogens

presented by

Jason M. Criscione

has been accepted towards fulfillment
of the requirements for the

MS. degree in Chemistry

 

 

 

200

te‘

MSU is an Affirmative Action/Equal Opportunity Institution

 

 

 

LIBRARY
Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 p:/C|RCIDaieDue.indd-p.l

BIOPHYSICAL AND STRUCTURAL STUDIES OF HRPA, A TYPE III
SECRETION PILIN FROM GAMMA-PROTEOBACTERIAL
PLANT PATHOGENS
By

Jason M. Criscione

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE
Department of Chemistry

2006

ABSTRACT

BIOPHYSICAL AND STRUCTURAL STUDIES OF HRPA, A TYPE III SECRETION
PILIN F ROM GAMMA-PROTEOBACTERIAL PLANT PATHOGENS

By
Jason M. Criscione
The type III secretion system plays a vital role for the virulence of Gram—negative
bacteria [8]. Gamma-proteobacterial plant pathogens utilize type III secretion to
transport cytosolic effector proteins across the cell wall of its target host cell [24]. This
essential cytosol-to-cytosol interaction is accomplished through the assembly of the type
III secretion apparatus (TTSA). The TTSA in proteobacterial plant pathogens is
comprised of a membrane-embedded base and a hollow extracellular filament termed the
Hrp pilus [16]. The base features a complex basal body that spans the periplasm and is
associated with both the inner and outer membranes (of the Gram-negative bacteria) [28].
The Hrp pilus elongates through addition of protein monomers (pilins) in helical repeat at
its tip, to invade the plant host cell and serve as a conduit for transport of effector proteins
[24]. The pilus is considered an attractive pharmaceutical target, both because it is

extracellular and because its elimination renders the bacteria non-virulent.

This work focuses on determining the structure and biophysical properties of two
unrelated HrpA pilins from proteobacterial plant pathogens, Erwinia amylovora and

Pseudomonas syringae.

ACKNOWLEDGMENTS

I would like to express my gratitude to my research advisor, Dr. William Wedemeyer.
As a research mentor, scientist, penseur, and friend, I hold you in the highest regard. You
are a person of true intellect, passion, and wisdom. You have served as a beacon to guide
me through my seemingly ever-winding journey through the world of science. I have
taken a great deal away from this relationship, and it has been my pleasure to work with

you.

I would like to thank my thesis committee members, Dr. Michael Feig and Dr. John

McCracken, for their support and a challenging final examination.

I would like to thank the Wedemeyer laboratory for their support. Special thanks go to
Terry Ball, Russell LaClair, and Joel Lwande. Terry provided me with essential
instruction on several biochemical techniques and methods. He also performed some of
the preliminary TEM and analytical ultracentrifugation experimentation. Russell
contributed to the project by collecting some of the preliminary CD and HN-HSQC data.
Joel provided me with information from his experience with several molecular biology

techniques. He also collected the initial data for thermal melt analysis.

I would like to thank several laboratories that facilitated the experiments performed

within this work. I am grateful to Dr. Babak Borhan and Marina Tanasova for the use of

their CD instrument. The analytical ultracentrifugation data were collected at the MSU

iii

Macromolecular Structure, Sequencing and Synthesis Facility, with the assistance of Dr.
Joseph Leykam. The TEM images were taken at the MSU Center for Advanced
Microscopy, with the assistance of Dr. Alicia Pastor-Lecha. The protein crystallization
was a collaborative effort with Dr. Jennifer Ekstrom. We would like to thank Dr. R. M.
Garavito and Dr. Nicole Webb for the use of their crystal screening robot. The solid-state
NMR spectra were obtained during my tenure in Dr. David Weliky’s laboratory. I would
like to thank both Dr. Weliky for his guidance and expertise and the senior member of the
group, Dr. Michele Bodner, for her instruction on experimental solid-state NMR

techniques and for collecting some of the Spectra displayed in this work.

iv

TABLE OF CONTENTS

LIST OF TABLES ............................................................................... vii
LIST OF FIGURES .............................................................................. viii
I. Introduction to Type III Secretion and the Role of the Hrp Pilus ........................ 1
LA. Probing the Structure and Function of HrpA from P. syringae and
E. amylovora ............................................................................. 4
II. Computational Modeling of HrpA ............................................................. 5
ILA. Multiple Sequence Alignment (MSA) ............................................... 5
ILB. Secondary-structure Predictions ....................................................... 7
II.C. Three-dimensional Modeling of the Hrp Pilus ...................................... 8
111. Protein Preparation ........................................................................... 12
III.A. Plasmid Constructs ....................................................................... 12
III.A.i. Truncated, Wild-type HrpA from P. syringae pv.
tomato [DC3000] ............................................................... 12
III.A.ii. Full-length, Wild-type HrpA from E. amylovora .......................... 13
“LB. Expression of Wild-type HrpA Pilins .............................................. 13
[II.C. Expression of the Truncated, Wild-type 15N-labeled HrpA from
P. syringae ............................................................................... 14
III.D. Expression of Single-Residue-Labeled HrpA from P. syringae ................ 15
“LE. Purification of HrpA Pilins .......................................................... 15
III.F. Cleavage of the N-terminal (His)6 Tag on HrpA from P. syringae ............ 21
IV. In Vitro Preparation of the Hrp Pilus ....................................................... 26
IV.A. Two Forms of HrpA: Monomer and Multimer ................................... 26
IV.A.i. 2,2,2-trifluoroethanol (TFE) ................................................... 27

IV.A.ii. Salts .............................................................................. 29

IV.A.iii. Stable Multimers of Uniform Dimensions ................................. 33
IV.A.iii.a. Adenosine Diphosphate (ADP) ....................................... 33
IV.A.iii.b. Dickerson-Drew Dodecamer of DNA ............................... 34

V. Secondary and Tertiary Structure of Monomeric and Multimeric HrpA ............... 37
VA. Analytical Ultracentn'fugation ........................................................ 38
V.B. Size Exclusion Chromatography ...................................................... 43
V.C. Solution-state 2D-NMR and HN-Heteronuclear Single Quantum

Correlation (HN-HSQC) .............................................................. 44
VD. Thermal Folding Transitions by UV Spectroscopy ................................ 46
V.E. Circular Dichroism (CD) .............................................................. 51
V.F. Solid-state NMR (SSNMR) ........................................................... 59

VI. Conclusions ................................................................................... 72

VII. Works in Progress .......................................................................... 74
VILA. Unfolding/Dissociation Kinetics of Multimeric HrpA from

P. syringae ............................................................................ 74

VILB. Preparation, Expression, Purification, and Fluorescence Microscopy of an
N-terminal GFP-tagged HrpA Construct ......................................... 76
VILC. Proline Mutations ..................................................................... 77

VII.D. Atomic-resolution Structural Studies: X-ray Crystallography and

Electron Fiber Diffraction ........................................................... 81
VII.D.i. Crystallization of Monomeric HrpA and X-ray Diffraction .............. 81
VILE. Fiber Diffraction Studies of Multimeric HrpA ................................... 83
Appendix. Buffers and Stock Solutions ........................................................ 87
Bibliography ....................................................................................... 88

vi

LIST OF TABLES

Table 3.1 Gradient elution scheme from 0% to 100% NiB buffer ........................ 16
Table 3.2 Composition of % NiB in the eluted fractions ................................... 17
Table 3.3 SDS-PAGE sample preparation .................................................... 17
Table 3.4 Size exclusion gel filtration system parameters .................................. 19
Table 5.1 JASCO J-810 acquisition parameters .............................................. 52

vii

LIST OF FIGURES

Figure 1.1 Model of the TTSA of a proteobacterial plant pathogen where (A) is the Hrp
pilus, (B) is the membrane-embedded base, (C) is the plant host cell, (D) is the
extracellular space, (E) is the outer bacterial membrane, (F) is the peptidoglycan layer,
(G) is the periplasmic space, (H) is the inner bacterial membrane, (I) is the bacterial
cytosol, (J) is the translocator protein, HrpA, chaperone complex, (K) is the effector
protein, HrpZ, chaperone complex, (L) is the unstructured HrpA monomer, and (M) is the
folded HrpA monomer associating with the tip of the growing pilus ........................ 2

Figure 2.1 Multiple Sequence Alignment (MSA) of HrpA from Pseudomonas syringae
pv. maculicola (PSPVM), Pseudomonas syringae pv. tomato (PSPVTOM), Pseudomonas
syringae pv. actinidiae (PSPVA), Pseudomonas syringae pv. delphinii (PSPVD),
Pseudomonas syringae pv. magnoliae (PSPVM_1), Pseudomonas syringae pv. theae
(PSPVTH), Pseudomonas syringae pv. tagetis (PSPVTA), Pseudomonas syringae
(HRPA_PS), Pseudomonas syringae pv. oryzae (PSPVO), Erwinia pyrifoliae
(ERW_PYR), Erwinia amylovora (ERW_AM and HRPA_ERWAM), Pantoea stewartii
subsp. stewartii (PSSST), Pectobacterium carotovorum subsp. carotovorum (ERW_PC).
Highly conserved, moderately conserved, and unconserved residues are represented by an
outlined grey background with a white letter, an outlined white background with a
boldface black letter, and a colorless background with a black letter, respectively ........ 6

Figure 2.2 Predicted secondary structure of HrpA from P. syringae using SAM-TZK,
where H and C represent Helix and Coil, respectively, and the intensity relates to the
probability ............................................................................................ 7

Figure 2.3 Model for the self-assembly of the Hrp pilus ..................................... 9

Figure 2.4 For a a-helix, contributions fi'om polarized amide groups along the peptide
backbone generate a net dipole moment (P) which points from the C-terminus to the N-
terminus, neglecting side chain contribution. The direction of the induced electric field
(E) is depicted by the dashed arrows .............................................................. 10

Figure 2.5 The electrostatic and hydrophobic clusters of the C-terminal helices
of HrpA .............................................................................................. 11

Figure 3.1 Coomassie-stained SDS-PAGE gel of purified HrpA from (a) P. syringae
and (b) E. amylovora .............................................................................. 18

Figure 3.2 MALDI-TOF-MS spectrum of the truncated, wild-type HrpA from P.
syringae. The peak centered at 9.8 kDa represents the truncated, wild-type HrpA and the
peak centered at 7.9 kDa is a consistently observed degradation product of

the HrpA ............................................................................................. 19

viii

Figure 3.3 Typical size exclusion chromatogram for HrpA from P. syringae ............ 20
Figure 3.4 Size exclusion chromatogram for isolation of cleaved HrpA .................. 22

Figure 3.5 SDS—PAGE gel depicting the time evolution of the thrombin digestion. Lane
1 contains molecular weight standards. Lane 2 contains HrpA diluted in the 1X cleavage
buffer pH 8.4. The band at 14.5 kDa represents the uncleaved HrpA and the band at ~9
kDa represents the consistently observed degradation product of HrpA that is consistent
in weight with the (His)6 tag cleaved HrpA. Lanes 3 and 4 contains the contents from
Lane 2 in the presence of thrombin (5 units/mg) at time t=0 and t=2 hours, respectively.

Lanes 5, 6, and 7 represent purified (His)6 tag cleaved HrpA collected from Size

exclusion fractions 10-12 .......................................................................... 23
Figure 3.6 HPLC chromatogram of purified, cleaved HrpA from P. syringae ........... 24
Figure 3.7 MALDI-TOF-MS spectrum of cleaved HrpA from P. syringae ............... 25

Figures 4.1 (a,b) TEM images of in vitro Hrp pilus assembly for (a) uncleaved and (b)

(His)6 tag cleaved wild-type, truncated HrpA from P. syringae in the presence
of TFE ................................................................................................ 29

Figures 4.2 (a-h) TEM images of truncated, wild-type HrpA from P. syringae in the
presence of 50 mM NaHgPO4 buffer pH 5.5 with (a) no salt (b) 300 mM NaF, (c) 300

mM NaZSO4, (d) 300 mM NaCl, (e) 300 mM NaNO3, (t) 300 mM NaBr, (g) 300 mM
NaI, and (h) 300 mM NaSCN ..................................................................... 32

Figures 4.3 (a,b) TEM images of truncated, wild-type HrpA from P. syringae in the
presence of 50 mM NaHzPO4 buffer pH 5.5 doped with ADP .............................. 34

Figures 4.4 (a-b) TEM images of truncated, wild-type HrpA from P. syringae in the

presence of 50 mM NaHzPO4 buffer pH 5.5 doped with the Dickerson-Drew dodecamer
of DNA .............................................................................................. 35

Figure 5.1 Sedimentation velocity of truncated, wild-type HrpA from
P. syringae .......................................................................................... 41

Figure 5.2 Sedimentation equilibrium of truncated, wild-type HrpA from
P. syringae .......................................................................................... 42

Figure 5.3 HSQC Spectrum of 15N-labeled truncated, wild-type HrpA from P. syringae.
F1 and F2 correspond to 15N and 1H chemical shifts, respectively ......................... 45

ix

Figure 5.4 Protein folding is observed by monitoring the absorbance differences
exhibited by the tyrosine chromophore at 287 nm as the tyrosine transitions from being
folded, or unexposed to the solvent (dotted line) to being unfolded, or exposed to the
solvent (solid line) ................................................................................. 47

Figure 5.5 Thermal melt analysis for the truncated, wild-type HrpA monomer from P.
syringae ............................................................................................. 49

Figure 5.6 Thermal melt analysis for the uncleaved truncated, wild-type Hrp pilus from
P. syringae .......................................................................................... 50

Figure 5.7 Thermal melt analysis for the (His)6 tag cleaved truncated, wild-type Hrp
pilus from P. syringae .............................................................................. 50

Figure 5.8 (a) CD spectrum of the truncated, wild-type HrpA monomer from P.
syringae in 50 mM NaHzPO4 buffer at pH 5.5 ................................................ 53

