Pine chitinase gene structure, expression and regulation


Material Information

Pine chitinase gene structure, expression and regulation analysis in pine cells and in heterologous systems
Physical Description:
xii, 139 leaves : ill. ; 29 cm.
Wu, Haiguo, 1967-
Publication Date:


Subjects / Keywords:
Plant cell walls   ( lcsh )
Plant cells and tissues   ( lcsh )
Plant molecular biology   ( lcsh )
Pine -- Genetics   ( lcsh )
Plant Molecular and Cellular Biology thesis, Ph. D   ( lcsh )
Dissertations, Academic -- Plant Molecular and Cellular Biology -- UF   ( lcsh )
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1996.
Includes bibliographical references (leaves 117-138).
Statement of Responsibility:
by Haiguo Wu.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 002218216
oclc - 36750808
notis - ALF8391
System ID:

This item is only available as the following downloads:

Full Text







This dissertation is dedicated to my Lord and Savior Jesus Christ, and to my

wife, Yang.


I would like to express my sincere appreciation to the members of my

graduate committee: Dr. John Davis, Dr. Ken Cline, Dr. Bill Gurley, Dr. Corby Kistler

and Dr. James Preston for their critical advice, suggestions and discussions, which

have greatly improved this dissertation. In particular, Dr. Gurley has provided me

with constant and enthusiastic support over the years.

I am also indebted to Dr. Don McCarty for allowing me to access his particle

gun, to Dr. Chien-Yuan Kao and Dr. Eva Czarnecka-Verner for teaching me how to

use this gun for bombardment experiments, and to Dr. Rosie Simmen for access to

the automated luminor for luciferase assays. I wish to extend my thanks to Dr.

Craig Echt for sequencing the initial genomic clone and providing part of the DNA

primers for this study, to Don Baldwin for helpful suggestions and discussions about

primer extension experiment.

I would like to thank all the past and present members of Dr. John Davis'

laboratory. My thanks go to Dr. Mark Lesney for helpful discussions and assistance

with the cell cultures, to Dr. Mick Popp, Dr. Yong Qian and Buddy Tignor for their

assistance and friendship. I would like to recognize the excellent technical support

of Tess Korhnak and Thea Edwards, whose collective contributions to this work

ensured the smooth and efficient running of the program.


My most special thanks go to my advisor, Dr. John Davis, for his invaluable

advice and guidance, for his encouragement and generosity, for his tremendous

support and unending confidence in me, for all his efforts on my behalf throughout

the course of this study. Working under his supervision has been a wonderful

experience in my life.

Finally, I would like to thank my wife, Yang, for her continual love,

encouragement and support, for without her and my faith in God, this work would

not have been completed.


ACKNOWLEDGMENTS ........................................... iii

LIST OF TABLES ............................................... vii

LIST OF FIGURES .............................................. viii

LIST OF ABBREVIATIONS ........................................ x

ABSTRACT ..................................................... xi

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

General Features of Plant Defense Responses ................... 1
Chitosan, General Elicitors and Specific Elicitors ............. ... 16
Pathogenesis-Related Proteins ............................... 20
Chitinase Structure, Function and Regulation .................... 27

INTRODUCTION ............................................... 35

MATERIALS AND METHODS ..................................... 41

Plant Materials ............................................ 41
Sequence and Sequence Analysis ............................ 42
Plant Transformation ............... ........................ 43
Elicitor Treatment and RNA Isolation ................... .... 44
Primer Extension .......................................... 47
Southern and Northern Analysis .............................. 47
Cloning of the Pschi4 cDNA .................................. 49
cDNA Expression and Generation of Antibody ................... 51
Protein Isolation and Western Blotting Analysis .................. 52
Particle Bombardment and Transient Expression ................. 55
Histochemical Assays in Transgenic Tobacco .................... 60

RESULTS ..................................................... 62

Pschi4 Gene Structure ...................................... 62
Pschi4 cDNA Cloning and Expression in Bacteria ................. 66
Pschi4 Expression ......................................... 73
Transient Assay of Pschi4 Promoter-GUS Constructs .............. 81
GUS Expression in Stably Transformed Tobacco Plants ............ 89
Developmental Regulation of Pschi4 Expression ................. 96

DISCUSSION ................................................. 101

SUMMARY AND FUTURE DIRECTIONS ........................... 110

APPENDIX ................................................... 114

REFERENCES ................................................ 117

BIOGRAPHICAL SKETCH ....................................... 139


Table page

1 Chitosan-inducted mRNA accumulation in transgenic tobacco plants .. 75

2 Summary of WP-GUS tobacco plants .......................... 90


Figure page

1 Gene-for-gene interactions specify plant disease
resistance or susceptibility ...................................... 5

2 Different actions between endochitinases and exochitinases ........ 28

3 Domain structure of three classes of chitinases ................... 30

4 Overall structure of the pine genomic
subclones gPschil and gPschi4 ............................. 38

5 Sites within Pschi4 that were used to design
oligonucleotide primers for this study .................. .... 48

6 Plasmid constructs ......................................... 56

7 Partial nucleotide sequence and translation product
encoded by the genomic clone containing Pschi4 ............... 63

8 Primer extension analysis to reveal the putative
transcription start site(s) ................................... 64

9 Domain structure of the putative Pschi4 protein from pine
with class I and II chitinase from tobacco ...................... 67

10 Sequence alignment of Pschi4 with tobacco chitinases ............ 68

11 Genomic Southern blot analysis of DNA from three pine species ..... 69

12 Cloning of Pschi4 cDNA by RT-PCR ........................... 70

13 Pschi4 cDNA expression in bacteria .................... .... 72

14 Transcripts accumulation in chitosan-treated pine cells ............. 74

15 Northern blot showing expression of Pschi4 in a
transgenic tobacco plant ................................... 76

16 Pschi4 protein expression in pine suspension cells ................ 78

17 Chitosan-induced Pschi4 protein expression
in pine suspension cells ................................... 80

18 Western blot analysis in tobacco suspension cells ................ 82

19 No chitosan-induction in transient assays in onion cells ............ 84

20 Promoter activity in onion cells ............................... 85

21 Promoter comparison in maize and in pine cells .................. 87

22 Promoter activity in pine cells ................................ 88

23 Particle bombardment per se induces promoter activity ............ 93

24 GUS activity was not induced by chitosan in
stably transformed tobacco plants ........................... 94

25 Mechanical wounding induced promoter activity .................. 95

26 Phosphate induced WP-GUS expression in transgenic tobacco ...... 97

27 X-gluc staining of tobacco pollen .............................. 98

28 Pschi4 protein expression in tobacco pollen ............... .... 100


2, 4-D 2,4-dichlorophenoxyacetic acid

2-iP N6-2-isopentenyl-adenine

BAP 6-benzyl-aminopurine

BCIP 5-bromo-4-chloro-3-indolyl phosphate

DMF dimethyl formamide

DTT dithiothreitol

EDTA ethylenediamine tetraacetate (disodium salt)

GUS p-glucuronidase

IPTG isopropyl 3-D-thiogalacto-pyranoside

LB Luria-Bertani medium

MS Murashige-Skoog medium

MUG 4-methylumbelliferyl p-D-glucuronide

NAA a-naphthalene acetic acid

NBT nitro blue tetrazolium

PVPP polyvinyl-polypyrrolidone

SDS sodium dodecyl sulfate

X-gluc 5-bromo-4-chloro-3-indolyl-p-D-glucuronide

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



Haiguo Wu

December 1996

Chairman: Dr. John M. Davis
Major Department: Plant Molecular and Cellular Biology

Chitinases are plant enzymes that hydrolyze chitin, which is the major

component of the cell walls of many pathogenic fungi but absent in higher plants.

Chitinases belong to Group III PR-proteins which are believed to play important

roles during plant-pathogen interactions. The present study describes the

structures of several genomic clones from pine trees that appear to encode

extracellular class II chitinase, and examines the expression of these genes in pine

cells as well as in transgenic tobacco plants. One of the genes, Pschi4, potentially

encodes a protein that shares 62% amino acid sequence identity through the

catalytic domain with class II chitinase from tobacco. The corresponding Pschi4

cDNA was cloned by RT-PCR. Nucleotide sequence analysis indicated that the

Pschi4 coding sequence is composed of three exons interrupted by two introns at

locations identical to those found in other chitinase genes that possess introns. In

contrast, Pschil contains a stop codon in the first exon and may be a pseudogene.

Pschi4 genes are conserved in several species of pine, and appear to comprise a

small multigene family. Treatment of pine cell suspension cultures with the general

elicitor chitosan induced Pschi4 expression at both mRNA and protein levels. The

regulatory sequences associated with the Pschi4 gene were sufficient to direct

chitosan- and wound-inducible expression of Pschi4 in transgenic tobacco plants,

which indicated that Pschi4 is an actively expressed member of the multigene

family. The region 5' to the putative transcription start site of Pschi4 was fused to

the GUS reporter gene to further analyze these inducible regulatory elements. The

-200 bp 5'-upstream sequence of Pschi4 was demonstrated to contain active

promoter sequences capable of wound-induced expression in transient assays

(particle bombardment) as well as in stably transformed tobacco plants. This region

was also sufficient to induce transcription in pollen of transgenic tobacco. The

putative pine promoter showed higher promoter activity than the widely used CaMV

35S in the transient expression in pine cells, which implies that the Pschi4 promoter

could be a good candidate to regulate transcription of other genes in transgenic

pine cells. The observation that the Pschi4 gene from pine (a gymnosperm) was

appropriately regulated by chitosan in tobacco (an angiosperm) suggests that the

signaling pathways that mediate chitosan-induced transcription are highly

conserved in the plant kingdom.


General Features of Plant Defense Responses

Plants are frequently challenged by pathogens that cause tissue damage and

disease. Resistance to these pathogenic organisms results from the rapid

deployment of a multicomponent defense response (Dixon et al., 1994). The

individual components of this response may include increased production of

antimicrobial phytoalexins (Dixon et al., 1983), hydrolytic enzymes (Stintzi et al.,

1993), lignin (Vance etal., 1980), hydroxyproline-rich glycoproteins (Vamer and Lin,

1989), and the development of a hypersensitive response (HR) around the infection

sites (Keen, 1990). Some of these components can also be induced in uninfected

tissue, at sites remote from pathogen infection. This physiologically acquired

resistance is termed systemic acquired resistance (SAR). This systemically induced

state can result in reduced severity of disease, detected when the tissues are

reinfected with any type of pathogen (bacterial, viral or fungal). Signals for

activation of host defense responses are thought to be initiated in recognition of

pathogen elicitors by the plant (Dixon et al., 1994).


The outcome (resistance or disease) of a plant-pathogen interaction depends

on both the plant and the pathogen. To be successful in infecting a host plant, the

pathogen must possess genes that function in pathogenicity. On the plant side,

some preformed antimicrobial compounds, termed phytoanticipin (Van Etten et a.,

1994), can serve as constitutive resistance factors (Osbourn, 1996). More

importantly, plants have the ability to develop an activated resistance, also called

an induced response. There are two relevant aspects in this induced process

(Alexander et a/., 1994). The first critical factor is the timing of the development of

the plant's active defense systems. If established quickly enough, these active

responses are usually effective in restricting the pathogen, resulting in disease

resistance. If initiated slowly or not at all, the pathogen may be successful in

infection, leading to disease. The second factor is the relevancy of the activated

responses for a particular pathogen. If the induced reactions have a deleterious

effect on the pathogen, the infection may be limited. If the elicited responses are

not relevant for the pathogen, or if the pathogen can use an alternative strategy to

escape the active responses, disease occurs.

Basic Incompatibility and Susceptibility

Basic incompatibility describes the failure of a fungus to cause disease on

any member of a plant species. This is a non-host, general resistance (Keen and

Dawson, 1992), and is proposed to result from the plant's ability to recognize the

general features of potential pathogens. The recognition of general elicitors, such


as chitin and chitosan, will be discussed later. In general, resistance is the rule and

susceptibility is the exception in the plant world (Staskawicz et al., 1995).

Basic susceptibility results in plant disease. Genetically, it is determined by

the pathogen's genes functioning in pathogenicity and host recognition (Keen and

Dawson, 1992). Some secondary metabolites from the pathogen act as host-

selective toxins (HSTs) which are low molecular weight compounds and are positive

agents of pathogenicity (Walton, 1996). Most known HSTs are made by fungi

(Walton, 1996). Successful infection by a fungal pathogen involves four steps: (1)

attachment; (2) germination of the fungal spores; (3) penetration and (4)

colonization of host tissues (Schafer, 1994). During the process, the pathogen may

detoxify host defense compounds such as phytoalexins and suppress defense

responses by modifying molecules in the signal transduction pathway (Keen and

Dawson, 1992).

Hypersensitive Response (HR)

Resistance in fungal-, bacterial- and viral-plant interactions is often

associated with the HR, in which a small number of cells that are at or near the site

of pathogen infection die rapidly. The protective cell suicide is considered as a very

strong defense response induced in plants by the pathogen itself (Stintzi et al.,

1993). The necrotic lesion which is formed around the infection site perhaps

depletes nutrients for the pathogen, and subsequently a very intense response is

induced in this region which confines the spread of pathogen (Lamb etal., 1989).


In many cases in which resistance occurs via an HR, the plant and the

pathogen have an apparent "gene-for-gene" relationship (Flor, 1942). In his classic

work, Flor defined the basic elements of gene-for-gene complementarity wherein

single plant disease resistance genes (R) are paired with single complementary

avirulence genes (Avr) in the pathogen resulting in the HR (Flor, 1942). The

functional alleles are proposed to be dominant and involved in recognition

(Ellingboe, 1981). As shown in Fig. 1, resistance occurs only when the plant

resistance gene (R) matches the pathogen avirulence gene (Avr). If either partner

lacks a functional dominant allele, recognition and resistance do not occur and the

plant becomes diseased (Fig. 1). However, a single plant may contain many

different resistance genes directed to a particular pathogenic species. Therefore,

a pathogen biotype must possess recessive alleles for all of the relevant avirulence

genes to successfully escape surveillance (Keen, 1990). A number of pathogen

avirulence genes were isolated in the 1980s (Keen, 1992). Some Avr genes are

proposed to be involved in the production of specific elicitors which will be

discussed later.

