Botanical Studies (2008) 49: 177-188.

*

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

TECHNICAL REPORT

Technical report on the molecular phylogeny of Camellia

with nrITS: the need for high quality DNA and PCR

amplification with Pfu-DNA polymerase

Kunjupillai VIJAYAN and Chih-Hua TSOU*

Institute of Plant and Microbial Biology, Academia Sinica, Taipei, Taiwan 115, ROC

(Received May 17, 2007; Accepted December 6, 2008)

ABSTRACT.

Internal transcribed spacer (ITS) of nrDNA has been widely employed for reconstructing

phylogenetic relationships in plants, especially at the species level. Previous attempts to reconstruct the

molecular phylogeny of Camellia based on nrITS, however, have not succeeded due to technical difficulties.

In order to identify the major factors responsible for these difficulties and also to assess the efficacy of

nrITS in elucidating the interspecific relationships of Camellia, the present investigation was conducted with

seven closely or distantly related species. The purity of the DNA was found to be one of the major factors

affecting the success of PCR amplification and the errors in the sequences. Therefore, an efficient protocol has

been developed for extracting genomic DNA from dried leaf samples of Camellia. The purity of the DNA,

extracted using this method, was quite good as revealed by the A260/A280 ratio, which ranged from 1.84 to 1.89.

Further investigation on the effect of DNA polymerases on PCR induced variations revealed that the PCR

error rate was much higher in Taq-amplified sequences than Pfu-amplified sequences. The effect of the error

on phylogenetic analysis was evident from the wide dispersal of Ta q-amplified sequences across the gene tree

while the Pfu-amplified sequences from the same sample joined together to form a single clade. Our extensive

study of Camellia based on Pfu-amplified ITS sequences showed well-resolved interspecies relationships.

Since the results of the molecular phylogenetic investigation of Camellia needs to be reported in a series, due

to the technical and scientific complexity of the work, in this first report, we provide technical and scientific

insights into the major factors responsible for the failure of the PCR amplification, the occurrence of high

sequencing errors, and their effect on the phylogenetic interpretations. The results further stress the potential

of nrITS in deducing the phylogenetic relationships in Camellia.

Keywords: Camellia; DNA isolation; ITS; PCR error; Pfu; Ta q polymerase.

INTRODUCTION

Molecular phylogeny based on DNA sequences

has recently been used extensively to resolve intricate

problems at various taxonomic levels. The internal

transcribed spacer (ITS) of the nuclear ribosomal DNA

has been the most widely used molecular marker in

resolving phylogeny at the generic and specific levels

in the angiosperms (Baldwin et al., 1995; Alvarez and

Wendel, 2003). More than two-thirds of the related papers

published during 1998 to 2002 and in 2005 included

nrITS in the analyses (Alvarez and Wendel, 2003; Feliner

and Rossello, 2007). The advantages as well as the

disadvantages of nrITS in the phylogenetic application

have also been well explored (Buckler et al., 1997; Alvarez

and Wendel, 2003; Feliner and Rossello, 2007). Although

general concerns have been raised about its sequence

complexity and the existence of infra-species or even

infra-individual variations, which impact the accuracy of

the phylogeny being deduced, its advantages at a specific

level of phylogeny are unsurpassed by any other markers.

Recently, it has become customary that whenever nrITS

is used for phylogenetic analysis, a segment of plastid or

mitochondrial sequences is also incorporated to make a

comparison.

The genus Camellia, the largest genus in the family

Theaceae, is equally important for both the horticultural

and beverage industries. It is mainly distributed in East

Asia, with Southwest China as the center of distribution

(Ming, 2000). Taxonomically, the most popular

classification systems proposed by Chang (1981, 1998)

and Ming (2000) are very different. For example, Chang

recognized 284 species and treated them in 22 sections,

but Ming recognized 119 species and 14 sections only. The

difficulty in reaching a well-accepted classification system

resides in the large number of the species having natural

hybridizations (Ming, 2000) and in the great number of

178

Botanical Studies, Vol. 49, 2008

variations in floral characters. For example, the number of

bracteoles, sepals, and petals is not fixed in many species,

and the degree of filamentous and stylar fusion is variable

within species and difficult to quantify. In addition, floral

size often varies within a single species. In the past decade,

while molecular taxonomy has been overwhelmingly

applied to the ranks of all living organisms, some effort

has undoubtedly been devoted to the genus Camellia

as well. The earliest report on the molecular taxonomy

of Camellia, to our knowledge, was a presentation in a

conference by Thakor and Parks (1997). In this study, four

cpDNA regions (rbcL, trnT-trnL, trnL-trnF, atpB-rbcL)

were sequenced from 20 species which showed great

morphological diversity, but the sequence variability was

as low as 0% to 3%. Later on, in Thakor ¡¦s unpublished

thesis, one more cpDNA region (rpL16 intron) was used,

but it showed high intra-specific variation (Xiao, 2001).

Thakor¡¦s study based on morphology and DNA sequences

suggested that Camellia was paraphyletic (Xiao, 2001).

Xiao (2001) continued the study of molecular phylogeny

in Camellia, in the same laboratory, with two intron

sequences from the RNA polymerase (RPB2) gene. His

extensive survey, which included 149 Camellia species

and around 2000 nucleotides, resulted in a well resolved

tree in which four sections of the genus Camellia formed

distinct clades while the remaining, around three-fifths

of the species, formed a few clades that could be defined

by the color of the petals and the geographic distribution

range (Xiao and Parks, 2003). Nevertheless, this work

contributed many new insights and was different from

previous treatments done by Sealy (1958), Chang (1998),

and Ming (2000). Meanwhile, a few small-scale studies

also appeared like that of Tang and Zhong (2002), who

constructed a molecular phylogeny of 15 golden camellias

with nrITS using DNA from fresh leaf samples. Orel et

al. (2003) studied interrelationships of 32 species with

cpDNA trnL-trnF sequences and ISSR markers and found

that the former showed limited variations while the latter

appeared to be influenced by phylogenetic as well as non-

phylogenetic factors, rendering the tree topology difficult

to interpret. Yang et al. (2006) reported their surveys on

the utility of four DNA regions from 21 Camellia species.

They claimed that the cpDNA trnL-F and rpL16 were too

conserved to be used while nrDNA-ITS was difficult to

sequence technically. A waxy gene, however, appeared

to be potentially useful. Although Yang et al. (2006) was

the first to mention the difficulties of nrITS sequencing

in Camellia, our personal contact with many laboratories

revealed that, due to difficulties in sequencing, most of

the laboratories engaged in the molecular phylogeny of

Camellia with nrITS have abandoned their efforts. Since

our laboratory is involved in the phylogenetic analysis

of the family Theaceae, and nrITS appears to have

more potential than many other currently available gene

sequences in resolving the phylogeny of Camellia, w e

decided to undertake the present investigation to identify

and resolve the major bottlenecks in the phylogenetic

studies of Camellia with nrITS.