Figure 5.8 (b) CD spectrum of the truncated, wild-type HrpA monomer from P.
syringae in 10% TFE and 50 mM NaH2P04 buffer at pH 5.5 ............................... 54

Figure 5.8 (c) CD spectrum at 5°C of the truncated, wild-type HrpA monomer from P.

syringae in 50 mM NaHzPO4 buffer pH5.5 with (A) 25 mM NaF, (B) 100 mM NaF, (C)
300 mM NaF, and (D) 600 mM NaF ............................................................ 55

Figure 5.9 (a) CD spectrum at 5°C of the truncated, wild-type HrpA from P. syringae

incorporated into the pilus in the presence of 10% TFE in 50 mM NaHzPO4 buffer at pH
5.5 .................................................................................................... 56

Figure 5.9 (b) CD spectrum at 5°C of the truncated, wild-type HrpA from P. syringae

incorporated into the pilus in the presence 50 mM NaHzPO4 buffer at pH 5.5 containing
(A) no salt, (B) 10 mM NaF, (C) 50 mM NaF, (D) 100 mM NaF, and
(E) 150 mM NaF ................................................................................... 57

Figure 5.10 (a) Increasing temperature ramp from 5-60°C and its associated pilus
stability observed at (A) 5°C, (B) 10°C, (C) 20°C, (D), 40°C, and (E) 60°C ............. 58

Figure 5.10 (b) Decreasing temperature ramp from 60-5°C and the refolding of the
dissociated pilus at (A) 5°C, (B) 10°C, (C) 20°C, (D), 40°C, and (E) 60°C ............... 59

Figures 5.11 (a-g) CPMAS signals for single isotopically labeled, truncated, wild-type
HrpA from E. amylovora (a-c) and P. syringae (d-g) pili. HrpA from E. amylovora was

labeled with 13C=O (a) Ala, (b) Gly, and (c) Leu, and HrpA from P. syringae was labeled
with l3C=O (d) Ala, (e) Gly, (f) Lys, and (g) Met ............................................. 68

Figure 5.12 REDOR-filtered signal from the I-I3C Ile-ISN Leu unique pair in the
truncated, wild-type HrpA from P. syringae prepared pili ................................... 69

Figure 5.13 Local 2° structure information of multimeric HrpA by homology. For
HrpA from E. amylovora, the “*” and the “-“ indicate residues involved in (1) an a-helix
and (2) an a-helix or B-sheet, respectively. For HrpA from P. syringae, the “+” indicates
residues involved in an or-helix. The “A” indicates residues involved in an a-helix that
occur at the same position in both sequences. Finally, the “>” indicates residues having
the same sequence position in both pilins, but are involved in (1) an a-helix or (2) an or-
helix or B-Sheet in P. syringae or E. amylovora, respectively ................................ 71

Figure 7.1 First-order exponential decay of Hrp pilus unfolding ........................... 76

Figures 7.2 (a-d) TEM images of (a) I89P mutant HrpA, (b) 1:1 molar ratio of I89P
mutant and truncated, wild-type HrpA, (c) IlOlP mutant HrpA, and (d) 1:1 molar ratio of

IlOlP mutant and truncated, wild-type HrpA in the presence of 50 mM NaHgPO4 buffer
pH 5.5 with 300 mM NaF ......................................................................... 79

Figure 7.3 Photograph of needle-like crystals of (His)6 tag cleaved, truncated, wild-type
HrpA from P. syringae obtained after two days of incubation at 5°C in 200 mM

ammonium sulfate and 30% PEG 4000 ......................................................... 82
Figure 7.4 TEM image for truncated, wild-type HrpA from P. syringae in the presence
of 200 mM ammonium sulfate and 30% PEG 4000 ........................................... 84
Figure 7.5 Electron diffraction image for the TEM sample area in Figure 7.4 ............ 85

xi

I. Introduction to Type III Secretion and the Role of the Hrp Pilus

The type III secretion system plays a vital role in the virulence of many Gram-negative
bacteria [8]. Both animal (e.g. Yersinia, Shigella, and Salmonella) and plant (e.g.
Pseudomonas syringae and Erwinia amylovora) pathogenic bacteria utilize the type III
secretion to transport effector proteins from their cytosol to that of their eukaryotic host
cell [24]. These effector proteins suppress the innate immune response of the eukaryotic
host to benefit the propagation of the Gram-negative bacteria [11]. Though these
effectors vary greatly across bacterial species, the molecular transport machinery, the
type III secretion apparatus (T TSA), is structurally and functionally conserved. More
than 20 conserved translocation proteins are involved in constructing this
macromolecular structure that spans the inner bacterial membrane, the periplasmic space,
the peptidoglycan layer, the outer bacterial membrane, the extracellular space, and the
host barrier [44] (Figure 1.1). Thus, virulence relies upon this essential cytosol-to-cytosol

interaction mediated by the assembly of the TTSA [8].

 

 

 

 

 

 

 

 

 

 

 

//"°
M
W’- A
—- D

ere/'-

E\p R RPRR" 0000.
Mint. ._..
22 _ g. 2- rt". a .... P a:

 

 

rams. B 68:32:33.6
| J\<S AQI/ K

Figure 1.1 Model of the TTSA of a proteobacterial plant pathogen where (A) is the Hrp
pilus, (B) is the membrane-embedded base, (C) is the plant host cell, (D) is the
extracellular space, (E) is the outer bacterial membrane, (F) is the peptidoglycan layer,
(G) is the periplasmic space, (H) is the inner bacterial membrane, (1) is the bacterial
cytosol, (J) is the translocator protein, HrpA, chaperone complex, (K) is the effector
protein, HrpZ, chaperone complex, (L) is the unstructured HrpA monomer, and (M) is the
folded HrpA monomer associating with the tip of the growing pilus.

 

The TTSA is termed a “needle complex” [28] and is comprised of (1) a membrane-
embedded base and (2) a hollow extracellular filament, termed either a needle or a
hypersensitive response and pathogenticity (Hrp) pilus in animal or plant pathogens,
respectively [16]. The Hrp pilus and the needle are structurally and functionally similar,
sharing outer diameters of roughly 8 nm and small structural proteins (7-9 kDa).

Although both filaments have been shown to elongate through assembly of monomeric

subunits [12] induced by bacterium-host cell-to-cell contact [6], a distinct difference is
observed in their lengths. This is probably because of the differences in cell barriers of

plant and animal cells, specifically the >100 nm thickness of plant cell wall [16].

The major structural protein of the Hrp pilus is encoded by the HrpA gene, an essential
gene for the type-III-mediated hypersensitive response and pathogenicity [8]. The
expressed and purified translocator protein, the HrpA pilin from Pseudomonas syringae
pathovar tomato strain [DC3000], has been shown to reassemble into Hrp pili in vitro
[36]. Immunocytochemistry has shown that the Hrp pilus elongates by addition of HrpA
pilin subunits at its distal end and that the effector protein HrpZ is only secreted from the
tip of the pilus [24]. Thus, the Hrp pilus is thought to penetrate the plant cell wall [5] and

serve as a conduit for transport of effector proteins [17].

The mechanism for the association of HrpA monomers to form the pilus is not well-
understood. In this work, we hypothesize that HrpA, albeit natively unstructured as a
monomer, adopts a stable fold that is predominantly a-helical upon incorporation into the
Hrp pilus. This hypothesis is consistent with the existing evidence that Hrp pilus
assembly requires that the translocation of HrpA to the membrane-embedded base be
assisted by a chaperone [33]. Dissociation of the chaperone upon reaching the base
enables the natively unfolded HrpA to flow through the 2 nm inner diameter of the
forming pilus. Upon reaching the tip of the pilus, the HrpA monomer folds to continue
the assembly [24]. It is important to elucidate this mechanism because the pilus is

considered to be an attractive pharmaceutical target, both because it is extracellular and

because its elimination renders the bacteria non-virulent. The development of a small-
molecule inhibitor would serve to prevent the spread of crop diseases, including wilting,

blight, and leaf Spots.

I.A. Probing the Structure and Function of HrpA from P. syringae and E. amylovora

This work focuses on two unrelated HrpA pilins from plant pathogens, Erwinia
amylovora and Pseudomonas syringae pv. tomato [DC3000]. Conventional structure-
determination methods, such as X-ray crystallography and solution-state NMR, have so
far failed to obtain a structure for these proteins. We wish to overcome this obstacle and
determine the three-dimensional structure of both the HrpA pilin and Hrp pilus to aid in
drug design. This work will attempt to address several fundamental questions regarding
the structure and function of both the pilin and the self-assembled pilus: (1) What is the
structure of the HrpA monomer? (2) What is the structure of the Hrp pilus? (3) What are
the physical/chemical forces that stabilize the pilus? (4) What mechanistic information is

contained with the kinetics of pilus formation?

The following chapter presents a computational model for the predicted structure of the
HrpA monomer and for the assembly of the Hrp pilus. Subsequent chapters Show the use
of several Spectroscopic methods designed Specifically to determine the structure and

function of both HrpA and the Hrp pilus.

II. Computational Modeling of HrpA

Currently, the structure of the Hrp pilus and the mechanism of its self-assembly are not
well-understood. In this work, we have designed a computational, 3D model of the pilus
to depict both the electrostatic and hydrophobic interactions that might initiate and
stabilize the growth of the pilus in vitro. Spectroscopic methods can then be designed to

test the validity of this model.
II.A. Multiple Sequence Alignment (MSA)

The full-length, wild-type HrpA from P. syringae and E. amylovora were individually

subjected to two iterations of PsiBLAST [2,3], respectively. Using the non-redundant

database and an E-value cutoff of 10'6 to obtain homologous sequences, 10 PsiBLAST

hits were obtained for HrpA from P. syringae; additional iterations yielded no new hits.

Lowering the E-value threshold to 10-3, 5 hits were obtained for HrpA from E.

amylovora; additional iterations yielded no new hits. The final basis set included the 9
most identical sequences to P. syringae and the 5 sequences from E. amylovora, all
derived from Gammaproteobacteria. These sequences were aligned using the best-
validated multiple sequence alignment server, T-Coffee [34] and depicted using ESPript
[14] (Figure 2.1). Residue conservation was assessed by the mean BLOSUM62 score

among residues at a given position and by the sequence entropy

Entropy at position is 2 p k log2 p k (1)

where pk is the fraction of amino-acid type k at position i, where the summation spans all
possible amino-acid types k. The lower the entropy, the higher the conservation at the

position; a zero entropy implies that only one residue appears at that position, pk=l.

I 1? 10 19 ‘9 59
m K'.':‘\F:‘I317.'5KLTIILCNSAVA uni-.ASEATLOKNILLGI. .css
rm It‘h‘iE‘JISLTSKLTIIL'SNSAVG 'UTEITVASNATLQKS11.1.6?. .525
"NA H'.'.‘.FAC...SKI.TIILL‘,N.‘.A‘.‘:; , "G‘u'll‘tVAEIIAYLGKNILCCT.Nils-LR

  
   
 

 

    
 
    
    
  

 

 

 

rm NW!FAMJSKLTNLGNSA‘”: updtfimzomzidut. .505
"MU. H‘l’kE’fiEL?SELT?.’L'.2.‘:3.M.. , 2 ‘:i‘.‘f.'xrl.'x5NA5:.Q.-IN:1.551..655.
rm H‘JAFJICL‘JSKLTlJLGllSh‘v'L‘ QDL'ST‘IASNASLQKN!LSCT. .623
am H'.’hF:‘.:'-‘..!’.‘3KI.TMI.GNL‘.M'.‘I ~t.‘.j:.::r‘.'AA.-Ixrz;sss:1.331. .assmv

‘QANT I 555M301“?LAISTIIXGIIC- 1.5
“EMU ZASAAHMQJRR'LAC I Ell-{Gilt}

‘SSchaarcvns

luflLl' -: 2
"we H9125?EZTSSIJQASTBVAN f

 

my: ..................... 1, .......................
m .................. y... a ..................... s
nu! . ............ . is. .n...... .. . ..1.lm
my: .. .. .... '11.. ............ .. ..... t.
mam ......................... .g ..................... s

 

Figure 2.1 Multiple Sequence Alignment (MSA) of HrpA from Pseudomonas syringae
pv. maculicola (PSPVM), Pseudomonas syringae pv. tomato (PSPVTOM), Pseudomonas
syringae pv. actinidiae (PSPVA), Pseudomonas syringae pv. delphim'i (PSPVD),
Pseudomonas syringae pv. magnoliae (PSPVM_1), Pseudomonas syringae pv. theae
(PSPVTH), Pseudomonas syringae pv. tagetis (PSPVTA), Pseudomonas syringae
(HRPA_PS), Pseudomonas syringae pv. oryzae (PSPVO), Erwinia pyrzfoliae
(ERW_PYR), Erwinia amylovora (ERW_AM and HRPA_ERWAM), Pantoea stewartii
subsp. stewartii (PSSST), Pectobacterium carotovorum subsp. carotovorum (ERW_PC).
Highly conserved, moderately conserved, and unconserved residues are represented by an
outlined grey background with a white letter, an outlined white background with a
boldface black letter, and a colorless background with a black letter, respectively.

II.B. Secondary—structure Predictions

Secondary-structure predictions were performed using the three best-validated [22]
secondary-structure servers, PsiPRED [18, 29], SABLE2 [1], and SAM-T2K [19]. The
results were visualized using our web-server
http://proteins.msu.edu/Servers/Secondary_Structure/visualize_secondary_structure_pred

ictions.html.

H " Blight-- "HELEN” ,, HHHHIIHHIIII 9‘13?”

