The Plant Cell, Vol. 11, 1081–1092, June 1999, www.plantcell.

org © 1999 American Society of Plant Physiologists

Structure of a Plant Cell Wall Fragment Complexed to Pectate

Lyase C

Robert D. Scavetta,a,1 Steven R. Herron,a Arland T. Hotchkiss,b Nobuhiro Kita,a,2 Noel T. Keen,c
Jacques A. E. Benen,d Harry C. M. Kester,d Jaap Visser,d and Frances Jurnaka,3
a Department of Physiology and Biophysics, 346-D Med Sci I, University of California, Irvine, California 92697-4560
b U.S.Department of Agriculture–Agricultural Research Service, Eastern Regional Research Center, 600 E. Mermaid Lane,
Wyndmoor, Pennsylvania 19038-8598
c Department of Plant Pathology, University of California, Riverside, California 92521
d Department of Molecular Genetics of Industrial Microorganisms, Wageningen Agricultural University, Dreijenlaan 2, 6703

HA Wageningen, The Netherlands

The three-dimensional structure of a complex between the pectate lyase C (PelC) R218K mutant and a plant cell wall
fragment has been determined by x-ray diffraction techniques to a resolution of 2.2 Å and refined to a crystallographic
R factor of 18.6%. The oligosaccharide substrate, a-D-GalpA-([1→4]-a-D-GalpA)3-(1→4)-D-GalpA, is composed of five
galacturonopyranose units (D-GalpA) linked by a-(1→4) glycosidic bonds. PelC is secreted by the plant pathogen Er-
winia chrysanthemi and degrades the pectate component of plant cell walls in soft rot diseases. The substrate has
been trapped in crystals by using the inactive R218K mutant. Four of the five saccharide units of the substrate are well
ordered and represent an atomic view of the pectate component in plant cell walls. The conformation of the pectate
fragment is a mix of 21 and 31 right-handed helices. The substrate binds in a cleft, interacting primarily with positively
charged groups: either lysine or arginine amino acids on PelC or the four Ca21 ions found in the complex. The observed
protein–oligosaccharide interactions provide a functional explanation for many of the invariant and conserved amino
acids in the pectate lyase family of proteins. Because the R218K PelC–galacturonopentaose complex represents an in-
termediate in the reaction pathway, the structure also reveals important details regarding the enzymatic mechanism.
Notably, the results suggest that an arginine, which is invariant in the pectate lyase superfamily, is the amino acid that
initiates proton abstraction during the b elimination cleavage of polygalacturonic acid.

INTRODUCTION

Pectate lyases are depolymerizing enzymes that degrade five members of the superfamily have been determined and
plant cell walls, causing tissue maceration and death. The include Erwinia chrysanthemi pectate lyase C (PelC) (Yoder
enzymes normally are secreted by phytopathogenic organ- et al., 1993; Yoder and Jurnak, 1995), E. chrysanthemi pec-
isms and are known to be the primary virulence agents in tate lyase E (PelE) (Lietzke et al., 1994), Bacillus subtilis pec-
soft rot diseases caused by Erwinia spp (Collmer and Keen, tate lyase (B. subtilis Pel) (Pickersgill et al., 1994), Aspergillus
1986; Kotoujansky, 1987; Barras et al., 1994). In the latter niger pectin lyase A (PLA) (Mayans et al., 1997), and A. niger
organisms, the enzymes exist as multiple, independently pectin lyase B (PLB) (Vitali et al., 1998). All share a similar
regulated isozymes that share amino acid sequence identity but an unusual structural motif, termed the parallel b helix, in
ranging from 27 to 80%. which the b strands are folded into a large, right-handed
Pectate lyases share sequence similarities with fungal coil. The enzyme structures differ in the size and conforma-
pectin lyases, plant pollen proteins, and plant style proteins tion of the loops that protrude from the parallel b helix core.
(Henrissat et al., 1996). The three-dimensional structures of As deduced from sequence similarity and site-directed mu-
tagenesis studies, the protruding loops on one side of the
parallel b helix form the pectolytic active site (Kita et al.,
1996). The structural differences of the loops are believed to
be related to subtle differences in the enzymatic and macer-
1 Current address: 6803 South Ivy Way, Inglewood, CO 80112.
2 Current
ation properties of the proteins.
address: Kanagawa Institute of Agricultural Science, 1617
Pectate lyases catalyze the cleavage of pectate, the de-
Kamikisawa, Hiratsuka, Kanagawa 259-12, Japan.
3 To whom correspondence should be addressed. E-mail jurnak@ esterified product of pectin, which is the major component that
uci.edu; fax 949-824-8540. maintains the structural integrity of cell walls in higher plants
1082 The Plant Cell

(Carpita and Gibeaut, 1993). The pectate backbone is com- The results of a recent nuclear magnetic resonance study
posed of blocks of polygalacturonic acid (PGA), which is a suggest that the Ca21–PGA complex in the plant cell wall is
helical homopolymer of D-galacturonic acid (GalpA) units much more complex than the simple eggbox model. This
linked by a-(1→4) glycosidic bonds. The blocks of PGA are complex contains both 21 and 31 helices of PGA as well as
separated by stretches in which (1 →2)-a-L-rhamnose resi- intermediate conformational states (Jarvis and Apperley,
dues alternate with GalpA (McNaught, 1997). Blocks of PGA 1995). Nuclear magnetic resonance, molecular modeling,
may contain as many as 200 Gal pA units and span 100 nm and molecular dynamic analyses of pectic disaccharides
(Thibault et al., 1993). Cations are necessary to neutralize PGA and trisaccharides also have reported that PGA has both 31
in solution and, as a consequence, influence its structure. and 21 helical conformations (Hricovini et al., 1991; DiNola et
In the presence of Ca21, PGA assumes a 21 helical confor- al., 1994; Gouvion et al., 1994). Disaccharide hydration and
mation in dilute polymer concentrations (Morris et al., 1982; sodium salt formation may shift the predicted PGA helical
Powell et al., 1982) and a 31 helix at high concentrations in conformation from 31 to 21 (DiNola et al., 1994; Gouvion et
either a gel or solid form (Walkinshaw and Arnott, 1981a, al., 1994; Catoire et al., 1997).
1981b). Because the PGA concentration in the plant cell wall All proteins in the pectate lyase superfamily are believed
upon demethylation of pectin lies near the critical conforma- to share a similar enzymatic mechanism, but the catalytic
tional transition point, considerable speculation exists as to roles of the amino acids in the active site region have not
the in situ structure of PGA. A popular view is the “eggbox been identified. For reasons of technical convenience, re-
model” in which Ca21 ions cross-link the uronic acid moi- cent studies have focused on PelC. The enzyme randomly
eties of neighboring antiparallel chains of PGA together. Al- cleaves PGA by a b elimination mechanism, generating pri-
though the eggbox model generally is depicted with PGA in marily a trimer end product with a 4,5-unsaturated bond in
a right-handed 21 helical conformation, the original literature the galacturonosyl residue ( a-L-4-eno-threohexosylpyrano-
suggests that cross-linking between Ca 21 and PGA in a syluronic acid [a-L-4-en-thrHexpA]) at the nonreducing end
right- or left-handed 31 helical conformation is feasible as (Preston et al., 1992). PelC has an in vitro pH optimum of 9.5
well (Grant et al., 1973; Kohn, 1975). and requires Ca21 for pectolytic activity. Structural studies

