Isolation and characterization of calmodulin-like domain protein kinase from soybean

MISSING IMAGE

Material Information

Title:
Isolation and characterization of calmodulin-like domain protein kinase from soybean
Physical Description:
ix, 112 leaves : ill. ; 29 cm.
Language:
English
Creator:
Lee, Jung-Youn, 1966-
Publication Date:

Subjects

Subjects / Keywords:
Protein kinases   ( lcsh )
Soybean -- Genetics   ( lcsh )
Plant Molecular and Cellular Biology thesis, Ph. D   ( lcsh )
Dissertations, Academic -- Plant Molecular and Cellular Biology -- UF   ( lcsh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1997.
Bibliography:
Includes bibliographical references (leaves 95-111).
Statement of Responsibility:
by Jung-Youn Lee.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 027599550
oclc - 37204846
System ID:
AA00014282:00001


This item is only available as the following downloads:


Full Text










ISOLATION AND CHARACTERIZATION OF CALMODULIN-LIKE
DOMAIN PROTEIN KINASE FROM SOYBEAN












By


JUNG-YOUN LEE


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY



UNIVERSITY OF FLORIDA


























To my parents and Jaewoong














ACKNOWLEDGMENTS


I would like to thank my adviser, Dr. Alice Harmon, for her support,
encouragement, and insight into science. Special thanks go to our former lab

member, Mrs. Candace MacCafferey, for her technical assistance. I thank my

committee members, Drs. Robert Ferl, Don McCarty, Bill Gurley, and David

Jones, for their advise and generous sharing of their resources. I owe

tremendously to my husband and colleague, Byung-Chun, for his bottomless
support and to our jolly son, Jaewoong, who always was willing to let mom

go to school. I also cannot thank enough my family in Korea, especially my

mother, who was not reluctant to fly over 30,000 miles.














TABLE OF CONTENTS


ACKNOWLEDGEMENTS ................................................ ..............................iii

LIST O F TA BLES....................................... ......................................................... vi

LIST O F FIG URES................................................................. ........................... vii

A BST R A C T .......................................................................... ..................................viii

CHAPTERS

1. LITERATURE REVIEW .................................................... ........................... 1

Introduction......................................... ................... .........................................1
Generation of Changes in Cytosolic Ca2+ Level............................ ............ 3
Regulation of Cytosolic Ca2............................. ............ .............................15

2. CLONING OF cDNA, HETEROLOGOUS EXPRESSION, AND
CHARACTERIZATION OF CDPKp AND y FROM SOYBEAN
AND THIEIR COMPARISON TO CDPKa......................................................24

Introduction..................... ....... .................................................24
Experimental Procedures ............................. ...........................................27
Results.................................................. ................................................. ....... 36
Discussion...................... ..........................................................................59

3. CALCIUM BINDING PROPERTIES OF THREE SOYBEAN CDPKS............65

Introduction......................................... .................... ............................... 65
Experimental Procedures..............................................................69
Results.................................................. ................................................. ....... 75
D iscu ssio n .......................................................................... ..................................87

4. SUMMARY AND CONCLUSIONS.....................................................................92

Summay of Results ................... .......................................................... 92
C on clusion s.............................................. .....................................................93





iv










REFER EN C ES ............................................ ...............................................................95

BIOGRAPHICAL SKETCH.............................. ....... .........................112




















































V













LIST OF TABLES


Table page

2-1. Activity of CDPK isoforms with various substrates....................................52

2-2. Kinetic parameters of CDPK isoforms ........................................................54

2-3. IC50 Values for Inhibitors of CDPK Isoenzymes...........................................57

3-1. Ca2+ dissociation and Hill constants of CDPK isoenzymes.......................76

3-2. K0.5s for Ca2+ of CDPKa, p, and y with various substrates..........................89













LIST OF FIGURES


Figure page

2-1. Nucleotide and deduced amino acid sequences of two new CDPK
isoenzmes from soybean............................... ..............................37

2-2. Alignment of amino acid residues of soybean CDPK isoforms with
known CDPKs from other plants.......................................................43

2-3. RNA blot hybridization analyses of CDPK encoding genes.......................47

2-4. Immunoblot of recombinant proteins with monoclonal antibodies........49

2-5. SDS-PAGE of purified recombinant proteins........................................50

2-6. Sequences of substrate peptides......................... ................................55

2-7. Effect of pH on activity................................................ ..... ...................... 57

2-8. Immunoblots of soybean cell extracts with Anti-Nty and anti-CLD
antibodies .......................................................... ...................................... 60

3-1. A schematic representation of flow dialysis system....................................72

3-2. Direct Ca2+ binding to CDPKa, p, and y................... ........................... 77

3-3. The effect of Ca2+ on kinase activity of CDPKs .............................................80

3-4. Ca2+-binding studies of CDPKa under various conditions......................82

3-5. Changes in K0.5 of CDPKa by different protein substrates as a function of
free Ca2+ concentration............... ............................. ................................ 84

3-6. The effect of Ca2+ on autophosphorylation of CDPKs..................... ........86













Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy


ISOLATION AND CHARACTERIZATION OF CALMODULIN-LIKE
DOMAIN PROTEIN KINASE ISOFORMS FROM SOYBEAN

By

Jung-Youn Lee

May, 1997




Chairman: Dr. Alice C. Harmon
Major Department: Plant Molecular and Cellular Biology


Calmodulin-like domain protein kinase (CDPK) is a calcium, but not

calmodulin, dependent protein kinase that was first characterized from

soybean. CDPKs are encoded by gene families in a wide variety of plant

species and are also found in protests. This paper reports the isolation of

cDNAs encoding two additional CDPK isoforms from soybean and studies of

biochemical characteristics of the recombinant proteins including CDPKa.

The new CDPKs with predicted molecular masses of 55 kDa and 60 kDa are
named CDPKp and CDPKy, respectively, and each showed 76% and 58%
amino acid sequence identity to CDPKa. RNA blot hybridization with specific
probes for each isoform revealed different expression patterns for each
transcript.








For determination of characteristics, the three isoforms were expressed

in Escherichia coli as glutathione S-transferase fusion proteins and were
highly purified. The CDPKs differed in biochemical properties such as
stability during storage, requirement of DTT for enzyme activity, and fold-
stimulation by calcium. All three isozymes phosphorylated histone IIIS,
syntide-2, autocamtide-2, and myosin light chain kinase substrate peptides,
but the kinetic parameters of Kmapp and Vmaxapp, and substrate preferences
were different. The pH optimum for all three enzymes was pH 7 to 8. The
three soybean CDPKs were inhibited by the general protein kinase inhibitors,
K252a and staurosporine at similar or higher IC50s than these for other

protein kinases. The protein kinase C specific inhibitor, calphostin C,
inhibited the CDPKs with IC50s higher (30- to 200-fold) than that for protein
kinase C.

The Ca2+-binding properties of CDPK isoenzymes were studied by flow

dialysis in buffer containing 50 mM HEPES, pH 7.5, and 100 mM KC1. The
Kds for Ca2+ of CDI'Ku, I1, and y were -45 giM, 1.5 iM, and 1 iM, respectively.
The concentration of Ca2+ required for half maximal activity (K0.5) of each
CDPK ranged from 0.1 pM to 1 gM. The Kd for Ca2+ of CDPKa decreased to ~1

iM in the presence of syntide-2 from 45 gM. In vitro substrates of CDPK,
histone IIIS and Serine acetyl transferase (SAT) affected the Ca2+ sensitivity of
CDPK activity to different degrees: the Ko.Ss were -0.1 jiM and -4 pM in the

presence of SAT and histone IRIS, respectively. These data demonstrated that
each CDPK isoform had unique Ca2+-binding property and kinetics.
These results show that members of the CDPK family differ in
biochemical properties and support the hypothesis that each isoform may
have a distinct role in Ca2+ signal transduction.













CHAPTER 1
LITERATURE REVIEW


Introduction

Ca2+ is a versatile intracellular signaling molecule that regulates a wide

array of physiological processes. The concentration of cytosolic Ca2+ changes

in response to a variety of stimuli and results in different cellular responses.

A particular physiological response is generated not only by the fluctuation in
the level of intracellular Ca2+ but also by the route of Ca2+ entry and its
intracellular localization. The role of Ca2+ as a second messenger in signal
transduction pathways in animals and plants has been extensively discussed
in numerous reviews (Berridge and Dupont, 1994; Bootman and Berridge,
1995; Bush, 1993, 1995; Clapham, 1995; Gilroy et al., 1993; Hepler and Wayne,
1985; Knight et al., 1995; Trewavas and Knight, 1994; Trewavas et al., 1996).
The cytosolic Ca2+ concentration is normally maintained at 10-100 nM,

and this is over 10,000-fold lower than extracellular Ca2+ concentration (1-2
mM). Cells keep cytosolic Ca2+ concentration low through numerous binding

proteins and specialized pumping proteins because Ca2+ is a cytotoxin. At

high concentration, Ca2+ will form precipitates inside the cell from the

reaction with inorganic phosphate. The cytosolic concentration of a closely
related divalent cation Mg2+ is higher (milimolar) than Ca2+ but many
cellular processes utilize Ca2+ as a signaling molecule. Mg2+ binds water
more tightly than Ca2+. Since the basal Ca2+ concentration is low level and it








has a lower affinity for water than Mg2+, it can act as an intracellular

messenger.
The increased Ca2+ concentration promotes formation of complexes

with target proteins. Ca2+-binding proteins often coordinate Ca2+ through -6

oxygens provided by glutamic acid, aspartic acid, serine, threonine, and

asparagine residues (Clapham, 1995; Hepler and Wayne, 1985; Kretsinger,

1987). Most Ca2+-modulated proteins contain EF-hand motifs. The EF-hand

consists of two perpendicularly arranged a-helices connected by a Ca2+

coordinating loop. EF-hand proteins can be grouped to Ca2+ sensors and Ca2+
buffers (Ikura, 1996). Ca2+ sensors or triggers change their conformations
upon binding Ca2+ so that they modulate effector molecules. Ca2+ buffers, on
the other hand, may simply bind Ca2+ to keep cytoplasmic Ca2+ level low.

Due to the number, affinity, and specificity of Ca2+-binding proteins in cells,

the Ca2+ signal can be highly localized. It is estimated that the effective range

of free Ca2+ is 0.1 grm lasting only ~30 [s before it is buffered, and that of

buffered Ca2+ is ~5 im lasting -1 s provided that the on-rate of a typical

calcium buffer is 108 M-ls-1 and the concentration of Ca2+ buffers is -300 iM

(Allbritton, et al., 1992). The slow diffusion of Ca2+ results from its binding to

slowly mobile or immobile buffers. Since the rapid buffering of Ca2+ makes it

a localized messenger, the effector molecules that require a high

concentration of Ca2+ for activation have to be located less than -0.5 Am from

a Ca2+ source.

As a second messenger, Ca2+ signals exhibit highly organized temporal
and spatial arrangement. The spatiotemporal aspect of Ca2+ signaling is

complex and the underlying mechanism is not yet fully defined (Berridge and
Dupont, 1994; Miyazaki, 1995).








Generation of Changes in Cytosolic Ca2+ Level

In plants, cytosolic Ca2+ levels have been reported to rise upon various
extracellular stimuli such as abscisic acid (Allan et al., 1994; Gilroy, 1996;
McAinsh et al., 1992; Zocchi and De Nish, 1996), auxin (Felle, 1988; Gehring et
al., 1990), gibberellic acid (Bush, 1996; Gilroy, 1996), cold shock (Knight et al.,
1991; Knight et al., 1996; Monroy and Dhindsa, 1995), light (Bowler et al., 1994;
Fa"on et al., 1993; Gehring et al., 1990; Nauhaus et al., 1993; Shacklock and
Trewavas, 1992), gravity (Gehring et al., 1990), mechanical stimulation
(Knight et al., 1991; Knight et al., 1992), osmotic stress (Cramer and Jones,
1996), anoxia (Sedbrook et al., 1996; Subbaiah et al., 1994), oxidative stress
(McAinsh et al., 1996; Price et al., 1994), elicitors (Knight et al., 1991), and
pathogen (Levine et al., 1996). Ca2+ is also involved in pollen tube growth
(Franklin-Tong et al., 1996; Malho and Trewavas, 1996; Pierson et al., 1994),
root nodule formation (Ehrhardt et al., 1996), ethylene (Philosoph-Hadas et
al., 1996; Raz and Fluhr, 1992), and carbon dioxide (Webb et al., 1996) signal
transduction pathways. Changes in cytosolic Ca2+ are essential to these

processes and the changes are highly variable in amplitude, kinetics, and
spatial distribution. Changes in cytosolic Ca2+ induced by different stimuli
may be transient, sustained, or oscillatory and the time that required for these
changes varies from seconds to hours. In plants, rapid and transient changes
in cytosolic Ca2+ could be induced by abscisic acid (ABA), auxin, cold, red
liglit, elicitor, and mechanical stimuli while slower and more complex
changes in cytosolic Ca2+ were observed in responses induced by auxin, ABA,
red light, gibberellic acid (GA). Changes in cytosolic Ca2+ also differ with
respect to spatial characteristics (Bush, 1996). A steady state Ca2+ gradient
across the plasma membrane is present in pollen tubes (Pierson et al., 1994)








and root hairs (Felle et al., 1992). It was suggested that these Ca2+ gradients
were formed based on the extracellular Ca2+ influx into cytoplasm through
the activity of Ca2+ channels at the apical plasma membrane. In addition, the
intracellular Ca2+ gradient might be regulated by specific pumps that extrude
Ca2+ from cytoplasm to lower the Ca2+ concentration to basal levels (Pierson

et al., 1994).

ABA and Stomatal Closure

The plant hormone ABA was first discovered as a naturally occurring

compound that accelerates leaf abscission and bud dormancy in woody plants.

ABA is also involved in other processes such as seed dormancy and
germination, the regulation of stomatal aperture, and the responses to
environmental stress. Plants respond to excessive water loss by controlling
the pore size of stomata. ABA stimulates stomatal closing by increasing efflux
of K+ and anions from guard cells (Giraudat, et al., 1994).
The aperture of stomatal pores through which the exchange of CO2 and

water vapor occurs is controlled by the two surrounding guard cells. Guard

cell volume changes in response to many signals such as CO2, humidity,
phytohormones, and light (Kearns, E.V. and Assmann, S.M., 1993). ABA is a
well known phytohormone to induce stomatal closure during water stress. It

has been reported that ABA induces rapid, transient increases (Gilroy, et al.,

1991; McAinsh et al., 1992) or repetitive increases (Schroeder and Hagiwara,
1990) of cytosolic Ca2+ prior to stomatal closure.
Increases of transient cytosolic Ca2+ were observed with both open and
closed guard cells of Commelina communis in response to ABA treatment
(Gilroy, et al., 1991; McAinsh et al., 1992). Fluorescence ratio imaging of ABA-
stimulated guard cell using microinjected calcium indicator Indo-I








demonstrated that cytosolic Ca2+ increases were not uniformly distributed but
localized in the periphery of the cell and the external boundaries of internal
Ca2+ stores. Gilroy et al. (1990) microinjected fluorescent Ca2+ indicator Fluo-

3 and caged Ca2+ and monitored changes in stomatal aperture of Commelina
communis. When cytosolic Ca2+ was elevated above 600 nM by UV
photolysis of caged Ca2+, stomatal closure was observed. When caged

inositol-1, 4, 5-trisphosphate (InsP3) was photoactivated or InsP3 was directly
microinjected, increases of cytosolic Ca2+ from intracellular Ca2+ stores were
observed followed by subsequent stomatal closure. However, the induction

of cytosolic Ca2+ increases by ABA treatment were highly variable while all
stomata were closed in response to ABA (Gilroy, et al., 1991; McAinsh et al.,
1992). These observations led to the question of whether ABA-induced
stomatal closure occurs through both Ca2+-dependent and Ca2+-independent
pathways. Allan et al. (1994) investigated the effects of growth temperature
on ABA-induced cytosolic Ca2+ changes and showed that ABA signal
transduction in guard cells resulting in stomatal closure could be mediated by

both Ca2+-dependent and Ca2+-independent pathways. The guard cells of

Commelina plants grown at 10-17 OC did not show an increase in cytosolic
Ca2+ by ABA treatment although all of them were closed. In contrast, when

the plants were grown at 25 OC an increase in cytosolic Ca2+ was always
accompanied by ABA-induced stomatal closures. Elevation of cytosolic Ca2+
in response to ABA treatment was also impaired when the plants were pre-
exposed to water stress.
ABA also has been implicated in inducing repetitive increases of
cytosolic Ca2+ in Vicia faba guard cells using patch damping and fluorescent
imaging with Fura-2, (Schroeder and Hagiwara, 1990). The increase of Ca2+
occurred simultaneously with the openings of nonspecific cation channel in








the plasma membrane. The result indicated that Ca2+ enters the cell through

these channels. The activation of these nonselective, Ca2+-permeable ion
channels occurred within 2 seconds of ABA treatment. When cytosolic Ca2+

was elevated to micromolar, inward K+ channel activity at the plasma
membrane was inhibited and caused anion efflux (Schroeder and Hagiwara,
1989; 1990) which resulted in membrane depolarization and activation of
outward K+ channels in the plasma membrane. The net efflux of ions from
the stomatal guard cells caused reduction in turgor and closing of stomata.
Increases in cytosolic Ca2+ in guard cells also showed the activation of
outward vacuolar K+ channels (Ward and Schroeder, 1994). The activation of
vacuolar K+ channels resulted in depolarization of vacuolar membrane
which is sufficient to activate slow vacuolar ion channels. These slow
vacuolar ion channels revealed a permeability ratio for Ca2+ to K+ of -3:1

suggesting a possible mechanism for Ca2+-induced Ca2+-release from vacuoles
during stomatal closure. Gilroy et al. (1990) showed InsP3 could induce

cytosolic Ca2+ increase and subsequent stomatal closure, when caged InsP3 was
microinjected into the guard cells and was activated by photolysis. Whether

InsP3 is the in vivo mediator of Ca2+ release from vacuoles requires further
examination. Luan et al. (1993) proposed that the inactivation of inward K+

channels on the plasma membrane in guard cells may be mediated by the
Ca2+/calmodulin dependent protein phosphatase 2B (PP2B) homolog using a

specific inhibitor of PP2B. Also, when the constitutively active (Ca2+-
independent) form, i.e., catalytic subunit of PP2B was introduced into the
guard cell, inward K+ channels were inhibited in the absence of Ca2+. Fairley-
Grcnot and Assmann (1991) reported that heterotrimeric G-protein may be
involved in stomatal closure through the regulation of K+ channels. When
nonhydrolyzable GDP analogue, GDPPS was introduced into the cytosol of








Vicia faba guard cells, inward K+ current through the plasma membrane was

activated. On the contrary, the GTP analogue, GTPyS inhibited the inward K+
current. However, in the presence of Ca2+ chelator BAPTA, GTPyS did not
inhibit the current implying cross talk between G-protein signaling and Ca2+
signaling in the regulation of stomatal aperture.


GA and Secretion in Aleurone Cells

The physiological role of endogenous GA is well documented in
germinating cereal seeds. When seeds imbibe water, GA is synthesized by the
embryo and translocated to the cells in the aleurone layer which surrounds

the starch-storing cells in endosperm. GA stimulates the synthesis of
hydrolases from the aleurone. The enzymes are then secreted into the

endosperm where starch breakdown occurs (Gilroy et al., 1993). The most

prominent hydrolase that is secreted upon stimulation by GA is the Ca2+

containing metaloprotein a-amylase. GA is perceived at the plasma
membrane and the signal is transduced by Ca2+-dependent and independent
pathways (Bush, 1996; Gilroy, 1996).

GA induced a sustained increase in cytosolic Ca2+ at the periphery of
the barley aleurone cell protoplasts from 100 nM to above 600 nM. The
increase started 1 to 4 hours after GA treatment and lasted up to 8 hours,

although individual protoplast differed in response kinetics (Gilroy, 1992;

1996). The increase of cytosolic Ca2+ and a-amylase secretion induced by GA
was dependent on extracellular Ca2+. Gilroy and Jones (1993) observed that
calmodulin levels were increased in aleurone layers by GA treatment before
hormone-induced a-amylase synthesis and secretion. Application of
calmodulin stimulated Ca2+ transport into ER of aleurone cells as observed in
those cells treated with GA implying calmodulin as a possible mediator in GA








signal transduction pathways. Wheat aleurone cells showed similar levels of

cytosolic Ca2+ increase, but the response was faster; the Ca2+ increase was
initiated within a few minutes and fully developed after 0.5 to 1.5 hours
(Bush, 1996). The increase in cytosolic Ca2+ induced by GA correlated with the
GA responses of barley aleurone protoplasts; changes in vacuolar
morphology, activation of a-amylase gene transcription, and amylase
secretion. However, mimicking the prolonged cytoplasmic Ca2+ increase by
microinjection or activating caged Ca2+ instead, did not mimic GA action in
activation of a-amylase gene. Only a-amylase secretion was blocked when
caged Ca2+ chelator was used in the GA stimulated protoplasts. These results
suggested that GA signal transduction is similar to ABA signaling in that
there appears to be Ca2+-dependent and Ca2+-independent pathways (Gilroy,
1996). The effect of GA on the increase of cytosolic Ca2+ was reversed by ABA
treatment and resulted in inhibition of a-amylase secretion (Gilroy, 1992).
However, microinjection of Ca2+/calmodulin into the barley aleurone
protoplasts blocked ABA inhibition of all of the GA-induced responses, i. g.,
both a-amylase secretion and gene transcription (Gilroy, 1996). These results
suggested involvement of other signaling elements integrated into the ABA-
induced inhibition of Ca2+-dependent and Ca2+-independent GA action.

