Transgenic approaches to modulating naringin content in Citrus paradisi


Material Information

Transgenic approaches to modulating naringin content in Citrus paradisi
Physical Description:
ix, 96 leaves : ill. ; 29 cm.
Luth, Diane
Publication Date:


Subjects / Keywords:
Glycosides   ( lcsh )
Grapefruit   ( lcsh )
Plant genetic engineering   ( lcsh )
Horticultural Science thesis, Ph.D   ( lcsh )
Dissertations, Academic -- Horticultural Science -- UF   ( lcsh )
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph.D.)--University of Florida, 1997.
Includes bibliographical references (leaves 83-95).
Statement of Responsibility:
by Diane Luth.
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 - 028630719
oclc - 38855521
System ID:

This item is only available as the following downloads:

Full Text








I would like to thank my advisor Dr. Moore for her enthusiasm, support,

and friendship throughout my studies. Dr. Cantliffe also assisted with my

financial support over the years. Other members of my thesis committee,

Dr. Cline, Dr. Davis, and Dr. Guy, provided helpful suggestions to improve the

quality of this work and challenged me with the excellent courses they taught.

Michael McCaffery and Cindy Ragland provided excellent technical assistance

as well as providing friendship. All members of the Moore lab provided some

form of assistance, including James Teague who assisted with the tissue culture.

Dr. Darnell and Steve Hiss also gave my project help at critical times.

In other locations, Dr. Martin kindly provided us with her Antirrhinum

probes to get this project started. Dr. Hood and Dr. Slightom provided us with

vectors, Dr. Hahlbrock and Dr. Mol provided us with antibodies, and Dr. Berhow

made the naringin analysis possible.

Lastly, I would like to thank my husband and children for their patience.


ACKNOWLEDGMENTS ................................................................................ ii

LIST O F TA B LES ............................................................................................ v

LIST O F FIG U R ES .......................................................................................... vi

A B ST R A C T ..................................................... ......... .......................................viii

INTRODUCTION ........................................................ ................................... 1

LITERATURE REVIEW ................................................................................. 3

Flavonoid Biosynthesis.............................................................. ........... 3
Flavonoid Compounds in Citrus..............................................................4
N a ring in ................................................................................. ............. 6
Regulation of Flavonoid Biosynthesis .................................... ............ 7
Genetic Manipulation of Secondary Metabolism ..........................................10
Sense Suppression ..........................................................................10
Antisense Suppression ..................................................................... 14
Citrus Transformation........................................................................ 18

CLONING OF CITRUS CHS AND CHI cDNAS ................................................26

Introd uctio n .............................................................................. .... ................. 26
Materials and Methods .........................................................................27
R N A B lot A analysis .................................................................................. 27
D NA Blot A analysis ................................... .................................... ..... 29
Construction of cDNA Library................................................................29
Screening of cDNA Library ....................................................................29
D NA Sequence A nalysis....................................... .................................31
R results and D discussion ............................................................................. .. 31
Chalcone Synthase ..........................................................................32
C halcone Isom erase .................................................... ......................... 33


Intro d uctio n ................................................................................................ ... 4 5
Materials and Methods ...........................................................................45

Transformation Plasmids ......................................................................45
P lant T ransform ation ............................................................................. 48
PCR Analysis of Transformants.............................................................50
DNA Blot Analysis ......... ................................................... ........... 50
RNA Blot A analysis ........................................................... ....... .......... 51
Protein Blot Analysis .......................................................................52
Naringin A analysis ........................................................................ .... 53
Results and Discussion ......................................................................... 53
Regeneration and Initial Screening of Transformants............................53
Evaluation of Transgenic Plants.............................................................57
Polymerase chain reaction ......... ................................... ............. 57
D NA blot analysis ........................................ ........... ......................... 57
RNA blot analysis .......................................................................58
Protein blot analysis...................................... ........................... 60
N aringin analysis .................................................... ......................... 60

SUMMARY AND CONCLUSIONS .................................................................... 76

Cloning of Citrus CHS and CHI cDNAs ....................................................... 76
Plant Transformation and Transgenic Analysis .............................................77
Future S studies ............................................................................... ............ 8 1

R E FE R E N C E S .................................................................................................. 83

BIOGRAPHICAL SKETCH ..............................................................................96


Table page

2.1. Flavanone glycosides in Citrus.......................................... ........ ..... 22

3.1. Primers used to sequence CHS and CHI cDNAs from Citrus paradisi......35

4.1. Agrobacterium-mediated transformation experiments completed on
grapefruit cv. Duncan .......... ......................................................... ... 62

4.2. Results of GUS histochemical staining of regenerated grapefruit shoots
following Agrobacterium-mediated transformation ..........................................63

4.3. Sum mary of transform ants .......... ......................................................... 64

4.4. Naringin analysis data on grapefruit obtained with HPLC analysis ...........65


Figure page

2.1. Flavonoid C -C -C 6 skeleton chain............................................................. 23

2.2. The biosynthesis of naringin in Citrus......... .......... ......................................24

2.3. Flavanone aglycone structure and two isomeric rhamnoglucose
diglycoside groups (neohesperidose and rutinoside) ....................................25

3.1. RNA Northern blot analysis of grapefruit................................................... 36

3.2. Nucleotide sequence for a CHS cDNA isolated from C. paradisi .............37

3.3. BLAST search results obtained when the cDNA putative CHS
sequence isolated from C. paradisi was aligned with nucleotide
sequences of known genes........... ......................................................... 38

3.4. Genomic DNA blot analysis of grapefruit probed with the grapefruit
C H S cD N A ........................................................................................... ... 39

3.5. RNA blot analysis of total RNA from various grapefruit tissues probed
with the grapefruit CHS cDNA .................................................... ...........40

3.6. Nucleotide sequence determined for a CHI cDNA isolated from C.
paradisi........ .............................. ...................................... ....... .......... 41

3.7. BLAST search results obtained when the cDNA putative CHI sequence
isolated from C. paradisi was aligned with nucleotide sequences of
know n genes. ............................................................... ..................... 42

3.8. Genomic DNA blot analysis of grapefruit probed with the grapefruit CHI
cD NA ........................................... ................... .................................. 43

3.9. RNA blot analysis of total RNA from various grapefruit tissues probed
with the grapefruit CHI cDNA. .................................................... ...........44

4.1. Plasmid vector system used in the construction of transformation
vectors ............................... ................................................................... 66

4.2. Polymerase chain reaction analysis detection of the presence of the
NPTII and GUS genes in transgenic grapefruit plants..................................67

4.3. DNA blot analysis of CHS transformed grapefruit probed with GUS.........68

4.4. DNA blot analysis of CHI transformed grapefruit probed with GUS .........69

4.5. DNA blot analysis of CHS transformed grapefruit probed with a CHS
cD N A .................................................................................. ............. 70

4.6. DNA blot analysis of CHI transformed grapefruit probed with a CHI
cD N A ................................................................................................... ... 7 1

4.7. RNA blot analysis of CHS transformed grapefruit probed with CMV.........72

4.8. RNA blot analysis of CHS transformed grapefruit probed with a CHS
c D N A ............................................................... .............................. ... 7 3

4.9. RNA blot analysis of CHI transformed grapefruit probed with CMV..........74

4.10. RNA blot analysis of CHI transformed grapefruit probed with a CHI
c D N A ..................................................................................................... ... 7 5

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



Diane Luth

December 1997

Chair: Dr. Gloria Moore
Major Department: Horticultural Science

The compound naringin imparts a bitter flavor to grapefruit (Citrus

paradise). The presence of excessive amounts of naringin can result in

commercially undesirable fruit and juice products. Naringin is a flavanone

glycoside derived from the flavonoid biosynthetic pathway. The first two steps of

the flavonoid pathway are catalyzed by the enzymes chalcone synthase (CHS)

and chalcone isomerase (CHI). It was reasoned that reducing the levels of these

enzymes might reduce naringin content in grapefruit tissues resulting in more

desirable fruit. The genes encoding these enzymes were isolated from a

grapefruit leaf cDNA library. Genomic DNA analysis indicates that both genes

might be present in several copies in the grapefruit genome. RNA for both

genes was present in all tissues tested. A transgenic approach to reduce CHS

and CHI levels by antisense RNA or cosuppression was attempted. Four

plasmids were constructed, one with each gene in either the sense or antisense

orientation, all under the control of the CaMV 35S promoter. The plasmids were

introduced into the grapefruit genome by Agrobacterium mediated

transformation. Several modifications of previously reported grapefruit

transformation techniques resulted in improved methodology. Transgenic plants

were obtained for each construct. Transcript levels and naringin content varied

among the transformants. Attempts to correlate RNA level to naringin content

were inconclusive. Suggestions for future directions are discussed.


Grapefruit (Citrus paradisi Macf.) is a crop of major economic

importance to the state of Florida. The origins of grapefruit are still unknown,

but may have occurred in Barbados (Scora et al. 1982). The crop was

introduced in Florida around 1823, and by 1995, approximately 127,300 acres of

commercial grapefruit were in production, resulting in 55 million 85 lb. box units

harvested (Florida Dept. of Citrus). Twenty one million of these went to fresh

market, with the rest utilized in processed juice products.

One of the major determinants of grapefruit quality is naringin content.

Naringin is a flavanone glycoside that imparts a bitter flavor that can make fruit

and juice undesirable or totally unacceptable to the consumer (Berhow and

Vandercook 1991). Flavanone glycosides are secondary metabolites produced

by the flavonoid pathway, a branch pathway of phenylpropanoid biosynthesis

(Hahlbrock and Grisebach 1979). They accumulate to high levels in the fruit peel

and juice, and are extracted as the fruit is pressed for commercial juice

production (Attaway 1977). Other flavonoids produced in Citrus include

anthocyanins, flavones, and other flavanones (Horowitz and Gentili 1969;

The first committed step in the flavonoid pathway is catalyzed by chalcone

synthase (CHS) followed by chalcone isomerase (CHI). Studies have shown
that transcriptional activation of the CHS and CHI genes is a major factor
regulating enzyme levels and hence product accumulation (Wingender

et al. 1989). In several crops, levels of CHS have been manipulated using a

transgenic approach (Napoli et al. 1990; van der Krol et al. 1990b; Elomaa et al.

1993). Both antisense and co-suppression methods resulted in a decrease or

elimination of some anthocyanin compounds produced later in the pathway. We

considered that isolating genes involved in biosynthesis of flavonoid compounds

and constructing transformation plasmids for sense and antisense manipulations

could be a powerful approach to decreasing levels of flavanone glycosides in

Citrus paradisi. Specifically we hypothesized an overall decrease in flavonoids

would include a decreased level of naringin produced.

The availability of heterologous probes for CHS and CHI provided a tool to

clone the genes from Citrus. Once cloned, these genes were used in the

construction of Agrobacterium-mediated transformation vectors and introduced

into the grapefruit cultivar "Duncan". Transformation of Citrus via Agrobacterium

provided a reliable method to integrate transgenes and obtain plants for analysis.

Once regenerated plants were available, these plants were analyzed for

evidence of transformation, gene expression, and levels of naringin.


Flavonoid Biosynthesis

The flavonoids represent a large group of water soluble phenolic

compounds that form an important class of secondary metabolites in plants.

They function in diverse processes including flower pigmentation, UV-protection,

plant defense against pathogens, and in legume nodulation (Koes et al. 1988).

All flavonoids are based on the C6-C3C6 skeleton with the 2 component aromatic

rings arising from different pathways (Figure 2.1). Ring A is derived from 3 C2

units formed by head to tail linkage from malonyl residues from the malonyl-

acetate pathway. Ring B, together with the three carbons linking A with B, is

derived from the phenylpropanoid residue 4-coumaroyl-CoA, from the shikimate

pathway (Bar-Peled et al. 1993). The oxidation level and other structural

features of the C3 carbon chain determine the different classes of flavonoids

(Goodwin and Mercer 1983).

Chalcone synthase (CHS) catalyzes the first committed step in the

biosynthesis of all flavonoid derivatives (Heller and Hahlbrock 1980; Sommer

and Saedler 1986). It catalyzes the condensation reaction between malonyl-

CoA and coumaroyl-CoA to form naringenin chalcone (Figure 2.2). CHS was
first isolated from young leaves and cell suspension cultures of parsley
(Kruezaler and Hahlbrock 1972).

The second enzyme in the flavonoid biosynthetic pathway, chalcone

flavanone isomerase (CHI) catalyzes the stereo-specific isomerization of

naringenin chalcone into the corresponding naringenin flavanone (van Tunen et

al. 1988). In the absence of CHI, the isomerization reaction can still occur, albeit
at a slower rate (Holton and Cornish 1995). Naringenin flavanone is the
precursor for a number of different classes of flavonoids including anthocyanins,
isoflavonoids (including phytoalexins), flavanones, flavones, and flavonols.

Flavonoid Compounds in Citrus

Citrus plants contain three types of flavonoids: anthocyanins, flavones,
and flavanones (Horowitz and Gentili 1977a; Lewinsohn et al. 1989a). In many
plants, anthocyanins typically make a large contribution to fruit and flower color,

but not in Citrus (Horowitz and Gentili 1977a). Only 5 anthocyanin compounds
have been described, and only 1 of them, a water-soluble red pigment of Citrus
sinensis cv. Moro (blood orange), is responsible for fruit color (Chandler 1958).
Forty-three flavones have been described, making them the most numerous, but

they occur in low concentrations and are difficult to purify. The flavones are
widely distributed in fruit, but usually make no contribution to the color of Citrus.
Although the flavones are neither bitter nor sweet, they can affect fruit taste. It is
believed the compounds raise the threshold at which bitterness is detected by
competing for sites on taste receptors (Horowitz and Gentili 1977b). Flavanone

glycosides are the major flavonoid constituent accumulating in large
concentrations as end products (Albach and Redman 1969; Horowitz and Gentili

1977a). They have been found in almost every Citrus tissue analyzed including
flowers, leaves, and fruit (Jourdan et al. 1985). Eleven different flavanone
glycosides have been described (Horowitz and Gentili 1977b). They have a
pronounced effect on fruit taste, but their function is unknown.

During the biosynthesis of flavanone glycosides in Citrus, the flavanones
are hydroxylated, glucosylated, and then rhamnosylated in a step-wise fashion
(Figure 2.2). The addition of a rhamnose to either the 2-hydroxyl or the

6-hydroxyl position on the glucose group forms either a neohesperidoside or
rutinoside respectively (Berhow and Smolensky 1995). Neohesperidosides are
bitter compounds and rutinosides are tasteless. The degree of bitterness of the
neohesperidosides is relative to the composition of the flavanone aglycone
structure. The relative bitterness of the 4 major nehesperidosides is naringin >
poncirin > neohesperidin > neoeriocitrin (Hasegawa et al.1996)(Figure 2.3).

Early work in citrus taxonomy recognized the usefulness of flavonoid
composition as a means of distinguishing different species of Citrus (Swingle
1943; Kefford 1959; Horowitz 1961; Albach and Redman 1969). Flavonoid
bitterness occurs in citrus species related to Citrus grandis (pummelo)

(Hasegawa et al. 1996). The bitter forms of flavanone glycosides are present at
all times in the fruit tissues (Maier 1969). Pummelo accumulates the flavanone
neohesperidosides exclusively while the tasteless flavanone-rutinoside
predominates in fruit related to Citrus sinensis (sweet orange) and Citrus limon
(lemon). Citrus aurantium (sour orange) and Citrus paradisi (grapefruit)
accumulate both the bitter neohesperidosides and the tasteless flavanone
rutinosides and are presumed hybrids (Berhow and Vandercook 1989). Citrus

paradisi is believed to be a hybrid between C. sinensis (which contributes the

rutinosides) and C. grandis (which contributes the neohesperodosides)(Scora et
al. 1982).


In a comprehensive study of citrus taxonomy (Swingle 1943), the
presence of naringin in C. grandis is noted as an "important chemical difference",

which in addition to morphological differences serves to make C. grandis one of
the most distinct species of Citrus. In grapefruit naringin is the most predominant
bitter flavonoid compound in the plant, reaching levels of 40-70% of the dry
weight of small green fruit (Jourdan et al. 1985). Excess amounts of naringin in

grapefruit adversely affects the quality of the fruit and juice (Hagen et al. 1966).
The naringin levels found in grapefruit juice far exceed the amounts of the
other neohesperidosides (Table 2.1)(Johnson 1988). Quantitative distributions
of naringin using radioimmunoassay procedures have been determined in seeds,

seedlings, young plants, branches, flowers and fruit of Citrus paradisi Macf., cv.

