Molecular Studies on Sweet Protein Mabinlin;
Thermal Stability
By
LEUNG Chun-wah
A Thesis Submitted in Partial fulfillment of the Requirement for the degree of Master ofPhilosophy in
Biology
©The Chinese University ofHong Kong July 2000
The Chinese University ofHong Kong holds the copyright ofthis thesis. Any person(s) intending to use a part or whole of the materials in the thesis in a proposed publication must seek copyright release from the Dean of the Graduate School.
Thesis Committee
Prof. Sun S. S. M. (Supervisor)
Prof. Lam H. M. (hitemal examiner)
Prof. Wong Y. S. (Internal examiner)
Prof. Wayne M. Becker (External examiner) Statement
All the experimental work reported in this thesis was performed by the author, unless specially stated otherwise in the text.
LEUNG Chun-wah Acknowledgment
First ofall, I would like to express my gratitude to my supervisor, Prof.
Sun S.S.M. for getting me into the field of Biotechnology. I am indebted to his invaluable experience, guidance, understanding and tolerance all the way through. I am also grateful to Prof. Sun S.S.M. for his insightful comments and patient reviews (even during his vacation) to the drafts ofthis thesis.
I would like to express my heartfelt appreciation for the valuable advice and encouragement to the members of my thesis committee, Prof.
Wong Y.S. and Prof. Lam H.M. My gratitude should also go to my extemal examiner, Prof. Wayne M. Becker for his patience ofreviewing my thesis.
I am also thankful for the generosity ofProf. Lam H.M. for translating the Chinese version of my abstract, providing the Agrobacterial strain and bringing about the vaccum infiltration technique in my research work.
Particularly, I would like to acknowledge Dr. Linda Ooi for her teachings and technical suggestions and Prof. Ooi V.E.C. for his caring. Their research experience and attitude is valuable to me in the course of my research study.
In addition, I am indebted to the continuous encouragement and technical assistance of my labmates (Marilyn Yu M.L., Simon Cheng M.K.，
Donna Lee C.，Danny Ng W.K., Angela Yu W.S., Rosalyn Chen L., Paulus
Fong M.K., Zhang Xiaodong, Stanley Cheung C.K., Jones Zhang S.Y.,
Gundam Wong H.W., Juon Juon Lee J.K., Yuan Dingyang, Smile Duan M.J.,
111 Xiao Guoying). The quality of my thesis was greatly enhanced by the effort
and cooperation from them.
I would also like to express special thanks to Dr. Lucia Liu P.S. for
her sharing of computer knowledge and constructive suggestions. My
appreciation should also go to many parties in the department (Mr. Jack Lo
C.S.，Miss May Yam K.M., Mr. Lung Chan Wl.，Mr. Chan Y.B.，Miss Grace
Leung S.W.，Mr. Eric Suen P.K., Mr. Arthur Lee CL.，Miss Maggie Kwan
P.C.，Miss Elizabeth Wilhnott, all technicians from team C and team A，
members from Prof. Lam H.M.'s lab, Prof. Fung M.C.'s lab, Prof. Ge Wei's
lab, Prof. Wong Y.S.'s lab, Prof. Kwan H.S.'s lab) for their lab-material
support, technical and emotional sharing.
Besides, the spiritual support and technical advice from my friendsan d
previous colleagues (flatmates from HKU, a-dozen-people from HKU A&P
Biotechnology, labmates from HKU Leung P.C.，s lab and HKIB, old friends
from CACWGC, impossible to name all) is indispensible for the completion of my thesis.
Last but not least, I would like to express my deepest gratitude to my dearest parents, sister (Jenny) and brother (Jimmy). Their unconditional love, infinite understanding, emotional support and caring have been the greatest help to finish my postgraduate study.
IV Abstract
Sweet plant proteins have the potential to substitute for sucrose as an
alternative natural sweetener in low-calorie and diabetic products. However,
most sweet plant proteins lose their sweetness upon high temperature
treatment and cleavage of disulfide bridges is a main factor of their heat-
instability. Mabinlin (MBL), a plant protein with four isoforms (MBL I，II，III
& IV) and 400 times sweeter than sucrose, shows high heat-stability in three
(MBL II，III & IV) of its four isoforms. Previous biochemical and molecular
studies also revealed that each MBL isoform consists of two subunit
polypeptides; the four MBLs possess highly similar amino acid sequences; and
all four disulfide bridges are at the same positions in both the heat-stable
(MBL II，III & IV) and unstable (MBL I) isoforms. The later structural
feature suggests that the disulfide bridges may not be the sole determinant of
heat stability in MBLs. A single amino acid difference, at position #47 in the
large subunit, however, was found between the two groups, i.e. Arg in MBL II,
III & IV while Gki in MBL I. hi view of these findings, we are attempting to
explore the underlying determinant(s) ofheat stability in the MBLs.
Site-directed mutagenesis was performed in the cDNAs encoding the heat-stable MBL III and unstable MBL I in an attempt to generate single amino acid mutations at position #47 of the large subunits, either to reverse
(by Arg to Gln or vice versa) or to maintain (by similar amino acids, Gly &
Asp) their heat-stability properties. Both the wild-type and mutant MBL III and MBL I cDNAs were used to construct chimeric genes under the control of the phaseolin regulatory regions of French bean. The chimeric genes were transferred into the Arabidopsis thaliana genome via Agrobacterium-rnQdmied transformation. Plants containing a single copy of the transgene were selected and propagated to obtain homozygous R3 generation for expression analysis, as a means of setting up a system to test the relationships between heat
stability and MBL primary structure.
For transgenic plants containing the chimeric genes MBL III Q)BI /
phas / MBLIII-wt, -GLN, & -LYS)，the activity of the marker gene was
detected by GUS assay. The integration of the MBL III transgenes was
proved by polymerase chain reaction (PCR), and the MBL III RNA transcripts
were detected by reverse transcritption-PCR (RT-PCR) and DNA sequencing.
For plants transformed with the chimeric genes MBL I (pBl / phas / MBLI-wt,
-ARG, -ASP), Ri seeds had been harvested for future analysis.
To detect the MBL protein in transgenic seeds, antisera with very
high titres against MBL were obtained. Tricine SDS-PAGE, protein
sequencing, isoelectric precipitation, and westem blot and immunodetection
had been attempted to demonstrate the stable accumulation of MBL III protein
in transgenic seeds. However, despite the proven presence of their transcripts,
detectable levels of MBL proteins have thus far not been confirmed in the
transgenic seeds. The similar size and immuno-cross-reactivity of the 2S
protein of the host plant Arabidopsis with the MBL protein compounds were
the difficulties in detecting the MBL. Further experiments that could separate
the two low molecular weight proteins and enhance the protein detection
sensitivity would be desirable, in addition to experiments to determine if the
MBLs are unstable in the transgenic Arabidopsis seeds.
VI 摘要
「甜蛋白」有潛力替代鹿糖成爲一種低卡路里和可用於高血糖病
患者食品的天然甜味劑。然而，絕大多數植物甜蛋白在高溫處理下會喪
失甜味，二硫鍵被打斷是熱不穩定性的一個主要因素°馬檳榔甜蛋白
(MBL)比簾糖甜400倍，它是一群有四種同工型(MBL I，II，III及IV)的
植物甜蛋白，其中n，111及〜三種同工型表現出高熱穩定性。以往的生
化及分子生物硏究發現每種MBL同工型有兩個多駄亞基0所有馬檳榔
甜蛋白都擁有高度類近的氨基酸序列，而且在熱穩定及不穩定型MBL
內四個二硫鍵的位置都是相同的’這結構的特點顯示二硫鍵並不是決定
熱穩定性的唯一因素。然而’大亞基第47位氣基酸在這熱穩定及不穩
定型兩類MBL中卻有分別（Arg在MBL II，III及IV ； Gb在MBL I)。
基於此發現，我們試圖探索MBL熱穩定性背後的關鍵決定因素。
利用特點誘變技術，可以改變熱穩定型MBL III和熱不穩定型
MBL I的cDNAs。目的是通過在大亞基第47位的氨基酸產生突變，逆
轉（Arg改爲Ghi或倒轉）或維持（如Lys和Asp等同類氨基酸）其氣基酸
及特性，以測試其對熱穩定性的影響。MBL III及I原來的及誘變後的 cDNAs都會受菜豆蛋白基因調控區嵌合並受其調控。嵌合的基因以農杆
菌感染轉化至擬南芥菜基因組內。
歸選出被插入單拷貝基因的轉化植物後，經過增殖再獲得純合的
第三代轉基因植物，目標是建立一套測定熱穩定性與MBL —元結構關
係的糸統。
嵌合基因及其表達載體PBI / phas / MBLIII-wt, -GLN及-LYS，pBI
/ phas / MBLI-wt, -ARG及-ASP的構建已完成。這些嵌合基因亦已轉化
Vll 至擬南芥菜基因組。而被轉化MBL III (pBI / phas / MBLIII-wt, -GLN,-
LYS)嵌合基因的轉基因純合植物，其標記基因的活性已用GUS測試來
確定；MBL III轉基因的整合通過聚合酶鏈式反應（PCR)來證實；RT-
PCR及DNA序列測定證明了 MBL III RNA轉錄產物的存在。
此外’被轉化 MBL I Q)BI / phas / MBLI-wt, -ARG, -ASP)的尺！種
子已被收獲及貯藏以待進一步的分析°
高效價的抗血淸被用以檢測轉基因種子內的MBL 0 Tricine-SDS-
PAGE，蛋白測序分析，等定點沉殿以及Westem蛋白印跡法則用來證明
MBL III蛋白的穩定表達。然而，到目前爲止，尙未能在轉基因植物種
子檢測到MBL蛋白。MBL檢測的另一困難是由於擬南芥菜內的2S蛋
白（AT2S3)與MBL的大小接近且能與其抗血淸產生交叉反應。因此’
將來除了檢察MBL在轉基因種子內的穩定性外，還需要進一步實驗來
分離MBL及AT2S3兩種低分子量蛋白及提高檢測的敏感度°
Vlll Table of Contents
Thesis committee
Statement
Acknowledgment
Abstract v
Table of contents ix
List of abbreviations xiv
List offigures xvii
List oftables xix
T,ITF,RATIIRE REVIEW
1.1 INTRODUCTION 1
1.2 ARTIFICML SWEETENERS 3
1.2.1 SACCHARm 3
1.2.2 CYCLAMATE 4
1.2.3 ASPARTAME 4
1.2.4 ACESULFAME-K 5
1.2.5 SUCRALOSE 5
1.3 NATURAL SWEET PLANT PROTEINS 7
1.3.1 THAUMAim 7
IX 1.3.2 MONELLESf 10
1.3.3 CURCULES[ 11
1.3.4 PENTADD^AND BRAZZEDSf 11
1.3.5 MlRACUUN 12
1.3.6 MAmNUN 12
1.4 GENETIC ENGINEERING OF SWEET PLANT PROTEIN 19
1.4.1 BlOTECHNOLOGICAL STUDIES ON lHAUMATE^ 20 1.4.1.1 Protein modification and sweetness 20 1.4.11 Transgenic expression in microbes 21 1.4.1.3 Transgenic expression in higher plants 23
1.4.2 BlOTECHNOLOGICAL STUDIES ON MONEUJN 24 1.4.2.1 Gene modification and transgenic expression in microbes 24 1.4.2.2 Transgenic expression in plants 25
1.4.3 TRANSGENIC EXPRESSION OF MABD^LD^J ^ PLANTS 26
1.5 PHASEOLIN AND ITS REGULATORY SEQUENCES 27
1.6 ARABIDOPSIS 29
1.6.1 ARABIDOPSIS THAUANA AS A MODEL PLANT 29
1.6.2 TRANSFORMATION METHODS 29 1.6.2.1 Direct DNA uptake 30 1.6.2.2 Agrobacterium-mtdA2iiQ^ transformation 31 1.6.2.3 Inplanta transformation 31
2 CFNERAT TNTttOmirT!ON AND HYPOTHESIS 21
2.1 GENERAL INTRODUCTION 33
2.2 HYPOTHESIS 34 2 MOLECULAR STUDIES ON SWKET PROTEIN MAR!NT JN ； THERMAL STABILITY 28
3.1 INTRODUCTION 38
3.2 MATEWALS 40
3.2.1 LABORATORY WARES 40
3.2.2 EQUIPMENTS 40
3.2.3 CHEMICALS 40
3.2.4 COMMERICAL KITS • 41
3.2.5 DNA PRIMERS 42
3.2.6 DNA PLASMIDS 43
3.2.7 BACTERIAL STRAE^S 43
3.2.8 PLANT MATERIALS 44
3.2.9 PROTEESf AND ANTIBODY 44
3.3 METHODS 45
3.3.1 TRANSFORMATION OF ARABlDOVm^WITH MBLIII AND MBLI GENES 45 3.3.1.1 Construction ofmutant MBLIII and MBLI genes containing single codon mutation by megaprimer PCR 45 3.3.1.2 Cloning ofPCR-amplified MBLIII and MBLI cDNAs into vector pD3-8 48 3.3.1.3 In vitro site-directed mutagensis (for the construction ofMBLIII and MBLI cDNAs containing single codon mutation) 49 3.3.1.4 Cloning ofthe wild-type and mutated MBLIII and MBLI cDNA into vector pTZ / phas 53 3.3.1.5 Confirmation of sequence fidelityan d mutated codon in MBLIII and MBLI cDNA by DNA sequencing 53 3.3.1.6 Transfer ofwild-type MBLIII and MBLI cDNA flanked by phaseolin regulatory sequence into Agrobacterium binary vector 55
XI 3.3.1.7 Transformation oiAgrobacterium with pBI / phas / MBLIII and pBI / phas / MBLI chimeric gene constructs 57 3.3.1.8 Vacuum infiltration transformation ofArabidopsis 5 8 3.3.1.9 Screening of homozygous transgenic Arabidopsis 59
3.3.2 EXPRESSION ANALYSIS OF MBLIII TRANSGENE 61 3.3.2.1 GUS assay of transgenic plants 61 3.3.2.2 Genomic DNA isolation from transgenic plants 61 3.3.2.3 PCR amplification oftransgene 62 3.3.2.4 Total RNA isolation from transgenic Arabidopsis 63 3.3.2.5 RT-PCR oftotal RNA from transgenic Arabidopsis 64 3.3.2.6 Verification ofthe presence ofmutagenic site and the fidelity of RNA transcript from transgenic Arabidopsis 65 3.3.2.7 Protein extraction and tricine SDS-PAGE ofputative transgenic protein from A rabidopsis 65 3.3.2.8 N-terminal amino acid sequencing 66 3.3.2.9 Isoelectric precipitation ofMBL 67 3,3.2. lOProduction ofpolyclonal antibody against purified MBL 67 3.3.2.1 lWestem-blotting and immunodectection ofArabidopsis protein by anti-MBL polyclonal antibody 69
3.4 RESULTS & DlSCUSSION 71
3.4.1 SrrE-SPECIFIC MUTATIONS OF ARGESfESfE RESIDUE D^ MBLIII CDNA AND GLUTAMESfE lN MBLI CDNA 71 3.4.1.1 Megaprimer PCR 71 3.4.1.2 Cloning into the seed-specific expression vector pD38 74 3.4.1.3 In vitro site-directed mutagenesis 76
3.4.2 CONSTRUCTION OF PLANT EXPRESSION VECTORS CONTAESfmG CHIMERIC MBLIII AND MBLI 80 3.4.2.1 Cloning ofMBLIII and MBLI cDNAs into the seed-specific expression vector pTZ / phas 80 3.4.2.2 Cloning into the plant expression vector pBI121 83
3.4.3 GENERATION OF HOMOZYGOUS TRANSGENIC ARABIDOPSIS 84
Xll 3.4.3.1 Screening oftransgenic Ri Arabidopsis 84 3.4.3.2 Screening oftransgenic R2 plants 86 3.4.3.3 Screening ofhomozygous R3 transgenic plants 88
3.4.4 DETECTION OF MBLIII TRANSGENE ESI ARABIDOPSIS 89 3.4.4.1 Gus Assay 89 3.4.4.2 Detection oftransgene integration 90
3.4.5 DETECTION OF MBLIII TRANSCRIPT DS[ TRANSGENIC ARABIDOPSIS 92 3.4.5.1 RT-PCR (Reverse-transcription polymerase chain reaction) 92 3.4.5.2 Verification ofthe presence ofthe mutant codon and sequence fidelity of the RT-PCR product 94
3.4.6 DETECTION OF MBL III PROTEO^ D^ TRANSGENIC AMBIDOPSlS 97 3.4.6.1 Expression of MBL protein 97 3.4.6.2 Isoelectric precipitation 101 3.4.6.3 Assay oftiters and quality ofprimary polyclonal antibody against purified MBL protein 103
3.4.6.4 Westem blot / Immunodetection 106
4 GENERAT. msniSSION ^9
Conclusion 112
References 113
Xlll List ofAbbreviations
Abbreviations used in this thesis without definition include:
a.a. Amino acid Ab Antibody ACK Acesulfame-K Amp-R Ampicillin resistant Amp-S Ampicillin sensitive AT2S3 Arabidopsis thaliana 2S albumin isoform 3 BCA Bicinochoninic acid bp Base pairs BSA Bovine serum albumin CaMV Cauliflower mosaic virus cDNA Complementary DNA CTAB Cetyltrimethylammonium bromide Da Dalton dATP 2’ deoxyadenosine 5'-triphosphate DEPC Diethylpyrocaibonate dH2O Distilled water DMF N,N-dimethyl formamide DNA Deoxyribonucleic acid dNTP Deoxyribonucleotide triphosphate DTT Dithiothreitol EDTA Ethylenediaminetetra acetic acid EMS Ethyl methanesulfonate FDA Food and Drug Adminstration
xw GRAS Generally recognised as safe GUS Glucuronidase Kan-R Kanamycin resistant kb Kilobases kDa Kilodaltons LB Left border (ofT-DNA in pBI121) or Luria-Bertani (medium) MBL Mabinlin M-MLV Moloney murine leukemia vims mRNA Messenger ribonucleic acid MS Murashige and Skoog (medium) MW Molecular weight NOS Nopaline synthase NPTII Neomycin phosphotransferase II nt Nucleotide OD Optical density (absorbance) PAGE Polyacrylamide gel electrophoresis PBS Phosphate buffer saline PCR Polymerase chain reaction PEG Polyethylene glycol pI Isoelectric point PVDF Polyvinylidene difluoride Ro The generation to be transformed Ri First generation ofRo R2 Second generation of Ro R3 Third generation of Ro RB Right border (ofT-DNA ofpBI121) RE Restriction enzyme RNA Ribonucleic acid
XV RNase Ribonuclease RT-PCR Reverse transcription-polymerase chain reaction SDS Sodium dodecyl sulfate TAE Tris-acetateyT)DTA TCA Trichloroacetic acid T-DNA Transfer DNA TE Tris-EDTA TEMED N,N,N,N-tetramethylethylenediamine Th Thaumatin Ti plasmid Tumor-inducing plasmid Tm Melting temperature Tricine N-[Tris-(hydroxymethyl)methyl]glycine Tris Tris (hydroxymethyl) amminomethane v/v Volume by volume w/v Weight by volume X-gluc 5-bromo-4-chloro-3-indolyl-P-glucuronic acid -ve Negative +ve Positive
XVI List of Figures
Figure 1-1: Amino acid sequence homology among mabinlin isoforms 14 Figure 1-2: Schematic representation of the disulfide structure in mabinlin 14 Figure 1-3: Amino acid sequence similarity between MBLII and 2S seed storage protein 3 (AT2S3) of Arabidopsis thaliana 15 Figure 1-4: Nucleotide sequence ofthe MBL cDNA and the deduced amino acid sequence 17 Figure 1-5: Comparisons of the precursor sequences between MBL and the 2S seed-storage proteins ofA. thaliana, B. napus and B. excelsa 18 Figure 2-1: The complete amino acid sequences of MBL isoforms I，II & III 35 Figure 2-2: Comparison of the nucleotide sequences ofthe cDNAs encoding the MBL isoforms (MBL I，II and III) 3 7 Figure 3-1: Strategy to construct mutagenic MBLIII and MBLI cDNAs 47 Figure 3-2: Strategy to clone MBLIII and I chimeric cDNAs into pD38 48 Figure 3-3: Strategy of in vito mutagenesis 51 Figure 3-4: Strategy to clone wild type and mutant MBLIII and MBLI cDNAs into seed-specific expression vector pTZ / phas 54 Figure 3-5: Strategy to clone mutant MBLIII and MBLI chimeric genes into plant expression vector pBI121 55 Figure 3-6: First PCR products of megaprimer PCR 73 Figure 3-7: Second PCR products of megaprimer PCR 74 Figure 3-8: Gel purification of phaseolin regulatory sequences 76 Figure 3-9: Gel electrophoresis ofblunt-ended PCR products ofMBLIII and MBLI cDNAs 77 Figure 3-10: Fast-screening of recombinant clones by "cracking" procedure 78 Figure 3-11: Screening for MBL mutants 79 Figure 3-12: Verification of the mutant 82 Figure 3-13: Cloning ofMBLIII and I cDNAs, flankedb y phaseolin regulatory sequences into plant expression vector 84
xvii Figure 3-14: Selection ofRl transgenic Arabidopsis 85 Figure 3-15: GUS staining of native and transgenic Arabidopsis 89 Figure 3-16: Gel electrophoresis ofgenomic DNA from transgenic Arabidopsis 91 Figure 3-17: PCR analysis of genomic DNA from native and transgenic Arabidopsis leaves 91 Figure 3-18: Gel electophoresis of total RNA isolated from maturing native and transgenic Arabidopsis siliques 93 Figure 3-19: RT-PCR of total RNA from maturing native and transgenic Arabidopsis siliques 93 Figure 3-20: Verification of the presence of the mutated codon sequence in transgenic RNA transcripts 95 Figure 3-21: Verification of the mutant codons in RNA transcripts 96 Figure 3-22: Protein profiles of native and transgenic Arabidopsis total seed protein extracts 98 Figure 3-23: Tricine-SDS-PAGE ofnative and transgenic Arabidopsis total seed protein extracts treated with p-mercaptoethanol 99 Figure 3-24: PVDF membrane with transferred native and transgenic seed protein extracts 99 Figure 3-25: Profiles of pI-precipitated proteins of native and transgenic Arabidopsis total seed protein extracts treated with P-mercaptoethanol 102 Figure 3-26: PVDF membrane with transferred pI-precipitated proteins of native and transgenic seeds 103 Figure 3-27: Titering of anti-MBL primary polyclonal rabbit antibodies 104 Figure 3-28: Westem blot of the pI-precipitated proteins from native and transgenic Arabidopsis 105 Figure 3-29: Westem blot of the pI-precipitated proteins from native and transgenic Arabidopsis 107 Figure 3-30: Western blot / immunodection ofnative and transgenic seed proteins 108
XV111 List ofTables
Table 2-1: Hypotheses to test the importance of47th amino acid ofthe large subunit on MBL thermal stability 36 Table 3-1: Name and brand of commercial kits used in different experiments 41 Table 3-2: Primers designed for megaprimer PCR 42 Table 3-3: Primers designed for in vitro mutagenesis 42 Table 3-4: Primers designed for checking DNA sequence 43 Table 3-5: Restriction enzyme site resulting from single codon mutation of MBL III and MBL I 52 Table 3-6: Chi-square analysis 0fR2 transgenic plants 87 Table 3-7: Screening of R3 plants with transgene homozygosity 88 Table 3-8: Restriction maps ofRT-PCR products ofthe three MBLIII transgenes 95 Table 3-9: N-terminal amino acid sequences ofthe 8 kDa protein from native and transgenic A rabidopsis seeds 10( Table 3-10: Biochemical properties of2S albumin in Arabidopsis and Mabinlin (Data from ExPasy search) 10
XIX 1 Literature Review
1.1 Introduction
Sweet tasting foodstuffs have been an integral part in of the diet ofhuman and other animal species for a long history. However, high sugar consumption adds to the problems of obesity, diabetes, coronary artery disease and dental caries throughout the world. To satisfy this desire for sweet taste, food scientists have been searching for sugar substitutes, both artificial and natural, which are soluble, stable, non-carious, low in calories and non-toxic.