Mechanical Signaling

Mechanical signals exert significant effects on the development and
morphology of plants. Externally applied mechanical signals such as touch
and wind induce immediate elevation of cytosolic Ca2+ in plants. Knight, et
al. (1991, 1992) demonstrated the effect of mechanical signals on changes of
cytosolic Ca2+ by utilizing transgenic tobacco plants expressing the Ca2+-
sensitive protein aequorin. Aequorin is a bioluminescent protein found in








the jelly fish Aequorea victoria. It consists of an apoprotein with a molecular
mass of 22 kDa and a small hydrophobic lumiphore, coelenterazine. Upon

binding Ca2+ aequorin emits a photon of blue light and the coelenterazine
become oxidized and inactivated (Blink et al., 1989). The aequorin-coding
sequence was fused to the cauliflower mosaic virus (CMV) 35S promoter and
constitutively expressed in tobacco plants (Knight, et al., 1991). Over 97% of
aequorin was found in soluble fractions of homogenates of transgenic
seedlings. When cotyledons of transgenic tobacco seedlings were touched and
monitored in a luminometer, an intracellular Ca2+ spike was observed with
each touch. A transient increase of cytosolic Ca2+ was also induced in
transformed seedlings by stimulating them with blasts of air from a syringe.
The mesophyll protoplasts and epidermal strips isolated from the transgenic
tobacco plants responded by increasing cytosolic Ca2+ up to 10 PM when
stimulated by injecting an isotonic medium (Haley et al., 1995).
Stimulation by wind or touch brings about the expression of TCH genes
(Braam and Davis, 1990). These TCH genes encode calmodulin and
calmodulin homologous proteins. Arabidopsis plants stimulated by touch

showed inhibited elongation of petiole and bolt compared to unstimulated
plants. TCH gene expression was also induced within 10 to 30 minutes by
treating cultured root cells of Arabidopsis with 100 mM CaCl2 (Braam, 1992).
Although changes in intracellular Ca2+ may transduce the mechanical
signals, the mechanism whereby touch and wind stimuli modify plant
growth and development is not understood.

Red Light Response

Red light induces transient increase of cytosolic Ca2+ and leads to the
swelling of etiolated wheat leaf protoplast (Shacklock et al., 1992). This








response can be mimicked by photorelease of caged Ca2+ or caged InsP3.
Transient elevation of cytosolic Ca2+ that lasted less than 1 minute was
sufficient to induce the increase of protoplast volume by 20% within 10
minutes. When red-light irradiation was followed by a subsequent far-red
light treatment, the effect of red-light on the transient Ca2+ change was
reversed. Fallon et al. (1993) observed that red-light exposure increases
phosphorylation of proteins in etiolated wheat leaf protoplasts and this
response is also induced by photolysis of caged Ca2+. Protein phosphorylation
was enhanced by increasing extracellular Ca2+, but decreased by increasing a
Ca2+ chelator, EGTA, (ethylene glycol bis(3-aminoethyl ether)-N,N,N",N"-

tetraactic add).
Using phytochrome deficient tomato cells, Neuhaus et al. (1993)
proposed that phytochrome signal transduction is mediated by
Ca2+/calmodulin-dependent and -independent pathways. Phytochrome is a
soluble protein with a chromophore which absorbs red and far-red light. This
photoreceptor exists in two forms; Pfr (far-red absorbing form) and Pr (red
light absorbing form), which are interconvertible by the appropriate
irradiation (Quail, 1991). A phytochrome-deficient tomato mutant was able to
develop the maturation of chloroplasts and biosynthesis of anthocyanin
pigment when microinjected with activated G-protein (Neuhaus et al., 1993).
However, Ca2+ and calmodulin could induce only the development of
chloroplasts while cyclic GMP could trigger anthocyanin synthesis in the
phytochrome-deficient tomato mutant (Bowler, et al., 1994). These mediators
of the phytochrome signal transduction pathway are shown to modulate the
expression of light responsive genes.








Ca2 Oscillations

In animal systems, Ca2+ signaling has been categorized in two groups:
signaling in electrically excitable cells, such as nerve and muscle, and in
inexcitable cells, such as epithelial or blood cells (Clapham, 1995; Ghosh and
Greenberg, 1995; Hardie, 1996; Putney, 1993). Both groups of cells utilize
mobilized Ca2+ from Ca2+-sequestering compartments inside cells. In

excitable cells, Ca2+ signaling is mediated by so called, Ca2+-induced Ca2+-
release (CICR) mechanism whereas in inexcitable cells, it is mediated by the
capacitative Ca2+ entry mechanism.
In inexcitable cells, Ca2+ signaling is typically a biphasic process:
extracellular signals stimulate an intracellular organelle to release stored Ca2+
into the cytoplasm, and this Ca2+ efflux from intracellular organelles induces
more Ca2+ influx into the cytoplasm from the extracellular space (Putney,
1993; Clapham, 1995). G-protein coupled receptors or receptor protein kinases
lead to activation of phospholipase C on the plasma membrane upon agonist
binding. Activated phospholipase C generates InsP3 which in turn, diffuses to
specific receptors on the ER. The InsP3 receptor is a ligand activated, Ca2+-
selective channel. The InsP3 receptor releases Ca2+ upon binding to InsP3
inducing transient cytoplasmic Ca2+ increase and depletion of the
intracellular Ca2+ store. The empty Ca2+ store generates a retrograde signal
that activates Ca2+ influx across the plasma membrane resulting in
capacitative Ca2+ entry (Putney, 1993). The capacitative Ca2+ entry is also
called, store-operated Ca2+ entry. An electrical current associated with this
entry is called the "calcium release-activated calcium current" (ICRAC). ICRAC
was shown to increase by a novel calcium influx factor, CIF (Clapham, 1993;
Randriamampita and Tsien, 1993; Parekh, et al., 1993). CIF is a small non-








peptide molecule generated by cells whose Ca2+ stores have been depleted and
move from organelle to cytoplasm. It can even move across the plasma

membrane and stimulate adjacent cells. However, the identification of CIF
still remains elusive. The mechanism by which the depleted stores signal the
Ca2+-influx channels (store-operated channels) is not clear, but the store
operated channels in inexcitable cells are not regulated by membrane

potential unlike the Ca2+ channels in excitable cells.
In excitable cells, the InsP3-triggered mechanism or the CICR mode of
the signaling pathway may operate. However, virtually all excitable cells
contain voltage-activated Ca2+ channels and many excitable cells express the
ryanodine receptor. Ryanodine receptors are gated either by
electromechanical coupling to the dihydropyridine receptor of plasma
membrane in skeletal muscle, or by Ca2+, or by cyclic ADP-ribose (cADPR) in

some other cell types. Depolarization of the cytoplasmic membrane activates

voltage-dependent Ca2+ channels thus enabling the flood of Ca2+ across the
membrane. However, the voltage-dependent Ca2+ channel activity is also

time-dependent. Ca2+ entering through voltage-dependent channels may

directly activate ryanodine receptors to release Ca2+ from intracellular stores

via CICR (Clapham, 1995). cADPR activates directly the cardiac but not
skeletal ryanodine receptor and calmodulin can modulate the action of
cADPR on ryanodine receptor activation (Ghosh, and Greenberg, 1995).
In excitable cells, CICR amplifies the magnitude and spatial distribution
of the transient Ca2+ signal for rapid, all-or-none responses. In inexcitable
cells, the capacitative Ca2+ entry amplifies the duration of the Ca2+ signal,
leading to sustained or tonic responses. When cells are activated through the
capacitative mechanism, the level of Ca2+ often oscillates with a frequency
that varies depending on the concentration of agonist. This regenerative








mechanism enables each Ca2+ pulse to spread throughout the cytosol as a
Ca2+ wave. It is thought that such frequency-encoded Ca2+ pulses may convey

more information than the simple static increase of cytosolic Ca2+ (Putney,

1993).
In plants, Ca2+ oscillations have been recently observed in the

responses stimulated by nodulation factors (Ehrhardt et al., 1996), osmotic
changes (Taylor, et al., 1996), and pistil S proteins (Franklin-Tong et al., 1996).

Cytosolic Ca2+ oscillation induced by extracellular Ca2+ in guard cells of
Commelina communis (McAinsh et al., 1995). Allen et al. (1995)
demonstrated the presence of both InsP3 receptors and ryanodine receptors in

plants. Vacuolar Ca2+ was released in response to ryanodine and cADPR.
Patch-clamping data showed that the cADPR receptor was voltage sensitive,

and spontaneous inactivation of the receptor was not observed. Moreover,

ryanodine receptors and InsP3 receptors were shown to colocalize in the
vacuolar membrane. It is suggested that the inactivation of these channels

may depend on the ligand metabolism in plants due to the high volume of

the vacuoles (-80%) which contains milimolar Ca2+, unlike in animal cells
where the termination of the ligand-gated Ca2+ signal arises from the

depletion of internal stores.
Ca2+ signaling is shown to be involved in mediating the self-

incompatibility response in pollen of Papaver rhoeas (Franklin-Tong.et al.,

1995). Self-incompatibility is a mechanism which regulates the acceptance or
rejection of pollen on the pistil. It is controlled by genes at the S-locus. Self
fertilization is inhibited when pollen carries identical S-allele to that of pistil.
Transient elevation of cytosolic Ca2+ was observed when challenged with self-
incompatibility (S-) protein and was followed by inhibition of pollen growth.
The increase of cytosolic Ca2+ was spatially localized in the intracellular








region associated with the nucleus and ER. Using photoactivated caged InsP3,
Franklin-Tong et al. (1996) demonstrated that InsP3 could induce

intracellular Ca2+ release starting behind the tip of the pollen tube and
expanding toward the tip as a form of Ca2+ wave. The cytosolic Ca2+ increase
induced by InsP3 showed slow transient kinetics increasing for 5 minutes and
lasting about 6 minutes. In parallel, Ca2+-dependent protein phosphorylation
of a 26 kD pollen protein was induced by the self-incompatibility response
implying the involvement of a Ca2+-regulated protein kinase in this
signaling (Rudd et al., 1996).
In response to compatible Rhizobium nodulation factor, alfalfa root
hairs showed an asymmetric Ca2+ oscillation in the form of baseline spikes
(Ehrhardt, et al., 1996). These Ca2+ spikes initiated 9 minutes after the root
hairs were treated with Nod factors and continued up to 3 hours. The
initiation of the Ca2+ elevation and spiking pattern were observed in the
nucleus region and propagated into the cytoplasm toward the root hair tip.
The Ca2+ spike induced by nodulation factor was specific to compatible host
plants, e. g., a nonlegume plant or alfalfa mutant lacking the nodulation
response failed to show changes in cytosolic Ca2+. In animals, the frequency
but not the amplitude of baseline Ca2+ spikes is determined by agonist
concentration. On the other hand, the agonist concentration regulates the
amplitude but not the frequency of sinusoidal Ca2+ oscillation. These
phenomena, which are observed in many cell types and in response to
different stimuli are thought to be regulated by distinct Ca2+-dependent
processes (Thomas et al., 1996).








Regulation of Cytosolic Ca2+

Like other organisms plant cells maintain the cytosolic free Ca2+ at very
low concentration under resting conditions (Bush et al., 1996). The transport
of Ca2+ across cellular membranes are controlled by Ca2+-influx and -efflux
transporters. Increase in cytosolic Ca2+ induced by a stimulus is accomplished
either by an influx of extracellular Ca2+ through Ca2+ channels in the plasma
membrane or by the release of Ca2+ from intracellular Ca2+ stores.
The role of Ca2+ as a second messenger is the transmission of
extracellular signals perceived by localized receptors to other parts of the cell
where effector molecules of the response reside. The increased Ca2+ in the
cytoplasm binds Ca2+ modulated proteins and activates them which, in turn
results in the regulation of target protein activities to generate specific
responses. The signal transduction cascade mediated by cytosolic Ca2+ starts
from the perception by the cell. This signal perception leads to the regulation
of Ca2+ transporters and results in changes of an intracellular Ca2+ level. It is
well known in animal systems that the generation of intracellular Ca2+ signal

is often regulated by a second messenger InsP3. In response to many stimuli,
both InsP3 and diacylglycerol are formed, and the InsP3 is released into
cytoplasm to mobilize Ca2+ from internal stores (Berridge, 1993). The
formation of InsP3 is mediated by G-protein-linked receptors or by receptor
tyrosine kinases. Signaling cross-talk between the second-messenger
generating systems generate the complexity of Ca2+ signals.

G-protein Mediated Cytosolic Ca2 Regulation

In animal cells, G-proteins couple the receptors and the signaling
systems which generate various second messenger in order to induce specific








cellular responses upon various stimuli and often induce changes in cytosolic

Ca2+. Many plasma membrane receptors when activated, generate more than

one second messenger involving cyclic AMP and/or InsP3 by activating
multiple G-proteins. Heterotrimeric G-proteins are part of the larger GTPase
superfamily that also includes small GTP-binding proteins. Heterotrimeric G-

proteins consist of an a subunit which binds and hydrolyzes GTP, and P and y
subunits. The P and y subunits forms an inseparable, functional monomer.

When GDP is bound, a subunit associates with the 3y subunits forming an
inactive heterotrimer that binds to the receptor. Upon stimulation, the
receptor becomes activated and changes its conformation. The GDP-bound a
subunit then responds with a conformational change that favors GTP
binding. Once bound to GTP, the a subunit dissociates from py subunits and

the receptor, and hydrolyzes GTP. The free a and py subunits each activate

target effectors. GTP hydrolysis is a time control that determines how long

both activated a and py subunits last (Neer, 1995; Rens-Domiano and Hamm,

1995).
Helerotrimeric G-proteins are an important component of signal

transduction pathways since they transfer information perceived at the cell

surface to the downstream effector molecules such as adenylate cyclases and

phospholipase C (PLC). PLC hydrolyzes phosphatidylinositol phosphates to

generate InsP3 and diacylglycerol. InsP3 then induces release of intracellular
Ca2+ and Ca2+ further modulates the activities of PLC (Berridge, 1993). Also

the fact that both a subunit and py subunits can modulate PLC greatly increase
the potential complexity of signal transduction. These multiple signaling
cross-talks make defining steps in signal transduction difficult. However,
these mechanisms provide complex and fine modulation of cytosolic Ca2+ for
divers cellular responses (Fyfe and Roberts, 1995).








In plants, the possible involvement of heterotrimeric G-proteins has

been implicated in diverse physiological processes. Among these is the
proposed G-protein regulation in guard cell physiology (Armstrong and Blatt,
1995; Fairley-Grenot and Assmann, 1991; Kelley, et al., 1995; Lee et al., 1993).
These authors used GTP analogues to monitor their effect on K+ channel
activity or stomatal opening. However, the results are not consistent and
quite complex rendering them difficult to interpret. For better understanding
of the role of G-protein in plants, isolation of the genes especially expressed in
guard cells and biochemical characterization seem to be prerequisite

(Assmann, 1996).


Ca2 Channels

In animals, it is well documented that Ca2+ signals are generated from

two sources; Ca2+ influx across the plasma membrane and Ca2+ release from
intracellular stores (Berridge and Dupont, 1994; Bootman and Berridge, 1995;
Clapham, 1995; Miyazaki, 1995). Ca2+ entry across the plasma membrane is

through the opening of plasma membrane Ca2+ channels that are voltage-

gated or receptor-operated. Ca2+ efflux from intracellular stores is controlled
by two types of Ca2+ release channels; InsP3 receptor channels and ryanodine
receptor channels. The coordination of these channels provide spatial and

temporal regulation of the elevation in cytosolic Ca2+ levels and can give rise

to highly localized Ca2+ signals; periodic opening and closing of the calcium
channels brings about repeated Ca2+ spikes or oscillations and Ca2+ waves.
The InsP3 receptor forms a homotetramer composed of subunits of
approximately 310 kDa, and one InsP3 binds each subunit. The primary
sequence of the InsP3 receptor shares no homology with proteins forming the
Ca2+ channels on the plasma membrane but shares partial homology with the








ryanodine receptors in muscles. Ligand binding and regulatory domains
reside in the cytoplasmic region. This region is enriched in basic residues and
binds heparin. The transmembrane domain has six or eight membrane-
spanning regions. Various isoforms are result from alternative splicing and
separate genes. ATP binds the receptor and enhances the channel activity.
The activity of the receptor is also modulated by protein kinase A, protein
kinase C, and calmodulin-dependent protein kinase II. Ca2+ itself has been
shown to bind the receptor. InsP3-induced Ca2+ flux is described by a bell-
shaped curve depending on the concentration of Ca2+; i. e., at low and high
Ca2+ levels, the InsP3 receptor is relatively insensitive to InsP3 suggesting that

InsP3-induced Ca2+ release is regulated by Ca2+ (Mikoshiba, 1993).
The ryanodine receptor-Ca2+ channel complex consists of a
homotetramer of 550 kDa, and several isoforms have been identified. Ca2+ is
the primary activating ligand of ryanodine receptors in skeletal muscle (RY1)
and cardiac muscle (RY2) among other modulators such as ryanodine,
caffeine, cyclic ADP-ribose, ATP, and calmodulin. However, the mechanisms
governing the activation of ryanodine receptors are isoform and cell type
dependent. RY1 is gated by electromechanical coupling to the plasma

membrane dihydrophyridine receptor, and RY2 is gated by Ca2+-induced Ca2+
release (Sitsapesan, et al., 1995). The activities of ryanodine receptors are also
regulated by phosphorylation by protein kinase A (Valdivia, et al., 1995) and

calmodulin-dependent protein kinase II (Takasawa, et al., 1995).
Observations suggesting the presence in plants of the Ca2+ oscillators
similar to those found in animals are accumulating (Bush, 1995). Three types
of Ca2+ channels have been identified by their electrophysical characteristics
in plants; voltage-gated, ligand-gated, and mechanically operated. Voltage-
gated Ca2+ channels are present both in the plasma membrane (Pineros and








Tester, 1995; Thuleau et al., 1994) and in the vacuolar membrane (Allen and
Sanders, 1994, 1995; Johannes and Sanders 1995; Ward and Schroeder, 1994).
InsP3-gated and cyclic ADP-ribose-gated Ca2+ channels are present in the
vacuolar membrane (Allen et al., 1995). Biswas et al. (1995) recently purified
a protein from microsomal fractions of mung bean using heparin affinity
chromatography which was shown to be reconstituted to yield InsP3-gated
Ca2+-release activity. It forms a homotetramer (110 kDa monomer) and is
markedly different from its mammalian counterpart (-250 kDa). The
presence of InsP3-elicited Ca2+ release channels at plant vacuoles may imply a
role for InsP3 as a second messenger in plants. At present there is no
molecular evidence for the existence of a ryanodine-like Ca2+ release channel
in plants. In storage roots of red beet, cyclic ADP-ribose has been shown to
elicit Ca2+ release from vacuole-enriched microsomes (Allen et al., 1995).
Muir and Sanders (1996) showed that the Ca2+ release induced by cyclic ADP-
ribose from red beet microsomes is comparably sensitive to modulation by
ryanodine receptor agonists and antagonists as shown in animal cells. In

addition to the ligand-gated Ca2+ release channels, another class of Ca2+
channel resides at the vacuolar membrane. These channels, which are
apparently not identified in animals, are activated by hyperpolarization of the
vacuolar membrane (Allen and Sanders, 1994; Johannes and Sanders, 1995).
However, the physiological role of these channels is not known.
Depolarization-activated (voltage-gated) channels in vacuoles are
known as the slowly activating vacuolar channels (SV channels), which
activates in response to the depolarizing potentials over one hundred
milliseconds (Allen and Sanders, 1995; Ward and Schroeder, 1994). SV
channels are gated open by a physiological range of cytosolic Ca2+ (0.1-1 JlM) as
well as by depolarization (Sanders et al., 1995). These channels seem to be








regulated by calmodulin (Bethke and Jones, 1994) and calcineurin (Allen and
Sanders, 1995). The opening of SV channels in barley aleurone cells was
sensitive to cytosolic Ca2+ in a range of relatively high concentration (600
nM-100 gM) and the activity was inhibited by calmodulin inhibitors (W7 and
trifluoperazine). Adding calmodulin reversed partially the effects of the
inhibitors. Calmodulin also sensitizes the channel to cytosolic Ca2+ in a
range of 2.5 to 10 jM (Bethke and Jones, 1994). GA treatment of aleurone
protoplasts increased the specific current compared to protoplasts treated with
CaCl2 or ABA. The activity of SV channels from broad bean guard cells was
not modulated by calmodulin (Allen and Sanders, 1995). But it was
potentially modulated by calcineurin, i. e., at high concentration calcineurin
showed strong inhibition of the channel activity. It was suggested that
modulation of SV channel activity by calcineurin may play a role as feedback
inhibition. The SV channel could mediate CICR because the channel can be
activated by increases in cytosolic Ca2+ (Ward and Schroeder, 1994) and
vacuoles contain high concentrations of Ca2+ which is virtually
inexhaustible. Therefore, tight regulation of CICR will be required in plants.
The model proposed by Allen and Sanders (1995) is that the initial rise of
cytosolic Ca2+ opens SV channels, which then releases Ca2+ that activates
Ca2+/calmodulin dependent phosphatase, calcineurin, leading to the feed
back inhibition of the channels. The presence of different types of Ca2+
release channels in plant cells may contribute to the generation of complex
temporal and spatial Ca2+ signals in response to various stimuli similar to
those demonstrated in animal cells.