Duncan (Jourdan et al. 1985). High levels of naringin were associated with very
young tissue and lower levels were found in older tissues. Seed coats of
ungerminated seeds and young shoots had high naringin concentrations,

whereas cotyledons and roots had very low concentrations. Light grown
seedlings contained nearly twice as much naringin as etiolated seedlings. In
young plants and branches, the naringin content was highest in developing
leaves and stem tissue. In flowers, the ovary had the highest levels of naringin
accounting for nearly 11% of the fresh weight. In the fruit, levels of naringin were

much higher in the peel than in the juice sacs (Jourdan et al. 1985). Much of the
naringin present in commercially processed juice is derived from peels and is
extracted along with the juice (Attaway 1977).
A later study using HPLC to detect naringin levels in Citrus aurantium
(sour orange) found results confirming the radioimunoassay data (Castillo et al.
1992). Naringin was found to be at maximum concentration during the

logarithmic phase of growth, gradually decreasing until the organs reached

maximum development. This decrease in the leaves, flower buds, and fruits was

due to turnover, and a dilution of the flavonoids caused by cell growth, with total

content per organ continuing to increase.

Studies on the enzymatic synthesis of naringin in Citrus include the
isolation or partial characterization of the 4 flavonoid pathway enzymes (Figure

2.2). Chalcone synthase and chalcone isomerase activity has been assayed by
several different groups (Hasegawa and Maier 1970; Raymond and Maier 1977;
Lewinsohn et al. 1989b). A 7-glucosyl transferase has been isolated from Citrus

paradisi (grapefruit) (Mclntosh et al. 1990). The enzyme has a MW of 55 kDa
and is believed to catalyze the 7-O-glucosylation of both naringenin and
hesperetin. A rhamnosyltransferase has been purified from pummelo leaves
described as a monomer with a molecular mass of 50 kD. It attaches rhamnose

to the C-2 hydroxyl of the glucose (neohesperidoide), (Bar-Peled et al. 1991).

Regulation of Flavonoid Biosynthesis

In Citrus, most studies involving flavonoid biosynthesis have been of a
biochemical nature, confirming that the Citrus flavonoid pathway is similar to that
described in other species (Lewinsohn et al. 1986; 1989a.) One genetic study
has been reported using segregation ratios to determine the inheritance of
flavanone neohesperidoside, and concluded 2 dominant multiple genes
controlled the compound formation (Matsumoto and Okudai 1991). No
molecular studies have been reported on regulation of any portion of the
flavonoid pathway, and no genes leading to naringin have been cloned.
Although naringin and the other flavanone glycosides occur in such high

concentrations, no functional role for these compounds in Citrus has been
determined. The metabolic fate of the compounds is also unknown.

The sole report of modifying specific levels of a flavanone glycoside
compound examined the individual effects of 4 different plant growth regulators
on naringin content in citrus (Berhow and Vandercook 1992). Immature fruit on
grapefruit trees were treated with 1000 ppm of benzyladenine (BA), abscisic acid
(ABA), gibberelic acid (GA3), and napthaleneacetic acid (NAA) applied to the fruit
as a lanolin paste. The BA and ABA treatments resulted in increased levels of
naringin content in the fruit and lowered brix (total soluble solids). Treatment
with NAA did not produce any measurable changes. The application of GA3 after
fruit set lowered the naringin content of the juice sacs and possibly of the albedo
(mesocarp) of the mature fruit.

In other plants expression of the gene encoding CHS can be easily
monitored by color formation in various tissues, and extensive studies have been
done at the genetic and biochemical level. CHS was first isolated from parsley
using a combination of differential screening and hybrid arrested and hybrid
selected translation, which identified a cDNA clone homologous to the CHS
mRNA (Kreuzaler et al. 1983). The parsley CHS clone was later used as a
molecular probe to isolate clones of 2 different CHS genes from petunia (Reif et
al. 1985). CHS is encoded by a single-copy gene in some plants such as
Petroselinum crispum (parsley) (Herrmann et al. 1988) and by multi-gene
families in others such as Petunia hybrida (petunia) (Koes et al. 1988). Twelve
different CHS genes have been found in the petunia genome, but only 4 of these
(CHSA, CHSB, CHSG, and CHSJ) are known to be expressed (Koes et al.
1989). All 4 genes are expressed in UV-irradiated seedlings, but only CHSA and
CHSJ are expressed in floral tissue. In maize, 2 genes encode CHS. Whp
controls CHS activity in pollen (Coe et al. 1981) and c2 is involved in

anthocyanin biosynthesis in seed (Dooner 1983). CHS has been isolated and
sequenced for a number of species. A comparison of CHS sequences from 8
different species revealed an identity higher than 66% at the nucleotide level and
80% at the amino acid level (Niesbach-Klisgen et al. 1987.) The CHS transcript
is G/C rich in monocotyledons (65-69.3%) but not in dicotyledons (45.3-53.9%).

Stress-induced expression of CHS has been studied in detail. Light,
elicitors, and wounding induce the expression of some CHS genes (Wingender
et al. 1989). In addition to exogenous signals, endogenous signals can also
regulate the expression of CHS genes. Intermediates in the phenylpropanoid
pathway can regulate the expression (Takeuchi et al. 1994). The presence of
para-coumaric acid and low concentrations of trans-cinnamic acid stimulate the
expression of a bean CHS, whereas high concentrations of trans-cinnamic acid
reduce the level of gene expression (Loake et al. 1991). CHS has also been
shown to display sugar-responsiveness in 2 cases. The presence of sugars was
found to increase the expression of a petunia CHS gene in transgenic
Arabidopsis (Tsukaya et al. 1991) and endogenous CHS transcripts in young
leaves of Camellia sinensis (black tea)(Takeuchi et al. 1994).

The first CHI cDNA was isolated from Phaseolus vulgaris (french bean)
using antibodies specific for the purified enzyme (Mehdy and Lamb 1987). Bean
has a single CHI gene that can be induced by fungal infection and mechanical
wounding (Grotewald and Peterson 1994). Petunia hybrida has two CHI genes
with different expression patterns: CHI-A is expressed in floral tissues and UV-
irradiated seedlings. The temporal and spatial expression is determined by
differential promoter usage (van Tunen et al. 1988; 1989). CHI-B is expressed in
immature anthers only (van Tunen et al. 1990a; b; 1991). CHI has been cloned
in a number of other species including Antirrhinum majus (snapdragon) (Martin

et al. 1991), Arabidopsis (Shirley et al. 1992), and Mediago sativa (alfalfa)
(McKhann and Hirsch 1994).

Genetic Manipulation of Secondary Metabolism

Molecular genetics provides an opportunity to manipulate levels of
secondary compounds by altering the expression of genes encoding key

enzymes or by diverting existing pathways. One approach is to block expression
of the gene for an enzyme at a particular step in the pathway, thereby blocking
biosynthesis at that step. Gene silencing has primarily been achieved in 2 ways:
through sense suppression or with antisense technology.

Sense Suppression

The silencing termed co-suppression or sense suppression involves the
coordinate silencing of either a transgene and a homologous endogenous gene
or two homologous transgene loci. This can also include multiple copies of

(partially) homologous genes. Sequence homology and repeats are thought to
interfere with transcriptional or post-transcriptional events, resulting in silencing
that can be general, developmental or random (Flavell 1994; Matzke and Matzke

Suppression of endogenous genes was first demonstrated for genes
involved in the synthesis of anthocyanin pigments in petunia flowers (Napoli et
al. 1990; van der Krol et al. 1990b) and for the polygalacturonase ripening gene
in tomato (Smith et al. 1990). An attempt to overexpress CHS in pigmented

petunia petals was made using a chimeric CHS gene fusing the cauliflower
mosaic virus 35S promoter (CaMV 35S) to the coding sequence of a CHS cDNA.

The introduced gene created a block in anthocyanin synthesis, and up to 50% of

the transgenic plants showed sectors of reduced or no anthocyanin pigment in

the flowers (Napoli et al. 1990). Two classes of patterns were observed, and the

lack of pigment was correlated with very low levels of mRNA from the newly

inserted and endogenous copies of the CHS genes. In white flowers the

message levels of the transgene and endogenous CHS gene were coordinately

and reversibly suppressed. The somatic and germinal stability of the novel color

patterns was variable, and somatic reversion of plants was associated with a

coordinate rise in the steady state levels of the endogenous and transgene

messages. These results were confirmed by a second group (van der Krol et al.

1990a), but the mechanism was not determined.

Silencing can occur either at the transcriptional or post-transcriptional
level (Elmayan and Vaucheret 1996; Parks et al. 1996). Evidence of

transcriptional silencing was found in petals of transgenic petunia plants that had

single and multiple copies of the maize Al gene driven by the CaMV 35S

promoter. The silencing resulted from a block in the transcription of the

transgene and correlated with methylation of the cytosines within the 35S
promoter sequence (Linn et al. 1990, Meyer et al. 1993; 1994). In general,

transcriptional silencing is thought to be caused by promoter inactivation, often

correlated with methylation (Stam et al. 1997). However methylation was not

involved when the chlorophyll a/b binding protein (CAB) was suppressed in

Arabidopsis (Brusslan et al. 1993).

One model to explain post-transcriptional silencing is the biochemical
switch model (Meins and Kunz 1994). The hypothesis is the product of gene
expression, RNA, accumulates until a threshold is reached, at which point RNA

degradation is initiated. In the case of transgenic plants, the cumulative

production of RNA from the endogenous gene and transgene would provoke

turnover of RNA. This accounts for the run-on transcripts that are synthesized,

but the absence of RNA in a steady-state pool. The analysis of

polygalacturonase gene silencing in tomato (Smith et al. 1990) supported such a
model. The silencing mechanism was not active until the stage in fruit

development when the silencing transgene and target endogenous gene were

both transcribed. This type of model has received additional support in recent

studies linking post-transcriptional gene silencing and virus resistance in

transgenic plants expressing virus-derived sequences (English et al. 1996).

Resistance has been achieved with untranslatable as well as translatable

transgenes in plants that accumulate low levels of the transgene mRNA and

protein product. A cytoplasmic post-transcriptional RNA degradation process

that elicits virus resistance is activated when the threshold level of transgene

derived transcripts is exceeded (Smith et al. 1994). Like the biochemical switch

model, this phenomenon requires that a threshold level of RNA be exceeded. A
specific study in tobacco to assess the number of tobacco etch virus (TEV) coat

protein transgenes involved in TEV resistance found 2-5 transgenes were

necessary to establish a highly resistant state (Goodwin et al. 1996). However,

there also appeared to be lower and upper limits in attaining the resistant state.

Plants containing one copy of the transgene, while not resistant, were able to

slowly recover from infection. Susceptible transgenic lines with a very high

transgene copy number may have experienced gene silencing at the nuclear

level. Silencing of non-viral transgenes post-transciptionally have prevented the
infection of a chimeric virus carrying non viral sequences (English et al. 1996).
Methylation occurred in the silenced non-viral genes and evidence exists that

transgene methylation could be the cause (Keshet et al. 1985) or the effect of

the silencing mechanism (Wassenegger et al. 1994).

Another model to explain post-transcriptional silencing was based on
similarities between the suppression of gene expression by sense and antisense

transgenes. Antisense RNA, generated by transcription of the "wrong" DNA

strand of the inserted sense transgene was proposed to be responsible for the

down regulation of the endogenous gene in the plants (Grierson et al. 1991).
These similarities include specificity for the target gene, the subsequent
reduction in endogenous RNA levels and production of similar endogenous

mRNA fragments. Grierson has proposed the production of antisense RNA
could arise from read-through transcription into the sense transgene from
external promoters near the site of T-DNA integration or from an adjacent
converging antibiotic resistance gene. However, specific pieces of evidence in

gene silencing and trans-inactivation examples argue strongly against a
chromosomal promoter being the source of antisense RNA. Read-through
transcription has been shown to be blocked by the NOS poly(A) signal/terminator
(Inglebrecht et al. 1991). The presence of this type of regulatory region in
transgene constructs should eliminate read-through transcripts occurring from
converging promoters (Matzke and Matzke 1993).

Post-transcriptional silencing does not always involve production of
excessive levels of transgene RNA. In the case of CHS in petunia (van Blokland
et al. 1994), the transcriptional activities of the suppressed genes was

comparable to those of the nonsuppressed genes. Recently a more detailed
genetic and molecular analysis was completed on the transformants generated
by van Blokland (1994). Silencing was associated with the presence of
multimeric T-DNA in transgenes driven by the CaMV 35S promoter and
promoterless CHS transgenes (Stam et al. 1997). All transformants exhibiting

silencing contained a T-DNA locus composed of 2 or more T-DNA's arranged as
inverted repeats. The suppressive activity of the transgenes may be associated

with a type of DNA modification which does not affect transcription but which
affects other processes, for example, the processing of primary transcripts.

Other properties of T-DNA transgene locus might be implicated, including
genomic position and copy number (Stam et al. 1997). The T-DNA structure has
been shown to affect transgene expression including the incidence of epigenetic
phenomena with multicopy inserts often showing lower expression than a single
inserted copy (Flavell 1994; Hobbs et al. 1993). Epigenetic effects can be
caused by DNA homology between the transgene and another locus, or by a

duplicated arrangement of the transforming DNA. The correlation between T-
DNA methylation and inactivation of T-DNA encoded genes has been verified in
many studies (Matzke and Matzke, 1991), but it is still unknown whether this

methylation is a cause or effect of gene silencing (Finnegan and McElroy 1994).
However silencing has also been associated with a single copy T-DNA locus
when the transgene was expressed from an enhanced CaMV 35S promoter
(Elmayan and Vaucheret 1996). When a regular CaMV 35S promoter was used
the frequency of silencing was much less (Palauqui and Vaucheret 1995).

Antisense Suppression

Antisense suppression is caused by blocking the informational flow from
DNA via RNA to protein by the introduction of an RNA strand complementary to
the sequence of the target RNA (van der Krol et al. 1988a). The first report on
artificial antisense regulation in plants involved examination of transient
expression levels of chloramphenicol acetyl transferase (CAT) in protoplasts
from carrot cell cultures (Ecker and Davis 1986). The sense CAT construct

contained the weak nopaline synthase gene promoter (NOS). To insure a
relatively high expression of the antisense gene, the antisense CATwas placed

under the control of CaMV 35S or phenylalanine ammonia-lyase (PAL) promoter.

Co-electroporation of the antisense and sense CAT genes reduced CAT activity

up to 95%.

The first wild-type plant genes artificially regulated by an antisense RNA

were the CHS-encoding genes from petunia and tobacco plants (van der Krol et

al. 1988b). An antisense CHS gene was made from the member of the petunia

CHS multigene family that accounts for 90% of the steady state mRNA level in

floral tissue. A 35S CaMV promoter and NOS 3' tail fragment providing a poly(A)

addition signal to the antisense transcript were included in the chimeric gene

fusion. The antisense gene was introduced into petunia and tobacco cells. The

regenerated plants from both species showed altered pigmentation of the

flowers. The mRNA steady-state levels in flowers were reduced, down to

approximately 1% of the wild-type. Levels were also reduced when the

experiment was done with a second member of the petunia multigene family,

CHS-J. This study showed that the antisense gene functioned on targets that

were not 100% homologous; the petunia CHS-A gene is 86% homologous to the

CHS-J gene, while sequence homology between tobacco and petunia CHS is


Position effect, rearrangement of the target gene, duplicated copies of the

antisense sequences, capability of the promoter, and specificity of the antisense

sequence for its target RNA have all been implicated in causing the variation in

antisense effect (Bourque 1995). Antisense polygalacturonase DNA present as

a single copy per diploid genome in tomato resulted in 50-95% down-regulation
of the enzyme activity in individual transformants. The variation might be

explained by the site of insertion (Smith et al. 1990). The selfed progeny

containing 2 copies of the antisense gene exhibited a higher reduction in enzyme

activity. However, other studies have not found homozygous plants to have

enhanced reduction in gene activity (Hall et al. 1993). While most studies have

used the strong CaMV 35S RNA promoter for the antisense transcript, other

promoters have been utilized including CAB (Sandier et al. 1988; Cannon et
al.1990), NOS (Ecker and Davis 1986; Pang et al. 1993), CaMV 19S (Delauney
et al. 1988), PAL (Ecker and Davis 1986), and CHS (Mol et al. 1989; van der
Krol et al. 1990c). The use of an RNA polymerase III promoter provides an
alternative approach for producing high levels of antisense transcripts (Bourque
1995). As the transcripts from RNA polymerase III are neither polyadenylated or

capped, the products are shorter than those transcribed by RNA polymerase II.
It is believed that the smaller products may function more efficiently in hybridizing
to target sequences (Bourque and Folk 1992).

