Botanical Studies (2009) 50: 115-125.

*

Corresponding author: E-mail: kyto@gate.sinica.edu.tw;

Tel: 886-2-26533161; Fax: 886-2-26515600.

INTRODUCTION

Isolation of high-quality RNA is a critical step in many

molecular biology experiments, such as cDNA synthesis,

cDNA library construction, RT-PCR (reverse transcription

-polymerase chain reaction), subtractive hybridization,

SAGE (serial analysis of gene expression) technology,

EST (expressed sequence tags) analysis, or DNA

microarray analysis (To, 2000; To, 2004). Various methods

have been developed to isolate high-quality RNA in

reasonable amounts from plant tissues which may contain

high levels of polyphenolic compounds, polysaccharides,

pigments and RNase. High salt concentrations in the

extraction buffer and an aqueous two-phase system

coupled with conventional phenol/chloroform extraction

and CsCl centrifugation have been used to isolate and

purify RNA from different tissues of pine trees, which

Isolation of functional RNA from different tissues

of tomato suitable for developmental profiling by

microarray analysis

Hsin-Mei WANG

1

, Wan-Chun YIN

2

, Chen-Kuen WANG

3

, and Kin-Ying TO

3,

*

1

Institute of Plant Biology, National Taiwan University, Taipei 106, Taiwan

2

Institute of Biotechnology, National Tsing Hua University, Hsinchu 300, Taiwan

3

Agricultural Biotechnology Research Center, Academia Sinica, Taipei 115, Taiwan

(Received August 19, 2008; Accepted November 28, 2008)

ABSTRACT.

An efficient and reproducible method is described for isolating high-quality RNA suitable for

microarray analysis from vegetative and reproductive tissues of tomato plants at different stages of growth.

This method was based on TRIzol method followed by lithium chloride (LiCl) precipitation and DNase treat-

ment. Using this method, high yields of high-quality and undegraded RNA were obtained as confirmed by

spectrophotometric method, gel electrophoresis and fluorescent quality control. The integrity and functionality

of RNA isolated by this procedure have been further demonstrated as probe in tomato cDNA microarrays for

identification of differentially expressed genes during fruit ripening, as template for cloning full-length cDNAs

encoding phytoene synthase (PSY), phytoene desaturase (PDS), £a-carotene desaturase (ZDS) and lycopene

£]-cyclase (LCY) in the carotenoid biosynthesis pathway by the reverse transcription-polymerase chain reaction

(RT-PCR), and as material for cDNA library construction.

Keywords: Carotenoid biosynthesis pathway; Fruit ripening; Microarray; RNA extraction; Solanum lycopersi-

cum; Tomato; TRIzol reagent.

Abbreviations: cDNA, complementary DNA; Cy3, Cy3-dUTP fluorescent dye; Cy5, Cy5-dUTP fluorescent

dye; DEPC, diethylpyrocarbonate; RT-PCR, reverse transcription-polymerase chain reaction.

Database Accession Nos: EF650010 (tomato cv. CL5915 phytoene synthase mRNA), EF650011 (tomato

cv. CL5915 phytoene desaturase mRNA), EF650012 (tomato cv. CL5915 £a-carotene desaturase mRNA), and

EF650013 (tomato cv. CL5915 lycopene £]-cyclase mRNA).

mOleCUlAR BIOlOgy

are especially rich in polyphenols (Schneiderbauer et

al., 1991). By using higher buffering capacity, alkaline

pH and polyvinylpyrrolidone (PVP), isolation of high-

quality RNA and DNA from cotton plants containing high

amounts of phenolic terpenoids and tannins has also been

reported (John, 1992). A rapid and facile modification of a

hexadecyltrimethyl ammonium bromide (CTAB) method

which allows for the preparation of total RNA from

recalcitrant materials such as pine needles without the use

of toxic chemicals has also been reported (Chang et al.,

1993). Based on the CTAB method with modifications,

an easy and efficient protocol was developed to isolate

high-quality total RNA from taxus and ginkgo (Liao et al.,

2004). Another method, using soluble PVP and ethanol

precipitation, has been reported and applied to several

recalcitrant materials such as ripening grape berries, dry

seeds of Albizia procera and radish, and leaf tissue of A.

procera and Griffonia simplicifolia (Salzman et al., 1999).

Extraction with phenol and polyvinyl polypyrolidone

(PVPP), followed by two purifications with LiCl plus a

116

Botanical Studies, Vol. 50, 2009

2-butoxyethanol treatment between the LiCl steps was

developed to isolate total RNA from pear plants which

contain considerable amounts of plant polyphenolic

compounds and polysaccharides (Malnoy et al., 2001).

Total RNA and genomic DNA have also been isolated

from an Australian native plant, Hakea actities, which is

grown in low nutrient conditions (Mason and Schmidt,

2002). Hot borate buffer at alkaline pH supplemented with

several adjuvants and followed by selective precipitations

has been used to isolate functional RNA from small

amounts of different grape and apple tissues (Moser et

al., 2004). A method consisting of a two-step extraction

with NaCl and trisodium citrate at high concentrations,

followed by isopropanol and LiCl precipitations, was

developed to isolate total RNA from siliques, dry seeds

and other tissues of Arabidopsis thaliana (Suzuki et al.,

2004).