The history of artificial sweeteners, in comparison with the natural ones, is just over a hundred years old. Several artificial sweeteners had been developed including the especially successful aspartame. Although evaluation of the physiological and toxicological properties of these sweeteners was carried out in parallel with the commercial interest and use of these artificial sweeteners, consumers are nevertheless concerned about the long-term safety of these artificial sweeteners. A natural and safer sweetener has long been a desirable food item ofthe general public and sweet plant proteins is one potential candidate.
With the recent advances in biotechnology, sweet plant proteins become more attractive for use as sweeteners, not only for mass production, but also for genetically engineering of foods for low caloric sweetness, and for understanding the underlying principle of sweetness.
To achieve these goals, a transgenic expression system is essential. This includes an optimal promoter sequence as well as terminator sequence to enhance the efficiency of transgene expression. Agriculture is the chief example of an inexpensive and successful biotechnology. Once a suitable transgenic plant is produced, no expensive apparatus and no specially trained personnel are needed to maintain the process. Moreover, post-translational modifications of plant proteins occur in planta, including possibly unknown steps, is more likely to take place in plants than in any other system, histead of producing sweet plant proteins in natural stands, plants can be used to produce transgenic proteins in greater quantity or in a more usefiil form. 1.2 Artificial sweeteners
Artificial sweeteners are sensory bioactive compounds with substantially higher sweetness than sugar (sucrose). They are used mainly in diet foods, in pharmaceutical applications to mask the unpleasant taste of drugs, in cosmetic preparations for mouth hygiene, and as a sugar substitute for diabetic patients. Many strict requirements must be satisfied before an artificial sweetener can be accepted, the most important of which is that it must not pose a health hazard. Several commercially well-known sweeteners, including saccharin, cyclamate, aspartame, and acesulfame K, were previously approved by the U.S. Food and Drug Administration (FDA) for use in foods. Other compounds such as alimate and sucralose are awaiting approval by the FDA. HoWever, several sweeteners, such as dulcin, P-4000, and perillaldoxime, used over a period of time, were later found to be toxic and their use prohibited. Therefore, the application of artificial sweeteners in food has always been controversial.
1.2.1 Saccharin
Saccharin was the firstknow n synthetic sweetener. It was discovered accidentally by a German chemist in 1878. The synthesis was started by any of the 4 compounds: toluene, phthalic anhydride, phthalic acid or 2-chlorotoluene, and the synthetic method was immediately patented. Its commercial production started in 1884. Saccharin is approximately 300 to 500 times sweeter than sucrose on a weight basis (Vincent et al., 1955) and is slowly absorbed by the body and excreted in the urine quickly without being metabolized (Renwick, 1985); therefore, it does not contribute calories to the diet. However, at higher concentrations, saccharin gives a bitter-sour off-taste (Larson-Powers and Pangbom, 1978) and thus limits its use in food industry. This off-taste can be masked by mixing with a neutral (tasteless) carrier, such as gelatin, or by combination with other intense sweeteners, such as cyclamate.
1.2.2 Cyclamate
Cyclamate, a derivative of amino-N-sulphanic acid, was discovered in 1937. It was marketed in the US in 1950 as sodium salt and in 1954 as calcium salt. Cyclamate is about 30 times sweeter than sugar on a weight basis, the least intense of the widely used alternative sweeteners. The ability of humans to metabolize and excrete cyclamate differs greatly, depending on an individual's intestinal flora. Like saccharin, cyclamate also has a significant bitter-sour off- taste (Larson-Powers and Pangbom, 1978). However, both the sodium and calcium salts can eliminate the bitter taste even at low dosages. Moreover, a blend of ten parts of cyclamate to one part saccharin exhibits a synergistic sweetening effect while masking the off-taste problem inherent to both components when used separately (Lipinski, 1990).
1.2.3 Aspartame
Aspartame is a dipeptide methyl esther ofaspartic acid and phenlyalanine. It was discovered in 1965 and has been marketed under different brand names including Nutrasweet, Equal, Usal, Cauderel, Pouss-suc, D-sucril, and Trisweet. Aspartame is approximately 180-200 times as sweet as sucrose on a weight basis (Vetsch, 1985) and its sweetness potency varies with its concentration and with the type offood it is used with. Aspartame can be rapidly hydrolysed in the body into two amino acids, aspartate and phenylalanine, and methanol. Therefore, it is used as a nutritive sweetener, contributing 4 KcaL^g to the diet. However, the calorie addition is negligible since only very minute quantities can give a
4 detectable sweet taste. Aspartame has been proved to be safe after extensive studies on different groups of people including pregnant women and diabetics. However, persons bom with phenylketonuria cannot metabolize phenylalanine since they lack the enzyme that catalyses the transformation of phenylalanine into tyrosine.
Alitame is another dipeptide sweetener being commercialized. It is 2,000 times sweeter than sucrose and can be metabolized normally and safely.
1.2.4 AcesuIfame-K
Acesulfame-K (ACK) is synthesized by the reaction of fluoro-，chloro-o r arylsulphonyl isocyanates with carbonyl compounds or alkynes. It was discovered in 1973. ACK is approximately 200 times sweeter than sucrose on a weight basis and its sweetness is similar to saccharin. Like saccharin, it is not metabolized by the body (Lipinski, 1985b) but is excreted in the urine unchanged. Moreover, it has excellent pH and heat stability and is even more heat-stable than sucrose. ACK elicits a sweet sensation rapidly without any induction period and persists slightly longer than does sucrose. No disturbing effects ofaftertaste were observed at low and medium (most commonly used) concentrations. Some off- taste can be detected at higher concentrations when it is used alone. ACK is also used in mixed-sweetener systems (Lipinski, 1985a).
1.2.5 Sucralose
Sucralose, a selectively chlorinated derivative of sucrose, was developed in 1988. It is 400 to 800 times sweeter than sucrose and has a high quality of sweet taste. The sweetness of this sweetener has a remarkably similar taste to sucrose. Sucralose has a very low acute toxicity, and is noncariogenic and virtually noncaloric (Jenner, 1989). It is exceptionally stable to the processing and storage conditions used for foods and beverages and is biodegradable and highly water-soluble. These advantages have caused sucralose to gain rapid acceptance by both food producers and consumers. L 3 Natural sweet plant proteins
hi addition to the artificial sweeteners, several naturally occurring proteins have been suggested as low-caloric sweeteners since they are many times sweeter than sugar. In general, substances with high molecular weight do not stimulate taste cells and hence have no taste. Likewise, most proteins are not known to elicit a sweet taste. Sweet proteins, however, are exceptions. Research has focused on sweet proteins since the 1960s because of the impending ban on cyclamate as a suspected carcinogen.
To date, seven plant proteins, namely, thaumatin, monellin, pentadin, brazzein, miracullin, curculin and mabinlin, are the only sweet or sweetness- inducing proteins known. It is interesting to note that all the above proteins are found in tropical plants. These natural sweet plant proteins have the potential to be used as harmless low-calorie sweeteners.
Recent developments in biotechnology offer opportunities to use the genes encoding these natural sweet proteins to genetically engineer plant products, such as fruitsan d vegetables, for improved taste quality. They can also be developed commercially as natural sugar substitutes. This will give consumers, especially those who prefer a natural and safer sugar-substitute, an alternative choice over the artificial sweeteners.
1.3.1 Thaumatin
Thaumatin was firstisolate d from the arils of the fruitso f the West African shrub Thaumatococcus danielli (Bannett) Benth of the Marantaceae family (van der Wel and Loeve, 1972). Traditionally, the fruito r fruitextrac t of this shrub has been used in West Afiica as a sweetening substance for sour fruit, com bread, and pahn wine.
Thaumatin is non-toxic, non-cariogenic, and has been used as a high- intensity sweetener and flavor potentiator in Japan and the UK (Higginbotham et a/.，1983). It is marketed under the trade name "Talin" and is offered as aluminum ion-adduct, which has been shown to improve colour, stability, filterability, purity, and thus higher quality sweetness, hi the highly competitive sweetener market, thaumatin, mostly from natural source, is thus the only sweet protein used as a commercial low-calorie sweetener and flavor enhancer in processed foods, soft drinks, confections (chewing gum) as well as animal feed (Etheridge, 1994).
At least fivedifferen t isoforms of thaumatin, Th I，II，III，a and b, have been isolated, depending on their charge characteristics, by ion exchange chromatography (van der Wel and Loeve, 1972; Higginbotham and Hough, 1977). These isoforms accounts for 2 % (w/w) ofthe dry weight ofthe total fruitan d 50 o/o ofthe dry weight of the aril, where they are located exclusively. The protein, with a molecular weight of 22,000 Da, consists of a single polypeptide chain of 207 amino acids with an isoelectric point of pH 12’ and eight disulfide bridges (Iyengar et al,, 1979). These bridges are responsible for the higher order structure of the molecule. They hold the protein in the correct formation for sweetness elicitation and stabilize the molecule against heat, pH denaturation, and proteolytic attack (van der Wel and Loeve, 1972; van der Wel and Bel, 1980; and van del Wel et al., 1984).
Both the C- and N-terminal amino acids of thaumatin are alanine. There are no unusual amino acids side-chain substitutions, linkages, or end-group modifications. Of all the common amino acids, only histidine is absent. Chemical modification oflysine residues in Th I showed that acetylation, but not methylation, changes the sweetness intensity, implying that the total net charge of thaumatin molecules probably plays a role in the sweetness perception (van der Wel et aL, 1982). Th I and II have been sequenced (Edens et al.’ 1982). Th II has the same total number of amino acids as Th I but with five sequence differences at amino acid positions #46, 63，67，76，and 113. Thus Th II has an extra arginine, glutamine, and aspartic acid, whereas Th I has two extra asparagines and a serine. The molecular weight of Th II，calculated from its amino acid sequence, is 22,293Da, slightly larger than Th I, which is 22,209 Da. Spectroscopic data (Korver et al, 1973) and sequence data (Eden et al,, 1982) predicted that the tertiaty structure ofthe thaumatin protein includes 10 % a-helix, 70 % P-sheet and P-bends, and 20 % random coils. The sweet taste ofthaumatin disappears completely on heating in an aqueous solution for a few minutes above 75 OQ upon splitting of the disulfide bridges in the molecule, and also when subjected to pH less than 2.5 at room temperature (van der Wel et al, 1972).
Antibodies to thamnatin cross-react with a wide variety of non-protein sweet substances (Hough and Edwardson, 1978)，suggesting that they may provide a novel method for measuring "sweetness" in vitro and may throw light on the nature of interactions between sweet substances and taste receptors. The fimction of thaumatin in plants remains to be elucidated, but the sequence similarity between thaumatin and a range of pathogenesis-related (PR) proteins has been noted (Comelissen et al., 1986; Richardson et al., 1987; Stintzi et al., 1991; Rodrigo et al., 1991; Rebmann et al, 1991; Graham et al- 1992; del Campillo et al.’ 1992; and Woloshuk et al, 1991). Thaumatin has been suggested for use as a sensitive and extremely convenient marker gene for easy identification oftransformed plants (Witty, 1989, 1992). 1.3.2 Monellin
Monellin was isolated from the red berries of the West Afiican plant Dioscoreophyllum cumminsii (Morris and Cagan, 1972). It has a sweetness about three thousand times sweeter than sugar on a weight basis.
Monellin is a basic protein (pI = 9.3) with a molecular weight of approximately 11,000 Da and is completely free of carbohydrate (Morris et al, 1972; 1973). Unlike thaumatin, monellin consists of two distinct, tightly and non-covalently bound polypeptides, an A-chain of 44 and a B-chain of 50 residues (Hudson and Biemann, 1976; Frank and Zuber, 1976). Structural studies proved that the sweetness of monellin requires dimer molecules and the individual subunits are not sweet (Bohak and Li, 1976). The secondary structure of monellin, as modeled by theoretical calculations, indicates that most of the monellin is in the P-helix form with the remainder in the a-helix form (Wlodawer and Hodgson, 1975). Monellin has been crystallized and x-ray analysis showed that monoclinic crystals contain four molecules in a unit cell (van der Wel, 1986).
Monellin has no disulfide bonds and is thermally less stable than thaumatin. Its sweetness can be eliminated by heating in aqueous solution to 55- 66 °C. Monellin becomes progressively less sweet with increasing pH, even at room temperature. Results from biochemical studies on taste sensation demonstrated that monellin can bind to bovine and to human taste tissue preparations (Cagan and Morris, 1979). These tissues can serve as models for studies on the structure-sweetness relationship in humans and several families of mammals (van der Wel, 1980). No known toxicity can be found in monellin yet and thus it is considered to be non-toxic (Krutosikova and Uher, 1992).
Although both thaumatin and monellin are potentially sweet and compete for the same sweet receptor (van der Wel and Arvidson, 1978; Brouwer, 1973)，
10 there is no similarity between their amino acid sequences. Moreover, their crystal structures appear to have no significant overall similarities except for the P-sheet structure motifs that are common in many proteins (de Vos et aI., 1985; Ogata et al., 1987).
1.3.3 Curculin
Curculin was isolated from the fruitso f Curculigo latifolia, a plant native to westem Malaysia (Yamashita et al., 1990). It is the only protein found that elicits a sweet taste and also has taste-modifying activities. It is 550 times sweeter than sucrose on a weight basis. Curculin is a diriier with two 12,000 Da polypeptides, each of which contains 114 amino acids. There are four cysteine residues in the two dimers, which are connected through two disulfide bridges. Molecular studies on curculin show that the protein is synthesized as a precursor composed of 158 amino acid residues, including a signal sequence of 22 residues and a carboxyl-teraiinal extension of22 residues (Abe et aI., 1992).
1.3.4 Pentadin and Brazzein
Pentadin and Brazzein are sweet proteins isolated from the fruitso f West African plant Pentadiplandra brazzeana Baillon (Pentadiplandraceae) (van der Wel et al, 1989; Ding and Goran, 1994). Pentadin has a molecular weight of about 12,000 Da and a sweetness of about 500 times that of sucrose on a weight basis.
Brazzein is a newly isolated high-potency thermostable sweet protein, composed of a single chain of 54 amino acids, with a molecular weight of 6,473 Da (Ding and Goran, 1994). It is the smallest protein sweetener yet discovered.
11 Its sweetness remains unchanged after incubation at 80 〇。for 4 hours. Brazzein has eight cysteine residues, the same as in mabinlin. These cysteines may tightly cross-link to form a disulfide-bonded network resulting in high thermostability. Some sequence similarity is found between brazzein and curculin, i.e. a 20 % identity in a 50-amino acid overlapping region.
1.3.5 Miraculin
Miraculin is a taste modifier found in the pulp of the red berries of the West Afhcan Richardella dulcifica (Schum) Daniell (formerly Synsepalum dulcificum). Miraculin has no detectable sweetness itself, but modifies the sour taste of another substance into a sweet taste after activation by acidic substances (higlett et al., 1965). For example, lemons elicit a sweet taste like that of oranges if eaten after the berry has been eaten.
Miraculin is a basic glycoprotein (Kurihara and Beidler, 1968; Brouwer et al., 1968) with a molecular weight of about 42,000 Da (Theerasilp et al., 1988). The amino acid sequence of miraculin consists of 191 amino acids and contains two glycosylated asparagines and seven cysteines forming three disulfide bridges (Theeraasilp et al., 1989).
1.3.6 Mabinlin
Mabinlin (MBL) was extracted from the mature seed of Capparis masaikai levi (local name: mabinlang) of the Caper family. Mabinlang grows in the subtropical region of the Yunnan Province of China and the natives usually chew its seeds, a Chinese medicine, for sweet taste.
12 MBL is a small water-soluble seed protein that accounts for 4 % of the dry weight of decorticated seed (Hu and He, 1983; Hu et al., 1985; Ding and Hu, 1986; Liu and Hu, 1988). MBL is a basic protein with a molecular weight of 12,400 Da and a pI of pH 11.3 (Hu and He, 1983). It is composed of two polypeptide chains, an A-chain of 33 amino acids and a B-chain of 72 amino acids (Liu et al,, 1993). Most of the amino acids found in the A chain are hydrophilic while the B chain also contains many hydrophilic amino acid residues. At least four isoforms, MBL I，II, III and IV，have been purified and isolated by ion-exchange chromatography (^irasawa et al，1994).
Among the four isoforms, MBL II has a much higher heat stability than the other components and has been extensively studied. MBL II is 400 times sweeter than sucrose on weight basis. Its sweetness remains unchanged even after heating for 48 hours at 85 °C (Hu et al., 1985). The sweetness of MBL III and IV，like MBL II，is unchanged after incubation for 1 hour at 80 °C, while
MBL I loses its sweetness under the same conditions QS[irasawa et al,’ 1994). The amino acid sequences ofall MBL isoforms have been determined (Figure 1- 1) ONJirasawa et al, 1994) and they share very high homology. The protein contains eight cysteine residues, which are responsible for two interchain disulfide bridges and two intrachain disulfide bridges (Figure 1-2) QS[irasawa et al, 1993). Cleavage of the disulfide bridges is commonly known to be a key determinant in the loss of sweetness in sweet proteins including thaumatin (van der Wel et al, 1984), brazzein (Kohmura et al., 1996) and mabinlin QS^irasawa et al, 1993). However, the positions of the four disulfide bridges are the same in MBL I，II，III and IV. It was found that Arg47 in the B-chain is the only residue present in all three heat-stable isoforms while it is a Ghi47 in the unstable homolog MBL I. The Arg47 residue was proposed to form a salt bridge with the carboxyl ends of the two polypeptides, which may contribute to the heat stability ofMBL II，III and IV ps[irasawa et al., 1994).
13 A-chain 10 20 30 33 i-i RRQ 'QQHQHLRA QRYIRRRAQRGGLVD II LWR Q-- --F - H P - - Q P - III IV
B>chain 20 30 36 i-l
Figure 1-1: Amino acid sequence homology among mabinlin isoforms The amino acid sequences of MBLI, II，III and IV are aligned and compared. A dash (-) indicates the same amino acid residue as in MBL I; • denotes the 8 cysteine residues responsible for disulfide bonding; (*) indicates the missing amino acid residue in the MBL homolog; and Q highlights the common R47 residue present in MBL II, III and IV while Q in MBL I.
17 32 A-chain S S
S S ^JLiU 121 ho B-chain S S s s J_i
Figure 1-2: Schematic representation of the disulfide structure in mabinlin
14 Comparisons of the amino acid sequences of MBL isoforms with those of the other sweet proteins, including thaumatin I and monelin (sweet proteins), curculin (a sweet protein with taste-modifying activity) and miraculin (a taste modifying protein), do not reveal any significant similarities between them. However, high amino acid homology was found between the sequences of MBL II and 2S seed storage proteins, especially the 2S albumin isoform 3 (AT2S3) of Arabidopsis thaliana (Figure 1-3)，although AT2S3 was not reported to have a sweet taste. It is interesting to note that both proteins consist of two chains and that the positions of all eight cysteine residues coincide in the two proteins. Plants of Capparis masaikai levi and Arabidopsis thaliana belong to different families but the same order (Rhoeadales). Both mabinlin and 2S albumin AT2S3 are contained in the seeds. Thus, both proteins may share the same evolutionary origin.
1 10 2a 30 33 MablnlinII A chain E LW!R C Q R Q F L Q H Q R L fl A C Q R F I H R R A Q F G G Q P D '• ' • • • • J| J J J j , • AT2S3 P R Q R C Q K E F Q Q S Q H L R A C Q R W M S K Q M R Q G R G G G ？ S L D D E F 41 50 M 70 10
~、 1 «b 20 . 30 M Mablntln(I，B chain E P R R P A L R Q C C N Q L R Q V 0 R P C V C P V L R Q A A Q Q V L Q R • • • • • • • • 參 • « _ • 囀 • * • • * * • • • • * • • « I • • I • • AT2S3 D E G P Q Q G Y Q L L Q Q C C N E L R 0 £ E P V C V C P T L K Q A A R A V • S L 81 90 10P 110 119
37 40 • SO 60 70 72 MiblnllnlI B chain Q I I Q G P Q Q L R R.L F D A A R N L P N I C N I P N I G A C P F R A W
AT2S3 Q G Q H G P F Q S R K I Y Q S A K Y L P N I C K I Q Q V G E C P P Q T T I P F 120 130 140 丨50 1M
Figure 1-3: Amino acid sequence similarity between MBL II and 2S seed storage protein 3 (AT2S3) of Arabidopsis thaliana
151 The cDNAs encoding the heat-stable mabinlin II and other isoforms were first sequenced and filed for patent in 1997 by Xiong et al. The structure of the MBL precursor was determined using the deduced amino acid sequence from the cDNA clone encoding the heat-stable mabinlin II fNirasawa et al, 1996) (Figure 1-4). When comparing the proteolytic cleavage sites during post-translational processing of MBL precursor with those 2S seed-storage proteins oiArabidopsis thaliana, Brassica napus, and Bertholletia excelsa, three individual cleavage sites were found to be conserved (Figure 1-5). Alanine is adjacent to the signal peptide cleavage sites of the three plants, yielding signal peptides of approximately the same length while asparagine is adjacent to the cleavage sites between the N-terminal extension peptides and the small subunits. For the cleavage sites between the linkers and the large subunits, asparagine is also the preceding amino acid except for one case (Brassica napus).
Artificial thermostable protein MBL II was successfully synthesized by Kohmura M. and Ariyoshi Y. in 1998. It is interesting to note that in the process of chemical synthesis, only the desired conformer was obtained from the folding of A and B chains.
16 CACACACTCACCCAAAACCCTAGCAATGGCGAAGCTCATCTTCCTCTTCGCGACCTTGGC 60 M A K L I F L F A T L A 12 TCTCTTCGTTCTCCTAGCGAACGCCTCCATCCAGACCACCGTTATCGAGGTCGATGAAGA 12 0 L F V L L A N A | S I Q T T V I E V D E B 32 AGAAGACAACCAACTGTGGAGATGTCAGAGGCAGTTCCTGCAGCACCAGCGACTCCGGGC 180 tt D N^Q L W R C Q R Q F L Q H Q R L R A 52 TTGCCAGCGGTTCATCCACCGACGAGCCCAGTTCGGCGGACAGCCCGATGAGCTTGAAGA 240 C Q R F I H R R A Q F G G Q P 0 ^ B L B D 72 CGAAGTCGAGGACGACA2^CGATGACGAAAACCAGCCAAGGCGACCGGCGCTCAGACAATG 300 E V E D D N 0 0 E N^Q P R R p A L R 0 C 92 CTGCAACCAGCTGCGTCAAGTGGACAGACCTTGTGTTTGCCCTGTCCTCAGACAAGCTGC 360 C N Q L R Q V D R P C V C P V L R Q A A 112 CCAGCAGGTGCTCCAACGACAAATAATCCAGGGTCCACAGCAGTTGAGGCGTCTCTTCGA 4 2 0 Q Q V L Q R Q I I 0 G P Q Q L R R L F D 132 TGCCGCAAGAAATTTGCCCAACATCTGCAACATACCCAACATCGGAACTTGCCCATTCAG 480 A A R M L P N I C N I P N I 6 T C P F R 152 AACATGGCCCTAGGCGAACCAACCAGTGGCTGACGGAGAGGTGTGTTGTAGAATCCCATG 540 T W t P • 155 TTGTAGTGTGTTAATAATATAGTTAGCATCGAGGCTAATGTCGAACTAGCACTACTCCTA 600 ATAAGAGGTTTCCAAGTTCTCTTAACTAATAAAAAAAAAAAAAAAAAAAAAAAAAA 656
Figure 1-4: Nucleotide sequence of the MBL cDNA and the deduced amino acid sequence Underlined amino acids denote sequences of the two peptide fragments used in the designing of primers for cDNA amplification and for DNA sequencing; arrows indicate the cleavage sites of post-translational processing of the MBL precursor; and * marks the stop codon.