Few studies have directly addressed mechanisms of action in antisense
suppression. As with sense suppression, theories exist that support antisense

RNA functioning at either the level of transcriptional/post-transcriptional control,
or translational control (Bourque, 1995). No evidence has been found for direct
transcriptional control through an interaction between antisense RNA and the

DNA template. A proposed post-transcriptional mechanism is the formation of
an RNA duplex between the antisense and target RNAs as the cause of the
repression. However, no evidence of duplex formation has been found. In
antisense CHS plants, no excess of antisense CHS mRNA was detected, and
the CHS mRNA isolated from white flowers was susceptible to digestion by
RNase A, a single strand specific RNase implying that a double stranded RNA

molecule could be unstable, or perhaps the duplex formation is not the sole
mechanism for the down regulation (van der Krol et al. 1990b). Some evidence
exists that antisense transcripts affect mRNA processing by interfering with

splicing of introns during maturation of mRNA (van der Krol et al. 1988c; Tieman
et al. 1992).

Translational control implies that antisense RNA and target RNA are

transported from the nucleus to the cytoplasm in the form of a stable duplex

(Bourque 1995), or that duplex formation occurs in the cytosol. In tobacco, the

amount of an expressed antisense alfalfa glutamine synthetase (GS) transcript

was correlated with the degree of inhibition in synthesis of the protein (Temple et

al. 1993). No change occurred in the steady-state level of endogenous tobacco

GS mRNA. It was hypothesized that a stable heteroduplex could contribute to

blocking of ribosome binding, or transport of the message out of the nucleus

(Bourque 1995).

Several groups have expressed various regions of target genes in their

antisense constructs with varying results. Using different regions of the NOS

template in reverse orientation behind a CAB gene promoter, it was concluded

that the 3' region was more effective for suppression than the full length gene

NOS gene (Sandier et al. 1988). A relative comparison of CHS subgenomic

fragments found the 3' portion more effective at affecting flower pigmentation,

but further division of the 3' segment into two smaller fragments found the

shorter 5'-most region contributing the inhibitory effect (van der Krol et al.

1990c). A similar study using GUS concluded that the full length gene or any 5'

terminal sequence longer than 68 base pairs were more effective at attaining the
inhibitory effect than the 3' terminal sequence or any 5' region less than 68 bp

(de Lange et al. 1993).

In summary, both sense and antisense approaches can silence gene

expression and could be used to block a pathway by inhibiting production of a
key enzyme. However, the results are neither always predictable nor stable.

Different mechanisms appear to be involved in activating silencing at the

transcriptional and post-transcriptional levels. Position effect, choice of

promoter, methylation, and copy number are only a few of the variables that can

influence silencing. While the results appear almost contradictory at times, they

also provide challenges for future genetic and molecular experiments that involve
the silencing phenomena.

Citrus Transformation

Most varieties of Citrus grown today for scion and rootstocks did not arise
from conventional breeding programs. Long juvenile periods, high degrees of
heterozygosity, nucellar embryony, and interspecific hybridization pose specific
problems in Citrus improvement (Davies et al. 1993). While some progress can

be made in selecting varieties with optimal characteristics, methods are limited in

what they can accomplish. Transformation of plants is seen as a way of bringing
about a more limited (and specific) genetic change while the majority of the plant
genome remains undisturbed (Dixon 1991).
The first published report of citrus transformation (Kobayashi et al. 1989)
described a polyethylene glycol-mediated transformation of protoplasts from

Citrus sinensis (navel orange). The protoplasts were isolated from suspension
cultures and transformed with a circular plasmid containing the selectable marker
gene aminoglycoside phosphotransferase II (APH 3'll). Using 9x105protoplasts,

10 colonies were regenerated on medium containing 50 mg/L kanamycin sulfate,

and 2 stable transformants were found. They were not able to regenerate plants
from the calli. This same approach was used with Citrusjambhiri (rough lemon)

(Vardi et al. 1990). Protoplast isolation was from calli, and selection for embryo
induction and successful regeneration of the NPTII transformants was done with
paromycin sulfate (a kanamycin analog). Nine stable transformed calli were

found, and 2 plants were regenerated.

The first Agrobacterim-mediated approach (Hidaka et al. 1990) reported
transformants from Citrus sinensis and Citrus reticulata (mandarin).
Embryogenic calli were used to initiate suspension cultures that were

co-cultivated with A. tumefaciens using binary or cointegrate vectors. Selection
was made for kanamycin or hygromycin resistant callus colonies.

Transformation frequencies varied with the calli used, estimates being 0.05% on
a per cell basis.

All methods described thus far required the use of embryogenic callus.
This is limiting as it has not been possible to induce callus growth in all citrus

cultivars (Gmitter and Moore 1986). A new Agrobacterium approach, utilizing

internodal stem segments as the explant, by-passed the need to develop and
maintain callus cultures (Moore et al. 1992). Seedlings of Citrus aurantifolia (key

lime) and Citrus sinsensis L. Osbeck x Poncirus trifoliata L. Raf. (Carrizo
citrange) were grown in sterile culture tubes. Explant material consisted of 1 cm

stem segments from 2 to 4 month old plants. Inoculation was accomplished by
placing a drop of Agrobacterium tumefaciens on top of a vertically oriented
explant. When 100lg/ml kanamycin was used for selection, shoots arose from

the cut ends with little or no callus production. Transformation frequencies were

reported to be 4-8%. The inability to root the plantlets was cited as the limiting
factor in the regeneration procedure.

Improvement was made to the stem segment system by Kaneyoshi et al.
(1994). Seeds of a citrus relative, Poncirus trifoliata, were peeled and grown in
the dark. The young etiolated seedlings were used after 20 days in culture.
Epicotyl segments of approximately 1 cm were soaked in a concentrated

Agrobacterium solution (5x108 cells ml-') and plated horizontally. Histochemical

R-glucuronidase (GUS) staining was positive for 55.4% to 87.7% of the
regenerated shoots. However, Poncirus trifoliata, (a Citrus relative), responds

differently in vitro than Citrus, and is easier to root (Gloria Moore, personal


An alternative approach to improving the stem explant method involved
extending the period of kanamycin selection and grafting the transformed shoot

apices onto rootstocks (Peia et al. 1995a). Using Carrizo citrange,

Agrobacterium inoculated explants were placed horizontally onto kanamycin

selection plates and kept in the dark for 8 weeks. Following a 1-4 month period

in the light the regenerated shoots were removed. Transformation frequency of

the primary regenerants was low, but maintaining the explants on selection and

appropriate media resulted in callus and new shoot growth. The shoots derived
from callus showed a higher frequency of GUS staining (55%), however, 46% of

the regenerated shoots were morphologically abnormal. This was presumed to

be caused by the extended exposure to kanamycin sulfate. The authors claimed

the higher frequency of transformation, while regenerating some abnormal

shoots, also provided more normal transformed shoots than previous methods.

Problems associated with rooting were also eliminated by using a shoot tip

grafting method. Nearly 100% of the grafted shoots survived. The same group

also used 6-12 month old greenhouse grown seedlings as explants with 7.9% of
the regenerated sweet orange shoots positive for GUS (Pefia et al.1995b).

Particle bombardment as a transformation method has been reported for

Citrus (Yao et. al 1996). Cells grown in suspension from embryogenic callus of

tangelo (C. reticulata Blanco X C. paradisi Macf.) were bombarded with a
plasmid containing the coding sequences of the uidA gene for GUS and nptll.
Transformed calli, embryos, and plantlets were obtained. Using 0.3-0.4 g fresh

weight cells per bombardment experiment, the frequency of transformation was

approximately 1000 GUS staining foci per bombardment. Approximately 13

positively staining call were obtained 2 months later. Only one plantlet was
analyzed for evidence of transformation.

The frequency of Agrobacterium-mediated transformation of Citrus using
stem explants has continued to improve and offers new potential for crop
improvement including adding disease resistance (Gutierrez et al. 1997). The

methods of direct DNA delivery and particle bombardment involve more

extensive regeneration methods, and offer no clear advantage over

Agrobacterium-mediated approaches. However these alternative techniques
may be necessary for Citrus cultivars not readily transformable by the
Agrobacterium approach.

Table 2.1. Flavanone glycosides in Citrus*.

From Johnson 1988.
** Concentration of flavanone glycoside in grapefruit juice
Analytical results of Hagen et al. 1965.


GLYCOSIDE in GFJ pg/ml **

Naringenin Naringin 306


Isokuranetin Poncirin 17


Hesperetin Neohesperidin 10.5


Eriodictyol Neoeriocitrin


Figure 2.1. Flavonoid C6-C3-Cskeleton chain.

Acetate Malonate
3x Malonyl-CoA C H

c on -scoA yntase

Chalcone synthase isA



Chalcone isomerase


OH 0

genin Chalcone


V Naringenin
O Flavanone

7-Glucosyl transferase

-. OH

glucosyM-Q 0 /
S Prunin

1-2 Rhamnosyl transferase

glucosyl-O o 1 o
1-2 "rnL ) Naringin
rhamnose W
OH 0

Figure 2.2. The biosynthesis of naringin in Citrus from Bar-Peled et al. 1993.



OH 0



HO r--O-C 2 H HO

rutinoside neohesperidoside
(6-0-a-L-rhamnosyl-p -D-glucoside) (2-0-a-L-rhamnosyl-p-D-glucoside)

Figure 2.3. Flavanone aglycone structure and two isomeric rhamnoglucose
diglycoside groups( neohesperidose and rutinoside). The flavanone
hesperidosides are bitter, and the bitterness is relative to the composition of the
flavanone aglycone structure. All bitter forms contain the neohesperidoside at
the R, position. Naringin (R2=OH, R3=H); Poncirin (R2=OMe); Neohesperidin
(R2=OMe, R3=OH); Neoriocitrin (R2=OH, R3=OH). The flavanone rutinosides are
tasteless. From Hawsegawa et al. 1996.



The bitter flavor compound naringin can accumulate in grapefruit in large

quantities, resulting in fruit and juice quality that is unacceptable to the consumer

(Hagen et al. 1966). The overall objective of our research was to reduce the

levels of naringin produced in C. paradisi. We were interested in using a

transgenic approach to decrease levels of key biosynthetic enzymes in the

naringin pathway.

The flavonoid biosynthetic pathway leading to the production of naringin
involves 4 enzymes: chalcone synthase, chalcone isomerase, 7-0 glucosyl

transferase, and 1-2 rhamnosyltransferase (Figure 2.2). These enzymes have

been isolated in Citrus (Raymond and Maier 1977; Mclntosh et al. 1990;

Lewinsohn et al. 1989 a,b; Bar-Peled et al. 1993), but the genes encoding them
have not. The most obvious strategy for altering an end product might be to

manipulate the genes for the enzymes furthest down the biosynthetic pathway of

the product. However, although various sugar transferases are common

enzymes in plant species, it has proven difficult in general to clone the genes for
these enzymes.
In several crops, the levels of chalcone synthase (CHS) have been

successfully manipulated in transgenic plants utilizing both antisense and co-

suppression methods (Napoli et al. 1990; van der Krol et al. 1990a,b; Elomaa et

al. 1993; Courtney-Gutterson et al. 1994). Targeting CHS in these studies

resulted in a decrease or elimination of some anthocyanin compounds produced
later in the pathway. Therefore, our strategy was to effect anti-sense or sense
suppression of the CHS and CHI genes and determine if naringin production was

affected. Heterologous probes were available to clone these 2 genes for use in

transgenic experiments. As an initial step toward this goal, the CHS and CHI
genes in grapefruit were molecularly isolated and characterized.

Materials and Methods

RNA Blot Analysis

RNA was isolated from young leaves of a mature grapefruit tree (C.
paradisi cv. Marsh) as follows: Total nucleic acids were isolated as described by
Cone (1989). The RNA was selectively precipitated with 2M LiCI2, resuspended

in H20, and quantified spectrophotometrically (Sambrook et al. 1989). Poly (A)+
RNA was isolated using the PolyA tract kit (Promega). A gel containing 1.0 %
agarose, 1.1M formaldehyde, and morpholine propanesulfonic (MOPS) buffer
was loaded with 10 pg total RNA, 0.1 pg poly (A)+ RNA, and an RNA marker was
electrophoresed in MOPS buffer containing 50 jIg ethidium bromide for 3 hours

at 60 volts. The gel was photographed and blotted using nylon filter (Nytran,
Schleicher and Schuell) and 20X SSC in an overnight capillary transfer. The

RNA was immobilized to the filter using the autocrosslink setting on the
Stratalinker (Stratagene).

The CHS and CHI coding regionsfrom Antirrhinum majus were used as
probes in the Northern analysis. Both were obtained from Dr. Cathie Martin at

John Innes Institute, Norwich, U.K. The 2.0 kb (kilobase) CHS coding sequence

was isolated from a 5.7 kb genomic clone with a Kpnl-EcoRI restriction digest

and gel purified. The 0.9 kb full length cDNA clone of CHI was excised from
pUC 1813 with an EcoRI restriction digest and gel purified.
Probes were prepared with approximately 25ng of DNA and 50 jCi of
a32P-dCTP using the random primer method and manufacturers' instructions

(BRL). Spun-column chromotography was used to separate labeled DNA from
unincorporated nucleotides (Sambrook et al. 1989). After hybridization, the filter

was rinsed in a small volume of 1X SSPE, 0.2% SDS, and washed twice with
0.25X SSPE, 0.2% SDS for 45 minutes at room temperature. A series of
washes took place at 420, 580, and 650 for 45 min to determine stringency levels.
Filters were autoradiographed for 2 days with an intensifying screen at -700C.

The probe was removed from the filters by washing 3 times in 500 ml of boiling

0.1XSSPE, 0.2%SDS for 10 minutes.

A second RNA blot was prepared using RNA from leaves, stems, roots,
and flowers of grapefruit. All parts except flowers were obtained from 3 month
old cv. Duncan seedlings that were grown in the greenhouse. Flowers were
taken from mature cv. Marsh trees. RNA was extracted using the TRIZOL
method (BRL). RNA was quantified with a spectrophotometer (Beckman) and
approximately 10 pg run on a gel and blotted as described earlier. RNA probes

for the CHS and CHI genes were prepared using an in vitro transcription system
(Promega Riboprobe). Using T3 RNA polymerase, 50 pCi a32 P CTP, and

linearized citrus cDNA for CHS, an antisense RNA probe was transcribed.
Hybridization and washing conditions were as described earlier. The blot was
stripped and an antisense RNA citrus CHI probe prepared in the same manner
was applied. Finally, the Northern blot was probed with CHS cDNA from Citrus
using standard procedures, removed, and probed with CHI cDNA.

DNA Blot Analysis

DNA for genomic Southern analysis was extracted from leaf tissue using
the method of Cone (1989). To remove excess polysaccharides, a high salt

(1.5 M) precipitation was substituted for the final ethanol precipitation. DNA
concentration was estimated using ethidium bromide stained gel electrophoresis.
Approximately 5 tg DNA was cut with appropriate restriction enzymes using the

manufacturer's buffers. The restricted DNA was electrophoresed through 0.8%
agarose gels in TAE buffer and transferred to Nytran membranes (Schleicher

and Schuell). The DNA was immobilized on the membrane by UV crosslinker
(Stratagene). Probes were prepared as described above. Filters were washed

briefly with 1X SSPE, 0.2% SDS at room temperature, then 2 times with 0.25X
SSPE, 0.2% SDS for 45 minutes at 650C.