Figure 1. Stereoview of the Ca21 Ions and TetraGalpA Superimposed upon the Simulated Annealed OMIT Electron Density Map of the PelC
R218K–Substrate Complex Contoured at 1.0s.
The Ca21 ions are represented by yellow spheres. TetraGalpA as well as the interacting amino acids are represented by rods by using the Inter-
national Union of Pure and Applied Chemistry coloring code: carbon atoms are gray; oxygen atoms, red; and nitrogen atoms, blue. The R218K
backbone is represented by green ribbons. Individual amino acids that are shown are labeled at the a-carbon.
 PelC Complexed to Plant Cell Wall Fragment 1083

Figure 2. Stereoview of the PelC R218K–(Ca21)3–4–PentaGalpA Complex.

The view and color scheme are the same as given in Figure 1, except that the tetraGalpA substrate is illustrated in cyan, the entire protein back-
bone is shown, and the individual amino acids are not labeled. The disulfide bonds are illustrated as yellow rods.

have shown that Ca21 is bound to the enzyme at a location binding. In addition, the results provide tentative identifica-
that was first suggested in a PelC–Lu 31 complex (Yoder et tion of the amino acid that initiates proton abstraction.
al., 1993) and later confirmed by structural studies of a B.
subtilis Pel–Ca21 complex (Pickersgill et al., 1994). The role
of Ca21 has not been established. In the b elimination reac-
tion, the reaction is initiated by proton abstraction from C-5 RESULTS
of the galacturonosyl residue on the reducing side of the
glycosidic scissile bond. The group or groups that initiate
proton abstraction and transfer the proton to the glycosidic Conformation of the PentaGalpA Substrate
oxygen have not been identified. Potential candidates in-
clude two invariant amino acids in the superfamily, Asp-131 Four of the five Gal pA units of the substrate used in the
and Arg-218 in PelC nomenclature, as well as four amino crystal diffusion experiment are visible as strong, well-
acids, Glu-166, Asp-170, Lys-190, and Arg-223, which are ordered electron density in difference Fourier maps, as
invariant within the pectate lyase subfamily (Henrissat et al., shown in Figure 1. As shown in Figure 2, the well-ordered
1996). Site-specific mutations at the latter PelC positions GalpA units interact with PelC in a groove encompassing
abolish pectolytic as well as maceration activity (Kita et al., the previously identified Ca21 binding site on the protein,
1996). Notably, the pectolytic region is devoid of conserved now termed the 4Ca21 site. The orientation of the tetraGalpA
histidine, serine, or tyrosine residues, which frequently are fragment is unambiguous. The reducing end, Gal pA1, is lo-
implicated in b elimination enzymatic mechanisms with a cated at the protein–solvent border, and the nonreducing
lower pH optimum. In this study, we have taken advantage end, GalpA4, lies near 4Ca21. Additional electron density,
of the impaired catalytic properties of one PelC mutant, corresponding to a fifth GalpA unit, is found at the nonre-
R218K, to form a stable substrate–enzyme complex that ducing end of the tetraGalpA fragment but is disordered and
could be studied by x-ray diffraction techniques. The results cannot be modeled.
provide an atomic view of a pectate fragment, a-D-GalpA-([1 Each of the well-ordered Gal pA rings refines to the con-
→4]-a-D-GalpA)3-(1→4)-D-GalpA (pentaGalpA), and the iden- ventional chair conformation, with all bond distances and
tification of the key amino acids involved in oligosaccharide angles consistent with single bonds. The pectate fragment
1084 The Plant Cell