MVAPAGLTSRLTHLGHSAVGGVGGALOGVNTVLSIATLOIIIILLGTGDSL
u o a o o o

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

In
I T N l 1 ' II
a a s
2
1
. . | I O
nun-u. ..-~ -- .umsuuuunlllhn 1 gig“
svoaonxasgssnausanghquozruxgornbvnlstgasxaosruxg
'0
‘9 I" O G O
'1 JLI I H l___:
3 ar 3 r a
1
l
‘ WFHHM CQCQC
strnrnhxclsr
3 0
~ l'-'l
T

Figure 2.2 Predicted secondary structure of HrpA from P. syringae using SAM-T2K,
where H and C represent Helix and Coil, respectively, and the intensity relates to the
probability.

The secondary-structure predictions generated by PsiPRED and SABLE2 were consistent
with that of SAM-T2K (Figure 2.2), which shows the presence of distinct a-helical

regions.

II.C. Three-dimensional Modeling of the Hrp Pilus

Jin et al. and Li et al. have shown, through immunocytochemistry, that the effector
proteins, HrpZ and AvrPto, are extruded from the tip of the Hrp pilus, an observation
which supports the conduit model [17, 24]. This observation also provides direct support
for a Shared mechanism in type III secretion and flagellar assembly, for the eight hrc
genes involved in pilus assembly are homologous to those involved in flagellar assembly
[10]. This hypothesis was experimentally confirmed by Li et al. who showed that newly

produced HrpA proteins are added to the tip of the growing Hrp pilus [24].

In 2004, K. Namba et al. reported a plausible 3D structural model for the Salmonella
typhimurium flagellum. This model was created using a partial structure obtained by X-
ray diffraction and an image-reconstruction from electron cryomicroscopy [32,37].
Although the flagellin protein is much larger than the pilin protein, its N and C-termini
associate as antiparallel a-helices in the innermost core of the flagellum. These terminal
(ii-helices are the most closely corresponding parts of the flagellin to the HrpA pilin. In

the Namba model, these helices are roughly parallel to the flagellar axis.

In this work, we combined our results from the MSA and the secondary-structure
predictions along with the flagellar model to generate a putative model for the self-

assembly of the Hrp pilus (Figure 2.3).

helix 3
helix 2 helix 4 .. . .

helix 1

NC

Figure 2.3 Model for the self-assembly of the Hrp pilus

According to this model, the folded form of the HrpA pilin in the pilus adopts an a-
helical hairpin conformation, with the turn centered on the TMKK (QASK) motif of
HrpA from P. syringae (E. amylovora). The helices are arranged roughly parallel to the
pilus axis, with the C-terminal half forming the inner wall and the N-terminal half the
outer wall. Since, each C-terminal helix Should contribute equally to the net dipole
moment of the pilus, an electric field is produced within the pilus that is directed along its

axis pointing from the tip to the base of the pilus (i.e. from the plant host cell to the outer

bacterial membrane). The electric field will help drive the transport of the negatively
charged effector proteins toward the host cell (Figure 2.4). Being an a-helix, the side
chains of the C-terminal half should point from the base to the tip of the pilus (i.e. away
from the bacterial cell and towards the host cell), thus inhibiting back-diffusion of the

effector proteins.

 

 

“at
ml

 

+--------------------------

 

 

 

Figure 2.4 For a a-helix, contributions from polarized amide groups along the peptide
backbone generate a net dipole moment (P) which points from the C-terminus to the N-
terminus, neglecting side chain contribution. The direction of the induced electric field
(E) is depicted by the dashed arrows.

When combined with the MSA, this model allows us to identify clusters of interacting

residues, since the helices of adjacent subunits are half-staggered. Using this reasoning,

10

we propose two clusters of interacting side chains for HrpA from P. syringae, one

electrostatic and the other hydrophobic. The electrostatic cluster (Figure 2.5) consists of

7 residues from 3 subunits: K79, K80, D84, and N87 of subunit 1; D95, K99, and K100

of subunit 2; and finally, K109 and the C-terminal carboxylate of subunit 3. The

hydrophobic cluster (Figure 2.5) consists of 7 residues from 2 subunits: I89, G92, and

S96 Ofsubunit 1 and A108, G110, 1111, and Y113 Ofsubunit 2.

—|

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

—|
l 3
as '
DB4 K99 K109
N87 K100 1
2 C
I89 A108
G92 G110
896 I111
Y113
1 C

 

 

 

C

Figure 2.5 The electrostatic and hydrophobic clusters of the C-terminal helices of HrpA

III. Protein Preparation

To test the computational model, the pilins must be expressed and purified for subsequent
experimentation. The following sections Show methods developed and optimized to
express and purify the HrpA pilins from Erwinia amylovora and Pseudomonas syringae
for physical and spectroscopic measurements. Methods for expressing isotopically

labeled HrpA pilins from each of these plant pathogens are discussed, as well as thrombin
digestion of the N-terminal (His)6 tag on the purified truncated, wild-type HrpA from P.

syringae for subsequent crystallization trials. Preparation of a novel GFP-tagged HrpA

construct is also described.

III.A. Plasmid Constructs

As previously stated, this work focuses on two unrelated Hrp pilins from plant pathogens,
Psuedomonas syringae pv. tomato and Erwinia amylovora. The two plasmids described
below were prepared in the laboratories of Dr. D. Arvidson and Dr. G. Sundin of MSU,

respectively.

III.A.i. Truncated, Wild-type HrpA from P. syringae pv. tomato [DC3000]

Truncated HrpA (residues 41-115) from the DC3000 strain of P. syringae pv. tomato was
cloned using the NdeI and HindIII restriction sites into a pET28b expression vector. This

truncated sequence not only represents the portion of the protein that aligns with the full-

12

length Erwinia amylovora sequence, but is also observed to be the shortest naturally

multimerizing sequence [3 6]. The construct was verified by DNA sequencing.

III.A.ii. Full-length, Wild-type HrpA from E. amylovora

The full-length HrpA from the Eal 10 strain of E. amylovora was cloned using the NdeI
and BamHI restriction sites into a pET28a expression vector from Novagen [45]. The

construct was verified by DNA sequencing.

[II.B. Expression of Wild-type HrpA Pilins

The pET28 expression vector codes for the lac repressor gene (lac!) and for antibiotic
resistance to kanamycin. Isopropyl-beta-D-thiogalactoside (IPTG), a lactose analogue,
induces the lac operon by initiating a conformational change in [ac]. This induction

enables the expression system to express the protein of interest.

Using sterile techniques, an aliquot of the BL21(DE3) Escherichia coli (E. coli) cells
transformed with the HrpA expression vector was plated on an agar plate containing
sterilized LB media and 50 ug/mL kanamycin, and incubated at 37°C for 24 hours.
Several 15 mL sterile Corning conical tubes, each containing 5 mL of sterilized LB
media and 50 ug/mL kanamycin, were inoculated with single colonies from the plate.
These starter cultures were grown to a translucent consistency on a New Brunswick

Scientific Series 25 Incubator shaker at 250 rpm and 37°C for 6 hours. These starter

l3

cultures were utilized to innoculate 1 L shake flasks of the LB media described above.

These 1 L cultures were grown to an OD600 of 1.5-2.0 at 250 rpm and 37°C for 18 hours.

These cultures were induced with 1 mM filter sterilized IPTG for 6 hours at 250 rpm and
37°C to obtain optimal expression of HrpA. The cells were harvested by centrifiigation
at 4°C. 1 L centrifuge bottles were Spun at 4000 rpm in a Sorvall SLC-4000 rotor for 15
minutes. The supernatant was decanted and the resulting wet cell pellet was divided into
5-6 gram aliquots and stored at -80°C for the subsequent extraction and purification of

HrpA.

III.C. Expression of the Truncated, Wild-type 15N-labeled HrpA from P. syringae

The truncated, wild-type 15N-labeled HrpA from P. syringae was expressed according to
the protocol described in section III.B. with the following modifications. After the 1 L
culture had reached an OD600 of 1.0, it was pelleted by centrifugation at 4000 rpm in a
Sorvall SLC-4000 rotor at 4°C for 15 minutes. The pellet was resuspended in 1 L of IX
M9 minimal medium [40] containing 50 ug/mL kanamycin and 1 g/L of 15N-labeled

ammonium chloride. The resulting culture was incubated with shaking at 37°C and 1
mM IPTG was immediately added to induce expression of the isotopically labeled HrpA.

After a 6 hour induction period, the cells were harvested by centrifugation as above.

14

III.D. Expression of Single-Residue-Labeled HrpA from P. syringae

The truncated, wild-type single-residue-labeled HrpA from P. syringae was expressed

according to the modified protocol described in section III.C with the following

modification. To accomplish single-residue labeling (e.g. 1-13C or 15N-labeled alanine),

100 mg of the isotopically labeled amino acid of interest (section V.F) were introduced to

the 1 L culture after the media switch and just prior to induction with 1 mM IPTG.

III.E. Purification of HrpA Pilins

The pET28 expression vector includes an N-terrninal hexahistidine [(His)6] tag and a

thrombin cleavage site (Leu-Val-Pro-Arg-Gly-Ser) immediately N-terminal to the HrpA

insert. Our standard purification protocol exploits the (His)5 tag.

The 5,-6 gram aliquots of wet cells were thawed and resuspended in 40 mL of Nickel A
(NiA) buffer (Appendix) in the presence of two EDTA-free protease inhibitor tablets.
Cell lysis was performed by pulse sonication at 50 W and the resulting lysate was
incubated on ice. The lysate was transferred to 50 mL Oak Ridge centrifuge tubes and
pelleted by centrifugation, using a Sorvall RC5C Plus centrifuge and a 88-34 rotor, at
11,000 rpm at 4°C for 20 minutes. The supernatant was heat-treated at 70°C to remove
protein impurities and then re-centrifuged at 12,000 rpm. The supernatant was retained

and kept on ice.

15

Extraction and isolation of the HrpA is facilitated by (His)6 tag, which has an inherent

affinity for chelating to Ni(ll). The first stage of purification is accomplished by a Ni-
affinity column chromatography. An F PLC column was packed with 15 mL of Ni-NTA
agarose and washed, by gravity filtration, with three column volumes of the 20 mM
imidazole NiA buffer. 120 mL of lysate supernatant were allowed to bind to the column
by gravity filtration. The column was washed with three column volumes of the
following buffers: 1) NiA, 2) NiA + 6M Urea, and 3) NiA + 600 mM NaCl (Appendix).
The column was installed onto a Pharmacia Biotech FPLC system for purification. The
His-tagged HrpA was eluted with a gradient of imidazole, contained in NiB buffer
(Appendix) according to the following program on the Pharmacia Biotech Programmer

GP-250 Plus (Table 3.1)

Table 3.1 Gradient elution scheme from 0% to 100% NiB buffer

 

 

 

 

 

 

 

 

 

Volume (Ml) Function Value
0 Concentration % NiB 0
0 lemin 0.4
40 Concentration % NiB 100
40 mL/min 0.4
45 Concentration % NiB 100
45 mL/min 0.4

 

 

 

The fractions were collected according to the following Table 3.2.

16

Table 3.2 Composition of % NiB in the eluted fractions

 

 

 

 

 

 

 

 

 

 

 

 

 

Fraction # % NiB
1 0-50%
2 50-55%
3 55-60%
4 60-65%
5 65-70%
6 70-75%
7 75-80%
8 80-85%
9 85-90%
10 90-95%
1 1 95-100%
12 >100% volume

 

 

 

 

The concentration of HrpA in each fraction was determined using a photometric BCA
assay. This assay extrapolates the protein concentration from a standard curve of the
absorbances at 562 nm of known quantities of BSA. Consistent protein yields of ~30

mg/L of culture were obtained for both HrpA constructs.

Each fraction found to contain significant HrpA was concentrated by centrifugation using
an Eppendorf Centrifuge 5810R and 4 mL Amicon 5000 molecular weight cut-off
(MWCO) concentrators. The fractions were centrifuged at 4000 rpm for 20 minutes at

4°C. 13 [IL aliquots of each fraction were used to generate samples for SDS-PAGE.

SDS-PAGE samples were prepared according to Table 3.3.

Table 3.3 SDS-PAGE sample preparation

 

13 [LL HrpA Elution Fractions
5 uL 4x Sample Loading Buffer
2 [IL 10x Reducing Agent

 

 

 

 

 

l7

These samples were heated for 10 min at 70°C and loaded onto 12% Bis/Tris acrylamide
gels from Invitrogen. The electrophoresis was run at 200 mV for 45 minutes. The gels
were the stained with Coomassie Blue stain overnight. Upon de-staining, the following

results were obtained (Figures 3.1 a-b).

   

(b)

Figure 3.1 Coomassie-stained SDS-PAGE gel of purified HrpA from (a) P. syringae
and (b) E. amylovora

The presence of HrpA was confirmed by N-terminal sequencing, MALDI-TOF-MS

(Figure 3.2), and chemiluminescent Western Blot analysis using an Invitrogen

WestemBreeze kit with a primary antibody for the N-terminal (His)6 tag.

18

 

1800

 

1600 ~

1400 e

1200 *

H

O

O

O
1

800 ‘

Intensity

600 —

400 ~

 

 

200 ~

0 q

GSPO 7500 8500 9500 10500
-200

 

 

 

m/z

 

 

 

Figure 3.2 MALDI-TOF-MS spectrum of the truncated, wild-type HrpA from P.
syringae. The peak centered at 9.8 kDa represents the truncated, wild-type HrpA and the
peak centered at 7.9 kDa is a consistently observed degradation product of the HrpA.

The fractions containing the 95% pure HrpA were combined and concentrated to 2 mL

aliquots for F PLC injection. Further purification and characterization were accomplished

by size exclusion chromatography according to the system parameters given in Table 3.4.