In general, fruits are considered as one of recalcitrant

tissues and different methods suitable for isolating fruit

RNA have also been developed. A method consisting

of a lysis buffer [50 £gM aurintricarboxylic acid (an

RNA inhibitor), 5 to 10% mercaptoethanol, 5% PVPP,

MOPS buffer, EDTA, urea and Triton], followed by

phenol/chloroform extraction and LiCl precipitation,

was developed to isolate total RNA from tomato fruit

(Rodrigues-Pousada et al., 1990). Another method, based

on LiCl precipitation followed by CsCl precipitation, has

been reported to isolate RNA from blackcurrant fruit, a

tissue that contains high levels of phenolic compounds

and polysaccharides and the high acidity (pH 2.5 to 2.9)

(Woodhead et al., 1997). A protocol consisting of a hot

lysis buffer containing 2% sodium dodecyl sulfate (SDS)

and phenol with a high buffering capacity (300 mM Tris/

Boric acid) and the subsequent use of a high concentration

of PVPP (8.5%) to prevent polyphenol oxidation was

developed to isolate RNA from the high acidity (pH 3.5 to

4.0) of raspberry fruit (Jones et al., 1997). An extraction

buffer containing guanidine chloride and phenol and

followed by a three-stage precipitation (NaCl, urea and

LiCl, and potassium acetate) has been used to isolate total

RNA from carotenoid-rich plant tissues of the Mexican

marigold (Chi-Manzanero et al., 2000). A modified

method of using CTAB, PVP and £]-mercaptoethanol in

an extraction buffer has also been reported to eliminate

polysaccharides and prevent the oxidation of phenolic

compounds in bilberry fruit (Jaakola et al., 2001). Isolation

of functional RNA from cactus fruit, which is considered

to contain high amounts of secondary metabolites and

polysaccharides, using a lysis buffer of 150 mM Tris

(pH 7.5) and 2% SDS, followed by phenol/chloroform

extraction and LiCl precipitation (Valderrama-Chairez et

al., 2002).

Tomato is an important crop worldwide. It provides

abundant sources of antioxidant micronutrients such as

£]-carotene, lutein, phytoene, phytofluene, £^-carotene,

vitamins C and E, and phenolic compounds; some of which

may contribute to the health-giving properties of tomato

(Dumas et al., 2003). We are interested in the tissue-

specific profiling of differentially expressed genes by

using microarray approach and engineering of carotenoid

biosynthesis pathway in tomato plants. Therefore, we

initially need to set up a method for RNA isolation from

different developmental tissues of tomato plants, especially

in ripening fruits. Although several methods as mentioned

above have been described to isolate high-quality RNA

from various plant species as well as plant tissues, only

one method used in Arabidopsis (Suzuki et al., 2004) has

been examined at the microarray-based level, which is

extremely sensitive to the quality of the RNA (Duggan

et al., 1999; Burgess, 2004; To, 2004). Our method is

originally based on the use of a reagent containing an

acid guanidine thiocyanate-phenol-chloroform mixture

(Chomczynski and Sacchi, 1987); this mixture is well-

known as TRIzol reagent (Invitrogen; Chomczynski,

1993). Here we report an efficient and reproducible

protocol, which is based on the TRIzol method followed

by LiCl precipitation and DNase treatment, to isolate high-

quality total RNA from various tissues and developmental

stages of tomato plants. The quality of the total RNA

obtained was then evaluated at the molecular level.

mATeRIAlS AND meTHODS

Plant material

Seeds of tomato (Solanum lycopersicum var. CL5915)

were kindly provided by the AVRDC-The World Vegetable

Center in Taiwan. To obtain synchronized growth, tomato

seeds were shaked gently (50 rpm) in a 50-ml tube

containing 10 ml water at room temperature for 2 days.

Those germinating seeds with visible roots were sown in

a mixture of peat and vermiculite, and grown in a walk-

in growth chamber under a cycle of 14-h (6:00 a.m. to

8:00 p.m., 25¢XC) illumination (250 £gmol/m

2

/s) and 10-h

(8:00 p.m. to 6:00 a.m., 20¢XC) darkness. Different tissues

including cotyledons, stems, leaves, roots, flowers and

fruits were harvested during different developmental

stages (young seedlings, 1-month-old; vegetative growth,

2-month-old; flowering stage, 3-month-old; fruiting stage,

4-month-old) and stored at -80¢XC until used.

Isolation of total RNA and mRNA from different

tissues in tomato plants

For total RNA isolation, the protocol was based on the

TRIzol reagent user manual provided by the manufacturer

(Invitrogen), followed by LiCl precipitation and DNase

treatment. Plant tissues were grinded to a fine powder

in liquid nitrogen with a pre-cooled pestle and mortar,

and then put separately into a 50-ml plastic screw-cap

centrifuge tube. To one gram samples, 10 ml extraction

buffer [TRIzol reagent: 38% phenol (USB Cooperation,

Cleveland, Ohio, USA) was equilibrated to pH 4.0 with

Tris-HCl buffer; 0.8 M guanidine thiocyanate; 0.4 M

ammonium thiocyanate; 0.1 M sodium acetate (pH 5.0);

5% glycerol] was added and mixed well. Samples were

WANG et al. ¡X RNA isolation from tomato tissues

117

incubated for 5 min at room temperature. After incubation,

0.2 ml chloroform was added for each one ml extraction

buffer and tubes were shaked vigorously with votex for

15 sec. Tubes were incubated at room temperature for 3

min and then centrifuged at 10,000 g at 4

o

C for 15 min.

Aqueous phase was carefully transferred into a clean

screw-cap centrifuge tube and then added 0.5 ml of

isopropanol for each one ml extraction buffer. Tubes were

covered and mixed by gentle inversion, and then sit at

room temperature for 10 min. After incubation, tubes were

centrifuged (10,000 g; 4¢XC; 10 min) and the supernatant

was discarded. For each gram of tissue, 50 £gl DEPC-

treated H

2

O was added to dissolve the pellet, 1/3 volume

of 8 M LiCl was added to each tube, and tubes were placed

at 4¢XC overnight. After centrifugation (10,000 g; 4¢XC; 10

min), supernatant was discarded and pellet was dissolved

in DEPC-treated H

2

O. For each gram of tissue, 0.5 £gl

RNase-free DNase (10 U/£gl; Roche Diagnostics GmbH,

Germany) was added and incubated at 37¢XC for 15 min.

Two volumes of isopropanol were then added and mixed.

Tubes were centrifuged at 10,000 g at 4¢XC for 10 min.

After centrifugation (10,000 g; 4

o

C; 10 min), supernatant

was discarded and pellet was washed with 75% ice-cold

ethanol, and then re-centrifuged (10,000 g; 4¢XC; 10 min).

