全 文 ：作物学报 ACTA AGRONOMICA SINICA 2010, 36(1): 85−91 http://www.chinacrops.org/zwxb/
ISSN 0496-3490; CODEN TSHPA9 E-mail: xbzw@chinajournal.net.cn

本研究由国家自然科学基金重点项目(30871558)，国家高技术研究发展计划(863计划)项目(2006AA10Z111)，江苏省自然科学基金项目(BK2008036)
和教育部高等学校学科创新引智计划项目(B08025)资助。
*
通讯作者(Corresponding author): 郭旺珍, E-mail: moelab@njau.edu.cn; Tel: 025-84395311
第一作者联系方式: E-mail: shanshi02@163.com; 2005201008@njau.edu.cn ** 共同第一作者
Received(收稿日期): 2009-04-15; Accepted(接受日期): 2009-06-25.
DOI: 10.3724/SP.J.1006.2010.00085
Molecular Cloning and Characterization of Two Fiber Elongation Genes Using
a Cotton Fiber Developmental Mutant (Gossypium hirsutum L.)
WANG Lei, ZHU Yi-Chao**, CAI Cai-Ping, ZHANG Tian-Zhen, and GUO Wang-Zhen*
National Key Laboratory of Crop Genetics and Germplasm Enhancement, Cotton Research Institute, Nanjing Agricultural University, Nanjing 210095,
China
Abstract: Cotton fibers are single-celled seed trichomes of major economic importance. Many important genes are expressed
during cotton fiber development and fiber developmental mutants can be used to preferentially detect the genes controlling fiber
development. The Ligon lintless mutant (Li1li1) is a fiber elongation developmental mutant with a dominant monogenetic mutation
characterized by short fibers and distorted leaf, stem and flower growth, and the recessive pure line (li1li1) exhibits normal fiber
developmental characteristics. The objectives of this study were to isolate genes preferentially or specifically expressed in fiber
elongation stage by comparing gene expression differences between Li1li1 and li1li1. RNAs isolated from 10 days post anthesis
(DPA) fibers and mixtures in Li1li1 and li1li1 were used to screen differential gene expression in fiber development using differen-
tial display reverse transcriptase polymerase chain reaction (DDRT-PCR). Two differential expression cDNA segments were iso-
lated, the corresponding full-length cDNAs were cloned and their primary function was analyzed. The two genes encoded 542 and
667 amino acid residues and functioned as glutamate decarboxylase (GhGAD) and vacuole-pyrophosphatase (GhVP1), respec-
tively. Transcriptional level assays showed the two genes were constitutively expressed in tested tissues with higher expression
levels during the fiber elongation stage. Furthermore, a BC1 mapping population derived from hybridization between G. hirsutum
acc. TM-1 and G. barbadense cv. Hai 7124, and TM-1 as the recurrent parent, was used for the location of GhGAD and GhVP1 on
chromosomes 12 and 8, respectively, using cleaved amplified polymorphic sequences (CAPs).
Keywords: Cotton; Fiber developmental mutant; DDRT-PCR; Glutamate decarboxylase (GhGAD); Vacuole-pyrophosphatase
(GhVP1)
两个棉纤维发育相关基因的克隆与特征分析
王 磊 朱一超** 蔡彩萍 张天真 郭旺珍*
南京农业大学作物遗传与种质创新国家重点实验室, 江苏南京 210095
摘 要: 棉纤维发育突变体是克隆棉纤维发育关键基因和阐明其发育分子机理的优异资源。陆地棉李氏超短纤维突
变体(Li1li1)是显性单基因突变体, 表现为显性纯合体(Li1Li1)致死, 显性杂合时(Li1li1)表型为茎秆扭曲、叶片卷曲和纤
维短至 6 mm, 而隐性纯合体(li1li1)则表现为株型和纤维发育都正常。本文对开花后 10 d 的李氏纤维发育正常材料
(li1li1)和超短纤维突变体(Li1li1)胚珠纤维混合体进行 mRNA差异显示反转录 PCR (DDRT-PCR)分析, 获得 2条在李氏
纤维发育正常材料中上调表达的差异片段。测序及 DNA 序列的生物信息学分析表明该差异片段分别与编码谷氨酸
脱羧酶和质子焦磷酸酶的基因有较高同源性。通过电子拼接, 5′RACE和全长 cDNA序列验证, 克隆了棉花的谷氨酸
脱羧酶(GhGAD)和质子焦磷酸酶(GhVP1)基因全长 cDNA, 进一步对其功能和染色体定位进行了初步分析。转录水平
分析表明, 这两个基因在棉花根、茎、叶和纤维中组成性表达, 在棉纤维中优势表达。利用本实验室陆地棉遗传标准
系 TM-1和海岛棉海 7124培育的含 140个单株的 BC1作图群体, 将 GhGAD 和 GhVP1分别定位在第 12条染色体和
第 8条染色体。
关键词: 棉花; 纤维突变体; DDRT-PCR; 谷氨酸脱羧酶; 质子焦磷酸酶
86 作 物 学 报 第 36卷