17 SignaU peptide MAB r A.thaL. L V C T^ C 21 B.napus L V S 21 B-exce. K I S A A A A G 22 N-terminal extention peptide MAB S I Q T |T V| I E - V D E E E D N" 3， A.thal. S I Y R T V V E F E E D D A.T )7 B.napus S I Y R T V V E F D E D D A T 37 B.exce. 一 F R A T V T T T V V - E E E )¾ Small subunit (A-chain for MAB) 一 .一 _ _ Q L W R C Q R Q F L Q H Q R L R A ^ A.thal P I G P K M R - K c R K E F Q K E Q H L R A C S A G P F R I P K c R K E F Q Q A Q H L R A c Q E E c R E Q M Q R Q Q M L S H c
7： 64 Linker MAB E D E D D N D n2 A.thal. D F E D «j B.napus P L D G D F E D •M- Q Large subunit (B-chain for MAB) Q MAB - - - - Q P R R P A L R E R 0 A.thal. P Q G Q Q Q E Q Q L F Q CT< E R P P B.napus P Q G P Q Q R P P L L Q Q B.exce. P R R G M E P H M S E - lc lc r I Q H G ？ p V p p V T R QQG M A A Q Q I 了 --Q - MAB c V c L K A A K A G 一 Q Q A^tl I Q 0 G Q :hal. V p T ^ K A s K A B.napus c c K E I M c p R R E G R M M M R CI 1}一 u^ Q QM KR s R L F D Q A T - X3 K I Y A]R ^N I p" P N V C D A.thal Q Q R I y Q T Q Q R R L P K V C R M M A _£] S R |C IP 94 p I G P] F R i3 p V D P F N I P s P S V S P F Q M P G P S 77 PI R - ^ M G S 42 C-terminal extention peptide MAB P X64 -"tha•l P * 171 I A G Figure 1-5: Comparisons of the precursor sequences between MBL and the 2S seed-storage proteins of A. thaliana’ B, napus and B. excelsa Boxes indicate homologies and blue arrows mark the conserved cleavage sites. 18 1,4 Genetic Engineering of Sweet Plant Protein Selective chemical and molecular modifications are useful methods to obtain information on the functional site groups of a protein molecule, especially of a protein whose primary and tertiary structures are known, hivestigation on the relationships between biological properties and specific residues is one of the most active applications of protein modification. When modification of a particular group results in reducing taste intensity, one can consider that group to be essential for the particular taste (or the active binding site). All seven sweet plant proteins are from tropical plants that cannot be induced to produce normal fruit outside the tropics (Most et al., 1979). Even in the controlled environment of a greenhouse, sweet protein production may eventually be not enough to satisfy the potentially large demands of food processors and consumers. Recombinant technology may be a means of producing large amounts of sweet protein or other usefiil forms in transgenic organisms. However, to produce high yield of sweet protein in transgenic microbes, the large investment in controlled growth facilities where the recombinant protein can be harvested and purified must be justified. Another consideration is that as sweet protein is intended for human consumption, a host organism with GRAS (Generally Recognized as Safe) status approved by FDA would be preferable. An alternative to the application of microbial biotechnology in sweet protein production is plant biotechnology. The cost-benefit, the figure with which to judge the benefit of making a recombinant protein from a transgenic organism, is much more attractive in a plant system. The steps of purifying a protein and then adding it to food later are simplified to a one-step procedure by growing a transgenic plant. Since plant can be eaten in the form in which it is 19 produced, no expensive purification or renaturation step or well-trained personnel to maintain the expensive apparatus is required. 1.4.1 Biotechnological Studies on Thaumatin 人 4. L1 Protein modification and sweetness Thaumatin contains eight disulfide bridges. Treatment ofthaumatin with reducing agents attacks the disulfide bonds and disrupts the tertiary structure of the molecule, resulting in a complete loss of the sweet taste (van der Wel and Loeve, 1972). After partial reduction ofadisulfide bond, thaumatin demonstrates a rapid autodigestion (van der Wel and Bel, 1976). This shows that the intact conformation of thaumatin is essential for its sweet taste. These bridges are responsible for the higher order structure ofthe molecule that holds the protein in the correct conformation for eliciting sweetness and that stabilizes the molecule to heat, pH denaturation and proteolytic attack. The dominant factor for the sweet taste is the tertiary structure as indicated by psychophysical, chemical and spectroscopic data (Korve et al., 1973; van der Wel et al.，1974b). Thaumatin comprises 3 domains. Domain II and III are stabilized by disulfide bridges and may play a role in the binding of thaumatin to the taste receptor (Korver et al.’ 1973). The binding of thaumatin to its receptor depends on how well they fit together geometrically and on specific local interactions such as electrostatic attraction or repulsion between charged regions. The extreme sweetness of thaumatin is associated with a high affinity for its receptor. Acetylation of 4 of the 10 lysine residues destroys the sweetness ofthaumatin (van der Wel and Bel, 1976). Lysine residues at positions 97 and 106 seem to be of particular 20 importance for the sweet taste. Modification of tyrosine residues by iodination also destroyed the sweetness of thaumatin. Chemical modifications showed that particular basic amino acids, notably Lys and Tyr, instead of the basic character ofthaumatin seems to be important and essential for its sweetness. Thaumatin dissolved in a phosphate buffer produces no sweetness but the sweet taste was immediately perceived on rinsing the tongue with water. Verrips (1992) proposed that phosphate blocks the interaction of thaumatin with the side of the receptor protein that is involved in the initiation of the polarization of the taste cell. Thaumatin may interact, via its lysine or arginine residue, with the membrane of the taste cell. He ftuthersuggeste d that this mechanism was similar to that ofprotein signal sequences for translocation in micro-organisms and plants. The positively charged amino acids of the signal sequences interact with the phosphate group ofthe phospholipids of membranes (Saier et al., 1989). Site-directed mutagenesis of specific amino acids has also been done on thaumatin. Mutation at position 67 (Lys67 ~> Glu or Ala) in Domain II reduces the sweetness. Over 40 different positions have been mutated to study the properties of sweetness, folding, and stability ofthaumatin. hiterestingly, most of the mutations that reduce the sweetness are localized on one side of the molecule, specifically, where domains II and III come together. L4,1.2 Transgenic expression in microbes biitial studies using microorganisms to produce thaumatin were conducted in E. coli with the gene under the transcriptional control of the lac or E. coli trp promoter/operators (Edens et al., 1982). The recombinant protein synthesized was preprothaumatin with extension at both the amino terminus and the carboxyl terminus. Cloning of preprothaumatin, prethaumatin, prothaumatin and mature 21 thaumatin was also tried in E. coli. All forms were expressed but at low levels (Edens et al., 1982). The expression of the gene for mature thaumatin was abnost undetectable. The protein produced from these studies was not biologically active since E. coli was found incapable of processing preprothaumatin and secreting the protein. Expression of thaumatin in S. cerevisiae under the control of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) promoter, a strong promoter in yeast, resulted in the synthesis of mature thaumatin with identical molecular size as the plant-derived one (Edens et al., 1984). However, the yield was quite low. Insertion of thaumatin gene into the rDNA locus of S. cerevisiae resulted in the integration ofabout,50-100 copies ofth'e gene and a more than ten- fold increase in production (Lopes et al.，1989). Another yeast system used to express thaumatin II gene was Kluyveromyces latic (Edens et cd., 1983). Neither system, however, gave a correctly folded protein. Huang et al. (1987) synthesized variants of the natural thaumatin genes using the preferred yeast codons. Using the powerful phosphoglycerokinase (PGK) promoter, the production of thaumatin was enhanced to 20 % of the total cellular protein. The expressed protein was not sweet, although the protein showed cross-reactivity with antibodies raised to thaumatin. Therefore, yeast appears to lack the ability to process thaumatin into a correctly-folded and ftmctionalprotein . When the inactive thaumatin derivative was purified and subjected to protein folding in vitro, the refolding resulted in an immunoreactive conformation and the sweet activity restored (Lee et al., 1988). The genetic engineering and expression of the thaumatin I gene in S. cerevisiae (Weickmann and Ghoshdasti, 1987) and the studies on synthetic thaumatin II gene in E. coli (Faus et al.’ 1996) were described. Studies on the expression ofthaumatin in other microbes including Bacillus subtilis (Illingworth 22 et al., 1988)，Streptomyces lividans (Illingworth et al., 1989) and Aspergillus oryzae (Hah and Blatt, 1990) were also reported. L4,L3 Transgenic expression in higher plants To evaluate the possibility of efficient expression of a biologically active thaumatin in different crop plants, Witty (1990) inserted the gene for preprothaumatin behind the CaMV35S promoter and used Agrobacterium rhizogenes A4T and the shuttle vector pWIT2 to transform Solanum tuberosum cv. "Iwa" as a model system. He used the hairy root transformation technique (Witty, 1990) to successfiilly regenerate hairy root tissue into whole plants. The characteristic sweet taste and aftertaste of thaumatin was found in both hairy root samples and in plant tissues obtained from regenerated potato plants. Although data on precise processing of the prothaumatin were not provided, the product was perceived as sweet tasting, a strong indication that processing and folding of the transgenic product were correct and biologically active thaumatin was expressed (Verrips, 1992). Due to its sweet fimction, the thaumatin has been proposed as a screenable marker for transformation, and a vector (pWIT2) containing the thaumatin expression cassette was constructed (Witty, 1989，1992). The strong native thaumatin promoter was also used in a similar procedure to try to overexpress the thaumatin gene. However, no general expression was seen in any tissue since thaumatin only accumulates in one specific organ, the aril of T. daniellii. 23 1.4.2 Biotechnological Studies on Monellin L4.2.1 Gene modification and transgenic expression in microbes Monellin consists of two peptide chains as in insulin, suggesting that it might be synthesized in microbes, as is insulin (Goeddel et al., 1979). However, microbes are usually unable to express the correctly-folded and biologically- active form of recombinant proteins. Based on this general thought, Kim et al. (1989) designed and synthesized linkers, using the crystal structure ofmonellin as reference, which were similar to the original loop structure akeady present in monellin. These linkers were used to connect the two peptide chains and to give rise to a single-chain monellin without significant alteration of its tertiary structure (Kim et al., 1989). E. coli was used to express these genes and all chimeric constructs were proved to be sweet. One ofthe newly designed proteins, lacking Phe of the A-chain, was found to be as sweet as the natural monellin but more heat and pH stable and more easily renatured after heat denaturation. The food yeast Candida utilis was also used to express the single-chain monellin (Kondo et al., 1997) under the control of the glyceraldehyde-3- phosphate decarboxylase promoter. The high copy number integration of this transgene into the yeast chromosome expressed a very high level of monellin, accounting for more than 50 % of the soluble protein. The high level expression with this regulatory-friendly-host and simple purification procedure pave a road for the commercial development of monellin, as for thaumatin, in the highly competitive low-calorie sweetener market. 24 1,4,2.2 Transgenic expression in plants Lettuce and tomato were transformed (Penarrubia et al., 1992) with a synthetic single-chain monellin gene (Kim et al., 1989)， encoding both polypeptide chains linked by a hinge sequence under the control of the constitutive promoter CaMV 35S or the fruit-ripening specific tomato promoter E8. These chimeric gene expressions resulted in the accumulation of modified monellin protein in the leaf tissues and fruit of these plants to significant levels, as detected by immuno-blotting assay. However, a sweet taste was detected only when the transgenic tomato fruits containing the monellin gene under the regulation of the E8 promoter was subjected to ethylene treatment. These fruits gave a faint sweet taste characteristic of the monellin protein. Preliminary observations suggest that monellin must comprise approximately 1 % (by weight) of the total tomato protein before its flavor can be recognized. These results suggest that by using appropriate promoters it may be possible to produce sweet protein in a variety of transgenic fruits and vegetables, and to enhance their flavor and quality. Although the foreign genes that have been introduced into plants may not be stable after propagation due to the inactivation of the transgenes (Finnegan and McEboy, 1989), the production of monellin was shown to be stable and heritable. On the other hand, the precise amount of monellin which is required to elicit a detectable level of sweetness in other transgenic fruits need to be determined and new strategies should also be developed to increase the production ofmonellin in transgenic plants. 25 1.4.3 Transgenic Expression of Mabinlin in Plants To date, no published data were found on transgenic expression of Mabinlin in microbes. Transgenic expression of mabinlin in potato microtubers was reported (Xiong and Sun, 1996). The cDNAs encoding MBL isoforms were successfully cloned and sequenced by these authors. MBLII was transformed into Solanum tuberosum microtubers, under the control of constitutively expressed CaMV35S promoter and the tube-specific patatin promoter, respectively. The average transformation efficiency was about 5 %. GUS staining, Southern hybridization, and RT-PCR, were carried out to prove the successful gene integration and the presence of RNA transcription in potato tubers, and protein expression was detected by the westem blot technique. 26 1.5 Phaseolin and its regulatory sequences Phaseolin is the major seed storage protein in French bean (Phaseolin vulgaris), accounting for 40-60 % of the total protein (Osbom, 1988). It is a globulin as it is salt soluble. Phaseolin consists of 3 subunits, a, P and y polypeptides with corresponding molecular weights ofabout51, 48 and 45.5 kDa. The synthesis of phaseolin, under strict tissue and temporal regulation, during embryo development in French bean was studied by Sun et al. (1978). It occurs only in the cotyledon of the bean. The three polypeptides are encoded by 16s mRNA species. The rapid protein accumulation in the developing cotyledon begins when the cotyledons are about 7 mm in len^ and continues until the cotyledons reach 17-19 mm in length. The P-phaseolin gene was firstlyisolate d from French bean by Sun et al. in 1981. This gene was found to contain 80 bp of5' untranslated region, 1263 bp of protein-encoding region, which is interrupted by 5 intervening sequences, and 135 bp of 3' untranslated region. Since phaseolin is an abundant seed storage protein, accounting for 50 % of the total protein in mature seeds in French bean (Ma et al., 1978)，the phaseolin promoter is a good candidate for heterologous expression of a target protein in transgenic plants，for both basic molecular studies on gene expression and for nutritional quality improvement of plant seeds. For example, the P-phaseolin promoter was shown to direct high-level expression of a phaseolin-coding sequence (Sengupta-Gopalan et al., 1985). The gene encoding a maize 15 kDa methionine-rich P-zein was placed under the control of the 5' flanking region of the French bean phaseolin gene (Hoffman et al., 1987) and transformed into tobacco plant. The methionine-rich zein protein was found to deposit in both the endosperm and embryo tissueso f tobacco seeds, and accumulate to 1.6 % as the total tobacco seed protein. The cDNA encoding 27 the Brazil nut methionine-rich protein was attached to the 5’ and 3' regulatory regions of the phaseolin gene and expressed in tobacco (Altenbach et al, 1989). It was found that the tobacco seeds were able to process the methionine-rich protein from its 17 kDa precursor polypeptide into the 9 kDa and 3 kDa subunits of the mature protein. A methionine-rich seed protein coding sequence from Brazil nut was used to replace the coding region of the phaseolin gene and the construct was transferred into canola (Altenbach et al., 1992). The Brazil nut protein was found to accumulate to high levels in transgenic canola seeds. 28 1.6 Arabidopsis 1.6.1 Arabidopsis thaliana as a model plant Arabidopsis thaliana is a flowering plant and an inconspicuous weed belonging to Brassicaceae. Yet it is a typical higher plant in most aspects, and is suitable for use in studies on molecular genetics (Meyerowitz, 1989; Danun and Halfer, 1993) because of its unusually small genome, hi addition, its small size, short generation time，and ability to grow well in constant fluorescent light make indoor maintenance of large population of Arabidopsis plants inexpensive and easy. The abundant seed production and small size of the seed produced after pollination make large-scale mutagenesis easy, as for example when plants are exposed to mutagens such as ethyl methanesulfonate (EMS). Many mutations have been identified, studied and mapped. According to the law of Mendelian inheritance, transgenic plants containing single transgene insertion (heterozygous), after self-pollination (Monohybrid cross), will give dominant phenotypes in three-fourths of its F1 offsprings. Furthermore, self-pollination of those F1 offsprings containing homozygous dominant genotypes will give dominant phenotypes in 100% of F2 offsprings. These F2 offsprings are said to be homozygous transgenic plants because they contain homozygous transgenes (the desired transgene present in both of its chromosomes). Therefore, Arabidopsis has become a desirable model organism for studies on a broad range ofbiological problems. 1.6.2 Transformation methods Simple and easy DNA-mediated transformation of Arabidopsis makes it valuable for studying factors involved in gene expression or fimction. Several 29 different methods had been employed for transformation of target DNA into Arahidopsis genome, including direct DNA uptake using particle bombardment, polyethylene glycol (PEG) treatment, and ^gro^ac/er/wm-mediated transformation by cocultivation of callus tissue, leaf, or root explants with Agrobacterium tumefaciem, or the vacuum infiltration method that involves direct application of Agrobacterium to the plant and recovery of transformants in the progeny. 1.6.2.1 DirectDNA uptake This method is especially important for monocots due to the low efficiency and failure ofusing Agro^acrer/ww-mediated transformation. Biolistic gene transfer ^)article bombardment) can deliver the DNA into plant cells (Seki et al., 1991a) using high-velocity acceleration of microprojectiles carrying foreign DNA, and the penetration of the cell wall and membrane by microprojectiles. Transient GUS expression was demonstrated in both the leaves and roots of Arabidopsis using particle gun delivery. Using the same transformation method, stable transformation ofArabidopsis was also reported by Seki etal.(l99lhy Polyethylene glycol (PEG) is a common chemical treatment used to facilitate DNA uptake into protoplasts of both dicots and monocots. The transformation of Arabidopsis using direct gene transfer method to protoplasts (Damm et al., 1989) was based on the PEG-mediated direct gene transfer method initially established for tobacco Q^egrutiu et al., 1987). 30 L6,2,2 Agrobacterium-mediated transformation This method was widely used after a simple and convenient leaf-disk method was devised (Horsch et aL, 1985) and proved to be effective in transferring genes into plants. This approach is based on the capability of the plant pathogenic Agrobacterium to integrate sequences of its tumor-inducing (Ti) plasmid DNA into the genomic DNA ofthe host plant. Leaf disks ofArabidopsis was inoculated with Agrobacterium containing Ti plasmid harboring a hygromycin resistant gene to generate transgenic plants (Lloyd et al., 1986). The Arabidopsis callus tissue was transformed by cultivation with Agrobacterium containing an avirulent Ti plasmid carrying the NPTII gene (Zhang and Somerville, 1987). A protocol was developed to transform Arabidopsis root explants by cultivation with Agrobacteria containing the kanamycin and herbicide resistance genes (Valvekens et al, 1988). The first non-tissue culture approach of Agrobacterium-mQdiaiQd transformation has also been developed in germinating seeds of Arabidopsis thaliana (Feldmann and Marks, 1987). As this method does not involve tissuecultur e for regeneration, it will not cause somatic mutations during the transformation process. Therefore, this approach is suitable for transformation of plant species recalcitrant to tissue culture and regeneration. However, the efficiency of whole plant transformation is quite low and variable. 1.6.2.3 In planta transformation The first in planta transformation method was described by Feldmann and Marks (1987) and consisted of imbibition of seeds with Agrobacteria. It was further developed by Chang et al. (1994) in which young inflorescences were cut 31 offand the wounded surfaces were inoculated with Agrobacterium bearing binary vectors. A transformation method using vacuum infiltration of Agrobacterium into mature Arabidopsis was also reported (Bechtold et ai, 1993，1995 and 1998). The possibility of up to 5 mechanisms involved in this transformation method is largely unknown. By this method, all the transformants are independent and hemizygous for T-DNA insertions, and the target cells for Agrobacterium transformation are assumed to be the zygote itself or the gametes. This simple non-tissue culture procedure yields stably transformed A. thaliana seeds in only 7-8 weeks. On average, 5.5 % of the newly regenerated shoots produces transformed progenies. More than 90 % of the transformed progeny exhibit Mendelian segregation pattems of th'e marker genes. Of those, 60 % contain one functional insert (Katavic et al, 1994). This plant transformation procedure is a reliable and reproducible method. As this method apparently does not require tissue wounding which is assumed to be a pre- requisite for Agrobacterium-mQdmtQd transformation, it was proposed that tissues that have stomata and are undergoing expansion may be more subjected to transformation during infiltration. 