Supernatant was discarded and RNA pellet was dissolved

in RNase-free water. Total RNA concentration was

determined by measuring absorbance at 260 and 280 nm.

Samples were stored at -80¢XC until further use.

The Oligotex mRNA spin-column protocol for isolation

of poly(A) mRNA from total RNA was based on the

manufacturer (Qiagen, Valencia, CA, USA).

labeling 1

st

strand cDNA with Cy3-dye for

mRNA quality control

In an Eppendorf tube, 2 £gl RNA sample containing 80

ng mRNA was mixed with 0.5 £gl oligo d(T)

23

N (2 £gg/£gl;

MD Bio Inc., Taiwan). The tube was incubated at 70¢XC

for 10 min and quickly chilled on ice. After incubation,

1 £gl 5X Superscript II buffer (Invitrogen), 0.5 £gl 0.1 M

DTT, 0.1 £gl 50X dNTPs (25 mM dNTPs except dTTP

at 10 mM), 0.5 £gL Cy3-dUTP (Amersham) and 0.4 £gl

Superscript II reverse transcriptase (200 U/£gl; Invitrogen)

were added into the tube and then incubated at 42¢XC for

2 h. The reaction was cleaned up by a Qiagen PCR clean

up kit, and the eluted DNA was dried and the pellet was

dissolved in 3 £gl TE buffer (pH 8.0). Three £gl of 2X

loading buffer was mixed with 3 £gl of labeled 1

st

strand

cDNA, and run on a 1.2% agarose gel in 1X TAE buffer

at 100 V for 30 min. A labeled DNA molecular weight

marker was used for comparison.

labeling

£f

/HindIII DNA with Cy3 fluorescent

dye as molecular size marker

In an Eppendorf tube containing 4 £gl Klenow 10X

buffer (Amersham), 4 £gl 10X dNTPs (0.25 mM each

except dTTP at 0.09 mM), 4 £gl £f/ HindIII DNA, 2 £gl Cy3-

dUTP, 2 £gl Klenow enzyme (5 U/£gl; Amersham) and 24

£gl H

2

O was incubated at 37

o

C for 1 h. After incubation,

reaction mixture was clean up by a Qiagen PCR clean up

kit. The eluted DNA was dried, resuspended in 10 £gl TE

(pH 8.0) and mixed with 10 £gl of 2X loading buffer. Ten £gl

of mixture was loaded into each well and electrophoresed

together with labeled samples on a 1.2% agarose gel in 1X

TAE buffer at 100 mA for 30 min.

Quality control gel analysis

After electrophoresis, the gel was trimmed to no more

than 25 mm ¡Ñ 75 mm, put on a glass slide (25 mm ¡Ñ

75 mm), and then placed in a 70¢XC baking oven until

completely dry. The dried gel was scanned by a scanner

(GenePix 4000B; Axon Instruments, Foster City, CA,

USA) under 350 nm and analyzed the data with GenePix

version 5.1 software.

cDNA synthesis and labeling for microarray

analysis

For the first strand cDNA synthesis, two 0.2-ml

Eppendorf tubes were labeled with "1" and "2", and 1 £gg

poly(A) mRNA of sample 1 and sample 2 were added

respectively. To each tube, 1 £gg oligo d(T)

23

N (MD Bio

Inc., Taiwan) was added and the final volume was adjusted

to 24.5 £gl by adding DEPC-treated H

2

O. The tubes were

incubated at 70¢XC for 10 min and then immediately

transferred to ice. To each tube, 8 £gl 5X Superscript buffer

(Invitrogen), 4 £gl 0.1 mM DTT, 2 £gl 10 mM dNTP and

1.5 £gl Superscript II reverse transcriptase (200 U/£gl;

Invitrogen) were added and then incubated at 42¢XC for 1

h. After incubation, 0.25 £gl RNase H (5 U/£gl; Amersham)

was added and then further incubated at 37¢XC for 30 min.

The reaction product was clean up by a Qiagen column

according to the manufacturer. The final volume was

adjusted to 28 £gl by H

2

O.

For the second strand cDNA synthesis and labeling, 28

£gl first-strand cDNA product from sample 1 and sample

2 were added to two 0.2-ml tubes labeled with "1" and

"2", respectively. To each tube, 4 £gl Klenow buffer and 1

£gl random primer (3 £gg/£gl; Invitrogen) were added and

incubated at 100¢XC for 2 min. Tubes were put at room

temperature for 5 min. To the tubes labeled with "1" and

"2", 1 £gl Cy3-dUTP (25 nmol; Amersham) and 1 £gl Cy5-

dUTP (25 nmol; Amersham) were added, respectively.

And then to each tube, 4 £gl 10X dNTPs (0.25 mM each

except for dTTP at 0.09 mM; Amersham) and 2 £gl Klenow

DNA polymerase (5 U/£gl; Amersham) were added and

incubated at 37¢XC for 3 h. The reaction product was

clean up according to the user manual of MinElute PCR

purification kit (Qiagen). DNA sample was dissolved in 10

£gl H

2

O.

microarray analysis

To set up cDNA microarray analysis, approximately

12,000 tomato cDNA clones including approximately

6,000 cDNA clones from four tomato root libraries (min-

eral deficiency, pre-anthesis stage, post-anthesis stage and

118

Botanical Studies, Vol. 50, 2009

germination seedlings) purchased from Clemson Univer-

sity Genomics Institute, USA, and approximately 6,000

subtractive clones or normal cDNA clones from individual

laboratories of Integrative Plant Stress Biology (iPSB)

group at the Agricultural Biotechnology Research Center,

Academia Sinica, Taipei, Taiwan (To, 2004), were ampli-

fied by PCR and purified, and then printed onto a slide (25

mm ¡Ñ 75 mm) by an arrayer (Cartesian SynQUAD, USA).