Cotton (Gossypium spp.) is the primary source of tex-
tile fiber throughout the world. The cotton fibers used in
textiles are derived from epidermal cells of maturing
seeds. Seed trichome features vary widely in the genus,
ranging from poor quantity and quality in most wild spe-
cies to excellent in cultivated species [1]. Four discrete yet
overlapping stages characterize cotton fiber development,
including initiation (–3 to 0 days post-anthesis, DPA),
cell elongation/primary wall synthesis (1–25 DPA), cell
wall deposition (16–40 DPA) and maturation (40–50
DPA) [2-4]. Fiber cells begin to initiate during initiation
and undergo elongation during development, increasing
in length by 1 000–3 000 times of the trichome cell dia-
meter.
Primary wall synthesis is complex and commences at
the first two stages of fiber development continuing
through 25 DPA. Concurrent with increased water uptake
through aquaporin activity, the sequestration of osmoti-
cally active solutes by the expanding vacuole is an im-
portant aspect of turgor-driven wall extension [5-6]. The
distinct phenomenon of fiber cell expansion is essentially
due to an elusive process of functional gene expression
and translation. Following the cloning of E6 from G. hir-
sutum [7], a number of structural genes that appear to play
a role in fiber elongation have been isolated. Therefore,
single-celled cotton fibers and their synchronous growth
have made them an attractive system for studying
trichome elongation.
Cotton fiber mutants are valuable materials to explore
genes controlling fiber development at the molecular
level. Griffee and Ligon were the first to identify the Li-
gon lintless mutant (Li1li1) and Kohel later determined its
genetic characteristics [8]. Li1li1 is a monogenic, dominant
mutant characterized by short fibers (no more than 6 mm
long; Fig. 1) and distorted leaf, stem and flower growth;
however, the dominant homozygote (Li1Li1) genotype is
lethal [8]. The pleiotropic effects of the Li1 gene suggest it
serves as a regulatory gene [9].
The objectives of the present research were to isolate
key genes in fiber development by comparing differential
gene expression between the Ligon lintless developmen-
tal mutant (Li1li1) and the wild-type line (li1li1); and to
further elucidate their roles in fiber developmental pro-
cesses.