131, and one carboxyl oxygen from each of Asp-129, Glu-

Table 1. Bond Angles (t) and Torsional Rotations (f and c)
about Glycosidic Bonds in the Refined Structures of the 166, and Asp-170. In the R218K complex with pentaGalpA,
PelC–(Ca21)3–4–PentaGalpA Complexesa the equivalent 4Ca21 coordinates to the same groups, with a
single exception—a carboxyl oxygen from GalpA4 replaces
Glycosidic Bond tb fc cd one of the water molecules. In addition to 4Ca21, three addi-
31 Helixe 117.08 80.08 89.08 tional Ca21 ions have been identified. Two of the additional
21 Helixf 117.08 80.08 161.08 Ca21 ions, 2Ca21 and 3Ca21, are fully occupied, and the
GalpA1–GalpA2 117.08 54.08 90.08 third, 1Ca21, has a partial occupancy of z50%. Each Ca21
GalpA2–GalpA3 116.68 117.08 157.08 ion bridges the carboxyl group of each GalpA unit to the
GalpA3–GalpA4 120.48 73.08 51.08 protein. In addition, 2Ca21 and 3Ca21 link the uronic acid
a Torsional angles were determined by looking from the nonreducing moieties of GalpA2, GalpA3, and GalpA4. The coordination
end side (with prime) down the bond of interest to the reducing end around each Ca21 ion is listed in Table 3. The observed Ca21
side and determining the angle of rotation created from the planes of positions are very different from the interstrand Ca 21 ions
O-59–C-19–O-4 and C-19–O-4–C-4 for f and C-19–O-4–C-4 and
O-4–C-4–C-5 for c. A cis configuration is taken to be 08, and a trans
configuration is taken to be 1808. A negative sign is a rotation from
cis in a counterclockwise direction, and a positive sign is a rotation
Table 2. Atomic Distances of 3.0 Å or Less between the Oxygen
from cis in a clockwise direction.
b t, the C –O–C 9 bond angle. Atoms of TetraGalpA and Amino Acids, Ca21 Ions, or
1 4
c f, the torsional rotation about the C –O bond. Water Molecules
1
d c, the torsional rotation about the O–C 9 bond.
4 GalpA Atomsa Interacting Atomsb Distance (Å)
e The values for the 3 helix are those determined by the program O
1
(Jones et al., 1991) from the model published by Walkinshaw and GalpA1
Arnott (1981b). Ring interaction Tyr-268
O-6A GA1Wat1 2.9
f The values for the 2 helix are those determined by the program O
1
O-6B 1Ca21 2.9
for a galacturonic acid model constructed using the parameters for
alginic acid by Atkins et al. (1973). GalpA2
O-2 Asp-162 O-d2 2.7
O-2 GA2Wat1 2.7
O-3 2Ca21 2.6
O-5 1Ca21 2.9
O-6A 1Ca21 2.6
folds into an unbent right-handed helical conformation, with
O-6A Arg-245 NH-1 2.8
the observed helical angles compared with the idealized 2 1
O-6B Arg-245 NH-2 3.0
and 31 helices in Table 1. Two of the three glycosidic bonds O-6B GA2Wat2 2.8
have c torsional angles that approximate the 31 helix observed GalpA3
in fiber diffraction studies of PGA–Ca 21 gels (Walkinshaw O-2 Arg-223 NH-2 2.9
and Arnott, 1981a, 1981b). One of the glycosidic bonds, be- O-3 Ser-196 O 2.8
tween GalpA2 and GalpA3, has a c torsional angle similar to O-3 Arg-223 NH-1 2.9
O-5 2Ca21 2.5
a 21 helix. Consequently, the overall appearance of the pec-
O-5 2CaWat2 3.0
tate fragment conformation is of a 31 helix, but with the mid-
dle segment distorted into 2 1 helix, as illustrated in Figures O-6A Lys-190 N-z 2.8
O-6A 2Ca21 2.3
3A and 3B. Given the greater number of GalpA3 contacts, as 2CaWat2
O-6A 2.9
listed in Table 2, the deviation from the 31 helical conforma- 3Ca21
O-6B 2.4
tion is probably a result of specific interactions with the en- O-6B 3CaWat3 2.8
zyme. The deviation is not likely due to other effects, such GalpA4
as pectate concentration, hydration, or cation type, which O-2 2CaWat2 2.9
are postulated to cause the transition between 21 and 31 he- O-3 GA4Wat1 2.8
lical conformations of pectate. If the R218K–(Ca 21)3–4– O-4 GA4Wat2 2.7
GalpA5 structures are representative of interactions that oc- O-4 Ser-308 O 2.8
O-5 3Ca21 2.4
cur within the plant cell wall, then endogenous proteins also
O-6A 4Ca21 2.5
are likely to distort the PGA conformation from any helical
O-6A 3Ca21 2.4
states observed under in vitro conditions. GA4Wat2
O-6A 2.9
O-6B 4CaWat1 2.6
a The positions of the atoms are indicated in Figure 4.
Coordination of Ca21 Ions
b xWaty refers to the Y water molecule associated with the X GalpA
In wild-type PelC, a Ca21 ion coordinates to seven ligands, in- unit. zCa21 refers to the Z position of the Ca21 ion as defined in Fig-
ure 4.
cluding two water molecules, both carboxyl oxygens of Asp-
 PelC Complexed to Plant Cell Wall Fragment 1085

Figure 3. Stereoview of the TetraGalpA Structure Superimposed upon Modeled Right-Hand 21 and 31 OligoGalpA Helices.
The tetraGalpA structure determined at 2.2 Å is shown in both (A) and (B) as cyan rods.
(A) The modeled 31 oligoGalpA helix (Walkinshaw and Arnott, 1981a) is shown in red.
(B) The modeled 21 oligoGalpA helix (Atkins et al., 1973) is shown in yellow.

postulated to link PGA helices together (Walkinshaw and gen bonds with the C-2 and C-3 hydroxyl groups of GalpA3,
Arnott, 1981a, 1981b; Morris et al., 1982; Powell et al., 1982). the orientation of which partially defines the galactose
In the present structure, the Ca 21 ions link not only the oli- epimer. The C-3 hydroxyl group also forms a hydrogen bond
gosaccharide to the protein but also adjacent uronic acid with a nonconserved Ser-196. In GalpA2, the C-2 hydroxyl
moieties within a single pectate strand. interacts with Asp-162, and in Gal pA1, the ring forms a
stacking interaction with Tyr-268. Both amino acids are con-
served but only in the PelC subfamily. In addition to interac-
tions with the protein and Ca 21 ions, the tetraGalpA
Protein–Ca21–TetraGalpA Interactions segment is highly solvated, forming many hydrogen bonds
with water molecules that increase in frequency from GalpA1
The protein–Ca21–tetraGalpA interactions are represented in to GalpA4. Collectively, the observed protein–tetraGalpA in-
Figure 4, and all relevant interatomic distances are summa- teractions provide a functional role for all invariant and con-
rized in Table 2. Electrostatic interactions dominate, with the served amino acids in the pectolytic region of the pectate
negatively charged uronic acid moieties primarily interacting lyases, except one, Arg-218.
with positively charged groups: either lysine or arginine on
PelC or the four Ca21 ions found in the complex. The car-
boxyl oxygens of GalpA2 and GalpA3 interact strongly with Position of Scissile Bond
Arg-245 and Lys-190, respectively, whereas a carboxyl oxy-
gen of GalpA4, at a distance of 3.2 Å from Lys-172, forms a Crystals of wild-type PelC, which are isomorphous with
weaker interaction. Lys-172 is highly conserved, and Lys- R218K crystals, cleave pentaGalpA when diffused into crys-
190 is invariant in the pectate lyases, but neither amino acid tals. Because the R218K mutant is catalytically inactive and
is found among the pectin lyases that bind a neutral methy- a saturated tetraGalpA has been observed, the R218K–
lated form of pectate. Arg-245 is conserved only among (Ca21)3–4–GalpA5 complexes represent a Michaelis complex
PelC subfamily members but not in the PelE subfamily in the reaction pathway. Can the scissile glycosidic bond be
whose members rapidly cleave the substrate to an unsatur- identified with certainty? PelC and subfamily members have
ated dimer. Several additional interactions between tet- been reported to cleave a pectate substrate, yielding a tri-
raGalpA and the protein were observed, but notably, the mer as the primary unsaturated end product (68 to 72%;
most specific ones involve GalpA3. Arg-223, another invari- Preston et al., 1992). In the crystal structure, an unsaturated
ant amino acid in the pectate lyase subfamily, forms hydro- trimeric end product would result only if the scissile bond
1086 The Plant Cell