32 2 General introduction and Hypothesis 2.1 General Introduction Among the natural sweet proteins discovered so far, only thamnatin has thus far reached a highly competitive low-calorie sweetener market. Although thaumatin has been available in food products in Japan and Europe for more than a decade, its limited source makes it very expensive. Many practical difQculties hinder the development of sweet proteins, including heat and pH instability of sweet proteins during food or beverage processing, limited supply of plant resources, and uncertainty in the regulation offood additives. The development of recombinant technology opens a new door for the production of sweet proteins in heterologous systems. Recombinant thaumatin was shown to accumulate to 20 % of the total protein in transgenic Saccharomyces cerevisiae (Lee et al, 1988). Thaumatin protein also showed strong expression in Solanum tuberosum (Zenmanek and Wasserman, 1995). However, the solubility and sweetness of the recombinant protein were not consistent. Therefore, the availability ofthaumatin to industry is still very limited and expensive. Mabinlin shows relatively higher stability to heat and acidic treatments than other sweet proteins; thus it has advantages over other sweet plant proteins for use as natural sweetener. However, it still loses sweetness after breakage of its disulfide bonds, similar to most sweet plant proteins. The only difference between the heat-stable MBL isoforms II，III and IV and heat unstable MBL I，at protein structure level, is the 47th amino acid in large-subunit and not the position of disulfide bridges (Nirasawa et al., 1993). This suggests that amino acid arginine at position 47 in the large subimit may play a role in heat stability. 33 2.2 Hypothesis Based on the information from previous studies on the sweet protein mabinlin, the amino acid arginine at position 47 in the large subunit will be tested for its contribution to the mabinlin heat stability. The following hypotheses are proposed for testing: 1. MBL I and MBL III，are highly similar in their sequences despite the three amino acid differences (Figure 2-1) (Xiong et al., 1997). They are the most suitable candidates of heat-unstable and heat-stable homologues, respectively, for use to test the hypothesis. 2. If the basic amino acid arginine of position 47 'of heat-stable MBL III is replaced by the neutral amino acid glutamine (as in heat-sensitive isoform I) or a similar-sized basic amino acid (e.g. Lys), the mutant wiU reverse or maintain its heat-stable property, respectively (Table 2-1). 3. Similarly, if the neutral amino acid glutamine at position 47 of heat-sensitive isoform I is replaced by the basic amino acid arginine or a similar-sized acidic amino acid (e.g. Asp"), the mutant will reverse or maintain its heat-sensitive properties, respectively (Table 2-1). 4. To test these hypotheses, wild-type and modified mabinlin genes (Figure 2-2) (Sun et al.，1997) under the control of a strong seed-specific phaseolin promoter and terminator will be constructed and expressed in Arabidopsis. The integration and expression of the wild type and mutant MBLs will be analysed in transgenic Arabidopsis seeds. These transgenic plants will establish a model system to test hypotheses 2 and 3. 5. If time permits, the heat stability of the mutant MBLs (Hypotheses 2 and 3) will be tested. 34 Signal peptide MBL + II M A K L I P L P A T L A L F V L L A N A S 1 L • L T III L • L T Small subunit ( A chain) E V D E D N W R D D E N Q D D E N Q Linker ？ R G • 6 » n*L Lp*l vD •. H R F I H R R A • R • Y[T|R • , • A- H • y 丘 R • • • • « Large subunit ( B chain) D E £ - D D N D D E N R P A L Q ,E •'- • 2 • • D E • G » • > E D G « • • c Q V D P 為& .N .N H • Q N K I . p • N & I V G .A. R c .？. F • R T A R Q K A V R Q T R « • • Figure 2-1: The complete amino acid sequences ofMBL isoforms I，11 & III indicates amino acid differ between MBL I & III. 35 Table 2-1: Hypotheses to test the importance of 47th amino acid of the large subunit on MBL thermal stability Mutant: MBL III-Gln: heat sensitive ？ MBL I-Arg:- heat stable ？ To test the CAA AGA importance of Q R 47^ amino acid (Ghi) (Arg+) i k i i MBLED-wt heat stable. MBL ^wl: heat sensitive. W^L.—. 一—‘一 一“"CAA DNA sequence CGrT - 47^ amino acid p• of large subunit (Arg+) (Gln) ^ r ，r Mutant: MBL III-Lys: heat stable ？ MBL I-Asp: heat To test the sensitive ？ importance of — —AAG ^ — the charge of D 47^ amino acid (LyO (Asp-) 36 Small subunit ( A chain) i 1 MBL II ATGGCGAAGCTCATCTTCCTCTTCGCGACCTTGGCTCTCTTCGTTCTCCTAGCGAACGCJ： I ATGGCGAAGCTCATCCTCCTCTTGACCACCTTGGCCCTCTTTGTTCTCCTGGCCAACGCC 111 ATGGCGAAGCTCATCCTCCTCTTGACCACCTTGGCCCTCTTTGTTCTCCTGGCCAACGCC TCCATCCAGACCACCGTTGTCGAGGTCGATGAAGAAGAAGACAAC CAA TCCATCTACCGCACCACCGTCGAGCTCGACGAAGAA——GACAACGACGATGAGAACCAO TCCATCTACCGCACCACCGTCGAGCTCGACGAAGAA——GACAACGACGATGAGAACCAO CTGTGGAGATGTCAGAGGCAGTTCCTGCAGCACCAGCGACTCCGGGCTTGCCAGCGGTTC CCCCTG TGCCGAAGGCAGTTCCAGCAGCACCAGCACCTCAGGGCTTGCCAGAGGTAC CCCCTG TGCCGAAGGCAGTTCCAGCAGCACCAGCAGGTCAGGGCTTGCCAGAGGTAC Linker i ATCCACCGACGAGCCCAGTTCGGCGGACAGCCCGATGAGCTTGAA GACGAAGTC ATCCGCCGCCGAGCCCAAAGAGGTGGATTGGTAGACGAGCTAGAGCTAGAAGAC GTC CTCCGCCGOCGAOCCCAAAGAGGTGGATTGGCAGACGAGCTTGAGCTAGAAGAC GTC Large subunit (B-chain)^ GAG—GACGACAACGATGACGAA AACCAGCCAAGGCGACCGGCGCTCAGACAA GAGGAA AACGAAGATGAAGACGAAAACCAGCAGAGGGGACCGGCGCTCCGACTA GAGGAA AACGAAGATGAAGACGAAAACCAGCAGAGGGGACCGGCGCTCCGACTA TGCTGCAACCAACTGCGTCAAGTGGACAGACCTTGTGTTTGCCCTGTCCTCAGACAAGCT TGCTGCAACCAACTGCGTCAGGTGAACAAACCCTGTGTTTGTCCCGTCCTCAGACAAOCT TGCTGCAACCAACTGCGTCAGGTGAACAAACCCTGTGTTTGTCCCGTCCTCAGACAAGCT GCCCAGCAGGTGCTCCAGCGACAAATAATCCAGGGTCCACAGCAGTTGAGGCGTCTCTTC GCCCACCAACAGTTGTACCAGGGACAAATCGAAGGTCCACGCCAGGTGAGGEA3CTATTT GCCCACCAACAGCTGTACCAGGGACAAATCGAAGGTCCACGCCAGGTGAG^GjcTATTC GATGCCGCAAGAAATTTGCCCAACATCTGCAACATACCCAACATCGGAGCTTGCCCATTC AGAGCCGCCAGGAACTTGCCCAACATCTGCAAAATCCCCGCCGTCGGACGCTGCCAGTTC AGAGCCGCTAGGAACTTGCCCAACATCTGCAAAATCCCCGCCGTCGGACGCTGCCAGTTC * * * AGAGCATGGCCCTAO ‘ . ACGAGATGG TAG ACGAGATGG TAG Figure 2-2: Comparison of the nucleotide sequences of the cDNAs encoding the MBL isoforms (MBL I，II and III) Black arrow indicates start codon (ATG); *** denotes stop codon;-一 shows a gap resulting from maximizing nucleotide matching; ]indicates the 47^ codon in the large subunit. 37 3 Molecular Studies on Sweet Protein Mabinlin : Thermal Stability 3.1 Introduction Mabinlin (MBL), a sweet plant protein from mabinlang seeds, has been purified and characterized for its chemical and physical properties, structure, and sweetfimction(Hu&He, 1983;Hue^a/., 1985;Ding&Hu^ 1986;Liu&Hu^ 1988; Liu et al., 1993; and Nirasawa et al., 1993，1994). At the molecular level, cDNAs encoding the MBL isoforms have been cloned, sequenced and characterized (Xiong et al.，1997). To study the relationship between the 47^ amino acid in the large subunit ofMBL and its thermal stability, single aunino acid modifications will be performed in cDNAs encoding the heat-stable MBL III and heat-unstable MBL Chimeric genes encoding wild type MBL III Q>BI / phas / MBLIII-wt), mutant MBL III Arg47^Ghi (pBl / phas / MBLIII-GLN), and mutant MBL III Arg47^Lys ^>BI / phas / MBLIII-LYS), under the control of the regulatory regions of the seed-specific phaseolin gene, will be constructed and transformed into Arabidopsis. These chimeric genes, without a signal peptide ofphaseolin, are expected to direct the expression of MBL III into the cytosol of transgenic Arabidopsis seeds and to reverse or maintain the heat-stability of the wild-type MBL III protein. To further analyze if the 47^ amino acid of MBL large-subunit and/or the charge of this amino acid would affect the thermal stability of MBL protein, chimeric genes encoding the wild type MBL I (pBI / phas / MBLI-wt), mutant MBL I Ghi47^Arg {jpBl / phas / MBLI-ARG), and mutant MBL I Gh47^Asp (pBI / phas / MBLI-ASP) under the control of the regulatory regions of the phaseolin gene will also be constructed and transformed into Arabidopsis. The 38 modified proteins are expected to reverse or maintain the heat-instabiUty of wild- type MBLI protein, respectively. 39 3.2 Materials 3.2.1 Laboratory Wares All general glasswares were cleaned and rinsed in distilled water. Before using for DNA experiments, they were autoclaved at 121 °C for 20 minutes at lkg/cm^. Before using for RNA experiments, they were wrapped by heavy duty aluminium foil and baked at 240。C for six hours. All plastic-wares were sterile and disposable, or autoclaved at 121。C for 20 minutes at lkg/cm^ before DNA experiments. Those for RNA experiments were immersed in 1 % DEPC-treated ultra-pure water and kept ovemight shaking. They were then autoclaved at 121 ®C for 20 minutes to remove residual DEPC. 3.2.2 Equipments MJ Research Peltier Thermal Cycler was used to perform PCR. Perkin Ehner Thermal Cycler and ABI Prism 310 Genetic Analyzer were used to perform automatic DNA seuqencing. Hewlett Packard N-terminal Protein Sequencer was used for protein sequencing. 3.2.3 Chemicals All chemicals and reagents were reagent grade, molecular grade or better. They were purchased from Sigma Chemical Co., Bio-Rad Co. or otherwise specified. Molecular enzymes were mostly purchased from Promega Co., NewEngland Biolabs Co. or otherwise specified. 40 3.2.4 Commerical kits Table 3-1: Name and brand of commercial kits used in different experiments Experiment Name of klt Brand Mini-preparation of Wizard Plus Minipreps DNA Promega plasmids Prification System Gel purification of Sephaglas™ Band Prep Kit Pharmacia DNA Site-directed Altered Sites II In vitro Mutagenesis Promega mutagenesis System T/A cloning pGEM®-l Easy Vector System Promega Automatic cycle dRhodamine Terminator Cycle Perkin Elmer sequencing Sequencing Kit Protein assay Bicinochoninic Acid BioRad Westem blot AURORA Westem Blot ICN Chemiluminescent Detection System 41 3.2.5 DNA primers Table 3-2: Primers designed for megaprimer PCR Name of primer Sequence (5，to 3，) Length (nt) GC/AT MBLFOR CCC CGT CTA CAT GGC GAA GCT 30 18/12 CAT CCT CCT MBLREV CCC CGT ATA CCT ACC ATC TCG 30 17/13 TGA ACT GGC MIII - GLN CTC TGA ATA G|TT G|CC TCA CCT 23 12/11 GG MIII-LYS CTC TGA ATA G|CT T|CC TCA CCT 23 12/11 GG MI-ARG TCT AAA TAG |^ CCT CAC CTG ^ 11/12 GC MI-ASP TCT AAA TAG |ATC| CCT CAC CTG 23 11/12 GC Table 3-3: Primers designed for in vitro mutagenesis Name of primer Sequence (5，to 3，) Length (nt) GC/AT pA-MIII GLN CCA CGC CAG GTG AGG ^^ 33 20/13 CTA TTC AGA GCC GCT pA-MIII LYS CCA CGC CAG GTG AGG AAG 33 20/13 CTA TTC AGAGCC GCT pA-MI ARG CCA CGC CAG GTG AGG AGA 32 19/13 CTA TTT AGA GCC GC pA - MI ASP CCA CGC CAG GTG AGG _ CTA 33 20/13 TTT AGA GCC GCC 42 Table 3-4: Primers designed for checking DNA sequence Name of primer Sequence (5，to 3，) Length (nt) GC/AT pPha-1 ATC ATC CAT CCA TCC AGA GT 20 9/11 pPha-2 AGA AGT GAG ATG GAG CTC AG 20 10/10 Note: Underline sequence : restriction enzyme site AccI to be added mutated codon sequence to generate single amino acid mutation 3.2.6 DNA plasmids The wild-type cDNAs of MBI I & III，containing in the plasmids pBluescript, were provided by Prof. Sun, S.S.M. (CUHK, Hong Kong). Plasmid pAlter-exl for site-directed mutagenesis was from Promega (Altered Sites II in vitro Mutagenesis System). The plasmid pD38 containing phaseoUn gene sequence and pTZ-19U sequence was also provided by Prof. Sun, S.S.M. The Agrobacterium binary vector pBI121 was purchased from Clontech, USA. Most other plasmids were purchased from Strategene or otherwise specified. 3.2.7 Bacterial strains Most plasmid DNAs were manipulated in E.coli DH5a cells, or otherwise specified. For site-directed mutagenesis, E.coli JM109 and ES1301 mutS were provided in the kit (Promega). For plant transformation, Agrobacterium tumefacien GV3101 / pMP90 were provided by Prof. Lam, H.M. (CUHK, Hong Kong). 43 3.2.8 Plant materials Arabidopsis thaliana ecotype Columbia-0 was provided by Prof. Lam, H.M. (CUHK, Hong Kong). 3.2.9 Protein and Antibody Purified MBLII protein was provided by Dr. Hu Zhong (Kunming histitute of Botany, China). Rabbit antiserum against purified MBL II linearized protein and large subunit plus small subunit was self-prepared. 44 3.3 Methods 3.3.1 Transformation oiArabidopsis with MBLIII and MBLI genes 5.3,1.1 Construction of mutant MBLIII and MBLI genes containing single codon mutation by megaprimer PCR pBK-MBLIII and pBK-MBLI (Xiong et al., 1997) harboring the fuU length MBLIII and MBLI cDNA, respectively (Figure 2-2) in pBluescript SK" vector (Strategene) were used as the starting materials. A 25p,l PCR reaction mixture containing 20 ng pBK-MBLIII / pBK-MBLI, 1.5 mM MgClj, 0.2 mM dNTPs, l[M MBLFOR primer, l|iM MIII-GLN, or MIII-LYS, and MI-ARG, or MI-ASP primer (Table 3-2), IX Tfl reaction buffer [50 mM Tris-HCL (pU 9.0), 20 mM pffl4)2SOJ and 1 unit Tfl thermostable DNA polymerase (Epicentre Technologies) was used to amplify and generate the corresponding MBLIII and MBLI megaprimer sequences, respectively (from start codon to 10 bp 3，downstream of codon sequence being mutated). The conditions for hot start PCR reaction were as follows: 1 cycle of95。C for 10 minutes (added DNA polymerase after 7 minutes), then 30 cycles of 95。C for 1 minute, 57。C for 1 minute, 72 T for 1 minute followed by 1 cycle of 72。C for 7 minutes. The double-stranded product of 1^ round PCR (417 bp) was purified and used as one of the primers (so called "megaprimer") in the second round of PCR along with the other flanking primer (MBLREV). A 25p,l PCR reaction mixture containing 20 ng pBK-MBLIII or pBK-MBLI, 1.5 mM MgCl2, 0.2 mM dNTPs, l^iM MBLREV primer (Table 3-2), 680 ng (0.1 ^M) of megaprimer {V' PCR purified DNA), IX Tfl reaction buffer and 1 unit Tfl thermostable DNA polymerase was used to amplify the 3' downstream DNA sequence of the mutagenic codon (from 11 bp 3，downstream of codon sequence being mutated to stop codon). The conditions for touchdown PCR reaction were as follows: 1 cycle 45 of95。C for 10 minutes (added DNA polymerase after 7 minutes), then 35 cycles of 95。C 1 minute, annealing temperature for 1 minute, 72。C for 1 minute (annealing stage for 13 cycles per temperature step beginning at 93� Cand decreasing at 3 ®C increments to 57 °C; followed by 22 additional cycles at 57。C)，then finally 1 cycle of 72 °C for 7 minutes. The resulting DNA product was MBLIII cDNA or MBLI cDNA sequence containing single codon mutation flanked by the two primers introducing AccI sites. This double-strand mutant DNA product fromth e 2°^ round ofPCR (494 bp) was purified. The purified double-stranded PCR product was used as the DNA template for reamplification. A 25 ^il PCR reaction mixture containing 20 ng DNA template, 1.5 mM MgCl2, 0.2 mM dNTPs, l^iM MBLFOR primer, l^iM MBLREV primer (Table 3-2)，IX Tfl reaction buffer and 1 unit Tfl thermostable DNA polymerase was used to reamplify mutant MBLIII cDNA and mutant MBLI cDNA flanked by 2 AccI restriction enzyme sites. The PCR conditions for reamplification were as follows: 1 cycle of95。C for 10 minutes (added DNA polymerase after 7 minutes), then 35 cycles of95 °C for 1 minute, annealing temperature for 1 minute, 72 °C for 1 minute (annealing stage of 11 cycles per temperature step beginning at 70。C and decreasing at 1 T increment to 60。C，followed by 24 additional cycles at 60。C)， then a final cycle of 72。C for 7 minutes. The wild-type MBLIII and wild-type MBLI cDNAs flanked by two AccI sites were amplified by the same conditions as used in the 1^ round PCR as described before, except using primers MBLREV instead ofprimers MIII~GLN or -LYS and MI-ARG or -ASP, respectively, in the reaction mixture. The resulting PCR product was 494 bp for both wild-type MBLIII and MBLI cDNAs and for reamplified mutant MBLIII GLN cDNA, MBLIII LYS cDNA, MBLI ARG cDNA, and MBLI ASP cDNA. These six PCR products were gel purified for later experimental use. 46 pBK-MBLIII or I (wild-type) 5'_ CGT V 5, mutagenic primer ^MBLFpR 5, round PCR V 5- 兴 ^ 5, mutangenic codon isolate 418bp DNA fragment (megaprimer) V MBLREV、, megaprimer _yv 5' 2"d round PCR V Acc I mutagenic codon Acc I * Mutant MBLIII or I nDNA ">e isolate 494bp DNA V If the yield is poor, PCR reamplification using MBLFOR and MBLREV primers Figure 3-1: Strategy to construct mutagenic MBLIII and MBLI cDNAs 47 3.3.L2 Cloning ofPCR-amplifiedMBLIII and MBLI cDNAs into vector pD3'8 As shown in Figure 3-2, the purified PCR products from Section 3.3.1.1 including both the wild-type and the mutant MBLIII and MBLI cDNAs flanked by AccI sites, were first digested with AccI and ligated to the AccI-cut pD3-8 (Doyle et al., 1986 and Slightom et al., 1983) to replace the phaseolin coding sequence. AccI AccI AccI Wild-type MBLIII or I cDNA AccI AccI mutagenic codon »<- Hindlll Mutant MBLIII or I cDNA 〜0.5 kb Acc I digestion & ligation AccI AccI pDMBLIII or Hindlll 5.8kb Figure 3-2: Strategy to clone MBLIII and I chimeric cDNAs into pD38 48 3,3,L3 In vitro site-directed mutagensis ffor the construction of MBLIII and MBLI cDNAs containing single codon mutation) Since the method of megaprimer PCR was inconsistent and inefficient in generating the mutagenic PCR products and not all six PCR products could be cloned into pD3-8, an alternative method was used to achieve the site-directed mutagenesis. The Altered sites II in vitro Mutagensis System ofPromega was used to create single codon mutation in MBLIII and MBLI genes, respectively. The strategy was shown in Figure 3-3. Plasmids pBK-MBLIII and pBK-MBLI (Xiong et al., 1997) which carry the full length MBLIII and MBLI cDNA, respectively (Figure 2-2) in pBluescript SK(-) vector (strategene) was used as DNA template. Primers MBL-FOR and MBL-REV were used to amplify the MBLIII cDNA and MBLI cDNA sequence from start to stop codon. The 25 ^1 PCR reaction mixture contained 20 ng pBK-MBLIII or pBK-MBLI plasmids DNA, 1.5 mM MgClj, 0.2 mM dNTPs, 1 jiM MBLFOR, 1 ^M MBLREV, IX buffer IV [75 mM Tris-HCl ^>H 8.8), 20 mM 0vJH4)2SO4 and 0.01 % Tween] and 1 unit thermoprime polymerase (Advanced Biotechnology). The PCR conditions were as follows: 1 cycle of 95 °C for 10 minutes (added polymerase after 7 minutes), then 35 cycles of95 °C for 1 minute, annealing temperature for 1 minute, 72 °C for 1 minute (annealing stage of 17 cycles per temperature step beginning at 94 °C and decreasing at 0.5。C increments to 57。C，followed by 18 additional cycles at 57 °C), then a final 1 cycle of72 °C for 7 minutes. The resulting PCR product was blunt-ended and flanked by the two primers introducing AccI sites. This 494-bp blunt-ended PCR fragment was purified and named MIII-wt (MBLIII-wild-type) or MI-wt (MBLI-wild-type). For making mutant MBLIII and mutant MBLI cDNAs, the purified blunt-ended PCR fragment was ligated into the StuI site ofpAlteMx I plasmid (Promega), generating plasmid pAher~ex I / MBLIII and pAlter-ex I / MBLI for site directed mutagenesis (Figure 3-3). The mutagenic primers, pA-MIII GLN, pA-MIII LYS, pA-MI ARG, and 49 pA-MI ASP (Table 3-3) were used to create the corresponding codon mutation at 47^ amino acid of the large subunit. The putative mutated plasmids, containing only one-strand of mutagenic sequence, were transformed into ES1301 mutS cells using ampicillin selection (125 ^g/ml). The plasmids were purified and transformed into JM 109 bacterial cells to produce double-strand DNA. The resulting plasmids, double stranded DNAs containing ampicillin resistance gene and MBLIII cDNA or MBLI cDNA with respective mutated codon, were selected by ampicillin (125 ^ig/ml). The resulting plasmids were named pAlter-ejc I / MBLIII"GLN, pAlteMJc I / MBLIII-LYS, pAlter^JC I / MBLI-ARG and pAlter-ejc I / MBLI-ASP, respectively. pBK-MBLIII or I (wild-type) MCS (Stul) 5L MBLREV、, 5 '\MBLFOR pAlter-ex I 5, 5.86 kb PCR with primers MBLFOR and MBLREV \/ Stul cut Accl Accl MBLIII or I cDNA Isolate 494 bp blunt-end pAlter-ex 丨（Stul cut) DNA fragment 5.86 kb V Blunt"6nd ligation Y 50 Accl Accl jMp pAlter-ex I/Mlll |^U 6麗 T4 DNA polymerase T4 DNA ligase \/ Transform ES1301 mut S cells Transform JM109 cells Ampicllin selection \/ Accl Accl Figure 3-3: Strategy of in vitro mutagenesis 51 To specifically confirm the presence of mutated codon in MBLIII and MBLI cDNA in pAIter-exl / MIII-GLN, MIII-LYS, MI-ARG, MI-ASP, restriction enzyme digestion was used (Table 3-5). Table 3-5: Restriction enzyme site resulting from single codon mutation of MBL III and MBL I Restriction enzyme (RE) cutting site Mutated codon Mutant: MBL I-ARG Modified DNA sequence GGAGAC RESite (BsmA I) Wild-type MBLm-wt MBL I-wt DNA sequence near mutated codon GGCGTC GG|CAA|C RESite (AcyI) (Nil) Mutant: MBL III-LYS MBL I-ASP Modified i ———T DNA sequence GG|AAG|CTATTC GG^^CT RESite (XmnI) (BstYI) 52 3,3,L4 Cloning ofthe wild-type and mutated MBLIII and MBLI cDNA into vector pTZ/ phas As shown in Figure 3-4，the wild-type MBLIII cDNA (MIII-wt) or wild- type MBLI cDNA (MI-wt) and the mutant MBLIII cDNA or the mutant MBLI cDNA contained in pAher~ex I，i.e. pAlter-ex I / MBLIII or MBLI, pAlter-ex I / MBLIII-GLN or -LYS and pAlter-ex I / MBLI-ARG or -ASP, respectively, from section 3.3.1.3 were digested with AccI and ligated to AccI~cut pTZ / phas to replace the phaseolin coding sequence. The resulting plasmids, pTZ / phas / MBLIII-wt, -GLN or -LYS, and pTZ / phas / MBLI-wt, -ARG or -ASP, were selected by ampicillin 100 (100 ngAnl). 3,3,L5 Confirmation of sequence fidelity and mutated codon in MBLIII and MBLI cDNA by DNA sequencing Sequencing was done on all the six constructs, pTZ / phas / MBLIII, and pTZ / phas / MBLI using two specific primers, pPha-1 and pPha-2 containing phaseolin promoter and terminator sequence respectively (Table 3-4)，to check the fidelity ofthe wild-type MBLIII, the wild-type MBLI, the mutated MBLIII cDNA， and the mutated MBLI cDNA sequences and their ligation junctions. The cycle sequencing reaction was performed by using ABI Prism™ dRhodamine Terminator cycle sequencing. To specifically re-confmn the presence of mutant codons in MBLIII and MBLI cDNAs, restriction enzyme digestion was used (Table 3-5). 