Ca2 Modulated Proteins

Elevation of cytosolic Ca2+ is detected by Ca2+ sensors that play roles in
altering activities of enzymes, pumps, and other targets. Ca2+ binding
proteins are known as EF-hand proteins, although not all Ca2+ binding
proteins contain an EF-hand motif (Kretsinger, 1996). To date there are 41
known subfamilies of EF-hand proteins (Kretsinger, 1996). The structural
motif of the EF-hand was first discovered from the crystal structure of
parvalbumin (Kretsinger, 1987). Parvalbumin contains three calcium-binding
domains designated AB, CD, and EF from N-terminus to C-terminus. The
name EF-hand (or calmodulin fold) is originated from the domain EF which
forms a a-helix-loop-a-helix structure. The Ca2+ is coordinated by the loop
which contains 12 amino acid residues and provides oxygen atoms from side
chains of five amino acids to coordinate the Ca2+. Five oxygen atoms are
provided by residues 1, 3, 5, 7, and 9, and two are provided by a glutamate at
position 12 (Babu et al., 1988). Residue 7 coordinates Ca2+ by a peptide
carbonyl group. The EF-hand usually works as a pair, but different EF-hand
subfamilies contain different numbers of EF-hands ranging from two to eight.
Ca2+ binding proteins are classified as trigger or buffer proteins based

on their affinity for Ca2+ (Ikura, 1996). Buffer proteins have higher affinity
than sensor proteins; thus, they are thought to respond to even a very small
fluctuation of cytosolic Ca2+ and sequester Ca2+ to maintain a low cytosolic
Ca2+ concentration. However, a possible trigger function for buffer proteins
remains to be elucidated. The best known Ca2+ trigger protein is calmodulin,
and its physiological roles are almost as diverse as those of Ca2+ (Cheung,
1980). Calmodulin (17 kDa) is involved in cyclic nucleotide metabolism,
phosphorylation cascade, cell division, Ca2+ transport system, and








dephosphorylation in animals (James et al., 1995). The main protein targets

of calmodulin are phosphodiesterase, adenylate cyclase, calmodulin-
dependent kinase I and II, elongation factor kinase, myosin light chain kinase,
phosphorylase kinase, plasma membrane Ca2+ pump, and calidneurin. Other
Ca2+ sensors include troponin C, calcineurin B, myosin light chains,

recovering, S100 proteins, and visinin. The molecular size of these proteins

range from 18 kDa to 22 kDa and contain 2 or 4 EF-hands (Ikura, 1996). Ca2+

buffer proteins include parvalbumin, calbindin D, and calretinin. Structural
studies showed that Ca2+-binding proteins that undergo large conformational
changes upon binding Ca2+ are all known to have a trigger function in the
activation of target proteins. In contrast, Ca2+ buffer protein calbindin D
showed little conformational change in response to Ca2+-binding. Troponin
C contains 2 pairs EF-hands comprising N-terminal and C-terminal domains
with each pair. The C-terminal domain of troponin C has high affinity and is
therefore always occupied by Ca2+. This domain is referred to the "structural
domain" in contrast to the N-terminal domain which serves as a regulatory
domain because it has low affinity and triggers Ca2+ signal leading to a specific

response. Calmodulin shows a closed conformation in the Ca2+-free state and

an open conformation in the Ca2+-bound stated. When bound Ca2+

calmodulin exposes its hydrophobic core (Yawaza, et al., 1987). This

hydrophobic region is essential for the interaction with target proteins (Ikura,
et al., 1992; Meador, et al., 1992, 1993). In plants, calmodulin and calcium-
dependent protein kinase (CDPK) are the best known Ca2+ modulated
proteins. Only few proteins have been reported to be regulated by
calmodulin. Among those are NAD kinase, NTP kinase, glutamate
decarboxylase, and Ca2+ ATPase both in plasma membrane and ER (Roberts
and Harmon, 1992). CDPK is a large gene family of which activity is regulated





23

by Ca2+ but not by calmodulin. CDPK has been implicated in many
physiological processes in plants (Roberts and Harmon, 1992). To understand
physiological roles of CDPK in plants identifying individual isoforms and

studying their biochemical properties are essential.













CHAPTER 2
CLONING OF cDNA, HETEROLOGOUS EXPRESSION, AND
CHARACTERIZATION OF CDPK3 AND y FROM SOYBEAN AND THEIR
COMPARISON TO CDPKa

Introduction

CDPKs (Ca2+-dependent protein kinases or calmodulin-like domain
protein kinases) are a large family of protein kinases regulated by calcium but
not by calmodulin (Hrabak et al., 1996; Roberts and Harmon, 1992). In plants,
CDPK was first described (Harmon et al., 1987), highly purified (Putnam-
Evans et al., 1990), and cloned (Harper et al., 1991) from soybean. CDPK is also
characterized in protests. Two CDPK homologs were purified from
Paramecium (Gunderson and Nelson, 1987; Son, et al., 1993) and molecular
cloning of another from Plasmodium has been reported (Zhao, et al., 1993).
The biochemical properties of CDPKs purified from soybean (Harmon, et al.,
1987; Putnam-Evans et al., 1990) and Paramecium (CaPK-1 and -2) (Gunderson
and Nelson, 1987; Son, et al., 1993) were similar in some aspects. Their
molecular masses were 50 to 52 kDa, and free Ca2+ (Ko.5 of -0.2 iM to 2 1iM)
was required for the enzyme activity and autophosphorylation. In contrast,
these CDPKs have different substrate specificities. CaPK-1 and -2 have strong
preference for casein over histone H1, but CDPK purified from soybean does
not phosphorylate casein. Also, unlike the soybean enzyme, Paramecium
CDPK required DTT (1,4-dithio-DL-threitol) for enzyme stability and could
not use Mg2+-GTP as a phosphate donor.









The catalytic domain of soybean CDPK (isoenzyme a) is related to that
of Ca2+/calmodulin dependent protein kinase type II (CaMK II). The C-
terminal regulatory domain, like calmodulin, has four EF-hands. CDPKs also

contain a short junction domain that links the catalytic and calmodulin-like
domains and functions as an autoinhibitor (Harmon et al., 1994; Harper et al.,

1994; Yoo and Harmon, 1996).
cDNAs encoding CDPKs have been cloned from carrot (Suen and Choi,
1991), Arabidopsis (Harper et al., 1991; Hong et al., 1996; Hrabak et al., 1996;
Urao et al., 1994), rice (Breviario et al., 1995; Kawasaki et al., 1993), maize
(Estruch et al., 1994; Takezawa, et al., 1996a), and mung bean (Botella et al.,
1996). All of these CDPKs contain catalytic, junction, and calmodulin-like
domains, and the amino acid sequence identity between CDPK isoforms
ranges from 50% to 95%. Recently two protein kinases have been described
that have catalytic domains related to those of CDPKs, but differ in their
regulatory domains. The CDPK-like protein kinase from carrot has no

predicted functional EF-hands in its carboxyl terminal domain (Lindzen and

Choi, 1995). CCaMK from lily anthers has a regulatory domain that contains

three EF-hands which are more similar to visinin than to calmodulin (Patil et

al., 1995). Its activity, unlike that of CDPKs, is stimulated by calmodulin (Patil
et al., 1995; Takezawa et al., 1996b).

In animals, cellular responses to calcium are brought about in part by
two families of protein kinases; protein kinase C (PKC) and the calmodulin-
dependent protein kinases. Activation of conventional PKCs (isoforms a, p,
and y) is dependent upon phosphatidyl serine, diacylglycerol, and the binding
of Ca2+ to the C2 domain. Activation of the calmodulin-dependent protein
kinases (phosphorylase kinase, myosin light chain kinases, and CaMKs I-IV)
occurs through the binding of calcium to calmodulin. PKCs and most of the









calmodulin-dependent kinases are multifunctional enzymes that have broad
substrate specificity (Dekker and Parker, 1994; Schulman, 1993). They are
widely distributed in various tissues and cell types, and they are activated by a
variety of external signals. An important question under active investigation
is how different stimuli that act through calcium as the second messenger
and a common set of calcium-regulated kinases are able to bring about specific
responses. One answer suggested by recent data is that cellular signals direct
tile translocation and compartmentalization of specific PKC isotypes and thus

target them to the locations of substrates involved in specific physiological
functions (Mochly-Rosen, 1995).
In plants, intracellular free Ca2+ concentrations change in response to

diverse stimuli such as growth regulators, red light, cold and salt stress, and
fungal elicitors (Bush, 1995). To date no homologs of PKC and only one

calmodulin-stimulated protein kinase (Patil et al., 1995; Takezawa et al.,
1996b) have been characterized. It is likely that CDPKs are involved in
mediating many of the diverse responses of plant cells to calcium. Specific

roles for the CDPK isozymes in these responses could arise from differences
in their biochemical properties, Ca2+ sensitivity, tissue distribution,
compartmentalization, and substrate specificity. While the number of genes

encoding CDPKs is growing, information regarding the properties of CDPK
isoenzymes from a single plant species, and data linking CDPK isozymes to

specific cellular responses are largely lacking. A single report made by Sheen
(1996) showed that one CDPK isoform, but not others, is involved in
expression of a reporter gene in response to stress.
To gain insight into how CDPK may play roles in transducing Ca2+
signals, I have undertaken a study of the biochemical properties of CDPK
isoforms from soybean. cDNA clones encoding two new soybean CDPK








isoenzymes, CDPKp and CDPKy were isolated and sequenced. These were

expressed in Escherichia coli (E. coli) and highly purified. The biochemical

and kinetic properties of the new CDPKs and the previously described CDPKa,
which is encoded by cDNA done SK5 (Harper et al., 1991), were compared.

Experimental Procedures


Materials

A soybean plumule cDNA library was generously provided by E.
Czarnecka-Varner and W. B. Gurley, University of Florida. A cDNA library

made from soybean cell suspension culture was a gift from Drs. R. Tenhaken

and C. Lamb, Salk institute. Membranes (Hybond-N+ and -N) for plaque and

RNA blotting were purchased from Amersham. Syntide-2 and

oligonuceotides were synthesized by the Protein Chemistry and DNA
Synthesis Core Laboratories, University of Florida. Autocamtide-2, skeletal
and smooth muscle myosin light chain kinase substrates were purchased

from BACHEM California. Protein kinase inhibitors were purchased from

Calbiochem (H8, ML7, KN62, and staurosporine), Kamiya biochemical

company (calphostin C, and K-252a), and from Seikagaku America, Inc.
(KN93).


Plant Materials

Soybean cell suspension cultures (Glycine max L.) were maintained
and prepared as described previously (Harmon et al., 1996). Soybean seeds

(Glycine max L. cv. Cobb) were imbibed for 4 hours, disinfected in 10%
Clorox for 10 minutes, rinsed several times with water, sown in perlite, and
grown for 3 weeks in a green house.









Isolation of cDNA Clones. DNA Sequencing, and Sequence Analysis

Partial cDNAs encoding new CDPK isoforms were isolated by screening
libraries with either cDNA probes or by PCR. A clone encoding CDPKp was
originally identified by screening a soybean plumule cDNA library with SK5
cDNA (Harper et al., 1991) as a probe. Plaque hybridization was performed at
65 OC overnight in 5x SSC (0.75 M NaC1, 75 mM Na3 citrate), 5x Denhardt's
solution (0.1% (w/v) BSA, 0.1% Ficoll, 0.1% PVP), 0.5% (w/v) SDS, and 20
ig/ml salmon sperm DNA according to the membrane manufacturer
(Amersham). Filters were washed twice at room temperature for 10 minutes
with 2x SSC, 0.1% (w/v) SDS, and once with lx SSC, 0.1% SDS at 65 OC for 15
minutes. A 0.9 kb cDNA clone having a unique restriction pattern was
identified. This cone was used to screen a cDNA library from soybean cell
suspension culture at high stringency (wash with 0.1x SSC, 0.1% SDS at 65 C
for 10 minutes). CDPKy was first identified by using a phage lysate of soybean
nodule cDNA library as a template for PCR. Two degenerate oligonudeotide
primers were used. The sense primer corresponded to the coding region for
the conserved DLKPEN in subdomain VI-b of the protein kinase domain and

was provided by Dr. J. Harper. The antisense primer, AS17 (5'-TCTAGAG-

GATCCATC(ACT)A(GAT)(AT)GG(TC)TT(AG)TC(AT)G(GA)(TA)GC-3') was
chosen from highly conserved sequence in the junction domains of CDPKs
from soybean, Arabidopsis and carrot (Harper et al., 1991, 1993; Suen and
Choi, 1991). The resulting 500 bp PCR product was subcloned in pUC 19 and
sequenced to show that it encoded a new isoform. This partial cDNA clone
was used as a specific probe for the isolation of full-length cDNA clones
encoding CDPKy by screening the cDNA library from soybean cell suspension
culture cells.








DNA sequencing was performed manually by the dideoxy method
(Sanger et al., 1977) using Sequenase version 2 (Amersham) according to the
manufacturer's protocol with some modification. Plasmid DNA templates
were first denatured in 1 M NaOH followed by neutralization in 1 M HCI
(Hsiao, 1991). The denatured templates were annealed to a primer, labeled
with [a-35S] dATP, and the polymerization was terminated by adding ddNTP.
The polymerization products were applied to glycerol-tolerant polyacrylamide
gel in 1.43 M Tris, 0.46 M taurine, and 8.6 mM EDTA (ethylenediamine
tetraacetic acid) (Pisa-Williamson and Fuller, 1992). The untreated gel was
covered with plastic wrap, dried, and exposed to X-ray film for 16 to 48 hours.
DNA sequence analysis was aided by DNAStar for the Macintosh.
GenBank Database searches were performed by the Blast Network Service
provided by the National Center for Biotechnology Information through
Interdisciplinary Center for Biotechnology Research at University of Florida.

RNA Analysis

Total RNA was isolated from soybean cell suspension cultures, eight-

day-old seedlings, leaves, stems, roots, shoot-tips, and petioles of three-week-
old soybean plants as described elsewhere (McCarty, 1986). RNA was blotted
onto nitrocellulose membranes (Hybond-N, Amersham) by downward
alkaline capillary transfer followed by electrophoresis in a denaturing
formaldehyde/agarose gel as described by Chomczynski (1992). Hybridization
was performed following standard protocols (Brown, 1992) in 50% formamide
at 42 oC. Specific DNA probes for each isoform (5' untranslated regions from
the cDNA clones encoding CDPKa and y and the 3' untranslated region from
cDNA clone encoding CDPKp) were generated using PCR. After hybridization
the blots were briefly washed several times in 2XSSC, 0.1% (w/v) SDS at room








temperature, twice for 10 minutes each in the same buffer, and finally twice

in 1XSSC, 0.1% (w/v) SDS at 42 OC for 15 minutes. Due to high background,
the blots hybridized with the CDPKp-specific probe were washed further in
0.1XSSC, 0.1% (w/v) SDS at 65 C for 10 minutes.

Construction of Plasmids

E. coli expression vector pGEX-KG (Guan and Dixon, 1991) was chosen

to produce glutathione S-transferase (GST) -fusion proteins of full-length
CDPKa, p, y, and an N-terminal deletion mutant of CDPKy containing amino
acid residues 66-538, CDPKy(66-538). pGST-CDPKa was generated by lighting
the full-length cDNA insert cut from pHisl530 (Harmon et al., 1996) into
pGEX-KG. For pGST-CDPKp, first, the 5'-end of the cDNA encoding CDPKp
was amplified by PCR using 5'-GCTCTAGACCATATGCAGAAGCATGGT-3'

(AS79) and 5'-GCCTTGTATCTGGACAACG-3' (AS62) as primers. The

amplified DNA of 160 bp was purified from an agarose gel for enzyme

digestion with Xbal and HindIII. Next, full-length cDNA clone of CDPKp in
pBlueScript was subjected to enzyme digestion with HindIII and XhoI, and 1.5
kbp fragment was isolated. These two fragments (160 bp and 1.5 kbp) were
then subcloned into XbaI/XhoI double-digested pGEX-KG in one ligation
reaction. For pGST-CDPKy, a fragment of the 5'-end of cDNA cone encoding

CDPKy was amplified by PCR using 5'-GCTCTAGACCATATGGTTACAGAC-

ATGCT-3' (AS77) and 5'-GGAATTCTTAAAGTGTGTGGAACTGCT-3' (AS75)
as primers. The fragment of 115 bp was purified from an agarose gel and was
digested with XbaI and SphI and ligated into the SphI-digested full-length
cDNA clone of CDPKy (1.5 kbp), and then the resulting 1.6 kbp fragment was
subjected to Xbal digestion in order to subclone into XbaI-digested pGEX-KG.
To make the N-terminal deletion mutant of CDPKy, A DNA fragment (200








bp) amplified using the primers 5'-GCTCTAGACCATATGGGTGITAGGCA-
AGAC-3' (AS78) and 5'-TAATCCATGGGTGCTCAA-3' (AS44) was digested
with XbaI and SstI, and subcloned into Xbal-digested pGEX-KG together with
SstI/XbaI double-digested cDNA cone of CDPKy (1.2 kbp). Constructs
expressing active recombinant proteins were selected and confirmed to be
error-free by DNA sequencing. The PCR primers AS77-79 were designed to
contain XbaI and NdeI restriction sites separated by one base pair to keep the
correct reading frames in any subcloning vectors using either sites. The two
bases at the 5' end of these primers were added for the efficient digestion with
XbaI.

In attempts to produce polyclonal antibodies specific to CDPKp and y,
N-terminal sequences of each isoform were chosen to produce GST-fusion
proteins. Because the N-terminal domain of CDPKp was short (23 amino acid
residues), part of catalytic domain was included. pGST-CDPKp was digested
with XbaI and HindIII to generate a 230 bp DNA fragment encoding 60 amino
acid residues of CDPKp N-terminus and subcloned into Xbal/HindIII double-
digested pGEX-KG. The fidelity was confirmed by expression and DNA
sequencing. Since the N-terminal domain of CDPKy contains 83 unique
amino acid residues, this region seemed an excellent choice for raising specific

antibodies. The fusion protein containing this region was created as follows.

As mentioned above, pGST-CDPKy was subcloned into pGEX-KG through

Xbal sites. The insert of pGST-CDPKy contains three XmnI sites and the
vector contains one XmnI site. Digestion with XmnI generates blunt ended
DNA. Because the most 5' side XmnI site in the insert was located at around
amino acid residue 80 of CDPKy, this restriction enzyme was utilized together
with XbaI to produce a 240 bp DNA fragment. Other small sized fragments
(30, 320, 630, 800 bp) resulting from the double digestion were separated from








this fragment in a 2% agarose gel. Due to the limited number of cloning sites

and their orientation in native pGEX-KG, the following modifications were
made. pGST-CDPKp was first digested with XhoI to cut 3'-end of the insert

followed by blunt-ending by Klenow enzyme treatment for 15 minutes at 25
C and for another 15 minutes at 75 C to inactivate the enzyme and then

digested with Xbal to cut out the insert from the rest of the plasmid. The

resulting pGEX-KG vector contained the partial multidoning sites originating

from pBlueScript, and was ligated with the purified 240 bp fragment described
above. The transformants were screened by digesting plasmids with XbaI and
HindIII which cuts the right side of 3' end of the 240 bp insert prior to
expression.

Expression and Purification of GST-Fusion Proteins

A colony of transformed E. coli cells (PR745) was grown overnight at
37 OC in 2 ml LB/ampicillin (0.1 mg/ml) and transferred to 500x dilution into
M9TB/ampicilin (Studier et al., 1990) and further cultured until OD600 0.5.