Our first major concern was the failure of PCR

amplification. Thus, we concentrated on the quality

of DNA as many contaminating agents¡Xlike

polysaccharides, phenolic compounds, tannins, resins,

and latex, present in the cell as secondary metabolites,

are present in the leaf samples¡Xusually coprecipitate

with DNA and interfere with the activity of the DNA

polymerase enzyme (Merlo and Kemp, 1976; Shioda and

Marakami-Musfushi, 1987; Fang et al., 1992; Panday

et al., 1996). In Camellia sinensis, i.e. tea, one of the

most common beverages, and most likely in many other

species of Camellia as well, a high content of secondary

metabolites is present in the leaves as 40% of the dry leaf

weight of Camellia sinensis comes from polyphenolic

compounds such as flavanols and their glycosides,

leucoanthocyanins, and phenolic acids, and more than

12% is polysaccharides (Graham, 1992). Therefore,

extraction of high quality genomic DNA from dried leaf

samples is challenging but a prerequisite for getting the

right sequences through PCR amplification. Initially we

tried a number of the existing protocols (Murray and

Thompson, 1980; Dellaporta et al., 1983; Rogers and

Bendich, 1988; Doyle and Doyle, 1990; Tai and Tanksley,

1990; Matsumoto et al., 1994; Takeuchi et al., 1994;

Wachira et al., 1995; Struwe et al., 1998), but none of

them was found efficient enough to isolate DNA suitable

for consistent PCR amplification. Therefore, we developed

a protocol suitable for extracting highly purified genomic

DNA, which is detailed in this report.

Our second concern was the possible occurrence of

intragenomic variations among the nrITS repeats, as our

attempts at direct sequencing of PCR product always

failed due to the large amount of noise (unreadable sites)

even though the DNA was of good quality. Thus, cloning

and sequencing of the PCR products became a necessity

to probe into the role of intragenomic heterogeneity

in it. Also proofreading DNA polymerase was tried to

examine the extent of the sequence variations caused by

the routinely used non-proofreading polymerase. Most

of the laboratories involved in phylogenetic studies with

large number of species and genera generally use the

thermostable polymerase from Thermus aquatics (Ta q )

for PCR amplifications due to its remarkable stability at

higher temperature, sensitivity, yield, and low economic

input. However, the PCR products amplified by Ta q

polymerase are subjected to nucleotide changes during

the amplification (Saiki et al., 1988; Tindall and Kunkel,

1988; Kwiatowski et al., 1991; Lundberg et al., 1991;

Kobayashi et al., 1999), and it was found that use of a high

fidelity DNA polymerase such as Pfu, Vent, DeepVent,

or UlTma could reduce the amplification error in PCR

products as compared to Ta q DNA polymerase (Flaman

et al., 1994; Cline et al., 1996). However, the higher cost

of these enzymes remains the major constraint on their

regular use. Considering this and the need of experimental

accuracy, it is prudent to carry out an initial polymerase

fidelity assay with the genomic DNA using non-

proofreading and proofreading polymerases. Therefore,

VIJAYAN and TSOU

¡X Technical report on molecular phylogeny of

Camellia

179

we made a comparison of the fidelity of Ta q polymerase

with that of Pfu polymerase, and the results showed

surprisingly high sequence variations in amplification

made by Ta q polymerases. During the last two years, we

have developed a protocol and succeeded in extending

the nrITS sequencing to about 100 species of Camellia.

The results obtained from the initial experiments were

insightful and useful in resolving intrageneric relationships

of Camellia. We intend to present our findings in a series

of reports to include both technical and scientific details of

the study. In this first report, we provide technical details

of the DNA extraction, methods of nrITS sequencing and

the major factors responsible for the failure of the PCR

amplification and the higher sequencing errors. We also

provide a snapshot of the potential of nrITS to resolve the

intrageneric relationships in Camellia.

MATERIALS AND METHODS

Plant material

Leaf samples of seven species belonging to four

sections, namely, C. assamica, C. euphlebia, C .

formosensis, C. microphylla, C. sasanqua, C. sinensis

var. sinensis, an d C. vietnanmensis were used for the

experiment (Table 1). The materials were kept in a sealable

bag containing silica gel immediately after detachment.

Voucher specimens were made and kept in the HAST

herbarium.

Solutions and reagents

¡E

Extraction buffer (Carlson et al., 1991). The extraction

buffer can be prepared by dissolving 2% (w/v)

cetyltrimethyl ammonium bromide (CTAB), 1.4

M sodium chloride, 20 mM EDTA, 100 mM Tris-

HCl, 1% PEG 8000. Just before use, add 1.5% (v/v)

£]-mercaptoethanol to the extraction buffer.

¡E

2% (w/v) Polyvinyl pyrrolidone (PVP; Sigma, St.

Louis, MO; MW 10,000)

¡E

Phenol: Chloroform: isoamyl alcohol (24:24:1)

¡E

Chloroform: isoamyl alcohol (24:1)

¡E

3 M Sodium acetate solution (pH 4.8)

¡E

Absolute alcohol

¡E

70% (v/v) ethanol

¡E

RNAse-A (10 mg/ml)

¡E

Tris-HCl (pH 8.0)

¡E

EDTA (pH 8.0)

¡E

GENECLEAN III-DNA purification kit.

DNA extraction protocol

One square cm of the dried leaf lamina was powdered

in a tube using a FASTPrep-FPI120 machine (BIO 101

Systems, New York). The extraction buffer along with

2% (w/v) PVP was added into the tube containing the leaf

powder, mixed thoroughly, and kept at 4¢XC for 5 days.

The slurry was subsequently incubated at 65¢XC for 1 h.

After incubation, an equal volume of chloroform: isoamyl

alcohol (24:1) was added, mixed well, and centrifuged at

14,000 rpm for two min. The supernatant was collected

in a fresh tube, and this step was repeated one more time.

To the supernatant, a 1/10 volume of 3 M-sodium acetate

and 2 volumes of pre-cooled ethyl alcohol were added

and mixed slowly by inverting the tube. The tubes were

kept at -20¢XC for 1h to precipitate the DNA before it was

spooled out into a fresh tube and air-dried at 37¢XC. The

air-dried DNA was dissolved in TE Buffer (10 mM Tris

HCl [pH 8.0] and 1 mM EDTA). RNA contamination

was removed by treating the DNA with bovine pancreatic

RNase-I at a final concentration of 40 £gg/ml at 37¢XC

for 30 min. The DNA, after RNAse treatment, was re-

extracted using a modified protocol of Struwe et al. (1998).

Table 1. List of samples, voucher information, and sequence accession numbers.

Species

Section

Collection locality

Voucher

(source origin) Accession number

Chang (1998) Ming (2000)

C. euphlebia

Chrysantha Archecamellia Xishanbana, Yunnan, China Wong Hong,

2004-3-4-a

Taq: EF544731-EF544734

Pfu: EF544735-EF544737

C. microphyla

Paracamellia Tuberculata Internatl. Camellia Gard.,

Zerjiang, China (cultivated)

Tsou-853T (J.