Since 1993, many plant resistance genes have also been cloned by

transposon tagging or positional mapping. For example, the tomato PTO gene

(Martin et al., 1993) and PRF gene (Salmeron et al., 1996), Arabidopsis RPS2 gene

(Mindrinos et al., 1994), and rice Xa21 gene (Song et al., 1995) were all isolated by

map-based cloning. Other resistance genes such as the tobacco N gene (Whitham

et al., 1994), tomato Cf-9 gene (Jones et al., 1994) and flax L6 gene (Lawrence et

Plant host cell

RR or Rr rr

(resistance) disease

disease disease

Figure 1. Gene-for-gene interactions specify plant disease resistance or
susceptibility. "R" denotes the dominant plant disease resistance (R) gene
and "A" indicates the corresponding pathogen avirulence (Avr) gene.
Resistance occurs only when both dominant alleles, R and Avr, are present in
plant and invading pathogen, respectively. All other combinations lead to
inability of recognition by plant host cells and result in disease.


al., 1995) were identified by transposon tagging. With the cloning of more Avr and

R genes, the basic tenet of specific recognition in the gene-for-gene hypothesis can

be directly tested. A consistent theme is that the R genes that have been cloned

appear to function in signal transduction pathways where they may rapidly activate

plant defense responses after pathogen recognition.

In the case of the hypersensitive response of tobacco to tobacco mosaic

virus (TMV), the areas of highly-induced responses can be easily detected under

UV light as rings, since cells surrounding the necrotic lesions exhibit bright blue

fluorescence. These cells have accumulated compounds of the phenylpropanoid

pathway (Legrand et al., 1976), some of which are fluorescent.


In association with the HR, plant cells produce many compounds that have

direct antimicrobial activity. One of these responses is the production of

phytoalexins. Phytoalexins are low-molecular-weight antimicrobial compounds

synthesized by plants in response to attempted infection by pathogens, exposure

to elicitor molecules, or other biotic and abiotic stresses (Dixon et al, 1983). More

than 350 phytoalexins have been chemically characterized from approximately 30

plant families (Kuc, 1995). Most of them have been isolated from dicots, but they

have also been isolated from monocots such as barley, corn, onion, rice, sorghum

and wheat (Kuc, 1995), and from pines (Lange et al., 1994). Phytoalexins are

isoflavonoid, terpenoid or other compounds of low molecular weight. Similarities


have been observed between phytoalexins from the same plant species, while

differences usually exist between phytoalexins from different genera and families

(Kuc and Rush, 1985). For example, isoflavonoid compounds are the major

phytoalexins in the Leguminosae and are rarely found in other plant species

(Dewick, 1988). On the other hand, terpenoid phytoalexins derived from the

isoprenoid pathway are the most abundant in Solanaceae such as tobacco, but

have not been reported in Leguminosae (Kuc, 1982a; Kuc, 1995).

Phytoalexins are synthesized de novo in response to infection as they are

usually not detected prior to infection (Kuc and Rush, 1985). The phytoalexin

precursors are produced from three major biosynthetic pathways in all plants:

shikimate, acetate-malonate and acetate-mevalonate pathways (Kuc and Rush,

1985; Kuc, 1995). Genes encoding individual enzymes in these pathways, such as

phenylalanine ammonia-lyase (PAL), chalcone synthase (CHS) and 3-hydroxyl-3-

methylglutaryl Coenzyme A reductase (HMGR), have been cloned from several

plant species, and their expression is regulated by environmental factors, including

pathogen infection (Choi et al., 1992; Fritze et al., 1991; Liang et al., 1989; OhI et

al., 1990; Stermer et al., 1990). The phytoalexins reported to date are not very

stable in plants and they are eventually degraded by the host plant and/or the

pathogen (Van Etten et al., 1982; Yoshikawa et a/., 1979). The detailed mechanism

by which phytoalexin biosynthesis and turnover is controlled remains somewhat



Cruickshank proposed earlier that in incompatible interactions,

accumulation of phytoalexins halts pathogen growth and thus confers resistance

(Cruickshank, 1963). In compatible interactions, the pathogen can either tolerate

the host phytoalexin, detoxify it, suppress its accumulation, or prevent the initial

elicitation (Cruickshank, 1963). Some studies have supported this proposal

(Kessmann and Barz, 1986; Kuc, 1995; Yamada et al., 1989).

Other Biochemical Responses

In addition to rapid cell death and production of phytoalexins at and in cells

surrounding infection sites, many other changes occur in the same cells (Lamb et

al., 1989). The most obvious observation is cell wall thickening and reinforcement

by deposition of various macromolecules such as callose, hydroxyproline-rich

glycoproteins (Varner and Lin, 1989), lignin (Lesney, 1989; Vance et al., 1980) and

cell wall bound phenolic compounds (Matern and Kneusel, 1988). These

compounds presumably serve as a physical barrier to prevent the pathogen from


Another important response at or near the infection site is the accumulation

of numerous pathogenesis-related (PR)-proteins. A few days after TMV infection,

PR-proteins may account for 10% of the total soluble proteins in tobacco leaves

(Jamet et al., 1985; Pierpoint, 1986). More details about PR-proteins will be

discussed later.


The above responses (HR, production of phytoalexins, cell wall lignification

and high concentrations of PR-proteins) are generally local responses and are very

effective in limiting pathogen growth.

Systemic Acquired Resistance (SAR)

Besides the local responses, many plants respond to the necrotizing

pathogen with a more thorough protection, the so-called systemic acquired

resistance (SAR). If part of a plant has already responded to an initial inoculation

hypersensitively, the uninoculated parts of this plant develop an increased state of

resistance evidenced by smaller lesions and greater restriction of the pathogen

upon subsequent infection by the same or even unrelated pathogens (McIntyre et

al., 1981; Ross, 1961; Ye et al., 1989). This type of plant immunity has been well

documented in tobacco (Ross, 1961) and cucumber (Kuc, 1982b). Systemic

acquired resistance can be detected a few days after inoculation and can last for

weeks to months (Lawton et al., 1993). Although the cellular intensity of the

systemic response is much lower than the local response, it still represents a

tremendous amplification in the plant defense response as it concerns the whole


A number of genes are associated with the appearance of SAR, which are

sometimes called SAR genes. Since PR-protein expression parallels the onset of

SAR (Bol and Van Kan, 1988), these genes are believed to be actively involved in

the development of SAR. PR genes have become sensitive markers in the search


for signals that are transmitted from necrotic lesions to distant parts. A more

detailed classification of PR-proteins will be discussed later.

Salicylic acid (SA) and SAR. Exogenously applied SA or acetylsalicylic acid

(aspirin) can induce SAR and the production of at least some PR-proteins in plants

(Ward et al., 1991; White, 1979). This was first discovered in Xanthi-nc tobacco

(Nicotiana tabacum) (White, 1979) which contains a dominant resistance gene (N)

that confers the host HR to specific races of tobacco mosaic virus (TMV). Injection

of solutions containing SA or aspirin into tobacco leaves prior to the inoculation of

TMV caused a dramatic reduction in lesion size (Wieringa-Brants and Schets, 1988)

and number (White, 1979). However, SA was found to be an endogenous signal

molecule only recently. The SA level increases transiently in the phloem of plants

just before the onset of SAR (Malamy et al., 1990; Metraux et al., 1990). The hybrid

of N. glutinosa X N. debneyi has been shown to contain a high constitutive level of

SA (in the absence of pathogens) (Yalpani et al., 1993b). This tobacco hybrid also

exhibits constitutive SAR and PR-protein expression, and is highly resistant to TMV

(Ahl Goy et al., 1992).

The role of SA in the SAR signaling pathway has been an active area of

research in the past few years. The direct evidence for SA involvement in SAR

induction comes from a transgenic experiment. Salicylate hydroxylase encoded by

a bacterial NahG gene catalyzes the hydroxylation of SA to catechol which does not

induce SAR. Transgenic tobacco plants harboring the NahG gene are unable to

accumulate SA and are defective in SAR induction upon TMV infection (Gaffney et


al., 1993). This apparent relationship supports the hypothesis that SA is involved

in the SAR signal transduction pathway. However, SA per se is probably not the

mobile signal from the infection site to other parts of the plant (Vemooij et al., 1994).

This is suggested by the observation that the removal of an infected leaf before any

measurable SA accumulation in the cucumber phloem still results in induction of

SAR genes in distant tissues (Rasmussen et al., 1991). This conclusion was

supported further by grafting experiments from the same group who did the NahG

transgenic experiment. When wild-type scion was grafted on top of NahG rootstock,

the top scion was able to develop SAR in response to TMV inoculation to the

(NahG) rootstock (Vernooij et al., 1994). Despite their inability to accumulate SA,

NahG tissues are fully capable of producing the transmittable mobile signals for

SAR. However, Raskin's group used "1 labelling and obtained evidence that SA

could be a mobile messenger in tobacco (Shulaev et al., 1995). They argued that

the SAR signaling mechanism in TMV-infected tobacco could be different from that

in Pseudomonas syringae-inoculated cucumber. Also, expression of salicylate

hydroxylase in NahG rootstock (Vernooij et al., 1994) does not completely block SA

accumulation. Small amounts of SA, above background levels, could escape and

be exported to phloem, and then transported to the upper wild-type scion. While no

consistent conclusion has been made about SAR signaling, the search for real

mobile signals) is still an active area of study.

The pathway of SA biosynthesis has been proposed in tobacco, in which SA

is derived from benzoic acid (Yalpani et al., 1993a). The last step is catalyzed by


benzoic acid-inducible benzoic acid 2-hydroxylase (BA2H). Activation of BA2H

leads to SA synthesis in tobacco. The level of benzoic acid increases dramatically

in TMV-inoculated tobacco tissue (Yalpani et a/., 1993a). Benzoic acid

accumulation has also been observed in pine injured by dothistromin, a toxin from

Dothistroma pini (Franich et a/., 1986).

To elucidate the mode of action of SA, Chen et al. identified an SA-binding

protein in tobacco (Chen and Klessig, 1991; Chen et al., 1993a) which turned out

to be a catalase (Chen et al., 1993b). SA may increase H202 levels by inhibiting

catalase activity which normally converts H202 to H20 and 02 (Chen et al., 1993b).

The active oxygen species (AOS), including H202, are thought to be important

molecules during SAR induction.

Active oxygen species (AOS). Active oxygen species (AOS) are generated

in pathogen-infected tissues and are widely involved in host defense responses

(Baker et al., 1993; Legendre et al., 1993; Orlandi et al., 1992). They are potentially

toxic intermediates to pathogens. The term oxidativee burst" describes the rapid

release of AOS during HR formation. This phenomenon has been observed in

many plant species such as potato (Doke, 1983), tomato (Vera-Estrella et al., 1992),

tobacco (Baker et al., 1993; Glazener et al., 1996; Keppler and Baker, 1989),

soybean (Apostol et a/., 1989; Levine et al., 1994) and spruce (Schwacke and

Hager, 1992). The predominant species detected in plant-pathogen interactions

include superoxide anion (02), hydrogen peroxide (H2 02), and hydroxyl radical

(OH-) (Mehdy, 1994). Injection of H202 into tobacco leaves induces the expression


of PR-1 genes locally, as is true with the treatment of leaves with glycolate and

paraquat, two substances that promote generation of H202 (Chen et al., 1993b).

Exogenously applied H202 (8 mM) induced hypersensitive cell death in soybean

suspension cells (Levine et a/., 1994; Levine et al., 1996). However, recent

evidence has shown that the relatively low amount of H202 (4-6 iM) generated

during the incompatible interactions between the plant and pathogen is not sufficient

to cause hypersensitive cell death (Glazener et al., 1996). More recently, Naton et

al. provided cytochemical evidence that intracellular accumulation of AOS in

infected parsley cells is related to rapid cell death (Naton et al., 1996). They argued

that intracellular AOS may be more important as possible mediators of rapid cell

death than extracellular AOS, which were measured previously by many

researchers including Glazener et al. (Glazener et al., 1996). Nevertheless, the

extracellular AOS may be involved in many processes associated with plant disease

resistance, e.g., crosslinking of cell wall proteins (Bradley et al., 1992), increased

lignification of cell walls at infection sites, or they may function as direct

antimicrobial compounds and as signal transducers leading to gene activation

(Sutherland, 1991).

The evidence suggests that H202 is involved in HR, that SA can inactivate

a catalase, and that SA is definitely involved in the signal transduction pathway

leading to SAR. However, recent experiments demonstrate that H202 is not a

second messenger downstream from SA and the inactivation of catalase by SA

does not result in SAR (Bi et al., 1995; Neuenschwander et a/., 1995). These data


suggest that H202 could be an important molecule involved in local responses, but

it is not the translocated signal and is not sufficient to induce SAR alone. The chain

of events in the SAR signal transduction pathway is still an active area of research.

Wounding Responses

Wounding damage to plant leaves results in activation of a variety of defense

genes locally as well as systemically (Brown and Ryan, 1984; Parsons et al., 1989;

Pena-Cortes et al., 1988). Considerable research has been conducted on induction

of PI genes, especially in tomato (Pearce et al., 1991; Ryan, 1990). These PI gene

products strongly inhibit the serine proteases found in insect gut tissue and their

expression is systemically induced by wounding. This suggests that Pis may serve

as a defense against insect feeding (Green and Ryan, 1972). Proteinase inhibitors

have also been found in tobacco and they are pathogen-inducible as well as wound-

inducible (Linthorst et al., 1993). In contrast to tomato, the tobacco PI genes can

only be induced locally (not systemically) by wounding or exposure to pathogens

(Linthorst et al., 1993). Some plant-derived chemicals such as oligosaccharides

and jasmonic acid have been found to be involved in the induction of these wound-

inducible PI genes (Farmer and Ryan, 1990; Farmer et al., 1992; Ryan, 1988;

Walker-Simmons, et al., 1983).