Expression of recombinant protein was induced by adding IPTG isopropyll 1-

thio-P-D-galactopyranoside) to 0.4 mM at room temperature. Cells expressing

CDPKp were collected and resuspended in lysis buffer containing 1 mM PMSF

(phenylmethylsulphonyl fluoride), 10 gg/ml leupeptin, 20 gg/ml aprotinin, 1
mM DTT, 50 mM Tris, pH 7.5, and 150 mM NaCl and centirifuged after

sonication (Yoo and Harmon, 1996). The supernatant was loaded onto a
glutathione-agarose column equilibrated with 50 mM Tris, pH 7.5, 150 mM
NaCl, and 1 mM DTT, and washed with the same buffer. Protein was eluted
with 50 mM Tris, pH 8.0, 1 mM DTT, and 10 mM glutathione. Fractions
containing kinase activity were pooled, and loaded onto Mono-Q equilibrated
in 20 mM Tris, pH 7.2, 1 mM CaCI2, 2% (w/v) betaine, and 1 mM DTT. CDPKp








was eluted with a gradient of 0-0.5 M NaCI in equilibration buffer. Fractions
containing kinase activity were pooled again, and further purified by another
Mono-Q chromatography in the absence of calcium but in the presence of 2.5
mM EDTA. The procedure for purification of CDPKa, y, and T(66-538) was as
follows. Cell lysis and affinity chromatography were performed in buffer A

(50 mM Tris, pH 7.5, 150 mM NaC1, 10 mM EDTA, and 2 mM DTT). The

debris in the lysed cells was pelleted by centrifugation, and the supernatant
was loaded onto a glutathione-agarose column. The column was washed
with buffer A and protein was eluted with buffer B (50 mM Tris, pH 8.0, 10
mM glutathione, 10 mM EDTA, and 2 mM DTT). Fractions containing kinase
activity were pooled and loaded onto a column of Mono-Q (Pharmacia)
equilibrated in buffer C (20 mM Tris, pH 8.0, 2.5 mM EDTA, 5% (w/v) betaine,

2 mM DTT). Enzymes were eluted with a gradient of 0-0.5 M KCl in buffer C.

Purified recombinant CDPKa, and p were dialyzed against 20 mM Tris, pH 8.0,
14.4 mM 2-mercaptoethanol and stored at -80 OC in 50% glycerol (v/v).
CDPK- was labile to freeze/thaw, so eluted pure fractions were stored at 4 OC.
The N-terminal domains of CDPKp and y expressed as GST fusion

proteins in E. coli were prepared and affinity purified on a glutathione-
agarose column in a same manner as described above. The protein eluates of

both constructs were came off during the wash step when loaded onto the
Mono-Q column. SDS-PAGE of the fractions showed that some higher and
smaller molecular weight contaminants were separated from main protein
bands of expected size for the constructs. Therefore, no further purification
was undertaken.

Concentrations of recombinant CDPKs were determined according to
the method of Bradford (Bradford, 1976) using Bio-Rad dye-binding assay.








Protein Kinase Assays

Enzyme activity assays were performed by a modification of the

procedure described previously (Harmon et al., 1996). The reaction mixture
contained 50 mM HEPES, pH 7.2, 10 mM MgCl2, 1 mM EGTA, 1.1 mM CaC12, 2
mM DTT, 0.1 mg/ml BSA, the indicated amount of substrates, 5 nM

recombinant CDPK, and 60 gM ATP (-500 cpm/pmol). The concentrations of
synthetic peptides (syntide-2, autocamtide-2, and skeletal and smooth muscle
myosin light chain kinase substrates) and histone IIIS in the assays were 100

iM and 0.5 mg/ml, respectively. Kinetic parameters (apparent Km and Vmax)
were determined from double reciprocal plots. Concentrations of synthetic

peptides were determined from amino acid composition analyses carried out
by Protein Chemistry Core Facility, University of Florida. Each kinetic

parameter was determined by averaging at least four independent assay
results.


Determination of ICso Values with Protein Kinase Inhibitors

Stocks of protein kinase inhibitors (1 mg/ml) were prepared in DMSO.
Various concentrations of inhibitors in the enzyme activity assay were
achieved by diluting stock solutions and the final dilutions were made in

0.1% (v/v) DMSO. Control assays (total 50 g1 in each reaction) were done in

the presence of 5 pl of 0.1% (v/v) DMSO. The enzyme activity assay mixture

was the same as described above and contained 100 gM syntide-2 as peptide
substrate. IC50 values were obtained by averaging enzyme activities from at
least four determinations.








Preparation of Hisf-CDPK Constructs and Purification of Expressed Proteins

Recombinant His6-CDPK constructs were generated by digesting pGST-

CDPKY(66-538) with NdeI and XhoI and subsequent subdoning into
NdeI/XhoI double-digested pET-15b. The transformants were screened for the
right insert size by digesting the plasmids with NdeI and Xhol. His6-

CDPKY(66-538) was purified as follows. The expression of transformed cells

and chromatography using a nickel chelation column were carried out

following the procedures described elsewhere (Yoo, et al., 1996). The eluates

were dialyzed against buffer A containing 20 mM Tris, pH 8.0, 2.5 mM EDTA,
and 100 mM KCI and then loaded onto Mono-Q column equilibrated with
buffer A. After washing in buffer A, the protein was eluted with a gradient
from 0.1-0.5 M KCI in buffer A. Fractions were analyzed by SDS-PAGE and
pooled for further purification by phenyl-Superose reverse phase
chromatography. Equilibration buffer contained 20 mM Tris, pH 8.0, 2.5 mM
EDTA, and 1.5 M ammonium sulfate. Following loadings of the pooled
fractions from the Mono-Q column onto a phenyl-Sepharose column, and

washing of the resin with equilibration buffer, the protein was eluted with a

reverse gradient of 100-0%. Highly purified His6-CDPKy resulting from this

chromatography was concentrated using a Centricon-30 (Amicon) filter.


Preparation of Microsomal Membrane

Cultured soybean cells were homogenated in Buffer A containing 2.5%
sucrose, 25 mM Tris, pH 7.8, and 0.5 mM CaCl2 using glass-glass homogenizer
and passed through cheese cloth. The filtrate was centrifuged at 8,000g for 15
minutes and the supernatant was ultracentrifuged at 50,000g for 30 minutes.
The supernatant was saved and the pellet was dissolved in buffer A without








sucrose (buffer B) and centrifuged at 5,000g for 30 minutes. The pellet was
dissolved in buffer B and the protein amount was measured by Bradford assay

(Bradford, 1976).

Other Procedures

SDS-PAGE was performed according to the method of Laemmli
(Laemmli, 1970). Electroblotting was performed in a buffer containing 25 mM
Tris, 192 mM glycine, pH 8.3, and 20% (v/v) methanol for at least 2 hours.
Blotted nitrocellulose membranes were blocked in 5% nonfat milk dissolved
in TBS (20 mM Tris, pH 7.5 and 500 mM NaC1) for 30 minutes and incubated
with the first antibody for 1-2 hours. The blots were washed in TBS
containing 0.05% Tween-20 several times followed by incubation with the
second antibody for 30-60 minutes. The blots were washed again and
subjected to color development in a solution containing 0.15 M NaHCO3, pH
9.6,4 mM MgCI2, 0.1 mg/ml NBT, and 50 ng/ml BCIP. Digital images of
polyacrylamide gels and nitrocellulose blots were obtained as previously
described (Harmon et al., 1994). Oligonucleotides were synthesized at the
Interdisciplinary Center for Biotechnology Research DNA Synthesis Lab,
University of Florida.


Results


cDNA Clones that Encode Two New CDPK Isoforms from Soybean

The nucleotide and deduced amino acid sequences of cDNA clones
encoding CDPK isoforms, CDPKp and CDPKy, are shown in Figure 2-1. The
cDNA encoding CDPKp contains an open reading frame 1470 nudeotides in
length without an in-frame stop codon at the 5' end. The size of the



























Figure 2-1. Nucleotide and deduced amino acid sequences of two new CDPK
isoenzmes from soybean. Nucleotide sequences are shown in the
numbered rows and the predicted amino acd sequences (single letter
code) of the longest open reading frames are shown beneath the
corresponding cDNA sequence. Possible poly (A) signals are
underlined.
a) CDPKp; b) CDPKy.











A. CDPKp

TCGGCTACCTTGGCTATAGCTTACGTGTTCAGAGGGAGAGGAGAGAGGACAAT 60


AGAAGCATTTTGC ATCTAA GCAACGTGTTGCCGTAAAACGGCGGCTAAGGG 120
K H G F A S K RN V LP Q T A R L R D

ACCACACGTTTCTGGGGG CT GGGC MTTCGGGACGACGTACCTGTGCA 180
H Y V L G K K L G Q G Q F G T T Y L C T

CCCACAAGGTGACGGGGAAGCTCTACCGCGTGCAAATCGATCCCGAAGAGGAACTTATGT 240
H V T G K L A C S I P K L M C

GCCAGGAGGATTACGATGATTGTGGAGGGAGATTCAGATCATGCACCATTGTCGGAGC 300
o E D Y D D V W R I I M H L H

ATCCAAACGTTGTCCAGATACAGGCACGTACGAGGATTCCGTGTTCGTGCACCTTTCA 360
P N V v o G T Y E D S v F V L V M

TGGAACTATGT GCTCGCC TTGACAGGATATTTCAGAAGGCATTACAGCG 420
E L C A G G E L F D R I I Q K G H Y S E

AGAGAGGGC TCCAAGTTGATAAAGACCATTGTTGGGGTGGTGGAGGCGTGCCACTCTC 480
R E A A K L I K TI V G V V E A C H L

TTGGGGTCATGCATAGGGATTCTCAGCCTAGAATTTCTTGTTTGATACCCCTGGCGAA 540
G V M H R D L K P E N F L F D T P G E D

ATGCCCAGATGAAGGCCACCGATTTGGCCTTCTCTGTCATTCTACAAGCCAGACAACCT 600
A QM K A T D F G L S V IL A R A F

TTCATGATGTAGTAGGAAGTCCCTACTATGTTGCCCCAGAGGTGTTGTGCAACAATATG 660
H D V V G S P Y Y V A P E V L C K Q Y G

GACCCTGAGTGGACGTATGGAGTCTGGTTTATCCTATACATCTTACTGAGTGGGGT 720
P E V D V W S A G V I L LLS G V P

CACCTTTCTCGGGC'IC1 ACCGAAGCAGGMTTTTCAGACAGATTTAATGGAGATTG 780
P F W A E T E A G I F R Q I L N G D L D

ATTTTGTTTCTGAACCGTGGCCAGTATTCCAGA TGCTAAAGATGGTA AGA 840
F V S E P W P S I S E N A K E L V K Q M

TGTTGATAGGGACCCTAAGA ATTCTTGCTCATGAAGTTTTATGTMACCCTTGGG 900
L D R D P K K R I S A E V L C N P W V

TTGTTTGACATTGACCTGCCACAMCCTGGACTCTC TTTTGCACCGCCTAAAGC 960
V D D I A P D K P L D S A V L T R L K H

ATTTCTCAGCATAATTAACTTAAGAAGATGGCATTACGGGTCATAGCAGAGGCTT 1020
F SAM N K L K K MA L R V I A E R L S

CAGAGGAAGAATAGGTGGATTG GAGTTGTTTAAAATGATTGACACAGACATAGTG 1080
K E E I G G L K E L F K M I D T D N S G

GGACAATAACTTTTGGGC GGGGGAAAAGTGTGGGCTCTAATCTCATGG 1140
T I T F E L K E G L V G S N L M

AATCTGAAATTAAATCACTTATGGAAGCGGCTGATATAGACAACATGGATCAATAGAC 1200
S E I K S L M E A A D I D N N G 5 I D Y

ATGGTGAATTTCTTGCTGCTACACTGCACTTGAATAAGATGGAAAGAGAGGAATTTGG 1260
G E F L A A T L H L N K M K R E K N L V

TTGCCTTTCGCCTATTTTATAAGATGGTAGTGGTTACATCACCATTACGAGCTTC 1320
SAA F A Y F D K D G S G Y I T I D E L Q

ACAGCCTTGTAAGGACTTCAGCCTAGGCTGATGCATC ATTGA ATCATCAGAA 1380
Q A C K D F S L G D V R L D E M I K E I

TTGATCAAGATAATGATGGGAGATTATTATGCGGAGTTMCAGCATGATGA GG 1440
D 0 D N D G R I D Y A E F A A M M K K G

GTGATCCMATATGTCCGAGAGCAGMCCATGACGGCAATTTGAACTTCAATATTGC 1500
D P N M G R S R T M K G N L N F N I A D

ATGCATTTGGMATGAAGACTCTTCGATATGCTACCTGATCTTTTGTACAGTTGGA 1560
A G M K D SS *

TGCATCATGd(TAGACMGAATCATTGTCAATITC~TrATCGCGWTGTTATMTCTGT 1620
TTTGTATTGGMACTTTGTTTACTTTTCTCAGCTAWCCTTTATCCCTTAACTTTACA 1680
C TCAGTGAAGTTGMATATATTCAGCATAATC ATATCAAAATTGGCTTGATGAA 1740
AAAAAAAAAAAA











B. CDPKy

GAACTCCACCTTATGTACTACTAGCATTTATCCCTTTTrTTTTTTTTATACATTCTTTCT 60
CTTTCACCTCTTCTAAGGCTCCACTTTTATAATTCCTTGTAATTTCTTCATCAAAGTAT 120
CAAACTTGTCTTCTCTCTCTCTCCTTTTACTTCCTCTCATATATGATATATTCTTCCAA 180
AATCCTCATCTAAGGGTTAATTTTGTTCCGGACAAGOTTTTATTATTATTATTATA 240
TTTAAATTTCAGAATGATTTAT CATAG TGATCCAGTTTAAATGTTATAGTT 300
TCCTTGAATTCTCATCTGGOTTCACCTAGT ACTTG TT ATGA TCA TTGT A TA 360
GTGGTGATTATCTGATACCAAACATAAAATTCACTCTTTTTCTTGTTTCTTGTCAC 420
AGACTTGGATTTTGAGTCTGGAGACATGGGTTGTCGTGC TAAAGTAGTCAGACCC 40
GCAACACAATGTTACAGACATGCTGGGACTGGTGTGTGCACAACCAAGAAGACTCATG 540
M V T D M L GL V VC T T K K T E

AACCTTTGGTCAACCAATCAAGACCACCTGCAAACAACCTTATCATTTACATGAAAAGC 600
P L V N S A P A N Q P H L H E K H

ATGCWTCATCCACTGCCTAGACAGTGCCTC TATCCATG GCCTCCTGGTCCAG 660
A A S T A O T V P 0 N M P W K P P G P A

CCCTTAGTCCTAAACCCTGTTGTTGGTGTTAGGCAAGACACAATTTTGGGAAGCAGTTTG 720
L S P K P V V G V R D T I L G Q FE

AGGATGTGAAGCAGTTCCACACACTTGGGAAGGAGTTGGGTAGAGGGAATTTGTGT 780
D V K 0 F 8 T L G K E L G R G O F G V T

C ATA T:TTr(;CACTArAGAATTCrACCGGAIrTCATATGCCTGCAGTCCATTTCCCAAGA 840
Y C T E N 5 T G 6 L A A C K S I S R R

GGAATTGCGACCAAATCTATAAGGAGGACATAAACAGGGAGATTCAGATTATGCAGC 900
K A S K S D K E 0 I K R E I 0 M Q H

ATTTGAGTGGTCAACCCAACATTGTTGAGTTCAMGGGCTATCAGGATAGGAGCTCAG 960
L S G Q P N i F K G A ED AR S V

TTCATTTGTGATGGAGCTTTTGT GGGGAACTTTTTGATAGGATTATTGCCAGG 1020
H V V L CA G G E L F D R I K G

GGCATTACAGTGAGAAGGCTGCTGCTTCATTTGCAGACAAATTGTAAATGTTGTTCATA 100
H1 Y S E K A A A S I C R Q I V N V V H I

TCTGTCATTTCATGGGTGTGATGCATAGGGATCGAAACCAGAGAATTTTTTGCTATCTA 1140
C R F M G V M 8 R D L K P 8 N F L L S S

GTAGGGACGAAAATGCACTCTTCAAGGCAACCGATTTTGGCTTGTCAGTTTCATTGAA 1200
R D E N A L L K A T D F G L V I E E

AAGGAAAGTATATCGGGATATAGTTGGTAGTGCTTACTATGTTGCTCCTGAGTTCTGC 1260
GK V Y R D I V GSA Y Y V A P E VL R

GGGCGAGATGTGGGAAAAATAGATATATGGAGTGCAGGTCATATTGTATATCTTAC 1320
R R C GK E I D I S A G V I L Y I L L

TTAGTGGAGTCCCTCCATTTTCGGGCTGGACTGAGAAGGGAATATTTGATGCCATATTGG 1380
S G V P P A E T K I A I L E

GGCACATATGATTTGAAGTCAACCATGCCTAACATCTCAGACATGCCAGGATC 1440
G H I D F E S 0 P w P N I S D s A K D L

TTTTCGTAAGATGCTTATACAGGATCCAAAGA CGCATTACCTCTGCTC TTCT 1500
V R K M L I O D P K K R I T S A 0 V L A

AGCACCCATGGATTAAGATGGAAGCTTCAGACACCGATAACAGTGCAGTCCTTT 1560
R P W I K D G N A S D K P 1 D S A V L S

CCAGAATGAAGCAATTTAGAGCAATGAATAAGCTAAAGAAACTTGCACTGAAGGTCATT 1620
R M K O F R A M N K L 8 E L A L K V I A

CT GAAGTATGTCTCAGAAGAATCCGGTTTGAAGMCAATGTTTACAATATGGACA 1680
E N M S A EE I G L K A T D T

CTACCAGGTGGTACAATCACCTATAAGAACTTAAGTCAGGATTGCATAGTTGGCT 1740
0D S G T I T S E E L K S G L H R L G S

CMAAGCTTACAAGGCTGAAGTGAAGCAACTTATGGAAGCTGCTGATTGA GGAAAT 1800
IALT E E V L M A 0 D G N G


Figure 2-1 continued









40



GCTCAATTGACTACATAGAATTCACACTCTACAATGCATAGACACAAATTAGAAAGAG 1060
S I D Y IE F I T A T M H R H L E R D

ATAAACC TTCGGCCTTCCTATTTTATAAGACAAAGGGGATTAACA 1920
D Q L F A YF D K D N S F I T R

GGAGTGATTGGATCAGCCATCGAA ATATGGTATGGGTGATGATGCAACAATCAAGG 190
D E L E S A M G M G D A T I K E

AAATCATATCTGAGTTATACAATTATATTTCT GTATAC ATGATGTAGAA 2040
I I S E V D T I I S E V D TD D G R I

TCAACTATGAAGAATTCTCTGCGATGATGAAGAGTGGGAACCAACAACAAGGCAAGCTAT 2100
N Y E E F S A M M K S G N 0 0 0 G K L F

TCTAAATCAACCCATCTAACCAGTCTGAGGGCCTGGTATCCAAAACCCCTCTAGAAAAA 2160
GGTTTATACCTCTGAGMGTACTTCCAAACAGTTTGTATCCTCATAGGCATGTAGCT 2220
TCTATAGAATGTTTTTGAGTTACTATGCTCCTCTGTCTTAGTACTCCCTTTTT 2280
TTCATTTCTGGGACCTTTTTTTGCTTTTCACTGAGATTCCATTTCCGAGCCTCTT 2340
TTTGTTTTGGACACAAAGTAITAACTCACATAGAAAAGGTGATATCTCATAAGATTAT 2400
ACAAGTTTTt CGCTTGIAAAAAAAAAAAAAAA


Figure 2-1 continued








transcript observed on an RNA blot probed with CDPKp-specific

sequence is approximately 1.8 kb (data not shown) which is similar to the size

of the cloned DNA (1.75 kb). Attempts to isolate longer cDNA clones from

the cDNA library yielded only clones of the same length, and attempts to
isolate clones containing additional 5' sequence by anchored PCR were not
successful. Taken together, these results suggest that the isolated cDNA

encodes full-length CDPKp. The predicted protein contains 490 amino acid
residues and has a molecular mass of 55 kD.

Four cDNA clones that hybridized to unique DNA probe generated

from PCR using nodule-enriched cDNA library lysate as template were

purified and sequenced. Three of the overlapping clones had inserts ranging
2.1 to 2.4 kb, and the fourth clone was 3.2 kb. Figure 2-1B shows the sequence
of the 2.4 kb clone. This clone and the two smaller ones contained an

identical open reading frame, 1,614 nucleotides in length, that encodes a

protein kinase 538 residues in length having a predicted molecular mass of 60

kD. This protein was named CDPKy. The 3.2 kb clone encoded a predicted

protein almost identical to CDPKy, but also contained an unidentified

upstream open reading frame that was presumed to have arisen as a cloning

artifact.


Deduced Amino Acid Sequences of Soybean CDPK Isoforms

CDPKp and CDPKy share 76% and 58% overall amino acid sequence

identity, respectively, with CDPKa. CDPKp shows highest identity to AtCPK4

(Hrabak et al., 1996) and Atcdpk2 (Urao et al., 1994) (80% and 79%,
respectively). CDPKy shows highest identity to OSCPK2 (Breviario et al., 1995)
and AtCPK9 (Hrabak et al., 1996), (77% and 74%, respectively), and to the








partial sequence of carrot CDPK (Suen and Choi, 1991) (82%, in the 425 residue
overlap).
The amino terminal domains of CDPKa, CDPKp, and CDPKy are 33, 23,
and 83 residues in length, respectively. The sequences of these domains
match no known proteins and they do not contain any known sequence

motifs. There is little similarity in sequence among the amino terminal

domains (10-185 residues in length) of 24 CDPKs in the GenBank database,

except for a high proportion of hydrophilic residues with more being basic
than acidic. The amino terminal domain of CDPKy is about 10% proline, as
are most CDPK amino termini that are greater than 74 amino residues in
length.
Figure 2-2 shows the amino acid sequence alignment and consensus
sequences of the catalytic (Panel A), junction (Panel B), and calmodulin-like
domains (Panel C) of CDPKs from different plant species. All kinase
subdomains and four EF-hands are highly conserved among them. CDPKy
contains additional eight amino acd residues between the third and fourth
EF-hands in the calmodulin-like domain which is the longest insert in that

region compared to the others. Alignment of the sequences by the CLUSTAL

algorithm shows that sequences flanking the fourth EF-hand are the least

conserved regions (Figure 2-2C). These regions of low conservation may
imply functional differences.