Y. gao)

Taq: EF544763-EF544767

Pfu: EF544768-EF544772

C. assamica

Thea

Thea Tea Improv. Res. Inst.

Nantou, Taiwan (cultivated)

Tsou-455S Taq: EF544721-EF544725

Pfu: EF544726-EF544730

C. formosensis

Thea

Thea Nantou, Taiwan

Tsou-318S Taq: EF544711-EF544715

Pfu: EF544716-EF544720

C. sinensis var. sinensis Thea

Thea West Tian-mu Mt., Zerjiang,

China

Tsou-894T Taq: EF544738-EF544741

Pfu: EF544742-EF544745

C. sasanqua

Oleifera Oleifera Yang-Ming Mt., Taipei,

Taiwan (cultivated)

Tsou-988T Taq: EF544755-EF544757

Pfu: EF544758-EF544762

C. vietnamensis

Oleifera Oleifera Internatl. Camellia Gard.,

Zerjiang, China (cultivated)

Tsou-856T (J.

Y. Gao)

Taq: EF544746-EF544749

Pfu: EF544750-EF544754

180

Botanical Studies, Vol. 49, 2008

An equal volume of Phenol-Chloroform-Isoamyl alcohol

(24:24:1) was added to the tube and mixed gently for 10

min. The upper aqueous layer, obtained after centrifuging

at 14,000 rpm for 2 min, was transferred into a new tube

and an equal volume of Chloroform-Isoamyl alcohol

(24:1) was added and mixed gently for 10 min. The tube

was subsequently centrifuged at 14,000 rpm for 2 min.

From the upper aqueous layer 300 £gl was transferred into

a new tube containing 900 £gl of sodium iodide solution

and 20 £gl of glass milk prepared from GENECLENE

R

kit (BIO101 Systems, New York), mixed gently for 20

min and centrifuged at 14,000 rpm for 30 s to pellet the

glass milk. The supernatant was poured off and 900 £gl of

NEW

TM

wash solution was added. The glass milk pellet

was resuspended by gently breaking the pellet with a

pipette tip and shaking the tube. The tube was centrifuged,

and the supernatant was poured off. The washing process

was repeated thrice with 900 £gl NEW

TM

wash solution,

and a fourth washing was done with 150 £gl NEW

TM

wash

solution. After centrifuging, the supernatant was removed

completely with a fine tipped pipette without disturbing

the pellet. The glass milk pellet was resuspended in TE

buffer and placed at 50¢XC for 10 min to elute the DNA

from the glass milk beads. The tube was centrifuged again

at 14,000 rpm for 2 min, and the supernatant was collected

into a new tube without disturbing the glass milk pellet.

DNA quantification

The quantity and quality of the DNA was estimated

with spectrophotometer (Hitachi U2000; Hitachi High

Technologies America, Inc. Life Sciences Division, San

Jose, CA, USA) as well as by electrophoresis on 1.0%

agarose gel in 1¡ÑTAE buffer.

PCR amplification with Taq and Pfu polymerases

The PCR amplification of the ITS1-5.8S-ITS2 regions

of the nrDNA was achieved with the primer pairs ITSleu1

(5¡¦-GTCCACTGAACCTTATCATTTAG-3¡¦ (Urbatsch et

al., 2000) and ITS4 (5¡¦-TCCTCCGCTTATTGATATGC-3¡¦)

(White et al., 1990). The PCR amplification reaction was

first carried out with Ta q polymerase (1U) in a reaction

mixture containing 5 mM Tris-HCl (pH 8.8 at 25¢XC), 10

mM NaCl, 0.01 mM EDTA, 0.1 mM DTT, 0.5% (v/v)

glycerol, 0.1% (v/v) Triton

R

X-100 (Promega corporation,

Madison, WI, USA), 2 mM MgCl

2

, 20 ng of Genomic

DNA, 100 £gM dNTPs, and 100 £gM primers. Then,

amplifications were carried out with Pfu polymerase

(Promega corporation, Madison, WI, USA) with the same

primers. The PCR cocktail contained Pfu 1U, 20 mM Tris-

HCl (pH 8.8 at 25¢XC), 10 mM KCl, 10 mM (NH

4

)

2

SO

4

,

2.0 mM MgSO

4

, 0.1% (v/v) Triton

R

X-100 and 0.1 mg/ml

nuclease-free BSA (Promega Corporation, Madison, WI,

USA), 20 ng of Genomic DNA, 100 £gM dNTPs, and 100

£gM primers. The PCR thermal cycler (GeneAmp PCR

system 2700, Applied Biosystem, Foster City, CA, USA)

was programmed as 94¢XC for 5 min for denaturing the

DNA followed by 30 cycles of 1 min at 94¢XC, 1 min at 50

¢XC, 1 min at 72¢XC, followed by a final extension of 7 min

at 72¢XC (Mast, 1998). PCR products were electrophoresed

on 1.0% agarose gel in TAE buffer, stained with ethidium

bromide (2 £gg/ml), and visualized under ultraviolet light.

Cloning and sequencing

PCR products were purified with QIAquick PCR

purification kit (QIAGEN Inc., Valencia, CA, USA). PCR

products of each sample were cloned to PCR II-TOPO

cloning vector and transformed into chemically competent

Ecoli- DH5£\

TM

-T1

R

cells (provided with the kit) following

the instructions of the manufacturer (Invitrogen Life

Technologies, Carlsbad, California, USA). Colonies were

cultured overnight at 37¢XC on LB (Luria-Bertani medium)

ampicillin/IPTG (isopropyl £]-D-1-thiogalactopyranoside)/

X-gal (5-bromo-4-chloro-3-idolyl beta-D-galactoside)

selective medium. Plasmids from white colonies were

isolated using Mini-M

TM

plasmid DNA extraction system

(VIOGENE, Sunnyvale, California, USA). The extracted

plasmids were digested with EcoR1 and tested on 1%

agarose gel for inserts. For each species, plasmids from

five colonies were sequenced using an ABI PRISM

dye terminator cycle sequencer, model 3700 (Applied

Biosystems, Foster City, CA, USA).

The full sequences were generated by aligning both

forward and reverse sequences of the same clone, and

any variation observed in the sequences was checked by

examining corresponding peaks in the chromatograms of

both forward and reverse sequences. Sequence alignments

were constructed with the help of the PILEUP program of

GCG, version 8.1 (Genetic Computer Group, 1994). Base

composition and length variability were estimated with

BioEdit, version 5.0.9 (Hall, 1999). Intragenomic variation

was calculated from the multiple clones as mean number of

nucleotide differences per site between pairs of sequences

according to Kimura¡¦s two-parameter model using MEGA

3.1 (Kumar et al., 2004). Standard errors of the sequence

divergence were estimated by applying 1000 bootstrap

replications. Total number of transitions and transversions

present within a set of sequences was estimated with

Tamura-Nei (gamma) algorithm (Tamura and Nei, 1993)

as implemented in MEGA 3.1. The PCR error rate was

calculated from the observed error frequency on the basis

of pair-wise comparison by taking into account the number

of doubling cycles using the formula (2 ¡Ñ observed

frequency) / (number of cycles) as described by first

Gelfand and White (1990) and later by Kwiatowski et al.