Table 3.4 Size exclusion gel filtration system parameters

 

 

 

Parameter Value
Column HiPrep 26/60 Sephacryl SlOO
Flow Rate 1.3 mL/min

 

Equilibrate w/ 0.00 Ml
Empty Loop w/ 100.00 Ml
Fraction Size 5.00 M]
Length of Elution 1.1

 

 

 

 

 

 

 

19

The Amersham Sciences HiPrep 26/60 Sephacryl S-100 column was calibrated using
bovine pancreatic ribonuclease A, chymotrypsinogen A, ovalbumin, and albumin by
plotting the logarithm of either the hydrodynamic radius (calculated from the PDB
structures 7rsa, lcgi, lova, and 1ao6, respectively) or molecular weight versus elution
volume. A typical size exclusion chromatogram for the HrpA from P. syringae is given
in Figure 3.3. The elution was performed with a 15 mM Tris-HCl and 100 mM NaCl

buffer at pH 7.5.

 

25

_1

154

Absorbance at 280 nm (mAU)
H
O

 

 

5 1
0 “Crux/er”
-5 .. ‘M‘M'NW
-10 1 r 1
100 150 200 250

Elution Volume (mL)

 

 

 

Figure 3.3 Typical Size exclusion chromatogram for HrpA from P. syringae

20

The fundamental equation used to determine the effective molecular weight from the

elution volume was derived from the above calibration to be

Elution Volume — Void Volume = -43.9 In (mol. wt.) + 208.25 (2)

According to Equation (2), the effective molecular weight of HrpA having an elution
peak centered at 156.862 mL was 28.9 kDa. The discrepancy observed in the effective _
and calculated molecular weights suggests that the protein is at least partially
unstructured. Fractions 13-18 were retained because they spanned the breadth of the
peak centered at 156.862 mL. These fractions were concentrated to ~20 mg/mL and

stored at -20°C to serve as HrpA stock solutions.

III.F. Cleavage of the N-tenninal (His)6 Tag on HrpA from P. syringae

Thrombin cleavage of the N-terminal l7-residue His-tag was deemed crucial for protein
crystallization and to control for the contributions of the N-tenninus to pilus assembly.
The following thrombin digestion protocol was developed and optimized for efficiency

and yield.

Thrombin was dissolved in 1x Thrombin Digestion buffer at pH 8.4 (Appendix) to a
concentration of 1 unit/11L. This pH was chosen because it enables high thrombin
activity and because it is below the isoelectric point (p1 ~ 8.77) of the cleaved HrpA, thus

ensuring solubility. Thrombin digestion was performed by combining the following

21

components into a 15 mL Corning conical tube: (1) 30 mg of HrpA, (2) thrombin at 5
units/mg of HrpA from P. syringae, and (3) 1x Thrombin Digestion buffer in 3:1 volume-
to-volume ratio with HrpA for solubility. The reaction conical was placed on a New
Brunswick Scientific InnOva 2000 platform shaker for 2 hours at 25°C. PMSF was
added to a final concentration of 1 mM to inhibit thrombin activity. The reaction mixture
was concentrated to 2 mL for subsequent size exclusion chromatography (Figure 3.4) as

described above. This chromatography was employed to separate the thrombin and the
(His)6 tag from the cleaved HrpA. SDS-PAGE analysis was performed to monitor the

progress of digestion at several times (Figure 3.5).

 

 

 

 

 

 

 

 

 

40
’5
g Cleaved HrpA
e
U
C
1‘!
a
L
0
ill
.1:
<
35 . . t ‘
80 180 280 380
Elution Volume (mL)

Figure 3.4 Size exclusion chromatogram for isolation of cleaved HrpA

22

 

Figure 3.5 SDS-PAGE gel depicting the time evolution of the thrombin digestion. Lane
1 contains molecular weight standards. Lane 2 contains HrpA diluted in the 1X cleavage
buffer pH 8.4. The band at 14.5 kDa represents the uncleaved HrpA and the band at ~9
kDa represents the consistently observed degradation product of HrpA that is consistent

in weight with the (His)6 tag cleaved HrpA. Lanes 3 and 4 contains the contents from
Lane 2 in the presence of thrombin (5 units/mg) at time t=0 and t=2 hours, respectively.

Lanes 5, 6, and 7 represent purified (His)5 tag cleaved HrpA collected from size
exclusion fractions 10- 12.

The digestion was confirmed by three methods: chemiluminescent Western Blot analysis
with a primary antibody for the N-terminal (His)5 tag, N-terminal sequencing, and
HPLC/MALDI-TOF-MS (Figure 3.6-7). The N-terrninal sequencing showed the

presence of 2 predominant species: (1) GSHMNILLG... with a molecular weight of 7.9

kDa and (2) MNTLLG. .. with a molecular weight of 7.7 kDa.

23

 

 

Absorbanoe

 

1.1

0.9 4

0.7

0.5 l

0.3 ~

0.1 1

 

 

 

 

0 20 4O 60 80 100
‘l’lme (min)

 

Figure 3.6 HPLC chromatogram of purified, cleaved HrpA from P. syringae

24

m Mum-_.....-

 

 

500

400*

300 1

200 1

Intensity

 

100*

'“"'“" "” amt..."

 

 

 

" 1 00 I I I T I I
5000 6000 7000 8000 9000 10000

m/z

 

 

 

Figure 3.7 MALDI-TOF-MS spectrum of cleaved HrpA fiom P. syringae

The spectrum obtained from MALDI-TOF-MS confirmed the digestion of HrpA and the
results from N-terminal sequencing, as mass peaks were observed at 7.7 kDa and 7.9

kDa. From these coincident pieces of data, it was determined that the thrombin digestion

was successful in preparing the (His)6 tag cleaved form of HrpA with high purity.

25

IV. In Vitro Preparation of the Hrp Pilus

The experimental determination of the structure and function of HrpA depends on
isolating stable forms of both its monomeric and multimeric states. Spectroscopic probes
of these two forms will provide insight not only into the structure and folding of HrpA,

but also into the formation of the Hrp pilus.

IV.A. Two Forms of HrpA: Monomer and Multimer

It has been hypothesized that, in the bacterial cytosol, the full-length HrpA pilin exists as
a mostly unstructured monomer linked in a chaperone complex [33]. Upon reaching the
membrane-embedded base, the monomer dissociates from the native chaperone complex.
This event is essential for pilus formation because only unstructured proteins can fit
through the inner diameter (2 nm) to translocate. When the monomer reaches the tip of
the growing pilus, it adopts a stable fold and adds to the assembly [24]. The resulting
non-covalent multimer of HrpA, termed the Hrp pilus, serves as a conduit for bacterial
effector proteins. The following work attempts to probe the mechanism for this self-

assembly.

To probe the self-assembly of HrpA, it was necessary to develop a means for reliably
producing the pilus in vitro. Roiné et al. have successfully Shown that the Hrp pilus can
be prepared in vitro from purified HrpA [36]. He et al. have shown that increasing the

pH from 5.0 to 7.5 induces pilus formation at pH>6.5 [23].

26

We have performed a more systematic study of co-solvents and their effects on Hrp pilus
formation. The secondary structure predictions presented in section II.B Show that the
stable folding of HrpA should be predominantly a-helical. The use of neutral helix-
favoring co-solvents, such as 2,2,2-trifluoroethanol (TFE) would be a reasonable
approach to induce the folding of monomeric HrpA [39]. Also, at relatively low pH
(pH<6.0), the surfaces of the cleaved and uncleaved HrpA monomers are positively
charged, for the isoelectric points are 8.77 and 9.6, respectively. Thus, it has been
hypothesized that the use of well-chosen anions might stabilize the electrostatic
interactions of the pilin and lower the free energy of the system enough to drive pilus

formation.

Our study shows that pilus formation is selective for helix-favoring co-solvents and for
the lyotropic anions of the Hofmeister series [4]. The denaturing effects of chaotropic
anions, even those existing within the same family as stabilizing lyotropic anions (i.e.
halide salts), are manifested in the absence of pili. The fact that HrpA does not self-
assemble in the presence of chaotropic anions suggests that the driving force for pilus
formation is the combination of the stabilization of electrostatic interactions within the

HrpA monomer and the hydrophobic interactions existing between adjacent monomers.

IV.A.i. 2,2,2-trifluoroethanol (TFE)

TF E across the concentration range of 10-20% (v/v) is a helix-favoring co-solvent [39].

TFE likely induces or-helical structure by disrupting the solvation sphere of the

27

unstructured protein, thus making it more likely that the amide hydrogens will form

intramolecular hydrogen-bonds.

To test the effects of TFE on Hrp pilus formation, 100 uL of 20 mg/mL cleaved and
uncleaved monomeric HrpA were exchanged from a pH 7.5 buffer containing 15 mM

Tris-HCl and 100 mM NaCl to a pH 5.5 buffer containing 10% (v/v) TFE and 50 mM

NaH2P04. Buffer exchange was accomplished in two repetitive stages of 50-fold

dilution in the 10% TFE buffer followed by concentration to the initial volume in 5 mL
Amicon 5000 MWCO centrifuge tubes. The resulting 100 11L were transferred to a 0.5
mL Eppendorf tube and frozen for 24 hours at -20°C. Upon thawing to 25°C, the

samples exhibited a high viscosity consistent with pilus formation.

These samples were plated on copper grids, negatively stained with 2% aqueous
phosphotungstic acid, and examined using transmission electron microscopy (TEM) at
the Michigan State University Center for Advanced Microscopy. Conventional light/dark
field images by electron scattering were obtained using a JEOL (Japan Electron Optics
Laboratories) 100CXII microscope with the help of Dr. Alicia Pastor-Lecha (Figures 4.1

a,b).

28

 

Figures 4.1 (a,b) TEM images of in vitro Hrp pilus assembly for (a) uncleaved and (b)
(His)5 tag cleaved wild-type, truncated HrpA from P. syringae in the presence of TFE.

These images verify that this protocol successfully generated pili at pH 5.5, a result that
disagrees with that of He et al. who found that pilus formation is induced only upon
increasing the pH above 6.5. The ability to produce pili at this relatively low pH has
enabled the examination of the electrostatic contributions to pilus formation. The

presence of pili in both the cleaved and uncleaved samples has proven that the N-terminal

(His)6 tag does not play any role in the self-assembly process.

IV.A.ii. Salts

The previously observed dependence of pilus formation on pH suggests that electrostatic
repulsion between the HrpA monomers contributes significantly to the free energy of
pilus formation. Hence, we hypothesized that anions would stabilize the positively

charged surface of monomeric HrpA at pH 5.5 by Manning-type condensation [27], in

29

which ions are quasi-permanently bound to a linear charge density, such as that of the
Hrp pilus. A set of anions was selected to examine the range of lyotropic to chaotropic

anions depicted in the Hofmeister series [4].

50 mM NaHzPO4 buffers at pH 5.5 were doped with 300 mM quantities of the following

salts: NaF, Na2SO4, NaCl, NaNO3, NaBr, NaI, and NaSCN. The sample preparation

described in section IV.A.i was followed for each buffer, and the resulting 100 11L were
transferred to 0.5 mL Eppendorf tubes and frozen for 24 hours at -20°C. Upon thawing

to 25°C, viscosity was observed only in the buffers containing 300 mM NaF, NaCl,
NaNO3, and Na2SO4. 20% dilutions were prepared for TEM according to the above

freeze/thaw cycle.

The samples were plated on copper grids and negatively stained with 2% aqueous

phosphotungstic acid for TEM. Conventional light/dark field images by electron

scattering were obtained (Figures 4.2 a-h).

30

 

(b) NaF

(a) No salt

 

(d) NaCl

 

(f) NaBr

(e) NaNO3

31

   

(g) Nal (h) NaSCN

Figures 4.2 (a-h) TEM images of truncated, wild-type HrpA from P. syringae in the
presence of 50 mM NaH2P04 buffer pH 5.5 with (a) no salt (b) 300 mM NaF, (c) 300

mM Na2SO4, (d) 300 mM NaCl, (e) 300 mM NaNO3, (t) 300 mM NaBr, (g) 300 mM
NaI, and (h) 300 mM NaSCN.

The samples containing the halide salts closely follow the lyotropic to chaotropic nature
of the Hofmeister series [4]. The sample containing F_ ions induced the formation of pili

having well-defined inner and outer diameters, 2 nm and 8 nm, respectively. The C1-
containing buffer elicited a “coral reef’-like structure, implying a moderate disorder to
the structure. The Br- and 1. containing buffers showed no structural forms in
comparison with the phosphate buffer control. The buffer containing the fully denaturing
anion, SCN', also Showed no structural forms when compared with the control.
According to the Hofmeister series, SO42- ions Often display similar lyotropic effects to
F- ions and N03- ions are reportedly similar to Cl- ions. The results of this work are
consistent with the known behaviors of these ions, as the N03- and SO42' buffers

induced the formation of pili and “raft”-like arrangements of pili, respectively. The

32

ability of S042- to form “raft”-like arrangements of pili may result from its divalent

nature, being able to coordinate with two pili simultaneously.

IV.A.iii. Stable Multimers of Uniform Dimensions

To probe the folding kinetics of HrpA or to obtain definitive secondary/tertiary structural
information via solid-state NMR, electron cryomicroscopy, or electron fiber diffraction, it
is desirable to produce stable multimers of uniform length. We hypothesized that this
would be feasible through polyanion driven Manning-type condensation. Two well-
chosen poly-anions were selected: ADP and the Dickerson-Drew dodecamer of DNA.
ADP was chosen to provide insight as to the validity of this hypothesis. The Dickerson-
Drew dodecamer has a unique palindromic sequence of 5’-CGCGAATTCGCG-3’ [9].
The reason for selecting the palindromic dodecamer of double-stranded DNA was two-
fold: (1) its polyanionic structure has consistent and uniform length and (2) the outer
diameter of the double-helix was slightly smaller than the inner diameter of the pili

observed above (Figure 4.2 a).

IV.A.iii.a. Adenosine Diphosphate (ADP)

50 mM NaHzPO4 buffer at pH 5.5 was doped with 278 mM ADP. The sample

preparation described in section IV.A.i was followed for the ADP-doped buffer. The
resulting 100 11L were transferred to a 0.5 mL Eppendorf tube and frozen for 24 hours at -

20°C. Upon thawing to 25°C, the sample exhibited a high viscosity consistent with pilus

33

formation. A 20% dilution was prepared for TEM according to the above freeze/thaw

cycle.