53 Accl Accl 3kb '1 pAlter-ex 1^^^ \Accl /MBLIIIor MBLI IMIII-GLN or LYS & / MI-ARG or ASP 6.36 kb Accl Accl Wild-type MBLill or I cDNA Accl mutagenic codon Accl 1 Mutant MBLIII or I cDNA \/ Isolate 500 bp fragment Hindlll Ligation V Accl Accl Hindlll f pTZ19u Hindlll Figure 3-4: Strategy to clone wild type and mutant MBLIII and MBLI cDNAs into seed-specific expression vector pTZ / phas 54 3.3.1.6 Transfer ofwild-type MBLIII andMBLIcDNA flanked by phaseolin regulatory sequence into Agrobacterium binary vector Recombinant plasmids pTZ / phas / MBLIII-wt, -GLN and -LYS, and pTZ / phas / MBLI-wt, -ARG and -ASP from session 3.3.1.4 were digested with HindIII, and the wild-type MBLIII, wild-type MBLI, mutant MBLIII, and mutant MBLI cDNAs flanked by phaseolin regulatory sequences were excised and collected. Each of these chimeric genes was inserted into the HindIII~cut Agrobacterium binary vector, pBI121 (Clontech, USA), which carries the neomycin phosphotransferase II O^PT II) selectable marker (kanamycin resistant) and P-glucurondiase (GUS) screenable marker genes. The resulting plasmids were selected by kanamycin (100 ^ig/ml) and named pBI / phas / MBLIII-wt, -GLN and -LYS and pBI / phas / MBLI-wt, -ARG and -ASP (Figure 3-5). Accl HindIII Accl pTZ19u HindIII HindIII cut 55 Hindlll Hindlll indlll Hindlll _ Wild-tvpe MBLIII or I mutagenic Hindlll codon Hindlll _ ~")f- MBLIIIorlMutant pBI121 (Hindlll cut) ( 15kb Isolate -3kb fragment Ligation V CL Hindll Figure 3-5: Strategy to clone mutant MBLIII and MBLI chimeric genes into plant expression vector pBI121 56 3,3,L7 Transformation of Agrobacterium with pBI/phas/MBLIII andpBI/ phas /MBLI chimeric gene constructs The ligated constructs, pBI / phas / MBLIII-wt, -GLN or -LYS and pBI / phas / MBLIII-wt, -ARG or -ASP and vector pBI 121 only were transformed into Agrobacterium GV 3101 / pMP 90 by electroporation. The method used was originally developed by Sambrook et al. (1989) and modified by Lam H.M. (CUHK, Hong Kong). An aliquot (40 ^il) ofGV 3101 / pMP 90 compentent Agrobacterium cells was defrozen on ice and mixed with 1 ^il of 50 ng pBI / phas / MBLIII or pBI / phas / MBLI plasmids (session 3.3.1.6) in an eppendorf tube. The suspension was gently-mixed and incubated on ice for 1 minute. It was transferred to a pre-chilled electroporation cuvette (BioRad) carefully and the suspension was gently shaked to the bottom ofthe cuvette without introducing any bubbles. The Gene Pulser apparatus (BioRad) was set with the following conditions: 25^iF, 2.5kV, 600 ohms. The cuvette containing Agrobacterium cells-DNA mixture was applied with a pulse. After a pulse, 1 ml SOC medium (2 % Bacto- tryptone, 0.5 % Bacto-yeast extract, 10 mM NaCl, 2.5 mM KC1, 10 mM MgCl2, 10 mM MgSO4 and 20 mM glucose) was added immediately to the cuvette and the cells were quickly resuspended. The cell suspension was transferred to a 17 X 100 mm polypropylene tube and shaked at 28。C for 2 hours. Then 5 ^il, 50 ^il and the remaining cell-pellet of the culture were spreaded on LB plus rifampicin (50 ^g/ml), gentamycin (25 ^g/ml), and kanamycin (50 ^g/ml). The plates were incubated at 28� Cfor 2 to 3 days to allow transformed Agrobacterial colonies containing pBI / phas / MBLIII and pBI / phas / MBLI cDNAs to develop and be selected. 57 3,3,L8 Vacuum infiltration transformation of Arabidopsis The method used was originally reported by Bechtold et al. (1993) and modified by Lam H.M. (CUHK, Hong Kong). Arabidopsis thaliana ecotype Columbia~0 seeds were sowed onto soil and kept at 4 °C for at least 2 days as imbibition period to synchronize germination. The seeds were then allowed to germinate and grow on plastic pots in growth chamber for four to six weeks at 22 °C, with a light (daylight, 400 klux) to dark cycle of 16 hours and 8 hours, respectively. Arabidopsis plants were irrigated weekly with IX Arabidopsis fertilizer (5 mM KNO3, 2.5 mM KPO4 Q)H 5.5), 2 mM MgSO4, 2 mM Ca_3)2, 0.5 mM FeEDTA, 70 ^iM boric acid, 14 ^M MnCl2, 5 ^M CuSO4, 1 ^iM ZnSO4, 0.2 ^M NaMoO4, 10 nM NaCl and 0.01 ^M CuCl2). Plants were ready for ^gro^^7C/er/wm-mediated transformation when the primary inflorescence (flower clusters) was 5 - 15 cm tall and the secondary inflorescence were appearing at the rosette. Any siliques in the plants were cut off right before infiltration. Five hundred ml dense culture of transformed Agrobacteria GV 3101 / pMP 90 containing pBI / phas / MBLIII cDNAs or pBI / phas / MBLI cDNAs or pBI121 was prepared in YEP medium (10 g/L Bacto peptone, 10 g/L yeast extract, 5 g/L NaCl) containing rifampicin (50 mg/L), gentamycin (25 mg/L) and kanamycin (50 mg/L), at 28。C with shaking. The culture was ready for use when its OD500 was greater than 2.0. It was spun at 7,000 rpm at 4 °C for 20 minutes. The bacterial pellet was resuspended in one litre of infiltration medium [2.2 g/L MS salts (Gibco), IX B5 vitamins (0.1 g/L myoinositol, 0.01 gfL thiamine-HCl, 0.001 g/L nicotine acid and 0.001 gA^ pyridoxine-HCl), 50 g/L sucrose, 0.5 gfL MES, 0.044 ^iM benzylamino-purine and 0.02 % Silwet™ L-77, (pH 5.7)].The resuspended culture was placed in a mbbermaid container inside a vacuum desiccator. 58 The pots containing the plants were inverted and immersed into Agrobacterial suspension in infiltration medium. The entire plant including the rosettes should be fully covered inside the medium to ensure successful transformation. Vacuum was applied at 400 mmHg for 10 minutes after bubbles started to arise from the plants. The infiltrated plants were air-dried briefly for several hours and set upright by hollow plastic cylinders to enclose the whole plant. They were grown under the same conditions as described before. When most of the siliques become mature after four to six months, the plants were stopped from watering. Seeds were harvested from individual infiltrated plants after they dried out. These plants were regarded as R^ plants. 3,3,1,9 Screening of homozygous transgenic Arabidopsis One hundred ^1 of seeds (roughly 2,500 seeds) from each R^ plant was sterilized under laminar hood before plating on selection plates for screening. Undiluted chlorox (sodium hyprochlorite, 5.25 %) was added to the eppendorf tubes containing the seeds and they were vortexed for 3 minutes. After further washing with 1 ml sterile distilled water for 3 times, 1 minutes each, the sterilized seeds were plated on circular plate containing MS medium [4.3 g/L MS salts (Gibco), 1 % sucrose, 0.5 gfL MES and 0.9 % bacto-agar, (pU 5.7)] with 50 mg/L kanamycin. The plates were placed at 4 °C for 2 days and then placed in growth chamber with the growing conditions as described in section 3.3.1.8. After one week, putative transformants appeared as deep green plantlets were transferred into a 9 cm X 9 cm MS square plate (to allow longer root development) containing 50 mg/L kanamycin for fiuthergrowt h in the growth chamber. After about one to two weeks, Ri plants showing true leaves were transferred to soil and allowed to grow under alternate light and dark cycles as describe in section 3.3.1.8. hidividual 59 transformant was enclosed with a hollow plastic cylinder to ensure self- fertilization. Mature seeds of Ri plants were harvested individually for each plant for both MBLIII and MBLI constructs. Only MBLIII constructs were kept going on for later experiments and procedures. About 50 seeds were sterilized using procedures as described before. Sterilized seeds were plated on circular plates containing MS / kanamycin 50, in array order and placed in growth chamber under growing condition as described in section 3.3.1.8. The number of green transformants and yellow plantlets were counted from each plate. These figures were used for selection of Rj transgenic plants harboring transgene insertion into one chromosome only (single gene insertion). Transgenic R2 plants in deep green and with acceptable chi-square values (green : yellow = 3 : 1) were selected from these plates. They were transferred to new MS / kanamycin 50 square plates for fiuther growth. When their roots were well developed after three weeks, these plants were transferred to soils and enclosed with hollow plastic cylinders as described before. After setf-fertilization ofindividual R2 plants, their mature seeds were harvested. Finally, seeds from Rj plants were selected for homozygosity. About 50 harvested seeds were sterilized and spread on MS / kanamycin 50 square plates in array order as described before. They were placed in growth chamber under growing condition described in section 3.3.1.8. Transgenic plant lines (R3), with all their plated seeds survived on the plate and grown into deep green plantlets, were allowed to grow fiuther following procedures as mentioned previously. These homozygous R3 plants were used for MBLIII expression analysis. 60 3.3.2 Expression analysis of MBLIII transgene 3.3.2.1 GUS assay of transgenic plants GUS assay was carried out following the standard protocol ofJefferson et al. (1987). One to several pieces of leaves were cut from each R3 transgenic plant and placed into a pypropylene tube containing the GUS staining buffer [100 mM Na- phosphate buffer (pH 7.0), 0.1 % Triton X-100, 0.5 mM K-ferricyanide (K3Fe(CN)6), 0.05 % X-glucuronide (X-gluc, dissolved in DMF) and 1 mM EDTA]. After overnight incubation at 37 °C, the leaf pieces were treated with 70 % ethanol to eliminate the interference of chlorophyll. The chlorophyll-removed plants were placed in 50 % glycerol for long time storage. 3.3.2.2 Genomic DNA isolation from transgenic plants Genomic DNA was extracted from fresh Arabidopsis leaves following the CTAB protocol (Doyle et al., 1990). Fresh leaves (about 0.75 g) were ground to power in liquid nitrogen in a chilled mortar and pestle. The powder was scraped directly into 50 ml sterilized falcon tube containing 5 ml preheated CTAB extraction buffer [2 % CTAB, 1.4 M NaCl, 20 mM EDTA, 0.1 M Tris HC1 (pU 8.0)，0.2 % P-mercaptoethanol]. It was swirled gently and incubated at 60。C for 30 minutes with occasional swirling. The plant lysate was mixed gently with 5 ml chloroform-isoamyl alcohol (24 : 1，v/v) and centrifuged at 1,600 X g in a swinging bucket rotor at room temperature for 10 minutes. The upper aqueous phase was transferred to a new sterile falcon tube by using sterile wide-bore pipette tips. It was mixed with two-third volume of cold isopropanol to precipitate the nucleic acids and then centrifuged at 1,600 X g at 4 °C for 10 minutes. The supematant was discarded and pellet was washed and resuspended with 10 ml washing buffer 61 (75 % ethanol, 10 mM ammonium acetate). The suspension was centrifuged again at 1600 X g at 4 °C for 10 minutes and the supematant discarded. The pellet was dried under vacuum for 10 minutes at room temperature. The dried DNA pellet was resuspended in 50 ^1 TE buffer [10 mM Tris-HCl ^)H 8.0)，1 mM EDTA]. The resuspended pellet was added with 5 [d ofRNase A (0.1 mg/ml, Promega) and incubated at 37 °C for 30 minutes. The DNA solution was transferred to a new eppendorf tube using wide-bore pipette tips and spun for five minutes. The clear supematant was stored in a new eppendorf and its DNA concentration was determined by spectrophotometric measurement at ODjgo- The quaHty of DNA was checked by gel electrophoresis. 3,3.2,3 PCR amplification of transgene Genomic DNA from homozygous and control Arabidopsis leaves was used as the starting materials for genomic PCR. A 25 ^il PCR reaction mixture containing 2 ^ig of genomic DNA, 1.5 mM MgClj, 0.2 mM dNTPs, 1 ^M MBLFOR primer and 1 ^M MBLREV primer (Table 3-2), 0.2 mM dNTPs, IX Taq reaction buffer and 2.5 units Taq polymerase (Promega) was used to amplify the MBLIII transgene. The PCR conditions were as follows: 1 cycle of 95 °C for 10 minutes (added DNA polymerase after 7 minutes), then 30 cycles of 95 °C for 1 minute, annealing temperature for 1 minute (annealing for 7 cycles per temperature step beginning at 72 °C and decreasing stepwise at 1 °C to 65 °C, followed by 23 additional cycles at 65 °C), 72 °C for 1 minute, then a final 1 cycle of72 °C for 7 minutes. The resulting DNA product would be MBLIII cDNA sequence with or without a single codon mutation (transgene). 62 5.3.2.4 Total RNA isolation from transgenic Arabidopsis Total RNA was extracted from developing siliques using the method of Altenbach et al (1989). Fresh plant siliques were harvested and frozen immediately by liquid nitrogen. About 0.1 g of chilled siliques was ground into fine powder by a baked mortar and pestle with liquid nitrogen. The powdered tissue was added to 0.5 ml pre-heated extraction buffer [0.1 M LiCl, 0.1 M Tris-HCl Q)H 8.0)，0.1 M EDTA, 1 % SDS, haIfvolume ofphenol, prewarmed to 80 °C] and vortexed for 30 seconds.. It was mixed with 0.25 ml chloroform/isoamyl alcohol (24 : 1) and vortexed again before being centrifuged at 14k rpm for 15 minutes. The upper clear aqueous layer was then mixed with one portion of4 M LiCl and incubated at 4 °C for one hour to precipitate nucleic acids. The mixture was subjected to centrifuge at 14k rpm at 4 °C for 30 minutes and the pellet was added with 100 ^il cold absolute ethanol to remove any residual phenol, which may inhibit enzymatic activities in later experiments. It was spun for 10 minutes and the pellet was dried under vacuum centrifugation. The pellet was resuspended in 0.25 ml DEPC-treated water and then added with 25 ^il 3 M sodium acetate (pU 5.5) and two volumes of absolute ethanol for precipitation of RNA. The mixture was incubated at 20 °C for one hour. It was centrifuged at 14k rpm at 4 °C for 30 minutes and the pellet was washed with 200 |il absolute ethanol. It was again centrifuged at 14k rpm at 4 °C for 20 minutes and the RNA pellet was dried under vacuum centrifugation. The dried RNA pellet was resuspended into 15 ^il DEPC-treated HjO and then subjected for DNA digestion. For each jiig RNA, 1 unit ofDNase I in 10 ^il IX DNase I reaction buffer was added and then incubated at 25 °C for 15 minutes. The DNase activity was inhibited by 63 adding 1 ^il 25 mM EDTA and incubated at 65 °C for 10 minutes. The amount of RNA was estimated by spectrophotometric measurement at OD260. 3.3,Z5 RT-PCR oftotalRNA from transgenic Arabidopsis First strand cDNA was synthesized by reverse transcription of total seed RNA. Fifteen jjl mixture containing 0.5 ^g total RNA and 0.5 ^ig primer MBLREV (Table 3-2) in DEPC-HjO was first denatured at 65 °C for 5 minutes followed by immediate chilling on ice. The RNA/primer template was used to produce 1^ strand cDNA by adding 10 jjI reaction mixture containing 200 units Moloney murine leukemia vims (M-MLV) reverse transcriptase, IX reverse transcription buffer [50 mM Tris-HCl Q)H 8.3), 75 mM KC1, 3 mM MgClj, 10 mM DTT], 0.5 mM dNTPs, 10 mM DTT and 40 units rRNasin RNase inhibitor in DEPC-treated HjO to achieve a final 21 [d volume. The mixture was mixed gently and incubated at 42 °C for 1 hour. Another set of control reaction was carried out by using the same reagents excluding only the M-MLV reverse transcriptase to eliminate false positive results fromcontaminate d genomic DNA. For second strand cDNA PCR, reverse transcription solution containing V^ strand cDNA was amplified in a 25 ^1 PCR reaction mixture. The mixture containing 2 ^il ” strand cDNA, 1.5 mM MgCl2, 0.5 mM dNTPs, l^iM primer MBLFOR, 1 |LiM primer MBLREV (Table 3-2), IX PCR reaction buffer and 2.5 units Taq polymerase was subjected to PCR amplification using PCR conditions as described in section 3.3.1.3. 64 3.3.2.6 Verification of the presence of mutagenic site and the fidelity ofRNA transcript from transgenic Arabidopsis To check whether the mutant codon is still present in the RNA transcript from the mutant MBL transgene, after electrophoresis, the 2°^ strand PCR product from RT-PCR of total RNA was gel-purified from agarose gel. One ^ig of each gel-purified DNA was subjected to 10 pl restriction digestion reaction containing 5 to 10 units of corresponding restriction enzymes (Table 3-5)，IX reaction buffer, and l^g acetyl BSA. 3.3.2.7 Protein extraction and tricine SDS-PAGE of putative transgenic protein from Arabidopsis Total seed protein was extracted from 0.1 g mature Arabidopsis seeds by 1 ml ofprotein extraction buffer [0.25 M NaCl, 0.05 M NaPO4 (pH 7)]. The seeds were thoroughly ground by a mortar and pestle after adding small pieces ofbroken Pasteur pipette tip. Cell debris was removed by centrifugation at 14k rpm for 5 minutes and the supernatant was saved and named as seed total protein extract (STPE). The protein quantity was determined by the method of Bradford or Bicinochoninic Acid (BCA) using bovine serum albumin (BSA) as a standard. Thirty |xg protein was separated by 16.5 % tricine-SDS-PAGE (Schagger etal., 1987). The 30^xg - STPE was mixed with 5 |il of2X sample buffer (8 % SDS, 24 % glycerol, 100 mM Tris-base, bromophenol blue) with or without P_ mercaptoethanol. The mixture was incubated at 40°C for 30 minutes before loading on the tricine-geL Anode buffer [0.2 M Tris-base, (pR 8.9)] and cathode buffer [0.1 M Tris-base, 0.1 M Tricine, 0.1。/。SDS, ^>H 8.25)] were used for running the electrophoresis. The tricine-gel was subjected to staining with Coomassie Brilliant Blue Solution (lg/ L Coomassie brilliant blue G-250, 1 % methanol) and destaining solution (methanol: 100 % glacial acetic acid : water, 20 : 65 6 : 55) to visualize the protein bands. For westem blot or amino-acid sequencing experiments, the gel was left unstained. 3.3,2,8 N-terminal amino acid sequencing Purified or transgenic protein sample was loaded on 16.5 % tricine SDS- PAGE gel and electrophoresed following the procedures as described in 3.3.2.7. The protein was then electro-blotted onto PVDF membrane and the membrane was then stained with Ponceau S dye. For blotting, the PVDF membrane was soaked briefly with methanol and then washed in distilled water for 2 minutes. The wet PVDF membrane and 3 MM filter paper were pre-soaked and equilibrated in transblotting buffer (48 mM Tris, 39 mM Glycine, 0.0375 % SDS, 20 % methanol) for 20 minutes, hi the semi-dry electrophoretic transfer cell (BioRad), one piece of pre-soaked filter paper was placed onto the platinum anode and the gel placed on top of it. The gel was then covered with the pre-treated membrane and another piece ofpre-soaked filter paper placed on top. Any bubbles were excluded by rolling a clean glass pipet across the sandwich before the cathode was covered on the stack. Electro-transfer was performed at constant voltage of20 V for 1 hour at room temperature. After electro-blotting, the membrane was stained briefly (2 minutes) with Ponceau S dye [0.2 % Ponceau S (Sigma), 3 % TCA and 3 % sulfosalicyclic acid] to confirm the presence oftransferred proteins on membranes. The desired protein bands were visualised and excised by a sterile blade. The bands were rinsed with methanol for 30 seconds briefly, immersed in 10 ml distilled water and sonicated for 10 minutes. The pre-treated bands were sequenced by the HP G1005A protein sequencing system (Hewlett Packard). 66 3,3,2,9 Isoelectric precipitation of MBL One ml of STPE (〜5 mg) from section 3.3.2.7 was transferred to 15 ml Falcon tube. Distilled water was added to make up to 4 ml. The solution was then adjusted to pH 12 using 2 M NaOH. The protein precipitate was collected by centrifiigation at 6k rpm for 10 minutes and the supematant was saved in another Falcon tube. The pellet was dissolved in minimal volume of re-suspension buffer [20 mM Tris, (pH 8.9)]. Proteins precipitated at pH 11.8 and pH 9.5 were also obtained by adding appropriate volumes of 1 M HC1 to the supematant using the above procedures. These different fractions of proteins, after treatment with sample buffer containing P-mercaptoethanol, were separated by 16.5 % tricine- SDS-PAGE, following the procedures as described in section 3.3.2.7. The gel was subjected to Coomassie blue staining or left unstained for visualization or westem blot, respectively. 3.3,2.7 0 Production of polyclonal antibody against purified MBL Four hundred ^g of purified MBL II protein (Hu, Kunming Institute of Botany, China) was mixed with one portion of2X sample buffer (section 3.3.2.7) before loading onto 16.5 % tricine-SDS-PAGE gel for electrophoresis. The gel was stained with Coomassie blue and destained until the gel background turned transparent. The linear protein band and the large and small-subunit protein bands were excised separately and ground into fine granules using small plastic pestle. The gel granules were resuspended into 0.5 ml of PBS and transferred into the barrel of2ml syringe (19 gauge) containing 0.5 ml of complete Freund Adjuvant (Sigma). Fine gel granules and adjuvant were mixed thoroughly by passing the content forth and back through the needle. 67 For primary immunization, about 1.5 ml of each mixture (about 100 n,g linearized MBL II or large and small subunit polypeptides) was injected subcutaneously on the back ofeach oftwo rabbits, respectively. After 10 days, the rabbits were further immunized with a booster injection. The booster injection contained the same amount ofpurified MBL II (linearized form and large and small subunits) prepared by the same procedures except using incomplete Freund Adjuvant instead of complete one. Two more booster injections were performed every ten-day afterwards. After 1 week from the second immunization, blood samples were taken from the marginal ear vein of two rabbits to check the production of specific antibodies. One week after the fourth immunization, total blood was collected fromth e two rabbits by heart puncture and transferred gently into two sterile falcon tubes, respectively. They were kept for 1 hour at room temperature to allow blood clotting and centrifuged at 3.5k rpm for 15 minutes at room temperature. The supernatant containing serum was pipetted out. Both sera were transferred into two beakers kept on ice and solid ammonium sulfate was added to 50 % saturation. The serum was stirred occasionally to confirm complete salt solubilization and the protein was allowed to precipitate for one hour. The serum was then centrifuged at 12k rpm for 20 minutes at 4 °C and the pellet was washed with 1.75 M ammonium sulfate solution for three times. The pellet was dissolved in minimal volume of 10 mM PBS, pH 7.0 and dialysed overnight against distilled water at 4 °C. The dialysed sample was centrifuged at 15k rpm for 5 minutes and the supematant was transferred into 1.5 ml sterile tubes in 500 ^1 aliquots and kept at -20 °C for storage. The protein concentration of the supematant containing the polyclonal antibody was checked by BCA protein assay using BSA as a standard. 