(1991). While calculating the PCR error, the variable sites

inherently present in the sequence were eliminated based

on the existence of patterns of Pfu-amplified sequences

present in each sample. For instance, in the aligned matrix

of the Pfu-amplified sequences of each species, if two or

more sequence types could be recognized, the variable

sites resulting from such differences in the sequence types,

as indicated in Figure 1, were identified and eliminated

from further analyses of errors.

The gene tree was generated with PAUP* 4.0b10

(Swofford, 2001) using a maximum likelihood algorithm.

VIJAYAN and TSOU

¡X Technical report on molecular phylogeny of

Camellia

181

Figure 1. Intragenomic variations inherently present in the sequences of Camellia assamica; portions highlighted with rectangles.

182

Botanical Studies, Vol. 49, 2008

To choose the nucleotide substitution model for the

maximum likelihood analysis, the computer program

Model Test 3.06 (Posada and Crandall, 1998) was used.

This program employs two statistical methods, a likelihood

ratio test (LRT) and an Akaike information criterion (AIC,

Akaike, 1974), to find out the best fitting model to be used

for the subsequent analysis with PAUP* 4.0b10. Based on

the results, the likelihood settings from the best-fit model

(GTR+G) were selected and implemented in a maximum

likelihood analysis with PAUP* 4.0b10 (Swofford,

2001). A heuristic search, containing 100 random taxon-

addition replicates, TBR branch swapping and MulTrees,

Collapse and Steepest Descent options were in effect, was

conducted with no upper limit imposed on the trees held

in memory. To ascertain the relative degree of support for

branches in the cladograms, jackknife (Farris et al., 1996)

support was estimated with 100 replicates using heuristic

searches and a random addition of sequences. Trees were

rooted using Pyreneria melanogaster as outgroup since

Pyrenaria (Tutcheria) is so far known as the sister genus

of Camellia (Prince and Parks, 2001).

RESULTS AND DISCUSSION

DNA quality

The DNA isolated from the dry leaf samples of

Camellia using our protocol was of high quality as is

evident from spectrophotometric and gel electrophoresis

analyses. The spectrophotometric analysis revealed that

the absorption ratio of the DNA at 260/280 nm was in the

range of 1.84 (C. vietnamensis) and 1.89 (C. formosensis).

A nearly 2.0 ratio indicates little contamination from

proteins, polyphenols, or polysaccharides in the DNA.

Electrophoresis of the DNA on agarose gel revealed

a conspicuous band nearly 22 kb in size with a little

shearing of the DNA. This may have resulted from the

purification with the GENECLEAN

R

III kit as shearing of

DNA with higher molecular weight (>10 kb) is a problem

often encountered with this kit (Application manual;

GENECLEAN

R

III kit, BIO101 Systems, New York).

This purification step was, however, essential as the DNA

obtained either from the CTAB extraction method alone

or from the method of Struwe et al. (1998) alone was not

amplified by PCR. Further, it should be noted that this

little shearing of the DNA has not interfered with the

subsequent PCR amplification of the DNA. The major

components we added into the extraction buffer were PVP,

CTAB, and PEG8000 to remove polyphenolic compounds,

polysaccharides, and proteins. PVP is reported to form

complex hydrogen bonds with phenolic compounds and

co-precipitates with cell debris upon cell lyses, and upon

centrifugation in the presence of chloroform the PVP

complexes accumulate at the interface between the organic

and the aqueous phases (Kim et al., 1997; Barnwell et

al., 1998). Similarly, CTAB, a cationic detergent which

solubilizes membranes, binds to fructans and other

polysaccharides to form complexes that are removed

during chloroform extraction. PEG-8000 also removes

proteins and polysacchriodes (Agudo et al., 1995).

Therefore, in our protocol most of the phenolics and

polysaccharides must have been removed in the first phase

of the extraction. In the second phase, the purification

steps with GENECLEAN

R

III kit removed the remaining

proteins, polyphenolics, and polysaccharides along with

tannins. This kit allows the DNA to bind onto the silica

particles present in the EZ-GLASSMILK under high salt

concentrations and to release from them under low salt

concentrations. The DNA, while adhering to the silica

particles, is washed many times to remove all contaminants

from it. Thus at the end, though the DNA was devoid of

most contaminants, the process left it slightly sheared.

Singh et al. (1999) also reported high degradation of DNA

upon its extraction from black tea, attributing it to the

extreme processing of the tea leaf and to the binding of

some of the phenolic compounds to the DNA upon cell

lyses as reported earlier (John, 1992). Nevertheless, in the

present study DNA shearing was insufficient to interfere

with PCR amplification.

PCR amplifications with Taq and Pfu

polymerases

The DNA extracted with our method was successfully

amplified with Ta q and Pfu DNA polymerases. Among

the 70 clones, i.e., five clones from each of the Ta q and

Pfu PCR amplifications for each of the seven samples, 63

clones were successfully sequenced (Table 1). However,

attempts to sequence the PCR products directly, without

cloning, were mostly unsuccessful. Sequences of the five

clones from each PCR amplification showed a certain

degree of intragenomic variability in the form of indels and

substitutions (Table 2). Those variations present among

the clones could be the main reason direct sequencing

of the PCR products failed. Further, it is obvious that

when sequences amplified with the Ta q polymerase

were compared with Pfu-amplified ones, the former

exhibited much higher intragenomic variability though

Taq polymerase is much commonly used in routine studies

than Pfu enzyme. The effect of the two DNA polymerases

on the sequence variability will be explicated in the next

paragraph.

The length of the ITS1-5.8S-ITS2 regions, of the seven

taxa studied, varied from 556 bp in Camellia sasanqua to

682 bp in C. microphylla when amplified with Ta q DNA

polymerase and varied from 634 bp in C. sinensis to 666

bp in C. microphylla when amplified with Pfu polymerase

(Table 1). The average length of sequences with Ta q was

643.52, and with Pfu 644.51. The average G+C content

of the sequences with Ta q was 66.12, and with Pfu 67.57.

Thus, both the length and G+C content of the sequences

amplified by Pfu polymerase were slightly higher.