The microarray slide was put in an 80¢XC oven for 5 h and

incubated in 250 ml pre-hybridization buffer (25% for-

mamide; 5X SSC; 0.1% SDS; 0.1 mg/ml BSA) for 1 h at

42¢XC. The slide was washed by immersing in 250 ml ex-

tra-pure H

2

O twice, rinsed in isopropanol, and then dried

by centrifugation (110 g, 10 min; Megafuge 2.0R, Heraeus

Instruments, Germany). For probe preparation, 12.75 £gl

hybridization buffer (25% formaide; 5X SSC; 0.1% SDS),

8.25 £gl labeled cDNA samples, 2 £gl poly(dA)

10

(10 £gg/£gl)

and 2 £gg human Cot-1 DNA (10 £gg/£gl; Invitrogen) were

added into an Eppendorf tube and mixed thoroughly. For

hybridization, the probe was denatured at 94¢XC for 5 min

and then incubated at room temperature for 20 min. Slide

was placed in the hybridization chamber (HybChamber,

GeneMachines), and the probe solution (25 £gl) was added

onto the center of the array. A cover glass (24 mm ¡Ñ 60

mm) (TaKaRa Space Cover Glass TX705, Takara Bio Inc.,

Japan) was carefully placed onto the slide, and 4 drops

of 10 £gl 3X SSC were added on the each lower edge of

the slide. After assembling the hybridization chamber, the

chamber was gently placed into 42¢XC water bath for 16 h

without shaking.

After hybridization, the slide was placed in a jar con-

taining the pre-warm (42¢XC) wash I solution (2X SSC;

0.1% SDS) until the cover glass moves freely away. Then

the jar was shaked (50 rpm) at room temperature for 5 min.

The slide was immediately transferred to wash II solu-

tion (0.1X SSC; 0.1% SDS) and shaked (50 rpm) at room

temperature for 5 min. The slide was repeatedly washed

2 more times in wash II solution. Afterwards, the slide

was placed in wash III solution (0.1X SSC) up and down

for 1 min, and repeatedly washed 4 more times in wash

III solution. After washing, the slide was rinsed with H

2

O

and then dried by centrifugation for 5 min at 110 g. Slides

were scanned (GenePix 4000B scanner; Axon Instruments,

Foster City, CA, USA) as soon as possible. Otherwise,

slides may be stored at -80¢XC for 1 to 2 weeks.

RT-PCR analysis

Total RNA samples from different developmental stag-

es of tomato fruits were treated with RNase-free DNase I

to eliminate genomic DNA contamination. After enzyme

removal by phenol/chloroform, approximately 200 ng of

RNA samples were used to perform reverse transcription-

polymerase chain reaction (RT-PCR) analysis, using a one-

step RT-PCR kit (Qiagen). For gene-specific amplification,

primers PSY-F1 (5¡¦-ATGTCTGTTGCCTTGTTATGG-3¡¦)

and PSY-R1 (5¡¦-TTATCTTTGAAGAGAGGCAGTTT-3¡¦)

specific for tomato phytoene synthase (PSY), prim-

ers ZDS-F1 (5¡¦-ATGGCTACTTCTTCAGCT-3¡¦) and

ZDS-R1 (5¡¦-TCAGACAAGACTCAACTC-3¡¦) specific

for tomato £a-carotene desaturase (ZDS), primers LCY-F1

(5¡¦-ATGGATACTTTGTTGAAA-3¡¦) and LCY-R1 (5¡¦

-TCATTCTTTATCCTGTAA-3¡¦) specific for tomato

lycopene £]-cyclase (LCY), and primers PDS-F1 (5¡¦

-ATGCCTCAAATTGGACTT-5¡¦) and PDS-R1 (5¡¦-CTA-

AACTACGCTTGCAAC-3¡¦) specific for tomato phytoene

desaturase (PDS) were synthesized to amplify full-length

reading frames of PSY (1239 bp), ZDS (1767 bp), LCY

(1503 bp) and PDS (1752 bp), respectively. Reverse tran-

scription was carried out at 50¢XC for 30 min. The PCR

mixtures (20 £gl) were initially denatured at 94¢XC for 5

min, and then subjected to 35 cycles (30 sec at 94¢XC, 30

sec at 55¢XC, 1.5 min at 72¢XC) with a final extension at 72

¢XC for 10 min. PCR products (2 £gl) were analyzed on 1%

TAE-agarose gel stained with ethidium bromide.

ReSUlTS AND DISCUSSION

RNA isolation from different tissues of tomato

plants

To analyze differentially expressed genes in different

tissues of tomato during development, we divided plant

development into 4 stages, namely the young seedling

stage (1-month-old plant), vegetative growth stage

(2-month-old plant), flowering stage (3-month-old plant)

and fruiting stage (4-month-old plant), as shown in Figure

1A. In addition, the typical 8 stages of fruit development

in tomato were also observed (Figure 1B). To simplify

the difficulty in collecting each stage of tomato fruits, we

roughly divided fruit into 3 classes during tomato fruit

ripening: green fruit represented immature and mature

green stages, orange fruit represented breaker, turning and

pink stages, and red fruit represented light-red, red and

over-ripen stages.

Using the TRIzol method coupled with LiCl

precipitation and DNase treatment as described, the yield

of total RNA ranged from 30 to 857 £gg/g of developing

vegetative tissues including leaf, stem, and root, as well

as cotyledon, and 1184 to 1313 £gg total RNA/g of flower

tissue (Table 1). For tomato fruits, the yield of total RNA

ranged from 118 to 183 £gg total RNA/g of green fruit,

15 to 34 £gg total RNA/g of orange fruit, and 20 to 32

£gg total RNA/g of red fruit (Table 1). Our results were

found similar or a little higher than the reported yields

of between 20 to 50 £gg total RNA per gram fresh weight

using protocol developed for tomato fruit (Rodrigues-

Pousada et al., 1990), and those reported yields of between

12 and 100 £gg total RNA per gram fresh weight using

protocols developed for other fruit tissues (Woodhead

et al., 1997; Jaakola et al., 2001; Valderrama-Chairez

et al., 2002). The ratio of A

260

/A

280

was 1.67 to 2.05 in

all samples. In general, the average yield of total RNA

in vegetative tissues was higher than in fruit tissues.