68 3,3,2.11 Western-blotting and immunodectection of Arabidopsisprotein by anti-MBL polyclonal antibody Westem blot was performed using the polyclonal antibody raised against purified MBL II protein. About 30 ^ig of transgenic Arabidopsis protein (STPE) was separated by 16.5 % tricine SDS-PAGE and blotted onto PVDF membrane as desired in session 3.3.2.8. After electrophoresis of the protein sample treated with or without p-mercaptoethanol, the gel containing the separated proteins was blotted with the pre-treated PVDF membrane by a semi-dry electrophoretic transfer cell (BioRad). After electroblotting, the membrane was steined briefly with Ponceau S dye to confmn the presence of the transferred proteins on membrane. The membrane was then subjected to immunodetection by the modified procedures of AURORA Westem Blot Chemiluminescent Detection System (ICN). The membranes were first rinsed briefly by distilled water briefly for only few seconds and then incubated in fixation buffer (40 % methanol, 10 % acetic acid) by very gentle agitation for 5 minutes. The blocking step was eliminated in this modified protocol. The membrane was completely dried for 1 hour at 37 ®C. The membrane was then incubated in freshlyprepare d blocking buffer (1 % BSA, IX PBS, 0.05 % Tween-20) containing appropriate titer (1: 5,000) anti-MBL polyclonal primary antibody, for 1 hour at room temperature. Unbound primary antibody was removed by rinsing the membrane with IX PBS for 10 seconds (2 times). Anti-rabbit secondary antibody-alkaline phosphatase conjugate in blocking buffer (1 : 8,000) was added to prime where primary antibody located, for 30 minutes incubation with gentle agitation. Similarly the unbound secondary antibody was removed by rinsing the membrane with IX PBS for 10 seconds (2 times). The membrane was then washed with at least 50 ml of IX Assay buffer [20 mM Tris-HCl (pU 9.8)，1 mM MgCl2 in distilled water, 4 °C] (2 times) and incubated in chemiluminescent substrate solution for 5 minutes. Excess substrate solution was drained from the 69 membrane and the membrane subjected to fQm exposure and fihn development 70 3.4 Results & Discussion 3.4.1 Site-specific mutations ofArginine residue in MBLIII cDNA and Glutamine in MBLI cDNA 3,4,1.1 Megaprimer PCR Site-specific mutations ofarginine (Table 2-1) at the 47^ amino acid in the large-subunit of MBL III and Glutamine in MBL I were performed by the ‘megaprimer，PCR method (Sarker and Sommer, 1990). hi this method, only three oligonucleotide primers and two rounds of PCR were needed to obtain a single codon mutation within the amplified DNA fragment (Figure 3-1). This was assumed to be the simplest, fastest and most versatile means to generate site- directed mutation. The first round PCR was performed using the mutant primer (MIII-GLN or -LYS and MI-ARG or -ASP) (Table 3-2) and one ofthe flanking primers (MBLFOR). The flanking primer, containing an Acc I restriction cutting site at 5，end, primed at the start codon of MBLIII cDNA and NBLI cDNA. The mutant primer, containing the desired base change, primed the DNA sequence flanking arginine and glutamine codon within the wild type MBLIII and MBLI cDNA targeted for mutation, respectively. After the 1^ PCR, the products were analysed by gel electrophoresis (Figure 3-6). Two major DNA bands, with molecular weight of about 400-bp and 100-bp, were visualized (lanes 2, 3 & 5). The 100-bp band is likely the primer- dimer, while the 400-bp band has a size ofthe predicted MBL amplified product. The double-stranded PCR product (418 bp) was gel-purified and used as one ofthe primers, megaprimer, in the second round of PCR along with the other flanking primer (MBLREV). Since the gel-purified megaprimer was so long when compared to the usual primers, its annealing temperature was very difficult to estimate and was too high for another flanking primer (MBLREV). Therefore, ‘touchdown，PCR (Hecker and Roux, 1996) was used to amplify the 2°^ PCR product. Instead of guessing an appropriate annealing temperature ofthe megaprimer, a wider net oftemperatures was cast by using progressively lower annealing temperatures over consecutive cycles. The goal is to select a broad range of annealing temperatures that begins above the estimated Tm and ends at the annealing temperature ofthe other flanking primer. Since the yield of the final mutant PCR product is usually low (Ling and Robinson, 1995), reamplification of mutant product was performed using two flanking primers (MBLFOR and MBLREV). The annealing temperature used in the PCR reamplification was higher than that in the \^ PCR since both flanking primers should be completely complementary to the DNA sequence flanking MBLIII cDNA and MBLI cDNA in the 2°^ PCR product. This higher annealing temperature could also avoid the amplification of contaminated DNA or non- specific sequence (Figure 3-7). After the 2°^ PCR, the 500-bp band was visualized and has a desired size ofthe MBL cDNA (lanes 2, 3 & 5). The 'megaprimer' method was adopted to perform site-directed mutagenesis since it is a time-saving strategy, requires only three primers and two PCRs and seems to introduce less additional spurious mutations. One may suggest that the 2°^ PCR can be made simpler if another flanking primer MBLREV were used instead of MBLFOR in the 1就 round PCR. However, if MBLREV and the mutant primer pair weree used, the product obtained in the 1对 PCR would be indistinguishable by size from the dimer-primer of the PCR . Moreover, it is claimed to be more efficient to gel-purify a 400-bp DNA fragment than a 100-bp one by using the commercial kit. The fmal mutant PCR products were MBLIII and MBLI cDNAs flankedb y two AccI sites containing a single amino-acid mutation. For MBLIII PCR product, 72 it was either R47Q, in which Arg 47 in the large subunit was replaced by Ghi or R47K, in which Arg 47 was replaced by Lys. For MBLIPCR product, it was either Q47R, in which Ghi 47 was replaced by Arg or Q47D, in which Gbi 47 was replaced by Asp. For these arginine and glutamine substitutions in MBLIII and I， respectively, the CGT codon in wild-type MBLIII cDNA was replaced by CAA and by AAG in the mutant primers, and the CAA codon in wild-type MBLI cDNA was replaced by AGA and by GAT. The usage frequencies of these codons are found to be optimal in Arabidopsis thaliana. 1 2 3 4 5 -418bp 100- Figure 3-6: First PCR products of megaprimer PCR the products from the 1^ PCR were separated on a 1.2 % agarose gel. Lanes 1, 100 BP DNA marker (Gibco); 2，3 and 5 contained 1，5, 19 |il of PCR product respectively; 4，low MW mass ladder (Gibco) (from top to bottom bands, with mass of 200, 160，80，40，20 and 10 ^g of DNA, respectively). 73 2 3 4 5 〜494 bp Figure 3-7: Second PCR products of megaprimer PCR The products from the 2"d PCR were separated on a 1.2 % agarose gel. Lanes 1，100 bp DNA marker (Gibco); 2, 3 and 5 contained 1，5, 19 |il of PCR product respectively; 4，low MW mass ladder (Gibco) (from top to bottom bands, with mass of 200, 160, 80，40, 20 and 10 ^ig of DNA, respectively). 3,4,L2 Cloning into the seed-specific expression vector pD38 Using megaprimer-PCR, the 494 bp MBLIII and MBLI coding sequence containing a single-amino acid mutation was obtained. The added-on AccI ends in the PCR-amplifled mutant MBLIII cDNAs and MBLI cDNAs allow the insertion of the sequence into pD38 (Figure 3-2), a seed-specific expression vector. The vector pD38 was chosen for the expression of the MBLIII and MBLI coding sequences because it contains the very strong seed-specific promoter from phaseolin, the major seed storage protein gene in French bean, which was expected to result in a high level expression of the mabinlin protein in the seeds of Arabidopsis thaliana. Another advantage of using this vector is that it contains 2 incompatible AccI sites flanking the phaseolin coding sequence. This would prevent self-ligation ofthe AccI-cut-pD38 and the MBLIII and MBLI cDNAs from cloning into pD38 in a reverse orientation. 74 To perform the construction, the pD38 was digested with AccI and then resolved by agarose gel electrophoresis (Figure 3-8A). Two DNA bands, with molecular weight of 5.3 and 1.7 kb, were observed (lanes 2 to 5). The DNA fragment containing the phaseolin regulatory sequence (-5.3 kb) was separated from the coding sequence and then gel-purified (Figure 3-8B, lane 1). However, the mutant MBLIII and MBLI cDNAs purified from the megaprimer PCR proved to be found very difficult to clone into the phaseolin regulatory sequences. The efficiency ofAccI digestion ofthe ends of PCR products was very low (0 % even after 24 hours incubation), bistead ofdirect restriction digestion, the PCR products were first cloned into the T-overhang vector (Marchuk et al, 1990; Hadjeb and Berkowitz, 1996). The template-independent terminal transferase activity ofDNA polymerase (e.g. Tfl polymerase in this case) usually causes the addition ofasingle adenosine at the 3，end of PCR-amplified fragment.Th e T-overhang vector, the blunt-end-cut plasmid pBluescript with a thymidine addition at the 3，end of both strands, can be cohesively ligated to the A-overhang PCR products. However, no successM transformants were obtained from cloning all the 2°^ PCR products of mutant MBLIII and MBLI into the T-vectors. The difficulties in cloning the mutant MBLIII and MBLI cDNAs into vector pD38 might be due to various factors. The problems could be in the 2"^ PCR products, the vectors in use, or the experimental process. Due to time constraints, another method, in vitro site-directed mutagenesis was tried in the construction of the mutant MBLIII and MBLI cDNAs. 75 2 3 4 5 Figure 3-8: GeI purification of phaseoIin regulatory sequences A: Seed-specific expression vector DNA pD38 was digested with AccI and the DNA fragments were separated by 1 % agarose gel electrophoresis. Lanes 1， 1 kb DNA marker (Gibco); 2, 3 & 4’ 2 ^ig ofpD38 were digested with AccI; 5, 0.2 ^g ofpD38 were digested with AccI. B: Lanes 1，the 5.3 kb fragment without the phaseolin coding sequence was gel purified; 2, 1 kb DNA marker (Gibco). 3,4,L3 In vitro site-directed mutagenesis Site-directed mutagenesis was performed using the kit 'Altered sites II in vitro Mutagenesis Systems' (Promega) (Figure 3-3). After the blunt-end PCR, 3 major DNA bands, with molecular weights of about 500 bp, 1.5 and 2.0 kb, were obtained (Figure 3-9, lanes 1 to 4). The 1.5 and 2.0 kb bands were the non-specific PCR products while the 500 bp band has the desired size ofthe MBL cDNA. The 500 bp-blunt-end PCR products, including the wild-type MBLIII and the wild-type MBLI cDNAs, flanked by AccI sites were gel purified and cloned in the mutagenic 76 vector pAlter-ex I. The successfiil clones were fast-screened by a method called ‘cracking，(Figure 3-10). The double-stranded plasmid was denatured and then hybridized with a mutagenic oligonucleotide, which is complementary to the corresponding DNA sequence but containing a mutated codon (Table 3-3). The ampicillin repair oligonucleotide was annealed to the single-stranded template at the same timet o let DNA polymerase extend and link the two oligonucleotides. The nick was then sealed with DNA ligase and the mutagenesis reaction containing the duplex structure was transformed into a repair minus strain ofE coli (ES 1301 mut S) to avoid selection against the desired mutation. The plasmids were purified from ES1301 mut S strain and transformed into JM109 to ensure proper segregation of mutant and wild-type plasmids. Only the mutant plasmids containing the ampicillin resistant gene can survive upon antibiotic selection and this selection results in a high yield ofmutants. �50 b0p Figure 3-9: Gel electrophoresis of blunt-ended PCR products of MBLIII and MBLI cDNAs The products from blunt-ended PCR were separated by 1 % agarose gel. Lanes 1，2 & 3，10 ^1 ofPCR product; 4, 1 ^il of PCR product; 5, low MW mass ladder (Gibco) (from top to bottom bands, with mass of 200, 160, 80，40 and 20 ^ig ofDNA, respectively); 6，1 kb DNA marker (Gibco). 77 Band shift Figure 3-10: Fast-screening of recombinant clones by "cracking" procedure After cloning with the pAlter-ex I，the successful clones were screened by gel electrophoresis using the "cracking" method. Bands with larger size contained the blunt-end insert. The double-strand plasmids, pAlter-ex I / MBLIII-GLN, / MBLIII-LYS, / MBLI-ARG and / MBLI-ASP, were purified from JM109 cells and expected to contain the same single amino-acid mutation as described in section 3.4.1.1. The mutations were preliminarily confirmed by restriction enzyme digestion. The mutant MBLIII-GLN with amino acid mutation R47Q (Table 2-1) would result in a deletion ofAcyI restriction site (Table 3-5) fromwild-typ e MBLIII cDNA (Figure 3-llA). Another mutant, MBLIII-LYS, with amino acid mutation R47K, would result in a XmnI and in a deletion of AcyI restriction site (Figure 3-llB). For MBLI constructs, the mutation Q47R would result in a BsmAI restriction site (Figure 3-1 lC) while mutation Q47D result in a BstYI site (Figure 3-llD). 78 MBLIII- MBLIII- MBLIII- MBLIII- wt GLN wt LYS A: Screening for MBLIII-GLN mutant. B: Screening forMBLIII-LY S mutant. The mutant plasmids were digested by The mutant plasmids were digested by AcyI. The disappearance of AcyI (arrow) XmnI. The addition ofXmnI (arrows) resulted in a larger fragment(1.2 7 kb) resulted in two smaller fragments(2. 5 & from wild-type 1.1 kb and 170 bp 3.2 kb) from wild-type 5.7 kb MBLI- MBLI- wt ASP C: Screening for MBLI-ARG mutant. D: Screening forMBLI-AS P mutant. The mutant plasmids were digested by The mutant plasmids were digested by BsmAI. The addition ofBsmAI (arrows) BstYI. The addition ofBstYI (arrows) resulted in two smaller fragments (1.25 kb resulted in two smaller fragments (1.25 & & 200 bp) from wild-type 1.45 kb 1.4 kb) from wild-type 2.65 kb Figure 3-11: Screening for MBL mutants After mutagenesis, the plasmid DNAs were digested by appropriate RE and analysed by gel electrophoresis. 79 3.4.2 Construction of plant expression vectors containing chimeric MBLIII and MBLI 3.4.2.1 Cloning of MBLIII and MBLI cDNAs into the seed-specific expression vector pTZ/phas The wild-type MBLIII and I cDNAs as well as the mutant MBLIII cDNA containing a single amino acid mutation were excised from the mutagenic plasmids. pAlter-ex I / MBLIII and / MBLI as well as pAlter-ex I / MBLIII-GLN, pAiter-ex I / MBLIII-LYS, pAlter-ex I / MBLI-ARG and pAlter-ex I / MBL-ASP, respectively, using AccI digestion (Figure 3-4). No difficulties were encountered in AccI digestion of mutagenic plasmids as compared to AccI digestion of mutagenic PCR products (session 3.4.1.1). The plasmid pTZ / phas, a seed-specific expression vector using phaseolin regulatory sequences, was prepared from a similar plasmid pD38, used in section 3.4.1.2. The only difference is the addition of a HindIII site 5’ upstream of the phaseolin promotor in pTZ / phas to facilitate insertion of target gene into the plant-expression vector. This vector was used in the current study. The wild-type and mutant MBLIII cDNAs and MBL cDNAs (6 constructs) were cloned into the AccI site ofpTZ / phas by replacement ofthe phaseolin coding region (Figure 3-5). The ligated plasmids of pTZ / phas / MBLIII-wt, -GLN and MBLIII-LYS, as well as pTZ / phas / MBLI-wt, -ARG and -ASP, were checked for the presence ofdesired single amino acid mutation by automated DNA sequencing (Figure 3-12). Moreover, the fidelity of the cloned cDNA sequence and the ligationjunctions (AccI sites) was also verified by DNA sequencing. 80 Q Q Q • C Q Q Q C Q Reverse sequence 5' ACG 3' Forward sequence 3’ •〒〜5' il lCiC MMMA II MBLIIIwt original codon dL ^A Ma u 1 mJti, 0 a Q Q C Q ‘ Q Reverse sequence 5’ TJG 3， Forward sequence 3，4 5，lj AAC il MBL III GLN mutant codon l |n ft cL liL/vi^ HiiiHI Q Q ： Q Reverse sequence 5，Qjj 3， Forward sequence 3，^^ 5， GAA MBL III LYS ‘il mutant codon ,L ： • I ... i : , , ： I ; ‘ ;S k: I ！ ‘…：：； :' ；.;|j ！ ll, , i j •； ； I lj ; ! ： j £M iiiWJi!ljJiM A: DNA sequencing results of MBL III 81 8-^^AAAT „ CACCT Q QCGT G Q ^¾" . 170 Reverse sequence 5'TTG3' Forward sequence 3 <~5， AAC MBLIwt original codon 1 A A A T A ^~ CT| C C T C A C C T 0 Q Reverse sequence 5’ jQj 3, Forward sequence 3: 5， AGA MBL丨 ARG mutant codon G Q CQ Q CT CT A A AT jQ AT|CCCT C ACC T Q 0. I/O Reverse sequence 5，Qpj 3， Forward sequence 3'4 5， TAG MBLI ASP mutant codon i™ i I B: DNA sequencing results of MBL I Figure 3-12: Verification of the mutant codons The codons (47^ amino acid of the large subunit) of wild-type and mutated MBL cDNAs cloned in seed-specific expression vector pTZ / phas were checked by automated DNA sequencing. Boxes indicate reverse sequence of the wild-type and mutant codons analysed by automated DNA sequencing machine. 82 3,4.2,2 Cloning into the plant expression vector pBI121 The DNA fragments flanked by HindIII site, containing the MBLIII cDNAs and MBLI cDNAs flanked by the phaseolin regulatory sequences (〜2.9 kb), were excised from pTZ / phas / MBLIII and MBLI (Figure 3-6). The 2.9 kb fragment was inserted into the plant expression vector pBI121 (-15 kb) (Figure 3- 13). These plant expression vectors (also called Agrobacterium binary vectors) containing the desired cDNAs were transformed into Agrobacterium and then into Arabidopsis. During vaccum infiltration, the Agrobacterium bound to the plant cells and the DNA flanked by T-DNA left and rightborder s were transferred into plant cells and then into the plant genome. 83 v^ Gel purification, ligation and transformation pBI I phas IMBL ~18kb 12kb Figure 3-13: Cloning of MBLIII and I cDNAs, flanked by phaseolin regulatory sequences, into plant expression vector Both the seed - specific expression vector containing MBL (pTZ / phas / MIII & / MI) and the plant expression vector pBI121 were digested with HindIII and separated by electrophoresis on 0.8 % agarose gels. The 2.9 kb fragment containing phaseolin regulatory sequences and MBL was purified and ligatedtopBI121. 3.4.3 Generation of homozygous transgenic Arabidopsis 3,4,3,1 Screening oftransgenic Rj Arabidopsis Upon vacuum infiltration, Arabidopsis (¾) plants, numbered la to ld, 2a to 2d，3a to 3d, 4a to 4d, take up the MBL chimeric transgenes or the plant- expression vector pBI121 (control) from Agrobacterium, by an unknown mechanism. For the chimeric transgenes, the MBLIII cDNA with the phaseolin 84 regulatory sequences flanked by the left and right borders of the T-DNA should integrate into plant genome and some should integrate into reproductive tissues to generate transgenic progenies. The seeds were harvested and selected on kanamycin selective medium. As the selectable marker fNPT II) confers a dominant phenotype of kanamycin resistance, and this new phenotype is only added to transformed plants but not to untransforaied ones. Therefore, only those seeds containing the transgene will germinate and develop into green plants (R!) (Figure 3-14). Those without the transgene would germinate but die eventually. Figure 3-14: Selection of Ri transgenic Arabidopsis Sterilized pBI / phas / MBLIII transformed Rj seeds were plated on MS / kan plate. After 2 days at 4 ®C, the seeds were incubated in a growth chamber at 22 ®C for about 2 weeks. Since independent vacuum infiltration was done for individual pots (numbered 1 to 4) containing 4 plants each (numbered a to d), respectively, transformation conditions were expected to be the same for each pot. Therefore deep green plants numbered with different first two-digits were selected for further 85 propagation into Rj. This is to enrich the collection of independent transformants with different insertion locations and expression levels ofthe transgenes. 3,4,3.2 Screening of transgenic R2plants Most transgenic Rj plants were expected to be hemizygous for the transgene (only one transgene integrated into a particular locus), although multiple insertions may occur. After self-fertilisation of these hemizygous R! plants (numbered la-l, la-2, la-3, la-4, 2a-l, etc), both homozygous and heterozygous R2 progenies would appear. Genetic analysis was used to identify progeny in which a transformation event happened in one locus only. According to Mendelian law, for monohybrid inheritance, the antibiotic resistant phenotype would appear in three-forths of the Rj progenies. Therefore，by chi-square analysis, deep green to pale yellow plantlets would appear in 3 to 1 ratio by growing Rj seeds under selective medium. T-test was used to check and validate those Rj plants showing acceptable ratio and the deep-green plantlets from those validated plants were allowed to propagate. Table 3-6 showed the summary of the chi-square analysis for the selection 0fR2 plants, including control pBI121, MBLIII wild-type, MBLIII Gki mutant and MBLIII Lys mutant constructs, showing single gene insertion. 86 Table 3-6: Chi-square analysis 0fR2 transgenic plants Plants Green: yellow Chi-square value VorX plantlets pBI:la-4 78:31 0.79 ~T~ pBI:lb-l 86:28 0 V pBI:lc-5 66: 10 9.33 X MIII wt:la-l 40:28 ^ X MIII wt:lb-2 52:6 13.42 X MIII wt:lc-3 120 : 28 2.91 ~T — MIIIGhi:la-l 48:20 ^ ~V~ MIIIGhi:lb-2 63:27 0.93 -…「- MIIIGhi:lb-7 141 : 79 16.98 X MIII Ghi:lc-8 72:27 0.21 V MIIIGhi:lc-10 82:33 0.74 V MIIIGhi:ld-14 58:23 0.60 V MIIILys:la-8 42:9 L^ ~V""" MIIILys:ld-3 65:22 0 V Note: Tranformants that can pass the assumption of "survival number : death number = 3 : 1” in the R2 progenies is indicated by a “V ” while those fail is indicated by an "X". 87 3.4,3.3 Screening ofhomozygous R^ transgenic plants The deep-green Rj plants with acceptable chi-square values were selected to propagate and self-fertilise to produce seeds. These deep-green Rj transformants were either heterozygous or homozygous. According to Mendelian law, those heterozygous Rj plants would produce green and yellow offsprings (R3) in 3 to 1 ratio after self-fertilisation. These plants therefore require an additional round of homozygous selection. Those homozygous Rj would produce all green R3 offsprings and they could further propagate and self-cross to obtain pure homozygous lines (R4). These homozygous R3 plants showing all green plantlets are indicated in Table 3-7. Table 3-7: Screening 0fR3 plants with transgene homozygosity Constructs pBI121 Mmwt MfflGln MHILyi R3 plants with 1a^4-1 ic-3-12 1a-1-2 1cM-2 all green 1a^-7 1c-3-13 1a-1-5 1d"44 plantlets 1b-1-3 1c-3-14 1b-2-2 1cM-5 1c-10-2 1cM"6 1cM-7 1cM-8 1cM-10 88 3.4.4 Detection of MBLIII transgene in Arabidopsis 3,4.4,1 Gus Assay The homozygous R3 transformants of the four constructs, i.e. pBI121 control, MBLIII wild-type, MBLIII-GLN mutant, MBLIII-Lys mutant, were assayed for GUS gene expression. The GUS gene can act as a reporter gene to verify the intergration of the reporter gene, hi the assay, an active GUS marker transgene will yield a deep-blue colored product in the transgenic tissue. Although GUS gene expression did not necessarily indicate that the MBL transgene was also expressed, GUS assay does indicate the occurrence of a transformation event and provides indirect evidence ofthe presence oftransgene. As shown in Figure 3-15, leaves oftransgenic^ra^/V/op5/5 turned deep blue while leaves ofnative Arabidopsis (untransformed) remained colorless upon GUS staining and chlorophyll removal. Figure 3-15: GUS staining of native and transgenic Arabidopsis Leaves were placed in GUS assay solution and incubated at 37^C. After ovemight incubation, 70 % ethanol was added to eliminate interference by chlorophyll. The leaves on the left of both photos were from native Arabidopsis plants while those on the right were from transgenic Arabidopsis plants. 89 3,4,4.2 Detection oftransgene integration Besides using the dominant kanamycin-resistant phenotype and GUS assay to indicate the presence of a transgene, a successfiil transformant should show stable transgene integration into the plant genome. This can be demonstrated by PCR on the transgene from total plant genomic DNA. Genomic DNA was thus isolated fromnativ e and transgenic Arabidopsis leaves by the CTAB method. The DNA samples were analysed by agarose gel electrophoresis (Figure 3-16). The uncut genomic DNA appeared as one major band with a relatively high molecular weight (over 15kb) while the HindIII-cut genomic DNA showed a smear ofvarious sizes ofDNA fragments. Genomic PCR was performed using genomic DNA as a template with a pair of specific primers, MBLFOR and MBLREV (Table 3-2). If the transgene was integrated into the plant genome, PCR should result in a DNA band with a molecular weight of500 bp, the expected size ofthe MBL insert, hi Figure 3-17， the genomic DNA ofMBLIII wild-type, MBLIII Gb and MBLIII Lys mutants all showed a dominant band of 500 bp in length. No such bands occurred in the lanes of native Arabidopsis or in the transformant containing only the plasmid pBI121. These results indicated the stable integration of the MBLIII transgene into the plant genome. 