Jasmonic acid (JA). Jasmonic acid (JA) and its methyl ester (MJ) have been

hypothesized to be a possible key component of intracellular signaling in response

to wounding (Farmer and Ryan, 1990; Farmer et al., 1992). Direct evidence of


jasmonate involvement in wound-induction came from the observation that JA levels

increase immediately and transiently after wounding (Albrecht et al., 1993). The

biosynthetic pathway for JA has been elucidated (Vick and Zimmerman, 1984). The

pathway starts from linolenic acid (LA) which is a component of the plasma

membrane. LA is converted to 13-S-hydroperoxylinolenic acid (13-S-HPLA) by

lipoxygenase (LOX). Then hydration and cyclization lead to the formation of 12-oxy-

phytodienoic acid (12-oxy-PDA), followed by three steps of 3-oxidation. It has been

demonstrated that three of the octadecanoid precursors of JA, i.e., LA, 13-S-HPLA

and 12-oxy-PDA, can activate the synthesis of PIs in tomato leaves when applied

to leaf surfaces (Farmer and Ryan, 1992). Inhibitors of LOX activity, such as

salicylhydroxamic acid, reduce JA biosynthesis (Staswick et al., 1991). It is

interesting that salicylic acid (SA), a key molecule in systemic acquired resistance,

has been shown to be an inhibitor of wound-induced PI accumulation (Doares et a.,

1995a; Pena-Cortes et al., 1993). The results suggest that SA could inhibit JA

synthesis by inhibiting the formation of 12-oxy-PDA (Pena-Cortes et al., 1993) or SA

could block JA action by inhibiting an as-yet-undefined step between JA and

transcriptional activation of PI genes (Doares et al., 1995a).

The wound-induced accumulation of PIs in tomato is proposed to be initiated

by the release of pectic polysaccharides from the plant cell walls (Bishop et al.,

1981). It has also been found that chitosan, a polymer of p3-1,4-glucosamine found

in fungal and insect cell walls, is a strong inducer of both JA and PI syntheses in

tomato leaves (Doares et al., 1995b; Walker-Simmons et al., 1983).

Chitosan, General Elicitors and Specific Elicitors


In searching for signal molecules that may be involved in the ability of plants

to recognize pathogens, efforts have focused on elicitorr" molecules. The term

elicitorr" was originally used to describe agents that induce the synthesis and

accumulation of antimicrobial compounds (phytoalexins) in plant cells (Keen, 1975),

but is now widely used to refer to molecules that stimulate any plant defense

response, from cellular changes such as the HR to molecular changes such as

transcriptional activation of defense-responsive genes (Dixon et al., 1994).

Elicitors have been divided into two groups: specific elicitors and general

elicitors (non-specific elicitors), based on whether the elicitor exerts similar effects

on all members of a plant species (general) or only on specific genotypes of a plant

species (specific).

Specific Elicitors

Specific elicitors are those involved in the interaction of a certain pathogen

biotype to a particular plant species, and usually refer to the avirulence gene

products (direct or indirect) that are produced in the gene-for-gene interactions. A

well-characterized specific elicitor is a small peptide of 28 amino acids which is


produced only by Cladosporium fulvum carrying the Avr9 avirulence gene. This

elicitor induces a defense response only in tomato plants containing the Cf-9

resistance gene (de Wit, 1992). The biological function of these specific elicitors in

pathogens is not clear. Do they reside in the pathogen only to elicit the plant

defense system against itself when it invades the plant? This doesn't seem to be

logical. It is most likely that specific elicitors do have biological functions in the

pathogen, as yet unknown, but the plant developed a defense system later by

recognizing so-called specific elicitors from the pathogen in order to survive. On the

other hand, the pathogen does not intend to produce "elicitors" against itself and

keeps modifying/changing the "elicitor"-related avirulence genes to escape the plant

recognition. Experimental evidence for this dynamic interaction is the Avr4 and

Avr9 genes in the fungus Cladosporium fulvum (de Wit et al., 1994 and refs.

therein). C. fulvum has exploited at least two different mechanisms to avoid specific

recognition by the host plant (tomato). In the case of Avr9, mutations from

avirulence to virulence involve the deletion of this gene in order to escape

recognition by Cf-9. In the case of the Avr4 gene, virulent races still contain

essentially the same gene except for a single point mutation, resulting in an amino

acid change from cysteine to tyrosine, which is sufficient to avoid Cf-4 recognition.

This may explain why there are so many pairs of Avr-R genes in a single pathogen

biotype and a single plant species. It has been found that the coat protein of TMV

serves as a specific elicitor to tobacco plants carrying the N' resistance gene (Culver

and Dawson, 1989), and the TMV replicase is specifically recognized by tobacco


plants carrying the N gene (Whitham et aL., 1994). This supports the view that

specific elicitors do have biological functions in the pathogen and have been

exploited by the plant.

General Elicitors and Chitosan

General elicitors. Different from specific elicitors, general elicitors are

involved in the general resistance of plants to broad ranges of pathogens or

potential pathogens. Disease is the exception, not the rule (Staskawicz et al.,

1995). Most plants recognize general features of potential pathogens and can

prevent disease due to an ability to recognize these elicitors. Biotic elicitors include

some oligosaccharides (Hahn et al., 1993), proteins (Ricci et al., 1993) and lipids

(Bostock et al., 1981), and some known abiotic elicitors include heavy metals and

UV radiation. Abiotic elicitors are thought to result in the release of biotic elicitors

from the plant cell walls (Hahn et al., 1993). Biotic elicitors could be of either

pathogen or plant origin, although in most cases, they are pathogen components.

General elicitors are usually present on or near the cell surface of the pathogen and

often serve as cell wall components. Structurally characterized fungal elicitors

include: glucan elicitors, oligogalacturonide elicitors, chitin and chitosan,

glycopeptides, and ergosterol (Granado et al., 1995; Hahn et al., 1993; Keen and

Dawson, 1992).

Chitosan. Chitosan is a deacetylated form of chitin which is a polysaccharide

composed of P-1,4-linked N-acetylglucosamine. Both chitin and chitosan are cell


wall components of many fungi (Bartnicki-Garcia, 1968). Many studies have

demonstrated that oligosaccharides derived from chitin and chitosan elicit defense

responses in various plants. Chitosan and its derived fragments elicit the

accumulation of phytoalexins in pea pods (Hadwiger and Beckman, 1980; Walker-

Simmons et al., 1983), suspension-cultured soybean cells (Kohle et al., 1984) and

parsley cells (Conrath et al., 1989). Relatively low concentrations of chitosan (as

low as 8 [pg/ml) directly inhibit the growth of certain fungal pathogens of pea

(Kendra et al., 1989). Chitosan oligomers with a degree of polymerization (DP)

between 6 and 11 are most active as phytoalexin elicitors in pea. Chitosan-derived

oligosaccharides are also capable of inducing the accumulation of proteinase

inhibitors in both tomato and potato leaves (Pena-Cortes et al., 1988; Walker-

Simmons et al., 1983; Walker-Simmons and Ryan, 1984) and the synthesis of

callose in suspension-cultured soybean (Kohle et al., 1985), parsley (Conrath et al.,

1989), tomato (Grosskopf et al., 1991) and Catharanthus roseus (Kauss et al.,

1989) cells. Oligosaccharide fragments of both chitosan and chitin have both been

shown to induce defense-related lignification of the walls of suspension-cultured

slash pine cells (Lesney 1989, 1990). Chitosan treatment induces pine cells to

synthesize hydrolytic enzymes including chitinase and glucanase (Popp et al.,

1996). Chitin oligomers with a degree of polymerization of 4 to 6 elicited lignification

in wounded wheat leaves (Barber et al., 1989). The deposition of both callose and

lignin are thought to enhance plant defense by strengthening the plant cell walls.


Active chitosan and chitin oligomers are thought to be released during plant-

pathogen interactions by plant enzymes such as chitinases. Chitinases belong to

a group of PR proteins (see below) whose synthesis is induced by pathogen

infection and also by chitosan elicitors. It is possible that low levels of constitutive

plant chitinases release active chitin and chitosan oligomers from invading fungi,

which in turn induce a defense response including accumulation of chitinases. This

reaction could serve as an "amplification" of signals and re-enforce the initial

defense response.

Pathogenesis-Related Proteins

Pathogenesis-related (PR) proteins are highly induced proteins during plant-

pathogen interactions. Expressed both locally and systemically, PR proteins

represent the major quantitative changes in soluble proteins during defense

responses. PR proteins have very distinct physicochemical properties, some of

which enable them to survive the harsh environment where they occur: (1) at acidic

pH (as low as pH 2.8), they are quite stable and soluble, whereas most other plant

proteins are denatured at this pH; (2) they are relatively resistant to proteolytic

enzymes of both endogenous and exogenous origin; (3) they are targeted to

compartments such as the vacuole, the cell wall or apoplast; (4) most of them are

monomers with low molecular weight (8-50 kD).


PR proteins were first reported by Van Loon and Van Kammen in the early

1970s. They detected four de novo synthesized proteins in tobacco plants reacting

hypersensitively to infection with TMV (Van Loon and Van Kammen, 1970). Since

then, many other proteins with similar physicochemical (see above) and induction

properties have been identified from tobacco and other plant species (Kombrink et

al., 1988; Stintzi et al., 1993; Vogeli et al., 1988). The tobacco (Nicotiana tabacum

Samsun)-TMV interaction leading to an HR is still the model for the study of PR

proteins and from which the highest number of PR proteins have been

characterized. PR proteins are induced in response to infection by pathogens of

viral, viroid, bacterial or fungal origins (Van Loon, 1985). However, some members

of PR families have been found to be induced by chemical treatment (elicitors),

stress or wounding where pathogens are not involved. The reason they are still

called "pathogenesis-related" is that they still represent a wide array of pathogen

defense-related proteins in a particular plant species. Furthermore, many of the

chemicals that induce PR proteins mimic natural compounds involved in pathogen-

plant interactions and/or the transduction pathways that are associated with

pathogen-plant signaling (Stintzi et al., 1993).

Classification of PR Proteins

By using different biochemical approaches, more than 30 PR proteins have

been isolated from TMV-infected tobacco plants (Stintzi et al., 1993). Based on

amino acid sequence similarity and serological properties, these tobacco PRs are


classified into five groups (Stintzi et al., 1993; Alexander et al., 1994). PR proteins

from other plant species can be put into each group, although there are exceptions.

Group PR-1. Tobacco PR-1 proteins represent most of the earlier work.

Three acidic isoforms of tobacco PR-1, i.e., PR-la, PR-1b and PR-1c, were the first

purified PRs (Jamet and Fritig, 1986) and are serologically related to each other in

tobacco and to PRs from other plant species (Nassuth and Sanger, 1986). The

amino acid sequence of the three PRs, deduced from the cloned cDNA (Cutt et al.,

1988), share more than 90% identity. Another clone, which was identified from a

cDNA library of TMV-infected tobacco, appeared to encode a basic isoform of PR-1

protein (Comelissen et al., 1987). The corresponding protein was actually purified

later and named PR-1g. All of the four PR-1 proteins from tobacco are localized

extracellularly. Tobacco PR-la and PR-1b, as well as three members of PR-1 from

tomato, have been shown to contain direct antifungal activities in in vitro assays

(Niderman et al., 1993; Stintzi et al., 1993), although the molecular mechanism

involved needs further investigation.

Group PR-2. The tobacco PR-2 proteins have been found to contain endo-

1,3-p-glucanase activity (Kauffmann et al., 1987) which produces oligomers of 2-6

glucose units from 1,3-p-glucans. Many 1,3-p-glucanases have been purified

(Boiler, 1988), and many genes encoding glucanases have also been cloned (Meins

et al., 1992). Glucanases are usually monomers with a molecular weight of 25-35

kD. Of the five members of PR-2 from tobacco, four of them (PR protein 2, N, 0,

Q') are acidic proteins and are targeted to extracellular space (Kauffmann et al.,


1987); one (gluc b) is basic and is localized in the vacuole (Van den Bulcke et al.,

1989). The specific activities of these glucanases are strikingly different towards a

particular substrate (Stintzi et al., 1993).

Group PR-3. Tobacco PR-3 proteins are chitinases. Chitinase structure,

function and regulation will be discussed in more detail later.

Group PR-4. Tobacco PR-4 group includes four proteins: rl, r2, sl, s2

(Kauffmann et al., 1990) that are all acidic and targeted to the extracellular space

(apoplast). They are small proteins (13-14.5 kD) and are serologically related to

each other. The biological function/activity of PR-4 proteins is not known.

Group PR-5. Two slightly acidic proteins (R & S) and two basic proteins (n-

osmotin and osmotin) from tobacco are included in PR-5 group. The acidic and

basic isoforms are localized in apoplastic space extracellularr) and vacuolar

compartment, respectively (Dore et al., 1991). Genes encoding tobacco PR-5

proteins have been isolated (Van Kan et al., 1989) and their corresponding cDNAs

have also been cloned (Payne et al., 1988). Sequence comparisons have shown

that there are no introns within these genes. The deduced amino acid sequences

of PR-5 proteins are approximately 60% similar to thaumatin, a protein isolated from

Thaumatococcus danielli. Therefore, PR-5 proteins are also called thaumatin-like

PRs (Stintzi et al., 1993). Thaumatin-like proteins have been characterized in many

other plant species such as maize (Richardson et al., 1987), tomato (King et al.,

1988), barley (Bryngelsson and Green, 1989) and potato (Pierpoint et a/., 1990).


Other groups of PR Proteins. Some PR proteins from tobacco and other

plant species cannot be put into the five major groups. To accommodate these

proteins and novel proteins identified in the future, scientists of the Commission on

Plant Gene Nomenclature (CPGN) have proposed a new Y category which

generally represents plant genes whose sequences are clearly conserved but

whose designations are not based on function. Thus, PR genes can be collectively

designated by Yprfollowed by a number (Van Loon et al., 1994). However, for PR

genes whose functions are known such as glucanases and chitinases, suggested

designations Glu or Chi are preferred to Ypr2 or Ypr3.

Regulation of PR Protein Expression

PR gene expression is generally induced by pathogens (fungal, bacterial or

viral), as is obvious from the definition; however, some elicitors and various stresses

can also induce PR gene expression. PR genes are differentially expressed in

response to various stress conditions and during different developmental stages.

Spraying plants with a salicylic acid solution induces expression of some PR genes

including PR-1 (acidic and basic), acidic PR-2 and basic PR-3 (Bol et al., 1990; Van

de Rhee et al., 1993). The tobacco basic PR proteins are highly induced by

wounding and ethephon, which produces ethylene in vivo, whereas acidic PRs are

not (Brederode et al., 1991). Moreover, the basic PRs are constitutively expressed

in roots and lower leaves of healthy plants in contrast to their acidic counterparts

(Memelink et al., 1990; Neale eta/., 1990; Van de Rhee et al., 1993). On the other


hand, tobacco acidic PRs are systemically elicited by TMV infection, while no or little

induction of basic PRs occurs in non-inoculated leaves (Brederode et al., 1991).