Agarose gel electrophoresis revealed that two major

bands of 28S and 18S rRNA were observed in root, stem,

WANG et al. ¡X RNA isolation from tomato tissues

119

flower and ripening fruit tissues, and 5 major RNA bands

were detected in cotyledon and developing leaf tissues,

indicating that the RNA was not degraded (Figure 2).

RNA samples from ripening fruits (Figure 2) were

selected to examine the quality of total RNA isolated. The

obtained mRNA were further labeled with Cy3 dye and

then checked for quality by gel electrophorsis analysis

(Figure 3). Abundant messages with high molecular sizes

(up to 4 kb) were detected, suggestive of high-quality

mRNA.

Figure 1. Different developmental stages in plants and fruits

of tomat o (Sol anu m lycopersicum c ult iva r C L5915). (A)

Growth stages in tomato plants were divided into four stages:

young seedling (1-month-old), vegetative growth (2-month-

old), f loweri ng (3-mont h-old) and fruiting (4 -m onth-old).

W hite a rrows indicate two cot yledons in a young seedling

pla nt. (B) Typical eight growth stages in tomato fruits.

120

Botanical Studies, Vol. 50, 2009

spots labeled with "a" to "h" in Figure 4A and Figure 4B).

Furthermore, our microarray data proved reproducible

upon comparison with the consistent expression pattern of

the repetitious £\-tubulin 3 (tubA3) sequence among differ-

ent blocks in our microarray slide (e.g., those spots labeled

with "

¡¹

" in Figure 4A appeared as green, those spots

labeled with "

¡¹

" in Figure 4B appeared as red). Interest-

ingly, we found from this study that the expression of the

£\-tubulin 3 gene was also affected by development (i.e.,

higher expression level of this gene in green fruit than

in orange fruit). Previously, a gene list including DnaJ-

like protein, translationally controlled tumor protein, two

£\-tubulins (tubA1 and tubA3), cyclophilin and glyceral-

dehydes-3-phosphate dehydrogenase (GAPDH) had been

compiled as so-called "housekeeping" gene candidates in

Figure 3. Quality control of poly(A) mRNA. Poly(A) mRNA

from green fruit, orange fruit and red fruit of tom ato were

isolated, equal amounts of 80 ng mRNA were used to label

t he 1

st

strand cDNA with a fluorescent dye Cy3, and the

labeled 1

st

strand cDNA product s were separate d on 1.2%

agarose gel in 1X TAE buffer. After electrophoresis, the gel

was t rim med, transferred onto a glass slide, and then com-

pletely dried in a 70¢XC oven. Images were obtained by scan-

ning the dried gel with a GenePix 4000B Scanner.

Figure 2. RNA gel elect rophoresis analysis of total RNA

from tomato. Total RNA from different tissues, as indicated,

were isolated. Equal amounts of 20 £gg RNA from each sam-

ple were separated at a 1.2% agarose gel.

microarray analysis

To genome-wide identify differentially expressed genes

during tomato fruit ripening, the obtained mRNA from

green fruit and orange fruit (Figure 1) were labeled with

different fluorescent dyes, as indicated in Figure 4A and

Figure 4B, and hybridized to our home-made microar-

ray chip containing approximately 12,000 PCR-amplified

fragments. After hybridization, signals were detected by a

highly sensitive laser scanner and analyzed by GeneSpring

software (see the legend in Figure 4). We believe our mi-

croarray data are convincing based on dye swap analysis.

For example, green spots in Figure 4A are considered

as up-regulated genes in green fruit as comparison with

orange fruit, since mRNA isolated from green fruit was

labeled with a green fluorescent dye Cy3 and mRNA iso-

lated from orange fruit was labeled with a red fluorescent

dye Cy5. Therefore, the gene labeled as "1" (green color)

in Figure 4A is believed to be more highly expressed in

green fruit than in orange fruit. However in Figure 4B,

those up-regulated genes in green fruit should appear as

red, since the method for probe labeling has been swapped

(mRNA isolated from green fruit and orange fruit were

labeled with a red fluorescent dye Cy5 and a green fluo-

rescent dye Cy3, respectively). Thus for a particular gene

"1" which appears as red Figure in 4B is equivalent to an

up-regulated gene in green fruit in Figure 4A. In addition,

genes having similar expression pattern between green and

red fruits appear as yellow (a third pseudo color represent-

ing similar expression levels between two samples was

generated by the GeneSpring software), no matter which

labeling method was employed (for example, 8 yellow

WANG et al. ¡X RNA isolation from tomato tissues

121

tomato plants using EST and bioinformatic approaches

(Coker and Davies, 2003); however, our microarray data

clearly showed that £\-tubulin 3 gene is not a housekeeping

gene in tomato plants, at least it is not suitable for studying

tomato fruit ripening.

By analyzing microarray data, we have identified 98

up-regulated genes and 37 down-regulated genes show-

ing at least 5-fold difference in orange fruit as compared

to green fruit. Top 10 putative differentially expressed

genes are listed in Table 2. Among them, a heat-inducible

gene encoding class II small heat shock protein 17.6 kDa

from tomato (LeHSP17.6) has been characterized and

no LeHSP17.6 mRNA could be detected in mature green

tomato fruit (Kadyrzhanova et al., 1998). More recently,

a novel tomato small heat shock protein gene (vis1) was

isolated and showed strong similarity with members of the

small heat shock proteins including LeHSP17.6, and vis1

transcripts were barely detected at early stages (e.g., ma-

ture green) of fruit development but rapidly accumulated

in fruit after onset of ripening with maximum accumula-

tion at the turning stage fruit (Ramakrisha et al., 2003).

Taken together, it is reasonable to expect that higher level

of LeSHP17.6 expression was detected in orange fruit

than in green fruit by our microarray analysis (Table 2).