RNA Expression Pattern of CDPK Isoforms

Total RNA from different organs of soybean plants or from
suspension-cultured cells was blot hybridized with DNA probes specific for
genes encoding each of CDPK isoenzymes (Figure 2-3). CDPKa, p, and y
transcripts were expressed in roots, leaves, petioles, stems, shoot tips of three-


















Figure 2-2 Alignment of amino acid residues of soybean CDPK isoforms with known CDPKs from other
plants. Multisequence alignment was performed using sequence analysis software CLUSTAL provided
by Genetics Computer Group. Gaps introduced for optimal alignment are indicated by dashes. The
three functional domains are compared separately and the positions of amino acid residues are
indicated in parentheses at the beginning of each panel. Residues and symbols in the consensus
sequence are: capital letters, absolutely conserved residues; lower case letters, residues conserved in
>70% of the isoforms; -, acidic residues; +, basic residues; #, aliphatic residues; $, serine or threonine.
The sequences are CDPKa (Harper et al., 1991), p, and y from soybean;VrCDPK (Botella et al., 1996) from
mung bean (Vigna radiata) ; AtCPK1 to AtCPK8 (Hrabak et al., 1996) and ATCDPK1 and 2 (Urao et al.,
1994) from Arabidopsis; SPK (Kawasaki et al., 1993) and OsCPK 2 and 11 (Breviario et al., 1995) from rice;
ZmCPK (Estruch et at, 1994) from maize; DcCDPK (Suen and Choi, 1991) from carrot
a) Catalytic domains. The twelve conserved kinase catalytic subdomains are designated by Roman
numerals; b) Junction Domains; c) Calmodulin-like Domains. Each Ca2+-coordinating loop of the four
EF-hands is indicated as I to IV.











A. Catalytic Domiain
I II III IV V
Conmnum y-g+-G-gqG-q-Tylct---~g--yACKsI-krkL----D eDvr.EiqIm-hl ----nv-~ ---yED---Vh#vME1C-GGELFdRI--+Ghy Er

CDPKa (34-295) YEVGRKLGQGQFGTTFECTRRASGGKFACKSIPKRKLLCKEDYEVWREIQ IMHLSEHANVVRIEGTYEDSTAVHLVMELCEGGELFDRIVQKGHYSER
CDPKX (24-285) YLGKKLGQGQFGTTYLCTHKVTGKLYACKSIPKRKLMCQEDYDDVWREIQIMHHLSEHPNWQIQGTYEDSVFVHLVMELCAGGELFDRIIQKGHYSER
CDPKY (84-344) HTLGKELGRGQFGVTYLCTENSTGLOYACKSISKRKLASKSDKED:I(REIQIMQHLSGQPNIVEFKGAYEDRSSVHVMELCAGGELFDRIIAKGHYSEK
SPK (73-334) YTLGRELGGQFGKTYLCTEISTGCQYACKTILKSNLRCVSDIEVRREIQ IMHLSGQKNIVTIKDTYEDEQAVHIVMELCAGGELFSKIQKRGHYSER
OsCPK2 (85-346) YSLGKELGRGQFGVTYLCTEIASGKYACKS ISKRKLVSKADKEDIRRE IQIMHLSGQQNIVEFRGAYEKSNVHVVMELCAGGELFDRIIAKGHYSER
OsCPK11 (79-340) YIIGRKLGQAQFGTTYLCTEINTGCEYACKTIPKRKLITKEDVEVRREIQIMHHLSGHKNVAIKDVYEDGQAVHIVMELCAGGELFDRIQEXGHYSER
ZmCDPK (93-354) YSMGKELGRGQFGVTHLCTHRTSGEKLACKTIAKRKLAAREDVDDVRREVQIMHHLSGPNVVGLRGAYEDKQSVHLVbELCAGGELFDRIIARGQYTME
Atcdpkl (11-272) YILGRELGRGEFGITYLCTDRETHEALACKSISKBRKLTAVDIEDVRREVAIMSTLPEHPNVVLKASYEDNENVHLVMELCEGGELFDRIVARGHYTER
Atcdpk2 (26-287) YLLGKKLGQGQFGTTYLCTEKSTSANYACKSIPKRKLVCREDYEDVWREIQIMHHLSEHNVVRIKGTYEDSVFVHIVMEVCEGGELFDRIVSXGHFSER
CPK1 (150-411) YSLGRKLGQGQFGTTFLCVEKTTGKEFACKSIAKRKLLTDEDVEDVRREIQIMHHLAGHPNVISIKGAYEDVVAVHLVECCAGGELFDRIIQRGHYTER
CPK2 (186-447) YSLGRKLGQGQFGTTFLCLEKGTGNEYACKSISKRKLLTDEDVEDVRREIQIMHHLAGHPNVISIKGAYEDWAVHLVMELCSGGELFDRIIQRGHY'ER
CPK3 (77-336) YEFGRELGRGQFGVTYLVTHKETKQQVACKSIPTRRLVHKDDIEDVRREVQIMHHLSGHRNIVDLKGAYEDRHSVNLIMELCEGGELFDRIISKGLYSER
CPK4 (25-286) YLLGKKLGQGQFGTTYLCTEKSSSANYACKSIPKRKLVCREDYEDVWREIQIMHHLSEHPNVVRIKGTYEDSVFVHIVMVCEGGELFDRIVSKGCFSER
CPK5 (97-357) YTLSRKLGGQFGTTYLCTEIASGVDYACKS ISKRKLISKEDVEDVRREIQIMHHLAGHGS IVTIKGAYEDSLYVHIVELCAGGELFDRI IQRGHYSER
CPK6 (85-346) YTLSRKLGQGQFGTTYLCTDIATGVDYACKSISKRKLISKEDVEDVRREIQIMHHLAGHKNIVTIKGAYEDPLYVHIVMELCAGGELFDR IIRGHYSER
CPK7 (59-321) YDLGREVGRGEFGITYLCTDKETGEKYACKSISKKLRTAVDIEDVRREVEIMKHMPKHPNVVSLKDSFEDDDAVHIVMELCEGGELFDRIVARGHYTER
CPK8 (57-318) YDLGREVGRGEFGITYLCTD ITGEKYACKSISKKKLRTAVDIEDVRREVEIMKHMPRHPNIVSLKDAFEDDDAVHIVMEXCEGGELFDRIVARGHYTER
CPK9 (91-352) YTLGKELGRGQFGVTYLCTENSTGKKYACKSISKKKLVTKADKDD RREIQ IMHLSGQPNIVEFKGAYEDEKAVNLVMELCAGGELFDRIIAKGHYTER
DcCDPK ( 1-235) ..........................- ACKS ILKRKLVSKNDKED LREIQILQHLSGQPNIVEFKGVFEDROSVHLVMELCAGGELFDRIIAQGHYSER


VI-a VI-b VII VIII IX
Cons. -Aa-#-+-i~-vv--CH~-GVmHRDIKPENFL ------- 1Ka-DFGLSvf---g ---- VGS-YYvAPEVL-+-yGpe-D#WsaGvilYilLsGvPPFWaete-G

CDPKa QAARLIKTIVEVVEACHSLGVMHRDLKPENFLFDTIDEDAKLKATDFGLSVFY:PGESFCDVVGSPYVAPEVLRKLYGPESDVWSAGVILYILLSGVPPFWAESEPG
CDPKP EAAKLIKTIVGVVEACHSLGVMHRDLKPENFLFDTPGEDAQMKATDFGLSVILQARQAFHDWGSPYYVAPEVLCKQYGPEVDVWSAGVILYILLSGVPPFWAETEAG
CDPKY AAASICRQIVNVVHICHFMGVMHRDLKPENFLLSSRDENALLKATDFGLSVFIEEGKVYRDIVGSAYYVAPEVLRRRCGKEIDIWSAGVILYILLSGVPPFWAETEKG
SPK KAAELIKIIVGIIETCHSHGVMHRDLKPENFLLLDADDEFSVKAIDFGLSVFFRPGQVFREVGSP YYIAPEVLEKRYGPEAD IWTAGVILYVLLTGVPPFWADTQSG
OsCPK2 AAATICRAVVNVVNICHFMGVMHRDLKPENFLLATKEENAMLKATDFGLSVFIEEGKMYRDIVGSAYYVAPEVLRRNYGKEIDVWSAGVILYILLSGVPPFWAETEKG
OsCPK11 KAAELIRIIVS IVAMCHSLGVMHRDLKPENFLLLDKDDDLSIKAIDFGLSVFFKPGQVFTELVGSPYYVAPEVLHKRYGPESDVWSAGVILYVLLSGVPPFWAETQQG
ZmCDPK GAAELLRAIVQIVHTCHSMGVMRDIKPENFLLLSKDEDAPLKATDFGLSVFFKEGELLRDIVGSAYYIAPEVLKRKYGPEADIWSVGVMLYIFLAGVPPFWAENENG
Atcdpkl AAAAVART IAEVVMMCHSNGVMHRDLPENFLFANKKENSPLKAIDFGLSVFFPGDKFTEIVGSPYYMAPEVLKRDYGPGVDVWSAGVI IYILLCGAPPFWAETEQG
Atcdpk2 EAVKLIKTILGVVEACHSLGVMHRLKPENFLFDSPKDDAKLKATDFGLSVFYKPGQYLYDVVGSPYYVAPEVLKKCYGPEIDVWSAGVILYILLSGVPPFWAETESG
CPK1 KAAELTRTIVGVVEACHSLGVMHRDLKPENFLFVSKHEDSLLKTIDFGLSMFFXPDDVFTDVVGSPYYVAPEVLRKRYGPEADVWSAGVIVYILLSGVPPFWAETEQG
CPK2 KAAELARTIVGVLEACHSLGVMHRDLKPENFLFVSREEDSLLKTIDFGLSMPFKPDEVFTDVGSPYYVAPEVLRKRYGPESDVWSAGVIVYILLSGVPPFWAETEQG
CPK3 AAADLCRQMVMVVHSCHSMGVMHRDLPENFLFLSKDENSPLKATDFGLSVFFKPGDKFKDLVGSAYYVAPEVLKRNYGPEADIWSAGVILYILLSGVPPFWGENETG
CPK4 EAAKLIKTILGVVEACHSLGVMHRDLKPENFLFDSPSDDAKLKATDFGLSVFYKPGQYLYDWGSPYYVAPEVLKKCYGPEIDVWSAGVILYILLSGVPPFWAETESG
CPK5 KAAELTKIIVGWEACHSLGVMHRDLKPENFLLVNKDDDFSLKAIDFGLSVFKPGQIFTDVVGSPYYVAPEVLLKRYGPEADVWTAGVILYILLSGVPPFWAETQQG
CPK6 KAAELTKIIVGVVEACHSLGVMHRDLKPENFLLVNKDDDFSLKAIDFGLSVFFXPGQIFKDVVGSPYYVAPEVLLKHYGPEADVWTAGVILYILLSGVPPFWAETQQG
CPK7 AAAAVMKT IVEWVVQ ICHKQGVMHRDLKPENFLFANKKETSALKAIDFGLSVFFPGEQFNE IVGSPYYMAPEVLRRNYGPEIDVWSAGVILYILLCGVPPFWAETEQG
CPK8 AAAAVMKTILEVVQICHKHGVMHRDLKPENFLFANKKETSALKAIDFGLSVFFiPGEGFNE IVGSPYYMAPEVLRRNYGPEVD IWSAGVILYILLCGVPPFWAXTEQG
CPK9 AAASVCRQIVNVVKICHFMGVLHRDLKPENFLLSSKDEKALIKATDFGLSVFIEEGKVYRDIVGSAYYVAPEVLRRRYGKEVDIWSAGIILYILLSGVPPFWAETEKG
DcCDPK AAAT ICRQIVNVVHVCHFMGVMHRDLKPENFLLSSKDKDAMLKATDFGLSVFIEEGKVYRNIVGSAYYVAPEVLRRSYGKE IDIWSAGVILYILLSGVPPFWAENEG














x XI
if~--l-gq-Df--~~pWP~-S-AKdL#-kMLP-- ~R-#~a--vL-hpW#--

IFRQILLGKLDFHSEPWPSISDSAKDLIRKMLDQNPKTRLTAHEVLRHPWIVDD
IFRQILNGDLDFVSEPWPSISENAKELVKQMLDRDPKKRISAHEVLCNPWVVDD
IFDAILEGHIDFESOPWPNISDSAKDLVRKMLIQDPKKRITSAQVLEHPWIKD-
iYEXVLDGRIDFKSNRWPRISDSAKDLIKKMLCPYPLERLKAHEVLKHPWICDN
IFDAILOGEIDFESQPWPSISESAKDLVRKMLTQDPKKRITSAQVLQHPWLRD-
IFDAVLKGHIDFQSDPWPKISDSAKDLIRKMLSHCPSERLKAHEVLREPWICEN
IFTAILRGQLDLSSEPWPHISPGAKDLVKKMLNINPKERLTAFQVLHWIKED
VALAILRGVLDFKRDPWPOISESAKSLVKQMLDPDPTKRLTAQQVLAHPWIQNA
IFRQILQGKLDFKSDPWPTISEAAKDLIYKMLERSPKKRISAHEALCHPWIVDE
IFEQVLHGDLDFSSDPWPSISESAKDLVRKMLVRDPKKRLTAHQVLCHPWVQVD
IFEQVLHGDLDFSSDPWPSISESAKDLVRKMLVRDPKRRLTAHQVLCHPWVQID
IFDAILQGQLDFSADPWPALSDGAKDLVRKMLKYDPKDRLTAAEVLNHPWIRED
IFRQILQGKIDFKSDPWPTISEGAKDLIYKMLDRSPKKRISAHEALCHPWIVDE
IFDAVLKGYIDFESDPWPVISDSAKDLIRRMLSSKPAERLTAHEVLRHPWICEN
IFDAVLKGYIDFDTDPWPVISDSAKDLIRKMLCSSPSERLTAHEVLRHPWICEN
VAQAIIRSVIDFKRDPWPRVSDSAKDLVRKMLEPDPKKRLTAAQVLEHTWILNA
VXQXIIRSVIDFKRDPWPRVSETAKDLVRKMLEPDPKKRLSAAOVLEHSWINA
IFDAILEGHIDFESQPWPSISSSAKDLVRRMLTADPKRRISAADVLQHPWLREG
IFDAILEGVIDFESEPWPSVSNSAKDLVRKMLTQDPRRRITSAQVLDHPWMREG


B, Junction Domain

Con.en.u -~A-d--d--VlsR#KqF--MNk#Kk-aL-vI

CDPKa (296-328) NIAPDKPLDSAVLSRLKQFSAMNKLKKMALRVI
CDPK) (286-317) -IAPDKPLDSAVLTRLKHFSAM4KLKKMALRVI
CDPKy (345-377) GNASDKPIDSAVLSRMKQFRAMNKLKKLALKVI
SPK (335-367) GVATNRALDPSVLPRLKQFSAMNRLKKXSLQII
OsCPK2 (347-379) GEASDKPLDPSVISRLKOFSAMNKLKKLALRVI
OsCPK11 (341-373) GVATDQAIDSAVLSRMKOFRAMNKLKKMALMVI
ZmCDPK (355-387) GDAPDTPLDNVVLDRLKQFRAMNQFKKAALRII
Atcdpkl (273-305) KKAPNVPLGDIVRSRLKQFSMMNRFKKKVLRVI
Atcdpk2 (288-320) QAAPDKPLDPAVLSRLKOFSQMNKIKKMALRVI
CPK1 (412-444) GVAPDKPLDSAVLSRMKQF SAMNKFKKMALRVI
CPK2 (448-480) GVAPDKPLDSAVLSRMKOFSAMNKFKKMALRVI
CPK3 (337-369) GEASDKPLDNAVLSRMKQFRAMNKLKKMALKVI
CPK4 (287-319) HAAPDKPLDPAVLSRLKQFSOQNKIKMALRVI
CPK5 (358-390) GVAPDRALDPAVLSRLKQFSAMNKLKKMALKVI
CPK6 (347-379) GVAPDRALDPAVLSRLKQFSAMNKLKKM&LKVI
CPK7 (321-353) KKAPNVSLGETVKARLKQFSVMNKLKKRALRVI
CPK8 (319-351) KKAPNVSLGETVKARLKQFSVMNKLKKRALRVI
CPK9 (353-385) GEASDKPIDSAVLSRMKQFRAMKKLKKLALKVI
DcCDPK (236-268) GEASDKPIDSAVLSRMKQFRAMNKLKQLALKVI


Con&.

CDPKm
CDPKA
CDPK7
SPK
OsCPK2
OsCPKl1
ZmCOPK
Atcdpkl
Atcdpk2
CPKI
CPK2
CPX3
CPK4
CPK5
CPK6
P?7
CPK8
CPK9
DcCDPK


Figure 2-2 continued










C. Calmdoulin-lik* Domian
--- I-- ------ II---
Consnus Ae-lseeEi-glk-#F--D-d--g-it#~ELK-GL--~~G~~~ --# --- #~~M-AAd-D-g-idy-EF#-ath--k#er-e---~~AF-f

CDPKa (329-508) AERLSEEE IGGLKELFKMIDTDNSGTITFDELKDGLKRVGS-ELMESEIKDLMAADIDKSGTIDYGEFIAATVHLNKLEREENLVSAFS F
CDPK" (318-498) AERLSEEE IGGLKELFMIDTDNSGT ITFEELKEGLKSVGS-NLMESEIKSLMEAADIDNNGSIDYGEFLAATLHLMREENLVAAFAYL
CDPKy (378-559) AENMSAEEIQGLKAMFTMDTDKSGT ITYEELKSGLHRLGS-KLTEAEVKQLMAADVDGNGSIDYIEFITATHRHKLERD DQLFKAF QY
SPK (368-534) AERLSEEEIVGLREMFKAMDTKNRSVVTFGELK-GLKRYSS-VFKDTEINDLMEAA-DDTTSTINWEEFIAAAVSLNKIEREKHLMAAFTYF
OsCPK2 (380-533) ASNLNEEE IKGLKQMFTMDTDNSGT ITYEELKAGLAKLGS-KLSETE IGDIEAAHNDNNVTIHYEEFIAATLPLNKIEREEHLLAAFTYF
OsCPKll (374-542) AERLSEEEIAGLREMFKAVDTKNRGVITFGELREGLRRFGA-EFKDAEVKQLMAADVDGNGSIDYVEFITATRHKLERDEHLFKAFQYF
ZmCDPK (388-550) AGCLSEEEITGLKEMFKNIDKDNSGTITLDELKHGLAKHGP-KLSDSEMEKMAADADGNGLIDYDEFVTATVHMNKLDREEHLYTAFQYF
Atcdpkl (306-495) AEHLSIQEVEVIKNMFSLMDDDKDGKITYPELKAGLQKGS-QLGEPEIKMLMEVADVDGNGFLDYGEFVAVIIHLQKIENDELFKLAFFF
Atcdpk2 (321-495) AERLSEEEIGGLKELFKMIDTDNSGTITFEELKAGLKRVGS-ELMESEIKSLAADIDNSGTIDYGEFLAATLHMNKMEREEILVAAFSDF
CPK1 (445-610) AESLSEEEIAGLKEMlNMIDADKSGQITFEELKAGLKRVGA-NLKESEILDLMOAADVDNSGTIDYKEFIAATLHLNKIEREDHLFAAFTYF
CPK2 (481-646) AESLSEEEIAGLKQMFMIDADNSGQITFEELKAGLKRVGA-NLKESEILDLMQAADVDNSGTIDYKEFIAATLHLNKIEREDHLFAAFSYF
CPK3 (370-526) AENLSEEEIIGLKEMFKSLDTDNNGIVTLEELRTGLPKLGS-KISEAEIRQLMAADMDGDGSIDYLEFISATMHMNRIEREDHLYTAFQFF
CPK4 (320-501) AERLSEEE IGGLKELFMIDTDNSGT ITFEELKAGLKRVGS-ELMESEIKSLMDAADIDNSGT IDYGEFLAATLH INKMREENLW AFSYF
PK5 (391-556) AESLSEEEIAGLREMFQAMDTDNSGAITFDELKAGLRKYGS-TLKDTEIHDLMDAADVDNSGTIDYSEFIAATIHLNKLEREEHLVAAFQYF
CPK6 (380-544) AESLSEEE IAGLRAFEAMDTDNSGAITFDELKAGLRRYGS-TLKDTEIRDLMEAADVDNSGTIDYSEFIAAT ILNKLEREEHLVSAFQYF
CPK7 (354-535) AEHLSVEEAAGIKEAFEMM2VNKRGKINLEELKYGLQKAG-QQIADTDLQILMEATDVDGDGTLNYSEFVAVSVHLKKMANDEHLHKAFNFF
CPK8 (352-533) AEHLSVEEVAGIKEAFEMMDSKKTGKINLEELKFGLHKLGQQIPDTD LQILMEAADVDGDGTLNYGEFVAVSVHLKKMANDEHLHKAFSFF
CPK9 (386-541) AENIDTEEIQGLKAMFANIDTDNSGTITYEELKEGLAKLGS-KLTEAEVKQLDAADVDGNGSIDYIEFITATMFRHRLESNENLYKAFQHF
DcCDPK (269-425) AESLSEEEIKGLKSMFAMDTDKSGTITYEELKSGLARLGS-KLSEVEVQLMDAADVDGNGTIDYLEFITATMHRHKLESYEH-QAFQYF