To identify the putative cis-acting elements responsible for PR gene

induction, efforts have been focused on making promoter deletions fused with the

reporter gene (uidA) which encodes bacterial p-glucuronidase (GUS), and testing

the relative GUS activities. The data on PR-3 proteins (chitinases) will be discussed

later. The analysis of upstream sequences of genes encoding tobacco PR-1, PR-2

(glucanase) and PR-5 indicates that PR promoters contain multiple cis-acting

elements (Albrecht et al., 1992; Van de Rhee and Bol, 1993; Van de Rhee et al.,

1993). A recent experiment has shown that the PR-la gene promoter of tobacco

contains several elements that can bind GT-1-like nuclear factors (GT-1 factor is

necessary for light-responsive expression of the pea rbcS-3A gene) (Buchel et al.,

1996), but yet, no common sequence motif has been identified from these assays.

However, a 10-bp element which is repeated four times in the 5'-non-translated

region of a barley p-1,3-glucanase gene has been found to be present in the non-

translated regions of over 30 stress- and pathogen-inducible promoters

(Goldsbrough et al., 1993). Gel mobility shift assays have provided preliminary

evidence that this element, known as the TCA motif (TCATCTTCTT), specifically

binds a tobacco nuclear protein, and this binding activity was greatly increased

when tobacco plants were pre-treated with salicylic acid (Goldsbrough et al., 1993).

It is suggested that the TCA motif could be important for induced expression. A 116

bp fragment between -168 and -52 of the parsley PR-2 promoter, which was shown


to be necessary for elicitor-mediated expression (Van de Locht et al., 1990), actually

contains a TCA element.

A 61-bp element of the tobacco p-1,3-glucanase B gene has been shown to

be an enhancer whose activity is independent of orientation. Analysis of point

mutations has identified the sequence AGCCGCC, named the AGC box, which is

essential for the enhancer activity (Hart et al., 1993). Nuclear extracts from tobacco

leaves contain one or more factors that can interact with this element specifically.

This binding activity is higher in nuclear extracts from ethylene-treated plants than

control plants which correlate with its postulated role in the regulation of the 0-1,3-

glucanase gene (Hart et al., 1993). Whether the same or a similar set of regulatory

proteins are involved in all PR gene induction, or each group of acidic/basic PR

genes are controlled by a unique set of factors, is not clear. It seems that at least

some factors are commonly involved in induction of certain PR proteins, such as

chitinases and glucanases. The purification of these nuclear factors and cloning of

their corresponding genes/cDNAs will help to understand the induction mechanisms

and to give some clues as to the signal transduction pathways involved in the

induced responses.

Chitinase Structure, Function and Regulation

Chitinase Structure and Function

Chitinases (PR-3 proteins) are enzymes that hydrolyze chitin, a linear

homopolymer of N-acetylglucosamine. Chitin is the major component of cell walls

of many fungi, but is not present in higher plants. On the other hand, chitinases

have been reported to be present in a variety of higher plants (Boiler et al., 1983;

Broglie and Broglie, 1993). It seems that there is no endogenous substrate for

chitinases in higher plants; therefore, it has been proposed that chitinases in higher

plants may play a role in protecting plants against chitin-containing fungi (Boiler,


There are two types of chitinases in plants: endochitinases and

exochitinases. Endochitinases randomly hydrolyze internal 13-1,4-linkages of chitin,

whereas exochitinases digest chitin from the non-reducing end of the polymer (Fig.

2). Therefore, the smallest substrate for endochitinases is a tetramer of N-

acetylglucosamine and that for exochitinases is a dimer (Fig. 2). Most of the

characterized plant chitinases are endochitinases (EC (Boiler et al. 1983;

Molano et al., 1979); however, exochitinases have also been purified from melon

(Roby and Esquerre-Tugaye, 1987) and carrot (Kurosaki et al., 1987).

Endo-type chitinases have been characterized from many plant species

including barley (Jacobsen et al., 1989; Kragh et al., 1991), bean (Boiler et al.,

4+ +^J






N-acetylglucosamine residue

Figure 2. Different substrate specificities for endochitinases and
exochitinases. (A). Endochitinases hydrolyze the internal p-1,4-bond of
chitin, whereas exochitinases digest the bond from the non-reducing
end. (B). The smallest substrate for endochitinases and exochitinases
are tetramer and dimer of N-acetylglucosamine, respectively.



1983), cucumber (Boiler and Metraux, 1988), maize (Nasser et al., 1990), pea

(Mauch et al., 1988a), potato (Kombrink et al., 1988), tobacco (Legrand et al.,

1987), tomato (Joosten and de Wit, 1989) and wheat (Ride and Barber, 1990). In

general, plant chitinases are proteins of 25-35 kD molecular weight which occur as

monomers and have either a high or low isoelectric point (basic or acidic chitinases)

(Boiler, 1988). Three classes of plant chitinases have been proposed based on the

primary structures (Fig. 3) (Broglie and Broglie, 1993; Collinge et al., 1993; Shinshi

et al., 1990). Class I chitinases are composed of an N-terminal signal sequence,

a cysteine-rich domain of approximately 40 amino acids, a variable length hinge

region, a highly conserved main structure (catalytic domain) and a C-terminal

vacuolar targeting sequence (usually 7 amino acids). Class II chitinases have a

high amino acid sequence identity to the main structure of class I chitinases, but

lack the cysteine-rich domain, the hinge region and C-terminal domain. Class III

chitinases show no sequence similarity to enzymes in class I or class II. The C-

terminal seven amino acids (GLLVDTM) of tobacco class I chitinase are necessary

and sufficient to direct the protein to the vacuole (Neuhaus eta/., 1991). In addition

to the difference in domain structure, class I and class II chitinases have other

distinctive properties and are located in separate cell compartments. Class I

chitinases are targeted to the vacuole, whereas class II chitinases are secreted into

the extracellular space. All identified class II chitinases are acidic proteins, but class

I and class III chitinases have been found to be either basic or acidic (Collinge et

al., 1993; Davis et al., 1991; Lawton et al., 1992).


signal peptide


vacuolar targeting

t catalytic domain *

Class I

Class II

Class III

Figure 3. Domain structure of three classes of chitinases. From left to right, the
domains in class I chitinases are: signal peptide, cysteine-rich, hinge, catalytic,
and vacuolar targeting. Class II chitinases share high homology with class I
chitinases in the catalytic domain, but lack the cysteine-rich, hinge and vacuolar
targeting regions. The catalytic domain of class I chitinases also contain a short
stretch of amino acids not found in class II chitinases. The catalytic domain of
class III chitinases are not related to either class I or class II chitinases.


Genes and/or cDNAs encoding chitinases have been cloned from a variety

of plant species, such as Arabidopsis (Samac et al., 1990), bean (Broglie et al.,

1986), cucumber (Metraux et al., 1989), poplar (Davis et al., 1991), potato

(Laflamme and Roxby, 1989), rice (Xu et al., 1996; Zhu and Lamb, 1991) and

tobacco (Payne et al., 1990; Shinshi et al., 1990). Constitutive expression (directed

by the CaMV 35S promoter) of a bean chitinase gene in transgenic tobacco plants

showed increased disease resistance against certain fungi (Broglie et al., 1991).

It was speculated that chitinases may inhibit fungal growth by direct lysis of hyphal

tips, particularly in combination with glucanases (Schlumbaum et al., 1986; Mauch

et al., 1988b; Sela-Buurlage et al., 1993). Chitinases could also function to amplify

defense responses in cells surrounding a site of infection by liberating chitin and

chitosan oligomers from fungal cell walls which may serve as general elicitors to

induce the expression of several defense-related genes (Boiler et al., 1983; Mauch

and Staehelin, 1989). In addition to potential roles in defense, chitinases may play

important roles during early embryo development (de Jong et al., 1992) and other

developmental processes (Neale et al., 1990). This is presumably because plant

cells contain substrates for chitinase that are not chitin per se, but may resemble

chitin structurally (Fisher and Long, 1992; Collinge et al., 1993).

Regulation of Chitinase Gene Expression

Environmental regulation. Chitinase gene expression is regulated by many

factors including pathogen invasion, treatment with elicitors and plant hormones


(ethylene), and mechanical wounding. Chitinase enzyme activity, protein and

mRNA levels have been shown to be elevated when the plant is under these

stresses (Boiler, 1988; Collinge et al., 1993; Joosten and de Wit, 1989). Induced

expression of chitinase is often co-ordinated with other PR proteins such as p-1,3-

glucanases (Joosten and de Wit, 1989; Kombrink et al., 1988; Shinshi et al., 1987;

Vogeli et al., 1988). Individual chitinase isoforms have been reported to be

differentially regulated in barley, pea and tobacco. In barley leaves and grain,

several chitinase isozymes have been found, but only one is induced in response

to pathogen infection (Kragh et al., 1990). In pea, at least two chitinases are

differentially regulated upon fungal infection (Mauch et al., 1988a). In tobacco, the

basic class I and acidic class II chitinases have been shown to be differentially

induced by various stresses such as virus infection, UV light and wounding

(Brederode et al, 1991; Memelink et al., 1990).

Developmental regulation. In addition to stress-induced regulation,

chitinases are also under developmental regulation in healthy Arabidopsis, rice and

tobacco plants. In normal tobacco, chitinase expression was found in roots,

developing flowers and lower older leaves (Memelink et al., 1990; Neale et al.,

1990; Shinshi et al., 1987). In Arabidopsis and rice, constitutive expression of

chitinase was found in roots (Samac et al., 1990; Zhu and Lamb, 1991). The

developmental regulation of chitinase in tobacco was proposed to be controlled by

auxin and cytokinin gradients within the plant (Shinshi et al., 1987). The finding that

chitinases accumulate in a tissue-specific manner during different developmental


stages has led to the view that chitinases may serve other functions in plants in

addition to their role in the defense response.

Promoter studies. The regulatory sequences that direct the expression of

chitinase have been studied in a few plant species. The first characterized promoter

is the bean chitinase 5B promoter (CH5B) which contains elements) for ethylene

induction (Broglie et al., 1986). The ethylene-responsive element was first identified

between positions -422 and -44 (Broglie et al., 1989) by analysis of deleted

chitinase genes in transgenic tobacco plants and was confirmed in a bean

protoplast system (Roby et al., 1991). This region has been further narrowed down

to sequences between nucleotides -305 and -236 by promoter-GUS deletion

analysis in transient expression assays (Broglie and Broglie, 1993). In addition, a

nuclear protein has been observed to bind to this DNA sequence in gel mobility shift

and DNase I protection assays (Broglie and Broglie, 1993). Similar research has

been conducted for Arabidopsis and tobacco chitinase genes (Fukuda and Shinshi,

1994; Samac and Shah, 1991). An acidic class III Arabidopsis chitinase promoter

was fused to the GUS reporter gene and transformed into Arabidopsis. Promoter

activity (GUS expression) was detected in roots, leaf vascular tissue, hydathodes,

guard cells, and anthers in healthy plants, which indicates its developmental

regulation. Induced expression was observed in mesophyll cells surrounding

lesions caused by fungal infection. Promoter deletion analysis demonstrated that

the region 192 bp upstream of the transcription start site is capable of both

developmental and induced expression (Samac and Shah, 1991). In tobacco, the


5'-upstream sequence from tobacco class I chitinase was fused to GUS and

introduced into tobacco. Promoter deletion analysis revealed that the region

between nucleotides -574 and -476 is sufficient for inducibility by a fungal elicitor.

Gel mobility shift assays further identified a sequence of 22 bp between -539 to -

518 specifically that binds to a nuclear protein from elicitor-treated cells, but not

from control cells. This 22 bp sequence contains a direct repeat of GTCAG

separated by three nucleotides (Fukuda and Shinshi, 1994).

From the above discussions, it is clear that angiosperm PR genes have been

well documented in the literature. Their gene products and promoter analyses have

been studied in detail. However, little information is available on defense responses

in gymnosperms. In this research, I chose chitinase gene as a model to study pine

defense responses, since chitinase structure seems to be conserved in

angiosperms, and previous work has shown that chitinase activity in pine

suspension cells increases upon chitosan treatment (Popp et al., in press).


Plants are constantly confronted by microbes and other pests that can cause

tissue damage. The cell walls of many of these microbes share common structural

features. Compounds like chitin and chitosan are found in the cell walls of many

fungi, as well as the exoskeleton of arthropods, but these compounds are not found

in plants. In most plants, a defense response is induced when they are treated with

chitin or chitosan (Kohle et al., 1984, 1985; Fritensky et al., 1985; Kombrink and

Hahlbrock, 1986; Lesney, 1989). One component of the defense response to

elicitors is the transcriptional activation of genes that encode pathogenesis-related

(PR) proteins. One class of PR proteins includes chitinases. Chitinases hydrolyze

chitin, a linear homopolymer of N-acetylglucosamine. Most of the characterized

plant chitinases are endochitinases (EC which randomly cleave internal

B-1,4 linkages in chitin and consequently release oligomers of N-acetylglucosamine

(Boller et al., 1983; Molano et al., 1979). Chitinase alone, or in combination with

glucanase, can directly inhibit fungal growth by causing lysis of hyphal tips (Mauch

et al., 1988b; Schlumbaum et al., 1986; Sela-Buurlage et al., 1993). Chitinase also

seems to amplify plant defense responses by releasing chitin and chitosan elicitors

from fungal cell walls (Boiler et al., 1983; Mauch and Staehelin, 1989). In addition

to potential roles in defense, chitinase may also serve other functions during



development (de Jong et al., 1992; Neale et al., 1990). Chitinase gene expression

is controlled at the transcriptional level and can be induced by general elicitors and

other stresses. Therefore, chitinase genes can be viewed as "reporters" that can

be used to study defense mechanisms as well as to define components of signal

transduction pathways involved in defense responses.

Pinus is an ancient as well as economically important genus in the plant

kingdom. Pine cells exhibit a pronounced defense response to chitin, chitosan, and

live pathogens (Lesney, 1989; Popp, 1993), and this defense response is

accompanied by secretion of chitinase (Popp et al., In press). To obtain a better

understanding of defense responses and gene regulation in conifers, a chitinase

gene was chosen as a model system for this study. As a first step toward this

overall goal, a chitinase gene from pine trees was isolated and characterized.

Alignment of class I and class II chitinase sequences from a number of

different plant species revealed the presence of highly conserved regions within the

catalytic domain. This feature was exploited as a strategy for cloning related

sequences from pines using PCR. A pair of primers were designed to anneal to

conserved regions of the chitinase catalytic domain. The upstream primer, 5'-


fold degenerate and was expected to anneal to the nucleotide sequence translated

as QTSH(Q)ET. The downstream primer, 5'-ATGGTACCCATCCA(AG)AACCA(AC

GT)A(AGT)(ACGT)GC-3', was 96-fold degenerate and designed to anneal to the

region encoding AI(LM)WFWM. Degenerate positions are shown in parentheses,


and restriction sites that were introduced into the primers for directional cloning are

underlined. Fragments about 400 bp in length were amplified from white pine,

cloned, sequenced, and found to show sequence similarity to known chitinase

genes in the expected region of the catalytic domain. The cloned PCR product was

used to screen a genomic library of eastern white pine. Six plaques were selected

due to their hybridization to the probe in duplicate plaque lifts. The resulting

recombinant phage were designated gPschil-6 (for genomic Pinus strobus

chitinase I through 6). Two of these genomic clones, gPschil and gPschi4, were

selected for detailed sequence analysis. Recombinant phage DNA was digested

with Sacl and the insert fragments were subcloned into pBlueScript. Physical

mapping and partial DNA sequence analysis suggested that gPschil and gPschi4

were likely to contain lengthy 5' flanking sequences and intact coding regions (Fig.

4). Most of the above work was accomplished by Dr. John M. Davis, part was done

by Dr. Michael P. Popp (SFRC, University of Florida). DNA sequencing was

performed by Dr. Craig S. Echt (USDA-Forest Service, Rhinelander, WI).

The first goal of my study was to identify a functional gene from the cloned

chitinase fragments, and to study its inducibility of expression by general elicitors.

The hypothesis to be tested was that pines possess defense-related genes that are

regulated at the transcriptional level. To begin, I introduced the entire Pschi4 gene,

including 4.5 kb of 5' upstream sequence, 0.9 kb of coding sequence and 1.5 kb of

3' downstream sequence, into tobacco via Agrobacterium-mediated transformation.

This gene was confirmed to be transcribed in transgenic tobacco plants by northern



sequenced region


1 kb

0 intron

Figure 4. Overall structure of the pine genomic subclones gPschil and gPschi4.
The transcriptional orientation and putative coding regions, including intron and
exon sequences, are indicated by arrows. The asterisk denotes the stop codon
within exon #1 of gPschil. Restriction endonuclease cleavage sites that were
used for subcloning are presented. Pschil was not mapped using BamHI and
Xbal, only with Sacl, so lacking of these ssites in Pschil does not imply
polymorphism between Pschil and Pschi4. The 668 bp Sacl-BamHI fragment in
the gPschi4 coding region was used as hybridization probe for the blots shown in
figures 14 and 15.



blot analysis. Accumulation of its mRNA was induced by the general elicitor

chitosan in both pine suspension cells and transgenic tobacco leaves, as well as by

mechanical wounding in transgenic tobacco plants. This suggests that at least part

of the signaling mechanism by which chitosan induces gene expression is

conserved in gymnosperms and angiosperms. A better understanding of the

structure and regulation of these chitinase genes should provide useful insights into

the evolution of defense responses in plants.

My second goal was to identify a functional chitinase promoter and dissect

the promoter for its chitosan/wound-responsive elementss. It has been found that

most transcriptional control elements reside within the 5'-upstream region of the

coding sequence in higher plants. However, this seems not to be necessarily true

for pine genes (Loopstra et al., 1995). Reports have indicated that the regulatory

sequences of pine genes could lie in the coding region or 3'-untranslated portion of

the gene (Loopstra et al., 1995; Loopstra, personal communication). Since the

widely used CaMV 35S promoter and a number of other promoters from

angiosperms do not promote high levels of transcription in conifers (Ellis, 1994 and

refs. cited therein), efforts have been focused on identifying an inducible coniferous

promoter to study gene expression and regulation in conifers, and on the

development of strategies to introduce foreign genes into conifers. The hypothesis

to be tested was that the regulatory elements of a pine gene could be identified

using assays for gene expression in pine cells and in angiosperm cells. The cloned


pine chitinase gene, Pschi4, seemed to be a good candidate to test this hypothesis

and to identify a functional promoter region of an inducible gene from pine.

In association with these goals, the Pschi4 cDNA was cloned and expressed

in E.coli. Antibody was then made against the purified recombinant protein. Pschi4

protein expression was monitored in pine suspension cells as well as in transgenic

tobacco plants. Also in association with these goals, a series of Pschi4 promoter-

GUS fusion constructs were made and tested in transient assays as well as in

stably transformed tobacco plants. Collectively, this research represents the first

studies on gene expression, regulation and promoter dissection of a defense-

related gene in pine trees.


Plant Materials

Pine materials. Seeds obtained by self-pollination of eastern white pine

(Pinus strobus) genotype P-18 were a gift from Dr. Don Riemenschneider, USDA-

Forest Service, Rhinelander, WI. Seeds were surface-sterilized in a 20% solution

of commercial bleach for 10 min, rinsed in sterile distilled water, stratified in the

refrigerator for one month, and then sown in commercial potting mix. After the

cotyledons had expanded fully, the above-ground portion of the seedlings was used

as a source of DNA for Southern blots. Needle tissue from loblolly pine (P. taeda)

genotype 7-56 was a gift from Dr. Les Pearson (Westvaco Corp., Summerville, SC).

Cell cultures of loblolly pine were derived from an individual seedling from family 10-

38. Cell cultures of slash pine (P. elliottii var. elliottii) genotype 52-56 were initiated

and maintained according to previously described methods (Lesney, 1989). The

starting material for the cell cultures was provided by Greg Powell and Dr. Tim

White (Cooperative Forest Genetics Research Program, University of Florida).

Tobacco materials. Tobacco plants (Nicotiana tabacum var. Turkish and

Nicotiana tabacum var. Samsun; gifts from Dr. F. Zettler and Dr. C. Kao, University


of Florida, respectively) were grown from seed on agar-solidified MS medium

(Sigma) and maintained in aseptic cultures. These tobacco plants were used as an

explant source for transformation.

Others. White onions were purchased from local grocery stores. Maize

suspension cell line PC-5 was kindly provided by Dr. L. C. Hannah (University of

Florida). These onion and maize cells as well as pine cells were used for particle


Sequencing and Sequence Analysis

After the insert from the recombinant phage DNA was subcloned into the

Sacl site of pBluescript, sequencing reactions were carried out from both ends. The

identified putative coding region, along with -750 bp of upstream sequence and

-350 bp of downstream sequence, were also sequenced by using an automated

DNA sequencer (ABI 373) with dye terminator chemistry (Dr. Craig Echt, USDA-

Forest Service, Rhinelander, WI). DNA and protein sequence analysis was

performed using the BLAST search algorithm (Altschul et al., 1990) and GCG

sequence analysis software (Genetics Computer Group, Madison, WI).

Plant Transformation

Plasmid construction. Tobacco plants were transformed with the putative

chitinase gene located in gPschi4. The two Sacl fragments containing different

regions of Pschi4 were subcloned separately into pBlueScript, then ligated together

as BamHI-Sacl and Sacl-Xbal fragments, respectively, into pBlueScript. The

BamHI-Sacl fragment contained the putative promoter region and extended ~200

bp into the 5' end of the coding region, and the Sacl-Xbal fragment contained the

rest of the coding region and 3' flanking sequence (Fig. 4). The entire 7 kb insert

was excised from pBlueScript by digestion with Kpni and Xbal and subcloned into

the Agrobacterium binary vector pCIB10 (Rothstein et al., 1987). The resulting

plasmid, pWC4KX18.9, was introduced into Agrobacterium LBA4404 by the freeze-

thaw method (An et al., 1988).

Transformation of tobacco plants. Infection of tobacco (var. turkish) leaf

disks by Agrobacterium, cocultivation and subsequent regeneration were carried out

using standard methods (Rogers et al., 1986) with some modifications. A single

transformed Agrobacterium colony was picked from an LB-agar plate and

transferred to 3-5 ml of LB medium with appropriate antibiotics and incubated at

280C for 24 to 48 hr. Cells were harvested and resuspended in 3-5 ml of sterile MS

salts (Sigma M-5519 plus 3% sucrose). Two ml of the freshly suspended

Agrobacterium was added to 25 ml of MS salts containing tobacco leaf disks in a

sterile 50-ml tissue culture tube and incubated with shaking for one hr. Leaf disks


were blotted dry on autoclaved paper towels, placed on MS-agar plates for 12-24

hr (co-cultivation), then transferred to fresh CIM plates (Callus-Inducing-Medium:

MS-agar containing 50 pg/ml Timentin, 100 pg/ml Kanamycin, 5 pM BAP, 0.5 pM

NAA). Leaf disks were transferred to fresh CIM every 4-5 days for the first 2 weeks

and transferred weekly thereafter. After callus was well-formed, NAA was

eliminated to promote shoot regeneration. If shoots were generated, they were

transferred to MS agar without hormones for rooting. Sixteen independent

kanamycin resistant plants were regenerated, and nine were selected at random to

test the expression of the transgene in response to chitosan treatment.

Elicitor Treatment and RNA Extraction

Treatment of transgenic tobacco leaves. Chitosan (Sigma) was dissolved in

0.1 N HCI with heating and the pH adjusted to 5.0 with NaOH prior to autoclaving

(Popp, 1993). Young leaves were excised from transgenic tobacco plants grown

on MS agar in culture vessels (Magenta GA-7), and incubated in a culture dish

(Fisher Scientific) containing a solution of 50 mM KCI or 50 mM KCI plus 60 pg/ml

chitosan for 24 hr in constant light before harvesting. A plant that contained a single

copy of the pine transgene, based on 3:1 segregation of kanamycin resistance in

its progeny, was selected for further analysis and designated as Chi4 tobacco. A

single fully expanded leaf was divided into four sections. One section was placed

in a petri dish lacking chitosan (50 mM KCI), a second section was placed in a


culture dish containing chitosan (50 mM KCI + 60 pg/ml chitosan), and both were

harvested after 24 hr incubation in constant light. A third section was immediately

placed in liquid nitrogen, and the fourth remained attached to the plant and was

mechanically wounded around its margin. Wounding was performed at 0, 2, and

4 hr, and the leaf was collected at 7 hr. The same experiment was conducted on

the control tobacco plant containing pCIB10 vector alone (designated as CIB10


RNA isolation from tobacco leaves. Total RNA was extracted from all

samples using previously described methods for poplar leaves (Davis et al., 1991)

with minor modifications. Briefly, ground tissue was transferred to 1 volume

extraction buffer (100 mM Tris-pH 8.0, 500 mM NaCI, 20 mM EDTA, 0.5% SDS,

0.5% p-mercaptoethanol, 0.1% PVPP) (1 ml buffer per gram tissue) and 1 volume

buffered phenol:chloroform. After thoroughly mixing and incubation on ice, the

mixture was centrifuged. The aqueous phase was re-extracted once with

phenol:chloroform (1:1), and RNA was precipitated by adding 1/5 volume of cold 10

M LiCI and incubating on ice overnight (at least 12 hrs). The pellet was

resuspended in 400 pl of RNase-free water and re-extracted with phenol:chloroform

in a microfuge tube. The RNA in the aqueous phase was precipitated by addition

of ethanol.

Treatment of pine suspension cells. Pine cultures (both slash pine 52-56 and

loblolly pine 10-38) were maintained on a 7 day subculture interval. Two days after

transfer, sterile chitosan was added to the flasks to a final concentration of 60 jLg/ml.


After 24 hr of incubation, cells were harvested by suction filtration and stored in

liquid nitrogen. Adjacent flasks that received no chitosan were designated as


A time course study was performed on slash pine 52-56 suspension cells.

One day after normal transfer, eight flasks of suspension cells were combined and

then re-distributed into 8 flasks to generate a uniform population of cells. After

another day, sterile chitosan was added to the flasks to a final concentration of 60

gig/ml and incubated for various periods of time (0, 0.5, 1.5, 3, 5, 8, 12, 24 hr). At

each time point, one flask of cells was harvested by suction filtration and stored in

liquid nitrogen.

RNA isolation from pine cells. Total RNA was extracted from pine

suspension cells using previously described methods (Schneiderbauer et al., 1991)

with some modifications. Pine cells were ground to a fine powder in liquid nitrogen

in a mortar, and 20 ml of cold acetone was added directly to the mortar to inactivate

RNase activity and extract phenolic compounds which are abundant in pine cells.

The mixture was transferred to a 50-ml falcon tube and centrifuged. The pellet was

saved and washed again with cold acetone until the solution became clear, instead

of green. The pellet was then dissolved in 12 ml of TNE buffer (100 mM Tris-CI, 10

mM NaCI, 10 mM EDTA, pH 8.0, add 0.1% Triton X-100 and 15 mM DTT just before

use) and extracted with phenol/chloroform. The supernatant was precipitated by

adding 1/5 of 10 M LiCI and incubating overnight on ice. Following centrifugation,


RNA pellet was re-dissolved in water, extracted with phenol/chloroform twice, and

precipitated by addition of ethanol.

Primer Extension

A synthetic oligonucleotide primer (5'-GCCAATAGCAACCTCATCGACATC

ATTC-3') complementary to positions 793 to 820 (Fig. 5, underlined and italicized)

of Pschi4 was end-labeled by T4 polynucleotide kinase and y-32P-ATP. Poly-A' RNA

isolated from transgenic tobacco plants containing or lacking Pschi4 was hybridized

with the end-labeled primer, and extended with M-MLV reverse transcriptase. Equal

quantities of radioactivity were loaded in each lane, and the products were analyzed

on a 9% polyacrylamide sequencing gel. A sequencing reaction (Sanger et al.,

1977) of the genomic clone was performed using the same end-labeled primer and

run on the same gel.

Southern and Northern Analysis

Genomic DNAs from white pine P-18, loblolly pine 7-56 and slash pine 52-56

were digested with restriction enzymes, fractionated by agarose gel electrophoresis

and transferred to Hybond N+ nylon membrane (Amersham) in a vacuum blotter

(Hoefer) with 0.2 N NaOH as the transfer solution. The Southern membrane was

hybridized with a 729 bp Pschi4 cDNA fragment (see below for the cloning of Pschi4






1101 gcttcctctaactcttctgcctccctgccatgccttaaatgttattaatcggattaggatgtatgggtttttacagGCGGGTGGCCAACGGCCCCAGACG


1301 ccgcttcgatttctagcaatagatatggaaaaaatcgaatgaatttcaagcctaatacacttaccgctctgtgggagcagGGACTACAACTACAAAGCTG




V D V G S N L D Y K N Q K P Y G T *

Figure 5. Sites within Pschi4 that were used to design oligonucleotide primers for
this study. The primer used for primer extension is complementary to the underlined,
italicized region. The primer sites used for RT-PCR to clone the Pschi4 cDNA are
shown in underlined boldface. A primer complementary to the double-underlined
sequence was used for cloning of the 5'-upstream sequence of Pschi4.


cDNA). Total RNA samples were quantified spectrophotometrically, fractionated on

formaldehyde agarose gels (Sambrook et al., 1989) and vacuum-blotted to Hybond-

N+ membrane with 50 mM NaOH as the transfer solution. Equal loading was

confirmed by ethidium bromide staining of the RNA prior to blotting. The northern

blot was hybridized with a 668 bp Sacl-BamHI fragment that is part of the coding

region of Pschi4 (Fig. 1). Hybridization and high stringency (650C; 1 mM EDTA, 1%

SDS, 40 mM NaHPO4 buffer, pH 7.2) washing were performed using reagents and

conditions that were previously described (Church and Gilbert, 1984).

Cloning of the Pschi4 cDNA

RNA template isolation. Leaves of the Chi4 tobacco plant were incubated

with 50 mM KCI + 60 pjg/ml chitosan for 24 hr. Total RNA was isolated as

previously described (Davis et al., 1991). RNA concentration was determined by

both spectrophotometry and confirmed by ethidium bromide staining after

electrophoresis through an agarose gel.

Reverse transcription (RT)-PCR. One jig of total RNA from chitosan-treated

leaves of the Chi4 tobacco plant was reverse transcribed as follows. Total RNA

was heated to 650C for 10 min in the presence of 1 pg of oligo dT and 2 jil of 5 x

AMV RT buffer. The sample was then placed on ice. A mixture of 80 units RNasin

(RNase inhibitor; Promega), 2 il 10 mM dNTPs and 19 units AMV reverse


transcriptase was added to yield a final volume of 10 ld. After 40 min incubation

at 420C, the RT reaction was heated to 980C for 8 min to denature the cDNA-RNA

hybrids, briefly centrifuged and chilled on ice. This fresh RT-cDNA was used as

template in a PCR without further purification. Primers (Forward: 5'-TCTGCACAAC


corresponding to nucleotide 864-887 and 1760-1782 in Fig. 5 (underlined and in

boldface), were synthesized by Craig Echt (USDA Forest Service, Rhinelander,

WI). RT-cDNA template was mixed with a solution containing primers (1 ig each),

10 X PCR buffer, MgCI2, dNTPs to a final volume of 49 pil. The mixture was

denatured at 940C for 2 min, and quickly cooled on ice. After a brief spin, 1 jl of

Taq polymerase (5 units) was added. The mixture was overlaid with one drop of

mineral oil prior to initiation of cycling. The PCR reaction was carried out as follows

in a Coy themocycler (model 50): 940C, 20 sec; 55C, 20 sec; 720C, 1 min 20 sec,

30 cycles with an additional extension of 5 min at 720C before cooling down to 40C.

Cloning of RT-PCR product. The initial RT-PCR product was polished by

adding 10 u of T4 DNA polymerase directly and incubating at 370C for 2 hr. The

product was electrophoresed in a 0.8% TAE agarose gel (Sambrook et al., 1989).

The DNA fragment of the expected size was excised from the gel and purified by

QIAEX II (Qiagen Inc., CA) according to the manufacturer's instructions. The

purified DNA was ligated with Smal-linearized pUC19 and transformed into E. coli.


TB1 cells. Positive clones were identified by restriction enzyme digestion and

further confirmed by DNA sequencing.

cDNA Expression and Generation of Antibody

Pschi4 cDNA expression in E. coli. The cloned cDNA was then subcloned

into the BamHI-EcoRI sites of expression vector pET24d (Novagen, Madison, WI).

Expression was performed in bacterial host BL21 (DE3) pLysS cells according to

the manufacturer's protocol. Briefly, a single colony was inoculated in 3 ml of LB

containing kanamycin (50 jpg/ml) and chloramphenicol (34 jig/ml), and grown at

370C until the OD600oo reached 0.6 to 1.0. Cells were collected and resuspended in

fresh medium containing antibiotics, and grown at 370C until the ODwo reached 0.6.

At this point, IPTG was added to a final concentration of 1 mM to induce target gene

expression by further incubation for 3 hr. Cells were then harvested and

resuspended in buffer A (20 mM Tris-CI, pH 7.5, 20% sucrose, 1 mM EDTA). After

centrifugation, cells were stored at -800C overnight. The recombinant protein was

expressed as an inclusion body in E.coli. To purify inclusion bodies from other

cellular proteins, frozen cells were resuspended in PBS (1 L of PBS contains 8 g

NaCI, 0.2 g KCI, 1.44 g Na2HPO4, 0.24 g KH2PO4), and then sequentially incubated

with lysozyme and 1% Triton to lyse cells and DNase I to degrade DNA. After

centrifugation, the pellet was washed with PBS plus 1% Triton and then suspended

in water.


Protein purification. Overexpressed Pschi4 protein present in the inclusion

body was purified by using the electro-eluter method (Bio-Rad laboratories,

Richmond, CA). Briefly, the resuspended insoluble pellet was mixed with SDS

loading dye, boiled for 5 min and resolved by SDS-polyacrylamide gel

electrophoresis. After separation, the gel was stained with Coomassie blue in the

absence of acetic acid. The expected protein band was cut out and loaded onto the

electro-eluter (Bio-Rad, Model 422) according to the manufacturer's instructions.

The purity of protein was examined on a 12% SDS-PAGE gel and further confirmed

by sequencing the N-terminal 26 amino acids (ICBR protein core laboratory,

University of Florida).

Rabbit anti-Pschi4 antiserum. Antiserum against r-Pschi4 protein was

prepared by Cocalico Biologicals Inc. (Reamstown, PA). One hundred tg of purified

recombinant Pschi4 proteins were mixed with complete Freund's adjuvant to inject

each of two New Zealand white rabbits. Four booster injections (each time with 50

pig antigen and incomplete Freund's adjuvant) were administered at biweekly

intervals to obtain a high titer antiserum.

Protein Isolation and Western Blotting Analysis

Protein Isolation from Pine and Tobacco Suspension Cells

Pine suspension cultures were established and maintained as described

previously (Lesney, 1989). Tobacco suspension cultures were initiated from Chi4


and CIB10 tobacco leaves. Specifically, tobacco leaf disks (diameter ca. 4 mm)

were induced to form callus on MS-agar plates containing the hormone 2, 4-D prior

to initiation of suspension cultures.

Pine and transgenic tobacco suspension cells were maintained in LM (Verma

et al., 1982) and M1 (for 1 L: 4.4 g Sigma's M-5524, 30 g sucrose, 1 mg thiamine,

1 mg pyridoxine, 1 mg pantothenic acid, 0.01 mg biotin, 1 mg nicotinic acid, 1 mg

L-cysteine, 0.2 g L-glutamine, 100 mg inositol, 10 gIM NAA, 5 IM 2-iP) medium,

respectively. They were transferred to fresh media at seven day intervals. Two

days after transfer, chitosan was added to a 250-ml culture flask containing 50 ml

of medium to a final concentration of 60 jig/ml. After 24 hr of incubation, cells were

harvested by suction filtration. Both cells and supernatant were saved. Cells were

ground in 20 mM of NaOAc (pH 5.2) and centrifuged. Proteins were concentrated

by dialyzing against solid sucrose and the concentrations were determined by using

Sigma's bicinchoninic acid protein assay kit according to the manufacturer's

instructions. Adjacent flasks that received no chitosan were designated as controls.

In a separate dose-effect experiment, pine cells were combined one day after

normal transfer to form a uniform population, and 0.6 ml of cells was incubated in

1 ml, 2 ml or 8 ml of LM medium with chitosan at different concentrations (0, 20, 30,

40, 60, 90, 120, 180, 240 pg/ml) for 23 hr. Media, which contain extracellular

proteins, were then collected. Proteins were quantified and dot-blotted onto a

nitrocellulose membrane.

Protein Isolation from Pollen of Pine and Tobacco

Pollen from pine trees was collected in late January and early February.

Pollen from transgenic tobacco plants (Chi4 and CIB10 tobacco F1) was collected

fresh from flowers on plants. Pollen was ground with a micropestle in solubilizing

buffer (22.5% p-mercaptoethanol, 9% SDS, 22.5% glycerol, 0.125 M Tris-CI, pH

6.8) in microcentrifuge tubes, frozen in liquid nitrogen, boiled in 1000C water and

ground again. This cycle was repeated for at least three times to release proteins

from pollen grains. The mixture was centrifuged and the supernatant was saved.

Western Blotting Analysis

Equal amounts of protein were loaded onto a 12% SDS-PAGE gel and

electro-transferred to a nitrocellulose membrane by using a Genie electrophoretic

blotter (Idea Scientific Co., Minnesota) according to the manufacturers instructions.

The membranes were probed to anti-Pschi4 antibody by standard western

blot procedures. Specifically, the membrane was incubated in blocking reagent

(TBST + 5% dried milk; TBST = 20 mM Tris-CI, pH 7.5,150 mM NaCI, 0.5% Tween

20) for 30 min and then incubated with primary antibody (1:50,000) for 6-12 hr.

After three washes with TBST, the membrane was incubated with goat anti-rabbit

AP (alkaline phosphatase)-conjugated antibody (1:5000 dilution) for 30 min. The

membrane was washed three times with TBST and once with TBS (same as TBST

except TBS lacks Tween 20). Color was developed in 25 ml of solution containing


2.5 ml Tris-CI, pH 9.5, 500 pl of 5 M NaCI, 125 il of 1 M MgCI2, 165 pl NBT (100 mg

dissolved in 2 ml of 70% DMF) and 65 l BCIP (100 mg dissolved in 2 ml of 100%

DMF). Preimmune serum was used as a control (1:50,000 dilution).

Particle Bombardment and Transient Expression

Plasmid Construction

The 5'-flanking region of Pschi4 was amplified from pWC4Sac6, which

contained a 6 kb Sacl fragment of gPschi4 (Fig. 4 and Fig. 6), by PCR using a pair

of primers (Forward: Universal primer F; Reverse: JOD5 = 5'-GCCAACGCAAATTG

TGGTGATGATCCC-3' complementary to the positions 735 to 761 in Fig. 5, double-

underlined). After T4 DNA polymerase treatment (to make blunt ends), the PCR

fragment was digested with BamHI, purified and cloned into pBlueScript BamHI-

EcoRV sites. The plasmid was then digested with Clal, filled by Klenow Fragment

of E. coli DNA polymerase I, digested with Xbal, and subcloned into pBI101.1

(Clontech) or pGUS (Hindlll-EcoRI fragment of pBI101.1 subcloned in pUC19) Xbal-

Smal sites (Fig. 6). The resulting constructs were named pWP16.7 and pWP-

GUS9.3 (white pine promoter-GUS, the number denotes the total size of plasmid

DNA), respectively.

Plasmid DNA of pWP-GUS9.3 was digested with Pstf and Safl and treated

with exonuclease III at 370C for various periods of time (25, 50, 75, 100, 125 sec).

Uni F M OI


pWP16.7 -4.5 kbI pWP-GUS9.3

pWP13.4 .......................... -1.2 kb | pWP-GUS6.0

pWP13.0 ............................. -0.8 kb | pWP-GUS5.8
pWP12.8 .............................. -0.6 kb -- pWP-GUS5.6
-0.5 kb -- 7 j pWP-GUS5.5

-0.4 kb ---- pWP-GUS5.2

pWP12.4 ................................... -0.2 kb pWP-GUS5.0

pBI101 ...................................... O0kb I GUS pGUS

(cloned in PBI101)


Internal Control

Ei Q-E -s = ro" E _
.0 ? .5 w E

GUS nos ter

ubiquitin promoter LUC nos ter

Figure 6 Plasmid constructs. These constructs were used for particle
bombardment (labeled as pWP-GUS#) and stable transformation of tobacco
(labeled as pWP#). Constructs were named according to the total size (in kb)
of each plasmid DNA. A nopine synthase terminator was fused to the 3' end
of both GUS and LUC.

(cloned in pUC19)


JOD5 f 010


At each time point, reactions were stopped by adding an equal volume of 2 x exo

III stop buffer (0.3 N NaCI, 7.5 mM EDTA) and incubated at 700C. Following

treatment with Klenow Fragment, the DNAs were self-ligated and transformed into

E. coli strain TB1 cells. This created a nested series of Pschi4 putative promoter-

GUS constructs in pUC19 (Fig. 6) with sizes of 1.2, 0.8, 0.6, 0.5, 0.4 and 0.2 kb.

These constructs, including pWP-GUS9.3, were tested by particle bombardment in

transient expression assays. These constructs were subcloned into the pBI101.3

HindIll site, except pWP-GUS9.3 which was subcloned differently (see above

paragraph) for stable transformation of tobacco.

To normalize for transformation efficiency in the transient assay, a Ubiquitin-

Luciferase construct (kindly provided by Dr. C. Kao in Dr. Don McCarty's laboratory)

was included in each bombardment mixture. Thus, expression data are expressed

as GUS/Ubi-LUC ratios.

Tissue Preparation

White onions were purchased from local grocery stores. The inner epidermis

of an outer layer of onion were peeled using a pair of fine forceps and placed on the

center of MS-agar plates which were to be used for particle bombardment.

Slash pine suspension cultures were maintained as described previously

(Lesney, 1989) and transferred to fresh medium at 7 day intervals. Three days after

transfer, cells from several flasks were pooled to create an experimental population.

Approximately one ml of cells was placed on the filter paper on MS-agar plate and


dried for 3-10 min. This procedure varied from plate to plate and needed to be

adjusted empirically. If cells were too wet, gold particles did not penetrate the

aqueous layer, which reduced the transformation efficiency; if cells were too dry,

their viability was reduced.

Particle Bombardment

Particle bombardment was performed essentially as described previously

(Taylor and Vasil, 1991) using a DuPont PDS-100 particle gun. Briefly, 37 pl of a

40 mg/ml gold stock solution was mixed with 5 pg of internal control plasmid DNA

(Ubi-LUC) and 5 4g of testing DNA (WP-GUS) in a total volume of 72 p1 in a 1.5-ml

microcentrifuge tube and vortexed briefly. Twenty [pl of 100 mM of free base

spermidine and 50 pl of 2.5 M CaCI2 were placed in separate drops on the side of

the tube to avoid pre-mixing of either solution with the gold/DNA solution. The tube

was then mixed immediately by vortexing for 20 sec, which allowed plasmid DNAs

to attach to the gold particle. The tube was centrifuged for 5 sec and supernatant

was removed. Two hundred (200) pl of 100% ethanol was added and sonicated

briefly. After centrifugation and removal of supernatant, 60 pl of 100% ethanol was

added. The tube was placed on ice until all samples were ready for bombardment.

Four pl of gold/DNA solution (sonicated again just before use) was used for each

individual bombardment shot. Each experiment represented 3-6 replicates.

Incubation and Extraction of Proteins

Following bombardment, petri dish plates were incubated at room

temperature in constant light for 24 hr. Bombarded onion or pine cells were ground

with mortar and pestle aided by addition of glass beads in 200-800 pl of GUS/LUC

extraction buffer (0.1 M potassium phosphate (pH 7.8), 2 mM EDTA (pH 8.0), 2 mM

DTT, 5% glycerol). The homogenates were centrifuged and supernatants were

transferred to clean tubes.

Quantification of Transient Expression

Quantitative measurement of GUS activities was performed essentially as

described by Jefferson et al. (1987), except that the substrate MUG was dissolved

in the extraction buffer described above. For the luciferase assay, an automated

luminometer (AutoLumat model #LB953, Wallac Inc., MA) and Promega's

Luciferase Assay Kit were used. Briefly, 10 pl aliquot of each extract was placed

in each disposable culture tube (Fisher brand, cat.# 14-961-26) and all tubes were

loaded into the luminometer. The instrument automatically injects 100 pl of

substrate luciferin dissolved in its assay buffer (prepared according to Promega's

instruction manual), counts the emitted photons for 60 sec and moves to the next

sample. The unit of measurement is the Relative Light Unit (RLU).

Analysis of Promoter Deletion Constructs in Stably Transformed Tobacco

Deletion constructs of the Pschi4 promoter were made and subcloned into

binary vector pBI101 as described above. These constructs were introduced into

tobacco (var. Samsun) using Agrobacterium-mediated transformation as described

above. Primary transgenic plants were initially used for chitosan and wounding

assays. In order to minimize the large variation in transgene expression and

inducibility, plants were allowed to flower and set seed. The seeds were placed on

MS-agar plates containing kanamycin (50 ig/ml) to determine transgene copy

number, and to generate a uniform population of plants for use in GUS assays.

Chitosan and wounding treatments were described previously (see above).

Histochemical Assays in Transgenic Tobacco

Histochemical GUS assays were performed essentially as described

previously (Jefferson et al., 1987). Tissue sections were floated in X-gluc solution

(0.5 mg/ml X-gluc in 50 mM of NaHPO4 pH 7.0) for 16 to 24 hr at 370C. Tissues

were then fixed in 5% formaldehyde, 5% acetic acid, 20% ethanol and washed in

80% ethanol. Tissue sections from transgenic tobacco containing the full-length

promoter-GUS (pWP16.7) were analyzed first. Untreated leaves, stems, and

reproductive organs (corolla, stigma, ovary, pollen) from pWP16.7 tobacco plants


were used. For other transgenic tobacco plants containing shorter promoter-GUS

constructs, only pollen grains were stained in X-gluc and examined microscopically.


Pschi4 Gene Structure

Nucleotide sequence analysis of Pschi4 revealed a complete coding

sequence. In contrast, Pschil contained a premature stop codon in the first exon

due to a T residue at nucleotide 1071 (Fig. 7). Pschi4 and Pschil are 90% identical

through the putative coding region (not including introns) and 83% identical through

the 5'-flanking sequence. In Pschi4, a putative TATA box is located at nucleotide

711 (Fig. 7). Primer extension analysis using RNA from transgenic tobacco

revealed two major bands in cells containing Pschi4 transcripts that were not

detected in cells lacking Pschi4 transcripts (Fig. 8). The nucleotide G at position

719 (Fig. 7) was considered to be the major putative transcription start site because

it was a longer and more abundant product. Primer extension products were not

seen when slash pine mRNA was used as template (data not shown), probably due

to poor annealing of the primer which was designed from the white pine (P. strobus)


Pschi4 contains several possible translation initiation sites. The ATG at

nucleotide 771 (Fig. 7) was considered to be the likely translation start site because

it is the first ATG downstream of the putative start of transcription, and it is flanked



-140 -131





1101 gcttcctctaactcttctgcctccctgccatgccttaaatgttattaatcggattaggatgtatgggtttttacagGCGGGTGGCCAACGGCCCCAGACG


1301 ccgcttcgatttctagcaatagatatggaaaaaatcgaatgaatttcaagcctaatacacttaccgctctgtgggagcagGGACTACAACTACAAAGCTG




V D V G S N L D Y K N Q K P Y G T *




Figure 7. Partial nucleotide sequence and translation product encoded by the
genomic clone containing Pschi4. The putative TATA box is double underlined, as
is the putative transcription start site. The likely signal peptidase cleavage site is
indicated (A). The annealing sites for the degenerate oligonucleotide primers are
shown in italicized boldface, and the position of the T in Pschil that results in a stop
codon is overlined. Introns are shown in lower cases. A potential glycosylation site
is shaded. Potential polyadenylation signals are underlined. The TCA-like cis-
element is underlined and in boldface. The complete sequence of Pschil and
Pschi4 that were deposited in the Genbank database are shown in the appendix.



A -
A O+
A 0 0
A 0t "
c CTAG > >


G ___


Figure 8. Primer extension analysis to reveal the putative transcription
start site(s). A primer complementary to positions 793 to 820 (Fig. 5)
was end-labeled, hybridized to poly-A RNA from transgenic tobacco
containing Pschi4 or pCIB10 vector alone, and extended by M-MLV
reverse transcriptase. A sequencing reaction using the same end-
labeled primer was run by side on a 9% polyacrylamide gel. The major
products are indicated by arrows and the nucleotides at which
transcription is initiated are marked with an asterisk.


by sequences that are the most similar to the consensus sequence for translation

initiation in eukaryotes (Kozak, 1991). In the 3' flanking region there are seven

potential polyadenylation signals (AATAAA).

A sequence was found between -140 and -131 (Fig. 7, underlined boldface)

that is similar in composition (TTACCTTCTA, identities underlined) and relative

location to a cis element implicated in wound and elicitor responsiveness of

proteinase inhibitor genes (GTACCTTGCC; Palm et al., 1990; Balandin et al.,

1995). This sequence (TTACCTTCTA, identities underlined) is also similar to the

10-bp TCA motif present in more than 30 pathogen-inducible promoters (TCA motif

consensus = TCATCTTCTT; Goldsbrough et al., 1993).

The chitinase coding region is divided into three exons of 313, 109 and 373

bp by two small introns of 93 and 96 bp (Fig. 1 and 7). The presence of introns was

first inferred by AT-richness (64.5% and 61.5% AT, respectively), orthodox

sequences at the putative splicing sites (GIGT ... AGIG), and conserved location in

other plant chitinase genes that possess introns. This was then confirmed by

cloning of the corresponding cDNA (see below). The deduced protein contains an

N-terminal 33 amino acids having the structure expected for a signal peptide, with

several positively charged amino acids near the N-terminus and an internal

hydrophobic region (Chrispeels, 1991). The predicted mature protein has a

molecular mass of 25.3 kD, assuming removal of the putative signal peptide. The

Pschi4 protein is expected to be acidic, with a predicted isoelectric point of 6.1.


The amino acid sequence of translated Pschi4 was aligned with tobacco

class I (Shinshi et al., 1990) and class II chitinases (Payne et al., 1990). The

catalytic domain of translated Pschi4 shares 64% and 62% amino acid sequence

identity with the catalytic domains of tobacco class I and II chitinases, respectively

(Fig. 9 and 10). Translated Pschi4 lacked the C-terminal sequence (GLLVDTM)

(Fig. 10) which is sufficient to target tobacco class I chitinase to the vacuole

(Neuhaus et al., 1991).

Genomic Southern blots were used to investigate the number of Pschi4

genes in pine and to assess their presence in different species of pines (Fig. 11).

All three species showed from two to four restriction fragments that accounted for

most of the hybridization to the Pschi4 cDNA probe. The slash and loblolly pine

patterns were more similar to one another than either was to white pine. A 2.5 kb

Sacl fragment, and a 5.5 kb BamHI fragment were predicted to be present in white

pine (Fig. 1), and were in fact observed (Fig. 11).

Pschi4 cDNA Cloning and Expression in Bacteria

RT-PCR technique was used to clone the Pschi4 cDNA. A 729-bp fragment

was predicted based on the sequence analysis (Fig. 5 between the two primers -

underlined boldface), assuming the two putative introns were spliced; otherwise, it

would be 917 bp. The 729-bp fragment was in fact amplified by RT-PCR from the

RNA isolated from Chi4 tobacco (Fig. 12). As a control, plasmid DNA containing the

1% S%




100 100


62 79

64 81

Figure 9. Domain structure of the putative Pschi4 protein from pine with class I
and II chitinase from tobacco. Class I chitinase from tobacco is vacuolar
(TobV; Shinshi et al., 1990), whereas class II chitinase from tobacco is
extracellular (TobE; Payne et al., 1990). Percent identity (1%) and similarity
(S%) values were calculated by comparing the amino acid sequence in the
catalytic domains (see Fig. 10 for detailed sequence alignment). From the left,
the domains in TobV are: signal peptide, cysteine-rich, hinge, catalytic, and
vacuolar targeting. Pschi4 and TobE lack the cysteine-rich, hinge, and
vacuolar targeting regions. The catalytic domain of TobV also contains a short
stretch of amino acids not found in Pschi4 or TobE.

0 00K MMI T-i



...... MEFS

..Q .................













GS ........



TD ........


VI-N----- I
VI-N----- L


....... SLC


---- I ---- A

....... RYR
....... .--Y




-M------ N-



DN--- S ----


Figure 10. Sequence alignment of Pschi4 with tobacco chitinases. The deduced
amino acid sequence of Pschi4 was aligned with tobacco class I and class II
chitinase sequences by using the GAP program of the GCG package. Identical
amino acids are indicated by hyphens. Dots represent gaps introduced to optimize
sequence alignment.

-P-N----- D





Sac I

BamH I



8.0 -
6.0 -




3.0 -


Figure 11. Genomic Sothern blot analysis of DNA from three pine species.
Fifteen pig of genomic DNA from white pine (WP), loblolly pine (LP) and slash
pine (SP) were digested with the restriction enzymes Sacl or BamHI,
fractionated, blotted, hybridized with 32P-labeled Pschi4 cDNA probe, and
washed at high stringency. The predicted 2.5 kb Sacl and 5.5 kb BamHI
fragments from white pine are indicated by arrows.

917 bp

729 bp

Figure 12. Cloning of Pschi4 cDNA by RT-PCR. One pig of oligo dT was
annealed to 1 pg of total RNA isolated from chitosan-treated Chi4 tobacco. The
first strand cDNA was synthesized by AMV reverse transcriptase, which was
used as the template for subsequent PCR, with a pair of primers spanning the
putative coding region (Fig. 5 underlined boldface). A DNA template of the
Pschi4 genomic clone was also used for PCR as a control.


Pschi4 genomic subclone was used as template for conventional PCR with the

same pair of primers. As expected, a 917-bp fragment was amplified (Fig. 12). The

amplified 729-bp fragment was cloned and its nucleotide sequence was determined.

Sequence analysis showed that the cloned fragment was identical to the original

genomic subclone except that the predicted two introns were correctly spliced (see

Fig. 5). Therefore, the 729-bp fragment was considered to be Pschi4 cDNA.

The cDNA was subcloned into expression vector pET24d, and

overexpressed in E.coli. As shown in Fig. 13A, more than 90% of the bacterial

protein was the target protein in the presence of IPTG, while this protein band was

weak or not detected in the absence of IPTG (Fig. 13A, control lane 1 and 2). The

overexpressed protein was further purified to higher homogeneity (Fig. 13B). Based

on the predicted amino acid sequence, the recombinant Pschi4 (r-Pschi4) has a

molecular weight of 26.9 kD, and this was actually observed in the SDS-PAGE gel

(Fig. 13 A&B). The N-terminal 26 amino acids of the purified protein were

determined by the protein sequencing core laboratory (ICBR, University of Florida)

to confirm that the protein was Pschi4. The result is shown in Fig. 13C and it

proved to be correct. Thus, the purified protein was used to generate anti-Pschi4


A. kD E 1 2 3 4 5

310 i0



31.0 4- purified r-Pschi4



Figure 13. Pschi4 cDNA expression in bacteria. (A). Crude extract of inclusion
body from E. coli. Lane 1. control #1: bacteria were taken out of shaker before
IPTG addition. Lane 2. control #2: bacteria were left in the shaker without IPTG
and grown for 3 h as lane 3-5. Lane 3-5: IPTG was added to these three flasks
and incubated for additional 3 h. (B). Purified recombinant Pschi4 protein from
bacteria. One, 4 or 8 p.l of proteins were loaded onto a 12% SDS-PAGE gel and
stained with Commassie blue. (C). The N-terminal sequence of r-Pschi4. The
underlined amino acids were from Pschi4 (see Fig. 5) and others were from the
vector sequence. The sequence was determined by the ICBR protein
sequencing core lab at University of Florida.

Pschi4 Expression

Chitosan-Induced Expression of Pschi4 at mRNA Level

In the absence of chitosan, Pschi4 transcripts were present at low or

undetectable levels in pine cells. In both slash and loblolly pine suspension cells,

transcripts related to Pschi4 accumulated after treatment with chitosan (Fig. 14).

A time course study was conducted in slash pine suspension cells. Pschi4-related

transcripts increased to detectable levels 3 h after chitosan treatment and remained

elevated up to 24 hr (Fig. 14B and data not shown). In parallel with transcript

accumulation, cell browning was observed after 2.5 hr, which is an indication of

lignification of cell walls (Lesney, 1989). Transcript induction was not observed at

0, 3, or 24 hr in the absence of chitosan (data not shown). The observed transcript

size is approximately 0.9 kb, which is consistent with the prediction from sequence

analysis (Fig. 7).

Pschi4 was introduced into tobacco under the control of its own regulatory

sequences by Agrobacterium-mediated transformation. Sixteen independent

transformants, designated as Pschi4.1 through Pschi4.16, were generated based

on kanamycin selection. Nine transgenic individuals were randomly selected to test

transgene expression. Chitosan-induced expression of Pschi4 was observed in

seven of these transgenic plants (Table 1). Fig. 15 shows a typical result of an

experiment in which the steady state level of mRNA increased in response to




+ chitosan

0 0.5 1.5 3 5 8 12 hr


Figure 14. Transcript accumulation in chitosan-treated pine cells. (A).
Loblolly pine 10-38 suspension cells were treated with or without chitosan for
24 h, and (B). Slash pine 52-56 suspension cells were treated with chitosan
for different time points as indicated before total RNA was isolated. Equal
amounts of RNA were subjected to denaturing gel electrophoresis, vacuum-
blotted to a nylon membrane, hybridized with the 668 bp Sacl-BamHI
fragment of Pschi4 (Fig. 4) and washed at high stringency. The bottom panel
shows the ethidium bromide-stained gel prior to blotting.

Table 1. Chitosan-induced mRNA accumulation in transgenic tobacco plants.

Nine individual transformants of Pschi4 were randomly selected for testing chitosan-
induced expression by northern blot analysis.

Transgenic CIB 4.2 4.6 4.7 4.8 4.9 4.10 4.11 4.12 4.14
lines 10
chitosan- no yes not yes yes not yes yes yes yes
induction sure sure

no no induction (actually no signal was detected in CIB10 tobacco plants)

yes chitosan-induction was observed

not sure inducibility is not sure because of unequal loading of RNA

* One transgenic line (4.10), specifically designated as Chi4 tobacco, was used
as a source for various studies (Fig. 8, Fig. 12, Fig. 15, Fig. 18 and Fig. 28).


vector only


Figure 15. Northern blot showing expression of Pschi4 in a transgenic tobacco
plant. A single leaf from tobacco transformed with pCIB10 alone (vector only)
or tobacco containing a genomic subclone with Pschi4 (vector + Pschi4) was
divided into four sections, and harvested immediately (untreated), or incubated
in 50 mM KCI (KCI), or 50 mM KCI plus 60 tg/ml chitosan (chitosan), or left on
the plant and mechanically wounded (wounded). Approximately 10 pg of total
RNA was loaded in each lane. Ethidium bromide staining indicated the last
lane (wounded) was underloaded.





chitosan treatment and mechanical wounding. This single tobacco plant (4.10

tobacco in Table 1) was used as sources for several experiments in the present

study (such as primer extension, cDNA cloning, initiation of suspension culture and

expression in pollen) and specifically designated as Chi4 tobacco. The pine

chitinase probe did not hybridize to any mRNA transcripts in tobacco transformed

with the pCIB10 vector alone (CIB10 tobacco). The probe appeared to be specific

for the pine chitinase transcripts and did not hybridize to endogenous tobacco

transcripts, since there is only 55% nucleotide sequence identity between Pschi4

and tobacco class II chitinase mRNA (Payne et al., 1990).

Pschi4 Protein Expression

Pschi4 protein expression in pine suspension cells. The expression of

Pschi4 protein in pine suspension cells was examined by western blot analysis.

Two cell lines of pine, slash pine genotype 52-56 and loblolly pine genotype 10-38,

were used. Sequence analysis predicted that Pschi4 would encode an extracellular

protein with an apparent molecular weight of 25.3 kD, assuming removal of the N-

terminal signal sequences. Fig. 16 shows that the anti-Pschi4 antibody could detect

Pschi4-related proteins in the supernatant (containing extracellular proteins) of both

slash and loblolly pine cells with the size slightly higher probably due to

glycosylation, since a potential glycosylation site (NST) was found in the sequence

(Fig. 7, shaded). Interestingly, this antibody could also recognize a protein

approximately of 32 kD in the cellular fraction of loblolly pine 10-38 cell line, but not

slash pine

0 0 0 0



loblolly pine

0 0 0 0

- 97.4
- 66.2


- 31.0

- 21.5

- 97.4
- 66.2

- 42.7

- 31.0

- 21.5

Figure 16. Pschi4 protein expression in pine suspension cells. Slash pine
52-56 and loblolly pine 10-38 suspension cells were treated with (ch) or
without (ck) chitosan. Total proteins were then isolated from media
supernatantt spnt) or cells (cytosol cyt), fractionated on a 12% SDS-PAGE
gel, and transferred to a nitrocellulose membrane. Duplicate membranes
were incubated with anti-Pschi4 (UF81) or preimmune (PI). The
overexpressed bacterial protein (not purified) was used as a positive control.


in slash pine cells (Fig. 16). This size is consistent with other vacuolar chitinases

reported (Broglie et al., 1986; Samac et al., 1990; Shinshi et al., 1987; Zhu and

Lamb, 1991). It is also of interest to note that chitosan did not show much induction

at the protein level in this blot.

However, slash pine cells did show a chitosan response at the protein level

in a separate study (Fig. 17). Equal amounts of suspension cells (0.6 ml) were

incubated in different volume of medium with various concentrations of chitosan.

Within the 1-ml incubation, Pschi4-related protein was low or not detected at

chitosan concentrations of 0, 180 and 240 ipg/ml; but accumulated to detectable

levels starting from 20 up to 120 pg/ml chitosan (Fig. 17). Cell browning, from light

to dark, was correlated with the chitosan concentrations from 20 to 240 gg/ml. In

fact, when chitosan was higher than 180 pg/ml in the small volume (1 ml) of cells,

cell cultures became very dark-brown, a phenomenon similar to the hypersensitive

cell death observed in soybean suspension cells exposed to high concentrations of

H202 (Levine et al., 1994; Levine et al., 1996). In the 8-ml incubation, cells became

dark-brown at the concentration of 60 gg/ml in the solution, and probably because

the ratio of net chitosan over cell numbers was much higher.

In early studies, proteins from extracellular and cellular fractions of loblolly

pine suspension cells were applied to a rotofor cell, an apparatus that separates

proteins based on isoelectric focusing (IEF). Extracellular proteins showed higher

1 ml 2 ml

0 0 20 40 60 120 180 240 30 60 90 (jg/ml chitosan)

841 321l

16 p 8 60,
32 pi 32 p
8 ml
64 p

Figure 17. Chitosan-induced Pschi4 protein expression in pine
suspension cells. Slash pine cells (0.6 ml of each) were incubated in 1
ml, 2 ml or 8 ml of LM medium with chitosan at the concentration (tpg/ml)
indicated. Proteins were isolated from the media, quantified and adjusted
to same concentration. From 1 to 64 pi of proteins were loaded on to a
nitrocellulose membrane and incubated with the anti-Pschi4 antibody.


chitinase activity in the acidic area (pH 4.6 to pH 6.4) after chitosan treatment (data

not shown).

Pschi4 protein expression in transgenic tobacco suspension cells. Tobacco

suspension culture was initiated from transgenic tobacco (Chi4 and CIB10 tobacco)

leaf disks. The protein level of Pschi4 in tobacco suspension cells was also

examined by western blot analysis. Fig. 18 shows that Pschi4 protein was present

in the supernatant extracellularr) fraction of the Chi4 tobacco suspension cells but

not in the cytosolic fraction. The size was approximately 25 kD, indicating it may not

be glycosylated as in pine. In contrast, nothing was detected in the CIB10 tobacco

suspension cells (Fig. 18). It is worthwhile to note that no chitosan responses were

observed at both the protein (Fig. 18) and mRNA levels (data not shown) in tobacco

suspension cells.

Transient Assay of Pschi4 Promoter-GUS Constructs

Plasmid Construction

The 5'-flanking sequence, putative promoter region, of Pschi4 was subcloned

and fused with the reporter gene uidA from bacteria which encodes p-glucuronidase

(GUS). A series of promoter deletion-GUS constructs were also made available by

time-dependent digestion by exonuclease III (Fig. 6). These constructs were made

in the vector pUC19 for transient assays and some of them were subcloned into the




0 0


00 00
CL CL' <
1P U,- $

Figure 18. Western blot analysis in tobacco suspension cells. Tobacco
suspension cell lines were initiated from leaves of transgenic tobacco
containing the Pschi4 or pCIB10 vector. Chi4 cells were treated with (CH) or
without (CK) chitosan (60 (pg/ml) for 24 h. Total proteins were isolated from
both medium supernatantt spnt) and cells (cytosol cyt). Western procedures
were the same as described in Fig. 16.


- 31.0

- 21.5

- 31.0

- 21.5


binary vector pBI101 for stable transformation into tobacco. All of them were named

according to the total size (kb) of the plasmid (Fig. 6).

Transient Expression in Onion Epidermis Cells

I was interested in identifying a functional pine promoter as well as defining

chitosan-responsive cis-element(s), if such elements) exist. Putative promoter-

GUS constructs were first tested by transient assays in onion cells (Fig. 19 and 20).

All constructs showed high promoter activity compared with the promoterless control

(Fig. 19A). Chitosan was included in the incubation, but did not induce higher

activity. Fig. 19 represents two typical results of these experiments. Since the

results from particle bombardment experiments showed large variations in onion,

both GUS and luciferase (LUX) activities are only comparable within a single

experiment. In the experiment shown in Fig. 20, more constructs were used and no

significant differences in activity were seen between constructs (Fig. 20), which

indicates that the smallest -200 bp promoter construct contained sufficient

regulatory elements to direct activated transcription in onion cells by transient


Promoter Comparison in Maize and Pine Suspension Cells

The commercially available CaMV 35S promoter (35S) has been

demonstrated to be highly expressed in angiosperms, however, there are conflicting

data regarding the efficiency of the 35S promoter in gymnosperms (Ellis, 1994).




* 1500.0




* + chitosan

-1200 -800 -500 -200 0

n no chitosan + chitosan

-2600 -80 -400

Figure 19. No chitosan-induction in transient assays in onion cells.
Onion epidermis cells were bombarded with 5 jLg of indicated WP-
GUS (see Fig. 6) and 5 p.g of Ubi-LUC, and cultured post-
bombardment with or without chitosan. (A) and (B) represent two
separate experiments. Data represent mean (+ SE) of five

D 2
0 o

no chitosan




0 01000



-4500 -2600


-800 -600


Figure 20. Promoter activity in onion cells. Onion cells were transformed by
particle bombardment as described in Fig. 19 and cultured in the absence of
chitosan. Data represent mean (+ SE) of six replicates.


Constructs containing either 35S-GUS or the white pine chitinase promoter (WP)-

GUS were bombarded into maize and pine suspension cells (Fig. 21). Both 35S

and WP showed significant promoter activity in maize cells, although 35S was much

higher than WP (Fig. 21A&C). In contrast, WP showed higher promoter activity

than 35S in pine cells (Fig. 21B&D). The overall GUS activity detected in pine cells

was lower than in maize (Fig. 21 C&D; scales are different). In fact, the internal

control, Ubi-LUC, also showed much lower activity in pine (average = 3.7 x 104

RLU) than in maize (average = 2.6 x 106 RLU). This is why the ratio of GUS/LUX

was higher in pine (Fig. 21B) than in maize (Fig. 21A).

Transient Assay in Pine Suspension Cells

Since no chitosan-induction was observed in onion cells (Fig. 19), it was

possible that some components that are present in pine could be missing in onion

cells. To test this idea, I repeatedly tested particle bombardment in pine suspension

cells. After optimization of the technique, GUS activity levels were sufficiently high

to obtain interpretable results (Fig. 21), but chitosan-induction was not observed

(Fig. 22A). Interestingly, the expression pattern was a little bit different in pine cells

from that in onion. The pWP-GUS9.3, containing the 4.5-kb WP promoter, showed

higher promoter activity than any of the others in pine cells (Fig. 22), while all

constructs showed similar activity in onion cells (Fig. 20). One small construct, -136

GUS, did not show any activity in pine cells compared with the -200 construct and

promoterless control (Fig. 22B).


A. B.

3600 9200
3000 -- ---
2 o 6900
DL1800 4600
o 1200
0x 2300
S600 F

0 II 0_
35S WP 35S WP
maize pine

C. D.

9000 220.0

s- 165.0 -------
0 6000 .- ....--- -----
D 3000 -
55.0 -- -

35S WP 35S WP
maize pine

Figure 21. Promoter comparison in maize and in pine cells. Maize and pine
suspension cells were bombarded with pBI221 (35S promoter) or pWP-
GUS9.3 (WP; 4.5 kb white pine chitinase promoter, see Fig. 6). Ubi-LUC
was an internal control in all cases. Data in (A) and (B) are expressed as
ratio of GUS activity (pmole MU) over luciferase activity (RLU relative light
unit). Data in (C) and (D) are expressed as GUS activity. All data represent
mean of two to five replicates.

u no chitosan

-4500 -1200 -700 -600

I + chitosan

-500 -400 -200

4000 O

-4500 -200


0 :


Figure 22. Promoter activity in pine cells. (A). Pine suspension cells were
bombarded with Ubi-LUC and promoter constructs (Fig. 6) as indicated
and cultured in the presence or absence of chitosan. (B). Bombarded
cells were cultured in the absence of chitosan. Each bar represents
mean (+ SE) of four (in panel A) or six (in panel B) replicates.


2 3000
1 2000

51 fl. ~ii 1

o o 3000

3 2000
x 1000





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 EDR2P8DP7_B90SIS INGEST_TIME 2012-02-20T21:28:13Z PACKAGE AA00009017_00001