Currently, we are picking up several clones including

LeSHP17.6 from the gene list for further characterization.

Figure 4. The dye swap method showing reliable reproduction and accuracy in tom ato microarray analysis. Poly(A) mRNA

from two samples of green fruit and orange fruit to be compared were labeled with fluorescent dyes Cy3 (green color) and Cy5

(red color), as indicated in panels A and B, by reverse transcription. T he fluorescent probes were then pooled and allowed to

hybridize under stringent conditions to the tomato array containing approximately 12,000 PCR-amplified gene fragments and

printed on a coated glass microscope slide by a computer-controlled, high-speed arrayer. After hybridization, highly sensitive

laser scanner was used to detect any expression signal in each spot on the slide. Monochrome images from the scanner were

imported into GeneSpring software in which the images pseudo-colored and merged. This image represents the relative expres-

sion in the t wo samples we examined (for example, spots in panel A with green color represent relatively high expression in

green fruit, while spots in the same panel with red color represent relatively high expression in orange fruit). A third color yel-

low was introduced to represent a similar expression level between two samples. The selected 8 spots labeled "1" to "8" repre-

sent up-regulated or down-regulated genes in one method of labeling probes, as shown in panel A; if the labeling method was

effective, the same spots should show up-regulation or down-regulation while another labeling method is employed, as shown

in panel B. This procedure is called the "dye swap" method. In parallel, the selected 8 spots labeled "a" to "h" represent sim ilar

expression pattern between two samples, as confirmed by dye swap method. Spots labeled with an asterisk (¡¹) in each block

represent the same DNA sequence of £\-tubulin 3 gene from Hordeum vulgare, and were used as an internal control for microa r-

ray analysis.

122

Botanical Studies, Vol. 50, 2009

gene cloning, sequence analysis and

construction of cDNA library from the RNA

samples

RT-PCR is a sensitive amplification procedure that

has been used to detect the presence of a gene in a plant

genome. If RNA is a poor template for reverse transcrip-

tion, it is very difficult or no chance to amplify a longer

DNA fragment such as full-length cDNA, and thus, a

good cDNA library cannot be prepared. To further prove

the integrity of the RNA we isolated, full-length cDNA

sequences encoding phytoene synthase (PSY), £a-carotene

desaturase (ZDS), lycopene £]-cyclase (LCY) and phy-

toene desaturase (PDS) in the carotenoid biosynthesis

pathway (To and Wang, 2006) were amplified by RT-PCR

from total RNA of tomato ripening fruits (Figure 5). All

PCR fragments with the expected sizes were detected in

RNA samples amplified by gene-specific primers (Figure

5). Furthermore, these PCR products were purified, cloned

into pGEM-T Easy cloning vector (Promega), and then

sequenced. DNA sequence analysis revealed clearly that

our PSY clone is 1239 bp in length and contains a reading

frame of 412 amino acids. In comparison with the pub-

lished tomato psy1 mRNA (1239 bp; Bartley et al., 1992),

a base change at nucleotide position 583 in the coding

region of our PSY clone is detected, where an adenine is

replaced by a guanine residue. As a result, the correspond-

ing amino acid at position 195 changes from methionine

(ATG) to valine (GTG). Different cultivars (CL5915 ver-

sus MicroTom) may be the possible explanation for this

observation. Thus, we concluded that our PSY clone is

encoded an enzyme catalyzing 2 molecules of geranyl-

geranyl diphosphate (GGDP) into phytoene. Nucleotide

sequence of our tomato cDNA clone encoding phytoene

synthase has been submitted into NCBI database with an

accession number EF650010. The second clone PDS (ac-

cession number EF650011) is 1752 bp in length and con-

tains a reading frame of 583 amino acids. In comparison

with the published tomato pds mRNA (1752 bp; Giuliano

et al., 1993), a base change at nucleotide position 705 in

the coding region of our PDS clone is detected, where an

adenosine is replaced by a guanosine residue. However,

the corresponding amino acid residue at 235 (proline)

is not altered. Thus, we conclude that our clone PDS is

encoded an enzyme catalyzing phytoene into £a-carotene.

The third clone ZDS (accession number EF650012) is

1767 bp in length and contains a reading frame of 588

amino acids. In comparison with the published tomato zds

mRNA (1767 bp; Bartley and Ishida, 1999), alterations

at nucleotide positions 709, 1133 and 1639 in the coding

region of our PSY clone are detected. A different codon

709 from thymidine is replaced by a cytidine residue and

the corresponding amino acid residue is also altered from

serine (TTC) to proline (TCC); a different codon 1133

from thymidine is replaced by a cytidine residue and the

Table 2. Top putative up-regulated or down-regulated genes in tomato orange fruits in comparison to tomato green fruits.

Up- or down-

regulation

Systematic name on array

Description

Fold change in

replicate

Up

LE023I12

Class II small heat shock protein Le-HSP17.6 (tomato)

27, 32

Up

LE022E08

FIN21.18 protein (Arabidopsis thaliana)

22, 32

Up

LE017D11

Metallothionein-like protein

20, 22

Up

LE023G18

Early light-induced protein-like protein (Retama raetam)

20, 47

Up

LE022G10

Heat shock protein 18p (tobacco)

17, 28

Up

LE017B21

Ribosomal protein S20 (Arabidopsis thaliana)

15, 28

Up

LE022H20

Type I small heat shock protein 17.6 kDa isoform (tomato)

15, 13

Up

LE009P11

Similar to AT4g27450 (Arabidopsis thaliana)

15, 13

Up

LE017H05

Zinc-finger protein (Petunia hybrida)

14, 24

Up

LE022K12

Mitochondrial small heat shock protein

14, 15

Down

CLEX1M5_1

Alpha-tubulin 3 (Hordeum vulgare)

16, 15

Down

CLEX1M5_12 Alpha-tubulin 3 (Hordeum vulgare)