The sample was plated on copper grids and negatively stained with 2% aqueous
phosphotungstic acid for TEM. Conventional light/dark field images by electron

scattering were obtained (Figures 4.3 a,b).

 

(a) (b)

Figures 4.3 (a,b) TEM images of truncated, wild-type HrpA from P. syringae in the
presence of 50 mM NaH2P04 buffer pH 5.5 doped with ADP.

The observation of pili in this sample exhibited evidence to support the hypothesis that

polyanion driven Manning-type condensation was reasonable for the HrpA system.

IV.A.iii.b. Dickerson-Drew Dodecamer of DNA

100 11L of 30 mg/mL uncleaved monomeric HrpA were exchanged from a pH 7.5 buffer

containing 15 mM Tris-HCl and 100 mM NaCl to a pH 5.5 buffer containing 50 mM

34

NaH2P04. The sample preparation described in section IV.A.i was followed, and the

resulting 100 uL were transferred to a 0.5 mL Eppendorf tube.

920 nanomoles of lyophilized Dickerson-Drew dodecamer were dissolved in 100 [AL
Tris-EDTA (TE) buffer (Appendix). The resulting DNA solution was displaced into the
100 11L of buffer exchanged HrpA until a final concentration of 4 mM DNA was
achieved. The resulting mixture was frozen at -20°C for 24 hours. Upon thawing to
25°C, a white precipitate was observed within the sample, which we hypothesized could
signify the presence of a large number of small pili. A 10% dilution was prepared for

TEM according to the above freeze/thaw cycle.

The sample was plated on copper grids and negatively stained with 2% aqueous
phosphotungstic acid for TEM. Conventional light/dark field images by electron

scattering were Obtained (Figures 4.4 a-b).

   

Figures 4.4 (a-b) TEM images of truncated, wild-type HrpA from P. syringae in the

presence of 50 mM NaH2P04 buffer pH 5.5 doped with the Dickerson-Drew dodecamer
of DNA.

35

From these images, it was determined that the pili adopted stable multimers having a 2
nm inner diameter, an 8 nm outer diameter, and a longest dimension ranging from 40 to
160 nm. Knowing that the dodecamer is approximately 4 nm in its longest dimension, it
was estimated that the smallest observable pilus condenses onto 10 units of DNA. From
these images, it was interpreted that the rigidity of the DNA only supports pilus growth to
the terminal endpoint at 160 nm. This terminal length is likely the point at which

bending and torsional forces, as depicted in Figures 4.1-4.3, affect pilus elongation.

36

V. Secondary and Tertiary Structure of Monomeric and Multimeric HrpA

General hydrodynamic and spectroscopic methods were utilized to probe the 2° and 3°

structure of HrpA. Analytical ultracentrifugation (AUC) and size exclusion
chromatography are hydrodynamic techniques that were used to answer the following
two questions: ( 1) Is HrpA monomeric or multimeric under native conditions? (2) Is
monomeric HrpA folded under native conditions? Thermal melt (TM) UV, circular
dichroism (CD), and liquid-state NMR spectroscopies provided complementary
techniques for determining the answer to the question regarding the folding state of

HrpA. CD and solid-state NMR (SSNMR) spectroscopies were employed to probe the

2° structure of multimeric HrpA. Finally, thermal melt UV spectroscopy was also used

to probe the unfolding transition of multimeric HrpA.

The hydrodynamic techniques Showed that, under native conditions, HrpA exists as a
mostly disordered monomer. This result was further supported by the data obtained from
the complementary spectroscopic methods, TM analysis, CD, and liquid-state NMR.
Structural analysis of multimeric HrpA, by CD and SSNMR, showed that upon self-
assembly into the Hrp pilus, HrpA adopts a predominantly a-helical structure. A
discussion of these hydrodynamic and spectroscopic techniques and of the interpretation

of the above results appears in the subsequent sections.

37

V.A. Analytical Ultracentrifugation

Analytical ultracentrifugation (AUC) is a hydrodynamic tool that can assay the mass,
density, and hydrodynamic volume of proteins under biologically relevant conditions
[22]. For studying proteins, the concentration is usually assayed by absorption at 280 nm
due to the tryptophan and tyrosine residues; however, other spectroscopic probes of
concentration can be used, such as fluorescence, refractrometry, and absorption at other
wavelengths (e.g. 215 nm). Real-time monitoring of the sedimentation of these
biomolecules in a centrifugal field enables extraction of both hydrodynamic and
thermodynamic information unique to the system. This is accomplished through two
complementary methods: (1) sedimentation velocity (SV) and (2) sedimentation

equilibrium (SE) [22].

In SV, the application of a strong centrifugal force causes the depletion of protein at the
meniscus and the formation of a concentration boundary that moves toward the base of
the centrifuge cell with time [22]. Analysis of the concentration boundary velocity yields
the sedimentation coefficient, 5, of the protein. The s-value is determined by the

Svedberg equation:

it M(1-VP)
- (3)
(1127‘ NAf

 

where p. is the radial velocity of the protein, a) is the angular velocity of the rotor, r is the

radial posmon of the boundary, a) r 18 the centnfugal acceleration, M 13 the molar mass, v

38

is the partial specific volume of the solute (i.e. protein), p is the solution density, N A is

Avagadro’s number and f is the frictional coefficient [22]. The measured sedimentation

coefficient, 3, is directly proportional to the buoyant mass of the protein by

M - M(l — vp) (4)

where Mb is the buoyant mass. Through this relation, SV is sensitive to the protein mass

and shape; thus, it enables one to distinguish multimeric states within the solution.

In SE, the centrifugal force is applied to establish an equilibrium between sedimentation
and diffusion [22]. A lesser centrifugal force is utilized to ensure an adequate
equlilbration time of ~5 days. At equilibrium, the Observed concentration distribution

conforms to a Boltzmann distribution given by

M bsz2
C,q(R> - exp —— (5)

2k BT

where R is the radius, a) is the angular velocity of the rotor, k3 is Boltzmann’s constant, T

is the temperature in Kelvin, and Mb is the buoyant mass from Equation (4). In

comparison with SV, the SE method is only sensitive to the protein mass and multimeric

states are differentiated by multiple distributions.

39

In this work, SV and SE data were collected for the wild-type, truncated HrpA from P.

syringae on a Beckman XLI analytical centrifuge. The samples for each experiment

consisted of FPLC-purified HrpA (1 mg/mL) in 450 uL of 50 mM NaHzPO4, 300 mM

NaCl at pH 5.5. The sample for SV was aliquotted to a centrifuge cell with quartz
windows and centrifuged at 60000 rpm with scans of absorbance at 230 nm recorded
every ten minutes for 12 hours at 4°C. One sedimenting Species was observed having a

sedimentation coefficient of 0.97 Svedbergs that was determined by linear fitting of the

logarithm of the boundary position versus mzt (Figure 5.1), where t is the time in seconds

and a) is the rotational speed in radians/sec; the boundary position was calculated by the

second-moment method [7, 13].

40

 

1.95 '-

log (Bormdary Position)
3.?
l

 

.8 -
' 5 y=9.73x1014x+1.82

   

 

 

I L I 1 I 1
l .8
0 5e+11 le+12 l.5e+12
Omega/‘2 "‘ t

 

Figure 5.1 Sedimentation velocity of truncated, wild-type HrpA fi'om P. syringae

The buoyant mass (Mb) of HrpA was calculated to be 1.2 kDa, assuming a solution

viscosity equal to that of water at 20°C, a protein solvation of 0.3 g/g, a solution density

of 1 mg/mL, and a protein specific volume of 0.73 mL/mg [7].

The sample for SE was centrifuged at 10000 rpm for five days at 4°C. A buoyant mass

of 1.8 kDa was obtained from the slope of the linear fit of the logarithm of the final

. 2 2 . . . .
concentration versus (1) r /2kBT (Figure 5.2), where r 18 the radlus, k3 IS Boltzmann’s

constant, and T is the temperature in Kelvin.

41

 

 

 

 

'0.3 I I l I I I I 1 I T I j
.- ’ -i
, 1.
-0.4 '— v‘ ' -2
1- "il " Cl
A q"'
< .
=9; -0.5 — f" _.
._ MI
0 l- 11' -l
.5 v4"
-0 6 - 1' -1
i 1 ‘-‘
g - ‘ ‘1 1 .
a l. ‘
v .0.7 —— - l —
°° ' 1
'9 u M; ' i
1 . y=0.04x—2.32
-0.8 — fl" ‘
0.9 L l l l l l L l l 1 l l
38 4O 42 44 46 48 50
Radius Squared (cm"2)

Figure 5.2 Sedimentation equilibrium of truncated, wild-type HrpA from P. syringae

The observed molecular weight was obtained from the relationship between the buoyant
mass and molar mass given by Equation (4), assuming a solution density of 1.0 mg/mL
and a protein partial specific volume of 0.73 mL/mg [7]. The observed molecular weight
for the truncated, wild-type HrpA from P. syringae was determined to be 6.69 kDa.
Given that the monomeric weight is known to be 9.8 kDa from MALDI-TOF-MS
(section III.E), we can estimate the protein partial specific volume to be 0.8156 mL/mg,

which is consistent with being unstructured.

42

V.B. Size Exclusion Chromatography

Size exclusion chromatography can be utilized both as a means of protein purification
and as a means of obtaining hydrodynamic properties of the protein of interest. This
technique provides insight into the hydrodynamic radius, the effective molar mass, and

the state of folding of the protein.

Size exclusion chromatography was performed according to section III.E. on an
Amersham Sciences HiPrep 26/60 S-100 HR sephacryl column calibrated with bovine
pancreatic ribonuclease A, chymotrypsinogen A, ovalbumin, and albumin by plotting the
logarithm of either the hydrodynamic radius (calculated from the PDB structures 7rsa,

lcgi, lova, and 1ao6, respectively) or molecular weight versus elution volume.

Through analysis of Figure 3.4 by Equation (1), the effective molecular weight of HrpA
having an elution peak centered at 156.862 mL was calculated to be 28.9 kDa. The
discrepancy observed in the effective molecular weight and theoretical molecular weight
(9.6 kDa) suggests that the protein is unstructured. The higher observed molecular
weight correlates with the increased amount of friction experienced by an unfolded

protein in comparison with a folded protein.

43

V.C. Solution-state 2D-NMR and HN-Heteronuclear Single Quantum Correlation

(HN-HSQC)

For small soluble proteins (<10 kDa), solution-state 2D-NMR provides an excellent set of
experiments for determining both folding and structure. The HN-HSQC pulse sequence

enables one to observe each of these fundamental frequencies as a correlation between
15N and 1H chemical shifts [26]. The distribution of 1H chemical Shifts is directly
related to the foldedness or unfoldedness of the protein of interest. A large dispersion of
1H chemical shifts (4-5 ppm) is indicative of a folded protein, while a narrow dispersion

(<2 ppm) indicates an unfolded protein. This incongruity in chemical shift ranges is
derived from the differences in local fields experienced by an amide proton in a folded
versus unfolded state. In the unfolded state, the amide protons are sufficiently exposed to
the solvent system and, thus, Share a similar chemical environment. However, in the
folded state, amide protons may experience very different chemical environments. In this
work, HN-HSQC was utilized for analysis of the state of folding of monomeric HrpA

from P. syringae.

15N-labeled wild-type, truncated HrpA from P. syringae was concentrated to 1 mM

within a minimal salt buffer, consisting of 10 mM NaH2P04 at pH 7.5 and 10% D20.

An HN-HSQC Spectrum was generated on a Varian INOVA 600 MHz spectrometer at

25°C using the standard probe and pulse sequence. 1024 t2 scans were collected and

signal averaged over 2 hours to ensure adequate baseline resolution (Figure 5.3).

44

P1 , 0
(ppm): ‘

 

114 : . V
.08 Q 0 l
116 «90
», ea
- .. a - O.
118 5 a.
.- 0A
. ‘. O i
1 O '
12° 1 O. P. i
i e ' I
I a.
' e 90
. th 0
124 E 0"
. o o
1 O
l 9" o
126 i o e
e
0 o
128 1 l?
3 e
1' .. 1 1111
8.2 7.8 7.4 7.0 6.6
F2 (m)

Figure 5.3 HSQC Spectrum of 15N-labeled truncated, wild-type HrpA from P. syringae.
F1 and F2 correspond to 15N and 1H chemical shifis, respectively.

The observed narrow dispersion of 1H chemical shifts (~0.6 ppm) provided another piece

of evidence in support of the hypothesis that the HrpA monomer is natively unstructured.

45

V.D. Thermal Folding Transitions by UV Spectroscopy

Determining the thermal stability of the folding of proteins containing aromatic residues
is facilitated by the application of thermal melt (TM) UV spectroscopy. Through

monitoring the absorbance at 287 run, one is able to observe changes in folding through

up
detection of the red-shifted fluctuations for the n—m transition of aromatic residues (e.g.

tyrosine for HrpA from P. syringae), a class of 280 nm chromophores. If the protein
were folded, the aromatic residue would not be exposed to the solvent, in contrast, if the
protein were unfolded the residue is completely exposed to the solvent system. The
absorbance signal for the unfolded protein would be blue-shifted and less intense because

of solvent quenching effects (Figure 5.4).

46

.E‘HQ :qfiv'fi 11m

Absorbance (Arb. Units)

 

 

 

 

 

Wavelength (nm) .

Figure 5.4 Protein folding is observed by monitoring the absorbance differences
exhibited by the tyrosine chromophore at 287 nm as the tyrosine transitions from being
folded, or unexposed to the solvent (dotted line) to being unfolded, or exposed to the
solvent (solid line)

The sample is taken across a reversible temperature ramp that is designed to denature and
possibly refold the protein of interest. Values for the thermal unfolding transition and the

enthalpy of folding are extracted from fitting these data to the simple two-state folding

transition with constant heat capacity model given by

 

AH T
_A€=£_fl-[ m]*[1__m_) (6,

47

where Tm and AHm represent the melting temperature and the folding enthalpy,

respectively [3 5].