-----II---- ------I----
Cons. Dkd-SqOIt-del~-a --D----------- #-~ # --#D-dnDGrI-0-EF-aMM --------------- -------- ----------

CDPKa DKDGSGYITLDEIQQACKDFGLDDI-------HIDDMIKEIDQDNDGQIDYGEFAAMM ----RKGNGGIGR-RTMRKTLNLR--DALGLVDNGSNOVIEGYFK*
CDPK DKDGSPYITIDELQQACKDFSLGDV------ HLDEMIKEIDQDNDGRIDYAEFAAMM-----KKGDPNMGRSRTMKGNLNFNIADAFGMKD----SS
CDPKY DKDNSGF ITRDELESAMKEYGMGDDAT IIEIISEVDTIISEVDTDHDGRINYEEFSAMM------KSGN---------QQQGKLF*
SPK DKDGSGFITVDKLOKACMERNMEDT ----- FLEEMILEVDQNNDGQIDYAEFVTM ------OSNNFGLGW-QTVESSLNVALREAPQVY*
OsCPK2 DKDNSGFITRDELESALIEHEMGDTSTIIKD-------IISEVDTDNDGRINYEEFCAMM4----RGGG-------QPMRLK
OsCPK11 DKDGSGYITVDKLQRACGEHNMEDS--------LLEEIISEVDQNNDGQIDYAEFVAMM------QGSNVGLGW-QTMESSLNVALRDAPV*
ZmCDPK DKDNSGYITKEELEHALKEQGLYDADKIKD -----IISDADSDNDGRIDYSEFVAMM-----RKGTAGAEP-MNIKKRRDIVL*
Atcdpkl DKDGSTYIELDELREA-----LADELGEPDA-SVLSDIMREVDTDKDGRINYDEFVTMMKAGTDWRKASROYSRERFKSLSINLMKDGSLHLHDALTGQTVPV*
Atcdpk2 DKDGSGYITIDELQSACTEFGLCDT----- PLDDMIKEIDLDNDGKIDFSEFTAMM---- RKGD-GVGRSRTMMbKNLNFNIADAFGVDG----EKSDD*
CPK1 DKDGSGYITPDELQOACEEFGVEDV----- RIEELMRDVDQDNDGRIDYNEFVAMM---- QKGSITGG---PVK8GLEKSFSIALKL'
CPK2 DKDESGFITPDELQQACEEFGVEDA----- RIEEMMRDVDODKDGRIDYNEFVAMM-----QKGSIMGG---PVKMGLENSISISLKH
CPK3 DNDNSGYITMEELELAMKKYNMGDDKSIKE------- IIAEVDTDRDGKINYEEFVAMM------KKG----NP-ELVPNRRRM
CPX4 DKDGSGYITIDELQQACTEFGLCDT----- PLDDMIKEIDLDNDGKIDFSEFTAMM -----KGD-GVGRSRTMRNNLNFNIAEAFGVEDTSSTAKSDDSPK*
CPK5 DKDGSGFITIDELQQACVEHGMADV------ FLEDIIKEVDQNNDGKIDYGEFVEM .-----QKGNAGVGR-RTHRNSLNISMRDA*
CPK6 DKDGSGYITIDELQQSCIEHGMTDV------ FLEDIIKEVDQDNDGRIDYEEFVAM ----QKGNAGVGR-RTMKNSLNISMRDV*
CPK7 DQNGSGYIEIDELREA-----LNDELDNTSSEEVIAAIMQDVDTDKDGRISYEEFVAMMKAGTDWRKASRQYSRERFNSLSLKMRDGSLQLEGET*
CPK8 DONQSDYIEIEELREA----LNDEVD-TNSEEVAAIMQDVDTDKDGRISYEEFAAMMKAGTDWRKASRQYSRERFNSLSLKLMREGSLQLEGEN*
CPK9 DKDSSGYITIDELESALKEYGMGDDATIKE------VLSDVDSDNDGRINYEEFCAMM----RSGNP-------QQQQPRLF*
DcCDPK DKDNSGFITKDELESAMKEYGMGDEATIKD-------IISEVDSDNDGRINYDEFCAMM-----RRAR--------NRR-KLFSVSLTS"

Figure 2-2 continued












B



C










Figure 2-4. RNA blot hybridization analyses of CDPK encoding genes. Total
RNA (10 'g) from three-week-old soybean plants (lanes 1-5), eight-day-
old seedlings, (lane 6), and soybean cell cultures (lanes 7-8), was
fractionated in a formaldehyde agarose gel, blotted, and hybridized with
specific probes for each enzyme as described in Experimental
Procedures. The samples were: lane 1, root; lane 2, stem; lane 3, shoot
tip; lane 4, petiole; lane 5, leaf; lane 6, whole seedling; lane 7, three-day-
old cell culture; and lane 8, seven-day-old cell culture.
a) CDPKa; b) CDPKp; c) CDPKy; d) Ethidium bromide stained gel
demonstrates that the equal amounts of RNA were loaded in each
lane.








week-old plants, in eight-day-old seedlings, and in suspension cell cultures of

different ages. However, the relative distribution of each transcript differed.
CDPKa mRNA was most highly expressed in three-day-old cell culture. This

result may imply that the expression of CDPKa is regulated by developmental
stages. CDPKp which was originally isolated from a plumule library, was
most highly expressed in leaves. CDPKy, which was originally isolated from a
library made from tissues enriched in nitrogen-fixing root nodules, was most
highly expressed in roots. It will be interesting to examine if the transcription
levels of each isoform in different tissues correlates with the translation
levels in order to better understand what physiological implications can be
made from these observations.

Expression of CDPK Fusion Proteins in E. coli and Purification

As a step towards determining if the three soybean CDPKs perform
distinct or overlapping roles in the cell, the kinetic properties of recombinant
enzymes expressed in E. coli were examined. The three full length enzymes

and an amino terminal deletion mutant of CDPKy, CDPKy(66-538), were

expressed as glutathione S-transferase (GST) fusion proteins. These

recombinant proteins were highly expressed in E. coli under experimental
conditions and showed CDPK activity when total cell extracts were used for
enzyme assays (data not shown). Interestingly, it was found that monoclonal
antibodies crossreact with CDPKp very weakly (Figure 2-4). Conversely,
polyclonal antibodies raised against the CLD domain of CDPKa (Dachmann et
al., 1996) recognized both CDPKp and y quite well.
Recombinant enzymes were highly purified by affinity
chromatography on glutathione-agarose and anion exchange chromatography
(Figure 2-5). The activities of CDPKa, and p were stable in the absence of












A B

94-

67-




43-





30-


1 2 3 123






Figure 2-4. Immunoblot of recombinant proteins with monoclonal
antibodies. Lane 1 contains GST-CDPKg; lane 2, GST-CDPKg(66538);
lane 3, GST-CDPKb.
a) Cells expressing recombinant proteins were resolved in 10 % SDS-
PAGE and stained with Coomassie blue or b) blotted to nitrocellulose
and probed with monoclonal antibodies directed against native
soybean cell CDPK.
















fm


43-





30-


1 2 3





Figure 2-5. SDS-PAGE of purified recombinant proteins. Purified
recombinant proteins (5 jig) were analyzed by electrophoresis in 10 %
SDS-polyacrylamide gel and stained with Coomassie blue. Lane 1
contains GST-CDPKr; lane 2, GST-CDPKT(66-538); lane 3, GST-CDPKp.
Molecular mass markers are as follows (in kilodaltons): phosphorylase
b, 94; BSA, 67; ovalbumin, 43; carbonic anhydrase, 30.








reducing agents and at low ionic strength, and these enzymes could be stored

in 50% (v/v) glycerol at -80 OC without appreciable loss of activity. CDPKyand
CDPKY(66-538), on the other hand, required reducing agents and 300 mM KC1
or NaCI during dialysis and could not be frozen without almost complete loss
of activity. These enzymes were also expressed as His6-fusion proteins, but
because His6-CDPK- required high ionic strength to maintain solubility and
activity, further purification of this protein by anion-exchange
chromatography was not possible. In addition, His6-CDPKy(66-538) bound to
phenyl-Sepharose at low ionic strength in the presence of Ca2+, but could be
eluted only by 6 M urea.

Substrate Specificities and Kinetic Parameters

The activities of soybean CDPKs with various synthetic peptides and
histone IIIS were determined in the presence or absence of Ca2+ (Table 2-1).
There was little difference in the activities of CDPKy and the amino terminal
deletion mutant CDPKY(66-538), showing that the N-terminal 66 residues of
this isoenzyme are not required for catalytic activity.

Histone IIS, a good substrate for CDPK purified from soybean cell
culture (Putnam-Evans et al., 1990), was phosphorylated by all three isozymes
(Table 2-1). Other proteins such as casein and bovine serum albumin were
not phosphorylated by any of the recombinant isoenzymes (data not shown).
Substrate peptides containing the motif basic-X-X-Ser/Thr, in which
the basic residue is arginine or lysine, x is any residue, and serine or
threonine is the phosphorylated residue, are good substrates for CDPKs
(Roberts and Harmon, 1992). The specific activities of the soybean CDPKs
with 100 uM of each of four peptides containing this motif, syntide-2,
autocamtide-2, and substrate peptides of skeletal and smooth muscle myosin












Table 2-1. Activity of CDPK isoforms with various substrates.


Enzyme activities were measured as described in Experimental Procedures with 0.5 mg/ml histone
IS or 100 pM synthetic peptide in the presence of 1 mM EGTA (-Ca2+) or 1 mM EGTA plus 1.1 mM
Ca2+ (+ Ca2+)a.
enzyme Ca2+ Histone IS Syntide-2 Autocamtide MLCKsk MLCKsm
pmollmin/mg
CDPKa + 0.045 1.67 0.81 0.15 0.14
ndb 0.075 0.025 0.005 0.005
CDPKP + 0.45 3.55 2.41 0.98 1.82
0.02 0.14 0.07 0.03 0.08
CDPKy + 0.25 1.86 1.82 1.92 157
ndb 0.03 0.01 0.004 0.03
'Y66-538) + 0.28 1.89 2.12 1.65 1.56
ndb 0.02 0.008 0.007 0.008


aThe standard error of each mean value of enzyme activities was less than 10% of the tabulated
mean value.
bnd, not detectable.








light chain kinases, were compared (Table 2-1). Phosphorylation of the

peptide substrates by CDPKy and CDPK(66-538) was stimulated 52 to 480-fold,
whereas the activities of CDPKa and CDPKp with these substrates was
stimulated 22- to 30-fold. All four peptides were good substrates for CDPKy
and CDPKY(66-538), and the maximal activity of these enzymes with each
peptide varied little (<1.4-fold). The activities of CDPKa and CDPKp varied by
as much as 12- and 3.6-fold, respectively. Syntide-2 and autocamtide were
good substrates for both CDPKa and CDPKp. In contrast, MLCK substrate
peptides were good substrates for CDPKp, but not for CDPKa.
To examine the basis for the difference in substrate preference, the
kinetic parameters for the CDPKs with two of the peptide substrates were
determined (Table 2-2). The apparent Km and Vmax of CDPKa and CDPKy
with syntide-2 as substrate were similar and about 2-fold lower than the

parameters for CDPKp. Comparison of the ratio of Vmax to Km, which is a
measure of catalytic efficiency, shows that syntide-2 is an equally good
substrate for all three isoenzymes. In contrast, skeletal muscle myosin light
chain kinase peptide is a good substrate for only CDPKy. The kinetic

parameters for the two peptides with CDPKy are similar and there is only a
two-fold difference in the catalytic efficiencies. While the apparent Vmaxs for
CDPKa and CDPKp were 3- and 2-fold higher, respectively, than that of
CDPKy, the apparent Kms were 140- and 10-fold higher, respectively.
Comparison of the ratios of Vmax to Km, shows that CDPKa has a strong
preference for syntide-2, CDPKp has less difference in preference, and CDPK-
has a slight preference for syntide-2. Syntide-2 and autocamtide-2 contain

branched-chain amino acid residues at positions P-5, P+1, and P+4, whereas
the MLCK peptides do not (Figure 2-6). One or more of these residues may be
determinants for phosphorylation of a protein by CDPKa.














Table 2-2. Kinetic parameters of CDPK isoforms.

Values of apparent Vmax and Km were determined from double-reciprocal plots.
Syntide-2 MLCKsk
Vmax Vmax
Enzyme Vmax Km Km Vmax Km Km
pmollminlmg pM pmol/min/mg pM
CDPKa 2.4 0.3 18.2 1.5 0.13 5.7 2.5 3700 1900 0.001
CDPKI 5.5 0.7 34.21.0 0.16 3.8 0.8 282 46 0.013
CDPKI 2.5 0.1 16.6 0.7 0.15 2.0 0.1 27 0.8 0.074



























Synt de-2
Autocomtlde-2
MLCsk
MLCsm


PLARTLSVAGLPGKK
KKALRRQETVDAL
AKRPORATSNVFS
KKRAARATSNVFA

-5 t t4
-1 +4


Figure 2-6. Sequences of Substrate Peptides. Residues are numbered relative to the
phosphorylated residue at position P. Residues amino terminal to P have
negative numbers and those carboxyl terminal to P have positive numbers.










Effect of pH

The effect of pH on phosphorylation of syntide 2 by the three CDPKs is

shown in Figure 2-7. All three isoenzymes showed maximal activity at pH 7
to 8, but CDPKa and p were more tolerant to alkaline conditions. The broad
pH optimum of CDPKa and p is similar to that of CDPK purified from
soybean (Putnam-Evans et al., 1990).

Effect of Protein Kinase Inhibitors

The effect of several classes of protein kinase inhibitors on the activity
of soybean CDPKs was determined. Control assays containing 0.01% (v/v)
DMSO did not affect kinase activity. Staurosporine, an inhibitor of broad
specificity that is suggested to interact with an essential region of catalytic
domain of protein kinases (Tamaoki, 1991), inhibited the CDPKs (Table 2-3)
with ICsos between 70 and 120 nM. Another general inhibitor of protein
kinases is K-252a (Kase et al., 1987), which competes with ATP. K-252a
inhibited the CDPKs (Table 2-3), with IC50s between 300 and 800 nM. The

IC50s for both of these inhibitors were one to two orders of magnitude higher
than those observed with PKC, PKA, or MLCK (Hashimoto et al., 1991; Kase et
al., 1987; Tamaoki, 1991).

Table 2-3. IC50 Values for Inhibitors of CDPK Isoenzymes.
Inhibitors CDPKa CDPK_ CDPKy
(pM)
Staurosporine 0.11 0.12 0.07
K-252a 0.8 0.8 0.3
Calphostin C 9.0 5.0 1.6


































40 -

20 1 1 1
C
100-

75"

50so

25-

0-
5 6 7 8 9 10

pH


Figure 2-7. Effect of pH on activity. Enzymes were assayed with syntide-2 as
described in Experimental Procedures. Buffers (50 mM) were: MES (pH
5.9), HEPES (pH 7.1, 7.9), Tris (pH 7.6, 7.7,8.8), and CHES (pH 8.9, 9.8).
a) CDPKa; b) CDPKB; c) CDPKy.








Highly effective inhibitors of CaMKII and myosin light chain kinase
had little to no effect on the CDPKs; concentrations of these inhibitors used in
the assays were 50 ;iM H8 (Hidaka et al., 1984), 10 iM KN62 (Hidaka et al.,
1991), 30 gM KN93 (Sumi et al., 1991), and 50 liM ML7 (Saitoh et al., 1987).
Calphostin C, which is reported to specifically inhibit protein kinase C
through interaction with its regulatory domain, inhibited all three CDPK
isoforms (Table 2-3), but the IC50 values were one to two orders of magnitude
greater than those for protein kinase C (Tamaoki, 1991).

Purification of Anti-Nti Antibodies

First, purification of polyclonal antibodies (Anti-Nty) raised against N-
terminal domain of CDPKy fused to GST were attempted from immunoblots.
(Olmsted, 1981; Smith and Fisher, 1984). But Anti-Nty could not be eluted
effectively due to the high affinity (data not shown). Therefore, an affinity
column was prepared by coupling GST expressed in E. coli to AminoLinkTM
Gel (Pierce) in order to absorb anti-GST antibodies. Partially purified Anti-
Nty did not crossreact with GST nor with recombinant CDPKa nor p, but did
crossreact with many protein bands in soybean cell extracts. When His6
fusion proteins of full-length and an N-terminal deletion mutant of CDPKy
were used as antigens on blots to elute antibodies, antibodies recognizing N-
terminal mutant were eluted. This result suggested that antibodies
recognizing amino acd residues from 66 to 83 of the N-terminal domain of
CDPKy could be affinity purified. Therefore, Anti-Nty were purified further
using affinity column made from highly purified His6-CDPK(66-538). When
soybean cell extracts were immunostained with affinity purified Anti-Nty,
nonspecific cross reactions were reduced, and very faint band of right size to
be CDPKy could be detectable (data not shown).








To test whether CDPKy is localized at membranes, microsome and
cytosol fractions were challenged with Anti-Nty and anti-CLD antibodies.
The blots (Figure 2-8) suggested that CDPKy is probably located both in
microsome and cytosol and probably at low abundance.

Discussion


Three Soybean CDPKs are Constitutively Expressed Multifunctional Protein
Kinases

The primary structures of soybean CDPKa, p, and y are similar to those
of CDPKs from a variety of plants. The most distinctive feature of the new
CDPKs is the eight amino acid insert prior to the fourth EF-hand that is
present in CDPKy, but not CDPKa and p. Whether this feature contributes to
the functional properties of this CDPKy remains to be determined. In contrast
to a CDPK from maize that is expressed only in pollen (Estruch et al., 1994),
the CDPKs examined in this study are present in numerous parts of soybean
plant. These CDPKs are also expressed in cell cultures and young plants
grown under standard conditions. Expression of a CDPK from Arabidopsis is
induced by environmental stress (Urao et al., 1994), and another from mung
bean is induced by mechanical strain or auxin (Botella et al., 1996).
We have shown that biochemical properties and substrate specificities
of soybean CDPK isoenzymes differ. Biochemical properties of CDPKa and p
are more similar to each other than to those of CDPKy as the sequence
identity between CDPKa and p is higher than that between CDPKa and 7 or p
and y. In contrast to CDPKa and CDPKp, CDPKy is not stable in the absence of
DTT, in buffers of low ionic strength, or when frozen. All three isoenzymes
phosphorylate peptides containing a basic-X-X-Ser motif, but each isoenzyme


















A B



97-
66-_

45-


31-


1 2 3 4 5 6






Figure 2-8. Immunoblots of soybean cell extracts with Anti-Nty or anti-CLD
antibodies. Lane 1 contains 20 ng of purified Hise-CDPKy; lane 2, 25 pg
of microsomal fraction prepared from soybean cells, and lane 3, 25 pg of
soluble fraction.
a) Cell extracts were resolved in 10 % SDS-PAGE and blotted onto
nitrocellulose membrane and probed with anti-CLD antibodies or b)
with Anti-Nty.








has a different value of apparent Km and Vmax for each of the peptides.
CDPKa selectively phosphorylates syntide-2, and CDPKy phosphorylates the
four peptides tested equally well.
Recently, Bachmann et al., (Bachmann et al., 1996) showed that serine-
543 in nitrate reductase is phosphorylated by partially purified CDPK from
spinach. Analysis of synthetic peptides in which residues surrounding the
phosphorylation site were varied, showed that motif preferred by the spinach
CDPK was hydrophobic-X-basic-X-X-Ser, where the hydrophobic residue at P-5
was leucine. These observations agree with the conclusion that the presence
of branched chain aliphatic amino acids at one or more of the positions P-5,
P+1, and P+4 is determinant for phosphorylation by some CDPK isoenzymes.
However, a systematic analysis of phosphorylation motifs is needed in order
to define the importance of the hydrophobic residues for each isoenzyme.
None of the recombinant CDPKs is identical to CDPK purified from
soybean cell culture. Unlike the recombinant CDPKs, the native enzyme
phosphorylates histone IlS very well (Harmon et al., 1996), but its peptide
substrate specificity is most similar to that of CDPKa. The amino acid
sequences of peptides derived from the native enzyme (Harper et al., 1991) are
not identical to sequences of CDPKa, p, or y, but 60 of 62 of the known residues
match the sequence of mung bean CDPK (Botella et al., 1996). While it is
possible that the properties of the recombinant enzymes differ from the
native soybean enzymes because of modifications due to their expression in a
prokaryotic system, the high identity of the known sequence from native
enzyme with the mung bean CDPK suggest that the native soybean cell
enzyme is a fourth isoenzyme that was purified to near homogeneity because
of its distinct biochemical properties.