15, 15

Down

CLEX1M5_10 Alpha-tubulin 3 (Hordeum vulgare)

14, 12

Down

LE001D09

Argonaute (AGO1)-like protein (Arabidopsis thaliana)

7, 11

Down

LE032C13

Unknown

6, 8

Down

LE030C01

Heat shock protein

6, 8

Down

LE011I11

Unknown protein F19K16.21 (Arabidopsis thaliana)

6, 5

Down

LE013M11

Glutathione peroxidase

6, 5

Down

LE004E24

Chlorophyll synthetase G4

6, 6

Down

LE001B10

Ethylene-responsive methionine synthase

5, 8

WANG et al. ¡X RNA isolation from tomato tissues

123

corresponding amino acid residue is altered from phenyl-

alanine (ATT) to serine (ATC); and a different codon 1639

from adenosine is replaced by a guanosine residue and the

corresponding amino acid residue is altered from alanine

(AGG) to glycine (G GG). To exclude the possibility of

3 alternations in our ZDS clone is due to the random oc-

currence of sequencing error in the Ta q DNA polymerase

during PCR amplification, two more ZDS clones were

picked up from the plate and then sequenced. These two

clones have identical sequence, and is consistent with our

previous ZDS clone we mentioned above. Thus, we be-

lieve that our clone ZDS is encoded an enzyme catalyzing

£a-carotene into lycopene. The fourth clone LCY (acces-

sion number EF650013) is 1503 bp in length and contains

a reading frame of 500 amino acids. In comparison with

the published tomato lcy mRNA (1503 bp; Cunningham et

al., 1996), a base change at nucleotide position 919 in the

coding region of our LCY clone is detected, where a thy-

midine was replaced by a cytidine residue. However, the

corresponding amino acid at position 307 (lysine) is not

altered. Thus, we conclude that our LCY clone is encoded

an enzyme catalyzing lycopene into £]-carotene. In brief,

DNA sequence analysis demonstrated clearly that our

cDNA clones are the same sizes and almost identical to the

corresponding sequences from tomato, suggesting that the

RNA samples we isolated are intact and functional.

In addition, we have also constructed a cDNA library

representing 8 developmental stages of tomato fruits (Fig-

ure 1B). Our tomato fruit cDNA library contains approxi-

mately 1.0 x 10

5

independent clones with inserts ranging

from 0.6 kb to 4 kb (data not shown). In conclusion, the

protocol for RNA isolation presented here allows for the

recovery of high-quality and functional RNA from differ-

ent tissues of tomato suitable for use in tissue-specific or

developmental-specific expression and regulation assays

such as Northern blot, RT-PCR, microarray analysis and

for the construction of cDNA libraries. Furthermore, toma-

to is considered as one of the model species in vegetative

crops and its transformation has been routinely carried out

for both basic and applied researches (Pfitzner, 1998; Bha-

tia et al., 2004; Cortina and Culianez-Macia, 2004), our

method should also be applied to isolate functional RNA

from different tissues in transgenic tomato plants and then

monitor transgene expression by molecular tools.

Acknowledgements. We gratefully acknowledge Dr.

Kenrick J. Deen for critical review on this manuscript. We

deeply appreciate Wei-Chou Huang, former assistant at

Agricultural Biotechnology Research Center, Academia

Sinica, for preparing and analyzing tomato microarray

chips, and our former laboratory members of Shaw-

Ching Soong and Shuh-Chun Chen for growing tomato

plants and technical assistance. We are also grateful to

the Institute of Plant and Microbial Biology, Academia

Sinica, for providing greenhouse facility. This work was

financially supported by grants from National Science

Council (NSC92-2313-B-001-001) and Academia Sinica

Genomics Research Program (94F007-4) of the Republic

of China to Dr. Kin-Ying To.

lITeRATURe CITeD

Bartley, G.E. and B.K. Ishida. 1999. Zeta-carotene desaturase

(Acces sion No. AF 195507) from tomato (P GR99-181).

Plant Physiol. 121: 1384.

Bartley, G.E., P.V. Viitanen, K.O. Bacot, and P.A. Scolnik. 1992.

A tomato gene expressed during fruit ripening encodes an

enzyme of the carotenoid biosynthesis pathway. J. Biol.

Chem. 267: 5036-5039.

Bhatia, P., N. Ashwath, T. Senaratna, and D. Midmore. 2004.

Tissue culture studies of tomato (Lycopersicon esculentum).

Plant Cell Tiss. Org. Cult. 78: 1-21.

Burge ss , J.K. 2004. Overvie w of m icroarrays in genom ic

analysis. In R. Rapley and S . Harbron (eds.), Molecular

Analysis and Genom e Discovery. John Wile y & Sons ,

Chichester, England, pp. 127-165.

F igu re 5. Gel electrophoresis of RT-PCR products.

App rox imat ely 200 ng of DNA-fre e RNA sa mple s fro m

different developmental stages of tomato fruits were ampli-

fied by gene-specific RT-PCR. Forward and reverse primers

were synthesi zed to a mplify full-length reading fram es of

1503 bp, 1752 bp, 1239 bp a nd 1767 bp encoding lycopene

£]-cyclase (LCY; lane 2), phytoene desat urase (PDS; lane 3),

phytoene synthase (PSY, lane 4) and £a-carotene desaturase

(ZDS; lane 5), respectively. PCR products (2 £gl out of 20 £gl)

were analyzed on 1% TAE-aga rose gel stained with ethidium

bromide.

124

Botanical Studies, Vol. 50, 2009

Chang, S., J. P uryear, and J . Cairney. 1993. A sim ple and

efficient method for isolating RNA from pine trees. Plant

Mol. Biol. Reptr. 11: 113-116.

Chi-Manzanero, B., M.L. Robert, and R. Rivera-Madrid. 2000.

Extraction of total RNA from a high pigment plant. Mol.

Biotechnol. 16: 17-21.

Chomczynski, P. 1993. A reagent for single-step simultaneous

isolation of RNA, DNA and proteins from cell and tissue

samples. BioTechniques 15: 532-536.