and the results, in Table 4, demonstrate that a pentaGalpA

Table 3. Ca21-Coordinating Ligandsa in the
substrate has two observed modes of binding on PelC. The
PelC R218K–(Ca21)3–4–PentaGalpA Complexes
primary mode yields, as products, an unsaturated trimer (4-
Ca21 Ligand Distance (Å) en-thrHexpA-[GalpA]2) and a saturated dimer at a frequency
1Ca21 b Lys-218 N-z 2.6 of 71%. A secondary binding mode occurs at a 29% fre-
GalpA1 O-6B 2.9 quency, yielding an unsaturated dimer (4-en-thrHexpA-GalpA)
GalpA2 O-5 2.9 and a saturated trimer. When reduced pentaGalpA, contain-
GalpA2 O-6A 2.6 ing a galactonic acid at the reducing end, is used as the sub-
1CaWat1 2.8 strate, the cleavage pattern produces a reduced, unsaturated
2Ca21 Asp-160 O-d2 2.3 tetramer ([4-en-thrHexpA-GalpA]2-L-Gal-onic, where L-Gal-
Asp-162 O-d2 2.4 onic refers to L-galactonic acid) and a saturated monomer at
GalpA2 O-3 2.6
a 68% frequency. In addition, the cleavage pattern produces
GalpA3 O-5 2.5
a reduced, unsaturated trimer (4-en-thrHexpA-GalpA-L-Gal-
GalpA3 O-6A 2.2
2CaWat1 2.6 onic) and a saturated dimer at a 32% frequency. The new
2CaWat2 2.2 pattern is indicative of a shift toward the nonreducing end in
3Ca21 Glu-166 O-d1 2.5 the position of the scissile bond, because the galactonic
Glu-166 O-d2 2.5 acid unit now lies outside the enzyme in the primary binding
GalpA3 O-6B 2.4 mode. Because the galactonic acid unit is open rather than
GalpA4 O-5 2.4 in a ring structure, the reduced saccharide cannot partici-
GalpA4 O-6A 2.4 pate in the same interactions and occupy the GalpA1 site on
3CaWat1 2.5
3CaWat2
the protein–substrate complex. The only bond position that
2.4
3CaWat3
is consistent with the observed primary-mode cleavage pat-
2.4
4Ca21 terns for the reduced and unreduced pentaGalpA substrate
Asp-129 O-d1 2.3
Asp-131 O-d1 2.5 is that between GalpA3 and GalpA4 in the crystals of the pro-
Asp-131 O-d2 2.4 tein–substrate complex.
Glu-166 O-d1 2.4
Asp-170 O-d2 2.4
GalpA4 O-6A 2.5
4CaWat1
DISCUSSION
2.2
a zCa21 refers to the Z position of the Ca21 ion as defined in Figure 4.
xWaty refers to the Y water molecule associated with either the X The b elimination reaction in pectolytic cleavage is believed
GalpA unit or the X Ca21 ion. The labels for the oxygen atoms are to involve three processes: neutralization of the carboxyl
defined in Figure 4. group adjacent to the scissile glycosidic bond, abstraction
b The observed electron density is best suited to a Ca 21 with a 50%
of the C-5 proton, and transfer of the proton to the glyco-
occupancy rather than to a water molecule. The 1Ca21 ion is in the sidic oxygen. In the R218K–(Ca21)3–4–pentaGalpA structures,
same location, relative to the uronic acid of GalpA1, as are the other
the carboxyl group of GalpA3 is neutralized by interactions
Ca21 ions that coordinate GalpA units. However, unfavorable con-
with 3Ca21 and 2Ca21 as well as by Lys-190, an invariant
tact with the well-ordered and fully occupied lysine, Lys-218, was
observed. It is not possible to determine whether 1Ca21 interacts
amino acid in the pectate lyase subfamily. Lys-190 also may
with an unprotonated lysine at pH 9.5 in the crystals in 50% of the serve an additional role, which is to partially protonate the
molecules. carboxylic acid group, stabilizing an enolic intermediate as
postulated and defined by Gerlt and colleagues (Gerlt et al.,
1991; Gerlt and Gassman, 1992, 1993). Either or both ef-
fects serve to decrease the pKa (the negative log of the dis-
occurred between GalpA3 and GalpA4. Moreover, only inter- sociation constant) of the a proton at C-5, making it more
actions with GalpA3 and GalpA4 involve highly conserved susceptible to an attack by a base.
and invariant amino acids within the pectate lyase family. It is more difficult to definitively identify the group(s) re-
GalpA3 forms the most protein interactions, which appear to sponsible for proton abstraction and transfer. In our struc-
cause the greatest distortion from the 3 1 helical conforma- ture, there are no amino acids, water molecules, or Ca 21
tion of the tetraGalpA. In contrast, there are fewer interac- ions within 3 to 4 Å of any C-5 atom or glycosidic oxygen for
tions with GalpA1 and GalpA2, and all involve amino acids any GalpA unit. If the wild-type PelC structure is superim-
that are conserved only within the PelC subfamily. posed upon the R218K–substrate structure, as shown in
To confirm the position of the scissile glycosidic bond, we Figure 5, there are minimal changes in the conformation of
investigated the enzymatic cleavage patterns of oligogalac- any side chain. However, one guanidinium nitrogen of the
turonates with different degrees of polymerization under op- wild-type amino acid Arg-218 is positioned within 2.6 Å of
timized assay conditions for PelC. The composition of both C-5 of GalpA3, and the other nitrogen, at a distance of 2.7 Å,
the saturated and unsaturated end products was analyzed, interacts with an oxygen of the carboxyl group. The latter in-
 PelC Complexed to Plant Cell Wall Fragment 1087

Figure 4. Schematic Representation of R218K and Ca21 Ion Interactions with TetraGalpA at a Distance of <3.0 Å.
GalpA1 is the reducing saccharide, and GalpA4 is the nonreducing terminus. The interactions are designated with dotted lines, and the distances
are given in Table 4. Oxygen atoms are represented by circles, with the corresponding number, and the carbon atoms are assumed at the inter-
section of bonds designated in boldface lines. Water molecules, which interact with tetraGalpA, are not shown but are listed in Table 4.