Subcellular Localization of CDPKs Remains Unsolved

While there is evidence for localization of CDPKs with microfilaments

in onion and spiderwort (Putnam-Evans et al., 1989) and with plasma
membranes of soybean (Borochov-Neori and Harmon, 1993), oat (Schaller et
al., 1992), and zucchini (Verhey et al., 1993), the basis for these associations is
not known. None of the CDPKs characterized to date contain membrane

spanning domains. Some CDPKs such as CPK1 (Harper et al., 1993) and CPK2
(Hrabak et al., 1996) from Arabidopsis and three rice CDPKs (Breviario et al.,
1995; Kawasaki et al., 1993) have putative myristoylation sites, and the
myristoylation of recombinant carrot CDPK has been reported (Farmer and
Choi, 1995), but it is not yet known if myristoylation is sufficient for
membrane localization. The localization of protein kinases to specific sites in
cells via interaction of the enzyme with targeting proteins is a common
theme in signal transduction in animals (Mochly-Rosen, 1995).
The amino terminal domain of CDPK- is a candidate for such
interaction since it is relatively long, and its deletion does not affect the
catalytic activity of the enzyme. If isoform specific antibodies can be obtained
they will be valuable tools to elucidate the subcellular localization of CDPK
isoforms. In this regard, the production and purification of antibodies specific

to CDPKy was attempted. Raising polyclonal antibody against GST tagged N-
terminal domain of CDPKy that was expressed in E. coli. was successful. The
polyclonal antibody only recognized recombinant CDPKy, but not a or p.
However, affinity purification of the antibody by conventional methods was
difficult due to its high affinity. These antibodies were shown to crossreact
with a band of the right size for CDPKy, but also recognized other bands in
crude extracts in either cytosolic or microsomal fractions of soybean cells.








Proving that the right sized band recognized by the antibodies corresponds to
CDPKy will not be an easy task because of extreme instability of the enzyme.

General Protein Kinase Inhibitors Shows Higher IC50s for CDPKs than Those
for Animal Protein Kinases

CDPKa, p, and T differed in their susceptibility to inhibition by protein
kinase inhibitors, e. g., staurosporine, K-252a, and calphostin C. CDPKy was
lthe most sensitive to inhibition by all three compounds. The IC50s for
staurosporine and K-252a were higher than those required for inhibition of
the animal protein kinases, PKA, PKC, MLCK, or CaMKII, but these
concentrations are similar to those needed for inhibition of various responses
of plant cells to stimuli (Hashimoto et al., 1991; Kase et al., 1987; Tamaoki,
1991). Calphostin C also inhibited the soybean CDPKs, but with IC50s 30- to
200-fold higher than that for PKC. These results show that the concentration
of calphostin C used in experiments with plant cells must be carefully
considered before it can be concluded that an observed effect resulted from
inhibition of a plant protein kinase C homolog. KN62 inhibits CaMKII

competitively with respect to calmodulin; however, it did not show any
significant inhibition of CDPKs at a concentration (10 gM) that inhibits over
80% of CaMKII activity. It will be beneficial to find CDPK-specific inhibitors,
either from pharmacological screening or from designing peptides, in order to
facilitate studies unraveling the physiological roles of CDPK in vivo.
The results demonstrated that three soybean CDPK isoforms differ in
biochemical and kinetic properties and in their RNA expression patterns.
Although all three CDPKs phosphorylate peptides having the same core
motif; each isoform differs in its selectivity for residues surrounding this
motif, basal activity, and fold-activation. These results support the hypothesis






64

that CDPK isoenzymes play distinct or overlapping roles and are not
redundant. Another factor that may contribute to distinct roles of CDPK
isoforms in transmitting Ca2+ signals generated by different stimuli would be
the targeting of CDPKs to specific sites differing calcium-sensitivity.
Characterization of these latter properties is described in following chapter.













CHAPTER 3
CALCIUM BINDING PROPERTIES OF THREE SOYBEAN CDPKS


Introduction

One of the best studied calcium binding proteins is calmodulin.
Calmodulin is an ubiquitous protein in eukaryotes that belongs to the super
family of EF-hand proteins (Kretsinger, 1987; Moncrief, 1990; Nakayama, 1994;
Kretsinger, 1996) and it is involved in numerous cellular processes that are
regulated by Ca2+ (Cheung, 1980).
The crystal structure of the Ca2+-calmodulin complex demonstrated
that it attains a dumbbell shape: at each end of the dumbbell is a pair of EF-
hands and they are connected by a long a helix (Babu et al., 1988; Babu et al.,
1985). Comparison of X-ray solution scattering data of calmodulin and
troponin C (Heidorn and Trewhella, 1988) and studies from site-directed

mutagenesis of the central helix of calmodulin (Persechini and Kretsinger,
1988) predicted that the central helix of calmodulin may form a flexible tether.
NMR spectroscopic studies provided the first solution structure of a ternary
complex, i.e., Ca2+-CaM and its target peptide (calmodulin binding domain of
skeletal myosin light chain kinase) (Ikura et al., 1992). Two crystal structures
of the ternary complex between Ca2+-CaM and a smooth muscle myosin light
chain kinase peptide (Meador et al., 1992) or a CaM-dependent protein kinase
II peptide (Meador et al., 1993) followed. These studies showed that the
central helix was indeed bent. The conformational flexibility of the central
helix brings the two globular domains closer to each other in order to interact








with a target peptide. These findings confirmed that calmodulin can
accommodate specific binding to a large number of targets with high affinity
through hydrophobic residues in the two lobes. These hydrophobic pockets
are exposed by conformational changes upon binding Ca2+ and form a tunnel
which wraps around the target peptide due to the flexibility of the central
helix (Finn and Forsen, 1995; James et al., 1995; Torok and Whitaker, 1994).
However, the binding of Ca2+/calmodulin to its numerous target enzymes
with high affinity is rather unusual because the binding sites do not show any
sequence conservation.
The Ca2+-binding properties of calmodulin have been the subject of
numerous studies for decades since the first report by Teo & Wang (1973).
The determination of Ca2+-dissociation constants of calmodulin were
reported using equilibrium dialysis (Cox et al., 1981; Crouch and Klee, 1980;
Potter et al., 1983), or flow dialysis (Haiech et al., 1981; Haiech et al., 1980;
Starovasnik et al., 1993), or a titration technique with a chromophoric Ca2+
chelator (Linse et al., 1991). Whether there is a cooperativity among Ca2+-
binding sites has been a subject of controversy, however it is now generally
agreed that there is positive cooperativity between the two EF-hands in each
pair (N-terminal and C-terminal domains of calmodulin) and that the C-
terminal domain has about 6-8 fold higher Ca2+-binding affinity than the N-
terminal domain (Linse et al., 1991; Porumb, 1994). The largest
conformational change occurs upon binding of two Ca2+ to the C-terminal
domain (Crouch and Klee, 1980). Notably each N-terminal or C-terminal
domain maintains its Ca2+-binding property when the domains are separated
by tryptic digestion (Linse et al., 1991). This result implies that each domain
binds Ca2+ independently, although a study employing a series of site-directed
mutageneses of the conserved EF-hands indicated that mutations in N-








terminal domain can affect the conformational change in C-terminal domain
(Beckingham, 1991; Maune, 1992). The apparent Ca2+-dissociation constants
(Kd) for the high-affinity (C-terminal) and the low-affinity (N-terminal)
domains are 1 piM and 10 iM, respectively (Lnse et al., 1991; Porumb, 1994).
However, it is note worthy that the interaction between the two domains of
calmodulin is required for the high affinity Ca2+-binding in the presence of a
target peptide (Yazawa et al., 1992).
The C-terminal domain of CDPK contains four EF-hands similar to
calmodulin (Harper et al., 1991). The activity of CDPK is regulated by Ca2+
through this C-terminal domain which was named calmodulin-like domain
(CLD) (Yoo et al., 1996). The Ca2+-binding properties and the effects of
mutating each EF-hand of PfCDPK from Plasmodium falcifarum have been
reported (Zhao et al., 1994). Equilibrium dialysis showed that PfCDPK is able
to bind four Ca2+ per molecule with mean Kd of 80 gM. The K0.5 for Ca2+ was
15 iM. The binding studies suggested at this concentration of free Ca2+ 1
mole of Ca2+ binds per mole of PfCDPK. The highly conserved glutamate
residue at position 12 in the first and second EF-hands of PfCDPK was crucial
for the structural change and for the enzyme activity. Similar studies
employing point mutations with Drosophila calmodulin revealed that the
mutations to the second and fourth Ca2+-binding sites resulted in more
deleterious effect than the mutations to the first and third Ca2+-binding sites
(Maune, 1992). However, the mutated calmodulin with altered Ca2+-binding
properties was able to activate substrate enzymes (Haiech et al., 1991). Yeast
calmodulin is similar to vertebrate calmodulin (60% amino acid sequence
identity) but only three EF-hands are functional (Starovasniket al., 1993).
Intriguingly, mutant yeast calmodulins in binding Ca2+ supported the growth
of yeast while deleting the calmodulin gene was fatal (Geiser et al., 1991).








Although yeast calmodulin is required for growth, it can perform its function
without the apparent ability to bind Ca2+.
CDPK is able to sense directly a Ca2+ signal through its calmodulin-like
domain. In addition, it is a kinase, an important signaling component, that
can act to amplify the signal. Since CDPKs are encoded by a large gene family
(Estelle, et al., 1996), one can hypothesize that Ca2+ can bring about certain
physiological responses in plants by utilizing different CDPK isoforms as the
signal mediator(s). CDPK isoforms with different Ca2+ sensitivities to decode
different Ca2+ signals.
There are large number of reports regarding the involvement of Ca2+
in various physiological responses of plants (Bush et al., 1996). However,
studies demonstrating the roles of CDPKs in plants are limited. Studies on
the differences in Ca2+ sensitivity of CDPK isoforms, therefore, will be
essential to give insight into how different CDPK isoforms may play roles in
transducing Ca2+ signals. In the present study the direct measurement of
Ca2+-binding properties by the flow dialysis method of three soybean CDPKs
were undertaken. The flow dialysis method was chosen because it produces a
complete data set from a single protein sample in a short time. The rate of
Ca2+ flux across the dialysis membrane to reach a steady state is within a

couple of minutes (Colowick and Womack, 1969; Porumb, 1994; Womack and
Colowick, 1973). In addition, this method has been used in numerous studies
of calmodulin (Haiech, et al., 1981, 1991; Starovasnik, et al., 1993;Yazawa, et al.,
1992). The effect of Ca2+ concentrations on the substrate phosphorylation and
autophosphorylation were also determined. The concentration of free Ca2+
in the activity assays was set by Ca2+ buffers.








Experimental Procedures


Materials

Analytical standards of 100 mM CaC12 and 1 M MgCl2 were purchased
from Orion and Fluka, respectively. Solutions of 45CaC12 (29.6 Ci/g, 10
mCi/mL) and [y-32p]ATP were supplied by DuPont NEN. Spectra/Por
Macrodialyzer and dialysis tubings (molecular cutoff -3,500) for flow dialysis
were obtained from Spectrum Co. Chelex 100 was purchased from Bio-Rad.
Fluo-3 pentapotassium salt was from Molecular Probes. All other chemicals
were reagent grade or higher.

Protein Purification

Molecular cloning of plasmid constructs for the expression of GST-
fusion proteins of soybean CDPK isozymes were described in Chapter 2.
Bacterial expressions of recombinant proteins were scaled up to several liters
to obtain enough protein for the flow dialysis experiments. The induction

conditions were optimized to get proteins in high yield: IPTG was added to 0.5
mM when OD600 reached -0.5 and cells were incubated for 3 hours of
induction at low temperature (-20 OC). Depending on the construct, -0.5-6 mg
of proteins/liter of bacterial culture could be obtained. Protein purification
procedures were the same as described in Chapter 1 except that the
equilibration buffer for ionic exchange chromatography contained only 20
mM Tris, pH 8.0.








Removal of Contaminating Calcium

Glass containers were avoided in the experiments involving calcium
binding measurements. Plastic bottles were filled with deionized water and
autoclaved. Small containers and graduated cylinders were soaked in 0.1 M
HCI and rinsed thoroughly with deionized water. Water and solutions used
for flow dialysis were decalcified by passing through a Chelex-100 column
(Crouch and Klee, 1980). The membranes for the flow dialysis were stored in
10 mM EGTA at 4 OC and thoroughly rinsed with deionized water followed by
final washes in Chelex-treated water prior to flow dialysis. Stock solutions of
enzymes were dialyzed in Chelex-treated buffer A (50 mM HEPES, pH 7.5, 100
mM KCI) in the presence or absence of 14.4 mM mercaptoethanol at 4 C
overnight and passed through Chelex column equilibrated with buffer A
(Stemmer and Klee, 1994). Following the Chelex-treatment the protein
samples were concentrated using a centrifugal concentrator (Centricon-10,
Amicon) which was prewashed with Chelex-treated buffer A.

Calcium Measurements

Contaminating Ca2+ in water, buffers, and protein samples was
measured using fluorescent Ca2+ indicator Fluo-3 (Minta et al., 1989), as
described elsewhere (Eberhard and Erne, 1991 and 1994). Calcium calibration
buffers from Molecular Probes were used for generating a standard curve
following the recommendations of the supplier. Briefly, Fluo-3 was dissolved
in DMSO to make 1 mM stock. Aliquots of 200 pL each in microcentrifuge
tubes were stored at -80 C. Fluo-3 (0.5-5 pM) was added to 2 ml sample for the
measurement of free calcium. A fluorometer cuvette was soaked in 0.1 M
HCI for at least 30 minutes and rinsed thoroughly with deionized water








before the fluorescence measurement. Free calcium concentrations in

Ca2+/EGTA buffers containing <1 iM free calcium were verified also by using
Fluo-3. The contaminating calcium concentration in Chelex-treated flow
dialysis buffer (50 mM HEPES and 100 mM KC1) and protein solutions was
less than 0.3 UM according to the fluorescence measurement with Fluo-3 and
atomic absorption spectrometry. Fluorescence was measured with Perkin-
Elmer LS5 spectrofluorometer. Concentrations of various CaC12 stock
solutions that were made by diluting 100 mM CaCl2 in Chelex-treated water
and kinase assay buffers containing > 1 UM free calcium were confirmed by
atomic absorption spectroscopy.

Calcium Binding Studies

The calcium binding properties of CDPK isoenzymes were studied
using flow dialysis method (Colowick and Womack, 1969; Womack and
Colowick, 1973; Porumb, 1994). The experiments were performed using a
dialysis apparatus (Spectra/Pore Macrodialyzer) at room temperature (24 2
OC). The apparatus consisted of two dialysis cells that were separated by a
dialysis membrane. The upper cell chamber contained 7 liM to 30 pM of
metal-free protein sample in 1 ml buffer A (50 mM HEPES, pH 7.5, 100 mM
KCI) and the lower chamber was filled with 1 ml of buffer A and
continuously pumped to the effluent collector (Figure 3-1).
The additions of solutions into the upper chamber were made through
a port using ultra-thin gel loading pipette tips (United Scientific Products).
The solutions were constantly mixed with a magnetic stirring bar (7 mm X 2
mm) in each chamber. The buffer was pumped from the lower chamber
using the multistaltic pump (Buchler Instruments) at a flow rate of 3 ml/min.
The Ca2+ titration was initiated by adding an aliquot of 1.5 UM to 7.5 uM





















protein sample

[7/ ~CaCI2 and
membrane -- CaCl2 addition

fraction pump buffer
collector F -

stirring bar





Figure 3-1. A schematic representation of flow dialysis system.
The flow dialysis apparatus (Spectra/PorMacroDialyzer) was supplied
from Spectrum. The upper and lower cell chambers each can hold I ml
of sample. The upper cell contained protein sample in dialysis buffer.
Injection of Ca2+ was made through an open port as indicated. The
lower cell contained dialysis buffer which was continuously pumped
from a buffer reservoir to the fraction collector by multistaltic pump as
explained in the Experimental procedures.








45CaC12 to the protein sample in the upper chamber and continued by adding
1.25 pl to 4 il of different CaCl2 stock solutions in each cycle. The time lag
from the moment when the buffer left the lower chamber until it reached the
fraction collector was measured and the correct timing of CaCI2 addition was
adjusted accordingly. The final chase-out was carried out with 4 mM CaCI2.
The effluent was collected in 1.5 ml fractions in microcentrifuge tubes every
30 seconds. Each was mixed with 2 ml of SintiVerse (Fisher) for the
radioactivity measurement in the Scintilation counter (Beckman). Steady
states of radioactivity were reached at the third fraction of each cycle. The
average counts from the last two fractions (3rd and 4th) of each cycle were
taken to calculate free calcium concentration from known initial calcium
concentration. The moles of bound calcium per mole of protein was
calculated by combining the concentration of bound calcium and protein
concentration as described by Porumb, 1994, i.e.,
[CaC] = [CaCltal X (volume) initial (cpm)-
(volume) i (cpm)
[CaCl 2bun = Ca aCl2tota [CaC 2]fre

(mole of bound CaCI 2 ) [CaCl 2]bund
mole of protein [protein]
where i and F represent i th cycle and final chase-out, respectively. It was
assumed that there is no loss of 45CaC12 by diffusion through the membrane
and thus the volume change in each cycle by the calcium addition was
substituted for 45CaC12 concentration change in the calculation. Control
experiments in the absence of protein sample confirmed that a steady state
was reached within 1.5 minutes and showed that the loss of 45CaC12 by
diffusion during flow dialysis through the membrane was negligible (-5%).








The calcium binding data were processed according to the theoretical
binding models, the Hill Model (Cornish-Bowden and Koshland, 1975;
Dahlquist, 1979) or the Klotz Adair Model (Fletcher et al., 1970) using
MacCurveFit program. The quality of the data, however, was not good
enough for the determination of macroscopic binding constants. Therefore,
the Hill model with fewer parameters was chosen for the binding isotherm
analyses. The Hill Model provides information regarding degrees of
cooperativity (Hill constant, a), maximum number of Ca2+-binding sites (n),
and apparent dissociation constant (K d (mol/liter)) according to the following
equation,

nX "
Kd +Xa

where r and X denote the average number of moles of Ca2+ bound per mole
of protein and the free Ca2+ concentration, respectively (Dahlquist, 1979;
Porumb, 1994; Cantor and Schimmel, 1980).

Protein Concentration Determination

Protein concentrations were measured according to Bradford (Bradford,
1976) assay using bovine serum albumin as a standard or by optical density at
280 nm from calculated extinction coefficients for each isoenzyme (Gill and
von Hippel, 1989). Protein concentrations determined by both methods were
in good agreement.

Protein Kinase Assays and Autophosphorylations

Kinase assays were performed in the presence or absence of calcium as
described in Chapter 2, with the following modifications. The assays were
carried out in a buffer containing 50 mM HEPES, pH 7.4, 100 mM KCI, 5 mM








MgCl2, 2 mM DTT, 60 pM [y-32P]ATP (about 500 cpm /pmol), 100 pM syntide-
2, and 10 ng of each isoenzyme at 24 2 OC for 6 minutes in the presence of
free calcium with varying concentrations from 0.1 jiM to 100 pM or in the
absence of calcium (10 mM EGTA). At the end of each reaction, 10 AL aliquots
were spotted onto precut (about 1.5 cm2) P-81 phosphocellulose paper and
washed 3 times for 1 minute each in 10% phosphoric acid. The remaining
radioactivity was measured by liquid scintillation counting; Ca2+/EGTA
buffers containing 0.1 gM to 1 pM free calcium in the presence of 0.2 mM free
EGTA were prepared according to Tsien and Pozzan (1989). Assay buffers
containing free calcium from 5 uM to 100 pM were made by direct dilutions
from 0.1 M CaCl2 standard solution purchased from Orion as recommended
by Bers et al. (1994). Activity assays of CDPKa using protein substrates were
performed in the same assay mixture except that 2 pg/ml of the enzyme was
used and 0.5 mg/ml of histone IIIS or 0.2 mg/ml serine acetyltransferase
(SAT) (Yoo and Harmon, submitted) substituted peptide substrates.
Autophosphorylation of each isoenzyme were performed in the
absence of the substrate at room temperature for 15 minutes. The total 25 gL
assay mixtures contained 2.5 pg enzymes in the same buffer used for the
activity assays.


Results


Direct Ca2t-Biding Studies

The binding of Ca2+ to CDPK isoenzymes measured by flow dialysis is
shown in Figure 3-2. The Ca2+-binding curves were analyzed by fitting the
curve according to the Hill model as described in Experimental Procedures.
The resulting parameters obtained by the Hill model are listed in Table 3-1.








Examination of the Ca2+-binding isotherms of CDPKa and y revealed rather

complex patterns compared to that of CDPKp.


Table 3-1. Ca2+ dissociation and Hill constants of CDPK isoenzymes
enzymes Kd (pM) a
CDPKa 44.7 (3)* 1.2(0.1)
CDPKp 1.5(0.1) 1.4(0.1)
CDPKy 1.1 (0.2) 0.6
*In parentheses, standard deviations of the fitted values are shown. The
parameters were obtained from at least two determinations. The goodness of
fit (R2) for each enzyme (CDPKa, p, and y) was 0.91, 0.95, and 0.93,
respectively.