Fig. 1 Fiber morphology of the wild-type line (li1li1, left) and the
mutant line (Li1li1, right)
1 Materials and methods
1.1 Materials
The following plant materials were chosen for study:
the Ligon lintless mutant (Li1li1) and its wild-type line
(li1li1); the genetic standard line Texas Marker-1 (TM-1);
sea island cotton variety Hai 7124; and 140 individuals
of the BC1 population derived from a cross between
TM-1 (recurrent parent) and Hai 7124. All varieties and
lines were planted under standard field conditions at the
Nanjing Agricultural University, Jiangsu Province,
China.
At 10 DPA, ovule and fiber mixtures were sampled
from the Li1li1 mutant and li1li1 wild type material for
differential gene expression analysis. Furthermore, ovule
and fiber mixtures at 0 and 6 DPA; fibers at 12, 18, and
24 DPA; and root, stem, and leaf tissues from li1li1 wild
type material were sampled for gene expression analysis.
All samples were isolated and immediately frozen in li-
quid nitrogen.
1.2 Total RNA isolation
Total RNAs from all samples were isolated following
the hot boric acid method [10].
1.3 Differential display reverse transcriptase- po-
lymerase chain reaction (DDRT-PCR) analysis
DDRT-PCR was performed to screen for genes in-
volved in fiber elongation following the method de-
scribed by Liang & Pardee [11]. The combinations of ran-
dom 10-base primers and an anchor primer (5′-AAGCT
TTTTTTTTTTG-3′) were used for DDRT- PCR analysis.
1.4 Gene cloning and sequence analysis
Based on the results of DDRT-PCR, two differential
display segments were selected for further study. Full
length cDNA sequences were obtained via 5′ RACE,
following the protocol of the 5′ RACE System for Rapid
Amplification of cDNA Ends (Invitrogen, Carlsbad,
Germany). The other reagents were purchased from Ta-
KaRa (Dalian, China), including Reverse Transcriptase
XL (AMV), Ribonuclease H, Ribonuclease Inhibitor,
Terminal Deoxynucleotidyl Transferase, dNTP Mixture,
and dCTP.
DNA sequence analysis and open reading frame (ORF)
searches were carried out using DNAMAN software
(Lynnon Biosoft, Quebec, Canada). Nucleic acid and
protein sequence BLAST searches were performed
against the GenBank database. Molecular weight and
isoelectric point, functional domain, and amino acid sig-
nal peptides were predicted online, using ExPASy
(http://cn.expasy.org/tools), NCBI CDS (http://www.ncbi.
nlm.nih.gov/Structure/cdd/wrpsb.cgi), SignalP 3.0 (http://
www.cbs.dtu.dk/services/SignalP/), respectively.
1.5 RT-PCR and quantitative real time RT-PCR
(qPCR)
Expression analysis was performed to confirm gene
transcriptional patterns using RT-PCR and three-step
第 1期 王 磊等: 两个棉纤维发育相关基因的克隆与特征分析 87


quantitative real time RT-PCR (qPCR). An in vitro tran-
scription procedure followed the M-MLV transcription
kit protocol (Promega, USA). RT-PCR reaction mixture
was prepared using 10×PCR buffer 2.5 μL, 25 mmol L−1
MgCl2 1.5 μL, 10 mmol L−1 dNTP Mix 0.5 μL, 10 μmol
L−1 primers 1 μL, 1 μL cDNA, 5 U μL−1 Taq 0.1 μL, and
ddH2O to a total volume of 25 μL. The PCR parameters
included a pre-denaturation at 95℃ for 2 min; followed
26 cycles of denaturation at 94℃ for 45 s; annealing at
58℃ for 45 s; extension at 72℃ for 60 s; and a final ex-
tension at 72℃ for 10 min. An internal standard (EF1α)
primer was used to adjust the relative PCR template
concentration.
qPCR was tracked on an iQ5 Multicolor Real-Time
PCR Detection System (BIO-RAD) using the iQ SYBR
Green Supermix (BIO-RAD). Each sample was PCR-
amplified using equal amounts of cDNA template in trip-
licate for at least three independent experiments. PCR
amplifications were as follows: an initial step for 10 min
at 95℃; 40 cycles of 10 s at 95℃, 20 s at 58℃ and 30 s
at 72℃; and completed by melting curve analysis to con-
firm PCR product specificity. The baseline and threshold
values were adjusted according to the manufacturer’s
instructions and its parenthesis software (iQ5 Optical
System Software, Version 2.0). Similar results were ob-
tained from relative gene expression data using the ΔCt
(i.e., ΔCt) method described by Winer et al.[12]. Specific
gene expression levels were not available (N/A) if Ct(gene)
> 30.
1.6 Primer design
Primer Premier 5 (Premier, Canada) designed the
specified primers, which were subsequently synthesized
by Genescript Biotech Co. (Nanjing, China). Primer se-
quences are provided in Table 1.
1.7 Chromosome localization
DNA templates were extracted from TM-1 and Hai
7124 (parents of the BC1 mapping population) [13]. PCR
products generated using ORF amplification primers
were digested by restriction enzymes and separated on
polyacrylamide gels. According to parental differences,
the amplified product polymorphisms were further ex-
amined in the BC1 mapping population. The polymorphic
loci were integrated into our backbone map [14] using
Joinmap 3.0 software [15]. The linkage map was drawn by
MapDraw software [16].
2 Results
2.1 Isolation of differential display genes by
DDRT-PCR
Ten DPA fiber and ovule mixtures from the Ligon lint-
less mutant (Li1li1) and the wild-type line (li1li1) were
selected to isolate fiber development differential expres-
sion genes via DDRT-PCR. Initially, five differentially
expressed segments were obtained. Among the five seg-