The thermal unfolding transition for the wild-type, truncated HrpA from P. syringae was

monitored by absorbance at 287 nm on a Shimadzu UV-l700 spectrophotometer under

computer control. HrpA was diluted to 1 mg/mL in 20 mM Mes, 100 mM NaHzPO4 and

10 mM NaF at pH 6.0. The sample was monitored along a reversible temperature ramp
from 2°-70°C and then back to 2°C using a Shimadzu S-1700 thermoelectric (Peltier) cell
holder coupled to a ThermoForma Model 2602 circulating water bath; a nitrogen
atmosphere was maintained to minimize condensation on cell windows. For analysis
purposes, the absorbance at 287 nm was normalized with the simultaneously collected

absorbance at 280 nm (Figure 5.5).

48

 

 

 

 

 

 

0.59 . . . . r .
o 10 20 30 4o 50 so 70 30

Temperature (Celsuis)

 

 

Figure 5.5 Thermal melt analysis for the truncated, wild-type HrpA monomer from P.
syringae
The absence of change in absorbance at 287 nm signifies that, under native conditions,

monomeric HrpA is unstructured.

From the EM data, it was known that buffer condition of 10% TFE in 50 mM NaHzPO4

at pH 5.5 supported pilus formation in both the His-tag cleaved and uncleaved forms of
the wild-type, truncated HrpA from P. syringae. Ten-fold dilution in fresh buffer was
performed on the 10% TFE buffered samples that had been previously prepared for EM
to yield working volumes of concentration 2 mg/mL for TM analysis. These samples
were frozen at -20°C for 24 hours to maintain intact pili post-dilution. TM data were

recorded according to the protocol from above (Figures 5.6-7)

49

 

 

'00
8&3.

0.0.0.0
£385:

rbance 287/280 nm
9 .0
N M
C” G)

0
$3
N
.5

n 0.22 4

As

 

 

_O
N
1

 

10 20 30 40 50 60 70 80
Temperature (Celsuis)

O

 

 

Figure 5.6 Thermal melt analysis for the uncleaved truncated, wild-type Hrp pilus from
P. syringae

 

 

0.3
5 0.295«
g 0.29 -
g 0.285 <

0.28
. 0.275 -
8 0.27.
3 0.2654
g 0.26 <
a 0.255 <
i 0.25 . r . 1 a f 2
j 0 10 20 30 40 50 60 70 80
: Tempentule [Celsuis]

 

 

 

 

 

Figure 5.7 Thermal melt analysis for the (His)6 tag cleaved truncated, wild-type Hrp
pilus from P. syringae.

Thermal melting of the Hrp pilus exhibited an irreversible unfolding transition at 30°C.

The noticeable difference existing between the cleaved and uncleaved traces was thought

50

to be related to the difference in stability of the pili upon dilution. Thus, there is clearly a
spectroscopic difference between the unstructured HrpA monomer and the folded Hrp
pilus. Given that there is only 1 aromatic residue, the C-terminal tyrosine, this residue is

likely to be buried in the pilus

V.E. Circular Dichroism (CD)

Plane-polarized light comprises two circularly polarized components of equal magnitude,
one rotating clockwise (dextrorotary) and one rotating counter-clockwise (levorotary). A
CD signal is generated when unequal absorption of these two components occurs as the
radiation passes through a sample that possesses a chiral chromophore, such as a protein.

The unequal absorption reorders the plane-polarized light into elliptically polarized light

[20]. Far-UV CD (170 -250 nm) is sensitive to the 2° structure of proteins and is widely
utilized as a spectroscopic means to estimate the relative percentages of 2° structure. In

this work, CD was employed to examine the 2° structure of both the monomer and

multimer of the wild-type, truncated HrpA from P. syringae.

CD spectra were generated on a JASCO J -810 spectropolarimeter according to the system

parameters outline in Table 5.1.

51

Table 5.1 JASCO J-810 acquisition parameters

 

System Parameter Value
Scanning Mode Continuous
Scanning Range 180-250 nm
Scanning Speed 2 nm/min
Band Width 1 nm

# of Scans 5

 

 

 

 

 

 

 

 

 

The temperature was controlled using a Hellma 165-QS water-jacketed, l-cm cuvette that

was coupled to a ThermoNeslab Model RTEl 11 circulating water bath.

1 mL solutions containing a 10 11M final concentration of monomeric HrpA were

prepared in each of the following buffer conditions: (1) 50 mM NaHzPO4 at pH 5.5, (2)
10% TFE in 50 mM NaHzPO4 at pH 5.5, and (3) 10, 25, 50, 100, 150, 300, and 600 mM

NaF in 50 mM NaHzPO4 at pH 5.5. The resulting CD spectra acquired at 5°C are given

in Figures 5.8 (a-c).

52

 

 

Elllpticlty (mdeg)

 

 

 

 

'20 fi 1
190 210 230 250

Wavelength (nm)

—.- -—1.—-. “_.-..-" ...- ”_..... fin . ... .

 

 

Figure 5.8 (a) CD spectrum of the truncated, wild-type HrpA monomer from P.
syringae in 50 mM NaHzPO4 buffer at pH 5.5

53

 

 

30*
20J

10‘

 

Elllptlclty (mdeg)

 

 

 

 

'40 r I ‘
190 210 230 250

Wavelength (nm)

 

 

 

Figure 5.8 (b) CD spectrum of the truncated, wild-type HrpA monomer from P.
syringae in 10% TFE and 50 mM NaH2P04 buffer at pH 5.5.

54

 

 

 

—B
C
D

 

 

Elllptlclty (mdeg)

 

 

 

 

190 200 210 220 230 240 250
Wavelength (nm)

 

 

Figure 5.8 (c) CD spectrum at 5°C of the truncated, wild-type HrpA monomer from P.

syringae in 50 mM NaH2P04 buffer pH5.5 with (A) 25 mM NaF, (B) 100 mM NaF, (C)
300 mM NaF, and (D) 600 mM NaF.

The CD spectra (Figures 5.8 (a-c)) of the wild-type, truncated HrpA monomer from P.

syringae were observed to follow the trace characteristic of a random coil [20], having a

55

single minimum at 204 nm. This is consistent with the monomer being unstructured

under native conditions.

Following the protocol for preparation of pili described in section IV.A.i-ii, 100 ILL

samples were prepared and diluted to 1 mL in the following buffer conditions: (1) 10%

TFE in 50 mM NaHzPO4 at pH 5.5 and (2) 1, 5, 10, 25, 50, 75, 100, 150, and 300 mM

NaF in 50 mM NaHzPO4 at pH 5.5. The resulting CD Spectra acquired at 5°C are given

in Figures 5.9 (a-b).

 

 

 

Ellipticity (mdeg)

 

 

 

'80 I I
190 210 230 V 250

Wavelength (nm)

 

 

 

Figure 5.9 (a) CD spectrum at 5°C of the truncated, wild-type HrpA from P. syringae

incorporated into the pilus in the presence of 10% TFE in 50 mM NaHzPO4 buffer at pH
5.5

56

 

 

 

 

 

Elliptlclty (mdeg)

 

 

 

 

190 200 210 220 230 240 250
Wavelength (nm)

 

 

 

Figure 5.9 (1)) CD spectrum at 5°C of the truncated, wild-type HrpA from P. syringae

incorporated into the pilus in the presence 50 mM NaHzPO4 buffer at pH 5.5 containing
(A) no salt, (B) 10 mM NaF, (C) 50 mM NaF, (D) 100 mM NaF, and (E) 150 mM NaF.

Under conditions identified by EM to support pilus formation, the CD spectra (Figures
5.9 (a-b)) were consistent with that of a folded protein [20], having a double minimum at
208 and 222 nm, which is consistent with the predicted (it-helical 2° structure presented in
section II. The results support the hypothesis that the monomer is natively unstructured,

but then folds upon incorporation in the pilus.

The thermal stability of pilus was monitored for the above solution of pili prepared with

300 mM NaF and 50 mM NaHzPO4 buffer at pH 5.5. CD spectra were acquired every

57

5°C across a reversible temperature ramp from 5-60°C and then back to 5°C (Figures

5.10 a-b).

 

Ellipticity (mdeg)

 

 

 

 

‘45 I I I f I
190 200 210 220 230 240 250

Wavelength (nm)

 

Figure 5.10 (a) Increasing temperature ramp from 5-60°C and its associated pilus
stability observed at (A) 5°C, (B) 10°C, (C) 20°C, (D), 40°C, and (E) 60°C.

58

 

 

 

 

 

 

 

I 5 .1
i A o
I
: 3" -5 i
E -10 1 A
l V -15 4 B
i g -20 . c
| g. -25 —D
a ’30 _E
-35 1 J.
-40
'45 m I I I

 

190 200 210 220 230 240 250
Wavelength (nm)

 

 

Figure 5.10 (b) Decreasing temperature ramp from 60-5°C and the refolding of the
dissociated pilus at (A) 5°C, (B) 10°C, (C) 20°C, (D), 40°C, and (E) 60°C.

In agreement with the TM analysis of the unfolding of the pilus presented in section V.D,

the CD data Show a thermal unfolding transition at ~30°C and an irreversible unfolding

of the pilus.
V.F. Solid-state NMR (SSNMR)

For membrane proteins or large soluble proteins (>100 kDa), Magic Angle Spinning
(MAS) solid-state NMR provides structural biochemists with a means for obtaining

information about local secondary structure. Taking advantage of the bacterial

. . . . . l3 . . .
expressron host’s ability to incorporate a Slngle 1- C labeled amino ac1d during

59

induction has facilitated the extraction of such information. Here, MAS is utilized to
average out dipolar couplings and chemical shift anisotropies (CSAS) to generate nearly

isotropic chemical shifts.

l3C chemical shift (CS) measurements by cross-polarization (CP) MAS experiments

enable one to probe local conformation at specific residues using the known correlation

of CS3 with local 2° structure [43]. The lD-CPMAS experiment is a blanket approach to

obtaining 2° structure because of the inability to label only one amino acid and the

inherent line broadening associated with SSNMR. Simply stated, if the incorporated
amino acid has more than one occurrence within the sequence, the resulting peak will

represent a distribution of the CSs contributed by each residue. For nearly homologous

. . . . . . . 0
proteins, this experiment presents an inexpenswe and efficrent means of extracting 2

structure for significant portions of the sequences.

The rotational-echo double resonance (REDOR) filtering sequence [15] is a robust

technique that enables one to obtain the 2° structure of a single, unique residue. Through
backbone labeling of an existing unique pair of adjacent residues (e.g. 13C=O and the

adjacent l5N-H), one utilizes this filtering sequence to examine the 13C CS of only one

residue. This information can then be directly compared to computational predictions of

2° structure to examine validity of the model of interest.

60

In this work, the contributions of both CPMAS and REDOR were examined on a Varian

400 MHz solid-state NMR spectrometer for comparison to the 2° structure predictions

presented in section II.B. Pili formation was induced (as in section IV.A) for isotopically

labeled HrpA from both P. syringae and E. amylovora (as in sections III.C and III.D).

For CPMAS, HrpA from P. syringae was expressed with 13C=O labeled Ala, Gly, Met,
or Lys, and HrpA from E. amylovora was expressed with 13C=O labeled Leu, Ala, or Gly
(Figure 5.10). For REDOR, HrpA from P. syringae was expressed with l3C=O Leu and

15N Ile to establish labeling of the Leu29/Ile30 unique residue pair (Figure 5.10). 40 uL

of the viscous samples were pipetted into a 2/3 volume 4 mm Teflon rotor. A quasi-solid
was formed by slow-spinning (500 Hz) the rotor at -80°C in a 4 mm HCX probe. The

Spinning speed for the CPMAS and REDOR experiments was then ramped Slowly to 8

kHz and the 1H decoupling field was set at 100 kHz for the duration of acquisition. The

resulting data were processed and phased (Figures 5.11 (a-g) and 5.12).

61

 

62

 

 

 

 

 

 

I I I I I I I _1 I I I I I T— r I m 1 I l
190 185 180 175 170 165 160
p:-

(b)

63

 

I I I I I I I I I I I I I I I I I I I
190 185 180 175 170 165 160
m

(C)

64

 

190 aeo 170 160
m

(d)

65

 

 

 

1.90 180 170 160
pan

(9)

66

 

 

 

 

 

 

[-
175

 

”D

T

67

125

 

 

 

 

 

 

I I I I I I I I I I I I I I I I I I I
190 185 180 175 170 165 160
D”

(9)

Figures 5.11 (a-g) CPMAS signals for single isotopically labeled, truncated, wild-type
HrpA from E. amylovora (a-c) and P. syringae (d-g) pili. HrpA from E. amylovora was

labeled with 13C=O (a) Ala, (b) Gly, and (c) Leu, and HrpA from P. syringae was labeled
with l3C=o (d) Ala, (e) Gly, (f) Lys, and (g) Met.

68

 

 

 

 

I I 1 I I I I I I I I I I I I j I I I
190 185 180 175 ‘ 170 165 160
Dill

Figure 5.12 REDOR-filtered signal from the 1-13C Ile-ISN Leu unique pair in the
truncated, wild-type HrpA from P. syringae prepared pili.

69

Our isotopic labeling scheme exploited the homology of these two HrpA pilins and

enabled us to obtain complementary, local 2° structure information Spanning the majority

of the individual sequences (Figure 5.13). The CS analysis of these spectra was
correlated with the table of known carbonyl shifts [43]. The observed CSs Show the

HrpA from both strains to be predominantly a-helical at the sites of the above residues.

These results are consistent with the local 2° structure predictions for these residues and

the presented CD data (Figures 5.8-10).

70

- the A vH-A a +i-Ae
I

3 19 3O 19 ‘9

 

  
  
 

  
  
 

 

 

 

 

M N'MF131751‘.LTNLGIISRVA fem «aluminum; tutucos
rm H'JIFACLYSKLTIILGIISA'V”. VCGA' '.'-:7 'A‘N Tier-Is: I...5'1..CE.‘S
m 11""A1 L?5KI.!)ILCN:‘-A‘i=‘ vcca (RTVASPIATL or». L::1..cc::1.3
um 1("211'53- nummnnsave vucx I...“ 85720111110501. .555
m1 MVRI‘SSLISPTLTtlLGEIS'I‘iCZ sccalgiwrvsai .IASLQKIIILSGT..GIJCS
rm .‘WA.’ A'JL; SKLTHLIIMISA‘C VCGA .51Vh5NAS-GI.llZLSG! .623
em ""h”-“'KIT\I.I.N A‘." "IJGHKIJSHAAHA:"3:35.!3121.Gti- .-
W Mitt-'51? Saucacnwm Innaheca‘ntasmmw;N.I.At:1:'11<;::52..<
em HPAE‘PSZ.SSIic.‘~ti.Gl.’i.tl .256.21.];25313'1{ASAANNQSIWLAGICKCHS
In,” i use .......................
Ill! 'n.. . . . ........ s.
um .. .. ...tus.
M g ..................... s 3

 

 

+++A *5” + +5} fi+>+ t”. +i+ 05+)
:9 1

    

Figure 5.13 Local 2° structure information of multimeric HrpA by homology. For
HrpA from E. amylovora, the “*” and the “-“ indicate residues involved in (1) an a-helix
and (2) an (it-helix or B-sheet, respectively. For HrpA from P. syringae, the “+” indicates
residues involved in an a-helix. The “"” indicates residues involved in an a-helix that
occur at the same position in both sequences. Finally, the “>” indicates residues having
the same sequence position in both pilins, but are involved in (1) an a-helix or (2) an a-
helix or B-sheet in P. syringae or E. amylovora, respectively.

71

VI. Conclusions

This work addresses the following topics: (1) the structure of monomeric and multimeric
HrpA, (2) the physical interactions stabilizing the self-assembly of HrpA, (3) the
agreement between the computational predictions and experimental observations, (4)
methods for preparing stable multimers of uniform length, and (5) methods for disrupting
pili to help guide development of small-molecule inhibitors. Through the use of various
co-solvents and sundry hydrodynamic and spectroscopic methods, a series of
complementary sets of data were obtained for interpretation and comparison to the

computational model identified in section 11.

Taken together, the data show that HrpA is intrinsically disordered as a monomer under

native conditions but, upon self-assembly in vitro into the Hrp pilus, adopts stable 2°

structure that is predominantly a-helical. These results are consistent with both the
computational model of self-assembly and the biological hypothesis for infection
mediated by type III secretion systems. EM images of the Hrp pili assembled in vitro
suggest that they are stable multimers of uniform outer diameter that could function as a
conduit for Gram-negative bacterial effector proteins in vivo. HrpA appears to self-
assemble only at pH values below its isoelectric point (i.e. when the HrpA monomer is
positively charged). However, the self-assembly can be inhibited by lowering the pH far
enough; for example, the pili will form in vitro at pH 7, but not at pH 5 [23]. This
suggests that overcoming electrostatic repulsion is a key step in pilus formation. Since

HrpA assembles into a long hollow cylinder of uniformly positive charge density,

72

Manning-type condensation of anions [27] within the cylinder will possibly occur to
stabilize the electrostatic interactions. It is evident that hydrophobic interactions are also
important for association, since different anions can promote or inhibit pilus formation in

vitro according to the Hofmeister series [4].

 

73

VII. Works in Progress

We have shown that, in pilus formation, folding is coupled to the association of the pilin
subunits. In future work, we intend to investigate the kinetics of this process (and its
reversal) as well as performing further structural studies. Kinetics will be probed by
fluorescence spectroscopy targeting the chromophore on the C-terminal aromatic residue,
as described below. Structural studies will be approached by several methods. First, an
N-tenninal GFP-tagged HrpA construct is being prepared to be expressed and purified for
fluorescence microscopy and TEM studies. This will allow us to monitor self-assembly

optically and to determine the number of monomers per unit length within the pilus.

Secondly, proline mutations of conserved residues predicted to have a-helical 2°

structure will provide information about residues thought to be important in pilus
formation. Thirdly, X-ray crystallography has the potential to yield atomic-resolution,
3D structures for the wild-type, truncated HrpA and the co-crystallizable GFP-tagged
HrpA. Finally, multimers of both the wild-type, truncated HrpA and the GFP-tagged
HrpA will be investigated by electron fiber diffraction to obtain high-resolution, 3D

structural information with respect to HrpA within the Hrp pilus.

VII.A. Unfolding/Dissociation Kinetics of Multimeric HrpA from P. syringae

The truncated, wild-type HrpA from P. syringae has a single aromatic residue, a tyrosine,

located at its C-terminus. The rate of unfolding of the truncated, wild-type multimer of

HrpA from P.syringae will be determined by fluorescence spectroscopy. Through

74

monitoring the solvent quenching of the tyrosine fluorophore, the rate constant associated
with this first-order process will be determined. The unfolding will be prompted by the
presence of denaturants, either guanidine-HCI (Gdn-HCl) or urea, across the range of

concentrations O-6M. The Arrhenius behavior of the observed rate constant will also be

monitored to determine the activation energy (Ea) associated with the unfolding transition

of the multimer.

A simple thermal unfolding experiment was designed to determine whether the rate of
unfolding occurred on a time—scale measurable by our instrumentation. Through manual
control of the Shimadzu UV-17OO spectrophotometer with the Peltier cell holder set to
the observed unfolding transition temperature of 30°C (section V.D), preliminary kinetics
data were collected on a multimeric HrpA sample as prepared previously for TM analysis
in section V.D. The rate of unfolding was monitored as the change in absorbance at 287
nm with elapsed time. A plot of absorbance versus time (Figure 7.1) was generated and

fit with the first-order exponential decay function given by

A(t) - A0 exp(- 5;) (7)

where A0 is a pre-exponential factor, t is time, and 1: is the lifetime.

75

 

 

0.156
0.154 $
0.1521
0.15 1
0.148 1
0.1461
0.144 4
0.142 ~ 0
0.14 «
0.133 l O
0.136 I e e o o o

0.134 . , , -

Absorbance

 

 

 

 

Time (min)

 

 

Figure 7.1 F irst-order exponential decay of Hrp pilus unfolding

The lifetime derived from the fit was 59 seconds. This result shows that the previously
described fluorescence measurements are feasible, for the rate of unfolding of the HrpA

multimer occurs on a time-scale detectable by the instrument.

VILB. Preparation, Expression, Purification, and Fluorescence Microscopy of an

N-terminal GFP-tagged HrpA Construct

The lnvitrogen pRSETb expression vector [46] containing a (His)6 tag, an enterokinase
(EK) cleavage site, and the BamHI fragment of the green-enhanced S65T mutant GFP on
the N-terminus, was donated generously by the laboratory of Dr. S. Ferguson-Miller.
Using a Stratagene QuikChange site-directed mutagenesis kit, the C-terminal TAATAA

stop codon sequence was mutated to TAGCAA to enable expression of a subsequent C-

76

terminal protein insert, and the NdeI restriction site of the pET28b expression vector N-
terminal to the truncated, wild-type HrpA from P. syringae to the XhoI restriction site.
The HrpA insert will be cloned into the pRSETb expression vector using the XhoI and
HindIII restriction sites. This N-terminal GFP-tagged HrpA construct will be verified by

DNA sequencing.

The GFP-tagged HrpA construct will be expressed by BL21:).DE3 E. coli cells and
purified according to the protocol outlined in section III. The monomeric form of the
GFP-tagged HrpA complex is hypothesized to facilitate protein crystallization efforts.
The use of GFP as a co-crystallizing agent could yield a crystal whose diffraction pattern
is easily solved by molecular replacement of the GF P, a protein of known structure. In
vitro preparation of multimeric HrpA will be verified by TEM. The resulting multimeric
sample will be utilized by two methods. First, the sample will be examined by
fluorescence microscopy. This experiment has the potential for determination of the
number of HrpA monomers present within a stable multimer of specified length. And
second, this multimer might facilitate high-resolution, 3D structure determination of
HrpA, within the pilus, by fiber diffraction utilizing the oversampling method [30, 31], as

described below. GFP would serve as a molecular replacement tool during analysis.

VILC. Proline Mutations

Collectively, 2° structure predictions, TEM, CD, and SSNMR have provided evidence

that shows the existence of a-helical structure in HrpA when assembled in the pilus. It

77

was hypothesized that a disturbance in the a-helical structure of HrpA would cause a
dissociation of the pilus into its unstructured, monomeric subunits. It is well-understood
that prolines disrupt a-helices [38]. A series of proline mutants were selected for several
highly conserved and predicted a-helical residues, as well as a residue that rests in the
random coil structure of the turn of the hairpin fold, in the truncated, wild-type HrpA
from P. syringae. A54P and A64P reside near the N—terminus, D84P represents the turn
of the predicted hairpin fold, and I89P and IlOlP reside near the C-terminus. These
mutants were generated using a Stratagene QuikChange Site-directed Mutagenesis kit

and verified by DNA sequencing.

Mutants 189P and I 101P were expressed in BL21z}.DE3 E. coli cells and purified
according to the protocol outlined in Chapter 3. Samples containing either 100% mutant
HrpA or a 1:1 molar equivalent mixture of the mutant and wild-type were prepared for
each mutant, according to the protocol for in vitro pilus preparation presented in Chapter
4. The samples were plated on copper grids and negative stained with 2% aqueous
phosphotungstic acid for TEM. Conventional light/dark field images by electron

scattering were obtained (Figures 7.2 a-d).

78

 

Figures 7.2 (a-d) TEM images of (a) I89P mutant HrpA, (b) 1:1 molar ratio of I89P
mutant and truncated, wild-type HrpA, (c) IlOlP mutant HrpA, and (d) 1:1 molar ratio of

llOlP mutant and truncated, wild-type HrpA in the presence of 50 mM NaH2P04 buffer
pH 5.5 with 300 mM NaF.

79

As predicted, the samples containing pure I89P and IlOlP exhibited a negative result for
pilus formation. The samples containing these mutants doped with wild-type HrpA were
not so consistent. The sample of I89P doped with wild-type HrpA showed a positive
result for pilus formation, as the formation of pili was not thwarted. By contrast, the
sample of IlOlP doped with wild-type HrpA exhibited a dominant negative result for

pilus formation.

One interpretation of these results is that the IlOlP mutant is able to associate with the
wild-type pilus and inhibit its growth, however, the 189P mutant is neither able to self-
assemble into the pilus nor associate with the wild-type pilus. To test the validity of this
interpretation, it is necessary to check whether the IlOlP HrpA can be incorporated into
the wild-type pilus. Preparation of samples containing both IlOlP HrpA and the GFP-
tagged HrpA will hopefully provide insight into this process. The will enable a more
definitive interpretation of the above TEM results. Applying the combination of TEM
and fluorescence microscopy will show whether the wild-type, in fact, associates with the

mutant.

Exploring the remaining mutants, in the manner described above, might provide some
insight as to which residues are crucial to pilus formation and stability. The results of
these experiments have the potential to guide development of small-molecule inhibitors

that might render this TTSA ineffective.

80

VII.D. Atomic-resolution Structural Studies: X-ray Crystallography and Electron Fiber

Diffraction

An atomic-resolution structure of folded HrpA would be useful in determining the
residues critical for folding and association. Several structure determination methods and

their associated preliminary work will be discussed in the subsequent sections.

VII.D.i. Crystallization of Monomeric HrpA and X-ray Diffraction

Protein crystallization and X-ray diffraction has the potential to yield atomic-resolution
three-dimensional structures of soluble proteins. Since crystallization requires that the
protein of interest be well-structured and folded, the unstructured HrpA monomer seems
unlikely to crystallize. It is not, however, beyond the realm of possibilities for the
crystallization of such a protein to occur. There could exist a set of conditions that would

induce HrpA to adopt some stable and folded intermediate state that would crystallize.

To pursue this possibility, we undertook preliminary crystallization trials of the truncated,

wild-type HrpA from P. syringae. The (His)6 tag cleaved form of the truncated, wild-
type HrpA was purified and concentrated to 10 mg/mL in 15 mM Tris-HCl, 100 mM
NaCl at pH 7.5 for crystallization condition screening. A Douglas Instruments Oryx-4

crystallization robot coupled to an XYZV Plateloader was used to set up sitting-drop

vapor diffusion across the conditions contained within the Synergy screen [25] and

81

Hampton Screens I and 11, provided by Hampton Research, Inc. 2 [LL drops, containing
50% protein stock solution, were prepared to sit on a pedestal in a closed, vapor diffusion
chamber having a well-solution of 75 [LL of the respective crystallizing agents. The
conditions were incubated at 5°C to assist in stabilizing the formation of a-helices. The
drops were visualized with a Leica M2 12 stereoscope and possible crystal hits were
recorded. Afier two days, conditions containing (1) 200 mM ammonium sulfate and 30%
PEG 4000 (Figure 7.3) and (2) 200 mM ammonium sulfate and 30% PEG 8000 (not
shown) yielded similar needle-like crystals that stained positively as protein crystals with

the application of Izit dye.

 

Figure 7.3 Photograph of needle-like crystals of (His)6 tag cleaved, truncated, wild-type
HrpA from P. syringae obtained after two days of incubation at 5°C in 200 mM
ammonium sulfate and 30% PEG 4000.

82

mm- myt 5331!:W

These conditions showed very rapid, clustered crystal growth from a large number of
nucleation sites. Current work focuses on optimizing these conditions to obtain

diffraction quality crystals for structural analysis.

VILE. Fiber Diffraction Studies of Multimeric HrpA

X-ray crystallography can provide atomic-resolution 3D structures of biomolecules only
if they are crystallized. For non-crystalline macromolecules or biomolecular assemblies
(e.g. fibrils), 3D structure determination is often an arduous task that requires many
spectroscopic tools. J. Miao et al. have shown that phase retrieval from oversampled
diffraction patterns of selected noncrystalline biological systems is feasible [30]. This
method is believed to be able to open the door for high-resolution, 3D structure
determination of complex and noncrystalline biological specimens [31]. In this work, we
wish to examine the feasibility of introducing either X-ray or electron beam fiber
diffraction as a means for determining the 3D high-resolution structure of the Hrp pilus as

prepared in vitro.

We hypothesized that either pili or noncrystalline pilus-like intermediates could be
prepared in a condition similar to the monomeric crystallization conditions presented in
Figure 7.3. 100 ML of 20 mg/mL uncleaved monomeric HrpA were exchanged from a
pH 7.5 buffer containing 15 mM Tris-HCl and 100 mM NaCl to an aqueous solution of
200 mM ammonium sulfate. Buffer exchange was accomplished in two repetitive stages

of 50-fold dilution in the phosphate buffer followed by concentration to the initial volume

83

., I‘m-It» 1' 1 '9? Imam]:

Wm ”‘5;

in 5 mL Amicon 5000 MWCO centrifuge tubes. The resulting 50 [LL were transferred to
a 0.5 mL Eppendorf tube. 72 [LL of 50% (w/v) PEG 4000 was pipetted to the HrpA
solution to yield a final concentration of 30% (w/v). The resulting mixture was
transferred to a 0.5 mL Eppendorf tube and frozen for 24 hours at -20°C. Upon thawing
to 25°C, the sample developed an inherent viscosity. A 10% dilution was prepared for

TEM according to the above freeze thaw cycle.

The unstained sample was plated on a copper grid for TEM analysis of its contents.

Conventional light/dark field images by electron scattering were obtained (Figure 7.4).

 

Figure 7.4 TEM image for truncated, wild-type HrpA from P. syringae in the presence
of 200 mM ammonium sulfate and 30% PEG 4000.

The sample area presented in Figure 7.4 was subjected to the incident electron beam. A

diffraction pattern was collected prior to ablation of the sample area (Figure 7.5).

 

Figure 7.5 Electron diffraction image for the TEM sample area in Figure 7.4

This preliminary result shows that obtaining structural information from electron and/or
X-ray fiber diffraction is possible. Current work focuses on extracting phase information
and diffraction coefficients from this diffraction pattern to eventually index and

characterize HrpA.

85

Appendix

86

Appendix. Buffers and Stock Solutions

NiA buffer:

1 X Thrombin Cleavage buffer:

 

 

50 mM NaH2P04

20 mM Tris-HCl

 

 

300 mM NaCl

150 mM NaCl

 

 

20 mM Irnidazole

2.5 mM CaC12

 

 

5 mM B-Mercaptoethanol

 

pH 8.4

 

 

2 EDTA-free Protease Inhibitor tablets

 

 

pH 8.0

 

 

NiA + 6 M Urea buffer:

Sterile Tris-EDTA (TE) buffer:

 

10 mM Tris-HCl

 

 

50 mM NaH2PO4

1 mM EDTA

 

 

pH 8.0

 

 

300 mM NaCl

 

20 mM Irnidazole

 

5 mM B-Mercaptoethanol

 

6 M Urea

5 X M9 Salts

 

 

2 EDTA-free Protease Inhibitor tablets

 

 

pH 8.0

30 g NaH2PO4

 

 

 

NiA + 600 mM NaCl buffer:

[53301904

 

5 g NH4C1

 

2.5 g NaCl

 

 

50 mM NaH2P04

15 mg CaClz

 

 

Autoclave to sterilize

 

 

300 mM NaCl

 

20 mM Irnidazole

 

5 mM B-Mercaptoethanol

 

600 mM NaCl

1 X M9 Minimal Media

 

 

2 EDTA-free Protease Inhibitor tablets

200 mL 5 X M9 Salts

 

 

 

pH 8.0

 

 

NiB buffer:

1 mM Sterile MgSO4

 

0.2% (v/v) Sterile Glucose

 

Antibiotic

 

 

Dilute to 1 L with Sterile H20

 

 

50 mM NaH2P04

 

300 mM NaCl

 

300 mM Irnidazole

 

5 mM B-Mercaptoethanol

 

2 EDTA-free Protease Inhibitor tablets

 

 

pH 8.0

 

 

 

 

 

 

 

10.

11.

12.

BIBLIOGRAPHY

Adamczak, R., Porollo, A., and Meller, J. (2005) Combining prediction of
secondary structure and solvent accessibility in proteins. Proteins: Struct. F unct.
Genet. 59, 467-475.

Altschul, S.F., Gish, W., Miller, W., Myers, E.W., and Lipman, DJ. (1990) Basic
Local Alignment Search Tool. J. Mol. Biol. 215, 403-410.

Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J .H., Zhang, Z. Miller, W., and
Lipman, DJ. (1997) Gapped BLAST and PSI-BLAST: a new generation of protein
database search algorithms. Nucl. Acids Res. 25, 3389-3402.

Baldwin, R.L. (1996) How Hofmeister ion interactions affect protein stability.
Biophys. J. 71, 2056-2063.

Brown, I.R., Mansfield, J .W., Taira, S., Roine, E., and Romantschuk, M. (2001)
Immunocytochemical localization of HrpA and HrpZ supports a role for the Hrp
pilus in the transfer of effector proteins from Pseudomonas syringae pv. tomato
across the host plant cell wall. Mol. Plant Microbe Interact. 14(3), 394-404.

Buttner, D. and Bonas, U. (2002) Port of entry — the type III secretion translocon.
TRENDS in Microbiol. 10(4), 186-192.

Cantor, CR. and Schimmel, P.R. Biophysical Chemistry. Part 1]. Techniques for the
Study of Biological Structure and Function. W.H. Freeman, New York, 1980.

Cornelis, GR. (2000) Type III secretion: a bacterial device for close combat with
cells of their eukaryotic host. Phil. Trans. R. Soc. Lond. B. 355, 681-693.

Dickerson, RE. and Drew, HR. (1981) Structure of a B-DNA dodecamer: II
Influence of base sequence on helix structure. .1. Mol. Biol. 149, 761-786.

Emerson, S.U., Tokuyasu, K., and Simon, M. (1970) Bacterial flagella: polarity of
elongation. Science. 169, 190-192.

Espinosa, A., and Alfano, J .R. (2004) Disabling surveillance: bacterial type III
secretion system effectors that suppress innate immunity. Cell. Microbiol. 6(11),
1027-1040.

Ghosh, P. (2004) Process of protein transport by the type III secretion system.
Microbiol. Mol. Biol. Rev. 68(4), 771-795.

88

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

Goldberg, RJ. (1953) Sedimentation in the ultracentrifuge. J. Phys. Chem. 57, 194-
202.

Gouet, P., Robert, X., and Courcelle, E. (2003) ESPript/ENDscript: extracting and
rendering seqeuence and 3D information from atomic structures of proteins. Nucl.
Acids Res. 31, 3320-3323.

Gullion, T. and Schaefer, J. (1989) Rotational-echo double-resonance NMR. J.
Mag. Reson. 81, 196-200.

He, S.Y., Nomura, K., and Whitman, TS. (2004) Type 111 protein secretion
mechanism in mammalian and plant pathogens. Biochimica et Biophysica Acta.
1694, 181-206.

Jin, Q. and He, S.Y. (2001) Role of the Hrp pilus in type 111 protein secretion in
Pseudomonas syringae. Science. 294, 2556-2558.

Jones, D.T., (1999) Protein secondary structure prediCtion based on position-specific
scoring matrices. J. Mol. Biol. 292, 195-202.

Karplus, K., Karchin, R., Barrett, C., Tu, S., Cline, M., Diekhans, M., Grate, L.,
Casper, J ., and Hughey, R. (2001) What is the value added by humen intervention
in protein structure prediction? Proteins: Struct. F unct. Genet. Suppl. 5., 86-91.

Kelly, S.M., Jess, T.J., and Price, NC. (2005) How to study proteins by circular
dichroism. Biochimica. et Biophysica Acta. 1751, 119-139.

Koh, I.Y.Y., Eyrich, V.A., Marti-Renom, M.A., Przybylski, D., Madhusudhan, M.S.,
Bswar, N., Grana, 0., Pazos, F ., Valencia, A., Sali, A., and Rost, B. (2003) EVA:
evaluation of protein structure prediction servers. Nucl. Acids Res. 31, 3311-3315.

Lebowitz, J ., Lewis, MS, and Schuck, P. (2002) Modern analytical
ultracentrifugation in protein science: a tutorial review. Protein Sci. 11, 2067-2079.

Lee, Y.H., Kolade, 0.0., Nomura, K., Arvidson, D.N., and He. S.Y. (2005) Use of
dominant-negative HrpA mutants to dissect Hrp pilus assembly and type III secretion
in Pseudomonas syringae pv. tomato. J. Biol. Chem. 280(22), 21409-21417.

Li, C., Brown, 1., Mansfield, J ., Stevens, 0, Boureau, T., Romantschuk, M., and
Taira, S. (2002) The Hrp pilus of Pseudomonas syringae elongates from its tip and
acts as a conduit for translocation of the effector protein HrpZ. EMBO J. 21(8),
1909-1915.

Majeed, S., Ofek, G., Belachew, A., Huang, 0, Zhou, T., and Kwong, PD. (2003)

Enhancing protein crystallization through precipitant synergy. Structure. 11, 1061 -
1070.

89

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

Mandal, PK. and Maj umdar, A. (2004) A comprehensive discussion of HSQC and
HMQC pulse sequences. Concepts Magn. Reson. Part A, 20A(l) 1—23.

Manning, GS. (1969) Limiting laws and counterion condensation in polyelectrolyte
solutions I. Colligative properties. J. Chem. Phys. 51(3), 924-933.

Marlovits, T.C., Kubori, T., Sukhan, A., Thomas, D.R., Galan, J .E., and Unger, V.M.
(2004) Structural insights inot the assembly of the type III secretion needle complex.
Science. 306, 1040- 1 042.

McGuffin, L.J., Bryson, K., and Jones, D.T. (2000) The PSIPRED protein structure
prediction server. Bioinformatics. 16, 404-405.

Miao, J ., Ishikawa, T., Anderson, E.H., and Hodgson, K.0. (2003) Phase retrieval
of diffraction patterns from noncrystalline samples using the oversampling method.
Phys. Rev. B. 67, 174104-(1-6).

Miao, J ., Kirz, J ., and Sayre, D. (2000) The oversampling phasing method. Acta
Cryst. D56,1312-1315.

Mimori-Kiyosue, Y., Yamashita, I., F ujiyoshi, Y., Yamaguchi, S., and Namba, K.
(1998) Role of the outermost subdomain of salmonella flagellin in the filament
structure revealed by electron cryomicroscopy. J Mol. Biol. 284, 521-530.

Mota, L.J., Sorg, 1., and Cornelis, GR. (2005) Type III secretion: the bacteria-
eukaryotic cell express. FEMS Microbiol. Lett. 252, 1-10.

Notredame, C., Higgins, D.G., and Heringa, J. (2000) T-Coffee: a novel method for
fast and accurate multiple sequence alignment. J. Mol. Biol. 302, 205-217.

Privalov, PL. and Gill, 8.]. (1998) Stability of protein structure and hydrophobic
interaction. Adv. Protein. Chem. 39, 191-234.

Roine, E., Saarinen, J ., Kalkkinen, N., and Romantschuk, M. (1997) Purified HrpA
of Pseudomonas syringae pv. tomato DC3000 reassembles into pili. FEBS. Lett.
417, 168-172.

Sarnatey, F.A., Matsunami, H., Imada, K., Nagashima, S., Shaikh, T.R., Thomas,
D.R., Chen, J .Z., DeRosier, D.J., Kitao, A., and Namba, K. (2004) Structure of the
bacterial flagellar hook and implication for the molecular universal joint mechanism.
Nature. 431, 1062-1068.

Schulman, BA. and Kim, PS. (1996) Proline scanning mutagenesis of a molten

globule reveals noncooperative formation of a protein’s overall topology. Nat.
Struct. Biol. 3, 682-687.

90

39.

40.

41.

42.

43.

44.

45.

46.

Shiraki, K., Nishikawa, K., and Goto, Y. (1995) Trifluoroethanol-induced
stabilization of the a-helical structure of B-Lactoglobulin: implication for non-
hierarchical protein folding. J. Mol. Biol. 245, 180-194.

Short Protocols in Molecular Biology, vol 1. Edited by Ausubel, F.M., Brent, R.,
Kingston, RE, Moore, D.D., Seidman, J .G., Smith, IA. and Struhl, K., John Wiley
and Sons, 2002, p.1-2.

Sigle, W. (2005) Analytical transmission electron microscopy. Annu. Rev. Mater.
Res. 35, 239-314.

Tampakaki, A.P., Fadouloglou, V.E., Gazi, A.D., Panopoulos, N.J., and Kokkinidis,
M. (2004) Conserved features of type III secretion. Cell. Microbiol. 6(9), 805-816.

Wishart, D.S., Sykes, B.D., and Richards, F .M. (1991) Relationship between
nuclear magnetic resonance chemical shift and protein secondary structure. .1. Mol.
Biol. 222, 311-333.

Yip, GK. and Strynadka, NC]. (2006) New structural insights into the bacterial
type III secretion system. TRENDS in Biochemical Sciences. 31(4), 223-230.

http://www.iiovagencom (last accessed June 2006)

http://www.invitrogen.com (last accessed June 2006)

91

   

u11111111111IIMEIIIIIHILIIIIII