teraction is likely to be responsible for the lowered pKa calcu- reaction, is modeled, a water molecule lies within 3 to 4 Å of
lated for Arg-218. By using the MEAD program (Bashford the glycosidic oxygen. The same water molecule, desig-
and Gerwert, 1992), the calculated pK a values for all PelC nated as 3CaWat2, coordinates strongly to 3Ca21 and possi-
arginine groups, except for Arg-218, fell within the range of bly is activated by the Ca21 ion. Additional experiments are
12.0 to 12.5. In contrast, the calculated pK a value for Arg- underway to test the novel enzymatic mechanism implied by
218 is 9.5, approximately the same as the pH optimum of the structural results.
the reaction. It is highly unusual for an arginine to act as a In summary, the structures of the R218K–(Ca21)3–4–penta-
general base during catalysis. However, as the H285R mu- GalpA complexes provide an atomic view of a pectate com-
tant of the acyl–acyl carrier protein thioesterase illustrates ponent of the plant cell wall, revealing a right-handed, mixed
(Yuan et al., 1995), it is not impossible. The site-specific mu- 21 and 31 helical conformation for the observed tetraGalpA
tation of the catalytic histidine to an arginine shifts the enzy- fragment and unanticipated Ca 21 positions. The complex
matic pH optimum from 8.5 to 12. In PelC, the orientation of represents a Michaelis complex in the reaction pathway,
Arg-218 suggests a catalytic role, which is consistent with and the details have led to a possible catalytic mechanism,
other known data, including the high pH optima for all involving a novel role for an arginine as a C-5 proton ab-
pectate lyase–catalyzed reactions in vitro, the catalytic im- stractor. Moreover, the structure provides a functional ex-
pairment of the R218K mutation, and the invariance of a planation for all of the invariant and conserved residues in
comparable arginine in the pectate lyase superfamily. the pectolytic active site region of the pectate lyases. Unfor-
In the structures presented in this study, no alternative at- tunately, the structure does not provide an explanation for
oms are close enough to the glycosidic oxygen between another set of invariant residues, the vWiDH amino acid se-
GalpA3 and GalpA4 to serve as a proton donor. When a par- quence, located in a second putative active site (Henrissat
tially flattened GalpA3 ring, expected during a b elimination et al., 1996), but one that is too small to accommodate long
1088 The Plant Cell

Table 4. End Product Analyses of PelC Cleavage of Oligogalacturonates

Oligomer Substratea Productsb Frequency (%) Rate (mkat/mg)

(GalpA)3 GalpA 1 4-en-thrHexpA-GalpA 100 0.17
(GalpA)4 GalpA 1 4-en-thrHexpA-(GalpA)2 25 3 to 4
(GalpA)2 1 4-en-thrHexpA-GalpA 75
(GalpA)5 (GalpA)2 1 4-en-thrHexpA-(GalpA)2 71 10.1
(GalpA)3 1 4-en-thrHexpA-GalpA 29
(GalpA)6 (GalpA)2 1 4-en-thrHexpA-(GalpA)3 9 16.1
(GalpA)3 1 4-en-thrHexpA-(GalpA)2 53
(GalpA)4 1 4-en-thrHexpA-GalpA 38
(GalpA)7 (GalpA)2 1 4-en-thrHexpA-(GalpA)4 4 22.7
(GalpA)3 1 4-en-thrHexpA-(GalpA)3 9
(GalpA)4 1 4-en-thrHexpA-(GalpA)2 55
(GalpA)5 1 4-en-thrHexpA-GalpA 32
L-Gal-onic-(GalpA)2 No cleavage 0
L-Gal-onic-(GalpA)3 GalpA 1 4-en-thrHexpA-GalpA-L-Gal-onic 100 0.1
L-Gal-onic-(GalpA)4 (GalpA)2 1 4-en-thrHexpA-GalpA-L-Gal-onic 32 1.2
GalpA 1 4-en-thrHexpA-(GalpA)2-L-Gal-onic 68
L-Gal-onic-(GalpA)5 (GalpA)4 1 4-en-thrHexpA-L-Gal-onic 2 4.0
(GalpA)3 1 4-en-thrHexpA-GalpA-L-Gal-onic 16
(GalpA)2 1 4-en-thrHexpA-(GalpA)2-L-Gal-onic 82
L-Gal-onic-(GalpA)6 (GalpA)5 1 4-en-thrHexpA-L-Gal-onic 8 4.0
(GalpA)4 1 4-en-thrHexpA-GalpA-L-Gal-onic 21
(GalpA)3 1 4-en-thrHexpA-(GalpA)2-L-Gal-onic 66
(GalpA)2 1 4-en-thrHexpA-(GalpA)3-L-Gal-onic 5
a L-Gal-onic refers to L-galactonic acid or the reduced form of galacturonic acid.
b 4-En-thrHexpA refers to a-L-eno-threohexosylpyranosyluronic acid or the 4,5-unstaurated form of galacturonic acid.

oligosaccharides. Despite the solvent accessibility in the 235 nm for unsaturated digalacturonate of 4600 M21 cm21 at pH 8.0
crystals, no GalpA units are found near the vWiDH region, (MacMillan and Vaughn, 1964). For the determination of bond cleav-
eliminating the possibility that the invariant amino acids are age frequencies, enzyme reactions were performed using the same
involved in pectolytic activity. reaction conditions, except that the buffer strength was lowered to
20 mM to prevent buffer component interference in the chromato-
graphic analysis. Aliquots were taken at timed intervals, and the re-
actions were stopped by lowering the pH to 4.5 by the addition of 0.1
METHODS volume 1% acetic acid. Reaction products were analyzed as de-
scribed previously (Hotchkiss et al., 1991; Hotchkiss and Hicks,
1993; Lieker et al., 1993; Benen et al., 1996) by using a Dionex BioLC
Preparation of Oligogalacturonates and Analyses of high-performance chromatography system (Sunnyvale, CA). Detec-
Reaction Products tion was done by pulsed amperometry and spectrophotometry at
235 nm. Quantitation of the saturated reaction products was done
Oligogalacturonates with two to seven GalpA units were prepared using the amperometric data and the amperometric response of a
from polygalacturonic acid (PGA), as described previously (Kester calibration mixture of oligogalacturonates with different degrees of po-
and Visser, 1990). Reduced oligogalacturonates were prepared ac- lymerization. The unsaturated oligogalacturonates were quantitated
cording to the method of Omran et al. (1986). Quantitation of the oli- using the spectrophotometric response.
gomers and analyses of the reaction products were conducted as A fast-atom bombardment mass spectrum of pentaGalpA dis-
described previously (Parenicova et al., 1998). Enzymatic reaction played the major difference between the molecular ion and hydro-
rates for oligogalacturonates with different degrees of polymerization gen, (M-H)2, at a mass-to-charge ratio (m/z) of 897.6. Relatively
were determined spectrophotometrically at 235 nm by using 0.5 mM minor amounts of (M-H) 2 ions at m/z ratios of 721.5, 545.4, and
oligogalacturonate in 0.1 M 2-amino-2-methyl-1-propanol buffer, pH 369.3 also were detected, representing tetraGalpA, triGalpA, and di-
9.5, in the presence of 1.0 mM CaCl2 at 258C. Enzyme activities were GalpA, respectively. Mass spectra were obtained with a ZAB 2-SE
expressed as mkat/mg by using the molar extinction coefficient at high-field magnetic sector mass spectrometer (VG Analytical,
 PelC Complexed to Plant Cell Wall Fragment 1089