Table 1 Primers for 5′RACE and RT-PCR
Gene No. Primer (5′→3′) Purpose
1-GSP1 TTCCAGCATAAACAAGA
1-GSP2 AGCTGCATCAACATGAATTGGGGTAT
1-GSP3 GGGTATCCCATCCAGTTTCTTTGTTCT
5′ RACE
1-F1 TTAGCCTTCCACTAATATCTTCCATCTCT
1-R1 TGCGAAAGGTTCGATTCAAATG
ORF amplification
1-F2 GGCTCCCCCTTGTGAAGAGTATTA
1-R2 CCCTTTGGGTCTCAATAGCAGTCT
RT-PCR
1-F3 GAGATGTTGCGTCGCTTCGGTT
GhGAD
1-R3 CCCTTTGGGTCTCAATAGCAGTCT
qPCR

2-GSP1 AGCCAACCCAAAAATC
2-GSP2 AGCCCACAATCCGATAGGAACAC
2-GSP3 GGTGGAGATAAGCAGTTGTCGTTTT
5′RACE
2-F1 TCACAGGATTCAAGCAATGGATA
2-R1 AATACAAAAGTTTCCCATATCCT
ORF amplification
2-F2 AACTGCTTATCTCCACCGTCTC
2-R2 TCTCCTATCACTGCTGCCTTGT
RT-PCR
2-F3 GAAGGCAGGGTAAAGCCAGACT
GhVP1
2-R3 GGATGTTTAGTGATGGACCTGAA
qPCR

3-F AGACCACCAAGTACTACTGCAC EF1α
3-R CCACCAATCTTGTACACATCC
Internal control
88 作 物 学 报 第 36卷


ments, two were expressed differentially and stably in
three independent differential display experiments, with
the expression level higher in 10 DPA wild fibers, com-
pared to mutant fibers (Fig. 2). Therefore, the two seg-
ments were excised and sequenced. BLAST analysis of
the two fragments showed high similarity to glutamate
decarboxylase and pyrophosphatase, respectively.



Fig. 2 Analysis of DDRT-PCR
Lane 1 and 3 show the ovule and fiber mixture of wild-type line (li1li1)
at 10 DPA; lane 2 and 4 show the ovule and fiber mixture of the mutant
line (Li1li1) at 10 DPA; the arrows show the segments of differential
display genes.

2.2 Molecular cloning of two new cotton fiber
elongation genes
The DNA sequences from differential display seg-
ments were used as a query to screen the G. hirsutum
EST database publicly available from NCBI (http://
www.ncbi.nlm.nih.gov/). The homolog ESTs were fur-
ther assembled into contigs. Based on the contigs se-
quence information, two full-length cDNAs were further
cloned.
2.2.1 Identification of a new cotton gene GhGAD
Sequence assembly analysis determined the longest
cDNA contig sequence at 1 131 bp in length. The 3′ ter-
minal was speculated to be complete due to the presence
of a poly-A stretch; however, the 5 terminal was incom-
plete. The differential expression level of this candidate
gene was investigated using RT-PCR primers designed
from the known assembled sequence information.
RT-PCR confirmed DDRT-PCR results (Fig. 3). Based on
the assembled sequence, 5 RACR primers were designed
to obtain the full-length glutamate decarboxylase gene. 5
RACE extended the gene to 1 770 bp, and identified a
1 626 bp-long ORF encoding 541 amino acids. BLASTx
showed the protein shared 92% identify with the gluta-
mate decarboxylase gene from Nicotiana tabacum.
Therefore, this gene was temporarily designated as
GhGAD.
The predicted correctness of the GhGAD cDNA se-
quence was validated using gene-specific primers de-
signed for full length ORF amplification. The sequence
analysis exhibited complete congruence with the pre-
dicted contig information. In addition, the putative pro-


Fig. 3 RT-PCR of two differential expression genes
1: the ovule and fiber mixture of wild-type line at 10 DPA; 2: the ovule
and fiber mixture of mutant line at 10 DPA.