Sequence variability

The substitution rate (cumulative transitions and

transversions) and indels were higher among Ta q -

amplified sequences than among Pfu-amplifed sequences

VIJAYAN and TSOU

¡X Technical report on molecular phylogeny of

Camellia

183

(Table 2). The average substitution+indel in sequences

amplified with Ta q polymerase was highest in C. euphebia

(15.8 per sequence) and was least in C. vietnamensis (3.5

per sequence) while the same in Pfu amplified sequences

was highest in C. sinensis (2.8 per sequence) and least

i n C. euphlebia (0.2 per sequence). When all the Taq -

and Pfu-amplified sequences of the same sample aligned

together and a consensus sequence was generated for each

species, the most common variations observed were C->T

and G->A in both Taq - and Pfu- amplified sequences.

Hofreiter et al. (2001) also found such biased substitutions

in the PCR products of DNA from ancient samples. They

found that a high frequency of deoxyadenosine residues

was incorporated to opposite positions where the template

carries deoxycytidine residues. One of the reasons for such

a high rate of G->A and C->T substitution is the tendency

of the Ta q polymerase to add deoxyadenosine residues

when it reaches the end of templates (Clark, 1988).

This has been shown to cause substitutions opposite to

deoxycytidine residues when degraded DNA or DNA with

high contaminants is used for PCR amplification (Kwok

et al., 1990; Paabo et al., 1990). Another reason for these

deamination-like substitutions in sequences amplified by

Ta q polymerase could be the preferential amplification

of templates with methylated residues. Methylation has

been implicated in nucleolar dominance phenomena, and

the resulting non-functional rDNA copies (pseudogenes)

therefore typically show patterns of deamination-like

substitutions (C->T and G->A) at methylation sensitive

sites relative to their active counterparts (Muri et al., 2001;

Marquez et al., 2003). Also noteworthy is that a higher

occurrence of C->T and G->A substitutions was noticed in

this study when G and C were consecutive, either in GC or

CG order, in both Ta q and Pfu-amplified sequences.

In order to prevent the inherently-present sequence

variations from influencing the rate of PCR/sequencing

errors, we identified and removed the variable sites

resulting from the intragenomic variations inherently

present in the nrDNA arrays. Out of the seven species

studied, three, namely C. assamica, C. microphylla, C.

sasanqua, showed definite patterns of variations. The

variations at the bases 12 to 15, 169 to173, 388, 390, 392,

473, and 620 (Figure 1) in C. assamica, at 9 to 12 in C.

microphylla, an d at 111 and 6 54 i n C. sasanqua were

considered variations inherently present in the nrITS

arrays, for they were consistently present in more than two

sequences amplified by Pfu-DNA polymerase. Therefore,

these variables sites were excluded from further analyses

of PCR/sequencing errors.

The mean pair-wise distance (kimura-2-parameters)

revealed that the highest distance was observed among

sequences amplified with Ta q polymerse (Table 2). The

number of indels present among the clones of each species

varied from 3.0 to 28.0 for Ta q and 0.0 to 5.0 for Pfu. This

high frequency of indels in Ta q-amplified sequences must

be due to PCR error. The PCR error rates also showed

considerable differences between Ta q - and Pfu-amplified

sequences. The error rate in the Ta q-amplified sequences

varied from 2.3 ¡Ñ 10

-4

/site/duplication in C. vietnamensis

to 1.4 ¡Ñ 10

-3

/site/duplication in C. euphlebia while in the

Pfu- amplified sequences it varied only from 0.0 ¡Ñ10

-5

/site/

duplication in C. euphlebia to 1.5 ¡Ñ 10

-4

/site/duplication

in C. sinensis var. sinensis (Table 3). This difference in

the error rates of Ta q and Pfu polymerases corroborates

the earlier reports that the error rate of Pfu polymerase

Table 2. Variations in length and G-C content of the sequences amplified with Taq and Pfu DNA polymerases.

Sample

PCR enzyme No. sequences

obtained

Range of

length (bp)

Range of G-C

content (%) Indels

Substitutions

ti* tv Total

C. euphlebia

Ta q

4

633-664 62.17-67.32

6 42 15 57

Pfu

3

647-649 66.72-66.77

1

0 0 0

C. microphylla

Ta q

5

650-682 66.86-68.46 28 14 9 23

Pfu

5

653-666 67.84-68.58

3

2 1 3

C. assamica

Ta q

5

640-658 62.97-68.13

6 17 4 21

Pfu

5

655-657 67.73-68.09

3

3 1 4

C. formosensis

Ta q

5

632-648 60.44-68.36

7 19 4 23

Pfu

5

646

67.96-68.27

0

1 0 1

C. sinensis var. sinensis Ta q

4

616-642 62.05-68.14 14 20 6 26

Pfu

4

634-639 67.87-68.45

5

4 2 6

C. sasanqua

Ta q

4

556-665 68.20-68.50 25 20 6 26

Pfu

5

657-661 67.37-68.08

3

3 0 3

C. vietnamensis

Ta q

4

635-664 66.67-67.62

5

7 2 9

Pfu

5

635

66.77-67.24

0

3 1 4

*ti: transition; tv: transversion.

184

Botanical Studies, Vol. 49, 2008

was 5 to 30 fold lower than the error rates of other proof

reading and non-proofreading enzymes (Cline et al., 1996;

Flaman et al., 1994). It has been noticed as well that the

Ta q polymerase has problems in amplifying regions with

high G+C contents (Innis et al., 1988). For instance, while

using Ta q polymerase to amplify a G+C-rich region of

the human 18S rRNA gene, Cariello et al. (1991) found

formation of deletions in the hairpin regions. This finding

supports an earlier report that using Ta q polymerase in

PCR amplification of sequences with secondary structures

may give rise to difficulties in the form of deletions and

amplification of unknown regions (Triglia et al., 1988;

Ochman et al., 1988). Since the ITS regions of Camellia

have an average of 66% GC content, which is towards

the high end of the range recorded for plants (Baldwin

et al., 1995), and do form secondary structures during

amplifications, the aforementioned problems in the PCR

amplification with Ta q polymerase can be anticipated. We

must note that the PCR error rates of the Pfu amplified

sequences in the seven Camellia samples (0.0 ¡Ñ 10

-5

, 6.0

¡Ñ 10

-5

, 8.1 ¡Ñ 10

-5

, 2.0 ¡Ñ 10

-4

, 1.5 ¡Ñ 10

-5

, 6.07 ¡Ñ 10

-5

, 8.4

¡Ñ 10

-5

) were higher than the commercially reported error

rate of 1.3 ¡Ñ 10

-6

for Pfu DNA polymerase (Slater et al.,

1998). This could be due to the high G+C content and

the tendency to form secondary structures in the nrITS

of Camellia, as the commercially reported error rate was

estimated from the E. coli lac I gene, which is devoid of

the above problems. Another problem with the Camellia

nrITS is the presence of mutational hot spots such as

mononucleotide stretches and microsatellites, which

are particularly prone to PCR induced errors (Lapegu et

al., 1999). An error rate of one in 1000 bp at each cycle

results in one error in 400 bp in the final product after

25 cycles (Saiki et al., 1988). Hence, Kobayashi et al.