Figure 5. Stereoview of the (Ca21)4–TetraGalpA Substrate Superimposed upon the Structure of Wild-Type PelC.
The color scheme is the same as that used in Figure 1, except that wild-type Arg-218 is present and highlighted in magenta. One of the guani-
dinium nitrogens of Arg-218 lies within 2.6 Å of C-5 of GalpA3, and the other nitrogen lies within 2.7 Å of O-6B of GalpA3.

Manchester, UK) by using a cesium gun for fast-atom bombardment Structure Determination
ionization at 8000 electron volts, glycerol–thioglycerol–triethylamine
(10:10:1) as the matrix, and cesium iodide for mass calibration. The The structure has been solved by difference Fourier methods. Struc-
pentaGalpA sample was dissolved in water. ture factors, by using the refined PelC model (Yoder and Jurnak,
1995) in which Arg-218 had been omitted, were used to calculate a
simulated annealed OMIT electron density map. Before water mole-
Preparation of Crystals cules were positioned, Ca21 ions were fitted into the three highest
peaks and refined without restraints. Four GalpA units were fitted to
The R218K mutant of pectate lyase C (PelC) was isolated from the the remaining clustered density with the program O (Jones et al.,
periplasm of Escherichia coli HMS174(DE3) cells harboring pRSET5A 1991). The model was refined using the method of slow-cooling sim-
constructs and purified as previously described (Kita et al., 1996). ulated annealing as implemented by X-PLOR (Brünger, 1996). The
Crystals were grown using conditions similar to that for wild-type reflection data, with structure factor amplitudes (F) greater than two
PelC crystals (Yoder et al., 1990). The R218K mutant crystals are iso- standard deviations, were randomly divided into two sets. The work-
morphous with wild-type PelC crystals and belong to space group ing set was composed of 90% of the data sampled at random, and
P212121 with unit cell parameters of a 5 72.14 Å, b 5 78.32 Å, and c 5 the test set was composed of the remaining 10% of the data used for
94.43 Å, with one molecule per asymmetric unit. The R218K crystals cross-validation of the refinement cycles (Brünger, 1993). The pa-
were transferred from ammonium sulfate to a solution at pH 9.5 con- rameter and topology files used in X-PLOR were those of Engh and
taining cryogenic agents, 7 mM Ca21, and 50 mM pentaGalpA, and Huber (1991) for the protein and those of Ha et al. (1988), as modified
after 30 hr, they were frozen in liquid nitrogen for data collection. by Weis et al. (1990), for the saccharide.
The conformation of each amino acid in the substrate binding re-
gion was adjusted by a series of refined OMIT maps in which a region
X-Ray Diffraction Data Collection of 8 Å around a residue had been omitted from the structure factor
calculations and refinement (Hodel et al., 1992). In subsequent differ-
X-ray diffraction data were collected to a resolution of 2.19 Å at ence maps, peaks .4s and satisfying reasonable distance and ge-
21708C by using a wavelength of 1.08 Å on a MARS imaging plate ometry criteria were assigned as water molecules by using MAPMAN
detector on Beam-Line 7-1 at the Stanford Synchrotron Radiation (Kleywegt and Jones, 1996). All water molecules were inspected vi-
Light Source (Stanford, CA). The data were processed using MOS- sually. In one location, a water molecule could not account ade-
FLM (Leslie, 1996). Data are included in Table 5. quately for the residual electron density. Because the density was
1090 The Plant Cell

Estimation of pKa Values

Table 5. Crystallographic Data Collection and Refinement Statistics
for the PelC R218K–(Ca21)3–4–PentaGalpA Complexes The pKa values for individual amino acids within the three-dimen-
Parameter Value sional structure of the R218K–(Ca21)3–4–GalpA5 complexes were de-
termined using macroscopic electrostatics with atomic detail
Resolution 2.2Å (MEAD), version 1.1.8 (Bashford and Gerwert, 1992). Standard partial
Total observations 93,630 atom charges were used, and the intrinsic pKa values were calcu-
Unique observations 27,518 lated from the MEAD program multiflex.
Percent completeness 95.1%
Average I/s 11.8
Rsyma 3.1%
Resolution range of refinement 2.2 to 10.0 Å ACKNOWLEDGMENTS
No. of reflections with F . 2s 26,662
Nonhydrogen protein atoms/asymmetric unit 2,647
Water molecules/asymmetric unit 328 The research was supported by the U.S. Department of Agriculture
Ca21 molecules/asymmetric unitb 3 to 4 (Grant No. 96-02966 to F.J.), the National Science Foundation (Grant
GalpA molecules/asymmetric unit 4 No. MCB9408999 to N.T.K. and F.J.), and the San Diego Supercom-
Rtestc for 10% data 23.3 puter Center. The research was conducted in part at the Stanford
Rworkd for 90% data 18.6 Synchrotron Radiation Laboratory, which is operated by the Office of
Root-mean-square deviation from ideal geometry Basic Energy Science of the U.S. Department of Energy.
Bond length 0.0006 Å
Bond angle 1.548
Impropers 1.178 Received August 10, 1998; accepted March 28, 1999.
Average thermal factors
Main chain 9.6 Å2
Side chain 10.0 Å2
All protein atoms 9.7 Å2 REFERENCES
Water molecules 22.6 Å2
All nonhydrogen atoms 11.1 Å2
Atkins, E.D.T., Nieduszynski, I.A., Mackie, W., Parker, K.D., and
aR
sym 5 100 5 ouIavg 2 Iobsu/oIavg, when Iavg is the average (avg) or Smolko, E.E. (1973). Structural components of alginic acid. II. The
the observed (obs) intensity of the reflection. crystalline structure of poly-a-guluronic acid. Results of x-ray dif-
b Three of the Ca 21 ions refined to a 100% occupancy, and the
fraction and polarized infrared studies. Biopolymers 12, 1879–
fourth Ca21 ion refined to a 50% occupancy. 1887.
test 5 ouFo 2 Fcu/ouFcu for the 10% of the reflections that were set
cR

aside for cross-validation and not used in the refinement. Barras, F., Van Gijsegem, F., and Chatterjee, A.K. (1994). Extra-
cellular enzymes and pathogenesis of soft rot Erwinia. Annu. Rev.
work 5 ouFo 2 Fcu/ouFcu for the 90% of the reflections used in the
dR

refinement calculations. Phytopathol. 32, 201–234.