tein was predicted using Expasy pI/MW software. The
theoretical iso-electric point was 8.78 with a calculated
molecular weight of 61.9 kD. The putative protein was
predicted with 2 trans-member domains (117–137 and
508–528 amino acid residues) using TMHMM (http://
www.cbs.dtu.dk/cgi-bin). The NCBI CDD prediction
detected a phosphopyridoxal binding domain (31–381
amino acid residues), glutamate decarboxylase b domain
(1–444 amino acid residues), and calmodulin binding site
at the C terminate. The full length GhGAD cDNA se-
quence and its deduced amino acid sequence are shown
in Fig. 4.
2.2.2 Identification of a new cotton gene GhVP1
Assembly analysis determined the longest cDNA con-
tig sequence for the second differential expression seg-
ment was 1 607 bp in length. The contig exhibited a
complete 3′ terminal due to the presence of a poly-A
stretch; however, the 5′ terminal was incomplete. The
differential expression level of the putative pyrophos-
phatase gene was verified using RT-PCR primers de-
signed from the known assembled sequence information.
RT-PCR confirmed DDRT-PCR results (Fig. 3). Subse-
quently, 5′ RACE primers were designed to obtain the
full-length pyrophosphatase gene. 5 RACE extended the
gene length to 2 510 bp, which contained a 2 001 bp-long
ORF encoding 666 amino acids. BLASTx showed that
the protein shared 87% identify with the vacuole- pyro-
phosphatase gene from Arabidopsis thaliana. This gene
was temporarily designated as GhVP1.
The predicted correctness of the GhVP1 cDNA se-
quence was validated using gene-specific primers de-
signed for full length ORF amplification. The sequence
analysis exhibited complete congruence with the pre-
dicted contig information. The theoretical iso-electric
point was 6.28 with a calculated molecular weight of
69.7 kD. The putative protein was predicted with 14
trans-member domains and a signal peptide with 18
amino acid reidues at the N terminate using TMHMM
(http://www.cbs.dtu.dk/cgi-bin) and Signal-NN (Neutral
Networks) method. The NCBI CDD prediction detected a
pyrophosphatase domain (42–653 amino acid residues).
The full length GhVP1 cDNA sequence and its deduced
amino acid sequence are shown in Fig. 5.
第 1期 王 磊等: 两个棉纤维发育相关基因的克隆与特征分析 89




Fig. 4 cDNA sequence and deduced amino acid sequence of the
full-length GhGAD cDNA
An upstream in-frame initiation codon is boxed.

2.3 Expression of GhGAD and GhVP1
To investigate GhGAD and GhVP1 expression patterns,
we processed qPCR in different tissues and organs from
the normal fiber developmental material (li1li1), including
root, stem, leaf, and several stages of fiber development.
The results suggested that GhGAD and GhVP1 were
constitutively expressed in all organs and tissues but dif-
ferent expression levels were observed. GhGAD demon-
strated the highest expression level in 12 DPA, and
GhVP1 in 18 DPA fibers (Fig. 6).
2.4 Chromosome localization of GhGAD and
GhVP1
CAPs markers from GhGAD and GhVP1 were deve-
loped and polymorphisms between the two mapping
parents, TM-1 and Hai 7124, were obtained. Firstly,
full-length ORF PCR amplification products of the two
genes were generated from mapping parents; Secondly,
BamH I and Hind III were used to digest respectively the
amplification products and polymorphic loci in GhGAD
and GhVP1 from TM-1 and Hai 7124 were detected. By


Fig. 5 cDNA sequence and deduced amino acid sequence of the
full-length GhVP1 cDNA
An upstream in-frame initiation codon is boxed.

integrating the polymorphic loci with the framework
makers established in our laboratory [14], GhGAD was
located between makers NAU943 and NAU5419 on
chromosome 12 (Fig. 7), and GhVP1 was located be-
tween makers NAU3558 and NAU1369 on chromosome
8 (Fig. 7).
3 Discussion
Cotton fiber mutant plants are invaluable materials to
investigate fiber-development and improve fiber quality.
of natural fiber mutants, the Ligon lintless mutant (Li1li1)
90 作 物 学 报 第 36卷




Fig. 6 Expression patterns of two differential expression genes in li1li1 wild type cotton
R: root; S: stem; L: leaf. 0 and 6 d show the ovule of li1li1 wild type at 0 and 6 DPA; 6 d, 12 d, 18 d, and 24 d show the fiber of li1li1 wild type at 12, 18,
and 24 DPA, respectively. One-way analyses of variance were performed with least significant difference (LSD) multiple comparison tests at 0.01
level. When two samples show different letters (A to D) above the bars, the difference between them is significant (P < 0.01).



Fig. 7 Chromosome mapping of two differential expression genes
in allotetraploid cotton

is a monogenic, dominant mutant characterized by short
fibers and distorted leaf, stem and flower growth [8].
Scanning electron microscopy studies have shown that
the genetic control of fiber initiation is very similar be-
tween TM-1 and Li1li1 mutant [9]. Due to premature ter-
mination of fiber elongation in the Li1li1 mutant, key fi-
ber elongation genes are readily screened by differential
expression analysis between mutant and wild type plants
with highly similar genetic backgrounds. In this study,
we characterized two differential expression genes highly
expressed in wild type fibers and ovules compared with
the mutant line at 10 DPA, a rapid fiber elongation period.
Furthermore, chromosome location for both genes was
confirmed, which provides a foundation for future re-
search in elucidating gene structure, function and evolu-
tionary relationships.
Glutamate decarboxylase (GAD) is a widespread en-
zyme operating in most organisms; and has been tho-
roughly studied in the mammalian neural system. GAD
catalyzes glutamic acid to γ-amidogen butyric acid
(GABA), which is a commonly occurring natural amino
acid [17]. GABA is a metabolism bypass product, which
stores nitrogen, adjusts pH, and promotes growth and
resistance in plant cells. GAD and GABA also participate
in plant resistance to environmental stress, including
mechanical damage, heat and cold shock, and water log-
ging [18]. Furthermore, GAD is known to play a role in
the entire life history of plants, where it exhibits a
calmodulin binding site at the C terminate, which differs
from GADs in animal and microbe species [19]. The pre-
sent study showed that GhGAD is highly expressed in
fiber cells at 12 DPA, as fibers are rapidly elongating,
suggesting GhGAD might also be primarily responsible
for fiber elongation.
In plants, vacuole-pyrophosphatase (V-PPase) is lo-
cated in the vacuole, and plays a vital role in saccharide
synthesis, and nucleotide and amino acid metabolism.
V-PPase can also serve as a substitute for ATP [20-22].
Furthermore, V-PPase can adjust the material concentra-
tion and bulk of the vacuole, which influences cell elon-
gation. Maeshima [23] hypothesized that V-PPase, which
includes 14 trans-member domains; including three con-
served regions (CS1, CS2, CS3), e-loop for substance
binding and CS1 with a region (KAADVGADLVGKVE)
for catalyzation; may participate in the hydrolyzation of
PPi and transport of H+. In general, V-PPase and
V-ATPase on cell members exert a co-effect, affecting the
whole proton member ΔμH+ system. Smart et al. found
that V-PPase and V-ATPase reached peak activity at 15
and 20 DPA in fiber development, respectively [24]. In
Arabidopsis, excessive V-PPase expression increased
plant resistance to drought and salt stress [25]. Based on
the above evidence and increased GhVP1 expression in
fiber cells at 18 DPA, GhVP1 may promote fiber cell
elongation through increased dissolved matter transport
and turgor pressure. The identification and characteriza-
tion of GhGAD and GhVP1 presents a unique opportu-
nity to investigate the functional mechanisms inherent in
fiber developmental processes. We have recently con-
structed plant expression and RNAi vectors for GhGAD
and GhVP1 using fiber specific expression promoters to
confirm GhGAD and GhVP1 function in cotton fiber de-
velopmental processes.
4 Conclusions
Two key genes of GhGAD and GhVP1 in fiber deve-
lopment isolated by comparing differential gene expres-
sion between the Ligon lintless developmental mutant
(Li1li1) and the wild line (li1li1) were constitutively ex-
pressed in all tissues with higher expression levels during
第 1期 王 磊等: 两个棉纤维发育相关基因的克隆与特征分析 91


the fiber elongation stage. GhGAD and GhVP1 were lo-
cated on chromosomes 12 and 8, respectively.
References
[1] Fryxell P A. The Natural History of the Cotton Tribe. College
Station, TX: Texas A&M University Press, 1979
[2] Basra A S, Malik C P. Development of cotton fiber. Inter Rev
Cytol, 1984, 89: 65–113
[3] Orford S J, Timmis J N. Abundant mRNAs specific to the devel-
oping cotton fiber. Theor Appl Genet, 1997, 94: 909–918
[4] Turley R B, Ferguson D L. Changes of ovule proteins during
early fiber developing in a normal and a fiberless line of cotton
(Gossypium hirsutum L.). J Plant Physiol, 1996, 149: 695–702
[5] Dhindsa R S, Beasley R B, Ting I P. Osmoregulation in cotton
fiber: Accumulation of potassium and malate during growth.
Plant Physiol, 1975, 56: 394–398
[6] Basra A S, Malik C P. Dark metabolism of CO2 during elongation
of two cottons differing in fiber lengths. J Exp Bot, 1983, 34:
1–9
[7] John M E, Crow L J. Gene expression in cotton (Gossypium hir-
sutum L.) fiber: Cloning of the mRNAs. Proc Natl Acad Sci USA,
1992, 89: 5769–5773
[8] Kohel R J. Linkage tests in upland cotton, Gossypium hirsutum L.
Crop Sci, 1972, 12: 66–69
[9] Karaca M, Saha S, Jenkins J N, Zipf A, Kohel R, Stelly D M.
Simple sequence repeat (SSR) markers linked to the Ligon lint-
less (Li1) mutant in cotton. J Hered, 2002, 93: 221–224
[10] Wan C Y, Wilkins T A. A modified hot borate method signifi-
cantly enhances the yield of high quality RNA from cotton (Gos-
sypium hirsutum L.). Anal Biochem, 1994, 223: 7−12
[11] Liang P, Pardee A B. Differential display of eukaryotic messenger
RNA by means of the polymerase chain reaction. Science, 1992,
257: 967–971
[12] Winer J, Jung C K, Shackel I, Williams P M. Development and
validation of real-time quantitative reverse transcriptase poly-
merase chain reaction for monitoring gene expression in cardiac
myocytes in vitro. Anal Biochem, 1999, 270: 41–49
[13] Han Z G, Guo W Z, Song X L. Genetic mapping of EST-derived
microsatellites from the diploid Gossypium arboretum in al-
lotetraploid cotton. Mol Genet Genomics, 2004, 272: 308–327
[14] Guo W Z, Cai C P, Wang C B, Zhao L, Wang L, Zhang T Z. A
preliminary analysis of genome structure and composition in
Gossypium hirsutum. BMC Genomics, 2008, 9: 314
[15] Van-Ooijen J W, Voorrips R E. JoinMapR Version 3.0: Software
for the Calculation of Genetic Linkage Maps, CPRO-DLO,
Wageningen, 2001
[16] Liu R-H(刘仁虎), Meng J-L(孟金陵). MapDraw: A Microsoft
Excel macro for drawing genetic linkage maps based on given
genetic linkage data. Hereditas (遗传), 2003, 25(3): 317–321 (in
Chinese with English abstract)
[17] Roberts E, Frankel S. Gamma-aminobutyric acid in brain: Its
formation from glutamic acid. J Biol Chem, 1950, 187: 55
[18] Satyanarayan V, Nair P M. Enzymolgy and possible roles of
4-aminobutyrate in higher plants. Phytochemistry, 1990, 29: 367–
375
[19] Baum G, Chen Y, Arazi T, Takatsuji H, Fromm H. A plant gluta-
mate decarboxylase containing a calmodulin binding domain. J
Biol Chem, 1993, 268: 19610–19617
[20] Rea P A. Vacuolar H+-translocating pyrophosphatases: A new
category of ion translocase. Trends Biochem Sci, 1992, 17:
348–353
[21] Rea P A, Poole R J. Vacuolar H+-translocating pyrophosphatase.
Plant Mol Biol, 1993, 44: 157–180
[22] Zhen R G, Kim E J, Rea P A. The molecular and biochemical
basis of pyrophosphate-energized proton translocation at the
vacuolar membrane. Adv Bot Res, 1997, 25: 297–337
[23] Maeshima M. Vacuolar H+-pyrophosphatase. Biochim Biophys
Acta, 2000, 1465: 37–51
[24] Smart L B, Vojdani F, Maeshima M, Wilkins T A. Genes involved
in osmoregulation during turgor-driven cell expansion of deve-
loping cotton fibers are differentially regulated. Plant Physiol,
1998, 116: 1539–1549
[25] Roberto A G, Jisheng L, Soledad U, Lien M D, Gethyn J A, Seth
L A, Gerald R F. Drought- and salt-tolerant plants result from
overexpression of the AVP1 H+-pump. Proc Natl Acad Sci USA,
2001, 98: 11444–11449