(1999) opined that to obtain sequence data free from

PCR error, sequences for several clones are to be made

and compared. In the light of these factors, we stress that

the use of Pfu is essential to reducing PCR errors in GC-

rich regions like the nrITS of Camellia. Furthermore, the

findings of this study may very well explain the failure of

previous attempts to sequence nrITS in Camellia if they

were performed by routinely used Taq polymerase on low

quality DNA.

Phylogenetic applications

Phylogenetic analyses were made to examine the

usefulness of nrITS in Camellia and the impact of different

DNA polymerases on the analyses. The consequence

of not eliminating PCR-induced variability from the

phylogenetic analysis was clear from the gene trees

(Figures 2, 3). From the gene tree obtained from both Ta q

and Pfu DNA polymerases (Figure 2), it could be observed

that those clones amplified with Ta q polymerase showed

different patterns of grouping. Clones of the same species

either joined the sequences of other species or remained

as singletons. For instance, the clone C. formosensis-4 Ta q

joined with the clade formed by clones of C. euphlebia.

Similarly, the clones C. assamica-3 Ta q , C. formosensis-

5 Ta q, C. sasanqua-1 Taq , an d C. microphylla-5 Taq

remained as isolates showing no definite relationship

Table 3. Sequence divergence calculated using Kimura-2-parameter model the PCR error rates in the sequences amplified with Ta q

and Pfu DNA polymerases.

Species

DNA polymerase No. of clones Kimura-2-parameter distance Rate of PCR error*

C. euphlebia

Ta q

4

0.097¡Ó0.009

1.4¡Ñ10

-3

Pfu

3

0.000¡Ó0.000

0.0¡Ñ10

-5

C. microphylla

Ta q

5

0.035¡Ó0.005

4.4¡Ñ10

-4

Pfu

5

0.005¡Ó0.002

6.0¡Ñ10

-5

C. assamica

Ta q

5

0.033¡Ó0.005

4.2¡Ñ10

-4

Pfu

5

0.006¡Ó0.002

8.1¡Ñ10

-5

C. formosensis

Ta q

5

0.085¡Ó0.010

4.7¡Ñ10

-4

Pfu

5

0.002¡Ó0.001

2.0¡Ñ10

-5

C. sinensis var. sinensis

Ta q

4

0.052¡Ó0.007

6.8¡Ñ10

-4

Pfu

4

0.009¡Ó0.003

1.5¡Ñ10

-4

C. sasanqua

Ta q

4

0.044¡Ó0.007

6.8¡Ñ10

-4

Pfu

5

0.006¡Ó0.002

6.07¡Ñ10

-5

C. vietnamensis

Ta q

4

0.017¡Ó0.004

2.5¡Ñ10

-4

Pfu

5

0.006¡Ó0.002

8.4¡Ñ10

-5

*

Rate of nucleotide substitution/site/per duplication, estimated after eliminating the variable sites that are inherently present in the

sequences.

VIJAYAN and TSOU

¡X Technical report on molecular phylogeny of

Camellia

185

with any other groups. In contrast, all clones of the

same species amplified with Pfu polymerase grouped

together into one clade. Thus, it is clear that the Ta q DNA

polymerase induced PCR errors in the sequences and

greatly affected the topology of the cladogram. When

the Pfu-amplified sequences were analyzed separately,

all clones of the same taxon grouped together, except

i n C. sinensis (Figure 3). In this gene tree depicting

the interspecies relationships of the seven species, four

sections were well demonstrated. Species of C. assamica,

C. formosensis, an d C. sinensis belonging to Section

Thea are so closely related that they are often treated as

three varieties under C. sinensis (Su et al., 2007). The 14

sequences of these three species formed one clade, with

the five sequences of C. formosensis and four of the five

sequences of C. assamica forming their subgroups. The

clones of C. sinensis were not held together, and this

was not surprising as C. sinensis has a long history of

domestication, naturalization, and cultivation. The two

species, C. sasanqua and C. vietnamensis, belonging to

the same section viz., sect. Oleifera, are grouped together

in a clade with 100% support and sequences of each

species further joined together as subgroups. As for C.

euphebia and C. microphylla, they belong to the sections

Chrysantha and Paracamellia, respectively (Chang, 1998),

and their sequences form distinct clades. In summary, the

four sections (Chrysantha, Oleifera, Paracamellia, and

Thea) as well as the species within the sections can be

well separated in the gene tree generated by the Pfu-DNA

polymerase amplified sequences.

Figure 2. Maximum likelihood 50% majority rule consensus

tree derived from the sequences of seven species of Camellia

amplified with both Taq and Pfu polymerases.

Figure 3. Maximum likelihood 50% majority rule consensus

tree derived from the sequences of seven species of Camellia

amplified with only Pfu polymerases.

186

Botanical Studies, Vol. 49, 2008

In short, this investigation has demonstrated that

molecular phylogeny with nrITS has great potential to

resolve many vexing problems associated with both

intersectional and intrasectional relationships of the

Camellia. However, this is possible only if high quality

DNA can be extracted from the dried leaf samples and

proofreading enzyme like Pfu polymerase can be used for

PCR amplification to reduce the PCR/sequencing errors.

The DNA extraction protocol developed in this study

could be of much use to this end. Furthermore, owing

to the possibility of pseudogenes being present in the

genome, sequences from multiple clones have to be used

to generate a gene tree to distinguish the functional genes

from the pseudogenes and to minimize the interference of

pseudogenes in the phylogenetic analysis.

Acknowledgement. We would like to thank Mong-Huai,

Su, Hong Wong and Wen-Ju Zhang for assistance with

sample collection and Yaw-wen Yang for technical advice.

This research was supported by the National Science

Council, Taiwan, Republic of China (Project NSC95-

2621-B-001-009-MY2).

LITERATURE CITED

Agudo, L.C., P. Gavidia, P. Perez-Bermudez, and J. Segura.

19 95 . P E G pr ec i pi t at i on, a re q ui re d s te p fo r P C R

amplification of DNA from wild plants of Digitalis obscura

L. BioTechniques 18: 766-768.

Akaike, H. 1974. A new look at the statistical model

identification. IEEE Trans. Contr. 19: 716-723.

Alvarez, I. and J.F. Wendel. 2003. Ribosomal ITS sequences and

plant phylogenetic inference. Mol. Phylogenet.. Evol. 29:

417-434.

Baldwin, B.G., M.J. Sanderson, J.M. Porter, M.F.

Wojciechowski, C.S. Campbell, and M.J. Donoghue. 1995.

The ITS regions of nuclear ribosomal DNA: a valuable

s o urce of e vid enc e on a ngio sp erm phy log eny. Ann .

Missouri Bot. Gard. 82: 247-277.

Barnwell, P., A. N. Blanchard, J.A. Bryant, N. Smirnoff, and A.F.