90 2 3 5 6 uncut 12.0 DNA Figure 3-16: Gel electrophoresis of genomic DNA from transgenic Arabidopsis >eaves Leaf genomic DNA samples, with or without HindIII digestion, were separated by 1 % agarose gel electrophoresis. Lanes 1， 3，5 contained uncut genomic DNA of MIII-wt, MIII-Ghi mutant, and MIII-Lys mutant respectively. Lanes 2, 4，6 represent DNA cut with HindIII, respectively. Lane 7 contained 100 bp DNA marker. 2 3 5 6 7 500 bp 100 Figure 3-17: PCR analysis of genomic DNA from native and transgenic Arabidopsis leaves PCR was performed using genomic DNA from native, pBI121control, and putative transgenic plant leaves and a pair of MBL specific primers MBLFOR and MBLREV. The PCR products were analysed by gel electrophoresis. Lanes 1: Native Arabidopsis, 2: control plasmid pBI121, 3: MIII-wt, 4: MIII-Gln, 5: MIII-Lys, 6: -ve control, genomic DNA replaced with dHjO. Lane 7 contained 100 bp DNA marker. 91 3.4.5 Detection of MBLIII transcript in transgenic Arabidopsis 3,4.5.1 RT-PCR (Reverse-transcription polymerase chain reaction) Since the transgene MBLIII was under the control of seed-specific regulatory sequences, mRNA of developing green siliques should contain the transcript ofMBLIII transgene ifthe gene is expressed. Total RNA was isolated from developing green siliques for analysis. The quality ofthe total RNA samples is shown in Figure 3-18. Two major RNA species, 28s and 18s ribosomalRNA , were observed. DNase I digestion on total RNA was performed to remove genomic DNA contamination which may give false positive results in second strand cDNA PCR. The overall RNA quality was sufficient for RT-PCR analysis although some apparent degradation was seen. RT-PCR was performed using total silique RNA as template and two specific primers, MBLFOR and MBLREV (Table 3-2). DNA bands ofabout 500 bp, the expeted size ofMBLIII product, appeared in the lanes ofMBLIII wild-type, MBLIII Gb and MBLIII Lys. The lanes containing the PCR reaction mixture with dHjO instead of RNA as template, native Arabidopsis total RNA, and RNA from transforaiant containing the control plasmid pBI121 did not show the expected band (Figure 3-19). These results established that MBLIII RNA transcripts are present in siliques of MBLIII transgenic constructs only. Negative controls were performed with all samples, with reverse transcriptase omitted in the first strand cDNA reaction mixture, to demonstrate that the MBLIII-specific PCR product was not due to genomic DNA contamination. The presence of MBLIII primers- and size-specific PCR product indicated that the transgene was transcribed and expressed as mRNA. 92 28s rRNA 18srRNA Figure 3-18: Gei electrophoresis of total RNA isolated from maturing native and transgenic Arahidopsis siIiques Silique total RNA samples were separated by 1 % agarose gel electrophoresis. Lanes 1，n?i^\tArabidopsis-, 2, pBI121; 3，MIII wt; 4，MIII Ghi mutant; 5， MIII Lys mutant. 2 3 4 5 6 7 8 9 10 11 12 500 bp Figure 3-19: RT-PCR of total RNA from maturing native and transgenic Arabidopsis siIiques Total RNA isolated from maturing siliques of native, pBI121control, and putative transgenic homozygous plants was used to prepare cDNA by the RT-PCR procedure. MBL-specific primers, MBLFOR and MBLREV (Table 3-2)，were used in the amplification. Lanes 1 & 2, -ve control; 3 & 4 native Arabidopsis; 5 & 6, pBI121; 7 & 8，MIII wt; 9 & 10，MIII Gb mutant; 11 & 1¾ MIII Lys mutant; odd lanes: with reverse transcriptase; even lanes: without reverse transcriptase. 93 3.4.5.2 Verification of the presence of the mutant codon and sequence fidelity ofthe RT-PCR product To check whether the mutant codon was still present in RNA transcript, restriction digestion was performed on the PCR product after second strand RT-PCR. The finalRT-PC R products (-500 bp band) of MBLIII wild-type, MBLIII Ghi, and MBLIII Lys were excised from the gel and purified. The purified DNA was subjected to restriction enzyme digestion, using AcyI and XmnI (Table 3-8). At the site of mutation (407 to 410 base pairs in PCR product), restriction sites AcyI and XmnI are unique to MBLIII wild-type and MBLIII Lys mutant, respectively. The expected fragment sizes upon these restriction digestions are shown in Table 3-8. The results ofthese experiments are shown in Figure 3-20. All the three MBLIII transgenic RNA transcripts were confirmed to contain the corresponding restriction enzyme sites, hence the presence ofthe expected mutant codon. The fidelity of the RNA transcripts was further confirmed by DNA sequencing. Since Taq polymerase possesses a template-independent polymerization activity that adds a single dATP to the 3’ ends of the DNA template, the second strand cDNA PCR products were cloned into T vector containing a complementary 3'-terminal thymidine overhang. The fidelityo f the cloned second strand cDNA and the mutagenic codon was checked by DNA sequencing. The total agreement ofthe DNA sequencing data with the MBLIII sequence and mutant codon ftoher indicates that the transgenes were successfully transcribed into mK^K 'mArabidopsis (Figure 3-21). 94 Although RT-PCR analysis may not be suitable to quantify the RNA expression and northern blot analysis will be a better procedure for quantification, it is extremely difficult, if not impossible, to tag the very small flowers of Arabidopsis for discrete developmental stages. Therefore, green siliques were collected without precise day-after-flowering information to perform the RNA analysis. Table 3-8: Restriction maps of RT-PCR products of the three MBLIII transgenes Restriction site AcyI digestion XmnI digestion -• • ^ • AcyI AcyI lll-Wt PCR ^ 1 238 + 169 + 87 bp undigested product 238 407 494 bp lll-Gln PCR AcyI 238 + 256 bp 1 undigested product 494 bp ^ lll AcyI XmnI -Lys PCR - I h- 238 + 256 bp 407 + 87 bp product 238 407 b 5 b S5 3oooo" oooo p )ooooo ooo 43 43 o o 2 21 o o 1 B Figure 3-20: Verification of the presence of the mutated codon sequence in transgenic RNA transcripts The RT-PCR products was gel- purified and subjected to RE digestions. A: Acy I digestion. Lanes 95 1，Mlll-wt; 2, MIII-Gln; 3，MIII-Lys and 4，100 bp DNA marker. B: Xmn I digestion. Lanes 1 to 4 were the same as in A. xr A GCG G T CrCJA^- Acj^:rr C.^CCI Q CCGT G 丨 140 15； 160 170 Reverse sequence 5, ACG 3 ’ Forward sequence 3'^ 5， r TGC MBLIIIwt U 1 liJI i,i! | original codon _ I II f II _l I 擺 _ Mi ilML :GGCTcr G ArtT kyyT^cv CAcn o GCor G c 15： 160 170 Reverse sequence 5，jjQ 3， Forward sequence 3，4 5, AAC iLU.I MBL III GLN mutant codon .^M (i\ .- '4 fiM riGiT;GCm7r GAAMT A^^C C f，CACi aCGGCGTG 150 lfiO 170 Reverse sequence 5，Qjj 3， Forward sequence 3，4 5' GAA MBL III LYS mutant codon k HI Figure 3-21: Verification of the mutant codons in RNA transcripts The RT-PCR products were gel-purified and cloned into T-vector for DNA sequencing. The codons (47^ amino acid ofthe large subunit) of wild-type and mutated MBL cDNAs were checked. Boxes indicate reverse sequence of the wild-type and mutant codons analysed by automated DNA sequencing machine. 96 3.4,6 Detection ofMBL III protein in transgenic Arabidopsis 3,4.6.1 Expression ofMBL protein Transgenic Arabidopsis plants were analysed for their accumulation of MBL. Total seed protein was separated by 16.5 % tricine-SDS-PAGE. As shown in Figure 3-22, the overall protein profiles for the native Arabidopsis control plant and the transgenic ones were quite similar except the difference in band intensity. A protein band of size between 15 and 17 kDa also showed different intensity even among different transgenic plants and different constructs. This band might contain the MBL protein although it was not the exact size as reported in previously (Hu and He, 1983). Large amount of protein accumulation leads to a shift of this protein band The protein extracts thus were treated with P-mercaptoethanol to fmd out if the suspected protein band can be denatured into two subunits as in MBL. To further identify the suspected protein bands，total seed protein was denatured by treatment with P-mercaptoethanol and was then analysed by 16.5 % tricine-SDS-PAGE. 97 2 3 5 6 k64322a _2 7 8 9 10 11 656940 •»>» S5 fi5 29 kDa 4 2 te«i mmmmrnt^iSk kDa mmmm 17 14.4 Figure 3-22: Protein profiles of native and transgenic Arahidopsis total seed protein extracts Total protein extracts from native and transgenic seeds were prepared and separated by 16.5 % tricine-SDS-PAGE in the absence of P-mercaptoethanol. Lanes 1: Native Arabidopsis - control; lanes 2 to 10: transgenic Arabidopsis, M III-wt (lanes 2-4), M III-Gln (lanes 5-7)，M III-Lys (lanes 8-10); and lane 11: purified MBL) As shown in Figure 3-23，all major protein bands with size between 14 to 17 kDa in Figure 3-22 disappeared and 2 new protein bands with lower moleculaQ r weight, 8 and 6 kDa, appeared The protein was electroblotted to polyvinylidene difluoride (PVDF) membrane (Figure 3-24). The 8 kDa protein bands were excised and subjected to N-terminal protein sequencing. Table 3-9 summarized the sequencing results. For all transgenic plants, the sequences of the N-terminal 10 to 14 amino acids matched exactly with that of the large subunit of 2S albumins isoform 3 in Arabidopsis thaliana (AT2S3) (Swiss-Prot, accession P15459). As the N-terminal sequences of AT2S3 and MBL are different (Figure 1-5), the sequence obtained is thus that of the Arabidopsis 2S protein (Table 3-9). 98 2 3 4 5 6 7 ^ ^i^a ^ MM. » W,^ •~~- rnmmm mmm I 1 11 - I Figure 3-23: Tricine-SDS-PAGE of native and transgenic Arabidopsis total seed protein extracts treated with P-mercaptoethanol Total protein extracts were treated with dissociation buffer containing p_ mercaptoethanol before electrophoresis. Lanes 1: Native Arabidopsis control; 2-3: M III-wt; 4-5: M III-GLN; 6-7: M III-LYS) 2 3 5 6 +t ^¾ 、々 【 等」^"^ i \ < ,|^»， 1" ,'is^ <»〜 giU^^^m^):�‘ ‘• ‘ 'iS .〜 'i Figure 3-24: PVDF membrane with transferred native and transgenic seed protein extracts Seed protein extracts of native and transgenic seeds, after dissociation with SDS and P-mercaptoethanol, were separated by electrophoresis and transferred to PVDF membrane. The membrane was stained with Ponceau S stain. Lanes 1: Purified MBL II; 2: native Arabidopsis\ 3-4: M III-wt; 5-6: M III-GLN; 7-8: M III-LYS) 99 Table 3-9: N-terminal amino acid sequences of the 8 kDa protein from native and transgenic Arahidopsis seeds N-terminal Amino acid of 8 kDa protein bands from native Published amino amino acid and transgenic Arabidopsis acid sequence of sequence # Native Mm-wt M ffl-Gln M ni-Lys AT2S3 large subunit 1 D D ~D D D 2 ~F F 1^ F ~F 3 ~E E ^" E ^ 4 ^ ^ ^ G ^ ~~ 5 ~P “ P P P ~P ~~ 6 ~Q ~Q ~Q ~Q ~Q 7 ~Q ~Q ~Q ~Q Q 8 ~G G ~G G ~G 9 Y Y ^ Y ^ ~~ 10 ~Q ~Q ~Q ~Q ^ 11 1^ L ~L L 12 ~L T ~L ~L 13 ~Q ~Q ~Q ~Q 14 ~Q ~Q ^ ~Q hi addition to the sequence similarity between MBL and the AT2S3 (Figure 1-3)，2S seed storage protein isoform 3 oiArahidopsis thaliana shares a number of other similarities with sweet protein mabinlin. Their mature proteins consist two subunits, a small and a large polypeptide chains of similar molecular weight and linked by disulfide bonds (Table 3-10). Since both chains are very similar in molecular size, it is very difficult, ifnot impossible, to separate them by 100 tricine-SDS-PAGE. The 8 kDa protein band excised for N-terminal sequence determination thus might contain both the large subunits of AT2S and MBL. As the N-terminal amino acid sequence of MBL may be blocked (we were not able to obtain N-teraiinal amino acid sequence of the purified MBL protein in a later experiment) or the level of the expressed MBL is just too low for sequencing, only the AT2S3 N-terminal sequence was obtained. Table 3-10: Biochemical properties of 2S albumin in Arabidopsis and Mabinlin (Data from ExPasy search) AT2S3 MBL Molecular Size 164 a.a. 155 a.a. Large subunit 83 a.a.(�9. kDa5 ) 73 a.a.(�8. kDa5 ) Small subunit 35a.a. (~ 4.1 kDa) 32 a.a. (~ 3.9 kDa) Isoelectric point -9.5 �12 3.4,6,2 Isoelectric precipitation The isoelectric point oiArabidopsis AT2S3 protein was reported to be pH 9.5 while that of MBL, pH 12 (Expasy search), hi a previous report, the pI of MBLin was determined as pH 11.8 (Hu and He, 1983). The distinct isoelectric points between Arabidopsis AT2S3 (pR 9.5) and mabinlang MBL (~pH12) (Table 3-10) were thus exploited as another means of searching for the expressed MBL proteins. Different fractions of protein were isolated from both the wild-type and transgenic seed total protein extracts (STPE) by isoelectric precipitation. Three pHs，9.5, 11.8 and 12, were used to precipitate the proteins. The precipitated proteins were denatured by treatment with sample buffer containing p- mercaptoethanol, and separated by 16.5 % tricine-SDS-PAGE. As shown in Figure 3-25, protein precipitates at pH 9.5 (lanes 2-5) and at pH 12 (lanes 13-16) contained most of the native Arabidopsis proteins. Proteins precipitated at pH 11.8 contained mainly the suspected protein bands, 8 and 6 kDa. The proteins were then electroblotted onto PVDF membrane (Figure 3-26) and subjected to westem blotting as described in section 3.4.6.4. ^ 2 3 5 6 ^ ^ :_ b ^ i^^ ^ 〜^« .,<^ft.: M 々rf:' m^M ‘^f^ 一_-一- 1 一 . ^ '4^^-r-^ym • m ^ _ ^^ r / ^ 々'供 kDa ^ ^^> "17 4 ^ ^ 14.4 M '、\ 10.7 6. 8.2 ^ ^ ^ 6.2 ^ '¢^^'「绍^ ^ ^i^a ^ .^f 0 2 3 4 5 6 一 J^^^^^l^J^ I • '*ttggtttgtg^ wV-'J-V^ar ^, ： 一 • '_-t>^,^^ $^ •_:_ . 5^'i /^¾ i 鄉0。^-^^^r^Y'^ 一 ^gKmm. •.: kDa V "17 fe 14.4 4 10.7 > 8.2 6. 6.2 Figure 3-25: Profiles of pI-precipitated proteins of native and transgenic Arabidopsis total seed protein extracts treated with p- mercaptoethanol Total protein extracts were precipitated at pH 9.5， 11.8 and 12 and then separated by 16.5 % tricine-SDS-PAGE after treatment with P-mercaptoethanol. Lanes 1 & 义 Native Arabidopsis STPE; 6 & 12, purified MBL II; 2-5, proteins precipitated at pH 9.5; 7,8,10,11, at pH 11.8; 13-16，at pH 12. Lanes 2,7,13，proteins precipitated from native Arabidopsis; 3，8, 14，from M III-wt; 4，9, 15, from M III-Ghi and 5，10, 16, from M III-Lys. ，，， 102 .减.'^^«1*響 、據.._艇‘ »!:*、 。、• 11 12 13 14 Figure 3-26: PVDF membrane with transferred pI-precipitated proteins of native and transgenic seeds Seed protein extracts of native and transgenic seeds, after dissociation with SDS and P-mercaptoethanol, were separated by electrophoresis and transferred to PVDF membrane. The membrane was stained with Ponceau S stain. Lanes 1 to 16 were the same as in Figure 3-25. 3,4.6J Assay of titers and quality of primary polyclonal antibody against purified MBL protein The protein bands representing linearized form (~ 14 kDa) and large and small subunits (〜8 kDa and 6 kDa) of pure MBL protein (Figure 3-28A) were used to raise polyclonal antibodies. To check the titer of antibodies, 2 jig of pure MBL II protein was dot-blotted onto a PVDF membrane. The membrane 103 containing the dots ofpurified MBL protein was air-dried and cut into strips. The strips were used for westem blot analysis following the procedures described in session 3.3.2.11. They were probed with different titers of two types of primary anti-MBL polyclonal antibodies dissolved in blocking buffer for one hour. Subsequent steps were the same as in session 3.3.2.11. As shown in Figure 3-1, the anti-linearized MBL II polyclonal antibody could detect purified MBL II protein up to a titer of 1 : 16,000. On the other hand, the anti-large and small subunit MBL II-antibody could only detect MBL II protein up to a titer of 1 : 8,000. Both antibodies were highly antigenic since they could recognize purified MBL II protein over a titer of 1 : 8,000. MBL III shows high similarity to MBL II protein in terms of amino acid sequence (Figure 2-1). Therefore, it was assumed that polyclonal antibodies raised from MBL II antigen could also recognize MBL III protein. Anti-linearized MBL II- antibody 500 1:1,000 1:2,000 l:4,0p0 1:8,000 1:16，00Q <-Titer of antibody Anti-large and small subunit MBL II- antibody Figure 3-1: Titering of anti-MBL primary polyclonal rabbit antibodies Two l^g ofpurified MBL II protein was dot-blotted onto PVDF membrane and subjected to immunodetection using different titers of primary polyclonal antibodies. 104 To further test whether the polyclonal antibodies can recognize linearized, large and small subunits ofMBL II protein, 4 ^ig ofpurified MBL II protein with P-mercaptoethanol treatment was used for tricine SDS-PAGE and western blotting. The procedure was the same as described in session 3.3.2.11. As shown in Figure 3-28A, the MBL II was incompletely reduced by P-mercaptoethanol, thus both a linearized MBL II (-14 kDa) and 2 subunits (〜8 kDa and ~6 kDa) were observed hi the immunodetection film, it was observed that the polyclonal antibody raised against linearized purified MBL II protein (Figure 3-28B) can recognize linearized 14 kDa, large subunit and small subunit of MBL II protein. The polyclonal antibody raised against large and small subunits (Figure 3-28C) recognizes the large subunit polypeptide better than the linearized 14 kDa and the small subunit polypeptide. Anti-linearized Anti-large and MBL II- small subunit antibody MBL II -antibody 17kDa 14.4kDa 10.7 kDa 8.2kDa 6.2kDa A B C Figure 3-28: Western bIot of the pI-precipitated proteins from native and transgenic Arabidopsis In theory, a good immunogen should have at least three chemical features. It must have an epitope that can be recognized by the cell-surface antibody found on B cells and at least one site recognized simultaneously by a class II protein and 105 by a T-cell receptor. The third parameter is that it must be degradable. These three features are the basic criteria to elicit a strong antibody response in the rabbit to be immunized The firsttw o parameters impose a minimum size limit, 5 kDa, on an immunogen. Since the small subunit polypeptide of purified MBL II is only about 6 kDa, it may not have suitable sites for the simultaneous binding of a class II protein and a T-cell receptor. Therefore, antibody raised from large and small subunits showed less ability to recognize the small subunit polypeptide and likely to the linearized 14 kDa protein too. 3.4.6,4 Western blot / Immunodetection PVDF membranes with transferred native and transgenic seed proteins pI- precipitated at pH 9.5，11.8 and 12 were subjeced to westem blot / immunodetection using anti-linearized MBL II primary antibody. From Figure 3- 29, no difference could be observed between the samples of native and transgenic Arabidopsis. The signals seemed to be very strong in the protein bands with molecular weight higher than purified MBL II (-14 kDa). Since Arabidopsis plant proteins including the 2S proteins might share some similar antigenic determinants with the sweet protein mabinlin, it is possible that the polyclonal antibodies raised against MBL II protein could react with proteins in other plant systems. Also, rabbits are herbivores that eat grasses and other plants, hence their biological systems might elicit antibodies against foreign antigens from the plants they eat hi future experiments, mice should be tried as host to raise polyclonal antibodies. (Personal communication with Prof. H.M. Lam, CUHK). 106 linearized MBL II (�1 4kDa) 9 10 11 12 13 14 15 16 linearized MBL II (~ 14 kDa) Figure 3-29: Western blot of the pI-precipitated proteins from native and transgenic Arabidopsis All the lanes were the same as in Figure 3- 25. hi another experiment, westem blotting was performed using the lower part of the PVDF membrane from Figure 3-24, which contained the native and transgenic seed protein extracts treated with P-mercaptoethanoL As shown in Figure 3-30，no difference was observed in any lanes. Since both the 2S protein of Arabidopsis and MBL consist of two subunit polypeptides with similar molecular wieghts, 8 and 6 kDa, and other similar biochemical and molecular properties (Table 3-10), cross-immuno reactivity is expected. To prevent such cross reactivity, monoclonal antibody is preferable for the immunodetection. 107 -i,^s.-.隱: - Figure 3-30: Western bIot I mmunodetection of native and transgenic seed proteins. Seed protein extracts of native and transgenic seeds, after dissociation with SDS and p-mercaptoethanol, were separated by electrophoresis and transferred to PVDF membrane. The lower portion of the membrane (Fig. 3-24) containing the two subunit polypeptides was excised and subjected to immunodetection. 108 4 General Discussion Thermostability is one of the indispensable concerns in the development of sweet proteins for food and beverage use. hivestigation of the underlying sweet principle not only improves our understanding of the thermostability of natural sweet proteins but also lead to a more logical approach to the design of synthetic sweeteners. To date, seven natural sweet plant proteins have been reported. However, most sweet proteins are heat-unstable. This characteristic makes their sales and marketing much more difficult. As reviewed in the literature, the cleavage ofthe disulfide bridges upon heating resulted in complete loss of the sweet activity in some of sweet proteins. The intensive disulfide, bonded network in those thermostable sweet proteins further proves the importance of disulfide bonds on thermostability (van der Wel et al, 1972: Kohmura et aL, 1996; Nirasawa, et al., 1993) A new observation on the thermostability of sweet protein was made by the research team of Nirasawa OSfirasawa et al., 1994). A single amino acid difference, instead ofthe four coincident disulfide bridges, was suggested by the team to discriminate the heat stability between heat-stable mabinlin II, III and IV and heat-unstable mabinlin I. To test this new hypothesis, we generated single specific amino acid alternations in both heat-stable MBLIII and heat-unstable MBLI and intended to analyse and compare their heat stability. Complementary DNAs encoding the two mabinlin isoforais were successfully and specifically engineered to contain these alternations by in vitro site-directed mutagenesis. Phaseolin promoter and terminator, a strong seed-specific expression system, were used to express both the wild-type and mutated mabinlin cDNAs in Arabidopsis. 