Construction of cDNA Library

Approximately 5 tg mRNA isolated from the Marsh leaves described

above was utilized in the library construction using a ZAPII-cDNA synthesis kit
(Stratagene). The cDNA was size fractionated through a Sephacryl-400 spin
column. The fraction containing the largest size cDNA was precipitated and
ligated into UNI-ZAPTM Vector arms. The library was packaged using the

Gigapack II packaging extracts and instructions (Stratagene). Packaged ligation
product was plated and titered using XL1-Blue MRF'cells at OD600= 0.5.

Screening of cDNA Library

The primary library was plated on ten large 150mm NZY plates to 50,000
pfu/plate with 600 jl of OD600 = 0.5 host cells/plate and 6.5 ml top agar/plate

following the Stratagene instructions. Duplicate lifts were made with nylon filters

(Magna Graph, MSE), and denatured for 2 minutes in 1.5 M NaCI and 0.5 M

NaOH, neutralized for 5 minutes in 1.5 M NaCI and 0.5 M Tris-HCI (pH 8.0), and

rinsed for 30 seconds in 0.2 M Tris-HCI, pH 7.5, and 2x SSC, and were
immobilized with a Stratalinker (Stratagene).

The filters were prehybridized as described for the Northern analysis,
except the 20 filters were placed in a round dish slightly larger than the filters
themselves. Due to the large number of filters used, the equivalent of 2 probes
was added (50 ng DNA labeled with 100 pCi of a 32P-dCTP) to fresh

hybridization buffer for the primary screening. The hybridization was 18 hours at
650. Filters were washed briefly with 1X SSPE, 0.2% SDS at room temperature,
then 2 times with 0.25X SSPE, 0.2% SDS for 45 minutes at 650C. The filters

were exposed to x-ray film with an intensifying screen for 4 days at -700 C.

The strongest hybridization signals on film were oriented to the
corresponding agar plates. An agarose plug containing the putative clone was
removed using the large end of a pasteur pipet and placed into a 1.5 ml
eppendorf tube containing 1 ml SM buffer and 201l chloroform. The tube was
vortexed and phage allowed to diffuse from the agar overnight at 40 C. For the

secondary screening, the phage was diluted and titered with host cells to obtain
two 100 mm NZY plates containing -450 plaques on one plate and -50 on the
other. A third round of screening was completed as above to obtain pure
plaques. These plaques were placed in 500 jlI SM buffer and 20 pl chloroform,
vortexed and kept at 40C.

The pBluescript phagemid was removed from the Uni-ZAP XR vector
using the in vivo Ex-Assist-SOLR system (Stratagene). Colonies resistant to
ampicillin (50mg/L) were grown up for plasmid mini-preps using an alkaline-lysis
method (Sambrook et al. 1989). Restriction digests were performed using Xhol
and EcoRI digests to screen for cDNA inserts.

DNA Sequence Analysis

Colonies selected for sequencing were grown in LB medium with 50 mg/L
ampicillin, and mini preps completed using an alkaline lysis PEG precipitation
method (Paithankar and Prasad 1991). Plasmids were sent to the University of
Florida and Iowa State University DNA Sequencing Core facilities for sequence
analysis using Applied Biosystems automated sequencers. Sequence analysis
was completed using the "GCG" Genetics Computer Group Sequence analysis
software package (Devereux et al. 1984). The 5' and 3' sequences of the
putative CHS and CHI clones were obtained using T7 and T3 primers from the
sequencing core which anneal to sites on the pBluescript polylinker. Additional
primers were designed using internal gene sequence (Table 3.1).

Results and Discussion

Before construction and screening of a cDNA library began, a Northern
blot containing RNA obtained from citrus leaf tissue was probed using the
Antirrhinum probe for CHS (Figure 3.1). One band of approximately 1.4 kb was
seen in the total RNA and poly (A)+ lanes. The results from the blot indicated
that the mRNA isolated was intact and of good quality for use in library
construction, the DNA probes and hybridization conditions would be satisfactory
for the library screening, and it should be possible to isolate the transcripts.
Similar results were obtained using the Antirrhinum CHI gene probe, however
the bands were slightly smeared making them difficult to size. A cDNA library
was then constructed from the leaf mRNA and screened with the CHS and CHI

Chalcone Synthase

Two of the clones believed to be the longest in length were selected from
the library screening for sequencing and further analysis. Partial sequence of
700 nucleotides from the 2 clones was identical, indicating that these 2 inserts

were probably the same clone; thereafter only 1 was sequenced to completion.
This CHS clone was 1349 nucleotides in length and contained an open reading
frame of 391 amino acids (Figure 3.2). Alignment of the predicted amino acid
sequence with known CHS proteins (see below) suggested that the clone

contained the entire coding sequence. The predicted molecular weight of the
amino acid sequence analyzed by the Peptide Sort program of GCG was 42.6

A BLAST (Basic Local Alignment Search Tool) (Altschul et al. 1990) was
utilized for searching sequence databases with the Citrus CHS sequence. The
results indicate that the cDNA we isolated was indeed Chalcone synthase
(Figure 3.3). Alignment of the C. paradisi CHS sequence with the Antirrhinum
majus sequence that was used as the heterologous probe showed 71%

nucleotide sequence identity and 83% amino acid identity. It also showed that
the first ATG of the citrus sequence lined up with the first ATG of Antirrhinum,
implying that the citrus cDNA was a full length clone. The highest amino acid
similarity was with Betula pendula (birch) at 85% identity.

Genomic DNA analysis using the citrus CHS cDNA probe showed multiple
bands, suggesting that CHS might be a multigene family with possibly 2 to 3
members (Figure 3.4). In other species, CHS gene families have been reported
to contain 1 (Herrmann et al. 1988) to 12 members (Koes et al. 1989).

Northern analysis was attempted on RNA extracted from flowers, leaves,
stems, and roots. An attempt to isolate RNA from peel and fruit was

unsuccessful. Due to nonspecific binding in the high molecular weight and
ribosomal RNA region we were not able to conclude expression patterns. The
blots were first probed with the RNA probes using the same stringency levels

used for screening our initial Northern blot with a DNA probe. Due to our
inexperience in using RNA probes, we did not anticipate the problems
associated with increased sensitivity and nonspecific binding. The use of an

RNA probe may require different hybridization and washing conditions (Murray et
al. 1991). The use of cDNA probes continued to display some binding to the
ribosomal and high molecular weight region (Figure 3.5). It was difficult to

determine if this was irreversibly bound from previous RNA probe, an artifact
from the labeling process, or was associated with the cDNA. A ubiquitous
expression pattern among tissues would not be surprising considering the
diversity of flavonoid derived compounds throughout the plant and what has
been reported for other species (McKhann and Hirsch 1994). If this is a multi-

gene family and additional CHS gene members are sequenced and cloned,
future studies can determine if particular members show tissue specific
expression patterns as has been shown in petunia and maize (Koes et al. 1989;

Chalcone Isomerase

One clone was selected from the library screening and analyzed.
Sequence analysis indicated that the clone was 862 nucleotides in length and
contained an open reading frame of 222 amino acids (Figure 3.6). The predicted
molecular weight of the amino acid sequence analyzed by the Peptide Sort

program of GCG was 23.9 kilodaltons.

A BLAST (Basic Local Alignment Search Tool) (Altschul et al. 1990) was
utilized for searching sequence databases with the Citrus CHI sequence. The

results indicate that the cDNA we isolated was indeed Chalcone isomerase

(Figure 3.7). Alignment of the Citrus paradisi CHI sequence with the A. majus
gene (used as the heterologous probe) showed 71% nucleotide sequence
identity. The highest amino acid similarity was with Vitis vinifera (table grape) at

76% identity.

Genomic DNA analysis utilizing the citrus CHI cDNA probe showed
multiple bands, indicating that, like CHS, CHI appears to be a multigene family
(Figure 3.8). This is also similar to what has been described in other species

(van Tunen et al. 1989)

Northern analysis was completed for CHI expression reprobing the same
blot containing tissue from plant parts described above. Unfortunately the
probing was done in the same fashion as CHS which resulted in non-specific
hybridization. Again, this precludes making conclusive statements about tissue
expression (Figure 3.9). Considering that CHS and CHI encode successive
steps in the same pathway, these genes may show similar patterns of
expression. Isolation of additional gene family members would provide useful

information to determine gene specific expression in tissues. Specific sequences
could then be used in future transgenic experiments to inhibit the flavonoid
pathway in certain tissues.

Table 3.1. Primers used to sequence CHS and CHI cDNAs from Citrus paradisi.



* The Position category refers to the nucleotide base pair followed by the strand
(5' or 3') on which the sequence is located. All base pair counts start at the 5'

2.4 kb

Figure 3.1. RNA Northern blot analysis of grapefruit. The probe was a 2.0 kb
fragment of the Chalcone synthase Antirrhinum majus genomic clone. Lane 1,
RNA marker; lane 2, total RNA; lane 3, poly (A)+ RNA.





























































































































Figure 3.2. Nucleotide sequence for a CHS cDNA isolated from C. paradisi.

A=Adenine,T=Thymine,C=Cytosine,G=Guanine, K= Guanine or Thymine.



Sequences Producing High-scoring Segment Pairs:

emb Z38096 GHCHS1
embZ 67988 CCCHSMR
emb X94706 JSPCHS2
emb X94995 JSPCHS1
emb X75969 VVCHS
dbj AB0018261AB001826
dbj AB002582 AB002582
dbj AB002815 AB002815
emb X14314 SASCHS3
emb X17577 MICHSY
emb Z340881RGACRSYN
embI 380981GHCHS3
emb|X58339 X58339
emb X89859 OSCHALSYN
gb S80554 S80554
emb V01538 PHCHAL

B.pendula Roth mRNA for chalcone sy...
G.hybrida mRNA for chalcone synthase
C.chinensis mRNA for chalcone synthase
Juglans sp. (J.nigra x J.regia) mRN...
Juglans sp. (J.nigra x J.regia) mRN...
V.vinifera CHS mRNA for chalcone sy...
Ipomoea purpurea mRNA for chalcone ...
Perilla frutescens mRNA for chalcon...
Perilla frutescens mRNA for chalcon...
Mustard mRNA for chalcone synthase ...
M.incana mRNA for chalcone synthase...
R.graveolens (AIG) mRNA for acridon...
G.hybrida mRNA for chalcone synthase
H.vulgare CHS gene for chalcone syn...
O.sativa mRNA for chalcone synthase
(tt4(85))=chalcone synthase [Arabid...
Parsley messenger RNA coding for ch...

Figure 3.3. BLAST search results obtained when the cDNA putative CHS

sequence isolated from C. paradisi was aligned with nucleotide sequences of

known genes. This is a partial list showing only the highest matches.



Score P(N)

^ *-~i


5.0 kb*t i [


Figure 3.4. Genomic DNA blot analysis of grapefruit probed with the grapefruit
CHS cDNA. Lane 1, 1 kb molecular weight marker; lane 2, BamHI digest; lane 3,
EcoRI digest; lane 4, Hind III digest; lane 5, Sail digest.

4 0

3 4 5

2.4 kb





Figure 3.5. RNA blot analysis of total RNA from various grapefruit tissues
probed with the grapefruit CHS cDNA. Lane 1, RNA molecular weight marker;
lane 2, flower; lane 3, leaves; lane 4, stems; lane 5, roots.



































Figure 3.6. Nucleotide sequence determined for a CHI cDNA isolated from C.

paradisi. A=Adenine,T=Thymine,C=Cytosine,G=Guanine.


















Sequences Producing High-scoring Segment Pairs:

emb X75963 VVCHI
emb Z67989 DCCHIMR
embIZ67980 CCCHIMR
emb Y00852 PHCHI
emb X14589 PHCHIA
emb X68978 MSCHI10
emb X68979 MSCHI5
dbj D63577 PUECHI
emb X16470 PVCHALIS
gb M91079 ALFMSCHI1
gb M91080 ALFMSCHI2
emb Z15046|PVCHISO

V.vinifera CHI mRNA for chalcone isom...
D.caryophyllus mRNA for chalcone isom...
C.chinensis mRNA for chalcone isomerase
Petunia hybrid mRNA for chalcone fla...
Petunia CHI-A gene for chalcone flava...
Antirrhinum majus chalcone isomerase ...
Malus sp. CHI mRNA for chalcone isome...
Malus sp. CHI mRNA for chalcone isome...
Pueraria lobata mRNA for chalcone fla...
Phaseolus vulgaris gene for chalcone ...
Medicago sativa (cultivar Iroquois) c...
Medicago sativa (cultivar Iroquois) c...
P.vulgaris gene for chalcone isomerase

Figure 3.7. BLAST search results obtained when the cDNA putative CHI
sequence isolated from C. paradisi was aligned with nucleotide sequences of
known genes. This is a partial list showing only the highest matches.



Score P(N)


5kb --

Figure 3.8. Genomic DNA blot analysis of grapefruit probed with the grapefruit
CHI cDNA. Lane 1, BamHI digest; lane 2 EcoRI digest; lane 3, Hind III digest.

3 4 5

2.4 kb-

Figure 3.9. RNA blot analysis of total RNA from various grapefruit tissues
probed with the grapefruit CHI cDNA.. Lane 1, RNA molecular weight marker;
lane 2, flower; lane 3, leaves; lane 4, stems; lane 5, roots.

A rII.



The bitter flavor of naringin can result in undesirable or unacceptable fruit
and juice quality (Hagen 1966). One approach to reducing the levels of naringin
is to reduce the expression levels of genes encoding biosynthetic enzymes of
the naringin pathway. A transgenic approach is possible whereby the
introduction of genes in the sense or antisense orientations might modify the
level of endogenous gene expression. CHS and CHI participate in the first 2
steps of the flavonoid biosynthetic pathway, leading to the production of naringin
(Bar-peled et al. 1993). As described in Chapter 2, Agrobacterium-mediated
transformation is currently the most efficient way to produce transgenic Citrus
plants. CHS and CHI cDNAs isolated from Citrus paradisi Macf. (grapefruit) (see
Chapter 3) were cloned into Agrobacterium transformation vectors in sense and
antisense orientations and introduced into C. paradisi. Regenerated plants were
evaluated for evidence of transformation, gene expression, and naringin content.

Materials and Methods

Transformation Plasmids

Four transformation plasmids were constructed, CHS and CHI each in
both the sense and antisense orientations: CHS-S (sense), CHS-A (antisense)
and CHI-S (sense), CHI-A (antisense). A system developed by Slightom (1991)

was chosen for transformation on the basis of its availability and demonstrated

utility in Citrus (Gutierrez-Espinosa 1995). A paired system, it utilizes pUC18-

expl7 and pGA482/GG plasmids. The former contains an expression cassette

that includes the CaMV 35S promoter, a 70 nucleotide 5' translational enhancer

from the cucumber mosic virus, an Ncol cloning site, and a CaMV

polyadenylation signal (Figure 4.1). The plasmid vector pGA482/GG contains

right and left T-DNA borders. Between the borders, there is a neomycin

phosphotransferase resistance gene (NPTII) to confer kanamycin resistance for

plant selection, a polylinker site for inserting the pUC18exp-17 cassette or other

DNA, and a CaMV 35S driven GUS reporter gene. Outside the T-DNA borders

there is a gentamicin resistance gene for growing the plasmid in Agrobacterium

cell lines, and a tetracycline resistance gene for growth in E. coli cell lines

(Figure 4.1).

The 1.4 kb CHS and 0.9 kb CHI cDNAs isolated from C. paradisi were
excised from the Uni-ZAPTM vector with an EcoRI and Xhol combined restriction

digest according to the manufacturer's instructions (BRL). The gene fragments

were gel purified and recovered with the Prep-a-Gene system (Bio-Rad). The

ends of the inserts were blunted using a fill-in procedure (Sambrook et al. 1989),

and buffers and enzymes were removed with the Prep-a Gene system. The

vector pUC18-expl7 was linearized for subcloning with Ncol and blunt-ended as

above. To prevent recircularization of the plasmid, it was treated with calf

intestinal alkaline phosphatase (CIAP). The CIAP was removed using the Prep-
a-Gene system.

A method for blunt-end ligations was followed using 1:1 and 1:3 molar
ratios of vector: insert (Damak and Bullock 1993). Thirty ng of each ligation
mixture was transformed into competent JM 83 bacterial cells using the heat

shock procedure (Sambrook et al. 1989) and the cells were plated on LB plates

containing 100 mg/L ampicillin. Colonies were screened for inserts by growing in
LB medium with 100 mg/L ampicillin using a quick check method (Akada 1994).