Weir. 1998. Isolation of DNA from highly mucilaginous

succulent plant Sedum telephium. Plant. Mol. Biol. Reptr.

16: 133-138.

Buckler, E.S., A. Ippolito, and T.P. Holtsford. 1997. The

evolution of ribosomal DNA: divergent paralogues and

phylogenetic implications. Genetics 145: 821-832.

Cariello, N.F., W.G. Thilly, J.A. Swenberg, and R.S. Thamas.

1991. De letion m utagenes is duri ng pol ymeras e cha in

reaction: dependence on DNA polymerase. Gene 99:

105-108.

Carlson, J.E., L.K. Tulsieram, J.C. Glaubitz, V.W. K Luk, C.

Kauffeldt, and R. Rutledge. 1991. Segregation of random

amplified DNA markers in F1 progeny of conifers. Theor.

Appl. Genet. 83: 194-200.

Chang, H.T. 1981. A taxonomy of the genus Camellia. Acta

S c i ent a rum Na tur al ium . Univ ers i ta t is . S u nya ts e ni,

Monographic series 1: 1-180.

Chang, H.T. 1998. F lora of Reipublicae Popularis Sinicae,

Delectis F lorae Republicae Popularis S inicae, Agendae

Academiae S inicae Edita, Tomus, Beijing Science P ress,

China, Vol. 49, pp. 101-113.

Clark, J.M. 1988. Novel non-templated addition reactions cata-

lyzed by prokaryotic and eukaryotic DNA polymerases.

Nucleic Acids Res. 16: 9677-9686.

Cline, J., J.C. Barman, and H.H. Hogrefe. 1996. PCR fidelity of

Pfu polymerase and other thermostable DNA polymerases.

Nucleic Acids Res. 24: 3546-3551.

Dellaporta, S.L., J. Wood, and J.B. Hicks. 1983. A plant DNA

mini preparation a. Vers ion II. Plant Mol. Biol. Rep. 1:

19-23.

Doyle, J.J. and J.L. Doyle. 1990. Isolation of plant DNA from

fresh tissue. Focus 12: 13-15.

Fang, G., S. Hamm er, and R. Groumet. 1992. A quick and

inexpensive method for removing polysaccharides from

plant genomic DNA. Biofeedback 13: 52-54.

Farris, J.S., V.A. Albert, M. Kallersjo, D. Lipscomb, and A.G.

Kluge. 1996. Parsimony jackknifing outperforms neighbor-

joining. Cladistics 12: 99-124.

Feliner, G.N. and Rossello. 2007. Better the devil you know.

Guidelines for insightful utilization of nrITS in species-

level evolutionary studies in plants. Mol. Phylologenet.

Evol. 44: 911-919.

F lama n, J .M., T. Frebourg, V. More au, F. Charbonnie r, C.

Martin, C. Ishioka, S.H. Friend, and R. Iggo. 1994. A rapid

PCR fidelity assay. Nucleic Acids Res. 22: 3259-3260.

Gelfand, D.H. and T.J. White. 1990. Thermostable DNA

polymerases. In M.A. Innis, D.H. Gelfand, J.J. Sninsky and

T.J. White (eds.), PCR Protocols: a Guide to Methods and

Applications. Academic Press, London, pp. 129-141.

Graham, H.N. 1992. Green tea composition, consumption, and

polyphenols chemistry. Prev. Med. 21: 334-350.

Hall, T.A. 1999. BioEdit: a user friendly biological sequence

alignment editor and analysis program for windows 95/98/

NT. Nucleic. Acid. Sym. Ser. 41: 95-98.

Hofreiter, M., V. J aenicke, D. S erre, A. Haeseler Av. and S.

Paabo. 2001. DNA sequences from multiple amplifications

reveal artifacts induced by cytosine deamination in ancient

DNA. Nucleic Acids Res. 29:4793-4799.

Innis, M.A., K. B. Mayamboo, D.H. Gelfand, and M.A. Brow.

1988. DNA se que nc ing with Ther mus aquat icus DNA

polym eras e and direct s equencing of DNA polym eras e

chain reaction-amplified DNA. Proc. Natl. Acad. Sci. USA

85: 9436-9440.

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

DNA from plants containing polyphenolics. Nucleic Acids.

Res. 20: 2381.

Kim, C.S., C.H. Lee, J.S . Shin, Y.S . Chung, and J.E. Hyung.

1997. A simple and rapid method of isolation of high quality

genomic DNA from fruit trees and conifers us ing PVP.

Nucleic Acids Res. 27: 66-58.

VIJAYAN and TSOU

¡X Technical report on molecular phylogeny of

Camellia

187

Kobayashi, N., K. Tamura, and T. Aotsuka.1999. PCR error

and molecular population genetics. Biochem. Genetics 3:

318-321.

Kumar, S., K. Tamura, and M. Nei. 2004. MEGA3: Integrated

software for molecular evolutionary genetics analysis and

sequence alignment. Briefings in Bioinformatics 5: 150-163.

Kwiatowski, J., D. Skarecky, S. Hernandez, D. Pham, F. Quijas,

and F.J. Ayala. 1991. High fidelity of polymerase chain

reaction. Mol. Biol. Evol. 8: 884-887.

Kwok, S ., D.E. Kellog, N. Mc Kinney, D. Spas ic, L . Goda,

C. Levenson, and J.J. S ninsky. 1990. Effects of primer-

template mismatches on polymerase chain reaction: human

immunodeficiency virus type I model studies. Nucleic Acids

Res. 18: 999-1005.

Lapegue, S.D., M.H. Pemonge, L. Gielly, P. Taberlt, and J. Petit.

1999. Amplification of oak DNA from ancient and moern

wood. Mol. Ecol. 8: 2137-2140.

Lun dberg, K.S., D.D. Sh oem aker, M.W.W. Adam s, J . M.

Short, J .A. S orge, and E.J . Marhur. 1991. High-fidelity

am pli fi cat ion us ing a the rmos tabl e DNA polyme ras e

isolated from Pyrococcus furiosus. Gene 108: 1-6.

Marquez, L.M., D.J. Miller, J.B. MacKenzie, and M.J.H. van

Oppen. 2003. Ps eudoge nes contribute to the e xtrem e

divers ity of nuclea r ribos omal DNA in the ha rd c ora l

Acropa. Mol. Bio. Evol. 20: 1077-1086.

Matsumoto, S., A. Takeuchi, M. Hayatsu, and S. Kondo. 1994.

Molecular cloning of phenyle alanine ammonia lyase cDNA

and classification of varieties and cultivars of tea lantsusing

the tea PAL cDNA as a probe. Theor. Appl. Genet. 89:

671-675.

Merlo, D.J. and J.D. Kemp. 1976. Effect of polysaccharides on

kinetics of DNA. Plant Physiol. 58: 1522-1526.

Ming, T. 2000. Monograph of the genus Camellia. Kunming

Institute of Botany, Chinese Academy of Sciences, Yunnan

Science and Technology Press, Kunming, P. R. China.