Chomczyns ki, P. and N. Sacchi. 1987. Single-step method of

RNA isolation by acid quanidinium thiocyanate-phenol-

chloroform extraction. Anal. Biochem. 162: 156-159.

Coker, J.S. and E. Davies. 2003. Selection of candidate

housekeeping controls in tomato plants using EST data.

BioTechniques 35: 740-748.

Cortina, C. and F.A. Culianez-Macia. 2004. Tomato

transformation and transgenic plant production. Plant Cell

Tiss. Org. Cult. 76: 269-275.

Cunningham, F., B. Pogson, Z. Sun, K. McDonald, D.

DellaPenna, and E. Gantt. 1996. Functional analysis of the

beta and epsilon lycopen cyclase enzymes of Arabidopsis

re veals a m echa nism for c ontrol of cycl ic ca rotenoid

formation. Plant Cell 8: 1613-1626.

Duggan, D.J., M. Bittner, Y. Chen, P. Meltzer, and J.M. Trent.

1999 Expression profiling using cDNA microarrays. Nat.

Genet. 21(Suppl. 1): 10-14.

Dumas , Y., M. Dadom o, G.D. Lucca, and P. Grolier. 2003.

Effects of environmental factors and agricultural techniques

on antioxidant content of tomatoes. J. Sci. Food Agric. 83:

369-382.

Giuliano, G., G.E. Bartley, and P.A. Scolnik. 1993. Regulation of

carotenoid biosynthesis during tomato development. Plant

Cell 5: 379-387.

Jaakola, L., A.M. Pirttila, M. Halonen, and A. Hohtola. 2001.

Isolation of high quality RNA from bilberry (Vaccinium

myrtillus L.) fruit. Mol. Biotechnol. 19: 201-203.

John, M.E. 1992. An efficient method for isolation of RNA and

DNA from plants containing polyphenolics. Nucleic Acids

Res. 20: 2381.

Jones, C.S ., P.P. Ia nnetta, M. Woodhead, H.V. Da vies , R.J.

McNicol, and M.A. Tayor. 1997. The is olation of RNA

from raspberry (Rubus idaeus) fruit. Mol. Biotechnol. 8:

219-221.

Kadyrzhanova, D.K., K.E. Vlachonasios, P. Ververidis, and D.R.

Dilley. 1998. Molecular cloning of a novel heat induced/

chilling tolerance related cDNA in tomato fruit by use of

mRNA differential display. Plant Mol. Biol. 36: 885-895.

Liao, Z., M. Chen, L. Guo, Y. Gong, F. Tang, X. Sun, and K.

Tang. 2004. Rapid isolation of high-quality total RNA from

Taxus and Ginkgo. Preparative Biochem. Biotechnol. 34:

209-214.

Malnoy, M., J.P. Reynoird, F. Mourgues, E. Chevreau, and P.

Simoneau. 2001. A method for isolating total RNA from

pear leaves. Plant Mol. Biol. Reptr. 19: 69a-69f.

Mas on, M.G. and S. Schmidt. 2002. Rapid isolation of total

RNA and genomic DNA from Hakea actities. Funct. Plant

Biol. 29: 1013-1016.

Moser, C., P. Gatto, M. Moser, M. Pindo, and R. Velasco. 2004.

Isolation of functional RNA from small amounts of different

grape and apple tissues. Mol. Biotechnol. 26: 95-99.

Pfitzner, A.J.P. 1998. Transformation of tomato. Meth. Mol.

Biol. 81: 359-363.

Ramakrishna, W., Z. Deng, C.K. Ding, A.K. Handa, and Jr.

R.H. Ozminkowski. 2003. A novel small heat shock protein

gene, vis1, contributes to pectin depolymerization and juice

viscosity in tomato fruit. Plant Physiol. 131: 725-735.

Rodrigues -Pous ada, R., M. Va n Mont agu, a nd D. Va n de r

Straeten. 1990. A protocol for preparation of total RNA

from fruit. Technique 2: 292-294.

Salzman, R.A., T. Fujita, K. Zhu-Salzman, P.M. Hasegawa, and

R.A. Bressan. 1999. An improved RNA isolation method for

plant tissues containing high levels of phenolic compounds

or carbohydrates. Plant Mol. Biol. Reptr. 17: 11-17.

Schneiderbauer, A., H. Jr. Sandermann, and D. Erns t. 1991.

Isolation of functional RNA from plant tis sues ri ch in

phenolic compounds. Anal. Biochem. 197: 91-95.

Suzuki, Y., T. Kawazu, and H. Koyama. 2004. RNA isolation

from siliques, dry seeds, and other tissues of Arabidopsis

thaliana. BioTechniques 37: 542-544.

To, K.Y. 2000. Identification of differential gene expression by

high throughput analysis. Comb. Chem. High T. Scr. 3:

235-241.

To, K.Y. 2004. Overview of differential gene expression by high-

throughput analysis. In R. Rapley and S. Harbron (eds.),

Molecular Analysis and Genome Discovery. John Wiley &

Sons, Chichester, England, pp. 167-190.

To, K.Y. and C.K. Wang. 2006. Molecular breeding of

flower color. In J.A. Teixeira da Silva (ed.), Floriculture,

Ornam enta l and P l ant Bi otec hnolog y: Adva nces an d

Topical Issues. Volume I. Global Science Books, Isleworth,

England, pp. 300-310.

Valderrama-Chairez, M.L., A. Cruz-Hernandez, and O. Paredes-

Lopez. 2002. Isolation of functional RNA from cactus fruit.

Plant Mol. Biol. Reptr. 20: 279-286.

Woodhead, M., M.A. Tayor, H.V. Davies, R.M. Brennan, and

R.J. McNicol. 1997. Isolation of RNA from blackcurrant

(Ribes nigrum L.) fruit. Mol. Biotechnol. 7: 1-4.

WANG et al. ¡X RNA isolation from tomato tissues

125