Bashford, D., and Gerwert, K. (1992). Electrostatic calculations of
the pKa values of ionizable groups in bacteriorhodopsin. J. Mol.
Biol. 224, 473–486.
located near to the carboxyl group of GalpA1, in the same relative lo- Benen, J.A.E., Kester, H.C.M., Parenicova, L., and Visser, J.
cation as the other Ca 21 ions to the coordinating GalpA units, the (1996). Kinetics and mode of action of Aspergillus niger polyga-
density was assigned as a fourth Ca21 ion, 1Ca21, with a lowered oc- lacturonase. In Progress in Biotechnology: Pectins and Pectin-
cupancy. After several cycles of refinements, the best fit of the den- ases, Vol. 14, J. Visser and A.G.J. Voragen, eds (Amsterdam:
sity appeared to be that for a Ca21 ion with 50% occupancy and a Elsevier), pp. 221–230.
thermal factor of 15 Å2. However, with this assignment, there remains
Brünger, A.T. (1993). Assessment of phase accuracy by cross vali-
an unfavorable contact with a well-ordered and fully occupied lysine,
dation: The free R value. Methods and applications. Acta Crystal-
Lys-218. It is not possible to determine whether 1Ca21 interacts with
logr. D49, 24–36.
an unprotonated lysine at pH 9.5 in the crystals in 50% of the mole-
cules. Brünger, A.T. (1996). X-PLOR Manual, Version 3.8. (New Haven,
The final atomic model consists of 352 amino acids, 328 water CT: Yale University).
molecules, 3.5 Ca21 ions, and an a-1,4–linked oligosaccharide con- Carpita, N.C., and Gibeaut, D.M. (1993). Structural models of pri-
sisting of four well-ordered GalpA units. The final model refined to a mary cell walls in flowering plants—Consistency of molecular
crystallographic R factor of 18.6% for all measured reflections with structure with the physical properties of the walls during growth.
structure–factor amplitudes F . 2s in the 2.2 to 10.0 Å range. For the Plant J. 3, 1–30.
final statistics calculations, all reflections .2s were used, including
the test data set. The model was checked throughout using Catoire, L., Derouet, C., Redon, A.M., Goldberg, R., and Dupenhoat,
PROCHECK (Laskowski et al., 1993), and with the exception of the C.H. (1997). An NMR study of the dynamic single-stranded con-
Lys-218–1Ca21 distance, no bad contacts were observed in the final formation of sodium pectate. Carbohydr. Res. 9, 19–29.
model. Refinement statistics are summarized in Table 5. All model Collmer, A., and Keen, N.T. (1986). The role of pectic enzymes in
figures were prepared using SETOR (Evans, 1993). plant pathogenesis. Annu. Rev. Phytopathol. 24, 383–409.
 PelC Complexed to Plant Cell Wall Fragment 1091

DiNola, A., Fabrizi, G., Lamba, D., and Segre, A.L. (1994). Solution Kester, H.C.M., and Visser, J. (1990). Purification and characteriza-
conformation of a pectic acid fragment by 1H-NMR and molecular tion of polygalacturonases produced by the fungus Aspergillus
dynamics. Biopolymers 34, 457–462. niger. Biotechnol. Appl. Biochem. 12, 150–160.

Engh, R.A., and Huber, R. (1991). Accurate bond and angle param- Kita, N., Boyd, C.M., Garrett, M.R., Jurnak, F., and Keen, N.T.
eters for x-ray protein structure refinement. Acta Crystallogr. A47, (1996). Differential effect of site-directed mutations in PelC of pec-
392–400. tate lyase activity, plant tissue maceration, and elicitor activity. J.
Biol. Chem. 271, 26529–26535.
Evans, S.V. (1993). SETOR. J. Mol. Graphics 11, 134.
Kleywegt, G.T., and Jones, T.A. (1996). Biomacromolecular XDL
Gerlt, J.A., and Gassman, P.G. (1992). Understanding enzyme-cat- Mapman and XDL Dataman—Programs for reformatting, analysis
alyzed proton abstraction from carbon acids: Details of stepwise and manipulation of electron density maps. Acta Crystallogr. D52,
mechanism for b-elimination reactions. J. Am. Chem. Soc. 114, 826–828.
5928–5934.
Kohn, R. (1975). Ion binding on polyuronates—Alginate and pectin.
Gerlt, J.A., and Gassman, P.G. (1993). An explanation for rapid Pure Appl. Chem. 42, 371–397.
enzyme-catalyzed proton abstraction from carbon acids: Impor-
tance of late transition states in concerted mechanisms. J. Am. Kotoujansky, A. (1987). Molecular genetics of soft rot Erwinias.
Chem. Soc. 115, 11552–11568. Annu. Rev. Phytopathol. 25, 405–430.