Muri, G., C.C. Fleming, and C. Schlotterer. 2001. Three

divergent rDNA clusters predate the species divergence in

Quercus petraea (Matt.) Liebl. and Quercus robur L. Mol.

Biol. Evol. 18: 112-119.

Murray, M.G. and W.F. Thompson. 1980. Rapid is olation of

high-molecular-weight plant DNA. Nucle Acids Res. 8:

4321-4325.

Ochm an, H., A.S . Ge rber, a nd D.L. Hart l. 1988. Genet ic

appl ica tion of an i nvers e polym eras e c hain re act ion.

Genetics 120: 621-623.

Orel, G., A. Marchent, and G. Richards. 2003. Evolutionary

relationships of yellow-flowered Camellia species from

southeast Asia. Inter. Camellia J. 35: 88-96.

Paabo, S., D.M. Irwin, and A.C. Wilson. 1990. DNA damage

prom otes jumping between templates during enzym atic

amplification. J. Biol. Chem. 265: 4718-4721.

Posada, D. and K.A. Crandall. 1998. MODELTEST: testing the

model of DNA substitution. Bioinformatics 14: 817-818.

Prince, L.M. and C.R. Parks. 2001. Phylogenetic relationships

of Theaceae inferred from chloroplast DNA sequences data.

Amer. J. Bot. 88: 2309-2320.

Rogers, S.U. and A.J. Bendich. 1988. Extraction of DNA from

m illigram am ounts of fresh, herbarium and mummified

plant tissues. Plant. Mol. Biol. 5: 69-76.

Saiki, R.K., D.H. Geifand, S. Stoffei, S.J. Scharf, R. Higuchi,

G.T. Horn, K.B. Mulfis, and H.A. Erlich. 1988. P rimer-

di r e ct e d e n z ym a t i c a m pl i f ic a t i on o f DNA wi th a

thermostable DNA polymerase. Science 239: 487-491.

S ealy, J.R. 1958. A Revis ion of genus Camellia. The Royal

Horticultural Society, pp. 168-191.

S hiod a, M. and K. Mara kam i-Muofus hi. 198 7. Se lec tiv e

inhibition of DNA polymerase by a polysaccharide purified

from slime of Physarum polycephalum. Biochem. Biophys.

Res. Commun. 146: 61-66.

Singh, M., Bandana, and P.S. Ahuja. 1999. Isolation and PCR

amplification of Genomic DNA from market samples of dry

tea. Plant. Mol. Biol. Reptr. 17: 171-178.

Slater, M., S . S elman, B.L. Mogilevsky, H. Ammons, and J.

Hartnett. 1998. P fu DNA polymera se: A high fideli ty

enzyme for nucleic acid amplification. Promega Notes 68:

7-12.

Struwe, L., M. Thiv, J.Kadereit, T. Motley, A.S .R. Pe pper,

J. Rova, K. Potgieter, P.W hi te, a nd V.A. Albert. 1998.

Saccifolium, an endemic of S ierra de La Neblina on the

Brazillian-Venezuelan frontier is related to a temperate

alpine lineage of Gentianaceae. Harvard Papers in Bot. 3:

199-214.

Su, M.H. 2007. Taxonom ic study of Camellia formosens is

(Mas amune et S uzuki). In M. H. Su, C. F. Hsieh et C.

H. Ts ou (The aceae) (eds .), P hD dis serta tion, Ins titute

of Ecology and Evolutionary Biology Colleage of L ife

Science, National Taiwan University. Taiwan.

Swofford, D.L. 2001. PAUP*: Phylogenetic Analys is Us ing

Parsimony (*and Other Methods). Version 4.0b10. Sinauer

Associates, Sunderland, Massachusetts.

Tai, T.H. and S.D. Tanks ley. 1990. A rapid and inexpensive

method for isolation of total DNA from dehydrated plant

tissue. Plant. Mol. Biol. Reptr. 8: 297-303.

Takeuchi, A., S. Matsumoto, and M. Hayatsu. 1994. Chalcone

synthase from C. sinensis: solation of cDNA and the organ

specific and sugar responsive expression of genes. Plant

Cell Physiol. 35: 1011-1018.

Tam ura, K. and M. Ne i. 1993. E st im ati on of the numbe r

o f nuc le ot ide s ubs t it uti ons i n t he con trol re gio n of

mitochondrial DNA in humans and chimpanzees. Mol. Biol.

Evol. 10: 512-526.

Tang, S. and Y. Zhong. 2002. A phylogenetic analysis of nrDNA

ITS sequences from S er. Chrysantha (Sect. Chrysantha,

Camellia, Theaceae). J. Genet. Mol. Biol. 103: 105-107.

Thakor, B.H. and C.R. Parks. 1997. Phylogenetic relationships

within the genus Camellia (Theace ae). S upplement to

Amer. J. Bot. 84: 237 (Abstract only).

Tindall, K.R. and T.A. Kunkel. 1988. Fidelity of DNA synthesis

188

Botanical Studies, Vol. 49, 2008

by the Thermus aquaticuc DNA polymerase. Biochemistry

27: 6008-6013.

Triglia, T., M.G. Peterson, and D.J. Kemp. 1988. A procedure

for in vitro amplification of DNA segments that lay out side

the boundaries of known sequences. Nucleic Acids Res. 16:

8186.

Urbat sch, L.E ., B.G. B aldwin, a nd M.J . Donoghue. 2000.

P hylogeny of cone flowers and relative s (Helianthease:

Ateraceae) based on nuclear DNA internal trans cribed

spacer (ITS) sequences and chloroplast DNA restriction site

data. Syst. Bot. 25: 1153-565.

Wachira, F.N., R. Waugh, C.A.Hackett, and W. Powell. 1995.

Detection of genetic diversity in tea using RAPD markers.

Genome 38: 201-210.

White, T.J., T. Bruns, S. Lee, and J. Taylor. 1990. Amplification

and direct sequencing of fungal ribosomal RNA genes for

phylogenetics. In M.A. Innis, D.H. Gelfand, J.J. Sninsky,

and T.J. White (eds.), PCR Protocols: a Guide to Methods

and Applications. Academic Press, London, pp. 315-322.

Xiao, T.J. 2001. Phylogenetic relationships of the genus

Camellia (Theaceae) bas ed on the RNA polyme ras e II

(RPB2) sequences. PhD dissertation, Univ. North Carolina,

Chapel Hill.

Xiao, T.J. and C.R. Parks. 2003. Molecular analysis of the genus

Camellia. Inter. Camellia J. 35: 57-65.

Yang, J.B., H.T. Li, S.X. Yang, D.Z. Li, and Y.Y. Yang. 2006.

T he a ppl ica ti on of fou r DNA se que nce s to s tudy ing

molecular phylogeny of Camellia (Theaceae). Acta Bot.

Yunn. 28: 108-114.