109 This was the first report of using Arabidopsis to express mabinlin. Arabidopsis was used as host plant because of its short life cycle. Both wild-type and mutated MBLIII and I cDNAs were introduced into plant genome by Agrobactenum-mediated transformation via vacuum infiltration. This technique is simple, efficient, and time saving when compared to the traditional tissue culture method. Transgenic plants of R3 generation carrying homozygous mabinlin genes were obtained in this study for molecular and biochemical study. This aims at using genetically homogenous seed material for expression of mabinlin genes. Positive signal of GUS staining and PCR amplification of MBL genomic DNA from transgenic leaves proved the stable integration of transgene into plant genome. Expression of MBL RNAs was also successftdly detected in transgenic siliques by RT-PCR. The presence of specific codon mutations in the RNA transcripts and mabinlin sequence fidelity were also confirmed by cDNA cloning and sequencing. These results indicated that the expression cassettes were working in the Arabidopsis genome, producing MBL mRNAs. Tricine SDS-PAGE showed that the protein profiles of native and transgenic Arabidopsis seeds displayed pattems with different banding intensity near the desired molecular size of mabinlin protein, suggesting additional protein with size comparable to mabinlin might be present in the transgenic seeds. The similar molecular size between mabinlin and the 2S seed storage protein of Arabidopsis made it much difficult to analyse and identify the expressed mabinlin. A number of modifications were attempted in protein extraction and westem blot procedures to overcome this difficulty. Thus, size-exclusion chromatography was tried to enrich the 14-kD protein fractioni n Arabidopsis plant extract for further studies; isoelectric focusing electrophoresis was tried to resolve the transgenic protein from native Arabidopsis 2S seed storage protein by their different pI points; and isoelectric precipitation was also tried to fractionatean d concentrate the transgenic protein and Arabidopsis native protein based on their different pIs. All these attempts did not yield a promising result, hi further study, monoclonal antibody specific only to MBL protein may be an approach worth of trying. hi addition to the detection method, low or absence of protein expression may be the key explanation of the negative westera-blot result. If this was the case, the major problem could be at the post-transcriptional or translational level. Gene silencing could hinder protein expression of transgene. Silencing occurs either at the transcriptional level, which frequently associates with DNA methylation, or post-transcriptionally, which involves RNA degradation (Vaucheret et al., 1998). Protein degradation after translation may also occur. Our laboratory is currently using in vitro transcriptionAranslation system to check whether the gene construct and mRNA are functional or not. If it works, gene constructs containing mabinlin wild-type and mutated genes can also be analysed. 111 Conclusion The genes for the sweet protein Mabinlin isoforai III，including both the wild-type gene (MBL III-wt) and its mutants (MBL III-Gln and -Lys), have been shown to be expressed at DNA and RNA levels in the seeds of transgenic Arabidopsis. Expression at the level of protein could not be confirmed because its biochemical properties and immuno-cross-reactivity is very similar to the 2S albumins in the host Arabidopsis (AT2S3). Further experiments to obtain more specific antibodies and/or to separate the MBL from AT2S3 will therefore be necessary to verify the protein expression, hence to perform the protein heat- stability test. 112 References Abe, K., Yamashita, H., Arai, S. and Kurihara, Y. 1992. Molecular cloning of curculin, a novel taste-modifying protein with a sweet taste. Biochem. et Biophys. Acta. 1130: 232-234 Altenbach, S.B., Kuo, C.C., Staraci, L.C., Pearson, K.W., Wainwright, C., Georgescu, A. and Townsend, J. 1992. Accumulation of a Brazil nut albumin in seeds of transgenic canola results in enhanced levels of seed protein methionine. PlantMol. Biol 18: 235-245 Altenbach, S.B., Pearson, K.W., Meeker, G., Staraci, L.C. and Sun, S.S.M. 1989. Enhancement of the methionine content of seed proteins by the expression of a chimeric gene encoding a methionine-rich protein in transgenic plants. PlantMol Biol. 13: 513-522 、 Bechtold, N. and Bouchez, D. 1995. In planta Agr0bacterium-mti&2AQ& Transformation of Adult Arabidopsis Thaliana Plants by Vacuum Mltration. In Gene Transfer to Plants, Potrykus, I. and Spangenberg, G. eds., Springer Lab Manual, pp. 19-23 Bechtold, N., Ellis, J. and Pelletier, G. 1993. In planta Agrobacterium-mQdmtQd gene transfer by infiltration ofsidultArabidopsis thaliana plants. C. R. Acad Sci. Paris, Life Sci. 316: 1194-1199 Bechtold, N. and Pelletier, G. 1998. In planta Agrobactehum-medmtQd transformaton of adult Arabidopsis thaliana plants by vacuum infiltration. hiArabidopsis Protocols, Martinez-Zapater, J. and Salinas, J. eds., Humana Press Inc., Totowa, NJ, pp. 259-266 Bohak, Z.V.I, and Li, S.L. 1976. The structure ofmonellin and its relation to the sweetness ofthe protein. Biochem. et Biophys. Acta. 427: 153-170 Brouwer, J.N, Hellekant, G., Kasahara, H., Van der Wel h. and Zotterman, Y. 1973. Electrophysiological study of the gus tastory effects of the sweet proteins monellin and thaumatin in monkey, guinea, pig and rat. Acta Physiol Scand 89: 550-557 Brouwer, J.N, Van del Wel H., Francke, A. and Henning, G.T. 1968. Miraculin, the sweetness-inducing protein from miracle fruit. Nature 220: 373-374 Cagan, R.H. and Morris, R.W. 1979. Biochemical studies of taste sensation: Binding to taste tissue of^H-labelled monellin, a sweet-tasting protein. Proc. Natl. Acad. Sci. USA 76 (4): 1692-1696 113 Chang, S.S.，Park, S.K. and Nam, H.G. 1994. Transformation oiArabidopsis by Argohacterium inoculation on wounds. Plant J. 5: 551-558 Comelissen, B.J.C.，Hooft, Van. Huijsduijnen, R.A.M. and Bol, J.F. 1986. A tobacco mosaic virus-induced tobacco proteins is homologous to the sweet- tasting protein thaumatin. Nature 321: 531-532 Damm, B. and Halfer, U. 1993. Transgenic Arahidopsis. In: Transgenic Plant Volume 2, Kung, S.D. and Wu, R. eds., Academic Press, hic. ppl69-188. Danun, B., Schmidt, R. and Wilhnitzer, L. 1989. Efficient transformation of Arahidopsis thaliana using direct gene transfer to protoplasts. Mol Gen. Genet 217: 6-12 de vos, A.M., Hatada, M., van der Wel H., Krabbendam, H., Peerdeman, A.F. and Kim, S.H. 1985. Three-dimensional structure of thaumatin I，an intensively sweet protein. Proc. Natl Sci. USA 82: 1406-1409 del Campillo and E., Lewis, L.N. 1992. Identification and kinetics of accumulation of proteins induced by ethylene in bean abscission zones. Plant Physiol 98: 955-961 Ding, M. and Goran, H. 1994. Brazzein, a new high-potency thermostable sweet protein from Pentadiplandra brazzeana B. FEBS Letters 355: 106-108 Ding，M. and Hu, Z. 1986. Studies on Mabinlin, the sweet protein from seeds of Capparis Masaikai. IV. Stability and denaturation. Acta. Bot. Yunnanica 8: 181-192 Doyle, J.D., Doyle, J.L. and Bailey, L.H. 1990. Isolation ofplant DNA from fresh tissue. Focus 12: 13-15 Doyle, DJ., Schuler, M.A., Godette, W.D., Zenger, B.，Beachy, R.N. and Slightom, J.L. 1986. The glycosylated seed storage proteins of glycine max and Phaseolus vulgaris: Structural homologies ofgenes and proteins. J. Biol. Chem. 261: 9228-9283 Edens，L.，Bom, I.，Ledeboer, A.M., Matt, J., Toonen, M.Y., Visser, C. and Venips, C.T. 1984. Synthesis and processing of the plant protein thaumatin in yeast. Cell 37: 629-633 Edens, L.，Heslinga, L.，Klok, R.，Ledeboer, A.M., Matt, J., Toonen, M.Y, Visser, C.,Verrips, C.T. 1982. Cloning of cDNA encoding the sweet tasting plant thaumatin and its expression in E. coli. Gene 18: 1-12 114 Edens, L.，Ledeboer, A.M., Verrips, C.T.，Van de Berg, J.A. 1983. Kluyveromyces yeast containing rDNA plasmid for cultivation in production of perprothamnatin or its modified forms. Patent application WO 8304051 Etheridge, K. 1994. Sales and marketing of Talin. hi Thaumatin, Witty, M. and Higginbotham, J.D. eds, CRC Press, Boca Raton, FL, pp 47-59 Faus，I.，Patino, C.，del Rio, J. L.，del Moral, C.，Barroso, H.S. and Rubio, V. 1996. Expression of a synthetic gene encoding the sweet-tasting protein thamnatin in Escherichia coli. Biochem. Biophy. Res. Commun. 229 121- 127 Feldmann, K.A. and Marks, M.D. 1987. Agrohacterium-mQ^mXtdi transformation of germinating seeds of Arabidopsis thaliana: A tissue culture approach. Mol Gen. Genet. 208: 1-9 Frank，G. and Zuber, H. 1976. The complete amino acid sequences of both subunits of the sweet protein monellin. Hoppe-Seyler's Zeitschrifft fur Physiologische Chemie 357: 585-592 Goeddel，D.V.，Kleid, D.G., Bolivar, F.，Heyncher, H.Z., Yansura, D.G., Crea, R.， Hirose, T., Kraszewski, A., Itakura, K. and Riggs, A.D. 1979. Expression in E coli of chemically synthesized genes for human insulin. Proc. Natl. Acad Sci. USA 76: 106-110 Graham, J.S., Burkhart, W.，Xiong, J., Gillikin, W. 1992. Complete amino acid sequence of soybean leaf P21 Similarity to the thaumatin-like polypeptides. Plant Physiol 98: 163-165 Hahm and Blatt 1990. Expression and secretion of thaumatin from Aspergillus oryzae. Agri. Bio. Chem. 54: 2513 Hecker，K.H. and Roux, K.H. 1996. High and low annealing temperatures increase both specificity and yield in TD and SD PCR. Biotechniques 20 478-485 Higginbotham, J.D. and Hough, C.A.M. 1977. Useful taste properties of amino acid and proteins, hi Sensory Properties ofFoods, Birth, G.G., Brennen, J.G. and Parker, K.J. eds., Applied Sciences, London, pp 129-149 Higginbotham, J.D., Snodiun, D.J., Eaton, K.K. and Daniel, J.W. 1983. Safety evaluation ofthaumatin (talin protein). FoodChem. Toxicol 21: 815-823 Hoffinan, L.M.，Donaldson, D.D., Bookland, R., Rashka, K. and Herman, E.M. 1987. Synthesis and protein body deposition of maize 15-kD zein in transgenic tobacco seeds. EMBOJ. 6: 3213-3221 115 Horsch, R.B., Fry, J.E., Hof&nan, Nl.，Eichholtz, D.，Rogers, S.G. and Fraley, R.T. 1985. A simple and general method for transferring genes into plants. Science 227: 1229-1231 Hough, C.A.M. and Edwardson, J.A. 1978. Antibodies to thaumatin as a model of the sweet taste receptor. Nature 271: 381-383 Hu, Z. and He, M. 1983. Studies on Mabinlin, a sweet protein from the seeds of Capparis Masaikai leve. I. Extraction, purification and certain characteristics. Acta Bot. Yunmnica 5: 207-212 Hu, Z., Peng, L.P. and He, M. 1985. Studies on mabinlin, the sweet protein from seeds of Capparis Masaikai. II. Comparison of mabinlin I to mabinlin II. Acta. Bot. Yunnanica 7: 1-10 Huang, S.，Elliott, R.C., Liu, P.S., Koduri, R.K., Weichmann, J.L., Lee, J.H., Blair, L.C., Chosh-Dastidav, P., Bradshaw, R.A., Bryan, K.M., Einarson, B., KendaU，、Rl.，Kolacz, K.H. and Saito, K. 1987. Specificity of cotranslational amino-terminal processing of proteins in yeast. Biochemistry 26: 8242-8246 Hudson, G. and Biemann, K. 1976. Mass spectrometric sequencing of proteins: The structure of subunits I of monellin. Biochem. Biophys. Res. Commun. 71:212-220 Illingworth, C.，Larson, G. and Helekant, G. 1989. Secretion of the sweet tasting plant protein thaumatin by Streptomyces lividans. J. industrial Microbiology 4: 37-42 Illingworth, C.，Larson, G. and Hellekant, G. 1988. Secretion ofthe sweet-tasting plant protein thaumatin by Bacillus subtillis P. Biotech. Lett 10: 587-592 toglett, G.E., Dowling, B., Albrecht, J. and Hoglan, F. 1965. Taste-modifying properties ofmiracle fruits. J. Agric. Food Chem. 13: 284-287 Iyengar, R.B., Smits, P., Van der Ouderaa F., van der Wel H., van Brouwerhaven J.，Ravestein, P., Richters, G. and van Wassenar P.D. 1979. The complete amino-acid sequence ofthe sweet protein thaumatin I. Eur. J. Biochem. 96: 193-204 Jefferson, R.A.，Kavanaugh, T.A. and Bevan, M.W. 1987. GUS fusion: p- glucuronidase as a sensitive and versatile gene fiision marker in higher plant EMBOJ. 6:3901 Jenner，M.R. 1989. Sucralose: unveiling its properties and applications. In Progress in Sweetners, Grenby, T.H. ed, New York: Elsevier, pp 121-141 116 Katavic，V.，Haughn, G.W.，Reed, D.，Martin, M. and Kunst, L. 1994. In planta transformation oiArabidopsis thaliana. Mol. Gen. Genet. 245: 363-370 Kim, S.H., Kang, C.H., Kim, R., Cho, J.M.，Lee, Y.B. and Lee, T.Y. 1989. Redesigning a sweet protein: increased stability and renaturability. Protein Engineering 2(8): 571-575 Kohmura M., Ota, M.，Izawa, H.，Ding M.，Hellekant, G. and Ariyoshi, Y. 1996. Assignment of the disulfide bonds in the sweet protein Brazzein. Biopolymers 38: 553-556 Kondo，K., Miura, Y.，Sone, H., Kobayashi, K. and Lijima, H. 1997. High level expression of a sweet protein, monellin, in the food yeast Candida utilis. Nature Biotechnology 15: 453-457 Korver，O.，Van Gorkom M. and van der Wel H. 1973. Spectrometric investigation of thaumatin I and II, the sweet tasting proteins from Thaumatococcus daniellii Benth. Eur. J. Biochem. 35: 554 Krutosikova, A., Uher, M. 1992. Naturally occuring sweet substances, hi Natural and Synthetic Sweet Substances, (Translation editor: Hudec, J.) pp.76-79 Kurihara, K.，Beidler, L.M. 1968. Taste-modifying protein from miracle fruit. Science 161: 1241-1243 Larson-Powers, N. and Pangbom, R.M. 1978. Paired comparison and time- intensity measurements of the sensory properties of beverages and gelatins containing sucrose or synthetic sweeteners. J. FoodScl 43: 41-46 Lee，J.H., Weichmann, J.L., Koduri, R.K., Ghosh-Dastidar, P., Saito, K., Blair, L.C.，Date, T., Lai, J.S.，Hollenberg, S.M. and Kendall, R.L. 1988. Expression of synthetic thaumatin genes in yeast. Biochemistry 11' 5101- 5107 Ling, M.F. and Robinson, B.H. 1995. A one-step polymerase chain reaction site- directed mutagenesis method for large gene-cassettes with high efficiency, yield, and fidelity. Analytical Biochemistry 230: 167-172 Lipinski, G.W.R. 1985a. The new intense sweetener acesulfame-K. Food Chem., 16: 259-269 Lipinski, G.W.R. 1985b. Acesulfame-K hi Alternative Sweeteners, Nabors, L.0., Gelardi, R.C. eds., Marcel Dekker, New York, pp89-102 ， Lipinski, G.W.R. 1990. Multiple sweeteners. FoodMark Technol 4: 22-25 117 Liu, X.Z. and Hu, Z. 1988. N- and C-terminal partial amino acid sequence determination of the sweet proteins from the seed of Capparis Masaikai. Acta. Bot. Yunnanica 10: 381-388 Liu, X.Z.，Maeda, S.J., Hu, Z.，Aiuchi, T.，Nakaya, K. and Kurihara, Y. 1993. Purification, complete amino acid sequence and structural characterization ofthe heat-stable sweet protein, mabinlin II. Eur. J. Biochem. 211: 281-287 Lloyd, A.M., Baamason, A., Rogers, S.G.，Byme, M.C., Fraley, R.T. and Horsch, R.B. 1986. Transformation of Arabidopsis thaliana with Agrobacterium tumefaciens. Science 234: 464-466 Lopes, T.S.，Klootwijk, J., Veenstra, A.E., van der Aar P.C., van Heerikhuizen H., Raue, A. and Planta, R.J. 1989. Gene 79: 199-206 Ma, Y. and Bliss, F.A. 1979. Seed protein of common bean. Crop Sci. 17: 431- 437 、 Meyerowitz, E.M. 1989. Arabidopsis, a useful weed. Cell 56: 263-269 Morris, J.A. and Cagan, R.H. 1972. Purification of monellin the sweet principle of Dioscoreophyllum cumminsii. Biochem. Biophys. Acta. 261: 114-122 Morris, J.A., Martenson, Deibler, G. and Cagan, R.H. 1973. Characterization of monellin, a protein that tastes sweet. J. Biol. Chem. 248: 534-539 Most, B.H., Summerfield, R.J. and Boxall, M. 1979. Tropical plants with sweetening properties, physiological and agronomic problems of protected cropping 2, Thaumatococcus daniellii. Econ. Bot. 32: 321 Nakajo, S., Akabane, T., Nakaya, K.，Nakamura, Y. and Kurihara, Y. 1992. An enzyme immunoassay and immunoblot analysis for curculin, a new type of taste-modifying protein: cross-reactivity of curculin and miraculin to both antibodies. Biochem. et Biophys. Acta. 1118: 293-297 Negrutiu, I.，Shillito, R.，Potrykus, I.，Biasini, G. and Sala, F. 1987. Hybrid genes in the analysis of transformation conditions. Setting up a simple method for direct gene transfer in plant protoplasts. PlantMol Biol 8: 363-373 Nirasawa, S.，Liu, X.Z.，Nishino, T. and Kurihara, Y. 1993. Disulfide bridge structure of the heat-stable sweet protein mabinlin II. Biochem. et Biophys. Acta 1202: 277-280 Nirasawa, S.，Nishino, T, Katahira, M, Uesugi, S.，Hu, Z. and Kurihara, Y. 1994. Structures of heat-stable and unstable homologues of the sweet protein mabinlin: The difference in the heat stability is due to replacement of a single amino acid residue. Eur. J. Biochem. 233: 989-995 118 Ogata, C., Hatada, M., Tomlinson, G., Shin, W.C. and Kim, S.H. 1987. Crystal structure ofthe intensively sweet protein monelin. Nature 32& 793-742 Osbom, T.C. 1988. Genetic control ofbean seed protein. CRC Crit. Rev. in Plant Sci. 7:93-116 Overbeeke, N. and Verrips, C.T. 1992. Expression of plant genes in yeast and bacteria . hi Plant Biotechnology. Fowler, M.W., Warren, G.S., and Moo- Young, M. eds., Pergamon Press, pp 181 Penamibia, L.’ Kim, R., Giovannoni, J.，Kim, S.H. and Fischer, R.L. 1992. Production ofthe sweet protein monellin in transgenic plants. Biotechnolo^ 10: 561-564 Rebmann, G., Mauch, F. and Dudler, R. 1991. Sequence of a wheat cDNA encoding a pathogen-induced thaumatin-like protein. Plant Mol Biol. 17. 283-285 Renwick, A.G. 1985. The disposition of saccharin in animals and men - a review. FoodChem. Toxicol. 23: 429-435 Richardson, M.，Vadles-Rodriguez, S.，and Blanco-Labra, A. 1987. A possible fUnction for thaumatin and a TMW-induced protein suggested by homology to a maize inhibitor. Nature 327: 432-433 Rodrigo, I.，Vera, P., Frank, R. and Cponejero, V. 1991. Identification of the viroid-induced tomato pathogensis-related (PR) protein P23 as the thaumatin-like tomato protein NP 24 associated with osmotic stress Plant Mol Biol 16: 931-934 Saier，M.H., Wemer, P.K. and Muller, M. 1989. tosertion of proteins into bacterial membranes: mechanism, characteristics, and comparisons with the eukaryotic process. Microbiological Reviews 53: 333-66 Sambrook, J., Fritsch, E.F. and Maniatis, T. 1989. In Molecular cloning: A Laboratory Manual, 2°^ ed., Cold Spring Habor Laboratory Press, Cold Spring Habor, NY. Sarkar，G. and Sommer, S.S. 1990. The megaprimer method of site-directed mutagenesis. Biotechniques 8 (4): 404-407 Schagger, H. and Von Jagow, G. 1987. Tricine-sodium dodecyl sulfate polyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa. Anal Biochem. 166: 386-379 Seki，M.，Komeda, Y., Iida, A.，Yamada, Y. and Morikawa, H. 1991a. Transient expression of P-glucuronidase in Arabidopsis thaliana leaves and roots and 119 Brassica napus stems using a pneumatic particle gun. Plant Mol Biol 17. 259-263 Seki, M., Shigemoto, N.，Komeda, Y.，hnamura, J. and Morikawa, H. 1991b. Transgenic Arabidopsis thaliana plants obtained by particle bombardment mediated transformation. Appl Microbiol Biotechnol 28: 228-230 Sengupta-Gopalan, C.’ Reichert, N.A.，Barker, R.F., Hall, T.C. and Kemp, J.D. 1985. Developmentally regulated expression of the bean P-phaseolin'gene in tobacco seed. Proc. Natl. Acad. Sci. USA 82: 3320-3324 Slightom, J.L., Sun, S.S.M. and Hall, T.C. 1983. Complete nucleotide sequence ofFrench bean storage protein gene: phaseolin. Proc. Natl. Acad. Sci USA 80: 1897-1901 Stintzi，A:，Heitz, T.，Kauffinann, S., Legrand, N. and Fritig, B. 1991. Mentification ofabasic pathogensis-related thaumatin-like protein ofvirus- infected tobacco as osmotin. Physiol Mol Plant Pathology 38: 131-146 Sun，S.S.M., Mutschler, M.A., Bliss, F.A. and Hall, T.C. 1978. Protein synthesis and accumulation in bean cotyledons during growth. Plant Physio 61 918- 923 Sun，S.S.M. and Xiong, L. 1996. Molecular cloning and transgenic expression of the sweet protein mabinlin in potato tubers. Plant Physiology 111(2): Suppl., 57 Sun，S.S.M., Xiong, L.，Hu, Z. and Chen, H. 1997. New isolated mabinlin nucleotide sequences. Patent application WO 9700945 Sun，S.S.M., Zuo, W., Tu, H.M. and Xiong, L. 1996. Plant proteins: engineering for improved quality. Ann. N. Y. Acad. Sci. 792: 37-42 Theerasil?，S.，Hitotsuya, H., Nakaku, S.，Nakaya, K., Nakamura, Y. and Kmihara，Y. 1989. Complete amino acid sequence and structure characterization of the taste-modifying protein, Miraculin. J. Biol Chem. 264: 6655-6659 Theerasilp, S. and Kuribara, Y. 1988. Complete purification and characterization ofthe taste-modifying protein, miraculin, from miracle fruit. J. Biol Chem 263: 11536-11539 Valvekens, D., Van Montagu, M. and Van Lijsebettens, M. 1988. Agrobacterium rwwe/ac/em-mediated transformatioin oiArabidopsis thaliana root explants by using kanamycin selection. Proc. Natl. Acad Sci. USA 85: 5536-5540 120 van der Wel H. 1974a. Miracle fruit, Katemfe and Serendipity berry, hi Symposium Sweeteners. Diglett, G.E. ed., AVI, Westport, CT, pp 194 van der Wel H. 1974b. Thaumatin and monelin, the intensely sweet tasting proteins from tropical African fruits,h i Biochemistry of Sensory Functions. Jaenicke, L. ed., Springer Verlag, Berlin, pp 235 van der Wel H. 1980. Physiological action and structural characteristics of the sweet-tasting proteins thaumatin and monellin. Trends in Biochem. Sci. 5: 122 van der Wel H. 1986. Developments in foodproteim. BJP Hudson ed., Appli. Sci. Publ. ltd., London, 4: 219-245 van der Wel H. and Arvidson, K. 1978. Chem. Senses Flavors 3: 291-299 van der Wel H. and Bel, W.J. 1976. Effect of acetylation and methylation the sweetness intensity ofthaumatin I. Chem. Senses Flavors 2: 211-218 van der Wel H. and Bel, W.J. 1980. Enzymatic properties of the sweet tasting proteins thaumatin and monellin after partial reduction. Eur. J. Biochem. 104: 413 van der Wel H.，Larson, G., Hladik, C.M., Hellekant, G. and Glaser, D. 1989. Isolation and characterization of pentadin, the sweet principle of Pentadiplandra brazzeana Baillon. Chem. Senses Flavors 14: 75-79 van der Wel H., Loeve, K. 1972. Isolation and characterization of Thaumatin I and II，the sweet-tasting proteins from Thaumatococcus daniellii Benth. Eur. J. Biochem. 31: 221-225 van der Wel H.，Lyengar, R.B.，van Brouwershaven, J., van Wassenaar, P.D.，Bel, W.J. and van der Ouderaa, FJ.G., 1984. Assignment of the disulfide bonds in the sweet tasting protein thaumatin I. Eur. J. Biochem. 144: 14 Vaucheret, H., Beclin, C.，Ehnayan, T, Feuerbach, F., Godon, C., Morel, J.B., Mourrain, P.，Palauqi, J.C. and Vemhettes, S. 1998. Transgene-induced gene silencing in plants. Plant J. 16: 651-659 Verrips C. T. 1992. Structure and protein engineering of thaumatin and other sweet proteins. In Plant Protein Engineering, Shewry, P.R. and Gutteridge, S. eds., Cambridge University Press, Pp 219-234 Vetsch, W. 1985. Aspartame: technical considerations and predicted use. Food Chem. 16: 245-258 121 Vincent, H.C., Lynch, M.J., Pohley, F.M. and Kirchmeyer, F.J. 1955. A taste panel study of cyclamate-Saccharin mixture and of its components. J. Am. Pharm. Assoc. 44: 442-446 Weichmann, LL. and Ghoshdasti, P. 1987. Recombinant thaumatin I analogues which can be re-folded to give native conformation eliciting a sweet taste sensation. Patent application WO 8703007 Witty, M. 1989. Thaumatin II: a marker gene for use in plants. Nucl. AcidRes. 17: 3312 Witty, M. 1990. Sensory evaluation of transgenic Solanum tuberosum producing r-thaumatin II. New Zealand. J. Crop Hort. Sci. 18: 77-80 Witty, M. 1992. Thaumatin II: A sweet marker gene for use in plants. Methods in Enzymology 216: 441-447 Wlodawer, A. and Hodgson, K.D. 1975. Crystallization and crystal data of monellin. Proc. Natl Acad. Sci. USA 172: 398-399 Woloshuk, C.P., Meulenhoff, J.S., Sela-Buurlage, M. and van der Elzen, P.J.M. 1991. Pathogen-induced proteins wi5i inhibitory activity toward Phytophthora infestans. Plant Cell 3: 619-628 Yamashita, H., Theerasilp, S,, Aiuchi, T., Nakaya, K.，Nakamura, Y. and Kurihara, Y. 1990. Purification and complete amino acid sequence of a new type of sweet protein with taste-modifying activity, curculin. J. Biol Chem. 265: 15770-15775 Zenmanek, E.C. and Wasserman, B.P. 1995. Issues and advances in the use of transgenic organisms for the production of thaumatin, the intensely sweet protein from Thaumatococcus danielli. Critical Reviews in Food Science andNutritio n 35(3): 455-466 Zhang, H. and Somerville, C.R. 1987. Transfer ofthe maize transposable element Mul into Arahidopsis thaliana. Plant Sci 48: 165-173 122 CUHK Libraries DD3flD3bSl