Mini-preps were completed on the plasmids containing inserts, and antisense
and sense orientations were identified. A Hindlll restriction digest was used to

determine the orientations of the CHS inserts, while Scal was used for CHI

Using the system of Slightom (1991), the next step was to cut the
expression cassette containing the gene of interest from Pucl8-expl7 with

Hindlll. The CHS cDNA contained an internal Hindlll site approximately 150 bp
upstream from the 3' end and required a partial digest to avoid cutting the

internal site. A dilution series of Hindlll digests was run on a 0.8 % agarose gel

to identify a correct partial digest. An enzyme dilution of 1:10 was selected and
scaled up to provide enough isolated DNA from the relevant band to be used for
cloning. The DNA was recovered from a 0.8% agarose gel using the Prep-a

Gene system. The vector pGA482/GG was linearized by cutting with Hindlll.
Ligations were set up in vector: insert ratios of 1:1 and 1:3 and transformed into
JM 83 as before.

Colonies were screened for inserts using PCR primers designed to
amplify portions of CHS and CHI. The CHS primers were
primers were 5'-GCACAATACTCAGAC-3' and 5-CCAGTACTGCCTCAG-3'.
Colonies were touched with the end of a sterile pipet tip and placed directly into
a PCR reaction mixture which contained 12.5 pl PCR Master (Boehringer), 15
pMoles of each primer, 10Opl H20 and a drop of mineral oil. The samples were
heated to 920 C for 2 min then subjected to 30 cycles of 2 min at 920 C, 2 min at
580 and 3 min at 720. A 10 l aliquot of the PCR reaction mixture was run on a

0.8% agarose gel to identify colonies with amplification products. Plasmid

mini-preps and restriction digests were done to confirm the results. Due to the

large plasmid size (>20 kb), and possibility of shearing, the alkaline lysis mini

prep procedure was replaced with a Wizard plasmid prep kit (Promega).
The 4 transformation vectors were then individually transformed into the

Agrobacterium strain EHA101 (Hood et al. 1986) using a heat shock method (An

1987). The disarmed strain of EHA101 carries the hypervirulent tumor inducing

plasmid pTiBo542 from which T-DNA sequences have been deleted. This

plasmid contains the necessary vir components for infection and integration into

the plant cell nucleus.

Plant Transformation

Seed coats were removed, and C. paradisi cv. Duncan seeds were

sterilized according to the method of Moore et al. (1992). After the seeds were

placed on germination medium, the tubes were were kept in the dark until the

germinated seedlings were used for transformation experiments (approximately

4-6 weeks).

The EHA101 Agrobacterium cultures were grown in YEP medium with 60

mg/L gentamicin and 25 mg/L kanamycin to OD62o=0.5-1.0. Cultures were spun

at 4000 x g for 10 min and the medium was decanted. The pellet was gently
resuspended in MS salts plus 100 pM acetosyringone to a cell density of 5 x 108-

5 x 1010 cells/ml (Lin et al. 1994).

The epicotyl portions of the etiolated seedlings were cut into 1 cm
segments as described by Kaneyoshi et al. (1994). The segments were soaked
in the Agrobacterium solution for 15 min minimum, blotted on sterile paper towels

to remove excess liquid, and placed horizontally on 100 cm petri plates
containing MS medium with 100lpM acetosyringone. The explants remained on

this medium for 2-3 days at 280 C to allow Agrobacterium infection to occur, then

were transferred to selection medium to induce shooting (MS containing 3mg/L
benzyladenine (BA), 50 mg/L kanamycin sulfate, 500mg/L Claforan, and 7g/L
Bacto-agar). The plates were maintained at 270C with 16 hours of cool-white

fluorescent light. When shoots appeared, the explants were transferred to MS

plates with the same level of kanamycin, but a lower level of BA (0.5mg/L).
Transfers to fresh medium were done every 3 weeks.

Regenerated shoots longer than 3mm were removed from the explants
and sections cut from their basal ends that were histochemically stained for GUS
(Moore et al. 1992). All shoots removed for GUS testing were placed on MS
medium without hormones. Those shoots that stained positive for GUS
remained on this medium for 3 weeks before transfer to MS medium with 5pM

NAA to induce rooting. Following a brief exposure to NAA (2 to 10 days), the
shoots were returned to hormone-free MS medium. Once roots developed, the

plants were transferred to soil cups that were sealed to ensure high humidity.
Following the appearance of new leaves, the plants were placed in peat pots

containing soil. Plastic bags covered the plants and the humidity was lowered
gradually over a week's time by increasing the number of holes cut in the bag.
After the bag was removed, the plants were moved to a growth chamber with 16
hours light at 300C and an 8 hour 280 dark cycle. The plants were retested for

histochemical GUS staining of the leaves approximately 4 months after the
transformations, and positive-staining plants were further analyzed.

PCR Analysis of Transformants

DNA was extracted from leaves using a high urea method (Cone 1989).
Approximately 5 ng DNA was added to a 25p1 reaction containing 15 pMoles of

each appropriate primer and 12.5 pl PCR Master (Boehringer). Primers for

amplification of a 350 bp NP71I gene fragment were

5'-CATCGCCATGGGTCACGACGA-3'. Primers used to amplify a 450 bp GUS
gene fragment were 5'-TTGGGCAGGCCAGCGTATCGT-3' and

5'-ATCACGCAGTTCAACGCTGAC-3'. The reactions were run as described
previously. Approximately 10 pl of each reaction was loaded on a 0.8% or 1.2%

agarose gel, which was run, stained with ethidium bromide, and photographed to
confirm absence or presence of bands.

DNA Blot Analysis

DNA for genomic Southern analysis was extracted from 0.1-0.3 g of leaf
tissue by scaling up the procedure used for PCR. To remove excess
polysaccharides, a high salt (1.5 M) precipitation was substituted for the final
ethanol precipitation. DNA concentration was estimated using gel

electrophoresis and spectrophotometric quantitation (Sambrook et al. 1989).
Approximately 5 jg DNA per sample was cut with Hpal (BRL) using the

manufacturer's buffer. The samples of approximately 5 pg were electrophoresed

through a 0.8% agarose gel in TAE buffer and transferred to a Nytran membrane
(Schleicher and Schuell). The DNA was immobilized onto the membrane by UV

crosslinking in a Stratalinker (Stratagene).

A 1.8 kb GUS gene fragment for use as a probe was excised from pBI221
(Clontech) with an Xbal/Sstl restriction digest. The probes were prepared using
25ng DNA, 50 pCi a32P-dCTP (NEN) and a random primer kit (Stratagene). The

blots were stripped and re-probed with their respective cDNA inserts that were
included in the transformation vector. High stringency hybridizations were used

for all Southern blots (overnight at 650) as described in chapter 3.

Autoradiographs were exposed at -700 C.

RNA Blot Analysis

RNA was extracted from 0.1g of leaf tissue using the Trizol reagent (BRL)
according to the manufacturer's instructions. The RNA was quantified using a
spectrophotometer, and approximately 5.g total RNA used for northern analysis.

The RNA was electrophoresed through gels containing 0.8% agarose and 2.63%
formaldehyde, blotted onto Nytran membranes (Schleicher and Schuell) and

immobilized by UV crosslinking in a Stratalinker (Stratagene).

The blots were first probed with the 70 bp CMV 5'untranslated region to
test expression of the transgene. The probe was prepared by substituting 2

CMV primers for random primers in a random primer labeling reaction using 50

(iCi a32 P-dCTP. The CMV primers were 5'-CATTTGGAGAGGACAGGGTA-3',
and 5'-TTTCAGCGCAACCAT-3'. Following a 30 minute labeling reaction at
370C, the labeled DNA was purified using a spin column method (Qiagen).

Hybridization and washes were at 600 C. Following development of the

autoradiographs, the CMV probe was removed and RNA probes for the CHS and
CHI genes were prepared using an in vitro transcription system (Promega
Riboprobe). Using T3 RNA polymerase and 50 pCi a32 P-CTP, an antisense

strand was generated from the cDNA that was expected to hybridize to both the
endogenous and transgene RNA. Following removal of the antisense strand
probe, the blot was probed with the respective cDNA used in the transformation.

Protein Blot Analysis

Total protein was extracted from frozen Citus leaf samples (0.1 g) by
grinding in buffer (2 ml) with a mortar and pestle. The extraction buffer contained

50 mM Tris pH 8.0, 150 mM NaCI, 1 mM EDTA, 1 mM DTT, and 0.2% SDS. The
samples were spun briefly at high speed in a microfuge and supernatants were

transferred to new tubes. The samples were boiled for 10 minutes. A protein
microassay (Bio-Rad) was performed to determine total protein concentrations.
The protein samples were adjusted to contain 40tg of protein in 40 pl of

extraction buffer. Loading buffer was added and the samples boiled for 5
minutes. Parsley and petunia were extracted as above and used for positive
controls with CHS and CHI respectively. The samples were loaded on a 12.5%
acrylamide mini gel (Biorad) and run for one hour at 200 volts. The gel was

electroblotted to a nitrocellulose membrane following the manufacturer's
directions (Bio-Rad). The membrane was blocked for one hour using a blocking
solution that contained phosphate-buffered saline solution, bovine serum

albumin, and Tween (1X PBS, 5% BSA, 0.05% Tween) (Sambrook et al. 1989).

The primary antibody consisted of either a CHS polyclonal antibody made from
parsley given to us by Dr. Klaus Halhbrock (Max Planck Institute), or CHI
antiserum produced from petunia protein donated by Dr. Joseph Mol (Free
University, Amsterdam). The primary plant antibody was diluted 1:1000 in fresh

blocking solution and allowed to incubate with the blot for 1 hour. The blot was

washed 3 times for 15 minutes using 1X PBS, 5% BSA, 0.05% Tween. The
secondary antibody anti-rabbit IgG (whole molecule) was diluted 1:20,000 in
blocking solution according to the manufacturer's instructions (Sigma). Following
a 1 hour incubation, the blot was washed 3 times in Tris-buffered saline solution
and rinsed in 150 mM NaCI, 50mMTris-CI (pH7.5). Immuno-localized alkaline

phosphatase detection was completed using BCIP/NPT (5-Bromo-4 chloro-3
inoyl phosphate/ Nitro blue tetrazolium stain (Sigma) (Sambrook et al. 1989).

Naringin Analysis

Leaves from transgenic and control plants were dried in paper bags
overnight at 600 C. A leaf sample weighing between 0.1-0.2 g was ground in a

mortar and pestle with sand. The sample was transferred to a test tube 0.5 ml

of a 1:1 DMSO:methanol mixture was added, and the sample was placed in a
500C water bath for 30 minutes. After removing the extract, the procedure was

repeated 3 additional times. The 4 extractions from each sample were combined
and filtered through a 0.45p filter (Nalgene). The samples were sent to Dr. Mark

Berhow at the USDA in Peoria Illinois who performed the naringin assay with
HPLC analysis. Dr. Berhow used a C-18 RP 5p Licrosorb column (250 x 4.5

mm) equilibrated at 20% methanol in 0.01M phosphoric acid. The gradient was
run to 100% MeOH over 50 minutes and 1001l of sample injected. Naringin

typically eluted from the column at approximately 25 minutes. The areas of the
peaks were converted to mg/g dry weight.

Results and Discussion

Regeneration and initial screening of transformants

A series of Agrobacterium-mediated transformation experiments were
performed using the 4 constructed transformation plasmids: CHS-S, CHS-A,
CHI-S, and CHI-A. All 4 constructs used resulted in successful transformation
events with regenerated shoots that stained positive for GUS (Table 4.1). An

individual experiment with the CHI-A construct produced the highest number of

surviving regenerated shoots. Many of the shoots regenerated from explants

using the CHS-A construct initially began to regenerate shoots that were smaller
than the control plant shoots, grew slower, had abnormal leaf curling and other

morphological anomalies, and died before they were large enough to
histochemically test for GUS. Although this could be associated with kanamycin

selection or the tissue culture process, this abnormal development was only
observed with the CHS-A construct. The possibility that the negative effect

occurred due to the antisense gene should be considered when using this
construct in future studies. It is enticing to speculate that these effects might

reflect the biological importance of flavonoids to Citrus.

Regenerated shoots were removed approximately 6-8 weeks after the
transformation date for histochemical GUS staining. The number of blue staining
shoots in an individual experiment ranged from 33-66% of all shoots examined

(Table 4.1). The majority of sections examined showed chimeric staining,
ranging from a few blue dots to sectors that were all blue. A total of 251 shoots

examined from the 4 constructs yielded 41% with tissue that stained at least 25%

blue. Only 11% of all shoots examined yielded sections that stained entirely
Low rooting efficiency has been reported as a major problem for in vitro
production of Citrus plants (Duran-Vila et al. 1989). Thus, rooting of transformed

Citrus continues to be a problem (Moore et al. 1992, Pefa et al. 1995a; b). The
recent reported success of micrografting Citrus transgenic shoots (Pefa et al.
1995b) may allow that technique to replace our current methods of rooting in soil
or medium for transformants used in scion production. This technique was not
routinely done in our lab and we did not use it for this study, as the probability of

plant death due to incorrect grafting seemed high.

Because of the high frequency of chimeric GUS staining (Table 4.2),

rooting was not attempted on medium containing kanamycin. After the shoots

were removed from the explant for GUS histochemical staining, they were placed

on antibiotic and hormone-free MS medium. Earlier reported studies returned

the excised shoots to the original shooting medium which contained kanamycin

until the results of the histochemical staining were completed. However,

because the exposure of the cut shoot to kanamycin (albeit brief), might interfere

with subsequent rooting, we chose to avoid it. The GUS positive shoots

remained on hormone-free MS for a 3 week period prior to being placed on an

auxin containing medium. This period of time may have diluted some effects of

BA from the shoot and made the exposure to NAA more effective. This rooting

method was developed using the putative transformants and changes were

made to the procedure during the course of this study. Initially the shoots were
placed on a high level of NAA (5iM) for 3 weeks. When only a few shoots

developed roots, another approach was tried. Fresh cuts were made on the
basal ends of the stems and the plants were exposed to 5pM NAA for 2-5 days,

then transferred to MS medium without hormones. The second method resulted

in root development within a 3 week period. Some of the plants required a

second 3 day exposure to NAA before rooting. Clearly this method requires

further study for optimization. However, roots appeared as early as one week

after exposure to NAA containing medium, thereby saving several month's time

and providing a higher survival rate with less effort than the soil rooting


Of the 260 shoots that were GUS positive, 84 survived the regeneration
process with continued shoot growth and root development. Many of the rooted

plants did not grow well in sealed soil cups in the growth room. The low light

levels from fluorescent bulbs in the tissue culture growth room, lower

temperature (26-280), and opaque lids on the Magenta boxes may have

contributed to poor plant growth and death. An alternative procedure was used
on the last group of 18 plants to regenerate roots. The rooted plants were
placed directly into small peat pots with soil, covered with plastic wrap and
placed in a growth chamber that contained a richer spectrum of light (fluorescent
and incandescent), adjustable lighting, and a higher temperature (300C). All 18

plants survived, and the plants grew faster and appeared healthier than those
transferred to soil cups in the growth room.

GUS histochemical staining was repeated on 66 surviving putative
transformed plants 4-6 months after the initial staining. Twenty seven of the
plants (40%) continued to stain blue. Two of the plants continued to have
chimeric staining and the 25 others stained entirely blue (Table 4.3). In order to
achieve a higher percentage of plants that continue to express GUS, less
chimeric and more solidly transformed plants are desired. This requires a better
selection system. Higher levels of kanamycin (100mg/L) have been used
successfully in Citrus (Kaneyoshi et al. 1994, Moore et al 1992; Peia et al.
1995) for selection. However, Pefia et al. (1995a; b) described morphological
abnormalities associated with an extended period of selection (8 weeks). One
approach to decreasing the number of chimeric transformation events and
avoiding plant abnormalities might involve selection on a higher level of
kanamycin for an initial period of 1 week or less, followed by transfer of the
explants to a lower level (50mg/L). By evaluating the results of histochemical
staining, different selection schemes could be proposed or improved.

Evaluation of Transgenic Plants

Polymerase chain reaction
PCR analysis was completed on 27 plants staining positive for the
histochemical GUS test. All plants amplified the expected size NPTII 350 bp
fragment and GUS 450 bp fragment except one, CHI-S1, which continued to
exhibit chimeric staining for GUS (Figure 4.2).

DNA blot analysis
Southern analysis was completed on those plants that were large enough
to take leaf samples and were positive for PCR analysis of GUS or NPT II. Hpal
was selected as the restriction enzyme of choice because neither of the cDNAs
used for transformation contains this restriction site, and the enzyme cuts once in
the polylinker portion of the pGA482GG T-DNA region (Figure 4.1). Hybridizing
bands should thus represent T-DNA and flanking citrus DNA sequences. The
length of the fragments would vary with the nearest Hpal site flanking the
integrated T-DNA. Probing the Southern blots with the GUS gene showed 1-3
copies of the gene integrated in both the CHS and CHI transformants (Figures
4.3 and 4.4). All bands were greater than the minimum expected 4.5 kb border
fragment. A few lanes showed bands only in the high molecular weight uncut
region. After examining the DNA gel photographs it was concluded this might be

caused by incomplete digestion of the genomic DNA. All plants that amplified
GUS or NPTII fragments with PCR analysis were positive when probed with GUS
on the Southern blot with one exception (CHI-S1). This particular plant was also
one that continued to exhibit chimeric staining for GUS (Table 4.3). Three other
plants (CHS-S3, CHI-A2, and CHI-S5) did not have GUS hybridizing bands on
the Southern blot, but this could be explained by the apparent low levels of DNA

loaded as shown in the gel photograph. This underloading of DNA was probably
caused by inaccurate spectrophotometric readings.

Multiple bands were produced when the Southern blots were probed with
the citrus cDNA fragments for CHS and CHI. These bands represent the

transgenes and endogenous copies. The cDNA probes used were not designed
to be gene specific, therefore endogenous bands may also represent members

of multigene families (Figures 4.5 and 4.6). The endogeneous bands appeared
darker than many of the transgene bands. A possible explanation is that if the
Hpal restriction sites were conserved among the members of the gene family,
multiple copies might only result in a more intense band on the Southern blot,

rather than multiple bands. Both blots showed between 2-5 bands with
transformant CHI-A6 having five bands. The control lanes showed one band for
CHS and 2 bands for CHI representing endogenous gene family members.

RNA blot analysis

Northern analyses were performed with CMV 5' untranslated region, CHS,
and CHI probes. The CMV 5' untranslated region probe was designed to tag the
transgene because the fragment was transcriptionally fused to the CHS and CHI
transgenes and not otherwise present in citrus. Probing 6 CHS transformants

revealed expression in 5 plants (all except CHS-A4). However plants CHS-A5,
CHS-S1, and CHS-S3 gave stronger signals at approximately 1.4 kb (Figure 4.

7). Probing the blot with an RNA antisense CHS probe to determine sense
expression resulted in nonspecific hybridization or other product (such as
carbohydrate), and could not reliably used for analysis. In retrospect poly A+
RNA should have been used for the template to eliminate the very high
backgrounds with nonspecific hybridization to rRNA when using total RNA

(Murray 1991). We thought we did not have enough tissue to isolate Poly (A)+

RNA, however the increased sensitivity of an RNA probe used on even a small

amount of poly A+ RNA usually improves the signal-to-noise ratio and may have

been a better approach. The CHS cDNA probe recognized both transgene and
endogenous transcripts and resulted in multiple transcripts expressed. The
transcripts appeared in the range of 0.8-1.4 kb but there still appeared to be
some non-specific hybridization occurring especially in the high molecular weight

region of approximately 3 kb. This may have been caused by the inability to

completely strip off the signal from using the RNA probe, or sub-optimal
hybridization and washing conditions (Figure 4.8). We were not able to make
any quantitative comparisons because the control RNA appears somewhat

Among the 10 CHI transformants analyzed, one antisense plant (CHI-A2)
and three sense plants (CHI-S2, CHI-S3, and CHI-S4) did not appear to express
the transgene, while the others gave a strong signal at 0.9kb (Figure 4.9).
Probing the blots with the CHI cDNA probe (Figure 4.10) revealed a consistent,
though not totally explainable pattern. Those plants which did not express the
transgene (including the control) all appear to produce 2 transcript bands, one

slightly larger than 1.0 kb, and one slightly less than 0.9kb, possibly representing

different transcripts associated with multigene family members. Those plants
that did express the transgene have what appears to be one transcript of the
same size as detected by the CMV probe (0.9). We are unsure at this time
whether the multiple bands represent multigene transcripts or artifacts resulting
from the incomplete removal of the RNA probe or non-optimization of the
hybridization process. The attempt to quantify the amount of RNA loaded in

each lane was unsuccessful, and resulted in some unequal loading as shown in
the gel. Further analysis would be necessary to conclude that there were
differences in expression levels relative to the control plant.

Protein blot analysis

Attempts were made at quantifying protein expression by western blots
using antibodies. Different blocking solutions and dilutions of antibodies were

tested, but non-specific binding of the primary antibodies occurred in the citrus
lanes giving multiple bands (data not shown). Bands appeared in the predicted

molecular weight region, as well as other regions. No non-specific binding
occurred in the positive control lanes for parsley (CHS) and petunia (CHI).
Attempts to produce a translation product from Citrus to confirm a positive citrus
band corresponding to the CHS and CHI cDNAs were unsuccessful, and the

analysis was discontinued. Future work in this area ideally would include the use
of antibodies to CHS and CHI made from Citrus proteins.

Naringin analysis
Twenty three transgenic and 4 control plants were analyzed by Dr. Mark
Berhow for naringin content using HPLC. Sampling was done on CHS-A1, A2,
A3, and A4, but not included because the plants were much younger and only
tiny new leaves were available on the plants. The results of the naringin analysis

ranged from 0.2 to 8.89 mg/g dry weight for the 19 transgenics (Table 4.1). The
range of naringin levels found in the control plants ranged from 1.8-8.89 mg/g
with a mean value of 4.3 and standard deviation of 2.35. No mean values were
computed for the transgenics because grouping individual transformation events
seemed inappropriate.

The lowest naringin level found (0.2 mg/g) was from the CHI-A6 plant.
This plant had multiple bands (4-5) on the Southern blot when probed with CHI
cDNA (Figure 4.6). Unfortunately, multiple attempts at isolating RNA for a

Northern blot from this plant were unsuccessful and no gene expression data is
available. All antisense plants analyzed had naringin values below the control
mean of 4.35 mg/g. The small number of both control and transgenic plants

analyzed make it impossible to statistically conclude there were any differences

in naringin due to the treatment. To remedy this, future experiments should
include a larger number of samples (both control and transgenic) to analyze.
This could result in a lower standard deviation value, and greater probability of

detecting a desired change among the transgenics. The transgenic plants
should also be micropropagated to include replications of each transformant so
that a mean value for each plant could be calculated, and statistical analysis be

Due to the time limitations associated with the project, sampling leaves for
naringin analysis was made while many of the plants were less than 5 months
old. Sampling young plants for naringin should be avoided if possible because
flavonoid concentrations in developing leaves can vary considerably.
Flavonoids are synthesized and accumulated at relatively high concentrations
during the cell division stage of new leaf development with little or no flavonoid

synthesis during the cell elongation stage of leaf growth and none in mature
leaves (Dr. Mark Berhow personal communication). The tiny leaves have high

concentrations of flavonoids which can get diluted 10-20 fold or more during the
course of leaf expansion. How old the leaf is and where it is located on the plant
relative to other leaves is important when comparing flavonoid leaves from one
plant to another.

While an attempt was made to select similar sized leaves from the same
areas of the plant, the plants were not of the same age. While some attempt at
quantifying naringin levels for the plants in this study was made, these plants
should be retested again after the period of leaf growth has passed.

Table 4.1. Agrobacterium-mediated transformation experiments completed on
grapefruit cv. Duncan.


CHS-A 7/18/95 440 79 40 (50.6)
CHS-A 7/19/95 160 31 12 (38.7)
CHS-A 10/06/95 340 30 18 (60.0)
CHS-A 10/21/95 200 12 4 (33.3)
CHS-A Total 1140 152 74 (48.7)

CHS-S 7/21/95 240 37 21 (56.8)
CHS-S 10/06/95 180 31 14 (45.2)
CHS-S Total 420 68 35 (51.5)

CHI-A 6/4/95 340 239 91 (38.1)
CHI-A 6/5/95 140 24 16 (66.7)
CHI-A Total 480 263 107 (40.7)

CHI-S 7/20/95 260 63 27 (42.9)
CHI-S 10/06/95 220 51 17 (33.3)
CHI-S Total 480 114 44 (38.6)

Table 4.2. Results of GUS histochemical staining of regenerated grapefruit
shoots following Agrobacterium-mediated transformation.

SHOOTS > 25% BLUE (%) SHOOTS (%)
CHS-A 7/18/96 40 16 (40.0) 3 (7.5)
CHS-A 7/19/95 12 3 (25.0) 0 (0.0)
CHS-A 10/06/95 18 14 (77.8) 7 (38.9)
CHS-A 10/21/95 4 0 (0.0) 0 (0.0)
CHS-A Total 74 33 (44.6) 10 (13.5)

CHS-S 7/21/95 21 11 (52.4) 1 (4.8)
CHS-S 10/06/95 14 5 (35.7) 3 (21.4)
CHS-S Total 35 16 (45.7) 4 (11.4)

CHI-A 6/4/95 91 25 (27.5) 9 (9.9)
CHI-A 6/5/95 16 11 (68.8) 5 (31.3)
CHI-A Total 107 36 (33.6) 14 (13.1)

CHI-S 7/20/95 27 14 (51.9) 2 (7.4)
CHI-S 10/06/95 17 10 (58.8) 1 (5.9)
CHI-S Total 44 24 (54.5) 3 (6.8)

Total 260 109 (41.9) 31 (11.9)

TABLE 4.3. Summary of transformants.

CHS-A 21 21 7 7
CHS-S 24 18 5 4
CHI-A 23 11 6 6
CHI-S 16 16 9 8
Total 84 66 27 25

Table 4.4. Naringin analysis data on grapefruit obtained with HPLC analysis.

CHS-A1 not included
CHS-A2 not included
CHS-A3 not included
CHS-A4 not included
CHS-A5 2.11
CHS-A6 3.30
CHS-S1 4.41
CHS-S2 6.30
CHS-S3 2.23
CHS-S4 2.46
CHI-A1 3.71
CHI-A2 2.40
CHI-A3 3.72
CHI-A4 2.13
CHI-A5 3.46
CHI-A6 0.20
CHI-S1 7.94
CHI-S2 1.28
CHI-S3 2.03
CHI-S4 8.89
CHI-S5 3.15
CHI-S6 6.66
CHI-S7 1.67
Control-1 1.84
Control-2 3.57
Control-3 4.41
Control-4 7.48

Hind III


Hind III


5' UTR

Poly A



Figure 4.1. Plasmid vector system used in the construction of transformation
vectors. The paired system of Slightom (1991) utilizes a HindIll cassette (top)
from pUC 18-exp 17 which is cloned into the Hindlll polylinker site of
pGA482/GG (bottom). The latter contains the T-DNA borders, reporter genes
and selectable markers used in Agrobacterium transformation.

Hind III


1 2 3 4 5 6 7 8 9 10 11 12 13 14 1516 17

4- 450 bp

350 bp

Figure 4.2. Polymerase chain reaction analysis detection of the presence of the
NPTII and GUS genes in transgenic grapefruit plants. PCR products of 350 bp
(NPTII) and 450 bp (GUS) were visualized on a 0.8% agarose-ethidium gel.
Lanes 1-5, amplification of NPTII fragments from plants staining GUS+. Lane 6,
no amplification from a chimeric staining plant. Lane 7, negative control
(non-transformed plant). Lane 8, positive NPTII control. Lane 9, 1 kb molecular
weight marker. Lane 10-14, amplification of GUS fragments from plants staining
GUS +. Lane 15, no amplification from chimeric staining plant. Lane 16,
negative control (non-transformed plant). Lane 17, positive GUS control.

1 2 3 4 5 6 7 8 9 10 11





4 kb 10 l

Figure 4.3. DNA blot analysis of CHS transformed grapefruit probed with GUS.
The DNA isolated from each plant was cut with Hpal and run on a 0.8% agarose
gel. The probe was a 1.8 kb GUS fragment isolated from pBI221. Lane 1, 1 kb
molecular weight marker; lane 2, non-transformed grapefruit plant; lane 3, CHS-
Al; lane 4, CHS-A2; lane 5, CHS-A3; lane 6, CHS-A4; lane7, CHS-A5; lane 8,
CHS-A6; lane 9, CHS-S1; lane 10, CHS-S2; lane 11, CHS-S3; lane 12, CHS-S4.
Lanes 3-8 were plants transformed with the antisense construct; lanes 9-12
transformed with the sense construct.

5 6 7 8 9 10 11 12 13 14

(_ il

9 P

4 kb ) : "

Figure 4.4. DNA blot analysis of CHI transformed grapefruit probed with GUS.
The DNA isolated from each plant was cut with Hpal and run on a 0.8% agarose
gel. The probe was a 1.8 kb GUS fragment. Lane 1, 1 kb molecular weight
marker; lane 2, nontransformed grapefruit plant; lane 3, CHI-A1; lane 4, CHI-A2;
lane 5, CHI-A3; lane 6, CHI-A4; lane 7, CHI-A5; lane 8, CHI-A6; lane 9, CHI-S1;
lane 10, CHI-S2; lane 11, CHI-S3; lane 12, CHI-S4; lane 13, CHI-S5; lane 14,
CHI-S6. Lanes 3-8 were plants transformed with the antisense construct; lanes
9-14 transformed with the sense construct.

1 2 3 4

1 2 3 4

K *^
/ f

5 6 7 8 9 10 11 12

* ;

nr Sg

5 kb --)

Figure 4.5. DNA blot analysis of CHS transformed grapefruit probed with a CHS
cDNA. The DNA isolated from each plant was cut with Hpal and run on a 0.8%
agarose gel. The probe was a 1.3 kb CHS cDNA fragment from grapefruit. Lane
1, 1 kb molecular weight marker; lane 2, non-transformed grapefruit plant; lane 3,
CHS-A1; lane 4, CHS-A2; lane 5, CHS-A3; lane 6, CHS-A4; lane 7, CHS-A5;
lane 8, CHS-A6; lane 9, CHS-S1; lane 10, CHS-S2; lane 11, CHS-S3; lane 12,
CHS-S4. Lanes 3-8 were plants transformed with the antisense construct; lanes
9-12 transformed with the sense construct.


r~ ~li~F~


dmy *jtlei **nia

1 2 3



5 6 7 8 9 10 11


12 13

5 kb -

Figure 4.6. DNA blot analysis of CHI transformed grapefruit probed with a CHI
cDNA. The DNA isolated from each plant was cut with Hpal and run on a 0.8%
agarose gel. The probe was a 0.9 kb CHI cDNA from grapefruit. Lane 1, 1 kb
molecular weight marker; lane 2, non-transformed grapefruit plant; lane 3,
CHI-A1; lane 4, CHI-A2; lane 5, CH I-A 3; lane 6, CHI-A 4; lane 7, CHI-A 5; lane
8, CHI-A 6; lane 9, CHI-S1; lane 10, CHI-S 2; lane 11, CHI-S3; lane 12, CHI-S4;
lane 13, CHI-S5; lane 14, CHI-S6. Lanes 3-8 were plants transformed with the
antisense construct; lanes 9-14 transformed with the sense construct.

1 2 3 4 5 6 7 8 9



Figure 4.7. RNA blot analysis of CHS transformed grapefruit probed with CMV.
Total RNA was run on a formaldehyde gel and blotted. The probe was the 70 bp
CMV fragment from the transformation plasmid. Lane 1, RNA molecular weight
marker; lane 2, blank; lane 3, CHS-A4; lane 4, CHS-A5; lane 5,CHS-A6; lane 6,
CHS-S1; lane 7, CHS-S2; lane 8, CHS-S3; lane 9, control, non-transformed

A Mh^f

1 2 3 4 5 6 7 8 9

*~ ~ '
yl.:l::' ,~g*~"~: j~
". '"I: .-r
;- ~il

2.4 kb -


b,: .; : ~4~
`:~~Jt~~E~~l~i~i3~!~~8Ei~i. ;:
'.""1 [' "' .;.*" ;'n~ E;1it~iP'~c`' z+f.x ..
i .;.1';:'!*.'~

Figure 4.8. RNA blot analysis of CHS transformed grapefruit probed with a CHS
cDNA. Total RNA was run on a formaldehyde gel and blotted. The probe was
the 1.3 kb CHS grapefruit cDNA. Lane 1, RNA molecular weight marker; lane 2,
blank; lane 3, CHS-A4; lane 4, CHS-A5; lane 5, CHS-A6; lane 6, CHS-S1; lane
7, CHS-S2; lane 8, CHS-S3; lane 9, control, non-transformed plant.

1 2 3 4 5 6 7 8 9 10 11 12

1 kb-P

Figure 4.9. RNA blot analysis of CHI transformed grapefruit probed with CMV.
Total RNA was run on a formaldehyde gel and blotted. The probe was the 70 bp
CMV fragment from the transformation plasmid. Lane 1, RNA molecular weight
marker; lane 2,CHI-A1; lane 3, CHI-A2; lane 4, CHI-A3; lane 5, CHI-A4; lane 6,
CHI-A5; lane 7, CHI-S1; lane 8, CHI-S2; lane 9, CHI-S3; lane 10, CHI-S4; lane
11, CHI-S6; lane 12, control, non-transformed plant.

1 2 3 4 5 6 7 8 9 10 11 12

1 kb -i

Figure 4.10. RNA blot analysis of CHI transformed grapefruit probed with a CHI
cDNA. Total RNA was run on a formaldehyde gel and blotted. The probe was a
0.9 kb CHI grapefruit cDNA. Lane 1, RNA molecular weight marker; lane
2,CHI-A1; lane 3, CHI-A2; lane 4, CHI-A3; lane 5, CHI-A4; lane 6, CHI-A5; lane
7, CHI-S1; lane 8, CHI-S2; lane 9, CHI-S3; lane 10, CHI-S4; lane 11, CHI-S6;
lane 12, control, non-transformed plant.


The bitter flavor of naringin can result in undesirable or unacceptable fruit
and juice quality in grapefruit (Hagen et al. 1966). The objective of this research

was to attempt to decrease levels of naringin in Citrus paradisi using a
transgenic approach. The transgenic approach utilized the genes for CHS and
CHI that encode enzymes in the first committed step and the following step in

flavonoid biosynthesis. We did not target naringin specifically, but hypothesized

that a decrease overall in the pathway might also result in decreased levels of
naringin. Transgenic plants were analyzed for evidence of transformation and
gene expression as well as naringin content.

Cloning of Citrus CHS and CHI cDNA

The biosynthetic genes encoding the enzymes leading to the production
of naringin had not been cloned in Citrus. The most obvious strategy for altering

an end product might be to manipulate the genes for the enzymes furthest down
the biosynthetic pathway of the product. However, although various sugar
transferases are common enzymes in plant species, it has proven difficult in
general to clone the genes for these enzymes.

CHS and CHI participate in the first two steps of the flavonoid biosynthetic
pathway leading to the production of naringin. The availability of CHS and CHI
genes cloned in Antirrhinum majus provided a tool for heterologous probing of a
citrus cDNA library. Using young leaves from a mature grapefruit tree

(cv. Marsh), RNA was isolated. Once it was determined by Northern blot

analysis that the heterologous probes did detect transcript from the citrus leaf
material, a cDNA library was constructed. Approximately 5 x 105 primary

recombinant phage were screened in succession first with the CHS probe, then
with the CHI probe. The clones with the longest putative cDNA insert were

selected for further analysis. Screening the library and analyzing the DNA

sequence obtained we were successful in cloning cDNAs of CHS and CHI from
Citrus paradisi.

Analysis of genomic DNA and Southern blots suggested that both CHS
and CHI in Citrus could be encoded by multigene families. This was not

surprising, given the large number of different species that contain multiple
members for these 2 enzymes (Koes et al 1989; van Tunen et al. 1989). The
expression analysis suggested that members of both gene families may be

expressed in all tissue parts examined (flowers, leaves, stems, and roots), which

also was not surprising given the many functions of flavonoid compounds in the
plant. Unexpected problems associated with the probing of total RNA lead to
nonspecific binding of RNA probes, with possible irreversible binding of the

probe to the blot. These blots were probed with CHS and CHI cDNAs, but the

analysis should be repeated, especially to determine relative levels of expression
among the different tissues.

Plant Transformation and Transgenic Analysis

Molecular genetics provides an opportunity to manipulate levels of
secondary compounds by altering the expression of genes encoding key
enzymes or diverting existing pathways. One approach is to block expression of

the gene product encoding an enzyme, thereby blocking biosynthesis at that

step. Gene silencing using this approach has primarily been achieved in 2 ways:
through sense suppression or with antisense technology.

We successfully constructed transformation plasmids with the sense and
antisense orientations of CHS and CHI using the pUC18 and pGA482/GG vector

system of Slightom (1991). The 4 plasmids were transformed into the

Agrobacterium cell line EHA101 and utilized in a series of seedling epicotyl
transformations using 1-3 month old seedlings of Duncan grapefruit.

All 4 plasmids used produced GUS positive shoots that developed into
plants. However, a number of the shoots transformed with the CHS antisense
construct displayed abnormal leaf morphology, stunting, and died before they
could be tested for evidence of transformation. If observed deleterious effects

were associated with the antisense effect it may reflect the biological importance
of flavonoids in Citrus. Flavonoid deficient plants have been reported as
intolerant to UV-B irradiation, producing non-functional pollen, and displaying
increased susceptibility to stress (Shirley et al. 1992,1995).

The number of regenerated shoots which stained positive for GUS ranged
from 33-68%. While we were pleased with such a high frequency, all plasmids
used resulted in a high number of shoots with chimeric staining for GUS. The
number of solidly staining blue shoots ranged from 0-38% in all experiments,

with an overall average of 11% solid blue. The chimeric staining is

understandable considering that the epicotyl method of transformation does not
involve single cells or undifferentiated calli. When all surviving plants were
retested for GUS, 25% stained solid blue. To increase the number of non-
chimeric transformed shoots it may be necessary to increase the level of

kanamycin selection. Higher initial levels of kanamycin (60-100mg/L) should be

investigated as a means of selecting solidly transformed plants.

Rooting citrus plants in vivo continues to be a laborious and slow process

(Duran-Vila et al. 1989). This study was successful in rooting shoots on medium;

however auxin requirements involved in the rooting process need to be

optimized. We found that a 3 week period of no hormones, followed by a brief
period on 5[tM NAA (up to a week) resulted in roots being produced. Once initial

roots were established, plants appeared to grow better in peat pots than in

sealed growth cups. Higher light and temperature may also have contributed to

better growth. Although we have shown that grapefruit can be rooted in medium,

micrografting (Pefa et al. 1995b) could be a useful technique and should be

investigated. Micrografting could potentially replace rooting of all transformants

except those specifically developed for rootstock use.

Nucleic acid analysis provided evidence for successful transformation with

all 4 constructs used. The use of PCR to determine integration of the GUS and

NPT II genes provided a simple technique that could screen large numbers of

plants using small amounts of tissue. The Southern blot analysis confirmed

integration of the T-DNA.

The Northern analysis completed was inconclusive due to problems with

non-specific binding of probes. This type of analysis would have provided
information on steady-state transcript levels. Our objective was to determine if

there was a correlation of gene expression to naringin produced. To investigate

at what level silencing might occur requires additional analysis. Researchers

often compare the results of transcription run-on experiments with the Northern

analysis to determine whether silencing has resulted from transcriptional or

post-transcriptional inactivation processes (Matzke and Matzke 1995).

The use of the RT-PCR (reverse transcriptase polymerase chain

reaction) technique could enhance this type of study. The tissue requirements

are less than for Northern blot analysis. RT-PCR has provided information on

gene expression (Simpson et al. 1992a), including expression of different

members of a multigene family (Simpson et al. 1992b). The technique has also
been modified to analyze expression in transgenic plants carrying antisense
gene constructs (Brown et al. 1993).

We were fortunate to be assisted in the naringin analysis by Dr. Berhow
who had extensive knowledge of quantifying naringin levels in C. paradisi using
HPLC. The naringin levels were measured while the plants were less than a

year old due to time constraints. High levels of naringin are associated with

young tissue, and these levels decrease by dilution as the plant grows (Castillo
et al. 1992). Unfortunately the naringin levels associated with our control plants

varied considerably, resulting in a large standard deviation. This made it difficult
to compare the transgenic values to the controls. The naringin levels should be
reanalyzed after the plants are at least one year old.

Alternative methods of flavonoid analysis were explored but not
recommended by Dr. Berhow. These included the Davis method (1947), which
colorimetrically determines the overall concentration of all flavanones, and TLC

methods (Hagen et al. 1965), which are not as sensitive or very good at
differentiating flavanone, flavone, and flavonol glycoside concentrations (Dr.
Mark Berhow, personal communication). However, experiments like ours did

not target naringin specifically, and may be viewed as an attempt to decrease

the level of total flavonoids, including naringin. Future experiments could
measure total flavonoids by the Davis method and compare this to HPLC data.
The Davis method might serve as a preliminary screening tool to decrease the
number and cost of plants analyzed by HPLC.

Future Studies

Certainly, this type of experiment must be repeated before we will be able
to conclude that a transgenic approach can successfully decrease the levels of
naringin in Citrus paradise. The small number of plants analyzed keeps us from

making generalized statements. While we were pleased to obtain transformants
using all 4 constructs, limiting the number of constructs may be the best

approach to obtain adequate numbers of plants. Due to the complexity of

silencing by cosuppression, antisense may be the better approach.

Including the fruit promoter in future transgenic experiments to decrease
naringin would confine gene silencing to the fruit level, thus preventing any
negative effects that occur to leaf morphology and plant development. This
might prevent the negative effect that was seen on leaf morphology when using

the antisense CHS construct. As part of this study we constructed a genomic
library from DNA isolated from leaves of C. paradisi (data not shown). Genomic
clones including the promoters could be isolated. Following the construction of a

genomic library from fruit tissue, molecular techniques are available to isolate a
fruit specific promoter.

Our attempt to measure protein levels was limited by not having CHS and
CHI antibodies from citrus protein. If possible these should be available for
future studies, however alternate methods of analysis have been used to

measure changes in enzyme levels. Suppression of CHS activity can be
monitored through the analysis of the precursors. Studies on control and
antisense CHS plants have shown a blockage in the pathway that leads to the
accumulation of coumaric acid. Antisense CHS experiments have included the

examination of total flavonoids from the transgenic and control plants plated on

TLC plates with coumaric acid as the standard (Courtney-Gutterson et al. 1994).

A CHS enzyme assay (O'Neill et al 1990) was successfully used by Elomaa et
al. 1993 using protein extracts from tissue of transformed and control Gerbera
hybrida to detect naringenin chalcone levels, indicating CHS activity levels.

These types of study could provide additional information to determine if enzyme

or product levels were inhibited at the target point.

Lastly, another approach at decreasing naringin content in future studies
could be to target the enzyme rhamnosyltransferase. Other researchers (Bar-
Peled et al. 1993) have suggested this step as the logical step to use antisense

technology. Although the rhamnosyltransferase gene has not been cloned, the
enzyme has been isolated, and antibody produced. Screening existing citrus
cDNA libraries with the antibody should be a simple approach to cloning the

rhamnosyltransferase gene.


Akada, R. (1994). Quick-check method to test the size of Escherichia coli
plasmids. Biotechniques 17, 58.

Albach, R. F., and Redman, G. H. (1969). Composition and inheritance of
flavanones in citrus fruit. Phytochemistry 8, 127-143.

Altschul, S. F., Gish, W., Miller, W., Myers, E. W., and Lipman, D. J. (1990).
Basic local alignment search tool. J. Mol. Biol. 215, 403-410.

An, G. (1987). Binary Ti vectors for plant transformation and promoter analysis.
Methods in Enzymology 153, 292-305.

Attaway, J. A. (1977). Factors influencing the flavor of grapefruit juice. Proc. Int.
Soc. Citriculture 3, 816-820.

Bar-Peled, M., Fluhr, R., and Gressel, J. (1993). Juvenile-specific localization
and accumulation of a rhamnosyltransferase and its bitter flavonoid in
foliage, flowers, and young citrus fruits. Plant Physiology 103, 1377-1384.

Bar-Peled, M., Lewinsohn, E., Fluhr, R., and Gressel, J. (1991). UDP-rhamnose:
Flavanone -7-O-glucoside-2'-O-rhamnosyltransferase. J. Biological
Chemistry 266, 20953-20959.

Berhow, M. A., and Smolensky, D. (1995). Developmental and substrate
specificity of hesperetin-7-O-glucosyltransferase activity in Citrus limon
tissues using high-performance liquid chromatographic analysis. Plant
Science 112, 139-147.

Berhow, M. A., and Vandercook, C. E. (1989). Biosynthesis of naringin and
prunin in detached grapefruit. Phytochemistry 28, 1627-1630.

Berhow, M. A., and Vandercook, C. E. (1992). The reduction of naringin content
of grapefruit by applications of gibberellic acid. Plant Growth Regulation
11, 75-80.

Berhow, M. A., and Vandercook, C. E. (1991). Sites of naringin biosynthesis in
grapefruit seedlings. Plant Physiology 138, 176-179.

Bourque, J. E. (1995). Antisense strategies for genetic manipulations in plants.
Plant Science 105, 125-149.

Bourque, J. E., and Folk, W. R. (1992). Suppression of gene expression in plant
cells utilizing antisense sequences transcribed by RNA polymerase IIl.
Plant Molecular Biology 19, 641-647.

Brown, J. W. S., Simpson, G., Clark, G., Lyon, J., Kumar, A., and Simpson, C.
(1993). Detection of antisense transcripts in transgenic plants by RT-PCR.
Plant Journal 4, 883-885.

Brusslan, J. A., Karlin-Neumann, G. A., Huang, L., and Tobin, E. M. (1993). An
Arabadopsis mutant with a reduced level of cabl40 RNA is a result of
cosuppression. Plant Cell 5, 667-677.

Cannon, M., Platz, J., M., O. L., Sookdeo, C., and Cannon, F. (1990). Organ
specific modulation of gene expression in transgenic plants using
antisense RNA. Plant Molecular Biology 15, 205-211.

Castillo, J., Benavente, O., and del Rio, J. A. (1992). Naringin and neohesperidin
levels during development of leaves, flower buds, and fruits of Citrus
aurantium. Plant Physiology 99, 67-73.

Chandler, B. V. (1958). Anthocyanins of blood oranges. Nature 182, 933.

Church, G. M., and Gilbert, W. (1984). Genomic Sequencing. Proc. Natl. Sci. 81,

Coe, E. H., McCormick, S. M., and Modena, S. A. (1981). White pollen in maize.
Heredity 72, 318-320.

Cone, K. (1989). Yet another rapid plant DNA prep. Maize Genetics Cooperative
Newsletter 63, 68.

Courtney-Gutterson, N., Napoli, C., Lemieux, C., Morgan, A., Firoozabady, E.,
and Robinson, K. E. P. (1994). Modification of flower color in florist's
chrysanthemum: Production of a white-flowering variety through
molecular genetics. Bio/Technology 12, 268-271.

Damak, S., and Bullock, D. W. (1993). A simple two-step method for efficient
blunt-end ligation of DNA fragments. Biotechniques 15, 448-452.

Davies, F. S., Krezdorn, A. H., and Jackson, L. K. (1993). Introduction to citrus
culture. FRC 3212 manual. Horticultural Sciences. University of Florida.
Gainesville. 141 p.

Davis, W. B. (1947). Determination of flavanones in citrus tissues. Analytical
Chemistry 19, 476-478.

de Lange, P., de Boer, G. J., Mol, J. M. N., and Kooter, J. M. (1993). Conditional
inhibition of 3-glucuronidase expression by the antisense gene fragments
in petunia protoplasts. Plant Molecular Biology 23, 45-55.

Delauney, A. J., Tabaeizadeh, Z., and Verma, D. P. S. (1988). Inhibition of gene
expression in transformed plants by antisense RNA. Plant Molecular
Biology 11, 301-310.

Devereux, J., Haeberli, P., and Smithies, O. (1984). A comprehensive set of
sequence analysis programs for the VAX. Nucleic Acids Research 12,

Dixon, R. A. (1991). Plant cell culture. In The practical approach (Oxford: IRL
Press Limited), pp. 236.

Dooner, H. K. (1983). Coordinate genetic regulation of flavonoid biosynthetic
enzymes in maize. Mol. Gen. Genet. 189, 136-141.

Duran-Vila, N., Ortega, V., and L., N. (1989). Morphogenesis and tissue cultures
of three citrus species. Plant Cell, Tissue and Organ Culture 16, 123-133.

Ecker, J., and Davis, R. (1986). Inhibition of gene expression in plant cells by
expression of antisense RNA. Proc. Natl. Acad. Sci. USA 83, 5372-5376.

Elmayan, T., and Vaucheret, H. (1996). Expression of single copies of a strongly
expressed 35S transgene can be silenced post-transcriptionally. Plant
Journal 9, 787-797.

Elomaa, P., Honkanen, J., Puska, R., Seppanen, P., Helsriutta, Y., Mehto, M.,
Kotilainen, M., Nevalainen, L., and Teeri, T. H. (1993). Agrobacterium-
mediated transfer of antisense chalcone synthase cDNA to Gerbera
hybrida inhibits flower pigmentation. Bio/Technology 11, 508-511.

English, J., Mueller, E., and Baulcombe, D. C. (1996). Suppression of virus
accumulation in transgenic plants exhibiting silencing of nuclear genes.
Plant Cell 8, 179-188.

Finnegan, J., and McElroy, D. (1994). Transgene inactivation: plants fight back!
Bio/Technology 12, 883-888.

Flavell, R. B. (1994). Inactivation of gene expression in plants as a consequence
of specific sequence duplication. Proc. Natl. Acad. Sci. 91, 3490-3496.

Gavish, H., Lewinsohn, E., Vardi, A., and Fluhr, R. (1989). Production of
flavanone-neohesperidosides in Citrus embryos. Plant Cell Reports 8,

Gmitter, F. G., and Moore, G. A. (1986). Plant regeneration from undeveloped
ovules and embryogenic calli of Citrus: embryo production, germination
and plant survival. Plant Cell Tissue Organ Culture 6, 139-147.

Goodwin, J., Chapman, K., Swaney, S., Parks, T. D., Wernsman, E. A., and
Dougherty, W. G. (1996). Genetic and biochemical dissection of
transgenic RNA-mediated virus resistance. Plant Cell 8, 95-105.

Goodwin, T. W., and Mercer, E. I. (1983). Introduction to Plant Biochemistry
(Oxford: Pergamon Press).

Grierson, D., Fray, R. G., Hamliton, A. J., Smith, C. J. S., and Watson, C. F.
(1991). Does co-suppression of sense transgenes involve antisense
RNA? Trends in Biotechnology 9, 122-123.

Grotewald, E., and Peterson, T. (1994). Isolation and characterization of a maize
gene encoding chalcone flavonone isomerase. Mol. Gen. Genet. 242, 1-8.

Gutierrez- Espinosa, M. A. (1995). Production of transgenic citrus plants
expressing the citrus tristeza virus coat protein gene. Ph.D. Dissertation.
University of Florida. Gainesville. 95p.

Hagen, R. E., Dunlap, W. J., Mizelle, J. W., Wender, S. H., Lime, B. J., Albach,
R. F., and Griffiths, F. P. (1965). A chromatograhic-fluorometric method
for determining naringin, naringenin rutinoside, and related flavanone
glycosides in grapefruit juice and juice sacs. Analytical Biochemistry 12,

Hagen, R. E., Dunlap, W. J., and Wender, S. H. (1966). Seasonal variation of
naringin and certain other flavanone glycosides in juice sacs of Texas
Ruby Red grapefruit. Food Science 31, 542-547.

Hahlbrock, K., and Grisebach, H. (1979). Enzymic controls in the biosynthesis of
lignin and flavonoids. Annual Review of Plant Physiology 30, 105-130.

Hahlbrock, K., and Scheel, D. (1989). Physiology and molecular biology of
phenylpropanoid metabolism. Annu. Rev. Plant Physiol. Plant Mol. Biol.
40, 347-369.

Hall, L. N., Tucker, G. A., Smith, C. J. S., Watson, C. F., Seymour, G. B.,
Bundick, Y., Boniwell, J. M., Fletcher, J. D., Ray, J. A., Schuch, W., Bird,
C. R., and Grierson, D. (1993). Antisense inhibition of pectin esterase
gene expression in transgenic tomatoes. Plant Journal 3, 469-481.

Harker, C. L., Ellis, T. H., and Coen, E. (1990). Identification and genetic
regulation of the chalcone synthase multigene family. Plant Cell 2, 185-

Hasegawa S., and Maier, V. P. (1970). Biosynthesis of trans-cinnamate from
phenylpyruvate and L-glutamate by cell free extracts of grapefruit.
Phytochemistry 9, 2483-2487.

Hasegawa, S., Merhow, M. A., and Fong, C. H. (1996). Analysis of bitter
principles in Citrus. Modern methods of plant analysis 18, 59-80.

Heller, W., and Hahlbrock, K. (1980). Highly purified flavanone synthase from
parsley catalyzes the formation of naringenin chalcone. Archives
Biochemistry and Biophysics 200, 617-619.

Herrmann, A., Schulz, W., and Hahlbrock, W. (1988). Two alleles of the single-
copy chalcone synthase gene in parsley differ by a transposon-like
element. Mol. Gen. Genet 212, 93-98.

Hidaka, T., Omura, M., Ugaki, M., Tomiyama, M., Kato, A., Ohshima, M., and
Motoyoshi, F. (1990). Agrobacterium-mediated transformation and
regeneration of Citrus spp. from suspension cells. Japanese Journal of
Breeding 10, 199-207.

Hobbs, S. L. A., Warkentin, T. D., and DeLong, C. M. 0. (1993). Transgene copy
number can be positively or negatively associated with transgene
expression. Plant Molecular Biology 21, 17-26.

Holton, T. A., and Cornish, E. C. (1995). Genetics and biochemistry of
anthocyanin synthesis. Plant Cell 7, 1071-1083.

Hood, E. E., Helmer, G. L., Fraley, R. T., and Chilton, M.-D. (1986). The
hypervirulence of Agrobacterium tumefaciens A281 is encoded in a region
of pTiBo542 outside of T-DNA. Journal of Bacteriology 168, 1292-1301.

Horowitz, R. M. (1961). The citrus flavonoids. In The Orange, W. B. Sinclair, ed.
(Davis: University of California).

Horowitz, R. M., and Gentili, B. (1977a). Flavonoid constituents of Citrus. In
Citrus Science and Technology, S. Nagy, P. E. Shaw and M. K. Veldhuis,
eds. (Westport: AVI), pp. 397-426.

Horowitz, R. M., and Gentili, B. (1977b). Flavonoids: Uses and byproducts. Proc.
Int. Soc. Citriculture 3, 743-746.

Horowitz, R. M., and Gentili, B. (1969). Taste and structure in phenolic
glycosides. J. Agric. Food Chemistry 17, 696-700.

Inglebrect, I., Breyne, P., Vancompernolle, K., Jacobs, A., Van Montagu, M., and
Depicker, A. (1991). Transcriptional interference in transgenic plants.
Gene 109, 239-242.

Johnson, R. L. (1988). Bitterness in fruit juices: Assay of bitter principles in citrus
juice. In Modern methods of plant analysis, H. F. Linskens and J. F.
Jackson, eds. (Berlin: Springer-Verlag).

Jourdan, P., Mclntosh, C. A., and Mansell, R. L. (1985). Naringin levels in citrus
tissues. Plant Physiology 77, 903-908.

Kaneyoshi (Hiramatsu), J., Kobayashi, S., Nakamura, Y., Shigemota, N., and
Doi, Y. (1994). A simple and efficient gene transfer system of trifoliate
orange (Poncirus trifoliata Raf.). Plant Cell Reports 13541-545.

Kefford, J. F. (1959). The chemical constituents of citrus fruits. Advances in Food
Research 9, 285.

Keshet, I., Yisraeli, J., and Cedar, H. (1985). Effect of regional DNA methylation
on gene expression. Proc. Natl. Acad. Sci. 82, 2560-2564.

Kobayashi, S., and Uchimaya, H. (1989). Expression and integration of a foreign
gene in orange (Citrus sinensis Osb.) protoplasts by direct DNA transfer.
Japanese Journal of Genetics 64, 91-97.

Koes, R. E., Spelt, C. E., Mol, J. N. M., and Gerats, A. G. M. (1988). The
chalcone synthase multigene family of Petunia hybrida (V30): sequence
homology, chromosomal location and evolutionary aspects. Plant
Molecular Biology 10, 377-385.

Koes, R. E., Spelt, C. E., van den Elzen, P. J. M., and Mol, J. N. M. (1989).
Cloning and molecular characterization of the chalcone synthase
multigene family of Petunia hybrida. Gene 81, 245-257.

Kreuzaler, F., and Hahlbrock, K. (1972). Enzymic synthesis of aromatic
compounds in higher plants: formation of naringenin (5.7.4'-
trihydroxyflavanone) from p-coumaroyl coenzyme A and malonyl
coenzyme 4. FEBS Letters 28, 69-72.

Kreuzaler, F., Ragg, H., Fautz, E., Kuhn, D. N., and Hahlbrock, K. (1983). UV-
induction of chalcone synthase mRNA in cell suspension cultures of
Petroselinum hortense. Proc. Natl. Acad. Sci. USA 80, 2591-2593.

Kuhn, B., Forkmann, G., and Seyffert, W. (1978). Genetic control of chalcone
isomerase activity in Callistephus chinensis. Planta 138, 199-203.

Lewinsohn, E., Berman, E., Mazur, Y., and Gressel, J. (1989b). (7) Glucosylation
and (1-6) rhamnosylation of exogenous flavanones by undifferentiated
Citrus cell cultures. Plant Science 61, 23-28.

Lewinsohn, E., Berman, E., Mazur, Y., and Gressel, J. (1986). Glucosylation of
exogenous flavanones by grapefruit (Citrus paradise) cell cultures.
Phytochemistry 25, 2531-2535.

Lewinsohn, E., Britsch, L., Mazur, L., and Gressel, J. (1989a). Flavanone
glycoside biosynthesis in Citrus. Plant Physiology 91, 1323-1328.

Lin, J.-J., Assad-Garcia, and Kuo, J. (1994). Effects of Agrobacterium cell
concentration on the transformation efficiency of tobacco and Arabadopsis
thaliana. Focus 16, 72-77.

Linn, F., Heidmann, I., Saedler, H., and Meyer, P. (1990). Epigenetic changes in
the expression of the maize Al gene in Petunia hybrida: Role of numbers
of integrated gene copies and state of methylation. Mol. Gen. Genet. 222,

Loake, G. J., Choudhary, A. D., Harrison, M. J., Mavandad, M., Lamb, C. J., and
Dixon, R. A. (1991). Phenylpropanoid intermediates regulate transient
expression of a chalcone synthase gene promoter. Plant Cell 3, 829-840.

Maier, V. P. (1969). Compositional studies of citrus: significance in processing,
identification, and flavor. Proceedings First International Citrus
Symposium 1, 235-241.

Martin, C., Prescott, A., Mackay, S., Bartlett, J., and Vrijlandt, E. (1991). Control
of anthocyanin biosynthesis in flowers of Antirrhinum majus. Plant Journal
1, 37-49.

Matsumoto, R., and Okudai, N. (1991). Inheritance of flavanone
neohesperidoside in Citrus. J. Japan. Soc. Hort. Sci. 60, 201-207.

Matzke, M. A., and Matzke, A. J. M. (1991). Differential inactivation and
methylation of a transgene in plants by two suppressor loci containing
homologous sequences. Plant Molecular Biology 16, 821-830.

Matzke, M., and Matzke, A. J. M. (1993). Genomic imprinting in plants: Parental
effects and trans-inactivation phenomena. Annual Review Plant
Physiology Plant Molecular Biology 44, 53-76.

Matzke, M. A., and Matzke, A. J. M. (1995). How and why do plants inactivate
homologous (trans)genes? Plant Physiology 107, 679-685.

Mclntosh, C. A., Latchinian, L., and Mansell, R. L. (1990). Flavanone-specific 7-
o-glucosyltransferase activity in Citrus paradisi seedlings: Purification and
characterization. Archives Biochemistry and Biophysics 282, 50-57.

McKhann, H. I., and Hirsch, A. M. (1994). Isolation of chalcone synthase and
chalcone isomerase cDNAs from alfalfa (Mediago sativa L.): highest
transcript levels occur in young roots and root tips. Plant Molecular
Biology 24, 767-777.

Mehdy, M. C., and Lamb, C. J. (1987). Chalcone isomerase cDNA cloning and
mRNA induction by fungal elicitor, wounding, and infection. EMBO 6,

Meins, F., and Kunz, C. (1994). Gene silencing in plants: a heuristic
autoregulation model. Curr. Topics Microbiol. 197, 105-119.

Meyer, P., Heidmann, I., and I., N. (1993). Differences in DNA methylation are
associated with a paramutation phenomenon in transgenic petunia. Plant
Journal 4, 89-100.

Meyer, P., Niedenhof, I., and Tenlohuis, M. (1994). Evidence for cytosine
methylation of non-symmetrical sequences in transgenic Petunia hybrida.
EMBO 13, 2084-2088.

Mol, J. N. M., Stuitje, A. R., and van der Krol, A. (1989). Genetic manipulation of
floral pigmentation genes. Plant Molecular Biology 13, 287-294.

Moore, G. A., Jacano, C. C., Neidigh, J. L., Lawrence, S. D., and Cline, K.
(1992). Agrobacterium-mediated transformation of Citrus stem segments
and regeneration of transgenic plants. Plant Cell Reports 11, 238-242.

Mueller, E., Gilbert, J., Davenport, G., Brigneti, G., and Baulcombe, D. C. (1995).
Homology-dependent resistance: transgenic virus resistance in plants
related to homology-dependent gene silencing. Plant Journal 7, 1001-13.

Murray, E. J. (1991). Gene Transfer and expression protocols. In Methods in
Molecular Biology, J. M. Walker, ed. (Clifton: Humana Press).

Napoli, C., Lemieux, C., and Jorgansen, R. (1990). Introduction of a chimeric
chalcone synthase gene into petunia results in reversible co-suppression
of homologous genes in trans. Plant Cell 2, 279-289.

Niesbach-Kl6sgen, U., Barzen, E., Bernhardt, J., Rohde, W., Schwarz-Summer,
Z., Reif, H. J., Wienand, U., and Saedler, H. (1987). Chalcone synthase
genes in plants: A tool to study evolutionary relationships. Molecular
Evolution 26, 213-225.

O'Neill, S. D. O., Tong, Y., Sporlein, B., Forkmann, G., and Yoder, J. I. (1990).
Molecular genetic analysis of chalcone synthase in Lycopersicon
esculentum and an anthocyanin-deficient mutant. Mol. Gen. Genet. 224,

Paithankar, K. R., and Prasad, K. S. N. (1991). Precipitation of DNA by
polyethylene glycol and ethanol. Nucleic Acids Research 19, 1346.

Palaqui, J.-C., and Vaucheret, H. (1995). Field trial analysis of nitrate reductase
co-suppression: A comparative study of 38 combinations of transgene
loci. Plant Molecular Biology 29, 149-159.

Pang, S. Z., Slightom, J. L., and Gonsalves, D. (1993). Different mechanisms
protect tobacco against tomato spotted wilt and impatiens necrotic spot
tospoviruses. Bio/Techology 11, 819-824.

Parks, Y. D., Papp, I., Moscone, E. A., Iglesia, V. A., Vauchreret, H., Matzke, A.
J. M., and Matzke, M. A. (1996). Gene silencing mediated by promoter
homology occurs at the level of transcription and results in meiotically
heritable alterations in methylation and gene activity. Plant Journal 9, 183-

Peia, L., Cervera, M., JuBrez, J., Navarro, A., Pina, J. A., Nuria, D. V., and
Navarro, L. (1995b). Agrobacterium-mediated transformation of sweet
orange and regeneration of transgenic plants. Plant Cell Reports 14, 616-

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 EQR6R6PQ1_L3BMJW INGEST_TIME 2013-10-10T00:57:17Z PACKAGE AA00017677_00001