Gerlt, J.A., Kozarich, J.W., Kenyon, G.L., and Gassman, P.G. Laskowski, R.A., McArthur, M.W., Moss, D.S., and Thornton,
(1991). Electrophilic catalysis can explain the unexpected acidity J.M. (1993). PROCHECK: A program to check the stereochemical
of carbon acids in enzyme catalyzed reactions. J. Am. Chem. Soc. quality of protein structures. J. Appl. Crystallogr. 26, 283–291.
113, 9667–9669. Leslie, A.G.W. (1996). MOSFLM User Guide. (Cambridge, UK: Med-
Gouvion, C., Mazeau, K., Heyraud, A., Taravel, F.R., and ical Research Council Laboratory of Molecular Biology).
Tvaroska, I. (1994). Conformational study of digalacturonic acid Lieker, H.-P., Thielecke, K., Buchholz, K., and Reilly, P.J. (1993).
and sodium digalacturonate in solution. Carbohydr. Res. 261, High performance anion-exchange chromatography of saturated
187–202. and unsaturated oligogalacturonic acids. Carbohydr. Res. 238,
Grant, G.T., Morris, E.R., Rees, D.A., Smith, P.J.C., and Thom, D. 307–311.
(1973). Biological interactions between polysaccharides and diva- Lietzke, S.E., Keen, N.T., Yoder, M.D., and Jurnak, F. (1994). The
lent cations. FEBS Lett. 32, 195–198. three-dimensional structure of pectate lyase E, a plant virulence
Ha, S.N., Giammona, A., Field, M., and Brady, J.W. (1988). A factor from Erwinia chrysanthemi. Plant Physiol. 106, 849–862.
revised potential energy surface for molecular mechanic studies MacMillan, J.D., and Vaughn, R.H. (1964). Purification and proper-
of carbohydrates. Carbohydr. Res. 180, 207–221. ties of a polygalacturonic acid trans-eliminase produced by
Henrissat, B., Heffron, S.E., Yoder, M.D., Lietzke, S.E., and Jurnak, Clostridium multifermentans. Biochemistry 3, 564–572.
F. (1996). Functional implications of structure-based sequence Mayans, O., Scott, M., Connerton, I., Gravesen, T., Benen, J.,
alignment of proteins in the extracellular pectate lyase superfam- Visser, J., Pickersgill, R., and Jenkins, J. (1997). Two crystal
ily. Plant Physiol. 107, 963–976. structures of pectin lyase A from Aspergillus reveal a pH driven
Hodel, A., Kim, S.H., and Brünger, A.T. (1992). Model bias in mac- conformational change and striking divergence in the substrate-
romolecular crystal structures. Acta Crystallogr. A48, 851–858. binding clefts of pectin and pectate lyases. Structure 5, 677–689.

McNaught, A.D. (1997). Nomenclature of carbohydrates. Adv. Car-

Hotchkiss, A.T., and Hicks, K.B. (1993). Analysis of pectate lyase-
bohydr. Chem. Biochem. 52, 43–171.
generated oligogalacturonic acids by high-performance anion-
exchange chromatography with pulsed amperometric detection. Morris, E.R., Powell, D.A., Gidley, M.J., and Rees, D.A. (1982). I.
Carbohydr. Res. 247, 1–7. Polymorphism between gel and solid states of calcium polygalac-
turonate. J. Mol. Biol. 155, 507–516.
Hotchkiss, A.T., Hicks, K.B., Doner, L.W., and Irwin, P.L. (1991).
Isolation of oligogalacturonic acids in gram quantities by prepara- Omran, H., Dörreich, K., and Gierscher, K. (1986). Some results of
tive HPLC. Carbohydr. Res. 215, 81–90. the isolation of discrete enzymes (exopolygalacturonase and
pectinesterase) from a commercial pectic enzyme preparation.
Hricovini, M., Slavomir, B., and Malovikova, A. (1991). Conforma-
Lebensm.-Wiss. Technol. 19, 457–463.
tions of (1→4)-linked a-D-galacturono-di- and -tri-saccharides in
solution analyzed by n.m.r. measurements and theoretical calcu- Parenicova, L., Benen, J.A.E., Kester, H.C.M., and Visser, J.
lations. Carbohydr. Res. 220, 23–31. (1998). PgaE encodes a fourth member of the endopolygalac-
turonase gene family from Aspergillus niger. Eur. J. Biochem. 251,
Jarvis, M.C., and Apperley, D.C. (1995). Chain conformation in
72–80.
concentrated pectic gels: Evidence from 13C-NMR. Carbohydr.
Res. 275, 131–145. Pickersgill, R., Jenkins, J., Harris, G., Nasser, W., and Robert-
Baudouy, J. (1994). The structure of Bacillus subtilis pectate lyase
Jones, T.A., Zou, J.-Y., Cowan, S.W., and Kjeldgaard, M. (1991).
in complex with calcium. Nat. Struct. Biol. 1, 717–723.
Improved methods for building protein models in electron density
maps and the location of errors in these models. Acta Crystallogr. Powell, D.A., Morris, E.R., Gidley, M.J., and Rees, D.A. (1982).
A47, 110–119. Conformations and interactions of pectins. II. Influence of residue
1092 The Plant Cell

sequence on chain association in calcium pectate gels. J. Mol. interactions of pectins. II. Models of junction zones in pectinic
Biol. 155, 517–531. acid and calcium pectate gels. J. Mol. Biol. 153, 1075–1085.
Weis, W.I., Brünger, A.T., Skehel, J.J., and Wiley, D.C. (1990).
Preston, J.F., Rice, J.D., Ingram, L.O., and Keen, N.T. (1992). Dif-
Refinement of the influenza virus hemagglutinin by simulated
ferential depolymerization mechanisms of pectate lyases secreted
annealing. J. Mol. Biol. 212, 737–761.
by Erwinia chrysanthemi EC16. J. Bacteriol. 174, 2039–2042.
Yoder, M.D., and Jurnak, F. (1995). The refined three-dimensional
Thibault, J.-F., Renard, C.M.G.C., Axelos, M.A.V., Roger, P., and structure of pectate lyase C from Erwinia chrysanthemi at 2.2 Å
Crepeau, M.-J. (1993). Studies of the length of homogalacturonic resolution—Implications for an enzymatic mechanism. Plant
regions in pectins by acid hydrolysis. Carbohydr. Res. 238, 271–286. Physiol. 107, 349–364.
Vitali, J., Schick, B., Kester, H.C.M., Visser, J., and Jurnak, F. Yoder, M.D., DeChaine, D.A., and Jurnak, F. (1990). Preliminary
(1998). The three-dimensional structure of Aspergillus niger pectin crystallographic analysis of pectate lyase C from Erwinia chrysan-
lyase B at 1.7 angstrom resolution. Plant Physiol. 116, 69–80. themi. J. Biol. Chem. 265, 11429–11431.
Yoder, M.D., Keen, N.T., and Jurnak, F. (1993). New domain motif:
Walkinshaw, M.D., and Arnott, S. (1981a). Conformations and
The structure of pectate lyase. Science 260, 1503–1507.
interactions of pectins. I. X-ray diffraction analysis of sodium pec-
Yuan, L., Nelson, B.A., and Caryl, G. (1995). The catalytic cysteine
tate in neutral acidified forms. J. Mol. Biol. 153, 1055–1073.
and histidine in the plant acyl–acyl carrier protein thioesterases. J.
Walkinshaw, M.D., and Arnott, S. (1981b). Conformations and Biol. Chem. 271, 3417–3419.