The Ca2+-binding curve of CDPKp was well fitted to the Hill model
with apparent Kd of 1.5 gM and Hill constant a of -1.4 suggesting positive
cooperativity (Table 3-1). Since soybean CDPKs contain predicted four perfect
E-F hand modules, 4 moles of Ca2+ were expected to bind per mole of enzyme.
But the maximum binding capacity according to the fitted Hill model was
slightly lower (n =3.4). Considering the residual bound Ca2+ in CDPK samples
after passage through Chelex-100 column being < 0.3 mole of Ca2+ per mole of
enzyme, the n could be increased to 3.7.
The acceptable Ca2+ contamination level in the protein sample is
suggested to be 0.025-0.1 mole of Ca2+ per mole of Ca2+-binding sites (Porumb,
1994). However, the success of decalcification depends on the affinity for Ca2+
of a protein of interest and stability after a harsh decalcification method like
trichloroacelic acid (TCA) precipitation which can be effectively used for
calmodulin (Haiech et al., 1981). For CDPKs, which are chimeric enzymes
containing catalytic domain and calmodulin-like domain, only mild and
limited treatments may be used to remove contaminating Ca2+ while keeping
the enzyme active. It is possible that the quality of experiments including the





















1 5-




E
34- 0
cc


2 0



1-



2 -




-9 -8 -7 -6 -5 -4 -3

Log [Ca2+ M


Figure 3-2. Direct Ca2+ binding to CDPKa, p, and 7.
Ca2+ binding data were obtained from flow dialysis as described in the
experimental procedures. The protein concentration was 20 pM in 50
mM HEPES, pH 7.5 containing 100 mM KCl at 25 2 C.
a) CDPKa; b) CDPKj; c) CDPKy. For CDPKy, 2 mM DTT was
supplemented for the enzyme stability. The average number of moles
of Ca2+ bound per mole of protein is plotted as a function of free Ca2+
without correction of contaminating Ca2+. The solid lines represent
computer fitted curve of the binding data (0) according to the Hill
model using MacCurveFit program.








protein purity, enzyme stability during dialysis, and accuracy of pipetting may
have caused this apparently lower number of binding sites. Errors in protein
measurement would also affect the number.
CDPK- is a highly unstable enzyme (Chapter 2). Full-length CDPKy
precipitated against flow dialysis buffer containing 100 mM KCI during
overnight dialysis. The protein remaining in the supernatant of dialysate was
recovered and subjected to the flow dialysis. However, the enzyme again
precipitated after the first injection of Ca2+ was made. Therefore, the N-
terminal deletion construct, CDPKX66-538) was chosen to replace the full-
length enzyme for the Ca2+-binding study. The enzyme activity of CDPK-(6-

538) was comparable to that of the full-length enzyme but it tolerated the
dialysis in the low salt buffer (Chapter 2). Although CDPK(66-538) could be
used for the flow dialysis without precipitation, it was not as stable as the
other isoenzymes. GST- CDPK(66-538) was freshly prepared each time
because the enzyme become unstable after overnight storage. To show that
Ca2+ binding property of CDPKy was not changed by the deletion of the N-
terminus in CDPKy(66-538), the effect of Ca2+ concentration on the kinase
activity was examined. The result confirmed that the response was identical
to that of the full-length enzyme (data not shown).
The Ca2+-binding data of CDPKy(66-538) was fitted to the Hill model by
setting n =4 due to the absence of apparent saturation of Ca2+ binding to the
enzyme. The resulting Kd and Hill constant a were -1 pM and -0.5 iM,
respectively. The Klotz-Adair model (Klotz, 1983; Porumb, 1994) was better in
the curve fitting for CDPKy. The resulting macroscopic dissociation
constants, however, showed high values of error, which made the result of
the analysis difficult to interpret. These results may imply that the behavior








of Ca2+ binding of CDPKy is complicated and requires further intensive
analyses with better data from flow dialysis or other methods.
CDPKa also showed no apparent saturation of Ca2+ binding (Figure 3-
2a). The Ca2+-binding data fitting for CDPKa was undertaken according to
the Hill model but the parameters were obtained by setting n to 4 as for
CDPKy above. CDPKa had a higher Kd (-45 pM) relative to the other CDPK
isoenzymes and nonspecific binding was observed at around 100 pM free Ca2+
making hard to predict the maximum number of binding with the Hill
Model without setting n to 4. Nonspecific binding of Ca2+ was detected for all
three enzymes. This may be attributed to the limitation of flow dialysis
method (Kakalis et al., 1995; Porumb, 1994). The examination of the Ca2+-
binding curve of CDPKa suggested that it is similar to that of PfCDPK (Zhao et
al., 1994). The Ca2+-binding to PfCDPK was measured by equilibrium dialysis
and the resulting Kd and n were reported to be 80 pM and 3.9 respectively.

Analysis of Ca2Z-Dependent Kinase Activity of CDPKs

To elucidate how the Ca2+-binding property of each CDPK isoenzyme

affects its kinase activity, changes of kinase activity in response to various
Ca2+ concentration were investigated (Figure 3-3). The Ca2+ concentrations

for half maximal kinase activity (Ko.5) for CDPKp and y were 0.5 and 1 pM,
respectively. These values were closely correlated to their Ca2+ binding
behavior, which suggested that 1 or 2 calcium ions bound per mole of each
enzyme may be sufficient for the half maximal enzyme activity. The kinase
activities of these two enzymes were saturated at a similar concentration, ~ 5
pM. At this Ca2+ concentration, Ca2+-binding to CDPKp was also saturated
(Figure 3-2b) implying that occupation of four Ca2+-binding sites was required
for the maximum activity of CDPKp.
























QB
S 100"





25

0

C


75-
100- T "r"



50

25


0 -7 -6 -6 -4
Log [Ca2 ] M


Figure 3-3. The effect of Ca2+ on kinase activity of CDPKs.
Enzyme activity assays were performed in the buffers with varied free
Ca2+ concentrations or in the presence of 1mM EGTA (indicated as zero
Ca2+ on abscissa) using syntide-2 as a peptide substrate as described in
the Experimental Procedures. Standard deviations are shown as
vertical bars from at least two independent experiments.
a) CDPKa; b) CDPKP; c) CDPKy.








The KO.5 of CDPKa was -0.1 gM (Figure 3-3a). This result was very
surprising because the Kd for Ca2+ of CDPKa was the highest (-45 gM) of the
three isoenzymes and the stoichiometric Ca2+ binding of CDPKa occurred at
calcium concentration above 10 pM (Figure 3-2a). Moreover, the saturation of
kinase activity of CDPKa appeared to be only about 0.3 pM. This value is also
far below the concentration of stoichiometric Ca2+ binding to the enzyme and
about ten-fold lower than that of the other two enzymes.


Ca+-Binding Studies of CDPKa in Various Conditions Resembling the
Enzyme Assay Mixture

The Ca2+-binding behavior of CDPKa was further examined in order to
understand what caused the apparent change in Ca2+ sensitivity of the
enzyme in the activity assay. Since there were additional components in the
assay mixture such as ATP, MgCl2, DTT, and a peptide substrate, syntide-2, the
effects of these components on the Kd of CDPKa were inspected. First, assay
components that were not present in the flow dialysis buffer were added
alone or as a combination to the dialysis buffer and the flow dialysis was

carried out in the same manner as described in Experimental Procedures.

MgCl2 and DTT did not show any effect on the Ca2+-binding of CDPKa (data
not shown). However, as shown in Figure 3-4, quite interesting results were
obtained in the presence of peptide substrate, syntide-2, and ATP. The Kd for
Ca2+ of CDPKa was markedly reduced to -1 WM from -45 pM. This striking
shift of Kd apparently resembled the influence on the Kd for Ca2+ of
calmodulin in the presence of a peptide containing the calmodulin binding
site (Porumb, 1994; Yazawa, 1992). The enhanced Ca2+-binding affinity of
CDPKa was mainly caused by the presence of syntide-2. The presence of ATP


















X 5-
S 4-
0






E *- -- -, ,- -
3

2 -





-8 -7 -6 -5 -4 -3
Log [Ca2+ ] M




Figure 3-4. Ca2+-binding studies of CDPKa under various conditions.
Flow dialysis was performed to examine the effect of syntide-2 and (or)
ATP on Ca2+-binding property of CDPKa. The sample in the upper
chamber contained 7 pM of CDPKa and 1.5 pM of 45CaC12 in a dialysis
buffer (50 mM HEPES, pH 7.5, 100 mM KC1, 5 mM MgC12, and 60 iM
ATP). The fitted curves of Ca2+-binding data in the presence (o) or the
absence (A) of 100 ILM syntide-2 are shown. The control experiment (-)
performed in a dialysis buffer containing 50 mM HEPES and 100 mM
KC1 is from Figure 3-2a.








alone slightly affected the Kd (Figure 3-4) but syntide-2 alone could cause the

shift even more dramatically than when combined with ATP (data not

shown).

Syntide 2 is the best substrate peptide tested for all three soybean CDPKs

(Chapter 2). It was asked whether other peptides which are not good

substrates for CDPKa but contain the motif, basic-X-X-ser/thr could shift the

Kd. Including H1-7 (RRKASGP) or skeletal muscle myosin light chain kinase

substrate (AKRPQRATSNVFS) in the Ca2+-binding experiments resulted in
the Kd shift similar to that caused by syntide-2 (data not shown). It can be
speculated from these results that these peptides are capable to bind the

enzyme as well as syntide-2 and cause the shift of Kd for Ca2+ but the
phosphorylation by CDPKa is not favorable. It was concluded that the low

K0.5 of CDPKa as shown in Figure 3-3a is not an experimental anomaly but
caused by syntide-2. If this titration curve (Figure 3-3a) is compared directly to

the binding curve in the presence of syntide-2 shown in Figure 3-4, it is clear

that kinase activity is well correlated to that of Ca2+-binding behavior in the

same condition as activity assay.


K0.5 of CDPK in the Presence of Protein Substrates

The observation that Kd for Ca2+ of CDPKa could be substantially decreased in

the presence of syntide-2 raised a question whether the phosphorylation by
CDPKa occurs always at low Ca2+ regardless of its substrate. Two protein
substrates, histone IIIS and SAT were chosen for the further examination
regarding this question. Histone HIS is not a favorable substrate of CDPKa but
the phosphorylation occurred sufficiently to detect the effect of Ca2+ on the
kinase activity. Interestingly, K0.5 for histone IIIS phosphorylation by CDPKa
turned out to be about 4 iM (Figure 3-5) which is forty-fold higher than that






















75-


S 50


25-



0 -7 -6 -5 -4
Log [Ca2 ]M




Figure 3-5. Changes in K0.5 of CDPKa by different protein substrates as a
function of free Ca2+ concentration.
The effect of different protein substrates on the activity of CDPKa in
response to Ca2+ concentration was determined. The assays were
performed as described in Experimental procedures by adding 0.5
mg/ml histone IIIS (o) or 0.2 mg/ml SAT (e) as a substrate.









for syntide-2 phosphorylation (K05 -0.1 pM). Saturation of histone IIIS
phosphorylation occurred at ~30 gM. These results appear to be consistent
with the Ca2+-binding behavior observed in the absence of syntide-2 (Figure
3-2A and Table 3-1) and imply that Kd for Ca2+ of CDPKa may depend on its
substrate. These results also confirm the validity of Ca2+-binding data
obtained in the absence of a substrate peptide.

Since it was observed that H1-7 or skeletal muscle myosin light chain
kinase substrate were able to change Kd for Ca2+ of CDPKa but were hardly
phosphorylated by the enzyme, it was asked whether the presence of these

peptides may affect the Ko.5 for histone IIIS phosphorylation by CDPKa.
When these peptides were added to the assay mixture, however, the K0.5 for
histone IIIS phosphorylation was not affected (data not shown). These results
imply the possibility that these peptides were unable to bind the enzyme in
the presence of histone IIIS.
A more remarkable observation was made when SAT was used as a
protein substrate for CDPKa. SAT is a novel in vitro substrate of CDPKs
identified by interaction cloning (Yoo and Harmon, submitted). As shown in

Figure 3-5, K0.5 for SAT is -0.1 pM, which is identical to that for syntide-2.

These results propose that certain substrates can induce the high affinity for
Ca2+ of CDPKa despite the intrinsically low affinity for Ca2+ of CDPKa. If this

is true in vivo, it could be a mechanism of regulating the activity of CDPKa in

response to different Ca2+ signals and for the selection of the specific substrate.


Autophosphorylation of CDPKs in Response to Ca2+

The effects of Ca2+ on autophosphorylation of CDPKs were investigated.
Maximum autophosphorylation of CDPKa was observed at very low Ca2+
concentration (-0.3 pM) (Figure 3-6). But the autophosphorylation is Ca2+-




























75

50

25




100

75

50 -

25

0
0 -7 -6 -5 -4

Log [Ca2 ] M




Figure 3-6. The effect of Ca2+ on autophosphorylation of CDPKs.
a) Autophosphorylation of CDPKa, b) p, and c) y was performed as
described in Experimental Procedures.








dependent because only basal level (less than 10 % of maximum
autophosphorylation activity) of autophosphorylation was observed in the
presence of 1 mM EGTA. For CDPKp and -, -5 VM of free Ca2+ was required
for the saturation of the autophosphorylation. It is intriguing that
autophosphorylation of CDPKa is highly sensitive to low Ca2+ because high
affinity Ca2+ binding site(s) was not detected when a substrate is absent.
Moreover, the contaminating Ca2+ was measured to be less than 0.3 mole of
Ca2+ per mole of enzyme in the flow dialysis buffer. Therefore, if there is
high affinity Ca2+ binding site(s) in CDPKa the Kd for that site(s) may be lower
than the Ca2+ contamination level in the flow dialysis but probably higher
than that of EGTA (-60 nM under the assay conditions).


Discussion


CDPK Isoforms Differ in Their Ca2+-Binding Properties

The Ca2+-binding properties of three soybean CDPK isoforms were
investigated to test the hypothesis that each CDPK isoenzyme may be able to
transduce distinct Ca2+ signals upon stimuli through its altered Ca2+-
sensitivity. The results clearly provided the biochemical evidence that each
isoenzyme differs in Ca2+-binding properties. Although there were some
experimental difficulties due to the enzyme stability during purification and
flow dialysis, the Ca2+-binding data obtained with recombinant CDPKs were
reproducible.
Fitting and interpreting data according to the simple ligand binding
models were not straightforward though the correlating coefficients were
high. It is possible that the data quality was not high enough and/or the Ca+-
binding behavior of CDPKs was too complicated to be predicted by the








theoretical binding models. For example, Hill constant a for CDPKy was

predicted to be 0.6 (Table 3-1). Hill constant a < 1 is interpreted as
anticooperativity (Porumb et al., 1994). Zhao et al. (1994) have ascribed the
quality of their Ca+-binding data for PfCDPK to the difficulty of analysis with
theoretical binding model. Fitting the Ca2+-binding data of CDPKp with Hill
model, however, was relatively satisfactory. The result suggested positive

cooperative binding. The saturation of Ca2+-binding to CDPKp was observed
at -10 gM of free Ca2+ unlike CDPKa and y. The nonspecific binding of Ca2+
to CDPKa and y at and above 100 gM was a probable limiting factor to get
saturation of these enzymes.
The mean Kds of CDPKp and y were similar (1.5 and 1.1 jiM,
respectively), though the shape of Ca2+-binding isotherms of these enzymes
were distinctive (Figure 3-2). These values of Kd are in good accordance with
that of calmodulin (1.5 to 5 pM). CDPKa, however, had a much higher Kd of
about 45 jM. The Kd for Ca2+ of CDPKa was comparable to the Kd of PfCDPK
(Zhao et al., 1994). Since the amino acid sequence comparison showed that

CDPKa and p are closer to each other than each of them to CDPKy, the result
was unexpected. Moreover, the sequence of PfCPK is highly diverged not

only from the sequence of CDPKa but also from the sequences of any of

CDPKs in plants. The sequence analysis of CLD domains of three soybean
CDPKs and calmodulin showed that the sequence identity between
calmodulin and soybean CDPKs were -40% but the amphiphilic
characteristics of calmodulin was highly conserved in CLD domains (data not
shown). In addition, quite a few residues in calmodulin known to interact
with a target peptide derived from smMLCK (Meador et al., 1993) were
consistently conserved in the CLD domains of the three CDPKs. These results
from the sequence analysis suggested that it is difficult to point out which








residues are responsible for the differences in Ca2+-binding properties,

especially, between CDPKa and p. It will be necessary to learn the actual
structures of CDPKs and the relationships between CLD and autoinhibitory
domain for the elucidation of the differences in Ca2+-binding properties

among CDPKs.


CDPKn Shows High Ca2+ Sensitivity in the Presence of Certain Substrates

The Kd and KO.5 for Ca2+ of CDPKp and y were almost equivalent (Table
3-1 and 3-2). In contrast, the K0.5 of CDPKa was over 400-fold lower than its
Kd. Further examination showed that the Kd for Ca2+ of CDPKa decreased to

a value near the Kd when syntide-2 was included in the direct Ca2+-binding
experiments. Also it was observed that autophosphorylation of CDPKa was

Ca2+-dependent but its requirement for free Ca2+ was very low (< 0.3 RlM).


Table 3-2. K0.5s for Ca2+ of CDPKa, p, and y with various substrates.

Syntide-2 SAT Histone IIIS

Enzyme K0.5 (1iM)

CDPKa 0.1 0.1 4

CDPKp 0.5 nda nda
CDPKy 1 nda nda
and, not determined.


CDPKa may contain at least one high affinity Ca2+-binding site which
was not detectable in the flow dialysis experiment. Since the contaminating
Ca2+ concentration in the buffer was -0.3 iM, it can be speculated that the
high affinity Ca2+-binding site may have been filled prior the initiation of
Ca2+-titration, if the Kd of this site was lower than 0.3 tM. The presence of








high affinity Ca2+-binding site in CDPKa could be tested in an alternative
method to the flow dialysis. Determination of Ca2+ dissociation constants
between 10-7 and 10-9 M has been obtained from Ca2+ titrations of the protein
in the presence of chromophoric Ca2+ chelators such as Quin2, BAPTA (bis(2-
aminophenoxy)ethane-N,N,N',N'-tetraacetic acid), or Fluo-3 (Eberhard and
Erne, 1991, 1994; Linse et al., 1991; Waltersson et al., 1993). The binding of
Ca2+ to these compounds was monitored by the absorbance at the specific

wave length.

The enhancement of Ca2+-binding affinity of CDPKa in the presence of
syntide-2 resembled that of calmodulin induced by the complex formation
with its numerous target proteins or peptides such as MLCK, cyclic nucleotide
phosphodiesterase (Olwin and Storm, 1985), caldesmon, mastoparan (Yazawa
et al., 1987), and calmodulin binding peptides derived from plasma
membrane calcium pump (Yazawa et al., 1992) and calcineurin (Stemmer and
Klee, 1994). CDPK contains an autoinhibitory domain that acts as a
pseudosubstrate (Harmon et al., 1994; Harper et al., 1994; Yoo and Harmon,
1996). The binding of Ca2+ to calmodulin-like domain is proposed to release
the autoinhibitory domain from the catalytic domain thus activating the
enzyme. The enzyme activity of CDPKa is only basal level in the presence of
EGTA. If it is presumed that CDPKa has a high affinity Ca2+-binding site and

the binding of Ca2+ to this site gives rise to a conformation which syntide-2

can bind to the catalytic domain resulting in freeing CDL binding site, the CLD
could form a complex with the CLD binding site. This complex formation
might have driven the shift of Ca2+-binding affinity, by analogy to the
enhanced Ca2+-binding affinity of calmodulin by forming a complex with a
target peptide. The fact that the enzyme activity of CDPKu was minimal in








the absence of Ca2+ could be interpreted that the Kd of the high affinity Ca2+-
binding site may be lower than that of EGTA.
It was intriguing that CDPKa showed comparably low K0.5 for Ca2+ in
phosphorylating a novel in vitro protein substrate, SAT, and syntide-2, while
the K0.5 for Ca2+ was much higher in the presence of another protein
substrate histone iIS. The inducible Ca2+ affinity of CDPKa could serve as
another mechanism controlling the CDPK activity in vivo for selecting
specific substrate(s) on top of the apparently narrow substrate preference of
the enzyme. Some substrate might be phosphorylated only at high free Ca2+
concentration in the cell if they are not effective in changing the Kd for Ca2+
of CDPKa. It may be possible that local Ca2+ concentration rises high enough
in certain cases so that CDPKa become active to phosphorylate such
substrates. In this case CDPKa may need to be localized in vicinity of the Ca2+
sources. The presence and the availability of the substrate that are capable to
enhance Ca2+-binding affinity of CDPKa may be an important factor for the
phosphorylation of those substrates because the resting level of free Ca2+ in
the cell may be high enough to activate the enzyme. It will require further

research to test these speculations but will provide much insight into this
likely novel regulatory mechanism of CDPKa.




Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E85PB2JIZ_X1QL3J INGEST_TIME 2013-09-28T02:04:28Z PACKAGE AA00014282_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES